diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,56325 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 20.0,
+  "eval_steps": 500,
+  "global_step": 80200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0024937655860349127,
+      "grad_norm": NaN,
+      "learning_rate": 1.9998753117206985e-05,
+      "loss": 7.3204,
+      "step": 10
+    },
+    {
+      "epoch": 0.004987531172069825,
+      "grad_norm": 30.45345687866211,
+      "learning_rate": 1.999625935162095e-05,
+      "loss": 6.3759,
+      "step": 20
+    },
+    {
+      "epoch": 0.007481296758104738,
+      "grad_norm": 52.34556579589844,
+      "learning_rate": 1.9993765586034916e-05,
+      "loss": 5.2081,
+      "step": 30
+    },
+    {
+      "epoch": 0.00997506234413965,
+      "grad_norm": 33.23439025878906,
+      "learning_rate": 1.999127182044888e-05,
+      "loss": 3.7536,
+      "step": 40
+    },
+    {
+      "epoch": 0.012468827930174564,
+      "grad_norm": 19.338409423828125,
+      "learning_rate": 1.9988778054862846e-05,
+      "loss": 2.4798,
+      "step": 50
+    },
+    {
+      "epoch": 0.014962593516209476,
+      "grad_norm": 5.575465679168701,
+      "learning_rate": 1.998628428927681e-05,
+      "loss": 1.4935,
+      "step": 60
+    },
+    {
+      "epoch": 0.017456359102244388,
+      "grad_norm": 1.394232153892517,
+      "learning_rate": 1.9983790523690774e-05,
+      "loss": 1.0447,
+      "step": 70
+    },
+    {
+      "epoch": 0.0199501246882793,
+      "grad_norm": 2.6528663635253906,
+      "learning_rate": 1.998129675810474e-05,
+      "loss": 1.1943,
+      "step": 80
+    },
+    {
+      "epoch": 0.022443890274314215,
+      "grad_norm": 2.9787819385528564,
+      "learning_rate": 1.9978802992518704e-05,
+      "loss": 1.2081,
+      "step": 90
+    },
+    {
+      "epoch": 0.02493765586034913,
+      "grad_norm": 1.774791955947876,
+      "learning_rate": 1.997630922693267e-05,
+      "loss": 0.9248,
+      "step": 100
+    },
+    {
+      "epoch": 0.02743142144638404,
+      "grad_norm": 1.5466320514678955,
+      "learning_rate": 1.9973815461346635e-05,
+      "loss": 1.0454,
+      "step": 110
+    },
+    {
+      "epoch": 0.029925187032418952,
+      "grad_norm": 2.4883105754852295,
+      "learning_rate": 1.99713216957606e-05,
+      "loss": 1.1251,
+      "step": 120
+    },
+    {
+      "epoch": 0.032418952618453865,
+      "grad_norm": 2.3588509559631348,
+      "learning_rate": 1.9968827930174565e-05,
+      "loss": 1.178,
+      "step": 130
+    },
+    {
+      "epoch": 0.034912718204488775,
+      "grad_norm": 2.097666025161743,
+      "learning_rate": 1.996633416458853e-05,
+      "loss": 0.937,
+      "step": 140
+    },
+    {
+      "epoch": 0.03740648379052369,
+      "grad_norm": 2.220046043395996,
+      "learning_rate": 1.9963840399002496e-05,
+      "loss": 0.9336,
+      "step": 150
+    },
+    {
+      "epoch": 0.0399002493765586,
+      "grad_norm": 1.8168127536773682,
+      "learning_rate": 1.996134663341646e-05,
+      "loss": 0.8292,
+      "step": 160
+    },
+    {
+      "epoch": 0.04239401496259352,
+      "grad_norm": 2.687063217163086,
+      "learning_rate": 1.9958852867830423e-05,
+      "loss": 0.9569,
+      "step": 170
+    },
+    {
+      "epoch": 0.04488778054862843,
+      "grad_norm": 2.103942632675171,
+      "learning_rate": 1.995635910224439e-05,
+      "loss": 0.7538,
+      "step": 180
+    },
+    {
+      "epoch": 0.04738154613466334,
+      "grad_norm": 2.371488571166992,
+      "learning_rate": 1.9953865336658357e-05,
+      "loss": 0.8276,
+      "step": 190
+    },
+    {
+      "epoch": 0.04987531172069826,
+      "grad_norm": 3.574516534805298,
+      "learning_rate": 1.995137157107232e-05,
+      "loss": 1.021,
+      "step": 200
+    },
+    {
+      "epoch": 0.05236907730673317,
+      "grad_norm": 3.9193525314331055,
+      "learning_rate": 1.9948877805486288e-05,
+      "loss": 0.8487,
+      "step": 210
+    },
+    {
+      "epoch": 0.05486284289276808,
+      "grad_norm": 2.934713363647461,
+      "learning_rate": 1.994638403990025e-05,
+      "loss": 0.7093,
+      "step": 220
+    },
+    {
+      "epoch": 0.057356608478802994,
+      "grad_norm": 2.5535857677459717,
+      "learning_rate": 1.994389027431422e-05,
+      "loss": 0.8764,
+      "step": 230
+    },
+    {
+      "epoch": 0.059850374064837904,
+      "grad_norm": 2.949345827102661,
+      "learning_rate": 1.9941396508728182e-05,
+      "loss": 0.8242,
+      "step": 240
+    },
+    {
+      "epoch": 0.06234413965087282,
+      "grad_norm": 2.376368522644043,
+      "learning_rate": 1.9938902743142146e-05,
+      "loss": 0.8442,
+      "step": 250
+    },
+    {
+      "epoch": 0.06483790523690773,
+      "grad_norm": 5.437434673309326,
+      "learning_rate": 1.9936408977556113e-05,
+      "loss": 1.0203,
+      "step": 260
+    },
+    {
+      "epoch": 0.06733167082294264,
+      "grad_norm": 2.630307674407959,
+      "learning_rate": 1.9933915211970076e-05,
+      "loss": 0.6734,
+      "step": 270
+    },
+    {
+      "epoch": 0.06982543640897755,
+      "grad_norm": 4.224475860595703,
+      "learning_rate": 1.993142144638404e-05,
+      "loss": 0.9501,
+      "step": 280
+    },
+    {
+      "epoch": 0.07231920199501247,
+      "grad_norm": 4.720646858215332,
+      "learning_rate": 1.9928927680798007e-05,
+      "loss": 0.6527,
+      "step": 290
+    },
+    {
+      "epoch": 0.07481296758104738,
+      "grad_norm": 4.588720321655273,
+      "learning_rate": 1.992643391521197e-05,
+      "loss": 0.7645,
+      "step": 300
+    },
+    {
+      "epoch": 0.0773067331670823,
+      "grad_norm": 2.956965923309326,
+      "learning_rate": 1.9923940149625938e-05,
+      "loss": 0.8471,
+      "step": 310
+    },
+    {
+      "epoch": 0.0798004987531172,
+      "grad_norm": 2.415419816970825,
+      "learning_rate": 1.99214463840399e-05,
+      "loss": 0.8739,
+      "step": 320
+    },
+    {
+      "epoch": 0.08229426433915212,
+      "grad_norm": 4.1891188621521,
+      "learning_rate": 1.9918952618453865e-05,
+      "loss": 0.8344,
+      "step": 330
+    },
+    {
+      "epoch": 0.08478802992518704,
+      "grad_norm": 4.513364315032959,
+      "learning_rate": 1.9916458852867832e-05,
+      "loss": 0.7266,
+      "step": 340
+    },
+    {
+      "epoch": 0.08728179551122195,
+      "grad_norm": 3.10957670211792,
+      "learning_rate": 1.9913965087281796e-05,
+      "loss": 0.8771,
+      "step": 350
+    },
+    {
+      "epoch": 0.08977556109725686,
+      "grad_norm": 2.839132070541382,
+      "learning_rate": 1.9911471321695763e-05,
+      "loss": 0.5522,
+      "step": 360
+    },
+    {
+      "epoch": 0.09226932668329177,
+      "grad_norm": 7.818932056427002,
+      "learning_rate": 1.9908977556109726e-05,
+      "loss": 0.7509,
+      "step": 370
+    },
+    {
+      "epoch": 0.09476309226932668,
+      "grad_norm": 3.391204357147217,
+      "learning_rate": 1.9906483790523693e-05,
+      "loss": 0.8433,
+      "step": 380
+    },
+    {
+      "epoch": 0.09725685785536159,
+      "grad_norm": 5.5061726570129395,
+      "learning_rate": 1.9903990024937657e-05,
+      "loss": 1.0054,
+      "step": 390
+    },
+    {
+      "epoch": 0.09975062344139651,
+      "grad_norm": 3.288151979446411,
+      "learning_rate": 1.9901496259351624e-05,
+      "loss": 0.563,
+      "step": 400
+    },
+    {
+      "epoch": 0.10224438902743142,
+      "grad_norm": 2.599865674972534,
+      "learning_rate": 1.9899002493765587e-05,
+      "loss": 0.5838,
+      "step": 410
+    },
+    {
+      "epoch": 0.10473815461346633,
+      "grad_norm": 10.21823787689209,
+      "learning_rate": 1.9896508728179554e-05,
+      "loss": 0.8896,
+      "step": 420
+    },
+    {
+      "epoch": 0.10723192019950124,
+      "grad_norm": 3.7210872173309326,
+      "learning_rate": 1.9894014962593518e-05,
+      "loss": 0.7539,
+      "step": 430
+    },
+    {
+      "epoch": 0.10972568578553615,
+      "grad_norm": 3.3652985095977783,
+      "learning_rate": 1.9891521197007485e-05,
+      "loss": 0.6975,
+      "step": 440
+    },
+    {
+      "epoch": 0.11221945137157108,
+      "grad_norm": 4.095602512359619,
+      "learning_rate": 1.988902743142145e-05,
+      "loss": 0.7403,
+      "step": 450
+    },
+    {
+      "epoch": 0.11471321695760599,
+      "grad_norm": 4.6530561447143555,
+      "learning_rate": 1.9886533665835412e-05,
+      "loss": 0.6642,
+      "step": 460
+    },
+    {
+      "epoch": 0.1172069825436409,
+      "grad_norm": 3.8766672611236572,
+      "learning_rate": 1.988403990024938e-05,
+      "loss": 0.6265,
+      "step": 470
+    },
+    {
+      "epoch": 0.11970074812967581,
+      "grad_norm": 6.619718551635742,
+      "learning_rate": 1.9881546134663343e-05,
+      "loss": 0.6444,
+      "step": 480
+    },
+    {
+      "epoch": 0.12219451371571072,
+      "grad_norm": 2.7176246643066406,
+      "learning_rate": 1.9879052369077306e-05,
+      "loss": 0.6325,
+      "step": 490
+    },
+    {
+      "epoch": 0.12468827930174564,
+      "grad_norm": 4.887167930603027,
+      "learning_rate": 1.9876558603491273e-05,
+      "loss": 0.5515,
+      "step": 500
+    },
+    {
+      "epoch": 0.12718204488778054,
+      "grad_norm": 3.827092170715332,
+      "learning_rate": 1.9874064837905237e-05,
+      "loss": 0.8736,
+      "step": 510
+    },
+    {
+      "epoch": 0.12967581047381546,
+      "grad_norm": 3.9637207984924316,
+      "learning_rate": 1.9871571072319204e-05,
+      "loss": 0.6267,
+      "step": 520
+    },
+    {
+      "epoch": 0.13216957605985039,
+      "grad_norm": 4.160803318023682,
+      "learning_rate": 1.9869077306733168e-05,
+      "loss": 0.6462,
+      "step": 530
+    },
+    {
+      "epoch": 0.13466334164588528,
+      "grad_norm": 4.465733528137207,
+      "learning_rate": 1.9866583541147135e-05,
+      "loss": 0.488,
+      "step": 540
+    },
+    {
+      "epoch": 0.1371571072319202,
+      "grad_norm": 3.197251081466675,
+      "learning_rate": 1.98640897755611e-05,
+      "loss": 0.7185,
+      "step": 550
+    },
+    {
+      "epoch": 0.1396508728179551,
+      "grad_norm": 3.279279947280884,
+      "learning_rate": 1.9861596009975065e-05,
+      "loss": 0.599,
+      "step": 560
+    },
+    {
+      "epoch": 0.14214463840399003,
+      "grad_norm": 7.450759410858154,
+      "learning_rate": 1.985910224438903e-05,
+      "loss": 0.7233,
+      "step": 570
+    },
+    {
+      "epoch": 0.14463840399002495,
+      "grad_norm": 4.170207500457764,
+      "learning_rate": 1.9856608478802996e-05,
+      "loss": 0.5774,
+      "step": 580
+    },
+    {
+      "epoch": 0.14713216957605985,
+      "grad_norm": 9.313362121582031,
+      "learning_rate": 1.985411471321696e-05,
+      "loss": 0.8636,
+      "step": 590
+    },
+    {
+      "epoch": 0.14962593516209477,
+      "grad_norm": 4.21543550491333,
+      "learning_rate": 1.9851620947630927e-05,
+      "loss": 0.648,
+      "step": 600
+    },
+    {
+      "epoch": 0.15211970074812967,
+      "grad_norm": 7.928882122039795,
+      "learning_rate": 1.984912718204489e-05,
+      "loss": 0.6985,
+      "step": 610
+    },
+    {
+      "epoch": 0.1546134663341646,
+      "grad_norm": 3.2070388793945312,
+      "learning_rate": 1.9846633416458854e-05,
+      "loss": 0.5127,
+      "step": 620
+    },
+    {
+      "epoch": 0.1571072319201995,
+      "grad_norm": 4.8116278648376465,
+      "learning_rate": 1.984413965087282e-05,
+      "loss": 0.6521,
+      "step": 630
+    },
+    {
+      "epoch": 0.1596009975062344,
+      "grad_norm": 6.319545269012451,
+      "learning_rate": 1.9841645885286784e-05,
+      "loss": 0.5693,
+      "step": 640
+    },
+    {
+      "epoch": 0.16209476309226933,
+      "grad_norm": 3.7397208213806152,
+      "learning_rate": 1.983915211970075e-05,
+      "loss": 0.6335,
+      "step": 650
+    },
+    {
+      "epoch": 0.16458852867830423,
+      "grad_norm": 4.853140830993652,
+      "learning_rate": 1.9836658354114715e-05,
+      "loss": 0.7663,
+      "step": 660
+    },
+    {
+      "epoch": 0.16708229426433915,
+      "grad_norm": 3.868441104888916,
+      "learning_rate": 1.983416458852868e-05,
+      "loss": 0.6261,
+      "step": 670
+    },
+    {
+      "epoch": 0.16957605985037408,
+      "grad_norm": 5.102773189544678,
+      "learning_rate": 1.9831670822942646e-05,
+      "loss": 0.7153,
+      "step": 680
+    },
+    {
+      "epoch": 0.17206982543640897,
+      "grad_norm": 4.242384433746338,
+      "learning_rate": 1.982917705735661e-05,
+      "loss": 0.6275,
+      "step": 690
+    },
+    {
+      "epoch": 0.1745635910224439,
+      "grad_norm": 3.3436532020568848,
+      "learning_rate": 1.9826683291770573e-05,
+      "loss": 0.5709,
+      "step": 700
+    },
+    {
+      "epoch": 0.1770573566084788,
+      "grad_norm": 5.803040027618408,
+      "learning_rate": 1.982418952618454e-05,
+      "loss": 0.5584,
+      "step": 710
+    },
+    {
+      "epoch": 0.17955112219451372,
+      "grad_norm": 5.347147464752197,
+      "learning_rate": 1.9821695760598503e-05,
+      "loss": 0.5785,
+      "step": 720
+    },
+    {
+      "epoch": 0.18204488778054864,
+      "grad_norm": 4.12513542175293,
+      "learning_rate": 1.981920199501247e-05,
+      "loss": 0.6152,
+      "step": 730
+    },
+    {
+      "epoch": 0.18453865336658354,
+      "grad_norm": 5.8045220375061035,
+      "learning_rate": 1.9816708229426434e-05,
+      "loss": 0.6319,
+      "step": 740
+    },
+    {
+      "epoch": 0.18703241895261846,
+      "grad_norm": 8.319589614868164,
+      "learning_rate": 1.98142144638404e-05,
+      "loss": 0.5286,
+      "step": 750
+    },
+    {
+      "epoch": 0.18952618453865336,
+      "grad_norm": 3.427260637283325,
+      "learning_rate": 1.9811720698254365e-05,
+      "loss": 0.6325,
+      "step": 760
+    },
+    {
+      "epoch": 0.19201995012468828,
+      "grad_norm": 12.969819068908691,
+      "learning_rate": 1.9809226932668332e-05,
+      "loss": 0.6051,
+      "step": 770
+    },
+    {
+      "epoch": 0.19451371571072318,
+      "grad_norm": 4.117243766784668,
+      "learning_rate": 1.9806733167082295e-05,
+      "loss": 0.7746,
+      "step": 780
+    },
+    {
+      "epoch": 0.1970074812967581,
+      "grad_norm": 8.8555326461792,
+      "learning_rate": 1.9804239401496262e-05,
+      "loss": 0.6756,
+      "step": 790
+    },
+    {
+      "epoch": 0.19950124688279303,
+      "grad_norm": 6.0134596824646,
+      "learning_rate": 1.9801745635910226e-05,
+      "loss": 0.479,
+      "step": 800
+    },
+    {
+      "epoch": 0.20199501246882792,
+      "grad_norm": 3.967231273651123,
+      "learning_rate": 1.9799251870324193e-05,
+      "loss": 0.5358,
+      "step": 810
+    },
+    {
+      "epoch": 0.20448877805486285,
+      "grad_norm": 3.7455625534057617,
+      "learning_rate": 1.9796758104738157e-05,
+      "loss": 0.5306,
+      "step": 820
+    },
+    {
+      "epoch": 0.20698254364089774,
+      "grad_norm": 4.493707656860352,
+      "learning_rate": 1.979426433915212e-05,
+      "loss": 0.6243,
+      "step": 830
+    },
+    {
+      "epoch": 0.20947630922693267,
+      "grad_norm": 4.1963653564453125,
+      "learning_rate": 1.9791770573566087e-05,
+      "loss": 0.6031,
+      "step": 840
+    },
+    {
+      "epoch": 0.2119700748129676,
+      "grad_norm": 3.4451446533203125,
+      "learning_rate": 1.978927680798005e-05,
+      "loss": 0.5625,
+      "step": 850
+    },
+    {
+      "epoch": 0.2144638403990025,
+      "grad_norm": 10.554472923278809,
+      "learning_rate": 1.9786783042394014e-05,
+      "loss": 0.7107,
+      "step": 860
+    },
+    {
+      "epoch": 0.2169576059850374,
+      "grad_norm": 6.821500778198242,
+      "learning_rate": 1.978428927680798e-05,
+      "loss": 0.5901,
+      "step": 870
+    },
+    {
+      "epoch": 0.2194513715710723,
+      "grad_norm": 8.539198875427246,
+      "learning_rate": 1.9781795511221945e-05,
+      "loss": 0.6157,
+      "step": 880
+    },
+    {
+      "epoch": 0.22194513715710723,
+      "grad_norm": 5.176815032958984,
+      "learning_rate": 1.9779301745635912e-05,
+      "loss": 0.5484,
+      "step": 890
+    },
+    {
+      "epoch": 0.22443890274314215,
+      "grad_norm": 7.525144577026367,
+      "learning_rate": 1.9776807980049876e-05,
+      "loss": 0.6059,
+      "step": 900
+    },
+    {
+      "epoch": 0.22693266832917705,
+      "grad_norm": 6.711145401000977,
+      "learning_rate": 1.9774314214463843e-05,
+      "loss": 0.6239,
+      "step": 910
+    },
+    {
+      "epoch": 0.22942643391521197,
+      "grad_norm": 4.520840644836426,
+      "learning_rate": 1.9771820448877806e-05,
+      "loss": 0.5336,
+      "step": 920
+    },
+    {
+      "epoch": 0.23192019950124687,
+      "grad_norm": 11.031719207763672,
+      "learning_rate": 1.9769326683291773e-05,
+      "loss": 0.5649,
+      "step": 930
+    },
+    {
+      "epoch": 0.2344139650872818,
+      "grad_norm": 7.578179836273193,
+      "learning_rate": 1.9766832917705737e-05,
+      "loss": 0.6233,
+      "step": 940
+    },
+    {
+      "epoch": 0.23690773067331672,
+      "grad_norm": 4.446279048919678,
+      "learning_rate": 1.9764339152119704e-05,
+      "loss": 0.6529,
+      "step": 950
+    },
+    {
+      "epoch": 0.23940149625935161,
+      "grad_norm": 5.39076042175293,
+      "learning_rate": 1.9761845386533668e-05,
+      "loss": 0.6355,
+      "step": 960
+    },
+    {
+      "epoch": 0.24189526184538654,
+      "grad_norm": 5.869710445404053,
+      "learning_rate": 1.9759351620947635e-05,
+      "loss": 0.6045,
+      "step": 970
+    },
+    {
+      "epoch": 0.24438902743142144,
+      "grad_norm": 4.218006610870361,
+      "learning_rate": 1.9756857855361598e-05,
+      "loss": 0.5569,
+      "step": 980
+    },
+    {
+      "epoch": 0.24688279301745636,
+      "grad_norm": 4.749696731567383,
+      "learning_rate": 1.9754364089775562e-05,
+      "loss": 0.5298,
+      "step": 990
+    },
+    {
+      "epoch": 0.24937655860349128,
+      "grad_norm": 7.573767185211182,
+      "learning_rate": 1.975187032418953e-05,
+      "loss": 0.5561,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2518703241895262,
+      "grad_norm": 5.48563289642334,
+      "learning_rate": 1.9749376558603492e-05,
+      "loss": 0.5115,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2543640897755611,
+      "grad_norm": 6.635676383972168,
+      "learning_rate": 1.974688279301746e-05,
+      "loss": 0.6161,
+      "step": 1020
+    },
+    {
+      "epoch": 0.256857855361596,
+      "grad_norm": 4.294478893280029,
+      "learning_rate": 1.9744389027431423e-05,
+      "loss": 0.5136,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2593516209476309,
+      "grad_norm": 2.8301312923431396,
+      "learning_rate": 1.9741895261845387e-05,
+      "loss": 0.451,
+      "step": 1040
+    },
+    {
+      "epoch": 0.26184538653366585,
+      "grad_norm": 4.692950248718262,
+      "learning_rate": 1.9739401496259354e-05,
+      "loss": 0.5172,
+      "step": 1050
+    },
+    {
+      "epoch": 0.26433915211970077,
+      "grad_norm": 9.1524076461792,
+      "learning_rate": 1.9736907730673317e-05,
+      "loss": 0.7668,
+      "step": 1060
+    },
+    {
+      "epoch": 0.26683291770573564,
+      "grad_norm": 3.0480797290802,
+      "learning_rate": 1.973441396508728e-05,
+      "loss": 0.5275,
+      "step": 1070
+    },
+    {
+      "epoch": 0.26932668329177056,
+      "grad_norm": 3.7727952003479004,
+      "learning_rate": 1.9731920199501248e-05,
+      "loss": 0.6045,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2718204488778055,
+      "grad_norm": 3.7919719219207764,
+      "learning_rate": 1.972942643391521e-05,
+      "loss": 0.568,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2743142144638404,
+      "grad_norm": 4.75301456451416,
+      "learning_rate": 1.972693266832918e-05,
+      "loss": 0.567,
+      "step": 1100
+    },
+    {
+      "epoch": 0.27680798004987534,
+      "grad_norm": 3.6362855434417725,
+      "learning_rate": 1.9724438902743142e-05,
+      "loss": 0.5015,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2793017456359102,
+      "grad_norm": 13.187034606933594,
+      "learning_rate": 1.972194513715711e-05,
+      "loss": 0.5433,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2817955112219451,
+      "grad_norm": 4.809338092803955,
+      "learning_rate": 1.9719451371571076e-05,
+      "loss": 0.5669,
+      "step": 1130
+    },
+    {
+      "epoch": 0.28428927680798005,
+      "grad_norm": 3.7646372318267822,
+      "learning_rate": 1.971695760598504e-05,
+      "loss": 0.5286,
+      "step": 1140
+    },
+    {
+      "epoch": 0.286783042394015,
+      "grad_norm": 3.752352237701416,
+      "learning_rate": 1.9714463840399007e-05,
+      "loss": 0.5286,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2892768079800499,
+      "grad_norm": 4.207645416259766,
+      "learning_rate": 1.971197007481297e-05,
+      "loss": 0.5052,
+      "step": 1160
+    },
+    {
+      "epoch": 0.29177057356608477,
+      "grad_norm": 3.126208782196045,
+      "learning_rate": 1.9709476309226934e-05,
+      "loss": 0.5188,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2942643391521197,
+      "grad_norm": 4.613558769226074,
+      "learning_rate": 1.97069825436409e-05,
+      "loss": 0.5405,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2967581047381546,
+      "grad_norm": 4.35943603515625,
+      "learning_rate": 1.9704488778054865e-05,
+      "loss": 0.4617,
+      "step": 1190
+    },
+    {
+      "epoch": 0.29925187032418954,
+      "grad_norm": 3.4332399368286133,
+      "learning_rate": 1.9701995012468828e-05,
+      "loss": 0.5354,
+      "step": 1200
+    },
+    {
+      "epoch": 0.30174563591022446,
+      "grad_norm": 3.738420009613037,
+      "learning_rate": 1.9699501246882795e-05,
+      "loss": 0.4887,
+      "step": 1210
+    },
+    {
+      "epoch": 0.30423940149625933,
+      "grad_norm": 6.625039577484131,
+      "learning_rate": 1.969700748129676e-05,
+      "loss": 0.5395,
+      "step": 1220
+    },
+    {
+      "epoch": 0.30673316708229426,
+      "grad_norm": 3.6015474796295166,
+      "learning_rate": 1.9694513715710726e-05,
+      "loss": 0.4949,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3092269326683292,
+      "grad_norm": 9.316793441772461,
+      "learning_rate": 1.969201995012469e-05,
+      "loss": 0.6079,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3117206982543641,
+      "grad_norm": 10.953588485717773,
+      "learning_rate": 1.9689526184538653e-05,
+      "loss": 0.524,
+      "step": 1250
+    },
+    {
+      "epoch": 0.314214463840399,
+      "grad_norm": 5.9117512702941895,
+      "learning_rate": 1.968703241895262e-05,
+      "loss": 0.6395,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3167082294264339,
+      "grad_norm": 3.799406051635742,
+      "learning_rate": 1.9684538653366584e-05,
+      "loss": 0.5604,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3192019950124688,
+      "grad_norm": 5.208715915679932,
+      "learning_rate": 1.968204488778055e-05,
+      "loss": 0.5423,
+      "step": 1280
+    },
+    {
+      "epoch": 0.32169576059850374,
+      "grad_norm": 8.043681144714355,
+      "learning_rate": 1.9679551122194514e-05,
+      "loss": 0.5482,
+      "step": 1290
+    },
+    {
+      "epoch": 0.32418952618453867,
+      "grad_norm": 4.668685436248779,
+      "learning_rate": 1.967705735660848e-05,
+      "loss": 0.5868,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3266832917705736,
+      "grad_norm": 5.136122703552246,
+      "learning_rate": 1.9674563591022445e-05,
+      "loss": 0.5217,
+      "step": 1310
+    },
+    {
+      "epoch": 0.32917705735660846,
+      "grad_norm": 3.8749747276306152,
+      "learning_rate": 1.9672069825436412e-05,
+      "loss": 0.48,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3316708229426434,
+      "grad_norm": 2.9240078926086426,
+      "learning_rate": 1.9669576059850376e-05,
+      "loss": 0.6384,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3341645885286783,
+      "grad_norm": 3.375347852706909,
+      "learning_rate": 1.9667082294264343e-05,
+      "loss": 0.5041,
+      "step": 1340
+    },
+    {
+      "epoch": 0.33665835411471323,
+      "grad_norm": 3.96232008934021,
+      "learning_rate": 1.9664588528678306e-05,
+      "loss": 0.4538,
+      "step": 1350
+    },
+    {
+      "epoch": 0.33915211970074816,
+      "grad_norm": 16.26675796508789,
+      "learning_rate": 1.966209476309227e-05,
+      "loss": 0.4686,
+      "step": 1360
+    },
+    {
+      "epoch": 0.341645885286783,
+      "grad_norm": 6.03434944152832,
+      "learning_rate": 1.9659600997506237e-05,
+      "loss": 0.5372,
+      "step": 1370
+    },
+    {
+      "epoch": 0.34413965087281795,
+      "grad_norm": 4.742402076721191,
+      "learning_rate": 1.96571072319202e-05,
+      "loss": 0.4793,
+      "step": 1380
+    },
+    {
+      "epoch": 0.34663341645885287,
+      "grad_norm": 4.454680442810059,
+      "learning_rate": 1.9654613466334167e-05,
+      "loss": 0.4901,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3491271820448878,
+      "grad_norm": 4.6647796630859375,
+      "learning_rate": 1.965211970074813e-05,
+      "loss": 0.5023,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3516209476309227,
+      "grad_norm": 5.384413719177246,
+      "learning_rate": 1.9649625935162095e-05,
+      "loss": 0.5912,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3541147132169576,
+      "grad_norm": 6.326955795288086,
+      "learning_rate": 1.964713216957606e-05,
+      "loss": 0.5661,
+      "step": 1420
+    },
+    {
+      "epoch": 0.3566084788029925,
+      "grad_norm": 3.840012550354004,
+      "learning_rate": 1.9644638403990025e-05,
+      "loss": 0.5107,
+      "step": 1430
+    },
+    {
+      "epoch": 0.35910224438902744,
+      "grad_norm": 5.383571147918701,
+      "learning_rate": 1.9642144638403992e-05,
+      "loss": 0.4882,
+      "step": 1440
+    },
+    {
+      "epoch": 0.36159600997506236,
+      "grad_norm": 2.377634048461914,
+      "learning_rate": 1.9639650872817956e-05,
+      "loss": 0.5656,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3640897755610973,
+      "grad_norm": 3.8716111183166504,
+      "learning_rate": 1.963715710723192e-05,
+      "loss": 0.5097,
+      "step": 1460
+    },
+    {
+      "epoch": 0.36658354114713215,
+      "grad_norm": 5.310560703277588,
+      "learning_rate": 1.9634663341645886e-05,
+      "loss": 0.4505,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3690773067331671,
+      "grad_norm": 7.158385753631592,
+      "learning_rate": 1.9632169576059853e-05,
+      "loss": 0.4722,
+      "step": 1480
+    },
+    {
+      "epoch": 0.371571072319202,
+      "grad_norm": 3.6055357456207275,
+      "learning_rate": 1.9629675810473817e-05,
+      "loss": 0.4542,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3740648379052369,
+      "grad_norm": 3.826148748397827,
+      "learning_rate": 1.9627182044887784e-05,
+      "loss": 0.5559,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3765586034912718,
+      "grad_norm": 5.25114631652832,
+      "learning_rate": 1.9624688279301748e-05,
+      "loss": 0.5079,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3790523690773067,
+      "grad_norm": 5.912806034088135,
+      "learning_rate": 1.9622194513715715e-05,
+      "loss": 0.5651,
+      "step": 1520
+    },
+    {
+      "epoch": 0.38154613466334164,
+      "grad_norm": 5.626669883728027,
+      "learning_rate": 1.961970074812968e-05,
+      "loss": 0.483,
+      "step": 1530
+    },
+    {
+      "epoch": 0.38403990024937656,
+      "grad_norm": 5.233978271484375,
+      "learning_rate": 1.9617206982543642e-05,
+      "loss": 0.472,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3865336658354115,
+      "grad_norm": 4.9565606117248535,
+      "learning_rate": 1.961471321695761e-05,
+      "loss": 0.5137,
+      "step": 1550
+    },
+    {
+      "epoch": 0.38902743142144636,
+      "grad_norm": 5.3332414627075195,
+      "learning_rate": 1.9612219451371573e-05,
+      "loss": 0.5708,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3915211970074813,
+      "grad_norm": 4.999671459197998,
+      "learning_rate": 1.9609725685785536e-05,
+      "loss": 0.5666,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3940149625935162,
+      "grad_norm": 10.932353019714355,
+      "learning_rate": 1.9607231920199503e-05,
+      "loss": 0.5769,
+      "step": 1580
+    },
+    {
+      "epoch": 0.39650872817955113,
+      "grad_norm": 7.1936211585998535,
+      "learning_rate": 1.9604738154613467e-05,
+      "loss": 0.5631,
+      "step": 1590
+    },
+    {
+      "epoch": 0.39900249376558605,
+      "grad_norm": 6.666138172149658,
+      "learning_rate": 1.9602244389027434e-05,
+      "loss": 0.5239,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4014962593516209,
+      "grad_norm": 3.0620758533477783,
+      "learning_rate": 1.9599750623441397e-05,
+      "loss": 0.5028,
+      "step": 1610
+    },
+    {
+      "epoch": 0.40399002493765584,
+      "grad_norm": 4.261651515960693,
+      "learning_rate": 1.959725685785536e-05,
+      "loss": 0.4711,
+      "step": 1620
+    },
+    {
+      "epoch": 0.40648379052369077,
+      "grad_norm": 3.4371354579925537,
+      "learning_rate": 1.9594763092269328e-05,
+      "loss": 0.4729,
+      "step": 1630
+    },
+    {
+      "epoch": 0.4089775561097257,
+      "grad_norm": 3.691063642501831,
+      "learning_rate": 1.959226932668329e-05,
+      "loss": 0.4844,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4114713216957606,
+      "grad_norm": 4.673091411590576,
+      "learning_rate": 1.958977556109726e-05,
+      "loss": 0.5131,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4139650872817955,
+      "grad_norm": 2.6619250774383545,
+      "learning_rate": 1.9587281795511222e-05,
+      "loss": 0.5196,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4164588528678304,
+      "grad_norm": 6.241903781890869,
+      "learning_rate": 1.958478802992519e-05,
+      "loss": 0.471,
+      "step": 1670
+    },
+    {
+      "epoch": 0.41895261845386533,
+      "grad_norm": 3.603950023651123,
+      "learning_rate": 1.9582294264339153e-05,
+      "loss": 0.4476,
+      "step": 1680
+    },
+    {
+      "epoch": 0.42144638403990026,
+      "grad_norm": 3.6078081130981445,
+      "learning_rate": 1.957980049875312e-05,
+      "loss": 0.5308,
+      "step": 1690
+    },
+    {
+      "epoch": 0.4239401496259352,
+      "grad_norm": 4.073026657104492,
+      "learning_rate": 1.9577306733167084e-05,
+      "loss": 0.4973,
+      "step": 1700
+    },
+    {
+      "epoch": 0.42643391521197005,
+      "grad_norm": 4.738262176513672,
+      "learning_rate": 1.957481296758105e-05,
+      "loss": 0.4709,
+      "step": 1710
+    },
+    {
+      "epoch": 0.428927680798005,
+      "grad_norm": 16.924922943115234,
+      "learning_rate": 1.9572319201995014e-05,
+      "loss": 0.5383,
+      "step": 1720
+    },
+    {
+      "epoch": 0.4314214463840399,
+      "grad_norm": 8.632024765014648,
+      "learning_rate": 1.956982543640898e-05,
+      "loss": 0.4519,
+      "step": 1730
+    },
+    {
+      "epoch": 0.4339152119700748,
+      "grad_norm": 8.484312057495117,
+      "learning_rate": 1.9567331670822945e-05,
+      "loss": 0.5642,
+      "step": 1740
+    },
+    {
+      "epoch": 0.43640897755610975,
+      "grad_norm": 4.789985179901123,
+      "learning_rate": 1.956483790523691e-05,
+      "loss": 0.4486,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4389027431421446,
+      "grad_norm": 4.4060587882995605,
+      "learning_rate": 1.9562344139650875e-05,
+      "loss": 0.5041,
+      "step": 1760
+    },
+    {
+      "epoch": 0.44139650872817954,
+      "grad_norm": 8.840462684631348,
+      "learning_rate": 1.955985037406484e-05,
+      "loss": 0.5466,
+      "step": 1770
+    },
+    {
+      "epoch": 0.44389027431421446,
+      "grad_norm": 3.662015438079834,
+      "learning_rate": 1.9557356608478803e-05,
+      "loss": 0.4922,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4463840399002494,
+      "grad_norm": 5.66322135925293,
+      "learning_rate": 1.955486284289277e-05,
+      "loss": 0.4851,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4488778054862843,
+      "grad_norm": 3.9291810989379883,
+      "learning_rate": 1.9552369077306733e-05,
+      "loss": 0.5185,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4513715710723192,
+      "grad_norm": 4.53109073638916,
+      "learning_rate": 1.95498753117207e-05,
+      "loss": 0.5114,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4538653366583541,
+      "grad_norm": 3.999675750732422,
+      "learning_rate": 1.9547381546134664e-05,
+      "loss": 0.4159,
+      "step": 1820
+    },
+    {
+      "epoch": 0.456359102244389,
+      "grad_norm": 8.522893905639648,
+      "learning_rate": 1.954488778054863e-05,
+      "loss": 0.74,
+      "step": 1830
+    },
+    {
+      "epoch": 0.45885286783042395,
+      "grad_norm": 3.3957653045654297,
+      "learning_rate": 1.9542394014962594e-05,
+      "loss": 0.5078,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4613466334164589,
+      "grad_norm": 2.780721664428711,
+      "learning_rate": 1.953990024937656e-05,
+      "loss": 0.5148,
+      "step": 1850
+    },
+    {
+      "epoch": 0.46384039900249374,
+      "grad_norm": 3.9767260551452637,
+      "learning_rate": 1.9537406483790525e-05,
+      "loss": 0.5263,
+      "step": 1860
+    },
+    {
+      "epoch": 0.46633416458852867,
+      "grad_norm": 15.764490127563477,
+      "learning_rate": 1.9534912718204492e-05,
+      "loss": 0.4691,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4688279301745636,
+      "grad_norm": 7.174173831939697,
+      "learning_rate": 1.9532418952618456e-05,
+      "loss": 0.5482,
+      "step": 1880
+    },
+    {
+      "epoch": 0.4713216957605985,
+      "grad_norm": 3.9777376651763916,
+      "learning_rate": 1.9529925187032423e-05,
+      "loss": 0.4499,
+      "step": 1890
+    },
+    {
+      "epoch": 0.47381546134663344,
+      "grad_norm": 4.089155197143555,
+      "learning_rate": 1.9527431421446386e-05,
+      "loss": 0.4241,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4763092269326683,
+      "grad_norm": 3.909698009490967,
+      "learning_rate": 1.952493765586035e-05,
+      "loss": 0.5389,
+      "step": 1910
+    },
+    {
+      "epoch": 0.47880299251870323,
+      "grad_norm": 4.601511478424072,
+      "learning_rate": 1.9522443890274317e-05,
+      "loss": 0.5299,
+      "step": 1920
+    },
+    {
+      "epoch": 0.48129675810473815,
+      "grad_norm": 3.4147753715515137,
+      "learning_rate": 1.951995012468828e-05,
+      "loss": 0.4071,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4837905236907731,
+      "grad_norm": 10.128273963928223,
+      "learning_rate": 1.9517456359102248e-05,
+      "loss": 0.5228,
+      "step": 1940
+    },
+    {
+      "epoch": 0.486284289276808,
+      "grad_norm": 6.3098464012146,
+      "learning_rate": 1.951496259351621e-05,
+      "loss": 0.5648,
+      "step": 1950
+    },
+    {
+      "epoch": 0.48877805486284287,
+      "grad_norm": 9.323540687561035,
+      "learning_rate": 1.9512468827930175e-05,
+      "loss": 0.5566,
+      "step": 1960
+    },
+    {
+      "epoch": 0.4912718204488778,
+      "grad_norm": 2.8909482955932617,
+      "learning_rate": 1.9509975062344142e-05,
+      "loss": 0.4574,
+      "step": 1970
+    },
+    {
+      "epoch": 0.4937655860349127,
+      "grad_norm": 8.425788879394531,
+      "learning_rate": 1.9507481296758105e-05,
+      "loss": 0.594,
+      "step": 1980
+    },
+    {
+      "epoch": 0.49625935162094764,
+      "grad_norm": 5.776350498199463,
+      "learning_rate": 1.950498753117207e-05,
+      "loss": 0.4679,
+      "step": 1990
+    },
+    {
+      "epoch": 0.49875311720698257,
+      "grad_norm": 5.987747669219971,
+      "learning_rate": 1.9502493765586036e-05,
+      "loss": 0.5371,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5012468827930174,
+      "grad_norm": 3.423344373703003,
+      "learning_rate": 1.95e-05,
+      "loss": 0.4901,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5037406483790524,
+      "grad_norm": 3.8344333171844482,
+      "learning_rate": 1.9497506234413967e-05,
+      "loss": 0.4937,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5062344139650873,
+      "grad_norm": 4.415685653686523,
+      "learning_rate": 1.949501246882793e-05,
+      "loss": 0.4658,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5087281795511222,
+      "grad_norm": 4.0992631912231445,
+      "learning_rate": 1.9492518703241897e-05,
+      "loss": 0.4264,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5112219451371571,
+      "grad_norm": 6.2230305671691895,
+      "learning_rate": 1.949002493765586e-05,
+      "loss": 0.5018,
+      "step": 2050
+    },
+    {
+      "epoch": 0.513715710723192,
+      "grad_norm": 7.8946661949157715,
+      "learning_rate": 1.9487531172069828e-05,
+      "loss": 0.477,
+      "step": 2060
+    },
+    {
+      "epoch": 0.516209476309227,
+      "grad_norm": 6.977099418640137,
+      "learning_rate": 1.948503740648379e-05,
+      "loss": 0.4574,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5187032418952618,
+      "grad_norm": 5.954398155212402,
+      "learning_rate": 1.948254364089776e-05,
+      "loss": 0.5273,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5211970074812967,
+      "grad_norm": 5.636336326599121,
+      "learning_rate": 1.9480049875311722e-05,
+      "loss": 0.4623,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5236907730673317,
+      "grad_norm": 7.925063610076904,
+      "learning_rate": 1.947755610972569e-05,
+      "loss": 0.4776,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5261845386533666,
+      "grad_norm": 6.0404276847839355,
+      "learning_rate": 1.9475062344139653e-05,
+      "loss": 0.5091,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5286783042394015,
+      "grad_norm": 5.379130840301514,
+      "learning_rate": 1.9472568578553616e-05,
+      "loss": 0.4522,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5311720698254364,
+      "grad_norm": 5.63524055480957,
+      "learning_rate": 1.9470074812967583e-05,
+      "loss": 0.5296,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5336658354114713,
+      "grad_norm": 3.867295026779175,
+      "learning_rate": 1.9467581047381547e-05,
+      "loss": 0.4351,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5361596009975063,
+      "grad_norm": 10.726486206054688,
+      "learning_rate": 1.9465087281795514e-05,
+      "loss": 0.4772,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5386533665835411,
+      "grad_norm": 5.092080593109131,
+      "learning_rate": 1.9462593516209478e-05,
+      "loss": 0.4877,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5411471321695761,
+      "grad_norm": 4.903257846832275,
+      "learning_rate": 1.946009975062344e-05,
+      "loss": 0.5226,
+      "step": 2170
+    },
+    {
+      "epoch": 0.543640897755611,
+      "grad_norm": 12.123847007751465,
+      "learning_rate": 1.9457605985037408e-05,
+      "loss": 0.4826,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5461346633416458,
+      "grad_norm": 7.251633644104004,
+      "learning_rate": 1.9455112219451372e-05,
+      "loss": 0.5337,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5486284289276808,
+      "grad_norm": 5.906688213348389,
+      "learning_rate": 1.945261845386534e-05,
+      "loss": 0.4971,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5511221945137157,
+      "grad_norm": 16.242473602294922,
+      "learning_rate": 1.9450124688279302e-05,
+      "loss": 0.5198,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5536159600997507,
+      "grad_norm": 5.047711372375488,
+      "learning_rate": 1.944763092269327e-05,
+      "loss": 0.5753,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5561097256857855,
+      "grad_norm": 12.766270637512207,
+      "learning_rate": 1.9445137157107233e-05,
+      "loss": 0.4736,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5586034912718204,
+      "grad_norm": 9.127399444580078,
+      "learning_rate": 1.94426433915212e-05,
+      "loss": 0.4879,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5610972568578554,
+      "grad_norm": 5.662039756774902,
+      "learning_rate": 1.9440149625935164e-05,
+      "loss": 0.4147,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5635910224438903,
+      "grad_norm": 5.060816764831543,
+      "learning_rate": 1.943765586034913e-05,
+      "loss": 0.4406,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5660847880299252,
+      "grad_norm": 7.149324893951416,
+      "learning_rate": 1.9435162094763094e-05,
+      "loss": 0.5203,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5685785536159601,
+      "grad_norm": 3.712177038192749,
+      "learning_rate": 1.9432668329177058e-05,
+      "loss": 0.4658,
+      "step": 2280
+    },
+    {
+      "epoch": 0.571072319201995,
+      "grad_norm": 3.989894151687622,
+      "learning_rate": 1.9430174563591025e-05,
+      "loss": 0.4428,
+      "step": 2290
+    },
+    {
+      "epoch": 0.57356608478803,
+      "grad_norm": 4.171450614929199,
+      "learning_rate": 1.942768079800499e-05,
+      "loss": 0.5034,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5760598503740648,
+      "grad_norm": 5.396224021911621,
+      "learning_rate": 1.9425187032418956e-05,
+      "loss": 0.4929,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5785536159600998,
+      "grad_norm": 5.547377586364746,
+      "learning_rate": 1.942269326683292e-05,
+      "loss": 0.4344,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5810473815461347,
+      "grad_norm": 9.413858413696289,
+      "learning_rate": 1.9420199501246883e-05,
+      "loss": 0.4321,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5835411471321695,
+      "grad_norm": 6.898621082305908,
+      "learning_rate": 1.941770573566085e-05,
+      "loss": 0.5428,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5860349127182045,
+      "grad_norm": 4.387313365936279,
+      "learning_rate": 1.9415211970074813e-05,
+      "loss": 0.4786,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5885286783042394,
+      "grad_norm": 8.502609252929688,
+      "learning_rate": 1.941271820448878e-05,
+      "loss": 0.4587,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5910224438902744,
+      "grad_norm": 5.494476795196533,
+      "learning_rate": 1.9410224438902744e-05,
+      "loss": 0.5174,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5935162094763092,
+      "grad_norm": 4.911831378936768,
+      "learning_rate": 1.9407730673316708e-05,
+      "loss": 0.5318,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5960099750623441,
+      "grad_norm": 3.805800676345825,
+      "learning_rate": 1.9405236907730675e-05,
+      "loss": 0.5523,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5985037406483791,
+      "grad_norm": 7.420222759246826,
+      "learning_rate": 1.9402743142144638e-05,
+      "loss": 0.492,
+      "step": 2400
+    },
+    {
+      "epoch": 0.600997506234414,
+      "grad_norm": 5.9345245361328125,
+      "learning_rate": 1.9400249376558605e-05,
+      "loss": 0.4199,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6034912718204489,
+      "grad_norm": 4.605949878692627,
+      "learning_rate": 1.9397755610972572e-05,
+      "loss": 0.5301,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6059850374064838,
+      "grad_norm": 11.447821617126465,
+      "learning_rate": 1.9395261845386536e-05,
+      "loss": 0.5429,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6084788029925187,
+      "grad_norm": 4.791404724121094,
+      "learning_rate": 1.9392768079800503e-05,
+      "loss": 0.5576,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6109725685785536,
+      "grad_norm": 4.776665210723877,
+      "learning_rate": 1.9390274314214466e-05,
+      "loss": 0.4874,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6134663341645885,
+      "grad_norm": 8.155165672302246,
+      "learning_rate": 1.938778054862843e-05,
+      "loss": 0.5492,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6159600997506235,
+      "grad_norm": 15.007184028625488,
+      "learning_rate": 1.9385286783042397e-05,
+      "loss": 0.4631,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6184538653366584,
+      "grad_norm": 3.208120107650757,
+      "learning_rate": 1.938279301745636e-05,
+      "loss": 0.4789,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6209476309226932,
+      "grad_norm": 5.233438014984131,
+      "learning_rate": 1.9380299251870324e-05,
+      "loss": 0.5402,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6234413965087282,
+      "grad_norm": 8.981165885925293,
+      "learning_rate": 1.937780548628429e-05,
+      "loss": 0.491,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6259351620947631,
+      "grad_norm": 8.05521011352539,
+      "learning_rate": 1.9375311720698255e-05,
+      "loss": 0.4621,
+      "step": 2510
+    },
+    {
+      "epoch": 0.628428927680798,
+      "grad_norm": 3.094724178314209,
+      "learning_rate": 1.9372817955112222e-05,
+      "loss": 0.4244,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6309226932668329,
+      "grad_norm": 5.453985691070557,
+      "learning_rate": 1.9370324189526186e-05,
+      "loss": 0.5035,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6334164588528678,
+      "grad_norm": 4.8853559494018555,
+      "learning_rate": 1.936783042394015e-05,
+      "loss": 0.4758,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6359102244389028,
+      "grad_norm": 4.26847505569458,
+      "learning_rate": 1.9365336658354116e-05,
+      "loss": 0.493,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6384039900249376,
+      "grad_norm": 7.400485038757324,
+      "learning_rate": 1.936284289276808e-05,
+      "loss": 0.465,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6408977556109726,
+      "grad_norm": 4.4760260581970215,
+      "learning_rate": 1.9360349127182047e-05,
+      "loss": 0.4429,
+      "step": 2570
+    },
+    {
+      "epoch": 0.6433915211970075,
+      "grad_norm": 4.798295974731445,
+      "learning_rate": 1.935785536159601e-05,
+      "loss": 0.4837,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6458852867830424,
+      "grad_norm": 11.160234451293945,
+      "learning_rate": 1.9355361596009977e-05,
+      "loss": 0.4985,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6483790523690773,
+      "grad_norm": 6.34352970123291,
+      "learning_rate": 1.935286783042394e-05,
+      "loss": 0.5137,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6508728179551122,
+      "grad_norm": 4.095518589019775,
+      "learning_rate": 1.9350374064837908e-05,
+      "loss": 0.4254,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6533665835411472,
+      "grad_norm": 8.039348602294922,
+      "learning_rate": 1.934788029925187e-05,
+      "loss": 0.4655,
+      "step": 2620
+    },
+    {
+      "epoch": 0.655860349127182,
+      "grad_norm": 4.135050296783447,
+      "learning_rate": 1.934538653366584e-05,
+      "loss": 0.5008,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6583541147132169,
+      "grad_norm": 7.032421112060547,
+      "learning_rate": 1.9342892768079802e-05,
+      "loss": 0.4439,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6608478802992519,
+      "grad_norm": 4.22358512878418,
+      "learning_rate": 1.934039900249377e-05,
+      "loss": 0.5356,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6633416458852868,
+      "grad_norm": 6.675922393798828,
+      "learning_rate": 1.9337905236907733e-05,
+      "loss": 0.5038,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6658354114713217,
+      "grad_norm": 4.241270542144775,
+      "learning_rate": 1.9335411471321697e-05,
+      "loss": 0.4468,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6683291770573566,
+      "grad_norm": 4.54661226272583,
+      "learning_rate": 1.9332917705735664e-05,
+      "loss": 0.4758,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6708229426433915,
+      "grad_norm": 3.0999062061309814,
+      "learning_rate": 1.9330423940149627e-05,
+      "loss": 0.566,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6733167082294265,
+      "grad_norm": 5.668433666229248,
+      "learning_rate": 1.932793017456359e-05,
+      "loss": 0.4979,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6758104738154613,
+      "grad_norm": 7.616546154022217,
+      "learning_rate": 1.9325436408977558e-05,
+      "loss": 0.3973,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6783042394014963,
+      "grad_norm": 4.385749816894531,
+      "learning_rate": 1.932294264339152e-05,
+      "loss": 0.5626,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6807980049875312,
+      "grad_norm": 12.706242561340332,
+      "learning_rate": 1.932044887780549e-05,
+      "loss": 0.5379,
+      "step": 2730
+    },
+    {
+      "epoch": 0.683291770573566,
+      "grad_norm": 4.375461101531982,
+      "learning_rate": 1.9317955112219452e-05,
+      "loss": 0.4326,
+      "step": 2740
+    },
+    {
+      "epoch": 0.685785536159601,
+      "grad_norm": 4.31388521194458,
+      "learning_rate": 1.931546134663342e-05,
+      "loss": 0.5309,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6882793017456359,
+      "grad_norm": 4.78411865234375,
+      "learning_rate": 1.9312967581047383e-05,
+      "loss": 0.4068,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6907730673316709,
+      "grad_norm": 10.348709106445312,
+      "learning_rate": 1.931047381546135e-05,
+      "loss": 0.4246,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6932668329177057,
+      "grad_norm": 3.510700225830078,
+      "learning_rate": 1.9307980049875313e-05,
+      "loss": 0.4845,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6957605985037406,
+      "grad_norm": 4.097596168518066,
+      "learning_rate": 1.930548628428928e-05,
+      "loss": 0.5605,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6982543640897756,
+      "grad_norm": 4.968656539916992,
+      "learning_rate": 1.9302992518703244e-05,
+      "loss": 0.467,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7007481296758105,
+      "grad_norm": 7.420864105224609,
+      "learning_rate": 1.930049875311721e-05,
+      "loss": 0.6162,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7032418952618454,
+      "grad_norm": 7.818281173706055,
+      "learning_rate": 1.9298004987531174e-05,
+      "loss": 0.4224,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7057356608478803,
+      "grad_norm": 6.321879863739014,
+      "learning_rate": 1.9295511221945138e-05,
+      "loss": 0.5507,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7082294264339152,
+      "grad_norm": 4.185654640197754,
+      "learning_rate": 1.9293017456359105e-05,
+      "loss": 0.4075,
+      "step": 2840
+    },
+    {
+      "epoch": 0.7107231920199502,
+      "grad_norm": 4.573608875274658,
+      "learning_rate": 1.929052369077307e-05,
+      "loss": 0.5097,
+      "step": 2850
+    },
+    {
+      "epoch": 0.713216957605985,
+      "grad_norm": 6.115342617034912,
+      "learning_rate": 1.9288029925187036e-05,
+      "loss": 0.4963,
+      "step": 2860
+    },
+    {
+      "epoch": 0.71571072319202,
+      "grad_norm": 4.295200824737549,
+      "learning_rate": 1.9285536159601e-05,
+      "loss": 0.4322,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7182044887780549,
+      "grad_norm": 2.7827277183532715,
+      "learning_rate": 1.9283042394014963e-05,
+      "loss": 0.4749,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7206982543640897,
+      "grad_norm": 4.560075283050537,
+      "learning_rate": 1.928054862842893e-05,
+      "loss": 0.4603,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7231920199501247,
+      "grad_norm": 6.625837326049805,
+      "learning_rate": 1.9278054862842894e-05,
+      "loss": 0.4779,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7256857855361596,
+      "grad_norm": 6.110123634338379,
+      "learning_rate": 1.9275561097256857e-05,
+      "loss": 0.5923,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7281795511221946,
+      "grad_norm": 5.3894429206848145,
+      "learning_rate": 1.9273067331670824e-05,
+      "loss": 0.4797,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7306733167082294,
+      "grad_norm": 6.485326290130615,
+      "learning_rate": 1.9270573566084788e-05,
+      "loss": 0.5242,
+      "step": 2930
+    },
+    {
+      "epoch": 0.7331670822942643,
+      "grad_norm": 8.976933479309082,
+      "learning_rate": 1.9268079800498755e-05,
+      "loss": 0.44,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7356608478802993,
+      "grad_norm": 8.325933456420898,
+      "learning_rate": 1.926558603491272e-05,
+      "loss": 0.5028,
+      "step": 2950
+    },
+    {
+      "epoch": 0.7381546134663342,
+      "grad_norm": 4.603588581085205,
+      "learning_rate": 1.9263092269326685e-05,
+      "loss": 0.4457,
+      "step": 2960
+    },
+    {
+      "epoch": 0.7406483790523691,
+      "grad_norm": 4.730616569519043,
+      "learning_rate": 1.926059850374065e-05,
+      "loss": 0.4515,
+      "step": 2970
+    },
+    {
+      "epoch": 0.743142144638404,
+      "grad_norm": 4.1613264083862305,
+      "learning_rate": 1.9258104738154616e-05,
+      "loss": 0.4357,
+      "step": 2980
+    },
+    {
+      "epoch": 0.7456359102244389,
+      "grad_norm": 4.212600231170654,
+      "learning_rate": 1.925561097256858e-05,
+      "loss": 0.4426,
+      "step": 2990
+    },
+    {
+      "epoch": 0.7481296758104738,
+      "grad_norm": 5.657212257385254,
+      "learning_rate": 1.9253117206982547e-05,
+      "loss": 0.4656,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7506234413965087,
+      "grad_norm": 6.499466419219971,
+      "learning_rate": 1.925062344139651e-05,
+      "loss": 0.421,
+      "step": 3010
+    },
+    {
+      "epoch": 0.7531172069825436,
+      "grad_norm": 4.284633159637451,
+      "learning_rate": 1.9248129675810477e-05,
+      "loss": 0.5375,
+      "step": 3020
+    },
+    {
+      "epoch": 0.7556109725685786,
+      "grad_norm": 4.252184867858887,
+      "learning_rate": 1.924563591022444e-05,
+      "loss": 0.4443,
+      "step": 3030
+    },
+    {
+      "epoch": 0.7581047381546134,
+      "grad_norm": 4.567617416381836,
+      "learning_rate": 1.9243142144638405e-05,
+      "loss": 0.4453,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7605985037406484,
+      "grad_norm": 3.3405144214630127,
+      "learning_rate": 1.924064837905237e-05,
+      "loss": 0.432,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7630922693266833,
+      "grad_norm": 7.207590579986572,
+      "learning_rate": 1.9238154613466335e-05,
+      "loss": 0.446,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7655860349127181,
+      "grad_norm": 7.11346960067749,
+      "learning_rate": 1.92356608478803e-05,
+      "loss": 0.4085,
+      "step": 3070
+    },
+    {
+      "epoch": 0.7680798004987531,
+      "grad_norm": 3.7571468353271484,
+      "learning_rate": 1.9233167082294266e-05,
+      "loss": 0.4873,
+      "step": 3080
+    },
+    {
+      "epoch": 0.770573566084788,
+      "grad_norm": 6.159874439239502,
+      "learning_rate": 1.923067331670823e-05,
+      "loss": 0.5184,
+      "step": 3090
+    },
+    {
+      "epoch": 0.773067331670823,
+      "grad_norm": 6.5550689697265625,
+      "learning_rate": 1.9228179551122196e-05,
+      "loss": 0.4748,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7755610972568578,
+      "grad_norm": 6.619629383087158,
+      "learning_rate": 1.922568578553616e-05,
+      "loss": 0.4853,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7780548628428927,
+      "grad_norm": 5.71950626373291,
+      "learning_rate": 1.9223192019950127e-05,
+      "loss": 0.4853,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7805486284289277,
+      "grad_norm": 4.437851905822754,
+      "learning_rate": 1.922069825436409e-05,
+      "loss": 0.4058,
+      "step": 3130
+    },
+    {
+      "epoch": 0.7830423940149626,
+      "grad_norm": 13.215971946716309,
+      "learning_rate": 1.9218204488778058e-05,
+      "loss": 0.51,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7855361596009975,
+      "grad_norm": 4.856098175048828,
+      "learning_rate": 1.921571072319202e-05,
+      "loss": 0.4832,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7880299251870324,
+      "grad_norm": 6.450615882873535,
+      "learning_rate": 1.9213216957605988e-05,
+      "loss": 0.4538,
+      "step": 3160
+    },
+    {
+      "epoch": 0.7905236907730673,
+      "grad_norm": 7.96874475479126,
+      "learning_rate": 1.9210723192019952e-05,
+      "loss": 0.4463,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7930174563591023,
+      "grad_norm": 5.789658069610596,
+      "learning_rate": 1.920822942643392e-05,
+      "loss": 0.3977,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7955112219451371,
+      "grad_norm": 4.556354999542236,
+      "learning_rate": 1.9205735660847882e-05,
+      "loss": 0.5665,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7980049875311721,
+      "grad_norm": 7.1619181632995605,
+      "learning_rate": 1.9203241895261846e-05,
+      "loss": 0.4884,
+      "step": 3200
+    },
+    {
+      "epoch": 0.800498753117207,
+      "grad_norm": 5.486998081207275,
+      "learning_rate": 1.9200748129675813e-05,
+      "loss": 0.5249,
+      "step": 3210
+    },
+    {
+      "epoch": 0.8029925187032418,
+      "grad_norm": 5.061128616333008,
+      "learning_rate": 1.9198254364089777e-05,
+      "loss": 0.5166,
+      "step": 3220
+    },
+    {
+      "epoch": 0.8054862842892768,
+      "grad_norm": 5.740884780883789,
+      "learning_rate": 1.9195760598503744e-05,
+      "loss": 0.4312,
+      "step": 3230
+    },
+    {
+      "epoch": 0.8079800498753117,
+      "grad_norm": 6.897968769073486,
+      "learning_rate": 1.9193266832917707e-05,
+      "loss": 0.4602,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8104738154613467,
+      "grad_norm": 11.832649230957031,
+      "learning_rate": 1.919077306733167e-05,
+      "loss": 0.4456,
+      "step": 3250
+    },
+    {
+      "epoch": 0.8129675810473815,
+      "grad_norm": 6.845297336578369,
+      "learning_rate": 1.9188279301745638e-05,
+      "loss": 0.5935,
+      "step": 3260
+    },
+    {
+      "epoch": 0.8154613466334164,
+      "grad_norm": 7.5892157554626465,
+      "learning_rate": 1.91857855361596e-05,
+      "loss": 0.5575,
+      "step": 3270
+    },
+    {
+      "epoch": 0.8179551122194514,
+      "grad_norm": 5.861032009124756,
+      "learning_rate": 1.9183291770573565e-05,
+      "loss": 0.4618,
+      "step": 3280
+    },
+    {
+      "epoch": 0.8204488778054863,
+      "grad_norm": 7.565516471862793,
+      "learning_rate": 1.9180798004987532e-05,
+      "loss": 0.5081,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8229426433915212,
+      "grad_norm": 3.7106189727783203,
+      "learning_rate": 1.9178304239401496e-05,
+      "loss": 0.4926,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8254364089775561,
+      "grad_norm": 4.886735916137695,
+      "learning_rate": 1.9175810473815463e-05,
+      "loss": 0.491,
+      "step": 3310
+    },
+    {
+      "epoch": 0.827930174563591,
+      "grad_norm": 4.354279041290283,
+      "learning_rate": 1.9173316708229426e-05,
+      "loss": 0.3565,
+      "step": 3320
+    },
+    {
+      "epoch": 0.830423940149626,
+      "grad_norm": 6.275864601135254,
+      "learning_rate": 1.9170822942643393e-05,
+      "loss": 0.4157,
+      "step": 3330
+    },
+    {
+      "epoch": 0.8329177057356608,
+      "grad_norm": 5.567061901092529,
+      "learning_rate": 1.916832917705736e-05,
+      "loss": 0.5054,
+      "step": 3340
+    },
+    {
+      "epoch": 0.8354114713216958,
+      "grad_norm": 5.522775650024414,
+      "learning_rate": 1.9165835411471324e-05,
+      "loss": 0.504,
+      "step": 3350
+    },
+    {
+      "epoch": 0.8379052369077307,
+      "grad_norm": 4.974456310272217,
+      "learning_rate": 1.916334164588529e-05,
+      "loss": 0.4118,
+      "step": 3360
+    },
+    {
+      "epoch": 0.8403990024937655,
+      "grad_norm": 3.5197205543518066,
+      "learning_rate": 1.9160847880299255e-05,
+      "loss": 0.4819,
+      "step": 3370
+    },
+    {
+      "epoch": 0.8428927680798005,
+      "grad_norm": 4.410624027252197,
+      "learning_rate": 1.9158354114713218e-05,
+      "loss": 0.583,
+      "step": 3380
+    },
+    {
+      "epoch": 0.8453865336658354,
+      "grad_norm": 6.812956809997559,
+      "learning_rate": 1.9155860349127185e-05,
+      "loss": 0.463,
+      "step": 3390
+    },
+    {
+      "epoch": 0.8478802992518704,
+      "grad_norm": 17.613037109375,
+      "learning_rate": 1.915336658354115e-05,
+      "loss": 0.4853,
+      "step": 3400
+    },
+    {
+      "epoch": 0.8503740648379052,
+      "grad_norm": 6.056275844573975,
+      "learning_rate": 1.9150872817955113e-05,
+      "loss": 0.5459,
+      "step": 3410
+    },
+    {
+      "epoch": 0.8528678304239401,
+      "grad_norm": 4.816941261291504,
+      "learning_rate": 1.914837905236908e-05,
+      "loss": 0.4013,
+      "step": 3420
+    },
+    {
+      "epoch": 0.8553615960099751,
+      "grad_norm": 5.128777027130127,
+      "learning_rate": 1.9145885286783043e-05,
+      "loss": 0.47,
+      "step": 3430
+    },
+    {
+      "epoch": 0.85785536159601,
+      "grad_norm": 5.4511566162109375,
+      "learning_rate": 1.914339152119701e-05,
+      "loss": 0.4487,
+      "step": 3440
+    },
+    {
+      "epoch": 0.8603491271820449,
+      "grad_norm": 7.571670055389404,
+      "learning_rate": 1.9140897755610974e-05,
+      "loss": 0.3869,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8628428927680798,
+      "grad_norm": 5.735408782958984,
+      "learning_rate": 1.9138403990024937e-05,
+      "loss": 0.4186,
+      "step": 3460
+    },
+    {
+      "epoch": 0.8653366583541147,
+      "grad_norm": 5.8194475173950195,
+      "learning_rate": 1.9135910224438904e-05,
+      "loss": 0.5033,
+      "step": 3470
+    },
+    {
+      "epoch": 0.8678304239401496,
+      "grad_norm": 6.551326274871826,
+      "learning_rate": 1.9133416458852868e-05,
+      "loss": 0.4501,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8703241895261845,
+      "grad_norm": 7.928218841552734,
+      "learning_rate": 1.9130922693266835e-05,
+      "loss": 0.492,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8728179551122195,
+      "grad_norm": 4.229366302490234,
+      "learning_rate": 1.91284289276808e-05,
+      "loss": 0.375,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8753117206982544,
+      "grad_norm": 7.293509006500244,
+      "learning_rate": 1.9125935162094766e-05,
+      "loss": 0.4393,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8778054862842892,
+      "grad_norm": 6.139492511749268,
+      "learning_rate": 1.912344139650873e-05,
+      "loss": 0.3927,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8802992518703242,
+      "grad_norm": 7.048768520355225,
+      "learning_rate": 1.9120947630922696e-05,
+      "loss": 0.4707,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8827930174563591,
+      "grad_norm": 4.429152488708496,
+      "learning_rate": 1.911845386533666e-05,
+      "loss": 0.4569,
+      "step": 3540
+    },
+    {
+      "epoch": 0.885286783042394,
+      "grad_norm": 7.218087196350098,
+      "learning_rate": 1.9115960099750627e-05,
+      "loss": 0.4156,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8877805486284289,
+      "grad_norm": 5.999967098236084,
+      "learning_rate": 1.911346633416459e-05,
+      "loss": 0.4679,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8902743142144638,
+      "grad_norm": 5.682243824005127,
+      "learning_rate": 1.9110972568578554e-05,
+      "loss": 0.4587,
+      "step": 3570
+    },
+    {
+      "epoch": 0.8927680798004988,
+      "grad_norm": 4.374421119689941,
+      "learning_rate": 1.910847880299252e-05,
+      "loss": 0.4885,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8952618453865336,
+      "grad_norm": 3.372025966644287,
+      "learning_rate": 1.9105985037406485e-05,
+      "loss": 0.4402,
+      "step": 3590
+    },
+    {
+      "epoch": 0.8977556109725686,
+      "grad_norm": 8.967430114746094,
+      "learning_rate": 1.9103491271820452e-05,
+      "loss": 0.4719,
+      "step": 3600
+    },
+    {
+      "epoch": 0.9002493765586035,
+      "grad_norm": 7.0953850746154785,
+      "learning_rate": 1.9100997506234415e-05,
+      "loss": 0.4882,
+      "step": 3610
+    },
+    {
+      "epoch": 0.9027431421446384,
+      "grad_norm": 7.988110542297363,
+      "learning_rate": 1.909850374064838e-05,
+      "loss": 0.3987,
+      "step": 3620
+    },
+    {
+      "epoch": 0.9052369077306733,
+      "grad_norm": 5.712813377380371,
+      "learning_rate": 1.9096009975062346e-05,
+      "loss": 0.4625,
+      "step": 3630
+    },
+    {
+      "epoch": 0.9077306733167082,
+      "grad_norm": 5.640379905700684,
+      "learning_rate": 1.909351620947631e-05,
+      "loss": 0.4713,
+      "step": 3640
+    },
+    {
+      "epoch": 0.9102244389027432,
+      "grad_norm": 6.33805513381958,
+      "learning_rate": 1.9091022443890277e-05,
+      "loss": 0.499,
+      "step": 3650
+    },
+    {
+      "epoch": 0.912718204488778,
+      "grad_norm": 4.548992156982422,
+      "learning_rate": 1.908852867830424e-05,
+      "loss": 0.5038,
+      "step": 3660
+    },
+    {
+      "epoch": 0.9152119700748129,
+      "grad_norm": 8.613363265991211,
+      "learning_rate": 1.9086034912718204e-05,
+      "loss": 0.484,
+      "step": 3670
+    },
+    {
+      "epoch": 0.9177057356608479,
+      "grad_norm": 4.469968318939209,
+      "learning_rate": 1.908354114713217e-05,
+      "loss": 0.4473,
+      "step": 3680
+    },
+    {
+      "epoch": 0.9201995012468828,
+      "grad_norm": 4.103412628173828,
+      "learning_rate": 1.9081047381546138e-05,
+      "loss": 0.4483,
+      "step": 3690
+    },
+    {
+      "epoch": 0.9226932668329177,
+      "grad_norm": 4.446389675140381,
+      "learning_rate": 1.90785536159601e-05,
+      "loss": 0.4354,
+      "step": 3700
+    },
+    {
+      "epoch": 0.9251870324189526,
+      "grad_norm": 3.4375393390655518,
+      "learning_rate": 1.907605985037407e-05,
+      "loss": 0.5089,
+      "step": 3710
+    },
+    {
+      "epoch": 0.9276807980049875,
+      "grad_norm": 4.128108024597168,
+      "learning_rate": 1.9073566084788032e-05,
+      "loss": 0.4881,
+      "step": 3720
+    },
+    {
+      "epoch": 0.9301745635910225,
+      "grad_norm": 6.5702104568481445,
+      "learning_rate": 1.9071072319202e-05,
+      "loss": 0.4994,
+      "step": 3730
+    },
+    {
+      "epoch": 0.9326683291770573,
+      "grad_norm": 4.294257164001465,
+      "learning_rate": 1.9068578553615963e-05,
+      "loss": 0.5258,
+      "step": 3740
+    },
+    {
+      "epoch": 0.9351620947630923,
+      "grad_norm": 10.037517547607422,
+      "learning_rate": 1.9066084788029926e-05,
+      "loss": 0.5119,
+      "step": 3750
+    },
+    {
+      "epoch": 0.9376558603491272,
+      "grad_norm": 12.020997047424316,
+      "learning_rate": 1.9063591022443893e-05,
+      "loss": 0.5925,
+      "step": 3760
+    },
+    {
+      "epoch": 0.940149625935162,
+      "grad_norm": 6.246609210968018,
+      "learning_rate": 1.9061097256857857e-05,
+      "loss": 0.4407,
+      "step": 3770
+    },
+    {
+      "epoch": 0.942643391521197,
+      "grad_norm": 6.553652286529541,
+      "learning_rate": 1.905860349127182e-05,
+      "loss": 0.4833,
+      "step": 3780
+    },
+    {
+      "epoch": 0.9451371571072319,
+      "grad_norm": 5.495224952697754,
+      "learning_rate": 1.9056109725685788e-05,
+      "loss": 0.4579,
+      "step": 3790
+    },
+    {
+      "epoch": 0.9476309226932669,
+      "grad_norm": 5.287566184997559,
+      "learning_rate": 1.905361596009975e-05,
+      "loss": 0.5213,
+      "step": 3800
+    },
+    {
+      "epoch": 0.9501246882793017,
+      "grad_norm": 5.307123184204102,
+      "learning_rate": 1.9051122194513718e-05,
+      "loss": 0.4347,
+      "step": 3810
+    },
+    {
+      "epoch": 0.9526184538653366,
+      "grad_norm": 6.510472774505615,
+      "learning_rate": 1.9048628428927682e-05,
+      "loss": 0.5203,
+      "step": 3820
+    },
+    {
+      "epoch": 0.9551122194513716,
+      "grad_norm": 4.413476467132568,
+      "learning_rate": 1.9046134663341645e-05,
+      "loss": 0.4359,
+      "step": 3830
+    },
+    {
+      "epoch": 0.9576059850374065,
+      "grad_norm": 4.353250503540039,
+      "learning_rate": 1.9043640897755612e-05,
+      "loss": 0.4169,
+      "step": 3840
+    },
+    {
+      "epoch": 0.9600997506234414,
+      "grad_norm": 6.984455108642578,
+      "learning_rate": 1.9041147132169576e-05,
+      "loss": 0.5156,
+      "step": 3850
+    },
+    {
+      "epoch": 0.9625935162094763,
+      "grad_norm": 4.519622325897217,
+      "learning_rate": 1.9038653366583543e-05,
+      "loss": 0.4545,
+      "step": 3860
+    },
+    {
+      "epoch": 0.9650872817955112,
+      "grad_norm": 4.484631061553955,
+      "learning_rate": 1.9036159600997507e-05,
+      "loss": 0.5115,
+      "step": 3870
+    },
+    {
+      "epoch": 0.9675810473815462,
+      "grad_norm": 6.899198532104492,
+      "learning_rate": 1.9033665835411474e-05,
+      "loss": 0.4979,
+      "step": 3880
+    },
+    {
+      "epoch": 0.970074812967581,
+      "grad_norm": 3.723076581954956,
+      "learning_rate": 1.9031172069825437e-05,
+      "loss": 0.464,
+      "step": 3890
+    },
+    {
+      "epoch": 0.972568578553616,
+      "grad_norm": 6.418664932250977,
+      "learning_rate": 1.9028678304239404e-05,
+      "loss": 0.4556,
+      "step": 3900
+    },
+    {
+      "epoch": 0.9750623441396509,
+      "grad_norm": 3.5927555561065674,
+      "learning_rate": 1.9026184538653368e-05,
+      "loss": 0.3993,
+      "step": 3910
+    },
+    {
+      "epoch": 0.9775561097256857,
+      "grad_norm": 4.519811153411865,
+      "learning_rate": 1.9023690773067335e-05,
+      "loss": 0.5431,
+      "step": 3920
+    },
+    {
+      "epoch": 0.9800498753117207,
+      "grad_norm": 11.50717830657959,
+      "learning_rate": 1.90211970074813e-05,
+      "loss": 0.4571,
+      "step": 3930
+    },
+    {
+      "epoch": 0.9825436408977556,
+      "grad_norm": 4.3499531745910645,
+      "learning_rate": 1.9018703241895265e-05,
+      "loss": 0.4586,
+      "step": 3940
+    },
+    {
+      "epoch": 0.9850374064837906,
+      "grad_norm": 6.324216365814209,
+      "learning_rate": 1.901620947630923e-05,
+      "loss": 0.4397,
+      "step": 3950
+    },
+    {
+      "epoch": 0.9875311720698254,
+      "grad_norm": 6.713108062744141,
+      "learning_rate": 1.9013715710723193e-05,
+      "loss": 0.4178,
+      "step": 3960
+    },
+    {
+      "epoch": 0.9900249376558603,
+      "grad_norm": 4.340822219848633,
+      "learning_rate": 1.901122194513716e-05,
+      "loss": 0.5256,
+      "step": 3970
+    },
+    {
+      "epoch": 0.9925187032418953,
+      "grad_norm": 15.54155445098877,
+      "learning_rate": 1.9008728179551123e-05,
+      "loss": 0.4169,
+      "step": 3980
+    },
+    {
+      "epoch": 0.9950124688279302,
+      "grad_norm": 3.4416344165802,
+      "learning_rate": 1.9006234413965087e-05,
+      "loss": 0.4568,
+      "step": 3990
+    },
+    {
+      "epoch": 0.9975062344139651,
+      "grad_norm": 4.241981506347656,
+      "learning_rate": 1.9003740648379054e-05,
+      "loss": 0.4386,
+      "step": 4000
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7881031036376953,
+      "learning_rate": 1.9001246882793018e-05,
+      "loss": 0.4039,
+      "step": 4010
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.46533510088920593,
+      "eval_runtime": 60.0033,
+      "eval_samples_per_second": 16.716,
+      "eval_steps_per_second": 16.716,
+      "step": 4010
+    },
+    {
+      "epoch": 1.0024937655860349,
+      "grad_norm": 5.6270012855529785,
+      "learning_rate": 1.8998753117206985e-05,
+      "loss": 0.4246,
+      "step": 4020
+    },
+    {
+      "epoch": 1.0049875311720697,
+      "grad_norm": 9.492588996887207,
+      "learning_rate": 1.8996259351620948e-05,
+      "loss": 0.5515,
+      "step": 4030
+    },
+    {
+      "epoch": 1.0074812967581048,
+      "grad_norm": 4.804108142852783,
+      "learning_rate": 1.8993765586034915e-05,
+      "loss": 0.4456,
+      "step": 4040
+    },
+    {
+      "epoch": 1.0099750623441397,
+      "grad_norm": 4.964878082275391,
+      "learning_rate": 1.899127182044888e-05,
+      "loss": 0.418,
+      "step": 4050
+    },
+    {
+      "epoch": 1.0124688279301746,
+      "grad_norm": 4.545513153076172,
+      "learning_rate": 1.8988778054862846e-05,
+      "loss": 0.4251,
+      "step": 4060
+    },
+    {
+      "epoch": 1.0149625935162094,
+      "grad_norm": 4.988970756530762,
+      "learning_rate": 1.898628428927681e-05,
+      "loss": 0.4635,
+      "step": 4070
+    },
+    {
+      "epoch": 1.0174563591022443,
+      "grad_norm": 3.7804431915283203,
+      "learning_rate": 1.8983790523690776e-05,
+      "loss": 0.4948,
+      "step": 4080
+    },
+    {
+      "epoch": 1.0199501246882794,
+      "grad_norm": 6.862154960632324,
+      "learning_rate": 1.898129675810474e-05,
+      "loss": 0.471,
+      "step": 4090
+    },
+    {
+      "epoch": 1.0224438902743143,
+      "grad_norm": 4.912306308746338,
+      "learning_rate": 1.8978802992518707e-05,
+      "loss": 0.4769,
+      "step": 4100
+    },
+    {
+      "epoch": 1.0249376558603491,
+      "grad_norm": 6.113458633422852,
+      "learning_rate": 1.897630922693267e-05,
+      "loss": 0.4927,
+      "step": 4110
+    },
+    {
+      "epoch": 1.027431421446384,
+      "grad_norm": 7.236551761627197,
+      "learning_rate": 1.8973815461346634e-05,
+      "loss": 0.4274,
+      "step": 4120
+    },
+    {
+      "epoch": 1.0299251870324189,
+      "grad_norm": 5.630270481109619,
+      "learning_rate": 1.89713216957606e-05,
+      "loss": 0.4913,
+      "step": 4130
+    },
+    {
+      "epoch": 1.032418952618454,
+      "grad_norm": 7.250363349914551,
+      "learning_rate": 1.8968827930174565e-05,
+      "loss": 0.3752,
+      "step": 4140
+    },
+    {
+      "epoch": 1.0349127182044888,
+      "grad_norm": 4.857931137084961,
+      "learning_rate": 1.8966334164588532e-05,
+      "loss": 0.4775,
+      "step": 4150
+    },
+    {
+      "epoch": 1.0374064837905237,
+      "grad_norm": 4.707160949707031,
+      "learning_rate": 1.8963840399002495e-05,
+      "loss": 0.5688,
+      "step": 4160
+    },
+    {
+      "epoch": 1.0399002493765586,
+      "grad_norm": 6.663334846496582,
+      "learning_rate": 1.896134663341646e-05,
+      "loss": 0.4619,
+      "step": 4170
+    },
+    {
+      "epoch": 1.0423940149625934,
+      "grad_norm": 8.353853225708008,
+      "learning_rate": 1.8958852867830426e-05,
+      "loss": 0.4525,
+      "step": 4180
+    },
+    {
+      "epoch": 1.0448877805486285,
+      "grad_norm": 5.910696506500244,
+      "learning_rate": 1.895635910224439e-05,
+      "loss": 0.4344,
+      "step": 4190
+    },
+    {
+      "epoch": 1.0473815461346634,
+      "grad_norm": 5.698975086212158,
+      "learning_rate": 1.8953865336658353e-05,
+      "loss": 0.4816,
+      "step": 4200
+    },
+    {
+      "epoch": 1.0498753117206983,
+      "grad_norm": 9.77688217163086,
+      "learning_rate": 1.895137157107232e-05,
+      "loss": 0.417,
+      "step": 4210
+    },
+    {
+      "epoch": 1.0523690773067331,
+      "grad_norm": 4.210202693939209,
+      "learning_rate": 1.8948877805486284e-05,
+      "loss": 0.5169,
+      "step": 4220
+    },
+    {
+      "epoch": 1.054862842892768,
+      "grad_norm": 4.911707401275635,
+      "learning_rate": 1.894638403990025e-05,
+      "loss": 0.4293,
+      "step": 4230
+    },
+    {
+      "epoch": 1.057356608478803,
+      "grad_norm": 5.228448867797852,
+      "learning_rate": 1.8943890274314215e-05,
+      "loss": 0.429,
+      "step": 4240
+    },
+    {
+      "epoch": 1.059850374064838,
+      "grad_norm": 5.143721103668213,
+      "learning_rate": 1.894139650872818e-05,
+      "loss": 0.4066,
+      "step": 4250
+    },
+    {
+      "epoch": 1.0623441396508728,
+      "grad_norm": 7.817266464233398,
+      "learning_rate": 1.8938902743142145e-05,
+      "loss": 0.6464,
+      "step": 4260
+    },
+    {
+      "epoch": 1.0648379052369077,
+      "grad_norm": 5.188505172729492,
+      "learning_rate": 1.8936408977556112e-05,
+      "loss": 0.4417,
+      "step": 4270
+    },
+    {
+      "epoch": 1.0673316708229426,
+      "grad_norm": 6.084823131561279,
+      "learning_rate": 1.8933915211970076e-05,
+      "loss": 0.47,
+      "step": 4280
+    },
+    {
+      "epoch": 1.0698254364089776,
+      "grad_norm": 4.737970352172852,
+      "learning_rate": 1.8931421446384043e-05,
+      "loss": 0.5196,
+      "step": 4290
+    },
+    {
+      "epoch": 1.0723192019950125,
+      "grad_norm": 6.13934326171875,
+      "learning_rate": 1.8928927680798006e-05,
+      "loss": 0.3904,
+      "step": 4300
+    },
+    {
+      "epoch": 1.0748129675810474,
+      "grad_norm": 7.485795497894287,
+      "learning_rate": 1.8926433915211973e-05,
+      "loss": 0.4556,
+      "step": 4310
+    },
+    {
+      "epoch": 1.0773067331670823,
+      "grad_norm": 6.431264877319336,
+      "learning_rate": 1.8923940149625937e-05,
+      "loss": 0.4713,
+      "step": 4320
+    },
+    {
+      "epoch": 1.0798004987531171,
+      "grad_norm": 3.5695624351501465,
+      "learning_rate": 1.89214463840399e-05,
+      "loss": 0.4109,
+      "step": 4330
+    },
+    {
+      "epoch": 1.0822942643391522,
+      "grad_norm": 4.416280746459961,
+      "learning_rate": 1.8918952618453868e-05,
+      "loss": 0.4734,
+      "step": 4340
+    },
+    {
+      "epoch": 1.084788029925187,
+      "grad_norm": 4.557003021240234,
+      "learning_rate": 1.891645885286783e-05,
+      "loss": 0.4291,
+      "step": 4350
+    },
+    {
+      "epoch": 1.087281795511222,
+      "grad_norm": 7.2662272453308105,
+      "learning_rate": 1.8913965087281798e-05,
+      "loss": 0.4176,
+      "step": 4360
+    },
+    {
+      "epoch": 1.0897755610972568,
+      "grad_norm": 7.033773422241211,
+      "learning_rate": 1.8911471321695762e-05,
+      "loss": 0.4534,
+      "step": 4370
+    },
+    {
+      "epoch": 1.0922693266832917,
+      "grad_norm": 4.475677490234375,
+      "learning_rate": 1.8908977556109726e-05,
+      "loss": 0.4394,
+      "step": 4380
+    },
+    {
+      "epoch": 1.0947630922693268,
+      "grad_norm": 4.8636698722839355,
+      "learning_rate": 1.8906483790523693e-05,
+      "loss": 0.4371,
+      "step": 4390
+    },
+    {
+      "epoch": 1.0972568578553616,
+      "grad_norm": 5.832028865814209,
+      "learning_rate": 1.8903990024937656e-05,
+      "loss": 0.444,
+      "step": 4400
+    },
+    {
+      "epoch": 1.0997506234413965,
+      "grad_norm": 7.826235771179199,
+      "learning_rate": 1.8901496259351623e-05,
+      "loss": 0.5365,
+      "step": 4410
+    },
+    {
+      "epoch": 1.1022443890274314,
+      "grad_norm": 6.200789928436279,
+      "learning_rate": 1.8899002493765587e-05,
+      "loss": 0.4081,
+      "step": 4420
+    },
+    {
+      "epoch": 1.1047381546134662,
+      "grad_norm": 5.708020210266113,
+      "learning_rate": 1.8896508728179554e-05,
+      "loss": 0.5594,
+      "step": 4430
+    },
+    {
+      "epoch": 1.1072319201995013,
+      "grad_norm": 5.138545513153076,
+      "learning_rate": 1.8894014962593517e-05,
+      "loss": 0.4128,
+      "step": 4440
+    },
+    {
+      "epoch": 1.1097256857855362,
+      "grad_norm": 5.702857971191406,
+      "learning_rate": 1.8891521197007484e-05,
+      "loss": 0.4568,
+      "step": 4450
+    },
+    {
+      "epoch": 1.112219451371571,
+      "grad_norm": 5.061423301696777,
+      "learning_rate": 1.8889027431421448e-05,
+      "loss": 0.4298,
+      "step": 4460
+    },
+    {
+      "epoch": 1.114713216957606,
+      "grad_norm": 4.579878330230713,
+      "learning_rate": 1.8886533665835415e-05,
+      "loss": 0.3446,
+      "step": 4470
+    },
+    {
+      "epoch": 1.1172069825436408,
+      "grad_norm": 4.493147373199463,
+      "learning_rate": 1.888403990024938e-05,
+      "loss": 0.4652,
+      "step": 4480
+    },
+    {
+      "epoch": 1.119700748129676,
+      "grad_norm": 3.8753387928009033,
+      "learning_rate": 1.8881546134663342e-05,
+      "loss": 0.4559,
+      "step": 4490
+    },
+    {
+      "epoch": 1.1221945137157108,
+      "grad_norm": 5.9976325035095215,
+      "learning_rate": 1.887905236907731e-05,
+      "loss": 0.4102,
+      "step": 4500
+    },
+    {
+      "epoch": 1.1246882793017456,
+      "grad_norm": 7.784470081329346,
+      "learning_rate": 1.8876558603491273e-05,
+      "loss": 0.4989,
+      "step": 4510
+    },
+    {
+      "epoch": 1.1271820448877805,
+      "grad_norm": 5.6073737144470215,
+      "learning_rate": 1.887406483790524e-05,
+      "loss": 0.3927,
+      "step": 4520
+    },
+    {
+      "epoch": 1.1296758104738154,
+      "grad_norm": 11.797880172729492,
+      "learning_rate": 1.8871571072319203e-05,
+      "loss": 0.5698,
+      "step": 4530
+    },
+    {
+      "epoch": 1.1321695760598505,
+      "grad_norm": 10.158886909484863,
+      "learning_rate": 1.8869077306733167e-05,
+      "loss": 0.3766,
+      "step": 4540
+    },
+    {
+      "epoch": 1.1346633416458853,
+      "grad_norm": 6.6217875480651855,
+      "learning_rate": 1.8866583541147134e-05,
+      "loss": 0.46,
+      "step": 4550
+    },
+    {
+      "epoch": 1.1371571072319202,
+      "grad_norm": 5.993943691253662,
+      "learning_rate": 1.8864089775561098e-05,
+      "loss": 0.512,
+      "step": 4560
+    },
+    {
+      "epoch": 1.139650872817955,
+      "grad_norm": 5.103926181793213,
+      "learning_rate": 1.886159600997506e-05,
+      "loss": 0.4591,
+      "step": 4570
+    },
+    {
+      "epoch": 1.14214463840399,
+      "grad_norm": 6.188269138336182,
+      "learning_rate": 1.885910224438903e-05,
+      "loss": 0.4811,
+      "step": 4580
+    },
+    {
+      "epoch": 1.144638403990025,
+      "grad_norm": 3.5611369609832764,
+      "learning_rate": 1.8856608478802992e-05,
+      "loss": 0.4138,
+      "step": 4590
+    },
+    {
+      "epoch": 1.14713216957606,
+      "grad_norm": 6.32553768157959,
+      "learning_rate": 1.885411471321696e-05,
+      "loss": 0.4718,
+      "step": 4600
+    },
+    {
+      "epoch": 1.1496259351620948,
+      "grad_norm": 5.74522590637207,
+      "learning_rate": 1.8851620947630923e-05,
+      "loss": 0.4678,
+      "step": 4610
+    },
+    {
+      "epoch": 1.1521197007481296,
+      "grad_norm": 3.8611743450164795,
+      "learning_rate": 1.884912718204489e-05,
+      "loss": 0.361,
+      "step": 4620
+    },
+    {
+      "epoch": 1.1546134663341645,
+      "grad_norm": 5.439576625823975,
+      "learning_rate": 1.8846633416458857e-05,
+      "loss": 0.4243,
+      "step": 4630
+    },
+    {
+      "epoch": 1.1571072319201996,
+      "grad_norm": 6.206503391265869,
+      "learning_rate": 1.884413965087282e-05,
+      "loss": 0.4304,
+      "step": 4640
+    },
+    {
+      "epoch": 1.1596009975062345,
+      "grad_norm": 10.82448959350586,
+      "learning_rate": 1.8841645885286787e-05,
+      "loss": 0.5127,
+      "step": 4650
+    },
+    {
+      "epoch": 1.1620947630922693,
+      "grad_norm": 5.439650535583496,
+      "learning_rate": 1.883915211970075e-05,
+      "loss": 0.4592,
+      "step": 4660
+    },
+    {
+      "epoch": 1.1645885286783042,
+      "grad_norm": 4.041094779968262,
+      "learning_rate": 1.8836658354114714e-05,
+      "loss": 0.4481,
+      "step": 4670
+    },
+    {
+      "epoch": 1.167082294264339,
+      "grad_norm": 4.277478218078613,
+      "learning_rate": 1.883416458852868e-05,
+      "loss": 0.4371,
+      "step": 4680
+    },
+    {
+      "epoch": 1.1695760598503742,
+      "grad_norm": 9.352392196655273,
+      "learning_rate": 1.8831670822942645e-05,
+      "loss": 0.4469,
+      "step": 4690
+    },
+    {
+      "epoch": 1.172069825436409,
+      "grad_norm": 6.215358257293701,
+      "learning_rate": 1.882917705735661e-05,
+      "loss": 0.4068,
+      "step": 4700
+    },
+    {
+      "epoch": 1.174563591022444,
+      "grad_norm": 8.063887596130371,
+      "learning_rate": 1.8826683291770576e-05,
+      "loss": 0.4997,
+      "step": 4710
+    },
+    {
+      "epoch": 1.1770573566084788,
+      "grad_norm": 9.79735279083252,
+      "learning_rate": 1.882418952618454e-05,
+      "loss": 0.4281,
+      "step": 4720
+    },
+    {
+      "epoch": 1.1795511221945136,
+      "grad_norm": 8.45605182647705,
+      "learning_rate": 1.8821695760598506e-05,
+      "loss": 0.4207,
+      "step": 4730
+    },
+    {
+      "epoch": 1.1820448877805487,
+      "grad_norm": 5.719277858734131,
+      "learning_rate": 1.881920199501247e-05,
+      "loss": 0.4669,
+      "step": 4740
+    },
+    {
+      "epoch": 1.1845386533665836,
+      "grad_norm": 5.201423645019531,
+      "learning_rate": 1.8816708229426434e-05,
+      "loss": 0.4345,
+      "step": 4750
+    },
+    {
+      "epoch": 1.1870324189526185,
+      "grad_norm": 4.76845121383667,
+      "learning_rate": 1.88142144638404e-05,
+      "loss": 0.4489,
+      "step": 4760
+    },
+    {
+      "epoch": 1.1895261845386533,
+      "grad_norm": 7.2256622314453125,
+      "learning_rate": 1.8811720698254364e-05,
+      "loss": 0.4855,
+      "step": 4770
+    },
+    {
+      "epoch": 1.1920199501246882,
+      "grad_norm": 6.201477527618408,
+      "learning_rate": 1.880922693266833e-05,
+      "loss": 0.4176,
+      "step": 4780
+    },
+    {
+      "epoch": 1.1945137157107233,
+      "grad_norm": 7.298521995544434,
+      "learning_rate": 1.8806733167082295e-05,
+      "loss": 0.4582,
+      "step": 4790
+    },
+    {
+      "epoch": 1.1970074812967582,
+      "grad_norm": 6.724678993225098,
+      "learning_rate": 1.8804239401496262e-05,
+      "loss": 0.502,
+      "step": 4800
+    },
+    {
+      "epoch": 1.199501246882793,
+      "grad_norm": 4.230656623840332,
+      "learning_rate": 1.8801745635910225e-05,
+      "loss": 0.4097,
+      "step": 4810
+    },
+    {
+      "epoch": 1.201995012468828,
+      "grad_norm": 4.771544456481934,
+      "learning_rate": 1.8799251870324192e-05,
+      "loss": 0.4729,
+      "step": 4820
+    },
+    {
+      "epoch": 1.2044887780548628,
+      "grad_norm": 5.5048651695251465,
+      "learning_rate": 1.8796758104738156e-05,
+      "loss": 0.4848,
+      "step": 4830
+    },
+    {
+      "epoch": 1.2069825436408976,
+      "grad_norm": 10.239465713500977,
+      "learning_rate": 1.8794264339152123e-05,
+      "loss": 0.3863,
+      "step": 4840
+    },
+    {
+      "epoch": 1.2094763092269327,
+      "grad_norm": 6.2510986328125,
+      "learning_rate": 1.8791770573566087e-05,
+      "loss": 0.4414,
+      "step": 4850
+    },
+    {
+      "epoch": 1.2119700748129676,
+      "grad_norm": 4.370057106018066,
+      "learning_rate": 1.8789276807980054e-05,
+      "loss": 0.4307,
+      "step": 4860
+    },
+    {
+      "epoch": 1.2144638403990025,
+      "grad_norm": 6.856778621673584,
+      "learning_rate": 1.8786783042394017e-05,
+      "loss": 0.5257,
+      "step": 4870
+    },
+    {
+      "epoch": 1.2169576059850373,
+      "grad_norm": 3.6856114864349365,
+      "learning_rate": 1.878428927680798e-05,
+      "loss": 0.433,
+      "step": 4880
+    },
+    {
+      "epoch": 1.2194513715710724,
+      "grad_norm": 4.310407638549805,
+      "learning_rate": 1.8781795511221948e-05,
+      "loss": 0.4211,
+      "step": 4890
+    },
+    {
+      "epoch": 1.2219451371571073,
+      "grad_norm": 4.885754585266113,
+      "learning_rate": 1.877930174563591e-05,
+      "loss": 0.3955,
+      "step": 4900
+    },
+    {
+      "epoch": 1.2244389027431422,
+      "grad_norm": 5.621823787689209,
+      "learning_rate": 1.8776807980049875e-05,
+      "loss": 0.4374,
+      "step": 4910
+    },
+    {
+      "epoch": 1.226932668329177,
+      "grad_norm": 8.302716255187988,
+      "learning_rate": 1.8774314214463842e-05,
+      "loss": 0.4245,
+      "step": 4920
+    },
+    {
+      "epoch": 1.229426433915212,
+      "grad_norm": 5.974333763122559,
+      "learning_rate": 1.8771820448877806e-05,
+      "loss": 0.547,
+      "step": 4930
+    },
+    {
+      "epoch": 1.2319201995012468,
+      "grad_norm": 6.3131303787231445,
+      "learning_rate": 1.8769326683291773e-05,
+      "loss": 0.4662,
+      "step": 4940
+    },
+    {
+      "epoch": 1.2344139650872819,
+      "grad_norm": 5.7204389572143555,
+      "learning_rate": 1.8766832917705736e-05,
+      "loss": 0.5053,
+      "step": 4950
+    },
+    {
+      "epoch": 1.2369077306733167,
+      "grad_norm": 4.9055047035217285,
+      "learning_rate": 1.87643391521197e-05,
+      "loss": 0.5201,
+      "step": 4960
+    },
+    {
+      "epoch": 1.2394014962593516,
+      "grad_norm": 3.9648923873901367,
+      "learning_rate": 1.8761845386533667e-05,
+      "loss": 0.4802,
+      "step": 4970
+    },
+    {
+      "epoch": 1.2418952618453865,
+      "grad_norm": 7.1016364097595215,
+      "learning_rate": 1.8759351620947634e-05,
+      "loss": 0.4384,
+      "step": 4980
+    },
+    {
+      "epoch": 1.2443890274314215,
+      "grad_norm": 4.486512184143066,
+      "learning_rate": 1.8756857855361598e-05,
+      "loss": 0.4456,
+      "step": 4990
+    },
+    {
+      "epoch": 1.2468827930174564,
+      "grad_norm": 5.411483287811279,
+      "learning_rate": 1.8754364089775565e-05,
+      "loss": 0.4317,
+      "step": 5000
+    },
+    {
+      "epoch": 1.2493765586034913,
+      "grad_norm": 6.580456733703613,
+      "learning_rate": 1.8751870324189528e-05,
+      "loss": 0.4559,
+      "step": 5010
+    },
+    {
+      "epoch": 1.2518703241895262,
+      "grad_norm": 3.885417938232422,
+      "learning_rate": 1.8749376558603495e-05,
+      "loss": 0.4265,
+      "step": 5020
+    },
+    {
+      "epoch": 1.254364089775561,
+      "grad_norm": 5.22758150100708,
+      "learning_rate": 1.874688279301746e-05,
+      "loss": 0.4173,
+      "step": 5030
+    },
+    {
+      "epoch": 1.2568578553615959,
+      "grad_norm": 6.1364827156066895,
+      "learning_rate": 1.8744389027431422e-05,
+      "loss": 0.4624,
+      "step": 5040
+    },
+    {
+      "epoch": 1.259351620947631,
+      "grad_norm": 6.770162105560303,
+      "learning_rate": 1.874189526184539e-05,
+      "loss": 0.4087,
+      "step": 5050
+    },
+    {
+      "epoch": 1.2618453865336658,
+      "grad_norm": 6.774170875549316,
+      "learning_rate": 1.8739401496259353e-05,
+      "loss": 0.4163,
+      "step": 5060
+    },
+    {
+      "epoch": 1.2643391521197007,
+      "grad_norm": 5.055146217346191,
+      "learning_rate": 1.8736907730673317e-05,
+      "loss": 0.4277,
+      "step": 5070
+    },
+    {
+      "epoch": 1.2668329177057356,
+      "grad_norm": 4.696500301361084,
+      "learning_rate": 1.8734413965087284e-05,
+      "loss": 0.3918,
+      "step": 5080
+    },
+    {
+      "epoch": 1.2693266832917707,
+      "grad_norm": 5.685618877410889,
+      "learning_rate": 1.8731920199501247e-05,
+      "loss": 0.5113,
+      "step": 5090
+    },
+    {
+      "epoch": 1.2718204488778055,
+      "grad_norm": 4.808736324310303,
+      "learning_rate": 1.8729426433915214e-05,
+      "loss": 0.4891,
+      "step": 5100
+    },
+    {
+      "epoch": 1.2743142144638404,
+      "grad_norm": 8.052289009094238,
+      "learning_rate": 1.8726932668329178e-05,
+      "loss": 0.4433,
+      "step": 5110
+    },
+    {
+      "epoch": 1.2768079800498753,
+      "grad_norm": 6.9519853591918945,
+      "learning_rate": 1.872443890274314e-05,
+      "loss": 0.4916,
+      "step": 5120
+    },
+    {
+      "epoch": 1.2793017456359101,
+      "grad_norm": 4.836465358734131,
+      "learning_rate": 1.872194513715711e-05,
+      "loss": 0.4411,
+      "step": 5130
+    },
+    {
+      "epoch": 1.281795511221945,
+      "grad_norm": 4.297245025634766,
+      "learning_rate": 1.8719451371571072e-05,
+      "loss": 0.4697,
+      "step": 5140
+    },
+    {
+      "epoch": 1.28428927680798,
+      "grad_norm": 5.085397720336914,
+      "learning_rate": 1.871695760598504e-05,
+      "loss": 0.4445,
+      "step": 5150
+    },
+    {
+      "epoch": 1.286783042394015,
+      "grad_norm": 4.748317241668701,
+      "learning_rate": 1.8714463840399003e-05,
+      "loss": 0.4488,
+      "step": 5160
+    },
+    {
+      "epoch": 1.2892768079800498,
+      "grad_norm": 5.959413528442383,
+      "learning_rate": 1.871197007481297e-05,
+      "loss": 0.4648,
+      "step": 5170
+    },
+    {
+      "epoch": 1.2917705735660847,
+      "grad_norm": 11.551535606384277,
+      "learning_rate": 1.8709476309226933e-05,
+      "loss": 0.445,
+      "step": 5180
+    },
+    {
+      "epoch": 1.2942643391521198,
+      "grad_norm": 6.094724655151367,
+      "learning_rate": 1.87069825436409e-05,
+      "loss": 0.4175,
+      "step": 5190
+    },
+    {
+      "epoch": 1.2967581047381547,
+      "grad_norm": 9.543522834777832,
+      "learning_rate": 1.8704488778054864e-05,
+      "loss": 0.5068,
+      "step": 5200
+    },
+    {
+      "epoch": 1.2992518703241895,
+      "grad_norm": 6.417718410491943,
+      "learning_rate": 1.870199501246883e-05,
+      "loss": 0.4341,
+      "step": 5210
+    },
+    {
+      "epoch": 1.3017456359102244,
+      "grad_norm": 4.876067638397217,
+      "learning_rate": 1.8699501246882795e-05,
+      "loss": 0.5283,
+      "step": 5220
+    },
+    {
+      "epoch": 1.3042394014962593,
+      "grad_norm": 6.962366580963135,
+      "learning_rate": 1.869700748129676e-05,
+      "loss": 0.5235,
+      "step": 5230
+    },
+    {
+      "epoch": 1.3067331670822941,
+      "grad_norm": 6.60381555557251,
+      "learning_rate": 1.8694513715710725e-05,
+      "loss": 0.5055,
+      "step": 5240
+    },
+    {
+      "epoch": 1.3092269326683292,
+      "grad_norm": 4.449953079223633,
+      "learning_rate": 1.869201995012469e-05,
+      "loss": 0.5745,
+      "step": 5250
+    },
+    {
+      "epoch": 1.311720698254364,
+      "grad_norm": 6.488658905029297,
+      "learning_rate": 1.8689526184538656e-05,
+      "loss": 0.4261,
+      "step": 5260
+    },
+    {
+      "epoch": 1.314214463840399,
+      "grad_norm": 4.1007795333862305,
+      "learning_rate": 1.868703241895262e-05,
+      "loss": 0.3836,
+      "step": 5270
+    },
+    {
+      "epoch": 1.3167082294264338,
+      "grad_norm": 7.238208293914795,
+      "learning_rate": 1.8684538653366583e-05,
+      "loss": 0.5217,
+      "step": 5280
+    },
+    {
+      "epoch": 1.319201995012469,
+      "grad_norm": 5.4013590812683105,
+      "learning_rate": 1.868204488778055e-05,
+      "loss": 0.4387,
+      "step": 5290
+    },
+    {
+      "epoch": 1.3216957605985038,
+      "grad_norm": 8.315073013305664,
+      "learning_rate": 1.8679551122194514e-05,
+      "loss": 0.4616,
+      "step": 5300
+    },
+    {
+      "epoch": 1.3241895261845387,
+      "grad_norm": 3.4126434326171875,
+      "learning_rate": 1.867705735660848e-05,
+      "loss": 0.4431,
+      "step": 5310
+    },
+    {
+      "epoch": 1.3266832917705735,
+      "grad_norm": 7.5681681632995605,
+      "learning_rate": 1.8674563591022444e-05,
+      "loss": 0.5109,
+      "step": 5320
+    },
+    {
+      "epoch": 1.3291770573566084,
+      "grad_norm": 5.397674560546875,
+      "learning_rate": 1.867206982543641e-05,
+      "loss": 0.4732,
+      "step": 5330
+    },
+    {
+      "epoch": 1.3316708229426433,
+      "grad_norm": 5.570555686950684,
+      "learning_rate": 1.8669576059850375e-05,
+      "loss": 0.4385,
+      "step": 5340
+    },
+    {
+      "epoch": 1.3341645885286784,
+      "grad_norm": 6.722954273223877,
+      "learning_rate": 1.8667082294264342e-05,
+      "loss": 0.4784,
+      "step": 5350
+    },
+    {
+      "epoch": 1.3366583541147132,
+      "grad_norm": 4.577895164489746,
+      "learning_rate": 1.8664588528678306e-05,
+      "loss": 0.4526,
+      "step": 5360
+    },
+    {
+      "epoch": 1.339152119700748,
+      "grad_norm": 5.092601776123047,
+      "learning_rate": 1.8662094763092273e-05,
+      "loss": 0.37,
+      "step": 5370
+    },
+    {
+      "epoch": 1.341645885286783,
+      "grad_norm": 6.236178398132324,
+      "learning_rate": 1.8659600997506236e-05,
+      "loss": 0.4303,
+      "step": 5380
+    },
+    {
+      "epoch": 1.344139650872818,
+      "grad_norm": 4.475374698638916,
+      "learning_rate": 1.8657107231920203e-05,
+      "loss": 0.4483,
+      "step": 5390
+    },
+    {
+      "epoch": 1.346633416458853,
+      "grad_norm": 4.651284217834473,
+      "learning_rate": 1.8654613466334167e-05,
+      "loss": 0.431,
+      "step": 5400
+    },
+    {
+      "epoch": 1.3491271820448878,
+      "grad_norm": 4.801163673400879,
+      "learning_rate": 1.865211970074813e-05,
+      "loss": 0.4817,
+      "step": 5410
+    },
+    {
+      "epoch": 1.3516209476309227,
+      "grad_norm": 6.051980972290039,
+      "learning_rate": 1.8649625935162097e-05,
+      "loss": 0.4447,
+      "step": 5420
+    },
+    {
+      "epoch": 1.3541147132169575,
+      "grad_norm": 6.429530143737793,
+      "learning_rate": 1.864713216957606e-05,
+      "loss": 0.5124,
+      "step": 5430
+    },
+    {
+      "epoch": 1.3566084788029924,
+      "grad_norm": 9.302543640136719,
+      "learning_rate": 1.8644638403990028e-05,
+      "loss": 0.4148,
+      "step": 5440
+    },
+    {
+      "epoch": 1.3591022443890275,
+      "grad_norm": 6.575656414031982,
+      "learning_rate": 1.864214463840399e-05,
+      "loss": 0.4759,
+      "step": 5450
+    },
+    {
+      "epoch": 1.3615960099750624,
+      "grad_norm": 6.7204084396362305,
+      "learning_rate": 1.8639650872817955e-05,
+      "loss": 0.4483,
+      "step": 5460
+    },
+    {
+      "epoch": 1.3640897755610972,
+      "grad_norm": 9.02580451965332,
+      "learning_rate": 1.8637157107231922e-05,
+      "loss": 0.4601,
+      "step": 5470
+    },
+    {
+      "epoch": 1.366583541147132,
+      "grad_norm": 4.408763885498047,
+      "learning_rate": 1.8634663341645886e-05,
+      "loss": 0.4205,
+      "step": 5480
+    },
+    {
+      "epoch": 1.3690773067331672,
+      "grad_norm": 4.800736427307129,
+      "learning_rate": 1.863216957605985e-05,
+      "loss": 0.4998,
+      "step": 5490
+    },
+    {
+      "epoch": 1.371571072319202,
+      "grad_norm": 7.011260509490967,
+      "learning_rate": 1.8629675810473817e-05,
+      "loss": 0.5127,
+      "step": 5500
+    },
+    {
+      "epoch": 1.374064837905237,
+      "grad_norm": 4.573014259338379,
+      "learning_rate": 1.862718204488778e-05,
+      "loss": 0.4685,
+      "step": 5510
+    },
+    {
+      "epoch": 1.3765586034912718,
+      "grad_norm": 6.397534370422363,
+      "learning_rate": 1.8624688279301747e-05,
+      "loss": 0.464,
+      "step": 5520
+    },
+    {
+      "epoch": 1.3790523690773067,
+      "grad_norm": 4.248035907745361,
+      "learning_rate": 1.862219451371571e-05,
+      "loss": 0.4275,
+      "step": 5530
+    },
+    {
+      "epoch": 1.3815461346633415,
+      "grad_norm": 4.859241008758545,
+      "learning_rate": 1.8619700748129678e-05,
+      "loss": 0.3563,
+      "step": 5540
+    },
+    {
+      "epoch": 1.3840399002493766,
+      "grad_norm": 4.7854905128479,
+      "learning_rate": 1.861720698254364e-05,
+      "loss": 0.4193,
+      "step": 5550
+    },
+    {
+      "epoch": 1.3865336658354115,
+      "grad_norm": 17.388259887695312,
+      "learning_rate": 1.861471321695761e-05,
+      "loss": 0.5405,
+      "step": 5560
+    },
+    {
+      "epoch": 1.3890274314214464,
+      "grad_norm": 25.828964233398438,
+      "learning_rate": 1.8612219451371572e-05,
+      "loss": 0.4526,
+      "step": 5570
+    },
+    {
+      "epoch": 1.3915211970074812,
+      "grad_norm": 12.78157901763916,
+      "learning_rate": 1.860972568578554e-05,
+      "loss": 0.45,
+      "step": 5580
+    },
+    {
+      "epoch": 1.3940149625935163,
+      "grad_norm": 5.025924205780029,
+      "learning_rate": 1.8607231920199503e-05,
+      "loss": 0.4668,
+      "step": 5590
+    },
+    {
+      "epoch": 1.3965087281795512,
+      "grad_norm": 3.784602403640747,
+      "learning_rate": 1.860473815461347e-05,
+      "loss": 0.4034,
+      "step": 5600
+    },
+    {
+      "epoch": 1.399002493765586,
+      "grad_norm": 5.713232517242432,
+      "learning_rate": 1.8602244389027433e-05,
+      "loss": 0.4374,
+      "step": 5610
+    },
+    {
+      "epoch": 1.401496259351621,
+      "grad_norm": 5.523266792297363,
+      "learning_rate": 1.8599750623441397e-05,
+      "loss": 0.4581,
+      "step": 5620
+    },
+    {
+      "epoch": 1.4039900249376558,
+      "grad_norm": 12.597986221313477,
+      "learning_rate": 1.8597256857855364e-05,
+      "loss": 0.4511,
+      "step": 5630
+    },
+    {
+      "epoch": 1.4064837905236907,
+      "grad_norm": 3.8616204261779785,
+      "learning_rate": 1.8594763092269327e-05,
+      "loss": 0.4322,
+      "step": 5640
+    },
+    {
+      "epoch": 1.4089775561097257,
+      "grad_norm": 6.135002613067627,
+      "learning_rate": 1.8592269326683294e-05,
+      "loss": 0.4043,
+      "step": 5650
+    },
+    {
+      "epoch": 1.4114713216957606,
+      "grad_norm": 4.191798210144043,
+      "learning_rate": 1.8589775561097258e-05,
+      "loss": 0.4508,
+      "step": 5660
+    },
+    {
+      "epoch": 1.4139650872817955,
+      "grad_norm": 5.872029781341553,
+      "learning_rate": 1.858728179551122e-05,
+      "loss": 0.4515,
+      "step": 5670
+    },
+    {
+      "epoch": 1.4164588528678304,
+      "grad_norm": 7.221251010894775,
+      "learning_rate": 1.858478802992519e-05,
+      "loss": 0.4471,
+      "step": 5680
+    },
+    {
+      "epoch": 1.4189526184538654,
+      "grad_norm": 3.4195024967193604,
+      "learning_rate": 1.8582294264339152e-05,
+      "loss": 0.3757,
+      "step": 5690
+    },
+    {
+      "epoch": 1.4214463840399003,
+      "grad_norm": 4.992902755737305,
+      "learning_rate": 1.857980049875312e-05,
+      "loss": 0.4098,
+      "step": 5700
+    },
+    {
+      "epoch": 1.4239401496259352,
+      "grad_norm": 7.851417064666748,
+      "learning_rate": 1.8577306733167083e-05,
+      "loss": 0.54,
+      "step": 5710
+    },
+    {
+      "epoch": 1.42643391521197,
+      "grad_norm": 5.539144992828369,
+      "learning_rate": 1.857481296758105e-05,
+      "loss": 0.4191,
+      "step": 5720
+    },
+    {
+      "epoch": 1.428927680798005,
+      "grad_norm": 9.11573314666748,
+      "learning_rate": 1.8572319201995014e-05,
+      "loss": 0.4237,
+      "step": 5730
+    },
+    {
+      "epoch": 1.4314214463840398,
+      "grad_norm": 5.815394401550293,
+      "learning_rate": 1.856982543640898e-05,
+      "loss": 0.4564,
+      "step": 5740
+    },
+    {
+      "epoch": 1.4339152119700749,
+      "grad_norm": 4.174201011657715,
+      "learning_rate": 1.8567331670822944e-05,
+      "loss": 0.4493,
+      "step": 5750
+    },
+    {
+      "epoch": 1.4364089775561097,
+      "grad_norm": 7.393326759338379,
+      "learning_rate": 1.856483790523691e-05,
+      "loss": 0.4579,
+      "step": 5760
+    },
+    {
+      "epoch": 1.4389027431421446,
+      "grad_norm": 4.805573463439941,
+      "learning_rate": 1.8562344139650875e-05,
+      "loss": 0.4525,
+      "step": 5770
+    },
+    {
+      "epoch": 1.4413965087281795,
+      "grad_norm": 4.3430495262146,
+      "learning_rate": 1.855985037406484e-05,
+      "loss": 0.4323,
+      "step": 5780
+    },
+    {
+      "epoch": 1.4438902743142146,
+      "grad_norm": 4.935792922973633,
+      "learning_rate": 1.8557356608478805e-05,
+      "loss": 0.4649,
+      "step": 5790
+    },
+    {
+      "epoch": 1.4463840399002494,
+      "grad_norm": 3.4280688762664795,
+      "learning_rate": 1.855486284289277e-05,
+      "loss": 0.4597,
+      "step": 5800
+    },
+    {
+      "epoch": 1.4488778054862843,
+      "grad_norm": 5.937455654144287,
+      "learning_rate": 1.8552369077306736e-05,
+      "loss": 0.4628,
+      "step": 5810
+    },
+    {
+      "epoch": 1.4513715710723192,
+      "grad_norm": 4.315957069396973,
+      "learning_rate": 1.85498753117207e-05,
+      "loss": 0.3815,
+      "step": 5820
+    },
+    {
+      "epoch": 1.453865336658354,
+      "grad_norm": 6.386058330535889,
+      "learning_rate": 1.8547381546134663e-05,
+      "loss": 0.5149,
+      "step": 5830
+    },
+    {
+      "epoch": 1.456359102244389,
+      "grad_norm": 7.211171627044678,
+      "learning_rate": 1.854488778054863e-05,
+      "loss": 0.399,
+      "step": 5840
+    },
+    {
+      "epoch": 1.458852867830424,
+      "grad_norm": 6.84633207321167,
+      "learning_rate": 1.8542394014962594e-05,
+      "loss": 0.5866,
+      "step": 5850
+    },
+    {
+      "epoch": 1.4613466334164589,
+      "grad_norm": 5.431811332702637,
+      "learning_rate": 1.853990024937656e-05,
+      "loss": 0.4732,
+      "step": 5860
+    },
+    {
+      "epoch": 1.4638403990024937,
+      "grad_norm": 4.845829486846924,
+      "learning_rate": 1.8537406483790524e-05,
+      "loss": 0.4998,
+      "step": 5870
+    },
+    {
+      "epoch": 1.4663341645885286,
+      "grad_norm": 5.635951042175293,
+      "learning_rate": 1.8534912718204488e-05,
+      "loss": 0.3866,
+      "step": 5880
+    },
+    {
+      "epoch": 1.4688279301745637,
+      "grad_norm": 8.562129974365234,
+      "learning_rate": 1.8532418952618455e-05,
+      "loss": 0.452,
+      "step": 5890
+    },
+    {
+      "epoch": 1.4713216957605986,
+      "grad_norm": 5.379266738891602,
+      "learning_rate": 1.852992518703242e-05,
+      "loss": 0.4434,
+      "step": 5900
+    },
+    {
+      "epoch": 1.4738154613466334,
+      "grad_norm": 3.9054388999938965,
+      "learning_rate": 1.8527431421446386e-05,
+      "loss": 0.4882,
+      "step": 5910
+    },
+    {
+      "epoch": 1.4763092269326683,
+      "grad_norm": 7.7161383628845215,
+      "learning_rate": 1.8524937655860353e-05,
+      "loss": 0.4105,
+      "step": 5920
+    },
+    {
+      "epoch": 1.4788029925187032,
+      "grad_norm": 5.12331485748291,
+      "learning_rate": 1.8522443890274316e-05,
+      "loss": 0.4601,
+      "step": 5930
+    },
+    {
+      "epoch": 1.481296758104738,
+      "grad_norm": 7.548189640045166,
+      "learning_rate": 1.8519950124688283e-05,
+      "loss": 0.4507,
+      "step": 5940
+    },
+    {
+      "epoch": 1.4837905236907731,
+      "grad_norm": 3.955315589904785,
+      "learning_rate": 1.8517456359102247e-05,
+      "loss": 0.4234,
+      "step": 5950
+    },
+    {
+      "epoch": 1.486284289276808,
+      "grad_norm": 6.11161470413208,
+      "learning_rate": 1.851496259351621e-05,
+      "loss": 0.4254,
+      "step": 5960
+    },
+    {
+      "epoch": 1.4887780548628429,
+      "grad_norm": 8.937747955322266,
+      "learning_rate": 1.8512468827930178e-05,
+      "loss": 0.4604,
+      "step": 5970
+    },
+    {
+      "epoch": 1.4912718204488777,
+      "grad_norm": 13.255558967590332,
+      "learning_rate": 1.850997506234414e-05,
+      "loss": 0.4284,
+      "step": 5980
+    },
+    {
+      "epoch": 1.4937655860349128,
+      "grad_norm": 6.2703776359558105,
+      "learning_rate": 1.8507481296758105e-05,
+      "loss": 0.3886,
+      "step": 5990
+    },
+    {
+      "epoch": 1.4962593516209477,
+      "grad_norm": 6.531227111816406,
+      "learning_rate": 1.8504987531172072e-05,
+      "loss": 0.4601,
+      "step": 6000
+    },
+    {
+      "epoch": 1.4987531172069826,
+      "grad_norm": 5.363514423370361,
+      "learning_rate": 1.8502493765586035e-05,
+      "loss": 0.5012,
+      "step": 6010
+    },
+    {
+      "epoch": 1.5012468827930174,
+      "grad_norm": 3.8353939056396484,
+      "learning_rate": 1.8500000000000002e-05,
+      "loss": 0.4911,
+      "step": 6020
+    },
+    {
+      "epoch": 1.5037406483790523,
+      "grad_norm": 5.923316955566406,
+      "learning_rate": 1.8497506234413966e-05,
+      "loss": 0.3982,
+      "step": 6030
+    },
+    {
+      "epoch": 1.5062344139650872,
+      "grad_norm": 4.840806484222412,
+      "learning_rate": 1.849501246882793e-05,
+      "loss": 0.4035,
+      "step": 6040
+    },
+    {
+      "epoch": 1.508728179551122,
+      "grad_norm": 4.171060085296631,
+      "learning_rate": 1.8492518703241897e-05,
+      "loss": 0.515,
+      "step": 6050
+    },
+    {
+      "epoch": 1.5112219451371571,
+      "grad_norm": 7.2029876708984375,
+      "learning_rate": 1.849002493765586e-05,
+      "loss": 0.4485,
+      "step": 6060
+    },
+    {
+      "epoch": 1.513715710723192,
+      "grad_norm": 4.763010025024414,
+      "learning_rate": 1.8487531172069827e-05,
+      "loss": 0.3859,
+      "step": 6070
+    },
+    {
+      "epoch": 1.516209476309227,
+      "grad_norm": 4.51248836517334,
+      "learning_rate": 1.848503740648379e-05,
+      "loss": 0.4476,
+      "step": 6080
+    },
+    {
+      "epoch": 1.518703241895262,
+      "grad_norm": 5.466489791870117,
+      "learning_rate": 1.8482543640897758e-05,
+      "loss": 0.4742,
+      "step": 6090
+    },
+    {
+      "epoch": 1.5211970074812968,
+      "grad_norm": 6.138864994049072,
+      "learning_rate": 1.848004987531172e-05,
+      "loss": 0.4634,
+      "step": 6100
+    },
+    {
+      "epoch": 1.5236907730673317,
+      "grad_norm": 4.4593892097473145,
+      "learning_rate": 1.847755610972569e-05,
+      "loss": 0.4097,
+      "step": 6110
+    },
+    {
+      "epoch": 1.5261845386533666,
+      "grad_norm": 5.734845161437988,
+      "learning_rate": 1.8475062344139652e-05,
+      "loss": 0.4536,
+      "step": 6120
+    },
+    {
+      "epoch": 1.5286783042394014,
+      "grad_norm": 5.039837837219238,
+      "learning_rate": 1.847256857855362e-05,
+      "loss": 0.4227,
+      "step": 6130
+    },
+    {
+      "epoch": 1.5311720698254363,
+      "grad_norm": 5.665532112121582,
+      "learning_rate": 1.8470074812967583e-05,
+      "loss": 0.3996,
+      "step": 6140
+    },
+    {
+      "epoch": 1.5336658354114712,
+      "grad_norm": 8.553827285766602,
+      "learning_rate": 1.846758104738155e-05,
+      "loss": 0.5689,
+      "step": 6150
+    },
+    {
+      "epoch": 1.5361596009975063,
+      "grad_norm": 10.087984085083008,
+      "learning_rate": 1.8465087281795513e-05,
+      "loss": 0.3943,
+      "step": 6160
+    },
+    {
+      "epoch": 1.5386533665835411,
+      "grad_norm": 4.186636924743652,
+      "learning_rate": 1.8462593516209477e-05,
+      "loss": 0.3991,
+      "step": 6170
+    },
+    {
+      "epoch": 1.5411471321695762,
+      "grad_norm": 6.2279839515686035,
+      "learning_rate": 1.8460099750623444e-05,
+      "loss": 0.4065,
+      "step": 6180
+    },
+    {
+      "epoch": 1.543640897755611,
+      "grad_norm": 4.494476795196533,
+      "learning_rate": 1.8457605985037408e-05,
+      "loss": 0.3763,
+      "step": 6190
+    },
+    {
+      "epoch": 1.546134663341646,
+      "grad_norm": 4.94999885559082,
+      "learning_rate": 1.845511221945137e-05,
+      "loss": 0.4988,
+      "step": 6200
+    },
+    {
+      "epoch": 1.5486284289276808,
+      "grad_norm": 5.871788024902344,
+      "learning_rate": 1.8452618453865338e-05,
+      "loss": 0.4231,
+      "step": 6210
+    },
+    {
+      "epoch": 1.5511221945137157,
+      "grad_norm": 5.616420269012451,
+      "learning_rate": 1.8450124688279302e-05,
+      "loss": 0.372,
+      "step": 6220
+    },
+    {
+      "epoch": 1.5536159600997506,
+      "grad_norm": 4.9915971755981445,
+      "learning_rate": 1.844763092269327e-05,
+      "loss": 0.3971,
+      "step": 6230
+    },
+    {
+      "epoch": 1.5561097256857854,
+      "grad_norm": 4.821728229522705,
+      "learning_rate": 1.8445137157107232e-05,
+      "loss": 0.4182,
+      "step": 6240
+    },
+    {
+      "epoch": 1.5586034912718203,
+      "grad_norm": 8.948670387268066,
+      "learning_rate": 1.8442643391521196e-05,
+      "loss": 0.4545,
+      "step": 6250
+    },
+    {
+      "epoch": 1.5610972568578554,
+      "grad_norm": 6.1959228515625,
+      "learning_rate": 1.8440149625935163e-05,
+      "loss": 0.397,
+      "step": 6260
+    },
+    {
+      "epoch": 1.5635910224438903,
+      "grad_norm": 4.38476037979126,
+      "learning_rate": 1.843765586034913e-05,
+      "loss": 0.3934,
+      "step": 6270
+    },
+    {
+      "epoch": 1.5660847880299253,
+      "grad_norm": 6.135573387145996,
+      "learning_rate": 1.8435162094763094e-05,
+      "loss": 0.4019,
+      "step": 6280
+    },
+    {
+      "epoch": 1.5685785536159602,
+      "grad_norm": 3.752264976501465,
+      "learning_rate": 1.843266832917706e-05,
+      "loss": 0.4041,
+      "step": 6290
+    },
+    {
+      "epoch": 1.571072319201995,
+      "grad_norm": 6.69175386428833,
+      "learning_rate": 1.8430174563591024e-05,
+      "loss": 0.3357,
+      "step": 6300
+    },
+    {
+      "epoch": 1.57356608478803,
+      "grad_norm": 3.5837090015411377,
+      "learning_rate": 1.842768079800499e-05,
+      "loss": 0.4172,
+      "step": 6310
+    },
+    {
+      "epoch": 1.5760598503740648,
+      "grad_norm": 4.377597332000732,
+      "learning_rate": 1.8425187032418955e-05,
+      "loss": 0.48,
+      "step": 6320
+    },
+    {
+      "epoch": 1.5785536159600997,
+      "grad_norm": 6.266040325164795,
+      "learning_rate": 1.842269326683292e-05,
+      "loss": 0.5284,
+      "step": 6330
+    },
+    {
+      "epoch": 1.5810473815461346,
+      "grad_norm": 6.752136707305908,
+      "learning_rate": 1.8420199501246886e-05,
+      "loss": 0.4682,
+      "step": 6340
+    },
+    {
+      "epoch": 1.5835411471321694,
+      "grad_norm": 4.733704090118408,
+      "learning_rate": 1.841770573566085e-05,
+      "loss": 0.345,
+      "step": 6350
+    },
+    {
+      "epoch": 1.5860349127182045,
+      "grad_norm": 6.0275397300720215,
+      "learning_rate": 1.8415211970074816e-05,
+      "loss": 0.5216,
+      "step": 6360
+    },
+    {
+      "epoch": 1.5885286783042394,
+      "grad_norm": 4.750929832458496,
+      "learning_rate": 1.841271820448878e-05,
+      "loss": 0.4312,
+      "step": 6370
+    },
+    {
+      "epoch": 1.5910224438902745,
+      "grad_norm": 3.941704511642456,
+      "learning_rate": 1.8410224438902743e-05,
+      "loss": 0.4652,
+      "step": 6380
+    },
+    {
+      "epoch": 1.5935162094763093,
+      "grad_norm": 6.078151226043701,
+      "learning_rate": 1.840773067331671e-05,
+      "loss": 0.4478,
+      "step": 6390
+    },
+    {
+      "epoch": 1.5960099750623442,
+      "grad_norm": 5.187417030334473,
+      "learning_rate": 1.8405236907730674e-05,
+      "loss": 0.4094,
+      "step": 6400
+    },
+    {
+      "epoch": 1.598503740648379,
+      "grad_norm": 6.688471794128418,
+      "learning_rate": 1.8402743142144638e-05,
+      "loss": 0.4707,
+      "step": 6410
+    },
+    {
+      "epoch": 1.600997506234414,
+      "grad_norm": 4.863025188446045,
+      "learning_rate": 1.8400249376558605e-05,
+      "loss": 0.4326,
+      "step": 6420
+    },
+    {
+      "epoch": 1.6034912718204488,
+      "grad_norm": 6.477893352508545,
+      "learning_rate": 1.8397755610972568e-05,
+      "loss": 0.4071,
+      "step": 6430
+    },
+    {
+      "epoch": 1.6059850374064837,
+      "grad_norm": 3.6920061111450195,
+      "learning_rate": 1.8395261845386535e-05,
+      "loss": 0.4258,
+      "step": 6440
+    },
+    {
+      "epoch": 1.6084788029925186,
+      "grad_norm": 5.195054531097412,
+      "learning_rate": 1.83927680798005e-05,
+      "loss": 0.4388,
+      "step": 6450
+    },
+    {
+      "epoch": 1.6109725685785536,
+      "grad_norm": 6.680331707000732,
+      "learning_rate": 1.8390274314214466e-05,
+      "loss": 0.4053,
+      "step": 6460
+    },
+    {
+      "epoch": 1.6134663341645885,
+      "grad_norm": 4.663398742675781,
+      "learning_rate": 1.838778054862843e-05,
+      "loss": 0.457,
+      "step": 6470
+    },
+    {
+      "epoch": 1.6159600997506236,
+      "grad_norm": 4.989374160766602,
+      "learning_rate": 1.8385286783042397e-05,
+      "loss": 0.451,
+      "step": 6480
+    },
+    {
+      "epoch": 1.6184538653366585,
+      "grad_norm": 7.583066940307617,
+      "learning_rate": 1.838279301745636e-05,
+      "loss": 0.3729,
+      "step": 6490
+    },
+    {
+      "epoch": 1.6209476309226933,
+      "grad_norm": 4.013853073120117,
+      "learning_rate": 1.8380299251870327e-05,
+      "loss": 0.4288,
+      "step": 6500
+    },
+    {
+      "epoch": 1.6234413965087282,
+      "grad_norm": 4.080069065093994,
+      "learning_rate": 1.837780548628429e-05,
+      "loss": 0.3728,
+      "step": 6510
+    },
+    {
+      "epoch": 1.625935162094763,
+      "grad_norm": 6.781628131866455,
+      "learning_rate": 1.8375311720698258e-05,
+      "loss": 0.4198,
+      "step": 6520
+    },
+    {
+      "epoch": 1.628428927680798,
+      "grad_norm": 5.727022647857666,
+      "learning_rate": 1.837281795511222e-05,
+      "loss": 0.5096,
+      "step": 6530
+    },
+    {
+      "epoch": 1.6309226932668328,
+      "grad_norm": 4.232419967651367,
+      "learning_rate": 1.8370324189526185e-05,
+      "loss": 0.4087,
+      "step": 6540
+    },
+    {
+      "epoch": 1.6334164588528677,
+      "grad_norm": 6.57977294921875,
+      "learning_rate": 1.8367830423940152e-05,
+      "loss": 0.5512,
+      "step": 6550
+    },
+    {
+      "epoch": 1.6359102244389028,
+      "grad_norm": 5.6456427574157715,
+      "learning_rate": 1.8365336658354116e-05,
+      "loss": 0.448,
+      "step": 6560
+    },
+    {
+      "epoch": 1.6384039900249376,
+      "grad_norm": 5.072854518890381,
+      "learning_rate": 1.8362842892768083e-05,
+      "loss": 0.4374,
+      "step": 6570
+    },
+    {
+      "epoch": 1.6408977556109727,
+      "grad_norm": 6.125659465789795,
+      "learning_rate": 1.8360349127182046e-05,
+      "loss": 0.4616,
+      "step": 6580
+    },
+    {
+      "epoch": 1.6433915211970076,
+      "grad_norm": 5.4351301193237305,
+      "learning_rate": 1.835785536159601e-05,
+      "loss": 0.561,
+      "step": 6590
+    },
+    {
+      "epoch": 1.6458852867830425,
+      "grad_norm": 5.115979194641113,
+      "learning_rate": 1.8355361596009977e-05,
+      "loss": 0.4454,
+      "step": 6600
+    },
+    {
+      "epoch": 1.6483790523690773,
+      "grad_norm": 5.563187599182129,
+      "learning_rate": 1.835286783042394e-05,
+      "loss": 0.4437,
+      "step": 6610
+    },
+    {
+      "epoch": 1.6508728179551122,
+      "grad_norm": 8.810982704162598,
+      "learning_rate": 1.8350374064837907e-05,
+      "loss": 0.472,
+      "step": 6620
+    },
+    {
+      "epoch": 1.653366583541147,
+      "grad_norm": 6.700620174407959,
+      "learning_rate": 1.834788029925187e-05,
+      "loss": 0.4538,
+      "step": 6630
+    },
+    {
+      "epoch": 1.655860349127182,
+      "grad_norm": 5.371676921844482,
+      "learning_rate": 1.8345386533665838e-05,
+      "loss": 0.4628,
+      "step": 6640
+    },
+    {
+      "epoch": 1.6583541147132168,
+      "grad_norm": 6.457888126373291,
+      "learning_rate": 1.8342892768079802e-05,
+      "loss": 0.3982,
+      "step": 6650
+    },
+    {
+      "epoch": 1.660847880299252,
+      "grad_norm": 3.3349645137786865,
+      "learning_rate": 1.834039900249377e-05,
+      "loss": 0.5806,
+      "step": 6660
+    },
+    {
+      "epoch": 1.6633416458852868,
+      "grad_norm": 4.4449334144592285,
+      "learning_rate": 1.8337905236907732e-05,
+      "loss": 0.3955,
+      "step": 6670
+    },
+    {
+      "epoch": 1.6658354114713219,
+      "grad_norm": 6.8672776222229,
+      "learning_rate": 1.83354114713217e-05,
+      "loss": 0.4466,
+      "step": 6680
+    },
+    {
+      "epoch": 1.6683291770573567,
+      "grad_norm": 8.786575317382812,
+      "learning_rate": 1.8332917705735663e-05,
+      "loss": 0.4113,
+      "step": 6690
+    },
+    {
+      "epoch": 1.6708229426433916,
+      "grad_norm": 5.177858352661133,
+      "learning_rate": 1.8330423940149627e-05,
+      "loss": 0.4509,
+      "step": 6700
+    },
+    {
+      "epoch": 1.6733167082294265,
+      "grad_norm": 8.495447158813477,
+      "learning_rate": 1.8327930174563594e-05,
+      "loss": 0.4024,
+      "step": 6710
+    },
+    {
+      "epoch": 1.6758104738154613,
+      "grad_norm": 3.8897151947021484,
+      "learning_rate": 1.8325436408977557e-05,
+      "loss": 0.4081,
+      "step": 6720
+    },
+    {
+      "epoch": 1.6783042394014962,
+      "grad_norm": 7.075688362121582,
+      "learning_rate": 1.8322942643391524e-05,
+      "loss": 0.4058,
+      "step": 6730
+    },
+    {
+      "epoch": 1.680798004987531,
+      "grad_norm": 5.916785717010498,
+      "learning_rate": 1.8320448877805488e-05,
+      "loss": 0.4553,
+      "step": 6740
+    },
+    {
+      "epoch": 1.683291770573566,
+      "grad_norm": 6.220627307891846,
+      "learning_rate": 1.831795511221945e-05,
+      "loss": 0.3365,
+      "step": 6750
+    },
+    {
+      "epoch": 1.685785536159601,
+      "grad_norm": 5.613668441772461,
+      "learning_rate": 1.831546134663342e-05,
+      "loss": 0.4703,
+      "step": 6760
+    },
+    {
+      "epoch": 1.688279301745636,
+      "grad_norm": 5.4366841316223145,
+      "learning_rate": 1.8312967581047382e-05,
+      "loss": 0.4233,
+      "step": 6770
+    },
+    {
+      "epoch": 1.690773067331671,
+      "grad_norm": 7.235218524932861,
+      "learning_rate": 1.8310473815461346e-05,
+      "loss": 0.4584,
+      "step": 6780
+    },
+    {
+      "epoch": 1.6932668329177059,
+      "grad_norm": 5.918466567993164,
+      "learning_rate": 1.8307980049875313e-05,
+      "loss": 0.4717,
+      "step": 6790
+    },
+    {
+      "epoch": 1.6957605985037407,
+      "grad_norm": 7.333720684051514,
+      "learning_rate": 1.8305486284289276e-05,
+      "loss": 0.4562,
+      "step": 6800
+    },
+    {
+      "epoch": 1.6982543640897756,
+      "grad_norm": 4.598586082458496,
+      "learning_rate": 1.8302992518703243e-05,
+      "loss": 0.3797,
+      "step": 6810
+    },
+    {
+      "epoch": 1.7007481296758105,
+      "grad_norm": 6.30976676940918,
+      "learning_rate": 1.8300498753117207e-05,
+      "loss": 0.5182,
+      "step": 6820
+    },
+    {
+      "epoch": 1.7032418952618453,
+      "grad_norm": 5.384555339813232,
+      "learning_rate": 1.8298004987531174e-05,
+      "loss": 0.3699,
+      "step": 6830
+    },
+    {
+      "epoch": 1.7057356608478802,
+      "grad_norm": 6.011791229248047,
+      "learning_rate": 1.8295511221945138e-05,
+      "loss": 0.4675,
+      "step": 6840
+    },
+    {
+      "epoch": 1.708229426433915,
+      "grad_norm": 5.3825883865356445,
+      "learning_rate": 1.8293017456359105e-05,
+      "loss": 0.4217,
+      "step": 6850
+    },
+    {
+      "epoch": 1.7107231920199502,
+      "grad_norm": 5.053082466125488,
+      "learning_rate": 1.829052369077307e-05,
+      "loss": 0.5013,
+      "step": 6860
+    },
+    {
+      "epoch": 1.713216957605985,
+      "grad_norm": 3.9529480934143066,
+      "learning_rate": 1.8288029925187035e-05,
+      "loss": 0.5152,
+      "step": 6870
+    },
+    {
+      "epoch": 1.7157107231920201,
+      "grad_norm": 7.375166416168213,
+      "learning_rate": 1.8285536159601e-05,
+      "loss": 0.3735,
+      "step": 6880
+    },
+    {
+      "epoch": 1.718204488778055,
+      "grad_norm": 6.015711307525635,
+      "learning_rate": 1.8283042394014966e-05,
+      "loss": 0.3918,
+      "step": 6890
+    },
+    {
+      "epoch": 1.7206982543640899,
+      "grad_norm": 4.770452499389648,
+      "learning_rate": 1.828054862842893e-05,
+      "loss": 0.4266,
+      "step": 6900
+    },
+    {
+      "epoch": 1.7231920199501247,
+      "grad_norm": 4.618214130401611,
+      "learning_rate": 1.8278054862842893e-05,
+      "loss": 0.4717,
+      "step": 6910
+    },
+    {
+      "epoch": 1.7256857855361596,
+      "grad_norm": 4.1790924072265625,
+      "learning_rate": 1.827556109725686e-05,
+      "loss": 0.4128,
+      "step": 6920
+    },
+    {
+      "epoch": 1.7281795511221945,
+      "grad_norm": 6.608766078948975,
+      "learning_rate": 1.8273067331670824e-05,
+      "loss": 0.4591,
+      "step": 6930
+    },
+    {
+      "epoch": 1.7306733167082293,
+      "grad_norm": 8.431743621826172,
+      "learning_rate": 1.827057356608479e-05,
+      "loss": 0.5575,
+      "step": 6940
+    },
+    {
+      "epoch": 1.7331670822942642,
+      "grad_norm": 5.013129711151123,
+      "learning_rate": 1.8268079800498754e-05,
+      "loss": 0.3743,
+      "step": 6950
+    },
+    {
+      "epoch": 1.7356608478802993,
+      "grad_norm": 4.251693248748779,
+      "learning_rate": 1.8265586034912718e-05,
+      "loss": 0.4133,
+      "step": 6960
+    },
+    {
+      "epoch": 1.7381546134663342,
+      "grad_norm": 4.549345970153809,
+      "learning_rate": 1.8263092269326685e-05,
+      "loss": 0.5037,
+      "step": 6970
+    },
+    {
+      "epoch": 1.7406483790523692,
+      "grad_norm": 6.431280612945557,
+      "learning_rate": 1.826059850374065e-05,
+      "loss": 0.3989,
+      "step": 6980
+    },
+    {
+      "epoch": 1.7431421446384041,
+      "grad_norm": 5.792124271392822,
+      "learning_rate": 1.8258104738154615e-05,
+      "loss": 0.4512,
+      "step": 6990
+    },
+    {
+      "epoch": 1.745635910224439,
+      "grad_norm": 8.254396438598633,
+      "learning_rate": 1.825561097256858e-05,
+      "loss": 0.4982,
+      "step": 7000
+    },
+    {
+      "epoch": 1.7481296758104738,
+      "grad_norm": 4.77825927734375,
+      "learning_rate": 1.8253117206982546e-05,
+      "loss": 0.4053,
+      "step": 7010
+    },
+    {
+      "epoch": 1.7506234413965087,
+      "grad_norm": 6.53203821182251,
+      "learning_rate": 1.825062344139651e-05,
+      "loss": 0.4921,
+      "step": 7020
+    },
+    {
+      "epoch": 1.7531172069825436,
+      "grad_norm": 4.8709940910339355,
+      "learning_rate": 1.8248129675810477e-05,
+      "loss": 0.4883,
+      "step": 7030
+    },
+    {
+      "epoch": 1.7556109725685785,
+      "grad_norm": 4.446110725402832,
+      "learning_rate": 1.824563591022444e-05,
+      "loss": 0.4667,
+      "step": 7040
+    },
+    {
+      "epoch": 1.7581047381546133,
+      "grad_norm": 6.364370346069336,
+      "learning_rate": 1.8243142144638407e-05,
+      "loss": 0.5304,
+      "step": 7050
+    },
+    {
+      "epoch": 1.7605985037406484,
+      "grad_norm": 5.594391822814941,
+      "learning_rate": 1.824064837905237e-05,
+      "loss": 0.4105,
+      "step": 7060
+    },
+    {
+      "epoch": 1.7630922693266833,
+      "grad_norm": 4.667792320251465,
+      "learning_rate": 1.8238154613466338e-05,
+      "loss": 0.445,
+      "step": 7070
+    },
+    {
+      "epoch": 1.7655860349127181,
+      "grad_norm": 30.49921417236328,
+      "learning_rate": 1.82356608478803e-05,
+      "loss": 0.5001,
+      "step": 7080
+    },
+    {
+      "epoch": 1.7680798004987532,
+      "grad_norm": 4.282893657684326,
+      "learning_rate": 1.8233167082294265e-05,
+      "loss": 0.428,
+      "step": 7090
+    },
+    {
+      "epoch": 1.770573566084788,
+      "grad_norm": 6.280821323394775,
+      "learning_rate": 1.8230673316708232e-05,
+      "loss": 0.4246,
+      "step": 7100
+    },
+    {
+      "epoch": 1.773067331670823,
+      "grad_norm": 5.7564005851745605,
+      "learning_rate": 1.8228179551122196e-05,
+      "loss": 0.5076,
+      "step": 7110
+    },
+    {
+      "epoch": 1.7755610972568578,
+      "grad_norm": 5.439352035522461,
+      "learning_rate": 1.822568578553616e-05,
+      "loss": 0.448,
+      "step": 7120
+    },
+    {
+      "epoch": 1.7780548628428927,
+      "grad_norm": 4.840354919433594,
+      "learning_rate": 1.8223192019950126e-05,
+      "loss": 0.4518,
+      "step": 7130
+    },
+    {
+      "epoch": 1.7805486284289276,
+      "grad_norm": 4.0967631340026855,
+      "learning_rate": 1.822069825436409e-05,
+      "loss": 0.441,
+      "step": 7140
+    },
+    {
+      "epoch": 1.7830423940149625,
+      "grad_norm": 5.563769340515137,
+      "learning_rate": 1.8218204488778057e-05,
+      "loss": 0.5182,
+      "step": 7150
+    },
+    {
+      "epoch": 1.7855361596009975,
+      "grad_norm": 6.383294582366943,
+      "learning_rate": 1.821571072319202e-05,
+      "loss": 0.4211,
+      "step": 7160
+    },
+    {
+      "epoch": 1.7880299251870324,
+      "grad_norm": 4.494612216949463,
+      "learning_rate": 1.8213216957605984e-05,
+      "loss": 0.426,
+      "step": 7170
+    },
+    {
+      "epoch": 1.7905236907730673,
+      "grad_norm": 4.303411483764648,
+      "learning_rate": 1.821072319201995e-05,
+      "loss": 0.3868,
+      "step": 7180
+    },
+    {
+      "epoch": 1.7930174563591024,
+      "grad_norm": 7.266903400421143,
+      "learning_rate": 1.8208229426433915e-05,
+      "loss": 0.4369,
+      "step": 7190
+    },
+    {
+      "epoch": 1.7955112219451372,
+      "grad_norm": 4.56803560256958,
+      "learning_rate": 1.8205735660847882e-05,
+      "loss": 0.4737,
+      "step": 7200
+    },
+    {
+      "epoch": 1.798004987531172,
+      "grad_norm": 5.030318260192871,
+      "learning_rate": 1.820324189526185e-05,
+      "loss": 0.5418,
+      "step": 7210
+    },
+    {
+      "epoch": 1.800498753117207,
+      "grad_norm": 5.952413558959961,
+      "learning_rate": 1.8200748129675813e-05,
+      "loss": 0.3784,
+      "step": 7220
+    },
+    {
+      "epoch": 1.8029925187032418,
+      "grad_norm": 13.113987922668457,
+      "learning_rate": 1.819825436408978e-05,
+      "loss": 0.4306,
+      "step": 7230
+    },
+    {
+      "epoch": 1.8054862842892767,
+      "grad_norm": 4.897136688232422,
+      "learning_rate": 1.8195760598503743e-05,
+      "loss": 0.4035,
+      "step": 7240
+    },
+    {
+      "epoch": 1.8079800498753116,
+      "grad_norm": 4.879326343536377,
+      "learning_rate": 1.8193266832917707e-05,
+      "loss": 0.4381,
+      "step": 7250
+    },
+    {
+      "epoch": 1.8104738154613467,
+      "grad_norm": 4.985284328460693,
+      "learning_rate": 1.8190773067331674e-05,
+      "loss": 0.4622,
+      "step": 7260
+    },
+    {
+      "epoch": 1.8129675810473815,
+      "grad_norm": 6.7815775871276855,
+      "learning_rate": 1.8188279301745637e-05,
+      "loss": 0.5456,
+      "step": 7270
+    },
+    {
+      "epoch": 1.8154613466334164,
+      "grad_norm": 7.435234069824219,
+      "learning_rate": 1.81857855361596e-05,
+      "loss": 0.4235,
+      "step": 7280
+    },
+    {
+      "epoch": 1.8179551122194515,
+      "grad_norm": 4.0493483543396,
+      "learning_rate": 1.8183291770573568e-05,
+      "loss": 0.4922,
+      "step": 7290
+    },
+    {
+      "epoch": 1.8204488778054864,
+      "grad_norm": 7.476971626281738,
+      "learning_rate": 1.818079800498753e-05,
+      "loss": 0.3972,
+      "step": 7300
+    },
+    {
+      "epoch": 1.8229426433915212,
+      "grad_norm": 5.938323497772217,
+      "learning_rate": 1.81783042394015e-05,
+      "loss": 0.4129,
+      "step": 7310
+    },
+    {
+      "epoch": 1.825436408977556,
+      "grad_norm": 6.04447078704834,
+      "learning_rate": 1.8175810473815462e-05,
+      "loss": 0.4718,
+      "step": 7320
+    },
+    {
+      "epoch": 1.827930174563591,
+      "grad_norm": 11.774287223815918,
+      "learning_rate": 1.8173316708229426e-05,
+      "loss": 0.4558,
+      "step": 7330
+    },
+    {
+      "epoch": 1.8304239401496258,
+      "grad_norm": 3.780583381652832,
+      "learning_rate": 1.8170822942643393e-05,
+      "loss": 0.4411,
+      "step": 7340
+    },
+    {
+      "epoch": 1.8329177057356607,
+      "grad_norm": 3.6571693420410156,
+      "learning_rate": 1.8168329177057356e-05,
+      "loss": 0.4096,
+      "step": 7350
+    },
+    {
+      "epoch": 1.8354114713216958,
+      "grad_norm": 7.564289093017578,
+      "learning_rate": 1.8165835411471323e-05,
+      "loss": 0.3931,
+      "step": 7360
+    },
+    {
+      "epoch": 1.8379052369077307,
+      "grad_norm": 7.691397666931152,
+      "learning_rate": 1.8163341645885287e-05,
+      "loss": 0.4236,
+      "step": 7370
+    },
+    {
+      "epoch": 1.8403990024937655,
+      "grad_norm": 5.345376968383789,
+      "learning_rate": 1.8160847880299254e-05,
+      "loss": 0.4377,
+      "step": 7380
+    },
+    {
+      "epoch": 1.8428927680798006,
+      "grad_norm": 11.432061195373535,
+      "learning_rate": 1.8158354114713218e-05,
+      "loss": 0.4218,
+      "step": 7390
+    },
+    {
+      "epoch": 1.8453865336658355,
+      "grad_norm": 4.74954080581665,
+      "learning_rate": 1.8155860349127185e-05,
+      "loss": 0.4077,
+      "step": 7400
+    },
+    {
+      "epoch": 1.8478802992518704,
+      "grad_norm": 4.802658557891846,
+      "learning_rate": 1.815336658354115e-05,
+      "loss": 0.4862,
+      "step": 7410
+    },
+    {
+      "epoch": 1.8503740648379052,
+      "grad_norm": 5.43306303024292,
+      "learning_rate": 1.8150872817955115e-05,
+      "loss": 0.4187,
+      "step": 7420
+    },
+    {
+      "epoch": 1.85286783042394,
+      "grad_norm": 3.525193929672241,
+      "learning_rate": 1.814837905236908e-05,
+      "loss": 0.3819,
+      "step": 7430
+    },
+    {
+      "epoch": 1.855361596009975,
+      "grad_norm": 6.635593891143799,
+      "learning_rate": 1.8145885286783046e-05,
+      "loss": 0.4183,
+      "step": 7440
+    },
+    {
+      "epoch": 1.8578553615960098,
+      "grad_norm": 6.29930305480957,
+      "learning_rate": 1.814339152119701e-05,
+      "loss": 0.5308,
+      "step": 7450
+    },
+    {
+      "epoch": 1.860349127182045,
+      "grad_norm": 5.674611568450928,
+      "learning_rate": 1.8140897755610973e-05,
+      "loss": 0.4003,
+      "step": 7460
+    },
+    {
+      "epoch": 1.8628428927680798,
+      "grad_norm": 9.636392593383789,
+      "learning_rate": 1.813840399002494e-05,
+      "loss": 0.4318,
+      "step": 7470
+    },
+    {
+      "epoch": 1.8653366583541147,
+      "grad_norm": 4.685451030731201,
+      "learning_rate": 1.8135910224438904e-05,
+      "loss": 0.4437,
+      "step": 7480
+    },
+    {
+      "epoch": 1.8678304239401498,
+      "grad_norm": 8.09288501739502,
+      "learning_rate": 1.8133416458852867e-05,
+      "loss": 0.5241,
+      "step": 7490
+    },
+    {
+      "epoch": 1.8703241895261846,
+      "grad_norm": 4.083009243011475,
+      "learning_rate": 1.8130922693266834e-05,
+      "loss": 0.3948,
+      "step": 7500
+    },
+    {
+      "epoch": 1.8728179551122195,
+      "grad_norm": 6.581864833831787,
+      "learning_rate": 1.8128428927680798e-05,
+      "loss": 0.5274,
+      "step": 7510
+    },
+    {
+      "epoch": 1.8753117206982544,
+      "grad_norm": 5.297054767608643,
+      "learning_rate": 1.8125935162094765e-05,
+      "loss": 0.4802,
+      "step": 7520
+    },
+    {
+      "epoch": 1.8778054862842892,
+      "grad_norm": 5.629668235778809,
+      "learning_rate": 1.812344139650873e-05,
+      "loss": 0.4183,
+      "step": 7530
+    },
+    {
+      "epoch": 1.880299251870324,
+      "grad_norm": 5.578089237213135,
+      "learning_rate": 1.8120947630922692e-05,
+      "loss": 0.4061,
+      "step": 7540
+    },
+    {
+      "epoch": 1.882793017456359,
+      "grad_norm": 6.019644260406494,
+      "learning_rate": 1.811845386533666e-05,
+      "loss": 0.4339,
+      "step": 7550
+    },
+    {
+      "epoch": 1.885286783042394,
+      "grad_norm": 6.0486016273498535,
+      "learning_rate": 1.8115960099750626e-05,
+      "loss": 0.3962,
+      "step": 7560
+    },
+    {
+      "epoch": 1.887780548628429,
+      "grad_norm": 12.667963027954102,
+      "learning_rate": 1.811346633416459e-05,
+      "loss": 0.4148,
+      "step": 7570
+    },
+    {
+      "epoch": 1.8902743142144638,
+      "grad_norm": 5.244442462921143,
+      "learning_rate": 1.8110972568578557e-05,
+      "loss": 0.4431,
+      "step": 7580
+    },
+    {
+      "epoch": 1.8927680798004989,
+      "grad_norm": 5.28915548324585,
+      "learning_rate": 1.810847880299252e-05,
+      "loss": 0.4288,
+      "step": 7590
+    },
+    {
+      "epoch": 1.8952618453865338,
+      "grad_norm": 5.8130717277526855,
+      "learning_rate": 1.8105985037406487e-05,
+      "loss": 0.4385,
+      "step": 7600
+    },
+    {
+      "epoch": 1.8977556109725686,
+      "grad_norm": 9.284071922302246,
+      "learning_rate": 1.810349127182045e-05,
+      "loss": 0.4476,
+      "step": 7610
+    },
+    {
+      "epoch": 1.9002493765586035,
+      "grad_norm": 4.337800979614258,
+      "learning_rate": 1.8100997506234415e-05,
+      "loss": 0.4499,
+      "step": 7620
+    },
+    {
+      "epoch": 1.9027431421446384,
+      "grad_norm": 5.436014652252197,
+      "learning_rate": 1.8098503740648382e-05,
+      "loss": 0.4046,
+      "step": 7630
+    },
+    {
+      "epoch": 1.9052369077306732,
+      "grad_norm": 5.804083347320557,
+      "learning_rate": 1.8096009975062345e-05,
+      "loss": 0.4042,
+      "step": 7640
+    },
+    {
+      "epoch": 1.907730673316708,
+      "grad_norm": 5.294909477233887,
+      "learning_rate": 1.8093516209476312e-05,
+      "loss": 0.4618,
+      "step": 7650
+    },
+    {
+      "epoch": 1.9102244389027432,
+      "grad_norm": 5.705154895782471,
+      "learning_rate": 1.8091022443890276e-05,
+      "loss": 0.3835,
+      "step": 7660
+    },
+    {
+      "epoch": 1.912718204488778,
+      "grad_norm": 7.614785194396973,
+      "learning_rate": 1.808852867830424e-05,
+      "loss": 0.4388,
+      "step": 7670
+    },
+    {
+      "epoch": 1.915211970074813,
+      "grad_norm": 9.722792625427246,
+      "learning_rate": 1.8086034912718207e-05,
+      "loss": 0.4568,
+      "step": 7680
+    },
+    {
+      "epoch": 1.917705735660848,
+      "grad_norm": 5.578179836273193,
+      "learning_rate": 1.808354114713217e-05,
+      "loss": 0.3962,
+      "step": 7690
+    },
+    {
+      "epoch": 1.9201995012468829,
+      "grad_norm": 5.2010178565979,
+      "learning_rate": 1.8081047381546134e-05,
+      "loss": 0.4123,
+      "step": 7700
+    },
+    {
+      "epoch": 1.9226932668329177,
+      "grad_norm": 4.397924423217773,
+      "learning_rate": 1.80785536159601e-05,
+      "loss": 0.4321,
+      "step": 7710
+    },
+    {
+      "epoch": 1.9251870324189526,
+      "grad_norm": 4.920018196105957,
+      "learning_rate": 1.8076059850374064e-05,
+      "loss": 0.4054,
+      "step": 7720
+    },
+    {
+      "epoch": 1.9276807980049875,
+      "grad_norm": 3.409426212310791,
+      "learning_rate": 1.807356608478803e-05,
+      "loss": 0.4464,
+      "step": 7730
+    },
+    {
+      "epoch": 1.9301745635910224,
+      "grad_norm": 7.432455062866211,
+      "learning_rate": 1.8071072319201995e-05,
+      "loss": 0.4521,
+      "step": 7740
+    },
+    {
+      "epoch": 1.9326683291770572,
+      "grad_norm": 4.5080485343933105,
+      "learning_rate": 1.8068578553615962e-05,
+      "loss": 0.4328,
+      "step": 7750
+    },
+    {
+      "epoch": 1.9351620947630923,
+      "grad_norm": 5.043903827667236,
+      "learning_rate": 1.8066084788029926e-05,
+      "loss": 0.4508,
+      "step": 7760
+    },
+    {
+      "epoch": 1.9376558603491272,
+      "grad_norm": 6.6283369064331055,
+      "learning_rate": 1.8063591022443893e-05,
+      "loss": 0.464,
+      "step": 7770
+    },
+    {
+      "epoch": 1.940149625935162,
+      "grad_norm": 4.977077007293701,
+      "learning_rate": 1.8061097256857856e-05,
+      "loss": 0.5318,
+      "step": 7780
+    },
+    {
+      "epoch": 1.9426433915211971,
+      "grad_norm": 4.694863796234131,
+      "learning_rate": 1.8058603491271823e-05,
+      "loss": 0.3441,
+      "step": 7790
+    },
+    {
+      "epoch": 1.945137157107232,
+      "grad_norm": 5.31052303314209,
+      "learning_rate": 1.8056109725685787e-05,
+      "loss": 0.4264,
+      "step": 7800
+    },
+    {
+      "epoch": 1.9476309226932669,
+      "grad_norm": 5.178094863891602,
+      "learning_rate": 1.8053615960099754e-05,
+      "loss": 0.3537,
+      "step": 7810
+    },
+    {
+      "epoch": 1.9501246882793017,
+      "grad_norm": 11.527572631835938,
+      "learning_rate": 1.8051122194513718e-05,
+      "loss": 0.5116,
+      "step": 7820
+    },
+    {
+      "epoch": 1.9526184538653366,
+      "grad_norm": 5.841403961181641,
+      "learning_rate": 1.804862842892768e-05,
+      "loss": 0.4658,
+      "step": 7830
+    },
+    {
+      "epoch": 1.9551122194513715,
+      "grad_norm": 8.75449275970459,
+      "learning_rate": 1.8046134663341648e-05,
+      "loss": 0.449,
+      "step": 7840
+    },
+    {
+      "epoch": 1.9576059850374063,
+      "grad_norm": 7.627124309539795,
+      "learning_rate": 1.8043640897755612e-05,
+      "loss": 0.4512,
+      "step": 7850
+    },
+    {
+      "epoch": 1.9600997506234414,
+      "grad_norm": 7.299831867218018,
+      "learning_rate": 1.804114713216958e-05,
+      "loss": 0.4311,
+      "step": 7860
+    },
+    {
+      "epoch": 1.9625935162094763,
+      "grad_norm": 5.300114154815674,
+      "learning_rate": 1.8038653366583542e-05,
+      "loss": 0.3713,
+      "step": 7870
+    },
+    {
+      "epoch": 1.9650872817955112,
+      "grad_norm": 5.638365745544434,
+      "learning_rate": 1.8036159600997506e-05,
+      "loss": 0.5088,
+      "step": 7880
+    },
+    {
+      "epoch": 1.9675810473815463,
+      "grad_norm": 6.054098606109619,
+      "learning_rate": 1.8033665835411473e-05,
+      "loss": 0.4084,
+      "step": 7890
+    },
+    {
+      "epoch": 1.9700748129675811,
+      "grad_norm": 7.920881271362305,
+      "learning_rate": 1.8031172069825437e-05,
+      "loss": 0.4036,
+      "step": 7900
+    },
+    {
+      "epoch": 1.972568578553616,
+      "grad_norm": 5.523858070373535,
+      "learning_rate": 1.8028678304239404e-05,
+      "loss": 0.4152,
+      "step": 7910
+    },
+    {
+      "epoch": 1.9750623441396509,
+      "grad_norm": 7.041183948516846,
+      "learning_rate": 1.8026184538653367e-05,
+      "loss": 0.314,
+      "step": 7920
+    },
+    {
+      "epoch": 1.9775561097256857,
+      "grad_norm": 5.018404960632324,
+      "learning_rate": 1.8023690773067334e-05,
+      "loss": 0.4854,
+      "step": 7930
+    },
+    {
+      "epoch": 1.9800498753117206,
+      "grad_norm": 6.31093168258667,
+      "learning_rate": 1.8021446384039903e-05,
+      "loss": 0.4654,
+      "step": 7940
+    },
+    {
+      "epoch": 1.9825436408977555,
+      "grad_norm": 6.725535869598389,
+      "learning_rate": 1.8018952618453866e-05,
+      "loss": 0.3975,
+      "step": 7950
+    },
+    {
+      "epoch": 1.9850374064837906,
+      "grad_norm": 7.007317066192627,
+      "learning_rate": 1.801645885286783e-05,
+      "loss": 0.4361,
+      "step": 7960
+    },
+    {
+      "epoch": 1.9875311720698254,
+      "grad_norm": 4.911192893981934,
+      "learning_rate": 1.8013965087281797e-05,
+      "loss": 0.4742,
+      "step": 7970
+    },
+    {
+      "epoch": 1.9900249376558603,
+      "grad_norm": 6.7747955322265625,
+      "learning_rate": 1.801147132169576e-05,
+      "loss": 0.4091,
+      "step": 7980
+    },
+    {
+      "epoch": 1.9925187032418954,
+      "grad_norm": 5.228945732116699,
+      "learning_rate": 1.8008977556109727e-05,
+      "loss": 0.4404,
+      "step": 7990
+    },
+    {
+      "epoch": 1.9950124688279303,
+      "grad_norm": 6.581811428070068,
+      "learning_rate": 1.800648379052369e-05,
+      "loss": 0.4312,
+      "step": 8000
+    },
+    {
+      "epoch": 1.9975062344139651,
+      "grad_norm": 5.50934362411499,
+      "learning_rate": 1.8003990024937658e-05,
+      "loss": 0.3873,
+      "step": 8010
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.261854648590088,
+      "learning_rate": 1.800149625935162e-05,
+      "loss": 0.4136,
+      "step": 8020
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.4465249180793762,
+      "eval_runtime": 59.866,
+      "eval_samples_per_second": 16.754,
+      "eval_steps_per_second": 16.754,
+      "step": 8020
+    },
+    {
+      "epoch": 2.002493765586035,
+      "grad_norm": 3.6723527908325195,
+      "learning_rate": 1.799900249376559e-05,
+      "loss": 0.3571,
+      "step": 8030
+    },
+    {
+      "epoch": 2.0049875311720697,
+      "grad_norm": 5.929795265197754,
+      "learning_rate": 1.7996508728179552e-05,
+      "loss": 0.5315,
+      "step": 8040
+    },
+    {
+      "epoch": 2.0074812967581046,
+      "grad_norm": 6.144455909729004,
+      "learning_rate": 1.799401496259352e-05,
+      "loss": 0.4404,
+      "step": 8050
+    },
+    {
+      "epoch": 2.0099750623441395,
+      "grad_norm": 6.373076438903809,
+      "learning_rate": 1.7991521197007483e-05,
+      "loss": 0.374,
+      "step": 8060
+    },
+    {
+      "epoch": 2.0124688279301743,
+      "grad_norm": 3.5335922241210938,
+      "learning_rate": 1.798902743142145e-05,
+      "loss": 0.3972,
+      "step": 8070
+    },
+    {
+      "epoch": 2.0149625935162097,
+      "grad_norm": 6.1185479164123535,
+      "learning_rate": 1.7986533665835413e-05,
+      "loss": 0.4131,
+      "step": 8080
+    },
+    {
+      "epoch": 2.0174563591022445,
+      "grad_norm": 5.0511345863342285,
+      "learning_rate": 1.7984039900249377e-05,
+      "loss": 0.4324,
+      "step": 8090
+    },
+    {
+      "epoch": 2.0199501246882794,
+      "grad_norm": 7.270425796508789,
+      "learning_rate": 1.7981546134663344e-05,
+      "loss": 0.3787,
+      "step": 8100
+    },
+    {
+      "epoch": 2.0224438902743143,
+      "grad_norm": 4.606263637542725,
+      "learning_rate": 1.7979052369077308e-05,
+      "loss": 0.3524,
+      "step": 8110
+    },
+    {
+      "epoch": 2.024937655860349,
+      "grad_norm": 5.3458452224731445,
+      "learning_rate": 1.7976558603491275e-05,
+      "loss": 0.4371,
+      "step": 8120
+    },
+    {
+      "epoch": 2.027431421446384,
+      "grad_norm": 4.736631870269775,
+      "learning_rate": 1.797406483790524e-05,
+      "loss": 0.3898,
+      "step": 8130
+    },
+    {
+      "epoch": 2.029925187032419,
+      "grad_norm": 5.234148025512695,
+      "learning_rate": 1.7971571072319202e-05,
+      "loss": 0.4307,
+      "step": 8140
+    },
+    {
+      "epoch": 2.0324189526184537,
+      "grad_norm": 6.201371669769287,
+      "learning_rate": 1.796907730673317e-05,
+      "loss": 0.4022,
+      "step": 8150
+    },
+    {
+      "epoch": 2.0349127182044886,
+      "grad_norm": 5.895164489746094,
+      "learning_rate": 1.7966583541147133e-05,
+      "loss": 0.4254,
+      "step": 8160
+    },
+    {
+      "epoch": 2.037406483790524,
+      "grad_norm": 4.677857398986816,
+      "learning_rate": 1.7964089775561096e-05,
+      "loss": 0.5048,
+      "step": 8170
+    },
+    {
+      "epoch": 2.039900249376559,
+      "grad_norm": 4.634955406188965,
+      "learning_rate": 1.7961596009975063e-05,
+      "loss": 0.4376,
+      "step": 8180
+    },
+    {
+      "epoch": 2.0423940149625937,
+      "grad_norm": 5.4965667724609375,
+      "learning_rate": 1.7959102244389027e-05,
+      "loss": 0.3972,
+      "step": 8190
+    },
+    {
+      "epoch": 2.0448877805486285,
+      "grad_norm": 5.892706394195557,
+      "learning_rate": 1.7956608478802994e-05,
+      "loss": 0.3795,
+      "step": 8200
+    },
+    {
+      "epoch": 2.0473815461346634,
+      "grad_norm": 6.747429370880127,
+      "learning_rate": 1.7954114713216957e-05,
+      "loss": 0.394,
+      "step": 8210
+    },
+    {
+      "epoch": 2.0498753117206983,
+      "grad_norm": 6.291106224060059,
+      "learning_rate": 1.7951620947630924e-05,
+      "loss": 0.4976,
+      "step": 8220
+    },
+    {
+      "epoch": 2.052369077306733,
+      "grad_norm": 5.840002536773682,
+      "learning_rate": 1.794912718204489e-05,
+      "loss": 0.4118,
+      "step": 8230
+    },
+    {
+      "epoch": 2.054862842892768,
+      "grad_norm": 9.487309455871582,
+      "learning_rate": 1.7946633416458855e-05,
+      "loss": 0.3606,
+      "step": 8240
+    },
+    {
+      "epoch": 2.057356608478803,
+      "grad_norm": 5.123715877532959,
+      "learning_rate": 1.7944139650872822e-05,
+      "loss": 0.4149,
+      "step": 8250
+    },
+    {
+      "epoch": 2.0598503740648377,
+      "grad_norm": 5.808225154876709,
+      "learning_rate": 1.7941645885286786e-05,
+      "loss": 0.4553,
+      "step": 8260
+    },
+    {
+      "epoch": 2.0623441396508726,
+      "grad_norm": 9.865265846252441,
+      "learning_rate": 1.793915211970075e-05,
+      "loss": 0.4352,
+      "step": 8270
+    },
+    {
+      "epoch": 2.064837905236908,
+      "grad_norm": 4.490272045135498,
+      "learning_rate": 1.7936658354114716e-05,
+      "loss": 0.4435,
+      "step": 8280
+    },
+    {
+      "epoch": 2.067331670822943,
+      "grad_norm": 4.552779197692871,
+      "learning_rate": 1.793416458852868e-05,
+      "loss": 0.3933,
+      "step": 8290
+    },
+    {
+      "epoch": 2.0698254364089776,
+      "grad_norm": 5.0049004554748535,
+      "learning_rate": 1.7931670822942644e-05,
+      "loss": 0.4585,
+      "step": 8300
+    },
+    {
+      "epoch": 2.0723192019950125,
+      "grad_norm": 5.8257904052734375,
+      "learning_rate": 1.792917705735661e-05,
+      "loss": 0.4453,
+      "step": 8310
+    },
+    {
+      "epoch": 2.0748129675810474,
+      "grad_norm": 5.575977802276611,
+      "learning_rate": 1.7926683291770574e-05,
+      "loss": 0.4536,
+      "step": 8320
+    },
+    {
+      "epoch": 2.0773067331670823,
+      "grad_norm": 4.55002498626709,
+      "learning_rate": 1.792418952618454e-05,
+      "loss": 0.4842,
+      "step": 8330
+    },
+    {
+      "epoch": 2.079800498753117,
+      "grad_norm": 6.223612308502197,
+      "learning_rate": 1.7921695760598505e-05,
+      "loss": 0.4209,
+      "step": 8340
+    },
+    {
+      "epoch": 2.082294264339152,
+      "grad_norm": 5.589142322540283,
+      "learning_rate": 1.791920199501247e-05,
+      "loss": 0.3957,
+      "step": 8350
+    },
+    {
+      "epoch": 2.084788029925187,
+      "grad_norm": 5.206849575042725,
+      "learning_rate": 1.7916708229426435e-05,
+      "loss": 0.4253,
+      "step": 8360
+    },
+    {
+      "epoch": 2.087281795511222,
+      "grad_norm": 7.368830680847168,
+      "learning_rate": 1.79142144638404e-05,
+      "loss": 0.391,
+      "step": 8370
+    },
+    {
+      "epoch": 2.089775561097257,
+      "grad_norm": 4.8193817138671875,
+      "learning_rate": 1.7911720698254366e-05,
+      "loss": 0.397,
+      "step": 8380
+    },
+    {
+      "epoch": 2.092269326683292,
+      "grad_norm": 5.361727237701416,
+      "learning_rate": 1.790922693266833e-05,
+      "loss": 0.4468,
+      "step": 8390
+    },
+    {
+      "epoch": 2.0947630922693268,
+      "grad_norm": 5.065185546875,
+      "learning_rate": 1.7906733167082297e-05,
+      "loss": 0.4142,
+      "step": 8400
+    },
+    {
+      "epoch": 2.0972568578553616,
+      "grad_norm": 4.190534591674805,
+      "learning_rate": 1.790423940149626e-05,
+      "loss": 0.3448,
+      "step": 8410
+    },
+    {
+      "epoch": 2.0997506234413965,
+      "grad_norm": 7.922725677490234,
+      "learning_rate": 1.7901745635910227e-05,
+      "loss": 0.3987,
+      "step": 8420
+    },
+    {
+      "epoch": 2.1022443890274314,
+      "grad_norm": 4.495791912078857,
+      "learning_rate": 1.789925187032419e-05,
+      "loss": 0.4028,
+      "step": 8430
+    },
+    {
+      "epoch": 2.1047381546134662,
+      "grad_norm": 5.718870639801025,
+      "learning_rate": 1.7896758104738158e-05,
+      "loss": 0.3954,
+      "step": 8440
+    },
+    {
+      "epoch": 2.107231920199501,
+      "grad_norm": 5.126749038696289,
+      "learning_rate": 1.789426433915212e-05,
+      "loss": 0.387,
+      "step": 8450
+    },
+    {
+      "epoch": 2.109725685785536,
+      "grad_norm": 6.903149604797363,
+      "learning_rate": 1.7891770573566085e-05,
+      "loss": 0.3494,
+      "step": 8460
+    },
+    {
+      "epoch": 2.112219451371571,
+      "grad_norm": 6.055934429168701,
+      "learning_rate": 1.7889276807980052e-05,
+      "loss": 0.4843,
+      "step": 8470
+    },
+    {
+      "epoch": 2.114713216957606,
+      "grad_norm": 5.180309295654297,
+      "learning_rate": 1.7886783042394016e-05,
+      "loss": 0.3912,
+      "step": 8480
+    },
+    {
+      "epoch": 2.117206982543641,
+      "grad_norm": 5.292367458343506,
+      "learning_rate": 1.7884289276807983e-05,
+      "loss": 0.4334,
+      "step": 8490
+    },
+    {
+      "epoch": 2.119700748129676,
+      "grad_norm": 6.745512008666992,
+      "learning_rate": 1.7881795511221946e-05,
+      "loss": 0.3954,
+      "step": 8500
+    },
+    {
+      "epoch": 2.1221945137157108,
+      "grad_norm": 6.295596122741699,
+      "learning_rate": 1.787930174563591e-05,
+      "loss": 0.3938,
+      "step": 8510
+    },
+    {
+      "epoch": 2.1246882793017456,
+      "grad_norm": 3.398513078689575,
+      "learning_rate": 1.7876807980049877e-05,
+      "loss": 0.3826,
+      "step": 8520
+    },
+    {
+      "epoch": 2.1271820448877805,
+      "grad_norm": 7.377808570861816,
+      "learning_rate": 1.787431421446384e-05,
+      "loss": 0.4062,
+      "step": 8530
+    },
+    {
+      "epoch": 2.1296758104738154,
+      "grad_norm": 5.2859272956848145,
+      "learning_rate": 1.7871820448877808e-05,
+      "loss": 0.3489,
+      "step": 8540
+    },
+    {
+      "epoch": 2.1321695760598502,
+      "grad_norm": 5.597499370574951,
+      "learning_rate": 1.786932668329177e-05,
+      "loss": 0.3924,
+      "step": 8550
+    },
+    {
+      "epoch": 2.134663341645885,
+      "grad_norm": 4.033814430236816,
+      "learning_rate": 1.7866832917705735e-05,
+      "loss": 0.401,
+      "step": 8560
+    },
+    {
+      "epoch": 2.1371571072319204,
+      "grad_norm": 10.128519058227539,
+      "learning_rate": 1.7864339152119702e-05,
+      "loss": 0.4635,
+      "step": 8570
+    },
+    {
+      "epoch": 2.1396508728179553,
+      "grad_norm": 3.9514145851135254,
+      "learning_rate": 1.786184538653367e-05,
+      "loss": 0.4217,
+      "step": 8580
+    },
+    {
+      "epoch": 2.14214463840399,
+      "grad_norm": 4.673630714416504,
+      "learning_rate": 1.7859351620947632e-05,
+      "loss": 0.4318,
+      "step": 8590
+    },
+    {
+      "epoch": 2.144638403990025,
+      "grad_norm": 5.571914196014404,
+      "learning_rate": 1.78568578553616e-05,
+      "loss": 0.414,
+      "step": 8600
+    },
+    {
+      "epoch": 2.14713216957606,
+      "grad_norm": 4.7045063972473145,
+      "learning_rate": 1.7854364089775563e-05,
+      "loss": 0.4471,
+      "step": 8610
+    },
+    {
+      "epoch": 2.1496259351620948,
+      "grad_norm": 4.17690372467041,
+      "learning_rate": 1.785187032418953e-05,
+      "loss": 0.4632,
+      "step": 8620
+    },
+    {
+      "epoch": 2.1521197007481296,
+      "grad_norm": 5.189255237579346,
+      "learning_rate": 1.7849376558603494e-05,
+      "loss": 0.5894,
+      "step": 8630
+    },
+    {
+      "epoch": 2.1546134663341645,
+      "grad_norm": 5.136035442352295,
+      "learning_rate": 1.7846882793017457e-05,
+      "loss": 0.4954,
+      "step": 8640
+    },
+    {
+      "epoch": 2.1571072319201994,
+      "grad_norm": 6.0098443031311035,
+      "learning_rate": 1.7844389027431424e-05,
+      "loss": 0.3998,
+      "step": 8650
+    },
+    {
+      "epoch": 2.1596009975062342,
+      "grad_norm": 6.5170979499816895,
+      "learning_rate": 1.7841895261845388e-05,
+      "loss": 0.4394,
+      "step": 8660
+    },
+    {
+      "epoch": 2.162094763092269,
+      "grad_norm": 7.620209217071533,
+      "learning_rate": 1.783940149625935e-05,
+      "loss": 0.4352,
+      "step": 8670
+    },
+    {
+      "epoch": 2.1645885286783044,
+      "grad_norm": 5.3294548988342285,
+      "learning_rate": 1.783690773067332e-05,
+      "loss": 0.4285,
+      "step": 8680
+    },
+    {
+      "epoch": 2.1670822942643393,
+      "grad_norm": 4.24174165725708,
+      "learning_rate": 1.7834413965087282e-05,
+      "loss": 0.4855,
+      "step": 8690
+    },
+    {
+      "epoch": 2.169576059850374,
+      "grad_norm": 4.900123119354248,
+      "learning_rate": 1.783192019950125e-05,
+      "loss": 0.5264,
+      "step": 8700
+    },
+    {
+      "epoch": 2.172069825436409,
+      "grad_norm": 5.917686939239502,
+      "learning_rate": 1.7829426433915213e-05,
+      "loss": 0.4571,
+      "step": 8710
+    },
+    {
+      "epoch": 2.174563591022444,
+      "grad_norm": 5.674629211425781,
+      "learning_rate": 1.7826932668329176e-05,
+      "loss": 0.44,
+      "step": 8720
+    },
+    {
+      "epoch": 2.1770573566084788,
+      "grad_norm": 6.6443986892700195,
+      "learning_rate": 1.7824438902743143e-05,
+      "loss": 0.3972,
+      "step": 8730
+    },
+    {
+      "epoch": 2.1795511221945136,
+      "grad_norm": 4.557736873626709,
+      "learning_rate": 1.7821945137157107e-05,
+      "loss": 0.4578,
+      "step": 8740
+    },
+    {
+      "epoch": 2.1820448877805485,
+      "grad_norm": 5.653411865234375,
+      "learning_rate": 1.7819451371571074e-05,
+      "loss": 0.4691,
+      "step": 8750
+    },
+    {
+      "epoch": 2.1845386533665834,
+      "grad_norm": 5.747959613800049,
+      "learning_rate": 1.7816957605985038e-05,
+      "loss": 0.4597,
+      "step": 8760
+    },
+    {
+      "epoch": 2.1870324189526187,
+      "grad_norm": 4.90444803237915,
+      "learning_rate": 1.7814463840399005e-05,
+      "loss": 0.4057,
+      "step": 8770
+    },
+    {
+      "epoch": 2.1895261845386536,
+      "grad_norm": 5.881532192230225,
+      "learning_rate": 1.7811970074812968e-05,
+      "loss": 0.4028,
+      "step": 8780
+    },
+    {
+      "epoch": 2.1920199501246884,
+      "grad_norm": 5.233574390411377,
+      "learning_rate": 1.7809476309226935e-05,
+      "loss": 0.3608,
+      "step": 8790
+    },
+    {
+      "epoch": 2.1945137157107233,
+      "grad_norm": 6.65619421005249,
+      "learning_rate": 1.78069825436409e-05,
+      "loss": 0.3862,
+      "step": 8800
+    },
+    {
+      "epoch": 2.197007481296758,
+      "grad_norm": 4.836643695831299,
+      "learning_rate": 1.7804488778054866e-05,
+      "loss": 0.4005,
+      "step": 8810
+    },
+    {
+      "epoch": 2.199501246882793,
+      "grad_norm": 5.234933376312256,
+      "learning_rate": 1.780199501246883e-05,
+      "loss": 0.4855,
+      "step": 8820
+    },
+    {
+      "epoch": 2.201995012468828,
+      "grad_norm": 4.512629985809326,
+      "learning_rate": 1.7799501246882796e-05,
+      "loss": 0.4217,
+      "step": 8830
+    },
+    {
+      "epoch": 2.2044887780548628,
+      "grad_norm": 7.3246541023254395,
+      "learning_rate": 1.779700748129676e-05,
+      "loss": 0.4968,
+      "step": 8840
+    },
+    {
+      "epoch": 2.2069825436408976,
+      "grad_norm": 5.89198112487793,
+      "learning_rate": 1.7794513715710724e-05,
+      "loss": 0.3842,
+      "step": 8850
+    },
+    {
+      "epoch": 2.2094763092269325,
+      "grad_norm": 5.893228530883789,
+      "learning_rate": 1.779201995012469e-05,
+      "loss": 0.3892,
+      "step": 8860
+    },
+    {
+      "epoch": 2.2119700748129674,
+      "grad_norm": 11.834216117858887,
+      "learning_rate": 1.7789526184538654e-05,
+      "loss": 0.4357,
+      "step": 8870
+    },
+    {
+      "epoch": 2.2144638403990027,
+      "grad_norm": 6.141737461090088,
+      "learning_rate": 1.7787032418952618e-05,
+      "loss": 0.3841,
+      "step": 8880
+    },
+    {
+      "epoch": 2.2169576059850375,
+      "grad_norm": 10.668390274047852,
+      "learning_rate": 1.7784538653366585e-05,
+      "loss": 0.4482,
+      "step": 8890
+    },
+    {
+      "epoch": 2.2194513715710724,
+      "grad_norm": 5.236949443817139,
+      "learning_rate": 1.778204488778055e-05,
+      "loss": 0.4517,
+      "step": 8900
+    },
+    {
+      "epoch": 2.2219451371571073,
+      "grad_norm": 6.584339618682861,
+      "learning_rate": 1.7779551122194516e-05,
+      "loss": 0.4163,
+      "step": 8910
+    },
+    {
+      "epoch": 2.224438902743142,
+      "grad_norm": 6.040124893188477,
+      "learning_rate": 1.777705735660848e-05,
+      "loss": 0.4095,
+      "step": 8920
+    },
+    {
+      "epoch": 2.226932668329177,
+      "grad_norm": 7.1335062980651855,
+      "learning_rate": 1.7774563591022446e-05,
+      "loss": 0.5042,
+      "step": 8930
+    },
+    {
+      "epoch": 2.229426433915212,
+      "grad_norm": 8.298208236694336,
+      "learning_rate": 1.777206982543641e-05,
+      "loss": 0.5001,
+      "step": 8940
+    },
+    {
+      "epoch": 2.2319201995012468,
+      "grad_norm": 4.262420177459717,
+      "learning_rate": 1.7769576059850377e-05,
+      "loss": 0.4444,
+      "step": 8950
+    },
+    {
+      "epoch": 2.2344139650872816,
+      "grad_norm": 3.8085227012634277,
+      "learning_rate": 1.776708229426434e-05,
+      "loss": 0.4736,
+      "step": 8960
+    },
+    {
+      "epoch": 2.236907730673317,
+      "grad_norm": 8.276735305786133,
+      "learning_rate": 1.7764588528678307e-05,
+      "loss": 0.4522,
+      "step": 8970
+    },
+    {
+      "epoch": 2.239401496259352,
+      "grad_norm": 6.127620220184326,
+      "learning_rate": 1.776209476309227e-05,
+      "loss": 0.3601,
+      "step": 8980
+    },
+    {
+      "epoch": 2.2418952618453867,
+      "grad_norm": 5.782129287719727,
+      "learning_rate": 1.7759600997506238e-05,
+      "loss": 0.4117,
+      "step": 8990
+    },
+    {
+      "epoch": 2.2443890274314215,
+      "grad_norm": 6.870619773864746,
+      "learning_rate": 1.77571072319202e-05,
+      "loss": 0.4816,
+      "step": 9000
+    },
+    {
+      "epoch": 2.2468827930174564,
+      "grad_norm": 3.876192331314087,
+      "learning_rate": 1.7754613466334165e-05,
+      "loss": 0.41,
+      "step": 9010
+    },
+    {
+      "epoch": 2.2493765586034913,
+      "grad_norm": 7.108855724334717,
+      "learning_rate": 1.7752119700748132e-05,
+      "loss": 0.4026,
+      "step": 9020
+    },
+    {
+      "epoch": 2.251870324189526,
+      "grad_norm": 5.940926551818848,
+      "learning_rate": 1.7749625935162096e-05,
+      "loss": 0.4307,
+      "step": 9030
+    },
+    {
+      "epoch": 2.254364089775561,
+      "grad_norm": 7.2069926261901855,
+      "learning_rate": 1.7747132169576063e-05,
+      "loss": 0.4925,
+      "step": 9040
+    },
+    {
+      "epoch": 2.256857855361596,
+      "grad_norm": 5.103030681610107,
+      "learning_rate": 1.7744638403990026e-05,
+      "loss": 0.4119,
+      "step": 9050
+    },
+    {
+      "epoch": 2.2593516209476308,
+      "grad_norm": 6.820674419403076,
+      "learning_rate": 1.774214463840399e-05,
+      "loss": 0.4423,
+      "step": 9060
+    },
+    {
+      "epoch": 2.2618453865336656,
+      "grad_norm": 5.859368801116943,
+      "learning_rate": 1.7739650872817957e-05,
+      "loss": 0.416,
+      "step": 9070
+    },
+    {
+      "epoch": 2.264339152119701,
+      "grad_norm": 3.5914974212646484,
+      "learning_rate": 1.773715710723192e-05,
+      "loss": 0.3593,
+      "step": 9080
+    },
+    {
+      "epoch": 2.266832917705736,
+      "grad_norm": 5.854361057281494,
+      "learning_rate": 1.7734663341645884e-05,
+      "loss": 0.3665,
+      "step": 9090
+    },
+    {
+      "epoch": 2.2693266832917707,
+      "grad_norm": 6.185330390930176,
+      "learning_rate": 1.773216957605985e-05,
+      "loss": 0.3968,
+      "step": 9100
+    },
+    {
+      "epoch": 2.2718204488778055,
+      "grad_norm": 9.060643196105957,
+      "learning_rate": 1.7729675810473815e-05,
+      "loss": 0.444,
+      "step": 9110
+    },
+    {
+      "epoch": 2.2743142144638404,
+      "grad_norm": 6.3430633544921875,
+      "learning_rate": 1.7727182044887782e-05,
+      "loss": 0.4807,
+      "step": 9120
+    },
+    {
+      "epoch": 2.2768079800498753,
+      "grad_norm": 6.883560657501221,
+      "learning_rate": 1.7724688279301746e-05,
+      "loss": 0.4042,
+      "step": 9130
+    },
+    {
+      "epoch": 2.27930174563591,
+      "grad_norm": 5.706015110015869,
+      "learning_rate": 1.7722194513715713e-05,
+      "loss": 0.4687,
+      "step": 9140
+    },
+    {
+      "epoch": 2.281795511221945,
+      "grad_norm": 6.107770919799805,
+      "learning_rate": 1.7719700748129676e-05,
+      "loss": 0.6099,
+      "step": 9150
+    },
+    {
+      "epoch": 2.28428927680798,
+      "grad_norm": 4.1198930740356445,
+      "learning_rate": 1.7717206982543643e-05,
+      "loss": 0.3808,
+      "step": 9160
+    },
+    {
+      "epoch": 2.286783042394015,
+      "grad_norm": 5.448758602142334,
+      "learning_rate": 1.7714713216957607e-05,
+      "loss": 0.3805,
+      "step": 9170
+    },
+    {
+      "epoch": 2.28927680798005,
+      "grad_norm": 4.109248638153076,
+      "learning_rate": 1.7712219451371574e-05,
+      "loss": 0.3963,
+      "step": 9180
+    },
+    {
+      "epoch": 2.291770573566085,
+      "grad_norm": 4.992825508117676,
+      "learning_rate": 1.7709725685785537e-05,
+      "loss": 0.3987,
+      "step": 9190
+    },
+    {
+      "epoch": 2.29426433915212,
+      "grad_norm": 4.800196647644043,
+      "learning_rate": 1.7707231920199504e-05,
+      "loss": 0.3984,
+      "step": 9200
+    },
+    {
+      "epoch": 2.2967581047381547,
+      "grad_norm": 7.4912896156311035,
+      "learning_rate": 1.7704738154613468e-05,
+      "loss": 0.462,
+      "step": 9210
+    },
+    {
+      "epoch": 2.2992518703241895,
+      "grad_norm": 7.3595991134643555,
+      "learning_rate": 1.770224438902743e-05,
+      "loss": 0.4066,
+      "step": 9220
+    },
+    {
+      "epoch": 2.3017456359102244,
+      "grad_norm": 8.947257995605469,
+      "learning_rate": 1.76997506234414e-05,
+      "loss": 0.4297,
+      "step": 9230
+    },
+    {
+      "epoch": 2.3042394014962593,
+      "grad_norm": 5.556568622589111,
+      "learning_rate": 1.7697256857855362e-05,
+      "loss": 0.4001,
+      "step": 9240
+    },
+    {
+      "epoch": 2.306733167082294,
+      "grad_norm": 5.276797294616699,
+      "learning_rate": 1.769476309226933e-05,
+      "loss": 0.4318,
+      "step": 9250
+    },
+    {
+      "epoch": 2.309226932668329,
+      "grad_norm": 7.36549711227417,
+      "learning_rate": 1.7692269326683293e-05,
+      "loss": 0.4791,
+      "step": 9260
+    },
+    {
+      "epoch": 2.311720698254364,
+      "grad_norm": 6.3212761878967285,
+      "learning_rate": 1.7689775561097257e-05,
+      "loss": 0.4505,
+      "step": 9270
+    },
+    {
+      "epoch": 2.314214463840399,
+      "grad_norm": 8.907316207885742,
+      "learning_rate": 1.7687281795511224e-05,
+      "loss": 0.5421,
+      "step": 9280
+    },
+    {
+      "epoch": 2.316708229426434,
+      "grad_norm": 5.004027843475342,
+      "learning_rate": 1.7684788029925187e-05,
+      "loss": 0.4232,
+      "step": 9290
+    },
+    {
+      "epoch": 2.319201995012469,
+      "grad_norm": 5.411210060119629,
+      "learning_rate": 1.7682294264339154e-05,
+      "loss": 0.374,
+      "step": 9300
+    },
+    {
+      "epoch": 2.321695760598504,
+      "grad_norm": 5.966613292694092,
+      "learning_rate": 1.7679800498753118e-05,
+      "loss": 0.4109,
+      "step": 9310
+    },
+    {
+      "epoch": 2.3241895261845387,
+      "grad_norm": 5.003303050994873,
+      "learning_rate": 1.7677306733167085e-05,
+      "loss": 0.4241,
+      "step": 9320
+    },
+    {
+      "epoch": 2.3266832917705735,
+      "grad_norm": 5.208148002624512,
+      "learning_rate": 1.767481296758105e-05,
+      "loss": 0.4793,
+      "step": 9330
+    },
+    {
+      "epoch": 2.3291770573566084,
+      "grad_norm": 8.17491626739502,
+      "learning_rate": 1.7672319201995015e-05,
+      "loss": 0.4423,
+      "step": 9340
+    },
+    {
+      "epoch": 2.3316708229426433,
+      "grad_norm": 8.450447082519531,
+      "learning_rate": 1.766982543640898e-05,
+      "loss": 0.4155,
+      "step": 9350
+    },
+    {
+      "epoch": 2.334164588528678,
+      "grad_norm": 6.473998546600342,
+      "learning_rate": 1.7667331670822946e-05,
+      "loss": 0.425,
+      "step": 9360
+    },
+    {
+      "epoch": 2.3366583541147135,
+      "grad_norm": 8.660719871520996,
+      "learning_rate": 1.766483790523691e-05,
+      "loss": 0.4596,
+      "step": 9370
+    },
+    {
+      "epoch": 2.3391521197007483,
+      "grad_norm": 5.478455543518066,
+      "learning_rate": 1.7662344139650873e-05,
+      "loss": 0.3841,
+      "step": 9380
+    },
+    {
+      "epoch": 2.341645885286783,
+      "grad_norm": 5.475340843200684,
+      "learning_rate": 1.765985037406484e-05,
+      "loss": 0.4179,
+      "step": 9390
+    },
+    {
+      "epoch": 2.344139650872818,
+      "grad_norm": 4.660368919372559,
+      "learning_rate": 1.7657356608478804e-05,
+      "loss": 0.4493,
+      "step": 9400
+    },
+    {
+      "epoch": 2.346633416458853,
+      "grad_norm": 3.002950668334961,
+      "learning_rate": 1.765486284289277e-05,
+      "loss": 0.3333,
+      "step": 9410
+    },
+    {
+      "epoch": 2.349127182044888,
+      "grad_norm": 7.417391777038574,
+      "learning_rate": 1.7652369077306734e-05,
+      "loss": 0.3882,
+      "step": 9420
+    },
+    {
+      "epoch": 2.3516209476309227,
+      "grad_norm": 5.637429714202881,
+      "learning_rate": 1.7649875311720698e-05,
+      "loss": 0.3217,
+      "step": 9430
+    },
+    {
+      "epoch": 2.3541147132169575,
+      "grad_norm": 9.37643814086914,
+      "learning_rate": 1.7647381546134665e-05,
+      "loss": 0.423,
+      "step": 9440
+    },
+    {
+      "epoch": 2.3566084788029924,
+      "grad_norm": 4.772799968719482,
+      "learning_rate": 1.764488778054863e-05,
+      "loss": 0.4789,
+      "step": 9450
+    },
+    {
+      "epoch": 2.3591022443890273,
+      "grad_norm": 6.806458950042725,
+      "learning_rate": 1.7642394014962592e-05,
+      "loss": 0.4245,
+      "step": 9460
+    },
+    {
+      "epoch": 2.361596009975062,
+      "grad_norm": 6.0944647789001465,
+      "learning_rate": 1.763990024937656e-05,
+      "loss": 0.4306,
+      "step": 9470
+    },
+    {
+      "epoch": 2.3640897755610975,
+      "grad_norm": 4.526205062866211,
+      "learning_rate": 1.7637406483790523e-05,
+      "loss": 0.3647,
+      "step": 9480
+    },
+    {
+      "epoch": 2.3665835411471323,
+      "grad_norm": 10.576577186584473,
+      "learning_rate": 1.763491271820449e-05,
+      "loss": 0.4645,
+      "step": 9490
+    },
+    {
+      "epoch": 2.369077306733167,
+      "grad_norm": 8.364253044128418,
+      "learning_rate": 1.7632418952618457e-05,
+      "loss": 0.4524,
+      "step": 9500
+    },
+    {
+      "epoch": 2.371571072319202,
+      "grad_norm": 6.188162326812744,
+      "learning_rate": 1.762992518703242e-05,
+      "loss": 0.446,
+      "step": 9510
+    },
+    {
+      "epoch": 2.374064837905237,
+      "grad_norm": 5.653889179229736,
+      "learning_rate": 1.7627431421446388e-05,
+      "loss": 0.3859,
+      "step": 9520
+    },
+    {
+      "epoch": 2.376558603491272,
+      "grad_norm": 7.959500789642334,
+      "learning_rate": 1.762493765586035e-05,
+      "loss": 0.4085,
+      "step": 9530
+    },
+    {
+      "epoch": 2.3790523690773067,
+      "grad_norm": 7.066885471343994,
+      "learning_rate": 1.7622443890274318e-05,
+      "loss": 0.4195,
+      "step": 9540
+    },
+    {
+      "epoch": 2.3815461346633415,
+      "grad_norm": 4.418666839599609,
+      "learning_rate": 1.7619950124688282e-05,
+      "loss": 0.3748,
+      "step": 9550
+    },
+    {
+      "epoch": 2.3840399002493764,
+      "grad_norm": 7.192332744598389,
+      "learning_rate": 1.7617456359102245e-05,
+      "loss": 0.4545,
+      "step": 9560
+    },
+    {
+      "epoch": 2.3865336658354117,
+      "grad_norm": 6.895399570465088,
+      "learning_rate": 1.7614962593516212e-05,
+      "loss": 0.4023,
+      "step": 9570
+    },
+    {
+      "epoch": 2.3890274314214466,
+      "grad_norm": 5.375655174255371,
+      "learning_rate": 1.7612468827930176e-05,
+      "loss": 0.4108,
+      "step": 9580
+    },
+    {
+      "epoch": 2.3915211970074814,
+      "grad_norm": 5.821678161621094,
+      "learning_rate": 1.760997506234414e-05,
+      "loss": 0.4258,
+      "step": 9590
+    },
+    {
+      "epoch": 2.3940149625935163,
+      "grad_norm": 7.174010276794434,
+      "learning_rate": 1.7607481296758107e-05,
+      "loss": 0.4199,
+      "step": 9600
+    },
+    {
+      "epoch": 2.396508728179551,
+      "grad_norm": 4.9364519119262695,
+      "learning_rate": 1.760498753117207e-05,
+      "loss": 0.4044,
+      "step": 9610
+    },
+    {
+      "epoch": 2.399002493765586,
+      "grad_norm": 6.391279220581055,
+      "learning_rate": 1.7602493765586037e-05,
+      "loss": 0.4368,
+      "step": 9620
+    },
+    {
+      "epoch": 2.401496259351621,
+      "grad_norm": 4.990833282470703,
+      "learning_rate": 1.76e-05,
+      "loss": 0.3174,
+      "step": 9630
+    },
+    {
+      "epoch": 2.403990024937656,
+      "grad_norm": 4.146638870239258,
+      "learning_rate": 1.7597506234413965e-05,
+      "loss": 0.3995,
+      "step": 9640
+    },
+    {
+      "epoch": 2.4064837905236907,
+      "grad_norm": 4.520384311676025,
+      "learning_rate": 1.759501246882793e-05,
+      "loss": 0.3627,
+      "step": 9650
+    },
+    {
+      "epoch": 2.4089775561097255,
+      "grad_norm": 5.92678165435791,
+      "learning_rate": 1.7592518703241895e-05,
+      "loss": 0.3737,
+      "step": 9660
+    },
+    {
+      "epoch": 2.4114713216957604,
+      "grad_norm": 8.433207511901855,
+      "learning_rate": 1.7590024937655862e-05,
+      "loss": 0.5038,
+      "step": 9670
+    },
+    {
+      "epoch": 2.4139650872817953,
+      "grad_norm": 6.0681986808776855,
+      "learning_rate": 1.7587531172069826e-05,
+      "loss": 0.4544,
+      "step": 9680
+    },
+    {
+      "epoch": 2.4164588528678306,
+      "grad_norm": 6.954349040985107,
+      "learning_rate": 1.7585037406483793e-05,
+      "loss": 0.4765,
+      "step": 9690
+    },
+    {
+      "epoch": 2.4189526184538654,
+      "grad_norm": 5.360884666442871,
+      "learning_rate": 1.7582543640897756e-05,
+      "loss": 0.4173,
+      "step": 9700
+    },
+    {
+      "epoch": 2.4214463840399003,
+      "grad_norm": 5.455905437469482,
+      "learning_rate": 1.7580049875311723e-05,
+      "loss": 0.3981,
+      "step": 9710
+    },
+    {
+      "epoch": 2.423940149625935,
+      "grad_norm": 7.694284439086914,
+      "learning_rate": 1.7577556109725687e-05,
+      "loss": 0.4159,
+      "step": 9720
+    },
+    {
+      "epoch": 2.42643391521197,
+      "grad_norm": 2.6241650581359863,
+      "learning_rate": 1.7575062344139654e-05,
+      "loss": 0.4415,
+      "step": 9730
+    },
+    {
+      "epoch": 2.428927680798005,
+      "grad_norm": 5.173274040222168,
+      "learning_rate": 1.7572568578553618e-05,
+      "loss": 0.3576,
+      "step": 9740
+    },
+    {
+      "epoch": 2.43142144638404,
+      "grad_norm": 5.52868127822876,
+      "learning_rate": 1.7570074812967585e-05,
+      "loss": 0.4224,
+      "step": 9750
+    },
+    {
+      "epoch": 2.4339152119700747,
+      "grad_norm": 6.820600509643555,
+      "learning_rate": 1.7567581047381548e-05,
+      "loss": 0.4329,
+      "step": 9760
+    },
+    {
+      "epoch": 2.43640897755611,
+      "grad_norm": 5.07393741607666,
+      "learning_rate": 1.7565087281795512e-05,
+      "loss": 0.4344,
+      "step": 9770
+    },
+    {
+      "epoch": 2.438902743142145,
+      "grad_norm": 6.3466572761535645,
+      "learning_rate": 1.756259351620948e-05,
+      "loss": 0.3851,
+      "step": 9780
+    },
+    {
+      "epoch": 2.4413965087281797,
+      "grad_norm": 4.636044025421143,
+      "learning_rate": 1.7560099750623442e-05,
+      "loss": 0.4259,
+      "step": 9790
+    },
+    {
+      "epoch": 2.4438902743142146,
+      "grad_norm": 4.414524078369141,
+      "learning_rate": 1.7557605985037406e-05,
+      "loss": 0.4189,
+      "step": 9800
+    },
+    {
+      "epoch": 2.4463840399002494,
+      "grad_norm": 5.910994529724121,
+      "learning_rate": 1.7555112219451373e-05,
+      "loss": 0.3701,
+      "step": 9810
+    },
+    {
+      "epoch": 2.4488778054862843,
+      "grad_norm": 5.326263427734375,
+      "learning_rate": 1.7552618453865337e-05,
+      "loss": 0.3784,
+      "step": 9820
+    },
+    {
+      "epoch": 2.451371571072319,
+      "grad_norm": 4.930222511291504,
+      "learning_rate": 1.7550124688279304e-05,
+      "loss": 0.4245,
+      "step": 9830
+    },
+    {
+      "epoch": 2.453865336658354,
+      "grad_norm": 14.357271194458008,
+      "learning_rate": 1.7547630922693267e-05,
+      "loss": 0.5032,
+      "step": 9840
+    },
+    {
+      "epoch": 2.456359102244389,
+      "grad_norm": 6.140951633453369,
+      "learning_rate": 1.7545137157107234e-05,
+      "loss": 0.3132,
+      "step": 9850
+    },
+    {
+      "epoch": 2.458852867830424,
+      "grad_norm": 6.1243743896484375,
+      "learning_rate": 1.7542643391521198e-05,
+      "loss": 0.4094,
+      "step": 9860
+    },
+    {
+      "epoch": 2.4613466334164587,
+      "grad_norm": 5.133067607879639,
+      "learning_rate": 1.7540149625935165e-05,
+      "loss": 0.4155,
+      "step": 9870
+    },
+    {
+      "epoch": 2.4638403990024935,
+      "grad_norm": 5.039419174194336,
+      "learning_rate": 1.753765586034913e-05,
+      "loss": 0.3812,
+      "step": 9880
+    },
+    {
+      "epoch": 2.466334164588529,
+      "grad_norm": 4.443388938903809,
+      "learning_rate": 1.7535162094763096e-05,
+      "loss": 0.3958,
+      "step": 9890
+    },
+    {
+      "epoch": 2.4688279301745637,
+      "grad_norm": 4.9401140213012695,
+      "learning_rate": 1.753266832917706e-05,
+      "loss": 0.3768,
+      "step": 9900
+    },
+    {
+      "epoch": 2.4713216957605986,
+      "grad_norm": 4.5980916023254395,
+      "learning_rate": 1.7530174563591026e-05,
+      "loss": 0.3657,
+      "step": 9910
+    },
+    {
+      "epoch": 2.4738154613466334,
+      "grad_norm": 4.5691423416137695,
+      "learning_rate": 1.752768079800499e-05,
+      "loss": 0.5407,
+      "step": 9920
+    },
+    {
+      "epoch": 2.4763092269326683,
+      "grad_norm": 10.96796989440918,
+      "learning_rate": 1.7525187032418953e-05,
+      "loss": 0.3909,
+      "step": 9930
+    },
+    {
+      "epoch": 2.478802992518703,
+      "grad_norm": 8.20064640045166,
+      "learning_rate": 1.752269326683292e-05,
+      "loss": 0.3977,
+      "step": 9940
+    },
+    {
+      "epoch": 2.481296758104738,
+      "grad_norm": 6.133816242218018,
+      "learning_rate": 1.7520199501246884e-05,
+      "loss": 0.4741,
+      "step": 9950
+    },
+    {
+      "epoch": 2.483790523690773,
+      "grad_norm": 3.481630325317383,
+      "learning_rate": 1.751770573566085e-05,
+      "loss": 0.359,
+      "step": 9960
+    },
+    {
+      "epoch": 2.4862842892768082,
+      "grad_norm": 6.158251762390137,
+      "learning_rate": 1.7515211970074815e-05,
+      "loss": 0.43,
+      "step": 9970
+    },
+    {
+      "epoch": 2.488778054862843,
+      "grad_norm": 6.90919828414917,
+      "learning_rate": 1.7512718204488778e-05,
+      "loss": 0.4154,
+      "step": 9980
+    },
+    {
+      "epoch": 2.491271820448878,
+      "grad_norm": 6.205567836761475,
+      "learning_rate": 1.7510224438902745e-05,
+      "loss": 0.4585,
+      "step": 9990
+    },
+    {
+      "epoch": 2.493765586034913,
+      "grad_norm": 8.581686973571777,
+      "learning_rate": 1.750773067331671e-05,
+      "loss": 0.4748,
+      "step": 10000
+    },
+    {
+      "epoch": 2.4962593516209477,
+      "grad_norm": 6.189614295959473,
+      "learning_rate": 1.7505236907730673e-05,
+      "loss": 0.4728,
+      "step": 10010
+    },
+    {
+      "epoch": 2.4987531172069826,
+      "grad_norm": 6.721256732940674,
+      "learning_rate": 1.750274314214464e-05,
+      "loss": 0.4508,
+      "step": 10020
+    },
+    {
+      "epoch": 2.5012468827930174,
+      "grad_norm": 7.564716815948486,
+      "learning_rate": 1.7500249376558603e-05,
+      "loss": 0.4478,
+      "step": 10030
+    },
+    {
+      "epoch": 2.5037406483790523,
+      "grad_norm": 6.12339973449707,
+      "learning_rate": 1.749775561097257e-05,
+      "loss": 0.3435,
+      "step": 10040
+    },
+    {
+      "epoch": 2.506234413965087,
+      "grad_norm": 6.196720123291016,
+      "learning_rate": 1.7495261845386534e-05,
+      "loss": 0.4308,
+      "step": 10050
+    },
+    {
+      "epoch": 2.508728179551122,
+      "grad_norm": 4.724903106689453,
+      "learning_rate": 1.74927680798005e-05,
+      "loss": 0.4072,
+      "step": 10060
+    },
+    {
+      "epoch": 2.511221945137157,
+      "grad_norm": 5.753084659576416,
+      "learning_rate": 1.7490274314214464e-05,
+      "loss": 0.3926,
+      "step": 10070
+    },
+    {
+      "epoch": 2.5137157107231918,
+      "grad_norm": 4.251994609832764,
+      "learning_rate": 1.748778054862843e-05,
+      "loss": 0.4337,
+      "step": 10080
+    },
+    {
+      "epoch": 2.516209476309227,
+      "grad_norm": 5.593048095703125,
+      "learning_rate": 1.7485286783042395e-05,
+      "loss": 0.3766,
+      "step": 10090
+    },
+    {
+      "epoch": 2.518703241895262,
+      "grad_norm": 3.456326723098755,
+      "learning_rate": 1.7482793017456362e-05,
+      "loss": 0.416,
+      "step": 10100
+    },
+    {
+      "epoch": 2.521197007481297,
+      "grad_norm": 6.803287982940674,
+      "learning_rate": 1.7480299251870326e-05,
+      "loss": 0.4372,
+      "step": 10110
+    },
+    {
+      "epoch": 2.5236907730673317,
+      "grad_norm": 6.90035343170166,
+      "learning_rate": 1.7477805486284293e-05,
+      "loss": 0.3842,
+      "step": 10120
+    },
+    {
+      "epoch": 2.5261845386533666,
+      "grad_norm": 7.092517375946045,
+      "learning_rate": 1.7475311720698256e-05,
+      "loss": 0.4817,
+      "step": 10130
+    },
+    {
+      "epoch": 2.5286783042394014,
+      "grad_norm": 7.831302642822266,
+      "learning_rate": 1.747281795511222e-05,
+      "loss": 0.3565,
+      "step": 10140
+    },
+    {
+      "epoch": 2.5311720698254363,
+      "grad_norm": 5.484219551086426,
+      "learning_rate": 1.7470324189526187e-05,
+      "loss": 0.4546,
+      "step": 10150
+    },
+    {
+      "epoch": 2.533665835411471,
+      "grad_norm": 9.50973129272461,
+      "learning_rate": 1.746783042394015e-05,
+      "loss": 0.3583,
+      "step": 10160
+    },
+    {
+      "epoch": 2.5361596009975065,
+      "grad_norm": 6.531426429748535,
+      "learning_rate": 1.7465336658354114e-05,
+      "loss": 0.3899,
+      "step": 10170
+    },
+    {
+      "epoch": 2.5386533665835413,
+      "grad_norm": 6.3029632568359375,
+      "learning_rate": 1.746284289276808e-05,
+      "loss": 0.4156,
+      "step": 10180
+    },
+    {
+      "epoch": 2.541147132169576,
+      "grad_norm": 6.3402252197265625,
+      "learning_rate": 1.7460349127182045e-05,
+      "loss": 0.5982,
+      "step": 10190
+    },
+    {
+      "epoch": 2.543640897755611,
+      "grad_norm": 6.966554164886475,
+      "learning_rate": 1.745785536159601e-05,
+      "loss": 0.4151,
+      "step": 10200
+    },
+    {
+      "epoch": 2.546134663341646,
+      "grad_norm": 4.213450908660889,
+      "learning_rate": 1.7455361596009975e-05,
+      "loss": 0.4014,
+      "step": 10210
+    },
+    {
+      "epoch": 2.548628428927681,
+      "grad_norm": 6.965415000915527,
+      "learning_rate": 1.7452867830423942e-05,
+      "loss": 0.5053,
+      "step": 10220
+    },
+    {
+      "epoch": 2.5511221945137157,
+      "grad_norm": 7.399175643920898,
+      "learning_rate": 1.7450374064837906e-05,
+      "loss": 0.4794,
+      "step": 10230
+    },
+    {
+      "epoch": 2.5536159600997506,
+      "grad_norm": 7.18726921081543,
+      "learning_rate": 1.7447880299251873e-05,
+      "loss": 0.4237,
+      "step": 10240
+    },
+    {
+      "epoch": 2.5561097256857854,
+      "grad_norm": 5.636864185333252,
+      "learning_rate": 1.7445386533665837e-05,
+      "loss": 0.4104,
+      "step": 10250
+    },
+    {
+      "epoch": 2.5586034912718203,
+      "grad_norm": 6.8820977210998535,
+      "learning_rate": 1.7442892768079804e-05,
+      "loss": 0.4435,
+      "step": 10260
+    },
+    {
+      "epoch": 2.561097256857855,
+      "grad_norm": 9.02462100982666,
+      "learning_rate": 1.7440399002493767e-05,
+      "loss": 0.4672,
+      "step": 10270
+    },
+    {
+      "epoch": 2.56359102244389,
+      "grad_norm": 5.171300411224365,
+      "learning_rate": 1.7437905236907734e-05,
+      "loss": 0.4258,
+      "step": 10280
+    },
+    {
+      "epoch": 2.5660847880299253,
+      "grad_norm": 11.12934684753418,
+      "learning_rate": 1.7435411471321698e-05,
+      "loss": 0.3967,
+      "step": 10290
+    },
+    {
+      "epoch": 2.56857855361596,
+      "grad_norm": 6.156098365783691,
+      "learning_rate": 1.743291770573566e-05,
+      "loss": 0.3973,
+      "step": 10300
+    },
+    {
+      "epoch": 2.571072319201995,
+      "grad_norm": 4.736686706542969,
+      "learning_rate": 1.743042394014963e-05,
+      "loss": 0.462,
+      "step": 10310
+    },
+    {
+      "epoch": 2.57356608478803,
+      "grad_norm": 8.570772171020508,
+      "learning_rate": 1.7427930174563592e-05,
+      "loss": 0.4494,
+      "step": 10320
+    },
+    {
+      "epoch": 2.576059850374065,
+      "grad_norm": 7.194964408874512,
+      "learning_rate": 1.742543640897756e-05,
+      "loss": 0.4351,
+      "step": 10330
+    },
+    {
+      "epoch": 2.5785536159600997,
+      "grad_norm": 4.6874918937683105,
+      "learning_rate": 1.7422942643391523e-05,
+      "loss": 0.4341,
+      "step": 10340
+    },
+    {
+      "epoch": 2.5810473815461346,
+      "grad_norm": 6.765628337860107,
+      "learning_rate": 1.7420448877805486e-05,
+      "loss": 0.4364,
+      "step": 10350
+    },
+    {
+      "epoch": 2.5835411471321694,
+      "grad_norm": 6.419005870819092,
+      "learning_rate": 1.7417955112219453e-05,
+      "loss": 0.4402,
+      "step": 10360
+    },
+    {
+      "epoch": 2.5860349127182047,
+      "grad_norm": 6.456396102905273,
+      "learning_rate": 1.7415461346633417e-05,
+      "loss": 0.4289,
+      "step": 10370
+    },
+    {
+      "epoch": 2.5885286783042396,
+      "grad_norm": 6.123867034912109,
+      "learning_rate": 1.741296758104738e-05,
+      "loss": 0.4403,
+      "step": 10380
+    },
+    {
+      "epoch": 2.5910224438902745,
+      "grad_norm": 5.706097602844238,
+      "learning_rate": 1.7410473815461347e-05,
+      "loss": 0.4294,
+      "step": 10390
+    },
+    {
+      "epoch": 2.5935162094763093,
+      "grad_norm": 6.1752543449401855,
+      "learning_rate": 1.740798004987531e-05,
+      "loss": 0.3477,
+      "step": 10400
+    },
+    {
+      "epoch": 2.596009975062344,
+      "grad_norm": 5.515731334686279,
+      "learning_rate": 1.7405486284289278e-05,
+      "loss": 0.4414,
+      "step": 10410
+    },
+    {
+      "epoch": 2.598503740648379,
+      "grad_norm": 6.281205654144287,
+      "learning_rate": 1.7402992518703242e-05,
+      "loss": 0.4038,
+      "step": 10420
+    },
+    {
+      "epoch": 2.600997506234414,
+      "grad_norm": 5.598005294799805,
+      "learning_rate": 1.740049875311721e-05,
+      "loss": 0.3969,
+      "step": 10430
+    },
+    {
+      "epoch": 2.603491271820449,
+      "grad_norm": 4.307186126708984,
+      "learning_rate": 1.7398004987531176e-05,
+      "loss": 0.4112,
+      "step": 10440
+    },
+    {
+      "epoch": 2.6059850374064837,
+      "grad_norm": 5.855963230133057,
+      "learning_rate": 1.739551122194514e-05,
+      "loss": 0.4137,
+      "step": 10450
+    },
+    {
+      "epoch": 2.6084788029925186,
+      "grad_norm": 6.094501972198486,
+      "learning_rate": 1.7393017456359106e-05,
+      "loss": 0.4384,
+      "step": 10460
+    },
+    {
+      "epoch": 2.6109725685785534,
+      "grad_norm": 7.285081386566162,
+      "learning_rate": 1.739052369077307e-05,
+      "loss": 0.4306,
+      "step": 10470
+    },
+    {
+      "epoch": 2.6134663341645883,
+      "grad_norm": 7.074899196624756,
+      "learning_rate": 1.7388029925187034e-05,
+      "loss": 0.4767,
+      "step": 10480
+    },
+    {
+      "epoch": 2.6159600997506236,
+      "grad_norm": 6.743325710296631,
+      "learning_rate": 1.7385536159601e-05,
+      "loss": 0.4486,
+      "step": 10490
+    },
+    {
+      "epoch": 2.6184538653366585,
+      "grad_norm": 5.235808372497559,
+      "learning_rate": 1.7383042394014964e-05,
+      "loss": 0.4621,
+      "step": 10500
+    },
+    {
+      "epoch": 2.6209476309226933,
+      "grad_norm": 5.081497669219971,
+      "learning_rate": 1.7380548628428928e-05,
+      "loss": 0.4273,
+      "step": 10510
+    },
+    {
+      "epoch": 2.623441396508728,
+      "grad_norm": 5.6099443435668945,
+      "learning_rate": 1.7378054862842895e-05,
+      "loss": 0.4259,
+      "step": 10520
+    },
+    {
+      "epoch": 2.625935162094763,
+      "grad_norm": 6.103354454040527,
+      "learning_rate": 1.737556109725686e-05,
+      "loss": 0.4568,
+      "step": 10530
+    },
+    {
+      "epoch": 2.628428927680798,
+      "grad_norm": 4.791126728057861,
+      "learning_rate": 1.7373067331670825e-05,
+      "loss": 0.4052,
+      "step": 10540
+    },
+    {
+      "epoch": 2.630922693266833,
+      "grad_norm": 4.903756618499756,
+      "learning_rate": 1.737057356608479e-05,
+      "loss": 0.4314,
+      "step": 10550
+    },
+    {
+      "epoch": 2.6334164588528677,
+      "grad_norm": 5.318016052246094,
+      "learning_rate": 1.7368079800498753e-05,
+      "loss": 0.4311,
+      "step": 10560
+    },
+    {
+      "epoch": 2.635910224438903,
+      "grad_norm": 10.588895797729492,
+      "learning_rate": 1.736558603491272e-05,
+      "loss": 0.4499,
+      "step": 10570
+    },
+    {
+      "epoch": 2.638403990024938,
+      "grad_norm": 4.9873833656311035,
+      "learning_rate": 1.7363092269326683e-05,
+      "loss": 0.4632,
+      "step": 10580
+    },
+    {
+      "epoch": 2.6408977556109727,
+      "grad_norm": 7.042355537414551,
+      "learning_rate": 1.736059850374065e-05,
+      "loss": 0.3997,
+      "step": 10590
+    },
+    {
+      "epoch": 2.6433915211970076,
+      "grad_norm": 6.466985702514648,
+      "learning_rate": 1.7358104738154614e-05,
+      "loss": 0.4844,
+      "step": 10600
+    },
+    {
+      "epoch": 2.6458852867830425,
+      "grad_norm": 4.656626224517822,
+      "learning_rate": 1.735561097256858e-05,
+      "loss": 0.3745,
+      "step": 10610
+    },
+    {
+      "epoch": 2.6483790523690773,
+      "grad_norm": 9.50407886505127,
+      "learning_rate": 1.7353117206982545e-05,
+      "loss": 0.5028,
+      "step": 10620
+    },
+    {
+      "epoch": 2.650872817955112,
+      "grad_norm": 5.126711845397949,
+      "learning_rate": 1.735062344139651e-05,
+      "loss": 0.4899,
+      "step": 10630
+    },
+    {
+      "epoch": 2.653366583541147,
+      "grad_norm": 10.905644416809082,
+      "learning_rate": 1.7348129675810475e-05,
+      "loss": 0.4768,
+      "step": 10640
+    },
+    {
+      "epoch": 2.655860349127182,
+      "grad_norm": 5.238419532775879,
+      "learning_rate": 1.7345635910224442e-05,
+      "loss": 0.3905,
+      "step": 10650
+    },
+    {
+      "epoch": 2.658354114713217,
+      "grad_norm": 5.836606979370117,
+      "learning_rate": 1.7343142144638406e-05,
+      "loss": 0.4586,
+      "step": 10660
+    },
+    {
+      "epoch": 2.6608478802992517,
+      "grad_norm": 6.579746723175049,
+      "learning_rate": 1.734064837905237e-05,
+      "loss": 0.5069,
+      "step": 10670
+    },
+    {
+      "epoch": 2.6633416458852865,
+      "grad_norm": 4.4686279296875,
+      "learning_rate": 1.7338154613466336e-05,
+      "loss": 0.3852,
+      "step": 10680
+    },
+    {
+      "epoch": 2.665835411471322,
+      "grad_norm": 6.548900604248047,
+      "learning_rate": 1.73356608478803e-05,
+      "loss": 0.3776,
+      "step": 10690
+    },
+    {
+      "epoch": 2.6683291770573567,
+      "grad_norm": 5.394494533538818,
+      "learning_rate": 1.7333167082294267e-05,
+      "loss": 0.3714,
+      "step": 10700
+    },
+    {
+      "epoch": 2.6708229426433916,
+      "grad_norm": 4.6077141761779785,
+      "learning_rate": 1.733067331670823e-05,
+      "loss": 0.3991,
+      "step": 10710
+    },
+    {
+      "epoch": 2.6733167082294265,
+      "grad_norm": 9.343939781188965,
+      "learning_rate": 1.7328179551122194e-05,
+      "loss": 0.3998,
+      "step": 10720
+    },
+    {
+      "epoch": 2.6758104738154613,
+      "grad_norm": 5.4641032218933105,
+      "learning_rate": 1.732568578553616e-05,
+      "loss": 0.3978,
+      "step": 10730
+    },
+    {
+      "epoch": 2.678304239401496,
+      "grad_norm": 6.154629230499268,
+      "learning_rate": 1.7323192019950125e-05,
+      "loss": 0.3814,
+      "step": 10740
+    },
+    {
+      "epoch": 2.680798004987531,
+      "grad_norm": 8.188713073730469,
+      "learning_rate": 1.7320698254364092e-05,
+      "loss": 0.454,
+      "step": 10750
+    },
+    {
+      "epoch": 2.683291770573566,
+      "grad_norm": 6.351066589355469,
+      "learning_rate": 1.7318204488778055e-05,
+      "loss": 0.4385,
+      "step": 10760
+    },
+    {
+      "epoch": 2.6857855361596013,
+      "grad_norm": 5.4914069175720215,
+      "learning_rate": 1.731571072319202e-05,
+      "loss": 0.4687,
+      "step": 10770
+    },
+    {
+      "epoch": 2.688279301745636,
+      "grad_norm": 10.653285026550293,
+      "learning_rate": 1.7313216957605986e-05,
+      "loss": 0.3617,
+      "step": 10780
+    },
+    {
+      "epoch": 2.690773067331671,
+      "grad_norm": 5.647527694702148,
+      "learning_rate": 1.7310723192019953e-05,
+      "loss": 0.4058,
+      "step": 10790
+    },
+    {
+      "epoch": 2.693266832917706,
+      "grad_norm": 8.179327964782715,
+      "learning_rate": 1.7308229426433917e-05,
+      "loss": 0.4148,
+      "step": 10800
+    },
+    {
+      "epoch": 2.6957605985037407,
+      "grad_norm": 4.55966854095459,
+      "learning_rate": 1.7305735660847884e-05,
+      "loss": 0.3592,
+      "step": 10810
+    },
+    {
+      "epoch": 2.6982543640897756,
+      "grad_norm": 7.562872409820557,
+      "learning_rate": 1.7303241895261847e-05,
+      "loss": 0.3956,
+      "step": 10820
+    },
+    {
+      "epoch": 2.7007481296758105,
+      "grad_norm": 12.162755012512207,
+      "learning_rate": 1.7300748129675814e-05,
+      "loss": 0.3981,
+      "step": 10830
+    },
+    {
+      "epoch": 2.7032418952618453,
+      "grad_norm": 9.855378150939941,
+      "learning_rate": 1.7298254364089778e-05,
+      "loss": 0.4338,
+      "step": 10840
+    },
+    {
+      "epoch": 2.70573566084788,
+      "grad_norm": 4.713982582092285,
+      "learning_rate": 1.729576059850374e-05,
+      "loss": 0.3565,
+      "step": 10850
+    },
+    {
+      "epoch": 2.708229426433915,
+      "grad_norm": 6.571569919586182,
+      "learning_rate": 1.729326683291771e-05,
+      "loss": 0.3863,
+      "step": 10860
+    },
+    {
+      "epoch": 2.71072319201995,
+      "grad_norm": 4.679405212402344,
+      "learning_rate": 1.7290773067331672e-05,
+      "loss": 0.3811,
+      "step": 10870
+    },
+    {
+      "epoch": 2.713216957605985,
+      "grad_norm": 6.7202982902526855,
+      "learning_rate": 1.7288279301745636e-05,
+      "loss": 0.3924,
+      "step": 10880
+    },
+    {
+      "epoch": 2.71571072319202,
+      "grad_norm": 6.362400531768799,
+      "learning_rate": 1.7285785536159603e-05,
+      "loss": 0.3774,
+      "step": 10890
+    },
+    {
+      "epoch": 2.718204488778055,
+      "grad_norm": 5.451074600219727,
+      "learning_rate": 1.7283291770573566e-05,
+      "loss": 0.391,
+      "step": 10900
+    },
+    {
+      "epoch": 2.72069825436409,
+      "grad_norm": 4.991129398345947,
+      "learning_rate": 1.7280798004987533e-05,
+      "loss": 0.4007,
+      "step": 10910
+    },
+    {
+      "epoch": 2.7231920199501247,
+      "grad_norm": 6.635371208190918,
+      "learning_rate": 1.7278304239401497e-05,
+      "loss": 0.4801,
+      "step": 10920
+    },
+    {
+      "epoch": 2.7256857855361596,
+      "grad_norm": 7.7395501136779785,
+      "learning_rate": 1.727581047381546e-05,
+      "loss": 0.4322,
+      "step": 10930
+    },
+    {
+      "epoch": 2.7281795511221945,
+      "grad_norm": 13.381205558776855,
+      "learning_rate": 1.7273316708229428e-05,
+      "loss": 0.468,
+      "step": 10940
+    },
+    {
+      "epoch": 2.7306733167082293,
+      "grad_norm": 7.249978542327881,
+      "learning_rate": 1.727082294264339e-05,
+      "loss": 0.4083,
+      "step": 10950
+    },
+    {
+      "epoch": 2.733167082294264,
+      "grad_norm": 7.9330573081970215,
+      "learning_rate": 1.7268329177057358e-05,
+      "loss": 0.4786,
+      "step": 10960
+    },
+    {
+      "epoch": 2.7356608478802995,
+      "grad_norm": 4.235700607299805,
+      "learning_rate": 1.7265835411471322e-05,
+      "loss": 0.4408,
+      "step": 10970
+    },
+    {
+      "epoch": 2.7381546134663344,
+      "grad_norm": 10.445588111877441,
+      "learning_rate": 1.726334164588529e-05,
+      "loss": 0.4088,
+      "step": 10980
+    },
+    {
+      "epoch": 2.7406483790523692,
+      "grad_norm": 4.6287031173706055,
+      "learning_rate": 1.7260847880299253e-05,
+      "loss": 0.4159,
+      "step": 10990
+    },
+    {
+      "epoch": 2.743142144638404,
+      "grad_norm": 6.074006080627441,
+      "learning_rate": 1.725835411471322e-05,
+      "loss": 0.4362,
+      "step": 11000
+    },
+    {
+      "epoch": 2.745635910224439,
+      "grad_norm": 6.392845153808594,
+      "learning_rate": 1.7255860349127183e-05,
+      "loss": 0.4406,
+      "step": 11010
+    },
+    {
+      "epoch": 2.748129675810474,
+      "grad_norm": 4.722134590148926,
+      "learning_rate": 1.725336658354115e-05,
+      "loss": 0.4392,
+      "step": 11020
+    },
+    {
+      "epoch": 2.7506234413965087,
+      "grad_norm": 5.2035698890686035,
+      "learning_rate": 1.7250872817955114e-05,
+      "loss": 0.4408,
+      "step": 11030
+    },
+    {
+      "epoch": 2.7531172069825436,
+      "grad_norm": 5.009521007537842,
+      "learning_rate": 1.724837905236908e-05,
+      "loss": 0.3727,
+      "step": 11040
+    },
+    {
+      "epoch": 2.7556109725685785,
+      "grad_norm": 8.132833480834961,
+      "learning_rate": 1.7245885286783044e-05,
+      "loss": 0.4183,
+      "step": 11050
+    },
+    {
+      "epoch": 2.7581047381546133,
+      "grad_norm": 4.388547897338867,
+      "learning_rate": 1.7243391521197008e-05,
+      "loss": 0.3871,
+      "step": 11060
+    },
+    {
+      "epoch": 2.760598503740648,
+      "grad_norm": 5.839278221130371,
+      "learning_rate": 1.7240897755610975e-05,
+      "loss": 0.3665,
+      "step": 11070
+    },
+    {
+      "epoch": 2.763092269326683,
+      "grad_norm": 6.53806734085083,
+      "learning_rate": 1.723840399002494e-05,
+      "loss": 0.4331,
+      "step": 11080
+    },
+    {
+      "epoch": 2.765586034912718,
+      "grad_norm": 9.082895278930664,
+      "learning_rate": 1.7235910224438902e-05,
+      "loss": 0.379,
+      "step": 11090
+    },
+    {
+      "epoch": 2.7680798004987532,
+      "grad_norm": 4.956233978271484,
+      "learning_rate": 1.723341645885287e-05,
+      "loss": 0.4789,
+      "step": 11100
+    },
+    {
+      "epoch": 2.770573566084788,
+      "grad_norm": 4.803463935852051,
+      "learning_rate": 1.7230922693266833e-05,
+      "loss": 0.4266,
+      "step": 11110
+    },
+    {
+      "epoch": 2.773067331670823,
+      "grad_norm": 5.308382511138916,
+      "learning_rate": 1.72284289276808e-05,
+      "loss": 0.4347,
+      "step": 11120
+    },
+    {
+      "epoch": 2.775561097256858,
+      "grad_norm": 4.9697418212890625,
+      "learning_rate": 1.7225935162094763e-05,
+      "loss": 0.4041,
+      "step": 11130
+    },
+    {
+      "epoch": 2.7780548628428927,
+      "grad_norm": 5.6570281982421875,
+      "learning_rate": 1.722344139650873e-05,
+      "loss": 0.4769,
+      "step": 11140
+    },
+    {
+      "epoch": 2.7805486284289276,
+      "grad_norm": 5.522329330444336,
+      "learning_rate": 1.7220947630922694e-05,
+      "loss": 0.4414,
+      "step": 11150
+    },
+    {
+      "epoch": 2.7830423940149625,
+      "grad_norm": 6.803591728210449,
+      "learning_rate": 1.721845386533666e-05,
+      "loss": 0.4085,
+      "step": 11160
+    },
+    {
+      "epoch": 2.7855361596009978,
+      "grad_norm": 13.538893699645996,
+      "learning_rate": 1.7215960099750625e-05,
+      "loss": 0.4206,
+      "step": 11170
+    },
+    {
+      "epoch": 2.7880299251870326,
+      "grad_norm": 5.024692535400391,
+      "learning_rate": 1.7213466334164592e-05,
+      "loss": 0.4413,
+      "step": 11180
+    },
+    {
+      "epoch": 2.7905236907730675,
+      "grad_norm": 6.203422546386719,
+      "learning_rate": 1.7210972568578555e-05,
+      "loss": 0.3727,
+      "step": 11190
+    },
+    {
+      "epoch": 2.7930174563591024,
+      "grad_norm": 4.57298469543457,
+      "learning_rate": 1.7208478802992522e-05,
+      "loss": 0.393,
+      "step": 11200
+    },
+    {
+      "epoch": 2.7955112219451372,
+      "grad_norm": 5.918140888214111,
+      "learning_rate": 1.7205985037406486e-05,
+      "loss": 0.3846,
+      "step": 11210
+    },
+    {
+      "epoch": 2.798004987531172,
+      "grad_norm": 3.546912670135498,
+      "learning_rate": 1.720349127182045e-05,
+      "loss": 0.4229,
+      "step": 11220
+    },
+    {
+      "epoch": 2.800498753117207,
+      "grad_norm": 4.094278335571289,
+      "learning_rate": 1.7200997506234417e-05,
+      "loss": 0.3498,
+      "step": 11230
+    },
+    {
+      "epoch": 2.802992518703242,
+      "grad_norm": 5.546658039093018,
+      "learning_rate": 1.719850374064838e-05,
+      "loss": 0.4516,
+      "step": 11240
+    },
+    {
+      "epoch": 2.8054862842892767,
+      "grad_norm": 3.7501320838928223,
+      "learning_rate": 1.7196009975062347e-05,
+      "loss": 0.366,
+      "step": 11250
+    },
+    {
+      "epoch": 2.8079800498753116,
+      "grad_norm": 7.085123062133789,
+      "learning_rate": 1.719351620947631e-05,
+      "loss": 0.3878,
+      "step": 11260
+    },
+    {
+      "epoch": 2.8104738154613464,
+      "grad_norm": 5.683539390563965,
+      "learning_rate": 1.7191022443890274e-05,
+      "loss": 0.4439,
+      "step": 11270
+    },
+    {
+      "epoch": 2.8129675810473813,
+      "grad_norm": 7.982929229736328,
+      "learning_rate": 1.718852867830424e-05,
+      "loss": 0.426,
+      "step": 11280
+    },
+    {
+      "epoch": 2.815461346633416,
+      "grad_norm": 5.907116889953613,
+      "learning_rate": 1.7186034912718205e-05,
+      "loss": 0.4181,
+      "step": 11290
+    },
+    {
+      "epoch": 2.8179551122194515,
+      "grad_norm": 6.01471471786499,
+      "learning_rate": 1.718354114713217e-05,
+      "loss": 0.4148,
+      "step": 11300
+    },
+    {
+      "epoch": 2.8204488778054864,
+      "grad_norm": 4.76801872253418,
+      "learning_rate": 1.7181047381546136e-05,
+      "loss": 0.3851,
+      "step": 11310
+    },
+    {
+      "epoch": 2.8229426433915212,
+      "grad_norm": 5.2211785316467285,
+      "learning_rate": 1.71785536159601e-05,
+      "loss": 0.4307,
+      "step": 11320
+    },
+    {
+      "epoch": 2.825436408977556,
+      "grad_norm": 3.697915554046631,
+      "learning_rate": 1.7176059850374066e-05,
+      "loss": 0.4286,
+      "step": 11330
+    },
+    {
+      "epoch": 2.827930174563591,
+      "grad_norm": 4.701713562011719,
+      "learning_rate": 1.717356608478803e-05,
+      "loss": 0.4104,
+      "step": 11340
+    },
+    {
+      "epoch": 2.830423940149626,
+      "grad_norm": 7.1458964347839355,
+      "learning_rate": 1.7171072319201997e-05,
+      "loss": 0.4007,
+      "step": 11350
+    },
+    {
+      "epoch": 2.8329177057356607,
+      "grad_norm": 4.9811272621154785,
+      "learning_rate": 1.716857855361596e-05,
+      "loss": 0.3228,
+      "step": 11360
+    },
+    {
+      "epoch": 2.835411471321696,
+      "grad_norm": 7.584052085876465,
+      "learning_rate": 1.7166084788029928e-05,
+      "loss": 0.4144,
+      "step": 11370
+    },
+    {
+      "epoch": 2.837905236907731,
+      "grad_norm": 6.256438732147217,
+      "learning_rate": 1.716359102244389e-05,
+      "loss": 0.459,
+      "step": 11380
+    },
+    {
+      "epoch": 2.8403990024937658,
+      "grad_norm": 6.128487586975098,
+      "learning_rate": 1.7161097256857858e-05,
+      "loss": 0.4185,
+      "step": 11390
+    },
+    {
+      "epoch": 2.8428927680798006,
+      "grad_norm": 5.535672664642334,
+      "learning_rate": 1.7158603491271822e-05,
+      "loss": 0.451,
+      "step": 11400
+    },
+    {
+      "epoch": 2.8453865336658355,
+      "grad_norm": 5.653360843658447,
+      "learning_rate": 1.715610972568579e-05,
+      "loss": 0.5224,
+      "step": 11410
+    },
+    {
+      "epoch": 2.8478802992518704,
+      "grad_norm": 8.358339309692383,
+      "learning_rate": 1.7153615960099752e-05,
+      "loss": 0.431,
+      "step": 11420
+    },
+    {
+      "epoch": 2.8503740648379052,
+      "grad_norm": 5.042673587799072,
+      "learning_rate": 1.7151122194513716e-05,
+      "loss": 0.4588,
+      "step": 11430
+    },
+    {
+      "epoch": 2.85286783042394,
+      "grad_norm": 5.333296298980713,
+      "learning_rate": 1.7148628428927683e-05,
+      "loss": 0.3947,
+      "step": 11440
+    },
+    {
+      "epoch": 2.855361596009975,
+      "grad_norm": 4.792226791381836,
+      "learning_rate": 1.7146134663341647e-05,
+      "loss": 0.3985,
+      "step": 11450
+    },
+    {
+      "epoch": 2.85785536159601,
+      "grad_norm": 6.703670501708984,
+      "learning_rate": 1.7143640897755614e-05,
+      "loss": 0.4351,
+      "step": 11460
+    },
+    {
+      "epoch": 2.8603491271820447,
+      "grad_norm": 6.484237194061279,
+      "learning_rate": 1.7141147132169577e-05,
+      "loss": 0.4915,
+      "step": 11470
+    },
+    {
+      "epoch": 2.8628428927680796,
+      "grad_norm": 7.9989447593688965,
+      "learning_rate": 1.713865336658354e-05,
+      "loss": 0.3769,
+      "step": 11480
+    },
+    {
+      "epoch": 2.8653366583541144,
+      "grad_norm": 5.715572834014893,
+      "learning_rate": 1.7136159600997508e-05,
+      "loss": 0.5566,
+      "step": 11490
+    },
+    {
+      "epoch": 2.8678304239401498,
+      "grad_norm": 5.8361945152282715,
+      "learning_rate": 1.713366583541147e-05,
+      "loss": 0.3843,
+      "step": 11500
+    },
+    {
+      "epoch": 2.8703241895261846,
+      "grad_norm": 7.280886650085449,
+      "learning_rate": 1.713117206982544e-05,
+      "loss": 0.4158,
+      "step": 11510
+    },
+    {
+      "epoch": 2.8728179551122195,
+      "grad_norm": 6.559328556060791,
+      "learning_rate": 1.7128678304239402e-05,
+      "loss": 0.4373,
+      "step": 11520
+    },
+    {
+      "epoch": 2.8753117206982544,
+      "grad_norm": 8.497318267822266,
+      "learning_rate": 1.712618453865337e-05,
+      "loss": 0.4206,
+      "step": 11530
+    },
+    {
+      "epoch": 2.8778054862842892,
+      "grad_norm": 4.384303092956543,
+      "learning_rate": 1.7123690773067333e-05,
+      "loss": 0.3942,
+      "step": 11540
+    },
+    {
+      "epoch": 2.880299251870324,
+      "grad_norm": 6.503232002258301,
+      "learning_rate": 1.71211970074813e-05,
+      "loss": 0.5279,
+      "step": 11550
+    },
+    {
+      "epoch": 2.882793017456359,
+      "grad_norm": 6.880457878112793,
+      "learning_rate": 1.7118703241895263e-05,
+      "loss": 0.4544,
+      "step": 11560
+    },
+    {
+      "epoch": 2.8852867830423943,
+      "grad_norm": 5.289419174194336,
+      "learning_rate": 1.711620947630923e-05,
+      "loss": 0.414,
+      "step": 11570
+    },
+    {
+      "epoch": 2.887780548628429,
+      "grad_norm": 7.233415126800537,
+      "learning_rate": 1.7113715710723194e-05,
+      "loss": 0.4409,
+      "step": 11580
+    },
+    {
+      "epoch": 2.890274314214464,
+      "grad_norm": 5.534732818603516,
+      "learning_rate": 1.7111221945137158e-05,
+      "loss": 0.4652,
+      "step": 11590
+    },
+    {
+      "epoch": 2.892768079800499,
+      "grad_norm": 6.039757251739502,
+      "learning_rate": 1.7108728179551125e-05,
+      "loss": 0.4492,
+      "step": 11600
+    },
+    {
+      "epoch": 2.8952618453865338,
+      "grad_norm": 8.882451057434082,
+      "learning_rate": 1.7106234413965088e-05,
+      "loss": 0.3906,
+      "step": 11610
+    },
+    {
+      "epoch": 2.8977556109725686,
+      "grad_norm": 6.894459247589111,
+      "learning_rate": 1.7103740648379055e-05,
+      "loss": 0.4368,
+      "step": 11620
+    },
+    {
+      "epoch": 2.9002493765586035,
+      "grad_norm": 4.845560073852539,
+      "learning_rate": 1.710124688279302e-05,
+      "loss": 0.4496,
+      "step": 11630
+    },
+    {
+      "epoch": 2.9027431421446384,
+      "grad_norm": 4.972646713256836,
+      "learning_rate": 1.7098753117206982e-05,
+      "loss": 0.4763,
+      "step": 11640
+    },
+    {
+      "epoch": 2.9052369077306732,
+      "grad_norm": 5.399471759796143,
+      "learning_rate": 1.709625935162095e-05,
+      "loss": 0.4365,
+      "step": 11650
+    },
+    {
+      "epoch": 2.907730673316708,
+      "grad_norm": 5.46916389465332,
+      "learning_rate": 1.7093765586034913e-05,
+      "loss": 0.4315,
+      "step": 11660
+    },
+    {
+      "epoch": 2.910224438902743,
+      "grad_norm": 9.235302925109863,
+      "learning_rate": 1.7091271820448877e-05,
+      "loss": 0.4335,
+      "step": 11670
+    },
+    {
+      "epoch": 2.912718204488778,
+      "grad_norm": 3.6637065410614014,
+      "learning_rate": 1.7088778054862844e-05,
+      "loss": 0.4421,
+      "step": 11680
+    },
+    {
+      "epoch": 2.9152119700748127,
+      "grad_norm": 4.475642681121826,
+      "learning_rate": 1.7086284289276807e-05,
+      "loss": 0.3812,
+      "step": 11690
+    },
+    {
+      "epoch": 2.917705735660848,
+      "grad_norm": 4.71533203125,
+      "learning_rate": 1.7083790523690774e-05,
+      "loss": 0.3723,
+      "step": 11700
+    },
+    {
+      "epoch": 2.920199501246883,
+      "grad_norm": 9.815446853637695,
+      "learning_rate": 1.7081296758104738e-05,
+      "loss": 0.4587,
+      "step": 11710
+    },
+    {
+      "epoch": 2.9226932668329177,
+      "grad_norm": 5.071043968200684,
+      "learning_rate": 1.7078802992518705e-05,
+      "loss": 0.4285,
+      "step": 11720
+    },
+    {
+      "epoch": 2.9251870324189526,
+      "grad_norm": 4.660521507263184,
+      "learning_rate": 1.7076309226932672e-05,
+      "loss": 0.4849,
+      "step": 11730
+    },
+    {
+      "epoch": 2.9276807980049875,
+      "grad_norm": 4.910328388214111,
+      "learning_rate": 1.7073815461346636e-05,
+      "loss": 0.4813,
+      "step": 11740
+    },
+    {
+      "epoch": 2.9301745635910224,
+      "grad_norm": 5.175450801849365,
+      "learning_rate": 1.7071321695760603e-05,
+      "loss": 0.4067,
+      "step": 11750
+    },
+    {
+      "epoch": 2.932668329177057,
+      "grad_norm": 5.369016170501709,
+      "learning_rate": 1.7068827930174566e-05,
+      "loss": 0.4048,
+      "step": 11760
+    },
+    {
+      "epoch": 2.9351620947630925,
+      "grad_norm": 4.3544745445251465,
+      "learning_rate": 1.706633416458853e-05,
+      "loss": 0.4286,
+      "step": 11770
+    },
+    {
+      "epoch": 2.9376558603491274,
+      "grad_norm": 7.268121242523193,
+      "learning_rate": 1.7063840399002497e-05,
+      "loss": 0.3988,
+      "step": 11780
+    },
+    {
+      "epoch": 2.9401496259351623,
+      "grad_norm": 5.836695194244385,
+      "learning_rate": 1.706134663341646e-05,
+      "loss": 0.4187,
+      "step": 11790
+    },
+    {
+      "epoch": 2.942643391521197,
+      "grad_norm": 4.910001754760742,
+      "learning_rate": 1.7058852867830424e-05,
+      "loss": 0.4232,
+      "step": 11800
+    },
+    {
+      "epoch": 2.945137157107232,
+      "grad_norm": 3.5782833099365234,
+      "learning_rate": 1.705635910224439e-05,
+      "loss": 0.5034,
+      "step": 11810
+    },
+    {
+      "epoch": 2.947630922693267,
+      "grad_norm": 4.067619323730469,
+      "learning_rate": 1.7053865336658355e-05,
+      "loss": 0.3851,
+      "step": 11820
+    },
+    {
+      "epoch": 2.9501246882793017,
+      "grad_norm": 9.091952323913574,
+      "learning_rate": 1.705137157107232e-05,
+      "loss": 0.5536,
+      "step": 11830
+    },
+    {
+      "epoch": 2.9526184538653366,
+      "grad_norm": 7.382187366485596,
+      "learning_rate": 1.7048877805486285e-05,
+      "loss": 0.3732,
+      "step": 11840
+    },
+    {
+      "epoch": 2.9551122194513715,
+      "grad_norm": 5.961577415466309,
+      "learning_rate": 1.704638403990025e-05,
+      "loss": 0.3849,
+      "step": 11850
+    },
+    {
+      "epoch": 2.9576059850374063,
+      "grad_norm": 7.97136116027832,
+      "learning_rate": 1.7043890274314216e-05,
+      "loss": 0.4156,
+      "step": 11860
+    },
+    {
+      "epoch": 2.960099750623441,
+      "grad_norm": 14.47574234008789,
+      "learning_rate": 1.704139650872818e-05,
+      "loss": 0.4306,
+      "step": 11870
+    },
+    {
+      "epoch": 2.962593516209476,
+      "grad_norm": 6.207291603088379,
+      "learning_rate": 1.7038902743142146e-05,
+      "loss": 0.3749,
+      "step": 11880
+    },
+    {
+      "epoch": 2.965087281795511,
+      "grad_norm": 6.804027080535889,
+      "learning_rate": 1.703640897755611e-05,
+      "loss": 0.4808,
+      "step": 11890
+    },
+    {
+      "epoch": 2.9675810473815463,
+      "grad_norm": 7.510690689086914,
+      "learning_rate": 1.7033915211970077e-05,
+      "loss": 0.5506,
+      "step": 11900
+    },
+    {
+      "epoch": 2.970074812967581,
+      "grad_norm": 10.168417930603027,
+      "learning_rate": 1.703142144638404e-05,
+      "loss": 0.5405,
+      "step": 11910
+    },
+    {
+      "epoch": 2.972568578553616,
+      "grad_norm": 5.433992385864258,
+      "learning_rate": 1.7028927680798008e-05,
+      "loss": 0.4205,
+      "step": 11920
+    },
+    {
+      "epoch": 2.975062344139651,
+      "grad_norm": 5.19559907913208,
+      "learning_rate": 1.702643391521197e-05,
+      "loss": 0.3868,
+      "step": 11930
+    },
+    {
+      "epoch": 2.9775561097256857,
+      "grad_norm": 6.645895004272461,
+      "learning_rate": 1.702394014962594e-05,
+      "loss": 0.4322,
+      "step": 11940
+    },
+    {
+      "epoch": 2.9800498753117206,
+      "grad_norm": 5.916768550872803,
+      "learning_rate": 1.7021446384039902e-05,
+      "loss": 0.3847,
+      "step": 11950
+    },
+    {
+      "epoch": 2.9825436408977555,
+      "grad_norm": 4.953275203704834,
+      "learning_rate": 1.701895261845387e-05,
+      "loss": 0.4585,
+      "step": 11960
+    },
+    {
+      "epoch": 2.985037406483791,
+      "grad_norm": 3.9088947772979736,
+      "learning_rate": 1.7016458852867833e-05,
+      "loss": 0.3925,
+      "step": 11970
+    },
+    {
+      "epoch": 2.9875311720698257,
+      "grad_norm": 12.187833786010742,
+      "learning_rate": 1.7013965087281796e-05,
+      "loss": 0.5033,
+      "step": 11980
+    },
+    {
+      "epoch": 2.9900249376558605,
+      "grad_norm": 6.370995998382568,
+      "learning_rate": 1.7011471321695763e-05,
+      "loss": 0.3995,
+      "step": 11990
+    },
+    {
+      "epoch": 2.9925187032418954,
+      "grad_norm": 5.4293365478515625,
+      "learning_rate": 1.7008977556109727e-05,
+      "loss": 0.4607,
+      "step": 12000
+    },
+    {
+      "epoch": 2.9950124688279303,
+      "grad_norm": 10.32349967956543,
+      "learning_rate": 1.700648379052369e-05,
+      "loss": 0.4499,
+      "step": 12010
+    },
+    {
+      "epoch": 2.997506234413965,
+      "grad_norm": 4.519608497619629,
+      "learning_rate": 1.7003990024937657e-05,
+      "loss": 0.4193,
+      "step": 12020
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.505052089691162,
+      "learning_rate": 1.700149625935162e-05,
+      "loss": 0.3652,
+      "step": 12030
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.43368199467658997,
+      "eval_runtime": 67.1343,
+      "eval_samples_per_second": 14.94,
+      "eval_steps_per_second": 14.94,
+      "step": 12030
+    },
+    {
+      "epoch": 3.002493765586035,
+      "grad_norm": 6.454890727996826,
+      "learning_rate": 1.6999002493765588e-05,
+      "loss": 0.4374,
+      "step": 12040
+    },
+    {
+      "epoch": 3.0049875311720697,
+      "grad_norm": 7.071195125579834,
+      "learning_rate": 1.699650872817955e-05,
+      "loss": 0.4221,
+      "step": 12050
+    },
+    {
+      "epoch": 3.0074812967581046,
+      "grad_norm": 5.097235202789307,
+      "learning_rate": 1.6994014962593515e-05,
+      "loss": 0.4258,
+      "step": 12060
+    },
+    {
+      "epoch": 3.0099750623441395,
+      "grad_norm": 4.156216144561768,
+      "learning_rate": 1.6991521197007482e-05,
+      "loss": 0.3844,
+      "step": 12070
+    },
+    {
+      "epoch": 3.0124688279301743,
+      "grad_norm": 4.2227325439453125,
+      "learning_rate": 1.698902743142145e-05,
+      "loss": 0.3943,
+      "step": 12080
+    },
+    {
+      "epoch": 3.0149625935162097,
+      "grad_norm": 5.322434425354004,
+      "learning_rate": 1.6986533665835413e-05,
+      "loss": 0.4466,
+      "step": 12090
+    },
+    {
+      "epoch": 3.0174563591022445,
+      "grad_norm": 7.377932548522949,
+      "learning_rate": 1.698403990024938e-05,
+      "loss": 0.4465,
+      "step": 12100
+    },
+    {
+      "epoch": 3.0199501246882794,
+      "grad_norm": 4.9225263595581055,
+      "learning_rate": 1.6981546134663343e-05,
+      "loss": 0.389,
+      "step": 12110
+    },
+    {
+      "epoch": 3.0224438902743143,
+      "grad_norm": 4.286352157592773,
+      "learning_rate": 1.697905236907731e-05,
+      "loss": 0.3605,
+      "step": 12120
+    },
+    {
+      "epoch": 3.024937655860349,
+      "grad_norm": 7.271639347076416,
+      "learning_rate": 1.6976558603491274e-05,
+      "loss": 0.4093,
+      "step": 12130
+    },
+    {
+      "epoch": 3.027431421446384,
+      "grad_norm": 9.17022705078125,
+      "learning_rate": 1.6974064837905238e-05,
+      "loss": 0.4649,
+      "step": 12140
+    },
+    {
+      "epoch": 3.029925187032419,
+      "grad_norm": 11.9248046875,
+      "learning_rate": 1.6971571072319205e-05,
+      "loss": 0.4102,
+      "step": 12150
+    },
+    {
+      "epoch": 3.0324189526184537,
+      "grad_norm": 5.974640369415283,
+      "learning_rate": 1.696907730673317e-05,
+      "loss": 0.4345,
+      "step": 12160
+    },
+    {
+      "epoch": 3.0349127182044886,
+      "grad_norm": 7.954996109008789,
+      "learning_rate": 1.6966583541147132e-05,
+      "loss": 0.3358,
+      "step": 12170
+    },
+    {
+      "epoch": 3.037406483790524,
+      "grad_norm": 7.6409220695495605,
+      "learning_rate": 1.6964339152119704e-05,
+      "loss": 0.4094,
+      "step": 12180
+    },
+    {
+      "epoch": 3.039900249376559,
+      "grad_norm": 6.415146827697754,
+      "learning_rate": 1.6961845386533667e-05,
+      "loss": 0.408,
+      "step": 12190
+    },
+    {
+      "epoch": 3.0423940149625937,
+      "grad_norm": 6.253342151641846,
+      "learning_rate": 1.6959351620947634e-05,
+      "loss": 0.385,
+      "step": 12200
+    },
+    {
+      "epoch": 3.0448877805486285,
+      "grad_norm": 6.785040378570557,
+      "learning_rate": 1.6956857855361598e-05,
+      "loss": 0.3796,
+      "step": 12210
+    },
+    {
+      "epoch": 3.0473815461346634,
+      "grad_norm": 5.031459331512451,
+      "learning_rate": 1.6954364089775565e-05,
+      "loss": 0.3762,
+      "step": 12220
+    },
+    {
+      "epoch": 3.0498753117206983,
+      "grad_norm": 5.665307521820068,
+      "learning_rate": 1.695187032418953e-05,
+      "loss": 0.4444,
+      "step": 12230
+    },
+    {
+      "epoch": 3.052369077306733,
+      "grad_norm": 9.208979606628418,
+      "learning_rate": 1.6949376558603492e-05,
+      "loss": 0.41,
+      "step": 12240
+    },
+    {
+      "epoch": 3.054862842892768,
+      "grad_norm": 6.723043441772461,
+      "learning_rate": 1.694688279301746e-05,
+      "loss": 0.378,
+      "step": 12250
+    },
+    {
+      "epoch": 3.057356608478803,
+      "grad_norm": 9.06165599822998,
+      "learning_rate": 1.6944389027431423e-05,
+      "loss": 0.3957,
+      "step": 12260
+    },
+    {
+      "epoch": 3.0598503740648377,
+      "grad_norm": 6.359433650970459,
+      "learning_rate": 1.6941895261845386e-05,
+      "loss": 0.3632,
+      "step": 12270
+    },
+    {
+      "epoch": 3.0623441396508726,
+      "grad_norm": 4.539393901824951,
+      "learning_rate": 1.6939401496259353e-05,
+      "loss": 0.4039,
+      "step": 12280
+    },
+    {
+      "epoch": 3.064837905236908,
+      "grad_norm": 7.049586772918701,
+      "learning_rate": 1.6936907730673317e-05,
+      "loss": 0.4395,
+      "step": 12290
+    },
+    {
+      "epoch": 3.067331670822943,
+      "grad_norm": 7.9074835777282715,
+      "learning_rate": 1.6934413965087284e-05,
+      "loss": 0.4092,
+      "step": 12300
+    },
+    {
+      "epoch": 3.0698254364089776,
+      "grad_norm": 6.730893611907959,
+      "learning_rate": 1.6931920199501248e-05,
+      "loss": 0.4422,
+      "step": 12310
+    },
+    {
+      "epoch": 3.0723192019950125,
+      "grad_norm": 4.9922194480896,
+      "learning_rate": 1.692942643391521e-05,
+      "loss": 0.3308,
+      "step": 12320
+    },
+    {
+      "epoch": 3.0748129675810474,
+      "grad_norm": 7.608273506164551,
+      "learning_rate": 1.6926932668329178e-05,
+      "loss": 0.363,
+      "step": 12330
+    },
+    {
+      "epoch": 3.0773067331670823,
+      "grad_norm": 7.005294322967529,
+      "learning_rate": 1.6924438902743142e-05,
+      "loss": 0.4175,
+      "step": 12340
+    },
+    {
+      "epoch": 3.079800498753117,
+      "grad_norm": 7.616174221038818,
+      "learning_rate": 1.692194513715711e-05,
+      "loss": 0.3848,
+      "step": 12350
+    },
+    {
+      "epoch": 3.082294264339152,
+      "grad_norm": 5.1623921394348145,
+      "learning_rate": 1.6919451371571072e-05,
+      "loss": 0.4422,
+      "step": 12360
+    },
+    {
+      "epoch": 3.084788029925187,
+      "grad_norm": 7.74609899520874,
+      "learning_rate": 1.691695760598504e-05,
+      "loss": 0.3759,
+      "step": 12370
+    },
+    {
+      "epoch": 3.087281795511222,
+      "grad_norm": 5.895921230316162,
+      "learning_rate": 1.6914463840399003e-05,
+      "loss": 0.4285,
+      "step": 12380
+    },
+    {
+      "epoch": 3.089775561097257,
+      "grad_norm": 10.510361671447754,
+      "learning_rate": 1.691197007481297e-05,
+      "loss": 0.3517,
+      "step": 12390
+    },
+    {
+      "epoch": 3.092269326683292,
+      "grad_norm": 6.013721466064453,
+      "learning_rate": 1.6909476309226934e-05,
+      "loss": 0.3976,
+      "step": 12400
+    },
+    {
+      "epoch": 3.0947630922693268,
+      "grad_norm": 7.516636371612549,
+      "learning_rate": 1.69069825436409e-05,
+      "loss": 0.4454,
+      "step": 12410
+    },
+    {
+      "epoch": 3.0972568578553616,
+      "grad_norm": 4.560348033905029,
+      "learning_rate": 1.6904488778054864e-05,
+      "loss": 0.4975,
+      "step": 12420
+    },
+    {
+      "epoch": 3.0997506234413965,
+      "grad_norm": 5.868087291717529,
+      "learning_rate": 1.690199501246883e-05,
+      "loss": 0.4223,
+      "step": 12430
+    },
+    {
+      "epoch": 3.1022443890274314,
+      "grad_norm": 5.38186502456665,
+      "learning_rate": 1.6899501246882795e-05,
+      "loss": 0.3995,
+      "step": 12440
+    },
+    {
+      "epoch": 3.1047381546134662,
+      "grad_norm": 6.7590436935424805,
+      "learning_rate": 1.689700748129676e-05,
+      "loss": 0.3922,
+      "step": 12450
+    },
+    {
+      "epoch": 3.107231920199501,
+      "grad_norm": 6.269115447998047,
+      "learning_rate": 1.6894513715710726e-05,
+      "loss": 0.4457,
+      "step": 12460
+    },
+    {
+      "epoch": 3.109725685785536,
+      "grad_norm": 17.736539840698242,
+      "learning_rate": 1.689201995012469e-05,
+      "loss": 0.4161,
+      "step": 12470
+    },
+    {
+      "epoch": 3.112219451371571,
+      "grad_norm": 6.468410491943359,
+      "learning_rate": 1.6889526184538653e-05,
+      "loss": 0.4411,
+      "step": 12480
+    },
+    {
+      "epoch": 3.114713216957606,
+      "grad_norm": 11.51278305053711,
+      "learning_rate": 1.688703241895262e-05,
+      "loss": 0.4001,
+      "step": 12490
+    },
+    {
+      "epoch": 3.117206982543641,
+      "grad_norm": 7.40097188949585,
+      "learning_rate": 1.6884538653366583e-05,
+      "loss": 0.3631,
+      "step": 12500
+    },
+    {
+      "epoch": 3.119700748129676,
+      "grad_norm": 4.9506683349609375,
+      "learning_rate": 1.688204488778055e-05,
+      "loss": 0.4302,
+      "step": 12510
+    },
+    {
+      "epoch": 3.1221945137157108,
+      "grad_norm": 4.265248775482178,
+      "learning_rate": 1.6879551122194514e-05,
+      "loss": 0.3701,
+      "step": 12520
+    },
+    {
+      "epoch": 3.1246882793017456,
+      "grad_norm": 13.634464263916016,
+      "learning_rate": 1.687705735660848e-05,
+      "loss": 0.3986,
+      "step": 12530
+    },
+    {
+      "epoch": 3.1271820448877805,
+      "grad_norm": 6.742231369018555,
+      "learning_rate": 1.6874563591022445e-05,
+      "loss": 0.4203,
+      "step": 12540
+    },
+    {
+      "epoch": 3.1296758104738154,
+      "grad_norm": 5.28926944732666,
+      "learning_rate": 1.687206982543641e-05,
+      "loss": 0.4416,
+      "step": 12550
+    },
+    {
+      "epoch": 3.1321695760598502,
+      "grad_norm": 5.512608051300049,
+      "learning_rate": 1.6869576059850375e-05,
+      "loss": 0.4417,
+      "step": 12560
+    },
+    {
+      "epoch": 3.134663341645885,
+      "grad_norm": 6.725086212158203,
+      "learning_rate": 1.6867082294264342e-05,
+      "loss": 0.3786,
+      "step": 12570
+    },
+    {
+      "epoch": 3.1371571072319204,
+      "grad_norm": 6.855806827545166,
+      "learning_rate": 1.6864588528678306e-05,
+      "loss": 0.4306,
+      "step": 12580
+    },
+    {
+      "epoch": 3.1396508728179553,
+      "grad_norm": 8.445453643798828,
+      "learning_rate": 1.6862094763092273e-05,
+      "loss": 0.4489,
+      "step": 12590
+    },
+    {
+      "epoch": 3.14214463840399,
+      "grad_norm": 4.596224784851074,
+      "learning_rate": 1.6859600997506236e-05,
+      "loss": 0.3834,
+      "step": 12600
+    },
+    {
+      "epoch": 3.144638403990025,
+      "grad_norm": 6.923195838928223,
+      "learning_rate": 1.68571072319202e-05,
+      "loss": 0.4399,
+      "step": 12610
+    },
+    {
+      "epoch": 3.14713216957606,
+      "grad_norm": 3.491319417953491,
+      "learning_rate": 1.6854613466334167e-05,
+      "loss": 0.3944,
+      "step": 12620
+    },
+    {
+      "epoch": 3.1496259351620948,
+      "grad_norm": 6.328956127166748,
+      "learning_rate": 1.685211970074813e-05,
+      "loss": 0.3583,
+      "step": 12630
+    },
+    {
+      "epoch": 3.1521197007481296,
+      "grad_norm": 6.093251705169678,
+      "learning_rate": 1.6849625935162098e-05,
+      "loss": 0.3517,
+      "step": 12640
+    },
+    {
+      "epoch": 3.1546134663341645,
+      "grad_norm": 6.187097549438477,
+      "learning_rate": 1.684713216957606e-05,
+      "loss": 0.4233,
+      "step": 12650
+    },
+    {
+      "epoch": 3.1571072319201994,
+      "grad_norm": 7.890842437744141,
+      "learning_rate": 1.6844638403990025e-05,
+      "loss": 0.3321,
+      "step": 12660
+    },
+    {
+      "epoch": 3.1596009975062342,
+      "grad_norm": 6.837558746337891,
+      "learning_rate": 1.6842144638403992e-05,
+      "loss": 0.4426,
+      "step": 12670
+    },
+    {
+      "epoch": 3.162094763092269,
+      "grad_norm": 5.910398960113525,
+      "learning_rate": 1.6839650872817956e-05,
+      "loss": 0.4121,
+      "step": 12680
+    },
+    {
+      "epoch": 3.1645885286783044,
+      "grad_norm": 6.540918827056885,
+      "learning_rate": 1.683715710723192e-05,
+      "loss": 0.4617,
+      "step": 12690
+    },
+    {
+      "epoch": 3.1670822942643393,
+      "grad_norm": 7.381392955780029,
+      "learning_rate": 1.6834663341645886e-05,
+      "loss": 0.4303,
+      "step": 12700
+    },
+    {
+      "epoch": 3.169576059850374,
+      "grad_norm": 7.120436668395996,
+      "learning_rate": 1.683216957605985e-05,
+      "loss": 0.4569,
+      "step": 12710
+    },
+    {
+      "epoch": 3.172069825436409,
+      "grad_norm": 4.337961196899414,
+      "learning_rate": 1.6829675810473817e-05,
+      "loss": 0.3563,
+      "step": 12720
+    },
+    {
+      "epoch": 3.174563591022444,
+      "grad_norm": 5.855359077453613,
+      "learning_rate": 1.682718204488778e-05,
+      "loss": 0.5174,
+      "step": 12730
+    },
+    {
+      "epoch": 3.1770573566084788,
+      "grad_norm": 5.234665393829346,
+      "learning_rate": 1.6824688279301747e-05,
+      "loss": 0.3649,
+      "step": 12740
+    },
+    {
+      "epoch": 3.1795511221945136,
+      "grad_norm": 7.155577182769775,
+      "learning_rate": 1.6822194513715714e-05,
+      "loss": 0.4493,
+      "step": 12750
+    },
+    {
+      "epoch": 3.1820448877805485,
+      "grad_norm": 6.0398406982421875,
+      "learning_rate": 1.6819700748129678e-05,
+      "loss": 0.337,
+      "step": 12760
+    },
+    {
+      "epoch": 3.1845386533665834,
+      "grad_norm": 6.6076250076293945,
+      "learning_rate": 1.681720698254364e-05,
+      "loss": 0.3503,
+      "step": 12770
+    },
+    {
+      "epoch": 3.1870324189526187,
+      "grad_norm": 6.856815814971924,
+      "learning_rate": 1.681471321695761e-05,
+      "loss": 0.4076,
+      "step": 12780
+    },
+    {
+      "epoch": 3.1895261845386536,
+      "grad_norm": 9.046415328979492,
+      "learning_rate": 1.6812219451371572e-05,
+      "loss": 0.4621,
+      "step": 12790
+    },
+    {
+      "epoch": 3.1920199501246884,
+      "grad_norm": 6.603618144989014,
+      "learning_rate": 1.680972568578554e-05,
+      "loss": 0.4491,
+      "step": 12800
+    },
+    {
+      "epoch": 3.1945137157107233,
+      "grad_norm": 8.234539985656738,
+      "learning_rate": 1.6807231920199503e-05,
+      "loss": 0.355,
+      "step": 12810
+    },
+    {
+      "epoch": 3.197007481296758,
+      "grad_norm": 6.2086005210876465,
+      "learning_rate": 1.6804738154613467e-05,
+      "loss": 0.3679,
+      "step": 12820
+    },
+    {
+      "epoch": 3.199501246882793,
+      "grad_norm": 5.592929363250732,
+      "learning_rate": 1.6802244389027434e-05,
+      "loss": 0.3855,
+      "step": 12830
+    },
+    {
+      "epoch": 3.201995012468828,
+      "grad_norm": 5.385773658752441,
+      "learning_rate": 1.6799750623441397e-05,
+      "loss": 0.4592,
+      "step": 12840
+    },
+    {
+      "epoch": 3.2044887780548628,
+      "grad_norm": 4.771883010864258,
+      "learning_rate": 1.679725685785536e-05,
+      "loss": 0.3648,
+      "step": 12850
+    },
+    {
+      "epoch": 3.2069825436408976,
+      "grad_norm": 7.125466823577881,
+      "learning_rate": 1.6794763092269328e-05,
+      "loss": 0.4146,
+      "step": 12860
+    },
+    {
+      "epoch": 3.2094763092269325,
+      "grad_norm": 5.639003753662109,
+      "learning_rate": 1.679226932668329e-05,
+      "loss": 0.4068,
+      "step": 12870
+    },
+    {
+      "epoch": 3.2119700748129674,
+      "grad_norm": 4.13544225692749,
+      "learning_rate": 1.678977556109726e-05,
+      "loss": 0.3519,
+      "step": 12880
+    },
+    {
+      "epoch": 3.2144638403990027,
+      "grad_norm": 5.362157344818115,
+      "learning_rate": 1.6787281795511222e-05,
+      "loss": 0.4477,
+      "step": 12890
+    },
+    {
+      "epoch": 3.2169576059850375,
+      "grad_norm": 5.619980335235596,
+      "learning_rate": 1.678478802992519e-05,
+      "loss": 0.416,
+      "step": 12900
+    },
+    {
+      "epoch": 3.2194513715710724,
+      "grad_norm": 6.641584396362305,
+      "learning_rate": 1.6782294264339153e-05,
+      "loss": 0.3819,
+      "step": 12910
+    },
+    {
+      "epoch": 3.2219451371571073,
+      "grad_norm": 6.697845935821533,
+      "learning_rate": 1.677980049875312e-05,
+      "loss": 0.3809,
+      "step": 12920
+    },
+    {
+      "epoch": 3.224438902743142,
+      "grad_norm": 9.737817764282227,
+      "learning_rate": 1.6777306733167083e-05,
+      "loss": 0.4195,
+      "step": 12930
+    },
+    {
+      "epoch": 3.226932668329177,
+      "grad_norm": 3.7950658798217773,
+      "learning_rate": 1.677481296758105e-05,
+      "loss": 0.4657,
+      "step": 12940
+    },
+    {
+      "epoch": 3.229426433915212,
+      "grad_norm": 6.923771858215332,
+      "learning_rate": 1.6772319201995014e-05,
+      "loss": 0.3954,
+      "step": 12950
+    },
+    {
+      "epoch": 3.2319201995012468,
+      "grad_norm": 8.065861701965332,
+      "learning_rate": 1.676982543640898e-05,
+      "loss": 0.4534,
+      "step": 12960
+    },
+    {
+      "epoch": 3.2344139650872816,
+      "grad_norm": 7.0419769287109375,
+      "learning_rate": 1.6767331670822944e-05,
+      "loss": 0.3881,
+      "step": 12970
+    },
+    {
+      "epoch": 3.236907730673317,
+      "grad_norm": 5.294747352600098,
+      "learning_rate": 1.6764837905236908e-05,
+      "loss": 0.4064,
+      "step": 12980
+    },
+    {
+      "epoch": 3.239401496259352,
+      "grad_norm": 4.490688800811768,
+      "learning_rate": 1.6762344139650875e-05,
+      "loss": 0.4715,
+      "step": 12990
+    },
+    {
+      "epoch": 3.2418952618453867,
+      "grad_norm": 6.766694068908691,
+      "learning_rate": 1.675985037406484e-05,
+      "loss": 0.3785,
+      "step": 13000
+    },
+    {
+      "epoch": 3.2443890274314215,
+      "grad_norm": 4.411302089691162,
+      "learning_rate": 1.6757356608478806e-05,
+      "loss": 0.4342,
+      "step": 13010
+    },
+    {
+      "epoch": 3.2468827930174564,
+      "grad_norm": 5.128880500793457,
+      "learning_rate": 1.675486284289277e-05,
+      "loss": 0.4306,
+      "step": 13020
+    },
+    {
+      "epoch": 3.2493765586034913,
+      "grad_norm": 8.443227767944336,
+      "learning_rate": 1.6752369077306733e-05,
+      "loss": 0.3925,
+      "step": 13030
+    },
+    {
+      "epoch": 3.251870324189526,
+      "grad_norm": 5.713132858276367,
+      "learning_rate": 1.67498753117207e-05,
+      "loss": 0.377,
+      "step": 13040
+    },
+    {
+      "epoch": 3.254364089775561,
+      "grad_norm": 6.96317720413208,
+      "learning_rate": 1.6747381546134664e-05,
+      "loss": 0.3748,
+      "step": 13050
+    },
+    {
+      "epoch": 3.256857855361596,
+      "grad_norm": 5.586208343505859,
+      "learning_rate": 1.6744887780548627e-05,
+      "loss": 0.3898,
+      "step": 13060
+    },
+    {
+      "epoch": 3.2593516209476308,
+      "grad_norm": 8.379228591918945,
+      "learning_rate": 1.6742394014962594e-05,
+      "loss": 0.4784,
+      "step": 13070
+    },
+    {
+      "epoch": 3.2618453865336656,
+      "grad_norm": 8.025406837463379,
+      "learning_rate": 1.6739900249376558e-05,
+      "loss": 0.3903,
+      "step": 13080
+    },
+    {
+      "epoch": 3.264339152119701,
+      "grad_norm": 8.072983741760254,
+      "learning_rate": 1.6737406483790525e-05,
+      "loss": 0.4254,
+      "step": 13090
+    },
+    {
+      "epoch": 3.266832917705736,
+      "grad_norm": 5.819389343261719,
+      "learning_rate": 1.6734912718204492e-05,
+      "loss": 0.4922,
+      "step": 13100
+    },
+    {
+      "epoch": 3.2693266832917707,
+      "grad_norm": 6.674468040466309,
+      "learning_rate": 1.6732418952618455e-05,
+      "loss": 0.438,
+      "step": 13110
+    },
+    {
+      "epoch": 3.2718204488778055,
+      "grad_norm": 5.349725246429443,
+      "learning_rate": 1.6729925187032422e-05,
+      "loss": 0.3625,
+      "step": 13120
+    },
+    {
+      "epoch": 3.2743142144638404,
+      "grad_norm": 5.366141319274902,
+      "learning_rate": 1.6727431421446386e-05,
+      "loss": 0.3778,
+      "step": 13130
+    },
+    {
+      "epoch": 3.2768079800498753,
+      "grad_norm": 4.81095552444458,
+      "learning_rate": 1.6724937655860353e-05,
+      "loss": 0.3809,
+      "step": 13140
+    },
+    {
+      "epoch": 3.27930174563591,
+      "grad_norm": 5.11724328994751,
+      "learning_rate": 1.6722443890274317e-05,
+      "loss": 0.3866,
+      "step": 13150
+    },
+    {
+      "epoch": 3.281795511221945,
+      "grad_norm": 4.886108875274658,
+      "learning_rate": 1.671995012468828e-05,
+      "loss": 0.3952,
+      "step": 13160
+    },
+    {
+      "epoch": 3.28428927680798,
+      "grad_norm": 6.775887966156006,
+      "learning_rate": 1.6717456359102247e-05,
+      "loss": 0.4382,
+      "step": 13170
+    },
+    {
+      "epoch": 3.286783042394015,
+      "grad_norm": 4.1350579261779785,
+      "learning_rate": 1.671496259351621e-05,
+      "loss": 0.3817,
+      "step": 13180
+    },
+    {
+      "epoch": 3.28927680798005,
+      "grad_norm": 10.852806091308594,
+      "learning_rate": 1.6712468827930174e-05,
+      "loss": 0.4278,
+      "step": 13190
+    },
+    {
+      "epoch": 3.291770573566085,
+      "grad_norm": 4.8792877197265625,
+      "learning_rate": 1.670997506234414e-05,
+      "loss": 0.3502,
+      "step": 13200
+    },
+    {
+      "epoch": 3.29426433915212,
+      "grad_norm": 6.0365400314331055,
+      "learning_rate": 1.6707481296758105e-05,
+      "loss": 0.3906,
+      "step": 13210
+    },
+    {
+      "epoch": 3.2967581047381547,
+      "grad_norm": 8.025360107421875,
+      "learning_rate": 1.6704987531172072e-05,
+      "loss": 0.4406,
+      "step": 13220
+    },
+    {
+      "epoch": 3.2992518703241895,
+      "grad_norm": 7.672976970672607,
+      "learning_rate": 1.6702493765586036e-05,
+      "loss": 0.4107,
+      "step": 13230
+    },
+    {
+      "epoch": 3.3017456359102244,
+      "grad_norm": 4.216901779174805,
+      "learning_rate": 1.67e-05,
+      "loss": 0.3429,
+      "step": 13240
+    },
+    {
+      "epoch": 3.3042394014962593,
+      "grad_norm": 5.8930583000183105,
+      "learning_rate": 1.6697506234413966e-05,
+      "loss": 0.4131,
+      "step": 13250
+    },
+    {
+      "epoch": 3.306733167082294,
+      "grad_norm": 6.364641189575195,
+      "learning_rate": 1.669501246882793e-05,
+      "loss": 0.4126,
+      "step": 13260
+    },
+    {
+      "epoch": 3.309226932668329,
+      "grad_norm": 4.867658615112305,
+      "learning_rate": 1.6692518703241897e-05,
+      "loss": 0.3936,
+      "step": 13270
+    },
+    {
+      "epoch": 3.311720698254364,
+      "grad_norm": 4.648925304412842,
+      "learning_rate": 1.669002493765586e-05,
+      "loss": 0.356,
+      "step": 13280
+    },
+    {
+      "epoch": 3.314214463840399,
+      "grad_norm": 5.959649562835693,
+      "learning_rate": 1.6687531172069828e-05,
+      "loss": 0.4527,
+      "step": 13290
+    },
+    {
+      "epoch": 3.316708229426434,
+      "grad_norm": 5.857553005218506,
+      "learning_rate": 1.668503740648379e-05,
+      "loss": 0.3464,
+      "step": 13300
+    },
+    {
+      "epoch": 3.319201995012469,
+      "grad_norm": 7.903529167175293,
+      "learning_rate": 1.6682543640897758e-05,
+      "loss": 0.3983,
+      "step": 13310
+    },
+    {
+      "epoch": 3.321695760598504,
+      "grad_norm": 7.403318881988525,
+      "learning_rate": 1.6680049875311722e-05,
+      "loss": 0.39,
+      "step": 13320
+    },
+    {
+      "epoch": 3.3241895261845387,
+      "grad_norm": 4.094770431518555,
+      "learning_rate": 1.667755610972569e-05,
+      "loss": 0.3534,
+      "step": 13330
+    },
+    {
+      "epoch": 3.3266832917705735,
+      "grad_norm": 6.073254585266113,
+      "learning_rate": 1.6675062344139652e-05,
+      "loss": 0.4348,
+      "step": 13340
+    },
+    {
+      "epoch": 3.3291770573566084,
+      "grad_norm": 5.4333271980285645,
+      "learning_rate": 1.667256857855362e-05,
+      "loss": 0.4438,
+      "step": 13350
+    },
+    {
+      "epoch": 3.3316708229426433,
+      "grad_norm": 8.86145305633545,
+      "learning_rate": 1.6670074812967583e-05,
+      "loss": 0.4223,
+      "step": 13360
+    },
+    {
+      "epoch": 3.334164588528678,
+      "grad_norm": 4.68287467956543,
+      "learning_rate": 1.6667581047381547e-05,
+      "loss": 0.419,
+      "step": 13370
+    },
+    {
+      "epoch": 3.3366583541147135,
+      "grad_norm": 5.4317827224731445,
+      "learning_rate": 1.6665087281795514e-05,
+      "loss": 0.4438,
+      "step": 13380
+    },
+    {
+      "epoch": 3.3391521197007483,
+      "grad_norm": 6.64384651184082,
+      "learning_rate": 1.6662842892768082e-05,
+      "loss": 0.4448,
+      "step": 13390
+    },
+    {
+      "epoch": 3.341645885286783,
+      "grad_norm": 6.038212776184082,
+      "learning_rate": 1.6660349127182046e-05,
+      "loss": 0.4357,
+      "step": 13400
+    },
+    {
+      "epoch": 3.344139650872818,
+      "grad_norm": 7.981716156005859,
+      "learning_rate": 1.6657855361596013e-05,
+      "loss": 0.4141,
+      "step": 13410
+    },
+    {
+      "epoch": 3.346633416458853,
+      "grad_norm": 6.937891006469727,
+      "learning_rate": 1.6655361596009976e-05,
+      "loss": 0.426,
+      "step": 13420
+    },
+    {
+      "epoch": 3.349127182044888,
+      "grad_norm": 8.279930114746094,
+      "learning_rate": 1.6652867830423943e-05,
+      "loss": 0.3797,
+      "step": 13430
+    },
+    {
+      "epoch": 3.3516209476309227,
+      "grad_norm": 6.905846118927002,
+      "learning_rate": 1.6650374064837907e-05,
+      "loss": 0.3517,
+      "step": 13440
+    },
+    {
+      "epoch": 3.3541147132169575,
+      "grad_norm": 4.052714824676514,
+      "learning_rate": 1.664788029925187e-05,
+      "loss": 0.3177,
+      "step": 13450
+    },
+    {
+      "epoch": 3.3566084788029924,
+      "grad_norm": 6.537426948547363,
+      "learning_rate": 1.6645386533665837e-05,
+      "loss": 0.4482,
+      "step": 13460
+    },
+    {
+      "epoch": 3.3591022443890273,
+      "grad_norm": 5.7491455078125,
+      "learning_rate": 1.66428927680798e-05,
+      "loss": 0.385,
+      "step": 13470
+    },
+    {
+      "epoch": 3.361596009975062,
+      "grad_norm": 7.766640663146973,
+      "learning_rate": 1.6640399002493768e-05,
+      "loss": 0.3945,
+      "step": 13480
+    },
+    {
+      "epoch": 3.3640897755610975,
+      "grad_norm": 5.237179279327393,
+      "learning_rate": 1.663790523690773e-05,
+      "loss": 0.4368,
+      "step": 13490
+    },
+    {
+      "epoch": 3.3665835411471323,
+      "grad_norm": 7.344296932220459,
+      "learning_rate": 1.6635411471321695e-05,
+      "loss": 0.3749,
+      "step": 13500
+    },
+    {
+      "epoch": 3.369077306733167,
+      "grad_norm": 5.677723407745361,
+      "learning_rate": 1.6632917705735662e-05,
+      "loss": 0.41,
+      "step": 13510
+    },
+    {
+      "epoch": 3.371571072319202,
+      "grad_norm": 6.3155412673950195,
+      "learning_rate": 1.6630423940149626e-05,
+      "loss": 0.4143,
+      "step": 13520
+    },
+    {
+      "epoch": 3.374064837905237,
+      "grad_norm": 6.052978038787842,
+      "learning_rate": 1.6627930174563593e-05,
+      "loss": 0.4444,
+      "step": 13530
+    },
+    {
+      "epoch": 3.376558603491272,
+      "grad_norm": 5.203304767608643,
+      "learning_rate": 1.6625436408977557e-05,
+      "loss": 0.3819,
+      "step": 13540
+    },
+    {
+      "epoch": 3.3790523690773067,
+      "grad_norm": 7.154673099517822,
+      "learning_rate": 1.6622942643391524e-05,
+      "loss": 0.421,
+      "step": 13550
+    },
+    {
+      "epoch": 3.3815461346633415,
+      "grad_norm": 6.324401378631592,
+      "learning_rate": 1.6620448877805487e-05,
+      "loss": 0.4142,
+      "step": 13560
+    },
+    {
+      "epoch": 3.3840399002493764,
+      "grad_norm": 5.293837070465088,
+      "learning_rate": 1.6617955112219454e-05,
+      "loss": 0.341,
+      "step": 13570
+    },
+    {
+      "epoch": 3.3865336658354117,
+      "grad_norm": 6.022460460662842,
+      "learning_rate": 1.6615461346633418e-05,
+      "loss": 0.4016,
+      "step": 13580
+    },
+    {
+      "epoch": 3.3890274314214466,
+      "grad_norm": 3.799043655395508,
+      "learning_rate": 1.6612967581047385e-05,
+      "loss": 0.4037,
+      "step": 13590
+    },
+    {
+      "epoch": 3.3915211970074814,
+      "grad_norm": 5.069158554077148,
+      "learning_rate": 1.661047381546135e-05,
+      "loss": 0.3429,
+      "step": 13600
+    },
+    {
+      "epoch": 3.3940149625935163,
+      "grad_norm": 6.114314079284668,
+      "learning_rate": 1.6607980049875315e-05,
+      "loss": 0.4867,
+      "step": 13610
+    },
+    {
+      "epoch": 3.396508728179551,
+      "grad_norm": 5.830687046051025,
+      "learning_rate": 1.660548628428928e-05,
+      "loss": 0.3991,
+      "step": 13620
+    },
+    {
+      "epoch": 3.399002493765586,
+      "grad_norm": 10.9564847946167,
+      "learning_rate": 1.6602992518703243e-05,
+      "loss": 0.398,
+      "step": 13630
+    },
+    {
+      "epoch": 3.401496259351621,
+      "grad_norm": 6.169339179992676,
+      "learning_rate": 1.660049875311721e-05,
+      "loss": 0.3501,
+      "step": 13640
+    },
+    {
+      "epoch": 3.403990024937656,
+      "grad_norm": 4.284657001495361,
+      "learning_rate": 1.6598004987531173e-05,
+      "loss": 0.5303,
+      "step": 13650
+    },
+    {
+      "epoch": 3.4064837905236907,
+      "grad_norm": 5.175209999084473,
+      "learning_rate": 1.6595511221945137e-05,
+      "loss": 0.4677,
+      "step": 13660
+    },
+    {
+      "epoch": 3.4089775561097255,
+      "grad_norm": 4.939456462860107,
+      "learning_rate": 1.6593017456359104e-05,
+      "loss": 0.5759,
+      "step": 13670
+    },
+    {
+      "epoch": 3.4114713216957604,
+      "grad_norm": 6.589729309082031,
+      "learning_rate": 1.6590523690773067e-05,
+      "loss": 0.374,
+      "step": 13680
+    },
+    {
+      "epoch": 3.4139650872817953,
+      "grad_norm": 5.515322685241699,
+      "learning_rate": 1.6588029925187034e-05,
+      "loss": 0.3992,
+      "step": 13690
+    },
+    {
+      "epoch": 3.4164588528678306,
+      "grad_norm": 7.204263687133789,
+      "learning_rate": 1.6585536159600998e-05,
+      "loss": 0.443,
+      "step": 13700
+    },
+    {
+      "epoch": 3.4189526184538654,
+      "grad_norm": 5.2953877449035645,
+      "learning_rate": 1.6583042394014962e-05,
+      "loss": 0.3856,
+      "step": 13710
+    },
+    {
+      "epoch": 3.4214463840399003,
+      "grad_norm": 5.203874111175537,
+      "learning_rate": 1.658054862842893e-05,
+      "loss": 0.45,
+      "step": 13720
+    },
+    {
+      "epoch": 3.423940149625935,
+      "grad_norm": 5.980413436889648,
+      "learning_rate": 1.6578054862842892e-05,
+      "loss": 0.4115,
+      "step": 13730
+    },
+    {
+      "epoch": 3.42643391521197,
+      "grad_norm": 4.752535343170166,
+      "learning_rate": 1.657556109725686e-05,
+      "loss": 0.4392,
+      "step": 13740
+    },
+    {
+      "epoch": 3.428927680798005,
+      "grad_norm": 5.2310919761657715,
+      "learning_rate": 1.6573067331670823e-05,
+      "loss": 0.361,
+      "step": 13750
+    },
+    {
+      "epoch": 3.43142144638404,
+      "grad_norm": 9.963449478149414,
+      "learning_rate": 1.657057356608479e-05,
+      "loss": 0.5536,
+      "step": 13760
+    },
+    {
+      "epoch": 3.4339152119700747,
+      "grad_norm": 5.837104320526123,
+      "learning_rate": 1.6568079800498757e-05,
+      "loss": 0.3722,
+      "step": 13770
+    },
+    {
+      "epoch": 3.43640897755611,
+      "grad_norm": 7.110990524291992,
+      "learning_rate": 1.656558603491272e-05,
+      "loss": 0.4229,
+      "step": 13780
+    },
+    {
+      "epoch": 3.438902743142145,
+      "grad_norm": 5.253129005432129,
+      "learning_rate": 1.6563092269326684e-05,
+      "loss": 0.3619,
+      "step": 13790
+    },
+    {
+      "epoch": 3.4413965087281797,
+      "grad_norm": 7.7098517417907715,
+      "learning_rate": 1.656059850374065e-05,
+      "loss": 0.452,
+      "step": 13800
+    },
+    {
+      "epoch": 3.4438902743142146,
+      "grad_norm": 6.151096820831299,
+      "learning_rate": 1.6558104738154615e-05,
+      "loss": 0.4055,
+      "step": 13810
+    },
+    {
+      "epoch": 3.4463840399002494,
+      "grad_norm": 6.894789218902588,
+      "learning_rate": 1.6555610972568582e-05,
+      "loss": 0.3736,
+      "step": 13820
+    },
+    {
+      "epoch": 3.4488778054862843,
+      "grad_norm": 5.148143291473389,
+      "learning_rate": 1.6553117206982545e-05,
+      "loss": 0.462,
+      "step": 13830
+    },
+    {
+      "epoch": 3.451371571072319,
+      "grad_norm": 5.808004856109619,
+      "learning_rate": 1.655062344139651e-05,
+      "loss": 0.3789,
+      "step": 13840
+    },
+    {
+      "epoch": 3.453865336658354,
+      "grad_norm": 4.1240339279174805,
+      "learning_rate": 1.6548129675810476e-05,
+      "loss": 0.4021,
+      "step": 13850
+    },
+    {
+      "epoch": 3.456359102244389,
+      "grad_norm": 4.346762180328369,
+      "learning_rate": 1.654563591022444e-05,
+      "loss": 0.3446,
+      "step": 13860
+    },
+    {
+      "epoch": 3.458852867830424,
+      "grad_norm": 7.146320819854736,
+      "learning_rate": 1.6543142144638403e-05,
+      "loss": 0.395,
+      "step": 13870
+    },
+    {
+      "epoch": 3.4613466334164587,
+      "grad_norm": 5.127925395965576,
+      "learning_rate": 1.654064837905237e-05,
+      "loss": 0.3873,
+      "step": 13880
+    },
+    {
+      "epoch": 3.4638403990024935,
+      "grad_norm": 6.493956565856934,
+      "learning_rate": 1.6538154613466334e-05,
+      "loss": 0.4049,
+      "step": 13890
+    },
+    {
+      "epoch": 3.466334164588529,
+      "grad_norm": 7.458160400390625,
+      "learning_rate": 1.65356608478803e-05,
+      "loss": 0.3721,
+      "step": 13900
+    },
+    {
+      "epoch": 3.4688279301745637,
+      "grad_norm": 6.495090007781982,
+      "learning_rate": 1.6533167082294265e-05,
+      "loss": 0.4702,
+      "step": 13910
+    },
+    {
+      "epoch": 3.4713216957605986,
+      "grad_norm": 6.47518253326416,
+      "learning_rate": 1.653067331670823e-05,
+      "loss": 0.4566,
+      "step": 13920
+    },
+    {
+      "epoch": 3.4738154613466334,
+      "grad_norm": 5.553424835205078,
+      "learning_rate": 1.6528179551122195e-05,
+      "loss": 0.3799,
+      "step": 13930
+    },
+    {
+      "epoch": 3.4763092269326683,
+      "grad_norm": 5.419945240020752,
+      "learning_rate": 1.6525685785536162e-05,
+      "loss": 0.4483,
+      "step": 13940
+    },
+    {
+      "epoch": 3.478802992518703,
+      "grad_norm": 5.477248191833496,
+      "learning_rate": 1.6523192019950126e-05,
+      "loss": 0.326,
+      "step": 13950
+    },
+    {
+      "epoch": 3.481296758104738,
+      "grad_norm": 7.121800899505615,
+      "learning_rate": 1.6520698254364093e-05,
+      "loss": 0.3673,
+      "step": 13960
+    },
+    {
+      "epoch": 3.483790523690773,
+      "grad_norm": 6.306591510772705,
+      "learning_rate": 1.6518204488778056e-05,
+      "loss": 0.4262,
+      "step": 13970
+    },
+    {
+      "epoch": 3.4862842892768082,
+      "grad_norm": 8.00887393951416,
+      "learning_rate": 1.6515710723192023e-05,
+      "loss": 0.4345,
+      "step": 13980
+    },
+    {
+      "epoch": 3.488778054862843,
+      "grad_norm": 5.474499702453613,
+      "learning_rate": 1.6513216957605987e-05,
+      "loss": 0.4994,
+      "step": 13990
+    },
+    {
+      "epoch": 3.491271820448878,
+      "grad_norm": 6.708861827850342,
+      "learning_rate": 1.651072319201995e-05,
+      "loss": 0.3884,
+      "step": 14000
+    },
+    {
+      "epoch": 3.493765586034913,
+      "grad_norm": 4.315401077270508,
+      "learning_rate": 1.6508229426433918e-05,
+      "loss": 0.3686,
+      "step": 14010
+    },
+    {
+      "epoch": 3.4962593516209477,
+      "grad_norm": 5.418401718139648,
+      "learning_rate": 1.650573566084788e-05,
+      "loss": 0.41,
+      "step": 14020
+    },
+    {
+      "epoch": 3.4987531172069826,
+      "grad_norm": 5.228391647338867,
+      "learning_rate": 1.6503241895261848e-05,
+      "loss": 0.4117,
+      "step": 14030
+    },
+    {
+      "epoch": 3.5012468827930174,
+      "grad_norm": 5.129116535186768,
+      "learning_rate": 1.6500748129675812e-05,
+      "loss": 0.3595,
+      "step": 14040
+    },
+    {
+      "epoch": 3.5037406483790523,
+      "grad_norm": 3.8561089038848877,
+      "learning_rate": 1.6498254364089775e-05,
+      "loss": 0.3515,
+      "step": 14050
+    },
+    {
+      "epoch": 3.506234413965087,
+      "grad_norm": 7.8990278244018555,
+      "learning_rate": 1.6495760598503742e-05,
+      "loss": 0.3853,
+      "step": 14060
+    },
+    {
+      "epoch": 3.508728179551122,
+      "grad_norm": 8.074807167053223,
+      "learning_rate": 1.6493266832917706e-05,
+      "loss": 0.4222,
+      "step": 14070
+    },
+    {
+      "epoch": 3.511221945137157,
+      "grad_norm": 7.334726333618164,
+      "learning_rate": 1.649077306733167e-05,
+      "loss": 0.4673,
+      "step": 14080
+    },
+    {
+      "epoch": 3.5137157107231918,
+      "grad_norm": 5.140750885009766,
+      "learning_rate": 1.6488279301745637e-05,
+      "loss": 0.3758,
+      "step": 14090
+    },
+    {
+      "epoch": 3.516209476309227,
+      "grad_norm": 5.326815605163574,
+      "learning_rate": 1.64857855361596e-05,
+      "loss": 0.3841,
+      "step": 14100
+    },
+    {
+      "epoch": 3.518703241895262,
+      "grad_norm": 4.542361259460449,
+      "learning_rate": 1.6483291770573567e-05,
+      "loss": 0.4115,
+      "step": 14110
+    },
+    {
+      "epoch": 3.521197007481297,
+      "grad_norm": 9.180794715881348,
+      "learning_rate": 1.6480798004987534e-05,
+      "loss": 0.3703,
+      "step": 14120
+    },
+    {
+      "epoch": 3.5236907730673317,
+      "grad_norm": 6.60845422744751,
+      "learning_rate": 1.6478304239401498e-05,
+      "loss": 0.396,
+      "step": 14130
+    },
+    {
+      "epoch": 3.5261845386533666,
+      "grad_norm": 13.184927940368652,
+      "learning_rate": 1.6475810473815465e-05,
+      "loss": 0.444,
+      "step": 14140
+    },
+    {
+      "epoch": 3.5286783042394014,
+      "grad_norm": 8.303811073303223,
+      "learning_rate": 1.647331670822943e-05,
+      "loss": 0.3708,
+      "step": 14150
+    },
+    {
+      "epoch": 3.5311720698254363,
+      "grad_norm": 7.675179958343506,
+      "learning_rate": 1.6470822942643392e-05,
+      "loss": 0.4134,
+      "step": 14160
+    },
+    {
+      "epoch": 3.533665835411471,
+      "grad_norm": 7.668274879455566,
+      "learning_rate": 1.646832917705736e-05,
+      "loss": 0.3957,
+      "step": 14170
+    },
+    {
+      "epoch": 3.5361596009975065,
+      "grad_norm": 19.888633728027344,
+      "learning_rate": 1.6465835411471323e-05,
+      "loss": 0.5228,
+      "step": 14180
+    },
+    {
+      "epoch": 3.5386533665835413,
+      "grad_norm": 6.9068474769592285,
+      "learning_rate": 1.646334164588529e-05,
+      "loss": 0.381,
+      "step": 14190
+    },
+    {
+      "epoch": 3.541147132169576,
+      "grad_norm": 9.011292457580566,
+      "learning_rate": 1.6460847880299253e-05,
+      "loss": 0.4341,
+      "step": 14200
+    },
+    {
+      "epoch": 3.543640897755611,
+      "grad_norm": 7.827251434326172,
+      "learning_rate": 1.6458354114713217e-05,
+      "loss": 0.4266,
+      "step": 14210
+    },
+    {
+      "epoch": 3.546134663341646,
+      "grad_norm": 7.010087966918945,
+      "learning_rate": 1.6455860349127184e-05,
+      "loss": 0.4038,
+      "step": 14220
+    },
+    {
+      "epoch": 3.548628428927681,
+      "grad_norm": 7.390041828155518,
+      "learning_rate": 1.6453366583541148e-05,
+      "loss": 0.4247,
+      "step": 14230
+    },
+    {
+      "epoch": 3.5511221945137157,
+      "grad_norm": 7.997394561767578,
+      "learning_rate": 1.645087281795511e-05,
+      "loss": 0.4505,
+      "step": 14240
+    },
+    {
+      "epoch": 3.5536159600997506,
+      "grad_norm": 7.814496040344238,
+      "learning_rate": 1.6448379052369078e-05,
+      "loss": 0.4635,
+      "step": 14250
+    },
+    {
+      "epoch": 3.5561097256857854,
+      "grad_norm": 4.237307548522949,
+      "learning_rate": 1.6445885286783042e-05,
+      "loss": 0.3517,
+      "step": 14260
+    },
+    {
+      "epoch": 3.5586034912718203,
+      "grad_norm": 5.595398426055908,
+      "learning_rate": 1.644339152119701e-05,
+      "loss": 0.3995,
+      "step": 14270
+    },
+    {
+      "epoch": 3.561097256857855,
+      "grad_norm": 6.063884258270264,
+      "learning_rate": 1.6440897755610972e-05,
+      "loss": 0.3919,
+      "step": 14280
+    },
+    {
+      "epoch": 3.56359102244389,
+      "grad_norm": 16.324724197387695,
+      "learning_rate": 1.643840399002494e-05,
+      "loss": 0.4988,
+      "step": 14290
+    },
+    {
+      "epoch": 3.5660847880299253,
+      "grad_norm": 5.327268600463867,
+      "learning_rate": 1.6435910224438903e-05,
+      "loss": 0.4688,
+      "step": 14300
+    },
+    {
+      "epoch": 3.56857855361596,
+      "grad_norm": 9.353160858154297,
+      "learning_rate": 1.643341645885287e-05,
+      "loss": 0.56,
+      "step": 14310
+    },
+    {
+      "epoch": 3.571072319201995,
+      "grad_norm": 4.720142364501953,
+      "learning_rate": 1.6430922693266834e-05,
+      "loss": 0.4159,
+      "step": 14320
+    },
+    {
+      "epoch": 3.57356608478803,
+      "grad_norm": 4.005600452423096,
+      "learning_rate": 1.64284289276808e-05,
+      "loss": 0.3677,
+      "step": 14330
+    },
+    {
+      "epoch": 3.576059850374065,
+      "grad_norm": 9.215933799743652,
+      "learning_rate": 1.6425935162094764e-05,
+      "loss": 0.432,
+      "step": 14340
+    },
+    {
+      "epoch": 3.5785536159600997,
+      "grad_norm": 4.696425437927246,
+      "learning_rate": 1.642344139650873e-05,
+      "loss": 0.3727,
+      "step": 14350
+    },
+    {
+      "epoch": 3.5810473815461346,
+      "grad_norm": 4.839138984680176,
+      "learning_rate": 1.6420947630922695e-05,
+      "loss": 0.4013,
+      "step": 14360
+    },
+    {
+      "epoch": 3.5835411471321694,
+      "grad_norm": 7.069340705871582,
+      "learning_rate": 1.641845386533666e-05,
+      "loss": 0.3633,
+      "step": 14370
+    },
+    {
+      "epoch": 3.5860349127182047,
+      "grad_norm": 5.56005334854126,
+      "learning_rate": 1.6415960099750626e-05,
+      "loss": 0.3376,
+      "step": 14380
+    },
+    {
+      "epoch": 3.5885286783042396,
+      "grad_norm": 5.112822532653809,
+      "learning_rate": 1.641346633416459e-05,
+      "loss": 0.4179,
+      "step": 14390
+    },
+    {
+      "epoch": 3.5910224438902745,
+      "grad_norm": 6.113282203674316,
+      "learning_rate": 1.6410972568578556e-05,
+      "loss": 0.4436,
+      "step": 14400
+    },
+    {
+      "epoch": 3.5935162094763093,
+      "grad_norm": 4.90158748626709,
+      "learning_rate": 1.640847880299252e-05,
+      "loss": 0.4283,
+      "step": 14410
+    },
+    {
+      "epoch": 3.596009975062344,
+      "grad_norm": 6.325465679168701,
+      "learning_rate": 1.6405985037406483e-05,
+      "loss": 0.4857,
+      "step": 14420
+    },
+    {
+      "epoch": 3.598503740648379,
+      "grad_norm": 8.354521751403809,
+      "learning_rate": 1.640349127182045e-05,
+      "loss": 0.4173,
+      "step": 14430
+    },
+    {
+      "epoch": 3.600997506234414,
+      "grad_norm": 5.772945880889893,
+      "learning_rate": 1.6400997506234414e-05,
+      "loss": 0.4367,
+      "step": 14440
+    },
+    {
+      "epoch": 3.603491271820449,
+      "grad_norm": 5.8793864250183105,
+      "learning_rate": 1.6398503740648378e-05,
+      "loss": 0.351,
+      "step": 14450
+    },
+    {
+      "epoch": 3.6059850374064837,
+      "grad_norm": 4.758648872375488,
+      "learning_rate": 1.6396009975062345e-05,
+      "loss": 0.4236,
+      "step": 14460
+    },
+    {
+      "epoch": 3.6084788029925186,
+      "grad_norm": 9.100720405578613,
+      "learning_rate": 1.639351620947631e-05,
+      "loss": 0.3892,
+      "step": 14470
+    },
+    {
+      "epoch": 3.6109725685785534,
+      "grad_norm": 5.36121129989624,
+      "learning_rate": 1.6391022443890275e-05,
+      "loss": 0.4256,
+      "step": 14480
+    },
+    {
+      "epoch": 3.6134663341645883,
+      "grad_norm": 6.1626505851745605,
+      "learning_rate": 1.6388528678304242e-05,
+      "loss": 0.4083,
+      "step": 14490
+    },
+    {
+      "epoch": 3.6159600997506236,
+      "grad_norm": 4.670628070831299,
+      "learning_rate": 1.6386034912718206e-05,
+      "loss": 0.4056,
+      "step": 14500
+    },
+    {
+      "epoch": 3.6184538653366585,
+      "grad_norm": 5.522136211395264,
+      "learning_rate": 1.6383541147132173e-05,
+      "loss": 0.4228,
+      "step": 14510
+    },
+    {
+      "epoch": 3.6209476309226933,
+      "grad_norm": 4.898807048797607,
+      "learning_rate": 1.6381047381546137e-05,
+      "loss": 0.3984,
+      "step": 14520
+    },
+    {
+      "epoch": 3.623441396508728,
+      "grad_norm": 6.272655963897705,
+      "learning_rate": 1.6378553615960104e-05,
+      "loss": 0.4146,
+      "step": 14530
+    },
+    {
+      "epoch": 3.625935162094763,
+      "grad_norm": 5.485193729400635,
+      "learning_rate": 1.6376059850374067e-05,
+      "loss": 0.3955,
+      "step": 14540
+    },
+    {
+      "epoch": 3.628428927680798,
+      "grad_norm": 5.901163578033447,
+      "learning_rate": 1.637356608478803e-05,
+      "loss": 0.4093,
+      "step": 14550
+    },
+    {
+      "epoch": 3.630922693266833,
+      "grad_norm": 7.169445991516113,
+      "learning_rate": 1.6371072319201998e-05,
+      "loss": 0.4289,
+      "step": 14560
+    },
+    {
+      "epoch": 3.6334164588528677,
+      "grad_norm": 7.727000713348389,
+      "learning_rate": 1.636857855361596e-05,
+      "loss": 0.4101,
+      "step": 14570
+    },
+    {
+      "epoch": 3.635910224438903,
+      "grad_norm": 5.395565509796143,
+      "learning_rate": 1.6366084788029925e-05,
+      "loss": 0.4015,
+      "step": 14580
+    },
+    {
+      "epoch": 3.638403990024938,
+      "grad_norm": 8.575839042663574,
+      "learning_rate": 1.6363591022443892e-05,
+      "loss": 0.4276,
+      "step": 14590
+    },
+    {
+      "epoch": 3.6408977556109727,
+      "grad_norm": 6.496227264404297,
+      "learning_rate": 1.6361097256857856e-05,
+      "loss": 0.3814,
+      "step": 14600
+    },
+    {
+      "epoch": 3.6433915211970076,
+      "grad_norm": 5.557234764099121,
+      "learning_rate": 1.6358603491271823e-05,
+      "loss": 0.3926,
+      "step": 14610
+    },
+    {
+      "epoch": 3.6458852867830425,
+      "grad_norm": 9.716835021972656,
+      "learning_rate": 1.6356109725685786e-05,
+      "loss": 0.3903,
+      "step": 14620
+    },
+    {
+      "epoch": 3.6483790523690773,
+      "grad_norm": 4.228976726531982,
+      "learning_rate": 1.635361596009975e-05,
+      "loss": 0.4135,
+      "step": 14630
+    },
+    {
+      "epoch": 3.650872817955112,
+      "grad_norm": 9.507262229919434,
+      "learning_rate": 1.6351122194513717e-05,
+      "loss": 0.4225,
+      "step": 14640
+    },
+    {
+      "epoch": 3.653366583541147,
+      "grad_norm": 8.070616722106934,
+      "learning_rate": 1.634862842892768e-05,
+      "loss": 0.3617,
+      "step": 14650
+    },
+    {
+      "epoch": 3.655860349127182,
+      "grad_norm": 7.822412490844727,
+      "learning_rate": 1.6346134663341647e-05,
+      "loss": 0.4306,
+      "step": 14660
+    },
+    {
+      "epoch": 3.658354114713217,
+      "grad_norm": 6.324841022491455,
+      "learning_rate": 1.634364089775561e-05,
+      "loss": 0.3662,
+      "step": 14670
+    },
+    {
+      "epoch": 3.6608478802992517,
+      "grad_norm": 4.228864669799805,
+      "learning_rate": 1.6341147132169578e-05,
+      "loss": 0.3822,
+      "step": 14680
+    },
+    {
+      "epoch": 3.6633416458852865,
+      "grad_norm": 6.306748867034912,
+      "learning_rate": 1.6338653366583542e-05,
+      "loss": 0.4123,
+      "step": 14690
+    },
+    {
+      "epoch": 3.665835411471322,
+      "grad_norm": 6.977389812469482,
+      "learning_rate": 1.633615960099751e-05,
+      "loss": 0.4305,
+      "step": 14700
+    },
+    {
+      "epoch": 3.6683291770573567,
+      "grad_norm": 5.803625583648682,
+      "learning_rate": 1.6333665835411472e-05,
+      "loss": 0.4697,
+      "step": 14710
+    },
+    {
+      "epoch": 3.6708229426433916,
+      "grad_norm": 5.12143087387085,
+      "learning_rate": 1.633117206982544e-05,
+      "loss": 0.3995,
+      "step": 14720
+    },
+    {
+      "epoch": 3.6733167082294265,
+      "grad_norm": 4.334492206573486,
+      "learning_rate": 1.6328678304239403e-05,
+      "loss": 0.4086,
+      "step": 14730
+    },
+    {
+      "epoch": 3.6758104738154613,
+      "grad_norm": 4.964009761810303,
+      "learning_rate": 1.6326184538653367e-05,
+      "loss": 0.3887,
+      "step": 14740
+    },
+    {
+      "epoch": 3.678304239401496,
+      "grad_norm": 5.872185707092285,
+      "learning_rate": 1.6323690773067334e-05,
+      "loss": 0.4069,
+      "step": 14750
+    },
+    {
+      "epoch": 3.680798004987531,
+      "grad_norm": 5.125484466552734,
+      "learning_rate": 1.6321197007481297e-05,
+      "loss": 0.4077,
+      "step": 14760
+    },
+    {
+      "epoch": 3.683291770573566,
+      "grad_norm": 6.067224502563477,
+      "learning_rate": 1.6318703241895264e-05,
+      "loss": 0.4064,
+      "step": 14770
+    },
+    {
+      "epoch": 3.6857855361596013,
+      "grad_norm": 5.0722455978393555,
+      "learning_rate": 1.6316209476309228e-05,
+      "loss": 0.3666,
+      "step": 14780
+    },
+    {
+      "epoch": 3.688279301745636,
+      "grad_norm": 8.505644798278809,
+      "learning_rate": 1.631371571072319e-05,
+      "loss": 0.4493,
+      "step": 14790
+    },
+    {
+      "epoch": 3.690773067331671,
+      "grad_norm": 4.582368850708008,
+      "learning_rate": 1.631122194513716e-05,
+      "loss": 0.4366,
+      "step": 14800
+    },
+    {
+      "epoch": 3.693266832917706,
+      "grad_norm": 8.571370124816895,
+      "learning_rate": 1.6308728179551122e-05,
+      "loss": 0.3407,
+      "step": 14810
+    },
+    {
+      "epoch": 3.6957605985037407,
+      "grad_norm": 4.467719554901123,
+      "learning_rate": 1.630623441396509e-05,
+      "loss": 0.4311,
+      "step": 14820
+    },
+    {
+      "epoch": 3.6982543640897756,
+      "grad_norm": 5.480850696563721,
+      "learning_rate": 1.6303740648379053e-05,
+      "loss": 0.3438,
+      "step": 14830
+    },
+    {
+      "epoch": 3.7007481296758105,
+      "grad_norm": 4.664899826049805,
+      "learning_rate": 1.630124688279302e-05,
+      "loss": 0.3888,
+      "step": 14840
+    },
+    {
+      "epoch": 3.7032418952618453,
+      "grad_norm": 5.578117370605469,
+      "learning_rate": 1.6298753117206983e-05,
+      "loss": 0.3614,
+      "step": 14850
+    },
+    {
+      "epoch": 3.70573566084788,
+      "grad_norm": 6.581024646759033,
+      "learning_rate": 1.629625935162095e-05,
+      "loss": 0.5117,
+      "step": 14860
+    },
+    {
+      "epoch": 3.708229426433915,
+      "grad_norm": 7.885854721069336,
+      "learning_rate": 1.6293765586034914e-05,
+      "loss": 0.3758,
+      "step": 14870
+    },
+    {
+      "epoch": 3.71072319201995,
+      "grad_norm": 6.261788368225098,
+      "learning_rate": 1.629127182044888e-05,
+      "loss": 0.4068,
+      "step": 14880
+    },
+    {
+      "epoch": 3.713216957605985,
+      "grad_norm": 5.979361057281494,
+      "learning_rate": 1.6288778054862845e-05,
+      "loss": 0.3442,
+      "step": 14890
+    },
+    {
+      "epoch": 3.71571072319202,
+      "grad_norm": 6.158860683441162,
+      "learning_rate": 1.628628428927681e-05,
+      "loss": 0.4162,
+      "step": 14900
+    },
+    {
+      "epoch": 3.718204488778055,
+      "grad_norm": 6.998167991638184,
+      "learning_rate": 1.6283790523690775e-05,
+      "loss": 0.4444,
+      "step": 14910
+    },
+    {
+      "epoch": 3.72069825436409,
+      "grad_norm": 4.483408451080322,
+      "learning_rate": 1.628129675810474e-05,
+      "loss": 0.4146,
+      "step": 14920
+    },
+    {
+      "epoch": 3.7231920199501247,
+      "grad_norm": 7.948692321777344,
+      "learning_rate": 1.6278802992518706e-05,
+      "loss": 0.4934,
+      "step": 14930
+    },
+    {
+      "epoch": 3.7256857855361596,
+      "grad_norm": 9.365506172180176,
+      "learning_rate": 1.627630922693267e-05,
+      "loss": 0.3799,
+      "step": 14940
+    },
+    {
+      "epoch": 3.7281795511221945,
+      "grad_norm": 4.377537727355957,
+      "learning_rate": 1.6273815461346633e-05,
+      "loss": 0.3687,
+      "step": 14950
+    },
+    {
+      "epoch": 3.7306733167082293,
+      "grad_norm": 14.354135513305664,
+      "learning_rate": 1.62713216957606e-05,
+      "loss": 0.4541,
+      "step": 14960
+    },
+    {
+      "epoch": 3.733167082294264,
+      "grad_norm": 5.008823871612549,
+      "learning_rate": 1.6268827930174564e-05,
+      "loss": 0.46,
+      "step": 14970
+    },
+    {
+      "epoch": 3.7356608478802995,
+      "grad_norm": 6.697189807891846,
+      "learning_rate": 1.626633416458853e-05,
+      "loss": 0.3777,
+      "step": 14980
+    },
+    {
+      "epoch": 3.7381546134663344,
+      "grad_norm": 6.032017707824707,
+      "learning_rate": 1.6263840399002494e-05,
+      "loss": 0.3748,
+      "step": 14990
+    },
+    {
+      "epoch": 3.7406483790523692,
+      "grad_norm": 5.817931652069092,
+      "learning_rate": 1.6261346633416458e-05,
+      "loss": 0.4145,
+      "step": 15000
+    },
+    {
+      "epoch": 3.743142144638404,
+      "grad_norm": 5.026968479156494,
+      "learning_rate": 1.6258852867830425e-05,
+      "loss": 0.3745,
+      "step": 15010
+    },
+    {
+      "epoch": 3.745635910224439,
+      "grad_norm": 4.957605361938477,
+      "learning_rate": 1.625635910224439e-05,
+      "loss": 0.427,
+      "step": 15020
+    },
+    {
+      "epoch": 3.748129675810474,
+      "grad_norm": 6.634207248687744,
+      "learning_rate": 1.6253865336658355e-05,
+      "loss": 0.3864,
+      "step": 15030
+    },
+    {
+      "epoch": 3.7506234413965087,
+      "grad_norm": 5.489914894104004,
+      "learning_rate": 1.625137157107232e-05,
+      "loss": 0.4067,
+      "step": 15040
+    },
+    {
+      "epoch": 3.7531172069825436,
+      "grad_norm": 4.863682270050049,
+      "learning_rate": 1.6248877805486286e-05,
+      "loss": 0.3701,
+      "step": 15050
+    },
+    {
+      "epoch": 3.7556109725685785,
+      "grad_norm": 5.958093643188477,
+      "learning_rate": 1.6246384039900253e-05,
+      "loss": 0.5015,
+      "step": 15060
+    },
+    {
+      "epoch": 3.7581047381546133,
+      "grad_norm": 6.1467084884643555,
+      "learning_rate": 1.6243890274314217e-05,
+      "loss": 0.3989,
+      "step": 15070
+    },
+    {
+      "epoch": 3.760598503740648,
+      "grad_norm": 5.685709476470947,
+      "learning_rate": 1.624139650872818e-05,
+      "loss": 0.4924,
+      "step": 15080
+    },
+    {
+      "epoch": 3.763092269326683,
+      "grad_norm": 5.435773849487305,
+      "learning_rate": 1.6238902743142147e-05,
+      "loss": 0.3388,
+      "step": 15090
+    },
+    {
+      "epoch": 3.765586034912718,
+      "grad_norm": 4.414902687072754,
+      "learning_rate": 1.623640897755611e-05,
+      "loss": 0.4236,
+      "step": 15100
+    },
+    {
+      "epoch": 3.7680798004987532,
+      "grad_norm": 4.539722919464111,
+      "learning_rate": 1.6233915211970078e-05,
+      "loss": 0.457,
+      "step": 15110
+    },
+    {
+      "epoch": 3.770573566084788,
+      "grad_norm": 7.175182342529297,
+      "learning_rate": 1.623142144638404e-05,
+      "loss": 0.3434,
+      "step": 15120
+    },
+    {
+      "epoch": 3.773067331670823,
+      "grad_norm": 7.918752670288086,
+      "learning_rate": 1.6228927680798005e-05,
+      "loss": 0.4374,
+      "step": 15130
+    },
+    {
+      "epoch": 3.775561097256858,
+      "grad_norm": 5.573683738708496,
+      "learning_rate": 1.6226433915211972e-05,
+      "loss": 0.4069,
+      "step": 15140
+    },
+    {
+      "epoch": 3.7780548628428927,
+      "grad_norm": 6.013986110687256,
+      "learning_rate": 1.6223940149625936e-05,
+      "loss": 0.372,
+      "step": 15150
+    },
+    {
+      "epoch": 3.7805486284289276,
+      "grad_norm": 7.387099742889404,
+      "learning_rate": 1.62214463840399e-05,
+      "loss": 0.4348,
+      "step": 15160
+    },
+    {
+      "epoch": 3.7830423940149625,
+      "grad_norm": 8.931721687316895,
+      "learning_rate": 1.6218952618453866e-05,
+      "loss": 0.4304,
+      "step": 15170
+    },
+    {
+      "epoch": 3.7855361596009978,
+      "grad_norm": 6.92478084564209,
+      "learning_rate": 1.621645885286783e-05,
+      "loss": 0.4486,
+      "step": 15180
+    },
+    {
+      "epoch": 3.7880299251870326,
+      "grad_norm": 5.480514049530029,
+      "learning_rate": 1.6213965087281797e-05,
+      "loss": 0.4168,
+      "step": 15190
+    },
+    {
+      "epoch": 3.7905236907730675,
+      "grad_norm": 6.616566181182861,
+      "learning_rate": 1.621147132169576e-05,
+      "loss": 0.3913,
+      "step": 15200
+    },
+    {
+      "epoch": 3.7930174563591024,
+      "grad_norm": 6.839771270751953,
+      "learning_rate": 1.6208977556109728e-05,
+      "loss": 0.4484,
+      "step": 15210
+    },
+    {
+      "epoch": 3.7955112219451372,
+      "grad_norm": 17.69823455810547,
+      "learning_rate": 1.620648379052369e-05,
+      "loss": 0.4469,
+      "step": 15220
+    },
+    {
+      "epoch": 3.798004987531172,
+      "grad_norm": 5.237950325012207,
+      "learning_rate": 1.6203990024937658e-05,
+      "loss": 0.4347,
+      "step": 15230
+    },
+    {
+      "epoch": 3.800498753117207,
+      "grad_norm": 11.648773193359375,
+      "learning_rate": 1.6201496259351622e-05,
+      "loss": 0.4735,
+      "step": 15240
+    },
+    {
+      "epoch": 3.802992518703242,
+      "grad_norm": 4.579110622406006,
+      "learning_rate": 1.619900249376559e-05,
+      "loss": 0.3873,
+      "step": 15250
+    },
+    {
+      "epoch": 3.8054862842892767,
+      "grad_norm": 5.944640636444092,
+      "learning_rate": 1.6196508728179553e-05,
+      "loss": 0.4153,
+      "step": 15260
+    },
+    {
+      "epoch": 3.8079800498753116,
+      "grad_norm": 6.9358696937561035,
+      "learning_rate": 1.619401496259352e-05,
+      "loss": 0.3617,
+      "step": 15270
+    },
+    {
+      "epoch": 3.8104738154613464,
+      "grad_norm": 6.954071521759033,
+      "learning_rate": 1.6191521197007483e-05,
+      "loss": 0.4271,
+      "step": 15280
+    },
+    {
+      "epoch": 3.8129675810473813,
+      "grad_norm": 5.045011520385742,
+      "learning_rate": 1.6189027431421447e-05,
+      "loss": 0.4202,
+      "step": 15290
+    },
+    {
+      "epoch": 3.815461346633416,
+      "grad_norm": 5.211480140686035,
+      "learning_rate": 1.6186533665835414e-05,
+      "loss": 0.4072,
+      "step": 15300
+    },
+    {
+      "epoch": 3.8179551122194515,
+      "grad_norm": 5.901054382324219,
+      "learning_rate": 1.6184039900249377e-05,
+      "loss": 0.3808,
+      "step": 15310
+    },
+    {
+      "epoch": 3.8204488778054864,
+      "grad_norm": 7.312894821166992,
+      "learning_rate": 1.6181546134663344e-05,
+      "loss": 0.4005,
+      "step": 15320
+    },
+    {
+      "epoch": 3.8229426433915212,
+      "grad_norm": 4.741137981414795,
+      "learning_rate": 1.6179052369077308e-05,
+      "loss": 0.4235,
+      "step": 15330
+    },
+    {
+      "epoch": 3.825436408977556,
+      "grad_norm": 14.889175415039062,
+      "learning_rate": 1.617655860349127e-05,
+      "loss": 0.5169,
+      "step": 15340
+    },
+    {
+      "epoch": 3.827930174563591,
+      "grad_norm": 6.416481018066406,
+      "learning_rate": 1.617406483790524e-05,
+      "loss": 0.4472,
+      "step": 15350
+    },
+    {
+      "epoch": 3.830423940149626,
+      "grad_norm": 4.6261701583862305,
+      "learning_rate": 1.6171571072319202e-05,
+      "loss": 0.4069,
+      "step": 15360
+    },
+    {
+      "epoch": 3.8329177057356607,
+      "grad_norm": 4.661010265350342,
+      "learning_rate": 1.6169077306733166e-05,
+      "loss": 0.4312,
+      "step": 15370
+    },
+    {
+      "epoch": 3.835411471321696,
+      "grad_norm": 5.187774658203125,
+      "learning_rate": 1.6166583541147133e-05,
+      "loss": 0.3947,
+      "step": 15380
+    },
+    {
+      "epoch": 3.837905236907731,
+      "grad_norm": 4.887078762054443,
+      "learning_rate": 1.6164089775561096e-05,
+      "loss": 0.3661,
+      "step": 15390
+    },
+    {
+      "epoch": 3.8403990024937658,
+      "grad_norm": 4.326541423797607,
+      "learning_rate": 1.6161596009975063e-05,
+      "loss": 0.4481,
+      "step": 15400
+    },
+    {
+      "epoch": 3.8428927680798006,
+      "grad_norm": 6.179752826690674,
+      "learning_rate": 1.615910224438903e-05,
+      "loss": 0.424,
+      "step": 15410
+    },
+    {
+      "epoch": 3.8453865336658355,
+      "grad_norm": 5.251521110534668,
+      "learning_rate": 1.6156608478802994e-05,
+      "loss": 0.4454,
+      "step": 15420
+    },
+    {
+      "epoch": 3.8478802992518704,
+      "grad_norm": 9.553689956665039,
+      "learning_rate": 1.615411471321696e-05,
+      "loss": 0.4171,
+      "step": 15430
+    },
+    {
+      "epoch": 3.8503740648379052,
+      "grad_norm": 5.729796409606934,
+      "learning_rate": 1.6151620947630925e-05,
+      "loss": 0.4229,
+      "step": 15440
+    },
+    {
+      "epoch": 3.85286783042394,
+      "grad_norm": 4.4921345710754395,
+      "learning_rate": 1.614912718204489e-05,
+      "loss": 0.3616,
+      "step": 15450
+    },
+    {
+      "epoch": 3.855361596009975,
+      "grad_norm": 5.6195502281188965,
+      "learning_rate": 1.6146633416458855e-05,
+      "loss": 0.3688,
+      "step": 15460
+    },
+    {
+      "epoch": 3.85785536159601,
+      "grad_norm": 6.036675930023193,
+      "learning_rate": 1.614413965087282e-05,
+      "loss": 0.379,
+      "step": 15470
+    },
+    {
+      "epoch": 3.8603491271820447,
+      "grad_norm": 6.581831455230713,
+      "learning_rate": 1.6141645885286786e-05,
+      "loss": 0.4892,
+      "step": 15480
+    },
+    {
+      "epoch": 3.8628428927680796,
+      "grad_norm": 9.268636703491211,
+      "learning_rate": 1.613915211970075e-05,
+      "loss": 0.4021,
+      "step": 15490
+    },
+    {
+      "epoch": 3.8653366583541144,
+      "grad_norm": 6.355315685272217,
+      "learning_rate": 1.6136658354114713e-05,
+      "loss": 0.4159,
+      "step": 15500
+    },
+    {
+      "epoch": 3.8678304239401498,
+      "grad_norm": 6.00802755355835,
+      "learning_rate": 1.613416458852868e-05,
+      "loss": 0.4285,
+      "step": 15510
+    },
+    {
+      "epoch": 3.8703241895261846,
+      "grad_norm": 5.517050743103027,
+      "learning_rate": 1.6131670822942644e-05,
+      "loss": 0.3094,
+      "step": 15520
+    },
+    {
+      "epoch": 3.8728179551122195,
+      "grad_norm": 5.958836078643799,
+      "learning_rate": 1.612917705735661e-05,
+      "loss": 0.5116,
+      "step": 15530
+    },
+    {
+      "epoch": 3.8753117206982544,
+      "grad_norm": 4.79234504699707,
+      "learning_rate": 1.6126683291770574e-05,
+      "loss": 0.3931,
+      "step": 15540
+    },
+    {
+      "epoch": 3.8778054862842892,
+      "grad_norm": 7.781925201416016,
+      "learning_rate": 1.6124189526184538e-05,
+      "loss": 0.4291,
+      "step": 15550
+    },
+    {
+      "epoch": 3.880299251870324,
+      "grad_norm": 6.349798679351807,
+      "learning_rate": 1.6121695760598505e-05,
+      "loss": 0.5231,
+      "step": 15560
+    },
+    {
+      "epoch": 3.882793017456359,
+      "grad_norm": 6.636193752288818,
+      "learning_rate": 1.611920199501247e-05,
+      "loss": 0.369,
+      "step": 15570
+    },
+    {
+      "epoch": 3.8852867830423943,
+      "grad_norm": 5.7847580909729,
+      "learning_rate": 1.6116708229426436e-05,
+      "loss": 0.4435,
+      "step": 15580
+    },
+    {
+      "epoch": 3.887780548628429,
+      "grad_norm": 6.807628154754639,
+      "learning_rate": 1.61142144638404e-05,
+      "loss": 0.4008,
+      "step": 15590
+    },
+    {
+      "epoch": 3.890274314214464,
+      "grad_norm": 8.10521125793457,
+      "learning_rate": 1.6111720698254366e-05,
+      "loss": 0.4364,
+      "step": 15600
+    },
+    {
+      "epoch": 3.892768079800499,
+      "grad_norm": 5.665589332580566,
+      "learning_rate": 1.610922693266833e-05,
+      "loss": 0.4302,
+      "step": 15610
+    },
+    {
+      "epoch": 3.8952618453865338,
+      "grad_norm": 6.3911943435668945,
+      "learning_rate": 1.6106733167082297e-05,
+      "loss": 0.4188,
+      "step": 15620
+    },
+    {
+      "epoch": 3.8977556109725686,
+      "grad_norm": 7.833658695220947,
+      "learning_rate": 1.610423940149626e-05,
+      "loss": 0.3424,
+      "step": 15630
+    },
+    {
+      "epoch": 3.9002493765586035,
+      "grad_norm": 5.939789772033691,
+      "learning_rate": 1.6101745635910228e-05,
+      "loss": 0.4221,
+      "step": 15640
+    },
+    {
+      "epoch": 3.9027431421446384,
+      "grad_norm": 6.181157112121582,
+      "learning_rate": 1.609925187032419e-05,
+      "loss": 0.4927,
+      "step": 15650
+    },
+    {
+      "epoch": 3.9052369077306732,
+      "grad_norm": 10.083945274353027,
+      "learning_rate": 1.6096758104738155e-05,
+      "loss": 0.4323,
+      "step": 15660
+    },
+    {
+      "epoch": 3.907730673316708,
+      "grad_norm": 7.5239129066467285,
+      "learning_rate": 1.6094264339152122e-05,
+      "loss": 0.4576,
+      "step": 15670
+    },
+    {
+      "epoch": 3.910224438902743,
+      "grad_norm": 7.4439311027526855,
+      "learning_rate": 1.6091770573566085e-05,
+      "loss": 0.373,
+      "step": 15680
+    },
+    {
+      "epoch": 3.912718204488778,
+      "grad_norm": 8.566959381103516,
+      "learning_rate": 1.6089276807980052e-05,
+      "loss": 0.4426,
+      "step": 15690
+    },
+    {
+      "epoch": 3.9152119700748127,
+      "grad_norm": 6.725310802459717,
+      "learning_rate": 1.6086783042394016e-05,
+      "loss": 0.4016,
+      "step": 15700
+    },
+    {
+      "epoch": 3.917705735660848,
+      "grad_norm": 7.9485392570495605,
+      "learning_rate": 1.608428927680798e-05,
+      "loss": 0.416,
+      "step": 15710
+    },
+    {
+      "epoch": 3.920199501246883,
+      "grad_norm": 7.6618170738220215,
+      "learning_rate": 1.6081795511221947e-05,
+      "loss": 0.3751,
+      "step": 15720
+    },
+    {
+      "epoch": 3.9226932668329177,
+      "grad_norm": 5.6373291015625,
+      "learning_rate": 1.607930174563591e-05,
+      "loss": 0.3605,
+      "step": 15730
+    },
+    {
+      "epoch": 3.9251870324189526,
+      "grad_norm": 4.732673168182373,
+      "learning_rate": 1.6076807980049874e-05,
+      "loss": 0.3484,
+      "step": 15740
+    },
+    {
+      "epoch": 3.9276807980049875,
+      "grad_norm": 6.598507404327393,
+      "learning_rate": 1.607431421446384e-05,
+      "loss": 0.381,
+      "step": 15750
+    },
+    {
+      "epoch": 3.9301745635910224,
+      "grad_norm": 4.610682010650635,
+      "learning_rate": 1.6071820448877808e-05,
+      "loss": 0.4001,
+      "step": 15760
+    },
+    {
+      "epoch": 3.932668329177057,
+      "grad_norm": 4.403365135192871,
+      "learning_rate": 1.606932668329177e-05,
+      "loss": 0.4298,
+      "step": 15770
+    },
+    {
+      "epoch": 3.9351620947630925,
+      "grad_norm": 5.305014133453369,
+      "learning_rate": 1.606683291770574e-05,
+      "loss": 0.361,
+      "step": 15780
+    },
+    {
+      "epoch": 3.9376558603491274,
+      "grad_norm": 7.3379998207092285,
+      "learning_rate": 1.6064339152119702e-05,
+      "loss": 0.4318,
+      "step": 15790
+    },
+    {
+      "epoch": 3.9401496259351623,
+      "grad_norm": 4.637630939483643,
+      "learning_rate": 1.606184538653367e-05,
+      "loss": 0.3809,
+      "step": 15800
+    },
+    {
+      "epoch": 3.942643391521197,
+      "grad_norm": 7.531750202178955,
+      "learning_rate": 1.6059351620947633e-05,
+      "loss": 0.3713,
+      "step": 15810
+    },
+    {
+      "epoch": 3.945137157107232,
+      "grad_norm": 5.172116756439209,
+      "learning_rate": 1.60568578553616e-05,
+      "loss": 0.5203,
+      "step": 15820
+    },
+    {
+      "epoch": 3.947630922693267,
+      "grad_norm": 4.275954723358154,
+      "learning_rate": 1.6054364089775563e-05,
+      "loss": 0.3671,
+      "step": 15830
+    },
+    {
+      "epoch": 3.9501246882793017,
+      "grad_norm": 9.31353759765625,
+      "learning_rate": 1.6051870324189527e-05,
+      "loss": 0.376,
+      "step": 15840
+    },
+    {
+      "epoch": 3.9526184538653366,
+      "grad_norm": 7.531866550445557,
+      "learning_rate": 1.6049376558603494e-05,
+      "loss": 0.3427,
+      "step": 15850
+    },
+    {
+      "epoch": 3.9551122194513715,
+      "grad_norm": 24.911481857299805,
+      "learning_rate": 1.6046882793017458e-05,
+      "loss": 0.3652,
+      "step": 15860
+    },
+    {
+      "epoch": 3.9576059850374063,
+      "grad_norm": 8.664227485656738,
+      "learning_rate": 1.604438902743142e-05,
+      "loss": 0.4131,
+      "step": 15870
+    },
+    {
+      "epoch": 3.960099750623441,
+      "grad_norm": 32.71113586425781,
+      "learning_rate": 1.6041895261845388e-05,
+      "loss": 0.3416,
+      "step": 15880
+    },
+    {
+      "epoch": 3.962593516209476,
+      "grad_norm": 5.102753162384033,
+      "learning_rate": 1.6039401496259352e-05,
+      "loss": 0.359,
+      "step": 15890
+    },
+    {
+      "epoch": 3.965087281795511,
+      "grad_norm": 4.086287021636963,
+      "learning_rate": 1.603690773067332e-05,
+      "loss": 0.3796,
+      "step": 15900
+    },
+    {
+      "epoch": 3.9675810473815463,
+      "grad_norm": 6.271583080291748,
+      "learning_rate": 1.6034413965087282e-05,
+      "loss": 0.4116,
+      "step": 15910
+    },
+    {
+      "epoch": 3.970074812967581,
+      "grad_norm": 6.0941877365112305,
+      "learning_rate": 1.6031920199501246e-05,
+      "loss": 0.4144,
+      "step": 15920
+    },
+    {
+      "epoch": 3.972568578553616,
+      "grad_norm": 6.220065593719482,
+      "learning_rate": 1.6029426433915213e-05,
+      "loss": 0.3997,
+      "step": 15930
+    },
+    {
+      "epoch": 3.975062344139651,
+      "grad_norm": 6.358421802520752,
+      "learning_rate": 1.6026932668329177e-05,
+      "loss": 0.415,
+      "step": 15940
+    },
+    {
+      "epoch": 3.9775561097256857,
+      "grad_norm": 5.868401050567627,
+      "learning_rate": 1.6024438902743144e-05,
+      "loss": 0.4425,
+      "step": 15950
+    },
+    {
+      "epoch": 3.9800498753117206,
+      "grad_norm": 4.294557571411133,
+      "learning_rate": 1.6021945137157107e-05,
+      "loss": 0.3949,
+      "step": 15960
+    },
+    {
+      "epoch": 3.9825436408977555,
+      "grad_norm": 8.493629455566406,
+      "learning_rate": 1.6019451371571074e-05,
+      "loss": 0.4398,
+      "step": 15970
+    },
+    {
+      "epoch": 3.985037406483791,
+      "grad_norm": 7.112602233886719,
+      "learning_rate": 1.6016957605985038e-05,
+      "loss": 0.4643,
+      "step": 15980
+    },
+    {
+      "epoch": 3.9875311720698257,
+      "grad_norm": 11.970419883728027,
+      "learning_rate": 1.6014463840399005e-05,
+      "loss": 0.3528,
+      "step": 15990
+    },
+    {
+      "epoch": 3.9900249376558605,
+      "grad_norm": 6.464308738708496,
+      "learning_rate": 1.601197007481297e-05,
+      "loss": 0.393,
+      "step": 16000
+    },
+    {
+      "epoch": 3.9925187032418954,
+      "grad_norm": 11.118416786193848,
+      "learning_rate": 1.6009476309226936e-05,
+      "loss": 0.3788,
+      "step": 16010
+    },
+    {
+      "epoch": 3.9950124688279303,
+      "grad_norm": 9.291110038757324,
+      "learning_rate": 1.60069825436409e-05,
+      "loss": 0.4067,
+      "step": 16020
+    },
+    {
+      "epoch": 3.997506234413965,
+      "grad_norm": 2.8626253604888916,
+      "learning_rate": 1.6004488778054866e-05,
+      "loss": 0.3647,
+      "step": 16030
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 4.152261734008789,
+      "learning_rate": 1.600199501246883e-05,
+      "loss": 0.3658,
+      "step": 16040
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 0.4268016815185547,
+      "eval_runtime": 59.9163,
+      "eval_samples_per_second": 16.74,
+      "eval_steps_per_second": 16.74,
+      "step": 16040
+    },
+    {
+      "epoch": 4.002493765586035,
+      "grad_norm": 7.153280735015869,
+      "learning_rate": 1.5999501246882793e-05,
+      "loss": 0.417,
+      "step": 16050
+    },
+    {
+      "epoch": 4.00498753117207,
+      "grad_norm": 6.5900163650512695,
+      "learning_rate": 1.599700748129676e-05,
+      "loss": 0.3892,
+      "step": 16060
+    },
+    {
+      "epoch": 4.007481296758105,
+      "grad_norm": 4.667524337768555,
+      "learning_rate": 1.5994513715710724e-05,
+      "loss": 0.4389,
+      "step": 16070
+    },
+    {
+      "epoch": 4.0099750623441395,
+      "grad_norm": 4.793924808502197,
+      "learning_rate": 1.5992019950124688e-05,
+      "loss": 0.4087,
+      "step": 16080
+    },
+    {
+      "epoch": 4.012468827930174,
+      "grad_norm": 4.576432228088379,
+      "learning_rate": 1.5989526184538655e-05,
+      "loss": 0.4334,
+      "step": 16090
+    },
+    {
+      "epoch": 4.014962593516209,
+      "grad_norm": 6.454722881317139,
+      "learning_rate": 1.5987032418952618e-05,
+      "loss": 0.3957,
+      "step": 16100
+    },
+    {
+      "epoch": 4.017456359102244,
+      "grad_norm": 5.114348411560059,
+      "learning_rate": 1.5984538653366585e-05,
+      "loss": 0.3999,
+      "step": 16110
+    },
+    {
+      "epoch": 4.019950124688279,
+      "grad_norm": 3.897921085357666,
+      "learning_rate": 1.598204488778055e-05,
+      "loss": 0.3734,
+      "step": 16120
+    },
+    {
+      "epoch": 4.022443890274314,
+      "grad_norm": 6.09066915512085,
+      "learning_rate": 1.5979551122194516e-05,
+      "loss": 0.339,
+      "step": 16130
+    },
+    {
+      "epoch": 4.024937655860349,
+      "grad_norm": 7.583807945251465,
+      "learning_rate": 1.597705735660848e-05,
+      "loss": 0.3507,
+      "step": 16140
+    },
+    {
+      "epoch": 4.027431421446384,
+      "grad_norm": 10.415247917175293,
+      "learning_rate": 1.5974563591022446e-05,
+      "loss": 0.4775,
+      "step": 16150
+    },
+    {
+      "epoch": 4.029925187032419,
+      "grad_norm": 8.367255210876465,
+      "learning_rate": 1.597206982543641e-05,
+      "loss": 0.4167,
+      "step": 16160
+    },
+    {
+      "epoch": 4.032418952618454,
+      "grad_norm": 4.611159801483154,
+      "learning_rate": 1.5969576059850377e-05,
+      "loss": 0.4007,
+      "step": 16170
+    },
+    {
+      "epoch": 4.034912718204489,
+      "grad_norm": 4.2518205642700195,
+      "learning_rate": 1.596708229426434e-05,
+      "loss": 0.3998,
+      "step": 16180
+    },
+    {
+      "epoch": 4.037406483790524,
+      "grad_norm": 5.522350788116455,
+      "learning_rate": 1.5964588528678308e-05,
+      "loss": 0.4691,
+      "step": 16190
+    },
+    {
+      "epoch": 4.039900249376559,
+      "grad_norm": 4.40548849105835,
+      "learning_rate": 1.596209476309227e-05,
+      "loss": 0.4,
+      "step": 16200
+    },
+    {
+      "epoch": 4.042394014962594,
+      "grad_norm": 6.4272141456604,
+      "learning_rate": 1.5959600997506235e-05,
+      "loss": 0.4412,
+      "step": 16210
+    },
+    {
+      "epoch": 4.0448877805486285,
+      "grad_norm": 5.885697364807129,
+      "learning_rate": 1.5957107231920202e-05,
+      "loss": 0.3978,
+      "step": 16220
+    },
+    {
+      "epoch": 4.047381546134663,
+      "grad_norm": 6.327160835266113,
+      "learning_rate": 1.5954613466334166e-05,
+      "loss": 0.4464,
+      "step": 16230
+    },
+    {
+      "epoch": 4.049875311720698,
+      "grad_norm": 6.883283615112305,
+      "learning_rate": 1.595211970074813e-05,
+      "loss": 0.4137,
+      "step": 16240
+    },
+    {
+      "epoch": 4.052369077306733,
+      "grad_norm": 6.365976810455322,
+      "learning_rate": 1.5949625935162096e-05,
+      "loss": 0.4346,
+      "step": 16250
+    },
+    {
+      "epoch": 4.054862842892768,
+      "grad_norm": 5.095077037811279,
+      "learning_rate": 1.594713216957606e-05,
+      "loss": 0.3615,
+      "step": 16260
+    },
+    {
+      "epoch": 4.057356608478803,
+      "grad_norm": 5.573093414306641,
+      "learning_rate": 1.5944638403990027e-05,
+      "loss": 0.3894,
+      "step": 16270
+    },
+    {
+      "epoch": 4.059850374064838,
+      "grad_norm": 7.349832057952881,
+      "learning_rate": 1.594214463840399e-05,
+      "loss": 0.405,
+      "step": 16280
+    },
+    {
+      "epoch": 4.062344139650873,
+      "grad_norm": 7.18329381942749,
+      "learning_rate": 1.5939650872817954e-05,
+      "loss": 0.3667,
+      "step": 16290
+    },
+    {
+      "epoch": 4.0648379052369075,
+      "grad_norm": 5.264151096343994,
+      "learning_rate": 1.593715710723192e-05,
+      "loss": 0.4239,
+      "step": 16300
+    },
+    {
+      "epoch": 4.067331670822942,
+      "grad_norm": 5.923305511474609,
+      "learning_rate": 1.5934663341645885e-05,
+      "loss": 0.4125,
+      "step": 16310
+    },
+    {
+      "epoch": 4.069825436408977,
+      "grad_norm": 7.834733486175537,
+      "learning_rate": 1.593216957605985e-05,
+      "loss": 0.4075,
+      "step": 16320
+    },
+    {
+      "epoch": 4.072319201995012,
+      "grad_norm": 6.473409175872803,
+      "learning_rate": 1.592967581047382e-05,
+      "loss": 0.3263,
+      "step": 16330
+    },
+    {
+      "epoch": 4.074812967581048,
+      "grad_norm": 4.875690460205078,
+      "learning_rate": 1.5927182044887782e-05,
+      "loss": 0.3769,
+      "step": 16340
+    },
+    {
+      "epoch": 4.077306733167083,
+      "grad_norm": 6.263331890106201,
+      "learning_rate": 1.592468827930175e-05,
+      "loss": 0.3858,
+      "step": 16350
+    },
+    {
+      "epoch": 4.079800498753118,
+      "grad_norm": 6.207515716552734,
+      "learning_rate": 1.5922194513715713e-05,
+      "loss": 0.5084,
+      "step": 16360
+    },
+    {
+      "epoch": 4.082294264339152,
+      "grad_norm": 3.502013683319092,
+      "learning_rate": 1.5919700748129676e-05,
+      "loss": 0.3862,
+      "step": 16370
+    },
+    {
+      "epoch": 4.084788029925187,
+      "grad_norm": 7.68639612197876,
+      "learning_rate": 1.5917206982543643e-05,
+      "loss": 0.4036,
+      "step": 16380
+    },
+    {
+      "epoch": 4.087281795511222,
+      "grad_norm": 7.196399688720703,
+      "learning_rate": 1.5914713216957607e-05,
+      "loss": 0.3566,
+      "step": 16390
+    },
+    {
+      "epoch": 4.089775561097257,
+      "grad_norm": 6.002182960510254,
+      "learning_rate": 1.5912219451371574e-05,
+      "loss": 0.3955,
+      "step": 16400
+    },
+    {
+      "epoch": 4.092269326683292,
+      "grad_norm": 8.02131462097168,
+      "learning_rate": 1.5909725685785538e-05,
+      "loss": 0.3704,
+      "step": 16410
+    },
+    {
+      "epoch": 4.094763092269327,
+      "grad_norm": 5.517106533050537,
+      "learning_rate": 1.59072319201995e-05,
+      "loss": 0.4051,
+      "step": 16420
+    },
+    {
+      "epoch": 4.097256857855362,
+      "grad_norm": 9.180191040039062,
+      "learning_rate": 1.590473815461347e-05,
+      "loss": 0.3857,
+      "step": 16430
+    },
+    {
+      "epoch": 4.0997506234413965,
+      "grad_norm": 6.0054030418396,
+      "learning_rate": 1.5902244389027432e-05,
+      "loss": 0.3396,
+      "step": 16440
+    },
+    {
+      "epoch": 4.102244389027431,
+      "grad_norm": 5.941941261291504,
+      "learning_rate": 1.5899750623441396e-05,
+      "loss": 0.373,
+      "step": 16450
+    },
+    {
+      "epoch": 4.104738154613466,
+      "grad_norm": 6.833188056945801,
+      "learning_rate": 1.5897256857855363e-05,
+      "loss": 0.3765,
+      "step": 16460
+    },
+    {
+      "epoch": 4.107231920199501,
+      "grad_norm": 5.7840986251831055,
+      "learning_rate": 1.5894763092269326e-05,
+      "loss": 0.396,
+      "step": 16470
+    },
+    {
+      "epoch": 4.109725685785536,
+      "grad_norm": 7.854373931884766,
+      "learning_rate": 1.5892269326683293e-05,
+      "loss": 0.3982,
+      "step": 16480
+    },
+    {
+      "epoch": 4.112219451371571,
+      "grad_norm": 3.8397293090820312,
+      "learning_rate": 1.5889775561097257e-05,
+      "loss": 0.3874,
+      "step": 16490
+    },
+    {
+      "epoch": 4.114713216957606,
+      "grad_norm": 5.715033054351807,
+      "learning_rate": 1.5887281795511224e-05,
+      "loss": 0.3209,
+      "step": 16500
+    },
+    {
+      "epoch": 4.117206982543641,
+      "grad_norm": 7.231847286224365,
+      "learning_rate": 1.5884788029925187e-05,
+      "loss": 0.4834,
+      "step": 16510
+    },
+    {
+      "epoch": 4.1197007481296755,
+      "grad_norm": 4.733031272888184,
+      "learning_rate": 1.5882294264339154e-05,
+      "loss": 0.2822,
+      "step": 16520
+    },
+    {
+      "epoch": 4.12219451371571,
+      "grad_norm": 5.9622273445129395,
+      "learning_rate": 1.5879800498753118e-05,
+      "loss": 0.413,
+      "step": 16530
+    },
+    {
+      "epoch": 4.124688279301745,
+      "grad_norm": 5.854028701782227,
+      "learning_rate": 1.5877306733167085e-05,
+      "loss": 0.3331,
+      "step": 16540
+    },
+    {
+      "epoch": 4.127182044887781,
+      "grad_norm": 8.097307205200195,
+      "learning_rate": 1.587481296758105e-05,
+      "loss": 0.5102,
+      "step": 16550
+    },
+    {
+      "epoch": 4.129675810473816,
+      "grad_norm": 4.774763584136963,
+      "learning_rate": 1.5872319201995016e-05,
+      "loss": 0.3981,
+      "step": 16560
+    },
+    {
+      "epoch": 4.132169576059851,
+      "grad_norm": 6.483943939208984,
+      "learning_rate": 1.586982543640898e-05,
+      "loss": 0.4533,
+      "step": 16570
+    },
+    {
+      "epoch": 4.134663341645886,
+      "grad_norm": 10.268522262573242,
+      "learning_rate": 1.5867331670822943e-05,
+      "loss": 0.3954,
+      "step": 16580
+    },
+    {
+      "epoch": 4.13715710723192,
+      "grad_norm": 4.830404281616211,
+      "learning_rate": 1.586483790523691e-05,
+      "loss": 0.3578,
+      "step": 16590
+    },
+    {
+      "epoch": 4.139650872817955,
+      "grad_norm": 3.815784215927124,
+      "learning_rate": 1.5862344139650874e-05,
+      "loss": 0.3744,
+      "step": 16600
+    },
+    {
+      "epoch": 4.14214463840399,
+      "grad_norm": 5.319235324859619,
+      "learning_rate": 1.585985037406484e-05,
+      "loss": 0.3813,
+      "step": 16610
+    },
+    {
+      "epoch": 4.144638403990025,
+      "grad_norm": 7.98597526550293,
+      "learning_rate": 1.5857356608478804e-05,
+      "loss": 0.3337,
+      "step": 16620
+    },
+    {
+      "epoch": 4.14713216957606,
+      "grad_norm": 5.068573951721191,
+      "learning_rate": 1.5854862842892768e-05,
+      "loss": 0.4009,
+      "step": 16630
+    },
+    {
+      "epoch": 4.149625935162095,
+      "grad_norm": 4.63468599319458,
+      "learning_rate": 1.5852369077306735e-05,
+      "loss": 0.3888,
+      "step": 16640
+    },
+    {
+      "epoch": 4.15211970074813,
+      "grad_norm": 4.852334976196289,
+      "learning_rate": 1.58498753117207e-05,
+      "loss": 0.3634,
+      "step": 16650
+    },
+    {
+      "epoch": 4.1546134663341645,
+      "grad_norm": 10.55074691772461,
+      "learning_rate": 1.5847381546134662e-05,
+      "loss": 0.4119,
+      "step": 16660
+    },
+    {
+      "epoch": 4.157107231920199,
+      "grad_norm": 5.653322219848633,
+      "learning_rate": 1.584488778054863e-05,
+      "loss": 0.3409,
+      "step": 16670
+    },
+    {
+      "epoch": 4.159600997506234,
+      "grad_norm": 7.006217002868652,
+      "learning_rate": 1.5842394014962596e-05,
+      "loss": 0.3994,
+      "step": 16680
+    },
+    {
+      "epoch": 4.162094763092269,
+      "grad_norm": 6.214487552642822,
+      "learning_rate": 1.583990024937656e-05,
+      "loss": 0.4093,
+      "step": 16690
+    },
+    {
+      "epoch": 4.164588528678304,
+      "grad_norm": 9.91429615020752,
+      "learning_rate": 1.5837406483790527e-05,
+      "loss": 0.4333,
+      "step": 16700
+    },
+    {
+      "epoch": 4.167082294264339,
+      "grad_norm": 5.225426197052002,
+      "learning_rate": 1.583491271820449e-05,
+      "loss": 0.418,
+      "step": 16710
+    },
+    {
+      "epoch": 4.169576059850374,
+      "grad_norm": 4.0191969871521,
+      "learning_rate": 1.5832418952618457e-05,
+      "loss": 0.3455,
+      "step": 16720
+    },
+    {
+      "epoch": 4.172069825436409,
+      "grad_norm": 4.602033615112305,
+      "learning_rate": 1.582992518703242e-05,
+      "loss": 0.3552,
+      "step": 16730
+    },
+    {
+      "epoch": 4.174563591022444,
+      "grad_norm": 4.991748332977295,
+      "learning_rate": 1.5827431421446384e-05,
+      "loss": 0.4175,
+      "step": 16740
+    },
+    {
+      "epoch": 4.177057356608479,
+      "grad_norm": 6.951719284057617,
+      "learning_rate": 1.582493765586035e-05,
+      "loss": 0.4059,
+      "step": 16750
+    },
+    {
+      "epoch": 4.179551122194514,
+      "grad_norm": 5.848910808563232,
+      "learning_rate": 1.5822443890274315e-05,
+      "loss": 0.3944,
+      "step": 16760
+    },
+    {
+      "epoch": 4.182044887780549,
+      "grad_norm": 29.50406837463379,
+      "learning_rate": 1.5819950124688282e-05,
+      "loss": 0.4481,
+      "step": 16770
+    },
+    {
+      "epoch": 4.184538653366584,
+      "grad_norm": 8.905052185058594,
+      "learning_rate": 1.5817456359102246e-05,
+      "loss": 0.3706,
+      "step": 16780
+    },
+    {
+      "epoch": 4.187032418952619,
+      "grad_norm": 4.365262031555176,
+      "learning_rate": 1.581496259351621e-05,
+      "loss": 0.4487,
+      "step": 16790
+    },
+    {
+      "epoch": 4.1895261845386536,
+      "grad_norm": 6.916237831115723,
+      "learning_rate": 1.5812468827930176e-05,
+      "loss": 0.4587,
+      "step": 16800
+    },
+    {
+      "epoch": 4.192019950124688,
+      "grad_norm": 5.828021049499512,
+      "learning_rate": 1.580997506234414e-05,
+      "loss": 0.3274,
+      "step": 16810
+    },
+    {
+      "epoch": 4.194513715710723,
+      "grad_norm": 5.9600138664245605,
+      "learning_rate": 1.5807481296758107e-05,
+      "loss": 0.4037,
+      "step": 16820
+    },
+    {
+      "epoch": 4.197007481296758,
+      "grad_norm": 6.082637310028076,
+      "learning_rate": 1.580498753117207e-05,
+      "loss": 0.4237,
+      "step": 16830
+    },
+    {
+      "epoch": 4.199501246882793,
+      "grad_norm": 4.901458740234375,
+      "learning_rate": 1.5802493765586034e-05,
+      "loss": 0.4596,
+      "step": 16840
+    },
+    {
+      "epoch": 4.201995012468828,
+      "grad_norm": 6.101640701293945,
+      "learning_rate": 1.58e-05,
+      "loss": 0.3973,
+      "step": 16850
+    },
+    {
+      "epoch": 4.204488778054863,
+      "grad_norm": 6.996503829956055,
+      "learning_rate": 1.5797506234413965e-05,
+      "loss": 0.3572,
+      "step": 16860
+    },
+    {
+      "epoch": 4.206982543640898,
+      "grad_norm": 7.564293384552002,
+      "learning_rate": 1.5795012468827932e-05,
+      "loss": 0.3673,
+      "step": 16870
+    },
+    {
+      "epoch": 4.2094763092269325,
+      "grad_norm": 5.015920162200928,
+      "learning_rate": 1.5792518703241895e-05,
+      "loss": 0.3798,
+      "step": 16880
+    },
+    {
+      "epoch": 4.211970074812967,
+      "grad_norm": 9.611790657043457,
+      "learning_rate": 1.5790024937655862e-05,
+      "loss": 0.3374,
+      "step": 16890
+    },
+    {
+      "epoch": 4.214463840399002,
+      "grad_norm": 8.28926944732666,
+      "learning_rate": 1.5787531172069826e-05,
+      "loss": 0.3995,
+      "step": 16900
+    },
+    {
+      "epoch": 4.216957605985037,
+      "grad_norm": 7.903942108154297,
+      "learning_rate": 1.5785037406483793e-05,
+      "loss": 0.4166,
+      "step": 16910
+    },
+    {
+      "epoch": 4.219451371571072,
+      "grad_norm": 6.265456199645996,
+      "learning_rate": 1.5782543640897757e-05,
+      "loss": 0.4312,
+      "step": 16920
+    },
+    {
+      "epoch": 4.221945137157107,
+      "grad_norm": 7.296082019805908,
+      "learning_rate": 1.5780049875311724e-05,
+      "loss": 0.3976,
+      "step": 16930
+    },
+    {
+      "epoch": 4.224438902743142,
+      "grad_norm": 5.318020820617676,
+      "learning_rate": 1.5777556109725687e-05,
+      "loss": 0.3702,
+      "step": 16940
+    },
+    {
+      "epoch": 4.2269326683291775,
+      "grad_norm": 7.501054763793945,
+      "learning_rate": 1.577506234413965e-05,
+      "loss": 0.4675,
+      "step": 16950
+    },
+    {
+      "epoch": 4.229426433915212,
+      "grad_norm": 6.426706790924072,
+      "learning_rate": 1.5772568578553618e-05,
+      "loss": 0.3556,
+      "step": 16960
+    },
+    {
+      "epoch": 4.231920199501247,
+      "grad_norm": 8.110666275024414,
+      "learning_rate": 1.577007481296758e-05,
+      "loss": 0.4167,
+      "step": 16970
+    },
+    {
+      "epoch": 4.234413965087282,
+      "grad_norm": 6.1675238609313965,
+      "learning_rate": 1.576758104738155e-05,
+      "loss": 0.3816,
+      "step": 16980
+    },
+    {
+      "epoch": 4.236907730673317,
+      "grad_norm": 5.011690616607666,
+      "learning_rate": 1.5765087281795512e-05,
+      "loss": 0.3866,
+      "step": 16990
+    },
+    {
+      "epoch": 4.239401496259352,
+      "grad_norm": 6.697452545166016,
+      "learning_rate": 1.5762593516209476e-05,
+      "loss": 0.5945,
+      "step": 17000
+    },
+    {
+      "epoch": 4.241895261845387,
+      "grad_norm": 11.016718864440918,
+      "learning_rate": 1.5760099750623443e-05,
+      "loss": 0.3831,
+      "step": 17010
+    },
+    {
+      "epoch": 4.2443890274314215,
+      "grad_norm": 7.073511123657227,
+      "learning_rate": 1.5757605985037406e-05,
+      "loss": 0.4155,
+      "step": 17020
+    },
+    {
+      "epoch": 4.246882793017456,
+      "grad_norm": 6.227389812469482,
+      "learning_rate": 1.5755112219451373e-05,
+      "loss": 0.3701,
+      "step": 17030
+    },
+    {
+      "epoch": 4.249376558603491,
+      "grad_norm": 5.904865264892578,
+      "learning_rate": 1.5752618453865337e-05,
+      "loss": 0.4464,
+      "step": 17040
+    },
+    {
+      "epoch": 4.251870324189526,
+      "grad_norm": 6.812474250793457,
+      "learning_rate": 1.5750124688279304e-05,
+      "loss": 0.399,
+      "step": 17050
+    },
+    {
+      "epoch": 4.254364089775561,
+      "grad_norm": 6.779056072235107,
+      "learning_rate": 1.5747630922693268e-05,
+      "loss": 0.3984,
+      "step": 17060
+    },
+    {
+      "epoch": 4.256857855361596,
+      "grad_norm": 5.1520891189575195,
+      "learning_rate": 1.5745137157107235e-05,
+      "loss": 0.4212,
+      "step": 17070
+    },
+    {
+      "epoch": 4.259351620947631,
+      "grad_norm": 10.970466613769531,
+      "learning_rate": 1.5742643391521198e-05,
+      "loss": 0.3745,
+      "step": 17080
+    },
+    {
+      "epoch": 4.261845386533666,
+      "grad_norm": 5.555710792541504,
+      "learning_rate": 1.5740149625935165e-05,
+      "loss": 0.4095,
+      "step": 17090
+    },
+    {
+      "epoch": 4.2643391521197005,
+      "grad_norm": 5.996181011199951,
+      "learning_rate": 1.573765586034913e-05,
+      "loss": 0.3816,
+      "step": 17100
+    },
+    {
+      "epoch": 4.266832917705735,
+      "grad_norm": 5.2007646560668945,
+      "learning_rate": 1.5735162094763096e-05,
+      "loss": 0.4572,
+      "step": 17110
+    },
+    {
+      "epoch": 4.26932668329177,
+      "grad_norm": 6.208393096923828,
+      "learning_rate": 1.573266832917706e-05,
+      "loss": 0.3948,
+      "step": 17120
+    },
+    {
+      "epoch": 4.271820448877805,
+      "grad_norm": 6.411615371704102,
+      "learning_rate": 1.5730174563591023e-05,
+      "loss": 0.4109,
+      "step": 17130
+    },
+    {
+      "epoch": 4.274314214463841,
+      "grad_norm": 7.031231880187988,
+      "learning_rate": 1.572768079800499e-05,
+      "loss": 0.3633,
+      "step": 17140
+    },
+    {
+      "epoch": 4.276807980049876,
+      "grad_norm": 7.267471790313721,
+      "learning_rate": 1.5725187032418954e-05,
+      "loss": 0.3945,
+      "step": 17150
+    },
+    {
+      "epoch": 4.279301745635911,
+      "grad_norm": 6.009880065917969,
+      "learning_rate": 1.5722693266832917e-05,
+      "loss": 0.3755,
+      "step": 17160
+    },
+    {
+      "epoch": 4.2817955112219455,
+      "grad_norm": 6.955780029296875,
+      "learning_rate": 1.5720199501246884e-05,
+      "loss": 0.4382,
+      "step": 17170
+    },
+    {
+      "epoch": 4.28428927680798,
+      "grad_norm": 4.450560569763184,
+      "learning_rate": 1.5717705735660848e-05,
+      "loss": 0.4039,
+      "step": 17180
+    },
+    {
+      "epoch": 4.286783042394015,
+      "grad_norm": 7.145969390869141,
+      "learning_rate": 1.5715211970074815e-05,
+      "loss": 0.4875,
+      "step": 17190
+    },
+    {
+      "epoch": 4.28927680798005,
+      "grad_norm": 5.937014102935791,
+      "learning_rate": 1.571271820448878e-05,
+      "loss": 0.3465,
+      "step": 17200
+    },
+    {
+      "epoch": 4.291770573566085,
+      "grad_norm": 6.380857944488525,
+      "learning_rate": 1.5710224438902742e-05,
+      "loss": 0.4347,
+      "step": 17210
+    },
+    {
+      "epoch": 4.29426433915212,
+      "grad_norm": 7.357115268707275,
+      "learning_rate": 1.570773067331671e-05,
+      "loss": 0.3523,
+      "step": 17220
+    },
+    {
+      "epoch": 4.296758104738155,
+      "grad_norm": 6.485621929168701,
+      "learning_rate": 1.5705236907730673e-05,
+      "loss": 0.3737,
+      "step": 17230
+    },
+    {
+      "epoch": 4.2992518703241895,
+      "grad_norm": 5.544325828552246,
+      "learning_rate": 1.570274314214464e-05,
+      "loss": 0.4713,
+      "step": 17240
+    },
+    {
+      "epoch": 4.301745635910224,
+      "grad_norm": 4.941150188446045,
+      "learning_rate": 1.5700249376558603e-05,
+      "loss": 0.4139,
+      "step": 17250
+    },
+    {
+      "epoch": 4.304239401496259,
+      "grad_norm": 5.890329837799072,
+      "learning_rate": 1.569775561097257e-05,
+      "loss": 0.4586,
+      "step": 17260
+    },
+    {
+      "epoch": 4.306733167082294,
+      "grad_norm": 4.7308669090271,
+      "learning_rate": 1.5695261845386537e-05,
+      "loss": 0.3211,
+      "step": 17270
+    },
+    {
+      "epoch": 4.309226932668329,
+      "grad_norm": 6.511252403259277,
+      "learning_rate": 1.56927680798005e-05,
+      "loss": 0.3643,
+      "step": 17280
+    },
+    {
+      "epoch": 4.311720698254364,
+      "grad_norm": 6.328287124633789,
+      "learning_rate": 1.5690274314214465e-05,
+      "loss": 0.377,
+      "step": 17290
+    },
+    {
+      "epoch": 4.314214463840399,
+      "grad_norm": 12.119094848632812,
+      "learning_rate": 1.568778054862843e-05,
+      "loss": 0.3672,
+      "step": 17300
+    },
+    {
+      "epoch": 4.316708229426434,
+      "grad_norm": 5.892948627471924,
+      "learning_rate": 1.5685286783042395e-05,
+      "loss": 0.3174,
+      "step": 17310
+    },
+    {
+      "epoch": 4.3192019950124685,
+      "grad_norm": 6.068002700805664,
+      "learning_rate": 1.5682793017456362e-05,
+      "loss": 0.4032,
+      "step": 17320
+    },
+    {
+      "epoch": 4.321695760598503,
+      "grad_norm": 4.890083312988281,
+      "learning_rate": 1.5680299251870326e-05,
+      "loss": 0.422,
+      "step": 17330
+    },
+    {
+      "epoch": 4.324189526184538,
+      "grad_norm": 4.682448863983154,
+      "learning_rate": 1.567780548628429e-05,
+      "loss": 0.3798,
+      "step": 17340
+    },
+    {
+      "epoch": 4.326683291770574,
+      "grad_norm": 9.495527267456055,
+      "learning_rate": 1.5675311720698257e-05,
+      "loss": 0.4184,
+      "step": 17350
+    },
+    {
+      "epoch": 4.329177057356609,
+      "grad_norm": 6.011453151702881,
+      "learning_rate": 1.567281795511222e-05,
+      "loss": 0.3809,
+      "step": 17360
+    },
+    {
+      "epoch": 4.331670822942644,
+      "grad_norm": 8.991615295410156,
+      "learning_rate": 1.5670324189526184e-05,
+      "loss": 0.3961,
+      "step": 17370
+    },
+    {
+      "epoch": 4.334164588528679,
+      "grad_norm": 7.84186315536499,
+      "learning_rate": 1.566783042394015e-05,
+      "loss": 0.4104,
+      "step": 17380
+    },
+    {
+      "epoch": 4.3366583541147135,
+      "grad_norm": 4.632195949554443,
+      "learning_rate": 1.5665336658354114e-05,
+      "loss": 0.3265,
+      "step": 17390
+    },
+    {
+      "epoch": 4.339152119700748,
+      "grad_norm": 7.049432754516602,
+      "learning_rate": 1.566284289276808e-05,
+      "loss": 0.3176,
+      "step": 17400
+    },
+    {
+      "epoch": 4.341645885286783,
+      "grad_norm": 5.684892177581787,
+      "learning_rate": 1.5660349127182045e-05,
+      "loss": 0.3997,
+      "step": 17410
+    },
+    {
+      "epoch": 4.344139650872818,
+      "grad_norm": 7.792070388793945,
+      "learning_rate": 1.5657855361596012e-05,
+      "loss": 0.4101,
+      "step": 17420
+    },
+    {
+      "epoch": 4.346633416458853,
+      "grad_norm": 6.015437126159668,
+      "learning_rate": 1.5655361596009976e-05,
+      "loss": 0.4351,
+      "step": 17430
+    },
+    {
+      "epoch": 4.349127182044888,
+      "grad_norm": 5.80635404586792,
+      "learning_rate": 1.5652867830423943e-05,
+      "loss": 0.3938,
+      "step": 17440
+    },
+    {
+      "epoch": 4.351620947630923,
+      "grad_norm": 6.462253570556641,
+      "learning_rate": 1.5650374064837906e-05,
+      "loss": 0.4272,
+      "step": 17450
+    },
+    {
+      "epoch": 4.3541147132169575,
+      "grad_norm": 5.280763149261475,
+      "learning_rate": 1.5647880299251873e-05,
+      "loss": 0.3625,
+      "step": 17460
+    },
+    {
+      "epoch": 4.356608478802992,
+      "grad_norm": 6.560273170471191,
+      "learning_rate": 1.5645386533665837e-05,
+      "loss": 0.403,
+      "step": 17470
+    },
+    {
+      "epoch": 4.359102244389027,
+      "grad_norm": 6.191250324249268,
+      "learning_rate": 1.5642892768079804e-05,
+      "loss": 0.3826,
+      "step": 17480
+    },
+    {
+      "epoch": 4.361596009975062,
+      "grad_norm": 7.668184280395508,
+      "learning_rate": 1.5640399002493767e-05,
+      "loss": 0.3455,
+      "step": 17490
+    },
+    {
+      "epoch": 4.364089775561097,
+      "grad_norm": 6.343785285949707,
+      "learning_rate": 1.563790523690773e-05,
+      "loss": 0.3462,
+      "step": 17500
+    },
+    {
+      "epoch": 4.366583541147132,
+      "grad_norm": 7.69766092300415,
+      "learning_rate": 1.5635411471321698e-05,
+      "loss": 0.4203,
+      "step": 17510
+    },
+    {
+      "epoch": 4.369077306733167,
+      "grad_norm": 5.314762115478516,
+      "learning_rate": 1.5632917705735662e-05,
+      "loss": 0.4044,
+      "step": 17520
+    },
+    {
+      "epoch": 4.371571072319202,
+      "grad_norm": 4.914552688598633,
+      "learning_rate": 1.563042394014963e-05,
+      "loss": 0.3755,
+      "step": 17530
+    },
+    {
+      "epoch": 4.374064837905237,
+      "grad_norm": 5.039554119110107,
+      "learning_rate": 1.5627930174563592e-05,
+      "loss": 0.3815,
+      "step": 17540
+    },
+    {
+      "epoch": 4.376558603491272,
+      "grad_norm": 7.485365390777588,
+      "learning_rate": 1.5625436408977556e-05,
+      "loss": 0.418,
+      "step": 17550
+    },
+    {
+      "epoch": 4.379052369077307,
+      "grad_norm": 5.805276393890381,
+      "learning_rate": 1.5622942643391523e-05,
+      "loss": 0.4342,
+      "step": 17560
+    },
+    {
+      "epoch": 4.381546134663342,
+      "grad_norm": 5.376327037811279,
+      "learning_rate": 1.562069825436409e-05,
+      "loss": 0.4123,
+      "step": 17570
+    },
+    {
+      "epoch": 4.384039900249377,
+      "grad_norm": 7.425755500793457,
+      "learning_rate": 1.5618204488778058e-05,
+      "loss": 0.3918,
+      "step": 17580
+    },
+    {
+      "epoch": 4.386533665835412,
+      "grad_norm": 7.2214813232421875,
+      "learning_rate": 1.5615710723192022e-05,
+      "loss": 0.405,
+      "step": 17590
+    },
+    {
+      "epoch": 4.389027431421447,
+      "grad_norm": 7.342413902282715,
+      "learning_rate": 1.5613216957605985e-05,
+      "loss": 0.4112,
+      "step": 17600
+    },
+    {
+      "epoch": 4.3915211970074814,
+      "grad_norm": 6.210078716278076,
+      "learning_rate": 1.5610723192019952e-05,
+      "loss": 0.4194,
+      "step": 17610
+    },
+    {
+      "epoch": 4.394014962593516,
+      "grad_norm": 8.023131370544434,
+      "learning_rate": 1.5608229426433916e-05,
+      "loss": 0.3781,
+      "step": 17620
+    },
+    {
+      "epoch": 4.396508728179551,
+      "grad_norm": 6.007671356201172,
+      "learning_rate": 1.560573566084788e-05,
+      "loss": 0.4539,
+      "step": 17630
+    },
+    {
+      "epoch": 4.399002493765586,
+      "grad_norm": 6.679064750671387,
+      "learning_rate": 1.5603241895261847e-05,
+      "loss": 0.4137,
+      "step": 17640
+    },
+    {
+      "epoch": 4.401496259351621,
+      "grad_norm": 3.1536433696746826,
+      "learning_rate": 1.560074812967581e-05,
+      "loss": 0.4058,
+      "step": 17650
+    },
+    {
+      "epoch": 4.403990024937656,
+      "grad_norm": 6.0214009284973145,
+      "learning_rate": 1.5598254364089777e-05,
+      "loss": 0.4396,
+      "step": 17660
+    },
+    {
+      "epoch": 4.406483790523691,
+      "grad_norm": 4.901978015899658,
+      "learning_rate": 1.559576059850374e-05,
+      "loss": 0.3561,
+      "step": 17670
+    },
+    {
+      "epoch": 4.4089775561097255,
+      "grad_norm": 7.766887664794922,
+      "learning_rate": 1.5593266832917705e-05,
+      "loss": 0.4944,
+      "step": 17680
+    },
+    {
+      "epoch": 4.41147132169576,
+      "grad_norm": 5.148433685302734,
+      "learning_rate": 1.559077306733167e-05,
+      "loss": 0.4223,
+      "step": 17690
+    },
+    {
+      "epoch": 4.413965087281795,
+      "grad_norm": 4.767563343048096,
+      "learning_rate": 1.558827930174564e-05,
+      "loss": 0.3936,
+      "step": 17700
+    },
+    {
+      "epoch": 4.41645885286783,
+      "grad_norm": 7.088444232940674,
+      "learning_rate": 1.5585785536159602e-05,
+      "loss": 0.4696,
+      "step": 17710
+    },
+    {
+      "epoch": 4.418952618453865,
+      "grad_norm": 7.782017707824707,
+      "learning_rate": 1.558329177057357e-05,
+      "loss": 0.3238,
+      "step": 17720
+    },
+    {
+      "epoch": 4.4214463840399,
+      "grad_norm": 8.426640510559082,
+      "learning_rate": 1.5580798004987533e-05,
+      "loss": 0.422,
+      "step": 17730
+    },
+    {
+      "epoch": 4.423940149625935,
+      "grad_norm": 6.744356155395508,
+      "learning_rate": 1.55783042394015e-05,
+      "loss": 0.3593,
+      "step": 17740
+    },
+    {
+      "epoch": 4.42643391521197,
+      "grad_norm": 4.268016338348389,
+      "learning_rate": 1.5575810473815463e-05,
+      "loss": 0.4009,
+      "step": 17750
+    },
+    {
+      "epoch": 4.428927680798005,
+      "grad_norm": 6.248506546020508,
+      "learning_rate": 1.5573316708229427e-05,
+      "loss": 0.3938,
+      "step": 17760
+    },
+    {
+      "epoch": 4.43142144638404,
+      "grad_norm": 5.938436031341553,
+      "learning_rate": 1.5570822942643394e-05,
+      "loss": 0.4448,
+      "step": 17770
+    },
+    {
+      "epoch": 4.433915211970075,
+      "grad_norm": 6.432234764099121,
+      "learning_rate": 1.5568329177057358e-05,
+      "loss": 0.3664,
+      "step": 17780
+    },
+    {
+      "epoch": 4.43640897755611,
+      "grad_norm": 5.563558101654053,
+      "learning_rate": 1.5565835411471325e-05,
+      "loss": 0.5319,
+      "step": 17790
+    },
+    {
+      "epoch": 4.438902743142145,
+      "grad_norm": 5.881470680236816,
+      "learning_rate": 1.5563341645885288e-05,
+      "loss": 0.435,
+      "step": 17800
+    },
+    {
+      "epoch": 4.44139650872818,
+      "grad_norm": 6.403785705566406,
+      "learning_rate": 1.5560847880299252e-05,
+      "loss": 0.3419,
+      "step": 17810
+    },
+    {
+      "epoch": 4.443890274314215,
+      "grad_norm": 6.167340278625488,
+      "learning_rate": 1.555835411471322e-05,
+      "loss": 0.4303,
+      "step": 17820
+    },
+    {
+      "epoch": 4.446384039900249,
+      "grad_norm": 8.766891479492188,
+      "learning_rate": 1.5555860349127182e-05,
+      "loss": 0.3443,
+      "step": 17830
+    },
+    {
+      "epoch": 4.448877805486284,
+      "grad_norm": 6.43876314163208,
+      "learning_rate": 1.5553366583541146e-05,
+      "loss": 0.3824,
+      "step": 17840
+    },
+    {
+      "epoch": 4.451371571072319,
+      "grad_norm": 14.504803657531738,
+      "learning_rate": 1.5550872817955113e-05,
+      "loss": 0.3503,
+      "step": 17850
+    },
+    {
+      "epoch": 4.453865336658354,
+      "grad_norm": 8.448104858398438,
+      "learning_rate": 1.5548379052369077e-05,
+      "loss": 0.4532,
+      "step": 17860
+    },
+    {
+      "epoch": 4.456359102244389,
+      "grad_norm": 7.163640022277832,
+      "learning_rate": 1.5545885286783044e-05,
+      "loss": 0.4343,
+      "step": 17870
+    },
+    {
+      "epoch": 4.458852867830424,
+      "grad_norm": 5.441543102264404,
+      "learning_rate": 1.5543391521197007e-05,
+      "loss": 0.3523,
+      "step": 17880
+    },
+    {
+      "epoch": 4.461346633416459,
+      "grad_norm": 4.966891765594482,
+      "learning_rate": 1.5540897755610974e-05,
+      "loss": 0.3701,
+      "step": 17890
+    },
+    {
+      "epoch": 4.4638403990024935,
+      "grad_norm": 5.674979209899902,
+      "learning_rate": 1.5538403990024938e-05,
+      "loss": 0.3415,
+      "step": 17900
+    },
+    {
+      "epoch": 4.466334164588528,
+      "grad_norm": 6.1280646324157715,
+      "learning_rate": 1.5535910224438905e-05,
+      "loss": 0.4511,
+      "step": 17910
+    },
+    {
+      "epoch": 4.468827930174563,
+      "grad_norm": 6.596588611602783,
+      "learning_rate": 1.553341645885287e-05,
+      "loss": 0.3627,
+      "step": 17920
+    },
+    {
+      "epoch": 4.471321695760598,
+      "grad_norm": 7.014939785003662,
+      "learning_rate": 1.5530922693266836e-05,
+      "loss": 0.3572,
+      "step": 17930
+    },
+    {
+      "epoch": 4.473815461346634,
+      "grad_norm": 5.796361923217773,
+      "learning_rate": 1.55284289276808e-05,
+      "loss": 0.448,
+      "step": 17940
+    },
+    {
+      "epoch": 4.476309226932669,
+      "grad_norm": 7.400503635406494,
+      "learning_rate": 1.5525935162094766e-05,
+      "loss": 0.3636,
+      "step": 17950
+    },
+    {
+      "epoch": 4.478802992518704,
+      "grad_norm": 6.007717132568359,
+      "learning_rate": 1.552344139650873e-05,
+      "loss": 0.3922,
+      "step": 17960
+    },
+    {
+      "epoch": 4.4812967581047385,
+      "grad_norm": 11.898872375488281,
+      "learning_rate": 1.5520947630922693e-05,
+      "loss": 0.4342,
+      "step": 17970
+    },
+    {
+      "epoch": 4.483790523690773,
+      "grad_norm": 6.466373443603516,
+      "learning_rate": 1.551845386533666e-05,
+      "loss": 0.4021,
+      "step": 17980
+    },
+    {
+      "epoch": 4.486284289276808,
+      "grad_norm": 5.270114898681641,
+      "learning_rate": 1.5515960099750624e-05,
+      "loss": 0.384,
+      "step": 17990
+    },
+    {
+      "epoch": 4.488778054862843,
+      "grad_norm": 5.372076034545898,
+      "learning_rate": 1.551346633416459e-05,
+      "loss": 0.4017,
+      "step": 18000
+    },
+    {
+      "epoch": 4.491271820448878,
+      "grad_norm": 7.527553558349609,
+      "learning_rate": 1.5510972568578555e-05,
+      "loss": 0.3786,
+      "step": 18010
+    },
+    {
+      "epoch": 4.493765586034913,
+      "grad_norm": 7.008097171783447,
+      "learning_rate": 1.5508478802992518e-05,
+      "loss": 0.4263,
+      "step": 18020
+    },
+    {
+      "epoch": 4.496259351620948,
+      "grad_norm": 5.688605308532715,
+      "learning_rate": 1.5505985037406485e-05,
+      "loss": 0.3271,
+      "step": 18030
+    },
+    {
+      "epoch": 4.498753117206983,
+      "grad_norm": 3.6691343784332275,
+      "learning_rate": 1.550349127182045e-05,
+      "loss": 0.3862,
+      "step": 18040
+    },
+    {
+      "epoch": 4.501246882793017,
+      "grad_norm": 5.9994635581970215,
+      "learning_rate": 1.5500997506234416e-05,
+      "loss": 0.4175,
+      "step": 18050
+    },
+    {
+      "epoch": 4.503740648379052,
+      "grad_norm": 5.447284698486328,
+      "learning_rate": 1.549850374064838e-05,
+      "loss": 0.4004,
+      "step": 18060
+    },
+    {
+      "epoch": 4.506234413965087,
+      "grad_norm": 6.669367790222168,
+      "learning_rate": 1.5496009975062347e-05,
+      "loss": 0.4212,
+      "step": 18070
+    },
+    {
+      "epoch": 4.508728179551122,
+      "grad_norm": 5.361328601837158,
+      "learning_rate": 1.549351620947631e-05,
+      "loss": 0.4352,
+      "step": 18080
+    },
+    {
+      "epoch": 4.511221945137157,
+      "grad_norm": 7.188773155212402,
+      "learning_rate": 1.5491022443890277e-05,
+      "loss": 0.3761,
+      "step": 18090
+    },
+    {
+      "epoch": 4.513715710723192,
+      "grad_norm": 4.965447902679443,
+      "learning_rate": 1.548852867830424e-05,
+      "loss": 0.3335,
+      "step": 18100
+    },
+    {
+      "epoch": 4.516209476309227,
+      "grad_norm": 8.553069114685059,
+      "learning_rate": 1.5486034912718208e-05,
+      "loss": 0.4081,
+      "step": 18110
+    },
+    {
+      "epoch": 4.5187032418952615,
+      "grad_norm": 6.796214580535889,
+      "learning_rate": 1.548354114713217e-05,
+      "loss": 0.3981,
+      "step": 18120
+    },
+    {
+      "epoch": 4.521197007481296,
+      "grad_norm": 5.6893815994262695,
+      "learning_rate": 1.5481047381546135e-05,
+      "loss": 0.5065,
+      "step": 18130
+    },
+    {
+      "epoch": 4.523690773067331,
+      "grad_norm": 6.233173370361328,
+      "learning_rate": 1.5478553615960102e-05,
+      "loss": 0.3679,
+      "step": 18140
+    },
+    {
+      "epoch": 4.526184538653366,
+      "grad_norm": 7.2343597412109375,
+      "learning_rate": 1.5476059850374066e-05,
+      "loss": 0.4395,
+      "step": 18150
+    },
+    {
+      "epoch": 4.528678304239402,
+      "grad_norm": 7.412199020385742,
+      "learning_rate": 1.5473566084788033e-05,
+      "loss": 0.4316,
+      "step": 18160
+    },
+    {
+      "epoch": 4.531172069825437,
+      "grad_norm": 4.937939643859863,
+      "learning_rate": 1.5471072319201996e-05,
+      "loss": 0.3744,
+      "step": 18170
+    },
+    {
+      "epoch": 4.533665835411472,
+      "grad_norm": 6.558340072631836,
+      "learning_rate": 1.546857855361596e-05,
+      "loss": 0.418,
+      "step": 18180
+    },
+    {
+      "epoch": 4.5361596009975065,
+      "grad_norm": 6.0610551834106445,
+      "learning_rate": 1.5466084788029927e-05,
+      "loss": 0.4406,
+      "step": 18190
+    },
+    {
+      "epoch": 4.538653366583541,
+      "grad_norm": 4.485851764678955,
+      "learning_rate": 1.546359102244389e-05,
+      "loss": 0.4511,
+      "step": 18200
+    },
+    {
+      "epoch": 4.541147132169576,
+      "grad_norm": 5.121599197387695,
+      "learning_rate": 1.5461097256857857e-05,
+      "loss": 0.3711,
+      "step": 18210
+    },
+    {
+      "epoch": 4.543640897755611,
+      "grad_norm": 4.987266540527344,
+      "learning_rate": 1.545860349127182e-05,
+      "loss": 0.3687,
+      "step": 18220
+    },
+    {
+      "epoch": 4.546134663341646,
+      "grad_norm": 6.394310474395752,
+      "learning_rate": 1.5456109725685785e-05,
+      "loss": 0.3217,
+      "step": 18230
+    },
+    {
+      "epoch": 4.548628428927681,
+      "grad_norm": 5.757366180419922,
+      "learning_rate": 1.5453615960099752e-05,
+      "loss": 0.3671,
+      "step": 18240
+    },
+    {
+      "epoch": 4.551122194513716,
+      "grad_norm": 6.678833961486816,
+      "learning_rate": 1.5451122194513715e-05,
+      "loss": 0.3765,
+      "step": 18250
+    },
+    {
+      "epoch": 4.553615960099751,
+      "grad_norm": 6.454516410827637,
+      "learning_rate": 1.5448628428927682e-05,
+      "loss": 0.4051,
+      "step": 18260
+    },
+    {
+      "epoch": 4.556109725685785,
+      "grad_norm": 5.4813337326049805,
+      "learning_rate": 1.5446134663341646e-05,
+      "loss": 0.3641,
+      "step": 18270
+    },
+    {
+      "epoch": 4.55860349127182,
+      "grad_norm": 5.699728488922119,
+      "learning_rate": 1.5443640897755613e-05,
+      "loss": 0.3532,
+      "step": 18280
+    },
+    {
+      "epoch": 4.561097256857855,
+      "grad_norm": 24.547840118408203,
+      "learning_rate": 1.544114713216958e-05,
+      "loss": 0.5092,
+      "step": 18290
+    },
+    {
+      "epoch": 4.56359102244389,
+      "grad_norm": 3.5861904621124268,
+      "learning_rate": 1.5438653366583544e-05,
+      "loss": 0.3255,
+      "step": 18300
+    },
+    {
+      "epoch": 4.566084788029925,
+      "grad_norm": 5.126884937286377,
+      "learning_rate": 1.5436159600997507e-05,
+      "loss": 0.3716,
+      "step": 18310
+    },
+    {
+      "epoch": 4.56857855361596,
+      "grad_norm": 6.409011363983154,
+      "learning_rate": 1.5433665835411474e-05,
+      "loss": 0.3852,
+      "step": 18320
+    },
+    {
+      "epoch": 4.571072319201995,
+      "grad_norm": 5.01663875579834,
+      "learning_rate": 1.5431172069825438e-05,
+      "loss": 0.4487,
+      "step": 18330
+    },
+    {
+      "epoch": 4.57356608478803,
+      "grad_norm": 5.1992506980896,
+      "learning_rate": 1.54286783042394e-05,
+      "loss": 0.384,
+      "step": 18340
+    },
+    {
+      "epoch": 4.576059850374065,
+      "grad_norm": 9.998324394226074,
+      "learning_rate": 1.542618453865337e-05,
+      "loss": 0.4018,
+      "step": 18350
+    },
+    {
+      "epoch": 4.5785536159601,
+      "grad_norm": 5.419407844543457,
+      "learning_rate": 1.5423690773067332e-05,
+      "loss": 0.3933,
+      "step": 18360
+    },
+    {
+      "epoch": 4.581047381546135,
+      "grad_norm": 3.5099966526031494,
+      "learning_rate": 1.54211970074813e-05,
+      "loss": 0.3373,
+      "step": 18370
+    },
+    {
+      "epoch": 4.58354114713217,
+      "grad_norm": 6.4149627685546875,
+      "learning_rate": 1.5418703241895263e-05,
+      "loss": 0.3542,
+      "step": 18380
+    },
+    {
+      "epoch": 4.586034912718205,
+      "grad_norm": 7.924310684204102,
+      "learning_rate": 1.5416209476309226e-05,
+      "loss": 0.4282,
+      "step": 18390
+    },
+    {
+      "epoch": 4.58852867830424,
+      "grad_norm": 19.269880294799805,
+      "learning_rate": 1.5413715710723193e-05,
+      "loss": 0.4397,
+      "step": 18400
+    },
+    {
+      "epoch": 4.5910224438902745,
+      "grad_norm": 8.860027313232422,
+      "learning_rate": 1.5411221945137157e-05,
+      "loss": 0.3427,
+      "step": 18410
+    },
+    {
+      "epoch": 4.593516209476309,
+      "grad_norm": 7.146231174468994,
+      "learning_rate": 1.5408728179551124e-05,
+      "loss": 0.4281,
+      "step": 18420
+    },
+    {
+      "epoch": 4.596009975062344,
+      "grad_norm": 5.25446081161499,
+      "learning_rate": 1.5406234413965088e-05,
+      "loss": 0.4066,
+      "step": 18430
+    },
+    {
+      "epoch": 4.598503740648379,
+      "grad_norm": 4.486170291900635,
+      "learning_rate": 1.5403740648379055e-05,
+      "loss": 0.3729,
+      "step": 18440
+    },
+    {
+      "epoch": 4.600997506234414,
+      "grad_norm": 5.795819282531738,
+      "learning_rate": 1.5401496259351623e-05,
+      "loss": 0.3982,
+      "step": 18450
+    },
+    {
+      "epoch": 4.603491271820449,
+      "grad_norm": 5.524128437042236,
+      "learning_rate": 1.5399002493765586e-05,
+      "loss": 0.3987,
+      "step": 18460
+    },
+    {
+      "epoch": 4.605985037406484,
+      "grad_norm": 9.092752456665039,
+      "learning_rate": 1.5396508728179553e-05,
+      "loss": 0.3542,
+      "step": 18470
+    },
+    {
+      "epoch": 4.6084788029925186,
+      "grad_norm": 4.337942123413086,
+      "learning_rate": 1.5394014962593517e-05,
+      "loss": 0.3433,
+      "step": 18480
+    },
+    {
+      "epoch": 4.610972568578553,
+      "grad_norm": 7.767664909362793,
+      "learning_rate": 1.539152119700748e-05,
+      "loss": 0.412,
+      "step": 18490
+    },
+    {
+      "epoch": 4.613466334164588,
+      "grad_norm": 7.6244964599609375,
+      "learning_rate": 1.5389027431421448e-05,
+      "loss": 0.4377,
+      "step": 18500
+    },
+    {
+      "epoch": 4.615960099750623,
+      "grad_norm": 5.876872539520264,
+      "learning_rate": 1.538653366583541e-05,
+      "loss": 0.4679,
+      "step": 18510
+    },
+    {
+      "epoch": 4.618453865336658,
+      "grad_norm": 7.280019283294678,
+      "learning_rate": 1.5384039900249378e-05,
+      "loss": 0.4518,
+      "step": 18520
+    },
+    {
+      "epoch": 4.620947630922693,
+      "grad_norm": 6.68482780456543,
+      "learning_rate": 1.5381546134663342e-05,
+      "loss": 0.351,
+      "step": 18530
+    },
+    {
+      "epoch": 4.623441396508728,
+      "grad_norm": 6.15846061706543,
+      "learning_rate": 1.537905236907731e-05,
+      "loss": 0.4252,
+      "step": 18540
+    },
+    {
+      "epoch": 4.625935162094763,
+      "grad_norm": 13.20067024230957,
+      "learning_rate": 1.5376558603491272e-05,
+      "loss": 0.3845,
+      "step": 18550
+    },
+    {
+      "epoch": 4.628428927680798,
+      "grad_norm": 5.230330467224121,
+      "learning_rate": 1.537406483790524e-05,
+      "loss": 0.4242,
+      "step": 18560
+    },
+    {
+      "epoch": 4.630922693266833,
+      "grad_norm": 6.149280071258545,
+      "learning_rate": 1.5371571072319203e-05,
+      "loss": 0.4017,
+      "step": 18570
+    },
+    {
+      "epoch": 4.633416458852868,
+      "grad_norm": 5.865719318389893,
+      "learning_rate": 1.536907730673317e-05,
+      "loss": 0.3697,
+      "step": 18580
+    },
+    {
+      "epoch": 4.635910224438903,
+      "grad_norm": 7.069497585296631,
+      "learning_rate": 1.5366583541147134e-05,
+      "loss": 0.3835,
+      "step": 18590
+    },
+    {
+      "epoch": 4.638403990024938,
+      "grad_norm": 10.992486000061035,
+      "learning_rate": 1.53640897755611e-05,
+      "loss": 0.4504,
+      "step": 18600
+    },
+    {
+      "epoch": 4.640897755610973,
+      "grad_norm": 9.323265075683594,
+      "learning_rate": 1.5361596009975064e-05,
+      "loss": 0.5272,
+      "step": 18610
+    },
+    {
+      "epoch": 4.643391521197008,
+      "grad_norm": 7.848715305328369,
+      "learning_rate": 1.5359102244389028e-05,
+      "loss": 0.4452,
+      "step": 18620
+    },
+    {
+      "epoch": 4.6458852867830425,
+      "grad_norm": 7.649922847747803,
+      "learning_rate": 1.5356608478802995e-05,
+      "loss": 0.3857,
+      "step": 18630
+    },
+    {
+      "epoch": 4.648379052369077,
+      "grad_norm": 4.88482141494751,
+      "learning_rate": 1.535411471321696e-05,
+      "loss": 0.3648,
+      "step": 18640
+    },
+    {
+      "epoch": 4.650872817955112,
+      "grad_norm": 8.439806938171387,
+      "learning_rate": 1.5351620947630922e-05,
+      "loss": 0.3763,
+      "step": 18650
+    },
+    {
+      "epoch": 4.653366583541147,
+      "grad_norm": 7.73594856262207,
+      "learning_rate": 1.534912718204489e-05,
+      "loss": 0.354,
+      "step": 18660
+    },
+    {
+      "epoch": 4.655860349127182,
+      "grad_norm": 7.768985271453857,
+      "learning_rate": 1.5346633416458853e-05,
+      "loss": 0.3676,
+      "step": 18670
+    },
+    {
+      "epoch": 4.658354114713217,
+      "grad_norm": 5.735325336456299,
+      "learning_rate": 1.534413965087282e-05,
+      "loss": 0.3333,
+      "step": 18680
+    },
+    {
+      "epoch": 4.660847880299252,
+      "grad_norm": 4.903052806854248,
+      "learning_rate": 1.5341645885286783e-05,
+      "loss": 0.467,
+      "step": 18690
+    },
+    {
+      "epoch": 4.6633416458852865,
+      "grad_norm": 6.769343376159668,
+      "learning_rate": 1.5339152119700747e-05,
+      "loss": 0.4424,
+      "step": 18700
+    },
+    {
+      "epoch": 4.665835411471321,
+      "grad_norm": 22.33620262145996,
+      "learning_rate": 1.5336658354114714e-05,
+      "loss": 0.4389,
+      "step": 18710
+    },
+    {
+      "epoch": 4.668329177057356,
+      "grad_norm": 5.3549909591674805,
+      "learning_rate": 1.533416458852868e-05,
+      "loss": 0.362,
+      "step": 18720
+    },
+    {
+      "epoch": 4.670822942643391,
+      "grad_norm": 4.449096202850342,
+      "learning_rate": 1.5331670822942645e-05,
+      "loss": 0.3292,
+      "step": 18730
+    },
+    {
+      "epoch": 4.673316708229427,
+      "grad_norm": 7.292250156402588,
+      "learning_rate": 1.532917705735661e-05,
+      "loss": 0.3926,
+      "step": 18740
+    },
+    {
+      "epoch": 4.675810473815462,
+      "grad_norm": 6.387401103973389,
+      "learning_rate": 1.5326683291770575e-05,
+      "loss": 0.4135,
+      "step": 18750
+    },
+    {
+      "epoch": 4.678304239401497,
+      "grad_norm": 5.179088115692139,
+      "learning_rate": 1.5324189526184542e-05,
+      "loss": 0.3263,
+      "step": 18760
+    },
+    {
+      "epoch": 4.6807980049875315,
+      "grad_norm": 6.1036529541015625,
+      "learning_rate": 1.5321695760598506e-05,
+      "loss": 0.3773,
+      "step": 18770
+    },
+    {
+      "epoch": 4.683291770573566,
+      "grad_norm": 6.317544937133789,
+      "learning_rate": 1.531920199501247e-05,
+      "loss": 0.3549,
+      "step": 18780
+    },
+    {
+      "epoch": 4.685785536159601,
+      "grad_norm": 4.250217437744141,
+      "learning_rate": 1.5316708229426437e-05,
+      "loss": 0.3675,
+      "step": 18790
+    },
+    {
+      "epoch": 4.688279301745636,
+      "grad_norm": 4.993629455566406,
+      "learning_rate": 1.53142144638404e-05,
+      "loss": 0.3732,
+      "step": 18800
+    },
+    {
+      "epoch": 4.690773067331671,
+      "grad_norm": 6.042954444885254,
+      "learning_rate": 1.5311720698254364e-05,
+      "loss": 0.4064,
+      "step": 18810
+    },
+    {
+      "epoch": 4.693266832917706,
+      "grad_norm": 7.933563709259033,
+      "learning_rate": 1.530922693266833e-05,
+      "loss": 0.4322,
+      "step": 18820
+    },
+    {
+      "epoch": 4.695760598503741,
+      "grad_norm": 5.294432163238525,
+      "learning_rate": 1.5306733167082294e-05,
+      "loss": 0.3844,
+      "step": 18830
+    },
+    {
+      "epoch": 4.698254364089776,
+      "grad_norm": 4.980920791625977,
+      "learning_rate": 1.530423940149626e-05,
+      "loss": 0.3597,
+      "step": 18840
+    },
+    {
+      "epoch": 4.7007481296758105,
+      "grad_norm": 5.460011005401611,
+      "learning_rate": 1.5301745635910225e-05,
+      "loss": 0.3212,
+      "step": 18850
+    },
+    {
+      "epoch": 4.703241895261845,
+      "grad_norm": 6.553736686706543,
+      "learning_rate": 1.529925187032419e-05,
+      "loss": 0.4003,
+      "step": 18860
+    },
+    {
+      "epoch": 4.70573566084788,
+      "grad_norm": 6.58586311340332,
+      "learning_rate": 1.5296758104738156e-05,
+      "loss": 0.4062,
+      "step": 18870
+    },
+    {
+      "epoch": 4.708229426433915,
+      "grad_norm": 6.710026741027832,
+      "learning_rate": 1.529426433915212e-05,
+      "loss": 0.4248,
+      "step": 18880
+    },
+    {
+      "epoch": 4.71072319201995,
+      "grad_norm": 6.377664089202881,
+      "learning_rate": 1.5291770573566086e-05,
+      "loss": 0.467,
+      "step": 18890
+    },
+    {
+      "epoch": 4.713216957605985,
+      "grad_norm": 8.970911979675293,
+      "learning_rate": 1.528927680798005e-05,
+      "loss": 0.4081,
+      "step": 18900
+    },
+    {
+      "epoch": 4.71571072319202,
+      "grad_norm": 5.463969707489014,
+      "learning_rate": 1.5286783042394017e-05,
+      "loss": 0.3779,
+      "step": 18910
+    },
+    {
+      "epoch": 4.7182044887780545,
+      "grad_norm": 5.64981746673584,
+      "learning_rate": 1.528428927680798e-05,
+      "loss": 0.4542,
+      "step": 18920
+    },
+    {
+      "epoch": 4.720698254364089,
+      "grad_norm": 7.127771377563477,
+      "learning_rate": 1.5281795511221947e-05,
+      "loss": 0.3373,
+      "step": 18930
+    },
+    {
+      "epoch": 4.723192019950124,
+      "grad_norm": 6.836137771606445,
+      "learning_rate": 1.527930174563591e-05,
+      "loss": 0.3651,
+      "step": 18940
+    },
+    {
+      "epoch": 4.725685785536159,
+      "grad_norm": 7.067853927612305,
+      "learning_rate": 1.5276807980049878e-05,
+      "loss": 0.3785,
+      "step": 18950
+    },
+    {
+      "epoch": 4.728179551122195,
+      "grad_norm": 8.48742389678955,
+      "learning_rate": 1.5274314214463842e-05,
+      "loss": 0.3731,
+      "step": 18960
+    },
+    {
+      "epoch": 4.73067331670823,
+      "grad_norm": 13.028044700622559,
+      "learning_rate": 1.527182044887781e-05,
+      "loss": 0.3846,
+      "step": 18970
+    },
+    {
+      "epoch": 4.733167082294265,
+      "grad_norm": 8.332368850708008,
+      "learning_rate": 1.5269326683291772e-05,
+      "loss": 0.4137,
+      "step": 18980
+    },
+    {
+      "epoch": 4.7356608478802995,
+      "grad_norm": 7.124736785888672,
+      "learning_rate": 1.5266832917705736e-05,
+      "loss": 0.386,
+      "step": 18990
+    },
+    {
+      "epoch": 4.738154613466334,
+      "grad_norm": 6.675353527069092,
+      "learning_rate": 1.5264339152119703e-05,
+      "loss": 0.4329,
+      "step": 19000
+    },
+    {
+      "epoch": 4.740648379052369,
+      "grad_norm": 5.873729705810547,
+      "learning_rate": 1.5261845386533667e-05,
+      "loss": 0.4582,
+      "step": 19010
+    },
+    {
+      "epoch": 4.743142144638404,
+      "grad_norm": 7.303366184234619,
+      "learning_rate": 1.525935162094763e-05,
+      "loss": 0.416,
+      "step": 19020
+    },
+    {
+      "epoch": 4.745635910224439,
+      "grad_norm": 4.469297885894775,
+      "learning_rate": 1.5256857855361597e-05,
+      "loss": 0.3637,
+      "step": 19030
+    },
+    {
+      "epoch": 4.748129675810474,
+      "grad_norm": 4.885001182556152,
+      "learning_rate": 1.5254364089775563e-05,
+      "loss": 0.3829,
+      "step": 19040
+    },
+    {
+      "epoch": 4.750623441396509,
+      "grad_norm": 12.459643363952637,
+      "learning_rate": 1.5251870324189528e-05,
+      "loss": 0.3955,
+      "step": 19050
+    },
+    {
+      "epoch": 4.753117206982544,
+      "grad_norm": 4.591975688934326,
+      "learning_rate": 1.5249376558603493e-05,
+      "loss": 0.3381,
+      "step": 19060
+    },
+    {
+      "epoch": 4.7556109725685785,
+      "grad_norm": 5.559258937835693,
+      "learning_rate": 1.5246882793017457e-05,
+      "loss": 0.4173,
+      "step": 19070
+    },
+    {
+      "epoch": 4.758104738154613,
+      "grad_norm": 8.134532928466797,
+      "learning_rate": 1.5244389027431424e-05,
+      "loss": 0.4132,
+      "step": 19080
+    },
+    {
+      "epoch": 4.760598503740648,
+      "grad_norm": 8.197823524475098,
+      "learning_rate": 1.5241895261845387e-05,
+      "loss": 0.4422,
+      "step": 19090
+    },
+    {
+      "epoch": 4.763092269326683,
+      "grad_norm": 6.592520713806152,
+      "learning_rate": 1.5239401496259354e-05,
+      "loss": 0.3964,
+      "step": 19100
+    },
+    {
+      "epoch": 4.765586034912718,
+      "grad_norm": 9.510995864868164,
+      "learning_rate": 1.5236907730673318e-05,
+      "loss": 0.4011,
+      "step": 19110
+    },
+    {
+      "epoch": 4.768079800498753,
+      "grad_norm": 8.02636432647705,
+      "learning_rate": 1.5234413965087282e-05,
+      "loss": 0.3814,
+      "step": 19120
+    },
+    {
+      "epoch": 4.770573566084788,
+      "grad_norm": 8.324048042297363,
+      "learning_rate": 1.5231920199501249e-05,
+      "loss": 0.4601,
+      "step": 19130
+    },
+    {
+      "epoch": 4.773067331670823,
+      "grad_norm": 4.599636554718018,
+      "learning_rate": 1.5229426433915214e-05,
+      "loss": 0.3584,
+      "step": 19140
+    },
+    {
+      "epoch": 4.775561097256858,
+      "grad_norm": 7.513518333435059,
+      "learning_rate": 1.5226932668329178e-05,
+      "loss": 0.4006,
+      "step": 19150
+    },
+    {
+      "epoch": 4.778054862842893,
+      "grad_norm": 6.876750946044922,
+      "learning_rate": 1.5224438902743145e-05,
+      "loss": 0.3725,
+      "step": 19160
+    },
+    {
+      "epoch": 4.780548628428928,
+      "grad_norm": 5.191885471343994,
+      "learning_rate": 1.5221945137157108e-05,
+      "loss": 0.3534,
+      "step": 19170
+    },
+    {
+      "epoch": 4.783042394014963,
+      "grad_norm": 7.336450099945068,
+      "learning_rate": 1.5219451371571075e-05,
+      "loss": 0.408,
+      "step": 19180
+    },
+    {
+      "epoch": 4.785536159600998,
+      "grad_norm": 4.654347896575928,
+      "learning_rate": 1.5216957605985039e-05,
+      "loss": 0.3971,
+      "step": 19190
+    },
+    {
+      "epoch": 4.788029925187033,
+      "grad_norm": 8.277083396911621,
+      "learning_rate": 1.5214463840399002e-05,
+      "loss": 0.4695,
+      "step": 19200
+    },
+    {
+      "epoch": 4.7905236907730675,
+      "grad_norm": 5.439416885375977,
+      "learning_rate": 1.521197007481297e-05,
+      "loss": 0.4059,
+      "step": 19210
+    },
+    {
+      "epoch": 4.793017456359102,
+      "grad_norm": 7.094707489013672,
+      "learning_rate": 1.5209476309226933e-05,
+      "loss": 0.4118,
+      "step": 19220
+    },
+    {
+      "epoch": 4.795511221945137,
+      "grad_norm": 4.1312575340271,
+      "learning_rate": 1.5206982543640898e-05,
+      "loss": 0.3257,
+      "step": 19230
+    },
+    {
+      "epoch": 4.798004987531172,
+      "grad_norm": 14.407641410827637,
+      "learning_rate": 1.5204488778054864e-05,
+      "loss": 0.3796,
+      "step": 19240
+    },
+    {
+      "epoch": 4.800498753117207,
+      "grad_norm": 5.280395984649658,
+      "learning_rate": 1.5201995012468829e-05,
+      "loss": 0.3183,
+      "step": 19250
+    },
+    {
+      "epoch": 4.802992518703242,
+      "grad_norm": 5.899741172790527,
+      "learning_rate": 1.5199501246882796e-05,
+      "loss": 0.3432,
+      "step": 19260
+    },
+    {
+      "epoch": 4.805486284289277,
+      "grad_norm": 6.163057804107666,
+      "learning_rate": 1.519700748129676e-05,
+      "loss": 0.3459,
+      "step": 19270
+    },
+    {
+      "epoch": 4.807980049875312,
+      "grad_norm": 6.30502986907959,
+      "learning_rate": 1.5194513715710723e-05,
+      "loss": 0.4026,
+      "step": 19280
+    },
+    {
+      "epoch": 4.8104738154613464,
+      "grad_norm": 8.228434562683105,
+      "learning_rate": 1.519201995012469e-05,
+      "loss": 0.4527,
+      "step": 19290
+    },
+    {
+      "epoch": 4.812967581047381,
+      "grad_norm": 7.36271333694458,
+      "learning_rate": 1.5189526184538654e-05,
+      "loss": 0.4226,
+      "step": 19300
+    },
+    {
+      "epoch": 4.815461346633416,
+      "grad_norm": 6.21150016784668,
+      "learning_rate": 1.5187032418952619e-05,
+      "loss": 0.3855,
+      "step": 19310
+    },
+    {
+      "epoch": 4.817955112219451,
+      "grad_norm": 5.929027557373047,
+      "learning_rate": 1.5184538653366584e-05,
+      "loss": 0.3232,
+      "step": 19320
+    },
+    {
+      "epoch": 4.820448877805486,
+      "grad_norm": 12.694462776184082,
+      "learning_rate": 1.518204488778055e-05,
+      "loss": 0.3749,
+      "step": 19330
+    },
+    {
+      "epoch": 4.822942643391521,
+      "grad_norm": 4.2973456382751465,
+      "learning_rate": 1.5179551122194515e-05,
+      "loss": 0.3592,
+      "step": 19340
+    },
+    {
+      "epoch": 4.825436408977556,
+      "grad_norm": 6.927724838256836,
+      "learning_rate": 1.517705735660848e-05,
+      "loss": 0.3819,
+      "step": 19350
+    },
+    {
+      "epoch": 4.8279301745635905,
+      "grad_norm": 8.140523910522461,
+      "learning_rate": 1.5174563591022444e-05,
+      "loss": 0.4302,
+      "step": 19360
+    },
+    {
+      "epoch": 4.830423940149626,
+      "grad_norm": 9.106082916259766,
+      "learning_rate": 1.5172069825436411e-05,
+      "loss": 0.4181,
+      "step": 19370
+    },
+    {
+      "epoch": 4.832917705735661,
+      "grad_norm": 5.513749599456787,
+      "learning_rate": 1.5169576059850375e-05,
+      "loss": 0.3663,
+      "step": 19380
+    },
+    {
+      "epoch": 4.835411471321696,
+      "grad_norm": 4.832230567932129,
+      "learning_rate": 1.5167082294264342e-05,
+      "loss": 0.3402,
+      "step": 19390
+    },
+    {
+      "epoch": 4.837905236907731,
+      "grad_norm": 5.5058183670043945,
+      "learning_rate": 1.5164588528678305e-05,
+      "loss": 0.4934,
+      "step": 19400
+    },
+    {
+      "epoch": 4.840399002493766,
+      "grad_norm": 6.715292930603027,
+      "learning_rate": 1.516209476309227e-05,
+      "loss": 0.4119,
+      "step": 19410
+    },
+    {
+      "epoch": 4.842892768079801,
+      "grad_norm": 7.778273105621338,
+      "learning_rate": 1.5159600997506236e-05,
+      "loss": 0.3779,
+      "step": 19420
+    },
+    {
+      "epoch": 4.8453865336658355,
+      "grad_norm": 5.197988510131836,
+      "learning_rate": 1.5157107231920201e-05,
+      "loss": 0.3754,
+      "step": 19430
+    },
+    {
+      "epoch": 4.84788029925187,
+      "grad_norm": 4.799652099609375,
+      "learning_rate": 1.5154613466334165e-05,
+      "loss": 0.3453,
+      "step": 19440
+    },
+    {
+      "epoch": 4.850374064837905,
+      "grad_norm": 7.939327716827393,
+      "learning_rate": 1.5152119700748132e-05,
+      "loss": 0.3741,
+      "step": 19450
+    },
+    {
+      "epoch": 4.85286783042394,
+      "grad_norm": 7.11137056350708,
+      "learning_rate": 1.5149625935162095e-05,
+      "loss": 0.3905,
+      "step": 19460
+    },
+    {
+      "epoch": 4.855361596009975,
+      "grad_norm": 6.2819061279296875,
+      "learning_rate": 1.5147132169576062e-05,
+      "loss": 0.3733,
+      "step": 19470
+    },
+    {
+      "epoch": 4.85785536159601,
+      "grad_norm": 5.845408916473389,
+      "learning_rate": 1.5144638403990026e-05,
+      "loss": 0.4008,
+      "step": 19480
+    },
+    {
+      "epoch": 4.860349127182045,
+      "grad_norm": 7.455866813659668,
+      "learning_rate": 1.5142144638403991e-05,
+      "loss": 0.418,
+      "step": 19490
+    },
+    {
+      "epoch": 4.86284289276808,
+      "grad_norm": 5.083822250366211,
+      "learning_rate": 1.5139650872817957e-05,
+      "loss": 0.3625,
+      "step": 19500
+    },
+    {
+      "epoch": 4.865336658354114,
+      "grad_norm": 6.3122968673706055,
+      "learning_rate": 1.5137157107231922e-05,
+      "loss": 0.396,
+      "step": 19510
+    },
+    {
+      "epoch": 4.867830423940149,
+      "grad_norm": 6.627301216125488,
+      "learning_rate": 1.5134663341645886e-05,
+      "loss": 0.3841,
+      "step": 19520
+    },
+    {
+      "epoch": 4.870324189526184,
+      "grad_norm": 7.174892425537109,
+      "learning_rate": 1.5132169576059853e-05,
+      "loss": 0.3484,
+      "step": 19530
+    },
+    {
+      "epoch": 4.87281795511222,
+      "grad_norm": 6.404763698577881,
+      "learning_rate": 1.5129675810473816e-05,
+      "loss": 0.4297,
+      "step": 19540
+    },
+    {
+      "epoch": 4.875311720698255,
+      "grad_norm": 9.780095100402832,
+      "learning_rate": 1.5127182044887783e-05,
+      "loss": 0.4377,
+      "step": 19550
+    },
+    {
+      "epoch": 4.87780548628429,
+      "grad_norm": 6.8640546798706055,
+      "learning_rate": 1.5124688279301747e-05,
+      "loss": 0.3958,
+      "step": 19560
+    },
+    {
+      "epoch": 4.8802992518703245,
+      "grad_norm": 4.705758094787598,
+      "learning_rate": 1.512219451371571e-05,
+      "loss": 0.4039,
+      "step": 19570
+    },
+    {
+      "epoch": 4.882793017456359,
+      "grad_norm": 4.98261833190918,
+      "learning_rate": 1.5119700748129677e-05,
+      "loss": 0.4121,
+      "step": 19580
+    },
+    {
+      "epoch": 4.885286783042394,
+      "grad_norm": 9.222187042236328,
+      "learning_rate": 1.5117206982543641e-05,
+      "loss": 0.412,
+      "step": 19590
+    },
+    {
+      "epoch": 4.887780548628429,
+      "grad_norm": 8.537120819091797,
+      "learning_rate": 1.5114713216957608e-05,
+      "loss": 0.4807,
+      "step": 19600
+    },
+    {
+      "epoch": 4.890274314214464,
+      "grad_norm": 9.445647239685059,
+      "learning_rate": 1.5112219451371573e-05,
+      "loss": 0.4003,
+      "step": 19610
+    },
+    {
+      "epoch": 4.892768079800499,
+      "grad_norm": 7.455517292022705,
+      "learning_rate": 1.5109725685785537e-05,
+      "loss": 0.3835,
+      "step": 19620
+    },
+    {
+      "epoch": 4.895261845386534,
+      "grad_norm": 7.285569667816162,
+      "learning_rate": 1.5107231920199504e-05,
+      "loss": 0.3211,
+      "step": 19630
+    },
+    {
+      "epoch": 4.897755610972569,
+      "grad_norm": 8.98646068572998,
+      "learning_rate": 1.5104738154613468e-05,
+      "loss": 0.4019,
+      "step": 19640
+    },
+    {
+      "epoch": 4.9002493765586035,
+      "grad_norm": 8.273704528808594,
+      "learning_rate": 1.5102244389027431e-05,
+      "loss": 0.4262,
+      "step": 19650
+    },
+    {
+      "epoch": 4.902743142144638,
+      "grad_norm": 3.785250663757324,
+      "learning_rate": 1.5099750623441398e-05,
+      "loss": 0.3724,
+      "step": 19660
+    },
+    {
+      "epoch": 4.905236907730673,
+      "grad_norm": 6.882282257080078,
+      "learning_rate": 1.5097256857855362e-05,
+      "loss": 0.4621,
+      "step": 19670
+    },
+    {
+      "epoch": 4.907730673316708,
+      "grad_norm": 9.958419799804688,
+      "learning_rate": 1.5094763092269329e-05,
+      "loss": 0.3655,
+      "step": 19680
+    },
+    {
+      "epoch": 4.910224438902743,
+      "grad_norm": 4.580639839172363,
+      "learning_rate": 1.5092269326683292e-05,
+      "loss": 0.3733,
+      "step": 19690
+    },
+    {
+      "epoch": 4.912718204488778,
+      "grad_norm": 6.3402276039123535,
+      "learning_rate": 1.5089775561097258e-05,
+      "loss": 0.3853,
+      "step": 19700
+    },
+    {
+      "epoch": 4.915211970074813,
+      "grad_norm": 6.479229927062988,
+      "learning_rate": 1.5087281795511223e-05,
+      "loss": 0.4365,
+      "step": 19710
+    },
+    {
+      "epoch": 4.917705735660848,
+      "grad_norm": 5.332212924957275,
+      "learning_rate": 1.5084788029925188e-05,
+      "loss": 0.4072,
+      "step": 19720
+    },
+    {
+      "epoch": 4.920199501246882,
+      "grad_norm": 5.466404438018799,
+      "learning_rate": 1.5082294264339152e-05,
+      "loss": 0.4019,
+      "step": 19730
+    },
+    {
+      "epoch": 4.922693266832917,
+      "grad_norm": 4.750015735626221,
+      "learning_rate": 1.5079800498753119e-05,
+      "loss": 0.3635,
+      "step": 19740
+    },
+    {
+      "epoch": 4.925187032418952,
+      "grad_norm": 5.45477294921875,
+      "learning_rate": 1.5077306733167083e-05,
+      "loss": 0.3239,
+      "step": 19750
+    },
+    {
+      "epoch": 4.927680798004987,
+      "grad_norm": 7.6153564453125,
+      "learning_rate": 1.507481296758105e-05,
+      "loss": 0.3935,
+      "step": 19760
+    },
+    {
+      "epoch": 4.930174563591023,
+      "grad_norm": 5.9966816902160645,
+      "learning_rate": 1.5072319201995013e-05,
+      "loss": 0.3912,
+      "step": 19770
+    },
+    {
+      "epoch": 4.932668329177058,
+      "grad_norm": 4.490943908691406,
+      "learning_rate": 1.5069825436408978e-05,
+      "loss": 0.3943,
+      "step": 19780
+    },
+    {
+      "epoch": 4.9351620947630925,
+      "grad_norm": 4.777584075927734,
+      "learning_rate": 1.5067331670822944e-05,
+      "loss": 0.3863,
+      "step": 19790
+    },
+    {
+      "epoch": 4.937655860349127,
+      "grad_norm": 6.837095737457275,
+      "learning_rate": 1.5064837905236909e-05,
+      "loss": 0.3762,
+      "step": 19800
+    },
+    {
+      "epoch": 4.940149625935162,
+      "grad_norm": 6.350588321685791,
+      "learning_rate": 1.5062344139650873e-05,
+      "loss": 0.4286,
+      "step": 19810
+    },
+    {
+      "epoch": 4.942643391521197,
+      "grad_norm": 4.62924337387085,
+      "learning_rate": 1.505985037406484e-05,
+      "loss": 0.3772,
+      "step": 19820
+    },
+    {
+      "epoch": 4.945137157107232,
+      "grad_norm": 6.749570369720459,
+      "learning_rate": 1.5057356608478803e-05,
+      "loss": 0.4314,
+      "step": 19830
+    },
+    {
+      "epoch": 4.947630922693267,
+      "grad_norm": 5.663025856018066,
+      "learning_rate": 1.505486284289277e-05,
+      "loss": 0.4405,
+      "step": 19840
+    },
+    {
+      "epoch": 4.950124688279302,
+      "grad_norm": 4.510965347290039,
+      "learning_rate": 1.5052369077306734e-05,
+      "loss": 0.351,
+      "step": 19850
+    },
+    {
+      "epoch": 4.952618453865337,
+      "grad_norm": 4.851863384246826,
+      "learning_rate": 1.50498753117207e-05,
+      "loss": 0.3867,
+      "step": 19860
+    },
+    {
+      "epoch": 4.9551122194513715,
+      "grad_norm": 5.330405235290527,
+      "learning_rate": 1.5047381546134665e-05,
+      "loss": 0.4789,
+      "step": 19870
+    },
+    {
+      "epoch": 4.957605985037406,
+      "grad_norm": 13.125707626342773,
+      "learning_rate": 1.504488778054863e-05,
+      "loss": 0.4699,
+      "step": 19880
+    },
+    {
+      "epoch": 4.960099750623441,
+      "grad_norm": 6.033151149749756,
+      "learning_rate": 1.5042394014962595e-05,
+      "loss": 0.381,
+      "step": 19890
+    },
+    {
+      "epoch": 4.962593516209476,
+      "grad_norm": 5.597021102905273,
+      "learning_rate": 1.503990024937656e-05,
+      "loss": 0.3887,
+      "step": 19900
+    },
+    {
+      "epoch": 4.965087281795511,
+      "grad_norm": 5.585844039916992,
+      "learning_rate": 1.5037406483790524e-05,
+      "loss": 0.4441,
+      "step": 19910
+    },
+    {
+      "epoch": 4.967581047381546,
+      "grad_norm": 6.1360602378845215,
+      "learning_rate": 1.5034912718204491e-05,
+      "loss": 0.4344,
+      "step": 19920
+    },
+    {
+      "epoch": 4.970074812967581,
+      "grad_norm": 13.40194320678711,
+      "learning_rate": 1.5032418952618455e-05,
+      "loss": 0.4552,
+      "step": 19930
+    },
+    {
+      "epoch": 4.9725685785536164,
+      "grad_norm": 7.084679126739502,
+      "learning_rate": 1.502992518703242e-05,
+      "loss": 0.3554,
+      "step": 19940
+    },
+    {
+      "epoch": 4.975062344139651,
+      "grad_norm": 7.618978977203369,
+      "learning_rate": 1.5027431421446385e-05,
+      "loss": 0.3493,
+      "step": 19950
+    },
+    {
+      "epoch": 4.977556109725686,
+      "grad_norm": 6.25748348236084,
+      "learning_rate": 1.502493765586035e-05,
+      "loss": 0.4517,
+      "step": 19960
+    },
+    {
+      "epoch": 4.980049875311721,
+      "grad_norm": 5.518866062164307,
+      "learning_rate": 1.5022443890274316e-05,
+      "loss": 0.3737,
+      "step": 19970
+    },
+    {
+      "epoch": 4.982543640897756,
+      "grad_norm": 6.079436779022217,
+      "learning_rate": 1.5019950124688281e-05,
+      "loss": 0.4093,
+      "step": 19980
+    },
+    {
+      "epoch": 4.985037406483791,
+      "grad_norm": 6.771421432495117,
+      "learning_rate": 1.5017456359102245e-05,
+      "loss": 0.3948,
+      "step": 19990
+    },
+    {
+      "epoch": 4.987531172069826,
+      "grad_norm": 5.409707069396973,
+      "learning_rate": 1.5014962593516212e-05,
+      "loss": 0.4326,
+      "step": 20000
+    },
+    {
+      "epoch": 4.9900249376558605,
+      "grad_norm": 4.681028842926025,
+      "learning_rate": 1.5012468827930176e-05,
+      "loss": 0.3327,
+      "step": 20010
+    },
+    {
+      "epoch": 4.992518703241895,
+      "grad_norm": 7.876126766204834,
+      "learning_rate": 1.5009975062344139e-05,
+      "loss": 0.3613,
+      "step": 20020
+    },
+    {
+      "epoch": 4.99501246882793,
+      "grad_norm": 5.726997375488281,
+      "learning_rate": 1.5007481296758106e-05,
+      "loss": 0.4506,
+      "step": 20030
+    },
+    {
+      "epoch": 4.997506234413965,
+      "grad_norm": 6.533457279205322,
+      "learning_rate": 1.500498753117207e-05,
+      "loss": 0.4232,
+      "step": 20040
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 21.448368072509766,
+      "learning_rate": 1.5002493765586037e-05,
+      "loss": 0.4523,
+      "step": 20050
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 0.42168691754341125,
+      "eval_runtime": 60.2711,
+      "eval_samples_per_second": 16.641,
+      "eval_steps_per_second": 16.641,
+      "step": 20050
+    },
+    {
+      "epoch": 5.002493765586035,
+      "grad_norm": 6.531921863555908,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.406,
+      "step": 20060
+    },
+    {
+      "epoch": 5.00498753117207,
+      "grad_norm": 5.814652442932129,
+      "learning_rate": 1.4997506234413966e-05,
+      "loss": 0.4352,
+      "step": 20070
+    },
+    {
+      "epoch": 5.007481296758105,
+      "grad_norm": 3.371870279312134,
+      "learning_rate": 1.4995012468827933e-05,
+      "loss": 0.3459,
+      "step": 20080
+    },
+    {
+      "epoch": 5.0099750623441395,
+      "grad_norm": 4.824948787689209,
+      "learning_rate": 1.4992518703241896e-05,
+      "loss": 0.3522,
+      "step": 20090
+    },
+    {
+      "epoch": 5.012468827930174,
+      "grad_norm": 10.411859512329102,
+      "learning_rate": 1.4990024937655863e-05,
+      "loss": 0.4159,
+      "step": 20100
+    },
+    {
+      "epoch": 5.014962593516209,
+      "grad_norm": 7.6682562828063965,
+      "learning_rate": 1.4987531172069827e-05,
+      "loss": 0.3963,
+      "step": 20110
+    },
+    {
+      "epoch": 5.017456359102244,
+      "grad_norm": 7.620563983917236,
+      "learning_rate": 1.498503740648379e-05,
+      "loss": 0.3902,
+      "step": 20120
+    },
+    {
+      "epoch": 5.019950124688279,
+      "grad_norm": 7.573136806488037,
+      "learning_rate": 1.4982543640897758e-05,
+      "loss": 0.5429,
+      "step": 20130
+    },
+    {
+      "epoch": 5.022443890274314,
+      "grad_norm": 5.5496697425842285,
+      "learning_rate": 1.4980049875311721e-05,
+      "loss": 0.3934,
+      "step": 20140
+    },
+    {
+      "epoch": 5.024937655860349,
+      "grad_norm": 4.304521560668945,
+      "learning_rate": 1.4977556109725686e-05,
+      "loss": 0.3509,
+      "step": 20150
+    },
+    {
+      "epoch": 5.027431421446384,
+      "grad_norm": 9.141407012939453,
+      "learning_rate": 1.4975062344139652e-05,
+      "loss": 0.433,
+      "step": 20160
+    },
+    {
+      "epoch": 5.029925187032419,
+      "grad_norm": 5.6662678718566895,
+      "learning_rate": 1.4972568578553617e-05,
+      "loss": 0.3718,
+      "step": 20170
+    },
+    {
+      "epoch": 5.032418952618454,
+      "grad_norm": 7.040060997009277,
+      "learning_rate": 1.4970074812967582e-05,
+      "loss": 0.3879,
+      "step": 20180
+    },
+    {
+      "epoch": 5.034912718204489,
+      "grad_norm": 5.835836410522461,
+      "learning_rate": 1.4967581047381548e-05,
+      "loss": 0.3697,
+      "step": 20190
+    },
+    {
+      "epoch": 5.037406483790524,
+      "grad_norm": 3.57354998588562,
+      "learning_rate": 1.4965087281795511e-05,
+      "loss": 0.3932,
+      "step": 20200
+    },
+    {
+      "epoch": 5.039900249376559,
+      "grad_norm": 7.62668514251709,
+      "learning_rate": 1.4962593516209478e-05,
+      "loss": 0.3542,
+      "step": 20210
+    },
+    {
+      "epoch": 5.042394014962594,
+      "grad_norm": 5.238462924957275,
+      "learning_rate": 1.4960099750623442e-05,
+      "loss": 0.3125,
+      "step": 20220
+    },
+    {
+      "epoch": 5.0448877805486285,
+      "grad_norm": 6.3393235206604,
+      "learning_rate": 1.4957605985037407e-05,
+      "loss": 0.4034,
+      "step": 20230
+    },
+    {
+      "epoch": 5.047381546134663,
+      "grad_norm": 7.231442451477051,
+      "learning_rate": 1.4955112219451373e-05,
+      "loss": 0.3211,
+      "step": 20240
+    },
+    {
+      "epoch": 5.049875311720698,
+      "grad_norm": 6.194590091705322,
+      "learning_rate": 1.4952618453865338e-05,
+      "loss": 0.4005,
+      "step": 20250
+    },
+    {
+      "epoch": 5.052369077306733,
+      "grad_norm": 5.869102478027344,
+      "learning_rate": 1.4950124688279303e-05,
+      "loss": 0.4237,
+      "step": 20260
+    },
+    {
+      "epoch": 5.054862842892768,
+      "grad_norm": 6.967468738555908,
+      "learning_rate": 1.4947630922693268e-05,
+      "loss": 0.4117,
+      "step": 20270
+    },
+    {
+      "epoch": 5.057356608478803,
+      "grad_norm": 4.7718658447265625,
+      "learning_rate": 1.4945137157107232e-05,
+      "loss": 0.3834,
+      "step": 20280
+    },
+    {
+      "epoch": 5.059850374064838,
+      "grad_norm": 6.20999813079834,
+      "learning_rate": 1.4942643391521199e-05,
+      "loss": 0.3369,
+      "step": 20290
+    },
+    {
+      "epoch": 5.062344139650873,
+      "grad_norm": 5.620708465576172,
+      "learning_rate": 1.4940149625935163e-05,
+      "loss": 0.3699,
+      "step": 20300
+    },
+    {
+      "epoch": 5.0648379052369075,
+      "grad_norm": 9.380097389221191,
+      "learning_rate": 1.493765586034913e-05,
+      "loss": 0.3487,
+      "step": 20310
+    },
+    {
+      "epoch": 5.067331670822942,
+      "grad_norm": 5.500340938568115,
+      "learning_rate": 1.4935162094763093e-05,
+      "loss": 0.3864,
+      "step": 20320
+    },
+    {
+      "epoch": 5.069825436408977,
+      "grad_norm": 6.254300117492676,
+      "learning_rate": 1.4932668329177059e-05,
+      "loss": 0.3982,
+      "step": 20330
+    },
+    {
+      "epoch": 5.072319201995012,
+      "grad_norm": 7.905352592468262,
+      "learning_rate": 1.4930174563591024e-05,
+      "loss": 0.4213,
+      "step": 20340
+    },
+    {
+      "epoch": 5.074812967581048,
+      "grad_norm": 5.519759654998779,
+      "learning_rate": 1.492768079800499e-05,
+      "loss": 0.3563,
+      "step": 20350
+    },
+    {
+      "epoch": 5.077306733167083,
+      "grad_norm": 7.680830955505371,
+      "learning_rate": 1.4925187032418953e-05,
+      "loss": 0.3744,
+      "step": 20360
+    },
+    {
+      "epoch": 5.079800498753118,
+      "grad_norm": 5.350339889526367,
+      "learning_rate": 1.492269326683292e-05,
+      "loss": 0.361,
+      "step": 20370
+    },
+    {
+      "epoch": 5.082294264339152,
+      "grad_norm": 6.773752212524414,
+      "learning_rate": 1.4920199501246884e-05,
+      "loss": 0.4192,
+      "step": 20380
+    },
+    {
+      "epoch": 5.084788029925187,
+      "grad_norm": 7.636809825897217,
+      "learning_rate": 1.491770573566085e-05,
+      "loss": 0.3706,
+      "step": 20390
+    },
+    {
+      "epoch": 5.087281795511222,
+      "grad_norm": 8.180363655090332,
+      "learning_rate": 1.4915211970074814e-05,
+      "loss": 0.3809,
+      "step": 20400
+    },
+    {
+      "epoch": 5.089775561097257,
+      "grad_norm": 6.33188533782959,
+      "learning_rate": 1.491271820448878e-05,
+      "loss": 0.3553,
+      "step": 20410
+    },
+    {
+      "epoch": 5.092269326683292,
+      "grad_norm": 9.772364616394043,
+      "learning_rate": 1.4910224438902745e-05,
+      "loss": 0.4478,
+      "step": 20420
+    },
+    {
+      "epoch": 5.094763092269327,
+      "grad_norm": 8.910597801208496,
+      "learning_rate": 1.490773067331671e-05,
+      "loss": 0.3909,
+      "step": 20430
+    },
+    {
+      "epoch": 5.097256857855362,
+      "grad_norm": 5.329483509063721,
+      "learning_rate": 1.4905236907730674e-05,
+      "loss": 0.3552,
+      "step": 20440
+    },
+    {
+      "epoch": 5.0997506234413965,
+      "grad_norm": 8.715051651000977,
+      "learning_rate": 1.490274314214464e-05,
+      "loss": 0.3806,
+      "step": 20450
+    },
+    {
+      "epoch": 5.102244389027431,
+      "grad_norm": 4.4914422035217285,
+      "learning_rate": 1.4900249376558604e-05,
+      "loss": 0.3125,
+      "step": 20460
+    },
+    {
+      "epoch": 5.104738154613466,
+      "grad_norm": 8.738079071044922,
+      "learning_rate": 1.4897755610972571e-05,
+      "loss": 0.3728,
+      "step": 20470
+    },
+    {
+      "epoch": 5.107231920199501,
+      "grad_norm": 6.785305023193359,
+      "learning_rate": 1.4895261845386535e-05,
+      "loss": 0.387,
+      "step": 20480
+    },
+    {
+      "epoch": 5.109725685785536,
+      "grad_norm": 6.179448127746582,
+      "learning_rate": 1.4892768079800499e-05,
+      "loss": 0.3828,
+      "step": 20490
+    },
+    {
+      "epoch": 5.112219451371571,
+      "grad_norm": 8.303686141967773,
+      "learning_rate": 1.4890274314214466e-05,
+      "loss": 0.4654,
+      "step": 20500
+    },
+    {
+      "epoch": 5.114713216957606,
+      "grad_norm": 5.6898956298828125,
+      "learning_rate": 1.4887780548628429e-05,
+      "loss": 0.3792,
+      "step": 20510
+    },
+    {
+      "epoch": 5.117206982543641,
+      "grad_norm": 7.065920829772949,
+      "learning_rate": 1.4885286783042394e-05,
+      "loss": 0.4403,
+      "step": 20520
+    },
+    {
+      "epoch": 5.1197007481296755,
+      "grad_norm": 7.726041316986084,
+      "learning_rate": 1.4882793017456361e-05,
+      "loss": 0.4045,
+      "step": 20530
+    },
+    {
+      "epoch": 5.12219451371571,
+      "grad_norm": 7.321938991546631,
+      "learning_rate": 1.4880299251870325e-05,
+      "loss": 0.365,
+      "step": 20540
+    },
+    {
+      "epoch": 5.124688279301745,
+      "grad_norm": 5.387010097503662,
+      "learning_rate": 1.4877805486284292e-05,
+      "loss": 0.4085,
+      "step": 20550
+    },
+    {
+      "epoch": 5.127182044887781,
+      "grad_norm": 5.158405303955078,
+      "learning_rate": 1.4875311720698256e-05,
+      "loss": 0.3001,
+      "step": 20560
+    },
+    {
+      "epoch": 5.129675810473816,
+      "grad_norm": 6.263799667358398,
+      "learning_rate": 1.487281795511222e-05,
+      "loss": 0.3872,
+      "step": 20570
+    },
+    {
+      "epoch": 5.132169576059851,
+      "grad_norm": 6.450216293334961,
+      "learning_rate": 1.4870324189526186e-05,
+      "loss": 0.3789,
+      "step": 20580
+    },
+    {
+      "epoch": 5.134663341645886,
+      "grad_norm": 7.4370198249816895,
+      "learning_rate": 1.486783042394015e-05,
+      "loss": 0.3508,
+      "step": 20590
+    },
+    {
+      "epoch": 5.13715710723192,
+      "grad_norm": 7.416337013244629,
+      "learning_rate": 1.4865336658354117e-05,
+      "loss": 0.3011,
+      "step": 20600
+    },
+    {
+      "epoch": 5.139650872817955,
+      "grad_norm": 7.354401588439941,
+      "learning_rate": 1.486284289276808e-05,
+      "loss": 0.3835,
+      "step": 20610
+    },
+    {
+      "epoch": 5.14214463840399,
+      "grad_norm": 8.334373474121094,
+      "learning_rate": 1.4860349127182046e-05,
+      "loss": 0.3471,
+      "step": 20620
+    },
+    {
+      "epoch": 5.144638403990025,
+      "grad_norm": 6.268391132354736,
+      "learning_rate": 1.4857855361596011e-05,
+      "loss": 0.4272,
+      "step": 20630
+    },
+    {
+      "epoch": 5.14713216957606,
+      "grad_norm": 4.680897235870361,
+      "learning_rate": 1.4855361596009976e-05,
+      "loss": 0.4065,
+      "step": 20640
+    },
+    {
+      "epoch": 5.149625935162095,
+      "grad_norm": 4.887267112731934,
+      "learning_rate": 1.485286783042394e-05,
+      "loss": 0.4085,
+      "step": 20650
+    },
+    {
+      "epoch": 5.15211970074813,
+      "grad_norm": 6.771895408630371,
+      "learning_rate": 1.4850374064837907e-05,
+      "loss": 0.3727,
+      "step": 20660
+    },
+    {
+      "epoch": 5.1546134663341645,
+      "grad_norm": 7.750133037567139,
+      "learning_rate": 1.484788029925187e-05,
+      "loss": 0.321,
+      "step": 20670
+    },
+    {
+      "epoch": 5.157107231920199,
+      "grad_norm": 4.861640930175781,
+      "learning_rate": 1.4845386533665838e-05,
+      "loss": 0.4132,
+      "step": 20680
+    },
+    {
+      "epoch": 5.159600997506234,
+      "grad_norm": 4.334446907043457,
+      "learning_rate": 1.4842892768079801e-05,
+      "loss": 0.3547,
+      "step": 20690
+    },
+    {
+      "epoch": 5.162094763092269,
+      "grad_norm": 5.8802618980407715,
+      "learning_rate": 1.4840399002493767e-05,
+      "loss": 0.4131,
+      "step": 20700
+    },
+    {
+      "epoch": 5.164588528678304,
+      "grad_norm": 7.876018524169922,
+      "learning_rate": 1.4837905236907732e-05,
+      "loss": 0.3721,
+      "step": 20710
+    },
+    {
+      "epoch": 5.167082294264339,
+      "grad_norm": 6.277840614318848,
+      "learning_rate": 1.4835411471321697e-05,
+      "loss": 0.4001,
+      "step": 20720
+    },
+    {
+      "epoch": 5.169576059850374,
+      "grad_norm": 6.03588342666626,
+      "learning_rate": 1.4832917705735661e-05,
+      "loss": 0.3861,
+      "step": 20730
+    },
+    {
+      "epoch": 5.172069825436409,
+      "grad_norm": 17.586366653442383,
+      "learning_rate": 1.4830423940149628e-05,
+      "loss": 0.3589,
+      "step": 20740
+    },
+    {
+      "epoch": 5.174563591022444,
+      "grad_norm": 5.083337783813477,
+      "learning_rate": 1.4827930174563592e-05,
+      "loss": 0.3565,
+      "step": 20750
+    },
+    {
+      "epoch": 5.177057356608479,
+      "grad_norm": 7.520929336547852,
+      "learning_rate": 1.4825436408977559e-05,
+      "loss": 0.3924,
+      "step": 20760
+    },
+    {
+      "epoch": 5.179551122194514,
+      "grad_norm": 10.510026931762695,
+      "learning_rate": 1.4822942643391522e-05,
+      "loss": 0.3726,
+      "step": 20770
+    },
+    {
+      "epoch": 5.182044887780549,
+      "grad_norm": 7.425801753997803,
+      "learning_rate": 1.4820448877805487e-05,
+      "loss": 0.3746,
+      "step": 20780
+    },
+    {
+      "epoch": 5.184538653366584,
+      "grad_norm": 5.459587574005127,
+      "learning_rate": 1.4817955112219453e-05,
+      "loss": 0.3441,
+      "step": 20790
+    },
+    {
+      "epoch": 5.187032418952619,
+      "grad_norm": 5.989594459533691,
+      "learning_rate": 1.4815461346633418e-05,
+      "loss": 0.43,
+      "step": 20800
+    },
+    {
+      "epoch": 5.1895261845386536,
+      "grad_norm": 4.831539154052734,
+      "learning_rate": 1.4812967581047383e-05,
+      "loss": 0.4618,
+      "step": 20810
+    },
+    {
+      "epoch": 5.192019950124688,
+      "grad_norm": 5.510180473327637,
+      "learning_rate": 1.4810473815461349e-05,
+      "loss": 0.3506,
+      "step": 20820
+    },
+    {
+      "epoch": 5.194513715710723,
+      "grad_norm": 6.887777805328369,
+      "learning_rate": 1.4807980049875312e-05,
+      "loss": 0.3266,
+      "step": 20830
+    },
+    {
+      "epoch": 5.197007481296758,
+      "grad_norm": 6.549706935882568,
+      "learning_rate": 1.480548628428928e-05,
+      "loss": 0.4234,
+      "step": 20840
+    },
+    {
+      "epoch": 5.199501246882793,
+      "grad_norm": 4.920103073120117,
+      "learning_rate": 1.4802992518703243e-05,
+      "loss": 0.3687,
+      "step": 20850
+    },
+    {
+      "epoch": 5.201995012468828,
+      "grad_norm": 6.742281436920166,
+      "learning_rate": 1.4800498753117207e-05,
+      "loss": 0.4335,
+      "step": 20860
+    },
+    {
+      "epoch": 5.204488778054863,
+      "grad_norm": 11.688139915466309,
+      "learning_rate": 1.4798004987531174e-05,
+      "loss": 0.3504,
+      "step": 20870
+    },
+    {
+      "epoch": 5.206982543640898,
+      "grad_norm": 7.675383567810059,
+      "learning_rate": 1.4795511221945139e-05,
+      "loss": 0.3989,
+      "step": 20880
+    },
+    {
+      "epoch": 5.2094763092269325,
+      "grad_norm": 5.377547740936279,
+      "learning_rate": 1.4793017456359104e-05,
+      "loss": 0.4201,
+      "step": 20890
+    },
+    {
+      "epoch": 5.211970074812967,
+      "grad_norm": 6.458754539489746,
+      "learning_rate": 1.479052369077307e-05,
+      "loss": 0.4486,
+      "step": 20900
+    },
+    {
+      "epoch": 5.214463840399002,
+      "grad_norm": 6.1860551834106445,
+      "learning_rate": 1.4788029925187033e-05,
+      "loss": 0.3515,
+      "step": 20910
+    },
+    {
+      "epoch": 5.216957605985037,
+      "grad_norm": 5.372089862823486,
+      "learning_rate": 1.4785536159601e-05,
+      "loss": 0.3294,
+      "step": 20920
+    },
+    {
+      "epoch": 5.219451371571072,
+      "grad_norm": 7.561584949493408,
+      "learning_rate": 1.4783042394014964e-05,
+      "loss": 0.3811,
+      "step": 20930
+    },
+    {
+      "epoch": 5.221945137157107,
+      "grad_norm": 7.178483963012695,
+      "learning_rate": 1.4780548628428927e-05,
+      "loss": 0.4044,
+      "step": 20940
+    },
+    {
+      "epoch": 5.224438902743142,
+      "grad_norm": 4.487502098083496,
+      "learning_rate": 1.4778054862842894e-05,
+      "loss": 0.334,
+      "step": 20950
+    },
+    {
+      "epoch": 5.2269326683291775,
+      "grad_norm": 6.48040246963501,
+      "learning_rate": 1.4775561097256858e-05,
+      "loss": 0.3088,
+      "step": 20960
+    },
+    {
+      "epoch": 5.229426433915212,
+      "grad_norm": 6.952301979064941,
+      "learning_rate": 1.4773067331670825e-05,
+      "loss": 0.4753,
+      "step": 20970
+    },
+    {
+      "epoch": 5.231920199501247,
+      "grad_norm": 5.854089260101318,
+      "learning_rate": 1.4770573566084789e-05,
+      "loss": 0.3515,
+      "step": 20980
+    },
+    {
+      "epoch": 5.234413965087282,
+      "grad_norm": 5.807803630828857,
+      "learning_rate": 1.4768079800498754e-05,
+      "loss": 0.4071,
+      "step": 20990
+    },
+    {
+      "epoch": 5.236907730673317,
+      "grad_norm": 8.918316841125488,
+      "learning_rate": 1.4765586034912721e-05,
+      "loss": 0.3717,
+      "step": 21000
+    },
+    {
+      "epoch": 5.239401496259352,
+      "grad_norm": 5.254445552825928,
+      "learning_rate": 1.4763092269326684e-05,
+      "loss": 0.3259,
+      "step": 21010
+    },
+    {
+      "epoch": 5.241895261845387,
+      "grad_norm": 5.781581401824951,
+      "learning_rate": 1.4760598503740648e-05,
+      "loss": 0.3459,
+      "step": 21020
+    },
+    {
+      "epoch": 5.2443890274314215,
+      "grad_norm": 5.134096622467041,
+      "learning_rate": 1.4758104738154615e-05,
+      "loss": 0.377,
+      "step": 21030
+    },
+    {
+      "epoch": 5.246882793017456,
+      "grad_norm": 3.860678195953369,
+      "learning_rate": 1.4755610972568579e-05,
+      "loss": 0.3596,
+      "step": 21040
+    },
+    {
+      "epoch": 5.249376558603491,
+      "grad_norm": 5.5030083656311035,
+      "learning_rate": 1.4753117206982546e-05,
+      "loss": 0.3723,
+      "step": 21050
+    },
+    {
+      "epoch": 5.251870324189526,
+      "grad_norm": 8.171494483947754,
+      "learning_rate": 1.475062344139651e-05,
+      "loss": 0.3947,
+      "step": 21060
+    },
+    {
+      "epoch": 5.254364089775561,
+      "grad_norm": 5.602912425994873,
+      "learning_rate": 1.4748129675810475e-05,
+      "loss": 0.4279,
+      "step": 21070
+    },
+    {
+      "epoch": 5.256857855361596,
+      "grad_norm": 6.1152472496032715,
+      "learning_rate": 1.474563591022444e-05,
+      "loss": 0.3779,
+      "step": 21080
+    },
+    {
+      "epoch": 5.259351620947631,
+      "grad_norm": 5.26813268661499,
+      "learning_rate": 1.4743142144638405e-05,
+      "loss": 0.3819,
+      "step": 21090
+    },
+    {
+      "epoch": 5.261845386533666,
+      "grad_norm": 4.3624491691589355,
+      "learning_rate": 1.474064837905237e-05,
+      "loss": 0.3442,
+      "step": 21100
+    },
+    {
+      "epoch": 5.2643391521197005,
+      "grad_norm": 5.86679220199585,
+      "learning_rate": 1.4738154613466336e-05,
+      "loss": 0.3749,
+      "step": 21110
+    },
+    {
+      "epoch": 5.266832917705735,
+      "grad_norm": 7.071092128753662,
+      "learning_rate": 1.47356608478803e-05,
+      "loss": 0.3797,
+      "step": 21120
+    },
+    {
+      "epoch": 5.26932668329177,
+      "grad_norm": 5.458619117736816,
+      "learning_rate": 1.4733167082294266e-05,
+      "loss": 0.4415,
+      "step": 21130
+    },
+    {
+      "epoch": 5.271820448877805,
+      "grad_norm": 5.8463544845581055,
+      "learning_rate": 1.473067331670823e-05,
+      "loss": 0.3957,
+      "step": 21140
+    },
+    {
+      "epoch": 5.274314214463841,
+      "grad_norm": 9.386983871459961,
+      "learning_rate": 1.4728179551122195e-05,
+      "loss": 0.3172,
+      "step": 21150
+    },
+    {
+      "epoch": 5.276807980049876,
+      "grad_norm": 5.963527679443359,
+      "learning_rate": 1.472568578553616e-05,
+      "loss": 0.3782,
+      "step": 21160
+    },
+    {
+      "epoch": 5.279301745635911,
+      "grad_norm": 8.052948951721191,
+      "learning_rate": 1.4723192019950126e-05,
+      "loss": 0.4264,
+      "step": 21170
+    },
+    {
+      "epoch": 5.2817955112219455,
+      "grad_norm": 9.773006439208984,
+      "learning_rate": 1.4720698254364091e-05,
+      "loss": 0.3145,
+      "step": 21180
+    },
+    {
+      "epoch": 5.28428927680798,
+      "grad_norm": 6.8744330406188965,
+      "learning_rate": 1.4718204488778057e-05,
+      "loss": 0.3495,
+      "step": 21190
+    },
+    {
+      "epoch": 5.286783042394015,
+      "grad_norm": 4.006411075592041,
+      "learning_rate": 1.471571072319202e-05,
+      "loss": 0.3099,
+      "step": 21200
+    },
+    {
+      "epoch": 5.28927680798005,
+      "grad_norm": 5.908966541290283,
+      "learning_rate": 1.4713216957605987e-05,
+      "loss": 0.4145,
+      "step": 21210
+    },
+    {
+      "epoch": 5.291770573566085,
+      "grad_norm": 5.07480001449585,
+      "learning_rate": 1.4710723192019951e-05,
+      "loss": 0.3299,
+      "step": 21220
+    },
+    {
+      "epoch": 5.29426433915212,
+      "grad_norm": 5.2527995109558105,
+      "learning_rate": 1.4708229426433916e-05,
+      "loss": 0.4364,
+      "step": 21230
+    },
+    {
+      "epoch": 5.296758104738155,
+      "grad_norm": 4.9071526527404785,
+      "learning_rate": 1.4705735660847882e-05,
+      "loss": 0.4228,
+      "step": 21240
+    },
+    {
+      "epoch": 5.2992518703241895,
+      "grad_norm": 7.707085609436035,
+      "learning_rate": 1.4703241895261847e-05,
+      "loss": 0.3715,
+      "step": 21250
+    },
+    {
+      "epoch": 5.301745635910224,
+      "grad_norm": 10.137189865112305,
+      "learning_rate": 1.4700748129675812e-05,
+      "loss": 0.3773,
+      "step": 21260
+    },
+    {
+      "epoch": 5.304239401496259,
+      "grad_norm": 6.27385950088501,
+      "learning_rate": 1.4698254364089777e-05,
+      "loss": 0.4286,
+      "step": 21270
+    },
+    {
+      "epoch": 5.306733167082294,
+      "grad_norm": 6.91831636428833,
+      "learning_rate": 1.4695760598503741e-05,
+      "loss": 0.3541,
+      "step": 21280
+    },
+    {
+      "epoch": 5.309226932668329,
+      "grad_norm": 4.474343299865723,
+      "learning_rate": 1.4693266832917708e-05,
+      "loss": 0.341,
+      "step": 21290
+    },
+    {
+      "epoch": 5.311720698254364,
+      "grad_norm": 6.206080913543701,
+      "learning_rate": 1.4690773067331672e-05,
+      "loss": 0.386,
+      "step": 21300
+    },
+    {
+      "epoch": 5.314214463840399,
+      "grad_norm": 5.673043251037598,
+      "learning_rate": 1.4688279301745639e-05,
+      "loss": 0.3751,
+      "step": 21310
+    },
+    {
+      "epoch": 5.316708229426434,
+      "grad_norm": 6.9373955726623535,
+      "learning_rate": 1.4685785536159602e-05,
+      "loss": 0.4205,
+      "step": 21320
+    },
+    {
+      "epoch": 5.3192019950124685,
+      "grad_norm": 5.714535713195801,
+      "learning_rate": 1.4683291770573566e-05,
+      "loss": 0.4316,
+      "step": 21330
+    },
+    {
+      "epoch": 5.321695760598503,
+      "grad_norm": 5.814162254333496,
+      "learning_rate": 1.4680798004987533e-05,
+      "loss": 0.4013,
+      "step": 21340
+    },
+    {
+      "epoch": 5.324189526184538,
+      "grad_norm": 5.710855484008789,
+      "learning_rate": 1.4678304239401498e-05,
+      "loss": 0.381,
+      "step": 21350
+    },
+    {
+      "epoch": 5.326683291770574,
+      "grad_norm": 4.715007781982422,
+      "learning_rate": 1.4675810473815462e-05,
+      "loss": 0.3644,
+      "step": 21360
+    },
+    {
+      "epoch": 5.329177057356609,
+      "grad_norm": 6.894988536834717,
+      "learning_rate": 1.4673316708229429e-05,
+      "loss": 0.3702,
+      "step": 21370
+    },
+    {
+      "epoch": 5.331670822942644,
+      "grad_norm": 5.765067100524902,
+      "learning_rate": 1.4670822942643392e-05,
+      "loss": 0.3833,
+      "step": 21380
+    },
+    {
+      "epoch": 5.334164588528679,
+      "grad_norm": 5.794302940368652,
+      "learning_rate": 1.466832917705736e-05,
+      "loss": 0.3816,
+      "step": 21390
+    },
+    {
+      "epoch": 5.3366583541147135,
+      "grad_norm": 7.873318195343018,
+      "learning_rate": 1.4665835411471323e-05,
+      "loss": 0.4156,
+      "step": 21400
+    },
+    {
+      "epoch": 5.339152119700748,
+      "grad_norm": 6.067409038543701,
+      "learning_rate": 1.4663341645885287e-05,
+      "loss": 0.3925,
+      "step": 21410
+    },
+    {
+      "epoch": 5.341645885286783,
+      "grad_norm": 4.952661514282227,
+      "learning_rate": 1.4660847880299254e-05,
+      "loss": 0.3794,
+      "step": 21420
+    },
+    {
+      "epoch": 5.344139650872818,
+      "grad_norm": 8.906375885009766,
+      "learning_rate": 1.4658354114713217e-05,
+      "loss": 0.4743,
+      "step": 21430
+    },
+    {
+      "epoch": 5.346633416458853,
+      "grad_norm": 5.5464630126953125,
+      "learning_rate": 1.4655860349127183e-05,
+      "loss": 0.3503,
+      "step": 21440
+    },
+    {
+      "epoch": 5.349127182044888,
+      "grad_norm": 4.894265651702881,
+      "learning_rate": 1.4653366583541148e-05,
+      "loss": 0.3889,
+      "step": 21450
+    },
+    {
+      "epoch": 5.351620947630923,
+      "grad_norm": 5.675085544586182,
+      "learning_rate": 1.4650872817955113e-05,
+      "loss": 0.4552,
+      "step": 21460
+    },
+    {
+      "epoch": 5.3541147132169575,
+      "grad_norm": 5.673407077789307,
+      "learning_rate": 1.464837905236908e-05,
+      "loss": 0.4368,
+      "step": 21470
+    },
+    {
+      "epoch": 5.356608478802992,
+      "grad_norm": 6.846253395080566,
+      "learning_rate": 1.4645885286783044e-05,
+      "loss": 0.3714,
+      "step": 21480
+    },
+    {
+      "epoch": 5.359102244389027,
+      "grad_norm": 7.071938991546631,
+      "learning_rate": 1.4643391521197007e-05,
+      "loss": 0.453,
+      "step": 21490
+    },
+    {
+      "epoch": 5.361596009975062,
+      "grad_norm": 5.694882392883301,
+      "learning_rate": 1.4640897755610974e-05,
+      "loss": 0.3199,
+      "step": 21500
+    },
+    {
+      "epoch": 5.364089775561097,
+      "grad_norm": 6.255917072296143,
+      "learning_rate": 1.4638403990024938e-05,
+      "loss": 0.3442,
+      "step": 21510
+    },
+    {
+      "epoch": 5.366583541147132,
+      "grad_norm": 5.694520950317383,
+      "learning_rate": 1.4635910224438903e-05,
+      "loss": 0.366,
+      "step": 21520
+    },
+    {
+      "epoch": 5.369077306733167,
+      "grad_norm": 4.848544597625732,
+      "learning_rate": 1.4633416458852869e-05,
+      "loss": 0.3746,
+      "step": 21530
+    },
+    {
+      "epoch": 5.371571072319202,
+      "grad_norm": 7.683907508850098,
+      "learning_rate": 1.4630922693266834e-05,
+      "loss": 0.401,
+      "step": 21540
+    },
+    {
+      "epoch": 5.374064837905237,
+      "grad_norm": 7.421435832977295,
+      "learning_rate": 1.46284289276808e-05,
+      "loss": 0.4114,
+      "step": 21550
+    },
+    {
+      "epoch": 5.376558603491272,
+      "grad_norm": 5.23777437210083,
+      "learning_rate": 1.4625935162094765e-05,
+      "loss": 0.3773,
+      "step": 21560
+    },
+    {
+      "epoch": 5.379052369077307,
+      "grad_norm": 10.150872230529785,
+      "learning_rate": 1.4623441396508728e-05,
+      "loss": 0.4342,
+      "step": 21570
+    },
+    {
+      "epoch": 5.381546134663342,
+      "grad_norm": 6.791860103607178,
+      "learning_rate": 1.4621197007481298e-05,
+      "loss": 0.3954,
+      "step": 21580
+    },
+    {
+      "epoch": 5.384039900249377,
+      "grad_norm": 7.470459461212158,
+      "learning_rate": 1.4618703241895262e-05,
+      "loss": 0.3082,
+      "step": 21590
+    },
+    {
+      "epoch": 5.386533665835412,
+      "grad_norm": 7.525590419769287,
+      "learning_rate": 1.4616209476309229e-05,
+      "loss": 0.4149,
+      "step": 21600
+    },
+    {
+      "epoch": 5.389027431421447,
+      "grad_norm": 8.42238998413086,
+      "learning_rate": 1.4613715710723192e-05,
+      "loss": 0.3672,
+      "step": 21610
+    },
+    {
+      "epoch": 5.3915211970074814,
+      "grad_norm": 5.296651363372803,
+      "learning_rate": 1.4611221945137158e-05,
+      "loss": 0.4263,
+      "step": 21620
+    },
+    {
+      "epoch": 5.394014962593516,
+      "grad_norm": 4.646917343139648,
+      "learning_rate": 1.4608728179551123e-05,
+      "loss": 0.357,
+      "step": 21630
+    },
+    {
+      "epoch": 5.396508728179551,
+      "grad_norm": 7.128079414367676,
+      "learning_rate": 1.4606234413965088e-05,
+      "loss": 0.3383,
+      "step": 21640
+    },
+    {
+      "epoch": 5.399002493765586,
+      "grad_norm": 6.30106782913208,
+      "learning_rate": 1.4603740648379054e-05,
+      "loss": 0.4503,
+      "step": 21650
+    },
+    {
+      "epoch": 5.401496259351621,
+      "grad_norm": 3.676443576812744,
+      "learning_rate": 1.4601246882793019e-05,
+      "loss": 0.3028,
+      "step": 21660
+    },
+    {
+      "epoch": 5.403990024937656,
+      "grad_norm": 6.699869155883789,
+      "learning_rate": 1.4598753117206983e-05,
+      "loss": 0.3205,
+      "step": 21670
+    },
+    {
+      "epoch": 5.406483790523691,
+      "grad_norm": 7.279172897338867,
+      "learning_rate": 1.459625935162095e-05,
+      "loss": 0.3449,
+      "step": 21680
+    },
+    {
+      "epoch": 5.4089775561097255,
+      "grad_norm": 5.297791004180908,
+      "learning_rate": 1.4593765586034913e-05,
+      "loss": 0.3866,
+      "step": 21690
+    },
+    {
+      "epoch": 5.41147132169576,
+      "grad_norm": 4.082034587860107,
+      "learning_rate": 1.4591271820448879e-05,
+      "loss": 0.3785,
+      "step": 21700
+    },
+    {
+      "epoch": 5.413965087281795,
+      "grad_norm": 4.643633842468262,
+      "learning_rate": 1.4588778054862844e-05,
+      "loss": 0.3734,
+      "step": 21710
+    },
+    {
+      "epoch": 5.41645885286783,
+      "grad_norm": 5.9492597579956055,
+      "learning_rate": 1.458628428927681e-05,
+      "loss": 0.4107,
+      "step": 21720
+    },
+    {
+      "epoch": 5.418952618453865,
+      "grad_norm": 6.517092704772949,
+      "learning_rate": 1.4583790523690774e-05,
+      "loss": 0.4124,
+      "step": 21730
+    },
+    {
+      "epoch": 5.4214463840399,
+      "grad_norm": 5.166103363037109,
+      "learning_rate": 1.458129675810474e-05,
+      "loss": 0.3856,
+      "step": 21740
+    },
+    {
+      "epoch": 5.423940149625935,
+      "grad_norm": 7.804215431213379,
+      "learning_rate": 1.4578802992518703e-05,
+      "loss": 0.3939,
+      "step": 21750
+    },
+    {
+      "epoch": 5.42643391521197,
+      "grad_norm": 9.893004417419434,
+      "learning_rate": 1.457630922693267e-05,
+      "loss": 0.3276,
+      "step": 21760
+    },
+    {
+      "epoch": 5.428927680798005,
+      "grad_norm": 6.463005065917969,
+      "learning_rate": 1.4573815461346634e-05,
+      "loss": 0.3526,
+      "step": 21770
+    },
+    {
+      "epoch": 5.43142144638404,
+      "grad_norm": 6.165153980255127,
+      "learning_rate": 1.4571321695760601e-05,
+      "loss": 0.4449,
+      "step": 21780
+    },
+    {
+      "epoch": 5.433915211970075,
+      "grad_norm": 4.93517541885376,
+      "learning_rate": 1.4568827930174565e-05,
+      "loss": 0.328,
+      "step": 21790
+    },
+    {
+      "epoch": 5.43640897755611,
+      "grad_norm": 5.643067359924316,
+      "learning_rate": 1.456633416458853e-05,
+      "loss": 0.4141,
+      "step": 21800
+    },
+    {
+      "epoch": 5.438902743142145,
+      "grad_norm": 7.786581039428711,
+      "learning_rate": 1.4563840399002495e-05,
+      "loss": 0.4098,
+      "step": 21810
+    },
+    {
+      "epoch": 5.44139650872818,
+      "grad_norm": 4.224101543426514,
+      "learning_rate": 1.456134663341646e-05,
+      "loss": 0.4034,
+      "step": 21820
+    },
+    {
+      "epoch": 5.443890274314215,
+      "grad_norm": 8.033632278442383,
+      "learning_rate": 1.4558852867830424e-05,
+      "loss": 0.3504,
+      "step": 21830
+    },
+    {
+      "epoch": 5.446384039900249,
+      "grad_norm": 6.644708156585693,
+      "learning_rate": 1.4556359102244391e-05,
+      "loss": 0.335,
+      "step": 21840
+    },
+    {
+      "epoch": 5.448877805486284,
+      "grad_norm": 8.7880859375,
+      "learning_rate": 1.4553865336658355e-05,
+      "loss": 0.3907,
+      "step": 21850
+    },
+    {
+      "epoch": 5.451371571072319,
+      "grad_norm": 8.938055038452148,
+      "learning_rate": 1.4551371571072322e-05,
+      "loss": 0.3681,
+      "step": 21860
+    },
+    {
+      "epoch": 5.453865336658354,
+      "grad_norm": 5.602493762969971,
+      "learning_rate": 1.4548877805486285e-05,
+      "loss": 0.4231,
+      "step": 21870
+    },
+    {
+      "epoch": 5.456359102244389,
+      "grad_norm": 6.839218616485596,
+      "learning_rate": 1.4546384039900249e-05,
+      "loss": 0.3974,
+      "step": 21880
+    },
+    {
+      "epoch": 5.458852867830424,
+      "grad_norm": 6.831507682800293,
+      "learning_rate": 1.4543890274314216e-05,
+      "loss": 0.4212,
+      "step": 21890
+    },
+    {
+      "epoch": 5.461346633416459,
+      "grad_norm": 6.885910987854004,
+      "learning_rate": 1.4541396508728181e-05,
+      "loss": 0.3996,
+      "step": 21900
+    },
+    {
+      "epoch": 5.4638403990024935,
+      "grad_norm": 9.154892921447754,
+      "learning_rate": 1.4538902743142145e-05,
+      "loss": 0.4248,
+      "step": 21910
+    },
+    {
+      "epoch": 5.466334164588528,
+      "grad_norm": 5.8164591789245605,
+      "learning_rate": 1.4536408977556112e-05,
+      "loss": 0.3866,
+      "step": 21920
+    },
+    {
+      "epoch": 5.468827930174563,
+      "grad_norm": 7.783123016357422,
+      "learning_rate": 1.4533915211970076e-05,
+      "loss": 0.3988,
+      "step": 21930
+    },
+    {
+      "epoch": 5.471321695760598,
+      "grad_norm": 8.292793273925781,
+      "learning_rate": 1.4531421446384043e-05,
+      "loss": 0.5072,
+      "step": 21940
+    },
+    {
+      "epoch": 5.473815461346634,
+      "grad_norm": 8.550527572631836,
+      "learning_rate": 1.4528927680798006e-05,
+      "loss": 0.4445,
+      "step": 21950
+    },
+    {
+      "epoch": 5.476309226932669,
+      "grad_norm": 7.110447883605957,
+      "learning_rate": 1.452643391521197e-05,
+      "loss": 0.3796,
+      "step": 21960
+    },
+    {
+      "epoch": 5.478802992518704,
+      "grad_norm": 5.753549575805664,
+      "learning_rate": 1.4523940149625937e-05,
+      "loss": 0.3745,
+      "step": 21970
+    },
+    {
+      "epoch": 5.4812967581047385,
+      "grad_norm": 12.874857902526855,
+      "learning_rate": 1.45214463840399e-05,
+      "loss": 0.3587,
+      "step": 21980
+    },
+    {
+      "epoch": 5.483790523690773,
+      "grad_norm": 7.264901161193848,
+      "learning_rate": 1.4518952618453867e-05,
+      "loss": 0.3727,
+      "step": 21990
+    },
+    {
+      "epoch": 5.486284289276808,
+      "grad_norm": 6.658206939697266,
+      "learning_rate": 1.4516458852867831e-05,
+      "loss": 0.4147,
+      "step": 22000
+    },
+    {
+      "epoch": 5.488778054862843,
+      "grad_norm": 8.003405570983887,
+      "learning_rate": 1.4513965087281796e-05,
+      "loss": 0.4492,
+      "step": 22010
+    },
+    {
+      "epoch": 5.491271820448878,
+      "grad_norm": 8.046963691711426,
+      "learning_rate": 1.4511471321695763e-05,
+      "loss": 0.3884,
+      "step": 22020
+    },
+    {
+      "epoch": 5.493765586034913,
+      "grad_norm": 4.5463385581970215,
+      "learning_rate": 1.4508977556109727e-05,
+      "loss": 0.3515,
+      "step": 22030
+    },
+    {
+      "epoch": 5.496259351620948,
+      "grad_norm": 7.268316268920898,
+      "learning_rate": 1.450648379052369e-05,
+      "loss": 0.3548,
+      "step": 22040
+    },
+    {
+      "epoch": 5.498753117206983,
+      "grad_norm": 6.258470058441162,
+      "learning_rate": 1.4503990024937658e-05,
+      "loss": 0.3813,
+      "step": 22050
+    },
+    {
+      "epoch": 5.501246882793017,
+      "grad_norm": 5.818022727966309,
+      "learning_rate": 1.4501496259351621e-05,
+      "loss": 0.4218,
+      "step": 22060
+    },
+    {
+      "epoch": 5.503740648379052,
+      "grad_norm": 5.728906631469727,
+      "learning_rate": 1.4499002493765588e-05,
+      "loss": 0.3664,
+      "step": 22070
+    },
+    {
+      "epoch": 5.506234413965087,
+      "grad_norm": 6.165427207946777,
+      "learning_rate": 1.4496508728179552e-05,
+      "loss": 0.3949,
+      "step": 22080
+    },
+    {
+      "epoch": 5.508728179551122,
+      "grad_norm": 6.8010783195495605,
+      "learning_rate": 1.4494014962593517e-05,
+      "loss": 0.3695,
+      "step": 22090
+    },
+    {
+      "epoch": 5.511221945137157,
+      "grad_norm": 5.022993087768555,
+      "learning_rate": 1.4491521197007482e-05,
+      "loss": 0.2954,
+      "step": 22100
+    },
+    {
+      "epoch": 5.513715710723192,
+      "grad_norm": 8.099823951721191,
+      "learning_rate": 1.4489027431421448e-05,
+      "loss": 0.3272,
+      "step": 22110
+    },
+    {
+      "epoch": 5.516209476309227,
+      "grad_norm": 7.62611198425293,
+      "learning_rate": 1.4486533665835411e-05,
+      "loss": 0.4664,
+      "step": 22120
+    },
+    {
+      "epoch": 5.5187032418952615,
+      "grad_norm": 9.035438537597656,
+      "learning_rate": 1.4484039900249378e-05,
+      "loss": 0.4082,
+      "step": 22130
+    },
+    {
+      "epoch": 5.521197007481296,
+      "grad_norm": 5.160449028015137,
+      "learning_rate": 1.4481546134663342e-05,
+      "loss": 0.4821,
+      "step": 22140
+    },
+    {
+      "epoch": 5.523690773067331,
+      "grad_norm": 5.530974388122559,
+      "learning_rate": 1.4479052369077309e-05,
+      "loss": 0.4418,
+      "step": 22150
+    },
+    {
+      "epoch": 5.526184538653366,
+      "grad_norm": 7.125683784484863,
+      "learning_rate": 1.4476558603491273e-05,
+      "loss": 0.358,
+      "step": 22160
+    },
+    {
+      "epoch": 5.528678304239402,
+      "grad_norm": 5.76719331741333,
+      "learning_rate": 1.4474064837905238e-05,
+      "loss": 0.3801,
+      "step": 22170
+    },
+    {
+      "epoch": 5.531172069825437,
+      "grad_norm": 5.448780536651611,
+      "learning_rate": 1.4471571072319203e-05,
+      "loss": 0.3539,
+      "step": 22180
+    },
+    {
+      "epoch": 5.533665835411472,
+      "grad_norm": 5.663511753082275,
+      "learning_rate": 1.4469077306733169e-05,
+      "loss": 0.451,
+      "step": 22190
+    },
+    {
+      "epoch": 5.5361596009975065,
+      "grad_norm": 5.449460983276367,
+      "learning_rate": 1.4466583541147132e-05,
+      "loss": 0.3879,
+      "step": 22200
+    },
+    {
+      "epoch": 5.538653366583541,
+      "grad_norm": 6.016726493835449,
+      "learning_rate": 1.44640897755611e-05,
+      "loss": 0.4061,
+      "step": 22210
+    },
+    {
+      "epoch": 5.541147132169576,
+      "grad_norm": 5.542261123657227,
+      "learning_rate": 1.4461596009975063e-05,
+      "loss": 0.3877,
+      "step": 22220
+    },
+    {
+      "epoch": 5.543640897755611,
+      "grad_norm": 8.105876922607422,
+      "learning_rate": 1.445910224438903e-05,
+      "loss": 0.3574,
+      "step": 22230
+    },
+    {
+      "epoch": 5.546134663341646,
+      "grad_norm": 6.1067914962768555,
+      "learning_rate": 1.4456608478802993e-05,
+      "loss": 0.4374,
+      "step": 22240
+    },
+    {
+      "epoch": 5.548628428927681,
+      "grad_norm": 7.3289265632629395,
+      "learning_rate": 1.4454114713216959e-05,
+      "loss": 0.3598,
+      "step": 22250
+    },
+    {
+      "epoch": 5.551122194513716,
+      "grad_norm": 6.727115154266357,
+      "learning_rate": 1.4451620947630924e-05,
+      "loss": 0.4083,
+      "step": 22260
+    },
+    {
+      "epoch": 5.553615960099751,
+      "grad_norm": 9.422988891601562,
+      "learning_rate": 1.444912718204489e-05,
+      "loss": 0.4603,
+      "step": 22270
+    },
+    {
+      "epoch": 5.556109725685785,
+      "grad_norm": 8.968878746032715,
+      "learning_rate": 1.4446633416458855e-05,
+      "loss": 0.4231,
+      "step": 22280
+    },
+    {
+      "epoch": 5.55860349127182,
+      "grad_norm": 5.541744232177734,
+      "learning_rate": 1.444413965087282e-05,
+      "loss": 0.3105,
+      "step": 22290
+    },
+    {
+      "epoch": 5.561097256857855,
+      "grad_norm": 5.500674724578857,
+      "learning_rate": 1.4441645885286784e-05,
+      "loss": 0.3357,
+      "step": 22300
+    },
+    {
+      "epoch": 5.56359102244389,
+      "grad_norm": 9.031492233276367,
+      "learning_rate": 1.443915211970075e-05,
+      "loss": 0.582,
+      "step": 22310
+    },
+    {
+      "epoch": 5.566084788029925,
+      "grad_norm": 9.218156814575195,
+      "learning_rate": 1.4436658354114714e-05,
+      "loss": 0.3842,
+      "step": 22320
+    },
+    {
+      "epoch": 5.56857855361596,
+      "grad_norm": 7.594570636749268,
+      "learning_rate": 1.4434164588528678e-05,
+      "loss": 0.4222,
+      "step": 22330
+    },
+    {
+      "epoch": 5.571072319201995,
+      "grad_norm": 5.0406270027160645,
+      "learning_rate": 1.4431670822942645e-05,
+      "loss": 0.4213,
+      "step": 22340
+    },
+    {
+      "epoch": 5.57356608478803,
+      "grad_norm": 5.456195831298828,
+      "learning_rate": 1.4429177057356608e-05,
+      "loss": 0.3552,
+      "step": 22350
+    },
+    {
+      "epoch": 5.576059850374065,
+      "grad_norm": 6.551260948181152,
+      "learning_rate": 1.4426683291770575e-05,
+      "loss": 0.5219,
+      "step": 22360
+    },
+    {
+      "epoch": 5.5785536159601,
+      "grad_norm": 6.80320405960083,
+      "learning_rate": 1.442418952618454e-05,
+      "loss": 0.3416,
+      "step": 22370
+    },
+    {
+      "epoch": 5.581047381546135,
+      "grad_norm": 5.928180694580078,
+      "learning_rate": 1.4421695760598504e-05,
+      "loss": 0.3299,
+      "step": 22380
+    },
+    {
+      "epoch": 5.58354114713217,
+      "grad_norm": 4.945070266723633,
+      "learning_rate": 1.4419201995012471e-05,
+      "loss": 0.3556,
+      "step": 22390
+    },
+    {
+      "epoch": 5.586034912718205,
+      "grad_norm": 6.342123031616211,
+      "learning_rate": 1.4416708229426435e-05,
+      "loss": 0.427,
+      "step": 22400
+    },
+    {
+      "epoch": 5.58852867830424,
+      "grad_norm": 6.5954060554504395,
+      "learning_rate": 1.4414214463840399e-05,
+      "loss": 0.4065,
+      "step": 22410
+    },
+    {
+      "epoch": 5.5910224438902745,
+      "grad_norm": 5.893113136291504,
+      "learning_rate": 1.4411720698254366e-05,
+      "loss": 0.4155,
+      "step": 22420
+    },
+    {
+      "epoch": 5.593516209476309,
+      "grad_norm": 6.298617839813232,
+      "learning_rate": 1.440922693266833e-05,
+      "loss": 0.3652,
+      "step": 22430
+    },
+    {
+      "epoch": 5.596009975062344,
+      "grad_norm": 5.96094274520874,
+      "learning_rate": 1.4406733167082296e-05,
+      "loss": 0.4139,
+      "step": 22440
+    },
+    {
+      "epoch": 5.598503740648379,
+      "grad_norm": 5.680459022521973,
+      "learning_rate": 1.440423940149626e-05,
+      "loss": 0.3219,
+      "step": 22450
+    },
+    {
+      "epoch": 5.600997506234414,
+      "grad_norm": 6.752685546875,
+      "learning_rate": 1.4401745635910225e-05,
+      "loss": 0.3695,
+      "step": 22460
+    },
+    {
+      "epoch": 5.603491271820449,
+      "grad_norm": 6.382385730743408,
+      "learning_rate": 1.439925187032419e-05,
+      "loss": 0.3797,
+      "step": 22470
+    },
+    {
+      "epoch": 5.605985037406484,
+      "grad_norm": 7.606115341186523,
+      "learning_rate": 1.4396758104738156e-05,
+      "loss": 0.3709,
+      "step": 22480
+    },
+    {
+      "epoch": 5.6084788029925186,
+      "grad_norm": 9.14214038848877,
+      "learning_rate": 1.4394264339152123e-05,
+      "loss": 0.3548,
+      "step": 22490
+    },
+    {
+      "epoch": 5.610972568578553,
+      "grad_norm": 4.893841743469238,
+      "learning_rate": 1.4391770573566086e-05,
+      "loss": 0.389,
+      "step": 22500
+    },
+    {
+      "epoch": 5.613466334164588,
+      "grad_norm": 7.275263786315918,
+      "learning_rate": 1.438927680798005e-05,
+      "loss": 0.3461,
+      "step": 22510
+    },
+    {
+      "epoch": 5.615960099750623,
+      "grad_norm": 5.975040435791016,
+      "learning_rate": 1.4386783042394017e-05,
+      "loss": 0.4767,
+      "step": 22520
+    },
+    {
+      "epoch": 5.618453865336658,
+      "grad_norm": 6.58006477355957,
+      "learning_rate": 1.438428927680798e-05,
+      "loss": 0.3466,
+      "step": 22530
+    },
+    {
+      "epoch": 5.620947630922693,
+      "grad_norm": 5.844588756561279,
+      "learning_rate": 1.4381795511221946e-05,
+      "loss": 0.3944,
+      "step": 22540
+    },
+    {
+      "epoch": 5.623441396508728,
+      "grad_norm": 12.84826374053955,
+      "learning_rate": 1.4379301745635911e-05,
+      "loss": 0.3679,
+      "step": 22550
+    },
+    {
+      "epoch": 5.625935162094763,
+      "grad_norm": 5.966850757598877,
+      "learning_rate": 1.4376807980049877e-05,
+      "loss": 0.3632,
+      "step": 22560
+    },
+    {
+      "epoch": 5.628428927680798,
+      "grad_norm": 5.235140323638916,
+      "learning_rate": 1.4374314214463842e-05,
+      "loss": 0.3889,
+      "step": 22570
+    },
+    {
+      "epoch": 5.630922693266833,
+      "grad_norm": 5.83263635635376,
+      "learning_rate": 1.4371820448877807e-05,
+      "loss": 0.3315,
+      "step": 22580
+    },
+    {
+      "epoch": 5.633416458852868,
+      "grad_norm": 4.937515735626221,
+      "learning_rate": 1.436932668329177e-05,
+      "loss": 0.3964,
+      "step": 22590
+    },
+    {
+      "epoch": 5.635910224438903,
+      "grad_norm": 6.0770392417907715,
+      "learning_rate": 1.4366832917705738e-05,
+      "loss": 0.3959,
+      "step": 22600
+    },
+    {
+      "epoch": 5.638403990024938,
+      "grad_norm": 7.637709140777588,
+      "learning_rate": 1.4364339152119701e-05,
+      "loss": 0.402,
+      "step": 22610
+    },
+    {
+      "epoch": 5.640897755610973,
+      "grad_norm": 5.632735252380371,
+      "learning_rate": 1.4361845386533667e-05,
+      "loss": 0.3721,
+      "step": 22620
+    },
+    {
+      "epoch": 5.643391521197008,
+      "grad_norm": 7.393962860107422,
+      "learning_rate": 1.4359351620947632e-05,
+      "loss": 0.3827,
+      "step": 22630
+    },
+    {
+      "epoch": 5.6458852867830425,
+      "grad_norm": 7.217513561248779,
+      "learning_rate": 1.4356857855361597e-05,
+      "loss": 0.3942,
+      "step": 22640
+    },
+    {
+      "epoch": 5.648379052369077,
+      "grad_norm": 8.804829597473145,
+      "learning_rate": 1.4354364089775563e-05,
+      "loss": 0.4314,
+      "step": 22650
+    },
+    {
+      "epoch": 5.650872817955112,
+      "grad_norm": 9.382487297058105,
+      "learning_rate": 1.4351870324189528e-05,
+      "loss": 0.552,
+      "step": 22660
+    },
+    {
+      "epoch": 5.653366583541147,
+      "grad_norm": 4.75145959854126,
+      "learning_rate": 1.4349376558603492e-05,
+      "loss": 0.3351,
+      "step": 22670
+    },
+    {
+      "epoch": 5.655860349127182,
+      "grad_norm": 5.727991104125977,
+      "learning_rate": 1.4346882793017459e-05,
+      "loss": 0.3271,
+      "step": 22680
+    },
+    {
+      "epoch": 5.658354114713217,
+      "grad_norm": 6.325864791870117,
+      "learning_rate": 1.4344389027431422e-05,
+      "loss": 0.3485,
+      "step": 22690
+    },
+    {
+      "epoch": 5.660847880299252,
+      "grad_norm": 6.61782169342041,
+      "learning_rate": 1.4341895261845386e-05,
+      "loss": 0.3918,
+      "step": 22700
+    },
+    {
+      "epoch": 5.6633416458852865,
+      "grad_norm": 8.966527938842773,
+      "learning_rate": 1.4339401496259353e-05,
+      "loss": 0.4352,
+      "step": 22710
+    },
+    {
+      "epoch": 5.665835411471321,
+      "grad_norm": 3.9728758335113525,
+      "learning_rate": 1.4336907730673318e-05,
+      "loss": 0.3993,
+      "step": 22720
+    },
+    {
+      "epoch": 5.668329177057356,
+      "grad_norm": 4.767123222351074,
+      "learning_rate": 1.4334413965087283e-05,
+      "loss": 0.3053,
+      "step": 22730
+    },
+    {
+      "epoch": 5.670822942643391,
+      "grad_norm": 5.514127254486084,
+      "learning_rate": 1.4331920199501249e-05,
+      "loss": 0.3162,
+      "step": 22740
+    },
+    {
+      "epoch": 5.673316708229427,
+      "grad_norm": 5.728281497955322,
+      "learning_rate": 1.4329426433915212e-05,
+      "loss": 0.3697,
+      "step": 22750
+    },
+    {
+      "epoch": 5.675810473815462,
+      "grad_norm": 5.388779640197754,
+      "learning_rate": 1.432693266832918e-05,
+      "loss": 0.3806,
+      "step": 22760
+    },
+    {
+      "epoch": 5.678304239401497,
+      "grad_norm": 5.20200252532959,
+      "learning_rate": 1.4324438902743143e-05,
+      "loss": 0.3255,
+      "step": 22770
+    },
+    {
+      "epoch": 5.6807980049875315,
+      "grad_norm": 6.948160648345947,
+      "learning_rate": 1.432194513715711e-05,
+      "loss": 0.3598,
+      "step": 22780
+    },
+    {
+      "epoch": 5.683291770573566,
+      "grad_norm": 5.765408515930176,
+      "learning_rate": 1.4319451371571074e-05,
+      "loss": 0.4121,
+      "step": 22790
+    },
+    {
+      "epoch": 5.685785536159601,
+      "grad_norm": 11.699581146240234,
+      "learning_rate": 1.4316957605985037e-05,
+      "loss": 0.4094,
+      "step": 22800
+    },
+    {
+      "epoch": 5.688279301745636,
+      "grad_norm": 5.797543048858643,
+      "learning_rate": 1.4314463840399004e-05,
+      "loss": 0.3325,
+      "step": 22810
+    },
+    {
+      "epoch": 5.690773067331671,
+      "grad_norm": 5.791079044342041,
+      "learning_rate": 1.4311970074812968e-05,
+      "loss": 0.3589,
+      "step": 22820
+    },
+    {
+      "epoch": 5.693266832917706,
+      "grad_norm": 5.654359817504883,
+      "learning_rate": 1.4309476309226933e-05,
+      "loss": 0.362,
+      "step": 22830
+    },
+    {
+      "epoch": 5.695760598503741,
+      "grad_norm": 5.24976110458374,
+      "learning_rate": 1.43069825436409e-05,
+      "loss": 0.3756,
+      "step": 22840
+    },
+    {
+      "epoch": 5.698254364089776,
+      "grad_norm": 6.083651542663574,
+      "learning_rate": 1.4304488778054864e-05,
+      "loss": 0.3565,
+      "step": 22850
+    },
+    {
+      "epoch": 5.7007481296758105,
+      "grad_norm": 6.845718860626221,
+      "learning_rate": 1.430199501246883e-05,
+      "loss": 0.3706,
+      "step": 22860
+    },
+    {
+      "epoch": 5.703241895261845,
+      "grad_norm": 4.934425354003906,
+      "learning_rate": 1.4299501246882794e-05,
+      "loss": 0.3646,
+      "step": 22870
+    },
+    {
+      "epoch": 5.70573566084788,
+      "grad_norm": 9.184797286987305,
+      "learning_rate": 1.4297007481296758e-05,
+      "loss": 0.4083,
+      "step": 22880
+    },
+    {
+      "epoch": 5.708229426433915,
+      "grad_norm": 9.234002113342285,
+      "learning_rate": 1.4294513715710725e-05,
+      "loss": 0.3188,
+      "step": 22890
+    },
+    {
+      "epoch": 5.71072319201995,
+      "grad_norm": 7.473569393157959,
+      "learning_rate": 1.4292019950124689e-05,
+      "loss": 0.4069,
+      "step": 22900
+    },
+    {
+      "epoch": 5.713216957605985,
+      "grad_norm": 5.984492301940918,
+      "learning_rate": 1.4289526184538654e-05,
+      "loss": 0.3863,
+      "step": 22910
+    },
+    {
+      "epoch": 5.71571072319202,
+      "grad_norm": 8.733806610107422,
+      "learning_rate": 1.428703241895262e-05,
+      "loss": 0.4223,
+      "step": 22920
+    },
+    {
+      "epoch": 5.7182044887780545,
+      "grad_norm": 7.522699356079102,
+      "learning_rate": 1.4284538653366585e-05,
+      "loss": 0.402,
+      "step": 22930
+    },
+    {
+      "epoch": 5.720698254364089,
+      "grad_norm": 6.460349082946777,
+      "learning_rate": 1.428204488778055e-05,
+      "loss": 0.3046,
+      "step": 22940
+    },
+    {
+      "epoch": 5.723192019950124,
+      "grad_norm": 6.2640204429626465,
+      "learning_rate": 1.4279551122194515e-05,
+      "loss": 0.3377,
+      "step": 22950
+    },
+    {
+      "epoch": 5.725685785536159,
+      "grad_norm": 3.072329521179199,
+      "learning_rate": 1.4277057356608479e-05,
+      "loss": 0.3212,
+      "step": 22960
+    },
+    {
+      "epoch": 5.728179551122195,
+      "grad_norm": 5.633548259735107,
+      "learning_rate": 1.4274563591022446e-05,
+      "loss": 0.3203,
+      "step": 22970
+    },
+    {
+      "epoch": 5.73067331670823,
+      "grad_norm": 8.383296012878418,
+      "learning_rate": 1.427206982543641e-05,
+      "loss": 0.4422,
+      "step": 22980
+    },
+    {
+      "epoch": 5.733167082294265,
+      "grad_norm": 7.479531764984131,
+      "learning_rate": 1.4269576059850376e-05,
+      "loss": 0.3443,
+      "step": 22990
+    },
+    {
+      "epoch": 5.7356608478802995,
+      "grad_norm": 8.796760559082031,
+      "learning_rate": 1.426708229426434e-05,
+      "loss": 0.4097,
+      "step": 23000
+    },
+    {
+      "epoch": 5.738154613466334,
+      "grad_norm": 8.372130393981934,
+      "learning_rate": 1.4264588528678305e-05,
+      "loss": 0.4402,
+      "step": 23010
+    },
+    {
+      "epoch": 5.740648379052369,
+      "grad_norm": 9.008794784545898,
+      "learning_rate": 1.426209476309227e-05,
+      "loss": 0.4109,
+      "step": 23020
+    },
+    {
+      "epoch": 5.743142144638404,
+      "grad_norm": 8.82776927947998,
+      "learning_rate": 1.4259600997506236e-05,
+      "loss": 0.3458,
+      "step": 23030
+    },
+    {
+      "epoch": 5.745635910224439,
+      "grad_norm": 6.604305267333984,
+      "learning_rate": 1.42571072319202e-05,
+      "loss": 0.3969,
+      "step": 23040
+    },
+    {
+      "epoch": 5.748129675810474,
+      "grad_norm": 6.345452308654785,
+      "learning_rate": 1.4254613466334167e-05,
+      "loss": 0.3854,
+      "step": 23050
+    },
+    {
+      "epoch": 5.750623441396509,
+      "grad_norm": 8.280367851257324,
+      "learning_rate": 1.425211970074813e-05,
+      "loss": 0.395,
+      "step": 23060
+    },
+    {
+      "epoch": 5.753117206982544,
+      "grad_norm": 4.906089782714844,
+      "learning_rate": 1.4249625935162097e-05,
+      "loss": 0.3929,
+      "step": 23070
+    },
+    {
+      "epoch": 5.7556109725685785,
+      "grad_norm": 8.1076021194458,
+      "learning_rate": 1.424713216957606e-05,
+      "loss": 0.3963,
+      "step": 23080
+    },
+    {
+      "epoch": 5.758104738154613,
+      "grad_norm": 5.047501087188721,
+      "learning_rate": 1.4244638403990026e-05,
+      "loss": 0.354,
+      "step": 23090
+    },
+    {
+      "epoch": 5.760598503740648,
+      "grad_norm": 5.581216812133789,
+      "learning_rate": 1.4242144638403991e-05,
+      "loss": 0.3966,
+      "step": 23100
+    },
+    {
+      "epoch": 5.763092269326683,
+      "grad_norm": 5.1039347648620605,
+      "learning_rate": 1.4239650872817957e-05,
+      "loss": 0.3249,
+      "step": 23110
+    },
+    {
+      "epoch": 5.765586034912718,
+      "grad_norm": 7.597653388977051,
+      "learning_rate": 1.423715710723192e-05,
+      "loss": 0.369,
+      "step": 23120
+    },
+    {
+      "epoch": 5.768079800498753,
+      "grad_norm": 4.918848037719727,
+      "learning_rate": 1.4234663341645887e-05,
+      "loss": 0.394,
+      "step": 23130
+    },
+    {
+      "epoch": 5.770573566084788,
+      "grad_norm": 7.739204406738281,
+      "learning_rate": 1.4232169576059851e-05,
+      "loss": 0.3303,
+      "step": 23140
+    },
+    {
+      "epoch": 5.773067331670823,
+      "grad_norm": 6.075505256652832,
+      "learning_rate": 1.4229675810473818e-05,
+      "loss": 0.3693,
+      "step": 23150
+    },
+    {
+      "epoch": 5.775561097256858,
+      "grad_norm": 10.600066184997559,
+      "learning_rate": 1.4227182044887782e-05,
+      "loss": 0.3751,
+      "step": 23160
+    },
+    {
+      "epoch": 5.778054862842893,
+      "grad_norm": 7.901242733001709,
+      "learning_rate": 1.4224688279301745e-05,
+      "loss": 0.3912,
+      "step": 23170
+    },
+    {
+      "epoch": 5.780548628428928,
+      "grad_norm": 7.340981483459473,
+      "learning_rate": 1.4222194513715712e-05,
+      "loss": 0.437,
+      "step": 23180
+    },
+    {
+      "epoch": 5.783042394014963,
+      "grad_norm": 6.0289626121521,
+      "learning_rate": 1.4219700748129678e-05,
+      "loss": 0.4226,
+      "step": 23190
+    },
+    {
+      "epoch": 5.785536159600998,
+      "grad_norm": 6.679872989654541,
+      "learning_rate": 1.4217206982543641e-05,
+      "loss": 0.4471,
+      "step": 23200
+    },
+    {
+      "epoch": 5.788029925187033,
+      "grad_norm": 7.326256275177002,
+      "learning_rate": 1.4214713216957608e-05,
+      "loss": 0.39,
+      "step": 23210
+    },
+    {
+      "epoch": 5.7905236907730675,
+      "grad_norm": 5.042798042297363,
+      "learning_rate": 1.4212219451371572e-05,
+      "loss": 0.3997,
+      "step": 23220
+    },
+    {
+      "epoch": 5.793017456359102,
+      "grad_norm": 6.334157943725586,
+      "learning_rate": 1.4209725685785539e-05,
+      "loss": 0.3387,
+      "step": 23230
+    },
+    {
+      "epoch": 5.795511221945137,
+      "grad_norm": 4.981834888458252,
+      "learning_rate": 1.4207231920199502e-05,
+      "loss": 0.5182,
+      "step": 23240
+    },
+    {
+      "epoch": 5.798004987531172,
+      "grad_norm": 4.364705562591553,
+      "learning_rate": 1.4204738154613466e-05,
+      "loss": 0.3756,
+      "step": 23250
+    },
+    {
+      "epoch": 5.800498753117207,
+      "grad_norm": 5.2104268074035645,
+      "learning_rate": 1.4202244389027433e-05,
+      "loss": 0.4008,
+      "step": 23260
+    },
+    {
+      "epoch": 5.802992518703242,
+      "grad_norm": 6.920286178588867,
+      "learning_rate": 1.4199750623441397e-05,
+      "loss": 0.4209,
+      "step": 23270
+    },
+    {
+      "epoch": 5.805486284289277,
+      "grad_norm": 7.996755599975586,
+      "learning_rate": 1.4197256857855364e-05,
+      "loss": 0.4087,
+      "step": 23280
+    },
+    {
+      "epoch": 5.807980049875312,
+      "grad_norm": 5.335063934326172,
+      "learning_rate": 1.4194763092269327e-05,
+      "loss": 0.3644,
+      "step": 23290
+    },
+    {
+      "epoch": 5.8104738154613464,
+      "grad_norm": 7.114324569702148,
+      "learning_rate": 1.4192269326683293e-05,
+      "loss": 0.3954,
+      "step": 23300
+    },
+    {
+      "epoch": 5.812967581047381,
+      "grad_norm": 4.345088005065918,
+      "learning_rate": 1.418977556109726e-05,
+      "loss": 0.4041,
+      "step": 23310
+    },
+    {
+      "epoch": 5.815461346633416,
+      "grad_norm": 6.6690287590026855,
+      "learning_rate": 1.4187281795511223e-05,
+      "loss": 0.3639,
+      "step": 23320
+    },
+    {
+      "epoch": 5.817955112219451,
+      "grad_norm": 7.393259048461914,
+      "learning_rate": 1.4184788029925187e-05,
+      "loss": 0.449,
+      "step": 23330
+    },
+    {
+      "epoch": 5.820448877805486,
+      "grad_norm": 6.419971466064453,
+      "learning_rate": 1.4182294264339154e-05,
+      "loss": 0.4075,
+      "step": 23340
+    },
+    {
+      "epoch": 5.822942643391521,
+      "grad_norm": 5.764027118682861,
+      "learning_rate": 1.4179800498753117e-05,
+      "loss": 0.3845,
+      "step": 23350
+    },
+    {
+      "epoch": 5.825436408977556,
+      "grad_norm": 7.242391109466553,
+      "learning_rate": 1.4177306733167084e-05,
+      "loss": 0.3426,
+      "step": 23360
+    },
+    {
+      "epoch": 5.8279301745635905,
+      "grad_norm": 7.269883155822754,
+      "learning_rate": 1.4174812967581048e-05,
+      "loss": 0.3911,
+      "step": 23370
+    },
+    {
+      "epoch": 5.830423940149626,
+      "grad_norm": 4.136662006378174,
+      "learning_rate": 1.4172319201995013e-05,
+      "loss": 0.5394,
+      "step": 23380
+    },
+    {
+      "epoch": 5.832917705735661,
+      "grad_norm": 7.696037769317627,
+      "learning_rate": 1.4169825436408979e-05,
+      "loss": 0.3823,
+      "step": 23390
+    },
+    {
+      "epoch": 5.835411471321696,
+      "grad_norm": 8.662434577941895,
+      "learning_rate": 1.4167331670822944e-05,
+      "loss": 0.448,
+      "step": 23400
+    },
+    {
+      "epoch": 5.837905236907731,
+      "grad_norm": 7.484658718109131,
+      "learning_rate": 1.4164837905236908e-05,
+      "loss": 0.4121,
+      "step": 23410
+    },
+    {
+      "epoch": 5.840399002493766,
+      "grad_norm": 4.754652976989746,
+      "learning_rate": 1.4162344139650875e-05,
+      "loss": 0.3833,
+      "step": 23420
+    },
+    {
+      "epoch": 5.842892768079801,
+      "grad_norm": 7.560851573944092,
+      "learning_rate": 1.4159850374064838e-05,
+      "loss": 0.4049,
+      "step": 23430
+    },
+    {
+      "epoch": 5.8453865336658355,
+      "grad_norm": 7.429394721984863,
+      "learning_rate": 1.4157356608478805e-05,
+      "loss": 0.409,
+      "step": 23440
+    },
+    {
+      "epoch": 5.84788029925187,
+      "grad_norm": 7.763380527496338,
+      "learning_rate": 1.4154862842892769e-05,
+      "loss": 0.3994,
+      "step": 23450
+    },
+    {
+      "epoch": 5.850374064837905,
+      "grad_norm": 6.119085311889648,
+      "learning_rate": 1.4152369077306734e-05,
+      "loss": 0.4111,
+      "step": 23460
+    },
+    {
+      "epoch": 5.85286783042394,
+      "grad_norm": 7.172895431518555,
+      "learning_rate": 1.41498753117207e-05,
+      "loss": 0.4037,
+      "step": 23470
+    },
+    {
+      "epoch": 5.855361596009975,
+      "grad_norm": 4.939626693725586,
+      "learning_rate": 1.4147381546134665e-05,
+      "loss": 0.4928,
+      "step": 23480
+    },
+    {
+      "epoch": 5.85785536159601,
+      "grad_norm": 5.627965450286865,
+      "learning_rate": 1.414488778054863e-05,
+      "loss": 0.4169,
+      "step": 23490
+    },
+    {
+      "epoch": 5.860349127182045,
+      "grad_norm": 7.699008464813232,
+      "learning_rate": 1.4142394014962595e-05,
+      "loss": 0.3669,
+      "step": 23500
+    },
+    {
+      "epoch": 5.86284289276808,
+      "grad_norm": 8.648443222045898,
+      "learning_rate": 1.4139900249376559e-05,
+      "loss": 0.3886,
+      "step": 23510
+    },
+    {
+      "epoch": 5.865336658354114,
+      "grad_norm": 6.288234710693359,
+      "learning_rate": 1.4137406483790526e-05,
+      "loss": 0.4166,
+      "step": 23520
+    },
+    {
+      "epoch": 5.867830423940149,
+      "grad_norm": 7.797109603881836,
+      "learning_rate": 1.413491271820449e-05,
+      "loss": 0.3971,
+      "step": 23530
+    },
+    {
+      "epoch": 5.870324189526184,
+      "grad_norm": 5.520691394805908,
+      "learning_rate": 1.4132418952618455e-05,
+      "loss": 0.4536,
+      "step": 23540
+    },
+    {
+      "epoch": 5.87281795511222,
+      "grad_norm": 4.677943229675293,
+      "learning_rate": 1.412992518703242e-05,
+      "loss": 0.3958,
+      "step": 23550
+    },
+    {
+      "epoch": 5.875311720698255,
+      "grad_norm": 5.682249069213867,
+      "learning_rate": 1.4127431421446386e-05,
+      "loss": 0.3522,
+      "step": 23560
+    },
+    {
+      "epoch": 5.87780548628429,
+      "grad_norm": 6.83863639831543,
+      "learning_rate": 1.412493765586035e-05,
+      "loss": 0.3656,
+      "step": 23570
+    },
+    {
+      "epoch": 5.8802992518703245,
+      "grad_norm": 6.313279628753662,
+      "learning_rate": 1.4122443890274316e-05,
+      "loss": 0.3988,
+      "step": 23580
+    },
+    {
+      "epoch": 5.882793017456359,
+      "grad_norm": 7.157898426055908,
+      "learning_rate": 1.411995012468828e-05,
+      "loss": 0.401,
+      "step": 23590
+    },
+    {
+      "epoch": 5.885286783042394,
+      "grad_norm": 6.3576531410217285,
+      "learning_rate": 1.4117456359102247e-05,
+      "loss": 0.4291,
+      "step": 23600
+    },
+    {
+      "epoch": 5.887780548628429,
+      "grad_norm": 6.434720993041992,
+      "learning_rate": 1.411496259351621e-05,
+      "loss": 0.3876,
+      "step": 23610
+    },
+    {
+      "epoch": 5.890274314214464,
+      "grad_norm": 6.252552509307861,
+      "learning_rate": 1.4112468827930174e-05,
+      "loss": 0.3471,
+      "step": 23620
+    },
+    {
+      "epoch": 5.892768079800499,
+      "grad_norm": 3.91813588142395,
+      "learning_rate": 1.4109975062344141e-05,
+      "loss": 0.3224,
+      "step": 23630
+    },
+    {
+      "epoch": 5.895261845386534,
+      "grad_norm": 6.144435882568359,
+      "learning_rate": 1.4107481296758105e-05,
+      "loss": 0.4012,
+      "step": 23640
+    },
+    {
+      "epoch": 5.897755610972569,
+      "grad_norm": 6.0497565269470215,
+      "learning_rate": 1.4104987531172072e-05,
+      "loss": 0.4003,
+      "step": 23650
+    },
+    {
+      "epoch": 5.9002493765586035,
+      "grad_norm": 4.732110500335693,
+      "learning_rate": 1.4102493765586037e-05,
+      "loss": 0.419,
+      "step": 23660
+    },
+    {
+      "epoch": 5.902743142144638,
+      "grad_norm": 13.11367416381836,
+      "learning_rate": 1.41e-05,
+      "loss": 0.3991,
+      "step": 23670
+    },
+    {
+      "epoch": 5.905236907730673,
+      "grad_norm": 5.816923141479492,
+      "learning_rate": 1.4097506234413968e-05,
+      "loss": 0.34,
+      "step": 23680
+    },
+    {
+      "epoch": 5.907730673316708,
+      "grad_norm": 7.775258541107178,
+      "learning_rate": 1.4095012468827931e-05,
+      "loss": 0.3784,
+      "step": 23690
+    },
+    {
+      "epoch": 5.910224438902743,
+      "grad_norm": 5.4291486740112305,
+      "learning_rate": 1.4092518703241898e-05,
+      "loss": 0.404,
+      "step": 23700
+    },
+    {
+      "epoch": 5.912718204488778,
+      "grad_norm": 7.315398216247559,
+      "learning_rate": 1.4090024937655862e-05,
+      "loss": 0.4278,
+      "step": 23710
+    },
+    {
+      "epoch": 5.915211970074813,
+      "grad_norm": 5.724961757659912,
+      "learning_rate": 1.4087531172069825e-05,
+      "loss": 0.4336,
+      "step": 23720
+    },
+    {
+      "epoch": 5.917705735660848,
+      "grad_norm": 8.703821182250977,
+      "learning_rate": 1.4085037406483792e-05,
+      "loss": 0.4091,
+      "step": 23730
+    },
+    {
+      "epoch": 5.920199501246882,
+      "grad_norm": 6.634580612182617,
+      "learning_rate": 1.4082543640897756e-05,
+      "loss": 0.3738,
+      "step": 23740
+    },
+    {
+      "epoch": 5.922693266832917,
+      "grad_norm": 7.790684223175049,
+      "learning_rate": 1.4080049875311721e-05,
+      "loss": 0.4328,
+      "step": 23750
+    },
+    {
+      "epoch": 5.925187032418952,
+      "grad_norm": 6.3984808921813965,
+      "learning_rate": 1.4077556109725687e-05,
+      "loss": 0.3641,
+      "step": 23760
+    },
+    {
+      "epoch": 5.927680798004987,
+      "grad_norm": 5.913004398345947,
+      "learning_rate": 1.4075062344139652e-05,
+      "loss": 0.3595,
+      "step": 23770
+    },
+    {
+      "epoch": 5.930174563591023,
+      "grad_norm": 9.154977798461914,
+      "learning_rate": 1.4072568578553619e-05,
+      "loss": 0.3679,
+      "step": 23780
+    },
+    {
+      "epoch": 5.932668329177058,
+      "grad_norm": 6.98581075668335,
+      "learning_rate": 1.4070074812967583e-05,
+      "loss": 0.3907,
+      "step": 23790
+    },
+    {
+      "epoch": 5.9351620947630925,
+      "grad_norm": 6.550581932067871,
+      "learning_rate": 1.4067581047381546e-05,
+      "loss": 0.4143,
+      "step": 23800
+    },
+    {
+      "epoch": 5.937655860349127,
+      "grad_norm": 7.637538909912109,
+      "learning_rate": 1.4065087281795513e-05,
+      "loss": 0.3854,
+      "step": 23810
+    },
+    {
+      "epoch": 5.940149625935162,
+      "grad_norm": 8.336382865905762,
+      "learning_rate": 1.4062593516209477e-05,
+      "loss": 0.3903,
+      "step": 23820
+    },
+    {
+      "epoch": 5.942643391521197,
+      "grad_norm": 5.107188701629639,
+      "learning_rate": 1.4060099750623442e-05,
+      "loss": 0.3879,
+      "step": 23830
+    },
+    {
+      "epoch": 5.945137157107232,
+      "grad_norm": 5.9631571769714355,
+      "learning_rate": 1.4057605985037407e-05,
+      "loss": 0.4691,
+      "step": 23840
+    },
+    {
+      "epoch": 5.947630922693267,
+      "grad_norm": 6.080106258392334,
+      "learning_rate": 1.4055112219451373e-05,
+      "loss": 0.4799,
+      "step": 23850
+    },
+    {
+      "epoch": 5.950124688279302,
+      "grad_norm": 9.913834571838379,
+      "learning_rate": 1.4052618453865338e-05,
+      "loss": 0.3768,
+      "step": 23860
+    },
+    {
+      "epoch": 5.952618453865337,
+      "grad_norm": 6.052908420562744,
+      "learning_rate": 1.4050124688279303e-05,
+      "loss": 0.3323,
+      "step": 23870
+    },
+    {
+      "epoch": 5.9551122194513715,
+      "grad_norm": 8.074835777282715,
+      "learning_rate": 1.4047630922693267e-05,
+      "loss": 0.4003,
+      "step": 23880
+    },
+    {
+      "epoch": 5.957605985037406,
+      "grad_norm": 4.824693202972412,
+      "learning_rate": 1.4045137157107234e-05,
+      "loss": 0.3868,
+      "step": 23890
+    },
+    {
+      "epoch": 5.960099750623441,
+      "grad_norm": 8.275466918945312,
+      "learning_rate": 1.4042643391521198e-05,
+      "loss": 0.3944,
+      "step": 23900
+    },
+    {
+      "epoch": 5.962593516209476,
+      "grad_norm": 7.059325695037842,
+      "learning_rate": 1.4040149625935163e-05,
+      "loss": 0.4608,
+      "step": 23910
+    },
+    {
+      "epoch": 5.965087281795511,
+      "grad_norm": 6.65533447265625,
+      "learning_rate": 1.4037655860349128e-05,
+      "loss": 0.3445,
+      "step": 23920
+    },
+    {
+      "epoch": 5.967581047381546,
+      "grad_norm": 6.2728590965271,
+      "learning_rate": 1.4035162094763093e-05,
+      "loss": 0.3764,
+      "step": 23930
+    },
+    {
+      "epoch": 5.970074812967581,
+      "grad_norm": 8.21418571472168,
+      "learning_rate": 1.4032668329177059e-05,
+      "loss": 0.3982,
+      "step": 23940
+    },
+    {
+      "epoch": 5.9725685785536164,
+      "grad_norm": 6.642632007598877,
+      "learning_rate": 1.4030423940149627e-05,
+      "loss": 0.3619,
+      "step": 23950
+    },
+    {
+      "epoch": 5.975062344139651,
+      "grad_norm": 7.222472667694092,
+      "learning_rate": 1.4027930174563592e-05,
+      "loss": 0.3871,
+      "step": 23960
+    },
+    {
+      "epoch": 5.977556109725686,
+      "grad_norm": 8.01220417022705,
+      "learning_rate": 1.4025436408977558e-05,
+      "loss": 0.3979,
+      "step": 23970
+    },
+    {
+      "epoch": 5.980049875311721,
+      "grad_norm": 5.2847981452941895,
+      "learning_rate": 1.4022942643391521e-05,
+      "loss": 0.4304,
+      "step": 23980
+    },
+    {
+      "epoch": 5.982543640897756,
+      "grad_norm": 10.303918838500977,
+      "learning_rate": 1.4020448877805488e-05,
+      "loss": 0.377,
+      "step": 23990
+    },
+    {
+      "epoch": 5.985037406483791,
+      "grad_norm": 10.450936317443848,
+      "learning_rate": 1.4017955112219452e-05,
+      "loss": 0.4502,
+      "step": 24000
+    },
+    {
+      "epoch": 5.987531172069826,
+      "grad_norm": 6.296121120452881,
+      "learning_rate": 1.4015461346633417e-05,
+      "loss": 0.4307,
+      "step": 24010
+    },
+    {
+      "epoch": 5.9900249376558605,
+      "grad_norm": 5.6589202880859375,
+      "learning_rate": 1.4012967581047383e-05,
+      "loss": 0.4116,
+      "step": 24020
+    },
+    {
+      "epoch": 5.992518703241895,
+      "grad_norm": 6.098679065704346,
+      "learning_rate": 1.4010473815461348e-05,
+      "loss": 0.336,
+      "step": 24030
+    },
+    {
+      "epoch": 5.99501246882793,
+      "grad_norm": 6.028469562530518,
+      "learning_rate": 1.4007980049875313e-05,
+      "loss": 0.3988,
+      "step": 24040
+    },
+    {
+      "epoch": 5.997506234413965,
+      "grad_norm": 7.266406536102295,
+      "learning_rate": 1.4005486284289278e-05,
+      "loss": 0.435,
+      "step": 24050
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 4.367632865905762,
+      "learning_rate": 1.4002992518703242e-05,
+      "loss": 0.3838,
+      "step": 24060
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 0.41945403814315796,
+      "eval_runtime": 59.8425,
+      "eval_samples_per_second": 16.761,
+      "eval_steps_per_second": 16.761,
+      "step": 24060
+    },
+    {
+      "epoch": 6.002493765586035,
+      "grad_norm": 5.5019001960754395,
+      "learning_rate": 1.4000498753117209e-05,
+      "loss": 0.3192,
+      "step": 24070
+    },
+    {
+      "epoch": 6.00498753117207,
+      "grad_norm": 6.230374813079834,
+      "learning_rate": 1.3998004987531173e-05,
+      "loss": 0.3463,
+      "step": 24080
+    },
+    {
+      "epoch": 6.007481296758105,
+      "grad_norm": 5.844031810760498,
+      "learning_rate": 1.3995511221945138e-05,
+      "loss": 0.4043,
+      "step": 24090
+    },
+    {
+      "epoch": 6.0099750623441395,
+      "grad_norm": 6.5853166580200195,
+      "learning_rate": 1.3993017456359103e-05,
+      "loss": 0.4222,
+      "step": 24100
+    },
+    {
+      "epoch": 6.012468827930174,
+      "grad_norm": 7.990192890167236,
+      "learning_rate": 1.3990523690773069e-05,
+      "loss": 0.3578,
+      "step": 24110
+    },
+    {
+      "epoch": 6.014962593516209,
+      "grad_norm": 5.267573356628418,
+      "learning_rate": 1.3988029925187034e-05,
+      "loss": 0.3111,
+      "step": 24120
+    },
+    {
+      "epoch": 6.017456359102244,
+      "grad_norm": 6.116258144378662,
+      "learning_rate": 1.3985536159601e-05,
+      "loss": 0.3556,
+      "step": 24130
+    },
+    {
+      "epoch": 6.019950124688279,
+      "grad_norm": 7.899449825286865,
+      "learning_rate": 1.3983042394014963e-05,
+      "loss": 0.3691,
+      "step": 24140
+    },
+    {
+      "epoch": 6.022443890274314,
+      "grad_norm": 8.339388847351074,
+      "learning_rate": 1.398054862842893e-05,
+      "loss": 0.3704,
+      "step": 24150
+    },
+    {
+      "epoch": 6.024937655860349,
+      "grad_norm": 5.403593063354492,
+      "learning_rate": 1.3978054862842893e-05,
+      "loss": 0.3668,
+      "step": 24160
+    },
+    {
+      "epoch": 6.027431421446384,
+      "grad_norm": 7.102303981781006,
+      "learning_rate": 1.397556109725686e-05,
+      "loss": 0.3266,
+      "step": 24170
+    },
+    {
+      "epoch": 6.029925187032419,
+      "grad_norm": 5.142783164978027,
+      "learning_rate": 1.3973067331670824e-05,
+      "loss": 0.3544,
+      "step": 24180
+    },
+    {
+      "epoch": 6.032418952618454,
+      "grad_norm": 7.026705265045166,
+      "learning_rate": 1.3970573566084788e-05,
+      "loss": 0.3973,
+      "step": 24190
+    },
+    {
+      "epoch": 6.034912718204489,
+      "grad_norm": 5.733627796173096,
+      "learning_rate": 1.3968079800498755e-05,
+      "loss": 0.3553,
+      "step": 24200
+    },
+    {
+      "epoch": 6.037406483790524,
+      "grad_norm": 6.362379550933838,
+      "learning_rate": 1.396558603491272e-05,
+      "loss": 0.3822,
+      "step": 24210
+    },
+    {
+      "epoch": 6.039900249376559,
+      "grad_norm": 7.283205509185791,
+      "learning_rate": 1.3963092269326684e-05,
+      "loss": 0.4559,
+      "step": 24220
+    },
+    {
+      "epoch": 6.042394014962594,
+      "grad_norm": 7.929378509521484,
+      "learning_rate": 1.396059850374065e-05,
+      "loss": 0.3449,
+      "step": 24230
+    },
+    {
+      "epoch": 6.0448877805486285,
+      "grad_norm": 7.110724449157715,
+      "learning_rate": 1.3958104738154614e-05,
+      "loss": 0.3746,
+      "step": 24240
+    },
+    {
+      "epoch": 6.047381546134663,
+      "grad_norm": 5.7137227058410645,
+      "learning_rate": 1.3955610972568581e-05,
+      "loss": 0.4771,
+      "step": 24250
+    },
+    {
+      "epoch": 6.049875311720698,
+      "grad_norm": 6.553785800933838,
+      "learning_rate": 1.3953117206982545e-05,
+      "loss": 0.3555,
+      "step": 24260
+    },
+    {
+      "epoch": 6.052369077306733,
+      "grad_norm": 5.964046955108643,
+      "learning_rate": 1.3950623441396509e-05,
+      "loss": 0.3619,
+      "step": 24270
+    },
+    {
+      "epoch": 6.054862842892768,
+      "grad_norm": 7.368553638458252,
+      "learning_rate": 1.3948129675810476e-05,
+      "loss": 0.3385,
+      "step": 24280
+    },
+    {
+      "epoch": 6.057356608478803,
+      "grad_norm": 9.049263000488281,
+      "learning_rate": 1.3945635910224439e-05,
+      "loss": 0.3605,
+      "step": 24290
+    },
+    {
+      "epoch": 6.059850374064838,
+      "grad_norm": 7.191588401794434,
+      "learning_rate": 1.3943142144638404e-05,
+      "loss": 0.4145,
+      "step": 24300
+    },
+    {
+      "epoch": 6.062344139650873,
+      "grad_norm": 10.2169828414917,
+      "learning_rate": 1.394064837905237e-05,
+      "loss": 0.3857,
+      "step": 24310
+    },
+    {
+      "epoch": 6.0648379052369075,
+      "grad_norm": 7.191344738006592,
+      "learning_rate": 1.3938154613466335e-05,
+      "loss": 0.3727,
+      "step": 24320
+    },
+    {
+      "epoch": 6.067331670822942,
+      "grad_norm": 6.484828472137451,
+      "learning_rate": 1.3935660847880302e-05,
+      "loss": 0.4169,
+      "step": 24330
+    },
+    {
+      "epoch": 6.069825436408977,
+      "grad_norm": 5.972647190093994,
+      "learning_rate": 1.3933167082294266e-05,
+      "loss": 0.2752,
+      "step": 24340
+    },
+    {
+      "epoch": 6.072319201995012,
+      "grad_norm": 7.331430912017822,
+      "learning_rate": 1.393067331670823e-05,
+      "loss": 0.3688,
+      "step": 24350
+    },
+    {
+      "epoch": 6.074812967581048,
+      "grad_norm": 4.472461223602295,
+      "learning_rate": 1.3928179551122196e-05,
+      "loss": 0.3835,
+      "step": 24360
+    },
+    {
+      "epoch": 6.077306733167083,
+      "grad_norm": 6.023663520812988,
+      "learning_rate": 1.392568578553616e-05,
+      "loss": 0.3425,
+      "step": 24370
+    },
+    {
+      "epoch": 6.079800498753118,
+      "grad_norm": 6.141242980957031,
+      "learning_rate": 1.3923192019950127e-05,
+      "loss": 0.3483,
+      "step": 24380
+    },
+    {
+      "epoch": 6.082294264339152,
+      "grad_norm": 6.454586029052734,
+      "learning_rate": 1.392069825436409e-05,
+      "loss": 0.3239,
+      "step": 24390
+    },
+    {
+      "epoch": 6.084788029925187,
+      "grad_norm": 5.868727684020996,
+      "learning_rate": 1.3918204488778056e-05,
+      "loss": 0.2889,
+      "step": 24400
+    },
+    {
+      "epoch": 6.087281795511222,
+      "grad_norm": 5.985288619995117,
+      "learning_rate": 1.3915710723192021e-05,
+      "loss": 0.3998,
+      "step": 24410
+    },
+    {
+      "epoch": 6.089775561097257,
+      "grad_norm": 8.732137680053711,
+      "learning_rate": 1.3913216957605986e-05,
+      "loss": 0.4191,
+      "step": 24420
+    },
+    {
+      "epoch": 6.092269326683292,
+      "grad_norm": 3.857243061065674,
+      "learning_rate": 1.391072319201995e-05,
+      "loss": 0.3174,
+      "step": 24430
+    },
+    {
+      "epoch": 6.094763092269327,
+      "grad_norm": 6.74662446975708,
+      "learning_rate": 1.3908229426433917e-05,
+      "loss": 0.4585,
+      "step": 24440
+    },
+    {
+      "epoch": 6.097256857855362,
+      "grad_norm": 8.382293701171875,
+      "learning_rate": 1.390573566084788e-05,
+      "loss": 0.4119,
+      "step": 24450
+    },
+    {
+      "epoch": 6.0997506234413965,
+      "grad_norm": 5.25961446762085,
+      "learning_rate": 1.3903241895261848e-05,
+      "loss": 0.355,
+      "step": 24460
+    },
+    {
+      "epoch": 6.102244389027431,
+      "grad_norm": 10.296069145202637,
+      "learning_rate": 1.3900748129675811e-05,
+      "loss": 0.4015,
+      "step": 24470
+    },
+    {
+      "epoch": 6.104738154613466,
+      "grad_norm": 6.3385009765625,
+      "learning_rate": 1.3898254364089777e-05,
+      "loss": 0.4538,
+      "step": 24480
+    },
+    {
+      "epoch": 6.107231920199501,
+      "grad_norm": 5.728212356567383,
+      "learning_rate": 1.3895760598503742e-05,
+      "loss": 0.3313,
+      "step": 24490
+    },
+    {
+      "epoch": 6.109725685785536,
+      "grad_norm": 7.13115930557251,
+      "learning_rate": 1.3893266832917707e-05,
+      "loss": 0.3547,
+      "step": 24500
+    },
+    {
+      "epoch": 6.112219451371571,
+      "grad_norm": 7.432462692260742,
+      "learning_rate": 1.3890773067331671e-05,
+      "loss": 0.3595,
+      "step": 24510
+    },
+    {
+      "epoch": 6.114713216957606,
+      "grad_norm": 7.006857872009277,
+      "learning_rate": 1.3888279301745638e-05,
+      "loss": 0.3668,
+      "step": 24520
+    },
+    {
+      "epoch": 6.117206982543641,
+      "grad_norm": 6.406911373138428,
+      "learning_rate": 1.3885785536159601e-05,
+      "loss": 0.3775,
+      "step": 24530
+    },
+    {
+      "epoch": 6.1197007481296755,
+      "grad_norm": 4.907394886016846,
+      "learning_rate": 1.3883291770573568e-05,
+      "loss": 0.3529,
+      "step": 24540
+    },
+    {
+      "epoch": 6.12219451371571,
+      "grad_norm": 5.443017482757568,
+      "learning_rate": 1.3880798004987532e-05,
+      "loss": 0.3293,
+      "step": 24550
+    },
+    {
+      "epoch": 6.124688279301745,
+      "grad_norm": 6.361752510070801,
+      "learning_rate": 1.3878304239401497e-05,
+      "loss": 0.3988,
+      "step": 24560
+    },
+    {
+      "epoch": 6.127182044887781,
+      "grad_norm": 7.4896464347839355,
+      "learning_rate": 1.3875810473815463e-05,
+      "loss": 0.3992,
+      "step": 24570
+    },
+    {
+      "epoch": 6.129675810473816,
+      "grad_norm": 5.954341888427734,
+      "learning_rate": 1.3873316708229428e-05,
+      "loss": 0.3524,
+      "step": 24580
+    },
+    {
+      "epoch": 6.132169576059851,
+      "grad_norm": 7.877580642700195,
+      "learning_rate": 1.3870822942643392e-05,
+      "loss": 0.3688,
+      "step": 24590
+    },
+    {
+      "epoch": 6.134663341645886,
+      "grad_norm": 11.095531463623047,
+      "learning_rate": 1.3868329177057359e-05,
+      "loss": 0.3773,
+      "step": 24600
+    },
+    {
+      "epoch": 6.13715710723192,
+      "grad_norm": 4.941054344177246,
+      "learning_rate": 1.3865835411471322e-05,
+      "loss": 0.337,
+      "step": 24610
+    },
+    {
+      "epoch": 6.139650872817955,
+      "grad_norm": 7.861473083496094,
+      "learning_rate": 1.386334164588529e-05,
+      "loss": 0.3896,
+      "step": 24620
+    },
+    {
+      "epoch": 6.14214463840399,
+      "grad_norm": 7.050281524658203,
+      "learning_rate": 1.3860847880299253e-05,
+      "loss": 0.3908,
+      "step": 24630
+    },
+    {
+      "epoch": 6.144638403990025,
+      "grad_norm": 8.201066970825195,
+      "learning_rate": 1.3858354114713217e-05,
+      "loss": 0.4176,
+      "step": 24640
+    },
+    {
+      "epoch": 6.14713216957606,
+      "grad_norm": 6.268393516540527,
+      "learning_rate": 1.3855860349127184e-05,
+      "loss": 0.3486,
+      "step": 24650
+    },
+    {
+      "epoch": 6.149625935162095,
+      "grad_norm": 7.479492664337158,
+      "learning_rate": 1.3853366583541147e-05,
+      "loss": 0.3568,
+      "step": 24660
+    },
+    {
+      "epoch": 6.15211970074813,
+      "grad_norm": 9.700624465942383,
+      "learning_rate": 1.3850872817955114e-05,
+      "loss": 0.5033,
+      "step": 24670
+    },
+    {
+      "epoch": 6.1546134663341645,
+      "grad_norm": 7.117500305175781,
+      "learning_rate": 1.384837905236908e-05,
+      "loss": 0.3423,
+      "step": 24680
+    },
+    {
+      "epoch": 6.157107231920199,
+      "grad_norm": 6.819540023803711,
+      "learning_rate": 1.3845885286783043e-05,
+      "loss": 0.4107,
+      "step": 24690
+    },
+    {
+      "epoch": 6.159600997506234,
+      "grad_norm": 7.1624755859375,
+      "learning_rate": 1.384339152119701e-05,
+      "loss": 0.3808,
+      "step": 24700
+    },
+    {
+      "epoch": 6.162094763092269,
+      "grad_norm": 5.549429416656494,
+      "learning_rate": 1.3840897755610974e-05,
+      "loss": 0.4034,
+      "step": 24710
+    },
+    {
+      "epoch": 6.164588528678304,
+      "grad_norm": 6.838397026062012,
+      "learning_rate": 1.3838403990024937e-05,
+      "loss": 0.4178,
+      "step": 24720
+    },
+    {
+      "epoch": 6.167082294264339,
+      "grad_norm": 6.023339748382568,
+      "learning_rate": 1.3835910224438904e-05,
+      "loss": 0.3682,
+      "step": 24730
+    },
+    {
+      "epoch": 6.169576059850374,
+      "grad_norm": 6.62280797958374,
+      "learning_rate": 1.3833416458852868e-05,
+      "loss": 0.3843,
+      "step": 24740
+    },
+    {
+      "epoch": 6.172069825436409,
+      "grad_norm": 6.568355083465576,
+      "learning_rate": 1.3830922693266835e-05,
+      "loss": 0.4233,
+      "step": 24750
+    },
+    {
+      "epoch": 6.174563591022444,
+      "grad_norm": 9.093072891235352,
+      "learning_rate": 1.3828428927680799e-05,
+      "loss": 0.3631,
+      "step": 24760
+    },
+    {
+      "epoch": 6.177057356608479,
+      "grad_norm": 8.72046184539795,
+      "learning_rate": 1.3825935162094764e-05,
+      "loss": 0.3435,
+      "step": 24770
+    },
+    {
+      "epoch": 6.179551122194514,
+      "grad_norm": 9.039979934692383,
+      "learning_rate": 1.3823441396508729e-05,
+      "loss": 0.3879,
+      "step": 24780
+    },
+    {
+      "epoch": 6.182044887780549,
+      "grad_norm": 8.933988571166992,
+      "learning_rate": 1.3820947630922694e-05,
+      "loss": 0.3646,
+      "step": 24790
+    },
+    {
+      "epoch": 6.184538653366584,
+      "grad_norm": 7.515913963317871,
+      "learning_rate": 1.3818453865336658e-05,
+      "loss": 0.3503,
+      "step": 24800
+    },
+    {
+      "epoch": 6.187032418952619,
+      "grad_norm": 6.617506504058838,
+      "learning_rate": 1.3815960099750625e-05,
+      "loss": 0.4107,
+      "step": 24810
+    },
+    {
+      "epoch": 6.1895261845386536,
+      "grad_norm": 5.950352668762207,
+      "learning_rate": 1.3813466334164589e-05,
+      "loss": 0.3318,
+      "step": 24820
+    },
+    {
+      "epoch": 6.192019950124688,
+      "grad_norm": 6.522132873535156,
+      "learning_rate": 1.3810972568578556e-05,
+      "loss": 0.3876,
+      "step": 24830
+    },
+    {
+      "epoch": 6.194513715710723,
+      "grad_norm": 5.656475067138672,
+      "learning_rate": 1.380847880299252e-05,
+      "loss": 0.3361,
+      "step": 24840
+    },
+    {
+      "epoch": 6.197007481296758,
+      "grad_norm": 9.664381980895996,
+      "learning_rate": 1.3805985037406485e-05,
+      "loss": 0.4696,
+      "step": 24850
+    },
+    {
+      "epoch": 6.199501246882793,
+      "grad_norm": 5.316076755523682,
+      "learning_rate": 1.380349127182045e-05,
+      "loss": 0.3133,
+      "step": 24860
+    },
+    {
+      "epoch": 6.201995012468828,
+      "grad_norm": 5.598194599151611,
+      "learning_rate": 1.3800997506234415e-05,
+      "loss": 0.4303,
+      "step": 24870
+    },
+    {
+      "epoch": 6.204488778054863,
+      "grad_norm": 9.467336654663086,
+      "learning_rate": 1.379850374064838e-05,
+      "loss": 0.3493,
+      "step": 24880
+    },
+    {
+      "epoch": 6.206982543640898,
+      "grad_norm": 5.858520030975342,
+      "learning_rate": 1.3796009975062346e-05,
+      "loss": 0.4155,
+      "step": 24890
+    },
+    {
+      "epoch": 6.2094763092269325,
+      "grad_norm": 6.202467441558838,
+      "learning_rate": 1.379351620947631e-05,
+      "loss": 0.3583,
+      "step": 24900
+    },
+    {
+      "epoch": 6.211970074812967,
+      "grad_norm": 10.094200134277344,
+      "learning_rate": 1.3791022443890276e-05,
+      "loss": 0.3416,
+      "step": 24910
+    },
+    {
+      "epoch": 6.214463840399002,
+      "grad_norm": 6.892770290374756,
+      "learning_rate": 1.378852867830424e-05,
+      "loss": 0.3961,
+      "step": 24920
+    },
+    {
+      "epoch": 6.216957605985037,
+      "grad_norm": 6.055887699127197,
+      "learning_rate": 1.3786034912718205e-05,
+      "loss": 0.3684,
+      "step": 24930
+    },
+    {
+      "epoch": 6.219451371571072,
+      "grad_norm": 7.503059387207031,
+      "learning_rate": 1.378354114713217e-05,
+      "loss": 0.3396,
+      "step": 24940
+    },
+    {
+      "epoch": 6.221945137157107,
+      "grad_norm": 6.7039642333984375,
+      "learning_rate": 1.3781047381546136e-05,
+      "loss": 0.3621,
+      "step": 24950
+    },
+    {
+      "epoch": 6.224438902743142,
+      "grad_norm": 6.32158088684082,
+      "learning_rate": 1.3778553615960101e-05,
+      "loss": 0.3421,
+      "step": 24960
+    },
+    {
+      "epoch": 6.2269326683291775,
+      "grad_norm": 7.217518329620361,
+      "learning_rate": 1.3776059850374067e-05,
+      "loss": 0.3611,
+      "step": 24970
+    },
+    {
+      "epoch": 6.229426433915212,
+      "grad_norm": 6.535889148712158,
+      "learning_rate": 1.377356608478803e-05,
+      "loss": 0.3765,
+      "step": 24980
+    },
+    {
+      "epoch": 6.231920199501247,
+      "grad_norm": 5.55719518661499,
+      "learning_rate": 1.3771072319201997e-05,
+      "loss": 0.3583,
+      "step": 24990
+    },
+    {
+      "epoch": 6.234413965087282,
+      "grad_norm": 8.752567291259766,
+      "learning_rate": 1.3768578553615961e-05,
+      "loss": 0.3583,
+      "step": 25000
+    },
+    {
+      "epoch": 6.236907730673317,
+      "grad_norm": 6.028888702392578,
+      "learning_rate": 1.3766084788029924e-05,
+      "loss": 0.3537,
+      "step": 25010
+    },
+    {
+      "epoch": 6.239401496259352,
+      "grad_norm": 8.308853149414062,
+      "learning_rate": 1.3763591022443891e-05,
+      "loss": 0.4142,
+      "step": 25020
+    },
+    {
+      "epoch": 6.241895261845387,
+      "grad_norm": 7.740601539611816,
+      "learning_rate": 1.3761097256857857e-05,
+      "loss": 0.3631,
+      "step": 25030
+    },
+    {
+      "epoch": 6.2443890274314215,
+      "grad_norm": 6.1072611808776855,
+      "learning_rate": 1.3758603491271822e-05,
+      "loss": 0.3523,
+      "step": 25040
+    },
+    {
+      "epoch": 6.246882793017456,
+      "grad_norm": 5.326240062713623,
+      "learning_rate": 1.3756109725685787e-05,
+      "loss": 0.3606,
+      "step": 25050
+    },
+    {
+      "epoch": 6.249376558603491,
+      "grad_norm": 7.631753921508789,
+      "learning_rate": 1.3753615960099751e-05,
+      "loss": 0.3653,
+      "step": 25060
+    },
+    {
+      "epoch": 6.251870324189526,
+      "grad_norm": 4.055186748504639,
+      "learning_rate": 1.3751122194513718e-05,
+      "loss": 0.3317,
+      "step": 25070
+    },
+    {
+      "epoch": 6.254364089775561,
+      "grad_norm": 6.139715671539307,
+      "learning_rate": 1.3748628428927682e-05,
+      "loss": 0.3835,
+      "step": 25080
+    },
+    {
+      "epoch": 6.256857855361596,
+      "grad_norm": 6.292752265930176,
+      "learning_rate": 1.3746134663341645e-05,
+      "loss": 0.2972,
+      "step": 25090
+    },
+    {
+      "epoch": 6.259351620947631,
+      "grad_norm": 7.624993324279785,
+      "learning_rate": 1.3743640897755612e-05,
+      "loss": 0.4233,
+      "step": 25100
+    },
+    {
+      "epoch": 6.261845386533666,
+      "grad_norm": 5.454896450042725,
+      "learning_rate": 1.3741147132169576e-05,
+      "loss": 0.3935,
+      "step": 25110
+    },
+    {
+      "epoch": 6.2643391521197005,
+      "grad_norm": 7.689428329467773,
+      "learning_rate": 1.3738653366583543e-05,
+      "loss": 0.3478,
+      "step": 25120
+    },
+    {
+      "epoch": 6.266832917705735,
+      "grad_norm": 5.5055365562438965,
+      "learning_rate": 1.3736159600997507e-05,
+      "loss": 0.3336,
+      "step": 25130
+    },
+    {
+      "epoch": 6.26932668329177,
+      "grad_norm": 4.818638324737549,
+      "learning_rate": 1.3733665835411472e-05,
+      "loss": 0.3446,
+      "step": 25140
+    },
+    {
+      "epoch": 6.271820448877805,
+      "grad_norm": 6.679276943206787,
+      "learning_rate": 1.3731172069825439e-05,
+      "loss": 0.3915,
+      "step": 25150
+    },
+    {
+      "epoch": 6.274314214463841,
+      "grad_norm": 7.1155009269714355,
+      "learning_rate": 1.3728678304239402e-05,
+      "loss": 0.3504,
+      "step": 25160
+    },
+    {
+      "epoch": 6.276807980049876,
+      "grad_norm": 7.2714009284973145,
+      "learning_rate": 1.372618453865337e-05,
+      "loss": 0.5088,
+      "step": 25170
+    },
+    {
+      "epoch": 6.279301745635911,
+      "grad_norm": 8.970192909240723,
+      "learning_rate": 1.3723690773067333e-05,
+      "loss": 0.4345,
+      "step": 25180
+    },
+    {
+      "epoch": 6.2817955112219455,
+      "grad_norm": 6.6837334632873535,
+      "learning_rate": 1.3721197007481297e-05,
+      "loss": 0.3889,
+      "step": 25190
+    },
+    {
+      "epoch": 6.28428927680798,
+      "grad_norm": 5.488166809082031,
+      "learning_rate": 1.3718703241895264e-05,
+      "loss": 0.3952,
+      "step": 25200
+    },
+    {
+      "epoch": 6.286783042394015,
+      "grad_norm": 7.671344757080078,
+      "learning_rate": 1.3716209476309227e-05,
+      "loss": 0.3589,
+      "step": 25210
+    },
+    {
+      "epoch": 6.28927680798005,
+      "grad_norm": 6.706274032592773,
+      "learning_rate": 1.3713715710723193e-05,
+      "loss": 0.4448,
+      "step": 25220
+    },
+    {
+      "epoch": 6.291770573566085,
+      "grad_norm": 8.258255004882812,
+      "learning_rate": 1.3711221945137158e-05,
+      "loss": 0.3784,
+      "step": 25230
+    },
+    {
+      "epoch": 6.29426433915212,
+      "grad_norm": 8.191130638122559,
+      "learning_rate": 1.3708728179551123e-05,
+      "loss": 0.4016,
+      "step": 25240
+    },
+    {
+      "epoch": 6.296758104738155,
+      "grad_norm": 6.618471622467041,
+      "learning_rate": 1.3706234413965089e-05,
+      "loss": 0.319,
+      "step": 25250
+    },
+    {
+      "epoch": 6.2992518703241895,
+      "grad_norm": 4.492340564727783,
+      "learning_rate": 1.3703740648379054e-05,
+      "loss": 0.3433,
+      "step": 25260
+    },
+    {
+      "epoch": 6.301745635910224,
+      "grad_norm": 7.213396072387695,
+      "learning_rate": 1.3701246882793017e-05,
+      "loss": 0.336,
+      "step": 25270
+    },
+    {
+      "epoch": 6.304239401496259,
+      "grad_norm": 6.88414192199707,
+      "learning_rate": 1.3698753117206984e-05,
+      "loss": 0.3996,
+      "step": 25280
+    },
+    {
+      "epoch": 6.306733167082294,
+      "grad_norm": 5.107346057891846,
+      "learning_rate": 1.3696259351620948e-05,
+      "loss": 0.3807,
+      "step": 25290
+    },
+    {
+      "epoch": 6.309226932668329,
+      "grad_norm": 6.0163445472717285,
+      "learning_rate": 1.3693765586034913e-05,
+      "loss": 0.3609,
+      "step": 25300
+    },
+    {
+      "epoch": 6.311720698254364,
+      "grad_norm": 6.904167175292969,
+      "learning_rate": 1.3691271820448879e-05,
+      "loss": 0.3868,
+      "step": 25310
+    },
+    {
+      "epoch": 6.314214463840399,
+      "grad_norm": 7.417514801025391,
+      "learning_rate": 1.3688778054862844e-05,
+      "loss": 0.384,
+      "step": 25320
+    },
+    {
+      "epoch": 6.316708229426434,
+      "grad_norm": 7.786649227142334,
+      "learning_rate": 1.368628428927681e-05,
+      "loss": 0.4802,
+      "step": 25330
+    },
+    {
+      "epoch": 6.3192019950124685,
+      "grad_norm": 7.041213512420654,
+      "learning_rate": 1.3683790523690775e-05,
+      "loss": 0.4231,
+      "step": 25340
+    },
+    {
+      "epoch": 6.321695760598503,
+      "grad_norm": 4.815100193023682,
+      "learning_rate": 1.3681296758104738e-05,
+      "loss": 0.3803,
+      "step": 25350
+    },
+    {
+      "epoch": 6.324189526184538,
+      "grad_norm": 5.091302871704102,
+      "learning_rate": 1.3678802992518705e-05,
+      "loss": 0.3602,
+      "step": 25360
+    },
+    {
+      "epoch": 6.326683291770574,
+      "grad_norm": 6.59751558303833,
+      "learning_rate": 1.3676309226932669e-05,
+      "loss": 0.3558,
+      "step": 25370
+    },
+    {
+      "epoch": 6.329177057356609,
+      "grad_norm": 6.500771522521973,
+      "learning_rate": 1.3673815461346636e-05,
+      "loss": 0.3743,
+      "step": 25380
+    },
+    {
+      "epoch": 6.331670822942644,
+      "grad_norm": 5.946366786956787,
+      "learning_rate": 1.36713216957606e-05,
+      "loss": 0.3757,
+      "step": 25390
+    },
+    {
+      "epoch": 6.334164588528679,
+      "grad_norm": 6.4445672035217285,
+      "learning_rate": 1.3668827930174565e-05,
+      "loss": 0.3488,
+      "step": 25400
+    },
+    {
+      "epoch": 6.3366583541147135,
+      "grad_norm": 8.107527732849121,
+      "learning_rate": 1.366633416458853e-05,
+      "loss": 0.3637,
+      "step": 25410
+    },
+    {
+      "epoch": 6.339152119700748,
+      "grad_norm": 7.021248817443848,
+      "learning_rate": 1.3663840399002495e-05,
+      "loss": 0.4392,
+      "step": 25420
+    },
+    {
+      "epoch": 6.341645885286783,
+      "grad_norm": 7.19385290145874,
+      "learning_rate": 1.3661346633416459e-05,
+      "loss": 0.3443,
+      "step": 25430
+    },
+    {
+      "epoch": 6.344139650872818,
+      "grad_norm": 8.480057716369629,
+      "learning_rate": 1.3658852867830426e-05,
+      "loss": 0.3249,
+      "step": 25440
+    },
+    {
+      "epoch": 6.346633416458853,
+      "grad_norm": 6.283059120178223,
+      "learning_rate": 1.365635910224439e-05,
+      "loss": 0.4301,
+      "step": 25450
+    },
+    {
+      "epoch": 6.349127182044888,
+      "grad_norm": 9.004962921142578,
+      "learning_rate": 1.3653865336658357e-05,
+      "loss": 0.3856,
+      "step": 25460
+    },
+    {
+      "epoch": 6.351620947630923,
+      "grad_norm": 6.921197891235352,
+      "learning_rate": 1.365137157107232e-05,
+      "loss": 0.3504,
+      "step": 25470
+    },
+    {
+      "epoch": 6.3541147132169575,
+      "grad_norm": 7.9730634689331055,
+      "learning_rate": 1.3648877805486284e-05,
+      "loss": 0.4598,
+      "step": 25480
+    },
+    {
+      "epoch": 6.356608478802992,
+      "grad_norm": 5.783198356628418,
+      "learning_rate": 1.3646384039900251e-05,
+      "loss": 0.388,
+      "step": 25490
+    },
+    {
+      "epoch": 6.359102244389027,
+      "grad_norm": 5.059525489807129,
+      "learning_rate": 1.3643890274314216e-05,
+      "loss": 0.327,
+      "step": 25500
+    },
+    {
+      "epoch": 6.361596009975062,
+      "grad_norm": 8.875046730041504,
+      "learning_rate": 1.364139650872818e-05,
+      "loss": 0.3387,
+      "step": 25510
+    },
+    {
+      "epoch": 6.364089775561097,
+      "grad_norm": 8.62806224822998,
+      "learning_rate": 1.3638902743142147e-05,
+      "loss": 0.3109,
+      "step": 25520
+    },
+    {
+      "epoch": 6.366583541147132,
+      "grad_norm": 6.7610344886779785,
+      "learning_rate": 1.363640897755611e-05,
+      "loss": 0.3855,
+      "step": 25530
+    },
+    {
+      "epoch": 6.369077306733167,
+      "grad_norm": 10.090450286865234,
+      "learning_rate": 1.3633915211970077e-05,
+      "loss": 0.3404,
+      "step": 25540
+    },
+    {
+      "epoch": 6.371571072319202,
+      "grad_norm": 7.940425872802734,
+      "learning_rate": 1.3631421446384041e-05,
+      "loss": 0.3752,
+      "step": 25550
+    },
+    {
+      "epoch": 6.374064837905237,
+      "grad_norm": 6.398936748504639,
+      "learning_rate": 1.3628927680798005e-05,
+      "loss": 0.3769,
+      "step": 25560
+    },
+    {
+      "epoch": 6.376558603491272,
+      "grad_norm": 4.926769733428955,
+      "learning_rate": 1.3626433915211972e-05,
+      "loss": 0.4024,
+      "step": 25570
+    },
+    {
+      "epoch": 6.379052369077307,
+      "grad_norm": 7.145235538482666,
+      "learning_rate": 1.3623940149625935e-05,
+      "loss": 0.3512,
+      "step": 25580
+    },
+    {
+      "epoch": 6.381546134663342,
+      "grad_norm": 8.193603515625,
+      "learning_rate": 1.36214463840399e-05,
+      "loss": 0.3556,
+      "step": 25590
+    },
+    {
+      "epoch": 6.384039900249377,
+      "grad_norm": 6.601048469543457,
+      "learning_rate": 1.3618952618453866e-05,
+      "loss": 0.442,
+      "step": 25600
+    },
+    {
+      "epoch": 6.386533665835412,
+      "grad_norm": 6.299713134765625,
+      "learning_rate": 1.3616458852867831e-05,
+      "loss": 0.3942,
+      "step": 25610
+    },
+    {
+      "epoch": 6.389027431421447,
+      "grad_norm": 5.013650417327881,
+      "learning_rate": 1.3613965087281798e-05,
+      "loss": 0.3576,
+      "step": 25620
+    },
+    {
+      "epoch": 6.3915211970074814,
+      "grad_norm": 6.1320905685424805,
+      "learning_rate": 1.3611471321695762e-05,
+      "loss": 0.426,
+      "step": 25630
+    },
+    {
+      "epoch": 6.394014962593516,
+      "grad_norm": 7.493283748626709,
+      "learning_rate": 1.3608977556109725e-05,
+      "loss": 0.4207,
+      "step": 25640
+    },
+    {
+      "epoch": 6.396508728179551,
+      "grad_norm": 7.163254261016846,
+      "learning_rate": 1.3606483790523692e-05,
+      "loss": 0.4041,
+      "step": 25650
+    },
+    {
+      "epoch": 6.399002493765586,
+      "grad_norm": 4.770473957061768,
+      "learning_rate": 1.3603990024937656e-05,
+      "loss": 0.3493,
+      "step": 25660
+    },
+    {
+      "epoch": 6.401496259351621,
+      "grad_norm": 6.536076545715332,
+      "learning_rate": 1.3601496259351623e-05,
+      "loss": 0.3204,
+      "step": 25670
+    },
+    {
+      "epoch": 6.403990024937656,
+      "grad_norm": 8.567519187927246,
+      "learning_rate": 1.3599002493765587e-05,
+      "loss": 0.4084,
+      "step": 25680
+    },
+    {
+      "epoch": 6.406483790523691,
+      "grad_norm": 10.568842887878418,
+      "learning_rate": 1.3596508728179552e-05,
+      "loss": 0.3572,
+      "step": 25690
+    },
+    {
+      "epoch": 6.4089775561097255,
+      "grad_norm": 8.251084327697754,
+      "learning_rate": 1.3594014962593517e-05,
+      "loss": 0.4384,
+      "step": 25700
+    },
+    {
+      "epoch": 6.41147132169576,
+      "grad_norm": 5.1175408363342285,
+      "learning_rate": 1.3591521197007483e-05,
+      "loss": 0.3512,
+      "step": 25710
+    },
+    {
+      "epoch": 6.413965087281795,
+      "grad_norm": 5.352699279785156,
+      "learning_rate": 1.3589027431421446e-05,
+      "loss": 0.3152,
+      "step": 25720
+    },
+    {
+      "epoch": 6.41645885286783,
+      "grad_norm": 11.443229675292969,
+      "learning_rate": 1.3586533665835413e-05,
+      "loss": 0.3842,
+      "step": 25730
+    },
+    {
+      "epoch": 6.418952618453865,
+      "grad_norm": 4.957744598388672,
+      "learning_rate": 1.3584039900249377e-05,
+      "loss": 0.353,
+      "step": 25740
+    },
+    {
+      "epoch": 6.4214463840399,
+      "grad_norm": 5.221795082092285,
+      "learning_rate": 1.3581546134663344e-05,
+      "loss": 0.3797,
+      "step": 25750
+    },
+    {
+      "epoch": 6.423940149625935,
+      "grad_norm": 5.243051052093506,
+      "learning_rate": 1.3579052369077307e-05,
+      "loss": 0.3356,
+      "step": 25760
+    },
+    {
+      "epoch": 6.42643391521197,
+      "grad_norm": 6.537060260772705,
+      "learning_rate": 1.3576558603491273e-05,
+      "loss": 0.4025,
+      "step": 25770
+    },
+    {
+      "epoch": 6.428927680798005,
+      "grad_norm": 7.9792094230651855,
+      "learning_rate": 1.3574064837905238e-05,
+      "loss": 0.4992,
+      "step": 25780
+    },
+    {
+      "epoch": 6.43142144638404,
+      "grad_norm": 6.797430992126465,
+      "learning_rate": 1.3571571072319203e-05,
+      "loss": 0.3187,
+      "step": 25790
+    },
+    {
+      "epoch": 6.433915211970075,
+      "grad_norm": 5.758587837219238,
+      "learning_rate": 1.3569077306733167e-05,
+      "loss": 0.3202,
+      "step": 25800
+    },
+    {
+      "epoch": 6.43640897755611,
+      "grad_norm": 6.541738986968994,
+      "learning_rate": 1.3566583541147134e-05,
+      "loss": 0.4289,
+      "step": 25810
+    },
+    {
+      "epoch": 6.438902743142145,
+      "grad_norm": 7.5989227294921875,
+      "learning_rate": 1.3564089775561098e-05,
+      "loss": 0.4379,
+      "step": 25820
+    },
+    {
+      "epoch": 6.44139650872818,
+      "grad_norm": 10.161792755126953,
+      "learning_rate": 1.3561596009975065e-05,
+      "loss": 0.344,
+      "step": 25830
+    },
+    {
+      "epoch": 6.443890274314215,
+      "grad_norm": 6.123249053955078,
+      "learning_rate": 1.3559102244389028e-05,
+      "loss": 0.3364,
+      "step": 25840
+    },
+    {
+      "epoch": 6.446384039900249,
+      "grad_norm": 4.593570709228516,
+      "learning_rate": 1.3556608478802994e-05,
+      "loss": 0.4108,
+      "step": 25850
+    },
+    {
+      "epoch": 6.448877805486284,
+      "grad_norm": 14.478041648864746,
+      "learning_rate": 1.3554114713216959e-05,
+      "loss": 0.2916,
+      "step": 25860
+    },
+    {
+      "epoch": 6.451371571072319,
+      "grad_norm": 9.636341094970703,
+      "learning_rate": 1.3551620947630924e-05,
+      "loss": 0.4033,
+      "step": 25870
+    },
+    {
+      "epoch": 6.453865336658354,
+      "grad_norm": 6.884788513183594,
+      "learning_rate": 1.354912718204489e-05,
+      "loss": 0.3478,
+      "step": 25880
+    },
+    {
+      "epoch": 6.456359102244389,
+      "grad_norm": 6.347228527069092,
+      "learning_rate": 1.3546633416458855e-05,
+      "loss": 0.3867,
+      "step": 25890
+    },
+    {
+      "epoch": 6.458852867830424,
+      "grad_norm": 6.127768516540527,
+      "learning_rate": 1.3544139650872818e-05,
+      "loss": 0.3709,
+      "step": 25900
+    },
+    {
+      "epoch": 6.461346633416459,
+      "grad_norm": 5.719086647033691,
+      "learning_rate": 1.3541645885286785e-05,
+      "loss": 0.3649,
+      "step": 25910
+    },
+    {
+      "epoch": 6.4638403990024935,
+      "grad_norm": 7.747276306152344,
+      "learning_rate": 1.3539152119700749e-05,
+      "loss": 0.3425,
+      "step": 25920
+    },
+    {
+      "epoch": 6.466334164588528,
+      "grad_norm": 7.665780544281006,
+      "learning_rate": 1.3536658354114713e-05,
+      "loss": 0.427,
+      "step": 25930
+    },
+    {
+      "epoch": 6.468827930174563,
+      "grad_norm": 4.468764305114746,
+      "learning_rate": 1.353416458852868e-05,
+      "loss": 0.3758,
+      "step": 25940
+    },
+    {
+      "epoch": 6.471321695760598,
+      "grad_norm": 8.336400985717773,
+      "learning_rate": 1.3531670822942643e-05,
+      "loss": 0.3451,
+      "step": 25950
+    },
+    {
+      "epoch": 6.473815461346634,
+      "grad_norm": 9.712798118591309,
+      "learning_rate": 1.352917705735661e-05,
+      "loss": 0.3795,
+      "step": 25960
+    },
+    {
+      "epoch": 6.476309226932669,
+      "grad_norm": 7.729135990142822,
+      "learning_rate": 1.3526683291770576e-05,
+      "loss": 0.3696,
+      "step": 25970
+    },
+    {
+      "epoch": 6.478802992518704,
+      "grad_norm": 6.578273296356201,
+      "learning_rate": 1.352418952618454e-05,
+      "loss": 0.4044,
+      "step": 25980
+    },
+    {
+      "epoch": 6.4812967581047385,
+      "grad_norm": 11.453405380249023,
+      "learning_rate": 1.3521695760598506e-05,
+      "loss": 0.3201,
+      "step": 25990
+    },
+    {
+      "epoch": 6.483790523690773,
+      "grad_norm": 6.865024566650391,
+      "learning_rate": 1.351920199501247e-05,
+      "loss": 0.3353,
+      "step": 26000
+    },
+    {
+      "epoch": 6.486284289276808,
+      "grad_norm": 6.444002151489258,
+      "learning_rate": 1.3516708229426433e-05,
+      "loss": 0.3806,
+      "step": 26010
+    },
+    {
+      "epoch": 6.488778054862843,
+      "grad_norm": 7.583583831787109,
+      "learning_rate": 1.35142144638404e-05,
+      "loss": 0.3622,
+      "step": 26020
+    },
+    {
+      "epoch": 6.491271820448878,
+      "grad_norm": 7.034835338592529,
+      "learning_rate": 1.3511720698254364e-05,
+      "loss": 0.4238,
+      "step": 26030
+    },
+    {
+      "epoch": 6.493765586034913,
+      "grad_norm": 7.059061050415039,
+      "learning_rate": 1.3509226932668331e-05,
+      "loss": 0.3675,
+      "step": 26040
+    },
+    {
+      "epoch": 6.496259351620948,
+      "grad_norm": 5.760257720947266,
+      "learning_rate": 1.3506733167082295e-05,
+      "loss": 0.3814,
+      "step": 26050
+    },
+    {
+      "epoch": 6.498753117206983,
+      "grad_norm": 6.137911319732666,
+      "learning_rate": 1.350423940149626e-05,
+      "loss": 0.3576,
+      "step": 26060
+    },
+    {
+      "epoch": 6.501246882793017,
+      "grad_norm": 7.457276821136475,
+      "learning_rate": 1.3501745635910225e-05,
+      "loss": 0.3941,
+      "step": 26070
+    },
+    {
+      "epoch": 6.503740648379052,
+      "grad_norm": 7.268845081329346,
+      "learning_rate": 1.349925187032419e-05,
+      "loss": 0.3924,
+      "step": 26080
+    },
+    {
+      "epoch": 6.506234413965087,
+      "grad_norm": 7.861654758453369,
+      "learning_rate": 1.3496758104738154e-05,
+      "loss": 0.3578,
+      "step": 26090
+    },
+    {
+      "epoch": 6.508728179551122,
+      "grad_norm": 8.676166534423828,
+      "learning_rate": 1.3494264339152121e-05,
+      "loss": 0.4239,
+      "step": 26100
+    },
+    {
+      "epoch": 6.511221945137157,
+      "grad_norm": 6.739658355712891,
+      "learning_rate": 1.3491770573566085e-05,
+      "loss": 0.4131,
+      "step": 26110
+    },
+    {
+      "epoch": 6.513715710723192,
+      "grad_norm": 6.20055627822876,
+      "learning_rate": 1.3489276807980052e-05,
+      "loss": 0.4278,
+      "step": 26120
+    },
+    {
+      "epoch": 6.516209476309227,
+      "grad_norm": 6.212075710296631,
+      "learning_rate": 1.3486783042394015e-05,
+      "loss": 0.3607,
+      "step": 26130
+    },
+    {
+      "epoch": 6.5187032418952615,
+      "grad_norm": 10.847241401672363,
+      "learning_rate": 1.348428927680798e-05,
+      "loss": 0.3624,
+      "step": 26140
+    },
+    {
+      "epoch": 6.521197007481296,
+      "grad_norm": 6.094544887542725,
+      "learning_rate": 1.3481795511221946e-05,
+      "loss": 0.4353,
+      "step": 26150
+    },
+    {
+      "epoch": 6.523690773067331,
+      "grad_norm": 7.994797706604004,
+      "learning_rate": 1.3479301745635911e-05,
+      "loss": 0.3961,
+      "step": 26160
+    },
+    {
+      "epoch": 6.526184538653366,
+      "grad_norm": 8.158559799194336,
+      "learning_rate": 1.3476807980049877e-05,
+      "loss": 0.3696,
+      "step": 26170
+    },
+    {
+      "epoch": 6.528678304239402,
+      "grad_norm": 7.66077995300293,
+      "learning_rate": 1.3474314214463842e-05,
+      "loss": 0.3302,
+      "step": 26180
+    },
+    {
+      "epoch": 6.531172069825437,
+      "grad_norm": 6.937994480133057,
+      "learning_rate": 1.3471820448877806e-05,
+      "loss": 0.4102,
+      "step": 26190
+    },
+    {
+      "epoch": 6.533665835411472,
+      "grad_norm": 7.02347993850708,
+      "learning_rate": 1.3469326683291773e-05,
+      "loss": 0.4596,
+      "step": 26200
+    },
+    {
+      "epoch": 6.5361596009975065,
+      "grad_norm": 7.245387077331543,
+      "learning_rate": 1.3466832917705736e-05,
+      "loss": 0.3513,
+      "step": 26210
+    },
+    {
+      "epoch": 6.538653366583541,
+      "grad_norm": 7.9023237228393555,
+      "learning_rate": 1.3464339152119702e-05,
+      "loss": 0.3554,
+      "step": 26220
+    },
+    {
+      "epoch": 6.541147132169576,
+      "grad_norm": 5.199001789093018,
+      "learning_rate": 1.3461845386533667e-05,
+      "loss": 0.4321,
+      "step": 26230
+    },
+    {
+      "epoch": 6.543640897755611,
+      "grad_norm": 5.115858554840088,
+      "learning_rate": 1.3459351620947632e-05,
+      "loss": 0.3597,
+      "step": 26240
+    },
+    {
+      "epoch": 6.546134663341646,
+      "grad_norm": 7.653583526611328,
+      "learning_rate": 1.3456857855361597e-05,
+      "loss": 0.4044,
+      "step": 26250
+    },
+    {
+      "epoch": 6.548628428927681,
+      "grad_norm": 5.315465450286865,
+      "learning_rate": 1.3454364089775563e-05,
+      "loss": 0.3737,
+      "step": 26260
+    },
+    {
+      "epoch": 6.551122194513716,
+      "grad_norm": 6.182226657867432,
+      "learning_rate": 1.3451870324189526e-05,
+      "loss": 0.3521,
+      "step": 26270
+    },
+    {
+      "epoch": 6.553615960099751,
+      "grad_norm": 5.8310546875,
+      "learning_rate": 1.3449376558603493e-05,
+      "loss": 0.3277,
+      "step": 26280
+    },
+    {
+      "epoch": 6.556109725685785,
+      "grad_norm": 6.5603437423706055,
+      "learning_rate": 1.3446882793017457e-05,
+      "loss": 0.3724,
+      "step": 26290
+    },
+    {
+      "epoch": 6.55860349127182,
+      "grad_norm": 7.382344722747803,
+      "learning_rate": 1.344438902743142e-05,
+      "loss": 0.3898,
+      "step": 26300
+    },
+    {
+      "epoch": 6.561097256857855,
+      "grad_norm": 5.281245708465576,
+      "learning_rate": 1.3441895261845388e-05,
+      "loss": 0.3697,
+      "step": 26310
+    },
+    {
+      "epoch": 6.56359102244389,
+      "grad_norm": 6.038678169250488,
+      "learning_rate": 1.3439401496259353e-05,
+      "loss": 0.3634,
+      "step": 26320
+    },
+    {
+      "epoch": 6.566084788029925,
+      "grad_norm": 8.70938491821289,
+      "learning_rate": 1.3436907730673318e-05,
+      "loss": 0.3867,
+      "step": 26330
+    },
+    {
+      "epoch": 6.56857855361596,
+      "grad_norm": 7.3120598793029785,
+      "learning_rate": 1.3434413965087284e-05,
+      "loss": 0.3791,
+      "step": 26340
+    },
+    {
+      "epoch": 6.571072319201995,
+      "grad_norm": 6.2208428382873535,
+      "learning_rate": 1.3431920199501247e-05,
+      "loss": 0.3998,
+      "step": 26350
+    },
+    {
+      "epoch": 6.57356608478803,
+      "grad_norm": 6.549139499664307,
+      "learning_rate": 1.3429426433915214e-05,
+      "loss": 0.3935,
+      "step": 26360
+    },
+    {
+      "epoch": 6.576059850374065,
+      "grad_norm": 4.998687267303467,
+      "learning_rate": 1.3426932668329178e-05,
+      "loss": 0.41,
+      "step": 26370
+    },
+    {
+      "epoch": 6.5785536159601,
+      "grad_norm": 5.387908458709717,
+      "learning_rate": 1.3424438902743145e-05,
+      "loss": 0.3421,
+      "step": 26380
+    },
+    {
+      "epoch": 6.581047381546135,
+      "grad_norm": 7.001126766204834,
+      "learning_rate": 1.3421945137157108e-05,
+      "loss": 0.3603,
+      "step": 26390
+    },
+    {
+      "epoch": 6.58354114713217,
+      "grad_norm": 10.226015090942383,
+      "learning_rate": 1.3419451371571072e-05,
+      "loss": 0.3955,
+      "step": 26400
+    },
+    {
+      "epoch": 6.586034912718205,
+      "grad_norm": 6.176059722900391,
+      "learning_rate": 1.3416957605985039e-05,
+      "loss": 0.302,
+      "step": 26410
+    },
+    {
+      "epoch": 6.58852867830424,
+      "grad_norm": 5.293604373931885,
+      "learning_rate": 1.3414463840399003e-05,
+      "loss": 0.3701,
+      "step": 26420
+    },
+    {
+      "epoch": 6.5910224438902745,
+      "grad_norm": 8.43692398071289,
+      "learning_rate": 1.3411970074812968e-05,
+      "loss": 0.3852,
+      "step": 26430
+    },
+    {
+      "epoch": 6.593516209476309,
+      "grad_norm": 7.703579425811768,
+      "learning_rate": 1.3409476309226935e-05,
+      "loss": 0.4242,
+      "step": 26440
+    },
+    {
+      "epoch": 6.596009975062344,
+      "grad_norm": 9.26103687286377,
+      "learning_rate": 1.3406982543640899e-05,
+      "loss": 0.3328,
+      "step": 26450
+    },
+    {
+      "epoch": 6.598503740648379,
+      "grad_norm": 7.32110071182251,
+      "learning_rate": 1.3404488778054866e-05,
+      "loss": 0.3679,
+      "step": 26460
+    },
+    {
+      "epoch": 6.600997506234414,
+      "grad_norm": 6.620180130004883,
+      "learning_rate": 1.340199501246883e-05,
+      "loss": 0.4228,
+      "step": 26470
+    },
+    {
+      "epoch": 6.603491271820449,
+      "grad_norm": 7.809380531311035,
+      "learning_rate": 1.3399501246882793e-05,
+      "loss": 0.4333,
+      "step": 26480
+    },
+    {
+      "epoch": 6.605985037406484,
+      "grad_norm": 5.4282331466674805,
+      "learning_rate": 1.339700748129676e-05,
+      "loss": 0.4154,
+      "step": 26490
+    },
+    {
+      "epoch": 6.6084788029925186,
+      "grad_norm": 6.508066177368164,
+      "learning_rate": 1.3394513715710723e-05,
+      "loss": 0.3998,
+      "step": 26500
+    },
+    {
+      "epoch": 6.610972568578553,
+      "grad_norm": 9.248714447021484,
+      "learning_rate": 1.3392019950124689e-05,
+      "loss": 0.3763,
+      "step": 26510
+    },
+    {
+      "epoch": 6.613466334164588,
+      "grad_norm": 7.227468967437744,
+      "learning_rate": 1.3389526184538654e-05,
+      "loss": 0.3659,
+      "step": 26520
+    },
+    {
+      "epoch": 6.615960099750623,
+      "grad_norm": 7.72911262512207,
+      "learning_rate": 1.338703241895262e-05,
+      "loss": 0.4427,
+      "step": 26530
+    },
+    {
+      "epoch": 6.618453865336658,
+      "grad_norm": 5.077463626861572,
+      "learning_rate": 1.3384538653366585e-05,
+      "loss": 0.3655,
+      "step": 26540
+    },
+    {
+      "epoch": 6.620947630922693,
+      "grad_norm": 5.688047885894775,
+      "learning_rate": 1.338204488778055e-05,
+      "loss": 0.3803,
+      "step": 26550
+    },
+    {
+      "epoch": 6.623441396508728,
+      "grad_norm": 5.621146202087402,
+      "learning_rate": 1.3379551122194514e-05,
+      "loss": 0.3938,
+      "step": 26560
+    },
+    {
+      "epoch": 6.625935162094763,
+      "grad_norm": 7.967491626739502,
+      "learning_rate": 1.337705735660848e-05,
+      "loss": 0.3353,
+      "step": 26570
+    },
+    {
+      "epoch": 6.628428927680798,
+      "grad_norm": 10.237777709960938,
+      "learning_rate": 1.3374563591022444e-05,
+      "loss": 0.3249,
+      "step": 26580
+    },
+    {
+      "epoch": 6.630922693266833,
+      "grad_norm": 13.385808944702148,
+      "learning_rate": 1.337206982543641e-05,
+      "loss": 0.4134,
+      "step": 26590
+    },
+    {
+      "epoch": 6.633416458852868,
+      "grad_norm": 4.2393341064453125,
+      "learning_rate": 1.3369576059850375e-05,
+      "loss": 0.4,
+      "step": 26600
+    },
+    {
+      "epoch": 6.635910224438903,
+      "grad_norm": 4.3933281898498535,
+      "learning_rate": 1.336708229426434e-05,
+      "loss": 0.3216,
+      "step": 26610
+    },
+    {
+      "epoch": 6.638403990024938,
+      "grad_norm": 4.819437503814697,
+      "learning_rate": 1.3364588528678305e-05,
+      "loss": 0.3578,
+      "step": 26620
+    },
+    {
+      "epoch": 6.640897755610973,
+      "grad_norm": 4.126856803894043,
+      "learning_rate": 1.336209476309227e-05,
+      "loss": 0.3625,
+      "step": 26630
+    },
+    {
+      "epoch": 6.643391521197008,
+      "grad_norm": 5.08698034286499,
+      "learning_rate": 1.3359600997506234e-05,
+      "loss": 0.417,
+      "step": 26640
+    },
+    {
+      "epoch": 6.6458852867830425,
+      "grad_norm": 5.38383674621582,
+      "learning_rate": 1.3357107231920201e-05,
+      "loss": 0.356,
+      "step": 26650
+    },
+    {
+      "epoch": 6.648379052369077,
+      "grad_norm": 8.648747444152832,
+      "learning_rate": 1.3354613466334165e-05,
+      "loss": 0.3886,
+      "step": 26660
+    },
+    {
+      "epoch": 6.650872817955112,
+      "grad_norm": 4.306643486022949,
+      "learning_rate": 1.3352119700748132e-05,
+      "loss": 0.3493,
+      "step": 26670
+    },
+    {
+      "epoch": 6.653366583541147,
+      "grad_norm": 6.597197532653809,
+      "learning_rate": 1.3349625935162096e-05,
+      "loss": 0.372,
+      "step": 26680
+    },
+    {
+      "epoch": 6.655860349127182,
+      "grad_norm": 6.69064474105835,
+      "learning_rate": 1.3347132169576061e-05,
+      "loss": 0.3965,
+      "step": 26690
+    },
+    {
+      "epoch": 6.658354114713217,
+      "grad_norm": 7.140894412994385,
+      "learning_rate": 1.3344638403990026e-05,
+      "loss": 0.3822,
+      "step": 26700
+    },
+    {
+      "epoch": 6.660847880299252,
+      "grad_norm": 5.859182357788086,
+      "learning_rate": 1.3342144638403992e-05,
+      "loss": 0.3458,
+      "step": 26710
+    },
+    {
+      "epoch": 6.6633416458852865,
+      "grad_norm": 6.949796199798584,
+      "learning_rate": 1.3339650872817955e-05,
+      "loss": 0.4071,
+      "step": 26720
+    },
+    {
+      "epoch": 6.665835411471321,
+      "grad_norm": 8.996414184570312,
+      "learning_rate": 1.3337157107231922e-05,
+      "loss": 0.3914,
+      "step": 26730
+    },
+    {
+      "epoch": 6.668329177057356,
+      "grad_norm": 9.598322868347168,
+      "learning_rate": 1.3334663341645886e-05,
+      "loss": 0.3846,
+      "step": 26740
+    },
+    {
+      "epoch": 6.670822942643391,
+      "grad_norm": 5.436387062072754,
+      "learning_rate": 1.3332169576059853e-05,
+      "loss": 0.3324,
+      "step": 26750
+    },
+    {
+      "epoch": 6.673316708229427,
+      "grad_norm": 8.324487686157227,
+      "learning_rate": 1.3329675810473816e-05,
+      "loss": 0.3996,
+      "step": 26760
+    },
+    {
+      "epoch": 6.675810473815462,
+      "grad_norm": 9.037124633789062,
+      "learning_rate": 1.3327182044887782e-05,
+      "loss": 0.4561,
+      "step": 26770
+    },
+    {
+      "epoch": 6.678304239401497,
+      "grad_norm": 7.689178466796875,
+      "learning_rate": 1.3324688279301747e-05,
+      "loss": 0.4511,
+      "step": 26780
+    },
+    {
+      "epoch": 6.6807980049875315,
+      "grad_norm": 6.272636890411377,
+      "learning_rate": 1.3322194513715712e-05,
+      "loss": 0.4475,
+      "step": 26790
+    },
+    {
+      "epoch": 6.683291770573566,
+      "grad_norm": 4.632606029510498,
+      "learning_rate": 1.3319700748129676e-05,
+      "loss": 0.4122,
+      "step": 26800
+    },
+    {
+      "epoch": 6.685785536159601,
+      "grad_norm": 7.478541851043701,
+      "learning_rate": 1.3317206982543643e-05,
+      "loss": 0.4107,
+      "step": 26810
+    },
+    {
+      "epoch": 6.688279301745636,
+      "grad_norm": 10.429938316345215,
+      "learning_rate": 1.3314713216957607e-05,
+      "loss": 0.4623,
+      "step": 26820
+    },
+    {
+      "epoch": 6.690773067331671,
+      "grad_norm": 7.3426995277404785,
+      "learning_rate": 1.3312219451371574e-05,
+      "loss": 0.4568,
+      "step": 26830
+    },
+    {
+      "epoch": 6.693266832917706,
+      "grad_norm": 7.244011402130127,
+      "learning_rate": 1.3309725685785537e-05,
+      "loss": 0.3298,
+      "step": 26840
+    },
+    {
+      "epoch": 6.695760598503741,
+      "grad_norm": 9.129273414611816,
+      "learning_rate": 1.33072319201995e-05,
+      "loss": 0.3867,
+      "step": 26850
+    },
+    {
+      "epoch": 6.698254364089776,
+      "grad_norm": 7.908605575561523,
+      "learning_rate": 1.3304738154613468e-05,
+      "loss": 0.394,
+      "step": 26860
+    },
+    {
+      "epoch": 6.7007481296758105,
+      "grad_norm": 5.406508445739746,
+      "learning_rate": 1.3302244389027431e-05,
+      "loss": 0.3463,
+      "step": 26870
+    },
+    {
+      "epoch": 6.703241895261845,
+      "grad_norm": 7.902821063995361,
+      "learning_rate": 1.3299750623441398e-05,
+      "loss": 0.4256,
+      "step": 26880
+    },
+    {
+      "epoch": 6.70573566084788,
+      "grad_norm": 5.451107978820801,
+      "learning_rate": 1.3297256857855362e-05,
+      "loss": 0.3999,
+      "step": 26890
+    },
+    {
+      "epoch": 6.708229426433915,
+      "grad_norm": 4.476869106292725,
+      "learning_rate": 1.3294763092269327e-05,
+      "loss": 0.3836,
+      "step": 26900
+    },
+    {
+      "epoch": 6.71072319201995,
+      "grad_norm": 6.899260520935059,
+      "learning_rate": 1.3292269326683294e-05,
+      "loss": 0.4067,
+      "step": 26910
+    },
+    {
+      "epoch": 6.713216957605985,
+      "grad_norm": 8.896599769592285,
+      "learning_rate": 1.3289775561097258e-05,
+      "loss": 0.4425,
+      "step": 26920
+    },
+    {
+      "epoch": 6.71571072319202,
+      "grad_norm": 7.202389717102051,
+      "learning_rate": 1.3287281795511222e-05,
+      "loss": 0.3334,
+      "step": 26930
+    },
+    {
+      "epoch": 6.7182044887780545,
+      "grad_norm": 5.247492790222168,
+      "learning_rate": 1.3284788029925189e-05,
+      "loss": 0.4,
+      "step": 26940
+    },
+    {
+      "epoch": 6.720698254364089,
+      "grad_norm": 10.521095275878906,
+      "learning_rate": 1.3282294264339152e-05,
+      "loss": 0.3652,
+      "step": 26950
+    },
+    {
+      "epoch": 6.723192019950124,
+      "grad_norm": 5.5208611488342285,
+      "learning_rate": 1.327980049875312e-05,
+      "loss": 0.4477,
+      "step": 26960
+    },
+    {
+      "epoch": 6.725685785536159,
+      "grad_norm": 8.8683500289917,
+      "learning_rate": 1.3277306733167083e-05,
+      "loss": 0.3449,
+      "step": 26970
+    },
+    {
+      "epoch": 6.728179551122195,
+      "grad_norm": 5.689904689788818,
+      "learning_rate": 1.3274812967581048e-05,
+      "loss": 0.4239,
+      "step": 26980
+    },
+    {
+      "epoch": 6.73067331670823,
+      "grad_norm": 7.4817328453063965,
+      "learning_rate": 1.3272319201995013e-05,
+      "loss": 0.4688,
+      "step": 26990
+    },
+    {
+      "epoch": 6.733167082294265,
+      "grad_norm": 8.645492553710938,
+      "learning_rate": 1.3269825436408979e-05,
+      "loss": 0.415,
+      "step": 27000
+    },
+    {
+      "epoch": 6.7356608478802995,
+      "grad_norm": 4.103111743927002,
+      "learning_rate": 1.3267331670822942e-05,
+      "loss": 0.3628,
+      "step": 27010
+    },
+    {
+      "epoch": 6.738154613466334,
+      "grad_norm": 6.457451820373535,
+      "learning_rate": 1.326483790523691e-05,
+      "loss": 0.4813,
+      "step": 27020
+    },
+    {
+      "epoch": 6.740648379052369,
+      "grad_norm": 11.222722053527832,
+      "learning_rate": 1.3262344139650873e-05,
+      "loss": 0.3365,
+      "step": 27030
+    },
+    {
+      "epoch": 6.743142144638404,
+      "grad_norm": 5.118739128112793,
+      "learning_rate": 1.325985037406484e-05,
+      "loss": 0.35,
+      "step": 27040
+    },
+    {
+      "epoch": 6.745635910224439,
+      "grad_norm": 8.133965492248535,
+      "learning_rate": 1.3257356608478804e-05,
+      "loss": 0.3907,
+      "step": 27050
+    },
+    {
+      "epoch": 6.748129675810474,
+      "grad_norm": 7.104856491088867,
+      "learning_rate": 1.3254862842892769e-05,
+      "loss": 0.4105,
+      "step": 27060
+    },
+    {
+      "epoch": 6.750623441396509,
+      "grad_norm": 5.757190227508545,
+      "learning_rate": 1.3252369077306734e-05,
+      "loss": 0.4462,
+      "step": 27070
+    },
+    {
+      "epoch": 6.753117206982544,
+      "grad_norm": 8.72276782989502,
+      "learning_rate": 1.32498753117207e-05,
+      "loss": 0.344,
+      "step": 27080
+    },
+    {
+      "epoch": 6.7556109725685785,
+      "grad_norm": 3.8614084720611572,
+      "learning_rate": 1.3247381546134663e-05,
+      "loss": 0.2905,
+      "step": 27090
+    },
+    {
+      "epoch": 6.758104738154613,
+      "grad_norm": 6.985355377197266,
+      "learning_rate": 1.324488778054863e-05,
+      "loss": 0.3524,
+      "step": 27100
+    },
+    {
+      "epoch": 6.760598503740648,
+      "grad_norm": 5.594140529632568,
+      "learning_rate": 1.3242394014962594e-05,
+      "loss": 0.4011,
+      "step": 27110
+    },
+    {
+      "epoch": 6.763092269326683,
+      "grad_norm": 7.396782875061035,
+      "learning_rate": 1.323990024937656e-05,
+      "loss": 0.378,
+      "step": 27120
+    },
+    {
+      "epoch": 6.765586034912718,
+      "grad_norm": 7.146440505981445,
+      "learning_rate": 1.3237406483790524e-05,
+      "loss": 0.3565,
+      "step": 27130
+    },
+    {
+      "epoch": 6.768079800498753,
+      "grad_norm": 4.529737949371338,
+      "learning_rate": 1.323491271820449e-05,
+      "loss": 0.3357,
+      "step": 27140
+    },
+    {
+      "epoch": 6.770573566084788,
+      "grad_norm": 5.635150909423828,
+      "learning_rate": 1.3232418952618455e-05,
+      "loss": 0.4166,
+      "step": 27150
+    },
+    {
+      "epoch": 6.773067331670823,
+      "grad_norm": 7.09439754486084,
+      "learning_rate": 1.322992518703242e-05,
+      "loss": 0.434,
+      "step": 27160
+    },
+    {
+      "epoch": 6.775561097256858,
+      "grad_norm": 7.409408092498779,
+      "learning_rate": 1.3227431421446386e-05,
+      "loss": 0.3876,
+      "step": 27170
+    },
+    {
+      "epoch": 6.778054862842893,
+      "grad_norm": 7.326381683349609,
+      "learning_rate": 1.3224937655860351e-05,
+      "loss": 0.3933,
+      "step": 27180
+    },
+    {
+      "epoch": 6.780548628428928,
+      "grad_norm": 6.994472980499268,
+      "learning_rate": 1.3222443890274315e-05,
+      "loss": 0.3826,
+      "step": 27190
+    },
+    {
+      "epoch": 6.783042394014963,
+      "grad_norm": 6.760201454162598,
+      "learning_rate": 1.3219950124688282e-05,
+      "loss": 0.4412,
+      "step": 27200
+    },
+    {
+      "epoch": 6.785536159600998,
+      "grad_norm": 5.957584381103516,
+      "learning_rate": 1.3217456359102245e-05,
+      "loss": 0.3916,
+      "step": 27210
+    },
+    {
+      "epoch": 6.788029925187033,
+      "grad_norm": 8.132131576538086,
+      "learning_rate": 1.3214962593516209e-05,
+      "loss": 0.4124,
+      "step": 27220
+    },
+    {
+      "epoch": 6.7905236907730675,
+      "grad_norm": 7.923279762268066,
+      "learning_rate": 1.3212468827930176e-05,
+      "loss": 0.3725,
+      "step": 27230
+    },
+    {
+      "epoch": 6.793017456359102,
+      "grad_norm": 8.802374839782715,
+      "learning_rate": 1.3209975062344141e-05,
+      "loss": 0.332,
+      "step": 27240
+    },
+    {
+      "epoch": 6.795511221945137,
+      "grad_norm": 8.764188766479492,
+      "learning_rate": 1.3207481296758106e-05,
+      "loss": 0.4053,
+      "step": 27250
+    },
+    {
+      "epoch": 6.798004987531172,
+      "grad_norm": 7.860824108123779,
+      "learning_rate": 1.3204987531172072e-05,
+      "loss": 0.4145,
+      "step": 27260
+    },
+    {
+      "epoch": 6.800498753117207,
+      "grad_norm": 5.447219371795654,
+      "learning_rate": 1.3202493765586035e-05,
+      "loss": 0.3885,
+      "step": 27270
+    },
+    {
+      "epoch": 6.802992518703242,
+      "grad_norm": 4.570130348205566,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 0.3414,
+      "step": 27280
+    },
+    {
+      "epoch": 6.805486284289277,
+      "grad_norm": 5.973016262054443,
+      "learning_rate": 1.3197506234413966e-05,
+      "loss": 0.3436,
+      "step": 27290
+    },
+    {
+      "epoch": 6.807980049875312,
+      "grad_norm": 6.451689720153809,
+      "learning_rate": 1.319501246882793e-05,
+      "loss": 0.3842,
+      "step": 27300
+    },
+    {
+      "epoch": 6.8104738154613464,
+      "grad_norm": 6.162374496459961,
+      "learning_rate": 1.3192518703241897e-05,
+      "loss": 0.386,
+      "step": 27310
+    },
+    {
+      "epoch": 6.812967581047381,
+      "grad_norm": 7.707452774047852,
+      "learning_rate": 1.319002493765586e-05,
+      "loss": 0.4001,
+      "step": 27320
+    },
+    {
+      "epoch": 6.815461346633416,
+      "grad_norm": 5.568814754486084,
+      "learning_rate": 1.3187531172069827e-05,
+      "loss": 0.4183,
+      "step": 27330
+    },
+    {
+      "epoch": 6.817955112219451,
+      "grad_norm": 6.596290588378906,
+      "learning_rate": 1.318503740648379e-05,
+      "loss": 0.4037,
+      "step": 27340
+    },
+    {
+      "epoch": 6.820448877805486,
+      "grad_norm": 6.735906600952148,
+      "learning_rate": 1.3182543640897756e-05,
+      "loss": 0.3821,
+      "step": 27350
+    },
+    {
+      "epoch": 6.822942643391521,
+      "grad_norm": 5.793524265289307,
+      "learning_rate": 1.3180049875311723e-05,
+      "loss": 0.3668,
+      "step": 27360
+    },
+    {
+      "epoch": 6.825436408977556,
+      "grad_norm": 5.449307918548584,
+      "learning_rate": 1.3177556109725687e-05,
+      "loss": 0.3029,
+      "step": 27370
+    },
+    {
+      "epoch": 6.8279301745635905,
+      "grad_norm": 7.30305814743042,
+      "learning_rate": 1.3175062344139654e-05,
+      "loss": 0.3419,
+      "step": 27380
+    },
+    {
+      "epoch": 6.830423940149626,
+      "grad_norm": 6.304130554199219,
+      "learning_rate": 1.3172568578553617e-05,
+      "loss": 0.3157,
+      "step": 27390
+    },
+    {
+      "epoch": 6.832917705735661,
+      "grad_norm": 4.693641185760498,
+      "learning_rate": 1.3170074812967581e-05,
+      "loss": 0.3667,
+      "step": 27400
+    },
+    {
+      "epoch": 6.835411471321696,
+      "grad_norm": 13.741803169250488,
+      "learning_rate": 1.3167581047381548e-05,
+      "loss": 0.3824,
+      "step": 27410
+    },
+    {
+      "epoch": 6.837905236907731,
+      "grad_norm": 5.390978813171387,
+      "learning_rate": 1.3165087281795512e-05,
+      "loss": 0.3571,
+      "step": 27420
+    },
+    {
+      "epoch": 6.840399002493766,
+      "grad_norm": 8.878130912780762,
+      "learning_rate": 1.3162593516209477e-05,
+      "loss": 0.3623,
+      "step": 27430
+    },
+    {
+      "epoch": 6.842892768079801,
+      "grad_norm": 9.977058410644531,
+      "learning_rate": 1.3160099750623442e-05,
+      "loss": 0.3624,
+      "step": 27440
+    },
+    {
+      "epoch": 6.8453865336658355,
+      "grad_norm": 4.6882452964782715,
+      "learning_rate": 1.3157605985037408e-05,
+      "loss": 0.3141,
+      "step": 27450
+    },
+    {
+      "epoch": 6.84788029925187,
+      "grad_norm": 14.156923294067383,
+      "learning_rate": 1.3155112219451373e-05,
+      "loss": 0.4383,
+      "step": 27460
+    },
+    {
+      "epoch": 6.850374064837905,
+      "grad_norm": 5.467523097991943,
+      "learning_rate": 1.3152618453865338e-05,
+      "loss": 0.367,
+      "step": 27470
+    },
+    {
+      "epoch": 6.85286783042394,
+      "grad_norm": 5.007584571838379,
+      "learning_rate": 1.3150124688279302e-05,
+      "loss": 0.3777,
+      "step": 27480
+    },
+    {
+      "epoch": 6.855361596009975,
+      "grad_norm": 9.654122352600098,
+      "learning_rate": 1.3147630922693269e-05,
+      "loss": 0.3868,
+      "step": 27490
+    },
+    {
+      "epoch": 6.85785536159601,
+      "grad_norm": 8.467918395996094,
+      "learning_rate": 1.3145137157107232e-05,
+      "loss": 0.3602,
+      "step": 27500
+    },
+    {
+      "epoch": 6.860349127182045,
+      "grad_norm": 8.106925010681152,
+      "learning_rate": 1.3142643391521198e-05,
+      "loss": 0.3881,
+      "step": 27510
+    },
+    {
+      "epoch": 6.86284289276808,
+      "grad_norm": 4.681951999664307,
+      "learning_rate": 1.3140149625935163e-05,
+      "loss": 0.4089,
+      "step": 27520
+    },
+    {
+      "epoch": 6.865336658354114,
+      "grad_norm": 10.276632308959961,
+      "learning_rate": 1.3137655860349128e-05,
+      "loss": 0.4051,
+      "step": 27530
+    },
+    {
+      "epoch": 6.867830423940149,
+      "grad_norm": 6.851044178009033,
+      "learning_rate": 1.3135162094763094e-05,
+      "loss": 0.3344,
+      "step": 27540
+    },
+    {
+      "epoch": 6.870324189526184,
+      "grad_norm": 8.850982666015625,
+      "learning_rate": 1.3132668329177059e-05,
+      "loss": 0.4303,
+      "step": 27550
+    },
+    {
+      "epoch": 6.87281795511222,
+      "grad_norm": 6.095160007476807,
+      "learning_rate": 1.3130174563591023e-05,
+      "loss": 0.4008,
+      "step": 27560
+    },
+    {
+      "epoch": 6.875311720698255,
+      "grad_norm": 4.623822212219238,
+      "learning_rate": 1.312768079800499e-05,
+      "loss": 0.4192,
+      "step": 27570
+    },
+    {
+      "epoch": 6.87780548628429,
+      "grad_norm": 4.292810440063477,
+      "learning_rate": 1.3125187032418953e-05,
+      "loss": 0.3421,
+      "step": 27580
+    },
+    {
+      "epoch": 6.8802992518703245,
+      "grad_norm": 5.254152774810791,
+      "learning_rate": 1.312269326683292e-05,
+      "loss": 0.4478,
+      "step": 27590
+    },
+    {
+      "epoch": 6.882793017456359,
+      "grad_norm": 10.666980743408203,
+      "learning_rate": 1.3120199501246884e-05,
+      "loss": 0.347,
+      "step": 27600
+    },
+    {
+      "epoch": 6.885286783042394,
+      "grad_norm": 5.066367149353027,
+      "learning_rate": 1.3117705735660849e-05,
+      "loss": 0.3701,
+      "step": 27610
+    },
+    {
+      "epoch": 6.887780548628429,
+      "grad_norm": 7.213781356811523,
+      "learning_rate": 1.3115211970074814e-05,
+      "loss": 0.4525,
+      "step": 27620
+    },
+    {
+      "epoch": 6.890274314214464,
+      "grad_norm": 8.0278959274292,
+      "learning_rate": 1.311271820448878e-05,
+      "loss": 0.4072,
+      "step": 27630
+    },
+    {
+      "epoch": 6.892768079800499,
+      "grad_norm": 4.485868453979492,
+      "learning_rate": 1.3110224438902743e-05,
+      "loss": 0.3232,
+      "step": 27640
+    },
+    {
+      "epoch": 6.895261845386534,
+      "grad_norm": 7.174385070800781,
+      "learning_rate": 1.310773067331671e-05,
+      "loss": 0.3655,
+      "step": 27650
+    },
+    {
+      "epoch": 6.897755610972569,
+      "grad_norm": 3.9336626529693604,
+      "learning_rate": 1.3105236907730674e-05,
+      "loss": 0.3454,
+      "step": 27660
+    },
+    {
+      "epoch": 6.9002493765586035,
+      "grad_norm": 6.965961933135986,
+      "learning_rate": 1.3102743142144641e-05,
+      "loss": 0.37,
+      "step": 27670
+    },
+    {
+      "epoch": 6.902743142144638,
+      "grad_norm": 6.3173675537109375,
+      "learning_rate": 1.3100249376558605e-05,
+      "loss": 0.3578,
+      "step": 27680
+    },
+    {
+      "epoch": 6.905236907730673,
+      "grad_norm": 5.358870983123779,
+      "learning_rate": 1.3097755610972568e-05,
+      "loss": 0.399,
+      "step": 27690
+    },
+    {
+      "epoch": 6.907730673316708,
+      "grad_norm": 7.68316125869751,
+      "learning_rate": 1.3095261845386535e-05,
+      "loss": 0.409,
+      "step": 27700
+    },
+    {
+      "epoch": 6.910224438902743,
+      "grad_norm": 7.902895927429199,
+      "learning_rate": 1.30927680798005e-05,
+      "loss": 0.3588,
+      "step": 27710
+    },
+    {
+      "epoch": 6.912718204488778,
+      "grad_norm": 5.715063571929932,
+      "learning_rate": 1.3090274314214464e-05,
+      "loss": 0.3029,
+      "step": 27720
+    },
+    {
+      "epoch": 6.915211970074813,
+      "grad_norm": 6.375357627868652,
+      "learning_rate": 1.3087780548628431e-05,
+      "loss": 0.3668,
+      "step": 27730
+    },
+    {
+      "epoch": 6.917705735660848,
+      "grad_norm": 8.482056617736816,
+      "learning_rate": 1.3085286783042395e-05,
+      "loss": 0.3974,
+      "step": 27740
+    },
+    {
+      "epoch": 6.920199501246882,
+      "grad_norm": 7.30125617980957,
+      "learning_rate": 1.3082793017456362e-05,
+      "loss": 0.368,
+      "step": 27750
+    },
+    {
+      "epoch": 6.922693266832917,
+      "grad_norm": 8.676459312438965,
+      "learning_rate": 1.3080299251870325e-05,
+      "loss": 0.4905,
+      "step": 27760
+    },
+    {
+      "epoch": 6.925187032418952,
+      "grad_norm": 6.748859405517578,
+      "learning_rate": 1.3077805486284289e-05,
+      "loss": 0.4028,
+      "step": 27770
+    },
+    {
+      "epoch": 6.927680798004987,
+      "grad_norm": 5.195030212402344,
+      "learning_rate": 1.3075311720698256e-05,
+      "loss": 0.3717,
+      "step": 27780
+    },
+    {
+      "epoch": 6.930174563591023,
+      "grad_norm": 7.407238960266113,
+      "learning_rate": 1.307281795511222e-05,
+      "loss": 0.3701,
+      "step": 27790
+    },
+    {
+      "epoch": 6.932668329177058,
+      "grad_norm": 9.330326080322266,
+      "learning_rate": 1.3070324189526185e-05,
+      "loss": 0.3282,
+      "step": 27800
+    },
+    {
+      "epoch": 6.9351620947630925,
+      "grad_norm": 6.410877704620361,
+      "learning_rate": 1.306783042394015e-05,
+      "loss": 0.4224,
+      "step": 27810
+    },
+    {
+      "epoch": 6.937655860349127,
+      "grad_norm": 8.255925178527832,
+      "learning_rate": 1.3065336658354116e-05,
+      "loss": 0.4419,
+      "step": 27820
+    },
+    {
+      "epoch": 6.940149625935162,
+      "grad_norm": 5.41157341003418,
+      "learning_rate": 1.3062842892768083e-05,
+      "loss": 0.3181,
+      "step": 27830
+    },
+    {
+      "epoch": 6.942643391521197,
+      "grad_norm": 6.762349605560303,
+      "learning_rate": 1.3060349127182046e-05,
+      "loss": 0.2942,
+      "step": 27840
+    },
+    {
+      "epoch": 6.945137157107232,
+      "grad_norm": 6.1585588455200195,
+      "learning_rate": 1.305785536159601e-05,
+      "loss": 0.3983,
+      "step": 27850
+    },
+    {
+      "epoch": 6.947630922693267,
+      "grad_norm": 6.742948532104492,
+      "learning_rate": 1.3055361596009977e-05,
+      "loss": 0.3911,
+      "step": 27860
+    },
+    {
+      "epoch": 6.950124688279302,
+      "grad_norm": 9.091787338256836,
+      "learning_rate": 1.305286783042394e-05,
+      "loss": 0.3564,
+      "step": 27870
+    },
+    {
+      "epoch": 6.952618453865337,
+      "grad_norm": 8.21529769897461,
+      "learning_rate": 1.3050374064837907e-05,
+      "loss": 0.3723,
+      "step": 27880
+    },
+    {
+      "epoch": 6.9551122194513715,
+      "grad_norm": 5.098814487457275,
+      "learning_rate": 1.3047880299251871e-05,
+      "loss": 0.2938,
+      "step": 27890
+    },
+    {
+      "epoch": 6.957605985037406,
+      "grad_norm": 7.9842000007629395,
+      "learning_rate": 1.3045386533665836e-05,
+      "loss": 0.3766,
+      "step": 27900
+    },
+    {
+      "epoch": 6.960099750623441,
+      "grad_norm": 6.199515342712402,
+      "learning_rate": 1.3042892768079802e-05,
+      "loss": 0.4068,
+      "step": 27910
+    },
+    {
+      "epoch": 6.962593516209476,
+      "grad_norm": 5.990458011627197,
+      "learning_rate": 1.3040399002493767e-05,
+      "loss": 0.3514,
+      "step": 27920
+    },
+    {
+      "epoch": 6.965087281795511,
+      "grad_norm": 6.094311714172363,
+      "learning_rate": 1.303790523690773e-05,
+      "loss": 0.3625,
+      "step": 27930
+    },
+    {
+      "epoch": 6.967581047381546,
+      "grad_norm": 6.9624481201171875,
+      "learning_rate": 1.3035411471321698e-05,
+      "loss": 0.4484,
+      "step": 27940
+    },
+    {
+      "epoch": 6.970074812967581,
+      "grad_norm": 6.599002838134766,
+      "learning_rate": 1.3032917705735661e-05,
+      "loss": 0.4132,
+      "step": 27950
+    },
+    {
+      "epoch": 6.9725685785536164,
+      "grad_norm": 5.155543327331543,
+      "learning_rate": 1.3030423940149628e-05,
+      "loss": 0.3469,
+      "step": 27960
+    },
+    {
+      "epoch": 6.975062344139651,
+      "grad_norm": 6.06123685836792,
+      "learning_rate": 1.3027930174563592e-05,
+      "loss": 0.4548,
+      "step": 27970
+    },
+    {
+      "epoch": 6.977556109725686,
+      "grad_norm": 8.714743614196777,
+      "learning_rate": 1.3025436408977557e-05,
+      "loss": 0.4578,
+      "step": 27980
+    },
+    {
+      "epoch": 6.980049875311721,
+      "grad_norm": 5.861020565032959,
+      "learning_rate": 1.3023192019950125e-05,
+      "loss": 0.337,
+      "step": 27990
+    },
+    {
+      "epoch": 6.982543640897756,
+      "grad_norm": 6.951167106628418,
+      "learning_rate": 1.302069825436409e-05,
+      "loss": 0.3824,
+      "step": 28000
+    },
+    {
+      "epoch": 6.985037406483791,
+      "grad_norm": 3.6698153018951416,
+      "learning_rate": 1.3018204488778056e-05,
+      "loss": 0.3376,
+      "step": 28010
+    },
+    {
+      "epoch": 6.987531172069826,
+      "grad_norm": 3.0680627822875977,
+      "learning_rate": 1.3015710723192021e-05,
+      "loss": 0.3163,
+      "step": 28020
+    },
+    {
+      "epoch": 6.9900249376558605,
+      "grad_norm": 11.675896644592285,
+      "learning_rate": 1.3013216957605985e-05,
+      "loss": 0.3303,
+      "step": 28030
+    },
+    {
+      "epoch": 6.992518703241895,
+      "grad_norm": 4.275930404663086,
+      "learning_rate": 1.3010723192019952e-05,
+      "loss": 0.3455,
+      "step": 28040
+    },
+    {
+      "epoch": 6.99501246882793,
+      "grad_norm": 7.872333526611328,
+      "learning_rate": 1.3008229426433916e-05,
+      "loss": 0.3802,
+      "step": 28050
+    },
+    {
+      "epoch": 6.997506234413965,
+      "grad_norm": 8.095891952514648,
+      "learning_rate": 1.3005735660847883e-05,
+      "loss": 0.4079,
+      "step": 28060
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 8.841901779174805,
+      "learning_rate": 1.3003241895261846e-05,
+      "loss": 0.4457,
+      "step": 28070
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 0.4166116714477539,
+      "eval_runtime": 59.9481,
+      "eval_samples_per_second": 16.731,
+      "eval_steps_per_second": 16.731,
+      "step": 28070
+    },
+    {
+      "epoch": 7.002493765586035,
+      "grad_norm": 10.384051322937012,
+      "learning_rate": 1.3000748129675811e-05,
+      "loss": 0.3682,
+      "step": 28080
+    },
+    {
+      "epoch": 7.00498753117207,
+      "grad_norm": 5.374267578125,
+      "learning_rate": 1.2998254364089777e-05,
+      "loss": 0.4221,
+      "step": 28090
+    },
+    {
+      "epoch": 7.007481296758105,
+      "grad_norm": 11.389906883239746,
+      "learning_rate": 1.2995760598503742e-05,
+      "loss": 0.3828,
+      "step": 28100
+    },
+    {
+      "epoch": 7.0099750623441395,
+      "grad_norm": 7.218733787536621,
+      "learning_rate": 1.2993266832917706e-05,
+      "loss": 0.3267,
+      "step": 28110
+    },
+    {
+      "epoch": 7.012468827930174,
+      "grad_norm": 5.1964263916015625,
+      "learning_rate": 1.2990773067331673e-05,
+      "loss": 0.3566,
+      "step": 28120
+    },
+    {
+      "epoch": 7.014962593516209,
+      "grad_norm": 9.020042419433594,
+      "learning_rate": 1.2988279301745636e-05,
+      "loss": 0.346,
+      "step": 28130
+    },
+    {
+      "epoch": 7.017456359102244,
+      "grad_norm": 13.949241638183594,
+      "learning_rate": 1.2985785536159603e-05,
+      "loss": 0.31,
+      "step": 28140
+    },
+    {
+      "epoch": 7.019950124688279,
+      "grad_norm": 6.440070152282715,
+      "learning_rate": 1.2983291770573567e-05,
+      "loss": 0.3707,
+      "step": 28150
+    },
+    {
+      "epoch": 7.022443890274314,
+      "grad_norm": 6.926358222961426,
+      "learning_rate": 1.2980798004987532e-05,
+      "loss": 0.3829,
+      "step": 28160
+    },
+    {
+      "epoch": 7.024937655860349,
+      "grad_norm": 6.67525053024292,
+      "learning_rate": 1.2978304239401498e-05,
+      "loss": 0.3468,
+      "step": 28170
+    },
+    {
+      "epoch": 7.027431421446384,
+      "grad_norm": 6.060240745544434,
+      "learning_rate": 1.2976059850374066e-05,
+      "loss": 0.3461,
+      "step": 28180
+    },
+    {
+      "epoch": 7.029925187032419,
+      "grad_norm": 9.439708709716797,
+      "learning_rate": 1.2973566084788031e-05,
+      "loss": 0.4009,
+      "step": 28190
+    },
+    {
+      "epoch": 7.032418952618454,
+      "grad_norm": 3.421870708465576,
+      "learning_rate": 1.2971072319201996e-05,
+      "loss": 0.3896,
+      "step": 28200
+    },
+    {
+      "epoch": 7.034912718204489,
+      "grad_norm": 5.836117744445801,
+      "learning_rate": 1.296857855361596e-05,
+      "loss": 0.3461,
+      "step": 28210
+    },
+    {
+      "epoch": 7.037406483790524,
+      "grad_norm": 7.276211738586426,
+      "learning_rate": 1.2966084788029927e-05,
+      "loss": 0.3559,
+      "step": 28220
+    },
+    {
+      "epoch": 7.039900249376559,
+      "grad_norm": 5.882770538330078,
+      "learning_rate": 1.296359102244389e-05,
+      "loss": 0.294,
+      "step": 28230
+    },
+    {
+      "epoch": 7.042394014962594,
+      "grad_norm": 4.949868679046631,
+      "learning_rate": 1.2961097256857858e-05,
+      "loss": 0.3668,
+      "step": 28240
+    },
+    {
+      "epoch": 7.0448877805486285,
+      "grad_norm": 5.131392955780029,
+      "learning_rate": 1.2958603491271821e-05,
+      "loss": 0.3404,
+      "step": 28250
+    },
+    {
+      "epoch": 7.047381546134663,
+      "grad_norm": 6.9679856300354,
+      "learning_rate": 1.2956109725685787e-05,
+      "loss": 0.4193,
+      "step": 28260
+    },
+    {
+      "epoch": 7.049875311720698,
+      "grad_norm": 6.4561767578125,
+      "learning_rate": 1.2953615960099752e-05,
+      "loss": 0.3525,
+      "step": 28270
+    },
+    {
+      "epoch": 7.052369077306733,
+      "grad_norm": 7.776483058929443,
+      "learning_rate": 1.2951122194513717e-05,
+      "loss": 0.3326,
+      "step": 28280
+    },
+    {
+      "epoch": 7.054862842892768,
+      "grad_norm": 5.461517810821533,
+      "learning_rate": 1.2948628428927681e-05,
+      "loss": 0.3554,
+      "step": 28290
+    },
+    {
+      "epoch": 7.057356608478803,
+      "grad_norm": 7.675981521606445,
+      "learning_rate": 1.2946134663341648e-05,
+      "loss": 0.4105,
+      "step": 28300
+    },
+    {
+      "epoch": 7.059850374064838,
+      "grad_norm": 3.51208233833313,
+      "learning_rate": 1.2943640897755611e-05,
+      "loss": 0.3293,
+      "step": 28310
+    },
+    {
+      "epoch": 7.062344139650873,
+      "grad_norm": 5.394303321838379,
+      "learning_rate": 1.2941147132169578e-05,
+      "loss": 0.4108,
+      "step": 28320
+    },
+    {
+      "epoch": 7.0648379052369075,
+      "grad_norm": 5.80884313583374,
+      "learning_rate": 1.2938653366583542e-05,
+      "loss": 0.2895,
+      "step": 28330
+    },
+    {
+      "epoch": 7.067331670822942,
+      "grad_norm": 11.093313217163086,
+      "learning_rate": 1.2936159600997507e-05,
+      "loss": 0.3812,
+      "step": 28340
+    },
+    {
+      "epoch": 7.069825436408977,
+      "grad_norm": 8.297402381896973,
+      "learning_rate": 1.2933665835411473e-05,
+      "loss": 0.4159,
+      "step": 28350
+    },
+    {
+      "epoch": 7.072319201995012,
+      "grad_norm": 8.410477638244629,
+      "learning_rate": 1.2931172069825438e-05,
+      "loss": 0.3703,
+      "step": 28360
+    },
+    {
+      "epoch": 7.074812967581048,
+      "grad_norm": 8.238855361938477,
+      "learning_rate": 1.2928678304239402e-05,
+      "loss": 0.4046,
+      "step": 28370
+    },
+    {
+      "epoch": 7.077306733167083,
+      "grad_norm": 9.850193977355957,
+      "learning_rate": 1.2926184538653369e-05,
+      "loss": 0.4449,
+      "step": 28380
+    },
+    {
+      "epoch": 7.079800498753118,
+      "grad_norm": 4.337007999420166,
+      "learning_rate": 1.2923690773067332e-05,
+      "loss": 0.3567,
+      "step": 28390
+    },
+    {
+      "epoch": 7.082294264339152,
+      "grad_norm": 7.723695278167725,
+      "learning_rate": 1.29211970074813e-05,
+      "loss": 0.3221,
+      "step": 28400
+    },
+    {
+      "epoch": 7.084788029925187,
+      "grad_norm": 3.753911256790161,
+      "learning_rate": 1.2918703241895263e-05,
+      "loss": 0.3524,
+      "step": 28410
+    },
+    {
+      "epoch": 7.087281795511222,
+      "grad_norm": 6.445272922515869,
+      "learning_rate": 1.2916209476309226e-05,
+      "loss": 0.3343,
+      "step": 28420
+    },
+    {
+      "epoch": 7.089775561097257,
+      "grad_norm": 7.809503078460693,
+      "learning_rate": 1.2913715710723193e-05,
+      "loss": 0.3879,
+      "step": 28430
+    },
+    {
+      "epoch": 7.092269326683292,
+      "grad_norm": 8.968701362609863,
+      "learning_rate": 1.2911221945137157e-05,
+      "loss": 0.4012,
+      "step": 28440
+    },
+    {
+      "epoch": 7.094763092269327,
+      "grad_norm": 8.712312698364258,
+      "learning_rate": 1.2908728179551124e-05,
+      "loss": 0.514,
+      "step": 28450
+    },
+    {
+      "epoch": 7.097256857855362,
+      "grad_norm": 7.140172481536865,
+      "learning_rate": 1.290623441396509e-05,
+      "loss": 0.3044,
+      "step": 28460
+    },
+    {
+      "epoch": 7.0997506234413965,
+      "grad_norm": 11.479339599609375,
+      "learning_rate": 1.2903740648379053e-05,
+      "loss": 0.3965,
+      "step": 28470
+    },
+    {
+      "epoch": 7.102244389027431,
+      "grad_norm": 7.283442974090576,
+      "learning_rate": 1.290124688279302e-05,
+      "loss": 0.3227,
+      "step": 28480
+    },
+    {
+      "epoch": 7.104738154613466,
+      "grad_norm": 8.901700019836426,
+      "learning_rate": 1.2898753117206984e-05,
+      "loss": 0.3847,
+      "step": 28490
+    },
+    {
+      "epoch": 7.107231920199501,
+      "grad_norm": 5.670577049255371,
+      "learning_rate": 1.2896259351620947e-05,
+      "loss": 0.4015,
+      "step": 28500
+    },
+    {
+      "epoch": 7.109725685785536,
+      "grad_norm": 5.779824733734131,
+      "learning_rate": 1.2893765586034914e-05,
+      "loss": 0.2875,
+      "step": 28510
+    },
+    {
+      "epoch": 7.112219451371571,
+      "grad_norm": 7.593159198760986,
+      "learning_rate": 1.2891271820448878e-05,
+      "loss": 0.3917,
+      "step": 28520
+    },
+    {
+      "epoch": 7.114713216957606,
+      "grad_norm": 5.073284149169922,
+      "learning_rate": 1.2888778054862845e-05,
+      "loss": 0.2831,
+      "step": 28530
+    },
+    {
+      "epoch": 7.117206982543641,
+      "grad_norm": 7.407405376434326,
+      "learning_rate": 1.2886284289276809e-05,
+      "loss": 0.3676,
+      "step": 28540
+    },
+    {
+      "epoch": 7.1197007481296755,
+      "grad_norm": 6.4434638023376465,
+      "learning_rate": 1.2883790523690774e-05,
+      "loss": 0.363,
+      "step": 28550
+    },
+    {
+      "epoch": 7.12219451371571,
+      "grad_norm": 4.871036052703857,
+      "learning_rate": 1.2881296758104739e-05,
+      "loss": 0.3373,
+      "step": 28560
+    },
+    {
+      "epoch": 7.124688279301745,
+      "grad_norm": 5.883360862731934,
+      "learning_rate": 1.2878802992518704e-05,
+      "loss": 0.3403,
+      "step": 28570
+    },
+    {
+      "epoch": 7.127182044887781,
+      "grad_norm": 7.858293533325195,
+      "learning_rate": 1.2876309226932668e-05,
+      "loss": 0.3905,
+      "step": 28580
+    },
+    {
+      "epoch": 7.129675810473816,
+      "grad_norm": 6.678670883178711,
+      "learning_rate": 1.2873815461346635e-05,
+      "loss": 0.3882,
+      "step": 28590
+    },
+    {
+      "epoch": 7.132169576059851,
+      "grad_norm": 8.354759216308594,
+      "learning_rate": 1.2871321695760599e-05,
+      "loss": 0.389,
+      "step": 28600
+    },
+    {
+      "epoch": 7.134663341645886,
+      "grad_norm": 8.125460624694824,
+      "learning_rate": 1.2868827930174566e-05,
+      "loss": 0.3412,
+      "step": 28610
+    },
+    {
+      "epoch": 7.13715710723192,
+      "grad_norm": 6.524233818054199,
+      "learning_rate": 1.286633416458853e-05,
+      "loss": 0.4105,
+      "step": 28620
+    },
+    {
+      "epoch": 7.139650872817955,
+      "grad_norm": 9.085260391235352,
+      "learning_rate": 1.2863840399002495e-05,
+      "loss": 0.3852,
+      "step": 28630
+    },
+    {
+      "epoch": 7.14214463840399,
+      "grad_norm": 6.906386852264404,
+      "learning_rate": 1.286134663341646e-05,
+      "loss": 0.3602,
+      "step": 28640
+    },
+    {
+      "epoch": 7.144638403990025,
+      "grad_norm": 11.312614440917969,
+      "learning_rate": 1.2858852867830425e-05,
+      "loss": 0.4129,
+      "step": 28650
+    },
+    {
+      "epoch": 7.14713216957606,
+      "grad_norm": 5.9242353439331055,
+      "learning_rate": 1.2856359102244389e-05,
+      "loss": 0.3578,
+      "step": 28660
+    },
+    {
+      "epoch": 7.149625935162095,
+      "grad_norm": 8.224637031555176,
+      "learning_rate": 1.2853865336658356e-05,
+      "loss": 0.3501,
+      "step": 28670
+    },
+    {
+      "epoch": 7.15211970074813,
+      "grad_norm": 5.648616313934326,
+      "learning_rate": 1.285137157107232e-05,
+      "loss": 0.385,
+      "step": 28680
+    },
+    {
+      "epoch": 7.1546134663341645,
+      "grad_norm": 5.693350791931152,
+      "learning_rate": 1.2848877805486286e-05,
+      "loss": 0.4725,
+      "step": 28690
+    },
+    {
+      "epoch": 7.157107231920199,
+      "grad_norm": 5.901285171508789,
+      "learning_rate": 1.284638403990025e-05,
+      "loss": 0.3856,
+      "step": 28700
+    },
+    {
+      "epoch": 7.159600997506234,
+      "grad_norm": 8.282814979553223,
+      "learning_rate": 1.2843890274314215e-05,
+      "loss": 0.3343,
+      "step": 28710
+    },
+    {
+      "epoch": 7.162094763092269,
+      "grad_norm": 6.745294094085693,
+      "learning_rate": 1.284139650872818e-05,
+      "loss": 0.3845,
+      "step": 28720
+    },
+    {
+      "epoch": 7.164588528678304,
+      "grad_norm": 5.898577690124512,
+      "learning_rate": 1.2838902743142146e-05,
+      "loss": 0.3966,
+      "step": 28730
+    },
+    {
+      "epoch": 7.167082294264339,
+      "grad_norm": 5.124606132507324,
+      "learning_rate": 1.2836408977556111e-05,
+      "loss": 0.4228,
+      "step": 28740
+    },
+    {
+      "epoch": 7.169576059850374,
+      "grad_norm": 5.206420421600342,
+      "learning_rate": 1.2833915211970077e-05,
+      "loss": 0.3557,
+      "step": 28750
+    },
+    {
+      "epoch": 7.172069825436409,
+      "grad_norm": 8.561698913574219,
+      "learning_rate": 1.283142144638404e-05,
+      "loss": 0.3579,
+      "step": 28760
+    },
+    {
+      "epoch": 7.174563591022444,
+      "grad_norm": 6.212121486663818,
+      "learning_rate": 1.2828927680798007e-05,
+      "loss": 0.3384,
+      "step": 28770
+    },
+    {
+      "epoch": 7.177057356608479,
+      "grad_norm": 7.147701263427734,
+      "learning_rate": 1.2826433915211971e-05,
+      "loss": 0.3883,
+      "step": 28780
+    },
+    {
+      "epoch": 7.179551122194514,
+      "grad_norm": 6.2454023361206055,
+      "learning_rate": 1.2823940149625934e-05,
+      "loss": 0.4135,
+      "step": 28790
+    },
+    {
+      "epoch": 7.182044887780549,
+      "grad_norm": 9.466367721557617,
+      "learning_rate": 1.2821446384039901e-05,
+      "loss": 0.4502,
+      "step": 28800
+    },
+    {
+      "epoch": 7.184538653366584,
+      "grad_norm": 8.00047779083252,
+      "learning_rate": 1.2818952618453867e-05,
+      "loss": 0.4377,
+      "step": 28810
+    },
+    {
+      "epoch": 7.187032418952619,
+      "grad_norm": 7.6417951583862305,
+      "learning_rate": 1.2816458852867832e-05,
+      "loss": 0.363,
+      "step": 28820
+    },
+    {
+      "epoch": 7.1895261845386536,
+      "grad_norm": 4.634132385253906,
+      "learning_rate": 1.2813965087281797e-05,
+      "loss": 0.3622,
+      "step": 28830
+    },
+    {
+      "epoch": 7.192019950124688,
+      "grad_norm": 7.929607391357422,
+      "learning_rate": 1.2811471321695761e-05,
+      "loss": 0.3553,
+      "step": 28840
+    },
+    {
+      "epoch": 7.194513715710723,
+      "grad_norm": 6.655643463134766,
+      "learning_rate": 1.2808977556109728e-05,
+      "loss": 0.3342,
+      "step": 28850
+    },
+    {
+      "epoch": 7.197007481296758,
+      "grad_norm": 7.349606037139893,
+      "learning_rate": 1.2806483790523692e-05,
+      "loss": 0.3685,
+      "step": 28860
+    },
+    {
+      "epoch": 7.199501246882793,
+      "grad_norm": 8.308873176574707,
+      "learning_rate": 1.2803990024937655e-05,
+      "loss": 0.3837,
+      "step": 28870
+    },
+    {
+      "epoch": 7.201995012468828,
+      "grad_norm": 9.069894790649414,
+      "learning_rate": 1.2801496259351622e-05,
+      "loss": 0.3609,
+      "step": 28880
+    },
+    {
+      "epoch": 7.204488778054863,
+      "grad_norm": 10.108824729919434,
+      "learning_rate": 1.2799002493765586e-05,
+      "loss": 0.4051,
+      "step": 28890
+    },
+    {
+      "epoch": 7.206982543640898,
+      "grad_norm": 9.224723815917969,
+      "learning_rate": 1.2796508728179553e-05,
+      "loss": 0.4147,
+      "step": 28900
+    },
+    {
+      "epoch": 7.2094763092269325,
+      "grad_norm": 5.563706398010254,
+      "learning_rate": 1.2794014962593517e-05,
+      "loss": 0.3391,
+      "step": 28910
+    },
+    {
+      "epoch": 7.211970074812967,
+      "grad_norm": 6.96822452545166,
+      "learning_rate": 1.2791521197007482e-05,
+      "loss": 0.3146,
+      "step": 28920
+    },
+    {
+      "epoch": 7.214463840399002,
+      "grad_norm": 8.976058006286621,
+      "learning_rate": 1.2789027431421449e-05,
+      "loss": 0.3729,
+      "step": 28930
+    },
+    {
+      "epoch": 7.216957605985037,
+      "grad_norm": 6.567900657653809,
+      "learning_rate": 1.2786533665835412e-05,
+      "loss": 0.3472,
+      "step": 28940
+    },
+    {
+      "epoch": 7.219451371571072,
+      "grad_norm": 6.142775535583496,
+      "learning_rate": 1.278403990024938e-05,
+      "loss": 0.3296,
+      "step": 28950
+    },
+    {
+      "epoch": 7.221945137157107,
+      "grad_norm": 6.999728202819824,
+      "learning_rate": 1.2781546134663343e-05,
+      "loss": 0.3493,
+      "step": 28960
+    },
+    {
+      "epoch": 7.224438902743142,
+      "grad_norm": 6.844429969787598,
+      "learning_rate": 1.2779052369077307e-05,
+      "loss": 0.3389,
+      "step": 28970
+    },
+    {
+      "epoch": 7.2269326683291775,
+      "grad_norm": 8.233739852905273,
+      "learning_rate": 1.2776558603491274e-05,
+      "loss": 0.3728,
+      "step": 28980
+    },
+    {
+      "epoch": 7.229426433915212,
+      "grad_norm": 6.4842848777771,
+      "learning_rate": 1.2774064837905237e-05,
+      "loss": 0.4287,
+      "step": 28990
+    },
+    {
+      "epoch": 7.231920199501247,
+      "grad_norm": 7.557778835296631,
+      "learning_rate": 1.2771571072319203e-05,
+      "loss": 0.3734,
+      "step": 29000
+    },
+    {
+      "epoch": 7.234413965087282,
+      "grad_norm": 6.296581268310547,
+      "learning_rate": 1.2769077306733168e-05,
+      "loss": 0.3677,
+      "step": 29010
+    },
+    {
+      "epoch": 7.236907730673317,
+      "grad_norm": 7.28203821182251,
+      "learning_rate": 1.2766583541147133e-05,
+      "loss": 0.3798,
+      "step": 29020
+    },
+    {
+      "epoch": 7.239401496259352,
+      "grad_norm": 5.902282238006592,
+      "learning_rate": 1.2764089775561099e-05,
+      "loss": 0.3399,
+      "step": 29030
+    },
+    {
+      "epoch": 7.241895261845387,
+      "grad_norm": 6.5836310386657715,
+      "learning_rate": 1.2761596009975064e-05,
+      "loss": 0.406,
+      "step": 29040
+    },
+    {
+      "epoch": 7.2443890274314215,
+      "grad_norm": 8.319708824157715,
+      "learning_rate": 1.2759102244389027e-05,
+      "loss": 0.3888,
+      "step": 29050
+    },
+    {
+      "epoch": 7.246882793017456,
+      "grad_norm": 8.096579551696777,
+      "learning_rate": 1.2756608478802994e-05,
+      "loss": 0.3393,
+      "step": 29060
+    },
+    {
+      "epoch": 7.249376558603491,
+      "grad_norm": 6.496395587921143,
+      "learning_rate": 1.2754114713216958e-05,
+      "loss": 0.3642,
+      "step": 29070
+    },
+    {
+      "epoch": 7.251870324189526,
+      "grad_norm": 7.759038925170898,
+      "learning_rate": 1.2751620947630923e-05,
+      "loss": 0.3334,
+      "step": 29080
+    },
+    {
+      "epoch": 7.254364089775561,
+      "grad_norm": 10.256757736206055,
+      "learning_rate": 1.2749127182044889e-05,
+      "loss": 0.3924,
+      "step": 29090
+    },
+    {
+      "epoch": 7.256857855361596,
+      "grad_norm": 10.726069450378418,
+      "learning_rate": 1.2746633416458854e-05,
+      "loss": 0.4015,
+      "step": 29100
+    },
+    {
+      "epoch": 7.259351620947631,
+      "grad_norm": 8.847235679626465,
+      "learning_rate": 1.274413965087282e-05,
+      "loss": 0.4174,
+      "step": 29110
+    },
+    {
+      "epoch": 7.261845386533666,
+      "grad_norm": 7.851325035095215,
+      "learning_rate": 1.2741645885286785e-05,
+      "loss": 0.3122,
+      "step": 29120
+    },
+    {
+      "epoch": 7.2643391521197005,
+      "grad_norm": 6.145222187042236,
+      "learning_rate": 1.2739152119700748e-05,
+      "loss": 0.3972,
+      "step": 29130
+    },
+    {
+      "epoch": 7.266832917705735,
+      "grad_norm": 7.113053798675537,
+      "learning_rate": 1.2736658354114715e-05,
+      "loss": 0.4018,
+      "step": 29140
+    },
+    {
+      "epoch": 7.26932668329177,
+      "grad_norm": 10.120803833007812,
+      "learning_rate": 1.2734164588528679e-05,
+      "loss": 0.3977,
+      "step": 29150
+    },
+    {
+      "epoch": 7.271820448877805,
+      "grad_norm": 5.162845134735107,
+      "learning_rate": 1.2731670822942644e-05,
+      "loss": 0.3468,
+      "step": 29160
+    },
+    {
+      "epoch": 7.274314214463841,
+      "grad_norm": 9.55453109741211,
+      "learning_rate": 1.272917705735661e-05,
+      "loss": 0.3615,
+      "step": 29170
+    },
+    {
+      "epoch": 7.276807980049876,
+      "grad_norm": 7.122071266174316,
+      "learning_rate": 1.2726683291770575e-05,
+      "loss": 0.3974,
+      "step": 29180
+    },
+    {
+      "epoch": 7.279301745635911,
+      "grad_norm": 6.449893474578857,
+      "learning_rate": 1.272418952618454e-05,
+      "loss": 0.3763,
+      "step": 29190
+    },
+    {
+      "epoch": 7.2817955112219455,
+      "grad_norm": 12.209590911865234,
+      "learning_rate": 1.2721695760598505e-05,
+      "loss": 0.32,
+      "step": 29200
+    },
+    {
+      "epoch": 7.28428927680798,
+      "grad_norm": 7.169990062713623,
+      "learning_rate": 1.2719201995012469e-05,
+      "loss": 0.3413,
+      "step": 29210
+    },
+    {
+      "epoch": 7.286783042394015,
+      "grad_norm": 8.662860870361328,
+      "learning_rate": 1.2716708229426436e-05,
+      "loss": 0.3932,
+      "step": 29220
+    },
+    {
+      "epoch": 7.28927680798005,
+      "grad_norm": 6.312235355377197,
+      "learning_rate": 1.27142144638404e-05,
+      "loss": 0.3372,
+      "step": 29230
+    },
+    {
+      "epoch": 7.291770573566085,
+      "grad_norm": 5.396263599395752,
+      "learning_rate": 1.2711720698254367e-05,
+      "loss": 0.2894,
+      "step": 29240
+    },
+    {
+      "epoch": 7.29426433915212,
+      "grad_norm": 7.881542205810547,
+      "learning_rate": 1.270922693266833e-05,
+      "loss": 0.4439,
+      "step": 29250
+    },
+    {
+      "epoch": 7.296758104738155,
+      "grad_norm": 6.539719104766846,
+      "learning_rate": 1.2706733167082294e-05,
+      "loss": 0.3688,
+      "step": 29260
+    },
+    {
+      "epoch": 7.2992518703241895,
+      "grad_norm": 9.538723945617676,
+      "learning_rate": 1.2704239401496261e-05,
+      "loss": 0.3785,
+      "step": 29270
+    },
+    {
+      "epoch": 7.301745635910224,
+      "grad_norm": 8.049639701843262,
+      "learning_rate": 1.2701745635910226e-05,
+      "loss": 0.3698,
+      "step": 29280
+    },
+    {
+      "epoch": 7.304239401496259,
+      "grad_norm": 11.048667907714844,
+      "learning_rate": 1.269925187032419e-05,
+      "loss": 0.477,
+      "step": 29290
+    },
+    {
+      "epoch": 7.306733167082294,
+      "grad_norm": 4.723964214324951,
+      "learning_rate": 1.2696758104738157e-05,
+      "loss": 0.278,
+      "step": 29300
+    },
+    {
+      "epoch": 7.309226932668329,
+      "grad_norm": 8.363397598266602,
+      "learning_rate": 1.269426433915212e-05,
+      "loss": 0.3963,
+      "step": 29310
+    },
+    {
+      "epoch": 7.311720698254364,
+      "grad_norm": 5.296679973602295,
+      "learning_rate": 1.2691770573566087e-05,
+      "loss": 0.3419,
+      "step": 29320
+    },
+    {
+      "epoch": 7.314214463840399,
+      "grad_norm": 4.19307804107666,
+      "learning_rate": 1.2689276807980051e-05,
+      "loss": 0.3923,
+      "step": 29330
+    },
+    {
+      "epoch": 7.316708229426434,
+      "grad_norm": 4.72712516784668,
+      "learning_rate": 1.2686783042394015e-05,
+      "loss": 0.3854,
+      "step": 29340
+    },
+    {
+      "epoch": 7.3192019950124685,
+      "grad_norm": 7.943241119384766,
+      "learning_rate": 1.2684289276807982e-05,
+      "loss": 0.4142,
+      "step": 29350
+    },
+    {
+      "epoch": 7.321695760598503,
+      "grad_norm": 7.628240585327148,
+      "learning_rate": 1.2681795511221945e-05,
+      "loss": 0.3044,
+      "step": 29360
+    },
+    {
+      "epoch": 7.324189526184538,
+      "grad_norm": 6.917629718780518,
+      "learning_rate": 1.267930174563591e-05,
+      "loss": 0.3286,
+      "step": 29370
+    },
+    {
+      "epoch": 7.326683291770574,
+      "grad_norm": 5.416843891143799,
+      "learning_rate": 1.2676807980049876e-05,
+      "loss": 0.442,
+      "step": 29380
+    },
+    {
+      "epoch": 7.329177057356609,
+      "grad_norm": 4.867097854614258,
+      "learning_rate": 1.2674314214463841e-05,
+      "loss": 0.3525,
+      "step": 29390
+    },
+    {
+      "epoch": 7.331670822942644,
+      "grad_norm": 8.583399772644043,
+      "learning_rate": 1.2671820448877808e-05,
+      "loss": 0.3817,
+      "step": 29400
+    },
+    {
+      "epoch": 7.334164588528679,
+      "grad_norm": 7.867173194885254,
+      "learning_rate": 1.2669326683291772e-05,
+      "loss": 0.3231,
+      "step": 29410
+    },
+    {
+      "epoch": 7.3366583541147135,
+      "grad_norm": 6.9553022384643555,
+      "learning_rate": 1.2666832917705735e-05,
+      "loss": 0.321,
+      "step": 29420
+    },
+    {
+      "epoch": 7.339152119700748,
+      "grad_norm": 6.242718696594238,
+      "learning_rate": 1.2664339152119702e-05,
+      "loss": 0.365,
+      "step": 29430
+    },
+    {
+      "epoch": 7.341645885286783,
+      "grad_norm": 8.64468765258789,
+      "learning_rate": 1.2661845386533666e-05,
+      "loss": 0.3114,
+      "step": 29440
+    },
+    {
+      "epoch": 7.344139650872818,
+      "grad_norm": 6.854546546936035,
+      "learning_rate": 1.2659351620947633e-05,
+      "loss": 0.3563,
+      "step": 29450
+    },
+    {
+      "epoch": 7.346633416458853,
+      "grad_norm": 10.145390510559082,
+      "learning_rate": 1.2656857855361597e-05,
+      "loss": 0.3541,
+      "step": 29460
+    },
+    {
+      "epoch": 7.349127182044888,
+      "grad_norm": 5.386590480804443,
+      "learning_rate": 1.2654364089775562e-05,
+      "loss": 0.3673,
+      "step": 29470
+    },
+    {
+      "epoch": 7.351620947630923,
+      "grad_norm": 5.273983955383301,
+      "learning_rate": 1.2651870324189527e-05,
+      "loss": 0.2825,
+      "step": 29480
+    },
+    {
+      "epoch": 7.3541147132169575,
+      "grad_norm": 4.6638689041137695,
+      "learning_rate": 1.2649376558603493e-05,
+      "loss": 0.4178,
+      "step": 29490
+    },
+    {
+      "epoch": 7.356608478802992,
+      "grad_norm": 5.493322372436523,
+      "learning_rate": 1.2646882793017456e-05,
+      "loss": 0.3265,
+      "step": 29500
+    },
+    {
+      "epoch": 7.359102244389027,
+      "grad_norm": 6.716048240661621,
+      "learning_rate": 1.2644389027431423e-05,
+      "loss": 0.3692,
+      "step": 29510
+    },
+    {
+      "epoch": 7.361596009975062,
+      "grad_norm": 6.258030891418457,
+      "learning_rate": 1.2641895261845387e-05,
+      "loss": 0.426,
+      "step": 29520
+    },
+    {
+      "epoch": 7.364089775561097,
+      "grad_norm": 5.933165550231934,
+      "learning_rate": 1.2639401496259354e-05,
+      "loss": 0.3618,
+      "step": 29530
+    },
+    {
+      "epoch": 7.366583541147132,
+      "grad_norm": 5.566700458526611,
+      "learning_rate": 1.2636907730673317e-05,
+      "loss": 0.4001,
+      "step": 29540
+    },
+    {
+      "epoch": 7.369077306733167,
+      "grad_norm": 7.390594959259033,
+      "learning_rate": 1.2634413965087283e-05,
+      "loss": 0.3826,
+      "step": 29550
+    },
+    {
+      "epoch": 7.371571072319202,
+      "grad_norm": 4.834929466247559,
+      "learning_rate": 1.2631920199501248e-05,
+      "loss": 0.3512,
+      "step": 29560
+    },
+    {
+      "epoch": 7.374064837905237,
+      "grad_norm": 8.851217269897461,
+      "learning_rate": 1.2629426433915213e-05,
+      "loss": 0.387,
+      "step": 29570
+    },
+    {
+      "epoch": 7.376558603491272,
+      "grad_norm": 14.366303443908691,
+      "learning_rate": 1.2626932668329177e-05,
+      "loss": 0.3875,
+      "step": 29580
+    },
+    {
+      "epoch": 7.379052369077307,
+      "grad_norm": 7.262747287750244,
+      "learning_rate": 1.2624438902743144e-05,
+      "loss": 0.3718,
+      "step": 29590
+    },
+    {
+      "epoch": 7.381546134663342,
+      "grad_norm": 5.579659938812256,
+      "learning_rate": 1.2621945137157108e-05,
+      "loss": 0.3919,
+      "step": 29600
+    },
+    {
+      "epoch": 7.384039900249377,
+      "grad_norm": 5.182689189910889,
+      "learning_rate": 1.2619451371571075e-05,
+      "loss": 0.3414,
+      "step": 29610
+    },
+    {
+      "epoch": 7.386533665835412,
+      "grad_norm": 4.612008094787598,
+      "learning_rate": 1.2616957605985038e-05,
+      "loss": 0.3101,
+      "step": 29620
+    },
+    {
+      "epoch": 7.389027431421447,
+      "grad_norm": 11.192936897277832,
+      "learning_rate": 1.2614463840399004e-05,
+      "loss": 0.3038,
+      "step": 29630
+    },
+    {
+      "epoch": 7.3915211970074814,
+      "grad_norm": 8.110973358154297,
+      "learning_rate": 1.2611970074812969e-05,
+      "loss": 0.451,
+      "step": 29640
+    },
+    {
+      "epoch": 7.394014962593516,
+      "grad_norm": 9.395048141479492,
+      "learning_rate": 1.2609476309226934e-05,
+      "loss": 0.3936,
+      "step": 29650
+    },
+    {
+      "epoch": 7.396508728179551,
+      "grad_norm": 6.499456882476807,
+      "learning_rate": 1.2606982543640898e-05,
+      "loss": 0.3217,
+      "step": 29660
+    },
+    {
+      "epoch": 7.399002493765586,
+      "grad_norm": 6.436877250671387,
+      "learning_rate": 1.2604488778054865e-05,
+      "loss": 0.4203,
+      "step": 29670
+    },
+    {
+      "epoch": 7.401496259351621,
+      "grad_norm": 4.4001688957214355,
+      "learning_rate": 1.2601995012468828e-05,
+      "loss": 0.4034,
+      "step": 29680
+    },
+    {
+      "epoch": 7.403990024937656,
+      "grad_norm": 7.0502800941467285,
+      "learning_rate": 1.2599501246882795e-05,
+      "loss": 0.4124,
+      "step": 29690
+    },
+    {
+      "epoch": 7.406483790523691,
+      "grad_norm": 10.673737525939941,
+      "learning_rate": 1.2597007481296759e-05,
+      "loss": 0.4948,
+      "step": 29700
+    },
+    {
+      "epoch": 7.4089775561097255,
+      "grad_norm": 6.177337646484375,
+      "learning_rate": 1.2594513715710723e-05,
+      "loss": 0.4471,
+      "step": 29710
+    },
+    {
+      "epoch": 7.41147132169576,
+      "grad_norm": 5.368765830993652,
+      "learning_rate": 1.259201995012469e-05,
+      "loss": 0.3887,
+      "step": 29720
+    },
+    {
+      "epoch": 7.413965087281795,
+      "grad_norm": 6.879129409790039,
+      "learning_rate": 1.2589526184538653e-05,
+      "loss": 0.3773,
+      "step": 29730
+    },
+    {
+      "epoch": 7.41645885286783,
+      "grad_norm": 8.17301082611084,
+      "learning_rate": 1.258703241895262e-05,
+      "loss": 0.3722,
+      "step": 29740
+    },
+    {
+      "epoch": 7.418952618453865,
+      "grad_norm": 8.106537818908691,
+      "learning_rate": 1.2584538653366586e-05,
+      "loss": 0.41,
+      "step": 29750
+    },
+    {
+      "epoch": 7.4214463840399,
+      "grad_norm": 6.589071273803711,
+      "learning_rate": 1.258204488778055e-05,
+      "loss": 0.4002,
+      "step": 29760
+    },
+    {
+      "epoch": 7.423940149625935,
+      "grad_norm": 8.137639999389648,
+      "learning_rate": 1.2579551122194516e-05,
+      "loss": 0.3859,
+      "step": 29770
+    },
+    {
+      "epoch": 7.42643391521197,
+      "grad_norm": 6.957583427429199,
+      "learning_rate": 1.257705735660848e-05,
+      "loss": 0.4258,
+      "step": 29780
+    },
+    {
+      "epoch": 7.428927680798005,
+      "grad_norm": 8.225882530212402,
+      "learning_rate": 1.2574563591022443e-05,
+      "loss": 0.3842,
+      "step": 29790
+    },
+    {
+      "epoch": 7.43142144638404,
+      "grad_norm": 6.6841888427734375,
+      "learning_rate": 1.257206982543641e-05,
+      "loss": 0.3717,
+      "step": 29800
+    },
+    {
+      "epoch": 7.433915211970075,
+      "grad_norm": 6.028921127319336,
+      "learning_rate": 1.2569576059850374e-05,
+      "loss": 0.3616,
+      "step": 29810
+    },
+    {
+      "epoch": 7.43640897755611,
+      "grad_norm": 5.823741436004639,
+      "learning_rate": 1.2567082294264341e-05,
+      "loss": 0.3755,
+      "step": 29820
+    },
+    {
+      "epoch": 7.438902743142145,
+      "grad_norm": 11.203161239624023,
+      "learning_rate": 1.2564588528678305e-05,
+      "loss": 0.3806,
+      "step": 29830
+    },
+    {
+      "epoch": 7.44139650872818,
+      "grad_norm": 5.957489967346191,
+      "learning_rate": 1.256209476309227e-05,
+      "loss": 0.3821,
+      "step": 29840
+    },
+    {
+      "epoch": 7.443890274314215,
+      "grad_norm": 4.782578468322754,
+      "learning_rate": 1.2559600997506235e-05,
+      "loss": 0.3689,
+      "step": 29850
+    },
+    {
+      "epoch": 7.446384039900249,
+      "grad_norm": 4.5286784172058105,
+      "learning_rate": 1.25571072319202e-05,
+      "loss": 0.3414,
+      "step": 29860
+    },
+    {
+      "epoch": 7.448877805486284,
+      "grad_norm": 11.462789535522461,
+      "learning_rate": 1.2554613466334164e-05,
+      "loss": 0.373,
+      "step": 29870
+    },
+    {
+      "epoch": 7.451371571072319,
+      "grad_norm": 8.337203979492188,
+      "learning_rate": 1.2552119700748131e-05,
+      "loss": 0.3389,
+      "step": 29880
+    },
+    {
+      "epoch": 7.453865336658354,
+      "grad_norm": 13.295976638793945,
+      "learning_rate": 1.2549625935162095e-05,
+      "loss": 0.3165,
+      "step": 29890
+    },
+    {
+      "epoch": 7.456359102244389,
+      "grad_norm": 6.52192497253418,
+      "learning_rate": 1.2547132169576062e-05,
+      "loss": 0.3477,
+      "step": 29900
+    },
+    {
+      "epoch": 7.458852867830424,
+      "grad_norm": 7.2272491455078125,
+      "learning_rate": 1.2544638403990025e-05,
+      "loss": 0.3611,
+      "step": 29910
+    },
+    {
+      "epoch": 7.461346633416459,
+      "grad_norm": 3.4752140045166016,
+      "learning_rate": 1.254214463840399e-05,
+      "loss": 0.3382,
+      "step": 29920
+    },
+    {
+      "epoch": 7.4638403990024935,
+      "grad_norm": 9.237256050109863,
+      "learning_rate": 1.2539650872817956e-05,
+      "loss": 0.3683,
+      "step": 29930
+    },
+    {
+      "epoch": 7.466334164588528,
+      "grad_norm": 9.527120590209961,
+      "learning_rate": 1.2537157107231921e-05,
+      "loss": 0.3784,
+      "step": 29940
+    },
+    {
+      "epoch": 7.468827930174563,
+      "grad_norm": 5.257556915283203,
+      "learning_rate": 1.2534663341645887e-05,
+      "loss": 0.3919,
+      "step": 29950
+    },
+    {
+      "epoch": 7.471321695760598,
+      "grad_norm": 5.737821102142334,
+      "learning_rate": 1.2532169576059852e-05,
+      "loss": 0.3766,
+      "step": 29960
+    },
+    {
+      "epoch": 7.473815461346634,
+      "grad_norm": 8.01321792602539,
+      "learning_rate": 1.2529675810473816e-05,
+      "loss": 0.3299,
+      "step": 29970
+    },
+    {
+      "epoch": 7.476309226932669,
+      "grad_norm": 5.301641464233398,
+      "learning_rate": 1.2527182044887783e-05,
+      "loss": 0.3851,
+      "step": 29980
+    },
+    {
+      "epoch": 7.478802992518704,
+      "grad_norm": 7.687638759613037,
+      "learning_rate": 1.2524688279301746e-05,
+      "loss": 0.3364,
+      "step": 29990
+    },
+    {
+      "epoch": 7.4812967581047385,
+      "grad_norm": 6.31342077255249,
+      "learning_rate": 1.2522194513715712e-05,
+      "loss": 0.3607,
+      "step": 30000
+    },
+    {
+      "epoch": 7.483790523690773,
+      "grad_norm": 5.950140953063965,
+      "learning_rate": 1.2519700748129677e-05,
+      "loss": 0.3616,
+      "step": 30010
+    },
+    {
+      "epoch": 7.486284289276808,
+      "grad_norm": 6.606958389282227,
+      "learning_rate": 1.2517206982543642e-05,
+      "loss": 0.3231,
+      "step": 30020
+    },
+    {
+      "epoch": 7.488778054862843,
+      "grad_norm": 8.758807182312012,
+      "learning_rate": 1.2514713216957607e-05,
+      "loss": 0.3508,
+      "step": 30030
+    },
+    {
+      "epoch": 7.491271820448878,
+      "grad_norm": 3.6969501972198486,
+      "learning_rate": 1.2512219451371573e-05,
+      "loss": 0.3668,
+      "step": 30040
+    },
+    {
+      "epoch": 7.493765586034913,
+      "grad_norm": 7.893050670623779,
+      "learning_rate": 1.2509725685785536e-05,
+      "loss": 0.3001,
+      "step": 30050
+    },
+    {
+      "epoch": 7.496259351620948,
+      "grad_norm": 8.5418119430542,
+      "learning_rate": 1.2507231920199503e-05,
+      "loss": 0.3428,
+      "step": 30060
+    },
+    {
+      "epoch": 7.498753117206983,
+      "grad_norm": 7.37095308303833,
+      "learning_rate": 1.2504738154613467e-05,
+      "loss": 0.36,
+      "step": 30070
+    },
+    {
+      "epoch": 7.501246882793017,
+      "grad_norm": 5.3542704582214355,
+      "learning_rate": 1.250224438902743e-05,
+      "loss": 0.354,
+      "step": 30080
+    },
+    {
+      "epoch": 7.503740648379052,
+      "grad_norm": 7.024536609649658,
+      "learning_rate": 1.2499750623441398e-05,
+      "loss": 0.3936,
+      "step": 30090
+    },
+    {
+      "epoch": 7.506234413965087,
+      "grad_norm": 8.495894432067871,
+      "learning_rate": 1.2497256857855363e-05,
+      "loss": 0.3381,
+      "step": 30100
+    },
+    {
+      "epoch": 7.508728179551122,
+      "grad_norm": 6.884313106536865,
+      "learning_rate": 1.2494763092269328e-05,
+      "loss": 0.381,
+      "step": 30110
+    },
+    {
+      "epoch": 7.511221945137157,
+      "grad_norm": 7.271829128265381,
+      "learning_rate": 1.2492269326683294e-05,
+      "loss": 0.4289,
+      "step": 30120
+    },
+    {
+      "epoch": 7.513715710723192,
+      "grad_norm": 14.023894309997559,
+      "learning_rate": 1.2489775561097257e-05,
+      "loss": 0.4139,
+      "step": 30130
+    },
+    {
+      "epoch": 7.516209476309227,
+      "grad_norm": 7.130033493041992,
+      "learning_rate": 1.2487281795511224e-05,
+      "loss": 0.3319,
+      "step": 30140
+    },
+    {
+      "epoch": 7.5187032418952615,
+      "grad_norm": 6.766006946563721,
+      "learning_rate": 1.2484788029925188e-05,
+      "loss": 0.3546,
+      "step": 30150
+    },
+    {
+      "epoch": 7.521197007481296,
+      "grad_norm": 10.241814613342285,
+      "learning_rate": 1.2482294264339151e-05,
+      "loss": 0.3268,
+      "step": 30160
+    },
+    {
+      "epoch": 7.523690773067331,
+      "grad_norm": 8.794346809387207,
+      "learning_rate": 1.2479800498753118e-05,
+      "loss": 0.3394,
+      "step": 30170
+    },
+    {
+      "epoch": 7.526184538653366,
+      "grad_norm": 8.435714721679688,
+      "learning_rate": 1.2477306733167082e-05,
+      "loss": 0.3174,
+      "step": 30180
+    },
+    {
+      "epoch": 7.528678304239402,
+      "grad_norm": 9.198299407958984,
+      "learning_rate": 1.2474812967581049e-05,
+      "loss": 0.36,
+      "step": 30190
+    },
+    {
+      "epoch": 7.531172069825437,
+      "grad_norm": 8.56997013092041,
+      "learning_rate": 1.2472319201995013e-05,
+      "loss": 0.3682,
+      "step": 30200
+    },
+    {
+      "epoch": 7.533665835411472,
+      "grad_norm": 9.016803741455078,
+      "learning_rate": 1.2469825436408978e-05,
+      "loss": 0.3727,
+      "step": 30210
+    },
+    {
+      "epoch": 7.5361596009975065,
+      "grad_norm": 7.877820014953613,
+      "learning_rate": 1.2467331670822945e-05,
+      "loss": 0.4129,
+      "step": 30220
+    },
+    {
+      "epoch": 7.538653366583541,
+      "grad_norm": 5.137962818145752,
+      "learning_rate": 1.2464837905236909e-05,
+      "loss": 0.3079,
+      "step": 30230
+    },
+    {
+      "epoch": 7.541147132169576,
+      "grad_norm": 6.477614879608154,
+      "learning_rate": 1.2462344139650876e-05,
+      "loss": 0.3845,
+      "step": 30240
+    },
+    {
+      "epoch": 7.543640897755611,
+      "grad_norm": 7.037804126739502,
+      "learning_rate": 1.245985037406484e-05,
+      "loss": 0.4319,
+      "step": 30250
+    },
+    {
+      "epoch": 7.546134663341646,
+      "grad_norm": 6.353123664855957,
+      "learning_rate": 1.2457356608478803e-05,
+      "loss": 0.3744,
+      "step": 30260
+    },
+    {
+      "epoch": 7.548628428927681,
+      "grad_norm": 18.240354537963867,
+      "learning_rate": 1.245486284289277e-05,
+      "loss": 0.3774,
+      "step": 30270
+    },
+    {
+      "epoch": 7.551122194513716,
+      "grad_norm": 9.531089782714844,
+      "learning_rate": 1.2452369077306733e-05,
+      "loss": 0.3673,
+      "step": 30280
+    },
+    {
+      "epoch": 7.553615960099751,
+      "grad_norm": 6.397697925567627,
+      "learning_rate": 1.2449875311720699e-05,
+      "loss": 0.4171,
+      "step": 30290
+    },
+    {
+      "epoch": 7.556109725685785,
+      "grad_norm": 5.065841197967529,
+      "learning_rate": 1.2447381546134664e-05,
+      "loss": 0.3296,
+      "step": 30300
+    },
+    {
+      "epoch": 7.55860349127182,
+      "grad_norm": 6.741978645324707,
+      "learning_rate": 1.244488778054863e-05,
+      "loss": 0.3183,
+      "step": 30310
+    },
+    {
+      "epoch": 7.561097256857855,
+      "grad_norm": 6.0341620445251465,
+      "learning_rate": 1.2442394014962595e-05,
+      "loss": 0.399,
+      "step": 30320
+    },
+    {
+      "epoch": 7.56359102244389,
+      "grad_norm": 5.0857977867126465,
+      "learning_rate": 1.243990024937656e-05,
+      "loss": 0.3163,
+      "step": 30330
+    },
+    {
+      "epoch": 7.566084788029925,
+      "grad_norm": 7.0515456199646,
+      "learning_rate": 1.2437406483790524e-05,
+      "loss": 0.3199,
+      "step": 30340
+    },
+    {
+      "epoch": 7.56857855361596,
+      "grad_norm": 8.328603744506836,
+      "learning_rate": 1.243491271820449e-05,
+      "loss": 0.3836,
+      "step": 30350
+    },
+    {
+      "epoch": 7.571072319201995,
+      "grad_norm": 7.779388427734375,
+      "learning_rate": 1.2432418952618454e-05,
+      "loss": 0.3254,
+      "step": 30360
+    },
+    {
+      "epoch": 7.57356608478803,
+      "grad_norm": 10.46296215057373,
+      "learning_rate": 1.242992518703242e-05,
+      "loss": 0.4414,
+      "step": 30370
+    },
+    {
+      "epoch": 7.576059850374065,
+      "grad_norm": 6.196642875671387,
+      "learning_rate": 1.2427431421446385e-05,
+      "loss": 0.3855,
+      "step": 30380
+    },
+    {
+      "epoch": 7.5785536159601,
+      "grad_norm": 9.153206825256348,
+      "learning_rate": 1.242493765586035e-05,
+      "loss": 0.2903,
+      "step": 30390
+    },
+    {
+      "epoch": 7.581047381546135,
+      "grad_norm": 6.691465377807617,
+      "learning_rate": 1.2422443890274315e-05,
+      "loss": 0.3715,
+      "step": 30400
+    },
+    {
+      "epoch": 7.58354114713217,
+      "grad_norm": 10.62185287475586,
+      "learning_rate": 1.241995012468828e-05,
+      "loss": 0.4213,
+      "step": 30410
+    },
+    {
+      "epoch": 7.586034912718205,
+      "grad_norm": 5.947132587432861,
+      "learning_rate": 1.2417456359102244e-05,
+      "loss": 0.3693,
+      "step": 30420
+    },
+    {
+      "epoch": 7.58852867830424,
+      "grad_norm": 7.0988993644714355,
+      "learning_rate": 1.2414962593516211e-05,
+      "loss": 0.3849,
+      "step": 30430
+    },
+    {
+      "epoch": 7.5910224438902745,
+      "grad_norm": 6.948726177215576,
+      "learning_rate": 1.2412468827930175e-05,
+      "loss": 0.3642,
+      "step": 30440
+    },
+    {
+      "epoch": 7.593516209476309,
+      "grad_norm": 5.161392688751221,
+      "learning_rate": 1.2409975062344142e-05,
+      "loss": 0.4056,
+      "step": 30450
+    },
+    {
+      "epoch": 7.596009975062344,
+      "grad_norm": 8.350500106811523,
+      "learning_rate": 1.2407481296758106e-05,
+      "loss": 0.4004,
+      "step": 30460
+    },
+    {
+      "epoch": 7.598503740648379,
+      "grad_norm": 7.8617753982543945,
+      "learning_rate": 1.2404987531172071e-05,
+      "loss": 0.3857,
+      "step": 30470
+    },
+    {
+      "epoch": 7.600997506234414,
+      "grad_norm": 6.594773292541504,
+      "learning_rate": 1.2402493765586036e-05,
+      "loss": 0.3984,
+      "step": 30480
+    },
+    {
+      "epoch": 7.603491271820449,
+      "grad_norm": 5.078848361968994,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 0.3225,
+      "step": 30490
+    },
+    {
+      "epoch": 7.605985037406484,
+      "grad_norm": 6.860848903656006,
+      "learning_rate": 1.2397506234413965e-05,
+      "loss": 0.3953,
+      "step": 30500
+    },
+    {
+      "epoch": 7.6084788029925186,
+      "grad_norm": 6.241046905517578,
+      "learning_rate": 1.2395012468827932e-05,
+      "loss": 0.3999,
+      "step": 30510
+    },
+    {
+      "epoch": 7.610972568578553,
+      "grad_norm": 4.877734661102295,
+      "learning_rate": 1.2392518703241896e-05,
+      "loss": 0.3535,
+      "step": 30520
+    },
+    {
+      "epoch": 7.613466334164588,
+      "grad_norm": 6.138383388519287,
+      "learning_rate": 1.2390024937655863e-05,
+      "loss": 0.4553,
+      "step": 30530
+    },
+    {
+      "epoch": 7.615960099750623,
+      "grad_norm": 6.23154354095459,
+      "learning_rate": 1.2387531172069826e-05,
+      "loss": 0.3939,
+      "step": 30540
+    },
+    {
+      "epoch": 7.618453865336658,
+      "grad_norm": 5.0998945236206055,
+      "learning_rate": 1.238503740648379e-05,
+      "loss": 0.3799,
+      "step": 30550
+    },
+    {
+      "epoch": 7.620947630922693,
+      "grad_norm": 5.4622392654418945,
+      "learning_rate": 1.2382543640897757e-05,
+      "loss": 0.4081,
+      "step": 30560
+    },
+    {
+      "epoch": 7.623441396508728,
+      "grad_norm": 4.260573863983154,
+      "learning_rate": 1.2380049875311722e-05,
+      "loss": 0.3986,
+      "step": 30570
+    },
+    {
+      "epoch": 7.625935162094763,
+      "grad_norm": 8.23079776763916,
+      "learning_rate": 1.2377556109725686e-05,
+      "loss": 0.3604,
+      "step": 30580
+    },
+    {
+      "epoch": 7.628428927680798,
+      "grad_norm": 8.109442710876465,
+      "learning_rate": 1.2375062344139653e-05,
+      "loss": 0.4082,
+      "step": 30590
+    },
+    {
+      "epoch": 7.630922693266833,
+      "grad_norm": 4.392514705657959,
+      "learning_rate": 1.2372568578553617e-05,
+      "loss": 0.3406,
+      "step": 30600
+    },
+    {
+      "epoch": 7.633416458852868,
+      "grad_norm": 6.927611827850342,
+      "learning_rate": 1.2370074812967584e-05,
+      "loss": 0.3685,
+      "step": 30610
+    },
+    {
+      "epoch": 7.635910224438903,
+      "grad_norm": 8.26883602142334,
+      "learning_rate": 1.2367581047381547e-05,
+      "loss": 0.4477,
+      "step": 30620
+    },
+    {
+      "epoch": 7.638403990024938,
+      "grad_norm": 6.3622517585754395,
+      "learning_rate": 1.236508728179551e-05,
+      "loss": 0.337,
+      "step": 30630
+    },
+    {
+      "epoch": 7.640897755610973,
+      "grad_norm": 9.278800010681152,
+      "learning_rate": 1.2362593516209478e-05,
+      "loss": 0.3987,
+      "step": 30640
+    },
+    {
+      "epoch": 7.643391521197008,
+      "grad_norm": 4.636536598205566,
+      "learning_rate": 1.2360099750623441e-05,
+      "loss": 0.3779,
+      "step": 30650
+    },
+    {
+      "epoch": 7.6458852867830425,
+      "grad_norm": 7.597597122192383,
+      "learning_rate": 1.2357605985037408e-05,
+      "loss": 0.3261,
+      "step": 30660
+    },
+    {
+      "epoch": 7.648379052369077,
+      "grad_norm": 6.5270771980285645,
+      "learning_rate": 1.2355112219451372e-05,
+      "loss": 0.3113,
+      "step": 30670
+    },
+    {
+      "epoch": 7.650872817955112,
+      "grad_norm": 8.069908142089844,
+      "learning_rate": 1.2352618453865337e-05,
+      "loss": 0.3637,
+      "step": 30680
+    },
+    {
+      "epoch": 7.653366583541147,
+      "grad_norm": 7.320400238037109,
+      "learning_rate": 1.2350124688279304e-05,
+      "loss": 0.3699,
+      "step": 30690
+    },
+    {
+      "epoch": 7.655860349127182,
+      "grad_norm": 15.375004768371582,
+      "learning_rate": 1.2347630922693268e-05,
+      "loss": 0.367,
+      "step": 30700
+    },
+    {
+      "epoch": 7.658354114713217,
+      "grad_norm": 7.086206436157227,
+      "learning_rate": 1.2345137157107232e-05,
+      "loss": 0.3479,
+      "step": 30710
+    },
+    {
+      "epoch": 7.660847880299252,
+      "grad_norm": 7.063462257385254,
+      "learning_rate": 1.2342643391521199e-05,
+      "loss": 0.3503,
+      "step": 30720
+    },
+    {
+      "epoch": 7.6633416458852865,
+      "grad_norm": 6.879456043243408,
+      "learning_rate": 1.2340149625935162e-05,
+      "loss": 0.3552,
+      "step": 30730
+    },
+    {
+      "epoch": 7.665835411471321,
+      "grad_norm": 7.064803600311279,
+      "learning_rate": 1.233765586034913e-05,
+      "loss": 0.3794,
+      "step": 30740
+    },
+    {
+      "epoch": 7.668329177057356,
+      "grad_norm": 6.3318023681640625,
+      "learning_rate": 1.2335162094763093e-05,
+      "loss": 0.3434,
+      "step": 30750
+    },
+    {
+      "epoch": 7.670822942643391,
+      "grad_norm": 6.186346054077148,
+      "learning_rate": 1.2332668329177058e-05,
+      "loss": 0.3624,
+      "step": 30760
+    },
+    {
+      "epoch": 7.673316708229427,
+      "grad_norm": 6.247686386108398,
+      "learning_rate": 1.2330174563591023e-05,
+      "loss": 0.3922,
+      "step": 30770
+    },
+    {
+      "epoch": 7.675810473815462,
+      "grad_norm": 5.863278865814209,
+      "learning_rate": 1.2327680798004989e-05,
+      "loss": 0.3569,
+      "step": 30780
+    },
+    {
+      "epoch": 7.678304239401497,
+      "grad_norm": 13.845823287963867,
+      "learning_rate": 1.2325187032418952e-05,
+      "loss": 0.3568,
+      "step": 30790
+    },
+    {
+      "epoch": 7.6807980049875315,
+      "grad_norm": 7.736221790313721,
+      "learning_rate": 1.232269326683292e-05,
+      "loss": 0.3471,
+      "step": 30800
+    },
+    {
+      "epoch": 7.683291770573566,
+      "grad_norm": 6.512248992919922,
+      "learning_rate": 1.2320199501246883e-05,
+      "loss": 0.3172,
+      "step": 30810
+    },
+    {
+      "epoch": 7.685785536159601,
+      "grad_norm": 6.663067817687988,
+      "learning_rate": 1.231770573566085e-05,
+      "loss": 0.3857,
+      "step": 30820
+    },
+    {
+      "epoch": 7.688279301745636,
+      "grad_norm": 8.590274810791016,
+      "learning_rate": 1.2315211970074814e-05,
+      "loss": 0.3427,
+      "step": 30830
+    },
+    {
+      "epoch": 7.690773067331671,
+      "grad_norm": 7.1273722648620605,
+      "learning_rate": 1.2312718204488779e-05,
+      "loss": 0.4224,
+      "step": 30840
+    },
+    {
+      "epoch": 7.693266832917706,
+      "grad_norm": 6.8802924156188965,
+      "learning_rate": 1.2310224438902744e-05,
+      "loss": 0.4097,
+      "step": 30850
+    },
+    {
+      "epoch": 7.695760598503741,
+      "grad_norm": 9.271326065063477,
+      "learning_rate": 1.230773067331671e-05,
+      "loss": 0.3757,
+      "step": 30860
+    },
+    {
+      "epoch": 7.698254364089776,
+      "grad_norm": 6.1901421546936035,
+      "learning_rate": 1.2305236907730673e-05,
+      "loss": 0.3571,
+      "step": 30870
+    },
+    {
+      "epoch": 7.7007481296758105,
+      "grad_norm": 8.706985473632812,
+      "learning_rate": 1.230274314214464e-05,
+      "loss": 0.335,
+      "step": 30880
+    },
+    {
+      "epoch": 7.703241895261845,
+      "grad_norm": 5.975154399871826,
+      "learning_rate": 1.2300249376558604e-05,
+      "loss": 0.3703,
+      "step": 30890
+    },
+    {
+      "epoch": 7.70573566084788,
+      "grad_norm": 8.364744186401367,
+      "learning_rate": 1.229775561097257e-05,
+      "loss": 0.3532,
+      "step": 30900
+    },
+    {
+      "epoch": 7.708229426433915,
+      "grad_norm": 6.633474826812744,
+      "learning_rate": 1.2295261845386534e-05,
+      "loss": 0.3447,
+      "step": 30910
+    },
+    {
+      "epoch": 7.71072319201995,
+      "grad_norm": 7.192882061004639,
+      "learning_rate": 1.22927680798005e-05,
+      "loss": 0.365,
+      "step": 30920
+    },
+    {
+      "epoch": 7.713216957605985,
+      "grad_norm": 4.311435222625732,
+      "learning_rate": 1.2290274314214465e-05,
+      "loss": 0.3817,
+      "step": 30930
+    },
+    {
+      "epoch": 7.71571072319202,
+      "grad_norm": 9.537071228027344,
+      "learning_rate": 1.228778054862843e-05,
+      "loss": 0.4124,
+      "step": 30940
+    },
+    {
+      "epoch": 7.7182044887780545,
+      "grad_norm": 6.788525581359863,
+      "learning_rate": 1.2285286783042396e-05,
+      "loss": 0.4284,
+      "step": 30950
+    },
+    {
+      "epoch": 7.720698254364089,
+      "grad_norm": 4.7663798332214355,
+      "learning_rate": 1.2282793017456361e-05,
+      "loss": 0.3825,
+      "step": 30960
+    },
+    {
+      "epoch": 7.723192019950124,
+      "grad_norm": 6.757378578186035,
+      "learning_rate": 1.2280299251870325e-05,
+      "loss": 0.3331,
+      "step": 30970
+    },
+    {
+      "epoch": 7.725685785536159,
+      "grad_norm": 6.984048366546631,
+      "learning_rate": 1.2277805486284292e-05,
+      "loss": 0.354,
+      "step": 30980
+    },
+    {
+      "epoch": 7.728179551122195,
+      "grad_norm": 6.14350700378418,
+      "learning_rate": 1.2275311720698255e-05,
+      "loss": 0.3543,
+      "step": 30990
+    },
+    {
+      "epoch": 7.73067331670823,
+      "grad_norm": 5.969043254852295,
+      "learning_rate": 1.2272817955112219e-05,
+      "loss": 0.3788,
+      "step": 31000
+    },
+    {
+      "epoch": 7.733167082294265,
+      "grad_norm": 6.313365936279297,
+      "learning_rate": 1.2270324189526186e-05,
+      "loss": 0.3608,
+      "step": 31010
+    },
+    {
+      "epoch": 7.7356608478802995,
+      "grad_norm": 5.8176703453063965,
+      "learning_rate": 1.226783042394015e-05,
+      "loss": 0.3599,
+      "step": 31020
+    },
+    {
+      "epoch": 7.738154613466334,
+      "grad_norm": 5.260715484619141,
+      "learning_rate": 1.2265336658354116e-05,
+      "loss": 0.3885,
+      "step": 31030
+    },
+    {
+      "epoch": 7.740648379052369,
+      "grad_norm": 9.272298812866211,
+      "learning_rate": 1.2262842892768082e-05,
+      "loss": 0.3606,
+      "step": 31040
+    },
+    {
+      "epoch": 7.743142144638404,
+      "grad_norm": 6.716904163360596,
+      "learning_rate": 1.2260349127182045e-05,
+      "loss": 0.3364,
+      "step": 31050
+    },
+    {
+      "epoch": 7.745635910224439,
+      "grad_norm": 7.895336151123047,
+      "learning_rate": 1.2257855361596012e-05,
+      "loss": 0.3685,
+      "step": 31060
+    },
+    {
+      "epoch": 7.748129675810474,
+      "grad_norm": 6.1272969245910645,
+      "learning_rate": 1.2255361596009976e-05,
+      "loss": 0.3745,
+      "step": 31070
+    },
+    {
+      "epoch": 7.750623441396509,
+      "grad_norm": 6.436134338378906,
+      "learning_rate": 1.225286783042394e-05,
+      "loss": 0.363,
+      "step": 31080
+    },
+    {
+      "epoch": 7.753117206982544,
+      "grad_norm": 7.552138328552246,
+      "learning_rate": 1.2250374064837907e-05,
+      "loss": 0.4172,
+      "step": 31090
+    },
+    {
+      "epoch": 7.7556109725685785,
+      "grad_norm": 4.2307658195495605,
+      "learning_rate": 1.224788029925187e-05,
+      "loss": 0.4331,
+      "step": 31100
+    },
+    {
+      "epoch": 7.758104738154613,
+      "grad_norm": 6.373022556304932,
+      "learning_rate": 1.2245386533665837e-05,
+      "loss": 0.3571,
+      "step": 31110
+    },
+    {
+      "epoch": 7.760598503740648,
+      "grad_norm": 6.442211151123047,
+      "learning_rate": 1.22428927680798e-05,
+      "loss": 0.3335,
+      "step": 31120
+    },
+    {
+      "epoch": 7.763092269326683,
+      "grad_norm": 5.978911399841309,
+      "learning_rate": 1.2240399002493766e-05,
+      "loss": 0.3363,
+      "step": 31130
+    },
+    {
+      "epoch": 7.765586034912718,
+      "grad_norm": 7.628538131713867,
+      "learning_rate": 1.2237905236907731e-05,
+      "loss": 0.389,
+      "step": 31140
+    },
+    {
+      "epoch": 7.768079800498753,
+      "grad_norm": 5.834782600402832,
+      "learning_rate": 1.2235411471321697e-05,
+      "loss": 0.3835,
+      "step": 31150
+    },
+    {
+      "epoch": 7.770573566084788,
+      "grad_norm": 9.812804222106934,
+      "learning_rate": 1.2232917705735664e-05,
+      "loss": 0.3542,
+      "step": 31160
+    },
+    {
+      "epoch": 7.773067331670823,
+      "grad_norm": 5.178422451019287,
+      "learning_rate": 1.2230423940149627e-05,
+      "loss": 0.3612,
+      "step": 31170
+    },
+    {
+      "epoch": 7.775561097256858,
+      "grad_norm": 5.95904016494751,
+      "learning_rate": 1.2227930174563591e-05,
+      "loss": 0.3285,
+      "step": 31180
+    },
+    {
+      "epoch": 7.778054862842893,
+      "grad_norm": 7.034958362579346,
+      "learning_rate": 1.2225436408977558e-05,
+      "loss": 0.3777,
+      "step": 31190
+    },
+    {
+      "epoch": 7.780548628428928,
+      "grad_norm": 13.055363655090332,
+      "learning_rate": 1.2222942643391522e-05,
+      "loss": 0.4765,
+      "step": 31200
+    },
+    {
+      "epoch": 7.783042394014963,
+      "grad_norm": 8.228713989257812,
+      "learning_rate": 1.2220448877805487e-05,
+      "loss": 0.393,
+      "step": 31210
+    },
+    {
+      "epoch": 7.785536159600998,
+      "grad_norm": 8.480477333068848,
+      "learning_rate": 1.2217955112219452e-05,
+      "loss": 0.352,
+      "step": 31220
+    },
+    {
+      "epoch": 7.788029925187033,
+      "grad_norm": 7.811897277832031,
+      "learning_rate": 1.2215461346633418e-05,
+      "loss": 0.4076,
+      "step": 31230
+    },
+    {
+      "epoch": 7.7905236907730675,
+      "grad_norm": 6.502315521240234,
+      "learning_rate": 1.2212967581047383e-05,
+      "loss": 0.3238,
+      "step": 31240
+    },
+    {
+      "epoch": 7.793017456359102,
+      "grad_norm": 6.6692657470703125,
+      "learning_rate": 1.2210473815461348e-05,
+      "loss": 0.3445,
+      "step": 31250
+    },
+    {
+      "epoch": 7.795511221945137,
+      "grad_norm": 5.5514655113220215,
+      "learning_rate": 1.2207980049875312e-05,
+      "loss": 0.3603,
+      "step": 31260
+    },
+    {
+      "epoch": 7.798004987531172,
+      "grad_norm": 5.353459358215332,
+      "learning_rate": 1.2205486284289279e-05,
+      "loss": 0.3117,
+      "step": 31270
+    },
+    {
+      "epoch": 7.800498753117207,
+      "grad_norm": 6.266322135925293,
+      "learning_rate": 1.2202992518703242e-05,
+      "loss": 0.3608,
+      "step": 31280
+    },
+    {
+      "epoch": 7.802992518703242,
+      "grad_norm": 5.57569694519043,
+      "learning_rate": 1.2200498753117208e-05,
+      "loss": 0.3953,
+      "step": 31290
+    },
+    {
+      "epoch": 7.805486284289277,
+      "grad_norm": 5.770155429840088,
+      "learning_rate": 1.2198004987531173e-05,
+      "loss": 0.4057,
+      "step": 31300
+    },
+    {
+      "epoch": 7.807980049875312,
+      "grad_norm": 8.527885437011719,
+      "learning_rate": 1.2195511221945138e-05,
+      "loss": 0.4245,
+      "step": 31310
+    },
+    {
+      "epoch": 7.8104738154613464,
+      "grad_norm": 8.572240829467773,
+      "learning_rate": 1.2193017456359104e-05,
+      "loss": 0.3845,
+      "step": 31320
+    },
+    {
+      "epoch": 7.812967581047381,
+      "grad_norm": 8.089426040649414,
+      "learning_rate": 1.2190523690773069e-05,
+      "loss": 0.3858,
+      "step": 31330
+    },
+    {
+      "epoch": 7.815461346633416,
+      "grad_norm": 5.3781657218933105,
+      "learning_rate": 1.2188029925187033e-05,
+      "loss": 0.4435,
+      "step": 31340
+    },
+    {
+      "epoch": 7.817955112219451,
+      "grad_norm": 7.371554374694824,
+      "learning_rate": 1.2185536159601e-05,
+      "loss": 0.3908,
+      "step": 31350
+    },
+    {
+      "epoch": 7.820448877805486,
+      "grad_norm": 6.417791366577148,
+      "learning_rate": 1.2183042394014963e-05,
+      "loss": 0.3638,
+      "step": 31360
+    },
+    {
+      "epoch": 7.822942643391521,
+      "grad_norm": 8.140090942382812,
+      "learning_rate": 1.2180548628428927e-05,
+      "loss": 0.4004,
+      "step": 31370
+    },
+    {
+      "epoch": 7.825436408977556,
+      "grad_norm": 28.9359073638916,
+      "learning_rate": 1.2178054862842894e-05,
+      "loss": 0.3789,
+      "step": 31380
+    },
+    {
+      "epoch": 7.8279301745635905,
+      "grad_norm": 5.224822044372559,
+      "learning_rate": 1.2175561097256859e-05,
+      "loss": 0.3685,
+      "step": 31390
+    },
+    {
+      "epoch": 7.830423940149626,
+      "grad_norm": 5.957163333892822,
+      "learning_rate": 1.2173067331670824e-05,
+      "loss": 0.2989,
+      "step": 31400
+    },
+    {
+      "epoch": 7.832917705735661,
+      "grad_norm": 7.902908802032471,
+      "learning_rate": 1.217057356608479e-05,
+      "loss": 0.3216,
+      "step": 31410
+    },
+    {
+      "epoch": 7.835411471321696,
+      "grad_norm": 9.915661811828613,
+      "learning_rate": 1.2168079800498753e-05,
+      "loss": 0.4022,
+      "step": 31420
+    },
+    {
+      "epoch": 7.837905236907731,
+      "grad_norm": 8.479707717895508,
+      "learning_rate": 1.216558603491272e-05,
+      "loss": 0.3614,
+      "step": 31430
+    },
+    {
+      "epoch": 7.840399002493766,
+      "grad_norm": 5.247765064239502,
+      "learning_rate": 1.2163092269326684e-05,
+      "loss": 0.354,
+      "step": 31440
+    },
+    {
+      "epoch": 7.842892768079801,
+      "grad_norm": 9.220551490783691,
+      "learning_rate": 1.2160598503740651e-05,
+      "loss": 0.4473,
+      "step": 31450
+    },
+    {
+      "epoch": 7.8453865336658355,
+      "grad_norm": 5.292842864990234,
+      "learning_rate": 1.2158104738154615e-05,
+      "loss": 0.3996,
+      "step": 31460
+    },
+    {
+      "epoch": 7.84788029925187,
+      "grad_norm": 9.123703956604004,
+      "learning_rate": 1.2155610972568578e-05,
+      "loss": 0.4442,
+      "step": 31470
+    },
+    {
+      "epoch": 7.850374064837905,
+      "grad_norm": 5.475888252258301,
+      "learning_rate": 1.2153117206982545e-05,
+      "loss": 0.3169,
+      "step": 31480
+    },
+    {
+      "epoch": 7.85286783042394,
+      "grad_norm": 7.926106929779053,
+      "learning_rate": 1.2150623441396509e-05,
+      "loss": 0.3636,
+      "step": 31490
+    },
+    {
+      "epoch": 7.855361596009975,
+      "grad_norm": 6.426756858825684,
+      "learning_rate": 1.2148129675810474e-05,
+      "loss": 0.3884,
+      "step": 31500
+    },
+    {
+      "epoch": 7.85785536159601,
+      "grad_norm": 7.212559700012207,
+      "learning_rate": 1.2145635910224441e-05,
+      "loss": 0.3312,
+      "step": 31510
+    },
+    {
+      "epoch": 7.860349127182045,
+      "grad_norm": 5.426643371582031,
+      "learning_rate": 1.2143142144638405e-05,
+      "loss": 0.3075,
+      "step": 31520
+    },
+    {
+      "epoch": 7.86284289276808,
+      "grad_norm": 6.562136650085449,
+      "learning_rate": 1.2140648379052372e-05,
+      "loss": 0.4217,
+      "step": 31530
+    },
+    {
+      "epoch": 7.865336658354114,
+      "grad_norm": 7.82149076461792,
+      "learning_rate": 1.2138154613466335e-05,
+      "loss": 0.347,
+      "step": 31540
+    },
+    {
+      "epoch": 7.867830423940149,
+      "grad_norm": 6.736440181732178,
+      "learning_rate": 1.2135660847880299e-05,
+      "loss": 0.3143,
+      "step": 31550
+    },
+    {
+      "epoch": 7.870324189526184,
+      "grad_norm": 7.551044940948486,
+      "learning_rate": 1.2133167082294266e-05,
+      "loss": 0.4345,
+      "step": 31560
+    },
+    {
+      "epoch": 7.87281795511222,
+      "grad_norm": 11.6002836227417,
+      "learning_rate": 1.213067331670823e-05,
+      "loss": 0.3961,
+      "step": 31570
+    },
+    {
+      "epoch": 7.875311720698255,
+      "grad_norm": 4.926860809326172,
+      "learning_rate": 1.2128179551122195e-05,
+      "loss": 0.3755,
+      "step": 31580
+    },
+    {
+      "epoch": 7.87780548628429,
+      "grad_norm": 13.205385208129883,
+      "learning_rate": 1.212568578553616e-05,
+      "loss": 0.3714,
+      "step": 31590
+    },
+    {
+      "epoch": 7.8802992518703245,
+      "grad_norm": 4.073211193084717,
+      "learning_rate": 1.2123192019950126e-05,
+      "loss": 0.2892,
+      "step": 31600
+    },
+    {
+      "epoch": 7.882793017456359,
+      "grad_norm": 11.72350788116455,
+      "learning_rate": 1.212069825436409e-05,
+      "loss": 0.3547,
+      "step": 31610
+    },
+    {
+      "epoch": 7.885286783042394,
+      "grad_norm": 6.860603332519531,
+      "learning_rate": 1.2118204488778056e-05,
+      "loss": 0.3574,
+      "step": 31620
+    },
+    {
+      "epoch": 7.887780548628429,
+      "grad_norm": 5.04127836227417,
+      "learning_rate": 1.211571072319202e-05,
+      "loss": 0.3101,
+      "step": 31630
+    },
+    {
+      "epoch": 7.890274314214464,
+      "grad_norm": 5.518341064453125,
+      "learning_rate": 1.2113216957605987e-05,
+      "loss": 0.4506,
+      "step": 31640
+    },
+    {
+      "epoch": 7.892768079800499,
+      "grad_norm": 6.156492710113525,
+      "learning_rate": 1.211072319201995e-05,
+      "loss": 0.3326,
+      "step": 31650
+    },
+    {
+      "epoch": 7.895261845386534,
+      "grad_norm": 7.646767616271973,
+      "learning_rate": 1.2108229426433917e-05,
+      "loss": 0.3646,
+      "step": 31660
+    },
+    {
+      "epoch": 7.897755610972569,
+      "grad_norm": 5.843283176422119,
+      "learning_rate": 1.2105735660847881e-05,
+      "loss": 0.418,
+      "step": 31670
+    },
+    {
+      "epoch": 7.9002493765586035,
+      "grad_norm": 7.041114330291748,
+      "learning_rate": 1.2103241895261846e-05,
+      "loss": 0.3759,
+      "step": 31680
+    },
+    {
+      "epoch": 7.902743142144638,
+      "grad_norm": 8.82856273651123,
+      "learning_rate": 1.2100748129675812e-05,
+      "loss": 0.478,
+      "step": 31690
+    },
+    {
+      "epoch": 7.905236907730673,
+      "grad_norm": 6.0212273597717285,
+      "learning_rate": 1.2098254364089777e-05,
+      "loss": 0.3257,
+      "step": 31700
+    },
+    {
+      "epoch": 7.907730673316708,
+      "grad_norm": 8.000700950622559,
+      "learning_rate": 1.209576059850374e-05,
+      "loss": 0.3212,
+      "step": 31710
+    },
+    {
+      "epoch": 7.910224438902743,
+      "grad_norm": 9.041226387023926,
+      "learning_rate": 1.2093266832917708e-05,
+      "loss": 0.3234,
+      "step": 31720
+    },
+    {
+      "epoch": 7.912718204488778,
+      "grad_norm": 10.657549858093262,
+      "learning_rate": 1.2090773067331671e-05,
+      "loss": 0.4932,
+      "step": 31730
+    },
+    {
+      "epoch": 7.915211970074813,
+      "grad_norm": 8.23529052734375,
+      "learning_rate": 1.2088279301745638e-05,
+      "loss": 0.3676,
+      "step": 31740
+    },
+    {
+      "epoch": 7.917705735660848,
+      "grad_norm": 5.820125102996826,
+      "learning_rate": 1.2085785536159602e-05,
+      "loss": 0.38,
+      "step": 31750
+    },
+    {
+      "epoch": 7.920199501246882,
+      "grad_norm": 7.681821346282959,
+      "learning_rate": 1.2083291770573567e-05,
+      "loss": 0.3368,
+      "step": 31760
+    },
+    {
+      "epoch": 7.922693266832917,
+      "grad_norm": 6.52007532119751,
+      "learning_rate": 1.2080798004987532e-05,
+      "loss": 0.3466,
+      "step": 31770
+    },
+    {
+      "epoch": 7.925187032418952,
+      "grad_norm": 8.937047958374023,
+      "learning_rate": 1.2078304239401498e-05,
+      "loss": 0.3921,
+      "step": 31780
+    },
+    {
+      "epoch": 7.927680798004987,
+      "grad_norm": 7.19578742980957,
+      "learning_rate": 1.2075810473815461e-05,
+      "loss": 0.3695,
+      "step": 31790
+    },
+    {
+      "epoch": 7.930174563591023,
+      "grad_norm": 9.430617332458496,
+      "learning_rate": 1.2073316708229428e-05,
+      "loss": 0.3753,
+      "step": 31800
+    },
+    {
+      "epoch": 7.932668329177058,
+      "grad_norm": 5.713561058044434,
+      "learning_rate": 1.2070822942643392e-05,
+      "loss": 0.4164,
+      "step": 31810
+    },
+    {
+      "epoch": 7.9351620947630925,
+      "grad_norm": 6.72670316696167,
+      "learning_rate": 1.2068329177057359e-05,
+      "loss": 0.4523,
+      "step": 31820
+    },
+    {
+      "epoch": 7.937655860349127,
+      "grad_norm": 6.967277526855469,
+      "learning_rate": 1.2065835411471323e-05,
+      "loss": 0.4086,
+      "step": 31830
+    },
+    {
+      "epoch": 7.940149625935162,
+      "grad_norm": 6.393032073974609,
+      "learning_rate": 1.2063341645885286e-05,
+      "loss": 0.3532,
+      "step": 31840
+    },
+    {
+      "epoch": 7.942643391521197,
+      "grad_norm": 8.138836860656738,
+      "learning_rate": 1.2060847880299253e-05,
+      "loss": 0.3699,
+      "step": 31850
+    },
+    {
+      "epoch": 7.945137157107232,
+      "grad_norm": 7.282432556152344,
+      "learning_rate": 1.2058354114713218e-05,
+      "loss": 0.3925,
+      "step": 31860
+    },
+    {
+      "epoch": 7.947630922693267,
+      "grad_norm": 5.286504745483398,
+      "learning_rate": 1.2055860349127182e-05,
+      "loss": 0.2947,
+      "step": 31870
+    },
+    {
+      "epoch": 7.950124688279302,
+      "grad_norm": 7.868513584136963,
+      "learning_rate": 1.2053366583541149e-05,
+      "loss": 0.501,
+      "step": 31880
+    },
+    {
+      "epoch": 7.952618453865337,
+      "grad_norm": 8.221051216125488,
+      "learning_rate": 1.2050872817955113e-05,
+      "loss": 0.4184,
+      "step": 31890
+    },
+    {
+      "epoch": 7.9551122194513715,
+      "grad_norm": 8.704934120178223,
+      "learning_rate": 1.204837905236908e-05,
+      "loss": 0.3372,
+      "step": 31900
+    },
+    {
+      "epoch": 7.957605985037406,
+      "grad_norm": 8.388035774230957,
+      "learning_rate": 1.2045885286783043e-05,
+      "loss": 0.2941,
+      "step": 31910
+    },
+    {
+      "epoch": 7.960099750623441,
+      "grad_norm": 6.82948637008667,
+      "learning_rate": 1.2043391521197007e-05,
+      "loss": 0.3935,
+      "step": 31920
+    },
+    {
+      "epoch": 7.962593516209476,
+      "grad_norm": 18.161638259887695,
+      "learning_rate": 1.2040897755610974e-05,
+      "loss": 0.4707,
+      "step": 31930
+    },
+    {
+      "epoch": 7.965087281795511,
+      "grad_norm": 7.017074108123779,
+      "learning_rate": 1.2038403990024938e-05,
+      "loss": 0.3647,
+      "step": 31940
+    },
+    {
+      "epoch": 7.967581047381546,
+      "grad_norm": 7.731515884399414,
+      "learning_rate": 1.2035910224438905e-05,
+      "loss": 0.3442,
+      "step": 31950
+    },
+    {
+      "epoch": 7.970074812967581,
+      "grad_norm": 5.6125898361206055,
+      "learning_rate": 1.2033416458852868e-05,
+      "loss": 0.4005,
+      "step": 31960
+    },
+    {
+      "epoch": 7.9725685785536164,
+      "grad_norm": 8.200444221496582,
+      "learning_rate": 1.2030922693266834e-05,
+      "loss": 0.4055,
+      "step": 31970
+    },
+    {
+      "epoch": 7.975062344139651,
+      "grad_norm": 8.561641693115234,
+      "learning_rate": 1.20284289276808e-05,
+      "loss": 0.3479,
+      "step": 31980
+    },
+    {
+      "epoch": 7.977556109725686,
+      "grad_norm": 6.790326118469238,
+      "learning_rate": 1.2025935162094764e-05,
+      "loss": 0.4134,
+      "step": 31990
+    },
+    {
+      "epoch": 7.980049875311721,
+      "grad_norm": 6.383502006530762,
+      "learning_rate": 1.2023441396508728e-05,
+      "loss": 0.3949,
+      "step": 32000
+    },
+    {
+      "epoch": 7.982543640897756,
+      "grad_norm": 5.95525598526001,
+      "learning_rate": 1.2020947630922695e-05,
+      "loss": 0.4402,
+      "step": 32010
+    },
+    {
+      "epoch": 7.985037406483791,
+      "grad_norm": 7.909287929534912,
+      "learning_rate": 1.2018453865336658e-05,
+      "loss": 0.349,
+      "step": 32020
+    },
+    {
+      "epoch": 7.987531172069826,
+      "grad_norm": 5.539024829864502,
+      "learning_rate": 1.2015960099750625e-05,
+      "loss": 0.3202,
+      "step": 32030
+    },
+    {
+      "epoch": 7.9900249376558605,
+      "grad_norm": 7.293623447418213,
+      "learning_rate": 1.2013466334164589e-05,
+      "loss": 0.3595,
+      "step": 32040
+    },
+    {
+      "epoch": 7.992518703241895,
+      "grad_norm": 6.55578088760376,
+      "learning_rate": 1.2010972568578554e-05,
+      "loss": 0.3922,
+      "step": 32050
+    },
+    {
+      "epoch": 7.99501246882793,
+      "grad_norm": 5.113552093505859,
+      "learning_rate": 1.200847880299252e-05,
+      "loss": 0.4029,
+      "step": 32060
+    },
+    {
+      "epoch": 7.997506234413965,
+      "grad_norm": 6.873347282409668,
+      "learning_rate": 1.2005985037406485e-05,
+      "loss": 0.3852,
+      "step": 32070
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 10.353362083435059,
+      "learning_rate": 1.2003491271820449e-05,
+      "loss": 0.3054,
+      "step": 32080
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 0.4143351912498474,
+      "eval_runtime": 59.8965,
+      "eval_samples_per_second": 16.746,
+      "eval_steps_per_second": 16.746,
+      "step": 32080
+    },
+    {
+      "epoch": 8.002493765586035,
+      "grad_norm": 6.222545623779297,
+      "learning_rate": 1.2000997506234416e-05,
+      "loss": 0.3814,
+      "step": 32090
+    },
+    {
+      "epoch": 8.00498753117207,
+      "grad_norm": 6.579477310180664,
+      "learning_rate": 1.1998503740648379e-05,
+      "loss": 0.3755,
+      "step": 32100
+    },
+    {
+      "epoch": 8.007481296758105,
+      "grad_norm": 4.621208667755127,
+      "learning_rate": 1.1996009975062346e-05,
+      "loss": 0.2821,
+      "step": 32110
+    },
+    {
+      "epoch": 8.00997506234414,
+      "grad_norm": 5.9014081954956055,
+      "learning_rate": 1.199351620947631e-05,
+      "loss": 0.3424,
+      "step": 32120
+    },
+    {
+      "epoch": 8.012468827930174,
+      "grad_norm": 8.920719146728516,
+      "learning_rate": 1.1991022443890275e-05,
+      "loss": 0.4108,
+      "step": 32130
+    },
+    {
+      "epoch": 8.01496259351621,
+      "grad_norm": 7.701635360717773,
+      "learning_rate": 1.198852867830424e-05,
+      "loss": 0.3543,
+      "step": 32140
+    },
+    {
+      "epoch": 8.017456359102244,
+      "grad_norm": 6.4233856201171875,
+      "learning_rate": 1.1986034912718206e-05,
+      "loss": 0.3799,
+      "step": 32150
+    },
+    {
+      "epoch": 8.019950124688279,
+      "grad_norm": 9.75025463104248,
+      "learning_rate": 1.1983541147132171e-05,
+      "loss": 0.4066,
+      "step": 32160
+    },
+    {
+      "epoch": 8.022443890274314,
+      "grad_norm": 7.719704627990723,
+      "learning_rate": 1.1981047381546136e-05,
+      "loss": 0.3938,
+      "step": 32170
+    },
+    {
+      "epoch": 8.024937655860349,
+      "grad_norm": 4.815377712249756,
+      "learning_rate": 1.19785536159601e-05,
+      "loss": 0.3635,
+      "step": 32180
+    },
+    {
+      "epoch": 8.027431421446384,
+      "grad_norm": 7.045759201049805,
+      "learning_rate": 1.197630922693267e-05,
+      "loss": 0.3618,
+      "step": 32190
+    },
+    {
+      "epoch": 8.029925187032418,
+      "grad_norm": 5.257409572601318,
+      "learning_rate": 1.1973815461346634e-05,
+      "loss": 0.3325,
+      "step": 32200
+    },
+    {
+      "epoch": 8.032418952618453,
+      "grad_norm": 12.353095054626465,
+      "learning_rate": 1.19713216957606e-05,
+      "loss": 0.356,
+      "step": 32210
+    },
+    {
+      "epoch": 8.034912718204488,
+      "grad_norm": 5.638147830963135,
+      "learning_rate": 1.1968827930174564e-05,
+      "loss": 0.3877,
+      "step": 32220
+    },
+    {
+      "epoch": 8.037406483790523,
+      "grad_norm": 4.99983024597168,
+      "learning_rate": 1.196633416458853e-05,
+      "loss": 0.357,
+      "step": 32230
+    },
+    {
+      "epoch": 8.039900249376558,
+      "grad_norm": 5.99961519241333,
+      "learning_rate": 1.1963840399002495e-05,
+      "loss": 0.3408,
+      "step": 32240
+    },
+    {
+      "epoch": 8.042394014962593,
+      "grad_norm": 8.61434555053711,
+      "learning_rate": 1.196134663341646e-05,
+      "loss": 0.3429,
+      "step": 32250
+    },
+    {
+      "epoch": 8.044887780548628,
+      "grad_norm": 6.741703033447266,
+      "learning_rate": 1.1958852867830424e-05,
+      "loss": 0.3562,
+      "step": 32260
+    },
+    {
+      "epoch": 8.047381546134662,
+      "grad_norm": 4.7175703048706055,
+      "learning_rate": 1.195635910224439e-05,
+      "loss": 0.3447,
+      "step": 32270
+    },
+    {
+      "epoch": 8.049875311720697,
+      "grad_norm": 6.885346412658691,
+      "learning_rate": 1.1953865336658354e-05,
+      "loss": 0.3637,
+      "step": 32280
+    },
+    {
+      "epoch": 8.052369077306734,
+      "grad_norm": 7.284684181213379,
+      "learning_rate": 1.1951371571072321e-05,
+      "loss": 0.3799,
+      "step": 32290
+    },
+    {
+      "epoch": 8.054862842892769,
+      "grad_norm": 8.595766067504883,
+      "learning_rate": 1.1948877805486285e-05,
+      "loss": 0.3967,
+      "step": 32300
+    },
+    {
+      "epoch": 8.057356608478804,
+      "grad_norm": 6.696188926696777,
+      "learning_rate": 1.194638403990025e-05,
+      "loss": 0.3993,
+      "step": 32310
+    },
+    {
+      "epoch": 8.059850374064839,
+      "grad_norm": 5.798731327056885,
+      "learning_rate": 1.1943890274314216e-05,
+      "loss": 0.3159,
+      "step": 32320
+    },
+    {
+      "epoch": 8.062344139650873,
+      "grad_norm": 6.5133891105651855,
+      "learning_rate": 1.1941396508728181e-05,
+      "loss": 0.3556,
+      "step": 32330
+    },
+    {
+      "epoch": 8.064837905236908,
+      "grad_norm": 8.15516471862793,
+      "learning_rate": 1.1938902743142146e-05,
+      "loss": 0.3533,
+      "step": 32340
+    },
+    {
+      "epoch": 8.067331670822943,
+      "grad_norm": 9.06305980682373,
+      "learning_rate": 1.1936408977556111e-05,
+      "loss": 0.419,
+      "step": 32350
+    },
+    {
+      "epoch": 8.069825436408978,
+      "grad_norm": 5.8989644050598145,
+      "learning_rate": 1.1933915211970075e-05,
+      "loss": 0.3469,
+      "step": 32360
+    },
+    {
+      "epoch": 8.072319201995013,
+      "grad_norm": 4.38079833984375,
+      "learning_rate": 1.1931421446384042e-05,
+      "loss": 0.3461,
+      "step": 32370
+    },
+    {
+      "epoch": 8.074812967581048,
+      "grad_norm": 7.34318208694458,
+      "learning_rate": 1.1928927680798006e-05,
+      "loss": 0.3793,
+      "step": 32380
+    },
+    {
+      "epoch": 8.077306733167083,
+      "grad_norm": 6.742260932922363,
+      "learning_rate": 1.192643391521197e-05,
+      "loss": 0.3649,
+      "step": 32390
+    },
+    {
+      "epoch": 8.079800498753118,
+      "grad_norm": 6.300445079803467,
+      "learning_rate": 1.1923940149625936e-05,
+      "loss": 0.3527,
+      "step": 32400
+    },
+    {
+      "epoch": 8.082294264339152,
+      "grad_norm": 12.198246002197266,
+      "learning_rate": 1.1921446384039902e-05,
+      "loss": 0.3832,
+      "step": 32410
+    },
+    {
+      "epoch": 8.084788029925187,
+      "grad_norm": 6.432736873626709,
+      "learning_rate": 1.1918952618453867e-05,
+      "loss": 0.3424,
+      "step": 32420
+    },
+    {
+      "epoch": 8.087281795511222,
+      "grad_norm": 7.7809271812438965,
+      "learning_rate": 1.1916458852867832e-05,
+      "loss": 0.3318,
+      "step": 32430
+    },
+    {
+      "epoch": 8.089775561097257,
+      "grad_norm": 8.33100700378418,
+      "learning_rate": 1.1913965087281796e-05,
+      "loss": 0.3166,
+      "step": 32440
+    },
+    {
+      "epoch": 8.092269326683292,
+      "grad_norm": 7.229568958282471,
+      "learning_rate": 1.1911471321695763e-05,
+      "loss": 0.4009,
+      "step": 32450
+    },
+    {
+      "epoch": 8.094763092269327,
+      "grad_norm": 6.71392822265625,
+      "learning_rate": 1.1908977556109726e-05,
+      "loss": 0.3876,
+      "step": 32460
+    },
+    {
+      "epoch": 8.097256857855362,
+      "grad_norm": 7.794308185577393,
+      "learning_rate": 1.190648379052369e-05,
+      "loss": 0.406,
+      "step": 32470
+    },
+    {
+      "epoch": 8.099750623441397,
+      "grad_norm": 6.342446327209473,
+      "learning_rate": 1.1903990024937657e-05,
+      "loss": 0.3472,
+      "step": 32480
+    },
+    {
+      "epoch": 8.102244389027431,
+      "grad_norm": 7.7976393699646,
+      "learning_rate": 1.190149625935162e-05,
+      "loss": 0.3918,
+      "step": 32490
+    },
+    {
+      "epoch": 8.104738154613466,
+      "grad_norm": 5.704341411590576,
+      "learning_rate": 1.1899002493765588e-05,
+      "loss": 0.3502,
+      "step": 32500
+    },
+    {
+      "epoch": 8.107231920199501,
+      "grad_norm": 6.822027206420898,
+      "learning_rate": 1.1896508728179551e-05,
+      "loss": 0.3749,
+      "step": 32510
+    },
+    {
+      "epoch": 8.109725685785536,
+      "grad_norm": 6.865316867828369,
+      "learning_rate": 1.1894014962593517e-05,
+      "loss": 0.3452,
+      "step": 32520
+    },
+    {
+      "epoch": 8.11221945137157,
+      "grad_norm": 9.309464454650879,
+      "learning_rate": 1.1891521197007484e-05,
+      "loss": 0.3609,
+      "step": 32530
+    },
+    {
+      "epoch": 8.114713216957606,
+      "grad_norm": 6.361128807067871,
+      "learning_rate": 1.1889027431421447e-05,
+      "loss": 0.3868,
+      "step": 32540
+    },
+    {
+      "epoch": 8.11720698254364,
+      "grad_norm": 6.661703586578369,
+      "learning_rate": 1.1886533665835411e-05,
+      "loss": 0.3508,
+      "step": 32550
+    },
+    {
+      "epoch": 8.119700748129675,
+      "grad_norm": 5.073139667510986,
+      "learning_rate": 1.1884039900249378e-05,
+      "loss": 0.4003,
+      "step": 32560
+    },
+    {
+      "epoch": 8.12219451371571,
+      "grad_norm": 5.183114051818848,
+      "learning_rate": 1.1881546134663342e-05,
+      "loss": 0.3226,
+      "step": 32570
+    },
+    {
+      "epoch": 8.124688279301745,
+      "grad_norm": 6.288895130157471,
+      "learning_rate": 1.1879052369077309e-05,
+      "loss": 0.4334,
+      "step": 32580
+    },
+    {
+      "epoch": 8.12718204488778,
+      "grad_norm": 6.090054988861084,
+      "learning_rate": 1.1876558603491272e-05,
+      "loss": 0.3291,
+      "step": 32590
+    },
+    {
+      "epoch": 8.129675810473815,
+      "grad_norm": 7.359978675842285,
+      "learning_rate": 1.1874064837905237e-05,
+      "loss": 0.4043,
+      "step": 32600
+    },
+    {
+      "epoch": 8.13216957605985,
+      "grad_norm": 9.150673866271973,
+      "learning_rate": 1.1871571072319203e-05,
+      "loss": 0.3534,
+      "step": 32610
+    },
+    {
+      "epoch": 8.134663341645885,
+      "grad_norm": 10.117431640625,
+      "learning_rate": 1.1869077306733168e-05,
+      "loss": 0.3602,
+      "step": 32620
+    },
+    {
+      "epoch": 8.13715710723192,
+      "grad_norm": 4.644679546356201,
+      "learning_rate": 1.1866583541147133e-05,
+      "loss": 0.3498,
+      "step": 32630
+    },
+    {
+      "epoch": 8.139650872817954,
+      "grad_norm": 6.087995529174805,
+      "learning_rate": 1.1864089775561099e-05,
+      "loss": 0.3325,
+      "step": 32640
+    },
+    {
+      "epoch": 8.14214463840399,
+      "grad_norm": 6.6635050773620605,
+      "learning_rate": 1.1861596009975062e-05,
+      "loss": 0.3532,
+      "step": 32650
+    },
+    {
+      "epoch": 8.144638403990024,
+      "grad_norm": 7.93565559387207,
+      "learning_rate": 1.185910224438903e-05,
+      "loss": 0.3327,
+      "step": 32660
+    },
+    {
+      "epoch": 8.147132169576059,
+      "grad_norm": 5.548285961151123,
+      "learning_rate": 1.1856608478802993e-05,
+      "loss": 0.409,
+      "step": 32670
+    },
+    {
+      "epoch": 8.149625935162096,
+      "grad_norm": 6.113910675048828,
+      "learning_rate": 1.1854114713216958e-05,
+      "loss": 0.3934,
+      "step": 32680
+    },
+    {
+      "epoch": 8.15211970074813,
+      "grad_norm": 8.297721862792969,
+      "learning_rate": 1.1851620947630924e-05,
+      "loss": 0.4034,
+      "step": 32690
+    },
+    {
+      "epoch": 8.154613466334165,
+      "grad_norm": 5.073780536651611,
+      "learning_rate": 1.1849127182044889e-05,
+      "loss": 0.326,
+      "step": 32700
+    },
+    {
+      "epoch": 8.1571072319202,
+      "grad_norm": 6.652329921722412,
+      "learning_rate": 1.1846633416458854e-05,
+      "loss": 0.3051,
+      "step": 32710
+    },
+    {
+      "epoch": 8.159600997506235,
+      "grad_norm": 8.228754043579102,
+      "learning_rate": 1.184413965087282e-05,
+      "loss": 0.346,
+      "step": 32720
+    },
+    {
+      "epoch": 8.16209476309227,
+      "grad_norm": 5.232443809509277,
+      "learning_rate": 1.1841645885286783e-05,
+      "loss": 0.3189,
+      "step": 32730
+    },
+    {
+      "epoch": 8.164588528678305,
+      "grad_norm": 6.608932018280029,
+      "learning_rate": 1.183915211970075e-05,
+      "loss": 0.291,
+      "step": 32740
+    },
+    {
+      "epoch": 8.16708229426434,
+      "grad_norm": 6.14049768447876,
+      "learning_rate": 1.1836658354114714e-05,
+      "loss": 0.3777,
+      "step": 32750
+    },
+    {
+      "epoch": 8.169576059850375,
+      "grad_norm": 6.535306453704834,
+      "learning_rate": 1.1834164588528679e-05,
+      "loss": 0.3807,
+      "step": 32760
+    },
+    {
+      "epoch": 8.17206982543641,
+      "grad_norm": 8.01197624206543,
+      "learning_rate": 1.1831670822942644e-05,
+      "loss": 0.3618,
+      "step": 32770
+    },
+    {
+      "epoch": 8.174563591022444,
+      "grad_norm": 6.140148639678955,
+      "learning_rate": 1.182917705735661e-05,
+      "loss": 0.396,
+      "step": 32780
+    },
+    {
+      "epoch": 8.17705735660848,
+      "grad_norm": 9.92581844329834,
+      "learning_rate": 1.1826683291770575e-05,
+      "loss": 0.4123,
+      "step": 32790
+    },
+    {
+      "epoch": 8.179551122194514,
+      "grad_norm": 5.655374526977539,
+      "learning_rate": 1.182418952618454e-05,
+      "loss": 0.3478,
+      "step": 32800
+    },
+    {
+      "epoch": 8.182044887780549,
+      "grad_norm": 6.091732501983643,
+      "learning_rate": 1.1821695760598504e-05,
+      "loss": 0.34,
+      "step": 32810
+    },
+    {
+      "epoch": 8.184538653366584,
+      "grad_norm": 6.486967086791992,
+      "learning_rate": 1.1819201995012471e-05,
+      "loss": 0.3117,
+      "step": 32820
+    },
+    {
+      "epoch": 8.187032418952619,
+      "grad_norm": 6.588410377502441,
+      "learning_rate": 1.1816708229426434e-05,
+      "loss": 0.3123,
+      "step": 32830
+    },
+    {
+      "epoch": 8.189526184538654,
+      "grad_norm": 5.499619007110596,
+      "learning_rate": 1.1814214463840401e-05,
+      "loss": 0.3842,
+      "step": 32840
+    },
+    {
+      "epoch": 8.192019950124688,
+      "grad_norm": 11.53024673461914,
+      "learning_rate": 1.1811720698254365e-05,
+      "loss": 0.437,
+      "step": 32850
+    },
+    {
+      "epoch": 8.194513715710723,
+      "grad_norm": 9.12915325164795,
+      "learning_rate": 1.1809226932668329e-05,
+      "loss": 0.3859,
+      "step": 32860
+    },
+    {
+      "epoch": 8.197007481296758,
+      "grad_norm": 6.353451728820801,
+      "learning_rate": 1.1806733167082296e-05,
+      "loss": 0.3672,
+      "step": 32870
+    },
+    {
+      "epoch": 8.199501246882793,
+      "grad_norm": 8.243021965026855,
+      "learning_rate": 1.1804239401496261e-05,
+      "loss": 0.351,
+      "step": 32880
+    },
+    {
+      "epoch": 8.201995012468828,
+      "grad_norm": 10.141286849975586,
+      "learning_rate": 1.1801745635910225e-05,
+      "loss": 0.3614,
+      "step": 32890
+    },
+    {
+      "epoch": 8.204488778054863,
+      "grad_norm": 6.647899627685547,
+      "learning_rate": 1.1799251870324192e-05,
+      "loss": 0.3553,
+      "step": 32900
+    },
+    {
+      "epoch": 8.206982543640898,
+      "grad_norm": 8.836320877075195,
+      "learning_rate": 1.1796758104738155e-05,
+      "loss": 0.3578,
+      "step": 32910
+    },
+    {
+      "epoch": 8.209476309226932,
+      "grad_norm": 5.279130458831787,
+      "learning_rate": 1.1794264339152122e-05,
+      "loss": 0.3746,
+      "step": 32920
+    },
+    {
+      "epoch": 8.211970074812967,
+      "grad_norm": 8.758357048034668,
+      "learning_rate": 1.1791770573566086e-05,
+      "loss": 0.3627,
+      "step": 32930
+    },
+    {
+      "epoch": 8.214463840399002,
+      "grad_norm": 7.450492858886719,
+      "learning_rate": 1.178927680798005e-05,
+      "loss": 0.3138,
+      "step": 32940
+    },
+    {
+      "epoch": 8.216957605985037,
+      "grad_norm": 7.956082820892334,
+      "learning_rate": 1.1786783042394016e-05,
+      "loss": 0.3486,
+      "step": 32950
+    },
+    {
+      "epoch": 8.219451371571072,
+      "grad_norm": 6.743360996246338,
+      "learning_rate": 1.178428927680798e-05,
+      "loss": 0.3642,
+      "step": 32960
+    },
+    {
+      "epoch": 8.221945137157107,
+      "grad_norm": 8.854829788208008,
+      "learning_rate": 1.1781795511221945e-05,
+      "loss": 0.3638,
+      "step": 32970
+    },
+    {
+      "epoch": 8.224438902743142,
+      "grad_norm": 8.033832550048828,
+      "learning_rate": 1.177930174563591e-05,
+      "loss": 0.3792,
+      "step": 32980
+    },
+    {
+      "epoch": 8.226932668329177,
+      "grad_norm": 3.877129554748535,
+      "learning_rate": 1.1776807980049876e-05,
+      "loss": 0.3361,
+      "step": 32990
+    },
+    {
+      "epoch": 8.229426433915211,
+      "grad_norm": 7.193696975708008,
+      "learning_rate": 1.1774314214463843e-05,
+      "loss": 0.3768,
+      "step": 33000
+    },
+    {
+      "epoch": 8.231920199501246,
+      "grad_norm": 6.511844635009766,
+      "learning_rate": 1.1771820448877807e-05,
+      "loss": 0.3266,
+      "step": 33010
+    },
+    {
+      "epoch": 8.234413965087281,
+      "grad_norm": 7.392646789550781,
+      "learning_rate": 1.176932668329177e-05,
+      "loss": 0.3634,
+      "step": 33020
+    },
+    {
+      "epoch": 8.236907730673316,
+      "grad_norm": 7.599452495574951,
+      "learning_rate": 1.1766832917705737e-05,
+      "loss": 0.4454,
+      "step": 33030
+    },
+    {
+      "epoch": 8.239401496259351,
+      "grad_norm": 8.29662036895752,
+      "learning_rate": 1.1764339152119701e-05,
+      "loss": 0.3168,
+      "step": 33040
+    },
+    {
+      "epoch": 8.241895261845386,
+      "grad_norm": 9.465046882629395,
+      "learning_rate": 1.1761845386533666e-05,
+      "loss": 0.4269,
+      "step": 33050
+    },
+    {
+      "epoch": 8.24438902743142,
+      "grad_norm": 5.553979396820068,
+      "learning_rate": 1.1759351620947632e-05,
+      "loss": 0.3853,
+      "step": 33060
+    },
+    {
+      "epoch": 8.246882793017456,
+      "grad_norm": 10.552316665649414,
+      "learning_rate": 1.1756857855361597e-05,
+      "loss": 0.3968,
+      "step": 33070
+    },
+    {
+      "epoch": 8.24937655860349,
+      "grad_norm": 6.383450984954834,
+      "learning_rate": 1.1754364089775562e-05,
+      "loss": 0.3565,
+      "step": 33080
+    },
+    {
+      "epoch": 8.251870324189527,
+      "grad_norm": 6.994113922119141,
+      "learning_rate": 1.1751870324189527e-05,
+      "loss": 0.3601,
+      "step": 33090
+    },
+    {
+      "epoch": 8.254364089775562,
+      "grad_norm": 5.046513080596924,
+      "learning_rate": 1.1749376558603491e-05,
+      "loss": 0.3747,
+      "step": 33100
+    },
+    {
+      "epoch": 8.256857855361597,
+      "grad_norm": 10.145176887512207,
+      "learning_rate": 1.1746882793017458e-05,
+      "loss": 0.4286,
+      "step": 33110
+    },
+    {
+      "epoch": 8.259351620947632,
+      "grad_norm": 8.742334365844727,
+      "learning_rate": 1.1744389027431422e-05,
+      "loss": 0.3592,
+      "step": 33120
+    },
+    {
+      "epoch": 8.261845386533667,
+      "grad_norm": 7.104617595672607,
+      "learning_rate": 1.1741895261845389e-05,
+      "loss": 0.3678,
+      "step": 33130
+    },
+    {
+      "epoch": 8.264339152119701,
+      "grad_norm": 7.787363529205322,
+      "learning_rate": 1.1739401496259352e-05,
+      "loss": 0.3511,
+      "step": 33140
+    },
+    {
+      "epoch": 8.266832917705736,
+      "grad_norm": 5.79642915725708,
+      "learning_rate": 1.1736907730673318e-05,
+      "loss": 0.3319,
+      "step": 33150
+    },
+    {
+      "epoch": 8.269326683291771,
+      "grad_norm": 7.905724048614502,
+      "learning_rate": 1.1734413965087283e-05,
+      "loss": 0.3393,
+      "step": 33160
+    },
+    {
+      "epoch": 8.271820448877806,
+      "grad_norm": 9.876928329467773,
+      "learning_rate": 1.1731920199501248e-05,
+      "loss": 0.3568,
+      "step": 33170
+    },
+    {
+      "epoch": 8.27431421446384,
+      "grad_norm": 9.139424324035645,
+      "learning_rate": 1.1729426433915212e-05,
+      "loss": 0.4312,
+      "step": 33180
+    },
+    {
+      "epoch": 8.276807980049876,
+      "grad_norm": 4.746652126312256,
+      "learning_rate": 1.1726932668329179e-05,
+      "loss": 0.3464,
+      "step": 33190
+    },
+    {
+      "epoch": 8.27930174563591,
+      "grad_norm": 7.954774379730225,
+      "learning_rate": 1.1724438902743142e-05,
+      "loss": 0.328,
+      "step": 33200
+    },
+    {
+      "epoch": 8.281795511221945,
+      "grad_norm": 8.069437980651855,
+      "learning_rate": 1.172194513715711e-05,
+      "loss": 0.4912,
+      "step": 33210
+    },
+    {
+      "epoch": 8.28428927680798,
+      "grad_norm": 5.376629829406738,
+      "learning_rate": 1.1719451371571073e-05,
+      "loss": 0.339,
+      "step": 33220
+    },
+    {
+      "epoch": 8.286783042394015,
+      "grad_norm": 5.5280632972717285,
+      "learning_rate": 1.1716957605985038e-05,
+      "loss": 0.3914,
+      "step": 33230
+    },
+    {
+      "epoch": 8.28927680798005,
+      "grad_norm": 6.899628162384033,
+      "learning_rate": 1.1714463840399004e-05,
+      "loss": 0.362,
+      "step": 33240
+    },
+    {
+      "epoch": 8.291770573566085,
+      "grad_norm": 5.153944969177246,
+      "learning_rate": 1.1711970074812969e-05,
+      "loss": 0.3784,
+      "step": 33250
+    },
+    {
+      "epoch": 8.29426433915212,
+      "grad_norm": 5.902993679046631,
+      "learning_rate": 1.1709476309226933e-05,
+      "loss": 0.3422,
+      "step": 33260
+    },
+    {
+      "epoch": 8.296758104738155,
+      "grad_norm": 5.776882648468018,
+      "learning_rate": 1.17069825436409e-05,
+      "loss": 0.3242,
+      "step": 33270
+    },
+    {
+      "epoch": 8.29925187032419,
+      "grad_norm": 6.157176494598389,
+      "learning_rate": 1.1704488778054863e-05,
+      "loss": 0.3903,
+      "step": 33280
+    },
+    {
+      "epoch": 8.301745635910224,
+      "grad_norm": 9.823055267333984,
+      "learning_rate": 1.170199501246883e-05,
+      "loss": 0.3959,
+      "step": 33290
+    },
+    {
+      "epoch": 8.30423940149626,
+      "grad_norm": 8.448189735412598,
+      "learning_rate": 1.1699501246882794e-05,
+      "loss": 0.3757,
+      "step": 33300
+    },
+    {
+      "epoch": 8.306733167082294,
+      "grad_norm": 8.766897201538086,
+      "learning_rate": 1.1697007481296757e-05,
+      "loss": 0.3782,
+      "step": 33310
+    },
+    {
+      "epoch": 8.309226932668329,
+      "grad_norm": 5.299570560455322,
+      "learning_rate": 1.1694513715710724e-05,
+      "loss": 0.3758,
+      "step": 33320
+    },
+    {
+      "epoch": 8.311720698254364,
+      "grad_norm": 5.184985637664795,
+      "learning_rate": 1.1692019950124688e-05,
+      "loss": 0.3417,
+      "step": 33330
+    },
+    {
+      "epoch": 8.314214463840399,
+      "grad_norm": 6.362977981567383,
+      "learning_rate": 1.1689526184538655e-05,
+      "loss": 0.341,
+      "step": 33340
+    },
+    {
+      "epoch": 8.316708229426434,
+      "grad_norm": 7.140589714050293,
+      "learning_rate": 1.168703241895262e-05,
+      "loss": 0.3829,
+      "step": 33350
+    },
+    {
+      "epoch": 8.319201995012468,
+      "grad_norm": 7.842562675476074,
+      "learning_rate": 1.1684538653366584e-05,
+      "loss": 0.293,
+      "step": 33360
+    },
+    {
+      "epoch": 8.321695760598503,
+      "grad_norm": 7.240875244140625,
+      "learning_rate": 1.1682044887780551e-05,
+      "loss": 0.4359,
+      "step": 33370
+    },
+    {
+      "epoch": 8.324189526184538,
+      "grad_norm": 10.326569557189941,
+      "learning_rate": 1.1679551122194515e-05,
+      "loss": 0.342,
+      "step": 33380
+    },
+    {
+      "epoch": 8.326683291770573,
+      "grad_norm": 8.610823631286621,
+      "learning_rate": 1.1677057356608478e-05,
+      "loss": 0.3688,
+      "step": 33390
+    },
+    {
+      "epoch": 8.329177057356608,
+      "grad_norm": 7.092180252075195,
+      "learning_rate": 1.1674563591022445e-05,
+      "loss": 0.3506,
+      "step": 33400
+    },
+    {
+      "epoch": 8.331670822942643,
+      "grad_norm": 6.74815559387207,
+      "learning_rate": 1.1672069825436409e-05,
+      "loss": 0.3515,
+      "step": 33410
+    },
+    {
+      "epoch": 8.334164588528678,
+      "grad_norm": 6.766687393188477,
+      "learning_rate": 1.1669576059850376e-05,
+      "loss": 0.3856,
+      "step": 33420
+    },
+    {
+      "epoch": 8.336658354114713,
+      "grad_norm": 7.77371883392334,
+      "learning_rate": 1.166708229426434e-05,
+      "loss": 0.3606,
+      "step": 33430
+    },
+    {
+      "epoch": 8.339152119700747,
+      "grad_norm": 8.904953956604004,
+      "learning_rate": 1.1664588528678305e-05,
+      "loss": 0.4668,
+      "step": 33440
+    },
+    {
+      "epoch": 8.341645885286782,
+      "grad_norm": 4.797825336456299,
+      "learning_rate": 1.166209476309227e-05,
+      "loss": 0.363,
+      "step": 33450
+    },
+    {
+      "epoch": 8.344139650872817,
+      "grad_norm": 9.200943946838379,
+      "learning_rate": 1.1659600997506235e-05,
+      "loss": 0.3227,
+      "step": 33460
+    },
+    {
+      "epoch": 8.346633416458852,
+      "grad_norm": 6.8330888748168945,
+      "learning_rate": 1.1657107231920199e-05,
+      "loss": 0.3053,
+      "step": 33470
+    },
+    {
+      "epoch": 8.349127182044889,
+      "grad_norm": 8.612521171569824,
+      "learning_rate": 1.1654613466334166e-05,
+      "loss": 0.3805,
+      "step": 33480
+    },
+    {
+      "epoch": 8.351620947630924,
+      "grad_norm": 6.210586071014404,
+      "learning_rate": 1.165211970074813e-05,
+      "loss": 0.396,
+      "step": 33490
+    },
+    {
+      "epoch": 8.354114713216958,
+      "grad_norm": 6.151739597320557,
+      "learning_rate": 1.1649625935162097e-05,
+      "loss": 0.4046,
+      "step": 33500
+    },
+    {
+      "epoch": 8.356608478802993,
+      "grad_norm": 7.85900354385376,
+      "learning_rate": 1.164713216957606e-05,
+      "loss": 0.3673,
+      "step": 33510
+    },
+    {
+      "epoch": 8.359102244389028,
+      "grad_norm": 6.287712097167969,
+      "learning_rate": 1.1644638403990026e-05,
+      "loss": 0.3073,
+      "step": 33520
+    },
+    {
+      "epoch": 8.361596009975063,
+      "grad_norm": 6.707120418548584,
+      "learning_rate": 1.1642144638403991e-05,
+      "loss": 0.3577,
+      "step": 33530
+    },
+    {
+      "epoch": 8.364089775561098,
+      "grad_norm": 6.422928810119629,
+      "learning_rate": 1.1639650872817956e-05,
+      "loss": 0.3452,
+      "step": 33540
+    },
+    {
+      "epoch": 8.366583541147133,
+      "grad_norm": 7.4128031730651855,
+      "learning_rate": 1.163715710723192e-05,
+      "loss": 0.4418,
+      "step": 33550
+    },
+    {
+      "epoch": 8.369077306733168,
+      "grad_norm": 7.202417850494385,
+      "learning_rate": 1.1634663341645887e-05,
+      "loss": 0.4487,
+      "step": 33560
+    },
+    {
+      "epoch": 8.371571072319203,
+      "grad_norm": 6.886559963226318,
+      "learning_rate": 1.163216957605985e-05,
+      "loss": 0.399,
+      "step": 33570
+    },
+    {
+      "epoch": 8.374064837905237,
+      "grad_norm": 6.747541427612305,
+      "learning_rate": 1.1629675810473817e-05,
+      "loss": 0.345,
+      "step": 33580
+    },
+    {
+      "epoch": 8.376558603491272,
+      "grad_norm": 5.691372871398926,
+      "learning_rate": 1.1627182044887781e-05,
+      "loss": 0.3636,
+      "step": 33590
+    },
+    {
+      "epoch": 8.379052369077307,
+      "grad_norm": 7.294488906860352,
+      "learning_rate": 1.1624688279301746e-05,
+      "loss": 0.3643,
+      "step": 33600
+    },
+    {
+      "epoch": 8.381546134663342,
+      "grad_norm": 6.717970371246338,
+      "learning_rate": 1.1622194513715712e-05,
+      "loss": 0.3814,
+      "step": 33610
+    },
+    {
+      "epoch": 8.384039900249377,
+      "grad_norm": 4.82089376449585,
+      "learning_rate": 1.1619700748129677e-05,
+      "loss": 0.3283,
+      "step": 33620
+    },
+    {
+      "epoch": 8.386533665835412,
+      "grad_norm": 10.288949966430664,
+      "learning_rate": 1.1617206982543642e-05,
+      "loss": 0.3937,
+      "step": 33630
+    },
+    {
+      "epoch": 8.389027431421447,
+      "grad_norm": 8.169456481933594,
+      "learning_rate": 1.1614713216957608e-05,
+      "loss": 0.3911,
+      "step": 33640
+    },
+    {
+      "epoch": 8.391521197007481,
+      "grad_norm": 4.798164367675781,
+      "learning_rate": 1.1612219451371571e-05,
+      "loss": 0.333,
+      "step": 33650
+    },
+    {
+      "epoch": 8.394014962593516,
+      "grad_norm": 8.796186447143555,
+      "learning_rate": 1.1609725685785538e-05,
+      "loss": 0.3268,
+      "step": 33660
+    },
+    {
+      "epoch": 8.396508728179551,
+      "grad_norm": 6.837301731109619,
+      "learning_rate": 1.1607231920199502e-05,
+      "loss": 0.3797,
+      "step": 33670
+    },
+    {
+      "epoch": 8.399002493765586,
+      "grad_norm": 4.7810282707214355,
+      "learning_rate": 1.1604738154613465e-05,
+      "loss": 0.3541,
+      "step": 33680
+    },
+    {
+      "epoch": 8.401496259351621,
+      "grad_norm": 7.039614200592041,
+      "learning_rate": 1.1602244389027432e-05,
+      "loss": 0.4047,
+      "step": 33690
+    },
+    {
+      "epoch": 8.403990024937656,
+      "grad_norm": 7.48544454574585,
+      "learning_rate": 1.1599750623441398e-05,
+      "loss": 0.3616,
+      "step": 33700
+    },
+    {
+      "epoch": 8.40648379052369,
+      "grad_norm": 12.1786470413208,
+      "learning_rate": 1.1597256857855363e-05,
+      "loss": 0.4326,
+      "step": 33710
+    },
+    {
+      "epoch": 8.408977556109726,
+      "grad_norm": 6.31995964050293,
+      "learning_rate": 1.1594763092269328e-05,
+      "loss": 0.3818,
+      "step": 33720
+    },
+    {
+      "epoch": 8.41147132169576,
+      "grad_norm": 9.140199661254883,
+      "learning_rate": 1.1592269326683292e-05,
+      "loss": 0.3978,
+      "step": 33730
+    },
+    {
+      "epoch": 8.413965087281795,
+      "grad_norm": 5.953067779541016,
+      "learning_rate": 1.1589775561097259e-05,
+      "loss": 0.3253,
+      "step": 33740
+    },
+    {
+      "epoch": 8.41645885286783,
+      "grad_norm": 6.9121246337890625,
+      "learning_rate": 1.1587281795511223e-05,
+      "loss": 0.3492,
+      "step": 33750
+    },
+    {
+      "epoch": 8.418952618453865,
+      "grad_norm": 6.274780750274658,
+      "learning_rate": 1.1584788029925186e-05,
+      "loss": 0.4546,
+      "step": 33760
+    },
+    {
+      "epoch": 8.4214463840399,
+      "grad_norm": 7.781808376312256,
+      "learning_rate": 1.1582294264339153e-05,
+      "loss": 0.3125,
+      "step": 33770
+    },
+    {
+      "epoch": 8.423940149625935,
+      "grad_norm": 9.689964294433594,
+      "learning_rate": 1.1579800498753117e-05,
+      "loss": 0.3515,
+      "step": 33780
+    },
+    {
+      "epoch": 8.42643391521197,
+      "grad_norm": 5.320454120635986,
+      "learning_rate": 1.1577306733167084e-05,
+      "loss": 0.2771,
+      "step": 33790
+    },
+    {
+      "epoch": 8.428927680798004,
+      "grad_norm": 7.566685199737549,
+      "learning_rate": 1.1574812967581047e-05,
+      "loss": 0.3667,
+      "step": 33800
+    },
+    {
+      "epoch": 8.43142144638404,
+      "grad_norm": 8.28944206237793,
+      "learning_rate": 1.1572319201995013e-05,
+      "loss": 0.3956,
+      "step": 33810
+    },
+    {
+      "epoch": 8.433915211970074,
+      "grad_norm": 7.253947734832764,
+      "learning_rate": 1.156982543640898e-05,
+      "loss": 0.3837,
+      "step": 33820
+    },
+    {
+      "epoch": 8.436408977556109,
+      "grad_norm": 9.780487060546875,
+      "learning_rate": 1.1567331670822943e-05,
+      "loss": 0.4626,
+      "step": 33830
+    },
+    {
+      "epoch": 8.438902743142144,
+      "grad_norm": 6.923801898956299,
+      "learning_rate": 1.156483790523691e-05,
+      "loss": 0.3675,
+      "step": 33840
+    },
+    {
+      "epoch": 8.441396508728179,
+      "grad_norm": 7.305504322052002,
+      "learning_rate": 1.1562344139650874e-05,
+      "loss": 0.3247,
+      "step": 33850
+    },
+    {
+      "epoch": 8.443890274314214,
+      "grad_norm": 5.790426731109619,
+      "learning_rate": 1.1559850374064838e-05,
+      "loss": 0.3559,
+      "step": 33860
+    },
+    {
+      "epoch": 8.446384039900249,
+      "grad_norm": 5.2143731117248535,
+      "learning_rate": 1.1557356608478805e-05,
+      "loss": 0.3547,
+      "step": 33870
+    },
+    {
+      "epoch": 8.448877805486283,
+      "grad_norm": 7.238703727722168,
+      "learning_rate": 1.1554862842892768e-05,
+      "loss": 0.3329,
+      "step": 33880
+    },
+    {
+      "epoch": 8.451371571072318,
+      "grad_norm": 5.136308193206787,
+      "learning_rate": 1.1552369077306734e-05,
+      "loss": 0.3662,
+      "step": 33890
+    },
+    {
+      "epoch": 8.453865336658355,
+      "grad_norm": 7.3576836585998535,
+      "learning_rate": 1.1549875311720699e-05,
+      "loss": 0.3306,
+      "step": 33900
+    },
+    {
+      "epoch": 8.45635910224439,
+      "grad_norm": 8.722187042236328,
+      "learning_rate": 1.1547381546134664e-05,
+      "loss": 0.3519,
+      "step": 33910
+    },
+    {
+      "epoch": 8.458852867830425,
+      "grad_norm": 8.820263862609863,
+      "learning_rate": 1.154488778054863e-05,
+      "loss": 0.4284,
+      "step": 33920
+    },
+    {
+      "epoch": 8.46134663341646,
+      "grad_norm": 8.29401683807373,
+      "learning_rate": 1.1542394014962595e-05,
+      "loss": 0.3871,
+      "step": 33930
+    },
+    {
+      "epoch": 8.463840399002494,
+      "grad_norm": 6.295302867889404,
+      "learning_rate": 1.1539900249376558e-05,
+      "loss": 0.3196,
+      "step": 33940
+    },
+    {
+      "epoch": 8.46633416458853,
+      "grad_norm": 7.778040885925293,
+      "learning_rate": 1.1537406483790525e-05,
+      "loss": 0.348,
+      "step": 33950
+    },
+    {
+      "epoch": 8.468827930174564,
+      "grad_norm": 9.188948631286621,
+      "learning_rate": 1.1534912718204489e-05,
+      "loss": 0.3504,
+      "step": 33960
+    },
+    {
+      "epoch": 8.471321695760599,
+      "grad_norm": 4.633731842041016,
+      "learning_rate": 1.1532418952618454e-05,
+      "loss": 0.3604,
+      "step": 33970
+    },
+    {
+      "epoch": 8.473815461346634,
+      "grad_norm": 7.2224884033203125,
+      "learning_rate": 1.152992518703242e-05,
+      "loss": 0.3979,
+      "step": 33980
+    },
+    {
+      "epoch": 8.476309226932669,
+      "grad_norm": 7.150679588317871,
+      "learning_rate": 1.1527431421446385e-05,
+      "loss": 0.3612,
+      "step": 33990
+    },
+    {
+      "epoch": 8.478802992518704,
+      "grad_norm": 6.690948486328125,
+      "learning_rate": 1.152493765586035e-05,
+      "loss": 0.3037,
+      "step": 34000
+    },
+    {
+      "epoch": 8.481296758104738,
+      "grad_norm": 5.986937046051025,
+      "learning_rate": 1.1522443890274316e-05,
+      "loss": 0.4867,
+      "step": 34010
+    },
+    {
+      "epoch": 8.483790523690773,
+      "grad_norm": 6.088142395019531,
+      "learning_rate": 1.151995012468828e-05,
+      "loss": 0.3606,
+      "step": 34020
+    },
+    {
+      "epoch": 8.486284289276808,
+      "grad_norm": 6.499013423919678,
+      "learning_rate": 1.1517456359102246e-05,
+      "loss": 0.3891,
+      "step": 34030
+    },
+    {
+      "epoch": 8.488778054862843,
+      "grad_norm": 9.472525596618652,
+      "learning_rate": 1.151496259351621e-05,
+      "loss": 0.3548,
+      "step": 34040
+    },
+    {
+      "epoch": 8.491271820448878,
+      "grad_norm": 7.18366813659668,
+      "learning_rate": 1.1512468827930175e-05,
+      "loss": 0.3939,
+      "step": 34050
+    },
+    {
+      "epoch": 8.493765586034913,
+      "grad_norm": 6.115909099578857,
+      "learning_rate": 1.150997506234414e-05,
+      "loss": 0.4296,
+      "step": 34060
+    },
+    {
+      "epoch": 8.496259351620948,
+      "grad_norm": 7.923805236816406,
+      "learning_rate": 1.1507481296758106e-05,
+      "loss": 0.389,
+      "step": 34070
+    },
+    {
+      "epoch": 8.498753117206983,
+      "grad_norm": 6.619093418121338,
+      "learning_rate": 1.1504987531172071e-05,
+      "loss": 0.3364,
+      "step": 34080
+    },
+    {
+      "epoch": 8.501246882793017,
+      "grad_norm": 4.969188213348389,
+      "learning_rate": 1.1502493765586036e-05,
+      "loss": 0.3985,
+      "step": 34090
+    },
+    {
+      "epoch": 8.503740648379052,
+      "grad_norm": 5.744391441345215,
+      "learning_rate": 1.15e-05,
+      "loss": 0.4074,
+      "step": 34100
+    },
+    {
+      "epoch": 8.506234413965087,
+      "grad_norm": 6.1294755935668945,
+      "learning_rate": 1.1497506234413967e-05,
+      "loss": 0.3872,
+      "step": 34110
+    },
+    {
+      "epoch": 8.508728179551122,
+      "grad_norm": 7.260706901550293,
+      "learning_rate": 1.149501246882793e-05,
+      "loss": 0.3354,
+      "step": 34120
+    },
+    {
+      "epoch": 8.511221945137157,
+      "grad_norm": 5.858883380889893,
+      "learning_rate": 1.1492518703241898e-05,
+      "loss": 0.3727,
+      "step": 34130
+    },
+    {
+      "epoch": 8.513715710723192,
+      "grad_norm": 5.901925086975098,
+      "learning_rate": 1.1490024937655861e-05,
+      "loss": 0.3703,
+      "step": 34140
+    },
+    {
+      "epoch": 8.516209476309227,
+      "grad_norm": 6.566878318786621,
+      "learning_rate": 1.1487531172069825e-05,
+      "loss": 0.3385,
+      "step": 34150
+    },
+    {
+      "epoch": 8.518703241895262,
+      "grad_norm": 6.557463645935059,
+      "learning_rate": 1.1485037406483792e-05,
+      "loss": 0.2862,
+      "step": 34160
+    },
+    {
+      "epoch": 8.521197007481296,
+      "grad_norm": 8.497570037841797,
+      "learning_rate": 1.1482543640897757e-05,
+      "loss": 0.3043,
+      "step": 34170
+    },
+    {
+      "epoch": 8.523690773067331,
+      "grad_norm": 6.737511157989502,
+      "learning_rate": 1.148004987531172e-05,
+      "loss": 0.3957,
+      "step": 34180
+    },
+    {
+      "epoch": 8.526184538653366,
+      "grad_norm": 7.414823055267334,
+      "learning_rate": 1.1477556109725688e-05,
+      "loss": 0.3243,
+      "step": 34190
+    },
+    {
+      "epoch": 8.528678304239401,
+      "grad_norm": 6.70835542678833,
+      "learning_rate": 1.1475311720698254e-05,
+      "loss": 0.2703,
+      "step": 34200
+    },
+    {
+      "epoch": 8.531172069825436,
+      "grad_norm": 6.923582077026367,
+      "learning_rate": 1.1472817955112221e-05,
+      "loss": 0.3621,
+      "step": 34210
+    },
+    {
+      "epoch": 8.53366583541147,
+      "grad_norm": 3.6266000270843506,
+      "learning_rate": 1.1470324189526185e-05,
+      "loss": 0.2875,
+      "step": 34220
+    },
+    {
+      "epoch": 8.536159600997506,
+      "grad_norm": 7.4206156730651855,
+      "learning_rate": 1.1467830423940152e-05,
+      "loss": 0.3646,
+      "step": 34230
+    },
+    {
+      "epoch": 8.53865336658354,
+      "grad_norm": 5.679503440856934,
+      "learning_rate": 1.1465336658354116e-05,
+      "loss": 0.355,
+      "step": 34240
+    },
+    {
+      "epoch": 8.541147132169575,
+      "grad_norm": 8.173449516296387,
+      "learning_rate": 1.1462842892768081e-05,
+      "loss": 0.3537,
+      "step": 34250
+    },
+    {
+      "epoch": 8.54364089775561,
+      "grad_norm": 7.642554759979248,
+      "learning_rate": 1.1460349127182046e-05,
+      "loss": 0.3652,
+      "step": 34260
+    },
+    {
+      "epoch": 8.546134663341645,
+      "grad_norm": 10.153780937194824,
+      "learning_rate": 1.1457855361596012e-05,
+      "loss": 0.4083,
+      "step": 34270
+    },
+    {
+      "epoch": 8.548628428927682,
+      "grad_norm": 7.6003289222717285,
+      "learning_rate": 1.1455361596009975e-05,
+      "loss": 0.4111,
+      "step": 34280
+    },
+    {
+      "epoch": 8.551122194513717,
+      "grad_norm": 8.038273811340332,
+      "learning_rate": 1.1452867830423942e-05,
+      "loss": 0.3452,
+      "step": 34290
+    },
+    {
+      "epoch": 8.553615960099751,
+      "grad_norm": 7.191328048706055,
+      "learning_rate": 1.1450374064837906e-05,
+      "loss": 0.3575,
+      "step": 34300
+    },
+    {
+      "epoch": 8.556109725685786,
+      "grad_norm": 8.192909240722656,
+      "learning_rate": 1.1447880299251873e-05,
+      "loss": 0.3853,
+      "step": 34310
+    },
+    {
+      "epoch": 8.558603491271821,
+      "grad_norm": 7.0611677169799805,
+      "learning_rate": 1.1445386533665836e-05,
+      "loss": 0.3626,
+      "step": 34320
+    },
+    {
+      "epoch": 8.561097256857856,
+      "grad_norm": 6.423704624176025,
+      "learning_rate": 1.14428927680798e-05,
+      "loss": 0.3413,
+      "step": 34330
+    },
+    {
+      "epoch": 8.563591022443891,
+      "grad_norm": 7.521974086761475,
+      "learning_rate": 1.1440399002493767e-05,
+      "loss": 0.3469,
+      "step": 34340
+    },
+    {
+      "epoch": 8.566084788029926,
+      "grad_norm": 4.919178485870361,
+      "learning_rate": 1.143790523690773e-05,
+      "loss": 0.3544,
+      "step": 34350
+    },
+    {
+      "epoch": 8.56857855361596,
+      "grad_norm": 6.528305530548096,
+      "learning_rate": 1.1435411471321696e-05,
+      "loss": 0.3007,
+      "step": 34360
+    },
+    {
+      "epoch": 8.571072319201996,
+      "grad_norm": 6.000070571899414,
+      "learning_rate": 1.1432917705735663e-05,
+      "loss": 0.3419,
+      "step": 34370
+    },
+    {
+      "epoch": 8.57356608478803,
+      "grad_norm": 6.5107245445251465,
+      "learning_rate": 1.1430423940149627e-05,
+      "loss": 0.3416,
+      "step": 34380
+    },
+    {
+      "epoch": 8.576059850374065,
+      "grad_norm": 8.171274185180664,
+      "learning_rate": 1.1427930174563594e-05,
+      "loss": 0.4224,
+      "step": 34390
+    },
+    {
+      "epoch": 8.5785536159601,
+      "grad_norm": 7.173635959625244,
+      "learning_rate": 1.1425436408977557e-05,
+      "loss": 0.3584,
+      "step": 34400
+    },
+    {
+      "epoch": 8.581047381546135,
+      "grad_norm": 6.059718608856201,
+      "learning_rate": 1.142294264339152e-05,
+      "loss": 0.3048,
+      "step": 34410
+    },
+    {
+      "epoch": 8.58354114713217,
+      "grad_norm": 5.417482376098633,
+      "learning_rate": 1.1420448877805488e-05,
+      "loss": 0.3443,
+      "step": 34420
+    },
+    {
+      "epoch": 8.586034912718205,
+      "grad_norm": 8.786724090576172,
+      "learning_rate": 1.1417955112219451e-05,
+      "loss": 0.4161,
+      "step": 34430
+    },
+    {
+      "epoch": 8.58852867830424,
+      "grad_norm": 8.032851219177246,
+      "learning_rate": 1.1415461346633417e-05,
+      "loss": 0.4395,
+      "step": 34440
+    },
+    {
+      "epoch": 8.591022443890274,
+      "grad_norm": 8.335599899291992,
+      "learning_rate": 1.1412967581047382e-05,
+      "loss": 0.3912,
+      "step": 34450
+    },
+    {
+      "epoch": 8.59351620947631,
+      "grad_norm": 8.087762832641602,
+      "learning_rate": 1.1410473815461347e-05,
+      "loss": 0.3549,
+      "step": 34460
+    },
+    {
+      "epoch": 8.596009975062344,
+      "grad_norm": 5.728715896606445,
+      "learning_rate": 1.1407980049875313e-05,
+      "loss": 0.3888,
+      "step": 34470
+    },
+    {
+      "epoch": 8.598503740648379,
+      "grad_norm": 8.289750099182129,
+      "learning_rate": 1.1405486284289278e-05,
+      "loss": 0.3327,
+      "step": 34480
+    },
+    {
+      "epoch": 8.600997506234414,
+      "grad_norm": 11.838362693786621,
+      "learning_rate": 1.1402992518703242e-05,
+      "loss": 0.3676,
+      "step": 34490
+    },
+    {
+      "epoch": 8.603491271820449,
+      "grad_norm": 7.733099937438965,
+      "learning_rate": 1.1400498753117209e-05,
+      "loss": 0.4015,
+      "step": 34500
+    },
+    {
+      "epoch": 8.605985037406484,
+      "grad_norm": 6.227921009063721,
+      "learning_rate": 1.1398004987531172e-05,
+      "loss": 0.3205,
+      "step": 34510
+    },
+    {
+      "epoch": 8.608478802992519,
+      "grad_norm": 10.058944702148438,
+      "learning_rate": 1.139551122194514e-05,
+      "loss": 0.3644,
+      "step": 34520
+    },
+    {
+      "epoch": 8.610972568578553,
+      "grad_norm": 5.577748775482178,
+      "learning_rate": 1.1393017456359103e-05,
+      "loss": 0.3545,
+      "step": 34530
+    },
+    {
+      "epoch": 8.613466334164588,
+      "grad_norm": 6.9346923828125,
+      "learning_rate": 1.1390523690773068e-05,
+      "loss": 0.3203,
+      "step": 34540
+    },
+    {
+      "epoch": 8.615960099750623,
+      "grad_norm": 7.139277935028076,
+      "learning_rate": 1.1388029925187033e-05,
+      "loss": 0.296,
+      "step": 34550
+    },
+    {
+      "epoch": 8.618453865336658,
+      "grad_norm": 6.940938949584961,
+      "learning_rate": 1.1385536159600999e-05,
+      "loss": 0.339,
+      "step": 34560
+    },
+    {
+      "epoch": 8.620947630922693,
+      "grad_norm": 7.233773708343506,
+      "learning_rate": 1.1383042394014962e-05,
+      "loss": 0.3677,
+      "step": 34570
+    },
+    {
+      "epoch": 8.623441396508728,
+      "grad_norm": 11.383943557739258,
+      "learning_rate": 1.138054862842893e-05,
+      "loss": 0.3711,
+      "step": 34580
+    },
+    {
+      "epoch": 8.625935162094763,
+      "grad_norm": 5.546501636505127,
+      "learning_rate": 1.1378054862842893e-05,
+      "loss": 0.4015,
+      "step": 34590
+    },
+    {
+      "epoch": 8.628428927680797,
+      "grad_norm": 8.203598022460938,
+      "learning_rate": 1.137556109725686e-05,
+      "loss": 0.32,
+      "step": 34600
+    },
+    {
+      "epoch": 8.630922693266832,
+      "grad_norm": 8.763114929199219,
+      "learning_rate": 1.1373067331670824e-05,
+      "loss": 0.2881,
+      "step": 34610
+    },
+    {
+      "epoch": 8.633416458852867,
+      "grad_norm": 9.64941120147705,
+      "learning_rate": 1.1370573566084789e-05,
+      "loss": 0.4229,
+      "step": 34620
+    },
+    {
+      "epoch": 8.635910224438902,
+      "grad_norm": 8.64614486694336,
+      "learning_rate": 1.1368079800498754e-05,
+      "loss": 0.3512,
+      "step": 34630
+    },
+    {
+      "epoch": 8.638403990024937,
+      "grad_norm": 9.003388404846191,
+      "learning_rate": 1.136558603491272e-05,
+      "loss": 0.3894,
+      "step": 34640
+    },
+    {
+      "epoch": 8.640897755610972,
+      "grad_norm": 6.2537336349487305,
+      "learning_rate": 1.1363092269326683e-05,
+      "loss": 0.3445,
+      "step": 34650
+    },
+    {
+      "epoch": 8.643391521197007,
+      "grad_norm": 7.811208248138428,
+      "learning_rate": 1.136059850374065e-05,
+      "loss": 0.4709,
+      "step": 34660
+    },
+    {
+      "epoch": 8.645885286783042,
+      "grad_norm": 7.517146587371826,
+      "learning_rate": 1.1358104738154614e-05,
+      "loss": 0.307,
+      "step": 34670
+    },
+    {
+      "epoch": 8.648379052369076,
+      "grad_norm": 7.594816207885742,
+      "learning_rate": 1.135561097256858e-05,
+      "loss": 0.3967,
+      "step": 34680
+    },
+    {
+      "epoch": 8.650872817955111,
+      "grad_norm": 6.5755510330200195,
+      "learning_rate": 1.1353117206982544e-05,
+      "loss": 0.3044,
+      "step": 34690
+    },
+    {
+      "epoch": 8.653366583541148,
+      "grad_norm": 10.061249732971191,
+      "learning_rate": 1.1350623441396508e-05,
+      "loss": 0.4416,
+      "step": 34700
+    },
+    {
+      "epoch": 8.655860349127183,
+      "grad_norm": 7.96301794052124,
+      "learning_rate": 1.1348129675810475e-05,
+      "loss": 0.4184,
+      "step": 34710
+    },
+    {
+      "epoch": 8.658354114713218,
+      "grad_norm": 7.919782638549805,
+      "learning_rate": 1.134563591022444e-05,
+      "loss": 0.3322,
+      "step": 34720
+    },
+    {
+      "epoch": 8.660847880299253,
+      "grad_norm": 8.549415588378906,
+      "learning_rate": 1.1343142144638406e-05,
+      "loss": 0.3588,
+      "step": 34730
+    },
+    {
+      "epoch": 8.663341645885287,
+      "grad_norm": 5.7566070556640625,
+      "learning_rate": 1.1340648379052371e-05,
+      "loss": 0.3391,
+      "step": 34740
+    },
+    {
+      "epoch": 8.665835411471322,
+      "grad_norm": 8.828804016113281,
+      "learning_rate": 1.1338154613466335e-05,
+      "loss": 0.3558,
+      "step": 34750
+    },
+    {
+      "epoch": 8.668329177057357,
+      "grad_norm": 7.40009069442749,
+      "learning_rate": 1.1335660847880302e-05,
+      "loss": 0.3556,
+      "step": 34760
+    },
+    {
+      "epoch": 8.670822942643392,
+      "grad_norm": 7.1716814041137695,
+      "learning_rate": 1.1333167082294265e-05,
+      "loss": 0.3363,
+      "step": 34770
+    },
+    {
+      "epoch": 8.673316708229427,
+      "grad_norm": 8.354803085327148,
+      "learning_rate": 1.1330673316708229e-05,
+      "loss": 0.3615,
+      "step": 34780
+    },
+    {
+      "epoch": 8.675810473815462,
+      "grad_norm": 5.095800399780273,
+      "learning_rate": 1.1328179551122196e-05,
+      "loss": 0.3083,
+      "step": 34790
+    },
+    {
+      "epoch": 8.678304239401497,
+      "grad_norm": 8.354507446289062,
+      "learning_rate": 1.132568578553616e-05,
+      "loss": 0.3613,
+      "step": 34800
+    },
+    {
+      "epoch": 8.680798004987532,
+      "grad_norm": 6.8850555419921875,
+      "learning_rate": 1.1323192019950126e-05,
+      "loss": 0.3627,
+      "step": 34810
+    },
+    {
+      "epoch": 8.683291770573566,
+      "grad_norm": 7.482059001922607,
+      "learning_rate": 1.132069825436409e-05,
+      "loss": 0.3885,
+      "step": 34820
+    },
+    {
+      "epoch": 8.685785536159601,
+      "grad_norm": 8.189881324768066,
+      "learning_rate": 1.1318204488778055e-05,
+      "loss": 0.3067,
+      "step": 34830
+    },
+    {
+      "epoch": 8.688279301745636,
+      "grad_norm": 6.003298759460449,
+      "learning_rate": 1.1315710723192022e-05,
+      "loss": 0.3853,
+      "step": 34840
+    },
+    {
+      "epoch": 8.690773067331671,
+      "grad_norm": 9.855851173400879,
+      "learning_rate": 1.1313216957605986e-05,
+      "loss": 0.426,
+      "step": 34850
+    },
+    {
+      "epoch": 8.693266832917706,
+      "grad_norm": 7.669548988342285,
+      "learning_rate": 1.131072319201995e-05,
+      "loss": 0.3384,
+      "step": 34860
+    },
+    {
+      "epoch": 8.69576059850374,
+      "grad_norm": 7.204586029052734,
+      "learning_rate": 1.1308229426433917e-05,
+      "loss": 0.3972,
+      "step": 34870
+    },
+    {
+      "epoch": 8.698254364089776,
+      "grad_norm": 5.998164176940918,
+      "learning_rate": 1.130573566084788e-05,
+      "loss": 0.3572,
+      "step": 34880
+    },
+    {
+      "epoch": 8.70074812967581,
+      "grad_norm": 7.780731201171875,
+      "learning_rate": 1.1303241895261847e-05,
+      "loss": 0.3469,
+      "step": 34890
+    },
+    {
+      "epoch": 8.703241895261845,
+      "grad_norm": 8.073392868041992,
+      "learning_rate": 1.130074812967581e-05,
+      "loss": 0.3723,
+      "step": 34900
+    },
+    {
+      "epoch": 8.70573566084788,
+      "grad_norm": 5.444243907928467,
+      "learning_rate": 1.1298254364089776e-05,
+      "loss": 0.3219,
+      "step": 34910
+    },
+    {
+      "epoch": 8.708229426433915,
+      "grad_norm": 10.334959030151367,
+      "learning_rate": 1.1295760598503741e-05,
+      "loss": 0.3386,
+      "step": 34920
+    },
+    {
+      "epoch": 8.71072319201995,
+      "grad_norm": 8.497328758239746,
+      "learning_rate": 1.1293266832917707e-05,
+      "loss": 0.3967,
+      "step": 34930
+    },
+    {
+      "epoch": 8.713216957605985,
+      "grad_norm": 6.893999099731445,
+      "learning_rate": 1.129077306733167e-05,
+      "loss": 0.2833,
+      "step": 34940
+    },
+    {
+      "epoch": 8.71571072319202,
+      "grad_norm": 6.12678337097168,
+      "learning_rate": 1.1288279301745637e-05,
+      "loss": 0.3801,
+      "step": 34950
+    },
+    {
+      "epoch": 8.718204488778055,
+      "grad_norm": 4.06450080871582,
+      "learning_rate": 1.1285785536159601e-05,
+      "loss": 0.3047,
+      "step": 34960
+    },
+    {
+      "epoch": 8.72069825436409,
+      "grad_norm": 6.966952323913574,
+      "learning_rate": 1.1283291770573568e-05,
+      "loss": 0.371,
+      "step": 34970
+    },
+    {
+      "epoch": 8.723192019950124,
+      "grad_norm": 6.376713752746582,
+      "learning_rate": 1.1280798004987532e-05,
+      "loss": 0.3878,
+      "step": 34980
+    },
+    {
+      "epoch": 8.72568578553616,
+      "grad_norm": 7.877644062042236,
+      "learning_rate": 1.1278304239401497e-05,
+      "loss": 0.3533,
+      "step": 34990
+    },
+    {
+      "epoch": 8.728179551122194,
+      "grad_norm": 4.689276218414307,
+      "learning_rate": 1.1275810473815462e-05,
+      "loss": 0.3081,
+      "step": 35000
+    },
+    {
+      "epoch": 8.730673316708229,
+      "grad_norm": 8.68629264831543,
+      "learning_rate": 1.1273316708229428e-05,
+      "loss": 0.3049,
+      "step": 35010
+    },
+    {
+      "epoch": 8.733167082294264,
+      "grad_norm": 6.382472991943359,
+      "learning_rate": 1.1270822942643393e-05,
+      "loss": 0.3957,
+      "step": 35020
+    },
+    {
+      "epoch": 8.735660847880299,
+      "grad_norm": 5.842182636260986,
+      "learning_rate": 1.1268329177057358e-05,
+      "loss": 0.3513,
+      "step": 35030
+    },
+    {
+      "epoch": 8.738154613466333,
+      "grad_norm": 5.1576948165893555,
+      "learning_rate": 1.1265835411471322e-05,
+      "loss": 0.3957,
+      "step": 35040
+    },
+    {
+      "epoch": 8.740648379052368,
+      "grad_norm": 12.842966079711914,
+      "learning_rate": 1.1263341645885289e-05,
+      "loss": 0.3318,
+      "step": 35050
+    },
+    {
+      "epoch": 8.743142144638403,
+      "grad_norm": 5.631074905395508,
+      "learning_rate": 1.1260847880299252e-05,
+      "loss": 0.2967,
+      "step": 35060
+    },
+    {
+      "epoch": 8.745635910224438,
+      "grad_norm": 6.541484355926514,
+      "learning_rate": 1.1258354114713218e-05,
+      "loss": 0.3186,
+      "step": 35070
+    },
+    {
+      "epoch": 8.748129675810475,
+      "grad_norm": 5.197242259979248,
+      "learning_rate": 1.1255860349127183e-05,
+      "loss": 0.3606,
+      "step": 35080
+    },
+    {
+      "epoch": 8.75062344139651,
+      "grad_norm": 7.91013240814209,
+      "learning_rate": 1.1253366583541148e-05,
+      "loss": 0.3609,
+      "step": 35090
+    },
+    {
+      "epoch": 8.753117206982544,
+      "grad_norm": 6.427094459533691,
+      "learning_rate": 1.1250872817955114e-05,
+      "loss": 0.4667,
+      "step": 35100
+    },
+    {
+      "epoch": 8.75561097256858,
+      "grad_norm": 5.502357482910156,
+      "learning_rate": 1.1248379052369079e-05,
+      "loss": 0.3279,
+      "step": 35110
+    },
+    {
+      "epoch": 8.758104738154614,
+      "grad_norm": 6.215767860412598,
+      "learning_rate": 1.1245885286783043e-05,
+      "loss": 0.3371,
+      "step": 35120
+    },
+    {
+      "epoch": 8.760598503740649,
+      "grad_norm": 6.710012435913086,
+      "learning_rate": 1.124339152119701e-05,
+      "loss": 0.3477,
+      "step": 35130
+    },
+    {
+      "epoch": 8.763092269326684,
+      "grad_norm": 7.127229690551758,
+      "learning_rate": 1.1240897755610973e-05,
+      "loss": 0.3687,
+      "step": 35140
+    },
+    {
+      "epoch": 8.765586034912719,
+      "grad_norm": 7.16987419128418,
+      "learning_rate": 1.1238403990024937e-05,
+      "loss": 0.3494,
+      "step": 35150
+    },
+    {
+      "epoch": 8.768079800498754,
+      "grad_norm": 8.195234298706055,
+      "learning_rate": 1.1235910224438904e-05,
+      "loss": 0.331,
+      "step": 35160
+    },
+    {
+      "epoch": 8.770573566084789,
+      "grad_norm": 5.69235372543335,
+      "learning_rate": 1.1233416458852869e-05,
+      "loss": 0.3815,
+      "step": 35170
+    },
+    {
+      "epoch": 8.773067331670823,
+      "grad_norm": 7.860400199890137,
+      "learning_rate": 1.1230922693266834e-05,
+      "loss": 0.3599,
+      "step": 35180
+    },
+    {
+      "epoch": 8.775561097256858,
+      "grad_norm": 16.83560562133789,
+      "learning_rate": 1.12284289276808e-05,
+      "loss": 0.3941,
+      "step": 35190
+    },
+    {
+      "epoch": 8.778054862842893,
+      "grad_norm": 6.330166816711426,
+      "learning_rate": 1.1225935162094763e-05,
+      "loss": 0.3688,
+      "step": 35200
+    },
+    {
+      "epoch": 8.780548628428928,
+      "grad_norm": 6.7685418128967285,
+      "learning_rate": 1.122344139650873e-05,
+      "loss": 0.3861,
+      "step": 35210
+    },
+    {
+      "epoch": 8.783042394014963,
+      "grad_norm": 8.012506484985352,
+      "learning_rate": 1.1220947630922694e-05,
+      "loss": 0.3795,
+      "step": 35220
+    },
+    {
+      "epoch": 8.785536159600998,
+      "grad_norm": 7.705891132354736,
+      "learning_rate": 1.1218453865336661e-05,
+      "loss": 0.439,
+      "step": 35230
+    },
+    {
+      "epoch": 8.788029925187033,
+      "grad_norm": 6.158017158508301,
+      "learning_rate": 1.1215960099750625e-05,
+      "loss": 0.3485,
+      "step": 35240
+    },
+    {
+      "epoch": 8.790523690773068,
+      "grad_norm": 10.359025001525879,
+      "learning_rate": 1.1213466334164588e-05,
+      "loss": 0.397,
+      "step": 35250
+    },
+    {
+      "epoch": 8.793017456359102,
+      "grad_norm": 6.772762298583984,
+      "learning_rate": 1.1210972568578555e-05,
+      "loss": 0.351,
+      "step": 35260
+    },
+    {
+      "epoch": 8.795511221945137,
+      "grad_norm": 8.312684059143066,
+      "learning_rate": 1.1208478802992519e-05,
+      "loss": 0.3252,
+      "step": 35270
+    },
+    {
+      "epoch": 8.798004987531172,
+      "grad_norm": 7.869873523712158,
+      "learning_rate": 1.1205985037406484e-05,
+      "loss": 0.3912,
+      "step": 35280
+    },
+    {
+      "epoch": 8.800498753117207,
+      "grad_norm": 7.570711612701416,
+      "learning_rate": 1.1203491271820451e-05,
+      "loss": 0.3725,
+      "step": 35290
+    },
+    {
+      "epoch": 8.802992518703242,
+      "grad_norm": 9.398818016052246,
+      "learning_rate": 1.1200997506234415e-05,
+      "loss": 0.3509,
+      "step": 35300
+    },
+    {
+      "epoch": 8.805486284289277,
+      "grad_norm": 6.553288459777832,
+      "learning_rate": 1.1198503740648382e-05,
+      "loss": 0.3134,
+      "step": 35310
+    },
+    {
+      "epoch": 8.807980049875312,
+      "grad_norm": 6.000061511993408,
+      "learning_rate": 1.1196009975062345e-05,
+      "loss": 0.3332,
+      "step": 35320
+    },
+    {
+      "epoch": 8.810473815461346,
+      "grad_norm": 10.576332092285156,
+      "learning_rate": 1.1193516209476309e-05,
+      "loss": 0.4149,
+      "step": 35330
+    },
+    {
+      "epoch": 8.812967581047381,
+      "grad_norm": 6.1524505615234375,
+      "learning_rate": 1.1191022443890276e-05,
+      "loss": 0.3426,
+      "step": 35340
+    },
+    {
+      "epoch": 8.815461346633416,
+      "grad_norm": 10.099903106689453,
+      "learning_rate": 1.118852867830424e-05,
+      "loss": 0.3464,
+      "step": 35350
+    },
+    {
+      "epoch": 8.817955112219451,
+      "grad_norm": 5.1967854499816895,
+      "learning_rate": 1.1186034912718205e-05,
+      "loss": 0.3209,
+      "step": 35360
+    },
+    {
+      "epoch": 8.820448877805486,
+      "grad_norm": 10.381093978881836,
+      "learning_rate": 1.118354114713217e-05,
+      "loss": 0.3733,
+      "step": 35370
+    },
+    {
+      "epoch": 8.82294264339152,
+      "grad_norm": 17.32126808166504,
+      "learning_rate": 1.1181047381546136e-05,
+      "loss": 0.3812,
+      "step": 35380
+    },
+    {
+      "epoch": 8.825436408977556,
+      "grad_norm": 11.370584487915039,
+      "learning_rate": 1.11785536159601e-05,
+      "loss": 0.3574,
+      "step": 35390
+    },
+    {
+      "epoch": 8.82793017456359,
+      "grad_norm": 6.021444797515869,
+      "learning_rate": 1.1176059850374066e-05,
+      "loss": 0.3345,
+      "step": 35400
+    },
+    {
+      "epoch": 8.830423940149625,
+      "grad_norm": 9.910523414611816,
+      "learning_rate": 1.117356608478803e-05,
+      "loss": 0.3779,
+      "step": 35410
+    },
+    {
+      "epoch": 8.83291770573566,
+      "grad_norm": 10.403486251831055,
+      "learning_rate": 1.1171072319201997e-05,
+      "loss": 0.2948,
+      "step": 35420
+    },
+    {
+      "epoch": 8.835411471321695,
+      "grad_norm": 5.244110107421875,
+      "learning_rate": 1.116857855361596e-05,
+      "loss": 0.3526,
+      "step": 35430
+    },
+    {
+      "epoch": 8.83790523690773,
+      "grad_norm": 8.06598949432373,
+      "learning_rate": 1.1166084788029926e-05,
+      "loss": 0.3763,
+      "step": 35440
+    },
+    {
+      "epoch": 8.840399002493765,
+      "grad_norm": 6.2660322189331055,
+      "learning_rate": 1.1163591022443891e-05,
+      "loss": 0.36,
+      "step": 35450
+    },
+    {
+      "epoch": 8.8428927680798,
+      "grad_norm": 7.570070266723633,
+      "learning_rate": 1.1161097256857856e-05,
+      "loss": 0.423,
+      "step": 35460
+    },
+    {
+      "epoch": 8.845386533665835,
+      "grad_norm": 5.027857780456543,
+      "learning_rate": 1.1158603491271822e-05,
+      "loss": 0.3392,
+      "step": 35470
+    },
+    {
+      "epoch": 8.84788029925187,
+      "grad_norm": 7.999138832092285,
+      "learning_rate": 1.1156109725685787e-05,
+      "loss": 0.3899,
+      "step": 35480
+    },
+    {
+      "epoch": 8.850374064837904,
+      "grad_norm": 9.944479942321777,
+      "learning_rate": 1.115361596009975e-05,
+      "loss": 0.3727,
+      "step": 35490
+    },
+    {
+      "epoch": 8.85286783042394,
+      "grad_norm": 7.199899673461914,
+      "learning_rate": 1.1151122194513718e-05,
+      "loss": 0.3394,
+      "step": 35500
+    },
+    {
+      "epoch": 8.855361596009976,
+      "grad_norm": 5.359086036682129,
+      "learning_rate": 1.1148628428927681e-05,
+      "loss": 0.4158,
+      "step": 35510
+    },
+    {
+      "epoch": 8.85785536159601,
+      "grad_norm": 5.6271820068359375,
+      "learning_rate": 1.1146134663341648e-05,
+      "loss": 0.391,
+      "step": 35520
+    },
+    {
+      "epoch": 8.860349127182046,
+      "grad_norm": 7.151126861572266,
+      "learning_rate": 1.1143640897755612e-05,
+      "loss": 0.3635,
+      "step": 35530
+    },
+    {
+      "epoch": 8.86284289276808,
+      "grad_norm": 6.151798725128174,
+      "learning_rate": 1.1141147132169577e-05,
+      "loss": 0.3526,
+      "step": 35540
+    },
+    {
+      "epoch": 8.865336658354115,
+      "grad_norm": 6.384893417358398,
+      "learning_rate": 1.1138653366583542e-05,
+      "loss": 0.3955,
+      "step": 35550
+    },
+    {
+      "epoch": 8.86783042394015,
+      "grad_norm": 11.764251708984375,
+      "learning_rate": 1.1136159600997508e-05,
+      "loss": 0.3387,
+      "step": 35560
+    },
+    {
+      "epoch": 8.870324189526185,
+      "grad_norm": 10.541848182678223,
+      "learning_rate": 1.1133665835411471e-05,
+      "loss": 0.3911,
+      "step": 35570
+    },
+    {
+      "epoch": 8.87281795511222,
+      "grad_norm": 6.850974082946777,
+      "learning_rate": 1.1131172069825438e-05,
+      "loss": 0.48,
+      "step": 35580
+    },
+    {
+      "epoch": 8.875311720698255,
+      "grad_norm": 7.8186211585998535,
+      "learning_rate": 1.1128678304239402e-05,
+      "loss": 0.4199,
+      "step": 35590
+    },
+    {
+      "epoch": 8.87780548628429,
+      "grad_norm": 7.9405012130737305,
+      "learning_rate": 1.1126184538653369e-05,
+      "loss": 0.3845,
+      "step": 35600
+    },
+    {
+      "epoch": 8.880299251870325,
+      "grad_norm": 7.980791091918945,
+      "learning_rate": 1.1123690773067333e-05,
+      "loss": 0.431,
+      "step": 35610
+    },
+    {
+      "epoch": 8.88279301745636,
+      "grad_norm": 8.639333724975586,
+      "learning_rate": 1.1121197007481296e-05,
+      "loss": 0.3792,
+      "step": 35620
+    },
+    {
+      "epoch": 8.885286783042394,
+      "grad_norm": 6.000261306762695,
+      "learning_rate": 1.1118703241895263e-05,
+      "loss": 0.3846,
+      "step": 35630
+    },
+    {
+      "epoch": 8.88778054862843,
+      "grad_norm": 8.1222505569458,
+      "learning_rate": 1.1116209476309228e-05,
+      "loss": 0.3591,
+      "step": 35640
+    },
+    {
+      "epoch": 8.890274314214464,
+      "grad_norm": 9.311159133911133,
+      "learning_rate": 1.1113715710723192e-05,
+      "loss": 0.3771,
+      "step": 35650
+    },
+    {
+      "epoch": 8.892768079800499,
+      "grad_norm": 5.505731105804443,
+      "learning_rate": 1.1111221945137159e-05,
+      "loss": 0.3531,
+      "step": 35660
+    },
+    {
+      "epoch": 8.895261845386534,
+      "grad_norm": 11.431441307067871,
+      "learning_rate": 1.1108728179551123e-05,
+      "loss": 0.3913,
+      "step": 35670
+    },
+    {
+      "epoch": 8.897755610972569,
+      "grad_norm": 6.601688385009766,
+      "learning_rate": 1.110623441396509e-05,
+      "loss": 0.374,
+      "step": 35680
+    },
+    {
+      "epoch": 8.900249376558603,
+      "grad_norm": 4.69558048248291,
+      "learning_rate": 1.1103740648379053e-05,
+      "loss": 0.3695,
+      "step": 35690
+    },
+    {
+      "epoch": 8.902743142144638,
+      "grad_norm": 6.6589131355285645,
+      "learning_rate": 1.1101246882793017e-05,
+      "loss": 0.3566,
+      "step": 35700
+    },
+    {
+      "epoch": 8.905236907730673,
+      "grad_norm": 7.764341831207275,
+      "learning_rate": 1.1098753117206984e-05,
+      "loss": 0.3723,
+      "step": 35710
+    },
+    {
+      "epoch": 8.907730673316708,
+      "grad_norm": 5.244799613952637,
+      "learning_rate": 1.1096259351620948e-05,
+      "loss": 0.3354,
+      "step": 35720
+    },
+    {
+      "epoch": 8.910224438902743,
+      "grad_norm": 7.864104270935059,
+      "learning_rate": 1.1093765586034915e-05,
+      "loss": 0.3841,
+      "step": 35730
+    },
+    {
+      "epoch": 8.912718204488778,
+      "grad_norm": 7.269486427307129,
+      "learning_rate": 1.1091271820448878e-05,
+      "loss": 0.4008,
+      "step": 35740
+    },
+    {
+      "epoch": 8.915211970074813,
+      "grad_norm": 7.298030376434326,
+      "learning_rate": 1.1088778054862843e-05,
+      "loss": 0.3713,
+      "step": 35750
+    },
+    {
+      "epoch": 8.917705735660848,
+      "grad_norm": 3.9201197624206543,
+      "learning_rate": 1.108628428927681e-05,
+      "loss": 0.3732,
+      "step": 35760
+    },
+    {
+      "epoch": 8.920199501246882,
+      "grad_norm": 5.9074249267578125,
+      "learning_rate": 1.1083790523690774e-05,
+      "loss": 0.3367,
+      "step": 35770
+    },
+    {
+      "epoch": 8.922693266832917,
+      "grad_norm": 16.71881866455078,
+      "learning_rate": 1.1081296758104738e-05,
+      "loss": 0.4005,
+      "step": 35780
+    },
+    {
+      "epoch": 8.925187032418952,
+      "grad_norm": 7.94209623336792,
+      "learning_rate": 1.1078802992518705e-05,
+      "loss": 0.3229,
+      "step": 35790
+    },
+    {
+      "epoch": 8.927680798004987,
+      "grad_norm": 6.1742143630981445,
+      "learning_rate": 1.1076309226932668e-05,
+      "loss": 0.3859,
+      "step": 35800
+    },
+    {
+      "epoch": 8.930174563591022,
+      "grad_norm": 14.812112808227539,
+      "learning_rate": 1.1073815461346635e-05,
+      "loss": 0.4756,
+      "step": 35810
+    },
+    {
+      "epoch": 8.932668329177057,
+      "grad_norm": 7.1429338455200195,
+      "learning_rate": 1.1071321695760599e-05,
+      "loss": 0.2895,
+      "step": 35820
+    },
+    {
+      "epoch": 8.935162094763092,
+      "grad_norm": 6.503314018249512,
+      "learning_rate": 1.1068827930174564e-05,
+      "loss": 0.3162,
+      "step": 35830
+    },
+    {
+      "epoch": 8.937655860349127,
+      "grad_norm": 8.817343711853027,
+      "learning_rate": 1.106633416458853e-05,
+      "loss": 0.374,
+      "step": 35840
+    },
+    {
+      "epoch": 8.940149625935161,
+      "grad_norm": 5.8551740646362305,
+      "learning_rate": 1.1063840399002495e-05,
+      "loss": 0.4058,
+      "step": 35850
+    },
+    {
+      "epoch": 8.942643391521196,
+      "grad_norm": 7.214171409606934,
+      "learning_rate": 1.1061346633416459e-05,
+      "loss": 0.388,
+      "step": 35860
+    },
+    {
+      "epoch": 8.945137157107231,
+      "grad_norm": 6.604674816131592,
+      "learning_rate": 1.1058852867830426e-05,
+      "loss": 0.3192,
+      "step": 35870
+    },
+    {
+      "epoch": 8.947630922693268,
+      "grad_norm": 9.637167930603027,
+      "learning_rate": 1.1056359102244389e-05,
+      "loss": 0.3364,
+      "step": 35880
+    },
+    {
+      "epoch": 8.950124688279303,
+      "grad_norm": 6.015578746795654,
+      "learning_rate": 1.1053865336658356e-05,
+      "loss": 0.3099,
+      "step": 35890
+    },
+    {
+      "epoch": 8.952618453865338,
+      "grad_norm": 6.540172100067139,
+      "learning_rate": 1.105137157107232e-05,
+      "loss": 0.4536,
+      "step": 35900
+    },
+    {
+      "epoch": 8.955112219451372,
+      "grad_norm": 7.110110759735107,
+      "learning_rate": 1.1048877805486285e-05,
+      "loss": 0.319,
+      "step": 35910
+    },
+    {
+      "epoch": 8.957605985037407,
+      "grad_norm": 5.485593318939209,
+      "learning_rate": 1.104638403990025e-05,
+      "loss": 0.3557,
+      "step": 35920
+    },
+    {
+      "epoch": 8.960099750623442,
+      "grad_norm": 6.532005786895752,
+      "learning_rate": 1.1043890274314216e-05,
+      "loss": 0.3922,
+      "step": 35930
+    },
+    {
+      "epoch": 8.962593516209477,
+      "grad_norm": 6.834729194641113,
+      "learning_rate": 1.104139650872818e-05,
+      "loss": 0.3233,
+      "step": 35940
+    },
+    {
+      "epoch": 8.965087281795512,
+      "grad_norm": 5.890754699707031,
+      "learning_rate": 1.1038902743142146e-05,
+      "loss": 0.3753,
+      "step": 35950
+    },
+    {
+      "epoch": 8.967581047381547,
+      "grad_norm": 6.0453619956970215,
+      "learning_rate": 1.103640897755611e-05,
+      "loss": 0.4137,
+      "step": 35960
+    },
+    {
+      "epoch": 8.970074812967582,
+      "grad_norm": 7.995392322540283,
+      "learning_rate": 1.1033915211970077e-05,
+      "loss": 0.3435,
+      "step": 35970
+    },
+    {
+      "epoch": 8.972568578553616,
+      "grad_norm": 7.464896202087402,
+      "learning_rate": 1.103142144638404e-05,
+      "loss": 0.3894,
+      "step": 35980
+    },
+    {
+      "epoch": 8.975062344139651,
+      "grad_norm": 9.544695854187012,
+      "learning_rate": 1.1028927680798006e-05,
+      "loss": 0.4271,
+      "step": 35990
+    },
+    {
+      "epoch": 8.977556109725686,
+      "grad_norm": 5.784707069396973,
+      "learning_rate": 1.1026433915211971e-05,
+      "loss": 0.3547,
+      "step": 36000
+    },
+    {
+      "epoch": 8.980049875311721,
+      "grad_norm": 4.10465669631958,
+      "learning_rate": 1.1023940149625936e-05,
+      "loss": 0.3881,
+      "step": 36010
+    },
+    {
+      "epoch": 8.982543640897756,
+      "grad_norm": 3.3419559001922607,
+      "learning_rate": 1.1021446384039902e-05,
+      "loss": 0.3459,
+      "step": 36020
+    },
+    {
+      "epoch": 8.98503740648379,
+      "grad_norm": 7.649290561676025,
+      "learning_rate": 1.1018952618453867e-05,
+      "loss": 0.2882,
+      "step": 36030
+    },
+    {
+      "epoch": 8.987531172069826,
+      "grad_norm": 4.2098798751831055,
+      "learning_rate": 1.101645885286783e-05,
+      "loss": 0.3061,
+      "step": 36040
+    },
+    {
+      "epoch": 8.99002493765586,
+      "grad_norm": 7.673419952392578,
+      "learning_rate": 1.1013965087281798e-05,
+      "loss": 0.3267,
+      "step": 36050
+    },
+    {
+      "epoch": 8.992518703241895,
+      "grad_norm": 8.080073356628418,
+      "learning_rate": 1.1011471321695761e-05,
+      "loss": 0.3203,
+      "step": 36060
+    },
+    {
+      "epoch": 8.99501246882793,
+      "grad_norm": 4.7040205001831055,
+      "learning_rate": 1.1008977556109725e-05,
+      "loss": 0.3322,
+      "step": 36070
+    },
+    {
+      "epoch": 8.997506234413965,
+      "grad_norm": 6.1027069091796875,
+      "learning_rate": 1.1006483790523692e-05,
+      "loss": 0.384,
+      "step": 36080
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 9.147186279296875,
+      "learning_rate": 1.1003990024937656e-05,
+      "loss": 0.4646,
+      "step": 36090
+    },
+    {
+      "epoch": 9.0,
+      "eval_loss": 0.41386765241622925,
+      "eval_runtime": 59.9541,
+      "eval_samples_per_second": 16.729,
+      "eval_steps_per_second": 16.729,
+      "step": 36090
+    },
+    {
+      "epoch": 9.002493765586035,
+      "grad_norm": 7.474572658538818,
+      "learning_rate": 1.1001496259351623e-05,
+      "loss": 0.3304,
+      "step": 36100
+    },
+    {
+      "epoch": 9.00498753117207,
+      "grad_norm": 4.692042827606201,
+      "learning_rate": 1.0999002493765588e-05,
+      "loss": 0.3193,
+      "step": 36110
+    },
+    {
+      "epoch": 9.007481296758105,
+      "grad_norm": 5.829125881195068,
+      "learning_rate": 1.0996508728179551e-05,
+      "loss": 0.3628,
+      "step": 36120
+    },
+    {
+      "epoch": 9.00997506234414,
+      "grad_norm": 7.040647029876709,
+      "learning_rate": 1.0994014962593518e-05,
+      "loss": 0.3537,
+      "step": 36130
+    },
+    {
+      "epoch": 9.012468827930174,
+      "grad_norm": 7.125100612640381,
+      "learning_rate": 1.0991521197007482e-05,
+      "loss": 0.3556,
+      "step": 36140
+    },
+    {
+      "epoch": 9.01496259351621,
+      "grad_norm": 8.841943740844727,
+      "learning_rate": 1.0989027431421446e-05,
+      "loss": 0.3286,
+      "step": 36150
+    },
+    {
+      "epoch": 9.017456359102244,
+      "grad_norm": 6.718105316162109,
+      "learning_rate": 1.0986533665835413e-05,
+      "loss": 0.3856,
+      "step": 36160
+    },
+    {
+      "epoch": 9.019950124688279,
+      "grad_norm": 6.22119665145874,
+      "learning_rate": 1.0984039900249376e-05,
+      "loss": 0.3323,
+      "step": 36170
+    },
+    {
+      "epoch": 9.022443890274314,
+      "grad_norm": 6.809352874755859,
+      "learning_rate": 1.0981546134663343e-05,
+      "loss": 0.3431,
+      "step": 36180
+    },
+    {
+      "epoch": 9.024937655860349,
+      "grad_norm": 8.36000919342041,
+      "learning_rate": 1.0979052369077307e-05,
+      "loss": 0.3702,
+      "step": 36190
+    },
+    {
+      "epoch": 9.027431421446384,
+      "grad_norm": 9.731013298034668,
+      "learning_rate": 1.0976558603491272e-05,
+      "loss": 0.3476,
+      "step": 36200
+    },
+    {
+      "epoch": 9.029925187032418,
+      "grad_norm": 5.753609657287598,
+      "learning_rate": 1.0974064837905238e-05,
+      "loss": 0.3427,
+      "step": 36210
+    },
+    {
+      "epoch": 9.032418952618453,
+      "grad_norm": 3.485929012298584,
+      "learning_rate": 1.0971571072319203e-05,
+      "loss": 0.3366,
+      "step": 36220
+    },
+    {
+      "epoch": 9.034912718204488,
+      "grad_norm": 5.159622669219971,
+      "learning_rate": 1.096907730673317e-05,
+      "loss": 0.3141,
+      "step": 36230
+    },
+    {
+      "epoch": 9.037406483790523,
+      "grad_norm": 6.331995010375977,
+      "learning_rate": 1.0966583541147134e-05,
+      "loss": 0.3536,
+      "step": 36240
+    },
+    {
+      "epoch": 9.039900249376558,
+      "grad_norm": 5.746108055114746,
+      "learning_rate": 1.0964089775561097e-05,
+      "loss": 0.3514,
+      "step": 36250
+    },
+    {
+      "epoch": 9.042394014962593,
+      "grad_norm": 6.630804061889648,
+      "learning_rate": 1.0961596009975064e-05,
+      "loss": 0.3299,
+      "step": 36260
+    },
+    {
+      "epoch": 9.044887780548628,
+      "grad_norm": 7.066774845123291,
+      "learning_rate": 1.0959102244389028e-05,
+      "loss": 0.2654,
+      "step": 36270
+    },
+    {
+      "epoch": 9.047381546134662,
+      "grad_norm": 9.622215270996094,
+      "learning_rate": 1.0956608478802993e-05,
+      "loss": 0.3501,
+      "step": 36280
+    },
+    {
+      "epoch": 9.049875311720697,
+      "grad_norm": 8.551993370056152,
+      "learning_rate": 1.0954114713216958e-05,
+      "loss": 0.3529,
+      "step": 36290
+    },
+    {
+      "epoch": 9.052369077306734,
+      "grad_norm": 6.227419376373291,
+      "learning_rate": 1.0951620947630924e-05,
+      "loss": 0.3803,
+      "step": 36300
+    },
+    {
+      "epoch": 9.054862842892769,
+      "grad_norm": 8.591386795043945,
+      "learning_rate": 1.0949127182044889e-05,
+      "loss": 0.3134,
+      "step": 36310
+    },
+    {
+      "epoch": 9.057356608478804,
+      "grad_norm": 5.2157673835754395,
+      "learning_rate": 1.0946633416458854e-05,
+      "loss": 0.3214,
+      "step": 36320
+    },
+    {
+      "epoch": 9.059850374064839,
+      "grad_norm": 5.971094131469727,
+      "learning_rate": 1.0944139650872818e-05,
+      "loss": 0.3343,
+      "step": 36330
+    },
+    {
+      "epoch": 9.062344139650873,
+      "grad_norm": 7.728214263916016,
+      "learning_rate": 1.0941645885286785e-05,
+      "loss": 0.4131,
+      "step": 36340
+    },
+    {
+      "epoch": 9.064837905236908,
+      "grad_norm": 9.236366271972656,
+      "learning_rate": 1.0939401496259351e-05,
+      "loss": 0.3518,
+      "step": 36350
+    },
+    {
+      "epoch": 9.067331670822943,
+      "grad_norm": 4.353968620300293,
+      "learning_rate": 1.0936907730673318e-05,
+      "loss": 0.352,
+      "step": 36360
+    },
+    {
+      "epoch": 9.069825436408978,
+      "grad_norm": 9.550552368164062,
+      "learning_rate": 1.0934413965087282e-05,
+      "loss": 0.4209,
+      "step": 36370
+    },
+    {
+      "epoch": 9.072319201995013,
+      "grad_norm": 8.467497825622559,
+      "learning_rate": 1.0931920199501247e-05,
+      "loss": 0.3342,
+      "step": 36380
+    },
+    {
+      "epoch": 9.074812967581048,
+      "grad_norm": 11.831032752990723,
+      "learning_rate": 1.0929675810473817e-05,
+      "loss": 0.3481,
+      "step": 36390
+    },
+    {
+      "epoch": 9.077306733167083,
+      "grad_norm": 5.57447624206543,
+      "learning_rate": 1.0927182044887781e-05,
+      "loss": 0.2991,
+      "step": 36400
+    },
+    {
+      "epoch": 9.079800498753118,
+      "grad_norm": 10.63701057434082,
+      "learning_rate": 1.0924688279301748e-05,
+      "loss": 0.3668,
+      "step": 36410
+    },
+    {
+      "epoch": 9.082294264339152,
+      "grad_norm": 9.0299711227417,
+      "learning_rate": 1.0922194513715712e-05,
+      "loss": 0.4303,
+      "step": 36420
+    },
+    {
+      "epoch": 9.084788029925187,
+      "grad_norm": 6.327706813812256,
+      "learning_rate": 1.0919700748129675e-05,
+      "loss": 0.3057,
+      "step": 36430
+    },
+    {
+      "epoch": 9.087281795511222,
+      "grad_norm": 10.897625923156738,
+      "learning_rate": 1.0917206982543642e-05,
+      "loss": 0.3726,
+      "step": 36440
+    },
+    {
+      "epoch": 9.089775561097257,
+      "grad_norm": 8.112853050231934,
+      "learning_rate": 1.0914713216957606e-05,
+      "loss": 0.3477,
+      "step": 36450
+    },
+    {
+      "epoch": 9.092269326683292,
+      "grad_norm": 7.813316345214844,
+      "learning_rate": 1.0912219451371573e-05,
+      "loss": 0.2972,
+      "step": 36460
+    },
+    {
+      "epoch": 9.094763092269327,
+      "grad_norm": 8.015786170959473,
+      "learning_rate": 1.0909725685785536e-05,
+      "loss": 0.3787,
+      "step": 36470
+    },
+    {
+      "epoch": 9.097256857855362,
+      "grad_norm": 6.349668502807617,
+      "learning_rate": 1.0907231920199502e-05,
+      "loss": 0.3598,
+      "step": 36480
+    },
+    {
+      "epoch": 9.099750623441397,
+      "grad_norm": 5.772064208984375,
+      "learning_rate": 1.0904738154613467e-05,
+      "loss": 0.32,
+      "step": 36490
+    },
+    {
+      "epoch": 9.102244389027431,
+      "grad_norm": 6.944436073303223,
+      "learning_rate": 1.0902244389027432e-05,
+      "loss": 0.3449,
+      "step": 36500
+    },
+    {
+      "epoch": 9.104738154613466,
+      "grad_norm": 6.625913619995117,
+      "learning_rate": 1.0899750623441396e-05,
+      "loss": 0.3236,
+      "step": 36510
+    },
+    {
+      "epoch": 9.107231920199501,
+      "grad_norm": 6.98844051361084,
+      "learning_rate": 1.0897256857855363e-05,
+      "loss": 0.3226,
+      "step": 36520
+    },
+    {
+      "epoch": 9.109725685785536,
+      "grad_norm": 10.177889823913574,
+      "learning_rate": 1.0894763092269327e-05,
+      "loss": 0.3706,
+      "step": 36530
+    },
+    {
+      "epoch": 9.11221945137157,
+      "grad_norm": 9.637530326843262,
+      "learning_rate": 1.0892269326683294e-05,
+      "loss": 0.3774,
+      "step": 36540
+    },
+    {
+      "epoch": 9.114713216957606,
+      "grad_norm": 6.966104030609131,
+      "learning_rate": 1.0889775561097257e-05,
+      "loss": 0.3834,
+      "step": 36550
+    },
+    {
+      "epoch": 9.11720698254364,
+      "grad_norm": 4.891960620880127,
+      "learning_rate": 1.0887281795511223e-05,
+      "loss": 0.3099,
+      "step": 36560
+    },
+    {
+      "epoch": 9.119700748129675,
+      "grad_norm": 12.025975227355957,
+      "learning_rate": 1.0884788029925188e-05,
+      "loss": 0.3532,
+      "step": 36570
+    },
+    {
+      "epoch": 9.12219451371571,
+      "grad_norm": 7.283741474151611,
+      "learning_rate": 1.0882294264339153e-05,
+      "loss": 0.3582,
+      "step": 36580
+    },
+    {
+      "epoch": 9.124688279301745,
+      "grad_norm": 16.89406967163086,
+      "learning_rate": 1.0879800498753118e-05,
+      "loss": 0.42,
+      "step": 36590
+    },
+    {
+      "epoch": 9.12718204488778,
+      "grad_norm": 9.610578536987305,
+      "learning_rate": 1.0877306733167084e-05,
+      "loss": 0.3352,
+      "step": 36600
+    },
+    {
+      "epoch": 9.129675810473815,
+      "grad_norm": 6.792508602142334,
+      "learning_rate": 1.0874812967581047e-05,
+      "loss": 0.3906,
+      "step": 36610
+    },
+    {
+      "epoch": 9.13216957605985,
+      "grad_norm": 6.052822113037109,
+      "learning_rate": 1.0872319201995014e-05,
+      "loss": 0.3852,
+      "step": 36620
+    },
+    {
+      "epoch": 9.134663341645885,
+      "grad_norm": 8.799402236938477,
+      "learning_rate": 1.0869825436408978e-05,
+      "loss": 0.4047,
+      "step": 36630
+    },
+    {
+      "epoch": 9.13715710723192,
+      "grad_norm": 8.114533424377441,
+      "learning_rate": 1.0867331670822943e-05,
+      "loss": 0.3871,
+      "step": 36640
+    },
+    {
+      "epoch": 9.139650872817954,
+      "grad_norm": 9.247639656066895,
+      "learning_rate": 1.0864837905236909e-05,
+      "loss": 0.3993,
+      "step": 36650
+    },
+    {
+      "epoch": 9.14214463840399,
+      "grad_norm": 7.904721736907959,
+      "learning_rate": 1.0862344139650874e-05,
+      "loss": 0.3441,
+      "step": 36660
+    },
+    {
+      "epoch": 9.144638403990024,
+      "grad_norm": 12.610846519470215,
+      "learning_rate": 1.085985037406484e-05,
+      "loss": 0.3699,
+      "step": 36670
+    },
+    {
+      "epoch": 9.147132169576059,
+      "grad_norm": 5.233782768249512,
+      "learning_rate": 1.0857356608478805e-05,
+      "loss": 0.3623,
+      "step": 36680
+    },
+    {
+      "epoch": 9.149625935162096,
+      "grad_norm": 6.99005651473999,
+      "learning_rate": 1.0854862842892768e-05,
+      "loss": 0.3476,
+      "step": 36690
+    },
+    {
+      "epoch": 9.15211970074813,
+      "grad_norm": 11.361775398254395,
+      "learning_rate": 1.0852369077306735e-05,
+      "loss": 0.412,
+      "step": 36700
+    },
+    {
+      "epoch": 9.154613466334165,
+      "grad_norm": 6.979437828063965,
+      "learning_rate": 1.0849875311720699e-05,
+      "loss": 0.4043,
+      "step": 36710
+    },
+    {
+      "epoch": 9.1571072319202,
+      "grad_norm": 6.4810943603515625,
+      "learning_rate": 1.0847381546134662e-05,
+      "loss": 0.3768,
+      "step": 36720
+    },
+    {
+      "epoch": 9.159600997506235,
+      "grad_norm": 10.649642944335938,
+      "learning_rate": 1.084488778054863e-05,
+      "loss": 0.3687,
+      "step": 36730
+    },
+    {
+      "epoch": 9.16209476309227,
+      "grad_norm": 10.065546035766602,
+      "learning_rate": 1.0842394014962595e-05,
+      "loss": 0.3891,
+      "step": 36740
+    },
+    {
+      "epoch": 9.164588528678305,
+      "grad_norm": 7.575584411621094,
+      "learning_rate": 1.083990024937656e-05,
+      "loss": 0.3816,
+      "step": 36750
+    },
+    {
+      "epoch": 9.16708229426434,
+      "grad_norm": 10.201583862304688,
+      "learning_rate": 1.0837406483790525e-05,
+      "loss": 0.3173,
+      "step": 36760
+    },
+    {
+      "epoch": 9.169576059850375,
+      "grad_norm": 10.348787307739258,
+      "learning_rate": 1.0834912718204489e-05,
+      "loss": 0.4488,
+      "step": 36770
+    },
+    {
+      "epoch": 9.17206982543641,
+      "grad_norm": 6.726578235626221,
+      "learning_rate": 1.0832418952618456e-05,
+      "loss": 0.3276,
+      "step": 36780
+    },
+    {
+      "epoch": 9.174563591022444,
+      "grad_norm": 8.635772705078125,
+      "learning_rate": 1.082992518703242e-05,
+      "loss": 0.3571,
+      "step": 36790
+    },
+    {
+      "epoch": 9.17705735660848,
+      "grad_norm": 9.382979393005371,
+      "learning_rate": 1.0827431421446383e-05,
+      "loss": 0.3251,
+      "step": 36800
+    },
+    {
+      "epoch": 9.179551122194514,
+      "grad_norm": 5.400996208190918,
+      "learning_rate": 1.082493765586035e-05,
+      "loss": 0.3253,
+      "step": 36810
+    },
+    {
+      "epoch": 9.182044887780549,
+      "grad_norm": 6.488340854644775,
+      "learning_rate": 1.0822443890274314e-05,
+      "loss": 0.3224,
+      "step": 36820
+    },
+    {
+      "epoch": 9.184538653366584,
+      "grad_norm": 6.428625583648682,
+      "learning_rate": 1.081995012468828e-05,
+      "loss": 0.3234,
+      "step": 36830
+    },
+    {
+      "epoch": 9.187032418952619,
+      "grad_norm": 7.193785667419434,
+      "learning_rate": 1.0817456359102244e-05,
+      "loss": 0.2956,
+      "step": 36840
+    },
+    {
+      "epoch": 9.189526184538654,
+      "grad_norm": 9.758343696594238,
+      "learning_rate": 1.081496259351621e-05,
+      "loss": 0.3444,
+      "step": 36850
+    },
+    {
+      "epoch": 9.192019950124688,
+      "grad_norm": 6.714698314666748,
+      "learning_rate": 1.0812468827930177e-05,
+      "loss": 0.4052,
+      "step": 36860
+    },
+    {
+      "epoch": 9.194513715710723,
+      "grad_norm": 8.280781745910645,
+      "learning_rate": 1.080997506234414e-05,
+      "loss": 0.3476,
+      "step": 36870
+    },
+    {
+      "epoch": 9.197007481296758,
+      "grad_norm": 9.618976593017578,
+      "learning_rate": 1.0807481296758107e-05,
+      "loss": 0.3929,
+      "step": 36880
+    },
+    {
+      "epoch": 9.199501246882793,
+      "grad_norm": 8.107794761657715,
+      "learning_rate": 1.0804987531172071e-05,
+      "loss": 0.3866,
+      "step": 36890
+    },
+    {
+      "epoch": 9.201995012468828,
+      "grad_norm": 6.555667400360107,
+      "learning_rate": 1.0802493765586035e-05,
+      "loss": 0.3293,
+      "step": 36900
+    },
+    {
+      "epoch": 9.204488778054863,
+      "grad_norm": 5.286886215209961,
+      "learning_rate": 1.0800000000000002e-05,
+      "loss": 0.3245,
+      "step": 36910
+    },
+    {
+      "epoch": 9.206982543640898,
+      "grad_norm": 16.370393753051758,
+      "learning_rate": 1.0797506234413965e-05,
+      "loss": 0.3701,
+      "step": 36920
+    },
+    {
+      "epoch": 9.209476309226932,
+      "grad_norm": 8.964117050170898,
+      "learning_rate": 1.079501246882793e-05,
+      "loss": 0.4376,
+      "step": 36930
+    },
+    {
+      "epoch": 9.211970074812967,
+      "grad_norm": 11.275566101074219,
+      "learning_rate": 1.0792518703241896e-05,
+      "loss": 0.3203,
+      "step": 36940
+    },
+    {
+      "epoch": 9.214463840399002,
+      "grad_norm": 7.305138111114502,
+      "learning_rate": 1.0790024937655861e-05,
+      "loss": 0.3401,
+      "step": 36950
+    },
+    {
+      "epoch": 9.216957605985037,
+      "grad_norm": 8.845014572143555,
+      "learning_rate": 1.0787531172069826e-05,
+      "loss": 0.3508,
+      "step": 36960
+    },
+    {
+      "epoch": 9.219451371571072,
+      "grad_norm": 7.507462024688721,
+      "learning_rate": 1.0785037406483792e-05,
+      "loss": 0.3717,
+      "step": 36970
+    },
+    {
+      "epoch": 9.221945137157107,
+      "grad_norm": 6.304407119750977,
+      "learning_rate": 1.0782543640897755e-05,
+      "loss": 0.4225,
+      "step": 36980
+    },
+    {
+      "epoch": 9.224438902743142,
+      "grad_norm": 6.4769511222839355,
+      "learning_rate": 1.0780049875311722e-05,
+      "loss": 0.3015,
+      "step": 36990
+    },
+    {
+      "epoch": 9.226932668329177,
+      "grad_norm": 5.64371395111084,
+      "learning_rate": 1.0777556109725686e-05,
+      "loss": 0.3666,
+      "step": 37000
+    },
+    {
+      "epoch": 9.229426433915211,
+      "grad_norm": 7.976513385772705,
+      "learning_rate": 1.0775062344139651e-05,
+      "loss": 0.3437,
+      "step": 37010
+    },
+    {
+      "epoch": 9.231920199501246,
+      "grad_norm": 8.061870574951172,
+      "learning_rate": 1.0772568578553617e-05,
+      "loss": 0.4176,
+      "step": 37020
+    },
+    {
+      "epoch": 9.234413965087281,
+      "grad_norm": 6.633289813995361,
+      "learning_rate": 1.0770074812967582e-05,
+      "loss": 0.3978,
+      "step": 37030
+    },
+    {
+      "epoch": 9.236907730673316,
+      "grad_norm": 6.083536624908447,
+      "learning_rate": 1.0767581047381547e-05,
+      "loss": 0.3067,
+      "step": 37040
+    },
+    {
+      "epoch": 9.239401496259351,
+      "grad_norm": 9.767627716064453,
+      "learning_rate": 1.0765087281795513e-05,
+      "loss": 0.449,
+      "step": 37050
+    },
+    {
+      "epoch": 9.241895261845386,
+      "grad_norm": 6.643617153167725,
+      "learning_rate": 1.0762593516209476e-05,
+      "loss": 0.4396,
+      "step": 37060
+    },
+    {
+      "epoch": 9.24438902743142,
+      "grad_norm": 13.932653427124023,
+      "learning_rate": 1.0760099750623443e-05,
+      "loss": 0.2836,
+      "step": 37070
+    },
+    {
+      "epoch": 9.246882793017456,
+      "grad_norm": 8.449307441711426,
+      "learning_rate": 1.0757605985037407e-05,
+      "loss": 0.3397,
+      "step": 37080
+    },
+    {
+      "epoch": 9.24937655860349,
+      "grad_norm": 4.122343063354492,
+      "learning_rate": 1.0755112219451374e-05,
+      "loss": 0.3449,
+      "step": 37090
+    },
+    {
+      "epoch": 9.251870324189527,
+      "grad_norm": 5.507016658782959,
+      "learning_rate": 1.0752618453865337e-05,
+      "loss": 0.3574,
+      "step": 37100
+    },
+    {
+      "epoch": 9.254364089775562,
+      "grad_norm": 4.903787136077881,
+      "learning_rate": 1.0750124688279303e-05,
+      "loss": 0.3682,
+      "step": 37110
+    },
+    {
+      "epoch": 9.256857855361597,
+      "grad_norm": 10.680002212524414,
+      "learning_rate": 1.0747630922693268e-05,
+      "loss": 0.2884,
+      "step": 37120
+    },
+    {
+      "epoch": 9.259351620947632,
+      "grad_norm": 7.1772284507751465,
+      "learning_rate": 1.0745137157107233e-05,
+      "loss": 0.3803,
+      "step": 37130
+    },
+    {
+      "epoch": 9.261845386533667,
+      "grad_norm": 5.805769920349121,
+      "learning_rate": 1.0742643391521197e-05,
+      "loss": 0.3874,
+      "step": 37140
+    },
+    {
+      "epoch": 9.264339152119701,
+      "grad_norm": 6.096296787261963,
+      "learning_rate": 1.0740149625935164e-05,
+      "loss": 0.2803,
+      "step": 37150
+    },
+    {
+      "epoch": 9.266832917705736,
+      "grad_norm": 9.108000755310059,
+      "learning_rate": 1.0737655860349128e-05,
+      "loss": 0.4142,
+      "step": 37160
+    },
+    {
+      "epoch": 9.269326683291771,
+      "grad_norm": 8.92357063293457,
+      "learning_rate": 1.0735162094763095e-05,
+      "loss": 0.3979,
+      "step": 37170
+    },
+    {
+      "epoch": 9.271820448877806,
+      "grad_norm": 7.647791862487793,
+      "learning_rate": 1.0732668329177058e-05,
+      "loss": 0.3712,
+      "step": 37180
+    },
+    {
+      "epoch": 9.27431421446384,
+      "grad_norm": 8.077733993530273,
+      "learning_rate": 1.0730174563591022e-05,
+      "loss": 0.334,
+      "step": 37190
+    },
+    {
+      "epoch": 9.276807980049876,
+      "grad_norm": 9.143230438232422,
+      "learning_rate": 1.0727680798004989e-05,
+      "loss": 0.3473,
+      "step": 37200
+    },
+    {
+      "epoch": 9.27930174563591,
+      "grad_norm": 9.733719825744629,
+      "learning_rate": 1.0725187032418954e-05,
+      "loss": 0.329,
+      "step": 37210
+    },
+    {
+      "epoch": 9.281795511221945,
+      "grad_norm": 6.865516185760498,
+      "learning_rate": 1.0722693266832918e-05,
+      "loss": 0.4986,
+      "step": 37220
+    },
+    {
+      "epoch": 9.28428927680798,
+      "grad_norm": 6.8000664710998535,
+      "learning_rate": 1.0720199501246885e-05,
+      "loss": 0.3473,
+      "step": 37230
+    },
+    {
+      "epoch": 9.286783042394015,
+      "grad_norm": 5.607472896575928,
+      "learning_rate": 1.0717705735660848e-05,
+      "loss": 0.3789,
+      "step": 37240
+    },
+    {
+      "epoch": 9.28927680798005,
+      "grad_norm": 8.08205795288086,
+      "learning_rate": 1.0715211970074815e-05,
+      "loss": 0.4234,
+      "step": 37250
+    },
+    {
+      "epoch": 9.291770573566085,
+      "grad_norm": 5.731466293334961,
+      "learning_rate": 1.0712718204488779e-05,
+      "loss": 0.3009,
+      "step": 37260
+    },
+    {
+      "epoch": 9.29426433915212,
+      "grad_norm": 8.150042533874512,
+      "learning_rate": 1.0710224438902743e-05,
+      "loss": 0.382,
+      "step": 37270
+    },
+    {
+      "epoch": 9.296758104738155,
+      "grad_norm": 4.996364116668701,
+      "learning_rate": 1.070773067331671e-05,
+      "loss": 0.319,
+      "step": 37280
+    },
+    {
+      "epoch": 9.29925187032419,
+      "grad_norm": 5.883243083953857,
+      "learning_rate": 1.0705236907730673e-05,
+      "loss": 0.3691,
+      "step": 37290
+    },
+    {
+      "epoch": 9.301745635910224,
+      "grad_norm": 8.430337905883789,
+      "learning_rate": 1.070274314214464e-05,
+      "loss": 0.3654,
+      "step": 37300
+    },
+    {
+      "epoch": 9.30423940149626,
+      "grad_norm": 5.940845012664795,
+      "learning_rate": 1.0700249376558604e-05,
+      "loss": 0.4136,
+      "step": 37310
+    },
+    {
+      "epoch": 9.306733167082294,
+      "grad_norm": 8.069353103637695,
+      "learning_rate": 1.069775561097257e-05,
+      "loss": 0.3543,
+      "step": 37320
+    },
+    {
+      "epoch": 9.309226932668329,
+      "grad_norm": 7.512514591217041,
+      "learning_rate": 1.0695261845386536e-05,
+      "loss": 0.3448,
+      "step": 37330
+    },
+    {
+      "epoch": 9.311720698254364,
+      "grad_norm": 8.438852310180664,
+      "learning_rate": 1.06927680798005e-05,
+      "loss": 0.3801,
+      "step": 37340
+    },
+    {
+      "epoch": 9.314214463840399,
+      "grad_norm": 6.524256229400635,
+      "learning_rate": 1.0690274314214463e-05,
+      "loss": 0.43,
+      "step": 37350
+    },
+    {
+      "epoch": 9.316708229426434,
+      "grad_norm": 7.598324298858643,
+      "learning_rate": 1.068778054862843e-05,
+      "loss": 0.3323,
+      "step": 37360
+    },
+    {
+      "epoch": 9.319201995012468,
+      "grad_norm": 6.724202632904053,
+      "learning_rate": 1.0685286783042394e-05,
+      "loss": 0.3265,
+      "step": 37370
+    },
+    {
+      "epoch": 9.321695760598503,
+      "grad_norm": 6.113863945007324,
+      "learning_rate": 1.0682793017456361e-05,
+      "loss": 0.3499,
+      "step": 37380
+    },
+    {
+      "epoch": 9.324189526184538,
+      "grad_norm": 5.505059719085693,
+      "learning_rate": 1.0680299251870325e-05,
+      "loss": 0.4184,
+      "step": 37390
+    },
+    {
+      "epoch": 9.326683291770573,
+      "grad_norm": 7.815418243408203,
+      "learning_rate": 1.067780548628429e-05,
+      "loss": 0.4034,
+      "step": 37400
+    },
+    {
+      "epoch": 9.329177057356608,
+      "grad_norm": 7.3198957443237305,
+      "learning_rate": 1.0675311720698255e-05,
+      "loss": 0.3815,
+      "step": 37410
+    },
+    {
+      "epoch": 9.331670822942643,
+      "grad_norm": 5.2639570236206055,
+      "learning_rate": 1.067281795511222e-05,
+      "loss": 0.4119,
+      "step": 37420
+    },
+    {
+      "epoch": 9.334164588528678,
+      "grad_norm": 7.291960716247559,
+      "learning_rate": 1.0670324189526184e-05,
+      "loss": 0.3337,
+      "step": 37430
+    },
+    {
+      "epoch": 9.336658354114713,
+      "grad_norm": 6.731927871704102,
+      "learning_rate": 1.0667830423940151e-05,
+      "loss": 0.3206,
+      "step": 37440
+    },
+    {
+      "epoch": 9.339152119700747,
+      "grad_norm": 6.022372722625732,
+      "learning_rate": 1.0665336658354115e-05,
+      "loss": 0.3708,
+      "step": 37450
+    },
+    {
+      "epoch": 9.341645885286782,
+      "grad_norm": 8.287287712097168,
+      "learning_rate": 1.0662842892768082e-05,
+      "loss": 0.3234,
+      "step": 37460
+    },
+    {
+      "epoch": 9.344139650872817,
+      "grad_norm": 5.861464023590088,
+      "learning_rate": 1.0660349127182045e-05,
+      "loss": 0.3365,
+      "step": 37470
+    },
+    {
+      "epoch": 9.346633416458852,
+      "grad_norm": 7.660949230194092,
+      "learning_rate": 1.065785536159601e-05,
+      "loss": 0.4347,
+      "step": 37480
+    },
+    {
+      "epoch": 9.349127182044889,
+      "grad_norm": 7.541104793548584,
+      "learning_rate": 1.0655361596009976e-05,
+      "loss": 0.3541,
+      "step": 37490
+    },
+    {
+      "epoch": 9.351620947630924,
+      "grad_norm": 6.571524620056152,
+      "learning_rate": 1.0652867830423941e-05,
+      "loss": 0.3529,
+      "step": 37500
+    },
+    {
+      "epoch": 9.354114713216958,
+      "grad_norm": 7.663899898529053,
+      "learning_rate": 1.0650374064837905e-05,
+      "loss": 0.3262,
+      "step": 37510
+    },
+    {
+      "epoch": 9.356608478802993,
+      "grad_norm": 5.076836109161377,
+      "learning_rate": 1.0647880299251872e-05,
+      "loss": 0.3693,
+      "step": 37520
+    },
+    {
+      "epoch": 9.359102244389028,
+      "grad_norm": 8.110295295715332,
+      "learning_rate": 1.0645386533665836e-05,
+      "loss": 0.3447,
+      "step": 37530
+    },
+    {
+      "epoch": 9.361596009975063,
+      "grad_norm": 9.976151466369629,
+      "learning_rate": 1.0642892768079803e-05,
+      "loss": 0.3858,
+      "step": 37540
+    },
+    {
+      "epoch": 9.364089775561098,
+      "grad_norm": 5.691299915313721,
+      "learning_rate": 1.0640399002493766e-05,
+      "loss": 0.3673,
+      "step": 37550
+    },
+    {
+      "epoch": 9.366583541147133,
+      "grad_norm": 6.755293369293213,
+      "learning_rate": 1.0637905236907732e-05,
+      "loss": 0.4159,
+      "step": 37560
+    },
+    {
+      "epoch": 9.369077306733168,
+      "grad_norm": 6.009609699249268,
+      "learning_rate": 1.0635411471321697e-05,
+      "loss": 0.3256,
+      "step": 37570
+    },
+    {
+      "epoch": 9.371571072319203,
+      "grad_norm": 8.494285583496094,
+      "learning_rate": 1.0632917705735662e-05,
+      "loss": 0.3694,
+      "step": 37580
+    },
+    {
+      "epoch": 9.374064837905237,
+      "grad_norm": 5.737974643707275,
+      "learning_rate": 1.0630423940149627e-05,
+      "loss": 0.3142,
+      "step": 37590
+    },
+    {
+      "epoch": 9.376558603491272,
+      "grad_norm": 6.128436088562012,
+      "learning_rate": 1.0627930174563593e-05,
+      "loss": 0.3198,
+      "step": 37600
+    },
+    {
+      "epoch": 9.379052369077307,
+      "grad_norm": 7.005685806274414,
+      "learning_rate": 1.0625436408977556e-05,
+      "loss": 0.2682,
+      "step": 37610
+    },
+    {
+      "epoch": 9.381546134663342,
+      "grad_norm": 9.262737274169922,
+      "learning_rate": 1.0622942643391523e-05,
+      "loss": 0.3833,
+      "step": 37620
+    },
+    {
+      "epoch": 9.384039900249377,
+      "grad_norm": 7.676625728607178,
+      "learning_rate": 1.0620448877805487e-05,
+      "loss": 0.3851,
+      "step": 37630
+    },
+    {
+      "epoch": 9.386533665835412,
+      "grad_norm": 7.627851486206055,
+      "learning_rate": 1.061795511221945e-05,
+      "loss": 0.3433,
+      "step": 37640
+    },
+    {
+      "epoch": 9.389027431421447,
+      "grad_norm": 7.219352722167969,
+      "learning_rate": 1.0615461346633418e-05,
+      "loss": 0.3822,
+      "step": 37650
+    },
+    {
+      "epoch": 9.391521197007481,
+      "grad_norm": 6.273348808288574,
+      "learning_rate": 1.0612967581047381e-05,
+      "loss": 0.3067,
+      "step": 37660
+    },
+    {
+      "epoch": 9.394014962593516,
+      "grad_norm": 7.28784704208374,
+      "learning_rate": 1.0610473815461348e-05,
+      "loss": 0.3483,
+      "step": 37670
+    },
+    {
+      "epoch": 9.396508728179551,
+      "grad_norm": 7.136936187744141,
+      "learning_rate": 1.0607980049875314e-05,
+      "loss": 0.3342,
+      "step": 37680
+    },
+    {
+      "epoch": 9.399002493765586,
+      "grad_norm": 9.429666519165039,
+      "learning_rate": 1.0605486284289277e-05,
+      "loss": 0.387,
+      "step": 37690
+    },
+    {
+      "epoch": 9.401496259351621,
+      "grad_norm": 10.183398246765137,
+      "learning_rate": 1.0602992518703244e-05,
+      "loss": 0.3281,
+      "step": 37700
+    },
+    {
+      "epoch": 9.403990024937656,
+      "grad_norm": 9.305542945861816,
+      "learning_rate": 1.0600498753117208e-05,
+      "loss": 0.422,
+      "step": 37710
+    },
+    {
+      "epoch": 9.40648379052369,
+      "grad_norm": 13.953811645507812,
+      "learning_rate": 1.0598004987531171e-05,
+      "loss": 0.4099,
+      "step": 37720
+    },
+    {
+      "epoch": 9.408977556109726,
+      "grad_norm": 7.635592460632324,
+      "learning_rate": 1.0595511221945138e-05,
+      "loss": 0.3641,
+      "step": 37730
+    },
+    {
+      "epoch": 9.41147132169576,
+      "grad_norm": 8.282495498657227,
+      "learning_rate": 1.0593017456359102e-05,
+      "loss": 0.3229,
+      "step": 37740
+    },
+    {
+      "epoch": 9.413965087281795,
+      "grad_norm": 6.074685096740723,
+      "learning_rate": 1.0590523690773069e-05,
+      "loss": 0.3637,
+      "step": 37750
+    },
+    {
+      "epoch": 9.41645885286783,
+      "grad_norm": 5.761679172515869,
+      "learning_rate": 1.0588029925187033e-05,
+      "loss": 0.3487,
+      "step": 37760
+    },
+    {
+      "epoch": 9.418952618453865,
+      "grad_norm": 9.500771522521973,
+      "learning_rate": 1.0585536159600998e-05,
+      "loss": 0.3125,
+      "step": 37770
+    },
+    {
+      "epoch": 9.4214463840399,
+      "grad_norm": 8.102692604064941,
+      "learning_rate": 1.0583042394014963e-05,
+      "loss": 0.3755,
+      "step": 37780
+    },
+    {
+      "epoch": 9.423940149625935,
+      "grad_norm": 7.283730506896973,
+      "learning_rate": 1.0580548628428929e-05,
+      "loss": 0.3766,
+      "step": 37790
+    },
+    {
+      "epoch": 9.42643391521197,
+      "grad_norm": 8.951366424560547,
+      "learning_rate": 1.0578054862842896e-05,
+      "loss": 0.4029,
+      "step": 37800
+    },
+    {
+      "epoch": 9.428927680798004,
+      "grad_norm": 5.814979076385498,
+      "learning_rate": 1.057556109725686e-05,
+      "loss": 0.3822,
+      "step": 37810
+    },
+    {
+      "epoch": 9.43142144638404,
+      "grad_norm": 5.397665977478027,
+      "learning_rate": 1.0573067331670823e-05,
+      "loss": 0.3395,
+      "step": 37820
+    },
+    {
+      "epoch": 9.433915211970074,
+      "grad_norm": 8.175272941589355,
+      "learning_rate": 1.057057356608479e-05,
+      "loss": 0.3406,
+      "step": 37830
+    },
+    {
+      "epoch": 9.436408977556109,
+      "grad_norm": 9.08521556854248,
+      "learning_rate": 1.0568079800498753e-05,
+      "loss": 0.3476,
+      "step": 37840
+    },
+    {
+      "epoch": 9.438902743142144,
+      "grad_norm": 5.0024518966674805,
+      "learning_rate": 1.0565586034912719e-05,
+      "loss": 0.4372,
+      "step": 37850
+    },
+    {
+      "epoch": 9.441396508728179,
+      "grad_norm": 9.28567886352539,
+      "learning_rate": 1.0563092269326684e-05,
+      "loss": 0.3301,
+      "step": 37860
+    },
+    {
+      "epoch": 9.443890274314214,
+      "grad_norm": 10.780661582946777,
+      "learning_rate": 1.056059850374065e-05,
+      "loss": 0.3487,
+      "step": 37870
+    },
+    {
+      "epoch": 9.446384039900249,
+      "grad_norm": 6.956487655639648,
+      "learning_rate": 1.0558104738154615e-05,
+      "loss": 0.3411,
+      "step": 37880
+    },
+    {
+      "epoch": 9.448877805486283,
+      "grad_norm": 8.598062515258789,
+      "learning_rate": 1.055561097256858e-05,
+      "loss": 0.406,
+      "step": 37890
+    },
+    {
+      "epoch": 9.451371571072318,
+      "grad_norm": 7.002991676330566,
+      "learning_rate": 1.0553117206982544e-05,
+      "loss": 0.3886,
+      "step": 37900
+    },
+    {
+      "epoch": 9.453865336658355,
+      "grad_norm": 6.1540846824646,
+      "learning_rate": 1.055062344139651e-05,
+      "loss": 0.3386,
+      "step": 37910
+    },
+    {
+      "epoch": 9.45635910224439,
+      "grad_norm": 4.934481620788574,
+      "learning_rate": 1.0548129675810474e-05,
+      "loss": 0.3195,
+      "step": 37920
+    },
+    {
+      "epoch": 9.458852867830425,
+      "grad_norm": 7.763730525970459,
+      "learning_rate": 1.054563591022444e-05,
+      "loss": 0.3492,
+      "step": 37930
+    },
+    {
+      "epoch": 9.46134663341646,
+      "grad_norm": 10.117830276489258,
+      "learning_rate": 1.0543142144638405e-05,
+      "loss": 0.3134,
+      "step": 37940
+    },
+    {
+      "epoch": 9.463840399002494,
+      "grad_norm": 6.80835485458374,
+      "learning_rate": 1.054064837905237e-05,
+      "loss": 0.3832,
+      "step": 37950
+    },
+    {
+      "epoch": 9.46633416458853,
+      "grad_norm": 6.438353538513184,
+      "learning_rate": 1.0538154613466335e-05,
+      "loss": 0.3753,
+      "step": 37960
+    },
+    {
+      "epoch": 9.468827930174564,
+      "grad_norm": 7.7765278816223145,
+      "learning_rate": 1.05356608478803e-05,
+      "loss": 0.342,
+      "step": 37970
+    },
+    {
+      "epoch": 9.471321695760599,
+      "grad_norm": 7.024895191192627,
+      "learning_rate": 1.0533167082294264e-05,
+      "loss": 0.3745,
+      "step": 37980
+    },
+    {
+      "epoch": 9.473815461346634,
+      "grad_norm": 9.783491134643555,
+      "learning_rate": 1.0530673316708231e-05,
+      "loss": 0.3311,
+      "step": 37990
+    },
+    {
+      "epoch": 9.476309226932669,
+      "grad_norm": 7.7284979820251465,
+      "learning_rate": 1.0528179551122195e-05,
+      "loss": 0.3344,
+      "step": 38000
+    },
+    {
+      "epoch": 9.478802992518704,
+      "grad_norm": 5.616584777832031,
+      "learning_rate": 1.0525685785536159e-05,
+      "loss": 0.3145,
+      "step": 38010
+    },
+    {
+      "epoch": 9.481296758104738,
+      "grad_norm": 7.88400936126709,
+      "learning_rate": 1.0523192019950126e-05,
+      "loss": 0.332,
+      "step": 38020
+    },
+    {
+      "epoch": 9.483790523690773,
+      "grad_norm": 4.7109222412109375,
+      "learning_rate": 1.0520698254364091e-05,
+      "loss": 0.3345,
+      "step": 38030
+    },
+    {
+      "epoch": 9.486284289276808,
+      "grad_norm": 7.7549543380737305,
+      "learning_rate": 1.0518204488778056e-05,
+      "loss": 0.3887,
+      "step": 38040
+    },
+    {
+      "epoch": 9.488778054862843,
+      "grad_norm": 13.061963081359863,
+      "learning_rate": 1.0515710723192022e-05,
+      "loss": 0.3815,
+      "step": 38050
+    },
+    {
+      "epoch": 9.491271820448878,
+      "grad_norm": 7.7246479988098145,
+      "learning_rate": 1.0513216957605985e-05,
+      "loss": 0.2914,
+      "step": 38060
+    },
+    {
+      "epoch": 9.493765586034913,
+      "grad_norm": 9.497514724731445,
+      "learning_rate": 1.0510723192019952e-05,
+      "loss": 0.3772,
+      "step": 38070
+    },
+    {
+      "epoch": 9.496259351620948,
+      "grad_norm": 10.470547676086426,
+      "learning_rate": 1.0508229426433916e-05,
+      "loss": 0.3788,
+      "step": 38080
+    },
+    {
+      "epoch": 9.498753117206983,
+      "grad_norm": 5.6805009841918945,
+      "learning_rate": 1.0505735660847883e-05,
+      "loss": 0.3683,
+      "step": 38090
+    },
+    {
+      "epoch": 9.501246882793017,
+      "grad_norm": 8.353499412536621,
+      "learning_rate": 1.0503241895261846e-05,
+      "loss": 0.3531,
+      "step": 38100
+    },
+    {
+      "epoch": 9.503740648379052,
+      "grad_norm": 5.944037914276123,
+      "learning_rate": 1.050074812967581e-05,
+      "loss": 0.3669,
+      "step": 38110
+    },
+    {
+      "epoch": 9.506234413965087,
+      "grad_norm": 10.917936325073242,
+      "learning_rate": 1.0498254364089777e-05,
+      "loss": 0.383,
+      "step": 38120
+    },
+    {
+      "epoch": 9.508728179551122,
+      "grad_norm": 8.01206111907959,
+      "learning_rate": 1.049576059850374e-05,
+      "loss": 0.3929,
+      "step": 38130
+    },
+    {
+      "epoch": 9.511221945137157,
+      "grad_norm": 8.861490249633789,
+      "learning_rate": 1.0493266832917706e-05,
+      "loss": 0.3805,
+      "step": 38140
+    },
+    {
+      "epoch": 9.513715710723192,
+      "grad_norm": 5.089669227600098,
+      "learning_rate": 1.0490773067331673e-05,
+      "loss": 0.3602,
+      "step": 38150
+    },
+    {
+      "epoch": 9.516209476309227,
+      "grad_norm": 7.066434383392334,
+      "learning_rate": 1.0488279301745637e-05,
+      "loss": 0.376,
+      "step": 38160
+    },
+    {
+      "epoch": 9.518703241895262,
+      "grad_norm": 6.961791038513184,
+      "learning_rate": 1.0485785536159604e-05,
+      "loss": 0.3763,
+      "step": 38170
+    },
+    {
+      "epoch": 9.521197007481296,
+      "grad_norm": 9.394706726074219,
+      "learning_rate": 1.0483291770573567e-05,
+      "loss": 0.3548,
+      "step": 38180
+    },
+    {
+      "epoch": 9.523690773067331,
+      "grad_norm": 8.510271072387695,
+      "learning_rate": 1.048079800498753e-05,
+      "loss": 0.3591,
+      "step": 38190
+    },
+    {
+      "epoch": 9.526184538653366,
+      "grad_norm": 6.186204433441162,
+      "learning_rate": 1.0478304239401498e-05,
+      "loss": 0.3024,
+      "step": 38200
+    },
+    {
+      "epoch": 9.528678304239401,
+      "grad_norm": 8.06693172454834,
+      "learning_rate": 1.0475810473815461e-05,
+      "loss": 0.3328,
+      "step": 38210
+    },
+    {
+      "epoch": 9.531172069825436,
+      "grad_norm": 5.462076663970947,
+      "learning_rate": 1.0473316708229427e-05,
+      "loss": 0.3593,
+      "step": 38220
+    },
+    {
+      "epoch": 9.53366583541147,
+      "grad_norm": 9.915395736694336,
+      "learning_rate": 1.0470822942643392e-05,
+      "loss": 0.3356,
+      "step": 38230
+    },
+    {
+      "epoch": 9.536159600997506,
+      "grad_norm": 7.561076641082764,
+      "learning_rate": 1.0468329177057357e-05,
+      "loss": 0.4131,
+      "step": 38240
+    },
+    {
+      "epoch": 9.53865336658354,
+      "grad_norm": 11.387033462524414,
+      "learning_rate": 1.0465835411471323e-05,
+      "loss": 0.3629,
+      "step": 38250
+    },
+    {
+      "epoch": 9.541147132169575,
+      "grad_norm": 6.235090255737305,
+      "learning_rate": 1.0463341645885288e-05,
+      "loss": 0.2799,
+      "step": 38260
+    },
+    {
+      "epoch": 9.54364089775561,
+      "grad_norm": 6.4282989501953125,
+      "learning_rate": 1.0460847880299252e-05,
+      "loss": 0.3212,
+      "step": 38270
+    },
+    {
+      "epoch": 9.546134663341645,
+      "grad_norm": 6.645716667175293,
+      "learning_rate": 1.0458354114713219e-05,
+      "loss": 0.3519,
+      "step": 38280
+    },
+    {
+      "epoch": 9.548628428927682,
+      "grad_norm": 8.45456314086914,
+      "learning_rate": 1.0455860349127182e-05,
+      "loss": 0.3729,
+      "step": 38290
+    },
+    {
+      "epoch": 9.551122194513717,
+      "grad_norm": 5.648975849151611,
+      "learning_rate": 1.045336658354115e-05,
+      "loss": 0.3289,
+      "step": 38300
+    },
+    {
+      "epoch": 9.553615960099751,
+      "grad_norm": 9.984889030456543,
+      "learning_rate": 1.0450872817955113e-05,
+      "loss": 0.4206,
+      "step": 38310
+    },
+    {
+      "epoch": 9.556109725685786,
+      "grad_norm": 4.996423721313477,
+      "learning_rate": 1.0448379052369078e-05,
+      "loss": 0.2972,
+      "step": 38320
+    },
+    {
+      "epoch": 9.558603491271821,
+      "grad_norm": 5.108618259429932,
+      "learning_rate": 1.0445885286783043e-05,
+      "loss": 0.3033,
+      "step": 38330
+    },
+    {
+      "epoch": 9.561097256857856,
+      "grad_norm": 6.292821884155273,
+      "learning_rate": 1.0443391521197009e-05,
+      "loss": 0.3105,
+      "step": 38340
+    },
+    {
+      "epoch": 9.563591022443891,
+      "grad_norm": 7.772547245025635,
+      "learning_rate": 1.0440897755610972e-05,
+      "loss": 0.3646,
+      "step": 38350
+    },
+    {
+      "epoch": 9.566084788029926,
+      "grad_norm": 6.700054168701172,
+      "learning_rate": 1.043840399002494e-05,
+      "loss": 0.3842,
+      "step": 38360
+    },
+    {
+      "epoch": 9.56857855361596,
+      "grad_norm": 5.910054683685303,
+      "learning_rate": 1.0435910224438903e-05,
+      "loss": 0.3718,
+      "step": 38370
+    },
+    {
+      "epoch": 9.571072319201996,
+      "grad_norm": 5.428488731384277,
+      "learning_rate": 1.043341645885287e-05,
+      "loss": 0.3454,
+      "step": 38380
+    },
+    {
+      "epoch": 9.57356608478803,
+      "grad_norm": 7.547824382781982,
+      "learning_rate": 1.0430922693266834e-05,
+      "loss": 0.3771,
+      "step": 38390
+    },
+    {
+      "epoch": 9.576059850374065,
+      "grad_norm": 5.2548909187316895,
+      "learning_rate": 1.0428428927680799e-05,
+      "loss": 0.3896,
+      "step": 38400
+    },
+    {
+      "epoch": 9.5785536159601,
+      "grad_norm": 6.828177452087402,
+      "learning_rate": 1.0425935162094764e-05,
+      "loss": 0.3847,
+      "step": 38410
+    },
+    {
+      "epoch": 9.581047381546135,
+      "grad_norm": 8.20153522491455,
+      "learning_rate": 1.042344139650873e-05,
+      "loss": 0.3524,
+      "step": 38420
+    },
+    {
+      "epoch": 9.58354114713217,
+      "grad_norm": 4.548335075378418,
+      "learning_rate": 1.0420947630922693e-05,
+      "loss": 0.3613,
+      "step": 38430
+    },
+    {
+      "epoch": 9.586034912718205,
+      "grad_norm": 6.978244304656982,
+      "learning_rate": 1.041845386533666e-05,
+      "loss": 0.3803,
+      "step": 38440
+    },
+    {
+      "epoch": 9.58852867830424,
+      "grad_norm": 7.369738578796387,
+      "learning_rate": 1.0415960099750624e-05,
+      "loss": 0.3134,
+      "step": 38450
+    },
+    {
+      "epoch": 9.591022443890274,
+      "grad_norm": 8.234667778015137,
+      "learning_rate": 1.041346633416459e-05,
+      "loss": 0.3537,
+      "step": 38460
+    },
+    {
+      "epoch": 9.59351620947631,
+      "grad_norm": 6.086751461029053,
+      "learning_rate": 1.0410972568578554e-05,
+      "loss": 0.3783,
+      "step": 38470
+    },
+    {
+      "epoch": 9.596009975062344,
+      "grad_norm": 7.9929046630859375,
+      "learning_rate": 1.0408478802992518e-05,
+      "loss": 0.3491,
+      "step": 38480
+    },
+    {
+      "epoch": 9.598503740648379,
+      "grad_norm": 11.648649215698242,
+      "learning_rate": 1.0405985037406485e-05,
+      "loss": 0.3418,
+      "step": 38490
+    },
+    {
+      "epoch": 9.600997506234414,
+      "grad_norm": 8.06251049041748,
+      "learning_rate": 1.040349127182045e-05,
+      "loss": 0.3483,
+      "step": 38500
+    },
+    {
+      "epoch": 9.603491271820449,
+      "grad_norm": 9.651390075683594,
+      "learning_rate": 1.0400997506234414e-05,
+      "loss": 0.3574,
+      "step": 38510
+    },
+    {
+      "epoch": 9.605985037406484,
+      "grad_norm": 4.3901777267456055,
+      "learning_rate": 1.0398503740648381e-05,
+      "loss": 0.3304,
+      "step": 38520
+    },
+    {
+      "epoch": 9.608478802992519,
+      "grad_norm": 5.272586345672607,
+      "learning_rate": 1.0396009975062345e-05,
+      "loss": 0.2917,
+      "step": 38530
+    },
+    {
+      "epoch": 9.610972568578553,
+      "grad_norm": 9.170848846435547,
+      "learning_rate": 1.0393516209476312e-05,
+      "loss": 0.2899,
+      "step": 38540
+    },
+    {
+      "epoch": 9.613466334164588,
+      "grad_norm": 7.961601257324219,
+      "learning_rate": 1.0391022443890275e-05,
+      "loss": 0.3714,
+      "step": 38550
+    },
+    {
+      "epoch": 9.615960099750623,
+      "grad_norm": 8.155771255493164,
+      "learning_rate": 1.0388528678304239e-05,
+      "loss": 0.3591,
+      "step": 38560
+    },
+    {
+      "epoch": 9.618453865336658,
+      "grad_norm": 7.194378852844238,
+      "learning_rate": 1.0386034912718206e-05,
+      "loss": 0.355,
+      "step": 38570
+    },
+    {
+      "epoch": 9.620947630922693,
+      "grad_norm": 8.259054183959961,
+      "learning_rate": 1.038354114713217e-05,
+      "loss": 0.2957,
+      "step": 38580
+    },
+    {
+      "epoch": 9.623441396508728,
+      "grad_norm": 6.669407844543457,
+      "learning_rate": 1.0381047381546136e-05,
+      "loss": 0.3607,
+      "step": 38590
+    },
+    {
+      "epoch": 9.625935162094763,
+      "grad_norm": 7.477216720581055,
+      "learning_rate": 1.03785536159601e-05,
+      "loss": 0.3427,
+      "step": 38600
+    },
+    {
+      "epoch": 9.628428927680797,
+      "grad_norm": 5.8216986656188965,
+      "learning_rate": 1.0376059850374065e-05,
+      "loss": 0.3458,
+      "step": 38610
+    },
+    {
+      "epoch": 9.630922693266832,
+      "grad_norm": 10.364824295043945,
+      "learning_rate": 1.0373566084788032e-05,
+      "loss": 0.3817,
+      "step": 38620
+    },
+    {
+      "epoch": 9.633416458852867,
+      "grad_norm": 5.567168235778809,
+      "learning_rate": 1.0371072319201996e-05,
+      "loss": 0.2853,
+      "step": 38630
+    },
+    {
+      "epoch": 9.635910224438902,
+      "grad_norm": 7.306951999664307,
+      "learning_rate": 1.036857855361596e-05,
+      "loss": 0.3538,
+      "step": 38640
+    },
+    {
+      "epoch": 9.638403990024937,
+      "grad_norm": 8.60983943939209,
+      "learning_rate": 1.0366084788029927e-05,
+      "loss": 0.3719,
+      "step": 38650
+    },
+    {
+      "epoch": 9.640897755610972,
+      "grad_norm": 8.791508674621582,
+      "learning_rate": 1.036359102244389e-05,
+      "loss": 0.4041,
+      "step": 38660
+    },
+    {
+      "epoch": 9.643391521197007,
+      "grad_norm": 6.479540824890137,
+      "learning_rate": 1.0361097256857857e-05,
+      "loss": 0.3139,
+      "step": 38670
+    },
+    {
+      "epoch": 9.645885286783042,
+      "grad_norm": 7.419809818267822,
+      "learning_rate": 1.035860349127182e-05,
+      "loss": 0.3653,
+      "step": 38680
+    },
+    {
+      "epoch": 9.648379052369076,
+      "grad_norm": 11.746686935424805,
+      "learning_rate": 1.0356109725685786e-05,
+      "loss": 0.3622,
+      "step": 38690
+    },
+    {
+      "epoch": 9.650872817955111,
+      "grad_norm": 9.593301773071289,
+      "learning_rate": 1.0353615960099751e-05,
+      "loss": 0.3942,
+      "step": 38700
+    },
+    {
+      "epoch": 9.653366583541148,
+      "grad_norm": 5.149587154388428,
+      "learning_rate": 1.0351122194513717e-05,
+      "loss": 0.3576,
+      "step": 38710
+    },
+    {
+      "epoch": 9.655860349127183,
+      "grad_norm": 5.903531551361084,
+      "learning_rate": 1.034862842892768e-05,
+      "loss": 0.3395,
+      "step": 38720
+    },
+    {
+      "epoch": 9.658354114713218,
+      "grad_norm": 7.770636558532715,
+      "learning_rate": 1.0346134663341647e-05,
+      "loss": 0.2986,
+      "step": 38730
+    },
+    {
+      "epoch": 9.660847880299253,
+      "grad_norm": 8.158554077148438,
+      "learning_rate": 1.0343640897755611e-05,
+      "loss": 0.3346,
+      "step": 38740
+    },
+    {
+      "epoch": 9.663341645885287,
+      "grad_norm": 5.318978786468506,
+      "learning_rate": 1.0341147132169578e-05,
+      "loss": 0.312,
+      "step": 38750
+    },
+    {
+      "epoch": 9.665835411471322,
+      "grad_norm": 7.235579490661621,
+      "learning_rate": 1.0338653366583542e-05,
+      "loss": 0.3416,
+      "step": 38760
+    },
+    {
+      "epoch": 9.668329177057357,
+      "grad_norm": 6.485595226287842,
+      "learning_rate": 1.0336159600997507e-05,
+      "loss": 0.335,
+      "step": 38770
+    },
+    {
+      "epoch": 9.670822942643392,
+      "grad_norm": 8.763544082641602,
+      "learning_rate": 1.0333665835411472e-05,
+      "loss": 0.3305,
+      "step": 38780
+    },
+    {
+      "epoch": 9.673316708229427,
+      "grad_norm": 11.1206693649292,
+      "learning_rate": 1.0331172069825438e-05,
+      "loss": 0.3349,
+      "step": 38790
+    },
+    {
+      "epoch": 9.675810473815462,
+      "grad_norm": 9.942346572875977,
+      "learning_rate": 1.0328678304239403e-05,
+      "loss": 0.3211,
+      "step": 38800
+    },
+    {
+      "epoch": 9.678304239401497,
+      "grad_norm": 7.735063552856445,
+      "learning_rate": 1.0326184538653368e-05,
+      "loss": 0.3628,
+      "step": 38810
+    },
+    {
+      "epoch": 9.680798004987532,
+      "grad_norm": 8.735753059387207,
+      "learning_rate": 1.0323690773067332e-05,
+      "loss": 0.3722,
+      "step": 38820
+    },
+    {
+      "epoch": 9.683291770573566,
+      "grad_norm": 8.924358367919922,
+      "learning_rate": 1.0321197007481299e-05,
+      "loss": 0.3788,
+      "step": 38830
+    },
+    {
+      "epoch": 9.685785536159601,
+      "grad_norm": 10.292500495910645,
+      "learning_rate": 1.0318703241895262e-05,
+      "loss": 0.4222,
+      "step": 38840
+    },
+    {
+      "epoch": 9.688279301745636,
+      "grad_norm": 5.393009662628174,
+      "learning_rate": 1.0316209476309228e-05,
+      "loss": 0.4147,
+      "step": 38850
+    },
+    {
+      "epoch": 9.690773067331671,
+      "grad_norm": 6.392873287200928,
+      "learning_rate": 1.0313715710723193e-05,
+      "loss": 0.2981,
+      "step": 38860
+    },
+    {
+      "epoch": 9.693266832917706,
+      "grad_norm": 5.524891376495361,
+      "learning_rate": 1.0311221945137158e-05,
+      "loss": 0.3877,
+      "step": 38870
+    },
+    {
+      "epoch": 9.69576059850374,
+      "grad_norm": 7.086117744445801,
+      "learning_rate": 1.0308728179551124e-05,
+      "loss": 0.3449,
+      "step": 38880
+    },
+    {
+      "epoch": 9.698254364089776,
+      "grad_norm": 6.8754072189331055,
+      "learning_rate": 1.0306234413965089e-05,
+      "loss": 0.365,
+      "step": 38890
+    },
+    {
+      "epoch": 9.70074812967581,
+      "grad_norm": 8.261271476745605,
+      "learning_rate": 1.0303740648379053e-05,
+      "loss": 0.2822,
+      "step": 38900
+    },
+    {
+      "epoch": 9.703241895261845,
+      "grad_norm": 8.90969467163086,
+      "learning_rate": 1.030124688279302e-05,
+      "loss": 0.4126,
+      "step": 38910
+    },
+    {
+      "epoch": 9.70573566084788,
+      "grad_norm": 6.020090579986572,
+      "learning_rate": 1.0298753117206983e-05,
+      "loss": 0.3471,
+      "step": 38920
+    },
+    {
+      "epoch": 9.708229426433915,
+      "grad_norm": 7.7385993003845215,
+      "learning_rate": 1.0296259351620947e-05,
+      "loss": 0.3681,
+      "step": 38930
+    },
+    {
+      "epoch": 9.71072319201995,
+      "grad_norm": 8.860572814941406,
+      "learning_rate": 1.0293765586034914e-05,
+      "loss": 0.435,
+      "step": 38940
+    },
+    {
+      "epoch": 9.713216957605985,
+      "grad_norm": 8.749921798706055,
+      "learning_rate": 1.0291271820448877e-05,
+      "loss": 0.3425,
+      "step": 38950
+    },
+    {
+      "epoch": 9.71571072319202,
+      "grad_norm": 8.579872131347656,
+      "learning_rate": 1.0288778054862844e-05,
+      "loss": 0.3983,
+      "step": 38960
+    },
+    {
+      "epoch": 9.718204488778055,
+      "grad_norm": 7.9669389724731445,
+      "learning_rate": 1.028628428927681e-05,
+      "loss": 0.3475,
+      "step": 38970
+    },
+    {
+      "epoch": 9.72069825436409,
+      "grad_norm": 7.299527645111084,
+      "learning_rate": 1.0283790523690773e-05,
+      "loss": 0.3911,
+      "step": 38980
+    },
+    {
+      "epoch": 9.723192019950124,
+      "grad_norm": 5.163363933563232,
+      "learning_rate": 1.028129675810474e-05,
+      "loss": 0.361,
+      "step": 38990
+    },
+    {
+      "epoch": 9.72568578553616,
+      "grad_norm": 8.54089641571045,
+      "learning_rate": 1.0278802992518704e-05,
+      "loss": 0.33,
+      "step": 39000
+    },
+    {
+      "epoch": 9.728179551122194,
+      "grad_norm": 7.269198894500732,
+      "learning_rate": 1.0276309226932668e-05,
+      "loss": 0.3695,
+      "step": 39010
+    },
+    {
+      "epoch": 9.730673316708229,
+      "grad_norm": 8.969745635986328,
+      "learning_rate": 1.0273815461346635e-05,
+      "loss": 0.365,
+      "step": 39020
+    },
+    {
+      "epoch": 9.733167082294264,
+      "grad_norm": 5.630902290344238,
+      "learning_rate": 1.0271321695760598e-05,
+      "loss": 0.3615,
+      "step": 39030
+    },
+    {
+      "epoch": 9.735660847880299,
+      "grad_norm": 8.235981941223145,
+      "learning_rate": 1.0268827930174565e-05,
+      "loss": 0.4206,
+      "step": 39040
+    },
+    {
+      "epoch": 9.738154613466333,
+      "grad_norm": 7.914098262786865,
+      "learning_rate": 1.0266334164588529e-05,
+      "loss": 0.343,
+      "step": 39050
+    },
+    {
+      "epoch": 9.740648379052368,
+      "grad_norm": 7.478903293609619,
+      "learning_rate": 1.0263840399002494e-05,
+      "loss": 0.3678,
+      "step": 39060
+    },
+    {
+      "epoch": 9.743142144638403,
+      "grad_norm": 7.806370735168457,
+      "learning_rate": 1.026134663341646e-05,
+      "loss": 0.3855,
+      "step": 39070
+    },
+    {
+      "epoch": 9.745635910224438,
+      "grad_norm": 6.9918341636657715,
+      "learning_rate": 1.0258852867830425e-05,
+      "loss": 0.3848,
+      "step": 39080
+    },
+    {
+      "epoch": 9.748129675810475,
+      "grad_norm": 7.807930946350098,
+      "learning_rate": 1.0256359102244392e-05,
+      "loss": 0.3518,
+      "step": 39090
+    },
+    {
+      "epoch": 9.75062344139651,
+      "grad_norm": 5.30518913269043,
+      "learning_rate": 1.0253865336658355e-05,
+      "loss": 0.3512,
+      "step": 39100
+    },
+    {
+      "epoch": 9.753117206982544,
+      "grad_norm": 7.166024208068848,
+      "learning_rate": 1.0251371571072319e-05,
+      "loss": 0.4185,
+      "step": 39110
+    },
+    {
+      "epoch": 9.75561097256858,
+      "grad_norm": 7.371334075927734,
+      "learning_rate": 1.0248877805486286e-05,
+      "loss": 0.385,
+      "step": 39120
+    },
+    {
+      "epoch": 9.758104738154614,
+      "grad_norm": 5.414678573608398,
+      "learning_rate": 1.024638403990025e-05,
+      "loss": 0.3971,
+      "step": 39130
+    },
+    {
+      "epoch": 9.760598503740649,
+      "grad_norm": 8.119851112365723,
+      "learning_rate": 1.0243890274314215e-05,
+      "loss": 0.3534,
+      "step": 39140
+    },
+    {
+      "epoch": 9.763092269326684,
+      "grad_norm": 10.847015380859375,
+      "learning_rate": 1.024139650872818e-05,
+      "loss": 0.4538,
+      "step": 39150
+    },
+    {
+      "epoch": 9.765586034912719,
+      "grad_norm": 7.289000511169434,
+      "learning_rate": 1.0238902743142145e-05,
+      "loss": 0.3739,
+      "step": 39160
+    },
+    {
+      "epoch": 9.768079800498754,
+      "grad_norm": 9.922390937805176,
+      "learning_rate": 1.023640897755611e-05,
+      "loss": 0.3072,
+      "step": 39170
+    },
+    {
+      "epoch": 9.770573566084789,
+      "grad_norm": 8.990649223327637,
+      "learning_rate": 1.0233915211970076e-05,
+      "loss": 0.3926,
+      "step": 39180
+    },
+    {
+      "epoch": 9.773067331670823,
+      "grad_norm": 6.73281717300415,
+      "learning_rate": 1.023142144638404e-05,
+      "loss": 0.3652,
+      "step": 39190
+    },
+    {
+      "epoch": 9.775561097256858,
+      "grad_norm": 6.121176242828369,
+      "learning_rate": 1.0228927680798007e-05,
+      "loss": 0.3424,
+      "step": 39200
+    },
+    {
+      "epoch": 9.778054862842893,
+      "grad_norm": 10.394766807556152,
+      "learning_rate": 1.022643391521197e-05,
+      "loss": 0.3365,
+      "step": 39210
+    },
+    {
+      "epoch": 9.780548628428928,
+      "grad_norm": 6.58123254776001,
+      "learning_rate": 1.0223940149625936e-05,
+      "loss": 0.3641,
+      "step": 39220
+    },
+    {
+      "epoch": 9.783042394014963,
+      "grad_norm": 6.802944183349609,
+      "learning_rate": 1.0221446384039901e-05,
+      "loss": 0.2725,
+      "step": 39230
+    },
+    {
+      "epoch": 9.785536159600998,
+      "grad_norm": 8.817499160766602,
+      "learning_rate": 1.0218952618453866e-05,
+      "loss": 0.3455,
+      "step": 39240
+    },
+    {
+      "epoch": 9.788029925187033,
+      "grad_norm": 9.854268074035645,
+      "learning_rate": 1.0216458852867832e-05,
+      "loss": 0.3698,
+      "step": 39250
+    },
+    {
+      "epoch": 9.790523690773068,
+      "grad_norm": 7.4926323890686035,
+      "learning_rate": 1.0213965087281797e-05,
+      "loss": 0.4717,
+      "step": 39260
+    },
+    {
+      "epoch": 9.793017456359102,
+      "grad_norm": 5.085393905639648,
+      "learning_rate": 1.021147132169576e-05,
+      "loss": 0.2991,
+      "step": 39270
+    },
+    {
+      "epoch": 9.795511221945137,
+      "grad_norm": 9.098230361938477,
+      "learning_rate": 1.0208977556109728e-05,
+      "loss": 0.2923,
+      "step": 39280
+    },
+    {
+      "epoch": 9.798004987531172,
+      "grad_norm": 10.33260440826416,
+      "learning_rate": 1.0206483790523691e-05,
+      "loss": 0.4181,
+      "step": 39290
+    },
+    {
+      "epoch": 9.800498753117207,
+      "grad_norm": 8.845526695251465,
+      "learning_rate": 1.0203990024937658e-05,
+      "loss": 0.4501,
+      "step": 39300
+    },
+    {
+      "epoch": 9.802992518703242,
+      "grad_norm": 11.282034873962402,
+      "learning_rate": 1.0201496259351622e-05,
+      "loss": 0.3389,
+      "step": 39310
+    },
+    {
+      "epoch": 9.805486284289277,
+      "grad_norm": 7.388114929199219,
+      "learning_rate": 1.0199002493765587e-05,
+      "loss": 0.3754,
+      "step": 39320
+    },
+    {
+      "epoch": 9.807980049875312,
+      "grad_norm": 7.895745754241943,
+      "learning_rate": 1.0196508728179552e-05,
+      "loss": 0.3507,
+      "step": 39330
+    },
+    {
+      "epoch": 9.810473815461346,
+      "grad_norm": 7.767431259155273,
+      "learning_rate": 1.0194014962593518e-05,
+      "loss": 0.4053,
+      "step": 39340
+    },
+    {
+      "epoch": 9.812967581047381,
+      "grad_norm": 6.963127613067627,
+      "learning_rate": 1.0191521197007481e-05,
+      "loss": 0.3367,
+      "step": 39350
+    },
+    {
+      "epoch": 9.815461346633416,
+      "grad_norm": 5.868590354919434,
+      "learning_rate": 1.0189027431421448e-05,
+      "loss": 0.3903,
+      "step": 39360
+    },
+    {
+      "epoch": 9.817955112219451,
+      "grad_norm": 8.267974853515625,
+      "learning_rate": 1.0186533665835412e-05,
+      "loss": 0.3503,
+      "step": 39370
+    },
+    {
+      "epoch": 9.820448877805486,
+      "grad_norm": 6.432572841644287,
+      "learning_rate": 1.0184039900249379e-05,
+      "loss": 0.3092,
+      "step": 39380
+    },
+    {
+      "epoch": 9.82294264339152,
+      "grad_norm": 9.071556091308594,
+      "learning_rate": 1.0181546134663343e-05,
+      "loss": 0.4478,
+      "step": 39390
+    },
+    {
+      "epoch": 9.825436408977556,
+      "grad_norm": 8.835619926452637,
+      "learning_rate": 1.0179052369077306e-05,
+      "loss": 0.3756,
+      "step": 39400
+    },
+    {
+      "epoch": 9.82793017456359,
+      "grad_norm": 10.740196228027344,
+      "learning_rate": 1.0176558603491273e-05,
+      "loss": 0.3349,
+      "step": 39410
+    },
+    {
+      "epoch": 9.830423940149625,
+      "grad_norm": 11.372931480407715,
+      "learning_rate": 1.0174064837905237e-05,
+      "loss": 0.403,
+      "step": 39420
+    },
+    {
+      "epoch": 9.83291770573566,
+      "grad_norm": 9.308483123779297,
+      "learning_rate": 1.0171571072319202e-05,
+      "loss": 0.3523,
+      "step": 39430
+    },
+    {
+      "epoch": 9.835411471321695,
+      "grad_norm": 6.578537464141846,
+      "learning_rate": 1.0169077306733169e-05,
+      "loss": 0.324,
+      "step": 39440
+    },
+    {
+      "epoch": 9.83790523690773,
+      "grad_norm": 6.254716396331787,
+      "learning_rate": 1.0166583541147133e-05,
+      "loss": 0.3546,
+      "step": 39450
+    },
+    {
+      "epoch": 9.840399002493765,
+      "grad_norm": 6.601047992706299,
+      "learning_rate": 1.01640897755611e-05,
+      "loss": 0.405,
+      "step": 39460
+    },
+    {
+      "epoch": 9.8428927680798,
+      "grad_norm": 9.484192848205566,
+      "learning_rate": 1.0161596009975063e-05,
+      "loss": 0.3502,
+      "step": 39470
+    },
+    {
+      "epoch": 9.845386533665835,
+      "grad_norm": 7.499188423156738,
+      "learning_rate": 1.0159102244389027e-05,
+      "loss": 0.3508,
+      "step": 39480
+    },
+    {
+      "epoch": 9.84788029925187,
+      "grad_norm": 5.97933292388916,
+      "learning_rate": 1.0156608478802994e-05,
+      "loss": 0.3444,
+      "step": 39490
+    },
+    {
+      "epoch": 9.850374064837904,
+      "grad_norm": 7.985541343688965,
+      "learning_rate": 1.0154114713216958e-05,
+      "loss": 0.3577,
+      "step": 39500
+    },
+    {
+      "epoch": 9.85286783042394,
+      "grad_norm": 9.508009910583496,
+      "learning_rate": 1.0151620947630923e-05,
+      "loss": 0.2855,
+      "step": 39510
+    },
+    {
+      "epoch": 9.855361596009976,
+      "grad_norm": 7.971591472625732,
+      "learning_rate": 1.0149127182044888e-05,
+      "loss": 0.3184,
+      "step": 39520
+    },
+    {
+      "epoch": 9.85785536159601,
+      "grad_norm": 6.7001848220825195,
+      "learning_rate": 1.0146633416458853e-05,
+      "loss": 0.3265,
+      "step": 39530
+    },
+    {
+      "epoch": 9.860349127182046,
+      "grad_norm": 5.2400689125061035,
+      "learning_rate": 1.0144139650872819e-05,
+      "loss": 0.3971,
+      "step": 39540
+    },
+    {
+      "epoch": 9.86284289276808,
+      "grad_norm": 7.525379657745361,
+      "learning_rate": 1.0141645885286784e-05,
+      "loss": 0.384,
+      "step": 39550
+    },
+    {
+      "epoch": 9.865336658354115,
+      "grad_norm": 6.4652628898620605,
+      "learning_rate": 1.0139152119700748e-05,
+      "loss": 0.4057,
+      "step": 39560
+    },
+    {
+      "epoch": 9.86783042394015,
+      "grad_norm": 7.187132835388184,
+      "learning_rate": 1.0136658354114715e-05,
+      "loss": 0.366,
+      "step": 39570
+    },
+    {
+      "epoch": 9.870324189526185,
+      "grad_norm": 7.20662784576416,
+      "learning_rate": 1.0134164588528678e-05,
+      "loss": 0.344,
+      "step": 39580
+    },
+    {
+      "epoch": 9.87281795511222,
+      "grad_norm": 5.796775817871094,
+      "learning_rate": 1.0131670822942645e-05,
+      "loss": 0.3455,
+      "step": 39590
+    },
+    {
+      "epoch": 9.875311720698255,
+      "grad_norm": 5.326415061950684,
+      "learning_rate": 1.0129177057356609e-05,
+      "loss": 0.3139,
+      "step": 39600
+    },
+    {
+      "epoch": 9.87780548628429,
+      "grad_norm": 9.870059967041016,
+      "learning_rate": 1.0126683291770574e-05,
+      "loss": 0.4385,
+      "step": 39610
+    },
+    {
+      "epoch": 9.880299251870325,
+      "grad_norm": 5.55562686920166,
+      "learning_rate": 1.012418952618454e-05,
+      "loss": 0.3718,
+      "step": 39620
+    },
+    {
+      "epoch": 9.88279301745636,
+      "grad_norm": 6.446163177490234,
+      "learning_rate": 1.0121695760598505e-05,
+      "loss": 0.4105,
+      "step": 39630
+    },
+    {
+      "epoch": 9.885286783042394,
+      "grad_norm": 7.71967887878418,
+      "learning_rate": 1.0119201995012469e-05,
+      "loss": 0.3544,
+      "step": 39640
+    },
+    {
+      "epoch": 9.88778054862843,
+      "grad_norm": 6.094079494476318,
+      "learning_rate": 1.0116708229426436e-05,
+      "loss": 0.3683,
+      "step": 39650
+    },
+    {
+      "epoch": 9.890274314214464,
+      "grad_norm": 6.991535186767578,
+      "learning_rate": 1.0114214463840399e-05,
+      "loss": 0.3176,
+      "step": 39660
+    },
+    {
+      "epoch": 9.892768079800499,
+      "grad_norm": 6.277289867401123,
+      "learning_rate": 1.0111720698254366e-05,
+      "loss": 0.3603,
+      "step": 39670
+    },
+    {
+      "epoch": 9.895261845386534,
+      "grad_norm": 5.89650821685791,
+      "learning_rate": 1.010922693266833e-05,
+      "loss": 0.3687,
+      "step": 39680
+    },
+    {
+      "epoch": 9.897755610972569,
+      "grad_norm": 8.910008430480957,
+      "learning_rate": 1.0106733167082295e-05,
+      "loss": 0.3066,
+      "step": 39690
+    },
+    {
+      "epoch": 9.900249376558603,
+      "grad_norm": 7.53981876373291,
+      "learning_rate": 1.010423940149626e-05,
+      "loss": 0.3586,
+      "step": 39700
+    },
+    {
+      "epoch": 9.902743142144638,
+      "grad_norm": 6.939689636230469,
+      "learning_rate": 1.0101745635910226e-05,
+      "loss": 0.3845,
+      "step": 39710
+    },
+    {
+      "epoch": 9.905236907730673,
+      "grad_norm": 7.977263927459717,
+      "learning_rate": 1.009925187032419e-05,
+      "loss": 0.4126,
+      "step": 39720
+    },
+    {
+      "epoch": 9.907730673316708,
+      "grad_norm": 7.981022357940674,
+      "learning_rate": 1.0096758104738156e-05,
+      "loss": 0.2967,
+      "step": 39730
+    },
+    {
+      "epoch": 9.910224438902743,
+      "grad_norm": 7.351691246032715,
+      "learning_rate": 1.009426433915212e-05,
+      "loss": 0.3309,
+      "step": 39740
+    },
+    {
+      "epoch": 9.912718204488778,
+      "grad_norm": 6.551850318908691,
+      "learning_rate": 1.0091770573566087e-05,
+      "loss": 0.3027,
+      "step": 39750
+    },
+    {
+      "epoch": 9.915211970074813,
+      "grad_norm": 10.421834945678711,
+      "learning_rate": 1.008927680798005e-05,
+      "loss": 0.3109,
+      "step": 39760
+    },
+    {
+      "epoch": 9.917705735660848,
+      "grad_norm": 9.007574081420898,
+      "learning_rate": 1.0086783042394014e-05,
+      "loss": 0.3231,
+      "step": 39770
+    },
+    {
+      "epoch": 9.920199501246882,
+      "grad_norm": 7.615650177001953,
+      "learning_rate": 1.0084289276807981e-05,
+      "loss": 0.3444,
+      "step": 39780
+    },
+    {
+      "epoch": 9.922693266832917,
+      "grad_norm": 5.678508758544922,
+      "learning_rate": 1.0081795511221946e-05,
+      "loss": 0.3623,
+      "step": 39790
+    },
+    {
+      "epoch": 9.925187032418952,
+      "grad_norm": 7.512736797332764,
+      "learning_rate": 1.0079301745635912e-05,
+      "loss": 0.2864,
+      "step": 39800
+    },
+    {
+      "epoch": 9.927680798004987,
+      "grad_norm": 6.589831829071045,
+      "learning_rate": 1.0076807980049877e-05,
+      "loss": 0.3472,
+      "step": 39810
+    },
+    {
+      "epoch": 9.930174563591022,
+      "grad_norm": 7.6342902183532715,
+      "learning_rate": 1.007431421446384e-05,
+      "loss": 0.3813,
+      "step": 39820
+    },
+    {
+      "epoch": 9.932668329177057,
+      "grad_norm": 11.522653579711914,
+      "learning_rate": 1.0071820448877808e-05,
+      "loss": 0.3478,
+      "step": 39830
+    },
+    {
+      "epoch": 9.935162094763092,
+      "grad_norm": 8.16284465789795,
+      "learning_rate": 1.0069326683291771e-05,
+      "loss": 0.3339,
+      "step": 39840
+    },
+    {
+      "epoch": 9.937655860349127,
+      "grad_norm": 7.553341865539551,
+      "learning_rate": 1.0066832917705735e-05,
+      "loss": 0.3129,
+      "step": 39850
+    },
+    {
+      "epoch": 9.940149625935161,
+      "grad_norm": 6.100545883178711,
+      "learning_rate": 1.0064339152119702e-05,
+      "loss": 0.3501,
+      "step": 39860
+    },
+    {
+      "epoch": 9.942643391521196,
+      "grad_norm": 5.523427486419678,
+      "learning_rate": 1.0061845386533666e-05,
+      "loss": 0.392,
+      "step": 39870
+    },
+    {
+      "epoch": 9.945137157107231,
+      "grad_norm": 6.449985980987549,
+      "learning_rate": 1.0059351620947633e-05,
+      "loss": 0.3228,
+      "step": 39880
+    },
+    {
+      "epoch": 9.947630922693268,
+      "grad_norm": 8.599003791809082,
+      "learning_rate": 1.0056857855361596e-05,
+      "loss": 0.3922,
+      "step": 39890
+    },
+    {
+      "epoch": 9.950124688279303,
+      "grad_norm": 7.818081855773926,
+      "learning_rate": 1.0054364089775561e-05,
+      "loss": 0.4501,
+      "step": 39900
+    },
+    {
+      "epoch": 9.952618453865338,
+      "grad_norm": 6.245206356048584,
+      "learning_rate": 1.0051870324189528e-05,
+      "loss": 0.3167,
+      "step": 39910
+    },
+    {
+      "epoch": 9.955112219451372,
+      "grad_norm": 9.523896217346191,
+      "learning_rate": 1.0049376558603492e-05,
+      "loss": 0.3944,
+      "step": 39920
+    },
+    {
+      "epoch": 9.957605985037407,
+      "grad_norm": 12.862037658691406,
+      "learning_rate": 1.0046882793017456e-05,
+      "loss": 0.3859,
+      "step": 39930
+    },
+    {
+      "epoch": 9.960099750623442,
+      "grad_norm": 11.528470993041992,
+      "learning_rate": 1.0044389027431423e-05,
+      "loss": 0.3911,
+      "step": 39940
+    },
+    {
+      "epoch": 9.962593516209477,
+      "grad_norm": 9.270523071289062,
+      "learning_rate": 1.0041895261845386e-05,
+      "loss": 0.4362,
+      "step": 39950
+    },
+    {
+      "epoch": 9.965087281795512,
+      "grad_norm": 9.051032066345215,
+      "learning_rate": 1.0039401496259353e-05,
+      "loss": 0.422,
+      "step": 39960
+    },
+    {
+      "epoch": 9.967581047381547,
+      "grad_norm": 4.276822566986084,
+      "learning_rate": 1.0036907730673317e-05,
+      "loss": 0.397,
+      "step": 39970
+    },
+    {
+      "epoch": 9.970074812967582,
+      "grad_norm": 4.984722137451172,
+      "learning_rate": 1.0034413965087282e-05,
+      "loss": 0.3636,
+      "step": 39980
+    },
+    {
+      "epoch": 9.972568578553616,
+      "grad_norm": 5.631036281585693,
+      "learning_rate": 1.0031920199501248e-05,
+      "loss": 0.3556,
+      "step": 39990
+    },
+    {
+      "epoch": 9.975062344139651,
+      "grad_norm": 8.992472648620605,
+      "learning_rate": 1.0029426433915213e-05,
+      "loss": 0.3561,
+      "step": 40000
+    },
+    {
+      "epoch": 9.977556109725686,
+      "grad_norm": 9.377120971679688,
+      "learning_rate": 1.0026932668329176e-05,
+      "loss": 0.444,
+      "step": 40010
+    },
+    {
+      "epoch": 9.980049875311721,
+      "grad_norm": NaN,
+      "learning_rate": 1.0024688279301746e-05,
+      "loss": 0.3719,
+      "step": 40020
+    },
+    {
+      "epoch": 9.982543640897756,
+      "grad_norm": 6.351677417755127,
+      "learning_rate": 1.002219451371571e-05,
+      "loss": 0.2969,
+      "step": 40030
+    },
+    {
+      "epoch": 9.98503740648379,
+      "grad_norm": 7.045421600341797,
+      "learning_rate": 1.0019700748129677e-05,
+      "loss": 0.3032,
+      "step": 40040
+    },
+    {
+      "epoch": 9.987531172069826,
+      "grad_norm": 7.283246994018555,
+      "learning_rate": 1.001720698254364e-05,
+      "loss": 0.4718,
+      "step": 40050
+    },
+    {
+      "epoch": 9.99002493765586,
+      "grad_norm": 6.027595520019531,
+      "learning_rate": 1.0014713216957608e-05,
+      "loss": 0.3524,
+      "step": 40060
+    },
+    {
+      "epoch": 9.992518703241895,
+      "grad_norm": 5.910287380218506,
+      "learning_rate": 1.0012219451371571e-05,
+      "loss": 0.3885,
+      "step": 40070
+    },
+    {
+      "epoch": 9.99501246882793,
+      "grad_norm": 6.286899089813232,
+      "learning_rate": 1.0009725685785537e-05,
+      "loss": 0.3322,
+      "step": 40080
+    },
+    {
+      "epoch": 9.997506234413965,
+      "grad_norm": 6.434454441070557,
+      "learning_rate": 1.0007231920199502e-05,
+      "loss": 0.3139,
+      "step": 40090
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 6.4550700187683105,
+      "learning_rate": 1.0004738154613467e-05,
+      "loss": 0.3902,
+      "step": 40100
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 0.4127441346645355,
+      "eval_runtime": 59.9961,
+      "eval_samples_per_second": 16.718,
+      "eval_steps_per_second": 16.718,
+      "step": 40100
+    },
+    {
+      "epoch": 10.002493765586035,
+      "grad_norm": 7.405590057373047,
+      "learning_rate": 1.0002244389027431e-05,
+      "loss": 0.3798,
+      "step": 40110
+    },
+    {
+      "epoch": 10.00498753117207,
+      "grad_norm": 6.414757251739502,
+      "learning_rate": 9.999750623441398e-06,
+      "loss": 0.3407,
+      "step": 40120
+    },
+    {
+      "epoch": 10.007481296758105,
+      "grad_norm": 12.407922744750977,
+      "learning_rate": 9.997256857855361e-06,
+      "loss": 0.3276,
+      "step": 40130
+    },
+    {
+      "epoch": 10.00997506234414,
+      "grad_norm": 7.273416996002197,
+      "learning_rate": 9.994763092269327e-06,
+      "loss": 0.3299,
+      "step": 40140
+    },
+    {
+      "epoch": 10.012468827930174,
+      "grad_norm": 5.8642449378967285,
+      "learning_rate": 9.992269326683292e-06,
+      "loss": 0.3227,
+      "step": 40150
+    },
+    {
+      "epoch": 10.01496259351621,
+      "grad_norm": 5.257193565368652,
+      "learning_rate": 9.989775561097257e-06,
+      "loss": 0.3175,
+      "step": 40160
+    },
+    {
+      "epoch": 10.017456359102244,
+      "grad_norm": 10.810057640075684,
+      "learning_rate": 9.987281795511223e-06,
+      "loss": 0.4082,
+      "step": 40170
+    },
+    {
+      "epoch": 10.019950124688279,
+      "grad_norm": 6.935123443603516,
+      "learning_rate": 9.984788029925188e-06,
+      "loss": 0.323,
+      "step": 40180
+    },
+    {
+      "epoch": 10.022443890274314,
+      "grad_norm": 5.4332051277160645,
+      "learning_rate": 9.982294264339153e-06,
+      "loss": 0.231,
+      "step": 40190
+    },
+    {
+      "epoch": 10.024937655860349,
+      "grad_norm": 8.775599479675293,
+      "learning_rate": 9.979800498753119e-06,
+      "loss": 0.3459,
+      "step": 40200
+    },
+    {
+      "epoch": 10.027431421446384,
+      "grad_norm": 4.293675422668457,
+      "learning_rate": 9.977306733167084e-06,
+      "loss": 0.3608,
+      "step": 40210
+    },
+    {
+      "epoch": 10.029925187032418,
+      "grad_norm": 7.175844669342041,
+      "learning_rate": 9.974812967581048e-06,
+      "loss": 0.3441,
+      "step": 40220
+    },
+    {
+      "epoch": 10.032418952618453,
+      "grad_norm": 6.6354522705078125,
+      "learning_rate": 9.972319201995013e-06,
+      "loss": 0.3302,
+      "step": 40230
+    },
+    {
+      "epoch": 10.034912718204488,
+      "grad_norm": 8.921165466308594,
+      "learning_rate": 9.969825436408978e-06,
+      "loss": 0.3587,
+      "step": 40240
+    },
+    {
+      "epoch": 10.037406483790523,
+      "grad_norm": 5.317387580871582,
+      "learning_rate": 9.967331670822943e-06,
+      "loss": 0.3295,
+      "step": 40250
+    },
+    {
+      "epoch": 10.039900249376558,
+      "grad_norm": 5.346241474151611,
+      "learning_rate": 9.964837905236909e-06,
+      "loss": 0.3452,
+      "step": 40260
+    },
+    {
+      "epoch": 10.042394014962593,
+      "grad_norm": 7.5005106925964355,
+      "learning_rate": 9.962344139650874e-06,
+      "loss": 0.3677,
+      "step": 40270
+    },
+    {
+      "epoch": 10.044887780548628,
+      "grad_norm": 6.900725841522217,
+      "learning_rate": 9.95985037406484e-06,
+      "loss": 0.3433,
+      "step": 40280
+    },
+    {
+      "epoch": 10.047381546134662,
+      "grad_norm": 6.401949405670166,
+      "learning_rate": 9.957356608478805e-06,
+      "loss": 0.3746,
+      "step": 40290
+    },
+    {
+      "epoch": 10.049875311720697,
+      "grad_norm": 8.203397750854492,
+      "learning_rate": 9.954862842892768e-06,
+      "loss": 0.2977,
+      "step": 40300
+    },
+    {
+      "epoch": 10.052369077306734,
+      "grad_norm": 4.425477027893066,
+      "learning_rate": 9.952369077306734e-06,
+      "loss": 0.3271,
+      "step": 40310
+    },
+    {
+      "epoch": 10.054862842892769,
+      "grad_norm": 9.94776439666748,
+      "learning_rate": 9.949875311720699e-06,
+      "loss": 0.3569,
+      "step": 40320
+    },
+    {
+      "epoch": 10.057356608478804,
+      "grad_norm": 6.8607892990112305,
+      "learning_rate": 9.947381546134664e-06,
+      "loss": 0.3441,
+      "step": 40330
+    },
+    {
+      "epoch": 10.059850374064839,
+      "grad_norm": 8.044713973999023,
+      "learning_rate": 9.94488778054863e-06,
+      "loss": 0.3992,
+      "step": 40340
+    },
+    {
+      "epoch": 10.062344139650873,
+      "grad_norm": 7.818645477294922,
+      "learning_rate": 9.942394014962595e-06,
+      "loss": 0.3718,
+      "step": 40350
+    },
+    {
+      "epoch": 10.064837905236908,
+      "grad_norm": 8.542047500610352,
+      "learning_rate": 9.93990024937656e-06,
+      "loss": 0.4638,
+      "step": 40360
+    },
+    {
+      "epoch": 10.067331670822943,
+      "grad_norm": 9.035058975219727,
+      "learning_rate": 9.937406483790526e-06,
+      "loss": 0.3159,
+      "step": 40370
+    },
+    {
+      "epoch": 10.069825436408978,
+      "grad_norm": 4.869987964630127,
+      "learning_rate": 9.934912718204489e-06,
+      "loss": 0.3242,
+      "step": 40380
+    },
+    {
+      "epoch": 10.072319201995013,
+      "grad_norm": 10.995007514953613,
+      "learning_rate": 9.932418952618454e-06,
+      "loss": 0.3318,
+      "step": 40390
+    },
+    {
+      "epoch": 10.074812967581048,
+      "grad_norm": 10.2716703414917,
+      "learning_rate": 9.92992518703242e-06,
+      "loss": 0.4521,
+      "step": 40400
+    },
+    {
+      "epoch": 10.077306733167083,
+      "grad_norm": 5.772310256958008,
+      "learning_rate": 9.927431421446385e-06,
+      "loss": 0.3316,
+      "step": 40410
+    },
+    {
+      "epoch": 10.079800498753118,
+      "grad_norm": 8.326214790344238,
+      "learning_rate": 9.924937655860349e-06,
+      "loss": 0.3413,
+      "step": 40420
+    },
+    {
+      "epoch": 10.082294264339152,
+      "grad_norm": 6.764336109161377,
+      "learning_rate": 9.922443890274314e-06,
+      "loss": 0.3069,
+      "step": 40430
+    },
+    {
+      "epoch": 10.084788029925187,
+      "grad_norm": 7.891993522644043,
+      "learning_rate": 9.91995012468828e-06,
+      "loss": 0.2665,
+      "step": 40440
+    },
+    {
+      "epoch": 10.087281795511222,
+      "grad_norm": 6.041927814483643,
+      "learning_rate": 9.917456359102246e-06,
+      "loss": 0.2921,
+      "step": 40450
+    },
+    {
+      "epoch": 10.089775561097257,
+      "grad_norm": 9.28152084350586,
+      "learning_rate": 9.914962593516212e-06,
+      "loss": 0.3558,
+      "step": 40460
+    },
+    {
+      "epoch": 10.092269326683292,
+      "grad_norm": 6.925419330596924,
+      "learning_rate": 9.912468827930175e-06,
+      "loss": 0.3164,
+      "step": 40470
+    },
+    {
+      "epoch": 10.094763092269327,
+      "grad_norm": 6.952348709106445,
+      "learning_rate": 9.90997506234414e-06,
+      "loss": 0.4215,
+      "step": 40480
+    },
+    {
+      "epoch": 10.097256857855362,
+      "grad_norm": 8.346542358398438,
+      "learning_rate": 9.907481296758106e-06,
+      "loss": 0.4134,
+      "step": 40490
+    },
+    {
+      "epoch": 10.099750623441397,
+      "grad_norm": 6.4831719398498535,
+      "learning_rate": 9.904987531172071e-06,
+      "loss": 0.3628,
+      "step": 40500
+    },
+    {
+      "epoch": 10.102244389027431,
+      "grad_norm": 6.676276206970215,
+      "learning_rate": 9.902493765586035e-06,
+      "loss": 0.3413,
+      "step": 40510
+    },
+    {
+      "epoch": 10.104738154613466,
+      "grad_norm": 9.890116691589355,
+      "learning_rate": 9.9e-06,
+      "loss": 0.3215,
+      "step": 40520
+    },
+    {
+      "epoch": 10.107231920199501,
+      "grad_norm": 5.857176780700684,
+      "learning_rate": 9.897506234413965e-06,
+      "loss": 0.3393,
+      "step": 40530
+    },
+    {
+      "epoch": 10.109725685785536,
+      "grad_norm": 6.540131092071533,
+      "learning_rate": 9.89501246882793e-06,
+      "loss": 0.3765,
+      "step": 40540
+    },
+    {
+      "epoch": 10.11221945137157,
+      "grad_norm": 9.703693389892578,
+      "learning_rate": 9.892518703241896e-06,
+      "loss": 0.3769,
+      "step": 40550
+    },
+    {
+      "epoch": 10.114713216957606,
+      "grad_norm": 9.611767768859863,
+      "learning_rate": 9.890024937655861e-06,
+      "loss": 0.3857,
+      "step": 40560
+    },
+    {
+      "epoch": 10.11720698254364,
+      "grad_norm": 6.961813926696777,
+      "learning_rate": 9.887531172069827e-06,
+      "loss": 0.3274,
+      "step": 40570
+    },
+    {
+      "epoch": 10.119700748129675,
+      "grad_norm": 7.749450206756592,
+      "learning_rate": 9.885037406483792e-06,
+      "loss": 0.3409,
+      "step": 40580
+    },
+    {
+      "epoch": 10.12219451371571,
+      "grad_norm": 10.376683235168457,
+      "learning_rate": 9.882543640897756e-06,
+      "loss": 0.36,
+      "step": 40590
+    },
+    {
+      "epoch": 10.124688279301745,
+      "grad_norm": 9.793608665466309,
+      "learning_rate": 9.880049875311721e-06,
+      "loss": 0.3467,
+      "step": 40600
+    },
+    {
+      "epoch": 10.12718204488778,
+      "grad_norm": 6.481242656707764,
+      "learning_rate": 9.877556109725686e-06,
+      "loss": 0.3184,
+      "step": 40610
+    },
+    {
+      "epoch": 10.129675810473815,
+      "grad_norm": 5.596631050109863,
+      "learning_rate": 9.875062344139651e-06,
+      "loss": 0.3533,
+      "step": 40620
+    },
+    {
+      "epoch": 10.13216957605985,
+      "grad_norm": 12.358193397521973,
+      "learning_rate": 9.872568578553617e-06,
+      "loss": 0.3539,
+      "step": 40630
+    },
+    {
+      "epoch": 10.134663341645885,
+      "grad_norm": 7.618841171264648,
+      "learning_rate": 9.870074812967582e-06,
+      "loss": 0.4053,
+      "step": 40640
+    },
+    {
+      "epoch": 10.13715710723192,
+      "grad_norm": 6.739760875701904,
+      "learning_rate": 9.867581047381547e-06,
+      "loss": 0.3235,
+      "step": 40650
+    },
+    {
+      "epoch": 10.139650872817954,
+      "grad_norm": 7.255684852600098,
+      "learning_rate": 9.865087281795513e-06,
+      "loss": 0.3134,
+      "step": 40660
+    },
+    {
+      "epoch": 10.14214463840399,
+      "grad_norm": 10.463494300842285,
+      "learning_rate": 9.862593516209476e-06,
+      "loss": 0.3657,
+      "step": 40670
+    },
+    {
+      "epoch": 10.144638403990024,
+      "grad_norm": 6.290800094604492,
+      "learning_rate": 9.860099750623442e-06,
+      "loss": 0.3472,
+      "step": 40680
+    },
+    {
+      "epoch": 10.147132169576059,
+      "grad_norm": 7.880048751831055,
+      "learning_rate": 9.857605985037407e-06,
+      "loss": 0.349,
+      "step": 40690
+    },
+    {
+      "epoch": 10.149625935162096,
+      "grad_norm": 8.012491226196289,
+      "learning_rate": 9.855112219451372e-06,
+      "loss": 0.3229,
+      "step": 40700
+    },
+    {
+      "epoch": 10.15211970074813,
+      "grad_norm": 9.101916313171387,
+      "learning_rate": 9.85286783042394e-06,
+      "loss": 0.3899,
+      "step": 40710
+    },
+    {
+      "epoch": 10.154613466334165,
+      "grad_norm": 10.565949440002441,
+      "learning_rate": 9.850374064837906e-06,
+      "loss": 0.342,
+      "step": 40720
+    },
+    {
+      "epoch": 10.1571072319202,
+      "grad_norm": 6.891229629516602,
+      "learning_rate": 9.847880299251871e-06,
+      "loss": 0.3361,
+      "step": 40730
+    },
+    {
+      "epoch": 10.159600997506235,
+      "grad_norm": 7.994174957275391,
+      "learning_rate": 9.845386533665836e-06,
+      "loss": 0.3542,
+      "step": 40740
+    },
+    {
+      "epoch": 10.16209476309227,
+      "grad_norm": 13.515098571777344,
+      "learning_rate": 9.842892768079802e-06,
+      "loss": 0.371,
+      "step": 40750
+    },
+    {
+      "epoch": 10.164588528678305,
+      "grad_norm": 6.52901554107666,
+      "learning_rate": 9.840399002493767e-06,
+      "loss": 0.3398,
+      "step": 40760
+    },
+    {
+      "epoch": 10.16708229426434,
+      "grad_norm": 6.790650844573975,
+      "learning_rate": 9.83790523690773e-06,
+      "loss": 0.3099,
+      "step": 40770
+    },
+    {
+      "epoch": 10.169576059850375,
+      "grad_norm": 4.857789516448975,
+      "learning_rate": 9.835411471321696e-06,
+      "loss": 0.3564,
+      "step": 40780
+    },
+    {
+      "epoch": 10.17206982543641,
+      "grad_norm": 10.275793075561523,
+      "learning_rate": 9.832917705735661e-06,
+      "loss": 0.4056,
+      "step": 40790
+    },
+    {
+      "epoch": 10.174563591022444,
+      "grad_norm": 5.9475812911987305,
+      "learning_rate": 9.830423940149627e-06,
+      "loss": 0.3462,
+      "step": 40800
+    },
+    {
+      "epoch": 10.17705735660848,
+      "grad_norm": 7.682375431060791,
+      "learning_rate": 9.827930174563592e-06,
+      "loss": 0.3232,
+      "step": 40810
+    },
+    {
+      "epoch": 10.179551122194514,
+      "grad_norm": 7.272604942321777,
+      "learning_rate": 9.825436408977557e-06,
+      "loss": 0.3526,
+      "step": 40820
+    },
+    {
+      "epoch": 10.182044887780549,
+      "grad_norm": 9.69630241394043,
+      "learning_rate": 9.822942643391523e-06,
+      "loss": 0.3904,
+      "step": 40830
+    },
+    {
+      "epoch": 10.184538653366584,
+      "grad_norm": 5.306740760803223,
+      "learning_rate": 9.820448877805488e-06,
+      "loss": 0.2868,
+      "step": 40840
+    },
+    {
+      "epoch": 10.187032418952619,
+      "grad_norm": 6.456069469451904,
+      "learning_rate": 9.817955112219451e-06,
+      "loss": 0.3426,
+      "step": 40850
+    },
+    {
+      "epoch": 10.189526184538654,
+      "grad_norm": 13.075380325317383,
+      "learning_rate": 9.815461346633417e-06,
+      "loss": 0.311,
+      "step": 40860
+    },
+    {
+      "epoch": 10.192019950124688,
+      "grad_norm": 5.357050895690918,
+      "learning_rate": 9.812967581047382e-06,
+      "loss": 0.4198,
+      "step": 40870
+    },
+    {
+      "epoch": 10.194513715710723,
+      "grad_norm": 7.080076694488525,
+      "learning_rate": 9.810473815461347e-06,
+      "loss": 0.3124,
+      "step": 40880
+    },
+    {
+      "epoch": 10.197007481296758,
+      "grad_norm": 4.402791500091553,
+      "learning_rate": 9.807980049875313e-06,
+      "loss": 0.3291,
+      "step": 40890
+    },
+    {
+      "epoch": 10.199501246882793,
+      "grad_norm": 6.491812705993652,
+      "learning_rate": 9.805486284289278e-06,
+      "loss": 0.3769,
+      "step": 40900
+    },
+    {
+      "epoch": 10.201995012468828,
+      "grad_norm": 7.707225322723389,
+      "learning_rate": 9.802992518703243e-06,
+      "loss": 0.395,
+      "step": 40910
+    },
+    {
+      "epoch": 10.204488778054863,
+      "grad_norm": 6.598991394042969,
+      "learning_rate": 9.800498753117209e-06,
+      "loss": 0.3356,
+      "step": 40920
+    },
+    {
+      "epoch": 10.206982543640898,
+      "grad_norm": 6.032766342163086,
+      "learning_rate": 9.798004987531174e-06,
+      "loss": 0.3243,
+      "step": 40930
+    },
+    {
+      "epoch": 10.209476309226932,
+      "grad_norm": 5.5279340744018555,
+      "learning_rate": 9.795511221945138e-06,
+      "loss": 0.3259,
+      "step": 40940
+    },
+    {
+      "epoch": 10.211970074812967,
+      "grad_norm": 5.94141149520874,
+      "learning_rate": 9.793017456359103e-06,
+      "loss": 0.3158,
+      "step": 40950
+    },
+    {
+      "epoch": 10.214463840399002,
+      "grad_norm": 8.342867851257324,
+      "learning_rate": 9.790523690773068e-06,
+      "loss": 0.3472,
+      "step": 40960
+    },
+    {
+      "epoch": 10.216957605985037,
+      "grad_norm": 6.11809778213501,
+      "learning_rate": 9.788029925187034e-06,
+      "loss": 0.3245,
+      "step": 40970
+    },
+    {
+      "epoch": 10.219451371571072,
+      "grad_norm": 3.928636074066162,
+      "learning_rate": 9.785536159600997e-06,
+      "loss": 0.2654,
+      "step": 40980
+    },
+    {
+      "epoch": 10.221945137157107,
+      "grad_norm": 5.511411190032959,
+      "learning_rate": 9.783042394014962e-06,
+      "loss": 0.378,
+      "step": 40990
+    },
+    {
+      "epoch": 10.224438902743142,
+      "grad_norm": 7.285804271697998,
+      "learning_rate": 9.78054862842893e-06,
+      "loss": 0.2728,
+      "step": 41000
+    },
+    {
+      "epoch": 10.226932668329177,
+      "grad_norm": 10.05148696899414,
+      "learning_rate": 9.778054862842895e-06,
+      "loss": 0.2886,
+      "step": 41010
+    },
+    {
+      "epoch": 10.229426433915211,
+      "grad_norm": 6.61144495010376,
+      "learning_rate": 9.775561097256858e-06,
+      "loss": 0.3763,
+      "step": 41020
+    },
+    {
+      "epoch": 10.231920199501246,
+      "grad_norm": 11.291173934936523,
+      "learning_rate": 9.773067331670824e-06,
+      "loss": 0.3521,
+      "step": 41030
+    },
+    {
+      "epoch": 10.234413965087281,
+      "grad_norm": 7.441661834716797,
+      "learning_rate": 9.770573566084789e-06,
+      "loss": 0.3958,
+      "step": 41040
+    },
+    {
+      "epoch": 10.236907730673316,
+      "grad_norm": 4.365013599395752,
+      "learning_rate": 9.768079800498754e-06,
+      "loss": 0.3623,
+      "step": 41050
+    },
+    {
+      "epoch": 10.239401496259351,
+      "grad_norm": 7.878891468048096,
+      "learning_rate": 9.765586034912718e-06,
+      "loss": 0.3634,
+      "step": 41060
+    },
+    {
+      "epoch": 10.241895261845386,
+      "grad_norm": 8.221386909484863,
+      "learning_rate": 9.763092269326683e-06,
+      "loss": 0.394,
+      "step": 41070
+    },
+    {
+      "epoch": 10.24438902743142,
+      "grad_norm": 6.871973991394043,
+      "learning_rate": 9.760598503740649e-06,
+      "loss": 0.3359,
+      "step": 41080
+    },
+    {
+      "epoch": 10.246882793017456,
+      "grad_norm": 7.841533184051514,
+      "learning_rate": 9.758104738154614e-06,
+      "loss": 0.316,
+      "step": 41090
+    },
+    {
+      "epoch": 10.24937655860349,
+      "grad_norm": 10.699889183044434,
+      "learning_rate": 9.755610972568579e-06,
+      "loss": 0.3793,
+      "step": 41100
+    },
+    {
+      "epoch": 10.251870324189527,
+      "grad_norm": 7.462698459625244,
+      "learning_rate": 9.753117206982544e-06,
+      "loss": 0.3795,
+      "step": 41110
+    },
+    {
+      "epoch": 10.254364089775562,
+      "grad_norm": 5.521877765655518,
+      "learning_rate": 9.75062344139651e-06,
+      "loss": 0.3306,
+      "step": 41120
+    },
+    {
+      "epoch": 10.256857855361597,
+      "grad_norm": 6.224538803100586,
+      "learning_rate": 9.748129675810475e-06,
+      "loss": 0.4206,
+      "step": 41130
+    },
+    {
+      "epoch": 10.259351620947632,
+      "grad_norm": 10.534749984741211,
+      "learning_rate": 9.74563591022444e-06,
+      "loss": 0.3515,
+      "step": 41140
+    },
+    {
+      "epoch": 10.261845386533667,
+      "grad_norm": 7.875677585601807,
+      "learning_rate": 9.743142144638404e-06,
+      "loss": 0.3535,
+      "step": 41150
+    },
+    {
+      "epoch": 10.264339152119701,
+      "grad_norm": 6.144274711608887,
+      "learning_rate": 9.74064837905237e-06,
+      "loss": 0.3605,
+      "step": 41160
+    },
+    {
+      "epoch": 10.266832917705736,
+      "grad_norm": 9.638225555419922,
+      "learning_rate": 9.738154613466335e-06,
+      "loss": 0.3261,
+      "step": 41170
+    },
+    {
+      "epoch": 10.269326683291771,
+      "grad_norm": 13.971467018127441,
+      "learning_rate": 9.7356608478803e-06,
+      "loss": 0.3333,
+      "step": 41180
+    },
+    {
+      "epoch": 10.271820448877806,
+      "grad_norm": 7.335572242736816,
+      "learning_rate": 9.733167082294265e-06,
+      "loss": 0.3326,
+      "step": 41190
+    },
+    {
+      "epoch": 10.27431421446384,
+      "grad_norm": 5.852632999420166,
+      "learning_rate": 9.73067331670823e-06,
+      "loss": 0.3612,
+      "step": 41200
+    },
+    {
+      "epoch": 10.276807980049876,
+      "grad_norm": 6.09345006942749,
+      "learning_rate": 9.728179551122196e-06,
+      "loss": 0.3995,
+      "step": 41210
+    },
+    {
+      "epoch": 10.27930174563591,
+      "grad_norm": 5.712091445922852,
+      "learning_rate": 9.725685785536161e-06,
+      "loss": 0.3078,
+      "step": 41220
+    },
+    {
+      "epoch": 10.281795511221945,
+      "grad_norm": 8.864590644836426,
+      "learning_rate": 9.723192019950125e-06,
+      "loss": 0.3629,
+      "step": 41230
+    },
+    {
+      "epoch": 10.28428927680798,
+      "grad_norm": 8.034436225891113,
+      "learning_rate": 9.72069825436409e-06,
+      "loss": 0.3913,
+      "step": 41240
+    },
+    {
+      "epoch": 10.286783042394015,
+      "grad_norm": 7.192776203155518,
+      "learning_rate": 9.718204488778055e-06,
+      "loss": 0.3908,
+      "step": 41250
+    },
+    {
+      "epoch": 10.28927680798005,
+      "grad_norm": 5.29744291305542,
+      "learning_rate": 9.71571072319202e-06,
+      "loss": 0.2776,
+      "step": 41260
+    },
+    {
+      "epoch": 10.291770573566085,
+      "grad_norm": 8.176728248596191,
+      "learning_rate": 9.713216957605986e-06,
+      "loss": 0.3514,
+      "step": 41270
+    },
+    {
+      "epoch": 10.29426433915212,
+      "grad_norm": 6.879700660705566,
+      "learning_rate": 9.710723192019951e-06,
+      "loss": 0.3185,
+      "step": 41280
+    },
+    {
+      "epoch": 10.296758104738155,
+      "grad_norm": 5.136245250701904,
+      "learning_rate": 9.708229426433917e-06,
+      "loss": 0.2969,
+      "step": 41290
+    },
+    {
+      "epoch": 10.29925187032419,
+      "grad_norm": 12.164958953857422,
+      "learning_rate": 9.705735660847882e-06,
+      "loss": 0.3606,
+      "step": 41300
+    },
+    {
+      "epoch": 10.301745635910224,
+      "grad_norm": 5.413372993469238,
+      "learning_rate": 9.703241895261846e-06,
+      "loss": 0.3706,
+      "step": 41310
+    },
+    {
+      "epoch": 10.30423940149626,
+      "grad_norm": 5.834495544433594,
+      "learning_rate": 9.700748129675811e-06,
+      "loss": 0.2906,
+      "step": 41320
+    },
+    {
+      "epoch": 10.306733167082294,
+      "grad_norm": 6.524380207061768,
+      "learning_rate": 9.698254364089776e-06,
+      "loss": 0.3777,
+      "step": 41330
+    },
+    {
+      "epoch": 10.309226932668329,
+      "grad_norm": 9.674921989440918,
+      "learning_rate": 9.695760598503741e-06,
+      "loss": 0.3734,
+      "step": 41340
+    },
+    {
+      "epoch": 10.311720698254364,
+      "grad_norm": 7.306479454040527,
+      "learning_rate": 9.693266832917707e-06,
+      "loss": 0.4354,
+      "step": 41350
+    },
+    {
+      "epoch": 10.314214463840399,
+      "grad_norm": 5.288915157318115,
+      "learning_rate": 9.690773067331672e-06,
+      "loss": 0.2979,
+      "step": 41360
+    },
+    {
+      "epoch": 10.316708229426434,
+      "grad_norm": 7.090877056121826,
+      "learning_rate": 9.688279301745637e-06,
+      "loss": 0.3617,
+      "step": 41370
+    },
+    {
+      "epoch": 10.319201995012468,
+      "grad_norm": 10.45274543762207,
+      "learning_rate": 9.685785536159603e-06,
+      "loss": 0.3738,
+      "step": 41380
+    },
+    {
+      "epoch": 10.321695760598503,
+      "grad_norm": 6.137414455413818,
+      "learning_rate": 9.683291770573568e-06,
+      "loss": 0.3855,
+      "step": 41390
+    },
+    {
+      "epoch": 10.324189526184538,
+      "grad_norm": 6.6708221435546875,
+      "learning_rate": 9.680798004987532e-06,
+      "loss": 0.3804,
+      "step": 41400
+    },
+    {
+      "epoch": 10.326683291770573,
+      "grad_norm": 7.918160438537598,
+      "learning_rate": 9.678304239401497e-06,
+      "loss": 0.3436,
+      "step": 41410
+    },
+    {
+      "epoch": 10.329177057356608,
+      "grad_norm": 5.83142614364624,
+      "learning_rate": 9.675810473815462e-06,
+      "loss": 0.3542,
+      "step": 41420
+    },
+    {
+      "epoch": 10.331670822942643,
+      "grad_norm": 7.670297145843506,
+      "learning_rate": 9.673316708229428e-06,
+      "loss": 0.3867,
+      "step": 41430
+    },
+    {
+      "epoch": 10.334164588528678,
+      "grad_norm": 5.303461074829102,
+      "learning_rate": 9.670822942643391e-06,
+      "loss": 0.3995,
+      "step": 41440
+    },
+    {
+      "epoch": 10.336658354114713,
+      "grad_norm": 6.452824115753174,
+      "learning_rate": 9.668329177057357e-06,
+      "loss": 0.3595,
+      "step": 41450
+    },
+    {
+      "epoch": 10.339152119700747,
+      "grad_norm": 14.923982620239258,
+      "learning_rate": 9.665835411471322e-06,
+      "loss": 0.3651,
+      "step": 41460
+    },
+    {
+      "epoch": 10.341645885286782,
+      "grad_norm": 5.03542947769165,
+      "learning_rate": 9.663341645885289e-06,
+      "loss": 0.3361,
+      "step": 41470
+    },
+    {
+      "epoch": 10.344139650872817,
+      "grad_norm": 7.917986869812012,
+      "learning_rate": 9.660847880299252e-06,
+      "loss": 0.3642,
+      "step": 41480
+    },
+    {
+      "epoch": 10.346633416458852,
+      "grad_norm": 8.379772186279297,
+      "learning_rate": 9.658354114713218e-06,
+      "loss": 0.3277,
+      "step": 41490
+    },
+    {
+      "epoch": 10.349127182044889,
+      "grad_norm": 7.017518997192383,
+      "learning_rate": 9.655860349127183e-06,
+      "loss": 0.3944,
+      "step": 41500
+    },
+    {
+      "epoch": 10.351620947630924,
+      "grad_norm": 5.158971309661865,
+      "learning_rate": 9.653366583541148e-06,
+      "loss": 0.3771,
+      "step": 41510
+    },
+    {
+      "epoch": 10.354114713216958,
+      "grad_norm": 5.925132751464844,
+      "learning_rate": 9.650872817955112e-06,
+      "loss": 0.3433,
+      "step": 41520
+    },
+    {
+      "epoch": 10.356608478802993,
+      "grad_norm": 8.094498634338379,
+      "learning_rate": 9.648379052369077e-06,
+      "loss": 0.3903,
+      "step": 41530
+    },
+    {
+      "epoch": 10.359102244389028,
+      "grad_norm": 5.886862277984619,
+      "learning_rate": 9.645885286783043e-06,
+      "loss": 0.3353,
+      "step": 41540
+    },
+    {
+      "epoch": 10.361596009975063,
+      "grad_norm": 7.800416469573975,
+      "learning_rate": 9.643391521197008e-06,
+      "loss": 0.3465,
+      "step": 41550
+    },
+    {
+      "epoch": 10.364089775561098,
+      "grad_norm": 5.397622108459473,
+      "learning_rate": 9.640897755610973e-06,
+      "loss": 0.4301,
+      "step": 41560
+    },
+    {
+      "epoch": 10.366583541147133,
+      "grad_norm": 8.980853080749512,
+      "learning_rate": 9.638403990024939e-06,
+      "loss": 0.3571,
+      "step": 41570
+    },
+    {
+      "epoch": 10.369077306733168,
+      "grad_norm": 9.865492820739746,
+      "learning_rate": 9.635910224438904e-06,
+      "loss": 0.322,
+      "step": 41580
+    },
+    {
+      "epoch": 10.371571072319203,
+      "grad_norm": 7.520500183105469,
+      "learning_rate": 9.63341645885287e-06,
+      "loss": 0.4147,
+      "step": 41590
+    },
+    {
+      "epoch": 10.374064837905237,
+      "grad_norm": 8.898337364196777,
+      "learning_rate": 9.630922693266833e-06,
+      "loss": 0.3443,
+      "step": 41600
+    },
+    {
+      "epoch": 10.376558603491272,
+      "grad_norm": 7.803848743438721,
+      "learning_rate": 9.628428927680798e-06,
+      "loss": 0.3786,
+      "step": 41610
+    },
+    {
+      "epoch": 10.379052369077307,
+      "grad_norm": 9.381965637207031,
+      "learning_rate": 9.625935162094763e-06,
+      "loss": 0.4683,
+      "step": 41620
+    },
+    {
+      "epoch": 10.381546134663342,
+      "grad_norm": 6.264208793640137,
+      "learning_rate": 9.623441396508729e-06,
+      "loss": 0.3297,
+      "step": 41630
+    },
+    {
+      "epoch": 10.384039900249377,
+      "grad_norm": 6.923156261444092,
+      "learning_rate": 9.620947630922694e-06,
+      "loss": 0.3056,
+      "step": 41640
+    },
+    {
+      "epoch": 10.386533665835412,
+      "grad_norm": 7.695561408996582,
+      "learning_rate": 9.61845386533666e-06,
+      "loss": 0.424,
+      "step": 41650
+    },
+    {
+      "epoch": 10.389027431421447,
+      "grad_norm": 5.747522354125977,
+      "learning_rate": 9.615960099750625e-06,
+      "loss": 0.3297,
+      "step": 41660
+    },
+    {
+      "epoch": 10.391521197007481,
+      "grad_norm": 5.526257038116455,
+      "learning_rate": 9.61346633416459e-06,
+      "loss": 0.3702,
+      "step": 41670
+    },
+    {
+      "epoch": 10.394014962593516,
+      "grad_norm": 7.8705244064331055,
+      "learning_rate": 9.610972568578555e-06,
+      "loss": 0.346,
+      "step": 41680
+    },
+    {
+      "epoch": 10.396508728179551,
+      "grad_norm": 7.783495903015137,
+      "learning_rate": 9.608478802992519e-06,
+      "loss": 0.3441,
+      "step": 41690
+    },
+    {
+      "epoch": 10.399002493765586,
+      "grad_norm": 8.003525733947754,
+      "learning_rate": 9.605985037406484e-06,
+      "loss": 0.3707,
+      "step": 41700
+    },
+    {
+      "epoch": 10.401496259351621,
+      "grad_norm": 7.88194465637207,
+      "learning_rate": 9.60349127182045e-06,
+      "loss": 0.3363,
+      "step": 41710
+    },
+    {
+      "epoch": 10.403990024937656,
+      "grad_norm": 7.859609127044678,
+      "learning_rate": 9.600997506234415e-06,
+      "loss": 0.3417,
+      "step": 41720
+    },
+    {
+      "epoch": 10.40648379052369,
+      "grad_norm": 6.967172145843506,
+      "learning_rate": 9.59850374064838e-06,
+      "loss": 0.3449,
+      "step": 41730
+    },
+    {
+      "epoch": 10.408977556109726,
+      "grad_norm": 6.901447772979736,
+      "learning_rate": 9.596009975062345e-06,
+      "loss": 0.4124,
+      "step": 41740
+    },
+    {
+      "epoch": 10.41147132169576,
+      "grad_norm": 7.131737232208252,
+      "learning_rate": 9.59351620947631e-06,
+      "loss": 0.3701,
+      "step": 41750
+    },
+    {
+      "epoch": 10.413965087281795,
+      "grad_norm": 8.898688316345215,
+      "learning_rate": 9.591022443890276e-06,
+      "loss": 0.323,
+      "step": 41760
+    },
+    {
+      "epoch": 10.41645885286783,
+      "grad_norm": 7.909594535827637,
+      "learning_rate": 9.58852867830424e-06,
+      "loss": 0.3552,
+      "step": 41770
+    },
+    {
+      "epoch": 10.418952618453865,
+      "grad_norm": 5.500792026519775,
+      "learning_rate": 9.586034912718205e-06,
+      "loss": 0.3423,
+      "step": 41780
+    },
+    {
+      "epoch": 10.4214463840399,
+      "grad_norm": 9.347502708435059,
+      "learning_rate": 9.58354114713217e-06,
+      "loss": 0.3671,
+      "step": 41790
+    },
+    {
+      "epoch": 10.423940149625935,
+      "grad_norm": 6.794683933258057,
+      "learning_rate": 9.581047381546136e-06,
+      "loss": 0.4234,
+      "step": 41800
+    },
+    {
+      "epoch": 10.42643391521197,
+      "grad_norm": 13.040794372558594,
+      "learning_rate": 9.5785536159601e-06,
+      "loss": 0.3615,
+      "step": 41810
+    },
+    {
+      "epoch": 10.428927680798004,
+      "grad_norm": 8.55950927734375,
+      "learning_rate": 9.576059850374066e-06,
+      "loss": 0.3848,
+      "step": 41820
+    },
+    {
+      "epoch": 10.43142144638404,
+      "grad_norm": 8.469449043273926,
+      "learning_rate": 9.573566084788032e-06,
+      "loss": 0.315,
+      "step": 41830
+    },
+    {
+      "epoch": 10.433915211970074,
+      "grad_norm": 6.907922744750977,
+      "learning_rate": 9.571072319201997e-06,
+      "loss": 0.4197,
+      "step": 41840
+    },
+    {
+      "epoch": 10.436408977556109,
+      "grad_norm": 5.317709445953369,
+      "learning_rate": 9.56857855361596e-06,
+      "loss": 0.3535,
+      "step": 41850
+    },
+    {
+      "epoch": 10.438902743142144,
+      "grad_norm": 6.60930061340332,
+      "learning_rate": 9.566084788029926e-06,
+      "loss": 0.3643,
+      "step": 41860
+    },
+    {
+      "epoch": 10.441396508728179,
+      "grad_norm": 9.468186378479004,
+      "learning_rate": 9.563591022443891e-06,
+      "loss": 0.3384,
+      "step": 41870
+    },
+    {
+      "epoch": 10.443890274314214,
+      "grad_norm": 5.716958522796631,
+      "learning_rate": 9.561097256857856e-06,
+      "loss": 0.3265,
+      "step": 41880
+    },
+    {
+      "epoch": 10.446384039900249,
+      "grad_norm": 7.9432172775268555,
+      "learning_rate": 9.558603491271822e-06,
+      "loss": 0.3969,
+      "step": 41890
+    },
+    {
+      "epoch": 10.448877805486283,
+      "grad_norm": 11.144307136535645,
+      "learning_rate": 9.556109725685785e-06,
+      "loss": 0.3584,
+      "step": 41900
+    },
+    {
+      "epoch": 10.451371571072318,
+      "grad_norm": 9.144835472106934,
+      "learning_rate": 9.55361596009975e-06,
+      "loss": 0.3963,
+      "step": 41910
+    },
+    {
+      "epoch": 10.453865336658355,
+      "grad_norm": 6.574597358703613,
+      "learning_rate": 9.551122194513716e-06,
+      "loss": 0.3258,
+      "step": 41920
+    },
+    {
+      "epoch": 10.45635910224439,
+      "grad_norm": 5.504093647003174,
+      "learning_rate": 9.548628428927681e-06,
+      "loss": 0.377,
+      "step": 41930
+    },
+    {
+      "epoch": 10.458852867830425,
+      "grad_norm": 5.041776180267334,
+      "learning_rate": 9.546134663341647e-06,
+      "loss": 0.3502,
+      "step": 41940
+    },
+    {
+      "epoch": 10.46134663341646,
+      "grad_norm": 8.293055534362793,
+      "learning_rate": 9.543640897755612e-06,
+      "loss": 0.3173,
+      "step": 41950
+    },
+    {
+      "epoch": 10.463840399002494,
+      "grad_norm": 6.408956527709961,
+      "learning_rate": 9.541147132169577e-06,
+      "loss": 0.3826,
+      "step": 41960
+    },
+    {
+      "epoch": 10.46633416458853,
+      "grad_norm": 9.970688819885254,
+      "learning_rate": 9.538653366583542e-06,
+      "loss": 0.3575,
+      "step": 41970
+    },
+    {
+      "epoch": 10.468827930174564,
+      "grad_norm": 6.766303062438965,
+      "learning_rate": 9.536159600997506e-06,
+      "loss": 0.3414,
+      "step": 41980
+    },
+    {
+      "epoch": 10.471321695760599,
+      "grad_norm": 5.876713275909424,
+      "learning_rate": 9.533665835411471e-06,
+      "loss": 0.3462,
+      "step": 41990
+    },
+    {
+      "epoch": 10.473815461346634,
+      "grad_norm": 8.713937759399414,
+      "learning_rate": 9.531172069825437e-06,
+      "loss": 0.3541,
+      "step": 42000
+    },
+    {
+      "epoch": 10.476309226932669,
+      "grad_norm": 5.201702117919922,
+      "learning_rate": 9.528678304239402e-06,
+      "loss": 0.3199,
+      "step": 42010
+    },
+    {
+      "epoch": 10.478802992518704,
+      "grad_norm": 12.989068031311035,
+      "learning_rate": 9.526184538653367e-06,
+      "loss": 0.3021,
+      "step": 42020
+    },
+    {
+      "epoch": 10.481296758104738,
+      "grad_norm": 7.327224254608154,
+      "learning_rate": 9.523690773067333e-06,
+      "loss": 0.3544,
+      "step": 42030
+    },
+    {
+      "epoch": 10.483790523690773,
+      "grad_norm": 10.310848236083984,
+      "learning_rate": 9.521197007481298e-06,
+      "loss": 0.3914,
+      "step": 42040
+    },
+    {
+      "epoch": 10.486284289276808,
+      "grad_norm": 6.748259544372559,
+      "learning_rate": 9.518703241895263e-06,
+      "loss": 0.3456,
+      "step": 42050
+    },
+    {
+      "epoch": 10.488778054862843,
+      "grad_norm": 12.408527374267578,
+      "learning_rate": 9.516209476309227e-06,
+      "loss": 0.3296,
+      "step": 42060
+    },
+    {
+      "epoch": 10.491271820448878,
+      "grad_norm": 8.86363697052002,
+      "learning_rate": 9.513715710723192e-06,
+      "loss": 0.3759,
+      "step": 42070
+    },
+    {
+      "epoch": 10.493765586034913,
+      "grad_norm": 9.177289962768555,
+      "learning_rate": 9.511221945137157e-06,
+      "loss": 0.375,
+      "step": 42080
+    },
+    {
+      "epoch": 10.496259351620948,
+      "grad_norm": 8.17337703704834,
+      "learning_rate": 9.508728179551123e-06,
+      "loss": 0.3493,
+      "step": 42090
+    },
+    {
+      "epoch": 10.498753117206983,
+      "grad_norm": 7.164012908935547,
+      "learning_rate": 9.506234413965088e-06,
+      "loss": 0.3652,
+      "step": 42100
+    },
+    {
+      "epoch": 10.501246882793017,
+      "grad_norm": 6.665464401245117,
+      "learning_rate": 9.503740648379053e-06,
+      "loss": 0.3838,
+      "step": 42110
+    },
+    {
+      "epoch": 10.503740648379052,
+      "grad_norm": 7.328868389129639,
+      "learning_rate": 9.501246882793019e-06,
+      "loss": 0.3464,
+      "step": 42120
+    },
+    {
+      "epoch": 10.506234413965087,
+      "grad_norm": 6.057427406311035,
+      "learning_rate": 9.498753117206984e-06,
+      "loss": 0.321,
+      "step": 42130
+    },
+    {
+      "epoch": 10.508728179551122,
+      "grad_norm": 5.329481601715088,
+      "learning_rate": 9.49625935162095e-06,
+      "loss": 0.2875,
+      "step": 42140
+    },
+    {
+      "epoch": 10.511221945137157,
+      "grad_norm": 7.551417827606201,
+      "learning_rate": 9.493765586034913e-06,
+      "loss": 0.3713,
+      "step": 42150
+    },
+    {
+      "epoch": 10.513715710723192,
+      "grad_norm": 5.047544956207275,
+      "learning_rate": 9.491271820448878e-06,
+      "loss": 0.4153,
+      "step": 42160
+    },
+    {
+      "epoch": 10.516209476309227,
+      "grad_norm": 5.810826301574707,
+      "learning_rate": 9.488778054862844e-06,
+      "loss": 0.3527,
+      "step": 42170
+    },
+    {
+      "epoch": 10.518703241895262,
+      "grad_norm": 5.0741753578186035,
+      "learning_rate": 9.486284289276809e-06,
+      "loss": 0.3569,
+      "step": 42180
+    },
+    {
+      "epoch": 10.521197007481296,
+      "grad_norm": 5.9377641677856445,
+      "learning_rate": 9.483790523690774e-06,
+      "loss": 0.3968,
+      "step": 42190
+    },
+    {
+      "epoch": 10.523690773067331,
+      "grad_norm": 8.227517127990723,
+      "learning_rate": 9.48129675810474e-06,
+      "loss": 0.3678,
+      "step": 42200
+    },
+    {
+      "epoch": 10.526184538653366,
+      "grad_norm": 7.43702507019043,
+      "learning_rate": 9.478802992518705e-06,
+      "loss": 0.3083,
+      "step": 42210
+    },
+    {
+      "epoch": 10.528678304239401,
+      "grad_norm": 6.5147199630737305,
+      "learning_rate": 9.47630922693267e-06,
+      "loss": 0.2658,
+      "step": 42220
+    },
+    {
+      "epoch": 10.531172069825436,
+      "grad_norm": 6.781185150146484,
+      "learning_rate": 9.473815461346634e-06,
+      "loss": 0.3596,
+      "step": 42230
+    },
+    {
+      "epoch": 10.53366583541147,
+      "grad_norm": 7.411608695983887,
+      "learning_rate": 9.471321695760599e-06,
+      "loss": 0.2931,
+      "step": 42240
+    },
+    {
+      "epoch": 10.536159600997506,
+      "grad_norm": 12.974010467529297,
+      "learning_rate": 9.468827930174564e-06,
+      "loss": 0.2816,
+      "step": 42250
+    },
+    {
+      "epoch": 10.53865336658354,
+      "grad_norm": 11.239740371704102,
+      "learning_rate": 9.46633416458853e-06,
+      "loss": 0.3805,
+      "step": 42260
+    },
+    {
+      "epoch": 10.541147132169575,
+      "grad_norm": 7.265082836151123,
+      "learning_rate": 9.463840399002493e-06,
+      "loss": 0.3101,
+      "step": 42270
+    },
+    {
+      "epoch": 10.54364089775561,
+      "grad_norm": 7.100124359130859,
+      "learning_rate": 9.461346633416459e-06,
+      "loss": 0.343,
+      "step": 42280
+    },
+    {
+      "epoch": 10.546134663341645,
+      "grad_norm": 5.751792907714844,
+      "learning_rate": 9.458852867830426e-06,
+      "loss": 0.3176,
+      "step": 42290
+    },
+    {
+      "epoch": 10.548628428927682,
+      "grad_norm": 6.716248989105225,
+      "learning_rate": 9.456359102244391e-06,
+      "loss": 0.3307,
+      "step": 42300
+    },
+    {
+      "epoch": 10.551122194513717,
+      "grad_norm": 9.329224586486816,
+      "learning_rate": 9.453865336658355e-06,
+      "loss": 0.3778,
+      "step": 42310
+    },
+    {
+      "epoch": 10.553615960099751,
+      "grad_norm": 8.02515697479248,
+      "learning_rate": 9.45137157107232e-06,
+      "loss": 0.4504,
+      "step": 42320
+    },
+    {
+      "epoch": 10.556109725685786,
+      "grad_norm": 8.72829818725586,
+      "learning_rate": 9.448877805486285e-06,
+      "loss": 0.4172,
+      "step": 42330
+    },
+    {
+      "epoch": 10.558603491271821,
+      "grad_norm": 6.4581780433654785,
+      "learning_rate": 9.44638403990025e-06,
+      "loss": 0.328,
+      "step": 42340
+    },
+    {
+      "epoch": 10.561097256857856,
+      "grad_norm": 10.020159721374512,
+      "learning_rate": 9.443890274314216e-06,
+      "loss": 0.4516,
+      "step": 42350
+    },
+    {
+      "epoch": 10.563591022443891,
+      "grad_norm": 8.494389533996582,
+      "learning_rate": 9.44139650872818e-06,
+      "loss": 0.3872,
+      "step": 42360
+    },
+    {
+      "epoch": 10.566084788029926,
+      "grad_norm": 7.352220058441162,
+      "learning_rate": 9.438902743142145e-06,
+      "loss": 0.2962,
+      "step": 42370
+    },
+    {
+      "epoch": 10.56857855361596,
+      "grad_norm": 7.202230930328369,
+      "learning_rate": 9.43640897755611e-06,
+      "loss": 0.3706,
+      "step": 42380
+    },
+    {
+      "epoch": 10.571072319201996,
+      "grad_norm": 6.778585433959961,
+      "learning_rate": 9.433915211970075e-06,
+      "loss": 0.3202,
+      "step": 42390
+    },
+    {
+      "epoch": 10.57356608478803,
+      "grad_norm": 10.143278121948242,
+      "learning_rate": 9.43142144638404e-06,
+      "loss": 0.3963,
+      "step": 42400
+    },
+    {
+      "epoch": 10.576059850374065,
+      "grad_norm": 9.119635581970215,
+      "learning_rate": 9.428927680798006e-06,
+      "loss": 0.3016,
+      "step": 42410
+    },
+    {
+      "epoch": 10.5785536159601,
+      "grad_norm": 6.856378555297852,
+      "learning_rate": 9.426433915211971e-06,
+      "loss": 0.3479,
+      "step": 42420
+    },
+    {
+      "epoch": 10.581047381546135,
+      "grad_norm": 5.599526882171631,
+      "learning_rate": 9.423940149625937e-06,
+      "loss": 0.3469,
+      "step": 42430
+    },
+    {
+      "epoch": 10.58354114713217,
+      "grad_norm": 6.967793941497803,
+      "learning_rate": 9.4214463840399e-06,
+      "loss": 0.3133,
+      "step": 42440
+    },
+    {
+      "epoch": 10.586034912718205,
+      "grad_norm": 5.5582404136657715,
+      "learning_rate": 9.418952618453865e-06,
+      "loss": 0.3838,
+      "step": 42450
+    },
+    {
+      "epoch": 10.58852867830424,
+      "grad_norm": 11.962347030639648,
+      "learning_rate": 9.41645885286783e-06,
+      "loss": 0.3827,
+      "step": 42460
+    },
+    {
+      "epoch": 10.591022443890274,
+      "grad_norm": 5.947192668914795,
+      "learning_rate": 9.413965087281796e-06,
+      "loss": 0.3638,
+      "step": 42470
+    },
+    {
+      "epoch": 10.59351620947631,
+      "grad_norm": 5.291638374328613,
+      "learning_rate": 9.411471321695761e-06,
+      "loss": 0.3622,
+      "step": 42480
+    },
+    {
+      "epoch": 10.596009975062344,
+      "grad_norm": 6.401669979095459,
+      "learning_rate": 9.408977556109727e-06,
+      "loss": 0.3045,
+      "step": 42490
+    },
+    {
+      "epoch": 10.598503740648379,
+      "grad_norm": 8.00327205657959,
+      "learning_rate": 9.406483790523692e-06,
+      "loss": 0.3862,
+      "step": 42500
+    },
+    {
+      "epoch": 10.600997506234414,
+      "grad_norm": 8.175292015075684,
+      "learning_rate": 9.403990024937657e-06,
+      "loss": 0.3123,
+      "step": 42510
+    },
+    {
+      "epoch": 10.603491271820449,
+      "grad_norm": 13.475317001342773,
+      "learning_rate": 9.401496259351621e-06,
+      "loss": 0.3399,
+      "step": 42520
+    },
+    {
+      "epoch": 10.605985037406484,
+      "grad_norm": 8.634927749633789,
+      "learning_rate": 9.399002493765586e-06,
+      "loss": 0.369,
+      "step": 42530
+    },
+    {
+      "epoch": 10.608478802992519,
+      "grad_norm": 8.839632034301758,
+      "learning_rate": 9.396508728179552e-06,
+      "loss": 0.3389,
+      "step": 42540
+    },
+    {
+      "epoch": 10.610972568578553,
+      "grad_norm": 4.931496620178223,
+      "learning_rate": 9.394014962593517e-06,
+      "loss": 0.3422,
+      "step": 42550
+    },
+    {
+      "epoch": 10.613466334164588,
+      "grad_norm": 5.307199001312256,
+      "learning_rate": 9.391521197007482e-06,
+      "loss": 0.2964,
+      "step": 42560
+    },
+    {
+      "epoch": 10.615960099750623,
+      "grad_norm": 11.918758392333984,
+      "learning_rate": 9.389027431421447e-06,
+      "loss": 0.3132,
+      "step": 42570
+    },
+    {
+      "epoch": 10.618453865336658,
+      "grad_norm": 10.152669906616211,
+      "learning_rate": 9.386533665835413e-06,
+      "loss": 0.3228,
+      "step": 42580
+    },
+    {
+      "epoch": 10.620947630922693,
+      "grad_norm": 8.567078590393066,
+      "learning_rate": 9.384039900249378e-06,
+      "loss": 0.4226,
+      "step": 42590
+    },
+    {
+      "epoch": 10.623441396508728,
+      "grad_norm": 9.373230934143066,
+      "learning_rate": 9.381546134663343e-06,
+      "loss": 0.3664,
+      "step": 42600
+    },
+    {
+      "epoch": 10.625935162094763,
+      "grad_norm": 7.585546016693115,
+      "learning_rate": 9.379052369077307e-06,
+      "loss": 0.3636,
+      "step": 42610
+    },
+    {
+      "epoch": 10.628428927680797,
+      "grad_norm": 6.881632328033447,
+      "learning_rate": 9.376558603491272e-06,
+      "loss": 0.304,
+      "step": 42620
+    },
+    {
+      "epoch": 10.630922693266832,
+      "grad_norm": 5.626585483551025,
+      "learning_rate": 9.374064837905238e-06,
+      "loss": 0.3563,
+      "step": 42630
+    },
+    {
+      "epoch": 10.633416458852867,
+      "grad_norm": 14.296299934387207,
+      "learning_rate": 9.371571072319203e-06,
+      "loss": 0.3678,
+      "step": 42640
+    },
+    {
+      "epoch": 10.635910224438902,
+      "grad_norm": 9.416000366210938,
+      "learning_rate": 9.369077306733168e-06,
+      "loss": 0.3737,
+      "step": 42650
+    },
+    {
+      "epoch": 10.638403990024937,
+      "grad_norm": 7.975756645202637,
+      "learning_rate": 9.366583541147134e-06,
+      "loss": 0.3162,
+      "step": 42660
+    },
+    {
+      "epoch": 10.640897755610972,
+      "grad_norm": 10.369644165039062,
+      "learning_rate": 9.364089775561099e-06,
+      "loss": 0.3617,
+      "step": 42670
+    },
+    {
+      "epoch": 10.643391521197007,
+      "grad_norm": 6.275656700134277,
+      "learning_rate": 9.361596009975064e-06,
+      "loss": 0.3393,
+      "step": 42680
+    },
+    {
+      "epoch": 10.645885286783042,
+      "grad_norm": 8.759428977966309,
+      "learning_rate": 9.359102244389028e-06,
+      "loss": 0.3428,
+      "step": 42690
+    },
+    {
+      "epoch": 10.648379052369076,
+      "grad_norm": 5.912599563598633,
+      "learning_rate": 9.356608478802993e-06,
+      "loss": 0.3031,
+      "step": 42700
+    },
+    {
+      "epoch": 10.650872817955111,
+      "grad_norm": 6.889318943023682,
+      "learning_rate": 9.354114713216958e-06,
+      "loss": 0.3897,
+      "step": 42710
+    },
+    {
+      "epoch": 10.653366583541148,
+      "grad_norm": 5.2232346534729,
+      "learning_rate": 9.351620947630924e-06,
+      "loss": 0.3764,
+      "step": 42720
+    },
+    {
+      "epoch": 10.655860349127183,
+      "grad_norm": 9.666282653808594,
+      "learning_rate": 9.349127182044887e-06,
+      "loss": 0.355,
+      "step": 42730
+    },
+    {
+      "epoch": 10.658354114713218,
+      "grad_norm": 6.856553077697754,
+      "learning_rate": 9.346633416458853e-06,
+      "loss": 0.3644,
+      "step": 42740
+    },
+    {
+      "epoch": 10.660847880299253,
+      "grad_norm": 7.544265270233154,
+      "learning_rate": 9.344139650872818e-06,
+      "loss": 0.3547,
+      "step": 42750
+    },
+    {
+      "epoch": 10.663341645885287,
+      "grad_norm": 9.202102661132812,
+      "learning_rate": 9.341645885286785e-06,
+      "loss": 0.3561,
+      "step": 42760
+    },
+    {
+      "epoch": 10.665835411471322,
+      "grad_norm": 5.572080135345459,
+      "learning_rate": 9.339152119700749e-06,
+      "loss": 0.379,
+      "step": 42770
+    },
+    {
+      "epoch": 10.668329177057357,
+      "grad_norm": 7.436700820922852,
+      "learning_rate": 9.336658354114714e-06,
+      "loss": 0.341,
+      "step": 42780
+    },
+    {
+      "epoch": 10.670822942643392,
+      "grad_norm": 7.51735782623291,
+      "learning_rate": 9.33416458852868e-06,
+      "loss": 0.3007,
+      "step": 42790
+    },
+    {
+      "epoch": 10.673316708229427,
+      "grad_norm": 7.93154239654541,
+      "learning_rate": 9.331670822942645e-06,
+      "loss": 0.3787,
+      "step": 42800
+    },
+    {
+      "epoch": 10.675810473815462,
+      "grad_norm": 7.7301106452941895,
+      "learning_rate": 9.329177057356608e-06,
+      "loss": 0.3315,
+      "step": 42810
+    },
+    {
+      "epoch": 10.678304239401497,
+      "grad_norm": 4.633336544036865,
+      "learning_rate": 9.326683291770573e-06,
+      "loss": 0.4015,
+      "step": 42820
+    },
+    {
+      "epoch": 10.680798004987532,
+      "grad_norm": 8.565386772155762,
+      "learning_rate": 9.324189526184539e-06,
+      "loss": 0.3873,
+      "step": 42830
+    },
+    {
+      "epoch": 10.683291770573566,
+      "grad_norm": 5.983166694641113,
+      "learning_rate": 9.321695760598504e-06,
+      "loss": 0.3298,
+      "step": 42840
+    },
+    {
+      "epoch": 10.685785536159601,
+      "grad_norm": 8.084731101989746,
+      "learning_rate": 9.31920199501247e-06,
+      "loss": 0.2891,
+      "step": 42850
+    },
+    {
+      "epoch": 10.688279301745636,
+      "grad_norm": 9.424635887145996,
+      "learning_rate": 9.316708229426435e-06,
+      "loss": 0.3609,
+      "step": 42860
+    },
+    {
+      "epoch": 10.690773067331671,
+      "grad_norm": 4.796233654022217,
+      "learning_rate": 9.3142144638404e-06,
+      "loss": 0.3446,
+      "step": 42870
+    },
+    {
+      "epoch": 10.693266832917706,
+      "grad_norm": 8.594659805297852,
+      "learning_rate": 9.311720698254365e-06,
+      "loss": 0.3575,
+      "step": 42880
+    },
+    {
+      "epoch": 10.69576059850374,
+      "grad_norm": 5.583627223968506,
+      "learning_rate": 9.30922693266833e-06,
+      "loss": 0.3207,
+      "step": 42890
+    },
+    {
+      "epoch": 10.698254364089776,
+      "grad_norm": 7.472324848175049,
+      "learning_rate": 9.306733167082294e-06,
+      "loss": 0.3412,
+      "step": 42900
+    },
+    {
+      "epoch": 10.70074812967581,
+      "grad_norm": 5.4451584815979,
+      "learning_rate": 9.30423940149626e-06,
+      "loss": 0.3852,
+      "step": 42910
+    },
+    {
+      "epoch": 10.703241895261845,
+      "grad_norm": 8.073676109313965,
+      "learning_rate": 9.301745635910225e-06,
+      "loss": 0.3938,
+      "step": 42920
+    },
+    {
+      "epoch": 10.70573566084788,
+      "grad_norm": 5.401284694671631,
+      "learning_rate": 9.29925187032419e-06,
+      "loss": 0.2982,
+      "step": 42930
+    },
+    {
+      "epoch": 10.708229426433915,
+      "grad_norm": 10.709132194519043,
+      "learning_rate": 9.296758104738155e-06,
+      "loss": 0.3338,
+      "step": 42940
+    },
+    {
+      "epoch": 10.71072319201995,
+      "grad_norm": 6.343472957611084,
+      "learning_rate": 9.29426433915212e-06,
+      "loss": 0.3251,
+      "step": 42950
+    },
+    {
+      "epoch": 10.713216957605985,
+      "grad_norm": 7.1593499183654785,
+      "learning_rate": 9.291770573566086e-06,
+      "loss": 0.332,
+      "step": 42960
+    },
+    {
+      "epoch": 10.71571072319202,
+      "grad_norm": 7.830443382263184,
+      "learning_rate": 9.289276807980051e-06,
+      "loss": 0.4038,
+      "step": 42970
+    },
+    {
+      "epoch": 10.718204488778055,
+      "grad_norm": 8.076725959777832,
+      "learning_rate": 9.286783042394015e-06,
+      "loss": 0.3343,
+      "step": 42980
+    },
+    {
+      "epoch": 10.72069825436409,
+      "grad_norm": 7.131729602813721,
+      "learning_rate": 9.28428927680798e-06,
+      "loss": 0.316,
+      "step": 42990
+    },
+    {
+      "epoch": 10.723192019950124,
+      "grad_norm": 6.504873752593994,
+      "learning_rate": 9.281795511221946e-06,
+      "loss": 0.3059,
+      "step": 43000
+    },
+    {
+      "epoch": 10.72568578553616,
+      "grad_norm": 6.572418212890625,
+      "learning_rate": 9.279301745635911e-06,
+      "loss": 0.3717,
+      "step": 43010
+    },
+    {
+      "epoch": 10.728179551122194,
+      "grad_norm": 7.904941558837891,
+      "learning_rate": 9.276807980049876e-06,
+      "loss": 0.405,
+      "step": 43020
+    },
+    {
+      "epoch": 10.730673316708229,
+      "grad_norm": 7.299516201019287,
+      "learning_rate": 9.274314214463842e-06,
+      "loss": 0.3487,
+      "step": 43030
+    },
+    {
+      "epoch": 10.733167082294264,
+      "grad_norm": 10.756522178649902,
+      "learning_rate": 9.271820448877807e-06,
+      "loss": 0.3513,
+      "step": 43040
+    },
+    {
+      "epoch": 10.735660847880299,
+      "grad_norm": 5.3038010597229,
+      "learning_rate": 9.269326683291772e-06,
+      "loss": 0.2722,
+      "step": 43050
+    },
+    {
+      "epoch": 10.738154613466333,
+      "grad_norm": 9.962454795837402,
+      "learning_rate": 9.266832917705736e-06,
+      "loss": 0.3812,
+      "step": 43060
+    },
+    {
+      "epoch": 10.740648379052368,
+      "grad_norm": 6.948879718780518,
+      "learning_rate": 9.264339152119701e-06,
+      "loss": 0.3602,
+      "step": 43070
+    },
+    {
+      "epoch": 10.743142144638403,
+      "grad_norm": 6.786881923675537,
+      "learning_rate": 9.261845386533666e-06,
+      "loss": 0.4163,
+      "step": 43080
+    },
+    {
+      "epoch": 10.745635910224438,
+      "grad_norm": 9.627817153930664,
+      "learning_rate": 9.259351620947632e-06,
+      "loss": 0.3901,
+      "step": 43090
+    },
+    {
+      "epoch": 10.748129675810475,
+      "grad_norm": 6.600256443023682,
+      "learning_rate": 9.256857855361597e-06,
+      "loss": 0.324,
+      "step": 43100
+    },
+    {
+      "epoch": 10.75062344139651,
+      "grad_norm": 6.746035575866699,
+      "learning_rate": 9.254364089775562e-06,
+      "loss": 0.3914,
+      "step": 43110
+    },
+    {
+      "epoch": 10.753117206982544,
+      "grad_norm": 8.608478546142578,
+      "learning_rate": 9.251870324189528e-06,
+      "loss": 0.3469,
+      "step": 43120
+    },
+    {
+      "epoch": 10.75561097256858,
+      "grad_norm": 7.662023544311523,
+      "learning_rate": 9.249376558603493e-06,
+      "loss": 0.3648,
+      "step": 43130
+    },
+    {
+      "epoch": 10.758104738154614,
+      "grad_norm": 7.992726802825928,
+      "learning_rate": 9.246882793017458e-06,
+      "loss": 0.3541,
+      "step": 43140
+    },
+    {
+      "epoch": 10.760598503740649,
+      "grad_norm": 6.671484470367432,
+      "learning_rate": 9.244389027431422e-06,
+      "loss": 0.3287,
+      "step": 43150
+    },
+    {
+      "epoch": 10.763092269326684,
+      "grad_norm": 8.085719108581543,
+      "learning_rate": 9.241895261845387e-06,
+      "loss": 0.3401,
+      "step": 43160
+    },
+    {
+      "epoch": 10.765586034912719,
+      "grad_norm": 5.610658168792725,
+      "learning_rate": 9.239401496259353e-06,
+      "loss": 0.3296,
+      "step": 43170
+    },
+    {
+      "epoch": 10.768079800498754,
+      "grad_norm": 8.575346946716309,
+      "learning_rate": 9.236907730673318e-06,
+      "loss": 0.3825,
+      "step": 43180
+    },
+    {
+      "epoch": 10.770573566084789,
+      "grad_norm": 9.167414665222168,
+      "learning_rate": 9.234413965087281e-06,
+      "loss": 0.3747,
+      "step": 43190
+    },
+    {
+      "epoch": 10.773067331670823,
+      "grad_norm": 9.802268028259277,
+      "learning_rate": 9.231920199501247e-06,
+      "loss": 0.323,
+      "step": 43200
+    },
+    {
+      "epoch": 10.775561097256858,
+      "grad_norm": 5.6160101890563965,
+      "learning_rate": 9.229426433915212e-06,
+      "loss": 0.34,
+      "step": 43210
+    },
+    {
+      "epoch": 10.778054862842893,
+      "grad_norm": 6.598016738891602,
+      "learning_rate": 9.226932668329179e-06,
+      "loss": 0.3541,
+      "step": 43220
+    },
+    {
+      "epoch": 10.780548628428928,
+      "grad_norm": 7.513404369354248,
+      "learning_rate": 9.224438902743143e-06,
+      "loss": 0.3587,
+      "step": 43230
+    },
+    {
+      "epoch": 10.783042394014963,
+      "grad_norm": 8.931615829467773,
+      "learning_rate": 9.221945137157108e-06,
+      "loss": 0.3301,
+      "step": 43240
+    },
+    {
+      "epoch": 10.785536159600998,
+      "grad_norm": 8.84310245513916,
+      "learning_rate": 9.219451371571073e-06,
+      "loss": 0.3577,
+      "step": 43250
+    },
+    {
+      "epoch": 10.788029925187033,
+      "grad_norm": 6.9853668212890625,
+      "learning_rate": 9.216957605985039e-06,
+      "loss": 0.3472,
+      "step": 43260
+    },
+    {
+      "epoch": 10.790523690773068,
+      "grad_norm": 8.114693641662598,
+      "learning_rate": 9.214463840399002e-06,
+      "loss": 0.3509,
+      "step": 43270
+    },
+    {
+      "epoch": 10.793017456359102,
+      "grad_norm": 6.722007751464844,
+      "learning_rate": 9.211970074812968e-06,
+      "loss": 0.3194,
+      "step": 43280
+    },
+    {
+      "epoch": 10.795511221945137,
+      "grad_norm": 11.021949768066406,
+      "learning_rate": 9.209476309226933e-06,
+      "loss": 0.3875,
+      "step": 43290
+    },
+    {
+      "epoch": 10.798004987531172,
+      "grad_norm": 9.12653636932373,
+      "learning_rate": 9.206982543640898e-06,
+      "loss": 0.3549,
+      "step": 43300
+    },
+    {
+      "epoch": 10.800498753117207,
+      "grad_norm": 6.999925136566162,
+      "learning_rate": 9.204488778054863e-06,
+      "loss": 0.3565,
+      "step": 43310
+    },
+    {
+      "epoch": 10.802992518703242,
+      "grad_norm": 8.389660835266113,
+      "learning_rate": 9.201995012468829e-06,
+      "loss": 0.2577,
+      "step": 43320
+    },
+    {
+      "epoch": 10.805486284289277,
+      "grad_norm": 6.072205543518066,
+      "learning_rate": 9.199501246882794e-06,
+      "loss": 0.3147,
+      "step": 43330
+    },
+    {
+      "epoch": 10.807980049875312,
+      "grad_norm": 7.625298500061035,
+      "learning_rate": 9.19700748129676e-06,
+      "loss": 0.2968,
+      "step": 43340
+    },
+    {
+      "epoch": 10.810473815461346,
+      "grad_norm": 9.655031204223633,
+      "learning_rate": 9.194513715710725e-06,
+      "loss": 0.3788,
+      "step": 43350
+    },
+    {
+      "epoch": 10.812967581047381,
+      "grad_norm": 5.24294900894165,
+      "learning_rate": 9.192019950124688e-06,
+      "loss": 0.3467,
+      "step": 43360
+    },
+    {
+      "epoch": 10.815461346633416,
+      "grad_norm": 3.961106777191162,
+      "learning_rate": 9.189526184538654e-06,
+      "loss": 0.2739,
+      "step": 43370
+    },
+    {
+      "epoch": 10.817955112219451,
+      "grad_norm": 9.00661563873291,
+      "learning_rate": 9.187032418952619e-06,
+      "loss": 0.3829,
+      "step": 43380
+    },
+    {
+      "epoch": 10.820448877805486,
+      "grad_norm": 5.253776550292969,
+      "learning_rate": 9.184538653366584e-06,
+      "loss": 0.3532,
+      "step": 43390
+    },
+    {
+      "epoch": 10.82294264339152,
+      "grad_norm": 6.8196587562561035,
+      "learning_rate": 9.18204488778055e-06,
+      "loss": 0.3576,
+      "step": 43400
+    },
+    {
+      "epoch": 10.825436408977556,
+      "grad_norm": 6.497467517852783,
+      "learning_rate": 9.179551122194515e-06,
+      "loss": 0.3487,
+      "step": 43410
+    },
+    {
+      "epoch": 10.82793017456359,
+      "grad_norm": 7.977984428405762,
+      "learning_rate": 9.17705735660848e-06,
+      "loss": 0.3008,
+      "step": 43420
+    },
+    {
+      "epoch": 10.830423940149625,
+      "grad_norm": 5.911102771759033,
+      "learning_rate": 9.174563591022445e-06,
+      "loss": 0.4166,
+      "step": 43430
+    },
+    {
+      "epoch": 10.83291770573566,
+      "grad_norm": 5.653079032897949,
+      "learning_rate": 9.172069825436409e-06,
+      "loss": 0.3284,
+      "step": 43440
+    },
+    {
+      "epoch": 10.835411471321695,
+      "grad_norm": 5.5909199714660645,
+      "learning_rate": 9.169576059850374e-06,
+      "loss": 0.3524,
+      "step": 43450
+    },
+    {
+      "epoch": 10.83790523690773,
+      "grad_norm": 5.932044982910156,
+      "learning_rate": 9.16708229426434e-06,
+      "loss": 0.371,
+      "step": 43460
+    },
+    {
+      "epoch": 10.840399002493765,
+      "grad_norm": 8.470308303833008,
+      "learning_rate": 9.164588528678305e-06,
+      "loss": 0.3592,
+      "step": 43470
+    },
+    {
+      "epoch": 10.8428927680798,
+      "grad_norm": 5.490266799926758,
+      "learning_rate": 9.16209476309227e-06,
+      "loss": 0.3547,
+      "step": 43480
+    },
+    {
+      "epoch": 10.845386533665835,
+      "grad_norm": 7.31662654876709,
+      "learning_rate": 9.159600997506236e-06,
+      "loss": 0.3529,
+      "step": 43490
+    },
+    {
+      "epoch": 10.84788029925187,
+      "grad_norm": 9.396953582763672,
+      "learning_rate": 9.157107231920201e-06,
+      "loss": 0.3525,
+      "step": 43500
+    },
+    {
+      "epoch": 10.850374064837904,
+      "grad_norm": 8.587018013000488,
+      "learning_rate": 9.154613466334166e-06,
+      "loss": 0.3494,
+      "step": 43510
+    },
+    {
+      "epoch": 10.85286783042394,
+      "grad_norm": 5.789772033691406,
+      "learning_rate": 9.15211970074813e-06,
+      "loss": 0.3412,
+      "step": 43520
+    },
+    {
+      "epoch": 10.855361596009976,
+      "grad_norm": 8.550657272338867,
+      "learning_rate": 9.149625935162095e-06,
+      "loss": 0.3953,
+      "step": 43530
+    },
+    {
+      "epoch": 10.85785536159601,
+      "grad_norm": 7.927114963531494,
+      "learning_rate": 9.14713216957606e-06,
+      "loss": 0.3944,
+      "step": 43540
+    },
+    {
+      "epoch": 10.860349127182046,
+      "grad_norm": 5.831830978393555,
+      "learning_rate": 9.144638403990026e-06,
+      "loss": 0.3161,
+      "step": 43550
+    },
+    {
+      "epoch": 10.86284289276808,
+      "grad_norm": 7.033876419067383,
+      "learning_rate": 9.14214463840399e-06,
+      "loss": 0.3668,
+      "step": 43560
+    },
+    {
+      "epoch": 10.865336658354115,
+      "grad_norm": 5.114386081695557,
+      "learning_rate": 9.139650872817956e-06,
+      "loss": 0.3579,
+      "step": 43570
+    },
+    {
+      "epoch": 10.86783042394015,
+      "grad_norm": 5.873979568481445,
+      "learning_rate": 9.137157107231922e-06,
+      "loss": 0.3652,
+      "step": 43580
+    },
+    {
+      "epoch": 10.870324189526185,
+      "grad_norm": 7.800948619842529,
+      "learning_rate": 9.134663341645887e-06,
+      "loss": 0.3274,
+      "step": 43590
+    },
+    {
+      "epoch": 10.87281795511222,
+      "grad_norm": 7.095271587371826,
+      "learning_rate": 9.132169576059852e-06,
+      "loss": 0.3438,
+      "step": 43600
+    },
+    {
+      "epoch": 10.875311720698255,
+      "grad_norm": 19.4151668548584,
+      "learning_rate": 9.129675810473816e-06,
+      "loss": 0.3501,
+      "step": 43610
+    },
+    {
+      "epoch": 10.87780548628429,
+      "grad_norm": 7.045140743255615,
+      "learning_rate": 9.127182044887781e-06,
+      "loss": 0.3194,
+      "step": 43620
+    },
+    {
+      "epoch": 10.880299251870325,
+      "grad_norm": 7.247674465179443,
+      "learning_rate": 9.124688279301747e-06,
+      "loss": 0.3335,
+      "step": 43630
+    },
+    {
+      "epoch": 10.88279301745636,
+      "grad_norm": 7.939310550689697,
+      "learning_rate": 9.122194513715712e-06,
+      "loss": 0.3741,
+      "step": 43640
+    },
+    {
+      "epoch": 10.885286783042394,
+      "grad_norm": 7.598067283630371,
+      "learning_rate": 9.119700748129676e-06,
+      "loss": 0.3175,
+      "step": 43650
+    },
+    {
+      "epoch": 10.88778054862843,
+      "grad_norm": 6.730564594268799,
+      "learning_rate": 9.11720698254364e-06,
+      "loss": 0.3614,
+      "step": 43660
+    },
+    {
+      "epoch": 10.890274314214464,
+      "grad_norm": 8.91694450378418,
+      "learning_rate": 9.114713216957606e-06,
+      "loss": 0.4158,
+      "step": 43670
+    },
+    {
+      "epoch": 10.892768079800499,
+      "grad_norm": 11.187231063842773,
+      "learning_rate": 9.112219451371571e-06,
+      "loss": 0.3443,
+      "step": 43680
+    },
+    {
+      "epoch": 10.895261845386534,
+      "grad_norm": 4.5806450843811035,
+      "learning_rate": 9.109725685785537e-06,
+      "loss": 0.3801,
+      "step": 43690
+    },
+    {
+      "epoch": 10.897755610972569,
+      "grad_norm": 8.682236671447754,
+      "learning_rate": 9.107231920199502e-06,
+      "loss": 0.3462,
+      "step": 43700
+    },
+    {
+      "epoch": 10.900249376558603,
+      "grad_norm": 6.818325042724609,
+      "learning_rate": 9.104738154613467e-06,
+      "loss": 0.3968,
+      "step": 43710
+    },
+    {
+      "epoch": 10.902743142144638,
+      "grad_norm": 6.7168426513671875,
+      "learning_rate": 9.102244389027433e-06,
+      "loss": 0.5048,
+      "step": 43720
+    },
+    {
+      "epoch": 10.905236907730673,
+      "grad_norm": 7.661280155181885,
+      "learning_rate": 9.099750623441396e-06,
+      "loss": 0.3603,
+      "step": 43730
+    },
+    {
+      "epoch": 10.907730673316708,
+      "grad_norm": 11.238551139831543,
+      "learning_rate": 9.097256857855362e-06,
+      "loss": 0.3616,
+      "step": 43740
+    },
+    {
+      "epoch": 10.910224438902743,
+      "grad_norm": 8.43984317779541,
+      "learning_rate": 9.094763092269327e-06,
+      "loss": 0.3914,
+      "step": 43750
+    },
+    {
+      "epoch": 10.912718204488778,
+      "grad_norm": 6.167079925537109,
+      "learning_rate": 9.092269326683292e-06,
+      "loss": 0.3538,
+      "step": 43760
+    },
+    {
+      "epoch": 10.915211970074813,
+      "grad_norm": 5.880906581878662,
+      "learning_rate": 9.089775561097258e-06,
+      "loss": 0.3578,
+      "step": 43770
+    },
+    {
+      "epoch": 10.917705735660848,
+      "grad_norm": 11.39787483215332,
+      "learning_rate": 9.087281795511223e-06,
+      "loss": 0.3781,
+      "step": 43780
+    },
+    {
+      "epoch": 10.920199501246882,
+      "grad_norm": 11.926407814025879,
+      "learning_rate": 9.084788029925188e-06,
+      "loss": 0.3413,
+      "step": 43790
+    },
+    {
+      "epoch": 10.922693266832917,
+      "grad_norm": 9.85400390625,
+      "learning_rate": 9.082294264339153e-06,
+      "loss": 0.4972,
+      "step": 43800
+    },
+    {
+      "epoch": 10.925187032418952,
+      "grad_norm": 8.283854484558105,
+      "learning_rate": 9.079800498753117e-06,
+      "loss": 0.3979,
+      "step": 43810
+    },
+    {
+      "epoch": 10.927680798004987,
+      "grad_norm": 9.599810600280762,
+      "learning_rate": 9.077306733167082e-06,
+      "loss": 0.3405,
+      "step": 43820
+    },
+    {
+      "epoch": 10.930174563591022,
+      "grad_norm": 8.065226554870605,
+      "learning_rate": 9.074812967581048e-06,
+      "loss": 0.4098,
+      "step": 43830
+    },
+    {
+      "epoch": 10.932668329177057,
+      "grad_norm": 6.116038799285889,
+      "learning_rate": 9.072319201995013e-06,
+      "loss": 0.3745,
+      "step": 43840
+    },
+    {
+      "epoch": 10.935162094763092,
+      "grad_norm": 8.363578796386719,
+      "learning_rate": 9.069825436408978e-06,
+      "loss": 0.3899,
+      "step": 43850
+    },
+    {
+      "epoch": 10.937655860349127,
+      "grad_norm": 5.87863302230835,
+      "learning_rate": 9.067331670822944e-06,
+      "loss": 0.3468,
+      "step": 43860
+    },
+    {
+      "epoch": 10.940149625935161,
+      "grad_norm": 7.009591102600098,
+      "learning_rate": 9.064837905236909e-06,
+      "loss": 0.3486,
+      "step": 43870
+    },
+    {
+      "epoch": 10.942643391521196,
+      "grad_norm": 8.262612342834473,
+      "learning_rate": 9.062344139650874e-06,
+      "loss": 0.3601,
+      "step": 43880
+    },
+    {
+      "epoch": 10.945137157107231,
+      "grad_norm": 11.304754257202148,
+      "learning_rate": 9.05985037406484e-06,
+      "loss": 0.3933,
+      "step": 43890
+    },
+    {
+      "epoch": 10.947630922693268,
+      "grad_norm": 7.655803680419922,
+      "learning_rate": 9.057356608478803e-06,
+      "loss": 0.3177,
+      "step": 43900
+    },
+    {
+      "epoch": 10.950124688279303,
+      "grad_norm": 7.3225226402282715,
+      "learning_rate": 9.054862842892768e-06,
+      "loss": 0.3793,
+      "step": 43910
+    },
+    {
+      "epoch": 10.952618453865338,
+      "grad_norm": 11.696869850158691,
+      "learning_rate": 9.052369077306734e-06,
+      "loss": 0.3915,
+      "step": 43920
+    },
+    {
+      "epoch": 10.955112219451372,
+      "grad_norm": 7.450771331787109,
+      "learning_rate": 9.049875311720699e-06,
+      "loss": 0.3423,
+      "step": 43930
+    },
+    {
+      "epoch": 10.957605985037407,
+      "grad_norm": 6.9573655128479,
+      "learning_rate": 9.047381546134664e-06,
+      "loss": 0.3169,
+      "step": 43940
+    },
+    {
+      "epoch": 10.960099750623442,
+      "grad_norm": 6.270671844482422,
+      "learning_rate": 9.04488778054863e-06,
+      "loss": 0.3519,
+      "step": 43950
+    },
+    {
+      "epoch": 10.962593516209477,
+      "grad_norm": 8.106279373168945,
+      "learning_rate": 9.042394014962595e-06,
+      "loss": 0.3458,
+      "step": 43960
+    },
+    {
+      "epoch": 10.965087281795512,
+      "grad_norm": 2.7840754985809326,
+      "learning_rate": 9.03990024937656e-06,
+      "loss": 0.3244,
+      "step": 43970
+    },
+    {
+      "epoch": 10.967581047381547,
+      "grad_norm": 6.383938789367676,
+      "learning_rate": 9.037406483790524e-06,
+      "loss": 0.3242,
+      "step": 43980
+    },
+    {
+      "epoch": 10.970074812967582,
+      "grad_norm": 8.362710952758789,
+      "learning_rate": 9.03491271820449e-06,
+      "loss": 0.3966,
+      "step": 43990
+    },
+    {
+      "epoch": 10.972568578553616,
+      "grad_norm": 8.788472175598145,
+      "learning_rate": 9.032418952618455e-06,
+      "loss": 0.3279,
+      "step": 44000
+    },
+    {
+      "epoch": 10.975062344139651,
+      "grad_norm": 7.121370792388916,
+      "learning_rate": 9.02992518703242e-06,
+      "loss": 0.3171,
+      "step": 44010
+    },
+    {
+      "epoch": 10.977556109725686,
+      "grad_norm": 8.656166076660156,
+      "learning_rate": 9.027431421446384e-06,
+      "loss": 0.3256,
+      "step": 44020
+    },
+    {
+      "epoch": 10.980049875311721,
+      "grad_norm": 7.024209022521973,
+      "learning_rate": 9.024937655860349e-06,
+      "loss": 0.394,
+      "step": 44030
+    },
+    {
+      "epoch": 10.982543640897756,
+      "grad_norm": 8.922717094421387,
+      "learning_rate": 9.022443890274316e-06,
+      "loss": 0.32,
+      "step": 44040
+    },
+    {
+      "epoch": 10.98503740648379,
+      "grad_norm": 7.146017074584961,
+      "learning_rate": 9.019950124688281e-06,
+      "loss": 0.3368,
+      "step": 44050
+    },
+    {
+      "epoch": 10.987531172069826,
+      "grad_norm": 6.3751749992370605,
+      "learning_rate": 9.017456359102245e-06,
+      "loss": 0.3405,
+      "step": 44060
+    },
+    {
+      "epoch": 10.99002493765586,
+      "grad_norm": 4.537290096282959,
+      "learning_rate": 9.01496259351621e-06,
+      "loss": 0.2856,
+      "step": 44070
+    },
+    {
+      "epoch": 10.992518703241895,
+      "grad_norm": 9.014129638671875,
+      "learning_rate": 9.012468827930175e-06,
+      "loss": 0.3719,
+      "step": 44080
+    },
+    {
+      "epoch": 10.99501246882793,
+      "grad_norm": 7.991206645965576,
+      "learning_rate": 9.00997506234414e-06,
+      "loss": 0.3403,
+      "step": 44090
+    },
+    {
+      "epoch": 10.997506234413965,
+      "grad_norm": 6.865902900695801,
+      "learning_rate": 9.007481296758106e-06,
+      "loss": 0.3167,
+      "step": 44100
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 8.729063034057617,
+      "learning_rate": 9.00498753117207e-06,
+      "loss": 0.3674,
+      "step": 44110
+    },
+    {
+      "epoch": 11.0,
+      "eval_loss": 0.4132729470729828,
+      "eval_runtime": 59.9748,
+      "eval_samples_per_second": 16.724,
+      "eval_steps_per_second": 16.724,
+      "step": 44110
+    },
+    {
+      "epoch": 11.002493765586035,
+      "grad_norm": 8.256260871887207,
+      "learning_rate": 9.002493765586035e-06,
+      "loss": 0.3417,
+      "step": 44120
+    },
+    {
+      "epoch": 11.00498753117207,
+      "grad_norm": 8.294591903686523,
+      "learning_rate": 9e-06,
+      "loss": 0.3305,
+      "step": 44130
+    },
+    {
+      "epoch": 11.007481296758105,
+      "grad_norm": 7.374525547027588,
+      "learning_rate": 8.997506234413966e-06,
+      "loss": 0.3561,
+      "step": 44140
+    },
+    {
+      "epoch": 11.00997506234414,
+      "grad_norm": 7.575953006744385,
+      "learning_rate": 8.995012468827931e-06,
+      "loss": 0.3326,
+      "step": 44150
+    },
+    {
+      "epoch": 11.012468827930174,
+      "grad_norm": 8.2825345993042,
+      "learning_rate": 8.992518703241896e-06,
+      "loss": 0.3432,
+      "step": 44160
+    },
+    {
+      "epoch": 11.01496259351621,
+      "grad_norm": 6.32546854019165,
+      "learning_rate": 8.990024937655861e-06,
+      "loss": 0.345,
+      "step": 44170
+    },
+    {
+      "epoch": 11.017456359102244,
+      "grad_norm": 8.19118595123291,
+      "learning_rate": 8.987531172069827e-06,
+      "loss": 0.3045,
+      "step": 44180
+    },
+    {
+      "epoch": 11.019950124688279,
+      "grad_norm": 6.752303600311279,
+      "learning_rate": 8.98503740648379e-06,
+      "loss": 0.3918,
+      "step": 44190
+    },
+    {
+      "epoch": 11.022443890274314,
+      "grad_norm": 5.932526588439941,
+      "learning_rate": 8.982543640897756e-06,
+      "loss": 0.2983,
+      "step": 44200
+    },
+    {
+      "epoch": 11.024937655860349,
+      "grad_norm": 10.784844398498535,
+      "learning_rate": 8.980049875311721e-06,
+      "loss": 0.328,
+      "step": 44210
+    },
+    {
+      "epoch": 11.027431421446384,
+      "grad_norm": 6.154462814331055,
+      "learning_rate": 8.977556109725686e-06,
+      "loss": 0.3924,
+      "step": 44220
+    },
+    {
+      "epoch": 11.029925187032418,
+      "grad_norm": 6.966713905334473,
+      "learning_rate": 8.975062344139652e-06,
+      "loss": 0.3309,
+      "step": 44230
+    },
+    {
+      "epoch": 11.032418952618453,
+      "grad_norm": 4.729964256286621,
+      "learning_rate": 8.972568578553617e-06,
+      "loss": 0.3863,
+      "step": 44240
+    },
+    {
+      "epoch": 11.034912718204488,
+      "grad_norm": 7.567785739898682,
+      "learning_rate": 8.970074812967582e-06,
+      "loss": 0.305,
+      "step": 44250
+    },
+    {
+      "epoch": 11.037406483790523,
+      "grad_norm": 7.73862886428833,
+      "learning_rate": 8.967581047381548e-06,
+      "loss": 0.4145,
+      "step": 44260
+    },
+    {
+      "epoch": 11.039900249376558,
+      "grad_norm": 6.4053168296813965,
+      "learning_rate": 8.965087281795511e-06,
+      "loss": 0.3216,
+      "step": 44270
+    },
+    {
+      "epoch": 11.042394014962593,
+      "grad_norm": 4.572265148162842,
+      "learning_rate": 8.962593516209476e-06,
+      "loss": 0.3534,
+      "step": 44280
+    },
+    {
+      "epoch": 11.044887780548628,
+      "grad_norm": 5.619957447052002,
+      "learning_rate": 8.960099750623442e-06,
+      "loss": 0.3327,
+      "step": 44290
+    },
+    {
+      "epoch": 11.047381546134662,
+      "grad_norm": 13.95361042022705,
+      "learning_rate": 8.957605985037407e-06,
+      "loss": 0.3672,
+      "step": 44300
+    },
+    {
+      "epoch": 11.049875311720697,
+      "grad_norm": 5.5569963455200195,
+      "learning_rate": 8.955112219451372e-06,
+      "loss": 0.4164,
+      "step": 44310
+    },
+    {
+      "epoch": 11.052369077306734,
+      "grad_norm": 5.730642318725586,
+      "learning_rate": 8.952618453865338e-06,
+      "loss": 0.2852,
+      "step": 44320
+    },
+    {
+      "epoch": 11.054862842892769,
+      "grad_norm": 10.969292640686035,
+      "learning_rate": 8.950124688279303e-06,
+      "loss": 0.3387,
+      "step": 44330
+    },
+    {
+      "epoch": 11.057356608478804,
+      "grad_norm": 6.42288064956665,
+      "learning_rate": 8.947630922693268e-06,
+      "loss": 0.3338,
+      "step": 44340
+    },
+    {
+      "epoch": 11.059850374064839,
+      "grad_norm": 7.737945556640625,
+      "learning_rate": 8.945137157107234e-06,
+      "loss": 0.3709,
+      "step": 44350
+    },
+    {
+      "epoch": 11.062344139650873,
+      "grad_norm": 8.051152229309082,
+      "learning_rate": 8.942643391521197e-06,
+      "loss": 0.356,
+      "step": 44360
+    },
+    {
+      "epoch": 11.064837905236908,
+      "grad_norm": 6.058350563049316,
+      "learning_rate": 8.940149625935163e-06,
+      "loss": 0.2939,
+      "step": 44370
+    },
+    {
+      "epoch": 11.067331670822943,
+      "grad_norm": 7.329736232757568,
+      "learning_rate": 8.937655860349128e-06,
+      "loss": 0.3819,
+      "step": 44380
+    },
+    {
+      "epoch": 11.069825436408978,
+      "grad_norm": 7.439964771270752,
+      "learning_rate": 8.935162094763093e-06,
+      "loss": 0.325,
+      "step": 44390
+    },
+    {
+      "epoch": 11.072319201995013,
+      "grad_norm": 7.003705978393555,
+      "learning_rate": 8.932668329177059e-06,
+      "loss": 0.3168,
+      "step": 44400
+    },
+    {
+      "epoch": 11.074812967581048,
+      "grad_norm": 5.3289570808410645,
+      "learning_rate": 8.930423940149627e-06,
+      "loss": 0.3069,
+      "step": 44410
+    },
+    {
+      "epoch": 11.077306733167083,
+      "grad_norm": 6.427891731262207,
+      "learning_rate": 8.927930174563592e-06,
+      "loss": 0.3028,
+      "step": 44420
+    },
+    {
+      "epoch": 11.079800498753118,
+      "grad_norm": 7.827418327331543,
+      "learning_rate": 8.925436408977557e-06,
+      "loss": 0.3543,
+      "step": 44430
+    },
+    {
+      "epoch": 11.082294264339152,
+      "grad_norm": 9.805115699768066,
+      "learning_rate": 8.922942643391523e-06,
+      "loss": 0.3216,
+      "step": 44440
+    },
+    {
+      "epoch": 11.084788029925187,
+      "grad_norm": 6.640481948852539,
+      "learning_rate": 8.920448877805486e-06,
+      "loss": 0.372,
+      "step": 44450
+    },
+    {
+      "epoch": 11.087281795511222,
+      "grad_norm": 7.233997821807861,
+      "learning_rate": 8.917955112219452e-06,
+      "loss": 0.3521,
+      "step": 44460
+    },
+    {
+      "epoch": 11.089775561097257,
+      "grad_norm": 7.673696517944336,
+      "learning_rate": 8.915461346633417e-06,
+      "loss": 0.3184,
+      "step": 44470
+    },
+    {
+      "epoch": 11.092269326683292,
+      "grad_norm": 7.302619934082031,
+      "learning_rate": 8.912967581047382e-06,
+      "loss": 0.3365,
+      "step": 44480
+    },
+    {
+      "epoch": 11.094763092269327,
+      "grad_norm": 9.098041534423828,
+      "learning_rate": 8.910473815461348e-06,
+      "loss": 0.3513,
+      "step": 44490
+    },
+    {
+      "epoch": 11.097256857855362,
+      "grad_norm": 6.597679615020752,
+      "learning_rate": 8.907980049875313e-06,
+      "loss": 0.3133,
+      "step": 44500
+    },
+    {
+      "epoch": 11.099750623441397,
+      "grad_norm": 6.860264301300049,
+      "learning_rate": 8.905486284289278e-06,
+      "loss": 0.4096,
+      "step": 44510
+    },
+    {
+      "epoch": 11.102244389027431,
+      "grad_norm": 4.556776523590088,
+      "learning_rate": 8.902992518703243e-06,
+      "loss": 0.3389,
+      "step": 44520
+    },
+    {
+      "epoch": 11.104738154613466,
+      "grad_norm": 9.30516529083252,
+      "learning_rate": 8.900498753117209e-06,
+      "loss": 0.3051,
+      "step": 44530
+    },
+    {
+      "epoch": 11.107231920199501,
+      "grad_norm": 10.449723243713379,
+      "learning_rate": 8.898004987531172e-06,
+      "loss": 0.3412,
+      "step": 44540
+    },
+    {
+      "epoch": 11.109725685785536,
+      "grad_norm": 7.47760009765625,
+      "learning_rate": 8.895511221945138e-06,
+      "loss": 0.3901,
+      "step": 44550
+    },
+    {
+      "epoch": 11.11221945137157,
+      "grad_norm": 12.093515396118164,
+      "learning_rate": 8.893017456359103e-06,
+      "loss": 0.3252,
+      "step": 44560
+    },
+    {
+      "epoch": 11.114713216957606,
+      "grad_norm": 9.427157402038574,
+      "learning_rate": 8.890523690773068e-06,
+      "loss": 0.3101,
+      "step": 44570
+    },
+    {
+      "epoch": 11.11720698254364,
+      "grad_norm": 10.620285034179688,
+      "learning_rate": 8.888029925187032e-06,
+      "loss": 0.3178,
+      "step": 44580
+    },
+    {
+      "epoch": 11.119700748129675,
+      "grad_norm": 7.5438079833984375,
+      "learning_rate": 8.885536159600999e-06,
+      "loss": 0.3176,
+      "step": 44590
+    },
+    {
+      "epoch": 11.12219451371571,
+      "grad_norm": 6.367938041687012,
+      "learning_rate": 8.883042394014964e-06,
+      "loss": 0.3195,
+      "step": 44600
+    },
+    {
+      "epoch": 11.124688279301745,
+      "grad_norm": 16.17655372619629,
+      "learning_rate": 8.88054862842893e-06,
+      "loss": 0.3953,
+      "step": 44610
+    },
+    {
+      "epoch": 11.12718204488778,
+      "grad_norm": 10.65559196472168,
+      "learning_rate": 8.878054862842893e-06,
+      "loss": 0.3251,
+      "step": 44620
+    },
+    {
+      "epoch": 11.129675810473815,
+      "grad_norm": 6.366180419921875,
+      "learning_rate": 8.875561097256859e-06,
+      "loss": 0.3519,
+      "step": 44630
+    },
+    {
+      "epoch": 11.13216957605985,
+      "grad_norm": 8.204825401306152,
+      "learning_rate": 8.873067331670824e-06,
+      "loss": 0.3497,
+      "step": 44640
+    },
+    {
+      "epoch": 11.134663341645885,
+      "grad_norm": 10.065892219543457,
+      "learning_rate": 8.870573566084789e-06,
+      "loss": 0.2901,
+      "step": 44650
+    },
+    {
+      "epoch": 11.13715710723192,
+      "grad_norm": 7.356030464172363,
+      "learning_rate": 8.868079800498753e-06,
+      "loss": 0.3841,
+      "step": 44660
+    },
+    {
+      "epoch": 11.139650872817954,
+      "grad_norm": 7.364969730377197,
+      "learning_rate": 8.865586034912718e-06,
+      "loss": 0.3247,
+      "step": 44670
+    },
+    {
+      "epoch": 11.14214463840399,
+      "grad_norm": 7.613313674926758,
+      "learning_rate": 8.863092269326683e-06,
+      "loss": 0.3508,
+      "step": 44680
+    },
+    {
+      "epoch": 11.144638403990024,
+      "grad_norm": 8.078596115112305,
+      "learning_rate": 8.860598503740649e-06,
+      "loss": 0.3305,
+      "step": 44690
+    },
+    {
+      "epoch": 11.147132169576059,
+      "grad_norm": 10.208447456359863,
+      "learning_rate": 8.858104738154614e-06,
+      "loss": 0.3537,
+      "step": 44700
+    },
+    {
+      "epoch": 11.149625935162096,
+      "grad_norm": 8.215570449829102,
+      "learning_rate": 8.85561097256858e-06,
+      "loss": 0.3275,
+      "step": 44710
+    },
+    {
+      "epoch": 11.15211970074813,
+      "grad_norm": 5.312885284423828,
+      "learning_rate": 8.853117206982545e-06,
+      "loss": 0.3202,
+      "step": 44720
+    },
+    {
+      "epoch": 11.154613466334165,
+      "grad_norm": 7.666956901550293,
+      "learning_rate": 8.85062344139651e-06,
+      "loss": 0.3238,
+      "step": 44730
+    },
+    {
+      "epoch": 11.1571072319202,
+      "grad_norm": 7.116825103759766,
+      "learning_rate": 8.848129675810474e-06,
+      "loss": 0.3231,
+      "step": 44740
+    },
+    {
+      "epoch": 11.159600997506235,
+      "grad_norm": 6.489451885223389,
+      "learning_rate": 8.845635910224439e-06,
+      "loss": 0.3411,
+      "step": 44750
+    },
+    {
+      "epoch": 11.16209476309227,
+      "grad_norm": 6.022216320037842,
+      "learning_rate": 8.843142144638404e-06,
+      "loss": 0.4006,
+      "step": 44760
+    },
+    {
+      "epoch": 11.164588528678305,
+      "grad_norm": 8.421282768249512,
+      "learning_rate": 8.84064837905237e-06,
+      "loss": 0.4011,
+      "step": 44770
+    },
+    {
+      "epoch": 11.16708229426434,
+      "grad_norm": 7.724409580230713,
+      "learning_rate": 8.838154613466335e-06,
+      "loss": 0.325,
+      "step": 44780
+    },
+    {
+      "epoch": 11.169576059850375,
+      "grad_norm": 8.80372142791748,
+      "learning_rate": 8.8356608478803e-06,
+      "loss": 0.3197,
+      "step": 44790
+    },
+    {
+      "epoch": 11.17206982543641,
+      "grad_norm": 10.359016418457031,
+      "learning_rate": 8.833167082294265e-06,
+      "loss": 0.3586,
+      "step": 44800
+    },
+    {
+      "epoch": 11.174563591022444,
+      "grad_norm": 8.572481155395508,
+      "learning_rate": 8.83067331670823e-06,
+      "loss": 0.3292,
+      "step": 44810
+    },
+    {
+      "epoch": 11.17705735660848,
+      "grad_norm": 13.053606033325195,
+      "learning_rate": 8.828179551122196e-06,
+      "loss": 0.3645,
+      "step": 44820
+    },
+    {
+      "epoch": 11.179551122194514,
+      "grad_norm": 10.115473747253418,
+      "learning_rate": 8.82568578553616e-06,
+      "loss": 0.2904,
+      "step": 44830
+    },
+    {
+      "epoch": 11.182044887780549,
+      "grad_norm": 8.798450469970703,
+      "learning_rate": 8.823192019950125e-06,
+      "loss": 0.334,
+      "step": 44840
+    },
+    {
+      "epoch": 11.184538653366584,
+      "grad_norm": 8.745047569274902,
+      "learning_rate": 8.82069825436409e-06,
+      "loss": 0.3467,
+      "step": 44850
+    },
+    {
+      "epoch": 11.187032418952619,
+      "grad_norm": 6.6508941650390625,
+      "learning_rate": 8.818204488778056e-06,
+      "loss": 0.4056,
+      "step": 44860
+    },
+    {
+      "epoch": 11.189526184538654,
+      "grad_norm": 7.657419204711914,
+      "learning_rate": 8.815710723192021e-06,
+      "loss": 0.3708,
+      "step": 44870
+    },
+    {
+      "epoch": 11.192019950124688,
+      "grad_norm": 7.383201599121094,
+      "learning_rate": 8.813216957605986e-06,
+      "loss": 0.3759,
+      "step": 44880
+    },
+    {
+      "epoch": 11.194513715710723,
+      "grad_norm": 5.854804992675781,
+      "learning_rate": 8.810723192019951e-06,
+      "loss": 0.3151,
+      "step": 44890
+    },
+    {
+      "epoch": 11.197007481296758,
+      "grad_norm": 6.18996000289917,
+      "learning_rate": 8.808229426433917e-06,
+      "loss": 0.3223,
+      "step": 44900
+    },
+    {
+      "epoch": 11.199501246882793,
+      "grad_norm": 7.372582912445068,
+      "learning_rate": 8.80573566084788e-06,
+      "loss": 0.3174,
+      "step": 44910
+    },
+    {
+      "epoch": 11.201995012468828,
+      "grad_norm": 4.5158233642578125,
+      "learning_rate": 8.803241895261846e-06,
+      "loss": 0.2967,
+      "step": 44920
+    },
+    {
+      "epoch": 11.204488778054863,
+      "grad_norm": 8.73290729522705,
+      "learning_rate": 8.800748129675811e-06,
+      "loss": 0.3015,
+      "step": 44930
+    },
+    {
+      "epoch": 11.206982543640898,
+      "grad_norm": 12.908029556274414,
+      "learning_rate": 8.798254364089776e-06,
+      "loss": 0.351,
+      "step": 44940
+    },
+    {
+      "epoch": 11.209476309226932,
+      "grad_norm": 6.993527889251709,
+      "learning_rate": 8.795760598503742e-06,
+      "loss": 0.3274,
+      "step": 44950
+    },
+    {
+      "epoch": 11.211970074812967,
+      "grad_norm": 8.653640747070312,
+      "learning_rate": 8.793266832917707e-06,
+      "loss": 0.3817,
+      "step": 44960
+    },
+    {
+      "epoch": 11.214463840399002,
+      "grad_norm": 4.629430770874023,
+      "learning_rate": 8.790773067331672e-06,
+      "loss": 0.3772,
+      "step": 44970
+    },
+    {
+      "epoch": 11.216957605985037,
+      "grad_norm": 7.76257848739624,
+      "learning_rate": 8.788279301745638e-06,
+      "loss": 0.3459,
+      "step": 44980
+    },
+    {
+      "epoch": 11.219451371571072,
+      "grad_norm": 8.239355087280273,
+      "learning_rate": 8.785785536159601e-06,
+      "loss": 0.3039,
+      "step": 44990
+    },
+    {
+      "epoch": 11.221945137157107,
+      "grad_norm": 8.355988502502441,
+      "learning_rate": 8.783291770573566e-06,
+      "loss": 0.363,
+      "step": 45000
+    },
+    {
+      "epoch": 11.224438902743142,
+      "grad_norm": 7.481101989746094,
+      "learning_rate": 8.780798004987532e-06,
+      "loss": 0.3703,
+      "step": 45010
+    },
+    {
+      "epoch": 11.226932668329177,
+      "grad_norm": 4.93237829208374,
+      "learning_rate": 8.778304239401497e-06,
+      "loss": 0.3426,
+      "step": 45020
+    },
+    {
+      "epoch": 11.229426433915211,
+      "grad_norm": 5.618884563446045,
+      "learning_rate": 8.775810473815462e-06,
+      "loss": 0.3202,
+      "step": 45030
+    },
+    {
+      "epoch": 11.231920199501246,
+      "grad_norm": 12.628522872924805,
+      "learning_rate": 8.773316708229426e-06,
+      "loss": 0.36,
+      "step": 45040
+    },
+    {
+      "epoch": 11.234413965087281,
+      "grad_norm": 7.585283279418945,
+      "learning_rate": 8.770822942643391e-06,
+      "loss": 0.3422,
+      "step": 45050
+    },
+    {
+      "epoch": 11.236907730673316,
+      "grad_norm": 7.396781921386719,
+      "learning_rate": 8.768329177057358e-06,
+      "loss": 0.3276,
+      "step": 45060
+    },
+    {
+      "epoch": 11.239401496259351,
+      "grad_norm": 13.765165328979492,
+      "learning_rate": 8.765835411471324e-06,
+      "loss": 0.3655,
+      "step": 45070
+    },
+    {
+      "epoch": 11.241895261845386,
+      "grad_norm": 7.234029769897461,
+      "learning_rate": 8.763341645885287e-06,
+      "loss": 0.3718,
+      "step": 45080
+    },
+    {
+      "epoch": 11.24438902743142,
+      "grad_norm": 4.461481094360352,
+      "learning_rate": 8.760847880299253e-06,
+      "loss": 0.3412,
+      "step": 45090
+    },
+    {
+      "epoch": 11.246882793017456,
+      "grad_norm": 8.242216110229492,
+      "learning_rate": 8.758354114713218e-06,
+      "loss": 0.3455,
+      "step": 45100
+    },
+    {
+      "epoch": 11.24937655860349,
+      "grad_norm": 10.74194622039795,
+      "learning_rate": 8.755860349127183e-06,
+      "loss": 0.3511,
+      "step": 45110
+    },
+    {
+      "epoch": 11.251870324189527,
+      "grad_norm": 9.051909446716309,
+      "learning_rate": 8.753366583541147e-06,
+      "loss": 0.3333,
+      "step": 45120
+    },
+    {
+      "epoch": 11.254364089775562,
+      "grad_norm": 9.297087669372559,
+      "learning_rate": 8.750872817955112e-06,
+      "loss": 0.3345,
+      "step": 45130
+    },
+    {
+      "epoch": 11.256857855361597,
+      "grad_norm": 7.317951679229736,
+      "learning_rate": 8.748379052369077e-06,
+      "loss": 0.3696,
+      "step": 45140
+    },
+    {
+      "epoch": 11.259351620947632,
+      "grad_norm": 6.456167221069336,
+      "learning_rate": 8.745885286783043e-06,
+      "loss": 0.2631,
+      "step": 45150
+    },
+    {
+      "epoch": 11.261845386533667,
+      "grad_norm": 9.548169136047363,
+      "learning_rate": 8.743391521197008e-06,
+      "loss": 0.3358,
+      "step": 45160
+    },
+    {
+      "epoch": 11.264339152119701,
+      "grad_norm": 4.220322608947754,
+      "learning_rate": 8.740897755610973e-06,
+      "loss": 0.309,
+      "step": 45170
+    },
+    {
+      "epoch": 11.266832917705736,
+      "grad_norm": 7.340099334716797,
+      "learning_rate": 8.738403990024939e-06,
+      "loss": 0.2948,
+      "step": 45180
+    },
+    {
+      "epoch": 11.269326683291771,
+      "grad_norm": 6.146076202392578,
+      "learning_rate": 8.735910224438904e-06,
+      "loss": 0.3575,
+      "step": 45190
+    },
+    {
+      "epoch": 11.271820448877806,
+      "grad_norm": 8.388291358947754,
+      "learning_rate": 8.733416458852868e-06,
+      "loss": 0.3306,
+      "step": 45200
+    },
+    {
+      "epoch": 11.27431421446384,
+      "grad_norm": 12.431069374084473,
+      "learning_rate": 8.730922693266833e-06,
+      "loss": 0.3423,
+      "step": 45210
+    },
+    {
+      "epoch": 11.276807980049876,
+      "grad_norm": 7.34416389465332,
+      "learning_rate": 8.728428927680798e-06,
+      "loss": 0.3473,
+      "step": 45220
+    },
+    {
+      "epoch": 11.27930174563591,
+      "grad_norm": 4.615060806274414,
+      "learning_rate": 8.725935162094764e-06,
+      "loss": 0.3393,
+      "step": 45230
+    },
+    {
+      "epoch": 11.281795511221945,
+      "grad_norm": 7.668536186218262,
+      "learning_rate": 8.723441396508729e-06,
+      "loss": 0.3442,
+      "step": 45240
+    },
+    {
+      "epoch": 11.28428927680798,
+      "grad_norm": 8.232738494873047,
+      "learning_rate": 8.720947630922694e-06,
+      "loss": 0.334,
+      "step": 45250
+    },
+    {
+      "epoch": 11.286783042394015,
+      "grad_norm": 9.045811653137207,
+      "learning_rate": 8.71845386533666e-06,
+      "loss": 0.428,
+      "step": 45260
+    },
+    {
+      "epoch": 11.28927680798005,
+      "grad_norm": 7.1297526359558105,
+      "learning_rate": 8.715960099750625e-06,
+      "loss": 0.3156,
+      "step": 45270
+    },
+    {
+      "epoch": 11.291770573566085,
+      "grad_norm": 8.763090133666992,
+      "learning_rate": 8.71346633416459e-06,
+      "loss": 0.3447,
+      "step": 45280
+    },
+    {
+      "epoch": 11.29426433915212,
+      "grad_norm": 8.615846633911133,
+      "learning_rate": 8.710972568578554e-06,
+      "loss": 0.3308,
+      "step": 45290
+    },
+    {
+      "epoch": 11.296758104738155,
+      "grad_norm": 10.320062637329102,
+      "learning_rate": 8.708478802992519e-06,
+      "loss": 0.3569,
+      "step": 45300
+    },
+    {
+      "epoch": 11.29925187032419,
+      "grad_norm": 7.310534954071045,
+      "learning_rate": 8.705985037406484e-06,
+      "loss": 0.362,
+      "step": 45310
+    },
+    {
+      "epoch": 11.301745635910224,
+      "grad_norm": 6.569928169250488,
+      "learning_rate": 8.70349127182045e-06,
+      "loss": 0.2893,
+      "step": 45320
+    },
+    {
+      "epoch": 11.30423940149626,
+      "grad_norm": 6.204464435577393,
+      "learning_rate": 8.700997506234415e-06,
+      "loss": 0.3239,
+      "step": 45330
+    },
+    {
+      "epoch": 11.306733167082294,
+      "grad_norm": 6.959688663482666,
+      "learning_rate": 8.69850374064838e-06,
+      "loss": 0.3394,
+      "step": 45340
+    },
+    {
+      "epoch": 11.309226932668329,
+      "grad_norm": 8.13005256652832,
+      "learning_rate": 8.696009975062346e-06,
+      "loss": 0.3523,
+      "step": 45350
+    },
+    {
+      "epoch": 11.311720698254364,
+      "grad_norm": 5.307433128356934,
+      "learning_rate": 8.693516209476311e-06,
+      "loss": 0.3485,
+      "step": 45360
+    },
+    {
+      "epoch": 11.314214463840399,
+      "grad_norm": 7.195748329162598,
+      "learning_rate": 8.691022443890274e-06,
+      "loss": 0.3992,
+      "step": 45370
+    },
+    {
+      "epoch": 11.316708229426434,
+      "grad_norm": 8.693116188049316,
+      "learning_rate": 8.68852867830424e-06,
+      "loss": 0.3582,
+      "step": 45380
+    },
+    {
+      "epoch": 11.319201995012468,
+      "grad_norm": 7.39258337020874,
+      "learning_rate": 8.686034912718205e-06,
+      "loss": 0.33,
+      "step": 45390
+    },
+    {
+      "epoch": 11.321695760598503,
+      "grad_norm": 7.933567047119141,
+      "learning_rate": 8.68354114713217e-06,
+      "loss": 0.3464,
+      "step": 45400
+    },
+    {
+      "epoch": 11.324189526184538,
+      "grad_norm": 8.168654441833496,
+      "learning_rate": 8.681047381546136e-06,
+      "loss": 0.3156,
+      "step": 45410
+    },
+    {
+      "epoch": 11.326683291770573,
+      "grad_norm": 10.021574974060059,
+      "learning_rate": 8.678553615960101e-06,
+      "loss": 0.3644,
+      "step": 45420
+    },
+    {
+      "epoch": 11.329177057356608,
+      "grad_norm": 7.775262832641602,
+      "learning_rate": 8.676059850374066e-06,
+      "loss": 0.3631,
+      "step": 45430
+    },
+    {
+      "epoch": 11.331670822942643,
+      "grad_norm": 8.023462295532227,
+      "learning_rate": 8.673566084788032e-06,
+      "loss": 0.3567,
+      "step": 45440
+    },
+    {
+      "epoch": 11.334164588528678,
+      "grad_norm": 10.410110473632812,
+      "learning_rate": 8.671072319201995e-06,
+      "loss": 0.3938,
+      "step": 45450
+    },
+    {
+      "epoch": 11.336658354114713,
+      "grad_norm": 10.069509506225586,
+      "learning_rate": 8.66857855361596e-06,
+      "loss": 0.4094,
+      "step": 45460
+    },
+    {
+      "epoch": 11.339152119700747,
+      "grad_norm": 5.57853364944458,
+      "learning_rate": 8.666084788029926e-06,
+      "loss": 0.3017,
+      "step": 45470
+    },
+    {
+      "epoch": 11.341645885286782,
+      "grad_norm": 7.023243427276611,
+      "learning_rate": 8.663591022443891e-06,
+      "loss": 0.3085,
+      "step": 45480
+    },
+    {
+      "epoch": 11.344139650872817,
+      "grad_norm": 6.86598014831543,
+      "learning_rate": 8.661097256857855e-06,
+      "loss": 0.3638,
+      "step": 45490
+    },
+    {
+      "epoch": 11.346633416458852,
+      "grad_norm": 8.568735122680664,
+      "learning_rate": 8.65860349127182e-06,
+      "loss": 0.3982,
+      "step": 45500
+    },
+    {
+      "epoch": 11.349127182044889,
+      "grad_norm": 8.5831880569458,
+      "learning_rate": 8.656109725685785e-06,
+      "loss": 0.3462,
+      "step": 45510
+    },
+    {
+      "epoch": 11.351620947630924,
+      "grad_norm": 6.069805145263672,
+      "learning_rate": 8.65361596009975e-06,
+      "loss": 0.2987,
+      "step": 45520
+    },
+    {
+      "epoch": 11.354114713216958,
+      "grad_norm": 7.621370315551758,
+      "learning_rate": 8.651122194513718e-06,
+      "loss": 0.3374,
+      "step": 45530
+    },
+    {
+      "epoch": 11.356608478802993,
+      "grad_norm": 6.071798324584961,
+      "learning_rate": 8.648628428927681e-06,
+      "loss": 0.3222,
+      "step": 45540
+    },
+    {
+      "epoch": 11.359102244389028,
+      "grad_norm": 5.18229341506958,
+      "learning_rate": 8.646134663341647e-06,
+      "loss": 0.3218,
+      "step": 45550
+    },
+    {
+      "epoch": 11.361596009975063,
+      "grad_norm": 6.7007293701171875,
+      "learning_rate": 8.643640897755612e-06,
+      "loss": 0.3675,
+      "step": 45560
+    },
+    {
+      "epoch": 11.364089775561098,
+      "grad_norm": 5.716126918792725,
+      "learning_rate": 8.641147132169577e-06,
+      "loss": 0.336,
+      "step": 45570
+    },
+    {
+      "epoch": 11.366583541147133,
+      "grad_norm": 6.14850378036499,
+      "learning_rate": 8.638653366583541e-06,
+      "loss": 0.3959,
+      "step": 45580
+    },
+    {
+      "epoch": 11.369077306733168,
+      "grad_norm": 4.6964898109436035,
+      "learning_rate": 8.636159600997506e-06,
+      "loss": 0.343,
+      "step": 45590
+    },
+    {
+      "epoch": 11.371571072319203,
+      "grad_norm": 8.84803581237793,
+      "learning_rate": 8.633665835411472e-06,
+      "loss": 0.332,
+      "step": 45600
+    },
+    {
+      "epoch": 11.374064837905237,
+      "grad_norm": 5.964016914367676,
+      "learning_rate": 8.631172069825437e-06,
+      "loss": 0.3045,
+      "step": 45610
+    },
+    {
+      "epoch": 11.376558603491272,
+      "grad_norm": 4.455854415893555,
+      "learning_rate": 8.628678304239402e-06,
+      "loss": 0.3272,
+      "step": 45620
+    },
+    {
+      "epoch": 11.379052369077307,
+      "grad_norm": 8.803267478942871,
+      "learning_rate": 8.626184538653367e-06,
+      "loss": 0.337,
+      "step": 45630
+    },
+    {
+      "epoch": 11.381546134663342,
+      "grad_norm": 9.714362144470215,
+      "learning_rate": 8.623690773067333e-06,
+      "loss": 0.3589,
+      "step": 45640
+    },
+    {
+      "epoch": 11.384039900249377,
+      "grad_norm": 4.950045108795166,
+      "learning_rate": 8.621197007481298e-06,
+      "loss": 0.3321,
+      "step": 45650
+    },
+    {
+      "epoch": 11.386533665835412,
+      "grad_norm": 8.148637771606445,
+      "learning_rate": 8.618703241895262e-06,
+      "loss": 0.3178,
+      "step": 45660
+    },
+    {
+      "epoch": 11.389027431421447,
+      "grad_norm": 21.125547409057617,
+      "learning_rate": 8.616209476309227e-06,
+      "loss": 0.3481,
+      "step": 45670
+    },
+    {
+      "epoch": 11.391521197007481,
+      "grad_norm": 7.620448112487793,
+      "learning_rate": 8.613715710723192e-06,
+      "loss": 0.3811,
+      "step": 45680
+    },
+    {
+      "epoch": 11.394014962593516,
+      "grad_norm": 7.3992743492126465,
+      "learning_rate": 8.611221945137158e-06,
+      "loss": 0.3509,
+      "step": 45690
+    },
+    {
+      "epoch": 11.396508728179551,
+      "grad_norm": 7.22592830657959,
+      "learning_rate": 8.608728179551123e-06,
+      "loss": 0.3888,
+      "step": 45700
+    },
+    {
+      "epoch": 11.399002493765586,
+      "grad_norm": 7.094658851623535,
+      "learning_rate": 8.606234413965088e-06,
+      "loss": 0.3421,
+      "step": 45710
+    },
+    {
+      "epoch": 11.401496259351621,
+      "grad_norm": 7.960940837860107,
+      "learning_rate": 8.603740648379054e-06,
+      "loss": 0.4286,
+      "step": 45720
+    },
+    {
+      "epoch": 11.403990024937656,
+      "grad_norm": 10.1074857711792,
+      "learning_rate": 8.601246882793019e-06,
+      "loss": 0.3917,
+      "step": 45730
+    },
+    {
+      "epoch": 11.40648379052369,
+      "grad_norm": 10.943230628967285,
+      "learning_rate": 8.598753117206984e-06,
+      "loss": 0.3824,
+      "step": 45740
+    },
+    {
+      "epoch": 11.408977556109726,
+      "grad_norm": 6.1284685134887695,
+      "learning_rate": 8.596259351620948e-06,
+      "loss": 0.2953,
+      "step": 45750
+    },
+    {
+      "epoch": 11.41147132169576,
+      "grad_norm": 6.4964776039123535,
+      "learning_rate": 8.593765586034913e-06,
+      "loss": 0.3705,
+      "step": 45760
+    },
+    {
+      "epoch": 11.413965087281795,
+      "grad_norm": 7.87790584564209,
+      "learning_rate": 8.591271820448878e-06,
+      "loss": 0.3237,
+      "step": 45770
+    },
+    {
+      "epoch": 11.41645885286783,
+      "grad_norm": 7.569821357727051,
+      "learning_rate": 8.588778054862844e-06,
+      "loss": 0.303,
+      "step": 45780
+    },
+    {
+      "epoch": 11.418952618453865,
+      "grad_norm": 8.668131828308105,
+      "learning_rate": 8.586284289276809e-06,
+      "loss": 0.2865,
+      "step": 45790
+    },
+    {
+      "epoch": 11.4214463840399,
+      "grad_norm": 8.034530639648438,
+      "learning_rate": 8.583790523690774e-06,
+      "loss": 0.3503,
+      "step": 45800
+    },
+    {
+      "epoch": 11.423940149625935,
+      "grad_norm": 8.05600357055664,
+      "learning_rate": 8.58129675810474e-06,
+      "loss": 0.3285,
+      "step": 45810
+    },
+    {
+      "epoch": 11.42643391521197,
+      "grad_norm": 7.00352668762207,
+      "learning_rate": 8.578802992518705e-06,
+      "loss": 0.3335,
+      "step": 45820
+    },
+    {
+      "epoch": 11.428927680798004,
+      "grad_norm": 12.199906349182129,
+      "learning_rate": 8.576309226932669e-06,
+      "loss": 0.4104,
+      "step": 45830
+    },
+    {
+      "epoch": 11.43142144638404,
+      "grad_norm": 8.1293306350708,
+      "learning_rate": 8.573815461346634e-06,
+      "loss": 0.3462,
+      "step": 45840
+    },
+    {
+      "epoch": 11.433915211970074,
+      "grad_norm": 9.158780097961426,
+      "learning_rate": 8.5713216957606e-06,
+      "loss": 0.3138,
+      "step": 45850
+    },
+    {
+      "epoch": 11.436408977556109,
+      "grad_norm": 13.985179901123047,
+      "learning_rate": 8.568827930174564e-06,
+      "loss": 0.3182,
+      "step": 45860
+    },
+    {
+      "epoch": 11.438902743142144,
+      "grad_norm": 7.04974889755249,
+      "learning_rate": 8.566334164588528e-06,
+      "loss": 0.3452,
+      "step": 45870
+    },
+    {
+      "epoch": 11.441396508728179,
+      "grad_norm": 7.684841632843018,
+      "learning_rate": 8.563840399002495e-06,
+      "loss": 0.3312,
+      "step": 45880
+    },
+    {
+      "epoch": 11.443890274314214,
+      "grad_norm": 7.361491680145264,
+      "learning_rate": 8.56134663341646e-06,
+      "loss": 0.3133,
+      "step": 45890
+    },
+    {
+      "epoch": 11.446384039900249,
+      "grad_norm": 8.644780158996582,
+      "learning_rate": 8.558852867830426e-06,
+      "loss": 0.3371,
+      "step": 45900
+    },
+    {
+      "epoch": 11.448877805486283,
+      "grad_norm": 8.315042495727539,
+      "learning_rate": 8.55635910224439e-06,
+      "loss": 0.3792,
+      "step": 45910
+    },
+    {
+      "epoch": 11.451371571072318,
+      "grad_norm": 8.369626998901367,
+      "learning_rate": 8.553865336658355e-06,
+      "loss": 0.3525,
+      "step": 45920
+    },
+    {
+      "epoch": 11.453865336658355,
+      "grad_norm": 12.835140228271484,
+      "learning_rate": 8.55137157107232e-06,
+      "loss": 0.3088,
+      "step": 45930
+    },
+    {
+      "epoch": 11.45635910224439,
+      "grad_norm": 10.077214241027832,
+      "learning_rate": 8.548877805486285e-06,
+      "loss": 0.3456,
+      "step": 45940
+    },
+    {
+      "epoch": 11.458852867830425,
+      "grad_norm": 8.558807373046875,
+      "learning_rate": 8.546384039900249e-06,
+      "loss": 0.2983,
+      "step": 45950
+    },
+    {
+      "epoch": 11.46134663341646,
+      "grad_norm": 10.858403205871582,
+      "learning_rate": 8.543890274314214e-06,
+      "loss": 0.3396,
+      "step": 45960
+    },
+    {
+      "epoch": 11.463840399002494,
+      "grad_norm": 6.8004469871521,
+      "learning_rate": 8.54139650872818e-06,
+      "loss": 0.3272,
+      "step": 45970
+    },
+    {
+      "epoch": 11.46633416458853,
+      "grad_norm": 6.694581985473633,
+      "learning_rate": 8.538902743142145e-06,
+      "loss": 0.3775,
+      "step": 45980
+    },
+    {
+      "epoch": 11.468827930174564,
+      "grad_norm": 8.675992012023926,
+      "learning_rate": 8.53640897755611e-06,
+      "loss": 0.3789,
+      "step": 45990
+    },
+    {
+      "epoch": 11.471321695760599,
+      "grad_norm": 6.141960620880127,
+      "learning_rate": 8.533915211970075e-06,
+      "loss": 0.3364,
+      "step": 46000
+    },
+    {
+      "epoch": 11.473815461346634,
+      "grad_norm": 14.198030471801758,
+      "learning_rate": 8.53142144638404e-06,
+      "loss": 0.3398,
+      "step": 46010
+    },
+    {
+      "epoch": 11.476309226932669,
+      "grad_norm": 6.926119327545166,
+      "learning_rate": 8.528927680798006e-06,
+      "loss": 0.2966,
+      "step": 46020
+    },
+    {
+      "epoch": 11.478802992518704,
+      "grad_norm": 5.021880626678467,
+      "learning_rate": 8.526433915211971e-06,
+      "loss": 0.3197,
+      "step": 46030
+    },
+    {
+      "epoch": 11.481296758104738,
+      "grad_norm": 6.237401008605957,
+      "learning_rate": 8.523940149625935e-06,
+      "loss": 0.3225,
+      "step": 46040
+    },
+    {
+      "epoch": 11.483790523690773,
+      "grad_norm": 11.097943305969238,
+      "learning_rate": 8.5214463840399e-06,
+      "loss": 0.4856,
+      "step": 46050
+    },
+    {
+      "epoch": 11.486284289276808,
+      "grad_norm": 8.584773063659668,
+      "learning_rate": 8.518952618453866e-06,
+      "loss": 0.3588,
+      "step": 46060
+    },
+    {
+      "epoch": 11.488778054862843,
+      "grad_norm": 5.824810981750488,
+      "learning_rate": 8.516458852867831e-06,
+      "loss": 0.2905,
+      "step": 46070
+    },
+    {
+      "epoch": 11.491271820448878,
+      "grad_norm": 8.207144737243652,
+      "learning_rate": 8.513965087281796e-06,
+      "loss": 0.3239,
+      "step": 46080
+    },
+    {
+      "epoch": 11.493765586034913,
+      "grad_norm": 8.661499977111816,
+      "learning_rate": 8.511471321695762e-06,
+      "loss": 0.3761,
+      "step": 46090
+    },
+    {
+      "epoch": 11.496259351620948,
+      "grad_norm": 6.3622941970825195,
+      "learning_rate": 8.508977556109727e-06,
+      "loss": 0.346,
+      "step": 46100
+    },
+    {
+      "epoch": 11.498753117206983,
+      "grad_norm": 5.697100639343262,
+      "learning_rate": 8.506483790523692e-06,
+      "loss": 0.3489,
+      "step": 46110
+    },
+    {
+      "epoch": 11.501246882793017,
+      "grad_norm": 10.397525787353516,
+      "learning_rate": 8.503990024937656e-06,
+      "loss": 0.4333,
+      "step": 46120
+    },
+    {
+      "epoch": 11.503740648379052,
+      "grad_norm": 11.356361389160156,
+      "learning_rate": 8.501496259351621e-06,
+      "loss": 0.3788,
+      "step": 46130
+    },
+    {
+      "epoch": 11.506234413965087,
+      "grad_norm": 10.53273868560791,
+      "learning_rate": 8.499002493765586e-06,
+      "loss": 0.4482,
+      "step": 46140
+    },
+    {
+      "epoch": 11.508728179551122,
+      "grad_norm": 7.545499324798584,
+      "learning_rate": 8.496508728179552e-06,
+      "loss": 0.3394,
+      "step": 46150
+    },
+    {
+      "epoch": 11.511221945137157,
+      "grad_norm": 5.596482276916504,
+      "learning_rate": 8.494014962593517e-06,
+      "loss": 0.3306,
+      "step": 46160
+    },
+    {
+      "epoch": 11.513715710723192,
+      "grad_norm": 7.381369590759277,
+      "learning_rate": 8.491521197007482e-06,
+      "loss": 0.3239,
+      "step": 46170
+    },
+    {
+      "epoch": 11.516209476309227,
+      "grad_norm": 9.224650382995605,
+      "learning_rate": 8.489027431421448e-06,
+      "loss": 0.3463,
+      "step": 46180
+    },
+    {
+      "epoch": 11.518703241895262,
+      "grad_norm": 6.10484504699707,
+      "learning_rate": 8.486533665835413e-06,
+      "loss": 0.457,
+      "step": 46190
+    },
+    {
+      "epoch": 11.521197007481296,
+      "grad_norm": 8.393863677978516,
+      "learning_rate": 8.484039900249377e-06,
+      "loss": 0.3499,
+      "step": 46200
+    },
+    {
+      "epoch": 11.523690773067331,
+      "grad_norm": 10.636434555053711,
+      "learning_rate": 8.481546134663342e-06,
+      "loss": 0.429,
+      "step": 46210
+    },
+    {
+      "epoch": 11.526184538653366,
+      "grad_norm": 8.618541717529297,
+      "learning_rate": 8.479052369077307e-06,
+      "loss": 0.355,
+      "step": 46220
+    },
+    {
+      "epoch": 11.528678304239401,
+      "grad_norm": 5.982786655426025,
+      "learning_rate": 8.476558603491272e-06,
+      "loss": 0.3373,
+      "step": 46230
+    },
+    {
+      "epoch": 11.531172069825436,
+      "grad_norm": 10.3162202835083,
+      "learning_rate": 8.474064837905238e-06,
+      "loss": 0.2873,
+      "step": 46240
+    },
+    {
+      "epoch": 11.53366583541147,
+      "grad_norm": 5.130282878875732,
+      "learning_rate": 8.471571072319203e-06,
+      "loss": 0.2964,
+      "step": 46250
+    },
+    {
+      "epoch": 11.536159600997506,
+      "grad_norm": 5.8889360427856445,
+      "learning_rate": 8.469077306733168e-06,
+      "loss": 0.3102,
+      "step": 46260
+    },
+    {
+      "epoch": 11.53865336658354,
+      "grad_norm": 4.879234313964844,
+      "learning_rate": 8.466583541147134e-06,
+      "loss": 0.3211,
+      "step": 46270
+    },
+    {
+      "epoch": 11.541147132169575,
+      "grad_norm": 7.243957042694092,
+      "learning_rate": 8.464089775561099e-06,
+      "loss": 0.3416,
+      "step": 46280
+    },
+    {
+      "epoch": 11.54364089775561,
+      "grad_norm": 8.841197967529297,
+      "learning_rate": 8.461596009975063e-06,
+      "loss": 0.3346,
+      "step": 46290
+    },
+    {
+      "epoch": 11.546134663341645,
+      "grad_norm": 7.920102119445801,
+      "learning_rate": 8.459102244389028e-06,
+      "loss": 0.3587,
+      "step": 46300
+    },
+    {
+      "epoch": 11.548628428927682,
+      "grad_norm": 8.432982444763184,
+      "learning_rate": 8.456608478802993e-06,
+      "loss": 0.3364,
+      "step": 46310
+    },
+    {
+      "epoch": 11.551122194513717,
+      "grad_norm": 7.410879135131836,
+      "learning_rate": 8.454114713216959e-06,
+      "loss": 0.4021,
+      "step": 46320
+    },
+    {
+      "epoch": 11.553615960099751,
+      "grad_norm": 9.919821739196777,
+      "learning_rate": 8.451620947630922e-06,
+      "loss": 0.4665,
+      "step": 46330
+    },
+    {
+      "epoch": 11.556109725685786,
+      "grad_norm": 6.146061420440674,
+      "learning_rate": 8.449127182044888e-06,
+      "loss": 0.3562,
+      "step": 46340
+    },
+    {
+      "epoch": 11.558603491271821,
+      "grad_norm": 6.502902507781982,
+      "learning_rate": 8.446633416458855e-06,
+      "loss": 0.3341,
+      "step": 46350
+    },
+    {
+      "epoch": 11.561097256857856,
+      "grad_norm": 7.523041725158691,
+      "learning_rate": 8.44413965087282e-06,
+      "loss": 0.3158,
+      "step": 46360
+    },
+    {
+      "epoch": 11.563591022443891,
+      "grad_norm": 5.467228412628174,
+      "learning_rate": 8.441645885286783e-06,
+      "loss": 0.3464,
+      "step": 46370
+    },
+    {
+      "epoch": 11.566084788029926,
+      "grad_norm": 7.694350242614746,
+      "learning_rate": 8.439152119700749e-06,
+      "loss": 0.3111,
+      "step": 46380
+    },
+    {
+      "epoch": 11.56857855361596,
+      "grad_norm": 13.799556732177734,
+      "learning_rate": 8.436658354114714e-06,
+      "loss": 0.3468,
+      "step": 46390
+    },
+    {
+      "epoch": 11.571072319201996,
+      "grad_norm": 8.76827621459961,
+      "learning_rate": 8.43416458852868e-06,
+      "loss": 0.3029,
+      "step": 46400
+    },
+    {
+      "epoch": 11.57356608478803,
+      "grad_norm": 11.704144477844238,
+      "learning_rate": 8.431670822942643e-06,
+      "loss": 0.3546,
+      "step": 46410
+    },
+    {
+      "epoch": 11.576059850374065,
+      "grad_norm": 8.31985855102539,
+      "learning_rate": 8.429177057356608e-06,
+      "loss": 0.3563,
+      "step": 46420
+    },
+    {
+      "epoch": 11.5785536159601,
+      "grad_norm": 6.618094444274902,
+      "learning_rate": 8.426683291770574e-06,
+      "loss": 0.2983,
+      "step": 46430
+    },
+    {
+      "epoch": 11.581047381546135,
+      "grad_norm": 13.60894775390625,
+      "learning_rate": 8.424189526184539e-06,
+      "loss": 0.3255,
+      "step": 46440
+    },
+    {
+      "epoch": 11.58354114713217,
+      "grad_norm": 5.429731369018555,
+      "learning_rate": 8.421695760598504e-06,
+      "loss": 0.3448,
+      "step": 46450
+    },
+    {
+      "epoch": 11.586034912718205,
+      "grad_norm": 5.1549224853515625,
+      "learning_rate": 8.41920199501247e-06,
+      "loss": 0.3547,
+      "step": 46460
+    },
+    {
+      "epoch": 11.58852867830424,
+      "grad_norm": 6.888156890869141,
+      "learning_rate": 8.416708229426435e-06,
+      "loss": 0.3126,
+      "step": 46470
+    },
+    {
+      "epoch": 11.591022443890274,
+      "grad_norm": 8.632163047790527,
+      "learning_rate": 8.4142144638404e-06,
+      "loss": 0.3216,
+      "step": 46480
+    },
+    {
+      "epoch": 11.59351620947631,
+      "grad_norm": 8.811448097229004,
+      "learning_rate": 8.411720698254365e-06,
+      "loss": 0.3131,
+      "step": 46490
+    },
+    {
+      "epoch": 11.596009975062344,
+      "grad_norm": 7.669614791870117,
+      "learning_rate": 8.409226932668329e-06,
+      "loss": 0.3222,
+      "step": 46500
+    },
+    {
+      "epoch": 11.598503740648379,
+      "grad_norm": 12.28158950805664,
+      "learning_rate": 8.406733167082294e-06,
+      "loss": 0.3609,
+      "step": 46510
+    },
+    {
+      "epoch": 11.600997506234414,
+      "grad_norm": 12.088496208190918,
+      "learning_rate": 8.40423940149626e-06,
+      "loss": 0.3294,
+      "step": 46520
+    },
+    {
+      "epoch": 11.603491271820449,
+      "grad_norm": 7.644925594329834,
+      "learning_rate": 8.401745635910225e-06,
+      "loss": 0.3499,
+      "step": 46530
+    },
+    {
+      "epoch": 11.605985037406484,
+      "grad_norm": 8.952886581420898,
+      "learning_rate": 8.39925187032419e-06,
+      "loss": 0.4748,
+      "step": 46540
+    },
+    {
+      "epoch": 11.608478802992519,
+      "grad_norm": 7.003511428833008,
+      "learning_rate": 8.396758104738156e-06,
+      "loss": 0.3068,
+      "step": 46550
+    },
+    {
+      "epoch": 11.610972568578553,
+      "grad_norm": 9.938292503356934,
+      "learning_rate": 8.394264339152121e-06,
+      "loss": 0.3804,
+      "step": 46560
+    },
+    {
+      "epoch": 11.613466334164588,
+      "grad_norm": 6.018657684326172,
+      "learning_rate": 8.391770573566086e-06,
+      "loss": 0.3036,
+      "step": 46570
+    },
+    {
+      "epoch": 11.615960099750623,
+      "grad_norm": 4.126202583312988,
+      "learning_rate": 8.38927680798005e-06,
+      "loss": 0.2909,
+      "step": 46580
+    },
+    {
+      "epoch": 11.618453865336658,
+      "grad_norm": 4.8967132568359375,
+      "learning_rate": 8.386783042394015e-06,
+      "loss": 0.2835,
+      "step": 46590
+    },
+    {
+      "epoch": 11.620947630922693,
+      "grad_norm": 7.265740871429443,
+      "learning_rate": 8.38428927680798e-06,
+      "loss": 0.3781,
+      "step": 46600
+    },
+    {
+      "epoch": 11.623441396508728,
+      "grad_norm": 8.042332649230957,
+      "learning_rate": 8.381795511221946e-06,
+      "loss": 0.3315,
+      "step": 46610
+    },
+    {
+      "epoch": 11.625935162094763,
+      "grad_norm": 9.90203857421875,
+      "learning_rate": 8.379301745635911e-06,
+      "loss": 0.3179,
+      "step": 46620
+    },
+    {
+      "epoch": 11.628428927680797,
+      "grad_norm": 5.170167446136475,
+      "learning_rate": 8.376807980049876e-06,
+      "loss": 0.3211,
+      "step": 46630
+    },
+    {
+      "epoch": 11.630922693266832,
+      "grad_norm": 11.844198226928711,
+      "learning_rate": 8.374314214463842e-06,
+      "loss": 0.3923,
+      "step": 46640
+    },
+    {
+      "epoch": 11.633416458852867,
+      "grad_norm": 6.6527862548828125,
+      "learning_rate": 8.371820448877807e-06,
+      "loss": 0.3308,
+      "step": 46650
+    },
+    {
+      "epoch": 11.635910224438902,
+      "grad_norm": 10.204750061035156,
+      "learning_rate": 8.36932668329177e-06,
+      "loss": 0.4115,
+      "step": 46660
+    },
+    {
+      "epoch": 11.638403990024937,
+      "grad_norm": 5.856804847717285,
+      "learning_rate": 8.366832917705736e-06,
+      "loss": 0.3647,
+      "step": 46670
+    },
+    {
+      "epoch": 11.640897755610972,
+      "grad_norm": 5.034014701843262,
+      "learning_rate": 8.364339152119701e-06,
+      "loss": 0.344,
+      "step": 46680
+    },
+    {
+      "epoch": 11.643391521197007,
+      "grad_norm": 6.951847076416016,
+      "learning_rate": 8.361845386533667e-06,
+      "loss": 0.3241,
+      "step": 46690
+    },
+    {
+      "epoch": 11.645885286783042,
+      "grad_norm": 6.871116638183594,
+      "learning_rate": 8.359351620947632e-06,
+      "loss": 0.3133,
+      "step": 46700
+    },
+    {
+      "epoch": 11.648379052369076,
+      "grad_norm": 9.318560600280762,
+      "learning_rate": 8.356857855361597e-06,
+      "loss": 0.3388,
+      "step": 46710
+    },
+    {
+      "epoch": 11.650872817955111,
+      "grad_norm": 10.321300506591797,
+      "learning_rate": 8.354364089775563e-06,
+      "loss": 0.3425,
+      "step": 46720
+    },
+    {
+      "epoch": 11.653366583541148,
+      "grad_norm": 6.536388874053955,
+      "learning_rate": 8.351870324189528e-06,
+      "loss": 0.4034,
+      "step": 46730
+    },
+    {
+      "epoch": 11.655860349127183,
+      "grad_norm": 7.95648717880249,
+      "learning_rate": 8.349376558603493e-06,
+      "loss": 0.387,
+      "step": 46740
+    },
+    {
+      "epoch": 11.658354114713218,
+      "grad_norm": 4.451998710632324,
+      "learning_rate": 8.346882793017457e-06,
+      "loss": 0.3272,
+      "step": 46750
+    },
+    {
+      "epoch": 11.660847880299253,
+      "grad_norm": 5.903833389282227,
+      "learning_rate": 8.344389027431422e-06,
+      "loss": 0.3689,
+      "step": 46760
+    },
+    {
+      "epoch": 11.663341645885287,
+      "grad_norm": 9.391934394836426,
+      "learning_rate": 8.341895261845387e-06,
+      "loss": 0.3681,
+      "step": 46770
+    },
+    {
+      "epoch": 11.665835411471322,
+      "grad_norm": 7.286145210266113,
+      "learning_rate": 8.339401496259353e-06,
+      "loss": 0.3198,
+      "step": 46780
+    },
+    {
+      "epoch": 11.668329177057357,
+      "grad_norm": 7.64100456237793,
+      "learning_rate": 8.336907730673316e-06,
+      "loss": 0.3116,
+      "step": 46790
+    },
+    {
+      "epoch": 11.670822942643392,
+      "grad_norm": 6.605118751525879,
+      "learning_rate": 8.334413965087282e-06,
+      "loss": 0.2764,
+      "step": 46800
+    },
+    {
+      "epoch": 11.673316708229427,
+      "grad_norm": 7.660946369171143,
+      "learning_rate": 8.331920199501249e-06,
+      "loss": 0.4052,
+      "step": 46810
+    },
+    {
+      "epoch": 11.675810473815462,
+      "grad_norm": 5.863259315490723,
+      "learning_rate": 8.329426433915214e-06,
+      "loss": 0.3599,
+      "step": 46820
+    },
+    {
+      "epoch": 11.678304239401497,
+      "grad_norm": 8.02147388458252,
+      "learning_rate": 8.326932668329178e-06,
+      "loss": 0.3401,
+      "step": 46830
+    },
+    {
+      "epoch": 11.680798004987532,
+      "grad_norm": 7.035722255706787,
+      "learning_rate": 8.324438902743143e-06,
+      "loss": 0.3427,
+      "step": 46840
+    },
+    {
+      "epoch": 11.683291770573566,
+      "grad_norm": 8.856318473815918,
+      "learning_rate": 8.321945137157108e-06,
+      "loss": 0.3217,
+      "step": 46850
+    },
+    {
+      "epoch": 11.685785536159601,
+      "grad_norm": 12.737922668457031,
+      "learning_rate": 8.319451371571073e-06,
+      "loss": 0.3705,
+      "step": 46860
+    },
+    {
+      "epoch": 11.688279301745636,
+      "grad_norm": 7.125262260437012,
+      "learning_rate": 8.316957605985037e-06,
+      "loss": 0.2677,
+      "step": 46870
+    },
+    {
+      "epoch": 11.690773067331671,
+      "grad_norm": 9.105042457580566,
+      "learning_rate": 8.314463840399002e-06,
+      "loss": 0.3248,
+      "step": 46880
+    },
+    {
+      "epoch": 11.693266832917706,
+      "grad_norm": 6.2453436851501465,
+      "learning_rate": 8.311970074812968e-06,
+      "loss": 0.3346,
+      "step": 46890
+    },
+    {
+      "epoch": 11.69576059850374,
+      "grad_norm": 9.779679298400879,
+      "learning_rate": 8.309476309226933e-06,
+      "loss": 0.3915,
+      "step": 46900
+    },
+    {
+      "epoch": 11.698254364089776,
+      "grad_norm": 8.361473083496094,
+      "learning_rate": 8.306982543640898e-06,
+      "loss": 0.3421,
+      "step": 46910
+    },
+    {
+      "epoch": 11.70074812967581,
+      "grad_norm": 8.903398513793945,
+      "learning_rate": 8.304488778054864e-06,
+      "loss": 0.3676,
+      "step": 46920
+    },
+    {
+      "epoch": 11.703241895261845,
+      "grad_norm": 7.4261016845703125,
+      "learning_rate": 8.301995012468829e-06,
+      "loss": 0.4291,
+      "step": 46930
+    },
+    {
+      "epoch": 11.70573566084788,
+      "grad_norm": 10.814020156860352,
+      "learning_rate": 8.299501246882794e-06,
+      "loss": 0.3552,
+      "step": 46940
+    },
+    {
+      "epoch": 11.708229426433915,
+      "grad_norm": 6.702065467834473,
+      "learning_rate": 8.297007481296758e-06,
+      "loss": 0.2939,
+      "step": 46950
+    },
+    {
+      "epoch": 11.71072319201995,
+      "grad_norm": 8.938770294189453,
+      "learning_rate": 8.294513715710723e-06,
+      "loss": 0.3835,
+      "step": 46960
+    },
+    {
+      "epoch": 11.713216957605985,
+      "grad_norm": 8.248211860656738,
+      "learning_rate": 8.292019950124688e-06,
+      "loss": 0.3454,
+      "step": 46970
+    },
+    {
+      "epoch": 11.71571072319202,
+      "grad_norm": 5.559350967407227,
+      "learning_rate": 8.289526184538654e-06,
+      "loss": 0.4123,
+      "step": 46980
+    },
+    {
+      "epoch": 11.718204488778055,
+      "grad_norm": 7.779847145080566,
+      "learning_rate": 8.287032418952619e-06,
+      "loss": 0.3581,
+      "step": 46990
+    },
+    {
+      "epoch": 11.72069825436409,
+      "grad_norm": 6.771767616271973,
+      "learning_rate": 8.284538653366584e-06,
+      "loss": 0.3246,
+      "step": 47000
+    },
+    {
+      "epoch": 11.723192019950124,
+      "grad_norm": 6.579812049865723,
+      "learning_rate": 8.28204488778055e-06,
+      "loss": 0.3924,
+      "step": 47010
+    },
+    {
+      "epoch": 11.72568578553616,
+      "grad_norm": 10.849081039428711,
+      "learning_rate": 8.279551122194515e-06,
+      "loss": 0.4271,
+      "step": 47020
+    },
+    {
+      "epoch": 11.728179551122194,
+      "grad_norm": 7.1727824211120605,
+      "learning_rate": 8.27705735660848e-06,
+      "loss": 0.3297,
+      "step": 47030
+    },
+    {
+      "epoch": 11.730673316708229,
+      "grad_norm": 9.346254348754883,
+      "learning_rate": 8.274563591022444e-06,
+      "loss": 0.336,
+      "step": 47040
+    },
+    {
+      "epoch": 11.733167082294264,
+      "grad_norm": 9.930221557617188,
+      "learning_rate": 8.27206982543641e-06,
+      "loss": 0.3422,
+      "step": 47050
+    },
+    {
+      "epoch": 11.735660847880299,
+      "grad_norm": 8.86354923248291,
+      "learning_rate": 8.269576059850375e-06,
+      "loss": 0.3921,
+      "step": 47060
+    },
+    {
+      "epoch": 11.738154613466333,
+      "grad_norm": 6.808560371398926,
+      "learning_rate": 8.26708229426434e-06,
+      "loss": 0.3291,
+      "step": 47070
+    },
+    {
+      "epoch": 11.740648379052368,
+      "grad_norm": 8.037238121032715,
+      "learning_rate": 8.264588528678305e-06,
+      "loss": 0.3975,
+      "step": 47080
+    },
+    {
+      "epoch": 11.743142144638403,
+      "grad_norm": 6.9091010093688965,
+      "learning_rate": 8.26209476309227e-06,
+      "loss": 0.2959,
+      "step": 47090
+    },
+    {
+      "epoch": 11.745635910224438,
+      "grad_norm": 7.109037399291992,
+      "learning_rate": 8.259600997506236e-06,
+      "loss": 0.3498,
+      "step": 47100
+    },
+    {
+      "epoch": 11.748129675810475,
+      "grad_norm": 13.113969802856445,
+      "learning_rate": 8.257107231920201e-06,
+      "loss": 0.3391,
+      "step": 47110
+    },
+    {
+      "epoch": 11.75062344139651,
+      "grad_norm": 10.038171768188477,
+      "learning_rate": 8.254613466334165e-06,
+      "loss": 0.3651,
+      "step": 47120
+    },
+    {
+      "epoch": 11.753117206982544,
+      "grad_norm": 9.342373847961426,
+      "learning_rate": 8.25211970074813e-06,
+      "loss": 0.3071,
+      "step": 47130
+    },
+    {
+      "epoch": 11.75561097256858,
+      "grad_norm": 6.506314277648926,
+      "learning_rate": 8.249625935162095e-06,
+      "loss": 0.4171,
+      "step": 47140
+    },
+    {
+      "epoch": 11.758104738154614,
+      "grad_norm": 9.478023529052734,
+      "learning_rate": 8.24713216957606e-06,
+      "loss": 0.3932,
+      "step": 47150
+    },
+    {
+      "epoch": 11.760598503740649,
+      "grad_norm": 6.497440814971924,
+      "learning_rate": 8.244638403990026e-06,
+      "loss": 0.3667,
+      "step": 47160
+    },
+    {
+      "epoch": 11.763092269326684,
+      "grad_norm": 7.87533712387085,
+      "learning_rate": 8.242144638403991e-06,
+      "loss": 0.3391,
+      "step": 47170
+    },
+    {
+      "epoch": 11.765586034912719,
+      "grad_norm": 5.2254557609558105,
+      "learning_rate": 8.239650872817957e-06,
+      "loss": 0.3359,
+      "step": 47180
+    },
+    {
+      "epoch": 11.768079800498754,
+      "grad_norm": 6.769067287445068,
+      "learning_rate": 8.237157107231922e-06,
+      "loss": 0.3649,
+      "step": 47190
+    },
+    {
+      "epoch": 11.770573566084789,
+      "grad_norm": 9.426133155822754,
+      "learning_rate": 8.234663341645886e-06,
+      "loss": 0.3654,
+      "step": 47200
+    },
+    {
+      "epoch": 11.773067331670823,
+      "grad_norm": 14.436060905456543,
+      "learning_rate": 8.23216957605985e-06,
+      "loss": 0.404,
+      "step": 47210
+    },
+    {
+      "epoch": 11.775561097256858,
+      "grad_norm": 5.8233137130737305,
+      "learning_rate": 8.229675810473816e-06,
+      "loss": 0.3201,
+      "step": 47220
+    },
+    {
+      "epoch": 11.778054862842893,
+      "grad_norm": 8.890005111694336,
+      "learning_rate": 8.227182044887781e-06,
+      "loss": 0.3244,
+      "step": 47230
+    },
+    {
+      "epoch": 11.780548628428928,
+      "grad_norm": 9.727415084838867,
+      "learning_rate": 8.224688279301747e-06,
+      "loss": 0.3729,
+      "step": 47240
+    },
+    {
+      "epoch": 11.783042394014963,
+      "grad_norm": 7.346099853515625,
+      "learning_rate": 8.22219451371571e-06,
+      "loss": 0.3708,
+      "step": 47250
+    },
+    {
+      "epoch": 11.785536159600998,
+      "grad_norm": 7.652705192565918,
+      "learning_rate": 8.219700748129676e-06,
+      "loss": 0.3199,
+      "step": 47260
+    },
+    {
+      "epoch": 11.788029925187033,
+      "grad_norm": 9.516809463500977,
+      "learning_rate": 8.217206982543641e-06,
+      "loss": 0.3202,
+      "step": 47270
+    },
+    {
+      "epoch": 11.790523690773068,
+      "grad_norm": 6.376964569091797,
+      "learning_rate": 8.214713216957608e-06,
+      "loss": 0.3306,
+      "step": 47280
+    },
+    {
+      "epoch": 11.793017456359102,
+      "grad_norm": 7.07741641998291,
+      "learning_rate": 8.212219451371572e-06,
+      "loss": 0.3668,
+      "step": 47290
+    },
+    {
+      "epoch": 11.795511221945137,
+      "grad_norm": 8.181482315063477,
+      "learning_rate": 8.209725685785537e-06,
+      "loss": 0.3843,
+      "step": 47300
+    },
+    {
+      "epoch": 11.798004987531172,
+      "grad_norm": 9.766548156738281,
+      "learning_rate": 8.207231920199502e-06,
+      "loss": 0.3964,
+      "step": 47310
+    },
+    {
+      "epoch": 11.800498753117207,
+      "grad_norm": 7.150549411773682,
+      "learning_rate": 8.204738154613468e-06,
+      "loss": 0.3973,
+      "step": 47320
+    },
+    {
+      "epoch": 11.802992518703242,
+      "grad_norm": 6.823990821838379,
+      "learning_rate": 8.202244389027431e-06,
+      "loss": 0.3975,
+      "step": 47330
+    },
+    {
+      "epoch": 11.805486284289277,
+      "grad_norm": 7.102343559265137,
+      "learning_rate": 8.199750623441396e-06,
+      "loss": 0.3856,
+      "step": 47340
+    },
+    {
+      "epoch": 11.807980049875312,
+      "grad_norm": 7.7780585289001465,
+      "learning_rate": 8.197256857855362e-06,
+      "loss": 0.3287,
+      "step": 47350
+    },
+    {
+      "epoch": 11.810473815461346,
+      "grad_norm": 16.928571701049805,
+      "learning_rate": 8.194763092269327e-06,
+      "loss": 0.315,
+      "step": 47360
+    },
+    {
+      "epoch": 11.812967581047381,
+      "grad_norm": 9.088364601135254,
+      "learning_rate": 8.192269326683292e-06,
+      "loss": 0.3788,
+      "step": 47370
+    },
+    {
+      "epoch": 11.815461346633416,
+      "grad_norm": 6.3990044593811035,
+      "learning_rate": 8.189775561097258e-06,
+      "loss": 0.3317,
+      "step": 47380
+    },
+    {
+      "epoch": 11.817955112219451,
+      "grad_norm": 8.615131378173828,
+      "learning_rate": 8.187281795511223e-06,
+      "loss": 0.449,
+      "step": 47390
+    },
+    {
+      "epoch": 11.820448877805486,
+      "grad_norm": 5.688783168792725,
+      "learning_rate": 8.184788029925188e-06,
+      "loss": 0.3203,
+      "step": 47400
+    },
+    {
+      "epoch": 11.82294264339152,
+      "grad_norm": 9.329352378845215,
+      "learning_rate": 8.182294264339152e-06,
+      "loss": 0.4166,
+      "step": 47410
+    },
+    {
+      "epoch": 11.825436408977556,
+      "grad_norm": 5.748128414154053,
+      "learning_rate": 8.179800498753117e-06,
+      "loss": 0.3375,
+      "step": 47420
+    },
+    {
+      "epoch": 11.82793017456359,
+      "grad_norm": 8.721724510192871,
+      "learning_rate": 8.177306733167083e-06,
+      "loss": 0.361,
+      "step": 47430
+    },
+    {
+      "epoch": 11.830423940149625,
+      "grad_norm": 8.815082550048828,
+      "learning_rate": 8.174812967581048e-06,
+      "loss": 0.3311,
+      "step": 47440
+    },
+    {
+      "epoch": 11.83291770573566,
+      "grad_norm": 9.213290214538574,
+      "learning_rate": 8.172319201995013e-06,
+      "loss": 0.3334,
+      "step": 47450
+    },
+    {
+      "epoch": 11.835411471321695,
+      "grad_norm": 10.43193531036377,
+      "learning_rate": 8.169825436408978e-06,
+      "loss": 0.4101,
+      "step": 47460
+    },
+    {
+      "epoch": 11.83790523690773,
+      "grad_norm": 7.8552350997924805,
+      "learning_rate": 8.167331670822944e-06,
+      "loss": 0.3305,
+      "step": 47470
+    },
+    {
+      "epoch": 11.840399002493765,
+      "grad_norm": 8.425485610961914,
+      "learning_rate": 8.164837905236909e-06,
+      "loss": 0.4038,
+      "step": 47480
+    },
+    {
+      "epoch": 11.8428927680798,
+      "grad_norm": 7.427910804748535,
+      "learning_rate": 8.162344139650874e-06,
+      "loss": 0.3453,
+      "step": 47490
+    },
+    {
+      "epoch": 11.845386533665835,
+      "grad_norm": 4.184894561767578,
+      "learning_rate": 8.159850374064838e-06,
+      "loss": 0.3295,
+      "step": 47500
+    },
+    {
+      "epoch": 11.84788029925187,
+      "grad_norm": 9.951261520385742,
+      "learning_rate": 8.157356608478803e-06,
+      "loss": 0.3587,
+      "step": 47510
+    },
+    {
+      "epoch": 11.850374064837904,
+      "grad_norm": 9.598414421081543,
+      "learning_rate": 8.154862842892769e-06,
+      "loss": 0.4135,
+      "step": 47520
+    },
+    {
+      "epoch": 11.85286783042394,
+      "grad_norm": 7.297958850860596,
+      "learning_rate": 8.152369077306734e-06,
+      "loss": 0.3501,
+      "step": 47530
+    },
+    {
+      "epoch": 11.855361596009976,
+      "grad_norm": 6.240098476409912,
+      "learning_rate": 8.1498753117207e-06,
+      "loss": 0.344,
+      "step": 47540
+    },
+    {
+      "epoch": 11.85785536159601,
+      "grad_norm": 7.912670135498047,
+      "learning_rate": 8.147381546134665e-06,
+      "loss": 0.2993,
+      "step": 47550
+    },
+    {
+      "epoch": 11.860349127182046,
+      "grad_norm": 7.632608890533447,
+      "learning_rate": 8.14488778054863e-06,
+      "loss": 0.3528,
+      "step": 47560
+    },
+    {
+      "epoch": 11.86284289276808,
+      "grad_norm": 12.331526756286621,
+      "learning_rate": 8.142394014962595e-06,
+      "loss": 0.4064,
+      "step": 47570
+    },
+    {
+      "epoch": 11.865336658354115,
+      "grad_norm": 10.794772148132324,
+      "learning_rate": 8.139900249376559e-06,
+      "loss": 0.3422,
+      "step": 47580
+    },
+    {
+      "epoch": 11.86783042394015,
+      "grad_norm": 9.130438804626465,
+      "learning_rate": 8.137406483790524e-06,
+      "loss": 0.3605,
+      "step": 47590
+    },
+    {
+      "epoch": 11.870324189526185,
+      "grad_norm": 6.942079067230225,
+      "learning_rate": 8.13491271820449e-06,
+      "loss": 0.3104,
+      "step": 47600
+    },
+    {
+      "epoch": 11.87281795511222,
+      "grad_norm": 12.40993595123291,
+      "learning_rate": 8.132418952618455e-06,
+      "loss": 0.2948,
+      "step": 47610
+    },
+    {
+      "epoch": 11.875311720698255,
+      "grad_norm": 6.655416011810303,
+      "learning_rate": 8.129925187032418e-06,
+      "loss": 0.3096,
+      "step": 47620
+    },
+    {
+      "epoch": 11.87780548628429,
+      "grad_norm": 6.1850175857543945,
+      "learning_rate": 8.127431421446385e-06,
+      "loss": 0.3626,
+      "step": 47630
+    },
+    {
+      "epoch": 11.880299251870325,
+      "grad_norm": 9.720369338989258,
+      "learning_rate": 8.12493765586035e-06,
+      "loss": 0.3307,
+      "step": 47640
+    },
+    {
+      "epoch": 11.88279301745636,
+      "grad_norm": 9.563785552978516,
+      "learning_rate": 8.122443890274316e-06,
+      "loss": 0.3062,
+      "step": 47650
+    },
+    {
+      "epoch": 11.885286783042394,
+      "grad_norm": 10.218947410583496,
+      "learning_rate": 8.11995012468828e-06,
+      "loss": 0.4228,
+      "step": 47660
+    },
+    {
+      "epoch": 11.88778054862843,
+      "grad_norm": 5.189842700958252,
+      "learning_rate": 8.117456359102245e-06,
+      "loss": 0.354,
+      "step": 47670
+    },
+    {
+      "epoch": 11.890274314214464,
+      "grad_norm": 6.639830112457275,
+      "learning_rate": 8.11496259351621e-06,
+      "loss": 0.3331,
+      "step": 47680
+    },
+    {
+      "epoch": 11.892768079800499,
+      "grad_norm": 8.736455917358398,
+      "learning_rate": 8.112468827930176e-06,
+      "loss": 0.3691,
+      "step": 47690
+    },
+    {
+      "epoch": 11.895261845386534,
+      "grad_norm": 7.995376110076904,
+      "learning_rate": 8.109975062344139e-06,
+      "loss": 0.3516,
+      "step": 47700
+    },
+    {
+      "epoch": 11.897755610972569,
+      "grad_norm": 7.362260341644287,
+      "learning_rate": 8.107481296758104e-06,
+      "loss": 0.4472,
+      "step": 47710
+    },
+    {
+      "epoch": 11.900249376558603,
+      "grad_norm": 9.83117961883545,
+      "learning_rate": 8.10498753117207e-06,
+      "loss": 0.3402,
+      "step": 47720
+    },
+    {
+      "epoch": 11.902743142144638,
+      "grad_norm": 7.758623123168945,
+      "learning_rate": 8.102493765586035e-06,
+      "loss": 0.3799,
+      "step": 47730
+    },
+    {
+      "epoch": 11.905236907730673,
+      "grad_norm": 9.175792694091797,
+      "learning_rate": 8.1e-06,
+      "loss": 0.391,
+      "step": 47740
+    },
+    {
+      "epoch": 11.907730673316708,
+      "grad_norm": 7.377160549163818,
+      "learning_rate": 8.097506234413966e-06,
+      "loss": 0.3799,
+      "step": 47750
+    },
+    {
+      "epoch": 11.910224438902743,
+      "grad_norm": 5.763824462890625,
+      "learning_rate": 8.095012468827931e-06,
+      "loss": 0.348,
+      "step": 47760
+    },
+    {
+      "epoch": 11.912718204488778,
+      "grad_norm": 8.172435760498047,
+      "learning_rate": 8.092518703241896e-06,
+      "loss": 0.4006,
+      "step": 47770
+    },
+    {
+      "epoch": 11.915211970074813,
+      "grad_norm": 9.261870384216309,
+      "learning_rate": 8.090024937655862e-06,
+      "loss": 0.3811,
+      "step": 47780
+    },
+    {
+      "epoch": 11.917705735660848,
+      "grad_norm": 11.933865547180176,
+      "learning_rate": 8.087531172069825e-06,
+      "loss": 0.4131,
+      "step": 47790
+    },
+    {
+      "epoch": 11.920199501246882,
+      "grad_norm": 8.930726051330566,
+      "learning_rate": 8.08503740648379e-06,
+      "loss": 0.3954,
+      "step": 47800
+    },
+    {
+      "epoch": 11.922693266832917,
+      "grad_norm": 7.120776653289795,
+      "learning_rate": 8.082543640897756e-06,
+      "loss": 0.3536,
+      "step": 47810
+    },
+    {
+      "epoch": 11.925187032418952,
+      "grad_norm": 6.882978439331055,
+      "learning_rate": 8.080049875311721e-06,
+      "loss": 0.351,
+      "step": 47820
+    },
+    {
+      "epoch": 11.927680798004987,
+      "grad_norm": 7.883481502532959,
+      "learning_rate": 8.077556109725686e-06,
+      "loss": 0.3385,
+      "step": 47830
+    },
+    {
+      "epoch": 11.930174563591022,
+      "grad_norm": 8.083521842956543,
+      "learning_rate": 8.075062344139652e-06,
+      "loss": 0.31,
+      "step": 47840
+    },
+    {
+      "epoch": 11.932668329177057,
+      "grad_norm": 7.6285080909729,
+      "learning_rate": 8.072568578553617e-06,
+      "loss": 0.3895,
+      "step": 47850
+    },
+    {
+      "epoch": 11.935162094763092,
+      "grad_norm": 5.246705532073975,
+      "learning_rate": 8.070074812967582e-06,
+      "loss": 0.3273,
+      "step": 47860
+    },
+    {
+      "epoch": 11.937655860349127,
+      "grad_norm": 7.005518913269043,
+      "learning_rate": 8.067581047381546e-06,
+      "loss": 0.3466,
+      "step": 47870
+    },
+    {
+      "epoch": 11.940149625935161,
+      "grad_norm": 10.184428215026855,
+      "learning_rate": 8.065087281795511e-06,
+      "loss": 0.3226,
+      "step": 47880
+    },
+    {
+      "epoch": 11.942643391521196,
+      "grad_norm": 6.404749870300293,
+      "learning_rate": 8.062593516209477e-06,
+      "loss": 0.3122,
+      "step": 47890
+    },
+    {
+      "epoch": 11.945137157107231,
+      "grad_norm": 9.22337818145752,
+      "learning_rate": 8.060099750623442e-06,
+      "loss": 0.4094,
+      "step": 47900
+    },
+    {
+      "epoch": 11.947630922693268,
+      "grad_norm": 6.901223182678223,
+      "learning_rate": 8.057605985037407e-06,
+      "loss": 0.3463,
+      "step": 47910
+    },
+    {
+      "epoch": 11.950124688279303,
+      "grad_norm": 7.0179033279418945,
+      "learning_rate": 8.055112219451373e-06,
+      "loss": 0.335,
+      "step": 47920
+    },
+    {
+      "epoch": 11.952618453865338,
+      "grad_norm": 8.986421585083008,
+      "learning_rate": 8.052618453865338e-06,
+      "loss": 0.3582,
+      "step": 47930
+    },
+    {
+      "epoch": 11.955112219451372,
+      "grad_norm": 7.0583577156066895,
+      "learning_rate": 8.050124688279303e-06,
+      "loss": 0.3613,
+      "step": 47940
+    },
+    {
+      "epoch": 11.957605985037407,
+      "grad_norm": 6.976990222930908,
+      "learning_rate": 8.047630922693267e-06,
+      "loss": 0.3075,
+      "step": 47950
+    },
+    {
+      "epoch": 11.960099750623442,
+      "grad_norm": 7.128075122833252,
+      "learning_rate": 8.045137157107232e-06,
+      "loss": 0.4033,
+      "step": 47960
+    },
+    {
+      "epoch": 11.962593516209477,
+      "grad_norm": 7.31989049911499,
+      "learning_rate": 8.042643391521197e-06,
+      "loss": 0.3093,
+      "step": 47970
+    },
+    {
+      "epoch": 11.965087281795512,
+      "grad_norm": 4.972882270812988,
+      "learning_rate": 8.040149625935163e-06,
+      "loss": 0.299,
+      "step": 47980
+    },
+    {
+      "epoch": 11.967581047381547,
+      "grad_norm": 11.064085006713867,
+      "learning_rate": 8.037655860349128e-06,
+      "loss": 0.3248,
+      "step": 47990
+    },
+    {
+      "epoch": 11.970074812967582,
+      "grad_norm": 13.1663179397583,
+      "learning_rate": 8.035162094763093e-06,
+      "loss": 0.3242,
+      "step": 48000
+    },
+    {
+      "epoch": 11.972568578553616,
+      "grad_norm": 5.935441970825195,
+      "learning_rate": 8.032668329177059e-06,
+      "loss": 0.3544,
+      "step": 48010
+    },
+    {
+      "epoch": 11.975062344139651,
+      "grad_norm": 7.8656840324401855,
+      "learning_rate": 8.030174563591024e-06,
+      "loss": 0.2541,
+      "step": 48020
+    },
+    {
+      "epoch": 11.977556109725686,
+      "grad_norm": 11.03922176361084,
+      "learning_rate": 8.02768079800499e-06,
+      "loss": 0.3548,
+      "step": 48030
+    },
+    {
+      "epoch": 11.980049875311721,
+      "grad_norm": 8.151224136352539,
+      "learning_rate": 8.025187032418953e-06,
+      "loss": 0.3088,
+      "step": 48040
+    },
+    {
+      "epoch": 11.982543640897756,
+      "grad_norm": 9.86962890625,
+      "learning_rate": 8.022693266832918e-06,
+      "loss": 0.3831,
+      "step": 48050
+    },
+    {
+      "epoch": 11.98503740648379,
+      "grad_norm": 6.20054292678833,
+      "learning_rate": 8.020199501246884e-06,
+      "loss": 0.3202,
+      "step": 48060
+    },
+    {
+      "epoch": 11.987531172069826,
+      "grad_norm": 7.702700614929199,
+      "learning_rate": 8.017705735660849e-06,
+      "loss": 0.3719,
+      "step": 48070
+    },
+    {
+      "epoch": 11.99002493765586,
+      "grad_norm": 8.15425968170166,
+      "learning_rate": 8.015211970074812e-06,
+      "loss": 0.3887,
+      "step": 48080
+    },
+    {
+      "epoch": 11.992518703241895,
+      "grad_norm": 4.551555633544922,
+      "learning_rate": 8.012718204488778e-06,
+      "loss": 0.3482,
+      "step": 48090
+    },
+    {
+      "epoch": 11.99501246882793,
+      "grad_norm": 5.3351640701293945,
+      "learning_rate": 8.010224438902745e-06,
+      "loss": 0.3449,
+      "step": 48100
+    },
+    {
+      "epoch": 11.997506234413965,
+      "grad_norm": 9.141931533813477,
+      "learning_rate": 8.00773067331671e-06,
+      "loss": 0.3044,
+      "step": 48110
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 4.466921806335449,
+      "learning_rate": 8.005236907730674e-06,
+      "loss": 0.3335,
+      "step": 48120
+    },
+    {
+      "epoch": 12.0,
+      "eval_loss": 0.41446128487586975,
+      "eval_runtime": 60.0916,
+      "eval_samples_per_second": 16.691,
+      "eval_steps_per_second": 16.691,
+      "step": 48120
+    },
+    {
+      "epoch": 12.002493765586035,
+      "grad_norm": 7.059330463409424,
+      "learning_rate": 8.002743142144639e-06,
+      "loss": 0.3357,
+      "step": 48130
+    },
+    {
+      "epoch": 12.00498753117207,
+      "grad_norm": 7.653445720672607,
+      "learning_rate": 8.000249376558604e-06,
+      "loss": 0.4164,
+      "step": 48140
+    },
+    {
+      "epoch": 12.007481296758105,
+      "grad_norm": 6.720803737640381,
+      "learning_rate": 7.99775561097257e-06,
+      "loss": 0.3236,
+      "step": 48150
+    },
+    {
+      "epoch": 12.00997506234414,
+      "grad_norm": 5.9595746994018555,
+      "learning_rate": 7.995261845386533e-06,
+      "loss": 0.3662,
+      "step": 48160
+    },
+    {
+      "epoch": 12.012468827930174,
+      "grad_norm": 6.095608234405518,
+      "learning_rate": 7.992768079800499e-06,
+      "loss": 0.3405,
+      "step": 48170
+    },
+    {
+      "epoch": 12.01496259351621,
+      "grad_norm": 8.565361022949219,
+      "learning_rate": 7.990274314214464e-06,
+      "loss": 0.2904,
+      "step": 48180
+    },
+    {
+      "epoch": 12.017456359102244,
+      "grad_norm": 6.547665119171143,
+      "learning_rate": 7.987780548628429e-06,
+      "loss": 0.3475,
+      "step": 48190
+    },
+    {
+      "epoch": 12.019950124688279,
+      "grad_norm": 10.47833251953125,
+      "learning_rate": 7.985286783042394e-06,
+      "loss": 0.3089,
+      "step": 48200
+    },
+    {
+      "epoch": 12.022443890274314,
+      "grad_norm": 15.010177612304688,
+      "learning_rate": 7.98279301745636e-06,
+      "loss": 0.3161,
+      "step": 48210
+    },
+    {
+      "epoch": 12.024937655860349,
+      "grad_norm": 9.223240852355957,
+      "learning_rate": 7.980299251870325e-06,
+      "loss": 0.3958,
+      "step": 48220
+    },
+    {
+      "epoch": 12.027431421446384,
+      "grad_norm": 7.6100382804870605,
+      "learning_rate": 7.97780548628429e-06,
+      "loss": 0.3337,
+      "step": 48230
+    },
+    {
+      "epoch": 12.029925187032418,
+      "grad_norm": 8.66632080078125,
+      "learning_rate": 7.975311720698256e-06,
+      "loss": 0.3571,
+      "step": 48240
+    },
+    {
+      "epoch": 12.032418952618453,
+      "grad_norm": 8.049689292907715,
+      "learning_rate": 7.97281795511222e-06,
+      "loss": 0.398,
+      "step": 48250
+    },
+    {
+      "epoch": 12.034912718204488,
+      "grad_norm": 9.214011192321777,
+      "learning_rate": 7.970324189526185e-06,
+      "loss": 0.3354,
+      "step": 48260
+    },
+    {
+      "epoch": 12.037406483790523,
+      "grad_norm": 7.5215229988098145,
+      "learning_rate": 7.96783042394015e-06,
+      "loss": 0.3055,
+      "step": 48270
+    },
+    {
+      "epoch": 12.039900249376558,
+      "grad_norm": 5.8552327156066895,
+      "learning_rate": 7.965336658354115e-06,
+      "loss": 0.3736,
+      "step": 48280
+    },
+    {
+      "epoch": 12.042394014962593,
+      "grad_norm": 8.036945343017578,
+      "learning_rate": 7.96284289276808e-06,
+      "loss": 0.3435,
+      "step": 48290
+    },
+    {
+      "epoch": 12.044887780548628,
+      "grad_norm": 9.113265037536621,
+      "learning_rate": 7.960349127182046e-06,
+      "loss": 0.3947,
+      "step": 48300
+    },
+    {
+      "epoch": 12.047381546134662,
+      "grad_norm": 8.261163711547852,
+      "learning_rate": 7.957855361596011e-06,
+      "loss": 0.3772,
+      "step": 48310
+    },
+    {
+      "epoch": 12.049875311720697,
+      "grad_norm": 8.655675888061523,
+      "learning_rate": 7.955361596009976e-06,
+      "loss": 0.4154,
+      "step": 48320
+    },
+    {
+      "epoch": 12.052369077306734,
+      "grad_norm": 7.443731307983398,
+      "learning_rate": 7.95286783042394e-06,
+      "loss": 0.3444,
+      "step": 48330
+    },
+    {
+      "epoch": 12.054862842892769,
+      "grad_norm": 8.582282066345215,
+      "learning_rate": 7.950374064837905e-06,
+      "loss": 0.3394,
+      "step": 48340
+    },
+    {
+      "epoch": 12.057356608478804,
+      "grad_norm": 4.728538513183594,
+      "learning_rate": 7.94788029925187e-06,
+      "loss": 0.334,
+      "step": 48350
+    },
+    {
+      "epoch": 12.059850374064839,
+      "grad_norm": 7.743102073669434,
+      "learning_rate": 7.945386533665836e-06,
+      "loss": 0.2972,
+      "step": 48360
+    },
+    {
+      "epoch": 12.062344139650873,
+      "grad_norm": 5.460146427154541,
+      "learning_rate": 7.942892768079801e-06,
+      "loss": 0.3959,
+      "step": 48370
+    },
+    {
+      "epoch": 12.064837905236908,
+      "grad_norm": 9.469954490661621,
+      "learning_rate": 7.940399002493767e-06,
+      "loss": 0.3682,
+      "step": 48380
+    },
+    {
+      "epoch": 12.067331670822943,
+      "grad_norm": 8.519420623779297,
+      "learning_rate": 7.937905236907732e-06,
+      "loss": 0.3083,
+      "step": 48390
+    },
+    {
+      "epoch": 12.069825436408978,
+      "grad_norm": 5.299350261688232,
+      "learning_rate": 7.935411471321697e-06,
+      "loss": 0.3441,
+      "step": 48400
+    },
+    {
+      "epoch": 12.072319201995013,
+      "grad_norm": 7.708430290222168,
+      "learning_rate": 7.932917705735661e-06,
+      "loss": 0.3613,
+      "step": 48410
+    },
+    {
+      "epoch": 12.074812967581048,
+      "grad_norm": 8.682879447937012,
+      "learning_rate": 7.930423940149626e-06,
+      "loss": 0.3806,
+      "step": 48420
+    },
+    {
+      "epoch": 12.077306733167083,
+      "grad_norm": 7.677736282348633,
+      "learning_rate": 7.927930174563591e-06,
+      "loss": 0.3095,
+      "step": 48430
+    },
+    {
+      "epoch": 12.079800498753118,
+      "grad_norm": 6.203550338745117,
+      "learning_rate": 7.925436408977557e-06,
+      "loss": 0.3583,
+      "step": 48440
+    },
+    {
+      "epoch": 12.082294264339152,
+      "grad_norm": 9.091981887817383,
+      "learning_rate": 7.922942643391522e-06,
+      "loss": 0.3517,
+      "step": 48450
+    },
+    {
+      "epoch": 12.084788029925187,
+      "grad_norm": 6.73786735534668,
+      "learning_rate": 7.920448877805487e-06,
+      "loss": 0.3773,
+      "step": 48460
+    },
+    {
+      "epoch": 12.087281795511222,
+      "grad_norm": 10.410232543945312,
+      "learning_rate": 7.917955112219453e-06,
+      "loss": 0.3852,
+      "step": 48470
+    },
+    {
+      "epoch": 12.089775561097257,
+      "grad_norm": 6.987138271331787,
+      "learning_rate": 7.915461346633418e-06,
+      "loss": 0.3285,
+      "step": 48480
+    },
+    {
+      "epoch": 12.092269326683292,
+      "grad_norm": 6.348522663116455,
+      "learning_rate": 7.912967581047383e-06,
+      "loss": 0.3437,
+      "step": 48490
+    },
+    {
+      "epoch": 12.094763092269327,
+      "grad_norm": 5.560189247131348,
+      "learning_rate": 7.910473815461347e-06,
+      "loss": 0.326,
+      "step": 48500
+    },
+    {
+      "epoch": 12.097256857855362,
+      "grad_norm": 6.605633735656738,
+      "learning_rate": 7.907980049875312e-06,
+      "loss": 0.3387,
+      "step": 48510
+    },
+    {
+      "epoch": 12.099750623441397,
+      "grad_norm": 11.42223072052002,
+      "learning_rate": 7.905486284289278e-06,
+      "loss": 0.4234,
+      "step": 48520
+    },
+    {
+      "epoch": 12.102244389027431,
+      "grad_norm": 7.262038707733154,
+      "learning_rate": 7.902992518703243e-06,
+      "loss": 0.3345,
+      "step": 48530
+    },
+    {
+      "epoch": 12.104738154613466,
+      "grad_norm": 10.022313117980957,
+      "learning_rate": 7.900498753117207e-06,
+      "loss": 0.3145,
+      "step": 48540
+    },
+    {
+      "epoch": 12.107231920199501,
+      "grad_norm": 8.85555362701416,
+      "learning_rate": 7.898004987531172e-06,
+      "loss": 0.2839,
+      "step": 48550
+    },
+    {
+      "epoch": 12.109725685785536,
+      "grad_norm": 7.570572376251221,
+      "learning_rate": 7.895511221945137e-06,
+      "loss": 0.3078,
+      "step": 48560
+    },
+    {
+      "epoch": 12.11221945137157,
+      "grad_norm": 7.913335800170898,
+      "learning_rate": 7.893017456359104e-06,
+      "loss": 0.314,
+      "step": 48570
+    },
+    {
+      "epoch": 12.114713216957606,
+      "grad_norm": 7.83607292175293,
+      "learning_rate": 7.890523690773068e-06,
+      "loss": 0.2858,
+      "step": 48580
+    },
+    {
+      "epoch": 12.11720698254364,
+      "grad_norm": 9.3693208694458,
+      "learning_rate": 7.888279301745636e-06,
+      "loss": 0.2934,
+      "step": 48590
+    },
+    {
+      "epoch": 12.119700748129675,
+      "grad_norm": 6.200734615325928,
+      "learning_rate": 7.885785536159601e-06,
+      "loss": 0.409,
+      "step": 48600
+    },
+    {
+      "epoch": 12.12219451371571,
+      "grad_norm": 4.377243518829346,
+      "learning_rate": 7.883291770573567e-06,
+      "loss": 0.2854,
+      "step": 48610
+    },
+    {
+      "epoch": 12.124688279301745,
+      "grad_norm": 9.177326202392578,
+      "learning_rate": 7.880798004987532e-06,
+      "loss": 0.3049,
+      "step": 48620
+    },
+    {
+      "epoch": 12.12718204488778,
+      "grad_norm": 7.598722457885742,
+      "learning_rate": 7.878304239401496e-06,
+      "loss": 0.3693,
+      "step": 48630
+    },
+    {
+      "epoch": 12.129675810473815,
+      "grad_norm": 8.089449882507324,
+      "learning_rate": 7.875810473815461e-06,
+      "loss": 0.3015,
+      "step": 48640
+    },
+    {
+      "epoch": 12.13216957605985,
+      "grad_norm": 6.351145267486572,
+      "learning_rate": 7.873316708229428e-06,
+      "loss": 0.3215,
+      "step": 48650
+    },
+    {
+      "epoch": 12.134663341645885,
+      "grad_norm": 4.465890884399414,
+      "learning_rate": 7.870822942643393e-06,
+      "loss": 0.3268,
+      "step": 48660
+    },
+    {
+      "epoch": 12.13715710723192,
+      "grad_norm": 7.8744707107543945,
+      "learning_rate": 7.868329177057359e-06,
+      "loss": 0.3502,
+      "step": 48670
+    },
+    {
+      "epoch": 12.139650872817954,
+      "grad_norm": 6.566856384277344,
+      "learning_rate": 7.865835411471322e-06,
+      "loss": 0.2754,
+      "step": 48680
+    },
+    {
+      "epoch": 12.14214463840399,
+      "grad_norm": 8.633282661437988,
+      "learning_rate": 7.863341645885287e-06,
+      "loss": 0.3319,
+      "step": 48690
+    },
+    {
+      "epoch": 12.144638403990024,
+      "grad_norm": 5.682807922363281,
+      "learning_rate": 7.860847880299253e-06,
+      "loss": 0.3287,
+      "step": 48700
+    },
+    {
+      "epoch": 12.147132169576059,
+      "grad_norm": 6.86955451965332,
+      "learning_rate": 7.858354114713218e-06,
+      "loss": 0.3137,
+      "step": 48710
+    },
+    {
+      "epoch": 12.149625935162096,
+      "grad_norm": 8.783101081848145,
+      "learning_rate": 7.855860349127182e-06,
+      "loss": 0.4136,
+      "step": 48720
+    },
+    {
+      "epoch": 12.15211970074813,
+      "grad_norm": 3.7242095470428467,
+      "learning_rate": 7.853366583541147e-06,
+      "loss": 0.2878,
+      "step": 48730
+    },
+    {
+      "epoch": 12.154613466334165,
+      "grad_norm": 7.640246391296387,
+      "learning_rate": 7.850872817955112e-06,
+      "loss": 0.3841,
+      "step": 48740
+    },
+    {
+      "epoch": 12.1571072319202,
+      "grad_norm": 11.29294490814209,
+      "learning_rate": 7.848379052369078e-06,
+      "loss": 0.3812,
+      "step": 48750
+    },
+    {
+      "epoch": 12.159600997506235,
+      "grad_norm": 7.655768871307373,
+      "learning_rate": 7.845885286783043e-06,
+      "loss": 0.3404,
+      "step": 48760
+    },
+    {
+      "epoch": 12.16209476309227,
+      "grad_norm": 6.344391822814941,
+      "learning_rate": 7.843391521197008e-06,
+      "loss": 0.3118,
+      "step": 48770
+    },
+    {
+      "epoch": 12.164588528678305,
+      "grad_norm": 9.420690536499023,
+      "learning_rate": 7.840897755610974e-06,
+      "loss": 0.3424,
+      "step": 48780
+    },
+    {
+      "epoch": 12.16708229426434,
+      "grad_norm": 7.383553981781006,
+      "learning_rate": 7.838403990024939e-06,
+      "loss": 0.3298,
+      "step": 48790
+    },
+    {
+      "epoch": 12.169576059850375,
+      "grad_norm": 6.917516708374023,
+      "learning_rate": 7.835910224438902e-06,
+      "loss": 0.3557,
+      "step": 48800
+    },
+    {
+      "epoch": 12.17206982543641,
+      "grad_norm": 7.802567005157471,
+      "learning_rate": 7.833416458852868e-06,
+      "loss": 0.4077,
+      "step": 48810
+    },
+    {
+      "epoch": 12.174563591022444,
+      "grad_norm": 6.911114692687988,
+      "learning_rate": 7.830922693266833e-06,
+      "loss": 0.2852,
+      "step": 48820
+    },
+    {
+      "epoch": 12.17705735660848,
+      "grad_norm": 10.203409194946289,
+      "learning_rate": 7.828428927680798e-06,
+      "loss": 0.3605,
+      "step": 48830
+    },
+    {
+      "epoch": 12.179551122194514,
+      "grad_norm": 9.16174030303955,
+      "learning_rate": 7.825935162094764e-06,
+      "loss": 0.3409,
+      "step": 48840
+    },
+    {
+      "epoch": 12.182044887780549,
+      "grad_norm": 6.887585639953613,
+      "learning_rate": 7.823441396508729e-06,
+      "loss": 0.3659,
+      "step": 48850
+    },
+    {
+      "epoch": 12.184538653366584,
+      "grad_norm": 7.380236625671387,
+      "learning_rate": 7.820947630922694e-06,
+      "loss": 0.3259,
+      "step": 48860
+    },
+    {
+      "epoch": 12.187032418952619,
+      "grad_norm": 5.912197589874268,
+      "learning_rate": 7.81845386533666e-06,
+      "loss": 0.2935,
+      "step": 48870
+    },
+    {
+      "epoch": 12.189526184538654,
+      "grad_norm": 10.71356201171875,
+      "learning_rate": 7.815960099750623e-06,
+      "loss": 0.3154,
+      "step": 48880
+    },
+    {
+      "epoch": 12.192019950124688,
+      "grad_norm": 9.142561912536621,
+      "learning_rate": 7.813466334164589e-06,
+      "loss": 0.3491,
+      "step": 48890
+    },
+    {
+      "epoch": 12.194513715710723,
+      "grad_norm": 5.923270225524902,
+      "learning_rate": 7.810972568578554e-06,
+      "loss": 0.3462,
+      "step": 48900
+    },
+    {
+      "epoch": 12.197007481296758,
+      "grad_norm": 8.043718338012695,
+      "learning_rate": 7.80847880299252e-06,
+      "loss": 0.3948,
+      "step": 48910
+    },
+    {
+      "epoch": 12.199501246882793,
+      "grad_norm": 10.630451202392578,
+      "learning_rate": 7.805985037406484e-06,
+      "loss": 0.2927,
+      "step": 48920
+    },
+    {
+      "epoch": 12.201995012468828,
+      "grad_norm": 8.33590316772461,
+      "learning_rate": 7.80349127182045e-06,
+      "loss": 0.3302,
+      "step": 48930
+    },
+    {
+      "epoch": 12.204488778054863,
+      "grad_norm": 4.248763084411621,
+      "learning_rate": 7.800997506234415e-06,
+      "loss": 0.2981,
+      "step": 48940
+    },
+    {
+      "epoch": 12.206982543640898,
+      "grad_norm": 7.462469577789307,
+      "learning_rate": 7.79850374064838e-06,
+      "loss": 0.353,
+      "step": 48950
+    },
+    {
+      "epoch": 12.209476309226932,
+      "grad_norm": 7.565711975097656,
+      "learning_rate": 7.796009975062346e-06,
+      "loss": 0.3257,
+      "step": 48960
+    },
+    {
+      "epoch": 12.211970074812967,
+      "grad_norm": 5.535632133483887,
+      "learning_rate": 7.79351620947631e-06,
+      "loss": 0.318,
+      "step": 48970
+    },
+    {
+      "epoch": 12.214463840399002,
+      "grad_norm": 8.999574661254883,
+      "learning_rate": 7.791022443890275e-06,
+      "loss": 0.3884,
+      "step": 48980
+    },
+    {
+      "epoch": 12.216957605985037,
+      "grad_norm": 7.833513259887695,
+      "learning_rate": 7.78852867830424e-06,
+      "loss": 0.3175,
+      "step": 48990
+    },
+    {
+      "epoch": 12.219451371571072,
+      "grad_norm": 6.90658712387085,
+      "learning_rate": 7.786034912718205e-06,
+      "loss": 0.2902,
+      "step": 49000
+    },
+    {
+      "epoch": 12.221945137157107,
+      "grad_norm": 10.725184440612793,
+      "learning_rate": 7.78354114713217e-06,
+      "loss": 0.3352,
+      "step": 49010
+    },
+    {
+      "epoch": 12.224438902743142,
+      "grad_norm": 5.924967288970947,
+      "learning_rate": 7.781047381546136e-06,
+      "loss": 0.3747,
+      "step": 49020
+    },
+    {
+      "epoch": 12.226932668329177,
+      "grad_norm": 15.080745697021484,
+      "learning_rate": 7.778553615960101e-06,
+      "loss": 0.3795,
+      "step": 49030
+    },
+    {
+      "epoch": 12.229426433915211,
+      "grad_norm": 9.706949234008789,
+      "learning_rate": 7.776059850374066e-06,
+      "loss": 0.3756,
+      "step": 49040
+    },
+    {
+      "epoch": 12.231920199501246,
+      "grad_norm": 6.478403091430664,
+      "learning_rate": 7.77356608478803e-06,
+      "loss": 0.3344,
+      "step": 49050
+    },
+    {
+      "epoch": 12.234413965087281,
+      "grad_norm": 20.941816329956055,
+      "learning_rate": 7.771072319201995e-06,
+      "loss": 0.3319,
+      "step": 49060
+    },
+    {
+      "epoch": 12.236907730673316,
+      "grad_norm": 4.482845306396484,
+      "learning_rate": 7.76857855361596e-06,
+      "loss": 0.3203,
+      "step": 49070
+    },
+    {
+      "epoch": 12.239401496259351,
+      "grad_norm": 5.751954078674316,
+      "learning_rate": 7.766084788029926e-06,
+      "loss": 0.3554,
+      "step": 49080
+    },
+    {
+      "epoch": 12.241895261845386,
+      "grad_norm": 7.152318477630615,
+      "learning_rate": 7.76359102244389e-06,
+      "loss": 0.3193,
+      "step": 49090
+    },
+    {
+      "epoch": 12.24438902743142,
+      "grad_norm": 7.085709571838379,
+      "learning_rate": 7.761097256857855e-06,
+      "loss": 0.3458,
+      "step": 49100
+    },
+    {
+      "epoch": 12.246882793017456,
+      "grad_norm": 7.26195764541626,
+      "learning_rate": 7.75860349127182e-06,
+      "loss": 0.3644,
+      "step": 49110
+    },
+    {
+      "epoch": 12.24937655860349,
+      "grad_norm": 5.610538959503174,
+      "learning_rate": 7.756109725685787e-06,
+      "loss": 0.3654,
+      "step": 49120
+    },
+    {
+      "epoch": 12.251870324189527,
+      "grad_norm": 6.579119682312012,
+      "learning_rate": 7.753615960099751e-06,
+      "loss": 0.3116,
+      "step": 49130
+    },
+    {
+      "epoch": 12.254364089775562,
+      "grad_norm": 9.59024429321289,
+      "learning_rate": 7.751122194513716e-06,
+      "loss": 0.3193,
+      "step": 49140
+    },
+    {
+      "epoch": 12.256857855361597,
+      "grad_norm": 7.97966194152832,
+      "learning_rate": 7.748628428927682e-06,
+      "loss": 0.3293,
+      "step": 49150
+    },
+    {
+      "epoch": 12.259351620947632,
+      "grad_norm": 9.137805938720703,
+      "learning_rate": 7.746134663341647e-06,
+      "loss": 0.409,
+      "step": 49160
+    },
+    {
+      "epoch": 12.261845386533667,
+      "grad_norm": 9.003076553344727,
+      "learning_rate": 7.743640897755612e-06,
+      "loss": 0.3663,
+      "step": 49170
+    },
+    {
+      "epoch": 12.264339152119701,
+      "grad_norm": 9.627756118774414,
+      "learning_rate": 7.741147132169576e-06,
+      "loss": 0.3485,
+      "step": 49180
+    },
+    {
+      "epoch": 12.266832917705736,
+      "grad_norm": 5.549960613250732,
+      "learning_rate": 7.738653366583541e-06,
+      "loss": 0.3514,
+      "step": 49190
+    },
+    {
+      "epoch": 12.269326683291771,
+      "grad_norm": 6.5951080322265625,
+      "learning_rate": 7.736159600997506e-06,
+      "loss": 0.3182,
+      "step": 49200
+    },
+    {
+      "epoch": 12.271820448877806,
+      "grad_norm": 10.914071083068848,
+      "learning_rate": 7.733665835411472e-06,
+      "loss": 0.3302,
+      "step": 49210
+    },
+    {
+      "epoch": 12.27431421446384,
+      "grad_norm": 8.802125930786133,
+      "learning_rate": 7.731172069825437e-06,
+      "loss": 0.2761,
+      "step": 49220
+    },
+    {
+      "epoch": 12.276807980049876,
+      "grad_norm": 7.288644790649414,
+      "learning_rate": 7.728678304239402e-06,
+      "loss": 0.365,
+      "step": 49230
+    },
+    {
+      "epoch": 12.27930174563591,
+      "grad_norm": 6.142082691192627,
+      "learning_rate": 7.726184538653368e-06,
+      "loss": 0.2988,
+      "step": 49240
+    },
+    {
+      "epoch": 12.281795511221945,
+      "grad_norm": 8.087503433227539,
+      "learning_rate": 7.723690773067333e-06,
+      "loss": 0.3527,
+      "step": 49250
+    },
+    {
+      "epoch": 12.28428927680798,
+      "grad_norm": 7.012362480163574,
+      "learning_rate": 7.721197007481297e-06,
+      "loss": 0.3552,
+      "step": 49260
+    },
+    {
+      "epoch": 12.286783042394015,
+      "grad_norm": 5.985617160797119,
+      "learning_rate": 7.718703241895262e-06,
+      "loss": 0.2818,
+      "step": 49270
+    },
+    {
+      "epoch": 12.28927680798005,
+      "grad_norm": 6.581480026245117,
+      "learning_rate": 7.716209476309227e-06,
+      "loss": 0.3897,
+      "step": 49280
+    },
+    {
+      "epoch": 12.291770573566085,
+      "grad_norm": 6.56619930267334,
+      "learning_rate": 7.713965087281795e-06,
+      "loss": 0.4206,
+      "step": 49290
+    },
+    {
+      "epoch": 12.29426433915212,
+      "grad_norm": 12.421707153320312,
+      "learning_rate": 7.71147132169576e-06,
+      "loss": 0.3616,
+      "step": 49300
+    },
+    {
+      "epoch": 12.296758104738155,
+      "grad_norm": 5.2989726066589355,
+      "learning_rate": 7.708977556109726e-06,
+      "loss": 0.3153,
+      "step": 49310
+    },
+    {
+      "epoch": 12.29925187032419,
+      "grad_norm": 8.248827934265137,
+      "learning_rate": 7.706483790523691e-06,
+      "loss": 0.4233,
+      "step": 49320
+    },
+    {
+      "epoch": 12.301745635910224,
+      "grad_norm": 9.170528411865234,
+      "learning_rate": 7.703990024937657e-06,
+      "loss": 0.3728,
+      "step": 49330
+    },
+    {
+      "epoch": 12.30423940149626,
+      "grad_norm": 8.818942070007324,
+      "learning_rate": 7.701496259351622e-06,
+      "loss": 0.3719,
+      "step": 49340
+    },
+    {
+      "epoch": 12.306733167082294,
+      "grad_norm": 4.5717997550964355,
+      "learning_rate": 7.699002493765587e-06,
+      "loss": 0.3448,
+      "step": 49350
+    },
+    {
+      "epoch": 12.309226932668329,
+      "grad_norm": 7.510899543762207,
+      "learning_rate": 7.696508728179551e-06,
+      "loss": 0.3549,
+      "step": 49360
+    },
+    {
+      "epoch": 12.311720698254364,
+      "grad_norm": 9.351420402526855,
+      "learning_rate": 7.694014962593516e-06,
+      "loss": 0.2777,
+      "step": 49370
+    },
+    {
+      "epoch": 12.314214463840399,
+      "grad_norm": 5.524614334106445,
+      "learning_rate": 7.691521197007482e-06,
+      "loss": 0.3373,
+      "step": 49380
+    },
+    {
+      "epoch": 12.316708229426434,
+      "grad_norm": 6.964102268218994,
+      "learning_rate": 7.689027431421447e-06,
+      "loss": 0.3343,
+      "step": 49390
+    },
+    {
+      "epoch": 12.319201995012468,
+      "grad_norm": 5.369339466094971,
+      "learning_rate": 7.686533665835412e-06,
+      "loss": 0.2794,
+      "step": 49400
+    },
+    {
+      "epoch": 12.321695760598503,
+      "grad_norm": 8.168196678161621,
+      "learning_rate": 7.684039900249377e-06,
+      "loss": 0.3399,
+      "step": 49410
+    },
+    {
+      "epoch": 12.324189526184538,
+      "grad_norm": 8.461885452270508,
+      "learning_rate": 7.681546134663343e-06,
+      "loss": 0.363,
+      "step": 49420
+    },
+    {
+      "epoch": 12.326683291770573,
+      "grad_norm": 8.73070240020752,
+      "learning_rate": 7.679052369077308e-06,
+      "loss": 0.352,
+      "step": 49430
+    },
+    {
+      "epoch": 12.329177057356608,
+      "grad_norm": 5.573559284210205,
+      "learning_rate": 7.676558603491272e-06,
+      "loss": 0.2903,
+      "step": 49440
+    },
+    {
+      "epoch": 12.331670822942643,
+      "grad_norm": 8.736505508422852,
+      "learning_rate": 7.674064837905237e-06,
+      "loss": 0.3443,
+      "step": 49450
+    },
+    {
+      "epoch": 12.334164588528678,
+      "grad_norm": 8.074347496032715,
+      "learning_rate": 7.671571072319202e-06,
+      "loss": 0.3055,
+      "step": 49460
+    },
+    {
+      "epoch": 12.336658354114713,
+      "grad_norm": 10.68688678741455,
+      "learning_rate": 7.669077306733168e-06,
+      "loss": 0.3402,
+      "step": 49470
+    },
+    {
+      "epoch": 12.339152119700747,
+      "grad_norm": 6.418087005615234,
+      "learning_rate": 7.666583541147133e-06,
+      "loss": 0.3485,
+      "step": 49480
+    },
+    {
+      "epoch": 12.341645885286782,
+      "grad_norm": 6.3780903816223145,
+      "learning_rate": 7.664089775561098e-06,
+      "loss": 0.3058,
+      "step": 49490
+    },
+    {
+      "epoch": 12.344139650872817,
+      "grad_norm": 6.528293132781982,
+      "learning_rate": 7.661596009975064e-06,
+      "loss": 0.3228,
+      "step": 49500
+    },
+    {
+      "epoch": 12.346633416458852,
+      "grad_norm": 12.134942054748535,
+      "learning_rate": 7.659102244389029e-06,
+      "loss": 0.3832,
+      "step": 49510
+    },
+    {
+      "epoch": 12.349127182044889,
+      "grad_norm": 9.033289909362793,
+      "learning_rate": 7.656608478802992e-06,
+      "loss": 0.3629,
+      "step": 49520
+    },
+    {
+      "epoch": 12.351620947630924,
+      "grad_norm": 5.786545276641846,
+      "learning_rate": 7.654114713216958e-06,
+      "loss": 0.386,
+      "step": 49530
+    },
+    {
+      "epoch": 12.354114713216958,
+      "grad_norm": 5.447564601898193,
+      "learning_rate": 7.651620947630923e-06,
+      "loss": 0.2737,
+      "step": 49540
+    },
+    {
+      "epoch": 12.356608478802993,
+      "grad_norm": 6.9369916915893555,
+      "learning_rate": 7.649127182044888e-06,
+      "loss": 0.4029,
+      "step": 49550
+    },
+    {
+      "epoch": 12.359102244389028,
+      "grad_norm": 5.646476745605469,
+      "learning_rate": 7.646633416458854e-06,
+      "loss": 0.3803,
+      "step": 49560
+    },
+    {
+      "epoch": 12.361596009975063,
+      "grad_norm": 8.804864883422852,
+      "learning_rate": 7.644139650872819e-06,
+      "loss": 0.3256,
+      "step": 49570
+    },
+    {
+      "epoch": 12.364089775561098,
+      "grad_norm": 4.971907615661621,
+      "learning_rate": 7.641645885286784e-06,
+      "loss": 0.3518,
+      "step": 49580
+    },
+    {
+      "epoch": 12.366583541147133,
+      "grad_norm": 7.365206718444824,
+      "learning_rate": 7.63915211970075e-06,
+      "loss": 0.3382,
+      "step": 49590
+    },
+    {
+      "epoch": 12.369077306733168,
+      "grad_norm": 4.726197719573975,
+      "learning_rate": 7.636658354114715e-06,
+      "loss": 0.325,
+      "step": 49600
+    },
+    {
+      "epoch": 12.371571072319203,
+      "grad_norm": 8.982644081115723,
+      "learning_rate": 7.634164588528679e-06,
+      "loss": 0.3807,
+      "step": 49610
+    },
+    {
+      "epoch": 12.374064837905237,
+      "grad_norm": 6.123873233795166,
+      "learning_rate": 7.631670822942644e-06,
+      "loss": 0.3543,
+      "step": 49620
+    },
+    {
+      "epoch": 12.376558603491272,
+      "grad_norm": 7.114482879638672,
+      "learning_rate": 7.629177057356609e-06,
+      "loss": 0.3948,
+      "step": 49630
+    },
+    {
+      "epoch": 12.379052369077307,
+      "grad_norm": 6.665993690490723,
+      "learning_rate": 7.6266832917705745e-06,
+      "loss": 0.2877,
+      "step": 49640
+    },
+    {
+      "epoch": 12.381546134663342,
+      "grad_norm": 8.79172134399414,
+      "learning_rate": 7.624189526184539e-06,
+      "loss": 0.3899,
+      "step": 49650
+    },
+    {
+      "epoch": 12.384039900249377,
+      "grad_norm": 6.153012752532959,
+      "learning_rate": 7.621695760598504e-06,
+      "loss": 0.318,
+      "step": 49660
+    },
+    {
+      "epoch": 12.386533665835412,
+      "grad_norm": 8.169590950012207,
+      "learning_rate": 7.6192019950124696e-06,
+      "loss": 0.3068,
+      "step": 49670
+    },
+    {
+      "epoch": 12.389027431421447,
+      "grad_norm": 8.219849586486816,
+      "learning_rate": 7.616708229426435e-06,
+      "loss": 0.3231,
+      "step": 49680
+    },
+    {
+      "epoch": 12.391521197007481,
+      "grad_norm": 7.9375152587890625,
+      "learning_rate": 7.614214463840399e-06,
+      "loss": 0.2944,
+      "step": 49690
+    },
+    {
+      "epoch": 12.394014962593516,
+      "grad_norm": 6.999446392059326,
+      "learning_rate": 7.611720698254365e-06,
+      "loss": 0.3445,
+      "step": 49700
+    },
+    {
+      "epoch": 12.396508728179551,
+      "grad_norm": 5.5128278732299805,
+      "learning_rate": 7.60922693266833e-06,
+      "loss": 0.3404,
+      "step": 49710
+    },
+    {
+      "epoch": 12.399002493765586,
+      "grad_norm": 17.4468994140625,
+      "learning_rate": 7.606733167082295e-06,
+      "loss": 0.3414,
+      "step": 49720
+    },
+    {
+      "epoch": 12.401496259351621,
+      "grad_norm": 7.349917888641357,
+      "learning_rate": 7.60423940149626e-06,
+      "loss": 0.4121,
+      "step": 49730
+    },
+    {
+      "epoch": 12.403990024937656,
+      "grad_norm": 7.664532661437988,
+      "learning_rate": 7.601745635910225e-06,
+      "loss": 0.3575,
+      "step": 49740
+    },
+    {
+      "epoch": 12.40648379052369,
+      "grad_norm": 7.233582496643066,
+      "learning_rate": 7.59925187032419e-06,
+      "loss": 0.3506,
+      "step": 49750
+    },
+    {
+      "epoch": 12.408977556109726,
+      "grad_norm": 5.654517650604248,
+      "learning_rate": 7.596758104738156e-06,
+      "loss": 0.3195,
+      "step": 49760
+    },
+    {
+      "epoch": 12.41147132169576,
+      "grad_norm": 10.009135246276855,
+      "learning_rate": 7.59426433915212e-06,
+      "loss": 0.2798,
+      "step": 49770
+    },
+    {
+      "epoch": 12.413965087281795,
+      "grad_norm": 6.770684719085693,
+      "learning_rate": 7.5917705735660854e-06,
+      "loss": 0.3402,
+      "step": 49780
+    },
+    {
+      "epoch": 12.41645885286783,
+      "grad_norm": 7.9065093994140625,
+      "learning_rate": 7.589276807980051e-06,
+      "loss": 0.2968,
+      "step": 49790
+    },
+    {
+      "epoch": 12.418952618453865,
+      "grad_norm": 6.457505226135254,
+      "learning_rate": 7.586783042394016e-06,
+      "loss": 0.2829,
+      "step": 49800
+    },
+    {
+      "epoch": 12.4214463840399,
+      "grad_norm": 10.425678253173828,
+      "learning_rate": 7.584289276807981e-06,
+      "loss": 0.3081,
+      "step": 49810
+    },
+    {
+      "epoch": 12.423940149625935,
+      "grad_norm": 9.779804229736328,
+      "learning_rate": 7.581795511221946e-06,
+      "loss": 0.3627,
+      "step": 49820
+    },
+    {
+      "epoch": 12.42643391521197,
+      "grad_norm": 8.749654769897461,
+      "learning_rate": 7.579301745635911e-06,
+      "loss": 0.31,
+      "step": 49830
+    },
+    {
+      "epoch": 12.428927680798004,
+      "grad_norm": 5.994139194488525,
+      "learning_rate": 7.5768079800498764e-06,
+      "loss": 0.3538,
+      "step": 49840
+    },
+    {
+      "epoch": 12.43142144638404,
+      "grad_norm": 9.143582344055176,
+      "learning_rate": 7.574314214463842e-06,
+      "loss": 0.3849,
+      "step": 49850
+    },
+    {
+      "epoch": 12.433915211970074,
+      "grad_norm": 5.614033222198486,
+      "learning_rate": 7.571820448877805e-06,
+      "loss": 0.3452,
+      "step": 49860
+    },
+    {
+      "epoch": 12.436408977556109,
+      "grad_norm": 5.022126197814941,
+      "learning_rate": 7.5693266832917715e-06,
+      "loss": 0.3065,
+      "step": 49870
+    },
+    {
+      "epoch": 12.438902743142144,
+      "grad_norm": 8.072562217712402,
+      "learning_rate": 7.566832917705737e-06,
+      "loss": 0.3097,
+      "step": 49880
+    },
+    {
+      "epoch": 12.441396508728179,
+      "grad_norm": 11.745954513549805,
+      "learning_rate": 7.564339152119702e-06,
+      "loss": 0.3353,
+      "step": 49890
+    },
+    {
+      "epoch": 12.443890274314214,
+      "grad_norm": 7.331192970275879,
+      "learning_rate": 7.561845386533666e-06,
+      "loss": 0.3727,
+      "step": 49900
+    },
+    {
+      "epoch": 12.446384039900249,
+      "grad_norm": 8.64783000946045,
+      "learning_rate": 7.559351620947631e-06,
+      "loss": 0.3649,
+      "step": 49910
+    },
+    {
+      "epoch": 12.448877805486283,
+      "grad_norm": 8.688132286071777,
+      "learning_rate": 7.556857855361596e-06,
+      "loss": 0.3288,
+      "step": 49920
+    },
+    {
+      "epoch": 12.451371571072318,
+      "grad_norm": 9.612068176269531,
+      "learning_rate": 7.5543640897755625e-06,
+      "loss": 0.3738,
+      "step": 49930
+    },
+    {
+      "epoch": 12.453865336658355,
+      "grad_norm": 10.88000202178955,
+      "learning_rate": 7.551870324189526e-06,
+      "loss": 0.3455,
+      "step": 49940
+    },
+    {
+      "epoch": 12.45635910224439,
+      "grad_norm": 6.210239887237549,
+      "learning_rate": 7.5493765586034915e-06,
+      "loss": 0.3381,
+      "step": 49950
+    },
+    {
+      "epoch": 12.458852867830425,
+      "grad_norm": 6.904968738555908,
+      "learning_rate": 7.546882793017457e-06,
+      "loss": 0.323,
+      "step": 49960
+    },
+    {
+      "epoch": 12.46134663341646,
+      "grad_norm": 8.996318817138672,
+      "learning_rate": 7.544389027431422e-06,
+      "loss": 0.3161,
+      "step": 49970
+    },
+    {
+      "epoch": 12.463840399002494,
+      "grad_norm": 12.361374855041504,
+      "learning_rate": 7.5418952618453865e-06,
+      "loss": 0.4072,
+      "step": 49980
+    },
+    {
+      "epoch": 12.46633416458853,
+      "grad_norm": 6.888818740844727,
+      "learning_rate": 7.539401496259352e-06,
+      "loss": 0.4031,
+      "step": 49990
+    },
+    {
+      "epoch": 12.468827930174564,
+      "grad_norm": 7.364432334899902,
+      "learning_rate": 7.536907730673317e-06,
+      "loss": 0.3472,
+      "step": 50000
+    },
+    {
+      "epoch": 12.471321695760599,
+      "grad_norm": 7.185758590698242,
+      "learning_rate": 7.5344139650872825e-06,
+      "loss": 0.3492,
+      "step": 50010
+    },
+    {
+      "epoch": 12.473815461346634,
+      "grad_norm": 4.996051788330078,
+      "learning_rate": 7.531920199501247e-06,
+      "loss": 0.3285,
+      "step": 50020
+    },
+    {
+      "epoch": 12.476309226932669,
+      "grad_norm": 9.784656524658203,
+      "learning_rate": 7.529426433915212e-06,
+      "loss": 0.3681,
+      "step": 50030
+    },
+    {
+      "epoch": 12.478802992518704,
+      "grad_norm": 9.993110656738281,
+      "learning_rate": 7.5269326683291776e-06,
+      "loss": 0.4345,
+      "step": 50040
+    },
+    {
+      "epoch": 12.481296758104738,
+      "grad_norm": 7.204555511474609,
+      "learning_rate": 7.524438902743143e-06,
+      "loss": 0.336,
+      "step": 50050
+    },
+    {
+      "epoch": 12.483790523690773,
+      "grad_norm": 5.9999165534973145,
+      "learning_rate": 7.521945137157108e-06,
+      "loss": 0.4092,
+      "step": 50060
+    },
+    {
+      "epoch": 12.486284289276808,
+      "grad_norm": 4.7937445640563965,
+      "learning_rate": 7.519451371571073e-06,
+      "loss": 0.3532,
+      "step": 50070
+    },
+    {
+      "epoch": 12.488778054862843,
+      "grad_norm": 7.650313854217529,
+      "learning_rate": 7.516957605985038e-06,
+      "loss": 0.277,
+      "step": 50080
+    },
+    {
+      "epoch": 12.491271820448878,
+      "grad_norm": 9.389827728271484,
+      "learning_rate": 7.514463840399003e-06,
+      "loss": 0.3082,
+      "step": 50090
+    },
+    {
+      "epoch": 12.493765586034913,
+      "grad_norm": 7.2986578941345215,
+      "learning_rate": 7.5119700748129686e-06,
+      "loss": 0.3464,
+      "step": 50100
+    },
+    {
+      "epoch": 12.496259351620948,
+      "grad_norm": 9.90869140625,
+      "learning_rate": 7.509476309226933e-06,
+      "loss": 0.3514,
+      "step": 50110
+    },
+    {
+      "epoch": 12.498753117206983,
+      "grad_norm": 10.251230239868164,
+      "learning_rate": 7.506982543640898e-06,
+      "loss": 0.3623,
+      "step": 50120
+    },
+    {
+      "epoch": 12.501246882793017,
+      "grad_norm": 6.3379974365234375,
+      "learning_rate": 7.504488778054864e-06,
+      "loss": 0.3323,
+      "step": 50130
+    },
+    {
+      "epoch": 12.503740648379052,
+      "grad_norm": 7.917457580566406,
+      "learning_rate": 7.501995012468829e-06,
+      "loss": 0.3967,
+      "step": 50140
+    },
+    {
+      "epoch": 12.506234413965087,
+      "grad_norm": 10.125364303588867,
+      "learning_rate": 7.499501246882793e-06,
+      "loss": 0.3394,
+      "step": 50150
+    },
+    {
+      "epoch": 12.508728179551122,
+      "grad_norm": 6.0608978271484375,
+      "learning_rate": 7.497007481296759e-06,
+      "loss": 0.3227,
+      "step": 50160
+    },
+    {
+      "epoch": 12.511221945137157,
+      "grad_norm": 6.0084919929504395,
+      "learning_rate": 7.494513715710724e-06,
+      "loss": 0.3547,
+      "step": 50170
+    },
+    {
+      "epoch": 12.513715710723192,
+      "grad_norm": 7.461294174194336,
+      "learning_rate": 7.492019950124689e-06,
+      "loss": 0.3247,
+      "step": 50180
+    },
+    {
+      "epoch": 12.516209476309227,
+      "grad_norm": 8.043869018554688,
+      "learning_rate": 7.489526184538654e-06,
+      "loss": 0.3308,
+      "step": 50190
+    },
+    {
+      "epoch": 12.518703241895262,
+      "grad_norm": 8.986414909362793,
+      "learning_rate": 7.487032418952619e-06,
+      "loss": 0.3246,
+      "step": 50200
+    },
+    {
+      "epoch": 12.521197007481296,
+      "grad_norm": 9.545452117919922,
+      "learning_rate": 7.4845386533665844e-06,
+      "loss": 0.3351,
+      "step": 50210
+    },
+    {
+      "epoch": 12.523690773067331,
+      "grad_norm": 7.211226940155029,
+      "learning_rate": 7.48204488778055e-06,
+      "loss": 0.3332,
+      "step": 50220
+    },
+    {
+      "epoch": 12.526184538653366,
+      "grad_norm": 15.852411270141602,
+      "learning_rate": 7.479551122194514e-06,
+      "loss": 0.3832,
+      "step": 50230
+    },
+    {
+      "epoch": 12.528678304239401,
+      "grad_norm": 11.03563117980957,
+      "learning_rate": 7.4770573566084795e-06,
+      "loss": 0.3479,
+      "step": 50240
+    },
+    {
+      "epoch": 12.531172069825436,
+      "grad_norm": 7.793381214141846,
+      "learning_rate": 7.474563591022445e-06,
+      "loss": 0.3877,
+      "step": 50250
+    },
+    {
+      "epoch": 12.53366583541147,
+      "grad_norm": 6.964505195617676,
+      "learning_rate": 7.47206982543641e-06,
+      "loss": 0.3602,
+      "step": 50260
+    },
+    {
+      "epoch": 12.536159600997506,
+      "grad_norm": 9.570211410522461,
+      "learning_rate": 7.469576059850374e-06,
+      "loss": 0.4639,
+      "step": 50270
+    },
+    {
+      "epoch": 12.53865336658354,
+      "grad_norm": 6.632506370544434,
+      "learning_rate": 7.46708229426434e-06,
+      "loss": 0.3231,
+      "step": 50280
+    },
+    {
+      "epoch": 12.541147132169575,
+      "grad_norm": 12.00216293334961,
+      "learning_rate": 7.464588528678305e-06,
+      "loss": 0.3751,
+      "step": 50290
+    },
+    {
+      "epoch": 12.54364089775561,
+      "grad_norm": 9.346277236938477,
+      "learning_rate": 7.4620947630922705e-06,
+      "loss": 0.384,
+      "step": 50300
+    },
+    {
+      "epoch": 12.546134663341645,
+      "grad_norm": 11.572287559509277,
+      "learning_rate": 7.459600997506236e-06,
+      "loss": 0.3448,
+      "step": 50310
+    },
+    {
+      "epoch": 12.548628428927682,
+      "grad_norm": 5.089780807495117,
+      "learning_rate": 7.4571072319201995e-06,
+      "loss": 0.3452,
+      "step": 50320
+    },
+    {
+      "epoch": 12.551122194513717,
+      "grad_norm": 5.670629024505615,
+      "learning_rate": 7.454613466334165e-06,
+      "loss": 0.2836,
+      "step": 50330
+    },
+    {
+      "epoch": 12.553615960099751,
+      "grad_norm": 8.412741661071777,
+      "learning_rate": 7.452119700748131e-06,
+      "loss": 0.3167,
+      "step": 50340
+    },
+    {
+      "epoch": 12.556109725685786,
+      "grad_norm": 6.057131767272949,
+      "learning_rate": 7.449625935162096e-06,
+      "loss": 0.3257,
+      "step": 50350
+    },
+    {
+      "epoch": 12.558603491271821,
+      "grad_norm": 13.145732879638672,
+      "learning_rate": 7.44713216957606e-06,
+      "loss": 0.3762,
+      "step": 50360
+    },
+    {
+      "epoch": 12.561097256857856,
+      "grad_norm": 6.84805965423584,
+      "learning_rate": 7.444638403990025e-06,
+      "loss": 0.3416,
+      "step": 50370
+    },
+    {
+      "epoch": 12.563591022443891,
+      "grad_norm": 9.685379028320312,
+      "learning_rate": 7.4421446384039905e-06,
+      "loss": 0.4172,
+      "step": 50380
+    },
+    {
+      "epoch": 12.566084788029926,
+      "grad_norm": 10.725334167480469,
+      "learning_rate": 7.439650872817956e-06,
+      "loss": 0.3302,
+      "step": 50390
+    },
+    {
+      "epoch": 12.56857855361596,
+      "grad_norm": 9.620137214660645,
+      "learning_rate": 7.43715710723192e-06,
+      "loss": 0.3222,
+      "step": 50400
+    },
+    {
+      "epoch": 12.571072319201996,
+      "grad_norm": 6.089214324951172,
+      "learning_rate": 7.4346633416458855e-06,
+      "loss": 0.2796,
+      "step": 50410
+    },
+    {
+      "epoch": 12.57356608478803,
+      "grad_norm": 9.723833084106445,
+      "learning_rate": 7.432169576059851e-06,
+      "loss": 0.3645,
+      "step": 50420
+    },
+    {
+      "epoch": 12.576059850374065,
+      "grad_norm": 7.900272369384766,
+      "learning_rate": 7.429675810473816e-06,
+      "loss": 0.3341,
+      "step": 50430
+    },
+    {
+      "epoch": 12.5785536159601,
+      "grad_norm": 5.3839521408081055,
+      "learning_rate": 7.427182044887781e-06,
+      "loss": 0.3581,
+      "step": 50440
+    },
+    {
+      "epoch": 12.581047381546135,
+      "grad_norm": 7.9673848152160645,
+      "learning_rate": 7.424688279301746e-06,
+      "loss": 0.3934,
+      "step": 50450
+    },
+    {
+      "epoch": 12.58354114713217,
+      "grad_norm": 7.145327091217041,
+      "learning_rate": 7.422194513715711e-06,
+      "loss": 0.3564,
+      "step": 50460
+    },
+    {
+      "epoch": 12.586034912718205,
+      "grad_norm": 8.317575454711914,
+      "learning_rate": 7.4197007481296766e-06,
+      "loss": 0.3919,
+      "step": 50470
+    },
+    {
+      "epoch": 12.58852867830424,
+      "grad_norm": 8.809738159179688,
+      "learning_rate": 7.417206982543641e-06,
+      "loss": 0.3832,
+      "step": 50480
+    },
+    {
+      "epoch": 12.591022443890274,
+      "grad_norm": 7.227417945861816,
+      "learning_rate": 7.414713216957606e-06,
+      "loss": 0.3157,
+      "step": 50490
+    },
+    {
+      "epoch": 12.59351620947631,
+      "grad_norm": 7.549991130828857,
+      "learning_rate": 7.412219451371572e-06,
+      "loss": 0.3329,
+      "step": 50500
+    },
+    {
+      "epoch": 12.596009975062344,
+      "grad_norm": 15.35284423828125,
+      "learning_rate": 7.409725685785537e-06,
+      "loss": 0.3495,
+      "step": 50510
+    },
+    {
+      "epoch": 12.598503740648379,
+      "grad_norm": 4.876797199249268,
+      "learning_rate": 7.407231920199501e-06,
+      "loss": 0.3429,
+      "step": 50520
+    },
+    {
+      "epoch": 12.600997506234414,
+      "grad_norm": 6.604551315307617,
+      "learning_rate": 7.404738154613467e-06,
+      "loss": 0.3698,
+      "step": 50530
+    },
+    {
+      "epoch": 12.603491271820449,
+      "grad_norm": 7.2110443115234375,
+      "learning_rate": 7.402244389027432e-06,
+      "loss": 0.331,
+      "step": 50540
+    },
+    {
+      "epoch": 12.605985037406484,
+      "grad_norm": 7.359330654144287,
+      "learning_rate": 7.399750623441397e-06,
+      "loss": 0.3591,
+      "step": 50550
+    },
+    {
+      "epoch": 12.608478802992519,
+      "grad_norm": 5.08826208114624,
+      "learning_rate": 7.397256857855363e-06,
+      "loss": 0.2935,
+      "step": 50560
+    },
+    {
+      "epoch": 12.610972568578553,
+      "grad_norm": 4.144226551055908,
+      "learning_rate": 7.394763092269327e-06,
+      "loss": 0.3926,
+      "step": 50570
+    },
+    {
+      "epoch": 12.613466334164588,
+      "grad_norm": 5.52124547958374,
+      "learning_rate": 7.392269326683292e-06,
+      "loss": 0.4041,
+      "step": 50580
+    },
+    {
+      "epoch": 12.615960099750623,
+      "grad_norm": 13.029586791992188,
+      "learning_rate": 7.389775561097258e-06,
+      "loss": 0.3669,
+      "step": 50590
+    },
+    {
+      "epoch": 12.618453865336658,
+      "grad_norm": 7.825071811676025,
+      "learning_rate": 7.387281795511223e-06,
+      "loss": 0.3541,
+      "step": 50600
+    },
+    {
+      "epoch": 12.620947630922693,
+      "grad_norm": 7.498974323272705,
+      "learning_rate": 7.3847880299251875e-06,
+      "loss": 0.2758,
+      "step": 50610
+    },
+    {
+      "epoch": 12.623441396508728,
+      "grad_norm": 10.242654800415039,
+      "learning_rate": 7.382294264339153e-06,
+      "loss": 0.302,
+      "step": 50620
+    },
+    {
+      "epoch": 12.625935162094763,
+      "grad_norm": 6.927699089050293,
+      "learning_rate": 7.379800498753118e-06,
+      "loss": 0.3352,
+      "step": 50630
+    },
+    {
+      "epoch": 12.628428927680797,
+      "grad_norm": 4.644331932067871,
+      "learning_rate": 7.3773067331670834e-06,
+      "loss": 0.3491,
+      "step": 50640
+    },
+    {
+      "epoch": 12.630922693266832,
+      "grad_norm": 5.701728343963623,
+      "learning_rate": 7.374812967581048e-06,
+      "loss": 0.2982,
+      "step": 50650
+    },
+    {
+      "epoch": 12.633416458852867,
+      "grad_norm": 10.834186553955078,
+      "learning_rate": 7.372319201995013e-06,
+      "loss": 0.3135,
+      "step": 50660
+    },
+    {
+      "epoch": 12.635910224438902,
+      "grad_norm": 9.567243576049805,
+      "learning_rate": 7.3698254364089785e-06,
+      "loss": 0.3627,
+      "step": 50670
+    },
+    {
+      "epoch": 12.638403990024937,
+      "grad_norm": 5.764795780181885,
+      "learning_rate": 7.367331670822944e-06,
+      "loss": 0.3256,
+      "step": 50680
+    },
+    {
+      "epoch": 12.640897755610972,
+      "grad_norm": 6.912472724914551,
+      "learning_rate": 7.364837905236908e-06,
+      "loss": 0.3154,
+      "step": 50690
+    },
+    {
+      "epoch": 12.643391521197007,
+      "grad_norm": 4.959378719329834,
+      "learning_rate": 7.362344139650874e-06,
+      "loss": 0.3111,
+      "step": 50700
+    },
+    {
+      "epoch": 12.645885286783042,
+      "grad_norm": 8.026573181152344,
+      "learning_rate": 7.359850374064839e-06,
+      "loss": 0.3211,
+      "step": 50710
+    },
+    {
+      "epoch": 12.648379052369076,
+      "grad_norm": 9.039690017700195,
+      "learning_rate": 7.357356608478804e-06,
+      "loss": 0.2915,
+      "step": 50720
+    },
+    {
+      "epoch": 12.650872817955111,
+      "grad_norm": 6.101404666900635,
+      "learning_rate": 7.354862842892768e-06,
+      "loss": 0.3123,
+      "step": 50730
+    },
+    {
+      "epoch": 12.653366583541148,
+      "grad_norm": 8.664443969726562,
+      "learning_rate": 7.352369077306733e-06,
+      "loss": 0.3825,
+      "step": 50740
+    },
+    {
+      "epoch": 12.655860349127183,
+      "grad_norm": 9.475874900817871,
+      "learning_rate": 7.349875311720699e-06,
+      "loss": 0.3463,
+      "step": 50750
+    },
+    {
+      "epoch": 12.658354114713218,
+      "grad_norm": 6.189550399780273,
+      "learning_rate": 7.347381546134665e-06,
+      "loss": 0.3556,
+      "step": 50760
+    },
+    {
+      "epoch": 12.660847880299253,
+      "grad_norm": 6.720332622528076,
+      "learning_rate": 7.344887780548628e-06,
+      "loss": 0.2882,
+      "step": 50770
+    },
+    {
+      "epoch": 12.663341645885287,
+      "grad_norm": 6.461765766143799,
+      "learning_rate": 7.3423940149625935e-06,
+      "loss": 0.3059,
+      "step": 50780
+    },
+    {
+      "epoch": 12.665835411471322,
+      "grad_norm": 5.625673294067383,
+      "learning_rate": 7.339900249376559e-06,
+      "loss": 0.3903,
+      "step": 50790
+    },
+    {
+      "epoch": 12.668329177057357,
+      "grad_norm": 4.355860233306885,
+      "learning_rate": 7.337406483790524e-06,
+      "loss": 0.332,
+      "step": 50800
+    },
+    {
+      "epoch": 12.670822942643392,
+      "grad_norm": 8.476668357849121,
+      "learning_rate": 7.33491271820449e-06,
+      "loss": 0.354,
+      "step": 50810
+    },
+    {
+      "epoch": 12.673316708229427,
+      "grad_norm": 8.867452621459961,
+      "learning_rate": 7.332418952618454e-06,
+      "loss": 0.2912,
+      "step": 50820
+    },
+    {
+      "epoch": 12.675810473815462,
+      "grad_norm": 6.675667762756348,
+      "learning_rate": 7.329925187032419e-06,
+      "loss": 0.369,
+      "step": 50830
+    },
+    {
+      "epoch": 12.678304239401497,
+      "grad_norm": 6.0077972412109375,
+      "learning_rate": 7.327680798004988e-06,
+      "loss": 0.3377,
+      "step": 50840
+    },
+    {
+      "epoch": 12.680798004987532,
+      "grad_norm": 10.627485275268555,
+      "learning_rate": 7.325187032418954e-06,
+      "loss": 0.3032,
+      "step": 50850
+    },
+    {
+      "epoch": 12.683291770573566,
+      "grad_norm": 6.776688098907471,
+      "learning_rate": 7.322693266832919e-06,
+      "loss": 0.3235,
+      "step": 50860
+    },
+    {
+      "epoch": 12.685785536159601,
+      "grad_norm": 8.06655216217041,
+      "learning_rate": 7.320199501246883e-06,
+      "loss": 0.3717,
+      "step": 50870
+    },
+    {
+      "epoch": 12.688279301745636,
+      "grad_norm": 5.898749351501465,
+      "learning_rate": 7.317705735660848e-06,
+      "loss": 0.3136,
+      "step": 50880
+    },
+    {
+      "epoch": 12.690773067331671,
+      "grad_norm": 5.939793109893799,
+      "learning_rate": 7.315211970074814e-06,
+      "loss": 0.2735,
+      "step": 50890
+    },
+    {
+      "epoch": 12.693266832917706,
+      "grad_norm": 9.688426971435547,
+      "learning_rate": 7.312718204488779e-06,
+      "loss": 0.3447,
+      "step": 50900
+    },
+    {
+      "epoch": 12.69576059850374,
+      "grad_norm": 8.256994247436523,
+      "learning_rate": 7.310224438902743e-06,
+      "loss": 0.3271,
+      "step": 50910
+    },
+    {
+      "epoch": 12.698254364089776,
+      "grad_norm": 8.417771339416504,
+      "learning_rate": 7.307730673316708e-06,
+      "loss": 0.4127,
+      "step": 50920
+    },
+    {
+      "epoch": 12.70074812967581,
+      "grad_norm": 7.872901916503906,
+      "learning_rate": 7.305236907730674e-06,
+      "loss": 0.3853,
+      "step": 50930
+    },
+    {
+      "epoch": 12.703241895261845,
+      "grad_norm": 7.196390151977539,
+      "learning_rate": 7.302743142144639e-06,
+      "loss": 0.3114,
+      "step": 50940
+    },
+    {
+      "epoch": 12.70573566084788,
+      "grad_norm": 7.429532051086426,
+      "learning_rate": 7.300249376558603e-06,
+      "loss": 0.3283,
+      "step": 50950
+    },
+    {
+      "epoch": 12.708229426433915,
+      "grad_norm": 11.154653549194336,
+      "learning_rate": 7.297755610972569e-06,
+      "loss": 0.3902,
+      "step": 50960
+    },
+    {
+      "epoch": 12.71072319201995,
+      "grad_norm": 4.860759258270264,
+      "learning_rate": 7.295261845386534e-06,
+      "loss": 0.3529,
+      "step": 50970
+    },
+    {
+      "epoch": 12.713216957605985,
+      "grad_norm": 4.447295665740967,
+      "learning_rate": 7.292768079800499e-06,
+      "loss": 0.3266,
+      "step": 50980
+    },
+    {
+      "epoch": 12.71571072319202,
+      "grad_norm": 8.30553913116455,
+      "learning_rate": 7.290274314214465e-06,
+      "loss": 0.3435,
+      "step": 50990
+    },
+    {
+      "epoch": 12.718204488778055,
+      "grad_norm": 8.856745719909668,
+      "learning_rate": 7.287780548628429e-06,
+      "loss": 0.3163,
+      "step": 51000
+    },
+    {
+      "epoch": 12.72069825436409,
+      "grad_norm": 8.923396110534668,
+      "learning_rate": 7.285286783042394e-06,
+      "loss": 0.3417,
+      "step": 51010
+    },
+    {
+      "epoch": 12.723192019950124,
+      "grad_norm": 9.019140243530273,
+      "learning_rate": 7.28279301745636e-06,
+      "loss": 0.3917,
+      "step": 51020
+    },
+    {
+      "epoch": 12.72568578553616,
+      "grad_norm": 8.857991218566895,
+      "learning_rate": 7.280299251870325e-06,
+      "loss": 0.3368,
+      "step": 51030
+    },
+    {
+      "epoch": 12.728179551122194,
+      "grad_norm": 7.459349632263184,
+      "learning_rate": 7.2778054862842895e-06,
+      "loss": 0.2827,
+      "step": 51040
+    },
+    {
+      "epoch": 12.730673316708229,
+      "grad_norm": 6.536787033081055,
+      "learning_rate": 7.275311720698255e-06,
+      "loss": 0.3304,
+      "step": 51050
+    },
+    {
+      "epoch": 12.733167082294264,
+      "grad_norm": 6.800075531005859,
+      "learning_rate": 7.27281795511222e-06,
+      "loss": 0.4169,
+      "step": 51060
+    },
+    {
+      "epoch": 12.735660847880299,
+      "grad_norm": 9.216288566589355,
+      "learning_rate": 7.270324189526185e-06,
+      "loss": 0.3732,
+      "step": 51070
+    },
+    {
+      "epoch": 12.738154613466333,
+      "grad_norm": 9.165617942810059,
+      "learning_rate": 7.26783042394015e-06,
+      "loss": 0.31,
+      "step": 51080
+    },
+    {
+      "epoch": 12.740648379052368,
+      "grad_norm": 7.988641262054443,
+      "learning_rate": 7.265336658354115e-06,
+      "loss": 0.3234,
+      "step": 51090
+    },
+    {
+      "epoch": 12.743142144638403,
+      "grad_norm": 6.585142135620117,
+      "learning_rate": 7.2628428927680805e-06,
+      "loss": 0.3289,
+      "step": 51100
+    },
+    {
+      "epoch": 12.745635910224438,
+      "grad_norm": 10.255621910095215,
+      "learning_rate": 7.260349127182046e-06,
+      "loss": 0.3875,
+      "step": 51110
+    },
+    {
+      "epoch": 12.748129675810475,
+      "grad_norm": 11.782647132873535,
+      "learning_rate": 7.25785536159601e-06,
+      "loss": 0.417,
+      "step": 51120
+    },
+    {
+      "epoch": 12.75062344139651,
+      "grad_norm": 9.408047676086426,
+      "learning_rate": 7.2553615960099756e-06,
+      "loss": 0.2934,
+      "step": 51130
+    },
+    {
+      "epoch": 12.753117206982544,
+      "grad_norm": 8.07828426361084,
+      "learning_rate": 7.252867830423941e-06,
+      "loss": 0.3938,
+      "step": 51140
+    },
+    {
+      "epoch": 12.75561097256858,
+      "grad_norm": 4.914299011230469,
+      "learning_rate": 7.250374064837906e-06,
+      "loss": 0.3387,
+      "step": 51150
+    },
+    {
+      "epoch": 12.758104738154614,
+      "grad_norm": 6.668241024017334,
+      "learning_rate": 7.247880299251871e-06,
+      "loss": 0.3265,
+      "step": 51160
+    },
+    {
+      "epoch": 12.760598503740649,
+      "grad_norm": 9.32829761505127,
+      "learning_rate": 7.245386533665836e-06,
+      "loss": 0.3249,
+      "step": 51170
+    },
+    {
+      "epoch": 12.763092269326684,
+      "grad_norm": 10.00203800201416,
+      "learning_rate": 7.242892768079801e-06,
+      "loss": 0.4209,
+      "step": 51180
+    },
+    {
+      "epoch": 12.765586034912719,
+      "grad_norm": 8.861324310302734,
+      "learning_rate": 7.2403990024937666e-06,
+      "loss": 0.3101,
+      "step": 51190
+    },
+    {
+      "epoch": 12.768079800498754,
+      "grad_norm": 8.35321044921875,
+      "learning_rate": 7.237905236907731e-06,
+      "loss": 0.3043,
+      "step": 51200
+    },
+    {
+      "epoch": 12.770573566084789,
+      "grad_norm": 5.797079086303711,
+      "learning_rate": 7.235411471321696e-06,
+      "loss": 0.3619,
+      "step": 51210
+    },
+    {
+      "epoch": 12.773067331670823,
+      "grad_norm": 8.274916648864746,
+      "learning_rate": 7.232917705735662e-06,
+      "loss": 0.2735,
+      "step": 51220
+    },
+    {
+      "epoch": 12.775561097256858,
+      "grad_norm": 6.340693473815918,
+      "learning_rate": 7.230423940149627e-06,
+      "loss": 0.3211,
+      "step": 51230
+    },
+    {
+      "epoch": 12.778054862842893,
+      "grad_norm": 6.9829792976379395,
+      "learning_rate": 7.227930174563592e-06,
+      "loss": 0.2852,
+      "step": 51240
+    },
+    {
+      "epoch": 12.780548628428928,
+      "grad_norm": 11.263195037841797,
+      "learning_rate": 7.225436408977557e-06,
+      "loss": 0.3557,
+      "step": 51250
+    },
+    {
+      "epoch": 12.783042394014963,
+      "grad_norm": 8.167370796203613,
+      "learning_rate": 7.222942643391522e-06,
+      "loss": 0.3264,
+      "step": 51260
+    },
+    {
+      "epoch": 12.785536159600998,
+      "grad_norm": 6.718631744384766,
+      "learning_rate": 7.220448877805487e-06,
+      "loss": 0.2878,
+      "step": 51270
+    },
+    {
+      "epoch": 12.788029925187033,
+      "grad_norm": 7.236240386962891,
+      "learning_rate": 7.217955112219453e-06,
+      "loss": 0.3249,
+      "step": 51280
+    },
+    {
+      "epoch": 12.790523690773068,
+      "grad_norm": 10.440741539001465,
+      "learning_rate": 7.215461346633416e-06,
+      "loss": 0.3534,
+      "step": 51290
+    },
+    {
+      "epoch": 12.793017456359102,
+      "grad_norm": 5.670083522796631,
+      "learning_rate": 7.2129675810473824e-06,
+      "loss": 0.3517,
+      "step": 51300
+    },
+    {
+      "epoch": 12.795511221945137,
+      "grad_norm": 15.356858253479004,
+      "learning_rate": 7.210473815461348e-06,
+      "loss": 0.4574,
+      "step": 51310
+    },
+    {
+      "epoch": 12.798004987531172,
+      "grad_norm": 9.100034713745117,
+      "learning_rate": 7.207980049875313e-06,
+      "loss": 0.3513,
+      "step": 51320
+    },
+    {
+      "epoch": 12.800498753117207,
+      "grad_norm": 5.6677937507629395,
+      "learning_rate": 7.205486284289277e-06,
+      "loss": 0.3663,
+      "step": 51330
+    },
+    {
+      "epoch": 12.802992518703242,
+      "grad_norm": 8.217962265014648,
+      "learning_rate": 7.202992518703242e-06,
+      "loss": 0.2736,
+      "step": 51340
+    },
+    {
+      "epoch": 12.805486284289277,
+      "grad_norm": 5.241400241851807,
+      "learning_rate": 7.200498753117207e-06,
+      "loss": 0.3087,
+      "step": 51350
+    },
+    {
+      "epoch": 12.807980049875312,
+      "grad_norm": 10.871718406677246,
+      "learning_rate": 7.1980049875311734e-06,
+      "loss": 0.3461,
+      "step": 51360
+    },
+    {
+      "epoch": 12.810473815461346,
+      "grad_norm": 6.515952110290527,
+      "learning_rate": 7.195511221945137e-06,
+      "loss": 0.3521,
+      "step": 51370
+    },
+    {
+      "epoch": 12.812967581047381,
+      "grad_norm": 9.03691577911377,
+      "learning_rate": 7.193017456359102e-06,
+      "loss": 0.3166,
+      "step": 51380
+    },
+    {
+      "epoch": 12.815461346633416,
+      "grad_norm": 8.544856071472168,
+      "learning_rate": 7.190523690773068e-06,
+      "loss": 0.3556,
+      "step": 51390
+    },
+    {
+      "epoch": 12.817955112219451,
+      "grad_norm": 6.503901481628418,
+      "learning_rate": 7.188029925187033e-06,
+      "loss": 0.3814,
+      "step": 51400
+    },
+    {
+      "epoch": 12.820448877805486,
+      "grad_norm": 6.1890387535095215,
+      "learning_rate": 7.1855361596009975e-06,
+      "loss": 0.3247,
+      "step": 51410
+    },
+    {
+      "epoch": 12.82294264339152,
+      "grad_norm": 5.103339672088623,
+      "learning_rate": 7.183042394014963e-06,
+      "loss": 0.4179,
+      "step": 51420
+    },
+    {
+      "epoch": 12.825436408977556,
+      "grad_norm": 9.049005508422852,
+      "learning_rate": 7.180548628428928e-06,
+      "loss": 0.3371,
+      "step": 51430
+    },
+    {
+      "epoch": 12.82793017456359,
+      "grad_norm": 8.748699188232422,
+      "learning_rate": 7.178054862842893e-06,
+      "loss": 0.3176,
+      "step": 51440
+    },
+    {
+      "epoch": 12.830423940149625,
+      "grad_norm": 8.368945121765137,
+      "learning_rate": 7.175561097256858e-06,
+      "loss": 0.3286,
+      "step": 51450
+    },
+    {
+      "epoch": 12.83291770573566,
+      "grad_norm": 7.695786952972412,
+      "learning_rate": 7.173067331670823e-06,
+      "loss": 0.3397,
+      "step": 51460
+    },
+    {
+      "epoch": 12.835411471321695,
+      "grad_norm": 8.735140800476074,
+      "learning_rate": 7.1705735660847885e-06,
+      "loss": 0.3353,
+      "step": 51470
+    },
+    {
+      "epoch": 12.83790523690773,
+      "grad_norm": 4.386609077453613,
+      "learning_rate": 7.168079800498754e-06,
+      "loss": 0.3446,
+      "step": 51480
+    },
+    {
+      "epoch": 12.840399002493765,
+      "grad_norm": 6.733331680297852,
+      "learning_rate": 7.165586034912719e-06,
+      "loss": 0.3685,
+      "step": 51490
+    },
+    {
+      "epoch": 12.8428927680798,
+      "grad_norm": 7.894585132598877,
+      "learning_rate": 7.1630922693266835e-06,
+      "loss": 0.332,
+      "step": 51500
+    },
+    {
+      "epoch": 12.845386533665835,
+      "grad_norm": 9.471677780151367,
+      "learning_rate": 7.160598503740649e-06,
+      "loss": 0.2614,
+      "step": 51510
+    },
+    {
+      "epoch": 12.84788029925187,
+      "grad_norm": 11.194404602050781,
+      "learning_rate": 7.158104738154614e-06,
+      "loss": 0.3671,
+      "step": 51520
+    },
+    {
+      "epoch": 12.850374064837904,
+      "grad_norm": 9.611897468566895,
+      "learning_rate": 7.1556109725685795e-06,
+      "loss": 0.3233,
+      "step": 51530
+    },
+    {
+      "epoch": 12.85286783042394,
+      "grad_norm": 8.516155242919922,
+      "learning_rate": 7.153117206982544e-06,
+      "loss": 0.3259,
+      "step": 51540
+    },
+    {
+      "epoch": 12.855361596009976,
+      "grad_norm": 6.97627067565918,
+      "learning_rate": 7.150623441396509e-06,
+      "loss": 0.3679,
+      "step": 51550
+    },
+    {
+      "epoch": 12.85785536159601,
+      "grad_norm": 9.25831127166748,
+      "learning_rate": 7.1481296758104746e-06,
+      "loss": 0.3705,
+      "step": 51560
+    },
+    {
+      "epoch": 12.860349127182046,
+      "grad_norm": 4.845349311828613,
+      "learning_rate": 7.14563591022444e-06,
+      "loss": 0.2548,
+      "step": 51570
+    },
+    {
+      "epoch": 12.86284289276808,
+      "grad_norm": 7.061251163482666,
+      "learning_rate": 7.143142144638404e-06,
+      "loss": 0.3352,
+      "step": 51580
+    },
+    {
+      "epoch": 12.865336658354115,
+      "grad_norm": 7.0790557861328125,
+      "learning_rate": 7.14064837905237e-06,
+      "loss": 0.2947,
+      "step": 51590
+    },
+    {
+      "epoch": 12.86783042394015,
+      "grad_norm": 9.186144828796387,
+      "learning_rate": 7.138154613466335e-06,
+      "loss": 0.3689,
+      "step": 51600
+    },
+    {
+      "epoch": 12.870324189526185,
+      "grad_norm": 5.841263771057129,
+      "learning_rate": 7.1356608478803e-06,
+      "loss": 0.3377,
+      "step": 51610
+    },
+    {
+      "epoch": 12.87281795511222,
+      "grad_norm": 15.965258598327637,
+      "learning_rate": 7.133167082294265e-06,
+      "loss": 0.3976,
+      "step": 51620
+    },
+    {
+      "epoch": 12.875311720698255,
+      "grad_norm": 7.985716342926025,
+      "learning_rate": 7.13067331670823e-06,
+      "loss": 0.3237,
+      "step": 51630
+    },
+    {
+      "epoch": 12.87780548628429,
+      "grad_norm": 7.872910499572754,
+      "learning_rate": 7.128179551122195e-06,
+      "loss": 0.3373,
+      "step": 51640
+    },
+    {
+      "epoch": 12.880299251870325,
+      "grad_norm": 8.896018028259277,
+      "learning_rate": 7.125685785536161e-06,
+      "loss": 0.3554,
+      "step": 51650
+    },
+    {
+      "epoch": 12.88279301745636,
+      "grad_norm": 8.754325866699219,
+      "learning_rate": 7.123192019950125e-06,
+      "loss": 0.3481,
+      "step": 51660
+    },
+    {
+      "epoch": 12.885286783042394,
+      "grad_norm": 9.110607147216797,
+      "learning_rate": 7.1206982543640904e-06,
+      "loss": 0.3519,
+      "step": 51670
+    },
+    {
+      "epoch": 12.88778054862843,
+      "grad_norm": 10.086724281311035,
+      "learning_rate": 7.118204488778056e-06,
+      "loss": 0.3255,
+      "step": 51680
+    },
+    {
+      "epoch": 12.890274314214464,
+      "grad_norm": 9.028765678405762,
+      "learning_rate": 7.115710723192021e-06,
+      "loss": 0.3527,
+      "step": 51690
+    },
+    {
+      "epoch": 12.892768079800499,
+      "grad_norm": 12.050107955932617,
+      "learning_rate": 7.113216957605985e-06,
+      "loss": 0.3747,
+      "step": 51700
+    },
+    {
+      "epoch": 12.895261845386534,
+      "grad_norm": 5.377803325653076,
+      "learning_rate": 7.110723192019951e-06,
+      "loss": 0.3493,
+      "step": 51710
+    },
+    {
+      "epoch": 12.897755610972569,
+      "grad_norm": 12.90459156036377,
+      "learning_rate": 7.108229426433916e-06,
+      "loss": 0.3973,
+      "step": 51720
+    },
+    {
+      "epoch": 12.900249376558603,
+      "grad_norm": 8.819035530090332,
+      "learning_rate": 7.1057356608478814e-06,
+      "loss": 0.3746,
+      "step": 51730
+    },
+    {
+      "epoch": 12.902743142144638,
+      "grad_norm": 7.119004726409912,
+      "learning_rate": 7.103241895261847e-06,
+      "loss": 0.3413,
+      "step": 51740
+    },
+    {
+      "epoch": 12.905236907730673,
+      "grad_norm": 9.566146850585938,
+      "learning_rate": 7.10074812967581e-06,
+      "loss": 0.3475,
+      "step": 51750
+    },
+    {
+      "epoch": 12.907730673316708,
+      "grad_norm": 6.812338829040527,
+      "learning_rate": 7.098254364089776e-06,
+      "loss": 0.3002,
+      "step": 51760
+    },
+    {
+      "epoch": 12.910224438902743,
+      "grad_norm": 6.894874095916748,
+      "learning_rate": 7.095760598503742e-06,
+      "loss": 0.3632,
+      "step": 51770
+    },
+    {
+      "epoch": 12.912718204488778,
+      "grad_norm": 10.196784019470215,
+      "learning_rate": 7.093266832917707e-06,
+      "loss": 0.3053,
+      "step": 51780
+    },
+    {
+      "epoch": 12.915211970074813,
+      "grad_norm": 7.024259090423584,
+      "learning_rate": 7.090773067331671e-06,
+      "loss": 0.3858,
+      "step": 51790
+    },
+    {
+      "epoch": 12.917705735660848,
+      "grad_norm": 8.830977439880371,
+      "learning_rate": 7.088279301745636e-06,
+      "loss": 0.319,
+      "step": 51800
+    },
+    {
+      "epoch": 12.920199501246882,
+      "grad_norm": 7.131579875946045,
+      "learning_rate": 7.085785536159601e-06,
+      "loss": 0.3245,
+      "step": 51810
+    },
+    {
+      "epoch": 12.922693266832917,
+      "grad_norm": 10.366734504699707,
+      "learning_rate": 7.083291770573567e-06,
+      "loss": 0.3072,
+      "step": 51820
+    },
+    {
+      "epoch": 12.925187032418952,
+      "grad_norm": 14.305132865905762,
+      "learning_rate": 7.080798004987531e-06,
+      "loss": 0.3282,
+      "step": 51830
+    },
+    {
+      "epoch": 12.927680798004987,
+      "grad_norm": 5.115872859954834,
+      "learning_rate": 7.0783042394014965e-06,
+      "loss": 0.3539,
+      "step": 51840
+    },
+    {
+      "epoch": 12.930174563591022,
+      "grad_norm": 6.633999347686768,
+      "learning_rate": 7.075810473815462e-06,
+      "loss": 0.4114,
+      "step": 51850
+    },
+    {
+      "epoch": 12.932668329177057,
+      "grad_norm": 9.143322944641113,
+      "learning_rate": 7.073316708229427e-06,
+      "loss": 0.3024,
+      "step": 51860
+    },
+    {
+      "epoch": 12.935162094763092,
+      "grad_norm": 8.794203758239746,
+      "learning_rate": 7.0708229426433915e-06,
+      "loss": 0.3265,
+      "step": 51870
+    },
+    {
+      "epoch": 12.937655860349127,
+      "grad_norm": 8.29458999633789,
+      "learning_rate": 7.068329177057357e-06,
+      "loss": 0.3302,
+      "step": 51880
+    },
+    {
+      "epoch": 12.940149625935161,
+      "grad_norm": 8.905904769897461,
+      "learning_rate": 7.065835411471322e-06,
+      "loss": 0.4036,
+      "step": 51890
+    },
+    {
+      "epoch": 12.942643391521196,
+      "grad_norm": 11.505474090576172,
+      "learning_rate": 7.0633416458852875e-06,
+      "loss": 0.3459,
+      "step": 51900
+    },
+    {
+      "epoch": 12.945137157107231,
+      "grad_norm": 7.7994561195373535,
+      "learning_rate": 7.060847880299252e-06,
+      "loss": 0.3477,
+      "step": 51910
+    },
+    {
+      "epoch": 12.947630922693268,
+      "grad_norm": 8.445225715637207,
+      "learning_rate": 7.058354114713217e-06,
+      "loss": 0.3153,
+      "step": 51920
+    },
+    {
+      "epoch": 12.950124688279303,
+      "grad_norm": 7.54745626449585,
+      "learning_rate": 7.0558603491271825e-06,
+      "loss": 0.2894,
+      "step": 51930
+    },
+    {
+      "epoch": 12.952618453865338,
+      "grad_norm": 8.576315879821777,
+      "learning_rate": 7.053366583541148e-06,
+      "loss": 0.3849,
+      "step": 51940
+    },
+    {
+      "epoch": 12.955112219451372,
+      "grad_norm": 11.285699844360352,
+      "learning_rate": 7.050872817955112e-06,
+      "loss": 0.3399,
+      "step": 51950
+    },
+    {
+      "epoch": 12.957605985037407,
+      "grad_norm": 5.247829914093018,
+      "learning_rate": 7.048379052369078e-06,
+      "loss": 0.299,
+      "step": 51960
+    },
+    {
+      "epoch": 12.960099750623442,
+      "grad_norm": 8.704336166381836,
+      "learning_rate": 7.045885286783043e-06,
+      "loss": 0.3415,
+      "step": 51970
+    },
+    {
+      "epoch": 12.962593516209477,
+      "grad_norm": 9.179566383361816,
+      "learning_rate": 7.043391521197008e-06,
+      "loss": 0.3734,
+      "step": 51980
+    },
+    {
+      "epoch": 12.965087281795512,
+      "grad_norm": 6.699599266052246,
+      "learning_rate": 7.0408977556109736e-06,
+      "loss": 0.3288,
+      "step": 51990
+    },
+    {
+      "epoch": 12.967581047381547,
+      "grad_norm": 6.417098045349121,
+      "learning_rate": 7.038403990024938e-06,
+      "loss": 0.2907,
+      "step": 52000
+    },
+    {
+      "epoch": 12.970074812967582,
+      "grad_norm": 7.011273384094238,
+      "learning_rate": 7.035910224438903e-06,
+      "loss": 0.3603,
+      "step": 52010
+    },
+    {
+      "epoch": 12.972568578553616,
+      "grad_norm": 8.827155113220215,
+      "learning_rate": 7.033416458852869e-06,
+      "loss": 0.3839,
+      "step": 52020
+    },
+    {
+      "epoch": 12.975062344139651,
+      "grad_norm": 7.616529941558838,
+      "learning_rate": 7.030922693266834e-06,
+      "loss": 0.3665,
+      "step": 52030
+    },
+    {
+      "epoch": 12.977556109725686,
+      "grad_norm": 8.937703132629395,
+      "learning_rate": 7.028428927680798e-06,
+      "loss": 0.3125,
+      "step": 52040
+    },
+    {
+      "epoch": 12.980049875311721,
+      "grad_norm": 9.48192310333252,
+      "learning_rate": 7.025935162094764e-06,
+      "loss": 0.445,
+      "step": 52050
+    },
+    {
+      "epoch": 12.982543640897756,
+      "grad_norm": 8.205970764160156,
+      "learning_rate": 7.023441396508729e-06,
+      "loss": 0.3578,
+      "step": 52060
+    },
+    {
+      "epoch": 12.98503740648379,
+      "grad_norm": 9.55613899230957,
+      "learning_rate": 7.020947630922694e-06,
+      "loss": 0.301,
+      "step": 52070
+    },
+    {
+      "epoch": 12.987531172069826,
+      "grad_norm": 7.20795202255249,
+      "learning_rate": 7.018453865336659e-06,
+      "loss": 0.3893,
+      "step": 52080
+    },
+    {
+      "epoch": 12.99002493765586,
+      "grad_norm": 12.070457458496094,
+      "learning_rate": 7.015960099750624e-06,
+      "loss": 0.3387,
+      "step": 52090
+    },
+    {
+      "epoch": 12.992518703241895,
+      "grad_norm": 4.415650844573975,
+      "learning_rate": 7.0134663341645894e-06,
+      "loss": 0.3183,
+      "step": 52100
+    },
+    {
+      "epoch": 12.99501246882793,
+      "grad_norm": 6.928340911865234,
+      "learning_rate": 7.010972568578555e-06,
+      "loss": 0.3319,
+      "step": 52110
+    },
+    {
+      "epoch": 12.997506234413965,
+      "grad_norm": 6.329419136047363,
+      "learning_rate": 7.008478802992519e-06,
+      "loss": 0.3946,
+      "step": 52120
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 4.517069339752197,
+      "learning_rate": 7.0059850374064845e-06,
+      "loss": 0.3482,
+      "step": 52130
+    },
+    {
+      "epoch": 13.0,
+      "eval_loss": 0.41498884558677673,
+      "eval_runtime": 60.065,
+      "eval_samples_per_second": 16.699,
+      "eval_steps_per_second": 16.699,
+      "step": 52130
+    },
+    {
+      "epoch": 13.002493765586035,
+      "grad_norm": 6.689090728759766,
+      "learning_rate": 7.00349127182045e-06,
+      "loss": 0.3395,
+      "step": 52140
+    },
+    {
+      "epoch": 13.00498753117207,
+      "grad_norm": 7.670643329620361,
+      "learning_rate": 7.000997506234415e-06,
+      "loss": 0.3143,
+      "step": 52150
+    },
+    {
+      "epoch": 13.007481296758105,
+      "grad_norm": 10.944851875305176,
+      "learning_rate": 6.998503740648379e-06,
+      "loss": 0.3196,
+      "step": 52160
+    },
+    {
+      "epoch": 13.00997506234414,
+      "grad_norm": 5.979905128479004,
+      "learning_rate": 6.996009975062344e-06,
+      "loss": 0.3221,
+      "step": 52170
+    },
+    {
+      "epoch": 13.012468827930174,
+      "grad_norm": 6.5112762451171875,
+      "learning_rate": 6.99351620947631e-06,
+      "loss": 0.2684,
+      "step": 52180
+    },
+    {
+      "epoch": 13.01496259351621,
+      "grad_norm": 5.4960761070251465,
+      "learning_rate": 6.9910224438902755e-06,
+      "loss": 0.2942,
+      "step": 52190
+    },
+    {
+      "epoch": 13.017456359102244,
+      "grad_norm": 6.578577995300293,
+      "learning_rate": 6.988528678304239e-06,
+      "loss": 0.3318,
+      "step": 52200
+    },
+    {
+      "epoch": 13.019950124688279,
+      "grad_norm": 7.574843406677246,
+      "learning_rate": 6.9860349127182044e-06,
+      "loss": 0.3365,
+      "step": 52210
+    },
+    {
+      "epoch": 13.022443890274314,
+      "grad_norm": 6.713075160980225,
+      "learning_rate": 6.98354114713217e-06,
+      "loss": 0.3522,
+      "step": 52220
+    },
+    {
+      "epoch": 13.024937655860349,
+      "grad_norm": 10.647099494934082,
+      "learning_rate": 6.981047381546135e-06,
+      "loss": 0.2822,
+      "step": 52230
+    },
+    {
+      "epoch": 13.027431421446384,
+      "grad_norm": 6.798612594604492,
+      "learning_rate": 6.978553615960101e-06,
+      "loss": 0.2812,
+      "step": 52240
+    },
+    {
+      "epoch": 13.029925187032418,
+      "grad_norm": 6.139517784118652,
+      "learning_rate": 6.976059850374065e-06,
+      "loss": 0.3224,
+      "step": 52250
+    },
+    {
+      "epoch": 13.032418952618453,
+      "grad_norm": 5.845915794372559,
+      "learning_rate": 6.97356608478803e-06,
+      "loss": 0.2987,
+      "step": 52260
+    },
+    {
+      "epoch": 13.034912718204488,
+      "grad_norm": 6.590796947479248,
+      "learning_rate": 6.9710723192019955e-06,
+      "loss": 0.3018,
+      "step": 52270
+    },
+    {
+      "epoch": 13.037406483790523,
+      "grad_norm": 6.91572904586792,
+      "learning_rate": 6.968578553615961e-06,
+      "loss": 0.3517,
+      "step": 52280
+    },
+    {
+      "epoch": 13.039900249376558,
+      "grad_norm": 5.153928756713867,
+      "learning_rate": 6.966084788029925e-06,
+      "loss": 0.2843,
+      "step": 52290
+    },
+    {
+      "epoch": 13.042394014962593,
+      "grad_norm": 8.661881446838379,
+      "learning_rate": 6.9635910224438905e-06,
+      "loss": 0.278,
+      "step": 52300
+    },
+    {
+      "epoch": 13.044887780548628,
+      "grad_norm": 6.513763904571533,
+      "learning_rate": 6.961097256857856e-06,
+      "loss": 0.3141,
+      "step": 52310
+    },
+    {
+      "epoch": 13.047381546134662,
+      "grad_norm": 7.926911354064941,
+      "learning_rate": 6.958603491271821e-06,
+      "loss": 0.3725,
+      "step": 52320
+    },
+    {
+      "epoch": 13.049875311720697,
+      "grad_norm": 10.43213176727295,
+      "learning_rate": 6.956109725685786e-06,
+      "loss": 0.3259,
+      "step": 52330
+    },
+    {
+      "epoch": 13.052369077306734,
+      "grad_norm": 5.205043792724609,
+      "learning_rate": 6.953615960099751e-06,
+      "loss": 0.3007,
+      "step": 52340
+    },
+    {
+      "epoch": 13.054862842892769,
+      "grad_norm": 6.522244930267334,
+      "learning_rate": 6.951122194513716e-06,
+      "loss": 0.2957,
+      "step": 52350
+    },
+    {
+      "epoch": 13.057356608478804,
+      "grad_norm": 6.536242961883545,
+      "learning_rate": 6.9486284289276815e-06,
+      "loss": 0.3116,
+      "step": 52360
+    },
+    {
+      "epoch": 13.059850374064839,
+      "grad_norm": 8.81574535369873,
+      "learning_rate": 6.946134663341646e-06,
+      "loss": 0.3632,
+      "step": 52370
+    },
+    {
+      "epoch": 13.062344139650873,
+      "grad_norm": 6.305959701538086,
+      "learning_rate": 6.943640897755611e-06,
+      "loss": 0.3337,
+      "step": 52380
+    },
+    {
+      "epoch": 13.064837905236908,
+      "grad_norm": 12.51152515411377,
+      "learning_rate": 6.941147132169577e-06,
+      "loss": 0.3609,
+      "step": 52390
+    },
+    {
+      "epoch": 13.067331670822943,
+      "grad_norm": 8.995133399963379,
+      "learning_rate": 6.938653366583542e-06,
+      "loss": 0.3537,
+      "step": 52400
+    },
+    {
+      "epoch": 13.069825436408978,
+      "grad_norm": 10.008258819580078,
+      "learning_rate": 6.936159600997506e-06,
+      "loss": 0.3284,
+      "step": 52410
+    },
+    {
+      "epoch": 13.072319201995013,
+      "grad_norm": 7.722633361816406,
+      "learning_rate": 6.933665835411472e-06,
+      "loss": 0.3823,
+      "step": 52420
+    },
+    {
+      "epoch": 13.074812967581048,
+      "grad_norm": 8.553972244262695,
+      "learning_rate": 6.931172069825437e-06,
+      "loss": 0.3226,
+      "step": 52430
+    },
+    {
+      "epoch": 13.077306733167083,
+      "grad_norm": 6.591207027435303,
+      "learning_rate": 6.928678304239402e-06,
+      "loss": 0.3421,
+      "step": 52440
+    },
+    {
+      "epoch": 13.079800498753118,
+      "grad_norm": 5.500464916229248,
+      "learning_rate": 6.926184538653367e-06,
+      "loss": 0.2842,
+      "step": 52450
+    },
+    {
+      "epoch": 13.082294264339152,
+      "grad_norm": 8.214122772216797,
+      "learning_rate": 6.923690773067332e-06,
+      "loss": 0.2846,
+      "step": 52460
+    },
+    {
+      "epoch": 13.084788029925187,
+      "grad_norm": 7.543227195739746,
+      "learning_rate": 6.921197007481297e-06,
+      "loss": 0.3326,
+      "step": 52470
+    },
+    {
+      "epoch": 13.087281795511222,
+      "grad_norm": 8.072737693786621,
+      "learning_rate": 6.918703241895263e-06,
+      "loss": 0.3319,
+      "step": 52480
+    },
+    {
+      "epoch": 13.089775561097257,
+      "grad_norm": 8.425067901611328,
+      "learning_rate": 6.916209476309228e-06,
+      "loss": 0.303,
+      "step": 52490
+    },
+    {
+      "epoch": 13.092269326683292,
+      "grad_norm": 8.168347358703613,
+      "learning_rate": 6.9137157107231925e-06,
+      "loss": 0.3316,
+      "step": 52500
+    },
+    {
+      "epoch": 13.094763092269327,
+      "grad_norm": 7.062584400177002,
+      "learning_rate": 6.911221945137158e-06,
+      "loss": 0.2915,
+      "step": 52510
+    },
+    {
+      "epoch": 13.097256857855362,
+      "grad_norm": 8.098386764526367,
+      "learning_rate": 6.908728179551123e-06,
+      "loss": 0.3826,
+      "step": 52520
+    },
+    {
+      "epoch": 13.099750623441397,
+      "grad_norm": 6.54594087600708,
+      "learning_rate": 6.9062344139650884e-06,
+      "loss": 0.3046,
+      "step": 52530
+    },
+    {
+      "epoch": 13.102244389027431,
+      "grad_norm": 8.44714069366455,
+      "learning_rate": 6.903740648379053e-06,
+      "loss": 0.296,
+      "step": 52540
+    },
+    {
+      "epoch": 13.104738154613466,
+      "grad_norm": 7.943593502044678,
+      "learning_rate": 6.901246882793018e-06,
+      "loss": 0.3459,
+      "step": 52550
+    },
+    {
+      "epoch": 13.107231920199501,
+      "grad_norm": 5.09817361831665,
+      "learning_rate": 6.8987531172069835e-06,
+      "loss": 0.2701,
+      "step": 52560
+    },
+    {
+      "epoch": 13.109725685785536,
+      "grad_norm": 10.518460273742676,
+      "learning_rate": 6.896259351620949e-06,
+      "loss": 0.3679,
+      "step": 52570
+    },
+    {
+      "epoch": 13.11221945137157,
+      "grad_norm": 7.73917293548584,
+      "learning_rate": 6.893765586034913e-06,
+      "loss": 0.2963,
+      "step": 52580
+    },
+    {
+      "epoch": 13.114713216957606,
+      "grad_norm": 8.970417976379395,
+      "learning_rate": 6.891271820448879e-06,
+      "loss": 0.316,
+      "step": 52590
+    },
+    {
+      "epoch": 13.11720698254364,
+      "grad_norm": 9.928048133850098,
+      "learning_rate": 6.888778054862844e-06,
+      "loss": 0.3444,
+      "step": 52600
+    },
+    {
+      "epoch": 13.119700748129675,
+      "grad_norm": 12.679057121276855,
+      "learning_rate": 6.886284289276809e-06,
+      "loss": 0.3914,
+      "step": 52610
+    },
+    {
+      "epoch": 13.12219451371571,
+      "grad_norm": 8.21277904510498,
+      "learning_rate": 6.883790523690773e-06,
+      "loss": 0.293,
+      "step": 52620
+    },
+    {
+      "epoch": 13.124688279301745,
+      "grad_norm": 8.76871109008789,
+      "learning_rate": 6.881296758104738e-06,
+      "loss": 0.3419,
+      "step": 52630
+    },
+    {
+      "epoch": 13.12718204488778,
+      "grad_norm": 7.2397236824035645,
+      "learning_rate": 6.878802992518704e-06,
+      "loss": 0.3769,
+      "step": 52640
+    },
+    {
+      "epoch": 13.129675810473815,
+      "grad_norm": 7.731866836547852,
+      "learning_rate": 6.87630922693267e-06,
+      "loss": 0.3589,
+      "step": 52650
+    },
+    {
+      "epoch": 13.13216957605985,
+      "grad_norm": 8.515292167663574,
+      "learning_rate": 6.873815461346633e-06,
+      "loss": 0.3328,
+      "step": 52660
+    },
+    {
+      "epoch": 13.134663341645885,
+      "grad_norm": 5.041269779205322,
+      "learning_rate": 6.8713216957605985e-06,
+      "loss": 0.3182,
+      "step": 52670
+    },
+    {
+      "epoch": 13.13715710723192,
+      "grad_norm": 7.612249851226807,
+      "learning_rate": 6.868827930174564e-06,
+      "loss": 0.3596,
+      "step": 52680
+    },
+    {
+      "epoch": 13.139650872817954,
+      "grad_norm": 8.448023796081543,
+      "learning_rate": 6.866334164588529e-06,
+      "loss": 0.3286,
+      "step": 52690
+    },
+    {
+      "epoch": 13.14214463840399,
+      "grad_norm": 8.699821472167969,
+      "learning_rate": 6.863840399002495e-06,
+      "loss": 0.3803,
+      "step": 52700
+    },
+    {
+      "epoch": 13.144638403990024,
+      "grad_norm": 9.197280883789062,
+      "learning_rate": 6.861346633416459e-06,
+      "loss": 0.3577,
+      "step": 52710
+    },
+    {
+      "epoch": 13.147132169576059,
+      "grad_norm": 5.9945597648620605,
+      "learning_rate": 6.858852867830424e-06,
+      "loss": 0.3208,
+      "step": 52720
+    },
+    {
+      "epoch": 13.149625935162096,
+      "grad_norm": 7.514776229858398,
+      "learning_rate": 6.8563591022443895e-06,
+      "loss": 0.3625,
+      "step": 52730
+    },
+    {
+      "epoch": 13.15211970074813,
+      "grad_norm": 17.048492431640625,
+      "learning_rate": 6.853865336658355e-06,
+      "loss": 0.3416,
+      "step": 52740
+    },
+    {
+      "epoch": 13.154613466334165,
+      "grad_norm": 9.108981132507324,
+      "learning_rate": 6.851371571072319e-06,
+      "loss": 0.3295,
+      "step": 52750
+    },
+    {
+      "epoch": 13.1571072319202,
+      "grad_norm": 4.699746131896973,
+      "learning_rate": 6.848877805486285e-06,
+      "loss": 0.3258,
+      "step": 52760
+    },
+    {
+      "epoch": 13.159600997506235,
+      "grad_norm": 9.46227741241455,
+      "learning_rate": 6.84638403990025e-06,
+      "loss": 0.3021,
+      "step": 52770
+    },
+    {
+      "epoch": 13.16209476309227,
+      "grad_norm": 10.8130521774292,
+      "learning_rate": 6.843890274314215e-06,
+      "loss": 0.3349,
+      "step": 52780
+    },
+    {
+      "epoch": 13.164588528678305,
+      "grad_norm": 6.772179126739502,
+      "learning_rate": 6.84139650872818e-06,
+      "loss": 0.3241,
+      "step": 52790
+    },
+    {
+      "epoch": 13.16708229426434,
+      "grad_norm": 8.732086181640625,
+      "learning_rate": 6.838902743142145e-06,
+      "loss": 0.2967,
+      "step": 52800
+    },
+    {
+      "epoch": 13.169576059850375,
+      "grad_norm": 8.218696594238281,
+      "learning_rate": 6.83640897755611e-06,
+      "loss": 0.3601,
+      "step": 52810
+    },
+    {
+      "epoch": 13.17206982543641,
+      "grad_norm": 11.002167701721191,
+      "learning_rate": 6.833915211970076e-06,
+      "loss": 0.3686,
+      "step": 52820
+    },
+    {
+      "epoch": 13.174563591022444,
+      "grad_norm": 7.543792724609375,
+      "learning_rate": 6.83142144638404e-06,
+      "loss": 0.3467,
+      "step": 52830
+    },
+    {
+      "epoch": 13.17705735660848,
+      "grad_norm": 7.169468402862549,
+      "learning_rate": 6.828927680798005e-06,
+      "loss": 0.3385,
+      "step": 52840
+    },
+    {
+      "epoch": 13.179551122194514,
+      "grad_norm": 11.047392845153809,
+      "learning_rate": 6.826433915211971e-06,
+      "loss": 0.3213,
+      "step": 52850
+    },
+    {
+      "epoch": 13.182044887780549,
+      "grad_norm": 7.974063396453857,
+      "learning_rate": 6.823940149625936e-06,
+      "loss": 0.3075,
+      "step": 52860
+    },
+    {
+      "epoch": 13.184538653366584,
+      "grad_norm": 9.911652565002441,
+      "learning_rate": 6.8214463840399005e-06,
+      "loss": 0.2636,
+      "step": 52870
+    },
+    {
+      "epoch": 13.187032418952619,
+      "grad_norm": 5.923473834991455,
+      "learning_rate": 6.818952618453866e-06,
+      "loss": 0.3013,
+      "step": 52880
+    },
+    {
+      "epoch": 13.189526184538654,
+      "grad_norm": 7.126490592956543,
+      "learning_rate": 6.816458852867831e-06,
+      "loss": 0.2964,
+      "step": 52890
+    },
+    {
+      "epoch": 13.192019950124688,
+      "grad_norm": 6.427547931671143,
+      "learning_rate": 6.813965087281796e-06,
+      "loss": 0.3058,
+      "step": 52900
+    },
+    {
+      "epoch": 13.194513715710723,
+      "grad_norm": 6.297787189483643,
+      "learning_rate": 6.811471321695761e-06,
+      "loss": 0.2955,
+      "step": 52910
+    },
+    {
+      "epoch": 13.197007481296758,
+      "grad_norm": 7.9449005126953125,
+      "learning_rate": 6.808977556109726e-06,
+      "loss": 0.3024,
+      "step": 52920
+    },
+    {
+      "epoch": 13.199501246882793,
+      "grad_norm": 10.155074119567871,
+      "learning_rate": 6.8064837905236915e-06,
+      "loss": 0.3011,
+      "step": 52930
+    },
+    {
+      "epoch": 13.201995012468828,
+      "grad_norm": 8.037057876586914,
+      "learning_rate": 6.803990024937657e-06,
+      "loss": 0.3507,
+      "step": 52940
+    },
+    {
+      "epoch": 13.204488778054863,
+      "grad_norm": 8.216534614562988,
+      "learning_rate": 6.801496259351622e-06,
+      "loss": 0.3815,
+      "step": 52950
+    },
+    {
+      "epoch": 13.206982543640898,
+      "grad_norm": 8.674677848815918,
+      "learning_rate": 6.799002493765587e-06,
+      "loss": 0.3702,
+      "step": 52960
+    },
+    {
+      "epoch": 13.209476309226932,
+      "grad_norm": 6.525129795074463,
+      "learning_rate": 6.796508728179552e-06,
+      "loss": 0.2832,
+      "step": 52970
+    },
+    {
+      "epoch": 13.211970074812967,
+      "grad_norm": 20.87378692626953,
+      "learning_rate": 6.794014962593517e-06,
+      "loss": 0.4099,
+      "step": 52980
+    },
+    {
+      "epoch": 13.214463840399002,
+      "grad_norm": 6.119201183319092,
+      "learning_rate": 6.7915211970074825e-06,
+      "loss": 0.3321,
+      "step": 52990
+    },
+    {
+      "epoch": 13.216957605985037,
+      "grad_norm": 9.405641555786133,
+      "learning_rate": 6.789027431421447e-06,
+      "loss": 0.3069,
+      "step": 53000
+    },
+    {
+      "epoch": 13.219451371571072,
+      "grad_norm": 9.708418846130371,
+      "learning_rate": 6.786533665835412e-06,
+      "loss": 0.3953,
+      "step": 53010
+    },
+    {
+      "epoch": 13.221945137157107,
+      "grad_norm": 6.832964897155762,
+      "learning_rate": 6.784039900249378e-06,
+      "loss": 0.3541,
+      "step": 53020
+    },
+    {
+      "epoch": 13.224438902743142,
+      "grad_norm": 15.42807674407959,
+      "learning_rate": 6.781546134663343e-06,
+      "loss": 0.4261,
+      "step": 53030
+    },
+    {
+      "epoch": 13.226932668329177,
+      "grad_norm": 17.721216201782227,
+      "learning_rate": 6.7790523690773065e-06,
+      "loss": 0.3629,
+      "step": 53040
+    },
+    {
+      "epoch": 13.229426433915211,
+      "grad_norm": 10.285332679748535,
+      "learning_rate": 6.776558603491273e-06,
+      "loss": 0.329,
+      "step": 53050
+    },
+    {
+      "epoch": 13.231920199501246,
+      "grad_norm": 4.8609113693237305,
+      "learning_rate": 6.774064837905238e-06,
+      "loss": 0.3147,
+      "step": 53060
+    },
+    {
+      "epoch": 13.234413965087281,
+      "grad_norm": 8.452110290527344,
+      "learning_rate": 6.771571072319203e-06,
+      "loss": 0.3458,
+      "step": 53070
+    },
+    {
+      "epoch": 13.236907730673316,
+      "grad_norm": 6.897256851196289,
+      "learning_rate": 6.769077306733167e-06,
+      "loss": 0.3171,
+      "step": 53080
+    },
+    {
+      "epoch": 13.239401496259351,
+      "grad_norm": 8.701833724975586,
+      "learning_rate": 6.766583541147132e-06,
+      "loss": 0.3182,
+      "step": 53090
+    },
+    {
+      "epoch": 13.241895261845386,
+      "grad_norm": 6.701672554016113,
+      "learning_rate": 6.7640897755610975e-06,
+      "loss": 0.4088,
+      "step": 53100
+    },
+    {
+      "epoch": 13.24438902743142,
+      "grad_norm": 9.619290351867676,
+      "learning_rate": 6.761596009975064e-06,
+      "loss": 0.354,
+      "step": 53110
+    },
+    {
+      "epoch": 13.246882793017456,
+      "grad_norm": 8.733819007873535,
+      "learning_rate": 6.759102244389027e-06,
+      "loss": 0.3971,
+      "step": 53120
+    },
+    {
+      "epoch": 13.24937655860349,
+      "grad_norm": 7.536902904510498,
+      "learning_rate": 6.756608478802993e-06,
+      "loss": 0.382,
+      "step": 53130
+    },
+    {
+      "epoch": 13.251870324189527,
+      "grad_norm": 8.806652069091797,
+      "learning_rate": 6.754114713216958e-06,
+      "loss": 0.359,
+      "step": 53140
+    },
+    {
+      "epoch": 13.254364089775562,
+      "grad_norm": 8.000337600708008,
+      "learning_rate": 6.751620947630923e-06,
+      "loss": 0.3648,
+      "step": 53150
+    },
+    {
+      "epoch": 13.256857855361597,
+      "grad_norm": 10.321428298950195,
+      "learning_rate": 6.749127182044888e-06,
+      "loss": 0.3462,
+      "step": 53160
+    },
+    {
+      "epoch": 13.259351620947632,
+      "grad_norm": 11.810160636901855,
+      "learning_rate": 6.746633416458853e-06,
+      "loss": 0.3129,
+      "step": 53170
+    },
+    {
+      "epoch": 13.261845386533667,
+      "grad_norm": 8.895573616027832,
+      "learning_rate": 6.744139650872818e-06,
+      "loss": 0.2583,
+      "step": 53180
+    },
+    {
+      "epoch": 13.264339152119701,
+      "grad_norm": 8.861416816711426,
+      "learning_rate": 6.741645885286784e-06,
+      "loss": 0.3389,
+      "step": 53190
+    },
+    {
+      "epoch": 13.266832917705736,
+      "grad_norm": 7.294126510620117,
+      "learning_rate": 6.739152119700749e-06,
+      "loss": 0.3496,
+      "step": 53200
+    },
+    {
+      "epoch": 13.269326683291771,
+      "grad_norm": 7.825159072875977,
+      "learning_rate": 6.736658354114713e-06,
+      "loss": 0.321,
+      "step": 53210
+    },
+    {
+      "epoch": 13.271820448877806,
+      "grad_norm": 7.070465564727783,
+      "learning_rate": 6.734164588528679e-06,
+      "loss": 0.2964,
+      "step": 53220
+    },
+    {
+      "epoch": 13.27431421446384,
+      "grad_norm": 11.65014934539795,
+      "learning_rate": 6.731670822942644e-06,
+      "loss": 0.2895,
+      "step": 53230
+    },
+    {
+      "epoch": 13.276807980049876,
+      "grad_norm": 4.547674655914307,
+      "learning_rate": 6.729177057356609e-06,
+      "loss": 0.3454,
+      "step": 53240
+    },
+    {
+      "epoch": 13.27930174563591,
+      "grad_norm": 6.833098411560059,
+      "learning_rate": 6.726683291770574e-06,
+      "loss": 0.3121,
+      "step": 53250
+    },
+    {
+      "epoch": 13.281795511221945,
+      "grad_norm": 11.146747589111328,
+      "learning_rate": 6.724189526184539e-06,
+      "loss": 0.36,
+      "step": 53260
+    },
+    {
+      "epoch": 13.28428927680798,
+      "grad_norm": 9.843538284301758,
+      "learning_rate": 6.721695760598504e-06,
+      "loss": 0.3221,
+      "step": 53270
+    },
+    {
+      "epoch": 13.286783042394015,
+      "grad_norm": 8.288082122802734,
+      "learning_rate": 6.71920199501247e-06,
+      "loss": 0.3221,
+      "step": 53280
+    },
+    {
+      "epoch": 13.28927680798005,
+      "grad_norm": 9.893383979797363,
+      "learning_rate": 6.716708229426434e-06,
+      "loss": 0.2934,
+      "step": 53290
+    },
+    {
+      "epoch": 13.291770573566085,
+      "grad_norm": 5.806131362915039,
+      "learning_rate": 6.7142144638403995e-06,
+      "loss": 0.401,
+      "step": 53300
+    },
+    {
+      "epoch": 13.29426433915212,
+      "grad_norm": 9.158594131469727,
+      "learning_rate": 6.711720698254365e-06,
+      "loss": 0.2943,
+      "step": 53310
+    },
+    {
+      "epoch": 13.296758104738155,
+      "grad_norm": 6.4060139656066895,
+      "learning_rate": 6.70922693266833e-06,
+      "loss": 0.3538,
+      "step": 53320
+    },
+    {
+      "epoch": 13.29925187032419,
+      "grad_norm": 10.887904167175293,
+      "learning_rate": 6.7067331670822946e-06,
+      "loss": 0.3591,
+      "step": 53330
+    },
+    {
+      "epoch": 13.301745635910224,
+      "grad_norm": 6.867051124572754,
+      "learning_rate": 6.70423940149626e-06,
+      "loss": 0.2973,
+      "step": 53340
+    },
+    {
+      "epoch": 13.30423940149626,
+      "grad_norm": 6.542517185211182,
+      "learning_rate": 6.701745635910225e-06,
+      "loss": 0.2912,
+      "step": 53350
+    },
+    {
+      "epoch": 13.306733167082294,
+      "grad_norm": 7.9812517166137695,
+      "learning_rate": 6.6992518703241905e-06,
+      "loss": 0.29,
+      "step": 53360
+    },
+    {
+      "epoch": 13.309226932668329,
+      "grad_norm": 7.906765460968018,
+      "learning_rate": 6.696758104738155e-06,
+      "loss": 0.2706,
+      "step": 53370
+    },
+    {
+      "epoch": 13.311720698254364,
+      "grad_norm": 8.329736709594727,
+      "learning_rate": 6.69426433915212e-06,
+      "loss": 0.3764,
+      "step": 53380
+    },
+    {
+      "epoch": 13.314214463840399,
+      "grad_norm": 9.357162475585938,
+      "learning_rate": 6.691770573566086e-06,
+      "loss": 0.3357,
+      "step": 53390
+    },
+    {
+      "epoch": 13.316708229426434,
+      "grad_norm": 7.998769760131836,
+      "learning_rate": 6.689276807980051e-06,
+      "loss": 0.3845,
+      "step": 53400
+    },
+    {
+      "epoch": 13.319201995012468,
+      "grad_norm": 9.348040580749512,
+      "learning_rate": 6.686783042394015e-06,
+      "loss": 0.389,
+      "step": 53410
+    },
+    {
+      "epoch": 13.321695760598503,
+      "grad_norm": 6.765722274780273,
+      "learning_rate": 6.684289276807981e-06,
+      "loss": 0.3126,
+      "step": 53420
+    },
+    {
+      "epoch": 13.324189526184538,
+      "grad_norm": 7.536268711090088,
+      "learning_rate": 6.681795511221946e-06,
+      "loss": 0.3189,
+      "step": 53430
+    },
+    {
+      "epoch": 13.326683291770573,
+      "grad_norm": 5.146828651428223,
+      "learning_rate": 6.679301745635911e-06,
+      "loss": 0.3335,
+      "step": 53440
+    },
+    {
+      "epoch": 13.329177057356608,
+      "grad_norm": 7.732426166534424,
+      "learning_rate": 6.676807980049877e-06,
+      "loss": 0.3094,
+      "step": 53450
+    },
+    {
+      "epoch": 13.331670822942643,
+      "grad_norm": 7.80210018157959,
+      "learning_rate": 6.674314214463841e-06,
+      "loss": 0.4274,
+      "step": 53460
+    },
+    {
+      "epoch": 13.334164588528678,
+      "grad_norm": 11.900932312011719,
+      "learning_rate": 6.671820448877806e-06,
+      "loss": 0.3221,
+      "step": 53470
+    },
+    {
+      "epoch": 13.336658354114713,
+      "grad_norm": 7.599810600280762,
+      "learning_rate": 6.669326683291772e-06,
+      "loss": 0.3771,
+      "step": 53480
+    },
+    {
+      "epoch": 13.339152119700747,
+      "grad_norm": 6.467700958251953,
+      "learning_rate": 6.666832917705737e-06,
+      "loss": 0.32,
+      "step": 53490
+    },
+    {
+      "epoch": 13.341645885286782,
+      "grad_norm": 8.103590965270996,
+      "learning_rate": 6.664339152119701e-06,
+      "loss": 0.3803,
+      "step": 53500
+    },
+    {
+      "epoch": 13.344139650872817,
+      "grad_norm": 8.192358016967773,
+      "learning_rate": 6.661845386533666e-06,
+      "loss": 0.3562,
+      "step": 53510
+    },
+    {
+      "epoch": 13.346633416458852,
+      "grad_norm": 6.621999263763428,
+      "learning_rate": 6.659351620947632e-06,
+      "loss": 0.306,
+      "step": 53520
+    },
+    {
+      "epoch": 13.349127182044889,
+      "grad_norm": 10.717852592468262,
+      "learning_rate": 6.656857855361597e-06,
+      "loss": 0.3724,
+      "step": 53530
+    },
+    {
+      "epoch": 13.351620947630924,
+      "grad_norm": 11.658101081848145,
+      "learning_rate": 6.654364089775561e-06,
+      "loss": 0.3673,
+      "step": 53540
+    },
+    {
+      "epoch": 13.354114713216958,
+      "grad_norm": 8.711099624633789,
+      "learning_rate": 6.651870324189526e-06,
+      "loss": 0.3506,
+      "step": 53550
+    },
+    {
+      "epoch": 13.356608478802993,
+      "grad_norm": 8.620141983032227,
+      "learning_rate": 6.649376558603492e-06,
+      "loss": 0.341,
+      "step": 53560
+    },
+    {
+      "epoch": 13.359102244389028,
+      "grad_norm": 6.421422958374023,
+      "learning_rate": 6.646882793017457e-06,
+      "loss": 0.3318,
+      "step": 53570
+    },
+    {
+      "epoch": 13.361596009975063,
+      "grad_norm": 7.4736127853393555,
+      "learning_rate": 6.644389027431421e-06,
+      "loss": 0.386,
+      "step": 53580
+    },
+    {
+      "epoch": 13.364089775561098,
+      "grad_norm": 8.752002716064453,
+      "learning_rate": 6.641895261845387e-06,
+      "loss": 0.3384,
+      "step": 53590
+    },
+    {
+      "epoch": 13.366583541147133,
+      "grad_norm": 4.710286617279053,
+      "learning_rate": 6.639401496259352e-06,
+      "loss": 0.3548,
+      "step": 53600
+    },
+    {
+      "epoch": 13.369077306733168,
+      "grad_norm": 6.772241115570068,
+      "learning_rate": 6.636907730673317e-06,
+      "loss": 0.3097,
+      "step": 53610
+    },
+    {
+      "epoch": 13.371571072319203,
+      "grad_norm": 9.710107803344727,
+      "learning_rate": 6.634413965087282e-06,
+      "loss": 0.3294,
+      "step": 53620
+    },
+    {
+      "epoch": 13.374064837905237,
+      "grad_norm": 7.724490165710449,
+      "learning_rate": 6.631920199501247e-06,
+      "loss": 0.3282,
+      "step": 53630
+    },
+    {
+      "epoch": 13.376558603491272,
+      "grad_norm": 11.170283317565918,
+      "learning_rate": 6.629426433915212e-06,
+      "loss": 0.4139,
+      "step": 53640
+    },
+    {
+      "epoch": 13.379052369077307,
+      "grad_norm": 11.032581329345703,
+      "learning_rate": 6.626932668329178e-06,
+      "loss": 0.3775,
+      "step": 53650
+    },
+    {
+      "epoch": 13.381546134663342,
+      "grad_norm": 7.674757480621338,
+      "learning_rate": 6.624438902743142e-06,
+      "loss": 0.2728,
+      "step": 53660
+    },
+    {
+      "epoch": 13.384039900249377,
+      "grad_norm": 8.27985954284668,
+      "learning_rate": 6.6219451371571075e-06,
+      "loss": 0.3075,
+      "step": 53670
+    },
+    {
+      "epoch": 13.386533665835412,
+      "grad_norm": 9.89207935333252,
+      "learning_rate": 6.619451371571073e-06,
+      "loss": 0.3486,
+      "step": 53680
+    },
+    {
+      "epoch": 13.389027431421447,
+      "grad_norm": 7.001489162445068,
+      "learning_rate": 6.616957605985038e-06,
+      "loss": 0.3218,
+      "step": 53690
+    },
+    {
+      "epoch": 13.391521197007481,
+      "grad_norm": 8.43419361114502,
+      "learning_rate": 6.614463840399003e-06,
+      "loss": 0.3296,
+      "step": 53700
+    },
+    {
+      "epoch": 13.394014962593516,
+      "grad_norm": 9.101119995117188,
+      "learning_rate": 6.611970074812968e-06,
+      "loss": 0.4188,
+      "step": 53710
+    },
+    {
+      "epoch": 13.396508728179551,
+      "grad_norm": 7.946018218994141,
+      "learning_rate": 6.609476309226933e-06,
+      "loss": 0.3638,
+      "step": 53720
+    },
+    {
+      "epoch": 13.399002493765586,
+      "grad_norm": 4.495571136474609,
+      "learning_rate": 6.6069825436408985e-06,
+      "loss": 0.2811,
+      "step": 53730
+    },
+    {
+      "epoch": 13.401496259351621,
+      "grad_norm": 9.637377738952637,
+      "learning_rate": 6.604488778054864e-06,
+      "loss": 0.3226,
+      "step": 53740
+    },
+    {
+      "epoch": 13.403990024937656,
+      "grad_norm": 12.018438339233398,
+      "learning_rate": 6.601995012468828e-06,
+      "loss": 0.3199,
+      "step": 53750
+    },
+    {
+      "epoch": 13.40648379052369,
+      "grad_norm": 8.298521995544434,
+      "learning_rate": 6.5995012468827936e-06,
+      "loss": 0.3609,
+      "step": 53760
+    },
+    {
+      "epoch": 13.408977556109726,
+      "grad_norm": 8.563523292541504,
+      "learning_rate": 6.597007481296759e-06,
+      "loss": 0.319,
+      "step": 53770
+    },
+    {
+      "epoch": 13.41147132169576,
+      "grad_norm": 6.873093128204346,
+      "learning_rate": 6.594513715710724e-06,
+      "loss": 0.3697,
+      "step": 53780
+    },
+    {
+      "epoch": 13.413965087281795,
+      "grad_norm": 6.433836936950684,
+      "learning_rate": 6.592019950124689e-06,
+      "loss": 0.3516,
+      "step": 53790
+    },
+    {
+      "epoch": 13.41645885286783,
+      "grad_norm": 7.9647908210754395,
+      "learning_rate": 6.589526184538654e-06,
+      "loss": 0.3636,
+      "step": 53800
+    },
+    {
+      "epoch": 13.418952618453865,
+      "grad_norm": 8.467369079589844,
+      "learning_rate": 6.587032418952619e-06,
+      "loss": 0.3561,
+      "step": 53810
+    },
+    {
+      "epoch": 13.4214463840399,
+      "grad_norm": 5.356767654418945,
+      "learning_rate": 6.584538653366585e-06,
+      "loss": 0.3067,
+      "step": 53820
+    },
+    {
+      "epoch": 13.423940149625935,
+      "grad_norm": 8.168974876403809,
+      "learning_rate": 6.582044887780549e-06,
+      "loss": 0.3501,
+      "step": 53830
+    },
+    {
+      "epoch": 13.42643391521197,
+      "grad_norm": 7.033174514770508,
+      "learning_rate": 6.579551122194514e-06,
+      "loss": 0.3376,
+      "step": 53840
+    },
+    {
+      "epoch": 13.428927680798004,
+      "grad_norm": 9.300447463989258,
+      "learning_rate": 6.57705735660848e-06,
+      "loss": 0.3486,
+      "step": 53850
+    },
+    {
+      "epoch": 13.43142144638404,
+      "grad_norm": 7.828090667724609,
+      "learning_rate": 6.574563591022445e-06,
+      "loss": 0.3739,
+      "step": 53860
+    },
+    {
+      "epoch": 13.433915211970074,
+      "grad_norm": 8.298700332641602,
+      "learning_rate": 6.5720698254364094e-06,
+      "loss": 0.3279,
+      "step": 53870
+    },
+    {
+      "epoch": 13.436408977556109,
+      "grad_norm": 6.854353427886963,
+      "learning_rate": 6.569576059850375e-06,
+      "loss": 0.2454,
+      "step": 53880
+    },
+    {
+      "epoch": 13.438902743142144,
+      "grad_norm": 10.509041786193848,
+      "learning_rate": 6.56708229426434e-06,
+      "loss": 0.4803,
+      "step": 53890
+    },
+    {
+      "epoch": 13.441396508728179,
+      "grad_norm": 9.381756782531738,
+      "learning_rate": 6.564588528678305e-06,
+      "loss": 0.3879,
+      "step": 53900
+    },
+    {
+      "epoch": 13.443890274314214,
+      "grad_norm": 7.465641498565674,
+      "learning_rate": 6.562094763092269e-06,
+      "loss": 0.3324,
+      "step": 53910
+    },
+    {
+      "epoch": 13.446384039900249,
+      "grad_norm": 9.815617561340332,
+      "learning_rate": 6.559600997506234e-06,
+      "loss": 0.3524,
+      "step": 53920
+    },
+    {
+      "epoch": 13.448877805486283,
+      "grad_norm": 7.948529243469238,
+      "learning_rate": 6.5571072319202004e-06,
+      "loss": 0.4092,
+      "step": 53930
+    },
+    {
+      "epoch": 13.451371571072318,
+      "grad_norm": 6.234860897064209,
+      "learning_rate": 6.554613466334166e-06,
+      "loss": 0.2804,
+      "step": 53940
+    },
+    {
+      "epoch": 13.453865336658355,
+      "grad_norm": 8.012157440185547,
+      "learning_rate": 6.552119700748131e-06,
+      "loss": 0.2756,
+      "step": 53950
+    },
+    {
+      "epoch": 13.45635910224439,
+      "grad_norm": 9.687308311462402,
+      "learning_rate": 6.549625935162095e-06,
+      "loss": 0.3588,
+      "step": 53960
+    },
+    {
+      "epoch": 13.458852867830425,
+      "grad_norm": 9.27532958984375,
+      "learning_rate": 6.54713216957606e-06,
+      "loss": 0.3675,
+      "step": 53970
+    },
+    {
+      "epoch": 13.46134663341646,
+      "grad_norm": 6.1434712409973145,
+      "learning_rate": 6.544638403990025e-06,
+      "loss": 0.3445,
+      "step": 53980
+    },
+    {
+      "epoch": 13.463840399002494,
+      "grad_norm": 8.947775840759277,
+      "learning_rate": 6.5421446384039915e-06,
+      "loss": 0.334,
+      "step": 53990
+    },
+    {
+      "epoch": 13.46633416458853,
+      "grad_norm": 9.55975341796875,
+      "learning_rate": 6.539650872817955e-06,
+      "loss": 0.4319,
+      "step": 54000
+    },
+    {
+      "epoch": 13.468827930174564,
+      "grad_norm": 7.890659332275391,
+      "learning_rate": 6.53715710723192e-06,
+      "loss": 0.3558,
+      "step": 54010
+    },
+    {
+      "epoch": 13.471321695760599,
+      "grad_norm": 9.459158897399902,
+      "learning_rate": 6.534663341645886e-06,
+      "loss": 0.347,
+      "step": 54020
+    },
+    {
+      "epoch": 13.473815461346634,
+      "grad_norm": 10.759072303771973,
+      "learning_rate": 6.532169576059851e-06,
+      "loss": 0.336,
+      "step": 54030
+    },
+    {
+      "epoch": 13.476309226932669,
+      "grad_norm": 5.313018798828125,
+      "learning_rate": 6.5296758104738155e-06,
+      "loss": 0.2731,
+      "step": 54040
+    },
+    {
+      "epoch": 13.478802992518704,
+      "grad_norm": 6.71405029296875,
+      "learning_rate": 6.527182044887781e-06,
+      "loss": 0.3615,
+      "step": 54050
+    },
+    {
+      "epoch": 13.481296758104738,
+      "grad_norm": 10.912115097045898,
+      "learning_rate": 6.524688279301746e-06,
+      "loss": 0.3431,
+      "step": 54060
+    },
+    {
+      "epoch": 13.483790523690773,
+      "grad_norm": 7.443094730377197,
+      "learning_rate": 6.522194513715711e-06,
+      "loss": 0.3877,
+      "step": 54070
+    },
+    {
+      "epoch": 13.486284289276808,
+      "grad_norm": 6.929405689239502,
+      "learning_rate": 6.519700748129676e-06,
+      "loss": 0.2966,
+      "step": 54080
+    },
+    {
+      "epoch": 13.488778054862843,
+      "grad_norm": 6.779537200927734,
+      "learning_rate": 6.517206982543641e-06,
+      "loss": 0.2756,
+      "step": 54090
+    },
+    {
+      "epoch": 13.491271820448878,
+      "grad_norm": 13.575571060180664,
+      "learning_rate": 6.5147132169576065e-06,
+      "loss": 0.3851,
+      "step": 54100
+    },
+    {
+      "epoch": 13.493765586034913,
+      "grad_norm": 6.581482410430908,
+      "learning_rate": 6.512219451371572e-06,
+      "loss": 0.3736,
+      "step": 54110
+    },
+    {
+      "epoch": 13.496259351620948,
+      "grad_norm": 9.323427200317383,
+      "learning_rate": 6.509725685785536e-06,
+      "loss": 0.4017,
+      "step": 54120
+    },
+    {
+      "epoch": 13.498753117206983,
+      "grad_norm": 6.501451015472412,
+      "learning_rate": 6.5072319201995016e-06,
+      "loss": 0.363,
+      "step": 54130
+    },
+    {
+      "epoch": 13.501246882793017,
+      "grad_norm": 7.850208282470703,
+      "learning_rate": 6.504738154613467e-06,
+      "loss": 0.351,
+      "step": 54140
+    },
+    {
+      "epoch": 13.503740648379052,
+      "grad_norm": 7.5729780197143555,
+      "learning_rate": 6.502244389027432e-06,
+      "loss": 0.2832,
+      "step": 54150
+    },
+    {
+      "epoch": 13.506234413965087,
+      "grad_norm": 5.65486478805542,
+      "learning_rate": 6.499750623441397e-06,
+      "loss": 0.3237,
+      "step": 54160
+    },
+    {
+      "epoch": 13.508728179551122,
+      "grad_norm": 10.24453067779541,
+      "learning_rate": 6.497256857855362e-06,
+      "loss": 0.3901,
+      "step": 54170
+    },
+    {
+      "epoch": 13.511221945137157,
+      "grad_norm": 7.5691962242126465,
+      "learning_rate": 6.494763092269327e-06,
+      "loss": 0.3493,
+      "step": 54180
+    },
+    {
+      "epoch": 13.513715710723192,
+      "grad_norm": 8.992122650146484,
+      "learning_rate": 6.4922693266832926e-06,
+      "loss": 0.3426,
+      "step": 54190
+    },
+    {
+      "epoch": 13.516209476309227,
+      "grad_norm": 6.614286422729492,
+      "learning_rate": 6.489775561097258e-06,
+      "loss": 0.3462,
+      "step": 54200
+    },
+    {
+      "epoch": 13.518703241895262,
+      "grad_norm": 8.444540977478027,
+      "learning_rate": 6.487281795511222e-06,
+      "loss": 0.2959,
+      "step": 54210
+    },
+    {
+      "epoch": 13.521197007481296,
+      "grad_norm": 4.639540672302246,
+      "learning_rate": 6.484788029925188e-06,
+      "loss": 0.3116,
+      "step": 54220
+    },
+    {
+      "epoch": 13.523690773067331,
+      "grad_norm": 5.426086902618408,
+      "learning_rate": 6.482294264339153e-06,
+      "loss": 0.3411,
+      "step": 54230
+    },
+    {
+      "epoch": 13.526184538653366,
+      "grad_norm": 12.386012077331543,
+      "learning_rate": 6.479800498753118e-06,
+      "loss": 0.3102,
+      "step": 54240
+    },
+    {
+      "epoch": 13.528678304239401,
+      "grad_norm": 6.084399700164795,
+      "learning_rate": 6.477306733167083e-06,
+      "loss": 0.3842,
+      "step": 54250
+    },
+    {
+      "epoch": 13.531172069825436,
+      "grad_norm": 7.526169300079346,
+      "learning_rate": 6.474812967581048e-06,
+      "loss": 0.3568,
+      "step": 54260
+    },
+    {
+      "epoch": 13.53366583541147,
+      "grad_norm": 7.161496639251709,
+      "learning_rate": 6.472319201995013e-06,
+      "loss": 0.3225,
+      "step": 54270
+    },
+    {
+      "epoch": 13.536159600997506,
+      "grad_norm": 5.644833087921143,
+      "learning_rate": 6.469825436408979e-06,
+      "loss": 0.293,
+      "step": 54280
+    },
+    {
+      "epoch": 13.53865336658354,
+      "grad_norm": 11.303448677062988,
+      "learning_rate": 6.467331670822943e-06,
+      "loss": 0.3523,
+      "step": 54290
+    },
+    {
+      "epoch": 13.541147132169575,
+      "grad_norm": 9.144119262695312,
+      "learning_rate": 6.4648379052369084e-06,
+      "loss": 0.3661,
+      "step": 54300
+    },
+    {
+      "epoch": 13.54364089775561,
+      "grad_norm": 8.02839469909668,
+      "learning_rate": 6.462344139650874e-06,
+      "loss": 0.3792,
+      "step": 54310
+    },
+    {
+      "epoch": 13.546134663341645,
+      "grad_norm": 9.855666160583496,
+      "learning_rate": 6.459850374064839e-06,
+      "loss": 0.3324,
+      "step": 54320
+    },
+    {
+      "epoch": 13.548628428927682,
+      "grad_norm": 8.013253211975098,
+      "learning_rate": 6.457356608478803e-06,
+      "loss": 0.3991,
+      "step": 54330
+    },
+    {
+      "epoch": 13.551122194513717,
+      "grad_norm": 8.09378433227539,
+      "learning_rate": 6.454862842892769e-06,
+      "loss": 0.3573,
+      "step": 54340
+    },
+    {
+      "epoch": 13.553615960099751,
+      "grad_norm": 8.810187339782715,
+      "learning_rate": 6.452369077306734e-06,
+      "loss": 0.335,
+      "step": 54350
+    },
+    {
+      "epoch": 13.556109725685786,
+      "grad_norm": 5.002645492553711,
+      "learning_rate": 6.4498753117206994e-06,
+      "loss": 0.2775,
+      "step": 54360
+    },
+    {
+      "epoch": 13.558603491271821,
+      "grad_norm": 5.791070938110352,
+      "learning_rate": 6.447381546134663e-06,
+      "loss": 0.2979,
+      "step": 54370
+    },
+    {
+      "epoch": 13.561097256857856,
+      "grad_norm": 11.853717803955078,
+      "learning_rate": 6.444887780548628e-06,
+      "loss": 0.3168,
+      "step": 54380
+    },
+    {
+      "epoch": 13.563591022443891,
+      "grad_norm": 7.837883949279785,
+      "learning_rate": 6.442394014962594e-06,
+      "loss": 0.4237,
+      "step": 54390
+    },
+    {
+      "epoch": 13.566084788029926,
+      "grad_norm": 8.088776588439941,
+      "learning_rate": 6.43990024937656e-06,
+      "loss": 0.29,
+      "step": 54400
+    },
+    {
+      "epoch": 13.56857855361596,
+      "grad_norm": 8.3961763381958,
+      "learning_rate": 6.4374064837905235e-06,
+      "loss": 0.3884,
+      "step": 54410
+    },
+    {
+      "epoch": 13.571072319201996,
+      "grad_norm": 8.64345645904541,
+      "learning_rate": 6.434912718204489e-06,
+      "loss": 0.335,
+      "step": 54420
+    },
+    {
+      "epoch": 13.57356608478803,
+      "grad_norm": 5.859470844268799,
+      "learning_rate": 6.432418952618454e-06,
+      "loss": 0.3534,
+      "step": 54430
+    },
+    {
+      "epoch": 13.576059850374065,
+      "grad_norm": 3.3717424869537354,
+      "learning_rate": 6.429925187032419e-06,
+      "loss": 0.2729,
+      "step": 54440
+    },
+    {
+      "epoch": 13.5785536159601,
+      "grad_norm": 6.788022518157959,
+      "learning_rate": 6.427431421446385e-06,
+      "loss": 0.311,
+      "step": 54450
+    },
+    {
+      "epoch": 13.581047381546135,
+      "grad_norm": 5.033174514770508,
+      "learning_rate": 6.424937655860349e-06,
+      "loss": 0.3251,
+      "step": 54460
+    },
+    {
+      "epoch": 13.58354114713217,
+      "grad_norm": 8.036091804504395,
+      "learning_rate": 6.4224438902743145e-06,
+      "loss": 0.4072,
+      "step": 54470
+    },
+    {
+      "epoch": 13.586034912718205,
+      "grad_norm": 8.163284301757812,
+      "learning_rate": 6.41995012468828e-06,
+      "loss": 0.318,
+      "step": 54480
+    },
+    {
+      "epoch": 13.58852867830424,
+      "grad_norm": 6.626342296600342,
+      "learning_rate": 6.417456359102245e-06,
+      "loss": 0.3064,
+      "step": 54490
+    },
+    {
+      "epoch": 13.591022443890274,
+      "grad_norm": 7.7200469970703125,
+      "learning_rate": 6.4149625935162095e-06,
+      "loss": 0.3844,
+      "step": 54500
+    },
+    {
+      "epoch": 13.59351620947631,
+      "grad_norm": 8.76436710357666,
+      "learning_rate": 6.412468827930175e-06,
+      "loss": 0.3757,
+      "step": 54510
+    },
+    {
+      "epoch": 13.596009975062344,
+      "grad_norm": 9.245163917541504,
+      "learning_rate": 6.40997506234414e-06,
+      "loss": 0.3857,
+      "step": 54520
+    },
+    {
+      "epoch": 13.598503740648379,
+      "grad_norm": 6.557892322540283,
+      "learning_rate": 6.4074812967581055e-06,
+      "loss": 0.3491,
+      "step": 54530
+    },
+    {
+      "epoch": 13.600997506234414,
+      "grad_norm": 9.482400894165039,
+      "learning_rate": 6.40498753117207e-06,
+      "loss": 0.3553,
+      "step": 54540
+    },
+    {
+      "epoch": 13.603491271820449,
+      "grad_norm": 8.862476348876953,
+      "learning_rate": 6.402493765586035e-06,
+      "loss": 0.334,
+      "step": 54550
+    },
+    {
+      "epoch": 13.605985037406484,
+      "grad_norm": 6.397747039794922,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.3245,
+      "step": 54560
+    },
+    {
+      "epoch": 13.608478802992519,
+      "grad_norm": 6.339082717895508,
+      "learning_rate": 6.397506234413966e-06,
+      "loss": 0.3779,
+      "step": 54570
+    },
+    {
+      "epoch": 13.610972568578553,
+      "grad_norm": 11.248029708862305,
+      "learning_rate": 6.39501246882793e-06,
+      "loss": 0.3455,
+      "step": 54580
+    },
+    {
+      "epoch": 13.613466334164588,
+      "grad_norm": 6.5518341064453125,
+      "learning_rate": 6.392518703241896e-06,
+      "loss": 0.2906,
+      "step": 54590
+    },
+    {
+      "epoch": 13.615960099750623,
+      "grad_norm": 6.682394981384277,
+      "learning_rate": 6.390024937655861e-06,
+      "loss": 0.314,
+      "step": 54600
+    },
+    {
+      "epoch": 13.618453865336658,
+      "grad_norm": 8.466720581054688,
+      "learning_rate": 6.387531172069826e-06,
+      "loss": 0.4433,
+      "step": 54610
+    },
+    {
+      "epoch": 13.620947630922693,
+      "grad_norm": 13.29272174835205,
+      "learning_rate": 6.385037406483791e-06,
+      "loss": 0.3942,
+      "step": 54620
+    },
+    {
+      "epoch": 13.623441396508728,
+      "grad_norm": 6.229150295257568,
+      "learning_rate": 6.382543640897756e-06,
+      "loss": 0.3084,
+      "step": 54630
+    },
+    {
+      "epoch": 13.625935162094763,
+      "grad_norm": 10.2169828414917,
+      "learning_rate": 6.380049875311721e-06,
+      "loss": 0.4638,
+      "step": 54640
+    },
+    {
+      "epoch": 13.628428927680797,
+      "grad_norm": 8.496891021728516,
+      "learning_rate": 6.377556109725687e-06,
+      "loss": 0.3698,
+      "step": 54650
+    },
+    {
+      "epoch": 13.630922693266832,
+      "grad_norm": 8.761347770690918,
+      "learning_rate": 6.375062344139651e-06,
+      "loss": 0.3581,
+      "step": 54660
+    },
+    {
+      "epoch": 13.633416458852867,
+      "grad_norm": 7.339252948760986,
+      "learning_rate": 6.372568578553616e-06,
+      "loss": 0.3113,
+      "step": 54670
+    },
+    {
+      "epoch": 13.635910224438902,
+      "grad_norm": 8.492396354675293,
+      "learning_rate": 6.370074812967582e-06,
+      "loss": 0.3409,
+      "step": 54680
+    },
+    {
+      "epoch": 13.638403990024937,
+      "grad_norm": 11.395689964294434,
+      "learning_rate": 6.367581047381547e-06,
+      "loss": 0.3892,
+      "step": 54690
+    },
+    {
+      "epoch": 13.640897755610972,
+      "grad_norm": 8.986297607421875,
+      "learning_rate": 6.365087281795512e-06,
+      "loss": 0.3154,
+      "step": 54700
+    },
+    {
+      "epoch": 13.643391521197007,
+      "grad_norm": 9.709697723388672,
+      "learning_rate": 6.362593516209477e-06,
+      "loss": 0.3331,
+      "step": 54710
+    },
+    {
+      "epoch": 13.645885286783042,
+      "grad_norm": 6.466365337371826,
+      "learning_rate": 6.360099750623442e-06,
+      "loss": 0.273,
+      "step": 54720
+    },
+    {
+      "epoch": 13.648379052369076,
+      "grad_norm": 10.639359474182129,
+      "learning_rate": 6.3576059850374074e-06,
+      "loss": 0.3514,
+      "step": 54730
+    },
+    {
+      "epoch": 13.650872817955111,
+      "grad_norm": 7.763825416564941,
+      "learning_rate": 6.355112219451373e-06,
+      "loss": 0.3726,
+      "step": 54740
+    },
+    {
+      "epoch": 13.653366583541148,
+      "grad_norm": 6.4132585525512695,
+      "learning_rate": 6.352618453865337e-06,
+      "loss": 0.3819,
+      "step": 54750
+    },
+    {
+      "epoch": 13.655860349127183,
+      "grad_norm": 9.225993156433105,
+      "learning_rate": 6.3501246882793025e-06,
+      "loss": 0.32,
+      "step": 54760
+    },
+    {
+      "epoch": 13.658354114713218,
+      "grad_norm": 7.557957172393799,
+      "learning_rate": 6.347630922693268e-06,
+      "loss": 0.3545,
+      "step": 54770
+    },
+    {
+      "epoch": 13.660847880299253,
+      "grad_norm": 6.421631813049316,
+      "learning_rate": 6.345137157107233e-06,
+      "loss": 0.2719,
+      "step": 54780
+    },
+    {
+      "epoch": 13.663341645885287,
+      "grad_norm": 7.9836602210998535,
+      "learning_rate": 6.342643391521197e-06,
+      "loss": 0.3194,
+      "step": 54790
+    },
+    {
+      "epoch": 13.665835411471322,
+      "grad_norm": 7.941399097442627,
+      "learning_rate": 6.340149625935162e-06,
+      "loss": 0.2969,
+      "step": 54800
+    },
+    {
+      "epoch": 13.668329177057357,
+      "grad_norm": 6.198439121246338,
+      "learning_rate": 6.337655860349128e-06,
+      "loss": 0.3156,
+      "step": 54810
+    },
+    {
+      "epoch": 13.670822942643392,
+      "grad_norm": 8.44516658782959,
+      "learning_rate": 6.3351620947630935e-06,
+      "loss": 0.3344,
+      "step": 54820
+    },
+    {
+      "epoch": 13.673316708229427,
+      "grad_norm": 8.17326831817627,
+      "learning_rate": 6.332668329177057e-06,
+      "loss": 0.2907,
+      "step": 54830
+    },
+    {
+      "epoch": 13.675810473815462,
+      "grad_norm": 8.938278198242188,
+      "learning_rate": 6.3301745635910225e-06,
+      "loss": 0.3873,
+      "step": 54840
+    },
+    {
+      "epoch": 13.678304239401497,
+      "grad_norm": 7.996182918548584,
+      "learning_rate": 6.327680798004988e-06,
+      "loss": 0.3435,
+      "step": 54850
+    },
+    {
+      "epoch": 13.680798004987532,
+      "grad_norm": 5.846700191497803,
+      "learning_rate": 6.325187032418953e-06,
+      "loss": 0.3513,
+      "step": 54860
+    },
+    {
+      "epoch": 13.683291770573566,
+      "grad_norm": 5.51074743270874,
+      "learning_rate": 6.3226932668329175e-06,
+      "loss": 0.3079,
+      "step": 54870
+    },
+    {
+      "epoch": 13.685785536159601,
+      "grad_norm": 6.8024492263793945,
+      "learning_rate": 6.320199501246883e-06,
+      "loss": 0.306,
+      "step": 54880
+    },
+    {
+      "epoch": 13.688279301745636,
+      "grad_norm": 11.52742862701416,
+      "learning_rate": 6.317705735660848e-06,
+      "loss": 0.3317,
+      "step": 54890
+    },
+    {
+      "epoch": 13.690773067331671,
+      "grad_norm": 14.795852661132812,
+      "learning_rate": 6.3152119700748135e-06,
+      "loss": 0.3375,
+      "step": 54900
+    },
+    {
+      "epoch": 13.693266832917706,
+      "grad_norm": 12.156822204589844,
+      "learning_rate": 6.312718204488778e-06,
+      "loss": 0.3397,
+      "step": 54910
+    },
+    {
+      "epoch": 13.69576059850374,
+      "grad_norm": 9.437907218933105,
+      "learning_rate": 6.310224438902743e-06,
+      "loss": 0.365,
+      "step": 54920
+    },
+    {
+      "epoch": 13.698254364089776,
+      "grad_norm": 9.907096862792969,
+      "learning_rate": 6.3077306733167085e-06,
+      "loss": 0.3039,
+      "step": 54930
+    },
+    {
+      "epoch": 13.70074812967581,
+      "grad_norm": 8.711502075195312,
+      "learning_rate": 6.305236907730674e-06,
+      "loss": 0.3617,
+      "step": 54940
+    },
+    {
+      "epoch": 13.703241895261845,
+      "grad_norm": 7.6768622398376465,
+      "learning_rate": 6.302743142144639e-06,
+      "loss": 0.3392,
+      "step": 54950
+    },
+    {
+      "epoch": 13.70573566084788,
+      "grad_norm": 5.637898921966553,
+      "learning_rate": 6.300249376558604e-06,
+      "loss": 0.3591,
+      "step": 54960
+    },
+    {
+      "epoch": 13.708229426433915,
+      "grad_norm": 8.709186553955078,
+      "learning_rate": 6.297755610972569e-06,
+      "loss": 0.3944,
+      "step": 54970
+    },
+    {
+      "epoch": 13.71072319201995,
+      "grad_norm": 9.957995414733887,
+      "learning_rate": 6.295261845386534e-06,
+      "loss": 0.2579,
+      "step": 54980
+    },
+    {
+      "epoch": 13.713216957605985,
+      "grad_norm": 7.942922115325928,
+      "learning_rate": 6.2927680798004996e-06,
+      "loss": 0.3825,
+      "step": 54990
+    },
+    {
+      "epoch": 13.71571072319202,
+      "grad_norm": 6.098124980926514,
+      "learning_rate": 6.290274314214464e-06,
+      "loss": 0.315,
+      "step": 55000
+    },
+    {
+      "epoch": 13.718204488778055,
+      "grad_norm": 6.805457592010498,
+      "learning_rate": 6.287780548628429e-06,
+      "loss": 0.3479,
+      "step": 55010
+    },
+    {
+      "epoch": 13.72069825436409,
+      "grad_norm": 7.8078765869140625,
+      "learning_rate": 6.285286783042395e-06,
+      "loss": 0.3516,
+      "step": 55020
+    },
+    {
+      "epoch": 13.723192019950124,
+      "grad_norm": 8.491788864135742,
+      "learning_rate": 6.28279301745636e-06,
+      "loss": 0.3239,
+      "step": 55030
+    },
+    {
+      "epoch": 13.72568578553616,
+      "grad_norm": 5.934401512145996,
+      "learning_rate": 6.280299251870324e-06,
+      "loss": 0.3531,
+      "step": 55040
+    },
+    {
+      "epoch": 13.728179551122194,
+      "grad_norm": 6.76987886428833,
+      "learning_rate": 6.27780548628429e-06,
+      "loss": 0.37,
+      "step": 55050
+    },
+    {
+      "epoch": 13.730673316708229,
+      "grad_norm": 9.285640716552734,
+      "learning_rate": 6.275311720698255e-06,
+      "loss": 0.3165,
+      "step": 55060
+    },
+    {
+      "epoch": 13.733167082294264,
+      "grad_norm": 6.509362697601318,
+      "learning_rate": 6.27281795511222e-06,
+      "loss": 0.3212,
+      "step": 55070
+    },
+    {
+      "epoch": 13.735660847880299,
+      "grad_norm": 9.682168006896973,
+      "learning_rate": 6.270324189526185e-06,
+      "loss": 0.3663,
+      "step": 55080
+    },
+    {
+      "epoch": 13.738154613466333,
+      "grad_norm": 12.535362243652344,
+      "learning_rate": 6.26783042394015e-06,
+      "loss": 0.3563,
+      "step": 55090
+    },
+    {
+      "epoch": 13.740648379052368,
+      "grad_norm": 7.953882217407227,
+      "learning_rate": 6.2653366583541154e-06,
+      "loss": 0.3366,
+      "step": 55100
+    },
+    {
+      "epoch": 13.743142144638403,
+      "grad_norm": 9.631477355957031,
+      "learning_rate": 6.262842892768081e-06,
+      "loss": 0.322,
+      "step": 55110
+    },
+    {
+      "epoch": 13.745635910224438,
+      "grad_norm": 7.036581516265869,
+      "learning_rate": 6.260349127182045e-06,
+      "loss": 0.3567,
+      "step": 55120
+    },
+    {
+      "epoch": 13.748129675810475,
+      "grad_norm": 10.336029052734375,
+      "learning_rate": 6.2578553615960105e-06,
+      "loss": 0.3196,
+      "step": 55130
+    },
+    {
+      "epoch": 13.75062344139651,
+      "grad_norm": 5.572134017944336,
+      "learning_rate": 6.255361596009976e-06,
+      "loss": 0.2873,
+      "step": 55140
+    },
+    {
+      "epoch": 13.753117206982544,
+      "grad_norm": 7.690764427185059,
+      "learning_rate": 6.252867830423941e-06,
+      "loss": 0.3743,
+      "step": 55150
+    },
+    {
+      "epoch": 13.75561097256858,
+      "grad_norm": 9.607911109924316,
+      "learning_rate": 6.250374064837906e-06,
+      "loss": 0.3554,
+      "step": 55160
+    },
+    {
+      "epoch": 13.758104738154614,
+      "grad_norm": 7.156585216522217,
+      "learning_rate": 6.247880299251871e-06,
+      "loss": 0.3376,
+      "step": 55170
+    },
+    {
+      "epoch": 13.760598503740649,
+      "grad_norm": 7.8262128829956055,
+      "learning_rate": 6.245386533665836e-06,
+      "loss": 0.3767,
+      "step": 55180
+    },
+    {
+      "epoch": 13.763092269326684,
+      "grad_norm": 7.737610340118408,
+      "learning_rate": 6.2428927680798015e-06,
+      "loss": 0.3563,
+      "step": 55190
+    },
+    {
+      "epoch": 13.765586034912719,
+      "grad_norm": 8.548540115356445,
+      "learning_rate": 6.240399002493767e-06,
+      "loss": 0.3358,
+      "step": 55200
+    },
+    {
+      "epoch": 13.768079800498754,
+      "grad_norm": 5.896496295928955,
+      "learning_rate": 6.2379052369077304e-06,
+      "loss": 0.3586,
+      "step": 55210
+    },
+    {
+      "epoch": 13.770573566084789,
+      "grad_norm": 6.512773036956787,
+      "learning_rate": 6.235411471321697e-06,
+      "loss": 0.3925,
+      "step": 55220
+    },
+    {
+      "epoch": 13.773067331670823,
+      "grad_norm": 8.241216659545898,
+      "learning_rate": 6.232917705735662e-06,
+      "loss": 0.3697,
+      "step": 55230
+    },
+    {
+      "epoch": 13.775561097256858,
+      "grad_norm": 7.041107654571533,
+      "learning_rate": 6.230423940149627e-06,
+      "loss": 0.3284,
+      "step": 55240
+    },
+    {
+      "epoch": 13.778054862842893,
+      "grad_norm": 7.809773921966553,
+      "learning_rate": 6.227930174563591e-06,
+      "loss": 0.3691,
+      "step": 55250
+    },
+    {
+      "epoch": 13.780548628428928,
+      "grad_norm": 6.446750640869141,
+      "learning_rate": 6.225436408977556e-06,
+      "loss": 0.3216,
+      "step": 55260
+    },
+    {
+      "epoch": 13.783042394014963,
+      "grad_norm": 6.692153453826904,
+      "learning_rate": 6.2229426433915215e-06,
+      "loss": 0.3639,
+      "step": 55270
+    },
+    {
+      "epoch": 13.785536159600998,
+      "grad_norm": 8.99974536895752,
+      "learning_rate": 6.220448877805488e-06,
+      "loss": 0.3577,
+      "step": 55280
+    },
+    {
+      "epoch": 13.788029925187033,
+      "grad_norm": 8.162135124206543,
+      "learning_rate": 6.217955112219451e-06,
+      "loss": 0.3082,
+      "step": 55290
+    },
+    {
+      "epoch": 13.790523690773068,
+      "grad_norm": 9.059871673583984,
+      "learning_rate": 6.2154613466334165e-06,
+      "loss": 0.3538,
+      "step": 55300
+    },
+    {
+      "epoch": 13.793017456359102,
+      "grad_norm": 8.866132736206055,
+      "learning_rate": 6.212967581047382e-06,
+      "loss": 0.3224,
+      "step": 55310
+    },
+    {
+      "epoch": 13.795511221945137,
+      "grad_norm": 6.690679550170898,
+      "learning_rate": 6.210473815461347e-06,
+      "loss": 0.3352,
+      "step": 55320
+    },
+    {
+      "epoch": 13.798004987531172,
+      "grad_norm": 7.613649368286133,
+      "learning_rate": 6.207980049875312e-06,
+      "loss": 0.3311,
+      "step": 55330
+    },
+    {
+      "epoch": 13.800498753117207,
+      "grad_norm": 5.005457878112793,
+      "learning_rate": 6.205486284289277e-06,
+      "loss": 0.3305,
+      "step": 55340
+    },
+    {
+      "epoch": 13.802992518703242,
+      "grad_norm": 5.639405250549316,
+      "learning_rate": 6.202992518703242e-06,
+      "loss": 0.3504,
+      "step": 55350
+    },
+    {
+      "epoch": 13.805486284289277,
+      "grad_norm": 4.40685510635376,
+      "learning_rate": 6.2004987531172075e-06,
+      "loss": 0.3078,
+      "step": 55360
+    },
+    {
+      "epoch": 13.807980049875312,
+      "grad_norm": 6.308961868286133,
+      "learning_rate": 6.198004987531172e-06,
+      "loss": 0.3049,
+      "step": 55370
+    },
+    {
+      "epoch": 13.810473815461346,
+      "grad_norm": 6.514129638671875,
+      "learning_rate": 6.195511221945137e-06,
+      "loss": 0.3479,
+      "step": 55380
+    },
+    {
+      "epoch": 13.812967581047381,
+      "grad_norm": 6.240500450134277,
+      "learning_rate": 6.193017456359103e-06,
+      "loss": 0.3622,
+      "step": 55390
+    },
+    {
+      "epoch": 13.815461346633416,
+      "grad_norm": 6.060596466064453,
+      "learning_rate": 6.190523690773068e-06,
+      "loss": 0.3123,
+      "step": 55400
+    },
+    {
+      "epoch": 13.817955112219451,
+      "grad_norm": 8.058761596679688,
+      "learning_rate": 6.188029925187032e-06,
+      "loss": 0.3519,
+      "step": 55410
+    },
+    {
+      "epoch": 13.820448877805486,
+      "grad_norm": 5.577200889587402,
+      "learning_rate": 6.185536159600998e-06,
+      "loss": 0.3323,
+      "step": 55420
+    },
+    {
+      "epoch": 13.82294264339152,
+      "grad_norm": 8.922237396240234,
+      "learning_rate": 6.183042394014963e-06,
+      "loss": 0.3886,
+      "step": 55430
+    },
+    {
+      "epoch": 13.825436408977556,
+      "grad_norm": 10.180780410766602,
+      "learning_rate": 6.180548628428928e-06,
+      "loss": 0.3128,
+      "step": 55440
+    },
+    {
+      "epoch": 13.82793017456359,
+      "grad_norm": 6.981229305267334,
+      "learning_rate": 6.178054862842894e-06,
+      "loss": 0.2746,
+      "step": 55450
+    },
+    {
+      "epoch": 13.830423940149625,
+      "grad_norm": 5.601588726043701,
+      "learning_rate": 6.175561097256858e-06,
+      "loss": 0.2593,
+      "step": 55460
+    },
+    {
+      "epoch": 13.83291770573566,
+      "grad_norm": 9.501214027404785,
+      "learning_rate": 6.173067331670823e-06,
+      "loss": 0.3332,
+      "step": 55470
+    },
+    {
+      "epoch": 13.835411471321695,
+      "grad_norm": 10.897354125976562,
+      "learning_rate": 6.170573566084789e-06,
+      "loss": 0.3535,
+      "step": 55480
+    },
+    {
+      "epoch": 13.83790523690773,
+      "grad_norm": 9.307845115661621,
+      "learning_rate": 6.168079800498754e-06,
+      "loss": 0.3206,
+      "step": 55490
+    },
+    {
+      "epoch": 13.840399002493765,
+      "grad_norm": 5.495110511779785,
+      "learning_rate": 6.1655860349127185e-06,
+      "loss": 0.3112,
+      "step": 55500
+    },
+    {
+      "epoch": 13.8428927680798,
+      "grad_norm": 9.099871635437012,
+      "learning_rate": 6.163092269326684e-06,
+      "loss": 0.3681,
+      "step": 55510
+    },
+    {
+      "epoch": 13.845386533665835,
+      "grad_norm": 7.402480602264404,
+      "learning_rate": 6.160598503740649e-06,
+      "loss": 0.3062,
+      "step": 55520
+    },
+    {
+      "epoch": 13.84788029925187,
+      "grad_norm": 6.1998982429504395,
+      "learning_rate": 6.1581047381546144e-06,
+      "loss": 0.3698,
+      "step": 55530
+    },
+    {
+      "epoch": 13.850374064837904,
+      "grad_norm": 5.2379021644592285,
+      "learning_rate": 6.155610972568579e-06,
+      "loss": 0.319,
+      "step": 55540
+    },
+    {
+      "epoch": 13.85286783042394,
+      "grad_norm": 7.226074695587158,
+      "learning_rate": 6.153117206982544e-06,
+      "loss": 0.35,
+      "step": 55550
+    },
+    {
+      "epoch": 13.855361596009976,
+      "grad_norm": 8.524081230163574,
+      "learning_rate": 6.1506234413965095e-06,
+      "loss": 0.3067,
+      "step": 55560
+    },
+    {
+      "epoch": 13.85785536159601,
+      "grad_norm": 7.00763463973999,
+      "learning_rate": 6.148129675810475e-06,
+      "loss": 0.3146,
+      "step": 55570
+    },
+    {
+      "epoch": 13.860349127182046,
+      "grad_norm": 10.624029159545898,
+      "learning_rate": 6.145635910224439e-06,
+      "loss": 0.3798,
+      "step": 55580
+    },
+    {
+      "epoch": 13.86284289276808,
+      "grad_norm": 8.03925609588623,
+      "learning_rate": 6.143142144638405e-06,
+      "loss": 0.3862,
+      "step": 55590
+    },
+    {
+      "epoch": 13.865336658354115,
+      "grad_norm": 12.083026885986328,
+      "learning_rate": 6.14064837905237e-06,
+      "loss": 0.3485,
+      "step": 55600
+    },
+    {
+      "epoch": 13.86783042394015,
+      "grad_norm": 7.343206405639648,
+      "learning_rate": 6.138154613466335e-06,
+      "loss": 0.3835,
+      "step": 55610
+    },
+    {
+      "epoch": 13.870324189526185,
+      "grad_norm": 11.571039199829102,
+      "learning_rate": 6.1356608478803e-06,
+      "loss": 0.3813,
+      "step": 55620
+    },
+    {
+      "epoch": 13.87281795511222,
+      "grad_norm": 5.217142581939697,
+      "learning_rate": 6.133167082294265e-06,
+      "loss": 0.3064,
+      "step": 55630
+    },
+    {
+      "epoch": 13.875311720698255,
+      "grad_norm": 7.893732070922852,
+      "learning_rate": 6.13067331670823e-06,
+      "loss": 0.3811,
+      "step": 55640
+    },
+    {
+      "epoch": 13.87780548628429,
+      "grad_norm": 7.880045413970947,
+      "learning_rate": 6.128179551122196e-06,
+      "loss": 0.3338,
+      "step": 55650
+    },
+    {
+      "epoch": 13.880299251870325,
+      "grad_norm": 10.907844543457031,
+      "learning_rate": 6.125685785536159e-06,
+      "loss": 0.4448,
+      "step": 55660
+    },
+    {
+      "epoch": 13.88279301745636,
+      "grad_norm": 6.203114032745361,
+      "learning_rate": 6.1231920199501245e-06,
+      "loss": 0.3407,
+      "step": 55670
+    },
+    {
+      "epoch": 13.885286783042394,
+      "grad_norm": 9.57861328125,
+      "learning_rate": 6.120698254364091e-06,
+      "loss": 0.3475,
+      "step": 55680
+    },
+    {
+      "epoch": 13.88778054862843,
+      "grad_norm": 9.820565223693848,
+      "learning_rate": 6.118204488778056e-06,
+      "loss": 0.3151,
+      "step": 55690
+    },
+    {
+      "epoch": 13.890274314214464,
+      "grad_norm": 7.9553046226501465,
+      "learning_rate": 6.115710723192021e-06,
+      "loss": 0.3343,
+      "step": 55700
+    },
+    {
+      "epoch": 13.892768079800499,
+      "grad_norm": 7.718328952789307,
+      "learning_rate": 6.113216957605985e-06,
+      "loss": 0.3281,
+      "step": 55710
+    },
+    {
+      "epoch": 13.895261845386534,
+      "grad_norm": 9.396270751953125,
+      "learning_rate": 6.11072319201995e-06,
+      "loss": 0.3393,
+      "step": 55720
+    },
+    {
+      "epoch": 13.897755610972569,
+      "grad_norm": 6.914912223815918,
+      "learning_rate": 6.1082294264339155e-06,
+      "loss": 0.3104,
+      "step": 55730
+    },
+    {
+      "epoch": 13.900249376558603,
+      "grad_norm": 7.916605472564697,
+      "learning_rate": 6.105735660847882e-06,
+      "loss": 0.3162,
+      "step": 55740
+    },
+    {
+      "epoch": 13.902743142144638,
+      "grad_norm": 7.331538200378418,
+      "learning_rate": 6.103241895261845e-06,
+      "loss": 0.3076,
+      "step": 55750
+    },
+    {
+      "epoch": 13.905236907730673,
+      "grad_norm": 6.395208835601807,
+      "learning_rate": 6.100748129675811e-06,
+      "loss": 0.2805,
+      "step": 55760
+    },
+    {
+      "epoch": 13.907730673316708,
+      "grad_norm": 9.674216270446777,
+      "learning_rate": 6.098254364089776e-06,
+      "loss": 0.303,
+      "step": 55770
+    },
+    {
+      "epoch": 13.910224438902743,
+      "grad_norm": 8.970970153808594,
+      "learning_rate": 6.095760598503741e-06,
+      "loss": 0.3443,
+      "step": 55780
+    },
+    {
+      "epoch": 13.912718204488778,
+      "grad_norm": 7.25998592376709,
+      "learning_rate": 6.093266832917706e-06,
+      "loss": 0.2639,
+      "step": 55790
+    },
+    {
+      "epoch": 13.915211970074813,
+      "grad_norm": 10.04069709777832,
+      "learning_rate": 6.090773067331671e-06,
+      "loss": 0.361,
+      "step": 55800
+    },
+    {
+      "epoch": 13.917705735660848,
+      "grad_norm": 10.420782089233398,
+      "learning_rate": 6.088279301745636e-06,
+      "loss": 0.3985,
+      "step": 55810
+    },
+    {
+      "epoch": 13.920199501246882,
+      "grad_norm": 7.295711040496826,
+      "learning_rate": 6.085785536159602e-06,
+      "loss": 0.2836,
+      "step": 55820
+    },
+    {
+      "epoch": 13.922693266832917,
+      "grad_norm": 10.063495635986328,
+      "learning_rate": 6.083291770573566e-06,
+      "loss": 0.3421,
+      "step": 55830
+    },
+    {
+      "epoch": 13.925187032418952,
+      "grad_norm": 11.42710018157959,
+      "learning_rate": 6.080798004987531e-06,
+      "loss": 0.3341,
+      "step": 55840
+    },
+    {
+      "epoch": 13.927680798004987,
+      "grad_norm": 5.669824123382568,
+      "learning_rate": 6.078304239401497e-06,
+      "loss": 0.3636,
+      "step": 55850
+    },
+    {
+      "epoch": 13.930174563591022,
+      "grad_norm": 9.360200881958008,
+      "learning_rate": 6.075810473815462e-06,
+      "loss": 0.3751,
+      "step": 55860
+    },
+    {
+      "epoch": 13.932668329177057,
+      "grad_norm": 8.900922775268555,
+      "learning_rate": 6.0733167082294265e-06,
+      "loss": 0.3607,
+      "step": 55870
+    },
+    {
+      "epoch": 13.935162094763092,
+      "grad_norm": 5.942974090576172,
+      "learning_rate": 6.070822942643392e-06,
+      "loss": 0.2925,
+      "step": 55880
+    },
+    {
+      "epoch": 13.937655860349127,
+      "grad_norm": 8.712531089782715,
+      "learning_rate": 6.068329177057357e-06,
+      "loss": 0.3076,
+      "step": 55890
+    },
+    {
+      "epoch": 13.940149625935161,
+      "grad_norm": 5.564552307128906,
+      "learning_rate": 6.065835411471322e-06,
+      "loss": 0.347,
+      "step": 55900
+    },
+    {
+      "epoch": 13.942643391521196,
+      "grad_norm": 11.01707935333252,
+      "learning_rate": 6.063341645885287e-06,
+      "loss": 0.3483,
+      "step": 55910
+    },
+    {
+      "epoch": 13.945137157107231,
+      "grad_norm": 6.945023536682129,
+      "learning_rate": 6.060847880299252e-06,
+      "loss": 0.3604,
+      "step": 55920
+    },
+    {
+      "epoch": 13.947630922693268,
+      "grad_norm": 6.796305179595947,
+      "learning_rate": 6.0583541147132175e-06,
+      "loss": 0.3503,
+      "step": 55930
+    },
+    {
+      "epoch": 13.950124688279303,
+      "grad_norm": 5.797430515289307,
+      "learning_rate": 6.055860349127183e-06,
+      "loss": 0.4618,
+      "step": 55940
+    },
+    {
+      "epoch": 13.952618453865338,
+      "grad_norm": 8.793232917785645,
+      "learning_rate": 6.053615960099751e-06,
+      "loss": 0.3292,
+      "step": 55950
+    },
+    {
+      "epoch": 13.955112219451372,
+      "grad_norm": 7.438122272491455,
+      "learning_rate": 6.051122194513716e-06,
+      "loss": 0.2957,
+      "step": 55960
+    },
+    {
+      "epoch": 13.957605985037407,
+      "grad_norm": 9.06357192993164,
+      "learning_rate": 6.048628428927681e-06,
+      "loss": 0.392,
+      "step": 55970
+    },
+    {
+      "epoch": 13.960099750623442,
+      "grad_norm": 11.190245628356934,
+      "learning_rate": 6.046134663341646e-06,
+      "loss": 0.3157,
+      "step": 55980
+    },
+    {
+      "epoch": 13.962593516209477,
+      "grad_norm": 11.194109916687012,
+      "learning_rate": 6.0436408977556115e-06,
+      "loss": 0.333,
+      "step": 55990
+    },
+    {
+      "epoch": 13.965087281795512,
+      "grad_norm": 10.485845565795898,
+      "learning_rate": 6.041147132169577e-06,
+      "loss": 0.3388,
+      "step": 56000
+    },
+    {
+      "epoch": 13.967581047381547,
+      "grad_norm": 9.132210731506348,
+      "learning_rate": 6.038653366583541e-06,
+      "loss": 0.3132,
+      "step": 56010
+    },
+    {
+      "epoch": 13.970074812967582,
+      "grad_norm": 9.017107963562012,
+      "learning_rate": 6.0361596009975065e-06,
+      "loss": 0.3127,
+      "step": 56020
+    },
+    {
+      "epoch": 13.972568578553616,
+      "grad_norm": 8.930654525756836,
+      "learning_rate": 6.033665835411472e-06,
+      "loss": 0.3004,
+      "step": 56030
+    },
+    {
+      "epoch": 13.975062344139651,
+      "grad_norm": 14.12149429321289,
+      "learning_rate": 6.031172069825437e-06,
+      "loss": 0.413,
+      "step": 56040
+    },
+    {
+      "epoch": 13.977556109725686,
+      "grad_norm": 5.871008396148682,
+      "learning_rate": 6.028678304239402e-06,
+      "loss": 0.2982,
+      "step": 56050
+    },
+    {
+      "epoch": 13.980049875311721,
+      "grad_norm": 6.392510890960693,
+      "learning_rate": 6.026184538653367e-06,
+      "loss": 0.3514,
+      "step": 56060
+    },
+    {
+      "epoch": 13.982543640897756,
+      "grad_norm": 10.883293151855469,
+      "learning_rate": 6.023690773067332e-06,
+      "loss": 0.3422,
+      "step": 56070
+    },
+    {
+      "epoch": 13.98503740648379,
+      "grad_norm": 5.680593013763428,
+      "learning_rate": 6.0211970074812976e-06,
+      "loss": 0.3912,
+      "step": 56080
+    },
+    {
+      "epoch": 13.987531172069826,
+      "grad_norm": 11.184502601623535,
+      "learning_rate": 6.018703241895262e-06,
+      "loss": 0.3154,
+      "step": 56090
+    },
+    {
+      "epoch": 13.99002493765586,
+      "grad_norm": 8.648889541625977,
+      "learning_rate": 6.016209476309227e-06,
+      "loss": 0.3405,
+      "step": 56100
+    },
+    {
+      "epoch": 13.992518703241895,
+      "grad_norm": 9.194254875183105,
+      "learning_rate": 6.013715710723193e-06,
+      "loss": 0.3542,
+      "step": 56110
+    },
+    {
+      "epoch": 13.99501246882793,
+      "grad_norm": 7.495959281921387,
+      "learning_rate": 6.011221945137158e-06,
+      "loss": 0.3882,
+      "step": 56120
+    },
+    {
+      "epoch": 13.997506234413965,
+      "grad_norm": 10.184008598327637,
+      "learning_rate": 6.008728179551123e-06,
+      "loss": 0.332,
+      "step": 56130
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 7.123004913330078,
+      "learning_rate": 6.006234413965088e-06,
+      "loss": 0.3622,
+      "step": 56140
+    },
+    {
+      "epoch": 14.0,
+      "eval_loss": 0.41418567299842834,
+      "eval_runtime": 60.1065,
+      "eval_samples_per_second": 16.687,
+      "eval_steps_per_second": 16.687,
+      "step": 56140
+    },
+    {
+      "epoch": 14.002493765586035,
+      "grad_norm": 7.25279426574707,
+      "learning_rate": 6.003740648379053e-06,
+      "loss": 0.3301,
+      "step": 56150
+    },
+    {
+      "epoch": 14.00498753117207,
+      "grad_norm": 5.500768661499023,
+      "learning_rate": 6.001246882793018e-06,
+      "loss": 0.3115,
+      "step": 56160
+    },
+    {
+      "epoch": 14.007481296758105,
+      "grad_norm": 6.643815994262695,
+      "learning_rate": 5.998753117206984e-06,
+      "loss": 0.2826,
+      "step": 56170
+    },
+    {
+      "epoch": 14.00997506234414,
+      "grad_norm": 10.706582069396973,
+      "learning_rate": 5.996259351620948e-06,
+      "loss": 0.3527,
+      "step": 56180
+    },
+    {
+      "epoch": 14.012468827930174,
+      "grad_norm": 15.78691291809082,
+      "learning_rate": 5.9937655860349134e-06,
+      "loss": 0.3879,
+      "step": 56190
+    },
+    {
+      "epoch": 14.01496259351621,
+      "grad_norm": 7.744325637817383,
+      "learning_rate": 5.991271820448879e-06,
+      "loss": 0.2975,
+      "step": 56200
+    },
+    {
+      "epoch": 14.017456359102244,
+      "grad_norm": 6.840442657470703,
+      "learning_rate": 5.988778054862844e-06,
+      "loss": 0.3212,
+      "step": 56210
+    },
+    {
+      "epoch": 14.019950124688279,
+      "grad_norm": 6.66912841796875,
+      "learning_rate": 5.986284289276808e-06,
+      "loss": 0.3231,
+      "step": 56220
+    },
+    {
+      "epoch": 14.022443890274314,
+      "grad_norm": 11.040656089782715,
+      "learning_rate": 5.983790523690774e-06,
+      "loss": 0.3638,
+      "step": 56230
+    },
+    {
+      "epoch": 14.024937655860349,
+      "grad_norm": 12.657490730285645,
+      "learning_rate": 5.981296758104739e-06,
+      "loss": 0.3794,
+      "step": 56240
+    },
+    {
+      "epoch": 14.027431421446384,
+      "grad_norm": 7.612829208374023,
+      "learning_rate": 5.9788029925187044e-06,
+      "loss": 0.379,
+      "step": 56250
+    },
+    {
+      "epoch": 14.029925187032418,
+      "grad_norm": 6.018895149230957,
+      "learning_rate": 5.976309226932668e-06,
+      "loss": 0.3329,
+      "step": 56260
+    },
+    {
+      "epoch": 14.032418952618453,
+      "grad_norm": 9.050536155700684,
+      "learning_rate": 5.973815461346633e-06,
+      "loss": 0.3569,
+      "step": 56270
+    },
+    {
+      "epoch": 14.034912718204488,
+      "grad_norm": 5.334194660186768,
+      "learning_rate": 5.971321695760599e-06,
+      "loss": 0.3331,
+      "step": 56280
+    },
+    {
+      "epoch": 14.037406483790523,
+      "grad_norm": 7.58427095413208,
+      "learning_rate": 5.968827930174565e-06,
+      "loss": 0.2786,
+      "step": 56290
+    },
+    {
+      "epoch": 14.039900249376558,
+      "grad_norm": 10.09249210357666,
+      "learning_rate": 5.9663341645885284e-06,
+      "loss": 0.3011,
+      "step": 56300
+    },
+    {
+      "epoch": 14.042394014962593,
+      "grad_norm": 12.218475341796875,
+      "learning_rate": 5.963840399002494e-06,
+      "loss": 0.3299,
+      "step": 56310
+    },
+    {
+      "epoch": 14.044887780548628,
+      "grad_norm": 6.630500316619873,
+      "learning_rate": 5.961346633416459e-06,
+      "loss": 0.3609,
+      "step": 56320
+    },
+    {
+      "epoch": 14.047381546134662,
+      "grad_norm": 5.265425205230713,
+      "learning_rate": 5.958852867830424e-06,
+      "loss": 0.2524,
+      "step": 56330
+    },
+    {
+      "epoch": 14.049875311720697,
+      "grad_norm": 10.030232429504395,
+      "learning_rate": 5.95635910224439e-06,
+      "loss": 0.3,
+      "step": 56340
+    },
+    {
+      "epoch": 14.052369077306734,
+      "grad_norm": 8.025049209594727,
+      "learning_rate": 5.953865336658354e-06,
+      "loss": 0.3647,
+      "step": 56350
+    },
+    {
+      "epoch": 14.054862842892769,
+      "grad_norm": 8.812301635742188,
+      "learning_rate": 5.9513715710723195e-06,
+      "loss": 0.4328,
+      "step": 56360
+    },
+    {
+      "epoch": 14.057356608478804,
+      "grad_norm": 8.302797317504883,
+      "learning_rate": 5.948877805486285e-06,
+      "loss": 0.3217,
+      "step": 56370
+    },
+    {
+      "epoch": 14.059850374064839,
+      "grad_norm": 7.739553928375244,
+      "learning_rate": 5.94638403990025e-06,
+      "loss": 0.352,
+      "step": 56380
+    },
+    {
+      "epoch": 14.062344139650873,
+      "grad_norm": 6.211906433105469,
+      "learning_rate": 5.9438902743142145e-06,
+      "loss": 0.3744,
+      "step": 56390
+    },
+    {
+      "epoch": 14.064837905236908,
+      "grad_norm": 9.784815788269043,
+      "learning_rate": 5.94139650872818e-06,
+      "loss": 0.3688,
+      "step": 56400
+    },
+    {
+      "epoch": 14.067331670822943,
+      "grad_norm": 7.506394386291504,
+      "learning_rate": 5.938902743142145e-06,
+      "loss": 0.3215,
+      "step": 56410
+    },
+    {
+      "epoch": 14.069825436408978,
+      "grad_norm": 9.98291301727295,
+      "learning_rate": 5.9364089775561105e-06,
+      "loss": 0.3251,
+      "step": 56420
+    },
+    {
+      "epoch": 14.072319201995013,
+      "grad_norm": NaN,
+      "learning_rate": 5.934164588528679e-06,
+      "loss": 0.35,
+      "step": 56430
+    },
+    {
+      "epoch": 14.074812967581048,
+      "grad_norm": 7.1763434410095215,
+      "learning_rate": 5.931670822942643e-06,
+      "loss": 0.3382,
+      "step": 56440
+    },
+    {
+      "epoch": 14.077306733167083,
+      "grad_norm": 11.202238082885742,
+      "learning_rate": 5.9291770573566085e-06,
+      "loss": 0.3186,
+      "step": 56450
+    },
+    {
+      "epoch": 14.079800498753118,
+      "grad_norm": 7.892131805419922,
+      "learning_rate": 5.926683291770574e-06,
+      "loss": 0.3167,
+      "step": 56460
+    },
+    {
+      "epoch": 14.082294264339152,
+      "grad_norm": 7.04252815246582,
+      "learning_rate": 5.924189526184539e-06,
+      "loss": 0.3132,
+      "step": 56470
+    },
+    {
+      "epoch": 14.084788029925187,
+      "grad_norm": 10.996591567993164,
+      "learning_rate": 5.921695760598504e-06,
+      "loss": 0.3149,
+      "step": 56480
+    },
+    {
+      "epoch": 14.087281795511222,
+      "grad_norm": 6.079100608825684,
+      "learning_rate": 5.919201995012469e-06,
+      "loss": 0.3049,
+      "step": 56490
+    },
+    {
+      "epoch": 14.089775561097257,
+      "grad_norm": 7.392295837402344,
+      "learning_rate": 5.916708229426434e-06,
+      "loss": 0.3318,
+      "step": 56500
+    },
+    {
+      "epoch": 14.092269326683292,
+      "grad_norm": 5.498250484466553,
+      "learning_rate": 5.9142144638403995e-06,
+      "loss": 0.2813,
+      "step": 56510
+    },
+    {
+      "epoch": 14.094763092269327,
+      "grad_norm": 8.986740112304688,
+      "learning_rate": 5.911720698254365e-06,
+      "loss": 0.285,
+      "step": 56520
+    },
+    {
+      "epoch": 14.097256857855362,
+      "grad_norm": 8.847763061523438,
+      "learning_rate": 5.909226932668329e-06,
+      "loss": 0.3253,
+      "step": 56530
+    },
+    {
+      "epoch": 14.099750623441397,
+      "grad_norm": 10.011739730834961,
+      "learning_rate": 5.906733167082295e-06,
+      "loss": 0.3322,
+      "step": 56540
+    },
+    {
+      "epoch": 14.102244389027431,
+      "grad_norm": 7.457201957702637,
+      "learning_rate": 5.90423940149626e-06,
+      "loss": 0.3628,
+      "step": 56550
+    },
+    {
+      "epoch": 14.104738154613466,
+      "grad_norm": 5.719525337219238,
+      "learning_rate": 5.901745635910225e-06,
+      "loss": 0.3435,
+      "step": 56560
+    },
+    {
+      "epoch": 14.107231920199501,
+      "grad_norm": 11.263422012329102,
+      "learning_rate": 5.89925187032419e-06,
+      "loss": 0.3875,
+      "step": 56570
+    },
+    {
+      "epoch": 14.109725685785536,
+      "grad_norm": 9.578998565673828,
+      "learning_rate": 5.896758104738155e-06,
+      "loss": 0.3096,
+      "step": 56580
+    },
+    {
+      "epoch": 14.11221945137157,
+      "grad_norm": 7.107484817504883,
+      "learning_rate": 5.89426433915212e-06,
+      "loss": 0.3913,
+      "step": 56590
+    },
+    {
+      "epoch": 14.114713216957606,
+      "grad_norm": 10.798450469970703,
+      "learning_rate": 5.891770573566086e-06,
+      "loss": 0.2936,
+      "step": 56600
+    },
+    {
+      "epoch": 14.11720698254364,
+      "grad_norm": 7.937857627868652,
+      "learning_rate": 5.88927680798005e-06,
+      "loss": 0.2655,
+      "step": 56610
+    },
+    {
+      "epoch": 14.119700748129675,
+      "grad_norm": 5.9058613777160645,
+      "learning_rate": 5.886783042394015e-06,
+      "loss": 0.3056,
+      "step": 56620
+    },
+    {
+      "epoch": 14.12219451371571,
+      "grad_norm": 5.481210231781006,
+      "learning_rate": 5.884289276807981e-06,
+      "loss": 0.378,
+      "step": 56630
+    },
+    {
+      "epoch": 14.124688279301745,
+      "grad_norm": 8.08363151550293,
+      "learning_rate": 5.881795511221946e-06,
+      "loss": 0.336,
+      "step": 56640
+    },
+    {
+      "epoch": 14.12718204488778,
+      "grad_norm": 10.191091537475586,
+      "learning_rate": 5.8793017456359105e-06,
+      "loss": 0.3545,
+      "step": 56650
+    },
+    {
+      "epoch": 14.129675810473815,
+      "grad_norm": 8.802706718444824,
+      "learning_rate": 5.876807980049876e-06,
+      "loss": 0.3284,
+      "step": 56660
+    },
+    {
+      "epoch": 14.13216957605985,
+      "grad_norm": 6.7587890625,
+      "learning_rate": 5.874314214463841e-06,
+      "loss": 0.3659,
+      "step": 56670
+    },
+    {
+      "epoch": 14.134663341645885,
+      "grad_norm": 6.1083526611328125,
+      "learning_rate": 5.871820448877806e-06,
+      "loss": 0.3007,
+      "step": 56680
+    },
+    {
+      "epoch": 14.13715710723192,
+      "grad_norm": 10.109125137329102,
+      "learning_rate": 5.869326683291771e-06,
+      "loss": 0.3703,
+      "step": 56690
+    },
+    {
+      "epoch": 14.139650872817954,
+      "grad_norm": 9.299257278442383,
+      "learning_rate": 5.866832917705736e-06,
+      "loss": 0.3109,
+      "step": 56700
+    },
+    {
+      "epoch": 14.14214463840399,
+      "grad_norm": 11.763932228088379,
+      "learning_rate": 5.8643391521197015e-06,
+      "loss": 0.4337,
+      "step": 56710
+    },
+    {
+      "epoch": 14.144638403990024,
+      "grad_norm": 9.687064170837402,
+      "learning_rate": 5.861845386533667e-06,
+      "loss": 0.3345,
+      "step": 56720
+    },
+    {
+      "epoch": 14.147132169576059,
+      "grad_norm": 7.107930660247803,
+      "learning_rate": 5.859351620947631e-06,
+      "loss": 0.3442,
+      "step": 56730
+    },
+    {
+      "epoch": 14.149625935162096,
+      "grad_norm": 8.584643363952637,
+      "learning_rate": 5.8568578553615966e-06,
+      "loss": 0.2821,
+      "step": 56740
+    },
+    {
+      "epoch": 14.15211970074813,
+      "grad_norm": 10.098069190979004,
+      "learning_rate": 5.854364089775562e-06,
+      "loss": 0.3944,
+      "step": 56750
+    },
+    {
+      "epoch": 14.154613466334165,
+      "grad_norm": 9.100006103515625,
+      "learning_rate": 5.851870324189527e-06,
+      "loss": 0.2938,
+      "step": 56760
+    },
+    {
+      "epoch": 14.1571072319202,
+      "grad_norm": 12.640045166015625,
+      "learning_rate": 5.8493765586034925e-06,
+      "loss": 0.3651,
+      "step": 56770
+    },
+    {
+      "epoch": 14.159600997506235,
+      "grad_norm": 8.430350303649902,
+      "learning_rate": 5.846882793017457e-06,
+      "loss": 0.2942,
+      "step": 56780
+    },
+    {
+      "epoch": 14.16209476309227,
+      "grad_norm": 7.6195855140686035,
+      "learning_rate": 5.844389027431422e-06,
+      "loss": 0.3382,
+      "step": 56790
+    },
+    {
+      "epoch": 14.164588528678305,
+      "grad_norm": 10.6289701461792,
+      "learning_rate": 5.8418952618453876e-06,
+      "loss": 0.3124,
+      "step": 56800
+    },
+    {
+      "epoch": 14.16708229426434,
+      "grad_norm": 10.420262336730957,
+      "learning_rate": 5.839401496259353e-06,
+      "loss": 0.3315,
+      "step": 56810
+    },
+    {
+      "epoch": 14.169576059850375,
+      "grad_norm": 8.555750846862793,
+      "learning_rate": 5.8369077306733165e-06,
+      "loss": 0.3292,
+      "step": 56820
+    },
+    {
+      "epoch": 14.17206982543641,
+      "grad_norm": 10.164884567260742,
+      "learning_rate": 5.834413965087282e-06,
+      "loss": 0.3674,
+      "step": 56830
+    },
+    {
+      "epoch": 14.174563591022444,
+      "grad_norm": 6.1933746337890625,
+      "learning_rate": 5.831920199501248e-06,
+      "loss": 0.3447,
+      "step": 56840
+    },
+    {
+      "epoch": 14.17705735660848,
+      "grad_norm": 6.741939544677734,
+      "learning_rate": 5.829426433915213e-06,
+      "loss": 0.3415,
+      "step": 56850
+    },
+    {
+      "epoch": 14.179551122194514,
+      "grad_norm": 9.930939674377441,
+      "learning_rate": 5.826932668329177e-06,
+      "loss": 0.3466,
+      "step": 56860
+    },
+    {
+      "epoch": 14.182044887780549,
+      "grad_norm": 6.0956854820251465,
+      "learning_rate": 5.824438902743142e-06,
+      "loss": 0.3609,
+      "step": 56870
+    },
+    {
+      "epoch": 14.184538653366584,
+      "grad_norm": 9.120026588439941,
+      "learning_rate": 5.8219451371571075e-06,
+      "loss": 0.3623,
+      "step": 56880
+    },
+    {
+      "epoch": 14.187032418952619,
+      "grad_norm": 5.484027862548828,
+      "learning_rate": 5.819451371571073e-06,
+      "loss": 0.334,
+      "step": 56890
+    },
+    {
+      "epoch": 14.189526184538654,
+      "grad_norm": 10.343612670898438,
+      "learning_rate": 5.816957605985037e-06,
+      "loss": 0.3805,
+      "step": 56900
+    },
+    {
+      "epoch": 14.192019950124688,
+      "grad_norm": 11.961114883422852,
+      "learning_rate": 5.814463840399003e-06,
+      "loss": 0.4391,
+      "step": 56910
+    },
+    {
+      "epoch": 14.194513715710723,
+      "grad_norm": 8.149081230163574,
+      "learning_rate": 5.811970074812968e-06,
+      "loss": 0.3533,
+      "step": 56920
+    },
+    {
+      "epoch": 14.197007481296758,
+      "grad_norm": 11.395288467407227,
+      "learning_rate": 5.809476309226933e-06,
+      "loss": 0.3897,
+      "step": 56930
+    },
+    {
+      "epoch": 14.199501246882793,
+      "grad_norm": 8.990917205810547,
+      "learning_rate": 5.806982543640898e-06,
+      "loss": 0.2918,
+      "step": 56940
+    },
+    {
+      "epoch": 14.201995012468828,
+      "grad_norm": 9.0519380569458,
+      "learning_rate": 5.804488778054863e-06,
+      "loss": 0.326,
+      "step": 56950
+    },
+    {
+      "epoch": 14.204488778054863,
+      "grad_norm": 5.42004919052124,
+      "learning_rate": 5.801995012468828e-06,
+      "loss": 0.3336,
+      "step": 56960
+    },
+    {
+      "epoch": 14.206982543640898,
+      "grad_norm": 24.50922966003418,
+      "learning_rate": 5.799501246882794e-06,
+      "loss": 0.3462,
+      "step": 56970
+    },
+    {
+      "epoch": 14.209476309226932,
+      "grad_norm": 7.2156877517700195,
+      "learning_rate": 5.797007481296758e-06,
+      "loss": 0.299,
+      "step": 56980
+    },
+    {
+      "epoch": 14.211970074812967,
+      "grad_norm": 8.618878364562988,
+      "learning_rate": 5.794513715710723e-06,
+      "loss": 0.3368,
+      "step": 56990
+    },
+    {
+      "epoch": 14.214463840399002,
+      "grad_norm": 8.678366661071777,
+      "learning_rate": 5.792019950124689e-06,
+      "loss": 0.3555,
+      "step": 57000
+    },
+    {
+      "epoch": 14.216957605985037,
+      "grad_norm": 10.28817081451416,
+      "learning_rate": 5.789526184538654e-06,
+      "loss": 0.3683,
+      "step": 57010
+    },
+    {
+      "epoch": 14.219451371571072,
+      "grad_norm": 11.527560234069824,
+      "learning_rate": 5.787032418952619e-06,
+      "loss": 0.3426,
+      "step": 57020
+    },
+    {
+      "epoch": 14.221945137157107,
+      "grad_norm": 5.7110724449157715,
+      "learning_rate": 5.784538653366584e-06,
+      "loss": 0.3402,
+      "step": 57030
+    },
+    {
+      "epoch": 14.224438902743142,
+      "grad_norm": 8.909784317016602,
+      "learning_rate": 5.782044887780549e-06,
+      "loss": 0.415,
+      "step": 57040
+    },
+    {
+      "epoch": 14.226932668329177,
+      "grad_norm": 9.159612655639648,
+      "learning_rate": 5.779551122194514e-06,
+      "loss": 0.3765,
+      "step": 57050
+    },
+    {
+      "epoch": 14.229426433915211,
+      "grad_norm": 7.4799346923828125,
+      "learning_rate": 5.77705735660848e-06,
+      "loss": 0.3718,
+      "step": 57060
+    },
+    {
+      "epoch": 14.231920199501246,
+      "grad_norm": 17.878170013427734,
+      "learning_rate": 5.774563591022444e-06,
+      "loss": 0.3199,
+      "step": 57070
+    },
+    {
+      "epoch": 14.234413965087281,
+      "grad_norm": 7.576568603515625,
+      "learning_rate": 5.7720698254364095e-06,
+      "loss": 0.3624,
+      "step": 57080
+    },
+    {
+      "epoch": 14.236907730673316,
+      "grad_norm": 8.394084930419922,
+      "learning_rate": 5.769576059850375e-06,
+      "loss": 0.3559,
+      "step": 57090
+    },
+    {
+      "epoch": 14.239401496259351,
+      "grad_norm": 9.45663070678711,
+      "learning_rate": 5.76708229426434e-06,
+      "loss": 0.3238,
+      "step": 57100
+    },
+    {
+      "epoch": 14.241895261845386,
+      "grad_norm": 7.425997734069824,
+      "learning_rate": 5.7645885286783046e-06,
+      "loss": 0.3416,
+      "step": 57110
+    },
+    {
+      "epoch": 14.24438902743142,
+      "grad_norm": 7.545619964599609,
+      "learning_rate": 5.76209476309227e-06,
+      "loss": 0.4245,
+      "step": 57120
+    },
+    {
+      "epoch": 14.246882793017456,
+      "grad_norm": 9.640861511230469,
+      "learning_rate": 5.759600997506235e-06,
+      "loss": 0.3163,
+      "step": 57130
+    },
+    {
+      "epoch": 14.24937655860349,
+      "grad_norm": 9.203802108764648,
+      "learning_rate": 5.7571072319202005e-06,
+      "loss": 0.4469,
+      "step": 57140
+    },
+    {
+      "epoch": 14.251870324189527,
+      "grad_norm": 7.76652193069458,
+      "learning_rate": 5.754613466334165e-06,
+      "loss": 0.3865,
+      "step": 57150
+    },
+    {
+      "epoch": 14.254364089775562,
+      "grad_norm": 7.3147358894348145,
+      "learning_rate": 5.75211970074813e-06,
+      "loss": 0.3029,
+      "step": 57160
+    },
+    {
+      "epoch": 14.256857855361597,
+      "grad_norm": 9.476174354553223,
+      "learning_rate": 5.7496259351620956e-06,
+      "loss": 0.3637,
+      "step": 57170
+    },
+    {
+      "epoch": 14.259351620947632,
+      "grad_norm": 6.6377434730529785,
+      "learning_rate": 5.747132169576061e-06,
+      "loss": 0.2983,
+      "step": 57180
+    },
+    {
+      "epoch": 14.261845386533667,
+      "grad_norm": 6.473023891448975,
+      "learning_rate": 5.744638403990025e-06,
+      "loss": 0.319,
+      "step": 57190
+    },
+    {
+      "epoch": 14.264339152119701,
+      "grad_norm": 5.787720680236816,
+      "learning_rate": 5.742144638403991e-06,
+      "loss": 0.325,
+      "step": 57200
+    },
+    {
+      "epoch": 14.266832917705736,
+      "grad_norm": 7.323090076446533,
+      "learning_rate": 5.739650872817956e-06,
+      "loss": 0.3233,
+      "step": 57210
+    },
+    {
+      "epoch": 14.269326683291771,
+      "grad_norm": 7.3568549156188965,
+      "learning_rate": 5.737157107231921e-06,
+      "loss": 0.3679,
+      "step": 57220
+    },
+    {
+      "epoch": 14.271820448877806,
+      "grad_norm": 8.636281967163086,
+      "learning_rate": 5.734663341645885e-06,
+      "loss": 0.2744,
+      "step": 57230
+    },
+    {
+      "epoch": 14.27431421446384,
+      "grad_norm": 6.717350482940674,
+      "learning_rate": 5.73216957605985e-06,
+      "loss": 0.293,
+      "step": 57240
+    },
+    {
+      "epoch": 14.276807980049876,
+      "grad_norm": 9.460376739501953,
+      "learning_rate": 5.729675810473816e-06,
+      "loss": 0.3502,
+      "step": 57250
+    },
+    {
+      "epoch": 14.27930174563591,
+      "grad_norm": 8.800605773925781,
+      "learning_rate": 5.727182044887782e-06,
+      "loss": 0.3078,
+      "step": 57260
+    },
+    {
+      "epoch": 14.281795511221945,
+      "grad_norm": 7.889537811279297,
+      "learning_rate": 5.724688279301747e-06,
+      "loss": 0.365,
+      "step": 57270
+    },
+    {
+      "epoch": 14.28428927680798,
+      "grad_norm": 10.018247604370117,
+      "learning_rate": 5.722194513715711e-06,
+      "loss": 0.2993,
+      "step": 57280
+    },
+    {
+      "epoch": 14.286783042394015,
+      "grad_norm": 5.913888931274414,
+      "learning_rate": 5.719700748129676e-06,
+      "loss": 0.2821,
+      "step": 57290
+    },
+    {
+      "epoch": 14.28927680798005,
+      "grad_norm": 6.76155424118042,
+      "learning_rate": 5.717206982543641e-06,
+      "loss": 0.3191,
+      "step": 57300
+    },
+    {
+      "epoch": 14.291770573566085,
+      "grad_norm": 6.1279215812683105,
+      "learning_rate": 5.714713216957607e-06,
+      "loss": 0.3588,
+      "step": 57310
+    },
+    {
+      "epoch": 14.29426433915212,
+      "grad_norm": 8.195113182067871,
+      "learning_rate": 5.712219451371571e-06,
+      "loss": 0.3062,
+      "step": 57320
+    },
+    {
+      "epoch": 14.296758104738155,
+      "grad_norm": 7.183683395385742,
+      "learning_rate": 5.709725685785536e-06,
+      "loss": 0.3615,
+      "step": 57330
+    },
+    {
+      "epoch": 14.29925187032419,
+      "grad_norm": 10.08060359954834,
+      "learning_rate": 5.707231920199502e-06,
+      "loss": 0.3116,
+      "step": 57340
+    },
+    {
+      "epoch": 14.301745635910224,
+      "grad_norm": 7.461446762084961,
+      "learning_rate": 5.704738154613467e-06,
+      "loss": 0.3036,
+      "step": 57350
+    },
+    {
+      "epoch": 14.30423940149626,
+      "grad_norm": 8.784198760986328,
+      "learning_rate": 5.702244389027431e-06,
+      "loss": 0.2936,
+      "step": 57360
+    },
+    {
+      "epoch": 14.306733167082294,
+      "grad_norm": 8.67398452758789,
+      "learning_rate": 5.699750623441397e-06,
+      "loss": 0.2893,
+      "step": 57370
+    },
+    {
+      "epoch": 14.309226932668329,
+      "grad_norm": 7.162827014923096,
+      "learning_rate": 5.697256857855362e-06,
+      "loss": 0.4105,
+      "step": 57380
+    },
+    {
+      "epoch": 14.311720698254364,
+      "grad_norm": 7.423861026763916,
+      "learning_rate": 5.694763092269327e-06,
+      "loss": 0.3044,
+      "step": 57390
+    },
+    {
+      "epoch": 14.314214463840399,
+      "grad_norm": 11.698079109191895,
+      "learning_rate": 5.692269326683292e-06,
+      "loss": 0.3955,
+      "step": 57400
+    },
+    {
+      "epoch": 14.316708229426434,
+      "grad_norm": 11.152626037597656,
+      "learning_rate": 5.689775561097257e-06,
+      "loss": 0.3488,
+      "step": 57410
+    },
+    {
+      "epoch": 14.319201995012468,
+      "grad_norm": 9.103079795837402,
+      "learning_rate": 5.687281795511222e-06,
+      "loss": 0.3548,
+      "step": 57420
+    },
+    {
+      "epoch": 14.321695760598503,
+      "grad_norm": 10.388227462768555,
+      "learning_rate": 5.684788029925188e-06,
+      "loss": 0.323,
+      "step": 57430
+    },
+    {
+      "epoch": 14.324189526184538,
+      "grad_norm": 9.910160064697266,
+      "learning_rate": 5.682294264339152e-06,
+      "loss": 0.3425,
+      "step": 57440
+    },
+    {
+      "epoch": 14.326683291770573,
+      "grad_norm": 8.148764610290527,
+      "learning_rate": 5.6798004987531175e-06,
+      "loss": 0.3473,
+      "step": 57450
+    },
+    {
+      "epoch": 14.329177057356608,
+      "grad_norm": 9.755147933959961,
+      "learning_rate": 5.677306733167083e-06,
+      "loss": 0.316,
+      "step": 57460
+    },
+    {
+      "epoch": 14.331670822942643,
+      "grad_norm": 5.160188674926758,
+      "learning_rate": 5.674812967581048e-06,
+      "loss": 0.3304,
+      "step": 57470
+    },
+    {
+      "epoch": 14.334164588528678,
+      "grad_norm": 9.043822288513184,
+      "learning_rate": 5.6723192019950125e-06,
+      "loss": 0.2585,
+      "step": 57480
+    },
+    {
+      "epoch": 14.336658354114713,
+      "grad_norm": 9.036372184753418,
+      "learning_rate": 5.669825436408978e-06,
+      "loss": 0.2661,
+      "step": 57490
+    },
+    {
+      "epoch": 14.339152119700747,
+      "grad_norm": 8.77037525177002,
+      "learning_rate": 5.667331670822943e-06,
+      "loss": 0.288,
+      "step": 57500
+    },
+    {
+      "epoch": 14.341645885286782,
+      "grad_norm": 5.4910712242126465,
+      "learning_rate": 5.6648379052369085e-06,
+      "loss": 0.3033,
+      "step": 57510
+    },
+    {
+      "epoch": 14.344139650872817,
+      "grad_norm": 11.051765441894531,
+      "learning_rate": 5.662344139650874e-06,
+      "loss": 0.3897,
+      "step": 57520
+    },
+    {
+      "epoch": 14.346633416458852,
+      "grad_norm": 6.824289798736572,
+      "learning_rate": 5.659850374064838e-06,
+      "loss": 0.3452,
+      "step": 57530
+    },
+    {
+      "epoch": 14.349127182044889,
+      "grad_norm": 7.920559406280518,
+      "learning_rate": 5.6573566084788036e-06,
+      "loss": 0.3175,
+      "step": 57540
+    },
+    {
+      "epoch": 14.351620947630924,
+      "grad_norm": 9.274563789367676,
+      "learning_rate": 5.654862842892769e-06,
+      "loss": 0.3346,
+      "step": 57550
+    },
+    {
+      "epoch": 14.354114713216958,
+      "grad_norm": 9.150640487670898,
+      "learning_rate": 5.652369077306734e-06,
+      "loss": 0.2985,
+      "step": 57560
+    },
+    {
+      "epoch": 14.356608478802993,
+      "grad_norm": 10.429580688476562,
+      "learning_rate": 5.649875311720699e-06,
+      "loss": 0.3712,
+      "step": 57570
+    },
+    {
+      "epoch": 14.359102244389028,
+      "grad_norm": 6.863627910614014,
+      "learning_rate": 5.647381546134664e-06,
+      "loss": 0.3518,
+      "step": 57580
+    },
+    {
+      "epoch": 14.361596009975063,
+      "grad_norm": 10.7223482131958,
+      "learning_rate": 5.644887780548629e-06,
+      "loss": 0.2925,
+      "step": 57590
+    },
+    {
+      "epoch": 14.364089775561098,
+      "grad_norm": 9.246960639953613,
+      "learning_rate": 5.6423940149625946e-06,
+      "loss": 0.3043,
+      "step": 57600
+    },
+    {
+      "epoch": 14.366583541147133,
+      "grad_norm": 8.545487403869629,
+      "learning_rate": 5.639900249376559e-06,
+      "loss": 0.3126,
+      "step": 57610
+    },
+    {
+      "epoch": 14.369077306733168,
+      "grad_norm": 7.441158771514893,
+      "learning_rate": 5.637406483790524e-06,
+      "loss": 0.3215,
+      "step": 57620
+    },
+    {
+      "epoch": 14.371571072319203,
+      "grad_norm": 8.528327941894531,
+      "learning_rate": 5.63491271820449e-06,
+      "loss": 0.352,
+      "step": 57630
+    },
+    {
+      "epoch": 14.374064837905237,
+      "grad_norm": 10.5371732711792,
+      "learning_rate": 5.632418952618455e-06,
+      "loss": 0.3339,
+      "step": 57640
+    },
+    {
+      "epoch": 14.376558603491272,
+      "grad_norm": 7.135348320007324,
+      "learning_rate": 5.6299251870324186e-06,
+      "loss": 0.4033,
+      "step": 57650
+    },
+    {
+      "epoch": 14.379052369077307,
+      "grad_norm": 8.768062591552734,
+      "learning_rate": 5.627431421446385e-06,
+      "loss": 0.3769,
+      "step": 57660
+    },
+    {
+      "epoch": 14.381546134663342,
+      "grad_norm": 9.260017395019531,
+      "learning_rate": 5.62493765586035e-06,
+      "loss": 0.3351,
+      "step": 57670
+    },
+    {
+      "epoch": 14.384039900249377,
+      "grad_norm": 10.839845657348633,
+      "learning_rate": 5.622443890274315e-06,
+      "loss": 0.317,
+      "step": 57680
+    },
+    {
+      "epoch": 14.386533665835412,
+      "grad_norm": 5.972354888916016,
+      "learning_rate": 5.619950124688279e-06,
+      "loss": 0.3857,
+      "step": 57690
+    },
+    {
+      "epoch": 14.389027431421447,
+      "grad_norm": 9.633734703063965,
+      "learning_rate": 5.617456359102244e-06,
+      "loss": 0.3434,
+      "step": 57700
+    },
+    {
+      "epoch": 14.391521197007481,
+      "grad_norm": 9.835312843322754,
+      "learning_rate": 5.61496259351621e-06,
+      "loss": 0.3031,
+      "step": 57710
+    },
+    {
+      "epoch": 14.394014962593516,
+      "grad_norm": 8.749693870544434,
+      "learning_rate": 5.612468827930176e-06,
+      "loss": 0.3306,
+      "step": 57720
+    },
+    {
+      "epoch": 14.396508728179551,
+      "grad_norm": 6.144033908843994,
+      "learning_rate": 5.609975062344139e-06,
+      "loss": 0.332,
+      "step": 57730
+    },
+    {
+      "epoch": 14.399002493765586,
+      "grad_norm": 6.014637470245361,
+      "learning_rate": 5.607481296758105e-06,
+      "loss": 0.3146,
+      "step": 57740
+    },
+    {
+      "epoch": 14.401496259351621,
+      "grad_norm": 7.245567321777344,
+      "learning_rate": 5.60498753117207e-06,
+      "loss": 0.3847,
+      "step": 57750
+    },
+    {
+      "epoch": 14.403990024937656,
+      "grad_norm": 4.784339904785156,
+      "learning_rate": 5.602493765586035e-06,
+      "loss": 0.3821,
+      "step": 57760
+    },
+    {
+      "epoch": 14.40648379052369,
+      "grad_norm": 8.494146347045898,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.3402,
+      "step": 57770
+    },
+    {
+      "epoch": 14.408977556109726,
+      "grad_norm": 7.598515510559082,
+      "learning_rate": 5.597506234413965e-06,
+      "loss": 0.325,
+      "step": 57780
+    },
+    {
+      "epoch": 14.41147132169576,
+      "grad_norm": 7.4690423011779785,
+      "learning_rate": 5.59501246882793e-06,
+      "loss": 0.3582,
+      "step": 57790
+    },
+    {
+      "epoch": 14.413965087281795,
+      "grad_norm": 8.942574501037598,
+      "learning_rate": 5.592518703241896e-06,
+      "loss": 0.3168,
+      "step": 57800
+    },
+    {
+      "epoch": 14.41645885286783,
+      "grad_norm": 8.781198501586914,
+      "learning_rate": 5.590024937655861e-06,
+      "loss": 0.2854,
+      "step": 57810
+    },
+    {
+      "epoch": 14.418952618453865,
+      "grad_norm": 5.599854469299316,
+      "learning_rate": 5.5875311720698254e-06,
+      "loss": 0.3373,
+      "step": 57820
+    },
+    {
+      "epoch": 14.4214463840399,
+      "grad_norm": 6.340407371520996,
+      "learning_rate": 5.585037406483791e-06,
+      "loss": 0.2954,
+      "step": 57830
+    },
+    {
+      "epoch": 14.423940149625935,
+      "grad_norm": 5.647294998168945,
+      "learning_rate": 5.582543640897756e-06,
+      "loss": 0.2989,
+      "step": 57840
+    },
+    {
+      "epoch": 14.42643391521197,
+      "grad_norm": 8.372790336608887,
+      "learning_rate": 5.580049875311721e-06,
+      "loss": 0.332,
+      "step": 57850
+    },
+    {
+      "epoch": 14.428927680798004,
+      "grad_norm": 8.006759643554688,
+      "learning_rate": 5.577556109725686e-06,
+      "loss": 0.362,
+      "step": 57860
+    },
+    {
+      "epoch": 14.43142144638404,
+      "grad_norm": 5.205109596252441,
+      "learning_rate": 5.575062344139651e-06,
+      "loss": 0.3863,
+      "step": 57870
+    },
+    {
+      "epoch": 14.433915211970074,
+      "grad_norm": 6.987797260284424,
+      "learning_rate": 5.5725685785536165e-06,
+      "loss": 0.3135,
+      "step": 57880
+    },
+    {
+      "epoch": 14.436408977556109,
+      "grad_norm": 9.735954284667969,
+      "learning_rate": 5.570074812967582e-06,
+      "loss": 0.2886,
+      "step": 57890
+    },
+    {
+      "epoch": 14.438902743142144,
+      "grad_norm": 10.121098518371582,
+      "learning_rate": 5.567581047381546e-06,
+      "loss": 0.3274,
+      "step": 57900
+    },
+    {
+      "epoch": 14.441396508728179,
+      "grad_norm": 8.132126808166504,
+      "learning_rate": 5.5650872817955115e-06,
+      "loss": 0.309,
+      "step": 57910
+    },
+    {
+      "epoch": 14.443890274314214,
+      "grad_norm": 9.95006275177002,
+      "learning_rate": 5.562593516209477e-06,
+      "loss": 0.3218,
+      "step": 57920
+    },
+    {
+      "epoch": 14.446384039900249,
+      "grad_norm": 9.619258880615234,
+      "learning_rate": 5.560099750623442e-06,
+      "loss": 0.3107,
+      "step": 57930
+    },
+    {
+      "epoch": 14.448877805486283,
+      "grad_norm": 8.964179039001465,
+      "learning_rate": 5.557605985037407e-06,
+      "loss": 0.4061,
+      "step": 57940
+    },
+    {
+      "epoch": 14.451371571072318,
+      "grad_norm": 15.57805347442627,
+      "learning_rate": 5.555112219451372e-06,
+      "loss": 0.3487,
+      "step": 57950
+    },
+    {
+      "epoch": 14.453865336658355,
+      "grad_norm": 10.295059204101562,
+      "learning_rate": 5.552618453865337e-06,
+      "loss": 0.2948,
+      "step": 57960
+    },
+    {
+      "epoch": 14.45635910224439,
+      "grad_norm": 6.323406219482422,
+      "learning_rate": 5.5501246882793026e-06,
+      "loss": 0.3311,
+      "step": 57970
+    },
+    {
+      "epoch": 14.458852867830425,
+      "grad_norm": 6.79067325592041,
+      "learning_rate": 5.547630922693267e-06,
+      "loss": 0.368,
+      "step": 57980
+    },
+    {
+      "epoch": 14.46134663341646,
+      "grad_norm": 7.148687839508057,
+      "learning_rate": 5.545137157107232e-06,
+      "loss": 0.312,
+      "step": 57990
+    },
+    {
+      "epoch": 14.463840399002494,
+      "grad_norm": 11.232911109924316,
+      "learning_rate": 5.542643391521198e-06,
+      "loss": 0.3485,
+      "step": 58000
+    },
+    {
+      "epoch": 14.46633416458853,
+      "grad_norm": 7.9350738525390625,
+      "learning_rate": 5.540149625935163e-06,
+      "loss": 0.2499,
+      "step": 58010
+    },
+    {
+      "epoch": 14.468827930174564,
+      "grad_norm": 7.820169448852539,
+      "learning_rate": 5.537655860349128e-06,
+      "loss": 0.3012,
+      "step": 58020
+    },
+    {
+      "epoch": 14.471321695760599,
+      "grad_norm": 6.683993816375732,
+      "learning_rate": 5.535162094763093e-06,
+      "loss": 0.3388,
+      "step": 58030
+    },
+    {
+      "epoch": 14.473815461346634,
+      "grad_norm": 8.00978946685791,
+      "learning_rate": 5.532668329177058e-06,
+      "loss": 0.2825,
+      "step": 58040
+    },
+    {
+      "epoch": 14.476309226932669,
+      "grad_norm": 7.9973015785217285,
+      "learning_rate": 5.530174563591023e-06,
+      "loss": 0.3373,
+      "step": 58050
+    },
+    {
+      "epoch": 14.478802992518704,
+      "grad_norm": 9.217852592468262,
+      "learning_rate": 5.527680798004989e-06,
+      "loss": 0.2765,
+      "step": 58060
+    },
+    {
+      "epoch": 14.481296758104738,
+      "grad_norm": 9.490992546081543,
+      "learning_rate": 5.525187032418953e-06,
+      "loss": 0.3424,
+      "step": 58070
+    },
+    {
+      "epoch": 14.483790523690773,
+      "grad_norm": 6.221439361572266,
+      "learning_rate": 5.522693266832918e-06,
+      "loss": 0.3139,
+      "step": 58080
+    },
+    {
+      "epoch": 14.486284289276808,
+      "grad_norm": 5.196888446807861,
+      "learning_rate": 5.520199501246884e-06,
+      "loss": 0.3762,
+      "step": 58090
+    },
+    {
+      "epoch": 14.488778054862843,
+      "grad_norm": 9.887178421020508,
+      "learning_rate": 5.517705735660849e-06,
+      "loss": 0.3322,
+      "step": 58100
+    },
+    {
+      "epoch": 14.491271820448878,
+      "grad_norm": 11.542625427246094,
+      "learning_rate": 5.515211970074813e-06,
+      "loss": 0.3423,
+      "step": 58110
+    },
+    {
+      "epoch": 14.493765586034913,
+      "grad_norm": 7.159249782562256,
+      "learning_rate": 5.512718204488778e-06,
+      "loss": 0.3121,
+      "step": 58120
+    },
+    {
+      "epoch": 14.496259351620948,
+      "grad_norm": 11.838788986206055,
+      "learning_rate": 5.510224438902744e-06,
+      "loss": 0.3516,
+      "step": 58130
+    },
+    {
+      "epoch": 14.498753117206983,
+      "grad_norm": 8.8449068069458,
+      "learning_rate": 5.5077306733167094e-06,
+      "loss": 0.3357,
+      "step": 58140
+    },
+    {
+      "epoch": 14.501246882793017,
+      "grad_norm": 10.590514183044434,
+      "learning_rate": 5.505236907730673e-06,
+      "loss": 0.2957,
+      "step": 58150
+    },
+    {
+      "epoch": 14.503740648379052,
+      "grad_norm": 9.069424629211426,
+      "learning_rate": 5.502743142144638e-06,
+      "loss": 0.2783,
+      "step": 58160
+    },
+    {
+      "epoch": 14.506234413965087,
+      "grad_norm": 7.590739727020264,
+      "learning_rate": 5.500249376558604e-06,
+      "loss": 0.2904,
+      "step": 58170
+    },
+    {
+      "epoch": 14.508728179551122,
+      "grad_norm": 6.409052848815918,
+      "learning_rate": 5.497755610972569e-06,
+      "loss": 0.3753,
+      "step": 58180
+    },
+    {
+      "epoch": 14.511221945137157,
+      "grad_norm": 7.4375481605529785,
+      "learning_rate": 5.4952618453865334e-06,
+      "loss": 0.2849,
+      "step": 58190
+    },
+    {
+      "epoch": 14.513715710723192,
+      "grad_norm": 8.864033699035645,
+      "learning_rate": 5.492768079800499e-06,
+      "loss": 0.32,
+      "step": 58200
+    },
+    {
+      "epoch": 14.516209476309227,
+      "grad_norm": 6.455445766448975,
+      "learning_rate": 5.490274314214464e-06,
+      "loss": 0.3475,
+      "step": 58210
+    },
+    {
+      "epoch": 14.518703241895262,
+      "grad_norm": 6.384631633758545,
+      "learning_rate": 5.487780548628429e-06,
+      "loss": 0.2812,
+      "step": 58220
+    },
+    {
+      "epoch": 14.521197007481296,
+      "grad_norm": 9.22549057006836,
+      "learning_rate": 5.485286783042394e-06,
+      "loss": 0.3304,
+      "step": 58230
+    },
+    {
+      "epoch": 14.523690773067331,
+      "grad_norm": 7.639939308166504,
+      "learning_rate": 5.482793017456359e-06,
+      "loss": 0.3435,
+      "step": 58240
+    },
+    {
+      "epoch": 14.526184538653366,
+      "grad_norm": 5.905279636383057,
+      "learning_rate": 5.4802992518703244e-06,
+      "loss": 0.2744,
+      "step": 58250
+    },
+    {
+      "epoch": 14.528678304239401,
+      "grad_norm": 8.444292068481445,
+      "learning_rate": 5.47780548628429e-06,
+      "loss": 0.3367,
+      "step": 58260
+    },
+    {
+      "epoch": 14.531172069825436,
+      "grad_norm": 9.172270774841309,
+      "learning_rate": 5.475311720698255e-06,
+      "loss": 0.3669,
+      "step": 58270
+    },
+    {
+      "epoch": 14.53366583541147,
+      "grad_norm": 15.167724609375,
+      "learning_rate": 5.4728179551122195e-06,
+      "loss": 0.4021,
+      "step": 58280
+    },
+    {
+      "epoch": 14.536159600997506,
+      "grad_norm": 8.261863708496094,
+      "learning_rate": 5.470324189526185e-06,
+      "loss": 0.2871,
+      "step": 58290
+    },
+    {
+      "epoch": 14.53865336658354,
+      "grad_norm": 10.16102123260498,
+      "learning_rate": 5.46783042394015e-06,
+      "loss": 0.3401,
+      "step": 58300
+    },
+    {
+      "epoch": 14.541147132169575,
+      "grad_norm": 8.043190956115723,
+      "learning_rate": 5.4653366583541155e-06,
+      "loss": 0.3537,
+      "step": 58310
+    },
+    {
+      "epoch": 14.54364089775561,
+      "grad_norm": 21.67076873779297,
+      "learning_rate": 5.46284289276808e-06,
+      "loss": 0.3269,
+      "step": 58320
+    },
+    {
+      "epoch": 14.546134663341645,
+      "grad_norm": 6.5939459800720215,
+      "learning_rate": 5.460349127182045e-06,
+      "loss": 0.3313,
+      "step": 58330
+    },
+    {
+      "epoch": 14.548628428927682,
+      "grad_norm": 6.585400581359863,
+      "learning_rate": 5.4578553615960105e-06,
+      "loss": 0.3449,
+      "step": 58340
+    },
+    {
+      "epoch": 14.551122194513717,
+      "grad_norm": 9.368444442749023,
+      "learning_rate": 5.455361596009976e-06,
+      "loss": 0.3536,
+      "step": 58350
+    },
+    {
+      "epoch": 14.553615960099751,
+      "grad_norm": 8.152172088623047,
+      "learning_rate": 5.45286783042394e-06,
+      "loss": 0.3483,
+      "step": 58360
+    },
+    {
+      "epoch": 14.556109725685786,
+      "grad_norm": 9.879807472229004,
+      "learning_rate": 5.450374064837906e-06,
+      "loss": 0.3036,
+      "step": 58370
+    },
+    {
+      "epoch": 14.558603491271821,
+      "grad_norm": 16.792516708374023,
+      "learning_rate": 5.447880299251871e-06,
+      "loss": 0.3782,
+      "step": 58380
+    },
+    {
+      "epoch": 14.561097256857856,
+      "grad_norm": 6.8559112548828125,
+      "learning_rate": 5.445386533665836e-06,
+      "loss": 0.3641,
+      "step": 58390
+    },
+    {
+      "epoch": 14.563591022443891,
+      "grad_norm": 12.230280876159668,
+      "learning_rate": 5.442892768079801e-06,
+      "loss": 0.3335,
+      "step": 58400
+    },
+    {
+      "epoch": 14.566084788029926,
+      "grad_norm": 5.788147926330566,
+      "learning_rate": 5.440399002493766e-06,
+      "loss": 0.3376,
+      "step": 58410
+    },
+    {
+      "epoch": 14.56857855361596,
+      "grad_norm": 9.005399703979492,
+      "learning_rate": 5.437905236907731e-06,
+      "loss": 0.3828,
+      "step": 58420
+    },
+    {
+      "epoch": 14.571072319201996,
+      "grad_norm": 6.153980255126953,
+      "learning_rate": 5.435411471321697e-06,
+      "loss": 0.3214,
+      "step": 58430
+    },
+    {
+      "epoch": 14.57356608478803,
+      "grad_norm": 7.769611835479736,
+      "learning_rate": 5.432917705735661e-06,
+      "loss": 0.2673,
+      "step": 58440
+    },
+    {
+      "epoch": 14.576059850374065,
+      "grad_norm": 7.903584957122803,
+      "learning_rate": 5.430423940149626e-06,
+      "loss": 0.3313,
+      "step": 58450
+    },
+    {
+      "epoch": 14.5785536159601,
+      "grad_norm": 11.687101364135742,
+      "learning_rate": 5.427930174563592e-06,
+      "loss": 0.3828,
+      "step": 58460
+    },
+    {
+      "epoch": 14.581047381546135,
+      "grad_norm": 9.845630645751953,
+      "learning_rate": 5.425436408977557e-06,
+      "loss": 0.3014,
+      "step": 58470
+    },
+    {
+      "epoch": 14.58354114713217,
+      "grad_norm": 7.4955244064331055,
+      "learning_rate": 5.4229426433915215e-06,
+      "loss": 0.3243,
+      "step": 58480
+    },
+    {
+      "epoch": 14.586034912718205,
+      "grad_norm": 9.153183937072754,
+      "learning_rate": 5.420448877805487e-06,
+      "loss": 0.3849,
+      "step": 58490
+    },
+    {
+      "epoch": 14.58852867830424,
+      "grad_norm": 8.650388717651367,
+      "learning_rate": 5.417955112219452e-06,
+      "loss": 0.3057,
+      "step": 58500
+    },
+    {
+      "epoch": 14.591022443890274,
+      "grad_norm": 5.724157333374023,
+      "learning_rate": 5.415461346633417e-06,
+      "loss": 0.2931,
+      "step": 58510
+    },
+    {
+      "epoch": 14.59351620947631,
+      "grad_norm": 8.662379264831543,
+      "learning_rate": 5.412967581047383e-06,
+      "loss": 0.324,
+      "step": 58520
+    },
+    {
+      "epoch": 14.596009975062344,
+      "grad_norm": 6.165246963500977,
+      "learning_rate": 5.410473815461346e-06,
+      "loss": 0.305,
+      "step": 58530
+    },
+    {
+      "epoch": 14.598503740648379,
+      "grad_norm": 10.333334922790527,
+      "learning_rate": 5.4079800498753125e-06,
+      "loss": 0.412,
+      "step": 58540
+    },
+    {
+      "epoch": 14.600997506234414,
+      "grad_norm": 12.94969654083252,
+      "learning_rate": 5.405486284289278e-06,
+      "loss": 0.3273,
+      "step": 58550
+    },
+    {
+      "epoch": 14.603491271820449,
+      "grad_norm": 8.342123985290527,
+      "learning_rate": 5.402992518703243e-06,
+      "loss": 0.3934,
+      "step": 58560
+    },
+    {
+      "epoch": 14.605985037406484,
+      "grad_norm": 6.839729309082031,
+      "learning_rate": 5.400498753117207e-06,
+      "loss": 0.303,
+      "step": 58570
+    },
+    {
+      "epoch": 14.608478802992519,
+      "grad_norm": 9.169561386108398,
+      "learning_rate": 5.398004987531172e-06,
+      "loss": 0.2673,
+      "step": 58580
+    },
+    {
+      "epoch": 14.610972568578553,
+      "grad_norm": 7.035985946655273,
+      "learning_rate": 5.395511221945137e-06,
+      "loss": 0.2992,
+      "step": 58590
+    },
+    {
+      "epoch": 14.613466334164588,
+      "grad_norm": 7.293179988861084,
+      "learning_rate": 5.3930174563591035e-06,
+      "loss": 0.3203,
+      "step": 58600
+    },
+    {
+      "epoch": 14.615960099750623,
+      "grad_norm": 8.904796600341797,
+      "learning_rate": 5.390523690773067e-06,
+      "loss": 0.3217,
+      "step": 58610
+    },
+    {
+      "epoch": 14.618453865336658,
+      "grad_norm": 9.282125473022461,
+      "learning_rate": 5.3880299251870324e-06,
+      "loss": 0.3283,
+      "step": 58620
+    },
+    {
+      "epoch": 14.620947630922693,
+      "grad_norm": 7.752005577087402,
+      "learning_rate": 5.385536159600998e-06,
+      "loss": 0.3145,
+      "step": 58630
+    },
+    {
+      "epoch": 14.623441396508728,
+      "grad_norm": 5.774524688720703,
+      "learning_rate": 5.383042394014963e-06,
+      "loss": 0.2747,
+      "step": 58640
+    },
+    {
+      "epoch": 14.625935162094763,
+      "grad_norm": 8.292061805725098,
+      "learning_rate": 5.3805486284289275e-06,
+      "loss": 0.2488,
+      "step": 58650
+    },
+    {
+      "epoch": 14.628428927680797,
+      "grad_norm": 9.226009368896484,
+      "learning_rate": 5.378054862842893e-06,
+      "loss": 0.2889,
+      "step": 58660
+    },
+    {
+      "epoch": 14.630922693266832,
+      "grad_norm": 8.59598159790039,
+      "learning_rate": 5.375561097256858e-06,
+      "loss": 0.3311,
+      "step": 58670
+    },
+    {
+      "epoch": 14.633416458852867,
+      "grad_norm": 11.77875804901123,
+      "learning_rate": 5.3730673316708234e-06,
+      "loss": 0.289,
+      "step": 58680
+    },
+    {
+      "epoch": 14.635910224438902,
+      "grad_norm": 8.944512367248535,
+      "learning_rate": 5.370573566084788e-06,
+      "loss": 0.3456,
+      "step": 58690
+    },
+    {
+      "epoch": 14.638403990024937,
+      "grad_norm": 8.320801734924316,
+      "learning_rate": 5.368079800498753e-06,
+      "loss": 0.3284,
+      "step": 58700
+    },
+    {
+      "epoch": 14.640897755610972,
+      "grad_norm": 20.068309783935547,
+      "learning_rate": 5.3655860349127185e-06,
+      "loss": 0.3567,
+      "step": 58710
+    },
+    {
+      "epoch": 14.643391521197007,
+      "grad_norm": 13.9030122756958,
+      "learning_rate": 5.363092269326684e-06,
+      "loss": 0.3672,
+      "step": 58720
+    },
+    {
+      "epoch": 14.645885286783042,
+      "grad_norm": 9.718421936035156,
+      "learning_rate": 5.360598503740648e-06,
+      "loss": 0.3213,
+      "step": 58730
+    },
+    {
+      "epoch": 14.648379052369076,
+      "grad_norm": 7.2954301834106445,
+      "learning_rate": 5.358104738154614e-06,
+      "loss": 0.3902,
+      "step": 58740
+    },
+    {
+      "epoch": 14.650872817955111,
+      "grad_norm": 7.279516696929932,
+      "learning_rate": 5.355610972568579e-06,
+      "loss": 0.2793,
+      "step": 58750
+    },
+    {
+      "epoch": 14.653366583541148,
+      "grad_norm": 10.923696517944336,
+      "learning_rate": 5.353117206982544e-06,
+      "loss": 0.3365,
+      "step": 58760
+    },
+    {
+      "epoch": 14.655860349127183,
+      "grad_norm": 9.616853713989258,
+      "learning_rate": 5.3506234413965095e-06,
+      "loss": 0.3174,
+      "step": 58770
+    },
+    {
+      "epoch": 14.658354114713218,
+      "grad_norm": 9.328225135803223,
+      "learning_rate": 5.348129675810474e-06,
+      "loss": 0.3395,
+      "step": 58780
+    },
+    {
+      "epoch": 14.660847880299253,
+      "grad_norm": 8.011582374572754,
+      "learning_rate": 5.345635910224439e-06,
+      "loss": 0.3506,
+      "step": 58790
+    },
+    {
+      "epoch": 14.663341645885287,
+      "grad_norm": 9.812850952148438,
+      "learning_rate": 5.343142144638405e-06,
+      "loss": 0.2907,
+      "step": 58800
+    },
+    {
+      "epoch": 14.665835411471322,
+      "grad_norm": 8.969762802124023,
+      "learning_rate": 5.34064837905237e-06,
+      "loss": 0.3821,
+      "step": 58810
+    },
+    {
+      "epoch": 14.668329177057357,
+      "grad_norm": 8.796773910522461,
+      "learning_rate": 5.338154613466334e-06,
+      "loss": 0.3282,
+      "step": 58820
+    },
+    {
+      "epoch": 14.670822942643392,
+      "grad_norm": 9.331019401550293,
+      "learning_rate": 5.3356608478803e-06,
+      "loss": 0.3352,
+      "step": 58830
+    },
+    {
+      "epoch": 14.673316708229427,
+      "grad_norm": 14.473604202270508,
+      "learning_rate": 5.333167082294265e-06,
+      "loss": 0.2999,
+      "step": 58840
+    },
+    {
+      "epoch": 14.675810473815462,
+      "grad_norm": 7.026119232177734,
+      "learning_rate": 5.33067331670823e-06,
+      "loss": 0.3559,
+      "step": 58850
+    },
+    {
+      "epoch": 14.678304239401497,
+      "grad_norm": 7.760469436645508,
+      "learning_rate": 5.328179551122195e-06,
+      "loss": 0.3282,
+      "step": 58860
+    },
+    {
+      "epoch": 14.680798004987532,
+      "grad_norm": 10.54293155670166,
+      "learning_rate": 5.32568578553616e-06,
+      "loss": 0.4874,
+      "step": 58870
+    },
+    {
+      "epoch": 14.683291770573566,
+      "grad_norm": 10.247196197509766,
+      "learning_rate": 5.323192019950125e-06,
+      "loss": 0.3434,
+      "step": 58880
+    },
+    {
+      "epoch": 14.685785536159601,
+      "grad_norm": 12.179352760314941,
+      "learning_rate": 5.320698254364091e-06,
+      "loss": 0.3833,
+      "step": 58890
+    },
+    {
+      "epoch": 14.688279301745636,
+      "grad_norm": 8.831155776977539,
+      "learning_rate": 5.318204488778055e-06,
+      "loss": 0.3422,
+      "step": 58900
+    },
+    {
+      "epoch": 14.690773067331671,
+      "grad_norm": 5.796462535858154,
+      "learning_rate": 5.3157107231920205e-06,
+      "loss": 0.3022,
+      "step": 58910
+    },
+    {
+      "epoch": 14.693266832917706,
+      "grad_norm": 9.8500337600708,
+      "learning_rate": 5.313216957605986e-06,
+      "loss": 0.3686,
+      "step": 58920
+    },
+    {
+      "epoch": 14.69576059850374,
+      "grad_norm": 9.79806137084961,
+      "learning_rate": 5.310723192019951e-06,
+      "loss": 0.3164,
+      "step": 58930
+    },
+    {
+      "epoch": 14.698254364089776,
+      "grad_norm": 7.471320629119873,
+      "learning_rate": 5.308229426433915e-06,
+      "loss": 0.3621,
+      "step": 58940
+    },
+    {
+      "epoch": 14.70074812967581,
+      "grad_norm": 15.296502113342285,
+      "learning_rate": 5.305735660847881e-06,
+      "loss": 0.3376,
+      "step": 58950
+    },
+    {
+      "epoch": 14.703241895261845,
+      "grad_norm": 8.871153831481934,
+      "learning_rate": 5.303241895261846e-06,
+      "loss": 0.3574,
+      "step": 58960
+    },
+    {
+      "epoch": 14.70573566084788,
+      "grad_norm": 9.778191566467285,
+      "learning_rate": 5.3007481296758115e-06,
+      "loss": 0.3363,
+      "step": 58970
+    },
+    {
+      "epoch": 14.708229426433915,
+      "grad_norm": 8.317192077636719,
+      "learning_rate": 5.298254364089775e-06,
+      "loss": 0.2979,
+      "step": 58980
+    },
+    {
+      "epoch": 14.71072319201995,
+      "grad_norm": 9.88344669342041,
+      "learning_rate": 5.2957605985037404e-06,
+      "loss": 0.3778,
+      "step": 58990
+    },
+    {
+      "epoch": 14.713216957605985,
+      "grad_norm": 8.39840030670166,
+      "learning_rate": 5.293266832917706e-06,
+      "loss": 0.3757,
+      "step": 59000
+    },
+    {
+      "epoch": 14.71571072319202,
+      "grad_norm": 11.54598331451416,
+      "learning_rate": 5.290773067331672e-06,
+      "loss": 0.3286,
+      "step": 59010
+    },
+    {
+      "epoch": 14.718204488778055,
+      "grad_norm": 6.494427680969238,
+      "learning_rate": 5.28852867830424e-06,
+      "loss": 0.3799,
+      "step": 59020
+    },
+    {
+      "epoch": 14.72069825436409,
+      "grad_norm": 9.357276916503906,
+      "learning_rate": 5.2860349127182055e-06,
+      "loss": 0.2867,
+      "step": 59030
+    },
+    {
+      "epoch": 14.723192019950124,
+      "grad_norm": 8.344010353088379,
+      "learning_rate": 5.28354114713217e-06,
+      "loss": 0.3489,
+      "step": 59040
+    },
+    {
+      "epoch": 14.72568578553616,
+      "grad_norm": 6.008798599243164,
+      "learning_rate": 5.281047381546135e-06,
+      "loss": 0.3094,
+      "step": 59050
+    },
+    {
+      "epoch": 14.728179551122194,
+      "grad_norm": 10.23482894897461,
+      "learning_rate": 5.2785536159601006e-06,
+      "loss": 0.3203,
+      "step": 59060
+    },
+    {
+      "epoch": 14.730673316708229,
+      "grad_norm": 6.847126007080078,
+      "learning_rate": 5.276059850374066e-06,
+      "loss": 0.25,
+      "step": 59070
+    },
+    {
+      "epoch": 14.733167082294264,
+      "grad_norm": 9.797203063964844,
+      "learning_rate": 5.2735660847880295e-06,
+      "loss": 0.3885,
+      "step": 59080
+    },
+    {
+      "epoch": 14.735660847880299,
+      "grad_norm": 8.682002067565918,
+      "learning_rate": 5.271072319201996e-06,
+      "loss": 0.3408,
+      "step": 59090
+    },
+    {
+      "epoch": 14.738154613466333,
+      "grad_norm": 7.708632946014404,
+      "learning_rate": 5.268578553615961e-06,
+      "loss": 0.3084,
+      "step": 59100
+    },
+    {
+      "epoch": 14.740648379052368,
+      "grad_norm": 9.106949806213379,
+      "learning_rate": 5.266084788029926e-06,
+      "loss": 0.2894,
+      "step": 59110
+    },
+    {
+      "epoch": 14.743142144638403,
+      "grad_norm": 8.195284843444824,
+      "learning_rate": 5.26359102244389e-06,
+      "loss": 0.3228,
+      "step": 59120
+    },
+    {
+      "epoch": 14.745635910224438,
+      "grad_norm": 10.221166610717773,
+      "learning_rate": 5.261097256857855e-06,
+      "loss": 0.3822,
+      "step": 59130
+    },
+    {
+      "epoch": 14.748129675810475,
+      "grad_norm": 8.522336959838867,
+      "learning_rate": 5.2586034912718205e-06,
+      "loss": 0.3341,
+      "step": 59140
+    },
+    {
+      "epoch": 14.75062344139651,
+      "grad_norm": 13.70825481414795,
+      "learning_rate": 5.256109725685787e-06,
+      "loss": 0.3543,
+      "step": 59150
+    },
+    {
+      "epoch": 14.753117206982544,
+      "grad_norm": 7.4037184715271,
+      "learning_rate": 5.25361596009975e-06,
+      "loss": 0.3357,
+      "step": 59160
+    },
+    {
+      "epoch": 14.75561097256858,
+      "grad_norm": 10.474981307983398,
+      "learning_rate": 5.2511221945137156e-06,
+      "loss": 0.3487,
+      "step": 59170
+    },
+    {
+      "epoch": 14.758104738154614,
+      "grad_norm": 5.967161655426025,
+      "learning_rate": 5.248628428927681e-06,
+      "loss": 0.3564,
+      "step": 59180
+    },
+    {
+      "epoch": 14.760598503740649,
+      "grad_norm": 7.465957164764404,
+      "learning_rate": 5.246134663341646e-06,
+      "loss": 0.3241,
+      "step": 59190
+    },
+    {
+      "epoch": 14.763092269326684,
+      "grad_norm": 7.333925247192383,
+      "learning_rate": 5.2436408977556115e-06,
+      "loss": 0.2976,
+      "step": 59200
+    },
+    {
+      "epoch": 14.765586034912719,
+      "grad_norm": 8.941734313964844,
+      "learning_rate": 5.241147132169576e-06,
+      "loss": 0.4209,
+      "step": 59210
+    },
+    {
+      "epoch": 14.768079800498754,
+      "grad_norm": 7.812376976013184,
+      "learning_rate": 5.238653366583541e-06,
+      "loss": 0.2961,
+      "step": 59220
+    },
+    {
+      "epoch": 14.770573566084789,
+      "grad_norm": 9.06418228149414,
+      "learning_rate": 5.236159600997507e-06,
+      "loss": 0.3249,
+      "step": 59230
+    },
+    {
+      "epoch": 14.773067331670823,
+      "grad_norm": 8.520550727844238,
+      "learning_rate": 5.233665835411472e-06,
+      "loss": 0.3469,
+      "step": 59240
+    },
+    {
+      "epoch": 14.775561097256858,
+      "grad_norm": 5.776762962341309,
+      "learning_rate": 5.231172069825436e-06,
+      "loss": 0.3048,
+      "step": 59250
+    },
+    {
+      "epoch": 14.778054862842893,
+      "grad_norm": 13.731014251708984,
+      "learning_rate": 5.228678304239402e-06,
+      "loss": 0.3533,
+      "step": 59260
+    },
+    {
+      "epoch": 14.780548628428928,
+      "grad_norm": 10.045135498046875,
+      "learning_rate": 5.226184538653367e-06,
+      "loss": 0.3062,
+      "step": 59270
+    },
+    {
+      "epoch": 14.783042394014963,
+      "grad_norm": 9.857339859008789,
+      "learning_rate": 5.223690773067332e-06,
+      "loss": 0.2885,
+      "step": 59280
+    },
+    {
+      "epoch": 14.785536159600998,
+      "grad_norm": 7.196331977844238,
+      "learning_rate": 5.221197007481297e-06,
+      "loss": 0.3024,
+      "step": 59290
+    },
+    {
+      "epoch": 14.788029925187033,
+      "grad_norm": 8.33714485168457,
+      "learning_rate": 5.218703241895262e-06,
+      "loss": 0.3089,
+      "step": 59300
+    },
+    {
+      "epoch": 14.790523690773068,
+      "grad_norm": 7.40169095993042,
+      "learning_rate": 5.216209476309227e-06,
+      "loss": 0.3327,
+      "step": 59310
+    },
+    {
+      "epoch": 14.793017456359102,
+      "grad_norm": 6.180658340454102,
+      "learning_rate": 5.213715710723193e-06,
+      "loss": 0.3209,
+      "step": 59320
+    },
+    {
+      "epoch": 14.795511221945137,
+      "grad_norm": 8.041952133178711,
+      "learning_rate": 5.211221945137157e-06,
+      "loss": 0.2623,
+      "step": 59330
+    },
+    {
+      "epoch": 14.798004987531172,
+      "grad_norm": 7.726354598999023,
+      "learning_rate": 5.2087281795511225e-06,
+      "loss": 0.3279,
+      "step": 59340
+    },
+    {
+      "epoch": 14.800498753117207,
+      "grad_norm": 7.853705406188965,
+      "learning_rate": 5.206234413965088e-06,
+      "loss": 0.3187,
+      "step": 59350
+    },
+    {
+      "epoch": 14.802992518703242,
+      "grad_norm": 6.197792053222656,
+      "learning_rate": 5.203740648379053e-06,
+      "loss": 0.4057,
+      "step": 59360
+    },
+    {
+      "epoch": 14.805486284289277,
+      "grad_norm": 8.7774658203125,
+      "learning_rate": 5.2012468827930175e-06,
+      "loss": 0.3922,
+      "step": 59370
+    },
+    {
+      "epoch": 14.807980049875312,
+      "grad_norm": 7.630876541137695,
+      "learning_rate": 5.198753117206983e-06,
+      "loss": 0.3802,
+      "step": 59380
+    },
+    {
+      "epoch": 14.810473815461346,
+      "grad_norm": 9.605398178100586,
+      "learning_rate": 5.196259351620948e-06,
+      "loss": 0.3564,
+      "step": 59390
+    },
+    {
+      "epoch": 14.812967581047381,
+      "grad_norm": 11.865791320800781,
+      "learning_rate": 5.1937655860349135e-06,
+      "loss": 0.4148,
+      "step": 59400
+    },
+    {
+      "epoch": 14.815461346633416,
+      "grad_norm": 12.097999572753906,
+      "learning_rate": 5.191271820448879e-06,
+      "loss": 0.3287,
+      "step": 59410
+    },
+    {
+      "epoch": 14.817955112219451,
+      "grad_norm": 7.8101277351379395,
+      "learning_rate": 5.188778054862843e-06,
+      "loss": 0.3158,
+      "step": 59420
+    },
+    {
+      "epoch": 14.820448877805486,
+      "grad_norm": 6.201448917388916,
+      "learning_rate": 5.1862842892768085e-06,
+      "loss": 0.3469,
+      "step": 59430
+    },
+    {
+      "epoch": 14.82294264339152,
+      "grad_norm": 11.883674621582031,
+      "learning_rate": 5.183790523690774e-06,
+      "loss": 0.3443,
+      "step": 59440
+    },
+    {
+      "epoch": 14.825436408977556,
+      "grad_norm": 10.622366905212402,
+      "learning_rate": 5.181296758104739e-06,
+      "loss": 0.3804,
+      "step": 59450
+    },
+    {
+      "epoch": 14.82793017456359,
+      "grad_norm": 10.955135345458984,
+      "learning_rate": 5.178802992518704e-06,
+      "loss": 0.3149,
+      "step": 59460
+    },
+    {
+      "epoch": 14.830423940149625,
+      "grad_norm": 6.3978495597839355,
+      "learning_rate": 5.176309226932669e-06,
+      "loss": 0.3854,
+      "step": 59470
+    },
+    {
+      "epoch": 14.83291770573566,
+      "grad_norm": 5.390768527984619,
+      "learning_rate": 5.173815461346634e-06,
+      "loss": 0.3576,
+      "step": 59480
+    },
+    {
+      "epoch": 14.835411471321695,
+      "grad_norm": 6.621118068695068,
+      "learning_rate": 5.1713216957605996e-06,
+      "loss": 0.328,
+      "step": 59490
+    },
+    {
+      "epoch": 14.83790523690773,
+      "grad_norm": 7.241116523742676,
+      "learning_rate": 5.168827930174564e-06,
+      "loss": 0.3244,
+      "step": 59500
+    },
+    {
+      "epoch": 14.840399002493765,
+      "grad_norm": 9.426288604736328,
+      "learning_rate": 5.166334164588529e-06,
+      "loss": 0.358,
+      "step": 59510
+    },
+    {
+      "epoch": 14.8428927680798,
+      "grad_norm": 7.703991413116455,
+      "learning_rate": 5.163840399002495e-06,
+      "loss": 0.304,
+      "step": 59520
+    },
+    {
+      "epoch": 14.845386533665835,
+      "grad_norm": 5.415951251983643,
+      "learning_rate": 5.16134663341646e-06,
+      "loss": 0.2659,
+      "step": 59530
+    },
+    {
+      "epoch": 14.84788029925187,
+      "grad_norm": 10.298295021057129,
+      "learning_rate": 5.1588528678304236e-06,
+      "loss": 0.3267,
+      "step": 59540
+    },
+    {
+      "epoch": 14.850374064837904,
+      "grad_norm": 8.1857328414917,
+      "learning_rate": 5.156359102244389e-06,
+      "loss": 0.4304,
+      "step": 59550
+    },
+    {
+      "epoch": 14.85286783042394,
+      "grad_norm": 8.54920482635498,
+      "learning_rate": 5.153865336658355e-06,
+      "loss": 0.4247,
+      "step": 59560
+    },
+    {
+      "epoch": 14.855361596009976,
+      "grad_norm": 6.647639751434326,
+      "learning_rate": 5.15137157107232e-06,
+      "loss": 0.3184,
+      "step": 59570
+    },
+    {
+      "epoch": 14.85785536159601,
+      "grad_norm": 6.457409858703613,
+      "learning_rate": 5.148877805486284e-06,
+      "loss": 0.3595,
+      "step": 59580
+    },
+    {
+      "epoch": 14.860349127182046,
+      "grad_norm": 9.045681953430176,
+      "learning_rate": 5.146384039900249e-06,
+      "loss": 0.3741,
+      "step": 59590
+    },
+    {
+      "epoch": 14.86284289276808,
+      "grad_norm": 8.095196723937988,
+      "learning_rate": 5.1438902743142146e-06,
+      "loss": 0.3488,
+      "step": 59600
+    },
+    {
+      "epoch": 14.865336658354115,
+      "grad_norm": 12.992335319519043,
+      "learning_rate": 5.14139650872818e-06,
+      "loss": 0.3655,
+      "step": 59610
+    },
+    {
+      "epoch": 14.86783042394015,
+      "grad_norm": 8.386945724487305,
+      "learning_rate": 5.138902743142144e-06,
+      "loss": 0.3483,
+      "step": 59620
+    },
+    {
+      "epoch": 14.870324189526185,
+      "grad_norm": 7.935437202453613,
+      "learning_rate": 5.13640897755611e-06,
+      "loss": 0.3158,
+      "step": 59630
+    },
+    {
+      "epoch": 14.87281795511222,
+      "grad_norm": 6.484121322631836,
+      "learning_rate": 5.133915211970075e-06,
+      "loss": 0.3386,
+      "step": 59640
+    },
+    {
+      "epoch": 14.875311720698255,
+      "grad_norm": 8.329977035522461,
+      "learning_rate": 5.13142144638404e-06,
+      "loss": 0.2833,
+      "step": 59650
+    },
+    {
+      "epoch": 14.87780548628429,
+      "grad_norm": 6.874395370483398,
+      "learning_rate": 5.128927680798006e-06,
+      "loss": 0.3502,
+      "step": 59660
+    },
+    {
+      "epoch": 14.880299251870325,
+      "grad_norm": 7.59706974029541,
+      "learning_rate": 5.12643391521197e-06,
+      "loss": 0.3216,
+      "step": 59670
+    },
+    {
+      "epoch": 14.88279301745636,
+      "grad_norm": 7.426278591156006,
+      "learning_rate": 5.123940149625935e-06,
+      "loss": 0.297,
+      "step": 59680
+    },
+    {
+      "epoch": 14.885286783042394,
+      "grad_norm": 10.72069263458252,
+      "learning_rate": 5.121446384039901e-06,
+      "loss": 0.3663,
+      "step": 59690
+    },
+    {
+      "epoch": 14.88778054862843,
+      "grad_norm": 8.7599458694458,
+      "learning_rate": 5.118952618453866e-06,
+      "loss": 0.308,
+      "step": 59700
+    },
+    {
+      "epoch": 14.890274314214464,
+      "grad_norm": 7.034759521484375,
+      "learning_rate": 5.1164588528678304e-06,
+      "loss": 0.2957,
+      "step": 59710
+    },
+    {
+      "epoch": 14.892768079800499,
+      "grad_norm": 8.246294975280762,
+      "learning_rate": 5.113965087281796e-06,
+      "loss": 0.3802,
+      "step": 59720
+    },
+    {
+      "epoch": 14.895261845386534,
+      "grad_norm": 7.399212837219238,
+      "learning_rate": 5.111471321695761e-06,
+      "loss": 0.3906,
+      "step": 59730
+    },
+    {
+      "epoch": 14.897755610972569,
+      "grad_norm": 6.495842933654785,
+      "learning_rate": 5.108977556109726e-06,
+      "loss": 0.3517,
+      "step": 59740
+    },
+    {
+      "epoch": 14.900249376558603,
+      "grad_norm": 8.81147289276123,
+      "learning_rate": 5.106483790523691e-06,
+      "loss": 0.307,
+      "step": 59750
+    },
+    {
+      "epoch": 14.902743142144638,
+      "grad_norm": 8.447113990783691,
+      "learning_rate": 5.103990024937656e-06,
+      "loss": 0.3051,
+      "step": 59760
+    },
+    {
+      "epoch": 14.905236907730673,
+      "grad_norm": 7.5276288986206055,
+      "learning_rate": 5.1014962593516215e-06,
+      "loss": 0.3735,
+      "step": 59770
+    },
+    {
+      "epoch": 14.907730673316708,
+      "grad_norm": 5.61832857131958,
+      "learning_rate": 5.099002493765587e-06,
+      "loss": 0.2694,
+      "step": 59780
+    },
+    {
+      "epoch": 14.910224438902743,
+      "grad_norm": 7.429715633392334,
+      "learning_rate": 5.096508728179551e-06,
+      "loss": 0.3523,
+      "step": 59790
+    },
+    {
+      "epoch": 14.912718204488778,
+      "grad_norm": 6.293092727661133,
+      "learning_rate": 5.0940149625935165e-06,
+      "loss": 0.2826,
+      "step": 59800
+    },
+    {
+      "epoch": 14.915211970074813,
+      "grad_norm": 8.379650115966797,
+      "learning_rate": 5.091521197007482e-06,
+      "loss": 0.2812,
+      "step": 59810
+    },
+    {
+      "epoch": 14.917705735660848,
+      "grad_norm": 8.945039749145508,
+      "learning_rate": 5.089027431421447e-06,
+      "loss": 0.393,
+      "step": 59820
+    },
+    {
+      "epoch": 14.920199501246882,
+      "grad_norm": 5.737639904022217,
+      "learning_rate": 5.086533665835412e-06,
+      "loss": 0.3738,
+      "step": 59830
+    },
+    {
+      "epoch": 14.922693266832917,
+      "grad_norm": 10.540114402770996,
+      "learning_rate": 5.084039900249377e-06,
+      "loss": 0.3052,
+      "step": 59840
+    },
+    {
+      "epoch": 14.925187032418952,
+      "grad_norm": 7.083829879760742,
+      "learning_rate": 5.081546134663342e-06,
+      "loss": 0.3163,
+      "step": 59850
+    },
+    {
+      "epoch": 14.927680798004987,
+      "grad_norm": 9.188494682312012,
+      "learning_rate": 5.0790523690773075e-06,
+      "loss": 0.3344,
+      "step": 59860
+    },
+    {
+      "epoch": 14.930174563591022,
+      "grad_norm": 5.866366386413574,
+      "learning_rate": 5.076558603491272e-06,
+      "loss": 0.3339,
+      "step": 59870
+    },
+    {
+      "epoch": 14.932668329177057,
+      "grad_norm": 7.339391708374023,
+      "learning_rate": 5.074064837905237e-06,
+      "loss": 0.312,
+      "step": 59880
+    },
+    {
+      "epoch": 14.935162094763092,
+      "grad_norm": 8.920500755310059,
+      "learning_rate": 5.071571072319203e-06,
+      "loss": 0.338,
+      "step": 59890
+    },
+    {
+      "epoch": 14.937655860349127,
+      "grad_norm": 11.074267387390137,
+      "learning_rate": 5.069077306733168e-06,
+      "loss": 0.3554,
+      "step": 59900
+    },
+    {
+      "epoch": 14.940149625935161,
+      "grad_norm": 7.53389310836792,
+      "learning_rate": 5.066583541147133e-06,
+      "loss": 0.3171,
+      "step": 59910
+    },
+    {
+      "epoch": 14.942643391521196,
+      "grad_norm": 5.2787346839904785,
+      "learning_rate": 5.064089775561098e-06,
+      "loss": 0.3658,
+      "step": 59920
+    },
+    {
+      "epoch": 14.945137157107231,
+      "grad_norm": 10.729007720947266,
+      "learning_rate": 5.061596009975063e-06,
+      "loss": 0.3204,
+      "step": 59930
+    },
+    {
+      "epoch": 14.947630922693268,
+      "grad_norm": 4.510223865509033,
+      "learning_rate": 5.059102244389028e-06,
+      "loss": 0.3533,
+      "step": 59940
+    },
+    {
+      "epoch": 14.950124688279303,
+      "grad_norm": 8.356785774230957,
+      "learning_rate": 5.056608478802994e-06,
+      "loss": 0.3236,
+      "step": 59950
+    },
+    {
+      "epoch": 14.952618453865338,
+      "grad_norm": 8.854043006896973,
+      "learning_rate": 5.054114713216958e-06,
+      "loss": 0.3763,
+      "step": 59960
+    },
+    {
+      "epoch": 14.955112219451372,
+      "grad_norm": 7.022368907928467,
+      "learning_rate": 5.051620947630923e-06,
+      "loss": 0.3308,
+      "step": 59970
+    },
+    {
+      "epoch": 14.957605985037407,
+      "grad_norm": 7.40346622467041,
+      "learning_rate": 5.049127182044889e-06,
+      "loss": 0.2865,
+      "step": 59980
+    },
+    {
+      "epoch": 14.960099750623442,
+      "grad_norm": 6.403690338134766,
+      "learning_rate": 5.046633416458854e-06,
+      "loss": 0.3002,
+      "step": 59990
+    },
+    {
+      "epoch": 14.962593516209477,
+      "grad_norm": 4.785618782043457,
+      "learning_rate": 5.044139650872818e-06,
+      "loss": 0.3558,
+      "step": 60000
+    },
+    {
+      "epoch": 14.965087281795512,
+      "grad_norm": 8.796896934509277,
+      "learning_rate": 5.041645885286783e-06,
+      "loss": 0.3605,
+      "step": 60010
+    },
+    {
+      "epoch": 14.967581047381547,
+      "grad_norm": 8.200934410095215,
+      "learning_rate": 5.039152119700749e-06,
+      "loss": 0.3184,
+      "step": 60020
+    },
+    {
+      "epoch": 14.970074812967582,
+      "grad_norm": 9.922554016113281,
+      "learning_rate": 5.036658354114714e-06,
+      "loss": 0.3724,
+      "step": 60030
+    },
+    {
+      "epoch": 14.972568578553616,
+      "grad_norm": 8.850462913513184,
+      "learning_rate": 5.034164588528678e-06,
+      "loss": 0.3298,
+      "step": 60040
+    },
+    {
+      "epoch": 14.975062344139651,
+      "grad_norm": 4.286106586456299,
+      "learning_rate": 5.031670822942643e-06,
+      "loss": 0.3485,
+      "step": 60050
+    },
+    {
+      "epoch": 14.977556109725686,
+      "grad_norm": 8.860762596130371,
+      "learning_rate": 5.029177057356609e-06,
+      "loss": 0.3658,
+      "step": 60060
+    },
+    {
+      "epoch": 14.980049875311721,
+      "grad_norm": 11.523205757141113,
+      "learning_rate": 5.026683291770574e-06,
+      "loss": 0.3549,
+      "step": 60070
+    },
+    {
+      "epoch": 14.982543640897756,
+      "grad_norm": 7.7817277908325195,
+      "learning_rate": 5.0241895261845384e-06,
+      "loss": 0.287,
+      "step": 60080
+    },
+    {
+      "epoch": 14.98503740648379,
+      "grad_norm": 9.745018005371094,
+      "learning_rate": 5.021695760598504e-06,
+      "loss": 0.3697,
+      "step": 60090
+    },
+    {
+      "epoch": 14.987531172069826,
+      "grad_norm": 7.144688606262207,
+      "learning_rate": 5.019201995012469e-06,
+      "loss": 0.3164,
+      "step": 60100
+    },
+    {
+      "epoch": 14.99002493765586,
+      "grad_norm": 7.69403600692749,
+      "learning_rate": 5.016708229426434e-06,
+      "loss": 0.3714,
+      "step": 60110
+    },
+    {
+      "epoch": 14.992518703241895,
+      "grad_norm": 7.915139198303223,
+      "learning_rate": 5.014214463840399e-06,
+      "loss": 0.3015,
+      "step": 60120
+    },
+    {
+      "epoch": 14.99501246882793,
+      "grad_norm": 7.1008381843566895,
+      "learning_rate": 5.011720698254364e-06,
+      "loss": 0.314,
+      "step": 60130
+    },
+    {
+      "epoch": 14.997506234413965,
+      "grad_norm": 9.205920219421387,
+      "learning_rate": 5.0092269326683294e-06,
+      "loss": 0.3682,
+      "step": 60140
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 5.531941890716553,
+      "learning_rate": 5.006733167082295e-06,
+      "loss": 0.2957,
+      "step": 60150
+    },
+    {
+      "epoch": 15.0,
+      "eval_loss": 0.4161190390586853,
+      "eval_runtime": 60.2534,
+      "eval_samples_per_second": 16.646,
+      "eval_steps_per_second": 16.646,
+      "step": 60150
+    },
+    {
+      "epoch": 15.002493765586035,
+      "grad_norm": 5.979081153869629,
+      "learning_rate": 5.00423940149626e-06,
+      "loss": 0.3142,
+      "step": 60160
+    },
+    {
+      "epoch": 15.00498753117207,
+      "grad_norm": 14.604164123535156,
+      "learning_rate": 5.0017456359102245e-06,
+      "loss": 0.3561,
+      "step": 60170
+    },
+    {
+      "epoch": 15.007481296758105,
+      "grad_norm": 8.22153377532959,
+      "learning_rate": 4.99925187032419e-06,
+      "loss": 0.379,
+      "step": 60180
+    },
+    {
+      "epoch": 15.00997506234414,
+      "grad_norm": 8.119734764099121,
+      "learning_rate": 4.996758104738155e-06,
+      "loss": 0.2916,
+      "step": 60190
+    },
+    {
+      "epoch": 15.012468827930174,
+      "grad_norm": 8.341341972351074,
+      "learning_rate": 4.9942643391521205e-06,
+      "loss": 0.345,
+      "step": 60200
+    },
+    {
+      "epoch": 15.01496259351621,
+      "grad_norm": 9.086451530456543,
+      "learning_rate": 4.991770573566086e-06,
+      "loss": 0.3281,
+      "step": 60210
+    },
+    {
+      "epoch": 15.017456359102244,
+      "grad_norm": 6.594851970672607,
+      "learning_rate": 4.98927680798005e-06,
+      "loss": 0.3928,
+      "step": 60220
+    },
+    {
+      "epoch": 15.019950124688279,
+      "grad_norm": 4.85838508605957,
+      "learning_rate": 4.9867830423940155e-06,
+      "loss": 0.2773,
+      "step": 60230
+    },
+    {
+      "epoch": 15.022443890274314,
+      "grad_norm": 6.930510520935059,
+      "learning_rate": 4.98428927680798e-06,
+      "loss": 0.3527,
+      "step": 60240
+    },
+    {
+      "epoch": 15.024937655860349,
+      "grad_norm": 10.528151512145996,
+      "learning_rate": 4.981795511221945e-06,
+      "loss": 0.3468,
+      "step": 60250
+    },
+    {
+      "epoch": 15.027431421446384,
+      "grad_norm": 13.538934707641602,
+      "learning_rate": 4.979301745635911e-06,
+      "loss": 0.3329,
+      "step": 60260
+    },
+    {
+      "epoch": 15.029925187032418,
+      "grad_norm": 9.08583927154541,
+      "learning_rate": 4.976807980049876e-06,
+      "loss": 0.3067,
+      "step": 60270
+    },
+    {
+      "epoch": 15.032418952618453,
+      "grad_norm": 6.1941423416137695,
+      "learning_rate": 4.97431421446384e-06,
+      "loss": 0.3442,
+      "step": 60280
+    },
+    {
+      "epoch": 15.034912718204488,
+      "grad_norm": 8.902884483337402,
+      "learning_rate": 4.971820448877806e-06,
+      "loss": 0.2784,
+      "step": 60290
+    },
+    {
+      "epoch": 15.037406483790523,
+      "grad_norm": 8.3753023147583,
+      "learning_rate": 4.969326683291771e-06,
+      "loss": 0.3317,
+      "step": 60300
+    },
+    {
+      "epoch": 15.039900249376558,
+      "grad_norm": 9.228056907653809,
+      "learning_rate": 4.966832917705736e-06,
+      "loss": 0.3551,
+      "step": 60310
+    },
+    {
+      "epoch": 15.042394014962593,
+      "grad_norm": 8.301530838012695,
+      "learning_rate": 4.964339152119701e-06,
+      "loss": 0.4634,
+      "step": 60320
+    },
+    {
+      "epoch": 15.044887780548628,
+      "grad_norm": 7.827813625335693,
+      "learning_rate": 4.961845386533666e-06,
+      "loss": 0.336,
+      "step": 60330
+    },
+    {
+      "epoch": 15.047381546134662,
+      "grad_norm": 8.87314510345459,
+      "learning_rate": 4.959351620947631e-06,
+      "loss": 0.3343,
+      "step": 60340
+    },
+    {
+      "epoch": 15.049875311720697,
+      "grad_norm": 5.845686912536621,
+      "learning_rate": 4.956857855361597e-06,
+      "loss": 0.3175,
+      "step": 60350
+    },
+    {
+      "epoch": 15.052369077306734,
+      "grad_norm": 7.405905246734619,
+      "learning_rate": 4.954364089775561e-06,
+      "loss": 0.354,
+      "step": 60360
+    },
+    {
+      "epoch": 15.054862842892769,
+      "grad_norm": 6.437865734100342,
+      "learning_rate": 4.9518703241895265e-06,
+      "loss": 0.3171,
+      "step": 60370
+    },
+    {
+      "epoch": 15.057356608478804,
+      "grad_norm": 10.79374885559082,
+      "learning_rate": 4.949376558603492e-06,
+      "loss": 0.317,
+      "step": 60380
+    },
+    {
+      "epoch": 15.059850374064839,
+      "grad_norm": 8.71488094329834,
+      "learning_rate": 4.946882793017457e-06,
+      "loss": 0.3546,
+      "step": 60390
+    },
+    {
+      "epoch": 15.062344139650873,
+      "grad_norm": 7.864600658416748,
+      "learning_rate": 4.944389027431422e-06,
+      "loss": 0.4108,
+      "step": 60400
+    },
+    {
+      "epoch": 15.064837905236908,
+      "grad_norm": 6.649560928344727,
+      "learning_rate": 4.941895261845387e-06,
+      "loss": 0.3197,
+      "step": 60410
+    },
+    {
+      "epoch": 15.067331670822943,
+      "grad_norm": 9.850674629211426,
+      "learning_rate": 4.939401496259352e-06,
+      "loss": 0.3384,
+      "step": 60420
+    },
+    {
+      "epoch": 15.069825436408978,
+      "grad_norm": 16.706798553466797,
+      "learning_rate": 4.9369077306733175e-06,
+      "loss": 0.3184,
+      "step": 60430
+    },
+    {
+      "epoch": 15.072319201995013,
+      "grad_norm": 8.220061302185059,
+      "learning_rate": 4.934413965087283e-06,
+      "loss": 0.308,
+      "step": 60440
+    },
+    {
+      "epoch": 15.074812967581048,
+      "grad_norm": 6.942609786987305,
+      "learning_rate": 4.931920199501247e-06,
+      "loss": 0.3105,
+      "step": 60450
+    },
+    {
+      "epoch": 15.077306733167083,
+      "grad_norm": 8.182219505310059,
+      "learning_rate": 4.9294264339152126e-06,
+      "loss": 0.2868,
+      "step": 60460
+    },
+    {
+      "epoch": 15.079800498753118,
+      "grad_norm": 7.964657783508301,
+      "learning_rate": 4.926932668329177e-06,
+      "loss": 0.3713,
+      "step": 60470
+    },
+    {
+      "epoch": 15.082294264339152,
+      "grad_norm": 6.614937782287598,
+      "learning_rate": 4.924438902743142e-06,
+      "loss": 0.3498,
+      "step": 60480
+    },
+    {
+      "epoch": 15.084788029925187,
+      "grad_norm": 9.722759246826172,
+      "learning_rate": 4.921945137157108e-06,
+      "loss": 0.3951,
+      "step": 60490
+    },
+    {
+      "epoch": 15.087281795511222,
+      "grad_norm": 7.471789836883545,
+      "learning_rate": 4.919451371571073e-06,
+      "loss": 0.289,
+      "step": 60500
+    },
+    {
+      "epoch": 15.089775561097257,
+      "grad_norm": 8.353428840637207,
+      "learning_rate": 4.9169576059850374e-06,
+      "loss": 0.3236,
+      "step": 60510
+    },
+    {
+      "epoch": 15.092269326683292,
+      "grad_norm": 6.462151527404785,
+      "learning_rate": 4.914463840399003e-06,
+      "loss": 0.3103,
+      "step": 60520
+    },
+    {
+      "epoch": 15.094763092269327,
+      "grad_norm": 10.315178871154785,
+      "learning_rate": 4.911970074812968e-06,
+      "loss": 0.3587,
+      "step": 60530
+    },
+    {
+      "epoch": 15.097256857855362,
+      "grad_norm": 7.671100616455078,
+      "learning_rate": 4.909476309226933e-06,
+      "loss": 0.3657,
+      "step": 60540
+    },
+    {
+      "epoch": 15.099750623441397,
+      "grad_norm": 8.20209789276123,
+      "learning_rate": 4.906982543640898e-06,
+      "loss": 0.3923,
+      "step": 60550
+    },
+    {
+      "epoch": 15.102244389027431,
+      "grad_norm": 5.816895008087158,
+      "learning_rate": 4.904488778054863e-06,
+      "loss": 0.3362,
+      "step": 60560
+    },
+    {
+      "epoch": 15.104738154613466,
+      "grad_norm": 6.219974994659424,
+      "learning_rate": 4.9019950124688284e-06,
+      "loss": 0.4108,
+      "step": 60570
+    },
+    {
+      "epoch": 15.107231920199501,
+      "grad_norm": 8.702142715454102,
+      "learning_rate": 4.899501246882794e-06,
+      "loss": 0.3341,
+      "step": 60580
+    },
+    {
+      "epoch": 15.109725685785536,
+      "grad_norm": 7.942946434020996,
+      "learning_rate": 4.897007481296758e-06,
+      "loss": 0.298,
+      "step": 60590
+    },
+    {
+      "epoch": 15.11221945137157,
+      "grad_norm": 14.992578506469727,
+      "learning_rate": 4.8945137157107235e-06,
+      "loss": 0.3463,
+      "step": 60600
+    },
+    {
+      "epoch": 15.114713216957606,
+      "grad_norm": 8.40830135345459,
+      "learning_rate": 4.892019950124689e-06,
+      "loss": 0.3821,
+      "step": 60610
+    },
+    {
+      "epoch": 15.11720698254364,
+      "grad_norm": 9.050780296325684,
+      "learning_rate": 4.889526184538654e-06,
+      "loss": 0.3323,
+      "step": 60620
+    },
+    {
+      "epoch": 15.119700748129675,
+      "grad_norm": 3.6410436630249023,
+      "learning_rate": 4.887032418952619e-06,
+      "loss": 0.321,
+      "step": 60630
+    },
+    {
+      "epoch": 15.12219451371571,
+      "grad_norm": 10.186809539794922,
+      "learning_rate": 4.884538653366584e-06,
+      "loss": 0.3119,
+      "step": 60640
+    },
+    {
+      "epoch": 15.124688279301745,
+      "grad_norm": 9.280686378479004,
+      "learning_rate": 4.882044887780549e-06,
+      "loss": 0.2915,
+      "step": 60650
+    },
+    {
+      "epoch": 15.12718204488778,
+      "grad_norm": 10.575470924377441,
+      "learning_rate": 4.879551122194514e-06,
+      "loss": 0.3587,
+      "step": 60660
+    },
+    {
+      "epoch": 15.129675810473815,
+      "grad_norm": 10.077107429504395,
+      "learning_rate": 4.87705735660848e-06,
+      "loss": 0.4142,
+      "step": 60670
+    },
+    {
+      "epoch": 15.13216957605985,
+      "grad_norm": 8.206668853759766,
+      "learning_rate": 4.874563591022444e-06,
+      "loss": 0.3306,
+      "step": 60680
+    },
+    {
+      "epoch": 15.134663341645885,
+      "grad_norm": 6.796830654144287,
+      "learning_rate": 4.87206982543641e-06,
+      "loss": 0.3437,
+      "step": 60690
+    },
+    {
+      "epoch": 15.13715710723192,
+      "grad_norm": 8.173311233520508,
+      "learning_rate": 4.869576059850374e-06,
+      "loss": 0.3,
+      "step": 60700
+    },
+    {
+      "epoch": 15.139650872817954,
+      "grad_norm": 11.791805267333984,
+      "learning_rate": 4.867082294264339e-06,
+      "loss": 0.2989,
+      "step": 60710
+    },
+    {
+      "epoch": 15.14214463840399,
+      "grad_norm": 8.304753303527832,
+      "learning_rate": 4.864588528678305e-06,
+      "loss": 0.3687,
+      "step": 60720
+    },
+    {
+      "epoch": 15.144638403990024,
+      "grad_norm": 6.197001934051514,
+      "learning_rate": 4.86209476309227e-06,
+      "loss": 0.3366,
+      "step": 60730
+    },
+    {
+      "epoch": 15.147132169576059,
+      "grad_norm": 11.391712188720703,
+      "learning_rate": 4.8596009975062345e-06,
+      "loss": 0.3422,
+      "step": 60740
+    },
+    {
+      "epoch": 15.149625935162096,
+      "grad_norm": 5.738236904144287,
+      "learning_rate": 4.8571072319202e-06,
+      "loss": 0.3359,
+      "step": 60750
+    },
+    {
+      "epoch": 15.15211970074813,
+      "grad_norm": 9.785381317138672,
+      "learning_rate": 4.854613466334165e-06,
+      "loss": 0.3786,
+      "step": 60760
+    },
+    {
+      "epoch": 15.154613466334165,
+      "grad_norm": 6.396745204925537,
+      "learning_rate": 4.85211970074813e-06,
+      "loss": 0.2676,
+      "step": 60770
+    },
+    {
+      "epoch": 15.1571072319202,
+      "grad_norm": 7.886439800262451,
+      "learning_rate": 4.849625935162095e-06,
+      "loss": 0.3038,
+      "step": 60780
+    },
+    {
+      "epoch": 15.159600997506235,
+      "grad_norm": 11.808472633361816,
+      "learning_rate": 4.84713216957606e-06,
+      "loss": 0.3218,
+      "step": 60790
+    },
+    {
+      "epoch": 15.16209476309227,
+      "grad_norm": 13.33260440826416,
+      "learning_rate": 4.8446384039900255e-06,
+      "loss": 0.3795,
+      "step": 60800
+    },
+    {
+      "epoch": 15.164588528678305,
+      "grad_norm": 6.950128078460693,
+      "learning_rate": 4.842144638403991e-06,
+      "loss": 0.3344,
+      "step": 60810
+    },
+    {
+      "epoch": 15.16708229426434,
+      "grad_norm": 10.497044563293457,
+      "learning_rate": 4.839650872817955e-06,
+      "loss": 0.3543,
+      "step": 60820
+    },
+    {
+      "epoch": 15.169576059850375,
+      "grad_norm": 10.250978469848633,
+      "learning_rate": 4.8371571072319206e-06,
+      "loss": 0.3404,
+      "step": 60830
+    },
+    {
+      "epoch": 15.17206982543641,
+      "grad_norm": 9.228592872619629,
+      "learning_rate": 4.834663341645886e-06,
+      "loss": 0.3418,
+      "step": 60840
+    },
+    {
+      "epoch": 15.174563591022444,
+      "grad_norm": 9.230294227600098,
+      "learning_rate": 4.832169576059851e-06,
+      "loss": 0.3317,
+      "step": 60850
+    },
+    {
+      "epoch": 15.17705735660848,
+      "grad_norm": 6.274941921234131,
+      "learning_rate": 4.829675810473816e-06,
+      "loss": 0.3361,
+      "step": 60860
+    },
+    {
+      "epoch": 15.179551122194514,
+      "grad_norm": 7.867391586303711,
+      "learning_rate": 4.827182044887781e-06,
+      "loss": 0.3096,
+      "step": 60870
+    },
+    {
+      "epoch": 15.182044887780549,
+      "grad_norm": 7.890642166137695,
+      "learning_rate": 4.824688279301745e-06,
+      "loss": 0.3399,
+      "step": 60880
+    },
+    {
+      "epoch": 15.184538653366584,
+      "grad_norm": 9.05803394317627,
+      "learning_rate": 4.822194513715711e-06,
+      "loss": 0.3148,
+      "step": 60890
+    },
+    {
+      "epoch": 15.187032418952619,
+      "grad_norm": 5.658694744110107,
+      "learning_rate": 4.819700748129677e-06,
+      "loss": 0.3686,
+      "step": 60900
+    },
+    {
+      "epoch": 15.189526184538654,
+      "grad_norm": 7.717238426208496,
+      "learning_rate": 4.817206982543641e-06,
+      "loss": 0.3431,
+      "step": 60910
+    },
+    {
+      "epoch": 15.192019950124688,
+      "grad_norm": 9.4716215133667,
+      "learning_rate": 4.814713216957607e-06,
+      "loss": 0.3377,
+      "step": 60920
+    },
+    {
+      "epoch": 15.194513715710723,
+      "grad_norm": 6.269632816314697,
+      "learning_rate": 4.812219451371571e-06,
+      "loss": 0.3173,
+      "step": 60930
+    },
+    {
+      "epoch": 15.197007481296758,
+      "grad_norm": 4.857752799987793,
+      "learning_rate": 4.8097256857855364e-06,
+      "loss": 0.2644,
+      "step": 60940
+    },
+    {
+      "epoch": 15.199501246882793,
+      "grad_norm": 4.956846714019775,
+      "learning_rate": 4.807231920199502e-06,
+      "loss": 0.4057,
+      "step": 60950
+    },
+    {
+      "epoch": 15.201995012468828,
+      "grad_norm": 11.810905456542969,
+      "learning_rate": 4.804738154613467e-06,
+      "loss": 0.3344,
+      "step": 60960
+    },
+    {
+      "epoch": 15.204488778054863,
+      "grad_norm": 10.063115119934082,
+      "learning_rate": 4.8022443890274315e-06,
+      "loss": 0.356,
+      "step": 60970
+    },
+    {
+      "epoch": 15.206982543640898,
+      "grad_norm": 8.853325843811035,
+      "learning_rate": 4.799750623441397e-06,
+      "loss": 0.2966,
+      "step": 60980
+    },
+    {
+      "epoch": 15.209476309226932,
+      "grad_norm": 7.260303974151611,
+      "learning_rate": 4.797256857855362e-06,
+      "loss": 0.2996,
+      "step": 60990
+    },
+    {
+      "epoch": 15.211970074812967,
+      "grad_norm": 9.30881118774414,
+      "learning_rate": 4.7947630922693274e-06,
+      "loss": 0.3598,
+      "step": 61000
+    },
+    {
+      "epoch": 15.214463840399002,
+      "grad_norm": 8.756742477416992,
+      "learning_rate": 4.792269326683292e-06,
+      "loss": 0.2864,
+      "step": 61010
+    },
+    {
+      "epoch": 15.216957605985037,
+      "grad_norm": 6.806258201599121,
+      "learning_rate": 4.789775561097257e-06,
+      "loss": 0.3255,
+      "step": 61020
+    },
+    {
+      "epoch": 15.219451371571072,
+      "grad_norm": 7.81227445602417,
+      "learning_rate": 4.7872817955112225e-06,
+      "loss": 0.42,
+      "step": 61030
+    },
+    {
+      "epoch": 15.221945137157107,
+      "grad_norm": 8.680373191833496,
+      "learning_rate": 4.784788029925188e-06,
+      "loss": 0.3207,
+      "step": 61040
+    },
+    {
+      "epoch": 15.224438902743142,
+      "grad_norm": 7.925756454467773,
+      "learning_rate": 4.782294264339152e-06,
+      "loss": 0.3674,
+      "step": 61050
+    },
+    {
+      "epoch": 15.226932668329177,
+      "grad_norm": 9.870626449584961,
+      "learning_rate": 4.779800498753118e-06,
+      "loss": 0.3326,
+      "step": 61060
+    },
+    {
+      "epoch": 15.229426433915211,
+      "grad_norm": 6.636346340179443,
+      "learning_rate": 4.777306733167082e-06,
+      "loss": 0.2297,
+      "step": 61070
+    },
+    {
+      "epoch": 15.231920199501246,
+      "grad_norm": 7.3385009765625,
+      "learning_rate": 4.774812967581048e-06,
+      "loss": 0.3006,
+      "step": 61080
+    },
+    {
+      "epoch": 15.234413965087281,
+      "grad_norm": 7.8916215896606445,
+      "learning_rate": 4.772319201995013e-06,
+      "loss": 0.3171,
+      "step": 61090
+    },
+    {
+      "epoch": 15.236907730673316,
+      "grad_norm": 7.509432792663574,
+      "learning_rate": 4.769825436408978e-06,
+      "loss": 0.3321,
+      "step": 61100
+    },
+    {
+      "epoch": 15.239401496259351,
+      "grad_norm": 5.825787544250488,
+      "learning_rate": 4.7673316708229425e-06,
+      "loss": 0.3242,
+      "step": 61110
+    },
+    {
+      "epoch": 15.241895261845386,
+      "grad_norm": 43.301937103271484,
+      "learning_rate": 4.764837905236908e-06,
+      "loss": 0.3342,
+      "step": 61120
+    },
+    {
+      "epoch": 15.24438902743142,
+      "grad_norm": 5.396152019500732,
+      "learning_rate": 4.762344139650873e-06,
+      "loss": 0.3322,
+      "step": 61130
+    },
+    {
+      "epoch": 15.246882793017456,
+      "grad_norm": 7.503886699676514,
+      "learning_rate": 4.759850374064838e-06,
+      "loss": 0.3069,
+      "step": 61140
+    },
+    {
+      "epoch": 15.24937655860349,
+      "grad_norm": 9.121578216552734,
+      "learning_rate": 4.757356608478804e-06,
+      "loss": 0.3596,
+      "step": 61150
+    },
+    {
+      "epoch": 15.251870324189527,
+      "grad_norm": 8.742789268493652,
+      "learning_rate": 4.754862842892768e-06,
+      "loss": 0.2804,
+      "step": 61160
+    },
+    {
+      "epoch": 15.254364089775562,
+      "grad_norm": 7.584442138671875,
+      "learning_rate": 4.7523690773067335e-06,
+      "loss": 0.2849,
+      "step": 61170
+    },
+    {
+      "epoch": 15.256857855361597,
+      "grad_norm": 6.909498691558838,
+      "learning_rate": 4.749875311720699e-06,
+      "loss": 0.2826,
+      "step": 61180
+    },
+    {
+      "epoch": 15.259351620947632,
+      "grad_norm": 6.300384044647217,
+      "learning_rate": 4.747381546134664e-06,
+      "loss": 0.3083,
+      "step": 61190
+    },
+    {
+      "epoch": 15.261845386533667,
+      "grad_norm": 7.162856101989746,
+      "learning_rate": 4.7448877805486286e-06,
+      "loss": 0.2551,
+      "step": 61200
+    },
+    {
+      "epoch": 15.264339152119701,
+      "grad_norm": 8.662135124206543,
+      "learning_rate": 4.742394014962594e-06,
+      "loss": 0.3813,
+      "step": 61210
+    },
+    {
+      "epoch": 15.266832917705736,
+      "grad_norm": 7.602468013763428,
+      "learning_rate": 4.739900249376559e-06,
+      "loss": 0.3187,
+      "step": 61220
+    },
+    {
+      "epoch": 15.269326683291771,
+      "grad_norm": 8.263647079467773,
+      "learning_rate": 4.7374064837905245e-06,
+      "loss": 0.3223,
+      "step": 61230
+    },
+    {
+      "epoch": 15.271820448877806,
+      "grad_norm": 8.883642196655273,
+      "learning_rate": 4.734912718204489e-06,
+      "loss": 0.4101,
+      "step": 61240
+    },
+    {
+      "epoch": 15.27431421446384,
+      "grad_norm": 13.972535133361816,
+      "learning_rate": 4.732418952618454e-06,
+      "loss": 0.3395,
+      "step": 61250
+    },
+    {
+      "epoch": 15.276807980049876,
+      "grad_norm": 12.901040077209473,
+      "learning_rate": 4.7299251870324196e-06,
+      "loss": 0.3364,
+      "step": 61260
+    },
+    {
+      "epoch": 15.27930174563591,
+      "grad_norm": 9.556396484375,
+      "learning_rate": 4.727431421446385e-06,
+      "loss": 0.3784,
+      "step": 61270
+    },
+    {
+      "epoch": 15.281795511221945,
+      "grad_norm": 9.603511810302734,
+      "learning_rate": 4.724937655860349e-06,
+      "loss": 0.321,
+      "step": 61280
+    },
+    {
+      "epoch": 15.28428927680798,
+      "grad_norm": 12.59118366241455,
+      "learning_rate": 4.722443890274315e-06,
+      "loss": 0.3617,
+      "step": 61290
+    },
+    {
+      "epoch": 15.286783042394015,
+      "grad_norm": 7.6207194328308105,
+      "learning_rate": 4.719950124688279e-06,
+      "loss": 0.2976,
+      "step": 61300
+    },
+    {
+      "epoch": 15.28927680798005,
+      "grad_norm": 10.782099723815918,
+      "learning_rate": 4.717456359102245e-06,
+      "loss": 0.2937,
+      "step": 61310
+    },
+    {
+      "epoch": 15.291770573566085,
+      "grad_norm": 7.862745761871338,
+      "learning_rate": 4.71496259351621e-06,
+      "loss": 0.2923,
+      "step": 61320
+    },
+    {
+      "epoch": 15.29426433915212,
+      "grad_norm": 10.322379112243652,
+      "learning_rate": 4.712468827930175e-06,
+      "loss": 0.3436,
+      "step": 61330
+    },
+    {
+      "epoch": 15.296758104738155,
+      "grad_norm": 6.147068500518799,
+      "learning_rate": 4.7099750623441395e-06,
+      "loss": 0.3373,
+      "step": 61340
+    },
+    {
+      "epoch": 15.29925187032419,
+      "grad_norm": 7.12548303604126,
+      "learning_rate": 4.707481296758105e-06,
+      "loss": 0.3147,
+      "step": 61350
+    },
+    {
+      "epoch": 15.301745635910224,
+      "grad_norm": 10.481398582458496,
+      "learning_rate": 4.70498753117207e-06,
+      "loss": 0.3268,
+      "step": 61360
+    },
+    {
+      "epoch": 15.30423940149626,
+      "grad_norm": 8.344328880310059,
+      "learning_rate": 4.7024937655860354e-06,
+      "loss": 0.3231,
+      "step": 61370
+    },
+    {
+      "epoch": 15.306733167082294,
+      "grad_norm": 5.839080333709717,
+      "learning_rate": 4.7e-06,
+      "loss": 0.2777,
+      "step": 61380
+    },
+    {
+      "epoch": 15.309226932668329,
+      "grad_norm": 14.974725723266602,
+      "learning_rate": 4.697506234413965e-06,
+      "loss": 0.3513,
+      "step": 61390
+    },
+    {
+      "epoch": 15.311720698254364,
+      "grad_norm": 6.981603145599365,
+      "learning_rate": 4.6950124688279305e-06,
+      "loss": 0.3106,
+      "step": 61400
+    },
+    {
+      "epoch": 15.314214463840399,
+      "grad_norm": 10.228157043457031,
+      "learning_rate": 4.692518703241896e-06,
+      "loss": 0.3035,
+      "step": 61410
+    },
+    {
+      "epoch": 15.316708229426434,
+      "grad_norm": 10.667278289794922,
+      "learning_rate": 4.690024937655861e-06,
+      "loss": 0.331,
+      "step": 61420
+    },
+    {
+      "epoch": 15.319201995012468,
+      "grad_norm": 7.606900215148926,
+      "learning_rate": 4.687531172069826e-06,
+      "loss": 0.2967,
+      "step": 61430
+    },
+    {
+      "epoch": 15.321695760598503,
+      "grad_norm": 8.036834716796875,
+      "learning_rate": 4.685037406483791e-06,
+      "loss": 0.3349,
+      "step": 61440
+    },
+    {
+      "epoch": 15.324189526184538,
+      "grad_norm": 7.838779449462891,
+      "learning_rate": 4.682543640897756e-06,
+      "loss": 0.3439,
+      "step": 61450
+    },
+    {
+      "epoch": 15.326683291770573,
+      "grad_norm": 13.692665100097656,
+      "learning_rate": 4.6800498753117215e-06,
+      "loss": 0.3764,
+      "step": 61460
+    },
+    {
+      "epoch": 15.329177057356608,
+      "grad_norm": 9.359086036682129,
+      "learning_rate": 4.677556109725686e-06,
+      "loss": 0.3118,
+      "step": 61470
+    },
+    {
+      "epoch": 15.331670822942643,
+      "grad_norm": 7.252995014190674,
+      "learning_rate": 4.675062344139651e-06,
+      "loss": 0.311,
+      "step": 61480
+    },
+    {
+      "epoch": 15.334164588528678,
+      "grad_norm": 7.640556812286377,
+      "learning_rate": 4.672568578553617e-06,
+      "loss": 0.3451,
+      "step": 61490
+    },
+    {
+      "epoch": 15.336658354114713,
+      "grad_norm": 8.669981956481934,
+      "learning_rate": 4.670074812967582e-06,
+      "loss": 0.3612,
+      "step": 61500
+    },
+    {
+      "epoch": 15.339152119700747,
+      "grad_norm": 12.157310485839844,
+      "learning_rate": 4.667581047381546e-06,
+      "loss": 0.3596,
+      "step": 61510
+    },
+    {
+      "epoch": 15.341645885286782,
+      "grad_norm": 7.998473644256592,
+      "learning_rate": 4.665087281795512e-06,
+      "loss": 0.3214,
+      "step": 61520
+    },
+    {
+      "epoch": 15.344139650872817,
+      "grad_norm": 11.854910850524902,
+      "learning_rate": 4.662593516209476e-06,
+      "loss": 0.3224,
+      "step": 61530
+    },
+    {
+      "epoch": 15.346633416458852,
+      "grad_norm": 8.815274238586426,
+      "learning_rate": 4.660099750623442e-06,
+      "loss": 0.317,
+      "step": 61540
+    },
+    {
+      "epoch": 15.349127182044889,
+      "grad_norm": 6.209719181060791,
+      "learning_rate": 4.657605985037407e-06,
+      "loss": 0.3578,
+      "step": 61550
+    },
+    {
+      "epoch": 15.351620947630924,
+      "grad_norm": 10.795744895935059,
+      "learning_rate": 4.655112219451372e-06,
+      "loss": 0.2813,
+      "step": 61560
+    },
+    {
+      "epoch": 15.354114713216958,
+      "grad_norm": 8.15768814086914,
+      "learning_rate": 4.6526184538653365e-06,
+      "loss": 0.3336,
+      "step": 61570
+    },
+    {
+      "epoch": 15.356608478802993,
+      "grad_norm": 8.229902267456055,
+      "learning_rate": 4.650124688279302e-06,
+      "loss": 0.291,
+      "step": 61580
+    },
+    {
+      "epoch": 15.359102244389028,
+      "grad_norm": 8.55030632019043,
+      "learning_rate": 4.647630922693267e-06,
+      "loss": 0.2709,
+      "step": 61590
+    },
+    {
+      "epoch": 15.361596009975063,
+      "grad_norm": 6.729324817657471,
+      "learning_rate": 4.6451371571072325e-06,
+      "loss": 0.3014,
+      "step": 61600
+    },
+    {
+      "epoch": 15.364089775561098,
+      "grad_norm": 9.249410629272461,
+      "learning_rate": 4.642643391521197e-06,
+      "loss": 0.3409,
+      "step": 61610
+    },
+    {
+      "epoch": 15.366583541147133,
+      "grad_norm": 8.171239852905273,
+      "learning_rate": 4.640149625935162e-06,
+      "loss": 0.2956,
+      "step": 61620
+    },
+    {
+      "epoch": 15.369077306733168,
+      "grad_norm": 8.610503196716309,
+      "learning_rate": 4.6376558603491276e-06,
+      "loss": 0.2898,
+      "step": 61630
+    },
+    {
+      "epoch": 15.371571072319203,
+      "grad_norm": 6.898585796356201,
+      "learning_rate": 4.635162094763093e-06,
+      "loss": 0.3064,
+      "step": 61640
+    },
+    {
+      "epoch": 15.374064837905237,
+      "grad_norm": 10.55229377746582,
+      "learning_rate": 4.632668329177058e-06,
+      "loss": 0.3469,
+      "step": 61650
+    },
+    {
+      "epoch": 15.376558603491272,
+      "grad_norm": 13.5460786819458,
+      "learning_rate": 4.630174563591023e-06,
+      "loss": 0.3784,
+      "step": 61660
+    },
+    {
+      "epoch": 15.379052369077307,
+      "grad_norm": 6.476283073425293,
+      "learning_rate": 4.627680798004988e-06,
+      "loss": 0.2643,
+      "step": 61670
+    },
+    {
+      "epoch": 15.381546134663342,
+      "grad_norm": 13.659258842468262,
+      "learning_rate": 4.625187032418953e-06,
+      "loss": 0.4079,
+      "step": 61680
+    },
+    {
+      "epoch": 15.384039900249377,
+      "grad_norm": 10.668188095092773,
+      "learning_rate": 4.6226932668329186e-06,
+      "loss": 0.3074,
+      "step": 61690
+    },
+    {
+      "epoch": 15.386533665835412,
+      "grad_norm": 8.238740921020508,
+      "learning_rate": 4.620199501246883e-06,
+      "loss": 0.304,
+      "step": 61700
+    },
+    {
+      "epoch": 15.389027431421447,
+      "grad_norm": 7.179644584655762,
+      "learning_rate": 4.617705735660848e-06,
+      "loss": 0.3141,
+      "step": 61710
+    },
+    {
+      "epoch": 15.391521197007481,
+      "grad_norm": 6.697994709014893,
+      "learning_rate": 4.615211970074814e-06,
+      "loss": 0.3026,
+      "step": 61720
+    },
+    {
+      "epoch": 15.394014962593516,
+      "grad_norm": 8.70664119720459,
+      "learning_rate": 4.612718204488779e-06,
+      "loss": 0.3417,
+      "step": 61730
+    },
+    {
+      "epoch": 15.396508728179551,
+      "grad_norm": 6.116442680358887,
+      "learning_rate": 4.610224438902743e-06,
+      "loss": 0.3057,
+      "step": 61740
+    },
+    {
+      "epoch": 15.399002493765586,
+      "grad_norm": 7.209672927856445,
+      "learning_rate": 4.607730673316709e-06,
+      "loss": 0.2968,
+      "step": 61750
+    },
+    {
+      "epoch": 15.401496259351621,
+      "grad_norm": 9.802582740783691,
+      "learning_rate": 4.605236907730673e-06,
+      "loss": 0.2957,
+      "step": 61760
+    },
+    {
+      "epoch": 15.403990024937656,
+      "grad_norm": 15.796855926513672,
+      "learning_rate": 4.6027431421446385e-06,
+      "loss": 0.4296,
+      "step": 61770
+    },
+    {
+      "epoch": 15.40648379052369,
+      "grad_norm": 6.593606948852539,
+      "learning_rate": 4.600249376558604e-06,
+      "loss": 0.3552,
+      "step": 61780
+    },
+    {
+      "epoch": 15.408977556109726,
+      "grad_norm": 6.938211441040039,
+      "learning_rate": 4.597755610972569e-06,
+      "loss": 0.3235,
+      "step": 61790
+    },
+    {
+      "epoch": 15.41147132169576,
+      "grad_norm": 9.084220886230469,
+      "learning_rate": 4.595261845386534e-06,
+      "loss": 0.3951,
+      "step": 61800
+    },
+    {
+      "epoch": 15.413965087281795,
+      "grad_norm": 8.550397872924805,
+      "learning_rate": 4.592768079800499e-06,
+      "loss": 0.3256,
+      "step": 61810
+    },
+    {
+      "epoch": 15.41645885286783,
+      "grad_norm": 8.361339569091797,
+      "learning_rate": 4.590274314214464e-06,
+      "loss": 0.3312,
+      "step": 61820
+    },
+    {
+      "epoch": 15.418952618453865,
+      "grad_norm": 6.701113700866699,
+      "learning_rate": 4.5877805486284295e-06,
+      "loss": 0.2941,
+      "step": 61830
+    },
+    {
+      "epoch": 15.4214463840399,
+      "grad_norm": 14.242850303649902,
+      "learning_rate": 4.585286783042394e-06,
+      "loss": 0.3327,
+      "step": 61840
+    },
+    {
+      "epoch": 15.423940149625935,
+      "grad_norm": 8.893102645874023,
+      "learning_rate": 4.582793017456359e-06,
+      "loss": 0.3021,
+      "step": 61850
+    },
+    {
+      "epoch": 15.42643391521197,
+      "grad_norm": 9.866777420043945,
+      "learning_rate": 4.580299251870325e-06,
+      "loss": 0.3488,
+      "step": 61860
+    },
+    {
+      "epoch": 15.428927680798004,
+      "grad_norm": 9.676253318786621,
+      "learning_rate": 4.57780548628429e-06,
+      "loss": 0.3901,
+      "step": 61870
+    },
+    {
+      "epoch": 15.43142144638404,
+      "grad_norm": 6.8704657554626465,
+      "learning_rate": 4.575311720698254e-06,
+      "loss": 0.3065,
+      "step": 61880
+    },
+    {
+      "epoch": 15.433915211970074,
+      "grad_norm": 8.320914268493652,
+      "learning_rate": 4.57281795511222e-06,
+      "loss": 0.3204,
+      "step": 61890
+    },
+    {
+      "epoch": 15.436408977556109,
+      "grad_norm": 9.73550033569336,
+      "learning_rate": 4.570324189526185e-06,
+      "loss": 0.3785,
+      "step": 61900
+    },
+    {
+      "epoch": 15.438902743142144,
+      "grad_norm": 7.0275139808654785,
+      "learning_rate": 4.56783042394015e-06,
+      "loss": 0.3277,
+      "step": 61910
+    },
+    {
+      "epoch": 15.441396508728179,
+      "grad_norm": 8.163436889648438,
+      "learning_rate": 4.565336658354116e-06,
+      "loss": 0.3182,
+      "step": 61920
+    },
+    {
+      "epoch": 15.443890274314214,
+      "grad_norm": 7.171289920806885,
+      "learning_rate": 4.56284289276808e-06,
+      "loss": 0.3612,
+      "step": 61930
+    },
+    {
+      "epoch": 15.446384039900249,
+      "grad_norm": 11.349479675292969,
+      "learning_rate": 4.560349127182045e-06,
+      "loss": 0.3601,
+      "step": 61940
+    },
+    {
+      "epoch": 15.448877805486283,
+      "grad_norm": 5.093303680419922,
+      "learning_rate": 4.557855361596011e-06,
+      "loss": 0.3162,
+      "step": 61950
+    },
+    {
+      "epoch": 15.451371571072318,
+      "grad_norm": 8.058874130249023,
+      "learning_rate": 4.555361596009976e-06,
+      "loss": 0.3508,
+      "step": 61960
+    },
+    {
+      "epoch": 15.453865336658355,
+      "grad_norm": 7.41152286529541,
+      "learning_rate": 4.5528678304239405e-06,
+      "loss": 0.256,
+      "step": 61970
+    },
+    {
+      "epoch": 15.45635910224439,
+      "grad_norm": 5.370427131652832,
+      "learning_rate": 4.550374064837906e-06,
+      "loss": 0.3207,
+      "step": 61980
+    },
+    {
+      "epoch": 15.458852867830425,
+      "grad_norm": 9.166321754455566,
+      "learning_rate": 4.54788029925187e-06,
+      "loss": 0.3825,
+      "step": 61990
+    },
+    {
+      "epoch": 15.46134663341646,
+      "grad_norm": 11.046178817749023,
+      "learning_rate": 4.5453865336658355e-06,
+      "loss": 0.3506,
+      "step": 62000
+    },
+    {
+      "epoch": 15.463840399002494,
+      "grad_norm": 10.814038276672363,
+      "learning_rate": 4.542892768079801e-06,
+      "loss": 0.3217,
+      "step": 62010
+    },
+    {
+      "epoch": 15.46633416458853,
+      "grad_norm": 6.165196895599365,
+      "learning_rate": 4.540399002493766e-06,
+      "loss": 0.321,
+      "step": 62020
+    },
+    {
+      "epoch": 15.468827930174564,
+      "grad_norm": 6.679711818695068,
+      "learning_rate": 4.537905236907731e-06,
+      "loss": 0.3212,
+      "step": 62030
+    },
+    {
+      "epoch": 15.471321695760599,
+      "grad_norm": 10.853338241577148,
+      "learning_rate": 4.535411471321696e-06,
+      "loss": 0.2962,
+      "step": 62040
+    },
+    {
+      "epoch": 15.473815461346634,
+      "grad_norm": 7.180294990539551,
+      "learning_rate": 4.532917705735661e-06,
+      "loss": 0.3367,
+      "step": 62050
+    },
+    {
+      "epoch": 15.476309226932669,
+      "grad_norm": 6.090845108032227,
+      "learning_rate": 4.5304239401496266e-06,
+      "loss": 0.3281,
+      "step": 62060
+    },
+    {
+      "epoch": 15.478802992518704,
+      "grad_norm": 9.395833015441895,
+      "learning_rate": 4.527930174563591e-06,
+      "loss": 0.3029,
+      "step": 62070
+    },
+    {
+      "epoch": 15.481296758104738,
+      "grad_norm": 8.523276329040527,
+      "learning_rate": 4.525436408977556e-06,
+      "loss": 0.3147,
+      "step": 62080
+    },
+    {
+      "epoch": 15.483790523690773,
+      "grad_norm": 6.579370498657227,
+      "learning_rate": 4.522942643391522e-06,
+      "loss": 0.3089,
+      "step": 62090
+    },
+    {
+      "epoch": 15.486284289276808,
+      "grad_norm": 9.35316276550293,
+      "learning_rate": 4.520448877805487e-06,
+      "loss": 0.3479,
+      "step": 62100
+    },
+    {
+      "epoch": 15.488778054862843,
+      "grad_norm": 8.047480583190918,
+      "learning_rate": 4.517955112219451e-06,
+      "loss": 0.3327,
+      "step": 62110
+    },
+    {
+      "epoch": 15.491271820448878,
+      "grad_norm": 9.682567596435547,
+      "learning_rate": 4.515461346633417e-06,
+      "loss": 0.2934,
+      "step": 62120
+    },
+    {
+      "epoch": 15.493765586034913,
+      "grad_norm": 6.29941463470459,
+      "learning_rate": 4.512967581047382e-06,
+      "loss": 0.3575,
+      "step": 62130
+    },
+    {
+      "epoch": 15.496259351620948,
+      "grad_norm": 6.268800735473633,
+      "learning_rate": 4.510473815461347e-06,
+      "loss": 0.301,
+      "step": 62140
+    },
+    {
+      "epoch": 15.498753117206983,
+      "grad_norm": 9.379264831542969,
+      "learning_rate": 4.507980049875313e-06,
+      "loss": 0.3548,
+      "step": 62150
+    },
+    {
+      "epoch": 15.501246882793017,
+      "grad_norm": 7.664111137390137,
+      "learning_rate": 4.505486284289277e-06,
+      "loss": 0.2977,
+      "step": 62160
+    },
+    {
+      "epoch": 15.503740648379052,
+      "grad_norm": 5.527115821838379,
+      "learning_rate": 4.502992518703242e-06,
+      "loss": 0.3337,
+      "step": 62170
+    },
+    {
+      "epoch": 15.506234413965087,
+      "grad_norm": 9.211461067199707,
+      "learning_rate": 4.500498753117207e-06,
+      "loss": 0.3509,
+      "step": 62180
+    },
+    {
+      "epoch": 15.508728179551122,
+      "grad_norm": 8.124340057373047,
+      "learning_rate": 4.498004987531173e-06,
+      "loss": 0.37,
+      "step": 62190
+    },
+    {
+      "epoch": 15.511221945137157,
+      "grad_norm": 8.361662864685059,
+      "learning_rate": 4.4955112219451375e-06,
+      "loss": 0.3075,
+      "step": 62200
+    },
+    {
+      "epoch": 15.513715710723192,
+      "grad_norm": 9.21060848236084,
+      "learning_rate": 4.493017456359103e-06,
+      "loss": 0.277,
+      "step": 62210
+    },
+    {
+      "epoch": 15.516209476309227,
+      "grad_norm": 9.97355842590332,
+      "learning_rate": 4.490523690773067e-06,
+      "loss": 0.3102,
+      "step": 62220
+    },
+    {
+      "epoch": 15.518703241895262,
+      "grad_norm": 7.273804664611816,
+      "learning_rate": 4.488029925187033e-06,
+      "loss": 0.273,
+      "step": 62230
+    },
+    {
+      "epoch": 15.521197007481296,
+      "grad_norm": 4.933453559875488,
+      "learning_rate": 4.485536159600998e-06,
+      "loss": 0.2666,
+      "step": 62240
+    },
+    {
+      "epoch": 15.523690773067331,
+      "grad_norm": 6.50958251953125,
+      "learning_rate": 4.483042394014963e-06,
+      "loss": 0.2831,
+      "step": 62250
+    },
+    {
+      "epoch": 15.526184538653366,
+      "grad_norm": 5.90592622756958,
+      "learning_rate": 4.480548628428928e-06,
+      "loss": 0.4132,
+      "step": 62260
+    },
+    {
+      "epoch": 15.528678304239401,
+      "grad_norm": 7.461799144744873,
+      "learning_rate": 4.478054862842893e-06,
+      "loss": 0.3894,
+      "step": 62270
+    },
+    {
+      "epoch": 15.531172069825436,
+      "grad_norm": 6.117193222045898,
+      "learning_rate": 4.475561097256858e-06,
+      "loss": 0.318,
+      "step": 62280
+    },
+    {
+      "epoch": 15.53366583541147,
+      "grad_norm": 7.674281120300293,
+      "learning_rate": 4.473067331670824e-06,
+      "loss": 0.3456,
+      "step": 62290
+    },
+    {
+      "epoch": 15.536159600997506,
+      "grad_norm": 7.840825080871582,
+      "learning_rate": 4.470573566084788e-06,
+      "loss": 0.3274,
+      "step": 62300
+    },
+    {
+      "epoch": 15.53865336658354,
+      "grad_norm": 9.947293281555176,
+      "learning_rate": 4.468079800498753e-06,
+      "loss": 0.3462,
+      "step": 62310
+    },
+    {
+      "epoch": 15.541147132169575,
+      "grad_norm": 6.473944664001465,
+      "learning_rate": 4.465586034912719e-06,
+      "loss": 0.2612,
+      "step": 62320
+    },
+    {
+      "epoch": 15.54364089775561,
+      "grad_norm": 10.87900447845459,
+      "learning_rate": 4.463092269326684e-06,
+      "loss": 0.3052,
+      "step": 62330
+    },
+    {
+      "epoch": 15.546134663341645,
+      "grad_norm": 7.721189022064209,
+      "learning_rate": 4.4605985037406484e-06,
+      "loss": 0.3771,
+      "step": 62340
+    },
+    {
+      "epoch": 15.548628428927682,
+      "grad_norm": 7.034101963043213,
+      "learning_rate": 4.458104738154614e-06,
+      "loss": 0.3067,
+      "step": 62350
+    },
+    {
+      "epoch": 15.551122194513717,
+      "grad_norm": 8.35641860961914,
+      "learning_rate": 4.455610972568579e-06,
+      "loss": 0.3294,
+      "step": 62360
+    },
+    {
+      "epoch": 15.553615960099751,
+      "grad_norm": 7.652080535888672,
+      "learning_rate": 4.453117206982544e-06,
+      "loss": 0.3738,
+      "step": 62370
+    },
+    {
+      "epoch": 15.556109725685786,
+      "grad_norm": 7.658141136169434,
+      "learning_rate": 4.450623441396509e-06,
+      "loss": 0.3161,
+      "step": 62380
+    },
+    {
+      "epoch": 15.558603491271821,
+      "grad_norm": 7.808701992034912,
+      "learning_rate": 4.448129675810474e-06,
+      "loss": 0.3701,
+      "step": 62390
+    },
+    {
+      "epoch": 15.561097256857856,
+      "grad_norm": 9.722882270812988,
+      "learning_rate": 4.4456359102244395e-06,
+      "loss": 0.3375,
+      "step": 62400
+    },
+    {
+      "epoch": 15.563591022443891,
+      "grad_norm": 8.67414379119873,
+      "learning_rate": 4.443142144638404e-06,
+      "loss": 0.3462,
+      "step": 62410
+    },
+    {
+      "epoch": 15.566084788029926,
+      "grad_norm": 5.231639385223389,
+      "learning_rate": 4.44064837905237e-06,
+      "loss": 0.3686,
+      "step": 62420
+    },
+    {
+      "epoch": 15.56857855361596,
+      "grad_norm": 11.116351127624512,
+      "learning_rate": 4.4381546134663345e-06,
+      "loss": 0.3446,
+      "step": 62430
+    },
+    {
+      "epoch": 15.571072319201996,
+      "grad_norm": 6.519227504730225,
+      "learning_rate": 4.4356608478803e-06,
+      "loss": 0.2998,
+      "step": 62440
+    },
+    {
+      "epoch": 15.57356608478803,
+      "grad_norm": 12.292588233947754,
+      "learning_rate": 4.433167082294264e-06,
+      "loss": 0.3128,
+      "step": 62450
+    },
+    {
+      "epoch": 15.576059850374065,
+      "grad_norm": 10.114727973937988,
+      "learning_rate": 4.43067331670823e-06,
+      "loss": 0.4101,
+      "step": 62460
+    },
+    {
+      "epoch": 15.5785536159601,
+      "grad_norm": 6.243036270141602,
+      "learning_rate": 4.428179551122195e-06,
+      "loss": 0.3902,
+      "step": 62470
+    },
+    {
+      "epoch": 15.581047381546135,
+      "grad_norm": 7.189294338226318,
+      "learning_rate": 4.42568578553616e-06,
+      "loss": 0.2728,
+      "step": 62480
+    },
+    {
+      "epoch": 15.58354114713217,
+      "grad_norm": 6.678802967071533,
+      "learning_rate": 4.423192019950125e-06,
+      "loss": 0.3568,
+      "step": 62490
+    },
+    {
+      "epoch": 15.586034912718205,
+      "grad_norm": 10.736851692199707,
+      "learning_rate": 4.42069825436409e-06,
+      "loss": 0.3335,
+      "step": 62500
+    },
+    {
+      "epoch": 15.58852867830424,
+      "grad_norm": 7.438121318817139,
+      "learning_rate": 4.418204488778055e-06,
+      "loss": 0.2941,
+      "step": 62510
+    },
+    {
+      "epoch": 15.591022443890274,
+      "grad_norm": 6.257914066314697,
+      "learning_rate": 4.415710723192021e-06,
+      "loss": 0.2712,
+      "step": 62520
+    },
+    {
+      "epoch": 15.59351620947631,
+      "grad_norm": 9.817953109741211,
+      "learning_rate": 4.413216957605985e-06,
+      "loss": 0.327,
+      "step": 62530
+    },
+    {
+      "epoch": 15.596009975062344,
+      "grad_norm": 6.332709312438965,
+      "learning_rate": 4.41072319201995e-06,
+      "loss": 0.3838,
+      "step": 62540
+    },
+    {
+      "epoch": 15.598503740648379,
+      "grad_norm": 7.172346115112305,
+      "learning_rate": 4.408229426433916e-06,
+      "loss": 0.3826,
+      "step": 62550
+    },
+    {
+      "epoch": 15.600997506234414,
+      "grad_norm": 8.06655502319336,
+      "learning_rate": 4.405735660847881e-06,
+      "loss": 0.2852,
+      "step": 62560
+    },
+    {
+      "epoch": 15.603491271820449,
+      "grad_norm": 12.748686790466309,
+      "learning_rate": 4.4032418952618455e-06,
+      "loss": 0.3728,
+      "step": 62570
+    },
+    {
+      "epoch": 15.605985037406484,
+      "grad_norm": 8.942939758300781,
+      "learning_rate": 4.400748129675811e-06,
+      "loss": 0.3794,
+      "step": 62580
+    },
+    {
+      "epoch": 15.608478802992519,
+      "grad_norm": 7.845804691314697,
+      "learning_rate": 4.398254364089775e-06,
+      "loss": 0.3231,
+      "step": 62590
+    },
+    {
+      "epoch": 15.610972568578553,
+      "grad_norm": 10.373620986938477,
+      "learning_rate": 4.395760598503741e-06,
+      "loss": 0.2595,
+      "step": 62600
+    },
+    {
+      "epoch": 15.613466334164588,
+      "grad_norm": 7.292490005493164,
+      "learning_rate": 4.393266832917706e-06,
+      "loss": 0.316,
+      "step": 62610
+    },
+    {
+      "epoch": 15.615960099750623,
+      "grad_norm": 8.094284057617188,
+      "learning_rate": 4.390773067331671e-06,
+      "loss": 0.3601,
+      "step": 62620
+    },
+    {
+      "epoch": 15.618453865336658,
+      "grad_norm": 8.893630981445312,
+      "learning_rate": 4.3882793017456365e-06,
+      "loss": 0.3487,
+      "step": 62630
+    },
+    {
+      "epoch": 15.620947630922693,
+      "grad_norm": 6.261669635772705,
+      "learning_rate": 4.385785536159601e-06,
+      "loss": 0.2983,
+      "step": 62640
+    },
+    {
+      "epoch": 15.623441396508728,
+      "grad_norm": 6.368238925933838,
+      "learning_rate": 4.383291770573566e-06,
+      "loss": 0.3524,
+      "step": 62650
+    },
+    {
+      "epoch": 15.625935162094763,
+      "grad_norm": 15.371926307678223,
+      "learning_rate": 4.380798004987532e-06,
+      "loss": 0.378,
+      "step": 62660
+    },
+    {
+      "epoch": 15.628428927680797,
+      "grad_norm": 5.758829593658447,
+      "learning_rate": 4.378304239401497e-06,
+      "loss": 0.3458,
+      "step": 62670
+    },
+    {
+      "epoch": 15.630922693266832,
+      "grad_norm": 7.6789326667785645,
+      "learning_rate": 4.375810473815461e-06,
+      "loss": 0.2688,
+      "step": 62680
+    },
+    {
+      "epoch": 15.633416458852867,
+      "grad_norm": 7.106253147125244,
+      "learning_rate": 4.373316708229427e-06,
+      "loss": 0.2851,
+      "step": 62690
+    },
+    {
+      "epoch": 15.635910224438902,
+      "grad_norm": 6.524157524108887,
+      "learning_rate": 4.370822942643392e-06,
+      "loss": 0.3196,
+      "step": 62700
+    },
+    {
+      "epoch": 15.638403990024937,
+      "grad_norm": 12.354263305664062,
+      "learning_rate": 4.368329177057357e-06,
+      "loss": 0.3764,
+      "step": 62710
+    },
+    {
+      "epoch": 15.640897755610972,
+      "grad_norm": 8.018882751464844,
+      "learning_rate": 4.365835411471322e-06,
+      "loss": 0.3431,
+      "step": 62720
+    },
+    {
+      "epoch": 15.643391521197007,
+      "grad_norm": 4.7391228675842285,
+      "learning_rate": 4.363341645885287e-06,
+      "loss": 0.2748,
+      "step": 62730
+    },
+    {
+      "epoch": 15.645885286783042,
+      "grad_norm": 7.503388404846191,
+      "learning_rate": 4.360847880299252e-06,
+      "loss": 0.3772,
+      "step": 62740
+    },
+    {
+      "epoch": 15.648379052369076,
+      "grad_norm": 8.94284725189209,
+      "learning_rate": 4.358354114713218e-06,
+      "loss": 0.3285,
+      "step": 62750
+    },
+    {
+      "epoch": 15.650872817955111,
+      "grad_norm": 7.088261127471924,
+      "learning_rate": 4.355860349127182e-06,
+      "loss": 0.2969,
+      "step": 62760
+    },
+    {
+      "epoch": 15.653366583541148,
+      "grad_norm": 5.1243414878845215,
+      "learning_rate": 4.3533665835411475e-06,
+      "loss": 0.3318,
+      "step": 62770
+    },
+    {
+      "epoch": 15.655860349127183,
+      "grad_norm": 9.318085670471191,
+      "learning_rate": 4.350872817955113e-06,
+      "loss": 0.3598,
+      "step": 62780
+    },
+    {
+      "epoch": 15.658354114713218,
+      "grad_norm": 7.903304576873779,
+      "learning_rate": 4.348379052369078e-06,
+      "loss": 0.3319,
+      "step": 62790
+    },
+    {
+      "epoch": 15.660847880299253,
+      "grad_norm": 7.943011283874512,
+      "learning_rate": 4.3458852867830425e-06,
+      "loss": 0.3095,
+      "step": 62800
+    },
+    {
+      "epoch": 15.663341645885287,
+      "grad_norm": 8.990947723388672,
+      "learning_rate": 4.343391521197008e-06,
+      "loss": 0.2832,
+      "step": 62810
+    },
+    {
+      "epoch": 15.665835411471322,
+      "grad_norm": 11.269744873046875,
+      "learning_rate": 4.340897755610972e-06,
+      "loss": 0.3504,
+      "step": 62820
+    },
+    {
+      "epoch": 15.668329177057357,
+      "grad_norm": 8.170928955078125,
+      "learning_rate": 4.3384039900249385e-06,
+      "loss": 0.2957,
+      "step": 62830
+    },
+    {
+      "epoch": 15.670822942643392,
+      "grad_norm": 10.878934860229492,
+      "learning_rate": 4.335910224438903e-06,
+      "loss": 0.3487,
+      "step": 62840
+    },
+    {
+      "epoch": 15.673316708229427,
+      "grad_norm": 7.941391944885254,
+      "learning_rate": 4.333416458852868e-06,
+      "loss": 0.3278,
+      "step": 62850
+    },
+    {
+      "epoch": 15.675810473815462,
+      "grad_norm": 8.739579200744629,
+      "learning_rate": 4.330922693266833e-06,
+      "loss": 0.3454,
+      "step": 62860
+    },
+    {
+      "epoch": 15.678304239401497,
+      "grad_norm": 7.752486705780029,
+      "learning_rate": 4.328428927680798e-06,
+      "loss": 0.3343,
+      "step": 62870
+    },
+    {
+      "epoch": 15.680798004987532,
+      "grad_norm": 10.62015151977539,
+      "learning_rate": 4.325935162094763e-06,
+      "loss": 0.3248,
+      "step": 62880
+    },
+    {
+      "epoch": 15.683291770573566,
+      "grad_norm": 7.1810832023620605,
+      "learning_rate": 4.323441396508729e-06,
+      "loss": 0.2936,
+      "step": 62890
+    },
+    {
+      "epoch": 15.685785536159601,
+      "grad_norm": 7.9837236404418945,
+      "learning_rate": 4.320947630922694e-06,
+      "loss": 0.2879,
+      "step": 62900
+    },
+    {
+      "epoch": 15.688279301745636,
+      "grad_norm": 9.112959861755371,
+      "learning_rate": 4.318453865336658e-06,
+      "loss": 0.2799,
+      "step": 62910
+    },
+    {
+      "epoch": 15.690773067331671,
+      "grad_norm": 8.914131164550781,
+      "learning_rate": 4.315960099750624e-06,
+      "loss": 0.3874,
+      "step": 62920
+    },
+    {
+      "epoch": 15.693266832917706,
+      "grad_norm": 10.754671096801758,
+      "learning_rate": 4.313466334164589e-06,
+      "loss": 0.3674,
+      "step": 62930
+    },
+    {
+      "epoch": 15.69576059850374,
+      "grad_norm": 8.270440101623535,
+      "learning_rate": 4.310972568578554e-06,
+      "loss": 0.3293,
+      "step": 62940
+    },
+    {
+      "epoch": 15.698254364089776,
+      "grad_norm": 9.554506301879883,
+      "learning_rate": 4.308478802992519e-06,
+      "loss": 0.3053,
+      "step": 62950
+    },
+    {
+      "epoch": 15.70074812967581,
+      "grad_norm": 8.808588981628418,
+      "learning_rate": 4.305985037406484e-06,
+      "loss": 0.3664,
+      "step": 62960
+    },
+    {
+      "epoch": 15.703241895261845,
+      "grad_norm": 16.313865661621094,
+      "learning_rate": 4.303491271820449e-06,
+      "loss": 0.3232,
+      "step": 62970
+    },
+    {
+      "epoch": 15.70573566084788,
+      "grad_norm": 18.93476104736328,
+      "learning_rate": 4.300997506234415e-06,
+      "loss": 0.3315,
+      "step": 62980
+    },
+    {
+      "epoch": 15.708229426433915,
+      "grad_norm": 7.082018852233887,
+      "learning_rate": 4.298503740648379e-06,
+      "loss": 0.3039,
+      "step": 62990
+    },
+    {
+      "epoch": 15.71072319201995,
+      "grad_norm": 5.019651889801025,
+      "learning_rate": 4.2960099750623445e-06,
+      "loss": 0.3773,
+      "step": 63000
+    },
+    {
+      "epoch": 15.713216957605985,
+      "grad_norm": 7.572514057159424,
+      "learning_rate": 4.29351620947631e-06,
+      "loss": 0.3187,
+      "step": 63010
+    },
+    {
+      "epoch": 15.71571072319202,
+      "grad_norm": 7.050946235656738,
+      "learning_rate": 4.291022443890275e-06,
+      "loss": 0.3341,
+      "step": 63020
+    },
+    {
+      "epoch": 15.718204488778055,
+      "grad_norm": 11.754009246826172,
+      "learning_rate": 4.2885286783042396e-06,
+      "loss": 0.3174,
+      "step": 63030
+    },
+    {
+      "epoch": 15.72069825436409,
+      "grad_norm": 6.041005611419678,
+      "learning_rate": 4.286034912718205e-06,
+      "loss": 0.2952,
+      "step": 63040
+    },
+    {
+      "epoch": 15.723192019950124,
+      "grad_norm": 9.302490234375,
+      "learning_rate": 4.283541147132169e-06,
+      "loss": 0.3406,
+      "step": 63050
+    },
+    {
+      "epoch": 15.72568578553616,
+      "grad_norm": 7.180544853210449,
+      "learning_rate": 4.2810473815461355e-06,
+      "loss": 0.3036,
+      "step": 63060
+    },
+    {
+      "epoch": 15.728179551122194,
+      "grad_norm": 10.246197700500488,
+      "learning_rate": 4.2785536159601e-06,
+      "loss": 0.3262,
+      "step": 63070
+    },
+    {
+      "epoch": 15.730673316708229,
+      "grad_norm": 10.846905708312988,
+      "learning_rate": 4.276059850374065e-06,
+      "loss": 0.3819,
+      "step": 63080
+    },
+    {
+      "epoch": 15.733167082294264,
+      "grad_norm": 7.604201316833496,
+      "learning_rate": 4.27356608478803e-06,
+      "loss": 0.3568,
+      "step": 63090
+    },
+    {
+      "epoch": 15.735660847880299,
+      "grad_norm": 9.36225414276123,
+      "learning_rate": 4.271072319201995e-06,
+      "loss": 0.3201,
+      "step": 63100
+    },
+    {
+      "epoch": 15.738154613466333,
+      "grad_norm": 15.31506061553955,
+      "learning_rate": 4.26857855361596e-06,
+      "loss": 0.3591,
+      "step": 63110
+    },
+    {
+      "epoch": 15.740648379052368,
+      "grad_norm": 10.249829292297363,
+      "learning_rate": 4.266084788029926e-06,
+      "loss": 0.3486,
+      "step": 63120
+    },
+    {
+      "epoch": 15.743142144638403,
+      "grad_norm": 8.24986457824707,
+      "learning_rate": 4.263591022443891e-06,
+      "loss": 0.3455,
+      "step": 63130
+    },
+    {
+      "epoch": 15.745635910224438,
+      "grad_norm": 6.340268135070801,
+      "learning_rate": 4.2610972568578554e-06,
+      "loss": 0.3159,
+      "step": 63140
+    },
+    {
+      "epoch": 15.748129675810475,
+      "grad_norm": 8.134374618530273,
+      "learning_rate": 4.258603491271821e-06,
+      "loss": 0.3524,
+      "step": 63150
+    },
+    {
+      "epoch": 15.75062344139651,
+      "grad_norm": 12.707883834838867,
+      "learning_rate": 4.256109725685786e-06,
+      "loss": 0.3194,
+      "step": 63160
+    },
+    {
+      "epoch": 15.753117206982544,
+      "grad_norm": 10.832149505615234,
+      "learning_rate": 4.253615960099751e-06,
+      "loss": 0.3122,
+      "step": 63170
+    },
+    {
+      "epoch": 15.75561097256858,
+      "grad_norm": 10.519184112548828,
+      "learning_rate": 4.251122194513716e-06,
+      "loss": 0.3642,
+      "step": 63180
+    },
+    {
+      "epoch": 15.758104738154614,
+      "grad_norm": 14.12321949005127,
+      "learning_rate": 4.248628428927681e-06,
+      "loss": 0.3388,
+      "step": 63190
+    },
+    {
+      "epoch": 15.760598503740649,
+      "grad_norm": 6.439551830291748,
+      "learning_rate": 4.2461346633416465e-06,
+      "loss": 0.385,
+      "step": 63200
+    },
+    {
+      "epoch": 15.763092269326684,
+      "grad_norm": 25.51148223876953,
+      "learning_rate": 4.243640897755612e-06,
+      "loss": 0.4235,
+      "step": 63210
+    },
+    {
+      "epoch": 15.765586034912719,
+      "grad_norm": 9.474204063415527,
+      "learning_rate": 4.241147132169576e-06,
+      "loss": 0.3963,
+      "step": 63220
+    },
+    {
+      "epoch": 15.768079800498754,
+      "grad_norm": 10.349353790283203,
+      "learning_rate": 4.2386533665835415e-06,
+      "loss": 0.293,
+      "step": 63230
+    },
+    {
+      "epoch": 15.770573566084789,
+      "grad_norm": 9.036044120788574,
+      "learning_rate": 4.236159600997507e-06,
+      "loss": 0.3493,
+      "step": 63240
+    },
+    {
+      "epoch": 15.773067331670823,
+      "grad_norm": 6.4520392417907715,
+      "learning_rate": 4.233665835411472e-06,
+      "loss": 0.3464,
+      "step": 63250
+    },
+    {
+      "epoch": 15.775561097256858,
+      "grad_norm": 6.277656078338623,
+      "learning_rate": 4.231172069825437e-06,
+      "loss": 0.4086,
+      "step": 63260
+    },
+    {
+      "epoch": 15.778054862842893,
+      "grad_norm": 7.3123979568481445,
+      "learning_rate": 4.228678304239402e-06,
+      "loss": 0.3339,
+      "step": 63270
+    },
+    {
+      "epoch": 15.780548628428928,
+      "grad_norm": 4.954773426055908,
+      "learning_rate": 4.226184538653366e-06,
+      "loss": 0.3325,
+      "step": 63280
+    },
+    {
+      "epoch": 15.783042394014963,
+      "grad_norm": 8.477773666381836,
+      "learning_rate": 4.223690773067332e-06,
+      "loss": 0.3338,
+      "step": 63290
+    },
+    {
+      "epoch": 15.785536159600998,
+      "grad_norm": 6.290766716003418,
+      "learning_rate": 4.221197007481297e-06,
+      "loss": 0.3108,
+      "step": 63300
+    },
+    {
+      "epoch": 15.788029925187033,
+      "grad_norm": 6.442891597747803,
+      "learning_rate": 4.218703241895262e-06,
+      "loss": 0.304,
+      "step": 63310
+    },
+    {
+      "epoch": 15.790523690773068,
+      "grad_norm": 7.096189022064209,
+      "learning_rate": 4.216209476309227e-06,
+      "loss": 0.2894,
+      "step": 63320
+    },
+    {
+      "epoch": 15.793017456359102,
+      "grad_norm": 7.581128120422363,
+      "learning_rate": 4.213715710723192e-06,
+      "loss": 0.3329,
+      "step": 63330
+    },
+    {
+      "epoch": 15.795511221945137,
+      "grad_norm": 7.5609130859375,
+      "learning_rate": 4.211221945137157e-06,
+      "loss": 0.233,
+      "step": 63340
+    },
+    {
+      "epoch": 15.798004987531172,
+      "grad_norm": 6.8791584968566895,
+      "learning_rate": 4.208728179551123e-06,
+      "loss": 0.3167,
+      "step": 63350
+    },
+    {
+      "epoch": 15.800498753117207,
+      "grad_norm": 8.420093536376953,
+      "learning_rate": 4.206234413965087e-06,
+      "loss": 0.3346,
+      "step": 63360
+    },
+    {
+      "epoch": 15.802992518703242,
+      "grad_norm": 8.641119003295898,
+      "learning_rate": 4.2037406483790525e-06,
+      "loss": 0.3995,
+      "step": 63370
+    },
+    {
+      "epoch": 15.805486284289277,
+      "grad_norm": 13.097655296325684,
+      "learning_rate": 4.201246882793018e-06,
+      "loss": 0.3696,
+      "step": 63380
+    },
+    {
+      "epoch": 15.807980049875312,
+      "grad_norm": 7.517343521118164,
+      "learning_rate": 4.198753117206983e-06,
+      "loss": 0.2953,
+      "step": 63390
+    },
+    {
+      "epoch": 15.810473815461346,
+      "grad_norm": 9.811420440673828,
+      "learning_rate": 4.196259351620948e-06,
+      "loss": 0.3546,
+      "step": 63400
+    },
+    {
+      "epoch": 15.812967581047381,
+      "grad_norm": 8.417027473449707,
+      "learning_rate": 4.193765586034913e-06,
+      "loss": 0.3077,
+      "step": 63410
+    },
+    {
+      "epoch": 15.815461346633416,
+      "grad_norm": 7.933226108551025,
+      "learning_rate": 4.191271820448878e-06,
+      "loss": 0.3254,
+      "step": 63420
+    },
+    {
+      "epoch": 15.817955112219451,
+      "grad_norm": 6.797752380371094,
+      "learning_rate": 4.1887780548628435e-06,
+      "loss": 0.2962,
+      "step": 63430
+    },
+    {
+      "epoch": 15.820448877805486,
+      "grad_norm": 6.070878505706787,
+      "learning_rate": 4.186284289276809e-06,
+      "loss": 0.2808,
+      "step": 63440
+    },
+    {
+      "epoch": 15.82294264339152,
+      "grad_norm": 10.812505722045898,
+      "learning_rate": 4.183790523690773e-06,
+      "loss": 0.334,
+      "step": 63450
+    },
+    {
+      "epoch": 15.825436408977556,
+      "grad_norm": 9.91736125946045,
+      "learning_rate": 4.1812967581047386e-06,
+      "loss": 0.2903,
+      "step": 63460
+    },
+    {
+      "epoch": 15.82793017456359,
+      "grad_norm": 7.868504524230957,
+      "learning_rate": 4.178802992518704e-06,
+      "loss": 0.3427,
+      "step": 63470
+    },
+    {
+      "epoch": 15.830423940149625,
+      "grad_norm": 10.431106567382812,
+      "learning_rate": 4.176309226932669e-06,
+      "loss": 0.3997,
+      "step": 63480
+    },
+    {
+      "epoch": 15.83291770573566,
+      "grad_norm": 9.199599266052246,
+      "learning_rate": 4.173815461346634e-06,
+      "loss": 0.3854,
+      "step": 63490
+    },
+    {
+      "epoch": 15.835411471321695,
+      "grad_norm": 9.323473930358887,
+      "learning_rate": 4.171321695760599e-06,
+      "loss": 0.3475,
+      "step": 63500
+    },
+    {
+      "epoch": 15.83790523690773,
+      "grad_norm": 6.273562431335449,
+      "learning_rate": 4.1688279301745634e-06,
+      "loss": 0.3192,
+      "step": 63510
+    },
+    {
+      "epoch": 15.840399002493765,
+      "grad_norm": 9.469862937927246,
+      "learning_rate": 4.166334164588529e-06,
+      "loss": 0.3289,
+      "step": 63520
+    },
+    {
+      "epoch": 15.8428927680798,
+      "grad_norm": 10.834332466125488,
+      "learning_rate": 4.163840399002494e-06,
+      "loss": 0.3935,
+      "step": 63530
+    },
+    {
+      "epoch": 15.845386533665835,
+      "grad_norm": 11.64710807800293,
+      "learning_rate": 4.161346633416459e-06,
+      "loss": 0.3366,
+      "step": 63540
+    },
+    {
+      "epoch": 15.84788029925187,
+      "grad_norm": 8.563414573669434,
+      "learning_rate": 4.158852867830424e-06,
+      "loss": 0.3576,
+      "step": 63550
+    },
+    {
+      "epoch": 15.850374064837904,
+      "grad_norm": 8.679342269897461,
+      "learning_rate": 4.156359102244389e-06,
+      "loss": 0.2798,
+      "step": 63560
+    },
+    {
+      "epoch": 15.85286783042394,
+      "grad_norm": 4.7781548500061035,
+      "learning_rate": 4.1538653366583544e-06,
+      "loss": 0.3337,
+      "step": 63570
+    },
+    {
+      "epoch": 15.855361596009976,
+      "grad_norm": 4.6593523025512695,
+      "learning_rate": 4.15137157107232e-06,
+      "loss": 0.3748,
+      "step": 63580
+    },
+    {
+      "epoch": 15.85785536159601,
+      "grad_norm": 7.632694244384766,
+      "learning_rate": 4.148877805486284e-06,
+      "loss": 0.3599,
+      "step": 63590
+    },
+    {
+      "epoch": 15.860349127182046,
+      "grad_norm": 6.817638397216797,
+      "learning_rate": 4.1463840399002495e-06,
+      "loss": 0.2949,
+      "step": 63600
+    },
+    {
+      "epoch": 15.86284289276808,
+      "grad_norm": 10.572402954101562,
+      "learning_rate": 4.143890274314215e-06,
+      "loss": 0.3316,
+      "step": 63610
+    },
+    {
+      "epoch": 15.865336658354115,
+      "grad_norm": 7.423572540283203,
+      "learning_rate": 4.14139650872818e-06,
+      "loss": 0.2964,
+      "step": 63620
+    },
+    {
+      "epoch": 15.86783042394015,
+      "grad_norm": 8.025785446166992,
+      "learning_rate": 4.1389027431421455e-06,
+      "loss": 0.2967,
+      "step": 63630
+    },
+    {
+      "epoch": 15.870324189526185,
+      "grad_norm": 5.359241485595703,
+      "learning_rate": 4.13640897755611e-06,
+      "loss": 0.3303,
+      "step": 63640
+    },
+    {
+      "epoch": 15.87281795511222,
+      "grad_norm": 7.528040885925293,
+      "learning_rate": 4.133915211970075e-06,
+      "loss": 0.3465,
+      "step": 63650
+    },
+    {
+      "epoch": 15.875311720698255,
+      "grad_norm": 8.721107482910156,
+      "learning_rate": 4.1314214463840405e-06,
+      "loss": 0.3032,
+      "step": 63660
+    },
+    {
+      "epoch": 15.87780548628429,
+      "grad_norm": 8.390384674072266,
+      "learning_rate": 4.128927680798006e-06,
+      "loss": 0.2707,
+      "step": 63670
+    },
+    {
+      "epoch": 15.880299251870325,
+      "grad_norm": 6.81118631362915,
+      "learning_rate": 4.12643391521197e-06,
+      "loss": 0.3659,
+      "step": 63680
+    },
+    {
+      "epoch": 15.88279301745636,
+      "grad_norm": 11.449249267578125,
+      "learning_rate": 4.123940149625936e-06,
+      "loss": 0.3118,
+      "step": 63690
+    },
+    {
+      "epoch": 15.885286783042394,
+      "grad_norm": 8.672457695007324,
+      "learning_rate": 4.1214463840399e-06,
+      "loss": 0.3703,
+      "step": 63700
+    },
+    {
+      "epoch": 15.88778054862843,
+      "grad_norm": 6.935692310333252,
+      "learning_rate": 4.118952618453866e-06,
+      "loss": 0.2715,
+      "step": 63710
+    },
+    {
+      "epoch": 15.890274314214464,
+      "grad_norm": 7.293999195098877,
+      "learning_rate": 4.116458852867831e-06,
+      "loss": 0.3367,
+      "step": 63720
+    },
+    {
+      "epoch": 15.892768079800499,
+      "grad_norm": 8.051675796508789,
+      "learning_rate": 4.113965087281796e-06,
+      "loss": 0.3378,
+      "step": 63730
+    },
+    {
+      "epoch": 15.895261845386534,
+      "grad_norm": 8.13101577758789,
+      "learning_rate": 4.1114713216957605e-06,
+      "loss": 0.3245,
+      "step": 63740
+    },
+    {
+      "epoch": 15.897755610972569,
+      "grad_norm": 8.935870170593262,
+      "learning_rate": 4.108977556109726e-06,
+      "loss": 0.3168,
+      "step": 63750
+    },
+    {
+      "epoch": 15.900249376558603,
+      "grad_norm": 8.20327091217041,
+      "learning_rate": 4.106483790523691e-06,
+      "loss": 0.3086,
+      "step": 63760
+    },
+    {
+      "epoch": 15.902743142144638,
+      "grad_norm": 6.9828667640686035,
+      "learning_rate": 4.103990024937656e-06,
+      "loss": 0.3724,
+      "step": 63770
+    },
+    {
+      "epoch": 15.905236907730673,
+      "grad_norm": 7.1580915451049805,
+      "learning_rate": 4.101496259351621e-06,
+      "loss": 0.3337,
+      "step": 63780
+    },
+    {
+      "epoch": 15.907730673316708,
+      "grad_norm": 7.152059078216553,
+      "learning_rate": 4.099002493765586e-06,
+      "loss": 0.3067,
+      "step": 63790
+    },
+    {
+      "epoch": 15.910224438902743,
+      "grad_norm": 8.175209999084473,
+      "learning_rate": 4.0965087281795515e-06,
+      "loss": 0.3518,
+      "step": 63800
+    },
+    {
+      "epoch": 15.912718204488778,
+      "grad_norm": 8.411026000976562,
+      "learning_rate": 4.094014962593517e-06,
+      "loss": 0.2738,
+      "step": 63810
+    },
+    {
+      "epoch": 15.915211970074813,
+      "grad_norm": 4.850775241851807,
+      "learning_rate": 4.091521197007481e-06,
+      "loss": 0.2802,
+      "step": 63820
+    },
+    {
+      "epoch": 15.917705735660848,
+      "grad_norm": 11.431254386901855,
+      "learning_rate": 4.0890274314214466e-06,
+      "loss": 0.3478,
+      "step": 63830
+    },
+    {
+      "epoch": 15.920199501246882,
+      "grad_norm": 7.01432991027832,
+      "learning_rate": 4.086533665835412e-06,
+      "loss": 0.3393,
+      "step": 63840
+    },
+    {
+      "epoch": 15.922693266832917,
+      "grad_norm": 6.793428421020508,
+      "learning_rate": 4.084039900249377e-06,
+      "loss": 0.3174,
+      "step": 63850
+    },
+    {
+      "epoch": 15.925187032418952,
+      "grad_norm": 7.512074947357178,
+      "learning_rate": 4.081546134663342e-06,
+      "loss": 0.2974,
+      "step": 63860
+    },
+    {
+      "epoch": 15.927680798004987,
+      "grad_norm": 8.08775806427002,
+      "learning_rate": 4.079052369077307e-06,
+      "loss": 0.3465,
+      "step": 63870
+    },
+    {
+      "epoch": 15.930174563591022,
+      "grad_norm": 8.87288761138916,
+      "learning_rate": 4.076558603491272e-06,
+      "loss": 0.329,
+      "step": 63880
+    },
+    {
+      "epoch": 15.932668329177057,
+      "grad_norm": 8.738507270812988,
+      "learning_rate": 4.0740648379052376e-06,
+      "loss": 0.3361,
+      "step": 63890
+    },
+    {
+      "epoch": 15.935162094763092,
+      "grad_norm": 6.354411602020264,
+      "learning_rate": 4.071571072319203e-06,
+      "loss": 0.3113,
+      "step": 63900
+    },
+    {
+      "epoch": 15.937655860349127,
+      "grad_norm": 10.517175674438477,
+      "learning_rate": 4.069077306733167e-06,
+      "loss": 0.3381,
+      "step": 63910
+    },
+    {
+      "epoch": 15.940149625935161,
+      "grad_norm": 8.709144592285156,
+      "learning_rate": 4.066583541147133e-06,
+      "loss": 0.3011,
+      "step": 63920
+    },
+    {
+      "epoch": 15.942643391521196,
+      "grad_norm": 7.812455654144287,
+      "learning_rate": 4.064089775561097e-06,
+      "loss": 0.3099,
+      "step": 63930
+    },
+    {
+      "epoch": 15.945137157107231,
+      "grad_norm": 7.7127685546875,
+      "learning_rate": 4.061596009975063e-06,
+      "loss": 0.3536,
+      "step": 63940
+    },
+    {
+      "epoch": 15.947630922693268,
+      "grad_norm": 8.810644149780273,
+      "learning_rate": 4.059102244389028e-06,
+      "loss": 0.3048,
+      "step": 63950
+    },
+    {
+      "epoch": 15.950124688279303,
+      "grad_norm": 9.955845832824707,
+      "learning_rate": 4.056608478802993e-06,
+      "loss": 0.3582,
+      "step": 63960
+    },
+    {
+      "epoch": 15.952618453865338,
+      "grad_norm": 8.85975456237793,
+      "learning_rate": 4.0541147132169575e-06,
+      "loss": 0.3893,
+      "step": 63970
+    },
+    {
+      "epoch": 15.955112219451372,
+      "grad_norm": 8.970422744750977,
+      "learning_rate": 4.051620947630923e-06,
+      "loss": 0.2967,
+      "step": 63980
+    },
+    {
+      "epoch": 15.957605985037407,
+      "grad_norm": 8.689909934997559,
+      "learning_rate": 4.049127182044888e-06,
+      "loss": 0.3129,
+      "step": 63990
+    },
+    {
+      "epoch": 15.960099750623442,
+      "grad_norm": 12.43597412109375,
+      "learning_rate": 4.0466334164588534e-06,
+      "loss": 0.4056,
+      "step": 64000
+    },
+    {
+      "epoch": 15.962593516209477,
+      "grad_norm": 5.898103713989258,
+      "learning_rate": 4.044139650872818e-06,
+      "loss": 0.3884,
+      "step": 64010
+    },
+    {
+      "epoch": 15.965087281795512,
+      "grad_norm": 9.561400413513184,
+      "learning_rate": 4.041645885286783e-06,
+      "loss": 0.2965,
+      "step": 64020
+    },
+    {
+      "epoch": 15.967581047381547,
+      "grad_norm": 13.149139404296875,
+      "learning_rate": 4.0391521197007485e-06,
+      "loss": 0.2631,
+      "step": 64030
+    },
+    {
+      "epoch": 15.970074812967582,
+      "grad_norm": 8.486005783081055,
+      "learning_rate": 4.036658354114714e-06,
+      "loss": 0.2731,
+      "step": 64040
+    },
+    {
+      "epoch": 15.972568578553616,
+      "grad_norm": 10.793266296386719,
+      "learning_rate": 4.034164588528678e-06,
+      "loss": 0.3073,
+      "step": 64050
+    },
+    {
+      "epoch": 15.975062344139651,
+      "grad_norm": 19.346826553344727,
+      "learning_rate": 4.031670822942644e-06,
+      "loss": 0.3753,
+      "step": 64060
+    },
+    {
+      "epoch": 15.977556109725686,
+      "grad_norm": 9.948132514953613,
+      "learning_rate": 4.029177057356609e-06,
+      "loss": 0.314,
+      "step": 64070
+    },
+    {
+      "epoch": 15.980049875311721,
+      "grad_norm": 8.224431991577148,
+      "learning_rate": 4.026683291770574e-06,
+      "loss": 0.3025,
+      "step": 64080
+    },
+    {
+      "epoch": 15.982543640897756,
+      "grad_norm": 9.23838996887207,
+      "learning_rate": 4.024189526184539e-06,
+      "loss": 0.4478,
+      "step": 64090
+    },
+    {
+      "epoch": 15.98503740648379,
+      "grad_norm": 11.388845443725586,
+      "learning_rate": 4.021695760598504e-06,
+      "loss": 0.3425,
+      "step": 64100
+    },
+    {
+      "epoch": 15.987531172069826,
+      "grad_norm": 7.132105350494385,
+      "learning_rate": 4.0192019950124685e-06,
+      "loss": 0.3889,
+      "step": 64110
+    },
+    {
+      "epoch": 15.99002493765586,
+      "grad_norm": 7.140593528747559,
+      "learning_rate": 4.016708229426435e-06,
+      "loss": 0.319,
+      "step": 64120
+    },
+    {
+      "epoch": 15.992518703241895,
+      "grad_norm": 7.605659484863281,
+      "learning_rate": 4.0142144638404e-06,
+      "loss": 0.3565,
+      "step": 64130
+    },
+    {
+      "epoch": 15.99501246882793,
+      "grad_norm": 9.850330352783203,
+      "learning_rate": 4.011720698254364e-06,
+      "loss": 0.363,
+      "step": 64140
+    },
+    {
+      "epoch": 15.997506234413965,
+      "grad_norm": 5.737972259521484,
+      "learning_rate": 4.00922693266833e-06,
+      "loss": 0.3383,
+      "step": 64150
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 9.490584373474121,
+      "learning_rate": 4.006733167082294e-06,
+      "loss": 0.3047,
+      "step": 64160
+    },
+    {
+      "epoch": 16.0,
+      "eval_loss": 0.41599616408348083,
+      "eval_runtime": 60.0234,
+      "eval_samples_per_second": 16.71,
+      "eval_steps_per_second": 16.71,
+      "step": 64160
+    },
+    {
+      "epoch": 16.002493765586035,
+      "grad_norm": 13.953182220458984,
+      "learning_rate": 4.0042394014962595e-06,
+      "loss": 0.3067,
+      "step": 64170
+    },
+    {
+      "epoch": 16.00498753117207,
+      "grad_norm": 7.221786022186279,
+      "learning_rate": 4.001745635910225e-06,
+      "loss": 0.3639,
+      "step": 64180
+    },
+    {
+      "epoch": 16.007481296758105,
+      "grad_norm": 8.095586776733398,
+      "learning_rate": 3.99925187032419e-06,
+      "loss": 0.3427,
+      "step": 64190
+    },
+    {
+      "epoch": 16.00997506234414,
+      "grad_norm": 5.289329528808594,
+      "learning_rate": 3.9967581047381546e-06,
+      "loss": 0.3658,
+      "step": 64200
+    },
+    {
+      "epoch": 16.012468827930174,
+      "grad_norm": 10.353299140930176,
+      "learning_rate": 3.99426433915212e-06,
+      "loss": 0.3201,
+      "step": 64210
+    },
+    {
+      "epoch": 16.01496259351621,
+      "grad_norm": 6.504826545715332,
+      "learning_rate": 3.991770573566085e-06,
+      "loss": 0.338,
+      "step": 64220
+    },
+    {
+      "epoch": 16.017456359102244,
+      "grad_norm": 10.51669692993164,
+      "learning_rate": 3.9892768079800505e-06,
+      "loss": 0.2674,
+      "step": 64230
+    },
+    {
+      "epoch": 16.01995012468828,
+      "grad_norm": 10.650725364685059,
+      "learning_rate": 3.986783042394015e-06,
+      "loss": 0.2787,
+      "step": 64240
+    },
+    {
+      "epoch": 16.022443890274314,
+      "grad_norm": 8.505513191223145,
+      "learning_rate": 3.98428927680798e-06,
+      "loss": 0.2972,
+      "step": 64250
+    },
+    {
+      "epoch": 16.02493765586035,
+      "grad_norm": 10.878944396972656,
+      "learning_rate": 3.9817955112219456e-06,
+      "loss": 0.3168,
+      "step": 64260
+    },
+    {
+      "epoch": 16.027431421446384,
+      "grad_norm": 7.452136993408203,
+      "learning_rate": 3.979301745635911e-06,
+      "loss": 0.319,
+      "step": 64270
+    },
+    {
+      "epoch": 16.02992518703242,
+      "grad_norm": 7.434098720550537,
+      "learning_rate": 3.976807980049875e-06,
+      "loss": 0.3388,
+      "step": 64280
+    },
+    {
+      "epoch": 16.032418952618453,
+      "grad_norm": 6.33979606628418,
+      "learning_rate": 3.974314214463841e-06,
+      "loss": 0.2748,
+      "step": 64290
+    },
+    {
+      "epoch": 16.034912718204488,
+      "grad_norm": 13.049262046813965,
+      "learning_rate": 3.971820448877806e-06,
+      "loss": 0.3333,
+      "step": 64300
+    },
+    {
+      "epoch": 16.037406483790523,
+      "grad_norm": 12.862238883972168,
+      "learning_rate": 3.969326683291771e-06,
+      "loss": 0.3516,
+      "step": 64310
+    },
+    {
+      "epoch": 16.039900249376558,
+      "grad_norm": 7.679028034210205,
+      "learning_rate": 3.966832917705736e-06,
+      "loss": 0.3192,
+      "step": 64320
+    },
+    {
+      "epoch": 16.042394014962593,
+      "grad_norm": 7.861089706420898,
+      "learning_rate": 3.964339152119701e-06,
+      "loss": 0.2909,
+      "step": 64330
+    },
+    {
+      "epoch": 16.044887780548628,
+      "grad_norm": 7.005455493927002,
+      "learning_rate": 3.9618453865336655e-06,
+      "loss": 0.3583,
+      "step": 64340
+    },
+    {
+      "epoch": 16.047381546134662,
+      "grad_norm": 11.43169116973877,
+      "learning_rate": 3.959351620947632e-06,
+      "loss": 0.3038,
+      "step": 64350
+    },
+    {
+      "epoch": 16.049875311720697,
+      "grad_norm": 8.389317512512207,
+      "learning_rate": 3.956857855361596e-06,
+      "loss": 0.2716,
+      "step": 64360
+    },
+    {
+      "epoch": 16.052369077306732,
+      "grad_norm": 8.883370399475098,
+      "learning_rate": 3.9543640897755614e-06,
+      "loss": 0.3067,
+      "step": 64370
+    },
+    {
+      "epoch": 16.054862842892767,
+      "grad_norm": 10.897636413574219,
+      "learning_rate": 3.951870324189527e-06,
+      "loss": 0.297,
+      "step": 64380
+    },
+    {
+      "epoch": 16.057356608478802,
+      "grad_norm": 9.456404685974121,
+      "learning_rate": 3.949376558603491e-06,
+      "loss": 0.3652,
+      "step": 64390
+    },
+    {
+      "epoch": 16.059850374064837,
+      "grad_norm": 8.108953475952148,
+      "learning_rate": 3.9468827930174565e-06,
+      "loss": 0.323,
+      "step": 64400
+    },
+    {
+      "epoch": 16.06234413965087,
+      "grad_norm": 5.497220039367676,
+      "learning_rate": 3.944389027431422e-06,
+      "loss": 0.2902,
+      "step": 64410
+    },
+    {
+      "epoch": 16.064837905236907,
+      "grad_norm": 8.865570068359375,
+      "learning_rate": 3.941895261845387e-06,
+      "loss": 0.356,
+      "step": 64420
+    },
+    {
+      "epoch": 16.06733167082294,
+      "grad_norm": 6.835111141204834,
+      "learning_rate": 3.939401496259352e-06,
+      "loss": 0.3206,
+      "step": 64430
+    },
+    {
+      "epoch": 16.069825436408976,
+      "grad_norm": 10.127936363220215,
+      "learning_rate": 3.936907730673317e-06,
+      "loss": 0.2834,
+      "step": 64440
+    },
+    {
+      "epoch": 16.07231920199501,
+      "grad_norm": 10.34428882598877,
+      "learning_rate": 3.934413965087282e-06,
+      "loss": 0.2795,
+      "step": 64450
+    },
+    {
+      "epoch": 16.074812967581046,
+      "grad_norm": 7.148595809936523,
+      "learning_rate": 3.9319201995012475e-06,
+      "loss": 0.3463,
+      "step": 64460
+    },
+    {
+      "epoch": 16.07730673316708,
+      "grad_norm": 7.980280876159668,
+      "learning_rate": 3.929426433915212e-06,
+      "loss": 0.3167,
+      "step": 64470
+    },
+    {
+      "epoch": 16.079800498753116,
+      "grad_norm": 11.756796836853027,
+      "learning_rate": 3.926932668329177e-06,
+      "loss": 0.3765,
+      "step": 64480
+    },
+    {
+      "epoch": 16.08229426433915,
+      "grad_norm": 11.97585391998291,
+      "learning_rate": 3.924438902743143e-06,
+      "loss": 0.3805,
+      "step": 64490
+    },
+    {
+      "epoch": 16.084788029925186,
+      "grad_norm": 11.348858833312988,
+      "learning_rate": 3.921945137157108e-06,
+      "loss": 0.2978,
+      "step": 64500
+    },
+    {
+      "epoch": 16.08728179551122,
+      "grad_norm": 6.33409309387207,
+      "learning_rate": 3.919451371571072e-06,
+      "loss": 0.3154,
+      "step": 64510
+    },
+    {
+      "epoch": 16.089775561097255,
+      "grad_norm": 10.497039794921875,
+      "learning_rate": 3.916957605985038e-06,
+      "loss": 0.2952,
+      "step": 64520
+    },
+    {
+      "epoch": 16.09226932668329,
+      "grad_norm": 6.439425945281982,
+      "learning_rate": 3.914463840399003e-06,
+      "loss": 0.3474,
+      "step": 64530
+    },
+    {
+      "epoch": 16.094763092269325,
+      "grad_norm": 11.366449356079102,
+      "learning_rate": 3.911970074812968e-06,
+      "loss": 0.3775,
+      "step": 64540
+    },
+    {
+      "epoch": 16.09725685785536,
+      "grad_norm": 7.353078365325928,
+      "learning_rate": 3.909476309226933e-06,
+      "loss": 0.3125,
+      "step": 64550
+    },
+    {
+      "epoch": 16.099750623441395,
+      "grad_norm": 7.988730430603027,
+      "learning_rate": 3.906982543640898e-06,
+      "loss": 0.3449,
+      "step": 64560
+    },
+    {
+      "epoch": 16.102244389027433,
+      "grad_norm": 7.281782627105713,
+      "learning_rate": 3.9044887780548625e-06,
+      "loss": 0.3189,
+      "step": 64570
+    },
+    {
+      "epoch": 16.104738154613468,
+      "grad_norm": 7.062814712524414,
+      "learning_rate": 3.901995012468829e-06,
+      "loss": 0.3316,
+      "step": 64580
+    },
+    {
+      "epoch": 16.107231920199503,
+      "grad_norm": 6.708259582519531,
+      "learning_rate": 3.899501246882793e-06,
+      "loss": 0.2937,
+      "step": 64590
+    },
+    {
+      "epoch": 16.109725685785538,
+      "grad_norm": 7.511795520782471,
+      "learning_rate": 3.8970074812967585e-06,
+      "loss": 0.3319,
+      "step": 64600
+    },
+    {
+      "epoch": 16.112219451371573,
+      "grad_norm": 7.629890441894531,
+      "learning_rate": 3.894513715710724e-06,
+      "loss": 0.3135,
+      "step": 64610
+    },
+    {
+      "epoch": 16.114713216957608,
+      "grad_norm": 12.183755874633789,
+      "learning_rate": 3.892019950124688e-06,
+      "loss": 0.3092,
+      "step": 64620
+    },
+    {
+      "epoch": 16.117206982543642,
+      "grad_norm": 12.015825271606445,
+      "learning_rate": 3.8895261845386536e-06,
+      "loss": 0.3199,
+      "step": 64630
+    },
+    {
+      "epoch": 16.119700748129677,
+      "grad_norm": 8.182219505310059,
+      "learning_rate": 3.887032418952619e-06,
+      "loss": 0.3227,
+      "step": 64640
+    },
+    {
+      "epoch": 16.122194513715712,
+      "grad_norm": 8.107385635375977,
+      "learning_rate": 3.884538653366584e-06,
+      "loss": 0.275,
+      "step": 64650
+    },
+    {
+      "epoch": 16.124688279301747,
+      "grad_norm": 12.364629745483398,
+      "learning_rate": 3.882044887780549e-06,
+      "loss": 0.3205,
+      "step": 64660
+    },
+    {
+      "epoch": 16.127182044887782,
+      "grad_norm": 9.689504623413086,
+      "learning_rate": 3.879551122194514e-06,
+      "loss": 0.3629,
+      "step": 64670
+    },
+    {
+      "epoch": 16.129675810473817,
+      "grad_norm": 9.650782585144043,
+      "learning_rate": 3.877057356608479e-06,
+      "loss": 0.3607,
+      "step": 64680
+    },
+    {
+      "epoch": 16.13216957605985,
+      "grad_norm": 8.96108341217041,
+      "learning_rate": 3.8745635910224446e-06,
+      "loss": 0.3243,
+      "step": 64690
+    },
+    {
+      "epoch": 16.134663341645886,
+      "grad_norm": 9.31535816192627,
+      "learning_rate": 3.872069825436409e-06,
+      "loss": 0.353,
+      "step": 64700
+    },
+    {
+      "epoch": 16.13715710723192,
+      "grad_norm": 9.662627220153809,
+      "learning_rate": 3.869576059850374e-06,
+      "loss": 0.2988,
+      "step": 64710
+    },
+    {
+      "epoch": 16.139650872817956,
+      "grad_norm": 6.727764129638672,
+      "learning_rate": 3.86708229426434e-06,
+      "loss": 0.2802,
+      "step": 64720
+    },
+    {
+      "epoch": 16.14214463840399,
+      "grad_norm": 11.656500816345215,
+      "learning_rate": 3.864588528678305e-06,
+      "loss": 0.319,
+      "step": 64730
+    },
+    {
+      "epoch": 16.144638403990026,
+      "grad_norm": 9.718652725219727,
+      "learning_rate": 3.862094763092269e-06,
+      "loss": 0.3288,
+      "step": 64740
+    },
+    {
+      "epoch": 16.14713216957606,
+      "grad_norm": 9.155292510986328,
+      "learning_rate": 3.859600997506235e-06,
+      "loss": 0.305,
+      "step": 64750
+    },
+    {
+      "epoch": 16.149625935162096,
+      "grad_norm": 8.969863891601562,
+      "learning_rate": 3.8571072319202e-06,
+      "loss": 0.3435,
+      "step": 64760
+    },
+    {
+      "epoch": 16.15211970074813,
+      "grad_norm": 9.428884506225586,
+      "learning_rate": 3.854613466334165e-06,
+      "loss": 0.3198,
+      "step": 64770
+    },
+    {
+      "epoch": 16.154613466334165,
+      "grad_norm": 10.018813133239746,
+      "learning_rate": 3.85211970074813e-06,
+      "loss": 0.3452,
+      "step": 64780
+    },
+    {
+      "epoch": 16.1571072319202,
+      "grad_norm": 11.803756713867188,
+      "learning_rate": 3.849625935162095e-06,
+      "loss": 0.3619,
+      "step": 64790
+    },
+    {
+      "epoch": 16.159600997506235,
+      "grad_norm": 14.001852989196777,
+      "learning_rate": 3.84713216957606e-06,
+      "loss": 0.3259,
+      "step": 64800
+    },
+    {
+      "epoch": 16.16209476309227,
+      "grad_norm": 7.694314956665039,
+      "learning_rate": 3.844638403990025e-06,
+      "loss": 0.3697,
+      "step": 64810
+    },
+    {
+      "epoch": 16.164588528678305,
+      "grad_norm": 7.822086811065674,
+      "learning_rate": 3.84214463840399e-06,
+      "loss": 0.2878,
+      "step": 64820
+    },
+    {
+      "epoch": 16.16708229426434,
+      "grad_norm": 8.987359046936035,
+      "learning_rate": 3.8396508728179555e-06,
+      "loss": 0.3268,
+      "step": 64830
+    },
+    {
+      "epoch": 16.169576059850375,
+      "grad_norm": 6.423337459564209,
+      "learning_rate": 3.83715710723192e-06,
+      "loss": 0.3347,
+      "step": 64840
+    },
+    {
+      "epoch": 16.17206982543641,
+      "grad_norm": 9.673188209533691,
+      "learning_rate": 3.834663341645885e-06,
+      "loss": 0.3471,
+      "step": 64850
+    },
+    {
+      "epoch": 16.174563591022444,
+      "grad_norm": 4.7321319580078125,
+      "learning_rate": 3.832169576059851e-06,
+      "loss": 0.3425,
+      "step": 64860
+    },
+    {
+      "epoch": 16.17705735660848,
+      "grad_norm": 10.942599296569824,
+      "learning_rate": 3.829675810473816e-06,
+      "loss": 0.2551,
+      "step": 64870
+    },
+    {
+      "epoch": 16.179551122194514,
+      "grad_norm": 9.56152629852295,
+      "learning_rate": 3.827182044887781e-06,
+      "loss": 0.3806,
+      "step": 64880
+    },
+    {
+      "epoch": 16.18204488778055,
+      "grad_norm": 10.860115051269531,
+      "learning_rate": 3.824688279301746e-06,
+      "loss": 0.3677,
+      "step": 64890
+    },
+    {
+      "epoch": 16.184538653366584,
+      "grad_norm": 6.14926815032959,
+      "learning_rate": 3.822194513715711e-06,
+      "loss": 0.3018,
+      "step": 64900
+    },
+    {
+      "epoch": 16.18703241895262,
+      "grad_norm": 9.40666675567627,
+      "learning_rate": 3.819700748129676e-06,
+      "loss": 0.307,
+      "step": 64910
+    },
+    {
+      "epoch": 16.189526184538654,
+      "grad_norm": 10.993182182312012,
+      "learning_rate": 3.817206982543642e-06,
+      "loss": 0.287,
+      "step": 64920
+    },
+    {
+      "epoch": 16.19201995012469,
+      "grad_norm": 8.050580978393555,
+      "learning_rate": 3.814713216957606e-06,
+      "loss": 0.3454,
+      "step": 64930
+    },
+    {
+      "epoch": 16.194513715710723,
+      "grad_norm": 7.049538612365723,
+      "learning_rate": 3.8122194513715714e-06,
+      "loss": 0.3349,
+      "step": 64940
+    },
+    {
+      "epoch": 16.197007481296758,
+      "grad_norm": 6.026520729064941,
+      "learning_rate": 3.8097256857855363e-06,
+      "loss": 0.2819,
+      "step": 64950
+    },
+    {
+      "epoch": 16.199501246882793,
+      "grad_norm": 9.586450576782227,
+      "learning_rate": 3.8072319201995016e-06,
+      "loss": 0.369,
+      "step": 64960
+    },
+    {
+      "epoch": 16.201995012468828,
+      "grad_norm": 12.329106330871582,
+      "learning_rate": 3.8047381546134665e-06,
+      "loss": 0.2958,
+      "step": 64970
+    },
+    {
+      "epoch": 16.204488778054863,
+      "grad_norm": 9.353446960449219,
+      "learning_rate": 3.8022443890274318e-06,
+      "loss": 0.3131,
+      "step": 64980
+    },
+    {
+      "epoch": 16.206982543640898,
+      "grad_norm": 7.740355014801025,
+      "learning_rate": 3.7997506234413967e-06,
+      "loss": 0.265,
+      "step": 64990
+    },
+    {
+      "epoch": 16.209476309226932,
+      "grad_norm": 7.058204174041748,
+      "learning_rate": 3.797256857855362e-06,
+      "loss": 0.3649,
+      "step": 65000
+    },
+    {
+      "epoch": 16.211970074812967,
+      "grad_norm": 9.42304801940918,
+      "learning_rate": 3.794763092269327e-06,
+      "loss": 0.3743,
+      "step": 65010
+    },
+    {
+      "epoch": 16.214463840399002,
+      "grad_norm": 10.909148216247559,
+      "learning_rate": 3.792269326683292e-06,
+      "loss": 0.3381,
+      "step": 65020
+    },
+    {
+      "epoch": 16.216957605985037,
+      "grad_norm": 6.674330711364746,
+      "learning_rate": 3.789775561097257e-06,
+      "loss": 0.3916,
+      "step": 65030
+    },
+    {
+      "epoch": 16.219451371571072,
+      "grad_norm": 7.515557765960693,
+      "learning_rate": 3.7872817955112224e-06,
+      "loss": 0.2461,
+      "step": 65040
+    },
+    {
+      "epoch": 16.221945137157107,
+      "grad_norm": 8.817490577697754,
+      "learning_rate": 3.7847880299251872e-06,
+      "loss": 0.282,
+      "step": 65050
+    },
+    {
+      "epoch": 16.22443890274314,
+      "grad_norm": 9.24267578125,
+      "learning_rate": 3.7822942643391526e-06,
+      "loss": 0.3451,
+      "step": 65060
+    },
+    {
+      "epoch": 16.226932668329177,
+      "grad_norm": 8.435280799865723,
+      "learning_rate": 3.7798004987531174e-06,
+      "loss": 0.3645,
+      "step": 65070
+    },
+    {
+      "epoch": 16.22942643391521,
+      "grad_norm": 8.729866027832031,
+      "learning_rate": 3.7773067331670827e-06,
+      "loss": 0.346,
+      "step": 65080
+    },
+    {
+      "epoch": 16.231920199501246,
+      "grad_norm": 9.877275466918945,
+      "learning_rate": 3.7748129675810476e-06,
+      "loss": 0.3218,
+      "step": 65090
+    },
+    {
+      "epoch": 16.23441396508728,
+      "grad_norm": 7.41887903213501,
+      "learning_rate": 3.772319201995013e-06,
+      "loss": 0.2789,
+      "step": 65100
+    },
+    {
+      "epoch": 16.236907730673316,
+      "grad_norm": 8.857915878295898,
+      "learning_rate": 3.7698254364089783e-06,
+      "loss": 0.347,
+      "step": 65110
+    },
+    {
+      "epoch": 16.23940149625935,
+      "grad_norm": 9.079187393188477,
+      "learning_rate": 3.767331670822943e-06,
+      "loss": 0.2908,
+      "step": 65120
+    },
+    {
+      "epoch": 16.241895261845386,
+      "grad_norm": 8.195780754089355,
+      "learning_rate": 3.7648379052369085e-06,
+      "loss": 0.3252,
+      "step": 65130
+    },
+    {
+      "epoch": 16.24438902743142,
+      "grad_norm": 11.435444831848145,
+      "learning_rate": 3.762344139650873e-06,
+      "loss": 0.377,
+      "step": 65140
+    },
+    {
+      "epoch": 16.246882793017456,
+      "grad_norm": 8.027785301208496,
+      "learning_rate": 3.7598503740648386e-06,
+      "loss": 0.2409,
+      "step": 65150
+    },
+    {
+      "epoch": 16.24937655860349,
+      "grad_norm": 9.426610946655273,
+      "learning_rate": 3.757356608478803e-06,
+      "loss": 0.3731,
+      "step": 65160
+    },
+    {
+      "epoch": 16.251870324189525,
+      "grad_norm": 11.091520309448242,
+      "learning_rate": 3.7548628428927684e-06,
+      "loss": 0.3764,
+      "step": 65170
+    },
+    {
+      "epoch": 16.25436408977556,
+      "grad_norm": 5.973906517028809,
+      "learning_rate": 3.7523690773067333e-06,
+      "loss": 0.2766,
+      "step": 65180
+    },
+    {
+      "epoch": 16.256857855361595,
+      "grad_norm": 8.147195816040039,
+      "learning_rate": 3.7498753117206986e-06,
+      "loss": 0.3152,
+      "step": 65190
+    },
+    {
+      "epoch": 16.25935162094763,
+      "grad_norm": 8.979331016540527,
+      "learning_rate": 3.7473815461346635e-06,
+      "loss": 0.3257,
+      "step": 65200
+    },
+    {
+      "epoch": 16.261845386533665,
+      "grad_norm": 11.531305313110352,
+      "learning_rate": 3.744887780548629e-06,
+      "loss": 0.3556,
+      "step": 65210
+    },
+    {
+      "epoch": 16.2643391521197,
+      "grad_norm": 11.863848686218262,
+      "learning_rate": 3.7423940149625937e-06,
+      "loss": 0.3515,
+      "step": 65220
+    },
+    {
+      "epoch": 16.266832917705734,
+      "grad_norm": 5.3083038330078125,
+      "learning_rate": 3.739900249376559e-06,
+      "loss": 0.2752,
+      "step": 65230
+    },
+    {
+      "epoch": 16.26932668329177,
+      "grad_norm": 7.9635820388793945,
+      "learning_rate": 3.737406483790524e-06,
+      "loss": 0.3552,
+      "step": 65240
+    },
+    {
+      "epoch": 16.271820448877804,
+      "grad_norm": 10.961655616760254,
+      "learning_rate": 3.734912718204489e-06,
+      "loss": 0.3359,
+      "step": 65250
+    },
+    {
+      "epoch": 16.27431421446384,
+      "grad_norm": 8.826502799987793,
+      "learning_rate": 3.732418952618454e-06,
+      "loss": 0.2981,
+      "step": 65260
+    },
+    {
+      "epoch": 16.276807980049874,
+      "grad_norm": 11.395182609558105,
+      "learning_rate": 3.7299251870324194e-06,
+      "loss": 0.3708,
+      "step": 65270
+    },
+    {
+      "epoch": 16.27930174563591,
+      "grad_norm": 7.6695146560668945,
+      "learning_rate": 3.7274314214463843e-06,
+      "loss": 0.3029,
+      "step": 65280
+    },
+    {
+      "epoch": 16.281795511221944,
+      "grad_norm": 10.41583251953125,
+      "learning_rate": 3.7249376558603496e-06,
+      "loss": 0.3226,
+      "step": 65290
+    },
+    {
+      "epoch": 16.28428927680798,
+      "grad_norm": 7.618106842041016,
+      "learning_rate": 3.7224438902743145e-06,
+      "loss": 0.3199,
+      "step": 65300
+    },
+    {
+      "epoch": 16.286783042394013,
+      "grad_norm": 8.326313018798828,
+      "learning_rate": 3.71995012468828e-06,
+      "loss": 0.3513,
+      "step": 65310
+    },
+    {
+      "epoch": 16.28927680798005,
+      "grad_norm": 5.363463878631592,
+      "learning_rate": 3.7174563591022443e-06,
+      "loss": 0.3408,
+      "step": 65320
+    },
+    {
+      "epoch": 16.291770573566083,
+      "grad_norm": 9.034497261047363,
+      "learning_rate": 3.71496259351621e-06,
+      "loss": 0.324,
+      "step": 65330
+    },
+    {
+      "epoch": 16.294264339152118,
+      "grad_norm": 13.3429536819458,
+      "learning_rate": 3.7124688279301744e-06,
+      "loss": 0.3495,
+      "step": 65340
+    },
+    {
+      "epoch": 16.296758104738153,
+      "grad_norm": 14.830777168273926,
+      "learning_rate": 3.7099750623441398e-06,
+      "loss": 0.3499,
+      "step": 65350
+    },
+    {
+      "epoch": 16.29925187032419,
+      "grad_norm": 6.612533092498779,
+      "learning_rate": 3.7074812967581055e-06,
+      "loss": 0.3221,
+      "step": 65360
+    },
+    {
+      "epoch": 16.301745635910226,
+      "grad_norm": 6.141602516174316,
+      "learning_rate": 3.70498753117207e-06,
+      "loss": 0.3303,
+      "step": 65370
+    },
+    {
+      "epoch": 16.30423940149626,
+      "grad_norm": 8.41512393951416,
+      "learning_rate": 3.7024937655860353e-06,
+      "loss": 0.2878,
+      "step": 65380
+    },
+    {
+      "epoch": 16.306733167082296,
+      "grad_norm": 5.508592128753662,
+      "learning_rate": 3.7e-06,
+      "loss": 0.3166,
+      "step": 65390
+    },
+    {
+      "epoch": 16.30922693266833,
+      "grad_norm": 9.821744918823242,
+      "learning_rate": 3.6975062344139655e-06,
+      "loss": 0.3125,
+      "step": 65400
+    },
+    {
+      "epoch": 16.311720698254366,
+      "grad_norm": 6.723392963409424,
+      "learning_rate": 3.6950124688279303e-06,
+      "loss": 0.3018,
+      "step": 65410
+    },
+    {
+      "epoch": 16.3142144638404,
+      "grad_norm": 8.72385311126709,
+      "learning_rate": 3.6925187032418957e-06,
+      "loss": 0.3337,
+      "step": 65420
+    },
+    {
+      "epoch": 16.316708229426435,
+      "grad_norm": 12.842667579650879,
+      "learning_rate": 3.6900249376558605e-06,
+      "loss": 0.3676,
+      "step": 65430
+    },
+    {
+      "epoch": 16.31920199501247,
+      "grad_norm": 9.1225004196167,
+      "learning_rate": 3.687531172069826e-06,
+      "loss": 0.3989,
+      "step": 65440
+    },
+    {
+      "epoch": 16.321695760598505,
+      "grad_norm": 11.330621719360352,
+      "learning_rate": 3.6850374064837907e-06,
+      "loss": 0.3592,
+      "step": 65450
+    },
+    {
+      "epoch": 16.32418952618454,
+      "grad_norm": 10.708224296569824,
+      "learning_rate": 3.682543640897756e-06,
+      "loss": 0.3368,
+      "step": 65460
+    },
+    {
+      "epoch": 16.326683291770575,
+      "grad_norm": 11.212067604064941,
+      "learning_rate": 3.680049875311721e-06,
+      "loss": 0.3081,
+      "step": 65470
+    },
+    {
+      "epoch": 16.32917705735661,
+      "grad_norm": 6.18818473815918,
+      "learning_rate": 3.6775561097256862e-06,
+      "loss": 0.3428,
+      "step": 65480
+    },
+    {
+      "epoch": 16.331670822942645,
+      "grad_norm": 7.7143635749816895,
+      "learning_rate": 3.675062344139651e-06,
+      "loss": 0.2853,
+      "step": 65490
+    },
+    {
+      "epoch": 16.33416458852868,
+      "grad_norm": 9.196212768554688,
+      "learning_rate": 3.6725685785536164e-06,
+      "loss": 0.3176,
+      "step": 65500
+    },
+    {
+      "epoch": 16.336658354114714,
+      "grad_norm": 9.68375301361084,
+      "learning_rate": 3.6700748129675813e-06,
+      "loss": 0.3151,
+      "step": 65510
+    },
+    {
+      "epoch": 16.33915211970075,
+      "grad_norm": 7.000097274780273,
+      "learning_rate": 3.6675810473815466e-06,
+      "loss": 0.3043,
+      "step": 65520
+    },
+    {
+      "epoch": 16.341645885286784,
+      "grad_norm": 6.249640464782715,
+      "learning_rate": 3.6650872817955115e-06,
+      "loss": 0.3183,
+      "step": 65530
+    },
+    {
+      "epoch": 16.34413965087282,
+      "grad_norm": 9.410320281982422,
+      "learning_rate": 3.662593516209477e-06,
+      "loss": 0.355,
+      "step": 65540
+    },
+    {
+      "epoch": 16.346633416458854,
+      "grad_norm": 8.34964656829834,
+      "learning_rate": 3.6600997506234413e-06,
+      "loss": 0.3565,
+      "step": 65550
+    },
+    {
+      "epoch": 16.34912718204489,
+      "grad_norm": 7.488079071044922,
+      "learning_rate": 3.657605985037407e-06,
+      "loss": 0.3327,
+      "step": 65560
+    },
+    {
+      "epoch": 16.351620947630924,
+      "grad_norm": 12.476749420166016,
+      "learning_rate": 3.6551122194513715e-06,
+      "loss": 0.2578,
+      "step": 65570
+    },
+    {
+      "epoch": 16.35411471321696,
+      "grad_norm": 7.019128322601318,
+      "learning_rate": 3.652618453865337e-06,
+      "loss": 0.332,
+      "step": 65580
+    },
+    {
+      "epoch": 16.356608478802993,
+      "grad_norm": 7.373568058013916,
+      "learning_rate": 3.6501246882793017e-06,
+      "loss": 0.3988,
+      "step": 65590
+    },
+    {
+      "epoch": 16.359102244389028,
+      "grad_norm": 8.25925064086914,
+      "learning_rate": 3.647630922693267e-06,
+      "loss": 0.359,
+      "step": 65600
+    },
+    {
+      "epoch": 16.361596009975063,
+      "grad_norm": 9.343754768371582,
+      "learning_rate": 3.6451371571072323e-06,
+      "loss": 0.3369,
+      "step": 65610
+    },
+    {
+      "epoch": 16.364089775561098,
+      "grad_norm": 7.88173770904541,
+      "learning_rate": 3.642892768079801e-06,
+      "loss": 0.3372,
+      "step": 65620
+    },
+    {
+      "epoch": 16.366583541147133,
+      "grad_norm": 12.04601001739502,
+      "learning_rate": 3.640399002493766e-06,
+      "loss": 0.3215,
+      "step": 65630
+    },
+    {
+      "epoch": 16.369077306733168,
+      "grad_norm": 11.845905303955078,
+      "learning_rate": 3.637905236907731e-06,
+      "loss": 0.4036,
+      "step": 65640
+    },
+    {
+      "epoch": 16.371571072319203,
+      "grad_norm": 11.93884563446045,
+      "learning_rate": 3.635411471321696e-06,
+      "loss": 0.3596,
+      "step": 65650
+    },
+    {
+      "epoch": 16.374064837905237,
+      "grad_norm": 8.246026992797852,
+      "learning_rate": 3.6329177057356614e-06,
+      "loss": 0.2996,
+      "step": 65660
+    },
+    {
+      "epoch": 16.376558603491272,
+      "grad_norm": 10.389656066894531,
+      "learning_rate": 3.6304239401496263e-06,
+      "loss": 0.3356,
+      "step": 65670
+    },
+    {
+      "epoch": 16.379052369077307,
+      "grad_norm": 9.619356155395508,
+      "learning_rate": 3.6279301745635916e-06,
+      "loss": 0.3053,
+      "step": 65680
+    },
+    {
+      "epoch": 16.381546134663342,
+      "grad_norm": 10.337359428405762,
+      "learning_rate": 3.625436408977556e-06,
+      "loss": 0.3775,
+      "step": 65690
+    },
+    {
+      "epoch": 16.384039900249377,
+      "grad_norm": 9.273616790771484,
+      "learning_rate": 3.6229426433915218e-06,
+      "loss": 0.3492,
+      "step": 65700
+    },
+    {
+      "epoch": 16.38653366583541,
+      "grad_norm": 7.120643138885498,
+      "learning_rate": 3.6204488778054862e-06,
+      "loss": 0.3044,
+      "step": 65710
+    },
+    {
+      "epoch": 16.389027431421447,
+      "grad_norm": 7.253455638885498,
+      "learning_rate": 3.6179551122194516e-06,
+      "loss": 0.3432,
+      "step": 65720
+    },
+    {
+      "epoch": 16.39152119700748,
+      "grad_norm": 9.472309112548828,
+      "learning_rate": 3.6154613466334164e-06,
+      "loss": 0.3314,
+      "step": 65730
+    },
+    {
+      "epoch": 16.394014962593516,
+      "grad_norm": 12.149874687194824,
+      "learning_rate": 3.6129675810473818e-06,
+      "loss": 0.4217,
+      "step": 65740
+    },
+    {
+      "epoch": 16.39650872817955,
+      "grad_norm": 9.527915954589844,
+      "learning_rate": 3.6104738154613466e-06,
+      "loss": 0.3714,
+      "step": 65750
+    },
+    {
+      "epoch": 16.399002493765586,
+      "grad_norm": 11.615216255187988,
+      "learning_rate": 3.607980049875312e-06,
+      "loss": 0.3939,
+      "step": 65760
+    },
+    {
+      "epoch": 16.40149625935162,
+      "grad_norm": 7.617849826812744,
+      "learning_rate": 3.605486284289277e-06,
+      "loss": 0.3533,
+      "step": 65770
+    },
+    {
+      "epoch": 16.403990024937656,
+      "grad_norm": 6.392154693603516,
+      "learning_rate": 3.602992518703242e-06,
+      "loss": 0.3103,
+      "step": 65780
+    },
+    {
+      "epoch": 16.40648379052369,
+      "grad_norm": 9.79951000213623,
+      "learning_rate": 3.6004987531172075e-06,
+      "loss": 0.3713,
+      "step": 65790
+    },
+    {
+      "epoch": 16.408977556109726,
+      "grad_norm": 8.584783554077148,
+      "learning_rate": 3.5980049875311723e-06,
+      "loss": 0.3699,
+      "step": 65800
+    },
+    {
+      "epoch": 16.41147132169576,
+      "grad_norm": 9.751562118530273,
+      "learning_rate": 3.5955112219451376e-06,
+      "loss": 0.3959,
+      "step": 65810
+    },
+    {
+      "epoch": 16.413965087281795,
+      "grad_norm": 6.995100975036621,
+      "learning_rate": 3.5930174563591025e-06,
+      "loss": 0.2217,
+      "step": 65820
+    },
+    {
+      "epoch": 16.41645885286783,
+      "grad_norm": 12.037008285522461,
+      "learning_rate": 3.590523690773068e-06,
+      "loss": 0.3115,
+      "step": 65830
+    },
+    {
+      "epoch": 16.418952618453865,
+      "grad_norm": 6.427266597747803,
+      "learning_rate": 3.5880299251870327e-06,
+      "loss": 0.3391,
+      "step": 65840
+    },
+    {
+      "epoch": 16.4214463840399,
+      "grad_norm": 6.5554890632629395,
+      "learning_rate": 3.585536159600998e-06,
+      "loss": 0.3549,
+      "step": 65850
+    },
+    {
+      "epoch": 16.423940149625935,
+      "grad_norm": 7.430528163909912,
+      "learning_rate": 3.583042394014963e-06,
+      "loss": 0.322,
+      "step": 65860
+    },
+    {
+      "epoch": 16.42643391521197,
+      "grad_norm": 6.291595935821533,
+      "learning_rate": 3.5805486284289282e-06,
+      "loss": 0.3099,
+      "step": 65870
+    },
+    {
+      "epoch": 16.428927680798004,
+      "grad_norm": 8.320535659790039,
+      "learning_rate": 3.578054862842893e-06,
+      "loss": 0.304,
+      "step": 65880
+    },
+    {
+      "epoch": 16.43142144638404,
+      "grad_norm": 8.501724243164062,
+      "learning_rate": 3.5755610972568584e-06,
+      "loss": 0.3474,
+      "step": 65890
+    },
+    {
+      "epoch": 16.433915211970074,
+      "grad_norm": 10.219943046569824,
+      "learning_rate": 3.573067331670823e-06,
+      "loss": 0.2741,
+      "step": 65900
+    },
+    {
+      "epoch": 16.43640897755611,
+      "grad_norm": 9.32536792755127,
+      "learning_rate": 3.5705735660847886e-06,
+      "loss": 0.3118,
+      "step": 65910
+    },
+    {
+      "epoch": 16.438902743142144,
+      "grad_norm": 7.996541976928711,
+      "learning_rate": 3.568079800498753e-06,
+      "loss": 0.3923,
+      "step": 65920
+    },
+    {
+      "epoch": 16.44139650872818,
+      "grad_norm": 7.270846366882324,
+      "learning_rate": 3.5655860349127184e-06,
+      "loss": 0.4512,
+      "step": 65930
+    },
+    {
+      "epoch": 16.443890274314214,
+      "grad_norm": 8.105086326599121,
+      "learning_rate": 3.5630922693266833e-06,
+      "loss": 0.325,
+      "step": 65940
+    },
+    {
+      "epoch": 16.44638403990025,
+      "grad_norm": 7.63092565536499,
+      "learning_rate": 3.5605985037406486e-06,
+      "loss": 0.3325,
+      "step": 65950
+    },
+    {
+      "epoch": 16.448877805486283,
+      "grad_norm": 8.287890434265137,
+      "learning_rate": 3.5581047381546135e-06,
+      "loss": 0.3664,
+      "step": 65960
+    },
+    {
+      "epoch": 16.45137157107232,
+      "grad_norm": 9.698472023010254,
+      "learning_rate": 3.555610972568579e-06,
+      "loss": 0.297,
+      "step": 65970
+    },
+    {
+      "epoch": 16.453865336658353,
+      "grad_norm": 8.163863182067871,
+      "learning_rate": 3.5531172069825437e-06,
+      "loss": 0.3469,
+      "step": 65980
+    },
+    {
+      "epoch": 16.456359102244388,
+      "grad_norm": 6.485888957977295,
+      "learning_rate": 3.550623441396509e-06,
+      "loss": 0.3444,
+      "step": 65990
+    },
+    {
+      "epoch": 16.458852867830423,
+      "grad_norm": 9.558843612670898,
+      "learning_rate": 3.548129675810474e-06,
+      "loss": 0.3527,
+      "step": 66000
+    },
+    {
+      "epoch": 16.461346633416458,
+      "grad_norm": 11.721412658691406,
+      "learning_rate": 3.545635910224439e-06,
+      "loss": 0.3955,
+      "step": 66010
+    },
+    {
+      "epoch": 16.463840399002493,
+      "grad_norm": 7.238114356994629,
+      "learning_rate": 3.5431421446384045e-06,
+      "loss": 0.3075,
+      "step": 66020
+    },
+    {
+      "epoch": 16.466334164588527,
+      "grad_norm": 8.826272964477539,
+      "learning_rate": 3.5406483790523694e-06,
+      "loss": 0.3622,
+      "step": 66030
+    },
+    {
+      "epoch": 16.468827930174562,
+      "grad_norm": 8.230854034423828,
+      "learning_rate": 3.5381546134663347e-06,
+      "loss": 0.3097,
+      "step": 66040
+    },
+    {
+      "epoch": 16.471321695760597,
+      "grad_norm": 7.81684684753418,
+      "learning_rate": 3.5356608478802996e-06,
+      "loss": 0.3132,
+      "step": 66050
+    },
+    {
+      "epoch": 16.473815461346632,
+      "grad_norm": 12.46484661102295,
+      "learning_rate": 3.533167082294265e-06,
+      "loss": 0.3048,
+      "step": 66060
+    },
+    {
+      "epoch": 16.476309226932667,
+      "grad_norm": 7.251391887664795,
+      "learning_rate": 3.5306733167082298e-06,
+      "loss": 0.313,
+      "step": 66070
+    },
+    {
+      "epoch": 16.478802992518702,
+      "grad_norm": 4.723079681396484,
+      "learning_rate": 3.528179551122195e-06,
+      "loss": 0.2885,
+      "step": 66080
+    },
+    {
+      "epoch": 16.481296758104737,
+      "grad_norm": 9.58585262298584,
+      "learning_rate": 3.52568578553616e-06,
+      "loss": 0.3646,
+      "step": 66090
+    },
+    {
+      "epoch": 16.48379052369077,
+      "grad_norm": 8.872747421264648,
+      "learning_rate": 3.5231920199501253e-06,
+      "loss": 0.3397,
+      "step": 66100
+    },
+    {
+      "epoch": 16.486284289276806,
+      "grad_norm": 7.260976314544678,
+      "learning_rate": 3.52069825436409e-06,
+      "loss": 0.2967,
+      "step": 66110
+    },
+    {
+      "epoch": 16.48877805486284,
+      "grad_norm": 8.966513633728027,
+      "learning_rate": 3.5182044887780555e-06,
+      "loss": 0.3234,
+      "step": 66120
+    },
+    {
+      "epoch": 16.491271820448876,
+      "grad_norm": 8.668418884277344,
+      "learning_rate": 3.51571072319202e-06,
+      "loss": 0.3109,
+      "step": 66130
+    },
+    {
+      "epoch": 16.49376558603491,
+      "grad_norm": 9.92924690246582,
+      "learning_rate": 3.5132169576059857e-06,
+      "loss": 0.3407,
+      "step": 66140
+    },
+    {
+      "epoch": 16.496259351620946,
+      "grad_norm": 8.693164825439453,
+      "learning_rate": 3.51072319201995e-06,
+      "loss": 0.3044,
+      "step": 66150
+    },
+    {
+      "epoch": 16.49875311720698,
+      "grad_norm": 14.11042594909668,
+      "learning_rate": 3.5082294264339154e-06,
+      "loss": 0.3371,
+      "step": 66160
+    },
+    {
+      "epoch": 16.50124688279302,
+      "grad_norm": 8.329866409301758,
+      "learning_rate": 3.5057356608478803e-06,
+      "loss": 0.3241,
+      "step": 66170
+    },
+    {
+      "epoch": 16.503740648379054,
+      "grad_norm": 7.5446553230285645,
+      "learning_rate": 3.5032418952618456e-06,
+      "loss": 0.3096,
+      "step": 66180
+    },
+    {
+      "epoch": 16.50623441396509,
+      "grad_norm": 5.738336563110352,
+      "learning_rate": 3.5007481296758105e-06,
+      "loss": 0.3996,
+      "step": 66190
+    },
+    {
+      "epoch": 16.508728179551124,
+      "grad_norm": 6.674221515655518,
+      "learning_rate": 3.498254364089776e-06,
+      "loss": 0.327,
+      "step": 66200
+    },
+    {
+      "epoch": 16.51122194513716,
+      "grad_norm": 6.403293132781982,
+      "learning_rate": 3.4957605985037407e-06,
+      "loss": 0.3152,
+      "step": 66210
+    },
+    {
+      "epoch": 16.513715710723194,
+      "grad_norm": 9.35428237915039,
+      "learning_rate": 3.493266832917706e-06,
+      "loss": 0.3089,
+      "step": 66220
+    },
+    {
+      "epoch": 16.51620947630923,
+      "grad_norm": 7.628061771392822,
+      "learning_rate": 3.490773067331671e-06,
+      "loss": 0.3204,
+      "step": 66230
+    },
+    {
+      "epoch": 16.518703241895263,
+      "grad_norm": 12.166457176208496,
+      "learning_rate": 3.4882793017456362e-06,
+      "loss": 0.354,
+      "step": 66240
+    },
+    {
+      "epoch": 16.521197007481298,
+      "grad_norm": 10.643746376037598,
+      "learning_rate": 3.485785536159601e-06,
+      "loss": 0.3516,
+      "step": 66250
+    },
+    {
+      "epoch": 16.523690773067333,
+      "grad_norm": 6.07993745803833,
+      "learning_rate": 3.4832917705735664e-06,
+      "loss": 0.3867,
+      "step": 66260
+    },
+    {
+      "epoch": 16.526184538653368,
+      "grad_norm": 7.434722423553467,
+      "learning_rate": 3.4807980049875317e-06,
+      "loss": 0.3612,
+      "step": 66270
+    },
+    {
+      "epoch": 16.528678304239403,
+      "grad_norm": 7.824409484863281,
+      "learning_rate": 3.4783042394014966e-06,
+      "loss": 0.3013,
+      "step": 66280
+    },
+    {
+      "epoch": 16.531172069825438,
+      "grad_norm": 10.760315895080566,
+      "learning_rate": 3.475810473815462e-06,
+      "loss": 0.3326,
+      "step": 66290
+    },
+    {
+      "epoch": 16.533665835411473,
+      "grad_norm": 9.23198413848877,
+      "learning_rate": 3.473316708229427e-06,
+      "loss": 0.4361,
+      "step": 66300
+    },
+    {
+      "epoch": 16.536159600997507,
+      "grad_norm": 7.750860214233398,
+      "learning_rate": 3.470822942643392e-06,
+      "loss": 0.3055,
+      "step": 66310
+    },
+    {
+      "epoch": 16.538653366583542,
+      "grad_norm": 7.906196594238281,
+      "learning_rate": 3.468329177057357e-06,
+      "loss": 0.282,
+      "step": 66320
+    },
+    {
+      "epoch": 16.541147132169577,
+      "grad_norm": 8.782937049865723,
+      "learning_rate": 3.4658354114713223e-06,
+      "loss": 0.3205,
+      "step": 66330
+    },
+    {
+      "epoch": 16.543640897755612,
+      "grad_norm": 10.13814926147461,
+      "learning_rate": 3.4633416458852868e-06,
+      "loss": 0.3365,
+      "step": 66340
+    },
+    {
+      "epoch": 16.546134663341647,
+      "grad_norm": 9.947074890136719,
+      "learning_rate": 3.4608478802992525e-06,
+      "loss": 0.3592,
+      "step": 66350
+    },
+    {
+      "epoch": 16.54862842892768,
+      "grad_norm": 12.302644729614258,
+      "learning_rate": 3.458354114713217e-06,
+      "loss": 0.3738,
+      "step": 66360
+    },
+    {
+      "epoch": 16.551122194513717,
+      "grad_norm": 8.22531509399414,
+      "learning_rate": 3.4558603491271823e-06,
+      "loss": 0.3024,
+      "step": 66370
+    },
+    {
+      "epoch": 16.55361596009975,
+      "grad_norm": 8.880837440490723,
+      "learning_rate": 3.453366583541147e-06,
+      "loss": 0.2816,
+      "step": 66380
+    },
+    {
+      "epoch": 16.556109725685786,
+      "grad_norm": 7.537718296051025,
+      "learning_rate": 3.4508728179551125e-06,
+      "loss": 0.3165,
+      "step": 66390
+    },
+    {
+      "epoch": 16.55860349127182,
+      "grad_norm": 11.703903198242188,
+      "learning_rate": 3.4483790523690774e-06,
+      "loss": 0.4277,
+      "step": 66400
+    },
+    {
+      "epoch": 16.561097256857856,
+      "grad_norm": 13.591885566711426,
+      "learning_rate": 3.4458852867830427e-06,
+      "loss": 0.2994,
+      "step": 66410
+    },
+    {
+      "epoch": 16.56359102244389,
+      "grad_norm": 8.424422264099121,
+      "learning_rate": 3.4433915211970076e-06,
+      "loss": 0.3531,
+      "step": 66420
+    },
+    {
+      "epoch": 16.566084788029926,
+      "grad_norm": 5.670681953430176,
+      "learning_rate": 3.440897755610973e-06,
+      "loss": 0.3566,
+      "step": 66430
+    },
+    {
+      "epoch": 16.56857855361596,
+      "grad_norm": 13.465998649597168,
+      "learning_rate": 3.4384039900249378e-06,
+      "loss": 0.297,
+      "step": 66440
+    },
+    {
+      "epoch": 16.571072319201996,
+      "grad_norm": 6.211828708648682,
+      "learning_rate": 3.435910224438903e-06,
+      "loss": 0.2967,
+      "step": 66450
+    },
+    {
+      "epoch": 16.57356608478803,
+      "grad_norm": 8.583992958068848,
+      "learning_rate": 3.433416458852868e-06,
+      "loss": 0.2802,
+      "step": 66460
+    },
+    {
+      "epoch": 16.576059850374065,
+      "grad_norm": 6.500486373901367,
+      "learning_rate": 3.4309226932668333e-06,
+      "loss": 0.3138,
+      "step": 66470
+    },
+    {
+      "epoch": 16.5785536159601,
+      "grad_norm": 9.81982707977295,
+      "learning_rate": 3.428428927680798e-06,
+      "loss": 0.3349,
+      "step": 66480
+    },
+    {
+      "epoch": 16.581047381546135,
+      "grad_norm": 9.09636402130127,
+      "learning_rate": 3.4259351620947635e-06,
+      "loss": 0.2904,
+      "step": 66490
+    },
+    {
+      "epoch": 16.58354114713217,
+      "grad_norm": 7.976047039031982,
+      "learning_rate": 3.4234413965087283e-06,
+      "loss": 0.3574,
+      "step": 66500
+    },
+    {
+      "epoch": 16.586034912718205,
+      "grad_norm": 6.352416038513184,
+      "learning_rate": 3.4209476309226937e-06,
+      "loss": 0.3534,
+      "step": 66510
+    },
+    {
+      "epoch": 16.58852867830424,
+      "grad_norm": 7.957899570465088,
+      "learning_rate": 3.418453865336659e-06,
+      "loss": 0.417,
+      "step": 66520
+    },
+    {
+      "epoch": 16.591022443890274,
+      "grad_norm": 7.401804447174072,
+      "learning_rate": 3.415960099750624e-06,
+      "loss": 0.3865,
+      "step": 66530
+    },
+    {
+      "epoch": 16.59351620947631,
+      "grad_norm": 10.400482177734375,
+      "learning_rate": 3.413466334164589e-06,
+      "loss": 0.3071,
+      "step": 66540
+    },
+    {
+      "epoch": 16.596009975062344,
+      "grad_norm": 8.039289474487305,
+      "learning_rate": 3.410972568578554e-06,
+      "loss": 0.3396,
+      "step": 66550
+    },
+    {
+      "epoch": 16.59850374064838,
+      "grad_norm": 6.777886390686035,
+      "learning_rate": 3.4084788029925194e-06,
+      "loss": 0.2646,
+      "step": 66560
+    },
+    {
+      "epoch": 16.600997506234414,
+      "grad_norm": 9.714869499206543,
+      "learning_rate": 3.405985037406484e-06,
+      "loss": 0.3047,
+      "step": 66570
+    },
+    {
+      "epoch": 16.60349127182045,
+      "grad_norm": 8.811583518981934,
+      "learning_rate": 3.4034912718204496e-06,
+      "loss": 0.2972,
+      "step": 66580
+    },
+    {
+      "epoch": 16.605985037406484,
+      "grad_norm": 8.421089172363281,
+      "learning_rate": 3.400997506234414e-06,
+      "loss": 0.3288,
+      "step": 66590
+    },
+    {
+      "epoch": 16.60847880299252,
+      "grad_norm": 7.008131980895996,
+      "learning_rate": 3.3985037406483793e-06,
+      "loss": 0.3063,
+      "step": 66600
+    },
+    {
+      "epoch": 16.610972568578553,
+      "grad_norm": 7.8703789710998535,
+      "learning_rate": 3.3960099750623442e-06,
+      "loss": 0.2479,
+      "step": 66610
+    },
+    {
+      "epoch": 16.61346633416459,
+      "grad_norm": 8.42544937133789,
+      "learning_rate": 3.3935162094763095e-06,
+      "loss": 0.3384,
+      "step": 66620
+    },
+    {
+      "epoch": 16.615960099750623,
+      "grad_norm": 6.888559341430664,
+      "learning_rate": 3.3910224438902744e-06,
+      "loss": 0.3679,
+      "step": 66630
+    },
+    {
+      "epoch": 16.618453865336658,
+      "grad_norm": 8.545351028442383,
+      "learning_rate": 3.3885286783042397e-06,
+      "loss": 0.3226,
+      "step": 66640
+    },
+    {
+      "epoch": 16.620947630922693,
+      "grad_norm": 7.288325309753418,
+      "learning_rate": 3.3860349127182046e-06,
+      "loss": 0.3422,
+      "step": 66650
+    },
+    {
+      "epoch": 16.623441396508728,
+      "grad_norm": 10.019088745117188,
+      "learning_rate": 3.38354114713217e-06,
+      "loss": 0.3509,
+      "step": 66660
+    },
+    {
+      "epoch": 16.625935162094763,
+      "grad_norm": 6.888891220092773,
+      "learning_rate": 3.381047381546135e-06,
+      "loss": 0.3493,
+      "step": 66670
+    },
+    {
+      "epoch": 16.628428927680797,
+      "grad_norm": 11.555747032165527,
+      "learning_rate": 3.3785536159601e-06,
+      "loss": 0.3701,
+      "step": 66680
+    },
+    {
+      "epoch": 16.630922693266832,
+      "grad_norm": 8.428287506103516,
+      "learning_rate": 3.376059850374065e-06,
+      "loss": 0.3779,
+      "step": 66690
+    },
+    {
+      "epoch": 16.633416458852867,
+      "grad_norm": 10.036828994750977,
+      "learning_rate": 3.3735660847880303e-06,
+      "loss": 0.3879,
+      "step": 66700
+    },
+    {
+      "epoch": 16.635910224438902,
+      "grad_norm": 7.905212879180908,
+      "learning_rate": 3.371072319201995e-06,
+      "loss": 0.2564,
+      "step": 66710
+    },
+    {
+      "epoch": 16.638403990024937,
+      "grad_norm": 7.539322853088379,
+      "learning_rate": 3.3685785536159605e-06,
+      "loss": 0.3989,
+      "step": 66720
+    },
+    {
+      "epoch": 16.640897755610972,
+      "grad_norm": 8.233596801757812,
+      "learning_rate": 3.3660847880299254e-06,
+      "loss": 0.3014,
+      "step": 66730
+    },
+    {
+      "epoch": 16.643391521197007,
+      "grad_norm": 7.139582633972168,
+      "learning_rate": 3.3635910224438907e-06,
+      "loss": 0.2907,
+      "step": 66740
+    },
+    {
+      "epoch": 16.64588528678304,
+      "grad_norm": 9.440067291259766,
+      "learning_rate": 3.361097256857855e-06,
+      "loss": 0.3126,
+      "step": 66750
+    },
+    {
+      "epoch": 16.648379052369076,
+      "grad_norm": 10.730769157409668,
+      "learning_rate": 3.358603491271821e-06,
+      "loss": 0.3447,
+      "step": 66760
+    },
+    {
+      "epoch": 16.65087281795511,
+      "grad_norm": 12.145304679870605,
+      "learning_rate": 3.356109725685786e-06,
+      "loss": 0.318,
+      "step": 66770
+    },
+    {
+      "epoch": 16.653366583541146,
+      "grad_norm": 13.246317863464355,
+      "learning_rate": 3.3536159600997507e-06,
+      "loss": 0.3085,
+      "step": 66780
+    },
+    {
+      "epoch": 16.65586034912718,
+      "grad_norm": 6.979218482971191,
+      "learning_rate": 3.3511221945137164e-06,
+      "loss": 0.328,
+      "step": 66790
+    },
+    {
+      "epoch": 16.658354114713216,
+      "grad_norm": 7.944563388824463,
+      "learning_rate": 3.348628428927681e-06,
+      "loss": 0.2856,
+      "step": 66800
+    },
+    {
+      "epoch": 16.66084788029925,
+      "grad_norm": 8.603992462158203,
+      "learning_rate": 3.346134663341646e-06,
+      "loss": 0.3337,
+      "step": 66810
+    },
+    {
+      "epoch": 16.663341645885286,
+      "grad_norm": 9.603078842163086,
+      "learning_rate": 3.343640897755611e-06,
+      "loss": 0.3376,
+      "step": 66820
+    },
+    {
+      "epoch": 16.66583541147132,
+      "grad_norm": 5.392066478729248,
+      "learning_rate": 3.3411471321695764e-06,
+      "loss": 0.3134,
+      "step": 66830
+    },
+    {
+      "epoch": 16.668329177057355,
+      "grad_norm": 6.164698600769043,
+      "learning_rate": 3.3386533665835413e-06,
+      "loss": 0.323,
+      "step": 66840
+    },
+    {
+      "epoch": 16.67082294264339,
+      "grad_norm": 8.777530670166016,
+      "learning_rate": 3.3361596009975066e-06,
+      "loss": 0.3715,
+      "step": 66850
+    },
+    {
+      "epoch": 16.673316708229425,
+      "grad_norm": 9.735917091369629,
+      "learning_rate": 3.3336658354114715e-06,
+      "loss": 0.3017,
+      "step": 66860
+    },
+    {
+      "epoch": 16.67581047381546,
+      "grad_norm": 5.572236061096191,
+      "learning_rate": 3.3311720698254368e-06,
+      "loss": 0.324,
+      "step": 66870
+    },
+    {
+      "epoch": 16.678304239401495,
+      "grad_norm": 8.303897857666016,
+      "learning_rate": 3.3286783042394016e-06,
+      "loss": 0.3146,
+      "step": 66880
+    },
+    {
+      "epoch": 16.68079800498753,
+      "grad_norm": 10.715437889099121,
+      "learning_rate": 3.326184538653367e-06,
+      "loss": 0.3026,
+      "step": 66890
+    },
+    {
+      "epoch": 16.683291770573565,
+      "grad_norm": 8.109511375427246,
+      "learning_rate": 3.323690773067332e-06,
+      "loss": 0.3878,
+      "step": 66900
+    },
+    {
+      "epoch": 16.6857855361596,
+      "grad_norm": 5.563307762145996,
+      "learning_rate": 3.321197007481297e-06,
+      "loss": 0.2698,
+      "step": 66910
+    },
+    {
+      "epoch": 16.688279301745634,
+      "grad_norm": 13.141777038574219,
+      "learning_rate": 3.318703241895262e-06,
+      "loss": 0.3028,
+      "step": 66920
+    },
+    {
+      "epoch": 16.69077306733167,
+      "grad_norm": 9.709403991699219,
+      "learning_rate": 3.3162094763092273e-06,
+      "loss": 0.3567,
+      "step": 66930
+    },
+    {
+      "epoch": 16.693266832917704,
+      "grad_norm": 9.365985870361328,
+      "learning_rate": 3.3137157107231922e-06,
+      "loss": 0.2793,
+      "step": 66940
+    },
+    {
+      "epoch": 16.69576059850374,
+      "grad_norm": 10.170108795166016,
+      "learning_rate": 3.3112219451371575e-06,
+      "loss": 0.2982,
+      "step": 66950
+    },
+    {
+      "epoch": 16.698254364089777,
+      "grad_norm": 5.353804588317871,
+      "learning_rate": 3.3087281795511224e-06,
+      "loss": 0.3412,
+      "step": 66960
+    },
+    {
+      "epoch": 16.70074812967581,
+      "grad_norm": 13.283953666687012,
+      "learning_rate": 3.3062344139650877e-06,
+      "loss": 0.2865,
+      "step": 66970
+    },
+    {
+      "epoch": 16.703241895261847,
+      "grad_norm": 9.242694854736328,
+      "learning_rate": 3.303740648379052e-06,
+      "loss": 0.2743,
+      "step": 66980
+    },
+    {
+      "epoch": 16.705735660847882,
+      "grad_norm": 9.898309707641602,
+      "learning_rate": 3.301246882793018e-06,
+      "loss": 0.3607,
+      "step": 66990
+    },
+    {
+      "epoch": 16.708229426433917,
+      "grad_norm": 7.244779109954834,
+      "learning_rate": 3.2987531172069824e-06,
+      "loss": 0.2872,
+      "step": 67000
+    },
+    {
+      "epoch": 16.71072319201995,
+      "grad_norm": 7.123824119567871,
+      "learning_rate": 3.2962593516209477e-06,
+      "loss": 0.2879,
+      "step": 67010
+    },
+    {
+      "epoch": 16.713216957605987,
+      "grad_norm": 11.419690132141113,
+      "learning_rate": 3.2937655860349134e-06,
+      "loss": 0.3193,
+      "step": 67020
+    },
+    {
+      "epoch": 16.71571072319202,
+      "grad_norm": 12.919515609741211,
+      "learning_rate": 3.291271820448878e-06,
+      "loss": 0.3836,
+      "step": 67030
+    },
+    {
+      "epoch": 16.718204488778056,
+      "grad_norm": 8.977980613708496,
+      "learning_rate": 3.2887780548628432e-06,
+      "loss": 0.303,
+      "step": 67040
+    },
+    {
+      "epoch": 16.72069825436409,
+      "grad_norm": 8.428853988647461,
+      "learning_rate": 3.286284289276808e-06,
+      "loss": 0.3334,
+      "step": 67050
+    },
+    {
+      "epoch": 16.723192019950126,
+      "grad_norm": 9.528043746948242,
+      "learning_rate": 3.2837905236907734e-06,
+      "loss": 0.3773,
+      "step": 67060
+    },
+    {
+      "epoch": 16.72568578553616,
+      "grad_norm": 8.615249633789062,
+      "learning_rate": 3.2812967581047383e-06,
+      "loss": 0.2869,
+      "step": 67070
+    },
+    {
+      "epoch": 16.728179551122196,
+      "grad_norm": 11.474807739257812,
+      "learning_rate": 3.2788029925187036e-06,
+      "loss": 0.3592,
+      "step": 67080
+    },
+    {
+      "epoch": 16.73067331670823,
+      "grad_norm": 7.433091640472412,
+      "learning_rate": 3.2763092269326685e-06,
+      "loss": 0.294,
+      "step": 67090
+    },
+    {
+      "epoch": 16.733167082294266,
+      "grad_norm": 9.934019088745117,
+      "learning_rate": 3.273815461346634e-06,
+      "loss": 0.3228,
+      "step": 67100
+    },
+    {
+      "epoch": 16.7356608478803,
+      "grad_norm": 8.578768730163574,
+      "learning_rate": 3.2713216957605987e-06,
+      "loss": 0.279,
+      "step": 67110
+    },
+    {
+      "epoch": 16.738154613466335,
+      "grad_norm": 7.551397800445557,
+      "learning_rate": 3.268827930174564e-06,
+      "loss": 0.3451,
+      "step": 67120
+    },
+    {
+      "epoch": 16.74064837905237,
+      "grad_norm": 7.463733196258545,
+      "learning_rate": 3.266334164588529e-06,
+      "loss": 0.3792,
+      "step": 67130
+    },
+    {
+      "epoch": 16.743142144638405,
+      "grad_norm": 9.001909255981445,
+      "learning_rate": 3.263840399002494e-06,
+      "loss": 0.3244,
+      "step": 67140
+    },
+    {
+      "epoch": 16.74563591022444,
+      "grad_norm": 10.928662300109863,
+      "learning_rate": 3.261346633416459e-06,
+      "loss": 0.3328,
+      "step": 67150
+    },
+    {
+      "epoch": 16.748129675810475,
+      "grad_norm": 5.639915466308594,
+      "learning_rate": 3.2588528678304244e-06,
+      "loss": 0.3326,
+      "step": 67160
+    },
+    {
+      "epoch": 16.75062344139651,
+      "grad_norm": 10.53235912322998,
+      "learning_rate": 3.2563591022443893e-06,
+      "loss": 0.361,
+      "step": 67170
+    },
+    {
+      "epoch": 16.753117206982544,
+      "grad_norm": 5.4268951416015625,
+      "learning_rate": 3.2538653366583546e-06,
+      "loss": 0.3213,
+      "step": 67180
+    },
+    {
+      "epoch": 16.75561097256858,
+      "grad_norm": 10.269506454467773,
+      "learning_rate": 3.2513715710723195e-06,
+      "loss": 0.3346,
+      "step": 67190
+    },
+    {
+      "epoch": 16.758104738154614,
+      "grad_norm": 21.99469757080078,
+      "learning_rate": 3.2488778054862848e-06,
+      "loss": 0.3419,
+      "step": 67200
+    },
+    {
+      "epoch": 16.76059850374065,
+      "grad_norm": 6.769282341003418,
+      "learning_rate": 3.2463840399002492e-06,
+      "loss": 0.321,
+      "step": 67210
+    },
+    {
+      "epoch": 16.763092269326684,
+      "grad_norm": 5.740752220153809,
+      "learning_rate": 3.243890274314215e-06,
+      "loss": 0.3051,
+      "step": 67220
+    },
+    {
+      "epoch": 16.76558603491272,
+      "grad_norm": 9.068950653076172,
+      "learning_rate": 3.2413965087281794e-06,
+      "loss": 0.4055,
+      "step": 67230
+    },
+    {
+      "epoch": 16.768079800498754,
+      "grad_norm": 11.04249095916748,
+      "learning_rate": 3.2389027431421448e-06,
+      "loss": 0.336,
+      "step": 67240
+    },
+    {
+      "epoch": 16.77057356608479,
+      "grad_norm": 9.966495513916016,
+      "learning_rate": 3.2364089775561096e-06,
+      "loss": 0.3491,
+      "step": 67250
+    },
+    {
+      "epoch": 16.773067331670823,
+      "grad_norm": 11.092144012451172,
+      "learning_rate": 3.233915211970075e-06,
+      "loss": 0.2806,
+      "step": 67260
+    },
+    {
+      "epoch": 16.77556109725686,
+      "grad_norm": 8.785165786743164,
+      "learning_rate": 3.2314214463840403e-06,
+      "loss": 0.3081,
+      "step": 67270
+    },
+    {
+      "epoch": 16.778054862842893,
+      "grad_norm": 9.333272933959961,
+      "learning_rate": 3.228927680798005e-06,
+      "loss": 0.364,
+      "step": 67280
+    },
+    {
+      "epoch": 16.780548628428928,
+      "grad_norm": 8.067471504211426,
+      "learning_rate": 3.2264339152119705e-06,
+      "loss": 0.3608,
+      "step": 67290
+    },
+    {
+      "epoch": 16.783042394014963,
+      "grad_norm": 4.908388614654541,
+      "learning_rate": 3.2239401496259353e-06,
+      "loss": 0.3417,
+      "step": 67300
+    },
+    {
+      "epoch": 16.785536159600998,
+      "grad_norm": 10.774919509887695,
+      "learning_rate": 3.2214463840399006e-06,
+      "loss": 0.3317,
+      "step": 67310
+    },
+    {
+      "epoch": 16.788029925187033,
+      "grad_norm": 19.7656192779541,
+      "learning_rate": 3.2189526184538655e-06,
+      "loss": 0.3345,
+      "step": 67320
+    },
+    {
+      "epoch": 16.790523690773068,
+      "grad_norm": 8.48336124420166,
+      "learning_rate": 3.216458852867831e-06,
+      "loss": 0.3249,
+      "step": 67330
+    },
+    {
+      "epoch": 16.793017456359102,
+      "grad_norm": 7.753775119781494,
+      "learning_rate": 3.2139650872817957e-06,
+      "loss": 0.295,
+      "step": 67340
+    },
+    {
+      "epoch": 16.795511221945137,
+      "grad_norm": 7.943778038024902,
+      "learning_rate": 3.211471321695761e-06,
+      "loss": 0.356,
+      "step": 67350
+    },
+    {
+      "epoch": 16.798004987531172,
+      "grad_norm": 8.604585647583008,
+      "learning_rate": 3.208977556109726e-06,
+      "loss": 0.3197,
+      "step": 67360
+    },
+    {
+      "epoch": 16.800498753117207,
+      "grad_norm": 7.526785373687744,
+      "learning_rate": 3.2064837905236912e-06,
+      "loss": 0.3185,
+      "step": 67370
+    },
+    {
+      "epoch": 16.802992518703242,
+      "grad_norm": 9.398643493652344,
+      "learning_rate": 3.203990024937656e-06,
+      "loss": 0.3266,
+      "step": 67380
+    },
+    {
+      "epoch": 16.805486284289277,
+      "grad_norm": 9.048074722290039,
+      "learning_rate": 3.2014962593516214e-06,
+      "loss": 0.3011,
+      "step": 67390
+    },
+    {
+      "epoch": 16.80798004987531,
+      "grad_norm": 8.05639362335205,
+      "learning_rate": 3.1990024937655863e-06,
+      "loss": 0.3413,
+      "step": 67400
+    },
+    {
+      "epoch": 16.810473815461346,
+      "grad_norm": 8.279003143310547,
+      "learning_rate": 3.1965087281795516e-06,
+      "loss": 0.3302,
+      "step": 67410
+    },
+    {
+      "epoch": 16.81296758104738,
+      "grad_norm": 10.437573432922363,
+      "learning_rate": 3.194014962593516e-06,
+      "loss": 0.3326,
+      "step": 67420
+    },
+    {
+      "epoch": 16.815461346633416,
+      "grad_norm": 10.088639259338379,
+      "learning_rate": 3.191521197007482e-06,
+      "loss": 0.2668,
+      "step": 67430
+    },
+    {
+      "epoch": 16.81795511221945,
+      "grad_norm": 11.739412307739258,
+      "learning_rate": 3.1890274314214463e-06,
+      "loss": 0.2719,
+      "step": 67440
+    },
+    {
+      "epoch": 16.820448877805486,
+      "grad_norm": 7.2401604652404785,
+      "learning_rate": 3.1865336658354116e-06,
+      "loss": 0.3168,
+      "step": 67450
+    },
+    {
+      "epoch": 16.82294264339152,
+      "grad_norm": 9.88704776763916,
+      "learning_rate": 3.1840399002493765e-06,
+      "loss": 0.4028,
+      "step": 67460
+    },
+    {
+      "epoch": 16.825436408977556,
+      "grad_norm": 10.055147171020508,
+      "learning_rate": 3.181546134663342e-06,
+      "loss": 0.3385,
+      "step": 67470
+    },
+    {
+      "epoch": 16.82793017456359,
+      "grad_norm": 11.214886665344238,
+      "learning_rate": 3.1790523690773067e-06,
+      "loss": 0.3381,
+      "step": 67480
+    },
+    {
+      "epoch": 16.830423940149625,
+      "grad_norm": 9.053804397583008,
+      "learning_rate": 3.176558603491272e-06,
+      "loss": 0.3107,
+      "step": 67490
+    },
+    {
+      "epoch": 16.83291770573566,
+      "grad_norm": 9.388381004333496,
+      "learning_rate": 3.174064837905237e-06,
+      "loss": 0.3519,
+      "step": 67500
+    },
+    {
+      "epoch": 16.835411471321695,
+      "grad_norm": 11.288336753845215,
+      "learning_rate": 3.171571072319202e-06,
+      "loss": 0.3049,
+      "step": 67510
+    },
+    {
+      "epoch": 16.83790523690773,
+      "grad_norm": 11.12149429321289,
+      "learning_rate": 3.1690773067331675e-06,
+      "loss": 0.3766,
+      "step": 67520
+    },
+    {
+      "epoch": 16.840399002493765,
+      "grad_norm": 6.969629287719727,
+      "learning_rate": 3.1665835411471324e-06,
+      "loss": 0.3209,
+      "step": 67530
+    },
+    {
+      "epoch": 16.8428927680798,
+      "grad_norm": 9.759892463684082,
+      "learning_rate": 3.1640897755610977e-06,
+      "loss": 0.2822,
+      "step": 67540
+    },
+    {
+      "epoch": 16.845386533665835,
+      "grad_norm": 9.30257797241211,
+      "learning_rate": 3.1615960099750626e-06,
+      "loss": 0.3563,
+      "step": 67550
+    },
+    {
+      "epoch": 16.84788029925187,
+      "grad_norm": 7.607780933380127,
+      "learning_rate": 3.159102244389028e-06,
+      "loss": 0.352,
+      "step": 67560
+    },
+    {
+      "epoch": 16.850374064837904,
+      "grad_norm": 6.663292407989502,
+      "learning_rate": 3.1566084788029928e-06,
+      "loss": 0.3008,
+      "step": 67570
+    },
+    {
+      "epoch": 16.85286783042394,
+      "grad_norm": 9.025524139404297,
+      "learning_rate": 3.154114713216958e-06,
+      "loss": 0.3307,
+      "step": 67580
+    },
+    {
+      "epoch": 16.855361596009974,
+      "grad_norm": 7.929129123687744,
+      "learning_rate": 3.151620947630923e-06,
+      "loss": 0.3064,
+      "step": 67590
+    },
+    {
+      "epoch": 16.85785536159601,
+      "grad_norm": 6.258866786956787,
+      "learning_rate": 3.1491271820448883e-06,
+      "loss": 0.2586,
+      "step": 67600
+    },
+    {
+      "epoch": 16.860349127182044,
+      "grad_norm": 5.975445747375488,
+      "learning_rate": 3.146633416458853e-06,
+      "loss": 0.2883,
+      "step": 67610
+    },
+    {
+      "epoch": 16.86284289276808,
+      "grad_norm": 6.100348472595215,
+      "learning_rate": 3.1441396508728185e-06,
+      "loss": 0.3816,
+      "step": 67620
+    },
+    {
+      "epoch": 16.865336658354114,
+      "grad_norm": 8.461503028869629,
+      "learning_rate": 3.1416458852867834e-06,
+      "loss": 0.3295,
+      "step": 67630
+    },
+    {
+      "epoch": 16.86783042394015,
+      "grad_norm": 11.094927787780762,
+      "learning_rate": 3.1391521197007487e-06,
+      "loss": 0.3346,
+      "step": 67640
+    },
+    {
+      "epoch": 16.870324189526183,
+      "grad_norm": 8.778679847717285,
+      "learning_rate": 3.136658354114713e-06,
+      "loss": 0.3169,
+      "step": 67650
+    },
+    {
+      "epoch": 16.872817955112218,
+      "grad_norm": 8.165081024169922,
+      "learning_rate": 3.134164588528679e-06,
+      "loss": 0.3515,
+      "step": 67660
+    },
+    {
+      "epoch": 16.875311720698253,
+      "grad_norm": 8.707517623901367,
+      "learning_rate": 3.1316708229426433e-06,
+      "loss": 0.3255,
+      "step": 67670
+    },
+    {
+      "epoch": 16.877805486284288,
+      "grad_norm": 12.160292625427246,
+      "learning_rate": 3.1291770573566086e-06,
+      "loss": 0.3691,
+      "step": 67680
+    },
+    {
+      "epoch": 16.880299251870323,
+      "grad_norm": 7.535975933074951,
+      "learning_rate": 3.1266832917705735e-06,
+      "loss": 0.3717,
+      "step": 67690
+    },
+    {
+      "epoch": 16.882793017456358,
+      "grad_norm": 9.464982032775879,
+      "learning_rate": 3.124189526184539e-06,
+      "loss": 0.3114,
+      "step": 67700
+    },
+    {
+      "epoch": 16.885286783042392,
+      "grad_norm": 6.380134582519531,
+      "learning_rate": 3.1216957605985037e-06,
+      "loss": 0.2951,
+      "step": 67710
+    },
+    {
+      "epoch": 16.887780548628427,
+      "grad_norm": 11.643170356750488,
+      "learning_rate": 3.119201995012469e-06,
+      "loss": 0.3128,
+      "step": 67720
+    },
+    {
+      "epoch": 16.890274314214462,
+      "grad_norm": 12.784978866577148,
+      "learning_rate": 3.116708229426434e-06,
+      "loss": 0.3273,
+      "step": 67730
+    },
+    {
+      "epoch": 16.892768079800497,
+      "grad_norm": 9.829355239868164,
+      "learning_rate": 3.1142144638403992e-06,
+      "loss": 0.3061,
+      "step": 67740
+    },
+    {
+      "epoch": 16.895261845386532,
+      "grad_norm": 10.183358192443848,
+      "learning_rate": 3.111720698254364e-06,
+      "loss": 0.3194,
+      "step": 67750
+    },
+    {
+      "epoch": 16.897755610972567,
+      "grad_norm": 9.071417808532715,
+      "learning_rate": 3.1092269326683294e-06,
+      "loss": 0.2804,
+      "step": 67760
+    },
+    {
+      "epoch": 16.900249376558605,
+      "grad_norm": 10.647289276123047,
+      "learning_rate": 3.1067331670822947e-06,
+      "loss": 0.3488,
+      "step": 67770
+    },
+    {
+      "epoch": 16.902743142144637,
+      "grad_norm": 9.39749813079834,
+      "learning_rate": 3.1042394014962596e-06,
+      "loss": 0.335,
+      "step": 67780
+    },
+    {
+      "epoch": 16.905236907730675,
+      "grad_norm": 9.534618377685547,
+      "learning_rate": 3.101745635910225e-06,
+      "loss": 0.2999,
+      "step": 67790
+    },
+    {
+      "epoch": 16.90773067331671,
+      "grad_norm": 13.632329940795898,
+      "learning_rate": 3.09925187032419e-06,
+      "loss": 0.3342,
+      "step": 67800
+    },
+    {
+      "epoch": 16.910224438902745,
+      "grad_norm": 7.592889308929443,
+      "learning_rate": 3.096758104738155e-06,
+      "loss": 0.2828,
+      "step": 67810
+    },
+    {
+      "epoch": 16.91271820448878,
+      "grad_norm": 6.196469783782959,
+      "learning_rate": 3.09426433915212e-06,
+      "loss": 0.3377,
+      "step": 67820
+    },
+    {
+      "epoch": 16.915211970074814,
+      "grad_norm": 9.526688575744629,
+      "learning_rate": 3.0917705735660853e-06,
+      "loss": 0.2684,
+      "step": 67830
+    },
+    {
+      "epoch": 16.91770573566085,
+      "grad_norm": 7.148454189300537,
+      "learning_rate": 3.08927680798005e-06,
+      "loss": 0.3152,
+      "step": 67840
+    },
+    {
+      "epoch": 16.920199501246884,
+      "grad_norm": 7.666889190673828,
+      "learning_rate": 3.0867830423940155e-06,
+      "loss": 0.3181,
+      "step": 67850
+    },
+    {
+      "epoch": 16.92269326683292,
+      "grad_norm": 6.633835792541504,
+      "learning_rate": 3.08428927680798e-06,
+      "loss": 0.3662,
+      "step": 67860
+    },
+    {
+      "epoch": 16.925187032418954,
+      "grad_norm": 9.805349349975586,
+      "learning_rate": 3.0817955112219457e-06,
+      "loss": 0.3224,
+      "step": 67870
+    },
+    {
+      "epoch": 16.92768079800499,
+      "grad_norm": 25.773767471313477,
+      "learning_rate": 3.07930174563591e-06,
+      "loss": 0.3182,
+      "step": 67880
+    },
+    {
+      "epoch": 16.930174563591024,
+      "grad_norm": 5.862217426300049,
+      "learning_rate": 3.0768079800498755e-06,
+      "loss": 0.362,
+      "step": 67890
+    },
+    {
+      "epoch": 16.93266832917706,
+      "grad_norm": 6.981509208679199,
+      "learning_rate": 3.0743142144638404e-06,
+      "loss": 0.2904,
+      "step": 67900
+    },
+    {
+      "epoch": 16.935162094763093,
+      "grad_norm": 7.372140407562256,
+      "learning_rate": 3.0718204488778057e-06,
+      "loss": 0.3158,
+      "step": 67910
+    },
+    {
+      "epoch": 16.93765586034913,
+      "grad_norm": 6.7870612144470215,
+      "learning_rate": 3.0693266832917706e-06,
+      "loss": 0.323,
+      "step": 67920
+    },
+    {
+      "epoch": 16.940149625935163,
+      "grad_norm": 11.441544532775879,
+      "learning_rate": 3.066832917705736e-06,
+      "loss": 0.3994,
+      "step": 67930
+    },
+    {
+      "epoch": 16.942643391521198,
+      "grad_norm": 7.740766525268555,
+      "learning_rate": 3.0643391521197008e-06,
+      "loss": 0.4086,
+      "step": 67940
+    },
+    {
+      "epoch": 16.945137157107233,
+      "grad_norm": 10.935362815856934,
+      "learning_rate": 3.061845386533666e-06,
+      "loss": 0.32,
+      "step": 67950
+    },
+    {
+      "epoch": 16.947630922693268,
+      "grad_norm": 9.322646141052246,
+      "learning_rate": 3.059351620947631e-06,
+      "loss": 0.3604,
+      "step": 67960
+    },
+    {
+      "epoch": 16.950124688279303,
+      "grad_norm": 8.889856338500977,
+      "learning_rate": 3.0568578553615963e-06,
+      "loss": 0.3302,
+      "step": 67970
+    },
+    {
+      "epoch": 16.952618453865338,
+      "grad_norm": 12.106937408447266,
+      "learning_rate": 3.054364089775561e-06,
+      "loss": 0.2517,
+      "step": 67980
+    },
+    {
+      "epoch": 16.955112219451372,
+      "grad_norm": 9.038848876953125,
+      "learning_rate": 3.0518703241895265e-06,
+      "loss": 0.3813,
+      "step": 67990
+    },
+    {
+      "epoch": 16.957605985037407,
+      "grad_norm": 10.314130783081055,
+      "learning_rate": 3.0493765586034918e-06,
+      "loss": 0.3414,
+      "step": 68000
+    },
+    {
+      "epoch": 16.960099750623442,
+      "grad_norm": 6.741665840148926,
+      "learning_rate": 3.0468827930174567e-06,
+      "loss": 0.3446,
+      "step": 68010
+    },
+    {
+      "epoch": 16.962593516209477,
+      "grad_norm": 7.708677768707275,
+      "learning_rate": 3.044389027431422e-06,
+      "loss": 0.2833,
+      "step": 68020
+    },
+    {
+      "epoch": 16.965087281795512,
+      "grad_norm": 7.930616855621338,
+      "learning_rate": 3.041895261845387e-06,
+      "loss": 0.2681,
+      "step": 68030
+    },
+    {
+      "epoch": 16.967581047381547,
+      "grad_norm": 6.562017440795898,
+      "learning_rate": 3.039401496259352e-06,
+      "loss": 0.3424,
+      "step": 68040
+    },
+    {
+      "epoch": 16.97007481296758,
+      "grad_norm": 9.841915130615234,
+      "learning_rate": 3.036907730673317e-06,
+      "loss": 0.3586,
+      "step": 68050
+    },
+    {
+      "epoch": 16.972568578553616,
+      "grad_norm": 13.928132057189941,
+      "learning_rate": 3.0344139650872824e-06,
+      "loss": 0.3287,
+      "step": 68060
+    },
+    {
+      "epoch": 16.97506234413965,
+      "grad_norm": 6.605895042419434,
+      "learning_rate": 3.0319201995012472e-06,
+      "loss": 0.3126,
+      "step": 68070
+    },
+    {
+      "epoch": 16.977556109725686,
+      "grad_norm": 8.77118968963623,
+      "learning_rate": 3.0294264339152126e-06,
+      "loss": 0.345,
+      "step": 68080
+    },
+    {
+      "epoch": 16.98004987531172,
+      "grad_norm": 10.565832138061523,
+      "learning_rate": 3.026932668329177e-06,
+      "loss": 0.3402,
+      "step": 68090
+    },
+    {
+      "epoch": 16.982543640897756,
+      "grad_norm": 7.508286952972412,
+      "learning_rate": 3.0244389027431428e-06,
+      "loss": 0.3443,
+      "step": 68100
+    },
+    {
+      "epoch": 16.98503740648379,
+      "grad_norm": 7.770997524261475,
+      "learning_rate": 3.0219451371571072e-06,
+      "loss": 0.3283,
+      "step": 68110
+    },
+    {
+      "epoch": 16.987531172069826,
+      "grad_norm": 10.09821891784668,
+      "learning_rate": 3.0194513715710725e-06,
+      "loss": 0.3236,
+      "step": 68120
+    },
+    {
+      "epoch": 16.99002493765586,
+      "grad_norm": 6.971126079559326,
+      "learning_rate": 3.0169576059850374e-06,
+      "loss": 0.3546,
+      "step": 68130
+    },
+    {
+      "epoch": 16.992518703241895,
+      "grad_norm": 8.345663070678711,
+      "learning_rate": 3.0144638403990027e-06,
+      "loss": 0.3018,
+      "step": 68140
+    },
+    {
+      "epoch": 16.99501246882793,
+      "grad_norm": 8.50364875793457,
+      "learning_rate": 3.0119700748129676e-06,
+      "loss": 0.342,
+      "step": 68150
+    },
+    {
+      "epoch": 16.997506234413965,
+      "grad_norm": 7.714181423187256,
+      "learning_rate": 3.009476309226933e-06,
+      "loss": 0.2599,
+      "step": 68160
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 9.013660430908203,
+      "learning_rate": 3.006982543640898e-06,
+      "loss": 0.2809,
+      "step": 68170
+    },
+    {
+      "epoch": 17.0,
+      "eval_loss": 0.4180484116077423,
+      "eval_runtime": 59.928,
+      "eval_samples_per_second": 16.737,
+      "eval_steps_per_second": 16.737,
+      "step": 68170
+    },
+    {
+      "epoch": 17.002493765586035,
+      "grad_norm": 9.489847183227539,
+      "learning_rate": 3.004488778054863e-06,
+      "loss": 0.3557,
+      "step": 68180
+    },
+    {
+      "epoch": 17.00498753117207,
+      "grad_norm": 7.423935890197754,
+      "learning_rate": 3.001995012468828e-06,
+      "loss": 0.3209,
+      "step": 68190
+    },
+    {
+      "epoch": 17.007481296758105,
+      "grad_norm": 5.265260219573975,
+      "learning_rate": 2.9995012468827933e-06,
+      "loss": 0.2767,
+      "step": 68200
+    },
+    {
+      "epoch": 17.00997506234414,
+      "grad_norm": 6.882249355316162,
+      "learning_rate": 2.997007481296758e-06,
+      "loss": 0.2949,
+      "step": 68210
+    },
+    {
+      "epoch": 17.012468827930174,
+      "grad_norm": 10.994675636291504,
+      "learning_rate": 2.9945137157107235e-06,
+      "loss": 0.2567,
+      "step": 68220
+    },
+    {
+      "epoch": 17.01496259351621,
+      "grad_norm": 6.603150367736816,
+      "learning_rate": 2.9920199501246884e-06,
+      "loss": 0.2465,
+      "step": 68230
+    },
+    {
+      "epoch": 17.017456359102244,
+      "grad_norm": 7.306191921234131,
+      "learning_rate": 2.9895261845386537e-06,
+      "loss": 0.3047,
+      "step": 68240
+    },
+    {
+      "epoch": 17.01995012468828,
+      "grad_norm": 6.446793556213379,
+      "learning_rate": 2.987032418952619e-06,
+      "loss": 0.3648,
+      "step": 68250
+    },
+    {
+      "epoch": 17.022443890274314,
+      "grad_norm": 9.686697006225586,
+      "learning_rate": 2.984538653366584e-06,
+      "loss": 0.4001,
+      "step": 68260
+    },
+    {
+      "epoch": 17.02493765586035,
+      "grad_norm": 8.045364379882812,
+      "learning_rate": 2.982044887780549e-06,
+      "loss": 0.3045,
+      "step": 68270
+    },
+    {
+      "epoch": 17.027431421446384,
+      "grad_norm": 11.373425483703613,
+      "learning_rate": 2.979551122194514e-06,
+      "loss": 0.3371,
+      "step": 68280
+    },
+    {
+      "epoch": 17.02992518703242,
+      "grad_norm": 9.176929473876953,
+      "learning_rate": 2.9770573566084794e-06,
+      "loss": 0.3362,
+      "step": 68290
+    },
+    {
+      "epoch": 17.032418952618453,
+      "grad_norm": 9.755634307861328,
+      "learning_rate": 2.974563591022444e-06,
+      "loss": 0.3246,
+      "step": 68300
+    },
+    {
+      "epoch": 17.034912718204488,
+      "grad_norm": 10.913490295410156,
+      "learning_rate": 2.9720698254364096e-06,
+      "loss": 0.3034,
+      "step": 68310
+    },
+    {
+      "epoch": 17.037406483790523,
+      "grad_norm": 7.916656494140625,
+      "learning_rate": 2.969576059850374e-06,
+      "loss": 0.2902,
+      "step": 68320
+    },
+    {
+      "epoch": 17.039900249376558,
+      "grad_norm": 12.00803279876709,
+      "learning_rate": 2.9670822942643394e-06,
+      "loss": 0.3386,
+      "step": 68330
+    },
+    {
+      "epoch": 17.042394014962593,
+      "grad_norm": 8.119451522827148,
+      "learning_rate": 2.9645885286783043e-06,
+      "loss": 0.3176,
+      "step": 68340
+    },
+    {
+      "epoch": 17.044887780548628,
+      "grad_norm": 11.683120727539062,
+      "learning_rate": 2.9620947630922696e-06,
+      "loss": 0.3665,
+      "step": 68350
+    },
+    {
+      "epoch": 17.047381546134662,
+      "grad_norm": 8.209905624389648,
+      "learning_rate": 2.9596009975062345e-06,
+      "loss": 0.3437,
+      "step": 68360
+    },
+    {
+      "epoch": 17.049875311720697,
+      "grad_norm": 9.127650260925293,
+      "learning_rate": 2.9571072319201998e-06,
+      "loss": 0.4074,
+      "step": 68370
+    },
+    {
+      "epoch": 17.052369077306732,
+      "grad_norm": 8.173192977905273,
+      "learning_rate": 2.9546134663341646e-06,
+      "loss": 0.2907,
+      "step": 68380
+    },
+    {
+      "epoch": 17.054862842892767,
+      "grad_norm": 8.60781192779541,
+      "learning_rate": 2.95211970074813e-06,
+      "loss": 0.3102,
+      "step": 68390
+    },
+    {
+      "epoch": 17.057356608478802,
+      "grad_norm": 12.1088228225708,
+      "learning_rate": 2.949625935162095e-06,
+      "loss": 0.3105,
+      "step": 68400
+    },
+    {
+      "epoch": 17.059850374064837,
+      "grad_norm": 6.979897499084473,
+      "learning_rate": 2.94713216957606e-06,
+      "loss": 0.3078,
+      "step": 68410
+    },
+    {
+      "epoch": 17.06234413965087,
+      "grad_norm": 9.642083168029785,
+      "learning_rate": 2.944638403990025e-06,
+      "loss": 0.3219,
+      "step": 68420
+    },
+    {
+      "epoch": 17.064837905236907,
+      "grad_norm": 7.635913848876953,
+      "learning_rate": 2.9421446384039903e-06,
+      "loss": 0.3145,
+      "step": 68430
+    },
+    {
+      "epoch": 17.06733167082294,
+      "grad_norm": 7.816982269287109,
+      "learning_rate": 2.9396508728179552e-06,
+      "loss": 0.3144,
+      "step": 68440
+    },
+    {
+      "epoch": 17.069825436408976,
+      "grad_norm": 9.944385528564453,
+      "learning_rate": 2.9371571072319205e-06,
+      "loss": 0.3679,
+      "step": 68450
+    },
+    {
+      "epoch": 17.07231920199501,
+      "grad_norm": 7.051234722137451,
+      "learning_rate": 2.9346633416458854e-06,
+      "loss": 0.3836,
+      "step": 68460
+    },
+    {
+      "epoch": 17.074812967581046,
+      "grad_norm": 6.5841965675354,
+      "learning_rate": 2.9321695760598507e-06,
+      "loss": 0.3354,
+      "step": 68470
+    },
+    {
+      "epoch": 17.07730673316708,
+      "grad_norm": 7.379906177520752,
+      "learning_rate": 2.9296758104738156e-06,
+      "loss": 0.3299,
+      "step": 68480
+    },
+    {
+      "epoch": 17.079800498753116,
+      "grad_norm": 6.527453899383545,
+      "learning_rate": 2.927182044887781e-06,
+      "loss": 0.2663,
+      "step": 68490
+    },
+    {
+      "epoch": 17.08229426433915,
+      "grad_norm": 9.604272842407227,
+      "learning_rate": 2.9246882793017462e-06,
+      "loss": 0.3929,
+      "step": 68500
+    },
+    {
+      "epoch": 17.084788029925186,
+      "grad_norm": 13.529383659362793,
+      "learning_rate": 2.922194513715711e-06,
+      "loss": 0.2815,
+      "step": 68510
+    },
+    {
+      "epoch": 17.08728179551122,
+      "grad_norm": 8.98625373840332,
+      "learning_rate": 2.9197007481296764e-06,
+      "loss": 0.303,
+      "step": 68520
+    },
+    {
+      "epoch": 17.089775561097255,
+      "grad_norm": 8.426309585571289,
+      "learning_rate": 2.917206982543641e-06,
+      "loss": 0.3067,
+      "step": 68530
+    },
+    {
+      "epoch": 17.09226932668329,
+      "grad_norm": 11.854998588562012,
+      "learning_rate": 2.9147132169576066e-06,
+      "loss": 0.3384,
+      "step": 68540
+    },
+    {
+      "epoch": 17.094763092269325,
+      "grad_norm": 7.317206382751465,
+      "learning_rate": 2.912219451371571e-06,
+      "loss": 0.3036,
+      "step": 68550
+    },
+    {
+      "epoch": 17.09725685785536,
+      "grad_norm": 7.078338146209717,
+      "learning_rate": 2.9097256857855364e-06,
+      "loss": 0.3772,
+      "step": 68560
+    },
+    {
+      "epoch": 17.099750623441395,
+      "grad_norm": 3.3603427410125732,
+      "learning_rate": 2.9072319201995013e-06,
+      "loss": 0.3848,
+      "step": 68570
+    },
+    {
+      "epoch": 17.102244389027433,
+      "grad_norm": 8.076667785644531,
+      "learning_rate": 2.9047381546134666e-06,
+      "loss": 0.3636,
+      "step": 68580
+    },
+    {
+      "epoch": 17.104738154613468,
+      "grad_norm": 8.68441104888916,
+      "learning_rate": 2.9022443890274315e-06,
+      "loss": 0.3029,
+      "step": 68590
+    },
+    {
+      "epoch": 17.107231920199503,
+      "grad_norm": 15.482074737548828,
+      "learning_rate": 2.899750623441397e-06,
+      "loss": 0.3457,
+      "step": 68600
+    },
+    {
+      "epoch": 17.109725685785538,
+      "grad_norm": 8.531909942626953,
+      "learning_rate": 2.8972568578553617e-06,
+      "loss": 0.3074,
+      "step": 68610
+    },
+    {
+      "epoch": 17.112219451371573,
+      "grad_norm": 6.706440448760986,
+      "learning_rate": 2.894763092269327e-06,
+      "loss": 0.3379,
+      "step": 68620
+    },
+    {
+      "epoch": 17.114713216957608,
+      "grad_norm": 8.066004753112793,
+      "learning_rate": 2.892269326683292e-06,
+      "loss": 0.3175,
+      "step": 68630
+    },
+    {
+      "epoch": 17.117206982543642,
+      "grad_norm": 7.285860061645508,
+      "learning_rate": 2.889775561097257e-06,
+      "loss": 0.2926,
+      "step": 68640
+    },
+    {
+      "epoch": 17.119700748129677,
+      "grad_norm": 11.93729019165039,
+      "learning_rate": 2.887531172069826e-06,
+      "loss": 0.4289,
+      "step": 68650
+    },
+    {
+      "epoch": 17.122194513715712,
+      "grad_norm": 9.996557235717773,
+      "learning_rate": 2.8850374064837904e-06,
+      "loss": 0.2979,
+      "step": 68660
+    },
+    {
+      "epoch": 17.124688279301747,
+      "grad_norm": 10.025423049926758,
+      "learning_rate": 2.8825436408977557e-06,
+      "loss": 0.3345,
+      "step": 68670
+    },
+    {
+      "epoch": 17.127182044887782,
+      "grad_norm": 7.561511993408203,
+      "learning_rate": 2.8800498753117214e-06,
+      "loss": 0.3385,
+      "step": 68680
+    },
+    {
+      "epoch": 17.129675810473817,
+      "grad_norm": 6.57540225982666,
+      "learning_rate": 2.877556109725686e-06,
+      "loss": 0.2567,
+      "step": 68690
+    },
+    {
+      "epoch": 17.13216957605985,
+      "grad_norm": 11.776047706604004,
+      "learning_rate": 2.875062344139651e-06,
+      "loss": 0.3308,
+      "step": 68700
+    },
+    {
+      "epoch": 17.134663341645886,
+      "grad_norm": 6.203807353973389,
+      "learning_rate": 2.872568578553616e-06,
+      "loss": 0.2472,
+      "step": 68710
+    },
+    {
+      "epoch": 17.13715710723192,
+      "grad_norm": 9.786406517028809,
+      "learning_rate": 2.8700748129675814e-06,
+      "loss": 0.3577,
+      "step": 68720
+    },
+    {
+      "epoch": 17.139650872817956,
+      "grad_norm": 6.93914794921875,
+      "learning_rate": 2.8675810473815462e-06,
+      "loss": 0.2713,
+      "step": 68730
+    },
+    {
+      "epoch": 17.14214463840399,
+      "grad_norm": 7.5632500648498535,
+      "learning_rate": 2.8650872817955116e-06,
+      "loss": 0.2773,
+      "step": 68740
+    },
+    {
+      "epoch": 17.144638403990026,
+      "grad_norm": 6.6398234367370605,
+      "learning_rate": 2.8625935162094764e-06,
+      "loss": 0.3466,
+      "step": 68750
+    },
+    {
+      "epoch": 17.14713216957606,
+      "grad_norm": 5.831292152404785,
+      "learning_rate": 2.8600997506234418e-06,
+      "loss": 0.348,
+      "step": 68760
+    },
+    {
+      "epoch": 17.149625935162096,
+      "grad_norm": 11.87356185913086,
+      "learning_rate": 2.8576059850374066e-06,
+      "loss": 0.2968,
+      "step": 68770
+    },
+    {
+      "epoch": 17.15211970074813,
+      "grad_norm": 10.019725799560547,
+      "learning_rate": 2.855112219451372e-06,
+      "loss": 0.3226,
+      "step": 68780
+    },
+    {
+      "epoch": 17.154613466334165,
+      "grad_norm": 9.507071495056152,
+      "learning_rate": 2.852618453865337e-06,
+      "loss": 0.3549,
+      "step": 68790
+    },
+    {
+      "epoch": 17.1571072319202,
+      "grad_norm": 10.056197166442871,
+      "learning_rate": 2.850124688279302e-06,
+      "loss": 0.2982,
+      "step": 68800
+    },
+    {
+      "epoch": 17.159600997506235,
+      "grad_norm": 15.976298332214355,
+      "learning_rate": 2.847630922693267e-06,
+      "loss": 0.3499,
+      "step": 68810
+    },
+    {
+      "epoch": 17.16209476309227,
+      "grad_norm": 11.574234962463379,
+      "learning_rate": 2.8451371571072323e-06,
+      "loss": 0.3486,
+      "step": 68820
+    },
+    {
+      "epoch": 17.164588528678305,
+      "grad_norm": 7.35728645324707,
+      "learning_rate": 2.8426433915211972e-06,
+      "loss": 0.3749,
+      "step": 68830
+    },
+    {
+      "epoch": 17.16708229426434,
+      "grad_norm": 9.311493873596191,
+      "learning_rate": 2.8401496259351625e-06,
+      "loss": 0.3784,
+      "step": 68840
+    },
+    {
+      "epoch": 17.169576059850375,
+      "grad_norm": 7.325751304626465,
+      "learning_rate": 2.837655860349127e-06,
+      "loss": 0.2806,
+      "step": 68850
+    },
+    {
+      "epoch": 17.17206982543641,
+      "grad_norm": 6.172769069671631,
+      "learning_rate": 2.8351620947630927e-06,
+      "loss": 0.2538,
+      "step": 68860
+    },
+    {
+      "epoch": 17.174563591022444,
+      "grad_norm": 10.258570671081543,
+      "learning_rate": 2.832668329177057e-06,
+      "loss": 0.3006,
+      "step": 68870
+    },
+    {
+      "epoch": 17.17705735660848,
+      "grad_norm": 6.926236152648926,
+      "learning_rate": 2.8301745635910225e-06,
+      "loss": 0.2841,
+      "step": 68880
+    },
+    {
+      "epoch": 17.179551122194514,
+      "grad_norm": 9.146723747253418,
+      "learning_rate": 2.8276807980049874e-06,
+      "loss": 0.3016,
+      "step": 68890
+    },
+    {
+      "epoch": 17.18204488778055,
+      "grad_norm": 6.017948150634766,
+      "learning_rate": 2.8251870324189527e-06,
+      "loss": 0.3461,
+      "step": 68900
+    },
+    {
+      "epoch": 17.184538653366584,
+      "grad_norm": 10.936685562133789,
+      "learning_rate": 2.8226932668329176e-06,
+      "loss": 0.3771,
+      "step": 68910
+    },
+    {
+      "epoch": 17.18703241895262,
+      "grad_norm": 8.554216384887695,
+      "learning_rate": 2.820199501246883e-06,
+      "loss": 0.2986,
+      "step": 68920
+    },
+    {
+      "epoch": 17.189526184538654,
+      "grad_norm": 8.152572631835938,
+      "learning_rate": 2.817705735660848e-06,
+      "loss": 0.3383,
+      "step": 68930
+    },
+    {
+      "epoch": 17.19201995012469,
+      "grad_norm": 8.737922668457031,
+      "learning_rate": 2.815211970074813e-06,
+      "loss": 0.316,
+      "step": 68940
+    },
+    {
+      "epoch": 17.194513715710723,
+      "grad_norm": 8.8933744430542,
+      "learning_rate": 2.8127182044887784e-06,
+      "loss": 0.3481,
+      "step": 68950
+    },
+    {
+      "epoch": 17.197007481296758,
+      "grad_norm": 9.593643188476562,
+      "learning_rate": 2.8102244389027433e-06,
+      "loss": 0.3012,
+      "step": 68960
+    },
+    {
+      "epoch": 17.199501246882793,
+      "grad_norm": 6.942617416381836,
+      "learning_rate": 2.8077306733167086e-06,
+      "loss": 0.2645,
+      "step": 68970
+    },
+    {
+      "epoch": 17.201995012468828,
+      "grad_norm": 9.486404418945312,
+      "learning_rate": 2.8052369077306735e-06,
+      "loss": 0.3266,
+      "step": 68980
+    },
+    {
+      "epoch": 17.204488778054863,
+      "grad_norm": 5.64856481552124,
+      "learning_rate": 2.802743142144639e-06,
+      "loss": 0.2926,
+      "step": 68990
+    },
+    {
+      "epoch": 17.206982543640898,
+      "grad_norm": 8.93766975402832,
+      "learning_rate": 2.8002493765586037e-06,
+      "loss": 0.2835,
+      "step": 69000
+    },
+    {
+      "epoch": 17.209476309226932,
+      "grad_norm": 7.289651393890381,
+      "learning_rate": 2.797755610972569e-06,
+      "loss": 0.3044,
+      "step": 69010
+    },
+    {
+      "epoch": 17.211970074812967,
+      "grad_norm": 9.757658004760742,
+      "learning_rate": 2.795261845386534e-06,
+      "loss": 0.304,
+      "step": 69020
+    },
+    {
+      "epoch": 17.214463840399002,
+      "grad_norm": 9.677502632141113,
+      "learning_rate": 2.792768079800499e-06,
+      "loss": 0.354,
+      "step": 69030
+    },
+    {
+      "epoch": 17.216957605985037,
+      "grad_norm": 7.527002811431885,
+      "learning_rate": 2.790274314214464e-06,
+      "loss": 0.3659,
+      "step": 69040
+    },
+    {
+      "epoch": 17.219451371571072,
+      "grad_norm": 15.7623872756958,
+      "learning_rate": 2.7877805486284294e-06,
+      "loss": 0.2914,
+      "step": 69050
+    },
+    {
+      "epoch": 17.221945137157107,
+      "grad_norm": 7.540277004241943,
+      "learning_rate": 2.7852867830423943e-06,
+      "loss": 0.3512,
+      "step": 69060
+    },
+    {
+      "epoch": 17.22443890274314,
+      "grad_norm": 8.1314697265625,
+      "learning_rate": 2.7827930174563596e-06,
+      "loss": 0.317,
+      "step": 69070
+    },
+    {
+      "epoch": 17.226932668329177,
+      "grad_norm": 8.093489646911621,
+      "learning_rate": 2.780299251870324e-06,
+      "loss": 0.3402,
+      "step": 69080
+    },
+    {
+      "epoch": 17.22942643391521,
+      "grad_norm": 4.8851094245910645,
+      "learning_rate": 2.7778054862842898e-06,
+      "loss": 0.3102,
+      "step": 69090
+    },
+    {
+      "epoch": 17.231920199501246,
+      "grad_norm": 6.918999671936035,
+      "learning_rate": 2.7753117206982542e-06,
+      "loss": 0.2954,
+      "step": 69100
+    },
+    {
+      "epoch": 17.23441396508728,
+      "grad_norm": 11.607989311218262,
+      "learning_rate": 2.7728179551122195e-06,
+      "loss": 0.3554,
+      "step": 69110
+    },
+    {
+      "epoch": 17.236907730673316,
+      "grad_norm": 8.533638000488281,
+      "learning_rate": 2.7703241895261844e-06,
+      "loss": 0.2778,
+      "step": 69120
+    },
+    {
+      "epoch": 17.23940149625935,
+      "grad_norm": 6.778941631317139,
+      "learning_rate": 2.7678304239401497e-06,
+      "loss": 0.3103,
+      "step": 69130
+    },
+    {
+      "epoch": 17.241895261845386,
+      "grad_norm": 8.85729694366455,
+      "learning_rate": 2.7653366583541146e-06,
+      "loss": 0.328,
+      "step": 69140
+    },
+    {
+      "epoch": 17.24438902743142,
+      "grad_norm": 11.336535453796387,
+      "learning_rate": 2.76284289276808e-06,
+      "loss": 0.3471,
+      "step": 69150
+    },
+    {
+      "epoch": 17.246882793017456,
+      "grad_norm": 8.672399520874023,
+      "learning_rate": 2.760349127182045e-06,
+      "loss": 0.4223,
+      "step": 69160
+    },
+    {
+      "epoch": 17.24937655860349,
+      "grad_norm": 6.419161796569824,
+      "learning_rate": 2.75785536159601e-06,
+      "loss": 0.3823,
+      "step": 69170
+    },
+    {
+      "epoch": 17.251870324189525,
+      "grad_norm": 6.270040035247803,
+      "learning_rate": 2.7553615960099754e-06,
+      "loss": 0.2993,
+      "step": 69180
+    },
+    {
+      "epoch": 17.25436408977556,
+      "grad_norm": 6.386322975158691,
+      "learning_rate": 2.7528678304239403e-06,
+      "loss": 0.3222,
+      "step": 69190
+    },
+    {
+      "epoch": 17.256857855361595,
+      "grad_norm": 10.8588285446167,
+      "learning_rate": 2.7503740648379056e-06,
+      "loss": 0.372,
+      "step": 69200
+    },
+    {
+      "epoch": 17.25935162094763,
+      "grad_norm": 9.645064353942871,
+      "learning_rate": 2.7478802992518705e-06,
+      "loss": 0.2823,
+      "step": 69210
+    },
+    {
+      "epoch": 17.261845386533665,
+      "grad_norm": 8.688511848449707,
+      "learning_rate": 2.745386533665836e-06,
+      "loss": 0.3246,
+      "step": 69220
+    },
+    {
+      "epoch": 17.2643391521197,
+      "grad_norm": 7.549156665802002,
+      "learning_rate": 2.7428927680798007e-06,
+      "loss": 0.2917,
+      "step": 69230
+    },
+    {
+      "epoch": 17.266832917705734,
+      "grad_norm": 7.3209357261657715,
+      "learning_rate": 2.740399002493766e-06,
+      "loss": 0.2893,
+      "step": 69240
+    },
+    {
+      "epoch": 17.26932668329177,
+      "grad_norm": 9.329798698425293,
+      "learning_rate": 2.737905236907731e-06,
+      "loss": 0.3382,
+      "step": 69250
+    },
+    {
+      "epoch": 17.271820448877804,
+      "grad_norm": 17.011377334594727,
+      "learning_rate": 2.7354114713216962e-06,
+      "loss": 0.2877,
+      "step": 69260
+    },
+    {
+      "epoch": 17.27431421446384,
+      "grad_norm": 13.201550483703613,
+      "learning_rate": 2.732917705735661e-06,
+      "loss": 0.3846,
+      "step": 69270
+    },
+    {
+      "epoch": 17.276807980049874,
+      "grad_norm": 7.851199150085449,
+      "learning_rate": 2.7304239401496264e-06,
+      "loss": 0.2899,
+      "step": 69280
+    },
+    {
+      "epoch": 17.27930174563591,
+      "grad_norm": 7.01190185546875,
+      "learning_rate": 2.7279301745635913e-06,
+      "loss": 0.3709,
+      "step": 69290
+    },
+    {
+      "epoch": 17.281795511221944,
+      "grad_norm": 8.432855606079102,
+      "learning_rate": 2.7254364089775566e-06,
+      "loss": 0.2868,
+      "step": 69300
+    },
+    {
+      "epoch": 17.28428927680798,
+      "grad_norm": 9.088372230529785,
+      "learning_rate": 2.722942643391521e-06,
+      "loss": 0.3208,
+      "step": 69310
+    },
+    {
+      "epoch": 17.286783042394013,
+      "grad_norm": 7.609044551849365,
+      "learning_rate": 2.720448877805487e-06,
+      "loss": 0.3291,
+      "step": 69320
+    },
+    {
+      "epoch": 17.28927680798005,
+      "grad_norm": 18.67131996154785,
+      "learning_rate": 2.7179551122194513e-06,
+      "loss": 0.38,
+      "step": 69330
+    },
+    {
+      "epoch": 17.291770573566083,
+      "grad_norm": 7.562769889831543,
+      "learning_rate": 2.7154613466334166e-06,
+      "loss": 0.3773,
+      "step": 69340
+    },
+    {
+      "epoch": 17.294264339152118,
+      "grad_norm": 6.86823844909668,
+      "learning_rate": 2.7129675810473815e-06,
+      "loss": 0.3447,
+      "step": 69350
+    },
+    {
+      "epoch": 17.296758104738153,
+      "grad_norm": 13.892622947692871,
+      "learning_rate": 2.7104738154613468e-06,
+      "loss": 0.3643,
+      "step": 69360
+    },
+    {
+      "epoch": 17.29925187032419,
+      "grad_norm": 7.478859901428223,
+      "learning_rate": 2.7079800498753117e-06,
+      "loss": 0.3812,
+      "step": 69370
+    },
+    {
+      "epoch": 17.301745635910226,
+      "grad_norm": 11.587136268615723,
+      "learning_rate": 2.705486284289277e-06,
+      "loss": 0.3247,
+      "step": 69380
+    },
+    {
+      "epoch": 17.30423940149626,
+      "grad_norm": 11.155722618103027,
+      "learning_rate": 2.702992518703242e-06,
+      "loss": 0.2928,
+      "step": 69390
+    },
+    {
+      "epoch": 17.306733167082296,
+      "grad_norm": 5.776709079742432,
+      "learning_rate": 2.700498753117207e-06,
+      "loss": 0.3003,
+      "step": 69400
+    },
+    {
+      "epoch": 17.30922693266833,
+      "grad_norm": 11.081796646118164,
+      "learning_rate": 2.698004987531172e-06,
+      "loss": 0.3558,
+      "step": 69410
+    },
+    {
+      "epoch": 17.311720698254366,
+      "grad_norm": 10.856132507324219,
+      "learning_rate": 2.6955112219451374e-06,
+      "loss": 0.3469,
+      "step": 69420
+    },
+    {
+      "epoch": 17.3142144638404,
+      "grad_norm": 7.837653636932373,
+      "learning_rate": 2.6930174563591027e-06,
+      "loss": 0.2866,
+      "step": 69430
+    },
+    {
+      "epoch": 17.316708229426435,
+      "grad_norm": 8.779632568359375,
+      "learning_rate": 2.6905236907730676e-06,
+      "loss": 0.3857,
+      "step": 69440
+    },
+    {
+      "epoch": 17.31920199501247,
+      "grad_norm": 9.35014820098877,
+      "learning_rate": 2.688029925187033e-06,
+      "loss": 0.2853,
+      "step": 69450
+    },
+    {
+      "epoch": 17.321695760598505,
+      "grad_norm": 4.780736923217773,
+      "learning_rate": 2.6855361596009978e-06,
+      "loss": 0.3128,
+      "step": 69460
+    },
+    {
+      "epoch": 17.32418952618454,
+      "grad_norm": 9.295073509216309,
+      "learning_rate": 2.683042394014963e-06,
+      "loss": 0.3148,
+      "step": 69470
+    },
+    {
+      "epoch": 17.326683291770575,
+      "grad_norm": 8.449419975280762,
+      "learning_rate": 2.680548628428928e-06,
+      "loss": 0.3382,
+      "step": 69480
+    },
+    {
+      "epoch": 17.32917705735661,
+      "grad_norm": 7.874120712280273,
+      "learning_rate": 2.6780548628428933e-06,
+      "loss": 0.3143,
+      "step": 69490
+    },
+    {
+      "epoch": 17.331670822942645,
+      "grad_norm": 6.140561103820801,
+      "learning_rate": 2.675561097256858e-06,
+      "loss": 0.3332,
+      "step": 69500
+    },
+    {
+      "epoch": 17.33416458852868,
+      "grad_norm": 9.675243377685547,
+      "learning_rate": 2.6730673316708235e-06,
+      "loss": 0.2659,
+      "step": 69510
+    },
+    {
+      "epoch": 17.336658354114714,
+      "grad_norm": 8.507969856262207,
+      "learning_rate": 2.670573566084788e-06,
+      "loss": 0.3464,
+      "step": 69520
+    },
+    {
+      "epoch": 17.33915211970075,
+      "grad_norm": 9.791665077209473,
+      "learning_rate": 2.6680798004987537e-06,
+      "loss": 0.3107,
+      "step": 69530
+    },
+    {
+      "epoch": 17.341645885286784,
+      "grad_norm": 6.693220138549805,
+      "learning_rate": 2.665586034912718e-06,
+      "loss": 0.3108,
+      "step": 69540
+    },
+    {
+      "epoch": 17.34413965087282,
+      "grad_norm": 9.014039993286133,
+      "learning_rate": 2.6630922693266834e-06,
+      "loss": 0.3689,
+      "step": 69550
+    },
+    {
+      "epoch": 17.346633416458854,
+      "grad_norm": 9.79101848602295,
+      "learning_rate": 2.6605985037406483e-06,
+      "loss": 0.3281,
+      "step": 69560
+    },
+    {
+      "epoch": 17.34912718204489,
+      "grad_norm": 6.341735363006592,
+      "learning_rate": 2.6581047381546136e-06,
+      "loss": 0.2585,
+      "step": 69570
+    },
+    {
+      "epoch": 17.351620947630924,
+      "grad_norm": 7.738804340362549,
+      "learning_rate": 2.6556109725685785e-06,
+      "loss": 0.3152,
+      "step": 69580
+    },
+    {
+      "epoch": 17.35411471321696,
+      "grad_norm": 8.257248878479004,
+      "learning_rate": 2.653117206982544e-06,
+      "loss": 0.3581,
+      "step": 69590
+    },
+    {
+      "epoch": 17.356608478802993,
+      "grad_norm": 9.338336944580078,
+      "learning_rate": 2.6506234413965087e-06,
+      "loss": 0.302,
+      "step": 69600
+    },
+    {
+      "epoch": 17.359102244389028,
+      "grad_norm": 9.739322662353516,
+      "learning_rate": 2.648129675810474e-06,
+      "loss": 0.2901,
+      "step": 69610
+    },
+    {
+      "epoch": 17.361596009975063,
+      "grad_norm": 9.487130165100098,
+      "learning_rate": 2.645635910224439e-06,
+      "loss": 0.3585,
+      "step": 69620
+    },
+    {
+      "epoch": 17.364089775561098,
+      "grad_norm": 10.598410606384277,
+      "learning_rate": 2.6431421446384042e-06,
+      "loss": 0.3382,
+      "step": 69630
+    },
+    {
+      "epoch": 17.366583541147133,
+      "grad_norm": 8.358076095581055,
+      "learning_rate": 2.640648379052369e-06,
+      "loss": 0.3341,
+      "step": 69640
+    },
+    {
+      "epoch": 17.369077306733168,
+      "grad_norm": 11.573534965515137,
+      "learning_rate": 2.6381546134663344e-06,
+      "loss": 0.3532,
+      "step": 69650
+    },
+    {
+      "epoch": 17.371571072319203,
+      "grad_norm": 10.417460441589355,
+      "learning_rate": 2.6356608478802997e-06,
+      "loss": 0.3544,
+      "step": 69660
+    },
+    {
+      "epoch": 17.374064837905237,
+      "grad_norm": 12.179265975952148,
+      "learning_rate": 2.6331670822942646e-06,
+      "loss": 0.304,
+      "step": 69670
+    },
+    {
+      "epoch": 17.376558603491272,
+      "grad_norm": 8.12159252166748,
+      "learning_rate": 2.63067331670823e-06,
+      "loss": 0.3341,
+      "step": 69680
+    },
+    {
+      "epoch": 17.379052369077307,
+      "grad_norm": 9.424428939819336,
+      "learning_rate": 2.628179551122195e-06,
+      "loss": 0.3425,
+      "step": 69690
+    },
+    {
+      "epoch": 17.381546134663342,
+      "grad_norm": 7.936733245849609,
+      "learning_rate": 2.62568578553616e-06,
+      "loss": 0.2796,
+      "step": 69700
+    },
+    {
+      "epoch": 17.384039900249377,
+      "grad_norm": 9.32441234588623,
+      "learning_rate": 2.623192019950125e-06,
+      "loss": 0.3235,
+      "step": 69710
+    },
+    {
+      "epoch": 17.38653366583541,
+      "grad_norm": 10.321310043334961,
+      "learning_rate": 2.6206982543640903e-06,
+      "loss": 0.3289,
+      "step": 69720
+    },
+    {
+      "epoch": 17.389027431421447,
+      "grad_norm": 7.7665181159973145,
+      "learning_rate": 2.618204488778055e-06,
+      "loss": 0.3161,
+      "step": 69730
+    },
+    {
+      "epoch": 17.39152119700748,
+      "grad_norm": 9.732200622558594,
+      "learning_rate": 2.6157107231920205e-06,
+      "loss": 0.2868,
+      "step": 69740
+    },
+    {
+      "epoch": 17.394014962593516,
+      "grad_norm": 6.528557300567627,
+      "learning_rate": 2.613216957605985e-06,
+      "loss": 0.3435,
+      "step": 69750
+    },
+    {
+      "epoch": 17.39650872817955,
+      "grad_norm": 8.758277893066406,
+      "learning_rate": 2.6107231920199507e-06,
+      "loss": 0.3674,
+      "step": 69760
+    },
+    {
+      "epoch": 17.399002493765586,
+      "grad_norm": 7.994776248931885,
+      "learning_rate": 2.608229426433915e-06,
+      "loss": 0.372,
+      "step": 69770
+    },
+    {
+      "epoch": 17.40149625935162,
+      "grad_norm": 9.376730918884277,
+      "learning_rate": 2.6057356608478805e-06,
+      "loss": 0.3216,
+      "step": 69780
+    },
+    {
+      "epoch": 17.403990024937656,
+      "grad_norm": 6.297952175140381,
+      "learning_rate": 2.6032418952618454e-06,
+      "loss": 0.3051,
+      "step": 69790
+    },
+    {
+      "epoch": 17.40648379052369,
+      "grad_norm": 8.456469535827637,
+      "learning_rate": 2.6007481296758107e-06,
+      "loss": 0.3467,
+      "step": 69800
+    },
+    {
+      "epoch": 17.408977556109726,
+      "grad_norm": 7.9176859855651855,
+      "learning_rate": 2.5982543640897756e-06,
+      "loss": 0.3259,
+      "step": 69810
+    },
+    {
+      "epoch": 17.41147132169576,
+      "grad_norm": 9.109759330749512,
+      "learning_rate": 2.595760598503741e-06,
+      "loss": 0.3349,
+      "step": 69820
+    },
+    {
+      "epoch": 17.413965087281795,
+      "grad_norm": 8.499098777770996,
+      "learning_rate": 2.5932668329177058e-06,
+      "loss": 0.3412,
+      "step": 69830
+    },
+    {
+      "epoch": 17.41645885286783,
+      "grad_norm": 8.873187065124512,
+      "learning_rate": 2.590773067331671e-06,
+      "loss": 0.2991,
+      "step": 69840
+    },
+    {
+      "epoch": 17.418952618453865,
+      "grad_norm": 6.6649489402771,
+      "learning_rate": 2.588279301745636e-06,
+      "loss": 0.362,
+      "step": 69850
+    },
+    {
+      "epoch": 17.4214463840399,
+      "grad_norm": 7.3186259269714355,
+      "learning_rate": 2.5857855361596013e-06,
+      "loss": 0.3563,
+      "step": 69860
+    },
+    {
+      "epoch": 17.423940149625935,
+      "grad_norm": 6.746584892272949,
+      "learning_rate": 2.583291770573566e-06,
+      "loss": 0.3346,
+      "step": 69870
+    },
+    {
+      "epoch": 17.42643391521197,
+      "grad_norm": 10.556224822998047,
+      "learning_rate": 2.5807980049875315e-06,
+      "loss": 0.367,
+      "step": 69880
+    },
+    {
+      "epoch": 17.428927680798004,
+      "grad_norm": 8.307790756225586,
+      "learning_rate": 2.5783042394014963e-06,
+      "loss": 0.337,
+      "step": 69890
+    },
+    {
+      "epoch": 17.43142144638404,
+      "grad_norm": 18.567258834838867,
+      "learning_rate": 2.5758104738154617e-06,
+      "loss": 0.3254,
+      "step": 69900
+    },
+    {
+      "epoch": 17.433915211970074,
+      "grad_norm": 8.430441856384277,
+      "learning_rate": 2.573316708229427e-06,
+      "loss": 0.2723,
+      "step": 69910
+    },
+    {
+      "epoch": 17.43640897755611,
+      "grad_norm": 6.945594310760498,
+      "learning_rate": 2.570822942643392e-06,
+      "loss": 0.3487,
+      "step": 69920
+    },
+    {
+      "epoch": 17.438902743142144,
+      "grad_norm": 9.82526969909668,
+      "learning_rate": 2.568329177057357e-06,
+      "loss": 0.3007,
+      "step": 69930
+    },
+    {
+      "epoch": 17.44139650872818,
+      "grad_norm": 9.881370544433594,
+      "learning_rate": 2.565835411471322e-06,
+      "loss": 0.3641,
+      "step": 69940
+    },
+    {
+      "epoch": 17.443890274314214,
+      "grad_norm": 10.655155181884766,
+      "learning_rate": 2.5633416458852874e-06,
+      "loss": 0.316,
+      "step": 69950
+    },
+    {
+      "epoch": 17.44638403990025,
+      "grad_norm": 7.941926002502441,
+      "learning_rate": 2.560847880299252e-06,
+      "loss": 0.2926,
+      "step": 69960
+    },
+    {
+      "epoch": 17.448877805486283,
+      "grad_norm": 10.348051071166992,
+      "learning_rate": 2.5583541147132175e-06,
+      "loss": 0.3244,
+      "step": 69970
+    },
+    {
+      "epoch": 17.45137157107232,
+      "grad_norm": 9.219032287597656,
+      "learning_rate": 2.555860349127182e-06,
+      "loss": 0.2996,
+      "step": 69980
+    },
+    {
+      "epoch": 17.453865336658353,
+      "grad_norm": 12.931209564208984,
+      "learning_rate": 2.5533665835411473e-06,
+      "loss": 0.3215,
+      "step": 69990
+    },
+    {
+      "epoch": 17.456359102244388,
+      "grad_norm": 12.621508598327637,
+      "learning_rate": 2.550872817955112e-06,
+      "loss": 0.3766,
+      "step": 70000
+    },
+    {
+      "epoch": 17.458852867830423,
+      "grad_norm": 13.947381019592285,
+      "learning_rate": 2.5483790523690775e-06,
+      "loss": 0.3613,
+      "step": 70010
+    },
+    {
+      "epoch": 17.461346633416458,
+      "grad_norm": 8.649821281433105,
+      "learning_rate": 2.5458852867830424e-06,
+      "loss": 0.3608,
+      "step": 70020
+    },
+    {
+      "epoch": 17.463840399002493,
+      "grad_norm": 13.746424674987793,
+      "learning_rate": 2.5433915211970077e-06,
+      "loss": 0.3476,
+      "step": 70030
+    },
+    {
+      "epoch": 17.466334164588527,
+      "grad_norm": 8.04629898071289,
+      "learning_rate": 2.5408977556109726e-06,
+      "loss": 0.2839,
+      "step": 70040
+    },
+    {
+      "epoch": 17.468827930174562,
+      "grad_norm": 13.645923614501953,
+      "learning_rate": 2.538403990024938e-06,
+      "loss": 0.3092,
+      "step": 70050
+    },
+    {
+      "epoch": 17.471321695760597,
+      "grad_norm": 6.63210391998291,
+      "learning_rate": 2.535910224438903e-06,
+      "loss": 0.3725,
+      "step": 70060
+    },
+    {
+      "epoch": 17.473815461346632,
+      "grad_norm": 6.891839981079102,
+      "learning_rate": 2.533416458852868e-06,
+      "loss": 0.2888,
+      "step": 70070
+    },
+    {
+      "epoch": 17.476309226932667,
+      "grad_norm": 9.221366882324219,
+      "learning_rate": 2.530922693266833e-06,
+      "loss": 0.3489,
+      "step": 70080
+    },
+    {
+      "epoch": 17.478802992518702,
+      "grad_norm": 10.869357109069824,
+      "learning_rate": 2.5284289276807983e-06,
+      "loss": 0.4087,
+      "step": 70090
+    },
+    {
+      "epoch": 17.481296758104737,
+      "grad_norm": 9.193957328796387,
+      "learning_rate": 2.525935162094763e-06,
+      "loss": 0.324,
+      "step": 70100
+    },
+    {
+      "epoch": 17.48379052369077,
+      "grad_norm": 7.817142009735107,
+      "learning_rate": 2.5234413965087285e-06,
+      "loss": 0.3491,
+      "step": 70110
+    },
+    {
+      "epoch": 17.486284289276806,
+      "grad_norm": 11.097153663635254,
+      "learning_rate": 2.5209476309226934e-06,
+      "loss": 0.2901,
+      "step": 70120
+    },
+    {
+      "epoch": 17.48877805486284,
+      "grad_norm": 8.653058052062988,
+      "learning_rate": 2.5184538653366587e-06,
+      "loss": 0.3366,
+      "step": 70130
+    },
+    {
+      "epoch": 17.491271820448876,
+      "grad_norm": 7.455108642578125,
+      "learning_rate": 2.5159600997506236e-06,
+      "loss": 0.342,
+      "step": 70140
+    },
+    {
+      "epoch": 17.49376558603491,
+      "grad_norm": 6.872528553009033,
+      "learning_rate": 2.513466334164589e-06,
+      "loss": 0.3033,
+      "step": 70150
+    },
+    {
+      "epoch": 17.496259351620946,
+      "grad_norm": 7.771061420440674,
+      "learning_rate": 2.510972568578554e-06,
+      "loss": 0.2978,
+      "step": 70160
+    },
+    {
+      "epoch": 17.49875311720698,
+      "grad_norm": 7.269931316375732,
+      "learning_rate": 2.508478802992519e-06,
+      "loss": 0.3072,
+      "step": 70170
+    },
+    {
+      "epoch": 17.50124688279302,
+      "grad_norm": 10.571673393249512,
+      "learning_rate": 2.5059850374064844e-06,
+      "loss": 0.4029,
+      "step": 70180
+    },
+    {
+      "epoch": 17.503740648379054,
+      "grad_norm": 7.55682373046875,
+      "learning_rate": 2.503491271820449e-06,
+      "loss": 0.3645,
+      "step": 70190
+    },
+    {
+      "epoch": 17.50623441396509,
+      "grad_norm": 8.619101524353027,
+      "learning_rate": 2.5009975062344146e-06,
+      "loss": 0.4596,
+      "step": 70200
+    },
+    {
+      "epoch": 17.508728179551124,
+      "grad_norm": 8.528571128845215,
+      "learning_rate": 2.498503740648379e-06,
+      "loss": 0.3011,
+      "step": 70210
+    },
+    {
+      "epoch": 17.51122194513716,
+      "grad_norm": 11.281834602355957,
+      "learning_rate": 2.4960099750623444e-06,
+      "loss": 0.3314,
+      "step": 70220
+    },
+    {
+      "epoch": 17.513715710723194,
+      "grad_norm": 6.625566482543945,
+      "learning_rate": 2.4935162094763092e-06,
+      "loss": 0.2803,
+      "step": 70230
+    },
+    {
+      "epoch": 17.51620947630923,
+      "grad_norm": 11.519903182983398,
+      "learning_rate": 2.4910224438902746e-06,
+      "loss": 0.3145,
+      "step": 70240
+    },
+    {
+      "epoch": 17.518703241895263,
+      "grad_norm": 7.888875484466553,
+      "learning_rate": 2.4885286783042394e-06,
+      "loss": 0.3021,
+      "step": 70250
+    },
+    {
+      "epoch": 17.521197007481298,
+      "grad_norm": 7.573606967926025,
+      "learning_rate": 2.4860349127182048e-06,
+      "loss": 0.2962,
+      "step": 70260
+    },
+    {
+      "epoch": 17.523690773067333,
+      "grad_norm": 8.137581825256348,
+      "learning_rate": 2.4835411471321696e-06,
+      "loss": 0.3396,
+      "step": 70270
+    },
+    {
+      "epoch": 17.526184538653368,
+      "grad_norm": 10.423111915588379,
+      "learning_rate": 2.481047381546135e-06,
+      "loss": 0.3194,
+      "step": 70280
+    },
+    {
+      "epoch": 17.528678304239403,
+      "grad_norm": 11.856714248657227,
+      "learning_rate": 2.4785536159601003e-06,
+      "loss": 0.3732,
+      "step": 70290
+    },
+    {
+      "epoch": 17.531172069825438,
+      "grad_norm": 8.993725776672363,
+      "learning_rate": 2.476059850374065e-06,
+      "loss": 0.288,
+      "step": 70300
+    },
+    {
+      "epoch": 17.533665835411473,
+      "grad_norm": 9.244604110717773,
+      "learning_rate": 2.47356608478803e-06,
+      "loss": 0.3439,
+      "step": 70310
+    },
+    {
+      "epoch": 17.536159600997507,
+      "grad_norm": 9.765801429748535,
+      "learning_rate": 2.4710723192019953e-06,
+      "loss": 0.3375,
+      "step": 70320
+    },
+    {
+      "epoch": 17.538653366583542,
+      "grad_norm": 5.98468542098999,
+      "learning_rate": 2.4685785536159602e-06,
+      "loss": 0.302,
+      "step": 70330
+    },
+    {
+      "epoch": 17.541147132169577,
+      "grad_norm": 8.877677917480469,
+      "learning_rate": 2.4660847880299255e-06,
+      "loss": 0.387,
+      "step": 70340
+    },
+    {
+      "epoch": 17.543640897755612,
+      "grad_norm": 10.188618659973145,
+      "learning_rate": 2.4635910224438904e-06,
+      "loss": 0.3714,
+      "step": 70350
+    },
+    {
+      "epoch": 17.546134663341647,
+      "grad_norm": 9.320550918579102,
+      "learning_rate": 2.4610972568578557e-06,
+      "loss": 0.3256,
+      "step": 70360
+    },
+    {
+      "epoch": 17.54862842892768,
+      "grad_norm": 7.385508060455322,
+      "learning_rate": 2.4586034912718206e-06,
+      "loss": 0.2825,
+      "step": 70370
+    },
+    {
+      "epoch": 17.551122194513717,
+      "grad_norm": 8.037924766540527,
+      "learning_rate": 2.456109725685786e-06,
+      "loss": 0.3273,
+      "step": 70380
+    },
+    {
+      "epoch": 17.55361596009975,
+      "grad_norm": 10.301573753356934,
+      "learning_rate": 2.453615960099751e-06,
+      "loss": 0.2664,
+      "step": 70390
+    },
+    {
+      "epoch": 17.556109725685786,
+      "grad_norm": 7.737729072570801,
+      "learning_rate": 2.451122194513716e-06,
+      "loss": 0.3422,
+      "step": 70400
+    },
+    {
+      "epoch": 17.55860349127182,
+      "grad_norm": 7.797543525695801,
+      "learning_rate": 2.448628428927681e-06,
+      "loss": 0.3644,
+      "step": 70410
+    },
+    {
+      "epoch": 17.561097256857856,
+      "grad_norm": 9.298747062683105,
+      "learning_rate": 2.446134663341646e-06,
+      "loss": 0.3983,
+      "step": 70420
+    },
+    {
+      "epoch": 17.56359102244389,
+      "grad_norm": 8.475441932678223,
+      "learning_rate": 2.443640897755611e-06,
+      "loss": 0.3913,
+      "step": 70430
+    },
+    {
+      "epoch": 17.566084788029926,
+      "grad_norm": 7.733828544616699,
+      "learning_rate": 2.441147132169576e-06,
+      "loss": 0.3371,
+      "step": 70440
+    },
+    {
+      "epoch": 17.56857855361596,
+      "grad_norm": 7.250517845153809,
+      "learning_rate": 2.4386533665835414e-06,
+      "loss": 0.3646,
+      "step": 70450
+    },
+    {
+      "epoch": 17.571072319201996,
+      "grad_norm": 13.300467491149902,
+      "learning_rate": 2.4361596009975063e-06,
+      "loss": 0.3736,
+      "step": 70460
+    },
+    {
+      "epoch": 17.57356608478803,
+      "grad_norm": 7.141026973724365,
+      "learning_rate": 2.4336658354114716e-06,
+      "loss": 0.2415,
+      "step": 70470
+    },
+    {
+      "epoch": 17.576059850374065,
+      "grad_norm": 6.321515083312988,
+      "learning_rate": 2.4311720698254365e-06,
+      "loss": 0.3182,
+      "step": 70480
+    },
+    {
+      "epoch": 17.5785536159601,
+      "grad_norm": 10.123273849487305,
+      "learning_rate": 2.428678304239402e-06,
+      "loss": 0.3007,
+      "step": 70490
+    },
+    {
+      "epoch": 17.581047381546135,
+      "grad_norm": 6.999142169952393,
+      "learning_rate": 2.4261845386533667e-06,
+      "loss": 0.3756,
+      "step": 70500
+    },
+    {
+      "epoch": 17.58354114713217,
+      "grad_norm": 11.76720905303955,
+      "learning_rate": 2.4236907730673316e-06,
+      "loss": 0.3423,
+      "step": 70510
+    },
+    {
+      "epoch": 17.586034912718205,
+      "grad_norm": 7.583934783935547,
+      "learning_rate": 2.421197007481297e-06,
+      "loss": 0.3457,
+      "step": 70520
+    },
+    {
+      "epoch": 17.58852867830424,
+      "grad_norm": 7.330530166625977,
+      "learning_rate": 2.418703241895262e-06,
+      "loss": 0.2704,
+      "step": 70530
+    },
+    {
+      "epoch": 17.591022443890274,
+      "grad_norm": 9.098861694335938,
+      "learning_rate": 2.416209476309227e-06,
+      "loss": 0.3421,
+      "step": 70540
+    },
+    {
+      "epoch": 17.59351620947631,
+      "grad_norm": 7.645537853240967,
+      "learning_rate": 2.4137157107231924e-06,
+      "loss": 0.3421,
+      "step": 70550
+    },
+    {
+      "epoch": 17.596009975062344,
+      "grad_norm": 7.081808090209961,
+      "learning_rate": 2.4112219451371573e-06,
+      "loss": 0.3081,
+      "step": 70560
+    },
+    {
+      "epoch": 17.59850374064838,
+      "grad_norm": 9.392647743225098,
+      "learning_rate": 2.4087281795511226e-06,
+      "loss": 0.322,
+      "step": 70570
+    },
+    {
+      "epoch": 17.600997506234414,
+      "grad_norm": 7.144293308258057,
+      "learning_rate": 2.4062344139650875e-06,
+      "loss": 0.3469,
+      "step": 70580
+    },
+    {
+      "epoch": 17.60349127182045,
+      "grad_norm": 4.9244489669799805,
+      "learning_rate": 2.4037406483790528e-06,
+      "loss": 0.2498,
+      "step": 70590
+    },
+    {
+      "epoch": 17.605985037406484,
+      "grad_norm": 7.220204830169678,
+      "learning_rate": 2.4012468827930177e-06,
+      "loss": 0.2706,
+      "step": 70600
+    },
+    {
+      "epoch": 17.60847880299252,
+      "grad_norm": 13.143561363220215,
+      "learning_rate": 2.398753117206983e-06,
+      "loss": 0.2832,
+      "step": 70610
+    },
+    {
+      "epoch": 17.610972568578553,
+      "grad_norm": 6.485062599182129,
+      "learning_rate": 2.396259351620948e-06,
+      "loss": 0.2975,
+      "step": 70620
+    },
+    {
+      "epoch": 17.61346633416459,
+      "grad_norm": 9.139724731445312,
+      "learning_rate": 2.3937655860349127e-06,
+      "loss": 0.3112,
+      "step": 70630
+    },
+    {
+      "epoch": 17.615960099750623,
+      "grad_norm": 9.77446460723877,
+      "learning_rate": 2.391271820448878e-06,
+      "loss": 0.3726,
+      "step": 70640
+    },
+    {
+      "epoch": 17.618453865336658,
+      "grad_norm": 8.16642951965332,
+      "learning_rate": 2.388778054862843e-06,
+      "loss": 0.3558,
+      "step": 70650
+    },
+    {
+      "epoch": 17.620947630922693,
+      "grad_norm": 8.044200897216797,
+      "learning_rate": 2.3862842892768082e-06,
+      "loss": 0.3816,
+      "step": 70660
+    },
+    {
+      "epoch": 17.623441396508728,
+      "grad_norm": 6.677599906921387,
+      "learning_rate": 2.383790523690773e-06,
+      "loss": 0.3632,
+      "step": 70670
+    },
+    {
+      "epoch": 17.625935162094763,
+      "grad_norm": 7.929919719696045,
+      "learning_rate": 2.3812967581047384e-06,
+      "loss": 0.3247,
+      "step": 70680
+    },
+    {
+      "epoch": 17.628428927680797,
+      "grad_norm": 5.376601219177246,
+      "learning_rate": 2.3788029925187033e-06,
+      "loss": 0.3405,
+      "step": 70690
+    },
+    {
+      "epoch": 17.630922693266832,
+      "grad_norm": 3.786297082901001,
+      "learning_rate": 2.3763092269326686e-06,
+      "loss": 0.2899,
+      "step": 70700
+    },
+    {
+      "epoch": 17.633416458852867,
+      "grad_norm": 6.230623722076416,
+      "learning_rate": 2.3738154613466335e-06,
+      "loss": 0.3461,
+      "step": 70710
+    },
+    {
+      "epoch": 17.635910224438902,
+      "grad_norm": 7.953461170196533,
+      "learning_rate": 2.3713216957605984e-06,
+      "loss": 0.2477,
+      "step": 70720
+    },
+    {
+      "epoch": 17.638403990024937,
+      "grad_norm": 6.597314357757568,
+      "learning_rate": 2.3688279301745637e-06,
+      "loss": 0.3126,
+      "step": 70730
+    },
+    {
+      "epoch": 17.640897755610972,
+      "grad_norm": 10.526558876037598,
+      "learning_rate": 2.3663341645885286e-06,
+      "loss": 0.3603,
+      "step": 70740
+    },
+    {
+      "epoch": 17.643391521197007,
+      "grad_norm": 6.605552673339844,
+      "learning_rate": 2.363840399002494e-06,
+      "loss": 0.3097,
+      "step": 70750
+    },
+    {
+      "epoch": 17.64588528678304,
+      "grad_norm": 7.1740031242370605,
+      "learning_rate": 2.361346633416459e-06,
+      "loss": 0.3014,
+      "step": 70760
+    },
+    {
+      "epoch": 17.648379052369076,
+      "grad_norm": 9.406676292419434,
+      "learning_rate": 2.358852867830424e-06,
+      "loss": 0.3107,
+      "step": 70770
+    },
+    {
+      "epoch": 17.65087281795511,
+      "grad_norm": 8.029032707214355,
+      "learning_rate": 2.3563591022443894e-06,
+      "loss": 0.3425,
+      "step": 70780
+    },
+    {
+      "epoch": 17.653366583541146,
+      "grad_norm": 13.621152877807617,
+      "learning_rate": 2.3538653366583543e-06,
+      "loss": 0.4274,
+      "step": 70790
+    },
+    {
+      "epoch": 17.65586034912718,
+      "grad_norm": 7.198687553405762,
+      "learning_rate": 2.3513715710723196e-06,
+      "loss": 0.3595,
+      "step": 70800
+    },
+    {
+      "epoch": 17.658354114713216,
+      "grad_norm": 7.494908332824707,
+      "learning_rate": 2.3488778054862845e-06,
+      "loss": 0.2798,
+      "step": 70810
+    },
+    {
+      "epoch": 17.66084788029925,
+      "grad_norm": 10.102776527404785,
+      "learning_rate": 2.34638403990025e-06,
+      "loss": 0.3385,
+      "step": 70820
+    },
+    {
+      "epoch": 17.663341645885286,
+      "grad_norm": 7.839235782623291,
+      "learning_rate": 2.3438902743142147e-06,
+      "loss": 0.3029,
+      "step": 70830
+    },
+    {
+      "epoch": 17.66583541147132,
+      "grad_norm": 7.807239532470703,
+      "learning_rate": 2.34139650872818e-06,
+      "loss": 0.2806,
+      "step": 70840
+    },
+    {
+      "epoch": 17.668329177057355,
+      "grad_norm": 7.2098069190979,
+      "learning_rate": 2.338902743142145e-06,
+      "loss": 0.3324,
+      "step": 70850
+    },
+    {
+      "epoch": 17.67082294264339,
+      "grad_norm": 12.253145217895508,
+      "learning_rate": 2.3364089775561098e-06,
+      "loss": 0.3326,
+      "step": 70860
+    },
+    {
+      "epoch": 17.673316708229425,
+      "grad_norm": 6.79581880569458,
+      "learning_rate": 2.333915211970075e-06,
+      "loss": 0.3108,
+      "step": 70870
+    },
+    {
+      "epoch": 17.67581047381546,
+      "grad_norm": 6.1217145919799805,
+      "learning_rate": 2.33142144638404e-06,
+      "loss": 0.2697,
+      "step": 70880
+    },
+    {
+      "epoch": 17.678304239401495,
+      "grad_norm": 5.935173034667969,
+      "learning_rate": 2.3289276807980053e-06,
+      "loss": 0.3026,
+      "step": 70890
+    },
+    {
+      "epoch": 17.68079800498753,
+      "grad_norm": 10.652816772460938,
+      "learning_rate": 2.32643391521197e-06,
+      "loss": 0.2878,
+      "step": 70900
+    },
+    {
+      "epoch": 17.683291770573565,
+      "grad_norm": 9.911921501159668,
+      "learning_rate": 2.3239401496259355e-06,
+      "loss": 0.305,
+      "step": 70910
+    },
+    {
+      "epoch": 17.6857855361596,
+      "grad_norm": 9.336223602294922,
+      "learning_rate": 2.3214463840399004e-06,
+      "loss": 0.3035,
+      "step": 70920
+    },
+    {
+      "epoch": 17.688279301745634,
+      "grad_norm": 10.193656921386719,
+      "learning_rate": 2.3189526184538657e-06,
+      "loss": 0.2473,
+      "step": 70930
+    },
+    {
+      "epoch": 17.69077306733167,
+      "grad_norm": 11.795038223266602,
+      "learning_rate": 2.3164588528678306e-06,
+      "loss": 0.3295,
+      "step": 70940
+    },
+    {
+      "epoch": 17.693266832917704,
+      "grad_norm": 10.375980377197266,
+      "learning_rate": 2.3139650872817955e-06,
+      "loss": 0.3085,
+      "step": 70950
+    },
+    {
+      "epoch": 17.69576059850374,
+      "grad_norm": 6.3128533363342285,
+      "learning_rate": 2.3114713216957608e-06,
+      "loss": 0.282,
+      "step": 70960
+    },
+    {
+      "epoch": 17.698254364089777,
+      "grad_norm": 5.7575907707214355,
+      "learning_rate": 2.3089775561097256e-06,
+      "loss": 0.3209,
+      "step": 70970
+    },
+    {
+      "epoch": 17.70074812967581,
+      "grad_norm": 8.107242584228516,
+      "learning_rate": 2.306483790523691e-06,
+      "loss": 0.3082,
+      "step": 70980
+    },
+    {
+      "epoch": 17.703241895261847,
+      "grad_norm": 9.448348999023438,
+      "learning_rate": 2.303990024937656e-06,
+      "loss": 0.3068,
+      "step": 70990
+    },
+    {
+      "epoch": 17.705735660847882,
+      "grad_norm": 15.375732421875,
+      "learning_rate": 2.301496259351621e-06,
+      "loss": 0.3056,
+      "step": 71000
+    },
+    {
+      "epoch": 17.708229426433917,
+      "grad_norm": 10.136244773864746,
+      "learning_rate": 2.299002493765586e-06,
+      "loss": 0.297,
+      "step": 71010
+    },
+    {
+      "epoch": 17.71072319201995,
+      "grad_norm": 7.518948078155518,
+      "learning_rate": 2.2965087281795514e-06,
+      "loss": 0.3659,
+      "step": 71020
+    },
+    {
+      "epoch": 17.713216957605987,
+      "grad_norm": 8.733294486999512,
+      "learning_rate": 2.2940149625935167e-06,
+      "loss": 0.3045,
+      "step": 71030
+    },
+    {
+      "epoch": 17.71571072319202,
+      "grad_norm": 11.162075996398926,
+      "learning_rate": 2.2915211970074815e-06,
+      "loss": 0.3342,
+      "step": 71040
+    },
+    {
+      "epoch": 17.718204488778056,
+      "grad_norm": 7.541927814483643,
+      "learning_rate": 2.289027431421447e-06,
+      "loss": 0.2797,
+      "step": 71050
+    },
+    {
+      "epoch": 17.72069825436409,
+      "grad_norm": 9.148726463317871,
+      "learning_rate": 2.2865336658354117e-06,
+      "loss": 0.2668,
+      "step": 71060
+    },
+    {
+      "epoch": 17.723192019950126,
+      "grad_norm": 10.433960914611816,
+      "learning_rate": 2.2840399002493766e-06,
+      "loss": 0.3384,
+      "step": 71070
+    },
+    {
+      "epoch": 17.72568578553616,
+      "grad_norm": 10.94124984741211,
+      "learning_rate": 2.281546134663342e-06,
+      "loss": 0.2648,
+      "step": 71080
+    },
+    {
+      "epoch": 17.728179551122196,
+      "grad_norm": 10.347514152526855,
+      "learning_rate": 2.279052369077307e-06,
+      "loss": 0.3198,
+      "step": 71090
+    },
+    {
+      "epoch": 17.73067331670823,
+      "grad_norm": 9.887333869934082,
+      "learning_rate": 2.276558603491272e-06,
+      "loss": 0.3412,
+      "step": 71100
+    },
+    {
+      "epoch": 17.733167082294266,
+      "grad_norm": 11.739676475524902,
+      "learning_rate": 2.274064837905237e-06,
+      "loss": 0.3266,
+      "step": 71110
+    },
+    {
+      "epoch": 17.7356608478803,
+      "grad_norm": 8.010821342468262,
+      "learning_rate": 2.2715710723192023e-06,
+      "loss": 0.2687,
+      "step": 71120
+    },
+    {
+      "epoch": 17.738154613466335,
+      "grad_norm": 7.8744730949401855,
+      "learning_rate": 2.2690773067331672e-06,
+      "loss": 0.3463,
+      "step": 71130
+    },
+    {
+      "epoch": 17.74064837905237,
+      "grad_norm": 6.5923237800598145,
+      "learning_rate": 2.2665835411471325e-06,
+      "loss": 0.3428,
+      "step": 71140
+    },
+    {
+      "epoch": 17.743142144638405,
+      "grad_norm": 12.174713134765625,
+      "learning_rate": 2.2640897755610974e-06,
+      "loss": 0.3093,
+      "step": 71150
+    },
+    {
+      "epoch": 17.74563591022444,
+      "grad_norm": 7.997066020965576,
+      "learning_rate": 2.2615960099750627e-06,
+      "loss": 0.3085,
+      "step": 71160
+    },
+    {
+      "epoch": 17.748129675810475,
+      "grad_norm": 10.162776947021484,
+      "learning_rate": 2.2591022443890276e-06,
+      "loss": 0.3193,
+      "step": 71170
+    },
+    {
+      "epoch": 17.75062344139651,
+      "grad_norm": 8.954973220825195,
+      "learning_rate": 2.2566084788029925e-06,
+      "loss": 0.3455,
+      "step": 71180
+    },
+    {
+      "epoch": 17.753117206982544,
+      "grad_norm": 5.909659385681152,
+      "learning_rate": 2.254114713216958e-06,
+      "loss": 0.3149,
+      "step": 71190
+    },
+    {
+      "epoch": 17.75561097256858,
+      "grad_norm": 7.491812229156494,
+      "learning_rate": 2.2516209476309227e-06,
+      "loss": 0.2545,
+      "step": 71200
+    },
+    {
+      "epoch": 17.758104738154614,
+      "grad_norm": 7.293992519378662,
+      "learning_rate": 2.249127182044888e-06,
+      "loss": 0.2647,
+      "step": 71210
+    },
+    {
+      "epoch": 17.76059850374065,
+      "grad_norm": 8.303019523620605,
+      "learning_rate": 2.246633416458853e-06,
+      "loss": 0.3833,
+      "step": 71220
+    },
+    {
+      "epoch": 17.763092269326684,
+      "grad_norm": 7.720460414886475,
+      "learning_rate": 2.244139650872818e-06,
+      "loss": 0.2887,
+      "step": 71230
+    },
+    {
+      "epoch": 17.76558603491272,
+      "grad_norm": 10.40861988067627,
+      "learning_rate": 2.241645885286783e-06,
+      "loss": 0.3591,
+      "step": 71240
+    },
+    {
+      "epoch": 17.768079800498754,
+      "grad_norm": 12.28356647491455,
+      "learning_rate": 2.2391521197007484e-06,
+      "loss": 0.3628,
+      "step": 71250
+    },
+    {
+      "epoch": 17.77057356608479,
+      "grad_norm": 10.298823356628418,
+      "learning_rate": 2.2366583541147133e-06,
+      "loss": 0.3126,
+      "step": 71260
+    },
+    {
+      "epoch": 17.773067331670823,
+      "grad_norm": 12.331887245178223,
+      "learning_rate": 2.2341645885286786e-06,
+      "loss": 0.3214,
+      "step": 71270
+    },
+    {
+      "epoch": 17.77556109725686,
+      "grad_norm": 5.967679977416992,
+      "learning_rate": 2.231670822942644e-06,
+      "loss": 0.3022,
+      "step": 71280
+    },
+    {
+      "epoch": 17.778054862842893,
+      "grad_norm": 8.2744722366333,
+      "learning_rate": 2.2291770573566088e-06,
+      "loss": 0.3448,
+      "step": 71290
+    },
+    {
+      "epoch": 17.780548628428928,
+      "grad_norm": 7.625812530517578,
+      "learning_rate": 2.2266832917705737e-06,
+      "loss": 0.2958,
+      "step": 71300
+    },
+    {
+      "epoch": 17.783042394014963,
+      "grad_norm": 9.751514434814453,
+      "learning_rate": 2.224189526184539e-06,
+      "loss": 0.3786,
+      "step": 71310
+    },
+    {
+      "epoch": 17.785536159600998,
+      "grad_norm": 8.716211318969727,
+      "learning_rate": 2.221695760598504e-06,
+      "loss": 0.2933,
+      "step": 71320
+    },
+    {
+      "epoch": 17.788029925187033,
+      "grad_norm": 11.199061393737793,
+      "learning_rate": 2.219201995012469e-06,
+      "loss": 0.3668,
+      "step": 71330
+    },
+    {
+      "epoch": 17.790523690773068,
+      "grad_norm": 14.230875015258789,
+      "learning_rate": 2.216708229426434e-06,
+      "loss": 0.3918,
+      "step": 71340
+    },
+    {
+      "epoch": 17.793017456359102,
+      "grad_norm": 7.556332111358643,
+      "learning_rate": 2.2142144638403994e-06,
+      "loss": 0.2992,
+      "step": 71350
+    },
+    {
+      "epoch": 17.795511221945137,
+      "grad_norm": 10.51162052154541,
+      "learning_rate": 2.2117206982543643e-06,
+      "loss": 0.2895,
+      "step": 71360
+    },
+    {
+      "epoch": 17.798004987531172,
+      "grad_norm": 8.932967185974121,
+      "learning_rate": 2.2092269326683296e-06,
+      "loss": 0.3283,
+      "step": 71370
+    },
+    {
+      "epoch": 17.800498753117207,
+      "grad_norm": 6.184909343719482,
+      "learning_rate": 2.2067331670822945e-06,
+      "loss": 0.3352,
+      "step": 71380
+    },
+    {
+      "epoch": 17.802992518703242,
+      "grad_norm": 16.87145233154297,
+      "learning_rate": 2.2042394014962593e-06,
+      "loss": 0.3586,
+      "step": 71390
+    },
+    {
+      "epoch": 17.805486284289277,
+      "grad_norm": 9.270705223083496,
+      "learning_rate": 2.2017456359102246e-06,
+      "loss": 0.2885,
+      "step": 71400
+    },
+    {
+      "epoch": 17.80798004987531,
+      "grad_norm": 8.402549743652344,
+      "learning_rate": 2.1992518703241895e-06,
+      "loss": 0.2626,
+      "step": 71410
+    },
+    {
+      "epoch": 17.810473815461346,
+      "grad_norm": 5.3837738037109375,
+      "learning_rate": 2.196758104738155e-06,
+      "loss": 0.3185,
+      "step": 71420
+    },
+    {
+      "epoch": 17.81296758104738,
+      "grad_norm": 9.095513343811035,
+      "learning_rate": 2.1942643391521197e-06,
+      "loss": 0.3384,
+      "step": 71430
+    },
+    {
+      "epoch": 17.815461346633416,
+      "grad_norm": 15.026154518127441,
+      "learning_rate": 2.191770573566085e-06,
+      "loss": 0.3384,
+      "step": 71440
+    },
+    {
+      "epoch": 17.81795511221945,
+      "grad_norm": 9.241385459899902,
+      "learning_rate": 2.18927680798005e-06,
+      "loss": 0.3365,
+      "step": 71450
+    },
+    {
+      "epoch": 17.820448877805486,
+      "grad_norm": 8.417593002319336,
+      "learning_rate": 2.1867830423940152e-06,
+      "loss": 0.3111,
+      "step": 71460
+    },
+    {
+      "epoch": 17.82294264339152,
+      "grad_norm": 6.55775260925293,
+      "learning_rate": 2.18428927680798e-06,
+      "loss": 0.3183,
+      "step": 71470
+    },
+    {
+      "epoch": 17.825436408977556,
+      "grad_norm": 5.751813888549805,
+      "learning_rate": 2.181795511221945e-06,
+      "loss": 0.248,
+      "step": 71480
+    },
+    {
+      "epoch": 17.82793017456359,
+      "grad_norm": 8.865701675415039,
+      "learning_rate": 2.1793017456359103e-06,
+      "loss": 0.3105,
+      "step": 71490
+    },
+    {
+      "epoch": 17.830423940149625,
+      "grad_norm": 8.64411449432373,
+      "learning_rate": 2.176807980049875e-06,
+      "loss": 0.3189,
+      "step": 71500
+    },
+    {
+      "epoch": 17.83291770573566,
+      "grad_norm": 10.088348388671875,
+      "learning_rate": 2.1743142144638405e-06,
+      "loss": 0.3171,
+      "step": 71510
+    },
+    {
+      "epoch": 17.835411471321695,
+      "grad_norm": 8.263297080993652,
+      "learning_rate": 2.171820448877806e-06,
+      "loss": 0.2926,
+      "step": 71520
+    },
+    {
+      "epoch": 17.83790523690773,
+      "grad_norm": 14.566165924072266,
+      "learning_rate": 2.1693266832917707e-06,
+      "loss": 0.3174,
+      "step": 71530
+    },
+    {
+      "epoch": 17.840399002493765,
+      "grad_norm": 7.461530685424805,
+      "learning_rate": 2.166832917705736e-06,
+      "loss": 0.3255,
+      "step": 71540
+    },
+    {
+      "epoch": 17.8428927680798,
+      "grad_norm": 5.400354385375977,
+      "learning_rate": 2.164339152119701e-06,
+      "loss": 0.263,
+      "step": 71550
+    },
+    {
+      "epoch": 17.845386533665835,
+      "grad_norm": 9.746865272521973,
+      "learning_rate": 2.1618453865336662e-06,
+      "loss": 0.3657,
+      "step": 71560
+    },
+    {
+      "epoch": 17.84788029925187,
+      "grad_norm": 7.5796380043029785,
+      "learning_rate": 2.159351620947631e-06,
+      "loss": 0.3111,
+      "step": 71570
+    },
+    {
+      "epoch": 17.850374064837904,
+      "grad_norm": 11.697222709655762,
+      "learning_rate": 2.1568578553615964e-06,
+      "loss": 0.2908,
+      "step": 71580
+    },
+    {
+      "epoch": 17.85286783042394,
+      "grad_norm": 6.474806308746338,
+      "learning_rate": 2.1543640897755613e-06,
+      "loss": 0.285,
+      "step": 71590
+    },
+    {
+      "epoch": 17.855361596009974,
+      "grad_norm": 9.604869842529297,
+      "learning_rate": 2.1518703241895266e-06,
+      "loss": 0.3314,
+      "step": 71600
+    },
+    {
+      "epoch": 17.85785536159601,
+      "grad_norm": 7.376259803771973,
+      "learning_rate": 2.1493765586034915e-06,
+      "loss": 0.257,
+      "step": 71610
+    },
+    {
+      "epoch": 17.860349127182044,
+      "grad_norm": 8.758426666259766,
+      "learning_rate": 2.1468827930174564e-06,
+      "loss": 0.3672,
+      "step": 71620
+    },
+    {
+      "epoch": 17.86284289276808,
+      "grad_norm": 8.611882209777832,
+      "learning_rate": 2.1443890274314217e-06,
+      "loss": 0.3306,
+      "step": 71630
+    },
+    {
+      "epoch": 17.865336658354114,
+      "grad_norm": 10.081382751464844,
+      "learning_rate": 2.1418952618453866e-06,
+      "loss": 0.3223,
+      "step": 71640
+    },
+    {
+      "epoch": 17.86783042394015,
+      "grad_norm": 11.713033676147461,
+      "learning_rate": 2.139401496259352e-06,
+      "loss": 0.3424,
+      "step": 71650
+    },
+    {
+      "epoch": 17.870324189526183,
+      "grad_norm": 8.003156661987305,
+      "learning_rate": 2.1369077306733168e-06,
+      "loss": 0.327,
+      "step": 71660
+    },
+    {
+      "epoch": 17.872817955112218,
+      "grad_norm": 9.68770694732666,
+      "learning_rate": 2.134413965087282e-06,
+      "loss": 0.3158,
+      "step": 71670
+    },
+    {
+      "epoch": 17.875311720698253,
+      "grad_norm": 4.524538516998291,
+      "learning_rate": 2.131920199501247e-06,
+      "loss": 0.3178,
+      "step": 71680
+    },
+    {
+      "epoch": 17.877805486284288,
+      "grad_norm": 8.126871109008789,
+      "learning_rate": 2.1294264339152123e-06,
+      "loss": 0.3448,
+      "step": 71690
+    },
+    {
+      "epoch": 17.880299251870323,
+      "grad_norm": 8.234221458435059,
+      "learning_rate": 2.126932668329177e-06,
+      "loss": 0.3894,
+      "step": 71700
+    },
+    {
+      "epoch": 17.882793017456358,
+      "grad_norm": 8.27507209777832,
+      "learning_rate": 2.124438902743142e-06,
+      "loss": 0.3331,
+      "step": 71710
+    },
+    {
+      "epoch": 17.885286783042392,
+      "grad_norm": 9.675071716308594,
+      "learning_rate": 2.1219451371571074e-06,
+      "loss": 0.3486,
+      "step": 71720
+    },
+    {
+      "epoch": 17.887780548628427,
+      "grad_norm": 7.789792537689209,
+      "learning_rate": 2.1194513715710722e-06,
+      "loss": 0.3582,
+      "step": 71730
+    },
+    {
+      "epoch": 17.890274314214462,
+      "grad_norm": 17.104684829711914,
+      "learning_rate": 2.1169576059850376e-06,
+      "loss": 0.3253,
+      "step": 71740
+    },
+    {
+      "epoch": 17.892768079800497,
+      "grad_norm": 10.092325210571289,
+      "learning_rate": 2.1144638403990024e-06,
+      "loss": 0.3024,
+      "step": 71750
+    },
+    {
+      "epoch": 17.895261845386532,
+      "grad_norm": 7.519723415374756,
+      "learning_rate": 2.1119700748129678e-06,
+      "loss": 0.3506,
+      "step": 71760
+    },
+    {
+      "epoch": 17.897755610972567,
+      "grad_norm": 9.717748641967773,
+      "learning_rate": 2.109476309226933e-06,
+      "loss": 0.2639,
+      "step": 71770
+    },
+    {
+      "epoch": 17.900249376558605,
+      "grad_norm": 7.854363441467285,
+      "learning_rate": 2.106982543640898e-06,
+      "loss": 0.3382,
+      "step": 71780
+    },
+    {
+      "epoch": 17.902743142144637,
+      "grad_norm": 6.1837158203125,
+      "learning_rate": 2.1044887780548633e-06,
+      "loss": 0.328,
+      "step": 71790
+    },
+    {
+      "epoch": 17.905236907730675,
+      "grad_norm": 9.821382522583008,
+      "learning_rate": 2.101995012468828e-06,
+      "loss": 0.3765,
+      "step": 71800
+    },
+    {
+      "epoch": 17.90773067331671,
+      "grad_norm": 8.692037582397461,
+      "learning_rate": 2.0995012468827935e-06,
+      "loss": 0.3051,
+      "step": 71810
+    },
+    {
+      "epoch": 17.910224438902745,
+      "grad_norm": 8.943056106567383,
+      "learning_rate": 2.0970074812967583e-06,
+      "loss": 0.339,
+      "step": 71820
+    },
+    {
+      "epoch": 17.91271820448878,
+      "grad_norm": 8.475810050964355,
+      "learning_rate": 2.0945137157107232e-06,
+      "loss": 0.3219,
+      "step": 71830
+    },
+    {
+      "epoch": 17.915211970074814,
+      "grad_norm": 13.694457054138184,
+      "learning_rate": 2.0920199501246885e-06,
+      "loss": 0.2674,
+      "step": 71840
+    },
+    {
+      "epoch": 17.91770573566085,
+      "grad_norm": 15.57154655456543,
+      "learning_rate": 2.0895261845386534e-06,
+      "loss": 0.3422,
+      "step": 71850
+    },
+    {
+      "epoch": 17.920199501246884,
+      "grad_norm": 7.596667289733887,
+      "learning_rate": 2.0870324189526187e-06,
+      "loss": 0.352,
+      "step": 71860
+    },
+    {
+      "epoch": 17.92269326683292,
+      "grad_norm": 7.299078941345215,
+      "learning_rate": 2.0845386533665836e-06,
+      "loss": 0.3473,
+      "step": 71870
+    },
+    {
+      "epoch": 17.925187032418954,
+      "grad_norm": 8.171175003051758,
+      "learning_rate": 2.082044887780549e-06,
+      "loss": 0.3868,
+      "step": 71880
+    },
+    {
+      "epoch": 17.92768079800499,
+      "grad_norm": 27.5074405670166,
+      "learning_rate": 2.079551122194514e-06,
+      "loss": 0.3894,
+      "step": 71890
+    },
+    {
+      "epoch": 17.930174563591024,
+      "grad_norm": 10.98017692565918,
+      "learning_rate": 2.077057356608479e-06,
+      "loss": 0.2976,
+      "step": 71900
+    },
+    {
+      "epoch": 17.93266832917706,
+      "grad_norm": 10.127460479736328,
+      "learning_rate": 2.074563591022444e-06,
+      "loss": 0.3238,
+      "step": 71910
+    },
+    {
+      "epoch": 17.935162094763093,
+      "grad_norm": 6.212870121002197,
+      "learning_rate": 2.0720698254364093e-06,
+      "loss": 0.3211,
+      "step": 71920
+    },
+    {
+      "epoch": 17.93765586034913,
+      "grad_norm": 8.205334663391113,
+      "learning_rate": 2.069576059850374e-06,
+      "loss": 0.3003,
+      "step": 71930
+    },
+    {
+      "epoch": 17.940149625935163,
+      "grad_norm": 8.863457679748535,
+      "learning_rate": 2.067082294264339e-06,
+      "loss": 0.3392,
+      "step": 71940
+    },
+    {
+      "epoch": 17.942643391521198,
+      "grad_norm": 8.939859390258789,
+      "learning_rate": 2.0645885286783044e-06,
+      "loss": 0.2978,
+      "step": 71950
+    },
+    {
+      "epoch": 17.945137157107233,
+      "grad_norm": 12.467900276184082,
+      "learning_rate": 2.0620947630922693e-06,
+      "loss": 0.3426,
+      "step": 71960
+    },
+    {
+      "epoch": 17.947630922693268,
+      "grad_norm": 9.684351921081543,
+      "learning_rate": 2.0596009975062346e-06,
+      "loss": 0.3648,
+      "step": 71970
+    },
+    {
+      "epoch": 17.950124688279303,
+      "grad_norm": 7.206076622009277,
+      "learning_rate": 2.0571072319201995e-06,
+      "loss": 0.3311,
+      "step": 71980
+    },
+    {
+      "epoch": 17.952618453865338,
+      "grad_norm": 8.474637985229492,
+      "learning_rate": 2.054613466334165e-06,
+      "loss": 0.2923,
+      "step": 71990
+    },
+    {
+      "epoch": 17.955112219451372,
+      "grad_norm": 5.08350944519043,
+      "learning_rate": 2.0521197007481297e-06,
+      "loss": 0.2908,
+      "step": 72000
+    },
+    {
+      "epoch": 17.957605985037407,
+      "grad_norm": 8.969040870666504,
+      "learning_rate": 2.049625935162095e-06,
+      "loss": 0.3462,
+      "step": 72010
+    },
+    {
+      "epoch": 17.960099750623442,
+      "grad_norm": 6.145038604736328,
+      "learning_rate": 2.0471321695760603e-06,
+      "loss": 0.3551,
+      "step": 72020
+    },
+    {
+      "epoch": 17.962593516209477,
+      "grad_norm": 6.762228012084961,
+      "learning_rate": 2.044638403990025e-06,
+      "loss": 0.3054,
+      "step": 72030
+    },
+    {
+      "epoch": 17.965087281795512,
+      "grad_norm": 8.712302207946777,
+      "learning_rate": 2.0421446384039905e-06,
+      "loss": 0.3249,
+      "step": 72040
+    },
+    {
+      "epoch": 17.967581047381547,
+      "grad_norm": 8.052441596984863,
+      "learning_rate": 2.0396508728179554e-06,
+      "loss": 0.2844,
+      "step": 72050
+    },
+    {
+      "epoch": 17.97007481296758,
+      "grad_norm": 11.43187427520752,
+      "learning_rate": 2.0371571072319203e-06,
+      "loss": 0.3358,
+      "step": 72060
+    },
+    {
+      "epoch": 17.972568578553616,
+      "grad_norm": 11.102864265441895,
+      "learning_rate": 2.0346633416458856e-06,
+      "loss": 0.3596,
+      "step": 72070
+    },
+    {
+      "epoch": 17.97506234413965,
+      "grad_norm": 9.357625961303711,
+      "learning_rate": 2.0321695760598505e-06,
+      "loss": 0.4051,
+      "step": 72080
+    },
+    {
+      "epoch": 17.977556109725686,
+      "grad_norm": 10.954850196838379,
+      "learning_rate": 2.0296758104738158e-06,
+      "loss": 0.3903,
+      "step": 72090
+    },
+    {
+      "epoch": 17.98004987531172,
+      "grad_norm": 7.4843645095825195,
+      "learning_rate": 2.0271820448877807e-06,
+      "loss": 0.3329,
+      "step": 72100
+    },
+    {
+      "epoch": 17.982543640897756,
+      "grad_norm": 6.2788286209106445,
+      "learning_rate": 2.024688279301746e-06,
+      "loss": 0.3574,
+      "step": 72110
+    },
+    {
+      "epoch": 17.98503740648379,
+      "grad_norm": 8.560807228088379,
+      "learning_rate": 2.022194513715711e-06,
+      "loss": 0.384,
+      "step": 72120
+    },
+    {
+      "epoch": 17.987531172069826,
+      "grad_norm": 6.629812717437744,
+      "learning_rate": 2.019700748129676e-06,
+      "loss": 0.3336,
+      "step": 72130
+    },
+    {
+      "epoch": 17.99002493765586,
+      "grad_norm": 7.1960859298706055,
+      "learning_rate": 2.017206982543641e-06,
+      "loss": 0.3042,
+      "step": 72140
+    },
+    {
+      "epoch": 17.992518703241895,
+      "grad_norm": 12.617613792419434,
+      "learning_rate": 2.014713216957606e-06,
+      "loss": 0.303,
+      "step": 72150
+    },
+    {
+      "epoch": 17.99501246882793,
+      "grad_norm": 8.579712867736816,
+      "learning_rate": 2.0122194513715712e-06,
+      "loss": 0.3387,
+      "step": 72160
+    },
+    {
+      "epoch": 17.997506234413965,
+      "grad_norm": 7.062646865844727,
+      "learning_rate": 2.009725685785536e-06,
+      "loss": 0.3563,
+      "step": 72170
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 9.439546585083008,
+      "learning_rate": 2.0072319201995014e-06,
+      "loss": 0.3308,
+      "step": 72180
+    },
+    {
+      "epoch": 18.0,
+      "eval_loss": 0.4180074632167816,
+      "eval_runtime": 60.175,
+      "eval_samples_per_second": 16.668,
+      "eval_steps_per_second": 16.668,
+      "step": 72180
+    },
+    {
+      "epoch": 18.002493765586035,
+      "grad_norm": 12.682416915893555,
+      "learning_rate": 2.0047381546134663e-06,
+      "loss": 0.315,
+      "step": 72190
+    },
+    {
+      "epoch": 18.00498753117207,
+      "grad_norm": 6.629142761230469,
+      "learning_rate": 2.0022443890274316e-06,
+      "loss": 0.2881,
+      "step": 72200
+    },
+    {
+      "epoch": 18.007481296758105,
+      "grad_norm": 11.43194580078125,
+      "learning_rate": 1.9997506234413965e-06,
+      "loss": 0.3674,
+      "step": 72210
+    },
+    {
+      "epoch": 18.00997506234414,
+      "grad_norm": 10.104514122009277,
+      "learning_rate": 1.997256857855362e-06,
+      "loss": 0.3005,
+      "step": 72220
+    },
+    {
+      "epoch": 18.012468827930174,
+      "grad_norm": 10.51015853881836,
+      "learning_rate": 1.9947630922693267e-06,
+      "loss": 0.3445,
+      "step": 72230
+    },
+    {
+      "epoch": 18.01496259351621,
+      "grad_norm": 7.913299083709717,
+      "learning_rate": 1.9922693266832916e-06,
+      "loss": 0.3416,
+      "step": 72240
+    },
+    {
+      "epoch": 18.017456359102244,
+      "grad_norm": 10.231574058532715,
+      "learning_rate": 1.989775561097257e-06,
+      "loss": 0.2745,
+      "step": 72250
+    },
+    {
+      "epoch": 18.01995012468828,
+      "grad_norm": 7.487124919891357,
+      "learning_rate": 1.9872817955112222e-06,
+      "loss": 0.3448,
+      "step": 72260
+    },
+    {
+      "epoch": 18.022443890274314,
+      "grad_norm": 8.64004898071289,
+      "learning_rate": 1.984788029925187e-06,
+      "loss": 0.3579,
+      "step": 72270
+    },
+    {
+      "epoch": 18.02493765586035,
+      "grad_norm": 10.440835952758789,
+      "learning_rate": 1.9822942643391524e-06,
+      "loss": 0.2995,
+      "step": 72280
+    },
+    {
+      "epoch": 18.027431421446384,
+      "grad_norm": 7.366189956665039,
+      "learning_rate": 1.9798004987531173e-06,
+      "loss": 0.276,
+      "step": 72290
+    },
+    {
+      "epoch": 18.02992518703242,
+      "grad_norm": 3.2743232250213623,
+      "learning_rate": 1.9773067331670826e-06,
+      "loss": 0.2778,
+      "step": 72300
+    },
+    {
+      "epoch": 18.032418952618453,
+      "grad_norm": 5.090819358825684,
+      "learning_rate": 1.9748129675810475e-06,
+      "loss": 0.296,
+      "step": 72310
+    },
+    {
+      "epoch": 18.034912718204488,
+      "grad_norm": 9.35080337524414,
+      "learning_rate": 1.972319201995013e-06,
+      "loss": 0.2632,
+      "step": 72320
+    },
+    {
+      "epoch": 18.037406483790523,
+      "grad_norm": 7.471287250518799,
+      "learning_rate": 1.9698254364089777e-06,
+      "loss": 0.3306,
+      "step": 72330
+    },
+    {
+      "epoch": 18.039900249376558,
+      "grad_norm": 8.039196014404297,
+      "learning_rate": 1.967331670822943e-06,
+      "loss": 0.3021,
+      "step": 72340
+    },
+    {
+      "epoch": 18.042394014962593,
+      "grad_norm": 8.003158569335938,
+      "learning_rate": 1.964837905236908e-06,
+      "loss": 0.3735,
+      "step": 72350
+    },
+    {
+      "epoch": 18.044887780548628,
+      "grad_norm": 7.758488178253174,
+      "learning_rate": 1.962344139650873e-06,
+      "loss": 0.3305,
+      "step": 72360
+    },
+    {
+      "epoch": 18.047381546134662,
+      "grad_norm": 9.10808277130127,
+      "learning_rate": 1.959850374064838e-06,
+      "loss": 0.3393,
+      "step": 72370
+    },
+    {
+      "epoch": 18.049875311720697,
+      "grad_norm": 6.801268100738525,
+      "learning_rate": 1.957356608478803e-06,
+      "loss": 0.2978,
+      "step": 72380
+    },
+    {
+      "epoch": 18.052369077306732,
+      "grad_norm": 7.056491851806641,
+      "learning_rate": 1.9548628428927683e-06,
+      "loss": 0.3622,
+      "step": 72390
+    },
+    {
+      "epoch": 18.054862842892767,
+      "grad_norm": 8.197697639465332,
+      "learning_rate": 1.952369077306733e-06,
+      "loss": 0.3272,
+      "step": 72400
+    },
+    {
+      "epoch": 18.057356608478802,
+      "grad_norm": 9.100240707397461,
+      "learning_rate": 1.9498753117206985e-06,
+      "loss": 0.2909,
+      "step": 72410
+    },
+    {
+      "epoch": 18.059850374064837,
+      "grad_norm": 9.016037940979004,
+      "learning_rate": 1.9473815461346634e-06,
+      "loss": 0.3508,
+      "step": 72420
+    },
+    {
+      "epoch": 18.06234413965087,
+      "grad_norm": 11.260120391845703,
+      "learning_rate": 1.9448877805486287e-06,
+      "loss": 0.3345,
+      "step": 72430
+    },
+    {
+      "epoch": 18.064837905236907,
+      "grad_norm": 8.461393356323242,
+      "learning_rate": 1.9423940149625936e-06,
+      "loss": 0.2626,
+      "step": 72440
+    },
+    {
+      "epoch": 18.06733167082294,
+      "grad_norm": 7.152003765106201,
+      "learning_rate": 1.939900249376559e-06,
+      "loss": 0.3907,
+      "step": 72450
+    },
+    {
+      "epoch": 18.069825436408976,
+      "grad_norm": 7.978116512298584,
+      "learning_rate": 1.9374064837905238e-06,
+      "loss": 0.3898,
+      "step": 72460
+    },
+    {
+      "epoch": 18.07231920199501,
+      "grad_norm": 6.364957332611084,
+      "learning_rate": 1.9349127182044886e-06,
+      "loss": 0.2275,
+      "step": 72470
+    },
+    {
+      "epoch": 18.074812967581046,
+      "grad_norm": 8.800970077514648,
+      "learning_rate": 1.932418952618454e-06,
+      "loss": 0.253,
+      "step": 72480
+    },
+    {
+      "epoch": 18.07730673316708,
+      "grad_norm": 5.478081226348877,
+      "learning_rate": 1.929925187032419e-06,
+      "loss": 0.3061,
+      "step": 72490
+    },
+    {
+      "epoch": 18.079800498753116,
+      "grad_norm": 6.752138614654541,
+      "learning_rate": 1.927431421446384e-06,
+      "loss": 0.3113,
+      "step": 72500
+    },
+    {
+      "epoch": 18.08229426433915,
+      "grad_norm": 10.513050079345703,
+      "learning_rate": 1.9249376558603495e-06,
+      "loss": 0.2699,
+      "step": 72510
+    },
+    {
+      "epoch": 18.084788029925186,
+      "grad_norm": 8.9693021774292,
+      "learning_rate": 1.9224438902743144e-06,
+      "loss": 0.3346,
+      "step": 72520
+    },
+    {
+      "epoch": 18.08728179551122,
+      "grad_norm": 7.213470458984375,
+      "learning_rate": 1.9199501246882797e-06,
+      "loss": 0.306,
+      "step": 72530
+    },
+    {
+      "epoch": 18.089775561097255,
+      "grad_norm": 9.323833465576172,
+      "learning_rate": 1.9174563591022445e-06,
+      "loss": 0.3102,
+      "step": 72540
+    },
+    {
+      "epoch": 18.09226932668329,
+      "grad_norm": 12.904609680175781,
+      "learning_rate": 1.91496259351621e-06,
+      "loss": 0.3915,
+      "step": 72550
+    },
+    {
+      "epoch": 18.094763092269325,
+      "grad_norm": 7.509303092956543,
+      "learning_rate": 1.9124688279301747e-06,
+      "loss": 0.2904,
+      "step": 72560
+    },
+    {
+      "epoch": 18.09725685785536,
+      "grad_norm": 7.649249076843262,
+      "learning_rate": 1.90997506234414e-06,
+      "loss": 0.2865,
+      "step": 72570
+    },
+    {
+      "epoch": 18.099750623441395,
+      "grad_norm": 11.983275413513184,
+      "learning_rate": 1.907481296758105e-06,
+      "loss": 0.3197,
+      "step": 72580
+    },
+    {
+      "epoch": 18.102244389027433,
+      "grad_norm": 10.603513717651367,
+      "learning_rate": 1.90498753117207e-06,
+      "loss": 0.3186,
+      "step": 72590
+    },
+    {
+      "epoch": 18.104738154613468,
+      "grad_norm": 8.808806419372559,
+      "learning_rate": 1.9024937655860351e-06,
+      "loss": 0.3585,
+      "step": 72600
+    },
+    {
+      "epoch": 18.107231920199503,
+      "grad_norm": 8.247468948364258,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.3955,
+      "step": 72610
+    },
+    {
+      "epoch": 18.109725685785538,
+      "grad_norm": 10.441139221191406,
+      "learning_rate": 1.8975062344139653e-06,
+      "loss": 0.355,
+      "step": 72620
+    },
+    {
+      "epoch": 18.112219451371573,
+      "grad_norm": 6.936132431030273,
+      "learning_rate": 1.8950124688279304e-06,
+      "loss": 0.3121,
+      "step": 72630
+    },
+    {
+      "epoch": 18.114713216957608,
+      "grad_norm": 7.062928676605225,
+      "learning_rate": 1.8925187032418953e-06,
+      "loss": 0.2723,
+      "step": 72640
+    },
+    {
+      "epoch": 18.117206982543642,
+      "grad_norm": 6.312658786773682,
+      "learning_rate": 1.8900249376558604e-06,
+      "loss": 0.2906,
+      "step": 72650
+    },
+    {
+      "epoch": 18.119700748129677,
+      "grad_norm": 8.172001838684082,
+      "learning_rate": 1.8877805486284289e-06,
+      "loss": 0.3375,
+      "step": 72660
+    },
+    {
+      "epoch": 18.122194513715712,
+      "grad_norm": 9.601921081542969,
+      "learning_rate": 1.885286783042394e-06,
+      "loss": 0.365,
+      "step": 72670
+    },
+    {
+      "epoch": 18.124688279301747,
+      "grad_norm": 7.265673637390137,
+      "learning_rate": 1.882793017456359e-06,
+      "loss": 0.2651,
+      "step": 72680
+    },
+    {
+      "epoch": 18.127182044887782,
+      "grad_norm": 9.785295486450195,
+      "learning_rate": 1.8802992518703244e-06,
+      "loss": 0.3108,
+      "step": 72690
+    },
+    {
+      "epoch": 18.129675810473817,
+      "grad_norm": 10.865586280822754,
+      "learning_rate": 1.8778054862842895e-06,
+      "loss": 0.2968,
+      "step": 72700
+    },
+    {
+      "epoch": 18.13216957605985,
+      "grad_norm": 5.767469882965088,
+      "learning_rate": 1.8753117206982546e-06,
+      "loss": 0.2573,
+      "step": 72710
+    },
+    {
+      "epoch": 18.134663341645886,
+      "grad_norm": 9.119926452636719,
+      "learning_rate": 1.8728179551122197e-06,
+      "loss": 0.336,
+      "step": 72720
+    },
+    {
+      "epoch": 18.13715710723192,
+      "grad_norm": 11.801755905151367,
+      "learning_rate": 1.8703241895261848e-06,
+      "loss": 0.3547,
+      "step": 72730
+    },
+    {
+      "epoch": 18.139650872817956,
+      "grad_norm": 8.99133014678955,
+      "learning_rate": 1.8678304239401499e-06,
+      "loss": 0.3888,
+      "step": 72740
+    },
+    {
+      "epoch": 18.14214463840399,
+      "grad_norm": 9.552180290222168,
+      "learning_rate": 1.865336658354115e-06,
+      "loss": 0.3176,
+      "step": 72750
+    },
+    {
+      "epoch": 18.144638403990026,
+      "grad_norm": 9.201723098754883,
+      "learning_rate": 1.86284289276808e-06,
+      "loss": 0.2666,
+      "step": 72760
+    },
+    {
+      "epoch": 18.14713216957606,
+      "grad_norm": 9.10981273651123,
+      "learning_rate": 1.8603491271820452e-06,
+      "loss": 0.2701,
+      "step": 72770
+    },
+    {
+      "epoch": 18.149625935162096,
+      "grad_norm": 5.472398281097412,
+      "learning_rate": 1.85785536159601e-06,
+      "loss": 0.3164,
+      "step": 72780
+    },
+    {
+      "epoch": 18.15211970074813,
+      "grad_norm": 10.77421760559082,
+      "learning_rate": 1.8553615960099752e-06,
+      "loss": 0.3434,
+      "step": 72790
+    },
+    {
+      "epoch": 18.154613466334165,
+      "grad_norm": 7.737188339233398,
+      "learning_rate": 1.8528678304239403e-06,
+      "loss": 0.2856,
+      "step": 72800
+    },
+    {
+      "epoch": 18.1571072319202,
+      "grad_norm": 10.025762557983398,
+      "learning_rate": 1.8503740648379054e-06,
+      "loss": 0.3094,
+      "step": 72810
+    },
+    {
+      "epoch": 18.159600997506235,
+      "grad_norm": 12.608803749084473,
+      "learning_rate": 1.8478802992518705e-06,
+      "loss": 0.3368,
+      "step": 72820
+    },
+    {
+      "epoch": 18.16209476309227,
+      "grad_norm": 17.044055938720703,
+      "learning_rate": 1.8453865336658356e-06,
+      "loss": 0.3715,
+      "step": 72830
+    },
+    {
+      "epoch": 18.164588528678305,
+      "grad_norm": 8.355241775512695,
+      "learning_rate": 1.8428927680798007e-06,
+      "loss": 0.329,
+      "step": 72840
+    },
+    {
+      "epoch": 18.16708229426434,
+      "grad_norm": 8.667624473571777,
+      "learning_rate": 1.8403990024937658e-06,
+      "loss": 0.3585,
+      "step": 72850
+    },
+    {
+      "epoch": 18.169576059850375,
+      "grad_norm": 9.941221237182617,
+      "learning_rate": 1.8379052369077309e-06,
+      "loss": 0.376,
+      "step": 72860
+    },
+    {
+      "epoch": 18.17206982543641,
+      "grad_norm": 8.516250610351562,
+      "learning_rate": 1.8354114713216957e-06,
+      "loss": 0.2555,
+      "step": 72870
+    },
+    {
+      "epoch": 18.174563591022444,
+      "grad_norm": 9.351677894592285,
+      "learning_rate": 1.8329177057356608e-06,
+      "loss": 0.312,
+      "step": 72880
+    },
+    {
+      "epoch": 18.17705735660848,
+      "grad_norm": 10.373080253601074,
+      "learning_rate": 1.830423940149626e-06,
+      "loss": 0.3313,
+      "step": 72890
+    },
+    {
+      "epoch": 18.179551122194514,
+      "grad_norm": 6.510061264038086,
+      "learning_rate": 1.827930174563591e-06,
+      "loss": 0.2775,
+      "step": 72900
+    },
+    {
+      "epoch": 18.18204488778055,
+      "grad_norm": 7.710317611694336,
+      "learning_rate": 1.8254364089775561e-06,
+      "loss": 0.3265,
+      "step": 72910
+    },
+    {
+      "epoch": 18.184538653366584,
+      "grad_norm": 9.246587753295898,
+      "learning_rate": 1.8229426433915212e-06,
+      "loss": 0.3061,
+      "step": 72920
+    },
+    {
+      "epoch": 18.18703241895262,
+      "grad_norm": 8.690896987915039,
+      "learning_rate": 1.8204488778054865e-06,
+      "loss": 0.3098,
+      "step": 72930
+    },
+    {
+      "epoch": 18.189526184538654,
+      "grad_norm": 4.682961463928223,
+      "learning_rate": 1.8179551122194516e-06,
+      "loss": 0.2886,
+      "step": 72940
+    },
+    {
+      "epoch": 18.19201995012469,
+      "grad_norm": 5.759263038635254,
+      "learning_rate": 1.8154613466334167e-06,
+      "loss": 0.2854,
+      "step": 72950
+    },
+    {
+      "epoch": 18.194513715710723,
+      "grad_norm": 7.35156774520874,
+      "learning_rate": 1.8129675810473818e-06,
+      "loss": 0.3615,
+      "step": 72960
+    },
+    {
+      "epoch": 18.197007481296758,
+      "grad_norm": 7.8746819496154785,
+      "learning_rate": 1.810473815461347e-06,
+      "loss": 0.3827,
+      "step": 72970
+    },
+    {
+      "epoch": 18.199501246882793,
+      "grad_norm": 10.189519882202148,
+      "learning_rate": 1.807980049875312e-06,
+      "loss": 0.3083,
+      "step": 72980
+    },
+    {
+      "epoch": 18.201995012468828,
+      "grad_norm": 8.979336738586426,
+      "learning_rate": 1.8054862842892771e-06,
+      "loss": 0.366,
+      "step": 72990
+    },
+    {
+      "epoch": 18.204488778054863,
+      "grad_norm": 6.781293869018555,
+      "learning_rate": 1.802992518703242e-06,
+      "loss": 0.348,
+      "step": 73000
+    },
+    {
+      "epoch": 18.206982543640898,
+      "grad_norm": 9.620187759399414,
+      "learning_rate": 1.8004987531172071e-06,
+      "loss": 0.3721,
+      "step": 73010
+    },
+    {
+      "epoch": 18.209476309226932,
+      "grad_norm": 8.865036964416504,
+      "learning_rate": 1.7980049875311722e-06,
+      "loss": 0.3411,
+      "step": 73020
+    },
+    {
+      "epoch": 18.211970074812967,
+      "grad_norm": 9.20656681060791,
+      "learning_rate": 1.7955112219451373e-06,
+      "loss": 0.3595,
+      "step": 73030
+    },
+    {
+      "epoch": 18.214463840399002,
+      "grad_norm": 11.974329948425293,
+      "learning_rate": 1.7930174563591024e-06,
+      "loss": 0.3608,
+      "step": 73040
+    },
+    {
+      "epoch": 18.216957605985037,
+      "grad_norm": 13.708403587341309,
+      "learning_rate": 1.7905236907730675e-06,
+      "loss": 0.3302,
+      "step": 73050
+    },
+    {
+      "epoch": 18.219451371571072,
+      "grad_norm": 8.671035766601562,
+      "learning_rate": 1.7880299251870326e-06,
+      "loss": 0.2768,
+      "step": 73060
+    },
+    {
+      "epoch": 18.221945137157107,
+      "grad_norm": 9.130983352661133,
+      "learning_rate": 1.7855361596009977e-06,
+      "loss": 0.3167,
+      "step": 73070
+    },
+    {
+      "epoch": 18.22443890274314,
+      "grad_norm": 9.449166297912598,
+      "learning_rate": 1.7830423940149628e-06,
+      "loss": 0.3409,
+      "step": 73080
+    },
+    {
+      "epoch": 18.226932668329177,
+      "grad_norm": 8.308408737182617,
+      "learning_rate": 1.7805486284289277e-06,
+      "loss": 0.2954,
+      "step": 73090
+    },
+    {
+      "epoch": 18.22942643391521,
+      "grad_norm": 11.468031883239746,
+      "learning_rate": 1.7780548628428928e-06,
+      "loss": 0.3183,
+      "step": 73100
+    },
+    {
+      "epoch": 18.231920199501246,
+      "grad_norm": 10.764180183410645,
+      "learning_rate": 1.7755610972568579e-06,
+      "loss": 0.3812,
+      "step": 73110
+    },
+    {
+      "epoch": 18.23441396508728,
+      "grad_norm": 8.669768333435059,
+      "learning_rate": 1.773067331670823e-06,
+      "loss": 0.3437,
+      "step": 73120
+    },
+    {
+      "epoch": 18.236907730673316,
+      "grad_norm": 8.164244651794434,
+      "learning_rate": 1.770573566084788e-06,
+      "loss": 0.2734,
+      "step": 73130
+    },
+    {
+      "epoch": 18.23940149625935,
+      "grad_norm": 6.5684943199157715,
+      "learning_rate": 1.7680798004987532e-06,
+      "loss": 0.3716,
+      "step": 73140
+    },
+    {
+      "epoch": 18.241895261845386,
+      "grad_norm": 8.023477554321289,
+      "learning_rate": 1.7655860349127183e-06,
+      "loss": 0.3461,
+      "step": 73150
+    },
+    {
+      "epoch": 18.24438902743142,
+      "grad_norm": 6.442568302154541,
+      "learning_rate": 1.7630922693266834e-06,
+      "loss": 0.3287,
+      "step": 73160
+    },
+    {
+      "epoch": 18.246882793017456,
+      "grad_norm": 10.371484756469727,
+      "learning_rate": 1.7605985037406485e-06,
+      "loss": 0.2998,
+      "step": 73170
+    },
+    {
+      "epoch": 18.24937655860349,
+      "grad_norm": 10.30907917022705,
+      "learning_rate": 1.7581047381546138e-06,
+      "loss": 0.3161,
+      "step": 73180
+    },
+    {
+      "epoch": 18.251870324189525,
+      "grad_norm": 11.629962921142578,
+      "learning_rate": 1.7556109725685789e-06,
+      "loss": 0.2857,
+      "step": 73190
+    },
+    {
+      "epoch": 18.25436408977556,
+      "grad_norm": 10.187129020690918,
+      "learning_rate": 1.753117206982544e-06,
+      "loss": 0.3485,
+      "step": 73200
+    },
+    {
+      "epoch": 18.256857855361595,
+      "grad_norm": 13.606742858886719,
+      "learning_rate": 1.750623441396509e-06,
+      "loss": 0.3258,
+      "step": 73210
+    },
+    {
+      "epoch": 18.25935162094763,
+      "grad_norm": 7.9029860496521,
+      "learning_rate": 1.748129675810474e-06,
+      "loss": 0.264,
+      "step": 73220
+    },
+    {
+      "epoch": 18.261845386533665,
+      "grad_norm": 6.414330959320068,
+      "learning_rate": 1.745635910224439e-06,
+      "loss": 0.3278,
+      "step": 73230
+    },
+    {
+      "epoch": 18.2643391521197,
+      "grad_norm": 4.603034019470215,
+      "learning_rate": 1.7431421446384042e-06,
+      "loss": 0.304,
+      "step": 73240
+    },
+    {
+      "epoch": 18.266832917705734,
+      "grad_norm": 9.417990684509277,
+      "learning_rate": 1.7406483790523692e-06,
+      "loss": 0.3797,
+      "step": 73250
+    },
+    {
+      "epoch": 18.26932668329177,
+      "grad_norm": 10.289613723754883,
+      "learning_rate": 1.7381546134663343e-06,
+      "loss": 0.4033,
+      "step": 73260
+    },
+    {
+      "epoch": 18.271820448877804,
+      "grad_norm": 18.14396858215332,
+      "learning_rate": 1.7356608478802994e-06,
+      "loss": 0.2768,
+      "step": 73270
+    },
+    {
+      "epoch": 18.27431421446384,
+      "grad_norm": 11.941176414489746,
+      "learning_rate": 1.7331670822942645e-06,
+      "loss": 0.431,
+      "step": 73280
+    },
+    {
+      "epoch": 18.276807980049874,
+      "grad_norm": 9.569537162780762,
+      "learning_rate": 1.7306733167082296e-06,
+      "loss": 0.2749,
+      "step": 73290
+    },
+    {
+      "epoch": 18.27930174563591,
+      "grad_norm": 7.719649791717529,
+      "learning_rate": 1.7281795511221947e-06,
+      "loss": 0.3432,
+      "step": 73300
+    },
+    {
+      "epoch": 18.281795511221944,
+      "grad_norm": 6.728184223175049,
+      "learning_rate": 1.7256857855361596e-06,
+      "loss": 0.2864,
+      "step": 73310
+    },
+    {
+      "epoch": 18.28428927680798,
+      "grad_norm": 7.6080708503723145,
+      "learning_rate": 1.7231920199501247e-06,
+      "loss": 0.3257,
+      "step": 73320
+    },
+    {
+      "epoch": 18.286783042394013,
+      "grad_norm": 8.406044006347656,
+      "learning_rate": 1.7206982543640898e-06,
+      "loss": 0.3528,
+      "step": 73330
+    },
+    {
+      "epoch": 18.28927680798005,
+      "grad_norm": 11.21180534362793,
+      "learning_rate": 1.718204488778055e-06,
+      "loss": 0.3166,
+      "step": 73340
+    },
+    {
+      "epoch": 18.291770573566083,
+      "grad_norm": 12.678295135498047,
+      "learning_rate": 1.71571072319202e-06,
+      "loss": 0.3484,
+      "step": 73350
+    },
+    {
+      "epoch": 18.294264339152118,
+      "grad_norm": 13.616421699523926,
+      "learning_rate": 1.7132169576059851e-06,
+      "loss": 0.3677,
+      "step": 73360
+    },
+    {
+      "epoch": 18.296758104738153,
+      "grad_norm": 8.696226119995117,
+      "learning_rate": 1.7107231920199502e-06,
+      "loss": 0.3082,
+      "step": 73370
+    },
+    {
+      "epoch": 18.29925187032419,
+      "grad_norm": 10.46649169921875,
+      "learning_rate": 1.7082294264339153e-06,
+      "loss": 0.3416,
+      "step": 73380
+    },
+    {
+      "epoch": 18.301745635910226,
+      "grad_norm": 5.726923942565918,
+      "learning_rate": 1.7057356608478804e-06,
+      "loss": 0.3065,
+      "step": 73390
+    },
+    {
+      "epoch": 18.30423940149626,
+      "grad_norm": 12.030797004699707,
+      "learning_rate": 1.7032418952618455e-06,
+      "loss": 0.328,
+      "step": 73400
+    },
+    {
+      "epoch": 18.306733167082296,
+      "grad_norm": 8.201338768005371,
+      "learning_rate": 1.7007481296758104e-06,
+      "loss": 0.3321,
+      "step": 73410
+    },
+    {
+      "epoch": 18.30922693266833,
+      "grad_norm": 9.528154373168945,
+      "learning_rate": 1.6982543640897755e-06,
+      "loss": 0.2487,
+      "step": 73420
+    },
+    {
+      "epoch": 18.311720698254366,
+      "grad_norm": 8.872268676757812,
+      "learning_rate": 1.695760598503741e-06,
+      "loss": 0.3468,
+      "step": 73430
+    },
+    {
+      "epoch": 18.3142144638404,
+      "grad_norm": 6.192471027374268,
+      "learning_rate": 1.693266832917706e-06,
+      "loss": 0.2942,
+      "step": 73440
+    },
+    {
+      "epoch": 18.316708229426435,
+      "grad_norm": 8.524904251098633,
+      "learning_rate": 1.690773067331671e-06,
+      "loss": 0.3305,
+      "step": 73450
+    },
+    {
+      "epoch": 18.31920199501247,
+      "grad_norm": 6.720682621002197,
+      "learning_rate": 1.688279301745636e-06,
+      "loss": 0.3114,
+      "step": 73460
+    },
+    {
+      "epoch": 18.321695760598505,
+      "grad_norm": 8.790159225463867,
+      "learning_rate": 1.6857855361596012e-06,
+      "loss": 0.3181,
+      "step": 73470
+    },
+    {
+      "epoch": 18.32418952618454,
+      "grad_norm": 6.875416278839111,
+      "learning_rate": 1.6832917705735663e-06,
+      "loss": 0.3115,
+      "step": 73480
+    },
+    {
+      "epoch": 18.326683291770575,
+      "grad_norm": 8.09900188446045,
+      "learning_rate": 1.6807980049875314e-06,
+      "loss": 0.3607,
+      "step": 73490
+    },
+    {
+      "epoch": 18.32917705735661,
+      "grad_norm": 8.227255821228027,
+      "learning_rate": 1.6783042394014965e-06,
+      "loss": 0.3341,
+      "step": 73500
+    },
+    {
+      "epoch": 18.331670822942645,
+      "grad_norm": 11.627192497253418,
+      "learning_rate": 1.6758104738154616e-06,
+      "loss": 0.3915,
+      "step": 73510
+    },
+    {
+      "epoch": 18.33416458852868,
+      "grad_norm": 11.05119800567627,
+      "learning_rate": 1.6733167082294267e-06,
+      "loss": 0.3357,
+      "step": 73520
+    },
+    {
+      "epoch": 18.336658354114714,
+      "grad_norm": 9.547077178955078,
+      "learning_rate": 1.6708229426433918e-06,
+      "loss": 0.3351,
+      "step": 73530
+    },
+    {
+      "epoch": 18.33915211970075,
+      "grad_norm": 10.876405715942383,
+      "learning_rate": 1.6683291770573567e-06,
+      "loss": 0.3085,
+      "step": 73540
+    },
+    {
+      "epoch": 18.341645885286784,
+      "grad_norm": 5.926941394805908,
+      "learning_rate": 1.6658354114713218e-06,
+      "loss": 0.2994,
+      "step": 73550
+    },
+    {
+      "epoch": 18.34413965087282,
+      "grad_norm": 9.938848495483398,
+      "learning_rate": 1.6633416458852869e-06,
+      "loss": 0.3295,
+      "step": 73560
+    },
+    {
+      "epoch": 18.346633416458854,
+      "grad_norm": 9.622965812683105,
+      "learning_rate": 1.660847880299252e-06,
+      "loss": 0.2535,
+      "step": 73570
+    },
+    {
+      "epoch": 18.34912718204489,
+      "grad_norm": 7.86170768737793,
+      "learning_rate": 1.658354114713217e-06,
+      "loss": 0.3655,
+      "step": 73580
+    },
+    {
+      "epoch": 18.351620947630924,
+      "grad_norm": 11.039167404174805,
+      "learning_rate": 1.6558603491271822e-06,
+      "loss": 0.3227,
+      "step": 73590
+    },
+    {
+      "epoch": 18.35411471321696,
+      "grad_norm": 9.131935119628906,
+      "learning_rate": 1.6533665835411473e-06,
+      "loss": 0.2782,
+      "step": 73600
+    },
+    {
+      "epoch": 18.356608478802993,
+      "grad_norm": 7.852529048919678,
+      "learning_rate": 1.6508728179551124e-06,
+      "loss": 0.2881,
+      "step": 73610
+    },
+    {
+      "epoch": 18.359102244389028,
+      "grad_norm": 6.67081356048584,
+      "learning_rate": 1.6483790523690775e-06,
+      "loss": 0.3281,
+      "step": 73620
+    },
+    {
+      "epoch": 18.361596009975063,
+      "grad_norm": 9.186594009399414,
+      "learning_rate": 1.6458852867830423e-06,
+      "loss": 0.3104,
+      "step": 73630
+    },
+    {
+      "epoch": 18.364089775561098,
+      "grad_norm": 10.543354988098145,
+      "learning_rate": 1.6433915211970074e-06,
+      "loss": 0.2973,
+      "step": 73640
+    },
+    {
+      "epoch": 18.366583541147133,
+      "grad_norm": 8.135270118713379,
+      "learning_rate": 1.6408977556109725e-06,
+      "loss": 0.3834,
+      "step": 73650
+    },
+    {
+      "epoch": 18.369077306733168,
+      "grad_norm": 8.696205139160156,
+      "learning_rate": 1.6384039900249376e-06,
+      "loss": 0.317,
+      "step": 73660
+    },
+    {
+      "epoch": 18.371571072319203,
+      "grad_norm": 8.284186363220215,
+      "learning_rate": 1.6359102244389027e-06,
+      "loss": 0.3482,
+      "step": 73670
+    },
+    {
+      "epoch": 18.374064837905237,
+      "grad_norm": 8.957206726074219,
+      "learning_rate": 1.633416458852868e-06,
+      "loss": 0.3858,
+      "step": 73680
+    },
+    {
+      "epoch": 18.376558603491272,
+      "grad_norm": 11.914627075195312,
+      "learning_rate": 1.6309226932668331e-06,
+      "loss": 0.335,
+      "step": 73690
+    },
+    {
+      "epoch": 18.379052369077307,
+      "grad_norm": 5.708254814147949,
+      "learning_rate": 1.6284289276807982e-06,
+      "loss": 0.365,
+      "step": 73700
+    },
+    {
+      "epoch": 18.381546134663342,
+      "grad_norm": 9.29157829284668,
+      "learning_rate": 1.6259351620947633e-06,
+      "loss": 0.29,
+      "step": 73710
+    },
+    {
+      "epoch": 18.384039900249377,
+      "grad_norm": 6.766468048095703,
+      "learning_rate": 1.6234413965087284e-06,
+      "loss": 0.2897,
+      "step": 73720
+    },
+    {
+      "epoch": 18.38653366583541,
+      "grad_norm": 7.3969621658325195,
+      "learning_rate": 1.6209476309226935e-06,
+      "loss": 0.3482,
+      "step": 73730
+    },
+    {
+      "epoch": 18.389027431421447,
+      "grad_norm": 6.613345623016357,
+      "learning_rate": 1.6184538653366586e-06,
+      "loss": 0.3101,
+      "step": 73740
+    },
+    {
+      "epoch": 18.39152119700748,
+      "grad_norm": 7.1980881690979,
+      "learning_rate": 1.6159600997506237e-06,
+      "loss": 0.2709,
+      "step": 73750
+    },
+    {
+      "epoch": 18.394014962593516,
+      "grad_norm": 12.448799133300781,
+      "learning_rate": 1.6134663341645886e-06,
+      "loss": 0.2926,
+      "step": 73760
+    },
+    {
+      "epoch": 18.39650872817955,
+      "grad_norm": 9.95907974243164,
+      "learning_rate": 1.6109725685785537e-06,
+      "loss": 0.303,
+      "step": 73770
+    },
+    {
+      "epoch": 18.399002493765586,
+      "grad_norm": 8.818389892578125,
+      "learning_rate": 1.6084788029925188e-06,
+      "loss": 0.2965,
+      "step": 73780
+    },
+    {
+      "epoch": 18.40149625935162,
+      "grad_norm": 6.657719612121582,
+      "learning_rate": 1.605985037406484e-06,
+      "loss": 0.3098,
+      "step": 73790
+    },
+    {
+      "epoch": 18.403990024937656,
+      "grad_norm": 8.187033653259277,
+      "learning_rate": 1.603491271820449e-06,
+      "loss": 0.2981,
+      "step": 73800
+    },
+    {
+      "epoch": 18.40648379052369,
+      "grad_norm": 8.11359977722168,
+      "learning_rate": 1.600997506234414e-06,
+      "loss": 0.3991,
+      "step": 73810
+    },
+    {
+      "epoch": 18.408977556109726,
+      "grad_norm": 5.73296594619751,
+      "learning_rate": 1.5985037406483792e-06,
+      "loss": 0.3214,
+      "step": 73820
+    },
+    {
+      "epoch": 18.41147132169576,
+      "grad_norm": 7.935540199279785,
+      "learning_rate": 1.5960099750623443e-06,
+      "loss": 0.3434,
+      "step": 73830
+    },
+    {
+      "epoch": 18.413965087281795,
+      "grad_norm": 6.838698863983154,
+      "learning_rate": 1.5935162094763094e-06,
+      "loss": 0.2924,
+      "step": 73840
+    },
+    {
+      "epoch": 18.41645885286783,
+      "grad_norm": 7.937646389007568,
+      "learning_rate": 1.5910224438902743e-06,
+      "loss": 0.2674,
+      "step": 73850
+    },
+    {
+      "epoch": 18.418952618453865,
+      "grad_norm": 8.589198112487793,
+      "learning_rate": 1.5885286783042394e-06,
+      "loss": 0.3232,
+      "step": 73860
+    },
+    {
+      "epoch": 18.4214463840399,
+      "grad_norm": 6.257779598236084,
+      "learning_rate": 1.5860349127182045e-06,
+      "loss": 0.2764,
+      "step": 73870
+    },
+    {
+      "epoch": 18.423940149625935,
+      "grad_norm": 9.784756660461426,
+      "learning_rate": 1.5835411471321696e-06,
+      "loss": 0.3288,
+      "step": 73880
+    },
+    {
+      "epoch": 18.42643391521197,
+      "grad_norm": 7.663664817810059,
+      "learning_rate": 1.5810473815461347e-06,
+      "loss": 0.2703,
+      "step": 73890
+    },
+    {
+      "epoch": 18.428927680798004,
+      "grad_norm": 7.855859756469727,
+      "learning_rate": 1.5785536159600998e-06,
+      "loss": 0.2777,
+      "step": 73900
+    },
+    {
+      "epoch": 18.43142144638404,
+      "grad_norm": 8.598638534545898,
+      "learning_rate": 1.5760598503740649e-06,
+      "loss": 0.3726,
+      "step": 73910
+    },
+    {
+      "epoch": 18.433915211970074,
+      "grad_norm": 12.342257499694824,
+      "learning_rate": 1.5735660847880302e-06,
+      "loss": 0.3803,
+      "step": 73920
+    },
+    {
+      "epoch": 18.43640897755611,
+      "grad_norm": 10.327555656433105,
+      "learning_rate": 1.5710723192019953e-06,
+      "loss": 0.324,
+      "step": 73930
+    },
+    {
+      "epoch": 18.438902743142144,
+      "grad_norm": 10.554641723632812,
+      "learning_rate": 1.5685785536159604e-06,
+      "loss": 0.2346,
+      "step": 73940
+    },
+    {
+      "epoch": 18.44139650872818,
+      "grad_norm": 10.217893600463867,
+      "learning_rate": 1.5660847880299255e-06,
+      "loss": 0.3869,
+      "step": 73950
+    },
+    {
+      "epoch": 18.443890274314214,
+      "grad_norm": 9.727376937866211,
+      "learning_rate": 1.5635910224438906e-06,
+      "loss": 0.3377,
+      "step": 73960
+    },
+    {
+      "epoch": 18.44638403990025,
+      "grad_norm": 8.947395324707031,
+      "learning_rate": 1.5610972568578557e-06,
+      "loss": 0.3863,
+      "step": 73970
+    },
+    {
+      "epoch": 18.448877805486283,
+      "grad_norm": 8.704399108886719,
+      "learning_rate": 1.5586034912718206e-06,
+      "loss": 0.3594,
+      "step": 73980
+    },
+    {
+      "epoch": 18.45137157107232,
+      "grad_norm": 7.907713413238525,
+      "learning_rate": 1.5561097256857857e-06,
+      "loss": 0.3043,
+      "step": 73990
+    },
+    {
+      "epoch": 18.453865336658353,
+      "grad_norm": 11.05506420135498,
+      "learning_rate": 1.5536159600997507e-06,
+      "loss": 0.307,
+      "step": 74000
+    },
+    {
+      "epoch": 18.456359102244388,
+      "grad_norm": 10.79157829284668,
+      "learning_rate": 1.5511221945137158e-06,
+      "loss": 0.3657,
+      "step": 74010
+    },
+    {
+      "epoch": 18.458852867830423,
+      "grad_norm": 9.126988410949707,
+      "learning_rate": 1.548628428927681e-06,
+      "loss": 0.2559,
+      "step": 74020
+    },
+    {
+      "epoch": 18.461346633416458,
+      "grad_norm": 15.90583324432373,
+      "learning_rate": 1.546134663341646e-06,
+      "loss": 0.387,
+      "step": 74030
+    },
+    {
+      "epoch": 18.463840399002493,
+      "grad_norm": 7.69087028503418,
+      "learning_rate": 1.5436408977556111e-06,
+      "loss": 0.3183,
+      "step": 74040
+    },
+    {
+      "epoch": 18.466334164588527,
+      "grad_norm": 10.746506690979004,
+      "learning_rate": 1.5411471321695762e-06,
+      "loss": 0.3444,
+      "step": 74050
+    },
+    {
+      "epoch": 18.468827930174562,
+      "grad_norm": 10.501643180847168,
+      "learning_rate": 1.5386533665835413e-06,
+      "loss": 0.3714,
+      "step": 74060
+    },
+    {
+      "epoch": 18.471321695760597,
+      "grad_norm": 5.846063137054443,
+      "learning_rate": 1.5361596009975062e-06,
+      "loss": 0.3788,
+      "step": 74070
+    },
+    {
+      "epoch": 18.473815461346632,
+      "grad_norm": 5.375149726867676,
+      "learning_rate": 1.5336658354114713e-06,
+      "loss": 0.2884,
+      "step": 74080
+    },
+    {
+      "epoch": 18.476309226932667,
+      "grad_norm": 8.232563972473145,
+      "learning_rate": 1.5311720698254364e-06,
+      "loss": 0.3549,
+      "step": 74090
+    },
+    {
+      "epoch": 18.478802992518702,
+      "grad_norm": 7.693140983581543,
+      "learning_rate": 1.5286783042394015e-06,
+      "loss": 0.3765,
+      "step": 74100
+    },
+    {
+      "epoch": 18.481296758104737,
+      "grad_norm": 8.750802040100098,
+      "learning_rate": 1.5261845386533666e-06,
+      "loss": 0.2949,
+      "step": 74110
+    },
+    {
+      "epoch": 18.48379052369077,
+      "grad_norm": 7.751341819763184,
+      "learning_rate": 1.5236907730673317e-06,
+      "loss": 0.3237,
+      "step": 74120
+    },
+    {
+      "epoch": 18.486284289276806,
+      "grad_norm": 8.8473482131958,
+      "learning_rate": 1.5211970074812968e-06,
+      "loss": 0.2869,
+      "step": 74130
+    },
+    {
+      "epoch": 18.48877805486284,
+      "grad_norm": 8.20888900756836,
+      "learning_rate": 1.518703241895262e-06,
+      "loss": 0.3399,
+      "step": 74140
+    },
+    {
+      "epoch": 18.491271820448876,
+      "grad_norm": 8.531835556030273,
+      "learning_rate": 1.516209476309227e-06,
+      "loss": 0.2784,
+      "step": 74150
+    },
+    {
+      "epoch": 18.49376558603491,
+      "grad_norm": 13.030848503112793,
+      "learning_rate": 1.513715710723192e-06,
+      "loss": 0.3464,
+      "step": 74160
+    },
+    {
+      "epoch": 18.496259351620946,
+      "grad_norm": 9.676426887512207,
+      "learning_rate": 1.5112219451371574e-06,
+      "loss": 0.3156,
+      "step": 74170
+    },
+    {
+      "epoch": 18.49875311720698,
+      "grad_norm": 5.307429313659668,
+      "learning_rate": 1.5087281795511225e-06,
+      "loss": 0.3121,
+      "step": 74180
+    },
+    {
+      "epoch": 18.50124688279302,
+      "grad_norm": 8.61061954498291,
+      "learning_rate": 1.5062344139650876e-06,
+      "loss": 0.4055,
+      "step": 74190
+    },
+    {
+      "epoch": 18.503740648379054,
+      "grad_norm": 7.381380558013916,
+      "learning_rate": 1.5037406483790525e-06,
+      "loss": 0.3907,
+      "step": 74200
+    },
+    {
+      "epoch": 18.50623441396509,
+      "grad_norm": 10.964561462402344,
+      "learning_rate": 1.5012468827930176e-06,
+      "loss": 0.3313,
+      "step": 74210
+    },
+    {
+      "epoch": 18.508728179551124,
+      "grad_norm": 8.3394136428833,
+      "learning_rate": 1.4987531172069827e-06,
+      "loss": 0.2691,
+      "step": 74220
+    },
+    {
+      "epoch": 18.51122194513716,
+      "grad_norm": 7.458098411560059,
+      "learning_rate": 1.4962593516209478e-06,
+      "loss": 0.307,
+      "step": 74230
+    },
+    {
+      "epoch": 18.513715710723194,
+      "grad_norm": 12.269042015075684,
+      "learning_rate": 1.4937655860349129e-06,
+      "loss": 0.299,
+      "step": 74240
+    },
+    {
+      "epoch": 18.51620947630923,
+      "grad_norm": 5.956721305847168,
+      "learning_rate": 1.491271820448878e-06,
+      "loss": 0.3303,
+      "step": 74250
+    },
+    {
+      "epoch": 18.518703241895263,
+      "grad_norm": 8.427618980407715,
+      "learning_rate": 1.488778054862843e-06,
+      "loss": 0.3256,
+      "step": 74260
+    },
+    {
+      "epoch": 18.521197007481298,
+      "grad_norm": 7.890618801116943,
+      "learning_rate": 1.4862842892768082e-06,
+      "loss": 0.3507,
+      "step": 74270
+    },
+    {
+      "epoch": 18.523690773067333,
+      "grad_norm": 10.247663497924805,
+      "learning_rate": 1.4837905236907733e-06,
+      "loss": 0.345,
+      "step": 74280
+    },
+    {
+      "epoch": 18.526184538653368,
+      "grad_norm": 11.79736614227295,
+      "learning_rate": 1.4812967581047384e-06,
+      "loss": 0.3271,
+      "step": 74290
+    },
+    {
+      "epoch": 18.528678304239403,
+      "grad_norm": 7.097879409790039,
+      "learning_rate": 1.4788029925187033e-06,
+      "loss": 0.2992,
+      "step": 74300
+    },
+    {
+      "epoch": 18.531172069825438,
+      "grad_norm": 9.422330856323242,
+      "learning_rate": 1.4763092269326684e-06,
+      "loss": 0.3164,
+      "step": 74310
+    },
+    {
+      "epoch": 18.533665835411473,
+      "grad_norm": 8.391748428344727,
+      "learning_rate": 1.4738154613466335e-06,
+      "loss": 0.3513,
+      "step": 74320
+    },
+    {
+      "epoch": 18.536159600997507,
+      "grad_norm": 12.254179954528809,
+      "learning_rate": 1.4713216957605986e-06,
+      "loss": 0.3117,
+      "step": 74330
+    },
+    {
+      "epoch": 18.538653366583542,
+      "grad_norm": 7.639408588409424,
+      "learning_rate": 1.4688279301745637e-06,
+      "loss": 0.2665,
+      "step": 74340
+    },
+    {
+      "epoch": 18.541147132169577,
+      "grad_norm": 13.742928504943848,
+      "learning_rate": 1.4663341645885288e-06,
+      "loss": 0.2916,
+      "step": 74350
+    },
+    {
+      "epoch": 18.543640897755612,
+      "grad_norm": 8.131094932556152,
+      "learning_rate": 1.4638403990024939e-06,
+      "loss": 0.3436,
+      "step": 74360
+    },
+    {
+      "epoch": 18.546134663341647,
+      "grad_norm": 8.458292961120605,
+      "learning_rate": 1.461346633416459e-06,
+      "loss": 0.3361,
+      "step": 74370
+    },
+    {
+      "epoch": 18.54862842892768,
+      "grad_norm": 7.697943210601807,
+      "learning_rate": 1.458852867830424e-06,
+      "loss": 0.2833,
+      "step": 74380
+    },
+    {
+      "epoch": 18.551122194513717,
+      "grad_norm": 11.632762908935547,
+      "learning_rate": 1.456359102244389e-06,
+      "loss": 0.2625,
+      "step": 74390
+    },
+    {
+      "epoch": 18.55361596009975,
+      "grad_norm": 8.714035034179688,
+      "learning_rate": 1.453865336658354e-06,
+      "loss": 0.3138,
+      "step": 74400
+    },
+    {
+      "epoch": 18.556109725685786,
+      "grad_norm": 10.529500961303711,
+      "learning_rate": 1.4513715710723191e-06,
+      "loss": 0.4054,
+      "step": 74410
+    },
+    {
+      "epoch": 18.55860349127182,
+      "grad_norm": 7.038276672363281,
+      "learning_rate": 1.4488778054862844e-06,
+      "loss": 0.318,
+      "step": 74420
+    },
+    {
+      "epoch": 18.561097256857856,
+      "grad_norm": 8.944663047790527,
+      "learning_rate": 1.4463840399002495e-06,
+      "loss": 0.3261,
+      "step": 74430
+    },
+    {
+      "epoch": 18.56359102244389,
+      "grad_norm": 6.774261474609375,
+      "learning_rate": 1.4438902743142146e-06,
+      "loss": 0.3315,
+      "step": 74440
+    },
+    {
+      "epoch": 18.566084788029926,
+      "grad_norm": 8.720541954040527,
+      "learning_rate": 1.4413965087281797e-06,
+      "loss": 0.3715,
+      "step": 74450
+    },
+    {
+      "epoch": 18.56857855361596,
+      "grad_norm": 9.304561614990234,
+      "learning_rate": 1.4389027431421448e-06,
+      "loss": 0.2913,
+      "step": 74460
+    },
+    {
+      "epoch": 18.571072319201996,
+      "grad_norm": 6.90985631942749,
+      "learning_rate": 1.43640897755611e-06,
+      "loss": 0.3142,
+      "step": 74470
+    },
+    {
+      "epoch": 18.57356608478803,
+      "grad_norm": 6.733709335327148,
+      "learning_rate": 1.433915211970075e-06,
+      "loss": 0.3388,
+      "step": 74480
+    },
+    {
+      "epoch": 18.576059850374065,
+      "grad_norm": 10.039854049682617,
+      "learning_rate": 1.4314214463840401e-06,
+      "loss": 0.3232,
+      "step": 74490
+    },
+    {
+      "epoch": 18.5785536159601,
+      "grad_norm": 9.462594032287598,
+      "learning_rate": 1.4289276807980052e-06,
+      "loss": 0.3591,
+      "step": 74500
+    },
+    {
+      "epoch": 18.581047381546135,
+      "grad_norm": 9.804282188415527,
+      "learning_rate": 1.4264339152119703e-06,
+      "loss": 0.3105,
+      "step": 74510
+    },
+    {
+      "epoch": 18.58354114713217,
+      "grad_norm": 11.309243202209473,
+      "learning_rate": 1.4239401496259352e-06,
+      "loss": 0.3292,
+      "step": 74520
+    },
+    {
+      "epoch": 18.586034912718205,
+      "grad_norm": 13.264601707458496,
+      "learning_rate": 1.4214463840399003e-06,
+      "loss": 0.3184,
+      "step": 74530
+    },
+    {
+      "epoch": 18.58852867830424,
+      "grad_norm": 8.009174346923828,
+      "learning_rate": 1.4189526184538654e-06,
+      "loss": 0.3198,
+      "step": 74540
+    },
+    {
+      "epoch": 18.591022443890274,
+      "grad_norm": 9.68816089630127,
+      "learning_rate": 1.4164588528678305e-06,
+      "loss": 0.2911,
+      "step": 74550
+    },
+    {
+      "epoch": 18.59351620947631,
+      "grad_norm": 15.687336921691895,
+      "learning_rate": 1.4139650872817956e-06,
+      "loss": 0.2892,
+      "step": 74560
+    },
+    {
+      "epoch": 18.596009975062344,
+      "grad_norm": 11.459478378295898,
+      "learning_rate": 1.4114713216957607e-06,
+      "loss": 0.361,
+      "step": 74570
+    },
+    {
+      "epoch": 18.59850374064838,
+      "grad_norm": 12.541219711303711,
+      "learning_rate": 1.4089775561097258e-06,
+      "loss": 0.3387,
+      "step": 74580
+    },
+    {
+      "epoch": 18.600997506234414,
+      "grad_norm": 8.407682418823242,
+      "learning_rate": 1.4064837905236909e-06,
+      "loss": 0.366,
+      "step": 74590
+    },
+    {
+      "epoch": 18.60349127182045,
+      "grad_norm": 9.860024452209473,
+      "learning_rate": 1.403990024937656e-06,
+      "loss": 0.3266,
+      "step": 74600
+    },
+    {
+      "epoch": 18.605985037406484,
+      "grad_norm": 13.05792236328125,
+      "learning_rate": 1.4014962593516209e-06,
+      "loss": 0.3454,
+      "step": 74610
+    },
+    {
+      "epoch": 18.60847880299252,
+      "grad_norm": 9.172198295593262,
+      "learning_rate": 1.399002493765586e-06,
+      "loss": 0.3367,
+      "step": 74620
+    },
+    {
+      "epoch": 18.610972568578553,
+      "grad_norm": 9.569443702697754,
+      "learning_rate": 1.396508728179551e-06,
+      "loss": 0.2912,
+      "step": 74630
+    },
+    {
+      "epoch": 18.61346633416459,
+      "grad_norm": 6.067528247833252,
+      "learning_rate": 1.3940149625935162e-06,
+      "loss": 0.2704,
+      "step": 74640
+    },
+    {
+      "epoch": 18.615960099750623,
+      "grad_norm": 12.0790433883667,
+      "learning_rate": 1.3915211970074813e-06,
+      "loss": 0.2749,
+      "step": 74650
+    },
+    {
+      "epoch": 18.618453865336658,
+      "grad_norm": 8.685276985168457,
+      "learning_rate": 1.3890274314214464e-06,
+      "loss": 0.2838,
+      "step": 74660
+    },
+    {
+      "epoch": 18.620947630922693,
+      "grad_norm": 8.051512718200684,
+      "learning_rate": 1.3865336658354117e-06,
+      "loss": 0.2934,
+      "step": 74670
+    },
+    {
+      "epoch": 18.623441396508728,
+      "grad_norm": 6.928664684295654,
+      "learning_rate": 1.3840399002493768e-06,
+      "loss": 0.2874,
+      "step": 74680
+    },
+    {
+      "epoch": 18.625935162094763,
+      "grad_norm": 5.946801662445068,
+      "learning_rate": 1.3815461346633419e-06,
+      "loss": 0.2963,
+      "step": 74690
+    },
+    {
+      "epoch": 18.628428927680797,
+      "grad_norm": 9.436132431030273,
+      "learning_rate": 1.379052369077307e-06,
+      "loss": 0.3327,
+      "step": 74700
+    },
+    {
+      "epoch": 18.630922693266832,
+      "grad_norm": 6.325564384460449,
+      "learning_rate": 1.376558603491272e-06,
+      "loss": 0.313,
+      "step": 74710
+    },
+    {
+      "epoch": 18.633416458852867,
+      "grad_norm": 5.90425443649292,
+      "learning_rate": 1.3740648379052372e-06,
+      "loss": 0.2912,
+      "step": 74720
+    },
+    {
+      "epoch": 18.635910224438902,
+      "grad_norm": 6.267174243927002,
+      "learning_rate": 1.3715710723192023e-06,
+      "loss": 0.3487,
+      "step": 74730
+    },
+    {
+      "epoch": 18.638403990024937,
+      "grad_norm": 6.616288185119629,
+      "learning_rate": 1.3690773067331672e-06,
+      "loss": 0.3146,
+      "step": 74740
+    },
+    {
+      "epoch": 18.640897755610972,
+      "grad_norm": 8.680601119995117,
+      "learning_rate": 1.3665835411471322e-06,
+      "loss": 0.3009,
+      "step": 74750
+    },
+    {
+      "epoch": 18.643391521197007,
+      "grad_norm": 12.029298782348633,
+      "learning_rate": 1.3640897755610973e-06,
+      "loss": 0.328,
+      "step": 74760
+    },
+    {
+      "epoch": 18.64588528678304,
+      "grad_norm": 6.389432907104492,
+      "learning_rate": 1.3615960099750624e-06,
+      "loss": 0.3464,
+      "step": 74770
+    },
+    {
+      "epoch": 18.648379052369076,
+      "grad_norm": 6.558785915374756,
+      "learning_rate": 1.3591022443890275e-06,
+      "loss": 0.2935,
+      "step": 74780
+    },
+    {
+      "epoch": 18.65087281795511,
+      "grad_norm": 8.406317710876465,
+      "learning_rate": 1.3566084788029926e-06,
+      "loss": 0.3135,
+      "step": 74790
+    },
+    {
+      "epoch": 18.653366583541146,
+      "grad_norm": 11.625472068786621,
+      "learning_rate": 1.3541147132169577e-06,
+      "loss": 0.3286,
+      "step": 74800
+    },
+    {
+      "epoch": 18.65586034912718,
+      "grad_norm": 7.691009044647217,
+      "learning_rate": 1.3516209476309228e-06,
+      "loss": 0.2337,
+      "step": 74810
+    },
+    {
+      "epoch": 18.658354114713216,
+      "grad_norm": 8.831992149353027,
+      "learning_rate": 1.349127182044888e-06,
+      "loss": 0.393,
+      "step": 74820
+    },
+    {
+      "epoch": 18.66084788029925,
+      "grad_norm": 10.607884407043457,
+      "learning_rate": 1.3466334164588528e-06,
+      "loss": 0.2846,
+      "step": 74830
+    },
+    {
+      "epoch": 18.663341645885286,
+      "grad_norm": 12.452134132385254,
+      "learning_rate": 1.344139650872818e-06,
+      "loss": 0.391,
+      "step": 74840
+    },
+    {
+      "epoch": 18.66583541147132,
+      "grad_norm": 7.653696537017822,
+      "learning_rate": 1.341645885286783e-06,
+      "loss": 0.2668,
+      "step": 74850
+    },
+    {
+      "epoch": 18.668329177057355,
+      "grad_norm": 4.616998195648193,
+      "learning_rate": 1.3391521197007481e-06,
+      "loss": 0.2979,
+      "step": 74860
+    },
+    {
+      "epoch": 18.67082294264339,
+      "grad_norm": 8.645465850830078,
+      "learning_rate": 1.3366583541147132e-06,
+      "loss": 0.3351,
+      "step": 74870
+    },
+    {
+      "epoch": 18.673316708229425,
+      "grad_norm": 11.444828987121582,
+      "learning_rate": 1.3341645885286783e-06,
+      "loss": 0.2851,
+      "step": 74880
+    },
+    {
+      "epoch": 18.67581047381546,
+      "grad_norm": 8.822684288024902,
+      "learning_rate": 1.3316708229426434e-06,
+      "loss": 0.3381,
+      "step": 74890
+    },
+    {
+      "epoch": 18.678304239401495,
+      "grad_norm": 6.625451564788818,
+      "learning_rate": 1.3291770573566085e-06,
+      "loss": 0.3038,
+      "step": 74900
+    },
+    {
+      "epoch": 18.68079800498753,
+      "grad_norm": 9.063295364379883,
+      "learning_rate": 1.3269326683291772e-06,
+      "loss": 0.2841,
+      "step": 74910
+    },
+    {
+      "epoch": 18.683291770573565,
+      "grad_norm": 10.658787727355957,
+      "learning_rate": 1.3244389027431423e-06,
+      "loss": 0.2965,
+      "step": 74920
+    },
+    {
+      "epoch": 18.6857855361596,
+      "grad_norm": 10.322769165039062,
+      "learning_rate": 1.3219451371571074e-06,
+      "loss": 0.3753,
+      "step": 74930
+    },
+    {
+      "epoch": 18.688279301745634,
+      "grad_norm": 5.894896030426025,
+      "learning_rate": 1.3194513715710725e-06,
+      "loss": 0.351,
+      "step": 74940
+    },
+    {
+      "epoch": 18.69077306733167,
+      "grad_norm": 11.83723258972168,
+      "learning_rate": 1.3169576059850376e-06,
+      "loss": 0.3247,
+      "step": 74950
+    },
+    {
+      "epoch": 18.693266832917704,
+      "grad_norm": 8.492645263671875,
+      "learning_rate": 1.314713216957606e-06,
+      "loss": 0.3656,
+      "step": 74960
+    },
+    {
+      "epoch": 18.69576059850374,
+      "grad_norm": 11.175936698913574,
+      "learning_rate": 1.3122194513715712e-06,
+      "loss": 0.348,
+      "step": 74970
+    },
+    {
+      "epoch": 18.698254364089777,
+      "grad_norm": 7.240017890930176,
+      "learning_rate": 1.309725685785536e-06,
+      "loss": 0.2931,
+      "step": 74980
+    },
+    {
+      "epoch": 18.70074812967581,
+      "grad_norm": 8.305476188659668,
+      "learning_rate": 1.3072319201995012e-06,
+      "loss": 0.3615,
+      "step": 74990
+    },
+    {
+      "epoch": 18.703241895261847,
+      "grad_norm": 12.025943756103516,
+      "learning_rate": 1.3047381546134663e-06,
+      "loss": 0.3484,
+      "step": 75000
+    },
+    {
+      "epoch": 18.705735660847882,
+      "grad_norm": 9.331719398498535,
+      "learning_rate": 1.3022443890274316e-06,
+      "loss": 0.3532,
+      "step": 75010
+    },
+    {
+      "epoch": 18.708229426433917,
+      "grad_norm": 11.926877975463867,
+      "learning_rate": 1.2997506234413967e-06,
+      "loss": 0.3444,
+      "step": 75020
+    },
+    {
+      "epoch": 18.71072319201995,
+      "grad_norm": 7.448859214782715,
+      "learning_rate": 1.2972568578553618e-06,
+      "loss": 0.2879,
+      "step": 75030
+    },
+    {
+      "epoch": 18.713216957605987,
+      "grad_norm": 11.024309158325195,
+      "learning_rate": 1.2947630922693269e-06,
+      "loss": 0.34,
+      "step": 75040
+    },
+    {
+      "epoch": 18.71571072319202,
+      "grad_norm": 6.629947185516357,
+      "learning_rate": 1.292269326683292e-06,
+      "loss": 0.3714,
+      "step": 75050
+    },
+    {
+      "epoch": 18.718204488778056,
+      "grad_norm": 7.899852275848389,
+      "learning_rate": 1.289775561097257e-06,
+      "loss": 0.3046,
+      "step": 75060
+    },
+    {
+      "epoch": 18.72069825436409,
+      "grad_norm": 8.794214248657227,
+      "learning_rate": 1.2872817955112222e-06,
+      "loss": 0.3517,
+      "step": 75070
+    },
+    {
+      "epoch": 18.723192019950126,
+      "grad_norm": 6.43766450881958,
+      "learning_rate": 1.2847880299251873e-06,
+      "loss": 0.3165,
+      "step": 75080
+    },
+    {
+      "epoch": 18.72568578553616,
+      "grad_norm": 6.629555702209473,
+      "learning_rate": 1.2822942643391523e-06,
+      "loss": 0.3499,
+      "step": 75090
+    },
+    {
+      "epoch": 18.728179551122196,
+      "grad_norm": 11.044578552246094,
+      "learning_rate": 1.2798004987531174e-06,
+      "loss": 0.3129,
+      "step": 75100
+    },
+    {
+      "epoch": 18.73067331670823,
+      "grad_norm": 9.229756355285645,
+      "learning_rate": 1.2773067331670823e-06,
+      "loss": 0.3121,
+      "step": 75110
+    },
+    {
+      "epoch": 18.733167082294266,
+      "grad_norm": 7.941111087799072,
+      "learning_rate": 1.2748129675810474e-06,
+      "loss": 0.4249,
+      "step": 75120
+    },
+    {
+      "epoch": 18.7356608478803,
+      "grad_norm": 9.484469413757324,
+      "learning_rate": 1.2723192019950125e-06,
+      "loss": 0.2897,
+      "step": 75130
+    },
+    {
+      "epoch": 18.738154613466335,
+      "grad_norm": 7.1095075607299805,
+      "learning_rate": 1.2698254364089776e-06,
+      "loss": 0.3664,
+      "step": 75140
+    },
+    {
+      "epoch": 18.74064837905237,
+      "grad_norm": 6.883621692657471,
+      "learning_rate": 1.2673316708229427e-06,
+      "loss": 0.2501,
+      "step": 75150
+    },
+    {
+      "epoch": 18.743142144638405,
+      "grad_norm": 8.864466667175293,
+      "learning_rate": 1.2648379052369078e-06,
+      "loss": 0.3146,
+      "step": 75160
+    },
+    {
+      "epoch": 18.74563591022444,
+      "grad_norm": 7.686639308929443,
+      "learning_rate": 1.262344139650873e-06,
+      "loss": 0.2874,
+      "step": 75170
+    },
+    {
+      "epoch": 18.748129675810475,
+      "grad_norm": 7.733112812042236,
+      "learning_rate": 1.259850374064838e-06,
+      "loss": 0.3358,
+      "step": 75180
+    },
+    {
+      "epoch": 18.75062344139651,
+      "grad_norm": 8.906079292297363,
+      "learning_rate": 1.2573566084788031e-06,
+      "loss": 0.3112,
+      "step": 75190
+    },
+    {
+      "epoch": 18.753117206982544,
+      "grad_norm": 11.127494812011719,
+      "learning_rate": 1.254862842892768e-06,
+      "loss": 0.3105,
+      "step": 75200
+    },
+    {
+      "epoch": 18.75561097256858,
+      "grad_norm": 8.906811714172363,
+      "learning_rate": 1.252369077306733e-06,
+      "loss": 0.3047,
+      "step": 75210
+    },
+    {
+      "epoch": 18.758104738154614,
+      "grad_norm": 6.8946027755737305,
+      "learning_rate": 1.2498753117206984e-06,
+      "loss": 0.3289,
+      "step": 75220
+    },
+    {
+      "epoch": 18.76059850374065,
+      "grad_norm": 8.487861633300781,
+      "learning_rate": 1.2473815461346635e-06,
+      "loss": 0.4042,
+      "step": 75230
+    },
+    {
+      "epoch": 18.763092269326684,
+      "grad_norm": 8.7031831741333,
+      "learning_rate": 1.2448877805486286e-06,
+      "loss": 0.3045,
+      "step": 75240
+    },
+    {
+      "epoch": 18.76558603491272,
+      "grad_norm": 8.133673667907715,
+      "learning_rate": 1.2423940149625937e-06,
+      "loss": 0.324,
+      "step": 75250
+    },
+    {
+      "epoch": 18.768079800498754,
+      "grad_norm": 8.087395668029785,
+      "learning_rate": 1.2399002493765588e-06,
+      "loss": 0.3004,
+      "step": 75260
+    },
+    {
+      "epoch": 18.77057356608479,
+      "grad_norm": 8.729161262512207,
+      "learning_rate": 1.2374064837905237e-06,
+      "loss": 0.3408,
+      "step": 75270
+    },
+    {
+      "epoch": 18.773067331670823,
+      "grad_norm": 9.859537124633789,
+      "learning_rate": 1.2349127182044888e-06,
+      "loss": 0.3635,
+      "step": 75280
+    },
+    {
+      "epoch": 18.77556109725686,
+      "grad_norm": 8.983450889587402,
+      "learning_rate": 1.2324189526184539e-06,
+      "loss": 0.2727,
+      "step": 75290
+    },
+    {
+      "epoch": 18.778054862842893,
+      "grad_norm": 7.337526798248291,
+      "learning_rate": 1.229925187032419e-06,
+      "loss": 0.2853,
+      "step": 75300
+    },
+    {
+      "epoch": 18.780548628428928,
+      "grad_norm": 9.413741111755371,
+      "learning_rate": 1.227431421446384e-06,
+      "loss": 0.3259,
+      "step": 75310
+    },
+    {
+      "epoch": 18.783042394014963,
+      "grad_norm": 8.917171478271484,
+      "learning_rate": 1.2249376558603494e-06,
+      "loss": 0.3087,
+      "step": 75320
+    },
+    {
+      "epoch": 18.785536159600998,
+      "grad_norm": 9.118664741516113,
+      "learning_rate": 1.2224438902743143e-06,
+      "loss": 0.3468,
+      "step": 75330
+    },
+    {
+      "epoch": 18.788029925187033,
+      "grad_norm": 11.285893440246582,
+      "learning_rate": 1.2199501246882794e-06,
+      "loss": 0.3248,
+      "step": 75340
+    },
+    {
+      "epoch": 18.790523690773068,
+      "grad_norm": 8.557662963867188,
+      "learning_rate": 1.2174563591022445e-06,
+      "loss": 0.3836,
+      "step": 75350
+    },
+    {
+      "epoch": 18.793017456359102,
+      "grad_norm": 9.565607070922852,
+      "learning_rate": 1.2149625935162096e-06,
+      "loss": 0.3616,
+      "step": 75360
+    },
+    {
+      "epoch": 18.795511221945137,
+      "grad_norm": 7.839760780334473,
+      "learning_rate": 1.2124688279301747e-06,
+      "loss": 0.4556,
+      "step": 75370
+    },
+    {
+      "epoch": 18.798004987531172,
+      "grad_norm": 8.867552757263184,
+      "learning_rate": 1.2099750623441398e-06,
+      "loss": 0.3541,
+      "step": 75380
+    },
+    {
+      "epoch": 18.800498753117207,
+      "grad_norm": 8.26724910736084,
+      "learning_rate": 1.2074812967581049e-06,
+      "loss": 0.3567,
+      "step": 75390
+    },
+    {
+      "epoch": 18.802992518703242,
+      "grad_norm": 11.374317169189453,
+      "learning_rate": 1.20498753117207e-06,
+      "loss": 0.2463,
+      "step": 75400
+    },
+    {
+      "epoch": 18.805486284289277,
+      "grad_norm": 6.54574728012085,
+      "learning_rate": 1.202493765586035e-06,
+      "loss": 0.3166,
+      "step": 75410
+    },
+    {
+      "epoch": 18.80798004987531,
+      "grad_norm": 6.992059230804443,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.3017,
+      "step": 75420
+    },
+    {
+      "epoch": 18.810473815461346,
+      "grad_norm": 6.233072757720947,
+      "learning_rate": 1.197506234413965e-06,
+      "loss": 0.3673,
+      "step": 75430
+    },
+    {
+      "epoch": 18.81296758104738,
+      "grad_norm": 7.485711097717285,
+      "learning_rate": 1.1950124688279301e-06,
+      "loss": 0.2989,
+      "step": 75440
+    },
+    {
+      "epoch": 18.815461346633416,
+      "grad_norm": 9.747830390930176,
+      "learning_rate": 1.1925187032418955e-06,
+      "loss": 0.325,
+      "step": 75450
+    },
+    {
+      "epoch": 18.81795511221945,
+      "grad_norm": 6.727051734924316,
+      "learning_rate": 1.1900249376558605e-06,
+      "loss": 0.3339,
+      "step": 75460
+    },
+    {
+      "epoch": 18.820448877805486,
+      "grad_norm": 7.6110968589782715,
+      "learning_rate": 1.1875311720698256e-06,
+      "loss": 0.2869,
+      "step": 75470
+    },
+    {
+      "epoch": 18.82294264339152,
+      "grad_norm": 9.239736557006836,
+      "learning_rate": 1.1850374064837907e-06,
+      "loss": 0.2949,
+      "step": 75480
+    },
+    {
+      "epoch": 18.825436408977556,
+      "grad_norm": 11.853121757507324,
+      "learning_rate": 1.1825436408977556e-06,
+      "loss": 0.3425,
+      "step": 75490
+    },
+    {
+      "epoch": 18.82793017456359,
+      "grad_norm": 9.181052207946777,
+      "learning_rate": 1.1800498753117207e-06,
+      "loss": 0.3708,
+      "step": 75500
+    },
+    {
+      "epoch": 18.830423940149625,
+      "grad_norm": 13.224994659423828,
+      "learning_rate": 1.1775561097256858e-06,
+      "loss": 0.3853,
+      "step": 75510
+    },
+    {
+      "epoch": 18.83291770573566,
+      "grad_norm": 8.818222045898438,
+      "learning_rate": 1.175062344139651e-06,
+      "loss": 0.3206,
+      "step": 75520
+    },
+    {
+      "epoch": 18.835411471321695,
+      "grad_norm": 10.314600944519043,
+      "learning_rate": 1.172568578553616e-06,
+      "loss": 0.3437,
+      "step": 75530
+    },
+    {
+      "epoch": 18.83790523690773,
+      "grad_norm": 9.383597373962402,
+      "learning_rate": 1.1700748129675811e-06,
+      "loss": 0.4307,
+      "step": 75540
+    },
+    {
+      "epoch": 18.840399002493765,
+      "grad_norm": 7.339836120605469,
+      "learning_rate": 1.1675810473815462e-06,
+      "loss": 0.3421,
+      "step": 75550
+    },
+    {
+      "epoch": 18.8428927680798,
+      "grad_norm": 8.67222785949707,
+      "learning_rate": 1.1650872817955113e-06,
+      "loss": 0.2738,
+      "step": 75560
+    },
+    {
+      "epoch": 18.845386533665835,
+      "grad_norm": 12.440013885498047,
+      "learning_rate": 1.1625935162094764e-06,
+      "loss": 0.3523,
+      "step": 75570
+    },
+    {
+      "epoch": 18.84788029925187,
+      "grad_norm": 8.511222839355469,
+      "learning_rate": 1.1600997506234415e-06,
+      "loss": 0.3205,
+      "step": 75580
+    },
+    {
+      "epoch": 18.850374064837904,
+      "grad_norm": 10.684443473815918,
+      "learning_rate": 1.1576059850374066e-06,
+      "loss": 0.3092,
+      "step": 75590
+    },
+    {
+      "epoch": 18.85286783042394,
+      "grad_norm": 8.839110374450684,
+      "learning_rate": 1.1551122194513717e-06,
+      "loss": 0.3351,
+      "step": 75600
+    },
+    {
+      "epoch": 18.855361596009974,
+      "grad_norm": 9.381013870239258,
+      "learning_rate": 1.1526184538653368e-06,
+      "loss": 0.2935,
+      "step": 75610
+    },
+    {
+      "epoch": 18.85785536159601,
+      "grad_norm": 10.822402954101562,
+      "learning_rate": 1.150124688279302e-06,
+      "loss": 0.3281,
+      "step": 75620
+    },
+    {
+      "epoch": 18.860349127182044,
+      "grad_norm": 8.065682411193848,
+      "learning_rate": 1.147630922693267e-06,
+      "loss": 0.336,
+      "step": 75630
+    },
+    {
+      "epoch": 18.86284289276808,
+      "grad_norm": 8.760025978088379,
+      "learning_rate": 1.145137157107232e-06,
+      "loss": 0.285,
+      "step": 75640
+    },
+    {
+      "epoch": 18.865336658354114,
+      "grad_norm": 10.8186616897583,
+      "learning_rate": 1.142643391521197e-06,
+      "loss": 0.4083,
+      "step": 75650
+    },
+    {
+      "epoch": 18.86783042394015,
+      "grad_norm": 10.44265079498291,
+      "learning_rate": 1.140149625935162e-06,
+      "loss": 0.2865,
+      "step": 75660
+    },
+    {
+      "epoch": 18.870324189526183,
+      "grad_norm": 12.26836109161377,
+      "learning_rate": 1.1376558603491272e-06,
+      "loss": 0.3483,
+      "step": 75670
+    },
+    {
+      "epoch": 18.872817955112218,
+      "grad_norm": 7.17544412612915,
+      "learning_rate": 1.1351620947630923e-06,
+      "loss": 0.3349,
+      "step": 75680
+    },
+    {
+      "epoch": 18.875311720698253,
+      "grad_norm": 10.033705711364746,
+      "learning_rate": 1.1326683291770576e-06,
+      "loss": 0.3098,
+      "step": 75690
+    },
+    {
+      "epoch": 18.877805486284288,
+      "grad_norm": 11.135791778564453,
+      "learning_rate": 1.1301745635910227e-06,
+      "loss": 0.2847,
+      "step": 75700
+    },
+    {
+      "epoch": 18.880299251870323,
+      "grad_norm": 6.80994176864624,
+      "learning_rate": 1.1276807980049876e-06,
+      "loss": 0.2888,
+      "step": 75710
+    },
+    {
+      "epoch": 18.882793017456358,
+      "grad_norm": 7.766765117645264,
+      "learning_rate": 1.1251870324189527e-06,
+      "loss": 0.3815,
+      "step": 75720
+    },
+    {
+      "epoch": 18.885286783042392,
+      "grad_norm": 6.553428649902344,
+      "learning_rate": 1.1226932668329178e-06,
+      "loss": 0.3442,
+      "step": 75730
+    },
+    {
+      "epoch": 18.887780548628427,
+      "grad_norm": 5.193823337554932,
+      "learning_rate": 1.1201995012468829e-06,
+      "loss": 0.2764,
+      "step": 75740
+    },
+    {
+      "epoch": 18.890274314214462,
+      "grad_norm": 12.545166015625,
+      "learning_rate": 1.117705735660848e-06,
+      "loss": 0.338,
+      "step": 75750
+    },
+    {
+      "epoch": 18.892768079800497,
+      "grad_norm": 9.6239595413208,
+      "learning_rate": 1.115211970074813e-06,
+      "loss": 0.3336,
+      "step": 75760
+    },
+    {
+      "epoch": 18.895261845386532,
+      "grad_norm": 8.860621452331543,
+      "learning_rate": 1.1127182044887782e-06,
+      "loss": 0.2912,
+      "step": 75770
+    },
+    {
+      "epoch": 18.897755610972567,
+      "grad_norm": 10.455620765686035,
+      "learning_rate": 1.1102244389027433e-06,
+      "loss": 0.3473,
+      "step": 75780
+    },
+    {
+      "epoch": 18.900249376558605,
+      "grad_norm": 7.781944751739502,
+      "learning_rate": 1.1077306733167084e-06,
+      "loss": 0.3008,
+      "step": 75790
+    },
+    {
+      "epoch": 18.902743142144637,
+      "grad_norm": 6.780787944793701,
+      "learning_rate": 1.1052369077306735e-06,
+      "loss": 0.2852,
+      "step": 75800
+    },
+    {
+      "epoch": 18.905236907730675,
+      "grad_norm": 11.432109832763672,
+      "learning_rate": 1.1027431421446383e-06,
+      "loss": 0.3317,
+      "step": 75810
+    },
+    {
+      "epoch": 18.90773067331671,
+      "grad_norm": 9.996076583862305,
+      "learning_rate": 1.1002493765586037e-06,
+      "loss": 0.2939,
+      "step": 75820
+    },
+    {
+      "epoch": 18.910224438902745,
+      "grad_norm": 9.775397300720215,
+      "learning_rate": 1.0977556109725688e-06,
+      "loss": 0.3415,
+      "step": 75830
+    },
+    {
+      "epoch": 18.91271820448878,
+      "grad_norm": 10.835949897766113,
+      "learning_rate": 1.0952618453865338e-06,
+      "loss": 0.2963,
+      "step": 75840
+    },
+    {
+      "epoch": 18.915211970074814,
+      "grad_norm": 8.529813766479492,
+      "learning_rate": 1.092768079800499e-06,
+      "loss": 0.2763,
+      "step": 75850
+    },
+    {
+      "epoch": 18.91770573566085,
+      "grad_norm": 7.813754558563232,
+      "learning_rate": 1.090274314214464e-06,
+      "loss": 0.2713,
+      "step": 75860
+    },
+    {
+      "epoch": 18.920199501246884,
+      "grad_norm": 10.183265686035156,
+      "learning_rate": 1.087780548628429e-06,
+      "loss": 0.3235,
+      "step": 75870
+    },
+    {
+      "epoch": 18.92269326683292,
+      "grad_norm": 8.408893585205078,
+      "learning_rate": 1.085286783042394e-06,
+      "loss": 0.3334,
+      "step": 75880
+    },
+    {
+      "epoch": 18.925187032418954,
+      "grad_norm": 11.72058391571045,
+      "learning_rate": 1.0827930174563591e-06,
+      "loss": 0.3118,
+      "step": 75890
+    },
+    {
+      "epoch": 18.92768079800499,
+      "grad_norm": 6.197012424468994,
+      "learning_rate": 1.0802992518703242e-06,
+      "loss": 0.2938,
+      "step": 75900
+    },
+    {
+      "epoch": 18.930174563591024,
+      "grad_norm": 7.970619201660156,
+      "learning_rate": 1.0778054862842893e-06,
+      "loss": 0.3396,
+      "step": 75910
+    },
+    {
+      "epoch": 18.93266832917706,
+      "grad_norm": 10.110791206359863,
+      "learning_rate": 1.0753117206982544e-06,
+      "loss": 0.3286,
+      "step": 75920
+    },
+    {
+      "epoch": 18.935162094763093,
+      "grad_norm": 10.215046882629395,
+      "learning_rate": 1.0728179551122195e-06,
+      "loss": 0.3226,
+      "step": 75930
+    },
+    {
+      "epoch": 18.93765586034913,
+      "grad_norm": 10.97467041015625,
+      "learning_rate": 1.0703241895261846e-06,
+      "loss": 0.2734,
+      "step": 75940
+    },
+    {
+      "epoch": 18.940149625935163,
+      "grad_norm": 8.248761177062988,
+      "learning_rate": 1.0678304239401497e-06,
+      "loss": 0.379,
+      "step": 75950
+    },
+    {
+      "epoch": 18.942643391521198,
+      "grad_norm": 8.312829971313477,
+      "learning_rate": 1.0653366583541148e-06,
+      "loss": 0.3253,
+      "step": 75960
+    },
+    {
+      "epoch": 18.945137157107233,
+      "grad_norm": 4.772835731506348,
+      "learning_rate": 1.06284289276808e-06,
+      "loss": 0.3596,
+      "step": 75970
+    },
+    {
+      "epoch": 18.947630922693268,
+      "grad_norm": 10.931731224060059,
+      "learning_rate": 1.060349127182045e-06,
+      "loss": 0.3222,
+      "step": 75980
+    },
+    {
+      "epoch": 18.950124688279303,
+      "grad_norm": 10.215840339660645,
+      "learning_rate": 1.05785536159601e-06,
+      "loss": 0.3021,
+      "step": 75990
+    },
+    {
+      "epoch": 18.952618453865338,
+      "grad_norm": 9.405722618103027,
+      "learning_rate": 1.0553615960099752e-06,
+      "loss": 0.2664,
+      "step": 76000
+    },
+    {
+      "epoch": 18.955112219451372,
+      "grad_norm": 7.368799209594727,
+      "learning_rate": 1.0528678304239403e-06,
+      "loss": 0.3743,
+      "step": 76010
+    },
+    {
+      "epoch": 18.957605985037407,
+      "grad_norm": 6.817213535308838,
+      "learning_rate": 1.0503740648379054e-06,
+      "loss": 0.3206,
+      "step": 76020
+    },
+    {
+      "epoch": 18.960099750623442,
+      "grad_norm": 12.117660522460938,
+      "learning_rate": 1.0478802992518703e-06,
+      "loss": 0.3007,
+      "step": 76030
+    },
+    {
+      "epoch": 18.962593516209477,
+      "grad_norm": 12.789581298828125,
+      "learning_rate": 1.0453865336658354e-06,
+      "loss": 0.2822,
+      "step": 76040
+    },
+    {
+      "epoch": 18.965087281795512,
+      "grad_norm": 11.194233894348145,
+      "learning_rate": 1.0428927680798005e-06,
+      "loss": 0.3346,
+      "step": 76050
+    },
+    {
+      "epoch": 18.967581047381547,
+      "grad_norm": 7.340683937072754,
+      "learning_rate": 1.0403990024937656e-06,
+      "loss": 0.2921,
+      "step": 76060
+    },
+    {
+      "epoch": 18.97007481296758,
+      "grad_norm": 8.430867195129395,
+      "learning_rate": 1.0379052369077309e-06,
+      "loss": 0.2891,
+      "step": 76070
+    },
+    {
+      "epoch": 18.972568578553616,
+      "grad_norm": 8.543846130371094,
+      "learning_rate": 1.035411471321696e-06,
+      "loss": 0.2746,
+      "step": 76080
+    },
+    {
+      "epoch": 18.97506234413965,
+      "grad_norm": 7.078313827514648,
+      "learning_rate": 1.0329177057356609e-06,
+      "loss": 0.2897,
+      "step": 76090
+    },
+    {
+      "epoch": 18.977556109725686,
+      "grad_norm": 5.388221263885498,
+      "learning_rate": 1.030423940149626e-06,
+      "loss": 0.2994,
+      "step": 76100
+    },
+    {
+      "epoch": 18.98004987531172,
+      "grad_norm": 7.963678359985352,
+      "learning_rate": 1.027930174563591e-06,
+      "loss": 0.3703,
+      "step": 76110
+    },
+    {
+      "epoch": 18.982543640897756,
+      "grad_norm": 8.849908828735352,
+      "learning_rate": 1.0254364089775562e-06,
+      "loss": 0.3534,
+      "step": 76120
+    },
+    {
+      "epoch": 18.98503740648379,
+      "grad_norm": 8.504463195800781,
+      "learning_rate": 1.0229426433915213e-06,
+      "loss": 0.3156,
+      "step": 76130
+    },
+    {
+      "epoch": 18.987531172069826,
+      "grad_norm": 7.680803298950195,
+      "learning_rate": 1.0204488778054864e-06,
+      "loss": 0.3605,
+      "step": 76140
+    },
+    {
+      "epoch": 18.99002493765586,
+      "grad_norm": 8.039253234863281,
+      "learning_rate": 1.0179551122194515e-06,
+      "loss": 0.3321,
+      "step": 76150
+    },
+    {
+      "epoch": 18.992518703241895,
+      "grad_norm": 6.414565563201904,
+      "learning_rate": 1.0154613466334166e-06,
+      "loss": 0.3486,
+      "step": 76160
+    },
+    {
+      "epoch": 18.99501246882793,
+      "grad_norm": 10.342339515686035,
+      "learning_rate": 1.0129675810473817e-06,
+      "loss": 0.3262,
+      "step": 76170
+    },
+    {
+      "epoch": 18.997506234413965,
+      "grad_norm": 10.92263412475586,
+      "learning_rate": 1.0104738154613468e-06,
+      "loss": 0.3949,
+      "step": 76180
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 9.860681533813477,
+      "learning_rate": 1.0079800498753119e-06,
+      "loss": 0.3274,
+      "step": 76190
+    },
+    {
+      "epoch": 19.0,
+      "eval_loss": 0.4184245765209198,
+      "eval_runtime": 60.1429,
+      "eval_samples_per_second": 16.677,
+      "eval_steps_per_second": 16.677,
+      "step": 76190
+    },
+    {
+      "epoch": 19.002493765586035,
+      "grad_norm": 6.654420852661133,
+      "learning_rate": 1.005486284289277e-06,
+      "loss": 0.3282,
+      "step": 76200
+    },
+    {
+      "epoch": 19.00498753117207,
+      "grad_norm": 11.368739128112793,
+      "learning_rate": 1.002992518703242e-06,
+      "loss": 0.3369,
+      "step": 76210
+    },
+    {
+      "epoch": 19.007481296758105,
+      "grad_norm": 6.13724422454834,
+      "learning_rate": 1.0004987531172071e-06,
+      "loss": 0.3149,
+      "step": 76220
+    },
+    {
+      "epoch": 19.00997506234414,
+      "grad_norm": 9.071732521057129,
+      "learning_rate": 9.980049875311722e-07,
+      "loss": 0.3517,
+      "step": 76230
+    },
+    {
+      "epoch": 19.012468827930174,
+      "grad_norm": 9.942281723022461,
+      "learning_rate": 9.955112219451373e-07,
+      "loss": 0.3144,
+      "step": 76240
+    },
+    {
+      "epoch": 19.01496259351621,
+      "grad_norm": 7.623038291931152,
+      "learning_rate": 9.930174563591022e-07,
+      "loss": 0.3624,
+      "step": 76250
+    },
+    {
+      "epoch": 19.017456359102244,
+      "grad_norm": 8.054025650024414,
+      "learning_rate": 9.905236907730673e-07,
+      "loss": 0.3465,
+      "step": 76260
+    },
+    {
+      "epoch": 19.01995012468828,
+      "grad_norm": 8.097540855407715,
+      "learning_rate": 9.880299251870324e-07,
+      "loss": 0.2926,
+      "step": 76270
+    },
+    {
+      "epoch": 19.022443890274314,
+      "grad_norm": 8.881487846374512,
+      "learning_rate": 9.855361596009975e-07,
+      "loss": 0.3272,
+      "step": 76280
+    },
+    {
+      "epoch": 19.02493765586035,
+      "grad_norm": 9.261902809143066,
+      "learning_rate": 9.830423940149626e-07,
+      "loss": 0.3085,
+      "step": 76290
+    },
+    {
+      "epoch": 19.027431421446384,
+      "grad_norm": 8.03958511352539,
+      "learning_rate": 9.805486284289277e-07,
+      "loss": 0.2904,
+      "step": 76300
+    },
+    {
+      "epoch": 19.02992518703242,
+      "grad_norm": 7.877828121185303,
+      "learning_rate": 9.780548628428928e-07,
+      "loss": 0.3262,
+      "step": 76310
+    },
+    {
+      "epoch": 19.032418952618453,
+      "grad_norm": 8.608955383300781,
+      "learning_rate": 9.75561097256858e-07,
+      "loss": 0.3105,
+      "step": 76320
+    },
+    {
+      "epoch": 19.034912718204488,
+      "grad_norm": 5.959877014160156,
+      "learning_rate": 9.73067331670823e-07,
+      "loss": 0.2964,
+      "step": 76330
+    },
+    {
+      "epoch": 19.037406483790523,
+      "grad_norm": 7.390963554382324,
+      "learning_rate": 9.705735660847881e-07,
+      "loss": 0.3047,
+      "step": 76340
+    },
+    {
+      "epoch": 19.039900249376558,
+      "grad_norm": 7.07940149307251,
+      "learning_rate": 9.680798004987532e-07,
+      "loss": 0.2936,
+      "step": 76350
+    },
+    {
+      "epoch": 19.042394014962593,
+      "grad_norm": 10.116901397705078,
+      "learning_rate": 9.655860349127183e-07,
+      "loss": 0.3208,
+      "step": 76360
+    },
+    {
+      "epoch": 19.044887780548628,
+      "grad_norm": 6.5279717445373535,
+      "learning_rate": 9.630922693266834e-07,
+      "loss": 0.3736,
+      "step": 76370
+    },
+    {
+      "epoch": 19.047381546134662,
+      "grad_norm": 7.244894981384277,
+      "learning_rate": 9.605985037406485e-07,
+      "loss": 0.3719,
+      "step": 76380
+    },
+    {
+      "epoch": 19.049875311720697,
+      "grad_norm": 9.443936347961426,
+      "learning_rate": 9.581047381546136e-07,
+      "loss": 0.2698,
+      "step": 76390
+    },
+    {
+      "epoch": 19.052369077306732,
+      "grad_norm": 7.093422889709473,
+      "learning_rate": 9.556109725685787e-07,
+      "loss": 0.3562,
+      "step": 76400
+    },
+    {
+      "epoch": 19.054862842892767,
+      "grad_norm": 8.13650131225586,
+      "learning_rate": 9.531172069825437e-07,
+      "loss": 0.3588,
+      "step": 76410
+    },
+    {
+      "epoch": 19.057356608478802,
+      "grad_norm": 8.89100170135498,
+      "learning_rate": 9.506234413965088e-07,
+      "loss": 0.2875,
+      "step": 76420
+    },
+    {
+      "epoch": 19.059850374064837,
+      "grad_norm": 12.49089527130127,
+      "learning_rate": 9.481296758104738e-07,
+      "loss": 0.2841,
+      "step": 76430
+    },
+    {
+      "epoch": 19.06234413965087,
+      "grad_norm": 8.83289623260498,
+      "learning_rate": 9.456359102244391e-07,
+      "loss": 0.3851,
+      "step": 76440
+    },
+    {
+      "epoch": 19.064837905236907,
+      "grad_norm": 8.544670104980469,
+      "learning_rate": 9.431421446384041e-07,
+      "loss": 0.309,
+      "step": 76450
+    },
+    {
+      "epoch": 19.06733167082294,
+      "grad_norm": 13.055142402648926,
+      "learning_rate": 9.406483790523692e-07,
+      "loss": 0.3448,
+      "step": 76460
+    },
+    {
+      "epoch": 19.069825436408976,
+      "grad_norm": 5.936143398284912,
+      "learning_rate": 9.381546134663343e-07,
+      "loss": 0.2878,
+      "step": 76470
+    },
+    {
+      "epoch": 19.07231920199501,
+      "grad_norm": 9.928505897521973,
+      "learning_rate": 9.356608478802994e-07,
+      "loss": 0.2742,
+      "step": 76480
+    },
+    {
+      "epoch": 19.074812967581046,
+      "grad_norm": 10.485602378845215,
+      "learning_rate": 9.331670822942644e-07,
+      "loss": 0.2967,
+      "step": 76490
+    },
+    {
+      "epoch": 19.07730673316708,
+      "grad_norm": 9.832148551940918,
+      "learning_rate": 9.306733167082295e-07,
+      "loss": 0.3564,
+      "step": 76500
+    },
+    {
+      "epoch": 19.079800498753116,
+      "grad_norm": 5.720829963684082,
+      "learning_rate": 9.281795511221946e-07,
+      "loss": 0.3261,
+      "step": 76510
+    },
+    {
+      "epoch": 19.08229426433915,
+      "grad_norm": 9.377291679382324,
+      "learning_rate": 9.256857855361597e-07,
+      "loss": 0.3256,
+      "step": 76520
+    },
+    {
+      "epoch": 19.084788029925186,
+      "grad_norm": 8.832295417785645,
+      "learning_rate": 9.231920199501248e-07,
+      "loss": 0.3498,
+      "step": 76530
+    },
+    {
+      "epoch": 19.08728179551122,
+      "grad_norm": 7.698011875152588,
+      "learning_rate": 9.206982543640898e-07,
+      "loss": 0.2729,
+      "step": 76540
+    },
+    {
+      "epoch": 19.089775561097255,
+      "grad_norm": 7.579693794250488,
+      "learning_rate": 9.182044887780549e-07,
+      "loss": 0.322,
+      "step": 76550
+    },
+    {
+      "epoch": 19.09226932668329,
+      "grad_norm": 8.33119010925293,
+      "learning_rate": 9.157107231920201e-07,
+      "loss": 0.2915,
+      "step": 76560
+    },
+    {
+      "epoch": 19.094763092269325,
+      "grad_norm": 11.253091812133789,
+      "learning_rate": 9.132169576059852e-07,
+      "loss": 0.3446,
+      "step": 76570
+    },
+    {
+      "epoch": 19.09725685785536,
+      "grad_norm": 9.968659400939941,
+      "learning_rate": 9.107231920199502e-07,
+      "loss": 0.2651,
+      "step": 76580
+    },
+    {
+      "epoch": 19.099750623441395,
+      "grad_norm": 10.114205360412598,
+      "learning_rate": 9.082294264339153e-07,
+      "loss": 0.3819,
+      "step": 76590
+    },
+    {
+      "epoch": 19.102244389027433,
+      "grad_norm": 10.0556058883667,
+      "learning_rate": 9.057356608478804e-07,
+      "loss": 0.3939,
+      "step": 76600
+    },
+    {
+      "epoch": 19.104738154613468,
+      "grad_norm": 6.652792930603027,
+      "learning_rate": 9.032418952618454e-07,
+      "loss": 0.4161,
+      "step": 76610
+    },
+    {
+      "epoch": 19.107231920199503,
+      "grad_norm": 8.197524070739746,
+      "learning_rate": 9.007481296758105e-07,
+      "loss": 0.3817,
+      "step": 76620
+    },
+    {
+      "epoch": 19.109725685785538,
+      "grad_norm": 9.194473266601562,
+      "learning_rate": 8.982543640897756e-07,
+      "loss": 0.3205,
+      "step": 76630
+    },
+    {
+      "epoch": 19.112219451371573,
+      "grad_norm": 9.719461441040039,
+      "learning_rate": 8.957605985037407e-07,
+      "loss": 0.3286,
+      "step": 76640
+    },
+    {
+      "epoch": 19.114713216957608,
+      "grad_norm": 5.662110328674316,
+      "learning_rate": 8.932668329177057e-07,
+      "loss": 0.3241,
+      "step": 76650
+    },
+    {
+      "epoch": 19.117206982543642,
+      "grad_norm": 7.443528652191162,
+      "learning_rate": 8.907730673316708e-07,
+      "loss": 0.3294,
+      "step": 76660
+    },
+    {
+      "epoch": 19.119700748129677,
+      "grad_norm": 8.850680351257324,
+      "learning_rate": 8.882793017456359e-07,
+      "loss": 0.4117,
+      "step": 76670
+    },
+    {
+      "epoch": 19.122194513715712,
+      "grad_norm": 9.90893840789795,
+      "learning_rate": 8.857855361596011e-07,
+      "loss": 0.3235,
+      "step": 76680
+    },
+    {
+      "epoch": 19.124688279301747,
+      "grad_norm": 6.257623672485352,
+      "learning_rate": 8.832917705735662e-07,
+      "loss": 0.2876,
+      "step": 76690
+    },
+    {
+      "epoch": 19.127182044887782,
+      "grad_norm": 10.78929615020752,
+      "learning_rate": 8.807980049875313e-07,
+      "loss": 0.2937,
+      "step": 76700
+    },
+    {
+      "epoch": 19.129675810473817,
+      "grad_norm": 12.467625617980957,
+      "learning_rate": 8.783042394014964e-07,
+      "loss": 0.2836,
+      "step": 76710
+    },
+    {
+      "epoch": 19.13216957605985,
+      "grad_norm": 9.467742919921875,
+      "learning_rate": 8.758104738154614e-07,
+      "loss": 0.2906,
+      "step": 76720
+    },
+    {
+      "epoch": 19.134663341645886,
+      "grad_norm": 7.464754581451416,
+      "learning_rate": 8.733167082294265e-07,
+      "loss": 0.3037,
+      "step": 76730
+    },
+    {
+      "epoch": 19.13715710723192,
+      "grad_norm": 9.207623481750488,
+      "learning_rate": 8.708229426433916e-07,
+      "loss": 0.3483,
+      "step": 76740
+    },
+    {
+      "epoch": 19.139650872817956,
+      "grad_norm": 5.893961429595947,
+      "learning_rate": 8.683291770573567e-07,
+      "loss": 0.327,
+      "step": 76750
+    },
+    {
+      "epoch": 19.14214463840399,
+      "grad_norm": 7.405910015106201,
+      "learning_rate": 8.658354114713217e-07,
+      "loss": 0.2723,
+      "step": 76760
+    },
+    {
+      "epoch": 19.144638403990026,
+      "grad_norm": 6.716821670532227,
+      "learning_rate": 8.633416458852868e-07,
+      "loss": 0.3184,
+      "step": 76770
+    },
+    {
+      "epoch": 19.14713216957606,
+      "grad_norm": 10.331329345703125,
+      "learning_rate": 8.608478802992519e-07,
+      "loss": 0.3315,
+      "step": 76780
+    },
+    {
+      "epoch": 19.149625935162096,
+      "grad_norm": 11.650618553161621,
+      "learning_rate": 8.58354114713217e-07,
+      "loss": 0.3256,
+      "step": 76790
+    },
+    {
+      "epoch": 19.15211970074813,
+      "grad_norm": 10.218818664550781,
+      "learning_rate": 8.558603491271821e-07,
+      "loss": 0.3823,
+      "step": 76800
+    },
+    {
+      "epoch": 19.154613466334165,
+      "grad_norm": 11.632461547851562,
+      "learning_rate": 8.533665835411473e-07,
+      "loss": 0.2996,
+      "step": 76810
+    },
+    {
+      "epoch": 19.1571072319202,
+      "grad_norm": 7.20107364654541,
+      "learning_rate": 8.508728179551124e-07,
+      "loss": 0.2705,
+      "step": 76820
+    },
+    {
+      "epoch": 19.159600997506235,
+      "grad_norm": 7.080248832702637,
+      "learning_rate": 8.483790523690774e-07,
+      "loss": 0.3582,
+      "step": 76830
+    },
+    {
+      "epoch": 19.16209476309227,
+      "grad_norm": 7.180111885070801,
+      "learning_rate": 8.458852867830425e-07,
+      "loss": 0.3788,
+      "step": 76840
+    },
+    {
+      "epoch": 19.164588528678305,
+      "grad_norm": 7.654664516448975,
+      "learning_rate": 8.433915211970076e-07,
+      "loss": 0.3894,
+      "step": 76850
+    },
+    {
+      "epoch": 19.16708229426434,
+      "grad_norm": 7.925827980041504,
+      "learning_rate": 8.408977556109727e-07,
+      "loss": 0.3759,
+      "step": 76860
+    },
+    {
+      "epoch": 19.169576059850375,
+      "grad_norm": 9.838397979736328,
+      "learning_rate": 8.384039900249377e-07,
+      "loss": 0.2958,
+      "step": 76870
+    },
+    {
+      "epoch": 19.17206982543641,
+      "grad_norm": 10.40937328338623,
+      "learning_rate": 8.359102244389028e-07,
+      "loss": 0.3657,
+      "step": 76880
+    },
+    {
+      "epoch": 19.174563591022444,
+      "grad_norm": 9.061084747314453,
+      "learning_rate": 8.334164588528679e-07,
+      "loss": 0.3918,
+      "step": 76890
+    },
+    {
+      "epoch": 19.17705735660848,
+      "grad_norm": 6.745436668395996,
+      "learning_rate": 8.30922693266833e-07,
+      "loss": 0.314,
+      "step": 76900
+    },
+    {
+      "epoch": 19.179551122194514,
+      "grad_norm": 9.674360275268555,
+      "learning_rate": 8.284289276807981e-07,
+      "loss": 0.3089,
+      "step": 76910
+    },
+    {
+      "epoch": 19.18204488778055,
+      "grad_norm": 7.983299732208252,
+      "learning_rate": 8.25935162094763e-07,
+      "loss": 0.3077,
+      "step": 76920
+    },
+    {
+      "epoch": 19.184538653366584,
+      "grad_norm": 8.201288223266602,
+      "learning_rate": 8.234413965087284e-07,
+      "loss": 0.3265,
+      "step": 76930
+    },
+    {
+      "epoch": 19.18703241895262,
+      "grad_norm": 6.199214935302734,
+      "learning_rate": 8.209476309226934e-07,
+      "loss": 0.2781,
+      "step": 76940
+    },
+    {
+      "epoch": 19.189526184538654,
+      "grad_norm": 7.515051364898682,
+      "learning_rate": 8.184538653366585e-07,
+      "loss": 0.2741,
+      "step": 76950
+    },
+    {
+      "epoch": 19.19201995012469,
+      "grad_norm": 7.1991658210754395,
+      "learning_rate": 8.159600997506235e-07,
+      "loss": 0.3119,
+      "step": 76960
+    },
+    {
+      "epoch": 19.194513715710723,
+      "grad_norm": 13.45710277557373,
+      "learning_rate": 8.134663341645886e-07,
+      "loss": 0.3671,
+      "step": 76970
+    },
+    {
+      "epoch": 19.197007481296758,
+      "grad_norm": 6.285548210144043,
+      "learning_rate": 8.109725685785537e-07,
+      "loss": 0.2633,
+      "step": 76980
+    },
+    {
+      "epoch": 19.199501246882793,
+      "grad_norm": 13.47431755065918,
+      "learning_rate": 8.087281795511223e-07,
+      "loss": 0.3932,
+      "step": 76990
+    },
+    {
+      "epoch": 19.201995012468828,
+      "grad_norm": 11.600540161132812,
+      "learning_rate": 8.062344139650874e-07,
+      "loss": 0.3936,
+      "step": 77000
+    },
+    {
+      "epoch": 19.204488778054863,
+      "grad_norm": 8.96855640411377,
+      "learning_rate": 8.037406483790524e-07,
+      "loss": 0.361,
+      "step": 77010
+    },
+    {
+      "epoch": 19.206982543640898,
+      "grad_norm": 10.361422538757324,
+      "learning_rate": 8.012468827930175e-07,
+      "loss": 0.3356,
+      "step": 77020
+    },
+    {
+      "epoch": 19.209476309226932,
+      "grad_norm": 7.201541423797607,
+      "learning_rate": 7.987531172069826e-07,
+      "loss": 0.3742,
+      "step": 77030
+    },
+    {
+      "epoch": 19.211970074812967,
+      "grad_norm": 8.441081047058105,
+      "learning_rate": 7.962593516209477e-07,
+      "loss": 0.3109,
+      "step": 77040
+    },
+    {
+      "epoch": 19.214463840399002,
+      "grad_norm": 7.7734174728393555,
+      "learning_rate": 7.937655860349128e-07,
+      "loss": 0.2973,
+      "step": 77050
+    },
+    {
+      "epoch": 19.216957605985037,
+      "grad_norm": 10.10339641571045,
+      "learning_rate": 7.912718204488778e-07,
+      "loss": 0.2633,
+      "step": 77060
+    },
+    {
+      "epoch": 19.219451371571072,
+      "grad_norm": 11.610758781433105,
+      "learning_rate": 7.887780548628429e-07,
+      "loss": 0.3264,
+      "step": 77070
+    },
+    {
+      "epoch": 19.221945137157107,
+      "grad_norm": 11.709192276000977,
+      "learning_rate": 7.86284289276808e-07,
+      "loss": 0.3189,
+      "step": 77080
+    },
+    {
+      "epoch": 19.22443890274314,
+      "grad_norm": 6.54279899597168,
+      "learning_rate": 7.837905236907731e-07,
+      "loss": 0.2883,
+      "step": 77090
+    },
+    {
+      "epoch": 19.226932668329177,
+      "grad_norm": 11.589969635009766,
+      "learning_rate": 7.812967581047382e-07,
+      "loss": 0.2831,
+      "step": 77100
+    },
+    {
+      "epoch": 19.22942643391521,
+      "grad_norm": 6.521005153656006,
+      "learning_rate": 7.788029925187034e-07,
+      "loss": 0.2952,
+      "step": 77110
+    },
+    {
+      "epoch": 19.231920199501246,
+      "grad_norm": 6.133444309234619,
+      "learning_rate": 7.763092269326684e-07,
+      "loss": 0.2859,
+      "step": 77120
+    },
+    {
+      "epoch": 19.23441396508728,
+      "grad_norm": 7.011719703674316,
+      "learning_rate": 7.738154613466335e-07,
+      "loss": 0.3177,
+      "step": 77130
+    },
+    {
+      "epoch": 19.236907730673316,
+      "grad_norm": 6.836624622344971,
+      "learning_rate": 7.713216957605986e-07,
+      "loss": 0.3955,
+      "step": 77140
+    },
+    {
+      "epoch": 19.23940149625935,
+      "grad_norm": 11.907731056213379,
+      "learning_rate": 7.688279301745637e-07,
+      "loss": 0.2778,
+      "step": 77150
+    },
+    {
+      "epoch": 19.241895261845386,
+      "grad_norm": 11.505409240722656,
+      "learning_rate": 7.663341645885288e-07,
+      "loss": 0.366,
+      "step": 77160
+    },
+    {
+      "epoch": 19.24438902743142,
+      "grad_norm": 10.92941951751709,
+      "learning_rate": 7.638403990024938e-07,
+      "loss": 0.3527,
+      "step": 77170
+    },
+    {
+      "epoch": 19.246882793017456,
+      "grad_norm": 8.509658813476562,
+      "learning_rate": 7.613466334164589e-07,
+      "loss": 0.269,
+      "step": 77180
+    },
+    {
+      "epoch": 19.24937655860349,
+      "grad_norm": 6.147039413452148,
+      "learning_rate": 7.58852867830424e-07,
+      "loss": 0.3095,
+      "step": 77190
+    },
+    {
+      "epoch": 19.251870324189525,
+      "grad_norm": 9.925485610961914,
+      "learning_rate": 7.563591022443891e-07,
+      "loss": 0.3432,
+      "step": 77200
+    },
+    {
+      "epoch": 19.25436408977556,
+      "grad_norm": 7.7526116371154785,
+      "learning_rate": 7.538653366583542e-07,
+      "loss": 0.3582,
+      "step": 77210
+    },
+    {
+      "epoch": 19.256857855361595,
+      "grad_norm": 9.207817077636719,
+      "learning_rate": 7.513715710723192e-07,
+      "loss": 0.2891,
+      "step": 77220
+    },
+    {
+      "epoch": 19.25935162094763,
+      "grad_norm": 8.330227851867676,
+      "learning_rate": 7.488778054862844e-07,
+      "loss": 0.2408,
+      "step": 77230
+    },
+    {
+      "epoch": 19.261845386533665,
+      "grad_norm": 7.16584587097168,
+      "learning_rate": 7.463840399002495e-07,
+      "loss": 0.2844,
+      "step": 77240
+    },
+    {
+      "epoch": 19.2643391521197,
+      "grad_norm": 7.896230220794678,
+      "learning_rate": 7.438902743142146e-07,
+      "loss": 0.3163,
+      "step": 77250
+    },
+    {
+      "epoch": 19.266832917705734,
+      "grad_norm": 6.485842704772949,
+      "learning_rate": 7.413965087281797e-07,
+      "loss": 0.3416,
+      "step": 77260
+    },
+    {
+      "epoch": 19.26932668329177,
+      "grad_norm": 10.570438385009766,
+      "learning_rate": 7.389027431421448e-07,
+      "loss": 0.3151,
+      "step": 77270
+    },
+    {
+      "epoch": 19.271820448877804,
+      "grad_norm": 10.489306449890137,
+      "learning_rate": 7.364089775561097e-07,
+      "loss": 0.3103,
+      "step": 77280
+    },
+    {
+      "epoch": 19.27431421446384,
+      "grad_norm": 17.813852310180664,
+      "learning_rate": 7.339152119700748e-07,
+      "loss": 0.3384,
+      "step": 77290
+    },
+    {
+      "epoch": 19.276807980049874,
+      "grad_norm": 6.28749418258667,
+      "learning_rate": 7.314214463840399e-07,
+      "loss": 0.3067,
+      "step": 77300
+    },
+    {
+      "epoch": 19.27930174563591,
+      "grad_norm": 10.992226600646973,
+      "learning_rate": 7.28927680798005e-07,
+      "loss": 0.3487,
+      "step": 77310
+    },
+    {
+      "epoch": 19.281795511221944,
+      "grad_norm": 8.586907386779785,
+      "learning_rate": 7.264339152119701e-07,
+      "loss": 0.2888,
+      "step": 77320
+    },
+    {
+      "epoch": 19.28428927680798,
+      "grad_norm": 8.104104995727539,
+      "learning_rate": 7.239401496259351e-07,
+      "loss": 0.2809,
+      "step": 77330
+    },
+    {
+      "epoch": 19.286783042394013,
+      "grad_norm": 9.069332122802734,
+      "learning_rate": 7.214463840399002e-07,
+      "loss": 0.2921,
+      "step": 77340
+    },
+    {
+      "epoch": 19.28927680798005,
+      "grad_norm": 8.148301124572754,
+      "learning_rate": 7.189526184538654e-07,
+      "loss": 0.4029,
+      "step": 77350
+    },
+    {
+      "epoch": 19.291770573566083,
+      "grad_norm": 7.358614921569824,
+      "learning_rate": 7.164588528678305e-07,
+      "loss": 0.3364,
+      "step": 77360
+    },
+    {
+      "epoch": 19.294264339152118,
+      "grad_norm": 11.08927059173584,
+      "learning_rate": 7.139650872817956e-07,
+      "loss": 0.3398,
+      "step": 77370
+    },
+    {
+      "epoch": 19.296758104738153,
+      "grad_norm": 10.33334732055664,
+      "learning_rate": 7.114713216957607e-07,
+      "loss": 0.326,
+      "step": 77380
+    },
+    {
+      "epoch": 19.29925187032419,
+      "grad_norm": 10.560935974121094,
+      "learning_rate": 7.089775561097257e-07,
+      "loss": 0.3557,
+      "step": 77390
+    },
+    {
+      "epoch": 19.301745635910226,
+      "grad_norm": 8.311334609985352,
+      "learning_rate": 7.064837905236908e-07,
+      "loss": 0.3314,
+      "step": 77400
+    },
+    {
+      "epoch": 19.30423940149626,
+      "grad_norm": 11.71918773651123,
+      "learning_rate": 7.039900249376559e-07,
+      "loss": 0.3246,
+      "step": 77410
+    },
+    {
+      "epoch": 19.306733167082296,
+      "grad_norm": 8.50728988647461,
+      "learning_rate": 7.01496259351621e-07,
+      "loss": 0.3343,
+      "step": 77420
+    },
+    {
+      "epoch": 19.30922693266833,
+      "grad_norm": 11.546276092529297,
+      "learning_rate": 6.990024937655861e-07,
+      "loss": 0.2922,
+      "step": 77430
+    },
+    {
+      "epoch": 19.311720698254366,
+      "grad_norm": 10.145132064819336,
+      "learning_rate": 6.965087281795511e-07,
+      "loss": 0.3306,
+      "step": 77440
+    },
+    {
+      "epoch": 19.3142144638404,
+      "grad_norm": 12.020730018615723,
+      "learning_rate": 6.940149625935162e-07,
+      "loss": 0.2949,
+      "step": 77450
+    },
+    {
+      "epoch": 19.316708229426435,
+      "grad_norm": 13.925124168395996,
+      "learning_rate": 6.915211970074813e-07,
+      "loss": 0.4027,
+      "step": 77460
+    },
+    {
+      "epoch": 19.31920199501247,
+      "grad_norm": 9.396332740783691,
+      "learning_rate": 6.890274314214464e-07,
+      "loss": 0.283,
+      "step": 77470
+    },
+    {
+      "epoch": 19.321695760598505,
+      "grad_norm": 5.274721622467041,
+      "learning_rate": 6.865336658354116e-07,
+      "loss": 0.3043,
+      "step": 77480
+    },
+    {
+      "epoch": 19.32418952618454,
+      "grad_norm": 7.85326623916626,
+      "learning_rate": 6.840399002493767e-07,
+      "loss": 0.3407,
+      "step": 77490
+    },
+    {
+      "epoch": 19.326683291770575,
+      "grad_norm": 12.338591575622559,
+      "learning_rate": 6.815461346633417e-07,
+      "loss": 0.2622,
+      "step": 77500
+    },
+    {
+      "epoch": 19.32917705735661,
+      "grad_norm": 7.619169235229492,
+      "learning_rate": 6.790523690773068e-07,
+      "loss": 0.2779,
+      "step": 77510
+    },
+    {
+      "epoch": 19.331670822942645,
+      "grad_norm": 8.566308975219727,
+      "learning_rate": 6.768079800498753e-07,
+      "loss": 0.3057,
+      "step": 77520
+    },
+    {
+      "epoch": 19.33416458852868,
+      "grad_norm": 11.934048652648926,
+      "learning_rate": 6.743142144638405e-07,
+      "loss": 0.3626,
+      "step": 77530
+    },
+    {
+      "epoch": 19.336658354114714,
+      "grad_norm": 12.807787895202637,
+      "learning_rate": 6.718204488778056e-07,
+      "loss": 0.297,
+      "step": 77540
+    },
+    {
+      "epoch": 19.33915211970075,
+      "grad_norm": 8.593514442443848,
+      "learning_rate": 6.693266832917707e-07,
+      "loss": 0.3174,
+      "step": 77550
+    },
+    {
+      "epoch": 19.341645885286784,
+      "grad_norm": 10.601576805114746,
+      "learning_rate": 6.668329177057358e-07,
+      "loss": 0.3191,
+      "step": 77560
+    },
+    {
+      "epoch": 19.34413965087282,
+      "grad_norm": 10.128413200378418,
+      "learning_rate": 6.643391521197009e-07,
+      "loss": 0.302,
+      "step": 77570
+    },
+    {
+      "epoch": 19.346633416458854,
+      "grad_norm": 8.060428619384766,
+      "learning_rate": 6.618453865336659e-07,
+      "loss": 0.2979,
+      "step": 77580
+    },
+    {
+      "epoch": 19.34912718204489,
+      "grad_norm": 5.3289079666137695,
+      "learning_rate": 6.59351620947631e-07,
+      "loss": 0.2762,
+      "step": 77590
+    },
+    {
+      "epoch": 19.351620947630924,
+      "grad_norm": 6.777313709259033,
+      "learning_rate": 6.568578553615961e-07,
+      "loss": 0.2989,
+      "step": 77600
+    },
+    {
+      "epoch": 19.35411471321696,
+      "grad_norm": 6.17332124710083,
+      "learning_rate": 6.543640897755612e-07,
+      "loss": 0.2851,
+      "step": 77610
+    },
+    {
+      "epoch": 19.356608478802993,
+      "grad_norm": 9.076419830322266,
+      "learning_rate": 6.518703241895261e-07,
+      "loss": 0.325,
+      "step": 77620
+    },
+    {
+      "epoch": 19.359102244389028,
+      "grad_norm": 9.705153465270996,
+      "learning_rate": 6.493765586034912e-07,
+      "loss": 0.2434,
+      "step": 77630
+    },
+    {
+      "epoch": 19.361596009975063,
+      "grad_norm": 11.839200973510742,
+      "learning_rate": 6.468827930174563e-07,
+      "loss": 0.3439,
+      "step": 77640
+    },
+    {
+      "epoch": 19.364089775561098,
+      "grad_norm": 10.060364723205566,
+      "learning_rate": 6.443890274314215e-07,
+      "loss": 0.3398,
+      "step": 77650
+    },
+    {
+      "epoch": 19.366583541147133,
+      "grad_norm": 4.9707560539245605,
+      "learning_rate": 6.418952618453866e-07,
+      "loss": 0.293,
+      "step": 77660
+    },
+    {
+      "epoch": 19.369077306733168,
+      "grad_norm": 7.344512462615967,
+      "learning_rate": 6.394014962593517e-07,
+      "loss": 0.3303,
+      "step": 77670
+    },
+    {
+      "epoch": 19.371571072319203,
+      "grad_norm": 8.75753402709961,
+      "learning_rate": 6.369077306733168e-07,
+      "loss": 0.3225,
+      "step": 77680
+    },
+    {
+      "epoch": 19.374064837905237,
+      "grad_norm": 8.907544136047363,
+      "learning_rate": 6.344139650872818e-07,
+      "loss": 0.4169,
+      "step": 77690
+    },
+    {
+      "epoch": 19.376558603491272,
+      "grad_norm": 8.724864959716797,
+      "learning_rate": 6.319201995012469e-07,
+      "loss": 0.4323,
+      "step": 77700
+    },
+    {
+      "epoch": 19.379052369077307,
+      "grad_norm": 10.523849487304688,
+      "learning_rate": 6.29426433915212e-07,
+      "loss": 0.3098,
+      "step": 77710
+    },
+    {
+      "epoch": 19.381546134663342,
+      "grad_norm": 6.705490589141846,
+      "learning_rate": 6.269326683291771e-07,
+      "loss": 0.2966,
+      "step": 77720
+    },
+    {
+      "epoch": 19.384039900249377,
+      "grad_norm": 9.971707344055176,
+      "learning_rate": 6.244389027431422e-07,
+      "loss": 0.3189,
+      "step": 77730
+    },
+    {
+      "epoch": 19.38653366583541,
+      "grad_norm": 10.387377738952637,
+      "learning_rate": 6.219451371571072e-07,
+      "loss": 0.3321,
+      "step": 77740
+    },
+    {
+      "epoch": 19.389027431421447,
+      "grad_norm": 8.39964485168457,
+      "learning_rate": 6.194513715710724e-07,
+      "loss": 0.2786,
+      "step": 77750
+    },
+    {
+      "epoch": 19.39152119700748,
+      "grad_norm": 7.76651668548584,
+      "learning_rate": 6.169576059850375e-07,
+      "loss": 0.3166,
+      "step": 77760
+    },
+    {
+      "epoch": 19.394014962593516,
+      "grad_norm": 8.83347225189209,
+      "learning_rate": 6.144638403990025e-07,
+      "loss": 0.3197,
+      "step": 77770
+    },
+    {
+      "epoch": 19.39650872817955,
+      "grad_norm": 11.635566711425781,
+      "learning_rate": 6.119700748129676e-07,
+      "loss": 0.3204,
+      "step": 77780
+    },
+    {
+      "epoch": 19.399002493765586,
+      "grad_norm": 8.75010871887207,
+      "learning_rate": 6.094763092269327e-07,
+      "loss": 0.3247,
+      "step": 77790
+    },
+    {
+      "epoch": 19.40149625935162,
+      "grad_norm": 12.465825080871582,
+      "learning_rate": 6.069825436408978e-07,
+      "loss": 0.3004,
+      "step": 77800
+    },
+    {
+      "epoch": 19.403990024937656,
+      "grad_norm": 10.757222175598145,
+      "learning_rate": 6.044887780548629e-07,
+      "loss": 0.3444,
+      "step": 77810
+    },
+    {
+      "epoch": 19.40648379052369,
+      "grad_norm": 4.095909118652344,
+      "learning_rate": 6.01995012468828e-07,
+      "loss": 0.3217,
+      "step": 77820
+    },
+    {
+      "epoch": 19.408977556109726,
+      "grad_norm": 9.890026092529297,
+      "learning_rate": 5.995012468827931e-07,
+      "loss": 0.3058,
+      "step": 77830
+    },
+    {
+      "epoch": 19.41147132169576,
+      "grad_norm": 9.338248252868652,
+      "learning_rate": 5.970074812967582e-07,
+      "loss": 0.2974,
+      "step": 77840
+    },
+    {
+      "epoch": 19.413965087281795,
+      "grad_norm": 7.386726379394531,
+      "learning_rate": 5.945137157107232e-07,
+      "loss": 0.4195,
+      "step": 77850
+    },
+    {
+      "epoch": 19.41645885286783,
+      "grad_norm": 10.347369194030762,
+      "learning_rate": 5.920199501246883e-07,
+      "loss": 0.3167,
+      "step": 77860
+    },
+    {
+      "epoch": 19.418952618453865,
+      "grad_norm": 5.733013153076172,
+      "learning_rate": 5.895261845386535e-07,
+      "loss": 0.2585,
+      "step": 77870
+    },
+    {
+      "epoch": 19.4214463840399,
+      "grad_norm": 8.385955810546875,
+      "learning_rate": 5.870324189526185e-07,
+      "loss": 0.2847,
+      "step": 77880
+    },
+    {
+      "epoch": 19.423940149625935,
+      "grad_norm": 7.1023736000061035,
+      "learning_rate": 5.845386533665836e-07,
+      "loss": 0.336,
+      "step": 77890
+    },
+    {
+      "epoch": 19.42643391521197,
+      "grad_norm": 9.88353157043457,
+      "learning_rate": 5.820448877805487e-07,
+      "loss": 0.3549,
+      "step": 77900
+    },
+    {
+      "epoch": 19.428927680798004,
+      "grad_norm": 8.863442420959473,
+      "learning_rate": 5.795511221945138e-07,
+      "loss": 0.3858,
+      "step": 77910
+    },
+    {
+      "epoch": 19.43142144638404,
+      "grad_norm": 8.624554634094238,
+      "learning_rate": 5.770573566084789e-07,
+      "loss": 0.3137,
+      "step": 77920
+    },
+    {
+      "epoch": 19.433915211970074,
+      "grad_norm": 9.726763725280762,
+      "learning_rate": 5.74563591022444e-07,
+      "loss": 0.345,
+      "step": 77930
+    },
+    {
+      "epoch": 19.43640897755611,
+      "grad_norm": 7.916418075561523,
+      "learning_rate": 5.720698254364091e-07,
+      "loss": 0.2833,
+      "step": 77940
+    },
+    {
+      "epoch": 19.438902743142144,
+      "grad_norm": 10.546299934387207,
+      "learning_rate": 5.695760598503742e-07,
+      "loss": 0.2614,
+      "step": 77950
+    },
+    {
+      "epoch": 19.44139650872818,
+      "grad_norm": 9.657865524291992,
+      "learning_rate": 5.670822942643392e-07,
+      "loss": 0.3221,
+      "step": 77960
+    },
+    {
+      "epoch": 19.443890274314214,
+      "grad_norm": 7.228270530700684,
+      "learning_rate": 5.645885286783043e-07,
+      "loss": 0.285,
+      "step": 77970
+    },
+    {
+      "epoch": 19.44638403990025,
+      "grad_norm": 12.523133277893066,
+      "learning_rate": 5.620947630922694e-07,
+      "loss": 0.2654,
+      "step": 77980
+    },
+    {
+      "epoch": 19.448877805486283,
+      "grad_norm": 9.408562660217285,
+      "learning_rate": 5.596009975062345e-07,
+      "loss": 0.2996,
+      "step": 77990
+    },
+    {
+      "epoch": 19.45137157107232,
+      "grad_norm": 8.884123802185059,
+      "learning_rate": 5.571072319201996e-07,
+      "loss": 0.3365,
+      "step": 78000
+    },
+    {
+      "epoch": 19.453865336658353,
+      "grad_norm": 7.461065769195557,
+      "learning_rate": 5.546134663341646e-07,
+      "loss": 0.3295,
+      "step": 78010
+    },
+    {
+      "epoch": 19.456359102244388,
+      "grad_norm": 7.785035610198975,
+      "learning_rate": 5.521197007481297e-07,
+      "loss": 0.235,
+      "step": 78020
+    },
+    {
+      "epoch": 19.458852867830423,
+      "grad_norm": 8.66875171661377,
+      "learning_rate": 5.496259351620948e-07,
+      "loss": 0.2687,
+      "step": 78030
+    },
+    {
+      "epoch": 19.461346633416458,
+      "grad_norm": 6.394353866577148,
+      "learning_rate": 5.471321695760598e-07,
+      "loss": 0.2604,
+      "step": 78040
+    },
+    {
+      "epoch": 19.463840399002493,
+      "grad_norm": 9.329657554626465,
+      "learning_rate": 5.44638403990025e-07,
+      "loss": 0.3127,
+      "step": 78050
+    },
+    {
+      "epoch": 19.466334164588527,
+      "grad_norm": 9.868875503540039,
+      "learning_rate": 5.421446384039901e-07,
+      "loss": 0.2944,
+      "step": 78060
+    },
+    {
+      "epoch": 19.468827930174562,
+      "grad_norm": 8.829019546508789,
+      "learning_rate": 5.396508728179551e-07,
+      "loss": 0.4614,
+      "step": 78070
+    },
+    {
+      "epoch": 19.471321695760597,
+      "grad_norm": 5.467095851898193,
+      "learning_rate": 5.371571072319202e-07,
+      "loss": 0.3467,
+      "step": 78080
+    },
+    {
+      "epoch": 19.473815461346632,
+      "grad_norm": 11.852470397949219,
+      "learning_rate": 5.346633416458853e-07,
+      "loss": 0.3606,
+      "step": 78090
+    },
+    {
+      "epoch": 19.476309226932667,
+      "grad_norm": 10.978060722351074,
+      "learning_rate": 5.321695760598504e-07,
+      "loss": 0.4001,
+      "step": 78100
+    },
+    {
+      "epoch": 19.478802992518702,
+      "grad_norm": 4.6936163902282715,
+      "learning_rate": 5.296758104738155e-07,
+      "loss": 0.2956,
+      "step": 78110
+    },
+    {
+      "epoch": 19.481296758104737,
+      "grad_norm": 8.054365158081055,
+      "learning_rate": 5.271820448877806e-07,
+      "loss": 0.3031,
+      "step": 78120
+    },
+    {
+      "epoch": 19.48379052369077,
+      "grad_norm": 7.315567493438721,
+      "learning_rate": 5.246882793017457e-07,
+      "loss": 0.292,
+      "step": 78130
+    },
+    {
+      "epoch": 19.486284289276806,
+      "grad_norm": 6.582238674163818,
+      "learning_rate": 5.221945137157108e-07,
+      "loss": 0.3452,
+      "step": 78140
+    },
+    {
+      "epoch": 19.48877805486284,
+      "grad_norm": 9.56821346282959,
+      "learning_rate": 5.197007481296758e-07,
+      "loss": 0.371,
+      "step": 78150
+    },
+    {
+      "epoch": 19.491271820448876,
+      "grad_norm": 10.5318603515625,
+      "learning_rate": 5.172069825436409e-07,
+      "loss": 0.3602,
+      "step": 78160
+    },
+    {
+      "epoch": 19.49376558603491,
+      "grad_norm": 7.174575328826904,
+      "learning_rate": 5.14713216957606e-07,
+      "loss": 0.2854,
+      "step": 78170
+    },
+    {
+      "epoch": 19.496259351620946,
+      "grad_norm": 9.72573184967041,
+      "learning_rate": 5.122194513715711e-07,
+      "loss": 0.2975,
+      "step": 78180
+    },
+    {
+      "epoch": 19.49875311720698,
+      "grad_norm": 9.575770378112793,
+      "learning_rate": 5.097256857855362e-07,
+      "loss": 0.303,
+      "step": 78190
+    },
+    {
+      "epoch": 19.50124688279302,
+      "grad_norm": 9.664167404174805,
+      "learning_rate": 5.072319201995013e-07,
+      "loss": 0.3959,
+      "step": 78200
+    },
+    {
+      "epoch": 19.503740648379054,
+      "grad_norm": 9.702856063842773,
+      "learning_rate": 5.047381546134664e-07,
+      "loss": 0.3454,
+      "step": 78210
+    },
+    {
+      "epoch": 19.50623441396509,
+      "grad_norm": 13.055885314941406,
+      "learning_rate": 5.022443890274315e-07,
+      "loss": 0.2945,
+      "step": 78220
+    },
+    {
+      "epoch": 19.508728179551124,
+      "grad_norm": 7.412993907928467,
+      "learning_rate": 4.997506234413965e-07,
+      "loss": 0.3167,
+      "step": 78230
+    },
+    {
+      "epoch": 19.51122194513716,
+      "grad_norm": 5.914247512817383,
+      "learning_rate": 4.972568578553617e-07,
+      "loss": 0.3067,
+      "step": 78240
+    },
+    {
+      "epoch": 19.513715710723194,
+      "grad_norm": 6.335524559020996,
+      "learning_rate": 4.947630922693268e-07,
+      "loss": 0.3002,
+      "step": 78250
+    },
+    {
+      "epoch": 19.51620947630923,
+      "grad_norm": 9.340645790100098,
+      "learning_rate": 4.922693266832918e-07,
+      "loss": 0.3535,
+      "step": 78260
+    },
+    {
+      "epoch": 19.518703241895263,
+      "grad_norm": 8.127272605895996,
+      "learning_rate": 4.897755610972569e-07,
+      "loss": 0.3276,
+      "step": 78270
+    },
+    {
+      "epoch": 19.521197007481298,
+      "grad_norm": 8.352450370788574,
+      "learning_rate": 4.87281795511222e-07,
+      "loss": 0.3708,
+      "step": 78280
+    },
+    {
+      "epoch": 19.523690773067333,
+      "grad_norm": 5.065845489501953,
+      "learning_rate": 4.847880299251871e-07,
+      "loss": 0.266,
+      "step": 78290
+    },
+    {
+      "epoch": 19.526184538653368,
+      "grad_norm": 9.564516067504883,
+      "learning_rate": 4.822942643391522e-07,
+      "loss": 0.2801,
+      "step": 78300
+    },
+    {
+      "epoch": 19.528678304239403,
+      "grad_norm": 8.529273986816406,
+      "learning_rate": 4.798004987531173e-07,
+      "loss": 0.3554,
+      "step": 78310
+    },
+    {
+      "epoch": 19.531172069825438,
+      "grad_norm": 9.940313339233398,
+      "learning_rate": 4.773067331670824e-07,
+      "loss": 0.2956,
+      "step": 78320
+    },
+    {
+      "epoch": 19.533665835411473,
+      "grad_norm": 7.539468288421631,
+      "learning_rate": 4.748129675810474e-07,
+      "loss": 0.3047,
+      "step": 78330
+    },
+    {
+      "epoch": 19.536159600997507,
+      "grad_norm": 9.497523307800293,
+      "learning_rate": 4.723192019950125e-07,
+      "loss": 0.2872,
+      "step": 78340
+    },
+    {
+      "epoch": 19.538653366583542,
+      "grad_norm": 8.197120666503906,
+      "learning_rate": 4.6982543640897756e-07,
+      "loss": 0.3116,
+      "step": 78350
+    },
+    {
+      "epoch": 19.541147132169577,
+      "grad_norm": 8.740259170532227,
+      "learning_rate": 4.673316708229427e-07,
+      "loss": 0.3306,
+      "step": 78360
+    },
+    {
+      "epoch": 19.543640897755612,
+      "grad_norm": 6.9125494956970215,
+      "learning_rate": 4.648379052369078e-07,
+      "loss": 0.3007,
+      "step": 78370
+    },
+    {
+      "epoch": 19.546134663341647,
+      "grad_norm": 7.752559185028076,
+      "learning_rate": 4.6234413965087285e-07,
+      "loss": 0.3,
+      "step": 78380
+    },
+    {
+      "epoch": 19.54862842892768,
+      "grad_norm": 10.674647331237793,
+      "learning_rate": 4.5985037406483795e-07,
+      "loss": 0.3157,
+      "step": 78390
+    },
+    {
+      "epoch": 19.551122194513717,
+      "grad_norm": 9.210230827331543,
+      "learning_rate": 4.57356608478803e-07,
+      "loss": 0.2928,
+      "step": 78400
+    },
+    {
+      "epoch": 19.55361596009975,
+      "grad_norm": 8.934961318969727,
+      "learning_rate": 4.548628428927681e-07,
+      "loss": 0.3272,
+      "step": 78410
+    },
+    {
+      "epoch": 19.556109725685786,
+      "grad_norm": 9.600099563598633,
+      "learning_rate": 4.523690773067332e-07,
+      "loss": 0.3249,
+      "step": 78420
+    },
+    {
+      "epoch": 19.55860349127182,
+      "grad_norm": 9.030034065246582,
+      "learning_rate": 4.4987531172069834e-07,
+      "loss": 0.283,
+      "step": 78430
+    },
+    {
+      "epoch": 19.561097256857856,
+      "grad_norm": 8.656168937683105,
+      "learning_rate": 4.473815461346634e-07,
+      "loss": 0.2666,
+      "step": 78440
+    },
+    {
+      "epoch": 19.56359102244389,
+      "grad_norm": 8.763611793518066,
+      "learning_rate": 4.448877805486285e-07,
+      "loss": 0.3127,
+      "step": 78450
+    },
+    {
+      "epoch": 19.566084788029926,
+      "grad_norm": 6.0083136558532715,
+      "learning_rate": 4.4239401496259353e-07,
+      "loss": 0.2952,
+      "step": 78460
+    },
+    {
+      "epoch": 19.56857855361596,
+      "grad_norm": 8.106850624084473,
+      "learning_rate": 4.399002493765586e-07,
+      "loss": 0.2619,
+      "step": 78470
+    },
+    {
+      "epoch": 19.571072319201996,
+      "grad_norm": 5.120926380157471,
+      "learning_rate": 4.3740648379052367e-07,
+      "loss": 0.3112,
+      "step": 78480
+    },
+    {
+      "epoch": 19.57356608478803,
+      "grad_norm": 8.02582836151123,
+      "learning_rate": 4.349127182044888e-07,
+      "loss": 0.3119,
+      "step": 78490
+    },
+    {
+      "epoch": 19.576059850374065,
+      "grad_norm": 14.72515869140625,
+      "learning_rate": 4.324189526184539e-07,
+      "loss": 0.3285,
+      "step": 78500
+    },
+    {
+      "epoch": 19.5785536159601,
+      "grad_norm": 10.0016508102417,
+      "learning_rate": 4.29925187032419e-07,
+      "loss": 0.3348,
+      "step": 78510
+    },
+    {
+      "epoch": 19.581047381546135,
+      "grad_norm": 8.692368507385254,
+      "learning_rate": 4.2743142144638406e-07,
+      "loss": 0.3112,
+      "step": 78520
+    },
+    {
+      "epoch": 19.58354114713217,
+      "grad_norm": 10.636957168579102,
+      "learning_rate": 4.2493765586034916e-07,
+      "loss": 0.3008,
+      "step": 78530
+    },
+    {
+      "epoch": 19.586034912718205,
+      "grad_norm": 14.530603408813477,
+      "learning_rate": 4.224438902743142e-07,
+      "loss": 0.3169,
+      "step": 78540
+    },
+    {
+      "epoch": 19.58852867830424,
+      "grad_norm": 10.313661575317383,
+      "learning_rate": 4.1995012468827936e-07,
+      "loss": 0.4105,
+      "step": 78550
+    },
+    {
+      "epoch": 19.591022443890274,
+      "grad_norm": 8.854100227355957,
+      "learning_rate": 4.1745635910224446e-07,
+      "loss": 0.3473,
+      "step": 78560
+    },
+    {
+      "epoch": 19.59351620947631,
+      "grad_norm": 5.632763385772705,
+      "learning_rate": 4.149625935162095e-07,
+      "loss": 0.3534,
+      "step": 78570
+    },
+    {
+      "epoch": 19.596009975062344,
+      "grad_norm": 10.139565467834473,
+      "learning_rate": 4.124688279301746e-07,
+      "loss": 0.2864,
+      "step": 78580
+    },
+    {
+      "epoch": 19.59850374064838,
+      "grad_norm": 10.44018268585205,
+      "learning_rate": 4.0997506234413964e-07,
+      "loss": 0.3235,
+      "step": 78590
+    },
+    {
+      "epoch": 19.600997506234414,
+      "grad_norm": 8.089249610900879,
+      "learning_rate": 4.0748129675810474e-07,
+      "loss": 0.3483,
+      "step": 78600
+    },
+    {
+      "epoch": 19.60349127182045,
+      "grad_norm": 8.919448852539062,
+      "learning_rate": 4.049875311720699e-07,
+      "loss": 0.2935,
+      "step": 78610
+    },
+    {
+      "epoch": 19.605985037406484,
+      "grad_norm": 12.397672653198242,
+      "learning_rate": 4.02493765586035e-07,
+      "loss": 0.2637,
+      "step": 78620
+    },
+    {
+      "epoch": 19.60847880299252,
+      "grad_norm": 7.087367057800293,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.3281,
+      "step": 78630
+    },
+    {
+      "epoch": 19.610972568578553,
+      "grad_norm": 10.792654037475586,
+      "learning_rate": 3.9750623441396513e-07,
+      "loss": 0.2798,
+      "step": 78640
+    },
+    {
+      "epoch": 19.61346633416459,
+      "grad_norm": 11.265054702758789,
+      "learning_rate": 3.950124688279302e-07,
+      "loss": 0.3599,
+      "step": 78650
+    },
+    {
+      "epoch": 19.615960099750623,
+      "grad_norm": 7.813275337219238,
+      "learning_rate": 3.925187032418953e-07,
+      "loss": 0.3727,
+      "step": 78660
+    },
+    {
+      "epoch": 19.618453865336658,
+      "grad_norm": 7.351382255554199,
+      "learning_rate": 3.9002493765586043e-07,
+      "loss": 0.337,
+      "step": 78670
+    },
+    {
+      "epoch": 19.620947630922693,
+      "grad_norm": 8.567827224731445,
+      "learning_rate": 3.8753117206982547e-07,
+      "loss": 0.3548,
+      "step": 78680
+    },
+    {
+      "epoch": 19.623441396508728,
+      "grad_norm": 8.134788513183594,
+      "learning_rate": 3.8503740648379057e-07,
+      "loss": 0.3234,
+      "step": 78690
+    },
+    {
+      "epoch": 19.625935162094763,
+      "grad_norm": 8.092358589172363,
+      "learning_rate": 3.8254364089775567e-07,
+      "loss": 0.4081,
+      "step": 78700
+    },
+    {
+      "epoch": 19.628428927680797,
+      "grad_norm": 10.603907585144043,
+      "learning_rate": 3.800498753117207e-07,
+      "loss": 0.3384,
+      "step": 78710
+    },
+    {
+      "epoch": 19.630922693266832,
+      "grad_norm": 7.839836120605469,
+      "learning_rate": 3.775561097256858e-07,
+      "loss": 0.3905,
+      "step": 78720
+    },
+    {
+      "epoch": 19.633416458852867,
+      "grad_norm": 10.831620216369629,
+      "learning_rate": 3.7506234413965086e-07,
+      "loss": 0.2838,
+      "step": 78730
+    },
+    {
+      "epoch": 19.635910224438902,
+      "grad_norm": 7.171167850494385,
+      "learning_rate": 3.72568578553616e-07,
+      "loss": 0.361,
+      "step": 78740
+    },
+    {
+      "epoch": 19.638403990024937,
+      "grad_norm": 8.736506462097168,
+      "learning_rate": 3.700748129675811e-07,
+      "loss": 0.3024,
+      "step": 78750
+    },
+    {
+      "epoch": 19.640897755610972,
+      "grad_norm": 8.326260566711426,
+      "learning_rate": 3.6758104738154615e-07,
+      "loss": 0.3674,
+      "step": 78760
+    },
+    {
+      "epoch": 19.643391521197007,
+      "grad_norm": 7.458507061004639,
+      "learning_rate": 3.6508728179551125e-07,
+      "loss": 0.3018,
+      "step": 78770
+    },
+    {
+      "epoch": 19.64588528678304,
+      "grad_norm": 10.653674125671387,
+      "learning_rate": 3.625935162094763e-07,
+      "loss": 0.3053,
+      "step": 78780
+    },
+    {
+      "epoch": 19.648379052369076,
+      "grad_norm": 7.049093723297119,
+      "learning_rate": 3.600997506234414e-07,
+      "loss": 0.3749,
+      "step": 78790
+    },
+    {
+      "epoch": 19.65087281795511,
+      "grad_norm": 8.122016906738281,
+      "learning_rate": 3.5760598503740654e-07,
+      "loss": 0.3234,
+      "step": 78800
+    },
+    {
+      "epoch": 19.653366583541146,
+      "grad_norm": 11.616135597229004,
+      "learning_rate": 3.5511221945137164e-07,
+      "loss": 0.2943,
+      "step": 78810
+    },
+    {
+      "epoch": 19.65586034912718,
+      "grad_norm": 12.034366607666016,
+      "learning_rate": 3.526184538653367e-07,
+      "loss": 0.322,
+      "step": 78820
+    },
+    {
+      "epoch": 19.658354114713216,
+      "grad_norm": 11.825723648071289,
+      "learning_rate": 3.501246882793018e-07,
+      "loss": 0.3468,
+      "step": 78830
+    },
+    {
+      "epoch": 19.66084788029925,
+      "grad_norm": 8.839386940002441,
+      "learning_rate": 3.4763092269326683e-07,
+      "loss": 0.3245,
+      "step": 78840
+    },
+    {
+      "epoch": 19.663341645885286,
+      "grad_norm": 5.619390487670898,
+      "learning_rate": 3.451371571072319e-07,
+      "loss": 0.2613,
+      "step": 78850
+    },
+    {
+      "epoch": 19.66583541147132,
+      "grad_norm": 7.329450607299805,
+      "learning_rate": 3.426433915211971e-07,
+      "loss": 0.3291,
+      "step": 78860
+    },
+    {
+      "epoch": 19.668329177057355,
+      "grad_norm": 8.157301902770996,
+      "learning_rate": 3.401496259351621e-07,
+      "loss": 0.3298,
+      "step": 78870
+    },
+    {
+      "epoch": 19.67082294264339,
+      "grad_norm": 18.877344131469727,
+      "learning_rate": 3.376558603491272e-07,
+      "loss": 0.279,
+      "step": 78880
+    },
+    {
+      "epoch": 19.673316708229425,
+      "grad_norm": 9.491035461425781,
+      "learning_rate": 3.351620947630923e-07,
+      "loss": 0.2876,
+      "step": 78890
+    },
+    {
+      "epoch": 19.67581047381546,
+      "grad_norm": 9.576622009277344,
+      "learning_rate": 3.3266832917705736e-07,
+      "loss": 0.2932,
+      "step": 78900
+    },
+    {
+      "epoch": 19.678304239401495,
+      "grad_norm": 8.893289566040039,
+      "learning_rate": 3.3017456359102246e-07,
+      "loss": 0.2582,
+      "step": 78910
+    },
+    {
+      "epoch": 19.68079800498753,
+      "grad_norm": 12.161534309387207,
+      "learning_rate": 3.276807980049876e-07,
+      "loss": 0.3097,
+      "step": 78920
+    },
+    {
+      "epoch": 19.683291770573565,
+      "grad_norm": 9.909988403320312,
+      "learning_rate": 3.2518703241895266e-07,
+      "loss": 0.326,
+      "step": 78930
+    },
+    {
+      "epoch": 19.6857855361596,
+      "grad_norm": 7.692135810852051,
+      "learning_rate": 3.2269326683291775e-07,
+      "loss": 0.3324,
+      "step": 78940
+    },
+    {
+      "epoch": 19.688279301745634,
+      "grad_norm": 9.485343933105469,
+      "learning_rate": 3.201995012468828e-07,
+      "loss": 0.2779,
+      "step": 78950
+    },
+    {
+      "epoch": 19.69077306733167,
+      "grad_norm": 9.124640464782715,
+      "learning_rate": 3.177057356608479e-07,
+      "loss": 0.3118,
+      "step": 78960
+    },
+    {
+      "epoch": 19.693266832917704,
+      "grad_norm": 8.47995376586914,
+      "learning_rate": 3.1521197007481294e-07,
+      "loss": 0.2873,
+      "step": 78970
+    },
+    {
+      "epoch": 19.69576059850374,
+      "grad_norm": 9.394947052001953,
+      "learning_rate": 3.127182044887781e-07,
+      "loss": 0.317,
+      "step": 78980
+    },
+    {
+      "epoch": 19.698254364089777,
+      "grad_norm": 9.908628463745117,
+      "learning_rate": 3.1022443890274314e-07,
+      "loss": 0.3296,
+      "step": 78990
+    },
+    {
+      "epoch": 19.70074812967581,
+      "grad_norm": 8.319961547851562,
+      "learning_rate": 3.077306733167083e-07,
+      "loss": 0.3183,
+      "step": 79000
+    },
+    {
+      "epoch": 19.703241895261847,
+      "grad_norm": 9.598522186279297,
+      "learning_rate": 3.0523690773067333e-07,
+      "loss": 0.3059,
+      "step": 79010
+    },
+    {
+      "epoch": 19.705735660847882,
+      "grad_norm": 7.940982341766357,
+      "learning_rate": 3.0274314214463843e-07,
+      "loss": 0.385,
+      "step": 79020
+    },
+    {
+      "epoch": 19.708229426433917,
+      "grad_norm": 9.714994430541992,
+      "learning_rate": 3.0024937655860353e-07,
+      "loss": 0.3216,
+      "step": 79030
+    },
+    {
+      "epoch": 19.71072319201995,
+      "grad_norm": 11.18153190612793,
+      "learning_rate": 2.9775561097256863e-07,
+      "loss": 0.3392,
+      "step": 79040
+    },
+    {
+      "epoch": 19.713216957605987,
+      "grad_norm": 8.196310043334961,
+      "learning_rate": 2.9526184538653367e-07,
+      "loss": 0.277,
+      "step": 79050
+    },
+    {
+      "epoch": 19.71571072319202,
+      "grad_norm": 12.826026916503906,
+      "learning_rate": 2.9276807980049877e-07,
+      "loss": 0.3442,
+      "step": 79060
+    },
+    {
+      "epoch": 19.718204488778056,
+      "grad_norm": 14.502406120300293,
+      "learning_rate": 2.9027431421446387e-07,
+      "loss": 0.3195,
+      "step": 79070
+    },
+    {
+      "epoch": 19.72069825436409,
+      "grad_norm": 10.214339256286621,
+      "learning_rate": 2.8778054862842897e-07,
+      "loss": 0.2983,
+      "step": 79080
+    },
+    {
+      "epoch": 19.723192019950126,
+      "grad_norm": 9.869833946228027,
+      "learning_rate": 2.8528678304239406e-07,
+      "loss": 0.3399,
+      "step": 79090
+    },
+    {
+      "epoch": 19.72568578553616,
+      "grad_norm": 10.579812049865723,
+      "learning_rate": 2.827930174563591e-07,
+      "loss": 0.2615,
+      "step": 79100
+    },
+    {
+      "epoch": 19.728179551122196,
+      "grad_norm": 9.650101661682129,
+      "learning_rate": 2.802992518703242e-07,
+      "loss": 0.2366,
+      "step": 79110
+    },
+    {
+      "epoch": 19.73067331670823,
+      "grad_norm": 7.501360893249512,
+      "learning_rate": 2.778054862842893e-07,
+      "loss": 0.2624,
+      "step": 79120
+    },
+    {
+      "epoch": 19.733167082294266,
+      "grad_norm": 7.877026081085205,
+      "learning_rate": 2.753117206982544e-07,
+      "loss": 0.302,
+      "step": 79130
+    },
+    {
+      "epoch": 19.7356608478803,
+      "grad_norm": 9.600006103515625,
+      "learning_rate": 2.7281795511221945e-07,
+      "loss": 0.3465,
+      "step": 79140
+    },
+    {
+      "epoch": 19.738154613466335,
+      "grad_norm": 8.802675247192383,
+      "learning_rate": 2.703241895261846e-07,
+      "loss": 0.2923,
+      "step": 79150
+    },
+    {
+      "epoch": 19.74064837905237,
+      "grad_norm": 16.73734474182129,
+      "learning_rate": 2.6783042394014964e-07,
+      "loss": 0.3215,
+      "step": 79160
+    },
+    {
+      "epoch": 19.743142144638405,
+      "grad_norm": 13.215644836425781,
+      "learning_rate": 2.6533665835411474e-07,
+      "loss": 0.3195,
+      "step": 79170
+    },
+    {
+      "epoch": 19.74563591022444,
+      "grad_norm": 6.847998142242432,
+      "learning_rate": 2.6284289276807984e-07,
+      "loss": 0.3243,
+      "step": 79180
+    },
+    {
+      "epoch": 19.748129675810475,
+      "grad_norm": 8.965968132019043,
+      "learning_rate": 2.6034912718204494e-07,
+      "loss": 0.3299,
+      "step": 79190
+    },
+    {
+      "epoch": 19.75062344139651,
+      "grad_norm": 8.921945571899414,
+      "learning_rate": 2.5785536159601e-07,
+      "loss": 0.3414,
+      "step": 79200
+    },
+    {
+      "epoch": 19.753117206982544,
+      "grad_norm": 8.51665210723877,
+      "learning_rate": 2.553615960099751e-07,
+      "loss": 0.3682,
+      "step": 79210
+    },
+    {
+      "epoch": 19.75561097256858,
+      "grad_norm": 9.166069984436035,
+      "learning_rate": 2.528678304239402e-07,
+      "loss": 0.3229,
+      "step": 79220
+    },
+    {
+      "epoch": 19.758104738154614,
+      "grad_norm": 16.574934005737305,
+      "learning_rate": 2.503740648379053e-07,
+      "loss": 0.3596,
+      "step": 79230
+    },
+    {
+      "epoch": 19.76059850374065,
+      "grad_norm": 8.440783500671387,
+      "learning_rate": 2.478802992518703e-07,
+      "loss": 0.259,
+      "step": 79240
+    },
+    {
+      "epoch": 19.763092269326684,
+      "grad_norm": 10.631123542785645,
+      "learning_rate": 2.453865336658354e-07,
+      "loss": 0.3265,
+      "step": 79250
+    },
+    {
+      "epoch": 19.76558603491272,
+      "grad_norm": 10.455692291259766,
+      "learning_rate": 2.428927680798005e-07,
+      "loss": 0.3613,
+      "step": 79260
+    },
+    {
+      "epoch": 19.768079800498754,
+      "grad_norm": 9.9364013671875,
+      "learning_rate": 2.403990024937656e-07,
+      "loss": 0.3156,
+      "step": 79270
+    },
+    {
+      "epoch": 19.77057356608479,
+      "grad_norm": 8.415620803833008,
+      "learning_rate": 2.3790523690773071e-07,
+      "loss": 0.3035,
+      "step": 79280
+    },
+    {
+      "epoch": 19.773067331670823,
+      "grad_norm": 9.595970153808594,
+      "learning_rate": 2.3541147132169579e-07,
+      "loss": 0.2897,
+      "step": 79290
+    },
+    {
+      "epoch": 19.77556109725686,
+      "grad_norm": 8.665029525756836,
+      "learning_rate": 2.3291770573566086e-07,
+      "loss": 0.4021,
+      "step": 79300
+    },
+    {
+      "epoch": 19.778054862842893,
+      "grad_norm": 8.567267417907715,
+      "learning_rate": 2.3042394014962595e-07,
+      "loss": 0.3056,
+      "step": 79310
+    },
+    {
+      "epoch": 19.780548628428928,
+      "grad_norm": 12.760359764099121,
+      "learning_rate": 2.2793017456359103e-07,
+      "loss": 0.3164,
+      "step": 79320
+    },
+    {
+      "epoch": 19.783042394014963,
+      "grad_norm": 11.60568904876709,
+      "learning_rate": 2.2543640897755612e-07,
+      "loss": 0.3371,
+      "step": 79330
+    },
+    {
+      "epoch": 19.785536159600998,
+      "grad_norm": 8.413047790527344,
+      "learning_rate": 2.2294264339152122e-07,
+      "loss": 0.312,
+      "step": 79340
+    },
+    {
+      "epoch": 19.788029925187033,
+      "grad_norm": 9.830360412597656,
+      "learning_rate": 2.204488778054863e-07,
+      "loss": 0.3139,
+      "step": 79350
+    },
+    {
+      "epoch": 19.790523690773068,
+      "grad_norm": 9.22851848602295,
+      "learning_rate": 2.1795511221945137e-07,
+      "loss": 0.3348,
+      "step": 79360
+    },
+    {
+      "epoch": 19.793017456359102,
+      "grad_norm": 9.71314811706543,
+      "learning_rate": 2.154613466334165e-07,
+      "loss": 0.4009,
+      "step": 79370
+    },
+    {
+      "epoch": 19.795511221945137,
+      "grad_norm": 12.098420143127441,
+      "learning_rate": 2.1296758104738156e-07,
+      "loss": 0.2935,
+      "step": 79380
+    },
+    {
+      "epoch": 19.798004987531172,
+      "grad_norm": 11.904885292053223,
+      "learning_rate": 2.1047381546134663e-07,
+      "loss": 0.3176,
+      "step": 79390
+    },
+    {
+      "epoch": 19.800498753117207,
+      "grad_norm": 10.853218078613281,
+      "learning_rate": 2.0798004987531176e-07,
+      "loss": 0.3419,
+      "step": 79400
+    },
+    {
+      "epoch": 19.802992518703242,
+      "grad_norm": 10.429359436035156,
+      "learning_rate": 2.0548628428927683e-07,
+      "loss": 0.28,
+      "step": 79410
+    },
+    {
+      "epoch": 19.805486284289277,
+      "grad_norm": 6.893282413482666,
+      "learning_rate": 2.029925187032419e-07,
+      "loss": 0.2803,
+      "step": 79420
+    },
+    {
+      "epoch": 19.80798004987531,
+      "grad_norm": 17.683412551879883,
+      "learning_rate": 2.0049875311720702e-07,
+      "loss": 0.3656,
+      "step": 79430
+    },
+    {
+      "epoch": 19.810473815461346,
+      "grad_norm": 6.478290557861328,
+      "learning_rate": 1.980049875311721e-07,
+      "loss": 0.3119,
+      "step": 79440
+    },
+    {
+      "epoch": 19.81296758104738,
+      "grad_norm": 11.704423904418945,
+      "learning_rate": 1.9551122194513717e-07,
+      "loss": 0.4239,
+      "step": 79450
+    },
+    {
+      "epoch": 19.815461346633416,
+      "grad_norm": 10.740715980529785,
+      "learning_rate": 1.9301745635910227e-07,
+      "loss": 0.3632,
+      "step": 79460
+    },
+    {
+      "epoch": 19.81795511221945,
+      "grad_norm": 7.998754501342773,
+      "learning_rate": 1.9052369077306736e-07,
+      "loss": 0.3205,
+      "step": 79470
+    },
+    {
+      "epoch": 19.820448877805486,
+      "grad_norm": 12.84117317199707,
+      "learning_rate": 1.8802992518703243e-07,
+      "loss": 0.322,
+      "step": 79480
+    },
+    {
+      "epoch": 19.82294264339152,
+      "grad_norm": 11.327634811401367,
+      "learning_rate": 1.8553615960099753e-07,
+      "loss": 0.3062,
+      "step": 79490
+    },
+    {
+      "epoch": 19.825436408977556,
+      "grad_norm": 16.317493438720703,
+      "learning_rate": 1.830423940149626e-07,
+      "loss": 0.3968,
+      "step": 79500
+    },
+    {
+      "epoch": 19.82793017456359,
+      "grad_norm": 9.288350105285645,
+      "learning_rate": 1.8054862842892768e-07,
+      "loss": 0.2814,
+      "step": 79510
+    },
+    {
+      "epoch": 19.830423940149625,
+      "grad_norm": 6.440745830535889,
+      "learning_rate": 1.780548628428928e-07,
+      "loss": 0.2889,
+      "step": 79520
+    },
+    {
+      "epoch": 19.83291770573566,
+      "grad_norm": 7.100279331207275,
+      "learning_rate": 1.7556109725685787e-07,
+      "loss": 0.2817,
+      "step": 79530
+    },
+    {
+      "epoch": 19.835411471321695,
+      "grad_norm": 8.271157264709473,
+      "learning_rate": 1.7306733167082294e-07,
+      "loss": 0.2249,
+      "step": 79540
+    },
+    {
+      "epoch": 19.83790523690773,
+      "grad_norm": 7.640126705169678,
+      "learning_rate": 1.7057356608478801e-07,
+      "loss": 0.3156,
+      "step": 79550
+    },
+    {
+      "epoch": 19.840399002493765,
+      "grad_norm": 11.971923828125,
+      "learning_rate": 1.6807980049875314e-07,
+      "loss": 0.3263,
+      "step": 79560
+    },
+    {
+      "epoch": 19.8428927680798,
+      "grad_norm": 6.482527256011963,
+      "learning_rate": 1.655860349127182e-07,
+      "loss": 0.2904,
+      "step": 79570
+    },
+    {
+      "epoch": 19.845386533665835,
+      "grad_norm": 7.9649200439453125,
+      "learning_rate": 1.6309226932668328e-07,
+      "loss": 0.309,
+      "step": 79580
+    },
+    {
+      "epoch": 19.84788029925187,
+      "grad_norm": 9.517223358154297,
+      "learning_rate": 1.605985037406484e-07,
+      "loss": 0.3661,
+      "step": 79590
+    },
+    {
+      "epoch": 19.850374064837904,
+      "grad_norm": 16.75897216796875,
+      "learning_rate": 1.5810473815461348e-07,
+      "loss": 0.3179,
+      "step": 79600
+    },
+    {
+      "epoch": 19.85286783042394,
+      "grad_norm": 7.259194850921631,
+      "learning_rate": 1.5561097256857858e-07,
+      "loss": 0.313,
+      "step": 79610
+    },
+    {
+      "epoch": 19.855361596009974,
+      "grad_norm": 6.734643936157227,
+      "learning_rate": 1.5311720698254367e-07,
+      "loss": 0.3129,
+      "step": 79620
+    },
+    {
+      "epoch": 19.85785536159601,
+      "grad_norm": 8.320770263671875,
+      "learning_rate": 1.5062344139650875e-07,
+      "loss": 0.2911,
+      "step": 79630
+    },
+    {
+      "epoch": 19.860349127182044,
+      "grad_norm": 9.448941230773926,
+      "learning_rate": 1.4812967581047384e-07,
+      "loss": 0.3353,
+      "step": 79640
+    },
+    {
+      "epoch": 19.86284289276808,
+      "grad_norm": 7.158919811248779,
+      "learning_rate": 1.4563591022443891e-07,
+      "loss": 0.3327,
+      "step": 79650
+    },
+    {
+      "epoch": 19.865336658354114,
+      "grad_norm": 6.863051891326904,
+      "learning_rate": 1.43142144638404e-07,
+      "loss": 0.2756,
+      "step": 79660
+    },
+    {
+      "epoch": 19.86783042394015,
+      "grad_norm": 9.206826210021973,
+      "learning_rate": 1.4064837905236908e-07,
+      "loss": 0.3184,
+      "step": 79670
+    },
+    {
+      "epoch": 19.870324189526183,
+      "grad_norm": 9.977852821350098,
+      "learning_rate": 1.3815461346633418e-07,
+      "loss": 0.3783,
+      "step": 79680
+    },
+    {
+      "epoch": 19.872817955112218,
+      "grad_norm": 8.145519256591797,
+      "learning_rate": 1.3566084788029925e-07,
+      "loss": 0.3138,
+      "step": 79690
+    },
+    {
+      "epoch": 19.875311720698253,
+      "grad_norm": 10.240920066833496,
+      "learning_rate": 1.3316708229426435e-07,
+      "loss": 0.3352,
+      "step": 79700
+    },
+    {
+      "epoch": 19.877805486284288,
+      "grad_norm": 4.372189998626709,
+      "learning_rate": 1.3067331670822942e-07,
+      "loss": 0.3623,
+      "step": 79710
+    },
+    {
+      "epoch": 19.880299251870323,
+      "grad_norm": 8.158557891845703,
+      "learning_rate": 1.2817955112219452e-07,
+      "loss": 0.3992,
+      "step": 79720
+    },
+    {
+      "epoch": 19.882793017456358,
+      "grad_norm": 9.305794715881348,
+      "learning_rate": 1.2568578553615962e-07,
+      "loss": 0.3528,
+      "step": 79730
+    },
+    {
+      "epoch": 19.885286783042392,
+      "grad_norm": 8.110395431518555,
+      "learning_rate": 1.231920199501247e-07,
+      "loss": 0.3128,
+      "step": 79740
+    },
+    {
+      "epoch": 19.887780548628427,
+      "grad_norm": 7.968203067779541,
+      "learning_rate": 1.206982543640898e-07,
+      "loss": 0.3007,
+      "step": 79750
+    },
+    {
+      "epoch": 19.890274314214462,
+      "grad_norm": 7.759973526000977,
+      "learning_rate": 1.1820448877805489e-07,
+      "loss": 0.2825,
+      "step": 79760
+    },
+    {
+      "epoch": 19.892768079800497,
+      "grad_norm": 10.173065185546875,
+      "learning_rate": 1.1571072319201996e-07,
+      "loss": 0.411,
+      "step": 79770
+    },
+    {
+      "epoch": 19.895261845386532,
+      "grad_norm": 7.4826555252075195,
+      "learning_rate": 1.1321695760598506e-07,
+      "loss": 0.2562,
+      "step": 79780
+    },
+    {
+      "epoch": 19.897755610972567,
+      "grad_norm": 9.044584274291992,
+      "learning_rate": 1.1072319201995014e-07,
+      "loss": 0.3396,
+      "step": 79790
+    },
+    {
+      "epoch": 19.900249376558605,
+      "grad_norm": 9.79257583618164,
+      "learning_rate": 1.0822942643391521e-07,
+      "loss": 0.323,
+      "step": 79800
+    },
+    {
+      "epoch": 19.902743142144637,
+      "grad_norm": 10.325493812561035,
+      "learning_rate": 1.0573566084788031e-07,
+      "loss": 0.2948,
+      "step": 79810
+    },
+    {
+      "epoch": 19.905236907730675,
+      "grad_norm": 8.43407917022705,
+      "learning_rate": 1.0324189526184538e-07,
+      "loss": 0.2562,
+      "step": 79820
+    },
+    {
+      "epoch": 19.90773067331671,
+      "grad_norm": 9.039451599121094,
+      "learning_rate": 1.0074812967581048e-07,
+      "loss": 0.319,
+      "step": 79830
+    },
+    {
+      "epoch": 19.910224438902745,
+      "grad_norm": 5.012162208557129,
+      "learning_rate": 9.825436408977558e-08,
+      "loss": 0.2856,
+      "step": 79840
+    },
+    {
+      "epoch": 19.91271820448878,
+      "grad_norm": 6.816104412078857,
+      "learning_rate": 9.576059850374065e-08,
+      "loss": 0.2679,
+      "step": 79850
+    },
+    {
+      "epoch": 19.915211970074814,
+      "grad_norm": 12.247861862182617,
+      "learning_rate": 9.326683291770575e-08,
+      "loss": 0.3458,
+      "step": 79860
+    },
+    {
+      "epoch": 19.91770573566085,
+      "grad_norm": 6.4964118003845215,
+      "learning_rate": 9.077306733167083e-08,
+      "loss": 0.3334,
+      "step": 79870
+    },
+    {
+      "epoch": 19.920199501246884,
+      "grad_norm": 9.710594177246094,
+      "learning_rate": 8.827930174563592e-08,
+      "loss": 0.3469,
+      "step": 79880
+    },
+    {
+      "epoch": 19.92269326683292,
+      "grad_norm": 12.090693473815918,
+      "learning_rate": 8.5785536159601e-08,
+      "loss": 0.3702,
+      "step": 79890
+    },
+    {
+      "epoch": 19.925187032418954,
+      "grad_norm": 8.259521484375,
+      "learning_rate": 8.32917705735661e-08,
+      "loss": 0.3142,
+      "step": 79900
+    },
+    {
+      "epoch": 19.92768079800499,
+      "grad_norm": 10.591052055358887,
+      "learning_rate": 8.079800498753117e-08,
+      "loss": 0.3415,
+      "step": 79910
+    },
+    {
+      "epoch": 19.930174563591024,
+      "grad_norm": 8.798907279968262,
+      "learning_rate": 7.830423940149627e-08,
+      "loss": 0.2823,
+      "step": 79920
+    },
+    {
+      "epoch": 19.93266832917706,
+      "grad_norm": 6.978452682495117,
+      "learning_rate": 7.581047381546135e-08,
+      "loss": 0.3098,
+      "step": 79930
+    },
+    {
+      "epoch": 19.935162094763093,
+      "grad_norm": 8.613764762878418,
+      "learning_rate": 7.331670822942644e-08,
+      "loss": 0.3456,
+      "step": 79940
+    },
+    {
+      "epoch": 19.93765586034913,
+      "grad_norm": 9.246316909790039,
+      "learning_rate": 7.082294264339154e-08,
+      "loss": 0.2926,
+      "step": 79950
+    },
+    {
+      "epoch": 19.940149625935163,
+      "grad_norm": 6.40701961517334,
+      "learning_rate": 6.832917705735662e-08,
+      "loss": 0.4304,
+      "step": 79960
+    },
+    {
+      "epoch": 19.942643391521198,
+      "grad_norm": 9.808752059936523,
+      "learning_rate": 6.58354114713217e-08,
+      "loss": 0.3531,
+      "step": 79970
+    },
+    {
+      "epoch": 19.945137157107233,
+      "grad_norm": 10.570954322814941,
+      "learning_rate": 6.334164588528679e-08,
+      "loss": 0.304,
+      "step": 79980
+    },
+    {
+      "epoch": 19.947630922693268,
+      "grad_norm": 6.992794036865234,
+      "learning_rate": 6.084788029925187e-08,
+      "loss": 0.333,
+      "step": 79990
+    },
+    {
+      "epoch": 19.950124688279303,
+      "grad_norm": 6.577144145965576,
+      "learning_rate": 5.835411471321696e-08,
+      "loss": 0.2929,
+      "step": 80000
+    },
+    {
+      "epoch": 19.952618453865338,
+      "grad_norm": 8.778485298156738,
+      "learning_rate": 5.5860349127182044e-08,
+      "loss": 0.3606,
+      "step": 80010
+    },
+    {
+      "epoch": 19.955112219451372,
+      "grad_norm": 6.685991287231445,
+      "learning_rate": 5.336658354114714e-08,
+      "loss": 0.3066,
+      "step": 80020
+    },
+    {
+      "epoch": 19.957605985037407,
+      "grad_norm": 11.452533721923828,
+      "learning_rate": 5.087281795511222e-08,
+      "loss": 0.284,
+      "step": 80030
+    },
+    {
+      "epoch": 19.960099750623442,
+      "grad_norm": 9.021291732788086,
+      "learning_rate": 4.8379052369077305e-08,
+      "loss": 0.3149,
+      "step": 80040
+    },
+    {
+      "epoch": 19.962593516209477,
+      "grad_norm": 8.216629981994629,
+      "learning_rate": 4.58852867830424e-08,
+      "loss": 0.3321,
+      "step": 80050
+    },
+    {
+      "epoch": 19.965087281795512,
+      "grad_norm": 9.343437194824219,
+      "learning_rate": 4.339152119700749e-08,
+      "loss": 0.325,
+      "step": 80060
+    },
+    {
+      "epoch": 19.967581047381547,
+      "grad_norm": 7.705197811126709,
+      "learning_rate": 4.089775561097257e-08,
+      "loss": 0.3533,
+      "step": 80070
+    },
+    {
+      "epoch": 19.97007481296758,
+      "grad_norm": 8.739502906799316,
+      "learning_rate": 3.840399002493766e-08,
+      "loss": 0.3282,
+      "step": 80080
+    },
+    {
+      "epoch": 19.972568578553616,
+      "grad_norm": 10.272692680358887,
+      "learning_rate": 3.591022443890275e-08,
+      "loss": 0.3355,
+      "step": 80090
+    },
+    {
+      "epoch": 19.97506234413965,
+      "grad_norm": 7.91961145401001,
+      "learning_rate": 3.341645885286783e-08,
+      "loss": 0.342,
+      "step": 80100
+    },
+    {
+      "epoch": 19.977556109725686,
+      "grad_norm": 7.7013936042785645,
+      "learning_rate": 3.0922693266832924e-08,
+      "loss": 0.3139,
+      "step": 80110
+    },
+    {
+      "epoch": 19.98004987531172,
+      "grad_norm": 8.244073867797852,
+      "learning_rate": 2.8428927680798006e-08,
+      "loss": 0.3602,
+      "step": 80120
+    },
+    {
+      "epoch": 19.982543640897756,
+      "grad_norm": 9.814148902893066,
+      "learning_rate": 2.5935162094763094e-08,
+      "loss": 0.3641,
+      "step": 80130
+    },
+    {
+      "epoch": 19.98503740648379,
+      "grad_norm": 7.1216020584106445,
+      "learning_rate": 2.344139650872818e-08,
+      "loss": 0.326,
+      "step": 80140
+    },
+    {
+      "epoch": 19.987531172069826,
+      "grad_norm": 9.763239860534668,
+      "learning_rate": 2.094763092269327e-08,
+      "loss": 0.3333,
+      "step": 80150
+    },
+    {
+      "epoch": 19.99002493765586,
+      "grad_norm": 7.955997943878174,
+      "learning_rate": 1.8453865336658355e-08,
+      "loss": 0.3693,
+      "step": 80160
+    },
+    {
+      "epoch": 19.992518703241895,
+      "grad_norm": 8.751893997192383,
+      "learning_rate": 1.5960099750623443e-08,
+      "loss": 0.3065,
+      "step": 80170
+    },
+    {
+      "epoch": 19.99501246882793,
+      "grad_norm": 11.452372550964355,
+      "learning_rate": 1.346633416458853e-08,
+      "loss": 0.3088,
+      "step": 80180
+    },
+    {
+      "epoch": 19.997506234413965,
+      "grad_norm": 7.998083591461182,
+      "learning_rate": 1.0972568578553617e-08,
+      "loss": 0.2875,
+      "step": 80190
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 9.69516372680664,
+      "learning_rate": 8.478802992518703e-09,
+      "loss": 0.3132,
+      "step": 80200
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 80200,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.9992048140288e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}