diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21043 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.747538882740141,
+  "eval_steps": 500,
+  "global_step": 30000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005825129609133803,
+      "grad_norm": 16.363956451416016,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.8717,
+      "step": 10
+    },
+    {
+      "epoch": 0.0011650259218267606,
+      "grad_norm": 13.027254104614258,
+      "learning_rate": 1.2666666666666667e-06,
+      "loss": 0.9411,
+      "step": 20
+    },
+    {
+      "epoch": 0.001747538882740141,
+      "grad_norm": 8.33905029296875,
+      "learning_rate": 1.9333333333333336e-06,
+      "loss": 0.6971,
+      "step": 30
+    },
+    {
+      "epoch": 0.0023300518436535213,
+      "grad_norm": 3.80228590965271,
+      "learning_rate": 2.6e-06,
+      "loss": 0.4438,
+      "step": 40
+    },
+    {
+      "epoch": 0.0029125648045669017,
+      "grad_norm": 3.3239755630493164,
+      "learning_rate": 3.2666666666666666e-06,
+      "loss": 0.2813,
+      "step": 50
+    },
+    {
+      "epoch": 0.003495077765480282,
+      "grad_norm": 2.004411458969116,
+      "learning_rate": 3.9333333333333335e-06,
+      "loss": 0.2361,
+      "step": 60
+    },
+    {
+      "epoch": 0.004077590726393662,
+      "grad_norm": 2.5876970291137695,
+      "learning_rate": 4.6e-06,
+      "loss": 0.1895,
+      "step": 70
+    },
+    {
+      "epoch": 0.0046601036873070425,
+      "grad_norm": 2.054030656814575,
+      "learning_rate": 5.266666666666667e-06,
+      "loss": 0.1678,
+      "step": 80
+    },
+    {
+      "epoch": 0.005242616648220423,
+      "grad_norm": 2.5571675300598145,
+      "learning_rate": 5.933333333333334e-06,
+      "loss": 0.1363,
+      "step": 90
+    },
+    {
+      "epoch": 0.005825129609133803,
+      "grad_norm": 2.476600408554077,
+      "learning_rate": 6.6e-06,
+      "loss": 0.1123,
+      "step": 100
+    },
+    {
+      "epoch": 0.006407642570047184,
+      "grad_norm": 1.5339707136154175,
+      "learning_rate": 7.266666666666668e-06,
+      "loss": 0.1119,
+      "step": 110
+    },
+    {
+      "epoch": 0.006990155530960564,
+      "grad_norm": 1.620774269104004,
+      "learning_rate": 7.933333333333334e-06,
+      "loss": 0.091,
+      "step": 120
+    },
+    {
+      "epoch": 0.007572668491873944,
+      "grad_norm": 1.8176084756851196,
+      "learning_rate": 8.599999999999999e-06,
+      "loss": 0.098,
+      "step": 130
+    },
+    {
+      "epoch": 0.008155181452787324,
+      "grad_norm": 1.9751254320144653,
+      "learning_rate": 9.266666666666667e-06,
+      "loss": 0.0937,
+      "step": 140
+    },
+    {
+      "epoch": 0.008737694413700705,
+      "grad_norm": 2.1015803813934326,
+      "learning_rate": 9.933333333333334e-06,
+      "loss": 0.0909,
+      "step": 150
+    },
+    {
+      "epoch": 0.009320207374614085,
+      "grad_norm": 1.4475528001785278,
+      "learning_rate": 1.06e-05,
+      "loss": 0.0868,
+      "step": 160
+    },
+    {
+      "epoch": 0.009902720335527465,
+      "grad_norm": 1.1314893960952759,
+      "learning_rate": 1.1266666666666667e-05,
+      "loss": 0.0783,
+      "step": 170
+    },
+    {
+      "epoch": 0.010485233296440846,
+      "grad_norm": 1.3180276155471802,
+      "learning_rate": 1.1933333333333333e-05,
+      "loss": 0.0701,
+      "step": 180
+    },
+    {
+      "epoch": 0.011067746257354226,
+      "grad_norm": 1.5609411001205444,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 0.0751,
+      "step": 190
+    },
+    {
+      "epoch": 0.011650259218267607,
+      "grad_norm": 1.6207369565963745,
+      "learning_rate": 1.3266666666666666e-05,
+      "loss": 0.072,
+      "step": 200
+    },
+    {
+      "epoch": 0.012232772179180987,
+      "grad_norm": 1.1604082584381104,
+      "learning_rate": 1.3933333333333334e-05,
+      "loss": 0.0822,
+      "step": 210
+    },
+    {
+      "epoch": 0.012815285140094368,
+      "grad_norm": 1.1570616960525513,
+      "learning_rate": 1.4599999999999999e-05,
+      "loss": 0.077,
+      "step": 220
+    },
+    {
+      "epoch": 0.013397798101007748,
+      "grad_norm": 1.152485728263855,
+      "learning_rate": 1.5266666666666667e-05,
+      "loss": 0.0611,
+      "step": 230
+    },
+    {
+      "epoch": 0.013980311061921128,
+      "grad_norm": 2.1177918910980225,
+      "learning_rate": 1.5933333333333332e-05,
+      "loss": 0.0761,
+      "step": 240
+    },
+    {
+      "epoch": 0.014562824022834509,
+      "grad_norm": 1.538111686706543,
+      "learning_rate": 1.66e-05,
+      "loss": 0.0659,
+      "step": 250
+    },
+    {
+      "epoch": 0.015145336983747888,
+      "grad_norm": 1.2610279321670532,
+      "learning_rate": 1.726666666666667e-05,
+      "loss": 0.059,
+      "step": 260
+    },
+    {
+      "epoch": 0.015727849944661268,
+      "grad_norm": 1.6613271236419678,
+      "learning_rate": 1.7933333333333337e-05,
+      "loss": 0.064,
+      "step": 270
+    },
+    {
+      "epoch": 0.01631036290557465,
+      "grad_norm": 1.3400022983551025,
+      "learning_rate": 1.86e-05,
+      "loss": 0.0624,
+      "step": 280
+    },
+    {
+      "epoch": 0.01689287586648803,
+      "grad_norm": 0.9490658640861511,
+      "learning_rate": 1.926666666666667e-05,
+      "loss": 0.0664,
+      "step": 290
+    },
+    {
+      "epoch": 0.01747538882740141,
+      "grad_norm": 1.3591103553771973,
+      "learning_rate": 1.9933333333333334e-05,
+      "loss": 0.0605,
+      "step": 300
+    },
+    {
+      "epoch": 0.01805790178831479,
+      "grad_norm": 1.059062123298645,
+      "learning_rate": 2.06e-05,
+      "loss": 0.0577,
+      "step": 310
+    },
+    {
+      "epoch": 0.01864041474922817,
+      "grad_norm": 1.0266015529632568,
+      "learning_rate": 2.1266666666666667e-05,
+      "loss": 0.054,
+      "step": 320
+    },
+    {
+      "epoch": 0.01922292771014155,
+      "grad_norm": 1.577591061592102,
+      "learning_rate": 2.1933333333333332e-05,
+      "loss": 0.0588,
+      "step": 330
+    },
+    {
+      "epoch": 0.01980544067105493,
+      "grad_norm": 1.5641520023345947,
+      "learning_rate": 2.26e-05,
+      "loss": 0.0664,
+      "step": 340
+    },
+    {
+      "epoch": 0.02038795363196831,
+      "grad_norm": 1.0901275873184204,
+      "learning_rate": 2.326666666666667e-05,
+      "loss": 0.048,
+      "step": 350
+    },
+    {
+      "epoch": 0.020970466592881692,
+      "grad_norm": 1.6798902750015259,
+      "learning_rate": 2.3933333333333337e-05,
+      "loss": 0.0597,
+      "step": 360
+    },
+    {
+      "epoch": 0.021552979553795072,
+      "grad_norm": 0.9511136412620544,
+      "learning_rate": 2.46e-05,
+      "loss": 0.0704,
+      "step": 370
+    },
+    {
+      "epoch": 0.022135492514708453,
+      "grad_norm": 0.9565727710723877,
+      "learning_rate": 2.5266666666666666e-05,
+      "loss": 0.0477,
+      "step": 380
+    },
+    {
+      "epoch": 0.022718005475621833,
+      "grad_norm": 1.4605202674865723,
+      "learning_rate": 2.5933333333333338e-05,
+      "loss": 0.0591,
+      "step": 390
+    },
+    {
+      "epoch": 0.023300518436535213,
+      "grad_norm": 1.688125491142273,
+      "learning_rate": 2.6600000000000003e-05,
+      "loss": 0.06,
+      "step": 400
+    },
+    {
+      "epoch": 0.023883031397448594,
+      "grad_norm": 1.2992297410964966,
+      "learning_rate": 2.7266666666666668e-05,
+      "loss": 0.0556,
+      "step": 410
+    },
+    {
+      "epoch": 0.024465544358361974,
+      "grad_norm": 1.2986173629760742,
+      "learning_rate": 2.7933333333333332e-05,
+      "loss": 0.0538,
+      "step": 420
+    },
+    {
+      "epoch": 0.025048057319275355,
+      "grad_norm": 1.1889179944992065,
+      "learning_rate": 2.86e-05,
+      "loss": 0.0621,
+      "step": 430
+    },
+    {
+      "epoch": 0.025630570280188735,
+      "grad_norm": 1.1551380157470703,
+      "learning_rate": 2.926666666666667e-05,
+      "loss": 0.0632,
+      "step": 440
+    },
+    {
+      "epoch": 0.026213083241102116,
+      "grad_norm": 1.21445894241333,
+      "learning_rate": 2.9933333333333337e-05,
+      "loss": 0.0549,
+      "step": 450
+    },
+    {
+      "epoch": 0.026795596202015496,
+      "grad_norm": 0.9449787735939026,
+      "learning_rate": 3.06e-05,
+      "loss": 0.0657,
+      "step": 460
+    },
+    {
+      "epoch": 0.027378109162928876,
+      "grad_norm": 0.9851808547973633,
+      "learning_rate": 3.126666666666666e-05,
+      "loss": 0.0492,
+      "step": 470
+    },
+    {
+      "epoch": 0.027960622123842257,
+      "grad_norm": 0.7608017921447754,
+      "learning_rate": 3.1933333333333335e-05,
+      "loss": 0.0496,
+      "step": 480
+    },
+    {
+      "epoch": 0.028543135084755637,
+      "grad_norm": 1.0604749917984009,
+      "learning_rate": 3.26e-05,
+      "loss": 0.0469,
+      "step": 490
+    },
+    {
+      "epoch": 0.029125648045669018,
+      "grad_norm": 0.817071259021759,
+      "learning_rate": 3.326666666666667e-05,
+      "loss": 0.0403,
+      "step": 500
+    },
+    {
+      "epoch": 0.029708161006582398,
+      "grad_norm": 1.0470209121704102,
+      "learning_rate": 3.3933333333333336e-05,
+      "loss": 0.0453,
+      "step": 510
+    },
+    {
+      "epoch": 0.030290673967495775,
+      "grad_norm": 1.0401445627212524,
+      "learning_rate": 3.46e-05,
+      "loss": 0.0423,
+      "step": 520
+    },
+    {
+      "epoch": 0.030873186928409155,
+      "grad_norm": 1.1210923194885254,
+      "learning_rate": 3.526666666666667e-05,
+      "loss": 0.0462,
+      "step": 530
+    },
+    {
+      "epoch": 0.031455699889322536,
+      "grad_norm": 1.0154286623001099,
+      "learning_rate": 3.593333333333334e-05,
+      "loss": 0.0426,
+      "step": 540
+    },
+    {
+      "epoch": 0.03203821285023592,
+      "grad_norm": 0.725735604763031,
+      "learning_rate": 3.66e-05,
+      "loss": 0.0378,
+      "step": 550
+    },
+    {
+      "epoch": 0.0326207258111493,
+      "grad_norm": 0.6761109828948975,
+      "learning_rate": 3.726666666666667e-05,
+      "loss": 0.0392,
+      "step": 560
+    },
+    {
+      "epoch": 0.03320323877206268,
+      "grad_norm": 0.8301836848258972,
+      "learning_rate": 3.793333333333334e-05,
+      "loss": 0.0332,
+      "step": 570
+    },
+    {
+      "epoch": 0.03378575173297606,
+      "grad_norm": 0.8133602142333984,
+      "learning_rate": 3.86e-05,
+      "loss": 0.0355,
+      "step": 580
+    },
+    {
+      "epoch": 0.03436826469388944,
+      "grad_norm": 1.0713573694229126,
+      "learning_rate": 3.926666666666667e-05,
+      "loss": 0.0438,
+      "step": 590
+    },
+    {
+      "epoch": 0.03495077765480282,
+      "grad_norm": 1.272591233253479,
+      "learning_rate": 3.993333333333333e-05,
+      "loss": 0.0375,
+      "step": 600
+    },
+    {
+      "epoch": 0.0355332906157162,
+      "grad_norm": 0.6690412759780884,
+      "learning_rate": 4.0600000000000004e-05,
+      "loss": 0.0429,
+      "step": 610
+    },
+    {
+      "epoch": 0.03611580357662958,
+      "grad_norm": 0.8729681968688965,
+      "learning_rate": 4.126666666666667e-05,
+      "loss": 0.041,
+      "step": 620
+    },
+    {
+      "epoch": 0.03669831653754296,
+      "grad_norm": 0.7829736471176147,
+      "learning_rate": 4.1933333333333334e-05,
+      "loss": 0.0353,
+      "step": 630
+    },
+    {
+      "epoch": 0.03728082949845634,
+      "grad_norm": 0.564466118812561,
+      "learning_rate": 4.26e-05,
+      "loss": 0.0335,
+      "step": 640
+    },
+    {
+      "epoch": 0.037863342459369724,
+      "grad_norm": 0.9312437772750854,
+      "learning_rate": 4.3266666666666664e-05,
+      "loss": 0.0388,
+      "step": 650
+    },
+    {
+      "epoch": 0.0384458554202831,
+      "grad_norm": 0.9627359509468079,
+      "learning_rate": 4.3933333333333335e-05,
+      "loss": 0.038,
+      "step": 660
+    },
+    {
+      "epoch": 0.039028368381196485,
+      "grad_norm": 1.1554635763168335,
+      "learning_rate": 4.46e-05,
+      "loss": 0.0416,
+      "step": 670
+    },
+    {
+      "epoch": 0.03961088134210986,
+      "grad_norm": 1.2683342695236206,
+      "learning_rate": 4.526666666666667e-05,
+      "loss": 0.0425,
+      "step": 680
+    },
+    {
+      "epoch": 0.040193394303023246,
+      "grad_norm": 0.7468678951263428,
+      "learning_rate": 4.5933333333333336e-05,
+      "loss": 0.0359,
+      "step": 690
+    },
+    {
+      "epoch": 0.04077590726393662,
+      "grad_norm": 0.7981159687042236,
+      "learning_rate": 4.660000000000001e-05,
+      "loss": 0.0387,
+      "step": 700
+    },
+    {
+      "epoch": 0.04135842022485,
+      "grad_norm": 0.746918797492981,
+      "learning_rate": 4.726666666666667e-05,
+      "loss": 0.0469,
+      "step": 710
+    },
+    {
+      "epoch": 0.041940933185763384,
+      "grad_norm": 0.6890844702720642,
+      "learning_rate": 4.793333333333334e-05,
+      "loss": 0.038,
+      "step": 720
+    },
+    {
+      "epoch": 0.04252344614667676,
+      "grad_norm": 1.093159556388855,
+      "learning_rate": 4.86e-05,
+      "loss": 0.0376,
+      "step": 730
+    },
+    {
+      "epoch": 0.043105959107590144,
+      "grad_norm": 0.7858609557151794,
+      "learning_rate": 4.926666666666667e-05,
+      "loss": 0.0332,
+      "step": 740
+    },
+    {
+      "epoch": 0.04368847206850352,
+      "grad_norm": 1.066215991973877,
+      "learning_rate": 4.993333333333334e-05,
+      "loss": 0.0353,
+      "step": 750
+    },
+    {
+      "epoch": 0.044270985029416905,
+      "grad_norm": 0.7255071401596069,
+      "learning_rate": 5.0600000000000003e-05,
+      "loss": 0.0291,
+      "step": 760
+    },
+    {
+      "epoch": 0.04485349799033028,
+      "grad_norm": 0.451518177986145,
+      "learning_rate": 5.1266666666666675e-05,
+      "loss": 0.0312,
+      "step": 770
+    },
+    {
+      "epoch": 0.045436010951243666,
+      "grad_norm": 0.6643555164337158,
+      "learning_rate": 5.193333333333333e-05,
+      "loss": 0.0314,
+      "step": 780
+    },
+    {
+      "epoch": 0.04601852391215704,
+      "grad_norm": 0.5722026824951172,
+      "learning_rate": 5.2600000000000005e-05,
+      "loss": 0.0291,
+      "step": 790
+    },
+    {
+      "epoch": 0.04660103687307043,
+      "grad_norm": 0.7939475178718567,
+      "learning_rate": 5.326666666666666e-05,
+      "loss": 0.0368,
+      "step": 800
+    },
+    {
+      "epoch": 0.047183549833983804,
+      "grad_norm": 0.8282892107963562,
+      "learning_rate": 5.3933333333333334e-05,
+      "loss": 0.0349,
+      "step": 810
+    },
+    {
+      "epoch": 0.04776606279489719,
+      "grad_norm": 1.4124369621276855,
+      "learning_rate": 5.4600000000000006e-05,
+      "loss": 0.0385,
+      "step": 820
+    },
+    {
+      "epoch": 0.048348575755810565,
+      "grad_norm": 0.6213952302932739,
+      "learning_rate": 5.5266666666666664e-05,
+      "loss": 0.0395,
+      "step": 830
+    },
+    {
+      "epoch": 0.04893108871672395,
+      "grad_norm": 0.5522814393043518,
+      "learning_rate": 5.5933333333333335e-05,
+      "loss": 0.0348,
+      "step": 840
+    },
+    {
+      "epoch": 0.049513601677637326,
+      "grad_norm": 0.5925297141075134,
+      "learning_rate": 5.66e-05,
+      "loss": 0.0351,
+      "step": 850
+    },
+    {
+      "epoch": 0.05009611463855071,
+      "grad_norm": 0.5774127244949341,
+      "learning_rate": 5.726666666666667e-05,
+      "loss": 0.0306,
+      "step": 860
+    },
+    {
+      "epoch": 0.050678627599464086,
+      "grad_norm": 0.7426440119743347,
+      "learning_rate": 5.7933333333333337e-05,
+      "loss": 0.0296,
+      "step": 870
+    },
+    {
+      "epoch": 0.05126114056037747,
+      "grad_norm": 1.0238310098648071,
+      "learning_rate": 5.86e-05,
+      "loss": 0.0309,
+      "step": 880
+    },
+    {
+      "epoch": 0.05184365352129085,
+      "grad_norm": 0.7471153140068054,
+      "learning_rate": 5.926666666666667e-05,
+      "loss": 0.0294,
+      "step": 890
+    },
+    {
+      "epoch": 0.05242616648220423,
+      "grad_norm": 0.8437319993972778,
+      "learning_rate": 5.9933333333333345e-05,
+      "loss": 0.035,
+      "step": 900
+    },
+    {
+      "epoch": 0.05300867944311761,
+      "grad_norm": 0.76613849401474,
+      "learning_rate": 6.06e-05,
+      "loss": 0.0288,
+      "step": 910
+    },
+    {
+      "epoch": 0.05359119240403099,
+      "grad_norm": 0.601531445980072,
+      "learning_rate": 6.126666666666667e-05,
+      "loss": 0.0377,
+      "step": 920
+    },
+    {
+      "epoch": 0.05417370536494437,
+      "grad_norm": 0.5623847246170044,
+      "learning_rate": 6.193333333333333e-05,
+      "loss": 0.0292,
+      "step": 930
+    },
+    {
+      "epoch": 0.05475621832585775,
+      "grad_norm": 0.5545303225517273,
+      "learning_rate": 6.26e-05,
+      "loss": 0.0271,
+      "step": 940
+    },
+    {
+      "epoch": 0.05533873128677113,
+      "grad_norm": 0.876323938369751,
+      "learning_rate": 6.326666666666667e-05,
+      "loss": 0.0356,
+      "step": 950
+    },
+    {
+      "epoch": 0.055921244247684514,
+      "grad_norm": 0.9015308022499084,
+      "learning_rate": 6.393333333333333e-05,
+      "loss": 0.0321,
+      "step": 960
+    },
+    {
+      "epoch": 0.05650375720859789,
+      "grad_norm": 0.3498276472091675,
+      "learning_rate": 6.460000000000001e-05,
+      "loss": 0.0327,
+      "step": 970
+    },
+    {
+      "epoch": 0.057086270169511275,
+      "grad_norm": 0.645955502986908,
+      "learning_rate": 6.526666666666666e-05,
+      "loss": 0.0327,
+      "step": 980
+    },
+    {
+      "epoch": 0.05766878313042465,
+      "grad_norm": 0.7454339265823364,
+      "learning_rate": 6.593333333333334e-05,
+      "loss": 0.0327,
+      "step": 990
+    },
+    {
+      "epoch": 0.058251296091338035,
+      "grad_norm": 0.9000582695007324,
+      "learning_rate": 6.66e-05,
+      "loss": 0.0286,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05883380905225141,
+      "grad_norm": 0.5894749760627747,
+      "learning_rate": 6.726666666666667e-05,
+      "loss": 0.0266,
+      "step": 1010
+    },
+    {
+      "epoch": 0.059416322013164796,
+      "grad_norm": 0.47711148858070374,
+      "learning_rate": 6.793333333333334e-05,
+      "loss": 0.0266,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05999883497407817,
+      "grad_norm": 0.39817479252815247,
+      "learning_rate": 6.860000000000001e-05,
+      "loss": 0.0296,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06058134793499155,
+      "grad_norm": 0.5175421237945557,
+      "learning_rate": 6.926666666666667e-05,
+      "loss": 0.0254,
+      "step": 1040
+    },
+    {
+      "epoch": 0.061163860895904934,
+      "grad_norm": 0.5396063923835754,
+      "learning_rate": 6.993333333333334e-05,
+      "loss": 0.0263,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06174637385681831,
+      "grad_norm": 0.7684789896011353,
+      "learning_rate": 7.06e-05,
+      "loss": 0.0308,
+      "step": 1060
+    },
+    {
+      "epoch": 0.062328886817731695,
+      "grad_norm": 0.4386548697948456,
+      "learning_rate": 7.126666666666667e-05,
+      "loss": 0.0296,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06291139977864507,
+      "grad_norm": 0.602109968662262,
+      "learning_rate": 7.193333333333334e-05,
+      "loss": 0.0307,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06349391273955846,
+      "grad_norm": 0.48234033584594727,
+      "learning_rate": 7.26e-05,
+      "loss": 0.0184,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06407642570047184,
+      "grad_norm": 0.5120774507522583,
+      "learning_rate": 7.326666666666667e-05,
+      "loss": 0.0243,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06465893866138521,
+      "grad_norm": 0.6919184327125549,
+      "learning_rate": 7.393333333333333e-05,
+      "loss": 0.0262,
+      "step": 1110
+    },
+    {
+      "epoch": 0.0652414516222986,
+      "grad_norm": 0.8763561844825745,
+      "learning_rate": 7.46e-05,
+      "loss": 0.0283,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06582396458321198,
+      "grad_norm": 0.6841975450515747,
+      "learning_rate": 7.526666666666668e-05,
+      "loss": 0.0316,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06640647754412536,
+      "grad_norm": 0.4978059232234955,
+      "learning_rate": 7.593333333333334e-05,
+      "loss": 0.0267,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06698899050503873,
+      "grad_norm": 0.7343912124633789,
+      "learning_rate": 7.66e-05,
+      "loss": 0.0304,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06757150346595212,
+      "grad_norm": 0.9718796014785767,
+      "learning_rate": 7.726666666666667e-05,
+      "loss": 0.0298,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0681540164268655,
+      "grad_norm": 0.7595483660697937,
+      "learning_rate": 7.793333333333333e-05,
+      "loss": 0.0252,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06873652938777888,
+      "grad_norm": 0.6774091720581055,
+      "learning_rate": 7.860000000000001e-05,
+      "loss": 0.026,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06931904234869225,
+      "grad_norm": 0.5943043231964111,
+      "learning_rate": 7.926666666666666e-05,
+      "loss": 0.0254,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06990155530960564,
+      "grad_norm": 0.32503804564476013,
+      "learning_rate": 7.993333333333334e-05,
+      "loss": 0.0269,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07048406827051902,
+      "grad_norm": 0.3729705214500427,
+      "learning_rate": 8.060000000000001e-05,
+      "loss": 0.0292,
+      "step": 1210
+    },
+    {
+      "epoch": 0.0710665812314324,
+      "grad_norm": 0.401933878660202,
+      "learning_rate": 8.126666666666667e-05,
+      "loss": 0.0243,
+      "step": 1220
+    },
+    {
+      "epoch": 0.07164909419234577,
+      "grad_norm": 0.6299405694007874,
+      "learning_rate": 8.193333333333334e-05,
+      "loss": 0.0255,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07223160715325916,
+      "grad_norm": 0.38478463888168335,
+      "learning_rate": 8.26e-05,
+      "loss": 0.0219,
+      "step": 1240
+    },
+    {
+      "epoch": 0.07281412011417254,
+      "grad_norm": 0.4080633819103241,
+      "learning_rate": 8.326666666666667e-05,
+      "loss": 0.0272,
+      "step": 1250
+    },
+    {
+      "epoch": 0.07339663307508593,
+      "grad_norm": 0.3903158903121948,
+      "learning_rate": 8.393333333333335e-05,
+      "loss": 0.0224,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0739791460359993,
+      "grad_norm": 0.6705195307731628,
+      "learning_rate": 8.46e-05,
+      "loss": 0.0261,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07456165899691268,
+      "grad_norm": 0.5832329392433167,
+      "learning_rate": 8.526666666666667e-05,
+      "loss": 0.0266,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07514417195782606,
+      "grad_norm": 0.42376530170440674,
+      "learning_rate": 8.593333333333333e-05,
+      "loss": 0.0286,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07572668491873945,
+      "grad_norm": 0.3618919849395752,
+      "learning_rate": 8.66e-05,
+      "loss": 0.0245,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07630919787965282,
+      "grad_norm": 0.5320606827735901,
+      "learning_rate": 8.726666666666667e-05,
+      "loss": 0.0248,
+      "step": 1310
+    },
+    {
+      "epoch": 0.0768917108405662,
+      "grad_norm": 0.6612217426300049,
+      "learning_rate": 8.793333333333333e-05,
+      "loss": 0.0274,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07747422380147959,
+      "grad_norm": 0.6843940019607544,
+      "learning_rate": 8.86e-05,
+      "loss": 0.0249,
+      "step": 1330
+    },
+    {
+      "epoch": 0.07805673676239297,
+      "grad_norm": 0.37553220987319946,
+      "learning_rate": 8.926666666666668e-05,
+      "loss": 0.0252,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07863924972330634,
+      "grad_norm": 0.5357182025909424,
+      "learning_rate": 8.993333333333334e-05,
+      "loss": 0.0251,
+      "step": 1350
+    },
+    {
+      "epoch": 0.07922176268421972,
+      "grad_norm": 0.8621771335601807,
+      "learning_rate": 9.06e-05,
+      "loss": 0.0265,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07980427564513311,
+      "grad_norm": 0.5881845355033875,
+      "learning_rate": 9.126666666666667e-05,
+      "loss": 0.0222,
+      "step": 1370
+    },
+    {
+      "epoch": 0.08038678860604649,
+      "grad_norm": 0.8289936184883118,
+      "learning_rate": 9.193333333333334e-05,
+      "loss": 0.0261,
+      "step": 1380
+    },
+    {
+      "epoch": 0.08096930156695986,
+      "grad_norm": 0.4810025095939636,
+      "learning_rate": 9.260000000000001e-05,
+      "loss": 0.0233,
+      "step": 1390
+    },
+    {
+      "epoch": 0.08155181452787325,
+      "grad_norm": 0.9317886233329773,
+      "learning_rate": 9.326666666666667e-05,
+      "loss": 0.0289,
+      "step": 1400
+    },
+    {
+      "epoch": 0.08213432748878663,
+      "grad_norm": 0.7096083760261536,
+      "learning_rate": 9.393333333333334e-05,
+      "loss": 0.0227,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0827168404497,
+      "grad_norm": 0.5615822672843933,
+      "learning_rate": 9.46e-05,
+      "loss": 0.0293,
+      "step": 1420
+    },
+    {
+      "epoch": 0.08329935341061338,
+      "grad_norm": 0.6307222247123718,
+      "learning_rate": 9.526666666666667e-05,
+      "loss": 0.0225,
+      "step": 1430
+    },
+    {
+      "epoch": 0.08388186637152677,
+      "grad_norm": 0.44919538497924805,
+      "learning_rate": 9.593333333333334e-05,
+      "loss": 0.0237,
+      "step": 1440
+    },
+    {
+      "epoch": 0.08446437933244015,
+      "grad_norm": 0.6162551641464233,
+      "learning_rate": 9.66e-05,
+      "loss": 0.0264,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08504689229335352,
+      "grad_norm": 0.5569806098937988,
+      "learning_rate": 9.726666666666667e-05,
+      "loss": 0.0252,
+      "step": 1460
+    },
+    {
+      "epoch": 0.0856294052542669,
+      "grad_norm": 0.5218840837478638,
+      "learning_rate": 9.793333333333333e-05,
+      "loss": 0.0272,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08621191821518029,
+      "grad_norm": 0.7213922142982483,
+      "learning_rate": 9.86e-05,
+      "loss": 0.0345,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08679443117609367,
+      "grad_norm": 0.35891667008399963,
+      "learning_rate": 9.926666666666668e-05,
+      "loss": 0.0255,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08737694413700704,
+      "grad_norm": 0.36590370535850525,
+      "learning_rate": 9.993333333333334e-05,
+      "loss": 0.033,
+      "step": 1500
+    },
+    {
+      "epoch": 0.08795945709792043,
+      "grad_norm": 0.5863376259803772,
+      "learning_rate": 9.999997539434007e-05,
+      "loss": 0.0236,
+      "step": 1510
+    },
+    {
+      "epoch": 0.08854197005883381,
+      "grad_norm": 0.7955098748207092,
+      "learning_rate": 9.999989033776898e-05,
+      "loss": 0.0255,
+      "step": 1520
+    },
+    {
+      "epoch": 0.0891244830197472,
+      "grad_norm": 0.45443108677864075,
+      "learning_rate": 9.999974452661641e-05,
+      "loss": 0.0259,
+      "step": 1530
+    },
+    {
+      "epoch": 0.08970699598066056,
+      "grad_norm": 0.5477753281593323,
+      "learning_rate": 9.999953796105959e-05,
+      "loss": 0.0254,
+      "step": 1540
+    },
+    {
+      "epoch": 0.09028950894157395,
+      "grad_norm": 0.2657114267349243,
+      "learning_rate": 9.999927064134949e-05,
+      "loss": 0.0247,
+      "step": 1550
+    },
+    {
+      "epoch": 0.09087202190248733,
+      "grad_norm": 0.4057621955871582,
+      "learning_rate": 9.999894256781095e-05,
+      "loss": 0.0223,
+      "step": 1560
+    },
+    {
+      "epoch": 0.09145453486340072,
+      "grad_norm": 0.4122306704521179,
+      "learning_rate": 9.99985537408426e-05,
+      "loss": 0.0216,
+      "step": 1570
+    },
+    {
+      "epoch": 0.09203704782431409,
+      "grad_norm": 0.2921590507030487,
+      "learning_rate": 9.999810416091688e-05,
+      "loss": 0.0231,
+      "step": 1580
+    },
+    {
+      "epoch": 0.09261956078522747,
+      "grad_norm": 0.33729857206344604,
+      "learning_rate": 9.99975938285801e-05,
+      "loss": 0.0202,
+      "step": 1590
+    },
+    {
+      "epoch": 0.09320207374614085,
+      "grad_norm": 0.6590836644172668,
+      "learning_rate": 9.999702274445236e-05,
+      "loss": 0.0209,
+      "step": 1600
+    },
+    {
+      "epoch": 0.09378458670705424,
+      "grad_norm": 0.5845506191253662,
+      "learning_rate": 9.999639090922756e-05,
+      "loss": 0.0204,
+      "step": 1610
+    },
+    {
+      "epoch": 0.09436709966796761,
+      "grad_norm": 0.37395179271698,
+      "learning_rate": 9.999569832367346e-05,
+      "loss": 0.0196,
+      "step": 1620
+    },
+    {
+      "epoch": 0.09494961262888099,
+      "grad_norm": 0.32569620013237,
+      "learning_rate": 9.999494498863162e-05,
+      "loss": 0.0217,
+      "step": 1630
+    },
+    {
+      "epoch": 0.09553212558979438,
+      "grad_norm": 0.5724361538887024,
+      "learning_rate": 9.99941309050174e-05,
+      "loss": 0.0223,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09611463855070776,
+      "grad_norm": 0.41537967324256897,
+      "learning_rate": 9.999325607381999e-05,
+      "loss": 0.0246,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09669715151162113,
+      "grad_norm": 0.473274290561676,
+      "learning_rate": 9.999232049610238e-05,
+      "loss": 0.0181,
+      "step": 1660
+    },
+    {
+      "epoch": 0.09727966447253451,
+      "grad_norm": 0.44749128818511963,
+      "learning_rate": 9.999132417300142e-05,
+      "loss": 0.0217,
+      "step": 1670
+    },
+    {
+      "epoch": 0.0978621774334479,
+      "grad_norm": 0.342777281999588,
+      "learning_rate": 9.99902671057277e-05,
+      "loss": 0.0212,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09844469039436128,
+      "grad_norm": 0.5124594569206238,
+      "learning_rate": 9.998914929556569e-05,
+      "loss": 0.024,
+      "step": 1690
+    },
+    {
+      "epoch": 0.09902720335527465,
+      "grad_norm": 0.6373164653778076,
+      "learning_rate": 9.998797074387361e-05,
+      "loss": 0.0239,
+      "step": 1700
+    },
+    {
+      "epoch": 0.09960971631618804,
+      "grad_norm": 0.5524207353591919,
+      "learning_rate": 9.99867314520835e-05,
+      "loss": 0.019,
+      "step": 1710
+    },
+    {
+      "epoch": 0.10019222927710142,
+      "grad_norm": 0.49288803339004517,
+      "learning_rate": 9.998543142170126e-05,
+      "loss": 0.0223,
+      "step": 1720
+    },
+    {
+      "epoch": 0.10077474223801479,
+      "grad_norm": 0.39513063430786133,
+      "learning_rate": 9.99840706543065e-05,
+      "loss": 0.021,
+      "step": 1730
+    },
+    {
+      "epoch": 0.10135725519892817,
+      "grad_norm": 0.38166385889053345,
+      "learning_rate": 9.998264915155274e-05,
+      "loss": 0.0197,
+      "step": 1740
+    },
+    {
+      "epoch": 0.10193976815984156,
+      "grad_norm": 0.4162251949310303,
+      "learning_rate": 9.998116691516718e-05,
+      "loss": 0.0215,
+      "step": 1750
+    },
+    {
+      "epoch": 0.10252228112075494,
+      "grad_norm": 0.4375555217266083,
+      "learning_rate": 9.997962394695091e-05,
+      "loss": 0.0243,
+      "step": 1760
+    },
+    {
+      "epoch": 0.10310479408166831,
+      "grad_norm": 0.4904995560646057,
+      "learning_rate": 9.997802024877875e-05,
+      "loss": 0.0222,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1036873070425817,
+      "grad_norm": 0.2985130846500397,
+      "learning_rate": 9.99763558225994e-05,
+      "loss": 0.0209,
+      "step": 1780
+    },
+    {
+      "epoch": 0.10426982000349508,
+      "grad_norm": 0.43920576572418213,
+      "learning_rate": 9.997463067043526e-05,
+      "loss": 0.0258,
+      "step": 1790
+    },
+    {
+      "epoch": 0.10485233296440846,
+      "grad_norm": 0.44892460107803345,
+      "learning_rate": 9.997284479438253e-05,
+      "loss": 0.02,
+      "step": 1800
+    },
+    {
+      "epoch": 0.10543484592532183,
+      "grad_norm": 0.48475006222724915,
+      "learning_rate": 9.997099819661127e-05,
+      "loss": 0.0214,
+      "step": 1810
+    },
+    {
+      "epoch": 0.10601735888623522,
+      "grad_norm": 0.5260090827941895,
+      "learning_rate": 9.996909087936524e-05,
+      "loss": 0.0217,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1065998718471486,
+      "grad_norm": 0.4060119092464447,
+      "learning_rate": 9.996712284496201e-05,
+      "loss": 0.02,
+      "step": 1830
+    },
+    {
+      "epoch": 0.10718238480806198,
+      "grad_norm": 0.5117402076721191,
+      "learning_rate": 9.996509409579293e-05,
+      "loss": 0.0205,
+      "step": 1840
+    },
+    {
+      "epoch": 0.10776489776897535,
+      "grad_norm": 0.36153656244277954,
+      "learning_rate": 9.996300463432312e-05,
+      "loss": 0.0197,
+      "step": 1850
+    },
+    {
+      "epoch": 0.10834741072988874,
+      "grad_norm": 0.5402187705039978,
+      "learning_rate": 9.996085446309148e-05,
+      "loss": 0.0233,
+      "step": 1860
+    },
+    {
+      "epoch": 0.10892992369080212,
+      "grad_norm": 0.5023298859596252,
+      "learning_rate": 9.995864358471066e-05,
+      "loss": 0.0234,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1095124366517155,
+      "grad_norm": 0.42993468046188354,
+      "learning_rate": 9.99563720018671e-05,
+      "loss": 0.0167,
+      "step": 1880
+    },
+    {
+      "epoch": 0.11009494961262888,
+      "grad_norm": 0.4474211633205414,
+      "learning_rate": 9.995403971732098e-05,
+      "loss": 0.0187,
+      "step": 1890
+    },
+    {
+      "epoch": 0.11067746257354226,
+      "grad_norm": 0.3085026144981384,
+      "learning_rate": 9.995164673390625e-05,
+      "loss": 0.0238,
+      "step": 1900
+    },
+    {
+      "epoch": 0.11125997553445564,
+      "grad_norm": 0.571968674659729,
+      "learning_rate": 9.994919305453059e-05,
+      "loss": 0.0211,
+      "step": 1910
+    },
+    {
+      "epoch": 0.11184248849536903,
+      "grad_norm": 0.439012348651886,
+      "learning_rate": 9.994667868217548e-05,
+      "loss": 0.0214,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1124250014562824,
+      "grad_norm": 0.33423957228660583,
+      "learning_rate": 9.99441036198961e-05,
+      "loss": 0.0203,
+      "step": 1930
+    },
+    {
+      "epoch": 0.11300751441719578,
+      "grad_norm": 0.3952094614505768,
+      "learning_rate": 9.99414678708214e-05,
+      "loss": 0.019,
+      "step": 1940
+    },
+    {
+      "epoch": 0.11359002737810917,
+      "grad_norm": 0.4013804793357849,
+      "learning_rate": 9.993877143815407e-05,
+      "loss": 0.0203,
+      "step": 1950
+    },
+    {
+      "epoch": 0.11417254033902255,
+      "grad_norm": 0.43464595079421997,
+      "learning_rate": 9.993601432517053e-05,
+      "loss": 0.0172,
+      "step": 1960
+    },
+    {
+      "epoch": 0.11475505329993592,
+      "grad_norm": 0.6095263957977295,
+      "learning_rate": 9.993319653522091e-05,
+      "loss": 0.0226,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1153375662608493,
+      "grad_norm": 0.4852878451347351,
+      "learning_rate": 9.993031807172911e-05,
+      "loss": 0.024,
+      "step": 1980
+    },
+    {
+      "epoch": 0.11592007922176269,
+      "grad_norm": 0.4639914929866791,
+      "learning_rate": 9.992737893819273e-05,
+      "loss": 0.0191,
+      "step": 1990
+    },
+    {
+      "epoch": 0.11650259218267607,
+      "grad_norm": 0.2555379271507263,
+      "learning_rate": 9.992437913818312e-05,
+      "loss": 0.0177,
+      "step": 2000
+    },
+    {
+      "epoch": 0.11708510514358944,
+      "grad_norm": 0.3941870927810669,
+      "learning_rate": 9.992131867534526e-05,
+      "loss": 0.0171,
+      "step": 2010
+    },
+    {
+      "epoch": 0.11766761810450282,
+      "grad_norm": 0.33638933300971985,
+      "learning_rate": 9.991819755339796e-05,
+      "loss": 0.022,
+      "step": 2020
+    },
+    {
+      "epoch": 0.11825013106541621,
+      "grad_norm": 0.2429458647966385,
+      "learning_rate": 9.991501577613365e-05,
+      "loss": 0.0214,
+      "step": 2030
+    },
+    {
+      "epoch": 0.11883264402632959,
+      "grad_norm": 0.3649020791053772,
+      "learning_rate": 9.99117733474185e-05,
+      "loss": 0.0232,
+      "step": 2040
+    },
+    {
+      "epoch": 0.11941515698724296,
+      "grad_norm": 0.22641423344612122,
+      "learning_rate": 9.990847027119234e-05,
+      "loss": 0.0187,
+      "step": 2050
+    },
+    {
+      "epoch": 0.11999766994815635,
+      "grad_norm": 0.2840929925441742,
+      "learning_rate": 9.990510655146877e-05,
+      "loss": 0.0162,
+      "step": 2060
+    },
+    {
+      "epoch": 0.12058018290906973,
+      "grad_norm": 0.33555904030799866,
+      "learning_rate": 9.990168219233496e-05,
+      "loss": 0.0166,
+      "step": 2070
+    },
+    {
+      "epoch": 0.1211626958699831,
+      "grad_norm": 0.3273191750049591,
+      "learning_rate": 9.989819719795188e-05,
+      "loss": 0.0164,
+      "step": 2080
+    },
+    {
+      "epoch": 0.12174520883089648,
+      "grad_norm": 0.29209548234939575,
+      "learning_rate": 9.989465157255412e-05,
+      "loss": 0.0167,
+      "step": 2090
+    },
+    {
+      "epoch": 0.12232772179180987,
+      "grad_norm": 0.6209645867347717,
+      "learning_rate": 9.989104532044994e-05,
+      "loss": 0.0171,
+      "step": 2100
+    },
+    {
+      "epoch": 0.12291023475272325,
+      "grad_norm": 0.5538815259933472,
+      "learning_rate": 9.988737844602128e-05,
+      "loss": 0.0168,
+      "step": 2110
+    },
+    {
+      "epoch": 0.12349274771363662,
+      "grad_norm": 0.39251554012298584,
+      "learning_rate": 9.988365095372372e-05,
+      "loss": 0.017,
+      "step": 2120
+    },
+    {
+      "epoch": 0.12407526067455,
+      "grad_norm": 0.3591337502002716,
+      "learning_rate": 9.987986284808654e-05,
+      "loss": 0.0169,
+      "step": 2130
+    },
+    {
+      "epoch": 0.12465777363546339,
+      "grad_norm": 0.5226990580558777,
+      "learning_rate": 9.987601413371264e-05,
+      "loss": 0.0162,
+      "step": 2140
+    },
+    {
+      "epoch": 0.12524028659637676,
+      "grad_norm": 0.4623308479785919,
+      "learning_rate": 9.987210481527855e-05,
+      "loss": 0.0215,
+      "step": 2150
+    },
+    {
+      "epoch": 0.12582279955729014,
+      "grad_norm": 0.4177090525627136,
+      "learning_rate": 9.98681348975345e-05,
+      "loss": 0.0181,
+      "step": 2160
+    },
+    {
+      "epoch": 0.12640531251820353,
+      "grad_norm": 0.43934598565101624,
+      "learning_rate": 9.986410438530427e-05,
+      "loss": 0.0174,
+      "step": 2170
+    },
+    {
+      "epoch": 0.1269878254791169,
+      "grad_norm": 0.4332578480243683,
+      "learning_rate": 9.986001328348534e-05,
+      "loss": 0.02,
+      "step": 2180
+    },
+    {
+      "epoch": 0.1275703384400303,
+      "grad_norm": 0.29921674728393555,
+      "learning_rate": 9.985586159704878e-05,
+      "loss": 0.0165,
+      "step": 2190
+    },
+    {
+      "epoch": 0.12815285140094368,
+      "grad_norm": 0.3050669729709625,
+      "learning_rate": 9.985164933103929e-05,
+      "loss": 0.0174,
+      "step": 2200
+    },
+    {
+      "epoch": 0.12873536436185706,
+      "grad_norm": 0.33571580052375793,
+      "learning_rate": 9.984737649057513e-05,
+      "loss": 0.0146,
+      "step": 2210
+    },
+    {
+      "epoch": 0.12931787732277042,
+      "grad_norm": 0.3501420319080353,
+      "learning_rate": 9.984304308084827e-05,
+      "loss": 0.0159,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1299003902836838,
+      "grad_norm": 0.350046843290329,
+      "learning_rate": 9.983864910712416e-05,
+      "loss": 0.0189,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1304829032445972,
+      "grad_norm": 0.45016878843307495,
+      "learning_rate": 9.98341945747419e-05,
+      "loss": 0.0164,
+      "step": 2240
+    },
+    {
+      "epoch": 0.13106541620551057,
+      "grad_norm": 0.19611197710037231,
+      "learning_rate": 9.98296794891142e-05,
+      "loss": 0.0213,
+      "step": 2250
+    },
+    {
+      "epoch": 0.13164792916642395,
+      "grad_norm": 0.4170686602592468,
+      "learning_rate": 9.982510385572725e-05,
+      "loss": 0.0177,
+      "step": 2260
+    },
+    {
+      "epoch": 0.13223044212733734,
+      "grad_norm": 0.49550408124923706,
+      "learning_rate": 9.982046768014094e-05,
+      "loss": 0.0161,
+      "step": 2270
+    },
+    {
+      "epoch": 0.13281295508825072,
+      "grad_norm": 0.4277210533618927,
+      "learning_rate": 9.981577096798863e-05,
+      "loss": 0.0205,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1333954680491641,
+      "grad_norm": 0.2560710906982422,
+      "learning_rate": 9.981101372497727e-05,
+      "loss": 0.0155,
+      "step": 2290
+    },
+    {
+      "epoch": 0.13397798101007746,
+      "grad_norm": 0.5669534802436829,
+      "learning_rate": 9.980619595688737e-05,
+      "loss": 0.0139,
+      "step": 2300
+    },
+    {
+      "epoch": 0.13456049397099085,
+      "grad_norm": 0.31216830015182495,
+      "learning_rate": 9.980131766957295e-05,
+      "loss": 0.0204,
+      "step": 2310
+    },
+    {
+      "epoch": 0.13514300693190423,
+      "grad_norm": 0.32178542017936707,
+      "learning_rate": 9.979637886896163e-05,
+      "loss": 0.0124,
+      "step": 2320
+    },
+    {
+      "epoch": 0.13572551989281761,
+      "grad_norm": 0.48147812485694885,
+      "learning_rate": 9.979137956105447e-05,
+      "loss": 0.0231,
+      "step": 2330
+    },
+    {
+      "epoch": 0.136308032853731,
+      "grad_norm": 0.5376265645027161,
+      "learning_rate": 9.978631975192613e-05,
+      "loss": 0.0174,
+      "step": 2340
+    },
+    {
+      "epoch": 0.13689054581464438,
+      "grad_norm": 0.5305449962615967,
+      "learning_rate": 9.978119944772475e-05,
+      "loss": 0.0207,
+      "step": 2350
+    },
+    {
+      "epoch": 0.13747305877555777,
+      "grad_norm": 0.5217538475990295,
+      "learning_rate": 9.977601865467197e-05,
+      "loss": 0.0166,
+      "step": 2360
+    },
+    {
+      "epoch": 0.13805557173647115,
+      "grad_norm": 0.4606058597564697,
+      "learning_rate": 9.977077737906297e-05,
+      "loss": 0.0174,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1386380846973845,
+      "grad_norm": 0.42594361305236816,
+      "learning_rate": 9.976547562726636e-05,
+      "loss": 0.0164,
+      "step": 2380
+    },
+    {
+      "epoch": 0.1392205976582979,
+      "grad_norm": 0.26246941089630127,
+      "learning_rate": 9.976011340572429e-05,
+      "loss": 0.0198,
+      "step": 2390
+    },
+    {
+      "epoch": 0.13980311061921127,
+      "grad_norm": 0.3303341567516327,
+      "learning_rate": 9.975469072095237e-05,
+      "loss": 0.0178,
+      "step": 2400
+    },
+    {
+      "epoch": 0.14038562358012466,
+      "grad_norm": 0.3827970325946808,
+      "learning_rate": 9.974920757953965e-05,
+      "loss": 0.0149,
+      "step": 2410
+    },
+    {
+      "epoch": 0.14096813654103804,
+      "grad_norm": 0.22160081565380096,
+      "learning_rate": 9.97436639881487e-05,
+      "loss": 0.0146,
+      "step": 2420
+    },
+    {
+      "epoch": 0.14155064950195143,
+      "grad_norm": 0.4541112184524536,
+      "learning_rate": 9.973805995351545e-05,
+      "loss": 0.0146,
+      "step": 2430
+    },
+    {
+      "epoch": 0.1421331624628648,
+      "grad_norm": 0.2938708961009979,
+      "learning_rate": 9.973239548244939e-05,
+      "loss": 0.0156,
+      "step": 2440
+    },
+    {
+      "epoch": 0.14271567542377817,
+      "grad_norm": 0.24117307364940643,
+      "learning_rate": 9.972667058183333e-05,
+      "loss": 0.0162,
+      "step": 2450
+    },
+    {
+      "epoch": 0.14329818838469155,
+      "grad_norm": 0.4317592680454254,
+      "learning_rate": 9.972088525862362e-05,
+      "loss": 0.0182,
+      "step": 2460
+    },
+    {
+      "epoch": 0.14388070134560493,
+      "grad_norm": 0.3063121438026428,
+      "learning_rate": 9.971503951984995e-05,
+      "loss": 0.015,
+      "step": 2470
+    },
+    {
+      "epoch": 0.14446321430651832,
+      "grad_norm": 0.28073713183403015,
+      "learning_rate": 9.970913337261543e-05,
+      "loss": 0.0158,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1450457272674317,
+      "grad_norm": 0.4092753827571869,
+      "learning_rate": 9.97031668240966e-05,
+      "loss": 0.0131,
+      "step": 2490
+    },
+    {
+      "epoch": 0.14562824022834508,
+      "grad_norm": 0.4061177968978882,
+      "learning_rate": 9.969713988154339e-05,
+      "loss": 0.0171,
+      "step": 2500
+    },
+    {
+      "epoch": 0.14621075318925847,
+      "grad_norm": 0.45309191942214966,
+      "learning_rate": 9.969105255227906e-05,
+      "loss": 0.0157,
+      "step": 2510
+    },
+    {
+      "epoch": 0.14679326615017185,
+      "grad_norm": 0.48115354776382446,
+      "learning_rate": 9.968490484370035e-05,
+      "loss": 0.0146,
+      "step": 2520
+    },
+    {
+      "epoch": 0.1473757791110852,
+      "grad_norm": 0.3232123851776123,
+      "learning_rate": 9.967869676327726e-05,
+      "loss": 0.0134,
+      "step": 2530
+    },
+    {
+      "epoch": 0.1479582920719986,
+      "grad_norm": 0.3205547630786896,
+      "learning_rate": 9.96724283185532e-05,
+      "loss": 0.0181,
+      "step": 2540
+    },
+    {
+      "epoch": 0.14854080503291198,
+      "grad_norm": 0.5272349119186401,
+      "learning_rate": 9.966609951714494e-05,
+      "loss": 0.0191,
+      "step": 2550
+    },
+    {
+      "epoch": 0.14912331799382536,
+      "grad_norm": 0.2887498736381531,
+      "learning_rate": 9.965971036674255e-05,
+      "loss": 0.0188,
+      "step": 2560
+    },
+    {
+      "epoch": 0.14970583095473874,
+      "grad_norm": 0.42976537346839905,
+      "learning_rate": 9.965326087510947e-05,
+      "loss": 0.0181,
+      "step": 2570
+    },
+    {
+      "epoch": 0.15028834391565213,
+      "grad_norm": 0.6130855679512024,
+      "learning_rate": 9.964675105008244e-05,
+      "loss": 0.0196,
+      "step": 2580
+    },
+    {
+      "epoch": 0.1508708568765655,
+      "grad_norm": 0.6346950531005859,
+      "learning_rate": 9.964018089957147e-05,
+      "loss": 0.0196,
+      "step": 2590
+    },
+    {
+      "epoch": 0.1514533698374789,
+      "grad_norm": 0.539539635181427,
+      "learning_rate": 9.963355043155996e-05,
+      "loss": 0.0165,
+      "step": 2600
+    },
+    {
+      "epoch": 0.15203588279839225,
+      "grad_norm": 0.3816796839237213,
+      "learning_rate": 9.962685965410455e-05,
+      "loss": 0.0156,
+      "step": 2610
+    },
+    {
+      "epoch": 0.15261839575930564,
+      "grad_norm": 0.409084677696228,
+      "learning_rate": 9.962010857533514e-05,
+      "loss": 0.0169,
+      "step": 2620
+    },
+    {
+      "epoch": 0.15320090872021902,
+      "grad_norm": 0.2978951036930084,
+      "learning_rate": 9.961329720345493e-05,
+      "loss": 0.0162,
+      "step": 2630
+    },
+    {
+      "epoch": 0.1537834216811324,
+      "grad_norm": 0.370525062084198,
+      "learning_rate": 9.96064255467404e-05,
+      "loss": 0.0136,
+      "step": 2640
+    },
+    {
+      "epoch": 0.1543659346420458,
+      "grad_norm": 0.4469973146915436,
+      "learning_rate": 9.959949361354126e-05,
+      "loss": 0.0193,
+      "step": 2650
+    },
+    {
+      "epoch": 0.15494844760295917,
+      "grad_norm": 0.2943956255912781,
+      "learning_rate": 9.959250141228045e-05,
+      "loss": 0.0184,
+      "step": 2660
+    },
+    {
+      "epoch": 0.15553096056387256,
+      "grad_norm": 0.22035115957260132,
+      "learning_rate": 9.958544895145414e-05,
+      "loss": 0.0168,
+      "step": 2670
+    },
+    {
+      "epoch": 0.15611347352478594,
+      "grad_norm": 0.340677410364151,
+      "learning_rate": 9.957833623963177e-05,
+      "loss": 0.0182,
+      "step": 2680
+    },
+    {
+      "epoch": 0.1566959864856993,
+      "grad_norm": 0.4560448229312897,
+      "learning_rate": 9.957116328545593e-05,
+      "loss": 0.0163,
+      "step": 2690
+    },
+    {
+      "epoch": 0.15727849944661268,
+      "grad_norm": 0.3833587169647217,
+      "learning_rate": 9.956393009764244e-05,
+      "loss": 0.0188,
+      "step": 2700
+    },
+    {
+      "epoch": 0.15786101240752606,
+      "grad_norm": 0.20285329222679138,
+      "learning_rate": 9.955663668498032e-05,
+      "loss": 0.016,
+      "step": 2710
+    },
+    {
+      "epoch": 0.15844352536843945,
+      "grad_norm": 0.4248286485671997,
+      "learning_rate": 9.954928305633173e-05,
+      "loss": 0.0141,
+      "step": 2720
+    },
+    {
+      "epoch": 0.15902603832935283,
+      "grad_norm": 0.34970593452453613,
+      "learning_rate": 9.954186922063204e-05,
+      "loss": 0.0171,
+      "step": 2730
+    },
+    {
+      "epoch": 0.15960855129026622,
+      "grad_norm": 0.33577001094818115,
+      "learning_rate": 9.953439518688974e-05,
+      "loss": 0.0179,
+      "step": 2740
+    },
+    {
+      "epoch": 0.1601910642511796,
+      "grad_norm": 0.3365096151828766,
+      "learning_rate": 9.952686096418652e-05,
+      "loss": 0.0137,
+      "step": 2750
+    },
+    {
+      "epoch": 0.16077357721209298,
+      "grad_norm": 0.2108718603849411,
+      "learning_rate": 9.951926656167714e-05,
+      "loss": 0.0172,
+      "step": 2760
+    },
+    {
+      "epoch": 0.16135609017300634,
+      "grad_norm": 0.4207550883293152,
+      "learning_rate": 9.951161198858953e-05,
+      "loss": 0.0162,
+      "step": 2770
+    },
+    {
+      "epoch": 0.16193860313391972,
+      "grad_norm": 0.4917376935482025,
+      "learning_rate": 9.950389725422471e-05,
+      "loss": 0.0155,
+      "step": 2780
+    },
+    {
+      "epoch": 0.1625211160948331,
+      "grad_norm": 0.35805121064186096,
+      "learning_rate": 9.949612236795682e-05,
+      "loss": 0.0128,
+      "step": 2790
+    },
+    {
+      "epoch": 0.1631036290557465,
+      "grad_norm": 0.44710585474967957,
+      "learning_rate": 9.948828733923305e-05,
+      "loss": 0.0156,
+      "step": 2800
+    },
+    {
+      "epoch": 0.16368614201665987,
+      "grad_norm": 0.2824556827545166,
+      "learning_rate": 9.948039217757374e-05,
+      "loss": 0.0161,
+      "step": 2810
+    },
+    {
+      "epoch": 0.16426865497757326,
+      "grad_norm": 0.483117014169693,
+      "learning_rate": 9.947243689257225e-05,
+      "loss": 0.0153,
+      "step": 2820
+    },
+    {
+      "epoch": 0.16485116793848664,
+      "grad_norm": 0.3247731328010559,
+      "learning_rate": 9.946442149389497e-05,
+      "loss": 0.0172,
+      "step": 2830
+    },
+    {
+      "epoch": 0.1654336808994,
+      "grad_norm": 0.3819945752620697,
+      "learning_rate": 9.945634599128139e-05,
+      "loss": 0.0177,
+      "step": 2840
+    },
+    {
+      "epoch": 0.16601619386031338,
+      "grad_norm": 0.4045076370239258,
+      "learning_rate": 9.944821039454402e-05,
+      "loss": 0.0154,
+      "step": 2850
+    },
+    {
+      "epoch": 0.16659870682122677,
+      "grad_norm": 0.4406282901763916,
+      "learning_rate": 9.944001471356835e-05,
+      "loss": 0.0145,
+      "step": 2860
+    },
+    {
+      "epoch": 0.16718121978214015,
+      "grad_norm": 0.28208911418914795,
+      "learning_rate": 9.94317589583129e-05,
+      "loss": 0.0138,
+      "step": 2870
+    },
+    {
+      "epoch": 0.16776373274305353,
+      "grad_norm": 0.35577231645584106,
+      "learning_rate": 9.942344313880922e-05,
+      "loss": 0.0165,
+      "step": 2880
+    },
+    {
+      "epoch": 0.16834624570396692,
+      "grad_norm": 0.28937041759490967,
+      "learning_rate": 9.941506726516179e-05,
+      "loss": 0.0155,
+      "step": 2890
+    },
+    {
+      "epoch": 0.1689287586648803,
+      "grad_norm": 0.39862045645713806,
+      "learning_rate": 9.94066313475481e-05,
+      "loss": 0.0166,
+      "step": 2900
+    },
+    {
+      "epoch": 0.16951127162579369,
+      "grad_norm": 0.3496819734573364,
+      "learning_rate": 9.939813539621857e-05,
+      "loss": 0.0202,
+      "step": 2910
+    },
+    {
+      "epoch": 0.17009378458670704,
+      "grad_norm": 0.396270751953125,
+      "learning_rate": 9.93895794214966e-05,
+      "loss": 0.0164,
+      "step": 2920
+    },
+    {
+      "epoch": 0.17067629754762043,
+      "grad_norm": 0.39987051486968994,
+      "learning_rate": 9.938096343377852e-05,
+      "loss": 0.016,
+      "step": 2930
+    },
+    {
+      "epoch": 0.1712588105085338,
+      "grad_norm": 0.33410680294036865,
+      "learning_rate": 9.937228744353353e-05,
+      "loss": 0.0166,
+      "step": 2940
+    },
+    {
+      "epoch": 0.1718413234694472,
+      "grad_norm": 0.27986234426498413,
+      "learning_rate": 9.936355146130379e-05,
+      "loss": 0.0142,
+      "step": 2950
+    },
+    {
+      "epoch": 0.17242383643036058,
+      "grad_norm": 0.3302169740200043,
+      "learning_rate": 9.935475549770435e-05,
+      "loss": 0.0174,
+      "step": 2960
+    },
+    {
+      "epoch": 0.17300634939127396,
+      "grad_norm": 0.41995158791542053,
+      "learning_rate": 9.934589956342315e-05,
+      "loss": 0.0153,
+      "step": 2970
+    },
+    {
+      "epoch": 0.17358886235218735,
+      "grad_norm": 0.3632122576236725,
+      "learning_rate": 9.933698366922093e-05,
+      "loss": 0.0132,
+      "step": 2980
+    },
+    {
+      "epoch": 0.17417137531310073,
+      "grad_norm": 0.3555792272090912,
+      "learning_rate": 9.93280078259314e-05,
+      "loss": 0.0121,
+      "step": 2990
+    },
+    {
+      "epoch": 0.17475388827401409,
+      "grad_norm": 0.32381510734558105,
+      "learning_rate": 9.931897204446104e-05,
+      "loss": 0.0158,
+      "step": 3000
+    },
+    {
+      "epoch": 0.17533640123492747,
+      "grad_norm": 0.32888510823249817,
+      "learning_rate": 9.930987633578915e-05,
+      "loss": 0.0146,
+      "step": 3010
+    },
+    {
+      "epoch": 0.17591891419584085,
+      "grad_norm": 0.39167606830596924,
+      "learning_rate": 9.93007207109679e-05,
+      "loss": 0.0187,
+      "step": 3020
+    },
+    {
+      "epoch": 0.17650142715675424,
+      "grad_norm": 0.634312093257904,
+      "learning_rate": 9.929150518112224e-05,
+      "loss": 0.0163,
+      "step": 3030
+    },
+    {
+      "epoch": 0.17708394011766762,
+      "grad_norm": 0.347868949174881,
+      "learning_rate": 9.928222975744991e-05,
+      "loss": 0.0141,
+      "step": 3040
+    },
+    {
+      "epoch": 0.177666453078581,
+      "grad_norm": 0.37778526544570923,
+      "learning_rate": 9.92728944512214e-05,
+      "loss": 0.0152,
+      "step": 3050
+    },
+    {
+      "epoch": 0.1782489660394944,
+      "grad_norm": 0.3820255398750305,
+      "learning_rate": 9.926349927378001e-05,
+      "loss": 0.0153,
+      "step": 3060
+    },
+    {
+      "epoch": 0.17883147900040777,
+      "grad_norm": 0.4140007495880127,
+      "learning_rate": 9.925404423654174e-05,
+      "loss": 0.0192,
+      "step": 3070
+    },
+    {
+      "epoch": 0.17941399196132113,
+      "grad_norm": 0.46465981006622314,
+      "learning_rate": 9.924452935099537e-05,
+      "loss": 0.0165,
+      "step": 3080
+    },
+    {
+      "epoch": 0.1799965049222345,
+      "grad_norm": 0.3738744556903839,
+      "learning_rate": 9.92349546287024e-05,
+      "loss": 0.0159,
+      "step": 3090
+    },
+    {
+      "epoch": 0.1805790178831479,
+      "grad_norm": 0.39833900332450867,
+      "learning_rate": 9.9225320081297e-05,
+      "loss": 0.0164,
+      "step": 3100
+    },
+    {
+      "epoch": 0.18116153084406128,
+      "grad_norm": 0.5395527482032776,
+      "learning_rate": 9.921562572048606e-05,
+      "loss": 0.018,
+      "step": 3110
+    },
+    {
+      "epoch": 0.18174404380497466,
+      "grad_norm": 0.4580261707305908,
+      "learning_rate": 9.920587155804913e-05,
+      "loss": 0.0159,
+      "step": 3120
+    },
+    {
+      "epoch": 0.18232655676588805,
+      "grad_norm": 0.42417433857917786,
+      "learning_rate": 9.919605760583845e-05,
+      "loss": 0.0187,
+      "step": 3130
+    },
+    {
+      "epoch": 0.18290906972680143,
+      "grad_norm": 0.33866560459136963,
+      "learning_rate": 9.91861838757789e-05,
+      "loss": 0.0164,
+      "step": 3140
+    },
+    {
+      "epoch": 0.1834915826877148,
+      "grad_norm": 0.4014289677143097,
+      "learning_rate": 9.917625037986798e-05,
+      "loss": 0.0162,
+      "step": 3150
+    },
+    {
+      "epoch": 0.18407409564862817,
+      "grad_norm": 0.2678658664226532,
+      "learning_rate": 9.916625713017583e-05,
+      "loss": 0.0142,
+      "step": 3160
+    },
+    {
+      "epoch": 0.18465660860954156,
+      "grad_norm": 0.24174220860004425,
+      "learning_rate": 9.915620413884519e-05,
+      "loss": 0.0155,
+      "step": 3170
+    },
+    {
+      "epoch": 0.18523912157045494,
+      "grad_norm": 0.46617671847343445,
+      "learning_rate": 9.914609141809139e-05,
+      "loss": 0.0162,
+      "step": 3180
+    },
+    {
+      "epoch": 0.18582163453136832,
+      "grad_norm": 0.4024803638458252,
+      "learning_rate": 9.913591898020235e-05,
+      "loss": 0.0172,
+      "step": 3190
+    },
+    {
+      "epoch": 0.1864041474922817,
+      "grad_norm": 0.6289913654327393,
+      "learning_rate": 9.912568683753853e-05,
+      "loss": 0.0209,
+      "step": 3200
+    },
+    {
+      "epoch": 0.1869866604531951,
+      "grad_norm": 0.346633642911911,
+      "learning_rate": 9.911539500253295e-05,
+      "loss": 0.0158,
+      "step": 3210
+    },
+    {
+      "epoch": 0.18756917341410848,
+      "grad_norm": 0.37537068128585815,
+      "learning_rate": 9.910504348769118e-05,
+      "loss": 0.0148,
+      "step": 3220
+    },
+    {
+      "epoch": 0.18815168637502183,
+      "grad_norm": 0.4973176121711731,
+      "learning_rate": 9.909463230559127e-05,
+      "loss": 0.0142,
+      "step": 3230
+    },
+    {
+      "epoch": 0.18873419933593522,
+      "grad_norm": 0.3805164396762848,
+      "learning_rate": 9.908416146888376e-05,
+      "loss": 0.0115,
+      "step": 3240
+    },
+    {
+      "epoch": 0.1893167122968486,
+      "grad_norm": 0.39441925287246704,
+      "learning_rate": 9.907363099029175e-05,
+      "loss": 0.0128,
+      "step": 3250
+    },
+    {
+      "epoch": 0.18989922525776198,
+      "grad_norm": 0.2503442168235779,
+      "learning_rate": 9.906304088261073e-05,
+      "loss": 0.0206,
+      "step": 3260
+    },
+    {
+      "epoch": 0.19048173821867537,
+      "grad_norm": 0.29791730642318726,
+      "learning_rate": 9.905239115870872e-05,
+      "loss": 0.0156,
+      "step": 3270
+    },
+    {
+      "epoch": 0.19106425117958875,
+      "grad_norm": 0.3576847016811371,
+      "learning_rate": 9.90416818315261e-05,
+      "loss": 0.0129,
+      "step": 3280
+    },
+    {
+      "epoch": 0.19164676414050213,
+      "grad_norm": 0.4002215266227722,
+      "learning_rate": 9.903091291407573e-05,
+      "loss": 0.015,
+      "step": 3290
+    },
+    {
+      "epoch": 0.19222927710141552,
+      "grad_norm": 0.44520682096481323,
+      "learning_rate": 9.902008441944286e-05,
+      "loss": 0.014,
+      "step": 3300
+    },
+    {
+      "epoch": 0.19281179006232887,
+      "grad_norm": 0.48155078291893005,
+      "learning_rate": 9.900919636078512e-05,
+      "loss": 0.0141,
+      "step": 3310
+    },
+    {
+      "epoch": 0.19339430302324226,
+      "grad_norm": 0.22092358767986298,
+      "learning_rate": 9.899824875133255e-05,
+      "loss": 0.014,
+      "step": 3320
+    },
+    {
+      "epoch": 0.19397681598415564,
+      "grad_norm": 0.48741286993026733,
+      "learning_rate": 9.898724160438749e-05,
+      "loss": 0.0142,
+      "step": 3330
+    },
+    {
+      "epoch": 0.19455932894506903,
+      "grad_norm": 0.34430861473083496,
+      "learning_rate": 9.89761749333247e-05,
+      "loss": 0.0138,
+      "step": 3340
+    },
+    {
+      "epoch": 0.1951418419059824,
+      "grad_norm": 0.3429625928401947,
+      "learning_rate": 9.896504875159122e-05,
+      "loss": 0.017,
+      "step": 3350
+    },
+    {
+      "epoch": 0.1957243548668958,
+      "grad_norm": 0.297616571187973,
+      "learning_rate": 9.89538630727064e-05,
+      "loss": 0.0131,
+      "step": 3360
+    },
+    {
+      "epoch": 0.19630686782780918,
+      "grad_norm": 0.39675647020339966,
+      "learning_rate": 9.894261791026189e-05,
+      "loss": 0.0207,
+      "step": 3370
+    },
+    {
+      "epoch": 0.19688938078872256,
+      "grad_norm": 0.29995664954185486,
+      "learning_rate": 9.893131327792165e-05,
+      "loss": 0.0162,
+      "step": 3380
+    },
+    {
+      "epoch": 0.19747189374963592,
+      "grad_norm": 0.33954164385795593,
+      "learning_rate": 9.891994918942182e-05,
+      "loss": 0.0161,
+      "step": 3390
+    },
+    {
+      "epoch": 0.1980544067105493,
+      "grad_norm": 0.5150769948959351,
+      "learning_rate": 9.890852565857092e-05,
+      "loss": 0.0167,
+      "step": 3400
+    },
+    {
+      "epoch": 0.1986369196714627,
+      "grad_norm": 0.5345335006713867,
+      "learning_rate": 9.889704269924954e-05,
+      "loss": 0.0161,
+      "step": 3410
+    },
+    {
+      "epoch": 0.19921943263237607,
+      "grad_norm": 0.30202922224998474,
+      "learning_rate": 9.888550032541059e-05,
+      "loss": 0.0187,
+      "step": 3420
+    },
+    {
+      "epoch": 0.19980194559328945,
+      "grad_norm": 0.3381350636482239,
+      "learning_rate": 9.887389855107916e-05,
+      "loss": 0.0146,
+      "step": 3430
+    },
+    {
+      "epoch": 0.20038445855420284,
+      "grad_norm": 0.5011172890663147,
+      "learning_rate": 9.886223739035248e-05,
+      "loss": 0.0169,
+      "step": 3440
+    },
+    {
+      "epoch": 0.20096697151511622,
+      "grad_norm": 0.269399493932724,
+      "learning_rate": 9.885051685739997e-05,
+      "loss": 0.0137,
+      "step": 3450
+    },
+    {
+      "epoch": 0.20154948447602958,
+      "grad_norm": 0.5072921514511108,
+      "learning_rate": 9.883873696646316e-05,
+      "loss": 0.0187,
+      "step": 3460
+    },
+    {
+      "epoch": 0.20213199743694296,
+      "grad_norm": 0.24462972581386566,
+      "learning_rate": 9.882689773185575e-05,
+      "loss": 0.0137,
+      "step": 3470
+    },
+    {
+      "epoch": 0.20271451039785635,
+      "grad_norm": 0.3352923095226288,
+      "learning_rate": 9.881499916796353e-05,
+      "loss": 0.0146,
+      "step": 3480
+    },
+    {
+      "epoch": 0.20329702335876973,
+      "grad_norm": 0.28809764981269836,
+      "learning_rate": 9.880304128924434e-05,
+      "loss": 0.0144,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2038795363196831,
+      "grad_norm": 0.4793824553489685,
+      "learning_rate": 9.879102411022817e-05,
+      "loss": 0.0179,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2044620492805965,
+      "grad_norm": 0.25081145763397217,
+      "learning_rate": 9.877894764551703e-05,
+      "loss": 0.0133,
+      "step": 3510
+    },
+    {
+      "epoch": 0.20504456224150988,
+      "grad_norm": 0.2768265902996063,
+      "learning_rate": 9.876681190978494e-05,
+      "loss": 0.0121,
+      "step": 3520
+    },
+    {
+      "epoch": 0.20562707520242327,
+      "grad_norm": 0.42935487627983093,
+      "learning_rate": 9.875461691777797e-05,
+      "loss": 0.0143,
+      "step": 3530
+    },
+    {
+      "epoch": 0.20620958816333662,
+      "grad_norm": 0.32817867398262024,
+      "learning_rate": 9.874236268431417e-05,
+      "loss": 0.0162,
+      "step": 3540
+    },
+    {
+      "epoch": 0.20679210112425,
+      "grad_norm": 0.3461551368236542,
+      "learning_rate": 9.873004922428361e-05,
+      "loss": 0.016,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2073746140851634,
+      "grad_norm": 0.5130866765975952,
+      "learning_rate": 9.871767655264829e-05,
+      "loss": 0.0157,
+      "step": 3560
+    },
+    {
+      "epoch": 0.20795712704607677,
+      "grad_norm": 0.34219974279403687,
+      "learning_rate": 9.87052446844422e-05,
+      "loss": 0.0181,
+      "step": 3570
+    },
+    {
+      "epoch": 0.20853964000699016,
+      "grad_norm": 0.3319697380065918,
+      "learning_rate": 9.869275363477122e-05,
+      "loss": 0.0159,
+      "step": 3580
+    },
+    {
+      "epoch": 0.20912215296790354,
+      "grad_norm": 0.37241482734680176,
+      "learning_rate": 9.868020341881312e-05,
+      "loss": 0.0132,
+      "step": 3590
+    },
+    {
+      "epoch": 0.20970466592881692,
+      "grad_norm": 0.34658288955688477,
+      "learning_rate": 9.866759405181765e-05,
+      "loss": 0.0163,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2102871788897303,
+      "grad_norm": 0.24060270190238953,
+      "learning_rate": 9.865492554910633e-05,
+      "loss": 0.0127,
+      "step": 3610
+    },
+    {
+      "epoch": 0.21086969185064366,
+      "grad_norm": 0.46318456530570984,
+      "learning_rate": 9.864219792607262e-05,
+      "loss": 0.016,
+      "step": 3620
+    },
+    {
+      "epoch": 0.21145220481155705,
+      "grad_norm": 0.30252811312675476,
+      "learning_rate": 9.862941119818177e-05,
+      "loss": 0.0168,
+      "step": 3630
+    },
+    {
+      "epoch": 0.21203471777247043,
+      "grad_norm": 0.39923396706581116,
+      "learning_rate": 9.861656538097086e-05,
+      "loss": 0.0165,
+      "step": 3640
+    },
+    {
+      "epoch": 0.21261723073338382,
+      "grad_norm": 0.3751615881919861,
+      "learning_rate": 9.860366049004877e-05,
+      "loss": 0.0188,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2131997436942972,
+      "grad_norm": 0.4151049852371216,
+      "learning_rate": 9.859069654109615e-05,
+      "loss": 0.0136,
+      "step": 3660
+    },
+    {
+      "epoch": 0.21378225665521058,
+      "grad_norm": 0.49764227867126465,
+      "learning_rate": 9.857767354986545e-05,
+      "loss": 0.0154,
+      "step": 3670
+    },
+    {
+      "epoch": 0.21436476961612397,
+      "grad_norm": 0.32790184020996094,
+      "learning_rate": 9.856459153218078e-05,
+      "loss": 0.0201,
+      "step": 3680
+    },
+    {
+      "epoch": 0.21494728257703735,
+      "grad_norm": 0.3363580107688904,
+      "learning_rate": 9.855145050393808e-05,
+      "loss": 0.0183,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2155297955379507,
+      "grad_norm": 0.36230507493019104,
+      "learning_rate": 9.85382504811049e-05,
+      "loss": 0.0165,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2161123084988641,
+      "grad_norm": 0.3681294023990631,
+      "learning_rate": 9.852499147972054e-05,
+      "loss": 0.0198,
+      "step": 3710
+    },
+    {
+      "epoch": 0.21669482145977748,
+      "grad_norm": 0.27238598465919495,
+      "learning_rate": 9.851167351589592e-05,
+      "loss": 0.0142,
+      "step": 3720
+    },
+    {
+      "epoch": 0.21727733442069086,
+      "grad_norm": 0.20715878903865814,
+      "learning_rate": 9.849829660581363e-05,
+      "loss": 0.0153,
+      "step": 3730
+    },
+    {
+      "epoch": 0.21785984738160424,
+      "grad_norm": 0.3693470358848572,
+      "learning_rate": 9.848486076572787e-05,
+      "loss": 0.0181,
+      "step": 3740
+    },
+    {
+      "epoch": 0.21844236034251763,
+      "grad_norm": 0.4169458746910095,
+      "learning_rate": 9.847136601196446e-05,
+      "loss": 0.0138,
+      "step": 3750
+    },
+    {
+      "epoch": 0.219024873303431,
+      "grad_norm": 0.4135362505912781,
+      "learning_rate": 9.845781236092078e-05,
+      "loss": 0.0146,
+      "step": 3760
+    },
+    {
+      "epoch": 0.2196073862643444,
+      "grad_norm": 0.40244174003601074,
+      "learning_rate": 9.844419982906583e-05,
+      "loss": 0.0203,
+      "step": 3770
+    },
+    {
+      "epoch": 0.22018989922525775,
+      "grad_norm": 0.2031620591878891,
+      "learning_rate": 9.843052843294008e-05,
+      "loss": 0.014,
+      "step": 3780
+    },
+    {
+      "epoch": 0.22077241218617114,
+      "grad_norm": 0.42439550161361694,
+      "learning_rate": 9.841679818915559e-05,
+      "loss": 0.0137,
+      "step": 3790
+    },
+    {
+      "epoch": 0.22135492514708452,
+      "grad_norm": 0.2717705965042114,
+      "learning_rate": 9.840300911439591e-05,
+      "loss": 0.0172,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2219374381079979,
+      "grad_norm": 0.21812152862548828,
+      "learning_rate": 9.838916122541603e-05,
+      "loss": 0.0123,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2225199510689113,
+      "grad_norm": 0.3498571813106537,
+      "learning_rate": 9.837525453904246e-05,
+      "loss": 0.0164,
+      "step": 3820
+    },
+    {
+      "epoch": 0.22310246402982467,
+      "grad_norm": 0.4606907069683075,
+      "learning_rate": 9.836128907217314e-05,
+      "loss": 0.0181,
+      "step": 3830
+    },
+    {
+      "epoch": 0.22368497699073805,
+      "grad_norm": 0.47948184609413147,
+      "learning_rate": 9.834726484177743e-05,
+      "loss": 0.0171,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2242674899516514,
+      "grad_norm": 0.3982204794883728,
+      "learning_rate": 9.833318186489609e-05,
+      "loss": 0.0165,
+      "step": 3850
+    },
+    {
+      "epoch": 0.2248500029125648,
+      "grad_norm": 0.3378174304962158,
+      "learning_rate": 9.831904015864126e-05,
+      "loss": 0.0159,
+      "step": 3860
+    },
+    {
+      "epoch": 0.22543251587347818,
+      "grad_norm": 0.3384462893009186,
+      "learning_rate": 9.830483974019645e-05,
+      "loss": 0.0144,
+      "step": 3870
+    },
+    {
+      "epoch": 0.22601502883439156,
+      "grad_norm": 0.3724633753299713,
+      "learning_rate": 9.82905806268165e-05,
+      "loss": 0.0135,
+      "step": 3880
+    },
+    {
+      "epoch": 0.22659754179530495,
+      "grad_norm": 0.28294652700424194,
+      "learning_rate": 9.82762628358276e-05,
+      "loss": 0.0131,
+      "step": 3890
+    },
+    {
+      "epoch": 0.22718005475621833,
+      "grad_norm": 0.35287946462631226,
+      "learning_rate": 9.826188638462718e-05,
+      "loss": 0.013,
+      "step": 3900
+    },
+    {
+      "epoch": 0.22776256771713171,
+      "grad_norm": 0.22180771827697754,
+      "learning_rate": 9.824745129068402e-05,
+      "loss": 0.0141,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2283450806780451,
+      "grad_norm": 0.3079909086227417,
+      "learning_rate": 9.82329575715381e-05,
+      "loss": 0.0133,
+      "step": 3920
+    },
+    {
+      "epoch": 0.22892759363895845,
+      "grad_norm": 0.25783053040504456,
+      "learning_rate": 9.821840524480066e-05,
+      "loss": 0.0165,
+      "step": 3930
+    },
+    {
+      "epoch": 0.22951010659987184,
+      "grad_norm": 0.3007711172103882,
+      "learning_rate": 9.820379432815414e-05,
+      "loss": 0.0161,
+      "step": 3940
+    },
+    {
+      "epoch": 0.23009261956078522,
+      "grad_norm": 0.39270880818367004,
+      "learning_rate": 9.81891248393522e-05,
+      "loss": 0.0163,
+      "step": 3950
+    },
+    {
+      "epoch": 0.2306751325216986,
+      "grad_norm": 0.3316587507724762,
+      "learning_rate": 9.817439679621963e-05,
+      "loss": 0.0142,
+      "step": 3960
+    },
+    {
+      "epoch": 0.231257645482612,
+      "grad_norm": 0.38680002093315125,
+      "learning_rate": 9.815961021665243e-05,
+      "loss": 0.0169,
+      "step": 3970
+    },
+    {
+      "epoch": 0.23184015844352537,
+      "grad_norm": 0.48907414078712463,
+      "learning_rate": 9.814476511861763e-05,
+      "loss": 0.0143,
+      "step": 3980
+    },
+    {
+      "epoch": 0.23242267140443876,
+      "grad_norm": 0.25044673681259155,
+      "learning_rate": 9.812986152015348e-05,
+      "loss": 0.0136,
+      "step": 3990
+    },
+    {
+      "epoch": 0.23300518436535214,
+      "grad_norm": 0.32564136385917664,
+      "learning_rate": 9.811489943936922e-05,
+      "loss": 0.0157,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2335876973262655,
+      "grad_norm": 0.5283328890800476,
+      "learning_rate": 9.809987889444522e-05,
+      "loss": 0.0151,
+      "step": 4010
+    },
+    {
+      "epoch": 0.23417021028717888,
+      "grad_norm": 0.4246067702770233,
+      "learning_rate": 9.808479990363282e-05,
+      "loss": 0.0133,
+      "step": 4020
+    },
+    {
+      "epoch": 0.23475272324809227,
+      "grad_norm": 0.3396816849708557,
+      "learning_rate": 9.806966248525445e-05,
+      "loss": 0.0158,
+      "step": 4030
+    },
+    {
+      "epoch": 0.23533523620900565,
+      "grad_norm": 0.10669705271720886,
+      "learning_rate": 9.805446665770348e-05,
+      "loss": 0.014,
+      "step": 4040
+    },
+    {
+      "epoch": 0.23591774916991903,
+      "grad_norm": 0.2744532823562622,
+      "learning_rate": 9.803921243944429e-05,
+      "loss": 0.0113,
+      "step": 4050
+    },
+    {
+      "epoch": 0.23650026213083242,
+      "grad_norm": 0.403071790933609,
+      "learning_rate": 9.802389984901218e-05,
+      "loss": 0.0167,
+      "step": 4060
+    },
+    {
+      "epoch": 0.2370827750917458,
+      "grad_norm": 0.18635526299476624,
+      "learning_rate": 9.80085289050134e-05,
+      "loss": 0.0127,
+      "step": 4070
+    },
+    {
+      "epoch": 0.23766528805265918,
+      "grad_norm": 0.4232059121131897,
+      "learning_rate": 9.799309962612508e-05,
+      "loss": 0.013,
+      "step": 4080
+    },
+    {
+      "epoch": 0.23824780101357254,
+      "grad_norm": 0.32140085101127625,
+      "learning_rate": 9.797761203109527e-05,
+      "loss": 0.0136,
+      "step": 4090
+    },
+    {
+      "epoch": 0.23883031397448592,
+      "grad_norm": 0.3512478470802307,
+      "learning_rate": 9.796206613874283e-05,
+      "loss": 0.0145,
+      "step": 4100
+    },
+    {
+      "epoch": 0.2394128269353993,
+      "grad_norm": 0.3515501618385315,
+      "learning_rate": 9.794646196795754e-05,
+      "loss": 0.0138,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2399953398963127,
+      "grad_norm": 0.2739296853542328,
+      "learning_rate": 9.793079953769987e-05,
+      "loss": 0.0135,
+      "step": 4120
+    },
+    {
+      "epoch": 0.24057785285722608,
+      "grad_norm": 0.27916398644447327,
+      "learning_rate": 9.79150788670012e-05,
+      "loss": 0.0146,
+      "step": 4130
+    },
+    {
+      "epoch": 0.24116036581813946,
+      "grad_norm": 0.22873377799987793,
+      "learning_rate": 9.78992999749636e-05,
+      "loss": 0.0143,
+      "step": 4140
+    },
+    {
+      "epoch": 0.24174287877905284,
+      "grad_norm": 0.37163299322128296,
+      "learning_rate": 9.788346288075994e-05,
+      "loss": 0.0134,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2423253917399662,
+      "grad_norm": 0.44292131066322327,
+      "learning_rate": 9.786756760363373e-05,
+      "loss": 0.0135,
+      "step": 4160
+    },
+    {
+      "epoch": 0.24290790470087958,
+      "grad_norm": 0.3150562644004822,
+      "learning_rate": 9.78516141628993e-05,
+      "loss": 0.0164,
+      "step": 4170
+    },
+    {
+      "epoch": 0.24349041766179297,
+      "grad_norm": 0.2714001536369324,
+      "learning_rate": 9.783560257794154e-05,
+      "loss": 0.0155,
+      "step": 4180
+    },
+    {
+      "epoch": 0.24407293062270635,
+      "grad_norm": 0.32230204343795776,
+      "learning_rate": 9.781953286821603e-05,
+      "loss": 0.0156,
+      "step": 4190
+    },
+    {
+      "epoch": 0.24465544358361974,
+      "grad_norm": 0.4852038621902466,
+      "learning_rate": 9.780340505324901e-05,
+      "loss": 0.0136,
+      "step": 4200
+    },
+    {
+      "epoch": 0.24523795654453312,
+      "grad_norm": 0.3366095721721649,
+      "learning_rate": 9.778721915263727e-05,
+      "loss": 0.0122,
+      "step": 4210
+    },
+    {
+      "epoch": 0.2458204695054465,
+      "grad_norm": 0.30512264370918274,
+      "learning_rate": 9.777097518604824e-05,
+      "loss": 0.0133,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2464029824663599,
+      "grad_norm": 0.3890034556388855,
+      "learning_rate": 9.775467317321984e-05,
+      "loss": 0.0119,
+      "step": 4230
+    },
+    {
+      "epoch": 0.24698549542727324,
+      "grad_norm": 0.33380448818206787,
+      "learning_rate": 9.773831313396055e-05,
+      "loss": 0.0147,
+      "step": 4240
+    },
+    {
+      "epoch": 0.24756800838818663,
+      "grad_norm": 0.3554973006248474,
+      "learning_rate": 9.77218950881494e-05,
+      "loss": 0.015,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2481505213491,
+      "grad_norm": 0.2725274860858917,
+      "learning_rate": 9.770541905573583e-05,
+      "loss": 0.0169,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2487330343100134,
+      "grad_norm": 0.26896587014198303,
+      "learning_rate": 9.768888505673976e-05,
+      "loss": 0.0143,
+      "step": 4270
+    },
+    {
+      "epoch": 0.24931554727092678,
+      "grad_norm": 0.3896014988422394,
+      "learning_rate": 9.767229311125162e-05,
+      "loss": 0.0136,
+      "step": 4280
+    },
+    {
+      "epoch": 0.24989806023184016,
+      "grad_norm": 0.224648579955101,
+      "learning_rate": 9.765564323943211e-05,
+      "loss": 0.0128,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2504805731927535,
+      "grad_norm": 0.2818295359611511,
+      "learning_rate": 9.763893546151244e-05,
+      "loss": 0.0142,
+      "step": 4300
+    },
+    {
+      "epoch": 0.2510630861536669,
+      "grad_norm": 0.2838699519634247,
+      "learning_rate": 9.762216979779412e-05,
+      "loss": 0.0107,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2516455991145803,
+      "grad_norm": 0.1902606189250946,
+      "learning_rate": 9.760534626864902e-05,
+      "loss": 0.0123,
+      "step": 4320
+    },
+    {
+      "epoch": 0.25222811207549367,
+      "grad_norm": 0.2964578866958618,
+      "learning_rate": 9.758846489451931e-05,
+      "loss": 0.0119,
+      "step": 4330
+    },
+    {
+      "epoch": 0.25281062503640706,
+      "grad_norm": 0.33075934648513794,
+      "learning_rate": 9.757152569591748e-05,
+      "loss": 0.0146,
+      "step": 4340
+    },
+    {
+      "epoch": 0.25339313799732044,
+      "grad_norm": 0.42020168900489807,
+      "learning_rate": 9.75545286934262e-05,
+      "loss": 0.0133,
+      "step": 4350
+    },
+    {
+      "epoch": 0.2539756509582338,
+      "grad_norm": 0.4201684296131134,
+      "learning_rate": 9.753747390769847e-05,
+      "loss": 0.0134,
+      "step": 4360
+    },
+    {
+      "epoch": 0.2545581639191472,
+      "grad_norm": 0.2482096552848816,
+      "learning_rate": 9.752036135945744e-05,
+      "loss": 0.0132,
+      "step": 4370
+    },
+    {
+      "epoch": 0.2551406768800606,
+      "grad_norm": 0.2185676395893097,
+      "learning_rate": 9.750319106949649e-05,
+      "loss": 0.0112,
+      "step": 4380
+    },
+    {
+      "epoch": 0.255723189840974,
+      "grad_norm": 0.4022974669933319,
+      "learning_rate": 9.748596305867913e-05,
+      "loss": 0.02,
+      "step": 4390
+    },
+    {
+      "epoch": 0.25630570280188736,
+      "grad_norm": 0.42810171842575073,
+      "learning_rate": 9.746867734793903e-05,
+      "loss": 0.0213,
+      "step": 4400
+    },
+    {
+      "epoch": 0.25688821576280074,
+      "grad_norm": 0.7219847440719604,
+      "learning_rate": 9.745133395827993e-05,
+      "loss": 0.0178,
+      "step": 4410
+    },
+    {
+      "epoch": 0.2574707287237141,
+      "grad_norm": 0.351438045501709,
+      "learning_rate": 9.743393291077572e-05,
+      "loss": 0.019,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2580532416846275,
+      "grad_norm": 0.4004956781864166,
+      "learning_rate": 9.741647422657028e-05,
+      "loss": 0.0179,
+      "step": 4430
+    },
+    {
+      "epoch": 0.25863575464554084,
+      "grad_norm": 0.5662891268730164,
+      "learning_rate": 9.739895792687758e-05,
+      "loss": 0.022,
+      "step": 4440
+    },
+    {
+      "epoch": 0.2592182676064542,
+      "grad_norm": 0.505552351474762,
+      "learning_rate": 9.738138403298157e-05,
+      "loss": 0.0155,
+      "step": 4450
+    },
+    {
+      "epoch": 0.2598007805673676,
+      "grad_norm": 0.4024660587310791,
+      "learning_rate": 9.736375256623619e-05,
+      "loss": 0.0168,
+      "step": 4460
+    },
+    {
+      "epoch": 0.260383293528281,
+      "grad_norm": 0.3369954228401184,
+      "learning_rate": 9.734606354806533e-05,
+      "loss": 0.0154,
+      "step": 4470
+    },
+    {
+      "epoch": 0.2609658064891944,
+      "grad_norm": 0.3174836337566376,
+      "learning_rate": 9.73283169999628e-05,
+      "loss": 0.0133,
+      "step": 4480
+    },
+    {
+      "epoch": 0.26154831945010776,
+      "grad_norm": 0.2838727533817291,
+      "learning_rate": 9.731051294349238e-05,
+      "loss": 0.0152,
+      "step": 4490
+    },
+    {
+      "epoch": 0.26213083241102114,
+      "grad_norm": 0.20824427902698517,
+      "learning_rate": 9.729265140028762e-05,
+      "loss": 0.0149,
+      "step": 4500
+    },
+    {
+      "epoch": 0.2627133453719345,
+      "grad_norm": 0.4184311032295227,
+      "learning_rate": 9.727473239205201e-05,
+      "loss": 0.0204,
+      "step": 4510
+    },
+    {
+      "epoch": 0.2632958583328479,
+      "grad_norm": 0.4343879818916321,
+      "learning_rate": 9.725675594055883e-05,
+      "loss": 0.0164,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2638783712937613,
+      "grad_norm": 0.25088533759117126,
+      "learning_rate": 9.723872206765116e-05,
+      "loss": 0.0157,
+      "step": 4530
+    },
+    {
+      "epoch": 0.2644608842546747,
+      "grad_norm": 0.28644782304763794,
+      "learning_rate": 9.722063079524185e-05,
+      "loss": 0.0133,
+      "step": 4540
+    },
+    {
+      "epoch": 0.26504339721558806,
+      "grad_norm": 0.4774155914783478,
+      "learning_rate": 9.720248214531351e-05,
+      "loss": 0.0143,
+      "step": 4550
+    },
+    {
+      "epoch": 0.26562591017650145,
+      "grad_norm": 0.36532267928123474,
+      "learning_rate": 9.718427613991848e-05,
+      "loss": 0.0134,
+      "step": 4560
+    },
+    {
+      "epoch": 0.26620842313741483,
+      "grad_norm": 0.2361028641462326,
+      "learning_rate": 9.716601280117873e-05,
+      "loss": 0.0133,
+      "step": 4570
+    },
+    {
+      "epoch": 0.2667909360983282,
+      "grad_norm": 0.35117876529693604,
+      "learning_rate": 9.714769215128596e-05,
+      "loss": 0.0145,
+      "step": 4580
+    },
+    {
+      "epoch": 0.26737344905924154,
+      "grad_norm": 0.28342658281326294,
+      "learning_rate": 9.712931421250152e-05,
+      "loss": 0.0134,
+      "step": 4590
+    },
+    {
+      "epoch": 0.2679559620201549,
+      "grad_norm": 0.2828042805194855,
+      "learning_rate": 9.711087900715627e-05,
+      "loss": 0.013,
+      "step": 4600
+    },
+    {
+      "epoch": 0.2685384749810683,
+      "grad_norm": 0.3468429148197174,
+      "learning_rate": 9.709238655765078e-05,
+      "loss": 0.0147,
+      "step": 4610
+    },
+    {
+      "epoch": 0.2691209879419817,
+      "grad_norm": 0.27148357033729553,
+      "learning_rate": 9.707383688645511e-05,
+      "loss": 0.0153,
+      "step": 4620
+    },
+    {
+      "epoch": 0.2697035009028951,
+      "grad_norm": 0.40684014558792114,
+      "learning_rate": 9.705523001610883e-05,
+      "loss": 0.0145,
+      "step": 4630
+    },
+    {
+      "epoch": 0.27028601386380846,
+      "grad_norm": 0.36176109313964844,
+      "learning_rate": 9.703656596922107e-05,
+      "loss": 0.0176,
+      "step": 4640
+    },
+    {
+      "epoch": 0.27086852682472184,
+      "grad_norm": 0.3847898840904236,
+      "learning_rate": 9.70178447684704e-05,
+      "loss": 0.0158,
+      "step": 4650
+    },
+    {
+      "epoch": 0.27145103978563523,
+      "grad_norm": 0.24188876152038574,
+      "learning_rate": 9.699906643660483e-05,
+      "loss": 0.0202,
+      "step": 4660
+    },
+    {
+      "epoch": 0.2720335527465486,
+      "grad_norm": 0.2264142483472824,
+      "learning_rate": 9.698023099644185e-05,
+      "loss": 0.0138,
+      "step": 4670
+    },
+    {
+      "epoch": 0.272616065707462,
+      "grad_norm": 0.24442389607429504,
+      "learning_rate": 9.696133847086823e-05,
+      "loss": 0.0107,
+      "step": 4680
+    },
+    {
+      "epoch": 0.2731985786683754,
+      "grad_norm": 0.22923371195793152,
+      "learning_rate": 9.694238888284022e-05,
+      "loss": 0.0142,
+      "step": 4690
+    },
+    {
+      "epoch": 0.27378109162928876,
+      "grad_norm": 0.269704133272171,
+      "learning_rate": 9.692338225538333e-05,
+      "loss": 0.0169,
+      "step": 4700
+    },
+    {
+      "epoch": 0.27436360459020215,
+      "grad_norm": 0.331922709941864,
+      "learning_rate": 9.690431861159241e-05,
+      "loss": 0.0138,
+      "step": 4710
+    },
+    {
+      "epoch": 0.27494611755111553,
+      "grad_norm": 0.3346977233886719,
+      "learning_rate": 9.688519797463161e-05,
+      "loss": 0.0153,
+      "step": 4720
+    },
+    {
+      "epoch": 0.2755286305120289,
+      "grad_norm": 0.3668387234210968,
+      "learning_rate": 9.686602036773426e-05,
+      "loss": 0.0144,
+      "step": 4730
+    },
+    {
+      "epoch": 0.2761111434729423,
+      "grad_norm": 0.43482881784439087,
+      "learning_rate": 9.684678581420302e-05,
+      "loss": 0.015,
+      "step": 4740
+    },
+    {
+      "epoch": 0.27669365643385563,
+      "grad_norm": 0.31008052825927734,
+      "learning_rate": 9.682749433740962e-05,
+      "loss": 0.0153,
+      "step": 4750
+    },
+    {
+      "epoch": 0.277276169394769,
+      "grad_norm": 0.310477614402771,
+      "learning_rate": 9.680814596079507e-05,
+      "loss": 0.0153,
+      "step": 4760
+    },
+    {
+      "epoch": 0.2778586823556824,
+      "grad_norm": 0.3365541398525238,
+      "learning_rate": 9.678874070786945e-05,
+      "loss": 0.0146,
+      "step": 4770
+    },
+    {
+      "epoch": 0.2784411953165958,
+      "grad_norm": 0.3239288926124573,
+      "learning_rate": 9.676927860221199e-05,
+      "loss": 0.0162,
+      "step": 4780
+    },
+    {
+      "epoch": 0.27902370827750916,
+      "grad_norm": 0.2856786251068115,
+      "learning_rate": 9.674975966747097e-05,
+      "loss": 0.0133,
+      "step": 4790
+    },
+    {
+      "epoch": 0.27960622123842255,
+      "grad_norm": 0.346770703792572,
+      "learning_rate": 9.673018392736374e-05,
+      "loss": 0.0149,
+      "step": 4800
+    },
+    {
+      "epoch": 0.28018873419933593,
+      "grad_norm": 0.34028011560440063,
+      "learning_rate": 9.671055140567667e-05,
+      "loss": 0.0133,
+      "step": 4810
+    },
+    {
+      "epoch": 0.2807712471602493,
+      "grad_norm": 0.32853594422340393,
+      "learning_rate": 9.669086212626511e-05,
+      "loss": 0.0104,
+      "step": 4820
+    },
+    {
+      "epoch": 0.2813537601211627,
+      "grad_norm": 0.3057480454444885,
+      "learning_rate": 9.667111611305341e-05,
+      "loss": 0.0151,
+      "step": 4830
+    },
+    {
+      "epoch": 0.2819362730820761,
+      "grad_norm": 0.314474880695343,
+      "learning_rate": 9.665131339003486e-05,
+      "loss": 0.0197,
+      "step": 4840
+    },
+    {
+      "epoch": 0.28251878604298947,
+      "grad_norm": 0.35622647404670715,
+      "learning_rate": 9.663145398127158e-05,
+      "loss": 0.014,
+      "step": 4850
+    },
+    {
+      "epoch": 0.28310129900390285,
+      "grad_norm": 0.5124862194061279,
+      "learning_rate": 9.661153791089467e-05,
+      "loss": 0.0138,
+      "step": 4860
+    },
+    {
+      "epoch": 0.28368381196481623,
+      "grad_norm": 0.4098750948905945,
+      "learning_rate": 9.659156520310402e-05,
+      "loss": 0.0136,
+      "step": 4870
+    },
+    {
+      "epoch": 0.2842663249257296,
+      "grad_norm": 0.24630773067474365,
+      "learning_rate": 9.657153588216835e-05,
+      "loss": 0.016,
+      "step": 4880
+    },
+    {
+      "epoch": 0.284848837886643,
+      "grad_norm": 0.45266249775886536,
+      "learning_rate": 9.655144997242516e-05,
+      "loss": 0.0131,
+      "step": 4890
+    },
+    {
+      "epoch": 0.28543135084755633,
+      "grad_norm": 0.3983039855957031,
+      "learning_rate": 9.653130749828075e-05,
+      "loss": 0.0165,
+      "step": 4900
+    },
+    {
+      "epoch": 0.2860138638084697,
+      "grad_norm": 0.35880064964294434,
+      "learning_rate": 9.65111084842101e-05,
+      "loss": 0.0151,
+      "step": 4910
+    },
+    {
+      "epoch": 0.2865963767693831,
+      "grad_norm": 0.25483396649360657,
+      "learning_rate": 9.649085295475695e-05,
+      "loss": 0.0126,
+      "step": 4920
+    },
+    {
+      "epoch": 0.2871788897302965,
+      "grad_norm": 0.3345501720905304,
+      "learning_rate": 9.647054093453365e-05,
+      "loss": 0.0148,
+      "step": 4930
+    },
+    {
+      "epoch": 0.28776140269120987,
+      "grad_norm": 0.44244930148124695,
+      "learning_rate": 9.645017244822123e-05,
+      "loss": 0.0136,
+      "step": 4940
+    },
+    {
+      "epoch": 0.28834391565212325,
+      "grad_norm": 0.38924461603164673,
+      "learning_rate": 9.642974752056931e-05,
+      "loss": 0.0148,
+      "step": 4950
+    },
+    {
+      "epoch": 0.28892642861303663,
+      "grad_norm": 0.2436865270137787,
+      "learning_rate": 9.640926617639613e-05,
+      "loss": 0.0121,
+      "step": 4960
+    },
+    {
+      "epoch": 0.28950894157395,
+      "grad_norm": 0.43023666739463806,
+      "learning_rate": 9.638872844058843e-05,
+      "loss": 0.0187,
+      "step": 4970
+    },
+    {
+      "epoch": 0.2900914545348634,
+      "grad_norm": 0.32516413927078247,
+      "learning_rate": 9.63681343381015e-05,
+      "loss": 0.0172,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2906739674957768,
+      "grad_norm": 0.2989577054977417,
+      "learning_rate": 9.634748389395914e-05,
+      "loss": 0.0123,
+      "step": 4990
+    },
+    {
+      "epoch": 0.29125648045669017,
+      "grad_norm": 0.27005958557128906,
+      "learning_rate": 9.632677713325353e-05,
+      "loss": 0.015,
+      "step": 5000
+    },
+    {
+      "epoch": 0.29183899341760355,
+      "grad_norm": 0.3099139630794525,
+      "learning_rate": 9.63060140811454e-05,
+      "loss": 0.0168,
+      "step": 5010
+    },
+    {
+      "epoch": 0.29242150637851694,
+      "grad_norm": 0.34511786699295044,
+      "learning_rate": 9.628519476286379e-05,
+      "loss": 0.0165,
+      "step": 5020
+    },
+    {
+      "epoch": 0.2930040193394303,
+      "grad_norm": 0.3462317883968353,
+      "learning_rate": 9.626431920370612e-05,
+      "loss": 0.0159,
+      "step": 5030
+    },
+    {
+      "epoch": 0.2935865323003437,
+      "grad_norm": 0.2514205276966095,
+      "learning_rate": 9.624338742903819e-05,
+      "loss": 0.0127,
+      "step": 5040
+    },
+    {
+      "epoch": 0.2941690452612571,
+      "grad_norm": 0.3226291239261627,
+      "learning_rate": 9.622239946429406e-05,
+      "loss": 0.0142,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2947515582221704,
+      "grad_norm": 0.330447256565094,
+      "learning_rate": 9.620135533497609e-05,
+      "loss": 0.0146,
+      "step": 5060
+    },
+    {
+      "epoch": 0.2953340711830838,
+      "grad_norm": 0.28347858786582947,
+      "learning_rate": 9.61802550666549e-05,
+      "loss": 0.0103,
+      "step": 5070
+    },
+    {
+      "epoch": 0.2959165841439972,
+      "grad_norm": 0.2465979903936386,
+      "learning_rate": 9.615909868496928e-05,
+      "loss": 0.0156,
+      "step": 5080
+    },
+    {
+      "epoch": 0.29649909710491057,
+      "grad_norm": 0.31589576601982117,
+      "learning_rate": 9.613788621562622e-05,
+      "loss": 0.015,
+      "step": 5090
+    },
+    {
+      "epoch": 0.29708161006582395,
+      "grad_norm": 0.4845690131187439,
+      "learning_rate": 9.611661768440093e-05,
+      "loss": 0.0163,
+      "step": 5100
+    },
+    {
+      "epoch": 0.29766412302673734,
+      "grad_norm": 0.33092549443244934,
+      "learning_rate": 9.609529311713661e-05,
+      "loss": 0.0145,
+      "step": 5110
+    },
+    {
+      "epoch": 0.2982466359876507,
+      "grad_norm": 0.29073378443717957,
+      "learning_rate": 9.607391253974466e-05,
+      "loss": 0.0173,
+      "step": 5120
+    },
+    {
+      "epoch": 0.2988291489485641,
+      "grad_norm": 0.38316112756729126,
+      "learning_rate": 9.605247597820448e-05,
+      "loss": 0.0116,
+      "step": 5130
+    },
+    {
+      "epoch": 0.2994116619094775,
+      "grad_norm": 0.2966541647911072,
+      "learning_rate": 9.603098345856354e-05,
+      "loss": 0.0132,
+      "step": 5140
+    },
+    {
+      "epoch": 0.2999941748703909,
+      "grad_norm": 0.28130510449409485,
+      "learning_rate": 9.600943500693725e-05,
+      "loss": 0.0106,
+      "step": 5150
+    },
+    {
+      "epoch": 0.30057668783130426,
+      "grad_norm": 0.2975674271583557,
+      "learning_rate": 9.598783064950902e-05,
+      "loss": 0.015,
+      "step": 5160
+    },
+    {
+      "epoch": 0.30115920079221764,
+      "grad_norm": 0.2338527888059616,
+      "learning_rate": 9.596617041253018e-05,
+      "loss": 0.0096,
+      "step": 5170
+    },
+    {
+      "epoch": 0.301741713753131,
+      "grad_norm": 0.3399932384490967,
+      "learning_rate": 9.594445432231996e-05,
+      "loss": 0.0115,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3023242267140444,
+      "grad_norm": 0.29272010922431946,
+      "learning_rate": 9.592268240526547e-05,
+      "loss": 0.0137,
+      "step": 5190
+    },
+    {
+      "epoch": 0.3029067396749578,
+      "grad_norm": 0.31858593225479126,
+      "learning_rate": 9.590085468782162e-05,
+      "loss": 0.0163,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3034892526358711,
+      "grad_norm": 0.38448017835617065,
+      "learning_rate": 9.587897119651116e-05,
+      "loss": 0.0128,
+      "step": 5210
+    },
+    {
+      "epoch": 0.3040717655967845,
+      "grad_norm": 0.22529268264770508,
+      "learning_rate": 9.585703195792459e-05,
+      "loss": 0.0129,
+      "step": 5220
+    },
+    {
+      "epoch": 0.3046542785576979,
+      "grad_norm": 0.3622327148914337,
+      "learning_rate": 9.583503699872016e-05,
+      "loss": 0.0131,
+      "step": 5230
+    },
+    {
+      "epoch": 0.3052367915186113,
+      "grad_norm": 0.2763410210609436,
+      "learning_rate": 9.581298634562381e-05,
+      "loss": 0.0116,
+      "step": 5240
+    },
+    {
+      "epoch": 0.30581930447952466,
+      "grad_norm": 0.2970474064350128,
+      "learning_rate": 9.579088002542917e-05,
+      "loss": 0.0126,
+      "step": 5250
+    },
+    {
+      "epoch": 0.30640181744043804,
+      "grad_norm": 0.2559505105018616,
+      "learning_rate": 9.57687180649975e-05,
+      "loss": 0.0146,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3069843304013514,
+      "grad_norm": 0.28003329038619995,
+      "learning_rate": 9.574650049125768e-05,
+      "loss": 0.0148,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3075668433622648,
+      "grad_norm": 0.26014652848243713,
+      "learning_rate": 9.572422733120614e-05,
+      "loss": 0.0149,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3081493563231782,
+      "grad_norm": 0.3492480516433716,
+      "learning_rate": 9.570189861190689e-05,
+      "loss": 0.0148,
+      "step": 5290
+    },
+    {
+      "epoch": 0.3087318692840916,
+      "grad_norm": 0.400206595659256,
+      "learning_rate": 9.56795143604914e-05,
+      "loss": 0.0142,
+      "step": 5300
+    },
+    {
+      "epoch": 0.30931438224500496,
+      "grad_norm": 0.34377798438072205,
+      "learning_rate": 9.565707460415869e-05,
+      "loss": 0.0136,
+      "step": 5310
+    },
+    {
+      "epoch": 0.30989689520591834,
+      "grad_norm": 0.34064170718193054,
+      "learning_rate": 9.563457937017515e-05,
+      "loss": 0.013,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3104794081668317,
+      "grad_norm": 0.22578756511211395,
+      "learning_rate": 9.56120286858746e-05,
+      "loss": 0.0122,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3110619211277451,
+      "grad_norm": 0.2612018585205078,
+      "learning_rate": 9.558942257865829e-05,
+      "loss": 0.0143,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3116444340886585,
+      "grad_norm": 0.26774755120277405,
+      "learning_rate": 9.556676107599472e-05,
+      "loss": 0.0125,
+      "step": 5350
+    },
+    {
+      "epoch": 0.3122269470495719,
+      "grad_norm": 0.21939046680927277,
+      "learning_rate": 9.554404420541978e-05,
+      "loss": 0.0125,
+      "step": 5360
+    },
+    {
+      "epoch": 0.3128094600104852,
+      "grad_norm": 0.34214887022972107,
+      "learning_rate": 9.55212719945366e-05,
+      "loss": 0.0147,
+      "step": 5370
+    },
+    {
+      "epoch": 0.3133919729713986,
+      "grad_norm": 0.27684247493743896,
+      "learning_rate": 9.549844447101559e-05,
+      "loss": 0.014,
+      "step": 5380
+    },
+    {
+      "epoch": 0.313974485932312,
+      "grad_norm": 0.31412631273269653,
+      "learning_rate": 9.547556166259433e-05,
+      "loss": 0.0136,
+      "step": 5390
+    },
+    {
+      "epoch": 0.31455699889322536,
+      "grad_norm": 0.2898443341255188,
+      "learning_rate": 9.545262359707756e-05,
+      "loss": 0.0131,
+      "step": 5400
+    },
+    {
+      "epoch": 0.31513951185413874,
+      "grad_norm": 0.24868787825107574,
+      "learning_rate": 9.542963030233724e-05,
+      "loss": 0.0142,
+      "step": 5410
+    },
+    {
+      "epoch": 0.3157220248150521,
+      "grad_norm": 0.31491348147392273,
+      "learning_rate": 9.540658180631237e-05,
+      "loss": 0.0133,
+      "step": 5420
+    },
+    {
+      "epoch": 0.3163045377759655,
+      "grad_norm": 0.20581558346748352,
+      "learning_rate": 9.538347813700904e-05,
+      "loss": 0.0158,
+      "step": 5430
+    },
+    {
+      "epoch": 0.3168870507368789,
+      "grad_norm": 0.2847931385040283,
+      "learning_rate": 9.536031932250036e-05,
+      "loss": 0.0129,
+      "step": 5440
+    },
+    {
+      "epoch": 0.3174695636977923,
+      "grad_norm": 0.33832216262817383,
+      "learning_rate": 9.533710539092653e-05,
+      "loss": 0.015,
+      "step": 5450
+    },
+    {
+      "epoch": 0.31805207665870566,
+      "grad_norm": 0.28036195039749146,
+      "learning_rate": 9.531383637049464e-05,
+      "loss": 0.011,
+      "step": 5460
+    },
+    {
+      "epoch": 0.31863458961961905,
+      "grad_norm": 0.30270060896873474,
+      "learning_rate": 9.529051228947875e-05,
+      "loss": 0.0124,
+      "step": 5470
+    },
+    {
+      "epoch": 0.31921710258053243,
+      "grad_norm": 0.34320932626724243,
+      "learning_rate": 9.52671331762198e-05,
+      "loss": 0.0129,
+      "step": 5480
+    },
+    {
+      "epoch": 0.3197996155414458,
+      "grad_norm": 0.40417709946632385,
+      "learning_rate": 9.524369905912565e-05,
+      "loss": 0.0109,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3203821285023592,
+      "grad_norm": 0.28732940554618835,
+      "learning_rate": 9.522020996667092e-05,
+      "loss": 0.0107,
+      "step": 5500
+    },
+    {
+      "epoch": 0.3209646414632726,
+      "grad_norm": 0.22060438990592957,
+      "learning_rate": 9.519666592739709e-05,
+      "loss": 0.0113,
+      "step": 5510
+    },
+    {
+      "epoch": 0.32154715442418597,
+      "grad_norm": 0.23726588487625122,
+      "learning_rate": 9.517306696991241e-05,
+      "loss": 0.0102,
+      "step": 5520
+    },
+    {
+      "epoch": 0.3221296673850993,
+      "grad_norm": 0.34877848625183105,
+      "learning_rate": 9.51494131228918e-05,
+      "loss": 0.0201,
+      "step": 5530
+    },
+    {
+      "epoch": 0.3227121803460127,
+      "grad_norm": 0.20062492787837982,
+      "learning_rate": 9.512570441507695e-05,
+      "loss": 0.014,
+      "step": 5540
+    },
+    {
+      "epoch": 0.32329469330692606,
+      "grad_norm": 0.3137679100036621,
+      "learning_rate": 9.510194087527615e-05,
+      "loss": 0.011,
+      "step": 5550
+    },
+    {
+      "epoch": 0.32387720626783945,
+      "grad_norm": 0.4614037275314331,
+      "learning_rate": 9.507812253236435e-05,
+      "loss": 0.0147,
+      "step": 5560
+    },
+    {
+      "epoch": 0.32445971922875283,
+      "grad_norm": 0.2742924392223358,
+      "learning_rate": 9.505424941528309e-05,
+      "loss": 0.0129,
+      "step": 5570
+    },
+    {
+      "epoch": 0.3250422321896662,
+      "grad_norm": 0.2699301242828369,
+      "learning_rate": 9.503032155304046e-05,
+      "loss": 0.013,
+      "step": 5580
+    },
+    {
+      "epoch": 0.3256247451505796,
+      "grad_norm": 0.4203640818595886,
+      "learning_rate": 9.500633897471106e-05,
+      "loss": 0.0138,
+      "step": 5590
+    },
+    {
+      "epoch": 0.326207258111493,
+      "grad_norm": 0.37082532048225403,
+      "learning_rate": 9.498230170943596e-05,
+      "loss": 0.0141,
+      "step": 5600
+    },
+    {
+      "epoch": 0.32678977107240637,
+      "grad_norm": 0.4199288785457611,
+      "learning_rate": 9.495820978642275e-05,
+      "loss": 0.0169,
+      "step": 5610
+    },
+    {
+      "epoch": 0.32737228403331975,
+      "grad_norm": 0.46307772397994995,
+      "learning_rate": 9.493406323494535e-05,
+      "loss": 0.0135,
+      "step": 5620
+    },
+    {
+      "epoch": 0.32795479699423313,
+      "grad_norm": 0.3684145510196686,
+      "learning_rate": 9.490986208434413e-05,
+      "loss": 0.0127,
+      "step": 5630
+    },
+    {
+      "epoch": 0.3285373099551465,
+      "grad_norm": 0.3294019401073456,
+      "learning_rate": 9.488560636402577e-05,
+      "loss": 0.0167,
+      "step": 5640
+    },
+    {
+      "epoch": 0.3291198229160599,
+      "grad_norm": 0.3000147342681885,
+      "learning_rate": 9.486129610346321e-05,
+      "loss": 0.0119,
+      "step": 5650
+    },
+    {
+      "epoch": 0.3297023358769733,
+      "grad_norm": 0.23165956139564514,
+      "learning_rate": 9.483693133219576e-05,
+      "loss": 0.0113,
+      "step": 5660
+    },
+    {
+      "epoch": 0.33028484883788667,
+      "grad_norm": 0.23021915555000305,
+      "learning_rate": 9.481251207982888e-05,
+      "loss": 0.0125,
+      "step": 5670
+    },
+    {
+      "epoch": 0.3308673617988,
+      "grad_norm": 0.2385621815919876,
+      "learning_rate": 9.47880383760343e-05,
+      "loss": 0.0139,
+      "step": 5680
+    },
+    {
+      "epoch": 0.3314498747597134,
+      "grad_norm": 0.2613489031791687,
+      "learning_rate": 9.476351025054983e-05,
+      "loss": 0.0164,
+      "step": 5690
+    },
+    {
+      "epoch": 0.33203238772062676,
+      "grad_norm": 0.23035407066345215,
+      "learning_rate": 9.473892773317952e-05,
+      "loss": 0.0114,
+      "step": 5700
+    },
+    {
+      "epoch": 0.33261490068154015,
+      "grad_norm": 0.2664507329463959,
+      "learning_rate": 9.471429085379338e-05,
+      "loss": 0.0183,
+      "step": 5710
+    },
+    {
+      "epoch": 0.33319741364245353,
+      "grad_norm": 0.32637298107147217,
+      "learning_rate": 9.468959964232757e-05,
+      "loss": 0.018,
+      "step": 5720
+    },
+    {
+      "epoch": 0.3337799266033669,
+      "grad_norm": 0.20291414856910706,
+      "learning_rate": 9.466485412878425e-05,
+      "loss": 0.0201,
+      "step": 5730
+    },
+    {
+      "epoch": 0.3343624395642803,
+      "grad_norm": 0.4235072433948517,
+      "learning_rate": 9.464005434323154e-05,
+      "loss": 0.0132,
+      "step": 5740
+    },
+    {
+      "epoch": 0.3349449525251937,
+      "grad_norm": 0.3145645260810852,
+      "learning_rate": 9.461520031580352e-05,
+      "loss": 0.013,
+      "step": 5750
+    },
+    {
+      "epoch": 0.33552746548610707,
+      "grad_norm": 0.2606288492679596,
+      "learning_rate": 9.459029207670019e-05,
+      "loss": 0.0136,
+      "step": 5760
+    },
+    {
+      "epoch": 0.33610997844702045,
+      "grad_norm": 0.3538146913051605,
+      "learning_rate": 9.456532965618737e-05,
+      "loss": 0.0135,
+      "step": 5770
+    },
+    {
+      "epoch": 0.33669249140793384,
+      "grad_norm": 0.3229866921901703,
+      "learning_rate": 9.454031308459681e-05,
+      "loss": 0.0134,
+      "step": 5780
+    },
+    {
+      "epoch": 0.3372750043688472,
+      "grad_norm": 0.28633013367652893,
+      "learning_rate": 9.451524239232595e-05,
+      "loss": 0.0148,
+      "step": 5790
+    },
+    {
+      "epoch": 0.3378575173297606,
+      "grad_norm": 0.3700575828552246,
+      "learning_rate": 9.449011760983809e-05,
+      "loss": 0.0139,
+      "step": 5800
+    },
+    {
+      "epoch": 0.338440030290674,
+      "grad_norm": 0.17904630303382874,
+      "learning_rate": 9.446493876766218e-05,
+      "loss": 0.0108,
+      "step": 5810
+    },
+    {
+      "epoch": 0.33902254325158737,
+      "grad_norm": 0.4766163229942322,
+      "learning_rate": 9.44397058963929e-05,
+      "loss": 0.0118,
+      "step": 5820
+    },
+    {
+      "epoch": 0.33960505621250076,
+      "grad_norm": 0.45538008213043213,
+      "learning_rate": 9.441441902669056e-05,
+      "loss": 0.0177,
+      "step": 5830
+    },
+    {
+      "epoch": 0.3401875691734141,
+      "grad_norm": 0.4499160051345825,
+      "learning_rate": 9.43890781892811e-05,
+      "loss": 0.0163,
+      "step": 5840
+    },
+    {
+      "epoch": 0.34077008213432747,
+      "grad_norm": 0.2667197287082672,
+      "learning_rate": 9.436368341495603e-05,
+      "loss": 0.0201,
+      "step": 5850
+    },
+    {
+      "epoch": 0.34135259509524085,
+      "grad_norm": 0.2519628405570984,
+      "learning_rate": 9.43382347345724e-05,
+      "loss": 0.0139,
+      "step": 5860
+    },
+    {
+      "epoch": 0.34193510805615424,
+      "grad_norm": 0.27849137783050537,
+      "learning_rate": 9.431273217905273e-05,
+      "loss": 0.0125,
+      "step": 5870
+    },
+    {
+      "epoch": 0.3425176210170676,
+      "grad_norm": 0.3382633626461029,
+      "learning_rate": 9.428717577938504e-05,
+      "loss": 0.0184,
+      "step": 5880
+    },
+    {
+      "epoch": 0.343100133977981,
+      "grad_norm": 0.32172083854675293,
+      "learning_rate": 9.426156556662276e-05,
+      "loss": 0.0118,
+      "step": 5890
+    },
+    {
+      "epoch": 0.3436826469388944,
+      "grad_norm": 0.36967912316322327,
+      "learning_rate": 9.423590157188474e-05,
+      "loss": 0.0156,
+      "step": 5900
+    },
+    {
+      "epoch": 0.34426515989980777,
+      "grad_norm": 0.33469265699386597,
+      "learning_rate": 9.421018382635513e-05,
+      "loss": 0.014,
+      "step": 5910
+    },
+    {
+      "epoch": 0.34484767286072115,
+      "grad_norm": 0.30567115545272827,
+      "learning_rate": 9.418441236128343e-05,
+      "loss": 0.0149,
+      "step": 5920
+    },
+    {
+      "epoch": 0.34543018582163454,
+      "grad_norm": 0.19063973426818848,
+      "learning_rate": 9.41585872079844e-05,
+      "loss": 0.0133,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3460126987825479,
+      "grad_norm": 0.4304380416870117,
+      "learning_rate": 9.413270839783802e-05,
+      "loss": 0.0125,
+      "step": 5940
+    },
+    {
+      "epoch": 0.3465952117434613,
+      "grad_norm": 0.2692255973815918,
+      "learning_rate": 9.41067759622895e-05,
+      "loss": 0.0128,
+      "step": 5950
+    },
+    {
+      "epoch": 0.3471777247043747,
+      "grad_norm": 0.2566480338573456,
+      "learning_rate": 9.408078993284917e-05,
+      "loss": 0.0128,
+      "step": 5960
+    },
+    {
+      "epoch": 0.3477602376652881,
+      "grad_norm": 0.30677348375320435,
+      "learning_rate": 9.405475034109254e-05,
+      "loss": 0.0111,
+      "step": 5970
+    },
+    {
+      "epoch": 0.34834275062620146,
+      "grad_norm": 0.31435564160346985,
+      "learning_rate": 9.402865721866015e-05,
+      "loss": 0.0129,
+      "step": 5980
+    },
+    {
+      "epoch": 0.3489252635871148,
+      "grad_norm": 0.3343210816383362,
+      "learning_rate": 9.400251059725762e-05,
+      "loss": 0.0104,
+      "step": 5990
+    },
+    {
+      "epoch": 0.34950777654802817,
+      "grad_norm": 0.3017050623893738,
+      "learning_rate": 9.397631050865554e-05,
+      "loss": 0.0133,
+      "step": 6000
+    },
+    {
+      "epoch": 0.35009028950894155,
+      "grad_norm": 0.20941412448883057,
+      "learning_rate": 9.395005698468949e-05,
+      "loss": 0.0098,
+      "step": 6010
+    },
+    {
+      "epoch": 0.35067280246985494,
+      "grad_norm": 0.29110249876976013,
+      "learning_rate": 9.392375005726e-05,
+      "loss": 0.0119,
+      "step": 6020
+    },
+    {
+      "epoch": 0.3512553154307683,
+      "grad_norm": 0.38054490089416504,
+      "learning_rate": 9.389738975833242e-05,
+      "loss": 0.01,
+      "step": 6030
+    },
+    {
+      "epoch": 0.3518378283916817,
+      "grad_norm": 0.24182026088237762,
+      "learning_rate": 9.387097611993707e-05,
+      "loss": 0.0146,
+      "step": 6040
+    },
+    {
+      "epoch": 0.3524203413525951,
+      "grad_norm": 0.20831067860126495,
+      "learning_rate": 9.384450917416894e-05,
+      "loss": 0.0107,
+      "step": 6050
+    },
+    {
+      "epoch": 0.3530028543135085,
+      "grad_norm": 0.17184917628765106,
+      "learning_rate": 9.381798895318792e-05,
+      "loss": 0.0117,
+      "step": 6060
+    },
+    {
+      "epoch": 0.35358536727442186,
+      "grad_norm": 0.30182504653930664,
+      "learning_rate": 9.379141548921853e-05,
+      "loss": 0.0108,
+      "step": 6070
+    },
+    {
+      "epoch": 0.35416788023533524,
+      "grad_norm": 0.2855520248413086,
+      "learning_rate": 9.376478881455009e-05,
+      "loss": 0.0151,
+      "step": 6080
+    },
+    {
+      "epoch": 0.3547503931962486,
+      "grad_norm": 0.29148533940315247,
+      "learning_rate": 9.373810896153647e-05,
+      "loss": 0.013,
+      "step": 6090
+    },
+    {
+      "epoch": 0.355332906157162,
+      "grad_norm": 0.3114771246910095,
+      "learning_rate": 9.371137596259623e-05,
+      "loss": 0.0101,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3559154191180754,
+      "grad_norm": 0.2501264214515686,
+      "learning_rate": 9.368458985021248e-05,
+      "loss": 0.0101,
+      "step": 6110
+    },
+    {
+      "epoch": 0.3564979320789888,
+      "grad_norm": 0.2828197479248047,
+      "learning_rate": 9.365775065693287e-05,
+      "loss": 0.0147,
+      "step": 6120
+    },
+    {
+      "epoch": 0.35708044503990216,
+      "grad_norm": 0.26596835255622864,
+      "learning_rate": 9.363085841536957e-05,
+      "loss": 0.0135,
+      "step": 6130
+    },
+    {
+      "epoch": 0.35766295800081555,
+      "grad_norm": 0.20203420519828796,
+      "learning_rate": 9.360391315819917e-05,
+      "loss": 0.0101,
+      "step": 6140
+    },
+    {
+      "epoch": 0.3582454709617289,
+      "grad_norm": 0.30759403109550476,
+      "learning_rate": 9.357691491816269e-05,
+      "loss": 0.0102,
+      "step": 6150
+    },
+    {
+      "epoch": 0.35882798392264226,
+      "grad_norm": 0.22800108790397644,
+      "learning_rate": 9.354986372806557e-05,
+      "loss": 0.0137,
+      "step": 6160
+    },
+    {
+      "epoch": 0.35941049688355564,
+      "grad_norm": 0.21857386827468872,
+      "learning_rate": 9.352275962077752e-05,
+      "loss": 0.0134,
+      "step": 6170
+    },
+    {
+      "epoch": 0.359993009844469,
+      "grad_norm": 0.40877413749694824,
+      "learning_rate": 9.349560262923262e-05,
+      "loss": 0.0144,
+      "step": 6180
+    },
+    {
+      "epoch": 0.3605755228053824,
+      "grad_norm": 0.45766687393188477,
+      "learning_rate": 9.346839278642913e-05,
+      "loss": 0.0147,
+      "step": 6190
+    },
+    {
+      "epoch": 0.3611580357662958,
+      "grad_norm": 0.31452977657318115,
+      "learning_rate": 9.344113012542963e-05,
+      "loss": 0.0129,
+      "step": 6200
+    },
+    {
+      "epoch": 0.3617405487272092,
+      "grad_norm": 0.2539997696876526,
+      "learning_rate": 9.341381467936079e-05,
+      "loss": 0.0115,
+      "step": 6210
+    },
+    {
+      "epoch": 0.36232306168812256,
+      "grad_norm": 0.26139014959335327,
+      "learning_rate": 9.338644648141346e-05,
+      "loss": 0.0116,
+      "step": 6220
+    },
+    {
+      "epoch": 0.36290557464903594,
+      "grad_norm": 0.1639491766691208,
+      "learning_rate": 9.335902556484257e-05,
+      "loss": 0.0134,
+      "step": 6230
+    },
+    {
+      "epoch": 0.36348808760994933,
+      "grad_norm": 0.2698086202144623,
+      "learning_rate": 9.333155196296712e-05,
+      "loss": 0.0119,
+      "step": 6240
+    },
+    {
+      "epoch": 0.3640706005708627,
+      "grad_norm": 0.38294240832328796,
+      "learning_rate": 9.330402570917016e-05,
+      "loss": 0.0126,
+      "step": 6250
+    },
+    {
+      "epoch": 0.3646531135317761,
+      "grad_norm": 0.36223292350769043,
+      "learning_rate": 9.327644683689865e-05,
+      "loss": 0.0138,
+      "step": 6260
+    },
+    {
+      "epoch": 0.3652356264926895,
+      "grad_norm": 0.30007612705230713,
+      "learning_rate": 9.324881537966354e-05,
+      "loss": 0.0124,
+      "step": 6270
+    },
+    {
+      "epoch": 0.36581813945360286,
+      "grad_norm": 0.23643670976161957,
+      "learning_rate": 9.322113137103964e-05,
+      "loss": 0.0119,
+      "step": 6280
+    },
+    {
+      "epoch": 0.36640065241451625,
+      "grad_norm": 0.24124610424041748,
+      "learning_rate": 9.319339484466564e-05,
+      "loss": 0.0144,
+      "step": 6290
+    },
+    {
+      "epoch": 0.3669831653754296,
+      "grad_norm": 0.2288600355386734,
+      "learning_rate": 9.316560583424403e-05,
+      "loss": 0.0109,
+      "step": 6300
+    },
+    {
+      "epoch": 0.36756567833634296,
+      "grad_norm": 0.26642534136772156,
+      "learning_rate": 9.313776437354108e-05,
+      "loss": 0.0138,
+      "step": 6310
+    },
+    {
+      "epoch": 0.36814819129725634,
+      "grad_norm": 0.26386332511901855,
+      "learning_rate": 9.31098704963868e-05,
+      "loss": 0.0106,
+      "step": 6320
+    },
+    {
+      "epoch": 0.36873070425816973,
+      "grad_norm": 0.29608970880508423,
+      "learning_rate": 9.308192423667485e-05,
+      "loss": 0.0101,
+      "step": 6330
+    },
+    {
+      "epoch": 0.3693132172190831,
+      "grad_norm": 0.3042888939380646,
+      "learning_rate": 9.305392562836262e-05,
+      "loss": 0.0114,
+      "step": 6340
+    },
+    {
+      "epoch": 0.3698957301799965,
+      "grad_norm": 0.30546969175338745,
+      "learning_rate": 9.3025874705471e-05,
+      "loss": 0.0114,
+      "step": 6350
+    },
+    {
+      "epoch": 0.3704782431409099,
+      "grad_norm": 0.3741450607776642,
+      "learning_rate": 9.299777150208457e-05,
+      "loss": 0.0133,
+      "step": 6360
+    },
+    {
+      "epoch": 0.37106075610182326,
+      "grad_norm": 0.3162546157836914,
+      "learning_rate": 9.296961605235133e-05,
+      "loss": 0.0138,
+      "step": 6370
+    },
+    {
+      "epoch": 0.37164326906273665,
+      "grad_norm": 0.3369394838809967,
+      "learning_rate": 9.29414083904828e-05,
+      "loss": 0.0128,
+      "step": 6380
+    },
+    {
+      "epoch": 0.37222578202365003,
+      "grad_norm": 0.2751036286354065,
+      "learning_rate": 9.291314855075397e-05,
+      "loss": 0.0126,
+      "step": 6390
+    },
+    {
+      "epoch": 0.3728082949845634,
+      "grad_norm": 0.19377347826957703,
+      "learning_rate": 9.288483656750322e-05,
+      "loss": 0.0118,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3733908079454768,
+      "grad_norm": 0.31796208024024963,
+      "learning_rate": 9.285647247513225e-05,
+      "loss": 0.0115,
+      "step": 6410
+    },
+    {
+      "epoch": 0.3739733209063902,
+      "grad_norm": 0.1682463437318802,
+      "learning_rate": 9.282805630810614e-05,
+      "loss": 0.0124,
+      "step": 6420
+    },
+    {
+      "epoch": 0.37455583386730357,
+      "grad_norm": 0.3048756718635559,
+      "learning_rate": 9.279958810095317e-05,
+      "loss": 0.0088,
+      "step": 6430
+    },
+    {
+      "epoch": 0.37513834682821695,
+      "grad_norm": 0.2317032665014267,
+      "learning_rate": 9.277106788826494e-05,
+      "loss": 0.0123,
+      "step": 6440
+    },
+    {
+      "epoch": 0.37572085978913033,
+      "grad_norm": 0.19936993718147278,
+      "learning_rate": 9.274249570469618e-05,
+      "loss": 0.0089,
+      "step": 6450
+    },
+    {
+      "epoch": 0.37630337275004366,
+      "grad_norm": 0.23571209609508514,
+      "learning_rate": 9.271387158496476e-05,
+      "loss": 0.0087,
+      "step": 6460
+    },
+    {
+      "epoch": 0.37688588571095705,
+      "grad_norm": 0.30482497811317444,
+      "learning_rate": 9.268519556385173e-05,
+      "loss": 0.0117,
+      "step": 6470
+    },
+    {
+      "epoch": 0.37746839867187043,
+      "grad_norm": 0.3876536190509796,
+      "learning_rate": 9.265646767620113e-05,
+      "loss": 0.0101,
+      "step": 6480
+    },
+    {
+      "epoch": 0.3780509116327838,
+      "grad_norm": 0.37219473719596863,
+      "learning_rate": 9.262768795692006e-05,
+      "loss": 0.0119,
+      "step": 6490
+    },
+    {
+      "epoch": 0.3786334245936972,
+      "grad_norm": 0.35567301511764526,
+      "learning_rate": 9.259885644097861e-05,
+      "loss": 0.0111,
+      "step": 6500
+    },
+    {
+      "epoch": 0.3792159375546106,
+      "grad_norm": 0.46365585923194885,
+      "learning_rate": 9.256997316340976e-05,
+      "loss": 0.0106,
+      "step": 6510
+    },
+    {
+      "epoch": 0.37979845051552397,
+      "grad_norm": 0.3406113088130951,
+      "learning_rate": 9.254103815930943e-05,
+      "loss": 0.0145,
+      "step": 6520
+    },
+    {
+      "epoch": 0.38038096347643735,
+      "grad_norm": 0.30493253469467163,
+      "learning_rate": 9.251205146383636e-05,
+      "loss": 0.014,
+      "step": 6530
+    },
+    {
+      "epoch": 0.38096347643735073,
+      "grad_norm": 0.21430696547031403,
+      "learning_rate": 9.248301311221216e-05,
+      "loss": 0.0121,
+      "step": 6540
+    },
+    {
+      "epoch": 0.3815459893982641,
+      "grad_norm": 0.3540270924568176,
+      "learning_rate": 9.245392313972115e-05,
+      "loss": 0.0128,
+      "step": 6550
+    },
+    {
+      "epoch": 0.3821285023591775,
+      "grad_norm": 0.2746662199497223,
+      "learning_rate": 9.242478158171037e-05,
+      "loss": 0.0132,
+      "step": 6560
+    },
+    {
+      "epoch": 0.3827110153200909,
+      "grad_norm": 0.3532452881336212,
+      "learning_rate": 9.239558847358957e-05,
+      "loss": 0.0148,
+      "step": 6570
+    },
+    {
+      "epoch": 0.38329352828100427,
+      "grad_norm": 0.2984524667263031,
+      "learning_rate": 9.236634385083114e-05,
+      "loss": 0.0162,
+      "step": 6580
+    },
+    {
+      "epoch": 0.38387604124191765,
+      "grad_norm": 0.3702338933944702,
+      "learning_rate": 9.233704774897005e-05,
+      "loss": 0.0126,
+      "step": 6590
+    },
+    {
+      "epoch": 0.38445855420283104,
+      "grad_norm": 0.38487353920936584,
+      "learning_rate": 9.230770020360382e-05,
+      "loss": 0.0136,
+      "step": 6600
+    },
+    {
+      "epoch": 0.38504106716374437,
+      "grad_norm": 0.43931037187576294,
+      "learning_rate": 9.227830125039247e-05,
+      "loss": 0.0134,
+      "step": 6610
+    },
+    {
+      "epoch": 0.38562358012465775,
+      "grad_norm": 0.20062223076820374,
+      "learning_rate": 9.224885092505853e-05,
+      "loss": 0.0122,
+      "step": 6620
+    },
+    {
+      "epoch": 0.38620609308557113,
+      "grad_norm": 0.40552452206611633,
+      "learning_rate": 9.221934926338691e-05,
+      "loss": 0.0165,
+      "step": 6630
+    },
+    {
+      "epoch": 0.3867886060464845,
+      "grad_norm": 0.41163545846939087,
+      "learning_rate": 9.21897963012249e-05,
+      "loss": 0.0118,
+      "step": 6640
+    },
+    {
+      "epoch": 0.3873711190073979,
+      "grad_norm": 0.22466735541820526,
+      "learning_rate": 9.216019207448217e-05,
+      "loss": 0.0106,
+      "step": 6650
+    },
+    {
+      "epoch": 0.3879536319683113,
+      "grad_norm": 0.23277828097343445,
+      "learning_rate": 9.21305366191306e-05,
+      "loss": 0.0109,
+      "step": 6660
+    },
+    {
+      "epoch": 0.38853614492922467,
+      "grad_norm": 0.33124229311943054,
+      "learning_rate": 9.21008299712044e-05,
+      "loss": 0.0117,
+      "step": 6670
+    },
+    {
+      "epoch": 0.38911865789013805,
+      "grad_norm": 0.2248697578907013,
+      "learning_rate": 9.207107216679993e-05,
+      "loss": 0.0125,
+      "step": 6680
+    },
+    {
+      "epoch": 0.38970117085105144,
+      "grad_norm": 0.31560787558555603,
+      "learning_rate": 9.204126324207574e-05,
+      "loss": 0.0123,
+      "step": 6690
+    },
+    {
+      "epoch": 0.3902836838119648,
+      "grad_norm": 0.37756285071372986,
+      "learning_rate": 9.201140323325248e-05,
+      "loss": 0.0137,
+      "step": 6700
+    },
+    {
+      "epoch": 0.3908661967728782,
+      "grad_norm": 0.3497389554977417,
+      "learning_rate": 9.198149217661287e-05,
+      "loss": 0.0128,
+      "step": 6710
+    },
+    {
+      "epoch": 0.3914487097337916,
+      "grad_norm": 0.3866865634918213,
+      "learning_rate": 9.195153010850167e-05,
+      "loss": 0.0131,
+      "step": 6720
+    },
+    {
+      "epoch": 0.392031222694705,
+      "grad_norm": 0.5670121908187866,
+      "learning_rate": 9.192151706532562e-05,
+      "loss": 0.0152,
+      "step": 6730
+    },
+    {
+      "epoch": 0.39261373565561836,
+      "grad_norm": 0.5314503312110901,
+      "learning_rate": 9.189145308355337e-05,
+      "loss": 0.0169,
+      "step": 6740
+    },
+    {
+      "epoch": 0.39319624861653174,
+      "grad_norm": 0.4168036878108978,
+      "learning_rate": 9.186133819971555e-05,
+      "loss": 0.013,
+      "step": 6750
+    },
+    {
+      "epoch": 0.3937787615774451,
+      "grad_norm": 0.36590367555618286,
+      "learning_rate": 9.183117245040455e-05,
+      "loss": 0.0121,
+      "step": 6760
+    },
+    {
+      "epoch": 0.39436127453835845,
+      "grad_norm": 0.29301807284355164,
+      "learning_rate": 9.180095587227458e-05,
+      "loss": 0.0108,
+      "step": 6770
+    },
+    {
+      "epoch": 0.39494378749927184,
+      "grad_norm": 0.45780086517333984,
+      "learning_rate": 9.177068850204167e-05,
+      "loss": 0.0124,
+      "step": 6780
+    },
+    {
+      "epoch": 0.3955263004601852,
+      "grad_norm": 0.2961581349372864,
+      "learning_rate": 9.174037037648351e-05,
+      "loss": 0.0117,
+      "step": 6790
+    },
+    {
+      "epoch": 0.3961088134210986,
+      "grad_norm": 0.27854636311531067,
+      "learning_rate": 9.171000153243947e-05,
+      "loss": 0.016,
+      "step": 6800
+    },
+    {
+      "epoch": 0.396691326382012,
+      "grad_norm": 0.2911928594112396,
+      "learning_rate": 9.167958200681057e-05,
+      "loss": 0.0095,
+      "step": 6810
+    },
+    {
+      "epoch": 0.3972738393429254,
+      "grad_norm": 0.28199344873428345,
+      "learning_rate": 9.164911183655943e-05,
+      "loss": 0.0112,
+      "step": 6820
+    },
+    {
+      "epoch": 0.39785635230383876,
+      "grad_norm": 0.23138020932674408,
+      "learning_rate": 9.161859105871013e-05,
+      "loss": 0.0099,
+      "step": 6830
+    },
+    {
+      "epoch": 0.39843886526475214,
+      "grad_norm": 0.2517688572406769,
+      "learning_rate": 9.158801971034832e-05,
+      "loss": 0.0112,
+      "step": 6840
+    },
+    {
+      "epoch": 0.3990213782256655,
+      "grad_norm": 0.2718874514102936,
+      "learning_rate": 9.155739782862107e-05,
+      "loss": 0.0084,
+      "step": 6850
+    },
+    {
+      "epoch": 0.3996038911865789,
+      "grad_norm": 0.42391732335090637,
+      "learning_rate": 9.152672545073687e-05,
+      "loss": 0.0142,
+      "step": 6860
+    },
+    {
+      "epoch": 0.4001864041474923,
+      "grad_norm": 0.23340462148189545,
+      "learning_rate": 9.149600261396551e-05,
+      "loss": 0.0129,
+      "step": 6870
+    },
+    {
+      "epoch": 0.4007689171084057,
+      "grad_norm": 0.24849651753902435,
+      "learning_rate": 9.146522935563817e-05,
+      "loss": 0.0165,
+      "step": 6880
+    },
+    {
+      "epoch": 0.40135143006931906,
+      "grad_norm": 0.29082268476486206,
+      "learning_rate": 9.143440571314723e-05,
+      "loss": 0.013,
+      "step": 6890
+    },
+    {
+      "epoch": 0.40193394303023244,
+      "grad_norm": 0.3280594050884247,
+      "learning_rate": 9.140353172394636e-05,
+      "loss": 0.0112,
+      "step": 6900
+    },
+    {
+      "epoch": 0.4025164559911458,
+      "grad_norm": 0.44642722606658936,
+      "learning_rate": 9.137260742555032e-05,
+      "loss": 0.0145,
+      "step": 6910
+    },
+    {
+      "epoch": 0.40309896895205916,
+      "grad_norm": 0.2751993238925934,
+      "learning_rate": 9.13416328555351e-05,
+      "loss": 0.0174,
+      "step": 6920
+    },
+    {
+      "epoch": 0.40368148191297254,
+      "grad_norm": 0.2839924693107605,
+      "learning_rate": 9.131060805153769e-05,
+      "loss": 0.011,
+      "step": 6930
+    },
+    {
+      "epoch": 0.4042639948738859,
+      "grad_norm": 0.26174360513687134,
+      "learning_rate": 9.127953305125618e-05,
+      "loss": 0.011,
+      "step": 6940
+    },
+    {
+      "epoch": 0.4048465078347993,
+      "grad_norm": 0.22545038163661957,
+      "learning_rate": 9.124840789244958e-05,
+      "loss": 0.0166,
+      "step": 6950
+    },
+    {
+      "epoch": 0.4054290207957127,
+      "grad_norm": 0.22442230582237244,
+      "learning_rate": 9.121723261293794e-05,
+      "loss": 0.0121,
+      "step": 6960
+    },
+    {
+      "epoch": 0.4060115337566261,
+      "grad_norm": 0.3519812524318695,
+      "learning_rate": 9.118600725060214e-05,
+      "loss": 0.0138,
+      "step": 6970
+    },
+    {
+      "epoch": 0.40659404671753946,
+      "grad_norm": 0.2668507397174835,
+      "learning_rate": 9.115473184338392e-05,
+      "loss": 0.0107,
+      "step": 6980
+    },
+    {
+      "epoch": 0.40717655967845284,
+      "grad_norm": 0.284104585647583,
+      "learning_rate": 9.11234064292859e-05,
+      "loss": 0.0135,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4077590726393662,
+      "grad_norm": 0.3436839282512665,
+      "learning_rate": 9.109203104637137e-05,
+      "loss": 0.0126,
+      "step": 7000
+    },
+    {
+      "epoch": 0.4083415856002796,
+      "grad_norm": 0.2720262408256531,
+      "learning_rate": 9.10606057327644e-05,
+      "loss": 0.0122,
+      "step": 7010
+    },
+    {
+      "epoch": 0.408924098561193,
+      "grad_norm": 0.21050935983657837,
+      "learning_rate": 9.10291305266497e-05,
+      "loss": 0.0107,
+      "step": 7020
+    },
+    {
+      "epoch": 0.4095066115221064,
+      "grad_norm": 0.33217161893844604,
+      "learning_rate": 9.099760546627261e-05,
+      "loss": 0.0137,
+      "step": 7030
+    },
+    {
+      "epoch": 0.41008912448301976,
+      "grad_norm": 0.18960629403591156,
+      "learning_rate": 9.096603058993907e-05,
+      "loss": 0.0144,
+      "step": 7040
+    },
+    {
+      "epoch": 0.41067163744393315,
+      "grad_norm": 0.5246853828430176,
+      "learning_rate": 9.093440593601552e-05,
+      "loss": 0.0174,
+      "step": 7050
+    },
+    {
+      "epoch": 0.41125415040484653,
+      "grad_norm": 0.27294328808784485,
+      "learning_rate": 9.090273154292889e-05,
+      "loss": 0.0124,
+      "step": 7060
+    },
+    {
+      "epoch": 0.4118366633657599,
+      "grad_norm": 0.30159419775009155,
+      "learning_rate": 9.087100744916656e-05,
+      "loss": 0.0133,
+      "step": 7070
+    },
+    {
+      "epoch": 0.41241917632667324,
+      "grad_norm": 0.21936914324760437,
+      "learning_rate": 9.08392336932763e-05,
+      "loss": 0.0097,
+      "step": 7080
+    },
+    {
+      "epoch": 0.4130016892875866,
+      "grad_norm": 0.37547966837882996,
+      "learning_rate": 9.080741031386619e-05,
+      "loss": 0.012,
+      "step": 7090
+    },
+    {
+      "epoch": 0.4135842022485,
+      "grad_norm": 0.4171476662158966,
+      "learning_rate": 9.077553734960468e-05,
+      "loss": 0.012,
+      "step": 7100
+    },
+    {
+      "epoch": 0.4141667152094134,
+      "grad_norm": 0.2655220925807953,
+      "learning_rate": 9.074361483922041e-05,
+      "loss": 0.0113,
+      "step": 7110
+    },
+    {
+      "epoch": 0.4147492281703268,
+      "grad_norm": 0.3095017671585083,
+      "learning_rate": 9.071164282150224e-05,
+      "loss": 0.0126,
+      "step": 7120
+    },
+    {
+      "epoch": 0.41533174113124016,
+      "grad_norm": 0.3188718259334564,
+      "learning_rate": 9.067962133529919e-05,
+      "loss": 0.0122,
+      "step": 7130
+    },
+    {
+      "epoch": 0.41591425409215355,
+      "grad_norm": 0.36818554997444153,
+      "learning_rate": 9.064755041952036e-05,
+      "loss": 0.0157,
+      "step": 7140
+    },
+    {
+      "epoch": 0.41649676705306693,
+      "grad_norm": 0.19709156453609467,
+      "learning_rate": 9.061543011313497e-05,
+      "loss": 0.015,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4170792800139803,
+      "grad_norm": 0.2513802647590637,
+      "learning_rate": 9.05832604551722e-05,
+      "loss": 0.0156,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4176617929748937,
+      "grad_norm": 0.25856226682662964,
+      "learning_rate": 9.055104148472124e-05,
+      "loss": 0.0117,
+      "step": 7170
+    },
+    {
+      "epoch": 0.4182443059358071,
+      "grad_norm": 0.21729667484760284,
+      "learning_rate": 9.051877324093114e-05,
+      "loss": 0.0123,
+      "step": 7180
+    },
+    {
+      "epoch": 0.41882681889672047,
+      "grad_norm": 0.2332047075033188,
+      "learning_rate": 9.04864557630109e-05,
+      "loss": 0.0144,
+      "step": 7190
+    },
+    {
+      "epoch": 0.41940933185763385,
+      "grad_norm": 0.24884118139743805,
+      "learning_rate": 9.045408909022928e-05,
+      "loss": 0.014,
+      "step": 7200
+    },
+    {
+      "epoch": 0.41999184481854723,
+      "grad_norm": 0.39284127950668335,
+      "learning_rate": 9.042167326191484e-05,
+      "loss": 0.0098,
+      "step": 7210
+    },
+    {
+      "epoch": 0.4205743577794606,
+      "grad_norm": 0.29154670238494873,
+      "learning_rate": 9.038920831745587e-05,
+      "loss": 0.0142,
+      "step": 7220
+    },
+    {
+      "epoch": 0.421156870740374,
+      "grad_norm": 0.23585759103298187,
+      "learning_rate": 9.035669429630036e-05,
+      "loss": 0.0113,
+      "step": 7230
+    },
+    {
+      "epoch": 0.42173938370128733,
+      "grad_norm": 0.1617414355278015,
+      "learning_rate": 9.032413123795588e-05,
+      "loss": 0.0103,
+      "step": 7240
+    },
+    {
+      "epoch": 0.4223218966622007,
+      "grad_norm": 0.3884938359260559,
+      "learning_rate": 9.029151918198962e-05,
+      "loss": 0.0115,
+      "step": 7250
+    },
+    {
+      "epoch": 0.4229044096231141,
+      "grad_norm": 0.2697516083717346,
+      "learning_rate": 9.025885816802833e-05,
+      "loss": 0.0117,
+      "step": 7260
+    },
+    {
+      "epoch": 0.4234869225840275,
+      "grad_norm": 0.3080799877643585,
+      "learning_rate": 9.02261482357582e-05,
+      "loss": 0.0121,
+      "step": 7270
+    },
+    {
+      "epoch": 0.42406943554494086,
+      "grad_norm": 0.2844873070716858,
+      "learning_rate": 9.019338942492485e-05,
+      "loss": 0.0112,
+      "step": 7280
+    },
+    {
+      "epoch": 0.42465194850585425,
+      "grad_norm": 0.44505390524864197,
+      "learning_rate": 9.01605817753334e-05,
+      "loss": 0.0124,
+      "step": 7290
+    },
+    {
+      "epoch": 0.42523446146676763,
+      "grad_norm": 0.2775717079639435,
+      "learning_rate": 9.012772532684818e-05,
+      "loss": 0.0114,
+      "step": 7300
+    },
+    {
+      "epoch": 0.425816974427681,
+      "grad_norm": 0.2966673970222473,
+      "learning_rate": 9.00948201193929e-05,
+      "loss": 0.0145,
+      "step": 7310
+    },
+    {
+      "epoch": 0.4263994873885944,
+      "grad_norm": 0.2556905746459961,
+      "learning_rate": 9.006186619295047e-05,
+      "loss": 0.012,
+      "step": 7320
+    },
+    {
+      "epoch": 0.4269820003495078,
+      "grad_norm": 0.24772870540618896,
+      "learning_rate": 9.002886358756305e-05,
+      "loss": 0.0134,
+      "step": 7330
+    },
+    {
+      "epoch": 0.42756451331042117,
+      "grad_norm": 0.29119929671287537,
+      "learning_rate": 8.999581234333188e-05,
+      "loss": 0.0111,
+      "step": 7340
+    },
+    {
+      "epoch": 0.42814702627133455,
+      "grad_norm": 0.26602277159690857,
+      "learning_rate": 8.996271250041734e-05,
+      "loss": 0.0101,
+      "step": 7350
+    },
+    {
+      "epoch": 0.42872953923224794,
+      "grad_norm": 0.2989645302295685,
+      "learning_rate": 8.99295640990389e-05,
+      "loss": 0.0106,
+      "step": 7360
+    },
+    {
+      "epoch": 0.4293120521931613,
+      "grad_norm": 0.2931496798992157,
+      "learning_rate": 8.989636717947494e-05,
+      "loss": 0.0114,
+      "step": 7370
+    },
+    {
+      "epoch": 0.4298945651540747,
+      "grad_norm": 0.25160327553749084,
+      "learning_rate": 8.986312178206291e-05,
+      "loss": 0.0132,
+      "step": 7380
+    },
+    {
+      "epoch": 0.43047707811498803,
+      "grad_norm": 0.19953225553035736,
+      "learning_rate": 8.982982794719904e-05,
+      "loss": 0.0138,
+      "step": 7390
+    },
+    {
+      "epoch": 0.4310595910759014,
+      "grad_norm": 0.2953711152076721,
+      "learning_rate": 8.979648571533852e-05,
+      "loss": 0.0125,
+      "step": 7400
+    },
+    {
+      "epoch": 0.4316421040368148,
+      "grad_norm": 0.4054790139198303,
+      "learning_rate": 8.976309512699528e-05,
+      "loss": 0.0125,
+      "step": 7410
+    },
+    {
+      "epoch": 0.4322246169977282,
+      "grad_norm": 0.22115576267242432,
+      "learning_rate": 8.972965622274205e-05,
+      "loss": 0.0122,
+      "step": 7420
+    },
+    {
+      "epoch": 0.43280712995864157,
+      "grad_norm": 0.33209818601608276,
+      "learning_rate": 8.969616904321026e-05,
+      "loss": 0.0117,
+      "step": 7430
+    },
+    {
+      "epoch": 0.43338964291955495,
+      "grad_norm": 0.22817012667655945,
+      "learning_rate": 8.966263362908998e-05,
+      "loss": 0.013,
+      "step": 7440
+    },
+    {
+      "epoch": 0.43397215588046834,
+      "grad_norm": 0.3235332667827606,
+      "learning_rate": 8.962905002112989e-05,
+      "loss": 0.0096,
+      "step": 7450
+    },
+    {
+      "epoch": 0.4345546688413817,
+      "grad_norm": 0.1812901496887207,
+      "learning_rate": 8.959541826013724e-05,
+      "loss": 0.0131,
+      "step": 7460
+    },
+    {
+      "epoch": 0.4351371818022951,
+      "grad_norm": 0.388325959444046,
+      "learning_rate": 8.956173838697781e-05,
+      "loss": 0.0116,
+      "step": 7470
+    },
+    {
+      "epoch": 0.4357196947632085,
+      "grad_norm": 0.35594281554222107,
+      "learning_rate": 8.952801044257581e-05,
+      "loss": 0.0121,
+      "step": 7480
+    },
+    {
+      "epoch": 0.43630220772412187,
+      "grad_norm": 0.289470911026001,
+      "learning_rate": 8.949423446791387e-05,
+      "loss": 0.0134,
+      "step": 7490
+    },
+    {
+      "epoch": 0.43688472068503525,
+      "grad_norm": 0.3230888545513153,
+      "learning_rate": 8.9460410504033e-05,
+      "loss": 0.0115,
+      "step": 7500
+    },
+    {
+      "epoch": 0.43746723364594864,
+      "grad_norm": 0.20693032443523407,
+      "learning_rate": 8.942653859203248e-05,
+      "loss": 0.0103,
+      "step": 7510
+    },
+    {
+      "epoch": 0.438049746606862,
+      "grad_norm": 0.17195037007331848,
+      "learning_rate": 8.93926187730699e-05,
+      "loss": 0.0134,
+      "step": 7520
+    },
+    {
+      "epoch": 0.4386322595677754,
+      "grad_norm": 0.1913907527923584,
+      "learning_rate": 8.935865108836102e-05,
+      "loss": 0.0118,
+      "step": 7530
+    },
+    {
+      "epoch": 0.4392147725286888,
+      "grad_norm": 0.23225539922714233,
+      "learning_rate": 8.932463557917981e-05,
+      "loss": 0.0101,
+      "step": 7540
+    },
+    {
+      "epoch": 0.4397972854896021,
+      "grad_norm": 0.2920669913291931,
+      "learning_rate": 8.929057228685829e-05,
+      "loss": 0.0099,
+      "step": 7550
+    },
+    {
+      "epoch": 0.4403797984505155,
+      "grad_norm": 0.34749093651771545,
+      "learning_rate": 8.925646125278658e-05,
+      "loss": 0.0116,
+      "step": 7560
+    },
+    {
+      "epoch": 0.4409623114114289,
+      "grad_norm": 0.3100294768810272,
+      "learning_rate": 8.92223025184128e-05,
+      "loss": 0.0112,
+      "step": 7570
+    },
+    {
+      "epoch": 0.44154482437234227,
+      "grad_norm": 0.3635091185569763,
+      "learning_rate": 8.918809612524303e-05,
+      "loss": 0.0147,
+      "step": 7580
+    },
+    {
+      "epoch": 0.44212733733325565,
+      "grad_norm": 0.2399435192346573,
+      "learning_rate": 8.91538421148413e-05,
+      "loss": 0.0112,
+      "step": 7590
+    },
+    {
+      "epoch": 0.44270985029416904,
+      "grad_norm": 0.25012388825416565,
+      "learning_rate": 8.91195405288294e-05,
+      "loss": 0.0128,
+      "step": 7600
+    },
+    {
+      "epoch": 0.4432923632550824,
+      "grad_norm": 0.48543089628219604,
+      "learning_rate": 8.908519140888705e-05,
+      "loss": 0.0127,
+      "step": 7610
+    },
+    {
+      "epoch": 0.4438748762159958,
+      "grad_norm": 0.21692843735218048,
+      "learning_rate": 8.905079479675164e-05,
+      "loss": 0.01,
+      "step": 7620
+    },
+    {
+      "epoch": 0.4444573891769092,
+      "grad_norm": 0.23970955610275269,
+      "learning_rate": 8.901635073421831e-05,
+      "loss": 0.0109,
+      "step": 7630
+    },
+    {
+      "epoch": 0.4450399021378226,
+      "grad_norm": 0.20894008874893188,
+      "learning_rate": 8.898185926313983e-05,
+      "loss": 0.0117,
+      "step": 7640
+    },
+    {
+      "epoch": 0.44562241509873596,
+      "grad_norm": 0.2764657735824585,
+      "learning_rate": 8.894732042542659e-05,
+      "loss": 0.0141,
+      "step": 7650
+    },
+    {
+      "epoch": 0.44620492805964934,
+      "grad_norm": 0.31210198998451233,
+      "learning_rate": 8.891273426304654e-05,
+      "loss": 0.0114,
+      "step": 7660
+    },
+    {
+      "epoch": 0.4467874410205627,
+      "grad_norm": 0.3830687701702118,
+      "learning_rate": 8.887810081802515e-05,
+      "loss": 0.0112,
+      "step": 7670
+    },
+    {
+      "epoch": 0.4473699539814761,
+      "grad_norm": 0.3342027962207794,
+      "learning_rate": 8.884342013244529e-05,
+      "loss": 0.0112,
+      "step": 7680
+    },
+    {
+      "epoch": 0.4479524669423895,
+      "grad_norm": 0.27329325675964355,
+      "learning_rate": 8.880869224844725e-05,
+      "loss": 0.0099,
+      "step": 7690
+    },
+    {
+      "epoch": 0.4485349799033028,
+      "grad_norm": 0.35470640659332275,
+      "learning_rate": 8.877391720822875e-05,
+      "loss": 0.0102,
+      "step": 7700
+    },
+    {
+      "epoch": 0.4491174928642162,
+      "grad_norm": 0.3574216663837433,
+      "learning_rate": 8.873909505404468e-05,
+      "loss": 0.0119,
+      "step": 7710
+    },
+    {
+      "epoch": 0.4497000058251296,
+      "grad_norm": 0.3599697947502136,
+      "learning_rate": 8.870422582820726e-05,
+      "loss": 0.0092,
+      "step": 7720
+    },
+    {
+      "epoch": 0.450282518786043,
+      "grad_norm": 0.2701171338558197,
+      "learning_rate": 8.866930957308588e-05,
+      "loss": 0.0138,
+      "step": 7730
+    },
+    {
+      "epoch": 0.45086503174695636,
+      "grad_norm": 0.31447646021842957,
+      "learning_rate": 8.86343463311071e-05,
+      "loss": 0.0102,
+      "step": 7740
+    },
+    {
+      "epoch": 0.45144754470786974,
+      "grad_norm": 0.30923134088516235,
+      "learning_rate": 8.859933614475452e-05,
+      "loss": 0.0119,
+      "step": 7750
+    },
+    {
+      "epoch": 0.4520300576687831,
+      "grad_norm": 0.1891762614250183,
+      "learning_rate": 8.85642790565689e-05,
+      "loss": 0.0103,
+      "step": 7760
+    },
+    {
+      "epoch": 0.4526125706296965,
+      "grad_norm": 0.2857305705547333,
+      "learning_rate": 8.852917510914782e-05,
+      "loss": 0.0096,
+      "step": 7770
+    },
+    {
+      "epoch": 0.4531950835906099,
+      "grad_norm": 0.2828695774078369,
+      "learning_rate": 8.84940243451459e-05,
+      "loss": 0.0129,
+      "step": 7780
+    },
+    {
+      "epoch": 0.4537775965515233,
+      "grad_norm": 0.24774837493896484,
+      "learning_rate": 8.845882680727469e-05,
+      "loss": 0.0102,
+      "step": 7790
+    },
+    {
+      "epoch": 0.45436010951243666,
+      "grad_norm": 0.21391414105892181,
+      "learning_rate": 8.842358253830246e-05,
+      "loss": 0.0087,
+      "step": 7800
+    },
+    {
+      "epoch": 0.45494262247335004,
+      "grad_norm": 0.23724718391895294,
+      "learning_rate": 8.838829158105434e-05,
+      "loss": 0.0104,
+      "step": 7810
+    },
+    {
+      "epoch": 0.45552513543426343,
+      "grad_norm": 0.26340359449386597,
+      "learning_rate": 8.835295397841217e-05,
+      "loss": 0.0142,
+      "step": 7820
+    },
+    {
+      "epoch": 0.4561076483951768,
+      "grad_norm": 0.32494908571243286,
+      "learning_rate": 8.831756977331447e-05,
+      "loss": 0.0152,
+      "step": 7830
+    },
+    {
+      "epoch": 0.4566901613560902,
+      "grad_norm": 0.42112550139427185,
+      "learning_rate": 8.828213900875638e-05,
+      "loss": 0.0124,
+      "step": 7840
+    },
+    {
+      "epoch": 0.4572726743170036,
+      "grad_norm": 0.3547781705856323,
+      "learning_rate": 8.824666172778964e-05,
+      "loss": 0.0107,
+      "step": 7850
+    },
+    {
+      "epoch": 0.4578551872779169,
+      "grad_norm": 0.32153594493865967,
+      "learning_rate": 8.821113797352245e-05,
+      "loss": 0.0091,
+      "step": 7860
+    },
+    {
+      "epoch": 0.4584377002388303,
+      "grad_norm": 0.25656503438949585,
+      "learning_rate": 8.817556778911956e-05,
+      "loss": 0.0109,
+      "step": 7870
+    },
+    {
+      "epoch": 0.4590202131997437,
+      "grad_norm": 0.2758748233318329,
+      "learning_rate": 8.81399512178021e-05,
+      "loss": 0.0119,
+      "step": 7880
+    },
+    {
+      "epoch": 0.45960272616065706,
+      "grad_norm": 0.25740793347358704,
+      "learning_rate": 8.810428830284752e-05,
+      "loss": 0.0116,
+      "step": 7890
+    },
+    {
+      "epoch": 0.46018523912157044,
+      "grad_norm": 0.23661264777183533,
+      "learning_rate": 8.806857908758967e-05,
+      "loss": 0.0109,
+      "step": 7900
+    },
+    {
+      "epoch": 0.46076775208248383,
+      "grad_norm": 0.33345550298690796,
+      "learning_rate": 8.80328236154186e-05,
+      "loss": 0.0129,
+      "step": 7910
+    },
+    {
+      "epoch": 0.4613502650433972,
+      "grad_norm": 0.15988755226135254,
+      "learning_rate": 8.799702192978056e-05,
+      "loss": 0.0085,
+      "step": 7920
+    },
+    {
+      "epoch": 0.4619327780043106,
+      "grad_norm": 0.2677052319049835,
+      "learning_rate": 8.7961174074178e-05,
+      "loss": 0.0112,
+      "step": 7930
+    },
+    {
+      "epoch": 0.462515290965224,
+      "grad_norm": 0.16391333937644958,
+      "learning_rate": 8.792528009216942e-05,
+      "loss": 0.0108,
+      "step": 7940
+    },
+    {
+      "epoch": 0.46309780392613736,
+      "grad_norm": 0.3000504672527313,
+      "learning_rate": 8.788934002736943e-05,
+      "loss": 0.0159,
+      "step": 7950
+    },
+    {
+      "epoch": 0.46368031688705075,
+      "grad_norm": 0.2887212038040161,
+      "learning_rate": 8.785335392344857e-05,
+      "loss": 0.0157,
+      "step": 7960
+    },
+    {
+      "epoch": 0.46426282984796413,
+      "grad_norm": 0.34124699234962463,
+      "learning_rate": 8.781732182413336e-05,
+      "loss": 0.0132,
+      "step": 7970
+    },
+    {
+      "epoch": 0.4648453428088775,
+      "grad_norm": 0.3974616229534149,
+      "learning_rate": 8.778124377320618e-05,
+      "loss": 0.0195,
+      "step": 7980
+    },
+    {
+      "epoch": 0.4654278557697909,
+      "grad_norm": 0.3733963668346405,
+      "learning_rate": 8.774511981450529e-05,
+      "loss": 0.0132,
+      "step": 7990
+    },
+    {
+      "epoch": 0.4660103687307043,
+      "grad_norm": 0.29485955834388733,
+      "learning_rate": 8.770894999192467e-05,
+      "loss": 0.0139,
+      "step": 8000
+    },
+    {
+      "epoch": 0.4665928816916176,
+      "grad_norm": 0.3780854642391205,
+      "learning_rate": 8.767273434941412e-05,
+      "loss": 0.0154,
+      "step": 8010
+    },
+    {
+      "epoch": 0.467175394652531,
+      "grad_norm": 0.2440689653158188,
+      "learning_rate": 8.7636472930979e-05,
+      "loss": 0.0133,
+      "step": 8020
+    },
+    {
+      "epoch": 0.4677579076134444,
+      "grad_norm": 0.14352597296237946,
+      "learning_rate": 8.76001657806804e-05,
+      "loss": 0.0122,
+      "step": 8030
+    },
+    {
+      "epoch": 0.46834042057435776,
+      "grad_norm": 0.2142922729253769,
+      "learning_rate": 8.756381294263491e-05,
+      "loss": 0.0096,
+      "step": 8040
+    },
+    {
+      "epoch": 0.46892293353527115,
+      "grad_norm": 0.3044984042644501,
+      "learning_rate": 8.752741446101464e-05,
+      "loss": 0.009,
+      "step": 8050
+    },
+    {
+      "epoch": 0.46950544649618453,
+      "grad_norm": 0.34259000420570374,
+      "learning_rate": 8.749097038004722e-05,
+      "loss": 0.0127,
+      "step": 8060
+    },
+    {
+      "epoch": 0.4700879594570979,
+      "grad_norm": 0.213138148188591,
+      "learning_rate": 8.745448074401561e-05,
+      "loss": 0.0098,
+      "step": 8070
+    },
+    {
+      "epoch": 0.4706704724180113,
+      "grad_norm": 0.3469372093677521,
+      "learning_rate": 8.741794559725818e-05,
+      "loss": 0.0117,
+      "step": 8080
+    },
+    {
+      "epoch": 0.4712529853789247,
+      "grad_norm": 0.2347586750984192,
+      "learning_rate": 8.738136498416856e-05,
+      "loss": 0.0087,
+      "step": 8090
+    },
+    {
+      "epoch": 0.47183549833983807,
+      "grad_norm": 0.30541104078292847,
+      "learning_rate": 8.734473894919564e-05,
+      "loss": 0.0083,
+      "step": 8100
+    },
+    {
+      "epoch": 0.47241801130075145,
+      "grad_norm": 0.227009117603302,
+      "learning_rate": 8.730806753684353e-05,
+      "loss": 0.0086,
+      "step": 8110
+    },
+    {
+      "epoch": 0.47300052426166483,
+      "grad_norm": 0.12994255125522614,
+      "learning_rate": 8.727135079167144e-05,
+      "loss": 0.0101,
+      "step": 8120
+    },
+    {
+      "epoch": 0.4735830372225782,
+      "grad_norm": 0.14903749525547028,
+      "learning_rate": 8.723458875829368e-05,
+      "loss": 0.0071,
+      "step": 8130
+    },
+    {
+      "epoch": 0.4741655501834916,
+      "grad_norm": 0.2669014632701874,
+      "learning_rate": 8.719778148137958e-05,
+      "loss": 0.0118,
+      "step": 8140
+    },
+    {
+      "epoch": 0.474748063144405,
+      "grad_norm": 0.2108674794435501,
+      "learning_rate": 8.716092900565346e-05,
+      "loss": 0.0107,
+      "step": 8150
+    },
+    {
+      "epoch": 0.47533057610531837,
+      "grad_norm": 0.17570310831069946,
+      "learning_rate": 8.712403137589456e-05,
+      "loss": 0.0113,
+      "step": 8160
+    },
+    {
+      "epoch": 0.4759130890662317,
+      "grad_norm": 0.23179100453853607,
+      "learning_rate": 8.708708863693697e-05,
+      "loss": 0.0103,
+      "step": 8170
+    },
+    {
+      "epoch": 0.4764956020271451,
+      "grad_norm": 0.20041556656360626,
+      "learning_rate": 8.70501008336696e-05,
+      "loss": 0.0117,
+      "step": 8180
+    },
+    {
+      "epoch": 0.47707811498805847,
+      "grad_norm": 0.3331192433834076,
+      "learning_rate": 8.701306801103611e-05,
+      "loss": 0.0125,
+      "step": 8190
+    },
+    {
+      "epoch": 0.47766062794897185,
+      "grad_norm": 0.3115651309490204,
+      "learning_rate": 8.697599021403489e-05,
+      "loss": 0.0115,
+      "step": 8200
+    },
+    {
+      "epoch": 0.47824314090988523,
+      "grad_norm": 0.18313908576965332,
+      "learning_rate": 8.693886748771896e-05,
+      "loss": 0.0118,
+      "step": 8210
+    },
+    {
+      "epoch": 0.4788256538707986,
+      "grad_norm": 0.3065651059150696,
+      "learning_rate": 8.690169987719593e-05,
+      "loss": 0.0101,
+      "step": 8220
+    },
+    {
+      "epoch": 0.479408166831712,
+      "grad_norm": 0.3847108781337738,
+      "learning_rate": 8.686448742762791e-05,
+      "loss": 0.0106,
+      "step": 8230
+    },
+    {
+      "epoch": 0.4799906797926254,
+      "grad_norm": 0.27982595562934875,
+      "learning_rate": 8.682723018423161e-05,
+      "loss": 0.0084,
+      "step": 8240
+    },
+    {
+      "epoch": 0.48057319275353877,
+      "grad_norm": 0.2973509728908539,
+      "learning_rate": 8.678992819227804e-05,
+      "loss": 0.0105,
+      "step": 8250
+    },
+    {
+      "epoch": 0.48115570571445215,
+      "grad_norm": 0.22349682450294495,
+      "learning_rate": 8.675258149709265e-05,
+      "loss": 0.0096,
+      "step": 8260
+    },
+    {
+      "epoch": 0.48173821867536554,
+      "grad_norm": 0.33618637919425964,
+      "learning_rate": 8.67151901440552e-05,
+      "loss": 0.0148,
+      "step": 8270
+    },
+    {
+      "epoch": 0.4823207316362789,
+      "grad_norm": 0.43434813618659973,
+      "learning_rate": 8.667775417859971e-05,
+      "loss": 0.0145,
+      "step": 8280
+    },
+    {
+      "epoch": 0.4829032445971923,
+      "grad_norm": 0.26474690437316895,
+      "learning_rate": 8.664027364621441e-05,
+      "loss": 0.0126,
+      "step": 8290
+    },
+    {
+      "epoch": 0.4834857575581057,
+      "grad_norm": 0.22505629062652588,
+      "learning_rate": 8.660274859244168e-05,
+      "loss": 0.0091,
+      "step": 8300
+    },
+    {
+      "epoch": 0.4840682705190191,
+      "grad_norm": 0.33556726574897766,
+      "learning_rate": 8.656517906287799e-05,
+      "loss": 0.0151,
+      "step": 8310
+    },
+    {
+      "epoch": 0.4846507834799324,
+      "grad_norm": 0.3187960386276245,
+      "learning_rate": 8.652756510317387e-05,
+      "loss": 0.0121,
+      "step": 8320
+    },
+    {
+      "epoch": 0.4852332964408458,
+      "grad_norm": 0.2117210030555725,
+      "learning_rate": 8.648990675903382e-05,
+      "loss": 0.0113,
+      "step": 8330
+    },
+    {
+      "epoch": 0.48581580940175917,
+      "grad_norm": 0.3122064769268036,
+      "learning_rate": 8.64522040762163e-05,
+      "loss": 0.0121,
+      "step": 8340
+    },
+    {
+      "epoch": 0.48639832236267255,
+      "grad_norm": 0.28426116704940796,
+      "learning_rate": 8.64144571005336e-05,
+      "loss": 0.0129,
+      "step": 8350
+    },
+    {
+      "epoch": 0.48698083532358594,
+      "grad_norm": 0.3730778992176056,
+      "learning_rate": 8.637666587785184e-05,
+      "loss": 0.0127,
+      "step": 8360
+    },
+    {
+      "epoch": 0.4875633482844993,
+      "grad_norm": 0.3090813457965851,
+      "learning_rate": 8.633883045409096e-05,
+      "loss": 0.0111,
+      "step": 8370
+    },
+    {
+      "epoch": 0.4881458612454127,
+      "grad_norm": 0.23262611031532288,
+      "learning_rate": 8.630095087522457e-05,
+      "loss": 0.0102,
+      "step": 8380
+    },
+    {
+      "epoch": 0.4887283742063261,
+      "grad_norm": 0.23348644375801086,
+      "learning_rate": 8.62630271872799e-05,
+      "loss": 0.0112,
+      "step": 8390
+    },
+    {
+      "epoch": 0.48931088716723947,
+      "grad_norm": 0.23587965965270996,
+      "learning_rate": 8.62250594363378e-05,
+      "loss": 0.008,
+      "step": 8400
+    },
+    {
+      "epoch": 0.48989340012815286,
+      "grad_norm": 0.37986883521080017,
+      "learning_rate": 8.61870476685327e-05,
+      "loss": 0.0109,
+      "step": 8410
+    },
+    {
+      "epoch": 0.49047591308906624,
+      "grad_norm": 0.4875761866569519,
+      "learning_rate": 8.614899193005248e-05,
+      "loss": 0.0115,
+      "step": 8420
+    },
+    {
+      "epoch": 0.4910584260499796,
+      "grad_norm": 0.2998208999633789,
+      "learning_rate": 8.611089226713842e-05,
+      "loss": 0.01,
+      "step": 8430
+    },
+    {
+      "epoch": 0.491640939010893,
+      "grad_norm": 0.20100432634353638,
+      "learning_rate": 8.607274872608521e-05,
+      "loss": 0.0099,
+      "step": 8440
+    },
+    {
+      "epoch": 0.4922234519718064,
+      "grad_norm": 0.3055593967437744,
+      "learning_rate": 8.603456135324088e-05,
+      "loss": 0.0096,
+      "step": 8450
+    },
+    {
+      "epoch": 0.4928059649327198,
+      "grad_norm": 0.22969605028629303,
+      "learning_rate": 8.599633019500665e-05,
+      "loss": 0.009,
+      "step": 8460
+    },
+    {
+      "epoch": 0.49338847789363316,
+      "grad_norm": 0.2377668172121048,
+      "learning_rate": 8.595805529783702e-05,
+      "loss": 0.0095,
+      "step": 8470
+    },
+    {
+      "epoch": 0.4939709908545465,
+      "grad_norm": 0.3378452956676483,
+      "learning_rate": 8.591973670823958e-05,
+      "loss": 0.0145,
+      "step": 8480
+    },
+    {
+      "epoch": 0.49455350381545987,
+      "grad_norm": 0.33285266160964966,
+      "learning_rate": 8.588137447277503e-05,
+      "loss": 0.0125,
+      "step": 8490
+    },
+    {
+      "epoch": 0.49513601677637326,
+      "grad_norm": 0.32284048199653625,
+      "learning_rate": 8.584296863805709e-05,
+      "loss": 0.009,
+      "step": 8500
+    },
+    {
+      "epoch": 0.49571852973728664,
+      "grad_norm": 0.3326222002506256,
+      "learning_rate": 8.580451925075248e-05,
+      "loss": 0.0119,
+      "step": 8510
+    },
+    {
+      "epoch": 0.4963010426982,
+      "grad_norm": 0.21906715631484985,
+      "learning_rate": 8.576602635758085e-05,
+      "loss": 0.0102,
+      "step": 8520
+    },
+    {
+      "epoch": 0.4968835556591134,
+      "grad_norm": 0.20839254558086395,
+      "learning_rate": 8.572749000531469e-05,
+      "loss": 0.0086,
+      "step": 8530
+    },
+    {
+      "epoch": 0.4974660686200268,
+      "grad_norm": 0.34974223375320435,
+      "learning_rate": 8.568891024077926e-05,
+      "loss": 0.0096,
+      "step": 8540
+    },
+    {
+      "epoch": 0.4980485815809402,
+      "grad_norm": 0.35410234332084656,
+      "learning_rate": 8.565028711085265e-05,
+      "loss": 0.0122,
+      "step": 8550
+    },
+    {
+      "epoch": 0.49863109454185356,
+      "grad_norm": 0.33047378063201904,
+      "learning_rate": 8.561162066246561e-05,
+      "loss": 0.0107,
+      "step": 8560
+    },
+    {
+      "epoch": 0.49921360750276694,
+      "grad_norm": 0.3621072471141815,
+      "learning_rate": 8.557291094260152e-05,
+      "loss": 0.0127,
+      "step": 8570
+    },
+    {
+      "epoch": 0.4997961204636803,
+      "grad_norm": 0.26117265224456787,
+      "learning_rate": 8.553415799829629e-05,
+      "loss": 0.0107,
+      "step": 8580
+    },
+    {
+      "epoch": 0.5003786334245937,
+      "grad_norm": 0.16914372146129608,
+      "learning_rate": 8.549536187663847e-05,
+      "loss": 0.0084,
+      "step": 8590
+    },
+    {
+      "epoch": 0.500961146385507,
+      "grad_norm": 0.2097463309764862,
+      "learning_rate": 8.545652262476897e-05,
+      "loss": 0.013,
+      "step": 8600
+    },
+    {
+      "epoch": 0.5015436593464204,
+      "grad_norm": 0.2855682671070099,
+      "learning_rate": 8.541764028988115e-05,
+      "loss": 0.0113,
+      "step": 8610
+    },
+    {
+      "epoch": 0.5021261723073338,
+      "grad_norm": 0.21892935037612915,
+      "learning_rate": 8.537871491922071e-05,
+      "loss": 0.0117,
+      "step": 8620
+    },
+    {
+      "epoch": 0.5027086852682472,
+      "grad_norm": 0.26727095246315,
+      "learning_rate": 8.533974656008566e-05,
+      "loss": 0.0112,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5032911982291606,
+      "grad_norm": 0.23189052939414978,
+      "learning_rate": 8.530073525982621e-05,
+      "loss": 0.009,
+      "step": 8640
+    },
+    {
+      "epoch": 0.503873711190074,
+      "grad_norm": 0.20634156465530396,
+      "learning_rate": 8.526168106584476e-05,
+      "loss": 0.0129,
+      "step": 8650
+    },
+    {
+      "epoch": 0.5044562241509873,
+      "grad_norm": 0.3063119947910309,
+      "learning_rate": 8.522258402559586e-05,
+      "loss": 0.0116,
+      "step": 8660
+    },
+    {
+      "epoch": 0.5050387371119007,
+      "grad_norm": 0.3161621391773224,
+      "learning_rate": 8.518344418658613e-05,
+      "loss": 0.0114,
+      "step": 8670
+    },
+    {
+      "epoch": 0.5056212500728141,
+      "grad_norm": 0.23569059371948242,
+      "learning_rate": 8.51442615963741e-05,
+      "loss": 0.0133,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5062037630337275,
+      "grad_norm": 0.24308477342128754,
+      "learning_rate": 8.510503630257034e-05,
+      "loss": 0.0104,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5067862759946409,
+      "grad_norm": 0.303740531206131,
+      "learning_rate": 8.50657683528373e-05,
+      "loss": 0.0096,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5073687889555543,
+      "grad_norm": 0.22782230377197266,
+      "learning_rate": 8.502645779488922e-05,
+      "loss": 0.0086,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5079513019164676,
+      "grad_norm": 0.3413344919681549,
+      "learning_rate": 8.498710467649213e-05,
+      "loss": 0.01,
+      "step": 8720
+    },
+    {
+      "epoch": 0.508533814877381,
+      "grad_norm": 0.2716794013977051,
+      "learning_rate": 8.494770904546381e-05,
+      "loss": 0.0123,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5091163278382944,
+      "grad_norm": 0.31317803263664246,
+      "learning_rate": 8.490827094967363e-05,
+      "loss": 0.0114,
+      "step": 8740
+    },
+    {
+      "epoch": 0.5096988407992078,
+      "grad_norm": 0.25290921330451965,
+      "learning_rate": 8.486879043704263e-05,
+      "loss": 0.0112,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5102813537601212,
+      "grad_norm": 0.3568704128265381,
+      "learning_rate": 8.482926755554332e-05,
+      "loss": 0.0113,
+      "step": 8760
+    },
+    {
+      "epoch": 0.5108638667210346,
+      "grad_norm": 0.30301928520202637,
+      "learning_rate": 8.478970235319975e-05,
+      "loss": 0.01,
+      "step": 8770
+    },
+    {
+      "epoch": 0.511446379681948,
+      "grad_norm": 0.40317073464393616,
+      "learning_rate": 8.475009487808737e-05,
+      "loss": 0.014,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5120288926428613,
+      "grad_norm": 0.38880282640457153,
+      "learning_rate": 8.471044517833298e-05,
+      "loss": 0.0131,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5126114056037747,
+      "grad_norm": 0.4629809856414795,
+      "learning_rate": 8.467075330211474e-05,
+      "loss": 0.011,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5131939185646881,
+      "grad_norm": 0.20804764330387115,
+      "learning_rate": 8.463101929766198e-05,
+      "loss": 0.0138,
+      "step": 8810
+    },
+    {
+      "epoch": 0.5137764315256015,
+      "grad_norm": 0.45887529850006104,
+      "learning_rate": 8.459124321325528e-05,
+      "loss": 0.0132,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5143589444865149,
+      "grad_norm": 0.360636830329895,
+      "learning_rate": 8.455142509722634e-05,
+      "loss": 0.0119,
+      "step": 8830
+    },
+    {
+      "epoch": 0.5149414574474283,
+      "grad_norm": 0.38743698596954346,
+      "learning_rate": 8.451156499795792e-05,
+      "loss": 0.01,
+      "step": 8840
+    },
+    {
+      "epoch": 0.5155239704083416,
+      "grad_norm": 0.3410528302192688,
+      "learning_rate": 8.447166296388381e-05,
+      "loss": 0.0118,
+      "step": 8850
+    },
+    {
+      "epoch": 0.516106483369255,
+      "grad_norm": 0.22457276284694672,
+      "learning_rate": 8.443171904348872e-05,
+      "loss": 0.0119,
+      "step": 8860
+    },
+    {
+      "epoch": 0.5166889963301683,
+      "grad_norm": 0.23236508667469025,
+      "learning_rate": 8.43917332853083e-05,
+      "loss": 0.0093,
+      "step": 8870
+    },
+    {
+      "epoch": 0.5172715092910817,
+      "grad_norm": 0.2859998643398285,
+      "learning_rate": 8.435170573792901e-05,
+      "loss": 0.0105,
+      "step": 8880
+    },
+    {
+      "epoch": 0.5178540222519951,
+      "grad_norm": 0.25001728534698486,
+      "learning_rate": 8.431163644998808e-05,
+      "loss": 0.0094,
+      "step": 8890
+    },
+    {
+      "epoch": 0.5184365352129084,
+      "grad_norm": 0.3896948993206024,
+      "learning_rate": 8.42715254701735e-05,
+      "loss": 0.0143,
+      "step": 8900
+    },
+    {
+      "epoch": 0.5190190481738218,
+      "grad_norm": 0.27904146909713745,
+      "learning_rate": 8.42313728472239e-05,
+      "loss": 0.0089,
+      "step": 8910
+    },
+    {
+      "epoch": 0.5196015611347352,
+      "grad_norm": 0.25483042001724243,
+      "learning_rate": 8.419117862992844e-05,
+      "loss": 0.0096,
+      "step": 8920
+    },
+    {
+      "epoch": 0.5201840740956486,
+      "grad_norm": 0.2366785705089569,
+      "learning_rate": 8.415094286712694e-05,
+      "loss": 0.0083,
+      "step": 8930
+    },
+    {
+      "epoch": 0.520766587056562,
+      "grad_norm": 0.2771153450012207,
+      "learning_rate": 8.411066560770965e-05,
+      "loss": 0.009,
+      "step": 8940
+    },
+    {
+      "epoch": 0.5213491000174754,
+      "grad_norm": 0.22708876430988312,
+      "learning_rate": 8.407034690061721e-05,
+      "loss": 0.011,
+      "step": 8950
+    },
+    {
+      "epoch": 0.5219316129783887,
+      "grad_norm": 0.32733431458473206,
+      "learning_rate": 8.402998679484067e-05,
+      "loss": 0.0115,
+      "step": 8960
+    },
+    {
+      "epoch": 0.5225141259393021,
+      "grad_norm": 0.2673744857311249,
+      "learning_rate": 8.398958533942135e-05,
+      "loss": 0.0074,
+      "step": 8970
+    },
+    {
+      "epoch": 0.5230966389002155,
+      "grad_norm": 0.2505076825618744,
+      "learning_rate": 8.394914258345083e-05,
+      "loss": 0.0122,
+      "step": 8980
+    },
+    {
+      "epoch": 0.5236791518611289,
+      "grad_norm": 0.42512375116348267,
+      "learning_rate": 8.390865857607089e-05,
+      "loss": 0.0114,
+      "step": 8990
+    },
+    {
+      "epoch": 0.5242616648220423,
+      "grad_norm": 0.4811088740825653,
+      "learning_rate": 8.386813336647339e-05,
+      "loss": 0.015,
+      "step": 9000
+    },
+    {
+      "epoch": 0.5248441777829557,
+      "grad_norm": 0.2737468481063843,
+      "learning_rate": 8.38275670039003e-05,
+      "loss": 0.0108,
+      "step": 9010
+    },
+    {
+      "epoch": 0.525426690743869,
+      "grad_norm": 0.21187454462051392,
+      "learning_rate": 8.378695953764356e-05,
+      "loss": 0.0085,
+      "step": 9020
+    },
+    {
+      "epoch": 0.5260092037047824,
+      "grad_norm": 0.30845895409584045,
+      "learning_rate": 8.374631101704509e-05,
+      "loss": 0.0099,
+      "step": 9030
+    },
+    {
+      "epoch": 0.5265917166656958,
+      "grad_norm": 0.20114067196846008,
+      "learning_rate": 8.370562149149665e-05,
+      "loss": 0.009,
+      "step": 9040
+    },
+    {
+      "epoch": 0.5271742296266092,
+      "grad_norm": 0.22177201509475708,
+      "learning_rate": 8.366489101043989e-05,
+      "loss": 0.0095,
+      "step": 9050
+    },
+    {
+      "epoch": 0.5277567425875226,
+      "grad_norm": 0.19219940900802612,
+      "learning_rate": 8.362411962336614e-05,
+      "loss": 0.0095,
+      "step": 9060
+    },
+    {
+      "epoch": 0.528339255548436,
+      "grad_norm": 0.44997864961624146,
+      "learning_rate": 8.358330737981651e-05,
+      "loss": 0.0105,
+      "step": 9070
+    },
+    {
+      "epoch": 0.5289217685093494,
+      "grad_norm": 0.2769324481487274,
+      "learning_rate": 8.35424543293817e-05,
+      "loss": 0.0116,
+      "step": 9080
+    },
+    {
+      "epoch": 0.5295042814702627,
+      "grad_norm": 0.28238800168037415,
+      "learning_rate": 8.350156052170206e-05,
+      "loss": 0.0101,
+      "step": 9090
+    },
+    {
+      "epoch": 0.5300867944311761,
+      "grad_norm": 0.2641945481300354,
+      "learning_rate": 8.346062600646738e-05,
+      "loss": 0.0097,
+      "step": 9100
+    },
+    {
+      "epoch": 0.5306693073920895,
+      "grad_norm": 0.27481865882873535,
+      "learning_rate": 8.341965083341697e-05,
+      "loss": 0.0134,
+      "step": 9110
+    },
+    {
+      "epoch": 0.5312518203530029,
+      "grad_norm": 0.3104117512702942,
+      "learning_rate": 8.337863505233953e-05,
+      "loss": 0.0105,
+      "step": 9120
+    },
+    {
+      "epoch": 0.5318343333139163,
+      "grad_norm": 0.38113418221473694,
+      "learning_rate": 8.33375787130731e-05,
+      "loss": 0.0083,
+      "step": 9130
+    },
+    {
+      "epoch": 0.5324168462748297,
+      "grad_norm": 0.21746212244033813,
+      "learning_rate": 8.3296481865505e-05,
+      "loss": 0.0106,
+      "step": 9140
+    },
+    {
+      "epoch": 0.532999359235743,
+      "grad_norm": 0.3854799270629883,
+      "learning_rate": 8.325534455957181e-05,
+      "loss": 0.0113,
+      "step": 9150
+    },
+    {
+      "epoch": 0.5335818721966564,
+      "grad_norm": 0.3229290544986725,
+      "learning_rate": 8.321416684525917e-05,
+      "loss": 0.01,
+      "step": 9160
+    },
+    {
+      "epoch": 0.5341643851575698,
+      "grad_norm": 0.28699108958244324,
+      "learning_rate": 8.317294877260193e-05,
+      "loss": 0.0116,
+      "step": 9170
+    },
+    {
+      "epoch": 0.5347468981184831,
+      "grad_norm": 0.38508835434913635,
+      "learning_rate": 8.313169039168394e-05,
+      "loss": 0.0116,
+      "step": 9180
+    },
+    {
+      "epoch": 0.5353294110793965,
+      "grad_norm": 0.19626887142658234,
+      "learning_rate": 8.3090391752638e-05,
+      "loss": 0.0087,
+      "step": 9190
+    },
+    {
+      "epoch": 0.5359119240403099,
+      "grad_norm": 0.21657134592533112,
+      "learning_rate": 8.304905290564586e-05,
+      "loss": 0.0115,
+      "step": 9200
+    },
+    {
+      "epoch": 0.5364944370012232,
+      "grad_norm": 0.3363923132419586,
+      "learning_rate": 8.300767390093813e-05,
+      "loss": 0.0107,
+      "step": 9210
+    },
+    {
+      "epoch": 0.5370769499621366,
+      "grad_norm": 0.3121386170387268,
+      "learning_rate": 8.296625478879417e-05,
+      "loss": 0.0158,
+      "step": 9220
+    },
+    {
+      "epoch": 0.53765946292305,
+      "grad_norm": 0.371357262134552,
+      "learning_rate": 8.292479561954214e-05,
+      "loss": 0.0112,
+      "step": 9230
+    },
+    {
+      "epoch": 0.5382419758839634,
+      "grad_norm": 0.37682169675827026,
+      "learning_rate": 8.288329644355883e-05,
+      "loss": 0.0114,
+      "step": 9240
+    },
+    {
+      "epoch": 0.5388244888448768,
+      "grad_norm": 0.28736403584480286,
+      "learning_rate": 8.284175731126965e-05,
+      "loss": 0.0134,
+      "step": 9250
+    },
+    {
+      "epoch": 0.5394070018057902,
+      "grad_norm": 0.32662689685821533,
+      "learning_rate": 8.280017827314854e-05,
+      "loss": 0.0098,
+      "step": 9260
+    },
+    {
+      "epoch": 0.5399895147667035,
+      "grad_norm": 0.1859537810087204,
+      "learning_rate": 8.2758559379718e-05,
+      "loss": 0.0138,
+      "step": 9270
+    },
+    {
+      "epoch": 0.5405720277276169,
+      "grad_norm": 0.3454076051712036,
+      "learning_rate": 8.271690068154887e-05,
+      "loss": 0.0105,
+      "step": 9280
+    },
+    {
+      "epoch": 0.5411545406885303,
+      "grad_norm": 0.25456947088241577,
+      "learning_rate": 8.26752022292604e-05,
+      "loss": 0.0093,
+      "step": 9290
+    },
+    {
+      "epoch": 0.5417370536494437,
+      "grad_norm": 0.3912484049797058,
+      "learning_rate": 8.263346407352018e-05,
+      "loss": 0.0113,
+      "step": 9300
+    },
+    {
+      "epoch": 0.5423195666103571,
+      "grad_norm": 0.289750337600708,
+      "learning_rate": 8.259168626504395e-05,
+      "loss": 0.0113,
+      "step": 9310
+    },
+    {
+      "epoch": 0.5429020795712705,
+      "grad_norm": 0.21054647862911224,
+      "learning_rate": 8.254986885459572e-05,
+      "loss": 0.0113,
+      "step": 9320
+    },
+    {
+      "epoch": 0.5434845925321838,
+      "grad_norm": 0.2637708783149719,
+      "learning_rate": 8.250801189298757e-05,
+      "loss": 0.0115,
+      "step": 9330
+    },
+    {
+      "epoch": 0.5440671054930972,
+      "grad_norm": 0.27710995078086853,
+      "learning_rate": 8.246611543107967e-05,
+      "loss": 0.0113,
+      "step": 9340
+    },
+    {
+      "epoch": 0.5446496184540106,
+      "grad_norm": 0.2379639893770218,
+      "learning_rate": 8.242417951978018e-05,
+      "loss": 0.0098,
+      "step": 9350
+    },
+    {
+      "epoch": 0.545232131414924,
+      "grad_norm": 0.2423040270805359,
+      "learning_rate": 8.238220421004517e-05,
+      "loss": 0.0124,
+      "step": 9360
+    },
+    {
+      "epoch": 0.5458146443758374,
+      "grad_norm": 0.31561440229415894,
+      "learning_rate": 8.234018955287859e-05,
+      "loss": 0.0177,
+      "step": 9370
+    },
+    {
+      "epoch": 0.5463971573367508,
+      "grad_norm": 0.4590918719768524,
+      "learning_rate": 8.229813559933224e-05,
+      "loss": 0.0155,
+      "step": 9380
+    },
+    {
+      "epoch": 0.5469796702976641,
+      "grad_norm": 0.40657955408096313,
+      "learning_rate": 8.22560424005056e-05,
+      "loss": 0.0149,
+      "step": 9390
+    },
+    {
+      "epoch": 0.5475621832585775,
+      "grad_norm": 0.3211154043674469,
+      "learning_rate": 8.22139100075459e-05,
+      "loss": 0.0092,
+      "step": 9400
+    },
+    {
+      "epoch": 0.5481446962194909,
+      "grad_norm": 0.2968524694442749,
+      "learning_rate": 8.217173847164799e-05,
+      "loss": 0.0114,
+      "step": 9410
+    },
+    {
+      "epoch": 0.5487272091804043,
+      "grad_norm": 0.3226020336151123,
+      "learning_rate": 8.212952784405422e-05,
+      "loss": 0.0115,
+      "step": 9420
+    },
+    {
+      "epoch": 0.5493097221413177,
+      "grad_norm": 0.250797837972641,
+      "learning_rate": 8.208727817605451e-05,
+      "loss": 0.0097,
+      "step": 9430
+    },
+    {
+      "epoch": 0.5498922351022311,
+      "grad_norm": 0.2678021788597107,
+      "learning_rate": 8.204498951898618e-05,
+      "loss": 0.0134,
+      "step": 9440
+    },
+    {
+      "epoch": 0.5504747480631444,
+      "grad_norm": 0.1832326352596283,
+      "learning_rate": 8.200266192423395e-05,
+      "loss": 0.009,
+      "step": 9450
+    },
+    {
+      "epoch": 0.5510572610240578,
+      "grad_norm": 0.3206072449684143,
+      "learning_rate": 8.196029544322981e-05,
+      "loss": 0.0112,
+      "step": 9460
+    },
+    {
+      "epoch": 0.5516397739849712,
+      "grad_norm": 0.17319335043430328,
+      "learning_rate": 8.191789012745305e-05,
+      "loss": 0.0106,
+      "step": 9470
+    },
+    {
+      "epoch": 0.5522222869458846,
+      "grad_norm": 0.2627083659172058,
+      "learning_rate": 8.187544602843014e-05,
+      "loss": 0.009,
+      "step": 9480
+    },
+    {
+      "epoch": 0.5528047999067979,
+      "grad_norm": 0.2164127081632614,
+      "learning_rate": 8.183296319773466e-05,
+      "loss": 0.0112,
+      "step": 9490
+    },
+    {
+      "epoch": 0.5533873128677113,
+      "grad_norm": 0.41261550784111023,
+      "learning_rate": 8.179044168698721e-05,
+      "loss": 0.0136,
+      "step": 9500
+    },
+    {
+      "epoch": 0.5539698258286246,
+      "grad_norm": 0.29752349853515625,
+      "learning_rate": 8.174788154785548e-05,
+      "loss": 0.011,
+      "step": 9510
+    },
+    {
+      "epoch": 0.554552338789538,
+      "grad_norm": 0.32278183102607727,
+      "learning_rate": 8.170528283205403e-05,
+      "loss": 0.0134,
+      "step": 9520
+    },
+    {
+      "epoch": 0.5551348517504514,
+      "grad_norm": 0.22981145977973938,
+      "learning_rate": 8.166264559134435e-05,
+      "loss": 0.0114,
+      "step": 9530
+    },
+    {
+      "epoch": 0.5557173647113648,
+      "grad_norm": 0.3100016415119171,
+      "learning_rate": 8.161996987753467e-05,
+      "loss": 0.0139,
+      "step": 9540
+    },
+    {
+      "epoch": 0.5562998776722782,
+      "grad_norm": 0.3087938725948334,
+      "learning_rate": 8.157725574248e-05,
+      "loss": 0.0135,
+      "step": 9550
+    },
+    {
+      "epoch": 0.5568823906331916,
+      "grad_norm": 0.21534736454486847,
+      "learning_rate": 8.153450323808209e-05,
+      "loss": 0.0109,
+      "step": 9560
+    },
+    {
+      "epoch": 0.5574649035941049,
+      "grad_norm": 0.38279926776885986,
+      "learning_rate": 8.149171241628923e-05,
+      "loss": 0.0126,
+      "step": 9570
+    },
+    {
+      "epoch": 0.5580474165550183,
+      "grad_norm": 0.3262600600719452,
+      "learning_rate": 8.144888332909631e-05,
+      "loss": 0.0123,
+      "step": 9580
+    },
+    {
+      "epoch": 0.5586299295159317,
+      "grad_norm": 0.17854773998260498,
+      "learning_rate": 8.14060160285447e-05,
+      "loss": 0.0083,
+      "step": 9590
+    },
+    {
+      "epoch": 0.5592124424768451,
+      "grad_norm": 0.2897261083126068,
+      "learning_rate": 8.136311056672223e-05,
+      "loss": 0.0123,
+      "step": 9600
+    },
+    {
+      "epoch": 0.5597949554377585,
+      "grad_norm": 0.42957374453544617,
+      "learning_rate": 8.132016699576308e-05,
+      "loss": 0.012,
+      "step": 9610
+    },
+    {
+      "epoch": 0.5603774683986719,
+      "grad_norm": 0.4951796233654022,
+      "learning_rate": 8.127718536784771e-05,
+      "loss": 0.0115,
+      "step": 9620
+    },
+    {
+      "epoch": 0.5609599813595852,
+      "grad_norm": 0.44225791096687317,
+      "learning_rate": 8.12341657352029e-05,
+      "loss": 0.0136,
+      "step": 9630
+    },
+    {
+      "epoch": 0.5615424943204986,
+      "grad_norm": 0.14812079071998596,
+      "learning_rate": 8.119110815010151e-05,
+      "loss": 0.0087,
+      "step": 9640
+    },
+    {
+      "epoch": 0.562125007281412,
+      "grad_norm": 0.46010565757751465,
+      "learning_rate": 8.114801266486261e-05,
+      "loss": 0.0136,
+      "step": 9650
+    },
+    {
+      "epoch": 0.5627075202423254,
+      "grad_norm": 0.48483937978744507,
+      "learning_rate": 8.110487933185122e-05,
+      "loss": 0.0107,
+      "step": 9660
+    },
+    {
+      "epoch": 0.5632900332032388,
+      "grad_norm": 0.27445095777511597,
+      "learning_rate": 8.106170820347847e-05,
+      "loss": 0.0112,
+      "step": 9670
+    },
+    {
+      "epoch": 0.5638725461641522,
+      "grad_norm": 0.4026414752006531,
+      "learning_rate": 8.101849933220132e-05,
+      "loss": 0.0126,
+      "step": 9680
+    },
+    {
+      "epoch": 0.5644550591250656,
+      "grad_norm": 0.24285726249217987,
+      "learning_rate": 8.097525277052264e-05,
+      "loss": 0.009,
+      "step": 9690
+    },
+    {
+      "epoch": 0.5650375720859789,
+      "grad_norm": 0.35967713594436646,
+      "learning_rate": 8.093196857099106e-05,
+      "loss": 0.0092,
+      "step": 9700
+    },
+    {
+      "epoch": 0.5656200850468923,
+      "grad_norm": 0.1624266505241394,
+      "learning_rate": 8.088864678620096e-05,
+      "loss": 0.0116,
+      "step": 9710
+    },
+    {
+      "epoch": 0.5662025980078057,
+      "grad_norm": 0.19601480662822723,
+      "learning_rate": 8.084528746879243e-05,
+      "loss": 0.0108,
+      "step": 9720
+    },
+    {
+      "epoch": 0.5667851109687191,
+      "grad_norm": 0.30270540714263916,
+      "learning_rate": 8.080189067145108e-05,
+      "loss": 0.011,
+      "step": 9730
+    },
+    {
+      "epoch": 0.5673676239296325,
+      "grad_norm": 0.16536946594715118,
+      "learning_rate": 8.075845644690813e-05,
+      "loss": 0.0104,
+      "step": 9740
+    },
+    {
+      "epoch": 0.5679501368905459,
+      "grad_norm": 0.29546692967414856,
+      "learning_rate": 8.07149848479403e-05,
+      "loss": 0.0122,
+      "step": 9750
+    },
+    {
+      "epoch": 0.5685326498514592,
+      "grad_norm": 0.3476848304271698,
+      "learning_rate": 8.067147592736962e-05,
+      "loss": 0.0098,
+      "step": 9760
+    },
+    {
+      "epoch": 0.5691151628123726,
+      "grad_norm": 0.16247175633907318,
+      "learning_rate": 8.062792973806357e-05,
+      "loss": 0.0087,
+      "step": 9770
+    },
+    {
+      "epoch": 0.569697675773286,
+      "grad_norm": 0.3933895230293274,
+      "learning_rate": 8.058434633293485e-05,
+      "loss": 0.0086,
+      "step": 9780
+    },
+    {
+      "epoch": 0.5702801887341994,
+      "grad_norm": 0.2233133465051651,
+      "learning_rate": 8.054072576494143e-05,
+      "loss": 0.01,
+      "step": 9790
+    },
+    {
+      "epoch": 0.5708627016951127,
+      "grad_norm": 0.3780955970287323,
+      "learning_rate": 8.04970680870864e-05,
+      "loss": 0.0076,
+      "step": 9800
+    },
+    {
+      "epoch": 0.571445214656026,
+      "grad_norm": 0.30459094047546387,
+      "learning_rate": 8.045337335241793e-05,
+      "loss": 0.0109,
+      "step": 9810
+    },
+    {
+      "epoch": 0.5720277276169394,
+      "grad_norm": 0.4079815447330475,
+      "learning_rate": 8.040964161402932e-05,
+      "loss": 0.0093,
+      "step": 9820
+    },
+    {
+      "epoch": 0.5726102405778528,
+      "grad_norm": 0.35216766595840454,
+      "learning_rate": 8.036587292505869e-05,
+      "loss": 0.015,
+      "step": 9830
+    },
+    {
+      "epoch": 0.5731927535387662,
+      "grad_norm": 0.483693927526474,
+      "learning_rate": 8.032206733868912e-05,
+      "loss": 0.0133,
+      "step": 9840
+    },
+    {
+      "epoch": 0.5737752664996796,
+      "grad_norm": 0.4553718864917755,
+      "learning_rate": 8.027822490814858e-05,
+      "loss": 0.0159,
+      "step": 9850
+    },
+    {
+      "epoch": 0.574357779460593,
+      "grad_norm": 0.31747058033943176,
+      "learning_rate": 8.023434568670972e-05,
+      "loss": 0.0147,
+      "step": 9860
+    },
+    {
+      "epoch": 0.5749402924215063,
+      "grad_norm": 0.30035489797592163,
+      "learning_rate": 8.019042972768992e-05,
+      "loss": 0.0104,
+      "step": 9870
+    },
+    {
+      "epoch": 0.5755228053824197,
+      "grad_norm": 0.2849234640598297,
+      "learning_rate": 8.014647708445124e-05,
+      "loss": 0.0109,
+      "step": 9880
+    },
+    {
+      "epoch": 0.5761053183433331,
+      "grad_norm": 0.4712446331977844,
+      "learning_rate": 8.010248781040027e-05,
+      "loss": 0.0132,
+      "step": 9890
+    },
+    {
+      "epoch": 0.5766878313042465,
+      "grad_norm": 0.38158559799194336,
+      "learning_rate": 8.005846195898814e-05,
+      "loss": 0.0108,
+      "step": 9900
+    },
+    {
+      "epoch": 0.5772703442651599,
+      "grad_norm": 0.3007536828517914,
+      "learning_rate": 8.00143995837104e-05,
+      "loss": 0.0112,
+      "step": 9910
+    },
+    {
+      "epoch": 0.5778528572260733,
+      "grad_norm": 0.2956334948539734,
+      "learning_rate": 7.997030073810698e-05,
+      "loss": 0.0114,
+      "step": 9920
+    },
+    {
+      "epoch": 0.5784353701869867,
+      "grad_norm": 0.18578597903251648,
+      "learning_rate": 7.992616547576217e-05,
+      "loss": 0.0084,
+      "step": 9930
+    },
+    {
+      "epoch": 0.5790178831479,
+      "grad_norm": 0.26690977811813354,
+      "learning_rate": 7.988199385030445e-05,
+      "loss": 0.0128,
+      "step": 9940
+    },
+    {
+      "epoch": 0.5796003961088134,
+      "grad_norm": 0.3819132447242737,
+      "learning_rate": 7.98377859154065e-05,
+      "loss": 0.0129,
+      "step": 9950
+    },
+    {
+      "epoch": 0.5801829090697268,
+      "grad_norm": 0.2008133828639984,
+      "learning_rate": 7.979354172478516e-05,
+      "loss": 0.0088,
+      "step": 9960
+    },
+    {
+      "epoch": 0.5807654220306402,
+      "grad_norm": 0.20811396837234497,
+      "learning_rate": 7.974926133220127e-05,
+      "loss": 0.0099,
+      "step": 9970
+    },
+    {
+      "epoch": 0.5813479349915536,
+      "grad_norm": 0.23433077335357666,
+      "learning_rate": 7.970494479145967e-05,
+      "loss": 0.0098,
+      "step": 9980
+    },
+    {
+      "epoch": 0.581930447952467,
+      "grad_norm": 0.4400034248828888,
+      "learning_rate": 7.966059215640919e-05,
+      "loss": 0.0127,
+      "step": 9990
+    },
+    {
+      "epoch": 0.5825129609133803,
+      "grad_norm": 0.2626042366027832,
+      "learning_rate": 7.961620348094241e-05,
+      "loss": 0.0115,
+      "step": 10000
+    },
+    {
+      "epoch": 0.5830954738742937,
+      "grad_norm": 0.39748772978782654,
+      "learning_rate": 7.957177881899577e-05,
+      "loss": 0.0109,
+      "step": 10010
+    },
+    {
+      "epoch": 0.5836779868352071,
+      "grad_norm": 0.23363058269023895,
+      "learning_rate": 7.952731822454944e-05,
+      "loss": 0.0117,
+      "step": 10020
+    },
+    {
+      "epoch": 0.5842604997961205,
+      "grad_norm": 0.3626881241798401,
+      "learning_rate": 7.948282175162722e-05,
+      "loss": 0.0118,
+      "step": 10030
+    },
+    {
+      "epoch": 0.5848430127570339,
+      "grad_norm": 0.26454007625579834,
+      "learning_rate": 7.943828945429652e-05,
+      "loss": 0.0134,
+      "step": 10040
+    },
+    {
+      "epoch": 0.5854255257179473,
+      "grad_norm": 0.22416579723358154,
+      "learning_rate": 7.939372138666827e-05,
+      "loss": 0.0086,
+      "step": 10050
+    },
+    {
+      "epoch": 0.5860080386788606,
+      "grad_norm": 0.18413110077381134,
+      "learning_rate": 7.934911760289692e-05,
+      "loss": 0.0106,
+      "step": 10060
+    },
+    {
+      "epoch": 0.586590551639774,
+      "grad_norm": 0.2637045681476593,
+      "learning_rate": 7.930447815718022e-05,
+      "loss": 0.0113,
+      "step": 10070
+    },
+    {
+      "epoch": 0.5871730646006874,
+      "grad_norm": 0.16207945346832275,
+      "learning_rate": 7.925980310375934e-05,
+      "loss": 0.0096,
+      "step": 10080
+    },
+    {
+      "epoch": 0.5877555775616008,
+      "grad_norm": 0.2489597648382187,
+      "learning_rate": 7.921509249691865e-05,
+      "loss": 0.0093,
+      "step": 10090
+    },
+    {
+      "epoch": 0.5883380905225142,
+      "grad_norm": 0.29985642433166504,
+      "learning_rate": 7.917034639098579e-05,
+      "loss": 0.0094,
+      "step": 10100
+    },
+    {
+      "epoch": 0.5889206034834275,
+      "grad_norm": 0.2591872215270996,
+      "learning_rate": 7.912556484033146e-05,
+      "loss": 0.0103,
+      "step": 10110
+    },
+    {
+      "epoch": 0.5895031164443408,
+      "grad_norm": 0.3224642872810364,
+      "learning_rate": 7.908074789936952e-05,
+      "loss": 0.0112,
+      "step": 10120
+    },
+    {
+      "epoch": 0.5900856294052542,
+      "grad_norm": 0.22806353867053986,
+      "learning_rate": 7.903589562255673e-05,
+      "loss": 0.0089,
+      "step": 10130
+    },
+    {
+      "epoch": 0.5906681423661676,
+      "grad_norm": 0.44386065006256104,
+      "learning_rate": 7.899100806439285e-05,
+      "loss": 0.0098,
+      "step": 10140
+    },
+    {
+      "epoch": 0.591250655327081,
+      "grad_norm": 0.5068503022193909,
+      "learning_rate": 7.894608527942049e-05,
+      "loss": 0.0094,
+      "step": 10150
+    },
+    {
+      "epoch": 0.5918331682879944,
+      "grad_norm": 0.4271147549152374,
+      "learning_rate": 7.89011273222251e-05,
+      "loss": 0.0099,
+      "step": 10160
+    },
+    {
+      "epoch": 0.5924156812489078,
+      "grad_norm": 0.2077079713344574,
+      "learning_rate": 7.885613424743482e-05,
+      "loss": 0.0102,
+      "step": 10170
+    },
+    {
+      "epoch": 0.5929981942098211,
+      "grad_norm": 0.2697822153568268,
+      "learning_rate": 7.881110610972044e-05,
+      "loss": 0.0109,
+      "step": 10180
+    },
+    {
+      "epoch": 0.5935807071707345,
+      "grad_norm": 0.17493903636932373,
+      "learning_rate": 7.876604296379544e-05,
+      "loss": 0.0089,
+      "step": 10190
+    },
+    {
+      "epoch": 0.5941632201316479,
+      "grad_norm": 0.19017846882343292,
+      "learning_rate": 7.872094486441579e-05,
+      "loss": 0.0096,
+      "step": 10200
+    },
+    {
+      "epoch": 0.5947457330925613,
+      "grad_norm": 0.1379752904176712,
+      "learning_rate": 7.867581186637991e-05,
+      "loss": 0.0106,
+      "step": 10210
+    },
+    {
+      "epoch": 0.5953282460534747,
+      "grad_norm": 0.23306509852409363,
+      "learning_rate": 7.863064402452866e-05,
+      "loss": 0.0099,
+      "step": 10220
+    },
+    {
+      "epoch": 0.5959107590143881,
+      "grad_norm": 0.23990528285503387,
+      "learning_rate": 7.858544139374524e-05,
+      "loss": 0.0093,
+      "step": 10230
+    },
+    {
+      "epoch": 0.5964932719753014,
+      "grad_norm": 0.23690129816532135,
+      "learning_rate": 7.854020402895508e-05,
+      "loss": 0.0091,
+      "step": 10240
+    },
+    {
+      "epoch": 0.5970757849362148,
+      "grad_norm": 0.408820778131485,
+      "learning_rate": 7.849493198512587e-05,
+      "loss": 0.0095,
+      "step": 10250
+    },
+    {
+      "epoch": 0.5976582978971282,
+      "grad_norm": 0.3510490655899048,
+      "learning_rate": 7.84496253172674e-05,
+      "loss": 0.0101,
+      "step": 10260
+    },
+    {
+      "epoch": 0.5982408108580416,
+      "grad_norm": 0.22155600786209106,
+      "learning_rate": 7.840428408043155e-05,
+      "loss": 0.0098,
+      "step": 10270
+    },
+    {
+      "epoch": 0.598823323818955,
+      "grad_norm": 0.28184619545936584,
+      "learning_rate": 7.835890832971218e-05,
+      "loss": 0.0169,
+      "step": 10280
+    },
+    {
+      "epoch": 0.5994058367798684,
+      "grad_norm": 0.37654370069503784,
+      "learning_rate": 7.831349812024512e-05,
+      "loss": 0.0121,
+      "step": 10290
+    },
+    {
+      "epoch": 0.5999883497407817,
+      "grad_norm": 0.30494412779808044,
+      "learning_rate": 7.826805350720807e-05,
+      "loss": 0.0096,
+      "step": 10300
+    },
+    {
+      "epoch": 0.6005708627016951,
+      "grad_norm": 0.16471631824970245,
+      "learning_rate": 7.822257454582049e-05,
+      "loss": 0.0098,
+      "step": 10310
+    },
+    {
+      "epoch": 0.6011533756626085,
+      "grad_norm": 0.2342802733182907,
+      "learning_rate": 7.817706129134363e-05,
+      "loss": 0.0113,
+      "step": 10320
+    },
+    {
+      "epoch": 0.6017358886235219,
+      "grad_norm": 0.2859920263290405,
+      "learning_rate": 7.813151379908036e-05,
+      "loss": 0.0083,
+      "step": 10330
+    },
+    {
+      "epoch": 0.6023184015844353,
+      "grad_norm": 0.2043689489364624,
+      "learning_rate": 7.808593212437523e-05,
+      "loss": 0.0083,
+      "step": 10340
+    },
+    {
+      "epoch": 0.6029009145453487,
+      "grad_norm": 0.1721859723329544,
+      "learning_rate": 7.804031632261421e-05,
+      "loss": 0.0086,
+      "step": 10350
+    },
+    {
+      "epoch": 0.603483427506262,
+      "grad_norm": 0.25944796204566956,
+      "learning_rate": 7.799466644922484e-05,
+      "loss": 0.0079,
+      "step": 10360
+    },
+    {
+      "epoch": 0.6040659404671754,
+      "grad_norm": 0.5207556486129761,
+      "learning_rate": 7.794898255967602e-05,
+      "loss": 0.011,
+      "step": 10370
+    },
+    {
+      "epoch": 0.6046484534280888,
+      "grad_norm": 0.42015981674194336,
+      "learning_rate": 7.790326470947795e-05,
+      "loss": 0.0115,
+      "step": 10380
+    },
+    {
+      "epoch": 0.6052309663890022,
+      "grad_norm": 0.22684963047504425,
+      "learning_rate": 7.785751295418217e-05,
+      "loss": 0.0096,
+      "step": 10390
+    },
+    {
+      "epoch": 0.6058134793499156,
+      "grad_norm": 0.2202410250902176,
+      "learning_rate": 7.781172734938136e-05,
+      "loss": 0.0116,
+      "step": 10400
+    },
+    {
+      "epoch": 0.606395992310829,
+      "grad_norm": 0.3481963574886322,
+      "learning_rate": 7.776590795070933e-05,
+      "loss": 0.0107,
+      "step": 10410
+    },
+    {
+      "epoch": 0.6069785052717422,
+      "grad_norm": 0.25396406650543213,
+      "learning_rate": 7.772005481384099e-05,
+      "loss": 0.0099,
+      "step": 10420
+    },
+    {
+      "epoch": 0.6075610182326556,
+      "grad_norm": 0.35001271963119507,
+      "learning_rate": 7.767416799449223e-05,
+      "loss": 0.0089,
+      "step": 10430
+    },
+    {
+      "epoch": 0.608143531193569,
+      "grad_norm": 0.24520301818847656,
+      "learning_rate": 7.762824754841984e-05,
+      "loss": 0.0107,
+      "step": 10440
+    },
+    {
+      "epoch": 0.6087260441544824,
+      "grad_norm": 0.39444899559020996,
+      "learning_rate": 7.758229353142152e-05,
+      "loss": 0.0103,
+      "step": 10450
+    },
+    {
+      "epoch": 0.6093085571153958,
+      "grad_norm": 0.2335900366306305,
+      "learning_rate": 7.753630599933572e-05,
+      "loss": 0.0097,
+      "step": 10460
+    },
+    {
+      "epoch": 0.6098910700763092,
+      "grad_norm": 0.28714701533317566,
+      "learning_rate": 7.749028500804161e-05,
+      "loss": 0.0095,
+      "step": 10470
+    },
+    {
+      "epoch": 0.6104735830372225,
+      "grad_norm": 0.29769057035446167,
+      "learning_rate": 7.744423061345906e-05,
+      "loss": 0.0085,
+      "step": 10480
+    },
+    {
+      "epoch": 0.6110560959981359,
+      "grad_norm": 0.27050989866256714,
+      "learning_rate": 7.73981428715485e-05,
+      "loss": 0.0095,
+      "step": 10490
+    },
+    {
+      "epoch": 0.6116386089590493,
+      "grad_norm": 0.2549913823604584,
+      "learning_rate": 7.735202183831085e-05,
+      "loss": 0.0096,
+      "step": 10500
+    },
+    {
+      "epoch": 0.6122211219199627,
+      "grad_norm": 0.24300307035446167,
+      "learning_rate": 7.730586756978757e-05,
+      "loss": 0.0126,
+      "step": 10510
+    },
+    {
+      "epoch": 0.6128036348808761,
+      "grad_norm": 0.1996365487575531,
+      "learning_rate": 7.72596801220604e-05,
+      "loss": 0.0096,
+      "step": 10520
+    },
+    {
+      "epoch": 0.6133861478417895,
+      "grad_norm": 0.32112377882003784,
+      "learning_rate": 7.721345955125147e-05,
+      "loss": 0.0123,
+      "step": 10530
+    },
+    {
+      "epoch": 0.6139686608027028,
+      "grad_norm": 0.23069436848163605,
+      "learning_rate": 7.71672059135231e-05,
+      "loss": 0.0089,
+      "step": 10540
+    },
+    {
+      "epoch": 0.6145511737636162,
+      "grad_norm": 0.17319975793361664,
+      "learning_rate": 7.712091926507787e-05,
+      "loss": 0.007,
+      "step": 10550
+    },
+    {
+      "epoch": 0.6151336867245296,
+      "grad_norm": 0.20370422303676605,
+      "learning_rate": 7.70745996621584e-05,
+      "loss": 0.0094,
+      "step": 10560
+    },
+    {
+      "epoch": 0.615716199685443,
+      "grad_norm": 0.2806628942489624,
+      "learning_rate": 7.702824716104735e-05,
+      "loss": 0.0095,
+      "step": 10570
+    },
+    {
+      "epoch": 0.6162987126463564,
+      "grad_norm": 0.3823358416557312,
+      "learning_rate": 7.698186181806743e-05,
+      "loss": 0.0085,
+      "step": 10580
+    },
+    {
+      "epoch": 0.6168812256072698,
+      "grad_norm": 0.3381858170032501,
+      "learning_rate": 7.693544368958116e-05,
+      "loss": 0.0095,
+      "step": 10590
+    },
+    {
+      "epoch": 0.6174637385681832,
+      "grad_norm": 0.24725757539272308,
+      "learning_rate": 7.688899283199096e-05,
+      "loss": 0.0101,
+      "step": 10600
+    },
+    {
+      "epoch": 0.6180462515290965,
+      "grad_norm": 0.23433803021907806,
+      "learning_rate": 7.684250930173901e-05,
+      "loss": 0.0079,
+      "step": 10610
+    },
+    {
+      "epoch": 0.6186287644900099,
+      "grad_norm": 0.2003704458475113,
+      "learning_rate": 7.679599315530718e-05,
+      "loss": 0.0085,
+      "step": 10620
+    },
+    {
+      "epoch": 0.6192112774509233,
+      "grad_norm": 0.26989614963531494,
+      "learning_rate": 7.674944444921695e-05,
+      "loss": 0.0104,
+      "step": 10630
+    },
+    {
+      "epoch": 0.6197937904118367,
+      "grad_norm": 0.22777387499809265,
+      "learning_rate": 7.670286324002944e-05,
+      "loss": 0.0081,
+      "step": 10640
+    },
+    {
+      "epoch": 0.6203763033727501,
+      "grad_norm": 0.29325246810913086,
+      "learning_rate": 7.665624958434514e-05,
+      "loss": 0.0081,
+      "step": 10650
+    },
+    {
+      "epoch": 0.6209588163336635,
+      "grad_norm": 0.29472509026527405,
+      "learning_rate": 7.66096035388041e-05,
+      "loss": 0.0098,
+      "step": 10660
+    },
+    {
+      "epoch": 0.6215413292945768,
+      "grad_norm": 0.25304609537124634,
+      "learning_rate": 7.656292516008563e-05,
+      "loss": 0.0069,
+      "step": 10670
+    },
+    {
+      "epoch": 0.6221238422554902,
+      "grad_norm": 0.3088533282279968,
+      "learning_rate": 7.651621450490837e-05,
+      "loss": 0.0105,
+      "step": 10680
+    },
+    {
+      "epoch": 0.6227063552164036,
+      "grad_norm": 0.22526845335960388,
+      "learning_rate": 7.646947163003017e-05,
+      "loss": 0.0082,
+      "step": 10690
+    },
+    {
+      "epoch": 0.623288868177317,
+      "grad_norm": 0.27041852474212646,
+      "learning_rate": 7.642269659224803e-05,
+      "loss": 0.0088,
+      "step": 10700
+    },
+    {
+      "epoch": 0.6238713811382304,
+      "grad_norm": 0.3110310435295105,
+      "learning_rate": 7.637588944839803e-05,
+      "loss": 0.0126,
+      "step": 10710
+    },
+    {
+      "epoch": 0.6244538940991438,
+      "grad_norm": 0.25566214323043823,
+      "learning_rate": 7.632905025535528e-05,
+      "loss": 0.008,
+      "step": 10720
+    },
+    {
+      "epoch": 0.625036407060057,
+      "grad_norm": 0.33295702934265137,
+      "learning_rate": 7.628217907003378e-05,
+      "loss": 0.0078,
+      "step": 10730
+    },
+    {
+      "epoch": 0.6256189200209704,
+      "grad_norm": 0.3524506986141205,
+      "learning_rate": 7.623527594938649e-05,
+      "loss": 0.0094,
+      "step": 10740
+    },
+    {
+      "epoch": 0.6262014329818838,
+      "grad_norm": 0.34373077750205994,
+      "learning_rate": 7.618834095040509e-05,
+      "loss": 0.0087,
+      "step": 10750
+    },
+    {
+      "epoch": 0.6267839459427972,
+      "grad_norm": 0.26415351033210754,
+      "learning_rate": 7.614137413012e-05,
+      "loss": 0.0099,
+      "step": 10760
+    },
+    {
+      "epoch": 0.6273664589037106,
+      "grad_norm": 0.28037315607070923,
+      "learning_rate": 7.609437554560042e-05,
+      "loss": 0.0092,
+      "step": 10770
+    },
+    {
+      "epoch": 0.627948971864624,
+      "grad_norm": 0.15280218422412872,
+      "learning_rate": 7.604734525395398e-05,
+      "loss": 0.0093,
+      "step": 10780
+    },
+    {
+      "epoch": 0.6285314848255373,
+      "grad_norm": 0.17172874510288239,
+      "learning_rate": 7.600028331232699e-05,
+      "loss": 0.0109,
+      "step": 10790
+    },
+    {
+      "epoch": 0.6291139977864507,
+      "grad_norm": 0.2339518666267395,
+      "learning_rate": 7.595318977790408e-05,
+      "loss": 0.008,
+      "step": 10800
+    },
+    {
+      "epoch": 0.6296965107473641,
+      "grad_norm": 0.2171650230884552,
+      "learning_rate": 7.590606470790836e-05,
+      "loss": 0.009,
+      "step": 10810
+    },
+    {
+      "epoch": 0.6302790237082775,
+      "grad_norm": 0.2459641844034195,
+      "learning_rate": 7.585890815960125e-05,
+      "loss": 0.0114,
+      "step": 10820
+    },
+    {
+      "epoch": 0.6308615366691909,
+      "grad_norm": 0.24623656272888184,
+      "learning_rate": 7.581172019028238e-05,
+      "loss": 0.0102,
+      "step": 10830
+    },
+    {
+      "epoch": 0.6314440496301043,
+      "grad_norm": 0.3801115155220032,
+      "learning_rate": 7.576450085728959e-05,
+      "loss": 0.0099,
+      "step": 10840
+    },
+    {
+      "epoch": 0.6320265625910176,
+      "grad_norm": 0.24757038056850433,
+      "learning_rate": 7.571725021799885e-05,
+      "loss": 0.0105,
+      "step": 10850
+    },
+    {
+      "epoch": 0.632609075551931,
+      "grad_norm": 0.20195166766643524,
+      "learning_rate": 7.566996832982409e-05,
+      "loss": 0.0103,
+      "step": 10860
+    },
+    {
+      "epoch": 0.6331915885128444,
+      "grad_norm": 0.20696143805980682,
+      "learning_rate": 7.56226552502173e-05,
+      "loss": 0.0081,
+      "step": 10870
+    },
+    {
+      "epoch": 0.6337741014737578,
+      "grad_norm": 0.34664294123649597,
+      "learning_rate": 7.557531103666833e-05,
+      "loss": 0.0092,
+      "step": 10880
+    },
+    {
+      "epoch": 0.6343566144346712,
+      "grad_norm": 0.3256179392337799,
+      "learning_rate": 7.552793574670485e-05,
+      "loss": 0.0092,
+      "step": 10890
+    },
+    {
+      "epoch": 0.6349391273955846,
+      "grad_norm": 0.20562410354614258,
+      "learning_rate": 7.54805294378923e-05,
+      "loss": 0.0091,
+      "step": 10900
+    },
+    {
+      "epoch": 0.6355216403564979,
+      "grad_norm": 0.27549681067466736,
+      "learning_rate": 7.543309216783384e-05,
+      "loss": 0.0083,
+      "step": 10910
+    },
+    {
+      "epoch": 0.6361041533174113,
+      "grad_norm": 0.19887396693229675,
+      "learning_rate": 7.53856239941702e-05,
+      "loss": 0.0083,
+      "step": 10920
+    },
+    {
+      "epoch": 0.6366866662783247,
+      "grad_norm": 0.2790752351284027,
+      "learning_rate": 7.533812497457972e-05,
+      "loss": 0.0103,
+      "step": 10930
+    },
+    {
+      "epoch": 0.6372691792392381,
+      "grad_norm": 0.1928553581237793,
+      "learning_rate": 7.529059516677814e-05,
+      "loss": 0.0082,
+      "step": 10940
+    },
+    {
+      "epoch": 0.6378516922001515,
+      "grad_norm": 0.24051916599273682,
+      "learning_rate": 7.524303462851872e-05,
+      "loss": 0.0082,
+      "step": 10950
+    },
+    {
+      "epoch": 0.6384342051610649,
+      "grad_norm": 0.25798922777175903,
+      "learning_rate": 7.519544341759192e-05,
+      "loss": 0.0075,
+      "step": 10960
+    },
+    {
+      "epoch": 0.6390167181219782,
+      "grad_norm": 0.17289404571056366,
+      "learning_rate": 7.514782159182562e-05,
+      "loss": 0.0083,
+      "step": 10970
+    },
+    {
+      "epoch": 0.6395992310828916,
+      "grad_norm": 0.26216280460357666,
+      "learning_rate": 7.510016920908481e-05,
+      "loss": 0.009,
+      "step": 10980
+    },
+    {
+      "epoch": 0.640181744043805,
+      "grad_norm": 0.30090492963790894,
+      "learning_rate": 7.505248632727158e-05,
+      "loss": 0.0107,
+      "step": 10990
+    },
+    {
+      "epoch": 0.6407642570047184,
+      "grad_norm": 0.28938353061676025,
+      "learning_rate": 7.50047730043252e-05,
+      "loss": 0.0129,
+      "step": 11000
+    },
+    {
+      "epoch": 0.6413467699656318,
+      "grad_norm": 0.20174451172351837,
+      "learning_rate": 7.495702929822183e-05,
+      "loss": 0.0089,
+      "step": 11010
+    },
+    {
+      "epoch": 0.6419292829265452,
+      "grad_norm": 0.2501292824745178,
+      "learning_rate": 7.490925526697455e-05,
+      "loss": 0.0078,
+      "step": 11020
+    },
+    {
+      "epoch": 0.6425117958874585,
+      "grad_norm": 0.3156457841396332,
+      "learning_rate": 7.486145096863334e-05,
+      "loss": 0.011,
+      "step": 11030
+    },
+    {
+      "epoch": 0.6430943088483719,
+      "grad_norm": 0.20288266241550446,
+      "learning_rate": 7.481361646128491e-05,
+      "loss": 0.0068,
+      "step": 11040
+    },
+    {
+      "epoch": 0.6436768218092852,
+      "grad_norm": 0.19690006971359253,
+      "learning_rate": 7.476575180305271e-05,
+      "loss": 0.0094,
+      "step": 11050
+    },
+    {
+      "epoch": 0.6442593347701986,
+      "grad_norm": 0.5085271000862122,
+      "learning_rate": 7.471785705209682e-05,
+      "loss": 0.0071,
+      "step": 11060
+    },
+    {
+      "epoch": 0.644841847731112,
+      "grad_norm": 0.47579434514045715,
+      "learning_rate": 7.466993226661387e-05,
+      "loss": 0.0103,
+      "step": 11070
+    },
+    {
+      "epoch": 0.6454243606920254,
+      "grad_norm": 0.3698206841945648,
+      "learning_rate": 7.462197750483699e-05,
+      "loss": 0.009,
+      "step": 11080
+    },
+    {
+      "epoch": 0.6460068736529387,
+      "grad_norm": 0.22008930146694183,
+      "learning_rate": 7.457399282503574e-05,
+      "loss": 0.0089,
+      "step": 11090
+    },
+    {
+      "epoch": 0.6465893866138521,
+      "grad_norm": 0.2831665575504303,
+      "learning_rate": 7.452597828551604e-05,
+      "loss": 0.0104,
+      "step": 11100
+    },
+    {
+      "epoch": 0.6471718995747655,
+      "grad_norm": 0.23014068603515625,
+      "learning_rate": 7.447793394462006e-05,
+      "loss": 0.0117,
+      "step": 11110
+    },
+    {
+      "epoch": 0.6477544125356789,
+      "grad_norm": 0.23992767930030823,
+      "learning_rate": 7.442985986072624e-05,
+      "loss": 0.0099,
+      "step": 11120
+    },
+    {
+      "epoch": 0.6483369254965923,
+      "grad_norm": 0.2168709635734558,
+      "learning_rate": 7.438175609224908e-05,
+      "loss": 0.0089,
+      "step": 11130
+    },
+    {
+      "epoch": 0.6489194384575057,
+      "grad_norm": 0.2596808969974518,
+      "learning_rate": 7.433362269763924e-05,
+      "loss": 0.0125,
+      "step": 11140
+    },
+    {
+      "epoch": 0.649501951418419,
+      "grad_norm": 0.2573094964027405,
+      "learning_rate": 7.428545973538329e-05,
+      "loss": 0.0136,
+      "step": 11150
+    },
+    {
+      "epoch": 0.6500844643793324,
+      "grad_norm": 0.2209426760673523,
+      "learning_rate": 7.42372672640038e-05,
+      "loss": 0.011,
+      "step": 11160
+    },
+    {
+      "epoch": 0.6506669773402458,
+      "grad_norm": 0.2567729353904724,
+      "learning_rate": 7.418904534205917e-05,
+      "loss": 0.01,
+      "step": 11170
+    },
+    {
+      "epoch": 0.6512494903011592,
+      "grad_norm": 0.20151248574256897,
+      "learning_rate": 7.414079402814356e-05,
+      "loss": 0.0096,
+      "step": 11180
+    },
+    {
+      "epoch": 0.6518320032620726,
+      "grad_norm": 0.25869089365005493,
+      "learning_rate": 7.409251338088687e-05,
+      "loss": 0.0132,
+      "step": 11190
+    },
+    {
+      "epoch": 0.652414516222986,
+      "grad_norm": 0.3204185366630554,
+      "learning_rate": 7.404420345895467e-05,
+      "loss": 0.0118,
+      "step": 11200
+    },
+    {
+      "epoch": 0.6529970291838993,
+      "grad_norm": 0.3233679533004761,
+      "learning_rate": 7.399586432104804e-05,
+      "loss": 0.0123,
+      "step": 11210
+    },
+    {
+      "epoch": 0.6535795421448127,
+      "grad_norm": 0.20079077780246735,
+      "learning_rate": 7.394749602590358e-05,
+      "loss": 0.0103,
+      "step": 11220
+    },
+    {
+      "epoch": 0.6541620551057261,
+      "grad_norm": 0.28856900334358215,
+      "learning_rate": 7.389909863229336e-05,
+      "loss": 0.0104,
+      "step": 11230
+    },
+    {
+      "epoch": 0.6547445680666395,
+      "grad_norm": 0.38460513949394226,
+      "learning_rate": 7.385067219902477e-05,
+      "loss": 0.0089,
+      "step": 11240
+    },
+    {
+      "epoch": 0.6553270810275529,
+      "grad_norm": 0.28212809562683105,
+      "learning_rate": 7.380221678494049e-05,
+      "loss": 0.0156,
+      "step": 11250
+    },
+    {
+      "epoch": 0.6559095939884663,
+      "grad_norm": 0.22180429100990295,
+      "learning_rate": 7.37537324489184e-05,
+      "loss": 0.0087,
+      "step": 11260
+    },
+    {
+      "epoch": 0.6564921069493797,
+      "grad_norm": 0.2083648145198822,
+      "learning_rate": 7.370521924987155e-05,
+      "loss": 0.0078,
+      "step": 11270
+    },
+    {
+      "epoch": 0.657074619910293,
+      "grad_norm": 0.22062233090400696,
+      "learning_rate": 7.365667724674806e-05,
+      "loss": 0.0074,
+      "step": 11280
+    },
+    {
+      "epoch": 0.6576571328712064,
+      "grad_norm": 0.7330291271209717,
+      "learning_rate": 7.360810649853105e-05,
+      "loss": 0.0118,
+      "step": 11290
+    },
+    {
+      "epoch": 0.6582396458321198,
+      "grad_norm": 0.27491968870162964,
+      "learning_rate": 7.355950706423853e-05,
+      "loss": 0.0119,
+      "step": 11300
+    },
+    {
+      "epoch": 0.6588221587930332,
+      "grad_norm": 0.23102988302707672,
+      "learning_rate": 7.351087900292342e-05,
+      "loss": 0.0116,
+      "step": 11310
+    },
+    {
+      "epoch": 0.6594046717539466,
+      "grad_norm": 0.2202250212430954,
+      "learning_rate": 7.346222237367339e-05,
+      "loss": 0.0084,
+      "step": 11320
+    },
+    {
+      "epoch": 0.65998718471486,
+      "grad_norm": 0.3032495081424713,
+      "learning_rate": 7.341353723561081e-05,
+      "loss": 0.011,
+      "step": 11330
+    },
+    {
+      "epoch": 0.6605696976757733,
+      "grad_norm": 0.3746453523635864,
+      "learning_rate": 7.336482364789277e-05,
+      "loss": 0.0138,
+      "step": 11340
+    },
+    {
+      "epoch": 0.6611522106366867,
+      "grad_norm": 0.3068520724773407,
+      "learning_rate": 7.331608166971082e-05,
+      "loss": 0.0094,
+      "step": 11350
+    },
+    {
+      "epoch": 0.6617347235976,
+      "grad_norm": 0.2291998416185379,
+      "learning_rate": 7.326731136029108e-05,
+      "loss": 0.0089,
+      "step": 11360
+    },
+    {
+      "epoch": 0.6623172365585134,
+      "grad_norm": 0.22003349661827087,
+      "learning_rate": 7.321851277889408e-05,
+      "loss": 0.0132,
+      "step": 11370
+    },
+    {
+      "epoch": 0.6628997495194268,
+      "grad_norm": 0.26796552538871765,
+      "learning_rate": 7.316968598481469e-05,
+      "loss": 0.0084,
+      "step": 11380
+    },
+    {
+      "epoch": 0.6634822624803401,
+      "grad_norm": 0.22081968188285828,
+      "learning_rate": 7.312083103738206e-05,
+      "loss": 0.0086,
+      "step": 11390
+    },
+    {
+      "epoch": 0.6640647754412535,
+      "grad_norm": 0.2831570506095886,
+      "learning_rate": 7.307194799595958e-05,
+      "loss": 0.0076,
+      "step": 11400
+    },
+    {
+      "epoch": 0.6646472884021669,
+      "grad_norm": 0.28629910945892334,
+      "learning_rate": 7.302303691994473e-05,
+      "loss": 0.0099,
+      "step": 11410
+    },
+    {
+      "epoch": 0.6652298013630803,
+      "grad_norm": 0.29212814569473267,
+      "learning_rate": 7.29740978687691e-05,
+      "loss": 0.0107,
+      "step": 11420
+    },
+    {
+      "epoch": 0.6658123143239937,
+      "grad_norm": 0.27409547567367554,
+      "learning_rate": 7.292513090189825e-05,
+      "loss": 0.0101,
+      "step": 11430
+    },
+    {
+      "epoch": 0.6663948272849071,
+      "grad_norm": 0.22072142362594604,
+      "learning_rate": 7.287613607883163e-05,
+      "loss": 0.0082,
+      "step": 11440
+    },
+    {
+      "epoch": 0.6669773402458204,
+      "grad_norm": 0.2923862338066101,
+      "learning_rate": 7.282711345910263e-05,
+      "loss": 0.0111,
+      "step": 11450
+    },
+    {
+      "epoch": 0.6675598532067338,
+      "grad_norm": 0.36427006125450134,
+      "learning_rate": 7.27780631022783e-05,
+      "loss": 0.0149,
+      "step": 11460
+    },
+    {
+      "epoch": 0.6681423661676472,
+      "grad_norm": 0.21705785393714905,
+      "learning_rate": 7.272898506795948e-05,
+      "loss": 0.0114,
+      "step": 11470
+    },
+    {
+      "epoch": 0.6687248791285606,
+      "grad_norm": 0.22277680039405823,
+      "learning_rate": 7.267987941578058e-05,
+      "loss": 0.0094,
+      "step": 11480
+    },
+    {
+      "epoch": 0.669307392089474,
+      "grad_norm": 0.16814929246902466,
+      "learning_rate": 7.263074620540962e-05,
+      "loss": 0.0068,
+      "step": 11490
+    },
+    {
+      "epoch": 0.6698899050503874,
+      "grad_norm": 0.2653045356273651,
+      "learning_rate": 7.25815854965481e-05,
+      "loss": 0.0102,
+      "step": 11500
+    },
+    {
+      "epoch": 0.6704724180113008,
+      "grad_norm": 0.2011665254831314,
+      "learning_rate": 7.253239734893089e-05,
+      "loss": 0.0075,
+      "step": 11510
+    },
+    {
+      "epoch": 0.6710549309722141,
+      "grad_norm": 0.17460496723651886,
+      "learning_rate": 7.248318182232623e-05,
+      "loss": 0.0091,
+      "step": 11520
+    },
+    {
+      "epoch": 0.6716374439331275,
+      "grad_norm": 0.4131033718585968,
+      "learning_rate": 7.243393897653565e-05,
+      "loss": 0.0105,
+      "step": 11530
+    },
+    {
+      "epoch": 0.6722199568940409,
+      "grad_norm": 0.24108853936195374,
+      "learning_rate": 7.23846688713938e-05,
+      "loss": 0.0081,
+      "step": 11540
+    },
+    {
+      "epoch": 0.6728024698549543,
+      "grad_norm": 0.19078513979911804,
+      "learning_rate": 7.233537156676854e-05,
+      "loss": 0.0092,
+      "step": 11550
+    },
+    {
+      "epoch": 0.6733849828158677,
+      "grad_norm": 0.3143669068813324,
+      "learning_rate": 7.228604712256076e-05,
+      "loss": 0.0087,
+      "step": 11560
+    },
+    {
+      "epoch": 0.6739674957767811,
+      "grad_norm": 0.15949471294879913,
+      "learning_rate": 7.223669559870426e-05,
+      "loss": 0.0085,
+      "step": 11570
+    },
+    {
+      "epoch": 0.6745500087376944,
+      "grad_norm": 0.2685379981994629,
+      "learning_rate": 7.218731705516585e-05,
+      "loss": 0.0082,
+      "step": 11580
+    },
+    {
+      "epoch": 0.6751325216986078,
+      "grad_norm": 0.30356070399284363,
+      "learning_rate": 7.21379115519451e-05,
+      "loss": 0.0096,
+      "step": 11590
+    },
+    {
+      "epoch": 0.6757150346595212,
+      "grad_norm": 0.30611756443977356,
+      "learning_rate": 7.20884791490743e-05,
+      "loss": 0.0095,
+      "step": 11600
+    },
+    {
+      "epoch": 0.6762975476204346,
+      "grad_norm": 0.24524232745170593,
+      "learning_rate": 7.203901990661857e-05,
+      "loss": 0.0095,
+      "step": 11610
+    },
+    {
+      "epoch": 0.676880060581348,
+      "grad_norm": 0.2544029951095581,
+      "learning_rate": 7.198953388467549e-05,
+      "loss": 0.0104,
+      "step": 11620
+    },
+    {
+      "epoch": 0.6774625735422614,
+      "grad_norm": 0.19939731061458588,
+      "learning_rate": 7.194002114337526e-05,
+      "loss": 0.0078,
+      "step": 11630
+    },
+    {
+      "epoch": 0.6780450865031747,
+      "grad_norm": 0.2947027385234833,
+      "learning_rate": 7.189048174288054e-05,
+      "loss": 0.0107,
+      "step": 11640
+    },
+    {
+      "epoch": 0.6786275994640881,
+      "grad_norm": 0.16887688636779785,
+      "learning_rate": 7.184091574338636e-05,
+      "loss": 0.0082,
+      "step": 11650
+    },
+    {
+      "epoch": 0.6792101124250015,
+      "grad_norm": 0.44667932391166687,
+      "learning_rate": 7.179132320512009e-05,
+      "loss": 0.0098,
+      "step": 11660
+    },
+    {
+      "epoch": 0.6797926253859148,
+      "grad_norm": 0.28363966941833496,
+      "learning_rate": 7.174170418834134e-05,
+      "loss": 0.0092,
+      "step": 11670
+    },
+    {
+      "epoch": 0.6803751383468282,
+      "grad_norm": 0.1710769087076187,
+      "learning_rate": 7.169205875334189e-05,
+      "loss": 0.0083,
+      "step": 11680
+    },
+    {
+      "epoch": 0.6809576513077416,
+      "grad_norm": 0.21662937104701996,
+      "learning_rate": 7.164238696044562e-05,
+      "loss": 0.0099,
+      "step": 11690
+    },
+    {
+      "epoch": 0.6815401642686549,
+      "grad_norm": 0.248279869556427,
+      "learning_rate": 7.159268887000846e-05,
+      "loss": 0.0081,
+      "step": 11700
+    },
+    {
+      "epoch": 0.6821226772295683,
+      "grad_norm": 0.1566261500120163,
+      "learning_rate": 7.154296454241827e-05,
+      "loss": 0.0115,
+      "step": 11710
+    },
+    {
+      "epoch": 0.6827051901904817,
+      "grad_norm": 0.17701533436775208,
+      "learning_rate": 7.149321403809479e-05,
+      "loss": 0.0077,
+      "step": 11720
+    },
+    {
+      "epoch": 0.6832877031513951,
+      "grad_norm": 0.2095070630311966,
+      "learning_rate": 7.144343741748956e-05,
+      "loss": 0.0086,
+      "step": 11730
+    },
+    {
+      "epoch": 0.6838702161123085,
+      "grad_norm": 0.20947040617465973,
+      "learning_rate": 7.139363474108589e-05,
+      "loss": 0.0067,
+      "step": 11740
+    },
+    {
+      "epoch": 0.6844527290732219,
+      "grad_norm": 0.16656440496444702,
+      "learning_rate": 7.134380606939874e-05,
+      "loss": 0.0075,
+      "step": 11750
+    },
+    {
+      "epoch": 0.6850352420341352,
+      "grad_norm": 0.2839418351650238,
+      "learning_rate": 7.12939514629746e-05,
+      "loss": 0.0104,
+      "step": 11760
+    },
+    {
+      "epoch": 0.6856177549950486,
+      "grad_norm": 0.40935373306274414,
+      "learning_rate": 7.124407098239155e-05,
+      "loss": 0.0064,
+      "step": 11770
+    },
+    {
+      "epoch": 0.686200267955962,
+      "grad_norm": 0.24508097767829895,
+      "learning_rate": 7.119416468825908e-05,
+      "loss": 0.0073,
+      "step": 11780
+    },
+    {
+      "epoch": 0.6867827809168754,
+      "grad_norm": 0.2682045102119446,
+      "learning_rate": 7.114423264121804e-05,
+      "loss": 0.0089,
+      "step": 11790
+    },
+    {
+      "epoch": 0.6873652938777888,
+      "grad_norm": 0.24415819346904755,
+      "learning_rate": 7.109427490194056e-05,
+      "loss": 0.011,
+      "step": 11800
+    },
+    {
+      "epoch": 0.6879478068387022,
+      "grad_norm": 0.3626476526260376,
+      "learning_rate": 7.104429153113001e-05,
+      "loss": 0.0087,
+      "step": 11810
+    },
+    {
+      "epoch": 0.6885303197996155,
+      "grad_norm": 0.2745492160320282,
+      "learning_rate": 7.099428258952092e-05,
+      "loss": 0.0102,
+      "step": 11820
+    },
+    {
+      "epoch": 0.6891128327605289,
+      "grad_norm": 0.31161826848983765,
+      "learning_rate": 7.094424813787883e-05,
+      "loss": 0.0087,
+      "step": 11830
+    },
+    {
+      "epoch": 0.6896953457214423,
+      "grad_norm": 0.26391202211380005,
+      "learning_rate": 7.089418823700034e-05,
+      "loss": 0.0074,
+      "step": 11840
+    },
+    {
+      "epoch": 0.6902778586823557,
+      "grad_norm": 0.19274213910102844,
+      "learning_rate": 7.084410294771298e-05,
+      "loss": 0.0116,
+      "step": 11850
+    },
+    {
+      "epoch": 0.6908603716432691,
+      "grad_norm": 0.4339475929737091,
+      "learning_rate": 7.079399233087504e-05,
+      "loss": 0.0117,
+      "step": 11860
+    },
+    {
+      "epoch": 0.6914428846041825,
+      "grad_norm": 0.20847979187965393,
+      "learning_rate": 7.074385644737568e-05,
+      "loss": 0.013,
+      "step": 11870
+    },
+    {
+      "epoch": 0.6920253975650958,
+      "grad_norm": 0.3464682102203369,
+      "learning_rate": 7.069369535813473e-05,
+      "loss": 0.0104,
+      "step": 11880
+    },
+    {
+      "epoch": 0.6926079105260092,
+      "grad_norm": 0.43526190519332886,
+      "learning_rate": 7.06435091241026e-05,
+      "loss": 0.0121,
+      "step": 11890
+    },
+    {
+      "epoch": 0.6931904234869226,
+      "grad_norm": 0.39423561096191406,
+      "learning_rate": 7.059329780626034e-05,
+      "loss": 0.0128,
+      "step": 11900
+    },
+    {
+      "epoch": 0.693772936447836,
+      "grad_norm": 0.1348189413547516,
+      "learning_rate": 7.054306146561944e-05,
+      "loss": 0.0093,
+      "step": 11910
+    },
+    {
+      "epoch": 0.6943554494087494,
+      "grad_norm": 0.21422705054283142,
+      "learning_rate": 7.049280016322178e-05,
+      "loss": 0.0083,
+      "step": 11920
+    },
+    {
+      "epoch": 0.6949379623696628,
+      "grad_norm": 0.310937762260437,
+      "learning_rate": 7.044251396013957e-05,
+      "loss": 0.0066,
+      "step": 11930
+    },
+    {
+      "epoch": 0.6955204753305761,
+      "grad_norm": 0.24765044450759888,
+      "learning_rate": 7.039220291747528e-05,
+      "loss": 0.0096,
+      "step": 11940
+    },
+    {
+      "epoch": 0.6961029882914895,
+      "grad_norm": 0.1896153688430786,
+      "learning_rate": 7.034186709636159e-05,
+      "loss": 0.0074,
+      "step": 11950
+    },
+    {
+      "epoch": 0.6966855012524029,
+      "grad_norm": 0.17196683585643768,
+      "learning_rate": 7.029150655796129e-05,
+      "loss": 0.0085,
+      "step": 11960
+    },
+    {
+      "epoch": 0.6972680142133163,
+      "grad_norm": 0.19206859171390533,
+      "learning_rate": 7.024112136346712e-05,
+      "loss": 0.0098,
+      "step": 11970
+    },
+    {
+      "epoch": 0.6978505271742296,
+      "grad_norm": 0.2963821589946747,
+      "learning_rate": 7.01907115741019e-05,
+      "loss": 0.0083,
+      "step": 11980
+    },
+    {
+      "epoch": 0.698433040135143,
+      "grad_norm": 0.28042685985565186,
+      "learning_rate": 7.014027725111825e-05,
+      "loss": 0.0077,
+      "step": 11990
+    },
+    {
+      "epoch": 0.6990155530960563,
+      "grad_norm": 0.2616901695728302,
+      "learning_rate": 7.008981845579864e-05,
+      "loss": 0.0076,
+      "step": 12000
+    },
+    {
+      "epoch": 0.6995980660569697,
+      "grad_norm": 0.3192943334579468,
+      "learning_rate": 7.003933524945528e-05,
+      "loss": 0.0109,
+      "step": 12010
+    },
+    {
+      "epoch": 0.7001805790178831,
+      "grad_norm": 0.2817870080471039,
+      "learning_rate": 6.998882769342998e-05,
+      "loss": 0.0107,
+      "step": 12020
+    },
+    {
+      "epoch": 0.7007630919787965,
+      "grad_norm": 0.2208189219236374,
+      "learning_rate": 6.993829584909423e-05,
+      "loss": 0.0097,
+      "step": 12030
+    },
+    {
+      "epoch": 0.7013456049397099,
+      "grad_norm": 0.27795520424842834,
+      "learning_rate": 6.988773977784895e-05,
+      "loss": 0.0072,
+      "step": 12040
+    },
+    {
+      "epoch": 0.7019281179006233,
+      "grad_norm": 0.4111139178276062,
+      "learning_rate": 6.983715954112454e-05,
+      "loss": 0.0093,
+      "step": 12050
+    },
+    {
+      "epoch": 0.7025106308615366,
+      "grad_norm": 0.3238416314125061,
+      "learning_rate": 6.978655520038079e-05,
+      "loss": 0.0102,
+      "step": 12060
+    },
+    {
+      "epoch": 0.70309314382245,
+      "grad_norm": 0.19110700488090515,
+      "learning_rate": 6.97359268171067e-05,
+      "loss": 0.0087,
+      "step": 12070
+    },
+    {
+      "epoch": 0.7036756567833634,
+      "grad_norm": 0.2528466284275055,
+      "learning_rate": 6.968527445282056e-05,
+      "loss": 0.0088,
+      "step": 12080
+    },
+    {
+      "epoch": 0.7042581697442768,
+      "grad_norm": 0.20163866877555847,
+      "learning_rate": 6.963459816906977e-05,
+      "loss": 0.0076,
+      "step": 12090
+    },
+    {
+      "epoch": 0.7048406827051902,
+      "grad_norm": 0.21941594779491425,
+      "learning_rate": 6.958389802743077e-05,
+      "loss": 0.0095,
+      "step": 12100
+    },
+    {
+      "epoch": 0.7054231956661036,
+      "grad_norm": 0.23076827824115753,
+      "learning_rate": 6.953317408950903e-05,
+      "loss": 0.0108,
+      "step": 12110
+    },
+    {
+      "epoch": 0.706005708627017,
+      "grad_norm": 0.3254426121711731,
+      "learning_rate": 6.94824264169389e-05,
+      "loss": 0.0096,
+      "step": 12120
+    },
+    {
+      "epoch": 0.7065882215879303,
+      "grad_norm": 0.17755886912345886,
+      "learning_rate": 6.94316550713836e-05,
+      "loss": 0.0082,
+      "step": 12130
+    },
+    {
+      "epoch": 0.7071707345488437,
+      "grad_norm": 0.24817058444023132,
+      "learning_rate": 6.938086011453513e-05,
+      "loss": 0.0077,
+      "step": 12140
+    },
+    {
+      "epoch": 0.7077532475097571,
+      "grad_norm": 0.32095474004745483,
+      "learning_rate": 6.93300416081141e-05,
+      "loss": 0.0116,
+      "step": 12150
+    },
+    {
+      "epoch": 0.7083357604706705,
+      "grad_norm": 0.2723959982395172,
+      "learning_rate": 6.927919961386984e-05,
+      "loss": 0.0148,
+      "step": 12160
+    },
+    {
+      "epoch": 0.7089182734315839,
+      "grad_norm": 0.1369193196296692,
+      "learning_rate": 6.922833419358013e-05,
+      "loss": 0.009,
+      "step": 12170
+    },
+    {
+      "epoch": 0.7095007863924973,
+      "grad_norm": 0.28748583793640137,
+      "learning_rate": 6.917744540905124e-05,
+      "loss": 0.0113,
+      "step": 12180
+    },
+    {
+      "epoch": 0.7100832993534106,
+      "grad_norm": 0.31490764021873474,
+      "learning_rate": 6.912653332211787e-05,
+      "loss": 0.0116,
+      "step": 12190
+    },
+    {
+      "epoch": 0.710665812314324,
+      "grad_norm": 0.21386760473251343,
+      "learning_rate": 6.9075597994643e-05,
+      "loss": 0.0077,
+      "step": 12200
+    },
+    {
+      "epoch": 0.7112483252752374,
+      "grad_norm": 0.32422274351119995,
+      "learning_rate": 6.902463948851787e-05,
+      "loss": 0.0113,
+      "step": 12210
+    },
+    {
+      "epoch": 0.7118308382361508,
+      "grad_norm": 0.1831922084093094,
+      "learning_rate": 6.897365786566184e-05,
+      "loss": 0.0106,
+      "step": 12220
+    },
+    {
+      "epoch": 0.7124133511970642,
+      "grad_norm": 0.35979321599006653,
+      "learning_rate": 6.892265318802242e-05,
+      "loss": 0.0133,
+      "step": 12230
+    },
+    {
+      "epoch": 0.7129958641579776,
+      "grad_norm": 0.2936689853668213,
+      "learning_rate": 6.887162551757507e-05,
+      "loss": 0.0088,
+      "step": 12240
+    },
+    {
+      "epoch": 0.7135783771188909,
+      "grad_norm": 0.19652894139289856,
+      "learning_rate": 6.882057491632326e-05,
+      "loss": 0.0085,
+      "step": 12250
+    },
+    {
+      "epoch": 0.7141608900798043,
+      "grad_norm": 0.2707644999027252,
+      "learning_rate": 6.876950144629824e-05,
+      "loss": 0.0095,
+      "step": 12260
+    },
+    {
+      "epoch": 0.7147434030407177,
+      "grad_norm": 0.48824790120124817,
+      "learning_rate": 6.87184051695591e-05,
+      "loss": 0.0111,
+      "step": 12270
+    },
+    {
+      "epoch": 0.7153259160016311,
+      "grad_norm": 0.31840792298316956,
+      "learning_rate": 6.866728614819268e-05,
+      "loss": 0.0099,
+      "step": 12280
+    },
+    {
+      "epoch": 0.7159084289625444,
+      "grad_norm": 0.29164719581604004,
+      "learning_rate": 6.861614444431337e-05,
+      "loss": 0.0087,
+      "step": 12290
+    },
+    {
+      "epoch": 0.7164909419234577,
+      "grad_norm": 0.18168750405311584,
+      "learning_rate": 6.856498012006318e-05,
+      "loss": 0.0079,
+      "step": 12300
+    },
+    {
+      "epoch": 0.7170734548843711,
+      "grad_norm": 0.28029417991638184,
+      "learning_rate": 6.851379323761157e-05,
+      "loss": 0.0084,
+      "step": 12310
+    },
+    {
+      "epoch": 0.7176559678452845,
+      "grad_norm": 0.21910573542118073,
+      "learning_rate": 6.846258385915545e-05,
+      "loss": 0.0083,
+      "step": 12320
+    },
+    {
+      "epoch": 0.7182384808061979,
+      "grad_norm": 0.1764867901802063,
+      "learning_rate": 6.841135204691902e-05,
+      "loss": 0.0086,
+      "step": 12330
+    },
+    {
+      "epoch": 0.7188209937671113,
+      "grad_norm": 0.2565914988517761,
+      "learning_rate": 6.836009786315377e-05,
+      "loss": 0.0099,
+      "step": 12340
+    },
+    {
+      "epoch": 0.7194035067280247,
+      "grad_norm": 0.30022677779197693,
+      "learning_rate": 6.830882137013839e-05,
+      "loss": 0.0087,
+      "step": 12350
+    },
+    {
+      "epoch": 0.719986019688938,
+      "grad_norm": 0.2706559896469116,
+      "learning_rate": 6.825752263017863e-05,
+      "loss": 0.0076,
+      "step": 12360
+    },
+    {
+      "epoch": 0.7205685326498514,
+      "grad_norm": 0.243755042552948,
+      "learning_rate": 6.820620170560731e-05,
+      "loss": 0.0066,
+      "step": 12370
+    },
+    {
+      "epoch": 0.7211510456107648,
+      "grad_norm": 0.35159483551979065,
+      "learning_rate": 6.815485865878418e-05,
+      "loss": 0.0119,
+      "step": 12380
+    },
+    {
+      "epoch": 0.7217335585716782,
+      "grad_norm": 0.1953379213809967,
+      "learning_rate": 6.810349355209588e-05,
+      "loss": 0.0097,
+      "step": 12390
+    },
+    {
+      "epoch": 0.7223160715325916,
+      "grad_norm": 0.3029601573944092,
+      "learning_rate": 6.805210644795588e-05,
+      "loss": 0.0105,
+      "step": 12400
+    },
+    {
+      "epoch": 0.722898584493505,
+      "grad_norm": 0.3221445381641388,
+      "learning_rate": 6.800069740880433e-05,
+      "loss": 0.0101,
+      "step": 12410
+    },
+    {
+      "epoch": 0.7234810974544184,
+      "grad_norm": 0.2511308193206787,
+      "learning_rate": 6.794926649710807e-05,
+      "loss": 0.0091,
+      "step": 12420
+    },
+    {
+      "epoch": 0.7240636104153317,
+      "grad_norm": 0.23041501641273499,
+      "learning_rate": 6.789781377536052e-05,
+      "loss": 0.0084,
+      "step": 12430
+    },
+    {
+      "epoch": 0.7246461233762451,
+      "grad_norm": 0.16476142406463623,
+      "learning_rate": 6.784633930608158e-05,
+      "loss": 0.0079,
+      "step": 12440
+    },
+    {
+      "epoch": 0.7252286363371585,
+      "grad_norm": 0.2456912249326706,
+      "learning_rate": 6.779484315181759e-05,
+      "loss": 0.0061,
+      "step": 12450
+    },
+    {
+      "epoch": 0.7258111492980719,
+      "grad_norm": 0.3024348318576813,
+      "learning_rate": 6.774332537514122e-05,
+      "loss": 0.0067,
+      "step": 12460
+    },
+    {
+      "epoch": 0.7263936622589853,
+      "grad_norm": 0.18799103796482086,
+      "learning_rate": 6.769178603865143e-05,
+      "loss": 0.0067,
+      "step": 12470
+    },
+    {
+      "epoch": 0.7269761752198987,
+      "grad_norm": 0.371861070394516,
+      "learning_rate": 6.764022520497337e-05,
+      "loss": 0.009,
+      "step": 12480
+    },
+    {
+      "epoch": 0.727558688180812,
+      "grad_norm": 0.2018270045518875,
+      "learning_rate": 6.758864293675833e-05,
+      "loss": 0.0077,
+      "step": 12490
+    },
+    {
+      "epoch": 0.7281412011417254,
+      "grad_norm": 0.2783985435962677,
+      "learning_rate": 6.753703929668362e-05,
+      "loss": 0.0081,
+      "step": 12500
+    },
+    {
+      "epoch": 0.7287237141026388,
+      "grad_norm": 0.23501135408878326,
+      "learning_rate": 6.748541434745253e-05,
+      "loss": 0.0067,
+      "step": 12510
+    },
+    {
+      "epoch": 0.7293062270635522,
+      "grad_norm": 0.24632450938224792,
+      "learning_rate": 6.743376815179424e-05,
+      "loss": 0.0074,
+      "step": 12520
+    },
+    {
+      "epoch": 0.7298887400244656,
+      "grad_norm": 0.22143439948558807,
+      "learning_rate": 6.738210077246376e-05,
+      "loss": 0.0072,
+      "step": 12530
+    },
+    {
+      "epoch": 0.730471252985379,
+      "grad_norm": 0.330000638961792,
+      "learning_rate": 6.733041227224181e-05,
+      "loss": 0.0069,
+      "step": 12540
+    },
+    {
+      "epoch": 0.7310537659462923,
+      "grad_norm": 0.18161383271217346,
+      "learning_rate": 6.72787027139348e-05,
+      "loss": 0.0102,
+      "step": 12550
+    },
+    {
+      "epoch": 0.7316362789072057,
+      "grad_norm": 0.2450011968612671,
+      "learning_rate": 6.72269721603747e-05,
+      "loss": 0.0067,
+      "step": 12560
+    },
+    {
+      "epoch": 0.7322187918681191,
+      "grad_norm": 0.1909734159708023,
+      "learning_rate": 6.717522067441904e-05,
+      "loss": 0.01,
+      "step": 12570
+    },
+    {
+      "epoch": 0.7328013048290325,
+      "grad_norm": 0.19598892331123352,
+      "learning_rate": 6.712344831895074e-05,
+      "loss": 0.0084,
+      "step": 12580
+    },
+    {
+      "epoch": 0.7333838177899459,
+      "grad_norm": 0.28475114703178406,
+      "learning_rate": 6.707165515687811e-05,
+      "loss": 0.0073,
+      "step": 12590
+    },
+    {
+      "epoch": 0.7339663307508592,
+      "grad_norm": 0.3548679053783417,
+      "learning_rate": 6.70198412511347e-05,
+      "loss": 0.0085,
+      "step": 12600
+    },
+    {
+      "epoch": 0.7345488437117725,
+      "grad_norm": 0.1787130981683731,
+      "learning_rate": 6.69680066646793e-05,
+      "loss": 0.0072,
+      "step": 12610
+    },
+    {
+      "epoch": 0.7351313566726859,
+      "grad_norm": 0.33021262288093567,
+      "learning_rate": 6.691615146049583e-05,
+      "loss": 0.0088,
+      "step": 12620
+    },
+    {
+      "epoch": 0.7357138696335993,
+      "grad_norm": 0.31082281470298767,
+      "learning_rate": 6.686427570159324e-05,
+      "loss": 0.0099,
+      "step": 12630
+    },
+    {
+      "epoch": 0.7362963825945127,
+      "grad_norm": 0.2861289978027344,
+      "learning_rate": 6.681237945100548e-05,
+      "loss": 0.0088,
+      "step": 12640
+    },
+    {
+      "epoch": 0.7368788955554261,
+      "grad_norm": 0.29931530356407166,
+      "learning_rate": 6.676046277179139e-05,
+      "loss": 0.0085,
+      "step": 12650
+    },
+    {
+      "epoch": 0.7374614085163395,
+      "grad_norm": 0.3150509297847748,
+      "learning_rate": 6.670852572703462e-05,
+      "loss": 0.0092,
+      "step": 12660
+    },
+    {
+      "epoch": 0.7380439214772528,
+      "grad_norm": 0.29703447222709656,
+      "learning_rate": 6.665656837984359e-05,
+      "loss": 0.0106,
+      "step": 12670
+    },
+    {
+      "epoch": 0.7386264344381662,
+      "grad_norm": 0.21371686458587646,
+      "learning_rate": 6.660459079335135e-05,
+      "loss": 0.0068,
+      "step": 12680
+    },
+    {
+      "epoch": 0.7392089473990796,
+      "grad_norm": 0.34527355432510376,
+      "learning_rate": 6.655259303071558e-05,
+      "loss": 0.0133,
+      "step": 12690
+    },
+    {
+      "epoch": 0.739791460359993,
+      "grad_norm": 0.2435365915298462,
+      "learning_rate": 6.650057515511848e-05,
+      "loss": 0.01,
+      "step": 12700
+    },
+    {
+      "epoch": 0.7403739733209064,
+      "grad_norm": 0.2495061308145523,
+      "learning_rate": 6.644853722976667e-05,
+      "loss": 0.0066,
+      "step": 12710
+    },
+    {
+      "epoch": 0.7409564862818198,
+      "grad_norm": 0.2159392535686493,
+      "learning_rate": 6.639647931789113e-05,
+      "loss": 0.0089,
+      "step": 12720
+    },
+    {
+      "epoch": 0.7415389992427331,
+      "grad_norm": 0.24808821082115173,
+      "learning_rate": 6.634440148274713e-05,
+      "loss": 0.0089,
+      "step": 12730
+    },
+    {
+      "epoch": 0.7421215122036465,
+      "grad_norm": 0.14140279591083527,
+      "learning_rate": 6.629230378761415e-05,
+      "loss": 0.0076,
+      "step": 12740
+    },
+    {
+      "epoch": 0.7427040251645599,
+      "grad_norm": 0.2241564840078354,
+      "learning_rate": 6.624018629579581e-05,
+      "loss": 0.0073,
+      "step": 12750
+    },
+    {
+      "epoch": 0.7432865381254733,
+      "grad_norm": 0.1419210433959961,
+      "learning_rate": 6.618804907061976e-05,
+      "loss": 0.0075,
+      "step": 12760
+    },
+    {
+      "epoch": 0.7438690510863867,
+      "grad_norm": 0.16972410678863525,
+      "learning_rate": 6.613589217543766e-05,
+      "loss": 0.0073,
+      "step": 12770
+    },
+    {
+      "epoch": 0.7444515640473001,
+      "grad_norm": 0.134632870554924,
+      "learning_rate": 6.608371567362504e-05,
+      "loss": 0.0115,
+      "step": 12780
+    },
+    {
+      "epoch": 0.7450340770082134,
+      "grad_norm": 0.34847497940063477,
+      "learning_rate": 6.60315196285813e-05,
+      "loss": 0.0106,
+      "step": 12790
+    },
+    {
+      "epoch": 0.7456165899691268,
+      "grad_norm": 0.17909187078475952,
+      "learning_rate": 6.597930410372954e-05,
+      "loss": 0.0088,
+      "step": 12800
+    },
+    {
+      "epoch": 0.7461991029300402,
+      "grad_norm": 0.3426502048969269,
+      "learning_rate": 6.592706916251653e-05,
+      "loss": 0.0103,
+      "step": 12810
+    },
+    {
+      "epoch": 0.7467816158909536,
+      "grad_norm": 0.23082861304283142,
+      "learning_rate": 6.587481486841267e-05,
+      "loss": 0.0093,
+      "step": 12820
+    },
+    {
+      "epoch": 0.747364128851867,
+      "grad_norm": 0.32361316680908203,
+      "learning_rate": 6.582254128491184e-05,
+      "loss": 0.0089,
+      "step": 12830
+    },
+    {
+      "epoch": 0.7479466418127804,
+      "grad_norm": 0.22431160509586334,
+      "learning_rate": 6.577024847553139e-05,
+      "loss": 0.0095,
+      "step": 12840
+    },
+    {
+      "epoch": 0.7485291547736938,
+      "grad_norm": 0.1638309359550476,
+      "learning_rate": 6.571793650381201e-05,
+      "loss": 0.0063,
+      "step": 12850
+    },
+    {
+      "epoch": 0.7491116677346071,
+      "grad_norm": 0.20186251401901245,
+      "learning_rate": 6.566560543331766e-05,
+      "loss": 0.0079,
+      "step": 12860
+    },
+    {
+      "epoch": 0.7496941806955205,
+      "grad_norm": 0.2446933388710022,
+      "learning_rate": 6.561325532763554e-05,
+      "loss": 0.0083,
+      "step": 12870
+    },
+    {
+      "epoch": 0.7502766936564339,
+      "grad_norm": 0.2417115867137909,
+      "learning_rate": 6.556088625037597e-05,
+      "loss": 0.0086,
+      "step": 12880
+    },
+    {
+      "epoch": 0.7508592066173473,
+      "grad_norm": 0.1608356386423111,
+      "learning_rate": 6.55084982651723e-05,
+      "loss": 0.007,
+      "step": 12890
+    },
+    {
+      "epoch": 0.7514417195782607,
+      "grad_norm": 0.23487161099910736,
+      "learning_rate": 6.54560914356809e-05,
+      "loss": 0.0063,
+      "step": 12900
+    },
+    {
+      "epoch": 0.7520242325391739,
+      "grad_norm": 0.2483849823474884,
+      "learning_rate": 6.540366582558097e-05,
+      "loss": 0.0071,
+      "step": 12910
+    },
+    {
+      "epoch": 0.7526067455000873,
+      "grad_norm": 0.18976159393787384,
+      "learning_rate": 6.53512214985746e-05,
+      "loss": 0.0092,
+      "step": 12920
+    },
+    {
+      "epoch": 0.7531892584610007,
+      "grad_norm": 0.284068763256073,
+      "learning_rate": 6.529875851838658e-05,
+      "loss": 0.008,
+      "step": 12930
+    },
+    {
+      "epoch": 0.7537717714219141,
+      "grad_norm": 0.2622430920600891,
+      "learning_rate": 6.52462769487644e-05,
+      "loss": 0.0058,
+      "step": 12940
+    },
+    {
+      "epoch": 0.7543542843828275,
+      "grad_norm": 0.18223126232624054,
+      "learning_rate": 6.519377685347807e-05,
+      "loss": 0.0087,
+      "step": 12950
+    },
+    {
+      "epoch": 0.7549367973437409,
+      "grad_norm": 0.3101232051849365,
+      "learning_rate": 6.514125829632021e-05,
+      "loss": 0.0089,
+      "step": 12960
+    },
+    {
+      "epoch": 0.7555193103046542,
+      "grad_norm": 0.20565840601921082,
+      "learning_rate": 6.508872134110577e-05,
+      "loss": 0.0087,
+      "step": 12970
+    },
+    {
+      "epoch": 0.7561018232655676,
+      "grad_norm": 0.2806341350078583,
+      "learning_rate": 6.503616605167213e-05,
+      "loss": 0.0116,
+      "step": 12980
+    },
+    {
+      "epoch": 0.756684336226481,
+      "grad_norm": 0.3076115548610687,
+      "learning_rate": 6.498359249187893e-05,
+      "loss": 0.0095,
+      "step": 12990
+    },
+    {
+      "epoch": 0.7572668491873944,
+      "grad_norm": 0.28373560309410095,
+      "learning_rate": 6.493100072560799e-05,
+      "loss": 0.01,
+      "step": 13000
+    },
+    {
+      "epoch": 0.7578493621483078,
+      "grad_norm": 0.1805787831544876,
+      "learning_rate": 6.487839081676327e-05,
+      "loss": 0.0106,
+      "step": 13010
+    },
+    {
+      "epoch": 0.7584318751092212,
+      "grad_norm": 0.25359657406806946,
+      "learning_rate": 6.482576282927076e-05,
+      "loss": 0.0083,
+      "step": 13020
+    },
+    {
+      "epoch": 0.7590143880701345,
+      "grad_norm": 0.24211549758911133,
+      "learning_rate": 6.477311682707845e-05,
+      "loss": 0.0086,
+      "step": 13030
+    },
+    {
+      "epoch": 0.7595969010310479,
+      "grad_norm": 0.17044241726398468,
+      "learning_rate": 6.472045287415616e-05,
+      "loss": 0.0087,
+      "step": 13040
+    },
+    {
+      "epoch": 0.7601794139919613,
+      "grad_norm": 0.24509020149707794,
+      "learning_rate": 6.466777103449558e-05,
+      "loss": 0.0091,
+      "step": 13050
+    },
+    {
+      "epoch": 0.7607619269528747,
+      "grad_norm": 0.1323293000459671,
+      "learning_rate": 6.461507137211012e-05,
+      "loss": 0.0085,
+      "step": 13060
+    },
+    {
+      "epoch": 0.7613444399137881,
+      "grad_norm": 0.1885659545660019,
+      "learning_rate": 6.456235395103483e-05,
+      "loss": 0.0052,
+      "step": 13070
+    },
+    {
+      "epoch": 0.7619269528747015,
+      "grad_norm": 0.28635963797569275,
+      "learning_rate": 6.450961883532634e-05,
+      "loss": 0.0072,
+      "step": 13080
+    },
+    {
+      "epoch": 0.7625094658356149,
+      "grad_norm": 0.2091861069202423,
+      "learning_rate": 6.445686608906283e-05,
+      "loss": 0.0091,
+      "step": 13090
+    },
+    {
+      "epoch": 0.7630919787965282,
+      "grad_norm": 0.5681166052818298,
+      "learning_rate": 6.44040957763438e-05,
+      "loss": 0.0091,
+      "step": 13100
+    },
+    {
+      "epoch": 0.7636744917574416,
+      "grad_norm": 0.2220182567834854,
+      "learning_rate": 6.435130796129018e-05,
+      "loss": 0.0083,
+      "step": 13110
+    },
+    {
+      "epoch": 0.764257004718355,
+      "grad_norm": 0.24739138782024384,
+      "learning_rate": 6.429850270804416e-05,
+      "loss": 0.0109,
+      "step": 13120
+    },
+    {
+      "epoch": 0.7648395176792684,
+      "grad_norm": 0.3352735638618469,
+      "learning_rate": 6.424568008076909e-05,
+      "loss": 0.0117,
+      "step": 13130
+    },
+    {
+      "epoch": 0.7654220306401818,
+      "grad_norm": 0.2567942440509796,
+      "learning_rate": 6.419284014364943e-05,
+      "loss": 0.0075,
+      "step": 13140
+    },
+    {
+      "epoch": 0.7660045436010952,
+      "grad_norm": 0.33313682675361633,
+      "learning_rate": 6.41399829608907e-05,
+      "loss": 0.0083,
+      "step": 13150
+    },
+    {
+      "epoch": 0.7665870565620085,
+      "grad_norm": 0.15305909514427185,
+      "learning_rate": 6.408710859671938e-05,
+      "loss": 0.0061,
+      "step": 13160
+    },
+    {
+      "epoch": 0.7671695695229219,
+      "grad_norm": 0.30298030376434326,
+      "learning_rate": 6.403421711538278e-05,
+      "loss": 0.0092,
+      "step": 13170
+    },
+    {
+      "epoch": 0.7677520824838353,
+      "grad_norm": 0.13110916316509247,
+      "learning_rate": 6.398130858114903e-05,
+      "loss": 0.0066,
+      "step": 13180
+    },
+    {
+      "epoch": 0.7683345954447487,
+      "grad_norm": 0.25777778029441833,
+      "learning_rate": 6.392838305830701e-05,
+      "loss": 0.0065,
+      "step": 13190
+    },
+    {
+      "epoch": 0.7689171084056621,
+      "grad_norm": 0.21617020666599274,
+      "learning_rate": 6.387544061116621e-05,
+      "loss": 0.0105,
+      "step": 13200
+    },
+    {
+      "epoch": 0.7694996213665755,
+      "grad_norm": 0.23031793534755707,
+      "learning_rate": 6.382248130405671e-05,
+      "loss": 0.0068,
+      "step": 13210
+    },
+    {
+      "epoch": 0.7700821343274887,
+      "grad_norm": 0.26593661308288574,
+      "learning_rate": 6.376950520132905e-05,
+      "loss": 0.0085,
+      "step": 13220
+    },
+    {
+      "epoch": 0.7706646472884021,
+      "grad_norm": 0.169608473777771,
+      "learning_rate": 6.371651236735418e-05,
+      "loss": 0.0085,
+      "step": 13230
+    },
+    {
+      "epoch": 0.7712471602493155,
+      "grad_norm": 0.1696559637784958,
+      "learning_rate": 6.366350286652341e-05,
+      "loss": 0.009,
+      "step": 13240
+    },
+    {
+      "epoch": 0.7718296732102289,
+      "grad_norm": 0.25189146399497986,
+      "learning_rate": 6.361047676324827e-05,
+      "loss": 0.0084,
+      "step": 13250
+    },
+    {
+      "epoch": 0.7724121861711423,
+      "grad_norm": 0.1725015938282013,
+      "learning_rate": 6.355743412196047e-05,
+      "loss": 0.0101,
+      "step": 13260
+    },
+    {
+      "epoch": 0.7729946991320557,
+      "grad_norm": 0.26044392585754395,
+      "learning_rate": 6.350437500711184e-05,
+      "loss": 0.0101,
+      "step": 13270
+    },
+    {
+      "epoch": 0.773577212092969,
+      "grad_norm": 0.16396208107471466,
+      "learning_rate": 6.345129948317418e-05,
+      "loss": 0.0078,
+      "step": 13280
+    },
+    {
+      "epoch": 0.7741597250538824,
+      "grad_norm": 0.20403556525707245,
+      "learning_rate": 6.33982076146393e-05,
+      "loss": 0.0095,
+      "step": 13290
+    },
+    {
+      "epoch": 0.7747422380147958,
+      "grad_norm": 0.17464278638362885,
+      "learning_rate": 6.334509946601879e-05,
+      "loss": 0.0093,
+      "step": 13300
+    },
+    {
+      "epoch": 0.7753247509757092,
+      "grad_norm": 0.19244590401649475,
+      "learning_rate": 6.329197510184406e-05,
+      "loss": 0.0073,
+      "step": 13310
+    },
+    {
+      "epoch": 0.7759072639366226,
+      "grad_norm": 0.2402058094739914,
+      "learning_rate": 6.323883458666624e-05,
+      "loss": 0.0089,
+      "step": 13320
+    },
+    {
+      "epoch": 0.776489776897536,
+      "grad_norm": 0.27657461166381836,
+      "learning_rate": 6.318567798505605e-05,
+      "loss": 0.0099,
+      "step": 13330
+    },
+    {
+      "epoch": 0.7770722898584493,
+      "grad_norm": 0.24560029804706573,
+      "learning_rate": 6.313250536160378e-05,
+      "loss": 0.008,
+      "step": 13340
+    },
+    {
+      "epoch": 0.7776548028193627,
+      "grad_norm": 0.18147815763950348,
+      "learning_rate": 6.307931678091917e-05,
+      "loss": 0.0066,
+      "step": 13350
+    },
+    {
+      "epoch": 0.7782373157802761,
+      "grad_norm": 0.2779025733470917,
+      "learning_rate": 6.302611230763138e-05,
+      "loss": 0.0082,
+      "step": 13360
+    },
+    {
+      "epoch": 0.7788198287411895,
+      "grad_norm": 0.20841482281684875,
+      "learning_rate": 6.297289200638887e-05,
+      "loss": 0.0072,
+      "step": 13370
+    },
+    {
+      "epoch": 0.7794023417021029,
+      "grad_norm": 0.1896149218082428,
+      "learning_rate": 6.29196559418593e-05,
+      "loss": 0.0088,
+      "step": 13380
+    },
+    {
+      "epoch": 0.7799848546630163,
+      "grad_norm": 0.22444932162761688,
+      "learning_rate": 6.286640417872951e-05,
+      "loss": 0.0088,
+      "step": 13390
+    },
+    {
+      "epoch": 0.7805673676239296,
+      "grad_norm": 0.15474310517311096,
+      "learning_rate": 6.281313678170542e-05,
+      "loss": 0.0066,
+      "step": 13400
+    },
+    {
+      "epoch": 0.781149880584843,
+      "grad_norm": 0.25766295194625854,
+      "learning_rate": 6.275985381551194e-05,
+      "loss": 0.0076,
+      "step": 13410
+    },
+    {
+      "epoch": 0.7817323935457564,
+      "grad_norm": 0.3598175346851349,
+      "learning_rate": 6.270655534489292e-05,
+      "loss": 0.007,
+      "step": 13420
+    },
+    {
+      "epoch": 0.7823149065066698,
+      "grad_norm": 0.11953185498714447,
+      "learning_rate": 6.265324143461098e-05,
+      "loss": 0.0085,
+      "step": 13430
+    },
+    {
+      "epoch": 0.7828974194675832,
+      "grad_norm": 0.14179261028766632,
+      "learning_rate": 6.259991214944758e-05,
+      "loss": 0.0123,
+      "step": 13440
+    },
+    {
+      "epoch": 0.7834799324284966,
+      "grad_norm": 0.15050460398197174,
+      "learning_rate": 6.254656755420283e-05,
+      "loss": 0.0075,
+      "step": 13450
+    },
+    {
+      "epoch": 0.78406244538941,
+      "grad_norm": 0.15919229388237,
+      "learning_rate": 6.249320771369545e-05,
+      "loss": 0.0082,
+      "step": 13460
+    },
+    {
+      "epoch": 0.7846449583503233,
+      "grad_norm": 0.17207765579223633,
+      "learning_rate": 6.243983269276263e-05,
+      "loss": 0.0077,
+      "step": 13470
+    },
+    {
+      "epoch": 0.7852274713112367,
+      "grad_norm": 0.2298387885093689,
+      "learning_rate": 6.238644255626012e-05,
+      "loss": 0.0059,
+      "step": 13480
+    },
+    {
+      "epoch": 0.7858099842721501,
+      "grad_norm": 0.3039219379425049,
+      "learning_rate": 6.233303736906193e-05,
+      "loss": 0.0129,
+      "step": 13490
+    },
+    {
+      "epoch": 0.7863924972330635,
+      "grad_norm": 0.21114885807037354,
+      "learning_rate": 6.227961719606039e-05,
+      "loss": 0.0094,
+      "step": 13500
+    },
+    {
+      "epoch": 0.7869750101939769,
+      "grad_norm": 0.26820486783981323,
+      "learning_rate": 6.222618210216608e-05,
+      "loss": 0.0103,
+      "step": 13510
+    },
+    {
+      "epoch": 0.7875575231548902,
+      "grad_norm": 0.15495796501636505,
+      "learning_rate": 6.217273215230767e-05,
+      "loss": 0.0067,
+      "step": 13520
+    },
+    {
+      "epoch": 0.7881400361158035,
+      "grad_norm": 0.23395368456840515,
+      "learning_rate": 6.211926741143188e-05,
+      "loss": 0.0071,
+      "step": 13530
+    },
+    {
+      "epoch": 0.7887225490767169,
+      "grad_norm": 0.258790522813797,
+      "learning_rate": 6.20657879445034e-05,
+      "loss": 0.0116,
+      "step": 13540
+    },
+    {
+      "epoch": 0.7893050620376303,
+      "grad_norm": 0.31089794635772705,
+      "learning_rate": 6.201229381650485e-05,
+      "loss": 0.011,
+      "step": 13550
+    },
+    {
+      "epoch": 0.7898875749985437,
+      "grad_norm": 0.2505391836166382,
+      "learning_rate": 6.195878509243661e-05,
+      "loss": 0.0089,
+      "step": 13560
+    },
+    {
+      "epoch": 0.7904700879594571,
+      "grad_norm": 0.16423818469047546,
+      "learning_rate": 6.190526183731685e-05,
+      "loss": 0.006,
+      "step": 13570
+    },
+    {
+      "epoch": 0.7910526009203704,
+      "grad_norm": 0.22283509373664856,
+      "learning_rate": 6.185172411618138e-05,
+      "loss": 0.0108,
+      "step": 13580
+    },
+    {
+      "epoch": 0.7916351138812838,
+      "grad_norm": 0.25103291869163513,
+      "learning_rate": 6.179817199408355e-05,
+      "loss": 0.0077,
+      "step": 13590
+    },
+    {
+      "epoch": 0.7922176268421972,
+      "grad_norm": 0.13040512800216675,
+      "learning_rate": 6.174460553609426e-05,
+      "loss": 0.0096,
+      "step": 13600
+    },
+    {
+      "epoch": 0.7928001398031106,
+      "grad_norm": 0.18640290200710297,
+      "learning_rate": 6.16910248073018e-05,
+      "loss": 0.0085,
+      "step": 13610
+    },
+    {
+      "epoch": 0.793382652764024,
+      "grad_norm": 0.2809637188911438,
+      "learning_rate": 6.16374298728118e-05,
+      "loss": 0.0078,
+      "step": 13620
+    },
+    {
+      "epoch": 0.7939651657249374,
+      "grad_norm": 0.29580461978912354,
+      "learning_rate": 6.158382079774716e-05,
+      "loss": 0.0083,
+      "step": 13630
+    },
+    {
+      "epoch": 0.7945476786858507,
+      "grad_norm": 0.17137892544269562,
+      "learning_rate": 6.153019764724799e-05,
+      "loss": 0.0089,
+      "step": 13640
+    },
+    {
+      "epoch": 0.7951301916467641,
+      "grad_norm": 0.27623164653778076,
+      "learning_rate": 6.147656048647144e-05,
+      "loss": 0.0076,
+      "step": 13650
+    },
+    {
+      "epoch": 0.7957127046076775,
+      "grad_norm": 0.18555782735347748,
+      "learning_rate": 6.142290938059173e-05,
+      "loss": 0.0096,
+      "step": 13660
+    },
+    {
+      "epoch": 0.7962952175685909,
+      "grad_norm": 0.32818278670310974,
+      "learning_rate": 6.136924439480001e-05,
+      "loss": 0.0064,
+      "step": 13670
+    },
+    {
+      "epoch": 0.7968777305295043,
+      "grad_norm": 0.2009228765964508,
+      "learning_rate": 6.13155655943043e-05,
+      "loss": 0.0103,
+      "step": 13680
+    },
+    {
+      "epoch": 0.7974602434904177,
+      "grad_norm": 0.34780409932136536,
+      "learning_rate": 6.126187304432941e-05,
+      "loss": 0.0076,
+      "step": 13690
+    },
+    {
+      "epoch": 0.798042756451331,
+      "grad_norm": 0.18259115517139435,
+      "learning_rate": 6.120816681011684e-05,
+      "loss": 0.0077,
+      "step": 13700
+    },
+    {
+      "epoch": 0.7986252694122444,
+      "grad_norm": 0.38433167338371277,
+      "learning_rate": 6.115444695692474e-05,
+      "loss": 0.0099,
+      "step": 13710
+    },
+    {
+      "epoch": 0.7992077823731578,
+      "grad_norm": 0.3076366186141968,
+      "learning_rate": 6.110071355002779e-05,
+      "loss": 0.0066,
+      "step": 13720
+    },
+    {
+      "epoch": 0.7997902953340712,
+      "grad_norm": 0.3600670397281647,
+      "learning_rate": 6.104696665471714e-05,
+      "loss": 0.008,
+      "step": 13730
+    },
+    {
+      "epoch": 0.8003728082949846,
+      "grad_norm": 0.19162212312221527,
+      "learning_rate": 6.099320633630036e-05,
+      "loss": 0.0065,
+      "step": 13740
+    },
+    {
+      "epoch": 0.800955321255898,
+      "grad_norm": 0.3683108389377594,
+      "learning_rate": 6.093943266010128e-05,
+      "loss": 0.0138,
+      "step": 13750
+    },
+    {
+      "epoch": 0.8015378342168114,
+      "grad_norm": 0.2973504960536957,
+      "learning_rate": 6.088564569146e-05,
+      "loss": 0.0075,
+      "step": 13760
+    },
+    {
+      "epoch": 0.8021203471777247,
+      "grad_norm": 0.31088486313819885,
+      "learning_rate": 6.083184549573274e-05,
+      "loss": 0.0084,
+      "step": 13770
+    },
+    {
+      "epoch": 0.8027028601386381,
+      "grad_norm": 0.20674611628055573,
+      "learning_rate": 6.077803213829184e-05,
+      "loss": 0.0077,
+      "step": 13780
+    },
+    {
+      "epoch": 0.8032853730995515,
+      "grad_norm": 0.21986396610736847,
+      "learning_rate": 6.0724205684525594e-05,
+      "loss": 0.0089,
+      "step": 13790
+    },
+    {
+      "epoch": 0.8038678860604649,
+      "grad_norm": 0.26583778858184814,
+      "learning_rate": 6.0670366199838215e-05,
+      "loss": 0.0095,
+      "step": 13800
+    },
+    {
+      "epoch": 0.8044503990213783,
+      "grad_norm": 0.2051122486591339,
+      "learning_rate": 6.061651374964974e-05,
+      "loss": 0.0115,
+      "step": 13810
+    },
+    {
+      "epoch": 0.8050329119822917,
+      "grad_norm": 0.1803383082151413,
+      "learning_rate": 6.0562648399396004e-05,
+      "loss": 0.008,
+      "step": 13820
+    },
+    {
+      "epoch": 0.805615424943205,
+      "grad_norm": 0.2943856120109558,
+      "learning_rate": 6.050877021452845e-05,
+      "loss": 0.009,
+      "step": 13830
+    },
+    {
+      "epoch": 0.8061979379041183,
+      "grad_norm": 0.2781359851360321,
+      "learning_rate": 6.04548792605142e-05,
+      "loss": 0.0094,
+      "step": 13840
+    },
+    {
+      "epoch": 0.8067804508650317,
+      "grad_norm": 0.2086310237646103,
+      "learning_rate": 6.04009756028358e-05,
+      "loss": 0.0073,
+      "step": 13850
+    },
+    {
+      "epoch": 0.8073629638259451,
+      "grad_norm": 0.18407326936721802,
+      "learning_rate": 6.0347059306991294e-05,
+      "loss": 0.0092,
+      "step": 13860
+    },
+    {
+      "epoch": 0.8079454767868585,
+      "grad_norm": 0.2851587235927582,
+      "learning_rate": 6.029313043849407e-05,
+      "loss": 0.007,
+      "step": 13870
+    },
+    {
+      "epoch": 0.8085279897477718,
+      "grad_norm": 0.16651418805122375,
+      "learning_rate": 6.023918906287279e-05,
+      "loss": 0.0079,
+      "step": 13880
+    },
+    {
+      "epoch": 0.8091105027086852,
+      "grad_norm": 0.31108665466308594,
+      "learning_rate": 6.01852352456713e-05,
+      "loss": 0.0082,
+      "step": 13890
+    },
+    {
+      "epoch": 0.8096930156695986,
+      "grad_norm": 0.19270960986614227,
+      "learning_rate": 6.013126905244858e-05,
+      "loss": 0.0086,
+      "step": 13900
+    },
+    {
+      "epoch": 0.810275528630512,
+      "grad_norm": 0.2899821400642395,
+      "learning_rate": 6.007729054877864e-05,
+      "loss": 0.0071,
+      "step": 13910
+    },
+    {
+      "epoch": 0.8108580415914254,
+      "grad_norm": 0.34771227836608887,
+      "learning_rate": 6.0023299800250466e-05,
+      "loss": 0.0086,
+      "step": 13920
+    },
+    {
+      "epoch": 0.8114405545523388,
+      "grad_norm": 0.3173035979270935,
+      "learning_rate": 5.9969296872467894e-05,
+      "loss": 0.0078,
+      "step": 13930
+    },
+    {
+      "epoch": 0.8120230675132522,
+      "grad_norm": 0.28133124113082886,
+      "learning_rate": 5.9915281831049585e-05,
+      "loss": 0.0093,
+      "step": 13940
+    },
+    {
+      "epoch": 0.8126055804741655,
+      "grad_norm": 0.2978377938270569,
+      "learning_rate": 5.98612547416289e-05,
+      "loss": 0.0103,
+      "step": 13950
+    },
+    {
+      "epoch": 0.8131880934350789,
+      "grad_norm": 0.23337078094482422,
+      "learning_rate": 5.9807215669853855e-05,
+      "loss": 0.0087,
+      "step": 13960
+    },
+    {
+      "epoch": 0.8137706063959923,
+      "grad_norm": 0.20659947395324707,
+      "learning_rate": 5.9753164681387e-05,
+      "loss": 0.007,
+      "step": 13970
+    },
+    {
+      "epoch": 0.8143531193569057,
+      "grad_norm": 0.2011047750711441,
+      "learning_rate": 5.969910184190539e-05,
+      "loss": 0.0061,
+      "step": 13980
+    },
+    {
+      "epoch": 0.8149356323178191,
+      "grad_norm": 0.16730518639087677,
+      "learning_rate": 5.9645027217100477e-05,
+      "loss": 0.0081,
+      "step": 13990
+    },
+    {
+      "epoch": 0.8155181452787325,
+      "grad_norm": 0.14032626152038574,
+      "learning_rate": 5.959094087267804e-05,
+      "loss": 0.0059,
+      "step": 14000
+    },
+    {
+      "epoch": 0.8161006582396458,
+      "grad_norm": 0.13825510442256927,
+      "learning_rate": 5.9536842874358065e-05,
+      "loss": 0.0081,
+      "step": 14010
+    },
+    {
+      "epoch": 0.8166831712005592,
+      "grad_norm": 0.20877620577812195,
+      "learning_rate": 5.9482733287874734e-05,
+      "loss": 0.0077,
+      "step": 14020
+    },
+    {
+      "epoch": 0.8172656841614726,
+      "grad_norm": 0.16830898821353912,
+      "learning_rate": 5.942861217897631e-05,
+      "loss": 0.0082,
+      "step": 14030
+    },
+    {
+      "epoch": 0.817848197122386,
+      "grad_norm": 0.2236877679824829,
+      "learning_rate": 5.9374479613425e-05,
+      "loss": 0.0071,
+      "step": 14040
+    },
+    {
+      "epoch": 0.8184307100832994,
+      "grad_norm": 0.1825820505619049,
+      "learning_rate": 5.932033565699704e-05,
+      "loss": 0.0063,
+      "step": 14050
+    },
+    {
+      "epoch": 0.8190132230442128,
+      "grad_norm": 0.3014568090438843,
+      "learning_rate": 5.926618037548237e-05,
+      "loss": 0.0084,
+      "step": 14060
+    },
+    {
+      "epoch": 0.8195957360051261,
+      "grad_norm": 0.23975959420204163,
+      "learning_rate": 5.9212013834684824e-05,
+      "loss": 0.0075,
+      "step": 14070
+    },
+    {
+      "epoch": 0.8201782489660395,
+      "grad_norm": 0.22208575904369354,
+      "learning_rate": 5.9157836100421824e-05,
+      "loss": 0.0072,
+      "step": 14080
+    },
+    {
+      "epoch": 0.8207607619269529,
+      "grad_norm": 0.26557594537734985,
+      "learning_rate": 5.910364723852444e-05,
+      "loss": 0.0065,
+      "step": 14090
+    },
+    {
+      "epoch": 0.8213432748878663,
+      "grad_norm": 0.26533761620521545,
+      "learning_rate": 5.904944731483723e-05,
+      "loss": 0.0071,
+      "step": 14100
+    },
+    {
+      "epoch": 0.8219257878487797,
+      "grad_norm": 0.30062136054039,
+      "learning_rate": 5.899523639521825e-05,
+      "loss": 0.0056,
+      "step": 14110
+    },
+    {
+      "epoch": 0.8225083008096931,
+      "grad_norm": 0.24028576910495758,
+      "learning_rate": 5.894101454553883e-05,
+      "loss": 0.0063,
+      "step": 14120
+    },
+    {
+      "epoch": 0.8230908137706064,
+      "grad_norm": 0.2715035080909729,
+      "learning_rate": 5.888678183168368e-05,
+      "loss": 0.0085,
+      "step": 14130
+    },
+    {
+      "epoch": 0.8236733267315198,
+      "grad_norm": 0.2137950360774994,
+      "learning_rate": 5.883253831955061e-05,
+      "loss": 0.0084,
+      "step": 14140
+    },
+    {
+      "epoch": 0.8242558396924332,
+      "grad_norm": 0.19830097258090973,
+      "learning_rate": 5.8778284075050625e-05,
+      "loss": 0.0065,
+      "step": 14150
+    },
+    {
+      "epoch": 0.8248383526533465,
+      "grad_norm": 0.22738951444625854,
+      "learning_rate": 5.872401916410777e-05,
+      "loss": 0.008,
+      "step": 14160
+    },
+    {
+      "epoch": 0.8254208656142599,
+      "grad_norm": 0.20681419968605042,
+      "learning_rate": 5.8669743652659014e-05,
+      "loss": 0.009,
+      "step": 14170
+    },
+    {
+      "epoch": 0.8260033785751733,
+      "grad_norm": 0.21602334082126617,
+      "learning_rate": 5.86154576066542e-05,
+      "loss": 0.0069,
+      "step": 14180
+    },
+    {
+      "epoch": 0.8265858915360866,
+      "grad_norm": 0.22629040479660034,
+      "learning_rate": 5.856116109205602e-05,
+      "loss": 0.0069,
+      "step": 14190
+    },
+    {
+      "epoch": 0.827168404497,
+      "grad_norm": 0.2837110459804535,
+      "learning_rate": 5.850685417483983e-05,
+      "loss": 0.0063,
+      "step": 14200
+    },
+    {
+      "epoch": 0.8277509174579134,
+      "grad_norm": 0.29922983050346375,
+      "learning_rate": 5.8452536920993684e-05,
+      "loss": 0.0074,
+      "step": 14210
+    },
+    {
+      "epoch": 0.8283334304188268,
+      "grad_norm": 0.212235227227211,
+      "learning_rate": 5.839820939651817e-05,
+      "loss": 0.0081,
+      "step": 14220
+    },
+    {
+      "epoch": 0.8289159433797402,
+      "grad_norm": 0.23015020787715912,
+      "learning_rate": 5.834387166742632e-05,
+      "loss": 0.0084,
+      "step": 14230
+    },
+    {
+      "epoch": 0.8294984563406536,
+      "grad_norm": 0.21302254498004913,
+      "learning_rate": 5.828952379974364e-05,
+      "loss": 0.008,
+      "step": 14240
+    },
+    {
+      "epoch": 0.8300809693015669,
+      "grad_norm": 0.20616202056407928,
+      "learning_rate": 5.8235165859507864e-05,
+      "loss": 0.0075,
+      "step": 14250
+    },
+    {
+      "epoch": 0.8306634822624803,
+      "grad_norm": 0.24378959834575653,
+      "learning_rate": 5.818079791276907e-05,
+      "loss": 0.0095,
+      "step": 14260
+    },
+    {
+      "epoch": 0.8312459952233937,
+      "grad_norm": 0.20431454479694366,
+      "learning_rate": 5.8126420025589415e-05,
+      "loss": 0.006,
+      "step": 14270
+    },
+    {
+      "epoch": 0.8318285081843071,
+      "grad_norm": 0.20747745037078857,
+      "learning_rate": 5.807203226404313e-05,
+      "loss": 0.0068,
+      "step": 14280
+    },
+    {
+      "epoch": 0.8324110211452205,
+      "grad_norm": 0.11964654922485352,
+      "learning_rate": 5.801763469421652e-05,
+      "loss": 0.006,
+      "step": 14290
+    },
+    {
+      "epoch": 0.8329935341061339,
+      "grad_norm": 0.22247470915317535,
+      "learning_rate": 5.796322738220774e-05,
+      "loss": 0.007,
+      "step": 14300
+    },
+    {
+      "epoch": 0.8335760470670472,
+      "grad_norm": 0.22033612430095673,
+      "learning_rate": 5.79088103941268e-05,
+      "loss": 0.0058,
+      "step": 14310
+    },
+    {
+      "epoch": 0.8341585600279606,
+      "grad_norm": 0.1835365742444992,
+      "learning_rate": 5.785438379609549e-05,
+      "loss": 0.0099,
+      "step": 14320
+    },
+    {
+      "epoch": 0.834741072988874,
+      "grad_norm": 0.25826510787010193,
+      "learning_rate": 5.779994765424724e-05,
+      "loss": 0.0096,
+      "step": 14330
+    },
+    {
+      "epoch": 0.8353235859497874,
+      "grad_norm": 0.28984546661376953,
+      "learning_rate": 5.7745502034727114e-05,
+      "loss": 0.0082,
+      "step": 14340
+    },
+    {
+      "epoch": 0.8359060989107008,
+      "grad_norm": 0.17987816035747528,
+      "learning_rate": 5.769104700369164e-05,
+      "loss": 0.0063,
+      "step": 14350
+    },
+    {
+      "epoch": 0.8364886118716142,
+      "grad_norm": 0.40107908844947815,
+      "learning_rate": 5.7636582627308854e-05,
+      "loss": 0.0092,
+      "step": 14360
+    },
+    {
+      "epoch": 0.8370711248325275,
+      "grad_norm": 0.19707366824150085,
+      "learning_rate": 5.7582108971758095e-05,
+      "loss": 0.0075,
+      "step": 14370
+    },
+    {
+      "epoch": 0.8376536377934409,
+      "grad_norm": 0.19297166168689728,
+      "learning_rate": 5.752762610322999e-05,
+      "loss": 0.0077,
+      "step": 14380
+    },
+    {
+      "epoch": 0.8382361507543543,
+      "grad_norm": 0.21698884665966034,
+      "learning_rate": 5.747313408792636e-05,
+      "loss": 0.0071,
+      "step": 14390
+    },
+    {
+      "epoch": 0.8388186637152677,
+      "grad_norm": 0.14300556480884552,
+      "learning_rate": 5.741863299206014e-05,
+      "loss": 0.0072,
+      "step": 14400
+    },
+    {
+      "epoch": 0.8394011766761811,
+      "grad_norm": 0.13605079054832458,
+      "learning_rate": 5.73641228818553e-05,
+      "loss": 0.008,
+      "step": 14410
+    },
+    {
+      "epoch": 0.8399836896370945,
+      "grad_norm": 0.25681909918785095,
+      "learning_rate": 5.730960382354676e-05,
+      "loss": 0.0084,
+      "step": 14420
+    },
+    {
+      "epoch": 0.8405662025980078,
+      "grad_norm": 0.16988423466682434,
+      "learning_rate": 5.725507588338035e-05,
+      "loss": 0.0061,
+      "step": 14430
+    },
+    {
+      "epoch": 0.8411487155589212,
+      "grad_norm": 0.2314169853925705,
+      "learning_rate": 5.7200539127612604e-05,
+      "loss": 0.0067,
+      "step": 14440
+    },
+    {
+      "epoch": 0.8417312285198346,
+      "grad_norm": 0.3801364004611969,
+      "learning_rate": 5.714599362251087e-05,
+      "loss": 0.0116,
+      "step": 14450
+    },
+    {
+      "epoch": 0.842313741480748,
+      "grad_norm": 0.34764859080314636,
+      "learning_rate": 5.709143943435307e-05,
+      "loss": 0.0076,
+      "step": 14460
+    },
+    {
+      "epoch": 0.8428962544416613,
+      "grad_norm": 0.3420166075229645,
+      "learning_rate": 5.7036876629427646e-05,
+      "loss": 0.01,
+      "step": 14470
+    },
+    {
+      "epoch": 0.8434787674025747,
+      "grad_norm": 0.3660932779312134,
+      "learning_rate": 5.698230527403361e-05,
+      "loss": 0.0129,
+      "step": 14480
+    },
+    {
+      "epoch": 0.844061280363488,
+      "grad_norm": 0.3269963264465332,
+      "learning_rate": 5.6927725434480274e-05,
+      "loss": 0.0082,
+      "step": 14490
+    },
+    {
+      "epoch": 0.8446437933244014,
+      "grad_norm": 0.22245074808597565,
+      "learning_rate": 5.687313717708728e-05,
+      "loss": 0.0051,
+      "step": 14500
+    },
+    {
+      "epoch": 0.8452263062853148,
+      "grad_norm": 0.17340850830078125,
+      "learning_rate": 5.681854056818453e-05,
+      "loss": 0.0063,
+      "step": 14510
+    },
+    {
+      "epoch": 0.8458088192462282,
+      "grad_norm": 0.2705160975456238,
+      "learning_rate": 5.676393567411205e-05,
+      "loss": 0.0061,
+      "step": 14520
+    },
+    {
+      "epoch": 0.8463913322071416,
+      "grad_norm": 0.2048708200454712,
+      "learning_rate": 5.670932256121991e-05,
+      "loss": 0.0079,
+      "step": 14530
+    },
+    {
+      "epoch": 0.846973845168055,
+      "grad_norm": 0.35470420122146606,
+      "learning_rate": 5.6654701295868216e-05,
+      "loss": 0.0089,
+      "step": 14540
+    },
+    {
+      "epoch": 0.8475563581289683,
+      "grad_norm": 0.1902502328157425,
+      "learning_rate": 5.660007194442697e-05,
+      "loss": 0.0094,
+      "step": 14550
+    },
+    {
+      "epoch": 0.8481388710898817,
+      "grad_norm": 0.14856259524822235,
+      "learning_rate": 5.6545434573275946e-05,
+      "loss": 0.0087,
+      "step": 14560
+    },
+    {
+      "epoch": 0.8487213840507951,
+      "grad_norm": 0.28809863328933716,
+      "learning_rate": 5.649078924880472e-05,
+      "loss": 0.0087,
+      "step": 14570
+    },
+    {
+      "epoch": 0.8493038970117085,
+      "grad_norm": 0.19070513546466827,
+      "learning_rate": 5.643613603741252e-05,
+      "loss": 0.007,
+      "step": 14580
+    },
+    {
+      "epoch": 0.8498864099726219,
+      "grad_norm": 0.28022000193595886,
+      "learning_rate": 5.6381475005508154e-05,
+      "loss": 0.0065,
+      "step": 14590
+    },
+    {
+      "epoch": 0.8504689229335353,
+      "grad_norm": 0.21863733232021332,
+      "learning_rate": 5.632680621950992e-05,
+      "loss": 0.0084,
+      "step": 14600
+    },
+    {
+      "epoch": 0.8510514358944486,
+      "grad_norm": 0.20701198279857635,
+      "learning_rate": 5.627212974584555e-05,
+      "loss": 0.0086,
+      "step": 14610
+    },
+    {
+      "epoch": 0.851633948855362,
+      "grad_norm": 0.2946155369281769,
+      "learning_rate": 5.62174456509521e-05,
+      "loss": 0.0073,
+      "step": 14620
+    },
+    {
+      "epoch": 0.8522164618162754,
+      "grad_norm": 0.16696076095104218,
+      "learning_rate": 5.616275400127594e-05,
+      "loss": 0.0093,
+      "step": 14630
+    },
+    {
+      "epoch": 0.8527989747771888,
+      "grad_norm": 0.22075411677360535,
+      "learning_rate": 5.610805486327254e-05,
+      "loss": 0.008,
+      "step": 14640
+    },
+    {
+      "epoch": 0.8533814877381022,
+      "grad_norm": 0.2011999487876892,
+      "learning_rate": 5.6053348303406536e-05,
+      "loss": 0.0057,
+      "step": 14650
+    },
+    {
+      "epoch": 0.8539640006990156,
+      "grad_norm": 0.22958886623382568,
+      "learning_rate": 5.599863438815156e-05,
+      "loss": 0.0065,
+      "step": 14660
+    },
+    {
+      "epoch": 0.854546513659929,
+      "grad_norm": 0.18867529928684235,
+      "learning_rate": 5.594391318399017e-05,
+      "loss": 0.0059,
+      "step": 14670
+    },
+    {
+      "epoch": 0.8551290266208423,
+      "grad_norm": 0.3065764605998993,
+      "learning_rate": 5.588918475741377e-05,
+      "loss": 0.0074,
+      "step": 14680
+    },
+    {
+      "epoch": 0.8557115395817557,
+      "grad_norm": 0.23066751658916473,
+      "learning_rate": 5.5834449174922586e-05,
+      "loss": 0.0066,
+      "step": 14690
+    },
+    {
+      "epoch": 0.8562940525426691,
+      "grad_norm": 0.23487553000450134,
+      "learning_rate": 5.57797065030255e-05,
+      "loss": 0.0071,
+      "step": 14700
+    },
+    {
+      "epoch": 0.8568765655035825,
+      "grad_norm": 0.18767152726650238,
+      "learning_rate": 5.5724956808240016e-05,
+      "loss": 0.0134,
+      "step": 14710
+    },
+    {
+      "epoch": 0.8574590784644959,
+      "grad_norm": 0.1619166135787964,
+      "learning_rate": 5.567020015709219e-05,
+      "loss": 0.0068,
+      "step": 14720
+    },
+    {
+      "epoch": 0.8580415914254093,
+      "grad_norm": 0.18800516426563263,
+      "learning_rate": 5.561543661611649e-05,
+      "loss": 0.0083,
+      "step": 14730
+    },
+    {
+      "epoch": 0.8586241043863226,
+      "grad_norm": 0.3032349646091461,
+      "learning_rate": 5.556066625185583e-05,
+      "loss": 0.0115,
+      "step": 14740
+    },
+    {
+      "epoch": 0.859206617347236,
+      "grad_norm": 0.5111093521118164,
+      "learning_rate": 5.550588913086131e-05,
+      "loss": 0.0088,
+      "step": 14750
+    },
+    {
+      "epoch": 0.8597891303081494,
+      "grad_norm": 0.17507800459861755,
+      "learning_rate": 5.545110531969234e-05,
+      "loss": 0.0083,
+      "step": 14760
+    },
+    {
+      "epoch": 0.8603716432690628,
+      "grad_norm": 0.25111520290374756,
+      "learning_rate": 5.539631488491641e-05,
+      "loss": 0.01,
+      "step": 14770
+    },
+    {
+      "epoch": 0.8609541562299761,
+      "grad_norm": 0.16584138572216034,
+      "learning_rate": 5.534151789310904e-05,
+      "loss": 0.0077,
+      "step": 14780
+    },
+    {
+      "epoch": 0.8615366691908894,
+      "grad_norm": 0.14699086546897888,
+      "learning_rate": 5.528671441085376e-05,
+      "loss": 0.0078,
+      "step": 14790
+    },
+    {
+      "epoch": 0.8621191821518028,
+      "grad_norm": 0.1341957300901413,
+      "learning_rate": 5.523190450474197e-05,
+      "loss": 0.0046,
+      "step": 14800
+    },
+    {
+      "epoch": 0.8627016951127162,
+      "grad_norm": 0.28935322165489197,
+      "learning_rate": 5.5177088241372865e-05,
+      "loss": 0.0069,
+      "step": 14810
+    },
+    {
+      "epoch": 0.8632842080736296,
+      "grad_norm": 0.18391087651252747,
+      "learning_rate": 5.512226568735338e-05,
+      "loss": 0.0064,
+      "step": 14820
+    },
+    {
+      "epoch": 0.863866721034543,
+      "grad_norm": 0.20363964140415192,
+      "learning_rate": 5.506743690929809e-05,
+      "loss": 0.0081,
+      "step": 14830
+    },
+    {
+      "epoch": 0.8644492339954564,
+      "grad_norm": 0.19138318300247192,
+      "learning_rate": 5.501260197382912e-05,
+      "loss": 0.0063,
+      "step": 14840
+    },
+    {
+      "epoch": 0.8650317469563698,
+      "grad_norm": 0.18149197101593018,
+      "learning_rate": 5.49577609475761e-05,
+      "loss": 0.0066,
+      "step": 14850
+    },
+    {
+      "epoch": 0.8656142599172831,
+      "grad_norm": 0.2031419724225998,
+      "learning_rate": 5.490291389717603e-05,
+      "loss": 0.0083,
+      "step": 14860
+    },
+    {
+      "epoch": 0.8661967728781965,
+      "grad_norm": 0.21548089385032654,
+      "learning_rate": 5.484806088927329e-05,
+      "loss": 0.0075,
+      "step": 14870
+    },
+    {
+      "epoch": 0.8667792858391099,
+      "grad_norm": 0.31034934520721436,
+      "learning_rate": 5.479320199051942e-05,
+      "loss": 0.0065,
+      "step": 14880
+    },
+    {
+      "epoch": 0.8673617988000233,
+      "grad_norm": 0.34767958521842957,
+      "learning_rate": 5.473833726757314e-05,
+      "loss": 0.0076,
+      "step": 14890
+    },
+    {
+      "epoch": 0.8679443117609367,
+      "grad_norm": 0.1996099352836609,
+      "learning_rate": 5.4683466787100304e-05,
+      "loss": 0.0074,
+      "step": 14900
+    },
+    {
+      "epoch": 0.86852682472185,
+      "grad_norm": 0.3034256100654602,
+      "learning_rate": 5.462859061577369e-05,
+      "loss": 0.0066,
+      "step": 14910
+    },
+    {
+      "epoch": 0.8691093376827634,
+      "grad_norm": 0.25283098220825195,
+      "learning_rate": 5.4573708820273026e-05,
+      "loss": 0.0091,
+      "step": 14920
+    },
+    {
+      "epoch": 0.8696918506436768,
+      "grad_norm": 0.2175549864768982,
+      "learning_rate": 5.451882146728488e-05,
+      "loss": 0.0062,
+      "step": 14930
+    },
+    {
+      "epoch": 0.8702743636045902,
+      "grad_norm": 0.35509759187698364,
+      "learning_rate": 5.446392862350255e-05,
+      "loss": 0.0079,
+      "step": 14940
+    },
+    {
+      "epoch": 0.8708568765655036,
+      "grad_norm": 0.17008253931999207,
+      "learning_rate": 5.440903035562603e-05,
+      "loss": 0.008,
+      "step": 14950
+    },
+    {
+      "epoch": 0.871439389526417,
+      "grad_norm": 0.23774291574954987,
+      "learning_rate": 5.435412673036188e-05,
+      "loss": 0.0062,
+      "step": 14960
+    },
+    {
+      "epoch": 0.8720219024873304,
+      "grad_norm": 0.20621567964553833,
+      "learning_rate": 5.429921781442318e-05,
+      "loss": 0.0058,
+      "step": 14970
+    },
+    {
+      "epoch": 0.8726044154482437,
+      "grad_norm": 0.1491243541240692,
+      "learning_rate": 5.424430367452945e-05,
+      "loss": 0.006,
+      "step": 14980
+    },
+    {
+      "epoch": 0.8731869284091571,
+      "grad_norm": 0.21246647834777832,
+      "learning_rate": 5.418938437740655e-05,
+      "loss": 0.009,
+      "step": 14990
+    },
+    {
+      "epoch": 0.8737694413700705,
+      "grad_norm": 0.23771041631698608,
+      "learning_rate": 5.413445998978658e-05,
+      "loss": 0.0062,
+      "step": 15000
+    },
+    {
+      "epoch": 0.8743519543309839,
+      "grad_norm": 0.2296302169561386,
+      "learning_rate": 5.407953057840789e-05,
+      "loss": 0.0069,
+      "step": 15010
+    },
+    {
+      "epoch": 0.8749344672918973,
+      "grad_norm": 0.17034557461738586,
+      "learning_rate": 5.4024596210014853e-05,
+      "loss": 0.0062,
+      "step": 15020
+    },
+    {
+      "epoch": 0.8755169802528107,
+      "grad_norm": 0.22645875811576843,
+      "learning_rate": 5.3969656951357935e-05,
+      "loss": 0.0098,
+      "step": 15030
+    },
+    {
+      "epoch": 0.876099493213724,
+      "grad_norm": 0.22713325917720795,
+      "learning_rate": 5.3914712869193504e-05,
+      "loss": 0.0085,
+      "step": 15040
+    },
+    {
+      "epoch": 0.8766820061746374,
+      "grad_norm": 0.17993152141571045,
+      "learning_rate": 5.385976403028381e-05,
+      "loss": 0.0072,
+      "step": 15050
+    },
+    {
+      "epoch": 0.8772645191355508,
+      "grad_norm": 0.17282229661941528,
+      "learning_rate": 5.3804810501396875e-05,
+      "loss": 0.0075,
+      "step": 15060
+    },
+    {
+      "epoch": 0.8778470320964642,
+      "grad_norm": 0.29393839836120605,
+      "learning_rate": 5.374985234930639e-05,
+      "loss": 0.0068,
+      "step": 15070
+    },
+    {
+      "epoch": 0.8784295450573776,
+      "grad_norm": 0.25646069645881653,
+      "learning_rate": 5.3694889640791724e-05,
+      "loss": 0.006,
+      "step": 15080
+    },
+    {
+      "epoch": 0.8790120580182909,
+      "grad_norm": 0.1746547818183899,
+      "learning_rate": 5.3639922442637736e-05,
+      "loss": 0.007,
+      "step": 15090
+    },
+    {
+      "epoch": 0.8795945709792042,
+      "grad_norm": 0.20840086042881012,
+      "learning_rate": 5.358495082163476e-05,
+      "loss": 0.0072,
+      "step": 15100
+    },
+    {
+      "epoch": 0.8801770839401176,
+      "grad_norm": 0.21731878817081451,
+      "learning_rate": 5.35299748445785e-05,
+      "loss": 0.0085,
+      "step": 15110
+    },
+    {
+      "epoch": 0.880759596901031,
+      "grad_norm": 0.20644234120845795,
+      "learning_rate": 5.3474994578269944e-05,
+      "loss": 0.0094,
+      "step": 15120
+    },
+    {
+      "epoch": 0.8813421098619444,
+      "grad_norm": 0.2079085409641266,
+      "learning_rate": 5.34200100895153e-05,
+      "loss": 0.0083,
+      "step": 15130
+    },
+    {
+      "epoch": 0.8819246228228578,
+      "grad_norm": 0.2427864521741867,
+      "learning_rate": 5.3365021445125916e-05,
+      "loss": 0.0057,
+      "step": 15140
+    },
+    {
+      "epoch": 0.8825071357837712,
+      "grad_norm": 0.22550378739833832,
+      "learning_rate": 5.331002871191817e-05,
+      "loss": 0.0065,
+      "step": 15150
+    },
+    {
+      "epoch": 0.8830896487446845,
+      "grad_norm": 0.28540095686912537,
+      "learning_rate": 5.3255031956713443e-05,
+      "loss": 0.0065,
+      "step": 15160
+    },
+    {
+      "epoch": 0.8836721617055979,
+      "grad_norm": 0.24852119386196136,
+      "learning_rate": 5.320003124633795e-05,
+      "loss": 0.0085,
+      "step": 15170
+    },
+    {
+      "epoch": 0.8842546746665113,
+      "grad_norm": 0.2663528025150299,
+      "learning_rate": 5.314502664762274e-05,
+      "loss": 0.006,
+      "step": 15180
+    },
+    {
+      "epoch": 0.8848371876274247,
+      "grad_norm": 0.23151522874832153,
+      "learning_rate": 5.3090018227403604e-05,
+      "loss": 0.0081,
+      "step": 15190
+    },
+    {
+      "epoch": 0.8854197005883381,
+      "grad_norm": 0.12143579870462418,
+      "learning_rate": 5.3035006052520955e-05,
+      "loss": 0.0061,
+      "step": 15200
+    },
+    {
+      "epoch": 0.8860022135492515,
+      "grad_norm": 0.29130372405052185,
+      "learning_rate": 5.297999018981976e-05,
+      "loss": 0.0064,
+      "step": 15210
+    },
+    {
+      "epoch": 0.8865847265101648,
+      "grad_norm": 0.18472237884998322,
+      "learning_rate": 5.2924970706149505e-05,
+      "loss": 0.0059,
+      "step": 15220
+    },
+    {
+      "epoch": 0.8871672394710782,
+      "grad_norm": 0.27427637577056885,
+      "learning_rate": 5.2869947668364014e-05,
+      "loss": 0.0068,
+      "step": 15230
+    },
+    {
+      "epoch": 0.8877497524319916,
+      "grad_norm": 0.3031053841114044,
+      "learning_rate": 5.2814921143321504e-05,
+      "loss": 0.0076,
+      "step": 15240
+    },
+    {
+      "epoch": 0.888332265392905,
+      "grad_norm": 0.2902502119541168,
+      "learning_rate": 5.275989119788436e-05,
+      "loss": 0.01,
+      "step": 15250
+    },
+    {
+      "epoch": 0.8889147783538184,
+      "grad_norm": 0.19721251726150513,
+      "learning_rate": 5.270485789891919e-05,
+      "loss": 0.0064,
+      "step": 15260
+    },
+    {
+      "epoch": 0.8894972913147318,
+      "grad_norm": 0.28069207072257996,
+      "learning_rate": 5.264982131329661e-05,
+      "loss": 0.0093,
+      "step": 15270
+    },
+    {
+      "epoch": 0.8900798042756451,
+      "grad_norm": 0.29210978746414185,
+      "learning_rate": 5.259478150789128e-05,
+      "loss": 0.0068,
+      "step": 15280
+    },
+    {
+      "epoch": 0.8906623172365585,
+      "grad_norm": 0.2540770471096039,
+      "learning_rate": 5.253973854958173e-05,
+      "loss": 0.0088,
+      "step": 15290
+    },
+    {
+      "epoch": 0.8912448301974719,
+      "grad_norm": 0.23548078536987305,
+      "learning_rate": 5.2484692505250376e-05,
+      "loss": 0.0071,
+      "step": 15300
+    },
+    {
+      "epoch": 0.8918273431583853,
+      "grad_norm": 0.19689851999282837,
+      "learning_rate": 5.242964344178333e-05,
+      "loss": 0.0074,
+      "step": 15310
+    },
+    {
+      "epoch": 0.8924098561192987,
+      "grad_norm": 0.2391650676727295,
+      "learning_rate": 5.237459142607041e-05,
+      "loss": 0.0107,
+      "step": 15320
+    },
+    {
+      "epoch": 0.8929923690802121,
+      "grad_norm": 0.2386670708656311,
+      "learning_rate": 5.231953652500498e-05,
+      "loss": 0.0086,
+      "step": 15330
+    },
+    {
+      "epoch": 0.8935748820411255,
+      "grad_norm": 0.27805930376052856,
+      "learning_rate": 5.2264478805483976e-05,
+      "loss": 0.0072,
+      "step": 15340
+    },
+    {
+      "epoch": 0.8941573950020388,
+      "grad_norm": 0.22928449511528015,
+      "learning_rate": 5.220941833440768e-05,
+      "loss": 0.0063,
+      "step": 15350
+    },
+    {
+      "epoch": 0.8947399079629522,
+      "grad_norm": 0.16681011021137238,
+      "learning_rate": 5.215435517867977e-05,
+      "loss": 0.0057,
+      "step": 15360
+    },
+    {
+      "epoch": 0.8953224209238656,
+      "grad_norm": 0.22742965817451477,
+      "learning_rate": 5.209928940520719e-05,
+      "loss": 0.005,
+      "step": 15370
+    },
+    {
+      "epoch": 0.895904933884779,
+      "grad_norm": 0.1593795120716095,
+      "learning_rate": 5.204422108090004e-05,
+      "loss": 0.0072,
+      "step": 15380
+    },
+    {
+      "epoch": 0.8964874468456924,
+      "grad_norm": 0.1953076720237732,
+      "learning_rate": 5.19891502726715e-05,
+      "loss": 0.0097,
+      "step": 15390
+    },
+    {
+      "epoch": 0.8970699598066056,
+      "grad_norm": 0.23075145483016968,
+      "learning_rate": 5.193407704743782e-05,
+      "loss": 0.0084,
+      "step": 15400
+    },
+    {
+      "epoch": 0.897652472767519,
+      "grad_norm": 0.14531663060188293,
+      "learning_rate": 5.187900147211815e-05,
+      "loss": 0.0069,
+      "step": 15410
+    },
+    {
+      "epoch": 0.8982349857284324,
+      "grad_norm": 0.26269760727882385,
+      "learning_rate": 5.182392361363453e-05,
+      "loss": 0.0106,
+      "step": 15420
+    },
+    {
+      "epoch": 0.8988174986893458,
+      "grad_norm": 0.12701228260993958,
+      "learning_rate": 5.176884353891172e-05,
+      "loss": 0.006,
+      "step": 15430
+    },
+    {
+      "epoch": 0.8994000116502592,
+      "grad_norm": 0.29145872592926025,
+      "learning_rate": 5.171376131487722e-05,
+      "loss": 0.0067,
+      "step": 15440
+    },
+    {
+      "epoch": 0.8999825246111726,
+      "grad_norm": 0.1868477463722229,
+      "learning_rate": 5.165867700846113e-05,
+      "loss": 0.0082,
+      "step": 15450
+    },
+    {
+      "epoch": 0.900565037572086,
+      "grad_norm": 0.26740896701812744,
+      "learning_rate": 5.160359068659606e-05,
+      "loss": 0.0063,
+      "step": 15460
+    },
+    {
+      "epoch": 0.9011475505329993,
+      "grad_norm": 0.2722265422344208,
+      "learning_rate": 5.154850241621712e-05,
+      "loss": 0.007,
+      "step": 15470
+    },
+    {
+      "epoch": 0.9017300634939127,
+      "grad_norm": 0.29693853855133057,
+      "learning_rate": 5.149341226426172e-05,
+      "loss": 0.006,
+      "step": 15480
+    },
+    {
+      "epoch": 0.9023125764548261,
+      "grad_norm": 0.13700908422470093,
+      "learning_rate": 5.14383202976696e-05,
+      "loss": 0.0087,
+      "step": 15490
+    },
+    {
+      "epoch": 0.9028950894157395,
+      "grad_norm": 0.20129533112049103,
+      "learning_rate": 5.1383226583382684e-05,
+      "loss": 0.0062,
+      "step": 15500
+    },
+    {
+      "epoch": 0.9034776023766529,
+      "grad_norm": 0.3930365741252899,
+      "learning_rate": 5.132813118834504e-05,
+      "loss": 0.0059,
+      "step": 15510
+    },
+    {
+      "epoch": 0.9040601153375663,
+      "grad_norm": 0.23254643380641937,
+      "learning_rate": 5.1273034179502777e-05,
+      "loss": 0.0074,
+      "step": 15520
+    },
+    {
+      "epoch": 0.9046426282984796,
+      "grad_norm": 0.19001874327659607,
+      "learning_rate": 5.1217935623803945e-05,
+      "loss": 0.0125,
+      "step": 15530
+    },
+    {
+      "epoch": 0.905225141259393,
+      "grad_norm": 0.2687726616859436,
+      "learning_rate": 5.1162835588198476e-05,
+      "loss": 0.0117,
+      "step": 15540
+    },
+    {
+      "epoch": 0.9058076542203064,
+      "grad_norm": 0.22594143450260162,
+      "learning_rate": 5.110773413963813e-05,
+      "loss": 0.008,
+      "step": 15550
+    },
+    {
+      "epoch": 0.9063901671812198,
+      "grad_norm": 0.2397158145904541,
+      "learning_rate": 5.105263134507636e-05,
+      "loss": 0.0067,
+      "step": 15560
+    },
+    {
+      "epoch": 0.9069726801421332,
+      "grad_norm": 0.2888263761997223,
+      "learning_rate": 5.099752727146824e-05,
+      "loss": 0.0061,
+      "step": 15570
+    },
+    {
+      "epoch": 0.9075551931030466,
+      "grad_norm": 0.29717153310775757,
+      "learning_rate": 5.094242198577042e-05,
+      "loss": 0.007,
+      "step": 15580
+    },
+    {
+      "epoch": 0.9081377060639599,
+      "grad_norm": 0.2734794616699219,
+      "learning_rate": 5.088731555494102e-05,
+      "loss": 0.0074,
+      "step": 15590
+    },
+    {
+      "epoch": 0.9087202190248733,
+      "grad_norm": 0.15560469031333923,
+      "learning_rate": 5.0832208045939556e-05,
+      "loss": 0.0066,
+      "step": 15600
+    },
+    {
+      "epoch": 0.9093027319857867,
+      "grad_norm": 0.270797461271286,
+      "learning_rate": 5.0777099525726844e-05,
+      "loss": 0.0051,
+      "step": 15610
+    },
+    {
+      "epoch": 0.9098852449467001,
+      "grad_norm": 0.2046782523393631,
+      "learning_rate": 5.072199006126494e-05,
+      "loss": 0.0067,
+      "step": 15620
+    },
+    {
+      "epoch": 0.9104677579076135,
+      "grad_norm": 0.16648422181606293,
+      "learning_rate": 5.0666879719517026e-05,
+      "loss": 0.0057,
+      "step": 15630
+    },
+    {
+      "epoch": 0.9110502708685269,
+      "grad_norm": 0.3262379467487335,
+      "learning_rate": 5.061176856744737e-05,
+      "loss": 0.007,
+      "step": 15640
+    },
+    {
+      "epoch": 0.9116327838294402,
+      "grad_norm": 0.31289246678352356,
+      "learning_rate": 5.0556656672021205e-05,
+      "loss": 0.0103,
+      "step": 15650
+    },
+    {
+      "epoch": 0.9122152967903536,
+      "grad_norm": 0.18275724351406097,
+      "learning_rate": 5.050154410020472e-05,
+      "loss": 0.0081,
+      "step": 15660
+    },
+    {
+      "epoch": 0.912797809751267,
+      "grad_norm": 0.24845468997955322,
+      "learning_rate": 5.044643091896485e-05,
+      "loss": 0.0069,
+      "step": 15670
+    },
+    {
+      "epoch": 0.9133803227121804,
+      "grad_norm": 0.25659939646720886,
+      "learning_rate": 5.039131719526932e-05,
+      "loss": 0.0067,
+      "step": 15680
+    },
+    {
+      "epoch": 0.9139628356730938,
+      "grad_norm": 0.15802839398384094,
+      "learning_rate": 5.03362029960865e-05,
+      "loss": 0.0068,
+      "step": 15690
+    },
+    {
+      "epoch": 0.9145453486340072,
+      "grad_norm": 0.2508903443813324,
+      "learning_rate": 5.028108838838533e-05,
+      "loss": 0.0063,
+      "step": 15700
+    },
+    {
+      "epoch": 0.9151278615949204,
+      "grad_norm": 0.13881438970565796,
+      "learning_rate": 5.0225973439135276e-05,
+      "loss": 0.0073,
+      "step": 15710
+    },
+    {
+      "epoch": 0.9157103745558338,
+      "grad_norm": 0.15649186074733734,
+      "learning_rate": 5.017085821530617e-05,
+      "loss": 0.0079,
+      "step": 15720
+    },
+    {
+      "epoch": 0.9162928875167472,
+      "grad_norm": 0.2362614870071411,
+      "learning_rate": 5.011574278386822e-05,
+      "loss": 0.006,
+      "step": 15730
+    },
+    {
+      "epoch": 0.9168754004776606,
+      "grad_norm": 0.26474374532699585,
+      "learning_rate": 5.006062721179189e-05,
+      "loss": 0.007,
+      "step": 15740
+    },
+    {
+      "epoch": 0.917457913438574,
+      "grad_norm": 0.3509806990623474,
+      "learning_rate": 5.000551156604777e-05,
+      "loss": 0.0057,
+      "step": 15750
+    },
+    {
+      "epoch": 0.9180404263994874,
+      "grad_norm": 0.2239517718553543,
+      "learning_rate": 4.9950395913606594e-05,
+      "loss": 0.0082,
+      "step": 15760
+    },
+    {
+      "epoch": 0.9186229393604007,
+      "grad_norm": 0.3379695415496826,
+      "learning_rate": 4.989528032143903e-05,
+      "loss": 0.0087,
+      "step": 15770
+    },
+    {
+      "epoch": 0.9192054523213141,
+      "grad_norm": 0.30121397972106934,
+      "learning_rate": 4.984016485651578e-05,
+      "loss": 0.0077,
+      "step": 15780
+    },
+    {
+      "epoch": 0.9197879652822275,
+      "grad_norm": 0.20878809690475464,
+      "learning_rate": 4.9785049585807274e-05,
+      "loss": 0.0068,
+      "step": 15790
+    },
+    {
+      "epoch": 0.9203704782431409,
+      "grad_norm": 0.28229114413261414,
+      "learning_rate": 4.972993457628381e-05,
+      "loss": 0.0073,
+      "step": 15800
+    },
+    {
+      "epoch": 0.9209529912040543,
+      "grad_norm": 0.3348454236984253,
+      "learning_rate": 4.9674819894915306e-05,
+      "loss": 0.0072,
+      "step": 15810
+    },
+    {
+      "epoch": 0.9215355041649677,
+      "grad_norm": 0.256380170583725,
+      "learning_rate": 4.9619705608671265e-05,
+      "loss": 0.0082,
+      "step": 15820
+    },
+    {
+      "epoch": 0.922118017125881,
+      "grad_norm": 0.199356347322464,
+      "learning_rate": 4.956459178452079e-05,
+      "loss": 0.0067,
+      "step": 15830
+    },
+    {
+      "epoch": 0.9227005300867944,
+      "grad_norm": 0.23109392821788788,
+      "learning_rate": 4.950947848943235e-05,
+      "loss": 0.0066,
+      "step": 15840
+    },
+    {
+      "epoch": 0.9232830430477078,
+      "grad_norm": 0.13141418993473053,
+      "learning_rate": 4.9454365790373805e-05,
+      "loss": 0.0089,
+      "step": 15850
+    },
+    {
+      "epoch": 0.9238655560086212,
+      "grad_norm": 0.17570920288562775,
+      "learning_rate": 4.939925375431226e-05,
+      "loss": 0.0049,
+      "step": 15860
+    },
+    {
+      "epoch": 0.9244480689695346,
+      "grad_norm": 0.28139644861221313,
+      "learning_rate": 4.9344142448214046e-05,
+      "loss": 0.0067,
+      "step": 15870
+    },
+    {
+      "epoch": 0.925030581930448,
+      "grad_norm": 0.26011669635772705,
+      "learning_rate": 4.928903193904461e-05,
+      "loss": 0.0085,
+      "step": 15880
+    },
+    {
+      "epoch": 0.9256130948913613,
+      "grad_norm": 0.17150840163230896,
+      "learning_rate": 4.923392229376842e-05,
+      "loss": 0.0061,
+      "step": 15890
+    },
+    {
+      "epoch": 0.9261956078522747,
+      "grad_norm": 0.20849215984344482,
+      "learning_rate": 4.9178813579348865e-05,
+      "loss": 0.0077,
+      "step": 15900
+    },
+    {
+      "epoch": 0.9267781208131881,
+      "grad_norm": 0.16399094462394714,
+      "learning_rate": 4.9123705862748253e-05,
+      "loss": 0.0084,
+      "step": 15910
+    },
+    {
+      "epoch": 0.9273606337741015,
+      "grad_norm": 0.13886746764183044,
+      "learning_rate": 4.9068599210927627e-05,
+      "loss": 0.0071,
+      "step": 15920
+    },
+    {
+      "epoch": 0.9279431467350149,
+      "grad_norm": 0.15196655690670013,
+      "learning_rate": 4.901349369084681e-05,
+      "loss": 0.0118,
+      "step": 15930
+    },
+    {
+      "epoch": 0.9285256596959283,
+      "grad_norm": 0.2778361737728119,
+      "learning_rate": 4.8958389369464156e-05,
+      "loss": 0.0057,
+      "step": 15940
+    },
+    {
+      "epoch": 0.9291081726568416,
+      "grad_norm": 0.46940287947654724,
+      "learning_rate": 4.890328631373666e-05,
+      "loss": 0.0071,
+      "step": 15950
+    },
+    {
+      "epoch": 0.929690685617755,
+      "grad_norm": 0.3208307921886444,
+      "learning_rate": 4.88481845906197e-05,
+      "loss": 0.0096,
+      "step": 15960
+    },
+    {
+      "epoch": 0.9302731985786684,
+      "grad_norm": 0.3008352518081665,
+      "learning_rate": 4.879308426706707e-05,
+      "loss": 0.0079,
+      "step": 15970
+    },
+    {
+      "epoch": 0.9308557115395818,
+      "grad_norm": 0.24216167628765106,
+      "learning_rate": 4.873798541003084e-05,
+      "loss": 0.0071,
+      "step": 15980
+    },
+    {
+      "epoch": 0.9314382245004952,
+      "grad_norm": 0.30538102984428406,
+      "learning_rate": 4.868288808646136e-05,
+      "loss": 0.0066,
+      "step": 15990
+    },
+    {
+      "epoch": 0.9320207374614086,
+      "grad_norm": 0.1646200716495514,
+      "learning_rate": 4.862779236330705e-05,
+      "loss": 0.0069,
+      "step": 16000
+    },
+    {
+      "epoch": 0.932603250422322,
+      "grad_norm": 0.293590247631073,
+      "learning_rate": 4.8572698307514395e-05,
+      "loss": 0.0058,
+      "step": 16010
+    },
+    {
+      "epoch": 0.9331857633832352,
+      "grad_norm": 0.3748309910297394,
+      "learning_rate": 4.85176059860279e-05,
+      "loss": 0.005,
+      "step": 16020
+    },
+    {
+      "epoch": 0.9337682763441486,
+      "grad_norm": 0.17028909921646118,
+      "learning_rate": 4.846251546578989e-05,
+      "loss": 0.0058,
+      "step": 16030
+    },
+    {
+      "epoch": 0.934350789305062,
+      "grad_norm": 0.29870548844337463,
+      "learning_rate": 4.840742681374058e-05,
+      "loss": 0.0074,
+      "step": 16040
+    },
+    {
+      "epoch": 0.9349333022659754,
+      "grad_norm": 0.24446365237236023,
+      "learning_rate": 4.835234009681787e-05,
+      "loss": 0.0072,
+      "step": 16050
+    },
+    {
+      "epoch": 0.9355158152268888,
+      "grad_norm": 0.1216355413198471,
+      "learning_rate": 4.8297255381957296e-05,
+      "loss": 0.0062,
+      "step": 16060
+    },
+    {
+      "epoch": 0.9360983281878021,
+      "grad_norm": 0.2844870388507843,
+      "learning_rate": 4.824217273609198e-05,
+      "loss": 0.0068,
+      "step": 16070
+    },
+    {
+      "epoch": 0.9366808411487155,
+      "grad_norm": 0.243734672665596,
+      "learning_rate": 4.8187092226152555e-05,
+      "loss": 0.0073,
+      "step": 16080
+    },
+    {
+      "epoch": 0.9372633541096289,
+      "grad_norm": 0.3168551027774811,
+      "learning_rate": 4.8132013919067016e-05,
+      "loss": 0.0058,
+      "step": 16090
+    },
+    {
+      "epoch": 0.9378458670705423,
+      "grad_norm": 0.12927831709384918,
+      "learning_rate": 4.807693788176071e-05,
+      "loss": 0.0067,
+      "step": 16100
+    },
+    {
+      "epoch": 0.9384283800314557,
+      "grad_norm": 0.3183070719242096,
+      "learning_rate": 4.8021864181156214e-05,
+      "loss": 0.0078,
+      "step": 16110
+    },
+    {
+      "epoch": 0.9390108929923691,
+      "grad_norm": 0.16012316942214966,
+      "learning_rate": 4.796679288417326e-05,
+      "loss": 0.0069,
+      "step": 16120
+    },
+    {
+      "epoch": 0.9395934059532824,
+      "grad_norm": 0.1911907196044922,
+      "learning_rate": 4.791172405772866e-05,
+      "loss": 0.005,
+      "step": 16130
+    },
+    {
+      "epoch": 0.9401759189141958,
+      "grad_norm": 0.24279716610908508,
+      "learning_rate": 4.785665776873625e-05,
+      "loss": 0.0066,
+      "step": 16140
+    },
+    {
+      "epoch": 0.9407584318751092,
+      "grad_norm": 0.3492344915866852,
+      "learning_rate": 4.7801594084106763e-05,
+      "loss": 0.0074,
+      "step": 16150
+    },
+    {
+      "epoch": 0.9413409448360226,
+      "grad_norm": 0.17826661467552185,
+      "learning_rate": 4.774653307074775e-05,
+      "loss": 0.01,
+      "step": 16160
+    },
+    {
+      "epoch": 0.941923457796936,
+      "grad_norm": 0.18824392557144165,
+      "learning_rate": 4.769147479556355e-05,
+      "loss": 0.0061,
+      "step": 16170
+    },
+    {
+      "epoch": 0.9425059707578494,
+      "grad_norm": 0.17664392292499542,
+      "learning_rate": 4.763641932545515e-05,
+      "loss": 0.0099,
+      "step": 16180
+    },
+    {
+      "epoch": 0.9430884837187627,
+      "grad_norm": 0.15552131831645966,
+      "learning_rate": 4.758136672732013e-05,
+      "loss": 0.0066,
+      "step": 16190
+    },
+    {
+      "epoch": 0.9436709966796761,
+      "grad_norm": 0.15324902534484863,
+      "learning_rate": 4.752631706805261e-05,
+      "loss": 0.0085,
+      "step": 16200
+    },
+    {
+      "epoch": 0.9442535096405895,
+      "grad_norm": 0.23809918761253357,
+      "learning_rate": 4.7471270414543114e-05,
+      "loss": 0.005,
+      "step": 16210
+    },
+    {
+      "epoch": 0.9448360226015029,
+      "grad_norm": 0.35456669330596924,
+      "learning_rate": 4.741622683367849e-05,
+      "loss": 0.0074,
+      "step": 16220
+    },
+    {
+      "epoch": 0.9454185355624163,
+      "grad_norm": 0.41411086916923523,
+      "learning_rate": 4.736118639234191e-05,
+      "loss": 0.0087,
+      "step": 16230
+    },
+    {
+      "epoch": 0.9460010485233297,
+      "grad_norm": 0.23977859318256378,
+      "learning_rate": 4.730614915741267e-05,
+      "loss": 0.0077,
+      "step": 16240
+    },
+    {
+      "epoch": 0.946583561484243,
+      "grad_norm": 0.15743732452392578,
+      "learning_rate": 4.7251115195766234e-05,
+      "loss": 0.0072,
+      "step": 16250
+    },
+    {
+      "epoch": 0.9471660744451564,
+      "grad_norm": 0.20586641132831573,
+      "learning_rate": 4.719608457427404e-05,
+      "loss": 0.0094,
+      "step": 16260
+    },
+    {
+      "epoch": 0.9477485874060698,
+      "grad_norm": 0.1869913935661316,
+      "learning_rate": 4.714105735980347e-05,
+      "loss": 0.0065,
+      "step": 16270
+    },
+    {
+      "epoch": 0.9483311003669832,
+      "grad_norm": 0.28948545455932617,
+      "learning_rate": 4.70860336192178e-05,
+      "loss": 0.0063,
+      "step": 16280
+    },
+    {
+      "epoch": 0.9489136133278966,
+      "grad_norm": 0.1451376974582672,
+      "learning_rate": 4.7031013419376034e-05,
+      "loss": 0.0088,
+      "step": 16290
+    },
+    {
+      "epoch": 0.94949612628881,
+      "grad_norm": 0.21628732979297638,
+      "learning_rate": 4.6975996827132926e-05,
+      "loss": 0.006,
+      "step": 16300
+    },
+    {
+      "epoch": 0.9500786392497234,
+      "grad_norm": 0.29419586062431335,
+      "learning_rate": 4.692098390933883e-05,
+      "loss": 0.0078,
+      "step": 16310
+    },
+    {
+      "epoch": 0.9506611522106367,
+      "grad_norm": 0.17986848950386047,
+      "learning_rate": 4.6865974732839615e-05,
+      "loss": 0.0055,
+      "step": 16320
+    },
+    {
+      "epoch": 0.95124366517155,
+      "grad_norm": 0.16125009953975677,
+      "learning_rate": 4.6810969364476615e-05,
+      "loss": 0.0074,
+      "step": 16330
+    },
+    {
+      "epoch": 0.9518261781324634,
+      "grad_norm": 0.12655679881572723,
+      "learning_rate": 4.675596787108653e-05,
+      "loss": 0.0057,
+      "step": 16340
+    },
+    {
+      "epoch": 0.9524086910933768,
+      "grad_norm": 0.16673727333545685,
+      "learning_rate": 4.670097031950139e-05,
+      "loss": 0.0065,
+      "step": 16350
+    },
+    {
+      "epoch": 0.9529912040542902,
+      "grad_norm": 0.16212335228919983,
+      "learning_rate": 4.664597677654838e-05,
+      "loss": 0.0064,
+      "step": 16360
+    },
+    {
+      "epoch": 0.9535737170152035,
+      "grad_norm": 0.3109503388404846,
+      "learning_rate": 4.659098730904986e-05,
+      "loss": 0.0064,
+      "step": 16370
+    },
+    {
+      "epoch": 0.9541562299761169,
+      "grad_norm": 0.20093172788619995,
+      "learning_rate": 4.6536001983823204e-05,
+      "loss": 0.0081,
+      "step": 16380
+    },
+    {
+      "epoch": 0.9547387429370303,
+      "grad_norm": 0.18221624195575714,
+      "learning_rate": 4.648102086768078e-05,
+      "loss": 0.0079,
+      "step": 16390
+    },
+    {
+      "epoch": 0.9553212558979437,
+      "grad_norm": 0.15910060703754425,
+      "learning_rate": 4.6426044027429786e-05,
+      "loss": 0.0071,
+      "step": 16400
+    },
+    {
+      "epoch": 0.9559037688588571,
+      "grad_norm": 0.17565271258354187,
+      "learning_rate": 4.637107152987234e-05,
+      "loss": 0.0074,
+      "step": 16410
+    },
+    {
+      "epoch": 0.9564862818197705,
+      "grad_norm": 0.2528422474861145,
+      "learning_rate": 4.631610344180515e-05,
+      "loss": 0.0051,
+      "step": 16420
+    },
+    {
+      "epoch": 0.9570687947806839,
+      "grad_norm": 0.14543932676315308,
+      "learning_rate": 4.626113983001965e-05,
+      "loss": 0.0056,
+      "step": 16430
+    },
+    {
+      "epoch": 0.9576513077415972,
+      "grad_norm": 0.2544792592525482,
+      "learning_rate": 4.620618076130182e-05,
+      "loss": 0.0066,
+      "step": 16440
+    },
+    {
+      "epoch": 0.9582338207025106,
+      "grad_norm": 0.17370153963565826,
+      "learning_rate": 4.6151226302432074e-05,
+      "loss": 0.006,
+      "step": 16450
+    },
+    {
+      "epoch": 0.958816333663424,
+      "grad_norm": 0.3153785467147827,
+      "learning_rate": 4.6096276520185304e-05,
+      "loss": 0.0061,
+      "step": 16460
+    },
+    {
+      "epoch": 0.9593988466243374,
+      "grad_norm": 0.18167081475257874,
+      "learning_rate": 4.604133148133066e-05,
+      "loss": 0.0065,
+      "step": 16470
+    },
+    {
+      "epoch": 0.9599813595852508,
+      "grad_norm": 0.22271473705768585,
+      "learning_rate": 4.598639125263155e-05,
+      "loss": 0.0064,
+      "step": 16480
+    },
+    {
+      "epoch": 0.9605638725461642,
+      "grad_norm": 0.26535695791244507,
+      "learning_rate": 4.593145590084553e-05,
+      "loss": 0.0076,
+      "step": 16490
+    },
+    {
+      "epoch": 0.9611463855070775,
+      "grad_norm": 0.14040829241275787,
+      "learning_rate": 4.58765254927242e-05,
+      "loss": 0.0061,
+      "step": 16500
+    },
+    {
+      "epoch": 0.9617288984679909,
+      "grad_norm": 0.30010026693344116,
+      "learning_rate": 4.582160009501323e-05,
+      "loss": 0.0055,
+      "step": 16510
+    },
+    {
+      "epoch": 0.9623114114289043,
+      "grad_norm": 0.16882310807704926,
+      "learning_rate": 4.5766679774452143e-05,
+      "loss": 0.0061,
+      "step": 16520
+    },
+    {
+      "epoch": 0.9628939243898177,
+      "grad_norm": 0.11858342587947845,
+      "learning_rate": 4.571176459777431e-05,
+      "loss": 0.0067,
+      "step": 16530
+    },
+    {
+      "epoch": 0.9634764373507311,
+      "grad_norm": 0.30074483156204224,
+      "learning_rate": 4.5656854631706846e-05,
+      "loss": 0.0069,
+      "step": 16540
+    },
+    {
+      "epoch": 0.9640589503116445,
+      "grad_norm": 0.3639950752258301,
+      "learning_rate": 4.560194994297054e-05,
+      "loss": 0.0074,
+      "step": 16550
+    },
+    {
+      "epoch": 0.9646414632725578,
+      "grad_norm": 0.26551637053489685,
+      "learning_rate": 4.554705059827974e-05,
+      "loss": 0.007,
+      "step": 16560
+    },
+    {
+      "epoch": 0.9652239762334712,
+      "grad_norm": 0.16254018247127533,
+      "learning_rate": 4.5492156664342365e-05,
+      "loss": 0.0071,
+      "step": 16570
+    },
+    {
+      "epoch": 0.9658064891943846,
+      "grad_norm": 0.23197409510612488,
+      "learning_rate": 4.5437268207859695e-05,
+      "loss": 0.0077,
+      "step": 16580
+    },
+    {
+      "epoch": 0.966389002155298,
+      "grad_norm": 0.24425451457500458,
+      "learning_rate": 4.538238529552641e-05,
+      "loss": 0.0058,
+      "step": 16590
+    },
+    {
+      "epoch": 0.9669715151162114,
+      "grad_norm": 0.22113890945911407,
+      "learning_rate": 4.5327507994030394e-05,
+      "loss": 0.0069,
+      "step": 16600
+    },
+    {
+      "epoch": 0.9675540280771248,
+      "grad_norm": 0.28939270973205566,
+      "learning_rate": 4.527263637005273e-05,
+      "loss": 0.0075,
+      "step": 16610
+    },
+    {
+      "epoch": 0.9681365410380381,
+      "grad_norm": 0.2888289988040924,
+      "learning_rate": 4.521777049026766e-05,
+      "loss": 0.0076,
+      "step": 16620
+    },
+    {
+      "epoch": 0.9687190539989515,
+      "grad_norm": 0.1812407225370407,
+      "learning_rate": 4.5162910421342386e-05,
+      "loss": 0.0057,
+      "step": 16630
+    },
+    {
+      "epoch": 0.9693015669598648,
+      "grad_norm": 0.10864023864269257,
+      "learning_rate": 4.510805622993706e-05,
+      "loss": 0.0062,
+      "step": 16640
+    },
+    {
+      "epoch": 0.9698840799207782,
+      "grad_norm": 0.20587857067584991,
+      "learning_rate": 4.505320798270467e-05,
+      "loss": 0.0074,
+      "step": 16650
+    },
+    {
+      "epoch": 0.9704665928816916,
+      "grad_norm": 0.2696254849433899,
+      "learning_rate": 4.499836574629105e-05,
+      "loss": 0.0056,
+      "step": 16660
+    },
+    {
+      "epoch": 0.971049105842605,
+      "grad_norm": 0.24635377526283264,
+      "learning_rate": 4.494352958733465e-05,
+      "loss": 0.0067,
+      "step": 16670
+    },
+    {
+      "epoch": 0.9716316188035183,
+      "grad_norm": 0.2275347113609314,
+      "learning_rate": 4.488869957246663e-05,
+      "loss": 0.0052,
+      "step": 16680
+    },
+    {
+      "epoch": 0.9722141317644317,
+      "grad_norm": 0.327776163816452,
+      "learning_rate": 4.4833875768310575e-05,
+      "loss": 0.0064,
+      "step": 16690
+    },
+    {
+      "epoch": 0.9727966447253451,
+      "grad_norm": 0.3159472644329071,
+      "learning_rate": 4.47790582414826e-05,
+      "loss": 0.0104,
+      "step": 16700
+    },
+    {
+      "epoch": 0.9733791576862585,
+      "grad_norm": 0.17145466804504395,
+      "learning_rate": 4.472424705859114e-05,
+      "loss": 0.0048,
+      "step": 16710
+    },
+    {
+      "epoch": 0.9739616706471719,
+      "grad_norm": 0.17889347672462463,
+      "learning_rate": 4.466944228623701e-05,
+      "loss": 0.0085,
+      "step": 16720
+    },
+    {
+      "epoch": 0.9745441836080853,
+      "grad_norm": 0.10761959105730057,
+      "learning_rate": 4.461464399101312e-05,
+      "loss": 0.0045,
+      "step": 16730
+    },
+    {
+      "epoch": 0.9751266965689986,
+      "grad_norm": 0.132523775100708,
+      "learning_rate": 4.45598522395046e-05,
+      "loss": 0.0044,
+      "step": 16740
+    },
+    {
+      "epoch": 0.975709209529912,
+      "grad_norm": 0.164432093501091,
+      "learning_rate": 4.4505067098288577e-05,
+      "loss": 0.0048,
+      "step": 16750
+    },
+    {
+      "epoch": 0.9762917224908254,
+      "grad_norm": 0.22944071888923645,
+      "learning_rate": 4.445028863393417e-05,
+      "loss": 0.0061,
+      "step": 16760
+    },
+    {
+      "epoch": 0.9768742354517388,
+      "grad_norm": 0.19161516427993774,
+      "learning_rate": 4.439551691300236e-05,
+      "loss": 0.0089,
+      "step": 16770
+    },
+    {
+      "epoch": 0.9774567484126522,
+      "grad_norm": 0.23397105932235718,
+      "learning_rate": 4.434075200204599e-05,
+      "loss": 0.0065,
+      "step": 16780
+    },
+    {
+      "epoch": 0.9780392613735656,
+      "grad_norm": 0.2313748300075531,
+      "learning_rate": 4.428599396760956e-05,
+      "loss": 0.0064,
+      "step": 16790
+    },
+    {
+      "epoch": 0.9786217743344789,
+      "grad_norm": 0.13568635284900665,
+      "learning_rate": 4.423124287622926e-05,
+      "loss": 0.0066,
+      "step": 16800
+    },
+    {
+      "epoch": 0.9792042872953923,
+      "grad_norm": 0.24646688997745514,
+      "learning_rate": 4.417649879443282e-05,
+      "loss": 0.0085,
+      "step": 16810
+    },
+    {
+      "epoch": 0.9797868002563057,
+      "grad_norm": 0.1909802109003067,
+      "learning_rate": 4.4121761788739444e-05,
+      "loss": 0.0057,
+      "step": 16820
+    },
+    {
+      "epoch": 0.9803693132172191,
+      "grad_norm": 0.2594318389892578,
+      "learning_rate": 4.4067031925659805e-05,
+      "loss": 0.0055,
+      "step": 16830
+    },
+    {
+      "epoch": 0.9809518261781325,
+      "grad_norm": 0.16822053492069244,
+      "learning_rate": 4.4012309271695816e-05,
+      "loss": 0.0049,
+      "step": 16840
+    },
+    {
+      "epoch": 0.9815343391390459,
+      "grad_norm": 0.1937374472618103,
+      "learning_rate": 4.395759389334067e-05,
+      "loss": 0.005,
+      "step": 16850
+    },
+    {
+      "epoch": 0.9821168520999592,
+      "grad_norm": 0.14757700264453888,
+      "learning_rate": 4.3902885857078685e-05,
+      "loss": 0.0046,
+      "step": 16860
+    },
+    {
+      "epoch": 0.9826993650608726,
+      "grad_norm": 0.1825265735387802,
+      "learning_rate": 4.384818522938531e-05,
+      "loss": 0.0054,
+      "step": 16870
+    },
+    {
+      "epoch": 0.983281878021786,
+      "grad_norm": 0.3273537755012512,
+      "learning_rate": 4.379349207672695e-05,
+      "loss": 0.0083,
+      "step": 16880
+    },
+    {
+      "epoch": 0.9838643909826994,
+      "grad_norm": 0.305160254240036,
+      "learning_rate": 4.373880646556098e-05,
+      "loss": 0.0094,
+      "step": 16890
+    },
+    {
+      "epoch": 0.9844469039436128,
+      "grad_norm": 0.20774319767951965,
+      "learning_rate": 4.3684128462335533e-05,
+      "loss": 0.0107,
+      "step": 16900
+    },
+    {
+      "epoch": 0.9850294169045262,
+      "grad_norm": 0.21791402995586395,
+      "learning_rate": 4.362945813348955e-05,
+      "loss": 0.0057,
+      "step": 16910
+    },
+    {
+      "epoch": 0.9856119298654396,
+      "grad_norm": 0.3064537048339844,
+      "learning_rate": 4.357479554545262e-05,
+      "loss": 0.0069,
+      "step": 16920
+    },
+    {
+      "epoch": 0.9861944428263529,
+      "grad_norm": 0.37489670515060425,
+      "learning_rate": 4.352014076464499e-05,
+      "loss": 0.0057,
+      "step": 16930
+    },
+    {
+      "epoch": 0.9867769557872663,
+      "grad_norm": 0.30794015526771545,
+      "learning_rate": 4.346549385747734e-05,
+      "loss": 0.0073,
+      "step": 16940
+    },
+    {
+      "epoch": 0.9873594687481796,
+      "grad_norm": 0.14218463003635406,
+      "learning_rate": 4.34108548903508e-05,
+      "loss": 0.006,
+      "step": 16950
+    },
+    {
+      "epoch": 0.987941981709093,
+      "grad_norm": 0.24668100476264954,
+      "learning_rate": 4.335622392965689e-05,
+      "loss": 0.009,
+      "step": 16960
+    },
+    {
+      "epoch": 0.9885244946700064,
+      "grad_norm": 0.2080397754907608,
+      "learning_rate": 4.330160104177738e-05,
+      "loss": 0.0075,
+      "step": 16970
+    },
+    {
+      "epoch": 0.9891070076309197,
+      "grad_norm": 0.2041449397802353,
+      "learning_rate": 4.32469862930842e-05,
+      "loss": 0.0049,
+      "step": 16980
+    },
+    {
+      "epoch": 0.9896895205918331,
+      "grad_norm": 0.20296189188957214,
+      "learning_rate": 4.319237974993946e-05,
+      "loss": 0.0082,
+      "step": 16990
+    },
+    {
+      "epoch": 0.9902720335527465,
+      "grad_norm": 0.3860246241092682,
+      "learning_rate": 4.3137781478695236e-05,
+      "loss": 0.0081,
+      "step": 17000
+    },
+    {
+      "epoch": 0.9908545465136599,
+      "grad_norm": 0.2490481734275818,
+      "learning_rate": 4.308319154569358e-05,
+      "loss": 0.0055,
+      "step": 17010
+    },
+    {
+      "epoch": 0.9914370594745733,
+      "grad_norm": 0.31504276394844055,
+      "learning_rate": 4.302861001726641e-05,
+      "loss": 0.0076,
+      "step": 17020
+    },
+    {
+      "epoch": 0.9920195724354867,
+      "grad_norm": 0.18228842318058014,
+      "learning_rate": 4.2974036959735415e-05,
+      "loss": 0.006,
+      "step": 17030
+    },
+    {
+      "epoch": 0.9926020853964,
+      "grad_norm": 0.22887305915355682,
+      "learning_rate": 4.2919472439412034e-05,
+      "loss": 0.0059,
+      "step": 17040
+    },
+    {
+      "epoch": 0.9931845983573134,
+      "grad_norm": 0.22207340598106384,
+      "learning_rate": 4.286491652259729e-05,
+      "loss": 0.0053,
+      "step": 17050
+    },
+    {
+      "epoch": 0.9937671113182268,
+      "grad_norm": 0.17153297364711761,
+      "learning_rate": 4.281036927558178e-05,
+      "loss": 0.0066,
+      "step": 17060
+    },
+    {
+      "epoch": 0.9943496242791402,
+      "grad_norm": 0.19197925925254822,
+      "learning_rate": 4.275583076464552e-05,
+      "loss": 0.0064,
+      "step": 17070
+    },
+    {
+      "epoch": 0.9949321372400536,
+      "grad_norm": 0.23953358829021454,
+      "learning_rate": 4.270130105605794e-05,
+      "loss": 0.0081,
+      "step": 17080
+    },
+    {
+      "epoch": 0.995514650200967,
+      "grad_norm": 0.16606943309307098,
+      "learning_rate": 4.264678021607782e-05,
+      "loss": 0.0047,
+      "step": 17090
+    },
+    {
+      "epoch": 0.9960971631618804,
+      "grad_norm": 0.22199587523937225,
+      "learning_rate": 4.25922683109531e-05,
+      "loss": 0.0074,
+      "step": 17100
+    },
+    {
+      "epoch": 0.9966796761227937,
+      "grad_norm": 0.27929314970970154,
+      "learning_rate": 4.25377654069209e-05,
+      "loss": 0.0086,
+      "step": 17110
+    },
+    {
+      "epoch": 0.9972621890837071,
+      "grad_norm": 0.2504722476005554,
+      "learning_rate": 4.248327157020737e-05,
+      "loss": 0.0076,
+      "step": 17120
+    },
+    {
+      "epoch": 0.9978447020446205,
+      "grad_norm": 0.2385127693414688,
+      "learning_rate": 4.242878686702763e-05,
+      "loss": 0.0093,
+      "step": 17130
+    },
+    {
+      "epoch": 0.9984272150055339,
+      "grad_norm": 0.16982513666152954,
+      "learning_rate": 4.23743113635858e-05,
+      "loss": 0.009,
+      "step": 17140
+    },
+    {
+      "epoch": 0.9990097279664473,
+      "grad_norm": 0.18596909940242767,
+      "learning_rate": 4.2319845126074705e-05,
+      "loss": 0.0061,
+      "step": 17150
+    },
+    {
+      "epoch": 0.9995922409273607,
+      "grad_norm": 0.21374642848968506,
+      "learning_rate": 4.226538822067598e-05,
+      "loss": 0.0066,
+      "step": 17160
+    },
+    {
+      "epoch": 1.000174753888274,
+      "grad_norm": 0.234579935669899,
+      "learning_rate": 4.2210940713559896e-05,
+      "loss": 0.0048,
+      "step": 17170
+    },
+    {
+      "epoch": 1.0007572668491873,
+      "grad_norm": 0.21201308071613312,
+      "learning_rate": 4.21565026708853e-05,
+      "loss": 0.0069,
+      "step": 17180
+    },
+    {
+      "epoch": 1.0013397798101007,
+      "grad_norm": 0.22076763212680817,
+      "learning_rate": 4.2102074158799535e-05,
+      "loss": 0.0062,
+      "step": 17190
+    },
+    {
+      "epoch": 1.001922292771014,
+      "grad_norm": 0.16819368302822113,
+      "learning_rate": 4.204765524343841e-05,
+      "loss": 0.0066,
+      "step": 17200
+    },
+    {
+      "epoch": 1.0025048057319275,
+      "grad_norm": 0.2129540890455246,
+      "learning_rate": 4.1993245990926036e-05,
+      "loss": 0.0098,
+      "step": 17210
+    },
+    {
+      "epoch": 1.0030873186928408,
+      "grad_norm": 0.14422816038131714,
+      "learning_rate": 4.193884646737474e-05,
+      "loss": 0.006,
+      "step": 17220
+    },
+    {
+      "epoch": 1.0036698316537542,
+      "grad_norm": 0.28163859248161316,
+      "learning_rate": 4.1884456738885124e-05,
+      "loss": 0.0071,
+      "step": 17230
+    },
+    {
+      "epoch": 1.0042523446146676,
+      "grad_norm": 0.23225603997707367,
+      "learning_rate": 4.1830076871545795e-05,
+      "loss": 0.0063,
+      "step": 17240
+    },
+    {
+      "epoch": 1.004834857575581,
+      "grad_norm": 0.1707264930009842,
+      "learning_rate": 4.177570693143346e-05,
+      "loss": 0.0093,
+      "step": 17250
+    },
+    {
+      "epoch": 1.0054173705364944,
+      "grad_norm": 0.1869906783103943,
+      "learning_rate": 4.172134698461271e-05,
+      "loss": 0.0056,
+      "step": 17260
+    },
+    {
+      "epoch": 1.0059998834974078,
+      "grad_norm": 0.44340330362319946,
+      "learning_rate": 4.166699709713599e-05,
+      "loss": 0.0076,
+      "step": 17270
+    },
+    {
+      "epoch": 1.0065823964583211,
+      "grad_norm": 0.2632408142089844,
+      "learning_rate": 4.1612657335043555e-05,
+      "loss": 0.0077,
+      "step": 17280
+    },
+    {
+      "epoch": 1.0071649094192345,
+      "grad_norm": 0.2275051772594452,
+      "learning_rate": 4.155832776436331e-05,
+      "loss": 0.0056,
+      "step": 17290
+    },
+    {
+      "epoch": 1.007747422380148,
+      "grad_norm": 0.2634100317955017,
+      "learning_rate": 4.1504008451110845e-05,
+      "loss": 0.0086,
+      "step": 17300
+    },
+    {
+      "epoch": 1.0083299353410613,
+      "grad_norm": 0.13787078857421875,
+      "learning_rate": 4.1449699461289225e-05,
+      "loss": 0.0047,
+      "step": 17310
+    },
+    {
+      "epoch": 1.0089124483019747,
+      "grad_norm": 0.23897790908813477,
+      "learning_rate": 4.139540086088901e-05,
+      "loss": 0.0065,
+      "step": 17320
+    },
+    {
+      "epoch": 1.009494961262888,
+      "grad_norm": 0.34493502974510193,
+      "learning_rate": 4.1341112715888106e-05,
+      "loss": 0.0075,
+      "step": 17330
+    },
+    {
+      "epoch": 1.0100774742238015,
+      "grad_norm": 0.30785462260246277,
+      "learning_rate": 4.128683509225172e-05,
+      "loss": 0.0081,
+      "step": 17340
+    },
+    {
+      "epoch": 1.0106599871847148,
+      "grad_norm": 0.1544915735721588,
+      "learning_rate": 4.123256805593231e-05,
+      "loss": 0.007,
+      "step": 17350
+    },
+    {
+      "epoch": 1.0112425001456282,
+      "grad_norm": 0.1695077270269394,
+      "learning_rate": 4.117831167286943e-05,
+      "loss": 0.0072,
+      "step": 17360
+    },
+    {
+      "epoch": 1.0118250131065416,
+      "grad_norm": 0.22155985236167908,
+      "learning_rate": 4.1124066008989685e-05,
+      "loss": 0.0055,
+      "step": 17370
+    },
+    {
+      "epoch": 1.012407526067455,
+      "grad_norm": 0.25313591957092285,
+      "learning_rate": 4.106983113020669e-05,
+      "loss": 0.006,
+      "step": 17380
+    },
+    {
+      "epoch": 1.0129900390283684,
+      "grad_norm": 0.20359335839748383,
+      "learning_rate": 4.101560710242094e-05,
+      "loss": 0.0049,
+      "step": 17390
+    },
+    {
+      "epoch": 1.0135725519892818,
+      "grad_norm": 0.1446719914674759,
+      "learning_rate": 4.096139399151971e-05,
+      "loss": 0.0045,
+      "step": 17400
+    },
+    {
+      "epoch": 1.0141550649501951,
+      "grad_norm": 0.17236465215682983,
+      "learning_rate": 4.090719186337709e-05,
+      "loss": 0.0061,
+      "step": 17410
+    },
+    {
+      "epoch": 1.0147375779111085,
+      "grad_norm": 0.1870393306016922,
+      "learning_rate": 4.0853000783853746e-05,
+      "loss": 0.0057,
+      "step": 17420
+    },
+    {
+      "epoch": 1.015320090872022,
+      "grad_norm": 0.1930486410856247,
+      "learning_rate": 4.0798820818796956e-05,
+      "loss": 0.0092,
+      "step": 17430
+    },
+    {
+      "epoch": 1.0159026038329353,
+      "grad_norm": 0.14709894359111786,
+      "learning_rate": 4.074465203404048e-05,
+      "loss": 0.0051,
+      "step": 17440
+    },
+    {
+      "epoch": 1.0164851167938487,
+      "grad_norm": 0.1655021756887436,
+      "learning_rate": 4.0690494495404495e-05,
+      "loss": 0.0056,
+      "step": 17450
+    },
+    {
+      "epoch": 1.017067629754762,
+      "grad_norm": 0.24993835389614105,
+      "learning_rate": 4.0636348268695535e-05,
+      "loss": 0.0055,
+      "step": 17460
+    },
+    {
+      "epoch": 1.0176501427156754,
+      "grad_norm": 0.28216642141342163,
+      "learning_rate": 4.058221341970637e-05,
+      "loss": 0.0057,
+      "step": 17470
+    },
+    {
+      "epoch": 1.0182326556765888,
+      "grad_norm": 0.18460363149642944,
+      "learning_rate": 4.0528090014215945e-05,
+      "loss": 0.0067,
+      "step": 17480
+    },
+    {
+      "epoch": 1.0188151686375022,
+      "grad_norm": 0.19767507910728455,
+      "learning_rate": 4.047397811798929e-05,
+      "loss": 0.0059,
+      "step": 17490
+    },
+    {
+      "epoch": 1.0193976815984156,
+      "grad_norm": 0.11736376583576202,
+      "learning_rate": 4.041987779677745e-05,
+      "loss": 0.0037,
+      "step": 17500
+    },
+    {
+      "epoch": 1.019980194559329,
+      "grad_norm": 0.10318326205015182,
+      "learning_rate": 4.036578911631745e-05,
+      "loss": 0.0055,
+      "step": 17510
+    },
+    {
+      "epoch": 1.0205627075202424,
+      "grad_norm": 0.11945056170225143,
+      "learning_rate": 4.031171214233211e-05,
+      "loss": 0.0067,
+      "step": 17520
+    },
+    {
+      "epoch": 1.0211452204811557,
+      "grad_norm": 0.261663556098938,
+      "learning_rate": 4.025764694053008e-05,
+      "loss": 0.0068,
+      "step": 17530
+    },
+    {
+      "epoch": 1.0217277334420691,
+      "grad_norm": 0.34986552596092224,
+      "learning_rate": 4.020359357660566e-05,
+      "loss": 0.0097,
+      "step": 17540
+    },
+    {
+      "epoch": 1.0223102464029825,
+      "grad_norm": 0.19175364077091217,
+      "learning_rate": 4.014955211623875e-05,
+      "loss": 0.0074,
+      "step": 17550
+    },
+    {
+      "epoch": 1.022892759363896,
+      "grad_norm": 0.3074852526187897,
+      "learning_rate": 4.0095522625094895e-05,
+      "loss": 0.0077,
+      "step": 17560
+    },
+    {
+      "epoch": 1.0234752723248093,
+      "grad_norm": 0.19195809960365295,
+      "learning_rate": 4.004150516882497e-05,
+      "loss": 0.0068,
+      "step": 17570
+    },
+    {
+      "epoch": 1.0240577852857227,
+      "grad_norm": 0.3464377820491791,
+      "learning_rate": 3.99874998130653e-05,
+      "loss": 0.0069,
+      "step": 17580
+    },
+    {
+      "epoch": 1.024640298246636,
+      "grad_norm": 0.2381429672241211,
+      "learning_rate": 3.9933506623437455e-05,
+      "loss": 0.0056,
+      "step": 17590
+    },
+    {
+      "epoch": 1.0252228112075494,
+      "grad_norm": 0.17219018936157227,
+      "learning_rate": 3.9879525665548276e-05,
+      "loss": 0.0045,
+      "step": 17600
+    },
+    {
+      "epoch": 1.0258053241684628,
+      "grad_norm": 0.15458042919635773,
+      "learning_rate": 3.9825557004989704e-05,
+      "loss": 0.0061,
+      "step": 17610
+    },
+    {
+      "epoch": 1.0263878371293762,
+      "grad_norm": 0.2317482978105545,
+      "learning_rate": 3.977160070733877e-05,
+      "loss": 0.0055,
+      "step": 17620
+    },
+    {
+      "epoch": 1.0269703500902896,
+      "grad_norm": 0.1832754909992218,
+      "learning_rate": 3.971765683815746e-05,
+      "loss": 0.0042,
+      "step": 17630
+    },
+    {
+      "epoch": 1.027552863051203,
+      "grad_norm": 0.23893572390079498,
+      "learning_rate": 3.966372546299265e-05,
+      "loss": 0.0066,
+      "step": 17640
+    },
+    {
+      "epoch": 1.0281353760121164,
+      "grad_norm": 0.21178510785102844,
+      "learning_rate": 3.960980664737604e-05,
+      "loss": 0.0053,
+      "step": 17650
+    },
+    {
+      "epoch": 1.0287178889730297,
+      "grad_norm": 0.2551325857639313,
+      "learning_rate": 3.955590045682408e-05,
+      "loss": 0.0061,
+      "step": 17660
+    },
+    {
+      "epoch": 1.0293004019339431,
+      "grad_norm": 0.3904256522655487,
+      "learning_rate": 3.950200695683788e-05,
+      "loss": 0.0098,
+      "step": 17670
+    },
+    {
+      "epoch": 1.0298829148948565,
+      "grad_norm": 0.2494555115699768,
+      "learning_rate": 3.944812621290314e-05,
+      "loss": 0.0059,
+      "step": 17680
+    },
+    {
+      "epoch": 1.0304654278557699,
+      "grad_norm": 0.17728951573371887,
+      "learning_rate": 3.939425829049002e-05,
+      "loss": 0.0053,
+      "step": 17690
+    },
+    {
+      "epoch": 1.0310479408166833,
+      "grad_norm": 0.30989453196525574,
+      "learning_rate": 3.9340403255053135e-05,
+      "loss": 0.0112,
+      "step": 17700
+    },
+    {
+      "epoch": 1.0316304537775967,
+      "grad_norm": 0.26468968391418457,
+      "learning_rate": 3.9286561172031407e-05,
+      "loss": 0.0068,
+      "step": 17710
+    },
+    {
+      "epoch": 1.03221296673851,
+      "grad_norm": 0.22202187776565552,
+      "learning_rate": 3.923273210684809e-05,
+      "loss": 0.0062,
+      "step": 17720
+    },
+    {
+      "epoch": 1.0327954796994234,
+      "grad_norm": 0.3105335831642151,
+      "learning_rate": 3.9178916124910555e-05,
+      "loss": 0.0096,
+      "step": 17730
+    },
+    {
+      "epoch": 1.0333779926603366,
+      "grad_norm": 0.2187647819519043,
+      "learning_rate": 3.9125113291610274e-05,
+      "loss": 0.0072,
+      "step": 17740
+    },
+    {
+      "epoch": 1.03396050562125,
+      "grad_norm": 0.1771087646484375,
+      "learning_rate": 3.907132367232278e-05,
+      "loss": 0.0051,
+      "step": 17750
+    },
+    {
+      "epoch": 1.0345430185821634,
+      "grad_norm": 0.19385531544685364,
+      "learning_rate": 3.901754733240752e-05,
+      "loss": 0.0055,
+      "step": 17760
+    },
+    {
+      "epoch": 1.0351255315430767,
+      "grad_norm": 0.2566831707954407,
+      "learning_rate": 3.896378433720785e-05,
+      "loss": 0.009,
+      "step": 17770
+    },
+    {
+      "epoch": 1.0357080445039901,
+      "grad_norm": 0.2099108099937439,
+      "learning_rate": 3.891003475205086e-05,
+      "loss": 0.0074,
+      "step": 17780
+    },
+    {
+      "epoch": 1.0362905574649035,
+      "grad_norm": 0.18527786433696747,
+      "learning_rate": 3.885629864224736e-05,
+      "loss": 0.0058,
+      "step": 17790
+    },
+    {
+      "epoch": 1.036873070425817,
+      "grad_norm": 0.19741323590278625,
+      "learning_rate": 3.880257607309178e-05,
+      "loss": 0.007,
+      "step": 17800
+    },
+    {
+      "epoch": 1.0374555833867303,
+      "grad_norm": 0.2898787558078766,
+      "learning_rate": 3.874886710986213e-05,
+      "loss": 0.0056,
+      "step": 17810
+    },
+    {
+      "epoch": 1.0380380963476437,
+      "grad_norm": 0.2288603037595749,
+      "learning_rate": 3.8695171817819834e-05,
+      "loss": 0.0048,
+      "step": 17820
+    },
+    {
+      "epoch": 1.038620609308557,
+      "grad_norm": 0.1266123205423355,
+      "learning_rate": 3.8641490262209774e-05,
+      "loss": 0.0059,
+      "step": 17830
+    },
+    {
+      "epoch": 1.0392031222694704,
+      "grad_norm": 0.21375329792499542,
+      "learning_rate": 3.8587822508260084e-05,
+      "loss": 0.0078,
+      "step": 17840
+    },
+    {
+      "epoch": 1.0397856352303838,
+      "grad_norm": 0.24600611627101898,
+      "learning_rate": 3.853416862118214e-05,
+      "loss": 0.013,
+      "step": 17850
+    },
+    {
+      "epoch": 1.0403681481912972,
+      "grad_norm": 0.13300442695617676,
+      "learning_rate": 3.848052866617049e-05,
+      "loss": 0.0057,
+      "step": 17860
+    },
+    {
+      "epoch": 1.0409506611522106,
+      "grad_norm": 0.25983357429504395,
+      "learning_rate": 3.84269027084027e-05,
+      "loss": 0.0068,
+      "step": 17870
+    },
+    {
+      "epoch": 1.041533174113124,
+      "grad_norm": 0.31247586011886597,
+      "learning_rate": 3.83732908130394e-05,
+      "loss": 0.0058,
+      "step": 17880
+    },
+    {
+      "epoch": 1.0421156870740373,
+      "grad_norm": 0.15630345046520233,
+      "learning_rate": 3.831969304522409e-05,
+      "loss": 0.0054,
+      "step": 17890
+    },
+    {
+      "epoch": 1.0426982000349507,
+      "grad_norm": 0.20724117755889893,
+      "learning_rate": 3.826610947008312e-05,
+      "loss": 0.0071,
+      "step": 17900
+    },
+    {
+      "epoch": 1.0432807129958641,
+      "grad_norm": 0.1527310013771057,
+      "learning_rate": 3.821254015272559e-05,
+      "loss": 0.0045,
+      "step": 17910
+    },
+    {
+      "epoch": 1.0438632259567775,
+      "grad_norm": 0.25436678528785706,
+      "learning_rate": 3.8158985158243214e-05,
+      "loss": 0.0096,
+      "step": 17920
+    },
+    {
+      "epoch": 1.0444457389176909,
+      "grad_norm": 0.16772925853729248,
+      "learning_rate": 3.810544455171044e-05,
+      "loss": 0.0083,
+      "step": 17930
+    },
+    {
+      "epoch": 1.0450282518786043,
+      "grad_norm": 0.15412017703056335,
+      "learning_rate": 3.8051918398184114e-05,
+      "loss": 0.0077,
+      "step": 17940
+    },
+    {
+      "epoch": 1.0456107648395176,
+      "grad_norm": 0.3007606565952301,
+      "learning_rate": 3.799840676270356e-05,
+      "loss": 0.0071,
+      "step": 17950
+    },
+    {
+      "epoch": 1.046193277800431,
+      "grad_norm": 0.32043027877807617,
+      "learning_rate": 3.794490971029048e-05,
+      "loss": 0.0062,
+      "step": 17960
+    },
+    {
+      "epoch": 1.0467757907613444,
+      "grad_norm": 0.1765252947807312,
+      "learning_rate": 3.789142730594881e-05,
+      "loss": 0.01,
+      "step": 17970
+    },
+    {
+      "epoch": 1.0473583037222578,
+      "grad_norm": 0.24289420247077942,
+      "learning_rate": 3.783795961466471e-05,
+      "loss": 0.0054,
+      "step": 17980
+    },
+    {
+      "epoch": 1.0479408166831712,
+      "grad_norm": 0.12668417394161224,
+      "learning_rate": 3.778450670140651e-05,
+      "loss": 0.0073,
+      "step": 17990
+    },
+    {
+      "epoch": 1.0485233296440846,
+      "grad_norm": 0.2219085544347763,
+      "learning_rate": 3.773106863112451e-05,
+      "loss": 0.0062,
+      "step": 18000
+    },
+    {
+      "epoch": 1.049105842604998,
+      "grad_norm": 0.3031097650527954,
+      "learning_rate": 3.7677645468751e-05,
+      "loss": 0.0072,
+      "step": 18010
+    },
+    {
+      "epoch": 1.0496883555659113,
+      "grad_norm": 0.2143469899892807,
+      "learning_rate": 3.762423727920018e-05,
+      "loss": 0.0087,
+      "step": 18020
+    },
+    {
+      "epoch": 1.0502708685268247,
+      "grad_norm": 0.3220011293888092,
+      "learning_rate": 3.7570844127367995e-05,
+      "loss": 0.0076,
+      "step": 18030
+    },
+    {
+      "epoch": 1.050853381487738,
+      "grad_norm": 0.24713626503944397,
+      "learning_rate": 3.751746607813221e-05,
+      "loss": 0.0068,
+      "step": 18040
+    },
+    {
+      "epoch": 1.0514358944486515,
+      "grad_norm": 0.15591208636760712,
+      "learning_rate": 3.746410319635217e-05,
+      "loss": 0.0089,
+      "step": 18050
+    },
+    {
+      "epoch": 1.0520184074095649,
+      "grad_norm": 0.24556222558021545,
+      "learning_rate": 3.7410755546868806e-05,
+      "loss": 0.0079,
+      "step": 18060
+    },
+    {
+      "epoch": 1.0526009203704783,
+      "grad_norm": 0.1497158706188202,
+      "learning_rate": 3.735742319450454e-05,
+      "loss": 0.0048,
+      "step": 18070
+    },
+    {
+      "epoch": 1.0531834333313916,
+      "grad_norm": 0.17919768393039703,
+      "learning_rate": 3.7304106204063184e-05,
+      "loss": 0.0061,
+      "step": 18080
+    },
+    {
+      "epoch": 1.053765946292305,
+      "grad_norm": 0.31438446044921875,
+      "learning_rate": 3.725080464032996e-05,
+      "loss": 0.0051,
+      "step": 18090
+    },
+    {
+      "epoch": 1.0543484592532184,
+      "grad_norm": 0.20861445367336273,
+      "learning_rate": 3.719751856807125e-05,
+      "loss": 0.0057,
+      "step": 18100
+    },
+    {
+      "epoch": 1.0549309722141318,
+      "grad_norm": 0.22839723527431488,
+      "learning_rate": 3.71442480520347e-05,
+      "loss": 0.0071,
+      "step": 18110
+    },
+    {
+      "epoch": 1.0555134851750452,
+      "grad_norm": 0.2174963355064392,
+      "learning_rate": 3.709099315694897e-05,
+      "loss": 0.0069,
+      "step": 18120
+    },
+    {
+      "epoch": 1.0560959981359586,
+      "grad_norm": 0.22441895306110382,
+      "learning_rate": 3.703775394752378e-05,
+      "loss": 0.0045,
+      "step": 18130
+    },
+    {
+      "epoch": 1.056678511096872,
+      "grad_norm": 0.3412150740623474,
+      "learning_rate": 3.698453048844983e-05,
+      "loss": 0.0084,
+      "step": 18140
+    },
+    {
+      "epoch": 1.0572610240577853,
+      "grad_norm": 0.13377846777439117,
+      "learning_rate": 3.693132284439861e-05,
+      "loss": 0.0043,
+      "step": 18150
+    },
+    {
+      "epoch": 1.0578435370186987,
+      "grad_norm": 0.2709502875804901,
+      "learning_rate": 3.6878131080022414e-05,
+      "loss": 0.0072,
+      "step": 18160
+    },
+    {
+      "epoch": 1.058426049979612,
+      "grad_norm": 0.14640028774738312,
+      "learning_rate": 3.682495525995429e-05,
+      "loss": 0.0043,
+      "step": 18170
+    },
+    {
+      "epoch": 1.0590085629405255,
+      "grad_norm": 0.26456964015960693,
+      "learning_rate": 3.6771795448807846e-05,
+      "loss": 0.0072,
+      "step": 18180
+    },
+    {
+      "epoch": 1.0595910759014389,
+      "grad_norm": 0.23677578568458557,
+      "learning_rate": 3.671865171117724e-05,
+      "loss": 0.0074,
+      "step": 18190
+    },
+    {
+      "epoch": 1.0601735888623522,
+      "grad_norm": 0.24363571405410767,
+      "learning_rate": 3.666552411163718e-05,
+      "loss": 0.0057,
+      "step": 18200
+    },
+    {
+      "epoch": 1.0607561018232656,
+      "grad_norm": 0.13092847168445587,
+      "learning_rate": 3.661241271474269e-05,
+      "loss": 0.0034,
+      "step": 18210
+    },
+    {
+      "epoch": 1.061338614784179,
+      "grad_norm": 0.23463112115859985,
+      "learning_rate": 3.6559317585029116e-05,
+      "loss": 0.0054,
+      "step": 18220
+    },
+    {
+      "epoch": 1.0619211277450924,
+      "grad_norm": 0.30116987228393555,
+      "learning_rate": 3.650623878701204e-05,
+      "loss": 0.0088,
+      "step": 18230
+    },
+    {
+      "epoch": 1.0625036407060058,
+      "grad_norm": 0.23397108912467957,
+      "learning_rate": 3.645317638518721e-05,
+      "loss": 0.0046,
+      "step": 18240
+    },
+    {
+      "epoch": 1.0630861536669192,
+      "grad_norm": 0.22369155287742615,
+      "learning_rate": 3.640013044403046e-05,
+      "loss": 0.0076,
+      "step": 18250
+    },
+    {
+      "epoch": 1.0636686666278325,
+      "grad_norm": 0.19742992520332336,
+      "learning_rate": 3.634710102799761e-05,
+      "loss": 0.0059,
+      "step": 18260
+    },
+    {
+      "epoch": 1.064251179588746,
+      "grad_norm": 0.16860133409500122,
+      "learning_rate": 3.6294088201524394e-05,
+      "loss": 0.0058,
+      "step": 18270
+    },
+    {
+      "epoch": 1.0648336925496593,
+      "grad_norm": 0.15085910260677338,
+      "learning_rate": 3.62410920290264e-05,
+      "loss": 0.0044,
+      "step": 18280
+    },
+    {
+      "epoch": 1.0654162055105727,
+      "grad_norm": 0.24134603142738342,
+      "learning_rate": 3.6188112574898954e-05,
+      "loss": 0.0061,
+      "step": 18290
+    },
+    {
+      "epoch": 1.065998718471486,
+      "grad_norm": 0.2757954001426697,
+      "learning_rate": 3.6135149903517115e-05,
+      "loss": 0.0067,
+      "step": 18300
+    },
+    {
+      "epoch": 1.0665812314323995,
+      "grad_norm": 0.22975055873394012,
+      "learning_rate": 3.608220407923552e-05,
+      "loss": 0.0061,
+      "step": 18310
+    },
+    {
+      "epoch": 1.0671637443933129,
+      "grad_norm": 0.10718982666730881,
+      "learning_rate": 3.602927516638833e-05,
+      "loss": 0.0029,
+      "step": 18320
+    },
+    {
+      "epoch": 1.067746257354226,
+      "grad_norm": 0.20827849209308624,
+      "learning_rate": 3.5976363229289165e-05,
+      "loss": 0.0065,
+      "step": 18330
+    },
+    {
+      "epoch": 1.0683287703151394,
+      "grad_norm": 0.12509961426258087,
+      "learning_rate": 3.5923468332231e-05,
+      "loss": 0.0055,
+      "step": 18340
+    },
+    {
+      "epoch": 1.0689112832760528,
+      "grad_norm": 0.10677243769168854,
+      "learning_rate": 3.587059053948616e-05,
+      "loss": 0.0046,
+      "step": 18350
+    },
+    {
+      "epoch": 1.0694937962369662,
+      "grad_norm": 0.21480847895145416,
+      "learning_rate": 3.5817729915306134e-05,
+      "loss": 0.0054,
+      "step": 18360
+    },
+    {
+      "epoch": 1.0700763091978795,
+      "grad_norm": 0.3381551504135132,
+      "learning_rate": 3.5764886523921566e-05,
+      "loss": 0.0077,
+      "step": 18370
+    },
+    {
+      "epoch": 1.070658822158793,
+      "grad_norm": 0.20705369114875793,
+      "learning_rate": 3.571206042954214e-05,
+      "loss": 0.0043,
+      "step": 18380
+    },
+    {
+      "epoch": 1.0712413351197063,
+      "grad_norm": 0.22307080030441284,
+      "learning_rate": 3.5659251696356566e-05,
+      "loss": 0.006,
+      "step": 18390
+    },
+    {
+      "epoch": 1.0718238480806197,
+      "grad_norm": 0.19851557910442352,
+      "learning_rate": 3.56064603885324e-05,
+      "loss": 0.0068,
+      "step": 18400
+    },
+    {
+      "epoch": 1.072406361041533,
+      "grad_norm": 0.1971447765827179,
+      "learning_rate": 3.5553686570216115e-05,
+      "loss": 0.008,
+      "step": 18410
+    },
+    {
+      "epoch": 1.0729888740024465,
+      "grad_norm": 0.17780564725399017,
+      "learning_rate": 3.550093030553284e-05,
+      "loss": 0.0058,
+      "step": 18420
+    },
+    {
+      "epoch": 1.0735713869633599,
+      "grad_norm": 0.21610446274280548,
+      "learning_rate": 3.544819165858642e-05,
+      "loss": 0.0057,
+      "step": 18430
+    },
+    {
+      "epoch": 1.0741538999242732,
+      "grad_norm": 0.11648005247116089,
+      "learning_rate": 3.539547069345926e-05,
+      "loss": 0.0049,
+      "step": 18440
+    },
+    {
+      "epoch": 1.0747364128851866,
+      "grad_norm": 0.1513436883687973,
+      "learning_rate": 3.5342767474212344e-05,
+      "loss": 0.006,
+      "step": 18450
+    },
+    {
+      "epoch": 1.0753189258461,
+      "grad_norm": 0.18297240138053894,
+      "learning_rate": 3.529008206488502e-05,
+      "loss": 0.0081,
+      "step": 18460
+    },
+    {
+      "epoch": 1.0759014388070134,
+      "grad_norm": 0.23625129461288452,
+      "learning_rate": 3.5237414529495055e-05,
+      "loss": 0.0051,
+      "step": 18470
+    },
+    {
+      "epoch": 1.0764839517679268,
+      "grad_norm": 0.16232188045978546,
+      "learning_rate": 3.5184764932038454e-05,
+      "loss": 0.0041,
+      "step": 18480
+    },
+    {
+      "epoch": 1.0770664647288402,
+      "grad_norm": 0.28743577003479004,
+      "learning_rate": 3.513213333648945e-05,
+      "loss": 0.0062,
+      "step": 18490
+    },
+    {
+      "epoch": 1.0776489776897535,
+      "grad_norm": 0.0983772873878479,
+      "learning_rate": 3.5079519806800374e-05,
+      "loss": 0.0063,
+      "step": 18500
+    },
+    {
+      "epoch": 1.078231490650667,
+      "grad_norm": 0.214277446269989,
+      "learning_rate": 3.502692440690165e-05,
+      "loss": 0.0065,
+      "step": 18510
+    },
+    {
+      "epoch": 1.0788140036115803,
+      "grad_norm": 0.2066417932510376,
+      "learning_rate": 3.497434720070165e-05,
+      "loss": 0.0086,
+      "step": 18520
+    },
+    {
+      "epoch": 1.0793965165724937,
+      "grad_norm": 0.3159111738204956,
+      "learning_rate": 3.4921788252086616e-05,
+      "loss": 0.0067,
+      "step": 18530
+    },
+    {
+      "epoch": 1.079979029533407,
+      "grad_norm": 0.2692568004131317,
+      "learning_rate": 3.486924762492065e-05,
+      "loss": 0.0102,
+      "step": 18540
+    },
+    {
+      "epoch": 1.0805615424943205,
+      "grad_norm": 0.1321927011013031,
+      "learning_rate": 3.4816725383045534e-05,
+      "loss": 0.0048,
+      "step": 18550
+    },
+    {
+      "epoch": 1.0811440554552338,
+      "grad_norm": 0.20547805726528168,
+      "learning_rate": 3.476422159028079e-05,
+      "loss": 0.0089,
+      "step": 18560
+    },
+    {
+      "epoch": 1.0817265684161472,
+      "grad_norm": 0.18560943007469177,
+      "learning_rate": 3.471173631042345e-05,
+      "loss": 0.0047,
+      "step": 18570
+    },
+    {
+      "epoch": 1.0823090813770606,
+      "grad_norm": 0.20201492309570312,
+      "learning_rate": 3.465926960724808e-05,
+      "loss": 0.0046,
+      "step": 18580
+    },
+    {
+      "epoch": 1.082891594337974,
+      "grad_norm": 0.2100198119878769,
+      "learning_rate": 3.460682154450666e-05,
+      "loss": 0.0049,
+      "step": 18590
+    },
+    {
+      "epoch": 1.0834741072988874,
+      "grad_norm": 0.44028934836387634,
+      "learning_rate": 3.4554392185928564e-05,
+      "loss": 0.0066,
+      "step": 18600
+    },
+    {
+      "epoch": 1.0840566202598008,
+      "grad_norm": 0.15994203090667725,
+      "learning_rate": 3.450198159522037e-05,
+      "loss": 0.0069,
+      "step": 18610
+    },
+    {
+      "epoch": 1.0846391332207141,
+      "grad_norm": 0.150681734085083,
+      "learning_rate": 3.444958983606592e-05,
+      "loss": 0.0055,
+      "step": 18620
+    },
+    {
+      "epoch": 1.0852216461816275,
+      "grad_norm": 0.1806386560201645,
+      "learning_rate": 3.439721697212612e-05,
+      "loss": 0.0062,
+      "step": 18630
+    },
+    {
+      "epoch": 1.085804159142541,
+      "grad_norm": 0.1805797964334488,
+      "learning_rate": 3.4344863067038954e-05,
+      "loss": 0.004,
+      "step": 18640
+    },
+    {
+      "epoch": 1.0863866721034543,
+      "grad_norm": 0.2374836951494217,
+      "learning_rate": 3.429252818441935e-05,
+      "loss": 0.0053,
+      "step": 18650
+    },
+    {
+      "epoch": 1.0869691850643677,
+      "grad_norm": 0.26689234375953674,
+      "learning_rate": 3.4240212387859094e-05,
+      "loss": 0.0044,
+      "step": 18660
+    },
+    {
+      "epoch": 1.087551698025281,
+      "grad_norm": 0.1469428688287735,
+      "learning_rate": 3.418791574092686e-05,
+      "loss": 0.012,
+      "step": 18670
+    },
+    {
+      "epoch": 1.0881342109861945,
+      "grad_norm": 0.16584601998329163,
+      "learning_rate": 3.413563830716796e-05,
+      "loss": 0.0061,
+      "step": 18680
+    },
+    {
+      "epoch": 1.0887167239471078,
+      "grad_norm": 0.14999045431613922,
+      "learning_rate": 3.408338015010445e-05,
+      "loss": 0.0056,
+      "step": 18690
+    },
+    {
+      "epoch": 1.0892992369080212,
+      "grad_norm": 0.1323958933353424,
+      "learning_rate": 3.40311413332349e-05,
+      "loss": 0.0076,
+      "step": 18700
+    },
+    {
+      "epoch": 1.0898817498689346,
+      "grad_norm": 0.15137185156345367,
+      "learning_rate": 3.3978921920034365e-05,
+      "loss": 0.0057,
+      "step": 18710
+    },
+    {
+      "epoch": 1.090464262829848,
+      "grad_norm": 0.20370100438594818,
+      "learning_rate": 3.392672197395441e-05,
+      "loss": 0.0046,
+      "step": 18720
+    },
+    {
+      "epoch": 1.0910467757907614,
+      "grad_norm": 0.1965283751487732,
+      "learning_rate": 3.387454155842287e-05,
+      "loss": 0.0047,
+      "step": 18730
+    },
+    {
+      "epoch": 1.0916292887516748,
+      "grad_norm": 0.19603459537029266,
+      "learning_rate": 3.382238073684386e-05,
+      "loss": 0.0051,
+      "step": 18740
+    },
+    {
+      "epoch": 1.0922118017125881,
+      "grad_norm": 0.1852051019668579,
+      "learning_rate": 3.377023957259771e-05,
+      "loss": 0.0047,
+      "step": 18750
+    },
+    {
+      "epoch": 1.0927943146735015,
+      "grad_norm": 0.15613462030887604,
+      "learning_rate": 3.3718118129040835e-05,
+      "loss": 0.0048,
+      "step": 18760
+    },
+    {
+      "epoch": 1.093376827634415,
+      "grad_norm": 0.24398194253444672,
+      "learning_rate": 3.3666016469505724e-05,
+      "loss": 0.0053,
+      "step": 18770
+    },
+    {
+      "epoch": 1.0939593405953283,
+      "grad_norm": 0.23396199941635132,
+      "learning_rate": 3.361393465730079e-05,
+      "loss": 0.0054,
+      "step": 18780
+    },
+    {
+      "epoch": 1.0945418535562417,
+      "grad_norm": 0.1656440943479538,
+      "learning_rate": 3.3561872755710366e-05,
+      "loss": 0.0071,
+      "step": 18790
+    },
+    {
+      "epoch": 1.095124366517155,
+      "grad_norm": 0.27500641345977783,
+      "learning_rate": 3.350983082799456e-05,
+      "loss": 0.0057,
+      "step": 18800
+    },
+    {
+      "epoch": 1.0957068794780684,
+      "grad_norm": 0.15668798983097076,
+      "learning_rate": 3.34578089373892e-05,
+      "loss": 0.0059,
+      "step": 18810
+    },
+    {
+      "epoch": 1.0962893924389818,
+      "grad_norm": 0.28912606835365295,
+      "learning_rate": 3.340580714710581e-05,
+      "loss": 0.0055,
+      "step": 18820
+    },
+    {
+      "epoch": 1.0968719053998952,
+      "grad_norm": 0.14135350286960602,
+      "learning_rate": 3.3353825520331466e-05,
+      "loss": 0.0052,
+      "step": 18830
+    },
+    {
+      "epoch": 1.0974544183608086,
+      "grad_norm": 0.31443122029304504,
+      "learning_rate": 3.330186412022876e-05,
+      "loss": 0.0053,
+      "step": 18840
+    },
+    {
+      "epoch": 1.098036931321722,
+      "grad_norm": 0.20390906929969788,
+      "learning_rate": 3.324992300993568e-05,
+      "loss": 0.0061,
+      "step": 18850
+    },
+    {
+      "epoch": 1.0986194442826354,
+      "grad_norm": 0.24383865296840668,
+      "learning_rate": 3.319800225256556e-05,
+      "loss": 0.0054,
+      "step": 18860
+    },
+    {
+      "epoch": 1.0992019572435487,
+      "grad_norm": 0.2769359052181244,
+      "learning_rate": 3.314610191120702e-05,
+      "loss": 0.0053,
+      "step": 18870
+    },
+    {
+      "epoch": 1.0997844702044621,
+      "grad_norm": 0.24231737852096558,
+      "learning_rate": 3.30942220489239e-05,
+      "loss": 0.0063,
+      "step": 18880
+    },
+    {
+      "epoch": 1.1003669831653755,
+      "grad_norm": 0.1262754201889038,
+      "learning_rate": 3.3042362728755086e-05,
+      "loss": 0.006,
+      "step": 18890
+    },
+    {
+      "epoch": 1.100949496126289,
+      "grad_norm": 0.303051233291626,
+      "learning_rate": 3.299052401371456e-05,
+      "loss": 0.0071,
+      "step": 18900
+    },
+    {
+      "epoch": 1.1015320090872023,
+      "grad_norm": 0.20265208184719086,
+      "learning_rate": 3.293870596679125e-05,
+      "loss": 0.0039,
+      "step": 18910
+    },
+    {
+      "epoch": 1.1021145220481157,
+      "grad_norm": 0.3367985486984253,
+      "learning_rate": 3.288690865094895e-05,
+      "loss": 0.0077,
+      "step": 18920
+    },
+    {
+      "epoch": 1.102697035009029,
+      "grad_norm": 0.2820715606212616,
+      "learning_rate": 3.283513212912632e-05,
+      "loss": 0.0075,
+      "step": 18930
+    },
+    {
+      "epoch": 1.1032795479699424,
+      "grad_norm": 0.12723077833652496,
+      "learning_rate": 3.2783376464236684e-05,
+      "loss": 0.0053,
+      "step": 18940
+    },
+    {
+      "epoch": 1.1038620609308558,
+      "grad_norm": 0.15983375906944275,
+      "learning_rate": 3.273164171916806e-05,
+      "loss": 0.0046,
+      "step": 18950
+    },
+    {
+      "epoch": 1.1044445738917692,
+      "grad_norm": 0.45043206214904785,
+      "learning_rate": 3.267992795678306e-05,
+      "loss": 0.0066,
+      "step": 18960
+    },
+    {
+      "epoch": 1.1050270868526826,
+      "grad_norm": 0.10983708500862122,
+      "learning_rate": 3.2628235239918744e-05,
+      "loss": 0.0057,
+      "step": 18970
+    },
+    {
+      "epoch": 1.105609599813596,
+      "grad_norm": 0.2402731031179428,
+      "learning_rate": 3.2576563631386695e-05,
+      "loss": 0.0051,
+      "step": 18980
+    },
+    {
+      "epoch": 1.1061921127745094,
+      "grad_norm": 0.25479036569595337,
+      "learning_rate": 3.252491319397275e-05,
+      "loss": 0.0035,
+      "step": 18990
+    },
+    {
+      "epoch": 1.1067746257354225,
+      "grad_norm": 0.2146436870098114,
+      "learning_rate": 3.247328399043706e-05,
+      "loss": 0.0044,
+      "step": 19000
+    },
+    {
+      "epoch": 1.107357138696336,
+      "grad_norm": 0.20285391807556152,
+      "learning_rate": 3.242167608351399e-05,
+      "loss": 0.0049,
+      "step": 19010
+    },
+    {
+      "epoch": 1.1079396516572493,
+      "grad_norm": 0.348293662071228,
+      "learning_rate": 3.2370089535911986e-05,
+      "loss": 0.0067,
+      "step": 19020
+    },
+    {
+      "epoch": 1.1085221646181627,
+      "grad_norm": 0.2555738389492035,
+      "learning_rate": 3.23185244103136e-05,
+      "loss": 0.0084,
+      "step": 19030
+    },
+    {
+      "epoch": 1.109104677579076,
+      "grad_norm": 0.2087748646736145,
+      "learning_rate": 3.22669807693753e-05,
+      "loss": 0.0052,
+      "step": 19040
+    },
+    {
+      "epoch": 1.1096871905399894,
+      "grad_norm": 0.31942400336265564,
+      "learning_rate": 3.2215458675727495e-05,
+      "loss": 0.0046,
+      "step": 19050
+    },
+    {
+      "epoch": 1.1102697035009028,
+      "grad_norm": 0.14269882440567017,
+      "learning_rate": 3.216395819197438e-05,
+      "loss": 0.0041,
+      "step": 19060
+    },
+    {
+      "epoch": 1.1108522164618162,
+      "grad_norm": 0.2789113223552704,
+      "learning_rate": 3.211247938069387e-05,
+      "loss": 0.005,
+      "step": 19070
+    },
+    {
+      "epoch": 1.1114347294227296,
+      "grad_norm": 0.16928580403327942,
+      "learning_rate": 3.206102230443759e-05,
+      "loss": 0.0049,
+      "step": 19080
+    },
+    {
+      "epoch": 1.112017242383643,
+      "grad_norm": 0.14172187447547913,
+      "learning_rate": 3.2009587025730764e-05,
+      "loss": 0.0046,
+      "step": 19090
+    },
+    {
+      "epoch": 1.1125997553445564,
+      "grad_norm": 0.19109465181827545,
+      "learning_rate": 3.195817360707207e-05,
+      "loss": 0.0047,
+      "step": 19100
+    },
+    {
+      "epoch": 1.1131822683054697,
+      "grad_norm": 0.37268689274787903,
+      "learning_rate": 3.19067821109337e-05,
+      "loss": 0.0059,
+      "step": 19110
+    },
+    {
+      "epoch": 1.1137647812663831,
+      "grad_norm": 0.2517755627632141,
+      "learning_rate": 3.185541259976114e-05,
+      "loss": 0.0047,
+      "step": 19120
+    },
+    {
+      "epoch": 1.1143472942272965,
+      "grad_norm": 0.2742716372013092,
+      "learning_rate": 3.180406513597316e-05,
+      "loss": 0.0062,
+      "step": 19130
+    },
+    {
+      "epoch": 1.1149298071882099,
+      "grad_norm": 0.16271816194057465,
+      "learning_rate": 3.1752739781961835e-05,
+      "loss": 0.0057,
+      "step": 19140
+    },
+    {
+      "epoch": 1.1155123201491233,
+      "grad_norm": 0.2321016490459442,
+      "learning_rate": 3.170143660009228e-05,
+      "loss": 0.0062,
+      "step": 19150
+    },
+    {
+      "epoch": 1.1160948331100367,
+      "grad_norm": 0.23313923180103302,
+      "learning_rate": 3.16501556527027e-05,
+      "loss": 0.0052,
+      "step": 19160
+    },
+    {
+      "epoch": 1.11667734607095,
+      "grad_norm": 0.15437953174114227,
+      "learning_rate": 3.1598897002104265e-05,
+      "loss": 0.0087,
+      "step": 19170
+    },
+    {
+      "epoch": 1.1172598590318634,
+      "grad_norm": 0.18482626974582672,
+      "learning_rate": 3.154766071058108e-05,
+      "loss": 0.005,
+      "step": 19180
+    },
+    {
+      "epoch": 1.1178423719927768,
+      "grad_norm": 0.3312392234802246,
+      "learning_rate": 3.149644684039008e-05,
+      "loss": 0.0054,
+      "step": 19190
+    },
+    {
+      "epoch": 1.1184248849536902,
+      "grad_norm": 0.1431121528148651,
+      "learning_rate": 3.144525545376095e-05,
+      "loss": 0.0053,
+      "step": 19200
+    },
+    {
+      "epoch": 1.1190073979146036,
+      "grad_norm": 0.18121092021465302,
+      "learning_rate": 3.139408661289603e-05,
+      "loss": 0.0043,
+      "step": 19210
+    },
+    {
+      "epoch": 1.119589910875517,
+      "grad_norm": 0.17589549720287323,
+      "learning_rate": 3.134294037997032e-05,
+      "loss": 0.0039,
+      "step": 19220
+    },
+    {
+      "epoch": 1.1201724238364303,
+      "grad_norm": 0.25184574723243713,
+      "learning_rate": 3.129181681713127e-05,
+      "loss": 0.0053,
+      "step": 19230
+    },
+    {
+      "epoch": 1.1207549367973437,
+      "grad_norm": 0.21516335010528564,
+      "learning_rate": 3.1240715986498855e-05,
+      "loss": 0.0062,
+      "step": 19240
+    },
+    {
+      "epoch": 1.121337449758257,
+      "grad_norm": 0.265697717666626,
+      "learning_rate": 3.1189637950165394e-05,
+      "loss": 0.0062,
+      "step": 19250
+    },
+    {
+      "epoch": 1.1219199627191705,
+      "grad_norm": 0.24929842352867126,
+      "learning_rate": 3.1138582770195544e-05,
+      "loss": 0.0053,
+      "step": 19260
+    },
+    {
+      "epoch": 1.1225024756800839,
+      "grad_norm": 0.18948212265968323,
+      "learning_rate": 3.108755050862615e-05,
+      "loss": 0.0043,
+      "step": 19270
+    },
+    {
+      "epoch": 1.1230849886409973,
+      "grad_norm": 0.21419095993041992,
+      "learning_rate": 3.1036541227466204e-05,
+      "loss": 0.0062,
+      "step": 19280
+    },
+    {
+      "epoch": 1.1236675016019106,
+      "grad_norm": 0.32477816939353943,
+      "learning_rate": 3.0985554988696786e-05,
+      "loss": 0.0058,
+      "step": 19290
+    },
+    {
+      "epoch": 1.124250014562824,
+      "grad_norm": 0.20524385571479797,
+      "learning_rate": 3.093459185427102e-05,
+      "loss": 0.0061,
+      "step": 19300
+    },
+    {
+      "epoch": 1.1248325275237374,
+      "grad_norm": 0.24362853169441223,
+      "learning_rate": 3.0883651886113905e-05,
+      "loss": 0.0035,
+      "step": 19310
+    },
+    {
+      "epoch": 1.1254150404846508,
+      "grad_norm": 0.2565145492553711,
+      "learning_rate": 3.0832735146122294e-05,
+      "loss": 0.0042,
+      "step": 19320
+    },
+    {
+      "epoch": 1.1259975534455642,
+      "grad_norm": 0.24529674649238586,
+      "learning_rate": 3.078184169616485e-05,
+      "loss": 0.0056,
+      "step": 19330
+    },
+    {
+      "epoch": 1.1265800664064776,
+      "grad_norm": 0.13250280916690826,
+      "learning_rate": 3.073097159808187e-05,
+      "loss": 0.0055,
+      "step": 19340
+    },
+    {
+      "epoch": 1.127162579367391,
+      "grad_norm": 0.19171005487442017,
+      "learning_rate": 3.068012491368537e-05,
+      "loss": 0.0038,
+      "step": 19350
+    },
+    {
+      "epoch": 1.1277450923283043,
+      "grad_norm": 0.22776301205158234,
+      "learning_rate": 3.062930170475885e-05,
+      "loss": 0.0045,
+      "step": 19360
+    },
+    {
+      "epoch": 1.1283276052892177,
+      "grad_norm": 0.238479346036911,
+      "learning_rate": 3.057850203305729e-05,
+      "loss": 0.0053,
+      "step": 19370
+    },
+    {
+      "epoch": 1.128910118250131,
+      "grad_norm": 0.7057229280471802,
+      "learning_rate": 3.052772596030708e-05,
+      "loss": 0.0057,
+      "step": 19380
+    },
+    {
+      "epoch": 1.1294926312110445,
+      "grad_norm": 0.30547603964805603,
+      "learning_rate": 3.0476973548205943e-05,
+      "loss": 0.0052,
+      "step": 19390
+    },
+    {
+      "epoch": 1.1300751441719579,
+      "grad_norm": 0.18138563632965088,
+      "learning_rate": 3.0426244858422847e-05,
+      "loss": 0.0053,
+      "step": 19400
+    },
+    {
+      "epoch": 1.1306576571328713,
+      "grad_norm": 0.118959441781044,
+      "learning_rate": 3.0375539952597943e-05,
+      "loss": 0.0073,
+      "step": 19410
+    },
+    {
+      "epoch": 1.1312401700937846,
+      "grad_norm": 0.2741149067878723,
+      "learning_rate": 3.0324858892342468e-05,
+      "loss": 0.0041,
+      "step": 19420
+    },
+    {
+      "epoch": 1.131822683054698,
+      "grad_norm": 0.16879117488861084,
+      "learning_rate": 3.0274201739238672e-05,
+      "loss": 0.0041,
+      "step": 19430
+    },
+    {
+      "epoch": 1.1324051960156114,
+      "grad_norm": 0.18931424617767334,
+      "learning_rate": 3.0223568554839786e-05,
+      "loss": 0.0053,
+      "step": 19440
+    },
+    {
+      "epoch": 1.1329877089765248,
+      "grad_norm": 0.18387170135974884,
+      "learning_rate": 3.0172959400669886e-05,
+      "loss": 0.0045,
+      "step": 19450
+    },
+    {
+      "epoch": 1.1335702219374382,
+      "grad_norm": 0.19439192116260529,
+      "learning_rate": 3.0122374338223902e-05,
+      "loss": 0.0045,
+      "step": 19460
+    },
+    {
+      "epoch": 1.1341527348983516,
+      "grad_norm": 0.3560427129268646,
+      "learning_rate": 3.0071813428967427e-05,
+      "loss": 0.0071,
+      "step": 19470
+    },
+    {
+      "epoch": 1.134735247859265,
+      "grad_norm": 0.23722222447395325,
+      "learning_rate": 3.0021276734336746e-05,
+      "loss": 0.0077,
+      "step": 19480
+    },
+    {
+      "epoch": 1.1353177608201783,
+      "grad_norm": 0.21614013612270355,
+      "learning_rate": 2.997076431573871e-05,
+      "loss": 0.0081,
+      "step": 19490
+    },
+    {
+      "epoch": 1.1359002737810917,
+      "grad_norm": 0.21244439482688904,
+      "learning_rate": 2.9920276234550636e-05,
+      "loss": 0.0046,
+      "step": 19500
+    },
+    {
+      "epoch": 1.136482786742005,
+      "grad_norm": 0.2121374011039734,
+      "learning_rate": 2.986981255212035e-05,
+      "loss": 0.0061,
+      "step": 19510
+    },
+    {
+      "epoch": 1.1370652997029185,
+      "grad_norm": 0.11649847030639648,
+      "learning_rate": 2.9819373329765977e-05,
+      "loss": 0.0041,
+      "step": 19520
+    },
+    {
+      "epoch": 1.1376478126638319,
+      "grad_norm": 0.17994952201843262,
+      "learning_rate": 2.9768958628775902e-05,
+      "loss": 0.0071,
+      "step": 19530
+    },
+    {
+      "epoch": 1.1382303256247452,
+      "grad_norm": 0.14891991019248962,
+      "learning_rate": 2.9718568510408762e-05,
+      "loss": 0.0048,
+      "step": 19540
+    },
+    {
+      "epoch": 1.1388128385856586,
+      "grad_norm": 0.24923191964626312,
+      "learning_rate": 2.9668203035893272e-05,
+      "loss": 0.0076,
+      "step": 19550
+    },
+    {
+      "epoch": 1.1393953515465718,
+      "grad_norm": 0.07182957977056503,
+      "learning_rate": 2.9617862266428287e-05,
+      "loss": 0.0047,
+      "step": 19560
+    },
+    {
+      "epoch": 1.1399778645074852,
+      "grad_norm": 0.20004503428936005,
+      "learning_rate": 2.9567546263182556e-05,
+      "loss": 0.0063,
+      "step": 19570
+    },
+    {
+      "epoch": 1.1405603774683986,
+      "grad_norm": 0.25040203332901,
+      "learning_rate": 2.951725508729476e-05,
+      "loss": 0.0068,
+      "step": 19580
+    },
+    {
+      "epoch": 1.141142890429312,
+      "grad_norm": 0.15449707210063934,
+      "learning_rate": 2.946698879987344e-05,
+      "loss": 0.0049,
+      "step": 19590
+    },
+    {
+      "epoch": 1.1417254033902253,
+      "grad_norm": 0.2365003377199173,
+      "learning_rate": 2.9416747461996853e-05,
+      "loss": 0.0046,
+      "step": 19600
+    },
+    {
+      "epoch": 1.1423079163511387,
+      "grad_norm": 0.2035025656223297,
+      "learning_rate": 2.9366531134712973e-05,
+      "loss": 0.003,
+      "step": 19610
+    },
+    {
+      "epoch": 1.142890429312052,
+      "grad_norm": 0.1350264698266983,
+      "learning_rate": 2.9316339879039367e-05,
+      "loss": 0.004,
+      "step": 19620
+    },
+    {
+      "epoch": 1.1434729422729655,
+      "grad_norm": 0.1691318154335022,
+      "learning_rate": 2.9266173755963167e-05,
+      "loss": 0.0045,
+      "step": 19630
+    },
+    {
+      "epoch": 1.1440554552338789,
+      "grad_norm": 0.1569589376449585,
+      "learning_rate": 2.9216032826440926e-05,
+      "loss": 0.0039,
+      "step": 19640
+    },
+    {
+      "epoch": 1.1446379681947922,
+      "grad_norm": 0.1513725370168686,
+      "learning_rate": 2.9165917151398592e-05,
+      "loss": 0.0036,
+      "step": 19650
+    },
+    {
+      "epoch": 1.1452204811557056,
+      "grad_norm": 0.1934337466955185,
+      "learning_rate": 2.9115826791731428e-05,
+      "loss": 0.0049,
+      "step": 19660
+    },
+    {
+      "epoch": 1.145802994116619,
+      "grad_norm": 0.2007901668548584,
+      "learning_rate": 2.906576180830398e-05,
+      "loss": 0.0065,
+      "step": 19670
+    },
+    {
+      "epoch": 1.1463855070775324,
+      "grad_norm": 0.07157131284475327,
+      "learning_rate": 2.9015722261949917e-05,
+      "loss": 0.0037,
+      "step": 19680
+    },
+    {
+      "epoch": 1.1469680200384458,
+      "grad_norm": 0.16172382235527039,
+      "learning_rate": 2.8965708213471986e-05,
+      "loss": 0.0043,
+      "step": 19690
+    },
+    {
+      "epoch": 1.1475505329993592,
+      "grad_norm": 0.2598986327648163,
+      "learning_rate": 2.8915719723641975e-05,
+      "loss": 0.0075,
+      "step": 19700
+    },
+    {
+      "epoch": 1.1481330459602725,
+      "grad_norm": 0.17781129479408264,
+      "learning_rate": 2.8865756853200604e-05,
+      "loss": 0.0072,
+      "step": 19710
+    },
+    {
+      "epoch": 1.148715558921186,
+      "grad_norm": 0.18315806984901428,
+      "learning_rate": 2.8815819662857503e-05,
+      "loss": 0.005,
+      "step": 19720
+    },
+    {
+      "epoch": 1.1492980718820993,
+      "grad_norm": 0.15250816941261292,
+      "learning_rate": 2.876590821329105e-05,
+      "loss": 0.0049,
+      "step": 19730
+    },
+    {
+      "epoch": 1.1498805848430127,
+      "grad_norm": 0.20587261021137238,
+      "learning_rate": 2.871602256514836e-05,
+      "loss": 0.0058,
+      "step": 19740
+    },
+    {
+      "epoch": 1.150463097803926,
+      "grad_norm": 0.08287015557289124,
+      "learning_rate": 2.8666162779045203e-05,
+      "loss": 0.0054,
+      "step": 19750
+    },
+    {
+      "epoch": 1.1510456107648395,
+      "grad_norm": 0.20474036037921906,
+      "learning_rate": 2.8616328915565904e-05,
+      "loss": 0.0055,
+      "step": 19760
+    },
+    {
+      "epoch": 1.1516281237257529,
+      "grad_norm": 0.2332179695367813,
+      "learning_rate": 2.856652103526334e-05,
+      "loss": 0.0053,
+      "step": 19770
+    },
+    {
+      "epoch": 1.1522106366866662,
+      "grad_norm": 0.16414222121238708,
+      "learning_rate": 2.8516739198658755e-05,
+      "loss": 0.0076,
+      "step": 19780
+    },
+    {
+      "epoch": 1.1527931496475796,
+      "grad_norm": 0.25006595253944397,
+      "learning_rate": 2.846698346624177e-05,
+      "loss": 0.0072,
+      "step": 19790
+    },
+    {
+      "epoch": 1.153375662608493,
+      "grad_norm": 0.133356973528862,
+      "learning_rate": 2.841725389847032e-05,
+      "loss": 0.0045,
+      "step": 19800
+    },
+    {
+      "epoch": 1.1539581755694064,
+      "grad_norm": 0.18339315056800842,
+      "learning_rate": 2.8367550555770506e-05,
+      "loss": 0.0059,
+      "step": 19810
+    },
+    {
+      "epoch": 1.1545406885303198,
+      "grad_norm": 0.14272843301296234,
+      "learning_rate": 2.831787349853655e-05,
+      "loss": 0.0034,
+      "step": 19820
+    },
+    {
+      "epoch": 1.1551232014912332,
+      "grad_norm": 0.16190747916698456,
+      "learning_rate": 2.8268222787130806e-05,
+      "loss": 0.0072,
+      "step": 19830
+    },
+    {
+      "epoch": 1.1557057144521465,
+      "grad_norm": 0.15783408284187317,
+      "learning_rate": 2.821859848188355e-05,
+      "loss": 0.0079,
+      "step": 19840
+    },
+    {
+      "epoch": 1.15628822741306,
+      "grad_norm": 0.10485533624887466,
+      "learning_rate": 2.816900064309299e-05,
+      "loss": 0.006,
+      "step": 19850
+    },
+    {
+      "epoch": 1.1568707403739733,
+      "grad_norm": 0.2073519378900528,
+      "learning_rate": 2.811942933102517e-05,
+      "loss": 0.0075,
+      "step": 19860
+    },
+    {
+      "epoch": 1.1574532533348867,
+      "grad_norm": 0.2148457020521164,
+      "learning_rate": 2.806988460591391e-05,
+      "loss": 0.0057,
+      "step": 19870
+    },
+    {
+      "epoch": 1.1580357662958,
+      "grad_norm": 0.20892345905303955,
+      "learning_rate": 2.802036652796074e-05,
+      "loss": 0.0047,
+      "step": 19880
+    },
+    {
+      "epoch": 1.1586182792567135,
+      "grad_norm": 0.21819831430912018,
+      "learning_rate": 2.797087515733478e-05,
+      "loss": 0.0061,
+      "step": 19890
+    },
+    {
+      "epoch": 1.1592007922176268,
+      "grad_norm": 0.16364489495754242,
+      "learning_rate": 2.7921410554172723e-05,
+      "loss": 0.0051,
+      "step": 19900
+    },
+    {
+      "epoch": 1.1597833051785402,
+      "grad_norm": 0.1934242695569992,
+      "learning_rate": 2.787197277857871e-05,
+      "loss": 0.005,
+      "step": 19910
+    },
+    {
+      "epoch": 1.1603658181394536,
+      "grad_norm": 0.21310240030288696,
+      "learning_rate": 2.782256189062429e-05,
+      "loss": 0.0048,
+      "step": 19920
+    },
+    {
+      "epoch": 1.160948331100367,
+      "grad_norm": 0.2592666447162628,
+      "learning_rate": 2.777317795034839e-05,
+      "loss": 0.0037,
+      "step": 19930
+    },
+    {
+      "epoch": 1.1615308440612804,
+      "grad_norm": 0.10502351820468903,
+      "learning_rate": 2.7723821017757112e-05,
+      "loss": 0.0046,
+      "step": 19940
+    },
+    {
+      "epoch": 1.1621133570221938,
+      "grad_norm": 0.237821564078331,
+      "learning_rate": 2.7674491152823822e-05,
+      "loss": 0.0061,
+      "step": 19950
+    },
+    {
+      "epoch": 1.1626958699831071,
+      "grad_norm": 0.31251534819602966,
+      "learning_rate": 2.7625188415488944e-05,
+      "loss": 0.0059,
+      "step": 19960
+    },
+    {
+      "epoch": 1.1632783829440205,
+      "grad_norm": 0.13228774070739746,
+      "learning_rate": 2.7575912865659924e-05,
+      "loss": 0.0045,
+      "step": 19970
+    },
+    {
+      "epoch": 1.163860895904934,
+      "grad_norm": 0.1851041167974472,
+      "learning_rate": 2.7526664563211245e-05,
+      "loss": 0.0047,
+      "step": 19980
+    },
+    {
+      "epoch": 1.1644434088658473,
+      "grad_norm": 0.208220973610878,
+      "learning_rate": 2.7477443567984224e-05,
+      "loss": 0.0059,
+      "step": 19990
+    },
+    {
+      "epoch": 1.1650259218267607,
+      "grad_norm": 0.10171040147542953,
+      "learning_rate": 2.7428249939787e-05,
+      "loss": 0.0047,
+      "step": 20000
+    },
+    {
+      "epoch": 1.165608434787674,
+      "grad_norm": 0.1045743077993393,
+      "learning_rate": 2.7379083738394484e-05,
+      "loss": 0.0041,
+      "step": 20010
+    },
+    {
+      "epoch": 1.1661909477485874,
+      "grad_norm": 0.1751803755760193,
+      "learning_rate": 2.732994502354823e-05,
+      "loss": 0.0048,
+      "step": 20020
+    },
+    {
+      "epoch": 1.1667734607095008,
+      "grad_norm": 0.1258053034543991,
+      "learning_rate": 2.72808338549564e-05,
+      "loss": 0.0053,
+      "step": 20030
+    },
+    {
+      "epoch": 1.1673559736704142,
+      "grad_norm": 0.17347893118858337,
+      "learning_rate": 2.723175029229374e-05,
+      "loss": 0.0052,
+      "step": 20040
+    },
+    {
+      "epoch": 1.1679384866313276,
+      "grad_norm": 0.31824564933776855,
+      "learning_rate": 2.718269439520138e-05,
+      "loss": 0.0039,
+      "step": 20050
+    },
+    {
+      "epoch": 1.168520999592241,
+      "grad_norm": 0.10639124363660812,
+      "learning_rate": 2.7133666223286858e-05,
+      "loss": 0.0051,
+      "step": 20060
+    },
+    {
+      "epoch": 1.1691035125531544,
+      "grad_norm": 0.20031382143497467,
+      "learning_rate": 2.7084665836124006e-05,
+      "loss": 0.0042,
+      "step": 20070
+    },
+    {
+      "epoch": 1.1696860255140678,
+      "grad_norm": 0.24361886084079742,
+      "learning_rate": 2.703569329325296e-05,
+      "loss": 0.0045,
+      "step": 20080
+    },
+    {
+      "epoch": 1.1702685384749811,
+      "grad_norm": 0.32514771819114685,
+      "learning_rate": 2.698674865417994e-05,
+      "loss": 0.0056,
+      "step": 20090
+    },
+    {
+      "epoch": 1.1708510514358945,
+      "grad_norm": 0.23065759241580963,
+      "learning_rate": 2.6937831978377325e-05,
+      "loss": 0.0064,
+      "step": 20100
+    },
+    {
+      "epoch": 1.171433564396808,
+      "grad_norm": 0.212798610329628,
+      "learning_rate": 2.6888943325283485e-05,
+      "loss": 0.0051,
+      "step": 20110
+    },
+    {
+      "epoch": 1.1720160773577213,
+      "grad_norm": 0.2243419736623764,
+      "learning_rate": 2.6840082754302733e-05,
+      "loss": 0.0042,
+      "step": 20120
+    },
+    {
+      "epoch": 1.1725985903186347,
+      "grad_norm": 0.26303279399871826,
+      "learning_rate": 2.679125032480525e-05,
+      "loss": 0.004,
+      "step": 20130
+    },
+    {
+      "epoch": 1.173181103279548,
+      "grad_norm": 0.13998739421367645,
+      "learning_rate": 2.674244609612708e-05,
+      "loss": 0.0047,
+      "step": 20140
+    },
+    {
+      "epoch": 1.1737636162404614,
+      "grad_norm": 0.12245301902294159,
+      "learning_rate": 2.6693670127569958e-05,
+      "loss": 0.0061,
+      "step": 20150
+    },
+    {
+      "epoch": 1.1743461292013748,
+      "grad_norm": 0.17332442104816437,
+      "learning_rate": 2.6644922478401268e-05,
+      "loss": 0.0058,
+      "step": 20160
+    },
+    {
+      "epoch": 1.1749286421622882,
+      "grad_norm": 0.12389328330755234,
+      "learning_rate": 2.6596203207854004e-05,
+      "loss": 0.0042,
+      "step": 20170
+    },
+    {
+      "epoch": 1.1755111551232016,
+      "grad_norm": 0.10340277850627899,
+      "learning_rate": 2.654751237512666e-05,
+      "loss": 0.0064,
+      "step": 20180
+    },
+    {
+      "epoch": 1.176093668084115,
+      "grad_norm": 0.20555490255355835,
+      "learning_rate": 2.649885003938323e-05,
+      "loss": 0.0048,
+      "step": 20190
+    },
+    {
+      "epoch": 1.1766761810450284,
+      "grad_norm": 0.20423851907253265,
+      "learning_rate": 2.6450216259753e-05,
+      "loss": 0.0122,
+      "step": 20200
+    },
+    {
+      "epoch": 1.1772586940059417,
+      "grad_norm": 0.2009936422109604,
+      "learning_rate": 2.640161109533063e-05,
+      "loss": 0.0057,
+      "step": 20210
+    },
+    {
+      "epoch": 1.1778412069668551,
+      "grad_norm": 0.20434145629405975,
+      "learning_rate": 2.6353034605175936e-05,
+      "loss": 0.0052,
+      "step": 20220
+    },
+    {
+      "epoch": 1.1784237199277685,
+      "grad_norm": 0.10659831762313843,
+      "learning_rate": 2.630448684831398e-05,
+      "loss": 0.0042,
+      "step": 20230
+    },
+    {
+      "epoch": 1.179006232888682,
+      "grad_norm": 0.40230634808540344,
+      "learning_rate": 2.625596788373482e-05,
+      "loss": 0.0054,
+      "step": 20240
+    },
+    {
+      "epoch": 1.1795887458495953,
+      "grad_norm": 0.17268060147762299,
+      "learning_rate": 2.620747777039363e-05,
+      "loss": 0.0039,
+      "step": 20250
+    },
+    {
+      "epoch": 1.1801712588105084,
+      "grad_norm": 0.14492864906787872,
+      "learning_rate": 2.6159016567210424e-05,
+      "loss": 0.0061,
+      "step": 20260
+    },
+    {
+      "epoch": 1.1807537717714218,
+      "grad_norm": 0.16161669790744781,
+      "learning_rate": 2.6110584333070153e-05,
+      "loss": 0.0056,
+      "step": 20270
+    },
+    {
+      "epoch": 1.1813362847323352,
+      "grad_norm": 0.2991522252559662,
+      "learning_rate": 2.6062181126822537e-05,
+      "loss": 0.0037,
+      "step": 20280
+    },
+    {
+      "epoch": 1.1819187976932486,
+      "grad_norm": 0.2766414284706116,
+      "learning_rate": 2.601380700728203e-05,
+      "loss": 0.0039,
+      "step": 20290
+    },
+    {
+      "epoch": 1.182501310654162,
+      "grad_norm": 0.19769255816936493,
+      "learning_rate": 2.596546203322777e-05,
+      "loss": 0.005,
+      "step": 20300
+    },
+    {
+      "epoch": 1.1830838236150754,
+      "grad_norm": 0.2686941623687744,
+      "learning_rate": 2.5917146263403457e-05,
+      "loss": 0.0044,
+      "step": 20310
+    },
+    {
+      "epoch": 1.1836663365759887,
+      "grad_norm": 0.18222245573997498,
+      "learning_rate": 2.5868859756517296e-05,
+      "loss": 0.0045,
+      "step": 20320
+    },
+    {
+      "epoch": 1.1842488495369021,
+      "grad_norm": 0.30395883321762085,
+      "learning_rate": 2.5820602571241947e-05,
+      "loss": 0.0051,
+      "step": 20330
+    },
+    {
+      "epoch": 1.1848313624978155,
+      "grad_norm": 0.1746487021446228,
+      "learning_rate": 2.577237476621442e-05,
+      "loss": 0.0077,
+      "step": 20340
+    },
+    {
+      "epoch": 1.185413875458729,
+      "grad_norm": 0.16367869079113007,
+      "learning_rate": 2.572417640003609e-05,
+      "loss": 0.0058,
+      "step": 20350
+    },
+    {
+      "epoch": 1.1859963884196423,
+      "grad_norm": 0.31145015358924866,
+      "learning_rate": 2.5676007531272473e-05,
+      "loss": 0.0061,
+      "step": 20360
+    },
+    {
+      "epoch": 1.1865789013805557,
+      "grad_norm": 0.22405989468097687,
+      "learning_rate": 2.5627868218453332e-05,
+      "loss": 0.005,
+      "step": 20370
+    },
+    {
+      "epoch": 1.187161414341469,
+      "grad_norm": 0.16974937915802002,
+      "learning_rate": 2.5579758520072443e-05,
+      "loss": 0.0039,
+      "step": 20380
+    },
+    {
+      "epoch": 1.1877439273023824,
+      "grad_norm": 0.34971392154693604,
+      "learning_rate": 2.5531678494587615e-05,
+      "loss": 0.0078,
+      "step": 20390
+    },
+    {
+      "epoch": 1.1883264402632958,
+      "grad_norm": 0.1582673043012619,
+      "learning_rate": 2.5483628200420644e-05,
+      "loss": 0.0035,
+      "step": 20400
+    },
+    {
+      "epoch": 1.1889089532242092,
+      "grad_norm": 0.2356518656015396,
+      "learning_rate": 2.5435607695957154e-05,
+      "loss": 0.0068,
+      "step": 20410
+    },
+    {
+      "epoch": 1.1894914661851226,
+      "grad_norm": 0.16864237189292908,
+      "learning_rate": 2.5387617039546586e-05,
+      "loss": 0.0038,
+      "step": 20420
+    },
+    {
+      "epoch": 1.190073979146036,
+      "grad_norm": 0.3198424279689789,
+      "learning_rate": 2.5339656289502106e-05,
+      "loss": 0.0056,
+      "step": 20430
+    },
+    {
+      "epoch": 1.1906564921069493,
+      "grad_norm": 0.2680354118347168,
+      "learning_rate": 2.529172550410056e-05,
+      "loss": 0.0045,
+      "step": 20440
+    },
+    {
+      "epoch": 1.1912390050678627,
+      "grad_norm": 0.1877739280462265,
+      "learning_rate": 2.524382474158234e-05,
+      "loss": 0.0052,
+      "step": 20450
+    },
+    {
+      "epoch": 1.1918215180287761,
+      "grad_norm": 0.10875384509563446,
+      "learning_rate": 2.5195954060151433e-05,
+      "loss": 0.0059,
+      "step": 20460
+    },
+    {
+      "epoch": 1.1924040309896895,
+      "grad_norm": 0.2688278555870056,
+      "learning_rate": 2.5148113517975212e-05,
+      "loss": 0.0052,
+      "step": 20470
+    },
+    {
+      "epoch": 1.1929865439506029,
+      "grad_norm": 0.23045240342617035,
+      "learning_rate": 2.5100303173184447e-05,
+      "loss": 0.0051,
+      "step": 20480
+    },
+    {
+      "epoch": 1.1935690569115163,
+      "grad_norm": 0.2154732495546341,
+      "learning_rate": 2.5052523083873196e-05,
+      "loss": 0.0032,
+      "step": 20490
+    },
+    {
+      "epoch": 1.1941515698724297,
+      "grad_norm": 0.2016867846250534,
+      "learning_rate": 2.500477330809881e-05,
+      "loss": 0.0051,
+      "step": 20500
+    },
+    {
+      "epoch": 1.194734082833343,
+      "grad_norm": 0.138419508934021,
+      "learning_rate": 2.4957053903881733e-05,
+      "loss": 0.0066,
+      "step": 20510
+    },
+    {
+      "epoch": 1.1953165957942564,
+      "grad_norm": 0.15247371792793274,
+      "learning_rate": 2.4909364929205576e-05,
+      "loss": 0.004,
+      "step": 20520
+    },
+    {
+      "epoch": 1.1958991087551698,
+      "grad_norm": 0.2212262898683548,
+      "learning_rate": 2.4861706442016923e-05,
+      "loss": 0.0051,
+      "step": 20530
+    },
+    {
+      "epoch": 1.1964816217160832,
+      "grad_norm": 0.5201049447059631,
+      "learning_rate": 2.481407850022533e-05,
+      "loss": 0.0064,
+      "step": 20540
+    },
+    {
+      "epoch": 1.1970641346769966,
+      "grad_norm": 0.16494764387607574,
+      "learning_rate": 2.4766481161703213e-05,
+      "loss": 0.0057,
+      "step": 20550
+    },
+    {
+      "epoch": 1.19764664763791,
+      "grad_norm": 0.20049002766609192,
+      "learning_rate": 2.4718914484285876e-05,
+      "loss": 0.0067,
+      "step": 20560
+    },
+    {
+      "epoch": 1.1982291605988233,
+      "grad_norm": 0.29536816477775574,
+      "learning_rate": 2.467137852577129e-05,
+      "loss": 0.006,
+      "step": 20570
+    },
+    {
+      "epoch": 1.1988116735597367,
+      "grad_norm": 0.156757652759552,
+      "learning_rate": 2.4623873343920123e-05,
+      "loss": 0.0062,
+      "step": 20580
+    },
+    {
+      "epoch": 1.19939418652065,
+      "grad_norm": 0.18742325901985168,
+      "learning_rate": 2.4576398996455656e-05,
+      "loss": 0.0046,
+      "step": 20590
+    },
+    {
+      "epoch": 1.1999766994815635,
+      "grad_norm": 0.31452953815460205,
+      "learning_rate": 2.452895554106368e-05,
+      "loss": 0.0068,
+      "step": 20600
+    },
+    {
+      "epoch": 1.2005592124424769,
+      "grad_norm": 0.17321531474590302,
+      "learning_rate": 2.4481543035392506e-05,
+      "loss": 0.0047,
+      "step": 20610
+    },
+    {
+      "epoch": 1.2011417254033903,
+      "grad_norm": 0.09139826148748398,
+      "learning_rate": 2.4434161537052774e-05,
+      "loss": 0.0039,
+      "step": 20620
+    },
+    {
+      "epoch": 1.2017242383643036,
+      "grad_norm": 0.23709987103939056,
+      "learning_rate": 2.4386811103617474e-05,
+      "loss": 0.0043,
+      "step": 20630
+    },
+    {
+      "epoch": 1.202306751325217,
+      "grad_norm": 0.2732355296611786,
+      "learning_rate": 2.4339491792621833e-05,
+      "loss": 0.0041,
+      "step": 20640
+    },
+    {
+      "epoch": 1.2028892642861304,
+      "grad_norm": 0.24540017545223236,
+      "learning_rate": 2.4292203661563312e-05,
+      "loss": 0.0037,
+      "step": 20650
+    },
+    {
+      "epoch": 1.2034717772470438,
+      "grad_norm": 0.16779111325740814,
+      "learning_rate": 2.424494676790141e-05,
+      "loss": 0.0065,
+      "step": 20660
+    },
+    {
+      "epoch": 1.2040542902079572,
+      "grad_norm": 0.18444012105464935,
+      "learning_rate": 2.4197721169057745e-05,
+      "loss": 0.0048,
+      "step": 20670
+    },
+    {
+      "epoch": 1.2046368031688706,
+      "grad_norm": 0.3186444640159607,
+      "learning_rate": 2.4150526922415857e-05,
+      "loss": 0.0039,
+      "step": 20680
+    },
+    {
+      "epoch": 1.205219316129784,
+      "grad_norm": 0.25987476110458374,
+      "learning_rate": 2.41033640853212e-05,
+      "loss": 0.0057,
+      "step": 20690
+    },
+    {
+      "epoch": 1.2058018290906973,
+      "grad_norm": 0.21384380757808685,
+      "learning_rate": 2.405623271508108e-05,
+      "loss": 0.0041,
+      "step": 20700
+    },
+    {
+      "epoch": 1.2063843420516107,
+      "grad_norm": 0.18099993467330933,
+      "learning_rate": 2.4009132868964522e-05,
+      "loss": 0.0051,
+      "step": 20710
+    },
+    {
+      "epoch": 1.206966855012524,
+      "grad_norm": 0.12152326852083206,
+      "learning_rate": 2.3962064604202327e-05,
+      "loss": 0.0044,
+      "step": 20720
+    },
+    {
+      "epoch": 1.2075493679734375,
+      "grad_norm": 0.2329428791999817,
+      "learning_rate": 2.3915027977986858e-05,
+      "loss": 0.0049,
+      "step": 20730
+    },
+    {
+      "epoch": 1.2081318809343509,
+      "grad_norm": 0.1266992688179016,
+      "learning_rate": 2.386802304747205e-05,
+      "loss": 0.0048,
+      "step": 20740
+    },
+    {
+      "epoch": 1.2087143938952642,
+      "grad_norm": 0.12381075322628021,
+      "learning_rate": 2.382104986977332e-05,
+      "loss": 0.0033,
+      "step": 20750
+    },
+    {
+      "epoch": 1.2092969068561776,
+      "grad_norm": 0.20725421607494354,
+      "learning_rate": 2.377410850196749e-05,
+      "loss": 0.0036,
+      "step": 20760
+    },
+    {
+      "epoch": 1.209879419817091,
+      "grad_norm": 0.19337572157382965,
+      "learning_rate": 2.3727199001092797e-05,
+      "loss": 0.0046,
+      "step": 20770
+    },
+    {
+      "epoch": 1.2104619327780044,
+      "grad_norm": 0.17137841880321503,
+      "learning_rate": 2.3680321424148677e-05,
+      "loss": 0.0059,
+      "step": 20780
+    },
+    {
+      "epoch": 1.2110444457389178,
+      "grad_norm": 0.19310294091701508,
+      "learning_rate": 2.3633475828095793e-05,
+      "loss": 0.0068,
+      "step": 20790
+    },
+    {
+      "epoch": 1.211626958699831,
+      "grad_norm": 0.22379134595394135,
+      "learning_rate": 2.358666226985599e-05,
+      "loss": 0.0065,
+      "step": 20800
+    },
+    {
+      "epoch": 1.2122094716607443,
+      "grad_norm": 0.20557384192943573,
+      "learning_rate": 2.353988080631213e-05,
+      "loss": 0.0041,
+      "step": 20810
+    },
+    {
+      "epoch": 1.2127919846216577,
+      "grad_norm": 0.18650870025157928,
+      "learning_rate": 2.3493131494308142e-05,
+      "loss": 0.0046,
+      "step": 20820
+    },
+    {
+      "epoch": 1.213374497582571,
+      "grad_norm": 0.25925755500793457,
+      "learning_rate": 2.344641439064881e-05,
+      "loss": 0.0038,
+      "step": 20830
+    },
+    {
+      "epoch": 1.2139570105434845,
+      "grad_norm": 0.31177136301994324,
+      "learning_rate": 2.339972955209984e-05,
+      "loss": 0.0041,
+      "step": 20840
+    },
+    {
+      "epoch": 1.2145395235043979,
+      "grad_norm": 0.18946701288223267,
+      "learning_rate": 2.335307703538771e-05,
+      "loss": 0.0045,
+      "step": 20850
+    },
+    {
+      "epoch": 1.2151220364653113,
+      "grad_norm": 0.11377757042646408,
+      "learning_rate": 2.330645689719962e-05,
+      "loss": 0.0026,
+      "step": 20860
+    },
+    {
+      "epoch": 1.2157045494262246,
+      "grad_norm": 0.15649139881134033,
+      "learning_rate": 2.3259869194183414e-05,
+      "loss": 0.0037,
+      "step": 20870
+    },
+    {
+      "epoch": 1.216287062387138,
+      "grad_norm": 0.23433303833007812,
+      "learning_rate": 2.321331398294759e-05,
+      "loss": 0.0036,
+      "step": 20880
+    },
+    {
+      "epoch": 1.2168695753480514,
+      "grad_norm": 0.3228027820587158,
+      "learning_rate": 2.3166791320061094e-05,
+      "loss": 0.0039,
+      "step": 20890
+    },
+    {
+      "epoch": 1.2174520883089648,
+      "grad_norm": 0.13117103278636932,
+      "learning_rate": 2.312030126205335e-05,
+      "loss": 0.0045,
+      "step": 20900
+    },
+    {
+      "epoch": 1.2180346012698782,
+      "grad_norm": 0.18662893772125244,
+      "learning_rate": 2.3073843865414162e-05,
+      "loss": 0.0047,
+      "step": 20910
+    },
+    {
+      "epoch": 1.2186171142307916,
+      "grad_norm": 0.31538307666778564,
+      "learning_rate": 2.302741918659363e-05,
+      "loss": 0.0059,
+      "step": 20920
+    },
+    {
+      "epoch": 1.219199627191705,
+      "grad_norm": 0.1715722680091858,
+      "learning_rate": 2.2981027282002154e-05,
+      "loss": 0.004,
+      "step": 20930
+    },
+    {
+      "epoch": 1.2197821401526183,
+      "grad_norm": 0.28882381319999695,
+      "learning_rate": 2.2934668208010236e-05,
+      "loss": 0.005,
+      "step": 20940
+    },
+    {
+      "epoch": 1.2203646531135317,
+      "grad_norm": 0.10931084305047989,
+      "learning_rate": 2.2888342020948555e-05,
+      "loss": 0.0045,
+      "step": 20950
+    },
+    {
+      "epoch": 1.220947166074445,
+      "grad_norm": 0.13048085570335388,
+      "learning_rate": 2.2842048777107783e-05,
+      "loss": 0.0044,
+      "step": 20960
+    },
+    {
+      "epoch": 1.2215296790353585,
+      "grad_norm": 0.17171211540699005,
+      "learning_rate": 2.2795788532738553e-05,
+      "loss": 0.0059,
+      "step": 20970
+    },
+    {
+      "epoch": 1.2221121919962719,
+      "grad_norm": 0.17031100392341614,
+      "learning_rate": 2.2749561344051468e-05,
+      "loss": 0.0051,
+      "step": 20980
+    },
+    {
+      "epoch": 1.2226947049571852,
+      "grad_norm": 0.2068527638912201,
+      "learning_rate": 2.2703367267216896e-05,
+      "loss": 0.0062,
+      "step": 20990
+    },
+    {
+      "epoch": 1.2232772179180986,
+      "grad_norm": 0.18442004919052124,
+      "learning_rate": 2.2657206358364997e-05,
+      "loss": 0.008,
+      "step": 21000
+    },
+    {
+      "epoch": 1.223859730879012,
+      "grad_norm": 0.15710611641407013,
+      "learning_rate": 2.261107867358563e-05,
+      "loss": 0.0031,
+      "step": 21010
+    },
+    {
+      "epoch": 1.2244422438399254,
+      "grad_norm": 0.14485406875610352,
+      "learning_rate": 2.2564984268928263e-05,
+      "loss": 0.0033,
+      "step": 21020
+    },
+    {
+      "epoch": 1.2250247568008388,
+      "grad_norm": 0.08650320768356323,
+      "learning_rate": 2.251892320040198e-05,
+      "loss": 0.0064,
+      "step": 21030
+    },
+    {
+      "epoch": 1.2256072697617522,
+      "grad_norm": 0.2814912497997284,
+      "learning_rate": 2.2472895523975312e-05,
+      "loss": 0.0033,
+      "step": 21040
+    },
+    {
+      "epoch": 1.2261897827226655,
+      "grad_norm": 0.14940081536769867,
+      "learning_rate": 2.2426901295576212e-05,
+      "loss": 0.0046,
+      "step": 21050
+    },
+    {
+      "epoch": 1.226772295683579,
+      "grad_norm": 0.2938425838947296,
+      "learning_rate": 2.2380940571092013e-05,
+      "loss": 0.0055,
+      "step": 21060
+    },
+    {
+      "epoch": 1.2273548086444923,
+      "grad_norm": 0.2667405903339386,
+      "learning_rate": 2.2335013406369302e-05,
+      "loss": 0.0036,
+      "step": 21070
+    },
+    {
+      "epoch": 1.2279373216054057,
+      "grad_norm": 0.10116902738809586,
+      "learning_rate": 2.228911985721397e-05,
+      "loss": 0.0086,
+      "step": 21080
+    },
+    {
+      "epoch": 1.228519834566319,
+      "grad_norm": 0.2843458354473114,
+      "learning_rate": 2.224325997939095e-05,
+      "loss": 0.0043,
+      "step": 21090
+    },
+    {
+      "epoch": 1.2291023475272325,
+      "grad_norm": 0.21746480464935303,
+      "learning_rate": 2.219743382862437e-05,
+      "loss": 0.0049,
+      "step": 21100
+    },
+    {
+      "epoch": 1.2296848604881458,
+      "grad_norm": 0.18752557039260864,
+      "learning_rate": 2.2151641460597294e-05,
+      "loss": 0.0045,
+      "step": 21110
+    },
+    {
+      "epoch": 1.2302673734490592,
+      "grad_norm": 0.11562065780162811,
+      "learning_rate": 2.210588293095177e-05,
+      "loss": 0.0038,
+      "step": 21120
+    },
+    {
+      "epoch": 1.2308498864099726,
+      "grad_norm": 0.2527402937412262,
+      "learning_rate": 2.2060158295288714e-05,
+      "loss": 0.0056,
+      "step": 21130
+    },
+    {
+      "epoch": 1.231432399370886,
+      "grad_norm": 0.20256440341472626,
+      "learning_rate": 2.2014467609167905e-05,
+      "loss": 0.0057,
+      "step": 21140
+    },
+    {
+      "epoch": 1.2320149123317994,
+      "grad_norm": 0.1765642911195755,
+      "learning_rate": 2.1968810928107806e-05,
+      "loss": 0.0043,
+      "step": 21150
+    },
+    {
+      "epoch": 1.2325974252927128,
+      "grad_norm": 0.1241622269153595,
+      "learning_rate": 2.1923188307585606e-05,
+      "loss": 0.0037,
+      "step": 21160
+    },
+    {
+      "epoch": 1.2331799382536262,
+      "grad_norm": 0.1839839369058609,
+      "learning_rate": 2.187759980303708e-05,
+      "loss": 0.0063,
+      "step": 21170
+    },
+    {
+      "epoch": 1.2337624512145395,
+      "grad_norm": 0.15432360768318176,
+      "learning_rate": 2.1832045469856545e-05,
+      "loss": 0.0039,
+      "step": 21180
+    },
+    {
+      "epoch": 1.234344964175453,
+      "grad_norm": 0.15211918950080872,
+      "learning_rate": 2.1786525363396844e-05,
+      "loss": 0.004,
+      "step": 21190
+    },
+    {
+      "epoch": 1.2349274771363663,
+      "grad_norm": 0.17500174045562744,
+      "learning_rate": 2.1741039538969183e-05,
+      "loss": 0.0042,
+      "step": 21200
+    },
+    {
+      "epoch": 1.2355099900972797,
+      "grad_norm": 0.16297830641269684,
+      "learning_rate": 2.169558805184313e-05,
+      "loss": 0.0041,
+      "step": 21210
+    },
+    {
+      "epoch": 1.236092503058193,
+      "grad_norm": 0.23848998546600342,
+      "learning_rate": 2.165017095724651e-05,
+      "loss": 0.0058,
+      "step": 21220
+    },
+    {
+      "epoch": 1.2366750160191065,
+      "grad_norm": 0.18026289343833923,
+      "learning_rate": 2.1604788310365405e-05,
+      "loss": 0.0058,
+      "step": 21230
+    },
+    {
+      "epoch": 1.2372575289800198,
+      "grad_norm": 0.17681477963924408,
+      "learning_rate": 2.155944016634398e-05,
+      "loss": 0.0046,
+      "step": 21240
+    },
+    {
+      "epoch": 1.2378400419409332,
+      "grad_norm": 0.1985267847776413,
+      "learning_rate": 2.1514126580284538e-05,
+      "loss": 0.0054,
+      "step": 21250
+    },
+    {
+      "epoch": 1.2384225549018466,
+      "grad_norm": 0.22486825287342072,
+      "learning_rate": 2.1468847607247344e-05,
+      "loss": 0.0064,
+      "step": 21260
+    },
+    {
+      "epoch": 1.23900506786276,
+      "grad_norm": 0.22276251018047333,
+      "learning_rate": 2.1423603302250623e-05,
+      "loss": 0.0058,
+      "step": 21270
+    },
+    {
+      "epoch": 1.2395875808236734,
+      "grad_norm": 0.165960431098938,
+      "learning_rate": 2.1378393720270468e-05,
+      "loss": 0.0057,
+      "step": 21280
+    },
+    {
+      "epoch": 1.2401700937845868,
+      "grad_norm": 0.2775128185749054,
+      "learning_rate": 2.133321891624076e-05,
+      "loss": 0.0037,
+      "step": 21290
+    },
+    {
+      "epoch": 1.2407526067455001,
+      "grad_norm": 0.23746441304683685,
+      "learning_rate": 2.1288078945053193e-05,
+      "loss": 0.0066,
+      "step": 21300
+    },
+    {
+      "epoch": 1.2413351197064135,
+      "grad_norm": 0.08691519498825073,
+      "learning_rate": 2.1242973861557064e-05,
+      "loss": 0.0028,
+      "step": 21310
+    },
+    {
+      "epoch": 1.241917632667327,
+      "grad_norm": 0.1924833208322525,
+      "learning_rate": 2.1197903720559304e-05,
+      "loss": 0.0032,
+      "step": 21320
+    },
+    {
+      "epoch": 1.2425001456282403,
+      "grad_norm": 0.13004125654697418,
+      "learning_rate": 2.115286857682438e-05,
+      "loss": 0.0041,
+      "step": 21330
+    },
+    {
+      "epoch": 1.2430826585891537,
+      "grad_norm": 0.14399629831314087,
+      "learning_rate": 2.1107868485074228e-05,
+      "loss": 0.0053,
+      "step": 21340
+    },
+    {
+      "epoch": 1.243665171550067,
+      "grad_norm": 0.1485465168952942,
+      "learning_rate": 2.1062903499988234e-05,
+      "loss": 0.0043,
+      "step": 21350
+    },
+    {
+      "epoch": 1.2442476845109804,
+      "grad_norm": 0.19372139871120453,
+      "learning_rate": 2.101797367620308e-05,
+      "loss": 0.0069,
+      "step": 21360
+    },
+    {
+      "epoch": 1.2448301974718938,
+      "grad_norm": 0.2578558325767517,
+      "learning_rate": 2.0973079068312713e-05,
+      "loss": 0.0054,
+      "step": 21370
+    },
+    {
+      "epoch": 1.2454127104328072,
+      "grad_norm": 0.2870299518108368,
+      "learning_rate": 2.0928219730868358e-05,
+      "loss": 0.0054,
+      "step": 21380
+    },
+    {
+      "epoch": 1.2459952233937206,
+      "grad_norm": 0.18724001944065094,
+      "learning_rate": 2.0883395718378302e-05,
+      "loss": 0.0041,
+      "step": 21390
+    },
+    {
+      "epoch": 1.246577736354634,
+      "grad_norm": 0.2278311401605606,
+      "learning_rate": 2.083860708530798e-05,
+      "loss": 0.0052,
+      "step": 21400
+    },
+    {
+      "epoch": 1.2471602493155474,
+      "grad_norm": 0.17008936405181885,
+      "learning_rate": 2.0793853886079796e-05,
+      "loss": 0.0039,
+      "step": 21410
+    },
+    {
+      "epoch": 1.2477427622764607,
+      "grad_norm": 0.13392511010169983,
+      "learning_rate": 2.074913617507309e-05,
+      "loss": 0.0044,
+      "step": 21420
+    },
+    {
+      "epoch": 1.2483252752373741,
+      "grad_norm": 0.2717956304550171,
+      "learning_rate": 2.070445400662411e-05,
+      "loss": 0.0054,
+      "step": 21430
+    },
+    {
+      "epoch": 1.2489077881982875,
+      "grad_norm": 0.11660441011190414,
+      "learning_rate": 2.0659807435025907e-05,
+      "loss": 0.0055,
+      "step": 21440
+    },
+    {
+      "epoch": 1.249490301159201,
+      "grad_norm": 0.14914101362228394,
+      "learning_rate": 2.061519651452825e-05,
+      "loss": 0.0075,
+      "step": 21450
+    },
+    {
+      "epoch": 1.2500728141201143,
+      "grad_norm": 0.193598672747612,
+      "learning_rate": 2.0570621299337656e-05,
+      "loss": 0.0037,
+      "step": 21460
+    },
+    {
+      "epoch": 1.2506553270810277,
+      "grad_norm": 0.17431007325649261,
+      "learning_rate": 2.0526081843617183e-05,
+      "loss": 0.0047,
+      "step": 21470
+    },
+    {
+      "epoch": 1.251237840041941,
+      "grad_norm": 0.1824505776166916,
+      "learning_rate": 2.0481578201486485e-05,
+      "loss": 0.0054,
+      "step": 21480
+    },
+    {
+      "epoch": 1.2518203530028544,
+      "grad_norm": 0.20421530306339264,
+      "learning_rate": 2.0437110427021677e-05,
+      "loss": 0.0044,
+      "step": 21490
+    },
+    {
+      "epoch": 1.2524028659637678,
+      "grad_norm": 0.18757197260856628,
+      "learning_rate": 2.0392678574255282e-05,
+      "loss": 0.0056,
+      "step": 21500
+    },
+    {
+      "epoch": 1.2529853789246812,
+      "grad_norm": 0.28430312871932983,
+      "learning_rate": 2.0348282697176217e-05,
+      "loss": 0.0043,
+      "step": 21510
+    },
+    {
+      "epoch": 1.2535678918855946,
+      "grad_norm": 0.13424529135227203,
+      "learning_rate": 2.030392284972964e-05,
+      "loss": 0.0037,
+      "step": 21520
+    },
+    {
+      "epoch": 1.254150404846508,
+      "grad_norm": 0.16240264475345612,
+      "learning_rate": 2.0259599085816973e-05,
+      "loss": 0.0031,
+      "step": 21530
+    },
+    {
+      "epoch": 1.2547329178074211,
+      "grad_norm": 0.29662689566612244,
+      "learning_rate": 2.0215311459295755e-05,
+      "loss": 0.0049,
+      "step": 21540
+    },
+    {
+      "epoch": 1.2553154307683345,
+      "grad_norm": 0.14077579975128174,
+      "learning_rate": 2.0171060023979606e-05,
+      "loss": 0.0031,
+      "step": 21550
+    },
+    {
+      "epoch": 1.255897943729248,
+      "grad_norm": 0.15870895981788635,
+      "learning_rate": 2.012684483363823e-05,
+      "loss": 0.0039,
+      "step": 21560
+    },
+    {
+      "epoch": 1.2564804566901613,
+      "grad_norm": 0.13118012249469757,
+      "learning_rate": 2.0082665941997237e-05,
+      "loss": 0.0046,
+      "step": 21570
+    },
+    {
+      "epoch": 1.2570629696510747,
+      "grad_norm": 0.3143213093280792,
+      "learning_rate": 2.0038523402738146e-05,
+      "loss": 0.0051,
+      "step": 21580
+    },
+    {
+      "epoch": 1.257645482611988,
+      "grad_norm": 0.1856813281774521,
+      "learning_rate": 1.99944172694983e-05,
+      "loss": 0.0057,
+      "step": 21590
+    },
+    {
+      "epoch": 1.2582279955729014,
+      "grad_norm": 0.1402192860841751,
+      "learning_rate": 1.99503475958708e-05,
+      "loss": 0.0037,
+      "step": 21600
+    },
+    {
+      "epoch": 1.2588105085338148,
+      "grad_norm": 0.18368394672870636,
+      "learning_rate": 1.9906314435404484e-05,
+      "loss": 0.0048,
+      "step": 21610
+    },
+    {
+      "epoch": 1.2593930214947282,
+      "grad_norm": 0.17350837588310242,
+      "learning_rate": 1.986231784160378e-05,
+      "loss": 0.0044,
+      "step": 21620
+    },
+    {
+      "epoch": 1.2599755344556416,
+      "grad_norm": 0.17285306751728058,
+      "learning_rate": 1.9818357867928693e-05,
+      "loss": 0.0056,
+      "step": 21630
+    },
+    {
+      "epoch": 1.260558047416555,
+      "grad_norm": 0.2504863739013672,
+      "learning_rate": 1.977443456779474e-05,
+      "loss": 0.0035,
+      "step": 21640
+    },
+    {
+      "epoch": 1.2611405603774684,
+      "grad_norm": 0.2436114102602005,
+      "learning_rate": 1.9730547994572858e-05,
+      "loss": 0.0048,
+      "step": 21650
+    },
+    {
+      "epoch": 1.2617230733383817,
+      "grad_norm": 0.25466442108154297,
+      "learning_rate": 1.9686698201589393e-05,
+      "loss": 0.0037,
+      "step": 21660
+    },
+    {
+      "epoch": 1.2623055862992951,
+      "grad_norm": 0.14095096290111542,
+      "learning_rate": 1.9642885242125964e-05,
+      "loss": 0.0034,
+      "step": 21670
+    },
+    {
+      "epoch": 1.2628880992602085,
+      "grad_norm": 0.2459566742181778,
+      "learning_rate": 1.9599109169419466e-05,
+      "loss": 0.0043,
+      "step": 21680
+    },
+    {
+      "epoch": 1.263470612221122,
+      "grad_norm": 0.09001509845256805,
+      "learning_rate": 1.9555370036661948e-05,
+      "loss": 0.0036,
+      "step": 21690
+    },
+    {
+      "epoch": 1.2640531251820353,
+      "grad_norm": 0.43285050988197327,
+      "learning_rate": 1.9511667897000574e-05,
+      "loss": 0.0056,
+      "step": 21700
+    },
+    {
+      "epoch": 1.2646356381429487,
+      "grad_norm": 0.23009738326072693,
+      "learning_rate": 1.946800280353755e-05,
+      "loss": 0.0046,
+      "step": 21710
+    },
+    {
+      "epoch": 1.265218151103862,
+      "grad_norm": 0.47730961441993713,
+      "learning_rate": 1.9424374809330114e-05,
+      "loss": 0.0046,
+      "step": 21720
+    },
+    {
+      "epoch": 1.2658006640647754,
+      "grad_norm": 0.20132310688495636,
+      "learning_rate": 1.9380783967390382e-05,
+      "loss": 0.0049,
+      "step": 21730
+    },
+    {
+      "epoch": 1.2663831770256888,
+      "grad_norm": 0.18645089864730835,
+      "learning_rate": 1.933723033068533e-05,
+      "loss": 0.0049,
+      "step": 21740
+    },
+    {
+      "epoch": 1.2669656899866022,
+      "grad_norm": 0.1737818717956543,
+      "learning_rate": 1.929371395213674e-05,
+      "loss": 0.0047,
+      "step": 21750
+    },
+    {
+      "epoch": 1.2675482029475156,
+      "grad_norm": 0.15621988475322723,
+      "learning_rate": 1.9250234884621092e-05,
+      "loss": 0.0045,
+      "step": 21760
+    },
+    {
+      "epoch": 1.268130715908429,
+      "grad_norm": 0.1186591163277626,
+      "learning_rate": 1.920679318096959e-05,
+      "loss": 0.0042,
+      "step": 21770
+    },
+    {
+      "epoch": 1.2687132288693423,
+      "grad_norm": 0.29013004899024963,
+      "learning_rate": 1.9163388893967982e-05,
+      "loss": 0.0043,
+      "step": 21780
+    },
+    {
+      "epoch": 1.2692957418302557,
+      "grad_norm": 0.2397008240222931,
+      "learning_rate": 1.9120022076356575e-05,
+      "loss": 0.0041,
+      "step": 21790
+    },
+    {
+      "epoch": 1.2698782547911691,
+      "grad_norm": 0.1569042056798935,
+      "learning_rate": 1.9076692780830114e-05,
+      "loss": 0.0042,
+      "step": 21800
+    },
+    {
+      "epoch": 1.2704607677520825,
+      "grad_norm": 0.20687006413936615,
+      "learning_rate": 1.903340106003782e-05,
+      "loss": 0.0059,
+      "step": 21810
+    },
+    {
+      "epoch": 1.2710432807129959,
+      "grad_norm": 0.1769949495792389,
+      "learning_rate": 1.8990146966583183e-05,
+      "loss": 0.004,
+      "step": 21820
+    },
+    {
+      "epoch": 1.2716257936739093,
+      "grad_norm": 0.16452525556087494,
+      "learning_rate": 1.8946930553024035e-05,
+      "loss": 0.005,
+      "step": 21830
+    },
+    {
+      "epoch": 1.2722083066348226,
+      "grad_norm": 0.16696324944496155,
+      "learning_rate": 1.8903751871872378e-05,
+      "loss": 0.0039,
+      "step": 21840
+    },
+    {
+      "epoch": 1.272790819595736,
+      "grad_norm": 0.2538956105709076,
+      "learning_rate": 1.8860610975594382e-05,
+      "loss": 0.0059,
+      "step": 21850
+    },
+    {
+      "epoch": 1.2733733325566494,
+      "grad_norm": 0.17405438423156738,
+      "learning_rate": 1.8817507916610307e-05,
+      "loss": 0.0046,
+      "step": 21860
+    },
+    {
+      "epoch": 1.2739558455175628,
+      "grad_norm": 0.22045010328292847,
+      "learning_rate": 1.8774442747294407e-05,
+      "loss": 0.0039,
+      "step": 21870
+    },
+    {
+      "epoch": 1.2745383584784762,
+      "grad_norm": 0.29201018810272217,
+      "learning_rate": 1.8731415519974965e-05,
+      "loss": 0.0043,
+      "step": 21880
+    },
+    {
+      "epoch": 1.2751208714393896,
+      "grad_norm": 0.17026935517787933,
+      "learning_rate": 1.86884262869341e-05,
+      "loss": 0.0037,
+      "step": 21890
+    },
+    {
+      "epoch": 1.275703384400303,
+      "grad_norm": 0.17703470587730408,
+      "learning_rate": 1.8645475100407788e-05,
+      "loss": 0.0032,
+      "step": 21900
+    },
+    {
+      "epoch": 1.2762858973612163,
+      "grad_norm": 0.33621203899383545,
+      "learning_rate": 1.8602562012585768e-05,
+      "loss": 0.0047,
+      "step": 21910
+    },
+    {
+      "epoch": 1.2768684103221297,
+      "grad_norm": 0.18372392654418945,
+      "learning_rate": 1.8559687075611466e-05,
+      "loss": 0.0029,
+      "step": 21920
+    },
+    {
+      "epoch": 1.277450923283043,
+      "grad_norm": 0.13207131624221802,
+      "learning_rate": 1.8516850341582017e-05,
+      "loss": 0.0043,
+      "step": 21930
+    },
+    {
+      "epoch": 1.2780334362439565,
+      "grad_norm": 0.125705748796463,
+      "learning_rate": 1.847405186254807e-05,
+      "loss": 0.0071,
+      "step": 21940
+    },
+    {
+      "epoch": 1.2786159492048699,
+      "grad_norm": 0.3394865393638611,
+      "learning_rate": 1.843129169051379e-05,
+      "loss": 0.0047,
+      "step": 21950
+    },
+    {
+      "epoch": 1.2791984621657833,
+      "grad_norm": 0.10356965661048889,
+      "learning_rate": 1.8388569877436863e-05,
+      "loss": 0.0035,
+      "step": 21960
+    },
+    {
+      "epoch": 1.2797809751266966,
+      "grad_norm": 0.09709061682224274,
+      "learning_rate": 1.834588647522828e-05,
+      "loss": 0.0032,
+      "step": 21970
+    },
+    {
+      "epoch": 1.28036348808761,
+      "grad_norm": 0.24591124057769775,
+      "learning_rate": 1.8303241535752435e-05,
+      "loss": 0.0074,
+      "step": 21980
+    },
+    {
+      "epoch": 1.2809460010485234,
+      "grad_norm": 0.17079505324363708,
+      "learning_rate": 1.8260635110826934e-05,
+      "loss": 0.006,
+      "step": 21990
+    },
+    {
+      "epoch": 1.2815285140094368,
+      "grad_norm": 0.19725094735622406,
+      "learning_rate": 1.8218067252222597e-05,
+      "loss": 0.0042,
+      "step": 22000
+    },
+    {
+      "epoch": 1.28211102697035,
+      "grad_norm": 0.22381342947483063,
+      "learning_rate": 1.817553801166339e-05,
+      "loss": 0.0043,
+      "step": 22010
+    },
+    {
+      "epoch": 1.2826935399312633,
+      "grad_norm": 0.18287421762943268,
+      "learning_rate": 1.8133047440826333e-05,
+      "loss": 0.0031,
+      "step": 22020
+    },
+    {
+      "epoch": 1.2832760528921767,
+      "grad_norm": 0.2575538456439972,
+      "learning_rate": 1.809059559134151e-05,
+      "loss": 0.0045,
+      "step": 22030
+    },
+    {
+      "epoch": 1.28385856585309,
+      "grad_norm": 0.11071810871362686,
+      "learning_rate": 1.80481825147919e-05,
+      "loss": 0.0039,
+      "step": 22040
+    },
+    {
+      "epoch": 1.2844410788140035,
+      "grad_norm": 0.2752107083797455,
+      "learning_rate": 1.8005808262713396e-05,
+      "loss": 0.0036,
+      "step": 22050
+    },
+    {
+      "epoch": 1.2850235917749169,
+      "grad_norm": 0.12737435102462769,
+      "learning_rate": 1.7963472886594714e-05,
+      "loss": 0.0037,
+      "step": 22060
+    },
+    {
+      "epoch": 1.2856061047358303,
+      "grad_norm": 0.10110799968242645,
+      "learning_rate": 1.7921176437877302e-05,
+      "loss": 0.0029,
+      "step": 22070
+    },
+    {
+      "epoch": 1.2861886176967436,
+      "grad_norm": 0.16161105036735535,
+      "learning_rate": 1.7878918967955366e-05,
+      "loss": 0.0039,
+      "step": 22080
+    },
+    {
+      "epoch": 1.286771130657657,
+      "grad_norm": 0.296579509973526,
+      "learning_rate": 1.7836700528175693e-05,
+      "loss": 0.0045,
+      "step": 22090
+    },
+    {
+      "epoch": 1.2873536436185704,
+      "grad_norm": 0.1514829844236374,
+      "learning_rate": 1.779452116983769e-05,
+      "loss": 0.0033,
+      "step": 22100
+    },
+    {
+      "epoch": 1.2879361565794838,
+      "grad_norm": 0.188397616147995,
+      "learning_rate": 1.775238094419325e-05,
+      "loss": 0.0064,
+      "step": 22110
+    },
+    {
+      "epoch": 1.2885186695403972,
+      "grad_norm": 0.19670860469341278,
+      "learning_rate": 1.7710279902446708e-05,
+      "loss": 0.0055,
+      "step": 22120
+    },
+    {
+      "epoch": 1.2891011825013106,
+      "grad_norm": 0.15476471185684204,
+      "learning_rate": 1.7668218095754795e-05,
+      "loss": 0.0032,
+      "step": 22130
+    },
+    {
+      "epoch": 1.289683695462224,
+      "grad_norm": 0.2178555727005005,
+      "learning_rate": 1.7626195575226594e-05,
+      "loss": 0.0036,
+      "step": 22140
+    },
+    {
+      "epoch": 1.2902662084231373,
+      "grad_norm": 0.2915867865085602,
+      "learning_rate": 1.758421239192343e-05,
+      "loss": 0.0043,
+      "step": 22150
+    },
+    {
+      "epoch": 1.2908487213840507,
+      "grad_norm": 0.17939074337482452,
+      "learning_rate": 1.7542268596858814e-05,
+      "loss": 0.0045,
+      "step": 22160
+    },
+    {
+      "epoch": 1.291431234344964,
+      "grad_norm": 0.25635042786598206,
+      "learning_rate": 1.750036424099841e-05,
+      "loss": 0.003,
+      "step": 22170
+    },
+    {
+      "epoch": 1.2920137473058775,
+      "grad_norm": 0.3532696068286896,
+      "learning_rate": 1.7458499375259956e-05,
+      "loss": 0.004,
+      "step": 22180
+    },
+    {
+      "epoch": 1.2925962602667909,
+      "grad_norm": 0.1677241027355194,
+      "learning_rate": 1.7416674050513245e-05,
+      "loss": 0.0046,
+      "step": 22190
+    },
+    {
+      "epoch": 1.2931787732277042,
+      "grad_norm": 0.20934894680976868,
+      "learning_rate": 1.737488831757997e-05,
+      "loss": 0.0037,
+      "step": 22200
+    },
+    {
+      "epoch": 1.2937612861886176,
+      "grad_norm": 0.23012606799602509,
+      "learning_rate": 1.733314222723373e-05,
+      "loss": 0.0033,
+      "step": 22210
+    },
+    {
+      "epoch": 1.294343799149531,
+      "grad_norm": 0.3446224331855774,
+      "learning_rate": 1.7291435830199954e-05,
+      "loss": 0.0053,
+      "step": 22220
+    },
+    {
+      "epoch": 1.2949263121104444,
+      "grad_norm": 0.23775896430015564,
+      "learning_rate": 1.7249769177155878e-05,
+      "loss": 0.0046,
+      "step": 22230
+    },
+    {
+      "epoch": 1.2955088250713578,
+      "grad_norm": 0.3229225277900696,
+      "learning_rate": 1.720814231873038e-05,
+      "loss": 0.0058,
+      "step": 22240
+    },
+    {
+      "epoch": 1.2960913380322712,
+      "grad_norm": 0.17829467356204987,
+      "learning_rate": 1.716655530550405e-05,
+      "loss": 0.0043,
+      "step": 22250
+    },
+    {
+      "epoch": 1.2966738509931846,
+      "grad_norm": 0.18781237304210663,
+      "learning_rate": 1.7125008188009017e-05,
+      "loss": 0.0074,
+      "step": 22260
+    },
+    {
+      "epoch": 1.297256363954098,
+      "grad_norm": 0.16207623481750488,
+      "learning_rate": 1.708350101672894e-05,
+      "loss": 0.004,
+      "step": 22270
+    },
+    {
+      "epoch": 1.2978388769150113,
+      "grad_norm": 0.12224821746349335,
+      "learning_rate": 1.7042033842098957e-05,
+      "loss": 0.0034,
+      "step": 22280
+    },
+    {
+      "epoch": 1.2984213898759247,
+      "grad_norm": 0.1529485434293747,
+      "learning_rate": 1.7000606714505567e-05,
+      "loss": 0.0049,
+      "step": 22290
+    },
+    {
+      "epoch": 1.299003902836838,
+      "grad_norm": 0.09538810700178146,
+      "learning_rate": 1.695921968428668e-05,
+      "loss": 0.0029,
+      "step": 22300
+    },
+    {
+      "epoch": 1.2995864157977515,
+      "grad_norm": 0.11703797429800034,
+      "learning_rate": 1.6917872801731416e-05,
+      "loss": 0.0035,
+      "step": 22310
+    },
+    {
+      "epoch": 1.3001689287586649,
+      "grad_norm": 0.19497032463550568,
+      "learning_rate": 1.687656611708014e-05,
+      "loss": 0.0043,
+      "step": 22320
+    },
+    {
+      "epoch": 1.3007514417195782,
+      "grad_norm": 0.15076172351837158,
+      "learning_rate": 1.6835299680524368e-05,
+      "loss": 0.0053,
+      "step": 22330
+    },
+    {
+      "epoch": 1.3013339546804916,
+      "grad_norm": 0.304424524307251,
+      "learning_rate": 1.679407354220669e-05,
+      "loss": 0.0042,
+      "step": 22340
+    },
+    {
+      "epoch": 1.301916467641405,
+      "grad_norm": 0.12780015170574188,
+      "learning_rate": 1.6752887752220793e-05,
+      "loss": 0.004,
+      "step": 22350
+    },
+    {
+      "epoch": 1.3024989806023184,
+      "grad_norm": 0.2522312104701996,
+      "learning_rate": 1.6711742360611277e-05,
+      "loss": 0.0031,
+      "step": 22360
+    },
+    {
+      "epoch": 1.3030814935632318,
+      "grad_norm": 0.20535795390605927,
+      "learning_rate": 1.6670637417373652e-05,
+      "loss": 0.0042,
+      "step": 22370
+    },
+    {
+      "epoch": 1.3036640065241452,
+      "grad_norm": 0.10855348408222198,
+      "learning_rate": 1.6629572972454333e-05,
+      "loss": 0.0063,
+      "step": 22380
+    },
+    {
+      "epoch": 1.3042465194850585,
+      "grad_norm": 0.16992110013961792,
+      "learning_rate": 1.6588549075750464e-05,
+      "loss": 0.0038,
+      "step": 22390
+    },
+    {
+      "epoch": 1.304829032445972,
+      "grad_norm": 0.22594119608402252,
+      "learning_rate": 1.6547565777109976e-05,
+      "loss": 0.0065,
+      "step": 22400
+    },
+    {
+      "epoch": 1.3054115454068853,
+      "grad_norm": 0.2591063380241394,
+      "learning_rate": 1.650662312633143e-05,
+      "loss": 0.007,
+      "step": 22410
+    },
+    {
+      "epoch": 1.3059940583677987,
+      "grad_norm": 0.15203727781772614,
+      "learning_rate": 1.6465721173164002e-05,
+      "loss": 0.0041,
+      "step": 22420
+    },
+    {
+      "epoch": 1.306576571328712,
+      "grad_norm": 0.14258986711502075,
+      "learning_rate": 1.6424859967307428e-05,
+      "loss": 0.0037,
+      "step": 22430
+    },
+    {
+      "epoch": 1.3071590842896255,
+      "grad_norm": 0.1147351786494255,
+      "learning_rate": 1.6384039558411903e-05,
+      "loss": 0.0034,
+      "step": 22440
+    },
+    {
+      "epoch": 1.3077415972505388,
+      "grad_norm": 0.29580917954444885,
+      "learning_rate": 1.634325999607811e-05,
+      "loss": 0.0099,
+      "step": 22450
+    },
+    {
+      "epoch": 1.3083241102114522,
+      "grad_norm": 0.33049604296684265,
+      "learning_rate": 1.6302521329857045e-05,
+      "loss": 0.0047,
+      "step": 22460
+    },
+    {
+      "epoch": 1.3089066231723656,
+      "grad_norm": 0.20380355417728424,
+      "learning_rate": 1.6261823609250027e-05,
+      "loss": 0.0045,
+      "step": 22470
+    },
+    {
+      "epoch": 1.309489136133279,
+      "grad_norm": 0.23785725235939026,
+      "learning_rate": 1.6221166883708627e-05,
+      "loss": 0.0039,
+      "step": 22480
+    },
+    {
+      "epoch": 1.3100716490941924,
+      "grad_norm": 0.156007781624794,
+      "learning_rate": 1.6180551202634603e-05,
+      "loss": 0.0061,
+      "step": 22490
+    },
+    {
+      "epoch": 1.3106541620551058,
+      "grad_norm": 0.28421878814697266,
+      "learning_rate": 1.613997661537981e-05,
+      "loss": 0.0035,
+      "step": 22500
+    },
+    {
+      "epoch": 1.3112366750160191,
+      "grad_norm": 0.18053607642650604,
+      "learning_rate": 1.609944317124624e-05,
+      "loss": 0.003,
+      "step": 22510
+    },
+    {
+      "epoch": 1.3118191879769325,
+      "grad_norm": 0.15965528786182404,
+      "learning_rate": 1.605895091948582e-05,
+      "loss": 0.0045,
+      "step": 22520
+    },
+    {
+      "epoch": 1.312401700937846,
+      "grad_norm": 0.2234203815460205,
+      "learning_rate": 1.6018499909300477e-05,
+      "loss": 0.0046,
+      "step": 22530
+    },
+    {
+      "epoch": 1.3129842138987593,
+      "grad_norm": 0.1207997128367424,
+      "learning_rate": 1.597809018984199e-05,
+      "loss": 0.0028,
+      "step": 22540
+    },
+    {
+      "epoch": 1.3135667268596727,
+      "grad_norm": 0.1352788507938385,
+      "learning_rate": 1.593772181021196e-05,
+      "loss": 0.0036,
+      "step": 22550
+    },
+    {
+      "epoch": 1.314149239820586,
+      "grad_norm": 0.13815782964229584,
+      "learning_rate": 1.5897394819461813e-05,
+      "loss": 0.0031,
+      "step": 22560
+    },
+    {
+      "epoch": 1.3147317527814995,
+      "grad_norm": 0.3121473491191864,
+      "learning_rate": 1.5857109266592625e-05,
+      "loss": 0.0045,
+      "step": 22570
+    },
+    {
+      "epoch": 1.3153142657424128,
+      "grad_norm": 0.12790603935718536,
+      "learning_rate": 1.581686520055514e-05,
+      "loss": 0.0032,
+      "step": 22580
+    },
+    {
+      "epoch": 1.3158967787033262,
+      "grad_norm": 0.11288014054298401,
+      "learning_rate": 1.5776662670249704e-05,
+      "loss": 0.0032,
+      "step": 22590
+    },
+    {
+      "epoch": 1.3164792916642396,
+      "grad_norm": 0.16248446702957153,
+      "learning_rate": 1.573650172452615e-05,
+      "loss": 0.0043,
+      "step": 22600
+    },
+    {
+      "epoch": 1.317061804625153,
+      "grad_norm": 0.2535873055458069,
+      "learning_rate": 1.5696382412183852e-05,
+      "loss": 0.0033,
+      "step": 22610
+    },
+    {
+      "epoch": 1.3176443175860664,
+      "grad_norm": 0.2139279693365097,
+      "learning_rate": 1.565630478197155e-05,
+      "loss": 0.0055,
+      "step": 22620
+    },
+    {
+      "epoch": 1.3182268305469798,
+      "grad_norm": 0.10027030110359192,
+      "learning_rate": 1.5616268882587332e-05,
+      "loss": 0.0025,
+      "step": 22630
+    },
+    {
+      "epoch": 1.3188093435078931,
+      "grad_norm": 0.08188876509666443,
+      "learning_rate": 1.5576274762678594e-05,
+      "loss": 0.0037,
+      "step": 22640
+    },
+    {
+      "epoch": 1.3193918564688065,
+      "grad_norm": 0.08680975437164307,
+      "learning_rate": 1.5536322470841952e-05,
+      "loss": 0.0049,
+      "step": 22650
+    },
+    {
+      "epoch": 1.31997436942972,
+      "grad_norm": 0.21435260772705078,
+      "learning_rate": 1.549641205562324e-05,
+      "loss": 0.0039,
+      "step": 22660
+    },
+    {
+      "epoch": 1.3205568823906333,
+      "grad_norm": 0.24836532771587372,
+      "learning_rate": 1.5456543565517334e-05,
+      "loss": 0.0054,
+      "step": 22670
+    },
+    {
+      "epoch": 1.3211393953515467,
+      "grad_norm": 0.10833306610584259,
+      "learning_rate": 1.541671704896825e-05,
+      "loss": 0.0063,
+      "step": 22680
+    },
+    {
+      "epoch": 1.32172190831246,
+      "grad_norm": 0.3324327766895294,
+      "learning_rate": 1.5376932554368938e-05,
+      "loss": 0.0039,
+      "step": 22690
+    },
+    {
+      "epoch": 1.3223044212733734,
+      "grad_norm": 0.24245776236057281,
+      "learning_rate": 1.5337190130061318e-05,
+      "loss": 0.0051,
+      "step": 22700
+    },
+    {
+      "epoch": 1.3228869342342868,
+      "grad_norm": 0.2800937592983246,
+      "learning_rate": 1.5297489824336165e-05,
+      "loss": 0.0061,
+      "step": 22710
+    },
+    {
+      "epoch": 1.3234694471952002,
+      "grad_norm": 0.12686367332935333,
+      "learning_rate": 1.5257831685433121e-05,
+      "loss": 0.0045,
+      "step": 22720
+    },
+    {
+      "epoch": 1.3240519601561136,
+      "grad_norm": 0.2106655240058899,
+      "learning_rate": 1.521821576154055e-05,
+      "loss": 0.0051,
+      "step": 22730
+    },
+    {
+      "epoch": 1.324634473117027,
+      "grad_norm": 0.21520201861858368,
+      "learning_rate": 1.5178642100795542e-05,
+      "loss": 0.005,
+      "step": 22740
+    },
+    {
+      "epoch": 1.3252169860779404,
+      "grad_norm": 0.2689678370952606,
+      "learning_rate": 1.5139110751283819e-05,
+      "loss": 0.0039,
+      "step": 22750
+    },
+    {
+      "epoch": 1.3257994990388537,
+      "grad_norm": 0.21697765588760376,
+      "learning_rate": 1.5099621761039683e-05,
+      "loss": 0.0031,
+      "step": 22760
+    },
+    {
+      "epoch": 1.3263820119997671,
+      "grad_norm": 0.12680985033512115,
+      "learning_rate": 1.5060175178046016e-05,
+      "loss": 0.0052,
+      "step": 22770
+    },
+    {
+      "epoch": 1.3269645249606805,
+      "grad_norm": 0.1977674812078476,
+      "learning_rate": 1.5020771050234117e-05,
+      "loss": 0.0049,
+      "step": 22780
+    },
+    {
+      "epoch": 1.3275470379215937,
+      "grad_norm": 0.180106520652771,
+      "learning_rate": 1.4981409425483716e-05,
+      "loss": 0.0049,
+      "step": 22790
+    },
+    {
+      "epoch": 1.328129550882507,
+      "grad_norm": 0.1909998655319214,
+      "learning_rate": 1.4942090351622883e-05,
+      "loss": 0.0043,
+      "step": 22800
+    },
+    {
+      "epoch": 1.3287120638434204,
+      "grad_norm": 0.15698198974132538,
+      "learning_rate": 1.490281387642804e-05,
+      "loss": 0.0045,
+      "step": 22810
+    },
+    {
+      "epoch": 1.3292945768043338,
+      "grad_norm": 0.13025203347206116,
+      "learning_rate": 1.4863580047623765e-05,
+      "loss": 0.0057,
+      "step": 22820
+    },
+    {
+      "epoch": 1.3298770897652472,
+      "grad_norm": 0.19189688563346863,
+      "learning_rate": 1.4824388912882897e-05,
+      "loss": 0.0034,
+      "step": 22830
+    },
+    {
+      "epoch": 1.3304596027261606,
+      "grad_norm": 0.19031046330928802,
+      "learning_rate": 1.4785240519826344e-05,
+      "loss": 0.0038,
+      "step": 22840
+    },
+    {
+      "epoch": 1.331042115687074,
+      "grad_norm": 0.15879002213478088,
+      "learning_rate": 1.4746134916023097e-05,
+      "loss": 0.0036,
+      "step": 22850
+    },
+    {
+      "epoch": 1.3316246286479874,
+      "grad_norm": 0.15148644149303436,
+      "learning_rate": 1.4707072148990141e-05,
+      "loss": 0.004,
+      "step": 22860
+    },
+    {
+      "epoch": 1.3322071416089007,
+      "grad_norm": 0.11583960801362991,
+      "learning_rate": 1.4668052266192423e-05,
+      "loss": 0.0025,
+      "step": 22870
+    },
+    {
+      "epoch": 1.3327896545698141,
+      "grad_norm": 0.15552952885627747,
+      "learning_rate": 1.4629075315042795e-05,
+      "loss": 0.0052,
+      "step": 22880
+    },
+    {
+      "epoch": 1.3333721675307275,
+      "grad_norm": 0.3102525472640991,
+      "learning_rate": 1.4590141342901925e-05,
+      "loss": 0.0044,
+      "step": 22890
+    },
+    {
+      "epoch": 1.333954680491641,
+      "grad_norm": 0.2149413526058197,
+      "learning_rate": 1.4551250397078253e-05,
+      "loss": 0.0033,
+      "step": 22900
+    },
+    {
+      "epoch": 1.3345371934525543,
+      "grad_norm": 0.45925623178482056,
+      "learning_rate": 1.4512402524827945e-05,
+      "loss": 0.0061,
+      "step": 22910
+    },
+    {
+      "epoch": 1.3351197064134677,
+      "grad_norm": 0.26168176531791687,
+      "learning_rate": 1.447359777335482e-05,
+      "loss": 0.0044,
+      "step": 22920
+    },
+    {
+      "epoch": 1.335702219374381,
+      "grad_norm": 0.11999425292015076,
+      "learning_rate": 1.4434836189810335e-05,
+      "loss": 0.004,
+      "step": 22930
+    },
+    {
+      "epoch": 1.3362847323352944,
+      "grad_norm": 0.16261479258537292,
+      "learning_rate": 1.4396117821293453e-05,
+      "loss": 0.0034,
+      "step": 22940
+    },
+    {
+      "epoch": 1.3368672452962078,
+      "grad_norm": 0.09627418965101242,
+      "learning_rate": 1.4357442714850633e-05,
+      "loss": 0.0031,
+      "step": 22950
+    },
+    {
+      "epoch": 1.3374497582571212,
+      "grad_norm": 0.2050800621509552,
+      "learning_rate": 1.43188109174758e-05,
+      "loss": 0.0037,
+      "step": 22960
+    },
+    {
+      "epoch": 1.3380322712180346,
+      "grad_norm": 0.14770030975341797,
+      "learning_rate": 1.4280222476110206e-05,
+      "loss": 0.0031,
+      "step": 22970
+    },
+    {
+      "epoch": 1.338614784178948,
+      "grad_norm": 0.34470927715301514,
+      "learning_rate": 1.4241677437642469e-05,
+      "loss": 0.0047,
+      "step": 22980
+    },
+    {
+      "epoch": 1.3391972971398614,
+      "grad_norm": 0.051038797944784164,
+      "learning_rate": 1.420317584890844e-05,
+      "loss": 0.0033,
+      "step": 22990
+    },
+    {
+      "epoch": 1.3397798101007747,
+      "grad_norm": 0.112837053835392,
+      "learning_rate": 1.4164717756691176e-05,
+      "loss": 0.0033,
+      "step": 23000
+    },
+    {
+      "epoch": 1.3403623230616881,
+      "grad_norm": 0.12880632281303406,
+      "learning_rate": 1.4126303207720882e-05,
+      "loss": 0.0052,
+      "step": 23010
+    },
+    {
+      "epoch": 1.3409448360226015,
+      "grad_norm": 0.15257976949214935,
+      "learning_rate": 1.408793224867484e-05,
+      "loss": 0.0037,
+      "step": 23020
+    },
+    {
+      "epoch": 1.341527348983515,
+      "grad_norm": 0.20780257880687714,
+      "learning_rate": 1.4049604926177423e-05,
+      "loss": 0.0044,
+      "step": 23030
+    },
+    {
+      "epoch": 1.3421098619444283,
+      "grad_norm": 0.28359290957450867,
+      "learning_rate": 1.4011321286799916e-05,
+      "loss": 0.0039,
+      "step": 23040
+    },
+    {
+      "epoch": 1.3426923749053417,
+      "grad_norm": 0.18815742433071136,
+      "learning_rate": 1.3973081377060565e-05,
+      "loss": 0.0067,
+      "step": 23050
+    },
+    {
+      "epoch": 1.343274887866255,
+      "grad_norm": 0.1938387155532837,
+      "learning_rate": 1.393488524342445e-05,
+      "loss": 0.0053,
+      "step": 23060
+    },
+    {
+      "epoch": 1.3438574008271684,
+      "grad_norm": 0.1650422215461731,
+      "learning_rate": 1.3896732932303486e-05,
+      "loss": 0.0056,
+      "step": 23070
+    },
+    {
+      "epoch": 1.3444399137880818,
+      "grad_norm": 0.10856599360704422,
+      "learning_rate": 1.3858624490056304e-05,
+      "loss": 0.0067,
+      "step": 23080
+    },
+    {
+      "epoch": 1.3450224267489952,
+      "grad_norm": 0.1495738923549652,
+      "learning_rate": 1.3820559962988299e-05,
+      "loss": 0.0042,
+      "step": 23090
+    },
+    {
+      "epoch": 1.3456049397099086,
+      "grad_norm": 0.21570543944835663,
+      "learning_rate": 1.3782539397351418e-05,
+      "loss": 0.0056,
+      "step": 23100
+    },
+    {
+      "epoch": 1.346187452670822,
+      "grad_norm": 0.3087175190448761,
+      "learning_rate": 1.3744562839344266e-05,
+      "loss": 0.0041,
+      "step": 23110
+    },
+    {
+      "epoch": 1.3467699656317353,
+      "grad_norm": 0.12588591873645782,
+      "learning_rate": 1.370663033511193e-05,
+      "loss": 0.0041,
+      "step": 23120
+    },
+    {
+      "epoch": 1.3473524785926487,
+      "grad_norm": 0.1964387446641922,
+      "learning_rate": 1.3668741930745964e-05,
+      "loss": 0.004,
+      "step": 23130
+    },
+    {
+      "epoch": 1.3479349915535621,
+      "grad_norm": 0.4997328817844391,
+      "learning_rate": 1.3630897672284382e-05,
+      "loss": 0.0052,
+      "step": 23140
+    },
+    {
+      "epoch": 1.3485175045144755,
+      "grad_norm": 0.18089736998081207,
+      "learning_rate": 1.3593097605711508e-05,
+      "loss": 0.0064,
+      "step": 23150
+    },
+    {
+      "epoch": 1.3491000174753889,
+      "grad_norm": 0.11540351063013077,
+      "learning_rate": 1.3555341776957992e-05,
+      "loss": 0.005,
+      "step": 23160
+    },
+    {
+      "epoch": 1.3496825304363023,
+      "grad_norm": 0.17794744670391083,
+      "learning_rate": 1.3517630231900724e-05,
+      "loss": 0.0031,
+      "step": 23170
+    },
+    {
+      "epoch": 1.3502650433972156,
+      "grad_norm": 0.08660774677991867,
+      "learning_rate": 1.3479963016362767e-05,
+      "loss": 0.006,
+      "step": 23180
+    },
+    {
+      "epoch": 1.350847556358129,
+      "grad_norm": 0.10138967633247375,
+      "learning_rate": 1.3442340176113377e-05,
+      "loss": 0.0039,
+      "step": 23190
+    },
+    {
+      "epoch": 1.3514300693190424,
+      "grad_norm": 0.17787504196166992,
+      "learning_rate": 1.340476175686784e-05,
+      "loss": 0.0042,
+      "step": 23200
+    },
+    {
+      "epoch": 1.3520125822799558,
+      "grad_norm": 0.4286711812019348,
+      "learning_rate": 1.3367227804287469e-05,
+      "loss": 0.0068,
+      "step": 23210
+    },
+    {
+      "epoch": 1.3525950952408692,
+      "grad_norm": 0.25379323959350586,
+      "learning_rate": 1.332973836397956e-05,
+      "loss": 0.0034,
+      "step": 23220
+    },
+    {
+      "epoch": 1.3531776082017826,
+      "grad_norm": 0.42579564452171326,
+      "learning_rate": 1.329229348149731e-05,
+      "loss": 0.0073,
+      "step": 23230
+    },
+    {
+      "epoch": 1.353760121162696,
+      "grad_norm": 0.2155422866344452,
+      "learning_rate": 1.3254893202339796e-05,
+      "loss": 0.0041,
+      "step": 23240
+    },
+    {
+      "epoch": 1.3543426341236091,
+      "grad_norm": 0.3034536838531494,
+      "learning_rate": 1.3217537571951872e-05,
+      "loss": 0.0052,
+      "step": 23250
+    },
+    {
+      "epoch": 1.3549251470845225,
+      "grad_norm": 0.6749097108840942,
+      "learning_rate": 1.3180226635724169e-05,
+      "loss": 0.0067,
+      "step": 23260
+    },
+    {
+      "epoch": 1.3555076600454359,
+      "grad_norm": 0.1049962043762207,
+      "learning_rate": 1.314296043899298e-05,
+      "loss": 0.0037,
+      "step": 23270
+    },
+    {
+      "epoch": 1.3560901730063493,
+      "grad_norm": 0.15737871825695038,
+      "learning_rate": 1.3105739027040248e-05,
+      "loss": 0.005,
+      "step": 23280
+    },
+    {
+      "epoch": 1.3566726859672626,
+      "grad_norm": 0.20393626391887665,
+      "learning_rate": 1.3068562445093491e-05,
+      "loss": 0.0047,
+      "step": 23290
+    },
+    {
+      "epoch": 1.357255198928176,
+      "grad_norm": 0.08219223469495773,
+      "learning_rate": 1.3031430738325778e-05,
+      "loss": 0.0038,
+      "step": 23300
+    },
+    {
+      "epoch": 1.3578377118890894,
+      "grad_norm": 0.14039932191371918,
+      "learning_rate": 1.299434395185563e-05,
+      "loss": 0.0034,
+      "step": 23310
+    },
+    {
+      "epoch": 1.3584202248500028,
+      "grad_norm": 0.21193276345729828,
+      "learning_rate": 1.295730213074699e-05,
+      "loss": 0.0034,
+      "step": 23320
+    },
+    {
+      "epoch": 1.3590027378109162,
+      "grad_norm": 0.18919794261455536,
+      "learning_rate": 1.2920305320009152e-05,
+      "loss": 0.0044,
+      "step": 23330
+    },
+    {
+      "epoch": 1.3595852507718296,
+      "grad_norm": 0.3358156383037567,
+      "learning_rate": 1.2883353564596729e-05,
+      "loss": 0.0035,
+      "step": 23340
+    },
+    {
+      "epoch": 1.360167763732743,
+      "grad_norm": 0.1414111852645874,
+      "learning_rate": 1.2846446909409604e-05,
+      "loss": 0.0038,
+      "step": 23350
+    },
+    {
+      "epoch": 1.3607502766936563,
+      "grad_norm": 0.11386580765247345,
+      "learning_rate": 1.280958539929284e-05,
+      "loss": 0.0024,
+      "step": 23360
+    },
+    {
+      "epoch": 1.3613327896545697,
+      "grad_norm": 0.15156708657741547,
+      "learning_rate": 1.277276907903664e-05,
+      "loss": 0.0032,
+      "step": 23370
+    },
+    {
+      "epoch": 1.361915302615483,
+      "grad_norm": 0.38691481947898865,
+      "learning_rate": 1.2735997993376298e-05,
+      "loss": 0.0043,
+      "step": 23380
+    },
+    {
+      "epoch": 1.3624978155763965,
+      "grad_norm": 0.24201275408267975,
+      "learning_rate": 1.2699272186992166e-05,
+      "loss": 0.0053,
+      "step": 23390
+    },
+    {
+      "epoch": 1.3630803285373099,
+      "grad_norm": 0.13019780814647675,
+      "learning_rate": 1.2662591704509547e-05,
+      "loss": 0.0037,
+      "step": 23400
+    },
+    {
+      "epoch": 1.3636628414982233,
+      "grad_norm": 0.14919491112232208,
+      "learning_rate": 1.2625956590498711e-05,
+      "loss": 0.0039,
+      "step": 23410
+    },
+    {
+      "epoch": 1.3642453544591366,
+      "grad_norm": 0.16809628903865814,
+      "learning_rate": 1.2589366889474757e-05,
+      "loss": 0.0039,
+      "step": 23420
+    },
+    {
+      "epoch": 1.36482786742005,
+      "grad_norm": 0.1353694498538971,
+      "learning_rate": 1.2552822645897621e-05,
+      "loss": 0.0038,
+      "step": 23430
+    },
+    {
+      "epoch": 1.3654103803809634,
+      "grad_norm": 0.2508366107940674,
+      "learning_rate": 1.2516323904172e-05,
+      "loss": 0.0035,
+      "step": 23440
+    },
+    {
+      "epoch": 1.3659928933418768,
+      "grad_norm": 0.15308399498462677,
+      "learning_rate": 1.2479870708647324e-05,
+      "loss": 0.004,
+      "step": 23450
+    },
+    {
+      "epoch": 1.3665754063027902,
+      "grad_norm": 0.23776814341545105,
+      "learning_rate": 1.2443463103617658e-05,
+      "loss": 0.0054,
+      "step": 23460
+    },
+    {
+      "epoch": 1.3671579192637036,
+      "grad_norm": 0.1886814534664154,
+      "learning_rate": 1.240710113332167e-05,
+      "loss": 0.0031,
+      "step": 23470
+    },
+    {
+      "epoch": 1.367740432224617,
+      "grad_norm": 0.21150608360767365,
+      "learning_rate": 1.2370784841942584e-05,
+      "loss": 0.005,
+      "step": 23480
+    },
+    {
+      "epoch": 1.3683229451855303,
+      "grad_norm": 0.2955150306224823,
+      "learning_rate": 1.2334514273608117e-05,
+      "loss": 0.0039,
+      "step": 23490
+    },
+    {
+      "epoch": 1.3689054581464437,
+      "grad_norm": 0.3890087306499481,
+      "learning_rate": 1.2298289472390417e-05,
+      "loss": 0.0057,
+      "step": 23500
+    },
+    {
+      "epoch": 1.369487971107357,
+      "grad_norm": 0.16496486961841583,
+      "learning_rate": 1.226211048230606e-05,
+      "loss": 0.0031,
+      "step": 23510
+    },
+    {
+      "epoch": 1.3700704840682705,
+      "grad_norm": 0.2794024348258972,
+      "learning_rate": 1.2225977347315921e-05,
+      "loss": 0.0043,
+      "step": 23520
+    },
+    {
+      "epoch": 1.3706529970291839,
+      "grad_norm": 0.12854237854480743,
+      "learning_rate": 1.2189890111325148e-05,
+      "loss": 0.0033,
+      "step": 23530
+    },
+    {
+      "epoch": 1.3712355099900972,
+      "grad_norm": 0.1840866506099701,
+      "learning_rate": 1.2153848818183161e-05,
+      "loss": 0.0033,
+      "step": 23540
+    },
+    {
+      "epoch": 1.3718180229510106,
+      "grad_norm": 0.13824740052223206,
+      "learning_rate": 1.2117853511683507e-05,
+      "loss": 0.0023,
+      "step": 23550
+    },
+    {
+      "epoch": 1.372400535911924,
+      "grad_norm": 0.15503467619419098,
+      "learning_rate": 1.2081904235563906e-05,
+      "loss": 0.0036,
+      "step": 23560
+    },
+    {
+      "epoch": 1.3729830488728374,
+      "grad_norm": 0.13043364882469177,
+      "learning_rate": 1.20460010335061e-05,
+      "loss": 0.0038,
+      "step": 23570
+    },
+    {
+      "epoch": 1.3735655618337508,
+      "grad_norm": 0.21112245321273804,
+      "learning_rate": 1.2010143949135865e-05,
+      "loss": 0.004,
+      "step": 23580
+    },
+    {
+      "epoch": 1.3741480747946642,
+      "grad_norm": 0.16092942655086517,
+      "learning_rate": 1.1974333026022938e-05,
+      "loss": 0.0033,
+      "step": 23590
+    },
+    {
+      "epoch": 1.3747305877555775,
+      "grad_norm": 0.1612042337656021,
+      "learning_rate": 1.1938568307680964e-05,
+      "loss": 0.0043,
+      "step": 23600
+    },
+    {
+      "epoch": 1.375313100716491,
+      "grad_norm": 0.3669988811016083,
+      "learning_rate": 1.1902849837567464e-05,
+      "loss": 0.0038,
+      "step": 23610
+    },
+    {
+      "epoch": 1.3758956136774043,
+      "grad_norm": 0.2579440474510193,
+      "learning_rate": 1.1867177659083739e-05,
+      "loss": 0.0034,
+      "step": 23620
+    },
+    {
+      "epoch": 1.3764781266383177,
+      "grad_norm": 0.1823093295097351,
+      "learning_rate": 1.1831551815574848e-05,
+      "loss": 0.0042,
+      "step": 23630
+    },
+    {
+      "epoch": 1.377060639599231,
+      "grad_norm": 0.5695813894271851,
+      "learning_rate": 1.1795972350329554e-05,
+      "loss": 0.0061,
+      "step": 23640
+    },
+    {
+      "epoch": 1.3776431525601445,
+      "grad_norm": 0.23355595767498016,
+      "learning_rate": 1.176043930658024e-05,
+      "loss": 0.0045,
+      "step": 23650
+    },
+    {
+      "epoch": 1.3782256655210579,
+      "grad_norm": 0.11696689575910568,
+      "learning_rate": 1.1724952727502942e-05,
+      "loss": 0.0048,
+      "step": 23660
+    },
+    {
+      "epoch": 1.3788081784819712,
+      "grad_norm": 0.12635931372642517,
+      "learning_rate": 1.1689512656217178e-05,
+      "loss": 0.0044,
+      "step": 23670
+    },
+    {
+      "epoch": 1.3793906914428846,
+      "grad_norm": 0.1664869636297226,
+      "learning_rate": 1.1654119135785962e-05,
+      "loss": 0.0036,
+      "step": 23680
+    },
+    {
+      "epoch": 1.379973204403798,
+      "grad_norm": 0.22471028566360474,
+      "learning_rate": 1.1618772209215794e-05,
+      "loss": 0.0035,
+      "step": 23690
+    },
+    {
+      "epoch": 1.3805557173647114,
+      "grad_norm": 0.2452670931816101,
+      "learning_rate": 1.1583471919456506e-05,
+      "loss": 0.0042,
+      "step": 23700
+    },
+    {
+      "epoch": 1.3811382303256248,
+      "grad_norm": 0.1166536808013916,
+      "learning_rate": 1.1548218309401266e-05,
+      "loss": 0.0025,
+      "step": 23710
+    },
+    {
+      "epoch": 1.3817207432865382,
+      "grad_norm": 0.18245767056941986,
+      "learning_rate": 1.1513011421886555e-05,
+      "loss": 0.0042,
+      "step": 23720
+    },
+    {
+      "epoch": 1.3823032562474515,
+      "grad_norm": 0.25724342465400696,
+      "learning_rate": 1.1477851299692056e-05,
+      "loss": 0.009,
+      "step": 23730
+    },
+    {
+      "epoch": 1.382885769208365,
+      "grad_norm": 0.43923506140708923,
+      "learning_rate": 1.144273798554063e-05,
+      "loss": 0.0033,
+      "step": 23740
+    },
+    {
+      "epoch": 1.3834682821692783,
+      "grad_norm": 0.2420254945755005,
+      "learning_rate": 1.1407671522098262e-05,
+      "loss": 0.005,
+      "step": 23750
+    },
+    {
+      "epoch": 1.3840507951301917,
+      "grad_norm": 0.15894566476345062,
+      "learning_rate": 1.1372651951974e-05,
+      "loss": 0.0038,
+      "step": 23760
+    },
+    {
+      "epoch": 1.384633308091105,
+      "grad_norm": 0.13341540098190308,
+      "learning_rate": 1.1337679317719952e-05,
+      "loss": 0.0043,
+      "step": 23770
+    },
+    {
+      "epoch": 1.3852158210520185,
+      "grad_norm": 0.24180126190185547,
+      "learning_rate": 1.1302753661831151e-05,
+      "loss": 0.0029,
+      "step": 23780
+    },
+    {
+      "epoch": 1.3857983340129318,
+      "grad_norm": 0.11061865836381912,
+      "learning_rate": 1.1267875026745561e-05,
+      "loss": 0.0034,
+      "step": 23790
+    },
+    {
+      "epoch": 1.3863808469738452,
+      "grad_norm": 0.1636645346879959,
+      "learning_rate": 1.1233043454844016e-05,
+      "loss": 0.004,
+      "step": 23800
+    },
+    {
+      "epoch": 1.3869633599347586,
+      "grad_norm": 0.18001937866210938,
+      "learning_rate": 1.1198258988450144e-05,
+      "loss": 0.0036,
+      "step": 23810
+    },
+    {
+      "epoch": 1.387545872895672,
+      "grad_norm": 0.1615382879972458,
+      "learning_rate": 1.1163521669830368e-05,
+      "loss": 0.004,
+      "step": 23820
+    },
+    {
+      "epoch": 1.3881283858565854,
+      "grad_norm": 0.1342018097639084,
+      "learning_rate": 1.1128831541193818e-05,
+      "loss": 0.0037,
+      "step": 23830
+    },
+    {
+      "epoch": 1.3887108988174988,
+      "grad_norm": 0.1134992241859436,
+      "learning_rate": 1.1094188644692255e-05,
+      "loss": 0.0033,
+      "step": 23840
+    },
+    {
+      "epoch": 1.3892934117784121,
+      "grad_norm": 0.17515751719474792,
+      "learning_rate": 1.1059593022420067e-05,
+      "loss": 0.0064,
+      "step": 23850
+    },
+    {
+      "epoch": 1.3898759247393255,
+      "grad_norm": 0.21904204785823822,
+      "learning_rate": 1.1025044716414184e-05,
+      "loss": 0.0038,
+      "step": 23860
+    },
+    {
+      "epoch": 1.390458437700239,
+      "grad_norm": 0.2655951678752899,
+      "learning_rate": 1.0990543768654082e-05,
+      "loss": 0.0044,
+      "step": 23870
+    },
+    {
+      "epoch": 1.3910409506611523,
+      "grad_norm": 0.28867703676223755,
+      "learning_rate": 1.0956090221061655e-05,
+      "loss": 0.0056,
+      "step": 23880
+    },
+    {
+      "epoch": 1.3916234636220657,
+      "grad_norm": 0.2656768262386322,
+      "learning_rate": 1.0921684115501208e-05,
+      "loss": 0.0051,
+      "step": 23890
+    },
+    {
+      "epoch": 1.392205976582979,
+      "grad_norm": 0.12237775325775146,
+      "learning_rate": 1.0887325493779405e-05,
+      "loss": 0.0033,
+      "step": 23900
+    },
+    {
+      "epoch": 1.3927884895438924,
+      "grad_norm": 0.25730660557746887,
+      "learning_rate": 1.0853014397645207e-05,
+      "loss": 0.004,
+      "step": 23910
+    },
+    {
+      "epoch": 1.3933710025048058,
+      "grad_norm": 0.19858750700950623,
+      "learning_rate": 1.0818750868789829e-05,
+      "loss": 0.0061,
+      "step": 23920
+    },
+    {
+      "epoch": 1.3939535154657192,
+      "grad_norm": 0.24854832887649536,
+      "learning_rate": 1.0784534948846703e-05,
+      "loss": 0.0041,
+      "step": 23930
+    },
+    {
+      "epoch": 1.3945360284266326,
+      "grad_norm": 0.2142823487520218,
+      "learning_rate": 1.0750366679391393e-05,
+      "loss": 0.0042,
+      "step": 23940
+    },
+    {
+      "epoch": 1.395118541387546,
+      "grad_norm": 0.21719197928905487,
+      "learning_rate": 1.0716246101941557e-05,
+      "loss": 0.0037,
+      "step": 23950
+    },
+    {
+      "epoch": 1.3957010543484594,
+      "grad_norm": 0.146333247423172,
+      "learning_rate": 1.0682173257956934e-05,
+      "loss": 0.0037,
+      "step": 23960
+    },
+    {
+      "epoch": 1.3962835673093728,
+      "grad_norm": 0.05000470578670502,
+      "learning_rate": 1.064814818883922e-05,
+      "loss": 0.0038,
+      "step": 23970
+    },
+    {
+      "epoch": 1.3968660802702861,
+      "grad_norm": 0.27781733870506287,
+      "learning_rate": 1.06141709359321e-05,
+      "loss": 0.0046,
+      "step": 23980
+    },
+    {
+      "epoch": 1.3974485932311995,
+      "grad_norm": 0.20302431285381317,
+      "learning_rate": 1.058024154052114e-05,
+      "loss": 0.0046,
+      "step": 23990
+    },
+    {
+      "epoch": 1.398031106192113,
+      "grad_norm": 0.17372949421405792,
+      "learning_rate": 1.0546360043833747e-05,
+      "loss": 0.0055,
+      "step": 24000
+    },
+    {
+      "epoch": 1.3986136191530263,
+      "grad_norm": 0.1331881880760193,
+      "learning_rate": 1.0512526487039138e-05,
+      "loss": 0.0063,
+      "step": 24010
+    },
+    {
+      "epoch": 1.3991961321139397,
+      "grad_norm": 0.18121671676635742,
+      "learning_rate": 1.0478740911248258e-05,
+      "loss": 0.0054,
+      "step": 24020
+    },
+    {
+      "epoch": 1.3997786450748528,
+      "grad_norm": 0.05608072131872177,
+      "learning_rate": 1.0445003357513793e-05,
+      "loss": 0.0061,
+      "step": 24030
+    },
+    {
+      "epoch": 1.4003611580357662,
+      "grad_norm": 0.12981872260570526,
+      "learning_rate": 1.0411313866830042e-05,
+      "loss": 0.0033,
+      "step": 24040
+    },
+    {
+      "epoch": 1.4009436709966796,
+      "grad_norm": 0.20175208151340485,
+      "learning_rate": 1.0377672480132915e-05,
+      "loss": 0.0034,
+      "step": 24050
+    },
+    {
+      "epoch": 1.401526183957593,
+      "grad_norm": 0.17746217548847198,
+      "learning_rate": 1.0344079238299864e-05,
+      "loss": 0.0036,
+      "step": 24060
+    },
+    {
+      "epoch": 1.4021086969185064,
+      "grad_norm": 0.18763774633407593,
+      "learning_rate": 1.0310534182149833e-05,
+      "loss": 0.0038,
+      "step": 24070
+    },
+    {
+      "epoch": 1.4026912098794198,
+      "grad_norm": 0.13987088203430176,
+      "learning_rate": 1.0277037352443258e-05,
+      "loss": 0.003,
+      "step": 24080
+    },
+    {
+      "epoch": 1.4032737228403331,
+      "grad_norm": 0.22446969151496887,
+      "learning_rate": 1.024358878988193e-05,
+      "loss": 0.006,
+      "step": 24090
+    },
+    {
+      "epoch": 1.4038562358012465,
+      "grad_norm": 0.09927909076213837,
+      "learning_rate": 1.0210188535108995e-05,
+      "loss": 0.0032,
+      "step": 24100
+    },
+    {
+      "epoch": 1.40443874876216,
+      "grad_norm": 0.14322426915168762,
+      "learning_rate": 1.0176836628708935e-05,
+      "loss": 0.003,
+      "step": 24110
+    },
+    {
+      "epoch": 1.4050212617230733,
+      "grad_norm": 0.20255301892757416,
+      "learning_rate": 1.0143533111207454e-05,
+      "loss": 0.0059,
+      "step": 24120
+    },
+    {
+      "epoch": 1.4056037746839867,
+      "grad_norm": 0.28289714455604553,
+      "learning_rate": 1.0110278023071446e-05,
+      "loss": 0.0026,
+      "step": 24130
+    },
+    {
+      "epoch": 1.4061862876449,
+      "grad_norm": 0.11662022769451141,
+      "learning_rate": 1.007707140470901e-05,
+      "loss": 0.0031,
+      "step": 24140
+    },
+    {
+      "epoch": 1.4067688006058134,
+      "grad_norm": 0.15197624266147614,
+      "learning_rate": 1.0043913296469298e-05,
+      "loss": 0.0025,
+      "step": 24150
+    },
+    {
+      "epoch": 1.4073513135667268,
+      "grad_norm": 0.13926014304161072,
+      "learning_rate": 1.001080373864255e-05,
+      "loss": 0.0045,
+      "step": 24160
+    },
+    {
+      "epoch": 1.4079338265276402,
+      "grad_norm": 0.17042744159698486,
+      "learning_rate": 9.977742771459991e-06,
+      "loss": 0.0028,
+      "step": 24170
+    },
+    {
+      "epoch": 1.4085163394885536,
+      "grad_norm": 0.18261465430259705,
+      "learning_rate": 9.944730435093802e-06,
+      "loss": 0.0041,
+      "step": 24180
+    },
+    {
+      "epoch": 1.409098852449467,
+      "grad_norm": 0.25808078050613403,
+      "learning_rate": 9.911766769657115e-06,
+      "loss": 0.0043,
+      "step": 24190
+    },
+    {
+      "epoch": 1.4096813654103804,
+      "grad_norm": 0.11658655852079391,
+      "learning_rate": 9.878851815203882e-06,
+      "loss": 0.003,
+      "step": 24200
+    },
+    {
+      "epoch": 1.4102638783712937,
+      "grad_norm": 0.2575288414955139,
+      "learning_rate": 9.845985611728864e-06,
+      "loss": 0.0031,
+      "step": 24210
+    },
+    {
+      "epoch": 1.4108463913322071,
+      "grad_norm": 0.11595244705677032,
+      "learning_rate": 9.813168199167606e-06,
+      "loss": 0.0037,
+      "step": 24220
+    },
+    {
+      "epoch": 1.4114289042931205,
+      "grad_norm": 0.13224220275878906,
+      "learning_rate": 9.780399617396341e-06,
+      "loss": 0.0032,
+      "step": 24230
+    },
+    {
+      "epoch": 1.412011417254034,
+      "grad_norm": 0.19940713047981262,
+      "learning_rate": 9.747679906232016e-06,
+      "loss": 0.004,
+      "step": 24240
+    },
+    {
+      "epoch": 1.4125939302149473,
+      "grad_norm": 0.2140982449054718,
+      "learning_rate": 9.715009105432138e-06,
+      "loss": 0.0029,
+      "step": 24250
+    },
+    {
+      "epoch": 1.4131764431758607,
+      "grad_norm": 0.13364163041114807,
+      "learning_rate": 9.682387254694835e-06,
+      "loss": 0.0034,
+      "step": 24260
+    },
+    {
+      "epoch": 1.413758956136774,
+      "grad_norm": 0.16118764877319336,
+      "learning_rate": 9.649814393658723e-06,
+      "loss": 0.004,
+      "step": 24270
+    },
+    {
+      "epoch": 1.4143414690976874,
+      "grad_norm": 0.1829240620136261,
+      "learning_rate": 9.617290561902881e-06,
+      "loss": 0.0034,
+      "step": 24280
+    },
+    {
+      "epoch": 1.4149239820586008,
+      "grad_norm": 0.14758509397506714,
+      "learning_rate": 9.584815798946861e-06,
+      "loss": 0.0032,
+      "step": 24290
+    },
+    {
+      "epoch": 1.4155064950195142,
+      "grad_norm": 0.26349347829818726,
+      "learning_rate": 9.55239014425055e-06,
+      "loss": 0.0034,
+      "step": 24300
+    },
+    {
+      "epoch": 1.4160890079804276,
+      "grad_norm": 0.0985504761338234,
+      "learning_rate": 9.520013637214176e-06,
+      "loss": 0.0032,
+      "step": 24310
+    },
+    {
+      "epoch": 1.416671520941341,
+      "grad_norm": 0.17721882462501526,
+      "learning_rate": 9.48768631717824e-06,
+      "loss": 0.0045,
+      "step": 24320
+    },
+    {
+      "epoch": 1.4172540339022544,
+      "grad_norm": 0.20563365519046783,
+      "learning_rate": 9.455408223423496e-06,
+      "loss": 0.0034,
+      "step": 24330
+    },
+    {
+      "epoch": 1.4178365468631677,
+      "grad_norm": 0.1904628723859787,
+      "learning_rate": 9.423179395170845e-06,
+      "loss": 0.0033,
+      "step": 24340
+    },
+    {
+      "epoch": 1.4184190598240811,
+      "grad_norm": 0.2277880311012268,
+      "learning_rate": 9.390999871581391e-06,
+      "loss": 0.0037,
+      "step": 24350
+    },
+    {
+      "epoch": 1.4190015727849945,
+      "grad_norm": 0.23124796152114868,
+      "learning_rate": 9.358869691756273e-06,
+      "loss": 0.0039,
+      "step": 24360
+    },
+    {
+      "epoch": 1.4195840857459079,
+      "grad_norm": 0.1538913995027542,
+      "learning_rate": 9.326788894736688e-06,
+      "loss": 0.0029,
+      "step": 24370
+    },
+    {
+      "epoch": 1.4201665987068213,
+      "grad_norm": 0.1608511507511139,
+      "learning_rate": 9.294757519503811e-06,
+      "loss": 0.0029,
+      "step": 24380
+    },
+    {
+      "epoch": 1.4207491116677347,
+      "grad_norm": 0.5165594220161438,
+      "learning_rate": 9.262775604978819e-06,
+      "loss": 0.007,
+      "step": 24390
+    },
+    {
+      "epoch": 1.421331624628648,
+      "grad_norm": 0.15663011372089386,
+      "learning_rate": 9.230843190022725e-06,
+      "loss": 0.003,
+      "step": 24400
+    },
+    {
+      "epoch": 1.4219141375895614,
+      "grad_norm": 0.13411806523799896,
+      "learning_rate": 9.198960313436445e-06,
+      "loss": 0.0044,
+      "step": 24410
+    },
+    {
+      "epoch": 1.4224966505504748,
+      "grad_norm": 0.20790554583072662,
+      "learning_rate": 9.16712701396067e-06,
+      "loss": 0.0057,
+      "step": 24420
+    },
+    {
+      "epoch": 1.4230791635113882,
+      "grad_norm": 0.10211014747619629,
+      "learning_rate": 9.135343330275864e-06,
+      "loss": 0.003,
+      "step": 24430
+    },
+    {
+      "epoch": 1.4236616764723016,
+      "grad_norm": 0.11212743818759918,
+      "learning_rate": 9.10360930100218e-06,
+      "loss": 0.0031,
+      "step": 24440
+    },
+    {
+      "epoch": 1.424244189433215,
+      "grad_norm": 0.22522151470184326,
+      "learning_rate": 9.07192496469949e-06,
+      "loss": 0.0044,
+      "step": 24450
+    },
+    {
+      "epoch": 1.4248267023941283,
+      "grad_norm": 0.105314701795578,
+      "learning_rate": 9.040290359867231e-06,
+      "loss": 0.0059,
+      "step": 24460
+    },
+    {
+      "epoch": 1.4254092153550417,
+      "grad_norm": 0.20832829177379608,
+      "learning_rate": 9.008705524944438e-06,
+      "loss": 0.0057,
+      "step": 24470
+    },
+    {
+      "epoch": 1.425991728315955,
+      "grad_norm": 0.2444380819797516,
+      "learning_rate": 8.977170498309651e-06,
+      "loss": 0.0026,
+      "step": 24480
+    },
+    {
+      "epoch": 1.4265742412768685,
+      "grad_norm": 0.1428886353969574,
+      "learning_rate": 8.945685318280916e-06,
+      "loss": 0.0033,
+      "step": 24490
+    },
+    {
+      "epoch": 1.4271567542377817,
+      "grad_norm": 0.0868973582983017,
+      "learning_rate": 8.914250023115672e-06,
+      "loss": 0.0033,
+      "step": 24500
+    },
+    {
+      "epoch": 1.427739267198695,
+      "grad_norm": 0.21695952117443085,
+      "learning_rate": 8.882864651010798e-06,
+      "loss": 0.0026,
+      "step": 24510
+    },
+    {
+      "epoch": 1.4283217801596084,
+      "grad_norm": 0.10803350061178207,
+      "learning_rate": 8.851529240102462e-06,
+      "loss": 0.0043,
+      "step": 24520
+    },
+    {
+      "epoch": 1.4289042931205218,
+      "grad_norm": 0.2090524286031723,
+      "learning_rate": 8.820243828466134e-06,
+      "loss": 0.0039,
+      "step": 24530
+    },
+    {
+      "epoch": 1.4294868060814352,
+      "grad_norm": 0.09620046615600586,
+      "learning_rate": 8.789008454116566e-06,
+      "loss": 0.0047,
+      "step": 24540
+    },
+    {
+      "epoch": 1.4300693190423486,
+      "grad_norm": 0.08627811074256897,
+      "learning_rate": 8.757823155007655e-06,
+      "loss": 0.0029,
+      "step": 24550
+    },
+    {
+      "epoch": 1.430651832003262,
+      "grad_norm": 0.15781013667583466,
+      "learning_rate": 8.72668796903251e-06,
+      "loss": 0.0028,
+      "step": 24560
+    },
+    {
+      "epoch": 1.4312343449641753,
+      "grad_norm": 0.11579106748104095,
+      "learning_rate": 8.695602934023301e-06,
+      "loss": 0.0027,
+      "step": 24570
+    },
+    {
+      "epoch": 1.4318168579250887,
+      "grad_norm": 0.0647030621767044,
+      "learning_rate": 8.664568087751274e-06,
+      "loss": 0.0029,
+      "step": 24580
+    },
+    {
+      "epoch": 1.432399370886002,
+      "grad_norm": 0.3143619894981384,
+      "learning_rate": 8.633583467926697e-06,
+      "loss": 0.005,
+      "step": 24590
+    },
+    {
+      "epoch": 1.4329818838469155,
+      "grad_norm": 0.15916238725185394,
+      "learning_rate": 8.602649112198796e-06,
+      "loss": 0.0056,
+      "step": 24600
+    },
+    {
+      "epoch": 1.4335643968078289,
+      "grad_norm": 0.0785314291715622,
+      "learning_rate": 8.571765058155745e-06,
+      "loss": 0.0048,
+      "step": 24610
+    },
+    {
+      "epoch": 1.4341469097687423,
+      "grad_norm": 0.1807757467031479,
+      "learning_rate": 8.540931343324583e-06,
+      "loss": 0.0043,
+      "step": 24620
+    },
+    {
+      "epoch": 1.4347294227296556,
+      "grad_norm": 0.2653921842575073,
+      "learning_rate": 8.51014800517117e-06,
+      "loss": 0.0043,
+      "step": 24630
+    },
+    {
+      "epoch": 1.435311935690569,
+      "grad_norm": 0.19354289770126343,
+      "learning_rate": 8.479415081100167e-06,
+      "loss": 0.0026,
+      "step": 24640
+    },
+    {
+      "epoch": 1.4358944486514824,
+      "grad_norm": 0.18783220648765564,
+      "learning_rate": 8.448732608454967e-06,
+      "loss": 0.0027,
+      "step": 24650
+    },
+    {
+      "epoch": 1.4364769616123958,
+      "grad_norm": 0.2452276051044464,
+      "learning_rate": 8.418100624517688e-06,
+      "loss": 0.0028,
+      "step": 24660
+    },
+    {
+      "epoch": 1.4370594745733092,
+      "grad_norm": 0.21859857439994812,
+      "learning_rate": 8.387519166509062e-06,
+      "loss": 0.004,
+      "step": 24670
+    },
+    {
+      "epoch": 1.4376419875342226,
+      "grad_norm": 0.06705216318368912,
+      "learning_rate": 8.356988271588445e-06,
+      "loss": 0.0032,
+      "step": 24680
+    },
+    {
+      "epoch": 1.438224500495136,
+      "grad_norm": 0.15397319197654724,
+      "learning_rate": 8.326507976853765e-06,
+      "loss": 0.0057,
+      "step": 24690
+    },
+    {
+      "epoch": 1.4388070134560493,
+      "grad_norm": 0.18246577680110931,
+      "learning_rate": 8.296078319341443e-06,
+      "loss": 0.0036,
+      "step": 24700
+    },
+    {
+      "epoch": 1.4393895264169627,
+      "grad_norm": 0.15032833814620972,
+      "learning_rate": 8.265699336026383e-06,
+      "loss": 0.0027,
+      "step": 24710
+    },
+    {
+      "epoch": 1.439972039377876,
+      "grad_norm": 0.09603174030780792,
+      "learning_rate": 8.235371063821923e-06,
+      "loss": 0.0051,
+      "step": 24720
+    },
+    {
+      "epoch": 1.4405545523387895,
+      "grad_norm": 0.12466403096914291,
+      "learning_rate": 8.205093539579767e-06,
+      "loss": 0.0033,
+      "step": 24730
+    },
+    {
+      "epoch": 1.4411370652997029,
+      "grad_norm": 0.31038451194763184,
+      "learning_rate": 8.174866800089963e-06,
+      "loss": 0.003,
+      "step": 24740
+    },
+    {
+      "epoch": 1.4417195782606163,
+      "grad_norm": 0.15687616169452667,
+      "learning_rate": 8.144690882080853e-06,
+      "loss": 0.0039,
+      "step": 24750
+    },
+    {
+      "epoch": 1.4423020912215296,
+      "grad_norm": 0.0916990339756012,
+      "learning_rate": 8.114565822219006e-06,
+      "loss": 0.003,
+      "step": 24760
+    },
+    {
+      "epoch": 1.442884604182443,
+      "grad_norm": 0.1866670697927475,
+      "learning_rate": 8.084491657109233e-06,
+      "loss": 0.0025,
+      "step": 24770
+    },
+    {
+      "epoch": 1.4434671171433564,
+      "grad_norm": 0.270288348197937,
+      "learning_rate": 8.054468423294476e-06,
+      "loss": 0.0041,
+      "step": 24780
+    },
+    {
+      "epoch": 1.4440496301042698,
+      "grad_norm": 0.18497419357299805,
+      "learning_rate": 8.024496157255785e-06,
+      "loss": 0.0035,
+      "step": 24790
+    },
+    {
+      "epoch": 1.4446321430651832,
+      "grad_norm": 0.10803688317537308,
+      "learning_rate": 7.994574895412293e-06,
+      "loss": 0.0046,
+      "step": 24800
+    },
+    {
+      "epoch": 1.4452146560260966,
+      "grad_norm": 0.10443883389234543,
+      "learning_rate": 7.964704674121149e-06,
+      "loss": 0.0023,
+      "step": 24810
+    },
+    {
+      "epoch": 1.44579716898701,
+      "grad_norm": 0.06344673037528992,
+      "learning_rate": 7.934885529677505e-06,
+      "loss": 0.0025,
+      "step": 24820
+    },
+    {
+      "epoch": 1.4463796819479233,
+      "grad_norm": 0.20122681558132172,
+      "learning_rate": 7.905117498314412e-06,
+      "loss": 0.0031,
+      "step": 24830
+    },
+    {
+      "epoch": 1.4469621949088367,
+      "grad_norm": 0.14655786752700806,
+      "learning_rate": 7.87540061620286e-06,
+      "loss": 0.003,
+      "step": 24840
+    },
+    {
+      "epoch": 1.44754470786975,
+      "grad_norm": 0.16399699449539185,
+      "learning_rate": 7.845734919451647e-06,
+      "loss": 0.0043,
+      "step": 24850
+    },
+    {
+      "epoch": 1.4481272208306635,
+      "grad_norm": 0.13855710625648499,
+      "learning_rate": 7.816120444107383e-06,
+      "loss": 0.0041,
+      "step": 24860
+    },
+    {
+      "epoch": 1.4487097337915769,
+      "grad_norm": 0.2229846566915512,
+      "learning_rate": 7.786557226154473e-06,
+      "loss": 0.0022,
+      "step": 24870
+    },
+    {
+      "epoch": 1.4492922467524902,
+      "grad_norm": 0.32148638367652893,
+      "learning_rate": 7.757045301514998e-06,
+      "loss": 0.0055,
+      "step": 24880
+    },
+    {
+      "epoch": 1.4498747597134036,
+      "grad_norm": 0.10996751487255096,
+      "learning_rate": 7.727584706048735e-06,
+      "loss": 0.0029,
+      "step": 24890
+    },
+    {
+      "epoch": 1.450457272674317,
+      "grad_norm": 0.20284827053546906,
+      "learning_rate": 7.698175475553077e-06,
+      "loss": 0.0026,
+      "step": 24900
+    },
+    {
+      "epoch": 1.4510397856352304,
+      "grad_norm": 0.11274287104606628,
+      "learning_rate": 7.668817645763021e-06,
+      "loss": 0.0028,
+      "step": 24910
+    },
+    {
+      "epoch": 1.4516222985961438,
+      "grad_norm": 0.09294421970844269,
+      "learning_rate": 7.639511252351088e-06,
+      "loss": 0.0032,
+      "step": 24920
+    },
+    {
+      "epoch": 1.4522048115570572,
+      "grad_norm": 0.055887654423713684,
+      "learning_rate": 7.610256330927324e-06,
+      "loss": 0.0031,
+      "step": 24930
+    },
+    {
+      "epoch": 1.4527873245179705,
+      "grad_norm": 0.2652095854282379,
+      "learning_rate": 7.581052917039211e-06,
+      "loss": 0.0071,
+      "step": 24940
+    },
+    {
+      "epoch": 1.453369837478884,
+      "grad_norm": 0.14466264843940735,
+      "learning_rate": 7.551901046171645e-06,
+      "loss": 0.0027,
+      "step": 24950
+    },
+    {
+      "epoch": 1.4539523504397973,
+      "grad_norm": 0.1966245025396347,
+      "learning_rate": 7.522800753746895e-06,
+      "loss": 0.0032,
+      "step": 24960
+    },
+    {
+      "epoch": 1.4545348634007107,
+      "grad_norm": 0.24737770855426788,
+      "learning_rate": 7.493752075124577e-06,
+      "loss": 0.0084,
+      "step": 24970
+    },
+    {
+      "epoch": 1.455117376361624,
+      "grad_norm": 0.14933402836322784,
+      "learning_rate": 7.464755045601557e-06,
+      "loss": 0.0034,
+      "step": 24980
+    },
+    {
+      "epoch": 1.4556998893225375,
+      "grad_norm": 0.35500356554985046,
+      "learning_rate": 7.435809700411972e-06,
+      "loss": 0.0031,
+      "step": 24990
+    },
+    {
+      "epoch": 1.4562824022834508,
+      "grad_norm": 0.11826625466346741,
+      "learning_rate": 7.40691607472715e-06,
+      "loss": 0.0024,
+      "step": 25000
+    },
+    {
+      "epoch": 1.4568649152443642,
+      "grad_norm": 0.26528671383857727,
+      "learning_rate": 7.3780742036555606e-06,
+      "loss": 0.0064,
+      "step": 25010
+    },
+    {
+      "epoch": 1.4574474282052776,
+      "grad_norm": 0.32382839918136597,
+      "learning_rate": 7.349284122242783e-06,
+      "loss": 0.0035,
+      "step": 25020
+    },
+    {
+      "epoch": 1.458029941166191,
+      "grad_norm": 0.1456727385520935,
+      "learning_rate": 7.320545865471512e-06,
+      "loss": 0.002,
+      "step": 25030
+    },
+    {
+      "epoch": 1.4586124541271044,
+      "grad_norm": 0.23837947845458984,
+      "learning_rate": 7.291859468261425e-06,
+      "loss": 0.0026,
+      "step": 25040
+    },
+    {
+      "epoch": 1.4591949670880178,
+      "grad_norm": 0.18037672340869904,
+      "learning_rate": 7.263224965469195e-06,
+      "loss": 0.0032,
+      "step": 25050
+    },
+    {
+      "epoch": 1.4597774800489312,
+      "grad_norm": 0.08026488870382309,
+      "learning_rate": 7.234642391888446e-06,
+      "loss": 0.0025,
+      "step": 25060
+    },
+    {
+      "epoch": 1.4603599930098445,
+      "grad_norm": 0.21792194247245789,
+      "learning_rate": 7.206111782249697e-06,
+      "loss": 0.0043,
+      "step": 25070
+    },
+    {
+      "epoch": 1.460942505970758,
+      "grad_norm": 0.08107858151197433,
+      "learning_rate": 7.1776331712203385e-06,
+      "loss": 0.0025,
+      "step": 25080
+    },
+    {
+      "epoch": 1.4615250189316713,
+      "grad_norm": 0.14643537998199463,
+      "learning_rate": 7.149206593404561e-06,
+      "loss": 0.0041,
+      "step": 25090
+    },
+    {
+      "epoch": 1.4621075318925847,
+      "grad_norm": 0.14042890071868896,
+      "learning_rate": 7.120832083343337e-06,
+      "loss": 0.0034,
+      "step": 25100
+    },
+    {
+      "epoch": 1.462690044853498,
+      "grad_norm": 0.09650041162967682,
+      "learning_rate": 7.092509675514369e-06,
+      "loss": 0.0036,
+      "step": 25110
+    },
+    {
+      "epoch": 1.4632725578144115,
+      "grad_norm": 0.09727079421281815,
+      "learning_rate": 7.064239404332063e-06,
+      "loss": 0.0049,
+      "step": 25120
+    },
+    {
+      "epoch": 1.4638550707753248,
+      "grad_norm": 0.046747226268053055,
+      "learning_rate": 7.03602130414745e-06,
+      "loss": 0.0025,
+      "step": 25130
+    },
+    {
+      "epoch": 1.4644375837362382,
+      "grad_norm": 0.24280959367752075,
+      "learning_rate": 7.007855409248198e-06,
+      "loss": 0.003,
+      "step": 25140
+    },
+    {
+      "epoch": 1.4650200966971516,
+      "grad_norm": 0.21582315862178802,
+      "learning_rate": 6.979741753858521e-06,
+      "loss": 0.0072,
+      "step": 25150
+    },
+    {
+      "epoch": 1.465602609658065,
+      "grad_norm": 0.14108766615390778,
+      "learning_rate": 6.951680372139158e-06,
+      "loss": 0.0035,
+      "step": 25160
+    },
+    {
+      "epoch": 1.4661851226189784,
+      "grad_norm": 0.09983344376087189,
+      "learning_rate": 6.923671298187334e-06,
+      "loss": 0.0036,
+      "step": 25170
+    },
+    {
+      "epoch": 1.4667676355798918,
+      "grad_norm": 0.09320273250341415,
+      "learning_rate": 6.895714566036704e-06,
+      "loss": 0.0025,
+      "step": 25180
+    },
+    {
+      "epoch": 1.4673501485408051,
+      "grad_norm": 0.3548181653022766,
+      "learning_rate": 6.86781020965736e-06,
+      "loss": 0.0034,
+      "step": 25190
+    },
+    {
+      "epoch": 1.4679326615017185,
+      "grad_norm": 0.08287105709314346,
+      "learning_rate": 6.839958262955709e-06,
+      "loss": 0.0024,
+      "step": 25200
+    },
+    {
+      "epoch": 1.468515174462632,
+      "grad_norm": 0.05883297324180603,
+      "learning_rate": 6.8121587597744894e-06,
+      "loss": 0.0026,
+      "step": 25210
+    },
+    {
+      "epoch": 1.4690976874235453,
+      "grad_norm": 0.28665056824684143,
+      "learning_rate": 6.784411733892732e-06,
+      "loss": 0.0033,
+      "step": 25220
+    },
+    {
+      "epoch": 1.4696802003844587,
+      "grad_norm": 0.1197194829583168,
+      "learning_rate": 6.756717219025666e-06,
+      "loss": 0.0028,
+      "step": 25230
+    },
+    {
+      "epoch": 1.470262713345372,
+      "grad_norm": 0.12402444332838058,
+      "learning_rate": 6.729075248824762e-06,
+      "loss": 0.0028,
+      "step": 25240
+    },
+    {
+      "epoch": 1.4708452263062854,
+      "grad_norm": 0.2753455340862274,
+      "learning_rate": 6.701485856877615e-06,
+      "loss": 0.004,
+      "step": 25250
+    },
+    {
+      "epoch": 1.4714277392671988,
+      "grad_norm": 0.06682582944631577,
+      "learning_rate": 6.673949076707925e-06,
+      "loss": 0.0028,
+      "step": 25260
+    },
+    {
+      "epoch": 1.472010252228112,
+      "grad_norm": 0.369096577167511,
+      "learning_rate": 6.646464941775499e-06,
+      "loss": 0.0048,
+      "step": 25270
+    },
+    {
+      "epoch": 1.4725927651890254,
+      "grad_norm": 0.21290220320224762,
+      "learning_rate": 6.619033485476128e-06,
+      "loss": 0.0032,
+      "step": 25280
+    },
+    {
+      "epoch": 1.4731752781499388,
+      "grad_norm": 0.08899926394224167,
+      "learning_rate": 6.59165474114164e-06,
+      "loss": 0.0025,
+      "step": 25290
+    },
+    {
+      "epoch": 1.4737577911108521,
+      "grad_norm": 0.1396801471710205,
+      "learning_rate": 6.564328742039783e-06,
+      "loss": 0.0027,
+      "step": 25300
+    },
+    {
+      "epoch": 1.4743403040717655,
+      "grad_norm": 0.15877139568328857,
+      "learning_rate": 6.537055521374219e-06,
+      "loss": 0.0038,
+      "step": 25310
+    },
+    {
+      "epoch": 1.474922817032679,
+      "grad_norm": 0.2479664534330368,
+      "learning_rate": 6.509835112284485e-06,
+      "loss": 0.0048,
+      "step": 25320
+    },
+    {
+      "epoch": 1.4755053299935923,
+      "grad_norm": 0.13254977762699127,
+      "learning_rate": 6.482667547845944e-06,
+      "loss": 0.0036,
+      "step": 25330
+    },
+    {
+      "epoch": 1.4760878429545057,
+      "grad_norm": 0.13986629247665405,
+      "learning_rate": 6.455552861069736e-06,
+      "loss": 0.0031,
+      "step": 25340
+    },
+    {
+      "epoch": 1.476670355915419,
+      "grad_norm": 0.17195871472358704,
+      "learning_rate": 6.428491084902788e-06,
+      "loss": 0.0037,
+      "step": 25350
+    },
+    {
+      "epoch": 1.4772528688763324,
+      "grad_norm": 0.23130644857883453,
+      "learning_rate": 6.401482252227698e-06,
+      "loss": 0.0021,
+      "step": 25360
+    },
+    {
+      "epoch": 1.4778353818372458,
+      "grad_norm": 0.15954571962356567,
+      "learning_rate": 6.37452639586274e-06,
+      "loss": 0.0051,
+      "step": 25370
+    },
+    {
+      "epoch": 1.4784178947981592,
+      "grad_norm": 0.2656429708003998,
+      "learning_rate": 6.347623548561826e-06,
+      "loss": 0.0036,
+      "step": 25380
+    },
+    {
+      "epoch": 1.4790004077590726,
+      "grad_norm": 0.16478660702705383,
+      "learning_rate": 6.320773743014441e-06,
+      "loss": 0.0034,
+      "step": 25390
+    },
+    {
+      "epoch": 1.479582920719986,
+      "grad_norm": 0.09280870109796524,
+      "learning_rate": 6.293977011845648e-06,
+      "loss": 0.0025,
+      "step": 25400
+    },
+    {
+      "epoch": 1.4801654336808994,
+      "grad_norm": 0.09120264649391174,
+      "learning_rate": 6.2672333876159835e-06,
+      "loss": 0.003,
+      "step": 25410
+    },
+    {
+      "epoch": 1.4807479466418128,
+      "grad_norm": 0.16470304131507874,
+      "learning_rate": 6.2405429028215e-06,
+      "loss": 0.0048,
+      "step": 25420
+    },
+    {
+      "epoch": 1.4813304596027261,
+      "grad_norm": 0.2831958532333374,
+      "learning_rate": 6.213905589893631e-06,
+      "loss": 0.0044,
+      "step": 25430
+    },
+    {
+      "epoch": 1.4819129725636395,
+      "grad_norm": 0.19644665718078613,
+      "learning_rate": 6.187321481199221e-06,
+      "loss": 0.0061,
+      "step": 25440
+    },
+    {
+      "epoch": 1.482495485524553,
+      "grad_norm": 0.09666575491428375,
+      "learning_rate": 6.1607906090404895e-06,
+      "loss": 0.0034,
+      "step": 25450
+    },
+    {
+      "epoch": 1.4830779984854663,
+      "grad_norm": 0.1819407343864441,
+      "learning_rate": 6.134313005654929e-06,
+      "loss": 0.0044,
+      "step": 25460
+    },
+    {
+      "epoch": 1.4836605114463797,
+      "grad_norm": 0.11287476867437363,
+      "learning_rate": 6.107888703215337e-06,
+      "loss": 0.0026,
+      "step": 25470
+    },
+    {
+      "epoch": 1.484243024407293,
+      "grad_norm": 0.13015146553516388,
+      "learning_rate": 6.081517733829722e-06,
+      "loss": 0.0033,
+      "step": 25480
+    },
+    {
+      "epoch": 1.4848255373682064,
+      "grad_norm": 0.23437786102294922,
+      "learning_rate": 6.055200129541294e-06,
+      "loss": 0.003,
+      "step": 25490
+    },
+    {
+      "epoch": 1.4854080503291198,
+      "grad_norm": 0.14347033202648163,
+      "learning_rate": 6.0289359223284434e-06,
+      "loss": 0.006,
+      "step": 25500
+    },
+    {
+      "epoch": 1.4859905632900332,
+      "grad_norm": 0.14470624923706055,
+      "learning_rate": 6.002725144104648e-06,
+      "loss": 0.0069,
+      "step": 25510
+    },
+    {
+      "epoch": 1.4865730762509466,
+      "grad_norm": 0.08264398574829102,
+      "learning_rate": 5.976567826718477e-06,
+      "loss": 0.0033,
+      "step": 25520
+    },
+    {
+      "epoch": 1.48715558921186,
+      "grad_norm": 0.24650938808918,
+      "learning_rate": 5.9504640019535315e-06,
+      "loss": 0.0059,
+      "step": 25530
+    },
+    {
+      "epoch": 1.4877381021727734,
+      "grad_norm": 0.25896814465522766,
+      "learning_rate": 5.924413701528414e-06,
+      "loss": 0.0025,
+      "step": 25540
+    },
+    {
+      "epoch": 1.4883206151336867,
+      "grad_norm": 0.1539149135351181,
+      "learning_rate": 5.898416957096703e-06,
+      "loss": 0.005,
+      "step": 25550
+    },
+    {
+      "epoch": 1.4889031280946001,
+      "grad_norm": 0.2849887013435364,
+      "learning_rate": 5.872473800246913e-06,
+      "loss": 0.0043,
+      "step": 25560
+    },
+    {
+      "epoch": 1.4894856410555135,
+      "grad_norm": 0.19475707411766052,
+      "learning_rate": 5.846584262502402e-06,
+      "loss": 0.0046,
+      "step": 25570
+    },
+    {
+      "epoch": 1.490068154016427,
+      "grad_norm": 0.1452820897102356,
+      "learning_rate": 5.820748375321411e-06,
+      "loss": 0.0025,
+      "step": 25580
+    },
+    {
+      "epoch": 1.4906506669773403,
+      "grad_norm": 0.19508476555347443,
+      "learning_rate": 5.794966170096977e-06,
+      "loss": 0.0037,
+      "step": 25590
+    },
+    {
+      "epoch": 1.4912331799382537,
+      "grad_norm": 0.3048933744430542,
+      "learning_rate": 5.769237678156897e-06,
+      "loss": 0.0044,
+      "step": 25600
+    },
+    {
+      "epoch": 1.491815692899167,
+      "grad_norm": 0.10157933831214905,
+      "learning_rate": 5.743562930763735e-06,
+      "loss": 0.0031,
+      "step": 25610
+    },
+    {
+      "epoch": 1.4923982058600804,
+      "grad_norm": 0.08195566385984421,
+      "learning_rate": 5.717941959114726e-06,
+      "loss": 0.0037,
+      "step": 25620
+    },
+    {
+      "epoch": 1.4929807188209938,
+      "grad_norm": 0.10283743590116501,
+      "learning_rate": 5.69237479434176e-06,
+      "loss": 0.0039,
+      "step": 25630
+    },
+    {
+      "epoch": 1.4935632317819072,
+      "grad_norm": 0.11221250146627426,
+      "learning_rate": 5.666861467511353e-06,
+      "loss": 0.0043,
+      "step": 25640
+    },
+    {
+      "epoch": 1.4941457447428206,
+      "grad_norm": 0.15689042210578918,
+      "learning_rate": 5.641402009624591e-06,
+      "loss": 0.0033,
+      "step": 25650
+    },
+    {
+      "epoch": 1.494728257703734,
+      "grad_norm": 0.22902750968933105,
+      "learning_rate": 5.615996451617145e-06,
+      "loss": 0.0032,
+      "step": 25660
+    },
+    {
+      "epoch": 1.4953107706646473,
+      "grad_norm": 0.17224305868148804,
+      "learning_rate": 5.590644824359148e-06,
+      "loss": 0.0027,
+      "step": 25670
+    },
+    {
+      "epoch": 1.4958932836255607,
+      "grad_norm": 0.1224016547203064,
+      "learning_rate": 5.56534715865521e-06,
+      "loss": 0.0035,
+      "step": 25680
+    },
+    {
+      "epoch": 1.4964757965864741,
+      "grad_norm": 0.18591760098934174,
+      "learning_rate": 5.5401034852443965e-06,
+      "loss": 0.0034,
+      "step": 25690
+    },
+    {
+      "epoch": 1.4970583095473875,
+      "grad_norm": 0.3415587246417999,
+      "learning_rate": 5.514913834800134e-06,
+      "loss": 0.0052,
+      "step": 25700
+    },
+    {
+      "epoch": 1.4976408225083009,
+      "grad_norm": 0.19143666326999664,
+      "learning_rate": 5.489778237930238e-06,
+      "loss": 0.003,
+      "step": 25710
+    },
+    {
+      "epoch": 1.4982233354692143,
+      "grad_norm": 0.1019892543554306,
+      "learning_rate": 5.464696725176827e-06,
+      "loss": 0.0029,
+      "step": 25720
+    },
+    {
+      "epoch": 1.4988058484301277,
+      "grad_norm": 0.4203032851219177,
+      "learning_rate": 5.439669327016294e-06,
+      "loss": 0.0042,
+      "step": 25730
+    },
+    {
+      "epoch": 1.4993883613910408,
+      "grad_norm": 0.17117981612682343,
+      "learning_rate": 5.4146960738592985e-06,
+      "loss": 0.0035,
+      "step": 25740
+    },
+    {
+      "epoch": 1.4999708743519542,
+      "grad_norm": 0.0628904178738594,
+      "learning_rate": 5.389776996050694e-06,
+      "loss": 0.0021,
+      "step": 25750
+    },
+    {
+      "epoch": 1.5005533873128676,
+      "grad_norm": 0.11588001251220703,
+      "learning_rate": 5.364912123869492e-06,
+      "loss": 0.0033,
+      "step": 25760
+    },
+    {
+      "epoch": 1.501135900273781,
+      "grad_norm": 0.16536039113998413,
+      "learning_rate": 5.3401014875288864e-06,
+      "loss": 0.0024,
+      "step": 25770
+    },
+    {
+      "epoch": 1.5017184132346943,
+      "grad_norm": 0.2869170308113098,
+      "learning_rate": 5.315345117176129e-06,
+      "loss": 0.0039,
+      "step": 25780
+    },
+    {
+      "epoch": 1.5023009261956077,
+      "grad_norm": 0.08399692177772522,
+      "learning_rate": 5.2906430428925405e-06,
+      "loss": 0.0031,
+      "step": 25790
+    },
+    {
+      "epoch": 1.5028834391565211,
+      "grad_norm": 0.12237927317619324,
+      "learning_rate": 5.265995294693471e-06,
+      "loss": 0.0022,
+      "step": 25800
+    },
+    {
+      "epoch": 1.5034659521174345,
+      "grad_norm": 0.21054264903068542,
+      "learning_rate": 5.241401902528254e-06,
+      "loss": 0.003,
+      "step": 25810
+    },
+    {
+      "epoch": 1.5040484650783479,
+      "grad_norm": 0.2158876359462738,
+      "learning_rate": 5.216862896280189e-06,
+      "loss": 0.0023,
+      "step": 25820
+    },
+    {
+      "epoch": 1.5046309780392613,
+      "grad_norm": 0.3090130686759949,
+      "learning_rate": 5.192378305766471e-06,
+      "loss": 0.0031,
+      "step": 25830
+    },
+    {
+      "epoch": 1.5052134910001747,
+      "grad_norm": 0.14856159687042236,
+      "learning_rate": 5.1679481607382065e-06,
+      "loss": 0.0032,
+      "step": 25840
+    },
+    {
+      "epoch": 1.505796003961088,
+      "grad_norm": 0.18629679083824158,
+      "learning_rate": 5.1435724908803105e-06,
+      "loss": 0.0043,
+      "step": 25850
+    },
+    {
+      "epoch": 1.5063785169220014,
+      "grad_norm": 0.057980045676231384,
+      "learning_rate": 5.119251325811514e-06,
+      "loss": 0.0041,
+      "step": 25860
+    },
+    {
+      "epoch": 1.5069610298829148,
+      "grad_norm": 0.23540790379047394,
+      "learning_rate": 5.094984695084348e-06,
+      "loss": 0.0032,
+      "step": 25870
+    },
+    {
+      "epoch": 1.5075435428438282,
+      "grad_norm": 0.16529834270477295,
+      "learning_rate": 5.070772628185039e-06,
+      "loss": 0.0033,
+      "step": 25880
+    },
+    {
+      "epoch": 1.5081260558047416,
+      "grad_norm": 0.07812958210706711,
+      "learning_rate": 5.046615154533535e-06,
+      "loss": 0.0036,
+      "step": 25890
+    },
+    {
+      "epoch": 1.508708568765655,
+      "grad_norm": 0.27338865399360657,
+      "learning_rate": 5.022512303483451e-06,
+      "loss": 0.004,
+      "step": 25900
+    },
+    {
+      "epoch": 1.5092910817265683,
+      "grad_norm": 0.08028960973024368,
+      "learning_rate": 4.998464104322015e-06,
+      "loss": 0.0035,
+      "step": 25910
+    },
+    {
+      "epoch": 1.5098735946874817,
+      "grad_norm": 0.16899392008781433,
+      "learning_rate": 4.974470586270047e-06,
+      "loss": 0.0027,
+      "step": 25920
+    },
+    {
+      "epoch": 1.510456107648395,
+      "grad_norm": 0.2159395068883896,
+      "learning_rate": 4.950531778481965e-06,
+      "loss": 0.0028,
+      "step": 25930
+    },
+    {
+      "epoch": 1.5110386206093085,
+      "grad_norm": 0.14772860705852509,
+      "learning_rate": 4.926647710045651e-06,
+      "loss": 0.0037,
+      "step": 25940
+    },
+    {
+      "epoch": 1.5116211335702219,
+      "grad_norm": 0.13572624325752258,
+      "learning_rate": 4.9028184099825125e-06,
+      "loss": 0.003,
+      "step": 25950
+    },
+    {
+      "epoch": 1.5122036465311353,
+      "grad_norm": 0.15965153276920319,
+      "learning_rate": 4.879043907247382e-06,
+      "loss": 0.0019,
+      "step": 25960
+    },
+    {
+      "epoch": 1.5127861594920486,
+      "grad_norm": 0.3129093050956726,
+      "learning_rate": 4.8553242307285414e-06,
+      "loss": 0.004,
+      "step": 25970
+    },
+    {
+      "epoch": 1.513368672452962,
+      "grad_norm": 0.06297149509191513,
+      "learning_rate": 4.8316594092476185e-06,
+      "loss": 0.0017,
+      "step": 25980
+    },
+    {
+      "epoch": 1.5139511854138754,
+      "grad_norm": 0.22972433269023895,
+      "learning_rate": 4.808049471559617e-06,
+      "loss": 0.0047,
+      "step": 25990
+    },
+    {
+      "epoch": 1.5145336983747888,
+      "grad_norm": 0.298460453748703,
+      "learning_rate": 4.784494446352833e-06,
+      "loss": 0.0033,
+      "step": 26000
+    },
+    {
+      "epoch": 1.5151162113357022,
+      "grad_norm": 0.09093844145536423,
+      "learning_rate": 4.760994362248833e-06,
+      "loss": 0.0027,
+      "step": 26010
+    },
+    {
+      "epoch": 1.5156987242966156,
+      "grad_norm": 0.1293647438287735,
+      "learning_rate": 4.737549247802442e-06,
+      "loss": 0.0033,
+      "step": 26020
+    },
+    {
+      "epoch": 1.516281237257529,
+      "grad_norm": 0.17233596742153168,
+      "learning_rate": 4.714159131501689e-06,
+      "loss": 0.0052,
+      "step": 26030
+    },
+    {
+      "epoch": 1.5168637502184423,
+      "grad_norm": 0.08634297549724579,
+      "learning_rate": 4.690824041767766e-06,
+      "loss": 0.003,
+      "step": 26040
+    },
+    {
+      "epoch": 1.5174462631793557,
+      "grad_norm": 0.11946024745702744,
+      "learning_rate": 4.6675440069550105e-06,
+      "loss": 0.0026,
+      "step": 26050
+    },
+    {
+      "epoch": 1.518028776140269,
+      "grad_norm": 0.2795431613922119,
+      "learning_rate": 4.64431905535086e-06,
+      "loss": 0.0035,
+      "step": 26060
+    },
+    {
+      "epoch": 1.5186112891011825,
+      "grad_norm": 0.1300330013036728,
+      "learning_rate": 4.621149215175808e-06,
+      "loss": 0.0052,
+      "step": 26070
+    },
+    {
+      "epoch": 1.5191938020620959,
+      "grad_norm": 0.13417604565620422,
+      "learning_rate": 4.5980345145834154e-06,
+      "loss": 0.004,
+      "step": 26080
+    },
+    {
+      "epoch": 1.5197763150230092,
+      "grad_norm": 0.29234766960144043,
+      "learning_rate": 4.574974981660213e-06,
+      "loss": 0.0028,
+      "step": 26090
+    },
+    {
+      "epoch": 1.5203588279839226,
+      "grad_norm": 0.11330074071884155,
+      "learning_rate": 4.551970644425707e-06,
+      "loss": 0.0037,
+      "step": 26100
+    },
+    {
+      "epoch": 1.520941340944836,
+      "grad_norm": 0.11604847013950348,
+      "learning_rate": 4.529021530832328e-06,
+      "loss": 0.0043,
+      "step": 26110
+    },
+    {
+      "epoch": 1.5215238539057494,
+      "grad_norm": 0.2992563545703888,
+      "learning_rate": 4.5061276687654306e-06,
+      "loss": 0.0041,
+      "step": 26120
+    },
+    {
+      "epoch": 1.5221063668666628,
+      "grad_norm": 0.12537245452404022,
+      "learning_rate": 4.4832890860431965e-06,
+      "loss": 0.0029,
+      "step": 26130
+    },
+    {
+      "epoch": 1.5226888798275762,
+      "grad_norm": 0.1463560163974762,
+      "learning_rate": 4.460505810416682e-06,
+      "loss": 0.0023,
+      "step": 26140
+    },
+    {
+      "epoch": 1.5232713927884896,
+      "grad_norm": 0.23925863206386566,
+      "learning_rate": 4.4377778695696984e-06,
+      "loss": 0.0045,
+      "step": 26150
+    },
+    {
+      "epoch": 1.523853905749403,
+      "grad_norm": 0.19198858737945557,
+      "learning_rate": 4.415105291118843e-06,
+      "loss": 0.0033,
+      "step": 26160
+    },
+    {
+      "epoch": 1.5244364187103163,
+      "grad_norm": 0.2662869989871979,
+      "learning_rate": 4.392488102613435e-06,
+      "loss": 0.0031,
+      "step": 26170
+    },
+    {
+      "epoch": 1.5250189316712297,
+      "grad_norm": 0.1913996785879135,
+      "learning_rate": 4.369926331535473e-06,
+      "loss": 0.0041,
+      "step": 26180
+    },
+    {
+      "epoch": 1.525601444632143,
+      "grad_norm": 0.08066805452108383,
+      "learning_rate": 4.347420005299668e-06,
+      "loss": 0.0024,
+      "step": 26190
+    },
+    {
+      "epoch": 1.5261839575930565,
+      "grad_norm": 0.2606082260608673,
+      "learning_rate": 4.324969151253317e-06,
+      "loss": 0.0033,
+      "step": 26200
+    },
+    {
+      "epoch": 1.5267664705539699,
+      "grad_norm": 0.16689768433570862,
+      "learning_rate": 4.302573796676312e-06,
+      "loss": 0.0034,
+      "step": 26210
+    },
+    {
+      "epoch": 1.5273489835148832,
+      "grad_norm": 0.14520810544490814,
+      "learning_rate": 4.2802339687811385e-06,
+      "loss": 0.0028,
+      "step": 26220
+    },
+    {
+      "epoch": 1.5279314964757966,
+      "grad_norm": 0.1750641018152237,
+      "learning_rate": 4.257949694712771e-06,
+      "loss": 0.0035,
+      "step": 26230
+    },
+    {
+      "epoch": 1.52851400943671,
+      "grad_norm": 0.17732131481170654,
+      "learning_rate": 4.235721001548726e-06,
+      "loss": 0.0027,
+      "step": 26240
+    },
+    {
+      "epoch": 1.5290965223976234,
+      "grad_norm": 0.175058513879776,
+      "learning_rate": 4.21354791629896e-06,
+      "loss": 0.004,
+      "step": 26250
+    },
+    {
+      "epoch": 1.5296790353585368,
+      "grad_norm": 0.189947709441185,
+      "learning_rate": 4.191430465905843e-06,
+      "loss": 0.0037,
+      "step": 26260
+    },
+    {
+      "epoch": 1.5302615483194502,
+      "grad_norm": 0.19996023178100586,
+      "learning_rate": 4.169368677244184e-06,
+      "loss": 0.003,
+      "step": 26270
+    },
+    {
+      "epoch": 1.5308440612803635,
+      "grad_norm": 0.0878337025642395,
+      "learning_rate": 4.147362577121122e-06,
+      "loss": 0.0026,
+      "step": 26280
+    },
+    {
+      "epoch": 1.531426574241277,
+      "grad_norm": 0.10889554768800735,
+      "learning_rate": 4.125412192276157e-06,
+      "loss": 0.0035,
+      "step": 26290
+    },
+    {
+      "epoch": 1.5320090872021903,
+      "grad_norm": 0.10601271688938141,
+      "learning_rate": 4.10351754938107e-06,
+      "loss": 0.0024,
+      "step": 26300
+    },
+    {
+      "epoch": 1.5325916001631037,
+      "grad_norm": 0.09924892336130142,
+      "learning_rate": 4.081678675039913e-06,
+      "loss": 0.0032,
+      "step": 26310
+    },
+    {
+      "epoch": 1.533174113124017,
+      "grad_norm": 0.13006731867790222,
+      "learning_rate": 4.0598955957889876e-06,
+      "loss": 0.0026,
+      "step": 26320
+    },
+    {
+      "epoch": 1.5337566260849305,
+      "grad_norm": 0.23325128853321075,
+      "learning_rate": 4.038168338096776e-06,
+      "loss": 0.0027,
+      "step": 26330
+    },
+    {
+      "epoch": 1.5343391390458438,
+      "grad_norm": 0.08871504664421082,
+      "learning_rate": 4.016496928363944e-06,
+      "loss": 0.0023,
+      "step": 26340
+    },
+    {
+      "epoch": 1.5349216520067572,
+      "grad_norm": 0.2486826479434967,
+      "learning_rate": 3.994881392923317e-06,
+      "loss": 0.0041,
+      "step": 26350
+    },
+    {
+      "epoch": 1.5355041649676706,
+      "grad_norm": 0.1687488555908203,
+      "learning_rate": 3.9733217580397945e-06,
+      "loss": 0.0026,
+      "step": 26360
+    },
+    {
+      "epoch": 1.536086677928584,
+      "grad_norm": 0.13622665405273438,
+      "learning_rate": 3.951818049910366e-06,
+      "loss": 0.003,
+      "step": 26370
+    },
+    {
+      "epoch": 1.5366691908894974,
+      "grad_norm": 0.17355699837207794,
+      "learning_rate": 3.930370294664071e-06,
+      "loss": 0.0039,
+      "step": 26380
+    },
+    {
+      "epoch": 1.5372517038504108,
+      "grad_norm": 0.18918152153491974,
+      "learning_rate": 3.908978518361938e-06,
+      "loss": 0.0027,
+      "step": 26390
+    },
+    {
+      "epoch": 1.5378342168113242,
+      "grad_norm": 0.19185519218444824,
+      "learning_rate": 3.887642746997017e-06,
+      "loss": 0.0023,
+      "step": 26400
+    },
+    {
+      "epoch": 1.5384167297722375,
+      "grad_norm": 0.12481196969747543,
+      "learning_rate": 3.866363006494255e-06,
+      "loss": 0.0035,
+      "step": 26410
+    },
+    {
+      "epoch": 1.538999242733151,
+      "grad_norm": 0.12779243290424347,
+      "learning_rate": 3.8451393227105725e-06,
+      "loss": 0.0031,
+      "step": 26420
+    },
+    {
+      "epoch": 1.5395817556940643,
+      "grad_norm": 0.06346990913152695,
+      "learning_rate": 3.823971721434727e-06,
+      "loss": 0.0035,
+      "step": 26430
+    },
+    {
+      "epoch": 1.5401642686549777,
+      "grad_norm": 0.08583824336528778,
+      "learning_rate": 3.8028602283873505e-06,
+      "loss": 0.0037,
+      "step": 26440
+    },
+    {
+      "epoch": 1.540746781615891,
+      "grad_norm": 0.1683126538991928,
+      "learning_rate": 3.781804869220912e-06,
+      "loss": 0.0057,
+      "step": 26450
+    },
+    {
+      "epoch": 1.5413292945768045,
+      "grad_norm": 0.10807096213102341,
+      "learning_rate": 3.760805669519646e-06,
+      "loss": 0.0039,
+      "step": 26460
+    },
+    {
+      "epoch": 1.5419118075377178,
+      "grad_norm": 0.10149163752794266,
+      "learning_rate": 3.7398626547995587e-06,
+      "loss": 0.0022,
+      "step": 26470
+    },
+    {
+      "epoch": 1.5424943204986312,
+      "grad_norm": 0.2520512044429779,
+      "learning_rate": 3.7189758505083995e-06,
+      "loss": 0.002,
+      "step": 26480
+    },
+    {
+      "epoch": 1.5430768334595446,
+      "grad_norm": 0.3238174319267273,
+      "learning_rate": 3.6981452820255837e-06,
+      "loss": 0.0025,
+      "step": 26490
+    },
+    {
+      "epoch": 1.543659346420458,
+      "grad_norm": 0.16209842264652252,
+      "learning_rate": 3.6773709746622363e-06,
+      "loss": 0.003,
+      "step": 26500
+    },
+    {
+      "epoch": 1.5442418593813714,
+      "grad_norm": 0.16427388787269592,
+      "learning_rate": 3.6566529536610995e-06,
+      "loss": 0.0026,
+      "step": 26510
+    },
+    {
+      "epoch": 1.5448243723422848,
+      "grad_norm": 0.18118585646152496,
+      "learning_rate": 3.635991244196513e-06,
+      "loss": 0.0057,
+      "step": 26520
+    },
+    {
+      "epoch": 1.5454068853031981,
+      "grad_norm": 0.08404374867677689,
+      "learning_rate": 3.615385871374405e-06,
+      "loss": 0.002,
+      "step": 26530
+    },
+    {
+      "epoch": 1.5459893982641115,
+      "grad_norm": 0.07793246954679489,
+      "learning_rate": 3.5948368602322367e-06,
+      "loss": 0.0046,
+      "step": 26540
+    },
+    {
+      "epoch": 1.546571911225025,
+      "grad_norm": 0.32492876052856445,
+      "learning_rate": 3.5743442357390187e-06,
+      "loss": 0.0033,
+      "step": 26550
+    },
+    {
+      "epoch": 1.5471544241859383,
+      "grad_norm": 0.25567471981048584,
+      "learning_rate": 3.5539080227951937e-06,
+      "loss": 0.0034,
+      "step": 26560
+    },
+    {
+      "epoch": 1.5477369371468517,
+      "grad_norm": 0.2852281630039215,
+      "learning_rate": 3.533528246232709e-06,
+      "loss": 0.0033,
+      "step": 26570
+    },
+    {
+      "epoch": 1.548319450107765,
+      "grad_norm": 0.17486290633678436,
+      "learning_rate": 3.5132049308149117e-06,
+      "loss": 0.0046,
+      "step": 26580
+    },
+    {
+      "epoch": 1.5489019630686784,
+      "grad_norm": 0.1875961422920227,
+      "learning_rate": 3.4929381012365425e-06,
+      "loss": 0.0027,
+      "step": 26590
+    },
+    {
+      "epoch": 1.5494844760295916,
+      "grad_norm": 0.12731048464775085,
+      "learning_rate": 3.472727782123697e-06,
+      "loss": 0.0038,
+      "step": 26600
+    },
+    {
+      "epoch": 1.550066988990505,
+      "grad_norm": 0.3298400044441223,
+      "learning_rate": 3.4525739980338425e-06,
+      "loss": 0.0045,
+      "step": 26610
+    },
+    {
+      "epoch": 1.5506495019514184,
+      "grad_norm": 0.15442371368408203,
+      "learning_rate": 3.432476773455712e-06,
+      "loss": 0.0026,
+      "step": 26620
+    },
+    {
+      "epoch": 1.5512320149123318,
+      "grad_norm": 0.20735491812229156,
+      "learning_rate": 3.4124361328093378e-06,
+      "loss": 0.0029,
+      "step": 26630
+    },
+    {
+      "epoch": 1.5518145278732451,
+      "grad_norm": 0.424357146024704,
+      "learning_rate": 3.392452100445975e-06,
+      "loss": 0.0061,
+      "step": 26640
+    },
+    {
+      "epoch": 1.5523970408341585,
+      "grad_norm": 0.214350625872612,
+      "learning_rate": 3.3725247006481163e-06,
+      "loss": 0.0063,
+      "step": 26650
+    },
+    {
+      "epoch": 1.552979553795072,
+      "grad_norm": 0.12077348679304123,
+      "learning_rate": 3.3526539576294313e-06,
+      "loss": 0.0026,
+      "step": 26660
+    },
+    {
+      "epoch": 1.5535620667559853,
+      "grad_norm": 0.03553283214569092,
+      "learning_rate": 3.332839895534745e-06,
+      "loss": 0.0028,
+      "step": 26670
+    },
+    {
+      "epoch": 1.5541445797168987,
+      "grad_norm": 0.2798008918762207,
+      "learning_rate": 3.3130825384400155e-06,
+      "loss": 0.0028,
+      "step": 26680
+    },
+    {
+      "epoch": 1.554727092677812,
+      "grad_norm": 0.11738568544387817,
+      "learning_rate": 3.293381910352278e-06,
+      "loss": 0.0032,
+      "step": 26690
+    },
+    {
+      "epoch": 1.5553096056387254,
+      "grad_norm": 0.16866131126880646,
+      "learning_rate": 3.273738035209678e-06,
+      "loss": 0.0028,
+      "step": 26700
+    },
+    {
+      "epoch": 1.5558921185996388,
+      "grad_norm": 0.32970762252807617,
+      "learning_rate": 3.254150936881356e-06,
+      "loss": 0.0044,
+      "step": 26710
+    },
+    {
+      "epoch": 1.5564746315605522,
+      "grad_norm": 0.4446662664413452,
+      "learning_rate": 3.234620639167496e-06,
+      "loss": 0.0034,
+      "step": 26720
+    },
+    {
+      "epoch": 1.5570571445214656,
+      "grad_norm": 0.1426686942577362,
+      "learning_rate": 3.2151471657992482e-06,
+      "loss": 0.0021,
+      "step": 26730
+    },
+    {
+      "epoch": 1.557639657482379,
+      "grad_norm": 0.10345635563135147,
+      "learning_rate": 3.195730540438718e-06,
+      "loss": 0.0048,
+      "step": 26740
+    },
+    {
+      "epoch": 1.5582221704432924,
+      "grad_norm": 0.1697050929069519,
+      "learning_rate": 3.1763707866789336e-06,
+      "loss": 0.0035,
+      "step": 26750
+    },
+    {
+      "epoch": 1.5588046834042057,
+      "grad_norm": 0.07912324368953705,
+      "learning_rate": 3.1570679280438165e-06,
+      "loss": 0.0035,
+      "step": 26760
+    },
+    {
+      "epoch": 1.5593871963651191,
+      "grad_norm": 0.08897348493337631,
+      "learning_rate": 3.1378219879881767e-06,
+      "loss": 0.003,
+      "step": 26770
+    },
+    {
+      "epoch": 1.5599697093260325,
+      "grad_norm": 0.18549391627311707,
+      "learning_rate": 3.11863298989763e-06,
+      "loss": 0.0024,
+      "step": 26780
+    },
+    {
+      "epoch": 1.560552222286946,
+      "grad_norm": 0.164683997631073,
+      "learning_rate": 3.0995009570886303e-06,
+      "loss": 0.0027,
+      "step": 26790
+    },
+    {
+      "epoch": 1.5611347352478593,
+      "grad_norm": 0.29790300130844116,
+      "learning_rate": 3.0804259128083936e-06,
+      "loss": 0.0042,
+      "step": 26800
+    },
+    {
+      "epoch": 1.5617172482087727,
+      "grad_norm": 0.09718704223632812,
+      "learning_rate": 3.06140788023489e-06,
+      "loss": 0.0037,
+      "step": 26810
+    },
+    {
+      "epoch": 1.562299761169686,
+      "grad_norm": 0.2995389699935913,
+      "learning_rate": 3.042446882476846e-06,
+      "loss": 0.0028,
+      "step": 26820
+    },
+    {
+      "epoch": 1.5628822741305994,
+      "grad_norm": 0.3960285484790802,
+      "learning_rate": 3.023542942573643e-06,
+      "loss": 0.0036,
+      "step": 26830
+    },
+    {
+      "epoch": 1.5634647870915128,
+      "grad_norm": 0.0684681311249733,
+      "learning_rate": 3.004696083495351e-06,
+      "loss": 0.0045,
+      "step": 26840
+    },
+    {
+      "epoch": 1.5640473000524262,
+      "grad_norm": 0.1886732280254364,
+      "learning_rate": 2.98590632814269e-06,
+      "loss": 0.0031,
+      "step": 26850
+    },
+    {
+      "epoch": 1.5646298130133396,
+      "grad_norm": 0.2736036479473114,
+      "learning_rate": 2.9671736993469702e-06,
+      "loss": 0.0068,
+      "step": 26860
+    },
+    {
+      "epoch": 1.565212325974253,
+      "grad_norm": 0.2760659158229828,
+      "learning_rate": 2.9484982198701216e-06,
+      "loss": 0.0032,
+      "step": 26870
+    },
+    {
+      "epoch": 1.5657948389351664,
+      "grad_norm": 0.31369152665138245,
+      "learning_rate": 2.929879912404604e-06,
+      "loss": 0.0047,
+      "step": 26880
+    },
+    {
+      "epoch": 1.5663773518960797,
+      "grad_norm": 0.060979146510362625,
+      "learning_rate": 2.9113187995734147e-06,
+      "loss": 0.0022,
+      "step": 26890
+    },
+    {
+      "epoch": 1.5669598648569931,
+      "grad_norm": 0.3609210252761841,
+      "learning_rate": 2.8928149039300524e-06,
+      "loss": 0.0032,
+      "step": 26900
+    },
+    {
+      "epoch": 1.5675423778179063,
+      "grad_norm": 0.11160704493522644,
+      "learning_rate": 2.8743682479584975e-06,
+      "loss": 0.0043,
+      "step": 26910
+    },
+    {
+      "epoch": 1.5681248907788197,
+      "grad_norm": 0.09147461503744125,
+      "learning_rate": 2.8559788540731824e-06,
+      "loss": 0.0023,
+      "step": 26920
+    },
+    {
+      "epoch": 1.568707403739733,
+      "grad_norm": 0.06852752715349197,
+      "learning_rate": 2.8376467446189493e-06,
+      "loss": 0.0021,
+      "step": 26930
+    },
+    {
+      "epoch": 1.5692899167006464,
+      "grad_norm": 0.1702806055545807,
+      "learning_rate": 2.8193719418710406e-06,
+      "loss": 0.0069,
+      "step": 26940
+    },
+    {
+      "epoch": 1.5698724296615598,
+      "grad_norm": 0.07028792053461075,
+      "learning_rate": 2.8011544680350667e-06,
+      "loss": 0.0031,
+      "step": 26950
+    },
+    {
+      "epoch": 1.5704549426224732,
+      "grad_norm": 0.24736250936985016,
+      "learning_rate": 2.7829943452469753e-06,
+      "loss": 0.0037,
+      "step": 26960
+    },
+    {
+      "epoch": 1.5710374555833866,
+      "grad_norm": 0.10018304735422134,
+      "learning_rate": 2.7648915955730213e-06,
+      "loss": 0.0033,
+      "step": 26970
+    },
+    {
+      "epoch": 1.5716199685443,
+      "grad_norm": 0.0906628742814064,
+      "learning_rate": 2.746846241009765e-06,
+      "loss": 0.0032,
+      "step": 26980
+    },
+    {
+      "epoch": 1.5722024815052134,
+      "grad_norm": 0.124571293592453,
+      "learning_rate": 2.7288583034839945e-06,
+      "loss": 0.0058,
+      "step": 26990
+    },
+    {
+      "epoch": 1.5727849944661267,
+      "grad_norm": 0.12843795120716095,
+      "learning_rate": 2.7109278048527752e-06,
+      "loss": 0.0034,
+      "step": 27000
+    },
+    {
+      "epoch": 1.5733675074270401,
+      "grad_norm": 0.16245298087596893,
+      "learning_rate": 2.6930547669033413e-06,
+      "loss": 0.0046,
+      "step": 27010
+    },
+    {
+      "epoch": 1.5739500203879535,
+      "grad_norm": 0.10931997746229172,
+      "learning_rate": 2.675239211353109e-06,
+      "loss": 0.0028,
+      "step": 27020
+    },
+    {
+      "epoch": 1.574532533348867,
+      "grad_norm": 0.13476353883743286,
+      "learning_rate": 2.6574811598496786e-06,
+      "loss": 0.0022,
+      "step": 27030
+    },
+    {
+      "epoch": 1.5751150463097803,
+      "grad_norm": 0.18880543112754822,
+      "learning_rate": 2.6397806339707456e-06,
+      "loss": 0.002,
+      "step": 27040
+    },
+    {
+      "epoch": 1.5756975592706937,
+      "grad_norm": 0.17403216660022736,
+      "learning_rate": 2.6221376552241217e-06,
+      "loss": 0.0024,
+      "step": 27050
+    },
+    {
+      "epoch": 1.576280072231607,
+      "grad_norm": 0.12779001891613007,
+      "learning_rate": 2.604552245047681e-06,
+      "loss": 0.0029,
+      "step": 27060
+    },
+    {
+      "epoch": 1.5768625851925204,
+      "grad_norm": 0.25050631165504456,
+      "learning_rate": 2.587024424809359e-06,
+      "loss": 0.0031,
+      "step": 27070
+    },
+    {
+      "epoch": 1.5774450981534338,
+      "grad_norm": 0.1696874052286148,
+      "learning_rate": 2.5695542158071186e-06,
+      "loss": 0.0042,
+      "step": 27080
+    },
+    {
+      "epoch": 1.5780276111143472,
+      "grad_norm": 0.22831350564956665,
+      "learning_rate": 2.5521416392689068e-06,
+      "loss": 0.0058,
+      "step": 27090
+    },
+    {
+      "epoch": 1.5786101240752606,
+      "grad_norm": 0.08078908920288086,
+      "learning_rate": 2.5347867163526384e-06,
+      "loss": 0.0044,
+      "step": 27100
+    },
+    {
+      "epoch": 1.579192637036174,
+      "grad_norm": 0.15348710119724274,
+      "learning_rate": 2.517489468146189e-06,
+      "loss": 0.0037,
+      "step": 27110
+    },
+    {
+      "epoch": 1.5797751499970873,
+      "grad_norm": 0.12756769359111786,
+      "learning_rate": 2.500249915667341e-06,
+      "loss": 0.0023,
+      "step": 27120
+    },
+    {
+      "epoch": 1.5803576629580007,
+      "grad_norm": 0.202960804104805,
+      "learning_rate": 2.4830680798637817e-06,
+      "loss": 0.004,
+      "step": 27130
+    },
+    {
+      "epoch": 1.5809401759189141,
+      "grad_norm": 0.12151604145765305,
+      "learning_rate": 2.465943981613056e-06,
+      "loss": 0.0039,
+      "step": 27140
+    },
+    {
+      "epoch": 1.5815226888798275,
+      "grad_norm": 0.11082343757152557,
+      "learning_rate": 2.448877641722569e-06,
+      "loss": 0.0037,
+      "step": 27150
+    },
+    {
+      "epoch": 1.5821052018407409,
+      "grad_norm": 0.456434428691864,
+      "learning_rate": 2.431869080929522e-06,
+      "loss": 0.0046,
+      "step": 27160
+    },
+    {
+      "epoch": 1.5826877148016543,
+      "grad_norm": 0.18470750749111176,
+      "learning_rate": 2.4149183199009216e-06,
+      "loss": 0.003,
+      "step": 27170
+    },
+    {
+      "epoch": 1.5832702277625677,
+      "grad_norm": 0.2179182916879654,
+      "learning_rate": 2.3980253792335427e-06,
+      "loss": 0.0062,
+      "step": 27180
+    },
+    {
+      "epoch": 1.583852740723481,
+      "grad_norm": 0.11319984495639801,
+      "learning_rate": 2.381190279453899e-06,
+      "loss": 0.0039,
+      "step": 27190
+    },
+    {
+      "epoch": 1.5844352536843944,
+      "grad_norm": 0.11839370429515839,
+      "learning_rate": 2.364413041018232e-06,
+      "loss": 0.003,
+      "step": 27200
+    },
+    {
+      "epoch": 1.5850177666453078,
+      "grad_norm": 0.21624349057674408,
+      "learning_rate": 2.347693684312463e-06,
+      "loss": 0.0019,
+      "step": 27210
+    },
+    {
+      "epoch": 1.5856002796062212,
+      "grad_norm": 0.19534026086330414,
+      "learning_rate": 2.331032229652186e-06,
+      "loss": 0.0036,
+      "step": 27220
+    },
+    {
+      "epoch": 1.5861827925671346,
+      "grad_norm": 0.1445130705833435,
+      "learning_rate": 2.314428697282628e-06,
+      "loss": 0.0029,
+      "step": 27230
+    },
+    {
+      "epoch": 1.586765305528048,
+      "grad_norm": 0.14219652116298676,
+      "learning_rate": 2.297883107378673e-06,
+      "loss": 0.0023,
+      "step": 27240
+    },
+    {
+      "epoch": 1.5873478184889613,
+      "grad_norm": 0.23550748825073242,
+      "learning_rate": 2.2813954800447513e-06,
+      "loss": 0.0035,
+      "step": 27250
+    },
+    {
+      "epoch": 1.5879303314498747,
+      "grad_norm": 0.17718768119812012,
+      "learning_rate": 2.2649658353148974e-06,
+      "loss": 0.004,
+      "step": 27260
+    },
+    {
+      "epoch": 1.588512844410788,
+      "grad_norm": 0.13269661366939545,
+      "learning_rate": 2.2485941931526645e-06,
+      "loss": 0.0037,
+      "step": 27270
+    },
+    {
+      "epoch": 1.5890953573717015,
+      "grad_norm": 0.16204704344272614,
+      "learning_rate": 2.232280573451151e-06,
+      "loss": 0.0032,
+      "step": 27280
+    },
+    {
+      "epoch": 1.5896778703326149,
+      "grad_norm": 0.08440358936786652,
+      "learning_rate": 2.216024996032945e-06,
+      "loss": 0.0031,
+      "step": 27290
+    },
+    {
+      "epoch": 1.5902603832935283,
+      "grad_norm": 0.05020131543278694,
+      "learning_rate": 2.1998274806501074e-06,
+      "loss": 0.0032,
+      "step": 27300
+    },
+    {
+      "epoch": 1.5908428962544416,
+      "grad_norm": 0.2244742214679718,
+      "learning_rate": 2.183688046984139e-06,
+      "loss": 0.0032,
+      "step": 27310
+    },
+    {
+      "epoch": 1.591425409215355,
+      "grad_norm": 0.15592506527900696,
+      "learning_rate": 2.167606714645981e-06,
+      "loss": 0.0028,
+      "step": 27320
+    },
+    {
+      "epoch": 1.5920079221762684,
+      "grad_norm": 0.08862439543008804,
+      "learning_rate": 2.151583503175958e-06,
+      "loss": 0.0028,
+      "step": 27330
+    },
+    {
+      "epoch": 1.5925904351371818,
+      "grad_norm": 0.13331052660942078,
+      "learning_rate": 2.1356184320437957e-06,
+      "loss": 0.0036,
+      "step": 27340
+    },
+    {
+      "epoch": 1.5931729480980952,
+      "grad_norm": 0.29033687710762024,
+      "learning_rate": 2.119711520648554e-06,
+      "loss": 0.005,
+      "step": 27350
+    },
+    {
+      "epoch": 1.5937554610590086,
+      "grad_norm": 0.12790866196155548,
+      "learning_rate": 2.103862788318628e-06,
+      "loss": 0.0033,
+      "step": 27360
+    },
+    {
+      "epoch": 1.594337974019922,
+      "grad_norm": 0.10604497790336609,
+      "learning_rate": 2.088072254311729e-06,
+      "loss": 0.0048,
+      "step": 27370
+    },
+    {
+      "epoch": 1.5949204869808353,
+      "grad_norm": 0.09602311998605728,
+      "learning_rate": 2.0723399378148435e-06,
+      "loss": 0.0026,
+      "step": 27380
+    },
+    {
+      "epoch": 1.5955029999417487,
+      "grad_norm": 0.11441168934106827,
+      "learning_rate": 2.0566658579442065e-06,
+      "loss": 0.0039,
+      "step": 27390
+    },
+    {
+      "epoch": 1.596085512902662,
+      "grad_norm": 0.0591399110853672,
+      "learning_rate": 2.0410500337453176e-06,
+      "loss": 0.0037,
+      "step": 27400
+    },
+    {
+      "epoch": 1.5966680258635755,
+      "grad_norm": 0.10361428558826447,
+      "learning_rate": 2.0254924841928645e-06,
+      "loss": 0.0028,
+      "step": 27410
+    },
+    {
+      "epoch": 1.5972505388244889,
+      "grad_norm": 0.16099022328853607,
+      "learning_rate": 2.009993228190754e-06,
+      "loss": 0.0053,
+      "step": 27420
+    },
+    {
+      "epoch": 1.5978330517854022,
+      "grad_norm": 0.08187589049339294,
+      "learning_rate": 1.9945522845720323e-06,
+      "loss": 0.0017,
+      "step": 27430
+    },
+    {
+      "epoch": 1.5984155647463156,
+      "grad_norm": 0.11524957418441772,
+      "learning_rate": 1.9791696720988963e-06,
+      "loss": 0.0023,
+      "step": 27440
+    },
+    {
+      "epoch": 1.598998077707229,
+      "grad_norm": 0.19680365920066833,
+      "learning_rate": 1.9638454094626833e-06,
+      "loss": 0.0031,
+      "step": 27450
+    },
+    {
+      "epoch": 1.5995805906681424,
+      "grad_norm": 0.18414273858070374,
+      "learning_rate": 1.9485795152838104e-06,
+      "loss": 0.0064,
+      "step": 27460
+    },
+    {
+      "epoch": 1.6001631036290558,
+      "grad_norm": 0.2880193591117859,
+      "learning_rate": 1.933372008111778e-06,
+      "loss": 0.0038,
+      "step": 27470
+    },
+    {
+      "epoch": 1.6007456165899692,
+      "grad_norm": 0.15873165428638458,
+      "learning_rate": 1.918222906425143e-06,
+      "loss": 0.0027,
+      "step": 27480
+    },
+    {
+      "epoch": 1.6013281295508826,
+      "grad_norm": 0.263339102268219,
+      "learning_rate": 1.903132228631488e-06,
+      "loss": 0.0045,
+      "step": 27490
+    },
+    {
+      "epoch": 1.601910642511796,
+      "grad_norm": 0.22501875460147858,
+      "learning_rate": 1.8880999930674215e-06,
+      "loss": 0.0039,
+      "step": 27500
+    },
+    {
+      "epoch": 1.6024931554727093,
+      "grad_norm": 0.3094823956489563,
+      "learning_rate": 1.8731262179985166e-06,
+      "loss": 0.0029,
+      "step": 27510
+    },
+    {
+      "epoch": 1.6030756684336227,
+      "grad_norm": 0.26423147320747375,
+      "learning_rate": 1.8582109216193244e-06,
+      "loss": 0.0039,
+      "step": 27520
+    },
+    {
+      "epoch": 1.603658181394536,
+      "grad_norm": 0.09433199465274811,
+      "learning_rate": 1.8433541220533368e-06,
+      "loss": 0.0019,
+      "step": 27530
+    },
+    {
+      "epoch": 1.6042406943554495,
+      "grad_norm": 0.14699028432369232,
+      "learning_rate": 1.8285558373529577e-06,
+      "loss": 0.0027,
+      "step": 27540
+    },
+    {
+      "epoch": 1.6048232073163629,
+      "grad_norm": 0.13233596086502075,
+      "learning_rate": 1.8138160854995145e-06,
+      "loss": 0.0023,
+      "step": 27550
+    },
+    {
+      "epoch": 1.6054057202772762,
+      "grad_norm": 0.1776420623064041,
+      "learning_rate": 1.7991348844031864e-06,
+      "loss": 0.0034,
+      "step": 27560
+    },
+    {
+      "epoch": 1.6059882332381896,
+      "grad_norm": 0.24940475821495056,
+      "learning_rate": 1.78451225190302e-06,
+      "loss": 0.004,
+      "step": 27570
+    },
+    {
+      "epoch": 1.606570746199103,
+      "grad_norm": 0.05166579410433769,
+      "learning_rate": 1.7699482057668914e-06,
+      "loss": 0.0025,
+      "step": 27580
+    },
+    {
+      "epoch": 1.6071532591600164,
+      "grad_norm": 0.16570842266082764,
+      "learning_rate": 1.7554427636914895e-06,
+      "loss": 0.0027,
+      "step": 27590
+    },
+    {
+      "epoch": 1.6077357721209298,
+      "grad_norm": 0.12059414386749268,
+      "learning_rate": 1.7409959433022871e-06,
+      "loss": 0.0028,
+      "step": 27600
+    },
+    {
+      "epoch": 1.6083182850818432,
+      "grad_norm": 0.1654326170682907,
+      "learning_rate": 1.7266077621535481e-06,
+      "loss": 0.0032,
+      "step": 27610
+    },
+    {
+      "epoch": 1.6089007980427565,
+      "grad_norm": 0.1670808345079422,
+      "learning_rate": 1.7122782377282597e-06,
+      "loss": 0.0029,
+      "step": 27620
+    },
+    {
+      "epoch": 1.60948331100367,
+      "grad_norm": 0.13403286039829254,
+      "learning_rate": 1.6980073874381497e-06,
+      "loss": 0.0026,
+      "step": 27630
+    },
+    {
+      "epoch": 1.6100658239645833,
+      "grad_norm": 0.08622683584690094,
+      "learning_rate": 1.6837952286236413e-06,
+      "loss": 0.0033,
+      "step": 27640
+    },
+    {
+      "epoch": 1.6106483369254967,
+      "grad_norm": 0.09247484058141708,
+      "learning_rate": 1.6696417785538488e-06,
+      "loss": 0.005,
+      "step": 27650
+    },
+    {
+      "epoch": 1.61123084988641,
+      "grad_norm": 0.21400822699069977,
+      "learning_rate": 1.6555470544265538e-06,
+      "loss": 0.0029,
+      "step": 27660
+    },
+    {
+      "epoch": 1.6118133628473235,
+      "grad_norm": 0.06220991909503937,
+      "learning_rate": 1.6415110733681738e-06,
+      "loss": 0.0036,
+      "step": 27670
+    },
+    {
+      "epoch": 1.6123958758082368,
+      "grad_norm": 0.10110336542129517,
+      "learning_rate": 1.6275338524337436e-06,
+      "loss": 0.0033,
+      "step": 27680
+    },
+    {
+      "epoch": 1.6129783887691502,
+      "grad_norm": 0.17017430067062378,
+      "learning_rate": 1.6136154086069056e-06,
+      "loss": 0.0021,
+      "step": 27690
+    },
+    {
+      "epoch": 1.6135609017300636,
+      "grad_norm": 0.09295425564050674,
+      "learning_rate": 1.5997557587998868e-06,
+      "loss": 0.0027,
+      "step": 27700
+    },
+    {
+      "epoch": 1.614143414690977,
+      "grad_norm": 0.1160450354218483,
+      "learning_rate": 1.5859549198534606e-06,
+      "loss": 0.0022,
+      "step": 27710
+    },
+    {
+      "epoch": 1.6147259276518904,
+      "grad_norm": 0.06652171164751053,
+      "learning_rate": 1.572212908536963e-06,
+      "loss": 0.0025,
+      "step": 27720
+    },
+    {
+      "epoch": 1.6153084406128038,
+      "grad_norm": 0.3194689452648163,
+      "learning_rate": 1.5585297415482202e-06,
+      "loss": 0.0053,
+      "step": 27730
+    },
+    {
+      "epoch": 1.6158909535737171,
+      "grad_norm": 0.1471574604511261,
+      "learning_rate": 1.5449054355135717e-06,
+      "loss": 0.0035,
+      "step": 27740
+    },
+    {
+      "epoch": 1.6164734665346305,
+      "grad_norm": 0.06871964782476425,
+      "learning_rate": 1.5313400069878415e-06,
+      "loss": 0.0032,
+      "step": 27750
+    },
+    {
+      "epoch": 1.617055979495544,
+      "grad_norm": 0.12766481935977936,
+      "learning_rate": 1.5178334724542887e-06,
+      "loss": 0.0032,
+      "step": 27760
+    },
+    {
+      "epoch": 1.6176384924564573,
+      "grad_norm": 0.3385268747806549,
+      "learning_rate": 1.504385848324641e-06,
+      "loss": 0.0034,
+      "step": 27770
+    },
+    {
+      "epoch": 1.6182210054173707,
+      "grad_norm": 0.15307725965976715,
+      "learning_rate": 1.490997150939033e-06,
+      "loss": 0.0033,
+      "step": 27780
+    },
+    {
+      "epoch": 1.618803518378284,
+      "grad_norm": 0.14600777626037598,
+      "learning_rate": 1.4776673965659793e-06,
+      "loss": 0.0022,
+      "step": 27790
+    },
+    {
+      "epoch": 1.6193860313391975,
+      "grad_norm": 0.14177826046943665,
+      "learning_rate": 1.4643966014023957e-06,
+      "loss": 0.0021,
+      "step": 27800
+    },
+    {
+      "epoch": 1.6199685443001108,
+      "grad_norm": 0.15549203753471375,
+      "learning_rate": 1.4511847815735503e-06,
+      "loss": 0.0041,
+      "step": 27810
+    },
+    {
+      "epoch": 1.6205510572610242,
+      "grad_norm": 0.18620546162128448,
+      "learning_rate": 1.438031953133051e-06,
+      "loss": 0.0034,
+      "step": 27820
+    },
+    {
+      "epoch": 1.6211335702219376,
+      "grad_norm": 0.11421048641204834,
+      "learning_rate": 1.4249381320628197e-06,
+      "loss": 0.0028,
+      "step": 27830
+    },
+    {
+      "epoch": 1.6217160831828508,
+      "grad_norm": 0.12926118075847626,
+      "learning_rate": 1.4119033342730902e-06,
+      "loss": 0.0026,
+      "step": 27840
+    },
+    {
+      "epoch": 1.6222985961437641,
+      "grad_norm": 0.13141915202140808,
+      "learning_rate": 1.3989275756023656e-06,
+      "loss": 0.0022,
+      "step": 27850
+    },
+    {
+      "epoch": 1.6228811091046775,
+      "grad_norm": 0.19641920924186707,
+      "learning_rate": 1.386010871817417e-06,
+      "loss": 0.0021,
+      "step": 27860
+    },
+    {
+      "epoch": 1.623463622065591,
+      "grad_norm": 0.26969674229621887,
+      "learning_rate": 1.3731532386132616e-06,
+      "loss": 0.0029,
+      "step": 27870
+    },
+    {
+      "epoch": 1.6240461350265043,
+      "grad_norm": 0.2197602242231369,
+      "learning_rate": 1.360354691613136e-06,
+      "loss": 0.0028,
+      "step": 27880
+    },
+    {
+      "epoch": 1.6246286479874177,
+      "grad_norm": 0.25574973225593567,
+      "learning_rate": 1.3476152463684776e-06,
+      "loss": 0.0036,
+      "step": 27890
+    },
+    {
+      "epoch": 1.625211160948331,
+      "grad_norm": 0.11012033373117447,
+      "learning_rate": 1.3349349183589155e-06,
+      "loss": 0.003,
+      "step": 27900
+    },
+    {
+      "epoch": 1.6257936739092445,
+      "grad_norm": 0.15684263408184052,
+      "learning_rate": 1.3223137229922356e-06,
+      "loss": 0.0032,
+      "step": 27910
+    },
+    {
+      "epoch": 1.6263761868701578,
+      "grad_norm": 0.30886030197143555,
+      "learning_rate": 1.3097516756043981e-06,
+      "loss": 0.0033,
+      "step": 27920
+    },
+    {
+      "epoch": 1.6269586998310712,
+      "grad_norm": 0.2484997808933258,
+      "learning_rate": 1.29724879145946e-06,
+      "loss": 0.0051,
+      "step": 27930
+    },
+    {
+      "epoch": 1.6275412127919846,
+      "grad_norm": 0.14981016516685486,
+      "learning_rate": 1.284805085749613e-06,
+      "loss": 0.0037,
+      "step": 27940
+    },
+    {
+      "epoch": 1.628123725752898,
+      "grad_norm": 0.03710196539759636,
+      "learning_rate": 1.2724205735951288e-06,
+      "loss": 0.0039,
+      "step": 27950
+    },
+    {
+      "epoch": 1.6287062387138114,
+      "grad_norm": 0.15386570990085602,
+      "learning_rate": 1.2600952700443591e-06,
+      "loss": 0.003,
+      "step": 27960
+    },
+    {
+      "epoch": 1.6292887516747248,
+      "grad_norm": 0.18887688219547272,
+      "learning_rate": 1.247829190073707e-06,
+      "loss": 0.0035,
+      "step": 27970
+    },
+    {
+      "epoch": 1.6298712646356381,
+      "grad_norm": 0.33417919278144836,
+      "learning_rate": 1.2356223485876173e-06,
+      "loss": 0.0028,
+      "step": 27980
+    },
+    {
+      "epoch": 1.6304537775965515,
+      "grad_norm": 0.06816980987787247,
+      "learning_rate": 1.2234747604185526e-06,
+      "loss": 0.0031,
+      "step": 27990
+    },
+    {
+      "epoch": 1.631036290557465,
+      "grad_norm": 0.25048476457595825,
+      "learning_rate": 1.2113864403269836e-06,
+      "loss": 0.0031,
+      "step": 28000
+    },
+    {
+      "epoch": 1.6316188035183783,
+      "grad_norm": 0.1470801681280136,
+      "learning_rate": 1.1993574030013555e-06,
+      "loss": 0.0028,
+      "step": 28010
+    },
+    {
+      "epoch": 1.6322013164792917,
+      "grad_norm": 0.1865183413028717,
+      "learning_rate": 1.1873876630580816e-06,
+      "loss": 0.0031,
+      "step": 28020
+    },
+    {
+      "epoch": 1.632783829440205,
+      "grad_norm": 0.12353920191526413,
+      "learning_rate": 1.1754772350415278e-06,
+      "loss": 0.0063,
+      "step": 28030
+    },
+    {
+      "epoch": 1.6333663424011184,
+      "grad_norm": 0.14580439031124115,
+      "learning_rate": 1.163626133423984e-06,
+      "loss": 0.0041,
+      "step": 28040
+    },
+    {
+      "epoch": 1.6339488553620318,
+      "grad_norm": 0.09690406173467636,
+      "learning_rate": 1.1518343726056591e-06,
+      "loss": 0.0022,
+      "step": 28050
+    },
+    {
+      "epoch": 1.6345313683229452,
+      "grad_norm": 0.1459706723690033,
+      "learning_rate": 1.1401019669146474e-06,
+      "loss": 0.0023,
+      "step": 28060
+    },
+    {
+      "epoch": 1.6351138812838586,
+      "grad_norm": 0.20409967005252838,
+      "learning_rate": 1.128428930606934e-06,
+      "loss": 0.0019,
+      "step": 28070
+    },
+    {
+      "epoch": 1.635696394244772,
+      "grad_norm": 0.05775986611843109,
+      "learning_rate": 1.1168152778663621e-06,
+      "loss": 0.0021,
+      "step": 28080
+    },
+    {
+      "epoch": 1.6362789072056854,
+      "grad_norm": 0.2404635101556778,
+      "learning_rate": 1.10526102280461e-06,
+      "loss": 0.0028,
+      "step": 28090
+    },
+    {
+      "epoch": 1.6368614201665987,
+      "grad_norm": 0.2647615671157837,
+      "learning_rate": 1.0937661794611864e-06,
+      "loss": 0.0041,
+      "step": 28100
+    },
+    {
+      "epoch": 1.6374439331275121,
+      "grad_norm": 0.112600177526474,
+      "learning_rate": 1.082330761803413e-06,
+      "loss": 0.0021,
+      "step": 28110
+    },
+    {
+      "epoch": 1.6380264460884255,
+      "grad_norm": 0.17186981439590454,
+      "learning_rate": 1.0709547837263966e-06,
+      "loss": 0.0039,
+      "step": 28120
+    },
+    {
+      "epoch": 1.638608959049339,
+      "grad_norm": 0.12882168591022491,
+      "learning_rate": 1.059638259053025e-06,
+      "loss": 0.002,
+      "step": 28130
+    },
+    {
+      "epoch": 1.6391914720102523,
+      "grad_norm": 0.04550999402999878,
+      "learning_rate": 1.0483812015339434e-06,
+      "loss": 0.002,
+      "step": 28140
+    },
+    {
+      "epoch": 1.6397739849711654,
+      "grad_norm": 0.11234162747859955,
+      "learning_rate": 1.037183624847543e-06,
+      "loss": 0.0027,
+      "step": 28150
+    },
+    {
+      "epoch": 1.6403564979320788,
+      "grad_norm": 0.11578910797834396,
+      "learning_rate": 1.0260455425999238e-06,
+      "loss": 0.0028,
+      "step": 28160
+    },
+    {
+      "epoch": 1.6409390108929922,
+      "grad_norm": 0.08051315695047379,
+      "learning_rate": 1.0149669683249208e-06,
+      "loss": 0.0029,
+      "step": 28170
+    },
+    {
+      "epoch": 1.6415215238539056,
+      "grad_norm": 0.10668817907571793,
+      "learning_rate": 1.003947915484027e-06,
+      "loss": 0.002,
+      "step": 28180
+    },
+    {
+      "epoch": 1.642104036814819,
+      "grad_norm": 0.13198666274547577,
+      "learning_rate": 9.92988397466449e-07,
+      "loss": 0.0019,
+      "step": 28190
+    },
+    {
+      "epoch": 1.6426865497757324,
+      "grad_norm": 0.19599559903144836,
+      "learning_rate": 9.820884275890286e-07,
+      "loss": 0.0024,
+      "step": 28200
+    },
+    {
+      "epoch": 1.6432690627366457,
+      "grad_norm": 0.07028232514858246,
+      "learning_rate": 9.712480190962546e-07,
+      "loss": 0.0024,
+      "step": 28210
+    },
+    {
+      "epoch": 1.6438515756975591,
+      "grad_norm": 0.38363486528396606,
+      "learning_rate": 9.604671851602464e-07,
+      "loss": 0.003,
+      "step": 28220
+    },
+    {
+      "epoch": 1.6444340886584725,
+      "grad_norm": 0.0691777691245079,
+      "learning_rate": 9.497459388807306e-07,
+      "loss": 0.0034,
+      "step": 28230
+    },
+    {
+      "epoch": 1.645016601619386,
+      "grad_norm": 0.13221475481987,
+      "learning_rate": 9.390842932850364e-07,
+      "loss": 0.0023,
+      "step": 28240
+    },
+    {
+      "epoch": 1.6455991145802993,
+      "grad_norm": 0.2454451620578766,
+      "learning_rate": 9.28482261328073e-07,
+      "loss": 0.0033,
+      "step": 28250
+    },
+    {
+      "epoch": 1.6461816275412127,
+      "grad_norm": 0.07785508036613464,
+      "learning_rate": 9.179398558923025e-07,
+      "loss": 0.0033,
+      "step": 28260
+    },
+    {
+      "epoch": 1.646764140502126,
+      "grad_norm": 0.22707445919513702,
+      "learning_rate": 9.074570897877388e-07,
+      "loss": 0.0033,
+      "step": 28270
+    },
+    {
+      "epoch": 1.6473466534630394,
+      "grad_norm": 0.29836714267730713,
+      "learning_rate": 8.970339757519375e-07,
+      "loss": 0.002,
+      "step": 28280
+    },
+    {
+      "epoch": 1.6479291664239528,
+      "grad_norm": 0.2177252322435379,
+      "learning_rate": 8.866705264499619e-07,
+      "loss": 0.0024,
+      "step": 28290
+    },
+    {
+      "epoch": 1.6485116793848662,
+      "grad_norm": 0.12949076294898987,
+      "learning_rate": 8.763667544743836e-07,
+      "loss": 0.0022,
+      "step": 28300
+    },
+    {
+      "epoch": 1.6490941923457796,
+      "grad_norm": 0.061454880982637405,
+      "learning_rate": 8.661226723452543e-07,
+      "loss": 0.0023,
+      "step": 28310
+    },
+    {
+      "epoch": 1.649676705306693,
+      "grad_norm": 0.21868835389614105,
+      "learning_rate": 8.559382925101001e-07,
+      "loss": 0.0032,
+      "step": 28320
+    },
+    {
+      "epoch": 1.6502592182676064,
+      "grad_norm": 0.1397545486688614,
+      "learning_rate": 8.458136273438943e-07,
+      "loss": 0.005,
+      "step": 28330
+    },
+    {
+      "epoch": 1.6508417312285197,
+      "grad_norm": 0.12138670682907104,
+      "learning_rate": 8.357486891490795e-07,
+      "loss": 0.0025,
+      "step": 28340
+    },
+    {
+      "epoch": 1.6514242441894331,
+      "grad_norm": 0.2001275271177292,
+      "learning_rate": 8.257434901554895e-07,
+      "loss": 0.003,
+      "step": 28350
+    },
+    {
+      "epoch": 1.6520067571503465,
+      "grad_norm": 0.15491336584091187,
+      "learning_rate": 8.157980425203937e-07,
+      "loss": 0.0019,
+      "step": 28360
+    },
+    {
+      "epoch": 1.65258927011126,
+      "grad_norm": 0.10495275259017944,
+      "learning_rate": 8.059123583284367e-07,
+      "loss": 0.003,
+      "step": 28370
+    },
+    {
+      "epoch": 1.6531717830721733,
+      "grad_norm": 0.17396701872348785,
+      "learning_rate": 7.960864495916654e-07,
+      "loss": 0.0063,
+      "step": 28380
+    },
+    {
+      "epoch": 1.6537542960330867,
+      "grad_norm": 0.152618870139122,
+      "learning_rate": 7.863203282494847e-07,
+      "loss": 0.0026,
+      "step": 28390
+    },
+    {
+      "epoch": 1.654336808994,
+      "grad_norm": 0.13482345640659332,
+      "learning_rate": 7.766140061686522e-07,
+      "loss": 0.0039,
+      "step": 28400
+    },
+    {
+      "epoch": 1.6549193219549134,
+      "grad_norm": 0.10104743391275406,
+      "learning_rate": 7.669674951432615e-07,
+      "loss": 0.0039,
+      "step": 28410
+    },
+    {
+      "epoch": 1.6555018349158268,
+      "grad_norm": 0.17268632352352142,
+      "learning_rate": 7.573808068947363e-07,
+      "loss": 0.0051,
+      "step": 28420
+    },
+    {
+      "epoch": 1.6560843478767402,
+      "grad_norm": 0.1587252914905548,
+      "learning_rate": 7.478539530718087e-07,
+      "loss": 0.003,
+      "step": 28430
+    },
+    {
+      "epoch": 1.6566668608376536,
+      "grad_norm": 0.16926072537899017,
+      "learning_rate": 7.383869452504965e-07,
+      "loss": 0.0041,
+      "step": 28440
+    },
+    {
+      "epoch": 1.657249373798567,
+      "grad_norm": 0.16699029505252838,
+      "learning_rate": 7.289797949341204e-07,
+      "loss": 0.0021,
+      "step": 28450
+    },
+    {
+      "epoch": 1.6578318867594803,
+      "grad_norm": 0.31210657954216003,
+      "learning_rate": 7.196325135532423e-07,
+      "loss": 0.0035,
+      "step": 28460
+    },
+    {
+      "epoch": 1.6584143997203937,
+      "grad_norm": 0.07599468529224396,
+      "learning_rate": 7.10345112465699e-07,
+      "loss": 0.004,
+      "step": 28470
+    },
+    {
+      "epoch": 1.6589969126813071,
+      "grad_norm": 0.14538788795471191,
+      "learning_rate": 7.011176029565525e-07,
+      "loss": 0.0028,
+      "step": 28480
+    },
+    {
+      "epoch": 1.6595794256422205,
+      "grad_norm": 0.31993383169174194,
+      "learning_rate": 6.919499962381004e-07,
+      "loss": 0.0036,
+      "step": 28490
+    },
+    {
+      "epoch": 1.6601619386031339,
+      "grad_norm": 0.20302484929561615,
+      "learning_rate": 6.828423034498488e-07,
+      "loss": 0.0026,
+      "step": 28500
+    },
+    {
+      "epoch": 1.6607444515640473,
+      "grad_norm": 0.10518087446689606,
+      "learning_rate": 6.737945356585007e-07,
+      "loss": 0.0034,
+      "step": 28510
+    },
+    {
+      "epoch": 1.6613269645249606,
+      "grad_norm": 0.3014754354953766,
+      "learning_rate": 6.648067038579508e-07,
+      "loss": 0.003,
+      "step": 28520
+    },
+    {
+      "epoch": 1.661909477485874,
+      "grad_norm": 0.07225207984447479,
+      "learning_rate": 6.558788189692578e-07,
+      "loss": 0.0023,
+      "step": 28530
+    },
+    {
+      "epoch": 1.6624919904467874,
+      "grad_norm": 0.29815351963043213,
+      "learning_rate": 6.470108918406492e-07,
+      "loss": 0.0047,
+      "step": 28540
+    },
+    {
+      "epoch": 1.6630745034077008,
+      "grad_norm": 0.1339595913887024,
+      "learning_rate": 6.382029332474893e-07,
+      "loss": 0.0031,
+      "step": 28550
+    },
+    {
+      "epoch": 1.6636570163686142,
+      "grad_norm": 0.2297477424144745,
+      "learning_rate": 6.294549538922778e-07,
+      "loss": 0.0022,
+      "step": 28560
+    },
+    {
+      "epoch": 1.6642395293295276,
+      "grad_norm": 0.047920141369104385,
+      "learning_rate": 6.207669644046344e-07,
+      "loss": 0.0018,
+      "step": 28570
+    },
+    {
+      "epoch": 1.664822042290441,
+      "grad_norm": 0.16528069972991943,
+      "learning_rate": 6.121389753412865e-07,
+      "loss": 0.0033,
+      "step": 28580
+    },
+    {
+      "epoch": 1.6654045552513543,
+      "grad_norm": 0.0906873270869255,
+      "learning_rate": 6.035709971860592e-07,
+      "loss": 0.0022,
+      "step": 28590
+    },
+    {
+      "epoch": 1.6659870682122677,
+      "grad_norm": 0.2947187125682831,
+      "learning_rate": 5.950630403498469e-07,
+      "loss": 0.0036,
+      "step": 28600
+    },
+    {
+      "epoch": 1.666569581173181,
+      "grad_norm": 0.11981963366270065,
+      "learning_rate": 5.866151151706189e-07,
+      "loss": 0.0026,
+      "step": 28610
+    },
+    {
+      "epoch": 1.6671520941340945,
+      "grad_norm": 0.07298427075147629,
+      "learning_rate": 5.782272319134086e-07,
+      "loss": 0.0069,
+      "step": 28620
+    },
+    {
+      "epoch": 1.6677346070950079,
+      "grad_norm": 0.1067531555891037,
+      "learning_rate": 5.698994007702796e-07,
+      "loss": 0.0033,
+      "step": 28630
+    },
+    {
+      "epoch": 1.6683171200559213,
+      "grad_norm": 0.13120315968990326,
+      "learning_rate": 5.616316318603321e-07,
+      "loss": 0.0032,
+      "step": 28640
+    },
+    {
+      "epoch": 1.6688996330168346,
+      "grad_norm": 0.09044377505779266,
+      "learning_rate": 5.534239352296799e-07,
+      "loss": 0.0026,
+      "step": 28650
+    },
+    {
+      "epoch": 1.669482145977748,
+      "grad_norm": 0.12720009684562683,
+      "learning_rate": 5.452763208514621e-07,
+      "loss": 0.0022,
+      "step": 28660
+    },
+    {
+      "epoch": 1.6700646589386614,
+      "grad_norm": 0.1352410465478897,
+      "learning_rate": 5.371887986257873e-07,
+      "loss": 0.0035,
+      "step": 28670
+    },
+    {
+      "epoch": 1.6706471718995748,
+      "grad_norm": 0.16141006350517273,
+      "learning_rate": 5.291613783797611e-07,
+      "loss": 0.0041,
+      "step": 28680
+    },
+    {
+      "epoch": 1.6712296848604882,
+      "grad_norm": 0.1771744340658188,
+      "learning_rate": 5.211940698674534e-07,
+      "loss": 0.0026,
+      "step": 28690
+    },
+    {
+      "epoch": 1.6718121978214016,
+      "grad_norm": 0.08664882928133011,
+      "learning_rate": 5.132868827698978e-07,
+      "loss": 0.0021,
+      "step": 28700
+    },
+    {
+      "epoch": 1.672394710782315,
+      "grad_norm": 0.11980684101581573,
+      "learning_rate": 5.054398266950755e-07,
+      "loss": 0.0026,
+      "step": 28710
+    },
+    {
+      "epoch": 1.6729772237432283,
+      "grad_norm": 0.08222827315330505,
+      "learning_rate": 4.976529111778872e-07,
+      "loss": 0.0022,
+      "step": 28720
+    },
+    {
+      "epoch": 1.6735597367041417,
+      "grad_norm": 0.049435198307037354,
+      "learning_rate": 4.899261456801862e-07,
+      "loss": 0.0024,
+      "step": 28730
+    },
+    {
+      "epoch": 1.674142249665055,
+      "grad_norm": 0.0769728273153305,
+      "learning_rate": 4.822595395907126e-07,
+      "loss": 0.0027,
+      "step": 28740
+    },
+    {
+      "epoch": 1.6747247626259685,
+      "grad_norm": 0.12158923596143723,
+      "learning_rate": 4.7465310222510886e-07,
+      "loss": 0.0021,
+      "step": 28750
+    },
+    {
+      "epoch": 1.6753072755868819,
+      "grad_norm": 0.12682054936885834,
+      "learning_rate": 4.6710684282593175e-07,
+      "loss": 0.0032,
+      "step": 28760
+    },
+    {
+      "epoch": 1.6758897885477952,
+      "grad_norm": 0.26456400752067566,
+      "learning_rate": 4.5962077056257993e-07,
+      "loss": 0.005,
+      "step": 28770
+    },
+    {
+      "epoch": 1.6764723015087086,
+      "grad_norm": 0.2087945193052292,
+      "learning_rate": 4.521948945313492e-07,
+      "loss": 0.0025,
+      "step": 28780
+    },
+    {
+      "epoch": 1.677054814469622,
+      "grad_norm": 0.287797749042511,
+      "learning_rate": 4.4482922375537195e-07,
+      "loss": 0.0038,
+      "step": 28790
+    },
+    {
+      "epoch": 1.6776373274305354,
+      "grad_norm": 0.2818441390991211,
+      "learning_rate": 4.375237671846333e-07,
+      "loss": 0.0024,
+      "step": 28800
+    },
+    {
+      "epoch": 1.6782198403914488,
+      "grad_norm": 0.16553948819637299,
+      "learning_rate": 4.302785336959547e-07,
+      "loss": 0.0026,
+      "step": 28810
+    },
+    {
+      "epoch": 1.6788023533523622,
+      "grad_norm": 0.07189997285604477,
+      "learning_rate": 4.2309353209297744e-07,
+      "loss": 0.0024,
+      "step": 28820
+    },
+    {
+      "epoch": 1.6793848663132755,
+      "grad_norm": 0.07663629204034805,
+      "learning_rate": 4.159687711061566e-07,
+      "loss": 0.002,
+      "step": 28830
+    },
+    {
+      "epoch": 1.679967379274189,
+      "grad_norm": 0.03444487228989601,
+      "learning_rate": 4.0890425939275055e-07,
+      "loss": 0.0028,
+      "step": 28840
+    },
+    {
+      "epoch": 1.6805498922351023,
+      "grad_norm": 0.1303909868001938,
+      "learning_rate": 4.0190000553679827e-07,
+      "loss": 0.0032,
+      "step": 28850
+    },
+    {
+      "epoch": 1.6811324051960157,
+      "grad_norm": 0.12830765545368195,
+      "learning_rate": 3.9495601804913627e-07,
+      "loss": 0.0035,
+      "step": 28860
+    },
+    {
+      "epoch": 1.681714918156929,
+      "grad_norm": 0.18251784145832062,
+      "learning_rate": 3.880723053673652e-07,
+      "loss": 0.0026,
+      "step": 28870
+    },
+    {
+      "epoch": 1.6822974311178425,
+      "grad_norm": 0.1985720545053482,
+      "learning_rate": 3.812488758558386e-07,
+      "loss": 0.0032,
+      "step": 28880
+    },
+    {
+      "epoch": 1.6828799440787559,
+      "grad_norm": 0.19910860061645508,
+      "learning_rate": 3.744857378056743e-07,
+      "loss": 0.0021,
+      "step": 28890
+    },
+    {
+      "epoch": 1.6834624570396692,
+      "grad_norm": 0.1611495316028595,
+      "learning_rate": 3.677828994347154e-07,
+      "loss": 0.0028,
+      "step": 28900
+    },
+    {
+      "epoch": 1.6840449700005826,
+      "grad_norm": 0.06089824438095093,
+      "learning_rate": 3.61140368887547e-07,
+      "loss": 0.0031,
+      "step": 28910
+    },
+    {
+      "epoch": 1.684627482961496,
+      "grad_norm": 0.07103503495454788,
+      "learning_rate": 3.545581542354681e-07,
+      "loss": 0.0051,
+      "step": 28920
+    },
+    {
+      "epoch": 1.6852099959224094,
+      "grad_norm": 0.2542659342288971,
+      "learning_rate": 3.480362634764922e-07,
+      "loss": 0.004,
+      "step": 28930
+    },
+    {
+      "epoch": 1.6857925088833228,
+      "grad_norm": 0.18553470075130463,
+      "learning_rate": 3.4157470453533014e-07,
+      "loss": 0.0028,
+      "step": 28940
+    },
+    {
+      "epoch": 1.6863750218442362,
+      "grad_norm": 0.060836032032966614,
+      "learning_rate": 3.3517348526339034e-07,
+      "loss": 0.0021,
+      "step": 28950
+    },
+    {
+      "epoch": 1.6869575348051495,
+      "grad_norm": 0.1184179037809372,
+      "learning_rate": 3.288326134387454e-07,
+      "loss": 0.0022,
+      "step": 28960
+    },
+    {
+      "epoch": 1.687540047766063,
+      "grad_norm": 0.1072288304567337,
+      "learning_rate": 3.225520967661655e-07,
+      "loss": 0.0026,
+      "step": 28970
+    },
+    {
+      "epoch": 1.6881225607269763,
+      "grad_norm": 0.09737038612365723,
+      "learning_rate": 3.163319428770628e-07,
+      "loss": 0.0023,
+      "step": 28980
+    },
+    {
+      "epoch": 1.6887050736878897,
+      "grad_norm": 0.28700464963912964,
+      "learning_rate": 3.1017215932951374e-07,
+      "loss": 0.0047,
+      "step": 28990
+    },
+    {
+      "epoch": 1.689287586648803,
+      "grad_norm": 0.17121894657611847,
+      "learning_rate": 3.040727536082366e-07,
+      "loss": 0.0026,
+      "step": 29000
+    },
+    {
+      "epoch": 1.6898700996097165,
+      "grad_norm": 0.17327986657619476,
+      "learning_rate": 2.980337331245864e-07,
+      "loss": 0.0043,
+      "step": 29010
+    },
+    {
+      "epoch": 1.6904526125706298,
+      "grad_norm": 0.20917299389839172,
+      "learning_rate": 2.9205510521653213e-07,
+      "loss": 0.0027,
+      "step": 29020
+    },
+    {
+      "epoch": 1.6910351255315432,
+      "grad_norm": 0.1458817720413208,
+      "learning_rate": 2.86136877148685e-07,
+      "loss": 0.0027,
+      "step": 29030
+    },
+    {
+      "epoch": 1.6916176384924566,
+      "grad_norm": 0.23862476646900177,
+      "learning_rate": 2.8027905611223704e-07,
+      "loss": 0.0036,
+      "step": 29040
+    },
+    {
+      "epoch": 1.69220015145337,
+      "grad_norm": 0.15557187795639038,
+      "learning_rate": 2.7448164922500573e-07,
+      "loss": 0.0028,
+      "step": 29050
+    },
+    {
+      "epoch": 1.6927826644142834,
+      "grad_norm": 0.0812695175409317,
+      "learning_rate": 2.687446635313784e-07,
+      "loss": 0.005,
+      "step": 29060
+    },
+    {
+      "epoch": 1.6933651773751968,
+      "grad_norm": 0.10532993823289871,
+      "learning_rate": 2.630681060023343e-07,
+      "loss": 0.0021,
+      "step": 29070
+    },
+    {
+      "epoch": 1.69394769033611,
+      "grad_norm": 0.16624745726585388,
+      "learning_rate": 2.5745198353542833e-07,
+      "loss": 0.0017,
+      "step": 29080
+    },
+    {
+      "epoch": 1.6945302032970233,
+      "grad_norm": 0.03405696526169777,
+      "learning_rate": 2.518963029547794e-07,
+      "loss": 0.0056,
+      "step": 29090
+    },
+    {
+      "epoch": 1.6951127162579367,
+      "grad_norm": 0.10715636610984802,
+      "learning_rate": 2.464010710110598e-07,
+      "loss": 0.0025,
+      "step": 29100
+    },
+    {
+      "epoch": 1.69569522921885,
+      "grad_norm": 0.20273993909358978,
+      "learning_rate": 2.4096629438150054e-07,
+      "loss": 0.0057,
+      "step": 29110
+    },
+    {
+      "epoch": 1.6962777421797635,
+      "grad_norm": 0.18573719263076782,
+      "learning_rate": 2.3559197966985802e-07,
+      "loss": 0.0031,
+      "step": 29120
+    },
+    {
+      "epoch": 1.6968602551406768,
+      "grad_norm": 0.2961921691894531,
+      "learning_rate": 2.3027813340644188e-07,
+      "loss": 0.0021,
+      "step": 29130
+    },
+    {
+      "epoch": 1.6974427681015902,
+      "grad_norm": 0.1231430396437645,
+      "learning_rate": 2.2502476204807055e-07,
+      "loss": 0.0028,
+      "step": 29140
+    },
+    {
+      "epoch": 1.6980252810625036,
+      "grad_norm": 0.20614241063594818,
+      "learning_rate": 2.1983187197808786e-07,
+      "loss": 0.0028,
+      "step": 29150
+    },
+    {
+      "epoch": 1.698607794023417,
+      "grad_norm": 0.0999142974615097,
+      "learning_rate": 2.1469946950634644e-07,
+      "loss": 0.0033,
+      "step": 29160
+    },
+    {
+      "epoch": 1.6991903069843304,
+      "grad_norm": 0.0777881070971489,
+      "learning_rate": 2.096275608691911e-07,
+      "loss": 0.0018,
+      "step": 29170
+    },
+    {
+      "epoch": 1.6997728199452438,
+      "grad_norm": 0.246847465634346,
+      "learning_rate": 2.046161522294754e-07,
+      "loss": 0.0034,
+      "step": 29180
+    },
+    {
+      "epoch": 1.7003553329061571,
+      "grad_norm": 0.07733192294836044,
+      "learning_rate": 1.9966524967653944e-07,
+      "loss": 0.0018,
+      "step": 29190
+    },
+    {
+      "epoch": 1.7009378458670705,
+      "grad_norm": 0.17589254677295685,
+      "learning_rate": 1.9477485922618222e-07,
+      "loss": 0.0032,
+      "step": 29200
+    },
+    {
+      "epoch": 1.701520358827984,
+      "grad_norm": 0.19928041100502014,
+      "learning_rate": 1.899449868207004e-07,
+      "loss": 0.0026,
+      "step": 29210
+    },
+    {
+      "epoch": 1.7021028717888973,
+      "grad_norm": 0.05861947685480118,
+      "learning_rate": 1.851756383288439e-07,
+      "loss": 0.004,
+      "step": 29220
+    },
+    {
+      "epoch": 1.7026853847498107,
+      "grad_norm": 0.14034606516361237,
+      "learning_rate": 1.8046681954581035e-07,
+      "loss": 0.0028,
+      "step": 29230
+    },
+    {
+      "epoch": 1.703267897710724,
+      "grad_norm": 0.060287412256002426,
+      "learning_rate": 1.7581853619327294e-07,
+      "loss": 0.0029,
+      "step": 29240
+    },
+    {
+      "epoch": 1.7038504106716374,
+      "grad_norm": 0.15230874717235565,
+      "learning_rate": 1.7123079391932472e-07,
+      "loss": 0.0026,
+      "step": 29250
+    },
+    {
+      "epoch": 1.7044329236325508,
+      "grad_norm": 0.13887517154216766,
+      "learning_rate": 1.6670359829850657e-07,
+      "loss": 0.0038,
+      "step": 29260
+    },
+    {
+      "epoch": 1.7050154365934642,
+      "grad_norm": 0.18385322391986847,
+      "learning_rate": 1.6223695483179037e-07,
+      "loss": 0.0039,
+      "step": 29270
+    },
+    {
+      "epoch": 1.7055979495543776,
+      "grad_norm": 0.18818265199661255,
+      "learning_rate": 1.5783086894656795e-07,
+      "loss": 0.0024,
+      "step": 29280
+    },
+    {
+      "epoch": 1.706180462515291,
+      "grad_norm": 0.13755087554454803,
+      "learning_rate": 1.5348534599665121e-07,
+      "loss": 0.004,
+      "step": 29290
+    },
+    {
+      "epoch": 1.7067629754762044,
+      "grad_norm": 0.14634089171886444,
+      "learning_rate": 1.4920039126225527e-07,
+      "loss": 0.0015,
+      "step": 29300
+    },
+    {
+      "epoch": 1.7073454884371178,
+      "grad_norm": 0.10681162774562836,
+      "learning_rate": 1.449760099500097e-07,
+      "loss": 0.0033,
+      "step": 29310
+    },
+    {
+      "epoch": 1.7079280013980311,
+      "grad_norm": 0.11402872204780579,
+      "learning_rate": 1.4081220719293075e-07,
+      "loss": 0.0024,
+      "step": 29320
+    },
+    {
+      "epoch": 1.7085105143589445,
+      "grad_norm": 0.213548943400383,
+      "learning_rate": 1.3670898805043797e-07,
+      "loss": 0.0035,
+      "step": 29330
+    },
+    {
+      "epoch": 1.709093027319858,
+      "grad_norm": 0.07849355787038803,
+      "learning_rate": 1.326663575083209e-07,
+      "loss": 0.0021,
+      "step": 29340
+    },
+    {
+      "epoch": 1.7096755402807713,
+      "grad_norm": 0.3689243197441101,
+      "learning_rate": 1.2868432047876688e-07,
+      "loss": 0.0049,
+      "step": 29350
+    },
+    {
+      "epoch": 1.7102580532416847,
+      "grad_norm": 0.19747138023376465,
+      "learning_rate": 1.2476288180032213e-07,
+      "loss": 0.0028,
+      "step": 29360
+    },
+    {
+      "epoch": 1.710840566202598,
+      "grad_norm": 0.066779226064682,
+      "learning_rate": 1.209020462379029e-07,
+      "loss": 0.0018,
+      "step": 29370
+    },
+    {
+      "epoch": 1.7114230791635114,
+      "grad_norm": 0.17843309044837952,
+      "learning_rate": 1.1710181848278435e-07,
+      "loss": 0.0034,
+      "step": 29380
+    },
+    {
+      "epoch": 1.7120055921244246,
+      "grad_norm": 0.37307751178741455,
+      "learning_rate": 1.133622031526116e-07,
+      "loss": 0.003,
+      "step": 29390
+    },
+    {
+      "epoch": 1.712588105085338,
+      "grad_norm": 0.2104889154434204,
+      "learning_rate": 1.0968320479136095e-07,
+      "loss": 0.0033,
+      "step": 29400
+    },
+    {
+      "epoch": 1.7131706180462514,
+      "grad_norm": 0.06660198420286179,
+      "learning_rate": 1.0606482786936767e-07,
+      "loss": 0.0031,
+      "step": 29410
+    },
+    {
+      "epoch": 1.7137531310071648,
+      "grad_norm": 0.05133553594350815,
+      "learning_rate": 1.0250707678329808e-07,
+      "loss": 0.0024,
+      "step": 29420
+    },
+    {
+      "epoch": 1.7143356439680781,
+      "grad_norm": 0.14742447435855865,
+      "learning_rate": 9.900995585615525e-08,
+      "loss": 0.0027,
+      "step": 29430
+    },
+    {
+      "epoch": 1.7149181569289915,
+      "grad_norm": 0.1072724238038063,
+      "learning_rate": 9.55734693372734e-08,
+      "loss": 0.003,
+      "step": 29440
+    },
+    {
+      "epoch": 1.715500669889905,
+      "grad_norm": 0.24509476125240326,
+      "learning_rate": 9.219762140231236e-08,
+      "loss": 0.0029,
+      "step": 29450
+    },
+    {
+      "epoch": 1.7160831828508183,
+      "grad_norm": 0.2387397289276123,
+      "learning_rate": 8.888241615322978e-08,
+      "loss": 0.0052,
+      "step": 29460
+    },
+    {
+      "epoch": 1.7166656958117317,
+      "grad_norm": 0.12913013994693756,
+      "learning_rate": 8.562785761833114e-08,
+      "loss": 0.0036,
+      "step": 29470
+    },
+    {
+      "epoch": 1.717248208772645,
+      "grad_norm": 0.3265467882156372,
+      "learning_rate": 8.243394975219753e-08,
+      "loss": 0.0039,
+      "step": 29480
+    },
+    {
+      "epoch": 1.7178307217335584,
+      "grad_norm": 0.163236603140831,
+      "learning_rate": 7.930069643573568e-08,
+      "loss": 0.0025,
+      "step": 29490
+    },
+    {
+      "epoch": 1.7184132346944718,
+      "grad_norm": 0.16275805234909058,
+      "learning_rate": 7.622810147614456e-08,
+      "loss": 0.0023,
+      "step": 29500
+    },
+    {
+      "epoch": 1.7189957476553852,
+      "grad_norm": 0.25081977248191833,
+      "learning_rate": 7.321616860690995e-08,
+      "loss": 0.0041,
+      "step": 29510
+    },
+    {
+      "epoch": 1.7195782606162986,
+      "grad_norm": 0.3237345516681671,
+      "learning_rate": 7.026490148782095e-08,
+      "loss": 0.0041,
+      "step": 29520
+    },
+    {
+      "epoch": 1.720160773577212,
+      "grad_norm": 0.05115390941500664,
+      "learning_rate": 6.737430370494236e-08,
+      "loss": 0.0029,
+      "step": 29530
+    },
+    {
+      "epoch": 1.7207432865381254,
+      "grad_norm": 0.13487602770328522,
+      "learning_rate": 6.454437877062569e-08,
+      "loss": 0.0027,
+      "step": 29540
+    },
+    {
+      "epoch": 1.7213257994990387,
+      "grad_norm": 0.07653814554214478,
+      "learning_rate": 6.177513012349256e-08,
+      "loss": 0.0026,
+      "step": 29550
+    },
+    {
+      "epoch": 1.7219083124599521,
+      "grad_norm": 0.22305920720100403,
+      "learning_rate": 5.9066561128445775e-08,
+      "loss": 0.0037,
+      "step": 29560
+    },
+    {
+      "epoch": 1.7224908254208655,
+      "grad_norm": 0.2577698826789856,
+      "learning_rate": 5.6418675076641556e-08,
+      "loss": 0.003,
+      "step": 29570
+    },
+    {
+      "epoch": 1.723073338381779,
+      "grad_norm": 0.3111644685268402,
+      "learning_rate": 5.383147518552845e-08,
+      "loss": 0.0021,
+      "step": 29580
+    },
+    {
+      "epoch": 1.7236558513426923,
+      "grad_norm": 0.08693022280931473,
+      "learning_rate": 5.1304964598786196e-08,
+      "loss": 0.0049,
+      "step": 29590
+    },
+    {
+      "epoch": 1.7242383643036057,
+      "grad_norm": 0.08577819913625717,
+      "learning_rate": 4.883914638636467e-08,
+      "loss": 0.004,
+      "step": 29600
+    },
+    {
+      "epoch": 1.724820877264519,
+      "grad_norm": 0.08518391847610474,
+      "learning_rate": 4.643402354446713e-08,
+      "loss": 0.0028,
+      "step": 29610
+    },
+    {
+      "epoch": 1.7254033902254324,
+      "grad_norm": 0.17423392832279205,
+      "learning_rate": 4.4089598995544766e-08,
+      "loss": 0.0033,
+      "step": 29620
+    },
+    {
+      "epoch": 1.7259859031863458,
+      "grad_norm": 0.12681856751441956,
+      "learning_rate": 4.180587558829663e-08,
+      "loss": 0.0027,
+      "step": 29630
+    },
+    {
+      "epoch": 1.7265684161472592,
+      "grad_norm": 0.32520073652267456,
+      "learning_rate": 3.958285609765855e-08,
+      "loss": 0.0024,
+      "step": 29640
+    },
+    {
+      "epoch": 1.7271509291081726,
+      "grad_norm": 0.08897735923528671,
+      "learning_rate": 3.74205432248087e-08,
+      "loss": 0.0032,
+      "step": 29650
+    },
+    {
+      "epoch": 1.727733442069086,
+      "grad_norm": 0.04691459611058235,
+      "learning_rate": 3.531893959716204e-08,
+      "loss": 0.0025,
+      "step": 29660
+    },
+    {
+      "epoch": 1.7283159550299994,
+      "grad_norm": 0.15256232023239136,
+      "learning_rate": 3.327804776837029e-08,
+      "loss": 0.0029,
+      "step": 29670
+    },
+    {
+      "epoch": 1.7288984679909127,
+      "grad_norm": 0.26718783378601074,
+      "learning_rate": 3.129787021829977e-08,
+      "loss": 0.0034,
+      "step": 29680
+    },
+    {
+      "epoch": 1.7294809809518261,
+      "grad_norm": 0.09881292283535004,
+      "learning_rate": 2.9378409353059133e-08,
+      "loss": 0.0023,
+      "step": 29690
+    },
+    {
+      "epoch": 1.7300634939127395,
+      "grad_norm": 0.13116058707237244,
+      "learning_rate": 2.7519667504971593e-08,
+      "loss": 0.0023,
+      "step": 29700
+    },
+    {
+      "epoch": 1.7306460068736529,
+      "grad_norm": 0.2654755711555481,
+      "learning_rate": 2.572164693258605e-08,
+      "loss": 0.0032,
+      "step": 29710
+    },
+    {
+      "epoch": 1.7312285198345663,
+      "grad_norm": 0.11338164657354355,
+      "learning_rate": 2.3984349820665996e-08,
+      "loss": 0.0032,
+      "step": 29720
+    },
+    {
+      "epoch": 1.7318110327954797,
+      "grad_norm": 0.0875202938914299,
+      "learning_rate": 2.2307778280189485e-08,
+      "loss": 0.0026,
+      "step": 29730
+    },
+    {
+      "epoch": 1.732393545756393,
+      "grad_norm": 0.15522614121437073,
+      "learning_rate": 2.0691934348354704e-08,
+      "loss": 0.0035,
+      "step": 29740
+    },
+    {
+      "epoch": 1.7329760587173064,
+      "grad_norm": 0.06897246837615967,
+      "learning_rate": 1.9136819988557763e-08,
+      "loss": 0.0029,
+      "step": 29750
+    },
+    {
+      "epoch": 1.7335585716782198,
+      "grad_norm": 0.07973351329565048,
+      "learning_rate": 1.7642437090414908e-08,
+      "loss": 0.0035,
+      "step": 29760
+    },
+    {
+      "epoch": 1.7341410846391332,
+      "grad_norm": 0.196388840675354,
+      "learning_rate": 1.6208787469734755e-08,
+      "loss": 0.0024,
+      "step": 29770
+    },
+    {
+      "epoch": 1.7347235976000466,
+      "grad_norm": 0.21949230134487152,
+      "learning_rate": 1.483587286854604e-08,
+      "loss": 0.0022,
+      "step": 29780
+    },
+    {
+      "epoch": 1.73530611056096,
+      "grad_norm": 0.0486697293817997,
+      "learning_rate": 1.3523694955064336e-08,
+      "loss": 0.0045,
+      "step": 29790
+    },
+    {
+      "epoch": 1.7358886235218733,
+      "grad_norm": 0.20237737894058228,
+      "learning_rate": 1.2272255323708681e-08,
+      "loss": 0.0041,
+      "step": 29800
+    },
+    {
+      "epoch": 1.7364711364827867,
+      "grad_norm": 0.08164387196302414,
+      "learning_rate": 1.1081555495096042e-08,
+      "loss": 0.0033,
+      "step": 29810
+    },
+    {
+      "epoch": 1.7370536494437,
+      "grad_norm": 0.14419779181480408,
+      "learning_rate": 9.951596916041306e-09,
+      "loss": 0.004,
+      "step": 29820
+    },
+    {
+      "epoch": 1.7376361624046135,
+      "grad_norm": 0.07511334121227264,
+      "learning_rate": 8.882380959551739e-09,
+      "loss": 0.0025,
+      "step": 29830
+    },
+    {
+      "epoch": 1.7382186753655269,
+      "grad_norm": 0.06141670048236847,
+      "learning_rate": 7.873908924821428e-09,
+      "loss": 0.0056,
+      "step": 29840
+    },
+    {
+      "epoch": 1.7388011883264403,
+      "grad_norm": 0.0919167771935463,
+      "learning_rate": 6.926182037242379e-09,
+      "loss": 0.0032,
+      "step": 29850
+    },
+    {
+      "epoch": 1.7393837012873536,
+      "grad_norm": 0.20604175329208374,
+      "learning_rate": 6.039201448387876e-09,
+      "loss": 0.003,
+      "step": 29860
+    },
+    {
+      "epoch": 1.739966214248267,
+      "grad_norm": 0.292091965675354,
+      "learning_rate": 5.212968236029125e-09,
+      "loss": 0.0043,
+      "step": 29870
+    },
+    {
+      "epoch": 1.7405487272091804,
+      "grad_norm": 0.05781623348593712,
+      "learning_rate": 4.447483404118602e-09,
+      "loss": 0.0031,
+      "step": 29880
+    },
+    {
+      "epoch": 1.7411312401700938,
+      "grad_norm": 0.08216499537229538,
+      "learning_rate": 3.742747882784503e-09,
+      "loss": 0.0021,
+      "step": 29890
+    },
+    {
+      "epoch": 1.7417137531310072,
+      "grad_norm": 0.12507128715515137,
+      "learning_rate": 3.0987625283473987e-09,
+      "loss": 0.0035,
+      "step": 29900
+    },
+    {
+      "epoch": 1.7422962660919206,
+      "grad_norm": 0.20986425876617432,
+      "learning_rate": 2.515528123320232e-09,
+      "loss": 0.0031,
+      "step": 29910
+    },
+    {
+      "epoch": 1.742878779052834,
+      "grad_norm": 0.1457216441631317,
+      "learning_rate": 1.9930453763750133e-09,
+      "loss": 0.0031,
+      "step": 29920
+    },
+    {
+      "epoch": 1.7434612920137473,
+      "grad_norm": 0.09422629326581955,
+      "learning_rate": 1.5313149223872281e-09,
+      "loss": 0.0034,
+      "step": 29930
+    },
+    {
+      "epoch": 1.7440438049746607,
+      "grad_norm": 0.12003514915704727,
+      "learning_rate": 1.1303373224025305e-09,
+      "loss": 0.0032,
+      "step": 29940
+    },
+    {
+      "epoch": 1.744626317935574,
+      "grad_norm": 0.03193308040499687,
+      "learning_rate": 7.901130636367438e-10,
+      "loss": 0.0022,
+      "step": 29950
+    },
+    {
+      "epoch": 1.7452088308964875,
+      "grad_norm": 0.12078665941953659,
+      "learning_rate": 5.106425595036158e-10,
+      "loss": 0.0038,
+      "step": 29960
+    },
+    {
+      "epoch": 1.7457913438574009,
+      "grad_norm": 0.07153503596782684,
+      "learning_rate": 2.9192614958706286e-10,
+      "loss": 0.0024,
+      "step": 29970
+    },
+    {
+      "epoch": 1.7463738568183143,
+      "grad_norm": 0.18771716952323914,
+      "learning_rate": 1.3396409964117064e-10,
+      "loss": 0.0023,
+      "step": 29980
+    },
+    {
+      "epoch": 1.7469563697792276,
+      "grad_norm": 0.171293243765831,
+      "learning_rate": 3.6756601606846574e-11,
+      "loss": 0.0025,
+      "step": 29990
+    },
+    {
+      "epoch": 1.747538882740141,
+      "grad_norm": 0.2965494692325592,
+      "learning_rate": 3.0377360626943076e-13,
+      "loss": 0.0036,
+      "step": 30000
+    },
+    {
+      "epoch": 1.747538882740141,
+      "step": 30000,
+      "total_flos": 0.0,
+      "train_loss": 0.011098173622476558,
+      "train_runtime": 35657.3089,
+      "train_samples_per_second": 26.923,
+      "train_steps_per_second": 0.841
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 30000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 10000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}