diff --git "a/checkpoint-31253/trainer_state.json" "b/checkpoint-31253/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-31253/trainer_state.json"
@@ -0,0 +1,21909 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 31253,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003199692829488369,
+      "grad_norm": 24.0,
+      "learning_rate": 1.1516314779270635e-07,
+      "loss": 0.5918,
+      "step": 10
+    },
+    {
+      "epoch": 0.0006399385658976738,
+      "grad_norm": 18.875,
+      "learning_rate": 2.431222008957134e-07,
+      "loss": 0.5934,
+      "step": 20
+    },
+    {
+      "epoch": 0.0009599078488465107,
+      "grad_norm": 15.4375,
+      "learning_rate": 3.7108125399872046e-07,
+      "loss": 0.6051,
+      "step": 30
+    },
+    {
+      "epoch": 0.0012798771317953477,
+      "grad_norm": 19.5,
+      "learning_rate": 4.990403071017275e-07,
+      "loss": 0.5723,
+      "step": 40
+    },
+    {
+      "epoch": 0.0015998464147441846,
+      "grad_norm": 11.75,
+      "learning_rate": 6.269993602047346e-07,
+      "loss": 0.6254,
+      "step": 50
+    },
+    {
+      "epoch": 0.0019198156976930215,
+      "grad_norm": 11.0625,
+      "learning_rate": 7.549584133077416e-07,
+      "loss": 0.5764,
+      "step": 60
+    },
+    {
+      "epoch": 0.0022397849806418584,
+      "grad_norm": 7.5,
+      "learning_rate": 8.829174664107486e-07,
+      "loss": 0.584,
+      "step": 70
+    },
+    {
+      "epoch": 0.0025597542635906953,
+      "grad_norm": 15.25,
+      "learning_rate": 1.0108765195137557e-06,
+      "loss": 0.5797,
+      "step": 80
+    },
+    {
+      "epoch": 0.002879723546539532,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.1388355726167627e-06,
+      "loss": 0.6012,
+      "step": 90
+    },
+    {
+      "epoch": 0.003199692829488369,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2667946257197696e-06,
+      "loss": 0.5617,
+      "step": 100
+    },
+    {
+      "epoch": 0.003519662112437206,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.394753678822777e-06,
+      "loss": 0.584,
+      "step": 110
+    },
+    {
+      "epoch": 0.003839631395386043,
+      "grad_norm": 18.0,
+      "learning_rate": 1.5227127319257839e-06,
+      "loss": 0.6004,
+      "step": 120
+    },
+    {
+      "epoch": 0.00415960067833488,
+      "grad_norm": 18.25,
+      "learning_rate": 1.650671785028791e-06,
+      "loss": 0.5977,
+      "step": 130
+    },
+    {
+      "epoch": 0.004479569961283717,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.778630838131798e-06,
+      "loss": 0.5578,
+      "step": 140
+    },
+    {
+      "epoch": 0.004799539244232554,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.906589891234805e-06,
+      "loss": 0.5926,
+      "step": 150
+    },
+    {
+      "epoch": 0.005119508527181391,
+      "grad_norm": 11.0,
+      "learning_rate": 2.0345489443378122e-06,
+      "loss": 0.6055,
+      "step": 160
+    },
+    {
+      "epoch": 0.0054394778101302275,
+      "grad_norm": 11.1875,
+      "learning_rate": 2.1625079974408194e-06,
+      "loss": 0.5736,
+      "step": 170
+    },
+    {
+      "epoch": 0.005759447093079064,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.290467050543826e-06,
+      "loss": 0.5375,
+      "step": 180
+    },
+    {
+      "epoch": 0.006079416376027901,
+      "grad_norm": 15.375,
+      "learning_rate": 2.4184261036468333e-06,
+      "loss": 0.6051,
+      "step": 190
+    },
+    {
+      "epoch": 0.006399385658976738,
+      "grad_norm": 10.375,
+      "learning_rate": 2.5463851567498404e-06,
+      "loss": 0.5988,
+      "step": 200
+    },
+    {
+      "epoch": 0.006719354941925575,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.674344209852847e-06,
+      "loss": 0.5555,
+      "step": 210
+    },
+    {
+      "epoch": 0.007039324224874412,
+      "grad_norm": 12.1875,
+      "learning_rate": 2.8023032629558543e-06,
+      "loss": 0.5713,
+      "step": 220
+    },
+    {
+      "epoch": 0.007359293507823249,
+      "grad_norm": 17.25,
+      "learning_rate": 2.930262316058861e-06,
+      "loss": 0.5605,
+      "step": 230
+    },
+    {
+      "epoch": 0.007679262790772086,
+      "grad_norm": 10.375,
+      "learning_rate": 3.0582213691618685e-06,
+      "loss": 0.565,
+      "step": 240
+    },
+    {
+      "epoch": 0.007999232073720922,
+      "grad_norm": 9.25,
+      "learning_rate": 3.1861804222648757e-06,
+      "loss": 0.5705,
+      "step": 250
+    },
+    {
+      "epoch": 0.00831920135666976,
+      "grad_norm": 25.125,
+      "learning_rate": 3.3141394753678824e-06,
+      "loss": 0.5398,
+      "step": 260
+    },
+    {
+      "epoch": 0.008639170639618596,
+      "grad_norm": 14.25,
+      "learning_rate": 3.4420985284708895e-06,
+      "loss": 0.5424,
+      "step": 270
+    },
+    {
+      "epoch": 0.008959139922567434,
+      "grad_norm": 24.0,
+      "learning_rate": 3.5700575815738963e-06,
+      "loss": 0.5867,
+      "step": 280
+    },
+    {
+      "epoch": 0.00927910920551627,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.698016634676904e-06,
+      "loss": 0.5211,
+      "step": 290
+    },
+    {
+      "epoch": 0.009599078488465107,
+      "grad_norm": 8.375,
+      "learning_rate": 3.8259756877799105e-06,
+      "loss": 0.5086,
+      "step": 300
+    },
+    {
+      "epoch": 0.009919047771413943,
+      "grad_norm": 13.125,
+      "learning_rate": 3.953934740882917e-06,
+      "loss": 0.5596,
+      "step": 310
+    },
+    {
+      "epoch": 0.010239017054362781,
+      "grad_norm": 6.875,
+      "learning_rate": 4.081893793985925e-06,
+      "loss": 0.565,
+      "step": 320
+    },
+    {
+      "epoch": 0.010558986337311617,
+      "grad_norm": 7.25,
+      "learning_rate": 4.209852847088932e-06,
+      "loss": 0.5412,
+      "step": 330
+    },
+    {
+      "epoch": 0.010878955620260455,
+      "grad_norm": 13.125,
+      "learning_rate": 4.337811900191939e-06,
+      "loss": 0.5969,
+      "step": 340
+    },
+    {
+      "epoch": 0.011198924903209291,
+      "grad_norm": 10.375,
+      "learning_rate": 4.465770953294946e-06,
+      "loss": 0.5518,
+      "step": 350
+    },
+    {
+      "epoch": 0.011518894186158129,
+      "grad_norm": 11.625,
+      "learning_rate": 4.5937300063979526e-06,
+      "loss": 0.5232,
+      "step": 360
+    },
+    {
+      "epoch": 0.011838863469106965,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.72168905950096e-06,
+      "loss": 0.5139,
+      "step": 370
+    },
+    {
+      "epoch": 0.012158832752055803,
+      "grad_norm": 11.625,
+      "learning_rate": 4.849648112603968e-06,
+      "loss": 0.5734,
+      "step": 380
+    },
+    {
+      "epoch": 0.012478802035004639,
+      "grad_norm": 9.25,
+      "learning_rate": 4.977607165706974e-06,
+      "loss": 0.4797,
+      "step": 390
+    },
+    {
+      "epoch": 0.012798771317953477,
+      "grad_norm": 8.75,
+      "learning_rate": 5.105566218809981e-06,
+      "loss": 0.4947,
+      "step": 400
+    },
+    {
+      "epoch": 0.013118740600902313,
+      "grad_norm": 17.25,
+      "learning_rate": 5.233525271912988e-06,
+      "loss": 0.5643,
+      "step": 410
+    },
+    {
+      "epoch": 0.01343870988385115,
+      "grad_norm": 15.625,
+      "learning_rate": 5.361484325015995e-06,
+      "loss": 0.5088,
+      "step": 420
+    },
+    {
+      "epoch": 0.013758679166799986,
+      "grad_norm": 10.5625,
+      "learning_rate": 5.489443378119003e-06,
+      "loss": 0.5064,
+      "step": 430
+    },
+    {
+      "epoch": 0.014078648449748824,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.61740243122201e-06,
+      "loss": 0.5547,
+      "step": 440
+    },
+    {
+      "epoch": 0.01439861773269766,
+      "grad_norm": 6.75,
+      "learning_rate": 5.745361484325016e-06,
+      "loss": 0.4967,
+      "step": 450
+    },
+    {
+      "epoch": 0.014718587015646498,
+      "grad_norm": 18.125,
+      "learning_rate": 5.873320537428023e-06,
+      "loss": 0.5059,
+      "step": 460
+    },
+    {
+      "epoch": 0.015038556298595334,
+      "grad_norm": 16.0,
+      "learning_rate": 6.001279590531031e-06,
+      "loss": 0.5262,
+      "step": 470
+    },
+    {
+      "epoch": 0.015358525581544172,
+      "grad_norm": 10.25,
+      "learning_rate": 6.129238643634038e-06,
+      "loss": 0.5311,
+      "step": 480
+    },
+    {
+      "epoch": 0.01567849486449301,
+      "grad_norm": 14.3125,
+      "learning_rate": 6.257197696737045e-06,
+      "loss": 0.5551,
+      "step": 490
+    },
+    {
+      "epoch": 0.015998464147441844,
+      "grad_norm": 12.6875,
+      "learning_rate": 6.385156749840052e-06,
+      "loss": 0.5479,
+      "step": 500
+    },
+    {
+      "epoch": 0.01631843343039068,
+      "grad_norm": 20.25,
+      "learning_rate": 6.513115802943058e-06,
+      "loss": 0.5518,
+      "step": 510
+    },
+    {
+      "epoch": 0.01663840271333952,
+      "grad_norm": 8.0,
+      "learning_rate": 6.641074856046066e-06,
+      "loss": 0.565,
+      "step": 520
+    },
+    {
+      "epoch": 0.016958371996288357,
+      "grad_norm": 19.125,
+      "learning_rate": 6.769033909149073e-06,
+      "loss": 0.5084,
+      "step": 530
+    },
+    {
+      "epoch": 0.01727834127923719,
+      "grad_norm": 9.5,
+      "learning_rate": 6.89699296225208e-06,
+      "loss": 0.5621,
+      "step": 540
+    },
+    {
+      "epoch": 0.01759831056218603,
+      "grad_norm": 12.5625,
+      "learning_rate": 7.024952015355086e-06,
+      "loss": 0.527,
+      "step": 550
+    },
+    {
+      "epoch": 0.017918279845134867,
+      "grad_norm": 11.875,
+      "learning_rate": 7.152911068458094e-06,
+      "loss": 0.5236,
+      "step": 560
+    },
+    {
+      "epoch": 0.018238249128083705,
+      "grad_norm": 9.8125,
+      "learning_rate": 7.280870121561101e-06,
+      "loss": 0.5033,
+      "step": 570
+    },
+    {
+      "epoch": 0.01855821841103254,
+      "grad_norm": 19.125,
+      "learning_rate": 7.408829174664108e-06,
+      "loss": 0.5205,
+      "step": 580
+    },
+    {
+      "epoch": 0.018878187693981377,
+      "grad_norm": 14.875,
+      "learning_rate": 7.5367882277671156e-06,
+      "loss": 0.583,
+      "step": 590
+    },
+    {
+      "epoch": 0.019198156976930215,
+      "grad_norm": 11.3125,
+      "learning_rate": 7.664747280870121e-06,
+      "loss": 0.4984,
+      "step": 600
+    },
+    {
+      "epoch": 0.019518126259879053,
+      "grad_norm": 17.875,
+      "learning_rate": 7.79270633397313e-06,
+      "loss": 0.4883,
+      "step": 610
+    },
+    {
+      "epoch": 0.019838095542827887,
+      "grad_norm": 26.75,
+      "learning_rate": 7.920665387076137e-06,
+      "loss": 0.534,
+      "step": 620
+    },
+    {
+      "epoch": 0.020158064825776725,
+      "grad_norm": 18.125,
+      "learning_rate": 8.048624440179143e-06,
+      "loss": 0.4973,
+      "step": 630
+    },
+    {
+      "epoch": 0.020478034108725562,
+      "grad_norm": 17.5,
+      "learning_rate": 8.176583493282152e-06,
+      "loss": 0.4695,
+      "step": 640
+    },
+    {
+      "epoch": 0.0207980033916744,
+      "grad_norm": 13.5625,
+      "learning_rate": 8.304542546385157e-06,
+      "loss": 0.5316,
+      "step": 650
+    },
+    {
+      "epoch": 0.021117972674623234,
+      "grad_norm": 44.0,
+      "learning_rate": 8.432501599488163e-06,
+      "loss": 0.4807,
+      "step": 660
+    },
+    {
+      "epoch": 0.021437941957572072,
+      "grad_norm": 41.5,
+      "learning_rate": 8.560460652591172e-06,
+      "loss": 0.4648,
+      "step": 670
+    },
+    {
+      "epoch": 0.02175791124052091,
+      "grad_norm": 20.375,
+      "learning_rate": 8.688419705694179e-06,
+      "loss": 0.4713,
+      "step": 680
+    },
+    {
+      "epoch": 0.022077880523469748,
+      "grad_norm": 46.75,
+      "learning_rate": 8.816378758797185e-06,
+      "loss": 0.3724,
+      "step": 690
+    },
+    {
+      "epoch": 0.022397849806418582,
+      "grad_norm": 26.875,
+      "learning_rate": 8.944337811900192e-06,
+      "loss": 0.449,
+      "step": 700
+    },
+    {
+      "epoch": 0.02271781908936742,
+      "grad_norm": 47.75,
+      "learning_rate": 9.072296865003199e-06,
+      "loss": 0.4127,
+      "step": 710
+    },
+    {
+      "epoch": 0.023037788372316258,
+      "grad_norm": 27.375,
+      "learning_rate": 9.200255918106207e-06,
+      "loss": 0.4557,
+      "step": 720
+    },
+    {
+      "epoch": 0.023357757655265095,
+      "grad_norm": 15.1875,
+      "learning_rate": 9.328214971209214e-06,
+      "loss": 0.4498,
+      "step": 730
+    },
+    {
+      "epoch": 0.02367772693821393,
+      "grad_norm": 39.0,
+      "learning_rate": 9.45617402431222e-06,
+      "loss": 0.4418,
+      "step": 740
+    },
+    {
+      "epoch": 0.023997696221162768,
+      "grad_norm": 33.75,
+      "learning_rate": 9.584133077415227e-06,
+      "loss": 0.4421,
+      "step": 750
+    },
+    {
+      "epoch": 0.024317665504111605,
+      "grad_norm": 28.125,
+      "learning_rate": 9.712092130518234e-06,
+      "loss": 0.4023,
+      "step": 760
+    },
+    {
+      "epoch": 0.024637634787060443,
+      "grad_norm": 14.5625,
+      "learning_rate": 9.840051183621242e-06,
+      "loss": 0.3801,
+      "step": 770
+    },
+    {
+      "epoch": 0.024957604070009277,
+      "grad_norm": 35.0,
+      "learning_rate": 9.968010236724249e-06,
+      "loss": 0.4164,
+      "step": 780
+    },
+    {
+      "epoch": 0.025277573352958115,
+      "grad_norm": 24.125,
+      "learning_rate": 1.0095969289827256e-05,
+      "loss": 0.4189,
+      "step": 790
+    },
+    {
+      "epoch": 0.025597542635906953,
+      "grad_norm": 35.0,
+      "learning_rate": 1.0223928342930263e-05,
+      "loss": 0.3312,
+      "step": 800
+    },
+    {
+      "epoch": 0.02591751191885579,
+      "grad_norm": 87.5,
+      "learning_rate": 1.0351887396033271e-05,
+      "loss": 0.3484,
+      "step": 810
+    },
+    {
+      "epoch": 0.026237481201804625,
+      "grad_norm": 31.75,
+      "learning_rate": 1.0479846449136278e-05,
+      "loss": 0.2955,
+      "step": 820
+    },
+    {
+      "epoch": 0.026557450484753463,
+      "grad_norm": 34.0,
+      "learning_rate": 1.0607805502239283e-05,
+      "loss": 0.238,
+      "step": 830
+    },
+    {
+      "epoch": 0.0268774197677023,
+      "grad_norm": 61.75,
+      "learning_rate": 1.0735764555342291e-05,
+      "loss": 0.2479,
+      "step": 840
+    },
+    {
+      "epoch": 0.02719738905065114,
+      "grad_norm": 37.25,
+      "learning_rate": 1.0863723608445298e-05,
+      "loss": 0.2883,
+      "step": 850
+    },
+    {
+      "epoch": 0.027517358333599973,
+      "grad_norm": 68.0,
+      "learning_rate": 1.0991682661548306e-05,
+      "loss": 0.3091,
+      "step": 860
+    },
+    {
+      "epoch": 0.02783732761654881,
+      "grad_norm": 54.25,
+      "learning_rate": 1.1119641714651313e-05,
+      "loss": 0.185,
+      "step": 870
+    },
+    {
+      "epoch": 0.02815729689949765,
+      "grad_norm": 24.125,
+      "learning_rate": 1.1247600767754318e-05,
+      "loss": 0.2128,
+      "step": 880
+    },
+    {
+      "epoch": 0.028477266182446486,
+      "grad_norm": 17.25,
+      "learning_rate": 1.1375559820857326e-05,
+      "loss": 0.2112,
+      "step": 890
+    },
+    {
+      "epoch": 0.02879723546539532,
+      "grad_norm": 67.5,
+      "learning_rate": 1.1503518873960333e-05,
+      "loss": 0.1948,
+      "step": 900
+    },
+    {
+      "epoch": 0.029117204748344158,
+      "grad_norm": 39.75,
+      "learning_rate": 1.1631477927063342e-05,
+      "loss": 0.1911,
+      "step": 910
+    },
+    {
+      "epoch": 0.029437174031292996,
+      "grad_norm": 35.25,
+      "learning_rate": 1.1759436980166348e-05,
+      "loss": 0.2865,
+      "step": 920
+    },
+    {
+      "epoch": 0.029757143314241834,
+      "grad_norm": 41.75,
+      "learning_rate": 1.1887396033269353e-05,
+      "loss": 0.188,
+      "step": 930
+    },
+    {
+      "epoch": 0.030077112597190668,
+      "grad_norm": 43.25,
+      "learning_rate": 1.2015355086372362e-05,
+      "loss": 0.1514,
+      "step": 940
+    },
+    {
+      "epoch": 0.030397081880139506,
+      "grad_norm": 62.75,
+      "learning_rate": 1.2143314139475368e-05,
+      "loss": 0.2148,
+      "step": 950
+    },
+    {
+      "epoch": 0.030717051163088344,
+      "grad_norm": 15.125,
+      "learning_rate": 1.2271273192578377e-05,
+      "loss": 0.2088,
+      "step": 960
+    },
+    {
+      "epoch": 0.03103702044603718,
+      "grad_norm": 23.125,
+      "learning_rate": 1.2399232245681384e-05,
+      "loss": 0.1531,
+      "step": 970
+    },
+    {
+      "epoch": 0.03135698972898602,
+      "grad_norm": 10.875,
+      "learning_rate": 1.2527191298784389e-05,
+      "loss": 0.2149,
+      "step": 980
+    },
+    {
+      "epoch": 0.031676959011934853,
+      "grad_norm": 26.0,
+      "learning_rate": 1.2655150351887397e-05,
+      "loss": 0.1947,
+      "step": 990
+    },
+    {
+      "epoch": 0.03199692829488369,
+      "grad_norm": 23.875,
+      "learning_rate": 1.2783109404990404e-05,
+      "loss": 0.2217,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03231689757783253,
+      "grad_norm": 23.875,
+      "learning_rate": 1.2911068458093412e-05,
+      "loss": 0.1784,
+      "step": 1010
+    },
+    {
+      "epoch": 0.03263686686078136,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.3039027511196419e-05,
+      "loss": 0.1642,
+      "step": 1020
+    },
+    {
+      "epoch": 0.032956836143730205,
+      "grad_norm": 177.0,
+      "learning_rate": 1.3166986564299424e-05,
+      "loss": 0.222,
+      "step": 1030
+    },
+    {
+      "epoch": 0.03327680542667904,
+      "grad_norm": 31.25,
+      "learning_rate": 1.3294945617402432e-05,
+      "loss": 0.1771,
+      "step": 1040
+    },
+    {
+      "epoch": 0.03359677470962787,
+      "grad_norm": 23.875,
+      "learning_rate": 1.3422904670505439e-05,
+      "loss": 0.1787,
+      "step": 1050
+    },
+    {
+      "epoch": 0.033916743992576714,
+      "grad_norm": 53.25,
+      "learning_rate": 1.3550863723608447e-05,
+      "loss": 0.2089,
+      "step": 1060
+    },
+    {
+      "epoch": 0.03423671327552555,
+      "grad_norm": 23.5,
+      "learning_rate": 1.3678822776711454e-05,
+      "loss": 0.2751,
+      "step": 1070
+    },
+    {
+      "epoch": 0.03455668255847438,
+      "grad_norm": 8.875,
+      "learning_rate": 1.380678182981446e-05,
+      "loss": 0.2051,
+      "step": 1080
+    },
+    {
+      "epoch": 0.034876651841423224,
+      "grad_norm": 32.75,
+      "learning_rate": 1.3934740882917468e-05,
+      "loss": 0.1436,
+      "step": 1090
+    },
+    {
+      "epoch": 0.03519662112437206,
+      "grad_norm": 41.25,
+      "learning_rate": 1.4062699936020474e-05,
+      "loss": 0.2113,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0355165904073209,
+      "grad_norm": 11.0,
+      "learning_rate": 1.4190658989123483e-05,
+      "loss": 0.2181,
+      "step": 1110
+    },
+    {
+      "epoch": 0.035836559690269734,
+      "grad_norm": 15.25,
+      "learning_rate": 1.431861804222649e-05,
+      "loss": 0.1552,
+      "step": 1120
+    },
+    {
+      "epoch": 0.03615652897321857,
+      "grad_norm": 9.125,
+      "learning_rate": 1.4446577095329494e-05,
+      "loss": 0.1851,
+      "step": 1130
+    },
+    {
+      "epoch": 0.03647649825616741,
+      "grad_norm": 16.375,
+      "learning_rate": 1.4574536148432503e-05,
+      "loss": 0.1531,
+      "step": 1140
+    },
+    {
+      "epoch": 0.036796467539116244,
+      "grad_norm": 10.875,
+      "learning_rate": 1.470249520153551e-05,
+      "loss": 0.1532,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03711643682206508,
+      "grad_norm": 4.6875,
+      "learning_rate": 1.4830454254638518e-05,
+      "loss": 0.2477,
+      "step": 1160
+    },
+    {
+      "epoch": 0.03743640610501392,
+      "grad_norm": 27.0,
+      "learning_rate": 1.4958413307741525e-05,
+      "loss": 0.1875,
+      "step": 1170
+    },
+    {
+      "epoch": 0.037756375387962754,
+      "grad_norm": 14.75,
+      "learning_rate": 1.508637236084453e-05,
+      "loss": 0.1873,
+      "step": 1180
+    },
+    {
+      "epoch": 0.038076344670911595,
+      "grad_norm": 26.625,
+      "learning_rate": 1.5214331413947538e-05,
+      "loss": 0.2008,
+      "step": 1190
+    },
+    {
+      "epoch": 0.03839631395386043,
+      "grad_norm": 43.5,
+      "learning_rate": 1.5342290467050545e-05,
+      "loss": 0.1827,
+      "step": 1200
+    },
+    {
+      "epoch": 0.038716283236809264,
+      "grad_norm": 56.25,
+      "learning_rate": 1.547024952015355e-05,
+      "loss": 0.2334,
+      "step": 1210
+    },
+    {
+      "epoch": 0.039036252519758105,
+      "grad_norm": 14.375,
+      "learning_rate": 1.559820857325656e-05,
+      "loss": 0.1578,
+      "step": 1220
+    },
+    {
+      "epoch": 0.03935622180270694,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.5726167626359565e-05,
+      "loss": 0.2213,
+      "step": 1230
+    },
+    {
+      "epoch": 0.039676191085655774,
+      "grad_norm": 23.25,
+      "learning_rate": 1.5854126679462575e-05,
+      "loss": 0.1362,
+      "step": 1240
+    },
+    {
+      "epoch": 0.039996160368604615,
+      "grad_norm": 17.625,
+      "learning_rate": 1.598208573256558e-05,
+      "loss": 0.1613,
+      "step": 1250
+    },
+    {
+      "epoch": 0.04031612965155345,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.611004478566859e-05,
+      "loss": 0.1239,
+      "step": 1260
+    },
+    {
+      "epoch": 0.04063609893450229,
+      "grad_norm": 15.625,
+      "learning_rate": 1.6238003838771595e-05,
+      "loss": 0.1669,
+      "step": 1270
+    },
+    {
+      "epoch": 0.040956068217451125,
+      "grad_norm": 27.75,
+      "learning_rate": 1.6365962891874602e-05,
+      "loss": 0.1449,
+      "step": 1280
+    },
+    {
+      "epoch": 0.04127603750039996,
+      "grad_norm": 35.0,
+      "learning_rate": 1.649392194497761e-05,
+      "loss": 0.1824,
+      "step": 1290
+    },
+    {
+      "epoch": 0.0415960067833488,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.6621880998080615e-05,
+      "loss": 0.2166,
+      "step": 1300
+    },
+    {
+      "epoch": 0.041915976066297635,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.6749840051183622e-05,
+      "loss": 0.1918,
+      "step": 1310
+    },
+    {
+      "epoch": 0.04223594534924647,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.687779910428663e-05,
+      "loss": 0.2087,
+      "step": 1320
+    },
+    {
+      "epoch": 0.04255591463219531,
+      "grad_norm": 17.625,
+      "learning_rate": 1.7005758157389636e-05,
+      "loss": 0.1757,
+      "step": 1330
+    },
+    {
+      "epoch": 0.042875883915144145,
+      "grad_norm": 12.75,
+      "learning_rate": 1.7133717210492646e-05,
+      "loss": 0.2006,
+      "step": 1340
+    },
+    {
+      "epoch": 0.043195853198092986,
+      "grad_norm": 75.0,
+      "learning_rate": 1.726167626359565e-05,
+      "loss": 0.1555,
+      "step": 1350
+    },
+    {
+      "epoch": 0.04351582248104182,
+      "grad_norm": 25.875,
+      "learning_rate": 1.738963531669866e-05,
+      "loss": 0.2205,
+      "step": 1360
+    },
+    {
+      "epoch": 0.043835791763990654,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.7517594369801666e-05,
+      "loss": 0.1482,
+      "step": 1370
+    },
+    {
+      "epoch": 0.044155761046939496,
+      "grad_norm": 4.875,
+      "learning_rate": 1.7645553422904673e-05,
+      "loss": 0.1595,
+      "step": 1380
+    },
+    {
+      "epoch": 0.04447573032988833,
+      "grad_norm": 20.125,
+      "learning_rate": 1.777351247600768e-05,
+      "loss": 0.2334,
+      "step": 1390
+    },
+    {
+      "epoch": 0.044795699612837164,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.7901471529110686e-05,
+      "loss": 0.1588,
+      "step": 1400
+    },
+    {
+      "epoch": 0.045115668895786006,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.8029430582213693e-05,
+      "loss": 0.174,
+      "step": 1410
+    },
+    {
+      "epoch": 0.04543563817873484,
+      "grad_norm": 35.25,
+      "learning_rate": 1.81573896353167e-05,
+      "loss": 0.1197,
+      "step": 1420
+    },
+    {
+      "epoch": 0.04575560746168368,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8285348688419706e-05,
+      "loss": 0.0996,
+      "step": 1430
+    },
+    {
+      "epoch": 0.046075576744632515,
+      "grad_norm": 12.875,
+      "learning_rate": 1.8413307741522716e-05,
+      "loss": 0.2596,
+      "step": 1440
+    },
+    {
+      "epoch": 0.04639554602758135,
+      "grad_norm": 53.5,
+      "learning_rate": 1.854126679462572e-05,
+      "loss": 0.1828,
+      "step": 1450
+    },
+    {
+      "epoch": 0.04671551531053019,
+      "grad_norm": 9.625,
+      "learning_rate": 1.866922584772873e-05,
+      "loss": 0.2171,
+      "step": 1460
+    },
+    {
+      "epoch": 0.047035484593479025,
+      "grad_norm": 7.375,
+      "learning_rate": 1.8797184900831736e-05,
+      "loss": 0.1726,
+      "step": 1470
+    },
+    {
+      "epoch": 0.04735545387642786,
+      "grad_norm": 10.375,
+      "learning_rate": 1.892514395393474e-05,
+      "loss": 0.1681,
+      "step": 1480
+    },
+    {
+      "epoch": 0.0476754231593767,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.905310300703775e-05,
+      "loss": 0.1735,
+      "step": 1490
+    },
+    {
+      "epoch": 0.047995392442325535,
+      "grad_norm": 16.25,
+      "learning_rate": 1.9181062060140757e-05,
+      "loss": 0.136,
+      "step": 1500
+    },
+    {
+      "epoch": 0.048315361725274376,
+      "grad_norm": 5.84375,
+      "learning_rate": 1.9309021113243763e-05,
+      "loss": 0.1463,
+      "step": 1510
+    },
+    {
+      "epoch": 0.04863533100822321,
+      "grad_norm": 12.5,
+      "learning_rate": 1.943698016634677e-05,
+      "loss": 0.1917,
+      "step": 1520
+    },
+    {
+      "epoch": 0.048955300291172045,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.9564939219449777e-05,
+      "loss": 0.1536,
+      "step": 1530
+    },
+    {
+      "epoch": 0.049275269574120886,
+      "grad_norm": 19.25,
+      "learning_rate": 1.9692898272552783e-05,
+      "loss": 0.1839,
+      "step": 1540
+    },
+    {
+      "epoch": 0.04959523885706972,
+      "grad_norm": 11.25,
+      "learning_rate": 1.982085732565579e-05,
+      "loss": 0.1862,
+      "step": 1550
+    },
+    {
+      "epoch": 0.049915208140018555,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.99488163787588e-05,
+      "loss": 0.1564,
+      "step": 1560
+    },
+    {
+      "epoch": 0.050235177422967396,
+      "grad_norm": 11.75,
+      "learning_rate": 1.9995958235095995e-05,
+      "loss": 0.1404,
+      "step": 1570
+    },
+    {
+      "epoch": 0.05055514670591623,
+      "grad_norm": 22.0,
+      "learning_rate": 1.998922196025598e-05,
+      "loss": 0.127,
+      "step": 1580
+    },
+    {
+      "epoch": 0.05087511598886507,
+      "grad_norm": 26.875,
+      "learning_rate": 1.9982485685415967e-05,
+      "loss": 0.1784,
+      "step": 1590
+    },
+    {
+      "epoch": 0.051195085271813906,
+      "grad_norm": 28.375,
+      "learning_rate": 1.997574941057595e-05,
+      "loss": 0.1716,
+      "step": 1600
+    },
+    {
+      "epoch": 0.05151505455476274,
+      "grad_norm": 26.0,
+      "learning_rate": 1.996901313573594e-05,
+      "loss": 0.2349,
+      "step": 1610
+    },
+    {
+      "epoch": 0.05183502383771158,
+      "grad_norm": 15.375,
+      "learning_rate": 1.9962276860895927e-05,
+      "loss": 0.1424,
+      "step": 1620
+    },
+    {
+      "epoch": 0.052154993120660416,
+      "grad_norm": 16.625,
+      "learning_rate": 1.995554058605591e-05,
+      "loss": 0.1324,
+      "step": 1630
+    },
+    {
+      "epoch": 0.05247496240360925,
+      "grad_norm": 5.5,
+      "learning_rate": 1.99488043112159e-05,
+      "loss": 0.131,
+      "step": 1640
+    },
+    {
+      "epoch": 0.05279493168655809,
+      "grad_norm": 6.25,
+      "learning_rate": 1.9942068036375887e-05,
+      "loss": 0.1536,
+      "step": 1650
+    },
+    {
+      "epoch": 0.053114900969506926,
+      "grad_norm": 27.875,
+      "learning_rate": 1.993533176153587e-05,
+      "loss": 0.193,
+      "step": 1660
+    },
+    {
+      "epoch": 0.05343487025245577,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.992859548669586e-05,
+      "loss": 0.162,
+      "step": 1670
+    },
+    {
+      "epoch": 0.0537548395354046,
+      "grad_norm": 9.0,
+      "learning_rate": 1.9921859211855846e-05,
+      "loss": 0.1653,
+      "step": 1680
+    },
+    {
+      "epoch": 0.054074808818353436,
+      "grad_norm": 20.875,
+      "learning_rate": 1.991512293701583e-05,
+      "loss": 0.1217,
+      "step": 1690
+    },
+    {
+      "epoch": 0.05439477810130228,
+      "grad_norm": 26.375,
+      "learning_rate": 1.990838666217582e-05,
+      "loss": 0.1305,
+      "step": 1700
+    },
+    {
+      "epoch": 0.05471474738425111,
+      "grad_norm": 21.125,
+      "learning_rate": 1.9901650387335806e-05,
+      "loss": 0.1377,
+      "step": 1710
+    },
+    {
+      "epoch": 0.055034716667199945,
+      "grad_norm": 58.0,
+      "learning_rate": 1.989491411249579e-05,
+      "loss": 0.1531,
+      "step": 1720
+    },
+    {
+      "epoch": 0.05535468595014879,
+      "grad_norm": 14.25,
+      "learning_rate": 1.988817783765578e-05,
+      "loss": 0.1842,
+      "step": 1730
+    },
+    {
+      "epoch": 0.05567465523309762,
+      "grad_norm": 30.0,
+      "learning_rate": 1.9881441562815763e-05,
+      "loss": 0.1714,
+      "step": 1740
+    },
+    {
+      "epoch": 0.05599462451604646,
+      "grad_norm": 39.75,
+      "learning_rate": 1.987470528797575e-05,
+      "loss": 0.1354,
+      "step": 1750
+    },
+    {
+      "epoch": 0.0563145937989953,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.9867969013135738e-05,
+      "loss": 0.1808,
+      "step": 1760
+    },
+    {
+      "epoch": 0.05663456308194413,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.9861232738295723e-05,
+      "loss": 0.1694,
+      "step": 1770
+    },
+    {
+      "epoch": 0.05695453236489297,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.985449646345571e-05,
+      "loss": 0.098,
+      "step": 1780
+    },
+    {
+      "epoch": 0.057274501647841806,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.9847760188615698e-05,
+      "loss": 0.0779,
+      "step": 1790
+    },
+    {
+      "epoch": 0.05759447093079064,
+      "grad_norm": 10.5625,
+      "learning_rate": 1.9841023913775682e-05,
+      "loss": 0.1768,
+      "step": 1800
+    },
+    {
+      "epoch": 0.05791444021373948,
+      "grad_norm": 18.875,
+      "learning_rate": 1.983428763893567e-05,
+      "loss": 0.157,
+      "step": 1810
+    },
+    {
+      "epoch": 0.058234409496688316,
+      "grad_norm": 18.0,
+      "learning_rate": 1.9827551364095658e-05,
+      "loss": 0.1839,
+      "step": 1820
+    },
+    {
+      "epoch": 0.05855437877963716,
+      "grad_norm": 23.875,
+      "learning_rate": 1.9820815089255642e-05,
+      "loss": 0.1748,
+      "step": 1830
+    },
+    {
+      "epoch": 0.05887434806258599,
+      "grad_norm": 15.0,
+      "learning_rate": 1.981407881441563e-05,
+      "loss": 0.1252,
+      "step": 1840
+    },
+    {
+      "epoch": 0.059194317345534826,
+      "grad_norm": 5.8125,
+      "learning_rate": 1.9807342539575614e-05,
+      "loss": 0.173,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05951428662848367,
+      "grad_norm": 32.25,
+      "learning_rate": 1.9800606264735602e-05,
+      "loss": 0.1857,
+      "step": 1860
+    },
+    {
+      "epoch": 0.0598342559114325,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.979386998989559e-05,
+      "loss": 0.1365,
+      "step": 1870
+    },
+    {
+      "epoch": 0.060154225194381336,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.9787133715055574e-05,
+      "loss": 0.176,
+      "step": 1880
+    },
+    {
+      "epoch": 0.06047419447733018,
+      "grad_norm": 6.125,
+      "learning_rate": 1.9780397440215562e-05,
+      "loss": 0.1885,
+      "step": 1890
+    },
+    {
+      "epoch": 0.06079416376027901,
+      "grad_norm": 18.25,
+      "learning_rate": 1.977366116537555e-05,
+      "loss": 0.2063,
+      "step": 1900
+    },
+    {
+      "epoch": 0.06111413304322785,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.9766924890535534e-05,
+      "loss": 0.1415,
+      "step": 1910
+    },
+    {
+      "epoch": 0.06143410232617669,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.976018861569552e-05,
+      "loss": 0.1379,
+      "step": 1920
+    },
+    {
+      "epoch": 0.06175407160912552,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.975345234085551e-05,
+      "loss": 0.167,
+      "step": 1930
+    },
+    {
+      "epoch": 0.06207404089207436,
+      "grad_norm": 20.875,
+      "learning_rate": 1.9746716066015494e-05,
+      "loss": 0.1686,
+      "step": 1940
+    },
+    {
+      "epoch": 0.0623940101750232,
+      "grad_norm": 38.25,
+      "learning_rate": 1.973997979117548e-05,
+      "loss": 0.1934,
+      "step": 1950
+    },
+    {
+      "epoch": 0.06271397945797204,
+      "grad_norm": 14.0,
+      "learning_rate": 1.9733243516335466e-05,
+      "loss": 0.1013,
+      "step": 1960
+    },
+    {
+      "epoch": 0.06303394874092087,
+      "grad_norm": 15.25,
+      "learning_rate": 1.9726507241495454e-05,
+      "loss": 0.1828,
+      "step": 1970
+    },
+    {
+      "epoch": 0.06335391802386971,
+      "grad_norm": 23.5,
+      "learning_rate": 1.971977096665544e-05,
+      "loss": 0.1755,
+      "step": 1980
+    },
+    {
+      "epoch": 0.06367388730681854,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.9713034691815426e-05,
+      "loss": 0.1267,
+      "step": 1990
+    },
+    {
+      "epoch": 0.06399385658976738,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.9706298416975413e-05,
+      "loss": 0.1271,
+      "step": 2000
+    },
+    {
+      "epoch": 0.06431382587271622,
+      "grad_norm": 18.25,
+      "learning_rate": 1.96995621421354e-05,
+      "loss": 0.1519,
+      "step": 2010
+    },
+    {
+      "epoch": 0.06463379515566506,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.9692825867295385e-05,
+      "loss": 0.1959,
+      "step": 2020
+    },
+    {
+      "epoch": 0.06495376443861389,
+      "grad_norm": 7.375,
+      "learning_rate": 1.9686089592455373e-05,
+      "loss": 0.1422,
+      "step": 2030
+    },
+    {
+      "epoch": 0.06527373372156273,
+      "grad_norm": 26.75,
+      "learning_rate": 1.967935331761536e-05,
+      "loss": 0.1793,
+      "step": 2040
+    },
+    {
+      "epoch": 0.06559370300451156,
+      "grad_norm": 2.75,
+      "learning_rate": 1.9672617042775345e-05,
+      "loss": 0.1432,
+      "step": 2050
+    },
+    {
+      "epoch": 0.06591367228746041,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9665880767935333e-05,
+      "loss": 0.1691,
+      "step": 2060
+    },
+    {
+      "epoch": 0.06623364157040924,
+      "grad_norm": 29.625,
+      "learning_rate": 1.9659144493095317e-05,
+      "loss": 0.1804,
+      "step": 2070
+    },
+    {
+      "epoch": 0.06655361085335808,
+      "grad_norm": 5.90625,
+      "learning_rate": 1.9652408218255305e-05,
+      "loss": 0.1773,
+      "step": 2080
+    },
+    {
+      "epoch": 0.06687358013630691,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9645671943415293e-05,
+      "loss": 0.1771,
+      "step": 2090
+    },
+    {
+      "epoch": 0.06719354941925575,
+      "grad_norm": 7.5,
+      "learning_rate": 1.9638935668575277e-05,
+      "loss": 0.1267,
+      "step": 2100
+    },
+    {
+      "epoch": 0.0675135187022046,
+      "grad_norm": 11.125,
+      "learning_rate": 1.9632199393735265e-05,
+      "loss": 0.1476,
+      "step": 2110
+    },
+    {
+      "epoch": 0.06783348798515343,
+      "grad_norm": 32.0,
+      "learning_rate": 1.9625463118895253e-05,
+      "loss": 0.1533,
+      "step": 2120
+    },
+    {
+      "epoch": 0.06815345726810226,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.9618726844055237e-05,
+      "loss": 0.1269,
+      "step": 2130
+    },
+    {
+      "epoch": 0.0684734265510511,
+      "grad_norm": 25.375,
+      "learning_rate": 1.9611990569215225e-05,
+      "loss": 0.1749,
+      "step": 2140
+    },
+    {
+      "epoch": 0.06879339583399993,
+      "grad_norm": 4.5625,
+      "learning_rate": 1.9605254294375212e-05,
+      "loss": 0.1499,
+      "step": 2150
+    },
+    {
+      "epoch": 0.06911336511694877,
+      "grad_norm": 24.125,
+      "learning_rate": 1.9598518019535197e-05,
+      "loss": 0.1549,
+      "step": 2160
+    },
+    {
+      "epoch": 0.06943333439989761,
+      "grad_norm": 32.25,
+      "learning_rate": 1.9591781744695185e-05,
+      "loss": 0.2125,
+      "step": 2170
+    },
+    {
+      "epoch": 0.06975330368284645,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.9585045469855172e-05,
+      "loss": 0.1628,
+      "step": 2180
+    },
+    {
+      "epoch": 0.07007327296579528,
+      "grad_norm": 6.1875,
+      "learning_rate": 1.9578309195015157e-05,
+      "loss": 0.2004,
+      "step": 2190
+    },
+    {
+      "epoch": 0.07039324224874412,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.9571572920175144e-05,
+      "loss": 0.1494,
+      "step": 2200
+    },
+    {
+      "epoch": 0.07071321153169295,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.9564836645335132e-05,
+      "loss": 0.1356,
+      "step": 2210
+    },
+    {
+      "epoch": 0.0710331808146418,
+      "grad_norm": 45.75,
+      "learning_rate": 1.955810037049512e-05,
+      "loss": 0.1338,
+      "step": 2220
+    },
+    {
+      "epoch": 0.07135315009759063,
+      "grad_norm": 10.5625,
+      "learning_rate": 1.9551364095655104e-05,
+      "loss": 0.1166,
+      "step": 2230
+    },
+    {
+      "epoch": 0.07167311938053947,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.9544627820815092e-05,
+      "loss": 0.1911,
+      "step": 2240
+    },
+    {
+      "epoch": 0.0719930886634883,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.953789154597508e-05,
+      "loss": 0.148,
+      "step": 2250
+    },
+    {
+      "epoch": 0.07231305794643714,
+      "grad_norm": 6.875,
+      "learning_rate": 1.9531155271135064e-05,
+      "loss": 0.1107,
+      "step": 2260
+    },
+    {
+      "epoch": 0.07263302722938599,
+      "grad_norm": 9.375,
+      "learning_rate": 1.9524418996295052e-05,
+      "loss": 0.2081,
+      "step": 2270
+    },
+    {
+      "epoch": 0.07295299651233482,
+      "grad_norm": 4.8125,
+      "learning_rate": 1.951768272145504e-05,
+      "loss": 0.1822,
+      "step": 2280
+    },
+    {
+      "epoch": 0.07327296579528365,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9510946446615024e-05,
+      "loss": 0.149,
+      "step": 2290
+    },
+    {
+      "epoch": 0.07359293507823249,
+      "grad_norm": 16.375,
+      "learning_rate": 1.950421017177501e-05,
+      "loss": 0.152,
+      "step": 2300
+    },
+    {
+      "epoch": 0.07391290436118132,
+      "grad_norm": 20.625,
+      "learning_rate": 1.9497473896934996e-05,
+      "loss": 0.1349,
+      "step": 2310
+    },
+    {
+      "epoch": 0.07423287364413016,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.9490737622094984e-05,
+      "loss": 0.1564,
+      "step": 2320
+    },
+    {
+      "epoch": 0.074552842927079,
+      "grad_norm": 8.25,
+      "learning_rate": 1.948400134725497e-05,
+      "loss": 0.1706,
+      "step": 2330
+    },
+    {
+      "epoch": 0.07487281221002784,
+      "grad_norm": 5.84375,
+      "learning_rate": 1.9477265072414956e-05,
+      "loss": 0.1003,
+      "step": 2340
+    },
+    {
+      "epoch": 0.07519278149297667,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.9470528797574943e-05,
+      "loss": 0.1849,
+      "step": 2350
+    },
+    {
+      "epoch": 0.07551275077592551,
+      "grad_norm": 8.25,
+      "learning_rate": 1.946379252273493e-05,
+      "loss": 0.1823,
+      "step": 2360
+    },
+    {
+      "epoch": 0.07583272005887434,
+      "grad_norm": 9.625,
+      "learning_rate": 1.9457056247894916e-05,
+      "loss": 0.1679,
+      "step": 2370
+    },
+    {
+      "epoch": 0.07615268934182319,
+      "grad_norm": 4.75,
+      "learning_rate": 1.9450319973054903e-05,
+      "loss": 0.1187,
+      "step": 2380
+    },
+    {
+      "epoch": 0.07647265862477202,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.944358369821489e-05,
+      "loss": 0.1311,
+      "step": 2390
+    },
+    {
+      "epoch": 0.07679262790772086,
+      "grad_norm": 11.5,
+      "learning_rate": 1.9436847423374875e-05,
+      "loss": 0.145,
+      "step": 2400
+    },
+    {
+      "epoch": 0.0771125971906697,
+      "grad_norm": 4.75,
+      "learning_rate": 1.9430111148534863e-05,
+      "loss": 0.1396,
+      "step": 2410
+    },
+    {
+      "epoch": 0.07743256647361853,
+      "grad_norm": 11.375,
+      "learning_rate": 1.9423374873694847e-05,
+      "loss": 0.155,
+      "step": 2420
+    },
+    {
+      "epoch": 0.07775253575656738,
+      "grad_norm": 12.5,
+      "learning_rate": 1.9416638598854835e-05,
+      "loss": 0.174,
+      "step": 2430
+    },
+    {
+      "epoch": 0.07807250503951621,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.9409902324014823e-05,
+      "loss": 0.1328,
+      "step": 2440
+    },
+    {
+      "epoch": 0.07839247432246504,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.9403166049174807e-05,
+      "loss": 0.1469,
+      "step": 2450
+    },
+    {
+      "epoch": 0.07871244360541388,
+      "grad_norm": 27.375,
+      "learning_rate": 1.9396429774334795e-05,
+      "loss": 0.1726,
+      "step": 2460
+    },
+    {
+      "epoch": 0.07903241288836271,
+      "grad_norm": 9.0,
+      "learning_rate": 1.9389693499494783e-05,
+      "loss": 0.205,
+      "step": 2470
+    },
+    {
+      "epoch": 0.07935238217131155,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9382957224654767e-05,
+      "loss": 0.1394,
+      "step": 2480
+    },
+    {
+      "epoch": 0.0796723514542604,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.9376220949814755e-05,
+      "loss": 0.1348,
+      "step": 2490
+    },
+    {
+      "epoch": 0.07999232073720923,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.9369484674974743e-05,
+      "loss": 0.1325,
+      "step": 2500
+    },
+    {
+      "epoch": 0.08031229002015806,
+      "grad_norm": 10.0,
+      "learning_rate": 1.9362748400134727e-05,
+      "loss": 0.1706,
+      "step": 2510
+    },
+    {
+      "epoch": 0.0806322593031069,
+      "grad_norm": 7.0,
+      "learning_rate": 1.9356012125294715e-05,
+      "loss": 0.1367,
+      "step": 2520
+    },
+    {
+      "epoch": 0.08095222858605573,
+      "grad_norm": 4.78125,
+      "learning_rate": 1.9349275850454702e-05,
+      "loss": 0.1703,
+      "step": 2530
+    },
+    {
+      "epoch": 0.08127219786900458,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9342539575614687e-05,
+      "loss": 0.17,
+      "step": 2540
+    },
+    {
+      "epoch": 0.08159216715195342,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.9335803300774674e-05,
+      "loss": 0.118,
+      "step": 2550
+    },
+    {
+      "epoch": 0.08191213643490225,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.932906702593466e-05,
+      "loss": 0.1729,
+      "step": 2560
+    },
+    {
+      "epoch": 0.08223210571785108,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9322330751094647e-05,
+      "loss": 0.1268,
+      "step": 2570
+    },
+    {
+      "epoch": 0.08255207500079992,
+      "grad_norm": 5.9375,
+      "learning_rate": 1.9315594476254634e-05,
+      "loss": 0.1394,
+      "step": 2580
+    },
+    {
+      "epoch": 0.08287204428374877,
+      "grad_norm": 29.75,
+      "learning_rate": 1.930885820141462e-05,
+      "loss": 0.1999,
+      "step": 2590
+    },
+    {
+      "epoch": 0.0831920135666976,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.9302121926574606e-05,
+      "loss": 0.1494,
+      "step": 2600
+    },
+    {
+      "epoch": 0.08351198284964644,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.9295385651734594e-05,
+      "loss": 0.1139,
+      "step": 2610
+    },
+    {
+      "epoch": 0.08383195213259527,
+      "grad_norm": 21.375,
+      "learning_rate": 1.928864937689458e-05,
+      "loss": 0.1486,
+      "step": 2620
+    },
+    {
+      "epoch": 0.0841519214155441,
+      "grad_norm": 25.75,
+      "learning_rate": 1.9281913102054566e-05,
+      "loss": 0.1235,
+      "step": 2630
+    },
+    {
+      "epoch": 0.08447189069849294,
+      "grad_norm": 37.75,
+      "learning_rate": 1.9275176827214554e-05,
+      "loss": 0.1411,
+      "step": 2640
+    },
+    {
+      "epoch": 0.08479185998144179,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.9268440552374538e-05,
+      "loss": 0.1276,
+      "step": 2650
+    },
+    {
+      "epoch": 0.08511182926439062,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9261704277534526e-05,
+      "loss": 0.1168,
+      "step": 2660
+    },
+    {
+      "epoch": 0.08543179854733945,
+      "grad_norm": 26.5,
+      "learning_rate": 1.925496800269451e-05,
+      "loss": 0.2217,
+      "step": 2670
+    },
+    {
+      "epoch": 0.08575176783028829,
+      "grad_norm": 14.5,
+      "learning_rate": 1.9248231727854498e-05,
+      "loss": 0.1944,
+      "step": 2680
+    },
+    {
+      "epoch": 0.08607173711323712,
+      "grad_norm": 15.0,
+      "learning_rate": 1.9241495453014486e-05,
+      "loss": 0.1108,
+      "step": 2690
+    },
+    {
+      "epoch": 0.08639170639618597,
+      "grad_norm": 35.25,
+      "learning_rate": 1.923475917817447e-05,
+      "loss": 0.116,
+      "step": 2700
+    },
+    {
+      "epoch": 0.0867116756791348,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.9228022903334458e-05,
+      "loss": 0.0653,
+      "step": 2710
+    },
+    {
+      "epoch": 0.08703164496208364,
+      "grad_norm": 24.875,
+      "learning_rate": 1.9221286628494446e-05,
+      "loss": 0.1254,
+      "step": 2720
+    },
+    {
+      "epoch": 0.08735161424503247,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.921455035365443e-05,
+      "loss": 0.1092,
+      "step": 2730
+    },
+    {
+      "epoch": 0.08767158352798131,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.9207814078814418e-05,
+      "loss": 0.145,
+      "step": 2740
+    },
+    {
+      "epoch": 0.08799155281093016,
+      "grad_norm": 13.625,
+      "learning_rate": 1.9201077803974405e-05,
+      "loss": 0.2331,
+      "step": 2750
+    },
+    {
+      "epoch": 0.08831152209387899,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.919434152913439e-05,
+      "loss": 0.1263,
+      "step": 2760
+    },
+    {
+      "epoch": 0.08863149137682783,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9187605254294378e-05,
+      "loss": 0.0827,
+      "step": 2770
+    },
+    {
+      "epoch": 0.08895146065977666,
+      "grad_norm": 10.625,
+      "learning_rate": 1.9180868979454362e-05,
+      "loss": 0.1144,
+      "step": 2780
+    },
+    {
+      "epoch": 0.0892714299427255,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.917413270461435e-05,
+      "loss": 0.1831,
+      "step": 2790
+    },
+    {
+      "epoch": 0.08959139922567433,
+      "grad_norm": 8.625,
+      "learning_rate": 1.9167396429774337e-05,
+      "loss": 0.1774,
+      "step": 2800
+    },
+    {
+      "epoch": 0.08991136850862318,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.916066015493432e-05,
+      "loss": 0.1444,
+      "step": 2810
+    },
+    {
+      "epoch": 0.09023133779157201,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.915392388009431e-05,
+      "loss": 0.1691,
+      "step": 2820
+    },
+    {
+      "epoch": 0.09055130707452085,
+      "grad_norm": 11.5,
+      "learning_rate": 1.9147187605254297e-05,
+      "loss": 0.119,
+      "step": 2830
+    },
+    {
+      "epoch": 0.09087127635746968,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.914045133041428e-05,
+      "loss": 0.1408,
+      "step": 2840
+    },
+    {
+      "epoch": 0.09119124564041851,
+      "grad_norm": 11.125,
+      "learning_rate": 1.913371505557427e-05,
+      "loss": 0.1178,
+      "step": 2850
+    },
+    {
+      "epoch": 0.09151121492336736,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9126978780734257e-05,
+      "loss": 0.0589,
+      "step": 2860
+    },
+    {
+      "epoch": 0.0918311842063162,
+      "grad_norm": 21.375,
+      "learning_rate": 1.912024250589424e-05,
+      "loss": 0.2153,
+      "step": 2870
+    },
+    {
+      "epoch": 0.09215115348926503,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.911350623105423e-05,
+      "loss": 0.1499,
+      "step": 2880
+    },
+    {
+      "epoch": 0.09247112277221387,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.9106769956214213e-05,
+      "loss": 0.1632,
+      "step": 2890
+    },
+    {
+      "epoch": 0.0927910920551627,
+      "grad_norm": 4.625,
+      "learning_rate": 1.91000336813742e-05,
+      "loss": 0.1638,
+      "step": 2900
+    },
+    {
+      "epoch": 0.09311106133811155,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.909329740653419e-05,
+      "loss": 0.0883,
+      "step": 2910
+    },
+    {
+      "epoch": 0.09343103062106038,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.9086561131694173e-05,
+      "loss": 0.1125,
+      "step": 2920
+    },
+    {
+      "epoch": 0.09375099990400922,
+      "grad_norm": 19.625,
+      "learning_rate": 1.907982485685416e-05,
+      "loss": 0.1405,
+      "step": 2930
+    },
+    {
+      "epoch": 0.09407096918695805,
+      "grad_norm": 27.0,
+      "learning_rate": 1.907308858201415e-05,
+      "loss": 0.176,
+      "step": 2940
+    },
+    {
+      "epoch": 0.09439093846990688,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.9066352307174133e-05,
+      "loss": 0.1551,
+      "step": 2950
+    },
+    {
+      "epoch": 0.09471090775285572,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.905961603233412e-05,
+      "loss": 0.1922,
+      "step": 2960
+    },
+    {
+      "epoch": 0.09503087703580457,
+      "grad_norm": 9.75,
+      "learning_rate": 1.905287975749411e-05,
+      "loss": 0.1658,
+      "step": 2970
+    },
+    {
+      "epoch": 0.0953508463187534,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.9046143482654093e-05,
+      "loss": 0.1546,
+      "step": 2980
+    },
+    {
+      "epoch": 0.09567081560170224,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.903940720781408e-05,
+      "loss": 0.1501,
+      "step": 2990
+    },
+    {
+      "epoch": 0.09599078488465107,
+      "grad_norm": 13.0,
+      "learning_rate": 1.903267093297407e-05,
+      "loss": 0.086,
+      "step": 3000
+    },
+    {
+      "epoch": 0.0963107541675999,
+      "grad_norm": 16.375,
+      "learning_rate": 1.9025934658134053e-05,
+      "loss": 0.135,
+      "step": 3010
+    },
+    {
+      "epoch": 0.09663072345054875,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.901919838329404e-05,
+      "loss": 0.1558,
+      "step": 3020
+    },
+    {
+      "epoch": 0.09695069273349759,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.9012462108454025e-05,
+      "loss": 0.1903,
+      "step": 3030
+    },
+    {
+      "epoch": 0.09727066201644642,
+      "grad_norm": 9.25,
+      "learning_rate": 1.9005725833614012e-05,
+      "loss": 0.1578,
+      "step": 3040
+    },
+    {
+      "epoch": 0.09759063129939526,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.8998989558774e-05,
+      "loss": 0.164,
+      "step": 3050
+    },
+    {
+      "epoch": 0.09791060058234409,
+      "grad_norm": 16.5,
+      "learning_rate": 1.8992253283933985e-05,
+      "loss": 0.1594,
+      "step": 3060
+    },
+    {
+      "epoch": 0.09823056986529294,
+      "grad_norm": 28.125,
+      "learning_rate": 1.8985517009093972e-05,
+      "loss": 0.1359,
+      "step": 3070
+    },
+    {
+      "epoch": 0.09855053914824177,
+      "grad_norm": 10.375,
+      "learning_rate": 1.897878073425396e-05,
+      "loss": 0.1359,
+      "step": 3080
+    },
+    {
+      "epoch": 0.0988705084311906,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.8972044459413944e-05,
+      "loss": 0.1366,
+      "step": 3090
+    },
+    {
+      "epoch": 0.09919047771413944,
+      "grad_norm": 5.21875,
+      "learning_rate": 1.8965308184573932e-05,
+      "loss": 0.1435,
+      "step": 3100
+    },
+    {
+      "epoch": 0.09951044699708828,
+      "grad_norm": 16.125,
+      "learning_rate": 1.895857190973392e-05,
+      "loss": 0.1572,
+      "step": 3110
+    },
+    {
+      "epoch": 0.09983041628003711,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.8951835634893904e-05,
+      "loss": 0.0966,
+      "step": 3120
+    },
+    {
+      "epoch": 0.10015038556298596,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.8945099360053892e-05,
+      "loss": 0.1203,
+      "step": 3130
+    },
+    {
+      "epoch": 0.10047035484593479,
+      "grad_norm": 24.625,
+      "learning_rate": 1.8938363085213876e-05,
+      "loss": 0.1025,
+      "step": 3140
+    },
+    {
+      "epoch": 0.10079032412888363,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.8931626810373864e-05,
+      "loss": 0.1449,
+      "step": 3150
+    },
+    {
+      "epoch": 0.10111029341183246,
+      "grad_norm": 15.75,
+      "learning_rate": 1.8924890535533852e-05,
+      "loss": 0.1695,
+      "step": 3160
+    },
+    {
+      "epoch": 0.1014302626947813,
+      "grad_norm": 37.5,
+      "learning_rate": 1.8918154260693836e-05,
+      "loss": 0.1122,
+      "step": 3170
+    },
+    {
+      "epoch": 0.10175023197773014,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.8911417985853824e-05,
+      "loss": 0.1035,
+      "step": 3180
+    },
+    {
+      "epoch": 0.10207020126067898,
+      "grad_norm": 20.5,
+      "learning_rate": 1.890468171101381e-05,
+      "loss": 0.1842,
+      "step": 3190
+    },
+    {
+      "epoch": 0.10239017054362781,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.8897945436173796e-05,
+      "loss": 0.1156,
+      "step": 3200
+    },
+    {
+      "epoch": 0.10271013982657665,
+      "grad_norm": 6.03125,
+      "learning_rate": 1.8891209161333784e-05,
+      "loss": 0.1611,
+      "step": 3210
+    },
+    {
+      "epoch": 0.10303010910952548,
+      "grad_norm": 8.5,
+      "learning_rate": 1.888447288649377e-05,
+      "loss": 0.1497,
+      "step": 3220
+    },
+    {
+      "epoch": 0.10335007839247433,
+      "grad_norm": 13.125,
+      "learning_rate": 1.8877736611653756e-05,
+      "loss": 0.1232,
+      "step": 3230
+    },
+    {
+      "epoch": 0.10367004767542316,
+      "grad_norm": 21.25,
+      "learning_rate": 1.8871000336813743e-05,
+      "loss": 0.1139,
+      "step": 3240
+    },
+    {
+      "epoch": 0.103990016958372,
+      "grad_norm": 25.75,
+      "learning_rate": 1.8864264061973728e-05,
+      "loss": 0.1872,
+      "step": 3250
+    },
+    {
+      "epoch": 0.10430998624132083,
+      "grad_norm": 6.28125,
+      "learning_rate": 1.8857527787133716e-05,
+      "loss": 0.0994,
+      "step": 3260
+    },
+    {
+      "epoch": 0.10462995552426967,
+      "grad_norm": 28.0,
+      "learning_rate": 1.8850791512293703e-05,
+      "loss": 0.0856,
+      "step": 3270
+    },
+    {
+      "epoch": 0.1049499248072185,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.8844055237453688e-05,
+      "loss": 0.1072,
+      "step": 3280
+    },
+    {
+      "epoch": 0.10526989409016735,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.8837318962613675e-05,
+      "loss": 0.1836,
+      "step": 3290
+    },
+    {
+      "epoch": 0.10558986337311618,
+      "grad_norm": 16.75,
+      "learning_rate": 1.8830582687773663e-05,
+      "loss": 0.171,
+      "step": 3300
+    },
+    {
+      "epoch": 0.10590983265606502,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.8823846412933647e-05,
+      "loss": 0.1256,
+      "step": 3310
+    },
+    {
+      "epoch": 0.10622980193901385,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.8817110138093635e-05,
+      "loss": 0.071,
+      "step": 3320
+    },
+    {
+      "epoch": 0.10654977122196269,
+      "grad_norm": 22.0,
+      "learning_rate": 1.8810373863253623e-05,
+      "loss": 0.2003,
+      "step": 3330
+    },
+    {
+      "epoch": 0.10686974050491153,
+      "grad_norm": 10.0,
+      "learning_rate": 1.8803637588413607e-05,
+      "loss": 0.0824,
+      "step": 3340
+    },
+    {
+      "epoch": 0.10718970978786037,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.8796901313573595e-05,
+      "loss": 0.1043,
+      "step": 3350
+    },
+    {
+      "epoch": 0.1075096790708092,
+      "grad_norm": 7.25,
+      "learning_rate": 1.879016503873358e-05,
+      "loss": 0.1802,
+      "step": 3360
+    },
+    {
+      "epoch": 0.10782964835375804,
+      "grad_norm": 4.8125,
+      "learning_rate": 1.8783428763893567e-05,
+      "loss": 0.1372,
+      "step": 3370
+    },
+    {
+      "epoch": 0.10814961763670687,
+      "grad_norm": 11.4375,
+      "learning_rate": 1.8776692489053555e-05,
+      "loss": 0.1097,
+      "step": 3380
+    },
+    {
+      "epoch": 0.10846958691965572,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.876995621421354e-05,
+      "loss": 0.1801,
+      "step": 3390
+    },
+    {
+      "epoch": 0.10878955620260455,
+      "grad_norm": 12.375,
+      "learning_rate": 1.8763219939373527e-05,
+      "loss": 0.1607,
+      "step": 3400
+    },
+    {
+      "epoch": 0.10910952548555339,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.8756483664533515e-05,
+      "loss": 0.1361,
+      "step": 3410
+    },
+    {
+      "epoch": 0.10942949476850222,
+      "grad_norm": 12.75,
+      "learning_rate": 1.87497473896935e-05,
+      "loss": 0.1432,
+      "step": 3420
+    },
+    {
+      "epoch": 0.10974946405145106,
+      "grad_norm": 18.75,
+      "learning_rate": 1.8743011114853487e-05,
+      "loss": 0.1197,
+      "step": 3430
+    },
+    {
+      "epoch": 0.11006943333439989,
+      "grad_norm": 12.0,
+      "learning_rate": 1.8736274840013474e-05,
+      "loss": 0.1405,
+      "step": 3440
+    },
+    {
+      "epoch": 0.11038940261734874,
+      "grad_norm": 6.59375,
+      "learning_rate": 1.872953856517346e-05,
+      "loss": 0.1271,
+      "step": 3450
+    },
+    {
+      "epoch": 0.11070937190029757,
+      "grad_norm": 27.125,
+      "learning_rate": 1.8722802290333447e-05,
+      "loss": 0.1116,
+      "step": 3460
+    },
+    {
+      "epoch": 0.11102934118324641,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.8716066015493434e-05,
+      "loss": 0.1374,
+      "step": 3470
+    },
+    {
+      "epoch": 0.11134931046619524,
+      "grad_norm": 2.5,
+      "learning_rate": 1.870932974065342e-05,
+      "loss": 0.1483,
+      "step": 3480
+    },
+    {
+      "epoch": 0.11166927974914408,
+      "grad_norm": 5.53125,
+      "learning_rate": 1.8702593465813406e-05,
+      "loss": 0.1863,
+      "step": 3490
+    },
+    {
+      "epoch": 0.11198924903209292,
+      "grad_norm": 6.34375,
+      "learning_rate": 1.869585719097339e-05,
+      "loss": 0.1741,
+      "step": 3500
+    },
+    {
+      "epoch": 0.11230921831504176,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.8689120916133382e-05,
+      "loss": 0.0774,
+      "step": 3510
+    },
+    {
+      "epoch": 0.1126291875979906,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.8682384641293366e-05,
+      "loss": 0.1872,
+      "step": 3520
+    },
+    {
+      "epoch": 0.11294915688093943,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.8675648366453354e-05,
+      "loss": 0.0897,
+      "step": 3530
+    },
+    {
+      "epoch": 0.11326912616388826,
+      "grad_norm": 12.25,
+      "learning_rate": 1.866891209161334e-05,
+      "loss": 0.1631,
+      "step": 3540
+    },
+    {
+      "epoch": 0.11358909544683711,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.8662175816773326e-05,
+      "loss": 0.1896,
+      "step": 3550
+    },
+    {
+      "epoch": 0.11390906472978594,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.8655439541933314e-05,
+      "loss": 0.1373,
+      "step": 3560
+    },
+    {
+      "epoch": 0.11422903401273478,
+      "grad_norm": 9.625,
+      "learning_rate": 1.86487032670933e-05,
+      "loss": 0.1341,
+      "step": 3570
+    },
+    {
+      "epoch": 0.11454900329568361,
+      "grad_norm": 12.25,
+      "learning_rate": 1.8641966992253286e-05,
+      "loss": 0.1644,
+      "step": 3580
+    },
+    {
+      "epoch": 0.11486897257863245,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.8635230717413274e-05,
+      "loss": 0.1105,
+      "step": 3590
+    },
+    {
+      "epoch": 0.11518894186158128,
+      "grad_norm": 5.71875,
+      "learning_rate": 1.8628494442573258e-05,
+      "loss": 0.0822,
+      "step": 3600
+    },
+    {
+      "epoch": 0.11550891114453013,
+      "grad_norm": 18.25,
+      "learning_rate": 1.8621758167733246e-05,
+      "loss": 0.1805,
+      "step": 3610
+    },
+    {
+      "epoch": 0.11582888042747896,
+      "grad_norm": 12.0,
+      "learning_rate": 1.8615021892893233e-05,
+      "loss": 0.1365,
+      "step": 3620
+    },
+    {
+      "epoch": 0.1161488497104278,
+      "grad_norm": 10.375,
+      "learning_rate": 1.8608285618053218e-05,
+      "loss": 0.1521,
+      "step": 3630
+    },
+    {
+      "epoch": 0.11646881899337663,
+      "grad_norm": 9.125,
+      "learning_rate": 1.8601549343213205e-05,
+      "loss": 0.175,
+      "step": 3640
+    },
+    {
+      "epoch": 0.11678878827632547,
+      "grad_norm": 9.25,
+      "learning_rate": 1.8594813068373193e-05,
+      "loss": 0.0963,
+      "step": 3650
+    },
+    {
+      "epoch": 0.11710875755927432,
+      "grad_norm": 16.75,
+      "learning_rate": 1.8588076793533178e-05,
+      "loss": 0.1028,
+      "step": 3660
+    },
+    {
+      "epoch": 0.11742872684222315,
+      "grad_norm": 4.25,
+      "learning_rate": 1.8581340518693165e-05,
+      "loss": 0.1375,
+      "step": 3670
+    },
+    {
+      "epoch": 0.11774869612517198,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.8574604243853153e-05,
+      "loss": 0.1694,
+      "step": 3680
+    },
+    {
+      "epoch": 0.11806866540812082,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.8567867969013137e-05,
+      "loss": 0.1823,
+      "step": 3690
+    },
+    {
+      "epoch": 0.11838863469106965,
+      "grad_norm": 11.0,
+      "learning_rate": 1.8561131694173125e-05,
+      "loss": 0.1314,
+      "step": 3700
+    },
+    {
+      "epoch": 0.1187086039740185,
+      "grad_norm": 12.375,
+      "learning_rate": 1.855439541933311e-05,
+      "loss": 0.105,
+      "step": 3710
+    },
+    {
+      "epoch": 0.11902857325696733,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.8547659144493097e-05,
+      "loss": 0.1525,
+      "step": 3720
+    },
+    {
+      "epoch": 0.11934854253991617,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.8540922869653085e-05,
+      "loss": 0.0756,
+      "step": 3730
+    },
+    {
+      "epoch": 0.119668511822865,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.853418659481307e-05,
+      "loss": 0.1043,
+      "step": 3740
+    },
+    {
+      "epoch": 0.11998848110581384,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.8527450319973057e-05,
+      "loss": 0.1945,
+      "step": 3750
+    },
+    {
+      "epoch": 0.12030845038876267,
+      "grad_norm": 16.125,
+      "learning_rate": 1.8520714045133045e-05,
+      "loss": 0.1316,
+      "step": 3760
+    },
+    {
+      "epoch": 0.12062841967171152,
+      "grad_norm": 13.875,
+      "learning_rate": 1.851397777029303e-05,
+      "loss": 0.1479,
+      "step": 3770
+    },
+    {
+      "epoch": 0.12094838895466035,
+      "grad_norm": 15.125,
+      "learning_rate": 1.8507241495453017e-05,
+      "loss": 0.1301,
+      "step": 3780
+    },
+    {
+      "epoch": 0.12126835823760919,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.8500505220613005e-05,
+      "loss": 0.159,
+      "step": 3790
+    },
+    {
+      "epoch": 0.12158832752055802,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.849376894577299e-05,
+      "loss": 0.1204,
+      "step": 3800
+    },
+    {
+      "epoch": 0.12190829680350686,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8487032670932977e-05,
+      "loss": 0.1535,
+      "step": 3810
+    },
+    {
+      "epoch": 0.1222282660864557,
+      "grad_norm": 6.15625,
+      "learning_rate": 1.848029639609296e-05,
+      "loss": 0.1375,
+      "step": 3820
+    },
+    {
+      "epoch": 0.12254823536940454,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.847356012125295e-05,
+      "loss": 0.147,
+      "step": 3830
+    },
+    {
+      "epoch": 0.12286820465235337,
+      "grad_norm": 29.375,
+      "learning_rate": 1.8466823846412936e-05,
+      "loss": 0.0999,
+      "step": 3840
+    },
+    {
+      "epoch": 0.12318817393530221,
+      "grad_norm": 5.96875,
+      "learning_rate": 1.846008757157292e-05,
+      "loss": 0.131,
+      "step": 3850
+    },
+    {
+      "epoch": 0.12350814321825104,
+      "grad_norm": 14.125,
+      "learning_rate": 1.845335129673291e-05,
+      "loss": 0.1024,
+      "step": 3860
+    },
+    {
+      "epoch": 0.12382811250119989,
+      "grad_norm": 6.09375,
+      "learning_rate": 1.8446615021892896e-05,
+      "loss": 0.1799,
+      "step": 3870
+    },
+    {
+      "epoch": 0.12414808178414873,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.843987874705288e-05,
+      "loss": 0.145,
+      "step": 3880
+    },
+    {
+      "epoch": 0.12446805106709756,
+      "grad_norm": 23.875,
+      "learning_rate": 1.843314247221287e-05,
+      "loss": 0.1232,
+      "step": 3890
+    },
+    {
+      "epoch": 0.1247880203500464,
+      "grad_norm": 5.5,
+      "learning_rate": 1.8426406197372856e-05,
+      "loss": 0.1572,
+      "step": 3900
+    },
+    {
+      "epoch": 0.12510798963299524,
+      "grad_norm": 13.625,
+      "learning_rate": 1.841966992253284e-05,
+      "loss": 0.0954,
+      "step": 3910
+    },
+    {
+      "epoch": 0.12542795891594408,
+      "grad_norm": 18.375,
+      "learning_rate": 1.8412933647692828e-05,
+      "loss": 0.1774,
+      "step": 3920
+    },
+    {
+      "epoch": 0.1257479281988929,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.8406197372852816e-05,
+      "loss": 0.1103,
+      "step": 3930
+    },
+    {
+      "epoch": 0.12606789748184175,
+      "grad_norm": 5.78125,
+      "learning_rate": 1.83994610980128e-05,
+      "loss": 0.1105,
+      "step": 3940
+    },
+    {
+      "epoch": 0.12638786676479058,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.8392724823172788e-05,
+      "loss": 0.0947,
+      "step": 3950
+    },
+    {
+      "epoch": 0.12670783604773941,
+      "grad_norm": 24.75,
+      "learning_rate": 1.8385988548332772e-05,
+      "loss": 0.1479,
+      "step": 3960
+    },
+    {
+      "epoch": 0.12702780533068825,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.837925227349276e-05,
+      "loss": 0.1065,
+      "step": 3970
+    },
+    {
+      "epoch": 0.12734777461363708,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.8372515998652748e-05,
+      "loss": 0.1766,
+      "step": 3980
+    },
+    {
+      "epoch": 0.12766774389658592,
+      "grad_norm": 18.375,
+      "learning_rate": 1.8365779723812732e-05,
+      "loss": 0.0801,
+      "step": 3990
+    },
+    {
+      "epoch": 0.12798771317953475,
+      "grad_norm": 20.375,
+      "learning_rate": 1.835904344897272e-05,
+      "loss": 0.1089,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1283076824624836,
+      "grad_norm": 26.5,
+      "learning_rate": 1.8352307174132708e-05,
+      "loss": 0.0943,
+      "step": 4010
+    },
+    {
+      "epoch": 0.12862765174543245,
+      "grad_norm": 33.75,
+      "learning_rate": 1.8345570899292692e-05,
+      "loss": 0.1652,
+      "step": 4020
+    },
+    {
+      "epoch": 0.12894762102838128,
+      "grad_norm": 17.0,
+      "learning_rate": 1.833883462445268e-05,
+      "loss": 0.1705,
+      "step": 4030
+    },
+    {
+      "epoch": 0.12926759031133012,
+      "grad_norm": 14.25,
+      "learning_rate": 1.8332098349612667e-05,
+      "loss": 0.1203,
+      "step": 4040
+    },
+    {
+      "epoch": 0.12958755959427895,
+      "grad_norm": 15.5,
+      "learning_rate": 1.8325362074772652e-05,
+      "loss": 0.0968,
+      "step": 4050
+    },
+    {
+      "epoch": 0.12990752887722778,
+      "grad_norm": 15.25,
+      "learning_rate": 1.831862579993264e-05,
+      "loss": 0.0765,
+      "step": 4060
+    },
+    {
+      "epoch": 0.13022749816017662,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8311889525092624e-05,
+      "loss": 0.1665,
+      "step": 4070
+    },
+    {
+      "epoch": 0.13054746744312545,
+      "grad_norm": 12.375,
+      "learning_rate": 1.830515325025261e-05,
+      "loss": 0.1342,
+      "step": 4080
+    },
+    {
+      "epoch": 0.1308674367260743,
+      "grad_norm": 28.25,
+      "learning_rate": 1.82984169754126e-05,
+      "loss": 0.1102,
+      "step": 4090
+    },
+    {
+      "epoch": 0.13118740600902312,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.8291680700572584e-05,
+      "loss": 0.1577,
+      "step": 4100
+    },
+    {
+      "epoch": 0.13150737529197198,
+      "grad_norm": 17.75,
+      "learning_rate": 1.828494442573257e-05,
+      "loss": 0.1468,
+      "step": 4110
+    },
+    {
+      "epoch": 0.13182734457492082,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.827820815089256e-05,
+      "loss": 0.1475,
+      "step": 4120
+    },
+    {
+      "epoch": 0.13214731385786965,
+      "grad_norm": 10.875,
+      "learning_rate": 1.8271471876052544e-05,
+      "loss": 0.124,
+      "step": 4130
+    },
+    {
+      "epoch": 0.1324672831408185,
+      "grad_norm": 17.375,
+      "learning_rate": 1.826473560121253e-05,
+      "loss": 0.1515,
+      "step": 4140
+    },
+    {
+      "epoch": 0.13278725242376732,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.825799932637252e-05,
+      "loss": 0.1272,
+      "step": 4150
+    },
+    {
+      "epoch": 0.13310722170671616,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.8251263051532503e-05,
+      "loss": 0.109,
+      "step": 4160
+    },
+    {
+      "epoch": 0.133427190989665,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.824452677669249e-05,
+      "loss": 0.1038,
+      "step": 4170
+    },
+    {
+      "epoch": 0.13374716027261382,
+      "grad_norm": 15.375,
+      "learning_rate": 1.8237790501852475e-05,
+      "loss": 0.135,
+      "step": 4180
+    },
+    {
+      "epoch": 0.13406712955556266,
+      "grad_norm": 15.5,
+      "learning_rate": 1.8231054227012463e-05,
+      "loss": 0.0941,
+      "step": 4190
+    },
+    {
+      "epoch": 0.1343870988385115,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.822431795217245e-05,
+      "loss": 0.1058,
+      "step": 4200
+    },
+    {
+      "epoch": 0.13470706812146033,
+      "grad_norm": 29.375,
+      "learning_rate": 1.8217581677332435e-05,
+      "loss": 0.133,
+      "step": 4210
+    },
+    {
+      "epoch": 0.1350270374044092,
+      "grad_norm": 18.0,
+      "learning_rate": 1.8210845402492423e-05,
+      "loss": 0.1107,
+      "step": 4220
+    },
+    {
+      "epoch": 0.13534700668735802,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.820410912765241e-05,
+      "loss": 0.1734,
+      "step": 4230
+    },
+    {
+      "epoch": 0.13566697597030686,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.8197372852812395e-05,
+      "loss": 0.1262,
+      "step": 4240
+    },
+    {
+      "epoch": 0.1359869452532557,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.8190636577972383e-05,
+      "loss": 0.0814,
+      "step": 4250
+    },
+    {
+      "epoch": 0.13630691453620453,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.818390030313237e-05,
+      "loss": 0.1544,
+      "step": 4260
+    },
+    {
+      "epoch": 0.13662688381915336,
+      "grad_norm": 4.90625,
+      "learning_rate": 1.8177164028292355e-05,
+      "loss": 0.0671,
+      "step": 4270
+    },
+    {
+      "epoch": 0.1369468531021022,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.8170427753452343e-05,
+      "loss": 0.0982,
+      "step": 4280
+    },
+    {
+      "epoch": 0.13726682238505103,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.8163691478612327e-05,
+      "loss": 0.1116,
+      "step": 4290
+    },
+    {
+      "epoch": 0.13758679166799986,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8156955203772315e-05,
+      "loss": 0.0687,
+      "step": 4300
+    },
+    {
+      "epoch": 0.1379067609509487,
+      "grad_norm": 21.375,
+      "learning_rate": 1.8150218928932302e-05,
+      "loss": 0.1176,
+      "step": 4310
+    },
+    {
+      "epoch": 0.13822673023389753,
+      "grad_norm": 10.5,
+      "learning_rate": 1.8143482654092287e-05,
+      "loss": 0.149,
+      "step": 4320
+    },
+    {
+      "epoch": 0.1385466995168464,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.8136746379252274e-05,
+      "loss": 0.1887,
+      "step": 4330
+    },
+    {
+      "epoch": 0.13886666879979523,
+      "grad_norm": 11.1875,
+      "learning_rate": 1.8130010104412262e-05,
+      "loss": 0.1482,
+      "step": 4340
+    },
+    {
+      "epoch": 0.13918663808274406,
+      "grad_norm": 23.625,
+      "learning_rate": 1.8123273829572247e-05,
+      "loss": 0.0944,
+      "step": 4350
+    },
+    {
+      "epoch": 0.1395066073656929,
+      "grad_norm": 36.75,
+      "learning_rate": 1.8116537554732234e-05,
+      "loss": 0.1426,
+      "step": 4360
+    },
+    {
+      "epoch": 0.13982657664864173,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.8109801279892222e-05,
+      "loss": 0.1638,
+      "step": 4370
+    },
+    {
+      "epoch": 0.14014654593159057,
+      "grad_norm": 9.375,
+      "learning_rate": 1.8103065005052206e-05,
+      "loss": 0.169,
+      "step": 4380
+    },
+    {
+      "epoch": 0.1404665152145394,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.8096328730212194e-05,
+      "loss": 0.129,
+      "step": 4390
+    },
+    {
+      "epoch": 0.14078648449748823,
+      "grad_norm": 4.75,
+      "learning_rate": 1.8089592455372182e-05,
+      "loss": 0.1339,
+      "step": 4400
+    },
+    {
+      "epoch": 0.14110645378043707,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.8082856180532166e-05,
+      "loss": 0.15,
+      "step": 4410
+    },
+    {
+      "epoch": 0.1414264230633859,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.8076119905692154e-05,
+      "loss": 0.1241,
+      "step": 4420
+    },
+    {
+      "epoch": 0.14174639234633477,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.8069383630852138e-05,
+      "loss": 0.1227,
+      "step": 4430
+    },
+    {
+      "epoch": 0.1420663616292836,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.8062647356012126e-05,
+      "loss": 0.1585,
+      "step": 4440
+    },
+    {
+      "epoch": 0.14238633091223243,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.8055911081172114e-05,
+      "loss": 0.1177,
+      "step": 4450
+    },
+    {
+      "epoch": 0.14270630019518127,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8049174806332098e-05,
+      "loss": 0.1297,
+      "step": 4460
+    },
+    {
+      "epoch": 0.1430262694781301,
+      "grad_norm": 10.875,
+      "learning_rate": 1.8042438531492086e-05,
+      "loss": 0.169,
+      "step": 4470
+    },
+    {
+      "epoch": 0.14334623876107894,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.8035702256652074e-05,
+      "loss": 0.179,
+      "step": 4480
+    },
+    {
+      "epoch": 0.14366620804402777,
+      "grad_norm": 10.125,
+      "learning_rate": 1.8028965981812058e-05,
+      "loss": 0.1445,
+      "step": 4490
+    },
+    {
+      "epoch": 0.1439861773269766,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.8022229706972046e-05,
+      "loss": 0.0873,
+      "step": 4500
+    },
+    {
+      "epoch": 0.14430614660992544,
+      "grad_norm": 22.625,
+      "learning_rate": 1.8015493432132033e-05,
+      "loss": 0.1543,
+      "step": 4510
+    },
+    {
+      "epoch": 0.14462611589287427,
+      "grad_norm": 19.25,
+      "learning_rate": 1.8008757157292018e-05,
+      "loss": 0.1299,
+      "step": 4520
+    },
+    {
+      "epoch": 0.1449460851758231,
+      "grad_norm": 27.5,
+      "learning_rate": 1.8002020882452005e-05,
+      "loss": 0.1449,
+      "step": 4530
+    },
+    {
+      "epoch": 0.14526605445877197,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.799528460761199e-05,
+      "loss": 0.0934,
+      "step": 4540
+    },
+    {
+      "epoch": 0.1455860237417208,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.7988548332771978e-05,
+      "loss": 0.1668,
+      "step": 4550
+    },
+    {
+      "epoch": 0.14590599302466964,
+      "grad_norm": 5.125,
+      "learning_rate": 1.7981812057931965e-05,
+      "loss": 0.1796,
+      "step": 4560
+    },
+    {
+      "epoch": 0.14622596230761847,
+      "grad_norm": 31.125,
+      "learning_rate": 1.797507578309195e-05,
+      "loss": 0.1353,
+      "step": 4570
+    },
+    {
+      "epoch": 0.1465459315905673,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.7968339508251937e-05,
+      "loss": 0.0967,
+      "step": 4580
+    },
+    {
+      "epoch": 0.14686590087351614,
+      "grad_norm": 4.65625,
+      "learning_rate": 1.7961603233411925e-05,
+      "loss": 0.1131,
+      "step": 4590
+    },
+    {
+      "epoch": 0.14718587015646498,
+      "grad_norm": 11.0,
+      "learning_rate": 1.795486695857191e-05,
+      "loss": 0.1127,
+      "step": 4600
+    },
+    {
+      "epoch": 0.1475058394394138,
+      "grad_norm": 22.875,
+      "learning_rate": 1.7948130683731897e-05,
+      "loss": 0.1579,
+      "step": 4610
+    },
+    {
+      "epoch": 0.14782580872236264,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.7941394408891885e-05,
+      "loss": 0.1329,
+      "step": 4620
+    },
+    {
+      "epoch": 0.14814577800531148,
+      "grad_norm": 10.625,
+      "learning_rate": 1.793465813405187e-05,
+      "loss": 0.1194,
+      "step": 4630
+    },
+    {
+      "epoch": 0.1484657472882603,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.7927921859211857e-05,
+      "loss": 0.1486,
+      "step": 4640
+    },
+    {
+      "epoch": 0.14878571657120918,
+      "grad_norm": 11.25,
+      "learning_rate": 1.792118558437184e-05,
+      "loss": 0.1315,
+      "step": 4650
+    },
+    {
+      "epoch": 0.149105685854158,
+      "grad_norm": 23.625,
+      "learning_rate": 1.791444930953183e-05,
+      "loss": 0.2333,
+      "step": 4660
+    },
+    {
+      "epoch": 0.14942565513710684,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.7907713034691817e-05,
+      "loss": 0.1039,
+      "step": 4670
+    },
+    {
+      "epoch": 0.14974562442005568,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.79009767598518e-05,
+      "loss": 0.1562,
+      "step": 4680
+    },
+    {
+      "epoch": 0.1500655937030045,
+      "grad_norm": 6.4375,
+      "learning_rate": 1.789424048501179e-05,
+      "loss": 0.1116,
+      "step": 4690
+    },
+    {
+      "epoch": 0.15038556298595335,
+      "grad_norm": 5.0625,
+      "learning_rate": 1.7887504210171777e-05,
+      "loss": 0.1069,
+      "step": 4700
+    },
+    {
+      "epoch": 0.15070553226890218,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.788076793533176e-05,
+      "loss": 0.0816,
+      "step": 4710
+    },
+    {
+      "epoch": 0.15102550155185102,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.787403166049175e-05,
+      "loss": 0.1824,
+      "step": 4720
+    },
+    {
+      "epoch": 0.15134547083479985,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.7867295385651736e-05,
+      "loss": 0.1502,
+      "step": 4730
+    },
+    {
+      "epoch": 0.15166544011774868,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.786055911081172e-05,
+      "loss": 0.1234,
+      "step": 4740
+    },
+    {
+      "epoch": 0.15198540940069755,
+      "grad_norm": 11.75,
+      "learning_rate": 1.785382283597171e-05,
+      "loss": 0.1784,
+      "step": 4750
+    },
+    {
+      "epoch": 0.15230537868364638,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.7847086561131693e-05,
+      "loss": 0.0886,
+      "step": 4760
+    },
+    {
+      "epoch": 0.15262534796659522,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.784035028629168e-05,
+      "loss": 0.1783,
+      "step": 4770
+    },
+    {
+      "epoch": 0.15294531724954405,
+      "grad_norm": 10.4375,
+      "learning_rate": 1.783361401145167e-05,
+      "loss": 0.1132,
+      "step": 4780
+    },
+    {
+      "epoch": 0.15326528653249288,
+      "grad_norm": 24.0,
+      "learning_rate": 1.7826877736611653e-05,
+      "loss": 0.1719,
+      "step": 4790
+    },
+    {
+      "epoch": 0.15358525581544172,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.7820141461771644e-05,
+      "loss": 0.1314,
+      "step": 4800
+    },
+    {
+      "epoch": 0.15390522509839055,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7813405186931628e-05,
+      "loss": 0.0683,
+      "step": 4810
+    },
+    {
+      "epoch": 0.1542251943813394,
+      "grad_norm": 7.5,
+      "learning_rate": 1.7806668912091616e-05,
+      "loss": 0.1135,
+      "step": 4820
+    },
+    {
+      "epoch": 0.15454516366428822,
+      "grad_norm": 18.125,
+      "learning_rate": 1.7799932637251604e-05,
+      "loss": 0.226,
+      "step": 4830
+    },
+    {
+      "epoch": 0.15486513294723706,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.7793196362411588e-05,
+      "loss": 0.096,
+      "step": 4840
+    },
+    {
+      "epoch": 0.1551851022301859,
+      "grad_norm": 34.5,
+      "learning_rate": 1.7786460087571576e-05,
+      "loss": 0.1353,
+      "step": 4850
+    },
+    {
+      "epoch": 0.15550507151313475,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.7779723812731563e-05,
+      "loss": 0.1998,
+      "step": 4860
+    },
+    {
+      "epoch": 0.15582504079608359,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.7772987537891548e-05,
+      "loss": 0.1153,
+      "step": 4870
+    },
+    {
+      "epoch": 0.15614501007903242,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.7766251263051536e-05,
+      "loss": 0.1819,
+      "step": 4880
+    },
+    {
+      "epoch": 0.15646497936198125,
+      "grad_norm": 4.5625,
+      "learning_rate": 1.775951498821152e-05,
+      "loss": 0.1191,
+      "step": 4890
+    },
+    {
+      "epoch": 0.1567849486449301,
+      "grad_norm": 40.5,
+      "learning_rate": 1.7752778713371508e-05,
+      "loss": 0.1025,
+      "step": 4900
+    },
+    {
+      "epoch": 0.15710491792787892,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7746042438531495e-05,
+      "loss": 0.0941,
+      "step": 4910
+    },
+    {
+      "epoch": 0.15742488721082776,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.773930616369148e-05,
+      "loss": 0.096,
+      "step": 4920
+    },
+    {
+      "epoch": 0.1577448564937766,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.7732569888851467e-05,
+      "loss": 0.2236,
+      "step": 4930
+    },
+    {
+      "epoch": 0.15806482577672543,
+      "grad_norm": 5.90625,
+      "learning_rate": 1.7725833614011455e-05,
+      "loss": 0.1338,
+      "step": 4940
+    },
+    {
+      "epoch": 0.15838479505967426,
+      "grad_norm": 8.875,
+      "learning_rate": 1.771909733917144e-05,
+      "loss": 0.1438,
+      "step": 4950
+    },
+    {
+      "epoch": 0.1587047643426231,
+      "grad_norm": 8.75,
+      "learning_rate": 1.7712361064331427e-05,
+      "loss": 0.0856,
+      "step": 4960
+    },
+    {
+      "epoch": 0.15902473362557196,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.7705624789491415e-05,
+      "loss": 0.1023,
+      "step": 4970
+    },
+    {
+      "epoch": 0.1593447029085208,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.76988885146514e-05,
+      "loss": 0.1126,
+      "step": 4980
+    },
+    {
+      "epoch": 0.15966467219146963,
+      "grad_norm": 9.75,
+      "learning_rate": 1.7692152239811387e-05,
+      "loss": 0.1572,
+      "step": 4990
+    },
+    {
+      "epoch": 0.15998464147441846,
+      "grad_norm": 9.0,
+      "learning_rate": 1.768541596497137e-05,
+      "loss": 0.0723,
+      "step": 5000
+    },
+    {
+      "epoch": 0.1603046107573673,
+      "grad_norm": 10.375,
+      "learning_rate": 1.767867969013136e-05,
+      "loss": 0.1359,
+      "step": 5010
+    },
+    {
+      "epoch": 0.16062458004031613,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.7671943415291347e-05,
+      "loss": 0.1012,
+      "step": 5020
+    },
+    {
+      "epoch": 0.16094454932326496,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.766520714045133e-05,
+      "loss": 0.1514,
+      "step": 5030
+    },
+    {
+      "epoch": 0.1612645186062138,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.765847086561132e-05,
+      "loss": 0.0964,
+      "step": 5040
+    },
+    {
+      "epoch": 0.16158448788916263,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.7651734590771307e-05,
+      "loss": 0.1257,
+      "step": 5050
+    },
+    {
+      "epoch": 0.16190445717211147,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.764499831593129e-05,
+      "loss": 0.0897,
+      "step": 5060
+    },
+    {
+      "epoch": 0.16222442645506033,
+      "grad_norm": 7.375,
+      "learning_rate": 1.763826204109128e-05,
+      "loss": 0.1221,
+      "step": 5070
+    },
+    {
+      "epoch": 0.16254439573800916,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.7631525766251267e-05,
+      "loss": 0.1566,
+      "step": 5080
+    },
+    {
+      "epoch": 0.162864365020958,
+      "grad_norm": 5.125,
+      "learning_rate": 1.762478949141125e-05,
+      "loss": 0.1231,
+      "step": 5090
+    },
+    {
+      "epoch": 0.16318433430390683,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.761805321657124e-05,
+      "loss": 0.1446,
+      "step": 5100
+    },
+    {
+      "epoch": 0.16350430358685566,
+      "grad_norm": 6.34375,
+      "learning_rate": 1.7611316941731223e-05,
+      "loss": 0.132,
+      "step": 5110
+    },
+    {
+      "epoch": 0.1638242728698045,
+      "grad_norm": 5.53125,
+      "learning_rate": 1.760458066689121e-05,
+      "loss": 0.1764,
+      "step": 5120
+    },
+    {
+      "epoch": 0.16414424215275333,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.75978443920512e-05,
+      "loss": 0.1182,
+      "step": 5130
+    },
+    {
+      "epoch": 0.16446421143570217,
+      "grad_norm": 2.875,
+      "learning_rate": 1.7591108117211183e-05,
+      "loss": 0.1019,
+      "step": 5140
+    },
+    {
+      "epoch": 0.164784180718651,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.758437184237117e-05,
+      "loss": 0.0536,
+      "step": 5150
+    },
+    {
+      "epoch": 0.16510415000159984,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.7577635567531158e-05,
+      "loss": 0.1366,
+      "step": 5160
+    },
+    {
+      "epoch": 0.16542411928454867,
+      "grad_norm": 11.125,
+      "learning_rate": 1.7570899292691143e-05,
+      "loss": 0.0998,
+      "step": 5170
+    },
+    {
+      "epoch": 0.16574408856749753,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.756416301785113e-05,
+      "loss": 0.1519,
+      "step": 5180
+    },
+    {
+      "epoch": 0.16606405785044637,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.7557426743011118e-05,
+      "loss": 0.1814,
+      "step": 5190
+    },
+    {
+      "epoch": 0.1663840271333952,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.7550690468171102e-05,
+      "loss": 0.1636,
+      "step": 5200
+    },
+    {
+      "epoch": 0.16670399641634404,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.754395419333109e-05,
+      "loss": 0.0829,
+      "step": 5210
+    },
+    {
+      "epoch": 0.16702396569929287,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.7537217918491078e-05,
+      "loss": 0.0843,
+      "step": 5220
+    },
+    {
+      "epoch": 0.1673439349822417,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.7530481643651062e-05,
+      "loss": 0.1332,
+      "step": 5230
+    },
+    {
+      "epoch": 0.16766390426519054,
+      "grad_norm": 10.875,
+      "learning_rate": 1.752374536881105e-05,
+      "loss": 0.1808,
+      "step": 5240
+    },
+    {
+      "epoch": 0.16798387354813937,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.7517009093971034e-05,
+      "loss": 0.0635,
+      "step": 5250
+    },
+    {
+      "epoch": 0.1683038428310882,
+      "grad_norm": 2.0,
+      "learning_rate": 1.7510272819131022e-05,
+      "loss": 0.1045,
+      "step": 5260
+    },
+    {
+      "epoch": 0.16862381211403704,
+      "grad_norm": 15.25,
+      "learning_rate": 1.750353654429101e-05,
+      "loss": 0.1292,
+      "step": 5270
+    },
+    {
+      "epoch": 0.16894378139698588,
+      "grad_norm": 4.25,
+      "learning_rate": 1.7496800269450994e-05,
+      "loss": 0.1462,
+      "step": 5280
+    },
+    {
+      "epoch": 0.16926375067993474,
+      "grad_norm": 4.78125,
+      "learning_rate": 1.7490063994610982e-05,
+      "loss": 0.0763,
+      "step": 5290
+    },
+    {
+      "epoch": 0.16958371996288357,
+      "grad_norm": 18.625,
+      "learning_rate": 1.748332771977097e-05,
+      "loss": 0.1251,
+      "step": 5300
+    },
+    {
+      "epoch": 0.1699036892458324,
+      "grad_norm": 22.25,
+      "learning_rate": 1.7476591444930954e-05,
+      "loss": 0.084,
+      "step": 5310
+    },
+    {
+      "epoch": 0.17022365852878124,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.7469855170090942e-05,
+      "loss": 0.1101,
+      "step": 5320
+    },
+    {
+      "epoch": 0.17054362781173008,
+      "grad_norm": 16.625,
+      "learning_rate": 1.746311889525093e-05,
+      "loss": 0.1222,
+      "step": 5330
+    },
+    {
+      "epoch": 0.1708635970946789,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.7456382620410914e-05,
+      "loss": 0.1125,
+      "step": 5340
+    },
+    {
+      "epoch": 0.17118356637762774,
+      "grad_norm": 5.8125,
+      "learning_rate": 1.74496463455709e-05,
+      "loss": 0.1216,
+      "step": 5350
+    },
+    {
+      "epoch": 0.17150353566057658,
+      "grad_norm": 25.5,
+      "learning_rate": 1.7442910070730886e-05,
+      "loss": 0.0724,
+      "step": 5360
+    },
+    {
+      "epoch": 0.1718235049435254,
+      "grad_norm": 23.875,
+      "learning_rate": 1.7436173795890874e-05,
+      "loss": 0.0847,
+      "step": 5370
+    },
+    {
+      "epoch": 0.17214347422647425,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.742943752105086e-05,
+      "loss": 0.1212,
+      "step": 5380
+    },
+    {
+      "epoch": 0.1724634435094231,
+      "grad_norm": 21.125,
+      "learning_rate": 1.7422701246210846e-05,
+      "loss": 0.1834,
+      "step": 5390
+    },
+    {
+      "epoch": 0.17278341279237194,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7415964971370833e-05,
+      "loss": 0.1676,
+      "step": 5400
+    },
+    {
+      "epoch": 0.17310338207532078,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.740922869653082e-05,
+      "loss": 0.107,
+      "step": 5410
+    },
+    {
+      "epoch": 0.1734233513582696,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.7402492421690806e-05,
+      "loss": 0.118,
+      "step": 5420
+    },
+    {
+      "epoch": 0.17374332064121845,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.7395756146850793e-05,
+      "loss": 0.1029,
+      "step": 5430
+    },
+    {
+      "epoch": 0.17406328992416728,
+      "grad_norm": 25.25,
+      "learning_rate": 1.738901987201078e-05,
+      "loss": 0.1236,
+      "step": 5440
+    },
+    {
+      "epoch": 0.17438325920711611,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.7382283597170765e-05,
+      "loss": 0.1597,
+      "step": 5450
+    },
+    {
+      "epoch": 0.17470322849006495,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.7375547322330753e-05,
+      "loss": 0.1246,
+      "step": 5460
+    },
+    {
+      "epoch": 0.17502319777301378,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.7368811047490737e-05,
+      "loss": 0.0773,
+      "step": 5470
+    },
+    {
+      "epoch": 0.17534316705596262,
+      "grad_norm": 5.21875,
+      "learning_rate": 1.7362074772650725e-05,
+      "loss": 0.1136,
+      "step": 5480
+    },
+    {
+      "epoch": 0.17566313633891145,
+      "grad_norm": 12.375,
+      "learning_rate": 1.7355338497810713e-05,
+      "loss": 0.0981,
+      "step": 5490
+    },
+    {
+      "epoch": 0.17598310562186031,
+      "grad_norm": 12.125,
+      "learning_rate": 1.7348602222970697e-05,
+      "loss": 0.1362,
+      "step": 5500
+    },
+    {
+      "epoch": 0.17630307490480915,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.7341865948130685e-05,
+      "loss": 0.1623,
+      "step": 5510
+    },
+    {
+      "epoch": 0.17662304418775798,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.7335129673290673e-05,
+      "loss": 0.1106,
+      "step": 5520
+    },
+    {
+      "epoch": 0.17694301347070682,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.7328393398450657e-05,
+      "loss": 0.1518,
+      "step": 5530
+    },
+    {
+      "epoch": 0.17726298275365565,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.7321657123610645e-05,
+      "loss": 0.1014,
+      "step": 5540
+    },
+    {
+      "epoch": 0.17758295203660449,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.7314920848770633e-05,
+      "loss": 0.0938,
+      "step": 5550
+    },
+    {
+      "epoch": 0.17790292131955332,
+      "grad_norm": 26.375,
+      "learning_rate": 1.7308184573930617e-05,
+      "loss": 0.0974,
+      "step": 5560
+    },
+    {
+      "epoch": 0.17822289060250215,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.7301448299090605e-05,
+      "loss": 0.105,
+      "step": 5570
+    },
+    {
+      "epoch": 0.178542859885451,
+      "grad_norm": 8.75,
+      "learning_rate": 1.729471202425059e-05,
+      "loss": 0.1591,
+      "step": 5580
+    },
+    {
+      "epoch": 0.17886282916839982,
+      "grad_norm": 9.125,
+      "learning_rate": 1.7287975749410577e-05,
+      "loss": 0.204,
+      "step": 5590
+    },
+    {
+      "epoch": 0.17918279845134866,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7281239474570564e-05,
+      "loss": 0.1568,
+      "step": 5600
+    },
+    {
+      "epoch": 0.17950276773429752,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.727450319973055e-05,
+      "loss": 0.172,
+      "step": 5610
+    },
+    {
+      "epoch": 0.17982273701724635,
+      "grad_norm": 12.125,
+      "learning_rate": 1.7267766924890537e-05,
+      "loss": 0.095,
+      "step": 5620
+    },
+    {
+      "epoch": 0.1801427063001952,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.7261030650050524e-05,
+      "loss": 0.1384,
+      "step": 5630
+    },
+    {
+      "epoch": 0.18046267558314402,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.725429437521051e-05,
+      "loss": 0.0906,
+      "step": 5640
+    },
+    {
+      "epoch": 0.18078264486609286,
+      "grad_norm": 2.75,
+      "learning_rate": 1.7247558100370496e-05,
+      "loss": 0.17,
+      "step": 5650
+    },
+    {
+      "epoch": 0.1811026141490417,
+      "grad_norm": 14.5,
+      "learning_rate": 1.7240821825530484e-05,
+      "loss": 0.1198,
+      "step": 5660
+    },
+    {
+      "epoch": 0.18142258343199053,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.723408555069047e-05,
+      "loss": 0.1267,
+      "step": 5670
+    },
+    {
+      "epoch": 0.18174255271493936,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.7227349275850456e-05,
+      "loss": 0.0867,
+      "step": 5680
+    },
+    {
+      "epoch": 0.1820625219978882,
+      "grad_norm": 4.75,
+      "learning_rate": 1.7220613001010444e-05,
+      "loss": 0.1107,
+      "step": 5690
+    },
+    {
+      "epoch": 0.18238249128083703,
+      "grad_norm": 71.5,
+      "learning_rate": 1.7213876726170428e-05,
+      "loss": 0.1062,
+      "step": 5700
+    },
+    {
+      "epoch": 0.1827024605637859,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.7207140451330416e-05,
+      "loss": 0.1304,
+      "step": 5710
+    },
+    {
+      "epoch": 0.18302242984673472,
+      "grad_norm": 27.75,
+      "learning_rate": 1.72004041764904e-05,
+      "loss": 0.1224,
+      "step": 5720
+    },
+    {
+      "epoch": 0.18334239912968356,
+      "grad_norm": 20.375,
+      "learning_rate": 1.7193667901650388e-05,
+      "loss": 0.1224,
+      "step": 5730
+    },
+    {
+      "epoch": 0.1836623684126324,
+      "grad_norm": 6.53125,
+      "learning_rate": 1.7186931626810376e-05,
+      "loss": 0.1302,
+      "step": 5740
+    },
+    {
+      "epoch": 0.18398233769558123,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.718019535197036e-05,
+      "loss": 0.104,
+      "step": 5750
+    },
+    {
+      "epoch": 0.18430230697853006,
+      "grad_norm": 5.78125,
+      "learning_rate": 1.7173459077130348e-05,
+      "loss": 0.1589,
+      "step": 5760
+    },
+    {
+      "epoch": 0.1846222762614789,
+      "grad_norm": 13.625,
+      "learning_rate": 1.7166722802290336e-05,
+      "loss": 0.1316,
+      "step": 5770
+    },
+    {
+      "epoch": 0.18494224554442773,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.715998652745032e-05,
+      "loss": 0.1645,
+      "step": 5780
+    },
+    {
+      "epoch": 0.18526221482737656,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.7153250252610308e-05,
+      "loss": 0.1323,
+      "step": 5790
+    },
+    {
+      "epoch": 0.1855821841103254,
+      "grad_norm": 6.875,
+      "learning_rate": 1.7146513977770295e-05,
+      "loss": 0.0834,
+      "step": 5800
+    },
+    {
+      "epoch": 0.18590215339327423,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.713977770293028e-05,
+      "loss": 0.075,
+      "step": 5810
+    },
+    {
+      "epoch": 0.1862221226762231,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.7133041428090268e-05,
+      "loss": 0.122,
+      "step": 5820
+    },
+    {
+      "epoch": 0.18654209195917193,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.7126305153250252e-05,
+      "loss": 0.0694,
+      "step": 5830
+    },
+    {
+      "epoch": 0.18686206124212076,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.711956887841024e-05,
+      "loss": 0.1691,
+      "step": 5840
+    },
+    {
+      "epoch": 0.1871820305250696,
+      "grad_norm": 6.625,
+      "learning_rate": 1.7112832603570227e-05,
+      "loss": 0.1055,
+      "step": 5850
+    },
+    {
+      "epoch": 0.18750199980801843,
+      "grad_norm": 5.1875,
+      "learning_rate": 1.710609632873021e-05,
+      "loss": 0.0935,
+      "step": 5860
+    },
+    {
+      "epoch": 0.18782196909096727,
+      "grad_norm": 23.75,
+      "learning_rate": 1.70993600538902e-05,
+      "loss": 0.1484,
+      "step": 5870
+    },
+    {
+      "epoch": 0.1881419383739161,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.7092623779050187e-05,
+      "loss": 0.1102,
+      "step": 5880
+    },
+    {
+      "epoch": 0.18846190765686494,
+      "grad_norm": 9.75,
+      "learning_rate": 1.708588750421017e-05,
+      "loss": 0.1075,
+      "step": 5890
+    },
+    {
+      "epoch": 0.18878187693981377,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.707915122937016e-05,
+      "loss": 0.1185,
+      "step": 5900
+    },
+    {
+      "epoch": 0.1891018462227626,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7072414954530147e-05,
+      "loss": 0.1335,
+      "step": 5910
+    },
+    {
+      "epoch": 0.18942181550571144,
+      "grad_norm": 5.375,
+      "learning_rate": 1.706567867969013e-05,
+      "loss": 0.0948,
+      "step": 5920
+    },
+    {
+      "epoch": 0.1897417847886603,
+      "grad_norm": 13.75,
+      "learning_rate": 1.705894240485012e-05,
+      "loss": 0.1593,
+      "step": 5930
+    },
+    {
+      "epoch": 0.19006175407160913,
+      "grad_norm": 17.625,
+      "learning_rate": 1.7052206130010103e-05,
+      "loss": 0.1549,
+      "step": 5940
+    },
+    {
+      "epoch": 0.19038172335455797,
+      "grad_norm": 5.0,
+      "learning_rate": 1.704546985517009e-05,
+      "loss": 0.1825,
+      "step": 5950
+    },
+    {
+      "epoch": 0.1907016926375068,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.703873358033008e-05,
+      "loss": 0.1993,
+      "step": 5960
+    },
+    {
+      "epoch": 0.19102166192045564,
+      "grad_norm": 10.625,
+      "learning_rate": 1.7031997305490063e-05,
+      "loss": 0.1825,
+      "step": 5970
+    },
+    {
+      "epoch": 0.19134163120340447,
+      "grad_norm": 7.375,
+      "learning_rate": 1.702526103065005e-05,
+      "loss": 0.1268,
+      "step": 5980
+    },
+    {
+      "epoch": 0.1916616004863533,
+      "grad_norm": 5.59375,
+      "learning_rate": 1.701852475581004e-05,
+      "loss": 0.0864,
+      "step": 5990
+    },
+    {
+      "epoch": 0.19198156976930214,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.7011788480970023e-05,
+      "loss": 0.1339,
+      "step": 6000
+    },
+    {
+      "epoch": 0.19230153905225097,
+      "grad_norm": 24.625,
+      "learning_rate": 1.700505220613001e-05,
+      "loss": 0.1293,
+      "step": 6010
+    },
+    {
+      "epoch": 0.1926215083351998,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.699831593129e-05,
+      "loss": 0.1562,
+      "step": 6020
+    },
+    {
+      "epoch": 0.19294147761814867,
+      "grad_norm": 4.40625,
+      "learning_rate": 1.6991579656449983e-05,
+      "loss": 0.1033,
+      "step": 6030
+    },
+    {
+      "epoch": 0.1932614469010975,
+      "grad_norm": 9.5,
+      "learning_rate": 1.698484338160997e-05,
+      "loss": 0.2078,
+      "step": 6040
+    },
+    {
+      "epoch": 0.19358141618404634,
+      "grad_norm": 11.375,
+      "learning_rate": 1.6978107106769955e-05,
+      "loss": 0.1351,
+      "step": 6050
+    },
+    {
+      "epoch": 0.19390138546699517,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.6971370831929943e-05,
+      "loss": 0.119,
+      "step": 6060
+    },
+    {
+      "epoch": 0.194221354749944,
+      "grad_norm": 20.75,
+      "learning_rate": 1.696463455708993e-05,
+      "loss": 0.1564,
+      "step": 6070
+    },
+    {
+      "epoch": 0.19454132403289284,
+      "grad_norm": 11.4375,
+      "learning_rate": 1.6957898282249915e-05,
+      "loss": 0.1463,
+      "step": 6080
+    },
+    {
+      "epoch": 0.19486129331584168,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.6951162007409906e-05,
+      "loss": 0.0718,
+      "step": 6090
+    },
+    {
+      "epoch": 0.1951812625987905,
+      "grad_norm": 1.125,
+      "learning_rate": 1.694442573256989e-05,
+      "loss": 0.0924,
+      "step": 6100
+    },
+    {
+      "epoch": 0.19550123188173935,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.6937689457729878e-05,
+      "loss": 0.1603,
+      "step": 6110
+    },
+    {
+      "epoch": 0.19582120116468818,
+      "grad_norm": 23.25,
+      "learning_rate": 1.6930953182889866e-05,
+      "loss": 0.0743,
+      "step": 6120
+    },
+    {
+      "epoch": 0.19614117044763701,
+      "grad_norm": 10.4375,
+      "learning_rate": 1.692421690804985e-05,
+      "loss": 0.0738,
+      "step": 6130
+    },
+    {
+      "epoch": 0.19646113973058588,
+      "grad_norm": 39.5,
+      "learning_rate": 1.6917480633209838e-05,
+      "loss": 0.1504,
+      "step": 6140
+    },
+    {
+      "epoch": 0.1967811090135347,
+      "grad_norm": 5.71875,
+      "learning_rate": 1.6910744358369826e-05,
+      "loss": 0.1192,
+      "step": 6150
+    },
+    {
+      "epoch": 0.19710107829648355,
+      "grad_norm": 8.375,
+      "learning_rate": 1.690400808352981e-05,
+      "loss": 0.1748,
+      "step": 6160
+    },
+    {
+      "epoch": 0.19742104757943238,
+      "grad_norm": 12.625,
+      "learning_rate": 1.6897271808689798e-05,
+      "loss": 0.0997,
+      "step": 6170
+    },
+    {
+      "epoch": 0.1977410168623812,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.6890535533849782e-05,
+      "loss": 0.1513,
+      "step": 6180
+    },
+    {
+      "epoch": 0.19806098614533005,
+      "grad_norm": 5.1875,
+      "learning_rate": 1.688379925900977e-05,
+      "loss": 0.1493,
+      "step": 6190
+    },
+    {
+      "epoch": 0.19838095542827888,
+      "grad_norm": 2.5,
+      "learning_rate": 1.6877062984169757e-05,
+      "loss": 0.0716,
+      "step": 6200
+    },
+    {
+      "epoch": 0.19870092471122772,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.6870326709329742e-05,
+      "loss": 0.1744,
+      "step": 6210
+    },
+    {
+      "epoch": 0.19902089399417655,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.686359043448973e-05,
+      "loss": 0.0735,
+      "step": 6220
+    },
+    {
+      "epoch": 0.19934086327712539,
+      "grad_norm": 4.5,
+      "learning_rate": 1.6856854159649717e-05,
+      "loss": 0.1173,
+      "step": 6230
+    },
+    {
+      "epoch": 0.19966083256007422,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.68501178848097e-05,
+      "loss": 0.0549,
+      "step": 6240
+    },
+    {
+      "epoch": 0.19998080184302308,
+      "grad_norm": 4.375,
+      "learning_rate": 1.684338160996969e-05,
+      "loss": 0.0844,
+      "step": 6250
+    },
+    {
+      "epoch": 0.20030077112597192,
+      "grad_norm": 13.625,
+      "learning_rate": 1.6836645335129677e-05,
+      "loss": 0.1093,
+      "step": 6260
+    },
+    {
+      "epoch": 0.20062074040892075,
+      "grad_norm": 6.40625,
+      "learning_rate": 1.682990906028966e-05,
+      "loss": 0.1535,
+      "step": 6270
+    },
+    {
+      "epoch": 0.20094070969186958,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.682317278544965e-05,
+      "loss": 0.1913,
+      "step": 6280
+    },
+    {
+      "epoch": 0.20126067897481842,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6816436510609633e-05,
+      "loss": 0.1162,
+      "step": 6290
+    },
+    {
+      "epoch": 0.20158064825776725,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.680970023576962e-05,
+      "loss": 0.1074,
+      "step": 6300
+    },
+    {
+      "epoch": 0.2019006175407161,
+      "grad_norm": 15.625,
+      "learning_rate": 1.680296396092961e-05,
+      "loss": 0.0973,
+      "step": 6310
+    },
+    {
+      "epoch": 0.20222058682366492,
+      "grad_norm": 23.125,
+      "learning_rate": 1.6796227686089593e-05,
+      "loss": 0.1102,
+      "step": 6320
+    },
+    {
+      "epoch": 0.20254055610661376,
+      "grad_norm": 5.3125,
+      "learning_rate": 1.678949141124958e-05,
+      "loss": 0.1171,
+      "step": 6330
+    },
+    {
+      "epoch": 0.2028605253895626,
+      "grad_norm": 6.34375,
+      "learning_rate": 1.678275513640957e-05,
+      "loss": 0.1504,
+      "step": 6340
+    },
+    {
+      "epoch": 0.20318049467251145,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.6776018861569553e-05,
+      "loss": 0.0638,
+      "step": 6350
+    },
+    {
+      "epoch": 0.2035004639554603,
+      "grad_norm": 7.625,
+      "learning_rate": 1.676928258672954e-05,
+      "loss": 0.1289,
+      "step": 6360
+    },
+    {
+      "epoch": 0.20382043323840912,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.676254631188953e-05,
+      "loss": 0.1183,
+      "step": 6370
+    },
+    {
+      "epoch": 0.20414040252135796,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.6755810037049513e-05,
+      "loss": 0.124,
+      "step": 6380
+    },
+    {
+      "epoch": 0.2044603718043068,
+      "grad_norm": 16.25,
+      "learning_rate": 1.67490737622095e-05,
+      "loss": 0.1014,
+      "step": 6390
+    },
+    {
+      "epoch": 0.20478034108725562,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.6742337487369485e-05,
+      "loss": 0.1125,
+      "step": 6400
+    },
+    {
+      "epoch": 0.20510031037020446,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.6735601212529473e-05,
+      "loss": 0.1733,
+      "step": 6410
+    },
+    {
+      "epoch": 0.2054202796531533,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.672886493768946e-05,
+      "loss": 0.1378,
+      "step": 6420
+    },
+    {
+      "epoch": 0.20574024893610213,
+      "grad_norm": 6.03125,
+      "learning_rate": 1.6722128662849445e-05,
+      "loss": 0.1147,
+      "step": 6430
+    },
+    {
+      "epoch": 0.20606021821905096,
+      "grad_norm": 19.5,
+      "learning_rate": 1.6715392388009433e-05,
+      "loss": 0.0789,
+      "step": 6440
+    },
+    {
+      "epoch": 0.2063801875019998,
+      "grad_norm": 6.0625,
+      "learning_rate": 1.670865611316942e-05,
+      "loss": 0.1286,
+      "step": 6450
+    },
+    {
+      "epoch": 0.20670015678494866,
+      "grad_norm": 21.125,
+      "learning_rate": 1.6701919838329405e-05,
+      "loss": 0.0632,
+      "step": 6460
+    },
+    {
+      "epoch": 0.2070201260678975,
+      "grad_norm": 16.125,
+      "learning_rate": 1.6695183563489392e-05,
+      "loss": 0.1671,
+      "step": 6470
+    },
+    {
+      "epoch": 0.20734009535084633,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.668844728864938e-05,
+      "loss": 0.1409,
+      "step": 6480
+    },
+    {
+      "epoch": 0.20766006463379516,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.6681711013809364e-05,
+      "loss": 0.1769,
+      "step": 6490
+    },
+    {
+      "epoch": 0.207980033916744,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.6674974738969352e-05,
+      "loss": 0.1058,
+      "step": 6500
+    },
+    {
+      "epoch": 0.20830000319969283,
+      "grad_norm": 13.75,
+      "learning_rate": 1.6668238464129337e-05,
+      "loss": 0.1337,
+      "step": 6510
+    },
+    {
+      "epoch": 0.20861997248264166,
+      "grad_norm": 17.125,
+      "learning_rate": 1.6661502189289324e-05,
+      "loss": 0.131,
+      "step": 6520
+    },
+    {
+      "epoch": 0.2089399417655905,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.6654765914449312e-05,
+      "loss": 0.0857,
+      "step": 6530
+    },
+    {
+      "epoch": 0.20925991104853933,
+      "grad_norm": 16.0,
+      "learning_rate": 1.6648029639609296e-05,
+      "loss": 0.1194,
+      "step": 6540
+    },
+    {
+      "epoch": 0.20957988033148817,
+      "grad_norm": 7.0,
+      "learning_rate": 1.6641293364769284e-05,
+      "loss": 0.1127,
+      "step": 6550
+    },
+    {
+      "epoch": 0.209899849614437,
+      "grad_norm": 10.125,
+      "learning_rate": 1.6634557089929272e-05,
+      "loss": 0.1289,
+      "step": 6560
+    },
+    {
+      "epoch": 0.21021981889738586,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.6627820815089256e-05,
+      "loss": 0.09,
+      "step": 6570
+    },
+    {
+      "epoch": 0.2105397881803347,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.6621084540249244e-05,
+      "loss": 0.0909,
+      "step": 6580
+    },
+    {
+      "epoch": 0.21085975746328353,
+      "grad_norm": 23.75,
+      "learning_rate": 1.661434826540923e-05,
+      "loss": 0.1099,
+      "step": 6590
+    },
+    {
+      "epoch": 0.21117972674623237,
+      "grad_norm": 4.84375,
+      "learning_rate": 1.6607611990569216e-05,
+      "loss": 0.1152,
+      "step": 6600
+    },
+    {
+      "epoch": 0.2114996960291812,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.6600875715729204e-05,
+      "loss": 0.164,
+      "step": 6610
+    },
+    {
+      "epoch": 0.21181966531213003,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.659413944088919e-05,
+      "loss": 0.14,
+      "step": 6620
+    },
+    {
+      "epoch": 0.21213963459507887,
+      "grad_norm": 13.75,
+      "learning_rate": 1.6587403166049176e-05,
+      "loss": 0.1046,
+      "step": 6630
+    },
+    {
+      "epoch": 0.2124596038780277,
+      "grad_norm": 13.75,
+      "learning_rate": 1.6580666891209164e-05,
+      "loss": 0.0868,
+      "step": 6640
+    },
+    {
+      "epoch": 0.21277957316097654,
+      "grad_norm": 5.5625,
+      "learning_rate": 1.6573930616369148e-05,
+      "loss": 0.2311,
+      "step": 6650
+    },
+    {
+      "epoch": 0.21309954244392537,
+      "grad_norm": 15.375,
+      "learning_rate": 1.6567194341529136e-05,
+      "loss": 0.1174,
+      "step": 6660
+    },
+    {
+      "epoch": 0.21341951172687423,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.6560458066689123e-05,
+      "loss": 0.0757,
+      "step": 6670
+    },
+    {
+      "epoch": 0.21373948100982307,
+      "grad_norm": 5.5625,
+      "learning_rate": 1.6553721791849108e-05,
+      "loss": 0.1379,
+      "step": 6680
+    },
+    {
+      "epoch": 0.2140594502927719,
+      "grad_norm": 7.25,
+      "learning_rate": 1.6546985517009095e-05,
+      "loss": 0.1872,
+      "step": 6690
+    },
+    {
+      "epoch": 0.21437941957572074,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.6540249242169083e-05,
+      "loss": 0.115,
+      "step": 6700
+    },
+    {
+      "epoch": 0.21469938885866957,
+      "grad_norm": 12.375,
+      "learning_rate": 1.6533512967329068e-05,
+      "loss": 0.1369,
+      "step": 6710
+    },
+    {
+      "epoch": 0.2150193581416184,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.6526776692489055e-05,
+      "loss": 0.1265,
+      "step": 6720
+    },
+    {
+      "epoch": 0.21533932742456724,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.6520040417649043e-05,
+      "loss": 0.0941,
+      "step": 6730
+    },
+    {
+      "epoch": 0.21565929670751607,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.6513304142809027e-05,
+      "loss": 0.153,
+      "step": 6740
+    },
+    {
+      "epoch": 0.2159792659904649,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.6506567867969015e-05,
+      "loss": 0.0875,
+      "step": 6750
+    },
+    {
+      "epoch": 0.21629923527341374,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.6499831593129e-05,
+      "loss": 0.0803,
+      "step": 6760
+    },
+    {
+      "epoch": 0.21661920455636258,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.6493095318288987e-05,
+      "loss": 0.0814,
+      "step": 6770
+    },
+    {
+      "epoch": 0.21693917383931144,
+      "grad_norm": 27.0,
+      "learning_rate": 1.6486359043448975e-05,
+      "loss": 0.1629,
+      "step": 6780
+    },
+    {
+      "epoch": 0.21725914312226027,
+      "grad_norm": 6.25,
+      "learning_rate": 1.647962276860896e-05,
+      "loss": 0.1529,
+      "step": 6790
+    },
+    {
+      "epoch": 0.2175791124052091,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.6472886493768947e-05,
+      "loss": 0.1408,
+      "step": 6800
+    },
+    {
+      "epoch": 0.21789908168815794,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.6466150218928935e-05,
+      "loss": 0.1601,
+      "step": 6810
+    },
+    {
+      "epoch": 0.21821905097110678,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.645941394408892e-05,
+      "loss": 0.1226,
+      "step": 6820
+    },
+    {
+      "epoch": 0.2185390202540556,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.6452677669248907e-05,
+      "loss": 0.0844,
+      "step": 6830
+    },
+    {
+      "epoch": 0.21885898953700444,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.6445941394408895e-05,
+      "loss": 0.248,
+      "step": 6840
+    },
+    {
+      "epoch": 0.21917895881995328,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.643920511956888e-05,
+      "loss": 0.0982,
+      "step": 6850
+    },
+    {
+      "epoch": 0.2194989281029021,
+      "grad_norm": 13.0,
+      "learning_rate": 1.6432468844728867e-05,
+      "loss": 0.117,
+      "step": 6860
+    },
+    {
+      "epoch": 0.21981889738585095,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.642573256988885e-05,
+      "loss": 0.1546,
+      "step": 6870
+    },
+    {
+      "epoch": 0.22013886666879978,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.641899629504884e-05,
+      "loss": 0.1001,
+      "step": 6880
+    },
+    {
+      "epoch": 0.22045883595174864,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6412260020208826e-05,
+      "loss": 0.1154,
+      "step": 6890
+    },
+    {
+      "epoch": 0.22077880523469748,
+      "grad_norm": 17.0,
+      "learning_rate": 1.640552374536881e-05,
+      "loss": 0.2123,
+      "step": 6900
+    },
+    {
+      "epoch": 0.2210987745176463,
+      "grad_norm": 8.875,
+      "learning_rate": 1.63987874705288e-05,
+      "loss": 0.1526,
+      "step": 6910
+    },
+    {
+      "epoch": 0.22141874380059515,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.6392051195688786e-05,
+      "loss": 0.1026,
+      "step": 6920
+    },
+    {
+      "epoch": 0.22173871308354398,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.638531492084877e-05,
+      "loss": 0.11,
+      "step": 6930
+    },
+    {
+      "epoch": 0.22205868236649282,
+      "grad_norm": 18.0,
+      "learning_rate": 1.637857864600876e-05,
+      "loss": 0.1496,
+      "step": 6940
+    },
+    {
+      "epoch": 0.22237865164944165,
+      "grad_norm": 9.25,
+      "learning_rate": 1.6371842371168746e-05,
+      "loss": 0.1105,
+      "step": 6950
+    },
+    {
+      "epoch": 0.22269862093239048,
+      "grad_norm": 15.875,
+      "learning_rate": 1.636510609632873e-05,
+      "loss": 0.1363,
+      "step": 6960
+    },
+    {
+      "epoch": 0.22301859021533932,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.6358369821488718e-05,
+      "loss": 0.0606,
+      "step": 6970
+    },
+    {
+      "epoch": 0.22333855949828815,
+      "grad_norm": 23.75,
+      "learning_rate": 1.6351633546648703e-05,
+      "loss": 0.0749,
+      "step": 6980
+    },
+    {
+      "epoch": 0.22365852878123701,
+      "grad_norm": 22.375,
+      "learning_rate": 1.634489727180869e-05,
+      "loss": 0.1165,
+      "step": 6990
+    },
+    {
+      "epoch": 0.22397849806418585,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.6338160996968678e-05,
+      "loss": 0.1138,
+      "step": 7000
+    },
+    {
+      "epoch": 0.22429846734713468,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.6331424722128662e-05,
+      "loss": 0.1439,
+      "step": 7010
+    },
+    {
+      "epoch": 0.22461843663008352,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.632468844728865e-05,
+      "loss": 0.0958,
+      "step": 7020
+    },
+    {
+      "epoch": 0.22493840591303235,
+      "grad_norm": 6.375,
+      "learning_rate": 1.6317952172448638e-05,
+      "loss": 0.2063,
+      "step": 7030
+    },
+    {
+      "epoch": 0.2252583751959812,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.6311215897608622e-05,
+      "loss": 0.1988,
+      "step": 7040
+    },
+    {
+      "epoch": 0.22557834447893002,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.630447962276861e-05,
+      "loss": 0.1764,
+      "step": 7050
+    },
+    {
+      "epoch": 0.22589831376187886,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.6297743347928598e-05,
+      "loss": 0.133,
+      "step": 7060
+    },
+    {
+      "epoch": 0.2262182830448277,
+      "grad_norm": 4.71875,
+      "learning_rate": 1.6291007073088582e-05,
+      "loss": 0.1223,
+      "step": 7070
+    },
+    {
+      "epoch": 0.22653825232777652,
+      "grad_norm": 4.5625,
+      "learning_rate": 1.628427079824857e-05,
+      "loss": 0.1131,
+      "step": 7080
+    },
+    {
+      "epoch": 0.22685822161072536,
+      "grad_norm": 8.625,
+      "learning_rate": 1.6277534523408557e-05,
+      "loss": 0.1424,
+      "step": 7090
+    },
+    {
+      "epoch": 0.22717819089367422,
+      "grad_norm": 11.375,
+      "learning_rate": 1.6270798248568542e-05,
+      "loss": 0.0843,
+      "step": 7100
+    },
+    {
+      "epoch": 0.22749816017662305,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.626406197372853e-05,
+      "loss": 0.1195,
+      "step": 7110
+    },
+    {
+      "epoch": 0.2278181294595719,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.6257325698888514e-05,
+      "loss": 0.1178,
+      "step": 7120
+    },
+    {
+      "epoch": 0.22813809874252072,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.62505894240485e-05,
+      "loss": 0.0507,
+      "step": 7130
+    },
+    {
+      "epoch": 0.22845806802546956,
+      "grad_norm": 10.25,
+      "learning_rate": 1.624385314920849e-05,
+      "loss": 0.1338,
+      "step": 7140
+    },
+    {
+      "epoch": 0.2287780373084184,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.6237116874368474e-05,
+      "loss": 0.1333,
+      "step": 7150
+    },
+    {
+      "epoch": 0.22909800659136723,
+      "grad_norm": 6.875,
+      "learning_rate": 1.623038059952846e-05,
+      "loss": 0.1183,
+      "step": 7160
+    },
+    {
+      "epoch": 0.22941797587431606,
+      "grad_norm": 9.375,
+      "learning_rate": 1.622364432468845e-05,
+      "loss": 0.1329,
+      "step": 7170
+    },
+    {
+      "epoch": 0.2297379451572649,
+      "grad_norm": 12.375,
+      "learning_rate": 1.6216908049848433e-05,
+      "loss": 0.1709,
+      "step": 7180
+    },
+    {
+      "epoch": 0.23005791444021373,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.621017177500842e-05,
+      "loss": 0.1209,
+      "step": 7190
+    },
+    {
+      "epoch": 0.23037788372316256,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.620343550016841e-05,
+      "loss": 0.1131,
+      "step": 7200
+    },
+    {
+      "epoch": 0.23069785300611143,
+      "grad_norm": 13.125,
+      "learning_rate": 1.6196699225328393e-05,
+      "loss": 0.1538,
+      "step": 7210
+    },
+    {
+      "epoch": 0.23101782228906026,
+      "grad_norm": 4.46875,
+      "learning_rate": 1.618996295048838e-05,
+      "loss": 0.1317,
+      "step": 7220
+    },
+    {
+      "epoch": 0.2313377915720091,
+      "grad_norm": 4.78125,
+      "learning_rate": 1.6183226675648365e-05,
+      "loss": 0.1416,
+      "step": 7230
+    },
+    {
+      "epoch": 0.23165776085495793,
+      "grad_norm": 16.375,
+      "learning_rate": 1.6176490400808353e-05,
+      "loss": 0.0833,
+      "step": 7240
+    },
+    {
+      "epoch": 0.23197773013790676,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.616975412596834e-05,
+      "loss": 0.0988,
+      "step": 7250
+    },
+    {
+      "epoch": 0.2322976994208556,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.6163017851128325e-05,
+      "loss": 0.137,
+      "step": 7260
+    },
+    {
+      "epoch": 0.23261766870380443,
+      "grad_norm": 15.5,
+      "learning_rate": 1.6156281576288313e-05,
+      "loss": 0.1676,
+      "step": 7270
+    },
+    {
+      "epoch": 0.23293763798675327,
+      "grad_norm": 20.875,
+      "learning_rate": 1.61495453014483e-05,
+      "loss": 0.1127,
+      "step": 7280
+    },
+    {
+      "epoch": 0.2332576072697021,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.6142809026608285e-05,
+      "loss": 0.111,
+      "step": 7290
+    },
+    {
+      "epoch": 0.23357757655265093,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.6136072751768273e-05,
+      "loss": 0.1173,
+      "step": 7300
+    },
+    {
+      "epoch": 0.2338975458355998,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.612933647692826e-05,
+      "loss": 0.0615,
+      "step": 7310
+    },
+    {
+      "epoch": 0.23421751511854863,
+      "grad_norm": 5.09375,
+      "learning_rate": 1.6122600202088245e-05,
+      "loss": 0.1078,
+      "step": 7320
+    },
+    {
+      "epoch": 0.23453748440149746,
+      "grad_norm": 4.46875,
+      "learning_rate": 1.6115863927248233e-05,
+      "loss": 0.0745,
+      "step": 7330
+    },
+    {
+      "epoch": 0.2348574536844463,
+      "grad_norm": 4.40625,
+      "learning_rate": 1.6109127652408217e-05,
+      "loss": 0.1212,
+      "step": 7340
+    },
+    {
+      "epoch": 0.23517742296739513,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.6102391377568205e-05,
+      "loss": 0.0647,
+      "step": 7350
+    },
+    {
+      "epoch": 0.23549739225034397,
+      "grad_norm": 17.0,
+      "learning_rate": 1.6095655102728192e-05,
+      "loss": 0.2942,
+      "step": 7360
+    },
+    {
+      "epoch": 0.2358173615332928,
+      "grad_norm": 6.125,
+      "learning_rate": 1.6088918827888177e-05,
+      "loss": 0.1488,
+      "step": 7370
+    },
+    {
+      "epoch": 0.23613733081624164,
+      "grad_norm": 12.0,
+      "learning_rate": 1.6082182553048168e-05,
+      "loss": 0.0842,
+      "step": 7380
+    },
+    {
+      "epoch": 0.23645730009919047,
+      "grad_norm": 11.25,
+      "learning_rate": 1.6075446278208152e-05,
+      "loss": 0.1328,
+      "step": 7390
+    },
+    {
+      "epoch": 0.2367772693821393,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.606871000336814e-05,
+      "loss": 0.1691,
+      "step": 7400
+    },
+    {
+      "epoch": 0.23709723866508814,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.6061973728528128e-05,
+      "loss": 0.1522,
+      "step": 7410
+    },
+    {
+      "epoch": 0.237417207948037,
+      "grad_norm": 10.625,
+      "learning_rate": 1.6055237453688112e-05,
+      "loss": 0.075,
+      "step": 7420
+    },
+    {
+      "epoch": 0.23773717723098584,
+      "grad_norm": 0.625,
+      "learning_rate": 1.60485011788481e-05,
+      "loss": 0.0629,
+      "step": 7430
+    },
+    {
+      "epoch": 0.23805714651393467,
+      "grad_norm": 27.375,
+      "learning_rate": 1.6041764904008088e-05,
+      "loss": 0.0742,
+      "step": 7440
+    },
+    {
+      "epoch": 0.2383771157968835,
+      "grad_norm": 17.75,
+      "learning_rate": 1.6035028629168072e-05,
+      "loss": 0.1081,
+      "step": 7450
+    },
+    {
+      "epoch": 0.23869708507983234,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.602829235432806e-05,
+      "loss": 0.2006,
+      "step": 7460
+    },
+    {
+      "epoch": 0.23901705436278117,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.6021556079488044e-05,
+      "loss": 0.0896,
+      "step": 7470
+    },
+    {
+      "epoch": 0.23933702364573,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.601481980464803e-05,
+      "loss": 0.0886,
+      "step": 7480
+    },
+    {
+      "epoch": 0.23965699292867884,
+      "grad_norm": 41.5,
+      "learning_rate": 1.600808352980802e-05,
+      "loss": 0.1366,
+      "step": 7490
+    },
+    {
+      "epoch": 0.23997696221162768,
+      "grad_norm": 7.0,
+      "learning_rate": 1.6001347254968004e-05,
+      "loss": 0.0731,
+      "step": 7500
+    },
+    {
+      "epoch": 0.2402969314945765,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.599461098012799e-05,
+      "loss": 0.1803,
+      "step": 7510
+    },
+    {
+      "epoch": 0.24061690077752534,
+      "grad_norm": 17.125,
+      "learning_rate": 1.598787470528798e-05,
+      "loss": 0.1282,
+      "step": 7520
+    },
+    {
+      "epoch": 0.2409368700604742,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.5981138430447964e-05,
+      "loss": 0.1225,
+      "step": 7530
+    },
+    {
+      "epoch": 0.24125683934342304,
+      "grad_norm": 11.75,
+      "learning_rate": 1.597440215560795e-05,
+      "loss": 0.1015,
+      "step": 7540
+    },
+    {
+      "epoch": 0.24157680862637188,
+      "grad_norm": 9.375,
+      "learning_rate": 1.596766588076794e-05,
+      "loss": 0.0981,
+      "step": 7550
+    },
+    {
+      "epoch": 0.2418967779093207,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.5960929605927923e-05,
+      "loss": 0.1125,
+      "step": 7560
+    },
+    {
+      "epoch": 0.24221674719226954,
+      "grad_norm": 11.25,
+      "learning_rate": 1.595419333108791e-05,
+      "loss": 0.171,
+      "step": 7570
+    },
+    {
+      "epoch": 0.24253671647521838,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.5947457056247895e-05,
+      "loss": 0.0962,
+      "step": 7580
+    },
+    {
+      "epoch": 0.2428566857581672,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.5940720781407883e-05,
+      "loss": 0.0968,
+      "step": 7590
+    },
+    {
+      "epoch": 0.24317665504111605,
+      "grad_norm": 17.625,
+      "learning_rate": 1.593398450656787e-05,
+      "loss": 0.0923,
+      "step": 7600
+    },
+    {
+      "epoch": 0.24349662432406488,
+      "grad_norm": 5.0625,
+      "learning_rate": 1.5927248231727855e-05,
+      "loss": 0.1091,
+      "step": 7610
+    },
+    {
+      "epoch": 0.24381659360701372,
+      "grad_norm": 20.875,
+      "learning_rate": 1.5920511956887843e-05,
+      "loss": 0.149,
+      "step": 7620
+    },
+    {
+      "epoch": 0.24413656288996258,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.591377568204783e-05,
+      "loss": 0.1362,
+      "step": 7630
+    },
+    {
+      "epoch": 0.2444565321729114,
+      "grad_norm": 16.875,
+      "learning_rate": 1.5907039407207815e-05,
+      "loss": 0.172,
+      "step": 7640
+    },
+    {
+      "epoch": 0.24477650145586025,
+      "grad_norm": 11.25,
+      "learning_rate": 1.5900303132367803e-05,
+      "loss": 0.1634,
+      "step": 7650
+    },
+    {
+      "epoch": 0.24509647073880908,
+      "grad_norm": 10.4375,
+      "learning_rate": 1.589356685752779e-05,
+      "loss": 0.0898,
+      "step": 7660
+    },
+    {
+      "epoch": 0.24541644002175791,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.5886830582687775e-05,
+      "loss": 0.2014,
+      "step": 7670
+    },
+    {
+      "epoch": 0.24573640930470675,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.5880094307847763e-05,
+      "loss": 0.1195,
+      "step": 7680
+    },
+    {
+      "epoch": 0.24605637858765558,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.5873358033007747e-05,
+      "loss": 0.0837,
+      "step": 7690
+    },
+    {
+      "epoch": 0.24637634787060442,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.5866621758167735e-05,
+      "loss": 0.1372,
+      "step": 7700
+    },
+    {
+      "epoch": 0.24669631715355325,
+      "grad_norm": 25.375,
+      "learning_rate": 1.5859885483327722e-05,
+      "loss": 0.15,
+      "step": 7710
+    },
+    {
+      "epoch": 0.24701628643650209,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5853149208487707e-05,
+      "loss": 0.1025,
+      "step": 7720
+    },
+    {
+      "epoch": 0.24733625571945092,
+      "grad_norm": 4.71875,
+      "learning_rate": 1.5846412933647695e-05,
+      "loss": 0.132,
+      "step": 7730
+    },
+    {
+      "epoch": 0.24765622500239978,
+      "grad_norm": 5.65625,
+      "learning_rate": 1.5839676658807682e-05,
+      "loss": 0.1472,
+      "step": 7740
+    },
+    {
+      "epoch": 0.24797619428534862,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.5832940383967667e-05,
+      "loss": 0.1148,
+      "step": 7750
+    },
+    {
+      "epoch": 0.24829616356829745,
+      "grad_norm": 9.75,
+      "learning_rate": 1.5826204109127654e-05,
+      "loss": 0.1518,
+      "step": 7760
+    },
+    {
+      "epoch": 0.24861613285124629,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.5819467834287642e-05,
+      "loss": 0.1138,
+      "step": 7770
+    },
+    {
+      "epoch": 0.24893610213419512,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5812731559447626e-05,
+      "loss": 0.1483,
+      "step": 7780
+    },
+    {
+      "epoch": 0.24925607141714395,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.5805995284607614e-05,
+      "loss": 0.0812,
+      "step": 7790
+    },
+    {
+      "epoch": 0.2495760407000928,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.57992590097676e-05,
+      "loss": 0.0918,
+      "step": 7800
+    },
+    {
+      "epoch": 0.24989600998304162,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.5792522734927586e-05,
+      "loss": 0.1659,
+      "step": 7810
+    },
+    {
+      "epoch": 0.2502159792659905,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5785786460087574e-05,
+      "loss": 0.1116,
+      "step": 7820
+    },
+    {
+      "epoch": 0.2505359485489393,
+      "grad_norm": 20.375,
+      "learning_rate": 1.577905018524756e-05,
+      "loss": 0.1521,
+      "step": 7830
+    },
+    {
+      "epoch": 0.25085591783188815,
+      "grad_norm": 20.5,
+      "learning_rate": 1.5772313910407546e-05,
+      "loss": 0.1637,
+      "step": 7840
+    },
+    {
+      "epoch": 0.25117588711483696,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.5765577635567534e-05,
+      "loss": 0.111,
+      "step": 7850
+    },
+    {
+      "epoch": 0.2514958563977858,
+      "grad_norm": 19.25,
+      "learning_rate": 1.5758841360727518e-05,
+      "loss": 0.1933,
+      "step": 7860
+    },
+    {
+      "epoch": 0.25181582568073463,
+      "grad_norm": 10.5,
+      "learning_rate": 1.5752105085887506e-05,
+      "loss": 0.1028,
+      "step": 7870
+    },
+    {
+      "epoch": 0.2521357949636835,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.5745368811047494e-05,
+      "loss": 0.1171,
+      "step": 7880
+    },
+    {
+      "epoch": 0.2524557642466323,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.5738632536207478e-05,
+      "loss": 0.1498,
+      "step": 7890
+    },
+    {
+      "epoch": 0.25277573352958116,
+      "grad_norm": 9.125,
+      "learning_rate": 1.5731896261367466e-05,
+      "loss": 0.1366,
+      "step": 7900
+    },
+    {
+      "epoch": 0.25309570281253,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5725159986527453e-05,
+      "loss": 0.1279,
+      "step": 7910
+    },
+    {
+      "epoch": 0.25341567209547883,
+      "grad_norm": 9.125,
+      "learning_rate": 1.5718423711687438e-05,
+      "loss": 0.0726,
+      "step": 7920
+    },
+    {
+      "epoch": 0.2537356413784277,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.5711687436847426e-05,
+      "loss": 0.1327,
+      "step": 7930
+    },
+    {
+      "epoch": 0.2540556106613765,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.570495116200741e-05,
+      "loss": 0.1418,
+      "step": 7940
+    },
+    {
+      "epoch": 0.25437557994432536,
+      "grad_norm": 20.125,
+      "learning_rate": 1.5698214887167398e-05,
+      "loss": 0.1674,
+      "step": 7950
+    },
+    {
+      "epoch": 0.25469554922727417,
+      "grad_norm": 22.0,
+      "learning_rate": 1.5691478612327385e-05,
+      "loss": 0.0886,
+      "step": 7960
+    },
+    {
+      "epoch": 0.255015518510223,
+      "grad_norm": 8.625,
+      "learning_rate": 1.568474233748737e-05,
+      "loss": 0.1446,
+      "step": 7970
+    },
+    {
+      "epoch": 0.25533548779317183,
+      "grad_norm": 7.875,
+      "learning_rate": 1.5678006062647357e-05,
+      "loss": 0.1127,
+      "step": 7980
+    },
+    {
+      "epoch": 0.2556554570761207,
+      "grad_norm": 5.875,
+      "learning_rate": 1.5671269787807345e-05,
+      "loss": 0.1043,
+      "step": 7990
+    },
+    {
+      "epoch": 0.2559754263590695,
+      "grad_norm": 30.5,
+      "learning_rate": 1.566453351296733e-05,
+      "loss": 0.1512,
+      "step": 8000
+    },
+    {
+      "epoch": 0.25629539564201836,
+      "grad_norm": 17.125,
+      "learning_rate": 1.5657797238127317e-05,
+      "loss": 0.1103,
+      "step": 8010
+    },
+    {
+      "epoch": 0.2566153649249672,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.5651060963287305e-05,
+      "loss": 0.1067,
+      "step": 8020
+    },
+    {
+      "epoch": 0.25693533420791603,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.564432468844729e-05,
+      "loss": 0.075,
+      "step": 8030
+    },
+    {
+      "epoch": 0.2572553034908649,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.5637588413607277e-05,
+      "loss": 0.0878,
+      "step": 8040
+    },
+    {
+      "epoch": 0.2575752727738137,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.563085213876726e-05,
+      "loss": 0.1468,
+      "step": 8050
+    },
+    {
+      "epoch": 0.25789524205676256,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.562411586392725e-05,
+      "loss": 0.1102,
+      "step": 8060
+    },
+    {
+      "epoch": 0.25821521133971137,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.5617379589087237e-05,
+      "loss": 0.118,
+      "step": 8070
+    },
+    {
+      "epoch": 0.25853518062266023,
+      "grad_norm": 16.125,
+      "learning_rate": 1.561064331424722e-05,
+      "loss": 0.1772,
+      "step": 8080
+    },
+    {
+      "epoch": 0.25885514990560904,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.560390703940721e-05,
+      "loss": 0.0989,
+      "step": 8090
+    },
+    {
+      "epoch": 0.2591751191885579,
+      "grad_norm": 21.375,
+      "learning_rate": 1.5597170764567197e-05,
+      "loss": 0.1573,
+      "step": 8100
+    },
+    {
+      "epoch": 0.25949508847150676,
+      "grad_norm": 4.6875,
+      "learning_rate": 1.559043448972718e-05,
+      "loss": 0.1529,
+      "step": 8110
+    },
+    {
+      "epoch": 0.25981505775445557,
+      "grad_norm": 7.125,
+      "learning_rate": 1.558369821488717e-05,
+      "loss": 0.1163,
+      "step": 8120
+    },
+    {
+      "epoch": 0.26013502703740443,
+      "grad_norm": 15.125,
+      "learning_rate": 1.5576961940047157e-05,
+      "loss": 0.1446,
+      "step": 8130
+    },
+    {
+      "epoch": 0.26045499632035324,
+      "grad_norm": 9.5,
+      "learning_rate": 1.557022566520714e-05,
+      "loss": 0.1785,
+      "step": 8140
+    },
+    {
+      "epoch": 0.2607749656033021,
+      "grad_norm": 11.125,
+      "learning_rate": 1.556348939036713e-05,
+      "loss": 0.1158,
+      "step": 8150
+    },
+    {
+      "epoch": 0.2610949348862509,
+      "grad_norm": 12.125,
+      "learning_rate": 1.5556753115527113e-05,
+      "loss": 0.1184,
+      "step": 8160
+    },
+    {
+      "epoch": 0.26141490416919977,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.55500168406871e-05,
+      "loss": 0.0967,
+      "step": 8170
+    },
+    {
+      "epoch": 0.2617348734521486,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.554328056584709e-05,
+      "loss": 0.055,
+      "step": 8180
+    },
+    {
+      "epoch": 0.26205484273509744,
+      "grad_norm": 13.5,
+      "learning_rate": 1.5536544291007073e-05,
+      "loss": 0.082,
+      "step": 8190
+    },
+    {
+      "epoch": 0.26237481201804624,
+      "grad_norm": 12.5,
+      "learning_rate": 1.552980801616706e-05,
+      "loss": 0.1091,
+      "step": 8200
+    },
+    {
+      "epoch": 0.2626947813009951,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.5523071741327048e-05,
+      "loss": 0.112,
+      "step": 8210
+    },
+    {
+      "epoch": 0.26301475058394397,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5516335466487033e-05,
+      "loss": 0.1267,
+      "step": 8220
+    },
+    {
+      "epoch": 0.2633347198668928,
+      "grad_norm": 4.5625,
+      "learning_rate": 1.550959919164702e-05,
+      "loss": 0.1499,
+      "step": 8230
+    },
+    {
+      "epoch": 0.26365468914984164,
+      "grad_norm": 16.0,
+      "learning_rate": 1.5502862916807008e-05,
+      "loss": 0.1749,
+      "step": 8240
+    },
+    {
+      "epoch": 0.26397465843279044,
+      "grad_norm": 15.875,
+      "learning_rate": 1.5496126641966992e-05,
+      "loss": 0.1058,
+      "step": 8250
+    },
+    {
+      "epoch": 0.2642946277157393,
+      "grad_norm": 25.75,
+      "learning_rate": 1.548939036712698e-05,
+      "loss": 0.1219,
+      "step": 8260
+    },
+    {
+      "epoch": 0.2646145969986881,
+      "grad_norm": 17.875,
+      "learning_rate": 1.5482654092286965e-05,
+      "loss": 0.0995,
+      "step": 8270
+    },
+    {
+      "epoch": 0.264934566281637,
+      "grad_norm": 34.25,
+      "learning_rate": 1.5475917817446952e-05,
+      "loss": 0.1514,
+      "step": 8280
+    },
+    {
+      "epoch": 0.2652545355645858,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.546918154260694e-05,
+      "loss": 0.2644,
+      "step": 8290
+    },
+    {
+      "epoch": 0.26557450484753464,
+      "grad_norm": 15.0,
+      "learning_rate": 1.5462445267766924e-05,
+      "loss": 0.1118,
+      "step": 8300
+    },
+    {
+      "epoch": 0.26589447413048345,
+      "grad_norm": 2.5,
+      "learning_rate": 1.5455708992926912e-05,
+      "loss": 0.0594,
+      "step": 8310
+    },
+    {
+      "epoch": 0.2662144434134323,
+      "grad_norm": 5.75,
+      "learning_rate": 1.54489727180869e-05,
+      "loss": 0.0984,
+      "step": 8320
+    },
+    {
+      "epoch": 0.2665344126963812,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.5442236443246884e-05,
+      "loss": 0.1146,
+      "step": 8330
+    },
+    {
+      "epoch": 0.26685438197933,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.5435500168406872e-05,
+      "loss": 0.078,
+      "step": 8340
+    },
+    {
+      "epoch": 0.26717435126227884,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.542876389356686e-05,
+      "loss": 0.1246,
+      "step": 8350
+    },
+    {
+      "epoch": 0.26749432054522765,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.5422027618726844e-05,
+      "loss": 0.1302,
+      "step": 8360
+    },
+    {
+      "epoch": 0.2678142898281765,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.5415291343886832e-05,
+      "loss": 0.0946,
+      "step": 8370
+    },
+    {
+      "epoch": 0.2681342591111253,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.5408555069046816e-05,
+      "loss": 0.0854,
+      "step": 8380
+    },
+    {
+      "epoch": 0.2684542283940742,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.5401818794206804e-05,
+      "loss": 0.0904,
+      "step": 8390
+    },
+    {
+      "epoch": 0.268774197677023,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.539508251936679e-05,
+      "loss": 0.1549,
+      "step": 8400
+    },
+    {
+      "epoch": 0.26909416695997185,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.5388346244526776e-05,
+      "loss": 0.1385,
+      "step": 8410
+    },
+    {
+      "epoch": 0.26941413624292065,
+      "grad_norm": 6.0625,
+      "learning_rate": 1.5381609969686764e-05,
+      "loss": 0.135,
+      "step": 8420
+    },
+    {
+      "epoch": 0.2697341055258695,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.537487369484675e-05,
+      "loss": 0.0887,
+      "step": 8430
+    },
+    {
+      "epoch": 0.2700540748088184,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.5368137420006736e-05,
+      "loss": 0.1285,
+      "step": 8440
+    },
+    {
+      "epoch": 0.2703740440917672,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.5361401145166723e-05,
+      "loss": 0.128,
+      "step": 8450
+    },
+    {
+      "epoch": 0.27069401337471605,
+      "grad_norm": 10.875,
+      "learning_rate": 1.535466487032671e-05,
+      "loss": 0.1024,
+      "step": 8460
+    },
+    {
+      "epoch": 0.27101398265766485,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.5347928595486696e-05,
+      "loss": 0.1384,
+      "step": 8470
+    },
+    {
+      "epoch": 0.2713339519406137,
+      "grad_norm": 9.875,
+      "learning_rate": 1.5341192320646683e-05,
+      "loss": 0.1007,
+      "step": 8480
+    },
+    {
+      "epoch": 0.2716539212235625,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.533445604580667e-05,
+      "loss": 0.0678,
+      "step": 8490
+    },
+    {
+      "epoch": 0.2719738905065114,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.5327719770966655e-05,
+      "loss": 0.1092,
+      "step": 8500
+    },
+    {
+      "epoch": 0.2722938597894602,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5320983496126643e-05,
+      "loss": 0.0776,
+      "step": 8510
+    },
+    {
+      "epoch": 0.27261382907240905,
+      "grad_norm": 16.0,
+      "learning_rate": 1.5314247221286627e-05,
+      "loss": 0.0872,
+      "step": 8520
+    },
+    {
+      "epoch": 0.27293379835535786,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.5307510946446615e-05,
+      "loss": 0.1059,
+      "step": 8530
+    },
+    {
+      "epoch": 0.2732537676383067,
+      "grad_norm": 9.125,
+      "learning_rate": 1.5300774671606603e-05,
+      "loss": 0.099,
+      "step": 8540
+    },
+    {
+      "epoch": 0.2735737369212556,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5294038396766587e-05,
+      "loss": 0.1089,
+      "step": 8550
+    },
+    {
+      "epoch": 0.2738937062042044,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.5287302121926575e-05,
+      "loss": 0.1797,
+      "step": 8560
+    },
+    {
+      "epoch": 0.27421367548715325,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.5280565847086563e-05,
+      "loss": 0.0968,
+      "step": 8570
+    },
+    {
+      "epoch": 0.27453364477010206,
+      "grad_norm": 15.0,
+      "learning_rate": 1.5273829572246547e-05,
+      "loss": 0.1877,
+      "step": 8580
+    },
+    {
+      "epoch": 0.2748536140530509,
+      "grad_norm": 17.625,
+      "learning_rate": 1.5267093297406535e-05,
+      "loss": 0.1551,
+      "step": 8590
+    },
+    {
+      "epoch": 0.2751735833359997,
+      "grad_norm": 20.25,
+      "learning_rate": 1.5260357022566523e-05,
+      "loss": 0.1506,
+      "step": 8600
+    },
+    {
+      "epoch": 0.2754935526189486,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.5253620747726507e-05,
+      "loss": 0.1837,
+      "step": 8610
+    },
+    {
+      "epoch": 0.2758135219018974,
+      "grad_norm": 7.25,
+      "learning_rate": 1.5246884472886495e-05,
+      "loss": 0.1349,
+      "step": 8620
+    },
+    {
+      "epoch": 0.27613349118484626,
+      "grad_norm": 6.28125,
+      "learning_rate": 1.524014819804648e-05,
+      "loss": 0.0888,
+      "step": 8630
+    },
+    {
+      "epoch": 0.27645346046779506,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5233411923206467e-05,
+      "loss": 0.1094,
+      "step": 8640
+    },
+    {
+      "epoch": 0.2767734297507439,
+      "grad_norm": 11.1875,
+      "learning_rate": 1.5226675648366453e-05,
+      "loss": 0.1154,
+      "step": 8650
+    },
+    {
+      "epoch": 0.2770933990336928,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.521993937352644e-05,
+      "loss": 0.1251,
+      "step": 8660
+    },
+    {
+      "epoch": 0.2774133683166416,
+      "grad_norm": 10.375,
+      "learning_rate": 1.5213203098686428e-05,
+      "loss": 0.1021,
+      "step": 8670
+    },
+    {
+      "epoch": 0.27773333759959046,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.5206466823846416e-05,
+      "loss": 0.0825,
+      "step": 8680
+    },
+    {
+      "epoch": 0.27805330688253926,
+      "grad_norm": 6.0,
+      "learning_rate": 1.5199730549006402e-05,
+      "loss": 0.1144,
+      "step": 8690
+    },
+    {
+      "epoch": 0.2783732761654881,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.5192994274166388e-05,
+      "loss": 0.0845,
+      "step": 8700
+    },
+    {
+      "epoch": 0.27869324544843693,
+      "grad_norm": 7.375,
+      "learning_rate": 1.5186257999326374e-05,
+      "loss": 0.137,
+      "step": 8710
+    },
+    {
+      "epoch": 0.2790132147313858,
+      "grad_norm": 10.5625,
+      "learning_rate": 1.5179521724486362e-05,
+      "loss": 0.1011,
+      "step": 8720
+    },
+    {
+      "epoch": 0.2793331840143346,
+      "grad_norm": 17.75,
+      "learning_rate": 1.5172785449646348e-05,
+      "loss": 0.0991,
+      "step": 8730
+    },
+    {
+      "epoch": 0.27965315329728346,
+      "grad_norm": 41.5,
+      "learning_rate": 1.5166049174806334e-05,
+      "loss": 0.1776,
+      "step": 8740
+    },
+    {
+      "epoch": 0.2799731225802323,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.5159312899966322e-05,
+      "loss": 0.0632,
+      "step": 8750
+    },
+    {
+      "epoch": 0.28029309186318113,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.5152576625126308e-05,
+      "loss": 0.0542,
+      "step": 8760
+    },
+    {
+      "epoch": 0.28061306114613,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.5145840350286294e-05,
+      "loss": 0.0905,
+      "step": 8770
+    },
+    {
+      "epoch": 0.2809330304290788,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.513910407544628e-05,
+      "loss": 0.1196,
+      "step": 8780
+    },
+    {
+      "epoch": 0.28125299971202766,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.5132367800606267e-05,
+      "loss": 0.1511,
+      "step": 8790
+    },
+    {
+      "epoch": 0.28157296899497647,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.5125631525766254e-05,
+      "loss": 0.1364,
+      "step": 8800
+    },
+    {
+      "epoch": 0.28189293827792533,
+      "grad_norm": 20.0,
+      "learning_rate": 1.511889525092624e-05,
+      "loss": 0.1594,
+      "step": 8810
+    },
+    {
+      "epoch": 0.28221290756087414,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.5112158976086226e-05,
+      "loss": 0.0953,
+      "step": 8820
+    },
+    {
+      "epoch": 0.282532876843823,
+      "grad_norm": 4.8125,
+      "learning_rate": 1.5105422701246213e-05,
+      "loss": 0.1271,
+      "step": 8830
+    },
+    {
+      "epoch": 0.2828528461267718,
+      "grad_norm": 5.71875,
+      "learning_rate": 1.50986864264062e-05,
+      "loss": 0.0866,
+      "step": 8840
+    },
+    {
+      "epoch": 0.28317281540972067,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.5091950151566185e-05,
+      "loss": 0.1211,
+      "step": 8850
+    },
+    {
+      "epoch": 0.28349278469266953,
+      "grad_norm": 1.75,
+      "learning_rate": 1.5085213876726173e-05,
+      "loss": 0.1184,
+      "step": 8860
+    },
+    {
+      "epoch": 0.28381275397561834,
+      "grad_norm": 19.875,
+      "learning_rate": 1.507847760188616e-05,
+      "loss": 0.0913,
+      "step": 8870
+    },
+    {
+      "epoch": 0.2841327232585672,
+      "grad_norm": 20.0,
+      "learning_rate": 1.5071741327046145e-05,
+      "loss": 0.0933,
+      "step": 8880
+    },
+    {
+      "epoch": 0.284452692541516,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.5065005052206131e-05,
+      "loss": 0.1482,
+      "step": 8890
+    },
+    {
+      "epoch": 0.28477266182446487,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.5058268777366119e-05,
+      "loss": 0.0946,
+      "step": 8900
+    },
+    {
+      "epoch": 0.2850926311074137,
+      "grad_norm": 14.25,
+      "learning_rate": 1.5051532502526105e-05,
+      "loss": 0.145,
+      "step": 8910
+    },
+    {
+      "epoch": 0.28541260039036254,
+      "grad_norm": 10.5,
+      "learning_rate": 1.5044796227686091e-05,
+      "loss": 0.106,
+      "step": 8920
+    },
+    {
+      "epoch": 0.28573256967331134,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.5038059952846079e-05,
+      "loss": 0.0602,
+      "step": 8930
+    },
+    {
+      "epoch": 0.2860525389562602,
+      "grad_norm": 7.375,
+      "learning_rate": 1.5031323678006065e-05,
+      "loss": 0.1149,
+      "step": 8940
+    },
+    {
+      "epoch": 0.286372508239209,
+      "grad_norm": 17.0,
+      "learning_rate": 1.5024587403166051e-05,
+      "loss": 0.1095,
+      "step": 8950
+    },
+    {
+      "epoch": 0.2866924775221579,
+      "grad_norm": 9.75,
+      "learning_rate": 1.5017851128326037e-05,
+      "loss": 0.0986,
+      "step": 8960
+    },
+    {
+      "epoch": 0.28701244680510674,
+      "grad_norm": 17.25,
+      "learning_rate": 1.5011114853486025e-05,
+      "loss": 0.1214,
+      "step": 8970
+    },
+    {
+      "epoch": 0.28733241608805554,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.500437857864601e-05,
+      "loss": 0.1167,
+      "step": 8980
+    },
+    {
+      "epoch": 0.2876523853710044,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.4997642303805997e-05,
+      "loss": 0.1413,
+      "step": 8990
+    },
+    {
+      "epoch": 0.2879723546539532,
+      "grad_norm": 28.875,
+      "learning_rate": 1.4990906028965983e-05,
+      "loss": 0.0858,
+      "step": 9000
+    },
+    {
+      "epoch": 0.2882923239369021,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.498416975412597e-05,
+      "loss": 0.0825,
+      "step": 9010
+    },
+    {
+      "epoch": 0.2886122932198509,
+      "grad_norm": 19.0,
+      "learning_rate": 1.4977433479285957e-05,
+      "loss": 0.1569,
+      "step": 9020
+    },
+    {
+      "epoch": 0.28893226250279974,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4970697204445943e-05,
+      "loss": 0.1458,
+      "step": 9030
+    },
+    {
+      "epoch": 0.28925223178574855,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.496396092960593e-05,
+      "loss": 0.1042,
+      "step": 9040
+    },
+    {
+      "epoch": 0.2895722010686974,
+      "grad_norm": 11.125,
+      "learning_rate": 1.4957224654765916e-05,
+      "loss": 0.1046,
+      "step": 9050
+    },
+    {
+      "epoch": 0.2898921703516462,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.4950488379925902e-05,
+      "loss": 0.1126,
+      "step": 9060
+    },
+    {
+      "epoch": 0.2902121396345951,
+      "grad_norm": 12.875,
+      "learning_rate": 1.4943752105085888e-05,
+      "loss": 0.0573,
+      "step": 9070
+    },
+    {
+      "epoch": 0.29053210891754394,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.4937015830245876e-05,
+      "loss": 0.1601,
+      "step": 9080
+    },
+    {
+      "epoch": 0.29085207820049275,
+      "grad_norm": 12.625,
+      "learning_rate": 1.4930279555405862e-05,
+      "loss": 0.119,
+      "step": 9090
+    },
+    {
+      "epoch": 0.2911720474834416,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.4923543280565848e-05,
+      "loss": 0.1089,
+      "step": 9100
+    },
+    {
+      "epoch": 0.2914920167663904,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.4916807005725834e-05,
+      "loss": 0.0874,
+      "step": 9110
+    },
+    {
+      "epoch": 0.2918119860493393,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.4910070730885822e-05,
+      "loss": 0.1135,
+      "step": 9120
+    },
+    {
+      "epoch": 0.2921319553322881,
+      "grad_norm": 9.0,
+      "learning_rate": 1.4903334456045808e-05,
+      "loss": 0.1416,
+      "step": 9130
+    },
+    {
+      "epoch": 0.29245192461523695,
+      "grad_norm": 17.25,
+      "learning_rate": 1.4896598181205794e-05,
+      "loss": 0.1287,
+      "step": 9140
+    },
+    {
+      "epoch": 0.29277189389818575,
+      "grad_norm": 9.875,
+      "learning_rate": 1.4889861906365782e-05,
+      "loss": 0.141,
+      "step": 9150
+    },
+    {
+      "epoch": 0.2930918631811346,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.4883125631525768e-05,
+      "loss": 0.1045,
+      "step": 9160
+    },
+    {
+      "epoch": 0.2934118324640834,
+      "grad_norm": 24.25,
+      "learning_rate": 1.4876389356685754e-05,
+      "loss": 0.1603,
+      "step": 9170
+    },
+    {
+      "epoch": 0.2937318017470323,
+      "grad_norm": 13.625,
+      "learning_rate": 1.486965308184574e-05,
+      "loss": 0.1114,
+      "step": 9180
+    },
+    {
+      "epoch": 0.29405177102998115,
+      "grad_norm": 5.15625,
+      "learning_rate": 1.4862916807005728e-05,
+      "loss": 0.1317,
+      "step": 9190
+    },
+    {
+      "epoch": 0.29437174031292995,
+      "grad_norm": 5.5,
+      "learning_rate": 1.4856180532165714e-05,
+      "loss": 0.0948,
+      "step": 9200
+    },
+    {
+      "epoch": 0.2946917095958788,
+      "grad_norm": 8.75,
+      "learning_rate": 1.48494442573257e-05,
+      "loss": 0.1485,
+      "step": 9210
+    },
+    {
+      "epoch": 0.2950116788788276,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.4842707982485688e-05,
+      "loss": 0.1273,
+      "step": 9220
+    },
+    {
+      "epoch": 0.2953316481617765,
+      "grad_norm": 9.375,
+      "learning_rate": 1.4835971707645674e-05,
+      "loss": 0.0799,
+      "step": 9230
+    },
+    {
+      "epoch": 0.2956516174447253,
+      "grad_norm": 4.9375,
+      "learning_rate": 1.482923543280566e-05,
+      "loss": 0.1229,
+      "step": 9240
+    },
+    {
+      "epoch": 0.29597158672767415,
+      "grad_norm": 18.75,
+      "learning_rate": 1.4822499157965646e-05,
+      "loss": 0.1457,
+      "step": 9250
+    },
+    {
+      "epoch": 0.29629155601062296,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.4815762883125633e-05,
+      "loss": 0.095,
+      "step": 9260
+    },
+    {
+      "epoch": 0.2966115252935718,
+      "grad_norm": 17.5,
+      "learning_rate": 1.480902660828562e-05,
+      "loss": 0.097,
+      "step": 9270
+    },
+    {
+      "epoch": 0.2969314945765206,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.4802290333445606e-05,
+      "loss": 0.1556,
+      "step": 9280
+    },
+    {
+      "epoch": 0.2972514638594695,
+      "grad_norm": 26.25,
+      "learning_rate": 1.4795554058605592e-05,
+      "loss": 0.1373,
+      "step": 9290
+    },
+    {
+      "epoch": 0.29757143314241835,
+      "grad_norm": 25.125,
+      "learning_rate": 1.478881778376558e-05,
+      "loss": 0.1193,
+      "step": 9300
+    },
+    {
+      "epoch": 0.29789140242536716,
+      "grad_norm": 14.125,
+      "learning_rate": 1.4782081508925565e-05,
+      "loss": 0.1143,
+      "step": 9310
+    },
+    {
+      "epoch": 0.298211371708316,
+      "grad_norm": 9.5,
+      "learning_rate": 1.4775345234085551e-05,
+      "loss": 0.0895,
+      "step": 9320
+    },
+    {
+      "epoch": 0.2985313409912648,
+      "grad_norm": 7.5,
+      "learning_rate": 1.4768608959245539e-05,
+      "loss": 0.0846,
+      "step": 9330
+    },
+    {
+      "epoch": 0.2988513102742137,
+      "grad_norm": 20.0,
+      "learning_rate": 1.4761872684405525e-05,
+      "loss": 0.1106,
+      "step": 9340
+    },
+    {
+      "epoch": 0.2991712795571625,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.4755136409565511e-05,
+      "loss": 0.1722,
+      "step": 9350
+    },
+    {
+      "epoch": 0.29949124884011136,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.4748400134725497e-05,
+      "loss": 0.0708,
+      "step": 9360
+    },
+    {
+      "epoch": 0.29981121812306016,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.4741663859885485e-05,
+      "loss": 0.1534,
+      "step": 9370
+    },
+    {
+      "epoch": 0.300131187406009,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.4734927585045471e-05,
+      "loss": 0.1129,
+      "step": 9380
+    },
+    {
+      "epoch": 0.3004511566889579,
+      "grad_norm": 9.25,
+      "learning_rate": 1.4728191310205457e-05,
+      "loss": 0.1174,
+      "step": 9390
+    },
+    {
+      "epoch": 0.3007711259719067,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.4721455035365445e-05,
+      "loss": 0.0873,
+      "step": 9400
+    },
+    {
+      "epoch": 0.30109109525485556,
+      "grad_norm": 11.375,
+      "learning_rate": 1.471471876052543e-05,
+      "loss": 0.0908,
+      "step": 9410
+    },
+    {
+      "epoch": 0.30141106453780436,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.4707982485685417e-05,
+      "loss": 0.154,
+      "step": 9420
+    },
+    {
+      "epoch": 0.3017310338207532,
+      "grad_norm": 16.875,
+      "learning_rate": 1.4701246210845403e-05,
+      "loss": 0.1004,
+      "step": 9430
+    },
+    {
+      "epoch": 0.30205100310370203,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.469450993600539e-05,
+      "loss": 0.0703,
+      "step": 9440
+    },
+    {
+      "epoch": 0.3023709723866509,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.4687773661165377e-05,
+      "loss": 0.1026,
+      "step": 9450
+    },
+    {
+      "epoch": 0.3026909416695997,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4681037386325363e-05,
+      "loss": 0.1203,
+      "step": 9460
+    },
+    {
+      "epoch": 0.30301091095254856,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.4674301111485349e-05,
+      "loss": 0.1584,
+      "step": 9470
+    },
+    {
+      "epoch": 0.30333088023549737,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.4667564836645337e-05,
+      "loss": 0.1524,
+      "step": 9480
+    },
+    {
+      "epoch": 0.30365084951844623,
+      "grad_norm": 6.40625,
+      "learning_rate": 1.4660828561805323e-05,
+      "loss": 0.1914,
+      "step": 9490
+    },
+    {
+      "epoch": 0.3039708188013951,
+      "grad_norm": 11.125,
+      "learning_rate": 1.4654092286965309e-05,
+      "loss": 0.1154,
+      "step": 9500
+    },
+    {
+      "epoch": 0.3042907880843439,
+      "grad_norm": 9.5,
+      "learning_rate": 1.4647356012125296e-05,
+      "loss": 0.1174,
+      "step": 9510
+    },
+    {
+      "epoch": 0.30461075736729276,
+      "grad_norm": 5.3125,
+      "learning_rate": 1.4640619737285282e-05,
+      "loss": 0.1947,
+      "step": 9520
+    },
+    {
+      "epoch": 0.30493072665024157,
+      "grad_norm": 5.78125,
+      "learning_rate": 1.4633883462445268e-05,
+      "loss": 0.0843,
+      "step": 9530
+    },
+    {
+      "epoch": 0.30525069593319043,
+      "grad_norm": 12.875,
+      "learning_rate": 1.4627147187605254e-05,
+      "loss": 0.1404,
+      "step": 9540
+    },
+    {
+      "epoch": 0.30557066521613924,
+      "grad_norm": 5.75,
+      "learning_rate": 1.4620410912765242e-05,
+      "loss": 0.1464,
+      "step": 9550
+    },
+    {
+      "epoch": 0.3058906344990881,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.4613674637925228e-05,
+      "loss": 0.1197,
+      "step": 9560
+    },
+    {
+      "epoch": 0.3062106037820369,
+      "grad_norm": 9.125,
+      "learning_rate": 1.4606938363085214e-05,
+      "loss": 0.1624,
+      "step": 9570
+    },
+    {
+      "epoch": 0.30653057306498577,
+      "grad_norm": 8.375,
+      "learning_rate": 1.46002020882452e-05,
+      "loss": 0.1093,
+      "step": 9580
+    },
+    {
+      "epoch": 0.3068505423479346,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.4593465813405188e-05,
+      "loss": 0.1241,
+      "step": 9590
+    },
+    {
+      "epoch": 0.30717051163088344,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.4586729538565174e-05,
+      "loss": 0.114,
+      "step": 9600
+    },
+    {
+      "epoch": 0.3074904809138323,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.457999326372516e-05,
+      "loss": 0.1178,
+      "step": 9610
+    },
+    {
+      "epoch": 0.3078104501967811,
+      "grad_norm": 14.125,
+      "learning_rate": 1.4573256988885148e-05,
+      "loss": 0.1257,
+      "step": 9620
+    },
+    {
+      "epoch": 0.30813041947972997,
+      "grad_norm": 15.0,
+      "learning_rate": 1.4566520714045134e-05,
+      "loss": 0.1045,
+      "step": 9630
+    },
+    {
+      "epoch": 0.3084503887626788,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.455978443920512e-05,
+      "loss": 0.1169,
+      "step": 9640
+    },
+    {
+      "epoch": 0.30877035804562764,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.4553048164365106e-05,
+      "loss": 0.1398,
+      "step": 9650
+    },
+    {
+      "epoch": 0.30909032732857644,
+      "grad_norm": 5.46875,
+      "learning_rate": 1.4546311889525094e-05,
+      "loss": 0.1164,
+      "step": 9660
+    },
+    {
+      "epoch": 0.3094102966115253,
+      "grad_norm": 6.53125,
+      "learning_rate": 1.453957561468508e-05,
+      "loss": 0.0888,
+      "step": 9670
+    },
+    {
+      "epoch": 0.3097302658944741,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4532839339845066e-05,
+      "loss": 0.0635,
+      "step": 9680
+    },
+    {
+      "epoch": 0.310050235177423,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.4526103065005054e-05,
+      "loss": 0.0842,
+      "step": 9690
+    },
+    {
+      "epoch": 0.3103702044603718,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.451936679016504e-05,
+      "loss": 0.148,
+      "step": 9700
+    },
+    {
+      "epoch": 0.31069017374332064,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.4512630515325026e-05,
+      "loss": 0.1911,
+      "step": 9710
+    },
+    {
+      "epoch": 0.3110101430262695,
+      "grad_norm": 10.4375,
+      "learning_rate": 1.4505894240485012e-05,
+      "loss": 0.0915,
+      "step": 9720
+    },
+    {
+      "epoch": 0.3113301123092183,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.4499157965645e-05,
+      "loss": 0.0696,
+      "step": 9730
+    },
+    {
+      "epoch": 0.31165008159216717,
+      "grad_norm": 18.5,
+      "learning_rate": 1.4492421690804985e-05,
+      "loss": 0.1597,
+      "step": 9740
+    },
+    {
+      "epoch": 0.311970050875116,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.4485685415964971e-05,
+      "loss": 0.1108,
+      "step": 9750
+    },
+    {
+      "epoch": 0.31229002015806484,
+      "grad_norm": 9.75,
+      "learning_rate": 1.4478949141124958e-05,
+      "loss": 0.1168,
+      "step": 9760
+    },
+    {
+      "epoch": 0.31260998944101365,
+      "grad_norm": 6.28125,
+      "learning_rate": 1.4472212866284945e-05,
+      "loss": 0.0774,
+      "step": 9770
+    },
+    {
+      "epoch": 0.3129299587239625,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.4465476591444931e-05,
+      "loss": 0.1198,
+      "step": 9780
+    },
+    {
+      "epoch": 0.3132499280069113,
+      "grad_norm": 33.0,
+      "learning_rate": 1.4458740316604917e-05,
+      "loss": 0.1354,
+      "step": 9790
+    },
+    {
+      "epoch": 0.3135698972898602,
+      "grad_norm": 5.78125,
+      "learning_rate": 1.4452004041764905e-05,
+      "loss": 0.1465,
+      "step": 9800
+    },
+    {
+      "epoch": 0.313889866572809,
+      "grad_norm": 14.375,
+      "learning_rate": 1.4445267766924891e-05,
+      "loss": 0.1162,
+      "step": 9810
+    },
+    {
+      "epoch": 0.31420983585575785,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.4438531492084877e-05,
+      "loss": 0.0746,
+      "step": 9820
+    },
+    {
+      "epoch": 0.3145298051387067,
+      "grad_norm": 4.375,
+      "learning_rate": 1.4431795217244863e-05,
+      "loss": 0.1316,
+      "step": 9830
+    },
+    {
+      "epoch": 0.3148497744216555,
+      "grad_norm": 5.78125,
+      "learning_rate": 1.4425058942404851e-05,
+      "loss": 0.1076,
+      "step": 9840
+    },
+    {
+      "epoch": 0.3151697437046044,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.4418322667564837e-05,
+      "loss": 0.1284,
+      "step": 9850
+    },
+    {
+      "epoch": 0.3154897129875532,
+      "grad_norm": 11.0,
+      "learning_rate": 1.4411586392724823e-05,
+      "loss": 0.0869,
+      "step": 9860
+    },
+    {
+      "epoch": 0.31580968227050205,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.440485011788481e-05,
+      "loss": 0.1006,
+      "step": 9870
+    },
+    {
+      "epoch": 0.31612965155345085,
+      "grad_norm": 4.59375,
+      "learning_rate": 1.4398113843044797e-05,
+      "loss": 0.0717,
+      "step": 9880
+    },
+    {
+      "epoch": 0.3164496208363997,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.4391377568204783e-05,
+      "loss": 0.1192,
+      "step": 9890
+    },
+    {
+      "epoch": 0.3167695901193485,
+      "grad_norm": 9.25,
+      "learning_rate": 1.4384641293364769e-05,
+      "loss": 0.1052,
+      "step": 9900
+    },
+    {
+      "epoch": 0.3170895594022974,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.4377905018524757e-05,
+      "loss": 0.1174,
+      "step": 9910
+    },
+    {
+      "epoch": 0.3174095286852462,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4371168743684743e-05,
+      "loss": 0.1466,
+      "step": 9920
+    },
+    {
+      "epoch": 0.31772949796819505,
+      "grad_norm": 4.34375,
+      "learning_rate": 1.4364432468844729e-05,
+      "loss": 0.108,
+      "step": 9930
+    },
+    {
+      "epoch": 0.3180494672511439,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.4357696194004715e-05,
+      "loss": 0.1376,
+      "step": 9940
+    },
+    {
+      "epoch": 0.3183694365340927,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4350959919164702e-05,
+      "loss": 0.0933,
+      "step": 9950
+    },
+    {
+      "epoch": 0.3186894058170416,
+      "grad_norm": 4.90625,
+      "learning_rate": 1.434422364432469e-05,
+      "loss": 0.1111,
+      "step": 9960
+    },
+    {
+      "epoch": 0.3190093750999904,
+      "grad_norm": 20.75,
+      "learning_rate": 1.4337487369484678e-05,
+      "loss": 0.1331,
+      "step": 9970
+    },
+    {
+      "epoch": 0.31932934438293925,
+      "grad_norm": 5.1875,
+      "learning_rate": 1.4330751094644664e-05,
+      "loss": 0.148,
+      "step": 9980
+    },
+    {
+      "epoch": 0.31964931366588806,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.432401481980465e-05,
+      "loss": 0.0841,
+      "step": 9990
+    },
+    {
+      "epoch": 0.3199692829488369,
+      "grad_norm": 16.125,
+      "learning_rate": 1.4317278544964636e-05,
+      "loss": 0.1145,
+      "step": 10000
+    },
+    {
+      "epoch": 0.3202892522317857,
+      "grad_norm": 5.21875,
+      "learning_rate": 1.4310542270124624e-05,
+      "loss": 0.1404,
+      "step": 10010
+    },
+    {
+      "epoch": 0.3206092215147346,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.430380599528461e-05,
+      "loss": 0.1227,
+      "step": 10020
+    },
+    {
+      "epoch": 0.32092919079768345,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.4297069720444596e-05,
+      "loss": 0.1495,
+      "step": 10030
+    },
+    {
+      "epoch": 0.32124916008063226,
+      "grad_norm": 5.125,
+      "learning_rate": 1.4290333445604584e-05,
+      "loss": 0.1115,
+      "step": 10040
+    },
+    {
+      "epoch": 0.3215691293635811,
+      "grad_norm": 8.375,
+      "learning_rate": 1.428359717076457e-05,
+      "loss": 0.0982,
+      "step": 10050
+    },
+    {
+      "epoch": 0.3218890986465299,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.4276860895924556e-05,
+      "loss": 0.1136,
+      "step": 10060
+    },
+    {
+      "epoch": 0.3222090679294788,
+      "grad_norm": 12.875,
+      "learning_rate": 1.4270124621084542e-05,
+      "loss": 0.166,
+      "step": 10070
+    },
+    {
+      "epoch": 0.3225290372124276,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.426338834624453e-05,
+      "loss": 0.0961,
+      "step": 10080
+    },
+    {
+      "epoch": 0.32284900649537646,
+      "grad_norm": 5.5625,
+      "learning_rate": 1.4256652071404516e-05,
+      "loss": 0.1153,
+      "step": 10090
+    },
+    {
+      "epoch": 0.32316897577832526,
+      "grad_norm": 13.75,
+      "learning_rate": 1.4249915796564502e-05,
+      "loss": 0.1336,
+      "step": 10100
+    },
+    {
+      "epoch": 0.3234889450612741,
+      "grad_norm": 15.125,
+      "learning_rate": 1.4243179521724488e-05,
+      "loss": 0.1708,
+      "step": 10110
+    },
+    {
+      "epoch": 0.32380891434422293,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.4236443246884475e-05,
+      "loss": 0.1428,
+      "step": 10120
+    },
+    {
+      "epoch": 0.3241288836271718,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.4229706972044461e-05,
+      "loss": 0.1125,
+      "step": 10130
+    },
+    {
+      "epoch": 0.32444885291012066,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.4222970697204447e-05,
+      "loss": 0.1269,
+      "step": 10140
+    },
+    {
+      "epoch": 0.32476882219306946,
+      "grad_norm": 17.5,
+      "learning_rate": 1.4216234422364435e-05,
+      "loss": 0.157,
+      "step": 10150
+    },
+    {
+      "epoch": 0.3250887914760183,
+      "grad_norm": 9.5,
+      "learning_rate": 1.4209498147524421e-05,
+      "loss": 0.0895,
+      "step": 10160
+    },
+    {
+      "epoch": 0.32540876075896713,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.4202761872684407e-05,
+      "loss": 0.0963,
+      "step": 10170
+    },
+    {
+      "epoch": 0.325728730041916,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.4196025597844393e-05,
+      "loss": 0.108,
+      "step": 10180
+    },
+    {
+      "epoch": 0.3260486993248648,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.4189289323004381e-05,
+      "loss": 0.1689,
+      "step": 10190
+    },
+    {
+      "epoch": 0.32636866860781366,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4182553048164367e-05,
+      "loss": 0.1628,
+      "step": 10200
+    },
+    {
+      "epoch": 0.32668863789076247,
+      "grad_norm": 5.625,
+      "learning_rate": 1.4175816773324353e-05,
+      "loss": 0.0784,
+      "step": 10210
+    },
+    {
+      "epoch": 0.32700860717371133,
+      "grad_norm": 17.75,
+      "learning_rate": 1.4169080498484339e-05,
+      "loss": 0.1304,
+      "step": 10220
+    },
+    {
+      "epoch": 0.32732857645666014,
+      "grad_norm": 5.90625,
+      "learning_rate": 1.4162344223644327e-05,
+      "loss": 0.1488,
+      "step": 10230
+    },
+    {
+      "epoch": 0.327648545739609,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.4155607948804313e-05,
+      "loss": 0.0688,
+      "step": 10240
+    },
+    {
+      "epoch": 0.32796851502255786,
+      "grad_norm": 5.84375,
+      "learning_rate": 1.4148871673964299e-05,
+      "loss": 0.0739,
+      "step": 10250
+    },
+    {
+      "epoch": 0.32828848430550667,
+      "grad_norm": 23.75,
+      "learning_rate": 1.4142135399124287e-05,
+      "loss": 0.1272,
+      "step": 10260
+    },
+    {
+      "epoch": 0.32860845358845553,
+      "grad_norm": 8.875,
+      "learning_rate": 1.4135399124284273e-05,
+      "loss": 0.1684,
+      "step": 10270
+    },
+    {
+      "epoch": 0.32892842287140434,
+      "grad_norm": 26.5,
+      "learning_rate": 1.4128662849444259e-05,
+      "loss": 0.1469,
+      "step": 10280
+    },
+    {
+      "epoch": 0.3292483921543532,
+      "grad_norm": 11.75,
+      "learning_rate": 1.4121926574604245e-05,
+      "loss": 0.1232,
+      "step": 10290
+    },
+    {
+      "epoch": 0.329568361437302,
+      "grad_norm": 17.0,
+      "learning_rate": 1.4115190299764233e-05,
+      "loss": 0.1062,
+      "step": 10300
+    },
+    {
+      "epoch": 0.32988833072025087,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.4108454024924219e-05,
+      "loss": 0.0957,
+      "step": 10310
+    },
+    {
+      "epoch": 0.3302083000031997,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.4101717750084205e-05,
+      "loss": 0.1465,
+      "step": 10320
+    },
+    {
+      "epoch": 0.33052826928614853,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.4094981475244192e-05,
+      "loss": 0.1535,
+      "step": 10330
+    },
+    {
+      "epoch": 0.33084823856909734,
+      "grad_norm": 9.75,
+      "learning_rate": 1.4088245200404178e-05,
+      "loss": 0.1099,
+      "step": 10340
+    },
+    {
+      "epoch": 0.3311682078520462,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.4081508925564164e-05,
+      "loss": 0.137,
+      "step": 10350
+    },
+    {
+      "epoch": 0.33148817713499507,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.407477265072415e-05,
+      "loss": 0.0814,
+      "step": 10360
+    },
+    {
+      "epoch": 0.33180814641794387,
+      "grad_norm": 10.25,
+      "learning_rate": 1.4068036375884138e-05,
+      "loss": 0.0976,
+      "step": 10370
+    },
+    {
+      "epoch": 0.33212811570089273,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.4061300101044124e-05,
+      "loss": 0.0583,
+      "step": 10380
+    },
+    {
+      "epoch": 0.33244808498384154,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.405456382620411e-05,
+      "loss": 0.1555,
+      "step": 10390
+    },
+    {
+      "epoch": 0.3327680542667904,
+      "grad_norm": 9.0,
+      "learning_rate": 1.4047827551364096e-05,
+      "loss": 0.2356,
+      "step": 10400
+    },
+    {
+      "epoch": 0.3330880235497392,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.4041091276524084e-05,
+      "loss": 0.0423,
+      "step": 10410
+    },
+    {
+      "epoch": 0.33340799283268807,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.403435500168407e-05,
+      "loss": 0.1153,
+      "step": 10420
+    },
+    {
+      "epoch": 0.3337279621156369,
+      "grad_norm": 18.5,
+      "learning_rate": 1.4027618726844056e-05,
+      "loss": 0.132,
+      "step": 10430
+    },
+    {
+      "epoch": 0.33404793139858574,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.4020882452004044e-05,
+      "loss": 0.0954,
+      "step": 10440
+    },
+    {
+      "epoch": 0.33436790068153455,
+      "grad_norm": 20.375,
+      "learning_rate": 1.401414617716403e-05,
+      "loss": 0.1082,
+      "step": 10450
+    },
+    {
+      "epoch": 0.3346878699644834,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4007409902324016e-05,
+      "loss": 0.1149,
+      "step": 10460
+    },
+    {
+      "epoch": 0.33500783924743227,
+      "grad_norm": 8.875,
+      "learning_rate": 1.4000673627484002e-05,
+      "loss": 0.1498,
+      "step": 10470
+    },
+    {
+      "epoch": 0.3353278085303811,
+      "grad_norm": 8.25,
+      "learning_rate": 1.399393735264399e-05,
+      "loss": 0.1202,
+      "step": 10480
+    },
+    {
+      "epoch": 0.33564777781332994,
+      "grad_norm": 34.5,
+      "learning_rate": 1.3987201077803976e-05,
+      "loss": 0.1411,
+      "step": 10490
+    },
+    {
+      "epoch": 0.33596774709627875,
+      "grad_norm": 9.375,
+      "learning_rate": 1.3980464802963962e-05,
+      "loss": 0.0587,
+      "step": 10500
+    },
+    {
+      "epoch": 0.3362877163792276,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.397372852812395e-05,
+      "loss": 0.0687,
+      "step": 10510
+    },
+    {
+      "epoch": 0.3366076856621764,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.3966992253283936e-05,
+      "loss": 0.1296,
+      "step": 10520
+    },
+    {
+      "epoch": 0.3369276549451253,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.3960255978443922e-05,
+      "loss": 0.1319,
+      "step": 10530
+    },
+    {
+      "epoch": 0.3372476242280741,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.3953519703603908e-05,
+      "loss": 0.1231,
+      "step": 10540
+    },
+    {
+      "epoch": 0.33756759351102295,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3946783428763895e-05,
+      "loss": 0.1108,
+      "step": 10550
+    },
+    {
+      "epoch": 0.33788756279397175,
+      "grad_norm": 10.375,
+      "learning_rate": 1.3940047153923881e-05,
+      "loss": 0.1038,
+      "step": 10560
+    },
+    {
+      "epoch": 0.3382075320769206,
+      "grad_norm": 11.125,
+      "learning_rate": 1.3933310879083868e-05,
+      "loss": 0.1357,
+      "step": 10570
+    },
+    {
+      "epoch": 0.3385275013598695,
+      "grad_norm": 12.875,
+      "learning_rate": 1.3926574604243854e-05,
+      "loss": 0.1081,
+      "step": 10580
+    },
+    {
+      "epoch": 0.3388474706428183,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.3919838329403841e-05,
+      "loss": 0.1027,
+      "step": 10590
+    },
+    {
+      "epoch": 0.33916743992576714,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3913102054563827e-05,
+      "loss": 0.1465,
+      "step": 10600
+    },
+    {
+      "epoch": 0.33948740920871595,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.3906365779723813e-05,
+      "loss": 0.1377,
+      "step": 10610
+    },
+    {
+      "epoch": 0.3398073784916648,
+      "grad_norm": 6.5,
+      "learning_rate": 1.3899629504883801e-05,
+      "loss": 0.1005,
+      "step": 10620
+    },
+    {
+      "epoch": 0.3401273477746136,
+      "grad_norm": 13.0,
+      "learning_rate": 1.3892893230043787e-05,
+      "loss": 0.146,
+      "step": 10630
+    },
+    {
+      "epoch": 0.3404473170575625,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.3886156955203773e-05,
+      "loss": 0.117,
+      "step": 10640
+    },
+    {
+      "epoch": 0.3407672863405113,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.387942068036376e-05,
+      "loss": 0.0816,
+      "step": 10650
+    },
+    {
+      "epoch": 0.34108725562346015,
+      "grad_norm": 18.25,
+      "learning_rate": 1.3872684405523747e-05,
+      "loss": 0.1183,
+      "step": 10660
+    },
+    {
+      "epoch": 0.34140722490640896,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.3865948130683733e-05,
+      "loss": 0.1237,
+      "step": 10670
+    },
+    {
+      "epoch": 0.3417271941893578,
+      "grad_norm": 4.59375,
+      "learning_rate": 1.3859211855843719e-05,
+      "loss": 0.1043,
+      "step": 10680
+    },
+    {
+      "epoch": 0.3420471634723067,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.3852475581003705e-05,
+      "loss": 0.0968,
+      "step": 10690
+    },
+    {
+      "epoch": 0.3423671327552555,
+      "grad_norm": 8.5,
+      "learning_rate": 1.3845739306163693e-05,
+      "loss": 0.1448,
+      "step": 10700
+    },
+    {
+      "epoch": 0.34268710203820435,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.3839003031323679e-05,
+      "loss": 0.1658,
+      "step": 10710
+    },
+    {
+      "epoch": 0.34300707132115316,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.3832266756483665e-05,
+      "loss": 0.1195,
+      "step": 10720
+    },
+    {
+      "epoch": 0.343327040604102,
+      "grad_norm": 14.5,
+      "learning_rate": 1.3825530481643653e-05,
+      "loss": 0.1541,
+      "step": 10730
+    },
+    {
+      "epoch": 0.3436470098870508,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.3818794206803639e-05,
+      "loss": 0.1049,
+      "step": 10740
+    },
+    {
+      "epoch": 0.3439669791699997,
+      "grad_norm": 6.75,
+      "learning_rate": 1.3812057931963625e-05,
+      "loss": 0.1166,
+      "step": 10750
+    },
+    {
+      "epoch": 0.3442869484529485,
+      "grad_norm": 6.25,
+      "learning_rate": 1.380532165712361e-05,
+      "loss": 0.1394,
+      "step": 10760
+    },
+    {
+      "epoch": 0.34460691773589736,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.3798585382283599e-05,
+      "loss": 0.1192,
+      "step": 10770
+    },
+    {
+      "epoch": 0.3449268870188462,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.3791849107443585e-05,
+      "loss": 0.1256,
+      "step": 10780
+    },
+    {
+      "epoch": 0.345246856301795,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.378511283260357e-05,
+      "loss": 0.1219,
+      "step": 10790
+    },
+    {
+      "epoch": 0.3455668255847439,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3778376557763558e-05,
+      "loss": 0.091,
+      "step": 10800
+    },
+    {
+      "epoch": 0.3458867948676927,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.3771640282923544e-05,
+      "loss": 0.0598,
+      "step": 10810
+    },
+    {
+      "epoch": 0.34620676415064155,
+      "grad_norm": 19.375,
+      "learning_rate": 1.376490400808353e-05,
+      "loss": 0.1141,
+      "step": 10820
+    },
+    {
+      "epoch": 0.34652673343359036,
+      "grad_norm": 17.125,
+      "learning_rate": 1.3758167733243516e-05,
+      "loss": 0.112,
+      "step": 10830
+    },
+    {
+      "epoch": 0.3468467027165392,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.3751431458403504e-05,
+      "loss": 0.1354,
+      "step": 10840
+    },
+    {
+      "epoch": 0.34716667199948803,
+      "grad_norm": 6.875,
+      "learning_rate": 1.374469518356349e-05,
+      "loss": 0.0969,
+      "step": 10850
+    },
+    {
+      "epoch": 0.3474866412824369,
+      "grad_norm": 2.125,
+      "learning_rate": 1.3737958908723476e-05,
+      "loss": 0.0671,
+      "step": 10860
+    },
+    {
+      "epoch": 0.3478066105653857,
+      "grad_norm": 4.46875,
+      "learning_rate": 1.3731222633883462e-05,
+      "loss": 0.1177,
+      "step": 10870
+    },
+    {
+      "epoch": 0.34812657984833456,
+      "grad_norm": 4.1875,
+      "learning_rate": 1.372448635904345e-05,
+      "loss": 0.1349,
+      "step": 10880
+    },
+    {
+      "epoch": 0.3484465491312834,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.3717750084203436e-05,
+      "loss": 0.0932,
+      "step": 10890
+    },
+    {
+      "epoch": 0.34876651841423223,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.3711013809363422e-05,
+      "loss": 0.1284,
+      "step": 10900
+    },
+    {
+      "epoch": 0.3490864876971811,
+      "grad_norm": 10.25,
+      "learning_rate": 1.370427753452341e-05,
+      "loss": 0.097,
+      "step": 10910
+    },
+    {
+      "epoch": 0.3494064569801299,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.3697541259683396e-05,
+      "loss": 0.1912,
+      "step": 10920
+    },
+    {
+      "epoch": 0.34972642626307876,
+      "grad_norm": 1.0,
+      "learning_rate": 1.3690804984843382e-05,
+      "loss": 0.096,
+      "step": 10930
+    },
+    {
+      "epoch": 0.35004639554602757,
+      "grad_norm": 4.5625,
+      "learning_rate": 1.3684068710003368e-05,
+      "loss": 0.099,
+      "step": 10940
+    },
+    {
+      "epoch": 0.35036636482897643,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.3677332435163356e-05,
+      "loss": 0.2057,
+      "step": 10950
+    },
+    {
+      "epoch": 0.35068633411192524,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.3670596160323342e-05,
+      "loss": 0.1023,
+      "step": 10960
+    },
+    {
+      "epoch": 0.3510063033948741,
+      "grad_norm": 10.0,
+      "learning_rate": 1.3663859885483328e-05,
+      "loss": 0.1038,
+      "step": 10970
+    },
+    {
+      "epoch": 0.3513262726778229,
+      "grad_norm": 17.125,
+      "learning_rate": 1.3657123610643316e-05,
+      "loss": 0.125,
+      "step": 10980
+    },
+    {
+      "epoch": 0.35164624196077177,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.3650387335803302e-05,
+      "loss": 0.0943,
+      "step": 10990
+    },
+    {
+      "epoch": 0.35196621124372063,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.3643651060963288e-05,
+      "loss": 0.1054,
+      "step": 11000
+    },
+    {
+      "epoch": 0.35228618052666943,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.3636914786123274e-05,
+      "loss": 0.0683,
+      "step": 11010
+    },
+    {
+      "epoch": 0.3526061498096183,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3630178511283261e-05,
+      "loss": 0.1136,
+      "step": 11020
+    },
+    {
+      "epoch": 0.3529261190925671,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.3623442236443247e-05,
+      "loss": 0.2113,
+      "step": 11030
+    },
+    {
+      "epoch": 0.35324608837551597,
+      "grad_norm": 7.625,
+      "learning_rate": 1.3616705961603233e-05,
+      "loss": 0.1611,
+      "step": 11040
+    },
+    {
+      "epoch": 0.35356605765846477,
+      "grad_norm": 8.0,
+      "learning_rate": 1.360996968676322e-05,
+      "loss": 0.0868,
+      "step": 11050
+    },
+    {
+      "epoch": 0.35388602694141363,
+      "grad_norm": 12.5,
+      "learning_rate": 1.3603233411923207e-05,
+      "loss": 0.09,
+      "step": 11060
+    },
+    {
+      "epoch": 0.35420599622436244,
+      "grad_norm": 19.375,
+      "learning_rate": 1.3596497137083193e-05,
+      "loss": 0.0934,
+      "step": 11070
+    },
+    {
+      "epoch": 0.3545259655073113,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.358976086224318e-05,
+      "loss": 0.0803,
+      "step": 11080
+    },
+    {
+      "epoch": 0.3548459347902601,
+      "grad_norm": 6.15625,
+      "learning_rate": 1.3583024587403167e-05,
+      "loss": 0.1367,
+      "step": 11090
+    },
+    {
+      "epoch": 0.35516590407320897,
+      "grad_norm": 6.15625,
+      "learning_rate": 1.3576288312563153e-05,
+      "loss": 0.0836,
+      "step": 11100
+    },
+    {
+      "epoch": 0.35548587335615783,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.356955203772314e-05,
+      "loss": 0.0831,
+      "step": 11110
+    },
+    {
+      "epoch": 0.35580584263910664,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.3562815762883125e-05,
+      "loss": 0.1429,
+      "step": 11120
+    },
+    {
+      "epoch": 0.3561258119220555,
+      "grad_norm": 4.71875,
+      "learning_rate": 1.3556079488043113e-05,
+      "loss": 0.1425,
+      "step": 11130
+    },
+    {
+      "epoch": 0.3564457812050043,
+      "grad_norm": 12.375,
+      "learning_rate": 1.3549343213203099e-05,
+      "loss": 0.1023,
+      "step": 11140
+    },
+    {
+      "epoch": 0.35676575048795317,
+      "grad_norm": 4.5,
+      "learning_rate": 1.3542606938363085e-05,
+      "loss": 0.0813,
+      "step": 11150
+    },
+    {
+      "epoch": 0.357085719770902,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.3535870663523071e-05,
+      "loss": 0.0891,
+      "step": 11160
+    },
+    {
+      "epoch": 0.35740568905385084,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.3529134388683059e-05,
+      "loss": 0.0813,
+      "step": 11170
+    },
+    {
+      "epoch": 0.35772565833679965,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.3522398113843045e-05,
+      "loss": 0.144,
+      "step": 11180
+    },
+    {
+      "epoch": 0.3580456276197485,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.3515661839003031e-05,
+      "loss": 0.0854,
+      "step": 11190
+    },
+    {
+      "epoch": 0.3583655969026973,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3508925564163019e-05,
+      "loss": 0.0996,
+      "step": 11200
+    },
+    {
+      "epoch": 0.3586855661856462,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.3502189289323005e-05,
+      "loss": 0.098,
+      "step": 11210
+    },
+    {
+      "epoch": 0.35900553546859504,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.349545301448299e-05,
+      "loss": 0.0949,
+      "step": 11220
+    },
+    {
+      "epoch": 0.35932550475154384,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3488716739642977e-05,
+      "loss": 0.1067,
+      "step": 11230
+    },
+    {
+      "epoch": 0.3596454740344927,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.3481980464802964e-05,
+      "loss": 0.0982,
+      "step": 11240
+    },
+    {
+      "epoch": 0.3599654433174415,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.3475244189962952e-05,
+      "loss": 0.1197,
+      "step": 11250
+    },
+    {
+      "epoch": 0.3602854126003904,
+      "grad_norm": 3.25,
+      "learning_rate": 1.346850791512294e-05,
+      "loss": 0.0754,
+      "step": 11260
+    },
+    {
+      "epoch": 0.3606053818833392,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.3461771640282926e-05,
+      "loss": 0.1188,
+      "step": 11270
+    },
+    {
+      "epoch": 0.36092535116628804,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3455035365442912e-05,
+      "loss": 0.1126,
+      "step": 11280
+    },
+    {
+      "epoch": 0.36124532044923685,
+      "grad_norm": 6.03125,
+      "learning_rate": 1.3448299090602898e-05,
+      "loss": 0.1263,
+      "step": 11290
+    },
+    {
+      "epoch": 0.3615652897321857,
+      "grad_norm": 5.90625,
+      "learning_rate": 1.3441562815762886e-05,
+      "loss": 0.1119,
+      "step": 11300
+    },
+    {
+      "epoch": 0.3618852590151345,
+      "grad_norm": 25.0,
+      "learning_rate": 1.3434826540922872e-05,
+      "loss": 0.107,
+      "step": 11310
+    },
+    {
+      "epoch": 0.3622052282980834,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.3428090266082858e-05,
+      "loss": 0.1341,
+      "step": 11320
+    },
+    {
+      "epoch": 0.36252519758103224,
+      "grad_norm": 9.25,
+      "learning_rate": 1.3421353991242844e-05,
+      "loss": 0.1262,
+      "step": 11330
+    },
+    {
+      "epoch": 0.36284516686398105,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.3414617716402832e-05,
+      "loss": 0.0878,
+      "step": 11340
+    },
+    {
+      "epoch": 0.3631651361469299,
+      "grad_norm": 2.25,
+      "learning_rate": 1.3407881441562818e-05,
+      "loss": 0.0722,
+      "step": 11350
+    },
+    {
+      "epoch": 0.3634851054298787,
+      "grad_norm": 29.625,
+      "learning_rate": 1.3401145166722804e-05,
+      "loss": 0.1658,
+      "step": 11360
+    },
+    {
+      "epoch": 0.3638050747128276,
+      "grad_norm": 4.875,
+      "learning_rate": 1.3394408891882791e-05,
+      "loss": 0.0945,
+      "step": 11370
+    },
+    {
+      "epoch": 0.3641250439957764,
+      "grad_norm": 6.25,
+      "learning_rate": 1.3387672617042778e-05,
+      "loss": 0.1189,
+      "step": 11380
+    },
+    {
+      "epoch": 0.36444501327872525,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.3380936342202764e-05,
+      "loss": 0.1067,
+      "step": 11390
+    },
+    {
+      "epoch": 0.36476498256167406,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.337420006736275e-05,
+      "loss": 0.1142,
+      "step": 11400
+    },
+    {
+      "epoch": 0.3650849518446229,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3367463792522737e-05,
+      "loss": 0.1694,
+      "step": 11410
+    },
+    {
+      "epoch": 0.3654049211275718,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.3360727517682723e-05,
+      "loss": 0.0988,
+      "step": 11420
+    },
+    {
+      "epoch": 0.3657248904105206,
+      "grad_norm": 5.4375,
+      "learning_rate": 1.335399124284271e-05,
+      "loss": 0.0628,
+      "step": 11430
+    },
+    {
+      "epoch": 0.36604485969346945,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3347254968002697e-05,
+      "loss": 0.1021,
+      "step": 11440
+    },
+    {
+      "epoch": 0.36636482897641826,
+      "grad_norm": 23.5,
+      "learning_rate": 1.3340518693162683e-05,
+      "loss": 0.1237,
+      "step": 11450
+    },
+    {
+      "epoch": 0.3666847982593671,
+      "grad_norm": 17.0,
+      "learning_rate": 1.333378241832267e-05,
+      "loss": 0.143,
+      "step": 11460
+    },
+    {
+      "epoch": 0.3670047675423159,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.3327046143482655e-05,
+      "loss": 0.2044,
+      "step": 11470
+    },
+    {
+      "epoch": 0.3673247368252648,
+      "grad_norm": 6.28125,
+      "learning_rate": 1.3320309868642643e-05,
+      "loss": 0.0776,
+      "step": 11480
+    },
+    {
+      "epoch": 0.3676447061082136,
+      "grad_norm": 5.15625,
+      "learning_rate": 1.3313573593802629e-05,
+      "loss": 0.0917,
+      "step": 11490
+    },
+    {
+      "epoch": 0.36796467539116245,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.3306837318962615e-05,
+      "loss": 0.1108,
+      "step": 11500
+    },
+    {
+      "epoch": 0.36828464467411126,
+      "grad_norm": 6.875,
+      "learning_rate": 1.3300101044122601e-05,
+      "loss": 0.0931,
+      "step": 11510
+    },
+    {
+      "epoch": 0.3686046139570601,
+      "grad_norm": 16.25,
+      "learning_rate": 1.3293364769282589e-05,
+      "loss": 0.1893,
+      "step": 11520
+    },
+    {
+      "epoch": 0.368924583240009,
+      "grad_norm": 5.40625,
+      "learning_rate": 1.3286628494442575e-05,
+      "loss": 0.0839,
+      "step": 11530
+    },
+    {
+      "epoch": 0.3692445525229578,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.3279892219602561e-05,
+      "loss": 0.0858,
+      "step": 11540
+    },
+    {
+      "epoch": 0.36956452180590665,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3273155944762549e-05,
+      "loss": 0.0876,
+      "step": 11550
+    },
+    {
+      "epoch": 0.36988449108885546,
+      "grad_norm": 17.375,
+      "learning_rate": 1.3266419669922535e-05,
+      "loss": 0.181,
+      "step": 11560
+    },
+    {
+      "epoch": 0.3702044603718043,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.325968339508252e-05,
+      "loss": 0.1222,
+      "step": 11570
+    },
+    {
+      "epoch": 0.37052442965475313,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.3252947120242507e-05,
+      "loss": 0.1103,
+      "step": 11580
+    },
+    {
+      "epoch": 0.370844398937702,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3246210845402495e-05,
+      "loss": 0.0888,
+      "step": 11590
+    },
+    {
+      "epoch": 0.3711643682206508,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.323947457056248e-05,
+      "loss": 0.1216,
+      "step": 11600
+    },
+    {
+      "epoch": 0.37148433750359966,
+      "grad_norm": 20.875,
+      "learning_rate": 1.3232738295722467e-05,
+      "loss": 0.1632,
+      "step": 11610
+    },
+    {
+      "epoch": 0.37180430678654847,
+      "grad_norm": 5.40625,
+      "learning_rate": 1.3226002020882454e-05,
+      "loss": 0.1715,
+      "step": 11620
+    },
+    {
+      "epoch": 0.37212427606949733,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.321926574604244e-05,
+      "loss": 0.1272,
+      "step": 11630
+    },
+    {
+      "epoch": 0.3724442453524462,
+      "grad_norm": 60.25,
+      "learning_rate": 1.3212529471202426e-05,
+      "loss": 0.1158,
+      "step": 11640
+    },
+    {
+      "epoch": 0.372764214635395,
+      "grad_norm": 9.5,
+      "learning_rate": 1.3205793196362413e-05,
+      "loss": 0.1397,
+      "step": 11650
+    },
+    {
+      "epoch": 0.37308418391834386,
+      "grad_norm": 22.5,
+      "learning_rate": 1.31990569215224e-05,
+      "loss": 0.1447,
+      "step": 11660
+    },
+    {
+      "epoch": 0.37340415320129267,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.3192320646682386e-05,
+      "loss": 0.11,
+      "step": 11670
+    },
+    {
+      "epoch": 0.3737241224842415,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3185584371842372e-05,
+      "loss": 0.1564,
+      "step": 11680
+    },
+    {
+      "epoch": 0.37404409176719033,
+      "grad_norm": 18.625,
+      "learning_rate": 1.3178848097002358e-05,
+      "loss": 0.1502,
+      "step": 11690
+    },
+    {
+      "epoch": 0.3743640610501392,
+      "grad_norm": 21.625,
+      "learning_rate": 1.3172111822162346e-05,
+      "loss": 0.1343,
+      "step": 11700
+    },
+    {
+      "epoch": 0.374684030333088,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3165375547322332e-05,
+      "loss": 0.0997,
+      "step": 11710
+    },
+    {
+      "epoch": 0.37500399961603686,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.3158639272482318e-05,
+      "loss": 0.087,
+      "step": 11720
+    },
+    {
+      "epoch": 0.37532396889898567,
+      "grad_norm": 12.125,
+      "learning_rate": 1.3151902997642306e-05,
+      "loss": 0.1206,
+      "step": 11730
+    },
+    {
+      "epoch": 0.37564393818193453,
+      "grad_norm": 21.125,
+      "learning_rate": 1.3145166722802292e-05,
+      "loss": 0.1031,
+      "step": 11740
+    },
+    {
+      "epoch": 0.3759639074648834,
+      "grad_norm": 9.75,
+      "learning_rate": 1.3138430447962278e-05,
+      "loss": 0.0723,
+      "step": 11750
+    },
+    {
+      "epoch": 0.3762838767478322,
+      "grad_norm": 5.90625,
+      "learning_rate": 1.3131694173122264e-05,
+      "loss": 0.1148,
+      "step": 11760
+    },
+    {
+      "epoch": 0.37660384603078106,
+      "grad_norm": 66.0,
+      "learning_rate": 1.3124957898282252e-05,
+      "loss": 0.1133,
+      "step": 11770
+    },
+    {
+      "epoch": 0.37692381531372987,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.3118221623442238e-05,
+      "loss": 0.0676,
+      "step": 11780
+    },
+    {
+      "epoch": 0.37724378459667873,
+      "grad_norm": 31.0,
+      "learning_rate": 1.3111485348602224e-05,
+      "loss": 0.1,
+      "step": 11790
+    },
+    {
+      "epoch": 0.37756375387962754,
+      "grad_norm": 1.375,
+      "learning_rate": 1.310474907376221e-05,
+      "loss": 0.0846,
+      "step": 11800
+    },
+    {
+      "epoch": 0.3778837231625764,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.3098012798922198e-05,
+      "loss": 0.0763,
+      "step": 11810
+    },
+    {
+      "epoch": 0.3782036924455252,
+      "grad_norm": 6.53125,
+      "learning_rate": 1.3091276524082184e-05,
+      "loss": 0.1412,
+      "step": 11820
+    },
+    {
+      "epoch": 0.37852366172847407,
+      "grad_norm": 18.5,
+      "learning_rate": 1.308454024924217e-05,
+      "loss": 0.16,
+      "step": 11830
+    },
+    {
+      "epoch": 0.3788436310114229,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.3077803974402157e-05,
+      "loss": 0.1328,
+      "step": 11840
+    },
+    {
+      "epoch": 0.37916360029437174,
+      "grad_norm": 20.0,
+      "learning_rate": 1.3071067699562143e-05,
+      "loss": 0.1119,
+      "step": 11850
+    },
+    {
+      "epoch": 0.3794835695773206,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.306433142472213e-05,
+      "loss": 0.1059,
+      "step": 11860
+    },
+    {
+      "epoch": 0.3798035388602694,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.3057595149882116e-05,
+      "loss": 0.1181,
+      "step": 11870
+    },
+    {
+      "epoch": 0.38012350814321827,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.3050858875042103e-05,
+      "loss": 0.1044,
+      "step": 11880
+    },
+    {
+      "epoch": 0.3804434774261671,
+      "grad_norm": 9.375,
+      "learning_rate": 1.304412260020209e-05,
+      "loss": 0.1582,
+      "step": 11890
+    },
+    {
+      "epoch": 0.38076344670911594,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.3037386325362075e-05,
+      "loss": 0.0781,
+      "step": 11900
+    },
+    {
+      "epoch": 0.38108341599206474,
+      "grad_norm": 9.0,
+      "learning_rate": 1.3030650050522063e-05,
+      "loss": 0.1283,
+      "step": 11910
+    },
+    {
+      "epoch": 0.3814033852750136,
+      "grad_norm": 28.5,
+      "learning_rate": 1.302391377568205e-05,
+      "loss": 0.152,
+      "step": 11920
+    },
+    {
+      "epoch": 0.3817233545579624,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.3017177500842035e-05,
+      "loss": 0.0666,
+      "step": 11930
+    },
+    {
+      "epoch": 0.3820433238409113,
+      "grad_norm": 24.5,
+      "learning_rate": 1.3010441226002021e-05,
+      "loss": 0.1481,
+      "step": 11940
+    },
+    {
+      "epoch": 0.3823632931238601,
+      "grad_norm": 5.40625,
+      "learning_rate": 1.3003704951162009e-05,
+      "loss": 0.1329,
+      "step": 11950
+    },
+    {
+      "epoch": 0.38268326240680894,
+      "grad_norm": 31.75,
+      "learning_rate": 1.2996968676321995e-05,
+      "loss": 0.1511,
+      "step": 11960
+    },
+    {
+      "epoch": 0.3830032316897578,
+      "grad_norm": 5.78125,
+      "learning_rate": 1.2990232401481981e-05,
+      "loss": 0.0712,
+      "step": 11970
+    },
+    {
+      "epoch": 0.3833232009727066,
+      "grad_norm": 5.75,
+      "learning_rate": 1.2983496126641967e-05,
+      "loss": 0.1224,
+      "step": 11980
+    },
+    {
+      "epoch": 0.3836431702556555,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.2976759851801955e-05,
+      "loss": 0.0876,
+      "step": 11990
+    },
+    {
+      "epoch": 0.3839631395386043,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.2970023576961941e-05,
+      "loss": 0.1448,
+      "step": 12000
+    },
+    {
+      "epoch": 0.38428310882155314,
+      "grad_norm": 10.0,
+      "learning_rate": 1.2963287302121927e-05,
+      "loss": 0.1001,
+      "step": 12010
+    },
+    {
+      "epoch": 0.38460307810450195,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.2956551027281915e-05,
+      "loss": 0.1316,
+      "step": 12020
+    },
+    {
+      "epoch": 0.3849230473874508,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.29498147524419e-05,
+      "loss": 0.1228,
+      "step": 12030
+    },
+    {
+      "epoch": 0.3852430166703996,
+      "grad_norm": 3.125,
+      "learning_rate": 1.2943078477601887e-05,
+      "loss": 0.0713,
+      "step": 12040
+    },
+    {
+      "epoch": 0.3855629859533485,
+      "grad_norm": 21.875,
+      "learning_rate": 1.2936342202761873e-05,
+      "loss": 0.1133,
+      "step": 12050
+    },
+    {
+      "epoch": 0.38588295523629734,
+      "grad_norm": 16.625,
+      "learning_rate": 1.292960592792186e-05,
+      "loss": 0.1818,
+      "step": 12060
+    },
+    {
+      "epoch": 0.38620292451924615,
+      "grad_norm": 17.25,
+      "learning_rate": 1.2922869653081847e-05,
+      "loss": 0.0688,
+      "step": 12070
+    },
+    {
+      "epoch": 0.386522893802195,
+      "grad_norm": 15.375,
+      "learning_rate": 1.2916133378241833e-05,
+      "loss": 0.0991,
+      "step": 12080
+    },
+    {
+      "epoch": 0.3868428630851438,
+      "grad_norm": 13.625,
+      "learning_rate": 1.290939710340182e-05,
+      "loss": 0.1509,
+      "step": 12090
+    },
+    {
+      "epoch": 0.3871628323680927,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.2902660828561806e-05,
+      "loss": 0.1587,
+      "step": 12100
+    },
+    {
+      "epoch": 0.3874828016510415,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.2895924553721792e-05,
+      "loss": 0.1229,
+      "step": 12110
+    },
+    {
+      "epoch": 0.38780277093399035,
+      "grad_norm": 5.5,
+      "learning_rate": 1.2889188278881778e-05,
+      "loss": 0.1131,
+      "step": 12120
+    },
+    {
+      "epoch": 0.38812274021693915,
+      "grad_norm": 12.375,
+      "learning_rate": 1.2882452004041766e-05,
+      "loss": 0.1133,
+      "step": 12130
+    },
+    {
+      "epoch": 0.388442709499888,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.2875715729201752e-05,
+      "loss": 0.0825,
+      "step": 12140
+    },
+    {
+      "epoch": 0.3887626787828368,
+      "grad_norm": 6.625,
+      "learning_rate": 1.2868979454361738e-05,
+      "loss": 0.1165,
+      "step": 12150
+    },
+    {
+      "epoch": 0.3890826480657857,
+      "grad_norm": 10.25,
+      "learning_rate": 1.2862243179521724e-05,
+      "loss": 0.1629,
+      "step": 12160
+    },
+    {
+      "epoch": 0.38940261734873455,
+      "grad_norm": 10.0,
+      "learning_rate": 1.2855506904681712e-05,
+      "loss": 0.1254,
+      "step": 12170
+    },
+    {
+      "epoch": 0.38972258663168335,
+      "grad_norm": 16.375,
+      "learning_rate": 1.2848770629841698e-05,
+      "loss": 0.1573,
+      "step": 12180
+    },
+    {
+      "epoch": 0.3900425559146322,
+      "grad_norm": 7.75,
+      "learning_rate": 1.2842034355001684e-05,
+      "loss": 0.1404,
+      "step": 12190
+    },
+    {
+      "epoch": 0.390362525197581,
+      "grad_norm": 9.0,
+      "learning_rate": 1.2835298080161672e-05,
+      "loss": 0.0921,
+      "step": 12200
+    },
+    {
+      "epoch": 0.3906824944805299,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.2828561805321658e-05,
+      "loss": 0.1462,
+      "step": 12210
+    },
+    {
+      "epoch": 0.3910024637634787,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.2821825530481644e-05,
+      "loss": 0.0841,
+      "step": 12220
+    },
+    {
+      "epoch": 0.39132243304642755,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.281508925564163e-05,
+      "loss": 0.1122,
+      "step": 12230
+    },
+    {
+      "epoch": 0.39164240232937636,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.2808352980801618e-05,
+      "loss": 0.046,
+      "step": 12240
+    },
+    {
+      "epoch": 0.3919623716123252,
+      "grad_norm": 4.65625,
+      "learning_rate": 1.2801616705961604e-05,
+      "loss": 0.1462,
+      "step": 12250
+    },
+    {
+      "epoch": 0.39228234089527403,
+      "grad_norm": 5.09375,
+      "learning_rate": 1.279488043112159e-05,
+      "loss": 0.0608,
+      "step": 12260
+    },
+    {
+      "epoch": 0.3926023101782229,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2788144156281576e-05,
+      "loss": 0.1465,
+      "step": 12270
+    },
+    {
+      "epoch": 0.39292227946117175,
+      "grad_norm": 31.5,
+      "learning_rate": 1.2781407881441564e-05,
+      "loss": 0.1733,
+      "step": 12280
+    },
+    {
+      "epoch": 0.39324224874412056,
+      "grad_norm": 6.21875,
+      "learning_rate": 1.277467160660155e-05,
+      "loss": 0.1205,
+      "step": 12290
+    },
+    {
+      "epoch": 0.3935622180270694,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.2767935331761536e-05,
+      "loss": 0.1399,
+      "step": 12300
+    },
+    {
+      "epoch": 0.39388218731001823,
+      "grad_norm": 11.5,
+      "learning_rate": 1.2761199056921523e-05,
+      "loss": 0.2015,
+      "step": 12310
+    },
+    {
+      "epoch": 0.3942021565929671,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.275446278208151e-05,
+      "loss": 0.1061,
+      "step": 12320
+    },
+    {
+      "epoch": 0.3945221258759159,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.2747726507241496e-05,
+      "loss": 0.1066,
+      "step": 12330
+    },
+    {
+      "epoch": 0.39484209515886476,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.2740990232401482e-05,
+      "loss": 0.1269,
+      "step": 12340
+    },
+    {
+      "epoch": 0.39516206444181357,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.273425395756147e-05,
+      "loss": 0.1048,
+      "step": 12350
+    },
+    {
+      "epoch": 0.3954820337247624,
+      "grad_norm": 5.78125,
+      "learning_rate": 1.2727517682721455e-05,
+      "loss": 0.1347,
+      "step": 12360
+    },
+    {
+      "epoch": 0.39580200300771123,
+      "grad_norm": 21.25,
+      "learning_rate": 1.2720781407881441e-05,
+      "loss": 0.1056,
+      "step": 12370
+    },
+    {
+      "epoch": 0.3961219722906601,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2714045133041429e-05,
+      "loss": 0.0858,
+      "step": 12380
+    },
+    {
+      "epoch": 0.39644194157360896,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.2707308858201415e-05,
+      "loss": 0.0857,
+      "step": 12390
+    },
+    {
+      "epoch": 0.39676191085655776,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.2700572583361401e-05,
+      "loss": 0.1692,
+      "step": 12400
+    },
+    {
+      "epoch": 0.3970818801395066,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.2693836308521387e-05,
+      "loss": 0.1272,
+      "step": 12410
+    },
+    {
+      "epoch": 0.39740184942245543,
+      "grad_norm": 13.375,
+      "learning_rate": 1.2687100033681375e-05,
+      "loss": 0.1234,
+      "step": 12420
+    },
+    {
+      "epoch": 0.3977218187054043,
+      "grad_norm": 6.15625,
+      "learning_rate": 1.2680363758841361e-05,
+      "loss": 0.107,
+      "step": 12430
+    },
+    {
+      "epoch": 0.3980417879883531,
+      "grad_norm": 11.1875,
+      "learning_rate": 1.2673627484001347e-05,
+      "loss": 0.1873,
+      "step": 12440
+    },
+    {
+      "epoch": 0.39836175727130196,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.2666891209161333e-05,
+      "loss": 0.085,
+      "step": 12450
+    },
+    {
+      "epoch": 0.39868172655425077,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.266015493432132e-05,
+      "loss": 0.1116,
+      "step": 12460
+    },
+    {
+      "epoch": 0.39900169583719963,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.2653418659481307e-05,
+      "loss": 0.111,
+      "step": 12470
+    },
+    {
+      "epoch": 0.39932166512014844,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.2646682384641293e-05,
+      "loss": 0.0514,
+      "step": 12480
+    },
+    {
+      "epoch": 0.3996416344030973,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.263994610980128e-05,
+      "loss": 0.0951,
+      "step": 12490
+    },
+    {
+      "epoch": 0.39996160368604616,
+      "grad_norm": 6.59375,
+      "learning_rate": 1.2633209834961267e-05,
+      "loss": 0.1671,
+      "step": 12500
+    },
+    {
+      "epoch": 0.40028157296899497,
+      "grad_norm": 30.125,
+      "learning_rate": 1.2626473560121253e-05,
+      "loss": 0.1168,
+      "step": 12510
+    },
+    {
+      "epoch": 0.40060154225194383,
+      "grad_norm": 22.625,
+      "learning_rate": 1.2619737285281239e-05,
+      "loss": 0.0853,
+      "step": 12520
+    },
+    {
+      "epoch": 0.40092151153489264,
+      "grad_norm": 20.25,
+      "learning_rate": 1.2613001010441226e-05,
+      "loss": 0.138,
+      "step": 12530
+    },
+    {
+      "epoch": 0.4012414808178415,
+      "grad_norm": 23.125,
+      "learning_rate": 1.2606264735601214e-05,
+      "loss": 0.0791,
+      "step": 12540
+    },
+    {
+      "epoch": 0.4015614501007903,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.2599528460761202e-05,
+      "loss": 0.1199,
+      "step": 12550
+    },
+    {
+      "epoch": 0.40188141938373917,
+      "grad_norm": 19.75,
+      "learning_rate": 1.2592792185921188e-05,
+      "loss": 0.1116,
+      "step": 12560
+    },
+    {
+      "epoch": 0.402201388666688,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.2586055911081174e-05,
+      "loss": 0.1268,
+      "step": 12570
+    },
+    {
+      "epoch": 0.40252135794963684,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.257931963624116e-05,
+      "loss": 0.1817,
+      "step": 12580
+    },
+    {
+      "epoch": 0.40284132723258564,
+      "grad_norm": 22.25,
+      "learning_rate": 1.2572583361401148e-05,
+      "loss": 0.0754,
+      "step": 12590
+    },
+    {
+      "epoch": 0.4031612965155345,
+      "grad_norm": 13.625,
+      "learning_rate": 1.2565847086561134e-05,
+      "loss": 0.1348,
+      "step": 12600
+    },
+    {
+      "epoch": 0.40348126579848337,
+      "grad_norm": 5.96875,
+      "learning_rate": 1.255911081172112e-05,
+      "loss": 0.0907,
+      "step": 12610
+    },
+    {
+      "epoch": 0.4038012350814322,
+      "grad_norm": 11.0,
+      "learning_rate": 1.2552374536881106e-05,
+      "loss": 0.1318,
+      "step": 12620
+    },
+    {
+      "epoch": 0.40412120436438104,
+      "grad_norm": 10.25,
+      "learning_rate": 1.2545638262041094e-05,
+      "loss": 0.1327,
+      "step": 12630
+    },
+    {
+      "epoch": 0.40444117364732984,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.253890198720108e-05,
+      "loss": 0.1387,
+      "step": 12640
+    },
+    {
+      "epoch": 0.4047611429302787,
+      "grad_norm": 15.25,
+      "learning_rate": 1.2532165712361066e-05,
+      "loss": 0.1375,
+      "step": 12650
+    },
+    {
+      "epoch": 0.4050811122132275,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.2525429437521054e-05,
+      "loss": 0.0695,
+      "step": 12660
+    },
+    {
+      "epoch": 0.4054010814961764,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.251869316268104e-05,
+      "loss": 0.1343,
+      "step": 12670
+    },
+    {
+      "epoch": 0.4057210507791252,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.2511956887841026e-05,
+      "loss": 0.0992,
+      "step": 12680
+    },
+    {
+      "epoch": 0.40604102006207404,
+      "grad_norm": 11.625,
+      "learning_rate": 1.2505220613001012e-05,
+      "loss": 0.1291,
+      "step": 12690
+    },
+    {
+      "epoch": 0.4063609893450229,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.2498484338161e-05,
+      "loss": 0.0713,
+      "step": 12700
+    },
+    {
+      "epoch": 0.4066809586279717,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.2491748063320985e-05,
+      "loss": 0.1136,
+      "step": 12710
+    },
+    {
+      "epoch": 0.4070009279109206,
+      "grad_norm": 10.25,
+      "learning_rate": 1.2485011788480971e-05,
+      "loss": 0.104,
+      "step": 12720
+    },
+    {
+      "epoch": 0.4073208971938694,
+      "grad_norm": 10.625,
+      "learning_rate": 1.247827551364096e-05,
+      "loss": 0.1258,
+      "step": 12730
+    },
+    {
+      "epoch": 0.40764086647681824,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.2471539238800945e-05,
+      "loss": 0.1554,
+      "step": 12740
+    },
+    {
+      "epoch": 0.40796083575976705,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2464802963960931e-05,
+      "loss": 0.1023,
+      "step": 12750
+    },
+    {
+      "epoch": 0.4082808050427159,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.2458066689120917e-05,
+      "loss": 0.1532,
+      "step": 12760
+    },
+    {
+      "epoch": 0.4086007743256647,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.2451330414280905e-05,
+      "loss": 0.1538,
+      "step": 12770
+    },
+    {
+      "epoch": 0.4089207436086136,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2444594139440891e-05,
+      "loss": 0.0604,
+      "step": 12780
+    },
+    {
+      "epoch": 0.4092407128915624,
+      "grad_norm": 1.0,
+      "learning_rate": 1.2437857864600877e-05,
+      "loss": 0.1201,
+      "step": 12790
+    },
+    {
+      "epoch": 0.40956068217451125,
+      "grad_norm": 10.5,
+      "learning_rate": 1.2431121589760863e-05,
+      "loss": 0.1449,
+      "step": 12800
+    },
+    {
+      "epoch": 0.4098806514574601,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.2424385314920851e-05,
+      "loss": 0.1129,
+      "step": 12810
+    },
+    {
+      "epoch": 0.4102006207404089,
+      "grad_norm": 28.875,
+      "learning_rate": 1.2417649040080837e-05,
+      "loss": 0.0656,
+      "step": 12820
+    },
+    {
+      "epoch": 0.4105205900233578,
+      "grad_norm": 14.75,
+      "learning_rate": 1.2410912765240823e-05,
+      "loss": 0.072,
+      "step": 12830
+    },
+    {
+      "epoch": 0.4108405593063066,
+      "grad_norm": 17.625,
+      "learning_rate": 1.240417649040081e-05,
+      "loss": 0.0477,
+      "step": 12840
+    },
+    {
+      "epoch": 0.41116052858925545,
+      "grad_norm": 18.75,
+      "learning_rate": 1.2397440215560797e-05,
+      "loss": 0.0945,
+      "step": 12850
+    },
+    {
+      "epoch": 0.41148049787220425,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.2390703940720783e-05,
+      "loss": 0.1411,
+      "step": 12860
+    },
+    {
+      "epoch": 0.4118004671551531,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.2383967665880769e-05,
+      "loss": 0.1358,
+      "step": 12870
+    },
+    {
+      "epoch": 0.4121204364381019,
+      "grad_norm": 18.125,
+      "learning_rate": 1.2377231391040757e-05,
+      "loss": 0.108,
+      "step": 12880
+    },
+    {
+      "epoch": 0.4124404057210508,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.2370495116200743e-05,
+      "loss": 0.141,
+      "step": 12890
+    },
+    {
+      "epoch": 0.4127603750039996,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.2363758841360729e-05,
+      "loss": 0.0725,
+      "step": 12900
+    },
+    {
+      "epoch": 0.41308034428694845,
+      "grad_norm": 16.125,
+      "learning_rate": 1.2357022566520715e-05,
+      "loss": 0.1443,
+      "step": 12910
+    },
+    {
+      "epoch": 0.4134003135698973,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.2350286291680702e-05,
+      "loss": 0.0807,
+      "step": 12920
+    },
+    {
+      "epoch": 0.4137202828528461,
+      "grad_norm": 15.0,
+      "learning_rate": 1.2343550016840688e-05,
+      "loss": 0.0965,
+      "step": 12930
+    },
+    {
+      "epoch": 0.414040252135795,
+      "grad_norm": 13.125,
+      "learning_rate": 1.2336813742000675e-05,
+      "loss": 0.1492,
+      "step": 12940
+    },
+    {
+      "epoch": 0.4143602214187438,
+      "grad_norm": 17.125,
+      "learning_rate": 1.2330077467160662e-05,
+      "loss": 0.217,
+      "step": 12950
+    },
+    {
+      "epoch": 0.41468019070169265,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.2323341192320648e-05,
+      "loss": 0.087,
+      "step": 12960
+    },
+    {
+      "epoch": 0.41500015998464146,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2316604917480634e-05,
+      "loss": 0.1075,
+      "step": 12970
+    },
+    {
+      "epoch": 0.4153201292675903,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.230986864264062e-05,
+      "loss": 0.0746,
+      "step": 12980
+    },
+    {
+      "epoch": 0.4156400985505391,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.2303132367800608e-05,
+      "loss": 0.0914,
+      "step": 12990
+    },
+    {
+      "epoch": 0.415960067833488,
+      "grad_norm": 22.5,
+      "learning_rate": 1.2296396092960594e-05,
+      "loss": 0.1077,
+      "step": 13000
+    },
+    {
+      "epoch": 0.4162800371164368,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.228965981812058e-05,
+      "loss": 0.1814,
+      "step": 13010
+    },
+    {
+      "epoch": 0.41660000639938566,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.2282923543280568e-05,
+      "loss": 0.135,
+      "step": 13020
+    },
+    {
+      "epoch": 0.4169199756823345,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.2276187268440554e-05,
+      "loss": 0.1697,
+      "step": 13030
+    },
+    {
+      "epoch": 0.4172399449652833,
+      "grad_norm": 26.5,
+      "learning_rate": 1.226945099360054e-05,
+      "loss": 0.0988,
+      "step": 13040
+    },
+    {
+      "epoch": 0.4175599142482322,
+      "grad_norm": 8.875,
+      "learning_rate": 1.2262714718760526e-05,
+      "loss": 0.1446,
+      "step": 13050
+    },
+    {
+      "epoch": 0.417879883531181,
+      "grad_norm": 17.5,
+      "learning_rate": 1.2255978443920514e-05,
+      "loss": 0.1545,
+      "step": 13060
+    },
+    {
+      "epoch": 0.41819985281412986,
+      "grad_norm": 5.1875,
+      "learning_rate": 1.22492421690805e-05,
+      "loss": 0.1427,
+      "step": 13070
+    },
+    {
+      "epoch": 0.41851982209707866,
+      "grad_norm": 9.25,
+      "learning_rate": 1.2242505894240486e-05,
+      "loss": 0.1168,
+      "step": 13080
+    },
+    {
+      "epoch": 0.4188397913800275,
+      "grad_norm": 6.15625,
+      "learning_rate": 1.2235769619400472e-05,
+      "loss": 0.0839,
+      "step": 13090
+    },
+    {
+      "epoch": 0.41915976066297633,
+      "grad_norm": 4.84375,
+      "learning_rate": 1.222903334456046e-05,
+      "loss": 0.1375,
+      "step": 13100
+    },
+    {
+      "epoch": 0.4194797299459252,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.2222297069720446e-05,
+      "loss": 0.1034,
+      "step": 13110
+    },
+    {
+      "epoch": 0.419799699228874,
+      "grad_norm": 14.625,
+      "learning_rate": 1.2215560794880432e-05,
+      "loss": 0.1141,
+      "step": 13120
+    },
+    {
+      "epoch": 0.42011966851182286,
+      "grad_norm": 6.875,
+      "learning_rate": 1.220882452004042e-05,
+      "loss": 0.1263,
+      "step": 13130
+    },
+    {
+      "epoch": 0.4204396377947717,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.2202088245200406e-05,
+      "loss": 0.1666,
+      "step": 13140
+    },
+    {
+      "epoch": 0.42075960707772053,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.2195351970360392e-05,
+      "loss": 0.1097,
+      "step": 13150
+    },
+    {
+      "epoch": 0.4210795763606694,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2188615695520378e-05,
+      "loss": 0.0806,
+      "step": 13160
+    },
+    {
+      "epoch": 0.4213995456436182,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.2181879420680365e-05,
+      "loss": 0.0913,
+      "step": 13170
+    },
+    {
+      "epoch": 0.42171951492656706,
+      "grad_norm": 24.875,
+      "learning_rate": 1.2175143145840351e-05,
+      "loss": 0.1109,
+      "step": 13180
+    },
+    {
+      "epoch": 0.42203948420951587,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2168406871000337e-05,
+      "loss": 0.0954,
+      "step": 13190
+    },
+    {
+      "epoch": 0.42235945349246473,
+      "grad_norm": 6.375,
+      "learning_rate": 1.2161670596160325e-05,
+      "loss": 0.097,
+      "step": 13200
+    },
+    {
+      "epoch": 0.42267942277541354,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.2154934321320311e-05,
+      "loss": 0.092,
+      "step": 13210
+    },
+    {
+      "epoch": 0.4229993920583624,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.2148198046480297e-05,
+      "loss": 0.1513,
+      "step": 13220
+    },
+    {
+      "epoch": 0.4233193613413112,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.2141461771640283e-05,
+      "loss": 0.0842,
+      "step": 13230
+    },
+    {
+      "epoch": 0.42363933062426007,
+      "grad_norm": 44.75,
+      "learning_rate": 1.2134725496800271e-05,
+      "loss": 0.1017,
+      "step": 13240
+    },
+    {
+      "epoch": 0.42395929990720893,
+      "grad_norm": 16.75,
+      "learning_rate": 1.2127989221960257e-05,
+      "loss": 0.0391,
+      "step": 13250
+    },
+    {
+      "epoch": 0.42427926919015774,
+      "grad_norm": 4.625,
+      "learning_rate": 1.2121252947120243e-05,
+      "loss": 0.0945,
+      "step": 13260
+    },
+    {
+      "epoch": 0.4245992384731066,
+      "grad_norm": 22.125,
+      "learning_rate": 1.2114516672280229e-05,
+      "loss": 0.0986,
+      "step": 13270
+    },
+    {
+      "epoch": 0.4249192077560554,
+      "grad_norm": 5.875,
+      "learning_rate": 1.2107780397440217e-05,
+      "loss": 0.1013,
+      "step": 13280
+    },
+    {
+      "epoch": 0.42523917703900427,
+      "grad_norm": 28.75,
+      "learning_rate": 1.2101044122600203e-05,
+      "loss": 0.1588,
+      "step": 13290
+    },
+    {
+      "epoch": 0.4255591463219531,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.2094307847760189e-05,
+      "loss": 0.1424,
+      "step": 13300
+    },
+    {
+      "epoch": 0.42587911560490194,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.2087571572920177e-05,
+      "loss": 0.1216,
+      "step": 13310
+    },
+    {
+      "epoch": 0.42619908488785074,
+      "grad_norm": 22.0,
+      "learning_rate": 1.2080835298080163e-05,
+      "loss": 0.1575,
+      "step": 13320
+    },
+    {
+      "epoch": 0.4265190541707996,
+      "grad_norm": 25.625,
+      "learning_rate": 1.2074099023240149e-05,
+      "loss": 0.0764,
+      "step": 13330
+    },
+    {
+      "epoch": 0.42683902345374847,
+      "grad_norm": 4.5,
+      "learning_rate": 1.2067362748400135e-05,
+      "loss": 0.1406,
+      "step": 13340
+    },
+    {
+      "epoch": 0.4271589927366973,
+      "grad_norm": 9.0,
+      "learning_rate": 1.2060626473560123e-05,
+      "loss": 0.103,
+      "step": 13350
+    },
+    {
+      "epoch": 0.42747896201964614,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.2053890198720109e-05,
+      "loss": 0.1186,
+      "step": 13360
+    },
+    {
+      "epoch": 0.42779893130259494,
+      "grad_norm": 13.25,
+      "learning_rate": 1.2047153923880095e-05,
+      "loss": 0.148,
+      "step": 13370
+    },
+    {
+      "epoch": 0.4281189005855438,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.204041764904008e-05,
+      "loss": 0.08,
+      "step": 13380
+    },
+    {
+      "epoch": 0.4284388698684926,
+      "grad_norm": 8.75,
+      "learning_rate": 1.2033681374200068e-05,
+      "loss": 0.0665,
+      "step": 13390
+    },
+    {
+      "epoch": 0.4287588391514415,
+      "grad_norm": 6.53125,
+      "learning_rate": 1.2026945099360054e-05,
+      "loss": 0.0673,
+      "step": 13400
+    },
+    {
+      "epoch": 0.4290788084343903,
+      "grad_norm": 24.875,
+      "learning_rate": 1.202020882452004e-05,
+      "loss": 0.0881,
+      "step": 13410
+    },
+    {
+      "epoch": 0.42939877771733914,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.2013472549680028e-05,
+      "loss": 0.087,
+      "step": 13420
+    },
+    {
+      "epoch": 0.42971874700028795,
+      "grad_norm": 17.0,
+      "learning_rate": 1.2006736274840014e-05,
+      "loss": 0.0621,
+      "step": 13430
+    },
+    {
+      "epoch": 0.4300387162832368,
+      "grad_norm": 16.125,
+      "learning_rate": 1.2e-05,
+      "loss": 0.0973,
+      "step": 13440
+    },
+    {
+      "epoch": 0.4303586855661857,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1993263725159986e-05,
+      "loss": 0.1739,
+      "step": 13450
+    },
+    {
+      "epoch": 0.4306786548491345,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.1986527450319974e-05,
+      "loss": 0.0507,
+      "step": 13460
+    },
+    {
+      "epoch": 0.43099862413208334,
+      "grad_norm": 6.75,
+      "learning_rate": 1.197979117547996e-05,
+      "loss": 0.1018,
+      "step": 13470
+    },
+    {
+      "epoch": 0.43131859341503215,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.1973054900639946e-05,
+      "loss": 0.1714,
+      "step": 13480
+    },
+    {
+      "epoch": 0.431638562697981,
+      "grad_norm": 52.5,
+      "learning_rate": 1.1966318625799934e-05,
+      "loss": 0.1345,
+      "step": 13490
+    },
+    {
+      "epoch": 0.4319585319809298,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.195958235095992e-05,
+      "loss": 0.1365,
+      "step": 13500
+    },
+    {
+      "epoch": 0.4322785012638787,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.1952846076119906e-05,
+      "loss": 0.0933,
+      "step": 13510
+    },
+    {
+      "epoch": 0.4325984705468275,
+      "grad_norm": 62.25,
+      "learning_rate": 1.1946109801279892e-05,
+      "loss": 0.1471,
+      "step": 13520
+    },
+    {
+      "epoch": 0.43291843982977635,
+      "grad_norm": 11.125,
+      "learning_rate": 1.193937352643988e-05,
+      "loss": 0.1052,
+      "step": 13530
+    },
+    {
+      "epoch": 0.43323840911272515,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.1932637251599866e-05,
+      "loss": 0.1031,
+      "step": 13540
+    },
+    {
+      "epoch": 0.433558378395674,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.1925900976759852e-05,
+      "loss": 0.1373,
+      "step": 13550
+    },
+    {
+      "epoch": 0.4338783476786229,
+      "grad_norm": 8.75,
+      "learning_rate": 1.1919164701919838e-05,
+      "loss": 0.071,
+      "step": 13560
+    },
+    {
+      "epoch": 0.4341983169615717,
+      "grad_norm": 4.84375,
+      "learning_rate": 1.1912428427079826e-05,
+      "loss": 0.1588,
+      "step": 13570
+    },
+    {
+      "epoch": 0.43451828624452055,
+      "grad_norm": 32.5,
+      "learning_rate": 1.1905692152239812e-05,
+      "loss": 0.1497,
+      "step": 13580
+    },
+    {
+      "epoch": 0.43483825552746935,
+      "grad_norm": 11.375,
+      "learning_rate": 1.1898955877399798e-05,
+      "loss": 0.0853,
+      "step": 13590
+    },
+    {
+      "epoch": 0.4351582248104182,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.1892219602559785e-05,
+      "loss": 0.0652,
+      "step": 13600
+    },
+    {
+      "epoch": 0.435478194093367,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.1885483327719771e-05,
+      "loss": 0.115,
+      "step": 13610
+    },
+    {
+      "epoch": 0.4357981633763159,
+      "grad_norm": 28.0,
+      "learning_rate": 1.1878747052879758e-05,
+      "loss": 0.0928,
+      "step": 13620
+    },
+    {
+      "epoch": 0.4361181326592647,
+      "grad_norm": 13.5,
+      "learning_rate": 1.1872010778039744e-05,
+      "loss": 0.1408,
+      "step": 13630
+    },
+    {
+      "epoch": 0.43643810194221355,
+      "grad_norm": 4.25,
+      "learning_rate": 1.1865274503199731e-05,
+      "loss": 0.1242,
+      "step": 13640
+    },
+    {
+      "epoch": 0.43675807122516236,
+      "grad_norm": 5.0625,
+      "learning_rate": 1.1858538228359717e-05,
+      "loss": 0.0633,
+      "step": 13650
+    },
+    {
+      "epoch": 0.4370780405081112,
+      "grad_norm": 20.375,
+      "learning_rate": 1.1851801953519703e-05,
+      "loss": 0.1442,
+      "step": 13660
+    },
+    {
+      "epoch": 0.4373980097910601,
+      "grad_norm": 13.0,
+      "learning_rate": 1.184506567867969e-05,
+      "loss": 0.1632,
+      "step": 13670
+    },
+    {
+      "epoch": 0.4377179790740089,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.1838329403839677e-05,
+      "loss": 0.1413,
+      "step": 13680
+    },
+    {
+      "epoch": 0.43803794835695775,
+      "grad_norm": 16.5,
+      "learning_rate": 1.1831593128999663e-05,
+      "loss": 0.0996,
+      "step": 13690
+    },
+    {
+      "epoch": 0.43835791763990656,
+      "grad_norm": 4.6875,
+      "learning_rate": 1.182485685415965e-05,
+      "loss": 0.1778,
+      "step": 13700
+    },
+    {
+      "epoch": 0.4386778869228554,
+      "grad_norm": 14.625,
+      "learning_rate": 1.1818120579319637e-05,
+      "loss": 0.103,
+      "step": 13710
+    },
+    {
+      "epoch": 0.4389978562058042,
+      "grad_norm": 16.875,
+      "learning_rate": 1.1811384304479623e-05,
+      "loss": 0.127,
+      "step": 13720
+    },
+    {
+      "epoch": 0.4393178254887531,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1804648029639609e-05,
+      "loss": 0.1439,
+      "step": 13730
+    },
+    {
+      "epoch": 0.4396377947717019,
+      "grad_norm": 11.5,
+      "learning_rate": 1.1797911754799595e-05,
+      "loss": 0.0937,
+      "step": 13740
+    },
+    {
+      "epoch": 0.43995776405465076,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.1791175479959583e-05,
+      "loss": 0.0906,
+      "step": 13750
+    },
+    {
+      "epoch": 0.44027773333759956,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.1784439205119569e-05,
+      "loss": 0.1649,
+      "step": 13760
+    },
+    {
+      "epoch": 0.4405977026205484,
+      "grad_norm": 4.53125,
+      "learning_rate": 1.1777702930279555e-05,
+      "loss": 0.1203,
+      "step": 13770
+    },
+    {
+      "epoch": 0.4409176719034973,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.1770966655439543e-05,
+      "loss": 0.0982,
+      "step": 13780
+    },
+    {
+      "epoch": 0.4412376411864461,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.1764230380599529e-05,
+      "loss": 0.1056,
+      "step": 13790
+    },
+    {
+      "epoch": 0.44155761046939496,
+      "grad_norm": 17.625,
+      "learning_rate": 1.1757494105759515e-05,
+      "loss": 0.1607,
+      "step": 13800
+    },
+    {
+      "epoch": 0.44187757975234376,
+      "grad_norm": 19.0,
+      "learning_rate": 1.17507578309195e-05,
+      "loss": 0.1308,
+      "step": 13810
+    },
+    {
+      "epoch": 0.4421975490352926,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.1744021556079489e-05,
+      "loss": 0.0923,
+      "step": 13820
+    },
+    {
+      "epoch": 0.44251751831824143,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.1737285281239476e-05,
+      "loss": 0.126,
+      "step": 13830
+    },
+    {
+      "epoch": 0.4428374876011903,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.1730549006399464e-05,
+      "loss": 0.0808,
+      "step": 13840
+    },
+    {
+      "epoch": 0.4431574568841391,
+      "grad_norm": 26.125,
+      "learning_rate": 1.172381273155945e-05,
+      "loss": 0.1405,
+      "step": 13850
+    },
+    {
+      "epoch": 0.44347742616708796,
+      "grad_norm": 14.625,
+      "learning_rate": 1.1717076456719436e-05,
+      "loss": 0.1164,
+      "step": 13860
+    },
+    {
+      "epoch": 0.44379739545003677,
+      "grad_norm": 12.25,
+      "learning_rate": 1.1710340181879422e-05,
+      "loss": 0.1179,
+      "step": 13870
+    },
+    {
+      "epoch": 0.44411736473298563,
+      "grad_norm": 6.09375,
+      "learning_rate": 1.170360390703941e-05,
+      "loss": 0.074,
+      "step": 13880
+    },
+    {
+      "epoch": 0.4444373340159345,
+      "grad_norm": 25.625,
+      "learning_rate": 1.1696867632199396e-05,
+      "loss": 0.097,
+      "step": 13890
+    },
+    {
+      "epoch": 0.4447573032988833,
+      "grad_norm": 6.0625,
+      "learning_rate": 1.1690131357359382e-05,
+      "loss": 0.1389,
+      "step": 13900
+    },
+    {
+      "epoch": 0.44507727258183216,
+      "grad_norm": 6.25,
+      "learning_rate": 1.1683395082519368e-05,
+      "loss": 0.1156,
+      "step": 13910
+    },
+    {
+      "epoch": 0.44539724186478097,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.1676658807679356e-05,
+      "loss": 0.0628,
+      "step": 13920
+    },
+    {
+      "epoch": 0.44571721114772983,
+      "grad_norm": 17.25,
+      "learning_rate": 1.1669922532839342e-05,
+      "loss": 0.2375,
+      "step": 13930
+    },
+    {
+      "epoch": 0.44603718043067864,
+      "grad_norm": 9.875,
+      "learning_rate": 1.1663186257999328e-05,
+      "loss": 0.086,
+      "step": 13940
+    },
+    {
+      "epoch": 0.4463571497136275,
+      "grad_norm": 17.875,
+      "learning_rate": 1.1656449983159316e-05,
+      "loss": 0.1142,
+      "step": 13950
+    },
+    {
+      "epoch": 0.4466771189965763,
+      "grad_norm": 16.875,
+      "learning_rate": 1.1649713708319302e-05,
+      "loss": 0.1143,
+      "step": 13960
+    },
+    {
+      "epoch": 0.44699708827952517,
+      "grad_norm": 15.625,
+      "learning_rate": 1.1642977433479288e-05,
+      "loss": 0.1356,
+      "step": 13970
+    },
+    {
+      "epoch": 0.44731705756247403,
+      "grad_norm": 13.0,
+      "learning_rate": 1.1636241158639274e-05,
+      "loss": 0.1781,
+      "step": 13980
+    },
+    {
+      "epoch": 0.44763702684542284,
+      "grad_norm": 7.625,
+      "learning_rate": 1.1629504883799261e-05,
+      "loss": 0.1361,
+      "step": 13990
+    },
+    {
+      "epoch": 0.4479569961283717,
+      "grad_norm": 11.75,
+      "learning_rate": 1.1622768608959247e-05,
+      "loss": 0.068,
+      "step": 14000
+    },
+    {
+      "epoch": 0.4482769654113205,
+      "grad_norm": 5.375,
+      "learning_rate": 1.1616032334119233e-05,
+      "loss": 0.1016,
+      "step": 14010
+    },
+    {
+      "epoch": 0.44859693469426937,
+      "grad_norm": 9.5,
+      "learning_rate": 1.160929605927922e-05,
+      "loss": 0.1316,
+      "step": 14020
+    },
+    {
+      "epoch": 0.4489169039772182,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.1602559784439207e-05,
+      "loss": 0.0897,
+      "step": 14030
+    },
+    {
+      "epoch": 0.44923687326016704,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.1595823509599193e-05,
+      "loss": 0.0831,
+      "step": 14040
+    },
+    {
+      "epoch": 0.44955684254311584,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.158908723475918e-05,
+      "loss": 0.1336,
+      "step": 14050
+    },
+    {
+      "epoch": 0.4498768118260647,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.1582350959919167e-05,
+      "loss": 0.0792,
+      "step": 14060
+    },
+    {
+      "epoch": 0.4501967811090135,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.1575614685079153e-05,
+      "loss": 0.1652,
+      "step": 14070
+    },
+    {
+      "epoch": 0.4505167503919624,
+      "grad_norm": 10.0,
+      "learning_rate": 1.1568878410239139e-05,
+      "loss": 0.0984,
+      "step": 14080
+    },
+    {
+      "epoch": 0.45083671967491123,
+      "grad_norm": 9.875,
+      "learning_rate": 1.1562142135399125e-05,
+      "loss": 0.1038,
+      "step": 14090
+    },
+    {
+      "epoch": 0.45115668895786004,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.1555405860559113e-05,
+      "loss": 0.0797,
+      "step": 14100
+    },
+    {
+      "epoch": 0.4514766582408089,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.1548669585719099e-05,
+      "loss": 0.1299,
+      "step": 14110
+    },
+    {
+      "epoch": 0.4517966275237577,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.1541933310879085e-05,
+      "loss": 0.1071,
+      "step": 14120
+    },
+    {
+      "epoch": 0.45211659680670657,
+      "grad_norm": 18.25,
+      "learning_rate": 1.1535197036039073e-05,
+      "loss": 0.1636,
+      "step": 14130
+    },
+    {
+      "epoch": 0.4524365660896554,
+      "grad_norm": 6.125,
+      "learning_rate": 1.1528460761199059e-05,
+      "loss": 0.1152,
+      "step": 14140
+    },
+    {
+      "epoch": 0.45275653537260424,
+      "grad_norm": 3.5,
+      "learning_rate": 1.1521724486359045e-05,
+      "loss": 0.1073,
+      "step": 14150
+    },
+    {
+      "epoch": 0.45307650465555305,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.1514988211519031e-05,
+      "loss": 0.1152,
+      "step": 14160
+    },
+    {
+      "epoch": 0.4533964739385019,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.1508251936679019e-05,
+      "loss": 0.1045,
+      "step": 14170
+    },
+    {
+      "epoch": 0.4537164432214507,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.1501515661839005e-05,
+      "loss": 0.1372,
+      "step": 14180
+    },
+    {
+      "epoch": 0.4540364125043996,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.149477938699899e-05,
+      "loss": 0.133,
+      "step": 14190
+    },
+    {
+      "epoch": 0.45435638178734844,
+      "grad_norm": 5.78125,
+      "learning_rate": 1.1488043112158977e-05,
+      "loss": 0.1355,
+      "step": 14200
+    },
+    {
+      "epoch": 0.45467635107029725,
+      "grad_norm": 5.125,
+      "learning_rate": 1.1481306837318964e-05,
+      "loss": 0.1443,
+      "step": 14210
+    },
+    {
+      "epoch": 0.4549963203532461,
+      "grad_norm": 20.875,
+      "learning_rate": 1.147457056247895e-05,
+      "loss": 0.1197,
+      "step": 14220
+    },
+    {
+      "epoch": 0.4553162896361949,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.1467834287638937e-05,
+      "loss": 0.0711,
+      "step": 14230
+    },
+    {
+      "epoch": 0.4556362589191438,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.1461098012798924e-05,
+      "loss": 0.0803,
+      "step": 14240
+    },
+    {
+      "epoch": 0.4559562282020926,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.145436173795891e-05,
+      "loss": 0.1262,
+      "step": 14250
+    },
+    {
+      "epoch": 0.45627619748504145,
+      "grad_norm": 21.5,
+      "learning_rate": 1.1447625463118896e-05,
+      "loss": 0.0693,
+      "step": 14260
+    },
+    {
+      "epoch": 0.45659616676799025,
+      "grad_norm": 15.125,
+      "learning_rate": 1.1440889188278882e-05,
+      "loss": 0.1039,
+      "step": 14270
+    },
+    {
+      "epoch": 0.4569161360509391,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.143415291343887e-05,
+      "loss": 0.1364,
+      "step": 14280
+    },
+    {
+      "epoch": 0.4572361053338879,
+      "grad_norm": 15.875,
+      "learning_rate": 1.1427416638598856e-05,
+      "loss": 0.0857,
+      "step": 14290
+    },
+    {
+      "epoch": 0.4575560746168368,
+      "grad_norm": 6.3125,
+      "learning_rate": 1.1420680363758842e-05,
+      "loss": 0.11,
+      "step": 14300
+    },
+    {
+      "epoch": 0.45787604389978565,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.141394408891883e-05,
+      "loss": 0.1195,
+      "step": 14310
+    },
+    {
+      "epoch": 0.45819601318273445,
+      "grad_norm": 18.25,
+      "learning_rate": 1.1407207814078816e-05,
+      "loss": 0.1222,
+      "step": 14320
+    },
+    {
+      "epoch": 0.4585159824656833,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.1400471539238802e-05,
+      "loss": 0.135,
+      "step": 14330
+    },
+    {
+      "epoch": 0.4588359517486321,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.1393735264398788e-05,
+      "loss": 0.1464,
+      "step": 14340
+    },
+    {
+      "epoch": 0.459155921031581,
+      "grad_norm": 5.1875,
+      "learning_rate": 1.1386998989558776e-05,
+      "loss": 0.0776,
+      "step": 14350
+    },
+    {
+      "epoch": 0.4594758903145298,
+      "grad_norm": 7.375,
+      "learning_rate": 1.1380262714718762e-05,
+      "loss": 0.1042,
+      "step": 14360
+    },
+    {
+      "epoch": 0.45979585959747865,
+      "grad_norm": 11.0,
+      "learning_rate": 1.1373526439878748e-05,
+      "loss": 0.1421,
+      "step": 14370
+    },
+    {
+      "epoch": 0.46011582888042746,
+      "grad_norm": 5.46875,
+      "learning_rate": 1.1366790165038734e-05,
+      "loss": 0.1339,
+      "step": 14380
+    },
+    {
+      "epoch": 0.4604357981633763,
+      "grad_norm": 15.75,
+      "learning_rate": 1.1360053890198722e-05,
+      "loss": 0.0889,
+      "step": 14390
+    },
+    {
+      "epoch": 0.4607557674463251,
+      "grad_norm": 11.125,
+      "learning_rate": 1.1353317615358708e-05,
+      "loss": 0.0998,
+      "step": 14400
+    },
+    {
+      "epoch": 0.461075736729274,
+      "grad_norm": 9.125,
+      "learning_rate": 1.1346581340518694e-05,
+      "loss": 0.1288,
+      "step": 14410
+    },
+    {
+      "epoch": 0.46139570601222285,
+      "grad_norm": 0.375,
+      "learning_rate": 1.1339845065678681e-05,
+      "loss": 0.0567,
+      "step": 14420
+    },
+    {
+      "epoch": 0.46171567529517166,
+      "grad_norm": 14.125,
+      "learning_rate": 1.1333108790838668e-05,
+      "loss": 0.1746,
+      "step": 14430
+    },
+    {
+      "epoch": 0.4620356445781205,
+      "grad_norm": 35.25,
+      "learning_rate": 1.1326372515998654e-05,
+      "loss": 0.1098,
+      "step": 14440
+    },
+    {
+      "epoch": 0.4623556138610693,
+      "grad_norm": 25.75,
+      "learning_rate": 1.131963624115864e-05,
+      "loss": 0.1504,
+      "step": 14450
+    },
+    {
+      "epoch": 0.4626755831440182,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.1312899966318627e-05,
+      "loss": 0.0965,
+      "step": 14460
+    },
+    {
+      "epoch": 0.462995552426967,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.1306163691478613e-05,
+      "loss": 0.1065,
+      "step": 14470
+    },
+    {
+      "epoch": 0.46331552170991586,
+      "grad_norm": 7.625,
+      "learning_rate": 1.12994274166386e-05,
+      "loss": 0.0949,
+      "step": 14480
+    },
+    {
+      "epoch": 0.46363549099286466,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.1292691141798585e-05,
+      "loss": 0.086,
+      "step": 14490
+    },
+    {
+      "epoch": 0.4639554602758135,
+      "grad_norm": 11.4375,
+      "learning_rate": 1.1285954866958573e-05,
+      "loss": 0.1052,
+      "step": 14500
+    },
+    {
+      "epoch": 0.46427542955876233,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.127921859211856e-05,
+      "loss": 0.1108,
+      "step": 14510
+    },
+    {
+      "epoch": 0.4645953988417112,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.1272482317278545e-05,
+      "loss": 0.1441,
+      "step": 14520
+    },
+    {
+      "epoch": 0.46491536812466006,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.1265746042438533e-05,
+      "loss": 0.1438,
+      "step": 14530
+    },
+    {
+      "epoch": 0.46523533740760886,
+      "grad_norm": 80.0,
+      "learning_rate": 1.1259009767598519e-05,
+      "loss": 0.1715,
+      "step": 14540
+    },
+    {
+      "epoch": 0.4655553066905577,
+      "grad_norm": 20.0,
+      "learning_rate": 1.1252273492758505e-05,
+      "loss": 0.1282,
+      "step": 14550
+    },
+    {
+      "epoch": 0.46587527597350653,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1245537217918491e-05,
+      "loss": 0.1233,
+      "step": 14560
+    },
+    {
+      "epoch": 0.4661952452564554,
+      "grad_norm": 10.5625,
+      "learning_rate": 1.1238800943078479e-05,
+      "loss": 0.1397,
+      "step": 14570
+    },
+    {
+      "epoch": 0.4665152145394042,
+      "grad_norm": 22.25,
+      "learning_rate": 1.1232064668238465e-05,
+      "loss": 0.0896,
+      "step": 14580
+    },
+    {
+      "epoch": 0.46683518382235306,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.1225328393398451e-05,
+      "loss": 0.1315,
+      "step": 14590
+    },
+    {
+      "epoch": 0.46715515310530187,
+      "grad_norm": 9.125,
+      "learning_rate": 1.1218592118558439e-05,
+      "loss": 0.1294,
+      "step": 14600
+    },
+    {
+      "epoch": 0.46747512238825073,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.1211855843718425e-05,
+      "loss": 0.1686,
+      "step": 14610
+    },
+    {
+      "epoch": 0.4677950916711996,
+      "grad_norm": 3.0,
+      "learning_rate": 1.120511956887841e-05,
+      "loss": 0.0702,
+      "step": 14620
+    },
+    {
+      "epoch": 0.4681150609541484,
+      "grad_norm": 9.375,
+      "learning_rate": 1.1198383294038397e-05,
+      "loss": 0.145,
+      "step": 14630
+    },
+    {
+      "epoch": 0.46843503023709726,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.1191647019198385e-05,
+      "loss": 0.097,
+      "step": 14640
+    },
+    {
+      "epoch": 0.46875499952004607,
+      "grad_norm": 5.8125,
+      "learning_rate": 1.118491074435837e-05,
+      "loss": 0.0526,
+      "step": 14650
+    },
+    {
+      "epoch": 0.46907496880299493,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.1178174469518357e-05,
+      "loss": 0.1244,
+      "step": 14660
+    },
+    {
+      "epoch": 0.46939493808594374,
+      "grad_norm": 12.875,
+      "learning_rate": 1.1171438194678343e-05,
+      "loss": 0.0983,
+      "step": 14670
+    },
+    {
+      "epoch": 0.4697149073688926,
+      "grad_norm": 9.25,
+      "learning_rate": 1.116470191983833e-05,
+      "loss": 0.0571,
+      "step": 14680
+    },
+    {
+      "epoch": 0.4700348766518414,
+      "grad_norm": 16.25,
+      "learning_rate": 1.1157965644998316e-05,
+      "loss": 0.0954,
+      "step": 14690
+    },
+    {
+      "epoch": 0.47035484593479027,
+      "grad_norm": 26.625,
+      "learning_rate": 1.1151229370158302e-05,
+      "loss": 0.0811,
+      "step": 14700
+    },
+    {
+      "epoch": 0.4706748152177391,
+      "grad_norm": 19.25,
+      "learning_rate": 1.114449309531829e-05,
+      "loss": 0.1657,
+      "step": 14710
+    },
+    {
+      "epoch": 0.47099478450068794,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.1137756820478276e-05,
+      "loss": 0.0761,
+      "step": 14720
+    },
+    {
+      "epoch": 0.4713147537836368,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.1131020545638262e-05,
+      "loss": 0.084,
+      "step": 14730
+    },
+    {
+      "epoch": 0.4716347230665856,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.1124284270798248e-05,
+      "loss": 0.1656,
+      "step": 14740
+    },
+    {
+      "epoch": 0.47195469234953447,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1117547995958236e-05,
+      "loss": 0.0599,
+      "step": 14750
+    },
+    {
+      "epoch": 0.4722746616324833,
+      "grad_norm": 23.375,
+      "learning_rate": 1.1110811721118222e-05,
+      "loss": 0.116,
+      "step": 14760
+    },
+    {
+      "epoch": 0.47259463091543213,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.1104075446278208e-05,
+      "loss": 0.1249,
+      "step": 14770
+    },
+    {
+      "epoch": 0.47291460019838094,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.1097339171438194e-05,
+      "loss": 0.0941,
+      "step": 14780
+    },
+    {
+      "epoch": 0.4732345694813298,
+      "grad_norm": 9.375,
+      "learning_rate": 1.1090602896598182e-05,
+      "loss": 0.0901,
+      "step": 14790
+    },
+    {
+      "epoch": 0.4735545387642786,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.1083866621758168e-05,
+      "loss": 0.0754,
+      "step": 14800
+    },
+    {
+      "epoch": 0.47387450804722747,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.1077130346918154e-05,
+      "loss": 0.0792,
+      "step": 14810
+    },
+    {
+      "epoch": 0.4741944773301763,
+      "grad_norm": 18.875,
+      "learning_rate": 1.1070394072078142e-05,
+      "loss": 0.1138,
+      "step": 14820
+    },
+    {
+      "epoch": 0.47451444661312514,
+      "grad_norm": 10.75,
+      "learning_rate": 1.1063657797238128e-05,
+      "loss": 0.1074,
+      "step": 14830
+    },
+    {
+      "epoch": 0.474834415896074,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.1056921522398114e-05,
+      "loss": 0.1801,
+      "step": 14840
+    },
+    {
+      "epoch": 0.4751543851790228,
+      "grad_norm": 9.0,
+      "learning_rate": 1.10501852475581e-05,
+      "loss": 0.0982,
+      "step": 14850
+    },
+    {
+      "epoch": 0.47547435446197167,
+      "grad_norm": 6.09375,
+      "learning_rate": 1.1043448972718088e-05,
+      "loss": 0.0937,
+      "step": 14860
+    },
+    {
+      "epoch": 0.4757943237449205,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.1036712697878074e-05,
+      "loss": 0.162,
+      "step": 14870
+    },
+    {
+      "epoch": 0.47611429302786934,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.102997642303806e-05,
+      "loss": 0.1362,
+      "step": 14880
+    },
+    {
+      "epoch": 0.47643426231081815,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.1023240148198047e-05,
+      "loss": 0.1303,
+      "step": 14890
+    },
+    {
+      "epoch": 0.476754231593767,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.1016503873358033e-05,
+      "loss": 0.0811,
+      "step": 14900
+    },
+    {
+      "epoch": 0.4770742008767158,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.100976759851802e-05,
+      "loss": 0.1527,
+      "step": 14910
+    },
+    {
+      "epoch": 0.4773941701596647,
+      "grad_norm": 11.1875,
+      "learning_rate": 1.1003031323678006e-05,
+      "loss": 0.1181,
+      "step": 14920
+    },
+    {
+      "epoch": 0.4777141394426135,
+      "grad_norm": 4.59375,
+      "learning_rate": 1.0996295048837993e-05,
+      "loss": 0.1016,
+      "step": 14930
+    },
+    {
+      "epoch": 0.47803410872556235,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.098955877399798e-05,
+      "loss": 0.1299,
+      "step": 14940
+    },
+    {
+      "epoch": 0.4783540780085112,
+      "grad_norm": 18.375,
+      "learning_rate": 1.0982822499157965e-05,
+      "loss": 0.0982,
+      "step": 14950
+    },
+    {
+      "epoch": 0.47867404729146,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.0976086224317951e-05,
+      "loss": 0.1061,
+      "step": 14960
+    },
+    {
+      "epoch": 0.4789940165744089,
+      "grad_norm": 11.875,
+      "learning_rate": 1.096934994947794e-05,
+      "loss": 0.0838,
+      "step": 14970
+    },
+    {
+      "epoch": 0.4793139858573577,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.0962613674637925e-05,
+      "loss": 0.1507,
+      "step": 14980
+    },
+    {
+      "epoch": 0.47963395514030654,
+      "grad_norm": 6.1875,
+      "learning_rate": 1.0955877399797911e-05,
+      "loss": 0.2222,
+      "step": 14990
+    },
+    {
+      "epoch": 0.47995392442325535,
+      "grad_norm": 7.125,
+      "learning_rate": 1.0949141124957899e-05,
+      "loss": 0.1062,
+      "step": 15000
+    },
+    {
+      "epoch": 0.4802738937062042,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0942404850117885e-05,
+      "loss": 0.1796,
+      "step": 15010
+    },
+    {
+      "epoch": 0.480593862989153,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.0935668575277871e-05,
+      "loss": 0.1774,
+      "step": 15020
+    },
+    {
+      "epoch": 0.4809138322721019,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.0928932300437857e-05,
+      "loss": 0.1015,
+      "step": 15030
+    },
+    {
+      "epoch": 0.4812338015550507,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.0922196025597845e-05,
+      "loss": 0.1353,
+      "step": 15040
+    },
+    {
+      "epoch": 0.48155377083799955,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.0915459750757831e-05,
+      "loss": 0.0966,
+      "step": 15050
+    },
+    {
+      "epoch": 0.4818737401209484,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.0908723475917817e-05,
+      "loss": 0.0704,
+      "step": 15060
+    },
+    {
+      "epoch": 0.4821937094038972,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.0901987201077805e-05,
+      "loss": 0.0683,
+      "step": 15070
+    },
+    {
+      "epoch": 0.4825136786868461,
+      "grad_norm": 8.5,
+      "learning_rate": 1.089525092623779e-05,
+      "loss": 0.1077,
+      "step": 15080
+    },
+    {
+      "epoch": 0.4828336479697949,
+      "grad_norm": 17.25,
+      "learning_rate": 1.0888514651397777e-05,
+      "loss": 0.1185,
+      "step": 15090
+    },
+    {
+      "epoch": 0.48315361725274375,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.0881778376557763e-05,
+      "loss": 0.1238,
+      "step": 15100
+    },
+    {
+      "epoch": 0.48347358653569256,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.087504210171775e-05,
+      "loss": 0.0841,
+      "step": 15110
+    },
+    {
+      "epoch": 0.4837935558186414,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.0868305826877738e-05,
+      "loss": 0.0856,
+      "step": 15120
+    },
+    {
+      "epoch": 0.4841135251015902,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.0861569552037724e-05,
+      "loss": 0.0966,
+      "step": 15130
+    },
+    {
+      "epoch": 0.4844334943845391,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0854833277197712e-05,
+      "loss": 0.1029,
+      "step": 15140
+    },
+    {
+      "epoch": 0.4847534636674879,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.0848097002357698e-05,
+      "loss": 0.1597,
+      "step": 15150
+    },
+    {
+      "epoch": 0.48507343295043676,
+      "grad_norm": 11.0,
+      "learning_rate": 1.0841360727517684e-05,
+      "loss": 0.1657,
+      "step": 15160
+    },
+    {
+      "epoch": 0.4853934022333856,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.0834624452677672e-05,
+      "loss": 0.1039,
+      "step": 15170
+    },
+    {
+      "epoch": 0.4857133715163344,
+      "grad_norm": 6.75,
+      "learning_rate": 1.0827888177837658e-05,
+      "loss": 0.0897,
+      "step": 15180
+    },
+    {
+      "epoch": 0.4860333407992833,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.0821151902997644e-05,
+      "loss": 0.16,
+      "step": 15190
+    },
+    {
+      "epoch": 0.4863533100822321,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.081441562815763e-05,
+      "loss": 0.1087,
+      "step": 15200
+    },
+    {
+      "epoch": 0.48667327936518096,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.0807679353317618e-05,
+      "loss": 0.0997,
+      "step": 15210
+    },
+    {
+      "epoch": 0.48699324864812976,
+      "grad_norm": 10.25,
+      "learning_rate": 1.0800943078477604e-05,
+      "loss": 0.1594,
+      "step": 15220
+    },
+    {
+      "epoch": 0.4873132179310786,
+      "grad_norm": 14.25,
+      "learning_rate": 1.079420680363759e-05,
+      "loss": 0.1314,
+      "step": 15230
+    },
+    {
+      "epoch": 0.48763318721402743,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.0787470528797578e-05,
+      "loss": 0.0725,
+      "step": 15240
+    },
+    {
+      "epoch": 0.4879531564969763,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0780734253957564e-05,
+      "loss": 0.1273,
+      "step": 15250
+    },
+    {
+      "epoch": 0.48827312577992515,
+      "grad_norm": 18.125,
+      "learning_rate": 1.077399797911755e-05,
+      "loss": 0.1245,
+      "step": 15260
+    },
+    {
+      "epoch": 0.48859309506287396,
+      "grad_norm": 11.1875,
+      "learning_rate": 1.0767261704277536e-05,
+      "loss": 0.1129,
+      "step": 15270
+    },
+    {
+      "epoch": 0.4889130643458228,
+      "grad_norm": 24.25,
+      "learning_rate": 1.0760525429437523e-05,
+      "loss": 0.106,
+      "step": 15280
+    },
+    {
+      "epoch": 0.48923303362877163,
+      "grad_norm": 21.75,
+      "learning_rate": 1.075378915459751e-05,
+      "loss": 0.1526,
+      "step": 15290
+    },
+    {
+      "epoch": 0.4895530029117205,
+      "grad_norm": 38.75,
+      "learning_rate": 1.0747052879757495e-05,
+      "loss": 0.137,
+      "step": 15300
+    },
+    {
+      "epoch": 0.4898729721946693,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.0740316604917482e-05,
+      "loss": 0.1571,
+      "step": 15310
+    },
+    {
+      "epoch": 0.49019294147761816,
+      "grad_norm": 6.0,
+      "learning_rate": 1.073358033007747e-05,
+      "loss": 0.1228,
+      "step": 15320
+    },
+    {
+      "epoch": 0.49051291076056697,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.0726844055237455e-05,
+      "loss": 0.0661,
+      "step": 15330
+    },
+    {
+      "epoch": 0.49083288004351583,
+      "grad_norm": 6.3125,
+      "learning_rate": 1.0720107780397441e-05,
+      "loss": 0.1023,
+      "step": 15340
+    },
+    {
+      "epoch": 0.49115284932646464,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.0713371505557429e-05,
+      "loss": 0.1107,
+      "step": 15350
+    },
+    {
+      "epoch": 0.4914728186094135,
+      "grad_norm": 5.4375,
+      "learning_rate": 1.0706635230717415e-05,
+      "loss": 0.0765,
+      "step": 15360
+    },
+    {
+      "epoch": 0.49179278789236236,
+      "grad_norm": 4.5,
+      "learning_rate": 1.0699898955877401e-05,
+      "loss": 0.1254,
+      "step": 15370
+    },
+    {
+      "epoch": 0.49211275717531117,
+      "grad_norm": 1.125,
+      "learning_rate": 1.0693162681037387e-05,
+      "loss": 0.112,
+      "step": 15380
+    },
+    {
+      "epoch": 0.49243272645826003,
+      "grad_norm": 24.0,
+      "learning_rate": 1.0686426406197375e-05,
+      "loss": 0.0961,
+      "step": 15390
+    },
+    {
+      "epoch": 0.49275269574120883,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0679690131357361e-05,
+      "loss": 0.146,
+      "step": 15400
+    },
+    {
+      "epoch": 0.4930726650241577,
+      "grad_norm": 13.875,
+      "learning_rate": 1.0672953856517347e-05,
+      "loss": 0.0843,
+      "step": 15410
+    },
+    {
+      "epoch": 0.4933926343071065,
+      "grad_norm": 19.75,
+      "learning_rate": 1.0666217581677335e-05,
+      "loss": 0.1355,
+      "step": 15420
+    },
+    {
+      "epoch": 0.49371260359005537,
+      "grad_norm": 16.75,
+      "learning_rate": 1.065948130683732e-05,
+      "loss": 0.1467,
+      "step": 15430
+    },
+    {
+      "epoch": 0.49403257287300417,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.0652745031997307e-05,
+      "loss": 0.0314,
+      "step": 15440
+    },
+    {
+      "epoch": 0.49435254215595303,
+      "grad_norm": 7.125,
+      "learning_rate": 1.0646008757157293e-05,
+      "loss": 0.1102,
+      "step": 15450
+    },
+    {
+      "epoch": 0.49467251143890184,
+      "grad_norm": 18.875,
+      "learning_rate": 1.063927248231728e-05,
+      "loss": 0.1862,
+      "step": 15460
+    },
+    {
+      "epoch": 0.4949924807218507,
+      "grad_norm": 30.125,
+      "learning_rate": 1.0632536207477267e-05,
+      "loss": 0.0807,
+      "step": 15470
+    },
+    {
+      "epoch": 0.49531245000479956,
+      "grad_norm": 4.46875,
+      "learning_rate": 1.0625799932637253e-05,
+      "loss": 0.1243,
+      "step": 15480
+    },
+    {
+      "epoch": 0.49563241928774837,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.0619063657797239e-05,
+      "loss": 0.1785,
+      "step": 15490
+    },
+    {
+      "epoch": 0.49595238857069723,
+      "grad_norm": 29.0,
+      "learning_rate": 1.0612327382957226e-05,
+      "loss": 0.131,
+      "step": 15500
+    },
+    {
+      "epoch": 0.49627235785364604,
+      "grad_norm": 32.0,
+      "learning_rate": 1.0605591108117213e-05,
+      "loss": 0.1062,
+      "step": 15510
+    },
+    {
+      "epoch": 0.4965923271365949,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0598854833277199e-05,
+      "loss": 0.0994,
+      "step": 15520
+    },
+    {
+      "epoch": 0.4969122964195437,
+      "grad_norm": 4.1875,
+      "learning_rate": 1.0592118558437186e-05,
+      "loss": 0.0746,
+      "step": 15530
+    },
+    {
+      "epoch": 0.49723226570249257,
+      "grad_norm": 6.5,
+      "learning_rate": 1.0585382283597172e-05,
+      "loss": 0.1101,
+      "step": 15540
+    },
+    {
+      "epoch": 0.4975522349854414,
+      "grad_norm": 20.125,
+      "learning_rate": 1.0578646008757158e-05,
+      "loss": 0.1653,
+      "step": 15550
+    },
+    {
+      "epoch": 0.49787220426839024,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.0571909733917144e-05,
+      "loss": 0.0993,
+      "step": 15560
+    },
+    {
+      "epoch": 0.49819217355133905,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.0565173459077132e-05,
+      "loss": 0.1378,
+      "step": 15570
+    },
+    {
+      "epoch": 0.4985121428342879,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.0558437184237118e-05,
+      "loss": 0.0934,
+      "step": 15580
+    },
+    {
+      "epoch": 0.49883211211723677,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.0551700909397104e-05,
+      "loss": 0.1055,
+      "step": 15590
+    },
+    {
+      "epoch": 0.4991520814001856,
+      "grad_norm": 16.5,
+      "learning_rate": 1.054496463455709e-05,
+      "loss": 0.138,
+      "step": 15600
+    },
+    {
+      "epoch": 0.49947205068313444,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.0538228359717078e-05,
+      "loss": 0.1034,
+      "step": 15610
+    },
+    {
+      "epoch": 0.49979201996608325,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.0531492084877064e-05,
+      "loss": 0.1236,
+      "step": 15620
+    },
+    {
+      "epoch": 0.5001119892490321,
+      "grad_norm": 20.0,
+      "learning_rate": 1.052475581003705e-05,
+      "loss": 0.0922,
+      "step": 15630
+    },
+    {
+      "epoch": 0.500431958531981,
+      "grad_norm": 26.25,
+      "learning_rate": 1.0518019535197038e-05,
+      "loss": 0.1105,
+      "step": 15640
+    },
+    {
+      "epoch": 0.5007519278149297,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.0511283260357024e-05,
+      "loss": 0.148,
+      "step": 15650
+    },
+    {
+      "epoch": 0.5010718970978786,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.050454698551701e-05,
+      "loss": 0.1202,
+      "step": 15660
+    },
+    {
+      "epoch": 0.5013918663808274,
+      "grad_norm": 15.25,
+      "learning_rate": 1.0497810710676996e-05,
+      "loss": 0.0687,
+      "step": 15670
+    },
+    {
+      "epoch": 0.5017118356637763,
+      "grad_norm": 4.0,
+      "learning_rate": 1.0491074435836984e-05,
+      "loss": 0.1277,
+      "step": 15680
+    },
+    {
+      "epoch": 0.5020318049467252,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.048433816099697e-05,
+      "loss": 0.1179,
+      "step": 15690
+    },
+    {
+      "epoch": 0.5023517742296739,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0477601886156956e-05,
+      "loss": 0.1262,
+      "step": 15700
+    },
+    {
+      "epoch": 0.5026717435126228,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.0470865611316943e-05,
+      "loss": 0.1401,
+      "step": 15710
+    },
+    {
+      "epoch": 0.5029917127955716,
+      "grad_norm": 4.6875,
+      "learning_rate": 1.046412933647693e-05,
+      "loss": 0.1198,
+      "step": 15720
+    },
+    {
+      "epoch": 0.5033116820785205,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.0457393061636916e-05,
+      "loss": 0.1152,
+      "step": 15730
+    },
+    {
+      "epoch": 0.5036316513614693,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.0450656786796902e-05,
+      "loss": 0.0833,
+      "step": 15740
+    },
+    {
+      "epoch": 0.5039516206444181,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.044392051195689e-05,
+      "loss": 0.1354,
+      "step": 15750
+    },
+    {
+      "epoch": 0.504271589927367,
+      "grad_norm": 4.65625,
+      "learning_rate": 1.0437184237116875e-05,
+      "loss": 0.0848,
+      "step": 15760
+    },
+    {
+      "epoch": 0.5045915592103158,
+      "grad_norm": 20.25,
+      "learning_rate": 1.0430447962276861e-05,
+      "loss": 0.1016,
+      "step": 15770
+    },
+    {
+      "epoch": 0.5049115284932646,
+      "grad_norm": 14.25,
+      "learning_rate": 1.0423711687436847e-05,
+      "loss": 0.1103,
+      "step": 15780
+    },
+    {
+      "epoch": 0.5052314977762135,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.0416975412596835e-05,
+      "loss": 0.0355,
+      "step": 15790
+    },
+    {
+      "epoch": 0.5055514670591623,
+      "grad_norm": 30.75,
+      "learning_rate": 1.0410239137756821e-05,
+      "loss": 0.1866,
+      "step": 15800
+    },
+    {
+      "epoch": 0.5058714363421112,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0403502862916807e-05,
+      "loss": 0.0898,
+      "step": 15810
+    },
+    {
+      "epoch": 0.50619140562506,
+      "grad_norm": 15.875,
+      "learning_rate": 1.0396766588076795e-05,
+      "loss": 0.1066,
+      "step": 15820
+    },
+    {
+      "epoch": 0.5065113749080088,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.0390030313236781e-05,
+      "loss": 0.1045,
+      "step": 15830
+    },
+    {
+      "epoch": 0.5068313441909577,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.0383294038396767e-05,
+      "loss": 0.1217,
+      "step": 15840
+    },
+    {
+      "epoch": 0.5071513134739065,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.0376557763556753e-05,
+      "loss": 0.0884,
+      "step": 15850
+    },
+    {
+      "epoch": 0.5074712827568554,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.0369821488716741e-05,
+      "loss": 0.0915,
+      "step": 15860
+    },
+    {
+      "epoch": 0.5077912520398041,
+      "grad_norm": 6.15625,
+      "learning_rate": 1.0363085213876727e-05,
+      "loss": 0.1297,
+      "step": 15870
+    },
+    {
+      "epoch": 0.508111221322753,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.0356348939036713e-05,
+      "loss": 0.0718,
+      "step": 15880
+    },
+    {
+      "epoch": 0.5084311906057019,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.0349612664196699e-05,
+      "loss": 0.0644,
+      "step": 15890
+    },
+    {
+      "epoch": 0.5087511598886507,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.0342876389356687e-05,
+      "loss": 0.108,
+      "step": 15900
+    },
+    {
+      "epoch": 0.5090711291715996,
+      "grad_norm": 12.75,
+      "learning_rate": 1.0336140114516673e-05,
+      "loss": 0.1508,
+      "step": 15910
+    },
+    {
+      "epoch": 0.5093910984545483,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.0329403839676659e-05,
+      "loss": 0.1289,
+      "step": 15920
+    },
+    {
+      "epoch": 0.5097110677374972,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.0322667564836647e-05,
+      "loss": 0.1498,
+      "step": 15930
+    },
+    {
+      "epoch": 0.510031037020446,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.0315931289996633e-05,
+      "loss": 0.1483,
+      "step": 15940
+    },
+    {
+      "epoch": 0.5103510063033949,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0309195015156619e-05,
+      "loss": 0.1524,
+      "step": 15950
+    },
+    {
+      "epoch": 0.5106709755863437,
+      "grad_norm": 11.75,
+      "learning_rate": 1.0302458740316605e-05,
+      "loss": 0.1435,
+      "step": 15960
+    },
+    {
+      "epoch": 0.5109909448692925,
+      "grad_norm": 26.0,
+      "learning_rate": 1.0295722465476592e-05,
+      "loss": 0.1148,
+      "step": 15970
+    },
+    {
+      "epoch": 0.5113109141522414,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.0288986190636578e-05,
+      "loss": 0.1448,
+      "step": 15980
+    },
+    {
+      "epoch": 0.5116308834351903,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0282249915796565e-05,
+      "loss": 0.08,
+      "step": 15990
+    },
+    {
+      "epoch": 0.511950852718139,
+      "grad_norm": 5.21875,
+      "learning_rate": 1.0275513640956552e-05,
+      "loss": 0.079,
+      "step": 16000
+    },
+    {
+      "epoch": 0.5122708220010879,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.0268777366116538e-05,
+      "loss": 0.1187,
+      "step": 16010
+    },
+    {
+      "epoch": 0.5125907912840367,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.0262041091276524e-05,
+      "loss": 0.0895,
+      "step": 16020
+    },
+    {
+      "epoch": 0.5129107605669856,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.025530481643651e-05,
+      "loss": 0.0903,
+      "step": 16030
+    },
+    {
+      "epoch": 0.5132307298499345,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.0248568541596498e-05,
+      "loss": 0.0901,
+      "step": 16040
+    },
+    {
+      "epoch": 0.5135506991328832,
+      "grad_norm": 4.59375,
+      "learning_rate": 1.0241832266756484e-05,
+      "loss": 0.1249,
+      "step": 16050
+    },
+    {
+      "epoch": 0.5138706684158321,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.023509599191647e-05,
+      "loss": 0.1342,
+      "step": 16060
+    },
+    {
+      "epoch": 0.5141906376987809,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.0228359717076456e-05,
+      "loss": 0.1455,
+      "step": 16070
+    },
+    {
+      "epoch": 0.5145106069817298,
+      "grad_norm": 15.25,
+      "learning_rate": 1.0221623442236444e-05,
+      "loss": 0.0825,
+      "step": 16080
+    },
+    {
+      "epoch": 0.5148305762646785,
+      "grad_norm": 14.5,
+      "learning_rate": 1.021488716739643e-05,
+      "loss": 0.1039,
+      "step": 16090
+    },
+    {
+      "epoch": 0.5151505455476274,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.0208150892556416e-05,
+      "loss": 0.1111,
+      "step": 16100
+    },
+    {
+      "epoch": 0.5154705148305763,
+      "grad_norm": 15.875,
+      "learning_rate": 1.0201414617716404e-05,
+      "loss": 0.1266,
+      "step": 16110
+    },
+    {
+      "epoch": 0.5157904841135251,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.019467834287639e-05,
+      "loss": 0.0752,
+      "step": 16120
+    },
+    {
+      "epoch": 0.516110453396474,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.0187942068036376e-05,
+      "loss": 0.0807,
+      "step": 16130
+    },
+    {
+      "epoch": 0.5164304226794227,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.0181205793196362e-05,
+      "loss": 0.1098,
+      "step": 16140
+    },
+    {
+      "epoch": 0.5167503919623716,
+      "grad_norm": 9.25,
+      "learning_rate": 1.017446951835635e-05,
+      "loss": 0.1115,
+      "step": 16150
+    },
+    {
+      "epoch": 0.5170703612453205,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0167733243516336e-05,
+      "loss": 0.106,
+      "step": 16160
+    },
+    {
+      "epoch": 0.5173903305282693,
+      "grad_norm": 12.25,
+      "learning_rate": 1.0160996968676322e-05,
+      "loss": 0.1005,
+      "step": 16170
+    },
+    {
+      "epoch": 0.5177102998112181,
+      "grad_norm": 21.75,
+      "learning_rate": 1.015426069383631e-05,
+      "loss": 0.1228,
+      "step": 16180
+    },
+    {
+      "epoch": 0.5180302690941669,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.0147524418996296e-05,
+      "loss": 0.1322,
+      "step": 16190
+    },
+    {
+      "epoch": 0.5183502383771158,
+      "grad_norm": 5.09375,
+      "learning_rate": 1.0140788144156282e-05,
+      "loss": 0.1218,
+      "step": 16200
+    },
+    {
+      "epoch": 0.5186702076600647,
+      "grad_norm": 12.625,
+      "learning_rate": 1.0134051869316268e-05,
+      "loss": 0.1476,
+      "step": 16210
+    },
+    {
+      "epoch": 0.5189901769430135,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.0127315594476255e-05,
+      "loss": 0.1353,
+      "step": 16220
+    },
+    {
+      "epoch": 0.5193101462259623,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.0120579319636241e-05,
+      "loss": 0.0939,
+      "step": 16230
+    },
+    {
+      "epoch": 0.5196301155089111,
+      "grad_norm": 16.5,
+      "learning_rate": 1.0113843044796227e-05,
+      "loss": 0.1778,
+      "step": 16240
+    },
+    {
+      "epoch": 0.51995008479186,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.0107106769956213e-05,
+      "loss": 0.0831,
+      "step": 16250
+    },
+    {
+      "epoch": 0.5202700540748089,
+      "grad_norm": 25.5,
+      "learning_rate": 1.0100370495116201e-05,
+      "loss": 0.1169,
+      "step": 16260
+    },
+    {
+      "epoch": 0.5205900233577576,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0093634220276187e-05,
+      "loss": 0.1802,
+      "step": 16270
+    },
+    {
+      "epoch": 0.5209099926407065,
+      "grad_norm": 318.0,
+      "learning_rate": 1.0086897945436173e-05,
+      "loss": 0.1238,
+      "step": 16280
+    },
+    {
+      "epoch": 0.5212299619236553,
+      "grad_norm": 12.25,
+      "learning_rate": 1.0080161670596161e-05,
+      "loss": 0.0891,
+      "step": 16290
+    },
+    {
+      "epoch": 0.5215499312066042,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.0073425395756147e-05,
+      "loss": 0.1026,
+      "step": 16300
+    },
+    {
+      "epoch": 0.521869900489553,
+      "grad_norm": 9.375,
+      "learning_rate": 1.0066689120916133e-05,
+      "loss": 0.1511,
+      "step": 16310
+    },
+    {
+      "epoch": 0.5221898697725018,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.0059952846076119e-05,
+      "loss": 0.1108,
+      "step": 16320
+    },
+    {
+      "epoch": 0.5225098390554507,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0053216571236107e-05,
+      "loss": 0.1441,
+      "step": 16330
+    },
+    {
+      "epoch": 0.5228298083383995,
+      "grad_norm": 36.5,
+      "learning_rate": 1.0046480296396093e-05,
+      "loss": 0.0913,
+      "step": 16340
+    },
+    {
+      "epoch": 0.5231497776213484,
+      "grad_norm": 6.59375,
+      "learning_rate": 1.0039744021556079e-05,
+      "loss": 0.1113,
+      "step": 16350
+    },
+    {
+      "epoch": 0.5234697469042972,
+      "grad_norm": 24.625,
+      "learning_rate": 1.0033007746716065e-05,
+      "loss": 0.1369,
+      "step": 16360
+    },
+    {
+      "epoch": 0.523789716187246,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0026271471876053e-05,
+      "loss": 0.1328,
+      "step": 16370
+    },
+    {
+      "epoch": 0.5241096854701949,
+      "grad_norm": 15.875,
+      "learning_rate": 1.0019535197036039e-05,
+      "loss": 0.0887,
+      "step": 16380
+    },
+    {
+      "epoch": 0.5244296547531437,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.0012798922196025e-05,
+      "loss": 0.1961,
+      "step": 16390
+    },
+    {
+      "epoch": 0.5247496240360925,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.0006062647356013e-05,
+      "loss": 0.109,
+      "step": 16400
+    },
+    {
+      "epoch": 0.5250695933190414,
+      "grad_norm": 25.25,
+      "learning_rate": 9.999326372516e-06,
+      "loss": 0.1243,
+      "step": 16410
+    },
+    {
+      "epoch": 0.5253895626019902,
+      "grad_norm": 15.5,
+      "learning_rate": 9.992590097675986e-06,
+      "loss": 0.0806,
+      "step": 16420
+    },
+    {
+      "epoch": 0.5257095318849391,
+      "grad_norm": 16.5,
+      "learning_rate": 9.985853822835972e-06,
+      "loss": 0.1409,
+      "step": 16430
+    },
+    {
+      "epoch": 0.5260295011678879,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.979117547995958e-06,
+      "loss": 0.0668,
+      "step": 16440
+    },
+    {
+      "epoch": 0.5263494704508367,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.972381273155946e-06,
+      "loss": 0.0781,
+      "step": 16450
+    },
+    {
+      "epoch": 0.5266694397337855,
+      "grad_norm": 10.0,
+      "learning_rate": 9.965644998315932e-06,
+      "loss": 0.1415,
+      "step": 16460
+    },
+    {
+      "epoch": 0.5269894090167344,
+      "grad_norm": 11.75,
+      "learning_rate": 9.958908723475918e-06,
+      "loss": 0.0876,
+      "step": 16470
+    },
+    {
+      "epoch": 0.5273093782996833,
+      "grad_norm": 14.125,
+      "learning_rate": 9.952172448635904e-06,
+      "loss": 0.1419,
+      "step": 16480
+    },
+    {
+      "epoch": 0.527629347582632,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.945436173795892e-06,
+      "loss": 0.1513,
+      "step": 16490
+    },
+    {
+      "epoch": 0.5279493168655809,
+      "grad_norm": 13.1875,
+      "learning_rate": 9.938699898955878e-06,
+      "loss": 0.1461,
+      "step": 16500
+    },
+    {
+      "epoch": 0.5282692861485297,
+      "grad_norm": 29.125,
+      "learning_rate": 9.931963624115864e-06,
+      "loss": 0.1361,
+      "step": 16510
+    },
+    {
+      "epoch": 0.5285892554314786,
+      "grad_norm": 12.4375,
+      "learning_rate": 9.925227349275852e-06,
+      "loss": 0.1173,
+      "step": 16520
+    },
+    {
+      "epoch": 0.5289092247144274,
+      "grad_norm": 11.4375,
+      "learning_rate": 9.918491074435838e-06,
+      "loss": 0.1368,
+      "step": 16530
+    },
+    {
+      "epoch": 0.5292291939973762,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.911754799595824e-06,
+      "loss": 0.1205,
+      "step": 16540
+    },
+    {
+      "epoch": 0.5295491632803251,
+      "grad_norm": 13.125,
+      "learning_rate": 9.90501852475581e-06,
+      "loss": 0.0839,
+      "step": 16550
+    },
+    {
+      "epoch": 0.529869132563274,
+      "grad_norm": 7.34375,
+      "learning_rate": 9.898282249915798e-06,
+      "loss": 0.0928,
+      "step": 16560
+    },
+    {
+      "epoch": 0.5301891018462228,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.891545975075784e-06,
+      "loss": 0.1458,
+      "step": 16570
+    },
+    {
+      "epoch": 0.5305090711291716,
+      "grad_norm": 13.1875,
+      "learning_rate": 9.88480970023577e-06,
+      "loss": 0.0947,
+      "step": 16580
+    },
+    {
+      "epoch": 0.5308290404121204,
+      "grad_norm": 16.25,
+      "learning_rate": 9.878073425395757e-06,
+      "loss": 0.1135,
+      "step": 16590
+    },
+    {
+      "epoch": 0.5311490096950693,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.871337150555744e-06,
+      "loss": 0.1173,
+      "step": 16600
+    },
+    {
+      "epoch": 0.5314689789780181,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.86460087571573e-06,
+      "loss": 0.1075,
+      "step": 16610
+    },
+    {
+      "epoch": 0.5317889482609669,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.857864600875716e-06,
+      "loss": 0.1312,
+      "step": 16620
+    },
+    {
+      "epoch": 0.5321089175439158,
+      "grad_norm": 7.75,
+      "learning_rate": 9.851128326035703e-06,
+      "loss": 0.1079,
+      "step": 16630
+    },
+    {
+      "epoch": 0.5324288868268646,
+      "grad_norm": 17.25,
+      "learning_rate": 9.84439205119569e-06,
+      "loss": 0.1423,
+      "step": 16640
+    },
+    {
+      "epoch": 0.5327488561098135,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.837655776355675e-06,
+      "loss": 0.1032,
+      "step": 16650
+    },
+    {
+      "epoch": 0.5330688253927623,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.830919501515661e-06,
+      "loss": 0.0872,
+      "step": 16660
+    },
+    {
+      "epoch": 0.5333887946757111,
+      "grad_norm": 15.5,
+      "learning_rate": 9.82418322667565e-06,
+      "loss": 0.1218,
+      "step": 16670
+    },
+    {
+      "epoch": 0.53370876395866,
+      "grad_norm": 9.625,
+      "learning_rate": 9.817446951835635e-06,
+      "loss": 0.1358,
+      "step": 16680
+    },
+    {
+      "epoch": 0.5340287332416088,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.810710676995621e-06,
+      "loss": 0.0717,
+      "step": 16690
+    },
+    {
+      "epoch": 0.5343487025245577,
+      "grad_norm": 48.25,
+      "learning_rate": 9.803974402155609e-06,
+      "loss": 0.1161,
+      "step": 16700
+    },
+    {
+      "epoch": 0.5346686718075064,
+      "grad_norm": 14.5625,
+      "learning_rate": 9.797238127315595e-06,
+      "loss": 0.0845,
+      "step": 16710
+    },
+    {
+      "epoch": 0.5349886410904553,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.790501852475581e-06,
+      "loss": 0.0844,
+      "step": 16720
+    },
+    {
+      "epoch": 0.5353086103734042,
+      "grad_norm": 18.5,
+      "learning_rate": 9.783765577635567e-06,
+      "loss": 0.1187,
+      "step": 16730
+    },
+    {
+      "epoch": 0.535628579656353,
+      "grad_norm": 9.3125,
+      "learning_rate": 9.777029302795555e-06,
+      "loss": 0.0853,
+      "step": 16740
+    },
+    {
+      "epoch": 0.5359485489393018,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.770293027955543e-06,
+      "loss": 0.0776,
+      "step": 16750
+    },
+    {
+      "epoch": 0.5362685182222506,
+      "grad_norm": 13.5625,
+      "learning_rate": 9.763556753115529e-06,
+      "loss": 0.0546,
+      "step": 16760
+    },
+    {
+      "epoch": 0.5365884875051995,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.756820478275515e-06,
+      "loss": 0.1381,
+      "step": 16770
+    },
+    {
+      "epoch": 0.5369084567881484,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.7500842034355e-06,
+      "loss": 0.1048,
+      "step": 16780
+    },
+    {
+      "epoch": 0.5372284260710972,
+      "grad_norm": 9.5625,
+      "learning_rate": 9.743347928595488e-06,
+      "loss": 0.1545,
+      "step": 16790
+    },
+    {
+      "epoch": 0.537548395354046,
+      "grad_norm": 8.0,
+      "learning_rate": 9.736611653755475e-06,
+      "loss": 0.0584,
+      "step": 16800
+    },
+    {
+      "epoch": 0.5378683646369948,
+      "grad_norm": 11.0,
+      "learning_rate": 9.72987537891546e-06,
+      "loss": 0.127,
+      "step": 16810
+    },
+    {
+      "epoch": 0.5381883339199437,
+      "grad_norm": 16.5,
+      "learning_rate": 9.723139104075448e-06,
+      "loss": 0.1087,
+      "step": 16820
+    },
+    {
+      "epoch": 0.5385083032028926,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.716402829235434e-06,
+      "loss": 0.1302,
+      "step": 16830
+    },
+    {
+      "epoch": 0.5388282724858413,
+      "grad_norm": 24.375,
+      "learning_rate": 9.70966655439542e-06,
+      "loss": 0.0874,
+      "step": 16840
+    },
+    {
+      "epoch": 0.5391482417687902,
+      "grad_norm": 17.75,
+      "learning_rate": 9.702930279555406e-06,
+      "loss": 0.0768,
+      "step": 16850
+    },
+    {
+      "epoch": 0.539468211051739,
+      "grad_norm": 10.625,
+      "learning_rate": 9.696194004715394e-06,
+      "loss": 0.1197,
+      "step": 16860
+    },
+    {
+      "epoch": 0.5397881803346879,
+      "grad_norm": 23.25,
+      "learning_rate": 9.68945772987538e-06,
+      "loss": 0.1135,
+      "step": 16870
+    },
+    {
+      "epoch": 0.5401081496176368,
+      "grad_norm": 5.375,
+      "learning_rate": 9.682721455035366e-06,
+      "loss": 0.1187,
+      "step": 16880
+    },
+    {
+      "epoch": 0.5404281189005855,
+      "grad_norm": 14.9375,
+      "learning_rate": 9.675985180195352e-06,
+      "loss": 0.1447,
+      "step": 16890
+    },
+    {
+      "epoch": 0.5407480881835344,
+      "grad_norm": 6.5,
+      "learning_rate": 9.66924890535534e-06,
+      "loss": 0.073,
+      "step": 16900
+    },
+    {
+      "epoch": 0.5410680574664832,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.662512630515326e-06,
+      "loss": 0.1211,
+      "step": 16910
+    },
+    {
+      "epoch": 0.5413880267494321,
+      "grad_norm": 10.625,
+      "learning_rate": 9.655776355675312e-06,
+      "loss": 0.103,
+      "step": 16920
+    },
+    {
+      "epoch": 0.5417079960323808,
+      "grad_norm": 2.25,
+      "learning_rate": 9.6490400808353e-06,
+      "loss": 0.1791,
+      "step": 16930
+    },
+    {
+      "epoch": 0.5420279653153297,
+      "grad_norm": 8.9375,
+      "learning_rate": 9.642303805995286e-06,
+      "loss": 0.1161,
+      "step": 16940
+    },
+    {
+      "epoch": 0.5423479345982786,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.635567531155272e-06,
+      "loss": 0.1118,
+      "step": 16950
+    },
+    {
+      "epoch": 0.5426679038812274,
+      "grad_norm": 10.8125,
+      "learning_rate": 9.628831256315258e-06,
+      "loss": 0.1129,
+      "step": 16960
+    },
+    {
+      "epoch": 0.5429878731641763,
+      "grad_norm": 16.25,
+      "learning_rate": 9.622094981475246e-06,
+      "loss": 0.1488,
+      "step": 16970
+    },
+    {
+      "epoch": 0.543307842447125,
+      "grad_norm": 7.34375,
+      "learning_rate": 9.615358706635232e-06,
+      "loss": 0.1121,
+      "step": 16980
+    },
+    {
+      "epoch": 0.5436278117300739,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.608622431795218e-06,
+      "loss": 0.1185,
+      "step": 16990
+    },
+    {
+      "epoch": 0.5439477810130228,
+      "grad_norm": 16.375,
+      "learning_rate": 9.601886156955204e-06,
+      "loss": 0.0769,
+      "step": 17000
+    },
+    {
+      "epoch": 0.5442677502959716,
+      "grad_norm": 12.25,
+      "learning_rate": 9.595149882115192e-06,
+      "loss": 0.1605,
+      "step": 17010
+    },
+    {
+      "epoch": 0.5445877195789204,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.588413607275178e-06,
+      "loss": 0.1395,
+      "step": 17020
+    },
+    {
+      "epoch": 0.5449076888618692,
+      "grad_norm": 21.5,
+      "learning_rate": 9.581677332435164e-06,
+      "loss": 0.1563,
+      "step": 17030
+    },
+    {
+      "epoch": 0.5452276581448181,
+      "grad_norm": 22.875,
+      "learning_rate": 9.574941057595151e-06,
+      "loss": 0.1045,
+      "step": 17040
+    },
+    {
+      "epoch": 0.545547627427767,
+      "grad_norm": 20.875,
+      "learning_rate": 9.568204782755137e-06,
+      "loss": 0.175,
+      "step": 17050
+    },
+    {
+      "epoch": 0.5458675967107157,
+      "grad_norm": 23.625,
+      "learning_rate": 9.561468507915123e-06,
+      "loss": 0.1244,
+      "step": 17060
+    },
+    {
+      "epoch": 0.5461875659936646,
+      "grad_norm": 16.5,
+      "learning_rate": 9.55473223307511e-06,
+      "loss": 0.0841,
+      "step": 17070
+    },
+    {
+      "epoch": 0.5465075352766134,
+      "grad_norm": 10.625,
+      "learning_rate": 9.547995958235097e-06,
+      "loss": 0.1042,
+      "step": 17080
+    },
+    {
+      "epoch": 0.5468275045595623,
+      "grad_norm": 9.75,
+      "learning_rate": 9.541259683395083e-06,
+      "loss": 0.0513,
+      "step": 17090
+    },
+    {
+      "epoch": 0.5471474738425112,
+      "grad_norm": 10.125,
+      "learning_rate": 9.53452340855507e-06,
+      "loss": 0.1127,
+      "step": 17100
+    },
+    {
+      "epoch": 0.5474674431254599,
+      "grad_norm": 16.125,
+      "learning_rate": 9.527787133715057e-06,
+      "loss": 0.1141,
+      "step": 17110
+    },
+    {
+      "epoch": 0.5477874124084088,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.521050858875043e-06,
+      "loss": 0.1045,
+      "step": 17120
+    },
+    {
+      "epoch": 0.5481073816913576,
+      "grad_norm": 4.875,
+      "learning_rate": 9.514314584035029e-06,
+      "loss": 0.0447,
+      "step": 17130
+    },
+    {
+      "epoch": 0.5484273509743065,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.507578309195015e-06,
+      "loss": 0.1104,
+      "step": 17140
+    },
+    {
+      "epoch": 0.5487473202572553,
+      "grad_norm": 19.25,
+      "learning_rate": 9.500842034355003e-06,
+      "loss": 0.1749,
+      "step": 17150
+    },
+    {
+      "epoch": 0.5490672895402041,
+      "grad_norm": 15.5625,
+      "learning_rate": 9.494105759514989e-06,
+      "loss": 0.0682,
+      "step": 17160
+    },
+    {
+      "epoch": 0.549387258823153,
+      "grad_norm": 10.125,
+      "learning_rate": 9.487369484674975e-06,
+      "loss": 0.0973,
+      "step": 17170
+    },
+    {
+      "epoch": 0.5497072281061018,
+      "grad_norm": 11.9375,
+      "learning_rate": 9.480633209834961e-06,
+      "loss": 0.1071,
+      "step": 17180
+    },
+    {
+      "epoch": 0.5500271973890507,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.473896934994949e-06,
+      "loss": 0.0739,
+      "step": 17190
+    },
+    {
+      "epoch": 0.5503471666719995,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.467160660154935e-06,
+      "loss": 0.1195,
+      "step": 17200
+    },
+    {
+      "epoch": 0.5506671359549483,
+      "grad_norm": 13.125,
+      "learning_rate": 9.46042438531492e-06,
+      "loss": 0.1357,
+      "step": 17210
+    },
+    {
+      "epoch": 0.5509871052378972,
+      "grad_norm": 6.75,
+      "learning_rate": 9.453688110474909e-06,
+      "loss": 0.0672,
+      "step": 17220
+    },
+    {
+      "epoch": 0.551307074520846,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.446951835634895e-06,
+      "loss": 0.1207,
+      "step": 17230
+    },
+    {
+      "epoch": 0.5516270438037948,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.44021556079488e-06,
+      "loss": 0.135,
+      "step": 17240
+    },
+    {
+      "epoch": 0.5519470130867437,
+      "grad_norm": 10.1875,
+      "learning_rate": 9.433479285954867e-06,
+      "loss": 0.1543,
+      "step": 17250
+    },
+    {
+      "epoch": 0.5522669823696925,
+      "grad_norm": 15.125,
+      "learning_rate": 9.426743011114854e-06,
+      "loss": 0.1163,
+      "step": 17260
+    },
+    {
+      "epoch": 0.5525869516526414,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.42000673627484e-06,
+      "loss": 0.084,
+      "step": 17270
+    },
+    {
+      "epoch": 0.5529069209355901,
+      "grad_norm": 32.0,
+      "learning_rate": 9.413270461434827e-06,
+      "loss": 0.0906,
+      "step": 17280
+    },
+    {
+      "epoch": 0.553226890218539,
+      "grad_norm": 28.75,
+      "learning_rate": 9.406534186594814e-06,
+      "loss": 0.092,
+      "step": 17290
+    },
+    {
+      "epoch": 0.5535468595014879,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.3997979117548e-06,
+      "loss": 0.074,
+      "step": 17300
+    },
+    {
+      "epoch": 0.5538668287844367,
+      "grad_norm": 9.4375,
+      "learning_rate": 9.393061636914786e-06,
+      "loss": 0.0844,
+      "step": 17310
+    },
+    {
+      "epoch": 0.5541867980673856,
+      "grad_norm": 22.0,
+      "learning_rate": 9.386325362074772e-06,
+      "loss": 0.1496,
+      "step": 17320
+    },
+    {
+      "epoch": 0.5545067673503343,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.37958908723476e-06,
+      "loss": 0.092,
+      "step": 17330
+    },
+    {
+      "epoch": 0.5548267366332832,
+      "grad_norm": 5.875,
+      "learning_rate": 9.372852812394746e-06,
+      "loss": 0.0977,
+      "step": 17340
+    },
+    {
+      "epoch": 0.555146705916232,
+      "grad_norm": 9.875,
+      "learning_rate": 9.366116537554732e-06,
+      "loss": 0.1711,
+      "step": 17350
+    },
+    {
+      "epoch": 0.5554666751991809,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.359380262714718e-06,
+      "loss": 0.1138,
+      "step": 17360
+    },
+    {
+      "epoch": 0.5557866444821297,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.352643987874706e-06,
+      "loss": 0.1155,
+      "step": 17370
+    },
+    {
+      "epoch": 0.5561066137650785,
+      "grad_norm": 13.875,
+      "learning_rate": 9.345907713034694e-06,
+      "loss": 0.1035,
+      "step": 17380
+    },
+    {
+      "epoch": 0.5564265830480274,
+      "grad_norm": 6.40625,
+      "learning_rate": 9.33917143819468e-06,
+      "loss": 0.1296,
+      "step": 17390
+    },
+    {
+      "epoch": 0.5567465523309763,
+      "grad_norm": 7.84375,
+      "learning_rate": 9.332435163354666e-06,
+      "loss": 0.0792,
+      "step": 17400
+    },
+    {
+      "epoch": 0.5570665216139251,
+      "grad_norm": 7.9375,
+      "learning_rate": 9.325698888514652e-06,
+      "loss": 0.1096,
+      "step": 17410
+    },
+    {
+      "epoch": 0.5573864908968739,
+      "grad_norm": 11.3125,
+      "learning_rate": 9.31896261367464e-06,
+      "loss": 0.1383,
+      "step": 17420
+    },
+    {
+      "epoch": 0.5577064601798227,
+      "grad_norm": 12.375,
+      "learning_rate": 9.312226338834626e-06,
+      "loss": 0.0889,
+      "step": 17430
+    },
+    {
+      "epoch": 0.5580264294627716,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.305490063994612e-06,
+      "loss": 0.1013,
+      "step": 17440
+    },
+    {
+      "epoch": 0.5583463987457205,
+      "grad_norm": 20.0,
+      "learning_rate": 9.2987537891546e-06,
+      "loss": 0.0904,
+      "step": 17450
+    },
+    {
+      "epoch": 0.5586663680286692,
+      "grad_norm": 14.4375,
+      "learning_rate": 9.292017514314585e-06,
+      "loss": 0.092,
+      "step": 17460
+    },
+    {
+      "epoch": 0.5589863373116181,
+      "grad_norm": 9.875,
+      "learning_rate": 9.285281239474571e-06,
+      "loss": 0.1124,
+      "step": 17470
+    },
+    {
+      "epoch": 0.5593063065945669,
+      "grad_norm": 6.96875,
+      "learning_rate": 9.278544964634558e-06,
+      "loss": 0.0693,
+      "step": 17480
+    },
+    {
+      "epoch": 0.5596262758775158,
+      "grad_norm": 9.8125,
+      "learning_rate": 9.271808689794545e-06,
+      "loss": 0.0827,
+      "step": 17490
+    },
+    {
+      "epoch": 0.5599462451604647,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.265072414954531e-06,
+      "loss": 0.0921,
+      "step": 17500
+    },
+    {
+      "epoch": 0.5602662144434134,
+      "grad_norm": 11.25,
+      "learning_rate": 9.258336140114517e-06,
+      "loss": 0.1227,
+      "step": 17510
+    },
+    {
+      "epoch": 0.5605861837263623,
+      "grad_norm": 20.0,
+      "learning_rate": 9.251599865274505e-06,
+      "loss": 0.2023,
+      "step": 17520
+    },
+    {
+      "epoch": 0.5609061530093111,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.244863590434491e-06,
+      "loss": 0.0955,
+      "step": 17530
+    },
+    {
+      "epoch": 0.56122612229226,
+      "grad_norm": 12.125,
+      "learning_rate": 9.238127315594477e-06,
+      "loss": 0.0827,
+      "step": 17540
+    },
+    {
+      "epoch": 0.5615460915752087,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.231391040754463e-06,
+      "loss": 0.1101,
+      "step": 17550
+    },
+    {
+      "epoch": 0.5618660608581576,
+      "grad_norm": 16.75,
+      "learning_rate": 9.224654765914451e-06,
+      "loss": 0.1245,
+      "step": 17560
+    },
+    {
+      "epoch": 0.5621860301411065,
+      "grad_norm": 6.15625,
+      "learning_rate": 9.217918491074437e-06,
+      "loss": 0.1227,
+      "step": 17570
+    },
+    {
+      "epoch": 0.5625059994240553,
+      "grad_norm": 37.25,
+      "learning_rate": 9.211182216234423e-06,
+      "loss": 0.1178,
+      "step": 17580
+    },
+    {
+      "epoch": 0.5628259687070041,
+      "grad_norm": 13.375,
+      "learning_rate": 9.204445941394409e-06,
+      "loss": 0.1369,
+      "step": 17590
+    },
+    {
+      "epoch": 0.5631459379899529,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.197709666554397e-06,
+      "loss": 0.1473,
+      "step": 17600
+    },
+    {
+      "epoch": 0.5634659072729018,
+      "grad_norm": 23.75,
+      "learning_rate": 9.190973391714383e-06,
+      "loss": 0.1329,
+      "step": 17610
+    },
+    {
+      "epoch": 0.5637858765558507,
+      "grad_norm": 18.125,
+      "learning_rate": 9.184237116874369e-06,
+      "loss": 0.0889,
+      "step": 17620
+    },
+    {
+      "epoch": 0.5641058458387995,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.177500842034357e-06,
+      "loss": 0.0962,
+      "step": 17630
+    },
+    {
+      "epoch": 0.5644258151217483,
+      "grad_norm": 27.0,
+      "learning_rate": 9.170764567194343e-06,
+      "loss": 0.1296,
+      "step": 17640
+    },
+    {
+      "epoch": 0.5647457844046971,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.164028292354329e-06,
+      "loss": 0.1092,
+      "step": 17650
+    },
+    {
+      "epoch": 0.565065753687646,
+      "grad_norm": 9.25,
+      "learning_rate": 9.157292017514315e-06,
+      "loss": 0.1075,
+      "step": 17660
+    },
+    {
+      "epoch": 0.5653857229705949,
+      "grad_norm": 16.75,
+      "learning_rate": 9.150555742674302e-06,
+      "loss": 0.1134,
+      "step": 17670
+    },
+    {
+      "epoch": 0.5657056922535436,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.143819467834289e-06,
+      "loss": 0.1229,
+      "step": 17680
+    },
+    {
+      "epoch": 0.5660256615364925,
+      "grad_norm": 12.75,
+      "learning_rate": 9.137083192994275e-06,
+      "loss": 0.1269,
+      "step": 17690
+    },
+    {
+      "epoch": 0.5663456308194413,
+      "grad_norm": 4.125,
+      "learning_rate": 9.130346918154262e-06,
+      "loss": 0.1302,
+      "step": 17700
+    },
+    {
+      "epoch": 0.5666656001023902,
+      "grad_norm": 8.75,
+      "learning_rate": 9.123610643314248e-06,
+      "loss": 0.2007,
+      "step": 17710
+    },
+    {
+      "epoch": 0.5669855693853391,
+      "grad_norm": 16.625,
+      "learning_rate": 9.116874368474234e-06,
+      "loss": 0.127,
+      "step": 17720
+    },
+    {
+      "epoch": 0.5673055386682878,
+      "grad_norm": 17.5,
+      "learning_rate": 9.11013809363422e-06,
+      "loss": 0.1337,
+      "step": 17730
+    },
+    {
+      "epoch": 0.5676255079512367,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.103401818794208e-06,
+      "loss": 0.0603,
+      "step": 17740
+    },
+    {
+      "epoch": 0.5679454772341855,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.096665543954194e-06,
+      "loss": 0.07,
+      "step": 17750
+    },
+    {
+      "epoch": 0.5682654465171344,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.08992926911418e-06,
+      "loss": 0.0582,
+      "step": 17760
+    },
+    {
+      "epoch": 0.5685854158000831,
+      "grad_norm": 18.0,
+      "learning_rate": 9.083192994274166e-06,
+      "loss": 0.1592,
+      "step": 17770
+    },
+    {
+      "epoch": 0.568905385083032,
+      "grad_norm": 11.625,
+      "learning_rate": 9.076456719434154e-06,
+      "loss": 0.1247,
+      "step": 17780
+    },
+    {
+      "epoch": 0.5692253543659809,
+      "grad_norm": 9.6875,
+      "learning_rate": 9.06972044459414e-06,
+      "loss": 0.0788,
+      "step": 17790
+    },
+    {
+      "epoch": 0.5695453236489297,
+      "grad_norm": 15.75,
+      "learning_rate": 9.062984169754126e-06,
+      "loss": 0.0767,
+      "step": 17800
+    },
+    {
+      "epoch": 0.5698652929318785,
+      "grad_norm": 3.125,
+      "learning_rate": 9.056247894914114e-06,
+      "loss": 0.0748,
+      "step": 17810
+    },
+    {
+      "epoch": 0.5701852622148273,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.0495116200741e-06,
+      "loss": 0.1468,
+      "step": 17820
+    },
+    {
+      "epoch": 0.5705052314977762,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.042775345234086e-06,
+      "loss": 0.0623,
+      "step": 17830
+    },
+    {
+      "epoch": 0.5708252007807251,
+      "grad_norm": 15.375,
+      "learning_rate": 9.036039070394072e-06,
+      "loss": 0.0719,
+      "step": 17840
+    },
+    {
+      "epoch": 0.5711451700636739,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.02930279555406e-06,
+      "loss": 0.1176,
+      "step": 17850
+    },
+    {
+      "epoch": 0.5714651393466227,
+      "grad_norm": 30.125,
+      "learning_rate": 9.022566520714046e-06,
+      "loss": 0.075,
+      "step": 17860
+    },
+    {
+      "epoch": 0.5717851086295715,
+      "grad_norm": 26.75,
+      "learning_rate": 9.015830245874032e-06,
+      "loss": 0.093,
+      "step": 17870
+    },
+    {
+      "epoch": 0.5721050779125204,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.009093971034018e-06,
+      "loss": 0.1431,
+      "step": 17880
+    },
+    {
+      "epoch": 0.5724250471954693,
+      "grad_norm": 6.9375,
+      "learning_rate": 9.002357696194006e-06,
+      "loss": 0.1061,
+      "step": 17890
+    },
+    {
+      "epoch": 0.572745016478418,
+      "grad_norm": 4.71875,
+      "learning_rate": 8.995621421353992e-06,
+      "loss": 0.1591,
+      "step": 17900
+    },
+    {
+      "epoch": 0.5730649857613669,
+      "grad_norm": 17.25,
+      "learning_rate": 8.988885146513978e-06,
+      "loss": 0.1671,
+      "step": 17910
+    },
+    {
+      "epoch": 0.5733849550443157,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.982148871673965e-06,
+      "loss": 0.087,
+      "step": 17920
+    },
+    {
+      "epoch": 0.5737049243272646,
+      "grad_norm": 10.0,
+      "learning_rate": 8.975412596833951e-06,
+      "loss": 0.0833,
+      "step": 17930
+    },
+    {
+      "epoch": 0.5740248936102135,
+      "grad_norm": 9.1875,
+      "learning_rate": 8.968676321993937e-06,
+      "loss": 0.0906,
+      "step": 17940
+    },
+    {
+      "epoch": 0.5743448628931622,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.961940047153923e-06,
+      "loss": 0.1227,
+      "step": 17950
+    },
+    {
+      "epoch": 0.5746648321761111,
+      "grad_norm": 10.4375,
+      "learning_rate": 8.955203772313911e-06,
+      "loss": 0.106,
+      "step": 17960
+    },
+    {
+      "epoch": 0.57498480145906,
+      "grad_norm": 8.625,
+      "learning_rate": 8.948467497473897e-06,
+      "loss": 0.0896,
+      "step": 17970
+    },
+    {
+      "epoch": 0.5753047707420088,
+      "grad_norm": 29.125,
+      "learning_rate": 8.941731222633883e-06,
+      "loss": 0.1356,
+      "step": 17980
+    },
+    {
+      "epoch": 0.5756247400249576,
+      "grad_norm": 8.875,
+      "learning_rate": 8.934994947793871e-06,
+      "loss": 0.1414,
+      "step": 17990
+    },
+    {
+      "epoch": 0.5759447093079064,
+      "grad_norm": 10.0625,
+      "learning_rate": 8.928258672953857e-06,
+      "loss": 0.1731,
+      "step": 18000
+    },
+    {
+      "epoch": 0.5762646785908553,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.921522398113843e-06,
+      "loss": 0.1153,
+      "step": 18010
+    },
+    {
+      "epoch": 0.5765846478738041,
+      "grad_norm": 10.25,
+      "learning_rate": 8.91478612327383e-06,
+      "loss": 0.1249,
+      "step": 18020
+    },
+    {
+      "epoch": 0.5769046171567529,
+      "grad_norm": 9.9375,
+      "learning_rate": 8.908049848433817e-06,
+      "loss": 0.0728,
+      "step": 18030
+    },
+    {
+      "epoch": 0.5772245864397018,
+      "grad_norm": 12.375,
+      "learning_rate": 8.901313573593805e-06,
+      "loss": 0.119,
+      "step": 18040
+    },
+    {
+      "epoch": 0.5775445557226506,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.89457729875379e-06,
+      "loss": 0.1275,
+      "step": 18050
+    },
+    {
+      "epoch": 0.5778645250055995,
+      "grad_norm": 6.28125,
+      "learning_rate": 8.887841023913777e-06,
+      "loss": 0.0936,
+      "step": 18060
+    },
+    {
+      "epoch": 0.5781844942885483,
+      "grad_norm": 6.84375,
+      "learning_rate": 8.881104749073763e-06,
+      "loss": 0.1294,
+      "step": 18070
+    },
+    {
+      "epoch": 0.5785044635714971,
+      "grad_norm": 17.25,
+      "learning_rate": 8.87436847423375e-06,
+      "loss": 0.1323,
+      "step": 18080
+    },
+    {
+      "epoch": 0.578824432854446,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.867632199393737e-06,
+      "loss": 0.0497,
+      "step": 18090
+    },
+    {
+      "epoch": 0.5791444021373948,
+      "grad_norm": 8.8125,
+      "learning_rate": 8.860895924553723e-06,
+      "loss": 0.0782,
+      "step": 18100
+    },
+    {
+      "epoch": 0.5794643714203437,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.854159649713709e-06,
+      "loss": 0.0648,
+      "step": 18110
+    },
+    {
+      "epoch": 0.5797843407032924,
+      "grad_norm": 11.8125,
+      "learning_rate": 8.847423374873696e-06,
+      "loss": 0.0994,
+      "step": 18120
+    },
+    {
+      "epoch": 0.5801043099862413,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.840687100033682e-06,
+      "loss": 0.1044,
+      "step": 18130
+    },
+    {
+      "epoch": 0.5804242792691902,
+      "grad_norm": 6.71875,
+      "learning_rate": 8.833950825193668e-06,
+      "loss": 0.1147,
+      "step": 18140
+    },
+    {
+      "epoch": 0.580744248552139,
+      "grad_norm": 4.5,
+      "learning_rate": 8.827214550353656e-06,
+      "loss": 0.1071,
+      "step": 18150
+    },
+    {
+      "epoch": 0.5810642178350879,
+      "grad_norm": 5.53125,
+      "learning_rate": 8.820478275513642e-06,
+      "loss": 0.0716,
+      "step": 18160
+    },
+    {
+      "epoch": 0.5813841871180366,
+      "grad_norm": 13.8125,
+      "learning_rate": 8.813742000673628e-06,
+      "loss": 0.1637,
+      "step": 18170
+    },
+    {
+      "epoch": 0.5817041564009855,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.807005725833614e-06,
+      "loss": 0.0763,
+      "step": 18180
+    },
+    {
+      "epoch": 0.5820241256839344,
+      "grad_norm": 35.25,
+      "learning_rate": 8.800269450993602e-06,
+      "loss": 0.1926,
+      "step": 18190
+    },
+    {
+      "epoch": 0.5823440949668832,
+      "grad_norm": 3.96875,
+      "learning_rate": 8.793533176153588e-06,
+      "loss": 0.1365,
+      "step": 18200
+    },
+    {
+      "epoch": 0.582664064249832,
+      "grad_norm": 17.625,
+      "learning_rate": 8.786796901313574e-06,
+      "loss": 0.1747,
+      "step": 18210
+    },
+    {
+      "epoch": 0.5829840335327808,
+      "grad_norm": 4.28125,
+      "learning_rate": 8.780060626473562e-06,
+      "loss": 0.0777,
+      "step": 18220
+    },
+    {
+      "epoch": 0.5833040028157297,
+      "grad_norm": 7.28125,
+      "learning_rate": 8.773324351633548e-06,
+      "loss": 0.172,
+      "step": 18230
+    },
+    {
+      "epoch": 0.5836239720986786,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.766588076793534e-06,
+      "loss": 0.0922,
+      "step": 18240
+    },
+    {
+      "epoch": 0.5839439413816274,
+      "grad_norm": 10.9375,
+      "learning_rate": 8.75985180195352e-06,
+      "loss": 0.1585,
+      "step": 18250
+    },
+    {
+      "epoch": 0.5842639106645762,
+      "grad_norm": 18.625,
+      "learning_rate": 8.753115527113508e-06,
+      "loss": 0.195,
+      "step": 18260
+    },
+    {
+      "epoch": 0.584583879947525,
+      "grad_norm": 39.75,
+      "learning_rate": 8.746379252273494e-06,
+      "loss": 0.1074,
+      "step": 18270
+    },
+    {
+      "epoch": 0.5849038492304739,
+      "grad_norm": 15.6875,
+      "learning_rate": 8.73964297743348e-06,
+      "loss": 0.1355,
+      "step": 18280
+    },
+    {
+      "epoch": 0.5852238185134228,
+      "grad_norm": 8.25,
+      "learning_rate": 8.732906702593466e-06,
+      "loss": 0.1012,
+      "step": 18290
+    },
+    {
+      "epoch": 0.5855437877963715,
+      "grad_norm": 12.75,
+      "learning_rate": 8.726170427753454e-06,
+      "loss": 0.1346,
+      "step": 18300
+    },
+    {
+      "epoch": 0.5858637570793204,
+      "grad_norm": 12.125,
+      "learning_rate": 8.71943415291344e-06,
+      "loss": 0.0888,
+      "step": 18310
+    },
+    {
+      "epoch": 0.5861837263622692,
+      "grad_norm": 4.6875,
+      "learning_rate": 8.712697878073426e-06,
+      "loss": 0.0843,
+      "step": 18320
+    },
+    {
+      "epoch": 0.5865036956452181,
+      "grad_norm": 11.1875,
+      "learning_rate": 8.705961603233413e-06,
+      "loss": 0.1001,
+      "step": 18330
+    },
+    {
+      "epoch": 0.5868236649281668,
+      "grad_norm": 12.0625,
+      "learning_rate": 8.6992253283934e-06,
+      "loss": 0.0906,
+      "step": 18340
+    },
+    {
+      "epoch": 0.5871436342111157,
+      "grad_norm": 13.5,
+      "learning_rate": 8.692489053553385e-06,
+      "loss": 0.0789,
+      "step": 18350
+    },
+    {
+      "epoch": 0.5874636034940646,
+      "grad_norm": 43.5,
+      "learning_rate": 8.685752778713372e-06,
+      "loss": 0.1321,
+      "step": 18360
+    },
+    {
+      "epoch": 0.5877835727770134,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.67901650387336e-06,
+      "loss": 0.1096,
+      "step": 18370
+    },
+    {
+      "epoch": 0.5881035420599623,
+      "grad_norm": 3.875,
+      "learning_rate": 8.672280229033345e-06,
+      "loss": 0.1238,
+      "step": 18380
+    },
+    {
+      "epoch": 0.588423511342911,
+      "grad_norm": 11.0625,
+      "learning_rate": 8.665543954193331e-06,
+      "loss": 0.1241,
+      "step": 18390
+    },
+    {
+      "epoch": 0.5887434806258599,
+      "grad_norm": 7.65625,
+      "learning_rate": 8.658807679353319e-06,
+      "loss": 0.0787,
+      "step": 18400
+    },
+    {
+      "epoch": 0.5890634499088088,
+      "grad_norm": 6.75,
+      "learning_rate": 8.652071404513305e-06,
+      "loss": 0.1444,
+      "step": 18410
+    },
+    {
+      "epoch": 0.5893834191917576,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.645335129673291e-06,
+      "loss": 0.0956,
+      "step": 18420
+    },
+    {
+      "epoch": 0.5897033884747064,
+      "grad_norm": 11.875,
+      "learning_rate": 8.638598854833277e-06,
+      "loss": 0.0662,
+      "step": 18430
+    },
+    {
+      "epoch": 0.5900233577576552,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.631862579993265e-06,
+      "loss": 0.0972,
+      "step": 18440
+    },
+    {
+      "epoch": 0.5903433270406041,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.625126305153251e-06,
+      "loss": 0.1032,
+      "step": 18450
+    },
+    {
+      "epoch": 0.590663296323553,
+      "grad_norm": 48.75,
+      "learning_rate": 8.618390030313237e-06,
+      "loss": 0.1035,
+      "step": 18460
+    },
+    {
+      "epoch": 0.5909832656065018,
+      "grad_norm": 12.25,
+      "learning_rate": 8.611653755473223e-06,
+      "loss": 0.1245,
+      "step": 18470
+    },
+    {
+      "epoch": 0.5913032348894506,
+      "grad_norm": 4.0625,
+      "learning_rate": 8.60491748063321e-06,
+      "loss": 0.1371,
+      "step": 18480
+    },
+    {
+      "epoch": 0.5916232041723994,
+      "grad_norm": 13.125,
+      "learning_rate": 8.598181205793197e-06,
+      "loss": 0.1607,
+      "step": 18490
+    },
+    {
+      "epoch": 0.5919431734553483,
+      "grad_norm": 12.375,
+      "learning_rate": 8.591444930953183e-06,
+      "loss": 0.14,
+      "step": 18500
+    },
+    {
+      "epoch": 0.5922631427382972,
+      "grad_norm": 4.4375,
+      "learning_rate": 8.58470865611317e-06,
+      "loss": 0.1604,
+      "step": 18510
+    },
+    {
+      "epoch": 0.5925831120212459,
+      "grad_norm": 12.6875,
+      "learning_rate": 8.577972381273157e-06,
+      "loss": 0.1063,
+      "step": 18520
+    },
+    {
+      "epoch": 0.5929030813041948,
+      "grad_norm": 7.6875,
+      "learning_rate": 8.571236106433143e-06,
+      "loss": 0.14,
+      "step": 18530
+    },
+    {
+      "epoch": 0.5932230505871436,
+      "grad_norm": 10.875,
+      "learning_rate": 8.564499831593129e-06,
+      "loss": 0.1498,
+      "step": 18540
+    },
+    {
+      "epoch": 0.5935430198700925,
+      "grad_norm": 10.625,
+      "learning_rate": 8.557763556753116e-06,
+      "loss": 0.0913,
+      "step": 18550
+    },
+    {
+      "epoch": 0.5938629891530413,
+      "grad_norm": 4.0,
+      "learning_rate": 8.551027281913102e-06,
+      "loss": 0.103,
+      "step": 18560
+    },
+    {
+      "epoch": 0.5941829584359901,
+      "grad_norm": 3.875,
+      "learning_rate": 8.544291007073089e-06,
+      "loss": 0.0771,
+      "step": 18570
+    },
+    {
+      "epoch": 0.594502927718939,
+      "grad_norm": 15.125,
+      "learning_rate": 8.537554732233075e-06,
+      "loss": 0.0844,
+      "step": 18580
+    },
+    {
+      "epoch": 0.5948228970018878,
+      "grad_norm": 20.0,
+      "learning_rate": 8.530818457393062e-06,
+      "loss": 0.118,
+      "step": 18590
+    },
+    {
+      "epoch": 0.5951428662848367,
+      "grad_norm": 19.875,
+      "learning_rate": 8.524082182553048e-06,
+      "loss": 0.1205,
+      "step": 18600
+    },
+    {
+      "epoch": 0.5954628355677855,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.517345907713034e-06,
+      "loss": 0.1637,
+      "step": 18610
+    },
+    {
+      "epoch": 0.5957828048507343,
+      "grad_norm": 13.8125,
+      "learning_rate": 8.510609632873022e-06,
+      "loss": 0.1474,
+      "step": 18620
+    },
+    {
+      "epoch": 0.5961027741336832,
+      "grad_norm": 12.875,
+      "learning_rate": 8.503873358033008e-06,
+      "loss": 0.0772,
+      "step": 18630
+    },
+    {
+      "epoch": 0.596422743416632,
+      "grad_norm": 7.0,
+      "learning_rate": 8.497137083192994e-06,
+      "loss": 0.1206,
+      "step": 18640
+    },
+    {
+      "epoch": 0.5967427126995808,
+      "grad_norm": 11.8125,
+      "learning_rate": 8.49040080835298e-06,
+      "loss": 0.0877,
+      "step": 18650
+    },
+    {
+      "epoch": 0.5970626819825297,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.483664533512968e-06,
+      "loss": 0.1119,
+      "step": 18660
+    },
+    {
+      "epoch": 0.5973826512654785,
+      "grad_norm": 30.625,
+      "learning_rate": 8.476928258672956e-06,
+      "loss": 0.088,
+      "step": 18670
+    },
+    {
+      "epoch": 0.5977026205484274,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.470191983832942e-06,
+      "loss": 0.0467,
+      "step": 18680
+    },
+    {
+      "epoch": 0.5980225898313762,
+      "grad_norm": 5.28125,
+      "learning_rate": 8.463455708992928e-06,
+      "loss": 0.0885,
+      "step": 18690
+    },
+    {
+      "epoch": 0.598342559114325,
+      "grad_norm": 1.421875,
+      "learning_rate": 8.456719434152914e-06,
+      "loss": 0.0802,
+      "step": 18700
+    },
+    {
+      "epoch": 0.5986625283972739,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.449983159312902e-06,
+      "loss": 0.0619,
+      "step": 18710
+    },
+    {
+      "epoch": 0.5989824976802227,
+      "grad_norm": 5.9375,
+      "learning_rate": 8.443246884472888e-06,
+      "loss": 0.0996,
+      "step": 18720
+    },
+    {
+      "epoch": 0.5993024669631716,
+      "grad_norm": 13.875,
+      "learning_rate": 8.436510609632874e-06,
+      "loss": 0.2121,
+      "step": 18730
+    },
+    {
+      "epoch": 0.5996224362461203,
+      "grad_norm": 10.5625,
+      "learning_rate": 8.429774334792861e-06,
+      "loss": 0.1205,
+      "step": 18740
+    },
+    {
+      "epoch": 0.5999424055290692,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.423038059952847e-06,
+      "loss": 0.1047,
+      "step": 18750
+    },
+    {
+      "epoch": 0.600262374812018,
+      "grad_norm": 13.1875,
+      "learning_rate": 8.416301785112833e-06,
+      "loss": 0.1272,
+      "step": 18760
+    },
+    {
+      "epoch": 0.6005823440949669,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.40956551027282e-06,
+      "loss": 0.084,
+      "step": 18770
+    },
+    {
+      "epoch": 0.6009023133779158,
+      "grad_norm": 19.875,
+      "learning_rate": 8.402829235432807e-06,
+      "loss": 0.1013,
+      "step": 18780
+    },
+    {
+      "epoch": 0.6012222826608645,
+      "grad_norm": 10.0625,
+      "learning_rate": 8.396092960592793e-06,
+      "loss": 0.0841,
+      "step": 18790
+    },
+    {
+      "epoch": 0.6015422519438134,
+      "grad_norm": 14.0,
+      "learning_rate": 8.38935668575278e-06,
+      "loss": 0.1079,
+      "step": 18800
+    },
+    {
+      "epoch": 0.6018622212267623,
+      "grad_norm": 8.875,
+      "learning_rate": 8.382620410912767e-06,
+      "loss": 0.0743,
+      "step": 18810
+    },
+    {
+      "epoch": 0.6021821905097111,
+      "grad_norm": 22.875,
+      "learning_rate": 8.375884136072753e-06,
+      "loss": 0.1192,
+      "step": 18820
+    },
+    {
+      "epoch": 0.6025021597926599,
+      "grad_norm": 23.75,
+      "learning_rate": 8.36914786123274e-06,
+      "loss": 0.0997,
+      "step": 18830
+    },
+    {
+      "epoch": 0.6028221290756087,
+      "grad_norm": 10.6875,
+      "learning_rate": 8.362411586392725e-06,
+      "loss": 0.0791,
+      "step": 18840
+    },
+    {
+      "epoch": 0.6031420983585576,
+      "grad_norm": 9.125,
+      "learning_rate": 8.355675311552713e-06,
+      "loss": 0.1361,
+      "step": 18850
+    },
+    {
+      "epoch": 0.6034620676415064,
+      "grad_norm": 11.9375,
+      "learning_rate": 8.348939036712699e-06,
+      "loss": 0.0868,
+      "step": 18860
+    },
+    {
+      "epoch": 0.6037820369244552,
+      "grad_norm": 9.0,
+      "learning_rate": 8.342202761872685e-06,
+      "loss": 0.1216,
+      "step": 18870
+    },
+    {
+      "epoch": 0.6041020062074041,
+      "grad_norm": 14.875,
+      "learning_rate": 8.335466487032671e-06,
+      "loss": 0.1298,
+      "step": 18880
+    },
+    {
+      "epoch": 0.6044219754903529,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.328730212192659e-06,
+      "loss": 0.1657,
+      "step": 18890
+    },
+    {
+      "epoch": 0.6047419447733018,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.321993937352645e-06,
+      "loss": 0.113,
+      "step": 18900
+    },
+    {
+      "epoch": 0.6050619140562506,
+      "grad_norm": 9.625,
+      "learning_rate": 8.315257662512631e-06,
+      "loss": 0.0905,
+      "step": 18910
+    },
+    {
+      "epoch": 0.6053818833391994,
+      "grad_norm": 18.375,
+      "learning_rate": 8.308521387672619e-06,
+      "loss": 0.1088,
+      "step": 18920
+    },
+    {
+      "epoch": 0.6057018526221483,
+      "grad_norm": 1.625,
+      "learning_rate": 8.301785112832605e-06,
+      "loss": 0.0995,
+      "step": 18930
+    },
+    {
+      "epoch": 0.6060218219050971,
+      "grad_norm": 9.875,
+      "learning_rate": 8.29504883799259e-06,
+      "loss": 0.1592,
+      "step": 18940
+    },
+    {
+      "epoch": 0.606341791188046,
+      "grad_norm": 13.25,
+      "learning_rate": 8.288312563152577e-06,
+      "loss": 0.1075,
+      "step": 18950
+    },
+    {
+      "epoch": 0.6066617604709947,
+      "grad_norm": 12.4375,
+      "learning_rate": 8.281576288312564e-06,
+      "loss": 0.1155,
+      "step": 18960
+    },
+    {
+      "epoch": 0.6069817297539436,
+      "grad_norm": 4.625,
+      "learning_rate": 8.27484001347255e-06,
+      "loss": 0.1474,
+      "step": 18970
+    },
+    {
+      "epoch": 0.6073016990368925,
+      "grad_norm": 9.5,
+      "learning_rate": 8.268103738632537e-06,
+      "loss": 0.0962,
+      "step": 18980
+    },
+    {
+      "epoch": 0.6076216683198413,
+      "grad_norm": 21.625,
+      "learning_rate": 8.261367463792523e-06,
+      "loss": 0.1287,
+      "step": 18990
+    },
+    {
+      "epoch": 0.6079416376027902,
+      "grad_norm": 8.625,
+      "learning_rate": 8.25463118895251e-06,
+      "loss": 0.0784,
+      "step": 19000
+    },
+    {
+      "epoch": 0.6082616068857389,
+      "grad_norm": 7.875,
+      "learning_rate": 8.247894914112496e-06,
+      "loss": 0.1569,
+      "step": 19010
+    },
+    {
+      "epoch": 0.6085815761686878,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.241158639272482e-06,
+      "loss": 0.0737,
+      "step": 19020
+    },
+    {
+      "epoch": 0.6089015454516367,
+      "grad_norm": 16.375,
+      "learning_rate": 8.23442236443247e-06,
+      "loss": 0.116,
+      "step": 19030
+    },
+    {
+      "epoch": 0.6092215147345855,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.227686089592456e-06,
+      "loss": 0.098,
+      "step": 19040
+    },
+    {
+      "epoch": 0.6095414840175343,
+      "grad_norm": 22.375,
+      "learning_rate": 8.220949814752442e-06,
+      "loss": 0.1514,
+      "step": 19050
+    },
+    {
+      "epoch": 0.6098614533004831,
+      "grad_norm": 24.5,
+      "learning_rate": 8.214213539912428e-06,
+      "loss": 0.1491,
+      "step": 19060
+    },
+    {
+      "epoch": 0.610181422583432,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.207477265072416e-06,
+      "loss": 0.1118,
+      "step": 19070
+    },
+    {
+      "epoch": 0.6105013918663809,
+      "grad_norm": 9.5,
+      "learning_rate": 8.200740990232402e-06,
+      "loss": 0.1323,
+      "step": 19080
+    },
+    {
+      "epoch": 0.6108213611493296,
+      "grad_norm": 19.5,
+      "learning_rate": 8.194004715392388e-06,
+      "loss": 0.1056,
+      "step": 19090
+    },
+    {
+      "epoch": 0.6111413304322785,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.187268440552376e-06,
+      "loss": 0.0918,
+      "step": 19100
+    },
+    {
+      "epoch": 0.6114612997152273,
+      "grad_norm": 22.25,
+      "learning_rate": 8.180532165712362e-06,
+      "loss": 0.1669,
+      "step": 19110
+    },
+    {
+      "epoch": 0.6117812689981762,
+      "grad_norm": 9.9375,
+      "learning_rate": 8.173795890872348e-06,
+      "loss": 0.0863,
+      "step": 19120
+    },
+    {
+      "epoch": 0.6121012382811251,
+      "grad_norm": 15.375,
+      "learning_rate": 8.167059616032334e-06,
+      "loss": 0.1289,
+      "step": 19130
+    },
+    {
+      "epoch": 0.6124212075640738,
+      "grad_norm": 6.84375,
+      "learning_rate": 8.160323341192322e-06,
+      "loss": 0.0885,
+      "step": 19140
+    },
+    {
+      "epoch": 0.6127411768470227,
+      "grad_norm": 5.46875,
+      "learning_rate": 8.153587066352308e-06,
+      "loss": 0.0948,
+      "step": 19150
+    },
+    {
+      "epoch": 0.6130611461299715,
+      "grad_norm": 9.125,
+      "learning_rate": 8.146850791512294e-06,
+      "loss": 0.1265,
+      "step": 19160
+    },
+    {
+      "epoch": 0.6133811154129204,
+      "grad_norm": 6.53125,
+      "learning_rate": 8.14011451667228e-06,
+      "loss": 0.1113,
+      "step": 19170
+    },
+    {
+      "epoch": 0.6137010846958691,
+      "grad_norm": 15.9375,
+      "learning_rate": 8.133378241832268e-06,
+      "loss": 0.1489,
+      "step": 19180
+    },
+    {
+      "epoch": 0.614021053978818,
+      "grad_norm": 4.5,
+      "learning_rate": 8.126641966992254e-06,
+      "loss": 0.0789,
+      "step": 19190
+    },
+    {
+      "epoch": 0.6143410232617669,
+      "grad_norm": 16.25,
+      "learning_rate": 8.11990569215224e-06,
+      "loss": 0.0689,
+      "step": 19200
+    },
+    {
+      "epoch": 0.6146609925447157,
+      "grad_norm": 10.625,
+      "learning_rate": 8.113169417312227e-06,
+      "loss": 0.067,
+      "step": 19210
+    },
+    {
+      "epoch": 0.6149809618276646,
+      "grad_norm": 14.25,
+      "learning_rate": 8.106433142472213e-06,
+      "loss": 0.1397,
+      "step": 19220
+    },
+    {
+      "epoch": 0.6153009311106133,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.0996968676322e-06,
+      "loss": 0.0818,
+      "step": 19230
+    },
+    {
+      "epoch": 0.6156209003935622,
+      "grad_norm": 8.875,
+      "learning_rate": 8.092960592792185e-06,
+      "loss": 0.1415,
+      "step": 19240
+    },
+    {
+      "epoch": 0.6159408696765111,
+      "grad_norm": 14.625,
+      "learning_rate": 8.086224317952173e-06,
+      "loss": 0.0961,
+      "step": 19250
+    },
+    {
+      "epoch": 0.6162608389594599,
+      "grad_norm": 6.71875,
+      "learning_rate": 8.07948804311216e-06,
+      "loss": 0.1307,
+      "step": 19260
+    },
+    {
+      "epoch": 0.6165808082424087,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.072751768272145e-06,
+      "loss": 0.1551,
+      "step": 19270
+    },
+    {
+      "epoch": 0.6169007775253575,
+      "grad_norm": 5.53125,
+      "learning_rate": 8.066015493432131e-06,
+      "loss": 0.0824,
+      "step": 19280
+    },
+    {
+      "epoch": 0.6172207468083064,
+      "grad_norm": 16.875,
+      "learning_rate": 8.059279218592119e-06,
+      "loss": 0.1849,
+      "step": 19290
+    },
+    {
+      "epoch": 0.6175407160912553,
+      "grad_norm": 13.5,
+      "learning_rate": 8.052542943752105e-06,
+      "loss": 0.0824,
+      "step": 19300
+    },
+    {
+      "epoch": 0.617860685374204,
+      "grad_norm": 9.9375,
+      "learning_rate": 8.045806668912091e-06,
+      "loss": 0.1108,
+      "step": 19310
+    },
+    {
+      "epoch": 0.6181806546571529,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.039070394072079e-06,
+      "loss": 0.0684,
+      "step": 19320
+    },
+    {
+      "epoch": 0.6185006239401017,
+      "grad_norm": 31.0,
+      "learning_rate": 8.032334119232067e-06,
+      "loss": 0.0614,
+      "step": 19330
+    },
+    {
+      "epoch": 0.6188205932230506,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.025597844392053e-06,
+      "loss": 0.0985,
+      "step": 19340
+    },
+    {
+      "epoch": 0.6191405625059995,
+      "grad_norm": 17.0,
+      "learning_rate": 8.018861569552039e-06,
+      "loss": 0.0931,
+      "step": 19350
+    },
+    {
+      "epoch": 0.6194605317889482,
+      "grad_norm": 10.625,
+      "learning_rate": 8.012125294712025e-06,
+      "loss": 0.1456,
+      "step": 19360
+    },
+    {
+      "epoch": 0.6197805010718971,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.005389019872012e-06,
+      "loss": 0.1399,
+      "step": 19370
+    },
+    {
+      "epoch": 0.620100470354846,
+      "grad_norm": 15.6875,
+      "learning_rate": 7.998652745031999e-06,
+      "loss": 0.1441,
+      "step": 19380
+    },
+    {
+      "epoch": 0.6204204396377948,
+      "grad_norm": 5.09375,
+      "learning_rate": 7.991916470191985e-06,
+      "loss": 0.0776,
+      "step": 19390
+    },
+    {
+      "epoch": 0.6207404089207436,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.98518019535197e-06,
+      "loss": 0.0576,
+      "step": 19400
+    },
+    {
+      "epoch": 0.6210603782036924,
+      "grad_norm": 5.46875,
+      "learning_rate": 7.978443920511958e-06,
+      "loss": 0.1035,
+      "step": 19410
+    },
+    {
+      "epoch": 0.6213803474866413,
+      "grad_norm": 12.375,
+      "learning_rate": 7.971707645671944e-06,
+      "loss": 0.1826,
+      "step": 19420
+    },
+    {
+      "epoch": 0.6217003167695901,
+      "grad_norm": 8.0,
+      "learning_rate": 7.96497137083193e-06,
+      "loss": 0.1464,
+      "step": 19430
+    },
+    {
+      "epoch": 0.622020286052539,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.958235095991918e-06,
+      "loss": 0.1282,
+      "step": 19440
+    },
+    {
+      "epoch": 0.6223402553354878,
+      "grad_norm": 15.6875,
+      "learning_rate": 7.951498821151904e-06,
+      "loss": 0.0828,
+      "step": 19450
+    },
+    {
+      "epoch": 0.6226602246184366,
+      "grad_norm": 26.25,
+      "learning_rate": 7.94476254631189e-06,
+      "loss": 0.171,
+      "step": 19460
+    },
+    {
+      "epoch": 0.6229801939013855,
+      "grad_norm": 20.75,
+      "learning_rate": 7.938026271471876e-06,
+      "loss": 0.0819,
+      "step": 19470
+    },
+    {
+      "epoch": 0.6233001631843343,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.931289996631864e-06,
+      "loss": 0.1237,
+      "step": 19480
+    },
+    {
+      "epoch": 0.6236201324672831,
+      "grad_norm": 14.8125,
+      "learning_rate": 7.92455372179185e-06,
+      "loss": 0.1599,
+      "step": 19490
+    },
+    {
+      "epoch": 0.623940101750232,
+      "grad_norm": 4.78125,
+      "learning_rate": 7.917817446951836e-06,
+      "loss": 0.0942,
+      "step": 19500
+    },
+    {
+      "epoch": 0.6242600710331808,
+      "grad_norm": 4.25,
+      "learning_rate": 7.911081172111824e-06,
+      "loss": 0.1062,
+      "step": 19510
+    },
+    {
+      "epoch": 0.6245800403161297,
+      "grad_norm": 10.3125,
+      "learning_rate": 7.90434489727181e-06,
+      "loss": 0.1284,
+      "step": 19520
+    },
+    {
+      "epoch": 0.6249000095990785,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.897608622431796e-06,
+      "loss": 0.0548,
+      "step": 19530
+    },
+    {
+      "epoch": 0.6252199788820273,
+      "grad_norm": 10.125,
+      "learning_rate": 7.890872347591782e-06,
+      "loss": 0.1392,
+      "step": 19540
+    },
+    {
+      "epoch": 0.6255399481649762,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.88413607275177e-06,
+      "loss": 0.0849,
+      "step": 19550
+    },
+    {
+      "epoch": 0.625859917447925,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.877399797911756e-06,
+      "loss": 0.0724,
+      "step": 19560
+    },
+    {
+      "epoch": 0.6261798867308739,
+      "grad_norm": 15.1875,
+      "learning_rate": 7.870663523071742e-06,
+      "loss": 0.1187,
+      "step": 19570
+    },
+    {
+      "epoch": 0.6264998560138226,
+      "grad_norm": 15.0,
+      "learning_rate": 7.863927248231728e-06,
+      "loss": 0.1365,
+      "step": 19580
+    },
+    {
+      "epoch": 0.6268198252967715,
+      "grad_norm": 12.6875,
+      "learning_rate": 7.857190973391716e-06,
+      "loss": 0.1425,
+      "step": 19590
+    },
+    {
+      "epoch": 0.6271397945797204,
+      "grad_norm": 17.875,
+      "learning_rate": 7.850454698551702e-06,
+      "loss": 0.1039,
+      "step": 19600
+    },
+    {
+      "epoch": 0.6274597638626692,
+      "grad_norm": 4.4375,
+      "learning_rate": 7.843718423711688e-06,
+      "loss": 0.0568,
+      "step": 19610
+    },
+    {
+      "epoch": 0.627779733145618,
+      "grad_norm": 19.625,
+      "learning_rate": 7.836982148871675e-06,
+      "loss": 0.0746,
+      "step": 19620
+    },
+    {
+      "epoch": 0.6280997024285668,
+      "grad_norm": 14.1875,
+      "learning_rate": 7.830245874031661e-06,
+      "loss": 0.1533,
+      "step": 19630
+    },
+    {
+      "epoch": 0.6284196717115157,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.823509599191647e-06,
+      "loss": 0.1254,
+      "step": 19640
+    },
+    {
+      "epoch": 0.6287396409944646,
+      "grad_norm": 16.125,
+      "learning_rate": 7.816773324351634e-06,
+      "loss": 0.0725,
+      "step": 19650
+    },
+    {
+      "epoch": 0.6290596102774134,
+      "grad_norm": 19.375,
+      "learning_rate": 7.810037049511621e-06,
+      "loss": 0.1433,
+      "step": 19660
+    },
+    {
+      "epoch": 0.6293795795603622,
+      "grad_norm": 31.125,
+      "learning_rate": 7.803300774671607e-06,
+      "loss": 0.1035,
+      "step": 19670
+    },
+    {
+      "epoch": 0.629699548843311,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.796564499831593e-06,
+      "loss": 0.1315,
+      "step": 19680
+    },
+    {
+      "epoch": 0.6300195181262599,
+      "grad_norm": 10.625,
+      "learning_rate": 7.78982822499158e-06,
+      "loss": 0.1518,
+      "step": 19690
+    },
+    {
+      "epoch": 0.6303394874092088,
+      "grad_norm": 4.65625,
+      "learning_rate": 7.783091950151567e-06,
+      "loss": 0.0929,
+      "step": 19700
+    },
+    {
+      "epoch": 0.6306594566921575,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.776355675311553e-06,
+      "loss": 0.1516,
+      "step": 19710
+    },
+    {
+      "epoch": 0.6309794259751064,
+      "grad_norm": 6.3125,
+      "learning_rate": 7.76961940047154e-06,
+      "loss": 0.092,
+      "step": 19720
+    },
+    {
+      "epoch": 0.6312993952580552,
+      "grad_norm": 7.34375,
+      "learning_rate": 7.762883125631527e-06,
+      "loss": 0.059,
+      "step": 19730
+    },
+    {
+      "epoch": 0.6316193645410041,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.756146850791513e-06,
+      "loss": 0.1217,
+      "step": 19740
+    },
+    {
+      "epoch": 0.631939333823953,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.749410575951499e-06,
+      "loss": 0.0466,
+      "step": 19750
+    },
+    {
+      "epoch": 0.6322593031069017,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.742674301111485e-06,
+      "loss": 0.1331,
+      "step": 19760
+    },
+    {
+      "epoch": 0.6325792723898506,
+      "grad_norm": 27.0,
+      "learning_rate": 7.735938026271473e-06,
+      "loss": 0.1202,
+      "step": 19770
+    },
+    {
+      "epoch": 0.6328992416727994,
+      "grad_norm": 3.9375,
+      "learning_rate": 7.729201751431459e-06,
+      "loss": 0.0823,
+      "step": 19780
+    },
+    {
+      "epoch": 0.6332192109557483,
+      "grad_norm": 17.625,
+      "learning_rate": 7.722465476591445e-06,
+      "loss": 0.1079,
+      "step": 19790
+    },
+    {
+      "epoch": 0.633539180238697,
+      "grad_norm": 8.9375,
+      "learning_rate": 7.715729201751433e-06,
+      "loss": 0.1229,
+      "step": 19800
+    },
+    {
+      "epoch": 0.6338591495216459,
+      "grad_norm": 14.0,
+      "learning_rate": 7.708992926911419e-06,
+      "loss": 0.1233,
+      "step": 19810
+    },
+    {
+      "epoch": 0.6341791188045948,
+      "grad_norm": 7.09375,
+      "learning_rate": 7.702256652071405e-06,
+      "loss": 0.082,
+      "step": 19820
+    },
+    {
+      "epoch": 0.6344990880875436,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.69552037723139e-06,
+      "loss": 0.0586,
+      "step": 19830
+    },
+    {
+      "epoch": 0.6348190573704924,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.688784102391378e-06,
+      "loss": 0.0743,
+      "step": 19840
+    },
+    {
+      "epoch": 0.6351390266534412,
+      "grad_norm": 7.15625,
+      "learning_rate": 7.682047827551365e-06,
+      "loss": 0.1212,
+      "step": 19850
+    },
+    {
+      "epoch": 0.6354589959363901,
+      "grad_norm": 17.875,
+      "learning_rate": 7.67531155271135e-06,
+      "loss": 0.0919,
+      "step": 19860
+    },
+    {
+      "epoch": 0.635778965219339,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.668575277871337e-06,
+      "loss": 0.1626,
+      "step": 19870
+    },
+    {
+      "epoch": 0.6360989345022878,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.661839003031324e-06,
+      "loss": 0.1455,
+      "step": 19880
+    },
+    {
+      "epoch": 0.6364189037852366,
+      "grad_norm": 20.125,
+      "learning_rate": 7.65510272819131e-06,
+      "loss": 0.1018,
+      "step": 19890
+    },
+    {
+      "epoch": 0.6367388730681854,
+      "grad_norm": 22.0,
+      "learning_rate": 7.648366453351296e-06,
+      "loss": 0.1173,
+      "step": 19900
+    },
+    {
+      "epoch": 0.6370588423511343,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.641630178511284e-06,
+      "loss": 0.1276,
+      "step": 19910
+    },
+    {
+      "epoch": 0.6373788116340832,
+      "grad_norm": 8.375,
+      "learning_rate": 7.63489390367127e-06,
+      "loss": 0.1154,
+      "step": 19920
+    },
+    {
+      "epoch": 0.6376987809170319,
+      "grad_norm": 10.25,
+      "learning_rate": 7.628157628831256e-06,
+      "loss": 0.0659,
+      "step": 19930
+    },
+    {
+      "epoch": 0.6380187501999808,
+      "grad_norm": 9.75,
+      "learning_rate": 7.621421353991243e-06,
+      "loss": 0.1202,
+      "step": 19940
+    },
+    {
+      "epoch": 0.6383387194829296,
+      "grad_norm": 14.9375,
+      "learning_rate": 7.614685079151229e-06,
+      "loss": 0.1326,
+      "step": 19950
+    },
+    {
+      "epoch": 0.6386586887658785,
+      "grad_norm": 16.125,
+      "learning_rate": 7.607948804311217e-06,
+      "loss": 0.102,
+      "step": 19960
+    },
+    {
+      "epoch": 0.6389786580488274,
+      "grad_norm": 20.75,
+      "learning_rate": 7.601212529471204e-06,
+      "loss": 0.1642,
+      "step": 19970
+    },
+    {
+      "epoch": 0.6392986273317761,
+      "grad_norm": 10.0,
+      "learning_rate": 7.59447625463119e-06,
+      "loss": 0.086,
+      "step": 19980
+    },
+    {
+      "epoch": 0.639618596614725,
+      "grad_norm": 10.0625,
+      "learning_rate": 7.587739979791177e-06,
+      "loss": 0.178,
+      "step": 19990
+    },
+    {
+      "epoch": 0.6399385658976738,
+      "grad_norm": 15.3125,
+      "learning_rate": 7.581003704951163e-06,
+      "loss": 0.136,
+      "step": 20000
+    },
+    {
+      "epoch": 0.6402585351806227,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.57426743011115e-06,
+      "loss": 0.0693,
+      "step": 20010
+    },
+    {
+      "epoch": 0.6405785044635715,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.5675311552711365e-06,
+      "loss": 0.0695,
+      "step": 20020
+    },
+    {
+      "epoch": 0.6408984737465203,
+      "grad_norm": 20.375,
+      "learning_rate": 7.5607948804311226e-06,
+      "loss": 0.0644,
+      "step": 20030
+    },
+    {
+      "epoch": 0.6412184430294692,
+      "grad_norm": 4.75,
+      "learning_rate": 7.5540586055911095e-06,
+      "loss": 0.1316,
+      "step": 20040
+    },
+    {
+      "epoch": 0.641538412312418,
+      "grad_norm": 6.1875,
+      "learning_rate": 7.5473223307510955e-06,
+      "loss": 0.1294,
+      "step": 20050
+    },
+    {
+      "epoch": 0.6418583815953669,
+      "grad_norm": 6.5625,
+      "learning_rate": 7.540586055911082e-06,
+      "loss": 0.049,
+      "step": 20060
+    },
+    {
+      "epoch": 0.6421783508783157,
+      "grad_norm": 9.0625,
+      "learning_rate": 7.5338497810710684e-06,
+      "loss": 0.0606,
+      "step": 20070
+    },
+    {
+      "epoch": 0.6424983201612645,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.527113506231055e-06,
+      "loss": 0.0997,
+      "step": 20080
+    },
+    {
+      "epoch": 0.6428182894442134,
+      "grad_norm": 31.375,
+      "learning_rate": 7.520377231391041e-06,
+      "loss": 0.1517,
+      "step": 20090
+    },
+    {
+      "epoch": 0.6431382587271622,
+      "grad_norm": 6.1875,
+      "learning_rate": 7.513640956551028e-06,
+      "loss": 0.1015,
+      "step": 20100
+    },
+    {
+      "epoch": 0.643458228010111,
+      "grad_norm": 17.75,
+      "learning_rate": 7.506904681711015e-06,
+      "loss": 0.1589,
+      "step": 20110
+    },
+    {
+      "epoch": 0.6437781972930599,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.500168406871001e-06,
+      "loss": 0.1131,
+      "step": 20120
+    },
+    {
+      "epoch": 0.6440981665760087,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.493432132030988e-06,
+      "loss": 0.1569,
+      "step": 20130
+    },
+    {
+      "epoch": 0.6444181358589576,
+      "grad_norm": 9.1875,
+      "learning_rate": 7.486695857190974e-06,
+      "loss": 0.0943,
+      "step": 20140
+    },
+    {
+      "epoch": 0.6447381051419063,
+      "grad_norm": 6.75,
+      "learning_rate": 7.479959582350961e-06,
+      "loss": 0.1397,
+      "step": 20150
+    },
+    {
+      "epoch": 0.6450580744248552,
+      "grad_norm": 12.875,
+      "learning_rate": 7.473223307510947e-06,
+      "loss": 0.1906,
+      "step": 20160
+    },
+    {
+      "epoch": 0.645378043707804,
+      "grad_norm": 8.375,
+      "learning_rate": 7.466487032670934e-06,
+      "loss": 0.1165,
+      "step": 20170
+    },
+    {
+      "epoch": 0.6456980129907529,
+      "grad_norm": 12.125,
+      "learning_rate": 7.45975075783092e-06,
+      "loss": 0.033,
+      "step": 20180
+    },
+    {
+      "epoch": 0.6460179822737018,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.453014482990907e-06,
+      "loss": 0.1061,
+      "step": 20190
+    },
+    {
+      "epoch": 0.6463379515566505,
+      "grad_norm": 11.0625,
+      "learning_rate": 7.446278208150893e-06,
+      "loss": 0.1324,
+      "step": 20200
+    },
+    {
+      "epoch": 0.6466579208395994,
+      "grad_norm": 5.75,
+      "learning_rate": 7.43954193331088e-06,
+      "loss": 0.1213,
+      "step": 20210
+    },
+    {
+      "epoch": 0.6469778901225482,
+      "grad_norm": 28.125,
+      "learning_rate": 7.432805658470867e-06,
+      "loss": 0.0962,
+      "step": 20220
+    },
+    {
+      "epoch": 0.6472978594054971,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.426069383630853e-06,
+      "loss": 0.1376,
+      "step": 20230
+    },
+    {
+      "epoch": 0.6476178286884459,
+      "grad_norm": 11.4375,
+      "learning_rate": 7.41933310879084e-06,
+      "loss": 0.1109,
+      "step": 20240
+    },
+    {
+      "epoch": 0.6479377979713947,
+      "grad_norm": 4.25,
+      "learning_rate": 7.412596833950826e-06,
+      "loss": 0.0387,
+      "step": 20250
+    },
+    {
+      "epoch": 0.6482577672543436,
+      "grad_norm": 10.125,
+      "learning_rate": 7.4058605591108125e-06,
+      "loss": 0.0841,
+      "step": 20260
+    },
+    {
+      "epoch": 0.6485777365372924,
+      "grad_norm": 9.4375,
+      "learning_rate": 7.3991242842707986e-06,
+      "loss": 0.0833,
+      "step": 20270
+    },
+    {
+      "epoch": 0.6488977058202413,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.3923880094307855e-06,
+      "loss": 0.0976,
+      "step": 20280
+    },
+    {
+      "epoch": 0.6492176751031901,
+      "grad_norm": 13.5625,
+      "learning_rate": 7.3856517345907715e-06,
+      "loss": 0.1339,
+      "step": 20290
+    },
+    {
+      "epoch": 0.6495376443861389,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.378915459750758e-06,
+      "loss": 0.0921,
+      "step": 20300
+    },
+    {
+      "epoch": 0.6498576136690878,
+      "grad_norm": 32.5,
+      "learning_rate": 7.372179184910745e-06,
+      "loss": 0.1333,
+      "step": 20310
+    },
+    {
+      "epoch": 0.6501775829520366,
+      "grad_norm": 21.625,
+      "learning_rate": 7.365442910070731e-06,
+      "loss": 0.1174,
+      "step": 20320
+    },
+    {
+      "epoch": 0.6504975522349854,
+      "grad_norm": 26.0,
+      "learning_rate": 7.358706635230718e-06,
+      "loss": 0.1552,
+      "step": 20330
+    },
+    {
+      "epoch": 0.6508175215179343,
+      "grad_norm": 6.84375,
+      "learning_rate": 7.351970360390704e-06,
+      "loss": 0.1214,
+      "step": 20340
+    },
+    {
+      "epoch": 0.6511374908008831,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.345234085550691e-06,
+      "loss": 0.1544,
+      "step": 20350
+    },
+    {
+      "epoch": 0.651457460083832,
+      "grad_norm": 12.6875,
+      "learning_rate": 7.338497810710677e-06,
+      "loss": 0.1477,
+      "step": 20360
+    },
+    {
+      "epoch": 0.6517774293667807,
+      "grad_norm": 3.75,
+      "learning_rate": 7.331761535870664e-06,
+      "loss": 0.046,
+      "step": 20370
+    },
+    {
+      "epoch": 0.6520973986497296,
+      "grad_norm": 16.875,
+      "learning_rate": 7.32502526103065e-06,
+      "loss": 0.1528,
+      "step": 20380
+    },
+    {
+      "epoch": 0.6524173679326785,
+      "grad_norm": 12.375,
+      "learning_rate": 7.318288986190637e-06,
+      "loss": 0.0999,
+      "step": 20390
+    },
+    {
+      "epoch": 0.6527373372156273,
+      "grad_norm": 8.75,
+      "learning_rate": 7.311552711350624e-06,
+      "loss": 0.1011,
+      "step": 20400
+    },
+    {
+      "epoch": 0.6530573064985762,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.30481643651061e-06,
+      "loss": 0.166,
+      "step": 20410
+    },
+    {
+      "epoch": 0.6533772757815249,
+      "grad_norm": 11.5,
+      "learning_rate": 7.298080161670597e-06,
+      "loss": 0.1344,
+      "step": 20420
+    },
+    {
+      "epoch": 0.6536972450644738,
+      "grad_norm": 13.5625,
+      "learning_rate": 7.291343886830583e-06,
+      "loss": 0.1874,
+      "step": 20430
+    },
+    {
+      "epoch": 0.6540172143474227,
+      "grad_norm": 9.8125,
+      "learning_rate": 7.28460761199057e-06,
+      "loss": 0.0561,
+      "step": 20440
+    },
+    {
+      "epoch": 0.6543371836303715,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.277871337150556e-06,
+      "loss": 0.1654,
+      "step": 20450
+    },
+    {
+      "epoch": 0.6546571529133203,
+      "grad_norm": 7.34375,
+      "learning_rate": 7.271135062310543e-06,
+      "loss": 0.0674,
+      "step": 20460
+    },
+    {
+      "epoch": 0.6549771221962691,
+      "grad_norm": 10.0625,
+      "learning_rate": 7.264398787470529e-06,
+      "loss": 0.1458,
+      "step": 20470
+    },
+    {
+      "epoch": 0.655297091479218,
+      "grad_norm": 9.5625,
+      "learning_rate": 7.257662512630516e-06,
+      "loss": 0.1359,
+      "step": 20480
+    },
+    {
+      "epoch": 0.6556170607621669,
+      "grad_norm": 12.5,
+      "learning_rate": 7.2509262377905025e-06,
+      "loss": 0.1537,
+      "step": 20490
+    },
+    {
+      "epoch": 0.6559370300451157,
+      "grad_norm": 21.75,
+      "learning_rate": 7.2441899629504885e-06,
+      "loss": 0.1127,
+      "step": 20500
+    },
+    {
+      "epoch": 0.6562569993280645,
+      "grad_norm": 13.875,
+      "learning_rate": 7.237453688110475e-06,
+      "loss": 0.1263,
+      "step": 20510
+    },
+    {
+      "epoch": 0.6565769686110133,
+      "grad_norm": 20.375,
+      "learning_rate": 7.2307174132704615e-06,
+      "loss": 0.0912,
+      "step": 20520
+    },
+    {
+      "epoch": 0.6568969378939622,
+      "grad_norm": 10.25,
+      "learning_rate": 7.223981138430448e-06,
+      "loss": 0.0811,
+      "step": 20530
+    },
+    {
+      "epoch": 0.6572169071769111,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.217244863590434e-06,
+      "loss": 0.209,
+      "step": 20540
+    },
+    {
+      "epoch": 0.6575368764598598,
+      "grad_norm": 7.96875,
+      "learning_rate": 7.210508588750421e-06,
+      "loss": 0.1284,
+      "step": 20550
+    },
+    {
+      "epoch": 0.6578568457428087,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.203772313910407e-06,
+      "loss": 0.0822,
+      "step": 20560
+    },
+    {
+      "epoch": 0.6581768150257575,
+      "grad_norm": 11.3125,
+      "learning_rate": 7.197036039070394e-06,
+      "loss": 0.1542,
+      "step": 20570
+    },
+    {
+      "epoch": 0.6584967843087064,
+      "grad_norm": 5.46875,
+      "learning_rate": 7.19029976423038e-06,
+      "loss": 0.1542,
+      "step": 20580
+    },
+    {
+      "epoch": 0.6588167535916551,
+      "grad_norm": 11.4375,
+      "learning_rate": 7.183563489390367e-06,
+      "loss": 0.0781,
+      "step": 20590
+    },
+    {
+      "epoch": 0.659136722874604,
+      "grad_norm": 5.1875,
+      "learning_rate": 7.176827214550354e-06,
+      "loss": 0.0887,
+      "step": 20600
+    },
+    {
+      "epoch": 0.6594566921575529,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.170090939710341e-06,
+      "loss": 0.0882,
+      "step": 20610
+    },
+    {
+      "epoch": 0.6597766614405017,
+      "grad_norm": 10.5,
+      "learning_rate": 7.163354664870328e-06,
+      "loss": 0.0617,
+      "step": 20620
+    },
+    {
+      "epoch": 0.6600966307234506,
+      "grad_norm": 10.0625,
+      "learning_rate": 7.156618390030315e-06,
+      "loss": 0.1362,
+      "step": 20630
+    },
+    {
+      "epoch": 0.6604166000063993,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.149882115190301e-06,
+      "loss": 0.0496,
+      "step": 20640
+    },
+    {
+      "epoch": 0.6607365692893482,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.143145840350288e-06,
+      "loss": 0.0986,
+      "step": 20650
+    },
+    {
+      "epoch": 0.6610565385722971,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.136409565510274e-06,
+      "loss": 0.0777,
+      "step": 20660
+    },
+    {
+      "epoch": 0.6613765078552459,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.1296732906702606e-06,
+      "loss": 0.062,
+      "step": 20670
+    },
+    {
+      "epoch": 0.6616964771381947,
+      "grad_norm": 10.75,
+      "learning_rate": 7.122937015830247e-06,
+      "loss": 0.1215,
+      "step": 20680
+    },
+    {
+      "epoch": 0.6620164464211435,
+      "grad_norm": 16.75,
+      "learning_rate": 7.1162007409902335e-06,
+      "loss": 0.1362,
+      "step": 20690
+    },
+    {
+      "epoch": 0.6623364157040924,
+      "grad_norm": 15.875,
+      "learning_rate": 7.1094644661502195e-06,
+      "loss": 0.1404,
+      "step": 20700
+    },
+    {
+      "epoch": 0.6626563849870413,
+      "grad_norm": 6.65625,
+      "learning_rate": 7.102728191310206e-06,
+      "loss": 0.1633,
+      "step": 20710
+    },
+    {
+      "epoch": 0.6629763542699901,
+      "grad_norm": 15.0,
+      "learning_rate": 7.095991916470193e-06,
+      "loss": 0.0841,
+      "step": 20720
+    },
+    {
+      "epoch": 0.6632963235529389,
+      "grad_norm": 13.6875,
+      "learning_rate": 7.089255641630179e-06,
+      "loss": 0.1324,
+      "step": 20730
+    },
+    {
+      "epoch": 0.6636162928358877,
+      "grad_norm": 7.5625,
+      "learning_rate": 7.082519366790166e-06,
+      "loss": 0.1023,
+      "step": 20740
+    },
+    {
+      "epoch": 0.6639362621188366,
+      "grad_norm": 11.875,
+      "learning_rate": 7.075783091950152e-06,
+      "loss": 0.0865,
+      "step": 20750
+    },
+    {
+      "epoch": 0.6642562314017855,
+      "grad_norm": 19.125,
+      "learning_rate": 7.069046817110139e-06,
+      "loss": 0.1169,
+      "step": 20760
+    },
+    {
+      "epoch": 0.6645762006847342,
+      "grad_norm": 11.5,
+      "learning_rate": 7.062310542270125e-06,
+      "loss": 0.1276,
+      "step": 20770
+    },
+    {
+      "epoch": 0.6648961699676831,
+      "grad_norm": 23.5,
+      "learning_rate": 7.055574267430112e-06,
+      "loss": 0.0881,
+      "step": 20780
+    },
+    {
+      "epoch": 0.6652161392506319,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.048837992590098e-06,
+      "loss": 0.1109,
+      "step": 20790
+    },
+    {
+      "epoch": 0.6655361085335808,
+      "grad_norm": 26.5,
+      "learning_rate": 7.042101717750085e-06,
+      "loss": 0.1002,
+      "step": 20800
+    },
+    {
+      "epoch": 0.6658560778165297,
+      "grad_norm": 12.75,
+      "learning_rate": 7.035365442910072e-06,
+      "loss": 0.1059,
+      "step": 20810
+    },
+    {
+      "epoch": 0.6661760470994784,
+      "grad_norm": 4.625,
+      "learning_rate": 7.028629168070058e-06,
+      "loss": 0.0575,
+      "step": 20820
+    },
+    {
+      "epoch": 0.6664960163824273,
+      "grad_norm": 13.6875,
+      "learning_rate": 7.021892893230045e-06,
+      "loss": 0.0989,
+      "step": 20830
+    },
+    {
+      "epoch": 0.6668159856653761,
+      "grad_norm": 9.0625,
+      "learning_rate": 7.015156618390031e-06,
+      "loss": 0.0852,
+      "step": 20840
+    },
+    {
+      "epoch": 0.667135954948325,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.008420343550018e-06,
+      "loss": 0.117,
+      "step": 20850
+    },
+    {
+      "epoch": 0.6674559242312738,
+      "grad_norm": 5.28125,
+      "learning_rate": 7.001684068710004e-06,
+      "loss": 0.0628,
+      "step": 20860
+    },
+    {
+      "epoch": 0.6677758935142226,
+      "grad_norm": 3.625,
+      "learning_rate": 6.994947793869991e-06,
+      "loss": 0.1088,
+      "step": 20870
+    },
+    {
+      "epoch": 0.6680958627971715,
+      "grad_norm": 5.03125,
+      "learning_rate": 6.988211519029977e-06,
+      "loss": 0.0809,
+      "step": 20880
+    },
+    {
+      "epoch": 0.6684158320801203,
+      "grad_norm": 0.3203125,
+      "learning_rate": 6.981475244189964e-06,
+      "loss": 0.1406,
+      "step": 20890
+    },
+    {
+      "epoch": 0.6687358013630691,
+      "grad_norm": 4.53125,
+      "learning_rate": 6.9747389693499505e-06,
+      "loss": 0.095,
+      "step": 20900
+    },
+    {
+      "epoch": 0.669055770646018,
+      "grad_norm": 11.5,
+      "learning_rate": 6.9680026945099366e-06,
+      "loss": 0.162,
+      "step": 20910
+    },
+    {
+      "epoch": 0.6693757399289668,
+      "grad_norm": 11.75,
+      "learning_rate": 6.9612664196699234e-06,
+      "loss": 0.1381,
+      "step": 20920
+    },
+    {
+      "epoch": 0.6696957092119157,
+      "grad_norm": 6.46875,
+      "learning_rate": 6.9545301448299095e-06,
+      "loss": 0.1148,
+      "step": 20930
+    },
+    {
+      "epoch": 0.6700156784948645,
+      "grad_norm": 6.3125,
+      "learning_rate": 6.947793869989896e-06,
+      "loss": 0.1229,
+      "step": 20940
+    },
+    {
+      "epoch": 0.6703356477778133,
+      "grad_norm": 9.9375,
+      "learning_rate": 6.941057595149882e-06,
+      "loss": 0.0953,
+      "step": 20950
+    },
+    {
+      "epoch": 0.6706556170607622,
+      "grad_norm": 6.75,
+      "learning_rate": 6.934321320309869e-06,
+      "loss": 0.1972,
+      "step": 20960
+    },
+    {
+      "epoch": 0.670975586343711,
+      "grad_norm": 24.5,
+      "learning_rate": 6.927585045469855e-06,
+      "loss": 0.109,
+      "step": 20970
+    },
+    {
+      "epoch": 0.6712955556266599,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.920848770629842e-06,
+      "loss": 0.0661,
+      "step": 20980
+    },
+    {
+      "epoch": 0.6716155249096086,
+      "grad_norm": 10.0,
+      "learning_rate": 6.914112495789828e-06,
+      "loss": 0.106,
+      "step": 20990
+    },
+    {
+      "epoch": 0.6719354941925575,
+      "grad_norm": 12.25,
+      "learning_rate": 6.907376220949815e-06,
+      "loss": 0.1246,
+      "step": 21000
+    },
+    {
+      "epoch": 0.6722554634755064,
+      "grad_norm": 15.4375,
+      "learning_rate": 6.900639946109802e-06,
+      "loss": 0.0964,
+      "step": 21010
+    },
+    {
+      "epoch": 0.6725754327584552,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.893903671269788e-06,
+      "loss": 0.0711,
+      "step": 21020
+    },
+    {
+      "epoch": 0.6728954020414041,
+      "grad_norm": 16.375,
+      "learning_rate": 6.887167396429775e-06,
+      "loss": 0.1346,
+      "step": 21030
+    },
+    {
+      "epoch": 0.6732153713243528,
+      "grad_norm": 1.359375,
+      "learning_rate": 6.880431121589761e-06,
+      "loss": 0.1528,
+      "step": 21040
+    },
+    {
+      "epoch": 0.6735353406073017,
+      "grad_norm": 10.0625,
+      "learning_rate": 6.873694846749748e-06,
+      "loss": 0.1028,
+      "step": 21050
+    },
+    {
+      "epoch": 0.6738553098902506,
+      "grad_norm": 6.75,
+      "learning_rate": 6.866958571909734e-06,
+      "loss": 0.1368,
+      "step": 21060
+    },
+    {
+      "epoch": 0.6741752791731994,
+      "grad_norm": 2.5,
+      "learning_rate": 6.860222297069721e-06,
+      "loss": 0.1235,
+      "step": 21070
+    },
+    {
+      "epoch": 0.6744952484561482,
+      "grad_norm": 8.875,
+      "learning_rate": 6.853486022229707e-06,
+      "loss": 0.0645,
+      "step": 21080
+    },
+    {
+      "epoch": 0.674815217739097,
+      "grad_norm": 19.0,
+      "learning_rate": 6.846749747389694e-06,
+      "loss": 0.1734,
+      "step": 21090
+    },
+    {
+      "epoch": 0.6751351870220459,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.840013472549681e-06,
+      "loss": 0.1741,
+      "step": 21100
+    },
+    {
+      "epoch": 0.6754551563049948,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.833277197709667e-06,
+      "loss": 0.1117,
+      "step": 21110
+    },
+    {
+      "epoch": 0.6757751255879435,
+      "grad_norm": 9.0625,
+      "learning_rate": 6.826540922869654e-06,
+      "loss": 0.0723,
+      "step": 21120
+    },
+    {
+      "epoch": 0.6760950948708924,
+      "grad_norm": 9.25,
+      "learning_rate": 6.81980464802964e-06,
+      "loss": 0.1361,
+      "step": 21130
+    },
+    {
+      "epoch": 0.6764150641538412,
+      "grad_norm": 1.9921875,
+      "learning_rate": 6.8130683731896265e-06,
+      "loss": 0.1702,
+      "step": 21140
+    },
+    {
+      "epoch": 0.6767350334367901,
+      "grad_norm": 11.0,
+      "learning_rate": 6.8063320983496126e-06,
+      "loss": 0.1383,
+      "step": 21150
+    },
+    {
+      "epoch": 0.677055002719739,
+      "grad_norm": 9.1875,
+      "learning_rate": 6.7995958235095994e-06,
+      "loss": 0.164,
+      "step": 21160
+    },
+    {
+      "epoch": 0.6773749720026877,
+      "grad_norm": 1.921875,
+      "learning_rate": 6.7928595486695855e-06,
+      "loss": 0.0882,
+      "step": 21170
+    },
+    {
+      "epoch": 0.6776949412856366,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.786123273829572e-06,
+      "loss": 0.0731,
+      "step": 21180
+    },
+    {
+      "epoch": 0.6780149105685854,
+      "grad_norm": 5.40625,
+      "learning_rate": 6.779386998989559e-06,
+      "loss": 0.1255,
+      "step": 21190
+    },
+    {
+      "epoch": 0.6783348798515343,
+      "grad_norm": 12.875,
+      "learning_rate": 6.772650724149545e-06,
+      "loss": 0.1471,
+      "step": 21200
+    },
+    {
+      "epoch": 0.678654849134483,
+      "grad_norm": 7.25,
+      "learning_rate": 6.765914449309532e-06,
+      "loss": 0.109,
+      "step": 21210
+    },
+    {
+      "epoch": 0.6789748184174319,
+      "grad_norm": 46.75,
+      "learning_rate": 6.759178174469518e-06,
+      "loss": 0.0797,
+      "step": 21220
+    },
+    {
+      "epoch": 0.6792947877003808,
+      "grad_norm": 20.875,
+      "learning_rate": 6.752441899629505e-06,
+      "loss": 0.1738,
+      "step": 21230
+    },
+    {
+      "epoch": 0.6796147569833296,
+      "grad_norm": 9.0625,
+      "learning_rate": 6.745705624789491e-06,
+      "loss": 0.0804,
+      "step": 21240
+    },
+    {
+      "epoch": 0.6799347262662785,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.738969349949479e-06,
+      "loss": 0.1467,
+      "step": 21250
+    },
+    {
+      "epoch": 0.6802546955492272,
+      "grad_norm": 15.4375,
+      "learning_rate": 6.732233075109466e-06,
+      "loss": 0.1002,
+      "step": 21260
+    },
+    {
+      "epoch": 0.6805746648321761,
+      "grad_norm": 11.375,
+      "learning_rate": 6.725496800269452e-06,
+      "loss": 0.0615,
+      "step": 21270
+    },
+    {
+      "epoch": 0.680894634115125,
+      "grad_norm": 14.25,
+      "learning_rate": 6.718760525429439e-06,
+      "loss": 0.0938,
+      "step": 21280
+    },
+    {
+      "epoch": 0.6812146033980738,
+      "grad_norm": 4.65625,
+      "learning_rate": 6.712024250589425e-06,
+      "loss": 0.1235,
+      "step": 21290
+    },
+    {
+      "epoch": 0.6815345726810226,
+      "grad_norm": 5.6875,
+      "learning_rate": 6.705287975749412e-06,
+      "loss": 0.1374,
+      "step": 21300
+    },
+    {
+      "epoch": 0.6818545419639714,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.698551700909398e-06,
+      "loss": 0.1177,
+      "step": 21310
+    },
+    {
+      "epoch": 0.6821745112469203,
+      "grad_norm": 15.4375,
+      "learning_rate": 6.691815426069385e-06,
+      "loss": 0.1472,
+      "step": 21320
+    },
+    {
+      "epoch": 0.6824944805298692,
+      "grad_norm": 27.625,
+      "learning_rate": 6.6850791512293715e-06,
+      "loss": 0.1317,
+      "step": 21330
+    },
+    {
+      "epoch": 0.6828144498128179,
+      "grad_norm": 10.5625,
+      "learning_rate": 6.6783428763893575e-06,
+      "loss": 0.0762,
+      "step": 21340
+    },
+    {
+      "epoch": 0.6831344190957668,
+      "grad_norm": 12.4375,
+      "learning_rate": 6.671606601549344e-06,
+      "loss": 0.0873,
+      "step": 21350
+    },
+    {
+      "epoch": 0.6834543883787156,
+      "grad_norm": 11.8125,
+      "learning_rate": 6.6648703267093304e-06,
+      "loss": 0.1275,
+      "step": 21360
+    },
+    {
+      "epoch": 0.6837743576616645,
+      "grad_norm": 11.75,
+      "learning_rate": 6.658134051869317e-06,
+      "loss": 0.1848,
+      "step": 21370
+    },
+    {
+      "epoch": 0.6840943269446134,
+      "grad_norm": 21.375,
+      "learning_rate": 6.651397777029303e-06,
+      "loss": 0.1494,
+      "step": 21380
+    },
+    {
+      "epoch": 0.6844142962275621,
+      "grad_norm": 8.625,
+      "learning_rate": 6.64466150218929e-06,
+      "loss": 0.1111,
+      "step": 21390
+    },
+    {
+      "epoch": 0.684734265510511,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.637925227349276e-06,
+      "loss": 0.1096,
+      "step": 21400
+    },
+    {
+      "epoch": 0.6850542347934598,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.631188952509263e-06,
+      "loss": 0.1192,
+      "step": 21410
+    },
+    {
+      "epoch": 0.6853742040764087,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.62445267766925e-06,
+      "loss": 0.0696,
+      "step": 21420
+    },
+    {
+      "epoch": 0.6856941733593575,
+      "grad_norm": 19.0,
+      "learning_rate": 6.617716402829236e-06,
+      "loss": 0.1002,
+      "step": 21430
+    },
+    {
+      "epoch": 0.6860141426423063,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.610980127989223e-06,
+      "loss": 0.0872,
+      "step": 21440
+    },
+    {
+      "epoch": 0.6863341119252552,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.604243853149209e-06,
+      "loss": 0.0812,
+      "step": 21450
+    },
+    {
+      "epoch": 0.686654081208204,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.597507578309196e-06,
+      "loss": 0.094,
+      "step": 21460
+    },
+    {
+      "epoch": 0.6869740504911529,
+      "grad_norm": 15.625,
+      "learning_rate": 6.590771303469182e-06,
+      "loss": 0.1171,
+      "step": 21470
+    },
+    {
+      "epoch": 0.6872940197741016,
+      "grad_norm": 6.28125,
+      "learning_rate": 6.584035028629169e-06,
+      "loss": 0.0997,
+      "step": 21480
+    },
+    {
+      "epoch": 0.6876139890570505,
+      "grad_norm": 12.9375,
+      "learning_rate": 6.577298753789155e-06,
+      "loss": 0.0879,
+      "step": 21490
+    },
+    {
+      "epoch": 0.6879339583399994,
+      "grad_norm": 2.5,
+      "learning_rate": 6.570562478949142e-06,
+      "loss": 0.1138,
+      "step": 21500
+    },
+    {
+      "epoch": 0.6882539276229482,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.563826204109129e-06,
+      "loss": 0.0852,
+      "step": 21510
+    },
+    {
+      "epoch": 0.688573896905897,
+      "grad_norm": 19.0,
+      "learning_rate": 6.557089929269115e-06,
+      "loss": 0.1887,
+      "step": 21520
+    },
+    {
+      "epoch": 0.6888938661888458,
+      "grad_norm": 14.0625,
+      "learning_rate": 6.550353654429102e-06,
+      "loss": 0.092,
+      "step": 21530
+    },
+    {
+      "epoch": 0.6892138354717947,
+      "grad_norm": 19.625,
+      "learning_rate": 6.543617379589088e-06,
+      "loss": 0.1611,
+      "step": 21540
+    },
+    {
+      "epoch": 0.6895338047547436,
+      "grad_norm": 26.25,
+      "learning_rate": 6.5368811047490745e-06,
+      "loss": 0.131,
+      "step": 21550
+    },
+    {
+      "epoch": 0.6898537740376924,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.530144829909061e-06,
+      "loss": 0.0562,
+      "step": 21560
+    },
+    {
+      "epoch": 0.6901737433206412,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.5234085550690475e-06,
+      "loss": 0.1075,
+      "step": 21570
+    },
+    {
+      "epoch": 0.69049371260359,
+      "grad_norm": 7.09375,
+      "learning_rate": 6.5166722802290335e-06,
+      "loss": 0.1253,
+      "step": 21580
+    },
+    {
+      "epoch": 0.6908136818865389,
+      "grad_norm": 22.875,
+      "learning_rate": 6.50993600538902e-06,
+      "loss": 0.107,
+      "step": 21590
+    },
+    {
+      "epoch": 0.6911336511694878,
+      "grad_norm": 15.625,
+      "learning_rate": 6.503199730549007e-06,
+      "loss": 0.1003,
+      "step": 21600
+    },
+    {
+      "epoch": 0.6914536204524365,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.496463455708993e-06,
+      "loss": 0.1314,
+      "step": 21610
+    },
+    {
+      "epoch": 0.6917735897353854,
+      "grad_norm": 5.625,
+      "learning_rate": 6.48972718086898e-06,
+      "loss": 0.1194,
+      "step": 21620
+    },
+    {
+      "epoch": 0.6920935590183342,
+      "grad_norm": 6.375,
+      "learning_rate": 6.482990906028966e-06,
+      "loss": 0.1135,
+      "step": 21630
+    },
+    {
+      "epoch": 0.6924135283012831,
+      "grad_norm": 0.482421875,
+      "learning_rate": 6.476254631188953e-06,
+      "loss": 0.0806,
+      "step": 21640
+    },
+    {
+      "epoch": 0.6927334975842319,
+      "grad_norm": 22.375,
+      "learning_rate": 6.469518356348939e-06,
+      "loss": 0.1119,
+      "step": 21650
+    },
+    {
+      "epoch": 0.6930534668671807,
+      "grad_norm": 17.0,
+      "learning_rate": 6.462782081508926e-06,
+      "loss": 0.0831,
+      "step": 21660
+    },
+    {
+      "epoch": 0.6933734361501296,
+      "grad_norm": 14.5625,
+      "learning_rate": 6.456045806668912e-06,
+      "loss": 0.1121,
+      "step": 21670
+    },
+    {
+      "epoch": 0.6936934054330784,
+      "grad_norm": 12.375,
+      "learning_rate": 6.449309531828899e-06,
+      "loss": 0.0832,
+      "step": 21680
+    },
+    {
+      "epoch": 0.6940133747160273,
+      "grad_norm": 18.5,
+      "learning_rate": 6.442573256988885e-06,
+      "loss": 0.1105,
+      "step": 21690
+    },
+    {
+      "epoch": 0.6943333439989761,
+      "grad_norm": 11.8125,
+      "learning_rate": 6.435836982148872e-06,
+      "loss": 0.1359,
+      "step": 21700
+    },
+    {
+      "epoch": 0.6946533132819249,
+      "grad_norm": 13.5,
+      "learning_rate": 6.429100707308859e-06,
+      "loss": 0.1036,
+      "step": 21710
+    },
+    {
+      "epoch": 0.6949732825648738,
+      "grad_norm": 22.625,
+      "learning_rate": 6.422364432468845e-06,
+      "loss": 0.1059,
+      "step": 21720
+    },
+    {
+      "epoch": 0.6952932518478226,
+      "grad_norm": 4.9375,
+      "learning_rate": 6.415628157628832e-06,
+      "loss": 0.0756,
+      "step": 21730
+    },
+    {
+      "epoch": 0.6956132211307714,
+      "grad_norm": 23.625,
+      "learning_rate": 6.408891882788818e-06,
+      "loss": 0.0888,
+      "step": 21740
+    },
+    {
+      "epoch": 0.6959331904137203,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.402155607948805e-06,
+      "loss": 0.0691,
+      "step": 21750
+    },
+    {
+      "epoch": 0.6962531596966691,
+      "grad_norm": 19.25,
+      "learning_rate": 6.395419333108791e-06,
+      "loss": 0.1483,
+      "step": 21760
+    },
+    {
+      "epoch": 0.696573128979618,
+      "grad_norm": 3.625,
+      "learning_rate": 6.388683058268778e-06,
+      "loss": 0.1069,
+      "step": 21770
+    },
+    {
+      "epoch": 0.6968930982625668,
+      "grad_norm": 12.4375,
+      "learning_rate": 6.381946783428764e-06,
+      "loss": 0.0811,
+      "step": 21780
+    },
+    {
+      "epoch": 0.6972130675455156,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.3752105085887505e-06,
+      "loss": 0.1532,
+      "step": 21790
+    },
+    {
+      "epoch": 0.6975330368284645,
+      "grad_norm": 9.625,
+      "learning_rate": 6.3684742337487374e-06,
+      "loss": 0.1015,
+      "step": 21800
+    },
+    {
+      "epoch": 0.6978530061114133,
+      "grad_norm": 9.125,
+      "learning_rate": 6.3617379589087235e-06,
+      "loss": 0.1437,
+      "step": 21810
+    },
+    {
+      "epoch": 0.6981729753943622,
+      "grad_norm": 6.3125,
+      "learning_rate": 6.35500168406871e-06,
+      "loss": 0.0685,
+      "step": 21820
+    },
+    {
+      "epoch": 0.6984929446773109,
+      "grad_norm": 10.875,
+      "learning_rate": 6.348265409228696e-06,
+      "loss": 0.1084,
+      "step": 21830
+    },
+    {
+      "epoch": 0.6988129139602598,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.341529134388683e-06,
+      "loss": 0.1048,
+      "step": 21840
+    },
+    {
+      "epoch": 0.6991328832432087,
+      "grad_norm": 5.84375,
+      "learning_rate": 6.334792859548669e-06,
+      "loss": 0.1149,
+      "step": 21850
+    },
+    {
+      "epoch": 0.6994528525261575,
+      "grad_norm": 1.875,
+      "learning_rate": 6.328056584708656e-06,
+      "loss": 0.1618,
+      "step": 21860
+    },
+    {
+      "epoch": 0.6997728218091063,
+      "grad_norm": 13.125,
+      "learning_rate": 6.321320309868642e-06,
+      "loss": 0.1054,
+      "step": 21870
+    },
+    {
+      "epoch": 0.7000927910920551,
+      "grad_norm": 5.78125,
+      "learning_rate": 6.314584035028629e-06,
+      "loss": 0.1341,
+      "step": 21880
+    },
+    {
+      "epoch": 0.700412760375004,
+      "grad_norm": 8.8125,
+      "learning_rate": 6.307847760188616e-06,
+      "loss": 0.0701,
+      "step": 21890
+    },
+    {
+      "epoch": 0.7007327296579529,
+      "grad_norm": 20.625,
+      "learning_rate": 6.301111485348603e-06,
+      "loss": 0.1398,
+      "step": 21900
+    },
+    {
+      "epoch": 0.7010526989409017,
+      "grad_norm": 17.875,
+      "learning_rate": 6.29437521050859e-06,
+      "loss": 0.1051,
+      "step": 21910
+    },
+    {
+      "epoch": 0.7013726682238505,
+      "grad_norm": 11.5,
+      "learning_rate": 6.287638935668577e-06,
+      "loss": 0.1059,
+      "step": 21920
+    },
+    {
+      "epoch": 0.7016926375067993,
+      "grad_norm": 42.75,
+      "learning_rate": 6.280902660828563e-06,
+      "loss": 0.0976,
+      "step": 21930
+    },
+    {
+      "epoch": 0.7020126067897482,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.27416638598855e-06,
+      "loss": 0.1205,
+      "step": 21940
+    },
+    {
+      "epoch": 0.7023325760726971,
+      "grad_norm": 0.455078125,
+      "learning_rate": 6.267430111148536e-06,
+      "loss": 0.082,
+      "step": 21950
+    },
+    {
+      "epoch": 0.7026525453556458,
+      "grad_norm": 12.6875,
+      "learning_rate": 6.2606938363085226e-06,
+      "loss": 0.1083,
+      "step": 21960
+    },
+    {
+      "epoch": 0.7029725146385947,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.253957561468509e-06,
+      "loss": 0.0909,
+      "step": 21970
+    },
+    {
+      "epoch": 0.7032924839215435,
+      "grad_norm": 5.46875,
+      "learning_rate": 6.2472212866284955e-06,
+      "loss": 0.0966,
+      "step": 21980
+    },
+    {
+      "epoch": 0.7036124532044924,
+      "grad_norm": 11.75,
+      "learning_rate": 6.2404850117884815e-06,
+      "loss": 0.1157,
+      "step": 21990
+    },
+    {
+      "epoch": 0.7039324224874413,
+      "grad_norm": 21.625,
+      "learning_rate": 6.2337487369484684e-06,
+      "loss": 0.1797,
+      "step": 22000
+    },
+    {
+      "epoch": 0.70425239177039,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.227012462108455e-06,
+      "loss": 0.1077,
+      "step": 22010
+    },
+    {
+      "epoch": 0.7045723610533389,
+      "grad_norm": 12.4375,
+      "learning_rate": 6.220276187268441e-06,
+      "loss": 0.1404,
+      "step": 22020
+    },
+    {
+      "epoch": 0.7048923303362877,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.213539912428428e-06,
+      "loss": 0.0437,
+      "step": 22030
+    },
+    {
+      "epoch": 0.7052122996192366,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.206803637588414e-06,
+      "loss": 0.1201,
+      "step": 22040
+    },
+    {
+      "epoch": 0.7055322689021853,
+      "grad_norm": 9.4375,
+      "learning_rate": 6.200067362748401e-06,
+      "loss": 0.131,
+      "step": 22050
+    },
+    {
+      "epoch": 0.7058522381851342,
+      "grad_norm": 6.96875,
+      "learning_rate": 6.193331087908387e-06,
+      "loss": 0.1306,
+      "step": 22060
+    },
+    {
+      "epoch": 0.7061722074680831,
+      "grad_norm": 16.75,
+      "learning_rate": 6.186594813068374e-06,
+      "loss": 0.2234,
+      "step": 22070
+    },
+    {
+      "epoch": 0.7064921767510319,
+      "grad_norm": 25.75,
+      "learning_rate": 6.17985853822836e-06,
+      "loss": 0.1111,
+      "step": 22080
+    },
+    {
+      "epoch": 0.7068121460339808,
+      "grad_norm": 14.9375,
+      "learning_rate": 6.173122263388347e-06,
+      "loss": 0.0629,
+      "step": 22090
+    },
+    {
+      "epoch": 0.7071321153169295,
+      "grad_norm": 13.1875,
+      "learning_rate": 6.166385988548333e-06,
+      "loss": 0.0609,
+      "step": 22100
+    },
+    {
+      "epoch": 0.7074520845998784,
+      "grad_norm": 11.875,
+      "learning_rate": 6.15964971370832e-06,
+      "loss": 0.217,
+      "step": 22110
+    },
+    {
+      "epoch": 0.7077720538828273,
+      "grad_norm": 14.0625,
+      "learning_rate": 6.152913438868307e-06,
+      "loss": 0.0941,
+      "step": 22120
+    },
+    {
+      "epoch": 0.7080920231657761,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.146177164028293e-06,
+      "loss": 0.1563,
+      "step": 22130
+    },
+    {
+      "epoch": 0.7084119924487249,
+      "grad_norm": 5.28125,
+      "learning_rate": 6.13944088918828e-06,
+      "loss": 0.1234,
+      "step": 22140
+    },
+    {
+      "epoch": 0.7087319617316737,
+      "grad_norm": 15.3125,
+      "learning_rate": 6.132704614348266e-06,
+      "loss": 0.1205,
+      "step": 22150
+    },
+    {
+      "epoch": 0.7090519310146226,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.125968339508253e-06,
+      "loss": 0.068,
+      "step": 22160
+    },
+    {
+      "epoch": 0.7093719002975715,
+      "grad_norm": 25.875,
+      "learning_rate": 6.119232064668239e-06,
+      "loss": 0.1535,
+      "step": 22170
+    },
+    {
+      "epoch": 0.7096918695805202,
+      "grad_norm": 14.0625,
+      "learning_rate": 6.112495789828226e-06,
+      "loss": 0.1171,
+      "step": 22180
+    },
+    {
+      "epoch": 0.7100118388634691,
+      "grad_norm": 13.9375,
+      "learning_rate": 6.105759514988212e-06,
+      "loss": 0.0756,
+      "step": 22190
+    },
+    {
+      "epoch": 0.7103318081464179,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.0990232401481986e-06,
+      "loss": 0.0698,
+      "step": 22200
+    },
+    {
+      "epoch": 0.7106517774293668,
+      "grad_norm": 26.75,
+      "learning_rate": 6.0922869653081855e-06,
+      "loss": 0.1055,
+      "step": 22210
+    },
+    {
+      "epoch": 0.7109717467123157,
+      "grad_norm": 14.3125,
+      "learning_rate": 6.0855506904681715e-06,
+      "loss": 0.0668,
+      "step": 22220
+    },
+    {
+      "epoch": 0.7112917159952644,
+      "grad_norm": 25.625,
+      "learning_rate": 6.078814415628158e-06,
+      "loss": 0.1934,
+      "step": 22230
+    },
+    {
+      "epoch": 0.7116116852782133,
+      "grad_norm": 12.1875,
+      "learning_rate": 6.0720781407881444e-06,
+      "loss": 0.1518,
+      "step": 22240
+    },
+    {
+      "epoch": 0.7119316545611621,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.065341865948131e-06,
+      "loss": 0.0687,
+      "step": 22250
+    },
+    {
+      "epoch": 0.712251623844111,
+      "grad_norm": 21.125,
+      "learning_rate": 6.058605591108117e-06,
+      "loss": 0.1328,
+      "step": 22260
+    },
+    {
+      "epoch": 0.7125715931270598,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.051869316268104e-06,
+      "loss": 0.1137,
+      "step": 22270
+    },
+    {
+      "epoch": 0.7128915624100086,
+      "grad_norm": 5.34375,
+      "learning_rate": 6.04513304142809e-06,
+      "loss": 0.0989,
+      "step": 22280
+    },
+    {
+      "epoch": 0.7132115316929575,
+      "grad_norm": 4.25,
+      "learning_rate": 6.038396766588077e-06,
+      "loss": 0.0996,
+      "step": 22290
+    },
+    {
+      "epoch": 0.7135315009759063,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.031660491748064e-06,
+      "loss": 0.1374,
+      "step": 22300
+    },
+    {
+      "epoch": 0.7138514702588552,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.02492421690805e-06,
+      "loss": 0.054,
+      "step": 22310
+    },
+    {
+      "epoch": 0.714171439541804,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.018187942068037e-06,
+      "loss": 0.0791,
+      "step": 22320
+    },
+    {
+      "epoch": 0.7144914088247528,
+      "grad_norm": 12.75,
+      "learning_rate": 6.011451667228023e-06,
+      "loss": 0.1519,
+      "step": 22330
+    },
+    {
+      "epoch": 0.7148113781077017,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.00471539238801e-06,
+      "loss": 0.1205,
+      "step": 22340
+    },
+    {
+      "epoch": 0.7151313473906505,
+      "grad_norm": 11.0,
+      "learning_rate": 5.997979117547996e-06,
+      "loss": 0.1396,
+      "step": 22350
+    },
+    {
+      "epoch": 0.7154513166735993,
+      "grad_norm": 10.375,
+      "learning_rate": 5.991242842707983e-06,
+      "loss": 0.1234,
+      "step": 22360
+    },
+    {
+      "epoch": 0.7157712859565482,
+      "grad_norm": 3.921875,
+      "learning_rate": 5.984506567867969e-06,
+      "loss": 0.0351,
+      "step": 22370
+    },
+    {
+      "epoch": 0.716091255239497,
+      "grad_norm": 6.375,
+      "learning_rate": 5.977770293027956e-06,
+      "loss": 0.0943,
+      "step": 22380
+    },
+    {
+      "epoch": 0.7164112245224459,
+      "grad_norm": 11.6875,
+      "learning_rate": 5.971034018187943e-06,
+      "loss": 0.1087,
+      "step": 22390
+    },
+    {
+      "epoch": 0.7167311938053946,
+      "grad_norm": 9.8125,
+      "learning_rate": 5.964297743347929e-06,
+      "loss": 0.079,
+      "step": 22400
+    },
+    {
+      "epoch": 0.7170511630883435,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.957561468507916e-06,
+      "loss": 0.0985,
+      "step": 22410
+    },
+    {
+      "epoch": 0.7173711323712924,
+      "grad_norm": 13.8125,
+      "learning_rate": 5.950825193667902e-06,
+      "loss": 0.0748,
+      "step": 22420
+    },
+    {
+      "epoch": 0.7176911016542412,
+      "grad_norm": 27.875,
+      "learning_rate": 5.9440889188278885e-06,
+      "loss": 0.1493,
+      "step": 22430
+    },
+    {
+      "epoch": 0.7180110709371901,
+      "grad_norm": 4.09375,
+      "learning_rate": 5.9373526439878746e-06,
+      "loss": 0.0855,
+      "step": 22440
+    },
+    {
+      "epoch": 0.7183310402201388,
+      "grad_norm": 12.9375,
+      "learning_rate": 5.9306163691478615e-06,
+      "loss": 0.1036,
+      "step": 22450
+    },
+    {
+      "epoch": 0.7186510095030877,
+      "grad_norm": 21.375,
+      "learning_rate": 5.9238800943078475e-06,
+      "loss": 0.1174,
+      "step": 22460
+    },
+    {
+      "epoch": 0.7189709787860366,
+      "grad_norm": 9.6875,
+      "learning_rate": 5.917143819467834e-06,
+      "loss": 0.0934,
+      "step": 22470
+    },
+    {
+      "epoch": 0.7192909480689854,
+      "grad_norm": 6.4375,
+      "learning_rate": 5.9104075446278204e-06,
+      "loss": 0.0783,
+      "step": 22480
+    },
+    {
+      "epoch": 0.7196109173519342,
+      "grad_norm": 2.0,
+      "learning_rate": 5.903671269787807e-06,
+      "loss": 0.1512,
+      "step": 22490
+    },
+    {
+      "epoch": 0.719930886634883,
+      "grad_norm": 21.125,
+      "learning_rate": 5.896934994947794e-06,
+      "loss": 0.0826,
+      "step": 22500
+    },
+    {
+      "epoch": 0.7202508559178319,
+      "grad_norm": 16.375,
+      "learning_rate": 5.89019872010778e-06,
+      "loss": 0.1183,
+      "step": 22510
+    },
+    {
+      "epoch": 0.7205708252007808,
+      "grad_norm": 9.8125,
+      "learning_rate": 5.883462445267767e-06,
+      "loss": 0.0709,
+      "step": 22520
+    },
+    {
+      "epoch": 0.7208907944837296,
+      "grad_norm": 12.875,
+      "learning_rate": 5.876726170427753e-06,
+      "loss": 0.1073,
+      "step": 22530
+    },
+    {
+      "epoch": 0.7212107637666784,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.86998989558774e-06,
+      "loss": 0.1094,
+      "step": 22540
+    },
+    {
+      "epoch": 0.7215307330496272,
+      "grad_norm": 13.625,
+      "learning_rate": 5.863253620747728e-06,
+      "loss": 0.0847,
+      "step": 22550
+    },
+    {
+      "epoch": 0.7218507023325761,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.856517345907714e-06,
+      "loss": 0.1061,
+      "step": 22560
+    },
+    {
+      "epoch": 0.722170671615525,
+      "grad_norm": 7.46875,
+      "learning_rate": 5.849781071067701e-06,
+      "loss": 0.1372,
+      "step": 22570
+    },
+    {
+      "epoch": 0.7224906408984737,
+      "grad_norm": 12.5,
+      "learning_rate": 5.843044796227687e-06,
+      "loss": 0.1453,
+      "step": 22580
+    },
+    {
+      "epoch": 0.7228106101814226,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.836308521387674e-06,
+      "loss": 0.07,
+      "step": 22590
+    },
+    {
+      "epoch": 0.7231305794643714,
+      "grad_norm": 16.625,
+      "learning_rate": 5.82957224654766e-06,
+      "loss": 0.1414,
+      "step": 22600
+    },
+    {
+      "epoch": 0.7234505487473203,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.822835971707647e-06,
+      "loss": 0.1263,
+      "step": 22610
+    },
+    {
+      "epoch": 0.723770518030269,
+      "grad_norm": 15.125,
+      "learning_rate": 5.8160996968676335e-06,
+      "loss": 0.191,
+      "step": 22620
+    },
+    {
+      "epoch": 0.7240904873132179,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.8093634220276195e-06,
+      "loss": 0.1746,
+      "step": 22630
+    },
+    {
+      "epoch": 0.7244104565961668,
+      "grad_norm": 5.40625,
+      "learning_rate": 5.802627147187606e-06,
+      "loss": 0.0994,
+      "step": 22640
+    },
+    {
+      "epoch": 0.7247304258791156,
+      "grad_norm": 8.6875,
+      "learning_rate": 5.7958908723475925e-06,
+      "loss": 0.1345,
+      "step": 22650
+    },
+    {
+      "epoch": 0.7250503951620645,
+      "grad_norm": 11.375,
+      "learning_rate": 5.789154597507579e-06,
+      "loss": 0.1087,
+      "step": 22660
+    },
+    {
+      "epoch": 0.7253703644450132,
+      "grad_norm": 9.0625,
+      "learning_rate": 5.782418322667565e-06,
+      "loss": 0.1405,
+      "step": 22670
+    },
+    {
+      "epoch": 0.7256903337279621,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.775682047827552e-06,
+      "loss": 0.1551,
+      "step": 22680
+    },
+    {
+      "epoch": 0.726010303010911,
+      "grad_norm": 36.5,
+      "learning_rate": 5.768945772987538e-06,
+      "loss": 0.1213,
+      "step": 22690
+    },
+    {
+      "epoch": 0.7263302722938598,
+      "grad_norm": 10.8125,
+      "learning_rate": 5.762209498147525e-06,
+      "loss": 0.1649,
+      "step": 22700
+    },
+    {
+      "epoch": 0.7266502415768086,
+      "grad_norm": 14.875,
+      "learning_rate": 5.755473223307512e-06,
+      "loss": 0.0815,
+      "step": 22710
+    },
+    {
+      "epoch": 0.7269702108597574,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.748736948467498e-06,
+      "loss": 0.0946,
+      "step": 22720
+    },
+    {
+      "epoch": 0.7272901801427063,
+      "grad_norm": 8.375,
+      "learning_rate": 5.742000673627485e-06,
+      "loss": 0.0957,
+      "step": 22730
+    },
+    {
+      "epoch": 0.7276101494256552,
+      "grad_norm": 23.875,
+      "learning_rate": 5.735264398787471e-06,
+      "loss": 0.0901,
+      "step": 22740
+    },
+    {
+      "epoch": 0.727930118708604,
+      "grad_norm": 5.78125,
+      "learning_rate": 5.728528123947458e-06,
+      "loss": 0.1153,
+      "step": 22750
+    },
+    {
+      "epoch": 0.7282500879915528,
+      "grad_norm": 18.375,
+      "learning_rate": 5.721791849107444e-06,
+      "loss": 0.153,
+      "step": 22760
+    },
+    {
+      "epoch": 0.7285700572745016,
+      "grad_norm": 6.0,
+      "learning_rate": 5.715055574267431e-06,
+      "loss": 0.1297,
+      "step": 22770
+    },
+    {
+      "epoch": 0.7288900265574505,
+      "grad_norm": 29.375,
+      "learning_rate": 5.708319299427417e-06,
+      "loss": 0.083,
+      "step": 22780
+    },
+    {
+      "epoch": 0.7292099958403994,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.701583024587404e-06,
+      "loss": 0.0606,
+      "step": 22790
+    },
+    {
+      "epoch": 0.7295299651233481,
+      "grad_norm": 10.875,
+      "learning_rate": 5.69484674974739e-06,
+      "loss": 0.1635,
+      "step": 22800
+    },
+    {
+      "epoch": 0.729849934406297,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.688110474907377e-06,
+      "loss": 0.0733,
+      "step": 22810
+    },
+    {
+      "epoch": 0.7301699036892458,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.681374200067364e-06,
+      "loss": 0.0844,
+      "step": 22820
+    },
+    {
+      "epoch": 0.7304898729721947,
+      "grad_norm": 10.125,
+      "learning_rate": 5.67463792522735e-06,
+      "loss": 0.1162,
+      "step": 22830
+    },
+    {
+      "epoch": 0.7308098422551436,
+      "grad_norm": 5.59375,
+      "learning_rate": 5.6679016503873366e-06,
+      "loss": 0.0697,
+      "step": 22840
+    },
+    {
+      "epoch": 0.7311298115380923,
+      "grad_norm": 10.4375,
+      "learning_rate": 5.661165375547323e-06,
+      "loss": 0.1829,
+      "step": 22850
+    },
+    {
+      "epoch": 0.7314497808210412,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.6544291007073095e-06,
+      "loss": 0.0806,
+      "step": 22860
+    },
+    {
+      "epoch": 0.73176975010399,
+      "grad_norm": 10.625,
+      "learning_rate": 5.6476928258672955e-06,
+      "loss": 0.1137,
+      "step": 22870
+    },
+    {
+      "epoch": 0.7320897193869389,
+      "grad_norm": 6.03125,
+      "learning_rate": 5.640956551027282e-06,
+      "loss": 0.1092,
+      "step": 22880
+    },
+    {
+      "epoch": 0.7324096886698876,
+      "grad_norm": 11.9375,
+      "learning_rate": 5.6342202761872685e-06,
+      "loss": 0.143,
+      "step": 22890
+    },
+    {
+      "epoch": 0.7327296579528365,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.627484001347255e-06,
+      "loss": 0.1039,
+      "step": 22900
+    },
+    {
+      "epoch": 0.7330496272357854,
+      "grad_norm": 14.6875,
+      "learning_rate": 5.620747726507242e-06,
+      "loss": 0.1587,
+      "step": 22910
+    },
+    {
+      "epoch": 0.7333695965187342,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.614011451667228e-06,
+      "loss": 0.0811,
+      "step": 22920
+    },
+    {
+      "epoch": 0.733689565801683,
+      "grad_norm": 14.625,
+      "learning_rate": 5.607275176827215e-06,
+      "loss": 0.0906,
+      "step": 22930
+    },
+    {
+      "epoch": 0.7340095350846318,
+      "grad_norm": 6.1875,
+      "learning_rate": 5.600538901987201e-06,
+      "loss": 0.088,
+      "step": 22940
+    },
+    {
+      "epoch": 0.7343295043675807,
+      "grad_norm": 18.25,
+      "learning_rate": 5.593802627147188e-06,
+      "loss": 0.1123,
+      "step": 22950
+    },
+    {
+      "epoch": 0.7346494736505296,
+      "grad_norm": 12.1875,
+      "learning_rate": 5.587066352307174e-06,
+      "loss": 0.123,
+      "step": 22960
+    },
+    {
+      "epoch": 0.7349694429334784,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.580330077467161e-06,
+      "loss": 0.125,
+      "step": 22970
+    },
+    {
+      "epoch": 0.7352894122164272,
+      "grad_norm": 5.5625,
+      "learning_rate": 5.573593802627147e-06,
+      "loss": 0.0638,
+      "step": 22980
+    },
+    {
+      "epoch": 0.735609381499376,
+      "grad_norm": 5.34375,
+      "learning_rate": 5.566857527787134e-06,
+      "loss": 0.1272,
+      "step": 22990
+    },
+    {
+      "epoch": 0.7359293507823249,
+      "grad_norm": 11.9375,
+      "learning_rate": 5.560121252947121e-06,
+      "loss": 0.0605,
+      "step": 23000
+    },
+    {
+      "epoch": 0.7362493200652738,
+      "grad_norm": 1.78125,
+      "learning_rate": 5.553384978107107e-06,
+      "loss": 0.0956,
+      "step": 23010
+    },
+    {
+      "epoch": 0.7365692893482225,
+      "grad_norm": 9.625,
+      "learning_rate": 5.546648703267094e-06,
+      "loss": 0.1145,
+      "step": 23020
+    },
+    {
+      "epoch": 0.7368892586311714,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.53991242842708e-06,
+      "loss": 0.0846,
+      "step": 23030
+    },
+    {
+      "epoch": 0.7372092279141202,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.533176153587067e-06,
+      "loss": 0.0686,
+      "step": 23040
+    },
+    {
+      "epoch": 0.7375291971970691,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.526439878747053e-06,
+      "loss": 0.1311,
+      "step": 23050
+    },
+    {
+      "epoch": 0.737849166480018,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.51970360390704e-06,
+      "loss": 0.0558,
+      "step": 23060
+    },
+    {
+      "epoch": 0.7381691357629667,
+      "grad_norm": 4.84375,
+      "learning_rate": 5.512967329067026e-06,
+      "loss": 0.1392,
+      "step": 23070
+    },
+    {
+      "epoch": 0.7384891050459156,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.5062310542270126e-06,
+      "loss": 0.0896,
+      "step": 23080
+    },
+    {
+      "epoch": 0.7388090743288644,
+      "grad_norm": 8.0,
+      "learning_rate": 5.4994947793869994e-06,
+      "loss": 0.0966,
+      "step": 23090
+    },
+    {
+      "epoch": 0.7391290436118133,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.4927585045469855e-06,
+      "loss": 0.082,
+      "step": 23100
+    },
+    {
+      "epoch": 0.7394490128947621,
+      "grad_norm": 7.125,
+      "learning_rate": 5.486022229706972e-06,
+      "loss": 0.0849,
+      "step": 23110
+    },
+    {
+      "epoch": 0.7397689821777109,
+      "grad_norm": 25.375,
+      "learning_rate": 5.479285954866958e-06,
+      "loss": 0.1424,
+      "step": 23120
+    },
+    {
+      "epoch": 0.7400889514606598,
+      "grad_norm": 0.8046875,
+      "learning_rate": 5.472549680026945e-06,
+      "loss": 0.0932,
+      "step": 23130
+    },
+    {
+      "epoch": 0.7404089207436086,
+      "grad_norm": 14.0625,
+      "learning_rate": 5.465813405186931e-06,
+      "loss": 0.0951,
+      "step": 23140
+    },
+    {
+      "epoch": 0.7407288900265574,
+      "grad_norm": 8.9375,
+      "learning_rate": 5.459077130346918e-06,
+      "loss": 0.0788,
+      "step": 23150
+    },
+    {
+      "epoch": 0.7410488593095063,
+      "grad_norm": 17.875,
+      "learning_rate": 5.452340855506904e-06,
+      "loss": 0.0854,
+      "step": 23160
+    },
+    {
+      "epoch": 0.7413688285924551,
+      "grad_norm": 1.78125,
+      "learning_rate": 5.445604580666891e-06,
+      "loss": 0.0994,
+      "step": 23170
+    },
+    {
+      "epoch": 0.741688797875404,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.438868305826878e-06,
+      "loss": 0.0653,
+      "step": 23180
+    },
+    {
+      "epoch": 0.7420087671583528,
+      "grad_norm": 25.125,
+      "learning_rate": 5.432132030986865e-06,
+      "loss": 0.093,
+      "step": 23190
+    },
+    {
+      "epoch": 0.7423287364413016,
+      "grad_norm": 9.375,
+      "learning_rate": 5.425395756146852e-06,
+      "loss": 0.1216,
+      "step": 23200
+    },
+    {
+      "epoch": 0.7426487057242505,
+      "grad_norm": 4.8125,
+      "learning_rate": 5.418659481306838e-06,
+      "loss": 0.1029,
+      "step": 23210
+    },
+    {
+      "epoch": 0.7429686750071993,
+      "grad_norm": 24.875,
+      "learning_rate": 5.411923206466825e-06,
+      "loss": 0.1356,
+      "step": 23220
+    },
+    {
+      "epoch": 0.7432886442901482,
+      "grad_norm": 8.75,
+      "learning_rate": 5.405186931626812e-06,
+      "loss": 0.1165,
+      "step": 23230
+    },
+    {
+      "epoch": 0.7436086135730969,
+      "grad_norm": 13.0,
+      "learning_rate": 5.398450656786798e-06,
+      "loss": 0.1202,
+      "step": 23240
+    },
+    {
+      "epoch": 0.7439285828560458,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.391714381946785e-06,
+      "loss": 0.0822,
+      "step": 23250
+    },
+    {
+      "epoch": 0.7442485521389947,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.384978107106771e-06,
+      "loss": 0.0837,
+      "step": 23260
+    },
+    {
+      "epoch": 0.7445685214219435,
+      "grad_norm": 0.294921875,
+      "learning_rate": 5.3782418322667575e-06,
+      "loss": 0.134,
+      "step": 23270
+    },
+    {
+      "epoch": 0.7448884907048924,
+      "grad_norm": 11.9375,
+      "learning_rate": 5.3715055574267435e-06,
+      "loss": 0.1752,
+      "step": 23280
+    },
+    {
+      "epoch": 0.7452084599878411,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.3647692825867304e-06,
+      "loss": 0.1131,
+      "step": 23290
+    },
+    {
+      "epoch": 0.74552842927079,
+      "grad_norm": 13.375,
+      "learning_rate": 5.3580330077467165e-06,
+      "loss": 0.0687,
+      "step": 23300
+    },
+    {
+      "epoch": 0.7458483985537389,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.351296732906703e-06,
+      "loss": 0.1376,
+      "step": 23310
+    },
+    {
+      "epoch": 0.7461683678366877,
+      "grad_norm": 19.75,
+      "learning_rate": 5.34456045806669e-06,
+      "loss": 0.167,
+      "step": 23320
+    },
+    {
+      "epoch": 0.7464883371196365,
+      "grad_norm": 17.125,
+      "learning_rate": 5.337824183226676e-06,
+      "loss": 0.1862,
+      "step": 23330
+    },
+    {
+      "epoch": 0.7468083064025853,
+      "grad_norm": 19.0,
+      "learning_rate": 5.331087908386663e-06,
+      "loss": 0.1813,
+      "step": 23340
+    },
+    {
+      "epoch": 0.7471282756855342,
+      "grad_norm": 5.625,
+      "learning_rate": 5.324351633546649e-06,
+      "loss": 0.1141,
+      "step": 23350
+    },
+    {
+      "epoch": 0.747448244968483,
+      "grad_norm": 15.875,
+      "learning_rate": 5.317615358706636e-06,
+      "loss": 0.0961,
+      "step": 23360
+    },
+    {
+      "epoch": 0.7477682142514319,
+      "grad_norm": 1.984375,
+      "learning_rate": 5.310879083866622e-06,
+      "loss": 0.125,
+      "step": 23370
+    },
+    {
+      "epoch": 0.7480881835343807,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.304142809026609e-06,
+      "loss": 0.1249,
+      "step": 23380
+    },
+    {
+      "epoch": 0.7484081528173295,
+      "grad_norm": 4.34375,
+      "learning_rate": 5.297406534186595e-06,
+      "loss": 0.0731,
+      "step": 23390
+    },
+    {
+      "epoch": 0.7487281221002784,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.290670259346582e-06,
+      "loss": 0.094,
+      "step": 23400
+    },
+    {
+      "epoch": 0.7490480913832273,
+      "grad_norm": 26.125,
+      "learning_rate": 5.283933984506569e-06,
+      "loss": 0.1594,
+      "step": 23410
+    },
+    {
+      "epoch": 0.749368060666176,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.277197709666555e-06,
+      "loss": 0.1418,
+      "step": 23420
+    },
+    {
+      "epoch": 0.7496880299491249,
+      "grad_norm": 14.375,
+      "learning_rate": 5.270461434826542e-06,
+      "loss": 0.1032,
+      "step": 23430
+    },
+    {
+      "epoch": 0.7500079992320737,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.263725159986528e-06,
+      "loss": 0.0911,
+      "step": 23440
+    },
+    {
+      "epoch": 0.7503279685150226,
+      "grad_norm": 5.5625,
+      "learning_rate": 5.256988885146515e-06,
+      "loss": 0.1399,
+      "step": 23450
+    },
+    {
+      "epoch": 0.7506479377979713,
+      "grad_norm": 14.1875,
+      "learning_rate": 5.250252610306501e-06,
+      "loss": 0.1017,
+      "step": 23460
+    },
+    {
+      "epoch": 0.7509679070809202,
+      "grad_norm": 9.6875,
+      "learning_rate": 5.243516335466488e-06,
+      "loss": 0.0914,
+      "step": 23470
+    },
+    {
+      "epoch": 0.7512878763638691,
+      "grad_norm": 16.5,
+      "learning_rate": 5.236780060626474e-06,
+      "loss": 0.1092,
+      "step": 23480
+    },
+    {
+      "epoch": 0.7516078456468179,
+      "grad_norm": 8.375,
+      "learning_rate": 5.230043785786461e-06,
+      "loss": 0.1256,
+      "step": 23490
+    },
+    {
+      "epoch": 0.7519278149297668,
+      "grad_norm": 15.25,
+      "learning_rate": 5.2233075109464475e-06,
+      "loss": 0.1031,
+      "step": 23500
+    },
+    {
+      "epoch": 0.7522477842127155,
+      "grad_norm": 19.875,
+      "learning_rate": 5.2165712361064335e-06,
+      "loss": 0.121,
+      "step": 23510
+    },
+    {
+      "epoch": 0.7525677534956644,
+      "grad_norm": 7.5,
+      "learning_rate": 5.20983496126642e-06,
+      "loss": 0.0723,
+      "step": 23520
+    },
+    {
+      "epoch": 0.7528877227786133,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.2030986864264064e-06,
+      "loss": 0.0531,
+      "step": 23530
+    },
+    {
+      "epoch": 0.7532076920615621,
+      "grad_norm": 20.625,
+      "learning_rate": 5.196362411586393e-06,
+      "loss": 0.113,
+      "step": 23540
+    },
+    {
+      "epoch": 0.7535276613445109,
+      "grad_norm": 18.5,
+      "learning_rate": 5.189626136746379e-06,
+      "loss": 0.0848,
+      "step": 23550
+    },
+    {
+      "epoch": 0.7538476306274597,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.182889861906366e-06,
+      "loss": 0.1387,
+      "step": 23560
+    },
+    {
+      "epoch": 0.7541675999104086,
+      "grad_norm": 13.8125,
+      "learning_rate": 5.176153587066352e-06,
+      "loss": 0.1026,
+      "step": 23570
+    },
+    {
+      "epoch": 0.7544875691933575,
+      "grad_norm": 7.8125,
+      "learning_rate": 5.169417312226339e-06,
+      "loss": 0.0791,
+      "step": 23580
+    },
+    {
+      "epoch": 0.7548075384763063,
+      "grad_norm": 17.875,
+      "learning_rate": 5.162681037386325e-06,
+      "loss": 0.1348,
+      "step": 23590
+    },
+    {
+      "epoch": 0.7551275077592551,
+      "grad_norm": 11.875,
+      "learning_rate": 5.155944762546312e-06,
+      "loss": 0.084,
+      "step": 23600
+    },
+    {
+      "epoch": 0.7554474770422039,
+      "grad_norm": 18.25,
+      "learning_rate": 5.149208487706299e-06,
+      "loss": 0.0931,
+      "step": 23610
+    },
+    {
+      "epoch": 0.7557674463251528,
+      "grad_norm": 31.625,
+      "learning_rate": 5.142472212866285e-06,
+      "loss": 0.1076,
+      "step": 23620
+    },
+    {
+      "epoch": 0.7560874156081017,
+      "grad_norm": 5.6875,
+      "learning_rate": 5.135735938026272e-06,
+      "loss": 0.0966,
+      "step": 23630
+    },
+    {
+      "epoch": 0.7564073848910504,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.128999663186258e-06,
+      "loss": 0.0546,
+      "step": 23640
+    },
+    {
+      "epoch": 0.7567273541739993,
+      "grad_norm": 14.25,
+      "learning_rate": 5.122263388346245e-06,
+      "loss": 0.1344,
+      "step": 23650
+    },
+    {
+      "epoch": 0.7570473234569481,
+      "grad_norm": 20.0,
+      "learning_rate": 5.115527113506231e-06,
+      "loss": 0.1221,
+      "step": 23660
+    },
+    {
+      "epoch": 0.757367292739897,
+      "grad_norm": 9.9375,
+      "learning_rate": 5.108790838666218e-06,
+      "loss": 0.1176,
+      "step": 23670
+    },
+    {
+      "epoch": 0.7576872620228458,
+      "grad_norm": 0.796875,
+      "learning_rate": 5.102054563826204e-06,
+      "loss": 0.0572,
+      "step": 23680
+    },
+    {
+      "epoch": 0.7580072313057946,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.095318288986191e-06,
+      "loss": 0.0788,
+      "step": 23690
+    },
+    {
+      "epoch": 0.7583272005887435,
+      "grad_norm": 14.75,
+      "learning_rate": 5.088582014146178e-06,
+      "loss": 0.1099,
+      "step": 23700
+    },
+    {
+      "epoch": 0.7586471698716923,
+      "grad_norm": 9.75,
+      "learning_rate": 5.081845739306164e-06,
+      "loss": 0.0943,
+      "step": 23710
+    },
+    {
+      "epoch": 0.7589671391546412,
+      "grad_norm": 6.0625,
+      "learning_rate": 5.0751094644661505e-06,
+      "loss": 0.1269,
+      "step": 23720
+    },
+    {
+      "epoch": 0.75928710843759,
+      "grad_norm": 15.1875,
+      "learning_rate": 5.068373189626137e-06,
+      "loss": 0.1426,
+      "step": 23730
+    },
+    {
+      "epoch": 0.7596070777205388,
+      "grad_norm": 4.71875,
+      "learning_rate": 5.0616369147861235e-06,
+      "loss": 0.0816,
+      "step": 23740
+    },
+    {
+      "epoch": 0.7599270470034877,
+      "grad_norm": 15.25,
+      "learning_rate": 5.0549006399461095e-06,
+      "loss": 0.1168,
+      "step": 23750
+    },
+    {
+      "epoch": 0.7602470162864365,
+      "grad_norm": 20.125,
+      "learning_rate": 5.048164365106096e-06,
+      "loss": 0.1414,
+      "step": 23760
+    },
+    {
+      "epoch": 0.7605669855693853,
+      "grad_norm": 12.25,
+      "learning_rate": 5.0414280902660824e-06,
+      "loss": 0.159,
+      "step": 23770
+    },
+    {
+      "epoch": 0.7608869548523342,
+      "grad_norm": 7.0625,
+      "learning_rate": 5.034691815426069e-06,
+      "loss": 0.1347,
+      "step": 23780
+    },
+    {
+      "epoch": 0.761206924135283,
+      "grad_norm": 16.875,
+      "learning_rate": 5.027955540586056e-06,
+      "loss": 0.1062,
+      "step": 23790
+    },
+    {
+      "epoch": 0.7615268934182319,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.021219265746042e-06,
+      "loss": 0.1231,
+      "step": 23800
+    },
+    {
+      "epoch": 0.7618468627011807,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.014482990906029e-06,
+      "loss": 0.1353,
+      "step": 23810
+    },
+    {
+      "epoch": 0.7621668319841295,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.007746716066015e-06,
+      "loss": 0.0671,
+      "step": 23820
+    },
+    {
+      "epoch": 0.7624868012670784,
+      "grad_norm": 5.71875,
+      "learning_rate": 5.001010441226002e-06,
+      "loss": 0.0989,
+      "step": 23830
+    },
+    {
+      "epoch": 0.7628067705500272,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.994274166385989e-06,
+      "loss": 0.0582,
+      "step": 23840
+    },
+    {
+      "epoch": 0.7631267398329761,
+      "grad_norm": 25.5,
+      "learning_rate": 4.987537891545976e-06,
+      "loss": 0.1007,
+      "step": 23850
+    },
+    {
+      "epoch": 0.7634467091159248,
+      "grad_norm": 25.75,
+      "learning_rate": 4.980801616705962e-06,
+      "loss": 0.1645,
+      "step": 23860
+    },
+    {
+      "epoch": 0.7637666783988737,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.974065341865949e-06,
+      "loss": 0.1078,
+      "step": 23870
+    },
+    {
+      "epoch": 0.7640866476818226,
+      "grad_norm": 13.5,
+      "learning_rate": 4.967329067025935e-06,
+      "loss": 0.1069,
+      "step": 23880
+    },
+    {
+      "epoch": 0.7644066169647714,
+      "grad_norm": 6.375,
+      "learning_rate": 4.960592792185922e-06,
+      "loss": 0.2056,
+      "step": 23890
+    },
+    {
+      "epoch": 0.7647265862477202,
+      "grad_norm": 20.875,
+      "learning_rate": 4.953856517345908e-06,
+      "loss": 0.1696,
+      "step": 23900
+    },
+    {
+      "epoch": 0.765046555530669,
+      "grad_norm": 14.8125,
+      "learning_rate": 4.947120242505895e-06,
+      "loss": 0.0932,
+      "step": 23910
+    },
+    {
+      "epoch": 0.7653665248136179,
+      "grad_norm": 1.65625,
+      "learning_rate": 4.940383967665881e-06,
+      "loss": 0.1406,
+      "step": 23920
+    },
+    {
+      "epoch": 0.7656864940965667,
+      "grad_norm": 16.875,
+      "learning_rate": 4.9336476928258676e-06,
+      "loss": 0.0738,
+      "step": 23930
+    },
+    {
+      "epoch": 0.7660064633795156,
+      "grad_norm": 33.5,
+      "learning_rate": 4.926911417985854e-06,
+      "loss": 0.1093,
+      "step": 23940
+    },
+    {
+      "epoch": 0.7663264326624644,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.9201751431458405e-06,
+      "loss": 0.1582,
+      "step": 23950
+    },
+    {
+      "epoch": 0.7666464019454132,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.913438868305827e-06,
+      "loss": 0.1022,
+      "step": 23960
+    },
+    {
+      "epoch": 0.7669663712283621,
+      "grad_norm": 9.0625,
+      "learning_rate": 4.9067025934658134e-06,
+      "loss": 0.1207,
+      "step": 23970
+    },
+    {
+      "epoch": 0.767286340511311,
+      "grad_norm": 20.0,
+      "learning_rate": 4.8999663186258e-06,
+      "loss": 0.044,
+      "step": 23980
+    },
+    {
+      "epoch": 0.7676063097942597,
+      "grad_norm": 14.5,
+      "learning_rate": 4.893230043785786e-06,
+      "loss": 0.0908,
+      "step": 23990
+    },
+    {
+      "epoch": 0.7679262790772086,
+      "grad_norm": 11.0,
+      "learning_rate": 4.886493768945773e-06,
+      "loss": 0.1493,
+      "step": 24000
+    },
+    {
+      "epoch": 0.7682462483601574,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.87975749410576e-06,
+      "loss": 0.0609,
+      "step": 24010
+    },
+    {
+      "epoch": 0.7685662176431063,
+      "grad_norm": 6.875,
+      "learning_rate": 4.873021219265747e-06,
+      "loss": 0.1127,
+      "step": 24020
+    },
+    {
+      "epoch": 0.7688861869260551,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.866284944425733e-06,
+      "loss": 0.0408,
+      "step": 24030
+    },
+    {
+      "epoch": 0.7692061562090039,
+      "grad_norm": 10.5,
+      "learning_rate": 4.85954866958572e-06,
+      "loss": 0.1193,
+      "step": 24040
+    },
+    {
+      "epoch": 0.7695261254919528,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.852812394745706e-06,
+      "loss": 0.1095,
+      "step": 24050
+    },
+    {
+      "epoch": 0.7698460947749016,
+      "grad_norm": 19.25,
+      "learning_rate": 4.846076119905693e-06,
+      "loss": 0.1393,
+      "step": 24060
+    },
+    {
+      "epoch": 0.7701660640578505,
+      "grad_norm": 20.875,
+      "learning_rate": 4.839339845065679e-06,
+      "loss": 0.113,
+      "step": 24070
+    },
+    {
+      "epoch": 0.7704860333407992,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.832603570225666e-06,
+      "loss": 0.0762,
+      "step": 24080
+    },
+    {
+      "epoch": 0.7708060026237481,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.825867295385652e-06,
+      "loss": 0.0885,
+      "step": 24090
+    },
+    {
+      "epoch": 0.771125971906697,
+      "grad_norm": 17.625,
+      "learning_rate": 4.819131020545639e-06,
+      "loss": 0.0782,
+      "step": 24100
+    },
+    {
+      "epoch": 0.7714459411896458,
+      "grad_norm": 16.0,
+      "learning_rate": 4.812394745705626e-06,
+      "loss": 0.1233,
+      "step": 24110
+    },
+    {
+      "epoch": 0.7717659104725947,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.805658470865612e-06,
+      "loss": 0.0857,
+      "step": 24120
+    },
+    {
+      "epoch": 0.7720858797555434,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.7989221960255986e-06,
+      "loss": 0.1749,
+      "step": 24130
+    },
+    {
+      "epoch": 0.7724058490384923,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 4.792185921185585e-06,
+      "loss": 0.1039,
+      "step": 24140
+    },
+    {
+      "epoch": 0.7727258183214412,
+      "grad_norm": 12.25,
+      "learning_rate": 4.7854496463455715e-06,
+      "loss": 0.0921,
+      "step": 24150
+    },
+    {
+      "epoch": 0.77304578760439,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.7787133715055575e-06,
+      "loss": 0.1247,
+      "step": 24160
+    },
+    {
+      "epoch": 0.7733657568873388,
+      "grad_norm": 1.9609375,
+      "learning_rate": 4.7719770966655444e-06,
+      "loss": 0.1465,
+      "step": 24170
+    },
+    {
+      "epoch": 0.7736857261702876,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.7652408218255305e-06,
+      "loss": 0.0654,
+      "step": 24180
+    },
+    {
+      "epoch": 0.7740056954532365,
+      "grad_norm": 21.25,
+      "learning_rate": 4.758504546985517e-06,
+      "loss": 0.0878,
+      "step": 24190
+    },
+    {
+      "epoch": 0.7743256647361854,
+      "grad_norm": 10.25,
+      "learning_rate": 4.751768272145504e-06,
+      "loss": 0.1133,
+      "step": 24200
+    },
+    {
+      "epoch": 0.7746456340191341,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.74503199730549e-06,
+      "loss": 0.1045,
+      "step": 24210
+    },
+    {
+      "epoch": 0.774965603302083,
+      "grad_norm": 21.25,
+      "learning_rate": 4.738295722465477e-06,
+      "loss": 0.1489,
+      "step": 24220
+    },
+    {
+      "epoch": 0.7752855725850318,
+      "grad_norm": 10.75,
+      "learning_rate": 4.731559447625463e-06,
+      "loss": 0.0739,
+      "step": 24230
+    },
+    {
+      "epoch": 0.7756055418679807,
+      "grad_norm": 12.875,
+      "learning_rate": 4.72482317278545e-06,
+      "loss": 0.0788,
+      "step": 24240
+    },
+    {
+      "epoch": 0.7759255111509296,
+      "grad_norm": 28.375,
+      "learning_rate": 4.718086897945436e-06,
+      "loss": 0.1271,
+      "step": 24250
+    },
+    {
+      "epoch": 0.7762454804338783,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.711350623105423e-06,
+      "loss": 0.0603,
+      "step": 24260
+    },
+    {
+      "epoch": 0.7765654497168272,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.704614348265409e-06,
+      "loss": 0.0914,
+      "step": 24270
+    },
+    {
+      "epoch": 0.776885418999776,
+      "grad_norm": 23.75,
+      "learning_rate": 4.697878073425396e-06,
+      "loss": 0.112,
+      "step": 24280
+    },
+    {
+      "epoch": 0.7772053882827249,
+      "grad_norm": 23.125,
+      "learning_rate": 4.691141798585383e-06,
+      "loss": 0.1202,
+      "step": 24290
+    },
+    {
+      "epoch": 0.7775253575656736,
+      "grad_norm": 12.25,
+      "learning_rate": 4.684405523745369e-06,
+      "loss": 0.1734,
+      "step": 24300
+    },
+    {
+      "epoch": 0.7778453268486225,
+      "grad_norm": 15.75,
+      "learning_rate": 4.677669248905356e-06,
+      "loss": 0.0789,
+      "step": 24310
+    },
+    {
+      "epoch": 0.7781652961315714,
+      "grad_norm": 35.25,
+      "learning_rate": 4.670932974065343e-06,
+      "loss": 0.1652,
+      "step": 24320
+    },
+    {
+      "epoch": 0.7784852654145202,
+      "grad_norm": 10.375,
+      "learning_rate": 4.664196699225329e-06,
+      "loss": 0.0908,
+      "step": 24330
+    },
+    {
+      "epoch": 0.7788052346974691,
+      "grad_norm": 17.625,
+      "learning_rate": 4.657460424385316e-06,
+      "loss": 0.1501,
+      "step": 24340
+    },
+    {
+      "epoch": 0.7791252039804178,
+      "grad_norm": 10.125,
+      "learning_rate": 4.650724149545302e-06,
+      "loss": 0.1478,
+      "step": 24350
+    },
+    {
+      "epoch": 0.7794451732633667,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.6439878747052885e-06,
+      "loss": 0.1584,
+      "step": 24360
+    },
+    {
+      "epoch": 0.7797651425463156,
+      "grad_norm": 14.1875,
+      "learning_rate": 4.637251599865275e-06,
+      "loss": 0.1025,
+      "step": 24370
+    },
+    {
+      "epoch": 0.7800851118292644,
+      "grad_norm": 25.875,
+      "learning_rate": 4.6305153250252615e-06,
+      "loss": 0.1103,
+      "step": 24380
+    },
+    {
+      "epoch": 0.7804050811122132,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.623779050185248e-06,
+      "loss": 0.0872,
+      "step": 24390
+    },
+    {
+      "epoch": 0.780725050395162,
+      "grad_norm": 12.4375,
+      "learning_rate": 4.617042775345234e-06,
+      "loss": 0.0856,
+      "step": 24400
+    },
+    {
+      "epoch": 0.7810450196781109,
+      "grad_norm": 1.4296875,
+      "learning_rate": 4.610306500505221e-06,
+      "loss": 0.1048,
+      "step": 24410
+    },
+    {
+      "epoch": 0.7813649889610598,
+      "grad_norm": 54.5,
+      "learning_rate": 4.603570225665207e-06,
+      "loss": 0.094,
+      "step": 24420
+    },
+    {
+      "epoch": 0.7816849582440085,
+      "grad_norm": 15.9375,
+      "learning_rate": 4.596833950825194e-06,
+      "loss": 0.1369,
+      "step": 24430
+    },
+    {
+      "epoch": 0.7820049275269574,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.59009767598518e-06,
+      "loss": 0.0601,
+      "step": 24440
+    },
+    {
+      "epoch": 0.7823248968099062,
+      "grad_norm": 11.75,
+      "learning_rate": 4.583361401145167e-06,
+      "loss": 0.1012,
+      "step": 24450
+    },
+    {
+      "epoch": 0.7826448660928551,
+      "grad_norm": 11.1875,
+      "learning_rate": 4.576625126305154e-06,
+      "loss": 0.0969,
+      "step": 24460
+    },
+    {
+      "epoch": 0.782964835375804,
+      "grad_norm": 25.375,
+      "learning_rate": 4.56988885146514e-06,
+      "loss": 0.1122,
+      "step": 24470
+    },
+    {
+      "epoch": 0.7832848046587527,
+      "grad_norm": 15.9375,
+      "learning_rate": 4.563152576625127e-06,
+      "loss": 0.053,
+      "step": 24480
+    },
+    {
+      "epoch": 0.7836047739417016,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.556416301785113e-06,
+      "loss": 0.0673,
+      "step": 24490
+    },
+    {
+      "epoch": 0.7839247432246504,
+      "grad_norm": 16.25,
+      "learning_rate": 4.5496800269451e-06,
+      "loss": 0.1106,
+      "step": 24500
+    },
+    {
+      "epoch": 0.7842447125075993,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.542943752105086e-06,
+      "loss": 0.0674,
+      "step": 24510
+    },
+    {
+      "epoch": 0.7845646817905481,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.536207477265073e-06,
+      "loss": 0.1191,
+      "step": 24520
+    },
+    {
+      "epoch": 0.7848846510734969,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.529471202425059e-06,
+      "loss": 0.0918,
+      "step": 24530
+    },
+    {
+      "epoch": 0.7852046203564458,
+      "grad_norm": 49.75,
+      "learning_rate": 4.522734927585046e-06,
+      "loss": 0.1386,
+      "step": 24540
+    },
+    {
+      "epoch": 0.7855245896393946,
+      "grad_norm": 24.25,
+      "learning_rate": 4.515998652745033e-06,
+      "loss": 0.1061,
+      "step": 24550
+    },
+    {
+      "epoch": 0.7858445589223435,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.509262377905019e-06,
+      "loss": 0.1245,
+      "step": 24560
+    },
+    {
+      "epoch": 0.7861645282052923,
+      "grad_norm": 13.8125,
+      "learning_rate": 4.5025261030650056e-06,
+      "loss": 0.1331,
+      "step": 24570
+    },
+    {
+      "epoch": 0.7864844974882411,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.495789828224992e-06,
+      "loss": 0.0716,
+      "step": 24580
+    },
+    {
+      "epoch": 0.78680446677119,
+      "grad_norm": 11.6875,
+      "learning_rate": 4.4890535533849785e-06,
+      "loss": 0.0899,
+      "step": 24590
+    },
+    {
+      "epoch": 0.7871244360541388,
+      "grad_norm": 17.875,
+      "learning_rate": 4.4823172785449645e-06,
+      "loss": 0.1161,
+      "step": 24600
+    },
+    {
+      "epoch": 0.7874444053370876,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.475581003704951e-06,
+      "loss": 0.1215,
+      "step": 24610
+    },
+    {
+      "epoch": 0.7877643746200365,
+      "grad_norm": 10.5,
+      "learning_rate": 4.4688447288649375e-06,
+      "loss": 0.0983,
+      "step": 24620
+    },
+    {
+      "epoch": 0.7880843439029853,
+      "grad_norm": 14.1875,
+      "learning_rate": 4.462108454024924e-06,
+      "loss": 0.112,
+      "step": 24630
+    },
+    {
+      "epoch": 0.7884043131859342,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.455372179184911e-06,
+      "loss": 0.0698,
+      "step": 24640
+    },
+    {
+      "epoch": 0.788724282468883,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.448635904344898e-06,
+      "loss": 0.101,
+      "step": 24650
+    },
+    {
+      "epoch": 0.7890442517518318,
+      "grad_norm": 13.1875,
+      "learning_rate": 4.441899629504884e-06,
+      "loss": 0.1169,
+      "step": 24660
+    },
+    {
+      "epoch": 0.7893642210347807,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.435163354664871e-06,
+      "loss": 0.076,
+      "step": 24670
+    },
+    {
+      "epoch": 0.7896841903177295,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.428427079824857e-06,
+      "loss": 0.107,
+      "step": 24680
+    },
+    {
+      "epoch": 0.7900041596006784,
+      "grad_norm": 8.75,
+      "learning_rate": 4.421690804984844e-06,
+      "loss": 0.0896,
+      "step": 24690
+    },
+    {
+      "epoch": 0.7903241288836271,
+      "grad_norm": 9.0,
+      "learning_rate": 4.41495453014483e-06,
+      "loss": 0.0668,
+      "step": 24700
+    },
+    {
+      "epoch": 0.790644098166576,
+      "grad_norm": 25.75,
+      "learning_rate": 4.408218255304817e-06,
+      "loss": 0.1462,
+      "step": 24710
+    },
+    {
+      "epoch": 0.7909640674495249,
+      "grad_norm": 7.46875,
+      "learning_rate": 4.401481980464804e-06,
+      "loss": 0.0954,
+      "step": 24720
+    },
+    {
+      "epoch": 0.7912840367324737,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.39474570562479e-06,
+      "loss": 0.1271,
+      "step": 24730
+    },
+    {
+      "epoch": 0.7916040060154225,
+      "grad_norm": 24.625,
+      "learning_rate": 4.388009430784777e-06,
+      "loss": 0.0901,
+      "step": 24740
+    },
+    {
+      "epoch": 0.7919239752983713,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.381273155944763e-06,
+      "loss": 0.101,
+      "step": 24750
+    },
+    {
+      "epoch": 0.7922439445813202,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.37453688110475e-06,
+      "loss": 0.1219,
+      "step": 24760
+    },
+    {
+      "epoch": 0.792563913864269,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.367800606264736e-06,
+      "loss": 0.119,
+      "step": 24770
+    },
+    {
+      "epoch": 0.7928838831472179,
+      "grad_norm": 15.5625,
+      "learning_rate": 4.361064331424723e-06,
+      "loss": 0.1503,
+      "step": 24780
+    },
+    {
+      "epoch": 0.7932038524301667,
+      "grad_norm": 17.375,
+      "learning_rate": 4.354328056584709e-06,
+      "loss": 0.1509,
+      "step": 24790
+    },
+    {
+      "epoch": 0.7935238217131155,
+      "grad_norm": 20.5,
+      "learning_rate": 4.3475917817446955e-06,
+      "loss": 0.1321,
+      "step": 24800
+    },
+    {
+      "epoch": 0.7938437909960644,
+      "grad_norm": 18.625,
+      "learning_rate": 4.340855506904682e-06,
+      "loss": 0.0568,
+      "step": 24810
+    },
+    {
+      "epoch": 0.7941637602790133,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.3341192320646685e-06,
+      "loss": 0.1381,
+      "step": 24820
+    },
+    {
+      "epoch": 0.794483729561962,
+      "grad_norm": 9.1875,
+      "learning_rate": 4.327382957224655e-06,
+      "loss": 0.13,
+      "step": 24830
+    },
+    {
+      "epoch": 0.7948036988449109,
+      "grad_norm": 14.5,
+      "learning_rate": 4.320646682384641e-06,
+      "loss": 0.1341,
+      "step": 24840
+    },
+    {
+      "epoch": 0.7951236681278597,
+      "grad_norm": 8.5,
+      "learning_rate": 4.313910407544628e-06,
+      "loss": 0.0861,
+      "step": 24850
+    },
+    {
+      "epoch": 0.7954436374108086,
+      "grad_norm": 5.25,
+      "learning_rate": 4.307174132704614e-06,
+      "loss": 0.1732,
+      "step": 24860
+    },
+    {
+      "epoch": 0.7957636066937575,
+      "grad_norm": 7.0625,
+      "learning_rate": 4.300437857864601e-06,
+      "loss": 0.073,
+      "step": 24870
+    },
+    {
+      "epoch": 0.7960835759767062,
+      "grad_norm": 15.4375,
+      "learning_rate": 4.293701583024587e-06,
+      "loss": 0.1574,
+      "step": 24880
+    },
+    {
+      "epoch": 0.7964035452596551,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.286965308184574e-06,
+      "loss": 0.1038,
+      "step": 24890
+    },
+    {
+      "epoch": 0.7967235145426039,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.280229033344561e-06,
+      "loss": 0.0847,
+      "step": 24900
+    },
+    {
+      "epoch": 0.7970434838255528,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.273492758504547e-06,
+      "loss": 0.1113,
+      "step": 24910
+    },
+    {
+      "epoch": 0.7973634531085015,
+      "grad_norm": 12.375,
+      "learning_rate": 4.266756483664534e-06,
+      "loss": 0.1403,
+      "step": 24920
+    },
+    {
+      "epoch": 0.7976834223914504,
+      "grad_norm": 10.1875,
+      "learning_rate": 4.26002020882452e-06,
+      "loss": 0.07,
+      "step": 24930
+    },
+    {
+      "epoch": 0.7980033916743993,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.253283933984507e-06,
+      "loss": 0.1548,
+      "step": 24940
+    },
+    {
+      "epoch": 0.7983233609573481,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.246547659144493e-06,
+      "loss": 0.1071,
+      "step": 24950
+    },
+    {
+      "epoch": 0.7986433302402969,
+      "grad_norm": 15.875,
+      "learning_rate": 4.23981138430448e-06,
+      "loss": 0.1146,
+      "step": 24960
+    },
+    {
+      "epoch": 0.7989632995232457,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.233075109464467e-06,
+      "loss": 0.1451,
+      "step": 24970
+    },
+    {
+      "epoch": 0.7992832688061946,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.226338834624454e-06,
+      "loss": 0.1184,
+      "step": 24980
+    },
+    {
+      "epoch": 0.7996032380891435,
+      "grad_norm": 7.34375,
+      "learning_rate": 4.21960255978444e-06,
+      "loss": 0.1455,
+      "step": 24990
+    },
+    {
+      "epoch": 0.7999232073720923,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.2128662849444265e-06,
+      "loss": 0.1296,
+      "step": 25000
+    },
+    {
+      "epoch": 0.8002431766550411,
+      "grad_norm": 8.875,
+      "learning_rate": 4.2061300101044126e-06,
+      "loss": 0.1127,
+      "step": 25010
+    },
+    {
+      "epoch": 0.8005631459379899,
+      "grad_norm": 12.25,
+      "learning_rate": 4.1993937352643994e-06,
+      "loss": 0.0406,
+      "step": 25020
+    },
+    {
+      "epoch": 0.8008831152209388,
+      "grad_norm": 9.0625,
+      "learning_rate": 4.1926574604243855e-06,
+      "loss": 0.1461,
+      "step": 25030
+    },
+    {
+      "epoch": 0.8012030845038877,
+      "grad_norm": 25.375,
+      "learning_rate": 4.185921185584372e-06,
+      "loss": 0.0801,
+      "step": 25040
+    },
+    {
+      "epoch": 0.8015230537868364,
+      "grad_norm": 17.625,
+      "learning_rate": 4.179184910744358e-06,
+      "loss": 0.0857,
+      "step": 25050
+    },
+    {
+      "epoch": 0.8018430230697853,
+      "grad_norm": 14.8125,
+      "learning_rate": 4.172448635904345e-06,
+      "loss": 0.118,
+      "step": 25060
+    },
+    {
+      "epoch": 0.8021629923527341,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.165712361064332e-06,
+      "loss": 0.1443,
+      "step": 25070
+    },
+    {
+      "epoch": 0.802482961635683,
+      "grad_norm": 29.5,
+      "learning_rate": 4.158976086224318e-06,
+      "loss": 0.0692,
+      "step": 25080
+    },
+    {
+      "epoch": 0.8028029309186319,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.152239811384305e-06,
+      "loss": 0.1291,
+      "step": 25090
+    },
+    {
+      "epoch": 0.8031229002015806,
+      "grad_norm": 20.5,
+      "learning_rate": 4.145503536544291e-06,
+      "loss": 0.0995,
+      "step": 25100
+    },
+    {
+      "epoch": 0.8034428694845295,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.138767261704278e-06,
+      "loss": 0.1184,
+      "step": 25110
+    },
+    {
+      "epoch": 0.8037628387674783,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.132030986864264e-06,
+      "loss": 0.0971,
+      "step": 25120
+    },
+    {
+      "epoch": 0.8040828080504272,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.125294712024251e-06,
+      "loss": 0.0952,
+      "step": 25130
+    },
+    {
+      "epoch": 0.804402777333376,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.118558437184237e-06,
+      "loss": 0.108,
+      "step": 25140
+    },
+    {
+      "epoch": 0.8047227466163248,
+      "grad_norm": 11.6875,
+      "learning_rate": 4.111822162344224e-06,
+      "loss": 0.1038,
+      "step": 25150
+    },
+    {
+      "epoch": 0.8050427158992737,
+      "grad_norm": 12.0625,
+      "learning_rate": 4.105085887504211e-06,
+      "loss": 0.1048,
+      "step": 25160
+    },
+    {
+      "epoch": 0.8053626851822225,
+      "grad_norm": 12.3125,
+      "learning_rate": 4.098349612664197e-06,
+      "loss": 0.075,
+      "step": 25170
+    },
+    {
+      "epoch": 0.8056826544651713,
+      "grad_norm": 23.875,
+      "learning_rate": 4.091613337824184e-06,
+      "loss": 0.0974,
+      "step": 25180
+    },
+    {
+      "epoch": 0.8060026237481202,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.08487706298417e-06,
+      "loss": 0.104,
+      "step": 25190
+    },
+    {
+      "epoch": 0.806322593031069,
+      "grad_norm": 16.5,
+      "learning_rate": 4.078140788144157e-06,
+      "loss": 0.0788,
+      "step": 25200
+    },
+    {
+      "epoch": 0.8066425623140179,
+      "grad_norm": 26.125,
+      "learning_rate": 4.071404513304143e-06,
+      "loss": 0.1206,
+      "step": 25210
+    },
+    {
+      "epoch": 0.8069625315969667,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.06466823846413e-06,
+      "loss": 0.0719,
+      "step": 25220
+    },
+    {
+      "epoch": 0.8072825008799155,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.057931963624116e-06,
+      "loss": 0.0406,
+      "step": 25230
+    },
+    {
+      "epoch": 0.8076024701628643,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.0511956887841025e-06,
+      "loss": 0.1575,
+      "step": 25240
+    },
+    {
+      "epoch": 0.8079224394458132,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.044459413944089e-06,
+      "loss": 0.0573,
+      "step": 25250
+    },
+    {
+      "epoch": 0.8082424087287621,
+      "grad_norm": 21.5,
+      "learning_rate": 4.0377231391040754e-06,
+      "loss": 0.1253,
+      "step": 25260
+    },
+    {
+      "epoch": 0.8085623780117108,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.030986864264062e-06,
+      "loss": 0.0968,
+      "step": 25270
+    },
+    {
+      "epoch": 0.8088823472946597,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.024250589424048e-06,
+      "loss": 0.1109,
+      "step": 25280
+    },
+    {
+      "epoch": 0.8092023165776085,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.017514314584035e-06,
+      "loss": 0.0831,
+      "step": 25290
+    },
+    {
+      "epoch": 0.8095222858605574,
+      "grad_norm": 32.5,
+      "learning_rate": 4.010778039744022e-06,
+      "loss": 0.1295,
+      "step": 25300
+    },
+    {
+      "epoch": 0.8098422551435063,
+      "grad_norm": 7.40625,
+      "learning_rate": 4.004041764904009e-06,
+      "loss": 0.0988,
+      "step": 25310
+    },
+    {
+      "epoch": 0.810162224426455,
+      "grad_norm": 9.1875,
+      "learning_rate": 3.997305490063995e-06,
+      "loss": 0.1058,
+      "step": 25320
+    },
+    {
+      "epoch": 0.8104821937094039,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.990569215223982e-06,
+      "loss": 0.1499,
+      "step": 25330
+    },
+    {
+      "epoch": 0.8108021629923527,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.983832940383968e-06,
+      "loss": 0.0452,
+      "step": 25340
+    },
+    {
+      "epoch": 0.8111221322753016,
+      "grad_norm": 13.5,
+      "learning_rate": 3.977096665543955e-06,
+      "loss": 0.0572,
+      "step": 25350
+    },
+    {
+      "epoch": 0.8114421015582504,
+      "grad_norm": 12.875,
+      "learning_rate": 3.970360390703941e-06,
+      "loss": 0.1075,
+      "step": 25360
+    },
+    {
+      "epoch": 0.8117620708411992,
+      "grad_norm": 8.0,
+      "learning_rate": 3.963624115863928e-06,
+      "loss": 0.0984,
+      "step": 25370
+    },
+    {
+      "epoch": 0.8120820401241481,
+      "grad_norm": 19.125,
+      "learning_rate": 3.956887841023914e-06,
+      "loss": 0.1493,
+      "step": 25380
+    },
+    {
+      "epoch": 0.812402009407097,
+      "grad_norm": 12.125,
+      "learning_rate": 3.950151566183901e-06,
+      "loss": 0.1212,
+      "step": 25390
+    },
+    {
+      "epoch": 0.8127219786900458,
+      "grad_norm": 8.625,
+      "learning_rate": 3.943415291343888e-06,
+      "loss": 0.1503,
+      "step": 25400
+    },
+    {
+      "epoch": 0.8130419479729946,
+      "grad_norm": 11.6875,
+      "learning_rate": 3.936679016503874e-06,
+      "loss": 0.1024,
+      "step": 25410
+    },
+    {
+      "epoch": 0.8133619172559434,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.929942741663861e-06,
+      "loss": 0.0917,
+      "step": 25420
+    },
+    {
+      "epoch": 0.8136818865388923,
+      "grad_norm": 11.0625,
+      "learning_rate": 3.923206466823847e-06,
+      "loss": 0.1651,
+      "step": 25430
+    },
+    {
+      "epoch": 0.8140018558218411,
+      "grad_norm": 4.125,
+      "learning_rate": 3.9164701919838335e-06,
+      "loss": 0.0405,
+      "step": 25440
+    },
+    {
+      "epoch": 0.8143218251047899,
+      "grad_norm": 13.625,
+      "learning_rate": 3.9097339171438195e-06,
+      "loss": 0.0846,
+      "step": 25450
+    },
+    {
+      "epoch": 0.8146417943877388,
+      "grad_norm": 17.875,
+      "learning_rate": 3.9029976423038064e-06,
+      "loss": 0.107,
+      "step": 25460
+    },
+    {
+      "epoch": 0.8149617636706876,
+      "grad_norm": 19.25,
+      "learning_rate": 3.8962613674637925e-06,
+      "loss": 0.1355,
+      "step": 25470
+    },
+    {
+      "epoch": 0.8152817329536365,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.889525092623779e-06,
+      "loss": 0.0642,
+      "step": 25480
+    },
+    {
+      "epoch": 0.8156017022365852,
+      "grad_norm": 8.5,
+      "learning_rate": 3.882788817783765e-06,
+      "loss": 0.0953,
+      "step": 25490
+    },
+    {
+      "epoch": 0.8159216715195341,
+      "grad_norm": 1.765625,
+      "learning_rate": 3.876052542943752e-06,
+      "loss": 0.1238,
+      "step": 25500
+    },
+    {
+      "epoch": 0.816241640802483,
+      "grad_norm": 19.5,
+      "learning_rate": 3.869316268103739e-06,
+      "loss": 0.1087,
+      "step": 25510
+    },
+    {
+      "epoch": 0.8165616100854318,
+      "grad_norm": 14.6875,
+      "learning_rate": 3.862579993263725e-06,
+      "loss": 0.1071,
+      "step": 25520
+    },
+    {
+      "epoch": 0.8168815793683807,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.855843718423712e-06,
+      "loss": 0.1191,
+      "step": 25530
+    },
+    {
+      "epoch": 0.8172015486513294,
+      "grad_norm": 27.0,
+      "learning_rate": 3.849107443583698e-06,
+      "loss": 0.1081,
+      "step": 25540
+    },
+    {
+      "epoch": 0.8175215179342783,
+      "grad_norm": 21.25,
+      "learning_rate": 3.842371168743685e-06,
+      "loss": 0.1368,
+      "step": 25550
+    },
+    {
+      "epoch": 0.8178414872172272,
+      "grad_norm": 10.6875,
+      "learning_rate": 3.835634893903671e-06,
+      "loss": 0.1391,
+      "step": 25560
+    },
+    {
+      "epoch": 0.818161456500176,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.828898619063658e-06,
+      "loss": 0.0959,
+      "step": 25570
+    },
+    {
+      "epoch": 0.8184814257831248,
+      "grad_norm": 11.75,
+      "learning_rate": 3.822162344223644e-06,
+      "loss": 0.1031,
+      "step": 25580
+    },
+    {
+      "epoch": 0.8188013950660736,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.815426069383631e-06,
+      "loss": 0.1009,
+      "step": 25590
+    },
+    {
+      "epoch": 0.8191213643490225,
+      "grad_norm": 41.0,
+      "learning_rate": 3.8086897945436174e-06,
+      "loss": 0.1584,
+      "step": 25600
+    },
+    {
+      "epoch": 0.8194413336319714,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.8019535197036047e-06,
+      "loss": 0.084,
+      "step": 25610
+    },
+    {
+      "epoch": 0.8197613029149202,
+      "grad_norm": 25.25,
+      "learning_rate": 3.795217244863591e-06,
+      "loss": 0.0917,
+      "step": 25620
+    },
+    {
+      "epoch": 0.820081272197869,
+      "grad_norm": 4.6875,
+      "learning_rate": 3.7884809700235776e-06,
+      "loss": 0.076,
+      "step": 25630
+    },
+    {
+      "epoch": 0.8204012414808178,
+      "grad_norm": 18.0,
+      "learning_rate": 3.781744695183564e-06,
+      "loss": 0.1561,
+      "step": 25640
+    },
+    {
+      "epoch": 0.8207212107637667,
+      "grad_norm": 10.375,
+      "learning_rate": 3.7750084203435505e-06,
+      "loss": 0.0642,
+      "step": 25650
+    },
+    {
+      "epoch": 0.8210411800467156,
+      "grad_norm": 21.0,
+      "learning_rate": 3.768272145503537e-06,
+      "loss": 0.0947,
+      "step": 25660
+    },
+    {
+      "epoch": 0.8213611493296643,
+      "grad_norm": 8.625,
+      "learning_rate": 3.7615358706635235e-06,
+      "loss": 0.1877,
+      "step": 25670
+    },
+    {
+      "epoch": 0.8216811186126132,
+      "grad_norm": 35.5,
+      "learning_rate": 3.75479959582351e-06,
+      "loss": 0.1457,
+      "step": 25680
+    },
+    {
+      "epoch": 0.822001087895562,
+      "grad_norm": 2.0,
+      "learning_rate": 3.7480633209834964e-06,
+      "loss": 0.062,
+      "step": 25690
+    },
+    {
+      "epoch": 0.8223210571785109,
+      "grad_norm": 10.9375,
+      "learning_rate": 3.7413270461434833e-06,
+      "loss": 0.1069,
+      "step": 25700
+    },
+    {
+      "epoch": 0.8226410264614596,
+      "grad_norm": 13.5,
+      "learning_rate": 3.7345907713034698e-06,
+      "loss": 0.0682,
+      "step": 25710
+    },
+    {
+      "epoch": 0.8229609957444085,
+      "grad_norm": 17.25,
+      "learning_rate": 3.7278544964634562e-06,
+      "loss": 0.0749,
+      "step": 25720
+    },
+    {
+      "epoch": 0.8232809650273574,
+      "grad_norm": 16.0,
+      "learning_rate": 3.7211182216234427e-06,
+      "loss": 0.0674,
+      "step": 25730
+    },
+    {
+      "epoch": 0.8236009343103062,
+      "grad_norm": 1.6875,
+      "learning_rate": 3.714381946783429e-06,
+      "loss": 0.1215,
+      "step": 25740
+    },
+    {
+      "epoch": 0.8239209035932551,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.7076456719434156e-06,
+      "loss": 0.13,
+      "step": 25750
+    },
+    {
+      "epoch": 0.8242408728762038,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.700909397103402e-06,
+      "loss": 0.0849,
+      "step": 25760
+    },
+    {
+      "epoch": 0.8245608421591527,
+      "grad_norm": 16.25,
+      "learning_rate": 3.6941731222633885e-06,
+      "loss": 0.0829,
+      "step": 25770
+    },
+    {
+      "epoch": 0.8248808114421016,
+      "grad_norm": 10.375,
+      "learning_rate": 3.687436847423375e-06,
+      "loss": 0.1161,
+      "step": 25780
+    },
+    {
+      "epoch": 0.8252007807250504,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.6807005725833615e-06,
+      "loss": 0.1281,
+      "step": 25790
+    },
+    {
+      "epoch": 0.8255207500079992,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.6739642977433484e-06,
+      "loss": 0.1703,
+      "step": 25800
+    },
+    {
+      "epoch": 0.825840719290948,
+      "grad_norm": 19.375,
+      "learning_rate": 3.667228022903335e-06,
+      "loss": 0.1104,
+      "step": 25810
+    },
+    {
+      "epoch": 0.8261606885738969,
+      "grad_norm": 1.6484375,
+      "learning_rate": 3.6604917480633213e-06,
+      "loss": 0.1361,
+      "step": 25820
+    },
+    {
+      "epoch": 0.8264806578568458,
+      "grad_norm": 13.0,
+      "learning_rate": 3.6537554732233078e-06,
+      "loss": 0.0433,
+      "step": 25830
+    },
+    {
+      "epoch": 0.8268006271397946,
+      "grad_norm": 24.75,
+      "learning_rate": 3.6470191983832942e-06,
+      "loss": 0.087,
+      "step": 25840
+    },
+    {
+      "epoch": 0.8271205964227434,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.6402829235432807e-06,
+      "loss": 0.1182,
+      "step": 25850
+    },
+    {
+      "epoch": 0.8274405657056922,
+      "grad_norm": 5.0,
+      "learning_rate": 3.633546648703267e-06,
+      "loss": 0.1062,
+      "step": 25860
+    },
+    {
+      "epoch": 0.8277605349886411,
+      "grad_norm": 12.5625,
+      "learning_rate": 3.6268103738632536e-06,
+      "loss": 0.1426,
+      "step": 25870
+    },
+    {
+      "epoch": 0.82808050427159,
+      "grad_norm": 3.625,
+      "learning_rate": 3.62007409902324e-06,
+      "loss": 0.1421,
+      "step": 25880
+    },
+    {
+      "epoch": 0.8284004735545387,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.613337824183227e-06,
+      "loss": 0.0257,
+      "step": 25890
+    },
+    {
+      "epoch": 0.8287204428374876,
+      "grad_norm": 19.125,
+      "learning_rate": 3.6066015493432134e-06,
+      "loss": 0.1668,
+      "step": 25900
+    },
+    {
+      "epoch": 0.8290404121204364,
+      "grad_norm": 8.25,
+      "learning_rate": 3.5998652745032e-06,
+      "loss": 0.1159,
+      "step": 25910
+    },
+    {
+      "epoch": 0.8293603814033853,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.5931289996631864e-06,
+      "loss": 0.0604,
+      "step": 25920
+    },
+    {
+      "epoch": 0.8296803506863342,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.5863927248231732e-06,
+      "loss": 0.1068,
+      "step": 25930
+    },
+    {
+      "epoch": 0.8300003199692829,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.5796564499831597e-06,
+      "loss": 0.1053,
+      "step": 25940
+    },
+    {
+      "epoch": 0.8303202892522318,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.572920175143146e-06,
+      "loss": 0.1022,
+      "step": 25950
+    },
+    {
+      "epoch": 0.8306402585351806,
+      "grad_norm": 14.5625,
+      "learning_rate": 3.566183900303133e-06,
+      "loss": 0.04,
+      "step": 25960
+    },
+    {
+      "epoch": 0.8309602278181295,
+      "grad_norm": 20.5,
+      "learning_rate": 3.5594476254631195e-06,
+      "loss": 0.1607,
+      "step": 25970
+    },
+    {
+      "epoch": 0.8312801971010783,
+      "grad_norm": 13.1875,
+      "learning_rate": 3.552711350623106e-06,
+      "loss": 0.1777,
+      "step": 25980
+    },
+    {
+      "epoch": 0.8316001663840271,
+      "grad_norm": 26.75,
+      "learning_rate": 3.5459750757830925e-06,
+      "loss": 0.0784,
+      "step": 25990
+    },
+    {
+      "epoch": 0.831920135666976,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.539238800943079e-06,
+      "loss": 0.0927,
+      "step": 26000
+    },
+    {
+      "epoch": 0.8322401049499248,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.5325025261030654e-06,
+      "loss": 0.1095,
+      "step": 26010
+    },
+    {
+      "epoch": 0.8325600742328736,
+      "grad_norm": 14.8125,
+      "learning_rate": 3.525766251263052e-06,
+      "loss": 0.0787,
+      "step": 26020
+    },
+    {
+      "epoch": 0.8328800435158225,
+      "grad_norm": 20.375,
+      "learning_rate": 3.5190299764230383e-06,
+      "loss": 0.0561,
+      "step": 26030
+    },
+    {
+      "epoch": 0.8332000127987713,
+      "grad_norm": 7.625,
+      "learning_rate": 3.5122937015830248e-06,
+      "loss": 0.0864,
+      "step": 26040
+    },
+    {
+      "epoch": 0.8335199820817202,
+      "grad_norm": 27.5,
+      "learning_rate": 3.5055574267430117e-06,
+      "loss": 0.1491,
+      "step": 26050
+    },
+    {
+      "epoch": 0.833839951364669,
+      "grad_norm": 10.625,
+      "learning_rate": 3.498821151902998e-06,
+      "loss": 0.0612,
+      "step": 26060
+    },
+    {
+      "epoch": 0.8341599206476178,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.4920848770629846e-06,
+      "loss": 0.1652,
+      "step": 26070
+    },
+    {
+      "epoch": 0.8344798899305667,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.485348602222971e-06,
+      "loss": 0.0912,
+      "step": 26080
+    },
+    {
+      "epoch": 0.8347998592135155,
+      "grad_norm": 38.0,
+      "learning_rate": 3.4786123273829575e-06,
+      "loss": 0.0419,
+      "step": 26090
+    },
+    {
+      "epoch": 0.8351198284964644,
+      "grad_norm": 11.125,
+      "learning_rate": 3.471876052542944e-06,
+      "loss": 0.0766,
+      "step": 26100
+    },
+    {
+      "epoch": 0.8354397977794131,
+      "grad_norm": 31.25,
+      "learning_rate": 3.4651397777029305e-06,
+      "loss": 0.121,
+      "step": 26110
+    },
+    {
+      "epoch": 0.835759767062362,
+      "grad_norm": 6.5,
+      "learning_rate": 3.458403502862917e-06,
+      "loss": 0.0931,
+      "step": 26120
+    },
+    {
+      "epoch": 0.8360797363453109,
+      "grad_norm": 14.8125,
+      "learning_rate": 3.4516672280229034e-06,
+      "loss": 0.0664,
+      "step": 26130
+    },
+    {
+      "epoch": 0.8363997056282597,
+      "grad_norm": 22.25,
+      "learning_rate": 3.44493095318289e-06,
+      "loss": 0.1157,
+      "step": 26140
+    },
+    {
+      "epoch": 0.8367196749112086,
+      "grad_norm": 4.8125,
+      "learning_rate": 3.4381946783428767e-06,
+      "loss": 0.1538,
+      "step": 26150
+    },
+    {
+      "epoch": 0.8370396441941573,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.4314584035028632e-06,
+      "loss": 0.1346,
+      "step": 26160
+    },
+    {
+      "epoch": 0.8373596134771062,
+      "grad_norm": 21.875,
+      "learning_rate": 3.4247221286628497e-06,
+      "loss": 0.1148,
+      "step": 26170
+    },
+    {
+      "epoch": 0.837679582760055,
+      "grad_norm": 17.25,
+      "learning_rate": 3.417985853822836e-06,
+      "loss": 0.1236,
+      "step": 26180
+    },
+    {
+      "epoch": 0.8379995520430039,
+      "grad_norm": 35.0,
+      "learning_rate": 3.4112495789828226e-06,
+      "loss": 0.1587,
+      "step": 26190
+    },
+    {
+      "epoch": 0.8383195213259527,
+      "grad_norm": 16.0,
+      "learning_rate": 3.404513304142809e-06,
+      "loss": 0.1405,
+      "step": 26200
+    },
+    {
+      "epoch": 0.8386394906089015,
+      "grad_norm": 13.5625,
+      "learning_rate": 3.3977770293027955e-06,
+      "loss": 0.1576,
+      "step": 26210
+    },
+    {
+      "epoch": 0.8389594598918504,
+      "grad_norm": 40.0,
+      "learning_rate": 3.391040754462782e-06,
+      "loss": 0.1091,
+      "step": 26220
+    },
+    {
+      "epoch": 0.8392794291747993,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.3843044796227685e-06,
+      "loss": 0.1204,
+      "step": 26230
+    },
+    {
+      "epoch": 0.839599398457748,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.3775682047827554e-06,
+      "loss": 0.0883,
+      "step": 26240
+    },
+    {
+      "epoch": 0.8399193677406969,
+      "grad_norm": 16.875,
+      "learning_rate": 3.370831929942742e-06,
+      "loss": 0.0634,
+      "step": 26250
+    },
+    {
+      "epoch": 0.8402393370236457,
+      "grad_norm": 12.4375,
+      "learning_rate": 3.3640956551027287e-06,
+      "loss": 0.1545,
+      "step": 26260
+    },
+    {
+      "epoch": 0.8405593063065946,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.357359380262715e-06,
+      "loss": 0.082,
+      "step": 26270
+    },
+    {
+      "epoch": 0.8408792755895435,
+      "grad_norm": 18.375,
+      "learning_rate": 3.3506231054227016e-06,
+      "loss": 0.1701,
+      "step": 26280
+    },
+    {
+      "epoch": 0.8411992448724922,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.343886830582688e-06,
+      "loss": 0.0834,
+      "step": 26290
+    },
+    {
+      "epoch": 0.8415192141554411,
+      "grad_norm": 9.25,
+      "learning_rate": 3.3371505557426746e-06,
+      "loss": 0.1042,
+      "step": 26300
+    },
+    {
+      "epoch": 0.8418391834383899,
+      "grad_norm": 10.6875,
+      "learning_rate": 3.3304142809026615e-06,
+      "loss": 0.1273,
+      "step": 26310
+    },
+    {
+      "epoch": 0.8421591527213388,
+      "grad_norm": 11.1875,
+      "learning_rate": 3.323678006062648e-06,
+      "loss": 0.0917,
+      "step": 26320
+    },
+    {
+      "epoch": 0.8424791220042875,
+      "grad_norm": 1.3125,
+      "learning_rate": 3.3169417312226344e-06,
+      "loss": 0.1099,
+      "step": 26330
+    },
+    {
+      "epoch": 0.8427990912872364,
+      "grad_norm": 7.0,
+      "learning_rate": 3.310205456382621e-06,
+      "loss": 0.0941,
+      "step": 26340
+    },
+    {
+      "epoch": 0.8431190605701853,
+      "grad_norm": 10.75,
+      "learning_rate": 3.3034691815426073e-06,
+      "loss": 0.108,
+      "step": 26350
+    },
+    {
+      "epoch": 0.8434390298531341,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.2967329067025938e-06,
+      "loss": 0.0746,
+      "step": 26360
+    },
+    {
+      "epoch": 0.843758999136083,
+      "grad_norm": 20.75,
+      "learning_rate": 3.2899966318625802e-06,
+      "loss": 0.188,
+      "step": 26370
+    },
+    {
+      "epoch": 0.8440789684190317,
+      "grad_norm": 15.0,
+      "learning_rate": 3.2832603570225667e-06,
+      "loss": 0.1475,
+      "step": 26380
+    },
+    {
+      "epoch": 0.8443989377019806,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.276524082182553e-06,
+      "loss": 0.1436,
+      "step": 26390
+    },
+    {
+      "epoch": 0.8447189069849295,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.26978780734254e-06,
+      "loss": 0.1557,
+      "step": 26400
+    },
+    {
+      "epoch": 0.8450388762678783,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.2630515325025265e-06,
+      "loss": 0.0868,
+      "step": 26410
+    },
+    {
+      "epoch": 0.8453588455508271,
+      "grad_norm": 4.75,
+      "learning_rate": 3.256315257662513e-06,
+      "loss": 0.0824,
+      "step": 26420
+    },
+    {
+      "epoch": 0.8456788148337759,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.2495789828224995e-06,
+      "loss": 0.0943,
+      "step": 26430
+    },
+    {
+      "epoch": 0.8459987841167248,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.242842707982486e-06,
+      "loss": 0.0855,
+      "step": 26440
+    },
+    {
+      "epoch": 0.8463187533996737,
+      "grad_norm": 13.8125,
+      "learning_rate": 3.2361064331424724e-06,
+      "loss": 0.1283,
+      "step": 26450
+    },
+    {
+      "epoch": 0.8466387226826224,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.229370158302459e-06,
+      "loss": 0.0688,
+      "step": 26460
+    },
+    {
+      "epoch": 0.8469586919655713,
+      "grad_norm": 5.625,
+      "learning_rate": 3.2226338834624453e-06,
+      "loss": 0.0882,
+      "step": 26470
+    },
+    {
+      "epoch": 0.8472786612485201,
+      "grad_norm": 12.375,
+      "learning_rate": 3.2158976086224318e-06,
+      "loss": 0.109,
+      "step": 26480
+    },
+    {
+      "epoch": 0.847598630531469,
+      "grad_norm": 12.125,
+      "learning_rate": 3.2091613337824182e-06,
+      "loss": 0.1799,
+      "step": 26490
+    },
+    {
+      "epoch": 0.8479185998144179,
+      "grad_norm": 14.9375,
+      "learning_rate": 3.202425058942405e-06,
+      "loss": 0.0885,
+      "step": 26500
+    },
+    {
+      "epoch": 0.8482385690973666,
+      "grad_norm": 6.625,
+      "learning_rate": 3.1956887841023916e-06,
+      "loss": 0.094,
+      "step": 26510
+    },
+    {
+      "epoch": 0.8485585383803155,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.188952509262378e-06,
+      "loss": 0.0746,
+      "step": 26520
+    },
+    {
+      "epoch": 0.8488785076632643,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.1822162344223645e-06,
+      "loss": 0.0866,
+      "step": 26530
+    },
+    {
+      "epoch": 0.8491984769462132,
+      "grad_norm": 8.5,
+      "learning_rate": 3.175479959582351e-06,
+      "loss": 0.1575,
+      "step": 26540
+    },
+    {
+      "epoch": 0.849518446229162,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.1687436847423375e-06,
+      "loss": 0.0354,
+      "step": 26550
+    },
+    {
+      "epoch": 0.8498384155121108,
+      "grad_norm": 10.0625,
+      "learning_rate": 3.162007409902324e-06,
+      "loss": 0.1632,
+      "step": 26560
+    },
+    {
+      "epoch": 0.8501583847950597,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.1552711350623104e-06,
+      "loss": 0.075,
+      "step": 26570
+    },
+    {
+      "epoch": 0.8504783540780085,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.1485348602222977e-06,
+      "loss": 0.1137,
+      "step": 26580
+    },
+    {
+      "epoch": 0.8507983233609574,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.141798585382284e-06,
+      "loss": 0.0632,
+      "step": 26590
+    },
+    {
+      "epoch": 0.8511182926439061,
+      "grad_norm": 18.5,
+      "learning_rate": 3.1350623105422706e-06,
+      "loss": 0.1179,
+      "step": 26600
+    },
+    {
+      "epoch": 0.851438261926855,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.128326035702257e-06,
+      "loss": 0.1367,
+      "step": 26610
+    },
+    {
+      "epoch": 0.8517582312098039,
+      "grad_norm": 23.75,
+      "learning_rate": 3.1215897608622436e-06,
+      "loss": 0.1312,
+      "step": 26620
+    },
+    {
+      "epoch": 0.8520782004927527,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.11485348602223e-06,
+      "loss": 0.1576,
+      "step": 26630
+    },
+    {
+      "epoch": 0.8523981697757015,
+      "grad_norm": 16.5,
+      "learning_rate": 3.1081172111822165e-06,
+      "loss": 0.1373,
+      "step": 26640
+    },
+    {
+      "epoch": 0.8527181390586503,
+      "grad_norm": 22.125,
+      "learning_rate": 3.101380936342203e-06,
+      "loss": 0.1213,
+      "step": 26650
+    },
+    {
+      "epoch": 0.8530381083415992,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 3.09464466150219e-06,
+      "loss": 0.0798,
+      "step": 26660
+    },
+    {
+      "epoch": 0.8533580776245481,
+      "grad_norm": 29.125,
+      "learning_rate": 3.0879083866621763e-06,
+      "loss": 0.1044,
+      "step": 26670
+    },
+    {
+      "epoch": 0.8536780469074969,
+      "grad_norm": 10.1875,
+      "learning_rate": 3.0811721118221628e-06,
+      "loss": 0.1142,
+      "step": 26680
+    },
+    {
+      "epoch": 0.8539980161904457,
+      "grad_norm": 12.125,
+      "learning_rate": 3.0744358369821492e-06,
+      "loss": 0.0464,
+      "step": 26690
+    },
+    {
+      "epoch": 0.8543179854733945,
+      "grad_norm": 9.0,
+      "learning_rate": 3.0676995621421357e-06,
+      "loss": 0.0788,
+      "step": 26700
+    },
+    {
+      "epoch": 0.8546379547563434,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.060963287302122e-06,
+      "loss": 0.1067,
+      "step": 26710
+    },
+    {
+      "epoch": 0.8549579240392923,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.0542270124621086e-06,
+      "loss": 0.161,
+      "step": 26720
+    },
+    {
+      "epoch": 0.855277893322241,
+      "grad_norm": 14.6875,
+      "learning_rate": 3.047490737622095e-06,
+      "loss": 0.1488,
+      "step": 26730
+    },
+    {
+      "epoch": 0.8555978626051899,
+      "grad_norm": 16.5,
+      "learning_rate": 3.0407544627820816e-06,
+      "loss": 0.0899,
+      "step": 26740
+    },
+    {
+      "epoch": 0.8559178318881387,
+      "grad_norm": 12.0,
+      "learning_rate": 3.0340181879420684e-06,
+      "loss": 0.1283,
+      "step": 26750
+    },
+    {
+      "epoch": 0.8562378011710876,
+      "grad_norm": 13.8125,
+      "learning_rate": 3.027281913102055e-06,
+      "loss": 0.1142,
+      "step": 26760
+    },
+    {
+      "epoch": 0.8565577704540364,
+      "grad_norm": 10.75,
+      "learning_rate": 3.0205456382620414e-06,
+      "loss": 0.1071,
+      "step": 26770
+    },
+    {
+      "epoch": 0.8568777397369852,
+      "grad_norm": 15.1875,
+      "learning_rate": 3.013809363422028e-06,
+      "loss": 0.1298,
+      "step": 26780
+    },
+    {
+      "epoch": 0.8571977090199341,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.0070730885820143e-06,
+      "loss": 0.081,
+      "step": 26790
+    },
+    {
+      "epoch": 0.857517678302883,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.0003368137420008e-06,
+      "loss": 0.1733,
+      "step": 26800
+    },
+    {
+      "epoch": 0.8578376475858318,
+      "grad_norm": 4.59375,
+      "learning_rate": 2.9936005389019872e-06,
+      "loss": 0.0857,
+      "step": 26810
+    },
+    {
+      "epoch": 0.8581576168687806,
+      "grad_norm": 18.125,
+      "learning_rate": 2.9868642640619737e-06,
+      "loss": 0.087,
+      "step": 26820
+    },
+    {
+      "epoch": 0.8584775861517294,
+      "grad_norm": 10.0625,
+      "learning_rate": 2.98012798922196e-06,
+      "loss": 0.143,
+      "step": 26830
+    },
+    {
+      "epoch": 0.8587975554346783,
+      "grad_norm": 12.0,
+      "learning_rate": 2.973391714381947e-06,
+      "loss": 0.1119,
+      "step": 26840
+    },
+    {
+      "epoch": 0.8591175247176271,
+      "grad_norm": 1.328125,
+      "learning_rate": 2.9666554395419335e-06,
+      "loss": 0.1347,
+      "step": 26850
+    },
+    {
+      "epoch": 0.8594374940005759,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.95991916470192e-06,
+      "loss": 0.1421,
+      "step": 26860
+    },
+    {
+      "epoch": 0.8597574632835248,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.9531828898619065e-06,
+      "loss": 0.0878,
+      "step": 26870
+    },
+    {
+      "epoch": 0.8600774325664736,
+      "grad_norm": 11.5625,
+      "learning_rate": 2.946446615021893e-06,
+      "loss": 0.1361,
+      "step": 26880
+    },
+    {
+      "epoch": 0.8603974018494225,
+      "grad_norm": 2.0,
+      "learning_rate": 2.9397103401818794e-06,
+      "loss": 0.0791,
+      "step": 26890
+    },
+    {
+      "epoch": 0.8607173711323713,
+      "grad_norm": 18.375,
+      "learning_rate": 2.9329740653418663e-06,
+      "loss": 0.1408,
+      "step": 26900
+    },
+    {
+      "epoch": 0.8610373404153201,
+      "grad_norm": 32.75,
+      "learning_rate": 2.926237790501853e-06,
+      "loss": 0.2028,
+      "step": 26910
+    },
+    {
+      "epoch": 0.861357309698269,
+      "grad_norm": 13.6875,
+      "learning_rate": 2.9195015156618396e-06,
+      "loss": 0.1018,
+      "step": 26920
+    },
+    {
+      "epoch": 0.8616772789812178,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.912765240821826e-06,
+      "loss": 0.1583,
+      "step": 26930
+    },
+    {
+      "epoch": 0.8619972482641667,
+      "grad_norm": 5.3125,
+      "learning_rate": 2.9060289659818126e-06,
+      "loss": 0.1422,
+      "step": 26940
+    },
+    {
+      "epoch": 0.8623172175471154,
+      "grad_norm": 10.375,
+      "learning_rate": 2.899292691141799e-06,
+      "loss": 0.1404,
+      "step": 26950
+    },
+    {
+      "epoch": 0.8626371868300643,
+      "grad_norm": 5.78125,
+      "learning_rate": 2.8925564163017855e-06,
+      "loss": 0.1187,
+      "step": 26960
+    },
+    {
+      "epoch": 0.8629571561130132,
+      "grad_norm": 10.4375,
+      "learning_rate": 2.885820141461772e-06,
+      "loss": 0.0961,
+      "step": 26970
+    },
+    {
+      "epoch": 0.863277125395962,
+      "grad_norm": 14.125,
+      "learning_rate": 2.8790838666217584e-06,
+      "loss": 0.0927,
+      "step": 26980
+    },
+    {
+      "epoch": 0.8635970946789108,
+      "grad_norm": 20.25,
+      "learning_rate": 2.872347591781745e-06,
+      "loss": 0.1489,
+      "step": 26990
+    },
+    {
+      "epoch": 0.8639170639618596,
+      "grad_norm": 11.6875,
+      "learning_rate": 2.8656113169417318e-06,
+      "loss": 0.1408,
+      "step": 27000
+    },
+    {
+      "epoch": 0.8642370332448085,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.8588750421017182e-06,
+      "loss": 0.081,
+      "step": 27010
+    },
+    {
+      "epoch": 0.8645570025277574,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.8521387672617047e-06,
+      "loss": 0.1263,
+      "step": 27020
+    },
+    {
+      "epoch": 0.8648769718107062,
+      "grad_norm": 16.25,
+      "learning_rate": 2.845402492421691e-06,
+      "loss": 0.1127,
+      "step": 27030
+    },
+    {
+      "epoch": 0.865196941093655,
+      "grad_norm": 16.25,
+      "learning_rate": 2.8386662175816776e-06,
+      "loss": 0.2059,
+      "step": 27040
+    },
+    {
+      "epoch": 0.8655169103766038,
+      "grad_norm": 18.875,
+      "learning_rate": 2.831929942741664e-06,
+      "loss": 0.1622,
+      "step": 27050
+    },
+    {
+      "epoch": 0.8658368796595527,
+      "grad_norm": 5.4375,
+      "learning_rate": 2.8251936679016506e-06,
+      "loss": 0.1294,
+      "step": 27060
+    },
+    {
+      "epoch": 0.8661568489425016,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.818457393061637e-06,
+      "loss": 0.1177,
+      "step": 27070
+    },
+    {
+      "epoch": 0.8664768182254503,
+      "grad_norm": 9.875,
+      "learning_rate": 2.8117211182216235e-06,
+      "loss": 0.1083,
+      "step": 27080
+    },
+    {
+      "epoch": 0.8667967875083992,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.80498484338161e-06,
+      "loss": 0.097,
+      "step": 27090
+    },
+    {
+      "epoch": 0.867116756791348,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.798248568541597e-06,
+      "loss": 0.1104,
+      "step": 27100
+    },
+    {
+      "epoch": 0.8674367260742969,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.7915122937015833e-06,
+      "loss": 0.0396,
+      "step": 27110
+    },
+    {
+      "epoch": 0.8677566953572458,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.7847760188615698e-06,
+      "loss": 0.1441,
+      "step": 27120
+    },
+    {
+      "epoch": 0.8680766646401945,
+      "grad_norm": 65.0,
+      "learning_rate": 2.7780397440215562e-06,
+      "loss": 0.1234,
+      "step": 27130
+    },
+    {
+      "epoch": 0.8683966339231434,
+      "grad_norm": 5.75,
+      "learning_rate": 2.7713034691815427e-06,
+      "loss": 0.0783,
+      "step": 27140
+    },
+    {
+      "epoch": 0.8687166032060922,
+      "grad_norm": 5.78125,
+      "learning_rate": 2.764567194341529e-06,
+      "loss": 0.1072,
+      "step": 27150
+    },
+    {
+      "epoch": 0.8690365724890411,
+      "grad_norm": 10.375,
+      "learning_rate": 2.7578309195015156e-06,
+      "loss": 0.0951,
+      "step": 27160
+    },
+    {
+      "epoch": 0.8693565417719898,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.751094644661502e-06,
+      "loss": 0.0661,
+      "step": 27170
+    },
+    {
+      "epoch": 0.8696765110549387,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.7443583698214886e-06,
+      "loss": 0.0462,
+      "step": 27180
+    },
+    {
+      "epoch": 0.8699964803378876,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.7376220949814754e-06,
+      "loss": 0.1362,
+      "step": 27190
+    },
+    {
+      "epoch": 0.8703164496208364,
+      "grad_norm": 10.8125,
+      "learning_rate": 2.730885820141462e-06,
+      "loss": 0.1367,
+      "step": 27200
+    },
+    {
+      "epoch": 0.8706364189037853,
+      "grad_norm": 11.3125,
+      "learning_rate": 2.7241495453014484e-06,
+      "loss": 0.099,
+      "step": 27210
+    },
+    {
+      "epoch": 0.870956388186734,
+      "grad_norm": 1.34375,
+      "learning_rate": 2.717413270461435e-06,
+      "loss": 0.0747,
+      "step": 27220
+    },
+    {
+      "epoch": 0.8712763574696829,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.7106769956214217e-06,
+      "loss": 0.1146,
+      "step": 27230
+    },
+    {
+      "epoch": 0.8715963267526318,
+      "grad_norm": 11.3125,
+      "learning_rate": 2.703940720781408e-06,
+      "loss": 0.1006,
+      "step": 27240
+    },
+    {
+      "epoch": 0.8719162960355806,
+      "grad_norm": 5.0625,
+      "learning_rate": 2.6972044459413947e-06,
+      "loss": 0.0773,
+      "step": 27250
+    },
+    {
+      "epoch": 0.8722362653185294,
+      "grad_norm": 10.375,
+      "learning_rate": 2.6904681711013815e-06,
+      "loss": 0.1592,
+      "step": 27260
+    },
+    {
+      "epoch": 0.8725562346014782,
+      "grad_norm": 19.375,
+      "learning_rate": 2.683731896261368e-06,
+      "loss": 0.1435,
+      "step": 27270
+    },
+    {
+      "epoch": 0.8728762038844271,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.6769956214213545e-06,
+      "loss": 0.1256,
+      "step": 27280
+    },
+    {
+      "epoch": 0.873196173167376,
+      "grad_norm": 20.375,
+      "learning_rate": 2.670259346581341e-06,
+      "loss": 0.1737,
+      "step": 27290
+    },
+    {
+      "epoch": 0.8735161424503247,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.6635230717413274e-06,
+      "loss": 0.1213,
+      "step": 27300
+    },
+    {
+      "epoch": 0.8738361117332736,
+      "grad_norm": 10.5,
+      "learning_rate": 2.656786796901314e-06,
+      "loss": 0.1409,
+      "step": 27310
+    },
+    {
+      "epoch": 0.8741560810162224,
+      "grad_norm": 16.75,
+      "learning_rate": 2.6500505220613003e-06,
+      "loss": 0.0825,
+      "step": 27320
+    },
+    {
+      "epoch": 0.8744760502991713,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.643314247221287e-06,
+      "loss": 0.1038,
+      "step": 27330
+    },
+    {
+      "epoch": 0.8747960195821202,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.6365779723812733e-06,
+      "loss": 0.0745,
+      "step": 27340
+    },
+    {
+      "epoch": 0.8751159888650689,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.62984169754126e-06,
+      "loss": 0.0833,
+      "step": 27350
+    },
+    {
+      "epoch": 0.8754359581480178,
+      "grad_norm": 12.6875,
+      "learning_rate": 2.6231054227012466e-06,
+      "loss": 0.0746,
+      "step": 27360
+    },
+    {
+      "epoch": 0.8757559274309666,
+      "grad_norm": 18.25,
+      "learning_rate": 2.616369147861233e-06,
+      "loss": 0.1527,
+      "step": 27370
+    },
+    {
+      "epoch": 0.8760758967139155,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.6096328730212195e-06,
+      "loss": 0.1121,
+      "step": 27380
+    },
+    {
+      "epoch": 0.8763958659968643,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.602896598181206e-06,
+      "loss": 0.1009,
+      "step": 27390
+    },
+    {
+      "epoch": 0.8767158352798131,
+      "grad_norm": 4.9375,
+      "learning_rate": 2.5961603233411925e-06,
+      "loss": 0.0946,
+      "step": 27400
+    },
+    {
+      "epoch": 0.877035804562762,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.589424048501179e-06,
+      "loss": 0.0971,
+      "step": 27410
+    },
+    {
+      "epoch": 0.8773557738457108,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.5826877736611654e-06,
+      "loss": 0.0833,
+      "step": 27420
+    },
+    {
+      "epoch": 0.8776757431286597,
+      "grad_norm": 10.875,
+      "learning_rate": 2.575951498821152e-06,
+      "loss": 0.1689,
+      "step": 27430
+    },
+    {
+      "epoch": 0.8779957124116085,
+      "grad_norm": 14.75,
+      "learning_rate": 2.5692152239811383e-06,
+      "loss": 0.1001,
+      "step": 27440
+    },
+    {
+      "epoch": 0.8783156816945573,
+      "grad_norm": 6.125,
+      "learning_rate": 2.5624789491411252e-06,
+      "loss": 0.0838,
+      "step": 27450
+    },
+    {
+      "epoch": 0.8786356509775062,
+      "grad_norm": 22.375,
+      "learning_rate": 2.5557426743011117e-06,
+      "loss": 0.1145,
+      "step": 27460
+    },
+    {
+      "epoch": 0.878955620260455,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.549006399461098e-06,
+      "loss": 0.0925,
+      "step": 27470
+    },
+    {
+      "epoch": 0.8792755895434038,
+      "grad_norm": 5.25,
+      "learning_rate": 2.5422701246210846e-06,
+      "loss": 0.1017,
+      "step": 27480
+    },
+    {
+      "epoch": 0.8795955588263527,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.535533849781071e-06,
+      "loss": 0.0581,
+      "step": 27490
+    },
+    {
+      "epoch": 0.8799155281093015,
+      "grad_norm": 26.125,
+      "learning_rate": 2.5287975749410575e-06,
+      "loss": 0.162,
+      "step": 27500
+    },
+    {
+      "epoch": 0.8802354973922504,
+      "grad_norm": 5.71875,
+      "learning_rate": 2.522061300101044e-06,
+      "loss": 0.1068,
+      "step": 27510
+    },
+    {
+      "epoch": 0.8805554666751991,
+      "grad_norm": 14.125,
+      "learning_rate": 2.5153250252610305e-06,
+      "loss": 0.145,
+      "step": 27520
+    },
+    {
+      "epoch": 0.880875435958148,
+      "grad_norm": 34.25,
+      "learning_rate": 2.508588750421017e-06,
+      "loss": 0.1379,
+      "step": 27530
+    },
+    {
+      "epoch": 0.8811954052410969,
+      "grad_norm": 6.25,
+      "learning_rate": 2.501852475581004e-06,
+      "loss": 0.0572,
+      "step": 27540
+    },
+    {
+      "epoch": 0.8815153745240457,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.4951162007409903e-06,
+      "loss": 0.1623,
+      "step": 27550
+    },
+    {
+      "epoch": 0.8818353438069946,
+      "grad_norm": 9.5625,
+      "learning_rate": 2.4883799259009768e-06,
+      "loss": 0.1115,
+      "step": 27560
+    },
+    {
+      "epoch": 0.8821553130899433,
+      "grad_norm": 13.375,
+      "learning_rate": 2.4816436510609636e-06,
+      "loss": 0.1245,
+      "step": 27570
+    },
+    {
+      "epoch": 0.8824752823728922,
+      "grad_norm": 20.625,
+      "learning_rate": 2.47490737622095e-06,
+      "loss": 0.141,
+      "step": 27580
+    },
+    {
+      "epoch": 0.882795251655841,
+      "grad_norm": 10.6875,
+      "learning_rate": 2.4681711013809366e-06,
+      "loss": 0.077,
+      "step": 27590
+    },
+    {
+      "epoch": 0.8831152209387899,
+      "grad_norm": 13.375,
+      "learning_rate": 2.461434826540923e-06,
+      "loss": 0.0968,
+      "step": 27600
+    },
+    {
+      "epoch": 0.8834351902217387,
+      "grad_norm": 23.25,
+      "learning_rate": 2.4546985517009095e-06,
+      "loss": 0.0824,
+      "step": 27610
+    },
+    {
+      "epoch": 0.8837551595046875,
+      "grad_norm": 11.5,
+      "learning_rate": 2.447962276860896e-06,
+      "loss": 0.146,
+      "step": 27620
+    },
+    {
+      "epoch": 0.8840751287876364,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.441226002020883e-06,
+      "loss": 0.0869,
+      "step": 27630
+    },
+    {
+      "epoch": 0.8843950980705853,
+      "grad_norm": 5.4375,
+      "learning_rate": 2.4344897271808693e-06,
+      "loss": 0.1064,
+      "step": 27640
+    },
+    {
+      "epoch": 0.8847150673535341,
+      "grad_norm": 16.5,
+      "learning_rate": 2.427753452340856e-06,
+      "loss": 0.1236,
+      "step": 27650
+    },
+    {
+      "epoch": 0.8850350366364829,
+      "grad_norm": 15.1875,
+      "learning_rate": 2.4210171775008423e-06,
+      "loss": 0.0924,
+      "step": 27660
+    },
+    {
+      "epoch": 0.8853550059194317,
+      "grad_norm": 1.5,
+      "learning_rate": 2.4142809026608287e-06,
+      "loss": 0.137,
+      "step": 27670
+    },
+    {
+      "epoch": 0.8856749752023806,
+      "grad_norm": 11.25,
+      "learning_rate": 2.407544627820815e-06,
+      "loss": 0.1282,
+      "step": 27680
+    },
+    {
+      "epoch": 0.8859949444853294,
+      "grad_norm": 7.375,
+      "learning_rate": 2.4008083529808017e-06,
+      "loss": 0.1409,
+      "step": 27690
+    },
+    {
+      "epoch": 0.8863149137682782,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.3940720781407885e-06,
+      "loss": 0.1107,
+      "step": 27700
+    },
+    {
+      "epoch": 0.8866348830512271,
+      "grad_norm": 4.625,
+      "learning_rate": 2.387335803300775e-06,
+      "loss": 0.1221,
+      "step": 27710
+    },
+    {
+      "epoch": 0.8869548523341759,
+      "grad_norm": 5.125,
+      "learning_rate": 2.3805995284607615e-06,
+      "loss": 0.0751,
+      "step": 27720
+    },
+    {
+      "epoch": 0.8872748216171248,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.373863253620748e-06,
+      "loss": 0.1011,
+      "step": 27730
+    },
+    {
+      "epoch": 0.8875947909000735,
+      "grad_norm": 16.125,
+      "learning_rate": 2.3671269787807344e-06,
+      "loss": 0.137,
+      "step": 27740
+    },
+    {
+      "epoch": 0.8879147601830224,
+      "grad_norm": 22.125,
+      "learning_rate": 2.360390703940721e-06,
+      "loss": 0.1568,
+      "step": 27750
+    },
+    {
+      "epoch": 0.8882347294659713,
+      "grad_norm": 4.53125,
+      "learning_rate": 2.3536544291007073e-06,
+      "loss": 0.0755,
+      "step": 27760
+    },
+    {
+      "epoch": 0.8885546987489201,
+      "grad_norm": 10.375,
+      "learning_rate": 2.346918154260694e-06,
+      "loss": 0.1248,
+      "step": 27770
+    },
+    {
+      "epoch": 0.888874668031869,
+      "grad_norm": 5.875,
+      "learning_rate": 2.3401818794206803e-06,
+      "loss": 0.0736,
+      "step": 27780
+    },
+    {
+      "epoch": 0.8891946373148177,
+      "grad_norm": 16.125,
+      "learning_rate": 2.333445604580667e-06,
+      "loss": 0.121,
+      "step": 27790
+    },
+    {
+      "epoch": 0.8895146065977666,
+      "grad_norm": 16.125,
+      "learning_rate": 2.3267093297406536e-06,
+      "loss": 0.208,
+      "step": 27800
+    },
+    {
+      "epoch": 0.8898345758807155,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.31997305490064e-06,
+      "loss": 0.0572,
+      "step": 27810
+    },
+    {
+      "epoch": 0.8901545451636643,
+      "grad_norm": 5.53125,
+      "learning_rate": 2.3132367800606265e-06,
+      "loss": 0.1395,
+      "step": 27820
+    },
+    {
+      "epoch": 0.8904745144466131,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.3065005052206134e-06,
+      "loss": 0.1638,
+      "step": 27830
+    },
+    {
+      "epoch": 0.8907944837295619,
+      "grad_norm": 5.71875,
+      "learning_rate": 2.2997642303806e-06,
+      "loss": 0.1019,
+      "step": 27840
+    },
+    {
+      "epoch": 0.8911144530125108,
+      "grad_norm": 9.75,
+      "learning_rate": 2.2930279555405864e-06,
+      "loss": 0.1173,
+      "step": 27850
+    },
+    {
+      "epoch": 0.8914344222954597,
+      "grad_norm": 11.75,
+      "learning_rate": 2.286291680700573e-06,
+      "loss": 0.0744,
+      "step": 27860
+    },
+    {
+      "epoch": 0.8917543915784085,
+      "grad_norm": 9.1875,
+      "learning_rate": 2.2795554058605593e-06,
+      "loss": 0.1343,
+      "step": 27870
+    },
+    {
+      "epoch": 0.8920743608613573,
+      "grad_norm": 5.28125,
+      "learning_rate": 2.2728191310205458e-06,
+      "loss": 0.1461,
+      "step": 27880
+    },
+    {
+      "epoch": 0.8923943301443061,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.2660828561805322e-06,
+      "loss": 0.0907,
+      "step": 27890
+    },
+    {
+      "epoch": 0.892714299427255,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.2593465813405187e-06,
+      "loss": 0.1232,
+      "step": 27900
+    },
+    {
+      "epoch": 0.8930342687102039,
+      "grad_norm": 20.25,
+      "learning_rate": 2.252610306500505e-06,
+      "loss": 0.1363,
+      "step": 27910
+    },
+    {
+      "epoch": 0.8933542379931526,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.245874031660492e-06,
+      "loss": 0.154,
+      "step": 27920
+    },
+    {
+      "epoch": 0.8936742072761015,
+      "grad_norm": 4.8125,
+      "learning_rate": 2.2391377568204785e-06,
+      "loss": 0.1451,
+      "step": 27930
+    },
+    {
+      "epoch": 0.8939941765590503,
+      "grad_norm": 40.25,
+      "learning_rate": 2.232401481980465e-06,
+      "loss": 0.2074,
+      "step": 27940
+    },
+    {
+      "epoch": 0.8943141458419992,
+      "grad_norm": 9.6875,
+      "learning_rate": 2.225665207140452e-06,
+      "loss": 0.1017,
+      "step": 27950
+    },
+    {
+      "epoch": 0.8946341151249481,
+      "grad_norm": 18.0,
+      "learning_rate": 2.2189289323004383e-06,
+      "loss": 0.1106,
+      "step": 27960
+    },
+    {
+      "epoch": 0.8949540844078968,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.2121926574604248e-06,
+      "loss": 0.0851,
+      "step": 27970
+    },
+    {
+      "epoch": 0.8952740536908457,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.2054563826204112e-06,
+      "loss": 0.112,
+      "step": 27980
+    },
+    {
+      "epoch": 0.8955940229737945,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.1987201077803977e-06,
+      "loss": 0.0813,
+      "step": 27990
+    },
+    {
+      "epoch": 0.8959139922567434,
+      "grad_norm": 23.875,
+      "learning_rate": 2.191983832940384e-06,
+      "loss": 0.1408,
+      "step": 28000
+    },
+    {
+      "epoch": 0.8962339615396921,
+      "grad_norm": 14.625,
+      "learning_rate": 2.1852475581003706e-06,
+      "loss": 0.1109,
+      "step": 28010
+    },
+    {
+      "epoch": 0.896553930822641,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.178511283260357e-06,
+      "loss": 0.0799,
+      "step": 28020
+    },
+    {
+      "epoch": 0.8968739001055899,
+      "grad_norm": 6.21875,
+      "learning_rate": 2.1717750084203436e-06,
+      "loss": 0.1101,
+      "step": 28030
+    },
+    {
+      "epoch": 0.8971938693885387,
+      "grad_norm": 13.4375,
+      "learning_rate": 2.16503873358033e-06,
+      "loss": 0.1114,
+      "step": 28040
+    },
+    {
+      "epoch": 0.8975138386714875,
+      "grad_norm": 19.5,
+      "learning_rate": 2.158302458740317e-06,
+      "loss": 0.0906,
+      "step": 28050
+    },
+    {
+      "epoch": 0.8978338079544363,
+      "grad_norm": 4.53125,
+      "learning_rate": 2.1515661839003034e-06,
+      "loss": 0.1327,
+      "step": 28060
+    },
+    {
+      "epoch": 0.8981537772373852,
+      "grad_norm": 23.625,
+      "learning_rate": 2.14482990906029e-06,
+      "loss": 0.1127,
+      "step": 28070
+    },
+    {
+      "epoch": 0.8984737465203341,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.1380936342202763e-06,
+      "loss": 0.0846,
+      "step": 28080
+    },
+    {
+      "epoch": 0.8987937158032829,
+      "grad_norm": 13.9375,
+      "learning_rate": 2.1313573593802628e-06,
+      "loss": 0.0603,
+      "step": 28090
+    },
+    {
+      "epoch": 0.8991136850862317,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.1246210845402493e-06,
+      "loss": 0.1015,
+      "step": 28100
+    },
+    {
+      "epoch": 0.8994336543691805,
+      "grad_norm": 32.25,
+      "learning_rate": 2.117884809700236e-06,
+      "loss": 0.0892,
+      "step": 28110
+    },
+    {
+      "epoch": 0.8997536236521294,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.1111485348602226e-06,
+      "loss": 0.0981,
+      "step": 28120
+    },
+    {
+      "epoch": 0.9000735929350783,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.104412260020209e-06,
+      "loss": 0.1675,
+      "step": 28130
+    },
+    {
+      "epoch": 0.900393562218027,
+      "grad_norm": 10.0,
+      "learning_rate": 2.0976759851801955e-06,
+      "loss": 0.1065,
+      "step": 28140
+    },
+    {
+      "epoch": 0.9007135315009759,
+      "grad_norm": 12.625,
+      "learning_rate": 2.090939710340182e-06,
+      "loss": 0.1469,
+      "step": 28150
+    },
+    {
+      "epoch": 0.9010335007839247,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.0842034355001685e-06,
+      "loss": 0.0984,
+      "step": 28160
+    },
+    {
+      "epoch": 0.9013534700668736,
+      "grad_norm": 10.1875,
+      "learning_rate": 2.077467160660155e-06,
+      "loss": 0.0784,
+      "step": 28170
+    },
+    {
+      "epoch": 0.9016734393498225,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.070730885820142e-06,
+      "loss": 0.1035,
+      "step": 28180
+    },
+    {
+      "epoch": 0.9019934086327712,
+      "grad_norm": 20.125,
+      "learning_rate": 2.0639946109801283e-06,
+      "loss": 0.1184,
+      "step": 28190
+    },
+    {
+      "epoch": 0.9023133779157201,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.0572583361401147e-06,
+      "loss": 0.1011,
+      "step": 28200
+    },
+    {
+      "epoch": 0.902633347198669,
+      "grad_norm": 9.125,
+      "learning_rate": 2.050522061300101e-06,
+      "loss": 0.0984,
+      "step": 28210
+    },
+    {
+      "epoch": 0.9029533164816178,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.0437857864600877e-06,
+      "loss": 0.1193,
+      "step": 28220
+    },
+    {
+      "epoch": 0.9032732857645666,
+      "grad_norm": 9.8125,
+      "learning_rate": 2.037049511620074e-06,
+      "loss": 0.1004,
+      "step": 28230
+    },
+    {
+      "epoch": 0.9035932550475154,
+      "grad_norm": 19.0,
+      "learning_rate": 2.0303132367800606e-06,
+      "loss": 0.1256,
+      "step": 28240
+    },
+    {
+      "epoch": 0.9039132243304643,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.023576961940047e-06,
+      "loss": 0.1225,
+      "step": 28250
+    },
+    {
+      "epoch": 0.9042331936134131,
+      "grad_norm": 8.875,
+      "learning_rate": 2.0168406871000335e-06,
+      "loss": 0.0965,
+      "step": 28260
+    },
+    {
+      "epoch": 0.9045531628963619,
+      "grad_norm": 5.59375,
+      "learning_rate": 2.0101044122600204e-06,
+      "loss": 0.0995,
+      "step": 28270
+    },
+    {
+      "epoch": 0.9048731321793108,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.003368137420007e-06,
+      "loss": 0.0994,
+      "step": 28280
+    },
+    {
+      "epoch": 0.9051931014622596,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.9966318625799934e-06,
+      "loss": 0.1405,
+      "step": 28290
+    },
+    {
+      "epoch": 0.9055130707452085,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.9898955877399802e-06,
+      "loss": 0.1157,
+      "step": 28300
+    },
+    {
+      "epoch": 0.9058330400281573,
+      "grad_norm": 123.0,
+      "learning_rate": 1.9831593128999667e-06,
+      "loss": 0.1767,
+      "step": 28310
+    },
+    {
+      "epoch": 0.9061530093111061,
+      "grad_norm": 8.75,
+      "learning_rate": 1.976423038059953e-06,
+      "loss": 0.1001,
+      "step": 28320
+    },
+    {
+      "epoch": 0.906472978594055,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.9696867632199396e-06,
+      "loss": 0.1721,
+      "step": 28330
+    },
+    {
+      "epoch": 0.9067929478770038,
+      "grad_norm": 3.625,
+      "learning_rate": 1.962950488379926e-06,
+      "loss": 0.0725,
+      "step": 28340
+    },
+    {
+      "epoch": 0.9071129171599527,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9562142135399126e-06,
+      "loss": 0.1292,
+      "step": 28350
+    },
+    {
+      "epoch": 0.9074328864429014,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.949477938699899e-06,
+      "loss": 0.0912,
+      "step": 28360
+    },
+    {
+      "epoch": 0.9077528557258503,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9427416638598855e-06,
+      "loss": 0.1139,
+      "step": 28370
+    },
+    {
+      "epoch": 0.9080728250087992,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.936005389019872e-06,
+      "loss": 0.1651,
+      "step": 28380
+    },
+    {
+      "epoch": 0.908392794291748,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.9292691141798584e-06,
+      "loss": 0.1372,
+      "step": 28390
+    },
+    {
+      "epoch": 0.9087127635746969,
+      "grad_norm": 27.875,
+      "learning_rate": 1.9225328393398453e-06,
+      "loss": 0.1306,
+      "step": 28400
+    },
+    {
+      "epoch": 0.9090327328576456,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.9157965644998318e-06,
+      "loss": 0.0986,
+      "step": 28410
+    },
+    {
+      "epoch": 0.9093527021405945,
+      "grad_norm": 23.25,
+      "learning_rate": 1.9090602896598182e-06,
+      "loss": 0.1372,
+      "step": 28420
+    },
+    {
+      "epoch": 0.9096726714235434,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9023240148198047e-06,
+      "loss": 0.1003,
+      "step": 28430
+    },
+    {
+      "epoch": 0.9099926407064922,
+      "grad_norm": 4.9375,
+      "learning_rate": 1.8955877399797914e-06,
+      "loss": 0.0856,
+      "step": 28440
+    },
+    {
+      "epoch": 0.910312609989441,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.888851465139778e-06,
+      "loss": 0.1232,
+      "step": 28450
+    },
+    {
+      "epoch": 0.9106325792723898,
+      "grad_norm": 4.90625,
+      "learning_rate": 1.8821151902997645e-06,
+      "loss": 0.1138,
+      "step": 28460
+    },
+    {
+      "epoch": 0.9109525485553387,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.875378915459751e-06,
+      "loss": 0.11,
+      "step": 28470
+    },
+    {
+      "epoch": 0.9112725178382876,
+      "grad_norm": 9.75,
+      "learning_rate": 1.8686426406197375e-06,
+      "loss": 0.0547,
+      "step": 28480
+    },
+    {
+      "epoch": 0.9115924871212364,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.861906365779724e-06,
+      "loss": 0.1488,
+      "step": 28490
+    },
+    {
+      "epoch": 0.9119124564041852,
+      "grad_norm": 12.625,
+      "learning_rate": 1.8551700909397106e-06,
+      "loss": 0.1034,
+      "step": 28500
+    },
+    {
+      "epoch": 0.912232425687134,
+      "grad_norm": 24.625,
+      "learning_rate": 1.848433816099697e-06,
+      "loss": 0.0859,
+      "step": 28510
+    },
+    {
+      "epoch": 0.9125523949700829,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.8416975412596835e-06,
+      "loss": 0.1189,
+      "step": 28520
+    },
+    {
+      "epoch": 0.9128723642530318,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.83496126641967e-06,
+      "loss": 0.0758,
+      "step": 28530
+    },
+    {
+      "epoch": 0.9131923335359805,
+      "grad_norm": 13.125,
+      "learning_rate": 1.8282249915796565e-06,
+      "loss": 0.0856,
+      "step": 28540
+    },
+    {
+      "epoch": 0.9135123028189294,
+      "grad_norm": 6.3125,
+      "learning_rate": 1.8214887167396431e-06,
+      "loss": 0.1443,
+      "step": 28550
+    },
+    {
+      "epoch": 0.9138322721018782,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.8147524418996296e-06,
+      "loss": 0.1287,
+      "step": 28560
+    },
+    {
+      "epoch": 0.9141522413848271,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.808016167059616e-06,
+      "loss": 0.1398,
+      "step": 28570
+    },
+    {
+      "epoch": 0.9144722106677758,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8012798922196025e-06,
+      "loss": 0.1299,
+      "step": 28580
+    },
+    {
+      "epoch": 0.9147921799507247,
+      "grad_norm": 22.125,
+      "learning_rate": 1.794543617379589e-06,
+      "loss": 0.1137,
+      "step": 28590
+    },
+    {
+      "epoch": 0.9151121492336736,
+      "grad_norm": 6.75,
+      "learning_rate": 1.7878073425395759e-06,
+      "loss": 0.1472,
+      "step": 28600
+    },
+    {
+      "epoch": 0.9154321185166224,
+      "grad_norm": 9.25,
+      "learning_rate": 1.7810710676995623e-06,
+      "loss": 0.1124,
+      "step": 28610
+    },
+    {
+      "epoch": 0.9157520877995713,
+      "grad_norm": 17.75,
+      "learning_rate": 1.7743347928595488e-06,
+      "loss": 0.1162,
+      "step": 28620
+    },
+    {
+      "epoch": 0.91607205708252,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7675985180195355e-06,
+      "loss": 0.0355,
+      "step": 28630
+    },
+    {
+      "epoch": 0.9163920263654689,
+      "grad_norm": 40.0,
+      "learning_rate": 1.760862243179522e-06,
+      "loss": 0.1198,
+      "step": 28640
+    },
+    {
+      "epoch": 0.9167119956484178,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7541259683395084e-06,
+      "loss": 0.0996,
+      "step": 28650
+    },
+    {
+      "epoch": 0.9170319649313666,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.7473896934994949e-06,
+      "loss": 0.0999,
+      "step": 28660
+    },
+    {
+      "epoch": 0.9173519342143154,
+      "grad_norm": 4.625,
+      "learning_rate": 1.7406534186594813e-06,
+      "loss": 0.1017,
+      "step": 28670
+    },
+    {
+      "epoch": 0.9176719034972642,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.733917143819468e-06,
+      "loss": 0.0565,
+      "step": 28680
+    },
+    {
+      "epoch": 0.9179918727802131,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7271808689794545e-06,
+      "loss": 0.0843,
+      "step": 28690
+    },
+    {
+      "epoch": 0.918311842063162,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.720444594139441e-06,
+      "loss": 0.138,
+      "step": 28700
+    },
+    {
+      "epoch": 0.9186318113461108,
+      "grad_norm": 4.75,
+      "learning_rate": 1.7137083192994274e-06,
+      "loss": 0.0942,
+      "step": 28710
+    },
+    {
+      "epoch": 0.9189517806290596,
+      "grad_norm": 6.25,
+      "learning_rate": 1.706972044459414e-06,
+      "loss": 0.1002,
+      "step": 28720
+    },
+    {
+      "epoch": 0.9192717499120084,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7002357696194006e-06,
+      "loss": 0.0911,
+      "step": 28730
+    },
+    {
+      "epoch": 0.9195917191949573,
+      "grad_norm": 24.5,
+      "learning_rate": 1.693499494779387e-06,
+      "loss": 0.1303,
+      "step": 28740
+    },
+    {
+      "epoch": 0.9199116884779062,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.6867632199393735e-06,
+      "loss": 0.1017,
+      "step": 28750
+    },
+    {
+      "epoch": 0.9202316577608549,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.6800269450993604e-06,
+      "loss": 0.0624,
+      "step": 28760
+    },
+    {
+      "epoch": 0.9205516270438038,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.6732906702593468e-06,
+      "loss": 0.1414,
+      "step": 28770
+    },
+    {
+      "epoch": 0.9208715963267526,
+      "grad_norm": 31.5,
+      "learning_rate": 1.6665543954193333e-06,
+      "loss": 0.087,
+      "step": 28780
+    },
+    {
+      "epoch": 0.9211915656097015,
+      "grad_norm": 13.25,
+      "learning_rate": 1.6598181205793198e-06,
+      "loss": 0.125,
+      "step": 28790
+    },
+    {
+      "epoch": 0.9215115348926503,
+      "grad_norm": 23.625,
+      "learning_rate": 1.6530818457393064e-06,
+      "loss": 0.1184,
+      "step": 28800
+    },
+    {
+      "epoch": 0.9218315041755991,
+      "grad_norm": 8.75,
+      "learning_rate": 1.646345570899293e-06,
+      "loss": 0.0559,
+      "step": 28810
+    },
+    {
+      "epoch": 0.922151473458548,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.6396092960592794e-06,
+      "loss": 0.0866,
+      "step": 28820
+    },
+    {
+      "epoch": 0.9224714427414968,
+      "grad_norm": 31.875,
+      "learning_rate": 1.6328730212192658e-06,
+      "loss": 0.1563,
+      "step": 28830
+    },
+    {
+      "epoch": 0.9227914120244457,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.6261367463792523e-06,
+      "loss": 0.0692,
+      "step": 28840
+    },
+    {
+      "epoch": 0.9231113813073945,
+      "grad_norm": 20.625,
+      "learning_rate": 1.619400471539239e-06,
+      "loss": 0.1056,
+      "step": 28850
+    },
+    {
+      "epoch": 0.9234313505903433,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.6126641966992254e-06,
+      "loss": 0.0678,
+      "step": 28860
+    },
+    {
+      "epoch": 0.9237513198732922,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.605927921859212e-06,
+      "loss": 0.1029,
+      "step": 28870
+    },
+    {
+      "epoch": 0.924071289156241,
+      "grad_norm": 12.875,
+      "learning_rate": 1.5991916470191984e-06,
+      "loss": 0.0639,
+      "step": 28880
+    },
+    {
+      "epoch": 0.9243912584391898,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.5924553721791848e-06,
+      "loss": 0.0996,
+      "step": 28890
+    },
+    {
+      "epoch": 0.9247112277221387,
+      "grad_norm": 21.75,
+      "learning_rate": 1.5857190973391715e-06,
+      "loss": 0.1423,
+      "step": 28900
+    },
+    {
+      "epoch": 0.9250311970050875,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.578982822499158e-06,
+      "loss": 0.0792,
+      "step": 28910
+    },
+    {
+      "epoch": 0.9253511662880364,
+      "grad_norm": 6.75,
+      "learning_rate": 1.5722465476591447e-06,
+      "loss": 0.0905,
+      "step": 28920
+    },
+    {
+      "epoch": 0.9256711355709852,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.5655102728191313e-06,
+      "loss": 0.0832,
+      "step": 28930
+    },
+    {
+      "epoch": 0.925991104853934,
+      "grad_norm": 11.875,
+      "learning_rate": 1.5587739979791178e-06,
+      "loss": 0.0703,
+      "step": 28940
+    },
+    {
+      "epoch": 0.9263110741368829,
+      "grad_norm": 59.25,
+      "learning_rate": 1.5520377231391043e-06,
+      "loss": 0.1043,
+      "step": 28950
+    },
+    {
+      "epoch": 0.9266310434198317,
+      "grad_norm": 4.46875,
+      "learning_rate": 1.5453014482990907e-06,
+      "loss": 0.0946,
+      "step": 28960
+    },
+    {
+      "epoch": 0.9269510127027806,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.5385651734590772e-06,
+      "loss": 0.0877,
+      "step": 28970
+    },
+    {
+      "epoch": 0.9272709819857293,
+      "grad_norm": 5.78125,
+      "learning_rate": 1.5318288986190639e-06,
+      "loss": 0.159,
+      "step": 28980
+    },
+    {
+      "epoch": 0.9275909512686782,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.5250926237790503e-06,
+      "loss": 0.0843,
+      "step": 28990
+    },
+    {
+      "epoch": 0.927910920551627,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.5183563489390368e-06,
+      "loss": 0.1604,
+      "step": 29000
+    },
+    {
+      "epoch": 0.9282308898345759,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.5116200740990233e-06,
+      "loss": 0.0657,
+      "step": 29010
+    },
+    {
+      "epoch": 0.9285508591175247,
+      "grad_norm": 8.0,
+      "learning_rate": 1.50488379925901e-06,
+      "loss": 0.108,
+      "step": 29020
+    },
+    {
+      "epoch": 0.9288708284004735,
+      "grad_norm": 16.875,
+      "learning_rate": 1.4981475244189964e-06,
+      "loss": 0.1829,
+      "step": 29030
+    },
+    {
+      "epoch": 0.9291907976834224,
+      "grad_norm": 6.4375,
+      "learning_rate": 1.4914112495789829e-06,
+      "loss": 0.0916,
+      "step": 29040
+    },
+    {
+      "epoch": 0.9295107669663712,
+      "grad_norm": 14.625,
+      "learning_rate": 1.4846749747389693e-06,
+      "loss": 0.115,
+      "step": 29050
+    },
+    {
+      "epoch": 0.9298307362493201,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.4779386998989558e-06,
+      "loss": 0.1315,
+      "step": 29060
+    },
+    {
+      "epoch": 0.9301507055322689,
+      "grad_norm": 20.625,
+      "learning_rate": 1.4712024250589425e-06,
+      "loss": 0.1696,
+      "step": 29070
+    },
+    {
+      "epoch": 0.9304706748152177,
+      "grad_norm": 63.5,
+      "learning_rate": 1.4644661502189292e-06,
+      "loss": 0.172,
+      "step": 29080
+    },
+    {
+      "epoch": 0.9307906440981666,
+      "grad_norm": 26.125,
+      "learning_rate": 1.4577298753789156e-06,
+      "loss": 0.1579,
+      "step": 29090
+    },
+    {
+      "epoch": 0.9311106133811154,
+      "grad_norm": 25.375,
+      "learning_rate": 1.4509936005389023e-06,
+      "loss": 0.1369,
+      "step": 29100
+    },
+    {
+      "epoch": 0.9314305826640642,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.4442573256988888e-06,
+      "loss": 0.1342,
+      "step": 29110
+    },
+    {
+      "epoch": 0.9317505519470131,
+      "grad_norm": 11.125,
+      "learning_rate": 1.4375210508588752e-06,
+      "loss": 0.0699,
+      "step": 29120
+    },
+    {
+      "epoch": 0.9320705212299619,
+      "grad_norm": 5.59375,
+      "learning_rate": 1.4307847760188617e-06,
+      "loss": 0.1644,
+      "step": 29130
+    },
+    {
+      "epoch": 0.9323904905129108,
+      "grad_norm": 17.375,
+      "learning_rate": 1.4240485011788482e-06,
+      "loss": 0.1777,
+      "step": 29140
+    },
+    {
+      "epoch": 0.9327104597958596,
+      "grad_norm": 25.875,
+      "learning_rate": 1.4173122263388348e-06,
+      "loss": 0.1518,
+      "step": 29150
+    },
+    {
+      "epoch": 0.9330304290788084,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.4105759514988213e-06,
+      "loss": 0.1291,
+      "step": 29160
+    },
+    {
+      "epoch": 0.9333503983617573,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.4038396766588078e-06,
+      "loss": 0.0807,
+      "step": 29170
+    },
+    {
+      "epoch": 0.9336703676447061,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3971034018187942e-06,
+      "loss": 0.0914,
+      "step": 29180
+    },
+    {
+      "epoch": 0.933990336927655,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.3903671269787807e-06,
+      "loss": 0.1007,
+      "step": 29190
+    },
+    {
+      "epoch": 0.9343103062106037,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.3836308521387674e-06,
+      "loss": 0.1125,
+      "step": 29200
+    },
+    {
+      "epoch": 0.9346302754935526,
+      "grad_norm": 22.875,
+      "learning_rate": 1.3768945772987538e-06,
+      "loss": 0.0648,
+      "step": 29210
+    },
+    {
+      "epoch": 0.9349502447765015,
+      "grad_norm": 5.46875,
+      "learning_rate": 1.3701583024587403e-06,
+      "loss": 0.0528,
+      "step": 29220
+    },
+    {
+      "epoch": 0.9352702140594503,
+      "grad_norm": 10.875,
+      "learning_rate": 1.3634220276187268e-06,
+      "loss": 0.1367,
+      "step": 29230
+    },
+    {
+      "epoch": 0.9355901833423992,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.3566857527787137e-06,
+      "loss": 0.0562,
+      "step": 29240
+    },
+    {
+      "epoch": 0.9359101526253479,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.3499494779387001e-06,
+      "loss": 0.0813,
+      "step": 29250
+    },
+    {
+      "epoch": 0.9362301219082968,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.3432132030986866e-06,
+      "loss": 0.1542,
+      "step": 29260
+    },
+    {
+      "epoch": 0.9365500911912457,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.336476928258673e-06,
+      "loss": 0.1144,
+      "step": 29270
+    },
+    {
+      "epoch": 0.9368700604741945,
+      "grad_norm": 25.0,
+      "learning_rate": 1.3297406534186597e-06,
+      "loss": 0.1216,
+      "step": 29280
+    },
+    {
+      "epoch": 0.9371900297571433,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.3230043785786462e-06,
+      "loss": 0.0869,
+      "step": 29290
+    },
+    {
+      "epoch": 0.9375099990400921,
+      "grad_norm": 12.625,
+      "learning_rate": 1.3162681037386327e-06,
+      "loss": 0.1606,
+      "step": 29300
+    },
+    {
+      "epoch": 0.937829968323041,
+      "grad_norm": 6.09375,
+      "learning_rate": 1.3095318288986191e-06,
+      "loss": 0.1795,
+      "step": 29310
+    },
+    {
+      "epoch": 0.9381499376059899,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.3027955540586058e-06,
+      "loss": 0.0819,
+      "step": 29320
+    },
+    {
+      "epoch": 0.9384699068889386,
+      "grad_norm": 22.375,
+      "learning_rate": 1.2960592792185923e-06,
+      "loss": 0.1286,
+      "step": 29330
+    },
+    {
+      "epoch": 0.9387898761718875,
+      "grad_norm": 17.875,
+      "learning_rate": 1.2893230043785787e-06,
+      "loss": 0.1253,
+      "step": 29340
+    },
+    {
+      "epoch": 0.9391098454548363,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2825867295385652e-06,
+      "loss": 0.047,
+      "step": 29350
+    },
+    {
+      "epoch": 0.9394298147377852,
+      "grad_norm": 19.375,
+      "learning_rate": 1.2758504546985517e-06,
+      "loss": 0.0671,
+      "step": 29360
+    },
+    {
+      "epoch": 0.9397497840207341,
+      "grad_norm": 18.125,
+      "learning_rate": 1.2691141798585383e-06,
+      "loss": 0.1547,
+      "step": 29370
+    },
+    {
+      "epoch": 0.9400697533036828,
+      "grad_norm": 12.0,
+      "learning_rate": 1.2623779050185248e-06,
+      "loss": 0.0987,
+      "step": 29380
+    },
+    {
+      "epoch": 0.9403897225866317,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.2556416301785113e-06,
+      "loss": 0.1345,
+      "step": 29390
+    },
+    {
+      "epoch": 0.9407096918695805,
+      "grad_norm": 8.5,
+      "learning_rate": 1.248905355338498e-06,
+      "loss": 0.0626,
+      "step": 29400
+    },
+    {
+      "epoch": 0.9410296611525294,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.2421690804984844e-06,
+      "loss": 0.0972,
+      "step": 29410
+    },
+    {
+      "epoch": 0.9413496304354781,
+      "grad_norm": 16.875,
+      "learning_rate": 1.2354328056584709e-06,
+      "loss": 0.1044,
+      "step": 29420
+    },
+    {
+      "epoch": 0.941669599718427,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.2286965308184573e-06,
+      "loss": 0.1222,
+      "step": 29430
+    },
+    {
+      "epoch": 0.9419895690013759,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.221960255978444e-06,
+      "loss": 0.1749,
+      "step": 29440
+    },
+    {
+      "epoch": 0.9423095382843247,
+      "grad_norm": 18.875,
+      "learning_rate": 1.2152239811384307e-06,
+      "loss": 0.1283,
+      "step": 29450
+    },
+    {
+      "epoch": 0.9426295075672736,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.2084877062984172e-06,
+      "loss": 0.0836,
+      "step": 29460
+    },
+    {
+      "epoch": 0.9429494768502223,
+      "grad_norm": 5.03125,
+      "learning_rate": 1.2017514314584036e-06,
+      "loss": 0.1439,
+      "step": 29470
+    },
+    {
+      "epoch": 0.9432694461331712,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.19501515661839e-06,
+      "loss": 0.0865,
+      "step": 29480
+    },
+    {
+      "epoch": 0.9435894154161201,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.1882788817783765e-06,
+      "loss": 0.1332,
+      "step": 29490
+    },
+    {
+      "epoch": 0.9439093846990689,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.1815426069383632e-06,
+      "loss": 0.1292,
+      "step": 29500
+    },
+    {
+      "epoch": 0.9442293539820177,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.1748063320983497e-06,
+      "loss": 0.1111,
+      "step": 29510
+    },
+    {
+      "epoch": 0.9445493232649665,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.1680700572583364e-06,
+      "loss": 0.0876,
+      "step": 29520
+    },
+    {
+      "epoch": 0.9448692925479154,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.1613337824183228e-06,
+      "loss": 0.0458,
+      "step": 29530
+    },
+    {
+      "epoch": 0.9451892618308643,
+      "grad_norm": 39.75,
+      "learning_rate": 1.1545975075783093e-06,
+      "loss": 0.1522,
+      "step": 29540
+    },
+    {
+      "epoch": 0.945509231113813,
+      "grad_norm": 5.875,
+      "learning_rate": 1.1478612327382958e-06,
+      "loss": 0.1341,
+      "step": 29550
+    },
+    {
+      "epoch": 0.9458292003967619,
+      "grad_norm": 31.875,
+      "learning_rate": 1.1411249578982824e-06,
+      "loss": 0.1406,
+      "step": 29560
+    },
+    {
+      "epoch": 0.9461491696797107,
+      "grad_norm": 2.75,
+      "learning_rate": 1.134388683058269e-06,
+      "loss": 0.1329,
+      "step": 29570
+    },
+    {
+      "epoch": 0.9464691389626596,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.1276524082182554e-06,
+      "loss": 0.0694,
+      "step": 29580
+    },
+    {
+      "epoch": 0.9467891082456085,
+      "grad_norm": 11.625,
+      "learning_rate": 1.1209161333782418e-06,
+      "loss": 0.1565,
+      "step": 29590
+    },
+    {
+      "epoch": 0.9471090775285572,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.1141798585382283e-06,
+      "loss": 0.0868,
+      "step": 29600
+    },
+    {
+      "epoch": 0.9474290468115061,
+      "grad_norm": 9.5,
+      "learning_rate": 1.107443583698215e-06,
+      "loss": 0.1018,
+      "step": 29610
+    },
+    {
+      "epoch": 0.9477490160944549,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.1007073088582014e-06,
+      "loss": 0.0625,
+      "step": 29620
+    },
+    {
+      "epoch": 0.9480689853774038,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.0939710340181881e-06,
+      "loss": 0.1285,
+      "step": 29630
+    },
+    {
+      "epoch": 0.9483889546603526,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.0872347591781746e-06,
+      "loss": 0.0586,
+      "step": 29640
+    },
+    {
+      "epoch": 0.9487089239433014,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.080498484338161e-06,
+      "loss": 0.049,
+      "step": 29650
+    },
+    {
+      "epoch": 0.9490288932262503,
+      "grad_norm": 24.125,
+      "learning_rate": 1.0737622094981475e-06,
+      "loss": 0.0841,
+      "step": 29660
+    },
+    {
+      "epoch": 0.9493488625091991,
+      "grad_norm": 16.875,
+      "learning_rate": 1.0670259346581342e-06,
+      "loss": 0.0813,
+      "step": 29670
+    },
+    {
+      "epoch": 0.949668831792148,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0602896598181206e-06,
+      "loss": 0.171,
+      "step": 29680
+    },
+    {
+      "epoch": 0.9499888010750968,
+      "grad_norm": 15.875,
+      "learning_rate": 1.0535533849781073e-06,
+      "loss": 0.1438,
+      "step": 29690
+    },
+    {
+      "epoch": 0.9503087703580456,
+      "grad_norm": 5.375,
+      "learning_rate": 1.0468171101380938e-06,
+      "loss": 0.1315,
+      "step": 29700
+    },
+    {
+      "epoch": 0.9506287396409945,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.0400808352980803e-06,
+      "loss": 0.0772,
+      "step": 29710
+    },
+    {
+      "epoch": 0.9509487089239433,
+      "grad_norm": 10.375,
+      "learning_rate": 1.0333445604580667e-06,
+      "loss": 0.0884,
+      "step": 29720
+    },
+    {
+      "epoch": 0.9512686782068921,
+      "grad_norm": 8.5,
+      "learning_rate": 1.0266082856180532e-06,
+      "loss": 0.0834,
+      "step": 29730
+    },
+    {
+      "epoch": 0.951588647489841,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.0198720107780399e-06,
+      "loss": 0.1144,
+      "step": 29740
+    },
+    {
+      "epoch": 0.9519086167727898,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.0131357359380263e-06,
+      "loss": 0.1354,
+      "step": 29750
+    },
+    {
+      "epoch": 0.9522285860557387,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0063994610980128e-06,
+      "loss": 0.1206,
+      "step": 29760
+    },
+    {
+      "epoch": 0.9525485553386875,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.996631862579995e-07,
+      "loss": 0.1125,
+      "step": 29770
+    },
+    {
+      "epoch": 0.9528685246216363,
+      "grad_norm": 7.625,
+      "learning_rate": 9.92926911417986e-07,
+      "loss": 0.1352,
+      "step": 29780
+    },
+    {
+      "epoch": 0.9531884939045852,
+      "grad_norm": 31.5,
+      "learning_rate": 9.861906365779724e-07,
+      "loss": 0.0936,
+      "step": 29790
+    },
+    {
+      "epoch": 0.953508463187534,
+      "grad_norm": 8.75,
+      "learning_rate": 9.79454361737959e-07,
+      "loss": 0.0617,
+      "step": 29800
+    },
+    {
+      "epoch": 0.9538284324704829,
+      "grad_norm": 17.125,
+      "learning_rate": 9.727180868979455e-07,
+      "loss": 0.078,
+      "step": 29810
+    },
+    {
+      "epoch": 0.9541484017534316,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.65981812057932e-07,
+      "loss": 0.0936,
+      "step": 29820
+    },
+    {
+      "epoch": 0.9544683710363805,
+      "grad_norm": 7.25,
+      "learning_rate": 9.592455372179185e-07,
+      "loss": 0.1529,
+      "step": 29830
+    },
+    {
+      "epoch": 0.9547883403193294,
+      "grad_norm": 9.375,
+      "learning_rate": 9.52509262377905e-07,
+      "loss": 0.0909,
+      "step": 29840
+    },
+    {
+      "epoch": 0.9551083096022782,
+      "grad_norm": 0.625,
+      "learning_rate": 9.457729875378917e-07,
+      "loss": 0.1272,
+      "step": 29850
+    },
+    {
+      "epoch": 0.955428278885227,
+      "grad_norm": 7.9375,
+      "learning_rate": 9.390367126978782e-07,
+      "loss": 0.0567,
+      "step": 29860
+    },
+    {
+      "epoch": 0.9557482481681758,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.323004378578646e-07,
+      "loss": 0.1161,
+      "step": 29870
+    },
+    {
+      "epoch": 0.9560682174511247,
+      "grad_norm": 16.25,
+      "learning_rate": 9.255641630178512e-07,
+      "loss": 0.1593,
+      "step": 29880
+    },
+    {
+      "epoch": 0.9563881867340736,
+      "grad_norm": 16.875,
+      "learning_rate": 9.188278881778377e-07,
+      "loss": 0.142,
+      "step": 29890
+    },
+    {
+      "epoch": 0.9567081560170224,
+      "grad_norm": 18.875,
+      "learning_rate": 9.120916133378243e-07,
+      "loss": 0.1522,
+      "step": 29900
+    },
+    {
+      "epoch": 0.9570281252999712,
+      "grad_norm": 1.9140625,
+      "learning_rate": 9.053553384978107e-07,
+      "loss": 0.1324,
+      "step": 29910
+    },
+    {
+      "epoch": 0.95734809458292,
+      "grad_norm": 6.53125,
+      "learning_rate": 8.986190636577973e-07,
+      "loss": 0.0888,
+      "step": 29920
+    },
+    {
+      "epoch": 0.9576680638658689,
+      "grad_norm": 36.0,
+      "learning_rate": 8.918827888177839e-07,
+      "loss": 0.2007,
+      "step": 29930
+    },
+    {
+      "epoch": 0.9579880331488178,
+      "grad_norm": 12.3125,
+      "learning_rate": 8.851465139777704e-07,
+      "loss": 0.0833,
+      "step": 29940
+    },
+    {
+      "epoch": 0.9583080024317665,
+      "grad_norm": 7.40625,
+      "learning_rate": 8.784102391377569e-07,
+      "loss": 0.1167,
+      "step": 29950
+    },
+    {
+      "epoch": 0.9586279717147154,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.716739642977435e-07,
+      "loss": 0.1334,
+      "step": 29960
+    },
+    {
+      "epoch": 0.9589479409976642,
+      "grad_norm": 15.0,
+      "learning_rate": 8.649376894577299e-07,
+      "loss": 0.0692,
+      "step": 29970
+    },
+    {
+      "epoch": 0.9592679102806131,
+      "grad_norm": 10.3125,
+      "learning_rate": 8.582014146177164e-07,
+      "loss": 0.1986,
+      "step": 29980
+    },
+    {
+      "epoch": 0.959587879563562,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.51465139777703e-07,
+      "loss": 0.0679,
+      "step": 29990
+    },
+    {
+      "epoch": 0.9599078488465107,
+      "grad_norm": 9.4375,
+      "learning_rate": 8.447288649376894e-07,
+      "loss": 0.1193,
+      "step": 30000
+    },
+    {
+      "epoch": 0.9602278181294596,
+      "grad_norm": 11.375,
+      "learning_rate": 8.379925900976761e-07,
+      "loss": 0.0968,
+      "step": 30010
+    },
+    {
+      "epoch": 0.9605477874124084,
+      "grad_norm": 10.9375,
+      "learning_rate": 8.312563152576626e-07,
+      "loss": 0.1227,
+      "step": 30020
+    },
+    {
+      "epoch": 0.9608677566953573,
+      "grad_norm": 18.125,
+      "learning_rate": 8.245200404176491e-07,
+      "loss": 0.1113,
+      "step": 30030
+    },
+    {
+      "epoch": 0.961187725978306,
+      "grad_norm": 18.125,
+      "learning_rate": 8.177837655776356e-07,
+      "loss": 0.1165,
+      "step": 30040
+    },
+    {
+      "epoch": 0.9615076952612549,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.110474907376222e-07,
+      "loss": 0.0833,
+      "step": 30050
+    },
+    {
+      "epoch": 0.9618276645442038,
+      "grad_norm": 6.84375,
+      "learning_rate": 8.043112158976086e-07,
+      "loss": 0.0866,
+      "step": 30060
+    },
+    {
+      "epoch": 0.9621476338271526,
+      "grad_norm": 6.71875,
+      "learning_rate": 7.975749410575952e-07,
+      "loss": 0.1212,
+      "step": 30070
+    },
+    {
+      "epoch": 0.9624676031101014,
+      "grad_norm": 10.75,
+      "learning_rate": 7.908386662175817e-07,
+      "loss": 0.1781,
+      "step": 30080
+    },
+    {
+      "epoch": 0.9627875723930502,
+      "grad_norm": 3.859375,
+      "learning_rate": 7.841023913775684e-07,
+      "loss": 0.1305,
+      "step": 30090
+    },
+    {
+      "epoch": 0.9631075416759991,
+      "grad_norm": 18.125,
+      "learning_rate": 7.773661165375548e-07,
+      "loss": 0.1295,
+      "step": 30100
+    },
+    {
+      "epoch": 0.963427510958948,
+      "grad_norm": 16.375,
+      "learning_rate": 7.706298416975414e-07,
+      "loss": 0.1152,
+      "step": 30110
+    },
+    {
+      "epoch": 0.9637474802418968,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.638935668575279e-07,
+      "loss": 0.137,
+      "step": 30120
+    },
+    {
+      "epoch": 0.9640674495248456,
+      "grad_norm": 7.28125,
+      "learning_rate": 7.571572920175143e-07,
+      "loss": 0.1233,
+      "step": 30130
+    },
+    {
+      "epoch": 0.9643874188077944,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.504210171775009e-07,
+      "loss": 0.129,
+      "step": 30140
+    },
+    {
+      "epoch": 0.9647073880907433,
+      "grad_norm": 9.8125,
+      "learning_rate": 7.436847423374874e-07,
+      "loss": 0.1409,
+      "step": 30150
+    },
+    {
+      "epoch": 0.9650273573736922,
+      "grad_norm": 13.75,
+      "learning_rate": 7.369484674974739e-07,
+      "loss": 0.082,
+      "step": 30160
+    },
+    {
+      "epoch": 0.9653473266566409,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.302121926574605e-07,
+      "loss": 0.0944,
+      "step": 30170
+    },
+    {
+      "epoch": 0.9656672959395898,
+      "grad_norm": 6.9375,
+      "learning_rate": 7.234759178174471e-07,
+      "loss": 0.0378,
+      "step": 30180
+    },
+    {
+      "epoch": 0.9659872652225386,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.167396429774335e-07,
+      "loss": 0.1357,
+      "step": 30190
+    },
+    {
+      "epoch": 0.9663072345054875,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.100033681374201e-07,
+      "loss": 0.1785,
+      "step": 30200
+    },
+    {
+      "epoch": 0.9666272037884364,
+      "grad_norm": 10.25,
+      "learning_rate": 7.032670932974066e-07,
+      "loss": 0.0899,
+      "step": 30210
+    },
+    {
+      "epoch": 0.9669471730713851,
+      "grad_norm": 7.5625,
+      "learning_rate": 6.965308184573931e-07,
+      "loss": 0.1482,
+      "step": 30220
+    },
+    {
+      "epoch": 0.967267142354334,
+      "grad_norm": 8.25,
+      "learning_rate": 6.897945436173796e-07,
+      "loss": 0.107,
+      "step": 30230
+    },
+    {
+      "epoch": 0.9675871116372828,
+      "grad_norm": 5.6875,
+      "learning_rate": 6.830582687773661e-07,
+      "loss": 0.0703,
+      "step": 30240
+    },
+    {
+      "epoch": 0.9679070809202317,
+      "grad_norm": 18.0,
+      "learning_rate": 6.763219939373527e-07,
+      "loss": 0.1616,
+      "step": 30250
+    },
+    {
+      "epoch": 0.9682270502031805,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.695857190973393e-07,
+      "loss": 0.1073,
+      "step": 30260
+    },
+    {
+      "epoch": 0.9685470194861293,
+      "grad_norm": 6.6875,
+      "learning_rate": 6.628494442573258e-07,
+      "loss": 0.1452,
+      "step": 30270
+    },
+    {
+      "epoch": 0.9688669887690782,
+      "grad_norm": 14.4375,
+      "learning_rate": 6.561131694173122e-07,
+      "loss": 0.2076,
+      "step": 30280
+    },
+    {
+      "epoch": 0.969186958052027,
+      "grad_norm": 14.5,
+      "learning_rate": 6.493768945772988e-07,
+      "loss": 0.1903,
+      "step": 30290
+    },
+    {
+      "epoch": 0.9695069273349758,
+      "grad_norm": 16.5,
+      "learning_rate": 6.426406197372853e-07,
+      "loss": 0.1945,
+      "step": 30300
+    },
+    {
+      "epoch": 0.9698268966179246,
+      "grad_norm": 8.875,
+      "learning_rate": 6.359043448972719e-07,
+      "loss": 0.1507,
+      "step": 30310
+    },
+    {
+      "epoch": 0.9701468659008735,
+      "grad_norm": 11.6875,
+      "learning_rate": 6.291680700572583e-07,
+      "loss": 0.1027,
+      "step": 30320
+    },
+    {
+      "epoch": 0.9704668351838224,
+      "grad_norm": 14.125,
+      "learning_rate": 6.224317952172449e-07,
+      "loss": 0.1336,
+      "step": 30330
+    },
+    {
+      "epoch": 0.9707868044667712,
+      "grad_norm": 16.875,
+      "learning_rate": 6.156955203772315e-07,
+      "loss": 0.0553,
+      "step": 30340
+    },
+    {
+      "epoch": 0.97110677374972,
+      "grad_norm": 20.625,
+      "learning_rate": 6.08959245537218e-07,
+      "loss": 0.0632,
+      "step": 30350
+    },
+    {
+      "epoch": 0.9714267430326688,
+      "grad_norm": 13.5,
+      "learning_rate": 6.022229706972045e-07,
+      "loss": 0.1034,
+      "step": 30360
+    },
+    {
+      "epoch": 0.9717467123156177,
+      "grad_norm": 15.0625,
+      "learning_rate": 5.954866958571911e-07,
+      "loss": 0.0737,
+      "step": 30370
+    },
+    {
+      "epoch": 0.9720666815985666,
+      "grad_norm": 24.375,
+      "learning_rate": 5.887504210171775e-07,
+      "loss": 0.1024,
+      "step": 30380
+    },
+    {
+      "epoch": 0.9723866508815153,
+      "grad_norm": 4.96875,
+      "learning_rate": 5.820141461771641e-07,
+      "loss": 0.1472,
+      "step": 30390
+    },
+    {
+      "epoch": 0.9727066201644642,
+      "grad_norm": 15.25,
+      "learning_rate": 5.752778713371506e-07,
+      "loss": 0.091,
+      "step": 30400
+    },
+    {
+      "epoch": 0.973026589447413,
+      "grad_norm": 30.875,
+      "learning_rate": 5.685415964971371e-07,
+      "loss": 0.105,
+      "step": 30410
+    },
+    {
+      "epoch": 0.9733465587303619,
+      "grad_norm": 0.5,
+      "learning_rate": 5.618053216571236e-07,
+      "loss": 0.1065,
+      "step": 30420
+    },
+    {
+      "epoch": 0.9736665280133108,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.550690468171102e-07,
+      "loss": 0.1332,
+      "step": 30430
+    },
+    {
+      "epoch": 0.9739864972962595,
+      "grad_norm": 40.5,
+      "learning_rate": 5.483327719770967e-07,
+      "loss": 0.1038,
+      "step": 30440
+    },
+    {
+      "epoch": 0.9743064665792084,
+      "grad_norm": 10.4375,
+      "learning_rate": 5.415964971370832e-07,
+      "loss": 0.1251,
+      "step": 30450
+    },
+    {
+      "epoch": 0.9746264358621572,
+      "grad_norm": 19.875,
+      "learning_rate": 5.348602222970698e-07,
+      "loss": 0.1756,
+      "step": 30460
+    },
+    {
+      "epoch": 0.9749464051451061,
+      "grad_norm": 15.9375,
+      "learning_rate": 5.281239474570563e-07,
+      "loss": 0.0858,
+      "step": 30470
+    },
+    {
+      "epoch": 0.9752663744280549,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.213876726170428e-07,
+      "loss": 0.0836,
+      "step": 30480
+    },
+    {
+      "epoch": 0.9755863437110037,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.146513977770294e-07,
+      "loss": 0.0787,
+      "step": 30490
+    },
+    {
+      "epoch": 0.9759063129939526,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.079151229370158e-07,
+      "loss": 0.1419,
+      "step": 30500
+    },
+    {
+      "epoch": 0.9762262822769014,
+      "grad_norm": 1.8203125,
+      "learning_rate": 5.011788480970024e-07,
+      "loss": 0.1126,
+      "step": 30510
+    },
+    {
+      "epoch": 0.9765462515598503,
+      "grad_norm": 6.0,
+      "learning_rate": 4.94442573256989e-07,
+      "loss": 0.0274,
+      "step": 30520
+    },
+    {
+      "epoch": 0.9768662208427991,
+      "grad_norm": 29.625,
+      "learning_rate": 4.877062984169755e-07,
+      "loss": 0.1164,
+      "step": 30530
+    },
+    {
+      "epoch": 0.9771861901257479,
+      "grad_norm": 54.75,
+      "learning_rate": 4.809700235769619e-07,
+      "loss": 0.0588,
+      "step": 30540
+    },
+    {
+      "epoch": 0.9775061594086968,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.7423374873694854e-07,
+      "loss": 0.1148,
+      "step": 30550
+    },
+    {
+      "epoch": 0.9778261286916456,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.6749747389693506e-07,
+      "loss": 0.1387,
+      "step": 30560
+    },
+    {
+      "epoch": 0.9781460979745944,
+      "grad_norm": 17.0,
+      "learning_rate": 4.607611990569216e-07,
+      "loss": 0.1025,
+      "step": 30570
+    },
+    {
+      "epoch": 0.9784660672575433,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.5402492421690804e-07,
+      "loss": 0.1132,
+      "step": 30580
+    },
+    {
+      "epoch": 0.9787860365404921,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.4728864937689467e-07,
+      "loss": 0.0996,
+      "step": 30590
+    },
+    {
+      "epoch": 0.979106005823441,
+      "grad_norm": 21.625,
+      "learning_rate": 4.4055237453688113e-07,
+      "loss": 0.1536,
+      "step": 30600
+    },
+    {
+      "epoch": 0.9794259751063897,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.3381609969686765e-07,
+      "loss": 0.1974,
+      "step": 30610
+    },
+    {
+      "epoch": 0.9797459443893386,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.2707982485685417e-07,
+      "loss": 0.1202,
+      "step": 30620
+    },
+    {
+      "epoch": 0.9800659136722875,
+      "grad_norm": 22.375,
+      "learning_rate": 4.2034355001684074e-07,
+      "loss": 0.1091,
+      "step": 30630
+    },
+    {
+      "epoch": 0.9803858829552363,
+      "grad_norm": 9.625,
+      "learning_rate": 4.1360727517682726e-07,
+      "loss": 0.0747,
+      "step": 30640
+    },
+    {
+      "epoch": 0.9807058522381852,
+      "grad_norm": 25.5,
+      "learning_rate": 4.068710003368138e-07,
+      "loss": 0.102,
+      "step": 30650
+    },
+    {
+      "epoch": 0.9810258215211339,
+      "grad_norm": 5.625,
+      "learning_rate": 4.001347254968003e-07,
+      "loss": 0.1312,
+      "step": 30660
+    },
+    {
+      "epoch": 0.9813457908040828,
+      "grad_norm": 10.4375,
+      "learning_rate": 3.9339845065678686e-07,
+      "loss": 0.1142,
+      "step": 30670
+    },
+    {
+      "epoch": 0.9816657600870317,
+      "grad_norm": 4.46875,
+      "learning_rate": 3.866621758167734e-07,
+      "loss": 0.0949,
+      "step": 30680
+    },
+    {
+      "epoch": 0.9819857293699805,
+      "grad_norm": 25.125,
+      "learning_rate": 3.799259009767599e-07,
+      "loss": 0.0841,
+      "step": 30690
+    },
+    {
+      "epoch": 0.9823056986529293,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.731896261367464e-07,
+      "loss": 0.1005,
+      "step": 30700
+    },
+    {
+      "epoch": 0.9826256679358781,
+      "grad_norm": 19.25,
+      "learning_rate": 3.66453351296733e-07,
+      "loss": 0.1297,
+      "step": 30710
+    },
+    {
+      "epoch": 0.982945637218827,
+      "grad_norm": 4.5625,
+      "learning_rate": 3.597170764567195e-07,
+      "loss": 0.114,
+      "step": 30720
+    },
+    {
+      "epoch": 0.9832656065017759,
+      "grad_norm": 19.625,
+      "learning_rate": 3.5298080161670597e-07,
+      "loss": 0.1473,
+      "step": 30730
+    },
+    {
+      "epoch": 0.9835855757847247,
+      "grad_norm": 18.25,
+      "learning_rate": 3.462445267766925e-07,
+      "loss": 0.1091,
+      "step": 30740
+    },
+    {
+      "epoch": 0.9839055450676735,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.3950825193667906e-07,
+      "loss": 0.0664,
+      "step": 30750
+    },
+    {
+      "epoch": 0.9842255143506223,
+      "grad_norm": 14.125,
+      "learning_rate": 3.327719770966656e-07,
+      "loss": 0.0747,
+      "step": 30760
+    },
+    {
+      "epoch": 0.9845454836335712,
+      "grad_norm": 16.25,
+      "learning_rate": 3.260357022566521e-07,
+      "loss": 0.0941,
+      "step": 30770
+    },
+    {
+      "epoch": 0.9848654529165201,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.192994274166386e-07,
+      "loss": 0.1096,
+      "step": 30780
+    },
+    {
+      "epoch": 0.9851854221994688,
+      "grad_norm": 19.5,
+      "learning_rate": 3.1256315257662513e-07,
+      "loss": 0.0965,
+      "step": 30790
+    },
+    {
+      "epoch": 0.9855053914824177,
+      "grad_norm": 11.5625,
+      "learning_rate": 3.0582687773661165e-07,
+      "loss": 0.1353,
+      "step": 30800
+    },
+    {
+      "epoch": 0.9858253607653665,
+      "grad_norm": 26.25,
+      "learning_rate": 2.990906028965982e-07,
+      "loss": 0.0604,
+      "step": 30810
+    },
+    {
+      "epoch": 0.9861453300483154,
+      "grad_norm": 9.375,
+      "learning_rate": 2.9235432805658474e-07,
+      "loss": 0.1302,
+      "step": 30820
+    },
+    {
+      "epoch": 0.9864652993312641,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.8561805321657125e-07,
+      "loss": 0.1513,
+      "step": 30830
+    },
+    {
+      "epoch": 0.986785268614213,
+      "grad_norm": 16.0,
+      "learning_rate": 2.7888177837655777e-07,
+      "loss": 0.0773,
+      "step": 30840
+    },
+    {
+      "epoch": 0.9871052378971619,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.721455035365443e-07,
+      "loss": 0.0452,
+      "step": 30850
+    },
+    {
+      "epoch": 0.9874252071801107,
+      "grad_norm": 22.125,
+      "learning_rate": 2.654092286965308e-07,
+      "loss": 0.1346,
+      "step": 30860
+    },
+    {
+      "epoch": 0.9877451764630596,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.586729538565174e-07,
+      "loss": 0.1433,
+      "step": 30870
+    },
+    {
+      "epoch": 0.9880651457460083,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.519366790165039e-07,
+      "loss": 0.1037,
+      "step": 30880
+    },
+    {
+      "epoch": 0.9883851150289572,
+      "grad_norm": 8.75,
+      "learning_rate": 2.452004041764904e-07,
+      "loss": 0.154,
+      "step": 30890
+    },
+    {
+      "epoch": 0.9887050843119061,
+      "grad_norm": 24.25,
+      "learning_rate": 2.3846412933647693e-07,
+      "loss": 0.1128,
+      "step": 30900
+    },
+    {
+      "epoch": 0.9890250535948549,
+      "grad_norm": 17.375,
+      "learning_rate": 2.3172785449646348e-07,
+      "loss": 0.0822,
+      "step": 30910
+    },
+    {
+      "epoch": 0.9893450228778037,
+      "grad_norm": 34.0,
+      "learning_rate": 2.2499157965645e-07,
+      "loss": 0.1088,
+      "step": 30920
+    },
+    {
+      "epoch": 0.9896649921607525,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.1825530481643654e-07,
+      "loss": 0.0829,
+      "step": 30930
+    },
+    {
+      "epoch": 0.9899849614437014,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.1151902997642303e-07,
+      "loss": 0.082,
+      "step": 30940
+    },
+    {
+      "epoch": 0.9903049307266503,
+      "grad_norm": 20.25,
+      "learning_rate": 2.0478275513640957e-07,
+      "loss": 0.1768,
+      "step": 30950
+    },
+    {
+      "epoch": 0.9906249000095991,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.980464802963961e-07,
+      "loss": 0.0725,
+      "step": 30960
+    },
+    {
+      "epoch": 0.9909448692925479,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.9131020545638264e-07,
+      "loss": 0.1199,
+      "step": 30970
+    },
+    {
+      "epoch": 0.9912648385754967,
+      "grad_norm": 13.625,
+      "learning_rate": 1.8457393061636915e-07,
+      "loss": 0.0619,
+      "step": 30980
+    },
+    {
+      "epoch": 0.9915848078584456,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.778376557763557e-07,
+      "loss": 0.1439,
+      "step": 30990
+    },
+    {
+      "epoch": 0.9919047771413945,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.7110138093634222e-07,
+      "loss": 0.1021,
+      "step": 31000
+    },
+    {
+      "epoch": 0.9922247464243432,
+      "grad_norm": 21.0,
+      "learning_rate": 1.6436510609632876e-07,
+      "loss": 0.1386,
+      "step": 31010
+    },
+    {
+      "epoch": 0.9925447157072921,
+      "grad_norm": 16.5,
+      "learning_rate": 1.5762883125631525e-07,
+      "loss": 0.1167,
+      "step": 31020
+    },
+    {
+      "epoch": 0.9928646849902409,
+      "grad_norm": 1.875,
+      "learning_rate": 1.508925564163018e-07,
+      "loss": 0.0717,
+      "step": 31030
+    },
+    {
+      "epoch": 0.9931846542731898,
+      "grad_norm": 27.5,
+      "learning_rate": 1.4415628157628834e-07,
+      "loss": 0.1387,
+      "step": 31040
+    },
+    {
+      "epoch": 0.9935046235561386,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.3742000673627486e-07,
+      "loss": 0.0809,
+      "step": 31050
+    },
+    {
+      "epoch": 0.9938245928390874,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3068373189626137e-07,
+      "loss": 0.0664,
+      "step": 31060
+    },
+    {
+      "epoch": 0.9941445621220363,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.2394745705624792e-07,
+      "loss": 0.1082,
+      "step": 31070
+    },
+    {
+      "epoch": 0.9944645314049851,
+      "grad_norm": 19.25,
+      "learning_rate": 1.1721118221623444e-07,
+      "loss": 0.1239,
+      "step": 31080
+    },
+    {
+      "epoch": 0.994784500687934,
+      "grad_norm": 10.0,
+      "learning_rate": 1.1047490737622097e-07,
+      "loss": 0.0653,
+      "step": 31090
+    },
+    {
+      "epoch": 0.9951044699708828,
+      "grad_norm": 13.5,
+      "learning_rate": 1.0373863253620749e-07,
+      "loss": 0.0619,
+      "step": 31100
+    },
+    {
+      "epoch": 0.9954244392538316,
+      "grad_norm": 22.875,
+      "learning_rate": 9.700235769619402e-08,
+      "loss": 0.1128,
+      "step": 31110
+    },
+    {
+      "epoch": 0.9957444085367805,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.026608285618055e-08,
+      "loss": 0.0826,
+      "step": 31120
+    },
+    {
+      "epoch": 0.9960643778197293,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.352980801616707e-08,
+      "loss": 0.1081,
+      "step": 31130
+    },
+    {
+      "epoch": 0.9963843471026781,
+      "grad_norm": 12.0625,
+      "learning_rate": 7.679353317615358e-08,
+      "loss": 0.1246,
+      "step": 31140
+    },
+    {
+      "epoch": 0.996704316385627,
+      "grad_norm": 9.0,
+      "learning_rate": 7.005725833614011e-08,
+      "loss": 0.0712,
+      "step": 31150
+    },
+    {
+      "epoch": 0.9970242856685758,
+      "grad_norm": 26.75,
+      "learning_rate": 6.332098349612665e-08,
+      "loss": 0.0889,
+      "step": 31160
+    },
+    {
+      "epoch": 0.9973442549515247,
+      "grad_norm": 12.0,
+      "learning_rate": 5.658470865611317e-08,
+      "loss": 0.0913,
+      "step": 31170
+    },
+    {
+      "epoch": 0.9976642242344735,
+      "grad_norm": 15.0,
+      "learning_rate": 4.98484338160997e-08,
+      "loss": 0.1362,
+      "step": 31180
+    },
+    {
+      "epoch": 0.9979841935174223,
+      "grad_norm": 4.875,
+      "learning_rate": 4.3112158976086225e-08,
+      "loss": 0.117,
+      "step": 31190
+    },
+    {
+      "epoch": 0.9983041628003712,
+      "grad_norm": 29.375,
+      "learning_rate": 3.6375884136072756e-08,
+      "loss": 0.1383,
+      "step": 31200
+    },
+    {
+      "epoch": 0.99862413208332,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.9639609296059284e-08,
+      "loss": 0.1093,
+      "step": 31210
+    },
+    {
+      "epoch": 0.9989441013662689,
+      "grad_norm": 13.5625,
+      "learning_rate": 2.2903334456045812e-08,
+      "loss": 0.1618,
+      "step": 31220
+    },
+    {
+      "epoch": 0.9992640706492176,
+      "grad_norm": 22.25,
+      "learning_rate": 1.6167059616032336e-08,
+      "loss": 0.0873,
+      "step": 31230
+    },
+    {
+      "epoch": 0.9995840399321665,
+      "grad_norm": 9.0,
+      "learning_rate": 9.430784776018862e-09,
+      "loss": 0.1089,
+      "step": 31240
+    },
+    {
+      "epoch": 0.9999040092151154,
+      "grad_norm": 33.25,
+      "learning_rate": 2.694509936005389e-09,
+      "loss": 0.0919,
+      "step": 31250
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 31253,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.105650020954463e+18,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}