diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,23282 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 33200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00030120481927710846,
+      "grad_norm": 34.184687640160305,
+      "learning_rate": 6.02409638554217e-08,
+      "loss": 1.9054,
+      "step": 10
+    },
+    {
+      "epoch": 0.0006024096385542169,
+      "grad_norm": 17.14476558690921,
+      "learning_rate": 1.204819277108434e-07,
+      "loss": 2.1905,
+      "step": 20
+    },
+    {
+      "epoch": 0.0009036144578313253,
+      "grad_norm": 33.892200398551054,
+      "learning_rate": 1.8072289156626505e-07,
+      "loss": 1.7507,
+      "step": 30
+    },
+    {
+      "epoch": 0.0012048192771084338,
+      "grad_norm": 22.97431512371491,
+      "learning_rate": 2.409638554216868e-07,
+      "loss": 1.8622,
+      "step": 40
+    },
+    {
+      "epoch": 0.0015060240963855422,
+      "grad_norm": 20.59352371925957,
+      "learning_rate": 3.0120481927710845e-07,
+      "loss": 1.9969,
+      "step": 50
+    },
+    {
+      "epoch": 0.0018072289156626507,
+      "grad_norm": 11.716491338439093,
+      "learning_rate": 3.614457831325301e-07,
+      "loss": 1.5547,
+      "step": 60
+    },
+    {
+      "epoch": 0.002108433734939759,
+      "grad_norm": 8.45595968752092,
+      "learning_rate": 4.216867469879518e-07,
+      "loss": 1.36,
+      "step": 70
+    },
+    {
+      "epoch": 0.0024096385542168677,
+      "grad_norm": 4.2568241543074,
+      "learning_rate": 4.819277108433736e-07,
+      "loss": 1.1727,
+      "step": 80
+    },
+    {
+      "epoch": 0.002710843373493976,
+      "grad_norm": 6.321702055594082,
+      "learning_rate": 5.421686746987952e-07,
+      "loss": 1.6102,
+      "step": 90
+    },
+    {
+      "epoch": 0.0030120481927710845,
+      "grad_norm": 5.267243903465454,
+      "learning_rate": 6.024096385542169e-07,
+      "loss": 1.2673,
+      "step": 100
+    },
+    {
+      "epoch": 0.0033132530120481927,
+      "grad_norm": 5.9212115360469015,
+      "learning_rate": 6.626506024096387e-07,
+      "loss": 1.2851,
+      "step": 110
+    },
+    {
+      "epoch": 0.0036144578313253013,
+      "grad_norm": 2.7909893738719256,
+      "learning_rate": 7.228915662650602e-07,
+      "loss": 1.2658,
+      "step": 120
+    },
+    {
+      "epoch": 0.00391566265060241,
+      "grad_norm": 4.700991426756883,
+      "learning_rate": 7.83132530120482e-07,
+      "loss": 1.2353,
+      "step": 130
+    },
+    {
+      "epoch": 0.004216867469879518,
+      "grad_norm": 2.982566988838653,
+      "learning_rate": 8.433734939759036e-07,
+      "loss": 1.1382,
+      "step": 140
+    },
+    {
+      "epoch": 0.004518072289156626,
+      "grad_norm": 4.751028531989113,
+      "learning_rate": 9.036144578313254e-07,
+      "loss": 1.2219,
+      "step": 150
+    },
+    {
+      "epoch": 0.004819277108433735,
+      "grad_norm": 4.058570665682753,
+      "learning_rate": 9.638554216867472e-07,
+      "loss": 1.4589,
+      "step": 160
+    },
+    {
+      "epoch": 0.0051204819277108436,
+      "grad_norm": 4.391684695638581,
+      "learning_rate": 1.0240963855421688e-06,
+      "loss": 1.2259,
+      "step": 170
+    },
+    {
+      "epoch": 0.005421686746987952,
+      "grad_norm": 2.749288984807557,
+      "learning_rate": 1.0843373493975905e-06,
+      "loss": 1.1787,
+      "step": 180
+    },
+    {
+      "epoch": 0.00572289156626506,
+      "grad_norm": 4.7615868068825575,
+      "learning_rate": 1.1445783132530121e-06,
+      "loss": 1.0488,
+      "step": 190
+    },
+    {
+      "epoch": 0.006024096385542169,
+      "grad_norm": 4.7129578011308535,
+      "learning_rate": 1.2048192771084338e-06,
+      "loss": 1.1802,
+      "step": 200
+    },
+    {
+      "epoch": 0.006325301204819277,
+      "grad_norm": 4.575454200356594,
+      "learning_rate": 1.2650602409638555e-06,
+      "loss": 1.2244,
+      "step": 210
+    },
+    {
+      "epoch": 0.006626506024096385,
+      "grad_norm": 4.406374457009423,
+      "learning_rate": 1.3253012048192773e-06,
+      "loss": 1.211,
+      "step": 220
+    },
+    {
+      "epoch": 0.0069277108433734936,
+      "grad_norm": 4.197678974172092,
+      "learning_rate": 1.385542168674699e-06,
+      "loss": 1.1268,
+      "step": 230
+    },
+    {
+      "epoch": 0.007228915662650603,
+      "grad_norm": 4.575232860771412,
+      "learning_rate": 1.4457831325301204e-06,
+      "loss": 1.1966,
+      "step": 240
+    },
+    {
+      "epoch": 0.007530120481927711,
+      "grad_norm": 4.099565929616023,
+      "learning_rate": 1.5060240963855425e-06,
+      "loss": 1.4051,
+      "step": 250
+    },
+    {
+      "epoch": 0.00783132530120482,
+      "grad_norm": 4.270337306856698,
+      "learning_rate": 1.566265060240964e-06,
+      "loss": 1.3423,
+      "step": 260
+    },
+    {
+      "epoch": 0.008132530120481927,
+      "grad_norm": 3.982952502682325,
+      "learning_rate": 1.6265060240963856e-06,
+      "loss": 1.3128,
+      "step": 270
+    },
+    {
+      "epoch": 0.008433734939759036,
+      "grad_norm": 4.004693789896972,
+      "learning_rate": 1.6867469879518073e-06,
+      "loss": 1.3504,
+      "step": 280
+    },
+    {
+      "epoch": 0.008734939759036145,
+      "grad_norm": 4.592468092346859,
+      "learning_rate": 1.7469879518072292e-06,
+      "loss": 1.2639,
+      "step": 290
+    },
+    {
+      "epoch": 0.009036144578313253,
+      "grad_norm": 4.060501115633273,
+      "learning_rate": 1.8072289156626508e-06,
+      "loss": 1.2481,
+      "step": 300
+    },
+    {
+      "epoch": 0.009337349397590362,
+      "grad_norm": 4.815336571061053,
+      "learning_rate": 1.8674698795180723e-06,
+      "loss": 1.2346,
+      "step": 310
+    },
+    {
+      "epoch": 0.00963855421686747,
+      "grad_norm": 4.053695477874738,
+      "learning_rate": 1.9277108433734943e-06,
+      "loss": 0.9776,
+      "step": 320
+    },
+    {
+      "epoch": 0.009939759036144578,
+      "grad_norm": 3.919326425143882,
+      "learning_rate": 1.987951807228916e-06,
+      "loss": 1.3553,
+      "step": 330
+    },
+    {
+      "epoch": 0.010240963855421687,
+      "grad_norm": 3.044358040142568,
+      "learning_rate": 2.0481927710843377e-06,
+      "loss": 1.2019,
+      "step": 340
+    },
+    {
+      "epoch": 0.010542168674698794,
+      "grad_norm": 2.8849085807871373,
+      "learning_rate": 2.1084337349397595e-06,
+      "loss": 1.1237,
+      "step": 350
+    },
+    {
+      "epoch": 0.010843373493975903,
+      "grad_norm": 4.338427499373051,
+      "learning_rate": 2.168674698795181e-06,
+      "loss": 1.2304,
+      "step": 360
+    },
+    {
+      "epoch": 0.011144578313253013,
+      "grad_norm": 4.230190025931408,
+      "learning_rate": 2.2289156626506024e-06,
+      "loss": 1.1195,
+      "step": 370
+    },
+    {
+      "epoch": 0.01144578313253012,
+      "grad_norm": 4.124234826305292,
+      "learning_rate": 2.2891566265060243e-06,
+      "loss": 1.279,
+      "step": 380
+    },
+    {
+      "epoch": 0.011746987951807229,
+      "grad_norm": 4.32999826606396,
+      "learning_rate": 2.349397590361446e-06,
+      "loss": 1.1222,
+      "step": 390
+    },
+    {
+      "epoch": 0.012048192771084338,
+      "grad_norm": 4.037705852563037,
+      "learning_rate": 2.4096385542168676e-06,
+      "loss": 1.171,
+      "step": 400
+    },
+    {
+      "epoch": 0.012349397590361445,
+      "grad_norm": 3.6215123183400806,
+      "learning_rate": 2.469879518072289e-06,
+      "loss": 1.3193,
+      "step": 410
+    },
+    {
+      "epoch": 0.012650602409638554,
+      "grad_norm": 4.0649621509685785,
+      "learning_rate": 2.530120481927711e-06,
+      "loss": 1.2043,
+      "step": 420
+    },
+    {
+      "epoch": 0.012951807228915663,
+      "grad_norm": 3.833813479206042,
+      "learning_rate": 2.590361445783133e-06,
+      "loss": 1.129,
+      "step": 430
+    },
+    {
+      "epoch": 0.01325301204819277,
+      "grad_norm": 3.951963927513688,
+      "learning_rate": 2.6506024096385547e-06,
+      "loss": 1.1101,
+      "step": 440
+    },
+    {
+      "epoch": 0.01355421686746988,
+      "grad_norm": 4.271519936836857,
+      "learning_rate": 2.710843373493976e-06,
+      "loss": 1.271,
+      "step": 450
+    },
+    {
+      "epoch": 0.013855421686746987,
+      "grad_norm": 2.6764256226165317,
+      "learning_rate": 2.771084337349398e-06,
+      "loss": 1.1224,
+      "step": 460
+    },
+    {
+      "epoch": 0.014156626506024096,
+      "grad_norm": 4.634735718452594,
+      "learning_rate": 2.83132530120482e-06,
+      "loss": 1.1163,
+      "step": 470
+    },
+    {
+      "epoch": 0.014457831325301205,
+      "grad_norm": 2.7293250507262155,
+      "learning_rate": 2.891566265060241e-06,
+      "loss": 0.9871,
+      "step": 480
+    },
+    {
+      "epoch": 0.014759036144578313,
+      "grad_norm": 3.866232843232568,
+      "learning_rate": 2.9518072289156627e-06,
+      "loss": 1.3567,
+      "step": 490
+    },
+    {
+      "epoch": 0.015060240963855422,
+      "grad_norm": 4.239742910872224,
+      "learning_rate": 3.012048192771085e-06,
+      "loss": 0.9524,
+      "step": 500
+    },
+    {
+      "epoch": 0.01536144578313253,
+      "grad_norm": 2.644873772813341,
+      "learning_rate": 3.072289156626506e-06,
+      "loss": 1.0942,
+      "step": 510
+    },
+    {
+      "epoch": 0.01566265060240964,
+      "grad_norm": 4.032437393625629,
+      "learning_rate": 3.132530120481928e-06,
+      "loss": 1.1671,
+      "step": 520
+    },
+    {
+      "epoch": 0.015963855421686747,
+      "grad_norm": 2.88176103182694,
+      "learning_rate": 3.1927710843373494e-06,
+      "loss": 1.0398,
+      "step": 530
+    },
+    {
+      "epoch": 0.016265060240963854,
+      "grad_norm": 4.0032840796024765,
+      "learning_rate": 3.2530120481927713e-06,
+      "loss": 1.0952,
+      "step": 540
+    },
+    {
+      "epoch": 0.016566265060240965,
+      "grad_norm": 3.902519176881758,
+      "learning_rate": 3.313253012048193e-06,
+      "loss": 1.1107,
+      "step": 550
+    },
+    {
+      "epoch": 0.016867469879518072,
+      "grad_norm": 4.340144191056268,
+      "learning_rate": 3.3734939759036146e-06,
+      "loss": 1.1925,
+      "step": 560
+    },
+    {
+      "epoch": 0.01716867469879518,
+      "grad_norm": 3.7467344008008316,
+      "learning_rate": 3.4337349397590364e-06,
+      "loss": 1.2664,
+      "step": 570
+    },
+    {
+      "epoch": 0.01746987951807229,
+      "grad_norm": 3.8308835387585694,
+      "learning_rate": 3.4939759036144583e-06,
+      "loss": 1.1655,
+      "step": 580
+    },
+    {
+      "epoch": 0.017771084337349398,
+      "grad_norm": 2.7014599545087865,
+      "learning_rate": 3.5542168674698798e-06,
+      "loss": 0.983,
+      "step": 590
+    },
+    {
+      "epoch": 0.018072289156626505,
+      "grad_norm": 2.6573476355219485,
+      "learning_rate": 3.6144578313253016e-06,
+      "loss": 1.0783,
+      "step": 600
+    },
+    {
+      "epoch": 0.018373493975903616,
+      "grad_norm": 4.238081108120648,
+      "learning_rate": 3.6746987951807235e-06,
+      "loss": 1.163,
+      "step": 610
+    },
+    {
+      "epoch": 0.018674698795180723,
+      "grad_norm": 4.306356593664175,
+      "learning_rate": 3.7349397590361445e-06,
+      "loss": 1.0491,
+      "step": 620
+    },
+    {
+      "epoch": 0.01897590361445783,
+      "grad_norm": 3.9183023904130625,
+      "learning_rate": 3.7951807228915664e-06,
+      "loss": 0.885,
+      "step": 630
+    },
+    {
+      "epoch": 0.01927710843373494,
+      "grad_norm": 2.754654672572441,
+      "learning_rate": 3.855421686746989e-06,
+      "loss": 0.9848,
+      "step": 640
+    },
+    {
+      "epoch": 0.01957831325301205,
+      "grad_norm": 2.9070743973349047,
+      "learning_rate": 3.91566265060241e-06,
+      "loss": 1.0814,
+      "step": 650
+    },
+    {
+      "epoch": 0.019879518072289156,
+      "grad_norm": 3.832054275721117,
+      "learning_rate": 3.975903614457832e-06,
+      "loss": 1.0426,
+      "step": 660
+    },
+    {
+      "epoch": 0.020180722891566263,
+      "grad_norm": 2.9112570734345375,
+      "learning_rate": 4.036144578313254e-06,
+      "loss": 1.0199,
+      "step": 670
+    },
+    {
+      "epoch": 0.020481927710843374,
+      "grad_norm": 3.884916864538904,
+      "learning_rate": 4.096385542168675e-06,
+      "loss": 1.104,
+      "step": 680
+    },
+    {
+      "epoch": 0.02078313253012048,
+      "grad_norm": 4.1678404584613205,
+      "learning_rate": 4.156626506024097e-06,
+      "loss": 1.0895,
+      "step": 690
+    },
+    {
+      "epoch": 0.02108433734939759,
+      "grad_norm": 3.864285436996809,
+      "learning_rate": 4.216867469879519e-06,
+      "loss": 1.2234,
+      "step": 700
+    },
+    {
+      "epoch": 0.0213855421686747,
+      "grad_norm": 5.16849448714143,
+      "learning_rate": 4.27710843373494e-06,
+      "loss": 1.1659,
+      "step": 710
+    },
+    {
+      "epoch": 0.021686746987951807,
+      "grad_norm": 4.095326962423422,
+      "learning_rate": 4.337349397590362e-06,
+      "loss": 1.1602,
+      "step": 720
+    },
+    {
+      "epoch": 0.021987951807228914,
+      "grad_norm": 2.952963894247243,
+      "learning_rate": 4.397590361445783e-06,
+      "loss": 0.9866,
+      "step": 730
+    },
+    {
+      "epoch": 0.022289156626506025,
+      "grad_norm": 4.342880994909028,
+      "learning_rate": 4.457831325301205e-06,
+      "loss": 1.0471,
+      "step": 740
+    },
+    {
+      "epoch": 0.022590361445783132,
+      "grad_norm": 4.188011930874192,
+      "learning_rate": 4.518072289156627e-06,
+      "loss": 1.0198,
+      "step": 750
+    },
+    {
+      "epoch": 0.02289156626506024,
+      "grad_norm": 3.9036198525868993,
+      "learning_rate": 4.578313253012049e-06,
+      "loss": 1.2074,
+      "step": 760
+    },
+    {
+      "epoch": 0.02319277108433735,
+      "grad_norm": 4.308339600081488,
+      "learning_rate": 4.63855421686747e-06,
+      "loss": 1.2214,
+      "step": 770
+    },
+    {
+      "epoch": 0.023493975903614458,
+      "grad_norm": 4.200789362568133,
+      "learning_rate": 4.698795180722892e-06,
+      "loss": 1.2115,
+      "step": 780
+    },
+    {
+      "epoch": 0.023795180722891565,
+      "grad_norm": 3.8186451735550135,
+      "learning_rate": 4.759036144578314e-06,
+      "loss": 1.0943,
+      "step": 790
+    },
+    {
+      "epoch": 0.024096385542168676,
+      "grad_norm": 3.7520522634243676,
+      "learning_rate": 4.819277108433735e-06,
+      "loss": 1.0688,
+      "step": 800
+    },
+    {
+      "epoch": 0.024397590361445783,
+      "grad_norm": 2.8150643850127186,
+      "learning_rate": 4.8795180722891575e-06,
+      "loss": 1.099,
+      "step": 810
+    },
+    {
+      "epoch": 0.02469879518072289,
+      "grad_norm": 3.0519094509153892,
+      "learning_rate": 4.939759036144578e-06,
+      "loss": 1.0965,
+      "step": 820
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 4.194014232032205,
+      "learning_rate": 5e-06,
+      "loss": 0.9908,
+      "step": 830
+    },
+    {
+      "epoch": 0.02530120481927711,
+      "grad_norm": 4.302589861076283,
+      "learning_rate": 5.060240963855422e-06,
+      "loss": 1.1164,
+      "step": 840
+    },
+    {
+      "epoch": 0.025602409638554216,
+      "grad_norm": 4.907180215758534,
+      "learning_rate": 5.120481927710844e-06,
+      "loss": 1.0914,
+      "step": 850
+    },
+    {
+      "epoch": 0.025903614457831327,
+      "grad_norm": 4.373876586118721,
+      "learning_rate": 5.180722891566266e-06,
+      "loss": 1.0364,
+      "step": 860
+    },
+    {
+      "epoch": 0.026204819277108434,
+      "grad_norm": 4.722558680398194,
+      "learning_rate": 5.240963855421687e-06,
+      "loss": 1.1332,
+      "step": 870
+    },
+    {
+      "epoch": 0.02650602409638554,
+      "grad_norm": 2.931284980772947,
+      "learning_rate": 5.301204819277109e-06,
+      "loss": 1.0539,
+      "step": 880
+    },
+    {
+      "epoch": 0.026807228915662652,
+      "grad_norm": 4.041583561605744,
+      "learning_rate": 5.361445783132531e-06,
+      "loss": 1.1339,
+      "step": 890
+    },
+    {
+      "epoch": 0.02710843373493976,
+      "grad_norm": 4.082971119962101,
+      "learning_rate": 5.421686746987952e-06,
+      "loss": 1.2437,
+      "step": 900
+    },
+    {
+      "epoch": 0.027409638554216867,
+      "grad_norm": 4.617617434680252,
+      "learning_rate": 5.4819277108433745e-06,
+      "loss": 1.092,
+      "step": 910
+    },
+    {
+      "epoch": 0.027710843373493974,
+      "grad_norm": 4.0349654871489,
+      "learning_rate": 5.542168674698796e-06,
+      "loss": 1.1967,
+      "step": 920
+    },
+    {
+      "epoch": 0.028012048192771085,
+      "grad_norm": 2.6027878449239426,
+      "learning_rate": 5.602409638554217e-06,
+      "loss": 1.0091,
+      "step": 930
+    },
+    {
+      "epoch": 0.028313253012048192,
+      "grad_norm": 4.485028893262391,
+      "learning_rate": 5.66265060240964e-06,
+      "loss": 1.0483,
+      "step": 940
+    },
+    {
+      "epoch": 0.0286144578313253,
+      "grad_norm": 4.522486623622015,
+      "learning_rate": 5.722891566265061e-06,
+      "loss": 1.2027,
+      "step": 950
+    },
+    {
+      "epoch": 0.02891566265060241,
+      "grad_norm": 2.8472374483590546,
+      "learning_rate": 5.783132530120482e-06,
+      "loss": 0.9967,
+      "step": 960
+    },
+    {
+      "epoch": 0.029216867469879518,
+      "grad_norm": 4.7385115596889635,
+      "learning_rate": 5.843373493975905e-06,
+      "loss": 1.1533,
+      "step": 970
+    },
+    {
+      "epoch": 0.029518072289156625,
+      "grad_norm": 3.076634710050929,
+      "learning_rate": 5.9036144578313255e-06,
+      "loss": 1.1005,
+      "step": 980
+    },
+    {
+      "epoch": 0.029819277108433736,
+      "grad_norm": 2.689123057316776,
+      "learning_rate": 5.963855421686747e-06,
+      "loss": 1.1132,
+      "step": 990
+    },
+    {
+      "epoch": 0.030120481927710843,
+      "grad_norm": 3.707408578720797,
+      "learning_rate": 6.02409638554217e-06,
+      "loss": 1.1946,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03042168674698795,
+      "grad_norm": 5.063000512597904,
+      "learning_rate": 6.084337349397591e-06,
+      "loss": 1.0708,
+      "step": 1010
+    },
+    {
+      "epoch": 0.03072289156626506,
+      "grad_norm": 4.06276247660459,
+      "learning_rate": 6.144578313253012e-06,
+      "loss": 0.9929,
+      "step": 1020
+    },
+    {
+      "epoch": 0.03102409638554217,
+      "grad_norm": 4.322190402987572,
+      "learning_rate": 6.2048192771084344e-06,
+      "loss": 1.0097,
+      "step": 1030
+    },
+    {
+      "epoch": 0.03132530120481928,
+      "grad_norm": 4.1985515286316195,
+      "learning_rate": 6.265060240963856e-06,
+      "loss": 1.1831,
+      "step": 1040
+    },
+    {
+      "epoch": 0.03162650602409638,
+      "grad_norm": 4.738056079854521,
+      "learning_rate": 6.325301204819277e-06,
+      "loss": 1.0026,
+      "step": 1050
+    },
+    {
+      "epoch": 0.031927710843373494,
+      "grad_norm": 7.908138501617392,
+      "learning_rate": 6.385542168674699e-06,
+      "loss": 1.0075,
+      "step": 1060
+    },
+    {
+      "epoch": 0.032228915662650605,
+      "grad_norm": 4.095549042393854,
+      "learning_rate": 6.445783132530121e-06,
+      "loss": 1.1996,
+      "step": 1070
+    },
+    {
+      "epoch": 0.03253012048192771,
+      "grad_norm": 4.607760763528683,
+      "learning_rate": 6.5060240963855425e-06,
+      "loss": 1.0659,
+      "step": 1080
+    },
+    {
+      "epoch": 0.03283132530120482,
+      "grad_norm": 3.599600156442471,
+      "learning_rate": 6.566265060240964e-06,
+      "loss": 0.9888,
+      "step": 1090
+    },
+    {
+      "epoch": 0.03313253012048193,
+      "grad_norm": 4.57889685904033,
+      "learning_rate": 6.626506024096386e-06,
+      "loss": 1.0804,
+      "step": 1100
+    },
+    {
+      "epoch": 0.033433734939759034,
+      "grad_norm": 2.8468610531916267,
+      "learning_rate": 6.686746987951808e-06,
+      "loss": 0.9436,
+      "step": 1110
+    },
+    {
+      "epoch": 0.033734939759036145,
+      "grad_norm": 4.028682492647136,
+      "learning_rate": 6.746987951807229e-06,
+      "loss": 1.042,
+      "step": 1120
+    },
+    {
+      "epoch": 0.034036144578313256,
+      "grad_norm": 4.487223110665026,
+      "learning_rate": 6.8072289156626514e-06,
+      "loss": 1.1325,
+      "step": 1130
+    },
+    {
+      "epoch": 0.03433734939759036,
+      "grad_norm": 4.528149873897056,
+      "learning_rate": 6.867469879518073e-06,
+      "loss": 1.2997,
+      "step": 1140
+    },
+    {
+      "epoch": 0.03463855421686747,
+      "grad_norm": 3.5686330335011553,
+      "learning_rate": 6.927710843373494e-06,
+      "loss": 0.9743,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03493975903614458,
+      "grad_norm": 3.2377711419187203,
+      "learning_rate": 6.987951807228917e-06,
+      "loss": 1.1301,
+      "step": 1160
+    },
+    {
+      "epoch": 0.035240963855421685,
+      "grad_norm": 4.098978665123173,
+      "learning_rate": 7.048192771084338e-06,
+      "loss": 1.1878,
+      "step": 1170
+    },
+    {
+      "epoch": 0.035542168674698796,
+      "grad_norm": 3.7882630819857774,
+      "learning_rate": 7.1084337349397595e-06,
+      "loss": 1.1188,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03584337349397591,
+      "grad_norm": 4.4587101564262515,
+      "learning_rate": 7.168674698795182e-06,
+      "loss": 1.1766,
+      "step": 1190
+    },
+    {
+      "epoch": 0.03614457831325301,
+      "grad_norm": 3.8989607392844507,
+      "learning_rate": 7.228915662650603e-06,
+      "loss": 1.0636,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03644578313253012,
+      "grad_norm": 3.839191587235316,
+      "learning_rate": 7.289156626506025e-06,
+      "loss": 1.196,
+      "step": 1210
+    },
+    {
+      "epoch": 0.03674698795180723,
+      "grad_norm": 3.78409926911461,
+      "learning_rate": 7.349397590361447e-06,
+      "loss": 1.0951,
+      "step": 1220
+    },
+    {
+      "epoch": 0.037048192771084336,
+      "grad_norm": 2.7074286505344913,
+      "learning_rate": 7.4096385542168684e-06,
+      "loss": 0.9479,
+      "step": 1230
+    },
+    {
+      "epoch": 0.03734939759036145,
+      "grad_norm": 4.381090966336082,
+      "learning_rate": 7.469879518072289e-06,
+      "loss": 1.1028,
+      "step": 1240
+    },
+    {
+      "epoch": 0.03765060240963856,
+      "grad_norm": 4.764045150202293,
+      "learning_rate": 7.530120481927712e-06,
+      "loss": 1.0773,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03795180722891566,
+      "grad_norm": 3.032725440916371,
+      "learning_rate": 7.590361445783133e-06,
+      "loss": 1.0261,
+      "step": 1260
+    },
+    {
+      "epoch": 0.03825301204819277,
+      "grad_norm": 4.035121309975671,
+      "learning_rate": 7.650602409638555e-06,
+      "loss": 1.062,
+      "step": 1270
+    },
+    {
+      "epoch": 0.03855421686746988,
+      "grad_norm": 3.5854891100719564,
+      "learning_rate": 7.710843373493977e-06,
+      "loss": 1.0601,
+      "step": 1280
+    },
+    {
+      "epoch": 0.03885542168674699,
+      "grad_norm": 3.8237871025587435,
+      "learning_rate": 7.771084337349398e-06,
+      "loss": 1.1306,
+      "step": 1290
+    },
+    {
+      "epoch": 0.0391566265060241,
+      "grad_norm": 4.115558747304762,
+      "learning_rate": 7.83132530120482e-06,
+      "loss": 1.0698,
+      "step": 1300
+    },
+    {
+      "epoch": 0.0394578313253012,
+      "grad_norm": 3.7410956006810547,
+      "learning_rate": 7.891566265060243e-06,
+      "loss": 1.146,
+      "step": 1310
+    },
+    {
+      "epoch": 0.03975903614457831,
+      "grad_norm": 3.6105166014436936,
+      "learning_rate": 7.951807228915663e-06,
+      "loss": 1.1621,
+      "step": 1320
+    },
+    {
+      "epoch": 0.04006024096385542,
+      "grad_norm": 3.973739829544077,
+      "learning_rate": 8.012048192771085e-06,
+      "loss": 1.2101,
+      "step": 1330
+    },
+    {
+      "epoch": 0.04036144578313253,
+      "grad_norm": 2.5733017305600114,
+      "learning_rate": 8.072289156626508e-06,
+      "loss": 1.0769,
+      "step": 1340
+    },
+    {
+      "epoch": 0.04066265060240964,
+      "grad_norm": 3.133485976297979,
+      "learning_rate": 8.132530120481928e-06,
+      "loss": 1.0562,
+      "step": 1350
+    },
+    {
+      "epoch": 0.04096385542168675,
+      "grad_norm": 4.01743747234379,
+      "learning_rate": 8.19277108433735e-06,
+      "loss": 1.0677,
+      "step": 1360
+    },
+    {
+      "epoch": 0.04126506024096385,
+      "grad_norm": 5.793222896431323,
+      "learning_rate": 8.253012048192773e-06,
+      "loss": 1.1337,
+      "step": 1370
+    },
+    {
+      "epoch": 0.04156626506024096,
+      "grad_norm": 4.267870315133769,
+      "learning_rate": 8.313253012048194e-06,
+      "loss": 1.1537,
+      "step": 1380
+    },
+    {
+      "epoch": 0.041867469879518074,
+      "grad_norm": 3.56150217463621,
+      "learning_rate": 8.373493975903614e-06,
+      "loss": 1.015,
+      "step": 1390
+    },
+    {
+      "epoch": 0.04216867469879518,
+      "grad_norm": 3.44868228410339,
+      "learning_rate": 8.433734939759038e-06,
+      "loss": 1.2518,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04246987951807229,
+      "grad_norm": 4.069954456635347,
+      "learning_rate": 8.493975903614459e-06,
+      "loss": 1.0554,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0427710843373494,
+      "grad_norm": 2.889978527209783,
+      "learning_rate": 8.55421686746988e-06,
+      "loss": 0.8878,
+      "step": 1420
+    },
+    {
+      "epoch": 0.0430722891566265,
+      "grad_norm": 3.8526229033524673,
+      "learning_rate": 8.614457831325302e-06,
+      "loss": 1.1049,
+      "step": 1430
+    },
+    {
+      "epoch": 0.043373493975903614,
+      "grad_norm": 3.992363809147142,
+      "learning_rate": 8.674698795180724e-06,
+      "loss": 1.1341,
+      "step": 1440
+    },
+    {
+      "epoch": 0.043674698795180725,
+      "grad_norm": 3.6448879341956895,
+      "learning_rate": 8.734939759036145e-06,
+      "loss": 1.1601,
+      "step": 1450
+    },
+    {
+      "epoch": 0.04397590361445783,
+      "grad_norm": 4.514705884392051,
+      "learning_rate": 8.795180722891567e-06,
+      "loss": 1.0113,
+      "step": 1460
+    },
+    {
+      "epoch": 0.04427710843373494,
+      "grad_norm": 4.0363185397570245,
+      "learning_rate": 8.855421686746989e-06,
+      "loss": 1.0451,
+      "step": 1470
+    },
+    {
+      "epoch": 0.04457831325301205,
+      "grad_norm": 3.925379146933244,
+      "learning_rate": 8.91566265060241e-06,
+      "loss": 1.1947,
+      "step": 1480
+    },
+    {
+      "epoch": 0.044879518072289154,
+      "grad_norm": 2.6181945942087608,
+      "learning_rate": 8.975903614457832e-06,
+      "loss": 1.0535,
+      "step": 1490
+    },
+    {
+      "epoch": 0.045180722891566265,
+      "grad_norm": 3.110805657773211,
+      "learning_rate": 9.036144578313254e-06,
+      "loss": 1.0878,
+      "step": 1500
+    },
+    {
+      "epoch": 0.045481927710843376,
+      "grad_norm": 3.4913255776936003,
+      "learning_rate": 9.096385542168675e-06,
+      "loss": 1.1069,
+      "step": 1510
+    },
+    {
+      "epoch": 0.04578313253012048,
+      "grad_norm": 2.7603777767228563,
+      "learning_rate": 9.156626506024097e-06,
+      "loss": 1.0823,
+      "step": 1520
+    },
+    {
+      "epoch": 0.04608433734939759,
+      "grad_norm": 4.280614356362033,
+      "learning_rate": 9.21686746987952e-06,
+      "loss": 1.1414,
+      "step": 1530
+    },
+    {
+      "epoch": 0.0463855421686747,
+      "grad_norm": 2.991791650219649,
+      "learning_rate": 9.27710843373494e-06,
+      "loss": 0.9234,
+      "step": 1540
+    },
+    {
+      "epoch": 0.046686746987951805,
+      "grad_norm": 4.513985825827475,
+      "learning_rate": 9.337349397590362e-06,
+      "loss": 1.0021,
+      "step": 1550
+    },
+    {
+      "epoch": 0.046987951807228916,
+      "grad_norm": 3.8650087083781033,
+      "learning_rate": 9.397590361445785e-06,
+      "loss": 1.0892,
+      "step": 1560
+    },
+    {
+      "epoch": 0.047289156626506026,
+      "grad_norm": 4.69885903360439,
+      "learning_rate": 9.457831325301205e-06,
+      "loss": 1.0488,
+      "step": 1570
+    },
+    {
+      "epoch": 0.04759036144578313,
+      "grad_norm": 4.178287377782754,
+      "learning_rate": 9.518072289156628e-06,
+      "loss": 1.1808,
+      "step": 1580
+    },
+    {
+      "epoch": 0.04789156626506024,
+      "grad_norm": 4.041913650511557,
+      "learning_rate": 9.57831325301205e-06,
+      "loss": 1.1977,
+      "step": 1590
+    },
+    {
+      "epoch": 0.04819277108433735,
+      "grad_norm": 2.4573674173962474,
+      "learning_rate": 9.63855421686747e-06,
+      "loss": 1.0869,
+      "step": 1600
+    },
+    {
+      "epoch": 0.048493975903614456,
+      "grad_norm": 3.7242981224219056,
+      "learning_rate": 9.698795180722893e-06,
+      "loss": 0.98,
+      "step": 1610
+    },
+    {
+      "epoch": 0.04879518072289157,
+      "grad_norm": 3.3696425145200015,
+      "learning_rate": 9.759036144578315e-06,
+      "loss": 1.1795,
+      "step": 1620
+    },
+    {
+      "epoch": 0.04909638554216868,
+      "grad_norm": 3.765251027998519,
+      "learning_rate": 9.819277108433736e-06,
+      "loss": 1.1529,
+      "step": 1630
+    },
+    {
+      "epoch": 0.04939759036144578,
+      "grad_norm": 4.380167625859237,
+      "learning_rate": 9.879518072289156e-06,
+      "loss": 1.0109,
+      "step": 1640
+    },
+    {
+      "epoch": 0.04969879518072289,
+      "grad_norm": 3.5901053438352757,
+      "learning_rate": 9.93975903614458e-06,
+      "loss": 1.0782,
+      "step": 1650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.697835290455968,
+      "learning_rate": 1e-05,
+      "loss": 1.1538,
+      "step": 1660
+    },
+    {
+      "epoch": 0.05030120481927711,
+      "grad_norm": 5.149787332226158,
+      "learning_rate": 9.999997519630741e-06,
+      "loss": 1.1527,
+      "step": 1670
+    },
+    {
+      "epoch": 0.05060240963855422,
+      "grad_norm": 4.2386612321855575,
+      "learning_rate": 9.999990078525426e-06,
+      "loss": 1.0683,
+      "step": 1680
+    },
+    {
+      "epoch": 0.05090361445783133,
+      "grad_norm": 3.6473005117732313,
+      "learning_rate": 9.999977676691436e-06,
+      "loss": 1.2378,
+      "step": 1690
+    },
+    {
+      "epoch": 0.05120481927710843,
+      "grad_norm": 3.7901644794972937,
+      "learning_rate": 9.999960314141076e-06,
+      "loss": 1.1079,
+      "step": 1700
+    },
+    {
+      "epoch": 0.05150602409638554,
+      "grad_norm": 3.68394417330946,
+      "learning_rate": 9.999937990891573e-06,
+      "loss": 1.0808,
+      "step": 1710
+    },
+    {
+      "epoch": 0.051807228915662654,
+      "grad_norm": 4.8171466442202036,
+      "learning_rate": 9.999910706965073e-06,
+      "loss": 1.0813,
+      "step": 1720
+    },
+    {
+      "epoch": 0.05210843373493976,
+      "grad_norm": 4.210089621064192,
+      "learning_rate": 9.999878462388647e-06,
+      "loss": 1.1659,
+      "step": 1730
+    },
+    {
+      "epoch": 0.05240963855421687,
+      "grad_norm": 2.401098250310252,
+      "learning_rate": 9.99984125719429e-06,
+      "loss": 1.0514,
+      "step": 1740
+    },
+    {
+      "epoch": 0.05271084337349398,
+      "grad_norm": 3.9199440778618113,
+      "learning_rate": 9.999799091418908e-06,
+      "loss": 1.1456,
+      "step": 1750
+    },
+    {
+      "epoch": 0.05301204819277108,
+      "grad_norm": 2.433777634678354,
+      "learning_rate": 9.99975196510434e-06,
+      "loss": 0.8795,
+      "step": 1760
+    },
+    {
+      "epoch": 0.053313253012048194,
+      "grad_norm": 3.815132735539651,
+      "learning_rate": 9.999699878297342e-06,
+      "loss": 1.2158,
+      "step": 1770
+    },
+    {
+      "epoch": 0.053614457831325305,
+      "grad_norm": 4.387848610690783,
+      "learning_rate": 9.999642831049591e-06,
+      "loss": 1.1993,
+      "step": 1780
+    },
+    {
+      "epoch": 0.05391566265060241,
+      "grad_norm": 4.701283174117646,
+      "learning_rate": 9.999580823417688e-06,
+      "loss": 1.0755,
+      "step": 1790
+    },
+    {
+      "epoch": 0.05421686746987952,
+      "grad_norm": 3.752459770285841,
+      "learning_rate": 9.99951385546315e-06,
+      "loss": 1.0101,
+      "step": 1800
+    },
+    {
+      "epoch": 0.05451807228915663,
+      "grad_norm": 5.337745045286134,
+      "learning_rate": 9.99944192725242e-06,
+      "loss": 1.1359,
+      "step": 1810
+    },
+    {
+      "epoch": 0.054819277108433734,
+      "grad_norm": 3.1352783689331547,
+      "learning_rate": 9.999365038856866e-06,
+      "loss": 1.0291,
+      "step": 1820
+    },
+    {
+      "epoch": 0.055120481927710845,
+      "grad_norm": 2.768514712096848,
+      "learning_rate": 9.999283190352768e-06,
+      "loss": 1.0069,
+      "step": 1830
+    },
+    {
+      "epoch": 0.05542168674698795,
+      "grad_norm": 3.858630636143342,
+      "learning_rate": 9.999196381821334e-06,
+      "loss": 1.1378,
+      "step": 1840
+    },
+    {
+      "epoch": 0.05572289156626506,
+      "grad_norm": 3.378119057719515,
+      "learning_rate": 9.99910461334869e-06,
+      "loss": 0.9906,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05602409638554217,
+      "grad_norm": 3.001219254715605,
+      "learning_rate": 9.999007885025883e-06,
+      "loss": 1.0809,
+      "step": 1860
+    },
+    {
+      "epoch": 0.056325301204819274,
+      "grad_norm": 3.5393439832915425,
+      "learning_rate": 9.998906196948884e-06,
+      "loss": 0.9938,
+      "step": 1870
+    },
+    {
+      "epoch": 0.056626506024096385,
+      "grad_norm": 3.7850574927214007,
+      "learning_rate": 9.998799549218581e-06,
+      "loss": 1.2124,
+      "step": 1880
+    },
+    {
+      "epoch": 0.056927710843373495,
+      "grad_norm": 2.4388623812023242,
+      "learning_rate": 9.998687941940784e-06,
+      "loss": 0.9052,
+      "step": 1890
+    },
+    {
+      "epoch": 0.0572289156626506,
+      "grad_norm": 2.5699202902016633,
+      "learning_rate": 9.998571375226224e-06,
+      "loss": 0.932,
+      "step": 1900
+    },
+    {
+      "epoch": 0.05753012048192771,
+      "grad_norm": 2.5661310794398946,
+      "learning_rate": 9.998449849190555e-06,
+      "loss": 1.1105,
+      "step": 1910
+    },
+    {
+      "epoch": 0.05783132530120482,
+      "grad_norm": 3.8414058457846245,
+      "learning_rate": 9.998323363954345e-06,
+      "loss": 1.1434,
+      "step": 1920
+    },
+    {
+      "epoch": 0.058132530120481925,
+      "grad_norm": 3.7746141534608326,
+      "learning_rate": 9.99819191964309e-06,
+      "loss": 1.1568,
+      "step": 1930
+    },
+    {
+      "epoch": 0.058433734939759036,
+      "grad_norm": 4.604015815450119,
+      "learning_rate": 9.998055516387199e-06,
+      "loss": 1.0162,
+      "step": 1940
+    },
+    {
+      "epoch": 0.058734939759036146,
+      "grad_norm": 3.960228642260117,
+      "learning_rate": 9.997914154322005e-06,
+      "loss": 0.9437,
+      "step": 1950
+    },
+    {
+      "epoch": 0.05903614457831325,
+      "grad_norm": 6.1648613306726965,
+      "learning_rate": 9.997767833587759e-06,
+      "loss": 1.182,
+      "step": 1960
+    },
+    {
+      "epoch": 0.05933734939759036,
+      "grad_norm": 2.549375334776112,
+      "learning_rate": 9.997616554329634e-06,
+      "loss": 0.9646,
+      "step": 1970
+    },
+    {
+      "epoch": 0.05963855421686747,
+      "grad_norm": 4.236515755123189,
+      "learning_rate": 9.997460316697722e-06,
+      "loss": 1.0929,
+      "step": 1980
+    },
+    {
+      "epoch": 0.059939759036144576,
+      "grad_norm": 5.0300970638199844,
+      "learning_rate": 9.997299120847035e-06,
+      "loss": 1.2331,
+      "step": 1990
+    },
+    {
+      "epoch": 0.060240963855421686,
+      "grad_norm": 3.1979172866204717,
+      "learning_rate": 9.9971329669375e-06,
+      "loss": 1.0689,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0605421686746988,
+      "grad_norm": 4.921490012301479,
+      "learning_rate": 9.996961855133966e-06,
+      "loss": 1.1924,
+      "step": 2010
+    },
+    {
+      "epoch": 0.0608433734939759,
+      "grad_norm": 5.084404905580949,
+      "learning_rate": 9.996785785606204e-06,
+      "loss": 1.1251,
+      "step": 2020
+    },
+    {
+      "epoch": 0.06114457831325301,
+      "grad_norm": 2.7468146978263994,
+      "learning_rate": 9.9966047585289e-06,
+      "loss": 1.0865,
+      "step": 2030
+    },
+    {
+      "epoch": 0.06144578313253012,
+      "grad_norm": 4.910435596688258,
+      "learning_rate": 9.996418774081658e-06,
+      "loss": 1.0844,
+      "step": 2040
+    },
+    {
+      "epoch": 0.061746987951807226,
+      "grad_norm": 3.4800396481633586,
+      "learning_rate": 9.996227832449003e-06,
+      "loss": 1.1514,
+      "step": 2050
+    },
+    {
+      "epoch": 0.06204819277108434,
+      "grad_norm": 3.65181405646942,
+      "learning_rate": 9.996031933820379e-06,
+      "loss": 1.2223,
+      "step": 2060
+    },
+    {
+      "epoch": 0.06234939759036145,
+      "grad_norm": 3.074297815221104,
+      "learning_rate": 9.995831078390145e-06,
+      "loss": 1.1549,
+      "step": 2070
+    },
+    {
+      "epoch": 0.06265060240963856,
+      "grad_norm": 3.8273029757375583,
+      "learning_rate": 9.995625266357579e-06,
+      "loss": 1.1299,
+      "step": 2080
+    },
+    {
+      "epoch": 0.06295180722891566,
+      "grad_norm": 3.7573449297064365,
+      "learning_rate": 9.995414497926876e-06,
+      "loss": 1.0365,
+      "step": 2090
+    },
+    {
+      "epoch": 0.06325301204819277,
+      "grad_norm": 10.118518034769624,
+      "learning_rate": 9.99519877330715e-06,
+      "loss": 1.0804,
+      "step": 2100
+    },
+    {
+      "epoch": 0.06355421686746988,
+      "grad_norm": 3.190883815111035,
+      "learning_rate": 9.994978092712435e-06,
+      "loss": 1.1677,
+      "step": 2110
+    },
+    {
+      "epoch": 0.06385542168674699,
+      "grad_norm": 3.9555260125968363,
+      "learning_rate": 9.994752456361674e-06,
+      "loss": 1.0777,
+      "step": 2120
+    },
+    {
+      "epoch": 0.06415662650602409,
+      "grad_norm": 2.905744251020082,
+      "learning_rate": 9.994521864478734e-06,
+      "loss": 1.051,
+      "step": 2130
+    },
+    {
+      "epoch": 0.06445783132530121,
+      "grad_norm": 2.320618136489521,
+      "learning_rate": 9.994286317292396e-06,
+      "loss": 1.0803,
+      "step": 2140
+    },
+    {
+      "epoch": 0.06475903614457831,
+      "grad_norm": 2.4877820734765046,
+      "learning_rate": 9.994045815036357e-06,
+      "loss": 0.9119,
+      "step": 2150
+    },
+    {
+      "epoch": 0.06506024096385542,
+      "grad_norm": 3.151171444449592,
+      "learning_rate": 9.993800357949231e-06,
+      "loss": 1.1343,
+      "step": 2160
+    },
+    {
+      "epoch": 0.06536144578313254,
+      "grad_norm": 2.506532317945994,
+      "learning_rate": 9.993549946274549e-06,
+      "loss": 1.1076,
+      "step": 2170
+    },
+    {
+      "epoch": 0.06566265060240964,
+      "grad_norm": 2.3254966136883524,
+      "learning_rate": 9.993294580260754e-06,
+      "loss": 1.0453,
+      "step": 2180
+    },
+    {
+      "epoch": 0.06596385542168674,
+      "grad_norm": 3.1805029580613837,
+      "learning_rate": 9.993034260161208e-06,
+      "loss": 1.117,
+      "step": 2190
+    },
+    {
+      "epoch": 0.06626506024096386,
+      "grad_norm": 2.9968452761944553,
+      "learning_rate": 9.992768986234187e-06,
+      "loss": 1.1851,
+      "step": 2200
+    },
+    {
+      "epoch": 0.06656626506024096,
+      "grad_norm": 2.434212792136939,
+      "learning_rate": 9.992498758742882e-06,
+      "loss": 1.0131,
+      "step": 2210
+    },
+    {
+      "epoch": 0.06686746987951807,
+      "grad_norm": 3.349590146400896,
+      "learning_rate": 9.992223577955398e-06,
+      "loss": 1.1821,
+      "step": 2220
+    },
+    {
+      "epoch": 0.06716867469879519,
+      "grad_norm": 3.764825932197075,
+      "learning_rate": 9.991943444144758e-06,
+      "loss": 1.0021,
+      "step": 2230
+    },
+    {
+      "epoch": 0.06746987951807229,
+      "grad_norm": 2.592028510712079,
+      "learning_rate": 9.991658357588892e-06,
+      "loss": 1.0018,
+      "step": 2240
+    },
+    {
+      "epoch": 0.0677710843373494,
+      "grad_norm": 3.1938995334707863,
+      "learning_rate": 9.991368318570648e-06,
+      "loss": 0.9243,
+      "step": 2250
+    },
+    {
+      "epoch": 0.06807228915662651,
+      "grad_norm": 3.595601219077997,
+      "learning_rate": 9.99107332737779e-06,
+      "loss": 1.0054,
+      "step": 2260
+    },
+    {
+      "epoch": 0.06837349397590362,
+      "grad_norm": 2.3885187420780536,
+      "learning_rate": 9.990773384302992e-06,
+      "loss": 1.0097,
+      "step": 2270
+    },
+    {
+      "epoch": 0.06867469879518072,
+      "grad_norm": 2.4930420792745798,
+      "learning_rate": 9.990468489643842e-06,
+      "loss": 1.0304,
+      "step": 2280
+    },
+    {
+      "epoch": 0.06897590361445784,
+      "grad_norm": 2.5449190333509093,
+      "learning_rate": 9.99015864370284e-06,
+      "loss": 1.007,
+      "step": 2290
+    },
+    {
+      "epoch": 0.06927710843373494,
+      "grad_norm": 3.5462243746153477,
+      "learning_rate": 9.9898438467874e-06,
+      "loss": 1.0589,
+      "step": 2300
+    },
+    {
+      "epoch": 0.06957831325301204,
+      "grad_norm": 2.7165842252202994,
+      "learning_rate": 9.989524099209846e-06,
+      "loss": 1.1116,
+      "step": 2310
+    },
+    {
+      "epoch": 0.06987951807228916,
+      "grad_norm": 3.1862255813692397,
+      "learning_rate": 9.989199401287415e-06,
+      "loss": 1.1819,
+      "step": 2320
+    },
+    {
+      "epoch": 0.07018072289156627,
+      "grad_norm": 2.6637679123123466,
+      "learning_rate": 9.988869753342256e-06,
+      "loss": 0.9118,
+      "step": 2330
+    },
+    {
+      "epoch": 0.07048192771084337,
+      "grad_norm": 3.2336481003567537,
+      "learning_rate": 9.988535155701427e-06,
+      "loss": 1.1251,
+      "step": 2340
+    },
+    {
+      "epoch": 0.07078313253012049,
+      "grad_norm": 3.5066824578309967,
+      "learning_rate": 9.9881956086969e-06,
+      "loss": 1.038,
+      "step": 2350
+    },
+    {
+      "epoch": 0.07108433734939759,
+      "grad_norm": 3.7795574130684857,
+      "learning_rate": 9.987851112665554e-06,
+      "loss": 1.0597,
+      "step": 2360
+    },
+    {
+      "epoch": 0.0713855421686747,
+      "grad_norm": 4.5414605529212295,
+      "learning_rate": 9.98750166794918e-06,
+      "loss": 1.0951,
+      "step": 2370
+    },
+    {
+      "epoch": 0.07168674698795181,
+      "grad_norm": 2.585582041219885,
+      "learning_rate": 9.987147274894482e-06,
+      "loss": 0.9117,
+      "step": 2380
+    },
+    {
+      "epoch": 0.07198795180722892,
+      "grad_norm": 4.080742697926455,
+      "learning_rate": 9.986787933853066e-06,
+      "loss": 0.9999,
+      "step": 2390
+    },
+    {
+      "epoch": 0.07228915662650602,
+      "grad_norm": 2.7035924145421633,
+      "learning_rate": 9.986423645181453e-06,
+      "loss": 1.0319,
+      "step": 2400
+    },
+    {
+      "epoch": 0.07259036144578314,
+      "grad_norm": 2.491404655175915,
+      "learning_rate": 9.986054409241073e-06,
+      "loss": 1.0551,
+      "step": 2410
+    },
+    {
+      "epoch": 0.07289156626506024,
+      "grad_norm": 3.6349460804068268,
+      "learning_rate": 9.985680226398261e-06,
+      "loss": 1.0337,
+      "step": 2420
+    },
+    {
+      "epoch": 0.07319277108433735,
+      "grad_norm": 3.10110982779257,
+      "learning_rate": 9.98530109702426e-06,
+      "loss": 1.1683,
+      "step": 2430
+    },
+    {
+      "epoch": 0.07349397590361446,
+      "grad_norm": 3.8403291806956057,
+      "learning_rate": 9.984917021495226e-06,
+      "loss": 1.1914,
+      "step": 2440
+    },
+    {
+      "epoch": 0.07379518072289157,
+      "grad_norm": 2.555025383200647,
+      "learning_rate": 9.984528000192216e-06,
+      "loss": 1.1094,
+      "step": 2450
+    },
+    {
+      "epoch": 0.07409638554216867,
+      "grad_norm": 3.605108875645977,
+      "learning_rate": 9.984134033501198e-06,
+      "loss": 1.0844,
+      "step": 2460
+    },
+    {
+      "epoch": 0.07439759036144579,
+      "grad_norm": 3.4843104645151812,
+      "learning_rate": 9.983735121813043e-06,
+      "loss": 1.0617,
+      "step": 2470
+    },
+    {
+      "epoch": 0.0746987951807229,
+      "grad_norm": 3.2471284257271167,
+      "learning_rate": 9.983331265523532e-06,
+      "loss": 1.1092,
+      "step": 2480
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 1.9449811473314356,
+      "learning_rate": 9.98292246503335e-06,
+      "loss": 0.998,
+      "step": 2490
+    },
+    {
+      "epoch": 0.07530120481927711,
+      "grad_norm": 3.941148943349073,
+      "learning_rate": 9.982508720748087e-06,
+      "loss": 1.2207,
+      "step": 2500
+    },
+    {
+      "epoch": 0.07560240963855422,
+      "grad_norm": 3.4412112592806383,
+      "learning_rate": 9.98209003307824e-06,
+      "loss": 1.0887,
+      "step": 2510
+    },
+    {
+      "epoch": 0.07590361445783132,
+      "grad_norm": 4.818457829096565,
+      "learning_rate": 9.981666402439206e-06,
+      "loss": 1.0562,
+      "step": 2520
+    },
+    {
+      "epoch": 0.07620481927710844,
+      "grad_norm": 3.80293562021422,
+      "learning_rate": 9.981237829251293e-06,
+      "loss": 1.0833,
+      "step": 2530
+    },
+    {
+      "epoch": 0.07650602409638554,
+      "grad_norm": 2.4087631061355363,
+      "learning_rate": 9.980804313939704e-06,
+      "loss": 1.0543,
+      "step": 2540
+    },
+    {
+      "epoch": 0.07680722891566265,
+      "grad_norm": 3.5487519559148266,
+      "learning_rate": 9.980365856934555e-06,
+      "loss": 0.9845,
+      "step": 2550
+    },
+    {
+      "epoch": 0.07710843373493977,
+      "grad_norm": 2.4850575123163727,
+      "learning_rate": 9.97992245867086e-06,
+      "loss": 0.9657,
+      "step": 2560
+    },
+    {
+      "epoch": 0.07740963855421687,
+      "grad_norm": 3.555607525747383,
+      "learning_rate": 9.97947411958853e-06,
+      "loss": 1.1084,
+      "step": 2570
+    },
+    {
+      "epoch": 0.07771084337349397,
+      "grad_norm": 2.40737881814081,
+      "learning_rate": 9.979020840132389e-06,
+      "loss": 1.1145,
+      "step": 2580
+    },
+    {
+      "epoch": 0.07801204819277109,
+      "grad_norm": 2.9290379516100717,
+      "learning_rate": 9.978562620752155e-06,
+      "loss": 1.1361,
+      "step": 2590
+    },
+    {
+      "epoch": 0.0783132530120482,
+      "grad_norm": 4.920614149434376,
+      "learning_rate": 9.978099461902449e-06,
+      "loss": 1.0711,
+      "step": 2600
+    },
+    {
+      "epoch": 0.0786144578313253,
+      "grad_norm": 2.396143608817259,
+      "learning_rate": 9.977631364042796e-06,
+      "loss": 0.9605,
+      "step": 2610
+    },
+    {
+      "epoch": 0.0789156626506024,
+      "grad_norm": 21.085877420807975,
+      "learning_rate": 9.977158327637614e-06,
+      "loss": 0.9583,
+      "step": 2620
+    },
+    {
+      "epoch": 0.07921686746987952,
+      "grad_norm": 2.294929049480722,
+      "learning_rate": 9.97668035315623e-06,
+      "loss": 1.0063,
+      "step": 2630
+    },
+    {
+      "epoch": 0.07951807228915662,
+      "grad_norm": 6.789153018446023,
+      "learning_rate": 9.976197441072859e-06,
+      "loss": 1.1275,
+      "step": 2640
+    },
+    {
+      "epoch": 0.07981927710843373,
+      "grad_norm": 7.742954844534115,
+      "learning_rate": 9.975709591866625e-06,
+      "loss": 1.1735,
+      "step": 2650
+    },
+    {
+      "epoch": 0.08012048192771085,
+      "grad_norm": 16.04533601383964,
+      "learning_rate": 9.975216806021546e-06,
+      "loss": 1.1783,
+      "step": 2660
+    },
+    {
+      "epoch": 0.08042168674698795,
+      "grad_norm": 2.1758648326982875,
+      "learning_rate": 9.974719084026539e-06,
+      "loss": 1.0405,
+      "step": 2670
+    },
+    {
+      "epoch": 0.08072289156626505,
+      "grad_norm": 6.064945451747976,
+      "learning_rate": 9.974216426375415e-06,
+      "loss": 1.0942,
+      "step": 2680
+    },
+    {
+      "epoch": 0.08102409638554217,
+      "grad_norm": 8.445513916077669,
+      "learning_rate": 9.973708833566888e-06,
+      "loss": 1.1507,
+      "step": 2690
+    },
+    {
+      "epoch": 0.08132530120481928,
+      "grad_norm": 7.596599663105275,
+      "learning_rate": 9.973196306104565e-06,
+      "loss": 1.0625,
+      "step": 2700
+    },
+    {
+      "epoch": 0.08162650602409638,
+      "grad_norm": 3.8107562681697886,
+      "learning_rate": 9.972678844496946e-06,
+      "loss": 1.054,
+      "step": 2710
+    },
+    {
+      "epoch": 0.0819277108433735,
+      "grad_norm": 2.60289693386569,
+      "learning_rate": 9.972156449257431e-06,
+      "loss": 1.1578,
+      "step": 2720
+    },
+    {
+      "epoch": 0.0822289156626506,
+      "grad_norm": 34.67041330752499,
+      "learning_rate": 9.971629120904311e-06,
+      "loss": 0.9634,
+      "step": 2730
+    },
+    {
+      "epoch": 0.0825301204819277,
+      "grad_norm": 4.124388244266611,
+      "learning_rate": 9.971096859960778e-06,
+      "loss": 1.083,
+      "step": 2740
+    },
+    {
+      "epoch": 0.08283132530120482,
+      "grad_norm": 2.5613783895358324,
+      "learning_rate": 9.970559666954912e-06,
+      "loss": 1.0909,
+      "step": 2750
+    },
+    {
+      "epoch": 0.08313253012048193,
+      "grad_norm": 2.4217897863699687,
+      "learning_rate": 9.970017542419685e-06,
+      "loss": 1.0532,
+      "step": 2760
+    },
+    {
+      "epoch": 0.08343373493975903,
+      "grad_norm": 3.489663688491791,
+      "learning_rate": 9.969470486892969e-06,
+      "loss": 1.059,
+      "step": 2770
+    },
+    {
+      "epoch": 0.08373493975903615,
+      "grad_norm": 3.441237385708804,
+      "learning_rate": 9.968918500917519e-06,
+      "loss": 1.1639,
+      "step": 2780
+    },
+    {
+      "epoch": 0.08403614457831325,
+      "grad_norm": 2.423276371430195,
+      "learning_rate": 9.96836158504099e-06,
+      "loss": 0.9466,
+      "step": 2790
+    },
+    {
+      "epoch": 0.08433734939759036,
+      "grad_norm": 3.289694596045945,
+      "learning_rate": 9.967799739815925e-06,
+      "loss": 0.9739,
+      "step": 2800
+    },
+    {
+      "epoch": 0.08463855421686747,
+      "grad_norm": 3.9050353977577235,
+      "learning_rate": 9.967232965799756e-06,
+      "loss": 1.0265,
+      "step": 2810
+    },
+    {
+      "epoch": 0.08493975903614458,
+      "grad_norm": 2.532904826528985,
+      "learning_rate": 9.966661263554807e-06,
+      "loss": 1.1686,
+      "step": 2820
+    },
+    {
+      "epoch": 0.08524096385542168,
+      "grad_norm": 2.3416221435459534,
+      "learning_rate": 9.96608463364829e-06,
+      "loss": 1.0797,
+      "step": 2830
+    },
+    {
+      "epoch": 0.0855421686746988,
+      "grad_norm": 3.0805759527658867,
+      "learning_rate": 9.965503076652309e-06,
+      "loss": 1.1602,
+      "step": 2840
+    },
+    {
+      "epoch": 0.0858433734939759,
+      "grad_norm": 2.887601186210108,
+      "learning_rate": 9.964916593143851e-06,
+      "loss": 1.1548,
+      "step": 2850
+    },
+    {
+      "epoch": 0.086144578313253,
+      "grad_norm": 3.3377312336528187,
+      "learning_rate": 9.964325183704801e-06,
+      "loss": 1.1498,
+      "step": 2860
+    },
+    {
+      "epoch": 0.08644578313253012,
+      "grad_norm": 3.7551887104103807,
+      "learning_rate": 9.963728848921918e-06,
+      "loss": 1.133,
+      "step": 2870
+    },
+    {
+      "epoch": 0.08674698795180723,
+      "grad_norm": 3.927673372140471,
+      "learning_rate": 9.963127589386856e-06,
+      "loss": 1.1079,
+      "step": 2880
+    },
+    {
+      "epoch": 0.08704819277108433,
+      "grad_norm": 3.646635001753062,
+      "learning_rate": 9.962521405696155e-06,
+      "loss": 1.1541,
+      "step": 2890
+    },
+    {
+      "epoch": 0.08734939759036145,
+      "grad_norm": 2.2738856304274138,
+      "learning_rate": 9.961910298451237e-06,
+      "loss": 1.0623,
+      "step": 2900
+    },
+    {
+      "epoch": 0.08765060240963855,
+      "grad_norm": 2.5961700941930217,
+      "learning_rate": 9.96129426825841e-06,
+      "loss": 1.054,
+      "step": 2910
+    },
+    {
+      "epoch": 0.08795180722891566,
+      "grad_norm": 3.151722234942918,
+      "learning_rate": 9.960673315728869e-06,
+      "loss": 1.0582,
+      "step": 2920
+    },
+    {
+      "epoch": 0.08825301204819277,
+      "grad_norm": 2.4318221200620447,
+      "learning_rate": 9.960047441478688e-06,
+      "loss": 1.0786,
+      "step": 2930
+    },
+    {
+      "epoch": 0.08855421686746988,
+      "grad_norm": 3.2436328140102466,
+      "learning_rate": 9.959416646128832e-06,
+      "loss": 0.9924,
+      "step": 2940
+    },
+    {
+      "epoch": 0.08885542168674698,
+      "grad_norm": 3.4902327614333366,
+      "learning_rate": 9.958780930305136e-06,
+      "loss": 1.0438,
+      "step": 2950
+    },
+    {
+      "epoch": 0.0891566265060241,
+      "grad_norm": 2.529586079380723,
+      "learning_rate": 9.95814029463833e-06,
+      "loss": 1.0668,
+      "step": 2960
+    },
+    {
+      "epoch": 0.0894578313253012,
+      "grad_norm": 3.5179299925095635,
+      "learning_rate": 9.957494739764015e-06,
+      "loss": 1.1369,
+      "step": 2970
+    },
+    {
+      "epoch": 0.08975903614457831,
+      "grad_norm": 3.209633091677148,
+      "learning_rate": 9.95684426632268e-06,
+      "loss": 1.0598,
+      "step": 2980
+    },
+    {
+      "epoch": 0.09006024096385543,
+      "grad_norm": 2.870985968805948,
+      "learning_rate": 9.956188874959686e-06,
+      "loss": 1.1513,
+      "step": 2990
+    },
+    {
+      "epoch": 0.09036144578313253,
+      "grad_norm": 3.458606378245846,
+      "learning_rate": 9.955528566325285e-06,
+      "loss": 1.0767,
+      "step": 3000
+    },
+    {
+      "epoch": 0.09066265060240963,
+      "grad_norm": 3.44900113672603,
+      "learning_rate": 9.954863341074597e-06,
+      "loss": 1.0477,
+      "step": 3010
+    },
+    {
+      "epoch": 0.09096385542168675,
+      "grad_norm": 2.4691131879834836,
+      "learning_rate": 9.954193199867622e-06,
+      "loss": 0.9465,
+      "step": 3020
+    },
+    {
+      "epoch": 0.09126506024096386,
+      "grad_norm": 3.452353799076586,
+      "learning_rate": 9.95351814336924e-06,
+      "loss": 1.0809,
+      "step": 3030
+    },
+    {
+      "epoch": 0.09156626506024096,
+      "grad_norm": 3.5876569387258,
+      "learning_rate": 9.95283817224921e-06,
+      "loss": 1.2361,
+      "step": 3040
+    },
+    {
+      "epoch": 0.09186746987951808,
+      "grad_norm": 2.3053326000738608,
+      "learning_rate": 9.952153287182161e-06,
+      "loss": 1.0254,
+      "step": 3050
+    },
+    {
+      "epoch": 0.09216867469879518,
+      "grad_norm": 3.312836075166249,
+      "learning_rate": 9.951463488847598e-06,
+      "loss": 1.0892,
+      "step": 3060
+    },
+    {
+      "epoch": 0.09246987951807228,
+      "grad_norm": 3.3874341809742066,
+      "learning_rate": 9.950768777929906e-06,
+      "loss": 0.9769,
+      "step": 3070
+    },
+    {
+      "epoch": 0.0927710843373494,
+      "grad_norm": 4.679167740714582,
+      "learning_rate": 9.950069155118341e-06,
+      "loss": 1.1189,
+      "step": 3080
+    },
+    {
+      "epoch": 0.0930722891566265,
+      "grad_norm": 3.6434603101544174,
+      "learning_rate": 9.949364621107032e-06,
+      "loss": 1.1087,
+      "step": 3090
+    },
+    {
+      "epoch": 0.09337349397590361,
+      "grad_norm": 4.565814160232158,
+      "learning_rate": 9.948655176594979e-06,
+      "loss": 1.1342,
+      "step": 3100
+    },
+    {
+      "epoch": 0.09367469879518073,
+      "grad_norm": 2.300775890404092,
+      "learning_rate": 9.947940822286058e-06,
+      "loss": 1.0865,
+      "step": 3110
+    },
+    {
+      "epoch": 0.09397590361445783,
+      "grad_norm": 3.7003026630602194,
+      "learning_rate": 9.947221558889011e-06,
+      "loss": 1.0942,
+      "step": 3120
+    },
+    {
+      "epoch": 0.09427710843373494,
+      "grad_norm": 3.2932522022342354,
+      "learning_rate": 9.946497387117456e-06,
+      "loss": 1.0463,
+      "step": 3130
+    },
+    {
+      "epoch": 0.09457831325301205,
+      "grad_norm": 2.2104973463777147,
+      "learning_rate": 9.945768307689878e-06,
+      "loss": 0.9945,
+      "step": 3140
+    },
+    {
+      "epoch": 0.09487951807228916,
+      "grad_norm": 3.454729841387448,
+      "learning_rate": 9.945034321329631e-06,
+      "loss": 1.0486,
+      "step": 3150
+    },
+    {
+      "epoch": 0.09518072289156626,
+      "grad_norm": 2.526763130559842,
+      "learning_rate": 9.944295428764939e-06,
+      "loss": 1.0415,
+      "step": 3160
+    },
+    {
+      "epoch": 0.09548192771084338,
+      "grad_norm": 3.3192206957471155,
+      "learning_rate": 9.94355163072889e-06,
+      "loss": 1.0743,
+      "step": 3170
+    },
+    {
+      "epoch": 0.09578313253012048,
+      "grad_norm": 3.6483424767916173,
+      "learning_rate": 9.942802927959444e-06,
+      "loss": 0.9687,
+      "step": 3180
+    },
+    {
+      "epoch": 0.09608433734939759,
+      "grad_norm": 3.550875633521195,
+      "learning_rate": 9.942049321199423e-06,
+      "loss": 1.1543,
+      "step": 3190
+    },
+    {
+      "epoch": 0.0963855421686747,
+      "grad_norm": 3.29660823020748,
+      "learning_rate": 9.941290811196518e-06,
+      "loss": 1.167,
+      "step": 3200
+    },
+    {
+      "epoch": 0.09668674698795181,
+      "grad_norm": 3.3986787350060603,
+      "learning_rate": 9.940527398703282e-06,
+      "loss": 1.0601,
+      "step": 3210
+    },
+    {
+      "epoch": 0.09698795180722891,
+      "grad_norm": 2.9548349101010243,
+      "learning_rate": 9.939759084477132e-06,
+      "loss": 1.1153,
+      "step": 3220
+    },
+    {
+      "epoch": 0.09728915662650603,
+      "grad_norm": 2.7975966103960115,
+      "learning_rate": 9.938985869280352e-06,
+      "loss": 1.0727,
+      "step": 3230
+    },
+    {
+      "epoch": 0.09759036144578313,
+      "grad_norm": 3.224219184432969,
+      "learning_rate": 9.938207753880082e-06,
+      "loss": 1.0709,
+      "step": 3240
+    },
+    {
+      "epoch": 0.09789156626506024,
+      "grad_norm": 4.89642920300496,
+      "learning_rate": 9.937424739048332e-06,
+      "loss": 1.0952,
+      "step": 3250
+    },
+    {
+      "epoch": 0.09819277108433735,
+      "grad_norm": 2.3077224024701293,
+      "learning_rate": 9.936636825561962e-06,
+      "loss": 1.077,
+      "step": 3260
+    },
+    {
+      "epoch": 0.09849397590361446,
+      "grad_norm": 3.525517406827571,
+      "learning_rate": 9.935844014202704e-06,
+      "loss": 1.09,
+      "step": 3270
+    },
+    {
+      "epoch": 0.09879518072289156,
+      "grad_norm": 3.3837534625306174,
+      "learning_rate": 9.935046305757144e-06,
+      "loss": 1.1552,
+      "step": 3280
+    },
+    {
+      "epoch": 0.09909638554216868,
+      "grad_norm": 3.3450320696159404,
+      "learning_rate": 9.934243701016722e-06,
+      "loss": 1.1078,
+      "step": 3290
+    },
+    {
+      "epoch": 0.09939759036144578,
+      "grad_norm": 3.2315444842214975,
+      "learning_rate": 9.933436200777744e-06,
+      "loss": 1.0719,
+      "step": 3300
+    },
+    {
+      "epoch": 0.09969879518072289,
+      "grad_norm": 3.681112830375248,
+      "learning_rate": 9.932623805841369e-06,
+      "loss": 1.0665,
+      "step": 3310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4855398408295777,
+      "learning_rate": 9.931806517013612e-06,
+      "loss": 0.9008,
+      "step": 3320
+    },
+    {
+      "epoch": 0.10030120481927711,
+      "grad_norm": 4.2067150888802365,
+      "learning_rate": 9.930984335105346e-06,
+      "loss": 1.0229,
+      "step": 3330
+    },
+    {
+      "epoch": 0.10060240963855421,
+      "grad_norm": 3.6811653466226653,
+      "learning_rate": 9.930157260932295e-06,
+      "loss": 0.9746,
+      "step": 3340
+    },
+    {
+      "epoch": 0.10090361445783133,
+      "grad_norm": 3.867774500184461,
+      "learning_rate": 9.929325295315039e-06,
+      "loss": 1.1569,
+      "step": 3350
+    },
+    {
+      "epoch": 0.10120481927710843,
+      "grad_norm": 3.2169003704498436,
+      "learning_rate": 9.928488439079012e-06,
+      "loss": 1.1525,
+      "step": 3360
+    },
+    {
+      "epoch": 0.10150602409638554,
+      "grad_norm": 2.429935006013882,
+      "learning_rate": 9.927646693054498e-06,
+      "loss": 0.9991,
+      "step": 3370
+    },
+    {
+      "epoch": 0.10180722891566266,
+      "grad_norm": 2.4064532640598353,
+      "learning_rate": 9.92680005807663e-06,
+      "loss": 1.0455,
+      "step": 3380
+    },
+    {
+      "epoch": 0.10210843373493976,
+      "grad_norm": 5.643497312447238,
+      "learning_rate": 9.925948534985404e-06,
+      "loss": 1.0037,
+      "step": 3390
+    },
+    {
+      "epoch": 0.10240963855421686,
+      "grad_norm": 2.4185176278194738,
+      "learning_rate": 9.925092124625649e-06,
+      "loss": 0.958,
+      "step": 3400
+    },
+    {
+      "epoch": 0.10271084337349398,
+      "grad_norm": 2.427364695140659,
+      "learning_rate": 9.92423082784705e-06,
+      "loss": 1.0585,
+      "step": 3410
+    },
+    {
+      "epoch": 0.10301204819277109,
+      "grad_norm": 5.887559987596885,
+      "learning_rate": 9.923364645504147e-06,
+      "loss": 1.0812,
+      "step": 3420
+    },
+    {
+      "epoch": 0.10331325301204819,
+      "grad_norm": 4.6406899191484685,
+      "learning_rate": 9.922493578456316e-06,
+      "loss": 1.0793,
+      "step": 3430
+    },
+    {
+      "epoch": 0.10361445783132531,
+      "grad_norm": 4.74836161300959,
+      "learning_rate": 9.921617627567786e-06,
+      "loss": 1.0573,
+      "step": 3440
+    },
+    {
+      "epoch": 0.10391566265060241,
+      "grad_norm": 3.8895481633008626,
+      "learning_rate": 9.920736793707627e-06,
+      "loss": 1.0817,
+      "step": 3450
+    },
+    {
+      "epoch": 0.10421686746987951,
+      "grad_norm": 2.2925826881912723,
+      "learning_rate": 9.91985107774976e-06,
+      "loss": 0.9974,
+      "step": 3460
+    },
+    {
+      "epoch": 0.10451807228915663,
+      "grad_norm": 2.500834179568333,
+      "learning_rate": 9.918960480572943e-06,
+      "loss": 1.161,
+      "step": 3470
+    },
+    {
+      "epoch": 0.10481927710843374,
+      "grad_norm": 3.140445484276692,
+      "learning_rate": 9.918065003060782e-06,
+      "loss": 0.933,
+      "step": 3480
+    },
+    {
+      "epoch": 0.10512048192771084,
+      "grad_norm": 2.0696270131923695,
+      "learning_rate": 9.917164646101723e-06,
+      "loss": 0.9928,
+      "step": 3490
+    },
+    {
+      "epoch": 0.10542168674698796,
+      "grad_norm": 3.5201557272388504,
+      "learning_rate": 9.916259410589051e-06,
+      "loss": 0.9544,
+      "step": 3500
+    },
+    {
+      "epoch": 0.10572289156626506,
+      "grad_norm": 3.4034684537817044,
+      "learning_rate": 9.915349297420895e-06,
+      "loss": 1.0712,
+      "step": 3510
+    },
+    {
+      "epoch": 0.10602409638554217,
+      "grad_norm": 3.8292368668618546,
+      "learning_rate": 9.914434307500223e-06,
+      "loss": 1.0634,
+      "step": 3520
+    },
+    {
+      "epoch": 0.10632530120481928,
+      "grad_norm": 3.217098721578885,
+      "learning_rate": 9.913514441734837e-06,
+      "loss": 1.1015,
+      "step": 3530
+    },
+    {
+      "epoch": 0.10662650602409639,
+      "grad_norm": 2.215923928809928,
+      "learning_rate": 9.91258970103738e-06,
+      "loss": 0.9373,
+      "step": 3540
+    },
+    {
+      "epoch": 0.10692771084337349,
+      "grad_norm": 3.249713627496291,
+      "learning_rate": 9.911660086325335e-06,
+      "loss": 1.0283,
+      "step": 3550
+    },
+    {
+      "epoch": 0.10722891566265061,
+      "grad_norm": 3.2075055711658504,
+      "learning_rate": 9.910725598521014e-06,
+      "loss": 1.1366,
+      "step": 3560
+    },
+    {
+      "epoch": 0.10753012048192771,
+      "grad_norm": 3.1526632646499437,
+      "learning_rate": 9.909786238551567e-06,
+      "loss": 1.0981,
+      "step": 3570
+    },
+    {
+      "epoch": 0.10783132530120482,
+      "grad_norm": 4.012541444087307,
+      "learning_rate": 9.90884200734898e-06,
+      "loss": 1.077,
+      "step": 3580
+    },
+    {
+      "epoch": 0.10813253012048193,
+      "grad_norm": 3.412507998567453,
+      "learning_rate": 9.907892905850067e-06,
+      "loss": 1.0265,
+      "step": 3590
+    },
+    {
+      "epoch": 0.10843373493975904,
+      "grad_norm": 3.3760447554613493,
+      "learning_rate": 9.906938934996477e-06,
+      "loss": 0.9877,
+      "step": 3600
+    },
+    {
+      "epoch": 0.10873493975903614,
+      "grad_norm": 3.152908418084381,
+      "learning_rate": 9.905980095734693e-06,
+      "loss": 1.0545,
+      "step": 3610
+    },
+    {
+      "epoch": 0.10903614457831326,
+      "grad_norm": 4.9422576933511335,
+      "learning_rate": 9.905016389016022e-06,
+      "loss": 1.0797,
+      "step": 3620
+    },
+    {
+      "epoch": 0.10933734939759036,
+      "grad_norm": 2.5713962627640394,
+      "learning_rate": 9.904047815796606e-06,
+      "loss": 1.0536,
+      "step": 3630
+    },
+    {
+      "epoch": 0.10963855421686747,
+      "grad_norm": 3.8361775301205445,
+      "learning_rate": 9.90307437703741e-06,
+      "loss": 1.0602,
+      "step": 3640
+    },
+    {
+      "epoch": 0.10993975903614457,
+      "grad_norm": 4.388280339546095,
+      "learning_rate": 9.902096073704232e-06,
+      "loss": 1.1544,
+      "step": 3650
+    },
+    {
+      "epoch": 0.11024096385542169,
+      "grad_norm": 3.53370160147256,
+      "learning_rate": 9.901112906767691e-06,
+      "loss": 1.1583,
+      "step": 3660
+    },
+    {
+      "epoch": 0.11054216867469879,
+      "grad_norm": 5.587251441249992,
+      "learning_rate": 9.900124877203234e-06,
+      "loss": 1.0497,
+      "step": 3670
+    },
+    {
+      "epoch": 0.1108433734939759,
+      "grad_norm": 2.499547777526994,
+      "learning_rate": 9.899131985991135e-06,
+      "loss": 0.9369,
+      "step": 3680
+    },
+    {
+      "epoch": 0.11114457831325301,
+      "grad_norm": 7.181643890173013,
+      "learning_rate": 9.898134234116484e-06,
+      "loss": 0.9935,
+      "step": 3690
+    },
+    {
+      "epoch": 0.11144578313253012,
+      "grad_norm": 4.832575337671559,
+      "learning_rate": 9.897131622569203e-06,
+      "loss": 1.1319,
+      "step": 3700
+    },
+    {
+      "epoch": 0.11174698795180722,
+      "grad_norm": 3.3856814879475814,
+      "learning_rate": 9.896124152344026e-06,
+      "loss": 1.0827,
+      "step": 3710
+    },
+    {
+      "epoch": 0.11204819277108434,
+      "grad_norm": 3.821240493677679,
+      "learning_rate": 9.895111824440518e-06,
+      "loss": 0.96,
+      "step": 3720
+    },
+    {
+      "epoch": 0.11234939759036144,
+      "grad_norm": 2.2862337649984044,
+      "learning_rate": 9.894094639863052e-06,
+      "loss": 1.075,
+      "step": 3730
+    },
+    {
+      "epoch": 0.11265060240963855,
+      "grad_norm": 3.7350976360281924,
+      "learning_rate": 9.89307259962083e-06,
+      "loss": 1.1295,
+      "step": 3740
+    },
+    {
+      "epoch": 0.11295180722891567,
+      "grad_norm": 5.33116955217106,
+      "learning_rate": 9.892045704727864e-06,
+      "loss": 1.0976,
+      "step": 3750
+    },
+    {
+      "epoch": 0.11325301204819277,
+      "grad_norm": 4.9096763724568175,
+      "learning_rate": 9.891013956202986e-06,
+      "loss": 1.0943,
+      "step": 3760
+    },
+    {
+      "epoch": 0.11355421686746987,
+      "grad_norm": 5.946510447864238,
+      "learning_rate": 9.889977355069846e-06,
+      "loss": 0.9954,
+      "step": 3770
+    },
+    {
+      "epoch": 0.11385542168674699,
+      "grad_norm": 2.5896161633818893,
+      "learning_rate": 9.8889359023569e-06,
+      "loss": 1.0303,
+      "step": 3780
+    },
+    {
+      "epoch": 0.1141566265060241,
+      "grad_norm": 2.682308487012938,
+      "learning_rate": 9.887889599097427e-06,
+      "loss": 0.9402,
+      "step": 3790
+    },
+    {
+      "epoch": 0.1144578313253012,
+      "grad_norm": 8.090386094644277,
+      "learning_rate": 9.886838446329513e-06,
+      "loss": 0.993,
+      "step": 3800
+    },
+    {
+      "epoch": 0.11475903614457832,
+      "grad_norm": 4.9136633248414,
+      "learning_rate": 9.885782445096054e-06,
+      "loss": 1.0911,
+      "step": 3810
+    },
+    {
+      "epoch": 0.11506024096385542,
+      "grad_norm": 11.66087514078214,
+      "learning_rate": 9.884721596444765e-06,
+      "loss": 1.0986,
+      "step": 3820
+    },
+    {
+      "epoch": 0.11536144578313252,
+      "grad_norm": 2.803940405000577,
+      "learning_rate": 9.88365590142816e-06,
+      "loss": 0.974,
+      "step": 3830
+    },
+    {
+      "epoch": 0.11566265060240964,
+      "grad_norm": 2.0741399724434197,
+      "learning_rate": 9.882585361103567e-06,
+      "loss": 0.9336,
+      "step": 3840
+    },
+    {
+      "epoch": 0.11596385542168675,
+      "grad_norm": 4.285118614701219,
+      "learning_rate": 9.881509976533122e-06,
+      "loss": 1.0348,
+      "step": 3850
+    },
+    {
+      "epoch": 0.11626506024096385,
+      "grad_norm": 6.401548377417817,
+      "learning_rate": 9.880429748783765e-06,
+      "loss": 1.084,
+      "step": 3860
+    },
+    {
+      "epoch": 0.11656626506024097,
+      "grad_norm": 2.238506957174242,
+      "learning_rate": 9.879344678927237e-06,
+      "loss": 0.9916,
+      "step": 3870
+    },
+    {
+      "epoch": 0.11686746987951807,
+      "grad_norm": 2.1766726826908824,
+      "learning_rate": 9.878254768040094e-06,
+      "loss": 0.9253,
+      "step": 3880
+    },
+    {
+      "epoch": 0.11716867469879517,
+      "grad_norm": 9.605422303557702,
+      "learning_rate": 9.877160017203685e-06,
+      "loss": 1.0702,
+      "step": 3890
+    },
+    {
+      "epoch": 0.11746987951807229,
+      "grad_norm": 6.72304912503531,
+      "learning_rate": 9.876060427504163e-06,
+      "loss": 1.0878,
+      "step": 3900
+    },
+    {
+      "epoch": 0.1177710843373494,
+      "grad_norm": 6.676548799781242,
+      "learning_rate": 9.874956000032486e-06,
+      "loss": 1.0945,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1180722891566265,
+      "grad_norm": 2.2583690591099885,
+      "learning_rate": 9.87384673588441e-06,
+      "loss": 0.9939,
+      "step": 3920
+    },
+    {
+      "epoch": 0.11837349397590362,
+      "grad_norm": 7.146130842348655,
+      "learning_rate": 9.872732636160485e-06,
+      "loss": 1.0195,
+      "step": 3930
+    },
+    {
+      "epoch": 0.11867469879518072,
+      "grad_norm": 7.463148927538527,
+      "learning_rate": 9.871613701966067e-06,
+      "loss": 1.0251,
+      "step": 3940
+    },
+    {
+      "epoch": 0.11897590361445783,
+      "grad_norm": 2.4988058155353463,
+      "learning_rate": 9.870489934411302e-06,
+      "loss": 1.0683,
+      "step": 3950
+    },
+    {
+      "epoch": 0.11927710843373494,
+      "grad_norm": 4.866698415262084,
+      "learning_rate": 9.869361334611134e-06,
+      "loss": 1.1394,
+      "step": 3960
+    },
+    {
+      "epoch": 0.11957831325301205,
+      "grad_norm": 6.302578574320336,
+      "learning_rate": 9.8682279036853e-06,
+      "loss": 1.0607,
+      "step": 3970
+    },
+    {
+      "epoch": 0.11987951807228915,
+      "grad_norm": 3.895202506220607,
+      "learning_rate": 9.867089642758329e-06,
+      "loss": 1.1389,
+      "step": 3980
+    },
+    {
+      "epoch": 0.12018072289156627,
+      "grad_norm": 3.3532605127940087,
+      "learning_rate": 9.865946552959546e-06,
+      "loss": 1.0315,
+      "step": 3990
+    },
+    {
+      "epoch": 0.12048192771084337,
+      "grad_norm": 2.287650829129172,
+      "learning_rate": 9.864798635423067e-06,
+      "loss": 1.0001,
+      "step": 4000
+    },
+    {
+      "epoch": 0.12078313253012048,
+      "grad_norm": 3.1202934651229777,
+      "learning_rate": 9.863645891287793e-06,
+      "loss": 1.033,
+      "step": 4010
+    },
+    {
+      "epoch": 0.1210843373493976,
+      "grad_norm": 3.2086966960342784,
+      "learning_rate": 9.862488321697414e-06,
+      "loss": 1.1562,
+      "step": 4020
+    },
+    {
+      "epoch": 0.1213855421686747,
+      "grad_norm": 2.417660159644811,
+      "learning_rate": 9.861325927800415e-06,
+      "loss": 0.9891,
+      "step": 4030
+    },
+    {
+      "epoch": 0.1216867469879518,
+      "grad_norm": 3.2444385489033545,
+      "learning_rate": 9.860158710750062e-06,
+      "loss": 1.1242,
+      "step": 4040
+    },
+    {
+      "epoch": 0.12198795180722892,
+      "grad_norm": 2.234151030168644,
+      "learning_rate": 9.858986671704405e-06,
+      "loss": 0.9638,
+      "step": 4050
+    },
+    {
+      "epoch": 0.12228915662650602,
+      "grad_norm": 3.235452061511477,
+      "learning_rate": 9.857809811826277e-06,
+      "loss": 0.9994,
+      "step": 4060
+    },
+    {
+      "epoch": 0.12259036144578313,
+      "grad_norm": 2.309966366479653,
+      "learning_rate": 9.856628132283303e-06,
+      "loss": 0.9584,
+      "step": 4070
+    },
+    {
+      "epoch": 0.12289156626506025,
+      "grad_norm": 2.814561986665927,
+      "learning_rate": 9.85544163424788e-06,
+      "loss": 1.0083,
+      "step": 4080
+    },
+    {
+      "epoch": 0.12319277108433735,
+      "grad_norm": 4.05930623486242,
+      "learning_rate": 9.854250318897188e-06,
+      "loss": 0.9834,
+      "step": 4090
+    },
+    {
+      "epoch": 0.12349397590361445,
+      "grad_norm": 3.755891333932205,
+      "learning_rate": 9.853054187413192e-06,
+      "loss": 1.1235,
+      "step": 4100
+    },
+    {
+      "epoch": 0.12379518072289157,
+      "grad_norm": 2.310425777436963,
+      "learning_rate": 9.851853240982628e-06,
+      "loss": 1.0613,
+      "step": 4110
+    },
+    {
+      "epoch": 0.12409638554216867,
+      "grad_norm": 3.083914087842926,
+      "learning_rate": 9.850647480797011e-06,
+      "loss": 1.1484,
+      "step": 4120
+    },
+    {
+      "epoch": 0.12439759036144578,
+      "grad_norm": 3.9021725996989582,
+      "learning_rate": 9.849436908052636e-06,
+      "loss": 1.0013,
+      "step": 4130
+    },
+    {
+      "epoch": 0.1246987951807229,
+      "grad_norm": 4.261045854576861,
+      "learning_rate": 9.84822152395057e-06,
+      "loss": 0.988,
+      "step": 4140
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 3.0533766104396562,
+      "learning_rate": 9.847001329696653e-06,
+      "loss": 1.1048,
+      "step": 4150
+    },
+    {
+      "epoch": 0.12530120481927712,
+      "grad_norm": 2.8009853883701235,
+      "learning_rate": 9.845776326501497e-06,
+      "loss": 0.9714,
+      "step": 4160
+    },
+    {
+      "epoch": 0.1256024096385542,
+      "grad_norm": 3.7077863547315464,
+      "learning_rate": 9.844546515580486e-06,
+      "loss": 1.0074,
+      "step": 4170
+    },
+    {
+      "epoch": 0.12590361445783133,
+      "grad_norm": 3.5851684616175374,
+      "learning_rate": 9.843311898153775e-06,
+      "loss": 0.9621,
+      "step": 4180
+    },
+    {
+      "epoch": 0.12620481927710844,
+      "grad_norm": 2.0936533248577427,
+      "learning_rate": 9.842072475446287e-06,
+      "loss": 0.9984,
+      "step": 4190
+    },
+    {
+      "epoch": 0.12650602409638553,
+      "grad_norm": 3.236837577173862,
+      "learning_rate": 9.840828248687714e-06,
+      "loss": 0.9921,
+      "step": 4200
+    },
+    {
+      "epoch": 0.12680722891566265,
+      "grad_norm": 4.302891239940509,
+      "learning_rate": 9.839579219112508e-06,
+      "loss": 1.0077,
+      "step": 4210
+    },
+    {
+      "epoch": 0.12710843373493977,
+      "grad_norm": 4.72174446683809,
+      "learning_rate": 9.838325387959894e-06,
+      "loss": 1.1164,
+      "step": 4220
+    },
+    {
+      "epoch": 0.12740963855421686,
+      "grad_norm": 3.53934622620043,
+      "learning_rate": 9.837066756473858e-06,
+      "loss": 1.1238,
+      "step": 4230
+    },
+    {
+      "epoch": 0.12771084337349398,
+      "grad_norm": 2.3670759836719353,
+      "learning_rate": 9.835803325903146e-06,
+      "loss": 1.0793,
+      "step": 4240
+    },
+    {
+      "epoch": 0.1280120481927711,
+      "grad_norm": 3.757187886266233,
+      "learning_rate": 9.834535097501271e-06,
+      "loss": 1.0342,
+      "step": 4250
+    },
+    {
+      "epoch": 0.12831325301204818,
+      "grad_norm": 2.2064521983256453,
+      "learning_rate": 9.833262072526502e-06,
+      "loss": 1.016,
+      "step": 4260
+    },
+    {
+      "epoch": 0.1286144578313253,
+      "grad_norm": 3.504463888375646,
+      "learning_rate": 9.831984252241864e-06,
+      "loss": 1.1306,
+      "step": 4270
+    },
+    {
+      "epoch": 0.12891566265060242,
+      "grad_norm": 3.4352599178275462,
+      "learning_rate": 9.83070163791515e-06,
+      "loss": 1.0684,
+      "step": 4280
+    },
+    {
+      "epoch": 0.1292168674698795,
+      "grad_norm": 2.4608514162669373,
+      "learning_rate": 9.829414230818897e-06,
+      "loss": 1.0528,
+      "step": 4290
+    },
+    {
+      "epoch": 0.12951807228915663,
+      "grad_norm": 3.697431973052916,
+      "learning_rate": 9.828122032230405e-06,
+      "loss": 1.0048,
+      "step": 4300
+    },
+    {
+      "epoch": 0.12981927710843374,
+      "grad_norm": 3.5673373327538562,
+      "learning_rate": 9.826825043431727e-06,
+      "loss": 1.0399,
+      "step": 4310
+    },
+    {
+      "epoch": 0.13012048192771083,
+      "grad_norm": 3.521636320056176,
+      "learning_rate": 9.825523265709667e-06,
+      "loss": 1.1146,
+      "step": 4320
+    },
+    {
+      "epoch": 0.13042168674698795,
+      "grad_norm": 2.212955928751166,
+      "learning_rate": 9.824216700355781e-06,
+      "loss": 1.0379,
+      "step": 4330
+    },
+    {
+      "epoch": 0.13072289156626507,
+      "grad_norm": 4.556411036688682,
+      "learning_rate": 9.822905348666373e-06,
+      "loss": 0.979,
+      "step": 4340
+    },
+    {
+      "epoch": 0.13102409638554216,
+      "grad_norm": 4.1305780843730515,
+      "learning_rate": 9.8215892119425e-06,
+      "loss": 1.1257,
+      "step": 4350
+    },
+    {
+      "epoch": 0.13132530120481928,
+      "grad_norm": 3.736857665673749,
+      "learning_rate": 9.820268291489962e-06,
+      "loss": 1.037,
+      "step": 4360
+    },
+    {
+      "epoch": 0.1316265060240964,
+      "grad_norm": 3.4496883762116344,
+      "learning_rate": 9.818942588619307e-06,
+      "loss": 1.1464,
+      "step": 4370
+    },
+    {
+      "epoch": 0.13192771084337349,
+      "grad_norm": 4.722706276570147,
+      "learning_rate": 9.81761210464583e-06,
+      "loss": 1.0105,
+      "step": 4380
+    },
+    {
+      "epoch": 0.1322289156626506,
+      "grad_norm": 3.674024436386717,
+      "learning_rate": 9.816276840889569e-06,
+      "loss": 0.9943,
+      "step": 4390
+    },
+    {
+      "epoch": 0.13253012048192772,
+      "grad_norm": 4.306251454283166,
+      "learning_rate": 9.8149367986753e-06,
+      "loss": 1.1416,
+      "step": 4400
+    },
+    {
+      "epoch": 0.1328313253012048,
+      "grad_norm": 2.318543131903423,
+      "learning_rate": 9.81359197933254e-06,
+      "loss": 1.0301,
+      "step": 4410
+    },
+    {
+      "epoch": 0.13313253012048193,
+      "grad_norm": 3.21024883306362,
+      "learning_rate": 9.812242384195553e-06,
+      "loss": 1.1388,
+      "step": 4420
+    },
+    {
+      "epoch": 0.13343373493975905,
+      "grad_norm": 4.890422036619223,
+      "learning_rate": 9.810888014603338e-06,
+      "loss": 1.1272,
+      "step": 4430
+    },
+    {
+      "epoch": 0.13373493975903614,
+      "grad_norm": 3.7271056320387834,
+      "learning_rate": 9.809528871899625e-06,
+      "loss": 0.9663,
+      "step": 4440
+    },
+    {
+      "epoch": 0.13403614457831325,
+      "grad_norm": 2.2886395005193227,
+      "learning_rate": 9.808164957432887e-06,
+      "loss": 1.0486,
+      "step": 4450
+    },
+    {
+      "epoch": 0.13433734939759037,
+      "grad_norm": 2.273780413684844,
+      "learning_rate": 9.806796272556329e-06,
+      "loss": 1.0658,
+      "step": 4460
+    },
+    {
+      "epoch": 0.13463855421686746,
+      "grad_norm": 3.4815418700381815,
+      "learning_rate": 9.805422818627887e-06,
+      "loss": 1.151,
+      "step": 4470
+    },
+    {
+      "epoch": 0.13493975903614458,
+      "grad_norm": 3.143963049514939,
+      "learning_rate": 9.80404459701023e-06,
+      "loss": 1.1618,
+      "step": 4480
+    },
+    {
+      "epoch": 0.1352409638554217,
+      "grad_norm": 3.8333974143775444,
+      "learning_rate": 9.80266160907076e-06,
+      "loss": 1.1307,
+      "step": 4490
+    },
+    {
+      "epoch": 0.1355421686746988,
+      "grad_norm": 3.796359402263128,
+      "learning_rate": 9.801273856181603e-06,
+      "loss": 1.0186,
+      "step": 4500
+    },
+    {
+      "epoch": 0.1358433734939759,
+      "grad_norm": 3.526380098069618,
+      "learning_rate": 9.799881339719615e-06,
+      "loss": 1.1202,
+      "step": 4510
+    },
+    {
+      "epoch": 0.13614457831325302,
+      "grad_norm": 2.3799394298701637,
+      "learning_rate": 9.798484061066379e-06,
+      "loss": 1.1168,
+      "step": 4520
+    },
+    {
+      "epoch": 0.1364457831325301,
+      "grad_norm": 2.3420014684851416,
+      "learning_rate": 9.797082021608199e-06,
+      "loss": 1.0472,
+      "step": 4530
+    },
+    {
+      "epoch": 0.13674698795180723,
+      "grad_norm": 3.8824369336877993,
+      "learning_rate": 9.795675222736109e-06,
+      "loss": 1.1207,
+      "step": 4540
+    },
+    {
+      "epoch": 0.13704819277108435,
+      "grad_norm": 3.7880901774233027,
+      "learning_rate": 9.794263665845862e-06,
+      "loss": 1.1036,
+      "step": 4550
+    },
+    {
+      "epoch": 0.13734939759036144,
+      "grad_norm": 2.2568634543924944,
+      "learning_rate": 9.792847352337926e-06,
+      "loss": 1.0919,
+      "step": 4560
+    },
+    {
+      "epoch": 0.13765060240963856,
+      "grad_norm": 3.1778188343088205,
+      "learning_rate": 9.791426283617497e-06,
+      "loss": 1.1278,
+      "step": 4570
+    },
+    {
+      "epoch": 0.13795180722891567,
+      "grad_norm": 3.224250552306772,
+      "learning_rate": 9.790000461094483e-06,
+      "loss": 0.982,
+      "step": 4580
+    },
+    {
+      "epoch": 0.13825301204819276,
+      "grad_norm": 3.1585896158608344,
+      "learning_rate": 9.788569886183511e-06,
+      "loss": 1.1179,
+      "step": 4590
+    },
+    {
+      "epoch": 0.13855421686746988,
+      "grad_norm": 3.462414772444446,
+      "learning_rate": 9.787134560303925e-06,
+      "loss": 1.0827,
+      "step": 4600
+    },
+    {
+      "epoch": 0.138855421686747,
+      "grad_norm": 2.0866169715331084,
+      "learning_rate": 9.785694484879778e-06,
+      "loss": 0.988,
+      "step": 4610
+    },
+    {
+      "epoch": 0.1391566265060241,
+      "grad_norm": 3.389979691122731,
+      "learning_rate": 9.784249661339838e-06,
+      "loss": 1.1036,
+      "step": 4620
+    },
+    {
+      "epoch": 0.1394578313253012,
+      "grad_norm": 5.210842898777085,
+      "learning_rate": 9.782800091117581e-06,
+      "loss": 1.0532,
+      "step": 4630
+    },
+    {
+      "epoch": 0.13975903614457832,
+      "grad_norm": 2.2035718957581456,
+      "learning_rate": 9.7813457756512e-06,
+      "loss": 0.9751,
+      "step": 4640
+    },
+    {
+      "epoch": 0.14006024096385541,
+      "grad_norm": 5.163817863609091,
+      "learning_rate": 9.779886716383585e-06,
+      "loss": 1.0447,
+      "step": 4650
+    },
+    {
+      "epoch": 0.14036144578313253,
+      "grad_norm": 2.1874523291711427,
+      "learning_rate": 9.778422914762342e-06,
+      "loss": 1.0163,
+      "step": 4660
+    },
+    {
+      "epoch": 0.14066265060240965,
+      "grad_norm": 3.3593565331746476,
+      "learning_rate": 9.776954372239778e-06,
+      "loss": 1.0752,
+      "step": 4670
+    },
+    {
+      "epoch": 0.14096385542168674,
+      "grad_norm": 6.561474234949798,
+      "learning_rate": 9.775481090272903e-06,
+      "loss": 1.0352,
+      "step": 4680
+    },
+    {
+      "epoch": 0.14126506024096386,
+      "grad_norm": 2.2764790120471474,
+      "learning_rate": 9.774003070323433e-06,
+      "loss": 1.0816,
+      "step": 4690
+    },
+    {
+      "epoch": 0.14156626506024098,
+      "grad_norm": 4.284699327442535,
+      "learning_rate": 9.772520313857777e-06,
+      "loss": 1.0362,
+      "step": 4700
+    },
+    {
+      "epoch": 0.14186746987951807,
+      "grad_norm": 5.622862962661047,
+      "learning_rate": 9.771032822347052e-06,
+      "loss": 1.1013,
+      "step": 4710
+    },
+    {
+      "epoch": 0.14216867469879518,
+      "grad_norm": 3.5193476260860708,
+      "learning_rate": 9.76954059726707e-06,
+      "loss": 1.0359,
+      "step": 4720
+    },
+    {
+      "epoch": 0.1424698795180723,
+      "grad_norm": 3.1635234057670396,
+      "learning_rate": 9.768043640098337e-06,
+      "loss": 1.0838,
+      "step": 4730
+    },
+    {
+      "epoch": 0.1427710843373494,
+      "grad_norm": 5.3147053949696845,
+      "learning_rate": 9.766541952326055e-06,
+      "loss": 1.0217,
+      "step": 4740
+    },
+    {
+      "epoch": 0.1430722891566265,
+      "grad_norm": 3.9502160134496602,
+      "learning_rate": 9.765035535440122e-06,
+      "loss": 1.0724,
+      "step": 4750
+    },
+    {
+      "epoch": 0.14337349397590363,
+      "grad_norm": 3.2204727664784274,
+      "learning_rate": 9.763524390935124e-06,
+      "loss": 1.0666,
+      "step": 4760
+    },
+    {
+      "epoch": 0.14367469879518072,
+      "grad_norm": 3.1741778820868247,
+      "learning_rate": 9.762008520310341e-06,
+      "loss": 0.963,
+      "step": 4770
+    },
+    {
+      "epoch": 0.14397590361445783,
+      "grad_norm": 4.733626504373592,
+      "learning_rate": 9.760487925069742e-06,
+      "loss": 0.9357,
+      "step": 4780
+    },
+    {
+      "epoch": 0.14427710843373495,
+      "grad_norm": 4.318346784048267,
+      "learning_rate": 9.75896260672198e-06,
+      "loss": 1.0624,
+      "step": 4790
+    },
+    {
+      "epoch": 0.14457831325301204,
+      "grad_norm": 2.327064879984706,
+      "learning_rate": 9.757432566780394e-06,
+      "loss": 1.0039,
+      "step": 4800
+    },
+    {
+      "epoch": 0.14487951807228916,
+      "grad_norm": 3.655317981523251,
+      "learning_rate": 9.755897806763015e-06,
+      "loss": 1.1628,
+      "step": 4810
+    },
+    {
+      "epoch": 0.14518072289156628,
+      "grad_norm": 3.1522492880450566,
+      "learning_rate": 9.754358328192546e-06,
+      "loss": 1.0239,
+      "step": 4820
+    },
+    {
+      "epoch": 0.14548192771084337,
+      "grad_norm": 3.351343494801427,
+      "learning_rate": 9.752814132596382e-06,
+      "loss": 1.1085,
+      "step": 4830
+    },
+    {
+      "epoch": 0.14578313253012049,
+      "grad_norm": 3.278068913765995,
+      "learning_rate": 9.75126522150659e-06,
+      "loss": 1.0416,
+      "step": 4840
+    },
+    {
+      "epoch": 0.1460843373493976,
+      "grad_norm": 3.2726257917049435,
+      "learning_rate": 9.74971159645992e-06,
+      "loss": 1.0441,
+      "step": 4850
+    },
+    {
+      "epoch": 0.1463855421686747,
+      "grad_norm": 2.3833415910525386,
+      "learning_rate": 9.748153258997795e-06,
+      "loss": 1.0152,
+      "step": 4860
+    },
+    {
+      "epoch": 0.1466867469879518,
+      "grad_norm": 3.231056424379859,
+      "learning_rate": 9.746590210666318e-06,
+      "loss": 1.1212,
+      "step": 4870
+    },
+    {
+      "epoch": 0.14698795180722893,
+      "grad_norm": 4.140471833383135,
+      "learning_rate": 9.745022453016265e-06,
+      "loss": 0.9856,
+      "step": 4880
+    },
+    {
+      "epoch": 0.14728915662650602,
+      "grad_norm": 2.0937714976689143,
+      "learning_rate": 9.743449987603082e-06,
+      "loss": 1.0324,
+      "step": 4890
+    },
+    {
+      "epoch": 0.14759036144578314,
+      "grad_norm": 4.2575104490906845,
+      "learning_rate": 9.741872815986888e-06,
+      "loss": 1.0921,
+      "step": 4900
+    },
+    {
+      "epoch": 0.14789156626506025,
+      "grad_norm": 2.9103859190839247,
+      "learning_rate": 9.740290939732467e-06,
+      "loss": 0.9322,
+      "step": 4910
+    },
+    {
+      "epoch": 0.14819277108433734,
+      "grad_norm": 2.721531202849386,
+      "learning_rate": 9.738704360409276e-06,
+      "loss": 1.1442,
+      "step": 4920
+    },
+    {
+      "epoch": 0.14849397590361446,
+      "grad_norm": 2.050446101251248,
+      "learning_rate": 9.737113079591438e-06,
+      "loss": 1.0194,
+      "step": 4930
+    },
+    {
+      "epoch": 0.14879518072289158,
+      "grad_norm": 3.3305624788137864,
+      "learning_rate": 9.735517098857735e-06,
+      "loss": 0.9831,
+      "step": 4940
+    },
+    {
+      "epoch": 0.14909638554216867,
+      "grad_norm": 2.420983075223732,
+      "learning_rate": 9.733916419791618e-06,
+      "loss": 1.0134,
+      "step": 4950
+    },
+    {
+      "epoch": 0.1493975903614458,
+      "grad_norm": 2.8134241849328574,
+      "learning_rate": 9.732311043981195e-06,
+      "loss": 1.0931,
+      "step": 4960
+    },
+    {
+      "epoch": 0.1496987951807229,
+      "grad_norm": 3.1472681936635483,
+      "learning_rate": 9.730700973019239e-06,
+      "loss": 1.0581,
+      "step": 4970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0810334766317093,
+      "learning_rate": 9.729086208503174e-06,
+      "loss": 1.0027,
+      "step": 4980
+    },
+    {
+      "epoch": 0.1503012048192771,
+      "grad_norm": 2.9206618313633395,
+      "learning_rate": 9.727466752035088e-06,
+      "loss": 0.9923,
+      "step": 4990
+    },
+    {
+      "epoch": 0.15060240963855423,
+      "grad_norm": 3.2699697753281587,
+      "learning_rate": 9.725842605221721e-06,
+      "loss": 1.1164,
+      "step": 5000
+    },
+    {
+      "epoch": 0.15090361445783132,
+      "grad_norm": 6.773061296964998,
+      "learning_rate": 9.724213769674466e-06,
+      "loss": 1.0413,
+      "step": 5010
+    },
+    {
+      "epoch": 0.15120481927710844,
+      "grad_norm": 3.348064035041207,
+      "learning_rate": 9.722580247009367e-06,
+      "loss": 1.0733,
+      "step": 5020
+    },
+    {
+      "epoch": 0.15150602409638556,
+      "grad_norm": 4.123785046042495,
+      "learning_rate": 9.720942038847123e-06,
+      "loss": 1.0695,
+      "step": 5030
+    },
+    {
+      "epoch": 0.15180722891566265,
+      "grad_norm": 6.6390263479865075,
+      "learning_rate": 9.719299146813074e-06,
+      "loss": 1.0102,
+      "step": 5040
+    },
+    {
+      "epoch": 0.15210843373493976,
+      "grad_norm": 3.183406080712184,
+      "learning_rate": 9.717651572537217e-06,
+      "loss": 1.0317,
+      "step": 5050
+    },
+    {
+      "epoch": 0.15240963855421688,
+      "grad_norm": 2.209915307656265,
+      "learning_rate": 9.715999317654186e-06,
+      "loss": 0.9702,
+      "step": 5060
+    },
+    {
+      "epoch": 0.15271084337349397,
+      "grad_norm": 3.71668545598457,
+      "learning_rate": 9.714342383803259e-06,
+      "loss": 1.1026,
+      "step": 5070
+    },
+    {
+      "epoch": 0.1530120481927711,
+      "grad_norm": 3.3454059076645373,
+      "learning_rate": 9.712680772628365e-06,
+      "loss": 1.024,
+      "step": 5080
+    },
+    {
+      "epoch": 0.1533132530120482,
+      "grad_norm": 3.7676266694645033,
+      "learning_rate": 9.711014485778063e-06,
+      "loss": 1.0529,
+      "step": 5090
+    },
+    {
+      "epoch": 0.1536144578313253,
+      "grad_norm": 2.3911391873701313,
+      "learning_rate": 9.709343524905558e-06,
+      "loss": 1.0187,
+      "step": 5100
+    },
+    {
+      "epoch": 0.15391566265060241,
+      "grad_norm": 4.183322203598176,
+      "learning_rate": 9.70766789166869e-06,
+      "loss": 0.9653,
+      "step": 5110
+    },
+    {
+      "epoch": 0.15421686746987953,
+      "grad_norm": 3.4195694386180646,
+      "learning_rate": 9.705987587729932e-06,
+      "loss": 1.0815,
+      "step": 5120
+    },
+    {
+      "epoch": 0.15451807228915662,
+      "grad_norm": 4.601648726229802,
+      "learning_rate": 9.7043026147564e-06,
+      "loss": 1.07,
+      "step": 5130
+    },
+    {
+      "epoch": 0.15481927710843374,
+      "grad_norm": 2.1251037089708418,
+      "learning_rate": 9.702612974419828e-06,
+      "loss": 1.0668,
+      "step": 5140
+    },
+    {
+      "epoch": 0.15512048192771086,
+      "grad_norm": 2.1782790790900224,
+      "learning_rate": 9.700918668396595e-06,
+      "loss": 1.0139,
+      "step": 5150
+    },
+    {
+      "epoch": 0.15542168674698795,
+      "grad_norm": 3.251835279628236,
+      "learning_rate": 9.699219698367698e-06,
+      "loss": 1.1551,
+      "step": 5160
+    },
+    {
+      "epoch": 0.15572289156626506,
+      "grad_norm": 3.53702536080039,
+      "learning_rate": 9.697516066018771e-06,
+      "loss": 0.9878,
+      "step": 5170
+    },
+    {
+      "epoch": 0.15602409638554218,
+      "grad_norm": 3.2199102356574096,
+      "learning_rate": 9.695807773040066e-06,
+      "loss": 1.0814,
+      "step": 5180
+    },
+    {
+      "epoch": 0.15632530120481927,
+      "grad_norm": 2.279979865855527,
+      "learning_rate": 9.694094821126463e-06,
+      "loss": 0.8722,
+      "step": 5190
+    },
+    {
+      "epoch": 0.1566265060240964,
+      "grad_norm": 3.414618660368347,
+      "learning_rate": 9.692377211977463e-06,
+      "loss": 1.1059,
+      "step": 5200
+    },
+    {
+      "epoch": 0.15692771084337348,
+      "grad_norm": 2.1338305587010895,
+      "learning_rate": 9.690654947297187e-06,
+      "loss": 0.9732,
+      "step": 5210
+    },
+    {
+      "epoch": 0.1572289156626506,
+      "grad_norm": 3.3926239553083364,
+      "learning_rate": 9.688928028794378e-06,
+      "loss": 1.0287,
+      "step": 5220
+    },
+    {
+      "epoch": 0.15753012048192772,
+      "grad_norm": 3.2978616424156866,
+      "learning_rate": 9.687196458182393e-06,
+      "loss": 1.0199,
+      "step": 5230
+    },
+    {
+      "epoch": 0.1578313253012048,
+      "grad_norm": 4.073536092982219,
+      "learning_rate": 9.685460237179205e-06,
+      "loss": 1.0905,
+      "step": 5240
+    },
+    {
+      "epoch": 0.15813253012048192,
+      "grad_norm": 3.332084115789668,
+      "learning_rate": 9.683719367507403e-06,
+      "loss": 1.0608,
+      "step": 5250
+    },
+    {
+      "epoch": 0.15843373493975904,
+      "grad_norm": 3.9475889479040536,
+      "learning_rate": 9.681973850894187e-06,
+      "loss": 1.0316,
+      "step": 5260
+    },
+    {
+      "epoch": 0.15873493975903613,
+      "grad_norm": 4.379157434561739,
+      "learning_rate": 9.680223689071364e-06,
+      "loss": 1.0919,
+      "step": 5270
+    },
+    {
+      "epoch": 0.15903614457831325,
+      "grad_norm": 3.0403670956453666,
+      "learning_rate": 9.678468883775357e-06,
+      "loss": 0.9766,
+      "step": 5280
+    },
+    {
+      "epoch": 0.15933734939759037,
+      "grad_norm": 3.8321049042223034,
+      "learning_rate": 9.676709436747192e-06,
+      "loss": 1.0237,
+      "step": 5290
+    },
+    {
+      "epoch": 0.15963855421686746,
+      "grad_norm": 3.410946507894495,
+      "learning_rate": 9.674945349732498e-06,
+      "loss": 1.0378,
+      "step": 5300
+    },
+    {
+      "epoch": 0.15993975903614457,
+      "grad_norm": 3.1167832390652057,
+      "learning_rate": 9.673176624481514e-06,
+      "loss": 1.073,
+      "step": 5310
+    },
+    {
+      "epoch": 0.1602409638554217,
+      "grad_norm": 2.627554451968262,
+      "learning_rate": 9.67140326274907e-06,
+      "loss": 0.8978,
+      "step": 5320
+    },
+    {
+      "epoch": 0.16054216867469878,
+      "grad_norm": 2.91273188611948,
+      "learning_rate": 9.669625266294607e-06,
+      "loss": 1.0324,
+      "step": 5330
+    },
+    {
+      "epoch": 0.1608433734939759,
+      "grad_norm": 3.6348902919380652,
+      "learning_rate": 9.667842636882161e-06,
+      "loss": 1.014,
+      "step": 5340
+    },
+    {
+      "epoch": 0.16114457831325302,
+      "grad_norm": 3.1362788924676175,
+      "learning_rate": 9.66605537628036e-06,
+      "loss": 1.1956,
+      "step": 5350
+    },
+    {
+      "epoch": 0.1614457831325301,
+      "grad_norm": 2.325065922275079,
+      "learning_rate": 9.664263486262435e-06,
+      "loss": 1.139,
+      "step": 5360
+    },
+    {
+      "epoch": 0.16174698795180723,
+      "grad_norm": 2.244733354608646,
+      "learning_rate": 9.662466968606204e-06,
+      "loss": 0.9976,
+      "step": 5370
+    },
+    {
+      "epoch": 0.16204819277108434,
+      "grad_norm": 2.2571621339890933,
+      "learning_rate": 9.660665825094075e-06,
+      "loss": 1.0131,
+      "step": 5380
+    },
+    {
+      "epoch": 0.16234939759036143,
+      "grad_norm": 3.162499695346246,
+      "learning_rate": 9.658860057513051e-06,
+      "loss": 1.0787,
+      "step": 5390
+    },
+    {
+      "epoch": 0.16265060240963855,
+      "grad_norm": 2.4532344817953757,
+      "learning_rate": 9.65704966765472e-06,
+      "loss": 1.0239,
+      "step": 5400
+    },
+    {
+      "epoch": 0.16295180722891567,
+      "grad_norm": 4.524706172091528,
+      "learning_rate": 9.655234657315255e-06,
+      "loss": 1.0448,
+      "step": 5410
+    },
+    {
+      "epoch": 0.16325301204819276,
+      "grad_norm": 3.6682300562837824,
+      "learning_rate": 9.653415028295416e-06,
+      "loss": 1.0477,
+      "step": 5420
+    },
+    {
+      "epoch": 0.16355421686746988,
+      "grad_norm": 3.296338353127837,
+      "learning_rate": 9.651590782400544e-06,
+      "loss": 1.0617,
+      "step": 5430
+    },
+    {
+      "epoch": 0.163855421686747,
+      "grad_norm": 4.166905656035678,
+      "learning_rate": 9.649761921440559e-06,
+      "loss": 1.1264,
+      "step": 5440
+    },
+    {
+      "epoch": 0.16415662650602408,
+      "grad_norm": 3.0669389796408577,
+      "learning_rate": 9.64792844722996e-06,
+      "loss": 1.1601,
+      "step": 5450
+    },
+    {
+      "epoch": 0.1644578313253012,
+      "grad_norm": 3.158083945931002,
+      "learning_rate": 9.646090361587828e-06,
+      "loss": 1.0575,
+      "step": 5460
+    },
+    {
+      "epoch": 0.16475903614457832,
+      "grad_norm": 3.9595012649954433,
+      "learning_rate": 9.644247666337811e-06,
+      "loss": 0.9377,
+      "step": 5470
+    },
+    {
+      "epoch": 0.1650602409638554,
+      "grad_norm": 7.823524146210229,
+      "learning_rate": 9.642400363308138e-06,
+      "loss": 1.1025,
+      "step": 5480
+    },
+    {
+      "epoch": 0.16536144578313253,
+      "grad_norm": 3.293888596495116,
+      "learning_rate": 9.640548454331605e-06,
+      "loss": 0.9792,
+      "step": 5490
+    },
+    {
+      "epoch": 0.16566265060240964,
+      "grad_norm": 2.2819294235665177,
+      "learning_rate": 9.638691941245581e-06,
+      "loss": 0.9978,
+      "step": 5500
+    },
+    {
+      "epoch": 0.16596385542168673,
+      "grad_norm": 3.3342689743151426,
+      "learning_rate": 9.636830825891999e-06,
+      "loss": 1.0293,
+      "step": 5510
+    },
+    {
+      "epoch": 0.16626506024096385,
+      "grad_norm": 3.3822531399355977,
+      "learning_rate": 9.63496511011736e-06,
+      "loss": 0.9436,
+      "step": 5520
+    },
+    {
+      "epoch": 0.16656626506024097,
+      "grad_norm": 2.255208646786481,
+      "learning_rate": 9.633094795772732e-06,
+      "loss": 1.0041,
+      "step": 5530
+    },
+    {
+      "epoch": 0.16686746987951806,
+      "grad_norm": 4.442125013498681,
+      "learning_rate": 9.631219884713744e-06,
+      "loss": 1.051,
+      "step": 5540
+    },
+    {
+      "epoch": 0.16716867469879518,
+      "grad_norm": 3.956265628275085,
+      "learning_rate": 9.62934037880058e-06,
+      "loss": 1.0325,
+      "step": 5550
+    },
+    {
+      "epoch": 0.1674698795180723,
+      "grad_norm": 4.556155077351476,
+      "learning_rate": 9.627456279897992e-06,
+      "loss": 1.09,
+      "step": 5560
+    },
+    {
+      "epoch": 0.16777108433734939,
+      "grad_norm": 3.899010336394591,
+      "learning_rate": 9.625567589875282e-06,
+      "loss": 1.0844,
+      "step": 5570
+    },
+    {
+      "epoch": 0.1680722891566265,
+      "grad_norm": 4.1025563838568075,
+      "learning_rate": 9.62367431060631e-06,
+      "loss": 1.0263,
+      "step": 5580
+    },
+    {
+      "epoch": 0.16837349397590362,
+      "grad_norm": 3.261371165963412,
+      "learning_rate": 9.621776443969488e-06,
+      "loss": 0.9759,
+      "step": 5590
+    },
+    {
+      "epoch": 0.1686746987951807,
+      "grad_norm": 3.8351126648010445,
+      "learning_rate": 9.619873991847782e-06,
+      "loss": 1.0331,
+      "step": 5600
+    },
+    {
+      "epoch": 0.16897590361445783,
+      "grad_norm": 10.385110842644648,
+      "learning_rate": 9.617966956128704e-06,
+      "loss": 1.026,
+      "step": 5610
+    },
+    {
+      "epoch": 0.16927710843373495,
+      "grad_norm": 3.9904039368640944,
+      "learning_rate": 9.616055338704313e-06,
+      "loss": 0.87,
+      "step": 5620
+    },
+    {
+      "epoch": 0.16957831325301204,
+      "grad_norm": 4.251155600976589,
+      "learning_rate": 9.614139141471221e-06,
+      "loss": 0.8925,
+      "step": 5630
+    },
+    {
+      "epoch": 0.16987951807228915,
+      "grad_norm": 5.092732483832327,
+      "learning_rate": 9.612218366330574e-06,
+      "loss": 1.0587,
+      "step": 5640
+    },
+    {
+      "epoch": 0.17018072289156627,
+      "grad_norm": 5.56784543715876,
+      "learning_rate": 9.610293015188067e-06,
+      "loss": 1.1863,
+      "step": 5650
+    },
+    {
+      "epoch": 0.17048192771084336,
+      "grad_norm": 2.085502556311897,
+      "learning_rate": 9.608363089953933e-06,
+      "loss": 0.9125,
+      "step": 5660
+    },
+    {
+      "epoch": 0.17078313253012048,
+      "grad_norm": 3.805266632887387,
+      "learning_rate": 9.606428592542942e-06,
+      "loss": 1.1807,
+      "step": 5670
+    },
+    {
+      "epoch": 0.1710843373493976,
+      "grad_norm": 2.176487423411828,
+      "learning_rate": 9.604489524874404e-06,
+      "loss": 1.0128,
+      "step": 5680
+    },
+    {
+      "epoch": 0.1713855421686747,
+      "grad_norm": 3.511119177564224,
+      "learning_rate": 9.602545888872153e-06,
+      "loss": 0.8843,
+      "step": 5690
+    },
+    {
+      "epoch": 0.1716867469879518,
+      "grad_norm": 4.65210599013388,
+      "learning_rate": 9.60059768646457e-06,
+      "loss": 1.0541,
+      "step": 5700
+    },
+    {
+      "epoch": 0.17198795180722892,
+      "grad_norm": 5.177770181975888,
+      "learning_rate": 9.598644919584557e-06,
+      "loss": 1.1533,
+      "step": 5710
+    },
+    {
+      "epoch": 0.172289156626506,
+      "grad_norm": 2.7868109404114447,
+      "learning_rate": 9.596687590169547e-06,
+      "loss": 0.9214,
+      "step": 5720
+    },
+    {
+      "epoch": 0.17259036144578313,
+      "grad_norm": 3.976355975955576,
+      "learning_rate": 9.594725700161503e-06,
+      "loss": 1.1111,
+      "step": 5730
+    },
+    {
+      "epoch": 0.17289156626506025,
+      "grad_norm": 3.8197839655454024,
+      "learning_rate": 9.592759251506905e-06,
+      "loss": 1.075,
+      "step": 5740
+    },
+    {
+      "epoch": 0.17319277108433734,
+      "grad_norm": 4.164754554800956,
+      "learning_rate": 9.590788246156762e-06,
+      "loss": 1.0444,
+      "step": 5750
+    },
+    {
+      "epoch": 0.17349397590361446,
+      "grad_norm": 2.055612662165044,
+      "learning_rate": 9.588812686066604e-06,
+      "loss": 1.0374,
+      "step": 5760
+    },
+    {
+      "epoch": 0.17379518072289157,
+      "grad_norm": 2.2807599206085913,
+      "learning_rate": 9.586832573196477e-06,
+      "loss": 1.0652,
+      "step": 5770
+    },
+    {
+      "epoch": 0.17409638554216866,
+      "grad_norm": 3.578007650323301,
+      "learning_rate": 9.584847909510947e-06,
+      "loss": 0.9844,
+      "step": 5780
+    },
+    {
+      "epoch": 0.17439759036144578,
+      "grad_norm": 2.2768377854701187,
+      "learning_rate": 9.582858696979092e-06,
+      "loss": 0.998,
+      "step": 5790
+    },
+    {
+      "epoch": 0.1746987951807229,
+      "grad_norm": 17.390522458209965,
+      "learning_rate": 9.580864937574503e-06,
+      "loss": 0.9881,
+      "step": 5800
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 2.999305191298459,
+      "learning_rate": 9.578866633275289e-06,
+      "loss": 1.0092,
+      "step": 5810
+    },
+    {
+      "epoch": 0.1753012048192771,
+      "grad_norm": 3.827981839453712,
+      "learning_rate": 9.576863786064056e-06,
+      "loss": 1.0557,
+      "step": 5820
+    },
+    {
+      "epoch": 0.17560240963855422,
+      "grad_norm": 3.397036893241917,
+      "learning_rate": 9.574856397927932e-06,
+      "loss": 1.0643,
+      "step": 5830
+    },
+    {
+      "epoch": 0.17590361445783131,
+      "grad_norm": 2.2348195275879577,
+      "learning_rate": 9.572844470858537e-06,
+      "loss": 1.0859,
+      "step": 5840
+    },
+    {
+      "epoch": 0.17620481927710843,
+      "grad_norm": 2.948518431736692,
+      "learning_rate": 9.570828006852002e-06,
+      "loss": 1.1199,
+      "step": 5850
+    },
+    {
+      "epoch": 0.17650602409638555,
+      "grad_norm": 2.07783596473763,
+      "learning_rate": 9.568807007908956e-06,
+      "loss": 0.9549,
+      "step": 5860
+    },
+    {
+      "epoch": 0.17680722891566264,
+      "grad_norm": 3.563921820764686,
+      "learning_rate": 9.56678147603453e-06,
+      "loss": 0.9696,
+      "step": 5870
+    },
+    {
+      "epoch": 0.17710843373493976,
+      "grad_norm": 5.512277139507565,
+      "learning_rate": 9.564751413238348e-06,
+      "loss": 1.1727,
+      "step": 5880
+    },
+    {
+      "epoch": 0.17740963855421688,
+      "grad_norm": 3.4600431387569155,
+      "learning_rate": 9.562716821534536e-06,
+      "loss": 1.0864,
+      "step": 5890
+    },
+    {
+      "epoch": 0.17771084337349397,
+      "grad_norm": 4.093428608264867,
+      "learning_rate": 9.560677702941708e-06,
+      "loss": 1.0945,
+      "step": 5900
+    },
+    {
+      "epoch": 0.17801204819277108,
+      "grad_norm": 2.09148292103574,
+      "learning_rate": 9.558634059482967e-06,
+      "loss": 0.8993,
+      "step": 5910
+    },
+    {
+      "epoch": 0.1783132530120482,
+      "grad_norm": 3.281318600861496,
+      "learning_rate": 9.556585893185916e-06,
+      "loss": 0.9695,
+      "step": 5920
+    },
+    {
+      "epoch": 0.1786144578313253,
+      "grad_norm": 3.9318585105629578,
+      "learning_rate": 9.554533206082633e-06,
+      "loss": 1.1362,
+      "step": 5930
+    },
+    {
+      "epoch": 0.1789156626506024,
+      "grad_norm": 3.518532574219804,
+      "learning_rate": 9.55247600020969e-06,
+      "loss": 1.0634,
+      "step": 5940
+    },
+    {
+      "epoch": 0.17921686746987953,
+      "grad_norm": 4.155557513315787,
+      "learning_rate": 9.550414277608138e-06,
+      "loss": 0.9822,
+      "step": 5950
+    },
+    {
+      "epoch": 0.17951807228915662,
+      "grad_norm": 3.7175217066617,
+      "learning_rate": 9.54834804032351e-06,
+      "loss": 1.0538,
+      "step": 5960
+    },
+    {
+      "epoch": 0.17981927710843373,
+      "grad_norm": 3.498416998598756,
+      "learning_rate": 9.546277290405816e-06,
+      "loss": 1.0399,
+      "step": 5970
+    },
+    {
+      "epoch": 0.18012048192771085,
+      "grad_norm": 3.458768120804074,
+      "learning_rate": 9.544202029909553e-06,
+      "loss": 1.0953,
+      "step": 5980
+    },
+    {
+      "epoch": 0.18042168674698794,
+      "grad_norm": 3.1602574880014482,
+      "learning_rate": 9.542122260893678e-06,
+      "loss": 0.946,
+      "step": 5990
+    },
+    {
+      "epoch": 0.18072289156626506,
+      "grad_norm": 4.6804582295334205,
+      "learning_rate": 9.540037985421635e-06,
+      "loss": 0.9814,
+      "step": 6000
+    },
+    {
+      "epoch": 0.18102409638554218,
+      "grad_norm": 3.6710468665569405,
+      "learning_rate": 9.537949205561328e-06,
+      "loss": 1.1269,
+      "step": 6010
+    },
+    {
+      "epoch": 0.18132530120481927,
+      "grad_norm": 4.293441762407758,
+      "learning_rate": 9.53585592338514e-06,
+      "loss": 1.0217,
+      "step": 6020
+    },
+    {
+      "epoch": 0.18162650602409638,
+      "grad_norm": 3.761636872276159,
+      "learning_rate": 9.533758140969913e-06,
+      "loss": 1.0593,
+      "step": 6030
+    },
+    {
+      "epoch": 0.1819277108433735,
+      "grad_norm": 2.0282710989656683,
+      "learning_rate": 9.53165586039696e-06,
+      "loss": 1.0833,
+      "step": 6040
+    },
+    {
+      "epoch": 0.1822289156626506,
+      "grad_norm": 3.6331857587787644,
+      "learning_rate": 9.52954908375205e-06,
+      "loss": 1.1333,
+      "step": 6050
+    },
+    {
+      "epoch": 0.1825301204819277,
+      "grad_norm": 2.1858602036458006,
+      "learning_rate": 9.527437813125418e-06,
+      "loss": 1.0902,
+      "step": 6060
+    },
+    {
+      "epoch": 0.18283132530120483,
+      "grad_norm": 4.342594825620618,
+      "learning_rate": 9.525322050611757e-06,
+      "loss": 1.0631,
+      "step": 6070
+    },
+    {
+      "epoch": 0.18313253012048192,
+      "grad_norm": 4.44971842951979,
+      "learning_rate": 9.523201798310216e-06,
+      "loss": 1.0879,
+      "step": 6080
+    },
+    {
+      "epoch": 0.18343373493975904,
+      "grad_norm": 3.7729594924080487,
+      "learning_rate": 9.521077058324396e-06,
+      "loss": 1.0204,
+      "step": 6090
+    },
+    {
+      "epoch": 0.18373493975903615,
+      "grad_norm": 3.510001474937845,
+      "learning_rate": 9.518947832762358e-06,
+      "loss": 0.9884,
+      "step": 6100
+    },
+    {
+      "epoch": 0.18403614457831324,
+      "grad_norm": 5.146451089418775,
+      "learning_rate": 9.516814123736604e-06,
+      "loss": 1.063,
+      "step": 6110
+    },
+    {
+      "epoch": 0.18433734939759036,
+      "grad_norm": 4.001807019595258,
+      "learning_rate": 9.51467593336409e-06,
+      "loss": 1.004,
+      "step": 6120
+    },
+    {
+      "epoch": 0.18463855421686748,
+      "grad_norm": 2.0613442946765694,
+      "learning_rate": 9.512533263766217e-06,
+      "loss": 0.9519,
+      "step": 6130
+    },
+    {
+      "epoch": 0.18493975903614457,
+      "grad_norm": 3.9775071093580885,
+      "learning_rate": 9.510386117068828e-06,
+      "loss": 0.9807,
+      "step": 6140
+    },
+    {
+      "epoch": 0.1852409638554217,
+      "grad_norm": 2.1522460573260003,
+      "learning_rate": 9.508234495402211e-06,
+      "loss": 0.9739,
+      "step": 6150
+    },
+    {
+      "epoch": 0.1855421686746988,
+      "grad_norm": 5.011296767337308,
+      "learning_rate": 9.506078400901092e-06,
+      "loss": 1.0179,
+      "step": 6160
+    },
+    {
+      "epoch": 0.1858433734939759,
+      "grad_norm": 16.124866692782845,
+      "learning_rate": 9.503917835704638e-06,
+      "loss": 1.0036,
+      "step": 6170
+    },
+    {
+      "epoch": 0.186144578313253,
+      "grad_norm": 2.0356647653715245,
+      "learning_rate": 9.501752801956442e-06,
+      "loss": 0.9634,
+      "step": 6180
+    },
+    {
+      "epoch": 0.18644578313253013,
+      "grad_norm": 7.487542173226552,
+      "learning_rate": 9.499583301804544e-06,
+      "loss": 0.9596,
+      "step": 6190
+    },
+    {
+      "epoch": 0.18674698795180722,
+      "grad_norm": 2.021399937359315,
+      "learning_rate": 9.497409337401405e-06,
+      "loss": 1.027,
+      "step": 6200
+    },
+    {
+      "epoch": 0.18704819277108434,
+      "grad_norm": 7.259619648806638,
+      "learning_rate": 9.495230910903921e-06,
+      "loss": 1.0857,
+      "step": 6210
+    },
+    {
+      "epoch": 0.18734939759036146,
+      "grad_norm": 6.860284296746412,
+      "learning_rate": 9.493048024473413e-06,
+      "loss": 1.0305,
+      "step": 6220
+    },
+    {
+      "epoch": 0.18765060240963854,
+      "grad_norm": 4.694009851591792,
+      "learning_rate": 9.490860680275623e-06,
+      "loss": 1.0267,
+      "step": 6230
+    },
+    {
+      "epoch": 0.18795180722891566,
+      "grad_norm": 6.55172407409244,
+      "learning_rate": 9.488668880480722e-06,
+      "loss": 0.9902,
+      "step": 6240
+    },
+    {
+      "epoch": 0.18825301204819278,
+      "grad_norm": 20.98587198042032,
+      "learning_rate": 9.486472627263302e-06,
+      "loss": 0.9025,
+      "step": 6250
+    },
+    {
+      "epoch": 0.18855421686746987,
+      "grad_norm": 2.122218322085193,
+      "learning_rate": 9.484271922802364e-06,
+      "loss": 1.0299,
+      "step": 6260
+    },
+    {
+      "epoch": 0.188855421686747,
+      "grad_norm": 4.352710660528723,
+      "learning_rate": 9.482066769281339e-06,
+      "loss": 1.0839,
+      "step": 6270
+    },
+    {
+      "epoch": 0.1891566265060241,
+      "grad_norm": 3.557693000462193,
+      "learning_rate": 9.47985716888806e-06,
+      "loss": 1.019,
+      "step": 6280
+    },
+    {
+      "epoch": 0.1894578313253012,
+      "grad_norm": 2.052304906280264,
+      "learning_rate": 9.47764312381478e-06,
+      "loss": 1.0287,
+      "step": 6290
+    },
+    {
+      "epoch": 0.1897590361445783,
+      "grad_norm": 2.1546406212728684,
+      "learning_rate": 9.475424636258154e-06,
+      "loss": 0.9923,
+      "step": 6300
+    },
+    {
+      "epoch": 0.19006024096385543,
+      "grad_norm": 3.432155846270581,
+      "learning_rate": 9.473201708419255e-06,
+      "loss": 0.9993,
+      "step": 6310
+    },
+    {
+      "epoch": 0.19036144578313252,
+      "grad_norm": 3.3836314481831162,
+      "learning_rate": 9.470974342503554e-06,
+      "loss": 1.0539,
+      "step": 6320
+    },
+    {
+      "epoch": 0.19066265060240964,
+      "grad_norm": 6.410959391767066,
+      "learning_rate": 9.468742540720925e-06,
+      "loss": 1.0704,
+      "step": 6330
+    },
+    {
+      "epoch": 0.19096385542168676,
+      "grad_norm": 3.3994587418544686,
+      "learning_rate": 9.466506305285648e-06,
+      "loss": 0.9905,
+      "step": 6340
+    },
+    {
+      "epoch": 0.19126506024096385,
+      "grad_norm": 3.122653146876219,
+      "learning_rate": 9.464265638416397e-06,
+      "loss": 0.9528,
+      "step": 6350
+    },
+    {
+      "epoch": 0.19156626506024096,
+      "grad_norm": 3.516254140668866,
+      "learning_rate": 9.462020542336243e-06,
+      "loss": 0.9787,
+      "step": 6360
+    },
+    {
+      "epoch": 0.19186746987951808,
+      "grad_norm": 2.210414613219828,
+      "learning_rate": 9.459771019272657e-06,
+      "loss": 1.0348,
+      "step": 6370
+    },
+    {
+      "epoch": 0.19216867469879517,
+      "grad_norm": 3.9327617237349215,
+      "learning_rate": 9.457517071457494e-06,
+      "loss": 1.1597,
+      "step": 6380
+    },
+    {
+      "epoch": 0.1924698795180723,
+      "grad_norm": 2.4111257611704393,
+      "learning_rate": 9.455258701127007e-06,
+      "loss": 1.0841,
+      "step": 6390
+    },
+    {
+      "epoch": 0.1927710843373494,
+      "grad_norm": 1.9627995460222205,
+      "learning_rate": 9.452995910521828e-06,
+      "loss": 1.0638,
+      "step": 6400
+    },
+    {
+      "epoch": 0.1930722891566265,
+      "grad_norm": 11.59027690688135,
+      "learning_rate": 9.450728701886985e-06,
+      "loss": 1.0772,
+      "step": 6410
+    },
+    {
+      "epoch": 0.19337349397590362,
+      "grad_norm": 5.395679302173612,
+      "learning_rate": 9.448457077471878e-06,
+      "loss": 1.0666,
+      "step": 6420
+    },
+    {
+      "epoch": 0.19367469879518073,
+      "grad_norm": 17.619792245169553,
+      "learning_rate": 9.4461810395303e-06,
+      "loss": 0.9663,
+      "step": 6430
+    },
+    {
+      "epoch": 0.19397590361445782,
+      "grad_norm": 5.3293150242242096,
+      "learning_rate": 9.443900590320413e-06,
+      "loss": 1.0672,
+      "step": 6440
+    },
+    {
+      "epoch": 0.19427710843373494,
+      "grad_norm": 2.084201526136716,
+      "learning_rate": 9.441615732104761e-06,
+      "loss": 0.9945,
+      "step": 6450
+    },
+    {
+      "epoch": 0.19457831325301206,
+      "grad_norm": 3.8016068016031275,
+      "learning_rate": 9.43932646715026e-06,
+      "loss": 1.0981,
+      "step": 6460
+    },
+    {
+      "epoch": 0.19487951807228915,
+      "grad_norm": 4.835719823581217,
+      "learning_rate": 9.4370327977282e-06,
+      "loss": 1.1119,
+      "step": 6470
+    },
+    {
+      "epoch": 0.19518072289156627,
+      "grad_norm": 2.344146263507332,
+      "learning_rate": 9.434734726114238e-06,
+      "loss": 0.8969,
+      "step": 6480
+    },
+    {
+      "epoch": 0.19548192771084338,
+      "grad_norm": 3.6141181910894185,
+      "learning_rate": 9.432432254588402e-06,
+      "loss": 0.9771,
+      "step": 6490
+    },
+    {
+      "epoch": 0.19578313253012047,
+      "grad_norm": 3.744360407510648,
+      "learning_rate": 9.430125385435083e-06,
+      "loss": 1.0365,
+      "step": 6500
+    },
+    {
+      "epoch": 0.1960843373493976,
+      "grad_norm": 4.024300378942041,
+      "learning_rate": 9.427814120943037e-06,
+      "loss": 1.0536,
+      "step": 6510
+    },
+    {
+      "epoch": 0.1963855421686747,
+      "grad_norm": 1.9838251481555649,
+      "learning_rate": 9.42549846340538e-06,
+      "loss": 1.0738,
+      "step": 6520
+    },
+    {
+      "epoch": 0.1966867469879518,
+      "grad_norm": 3.8774275284407276,
+      "learning_rate": 9.423178415119585e-06,
+      "loss": 1.0219,
+      "step": 6530
+    },
+    {
+      "epoch": 0.19698795180722892,
+      "grad_norm": 3.4086775694668896,
+      "learning_rate": 9.420853978387482e-06,
+      "loss": 1.0954,
+      "step": 6540
+    },
+    {
+      "epoch": 0.19728915662650603,
+      "grad_norm": 3.190666965732196,
+      "learning_rate": 9.418525155515257e-06,
+      "loss": 0.9372,
+      "step": 6550
+    },
+    {
+      "epoch": 0.19759036144578312,
+      "grad_norm": 3.9995234218922073,
+      "learning_rate": 9.416191948813446e-06,
+      "loss": 1.0769,
+      "step": 6560
+    },
+    {
+      "epoch": 0.19789156626506024,
+      "grad_norm": 4.123718560173802,
+      "learning_rate": 9.413854360596934e-06,
+      "loss": 1.0564,
+      "step": 6570
+    },
+    {
+      "epoch": 0.19819277108433736,
+      "grad_norm": 2.0993768228124416,
+      "learning_rate": 9.411512393184955e-06,
+      "loss": 1.0118,
+      "step": 6580
+    },
+    {
+      "epoch": 0.19849397590361445,
+      "grad_norm": 3.010062786392957,
+      "learning_rate": 9.409166048901084e-06,
+      "loss": 0.963,
+      "step": 6590
+    },
+    {
+      "epoch": 0.19879518072289157,
+      "grad_norm": 3.1623236246645154,
+      "learning_rate": 9.406815330073244e-06,
+      "loss": 1.1061,
+      "step": 6600
+    },
+    {
+      "epoch": 0.19909638554216869,
+      "grad_norm": 3.0914725786047526,
+      "learning_rate": 9.404460239033696e-06,
+      "loss": 1.1769,
+      "step": 6610
+    },
+    {
+      "epoch": 0.19939759036144578,
+      "grad_norm": 2.96996989029304,
+      "learning_rate": 9.402100778119034e-06,
+      "loss": 1.0114,
+      "step": 6620
+    },
+    {
+      "epoch": 0.1996987951807229,
+      "grad_norm": 3.859782714986875,
+      "learning_rate": 9.399736949670195e-06,
+      "loss": 1.1542,
+      "step": 6630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.306360658537134,
+      "learning_rate": 9.397368756032445e-06,
+      "loss": 1.0491,
+      "step": 6640
+    },
+    {
+      "epoch": 0.2003012048192771,
+      "grad_norm": 2.8393886148069942,
+      "learning_rate": 9.394996199555384e-06,
+      "loss": 1.0969,
+      "step": 6650
+    },
+    {
+      "epoch": 0.20060240963855422,
+      "grad_norm": 3.3816692979942027,
+      "learning_rate": 9.392619282592935e-06,
+      "loss": 1.0189,
+      "step": 6660
+    },
+    {
+      "epoch": 0.20090361445783134,
+      "grad_norm": 2.2610058631841583,
+      "learning_rate": 9.390238007503353e-06,
+      "loss": 1.0917,
+      "step": 6670
+    },
+    {
+      "epoch": 0.20120481927710843,
+      "grad_norm": 3.2319729498456633,
+      "learning_rate": 9.387852376649211e-06,
+      "loss": 0.9724,
+      "step": 6680
+    },
+    {
+      "epoch": 0.20150602409638554,
+      "grad_norm": 3.1895675787296627,
+      "learning_rate": 9.385462392397411e-06,
+      "loss": 0.8884,
+      "step": 6690
+    },
+    {
+      "epoch": 0.20180722891566266,
+      "grad_norm": 2.017784678080749,
+      "learning_rate": 9.38306805711917e-06,
+      "loss": 0.9923,
+      "step": 6700
+    },
+    {
+      "epoch": 0.20210843373493975,
+      "grad_norm": 3.7886368832812853,
+      "learning_rate": 9.380669373190021e-06,
+      "loss": 0.9918,
+      "step": 6710
+    },
+    {
+      "epoch": 0.20240963855421687,
+      "grad_norm": 3.044996454038079,
+      "learning_rate": 9.378266342989815e-06,
+      "loss": 0.944,
+      "step": 6720
+    },
+    {
+      "epoch": 0.202710843373494,
+      "grad_norm": 3.2340893547627925,
+      "learning_rate": 9.375858968902709e-06,
+      "loss": 1.023,
+      "step": 6730
+    },
+    {
+      "epoch": 0.20301204819277108,
+      "grad_norm": 3.386251844194327,
+      "learning_rate": 9.373447253317178e-06,
+      "loss": 1.0739,
+      "step": 6740
+    },
+    {
+      "epoch": 0.2033132530120482,
+      "grad_norm": 1.9994776439159567,
+      "learning_rate": 9.371031198625998e-06,
+      "loss": 0.9843,
+      "step": 6750
+    },
+    {
+      "epoch": 0.2036144578313253,
+      "grad_norm": 3.8385337567154414,
+      "learning_rate": 9.368610807226249e-06,
+      "loss": 1.123,
+      "step": 6760
+    },
+    {
+      "epoch": 0.2039156626506024,
+      "grad_norm": 1.9827397135204707,
+      "learning_rate": 9.366186081519322e-06,
+      "loss": 0.9901,
+      "step": 6770
+    },
+    {
+      "epoch": 0.20421686746987952,
+      "grad_norm": 3.132873673691564,
+      "learning_rate": 9.3637570239109e-06,
+      "loss": 0.8976,
+      "step": 6780
+    },
+    {
+      "epoch": 0.20451807228915664,
+      "grad_norm": 2.9914579709001337,
+      "learning_rate": 9.36132363681097e-06,
+      "loss": 1.1023,
+      "step": 6790
+    },
+    {
+      "epoch": 0.20481927710843373,
+      "grad_norm": 3.602994062694188,
+      "learning_rate": 9.358885922633807e-06,
+      "loss": 1.1215,
+      "step": 6800
+    },
+    {
+      "epoch": 0.20512048192771085,
+      "grad_norm": 2.9406954201360453,
+      "learning_rate": 9.356443883797987e-06,
+      "loss": 0.9552,
+      "step": 6810
+    },
+    {
+      "epoch": 0.20542168674698796,
+      "grad_norm": 3.267769576224149,
+      "learning_rate": 9.353997522726372e-06,
+      "loss": 1.0854,
+      "step": 6820
+    },
+    {
+      "epoch": 0.20572289156626505,
+      "grad_norm": 2.8848588751589426,
+      "learning_rate": 9.351546841846113e-06,
+      "loss": 0.9388,
+      "step": 6830
+    },
+    {
+      "epoch": 0.20602409638554217,
+      "grad_norm": 2.1700260524078407,
+      "learning_rate": 9.349091843588647e-06,
+      "loss": 1.0435,
+      "step": 6840
+    },
+    {
+      "epoch": 0.2063253012048193,
+      "grad_norm": 3.603632113043179,
+      "learning_rate": 9.346632530389697e-06,
+      "loss": 1.0195,
+      "step": 6850
+    },
+    {
+      "epoch": 0.20662650602409638,
+      "grad_norm": 1.91925522962616,
+      "learning_rate": 9.344168904689266e-06,
+      "loss": 0.9093,
+      "step": 6860
+    },
+    {
+      "epoch": 0.2069277108433735,
+      "grad_norm": 3.102453015435755,
+      "learning_rate": 9.341700968931628e-06,
+      "loss": 1.0657,
+      "step": 6870
+    },
+    {
+      "epoch": 0.20722891566265061,
+      "grad_norm": 3.162507393683283,
+      "learning_rate": 9.339228725565346e-06,
+      "loss": 1.059,
+      "step": 6880
+    },
+    {
+      "epoch": 0.2075301204819277,
+      "grad_norm": 3.121578902753025,
+      "learning_rate": 9.33675217704325e-06,
+      "loss": 1.0669,
+      "step": 6890
+    },
+    {
+      "epoch": 0.20783132530120482,
+      "grad_norm": 2.7793956733245264,
+      "learning_rate": 9.33427132582244e-06,
+      "loss": 0.9935,
+      "step": 6900
+    },
+    {
+      "epoch": 0.20813253012048194,
+      "grad_norm": 2.085517781341558,
+      "learning_rate": 9.331786174364286e-06,
+      "loss": 0.9213,
+      "step": 6910
+    },
+    {
+      "epoch": 0.20843373493975903,
+      "grad_norm": 2.010153016809384,
+      "learning_rate": 9.329296725134426e-06,
+      "loss": 0.9949,
+      "step": 6920
+    },
+    {
+      "epoch": 0.20873493975903615,
+      "grad_norm": 4.054808349818256,
+      "learning_rate": 9.326802980602764e-06,
+      "loss": 1.0556,
+      "step": 6930
+    },
+    {
+      "epoch": 0.20903614457831327,
+      "grad_norm": 4.505837092533477,
+      "learning_rate": 9.324304943243461e-06,
+      "loss": 0.9794,
+      "step": 6940
+    },
+    {
+      "epoch": 0.20933734939759036,
+      "grad_norm": 2.979603769778444,
+      "learning_rate": 9.32180261553494e-06,
+      "loss": 1.0025,
+      "step": 6950
+    },
+    {
+      "epoch": 0.20963855421686747,
+      "grad_norm": 2.144729692647981,
+      "learning_rate": 9.319295999959877e-06,
+      "loss": 0.9761,
+      "step": 6960
+    },
+    {
+      "epoch": 0.2099397590361446,
+      "grad_norm": 3.486174659452024,
+      "learning_rate": 9.316785099005208e-06,
+      "loss": 0.9381,
+      "step": 6970
+    },
+    {
+      "epoch": 0.21024096385542168,
+      "grad_norm": 3.184600253664155,
+      "learning_rate": 9.314269915162115e-06,
+      "loss": 1.0594,
+      "step": 6980
+    },
+    {
+      "epoch": 0.2105421686746988,
+      "grad_norm": 6.215920268789802,
+      "learning_rate": 9.311750450926034e-06,
+      "loss": 1.0289,
+      "step": 6990
+    },
+    {
+      "epoch": 0.21084337349397592,
+      "grad_norm": 3.3514667059868275,
+      "learning_rate": 9.309226708796643e-06,
+      "loss": 1.0761,
+      "step": 7000
+    },
+    {
+      "epoch": 0.211144578313253,
+      "grad_norm": 3.25441268151538,
+      "learning_rate": 9.30669869127787e-06,
+      "loss": 1.101,
+      "step": 7010
+    },
+    {
+      "epoch": 0.21144578313253012,
+      "grad_norm": 3.5027058202888646,
+      "learning_rate": 9.304166400877882e-06,
+      "loss": 0.9201,
+      "step": 7020
+    },
+    {
+      "epoch": 0.21174698795180724,
+      "grad_norm": 2.9819940458452736,
+      "learning_rate": 9.301629840109082e-06,
+      "loss": 1.0614,
+      "step": 7030
+    },
+    {
+      "epoch": 0.21204819277108433,
+      "grad_norm": 1.924930162325777,
+      "learning_rate": 9.299089011488115e-06,
+      "loss": 1.0137,
+      "step": 7040
+    },
+    {
+      "epoch": 0.21234939759036145,
+      "grad_norm": 2.3023141591532643,
+      "learning_rate": 9.296543917535859e-06,
+      "loss": 0.9356,
+      "step": 7050
+    },
+    {
+      "epoch": 0.21265060240963857,
+      "grad_norm": 2.075460650189826,
+      "learning_rate": 9.293994560777419e-06,
+      "loss": 1.013,
+      "step": 7060
+    },
+    {
+      "epoch": 0.21295180722891566,
+      "grad_norm": 2.044014707663297,
+      "learning_rate": 9.291440943742138e-06,
+      "loss": 0.8399,
+      "step": 7070
+    },
+    {
+      "epoch": 0.21325301204819277,
+      "grad_norm": 3.2085697316677417,
+      "learning_rate": 9.288883068963581e-06,
+      "loss": 0.9995,
+      "step": 7080
+    },
+    {
+      "epoch": 0.2135542168674699,
+      "grad_norm": 3.2393919790610703,
+      "learning_rate": 9.286320938979533e-06,
+      "loss": 1.0997,
+      "step": 7090
+    },
+    {
+      "epoch": 0.21385542168674698,
+      "grad_norm": 3.1739107295653097,
+      "learning_rate": 9.283754556332012e-06,
+      "loss": 1.0699,
+      "step": 7100
+    },
+    {
+      "epoch": 0.2141566265060241,
+      "grad_norm": 3.0110951714632757,
+      "learning_rate": 9.281183923567245e-06,
+      "loss": 1.0548,
+      "step": 7110
+    },
+    {
+      "epoch": 0.21445783132530122,
+      "grad_norm": 4.18574638433286,
+      "learning_rate": 9.278609043235678e-06,
+      "loss": 1.0935,
+      "step": 7120
+    },
+    {
+      "epoch": 0.2147590361445783,
+      "grad_norm": 3.326876881645158,
+      "learning_rate": 9.276029917891976e-06,
+      "loss": 1.0521,
+      "step": 7130
+    },
+    {
+      "epoch": 0.21506024096385543,
+      "grad_norm": 3.210473294446714,
+      "learning_rate": 9.27344655009501e-06,
+      "loss": 1.0428,
+      "step": 7140
+    },
+    {
+      "epoch": 0.21536144578313254,
+      "grad_norm": 2.1345313470502436,
+      "learning_rate": 9.27085894240786e-06,
+      "loss": 1.0272,
+      "step": 7150
+    },
+    {
+      "epoch": 0.21566265060240963,
+      "grad_norm": 2.2664811897689847,
+      "learning_rate": 9.268267097397824e-06,
+      "loss": 0.9575,
+      "step": 7160
+    },
+    {
+      "epoch": 0.21596385542168675,
+      "grad_norm": 3.184695167512654,
+      "learning_rate": 9.265671017636384e-06,
+      "loss": 1.0909,
+      "step": 7170
+    },
+    {
+      "epoch": 0.21626506024096387,
+      "grad_norm": 2.1459242715287945,
+      "learning_rate": 9.263070705699241e-06,
+      "loss": 1.08,
+      "step": 7180
+    },
+    {
+      "epoch": 0.21656626506024096,
+      "grad_norm": 2.2353506779574825,
+      "learning_rate": 9.260466164166289e-06,
+      "loss": 1.0423,
+      "step": 7190
+    },
+    {
+      "epoch": 0.21686746987951808,
+      "grad_norm": 3.0059533014163793,
+      "learning_rate": 9.257857395621614e-06,
+      "loss": 1.008,
+      "step": 7200
+    },
+    {
+      "epoch": 0.2171686746987952,
+      "grad_norm": 3.804731492845275,
+      "learning_rate": 9.255244402653504e-06,
+      "loss": 0.9525,
+      "step": 7210
+    },
+    {
+      "epoch": 0.21746987951807228,
+      "grad_norm": 2.1694769108362753,
+      "learning_rate": 9.25262718785443e-06,
+      "loss": 1.033,
+      "step": 7220
+    },
+    {
+      "epoch": 0.2177710843373494,
+      "grad_norm": 3.279818738883482,
+      "learning_rate": 9.250005753821059e-06,
+      "loss": 1.0243,
+      "step": 7230
+    },
+    {
+      "epoch": 0.21807228915662652,
+      "grad_norm": 3.483345330928486,
+      "learning_rate": 9.247380103154238e-06,
+      "loss": 1.1496,
+      "step": 7240
+    },
+    {
+      "epoch": 0.2183734939759036,
+      "grad_norm": 3.060128588511046,
+      "learning_rate": 9.244750238459002e-06,
+      "loss": 1.0503,
+      "step": 7250
+    },
+    {
+      "epoch": 0.21867469879518073,
+      "grad_norm": 3.4930189068193735,
+      "learning_rate": 9.242116162344564e-06,
+      "loss": 1.087,
+      "step": 7260
+    },
+    {
+      "epoch": 0.21897590361445782,
+      "grad_norm": 1.973294434777584,
+      "learning_rate": 9.239477877424317e-06,
+      "loss": 1.0159,
+      "step": 7270
+    },
+    {
+      "epoch": 0.21927710843373494,
+      "grad_norm": 3.195959866031689,
+      "learning_rate": 9.23683538631583e-06,
+      "loss": 1.0611,
+      "step": 7280
+    },
+    {
+      "epoch": 0.21957831325301205,
+      "grad_norm": 3.6241550357476258,
+      "learning_rate": 9.234188691640846e-06,
+      "loss": 0.9758,
+      "step": 7290
+    },
+    {
+      "epoch": 0.21987951807228914,
+      "grad_norm": 4.35318569238192,
+      "learning_rate": 9.231537796025275e-06,
+      "loss": 0.8874,
+      "step": 7300
+    },
+    {
+      "epoch": 0.22018072289156626,
+      "grad_norm": 3.818817870724282,
+      "learning_rate": 9.228882702099194e-06,
+      "loss": 1.1088,
+      "step": 7310
+    },
+    {
+      "epoch": 0.22048192771084338,
+      "grad_norm": 4.832753181216332,
+      "learning_rate": 9.226223412496852e-06,
+      "loss": 1.0271,
+      "step": 7320
+    },
+    {
+      "epoch": 0.22078313253012047,
+      "grad_norm": 3.563928636536491,
+      "learning_rate": 9.22355992985666e-06,
+      "loss": 0.9431,
+      "step": 7330
+    },
+    {
+      "epoch": 0.22108433734939759,
+      "grad_norm": 2.106032723927942,
+      "learning_rate": 9.22089225682118e-06,
+      "loss": 0.9198,
+      "step": 7340
+    },
+    {
+      "epoch": 0.2213855421686747,
+      "grad_norm": 3.6012656698254513,
+      "learning_rate": 9.218220396037142e-06,
+      "loss": 0.9619,
+      "step": 7350
+    },
+    {
+      "epoch": 0.2216867469879518,
+      "grad_norm": 4.2692913382684745,
+      "learning_rate": 9.215544350155423e-06,
+      "loss": 1.1123,
+      "step": 7360
+    },
+    {
+      "epoch": 0.2219879518072289,
+      "grad_norm": 3.1337403317924464,
+      "learning_rate": 9.212864121831059e-06,
+      "loss": 0.961,
+      "step": 7370
+    },
+    {
+      "epoch": 0.22228915662650603,
+      "grad_norm": 4.758260061052883,
+      "learning_rate": 9.21017971372323e-06,
+      "loss": 0.9874,
+      "step": 7380
+    },
+    {
+      "epoch": 0.22259036144578312,
+      "grad_norm": 5.2849022294711085,
+      "learning_rate": 9.207491128495267e-06,
+      "loss": 1.024,
+      "step": 7390
+    },
+    {
+      "epoch": 0.22289156626506024,
+      "grad_norm": 2.1049321654664546,
+      "learning_rate": 9.204798368814645e-06,
+      "loss": 0.9812,
+      "step": 7400
+    },
+    {
+      "epoch": 0.22319277108433735,
+      "grad_norm": 4.772520821330163,
+      "learning_rate": 9.202101437352974e-06,
+      "loss": 1.0492,
+      "step": 7410
+    },
+    {
+      "epoch": 0.22349397590361444,
+      "grad_norm": 4.363365337409341,
+      "learning_rate": 9.199400336786016e-06,
+      "loss": 1.0548,
+      "step": 7420
+    },
+    {
+      "epoch": 0.22379518072289156,
+      "grad_norm": 3.7654678746095813,
+      "learning_rate": 9.196695069793652e-06,
+      "loss": 1.0793,
+      "step": 7430
+    },
+    {
+      "epoch": 0.22409638554216868,
+      "grad_norm": 5.043492783400528,
+      "learning_rate": 9.193985639059917e-06,
+      "loss": 1.0579,
+      "step": 7440
+    },
+    {
+      "epoch": 0.22439759036144577,
+      "grad_norm": 4.163967740388339,
+      "learning_rate": 9.19127204727296e-06,
+      "loss": 0.9827,
+      "step": 7450
+    },
+    {
+      "epoch": 0.2246987951807229,
+      "grad_norm": 3.6302790775841434,
+      "learning_rate": 9.188554297125065e-06,
+      "loss": 0.9875,
+      "step": 7460
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 2.296005109334423,
+      "learning_rate": 9.185832391312644e-06,
+      "loss": 0.9688,
+      "step": 7470
+    },
+    {
+      "epoch": 0.2253012048192771,
+      "grad_norm": 9.846179925506112,
+      "learning_rate": 9.183106332536228e-06,
+      "loss": 1.0282,
+      "step": 7480
+    },
+    {
+      "epoch": 0.2256024096385542,
+      "grad_norm": 5.056630245306455,
+      "learning_rate": 9.18037612350047e-06,
+      "loss": 0.9888,
+      "step": 7490
+    },
+    {
+      "epoch": 0.22590361445783133,
+      "grad_norm": 5.431301453348206,
+      "learning_rate": 9.177641766914143e-06,
+      "loss": 1.021,
+      "step": 7500
+    },
+    {
+      "epoch": 0.22620481927710842,
+      "grad_norm": 3.301911571634357,
+      "learning_rate": 9.174903265490128e-06,
+      "loss": 0.955,
+      "step": 7510
+    },
+    {
+      "epoch": 0.22650602409638554,
+      "grad_norm": 3.5962629099331647,
+      "learning_rate": 9.172160621945428e-06,
+      "loss": 0.9391,
+      "step": 7520
+    },
+    {
+      "epoch": 0.22680722891566266,
+      "grad_norm": 2.269448713410476,
+      "learning_rate": 9.169413839001147e-06,
+      "loss": 1.0487,
+      "step": 7530
+    },
+    {
+      "epoch": 0.22710843373493975,
+      "grad_norm": 3.8344296994307996,
+      "learning_rate": 9.1666629193825e-06,
+      "loss": 0.9021,
+      "step": 7540
+    },
+    {
+      "epoch": 0.22740963855421686,
+      "grad_norm": 4.164141091617803,
+      "learning_rate": 9.163907865818806e-06,
+      "loss": 1.0118,
+      "step": 7550
+    },
+    {
+      "epoch": 0.22771084337349398,
+      "grad_norm": 2.0103672358412963,
+      "learning_rate": 9.161148681043486e-06,
+      "loss": 0.9915,
+      "step": 7560
+    },
+    {
+      "epoch": 0.22801204819277107,
+      "grad_norm": 3.8544760596594614,
+      "learning_rate": 9.158385367794056e-06,
+      "loss": 1.0487,
+      "step": 7570
+    },
+    {
+      "epoch": 0.2283132530120482,
+      "grad_norm": 3.237976024176194,
+      "learning_rate": 9.155617928812134e-06,
+      "loss": 0.8809,
+      "step": 7580
+    },
+    {
+      "epoch": 0.2286144578313253,
+      "grad_norm": 3.757167149271603,
+      "learning_rate": 9.152846366843428e-06,
+      "loss": 1.1089,
+      "step": 7590
+    },
+    {
+      "epoch": 0.2289156626506024,
+      "grad_norm": 2.2128471871795075,
+      "learning_rate": 9.150070684637736e-06,
+      "loss": 1.0245,
+      "step": 7600
+    },
+    {
+      "epoch": 0.22921686746987951,
+      "grad_norm": 4.482785826969745,
+      "learning_rate": 9.147290884948943e-06,
+      "loss": 1.0492,
+      "step": 7610
+    },
+    {
+      "epoch": 0.22951807228915663,
+      "grad_norm": 2.0738750013951046,
+      "learning_rate": 9.144506970535023e-06,
+      "loss": 0.9468,
+      "step": 7620
+    },
+    {
+      "epoch": 0.22981927710843372,
+      "grad_norm": 3.7297022255307763,
+      "learning_rate": 9.14171894415803e-06,
+      "loss": 0.9987,
+      "step": 7630
+    },
+    {
+      "epoch": 0.23012048192771084,
+      "grad_norm": 3.190077549805851,
+      "learning_rate": 9.138926808584098e-06,
+      "loss": 1.0469,
+      "step": 7640
+    },
+    {
+      "epoch": 0.23042168674698796,
+      "grad_norm": 1.9866370370988622,
+      "learning_rate": 9.136130566583436e-06,
+      "loss": 0.8609,
+      "step": 7650
+    },
+    {
+      "epoch": 0.23072289156626505,
+      "grad_norm": 3.07394379390921,
+      "learning_rate": 9.133330220930333e-06,
+      "loss": 1.0099,
+      "step": 7660
+    },
+    {
+      "epoch": 0.23102409638554217,
+      "grad_norm": 3.0863871770055353,
+      "learning_rate": 9.130525774403143e-06,
+      "loss": 1.0596,
+      "step": 7670
+    },
+    {
+      "epoch": 0.23132530120481928,
+      "grad_norm": 3.0317240789476902,
+      "learning_rate": 9.12771722978429e-06,
+      "loss": 1.0161,
+      "step": 7680
+    },
+    {
+      "epoch": 0.23162650602409637,
+      "grad_norm": 3.18180990792581,
+      "learning_rate": 9.124904589860269e-06,
+      "loss": 1.0686,
+      "step": 7690
+    },
+    {
+      "epoch": 0.2319277108433735,
+      "grad_norm": 3.1169520703622404,
+      "learning_rate": 9.122087857421628e-06,
+      "loss": 1.0192,
+      "step": 7700
+    },
+    {
+      "epoch": 0.2322289156626506,
+      "grad_norm": 2.9204592873855035,
+      "learning_rate": 9.119267035262989e-06,
+      "loss": 0.8653,
+      "step": 7710
+    },
+    {
+      "epoch": 0.2325301204819277,
+      "grad_norm": 4.344511588071381,
+      "learning_rate": 9.116442126183018e-06,
+      "loss": 1.0525,
+      "step": 7720
+    },
+    {
+      "epoch": 0.23283132530120482,
+      "grad_norm": 2.9238280827556165,
+      "learning_rate": 9.113613132984446e-06,
+      "loss": 0.9894,
+      "step": 7730
+    },
+    {
+      "epoch": 0.23313253012048193,
+      "grad_norm": 3.391027204375839,
+      "learning_rate": 9.110780058474052e-06,
+      "loss": 1.0166,
+      "step": 7740
+    },
+    {
+      "epoch": 0.23343373493975902,
+      "grad_norm": 3.2196892544578857,
+      "learning_rate": 9.107942905462662e-06,
+      "loss": 0.9522,
+      "step": 7750
+    },
+    {
+      "epoch": 0.23373493975903614,
+      "grad_norm": 3.7064446670426556,
+      "learning_rate": 9.10510167676515e-06,
+      "loss": 1.0051,
+      "step": 7760
+    },
+    {
+      "epoch": 0.23403614457831326,
+      "grad_norm": 2.9697209146836476,
+      "learning_rate": 9.102256375200438e-06,
+      "loss": 1.0173,
+      "step": 7770
+    },
+    {
+      "epoch": 0.23433734939759035,
+      "grad_norm": 2.2055470227618548,
+      "learning_rate": 9.099407003591485e-06,
+      "loss": 0.9593,
+      "step": 7780
+    },
+    {
+      "epoch": 0.23463855421686747,
+      "grad_norm": 3.247569268304636,
+      "learning_rate": 9.096553564765286e-06,
+      "loss": 0.9129,
+      "step": 7790
+    },
+    {
+      "epoch": 0.23493975903614459,
+      "grad_norm": 3.5284419683223796,
+      "learning_rate": 9.093696061552876e-06,
+      "loss": 1.026,
+      "step": 7800
+    },
+    {
+      "epoch": 0.23524096385542168,
+      "grad_norm": 2.9504014391893354,
+      "learning_rate": 9.090834496789319e-06,
+      "loss": 0.9459,
+      "step": 7810
+    },
+    {
+      "epoch": 0.2355421686746988,
+      "grad_norm": 2.8731903409884842,
+      "learning_rate": 9.08796887331371e-06,
+      "loss": 1.0403,
+      "step": 7820
+    },
+    {
+      "epoch": 0.2358433734939759,
+      "grad_norm": 3.9923686182409535,
+      "learning_rate": 9.085099193969172e-06,
+      "loss": 0.9725,
+      "step": 7830
+    },
+    {
+      "epoch": 0.236144578313253,
+      "grad_norm": 2.887965087195908,
+      "learning_rate": 9.082225461602849e-06,
+      "loss": 1.0674,
+      "step": 7840
+    },
+    {
+      "epoch": 0.23644578313253012,
+      "grad_norm": 3.461852254021425,
+      "learning_rate": 9.079347679065909e-06,
+      "loss": 1.1639,
+      "step": 7850
+    },
+    {
+      "epoch": 0.23674698795180724,
+      "grad_norm": 3.2103801236890024,
+      "learning_rate": 9.076465849213538e-06,
+      "loss": 1.0201,
+      "step": 7860
+    },
+    {
+      "epoch": 0.23704819277108433,
+      "grad_norm": 3.0429166797677483,
+      "learning_rate": 9.073579974904936e-06,
+      "loss": 0.9673,
+      "step": 7870
+    },
+    {
+      "epoch": 0.23734939759036144,
+      "grad_norm": 2.306017900996041,
+      "learning_rate": 9.070690059003316e-06,
+      "loss": 0.9704,
+      "step": 7880
+    },
+    {
+      "epoch": 0.23765060240963856,
+      "grad_norm": 2.98592283815459,
+      "learning_rate": 9.067796104375903e-06,
+      "loss": 1.1364,
+      "step": 7890
+    },
+    {
+      "epoch": 0.23795180722891565,
+      "grad_norm": 3.4606592270355945,
+      "learning_rate": 9.064898113893925e-06,
+      "loss": 1.0417,
+      "step": 7900
+    },
+    {
+      "epoch": 0.23825301204819277,
+      "grad_norm": 3.372271891702563,
+      "learning_rate": 9.061996090432618e-06,
+      "loss": 1.0594,
+      "step": 7910
+    },
+    {
+      "epoch": 0.2385542168674699,
+      "grad_norm": 3.229179997877593,
+      "learning_rate": 9.059090036871221e-06,
+      "loss": 1.1316,
+      "step": 7920
+    },
+    {
+      "epoch": 0.23885542168674698,
+      "grad_norm": 3.2308663576330185,
+      "learning_rate": 9.056179956092961e-06,
+      "loss": 1.0451,
+      "step": 7930
+    },
+    {
+      "epoch": 0.2391566265060241,
+      "grad_norm": 3.3957658720642376,
+      "learning_rate": 9.053265850985075e-06,
+      "loss": 1.0646,
+      "step": 7940
+    },
+    {
+      "epoch": 0.2394578313253012,
+      "grad_norm": 3.176042552942464,
+      "learning_rate": 9.050347724438785e-06,
+      "loss": 0.9957,
+      "step": 7950
+    },
+    {
+      "epoch": 0.2397590361445783,
+      "grad_norm": 4.48528608034301,
+      "learning_rate": 9.047425579349298e-06,
+      "loss": 0.9992,
+      "step": 7960
+    },
+    {
+      "epoch": 0.24006024096385542,
+      "grad_norm": 2.1282679119321743,
+      "learning_rate": 9.044499418615817e-06,
+      "loss": 1.0291,
+      "step": 7970
+    },
+    {
+      "epoch": 0.24036144578313254,
+      "grad_norm": 2.736142662871743,
+      "learning_rate": 9.041569245141528e-06,
+      "loss": 0.9823,
+      "step": 7980
+    },
+    {
+      "epoch": 0.24066265060240963,
+      "grad_norm": 3.3382841573409188,
+      "learning_rate": 9.038635061833592e-06,
+      "loss": 1.0627,
+      "step": 7990
+    },
+    {
+      "epoch": 0.24096385542168675,
+      "grad_norm": 2.024815484120407,
+      "learning_rate": 9.035696871603155e-06,
+      "loss": 0.9441,
+      "step": 8000
+    },
+    {
+      "epoch": 0.24126506024096386,
+      "grad_norm": 4.289359937594581,
+      "learning_rate": 9.032754677365335e-06,
+      "loss": 0.9866,
+      "step": 8010
+    },
+    {
+      "epoch": 0.24156626506024095,
+      "grad_norm": 4.359534389293543,
+      "learning_rate": 9.029808482039221e-06,
+      "loss": 1.0458,
+      "step": 8020
+    },
+    {
+      "epoch": 0.24186746987951807,
+      "grad_norm": 3.268358974643368,
+      "learning_rate": 9.026858288547877e-06,
+      "loss": 1.0405,
+      "step": 8030
+    },
+    {
+      "epoch": 0.2421686746987952,
+      "grad_norm": 2.2008962339669287,
+      "learning_rate": 9.02390409981833e-06,
+      "loss": 1.0244,
+      "step": 8040
+    },
+    {
+      "epoch": 0.24246987951807228,
+      "grad_norm": 2.118757535566222,
+      "learning_rate": 9.02094591878157e-06,
+      "loss": 0.9002,
+      "step": 8050
+    },
+    {
+      "epoch": 0.2427710843373494,
+      "grad_norm": 3.2090755214164566,
+      "learning_rate": 9.01798374837255e-06,
+      "loss": 1.1051,
+      "step": 8060
+    },
+    {
+      "epoch": 0.24307228915662651,
+      "grad_norm": 3.042220359326803,
+      "learning_rate": 9.015017591530183e-06,
+      "loss": 0.9293,
+      "step": 8070
+    },
+    {
+      "epoch": 0.2433734939759036,
+      "grad_norm": 3.426262304434553,
+      "learning_rate": 9.01204745119733e-06,
+      "loss": 1.1623,
+      "step": 8080
+    },
+    {
+      "epoch": 0.24367469879518072,
+      "grad_norm": 2.0858619438910204,
+      "learning_rate": 9.009073330320814e-06,
+      "loss": 0.8809,
+      "step": 8090
+    },
+    {
+      "epoch": 0.24397590361445784,
+      "grad_norm": 3.2517862161832367,
+      "learning_rate": 9.0060952318514e-06,
+      "loss": 0.9687,
+      "step": 8100
+    },
+    {
+      "epoch": 0.24427710843373493,
+      "grad_norm": 5.385008730635654,
+      "learning_rate": 9.003113158743799e-06,
+      "loss": 0.9973,
+      "step": 8110
+    },
+    {
+      "epoch": 0.24457831325301205,
+      "grad_norm": 3.260874636870826,
+      "learning_rate": 9.000127113956673e-06,
+      "loss": 1.1001,
+      "step": 8120
+    },
+    {
+      "epoch": 0.24487951807228917,
+      "grad_norm": 3.748508357440099,
+      "learning_rate": 8.997137100452617e-06,
+      "loss": 1.0199,
+      "step": 8130
+    },
+    {
+      "epoch": 0.24518072289156626,
+      "grad_norm": 2.3309099545563647,
+      "learning_rate": 8.994143121198165e-06,
+      "loss": 1.0364,
+      "step": 8140
+    },
+    {
+      "epoch": 0.24548192771084337,
+      "grad_norm": 3.456354645387568,
+      "learning_rate": 8.991145179163786e-06,
+      "loss": 0.9503,
+      "step": 8150
+    },
+    {
+      "epoch": 0.2457831325301205,
+      "grad_norm": 2.814700506762087,
+      "learning_rate": 8.988143277323886e-06,
+      "loss": 1.0362,
+      "step": 8160
+    },
+    {
+      "epoch": 0.24608433734939758,
+      "grad_norm": 2.12355078102478,
+      "learning_rate": 8.98513741865679e-06,
+      "loss": 1.0603,
+      "step": 8170
+    },
+    {
+      "epoch": 0.2463855421686747,
+      "grad_norm": 2.0907205765314076,
+      "learning_rate": 8.982127606144757e-06,
+      "loss": 1.0424,
+      "step": 8180
+    },
+    {
+      "epoch": 0.24668674698795182,
+      "grad_norm": 2.0393725120506816,
+      "learning_rate": 8.979113842773963e-06,
+      "loss": 0.9431,
+      "step": 8190
+    },
+    {
+      "epoch": 0.2469879518072289,
+      "grad_norm": 3.5145116975187687,
+      "learning_rate": 8.976096131534509e-06,
+      "loss": 0.9675,
+      "step": 8200
+    },
+    {
+      "epoch": 0.24728915662650602,
+      "grad_norm": 3.4096899005064016,
+      "learning_rate": 8.973074475420407e-06,
+      "loss": 0.9713,
+      "step": 8210
+    },
+    {
+      "epoch": 0.24759036144578314,
+      "grad_norm": 3.4456131377311445,
+      "learning_rate": 8.970048877429589e-06,
+      "loss": 1.0266,
+      "step": 8220
+    },
+    {
+      "epoch": 0.24789156626506023,
+      "grad_norm": 3.28229849253706,
+      "learning_rate": 8.967019340563895e-06,
+      "loss": 1.0719,
+      "step": 8230
+    },
+    {
+      "epoch": 0.24819277108433735,
+      "grad_norm": 3.029145331734003,
+      "learning_rate": 8.963985867829071e-06,
+      "loss": 1.0857,
+      "step": 8240
+    },
+    {
+      "epoch": 0.24849397590361447,
+      "grad_norm": 3.5855129324830393,
+      "learning_rate": 8.960948462234774e-06,
+      "loss": 0.9337,
+      "step": 8250
+    },
+    {
+      "epoch": 0.24879518072289156,
+      "grad_norm": 3.0901780732506396,
+      "learning_rate": 8.957907126794554e-06,
+      "loss": 1.0781,
+      "step": 8260
+    },
+    {
+      "epoch": 0.24909638554216867,
+      "grad_norm": 3.121913022289746,
+      "learning_rate": 8.954861864525868e-06,
+      "loss": 1.0861,
+      "step": 8270
+    },
+    {
+      "epoch": 0.2493975903614458,
+      "grad_norm": 2.1150872922991755,
+      "learning_rate": 8.951812678450064e-06,
+      "loss": 0.953,
+      "step": 8280
+    },
+    {
+      "epoch": 0.24969879518072288,
+      "grad_norm": 3.101876179688924,
+      "learning_rate": 8.948759571592387e-06,
+      "loss": 0.9801,
+      "step": 8290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9124587096697256,
+      "learning_rate": 8.94570254698197e-06,
+      "loss": 1.0535,
+      "step": 8300
+    },
+    {
+      "epoch": 0.2503012048192771,
+      "grad_norm": 3.407308686234648,
+      "learning_rate": 8.94264160765183e-06,
+      "loss": 0.996,
+      "step": 8310
+    },
+    {
+      "epoch": 0.25060240963855424,
+      "grad_norm": 2.155869125435714,
+      "learning_rate": 8.939576756638874e-06,
+      "loss": 0.9361,
+      "step": 8320
+    },
+    {
+      "epoch": 0.2509036144578313,
+      "grad_norm": 1.8777188079853016,
+      "learning_rate": 8.936507996983887e-06,
+      "loss": 0.9956,
+      "step": 8330
+    },
+    {
+      "epoch": 0.2512048192771084,
+      "grad_norm": 3.155861948679197,
+      "learning_rate": 8.933435331731528e-06,
+      "loss": 0.9806,
+      "step": 8340
+    },
+    {
+      "epoch": 0.25150602409638556,
+      "grad_norm": 2.017966739523534,
+      "learning_rate": 8.93035876393034e-06,
+      "loss": 1.0198,
+      "step": 8350
+    },
+    {
+      "epoch": 0.25180722891566265,
+      "grad_norm": 5.9755373662877265,
+      "learning_rate": 8.92727829663273e-06,
+      "loss": 1.0251,
+      "step": 8360
+    },
+    {
+      "epoch": 0.25210843373493974,
+      "grad_norm": 4.003781774292372,
+      "learning_rate": 8.924193932894975e-06,
+      "loss": 1.0323,
+      "step": 8370
+    },
+    {
+      "epoch": 0.2524096385542169,
+      "grad_norm": 4.131019163865079,
+      "learning_rate": 8.921105675777224e-06,
+      "loss": 1.1216,
+      "step": 8380
+    },
+    {
+      "epoch": 0.252710843373494,
+      "grad_norm": 3.3247654634034123,
+      "learning_rate": 8.91801352834348e-06,
+      "loss": 1.0649,
+      "step": 8390
+    },
+    {
+      "epoch": 0.25301204819277107,
+      "grad_norm": 3.729922888326644,
+      "learning_rate": 8.91491749366161e-06,
+      "loss": 0.9558,
+      "step": 8400
+    },
+    {
+      "epoch": 0.2533132530120482,
+      "grad_norm": 10.224188820758755,
+      "learning_rate": 8.91181757480334e-06,
+      "loss": 0.9427,
+      "step": 8410
+    },
+    {
+      "epoch": 0.2536144578313253,
+      "grad_norm": 2.038080876914104,
+      "learning_rate": 8.908713774844246e-06,
+      "loss": 1.0966,
+      "step": 8420
+    },
+    {
+      "epoch": 0.2539156626506024,
+      "grad_norm": 3.159311381360028,
+      "learning_rate": 8.905606096863758e-06,
+      "loss": 0.9847,
+      "step": 8430
+    },
+    {
+      "epoch": 0.25421686746987954,
+      "grad_norm": 4.105394915861901,
+      "learning_rate": 8.90249454394515e-06,
+      "loss": 0.9863,
+      "step": 8440
+    },
+    {
+      "epoch": 0.2545180722891566,
+      "grad_norm": 2.987840420686699,
+      "learning_rate": 8.899379119175539e-06,
+      "loss": 0.9878,
+      "step": 8450
+    },
+    {
+      "epoch": 0.2548192771084337,
+      "grad_norm": 3.418140227575701,
+      "learning_rate": 8.896259825645892e-06,
+      "loss": 0.9998,
+      "step": 8460
+    },
+    {
+      "epoch": 0.25512048192771086,
+      "grad_norm": 6.400861492734381,
+      "learning_rate": 8.893136666451008e-06,
+      "loss": 0.9427,
+      "step": 8470
+    },
+    {
+      "epoch": 0.25542168674698795,
+      "grad_norm": 2.103983548502195,
+      "learning_rate": 8.890009644689516e-06,
+      "loss": 1.0309,
+      "step": 8480
+    },
+    {
+      "epoch": 0.25572289156626504,
+      "grad_norm": 2.3373928055118256,
+      "learning_rate": 8.886878763463893e-06,
+      "loss": 0.9713,
+      "step": 8490
+    },
+    {
+      "epoch": 0.2560240963855422,
+      "grad_norm": 2.2378882334219425,
+      "learning_rate": 8.883744025880429e-06,
+      "loss": 1.0063,
+      "step": 8500
+    },
+    {
+      "epoch": 0.2563253012048193,
+      "grad_norm": 2.399024881939655,
+      "learning_rate": 8.880605435049247e-06,
+      "loss": 1.0015,
+      "step": 8510
+    },
+    {
+      "epoch": 0.25662650602409637,
+      "grad_norm": 3.5593274010218217,
+      "learning_rate": 8.877462994084296e-06,
+      "loss": 0.963,
+      "step": 8520
+    },
+    {
+      "epoch": 0.2569277108433735,
+      "grad_norm": 3.3842963179891226,
+      "learning_rate": 8.874316706103339e-06,
+      "loss": 0.9556,
+      "step": 8530
+    },
+    {
+      "epoch": 0.2572289156626506,
+      "grad_norm": 3.6922453483063364,
+      "learning_rate": 8.871166574227961e-06,
+      "loss": 1.0794,
+      "step": 8540
+    },
+    {
+      "epoch": 0.2575301204819277,
+      "grad_norm": 2.051465817210662,
+      "learning_rate": 8.868012601583553e-06,
+      "loss": 0.9485,
+      "step": 8550
+    },
+    {
+      "epoch": 0.25783132530120484,
+      "grad_norm": 3.4485292723014718,
+      "learning_rate": 8.864854791299328e-06,
+      "loss": 1.0427,
+      "step": 8560
+    },
+    {
+      "epoch": 0.25813253012048193,
+      "grad_norm": 1.9121000707312048,
+      "learning_rate": 8.861693146508295e-06,
+      "loss": 0.8302,
+      "step": 8570
+    },
+    {
+      "epoch": 0.258433734939759,
+      "grad_norm": 3.426783290699053,
+      "learning_rate": 8.858527670347277e-06,
+      "loss": 0.9892,
+      "step": 8580
+    },
+    {
+      "epoch": 0.25873493975903616,
+      "grad_norm": 2.110351300097445,
+      "learning_rate": 8.85535836595689e-06,
+      "loss": 0.9055,
+      "step": 8590
+    },
+    {
+      "epoch": 0.25903614457831325,
+      "grad_norm": 3.499859287241071,
+      "learning_rate": 8.852185236481553e-06,
+      "loss": 1.008,
+      "step": 8600
+    },
+    {
+      "epoch": 0.25933734939759034,
+      "grad_norm": 1.967989273129067,
+      "learning_rate": 8.849008285069482e-06,
+      "loss": 0.9881,
+      "step": 8610
+    },
+    {
+      "epoch": 0.2596385542168675,
+      "grad_norm": 3.7193727294787355,
+      "learning_rate": 8.84582751487268e-06,
+      "loss": 0.9405,
+      "step": 8620
+    },
+    {
+      "epoch": 0.2599397590361446,
+      "grad_norm": 3.205138528859601,
+      "learning_rate": 8.84264292904694e-06,
+      "loss": 0.8871,
+      "step": 8630
+    },
+    {
+      "epoch": 0.26024096385542167,
+      "grad_norm": 3.105480829447051,
+      "learning_rate": 8.839454530751842e-06,
+      "loss": 1.0926,
+      "step": 8640
+    },
+    {
+      "epoch": 0.2605421686746988,
+      "grad_norm": 3.7813810151424043,
+      "learning_rate": 8.83626232315075e-06,
+      "loss": 0.8883,
+      "step": 8650
+    },
+    {
+      "epoch": 0.2608433734939759,
+      "grad_norm": 3.8266788365773072,
+      "learning_rate": 8.833066309410802e-06,
+      "loss": 1.0007,
+      "step": 8660
+    },
+    {
+      "epoch": 0.261144578313253,
+      "grad_norm": 4.250171524841646,
+      "learning_rate": 8.829866492702918e-06,
+      "loss": 1.0061,
+      "step": 8670
+    },
+    {
+      "epoch": 0.26144578313253014,
+      "grad_norm": 6.36063003016339,
+      "learning_rate": 8.826662876201788e-06,
+      "loss": 0.9953,
+      "step": 8680
+    },
+    {
+      "epoch": 0.26174698795180723,
+      "grad_norm": 3.564978005434069,
+      "learning_rate": 8.823455463085873e-06,
+      "loss": 0.9286,
+      "step": 8690
+    },
+    {
+      "epoch": 0.2620481927710843,
+      "grad_norm": 3.1135254464036253,
+      "learning_rate": 8.820244256537402e-06,
+      "loss": 1.0545,
+      "step": 8700
+    },
+    {
+      "epoch": 0.26234939759036147,
+      "grad_norm": 3.647317125831831,
+      "learning_rate": 8.817029259742365e-06,
+      "loss": 1.0226,
+      "step": 8710
+    },
+    {
+      "epoch": 0.26265060240963856,
+      "grad_norm": 2.8942736143999666,
+      "learning_rate": 8.813810475890512e-06,
+      "loss": 0.887,
+      "step": 8720
+    },
+    {
+      "epoch": 0.26295180722891565,
+      "grad_norm": 3.3744023346923755,
+      "learning_rate": 8.810587908175354e-06,
+      "loss": 0.9674,
+      "step": 8730
+    },
+    {
+      "epoch": 0.2632530120481928,
+      "grad_norm": 3.605368502463631,
+      "learning_rate": 8.807361559794152e-06,
+      "loss": 0.9999,
+      "step": 8740
+    },
+    {
+      "epoch": 0.2635542168674699,
+      "grad_norm": 2.7144431832911438,
+      "learning_rate": 8.804131433947924e-06,
+      "loss": 0.9657,
+      "step": 8750
+    },
+    {
+      "epoch": 0.26385542168674697,
+      "grad_norm": 9.828698819825334,
+      "learning_rate": 8.80089753384143e-06,
+      "loss": 1.0306,
+      "step": 8760
+    },
+    {
+      "epoch": 0.2641566265060241,
+      "grad_norm": 3.5497797003483016,
+      "learning_rate": 8.797659862683177e-06,
+      "loss": 1.067,
+      "step": 8770
+    },
+    {
+      "epoch": 0.2644578313253012,
+      "grad_norm": 4.204949003562374,
+      "learning_rate": 8.794418423685409e-06,
+      "loss": 0.883,
+      "step": 8780
+    },
+    {
+      "epoch": 0.2647590361445783,
+      "grad_norm": 2.1131690002869035,
+      "learning_rate": 8.79117322006412e-06,
+      "loss": 0.9446,
+      "step": 8790
+    },
+    {
+      "epoch": 0.26506024096385544,
+      "grad_norm": 3.2653994029648103,
+      "learning_rate": 8.787924255039023e-06,
+      "loss": 1.0977,
+      "step": 8800
+    },
+    {
+      "epoch": 0.26536144578313253,
+      "grad_norm": 3.4847198927088034,
+      "learning_rate": 8.784671531833576e-06,
+      "loss": 1.0903,
+      "step": 8810
+    },
+    {
+      "epoch": 0.2656626506024096,
+      "grad_norm": 2.059740098020506,
+      "learning_rate": 8.781415053674961e-06,
+      "loss": 1.022,
+      "step": 8820
+    },
+    {
+      "epoch": 0.26596385542168677,
+      "grad_norm": 2.1593334065604086,
+      "learning_rate": 8.778154823794084e-06,
+      "loss": 0.9914,
+      "step": 8830
+    },
+    {
+      "epoch": 0.26626506024096386,
+      "grad_norm": 2.0773201565650523,
+      "learning_rate": 8.774890845425574e-06,
+      "loss": 1.075,
+      "step": 8840
+    },
+    {
+      "epoch": 0.26656626506024095,
+      "grad_norm": 2.936156340255884,
+      "learning_rate": 8.771623121807782e-06,
+      "loss": 0.9368,
+      "step": 8850
+    },
+    {
+      "epoch": 0.2668674698795181,
+      "grad_norm": 3.089892911981274,
+      "learning_rate": 8.76835165618277e-06,
+      "loss": 1.0487,
+      "step": 8860
+    },
+    {
+      "epoch": 0.2671686746987952,
+      "grad_norm": 2.246379326838362,
+      "learning_rate": 8.765076451796316e-06,
+      "loss": 0.9794,
+      "step": 8870
+    },
+    {
+      "epoch": 0.2674698795180723,
+      "grad_norm": 3.1692235943927582,
+      "learning_rate": 8.761797511897907e-06,
+      "loss": 1.0307,
+      "step": 8880
+    },
+    {
+      "epoch": 0.2677710843373494,
+      "grad_norm": 2.948244820205797,
+      "learning_rate": 8.758514839740735e-06,
+      "loss": 0.9876,
+      "step": 8890
+    },
+    {
+      "epoch": 0.2680722891566265,
+      "grad_norm": 3.4051597829304416,
+      "learning_rate": 8.755228438581697e-06,
+      "loss": 0.9459,
+      "step": 8900
+    },
+    {
+      "epoch": 0.2683734939759036,
+      "grad_norm": 2.2249274490023856,
+      "learning_rate": 8.751938311681386e-06,
+      "loss": 0.9304,
+      "step": 8910
+    },
+    {
+      "epoch": 0.26867469879518074,
+      "grad_norm": 5.608149002370423,
+      "learning_rate": 8.748644462304096e-06,
+      "loss": 1.0476,
+      "step": 8920
+    },
+    {
+      "epoch": 0.26897590361445783,
+      "grad_norm": 4.532022390089638,
+      "learning_rate": 8.74534689371781e-06,
+      "loss": 1.091,
+      "step": 8930
+    },
+    {
+      "epoch": 0.2692771084337349,
+      "grad_norm": 2.0657552554643455,
+      "learning_rate": 8.742045609194207e-06,
+      "loss": 1.0839,
+      "step": 8940
+    },
+    {
+      "epoch": 0.26957831325301207,
+      "grad_norm": 3.262085649651936,
+      "learning_rate": 8.738740612008644e-06,
+      "loss": 1.1189,
+      "step": 8950
+    },
+    {
+      "epoch": 0.26987951807228916,
+      "grad_norm": 3.0214716038681972,
+      "learning_rate": 8.735431905440168e-06,
+      "loss": 0.9694,
+      "step": 8960
+    },
+    {
+      "epoch": 0.27018072289156625,
+      "grad_norm": 2.8653234520295574,
+      "learning_rate": 8.732119492771506e-06,
+      "loss": 0.9334,
+      "step": 8970
+    },
+    {
+      "epoch": 0.2704819277108434,
+      "grad_norm": 3.31071343105019,
+      "learning_rate": 8.728803377289061e-06,
+      "loss": 1.0991,
+      "step": 8980
+    },
+    {
+      "epoch": 0.2707831325301205,
+      "grad_norm": 3.243179632332493,
+      "learning_rate": 8.725483562282906e-06,
+      "loss": 1.0835,
+      "step": 8990
+    },
+    {
+      "epoch": 0.2710843373493976,
+      "grad_norm": 3.761598483639871,
+      "learning_rate": 8.722160051046793e-06,
+      "loss": 0.9827,
+      "step": 9000
+    },
+    {
+      "epoch": 0.2713855421686747,
+      "grad_norm": 1.9217945655360718,
+      "learning_rate": 8.718832846878131e-06,
+      "loss": 0.9949,
+      "step": 9010
+    },
+    {
+      "epoch": 0.2716867469879518,
+      "grad_norm": 2.887794643400935,
+      "learning_rate": 8.715501953078e-06,
+      "loss": 1.01,
+      "step": 9020
+    },
+    {
+      "epoch": 0.2719879518072289,
+      "grad_norm": 2.850209938883259,
+      "learning_rate": 8.712167372951138e-06,
+      "loss": 1.0425,
+      "step": 9030
+    },
+    {
+      "epoch": 0.27228915662650605,
+      "grad_norm": 2.0174305923281577,
+      "learning_rate": 8.708829109805946e-06,
+      "loss": 0.9925,
+      "step": 9040
+    },
+    {
+      "epoch": 0.27259036144578314,
+      "grad_norm": 2.8949800495869997,
+      "learning_rate": 8.705487166954466e-06,
+      "loss": 1.1293,
+      "step": 9050
+    },
+    {
+      "epoch": 0.2728915662650602,
+      "grad_norm": 2.195829337105026,
+      "learning_rate": 8.702141547712401e-06,
+      "loss": 1.0801,
+      "step": 9060
+    },
+    {
+      "epoch": 0.27319277108433737,
+      "grad_norm": 3.0799602793477816,
+      "learning_rate": 8.698792255399104e-06,
+      "loss": 1.0336,
+      "step": 9070
+    },
+    {
+      "epoch": 0.27349397590361446,
+      "grad_norm": 4.01655154679231,
+      "learning_rate": 8.695439293337564e-06,
+      "loss": 0.9759,
+      "step": 9080
+    },
+    {
+      "epoch": 0.27379518072289155,
+      "grad_norm": 4.476127710746531,
+      "learning_rate": 8.692082664854417e-06,
+      "loss": 1.013,
+      "step": 9090
+    },
+    {
+      "epoch": 0.2740963855421687,
+      "grad_norm": 3.316655101014646,
+      "learning_rate": 8.688722373279933e-06,
+      "loss": 0.9287,
+      "step": 9100
+    },
+    {
+      "epoch": 0.2743975903614458,
+      "grad_norm": 2.2120599919290305,
+      "learning_rate": 8.685358421948014e-06,
+      "loss": 0.9453,
+      "step": 9110
+    },
+    {
+      "epoch": 0.2746987951807229,
+      "grad_norm": 2.114406706854489,
+      "learning_rate": 8.681990814196202e-06,
+      "loss": 0.9044,
+      "step": 9120
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 2.0479542650125167,
+      "learning_rate": 8.67861955336566e-06,
+      "loss": 1.0563,
+      "step": 9130
+    },
+    {
+      "epoch": 0.2753012048192771,
+      "grad_norm": 1.913614323403774,
+      "learning_rate": 8.675244642801172e-06,
+      "loss": 0.9772,
+      "step": 9140
+    },
+    {
+      "epoch": 0.2756024096385542,
+      "grad_norm": 1.974782501030137,
+      "learning_rate": 8.671866085851156e-06,
+      "loss": 0.9173,
+      "step": 9150
+    },
+    {
+      "epoch": 0.27590361445783135,
+      "grad_norm": 3.9898567048833673,
+      "learning_rate": 8.668483885867633e-06,
+      "loss": 0.9698,
+      "step": 9160
+    },
+    {
+      "epoch": 0.27620481927710844,
+      "grad_norm": 4.0756619027536045,
+      "learning_rate": 8.665098046206248e-06,
+      "loss": 0.9545,
+      "step": 9170
+    },
+    {
+      "epoch": 0.2765060240963855,
+      "grad_norm": 4.6695325960080565,
+      "learning_rate": 8.661708570226253e-06,
+      "loss": 1.0066,
+      "step": 9180
+    },
+    {
+      "epoch": 0.2768072289156627,
+      "grad_norm": 3.3586111674870347,
+      "learning_rate": 8.658315461290507e-06,
+      "loss": 0.9848,
+      "step": 9190
+    },
+    {
+      "epoch": 0.27710843373493976,
+      "grad_norm": 2.316396336595485,
+      "learning_rate": 8.654918722765479e-06,
+      "loss": 1.0356,
+      "step": 9200
+    },
+    {
+      "epoch": 0.27740963855421685,
+      "grad_norm": 3.6734082611914354,
+      "learning_rate": 8.651518358021234e-06,
+      "loss": 0.9983,
+      "step": 9210
+    },
+    {
+      "epoch": 0.277710843373494,
+      "grad_norm": 3.555579508427273,
+      "learning_rate": 8.648114370431436e-06,
+      "loss": 1.0605,
+      "step": 9220
+    },
+    {
+      "epoch": 0.2780120481927711,
+      "grad_norm": 3.0162969039502086,
+      "learning_rate": 8.644706763373341e-06,
+      "loss": 0.9616,
+      "step": 9230
+    },
+    {
+      "epoch": 0.2783132530120482,
+      "grad_norm": 3.1072813195691857,
+      "learning_rate": 8.641295540227801e-06,
+      "loss": 1.0275,
+      "step": 9240
+    },
+    {
+      "epoch": 0.2786144578313253,
+      "grad_norm": 3.152335395360339,
+      "learning_rate": 8.637880704379254e-06,
+      "loss": 1.0587,
+      "step": 9250
+    },
+    {
+      "epoch": 0.2789156626506024,
+      "grad_norm": 2.1320231425099596,
+      "learning_rate": 8.634462259215719e-06,
+      "loss": 0.9809,
+      "step": 9260
+    },
+    {
+      "epoch": 0.2792168674698795,
+      "grad_norm": 2.123634624614154,
+      "learning_rate": 8.631040208128801e-06,
+      "loss": 0.9439,
+      "step": 9270
+    },
+    {
+      "epoch": 0.27951807228915665,
+      "grad_norm": 3.7308636123001953,
+      "learning_rate": 8.627614554513678e-06,
+      "loss": 0.9259,
+      "step": 9280
+    },
+    {
+      "epoch": 0.27981927710843374,
+      "grad_norm": 2.209359435710404,
+      "learning_rate": 8.624185301769106e-06,
+      "loss": 1.0067,
+      "step": 9290
+    },
+    {
+      "epoch": 0.28012048192771083,
+      "grad_norm": 4.388990797817878,
+      "learning_rate": 8.62075245329741e-06,
+      "loss": 1.0468,
+      "step": 9300
+    },
+    {
+      "epoch": 0.280421686746988,
+      "grad_norm": 2.0315716687217247,
+      "learning_rate": 8.617316012504482e-06,
+      "loss": 0.9971,
+      "step": 9310
+    },
+    {
+      "epoch": 0.28072289156626506,
+      "grad_norm": 1.9933640334109042,
+      "learning_rate": 8.613875982799778e-06,
+      "loss": 0.9349,
+      "step": 9320
+    },
+    {
+      "epoch": 0.28102409638554215,
+      "grad_norm": 4.010703901407647,
+      "learning_rate": 8.610432367596318e-06,
+      "loss": 0.9986,
+      "step": 9330
+    },
+    {
+      "epoch": 0.2813253012048193,
+      "grad_norm": 2.0715506637193792,
+      "learning_rate": 8.606985170310673e-06,
+      "loss": 1.0251,
+      "step": 9340
+    },
+    {
+      "epoch": 0.2816265060240964,
+      "grad_norm": 3.521705097589883,
+      "learning_rate": 8.603534394362978e-06,
+      "loss": 1.0081,
+      "step": 9350
+    },
+    {
+      "epoch": 0.2819277108433735,
+      "grad_norm": 2.0308301769499537,
+      "learning_rate": 8.600080043176908e-06,
+      "loss": 0.9123,
+      "step": 9360
+    },
+    {
+      "epoch": 0.2822289156626506,
+      "grad_norm": 3.4120511236315316,
+      "learning_rate": 8.59662212017969e-06,
+      "loss": 1.0579,
+      "step": 9370
+    },
+    {
+      "epoch": 0.2825301204819277,
+      "grad_norm": 3.558088171590651,
+      "learning_rate": 8.593160628802093e-06,
+      "loss": 1.1084,
+      "step": 9380
+    },
+    {
+      "epoch": 0.2828313253012048,
+      "grad_norm": 5.2077730453201125,
+      "learning_rate": 8.589695572478432e-06,
+      "loss": 1.0583,
+      "step": 9390
+    },
+    {
+      "epoch": 0.28313253012048195,
+      "grad_norm": 3.829233159935352,
+      "learning_rate": 8.586226954646552e-06,
+      "loss": 0.9504,
+      "step": 9400
+    },
+    {
+      "epoch": 0.28343373493975904,
+      "grad_norm": 2.1213041714506318,
+      "learning_rate": 8.582754778747836e-06,
+      "loss": 1.0081,
+      "step": 9410
+    },
+    {
+      "epoch": 0.28373493975903613,
+      "grad_norm": 3.7570802554284968,
+      "learning_rate": 8.57927904822719e-06,
+      "loss": 1.0694,
+      "step": 9420
+    },
+    {
+      "epoch": 0.2840361445783133,
+      "grad_norm": 3.680727169065133,
+      "learning_rate": 8.57579976653306e-06,
+      "loss": 1.0114,
+      "step": 9430
+    },
+    {
+      "epoch": 0.28433734939759037,
+      "grad_norm": 2.125366849338149,
+      "learning_rate": 8.5723169371174e-06,
+      "loss": 0.9267,
+      "step": 9440
+    },
+    {
+      "epoch": 0.28463855421686746,
+      "grad_norm": 3.5634554268286824,
+      "learning_rate": 8.568830563435695e-06,
+      "loss": 1.0245,
+      "step": 9450
+    },
+    {
+      "epoch": 0.2849397590361446,
+      "grad_norm": 3.383819312317253,
+      "learning_rate": 8.565340648946942e-06,
+      "loss": 1.0031,
+      "step": 9460
+    },
+    {
+      "epoch": 0.2852409638554217,
+      "grad_norm": 5.018489228856675,
+      "learning_rate": 8.561847197113652e-06,
+      "loss": 1.0384,
+      "step": 9470
+    },
+    {
+      "epoch": 0.2855421686746988,
+      "grad_norm": 2.106375987821443,
+      "learning_rate": 8.558350211401847e-06,
+      "loss": 1.0437,
+      "step": 9480
+    },
+    {
+      "epoch": 0.2858433734939759,
+      "grad_norm": 2.051399084986709,
+      "learning_rate": 8.55484969528105e-06,
+      "loss": 0.8726,
+      "step": 9490
+    },
+    {
+      "epoch": 0.286144578313253,
+      "grad_norm": 5.571725247806883,
+      "learning_rate": 8.55134565222429e-06,
+      "loss": 1.0646,
+      "step": 9500
+    },
+    {
+      "epoch": 0.2864457831325301,
+      "grad_norm": 3.740064004961406,
+      "learning_rate": 8.5478380857081e-06,
+      "loss": 1.0561,
+      "step": 9510
+    },
+    {
+      "epoch": 0.28674698795180725,
+      "grad_norm": 2.170774936138277,
+      "learning_rate": 8.544326999212498e-06,
+      "loss": 0.945,
+      "step": 9520
+    },
+    {
+      "epoch": 0.28704819277108434,
+      "grad_norm": 3.5397619884652887,
+      "learning_rate": 8.540812396221004e-06,
+      "loss": 1.081,
+      "step": 9530
+    },
+    {
+      "epoch": 0.28734939759036143,
+      "grad_norm": 4.106927824872186,
+      "learning_rate": 8.53729428022062e-06,
+      "loss": 0.9765,
+      "step": 9540
+    },
+    {
+      "epoch": 0.2876506024096386,
+      "grad_norm": 2.1156632875564267,
+      "learning_rate": 8.533772654701841e-06,
+      "loss": 1.0164,
+      "step": 9550
+    },
+    {
+      "epoch": 0.28795180722891567,
+      "grad_norm": 2.22827357268054,
+      "learning_rate": 8.530247523158636e-06,
+      "loss": 0.918,
+      "step": 9560
+    },
+    {
+      "epoch": 0.28825301204819276,
+      "grad_norm": 4.056245021986871,
+      "learning_rate": 8.526718889088462e-06,
+      "loss": 0.8682,
+      "step": 9570
+    },
+    {
+      "epoch": 0.2885542168674699,
+      "grad_norm": 3.9995029052963624,
+      "learning_rate": 8.523186755992238e-06,
+      "loss": 1.0473,
+      "step": 9580
+    },
+    {
+      "epoch": 0.288855421686747,
+      "grad_norm": 3.9407262805197716,
+      "learning_rate": 8.519651127374365e-06,
+      "loss": 1.0374,
+      "step": 9590
+    },
+    {
+      "epoch": 0.2891566265060241,
+      "grad_norm": 3.5140684558085034,
+      "learning_rate": 8.516112006742708e-06,
+      "loss": 0.9466,
+      "step": 9600
+    },
+    {
+      "epoch": 0.28945783132530123,
+      "grad_norm": 4.126326268264751,
+      "learning_rate": 8.5125693976086e-06,
+      "loss": 1.1305,
+      "step": 9610
+    },
+    {
+      "epoch": 0.2897590361445783,
+      "grad_norm": 1.9779681956532646,
+      "learning_rate": 8.50902330348683e-06,
+      "loss": 0.8741,
+      "step": 9620
+    },
+    {
+      "epoch": 0.2900602409638554,
+      "grad_norm": 4.815731857352997,
+      "learning_rate": 8.505473727895645e-06,
+      "loss": 1.0245,
+      "step": 9630
+    },
+    {
+      "epoch": 0.29036144578313255,
+      "grad_norm": 4.386853244257249,
+      "learning_rate": 8.501920674356755e-06,
+      "loss": 0.9732,
+      "step": 9640
+    },
+    {
+      "epoch": 0.29066265060240964,
+      "grad_norm": 7.675980435686392,
+      "learning_rate": 8.498364146395308e-06,
+      "loss": 1.0894,
+      "step": 9650
+    },
+    {
+      "epoch": 0.29096385542168673,
+      "grad_norm": 7.4362537814990395,
+      "learning_rate": 8.494804147539908e-06,
+      "loss": 1.0496,
+      "step": 9660
+    },
+    {
+      "epoch": 0.2912650602409639,
+      "grad_norm": 2.1190568530914233,
+      "learning_rate": 8.491240681322597e-06,
+      "loss": 0.9939,
+      "step": 9670
+    },
+    {
+      "epoch": 0.29156626506024097,
+      "grad_norm": 4.504454100292339,
+      "learning_rate": 8.487673751278862e-06,
+      "loss": 0.9213,
+      "step": 9680
+    },
+    {
+      "epoch": 0.29186746987951806,
+      "grad_norm": 1.9990914564932507,
+      "learning_rate": 8.484103360947624e-06,
+      "loss": 0.8883,
+      "step": 9690
+    },
+    {
+      "epoch": 0.2921686746987952,
+      "grad_norm": 1.96311413820024,
+      "learning_rate": 8.480529513871237e-06,
+      "loss": 1.0131,
+      "step": 9700
+    },
+    {
+      "epoch": 0.2924698795180723,
+      "grad_norm": 1.9959365911392968,
+      "learning_rate": 8.476952213595486e-06,
+      "loss": 0.9971,
+      "step": 9710
+    },
+    {
+      "epoch": 0.2927710843373494,
+      "grad_norm": 4.261444775892076,
+      "learning_rate": 8.473371463669582e-06,
+      "loss": 1.091,
+      "step": 9720
+    },
+    {
+      "epoch": 0.29307228915662653,
+      "grad_norm": 4.291539127103698,
+      "learning_rate": 8.469787267646156e-06,
+      "loss": 1.0118,
+      "step": 9730
+    },
+    {
+      "epoch": 0.2933734939759036,
+      "grad_norm": 2.058077374442829,
+      "learning_rate": 8.466199629081259e-06,
+      "loss": 1.0093,
+      "step": 9740
+    },
+    {
+      "epoch": 0.2936746987951807,
+      "grad_norm": 7.278764807615133,
+      "learning_rate": 8.462608551534361e-06,
+      "loss": 1.0131,
+      "step": 9750
+    },
+    {
+      "epoch": 0.29397590361445786,
+      "grad_norm": 2.192933240945323,
+      "learning_rate": 8.45901403856834e-06,
+      "loss": 0.9013,
+      "step": 9760
+    },
+    {
+      "epoch": 0.29427710843373495,
+      "grad_norm": 4.517739760111507,
+      "learning_rate": 8.455416093749484e-06,
+      "loss": 1.0373,
+      "step": 9770
+    },
+    {
+      "epoch": 0.29457831325301204,
+      "grad_norm": 4.809282144037537,
+      "learning_rate": 8.451814720647489e-06,
+      "loss": 1.0661,
+      "step": 9780
+    },
+    {
+      "epoch": 0.2948795180722892,
+      "grad_norm": 2.220220596520115,
+      "learning_rate": 8.448209922835441e-06,
+      "loss": 0.9799,
+      "step": 9790
+    },
+    {
+      "epoch": 0.29518072289156627,
+      "grad_norm": 3.3384467708607146,
+      "learning_rate": 8.444601703889839e-06,
+      "loss": 1.0893,
+      "step": 9800
+    },
+    {
+      "epoch": 0.29548192771084336,
+      "grad_norm": 3.844421196291859,
+      "learning_rate": 8.440990067390567e-06,
+      "loss": 1.0661,
+      "step": 9810
+    },
+    {
+      "epoch": 0.2957831325301205,
+      "grad_norm": 5.606851916002022,
+      "learning_rate": 8.437375016920901e-06,
+      "loss": 0.9668,
+      "step": 9820
+    },
+    {
+      "epoch": 0.2960843373493976,
+      "grad_norm": 1.9419333318766125,
+      "learning_rate": 8.433756556067506e-06,
+      "loss": 0.9926,
+      "step": 9830
+    },
+    {
+      "epoch": 0.2963855421686747,
+      "grad_norm": 3.5390962538094497,
+      "learning_rate": 8.430134688420428e-06,
+      "loss": 1.0604,
+      "step": 9840
+    },
+    {
+      "epoch": 0.29668674698795183,
+      "grad_norm": 2.08481584778143,
+      "learning_rate": 8.426509417573099e-06,
+      "loss": 1.0084,
+      "step": 9850
+    },
+    {
+      "epoch": 0.2969879518072289,
+      "grad_norm": 5.1937764318739505,
+      "learning_rate": 8.422880747122317e-06,
+      "loss": 0.9945,
+      "step": 9860
+    },
+    {
+      "epoch": 0.297289156626506,
+      "grad_norm": 2.122275842944583,
+      "learning_rate": 8.419248680668263e-06,
+      "loss": 0.9689,
+      "step": 9870
+    },
+    {
+      "epoch": 0.29759036144578316,
+      "grad_norm": 5.668450624299819,
+      "learning_rate": 8.415613221814483e-06,
+      "loss": 0.9907,
+      "step": 9880
+    },
+    {
+      "epoch": 0.29789156626506025,
+      "grad_norm": 4.105107555390713,
+      "learning_rate": 8.411974374167891e-06,
+      "loss": 1.0722,
+      "step": 9890
+    },
+    {
+      "epoch": 0.29819277108433734,
+      "grad_norm": 4.038784462134157,
+      "learning_rate": 8.408332141338755e-06,
+      "loss": 0.9445,
+      "step": 9900
+    },
+    {
+      "epoch": 0.2984939759036145,
+      "grad_norm": 5.71876423304024,
+      "learning_rate": 8.404686526940715e-06,
+      "loss": 0.9854,
+      "step": 9910
+    },
+    {
+      "epoch": 0.2987951807228916,
+      "grad_norm": 7.343845231655897,
+      "learning_rate": 8.401037534590754e-06,
+      "loss": 1.0499,
+      "step": 9920
+    },
+    {
+      "epoch": 0.29909638554216866,
+      "grad_norm": 7.343094953013136,
+      "learning_rate": 8.397385167909214e-06,
+      "loss": 0.9208,
+      "step": 9930
+    },
+    {
+      "epoch": 0.2993975903614458,
+      "grad_norm": 2.0254793586076727,
+      "learning_rate": 8.393729430519781e-06,
+      "loss": 1.038,
+      "step": 9940
+    },
+    {
+      "epoch": 0.2996987951807229,
+      "grad_norm": 2.0875635754460737,
+      "learning_rate": 8.39007032604949e-06,
+      "loss": 0.8775,
+      "step": 9950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.629135068376105,
+      "learning_rate": 8.386407858128707e-06,
+      "loss": 1.0253,
+      "step": 9960
+    },
+    {
+      "epoch": 0.30030120481927713,
+      "grad_norm": 9.143103512430542,
+      "learning_rate": 8.382742030391144e-06,
+      "loss": 1.1477,
+      "step": 9970
+    },
+    {
+      "epoch": 0.3006024096385542,
+      "grad_norm": 2.1611988930239674,
+      "learning_rate": 8.379072846473845e-06,
+      "loss": 0.8932,
+      "step": 9980
+    },
+    {
+      "epoch": 0.3009036144578313,
+      "grad_norm": 2.1482285938635153,
+      "learning_rate": 8.37540031001718e-06,
+      "loss": 0.989,
+      "step": 9990
+    },
+    {
+      "epoch": 0.30120481927710846,
+      "grad_norm": 3.8113924866377142,
+      "learning_rate": 8.371724424664851e-06,
+      "loss": 1.0152,
+      "step": 10000
+    },
+    {
+      "epoch": 0.30150602409638555,
+      "grad_norm": 4.327424585719604,
+      "learning_rate": 8.368045194063876e-06,
+      "loss": 1.0561,
+      "step": 10010
+    },
+    {
+      "epoch": 0.30180722891566264,
+      "grad_norm": 2.10279887823453,
+      "learning_rate": 8.364362621864595e-06,
+      "loss": 0.9944,
+      "step": 10020
+    },
+    {
+      "epoch": 0.3021084337349398,
+      "grad_norm": 2.037386564834821,
+      "learning_rate": 8.360676711720666e-06,
+      "loss": 0.9977,
+      "step": 10030
+    },
+    {
+      "epoch": 0.3024096385542169,
+      "grad_norm": 2.2070496569550406,
+      "learning_rate": 8.356987467289055e-06,
+      "loss": 1.0271,
+      "step": 10040
+    },
+    {
+      "epoch": 0.30271084337349397,
+      "grad_norm": 2.1707092869165097,
+      "learning_rate": 8.353294892230038e-06,
+      "loss": 0.8483,
+      "step": 10050
+    },
+    {
+      "epoch": 0.3030120481927711,
+      "grad_norm": 3.7454069955763925,
+      "learning_rate": 8.349598990207192e-06,
+      "loss": 0.9188,
+      "step": 10060
+    },
+    {
+      "epoch": 0.3033132530120482,
+      "grad_norm": 3.2981642962310995,
+      "learning_rate": 8.345899764887402e-06,
+      "loss": 0.9694,
+      "step": 10070
+    },
+    {
+      "epoch": 0.3036144578313253,
+      "grad_norm": 3.213099280209126,
+      "learning_rate": 8.342197219940844e-06,
+      "loss": 1.1052,
+      "step": 10080
+    },
+    {
+      "epoch": 0.30391566265060244,
+      "grad_norm": 3.2864018134132014,
+      "learning_rate": 8.338491359040988e-06,
+      "loss": 1.0089,
+      "step": 10090
+    },
+    {
+      "epoch": 0.3042168674698795,
+      "grad_norm": 2.0954484427271014,
+      "learning_rate": 8.334782185864598e-06,
+      "loss": 1.0294,
+      "step": 10100
+    },
+    {
+      "epoch": 0.3045180722891566,
+      "grad_norm": 3.303121187870742,
+      "learning_rate": 8.331069704091722e-06,
+      "loss": 1.0381,
+      "step": 10110
+    },
+    {
+      "epoch": 0.30481927710843376,
+      "grad_norm": 3.490692244009057,
+      "learning_rate": 8.327353917405685e-06,
+      "loss": 0.9893,
+      "step": 10120
+    },
+    {
+      "epoch": 0.30512048192771085,
+      "grad_norm": 4.715938395729553,
+      "learning_rate": 8.323634829493101e-06,
+      "loss": 0.9515,
+      "step": 10130
+    },
+    {
+      "epoch": 0.30542168674698794,
+      "grad_norm": 4.041702284257836,
+      "learning_rate": 8.319912444043856e-06,
+      "loss": 1.0521,
+      "step": 10140
+    },
+    {
+      "epoch": 0.3057228915662651,
+      "grad_norm": 2.1308567518259305,
+      "learning_rate": 8.316186764751101e-06,
+      "loss": 0.9391,
+      "step": 10150
+    },
+    {
+      "epoch": 0.3060240963855422,
+      "grad_norm": 6.398420670580796,
+      "learning_rate": 8.312457795311263e-06,
+      "loss": 1.0661,
+      "step": 10160
+    },
+    {
+      "epoch": 0.30632530120481927,
+      "grad_norm": 6.049901202327995,
+      "learning_rate": 8.308725539424032e-06,
+      "loss": 0.8529,
+      "step": 10170
+    },
+    {
+      "epoch": 0.3066265060240964,
+      "grad_norm": 2.1836920028458158,
+      "learning_rate": 8.304990000792354e-06,
+      "loss": 1.0632,
+      "step": 10180
+    },
+    {
+      "epoch": 0.3069277108433735,
+      "grad_norm": 4.315187827797635,
+      "learning_rate": 8.301251183122437e-06,
+      "loss": 1.1366,
+      "step": 10190
+    },
+    {
+      "epoch": 0.3072289156626506,
+      "grad_norm": 3.8805494967232863,
+      "learning_rate": 8.297509090123739e-06,
+      "loss": 1.0369,
+      "step": 10200
+    },
+    {
+      "epoch": 0.30753012048192774,
+      "grad_norm": 4.9133585436285605,
+      "learning_rate": 8.29376372550897e-06,
+      "loss": 1.1049,
+      "step": 10210
+    },
+    {
+      "epoch": 0.30783132530120483,
+      "grad_norm": 3.019274722378204,
+      "learning_rate": 8.290015092994083e-06,
+      "loss": 1.0532,
+      "step": 10220
+    },
+    {
+      "epoch": 0.3081325301204819,
+      "grad_norm": 2.176444809228001,
+      "learning_rate": 8.28626319629828e-06,
+      "loss": 1.0581,
+      "step": 10230
+    },
+    {
+      "epoch": 0.30843373493975906,
+      "grad_norm": 4.955783849926732,
+      "learning_rate": 8.282508039143992e-06,
+      "loss": 1.0501,
+      "step": 10240
+    },
+    {
+      "epoch": 0.30873493975903615,
+      "grad_norm": 4.108722335550171,
+      "learning_rate": 8.27874962525689e-06,
+      "loss": 0.9323,
+      "step": 10250
+    },
+    {
+      "epoch": 0.30903614457831324,
+      "grad_norm": 3.3746658696795206,
+      "learning_rate": 8.27498795836588e-06,
+      "loss": 0.9167,
+      "step": 10260
+    },
+    {
+      "epoch": 0.3093373493975904,
+      "grad_norm": 2.2109002059876834,
+      "learning_rate": 8.271223042203086e-06,
+      "loss": 0.8523,
+      "step": 10270
+    },
+    {
+      "epoch": 0.3096385542168675,
+      "grad_norm": 1.9938075662060806,
+      "learning_rate": 8.267454880503862e-06,
+      "loss": 0.962,
+      "step": 10280
+    },
+    {
+      "epoch": 0.30993975903614457,
+      "grad_norm": 4.895058685699006,
+      "learning_rate": 8.26368347700678e-06,
+      "loss": 1.0655,
+      "step": 10290
+    },
+    {
+      "epoch": 0.3102409638554217,
+      "grad_norm": 4.683607734527515,
+      "learning_rate": 8.259908835453635e-06,
+      "loss": 1.0902,
+      "step": 10300
+    },
+    {
+      "epoch": 0.3105421686746988,
+      "grad_norm": 4.551887962361471,
+      "learning_rate": 8.256130959589424e-06,
+      "loss": 1.0172,
+      "step": 10310
+    },
+    {
+      "epoch": 0.3108433734939759,
+      "grad_norm": 6.323504816735507,
+      "learning_rate": 8.252349853162358e-06,
+      "loss": 0.9511,
+      "step": 10320
+    },
+    {
+      "epoch": 0.31114457831325304,
+      "grad_norm": 2.2244199490191026,
+      "learning_rate": 8.248565519923855e-06,
+      "loss": 0.9199,
+      "step": 10330
+    },
+    {
+      "epoch": 0.31144578313253013,
+      "grad_norm": 3.4041532317646626,
+      "learning_rate": 8.244777963628532e-06,
+      "loss": 0.9874,
+      "step": 10340
+    },
+    {
+      "epoch": 0.3117469879518072,
+      "grad_norm": 2.169175654323326,
+      "learning_rate": 8.240987188034202e-06,
+      "loss": 1.0389,
+      "step": 10350
+    },
+    {
+      "epoch": 0.31204819277108437,
+      "grad_norm": 2.173393683751808,
+      "learning_rate": 8.237193196901876e-06,
+      "loss": 0.9878,
+      "step": 10360
+    },
+    {
+      "epoch": 0.31234939759036146,
+      "grad_norm": 1.8960616181589895,
+      "learning_rate": 8.233395993995757e-06,
+      "loss": 0.8596,
+      "step": 10370
+    },
+    {
+      "epoch": 0.31265060240963854,
+      "grad_norm": 4.881679487431638,
+      "learning_rate": 8.229595583083225e-06,
+      "loss": 0.9993,
+      "step": 10380
+    },
+    {
+      "epoch": 0.31295180722891563,
+      "grad_norm": 4.0622475220613365,
+      "learning_rate": 8.225791967934853e-06,
+      "loss": 1.0734,
+      "step": 10390
+    },
+    {
+      "epoch": 0.3132530120481928,
+      "grad_norm": 3.463187950518435,
+      "learning_rate": 8.221985152324385e-06,
+      "loss": 0.9813,
+      "step": 10400
+    },
+    {
+      "epoch": 0.31355421686746987,
+      "grad_norm": 3.414810584626675,
+      "learning_rate": 8.21817514002875e-06,
+      "loss": 1.0017,
+      "step": 10410
+    },
+    {
+      "epoch": 0.31385542168674696,
+      "grad_norm": 1.946505225650699,
+      "learning_rate": 8.21436193482804e-06,
+      "loss": 0.8486,
+      "step": 10420
+    },
+    {
+      "epoch": 0.3141566265060241,
+      "grad_norm": 4.058793686487053,
+      "learning_rate": 8.210545540505519e-06,
+      "loss": 1.0272,
+      "step": 10430
+    },
+    {
+      "epoch": 0.3144578313253012,
+      "grad_norm": 4.221866633099206,
+      "learning_rate": 8.20672596084761e-06,
+      "loss": 1.0307,
+      "step": 10440
+    },
+    {
+      "epoch": 0.3147590361445783,
+      "grad_norm": 4.250681700188214,
+      "learning_rate": 8.202903199643903e-06,
+      "loss": 0.9789,
+      "step": 10450
+    },
+    {
+      "epoch": 0.31506024096385543,
+      "grad_norm": 4.088554887712706,
+      "learning_rate": 8.199077260687145e-06,
+      "loss": 1.1202,
+      "step": 10460
+    },
+    {
+      "epoch": 0.3153614457831325,
+      "grad_norm": 4.336915914235672,
+      "learning_rate": 8.195248147773228e-06,
+      "loss": 0.9602,
+      "step": 10470
+    },
+    {
+      "epoch": 0.3156626506024096,
+      "grad_norm": 2.086010663424,
+      "learning_rate": 8.191415864701198e-06,
+      "loss": 0.9813,
+      "step": 10480
+    },
+    {
+      "epoch": 0.31596385542168676,
+      "grad_norm": 4.094892650793342,
+      "learning_rate": 8.187580415273247e-06,
+      "loss": 0.9845,
+      "step": 10490
+    },
+    {
+      "epoch": 0.31626506024096385,
+      "grad_norm": 5.0111439904767,
+      "learning_rate": 8.183741803294707e-06,
+      "loss": 1.0929,
+      "step": 10500
+    },
+    {
+      "epoch": 0.31656626506024094,
+      "grad_norm": 2.1263418445804843,
+      "learning_rate": 8.179900032574049e-06,
+      "loss": 0.9088,
+      "step": 10510
+    },
+    {
+      "epoch": 0.3168674698795181,
+      "grad_norm": 8.187585461551897,
+      "learning_rate": 8.176055106922873e-06,
+      "loss": 1.0677,
+      "step": 10520
+    },
+    {
+      "epoch": 0.31716867469879517,
+      "grad_norm": 2.2516961374446334,
+      "learning_rate": 8.172207030155922e-06,
+      "loss": 0.9015,
+      "step": 10530
+    },
+    {
+      "epoch": 0.31746987951807226,
+      "grad_norm": 4.516664383443458,
+      "learning_rate": 8.168355806091045e-06,
+      "loss": 1.0555,
+      "step": 10540
+    },
+    {
+      "epoch": 0.3177710843373494,
+      "grad_norm": 6.12958491028471,
+      "learning_rate": 8.164501438549234e-06,
+      "loss": 0.9436,
+      "step": 10550
+    },
+    {
+      "epoch": 0.3180722891566265,
+      "grad_norm": 4.2743258425990485,
+      "learning_rate": 8.160643931354586e-06,
+      "loss": 1.0102,
+      "step": 10560
+    },
+    {
+      "epoch": 0.3183734939759036,
+      "grad_norm": 2.100559407650895,
+      "learning_rate": 8.156783288334323e-06,
+      "loss": 0.9872,
+      "step": 10570
+    },
+    {
+      "epoch": 0.31867469879518073,
+      "grad_norm": 4.482796122449825,
+      "learning_rate": 8.152919513318766e-06,
+      "loss": 1.0186,
+      "step": 10580
+    },
+    {
+      "epoch": 0.3189759036144578,
+      "grad_norm": 4.864044671249022,
+      "learning_rate": 8.149052610141357e-06,
+      "loss": 0.9591,
+      "step": 10590
+    },
+    {
+      "epoch": 0.3192771084337349,
+      "grad_norm": 1.899935810099827,
+      "learning_rate": 8.145182582638632e-06,
+      "loss": 0.9434,
+      "step": 10600
+    },
+    {
+      "epoch": 0.31957831325301206,
+      "grad_norm": 5.821848522611604,
+      "learning_rate": 8.141309434650228e-06,
+      "loss": 1.0552,
+      "step": 10610
+    },
+    {
+      "epoch": 0.31987951807228915,
+      "grad_norm": 6.2963056165551805,
+      "learning_rate": 8.137433170018885e-06,
+      "loss": 1.02,
+      "step": 10620
+    },
+    {
+      "epoch": 0.32018072289156624,
+      "grad_norm": 5.508301247500946,
+      "learning_rate": 8.133553792590427e-06,
+      "loss": 1.076,
+      "step": 10630
+    },
+    {
+      "epoch": 0.3204819277108434,
+      "grad_norm": 1.885883373268772,
+      "learning_rate": 8.12967130621377e-06,
+      "loss": 0.9259,
+      "step": 10640
+    },
+    {
+      "epoch": 0.3207831325301205,
+      "grad_norm": 2.1958518354758754,
+      "learning_rate": 8.125785714740912e-06,
+      "loss": 1.0185,
+      "step": 10650
+    },
+    {
+      "epoch": 0.32108433734939756,
+      "grad_norm": 2.5813327610862618,
+      "learning_rate": 8.121897022026935e-06,
+      "loss": 1.0261,
+      "step": 10660
+    },
+    {
+      "epoch": 0.3213855421686747,
+      "grad_norm": 12.632784671316534,
+      "learning_rate": 8.118005231929999e-06,
+      "loss": 0.9618,
+      "step": 10670
+    },
+    {
+      "epoch": 0.3216867469879518,
+      "grad_norm": 33.71508467064113,
+      "learning_rate": 8.11411034831133e-06,
+      "loss": 1.0002,
+      "step": 10680
+    },
+    {
+      "epoch": 0.3219879518072289,
+      "grad_norm": 2.2510074427985955,
+      "learning_rate": 8.110212375035234e-06,
+      "loss": 1.0596,
+      "step": 10690
+    },
+    {
+      "epoch": 0.32228915662650603,
+      "grad_norm": 12.161236885052919,
+      "learning_rate": 8.106311315969069e-06,
+      "loss": 1.0568,
+      "step": 10700
+    },
+    {
+      "epoch": 0.3225903614457831,
+      "grad_norm": 5.371988165263253,
+      "learning_rate": 8.102407174983266e-06,
+      "loss": 0.9959,
+      "step": 10710
+    },
+    {
+      "epoch": 0.3228915662650602,
+      "grad_norm": 4.483323978123669,
+      "learning_rate": 8.09849995595131e-06,
+      "loss": 1.1077,
+      "step": 10720
+    },
+    {
+      "epoch": 0.32319277108433736,
+      "grad_norm": 6.201584399794071,
+      "learning_rate": 8.094589662749737e-06,
+      "loss": 1.0778,
+      "step": 10730
+    },
+    {
+      "epoch": 0.32349397590361445,
+      "grad_norm": 7.385309985399349,
+      "learning_rate": 8.090676299258138e-06,
+      "loss": 0.9272,
+      "step": 10740
+    },
+    {
+      "epoch": 0.32379518072289154,
+      "grad_norm": 5.064938597035172,
+      "learning_rate": 8.086759869359145e-06,
+      "loss": 1.0434,
+      "step": 10750
+    },
+    {
+      "epoch": 0.3240963855421687,
+      "grad_norm": 2.061147108501892,
+      "learning_rate": 8.082840376938436e-06,
+      "loss": 0.9781,
+      "step": 10760
+    },
+    {
+      "epoch": 0.3243975903614458,
+      "grad_norm": 5.756402561330784,
+      "learning_rate": 8.078917825884728e-06,
+      "loss": 0.9888,
+      "step": 10770
+    },
+    {
+      "epoch": 0.32469879518072287,
+      "grad_norm": 4.540292910517579,
+      "learning_rate": 8.07499222008977e-06,
+      "loss": 1.0483,
+      "step": 10780
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 5.000002960734428,
+      "learning_rate": 8.071063563448341e-06,
+      "loss": 1.0091,
+      "step": 10790
+    },
+    {
+      "epoch": 0.3253012048192771,
+      "grad_norm": 2.042460734798289,
+      "learning_rate": 8.06713185985825e-06,
+      "loss": 0.8824,
+      "step": 10800
+    },
+    {
+      "epoch": 0.3256024096385542,
+      "grad_norm": 2.0853705185850457,
+      "learning_rate": 8.063197113220327e-06,
+      "loss": 0.9518,
+      "step": 10810
+    },
+    {
+      "epoch": 0.32590361445783134,
+      "grad_norm": 2.1195196765705737,
+      "learning_rate": 8.059259327438425e-06,
+      "loss": 1.0,
+      "step": 10820
+    },
+    {
+      "epoch": 0.3262048192771084,
+      "grad_norm": 5.880477193857654,
+      "learning_rate": 8.055318506419403e-06,
+      "loss": 1.0192,
+      "step": 10830
+    },
+    {
+      "epoch": 0.3265060240963855,
+      "grad_norm": 1.9781978177363087,
+      "learning_rate": 8.051374654073142e-06,
+      "loss": 0.9261,
+      "step": 10840
+    },
+    {
+      "epoch": 0.32680722891566266,
+      "grad_norm": 2.0116533404843024,
+      "learning_rate": 8.047427774312528e-06,
+      "loss": 0.9988,
+      "step": 10850
+    },
+    {
+      "epoch": 0.32710843373493975,
+      "grad_norm": 4.829421479149449,
+      "learning_rate": 8.043477871053445e-06,
+      "loss": 0.955,
+      "step": 10860
+    },
+    {
+      "epoch": 0.32740963855421684,
+      "grad_norm": 6.453665494173318,
+      "learning_rate": 8.03952494821478e-06,
+      "loss": 0.9507,
+      "step": 10870
+    },
+    {
+      "epoch": 0.327710843373494,
+      "grad_norm": 3.8480244595337014,
+      "learning_rate": 8.03556900971842e-06,
+      "loss": 1.0797,
+      "step": 10880
+    },
+    {
+      "epoch": 0.3280120481927711,
+      "grad_norm": 4.564222918930517,
+      "learning_rate": 8.031610059489237e-06,
+      "loss": 0.9727,
+      "step": 10890
+    },
+    {
+      "epoch": 0.32831325301204817,
+      "grad_norm": 2.3236021716733184,
+      "learning_rate": 8.027648101455095e-06,
+      "loss": 1.0288,
+      "step": 10900
+    },
+    {
+      "epoch": 0.3286144578313253,
+      "grad_norm": 4.169807812805349,
+      "learning_rate": 8.023683139546841e-06,
+      "loss": 1.0955,
+      "step": 10910
+    },
+    {
+      "epoch": 0.3289156626506024,
+      "grad_norm": 2.003645238689451,
+      "learning_rate": 8.019715177698305e-06,
+      "loss": 1.0459,
+      "step": 10920
+    },
+    {
+      "epoch": 0.3292168674698795,
+      "grad_norm": 2.179700787095523,
+      "learning_rate": 8.015744219846289e-06,
+      "loss": 0.8858,
+      "step": 10930
+    },
+    {
+      "epoch": 0.32951807228915664,
+      "grad_norm": 4.444688268717414,
+      "learning_rate": 8.011770269930572e-06,
+      "loss": 0.9368,
+      "step": 10940
+    },
+    {
+      "epoch": 0.32981927710843373,
+      "grad_norm": 8.984812974122574,
+      "learning_rate": 8.007793331893898e-06,
+      "loss": 1.0194,
+      "step": 10950
+    },
+    {
+      "epoch": 0.3301204819277108,
+      "grad_norm": 1.9534289762852177,
+      "learning_rate": 8.003813409681978e-06,
+      "loss": 0.8108,
+      "step": 10960
+    },
+    {
+      "epoch": 0.33042168674698796,
+      "grad_norm": 5.295910267661215,
+      "learning_rate": 7.999830507243478e-06,
+      "loss": 0.8803,
+      "step": 10970
+    },
+    {
+      "epoch": 0.33072289156626505,
+      "grad_norm": 2.123748614518604,
+      "learning_rate": 7.995844628530034e-06,
+      "loss": 0.9602,
+      "step": 10980
+    },
+    {
+      "epoch": 0.33102409638554214,
+      "grad_norm": 12.124457803955242,
+      "learning_rate": 7.991855777496219e-06,
+      "loss": 1.0699,
+      "step": 10990
+    },
+    {
+      "epoch": 0.3313253012048193,
+      "grad_norm": 7.682101489526108,
+      "learning_rate": 7.987863958099564e-06,
+      "loss": 1.0692,
+      "step": 11000
+    },
+    {
+      "epoch": 0.3316265060240964,
+      "grad_norm": 7.738741702593449,
+      "learning_rate": 7.983869174300544e-06,
+      "loss": 1.0043,
+      "step": 11010
+    },
+    {
+      "epoch": 0.33192771084337347,
+      "grad_norm": 6.092864158051944,
+      "learning_rate": 7.979871430062577e-06,
+      "loss": 1.0381,
+      "step": 11020
+    },
+    {
+      "epoch": 0.3322289156626506,
+      "grad_norm": 4.176544518300791,
+      "learning_rate": 7.975870729352013e-06,
+      "loss": 1.0028,
+      "step": 11030
+    },
+    {
+      "epoch": 0.3325301204819277,
+      "grad_norm": 4.871432857202954,
+      "learning_rate": 7.971867076138139e-06,
+      "loss": 0.9203,
+      "step": 11040
+    },
+    {
+      "epoch": 0.3328313253012048,
+      "grad_norm": 4.1801337367747635,
+      "learning_rate": 7.96786047439317e-06,
+      "loss": 1.0804,
+      "step": 11050
+    },
+    {
+      "epoch": 0.33313253012048194,
+      "grad_norm": 11.014746998172521,
+      "learning_rate": 7.963850928092247e-06,
+      "loss": 1.0475,
+      "step": 11060
+    },
+    {
+      "epoch": 0.33343373493975903,
+      "grad_norm": 4.525575667256446,
+      "learning_rate": 7.959838441213432e-06,
+      "loss": 0.992,
+      "step": 11070
+    },
+    {
+      "epoch": 0.3337349397590361,
+      "grad_norm": 4.410829555085609,
+      "learning_rate": 7.955823017737706e-06,
+      "loss": 1.0486,
+      "step": 11080
+    },
+    {
+      "epoch": 0.33403614457831327,
+      "grad_norm": 3.4309869480769413,
+      "learning_rate": 7.951804661648958e-06,
+      "loss": 0.9851,
+      "step": 11090
+    },
+    {
+      "epoch": 0.33433734939759036,
+      "grad_norm": 7.16301588450031,
+      "learning_rate": 7.947783376933995e-06,
+      "loss": 0.9675,
+      "step": 11100
+    },
+    {
+      "epoch": 0.33463855421686745,
+      "grad_norm": 1.7946658805824416,
+      "learning_rate": 7.943759167582524e-06,
+      "loss": 0.8823,
+      "step": 11110
+    },
+    {
+      "epoch": 0.3349397590361446,
+      "grad_norm": 5.01038703406234,
+      "learning_rate": 7.939732037587155e-06,
+      "loss": 1.0359,
+      "step": 11120
+    },
+    {
+      "epoch": 0.3352409638554217,
+      "grad_norm": 4.131627029526289,
+      "learning_rate": 7.935701990943395e-06,
+      "loss": 0.9927,
+      "step": 11130
+    },
+    {
+      "epoch": 0.33554216867469877,
+      "grad_norm": 3.362435113496702,
+      "learning_rate": 7.931669031649648e-06,
+      "loss": 1.102,
+      "step": 11140
+    },
+    {
+      "epoch": 0.3358433734939759,
+      "grad_norm": 2.0904828288683888,
+      "learning_rate": 7.927633163707202e-06,
+      "loss": 0.9637,
+      "step": 11150
+    },
+    {
+      "epoch": 0.336144578313253,
+      "grad_norm": 1.9167001134841364,
+      "learning_rate": 7.923594391120237e-06,
+      "loss": 1.0139,
+      "step": 11160
+    },
+    {
+      "epoch": 0.3364457831325301,
+      "grad_norm": 3.8166751157117487,
+      "learning_rate": 7.919552717895808e-06,
+      "loss": 1.1167,
+      "step": 11170
+    },
+    {
+      "epoch": 0.33674698795180724,
+      "grad_norm": 1.8530000899009555,
+      "learning_rate": 7.915508148043857e-06,
+      "loss": 0.998,
+      "step": 11180
+    },
+    {
+      "epoch": 0.33704819277108433,
+      "grad_norm": 5.089422994663483,
+      "learning_rate": 7.911460685577193e-06,
+      "loss": 1.0284,
+      "step": 11190
+    },
+    {
+      "epoch": 0.3373493975903614,
+      "grad_norm": 5.340756030145489,
+      "learning_rate": 7.907410334511494e-06,
+      "loss": 0.9679,
+      "step": 11200
+    },
+    {
+      "epoch": 0.33765060240963857,
+      "grad_norm": 4.2778241487895645,
+      "learning_rate": 7.90335709886531e-06,
+      "loss": 1.0368,
+      "step": 11210
+    },
+    {
+      "epoch": 0.33795180722891566,
+      "grad_norm": 1.7482343695076692,
+      "learning_rate": 7.899300982660048e-06,
+      "loss": 0.9494,
+      "step": 11220
+    },
+    {
+      "epoch": 0.33825301204819275,
+      "grad_norm": 3.5537222269878637,
+      "learning_rate": 7.895241989919971e-06,
+      "loss": 1.0937,
+      "step": 11230
+    },
+    {
+      "epoch": 0.3385542168674699,
+      "grad_norm": 4.70721515455698,
+      "learning_rate": 7.891180124672205e-06,
+      "loss": 0.9367,
+      "step": 11240
+    },
+    {
+      "epoch": 0.338855421686747,
+      "grad_norm": 4.040149729274763,
+      "learning_rate": 7.887115390946716e-06,
+      "loss": 0.975,
+      "step": 11250
+    },
+    {
+      "epoch": 0.3391566265060241,
+      "grad_norm": 4.191861338628631,
+      "learning_rate": 7.883047792776324e-06,
+      "loss": 1.0594,
+      "step": 11260
+    },
+    {
+      "epoch": 0.3394578313253012,
+      "grad_norm": 5.162182052647102,
+      "learning_rate": 7.878977334196684e-06,
+      "loss": 1.0286,
+      "step": 11270
+    },
+    {
+      "epoch": 0.3397590361445783,
+      "grad_norm": 3.785017852331962,
+      "learning_rate": 7.874904019246294e-06,
+      "loss": 0.9398,
+      "step": 11280
+    },
+    {
+      "epoch": 0.3400602409638554,
+      "grad_norm": 4.492677914363335,
+      "learning_rate": 7.870827851966482e-06,
+      "loss": 1.0652,
+      "step": 11290
+    },
+    {
+      "epoch": 0.34036144578313254,
+      "grad_norm": 3.6401117799449745,
+      "learning_rate": 7.86674883640141e-06,
+      "loss": 0.9585,
+      "step": 11300
+    },
+    {
+      "epoch": 0.34066265060240963,
+      "grad_norm": 3.913573078107745,
+      "learning_rate": 7.862666976598063e-06,
+      "loss": 1.0016,
+      "step": 11310
+    },
+    {
+      "epoch": 0.3409638554216867,
+      "grad_norm": 3.6471231467900225,
+      "learning_rate": 7.85858227660625e-06,
+      "loss": 1.1002,
+      "step": 11320
+    },
+    {
+      "epoch": 0.34126506024096387,
+      "grad_norm": 2.0450868462730947,
+      "learning_rate": 7.854494740478596e-06,
+      "loss": 0.9459,
+      "step": 11330
+    },
+    {
+      "epoch": 0.34156626506024096,
+      "grad_norm": 5.270128040029897,
+      "learning_rate": 7.85040437227054e-06,
+      "loss": 0.8647,
+      "step": 11340
+    },
+    {
+      "epoch": 0.34186746987951805,
+      "grad_norm": 2.7680072621035454,
+      "learning_rate": 7.846311176040331e-06,
+      "loss": 1.0631,
+      "step": 11350
+    },
+    {
+      "epoch": 0.3421686746987952,
+      "grad_norm": 3.672272147127404,
+      "learning_rate": 7.842215155849024e-06,
+      "loss": 1.0353,
+      "step": 11360
+    },
+    {
+      "epoch": 0.3424698795180723,
+      "grad_norm": 7.606629385178783,
+      "learning_rate": 7.83811631576048e-06,
+      "loss": 0.9552,
+      "step": 11370
+    },
+    {
+      "epoch": 0.3427710843373494,
+      "grad_norm": 3.535222873623303,
+      "learning_rate": 7.834014659841349e-06,
+      "loss": 1.0624,
+      "step": 11380
+    },
+    {
+      "epoch": 0.3430722891566265,
+      "grad_norm": 2.0884459939916273,
+      "learning_rate": 7.829910192161083e-06,
+      "loss": 0.9438,
+      "step": 11390
+    },
+    {
+      "epoch": 0.3433734939759036,
+      "grad_norm": 4.195423300204089,
+      "learning_rate": 7.825802916791917e-06,
+      "loss": 1.0971,
+      "step": 11400
+    },
+    {
+      "epoch": 0.3436746987951807,
+      "grad_norm": 1.766631150711695,
+      "learning_rate": 7.821692837808877e-06,
+      "loss": 0.915,
+      "step": 11410
+    },
+    {
+      "epoch": 0.34397590361445785,
+      "grad_norm": 1.9538141950054282,
+      "learning_rate": 7.817579959289768e-06,
+      "loss": 0.8541,
+      "step": 11420
+    },
+    {
+      "epoch": 0.34427710843373494,
+      "grad_norm": 4.066235068672363,
+      "learning_rate": 7.813464285315171e-06,
+      "loss": 0.9872,
+      "step": 11430
+    },
+    {
+      "epoch": 0.344578313253012,
+      "grad_norm": 3.5828437974576963,
+      "learning_rate": 7.809345819968449e-06,
+      "loss": 1.1013,
+      "step": 11440
+    },
+    {
+      "epoch": 0.34487951807228917,
+      "grad_norm": 1.8901758058197726,
+      "learning_rate": 7.805224567335719e-06,
+      "loss": 0.9521,
+      "step": 11450
+    },
+    {
+      "epoch": 0.34518072289156626,
+      "grad_norm": 2.072625529214965,
+      "learning_rate": 7.801100531505877e-06,
+      "loss": 0.9442,
+      "step": 11460
+    },
+    {
+      "epoch": 0.34548192771084335,
+      "grad_norm": 2.0061827751587136,
+      "learning_rate": 7.796973716570576e-06,
+      "loss": 1.0048,
+      "step": 11470
+    },
+    {
+      "epoch": 0.3457831325301205,
+      "grad_norm": 5.362888952636561,
+      "learning_rate": 7.792844126624226e-06,
+      "loss": 0.9931,
+      "step": 11480
+    },
+    {
+      "epoch": 0.3460843373493976,
+      "grad_norm": 4.733275942529403,
+      "learning_rate": 7.788711765763987e-06,
+      "loss": 0.8837,
+      "step": 11490
+    },
+    {
+      "epoch": 0.3463855421686747,
+      "grad_norm": 5.216588717287876,
+      "learning_rate": 7.784576638089773e-06,
+      "loss": 1.0845,
+      "step": 11500
+    },
+    {
+      "epoch": 0.3466867469879518,
+      "grad_norm": 5.221184937904206,
+      "learning_rate": 7.780438747704246e-06,
+      "loss": 0.9466,
+      "step": 11510
+    },
+    {
+      "epoch": 0.3469879518072289,
+      "grad_norm": 4.503741386955517,
+      "learning_rate": 7.776298098712798e-06,
+      "loss": 1.0021,
+      "step": 11520
+    },
+    {
+      "epoch": 0.347289156626506,
+      "grad_norm": 1.8923775713854136,
+      "learning_rate": 7.772154695223568e-06,
+      "loss": 0.8952,
+      "step": 11530
+    },
+    {
+      "epoch": 0.34759036144578315,
+      "grad_norm": 5.6942666030655005,
+      "learning_rate": 7.768008541347423e-06,
+      "loss": 0.9808,
+      "step": 11540
+    },
+    {
+      "epoch": 0.34789156626506024,
+      "grad_norm": 4.077644152619464,
+      "learning_rate": 7.76385964119796e-06,
+      "loss": 0.8846,
+      "step": 11550
+    },
+    {
+      "epoch": 0.3481927710843373,
+      "grad_norm": 2.09267718532834,
+      "learning_rate": 7.759707998891501e-06,
+      "loss": 0.8713,
+      "step": 11560
+    },
+    {
+      "epoch": 0.3484939759036145,
+      "grad_norm": 10.34967959913251,
+      "learning_rate": 7.75555361854709e-06,
+      "loss": 0.9726,
+      "step": 11570
+    },
+    {
+      "epoch": 0.34879518072289156,
+      "grad_norm": 5.35403644832154,
+      "learning_rate": 7.751396504286482e-06,
+      "loss": 0.9894,
+      "step": 11580
+    },
+    {
+      "epoch": 0.34909638554216865,
+      "grad_norm": 1.8889949713805092,
+      "learning_rate": 7.747236660234152e-06,
+      "loss": 0.9279,
+      "step": 11590
+    },
+    {
+      "epoch": 0.3493975903614458,
+      "grad_norm": 3.626748629240793,
+      "learning_rate": 7.743074090517277e-06,
+      "loss": 0.9579,
+      "step": 11600
+    },
+    {
+      "epoch": 0.3496987951807229,
+      "grad_norm": 2.2815070727483175,
+      "learning_rate": 7.738908799265744e-06,
+      "loss": 1.0396,
+      "step": 11610
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9994521502929223,
+      "learning_rate": 7.734740790612137e-06,
+      "loss": 0.8925,
+      "step": 11620
+    },
+    {
+      "epoch": 0.3503012048192771,
+      "grad_norm": 3.6622482362392077,
+      "learning_rate": 7.73057006869173e-06,
+      "loss": 1.1075,
+      "step": 11630
+    },
+    {
+      "epoch": 0.3506024096385542,
+      "grad_norm": 2.069341044213193,
+      "learning_rate": 7.726396637642503e-06,
+      "loss": 0.8838,
+      "step": 11640
+    },
+    {
+      "epoch": 0.3509036144578313,
+      "grad_norm": 1.8822289172236946,
+      "learning_rate": 7.722220501605114e-06,
+      "loss": 0.9637,
+      "step": 11650
+    },
+    {
+      "epoch": 0.35120481927710845,
+      "grad_norm": 2.0548049156522814,
+      "learning_rate": 7.718041664722904e-06,
+      "loss": 0.8996,
+      "step": 11660
+    },
+    {
+      "epoch": 0.35150602409638554,
+      "grad_norm": 3.1647212555037894,
+      "learning_rate": 7.713860131141898e-06,
+      "loss": 0.9324,
+      "step": 11670
+    },
+    {
+      "epoch": 0.35180722891566263,
+      "grad_norm": 2.0307667265535407,
+      "learning_rate": 7.709675905010796e-06,
+      "loss": 0.8733,
+      "step": 11680
+    },
+    {
+      "epoch": 0.3521084337349398,
+      "grad_norm": 3.3587937751270207,
+      "learning_rate": 7.705488990480967e-06,
+      "loss": 1.0552,
+      "step": 11690
+    },
+    {
+      "epoch": 0.35240963855421686,
+      "grad_norm": 3.6672200860743716,
+      "learning_rate": 7.701299391706449e-06,
+      "loss": 0.9027,
+      "step": 11700
+    },
+    {
+      "epoch": 0.35271084337349395,
+      "grad_norm": 2.055809584161785,
+      "learning_rate": 7.697107112843944e-06,
+      "loss": 0.9677,
+      "step": 11710
+    },
+    {
+      "epoch": 0.3530120481927711,
+      "grad_norm": 5.681151115408283,
+      "learning_rate": 7.692912158052807e-06,
+      "loss": 1.1032,
+      "step": 11720
+    },
+    {
+      "epoch": 0.3533132530120482,
+      "grad_norm": 1.9187476703936253,
+      "learning_rate": 7.688714531495061e-06,
+      "loss": 0.9076,
+      "step": 11730
+    },
+    {
+      "epoch": 0.3536144578313253,
+      "grad_norm": 1.993897678107153,
+      "learning_rate": 7.684514237335365e-06,
+      "loss": 0.9445,
+      "step": 11740
+    },
+    {
+      "epoch": 0.3539156626506024,
+      "grad_norm": 3.8781429686090974,
+      "learning_rate": 7.680311279741033e-06,
+      "loss": 0.962,
+      "step": 11750
+    },
+    {
+      "epoch": 0.3542168674698795,
+      "grad_norm": 4.598950729399993,
+      "learning_rate": 7.676105662882023e-06,
+      "loss": 0.9035,
+      "step": 11760
+    },
+    {
+      "epoch": 0.3545180722891566,
+      "grad_norm": 4.469712832430609,
+      "learning_rate": 7.67189739093092e-06,
+      "loss": 0.9874,
+      "step": 11770
+    },
+    {
+      "epoch": 0.35481927710843375,
+      "grad_norm": 2.0608164849855215,
+      "learning_rate": 7.66768646806296e-06,
+      "loss": 0.9224,
+      "step": 11780
+    },
+    {
+      "epoch": 0.35512048192771084,
+      "grad_norm": 2.2705628156118784,
+      "learning_rate": 7.663472898455996e-06,
+      "loss": 1.0113,
+      "step": 11790
+    },
+    {
+      "epoch": 0.35542168674698793,
+      "grad_norm": 6.425444929730967,
+      "learning_rate": 7.659256686290513e-06,
+      "loss": 1.0422,
+      "step": 11800
+    },
+    {
+      "epoch": 0.3557228915662651,
+      "grad_norm": 2.0760865511603757,
+      "learning_rate": 7.655037835749615e-06,
+      "loss": 0.9742,
+      "step": 11810
+    },
+    {
+      "epoch": 0.35602409638554217,
+      "grad_norm": 3.222186139567789,
+      "learning_rate": 7.650816351019024e-06,
+      "loss": 1.0131,
+      "step": 11820
+    },
+    {
+      "epoch": 0.35632530120481926,
+      "grad_norm": 5.848832526960766,
+      "learning_rate": 7.646592236287078e-06,
+      "loss": 1.0364,
+      "step": 11830
+    },
+    {
+      "epoch": 0.3566265060240964,
+      "grad_norm": 4.3173339914073585,
+      "learning_rate": 7.642365495744724e-06,
+      "loss": 0.9846,
+      "step": 11840
+    },
+    {
+      "epoch": 0.3569277108433735,
+      "grad_norm": 1.9894810652672887,
+      "learning_rate": 7.638136133585511e-06,
+      "loss": 1.0054,
+      "step": 11850
+    },
+    {
+      "epoch": 0.3572289156626506,
+      "grad_norm": 2.490300414940136,
+      "learning_rate": 7.633904154005592e-06,
+      "loss": 0.9597,
+      "step": 11860
+    },
+    {
+      "epoch": 0.3575301204819277,
+      "grad_norm": 1.9732180468952747,
+      "learning_rate": 7.629669561203715e-06,
+      "loss": 0.9413,
+      "step": 11870
+    },
+    {
+      "epoch": 0.3578313253012048,
+      "grad_norm": 4.519632081109088,
+      "learning_rate": 7.62543235938122e-06,
+      "loss": 1.0379,
+      "step": 11880
+    },
+    {
+      "epoch": 0.3581325301204819,
+      "grad_norm": 7.541543856491166,
+      "learning_rate": 7.621192552742041e-06,
+      "loss": 0.9894,
+      "step": 11890
+    },
+    {
+      "epoch": 0.35843373493975905,
+      "grad_norm": 2.1839219380733037,
+      "learning_rate": 7.616950145492688e-06,
+      "loss": 0.9952,
+      "step": 11900
+    },
+    {
+      "epoch": 0.35873493975903614,
+      "grad_norm": 9.463629219698094,
+      "learning_rate": 7.612705141842258e-06,
+      "loss": 0.9865,
+      "step": 11910
+    },
+    {
+      "epoch": 0.35903614457831323,
+      "grad_norm": 7.0291374448550314,
+      "learning_rate": 7.608457546002423e-06,
+      "loss": 1.0408,
+      "step": 11920
+    },
+    {
+      "epoch": 0.3593373493975904,
+      "grad_norm": 15.025469769876697,
+      "learning_rate": 7.604207362187423e-06,
+      "loss": 1.0129,
+      "step": 11930
+    },
+    {
+      "epoch": 0.35963855421686747,
+      "grad_norm": 8.063578405889139,
+      "learning_rate": 7.5999545946140696e-06,
+      "loss": 1.0775,
+      "step": 11940
+    },
+    {
+      "epoch": 0.35993975903614456,
+      "grad_norm": 6.312545184922915,
+      "learning_rate": 7.595699247501735e-06,
+      "loss": 0.9485,
+      "step": 11950
+    },
+    {
+      "epoch": 0.3602409638554217,
+      "grad_norm": 5.036637979316993,
+      "learning_rate": 7.591441325072355e-06,
+      "loss": 0.9692,
+      "step": 11960
+    },
+    {
+      "epoch": 0.3605421686746988,
+      "grad_norm": 5.121623705611691,
+      "learning_rate": 7.5871808315504134e-06,
+      "loss": 0.9424,
+      "step": 11970
+    },
+    {
+      "epoch": 0.3608433734939759,
+      "grad_norm": 4.708061740133258,
+      "learning_rate": 7.5829177711629506e-06,
+      "loss": 0.9425,
+      "step": 11980
+    },
+    {
+      "epoch": 0.36114457831325303,
+      "grad_norm": 4.403096358482696,
+      "learning_rate": 7.578652148139553e-06,
+      "loss": 0.9203,
+      "step": 11990
+    },
+    {
+      "epoch": 0.3614457831325301,
+      "grad_norm": 4.495083693777487,
+      "learning_rate": 7.574383966712348e-06,
+      "loss": 1.028,
+      "step": 12000
+    },
+    {
+      "epoch": 0.3617469879518072,
+      "grad_norm": 3.7829270017291474,
+      "learning_rate": 7.570113231116005e-06,
+      "loss": 1.041,
+      "step": 12010
+    },
+    {
+      "epoch": 0.36204819277108435,
+      "grad_norm": 4.045104830847792,
+      "learning_rate": 7.56583994558772e-06,
+      "loss": 1.0033,
+      "step": 12020
+    },
+    {
+      "epoch": 0.36234939759036144,
+      "grad_norm": 5.584787634293018,
+      "learning_rate": 7.561564114367226e-06,
+      "loss": 0.9923,
+      "step": 12030
+    },
+    {
+      "epoch": 0.36265060240963853,
+      "grad_norm": 3.878966372313198,
+      "learning_rate": 7.557285741696777e-06,
+      "loss": 1.1086,
+      "step": 12040
+    },
+    {
+      "epoch": 0.3629518072289157,
+      "grad_norm": 3.2923754142678288,
+      "learning_rate": 7.553004831821153e-06,
+      "loss": 0.9805,
+      "step": 12050
+    },
+    {
+      "epoch": 0.36325301204819277,
+      "grad_norm": 3.4821449007447116,
+      "learning_rate": 7.5487213889876475e-06,
+      "loss": 1.0026,
+      "step": 12060
+    },
+    {
+      "epoch": 0.36355421686746986,
+      "grad_norm": 3.4609487186400054,
+      "learning_rate": 7.544435417446068e-06,
+      "loss": 0.7465,
+      "step": 12070
+    },
+    {
+      "epoch": 0.363855421686747,
+      "grad_norm": 3.3833798650840685,
+      "learning_rate": 7.540146921448735e-06,
+      "loss": 0.9664,
+      "step": 12080
+    },
+    {
+      "epoch": 0.3641566265060241,
+      "grad_norm": 1.984224008796806,
+      "learning_rate": 7.535855905250464e-06,
+      "loss": 0.9476,
+      "step": 12090
+    },
+    {
+      "epoch": 0.3644578313253012,
+      "grad_norm": 4.128395853880073,
+      "learning_rate": 7.531562373108583e-06,
+      "loss": 0.9898,
+      "step": 12100
+    },
+    {
+      "epoch": 0.36475903614457833,
+      "grad_norm": 3.523795842940874,
+      "learning_rate": 7.527266329282905e-06,
+      "loss": 0.9142,
+      "step": 12110
+    },
+    {
+      "epoch": 0.3650602409638554,
+      "grad_norm": 1.998039113548706,
+      "learning_rate": 7.5229677780357435e-06,
+      "loss": 1.0372,
+      "step": 12120
+    },
+    {
+      "epoch": 0.3653614457831325,
+      "grad_norm": 3.7965229995688725,
+      "learning_rate": 7.5186667236318936e-06,
+      "loss": 0.8416,
+      "step": 12130
+    },
+    {
+      "epoch": 0.36566265060240966,
+      "grad_norm": 3.481348140636192,
+      "learning_rate": 7.514363170338639e-06,
+      "loss": 1.0853,
+      "step": 12140
+    },
+    {
+      "epoch": 0.36596385542168675,
+      "grad_norm": 2.9042009507627604,
+      "learning_rate": 7.510057122425737e-06,
+      "loss": 0.9719,
+      "step": 12150
+    },
+    {
+      "epoch": 0.36626506024096384,
+      "grad_norm": 5.666203153319428,
+      "learning_rate": 7.505748584165426e-06,
+      "loss": 0.9615,
+      "step": 12160
+    },
+    {
+      "epoch": 0.366566265060241,
+      "grad_norm": 2.0590145327402185,
+      "learning_rate": 7.501437559832413e-06,
+      "loss": 0.9807,
+      "step": 12170
+    },
+    {
+      "epoch": 0.36686746987951807,
+      "grad_norm": 2.077984305278314,
+      "learning_rate": 7.497124053703868e-06,
+      "loss": 0.9644,
+      "step": 12180
+    },
+    {
+      "epoch": 0.36716867469879516,
+      "grad_norm": 3.243407402314189,
+      "learning_rate": 7.492808070059428e-06,
+      "loss": 0.9872,
+      "step": 12190
+    },
+    {
+      "epoch": 0.3674698795180723,
+      "grad_norm": 3.755668320775745,
+      "learning_rate": 7.488489613181186e-06,
+      "loss": 1.0369,
+      "step": 12200
+    },
+    {
+      "epoch": 0.3677710843373494,
+      "grad_norm": 3.0652712536883095,
+      "learning_rate": 7.48416868735369e-06,
+      "loss": 0.9233,
+      "step": 12210
+    },
+    {
+      "epoch": 0.3680722891566265,
+      "grad_norm": 3.4261592542583794,
+      "learning_rate": 7.479845296863934e-06,
+      "loss": 1.1334,
+      "step": 12220
+    },
+    {
+      "epoch": 0.36837349397590363,
+      "grad_norm": 5.704405789481728,
+      "learning_rate": 7.4755194460013615e-06,
+      "loss": 1.0497,
+      "step": 12230
+    },
+    {
+      "epoch": 0.3686746987951807,
+      "grad_norm": 3.1187485791328675,
+      "learning_rate": 7.4711911390578565e-06,
+      "loss": 1.0918,
+      "step": 12240
+    },
+    {
+      "epoch": 0.3689759036144578,
+      "grad_norm": 1.9253129895742902,
+      "learning_rate": 7.466860380327738e-06,
+      "loss": 1.0079,
+      "step": 12250
+    },
+    {
+      "epoch": 0.36927710843373496,
+      "grad_norm": 4.269061577068835,
+      "learning_rate": 7.462527174107757e-06,
+      "loss": 1.0405,
+      "step": 12260
+    },
+    {
+      "epoch": 0.36957831325301205,
+      "grad_norm": 3.8701615824129516,
+      "learning_rate": 7.458191524697095e-06,
+      "loss": 0.9703,
+      "step": 12270
+    },
+    {
+      "epoch": 0.36987951807228914,
+      "grad_norm": 3.3658365668570505,
+      "learning_rate": 7.453853436397358e-06,
+      "loss": 0.971,
+      "step": 12280
+    },
+    {
+      "epoch": 0.3701807228915663,
+      "grad_norm": 2.1062448579153044,
+      "learning_rate": 7.449512913512569e-06,
+      "loss": 0.9751,
+      "step": 12290
+    },
+    {
+      "epoch": 0.3704819277108434,
+      "grad_norm": 2.0748820681020748,
+      "learning_rate": 7.445169960349167e-06,
+      "loss": 1.0067,
+      "step": 12300
+    },
+    {
+      "epoch": 0.37078313253012046,
+      "grad_norm": 3.859794676968757,
+      "learning_rate": 7.440824581216005e-06,
+      "loss": 0.9573,
+      "step": 12310
+    },
+    {
+      "epoch": 0.3710843373493976,
+      "grad_norm": 3.466358057065326,
+      "learning_rate": 7.436476780424339e-06,
+      "loss": 0.9606,
+      "step": 12320
+    },
+    {
+      "epoch": 0.3713855421686747,
+      "grad_norm": 2.1083525878028824,
+      "learning_rate": 7.432126562287833e-06,
+      "loss": 0.8662,
+      "step": 12330
+    },
+    {
+      "epoch": 0.3716867469879518,
+      "grad_norm": 3.797578592911251,
+      "learning_rate": 7.427773931122541e-06,
+      "loss": 1.0214,
+      "step": 12340
+    },
+    {
+      "epoch": 0.37198795180722893,
+      "grad_norm": 3.6345295936608104,
+      "learning_rate": 7.4234188912469185e-06,
+      "loss": 1.0191,
+      "step": 12350
+    },
+    {
+      "epoch": 0.372289156626506,
+      "grad_norm": 3.74131550894578,
+      "learning_rate": 7.419061446981809e-06,
+      "loss": 0.9515,
+      "step": 12360
+    },
+    {
+      "epoch": 0.3725903614457831,
+      "grad_norm": 4.015919698762564,
+      "learning_rate": 7.41470160265044e-06,
+      "loss": 0.9016,
+      "step": 12370
+    },
+    {
+      "epoch": 0.37289156626506026,
+      "grad_norm": 1.8975136407931068,
+      "learning_rate": 7.410339362578422e-06,
+      "loss": 0.9815,
+      "step": 12380
+    },
+    {
+      "epoch": 0.37319277108433735,
+      "grad_norm": 2.027588364097527,
+      "learning_rate": 7.40597473109374e-06,
+      "loss": 1.0008,
+      "step": 12390
+    },
+    {
+      "epoch": 0.37349397590361444,
+      "grad_norm": 4.210884740114786,
+      "learning_rate": 7.4016077125267524e-06,
+      "loss": 1.0322,
+      "step": 12400
+    },
+    {
+      "epoch": 0.3737951807228916,
+      "grad_norm": 1.900718721757264,
+      "learning_rate": 7.397238311210189e-06,
+      "loss": 1.0023,
+      "step": 12410
+    },
+    {
+      "epoch": 0.3740963855421687,
+      "grad_norm": 4.013336922349029,
+      "learning_rate": 7.392866531479142e-06,
+      "loss": 1.0003,
+      "step": 12420
+    },
+    {
+      "epoch": 0.37439759036144576,
+      "grad_norm": 2.1955865782421173,
+      "learning_rate": 7.38849237767106e-06,
+      "loss": 0.9563,
+      "step": 12430
+    },
+    {
+      "epoch": 0.3746987951807229,
+      "grad_norm": 2.0046519737550335,
+      "learning_rate": 7.384115854125748e-06,
+      "loss": 0.9102,
+      "step": 12440
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 3.9016136604452996,
+      "learning_rate": 7.379736965185369e-06,
+      "loss": 0.9413,
+      "step": 12450
+    },
+    {
+      "epoch": 0.3753012048192771,
+      "grad_norm": 3.2483313959778872,
+      "learning_rate": 7.375355715194423e-06,
+      "loss": 1.0819,
+      "step": 12460
+    },
+    {
+      "epoch": 0.37560240963855424,
+      "grad_norm": 3.204785265080706,
+      "learning_rate": 7.37097210849976e-06,
+      "loss": 1.0078,
+      "step": 12470
+    },
+    {
+      "epoch": 0.3759036144578313,
+      "grad_norm": 3.3390830579513957,
+      "learning_rate": 7.366586149450565e-06,
+      "loss": 1.0894,
+      "step": 12480
+    },
+    {
+      "epoch": 0.3762048192771084,
+      "grad_norm": 3.2214701749722265,
+      "learning_rate": 7.362197842398355e-06,
+      "loss": 0.9376,
+      "step": 12490
+    },
+    {
+      "epoch": 0.37650602409638556,
+      "grad_norm": 4.187817334567737,
+      "learning_rate": 7.3578071916969815e-06,
+      "loss": 1.0763,
+      "step": 12500
+    },
+    {
+      "epoch": 0.37680722891566265,
+      "grad_norm": 1.9445817326027879,
+      "learning_rate": 7.353414201702618e-06,
+      "loss": 0.8968,
+      "step": 12510
+    },
+    {
+      "epoch": 0.37710843373493974,
+      "grad_norm": 3.3085143469005405,
+      "learning_rate": 7.349018876773757e-06,
+      "loss": 1.0455,
+      "step": 12520
+    },
+    {
+      "epoch": 0.3774096385542169,
+      "grad_norm": 1.8688818785972638,
+      "learning_rate": 7.344621221271213e-06,
+      "loss": 0.958,
+      "step": 12530
+    },
+    {
+      "epoch": 0.377710843373494,
+      "grad_norm": 3.2987717394592617,
+      "learning_rate": 7.340221239558108e-06,
+      "loss": 0.9878,
+      "step": 12540
+    },
+    {
+      "epoch": 0.37801204819277107,
+      "grad_norm": 3.082266542193966,
+      "learning_rate": 7.335818935999874e-06,
+      "loss": 1.0674,
+      "step": 12550
+    },
+    {
+      "epoch": 0.3783132530120482,
+      "grad_norm": 3.2534423955558274,
+      "learning_rate": 7.331414314964248e-06,
+      "loss": 1.1027,
+      "step": 12560
+    },
+    {
+      "epoch": 0.3786144578313253,
+      "grad_norm": 3.43243228305782,
+      "learning_rate": 7.3270073808212615e-06,
+      "loss": 0.958,
+      "step": 12570
+    },
+    {
+      "epoch": 0.3789156626506024,
+      "grad_norm": 2.832033942525752,
+      "learning_rate": 7.322598137943249e-06,
+      "loss": 0.9465,
+      "step": 12580
+    },
+    {
+      "epoch": 0.37921686746987954,
+      "grad_norm": 1.937079304513958,
+      "learning_rate": 7.318186590704824e-06,
+      "loss": 0.9161,
+      "step": 12590
+    },
+    {
+      "epoch": 0.3795180722891566,
+      "grad_norm": 3.182373336700091,
+      "learning_rate": 7.3137727434828995e-06,
+      "loss": 1.0651,
+      "step": 12600
+    },
+    {
+      "epoch": 0.3798192771084337,
+      "grad_norm": 2.1070515774968333,
+      "learning_rate": 7.309356600656659e-06,
+      "loss": 0.998,
+      "step": 12610
+    },
+    {
+      "epoch": 0.38012048192771086,
+      "grad_norm": 2.9732502478244887,
+      "learning_rate": 7.3049381666075715e-06,
+      "loss": 0.9811,
+      "step": 12620
+    },
+    {
+      "epoch": 0.38042168674698795,
+      "grad_norm": 3.514134853004432,
+      "learning_rate": 7.300517445719375e-06,
+      "loss": 0.9795,
+      "step": 12630
+    },
+    {
+      "epoch": 0.38072289156626504,
+      "grad_norm": 3.107817337048043,
+      "learning_rate": 7.296094442378078e-06,
+      "loss": 0.9908,
+      "step": 12640
+    },
+    {
+      "epoch": 0.3810240963855422,
+      "grad_norm": 3.4623836273364055,
+      "learning_rate": 7.2916691609719525e-06,
+      "loss": 0.947,
+      "step": 12650
+    },
+    {
+      "epoch": 0.3813253012048193,
+      "grad_norm": 3.086429428881401,
+      "learning_rate": 7.2872416058915305e-06,
+      "loss": 0.9981,
+      "step": 12660
+    },
+    {
+      "epoch": 0.38162650602409637,
+      "grad_norm": 3.8489801730440503,
+      "learning_rate": 7.2828117815296025e-06,
+      "loss": 0.9235,
+      "step": 12670
+    },
+    {
+      "epoch": 0.3819277108433735,
+      "grad_norm": 3.4102823216410205,
+      "learning_rate": 7.278379692281209e-06,
+      "loss": 0.9588,
+      "step": 12680
+    },
+    {
+      "epoch": 0.3822289156626506,
+      "grad_norm": 2.949334287853416,
+      "learning_rate": 7.2739453425436345e-06,
+      "loss": 0.917,
+      "step": 12690
+    },
+    {
+      "epoch": 0.3825301204819277,
+      "grad_norm": 3.1229075873163845,
+      "learning_rate": 7.269508736716411e-06,
+      "loss": 1.0317,
+      "step": 12700
+    },
+    {
+      "epoch": 0.38283132530120484,
+      "grad_norm": 3.351920270635086,
+      "learning_rate": 7.2650698792013065e-06,
+      "loss": 1.0268,
+      "step": 12710
+    },
+    {
+      "epoch": 0.38313253012048193,
+      "grad_norm": 2.929317217211371,
+      "learning_rate": 7.260628774402322e-06,
+      "loss": 1.0022,
+      "step": 12720
+    },
+    {
+      "epoch": 0.383433734939759,
+      "grad_norm": 3.613800636085388,
+      "learning_rate": 7.256185426725691e-06,
+      "loss": 1.0318,
+      "step": 12730
+    },
+    {
+      "epoch": 0.38373493975903616,
+      "grad_norm": 3.2014664228395064,
+      "learning_rate": 7.25173984057987e-06,
+      "loss": 1.0667,
+      "step": 12740
+    },
+    {
+      "epoch": 0.38403614457831325,
+      "grad_norm": 1.9295027200797479,
+      "learning_rate": 7.247292020375537e-06,
+      "loss": 0.9882,
+      "step": 12750
+    },
+    {
+      "epoch": 0.38433734939759034,
+      "grad_norm": 3.2515691317224134,
+      "learning_rate": 7.242841970525587e-06,
+      "loss": 0.9574,
+      "step": 12760
+    },
+    {
+      "epoch": 0.3846385542168675,
+      "grad_norm": 2.0144366573518604,
+      "learning_rate": 7.238389695445127e-06,
+      "loss": 1.0104,
+      "step": 12770
+    },
+    {
+      "epoch": 0.3849397590361446,
+      "grad_norm": 1.9938680368796105,
+      "learning_rate": 7.23393519955147e-06,
+      "loss": 0.8391,
+      "step": 12780
+    },
+    {
+      "epoch": 0.38524096385542167,
+      "grad_norm": 3.0479896541727065,
+      "learning_rate": 7.229478487264134e-06,
+      "loss": 1.0009,
+      "step": 12790
+    },
+    {
+      "epoch": 0.3855421686746988,
+      "grad_norm": 7.5561014652032235,
+      "learning_rate": 7.225019563004839e-06,
+      "loss": 0.914,
+      "step": 12800
+    },
+    {
+      "epoch": 0.3858433734939759,
+      "grad_norm": 3.177648156582394,
+      "learning_rate": 7.220558431197493e-06,
+      "loss": 0.9036,
+      "step": 12810
+    },
+    {
+      "epoch": 0.386144578313253,
+      "grad_norm": 3.15907668264104,
+      "learning_rate": 7.216095096268199e-06,
+      "loss": 0.9943,
+      "step": 12820
+    },
+    {
+      "epoch": 0.38644578313253014,
+      "grad_norm": 2.9532276163876383,
+      "learning_rate": 7.2116295626452434e-06,
+      "loss": 1.0642,
+      "step": 12830
+    },
+    {
+      "epoch": 0.38674698795180723,
+      "grad_norm": 3.0071648243174125,
+      "learning_rate": 7.207161834759097e-06,
+      "loss": 0.9418,
+      "step": 12840
+    },
+    {
+      "epoch": 0.3870481927710843,
+      "grad_norm": 2.9061077102466104,
+      "learning_rate": 7.202691917042406e-06,
+      "loss": 1.0247,
+      "step": 12850
+    },
+    {
+      "epoch": 0.38734939759036147,
+      "grad_norm": 4.312299457555282,
+      "learning_rate": 7.198219813929985e-06,
+      "loss": 1.142,
+      "step": 12860
+    },
+    {
+      "epoch": 0.38765060240963856,
+      "grad_norm": 3.9434226113424566,
+      "learning_rate": 7.193745529858827e-06,
+      "loss": 1.0536,
+      "step": 12870
+    },
+    {
+      "epoch": 0.38795180722891565,
+      "grad_norm": 3.345993349978265,
+      "learning_rate": 7.189269069268077e-06,
+      "loss": 0.9771,
+      "step": 12880
+    },
+    {
+      "epoch": 0.3882530120481928,
+      "grad_norm": 3.1833847360222634,
+      "learning_rate": 7.184790436599047e-06,
+      "loss": 0.9317,
+      "step": 12890
+    },
+    {
+      "epoch": 0.3885542168674699,
+      "grad_norm": 1.9698557559954264,
+      "learning_rate": 7.1803096362952066e-06,
+      "loss": 0.9526,
+      "step": 12900
+    },
+    {
+      "epoch": 0.38885542168674697,
+      "grad_norm": 4.984760388967978,
+      "learning_rate": 7.175826672802163e-06,
+      "loss": 1.0319,
+      "step": 12910
+    },
+    {
+      "epoch": 0.3891566265060241,
+      "grad_norm": 3.751405052016957,
+      "learning_rate": 7.171341550567685e-06,
+      "loss": 1.0487,
+      "step": 12920
+    },
+    {
+      "epoch": 0.3894578313253012,
+      "grad_norm": 3.7328836878826506,
+      "learning_rate": 7.166854274041672e-06,
+      "loss": 0.9844,
+      "step": 12930
+    },
+    {
+      "epoch": 0.3897590361445783,
+      "grad_norm": 3.3695214026996143,
+      "learning_rate": 7.162364847676169e-06,
+      "loss": 0.9598,
+      "step": 12940
+    },
+    {
+      "epoch": 0.39006024096385544,
+      "grad_norm": 2.142318438489866,
+      "learning_rate": 7.157873275925347e-06,
+      "loss": 0.8418,
+      "step": 12950
+    },
+    {
+      "epoch": 0.39036144578313253,
+      "grad_norm": 1.8798550671654182,
+      "learning_rate": 7.1533795632455105e-06,
+      "loss": 0.9877,
+      "step": 12960
+    },
+    {
+      "epoch": 0.3906626506024096,
+      "grad_norm": 1.9725118116370421,
+      "learning_rate": 7.1488837140950865e-06,
+      "loss": 0.9163,
+      "step": 12970
+    },
+    {
+      "epoch": 0.39096385542168677,
+      "grad_norm": 4.642643795325702,
+      "learning_rate": 7.144385732934618e-06,
+      "loss": 1.037,
+      "step": 12980
+    },
+    {
+      "epoch": 0.39126506024096386,
+      "grad_norm": 3.4401784850680848,
+      "learning_rate": 7.13988562422677e-06,
+      "loss": 1.0857,
+      "step": 12990
+    },
+    {
+      "epoch": 0.39156626506024095,
+      "grad_norm": 4.453893132851004,
+      "learning_rate": 7.135383392436314e-06,
+      "loss": 1.0527,
+      "step": 13000
+    },
+    {
+      "epoch": 0.3918674698795181,
+      "grad_norm": 3.4785194656641596,
+      "learning_rate": 7.130879042030129e-06,
+      "loss": 0.9375,
+      "step": 13010
+    },
+    {
+      "epoch": 0.3921686746987952,
+      "grad_norm": 4.25105276360076,
+      "learning_rate": 7.126372577477196e-06,
+      "loss": 1.07,
+      "step": 13020
+    },
+    {
+      "epoch": 0.3924698795180723,
+      "grad_norm": 3.8572850912672454,
+      "learning_rate": 7.121864003248593e-06,
+      "loss": 1.0047,
+      "step": 13030
+    },
+    {
+      "epoch": 0.3927710843373494,
+      "grad_norm": 3.41181115848503,
+      "learning_rate": 7.117353323817491e-06,
+      "loss": 1.0365,
+      "step": 13040
+    },
+    {
+      "epoch": 0.3930722891566265,
+      "grad_norm": 3.953355678442684,
+      "learning_rate": 7.112840543659154e-06,
+      "loss": 0.9804,
+      "step": 13050
+    },
+    {
+      "epoch": 0.3933734939759036,
+      "grad_norm": 2.0763038861700864,
+      "learning_rate": 7.10832566725092e-06,
+      "loss": 0.9337,
+      "step": 13060
+    },
+    {
+      "epoch": 0.39367469879518074,
+      "grad_norm": 4.7423588529422025,
+      "learning_rate": 7.103808699072219e-06,
+      "loss": 1.0017,
+      "step": 13070
+    },
+    {
+      "epoch": 0.39397590361445783,
+      "grad_norm": 4.681127159543774,
+      "learning_rate": 7.099289643604549e-06,
+      "loss": 0.9772,
+      "step": 13080
+    },
+    {
+      "epoch": 0.3942771084337349,
+      "grad_norm": 2.075029851200195,
+      "learning_rate": 7.0947685053314785e-06,
+      "loss": 0.8014,
+      "step": 13090
+    },
+    {
+      "epoch": 0.39457831325301207,
+      "grad_norm": 8.081133957333527,
+      "learning_rate": 7.090245288738648e-06,
+      "loss": 0.9404,
+      "step": 13100
+    },
+    {
+      "epoch": 0.39487951807228916,
+      "grad_norm": 8.287050295038267,
+      "learning_rate": 7.085719998313752e-06,
+      "loss": 1.0125,
+      "step": 13110
+    },
+    {
+      "epoch": 0.39518072289156625,
+      "grad_norm": 5.2170357109722545,
+      "learning_rate": 7.081192638546551e-06,
+      "loss": 0.9959,
+      "step": 13120
+    },
+    {
+      "epoch": 0.3954819277108434,
+      "grad_norm": 3.4583333334809483,
+      "learning_rate": 7.076663213928852e-06,
+      "loss": 1.0156,
+      "step": 13130
+    },
+    {
+      "epoch": 0.3957831325301205,
+      "grad_norm": 5.98502170032457,
+      "learning_rate": 7.072131728954514e-06,
+      "loss": 0.982,
+      "step": 13140
+    },
+    {
+      "epoch": 0.3960843373493976,
+      "grad_norm": 3.8811049248944207,
+      "learning_rate": 7.0675981881194415e-06,
+      "loss": 0.9698,
+      "step": 13150
+    },
+    {
+      "epoch": 0.3963855421686747,
+      "grad_norm": 3.8526087519677605,
+      "learning_rate": 7.063062595921573e-06,
+      "loss": 0.9461,
+      "step": 13160
+    },
+    {
+      "epoch": 0.3966867469879518,
+      "grad_norm": 4.52942933395037,
+      "learning_rate": 7.0585249568608885e-06,
+      "loss": 1.0396,
+      "step": 13170
+    },
+    {
+      "epoch": 0.3969879518072289,
+      "grad_norm": 4.504220498757502,
+      "learning_rate": 7.053985275439395e-06,
+      "loss": 0.998,
+      "step": 13180
+    },
+    {
+      "epoch": 0.39728915662650605,
+      "grad_norm": 4.933275566569705,
+      "learning_rate": 7.049443556161127e-06,
+      "loss": 1.0248,
+      "step": 13190
+    },
+    {
+      "epoch": 0.39759036144578314,
+      "grad_norm": 7.435038837767327,
+      "learning_rate": 7.044899803532141e-06,
+      "loss": 1.0102,
+      "step": 13200
+    },
+    {
+      "epoch": 0.3978915662650602,
+      "grad_norm": 1.9475239073919097,
+      "learning_rate": 7.040354022060512e-06,
+      "loss": 0.9236,
+      "step": 13210
+    },
+    {
+      "epoch": 0.39819277108433737,
+      "grad_norm": 5.799189333567071,
+      "learning_rate": 7.035806216256326e-06,
+      "loss": 0.8302,
+      "step": 13220
+    },
+    {
+      "epoch": 0.39849397590361446,
+      "grad_norm": 3.474951174700604,
+      "learning_rate": 7.031256390631675e-06,
+      "loss": 1.0335,
+      "step": 13230
+    },
+    {
+      "epoch": 0.39879518072289155,
+      "grad_norm": 2.8426556346425205,
+      "learning_rate": 7.0267045497006626e-06,
+      "loss": 1.0369,
+      "step": 13240
+    },
+    {
+      "epoch": 0.3990963855421687,
+      "grad_norm": 1.9042082433939282,
+      "learning_rate": 7.022150697979385e-06,
+      "loss": 0.9762,
+      "step": 13250
+    },
+    {
+      "epoch": 0.3993975903614458,
+      "grad_norm": 3.0877693825088075,
+      "learning_rate": 7.017594839985937e-06,
+      "loss": 0.9054,
+      "step": 13260
+    },
+    {
+      "epoch": 0.3996987951807229,
+      "grad_norm": 1.8666653249253085,
+      "learning_rate": 7.013036980240401e-06,
+      "loss": 0.9911,
+      "step": 13270
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.647999584328929,
+      "learning_rate": 7.008477123264849e-06,
+      "loss": 1.0216,
+      "step": 13280
+    },
+    {
+      "epoch": 0.4003012048192771,
+      "grad_norm": 2.7021011233799324,
+      "learning_rate": 7.00391527358333e-06,
+      "loss": 1.0102,
+      "step": 13290
+    },
+    {
+      "epoch": 0.4006024096385542,
+      "grad_norm": 1.861482437705702,
+      "learning_rate": 6.999351435721875e-06,
+      "loss": 0.959,
+      "step": 13300
+    },
+    {
+      "epoch": 0.40090361445783135,
+      "grad_norm": 2.969205698521233,
+      "learning_rate": 6.994785614208484e-06,
+      "loss": 1.023,
+      "step": 13310
+    },
+    {
+      "epoch": 0.40120481927710844,
+      "grad_norm": 3.2800693516097525,
+      "learning_rate": 6.990217813573126e-06,
+      "loss": 1.0551,
+      "step": 13320
+    },
+    {
+      "epoch": 0.4015060240963855,
+      "grad_norm": 3.307627142993508,
+      "learning_rate": 6.9856480383477355e-06,
+      "loss": 1.0215,
+      "step": 13330
+    },
+    {
+      "epoch": 0.4018072289156627,
+      "grad_norm": 2.916283279567481,
+      "learning_rate": 6.981076293066204e-06,
+      "loss": 0.9557,
+      "step": 13340
+    },
+    {
+      "epoch": 0.40210843373493976,
+      "grad_norm": 1.9335183091976946,
+      "learning_rate": 6.976502582264377e-06,
+      "loss": 0.8666,
+      "step": 13350
+    },
+    {
+      "epoch": 0.40240963855421685,
+      "grad_norm": 3.7893028571634293,
+      "learning_rate": 6.971926910480052e-06,
+      "loss": 1.0476,
+      "step": 13360
+    },
+    {
+      "epoch": 0.402710843373494,
+      "grad_norm": 6.308427673739527,
+      "learning_rate": 6.967349282252973e-06,
+      "loss": 0.8638,
+      "step": 13370
+    },
+    {
+      "epoch": 0.4030120481927711,
+      "grad_norm": 1.8787099051799305,
+      "learning_rate": 6.962769702124819e-06,
+      "loss": 0.8845,
+      "step": 13380
+    },
+    {
+      "epoch": 0.4033132530120482,
+      "grad_norm": 3.0708435846186646,
+      "learning_rate": 6.958188174639214e-06,
+      "loss": 1.0184,
+      "step": 13390
+    },
+    {
+      "epoch": 0.4036144578313253,
+      "grad_norm": 3.7886839727098587,
+      "learning_rate": 6.953604704341707e-06,
+      "loss": 0.9131,
+      "step": 13400
+    },
+    {
+      "epoch": 0.4039156626506024,
+      "grad_norm": 2.8716752329837347,
+      "learning_rate": 6.9490192957797795e-06,
+      "loss": 1.0776,
+      "step": 13410
+    },
+    {
+      "epoch": 0.4042168674698795,
+      "grad_norm": 2.15047239622301,
+      "learning_rate": 6.9444319535028345e-06,
+      "loss": 0.927,
+      "step": 13420
+    },
+    {
+      "epoch": 0.40451807228915665,
+      "grad_norm": 3.240249287432723,
+      "learning_rate": 6.939842682062191e-06,
+      "loss": 0.9992,
+      "step": 13430
+    },
+    {
+      "epoch": 0.40481927710843374,
+      "grad_norm": 1.7911534969340555,
+      "learning_rate": 6.9352514860110876e-06,
+      "loss": 0.9444,
+      "step": 13440
+    },
+    {
+      "epoch": 0.40512048192771083,
+      "grad_norm": 2.0337847761318852,
+      "learning_rate": 6.930658369904664e-06,
+      "loss": 0.8718,
+      "step": 13450
+    },
+    {
+      "epoch": 0.405421686746988,
+      "grad_norm": 3.8277504093569634,
+      "learning_rate": 6.926063338299972e-06,
+      "loss": 1.0422,
+      "step": 13460
+    },
+    {
+      "epoch": 0.40572289156626506,
+      "grad_norm": 2.037541183555164,
+      "learning_rate": 6.921466395755963e-06,
+      "loss": 0.8345,
+      "step": 13470
+    },
+    {
+      "epoch": 0.40602409638554215,
+      "grad_norm": 1.857758733614469,
+      "learning_rate": 6.916867546833481e-06,
+      "loss": 1.0182,
+      "step": 13480
+    },
+    {
+      "epoch": 0.4063253012048193,
+      "grad_norm": 3.687087016747306,
+      "learning_rate": 6.912266796095267e-06,
+      "loss": 0.8923,
+      "step": 13490
+    },
+    {
+      "epoch": 0.4066265060240964,
+      "grad_norm": 3.2284327166784608,
+      "learning_rate": 6.907664148105939e-06,
+      "loss": 0.9552,
+      "step": 13500
+    },
+    {
+      "epoch": 0.4069277108433735,
+      "grad_norm": 3.3924336900550047,
+      "learning_rate": 6.90305960743201e-06,
+      "loss": 1.038,
+      "step": 13510
+    },
+    {
+      "epoch": 0.4072289156626506,
+      "grad_norm": 3.284674519110202,
+      "learning_rate": 6.8984531786418615e-06,
+      "loss": 0.9603,
+      "step": 13520
+    },
+    {
+      "epoch": 0.4075301204819277,
+      "grad_norm": 1.8492992696305701,
+      "learning_rate": 6.893844866305751e-06,
+      "loss": 0.8277,
+      "step": 13530
+    },
+    {
+      "epoch": 0.4078313253012048,
+      "grad_norm": 3.952060181626697,
+      "learning_rate": 6.889234674995806e-06,
+      "loss": 1.0991,
+      "step": 13540
+    },
+    {
+      "epoch": 0.40813253012048195,
+      "grad_norm": 4.09744931047574,
+      "learning_rate": 6.8846226092860145e-06,
+      "loss": 0.9109,
+      "step": 13550
+    },
+    {
+      "epoch": 0.40843373493975904,
+      "grad_norm": 8.374704682984751,
+      "learning_rate": 6.880008673752231e-06,
+      "loss": 1.0218,
+      "step": 13560
+    },
+    {
+      "epoch": 0.40873493975903613,
+      "grad_norm": 3.760826936335024,
+      "learning_rate": 6.87539287297216e-06,
+      "loss": 0.9871,
+      "step": 13570
+    },
+    {
+      "epoch": 0.4090361445783133,
+      "grad_norm": 3.3006168338048543,
+      "learning_rate": 6.8707752115253556e-06,
+      "loss": 0.9671,
+      "step": 13580
+    },
+    {
+      "epoch": 0.40933734939759037,
+      "grad_norm": 3.7756222005743343,
+      "learning_rate": 6.866155693993222e-06,
+      "loss": 1.0525,
+      "step": 13590
+    },
+    {
+      "epoch": 0.40963855421686746,
+      "grad_norm": 3.3836339010150143,
+      "learning_rate": 6.861534324959e-06,
+      "loss": 0.8573,
+      "step": 13600
+    },
+    {
+      "epoch": 0.4099397590361446,
+      "grad_norm": 2.9025089389335976,
+      "learning_rate": 6.856911109007773e-06,
+      "loss": 1.0036,
+      "step": 13610
+    },
+    {
+      "epoch": 0.4102409638554217,
+      "grad_norm": 3.9996824449944297,
+      "learning_rate": 6.852286050726456e-06,
+      "loss": 1.0109,
+      "step": 13620
+    },
+    {
+      "epoch": 0.4105421686746988,
+      "grad_norm": 3.737338629163707,
+      "learning_rate": 6.847659154703785e-06,
+      "loss": 0.9516,
+      "step": 13630
+    },
+    {
+      "epoch": 0.4108433734939759,
+      "grad_norm": 4.409322556657848,
+      "learning_rate": 6.843030425530328e-06,
+      "loss": 0.9622,
+      "step": 13640
+    },
+    {
+      "epoch": 0.411144578313253,
+      "grad_norm": 2.079999565648441,
+      "learning_rate": 6.838399867798467e-06,
+      "loss": 0.9414,
+      "step": 13650
+    },
+    {
+      "epoch": 0.4114457831325301,
+      "grad_norm": 4.089245574983412,
+      "learning_rate": 6.8337674861023975e-06,
+      "loss": 0.9327,
+      "step": 13660
+    },
+    {
+      "epoch": 0.41174698795180725,
+      "grad_norm": 3.3804448567114176,
+      "learning_rate": 6.829133285038128e-06,
+      "loss": 0.9547,
+      "step": 13670
+    },
+    {
+      "epoch": 0.41204819277108434,
+      "grad_norm": 2.2023297085482554,
+      "learning_rate": 6.82449726920347e-06,
+      "loss": 0.9069,
+      "step": 13680
+    },
+    {
+      "epoch": 0.41234939759036143,
+      "grad_norm": 5.184320269609493,
+      "learning_rate": 6.819859443198038e-06,
+      "loss": 0.9814,
+      "step": 13690
+    },
+    {
+      "epoch": 0.4126506024096386,
+      "grad_norm": 7.209148464071631,
+      "learning_rate": 6.815219811623237e-06,
+      "loss": 0.8775,
+      "step": 13700
+    },
+    {
+      "epoch": 0.41295180722891567,
+      "grad_norm": 5.309689062332096,
+      "learning_rate": 6.810578379082269e-06,
+      "loss": 0.9626,
+      "step": 13710
+    },
+    {
+      "epoch": 0.41325301204819276,
+      "grad_norm": 3.3671534143446693,
+      "learning_rate": 6.80593515018012e-06,
+      "loss": 0.8824,
+      "step": 13720
+    },
+    {
+      "epoch": 0.4135542168674699,
+      "grad_norm": 5.926113381160186,
+      "learning_rate": 6.8012901295235575e-06,
+      "loss": 1.0967,
+      "step": 13730
+    },
+    {
+      "epoch": 0.413855421686747,
+      "grad_norm": 3.594663545123508,
+      "learning_rate": 6.796643321721133e-06,
+      "loss": 0.9188,
+      "step": 13740
+    },
+    {
+      "epoch": 0.4141566265060241,
+      "grad_norm": 1.9681201373343726,
+      "learning_rate": 6.7919947313831604e-06,
+      "loss": 0.9908,
+      "step": 13750
+    },
+    {
+      "epoch": 0.41445783132530123,
+      "grad_norm": 3.3151046083210063,
+      "learning_rate": 6.787344363121731e-06,
+      "loss": 1.009,
+      "step": 13760
+    },
+    {
+      "epoch": 0.4147590361445783,
+      "grad_norm": 3.8683175992246754,
+      "learning_rate": 6.782692221550695e-06,
+      "loss": 0.9059,
+      "step": 13770
+    },
+    {
+      "epoch": 0.4150602409638554,
+      "grad_norm": 3.5903375766684245,
+      "learning_rate": 6.778038311285666e-06,
+      "loss": 0.9906,
+      "step": 13780
+    },
+    {
+      "epoch": 0.41536144578313255,
+      "grad_norm": 4.143125510768994,
+      "learning_rate": 6.77338263694401e-06,
+      "loss": 1.0809,
+      "step": 13790
+    },
+    {
+      "epoch": 0.41566265060240964,
+      "grad_norm": 3.817530715680697,
+      "learning_rate": 6.768725203144843e-06,
+      "loss": 1.0001,
+      "step": 13800
+    },
+    {
+      "epoch": 0.41596385542168673,
+      "grad_norm": 3.932977811242661,
+      "learning_rate": 6.764066014509028e-06,
+      "loss": 1.0326,
+      "step": 13810
+    },
+    {
+      "epoch": 0.4162650602409639,
+      "grad_norm": 1.8882146891824592,
+      "learning_rate": 6.759405075659165e-06,
+      "loss": 0.8824,
+      "step": 13820
+    },
+    {
+      "epoch": 0.41656626506024097,
+      "grad_norm": 3.63069039946318,
+      "learning_rate": 6.7547423912196e-06,
+      "loss": 0.9037,
+      "step": 13830
+    },
+    {
+      "epoch": 0.41686746987951806,
+      "grad_norm": 4.064663161840107,
+      "learning_rate": 6.7500779658163996e-06,
+      "loss": 1.0934,
+      "step": 13840
+    },
+    {
+      "epoch": 0.4171686746987952,
+      "grad_norm": 2.0403234567177657,
+      "learning_rate": 6.745411804077365e-06,
+      "loss": 0.9486,
+      "step": 13850
+    },
+    {
+      "epoch": 0.4174698795180723,
+      "grad_norm": 26.80763155624957,
+      "learning_rate": 6.740743910632017e-06,
+      "loss": 1.0734,
+      "step": 13860
+    },
+    {
+      "epoch": 0.4177710843373494,
+      "grad_norm": 4.35823590882377,
+      "learning_rate": 6.736074290111596e-06,
+      "loss": 1.0046,
+      "step": 13870
+    },
+    {
+      "epoch": 0.41807228915662653,
+      "grad_norm": 3.5932599421189386,
+      "learning_rate": 6.731402947149053e-06,
+      "loss": 0.9644,
+      "step": 13880
+    },
+    {
+      "epoch": 0.4183734939759036,
+      "grad_norm": 4.97566959388549,
+      "learning_rate": 6.7267298863790545e-06,
+      "loss": 0.9288,
+      "step": 13890
+    },
+    {
+      "epoch": 0.4186746987951807,
+      "grad_norm": 2.1059425318909795,
+      "learning_rate": 6.722055112437963e-06,
+      "loss": 0.8103,
+      "step": 13900
+    },
+    {
+      "epoch": 0.41897590361445786,
+      "grad_norm": 1.8755405253775557,
+      "learning_rate": 6.717378629963847e-06,
+      "loss": 0.9401,
+      "step": 13910
+    },
+    {
+      "epoch": 0.41927710843373495,
+      "grad_norm": 4.448628855973795,
+      "learning_rate": 6.712700443596467e-06,
+      "loss": 0.9688,
+      "step": 13920
+    },
+    {
+      "epoch": 0.41957831325301204,
+      "grad_norm": 4.087419002340595,
+      "learning_rate": 6.708020557977274e-06,
+      "loss": 0.9583,
+      "step": 13930
+    },
+    {
+      "epoch": 0.4198795180722892,
+      "grad_norm": 3.9225320035818037,
+      "learning_rate": 6.703338977749408e-06,
+      "loss": 1.0504,
+      "step": 13940
+    },
+    {
+      "epoch": 0.42018072289156627,
+      "grad_norm": 3.8556715938444386,
+      "learning_rate": 6.698655707557686e-06,
+      "loss": 1.0062,
+      "step": 13950
+    },
+    {
+      "epoch": 0.42048192771084336,
+      "grad_norm": 3.6223976107972544,
+      "learning_rate": 6.6939707520486065e-06,
+      "loss": 1.105,
+      "step": 13960
+    },
+    {
+      "epoch": 0.4207831325301205,
+      "grad_norm": 4.7489882357330595,
+      "learning_rate": 6.689284115870334e-06,
+      "loss": 0.9605,
+      "step": 13970
+    },
+    {
+      "epoch": 0.4210843373493976,
+      "grad_norm": 5.3454934516173305,
+      "learning_rate": 6.684595803672705e-06,
+      "loss": 1.007,
+      "step": 13980
+    },
+    {
+      "epoch": 0.4213855421686747,
+      "grad_norm": 2.242825259302051,
+      "learning_rate": 6.679905820107217e-06,
+      "loss": 1.0303,
+      "step": 13990
+    },
+    {
+      "epoch": 0.42168674698795183,
+      "grad_norm": 3.867493692013836,
+      "learning_rate": 6.675214169827029e-06,
+      "loss": 1.0206,
+      "step": 14000
+    },
+    {
+      "epoch": 0.4219879518072289,
+      "grad_norm": 1.9022218152561652,
+      "learning_rate": 6.6705208574869504e-06,
+      "loss": 0.878,
+      "step": 14010
+    },
+    {
+      "epoch": 0.422289156626506,
+      "grad_norm": 7.351907608382645,
+      "learning_rate": 6.665825887743437e-06,
+      "loss": 0.9591,
+      "step": 14020
+    },
+    {
+      "epoch": 0.42259036144578316,
+      "grad_norm": 4.134218650532387,
+      "learning_rate": 6.6611292652545955e-06,
+      "loss": 0.9686,
+      "step": 14030
+    },
+    {
+      "epoch": 0.42289156626506025,
+      "grad_norm": 5.545679526427971,
+      "learning_rate": 6.656430994680168e-06,
+      "loss": 1.0809,
+      "step": 14040
+    },
+    {
+      "epoch": 0.42319277108433734,
+      "grad_norm": 5.354084179744823,
+      "learning_rate": 6.651731080681532e-06,
+      "loss": 0.9496,
+      "step": 14050
+    },
+    {
+      "epoch": 0.4234939759036145,
+      "grad_norm": 5.0075043787750895,
+      "learning_rate": 6.6470295279217e-06,
+      "loss": 0.895,
+      "step": 14060
+    },
+    {
+      "epoch": 0.4237951807228916,
+      "grad_norm": 4.7102636641490125,
+      "learning_rate": 6.6423263410653025e-06,
+      "loss": 0.8286,
+      "step": 14070
+    },
+    {
+      "epoch": 0.42409638554216866,
+      "grad_norm": 10.220190685371245,
+      "learning_rate": 6.637621524778597e-06,
+      "loss": 1.0195,
+      "step": 14080
+    },
+    {
+      "epoch": 0.4243975903614458,
+      "grad_norm": 4.055966764848486,
+      "learning_rate": 6.632915083729457e-06,
+      "loss": 1.0562,
+      "step": 14090
+    },
+    {
+      "epoch": 0.4246987951807229,
+      "grad_norm": 7.568283656818405,
+      "learning_rate": 6.628207022587367e-06,
+      "loss": 0.9617,
+      "step": 14100
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 11.926916376967057,
+      "learning_rate": 6.6234973460234184e-06,
+      "loss": 0.9462,
+      "step": 14110
+    },
+    {
+      "epoch": 0.42530120481927713,
+      "grad_norm": 5.050840217737689,
+      "learning_rate": 6.618786058710306e-06,
+      "loss": 0.8864,
+      "step": 14120
+    },
+    {
+      "epoch": 0.4256024096385542,
+      "grad_norm": 5.256991626041585,
+      "learning_rate": 6.614073165322322e-06,
+      "loss": 0.98,
+      "step": 14130
+    },
+    {
+      "epoch": 0.4259036144578313,
+      "grad_norm": 3.9388850836709355,
+      "learning_rate": 6.6093586705353545e-06,
+      "loss": 1.0085,
+      "step": 14140
+    },
+    {
+      "epoch": 0.42620481927710846,
+      "grad_norm": 4.248981036001199,
+      "learning_rate": 6.604642579026878e-06,
+      "loss": 0.8712,
+      "step": 14150
+    },
+    {
+      "epoch": 0.42650602409638555,
+      "grad_norm": 4.710615070800449,
+      "learning_rate": 6.599924895475951e-06,
+      "loss": 1.0189,
+      "step": 14160
+    },
+    {
+      "epoch": 0.42680722891566264,
+      "grad_norm": 4.079255733779571,
+      "learning_rate": 6.595205624563214e-06,
+      "loss": 1.032,
+      "step": 14170
+    },
+    {
+      "epoch": 0.4271084337349398,
+      "grad_norm": 4.532412318187093,
+      "learning_rate": 6.59048477097088e-06,
+      "loss": 0.9976,
+      "step": 14180
+    },
+    {
+      "epoch": 0.4274096385542169,
+      "grad_norm": 2.2398021659582965,
+      "learning_rate": 6.585762339382732e-06,
+      "loss": 0.9661,
+      "step": 14190
+    },
+    {
+      "epoch": 0.42771084337349397,
+      "grad_norm": 5.315546724125617,
+      "learning_rate": 6.58103833448412e-06,
+      "loss": 1.0394,
+      "step": 14200
+    },
+    {
+      "epoch": 0.4280120481927711,
+      "grad_norm": 1.8260965219107057,
+      "learning_rate": 6.5763127609619556e-06,
+      "loss": 0.8876,
+      "step": 14210
+    },
+    {
+      "epoch": 0.4283132530120482,
+      "grad_norm": 1.9870117025007035,
+      "learning_rate": 6.571585623504707e-06,
+      "loss": 0.8464,
+      "step": 14220
+    },
+    {
+      "epoch": 0.4286144578313253,
+      "grad_norm": 1.9650537200759508,
+      "learning_rate": 6.56685692680239e-06,
+      "loss": 0.9696,
+      "step": 14230
+    },
+    {
+      "epoch": 0.42891566265060244,
+      "grad_norm": 2.2143800408467698,
+      "learning_rate": 6.562126675546571e-06,
+      "loss": 1.0792,
+      "step": 14240
+    },
+    {
+      "epoch": 0.4292168674698795,
+      "grad_norm": 4.6673063210425,
+      "learning_rate": 6.557394874430357e-06,
+      "loss": 0.9912,
+      "step": 14250
+    },
+    {
+      "epoch": 0.4295180722891566,
+      "grad_norm": 4.063286718378896,
+      "learning_rate": 6.552661528148399e-06,
+      "loss": 0.9883,
+      "step": 14260
+    },
+    {
+      "epoch": 0.42981927710843376,
+      "grad_norm": 5.224078068726723,
+      "learning_rate": 6.547926641396867e-06,
+      "loss": 0.9713,
+      "step": 14270
+    },
+    {
+      "epoch": 0.43012048192771085,
+      "grad_norm": 1.8485432679010896,
+      "learning_rate": 6.543190218873474e-06,
+      "loss": 0.9844,
+      "step": 14280
+    },
+    {
+      "epoch": 0.43042168674698794,
+      "grad_norm": 4.35081591142146,
+      "learning_rate": 6.538452265277448e-06,
+      "loss": 1.0075,
+      "step": 14290
+    },
+    {
+      "epoch": 0.4307228915662651,
+      "grad_norm": 2.046262765455691,
+      "learning_rate": 6.533712785309541e-06,
+      "loss": 0.8561,
+      "step": 14300
+    },
+    {
+      "epoch": 0.4310240963855422,
+      "grad_norm": 2.110252868904274,
+      "learning_rate": 6.528971783672018e-06,
+      "loss": 1.0265,
+      "step": 14310
+    },
+    {
+      "epoch": 0.43132530120481927,
+      "grad_norm": 1.9507659317012978,
+      "learning_rate": 6.524229265068649e-06,
+      "loss": 0.876,
+      "step": 14320
+    },
+    {
+      "epoch": 0.4316265060240964,
+      "grad_norm": 4.52830654049744,
+      "learning_rate": 6.5194852342047164e-06,
+      "loss": 0.9584,
+      "step": 14330
+    },
+    {
+      "epoch": 0.4319277108433735,
+      "grad_norm": 2.03685456762182,
+      "learning_rate": 6.5147396957869954e-06,
+      "loss": 1.0511,
+      "step": 14340
+    },
+    {
+      "epoch": 0.4322289156626506,
+      "grad_norm": 2.301853573936113,
+      "learning_rate": 6.5099926545237655e-06,
+      "loss": 0.9472,
+      "step": 14350
+    },
+    {
+      "epoch": 0.43253012048192774,
+      "grad_norm": 63.68246590306157,
+      "learning_rate": 6.505244115124792e-06,
+      "loss": 0.9098,
+      "step": 14360
+    },
+    {
+      "epoch": 0.43283132530120483,
+      "grad_norm": 1.9576463980977339,
+      "learning_rate": 6.500494082301325e-06,
+      "loss": 0.8991,
+      "step": 14370
+    },
+    {
+      "epoch": 0.4331325301204819,
+      "grad_norm": 18.835906175346544,
+      "learning_rate": 6.4957425607661e-06,
+      "loss": 0.9087,
+      "step": 14380
+    },
+    {
+      "epoch": 0.43343373493975906,
+      "grad_norm": 5.337406422650925,
+      "learning_rate": 6.490989555233328e-06,
+      "loss": 0.9445,
+      "step": 14390
+    },
+    {
+      "epoch": 0.43373493975903615,
+      "grad_norm": 4.558817840401622,
+      "learning_rate": 6.486235070418693e-06,
+      "loss": 1.0095,
+      "step": 14400
+    },
+    {
+      "epoch": 0.43403614457831324,
+      "grad_norm": 5.982268811359151,
+      "learning_rate": 6.481479111039346e-06,
+      "loss": 0.9943,
+      "step": 14410
+    },
+    {
+      "epoch": 0.4343373493975904,
+      "grad_norm": 4.837635700430477,
+      "learning_rate": 6.4767216818139025e-06,
+      "loss": 0.9011,
+      "step": 14420
+    },
+    {
+      "epoch": 0.4346385542168675,
+      "grad_norm": 1.9525336026031908,
+      "learning_rate": 6.4719627874624315e-06,
+      "loss": 0.9556,
+      "step": 14430
+    },
+    {
+      "epoch": 0.43493975903614457,
+      "grad_norm": 6.444628382446914,
+      "learning_rate": 6.467202432706462e-06,
+      "loss": 1.0309,
+      "step": 14440
+    },
+    {
+      "epoch": 0.4352409638554217,
+      "grad_norm": 8.989013397893679,
+      "learning_rate": 6.462440622268968e-06,
+      "loss": 0.9094,
+      "step": 14450
+    },
+    {
+      "epoch": 0.4355421686746988,
+      "grad_norm": 5.491630015275027,
+      "learning_rate": 6.457677360874369e-06,
+      "loss": 0.8831,
+      "step": 14460
+    },
+    {
+      "epoch": 0.4358433734939759,
+      "grad_norm": 5.207365934374556,
+      "learning_rate": 6.452912653248524e-06,
+      "loss": 0.8662,
+      "step": 14470
+    },
+    {
+      "epoch": 0.43614457831325304,
+      "grad_norm": 4.235258432845556,
+      "learning_rate": 6.448146504118727e-06,
+      "loss": 0.913,
+      "step": 14480
+    },
+    {
+      "epoch": 0.43644578313253013,
+      "grad_norm": 1.9979824771424923,
+      "learning_rate": 6.443378918213702e-06,
+      "loss": 0.954,
+      "step": 14490
+    },
+    {
+      "epoch": 0.4367469879518072,
+      "grad_norm": 2.0219849962315513,
+      "learning_rate": 6.438609900263597e-06,
+      "loss": 0.8747,
+      "step": 14500
+    },
+    {
+      "epoch": 0.43704819277108437,
+      "grad_norm": 4.536709248678018,
+      "learning_rate": 6.433839454999984e-06,
+      "loss": 1.0589,
+      "step": 14510
+    },
+    {
+      "epoch": 0.43734939759036146,
+      "grad_norm": 1.8603159043406443,
+      "learning_rate": 6.429067587155846e-06,
+      "loss": 0.9545,
+      "step": 14520
+    },
+    {
+      "epoch": 0.43765060240963854,
+      "grad_norm": 5.445589542955186,
+      "learning_rate": 6.424294301465587e-06,
+      "loss": 1.0,
+      "step": 14530
+    },
+    {
+      "epoch": 0.43795180722891563,
+      "grad_norm": 3.647002525829796,
+      "learning_rate": 6.419519602665005e-06,
+      "loss": 0.9889,
+      "step": 14540
+    },
+    {
+      "epoch": 0.4382530120481928,
+      "grad_norm": 3.8125231851329655,
+      "learning_rate": 6.414743495491309e-06,
+      "loss": 0.8806,
+      "step": 14550
+    },
+    {
+      "epoch": 0.43855421686746987,
+      "grad_norm": 11.404506189201793,
+      "learning_rate": 6.409965984683104e-06,
+      "loss": 1.0634,
+      "step": 14560
+    },
+    {
+      "epoch": 0.43885542168674696,
+      "grad_norm": 2.113446004546863,
+      "learning_rate": 6.4051870749803835e-06,
+      "loss": 1.0219,
+      "step": 14570
+    },
+    {
+      "epoch": 0.4391566265060241,
+      "grad_norm": 4.272081146701065,
+      "learning_rate": 6.4004067711245366e-06,
+      "loss": 0.8988,
+      "step": 14580
+    },
+    {
+      "epoch": 0.4394578313253012,
+      "grad_norm": 3.791666887388401,
+      "learning_rate": 6.395625077858324e-06,
+      "loss": 0.8926,
+      "step": 14590
+    },
+    {
+      "epoch": 0.4397590361445783,
+      "grad_norm": 5.541461091861432,
+      "learning_rate": 6.390841999925897e-06,
+      "loss": 1.0429,
+      "step": 14600
+    },
+    {
+      "epoch": 0.44006024096385543,
+      "grad_norm": 21.20771958021225,
+      "learning_rate": 6.386057542072775e-06,
+      "loss": 1.0525,
+      "step": 14610
+    },
+    {
+      "epoch": 0.4403614457831325,
+      "grad_norm": 5.5831870449008365,
+      "learning_rate": 6.381271709045844e-06,
+      "loss": 0.8978,
+      "step": 14620
+    },
+    {
+      "epoch": 0.4406626506024096,
+      "grad_norm": 6.211803700001257,
+      "learning_rate": 6.376484505593359e-06,
+      "loss": 0.9337,
+      "step": 14630
+    },
+    {
+      "epoch": 0.44096385542168676,
+      "grad_norm": 4.327191340451297,
+      "learning_rate": 6.371695936464932e-06,
+      "loss": 1.0032,
+      "step": 14640
+    },
+    {
+      "epoch": 0.44126506024096385,
+      "grad_norm": 4.017862110781731,
+      "learning_rate": 6.366906006411534e-06,
+      "loss": 1.0761,
+      "step": 14650
+    },
+    {
+      "epoch": 0.44156626506024094,
+      "grad_norm": 2.0834143928322355,
+      "learning_rate": 6.3621147201854795e-06,
+      "loss": 1.015,
+      "step": 14660
+    },
+    {
+      "epoch": 0.4418674698795181,
+      "grad_norm": 5.110735498556025,
+      "learning_rate": 6.357322082540434e-06,
+      "loss": 0.8988,
+      "step": 14670
+    },
+    {
+      "epoch": 0.44216867469879517,
+      "grad_norm": 6.354116123529262,
+      "learning_rate": 6.352528098231401e-06,
+      "loss": 0.9398,
+      "step": 14680
+    },
+    {
+      "epoch": 0.44246987951807226,
+      "grad_norm": 2.159530138316348,
+      "learning_rate": 6.347732772014721e-06,
+      "loss": 1.0521,
+      "step": 14690
+    },
+    {
+      "epoch": 0.4427710843373494,
+      "grad_norm": 1.9933712179887328,
+      "learning_rate": 6.342936108648066e-06,
+      "loss": 0.8788,
+      "step": 14700
+    },
+    {
+      "epoch": 0.4430722891566265,
+      "grad_norm": 1.7494841584547256,
+      "learning_rate": 6.338138112890434e-06,
+      "loss": 0.8607,
+      "step": 14710
+    },
+    {
+      "epoch": 0.4433734939759036,
+      "grad_norm": 9.023499609529305,
+      "learning_rate": 6.333338789502148e-06,
+      "loss": 0.8774,
+      "step": 14720
+    },
+    {
+      "epoch": 0.44367469879518073,
+      "grad_norm": 5.1989634372684215,
+      "learning_rate": 6.328538143244844e-06,
+      "loss": 0.9683,
+      "step": 14730
+    },
+    {
+      "epoch": 0.4439759036144578,
+      "grad_norm": 5.524593945218239,
+      "learning_rate": 6.32373617888147e-06,
+      "loss": 1.0263,
+      "step": 14740
+    },
+    {
+      "epoch": 0.4442771084337349,
+      "grad_norm": 1.8271075356600708,
+      "learning_rate": 6.318932901176288e-06,
+      "loss": 1.0184,
+      "step": 14750
+    },
+    {
+      "epoch": 0.44457831325301206,
+      "grad_norm": 4.9739667705548865,
+      "learning_rate": 6.314128314894856e-06,
+      "loss": 1.0007,
+      "step": 14760
+    },
+    {
+      "epoch": 0.44487951807228915,
+      "grad_norm": 7.536616587212736,
+      "learning_rate": 6.309322424804034e-06,
+      "loss": 0.9423,
+      "step": 14770
+    },
+    {
+      "epoch": 0.44518072289156624,
+      "grad_norm": 2.150751980409594,
+      "learning_rate": 6.3045152356719765e-06,
+      "loss": 0.9498,
+      "step": 14780
+    },
+    {
+      "epoch": 0.4454819277108434,
+      "grad_norm": 2.1021294109207216,
+      "learning_rate": 6.299706752268122e-06,
+      "loss": 0.9776,
+      "step": 14790
+    },
+    {
+      "epoch": 0.4457831325301205,
+      "grad_norm": 5.167209206488193,
+      "learning_rate": 6.294896979363199e-06,
+      "loss": 1.0821,
+      "step": 14800
+    },
+    {
+      "epoch": 0.44608433734939756,
+      "grad_norm": 1.8790346506081124,
+      "learning_rate": 6.290085921729211e-06,
+      "loss": 0.8127,
+      "step": 14810
+    },
+    {
+      "epoch": 0.4463855421686747,
+      "grad_norm": 3.6596766848390576,
+      "learning_rate": 6.285273584139437e-06,
+      "loss": 1.0506,
+      "step": 14820
+    },
+    {
+      "epoch": 0.4466867469879518,
+      "grad_norm": 3.4487164190170643,
+      "learning_rate": 6.280459971368431e-06,
+      "loss": 0.9859,
+      "step": 14830
+    },
+    {
+      "epoch": 0.4469879518072289,
+      "grad_norm": 1.8527260757228985,
+      "learning_rate": 6.275645088192003e-06,
+      "loss": 0.9228,
+      "step": 14840
+    },
+    {
+      "epoch": 0.44728915662650603,
+      "grad_norm": 3.2891709546322856,
+      "learning_rate": 6.270828939387232e-06,
+      "loss": 1.0175,
+      "step": 14850
+    },
+    {
+      "epoch": 0.4475903614457831,
+      "grad_norm": 4.9358873030850114,
+      "learning_rate": 6.266011529732445e-06,
+      "loss": 0.9466,
+      "step": 14860
+    },
+    {
+      "epoch": 0.4478915662650602,
+      "grad_norm": 4.0035091825160345,
+      "learning_rate": 6.261192864007226e-06,
+      "loss": 0.987,
+      "step": 14870
+    },
+    {
+      "epoch": 0.44819277108433736,
+      "grad_norm": 4.409290171950485,
+      "learning_rate": 6.256372946992405e-06,
+      "loss": 0.8875,
+      "step": 14880
+    },
+    {
+      "epoch": 0.44849397590361445,
+      "grad_norm": 3.753069019045162,
+      "learning_rate": 6.251551783470048e-06,
+      "loss": 0.9529,
+      "step": 14890
+    },
+    {
+      "epoch": 0.44879518072289154,
+      "grad_norm": 4.315050145797351,
+      "learning_rate": 6.246729378223465e-06,
+      "loss": 0.97,
+      "step": 14900
+    },
+    {
+      "epoch": 0.4490963855421687,
+      "grad_norm": 23.925305437786783,
+      "learning_rate": 6.2419057360371915e-06,
+      "loss": 1.0131,
+      "step": 14910
+    },
+    {
+      "epoch": 0.4493975903614458,
+      "grad_norm": 3.74052244546966,
+      "learning_rate": 6.237080861696994e-06,
+      "loss": 0.9892,
+      "step": 14920
+    },
+    {
+      "epoch": 0.44969879518072287,
+      "grad_norm": 3.4708257534340063,
+      "learning_rate": 6.232254759989861e-06,
+      "loss": 0.9847,
+      "step": 14930
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.468240886606056,
+      "learning_rate": 6.227427435703997e-06,
+      "loss": 1.0243,
+      "step": 14940
+    },
+    {
+      "epoch": 0.4503012048192771,
+      "grad_norm": 2.0336268064530123,
+      "learning_rate": 6.222598893628822e-06,
+      "loss": 0.868,
+      "step": 14950
+    },
+    {
+      "epoch": 0.4506024096385542,
+      "grad_norm": 5.483948935970706,
+      "learning_rate": 6.2177691385549595e-06,
+      "loss": 0.9283,
+      "step": 14960
+    },
+    {
+      "epoch": 0.45090361445783134,
+      "grad_norm": 5.076421463812874,
+      "learning_rate": 6.212938175274246e-06,
+      "loss": 1.0003,
+      "step": 14970
+    },
+    {
+      "epoch": 0.4512048192771084,
+      "grad_norm": 2.0800752070054016,
+      "learning_rate": 6.2081060085797065e-06,
+      "loss": 0.8448,
+      "step": 14980
+    },
+    {
+      "epoch": 0.4515060240963855,
+      "grad_norm": 2.0948853592958185,
+      "learning_rate": 6.2032726432655655e-06,
+      "loss": 0.9675,
+      "step": 14990
+    },
+    {
+      "epoch": 0.45180722891566266,
+      "grad_norm": 1.9638910363683588,
+      "learning_rate": 6.198438084127234e-06,
+      "loss": 0.8365,
+      "step": 15000
+    },
+    {
+      "epoch": 0.45210843373493975,
+      "grad_norm": 4.629893900801348,
+      "learning_rate": 6.1936023359613095e-06,
+      "loss": 0.9763,
+      "step": 15010
+    },
+    {
+      "epoch": 0.45240963855421684,
+      "grad_norm": 2.1526912677468117,
+      "learning_rate": 6.188765403565569e-06,
+      "loss": 0.8948,
+      "step": 15020
+    },
+    {
+      "epoch": 0.452710843373494,
+      "grad_norm": 4.639081635265701,
+      "learning_rate": 6.183927291738963e-06,
+      "loss": 0.8947,
+      "step": 15030
+    },
+    {
+      "epoch": 0.4530120481927711,
+      "grad_norm": 8.3365809018225,
+      "learning_rate": 6.179088005281612e-06,
+      "loss": 0.9226,
+      "step": 15040
+    },
+    {
+      "epoch": 0.45331325301204817,
+      "grad_norm": 16.48978218950864,
+      "learning_rate": 6.174247548994805e-06,
+      "loss": 0.9166,
+      "step": 15050
+    },
+    {
+      "epoch": 0.4536144578313253,
+      "grad_norm": 4.254658787892794,
+      "learning_rate": 6.169405927680989e-06,
+      "loss": 1.0038,
+      "step": 15060
+    },
+    {
+      "epoch": 0.4539156626506024,
+      "grad_norm": 5.272768246680619,
+      "learning_rate": 6.164563146143767e-06,
+      "loss": 0.8653,
+      "step": 15070
+    },
+    {
+      "epoch": 0.4542168674698795,
+      "grad_norm": 13.789656716662838,
+      "learning_rate": 6.159719209187896e-06,
+      "loss": 0.9806,
+      "step": 15080
+    },
+    {
+      "epoch": 0.45451807228915664,
+      "grad_norm": 2.1193314572411652,
+      "learning_rate": 6.154874121619271e-06,
+      "loss": 0.9901,
+      "step": 15090
+    },
+    {
+      "epoch": 0.45481927710843373,
+      "grad_norm": 5.486995531643311,
+      "learning_rate": 6.1500278882449415e-06,
+      "loss": 1.0283,
+      "step": 15100
+    },
+    {
+      "epoch": 0.4551204819277108,
+      "grad_norm": 2.025029285646546,
+      "learning_rate": 6.145180513873083e-06,
+      "loss": 0.8871,
+      "step": 15110
+    },
+    {
+      "epoch": 0.45542168674698796,
+      "grad_norm": 1.8356724233464343,
+      "learning_rate": 6.140332003313006e-06,
+      "loss": 0.9598,
+      "step": 15120
+    },
+    {
+      "epoch": 0.45572289156626505,
+      "grad_norm": 1.9119626614528773,
+      "learning_rate": 6.135482361375152e-06,
+      "loss": 0.9334,
+      "step": 15130
+    },
+    {
+      "epoch": 0.45602409638554214,
+      "grad_norm": 1.9457899346466425,
+      "learning_rate": 6.13063159287108e-06,
+      "loss": 0.8897,
+      "step": 15140
+    },
+    {
+      "epoch": 0.4563253012048193,
+      "grad_norm": 6.346461425624317,
+      "learning_rate": 6.125779702613471e-06,
+      "loss": 0.9209,
+      "step": 15150
+    },
+    {
+      "epoch": 0.4566265060240964,
+      "grad_norm": 11.082672495907039,
+      "learning_rate": 6.120926695416114e-06,
+      "loss": 0.9767,
+      "step": 15160
+    },
+    {
+      "epoch": 0.45692771084337347,
+      "grad_norm": 1.9417234750492238,
+      "learning_rate": 6.1160725760939125e-06,
+      "loss": 0.9292,
+      "step": 15170
+    },
+    {
+      "epoch": 0.4572289156626506,
+      "grad_norm": 4.910654199287689,
+      "learning_rate": 6.111217349462864e-06,
+      "loss": 0.9339,
+      "step": 15180
+    },
+    {
+      "epoch": 0.4575301204819277,
+      "grad_norm": 4.090750023657973,
+      "learning_rate": 6.106361020340075e-06,
+      "loss": 0.9759,
+      "step": 15190
+    },
+    {
+      "epoch": 0.4578313253012048,
+      "grad_norm": 10.45340995527639,
+      "learning_rate": 6.101503593543742e-06,
+      "loss": 1.0505,
+      "step": 15200
+    },
+    {
+      "epoch": 0.45813253012048194,
+      "grad_norm": 4.393493650041436,
+      "learning_rate": 6.096645073893147e-06,
+      "loss": 1.0164,
+      "step": 15210
+    },
+    {
+      "epoch": 0.45843373493975903,
+      "grad_norm": 2.932003843284258,
+      "learning_rate": 6.091785466208662e-06,
+      "loss": 0.9914,
+      "step": 15220
+    },
+    {
+      "epoch": 0.4587349397590361,
+      "grad_norm": 5.511367512231397,
+      "learning_rate": 6.086924775311731e-06,
+      "loss": 0.9843,
+      "step": 15230
+    },
+    {
+      "epoch": 0.45903614457831327,
+      "grad_norm": 3.622507160094181,
+      "learning_rate": 6.082063006024883e-06,
+      "loss": 0.9724,
+      "step": 15240
+    },
+    {
+      "epoch": 0.45933734939759036,
+      "grad_norm": 4.42097782859806,
+      "learning_rate": 6.077200163171707e-06,
+      "loss": 1.0642,
+      "step": 15250
+    },
+    {
+      "epoch": 0.45963855421686745,
+      "grad_norm": 1.9842004474371322,
+      "learning_rate": 6.072336251576864e-06,
+      "loss": 0.8821,
+      "step": 15260
+    },
+    {
+      "epoch": 0.4599397590361446,
+      "grad_norm": 4.8517110679655095,
+      "learning_rate": 6.067471276066071e-06,
+      "loss": 0.9392,
+      "step": 15270
+    },
+    {
+      "epoch": 0.4602409638554217,
+      "grad_norm": 4.067247531850998,
+      "learning_rate": 6.062605241466102e-06,
+      "loss": 0.9131,
+      "step": 15280
+    },
+    {
+      "epoch": 0.46054216867469877,
+      "grad_norm": 3.628697422434493,
+      "learning_rate": 6.057738152604783e-06,
+      "loss": 0.9612,
+      "step": 15290
+    },
+    {
+      "epoch": 0.4608433734939759,
+      "grad_norm": 1.8238831228622823,
+      "learning_rate": 6.0528700143109856e-06,
+      "loss": 0.8898,
+      "step": 15300
+    },
+    {
+      "epoch": 0.461144578313253,
+      "grad_norm": 1.9506638259955418,
+      "learning_rate": 6.048000831414621e-06,
+      "loss": 0.9569,
+      "step": 15310
+    },
+    {
+      "epoch": 0.4614457831325301,
+      "grad_norm": 2.038623106823851,
+      "learning_rate": 6.043130608746638e-06,
+      "loss": 1.0339,
+      "step": 15320
+    },
+    {
+      "epoch": 0.46174698795180724,
+      "grad_norm": 4.081124055135755,
+      "learning_rate": 6.0382593511390175e-06,
+      "loss": 0.9014,
+      "step": 15330
+    },
+    {
+      "epoch": 0.46204819277108433,
+      "grad_norm": 4.2039738385113,
+      "learning_rate": 6.033387063424765e-06,
+      "loss": 1.0432,
+      "step": 15340
+    },
+    {
+      "epoch": 0.4623493975903614,
+      "grad_norm": 4.771012156477197,
+      "learning_rate": 6.0285137504379136e-06,
+      "loss": 0.8905,
+      "step": 15350
+    },
+    {
+      "epoch": 0.46265060240963857,
+      "grad_norm": 3.8698912276052826,
+      "learning_rate": 6.023639417013503e-06,
+      "loss": 0.9479,
+      "step": 15360
+    },
+    {
+      "epoch": 0.46295180722891566,
+      "grad_norm": 3.653400357953905,
+      "learning_rate": 6.018764067987597e-06,
+      "loss": 0.9512,
+      "step": 15370
+    },
+    {
+      "epoch": 0.46325301204819275,
+      "grad_norm": 4.353461797388446,
+      "learning_rate": 6.013887708197261e-06,
+      "loss": 1.0287,
+      "step": 15380
+    },
+    {
+      "epoch": 0.4635542168674699,
+      "grad_norm": 1.9962682766637123,
+      "learning_rate": 6.009010342480563e-06,
+      "loss": 0.9262,
+      "step": 15390
+    },
+    {
+      "epoch": 0.463855421686747,
+      "grad_norm": 3.393075109953769,
+      "learning_rate": 6.004131975676571e-06,
+      "loss": 0.9228,
+      "step": 15400
+    },
+    {
+      "epoch": 0.4641566265060241,
+      "grad_norm": 3.631344542832589,
+      "learning_rate": 5.999252612625345e-06,
+      "loss": 0.968,
+      "step": 15410
+    },
+    {
+      "epoch": 0.4644578313253012,
+      "grad_norm": 3.8199424808707336,
+      "learning_rate": 5.994372258167935e-06,
+      "loss": 0.9909,
+      "step": 15420
+    },
+    {
+      "epoch": 0.4647590361445783,
+      "grad_norm": 2.020212908749961,
+      "learning_rate": 5.9894909171463715e-06,
+      "loss": 0.907,
+      "step": 15430
+    },
+    {
+      "epoch": 0.4650602409638554,
+      "grad_norm": 4.435765129156293,
+      "learning_rate": 5.984608594403667e-06,
+      "loss": 0.9393,
+      "step": 15440
+    },
+    {
+      "epoch": 0.46536144578313254,
+      "grad_norm": 4.272774995977806,
+      "learning_rate": 5.979725294783808e-06,
+      "loss": 1.0173,
+      "step": 15450
+    },
+    {
+      "epoch": 0.46566265060240963,
+      "grad_norm": 3.4351057816326476,
+      "learning_rate": 5.974841023131745e-06,
+      "loss": 0.9682,
+      "step": 15460
+    },
+    {
+      "epoch": 0.4659638554216867,
+      "grad_norm": 3.343332344783758,
+      "learning_rate": 5.969955784293403e-06,
+      "loss": 0.9354,
+      "step": 15470
+    },
+    {
+      "epoch": 0.46626506024096387,
+      "grad_norm": 5.375685723600656,
+      "learning_rate": 5.965069583115655e-06,
+      "loss": 0.986,
+      "step": 15480
+    },
+    {
+      "epoch": 0.46656626506024096,
+      "grad_norm": 3.7740604287826423,
+      "learning_rate": 5.9601824244463354e-06,
+      "loss": 0.98,
+      "step": 15490
+    },
+    {
+      "epoch": 0.46686746987951805,
+      "grad_norm": 1.958611231889104,
+      "learning_rate": 5.95529431313423e-06,
+      "loss": 0.8225,
+      "step": 15500
+    },
+    {
+      "epoch": 0.4671686746987952,
+      "grad_norm": 3.752224755808922,
+      "learning_rate": 5.9504052540290655e-06,
+      "loss": 1.0359,
+      "step": 15510
+    },
+    {
+      "epoch": 0.4674698795180723,
+      "grad_norm": 1.9932450938425532,
+      "learning_rate": 5.945515251981511e-06,
+      "loss": 0.8257,
+      "step": 15520
+    },
+    {
+      "epoch": 0.4677710843373494,
+      "grad_norm": 6.141550273844298,
+      "learning_rate": 5.94062431184317e-06,
+      "loss": 0.9846,
+      "step": 15530
+    },
+    {
+      "epoch": 0.4680722891566265,
+      "grad_norm": 4.069407210066842,
+      "learning_rate": 5.935732438466578e-06,
+      "loss": 0.9558,
+      "step": 15540
+    },
+    {
+      "epoch": 0.4683734939759036,
+      "grad_norm": 3.8240966984099583,
+      "learning_rate": 5.930839636705196e-06,
+      "loss": 0.9485,
+      "step": 15550
+    },
+    {
+      "epoch": 0.4686746987951807,
+      "grad_norm": 2.1199639876192333,
+      "learning_rate": 5.925945911413405e-06,
+      "loss": 0.9409,
+      "step": 15560
+    },
+    {
+      "epoch": 0.46897590361445785,
+      "grad_norm": 4.253009603524578,
+      "learning_rate": 5.921051267446507e-06,
+      "loss": 1.066,
+      "step": 15570
+    },
+    {
+      "epoch": 0.46927710843373494,
+      "grad_norm": 4.834378079812853,
+      "learning_rate": 5.916155709660707e-06,
+      "loss": 0.9991,
+      "step": 15580
+    },
+    {
+      "epoch": 0.469578313253012,
+      "grad_norm": 2.1958271799561264,
+      "learning_rate": 5.911259242913124e-06,
+      "loss": 0.9344,
+      "step": 15590
+    },
+    {
+      "epoch": 0.46987951807228917,
+      "grad_norm": 4.251159210640614,
+      "learning_rate": 5.906361872061776e-06,
+      "loss": 0.8428,
+      "step": 15600
+    },
+    {
+      "epoch": 0.47018072289156626,
+      "grad_norm": 3.4579718127619654,
+      "learning_rate": 5.901463601965578e-06,
+      "loss": 1.0252,
+      "step": 15610
+    },
+    {
+      "epoch": 0.47048192771084335,
+      "grad_norm": 4.123233913463495,
+      "learning_rate": 5.896564437484338e-06,
+      "loss": 0.9784,
+      "step": 15620
+    },
+    {
+      "epoch": 0.4707831325301205,
+      "grad_norm": 3.805917111735305,
+      "learning_rate": 5.89166438347875e-06,
+      "loss": 0.9682,
+      "step": 15630
+    },
+    {
+      "epoch": 0.4710843373493976,
+      "grad_norm": 3.472449173403724,
+      "learning_rate": 5.886763444810391e-06,
+      "loss": 0.9287,
+      "step": 15640
+    },
+    {
+      "epoch": 0.4713855421686747,
+      "grad_norm": 3.505798023665673,
+      "learning_rate": 5.8818616263417175e-06,
+      "loss": 1.0114,
+      "step": 15650
+    },
+    {
+      "epoch": 0.4716867469879518,
+      "grad_norm": 2.2101309910312774,
+      "learning_rate": 5.876958932936056e-06,
+      "loss": 0.9078,
+      "step": 15660
+    },
+    {
+      "epoch": 0.4719879518072289,
+      "grad_norm": 5.024115659027036,
+      "learning_rate": 5.872055369457603e-06,
+      "loss": 0.9819,
+      "step": 15670
+    },
+    {
+      "epoch": 0.472289156626506,
+      "grad_norm": 1.9388970415560967,
+      "learning_rate": 5.867150940771418e-06,
+      "loss": 0.972,
+      "step": 15680
+    },
+    {
+      "epoch": 0.47259036144578315,
+      "grad_norm": 1.923641231368815,
+      "learning_rate": 5.862245651743419e-06,
+      "loss": 0.9894,
+      "step": 15690
+    },
+    {
+      "epoch": 0.47289156626506024,
+      "grad_norm": 3.5761198625957924,
+      "learning_rate": 5.857339507240376e-06,
+      "loss": 1.0599,
+      "step": 15700
+    },
+    {
+      "epoch": 0.4731927710843373,
+      "grad_norm": 3.7466601545245815,
+      "learning_rate": 5.85243251212991e-06,
+      "loss": 1.0256,
+      "step": 15710
+    },
+    {
+      "epoch": 0.4734939759036145,
+      "grad_norm": 1.8344281968594405,
+      "learning_rate": 5.8475246712804845e-06,
+      "loss": 0.9542,
+      "step": 15720
+    },
+    {
+      "epoch": 0.47379518072289156,
+      "grad_norm": 2.821146548540811,
+      "learning_rate": 5.8426159895614025e-06,
+      "loss": 0.9607,
+      "step": 15730
+    },
+    {
+      "epoch": 0.47409638554216865,
+      "grad_norm": 2.0109465472597425,
+      "learning_rate": 5.837706471842802e-06,
+      "loss": 0.9255,
+      "step": 15740
+    },
+    {
+      "epoch": 0.4743975903614458,
+      "grad_norm": 3.793950849129281,
+      "learning_rate": 5.832796122995648e-06,
+      "loss": 0.9613,
+      "step": 15750
+    },
+    {
+      "epoch": 0.4746987951807229,
+      "grad_norm": 3.5195351231947214,
+      "learning_rate": 5.8278849478917345e-06,
+      "loss": 0.9551,
+      "step": 15760
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 3.44390448833659,
+      "learning_rate": 5.82297295140367e-06,
+      "loss": 0.9127,
+      "step": 15770
+    },
+    {
+      "epoch": 0.4753012048192771,
+      "grad_norm": 2.064193688758877,
+      "learning_rate": 5.818060138404882e-06,
+      "loss": 0.9144,
+      "step": 15780
+    },
+    {
+      "epoch": 0.4756024096385542,
+      "grad_norm": 3.930204164602873,
+      "learning_rate": 5.813146513769607e-06,
+      "loss": 1.0549,
+      "step": 15790
+    },
+    {
+      "epoch": 0.4759036144578313,
+      "grad_norm": 1.8105852580327773,
+      "learning_rate": 5.808232082372884e-06,
+      "loss": 0.8884,
+      "step": 15800
+    },
+    {
+      "epoch": 0.47620481927710845,
+      "grad_norm": 3.401223624686546,
+      "learning_rate": 5.803316849090559e-06,
+      "loss": 0.8972,
+      "step": 15810
+    },
+    {
+      "epoch": 0.47650602409638554,
+      "grad_norm": 4.218879987468935,
+      "learning_rate": 5.798400818799266e-06,
+      "loss": 0.9108,
+      "step": 15820
+    },
+    {
+      "epoch": 0.47680722891566263,
+      "grad_norm": 5.303922998277883,
+      "learning_rate": 5.793483996376433e-06,
+      "loss": 0.9926,
+      "step": 15830
+    },
+    {
+      "epoch": 0.4771084337349398,
+      "grad_norm": 3.2250696460266854,
+      "learning_rate": 5.788566386700276e-06,
+      "loss": 0.9818,
+      "step": 15840
+    },
+    {
+      "epoch": 0.47740963855421686,
+      "grad_norm": 2.0605598980504443,
+      "learning_rate": 5.7836479946497905e-06,
+      "loss": 0.9629,
+      "step": 15850
+    },
+    {
+      "epoch": 0.47771084337349395,
+      "grad_norm": 1.990069257829727,
+      "learning_rate": 5.778728825104745e-06,
+      "loss": 0.9655,
+      "step": 15860
+    },
+    {
+      "epoch": 0.4780120481927711,
+      "grad_norm": 4.242590643192938,
+      "learning_rate": 5.773808882945684e-06,
+      "loss": 1.0566,
+      "step": 15870
+    },
+    {
+      "epoch": 0.4783132530120482,
+      "grad_norm": 3.201563490514386,
+      "learning_rate": 5.768888173053917e-06,
+      "loss": 0.9913,
+      "step": 15880
+    },
+    {
+      "epoch": 0.4786144578313253,
+      "grad_norm": 3.6375935532616106,
+      "learning_rate": 5.763966700311516e-06,
+      "loss": 1.0634,
+      "step": 15890
+    },
+    {
+      "epoch": 0.4789156626506024,
+      "grad_norm": 3.42465631670704,
+      "learning_rate": 5.759044469601307e-06,
+      "loss": 0.9021,
+      "step": 15900
+    },
+    {
+      "epoch": 0.4792168674698795,
+      "grad_norm": 3.5933105683523103,
+      "learning_rate": 5.7541214858068705e-06,
+      "loss": 0.9248,
+      "step": 15910
+    },
+    {
+      "epoch": 0.4795180722891566,
+      "grad_norm": 3.818194076790111,
+      "learning_rate": 5.749197753812533e-06,
+      "loss": 0.9539,
+      "step": 15920
+    },
+    {
+      "epoch": 0.47981927710843375,
+      "grad_norm": 2.9683401007502197,
+      "learning_rate": 5.744273278503365e-06,
+      "loss": 0.9131,
+      "step": 15930
+    },
+    {
+      "epoch": 0.48012048192771084,
+      "grad_norm": 3.976873236658985,
+      "learning_rate": 5.739348064765173e-06,
+      "loss": 0.8902,
+      "step": 15940
+    },
+    {
+      "epoch": 0.48042168674698793,
+      "grad_norm": 1.8917758808628107,
+      "learning_rate": 5.734422117484496e-06,
+      "loss": 0.9426,
+      "step": 15950
+    },
+    {
+      "epoch": 0.4807228915662651,
+      "grad_norm": 3.718175782520566,
+      "learning_rate": 5.729495441548603e-06,
+      "loss": 0.9244,
+      "step": 15960
+    },
+    {
+      "epoch": 0.48102409638554217,
+      "grad_norm": 4.92755103394381,
+      "learning_rate": 5.724568041845482e-06,
+      "loss": 1.0401,
+      "step": 15970
+    },
+    {
+      "epoch": 0.48132530120481926,
+      "grad_norm": 3.991297733931076,
+      "learning_rate": 5.719639923263841e-06,
+      "loss": 0.914,
+      "step": 15980
+    },
+    {
+      "epoch": 0.4816265060240964,
+      "grad_norm": 3.683504880134318,
+      "learning_rate": 5.714711090693105e-06,
+      "loss": 1.0589,
+      "step": 15990
+    },
+    {
+      "epoch": 0.4819277108433735,
+      "grad_norm": 2.103618704503937,
+      "learning_rate": 5.7097815490234e-06,
+      "loss": 1.0494,
+      "step": 16000
+    },
+    {
+      "epoch": 0.4822289156626506,
+      "grad_norm": 3.148525504100099,
+      "learning_rate": 5.704851303145562e-06,
+      "loss": 0.9384,
+      "step": 16010
+    },
+    {
+      "epoch": 0.4825301204819277,
+      "grad_norm": 3.7542549837016566,
+      "learning_rate": 5.699920357951122e-06,
+      "loss": 0.9955,
+      "step": 16020
+    },
+    {
+      "epoch": 0.4828313253012048,
+      "grad_norm": 4.263674016496408,
+      "learning_rate": 5.694988718332305e-06,
+      "loss": 1.0508,
+      "step": 16030
+    },
+    {
+      "epoch": 0.4831325301204819,
+      "grad_norm": 2.0369555452543433,
+      "learning_rate": 5.690056389182029e-06,
+      "loss": 0.9159,
+      "step": 16040
+    },
+    {
+      "epoch": 0.48343373493975905,
+      "grad_norm": 2.018208089452482,
+      "learning_rate": 5.68512337539389e-06,
+      "loss": 0.9311,
+      "step": 16050
+    },
+    {
+      "epoch": 0.48373493975903614,
+      "grad_norm": 3.6863006912289906,
+      "learning_rate": 5.680189681862169e-06,
+      "loss": 0.9928,
+      "step": 16060
+    },
+    {
+      "epoch": 0.48403614457831323,
+      "grad_norm": 3.883766819311878,
+      "learning_rate": 5.675255313481816e-06,
+      "loss": 1.011,
+      "step": 16070
+    },
+    {
+      "epoch": 0.4843373493975904,
+      "grad_norm": 4.022219241501254,
+      "learning_rate": 5.670320275148453e-06,
+      "loss": 1.0192,
+      "step": 16080
+    },
+    {
+      "epoch": 0.48463855421686747,
+      "grad_norm": 3.625969394490103,
+      "learning_rate": 5.66538457175837e-06,
+      "loss": 0.8921,
+      "step": 16090
+    },
+    {
+      "epoch": 0.48493975903614456,
+      "grad_norm": 3.934005534709987,
+      "learning_rate": 5.660448208208513e-06,
+      "loss": 1.0507,
+      "step": 16100
+    },
+    {
+      "epoch": 0.4852409638554217,
+      "grad_norm": 1.9434506662927051,
+      "learning_rate": 5.655511189396482e-06,
+      "loss": 0.8856,
+      "step": 16110
+    },
+    {
+      "epoch": 0.4855421686746988,
+      "grad_norm": 3.2593475606632722,
+      "learning_rate": 5.650573520220528e-06,
+      "loss": 0.8519,
+      "step": 16120
+    },
+    {
+      "epoch": 0.4858433734939759,
+      "grad_norm": 3.5373579746777466,
+      "learning_rate": 5.645635205579551e-06,
+      "loss": 0.9807,
+      "step": 16130
+    },
+    {
+      "epoch": 0.48614457831325303,
+      "grad_norm": 3.4058108606170596,
+      "learning_rate": 5.640696250373089e-06,
+      "loss": 0.9153,
+      "step": 16140
+    },
+    {
+      "epoch": 0.4864457831325301,
+      "grad_norm": 2.1061104208189803,
+      "learning_rate": 5.635756659501312e-06,
+      "loss": 1.0289,
+      "step": 16150
+    },
+    {
+      "epoch": 0.4867469879518072,
+      "grad_norm": 3.7143925075194324,
+      "learning_rate": 5.630816437865026e-06,
+      "loss": 0.9853,
+      "step": 16160
+    },
+    {
+      "epoch": 0.48704819277108435,
+      "grad_norm": 3.4675259291475284,
+      "learning_rate": 5.62587559036566e-06,
+      "loss": 0.9042,
+      "step": 16170
+    },
+    {
+      "epoch": 0.48734939759036144,
+      "grad_norm": 3.258157402874766,
+      "learning_rate": 5.6209341219052625e-06,
+      "loss": 1.0681,
+      "step": 16180
+    },
+    {
+      "epoch": 0.48765060240963853,
+      "grad_norm": 3.5750396020214454,
+      "learning_rate": 5.615992037386505e-06,
+      "loss": 0.9625,
+      "step": 16190
+    },
+    {
+      "epoch": 0.4879518072289157,
+      "grad_norm": 5.736958849615456,
+      "learning_rate": 5.61104934171266e-06,
+      "loss": 1.0082,
+      "step": 16200
+    },
+    {
+      "epoch": 0.48825301204819277,
+      "grad_norm": 3.370647439497055,
+      "learning_rate": 5.606106039787614e-06,
+      "loss": 0.9527,
+      "step": 16210
+    },
+    {
+      "epoch": 0.48855421686746986,
+      "grad_norm": 4.12517330758,
+      "learning_rate": 5.601162136515853e-06,
+      "loss": 0.96,
+      "step": 16220
+    },
+    {
+      "epoch": 0.488855421686747,
+      "grad_norm": 3.697771225010867,
+      "learning_rate": 5.596217636802459e-06,
+      "loss": 1.0006,
+      "step": 16230
+    },
+    {
+      "epoch": 0.4891566265060241,
+      "grad_norm": 3.281437853463002,
+      "learning_rate": 5.591272545553105e-06,
+      "loss": 0.9578,
+      "step": 16240
+    },
+    {
+      "epoch": 0.4894578313253012,
+      "grad_norm": 4.635097307896269,
+      "learning_rate": 5.586326867674052e-06,
+      "loss": 0.9059,
+      "step": 16250
+    },
+    {
+      "epoch": 0.48975903614457833,
+      "grad_norm": 14.395878411842967,
+      "learning_rate": 5.5813806080721465e-06,
+      "loss": 1.0381,
+      "step": 16260
+    },
+    {
+      "epoch": 0.4900602409638554,
+      "grad_norm": 3.193713434679751,
+      "learning_rate": 5.576433771654803e-06,
+      "loss": 0.8877,
+      "step": 16270
+    },
+    {
+      "epoch": 0.4903614457831325,
+      "grad_norm": 4.207042639185046,
+      "learning_rate": 5.571486363330019e-06,
+      "loss": 0.998,
+      "step": 16280
+    },
+    {
+      "epoch": 0.49066265060240966,
+      "grad_norm": 3.5938725066185464,
+      "learning_rate": 5.566538388006351e-06,
+      "loss": 0.9872,
+      "step": 16290
+    },
+    {
+      "epoch": 0.49096385542168675,
+      "grad_norm": 4.012319826138452,
+      "learning_rate": 5.56158985059292e-06,
+      "loss": 1.0337,
+      "step": 16300
+    },
+    {
+      "epoch": 0.49126506024096384,
+      "grad_norm": 2.113113287905368,
+      "learning_rate": 5.556640755999412e-06,
+      "loss": 0.9452,
+      "step": 16310
+    },
+    {
+      "epoch": 0.491566265060241,
+      "grad_norm": 3.413853496438198,
+      "learning_rate": 5.551691109136053e-06,
+      "loss": 0.9903,
+      "step": 16320
+    },
+    {
+      "epoch": 0.49186746987951807,
+      "grad_norm": 5.3313796303409,
+      "learning_rate": 5.54674091491363e-06,
+      "loss": 0.968,
+      "step": 16330
+    },
+    {
+      "epoch": 0.49216867469879516,
+      "grad_norm": 7.5749167631122765,
+      "learning_rate": 5.5417901782434624e-06,
+      "loss": 0.918,
+      "step": 16340
+    },
+    {
+      "epoch": 0.4924698795180723,
+      "grad_norm": 3.886130707996784,
+      "learning_rate": 5.536838904037415e-06,
+      "loss": 1.0647,
+      "step": 16350
+    },
+    {
+      "epoch": 0.4927710843373494,
+      "grad_norm": 3.3551181609478564,
+      "learning_rate": 5.531887097207881e-06,
+      "loss": 0.9375,
+      "step": 16360
+    },
+    {
+      "epoch": 0.4930722891566265,
+      "grad_norm": 4.268081130671484,
+      "learning_rate": 5.526934762667783e-06,
+      "loss": 1.0252,
+      "step": 16370
+    },
+    {
+      "epoch": 0.49337349397590363,
+      "grad_norm": 2.0952551160850095,
+      "learning_rate": 5.521981905330572e-06,
+      "loss": 0.9003,
+      "step": 16380
+    },
+    {
+      "epoch": 0.4936746987951807,
+      "grad_norm": 4.735644970986755,
+      "learning_rate": 5.517028530110212e-06,
+      "loss": 1.0146,
+      "step": 16390
+    },
+    {
+      "epoch": 0.4939759036144578,
+      "grad_norm": 3.3887005523926694,
+      "learning_rate": 5.512074641921185e-06,
+      "loss": 0.9946,
+      "step": 16400
+    },
+    {
+      "epoch": 0.49427710843373496,
+      "grad_norm": 3.3361038650093424,
+      "learning_rate": 5.507120245678476e-06,
+      "loss": 1.0096,
+      "step": 16410
+    },
+    {
+      "epoch": 0.49457831325301205,
+      "grad_norm": 3.516599104133175,
+      "learning_rate": 5.5021653462975795e-06,
+      "loss": 0.9707,
+      "step": 16420
+    },
+    {
+      "epoch": 0.49487951807228914,
+      "grad_norm": 3.796649488608317,
+      "learning_rate": 5.49720994869449e-06,
+      "loss": 0.9433,
+      "step": 16430
+    },
+    {
+      "epoch": 0.4951807228915663,
+      "grad_norm": 2.161848452188897,
+      "learning_rate": 5.492254057785689e-06,
+      "loss": 1.0216,
+      "step": 16440
+    },
+    {
+      "epoch": 0.4954819277108434,
+      "grad_norm": 3.6082558748692275,
+      "learning_rate": 5.487297678488156e-06,
+      "loss": 1.0332,
+      "step": 16450
+    },
+    {
+      "epoch": 0.49578313253012046,
+      "grad_norm": 3.6447919883275954,
+      "learning_rate": 5.4823408157193516e-06,
+      "loss": 0.9219,
+      "step": 16460
+    },
+    {
+      "epoch": 0.4960843373493976,
+      "grad_norm": 3.8284824445728423,
+      "learning_rate": 5.477383474397213e-06,
+      "loss": 0.9927,
+      "step": 16470
+    },
+    {
+      "epoch": 0.4963855421686747,
+      "grad_norm": 3.8326747018520066,
+      "learning_rate": 5.472425659440157e-06,
+      "loss": 0.8893,
+      "step": 16480
+    },
+    {
+      "epoch": 0.4966867469879518,
+      "grad_norm": 1.911047180054358,
+      "learning_rate": 5.4674673757670685e-06,
+      "loss": 0.9214,
+      "step": 16490
+    },
+    {
+      "epoch": 0.49698795180722893,
+      "grad_norm": 1.88254732748915,
+      "learning_rate": 5.462508628297295e-06,
+      "loss": 0.9643,
+      "step": 16500
+    },
+    {
+      "epoch": 0.497289156626506,
+      "grad_norm": 2.082898839883877,
+      "learning_rate": 5.457549421950651e-06,
+      "loss": 0.9658,
+      "step": 16510
+    },
+    {
+      "epoch": 0.4975903614457831,
+      "grad_norm": 4.889244313390055,
+      "learning_rate": 5.4525897616473955e-06,
+      "loss": 0.9504,
+      "step": 16520
+    },
+    {
+      "epoch": 0.49789156626506026,
+      "grad_norm": 4.4994284167839105,
+      "learning_rate": 5.447629652308249e-06,
+      "loss": 0.9766,
+      "step": 16530
+    },
+    {
+      "epoch": 0.49819277108433735,
+      "grad_norm": 3.539972699302177,
+      "learning_rate": 5.442669098854371e-06,
+      "loss": 1.0098,
+      "step": 16540
+    },
+    {
+      "epoch": 0.49849397590361444,
+      "grad_norm": 3.095128366613158,
+      "learning_rate": 5.437708106207362e-06,
+      "loss": 0.9201,
+      "step": 16550
+    },
+    {
+      "epoch": 0.4987951807228916,
+      "grad_norm": 11.814963135069243,
+      "learning_rate": 5.432746679289263e-06,
+      "loss": 0.9717,
+      "step": 16560
+    },
+    {
+      "epoch": 0.4990963855421687,
+      "grad_norm": 3.868649908326447,
+      "learning_rate": 5.4277848230225375e-06,
+      "loss": 1.0115,
+      "step": 16570
+    },
+    {
+      "epoch": 0.49939759036144576,
+      "grad_norm": 8.5980024382822,
+      "learning_rate": 5.422822542330084e-06,
+      "loss": 0.952,
+      "step": 16580
+    },
+    {
+      "epoch": 0.4996987951807229,
+      "grad_norm": 3.3459203711087873,
+      "learning_rate": 5.417859842135214e-06,
+      "loss": 0.7811,
+      "step": 16590
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.8414494077082875,
+      "learning_rate": 5.412896727361663e-06,
+      "loss": 0.865,
+      "step": 16600
+    },
+    {
+      "epoch": 0.5003012048192771,
+      "grad_norm": 2.096825972352622,
+      "learning_rate": 5.40793320293357e-06,
+      "loss": 0.8719,
+      "step": 16610
+    },
+    {
+      "epoch": 0.5006024096385542,
+      "grad_norm": 3.3793756859421085,
+      "learning_rate": 5.402969273775488e-06,
+      "loss": 1.0394,
+      "step": 16620
+    },
+    {
+      "epoch": 0.5009036144578313,
+      "grad_norm": 3.5366354603671377,
+      "learning_rate": 5.398004944812368e-06,
+      "loss": 0.9838,
+      "step": 16630
+    },
+    {
+      "epoch": 0.5012048192771085,
+      "grad_norm": 3.443967399642233,
+      "learning_rate": 5.393040220969553e-06,
+      "loss": 0.882,
+      "step": 16640
+    },
+    {
+      "epoch": 0.5015060240963856,
+      "grad_norm": 3.883316324592676,
+      "learning_rate": 5.388075107172788e-06,
+      "loss": 1.0068,
+      "step": 16650
+    },
+    {
+      "epoch": 0.5018072289156627,
+      "grad_norm": 1.9661484881701687,
+      "learning_rate": 5.383109608348195e-06,
+      "loss": 0.9525,
+      "step": 16660
+    },
+    {
+      "epoch": 0.5021084337349397,
+      "grad_norm": 4.397328415581478,
+      "learning_rate": 5.378143729422285e-06,
+      "loss": 0.8473,
+      "step": 16670
+    },
+    {
+      "epoch": 0.5024096385542168,
+      "grad_norm": 3.7685085101788474,
+      "learning_rate": 5.373177475321941e-06,
+      "loss": 0.8884,
+      "step": 16680
+    },
+    {
+      "epoch": 0.5027108433734939,
+      "grad_norm": 3.52299888030089,
+      "learning_rate": 5.3682108509744215e-06,
+      "loss": 0.952,
+      "step": 16690
+    },
+    {
+      "epoch": 0.5030120481927711,
+      "grad_norm": 3.1985604354976,
+      "learning_rate": 5.363243861307352e-06,
+      "loss": 0.9665,
+      "step": 16700
+    },
+    {
+      "epoch": 0.5033132530120482,
+      "grad_norm": 2.0732364114962256,
+      "learning_rate": 5.358276511248721e-06,
+      "loss": 0.8739,
+      "step": 16710
+    },
+    {
+      "epoch": 0.5036144578313253,
+      "grad_norm": 5.338086958389926,
+      "learning_rate": 5.3533088057268726e-06,
+      "loss": 1.0419,
+      "step": 16720
+    },
+    {
+      "epoch": 0.5039156626506024,
+      "grad_norm": 1.8328216310168854,
+      "learning_rate": 5.348340749670502e-06,
+      "loss": 0.8538,
+      "step": 16730
+    },
+    {
+      "epoch": 0.5042168674698795,
+      "grad_norm": 5.621038149996067,
+      "learning_rate": 5.343372348008657e-06,
+      "loss": 0.9176,
+      "step": 16740
+    },
+    {
+      "epoch": 0.5045180722891566,
+      "grad_norm": 6.444314752666458,
+      "learning_rate": 5.338403605670726e-06,
+      "loss": 1.0439,
+      "step": 16750
+    },
+    {
+      "epoch": 0.5048192771084338,
+      "grad_norm": 4.368052837782761,
+      "learning_rate": 5.333434527586435e-06,
+      "loss": 0.9521,
+      "step": 16760
+    },
+    {
+      "epoch": 0.5051204819277109,
+      "grad_norm": 4.8500848545911985,
+      "learning_rate": 5.3284651186858415e-06,
+      "loss": 1.0498,
+      "step": 16770
+    },
+    {
+      "epoch": 0.505421686746988,
+      "grad_norm": 3.3613658046037673,
+      "learning_rate": 5.323495383899336e-06,
+      "loss": 0.9349,
+      "step": 16780
+    },
+    {
+      "epoch": 0.505722891566265,
+      "grad_norm": 3.586912532561054,
+      "learning_rate": 5.318525328157629e-06,
+      "loss": 0.9039,
+      "step": 16790
+    },
+    {
+      "epoch": 0.5060240963855421,
+      "grad_norm": 4.141339792328036,
+      "learning_rate": 5.313554956391748e-06,
+      "loss": 0.9305,
+      "step": 16800
+    },
+    {
+      "epoch": 0.5063253012048192,
+      "grad_norm": 9.439351804482932,
+      "learning_rate": 5.308584273533038e-06,
+      "loss": 1.0657,
+      "step": 16810
+    },
+    {
+      "epoch": 0.5066265060240964,
+      "grad_norm": 3.882519429169925,
+      "learning_rate": 5.3036132845131485e-06,
+      "loss": 0.9061,
+      "step": 16820
+    },
+    {
+      "epoch": 0.5069277108433735,
+      "grad_norm": 5.082502222873879,
+      "learning_rate": 5.298641994264038e-06,
+      "loss": 0.9464,
+      "step": 16830
+    },
+    {
+      "epoch": 0.5072289156626506,
+      "grad_norm": 3.678118838009688,
+      "learning_rate": 5.293670407717957e-06,
+      "loss": 0.9193,
+      "step": 16840
+    },
+    {
+      "epoch": 0.5075301204819277,
+      "grad_norm": 1.928823152489833,
+      "learning_rate": 5.288698529807455e-06,
+      "loss": 0.9423,
+      "step": 16850
+    },
+    {
+      "epoch": 0.5078313253012048,
+      "grad_norm": 5.346641251646554,
+      "learning_rate": 5.2837263654653715e-06,
+      "loss": 0.8375,
+      "step": 16860
+    },
+    {
+      "epoch": 0.5081325301204819,
+      "grad_norm": 7.322236295268061,
+      "learning_rate": 5.278753919624824e-06,
+      "loss": 0.9388,
+      "step": 16870
+    },
+    {
+      "epoch": 0.5084337349397591,
+      "grad_norm": 1.8586260821774663,
+      "learning_rate": 5.273781197219217e-06,
+      "loss": 0.8868,
+      "step": 16880
+    },
+    {
+      "epoch": 0.5087349397590362,
+      "grad_norm": 5.532900617505604,
+      "learning_rate": 5.268808203182222e-06,
+      "loss": 1.0514,
+      "step": 16890
+    },
+    {
+      "epoch": 0.5090361445783133,
+      "grad_norm": 4.011795165947062,
+      "learning_rate": 5.263834942447787e-06,
+      "loss": 0.9166,
+      "step": 16900
+    },
+    {
+      "epoch": 0.5093373493975903,
+      "grad_norm": 3.7516058698720944,
+      "learning_rate": 5.25886141995012e-06,
+      "loss": 0.964,
+      "step": 16910
+    },
+    {
+      "epoch": 0.5096385542168674,
+      "grad_norm": 4.206936197114627,
+      "learning_rate": 5.253887640623688e-06,
+      "loss": 0.949,
+      "step": 16920
+    },
+    {
+      "epoch": 0.5099397590361445,
+      "grad_norm": 5.066907135959254,
+      "learning_rate": 5.248913609403218e-06,
+      "loss": 1.0073,
+      "step": 16930
+    },
+    {
+      "epoch": 0.5102409638554217,
+      "grad_norm": 3.3512241583371454,
+      "learning_rate": 5.243939331223681e-06,
+      "loss": 1.0165,
+      "step": 16940
+    },
+    {
+      "epoch": 0.5105421686746988,
+      "grad_norm": 3.927804930781103,
+      "learning_rate": 5.238964811020299e-06,
+      "loss": 1.0325,
+      "step": 16950
+    },
+    {
+      "epoch": 0.5108433734939759,
+      "grad_norm": 1.9899819718907672,
+      "learning_rate": 5.233990053728525e-06,
+      "loss": 0.9605,
+      "step": 16960
+    },
+    {
+      "epoch": 0.511144578313253,
+      "grad_norm": 3.3601632658216025,
+      "learning_rate": 5.229015064284057e-06,
+      "loss": 1.0048,
+      "step": 16970
+    },
+    {
+      "epoch": 0.5114457831325301,
+      "grad_norm": 4.047990702246386,
+      "learning_rate": 5.224039847622821e-06,
+      "loss": 1.0879,
+      "step": 16980
+    },
+    {
+      "epoch": 0.5117469879518072,
+      "grad_norm": 3.7722216904260333,
+      "learning_rate": 5.219064408680964e-06,
+      "loss": 1.0098,
+      "step": 16990
+    },
+    {
+      "epoch": 0.5120481927710844,
+      "grad_norm": 1.93704875953522,
+      "learning_rate": 5.214088752394857e-06,
+      "loss": 0.9124,
+      "step": 17000
+    },
+    {
+      "epoch": 0.5123493975903615,
+      "grad_norm": 7.500304067179252,
+      "learning_rate": 5.209112883701086e-06,
+      "loss": 1.0047,
+      "step": 17010
+    },
+    {
+      "epoch": 0.5126506024096386,
+      "grad_norm": 3.47159955480306,
+      "learning_rate": 5.204136807536446e-06,
+      "loss": 1.0444,
+      "step": 17020
+    },
+    {
+      "epoch": 0.5129518072289156,
+      "grad_norm": 3.5945005438563125,
+      "learning_rate": 5.1991605288379435e-06,
+      "loss": 0.9632,
+      "step": 17030
+    },
+    {
+      "epoch": 0.5132530120481927,
+      "grad_norm": 3.7044643169981386,
+      "learning_rate": 5.194184052542779e-06,
+      "loss": 0.844,
+      "step": 17040
+    },
+    {
+      "epoch": 0.5135542168674698,
+      "grad_norm": 4.019279915658249,
+      "learning_rate": 5.189207383588353e-06,
+      "loss": 0.9806,
+      "step": 17050
+    },
+    {
+      "epoch": 0.513855421686747,
+      "grad_norm": 3.7902114831517504,
+      "learning_rate": 5.184230526912255e-06,
+      "loss": 0.9074,
+      "step": 17060
+    },
+    {
+      "epoch": 0.5141566265060241,
+      "grad_norm": 3.3532788051860853,
+      "learning_rate": 5.1792534874522614e-06,
+      "loss": 1.0144,
+      "step": 17070
+    },
+    {
+      "epoch": 0.5144578313253012,
+      "grad_norm": 3.712019236827946,
+      "learning_rate": 5.174276270146335e-06,
+      "loss": 1.0634,
+      "step": 17080
+    },
+    {
+      "epoch": 0.5147590361445783,
+      "grad_norm": 1.9844335581390444,
+      "learning_rate": 5.169298879932605e-06,
+      "loss": 0.8989,
+      "step": 17090
+    },
+    {
+      "epoch": 0.5150602409638554,
+      "grad_norm": 5.182718294180661,
+      "learning_rate": 5.1643213217493815e-06,
+      "loss": 0.9546,
+      "step": 17100
+    },
+    {
+      "epoch": 0.5153614457831325,
+      "grad_norm": 3.5507685142494854,
+      "learning_rate": 5.1593436005351364e-06,
+      "loss": 1.0191,
+      "step": 17110
+    },
+    {
+      "epoch": 0.5156626506024097,
+      "grad_norm": 2.05952870693458,
+      "learning_rate": 5.154365721228502e-06,
+      "loss": 0.9325,
+      "step": 17120
+    },
+    {
+      "epoch": 0.5159638554216868,
+      "grad_norm": 3.634311089964099,
+      "learning_rate": 5.149387688768274e-06,
+      "loss": 0.9314,
+      "step": 17130
+    },
+    {
+      "epoch": 0.5162650602409639,
+      "grad_norm": 4.304533662235416,
+      "learning_rate": 5.144409508093391e-06,
+      "loss": 1.0373,
+      "step": 17140
+    },
+    {
+      "epoch": 0.516566265060241,
+      "grad_norm": 3.5427747846881616,
+      "learning_rate": 5.139431184142949e-06,
+      "loss": 0.947,
+      "step": 17150
+    },
+    {
+      "epoch": 0.516867469879518,
+      "grad_norm": 4.153543241500085,
+      "learning_rate": 5.1344527218561754e-06,
+      "loss": 0.9852,
+      "step": 17160
+    },
+    {
+      "epoch": 0.5171686746987951,
+      "grad_norm": 3.9695888760257345,
+      "learning_rate": 5.129474126172444e-06,
+      "loss": 0.9578,
+      "step": 17170
+    },
+    {
+      "epoch": 0.5174698795180723,
+      "grad_norm": 3.1424892784401597,
+      "learning_rate": 5.124495402031255e-06,
+      "loss": 0.9758,
+      "step": 17180
+    },
+    {
+      "epoch": 0.5177710843373494,
+      "grad_norm": 4.4720130227836865,
+      "learning_rate": 5.119516554372239e-06,
+      "loss": 0.8981,
+      "step": 17190
+    },
+    {
+      "epoch": 0.5180722891566265,
+      "grad_norm": 4.455717293784039,
+      "learning_rate": 5.114537588135149e-06,
+      "loss": 1.0093,
+      "step": 17200
+    },
+    {
+      "epoch": 0.5183734939759036,
+      "grad_norm": 1.8445645487163518,
+      "learning_rate": 5.109558508259852e-06,
+      "loss": 0.8968,
+      "step": 17210
+    },
+    {
+      "epoch": 0.5186746987951807,
+      "grad_norm": 3.4562959811878606,
+      "learning_rate": 5.104579319686335e-06,
+      "loss": 0.9239,
+      "step": 17220
+    },
+    {
+      "epoch": 0.5189759036144578,
+      "grad_norm": 3.4755156299950176,
+      "learning_rate": 5.0996000273546845e-06,
+      "loss": 0.9662,
+      "step": 17230
+    },
+    {
+      "epoch": 0.519277108433735,
+      "grad_norm": 4.054513837691121,
+      "learning_rate": 5.094620636205096e-06,
+      "loss": 0.9137,
+      "step": 17240
+    },
+    {
+      "epoch": 0.5195783132530121,
+      "grad_norm": 1.912461255336725,
+      "learning_rate": 5.089641151177861e-06,
+      "loss": 0.9686,
+      "step": 17250
+    },
+    {
+      "epoch": 0.5198795180722892,
+      "grad_norm": 1.8423595709599472,
+      "learning_rate": 5.084661577213363e-06,
+      "loss": 0.9428,
+      "step": 17260
+    },
+    {
+      "epoch": 0.5201807228915662,
+      "grad_norm": 1.8158581187420892,
+      "learning_rate": 5.079681919252076e-06,
+      "loss": 1.0003,
+      "step": 17270
+    },
+    {
+      "epoch": 0.5204819277108433,
+      "grad_norm": 4.787825750588715,
+      "learning_rate": 5.074702182234554e-06,
+      "loss": 1.0396,
+      "step": 17280
+    },
+    {
+      "epoch": 0.5207831325301204,
+      "grad_norm": 3.502506610642177,
+      "learning_rate": 5.0697223711014355e-06,
+      "loss": 0.9102,
+      "step": 17290
+    },
+    {
+      "epoch": 0.5210843373493976,
+      "grad_norm": 1.833976135044235,
+      "learning_rate": 5.0647424907934265e-06,
+      "loss": 0.9133,
+      "step": 17300
+    },
+    {
+      "epoch": 0.5213855421686747,
+      "grad_norm": 2.0346727920494883,
+      "learning_rate": 5.059762546251304e-06,
+      "loss": 1.0013,
+      "step": 17310
+    },
+    {
+      "epoch": 0.5216867469879518,
+      "grad_norm": 4.690418242534549,
+      "learning_rate": 5.054782542415908e-06,
+      "loss": 0.9793,
+      "step": 17320
+    },
+    {
+      "epoch": 0.5219879518072289,
+      "grad_norm": 4.718099340990269,
+      "learning_rate": 5.049802484228139e-06,
+      "loss": 0.9517,
+      "step": 17330
+    },
+    {
+      "epoch": 0.522289156626506,
+      "grad_norm": 1.774718518131299,
+      "learning_rate": 5.0448223766289485e-06,
+      "loss": 0.8892,
+      "step": 17340
+    },
+    {
+      "epoch": 0.5225903614457831,
+      "grad_norm": 3.3161270959024085,
+      "learning_rate": 5.03984222455934e-06,
+      "loss": 0.9927,
+      "step": 17350
+    },
+    {
+      "epoch": 0.5228915662650603,
+      "grad_norm": 3.6248939785901673,
+      "learning_rate": 5.03486203296036e-06,
+      "loss": 0.9208,
+      "step": 17360
+    },
+    {
+      "epoch": 0.5231927710843374,
+      "grad_norm": 2.1425751026089297,
+      "learning_rate": 5.029881806773096e-06,
+      "loss": 0.9274,
+      "step": 17370
+    },
+    {
+      "epoch": 0.5234939759036145,
+      "grad_norm": 4.163664848576764,
+      "learning_rate": 5.024901550938665e-06,
+      "loss": 0.987,
+      "step": 17380
+    },
+    {
+      "epoch": 0.5237951807228916,
+      "grad_norm": 6.647738532957064,
+      "learning_rate": 5.019921270398215e-06,
+      "loss": 1.0385,
+      "step": 17390
+    },
+    {
+      "epoch": 0.5240963855421686,
+      "grad_norm": 5.3044873464190045,
+      "learning_rate": 5.014940970092926e-06,
+      "loss": 0.9234,
+      "step": 17400
+    },
+    {
+      "epoch": 0.5243975903614457,
+      "grad_norm": 5.20998974799349,
+      "learning_rate": 5.009960654963985e-06,
+      "loss": 0.98,
+      "step": 17410
+    },
+    {
+      "epoch": 0.5246987951807229,
+      "grad_norm": 3.4715745569717775,
+      "learning_rate": 5.004980329952606e-06,
+      "loss": 0.9598,
+      "step": 17420
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 5.251536276234192,
+      "learning_rate": 5e-06,
+      "loss": 1.0854,
+      "step": 17430
+    },
+    {
+      "epoch": 0.5253012048192771,
+      "grad_norm": 4.42136930557203,
+      "learning_rate": 4.995019670047397e-06,
+      "loss": 0.9094,
+      "step": 17440
+    },
+    {
+      "epoch": 0.5256024096385542,
+      "grad_norm": 4.001410215815066,
+      "learning_rate": 4.990039345036015e-06,
+      "loss": 0.9663,
+      "step": 17450
+    },
+    {
+      "epoch": 0.5259036144578313,
+      "grad_norm": 3.463306758529009,
+      "learning_rate": 4.985059029907075e-06,
+      "loss": 1.0536,
+      "step": 17460
+    },
+    {
+      "epoch": 0.5262048192771084,
+      "grad_norm": 3.522453864262608,
+      "learning_rate": 4.980078729601786e-06,
+      "loss": 0.9821,
+      "step": 17470
+    },
+    {
+      "epoch": 0.5265060240963856,
+      "grad_norm": 4.056603488992403,
+      "learning_rate": 4.9750984490613385e-06,
+      "loss": 0.9351,
+      "step": 17480
+    },
+    {
+      "epoch": 0.5268072289156627,
+      "grad_norm": 4.868537127841475,
+      "learning_rate": 4.970118193226906e-06,
+      "loss": 0.9887,
+      "step": 17490
+    },
+    {
+      "epoch": 0.5271084337349398,
+      "grad_norm": 1.9246294326348845,
+      "learning_rate": 4.965137967039641e-06,
+      "loss": 0.9296,
+      "step": 17500
+    },
+    {
+      "epoch": 0.5274096385542169,
+      "grad_norm": 2.225110752535346,
+      "learning_rate": 4.960157775440661e-06,
+      "loss": 0.9025,
+      "step": 17510
+    },
+    {
+      "epoch": 0.5277108433734939,
+      "grad_norm": 3.37465435685935,
+      "learning_rate": 4.955177623371052e-06,
+      "loss": 1.0208,
+      "step": 17520
+    },
+    {
+      "epoch": 0.528012048192771,
+      "grad_norm": 2.0000699800782207,
+      "learning_rate": 4.950197515771862e-06,
+      "loss": 0.9341,
+      "step": 17530
+    },
+    {
+      "epoch": 0.5283132530120482,
+      "grad_norm": 2.0211153004119358,
+      "learning_rate": 4.945217457584093e-06,
+      "loss": 0.9828,
+      "step": 17540
+    },
+    {
+      "epoch": 0.5286144578313253,
+      "grad_norm": 3.4947840076925014,
+      "learning_rate": 4.940237453748698e-06,
+      "loss": 0.9797,
+      "step": 17550
+    },
+    {
+      "epoch": 0.5289156626506024,
+      "grad_norm": 2.7483176433858088,
+      "learning_rate": 4.9352575092065735e-06,
+      "loss": 0.9642,
+      "step": 17560
+    },
+    {
+      "epoch": 0.5292168674698795,
+      "grad_norm": 1.8624396770427465,
+      "learning_rate": 4.930277628898565e-06,
+      "loss": 0.8357,
+      "step": 17570
+    },
+    {
+      "epoch": 0.5295180722891566,
+      "grad_norm": 2.142716430230594,
+      "learning_rate": 4.925297817765447e-06,
+      "loss": 0.9222,
+      "step": 17580
+    },
+    {
+      "epoch": 0.5298192771084337,
+      "grad_norm": 3.397721034908635,
+      "learning_rate": 4.920318080747926e-06,
+      "loss": 0.9685,
+      "step": 17590
+    },
+    {
+      "epoch": 0.5301204819277109,
+      "grad_norm": 3.7788232985839167,
+      "learning_rate": 4.915338422786639e-06,
+      "loss": 0.9021,
+      "step": 17600
+    },
+    {
+      "epoch": 0.530421686746988,
+      "grad_norm": 2.105592567465183,
+      "learning_rate": 4.910358848822141e-06,
+      "loss": 0.9367,
+      "step": 17610
+    },
+    {
+      "epoch": 0.5307228915662651,
+      "grad_norm": 3.861617065581075,
+      "learning_rate": 4.905379363794907e-06,
+      "loss": 1.0091,
+      "step": 17620
+    },
+    {
+      "epoch": 0.5310240963855422,
+      "grad_norm": 4.0934314679798245,
+      "learning_rate": 4.900399972645316e-06,
+      "loss": 0.9582,
+      "step": 17630
+    },
+    {
+      "epoch": 0.5313253012048192,
+      "grad_norm": 3.2743123976308115,
+      "learning_rate": 4.8954206803136675e-06,
+      "loss": 1.0131,
+      "step": 17640
+    },
+    {
+      "epoch": 0.5316265060240963,
+      "grad_norm": 3.91844832163504,
+      "learning_rate": 4.89044149174015e-06,
+      "loss": 1.015,
+      "step": 17650
+    },
+    {
+      "epoch": 0.5319277108433735,
+      "grad_norm": 3.220576979788481,
+      "learning_rate": 4.885462411864854e-06,
+      "loss": 0.9799,
+      "step": 17660
+    },
+    {
+      "epoch": 0.5322289156626506,
+      "grad_norm": 3.4419367482701557,
+      "learning_rate": 4.880483445627762e-06,
+      "loss": 0.9709,
+      "step": 17670
+    },
+    {
+      "epoch": 0.5325301204819277,
+      "grad_norm": 3.1698518987875444,
+      "learning_rate": 4.875504597968746e-06,
+      "loss": 0.885,
+      "step": 17680
+    },
+    {
+      "epoch": 0.5328313253012048,
+      "grad_norm": 5.532121312291464,
+      "learning_rate": 4.8705258738275575e-06,
+      "loss": 0.8918,
+      "step": 17690
+    },
+    {
+      "epoch": 0.5331325301204819,
+      "grad_norm": 4.087596559666748,
+      "learning_rate": 4.8655472781438245e-06,
+      "loss": 0.995,
+      "step": 17700
+    },
+    {
+      "epoch": 0.533433734939759,
+      "grad_norm": 2.041813382381951,
+      "learning_rate": 4.860568815857052e-06,
+      "loss": 1.04,
+      "step": 17710
+    },
+    {
+      "epoch": 0.5337349397590362,
+      "grad_norm": 4.325577814191232,
+      "learning_rate": 4.855590491906611e-06,
+      "loss": 1.0045,
+      "step": 17720
+    },
+    {
+      "epoch": 0.5340361445783133,
+      "grad_norm": 4.915812484305553,
+      "learning_rate": 4.850612311231729e-06,
+      "loss": 0.969,
+      "step": 17730
+    },
+    {
+      "epoch": 0.5343373493975904,
+      "grad_norm": 3.9105916475757168,
+      "learning_rate": 4.845634278771499e-06,
+      "loss": 0.8759,
+      "step": 17740
+    },
+    {
+      "epoch": 0.5346385542168675,
+      "grad_norm": 4.108462135369391,
+      "learning_rate": 4.840656399464866e-06,
+      "loss": 0.9976,
+      "step": 17750
+    },
+    {
+      "epoch": 0.5349397590361445,
+      "grad_norm": 1.9324226196798016,
+      "learning_rate": 4.83567867825062e-06,
+      "loss": 0.8654,
+      "step": 17760
+    },
+    {
+      "epoch": 0.5352409638554216,
+      "grad_norm": 3.580893486094657,
+      "learning_rate": 4.830701120067395e-06,
+      "loss": 0.9513,
+      "step": 17770
+    },
+    {
+      "epoch": 0.5355421686746988,
+      "grad_norm": 1.9065088849623395,
+      "learning_rate": 4.825723729853666e-06,
+      "loss": 0.9402,
+      "step": 17780
+    },
+    {
+      "epoch": 0.5358433734939759,
+      "grad_norm": 3.5789673775529853,
+      "learning_rate": 4.820746512547739e-06,
+      "loss": 0.9909,
+      "step": 17790
+    },
+    {
+      "epoch": 0.536144578313253,
+      "grad_norm": 1.797134767555618,
+      "learning_rate": 4.815769473087748e-06,
+      "loss": 0.9261,
+      "step": 17800
+    },
+    {
+      "epoch": 0.5364457831325301,
+      "grad_norm": 4.699788177369744,
+      "learning_rate": 4.81079261641165e-06,
+      "loss": 0.9696,
+      "step": 17810
+    },
+    {
+      "epoch": 0.5367469879518072,
+      "grad_norm": 4.0347822848961785,
+      "learning_rate": 4.805815947457223e-06,
+      "loss": 1.0048,
+      "step": 17820
+    },
+    {
+      "epoch": 0.5370481927710843,
+      "grad_norm": 1.8877736285879627,
+      "learning_rate": 4.800839471162058e-06,
+      "loss": 0.9339,
+      "step": 17830
+    },
+    {
+      "epoch": 0.5373493975903615,
+      "grad_norm": 4.061992285499995,
+      "learning_rate": 4.795863192463555e-06,
+      "loss": 0.9045,
+      "step": 17840
+    },
+    {
+      "epoch": 0.5376506024096386,
+      "grad_norm": 3.2819153245763713,
+      "learning_rate": 4.790887116298915e-06,
+      "loss": 1.0393,
+      "step": 17850
+    },
+    {
+      "epoch": 0.5379518072289157,
+      "grad_norm": 5.540904481275622,
+      "learning_rate": 4.785911247605144e-06,
+      "loss": 1.0325,
+      "step": 17860
+    },
+    {
+      "epoch": 0.5382530120481928,
+      "grad_norm": 3.9915915146618874,
+      "learning_rate": 4.780935591319038e-06,
+      "loss": 1.0207,
+      "step": 17870
+    },
+    {
+      "epoch": 0.5385542168674698,
+      "grad_norm": 3.3171207025582246,
+      "learning_rate": 4.7759601523771795e-06,
+      "loss": 0.9246,
+      "step": 17880
+    },
+    {
+      "epoch": 0.5388554216867469,
+      "grad_norm": 6.787166761029138,
+      "learning_rate": 4.770984935715943e-06,
+      "loss": 1.0477,
+      "step": 17890
+    },
+    {
+      "epoch": 0.5391566265060241,
+      "grad_norm": 1.9347569368374558,
+      "learning_rate": 4.766009946271477e-06,
+      "loss": 0.8728,
+      "step": 17900
+    },
+    {
+      "epoch": 0.5394578313253012,
+      "grad_norm": 3.3782462151839545,
+      "learning_rate": 4.7610351889797045e-06,
+      "loss": 0.9746,
+      "step": 17910
+    },
+    {
+      "epoch": 0.5397590361445783,
+      "grad_norm": 6.551315258730182,
+      "learning_rate": 4.75606066877632e-06,
+      "loss": 1.0123,
+      "step": 17920
+    },
+    {
+      "epoch": 0.5400602409638554,
+      "grad_norm": 4.852048589451806,
+      "learning_rate": 4.751086390596784e-06,
+      "loss": 0.9505,
+      "step": 17930
+    },
+    {
+      "epoch": 0.5403614457831325,
+      "grad_norm": 4.129868979814431,
+      "learning_rate": 4.7461123593763145e-06,
+      "loss": 0.9803,
+      "step": 17940
+    },
+    {
+      "epoch": 0.5406626506024096,
+      "grad_norm": 4.492534192590365,
+      "learning_rate": 4.741138580049881e-06,
+      "loss": 0.973,
+      "step": 17950
+    },
+    {
+      "epoch": 0.5409638554216868,
+      "grad_norm": 6.711979410474132,
+      "learning_rate": 4.736165057552214e-06,
+      "loss": 0.9755,
+      "step": 17960
+    },
+    {
+      "epoch": 0.5412650602409639,
+      "grad_norm": 2.143347947839009,
+      "learning_rate": 4.73119179681778e-06,
+      "loss": 0.9958,
+      "step": 17970
+    },
+    {
+      "epoch": 0.541566265060241,
+      "grad_norm": 2.050007419900864,
+      "learning_rate": 4.726218802780786e-06,
+      "loss": 0.9691,
+      "step": 17980
+    },
+    {
+      "epoch": 0.5418674698795181,
+      "grad_norm": 3.56161063519906,
+      "learning_rate": 4.721246080375177e-06,
+      "loss": 0.9714,
+      "step": 17990
+    },
+    {
+      "epoch": 0.5421686746987951,
+      "grad_norm": 3.5959178734003974,
+      "learning_rate": 4.71627363453463e-06,
+      "loss": 0.8454,
+      "step": 18000
+    },
+    {
+      "epoch": 0.5424698795180722,
+      "grad_norm": 4.890303472159129,
+      "learning_rate": 4.711301470192546e-06,
+      "loss": 0.9353,
+      "step": 18010
+    },
+    {
+      "epoch": 0.5427710843373494,
+      "grad_norm": 1.8706983160096207,
+      "learning_rate": 4.706329592282044e-06,
+      "loss": 0.9297,
+      "step": 18020
+    },
+    {
+      "epoch": 0.5430722891566265,
+      "grad_norm": 3.940535617761609,
+      "learning_rate": 4.701358005735963e-06,
+      "loss": 0.9424,
+      "step": 18030
+    },
+    {
+      "epoch": 0.5433734939759036,
+      "grad_norm": 6.2321792966622285,
+      "learning_rate": 4.696386715486852e-06,
+      "loss": 0.8667,
+      "step": 18040
+    },
+    {
+      "epoch": 0.5436746987951807,
+      "grad_norm": 2.0485982332257295,
+      "learning_rate": 4.691415726466964e-06,
+      "loss": 0.9719,
+      "step": 18050
+    },
+    {
+      "epoch": 0.5439759036144578,
+      "grad_norm": 2.1319835740570885,
+      "learning_rate": 4.686445043608252e-06,
+      "loss": 0.8765,
+      "step": 18060
+    },
+    {
+      "epoch": 0.5442771084337349,
+      "grad_norm": 3.869973375080356,
+      "learning_rate": 4.681474671842372e-06,
+      "loss": 1.0161,
+      "step": 18070
+    },
+    {
+      "epoch": 0.5445783132530121,
+      "grad_norm": 3.9407887892231024,
+      "learning_rate": 4.676504616100665e-06,
+      "loss": 0.9007,
+      "step": 18080
+    },
+    {
+      "epoch": 0.5448795180722892,
+      "grad_norm": 4.208148187578163,
+      "learning_rate": 4.671534881314159e-06,
+      "loss": 0.9004,
+      "step": 18090
+    },
+    {
+      "epoch": 0.5451807228915663,
+      "grad_norm": 6.110829514591843,
+      "learning_rate": 4.666565472413566e-06,
+      "loss": 0.9694,
+      "step": 18100
+    },
+    {
+      "epoch": 0.5454819277108434,
+      "grad_norm": 4.11019509226744,
+      "learning_rate": 4.6615963943292746e-06,
+      "loss": 0.9301,
+      "step": 18110
+    },
+    {
+      "epoch": 0.5457831325301205,
+      "grad_norm": 4.8827685283160625,
+      "learning_rate": 4.656627651991345e-06,
+      "loss": 0.9501,
+      "step": 18120
+    },
+    {
+      "epoch": 0.5460843373493975,
+      "grad_norm": 1.9276015787098681,
+      "learning_rate": 4.651659250329498e-06,
+      "loss": 0.9188,
+      "step": 18130
+    },
+    {
+      "epoch": 0.5463855421686747,
+      "grad_norm": 5.253274346868863,
+      "learning_rate": 4.64669119427313e-06,
+      "loss": 0.9809,
+      "step": 18140
+    },
+    {
+      "epoch": 0.5466867469879518,
+      "grad_norm": 2.0747414953348593,
+      "learning_rate": 4.6417234887512806e-06,
+      "loss": 0.84,
+      "step": 18150
+    },
+    {
+      "epoch": 0.5469879518072289,
+      "grad_norm": 4.081615293667149,
+      "learning_rate": 4.636756138692649e-06,
+      "loss": 0.9868,
+      "step": 18160
+    },
+    {
+      "epoch": 0.547289156626506,
+      "grad_norm": 1.9149772773466076,
+      "learning_rate": 4.631789149025579e-06,
+      "loss": 0.9231,
+      "step": 18170
+    },
+    {
+      "epoch": 0.5475903614457831,
+      "grad_norm": 1.9362830048849156,
+      "learning_rate": 4.62682252467806e-06,
+      "loss": 0.9863,
+      "step": 18180
+    },
+    {
+      "epoch": 0.5478915662650602,
+      "grad_norm": 1.8451057636135912,
+      "learning_rate": 4.6218562705777185e-06,
+      "loss": 0.7929,
+      "step": 18190
+    },
+    {
+      "epoch": 0.5481927710843374,
+      "grad_norm": 4.306535903769046,
+      "learning_rate": 4.616890391651806e-06,
+      "loss": 0.8187,
+      "step": 18200
+    },
+    {
+      "epoch": 0.5484939759036145,
+      "grad_norm": 6.2784166812719375,
+      "learning_rate": 4.611924892827214e-06,
+      "loss": 1.0032,
+      "step": 18210
+    },
+    {
+      "epoch": 0.5487951807228916,
+      "grad_norm": 5.140685994476185,
+      "learning_rate": 4.606959779030448e-06,
+      "loss": 0.9382,
+      "step": 18220
+    },
+    {
+      "epoch": 0.5490963855421687,
+      "grad_norm": 1.9687472287786298,
+      "learning_rate": 4.6019950551876344e-06,
+      "loss": 0.9879,
+      "step": 18230
+    },
+    {
+      "epoch": 0.5493975903614458,
+      "grad_norm": 6.239333787072131,
+      "learning_rate": 4.597030726224512e-06,
+      "loss": 0.8209,
+      "step": 18240
+    },
+    {
+      "epoch": 0.5496987951807228,
+      "grad_norm": 1.9971325593798464,
+      "learning_rate": 4.5920667970664305e-06,
+      "loss": 0.9806,
+      "step": 18250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.065120900266769,
+      "learning_rate": 4.587103272638339e-06,
+      "loss": 0.9718,
+      "step": 18260
+    },
+    {
+      "epoch": 0.5503012048192771,
+      "grad_norm": 1.8479007403994274,
+      "learning_rate": 4.582140157864786e-06,
+      "loss": 0.9457,
+      "step": 18270
+    },
+    {
+      "epoch": 0.5506024096385542,
+      "grad_norm": 5.232467134900734,
+      "learning_rate": 4.577177457669918e-06,
+      "loss": 0.9836,
+      "step": 18280
+    },
+    {
+      "epoch": 0.5509036144578313,
+      "grad_norm": 3.686243550061335,
+      "learning_rate": 4.572215176977465e-06,
+      "loss": 1.0281,
+      "step": 18290
+    },
+    {
+      "epoch": 0.5512048192771084,
+      "grad_norm": 1.9160010285583549,
+      "learning_rate": 4.56725332071074e-06,
+      "loss": 0.8726,
+      "step": 18300
+    },
+    {
+      "epoch": 0.5515060240963855,
+      "grad_norm": 3.9259430418582584,
+      "learning_rate": 4.562291893792639e-06,
+      "loss": 0.9226,
+      "step": 18310
+    },
+    {
+      "epoch": 0.5518072289156627,
+      "grad_norm": 1.958130002078843,
+      "learning_rate": 4.557330901145631e-06,
+      "loss": 0.9502,
+      "step": 18320
+    },
+    {
+      "epoch": 0.5521084337349398,
+      "grad_norm": 4.683451039328426,
+      "learning_rate": 4.552370347691752e-06,
+      "loss": 1.0428,
+      "step": 18330
+    },
+    {
+      "epoch": 0.5524096385542169,
+      "grad_norm": 7.304841244404751,
+      "learning_rate": 4.5474102383526045e-06,
+      "loss": 0.9516,
+      "step": 18340
+    },
+    {
+      "epoch": 0.552710843373494,
+      "grad_norm": 3.9484063329996713,
+      "learning_rate": 4.542450578049351e-06,
+      "loss": 0.9295,
+      "step": 18350
+    },
+    {
+      "epoch": 0.553012048192771,
+      "grad_norm": 1.9830861189302107,
+      "learning_rate": 4.5374913717027065e-06,
+      "loss": 0.9482,
+      "step": 18360
+    },
+    {
+      "epoch": 0.5533132530120481,
+      "grad_norm": 4.071109109673892,
+      "learning_rate": 4.532532624232934e-06,
+      "loss": 0.8806,
+      "step": 18370
+    },
+    {
+      "epoch": 0.5536144578313253,
+      "grad_norm": 3.877351017673144,
+      "learning_rate": 4.527574340559844e-06,
+      "loss": 0.9399,
+      "step": 18380
+    },
+    {
+      "epoch": 0.5539156626506024,
+      "grad_norm": 5.156381182005516,
+      "learning_rate": 4.522616525602789e-06,
+      "loss": 0.9347,
+      "step": 18390
+    },
+    {
+      "epoch": 0.5542168674698795,
+      "grad_norm": 1.9441536621399538,
+      "learning_rate": 4.517659184280651e-06,
+      "loss": 0.8091,
+      "step": 18400
+    },
+    {
+      "epoch": 0.5545180722891566,
+      "grad_norm": 3.68109023439261,
+      "learning_rate": 4.512702321511845e-06,
+      "loss": 0.9961,
+      "step": 18410
+    },
+    {
+      "epoch": 0.5548192771084337,
+      "grad_norm": 1.922771806218246,
+      "learning_rate": 4.507745942214311e-06,
+      "loss": 0.836,
+      "step": 18420
+    },
+    {
+      "epoch": 0.5551204819277108,
+      "grad_norm": 2.0451128315073124,
+      "learning_rate": 4.502790051305512e-06,
+      "loss": 0.8536,
+      "step": 18430
+    },
+    {
+      "epoch": 0.555421686746988,
+      "grad_norm": 2.0886689680083013,
+      "learning_rate": 4.497834653702422e-06,
+      "loss": 0.9008,
+      "step": 18440
+    },
+    {
+      "epoch": 0.5557228915662651,
+      "grad_norm": 4.115499394418487,
+      "learning_rate": 4.492879754321525e-06,
+      "loss": 0.9723,
+      "step": 18450
+    },
+    {
+      "epoch": 0.5560240963855422,
+      "grad_norm": 1.9508285900298126,
+      "learning_rate": 4.487925358078817e-06,
+      "loss": 0.8829,
+      "step": 18460
+    },
+    {
+      "epoch": 0.5563253012048193,
+      "grad_norm": 1.9166029511192284,
+      "learning_rate": 4.4829714698897885e-06,
+      "loss": 0.9262,
+      "step": 18470
+    },
+    {
+      "epoch": 0.5566265060240964,
+      "grad_norm": 4.334708285468265,
+      "learning_rate": 4.478018094669429e-06,
+      "loss": 1.0462,
+      "step": 18480
+    },
+    {
+      "epoch": 0.5569277108433734,
+      "grad_norm": 6.100242190534848,
+      "learning_rate": 4.473065237332217e-06,
+      "loss": 1.0086,
+      "step": 18490
+    },
+    {
+      "epoch": 0.5572289156626506,
+      "grad_norm": 1.8737237748603184,
+      "learning_rate": 4.468112902792121e-06,
+      "loss": 1.0117,
+      "step": 18500
+    },
+    {
+      "epoch": 0.5575301204819277,
+      "grad_norm": 1.9908048867633226,
+      "learning_rate": 4.4631610959625885e-06,
+      "loss": 1.01,
+      "step": 18510
+    },
+    {
+      "epoch": 0.5578313253012048,
+      "grad_norm": 2.067508599090194,
+      "learning_rate": 4.4582098217565375e-06,
+      "loss": 0.946,
+      "step": 18520
+    },
+    {
+      "epoch": 0.5581325301204819,
+      "grad_norm": 2.040770869396424,
+      "learning_rate": 4.453259085086371e-06,
+      "loss": 0.9013,
+      "step": 18530
+    },
+    {
+      "epoch": 0.558433734939759,
+      "grad_norm": 4.935813901130578,
+      "learning_rate": 4.4483088908639475e-06,
+      "loss": 0.8959,
+      "step": 18540
+    },
+    {
+      "epoch": 0.5587349397590361,
+      "grad_norm": 4.236655298417486,
+      "learning_rate": 4.44335924400059e-06,
+      "loss": 0.9883,
+      "step": 18550
+    },
+    {
+      "epoch": 0.5590361445783133,
+      "grad_norm": 4.283614854202966,
+      "learning_rate": 4.43841014940708e-06,
+      "loss": 1.0008,
+      "step": 18560
+    },
+    {
+      "epoch": 0.5593373493975904,
+      "grad_norm": 4.698810672546298,
+      "learning_rate": 4.4334616119936516e-06,
+      "loss": 1.0652,
+      "step": 18570
+    },
+    {
+      "epoch": 0.5596385542168675,
+      "grad_norm": 6.206177154210965,
+      "learning_rate": 4.4285136366699835e-06,
+      "loss": 0.8549,
+      "step": 18580
+    },
+    {
+      "epoch": 0.5599397590361446,
+      "grad_norm": 1.8802465834282318,
+      "learning_rate": 4.4235662283451975e-06,
+      "loss": 0.9175,
+      "step": 18590
+    },
+    {
+      "epoch": 0.5602409638554217,
+      "grad_norm": 3.9318789018894313,
+      "learning_rate": 4.418619391927854e-06,
+      "loss": 0.9142,
+      "step": 18600
+    },
+    {
+      "epoch": 0.5605421686746987,
+      "grad_norm": 3.9094542093330675,
+      "learning_rate": 4.4136731323259494e-06,
+      "loss": 0.86,
+      "step": 18610
+    },
+    {
+      "epoch": 0.560843373493976,
+      "grad_norm": 3.836422362284994,
+      "learning_rate": 4.408727454446897e-06,
+      "loss": 0.8367,
+      "step": 18620
+    },
+    {
+      "epoch": 0.561144578313253,
+      "grad_norm": 4.114832915900658,
+      "learning_rate": 4.403782363197543e-06,
+      "loss": 0.9804,
+      "step": 18630
+    },
+    {
+      "epoch": 0.5614457831325301,
+      "grad_norm": 5.596184527159677,
+      "learning_rate": 4.398837863484149e-06,
+      "loss": 0.95,
+      "step": 18640
+    },
+    {
+      "epoch": 0.5617469879518072,
+      "grad_norm": 1.9534028234204528,
+      "learning_rate": 4.393893960212387e-06,
+      "loss": 0.9503,
+      "step": 18650
+    },
+    {
+      "epoch": 0.5620481927710843,
+      "grad_norm": 3.7464535944102817,
+      "learning_rate": 4.388950658287343e-06,
+      "loss": 0.8979,
+      "step": 18660
+    },
+    {
+      "epoch": 0.5623493975903614,
+      "grad_norm": 2.111862104385048,
+      "learning_rate": 4.384007962613496e-06,
+      "loss": 0.9893,
+      "step": 18670
+    },
+    {
+      "epoch": 0.5626506024096386,
+      "grad_norm": 3.579984988879612,
+      "learning_rate": 4.379065878094738e-06,
+      "loss": 0.9144,
+      "step": 18680
+    },
+    {
+      "epoch": 0.5629518072289157,
+      "grad_norm": 3.8163492778300805,
+      "learning_rate": 4.374124409634342e-06,
+      "loss": 0.9661,
+      "step": 18690
+    },
+    {
+      "epoch": 0.5632530120481928,
+      "grad_norm": 3.7980937513766158,
+      "learning_rate": 4.369183562134975e-06,
+      "loss": 1.0123,
+      "step": 18700
+    },
+    {
+      "epoch": 0.5635542168674699,
+      "grad_norm": 4.803940294654779,
+      "learning_rate": 4.364243340498689e-06,
+      "loss": 1.025,
+      "step": 18710
+    },
+    {
+      "epoch": 0.563855421686747,
+      "grad_norm": 2.1392657146156475,
+      "learning_rate": 4.359303749626913e-06,
+      "loss": 0.9972,
+      "step": 18720
+    },
+    {
+      "epoch": 0.564156626506024,
+      "grad_norm": 4.637960149146857,
+      "learning_rate": 4.3543647944204495e-06,
+      "loss": 0.979,
+      "step": 18730
+    },
+    {
+      "epoch": 0.5644578313253013,
+      "grad_norm": 3.770254882330825,
+      "learning_rate": 4.349426479779472e-06,
+      "loss": 0.9557,
+      "step": 18740
+    },
+    {
+      "epoch": 0.5647590361445783,
+      "grad_norm": 4.127228391502067,
+      "learning_rate": 4.34448881060352e-06,
+      "loss": 0.9658,
+      "step": 18750
+    },
+    {
+      "epoch": 0.5650602409638554,
+      "grad_norm": 4.262381967244327,
+      "learning_rate": 4.33955179179149e-06,
+      "loss": 1.0109,
+      "step": 18760
+    },
+    {
+      "epoch": 0.5653614457831325,
+      "grad_norm": 3.4941324642398746,
+      "learning_rate": 4.334615428241629e-06,
+      "loss": 0.9235,
+      "step": 18770
+    },
+    {
+      "epoch": 0.5656626506024096,
+      "grad_norm": 5.75497366241385,
+      "learning_rate": 4.329679724851548e-06,
+      "loss": 1.0059,
+      "step": 18780
+    },
+    {
+      "epoch": 0.5659638554216867,
+      "grad_norm": 3.7544708516209657,
+      "learning_rate": 4.324744686518187e-06,
+      "loss": 0.8923,
+      "step": 18790
+    },
+    {
+      "epoch": 0.5662650602409639,
+      "grad_norm": 3.5842199961136165,
+      "learning_rate": 4.3198103181378335e-06,
+      "loss": 1.0152,
+      "step": 18800
+    },
+    {
+      "epoch": 0.566566265060241,
+      "grad_norm": 3.7537876702822475,
+      "learning_rate": 4.31487662460611e-06,
+      "loss": 1.0748,
+      "step": 18810
+    },
+    {
+      "epoch": 0.5668674698795181,
+      "grad_norm": 1.9374060537841604,
+      "learning_rate": 4.309943610817972e-06,
+      "loss": 0.9367,
+      "step": 18820
+    },
+    {
+      "epoch": 0.5671686746987952,
+      "grad_norm": 4.621492949057501,
+      "learning_rate": 4.305011281667697e-06,
+      "loss": 0.8945,
+      "step": 18830
+    },
+    {
+      "epoch": 0.5674698795180723,
+      "grad_norm": 3.9082665610282983,
+      "learning_rate": 4.300079642048878e-06,
+      "loss": 0.9052,
+      "step": 18840
+    },
+    {
+      "epoch": 0.5677710843373494,
+      "grad_norm": 4.924352913638754,
+      "learning_rate": 4.29514869685444e-06,
+      "loss": 0.9619,
+      "step": 18850
+    },
+    {
+      "epoch": 0.5680722891566266,
+      "grad_norm": 3.704516720818371,
+      "learning_rate": 4.290218450976602e-06,
+      "loss": 0.97,
+      "step": 18860
+    },
+    {
+      "epoch": 0.5683734939759036,
+      "grad_norm": 2.089249845955756,
+      "learning_rate": 4.285288909306898e-06,
+      "loss": 0.8702,
+      "step": 18870
+    },
+    {
+      "epoch": 0.5686746987951807,
+      "grad_norm": 20.96076288468202,
+      "learning_rate": 4.28036007673616e-06,
+      "loss": 0.951,
+      "step": 18880
+    },
+    {
+      "epoch": 0.5689759036144578,
+      "grad_norm": 4.633173693165347,
+      "learning_rate": 4.27543195815452e-06,
+      "loss": 0.9514,
+      "step": 18890
+    },
+    {
+      "epoch": 0.5692771084337349,
+      "grad_norm": 4.299784221202882,
+      "learning_rate": 4.270504558451399e-06,
+      "loss": 0.8395,
+      "step": 18900
+    },
+    {
+      "epoch": 0.569578313253012,
+      "grad_norm": 1.9314982054027046,
+      "learning_rate": 4.265577882515506e-06,
+      "loss": 0.8127,
+      "step": 18910
+    },
+    {
+      "epoch": 0.5698795180722892,
+      "grad_norm": 4.698112498219021,
+      "learning_rate": 4.2606519352348276e-06,
+      "loss": 0.948,
+      "step": 18920
+    },
+    {
+      "epoch": 0.5701807228915663,
+      "grad_norm": 6.041828060365895,
+      "learning_rate": 4.255726721496637e-06,
+      "loss": 0.9664,
+      "step": 18930
+    },
+    {
+      "epoch": 0.5704819277108434,
+      "grad_norm": 5.2524601599322605,
+      "learning_rate": 4.2508022461874695e-06,
+      "loss": 0.9306,
+      "step": 18940
+    },
+    {
+      "epoch": 0.5707831325301205,
+      "grad_norm": 1.9830774556820767,
+      "learning_rate": 4.245878514193131e-06,
+      "loss": 0.9587,
+      "step": 18950
+    },
+    {
+      "epoch": 0.5710843373493976,
+      "grad_norm": 4.384318913526011,
+      "learning_rate": 4.240955530398695e-06,
+      "loss": 0.8796,
+      "step": 18960
+    },
+    {
+      "epoch": 0.5713855421686747,
+      "grad_norm": 4.287595994126813,
+      "learning_rate": 4.236033299688486e-06,
+      "loss": 0.8963,
+      "step": 18970
+    },
+    {
+      "epoch": 0.5716867469879519,
+      "grad_norm": 7.136055525295403,
+      "learning_rate": 4.2311118269460845e-06,
+      "loss": 1.0356,
+      "step": 18980
+    },
+    {
+      "epoch": 0.571987951807229,
+      "grad_norm": 4.962219512409365,
+      "learning_rate": 4.226191117054316e-06,
+      "loss": 0.9524,
+      "step": 18990
+    },
+    {
+      "epoch": 0.572289156626506,
+      "grad_norm": 2.0172963730938362,
+      "learning_rate": 4.2212711748952575e-06,
+      "loss": 0.9047,
+      "step": 19000
+    },
+    {
+      "epoch": 0.5725903614457831,
+      "grad_norm": 2.077594005170364,
+      "learning_rate": 4.216352005350212e-06,
+      "loss": 0.9941,
+      "step": 19010
+    },
+    {
+      "epoch": 0.5728915662650602,
+      "grad_norm": 4.217654263507011,
+      "learning_rate": 4.211433613299725e-06,
+      "loss": 0.8907,
+      "step": 19020
+    },
+    {
+      "epoch": 0.5731927710843373,
+      "grad_norm": 6.467371706436814,
+      "learning_rate": 4.206516003623568e-06,
+      "loss": 1.0311,
+      "step": 19030
+    },
+    {
+      "epoch": 0.5734939759036145,
+      "grad_norm": 4.6591531707198905,
+      "learning_rate": 4.201599181200736e-06,
+      "loss": 0.8733,
+      "step": 19040
+    },
+    {
+      "epoch": 0.5737951807228916,
+      "grad_norm": 2.0132631749228063,
+      "learning_rate": 4.196683150909443e-06,
+      "loss": 0.9221,
+      "step": 19050
+    },
+    {
+      "epoch": 0.5740963855421687,
+      "grad_norm": 1.880468359992078,
+      "learning_rate": 4.191767917627115e-06,
+      "loss": 0.8268,
+      "step": 19060
+    },
+    {
+      "epoch": 0.5743975903614458,
+      "grad_norm": 5.970993317806257,
+      "learning_rate": 4.1868534862303935e-06,
+      "loss": 1.0026,
+      "step": 19070
+    },
+    {
+      "epoch": 0.5746987951807229,
+      "grad_norm": 1.850177492612431,
+      "learning_rate": 4.1819398615951205e-06,
+      "loss": 0.8882,
+      "step": 19080
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 5.090666853571762,
+      "learning_rate": 4.17702704859633e-06,
+      "loss": 0.8254,
+      "step": 19090
+    },
+    {
+      "epoch": 0.5753012048192772,
+      "grad_norm": 4.473504116706102,
+      "learning_rate": 4.172115052108268e-06,
+      "loss": 0.9926,
+      "step": 19100
+    },
+    {
+      "epoch": 0.5756024096385542,
+      "grad_norm": 6.458336484469976,
+      "learning_rate": 4.167203877004353e-06,
+      "loss": 0.8647,
+      "step": 19110
+    },
+    {
+      "epoch": 0.5759036144578313,
+      "grad_norm": 3.926443743927527,
+      "learning_rate": 4.162293528157201e-06,
+      "loss": 0.9696,
+      "step": 19120
+    },
+    {
+      "epoch": 0.5762048192771084,
+      "grad_norm": 4.123870508627569,
+      "learning_rate": 4.157384010438598e-06,
+      "loss": 0.8732,
+      "step": 19130
+    },
+    {
+      "epoch": 0.5765060240963855,
+      "grad_norm": 5.650853444971126,
+      "learning_rate": 4.152475328719517e-06,
+      "loss": 0.9049,
+      "step": 19140
+    },
+    {
+      "epoch": 0.5768072289156626,
+      "grad_norm": 4.038437621321253,
+      "learning_rate": 4.147567487870092e-06,
+      "loss": 0.9814,
+      "step": 19150
+    },
+    {
+      "epoch": 0.5771084337349398,
+      "grad_norm": 2.095143797451166,
+      "learning_rate": 4.142660492759627e-06,
+      "loss": 0.8653,
+      "step": 19160
+    },
+    {
+      "epoch": 0.5774096385542169,
+      "grad_norm": 9.477101458372529,
+      "learning_rate": 4.137754348256582e-06,
+      "loss": 0.9629,
+      "step": 19170
+    },
+    {
+      "epoch": 0.577710843373494,
+      "grad_norm": 4.636146793154455,
+      "learning_rate": 4.1328490592285835e-06,
+      "loss": 0.9698,
+      "step": 19180
+    },
+    {
+      "epoch": 0.5780120481927711,
+      "grad_norm": 3.966840597990876,
+      "learning_rate": 4.127944630542399e-06,
+      "loss": 0.9285,
+      "step": 19190
+    },
+    {
+      "epoch": 0.5783132530120482,
+      "grad_norm": 3.870866783235835,
+      "learning_rate": 4.123041067063945e-06,
+      "loss": 0.982,
+      "step": 19200
+    },
+    {
+      "epoch": 0.5786144578313253,
+      "grad_norm": 3.9939869149013063,
+      "learning_rate": 4.118138373658283e-06,
+      "loss": 0.9778,
+      "step": 19210
+    },
+    {
+      "epoch": 0.5789156626506025,
+      "grad_norm": 4.736026530445284,
+      "learning_rate": 4.1132365551896104e-06,
+      "loss": 0.9062,
+      "step": 19220
+    },
+    {
+      "epoch": 0.5792168674698795,
+      "grad_norm": 4.063967782259461,
+      "learning_rate": 4.108335616521253e-06,
+      "loss": 1.0021,
+      "step": 19230
+    },
+    {
+      "epoch": 0.5795180722891566,
+      "grad_norm": 8.251032895747358,
+      "learning_rate": 4.103435562515662e-06,
+      "loss": 0.9911,
+      "step": 19240
+    },
+    {
+      "epoch": 0.5798192771084337,
+      "grad_norm": 4.472907905804807,
+      "learning_rate": 4.098536398034423e-06,
+      "loss": 0.9788,
+      "step": 19250
+    },
+    {
+      "epoch": 0.5801204819277108,
+      "grad_norm": 5.11668912041138,
+      "learning_rate": 4.093638127938225e-06,
+      "loss": 0.9856,
+      "step": 19260
+    },
+    {
+      "epoch": 0.5804216867469879,
+      "grad_norm": 2.0103905517811356,
+      "learning_rate": 4.088740757086877e-06,
+      "loss": 0.8786,
+      "step": 19270
+    },
+    {
+      "epoch": 0.5807228915662651,
+      "grad_norm": 6.724249971609902,
+      "learning_rate": 4.0838442903392936e-06,
+      "loss": 0.98,
+      "step": 19280
+    },
+    {
+      "epoch": 0.5810240963855422,
+      "grad_norm": 1.8756102514139061,
+      "learning_rate": 4.078948732553495e-06,
+      "loss": 0.8569,
+      "step": 19290
+    },
+    {
+      "epoch": 0.5813253012048193,
+      "grad_norm": 2.018430839661244,
+      "learning_rate": 4.074054088586596e-06,
+      "loss": 0.9565,
+      "step": 19300
+    },
+    {
+      "epoch": 0.5816265060240964,
+      "grad_norm": 2.063677566455654,
+      "learning_rate": 4.069160363294805e-06,
+      "loss": 1.0176,
+      "step": 19310
+    },
+    {
+      "epoch": 0.5819277108433735,
+      "grad_norm": 5.305992921836024,
+      "learning_rate": 4.064267561533423e-06,
+      "loss": 0.9794,
+      "step": 19320
+    },
+    {
+      "epoch": 0.5822289156626506,
+      "grad_norm": 5.892546927954209,
+      "learning_rate": 4.059375688156833e-06,
+      "loss": 0.8623,
+      "step": 19330
+    },
+    {
+      "epoch": 0.5825301204819278,
+      "grad_norm": 2.0344855813127674,
+      "learning_rate": 4.05448474801849e-06,
+      "loss": 0.8925,
+      "step": 19340
+    },
+    {
+      "epoch": 0.5828313253012049,
+      "grad_norm": 4.281680194569213,
+      "learning_rate": 4.049594745970935e-06,
+      "loss": 0.8764,
+      "step": 19350
+    },
+    {
+      "epoch": 0.5831325301204819,
+      "grad_norm": 1.984358877697064,
+      "learning_rate": 4.044705686865771e-06,
+      "loss": 0.957,
+      "step": 19360
+    },
+    {
+      "epoch": 0.583433734939759,
+      "grad_norm": 5.077279148504344,
+      "learning_rate": 4.039817575553665e-06,
+      "loss": 0.9986,
+      "step": 19370
+    },
+    {
+      "epoch": 0.5837349397590361,
+      "grad_norm": 3.8524178742577386,
+      "learning_rate": 4.034930416884346e-06,
+      "loss": 1.0043,
+      "step": 19380
+    },
+    {
+      "epoch": 0.5840361445783132,
+      "grad_norm": 5.127115601818105,
+      "learning_rate": 4.030044215706599e-06,
+      "loss": 0.9809,
+      "step": 19390
+    },
+    {
+      "epoch": 0.5843373493975904,
+      "grad_norm": 5.1176030232461125,
+      "learning_rate": 4.025158976868256e-06,
+      "loss": 0.9775,
+      "step": 19400
+    },
+    {
+      "epoch": 0.5846385542168675,
+      "grad_norm": 1.9987359372376363,
+      "learning_rate": 4.020274705216195e-06,
+      "loss": 0.9231,
+      "step": 19410
+    },
+    {
+      "epoch": 0.5849397590361446,
+      "grad_norm": 4.5078001613848535,
+      "learning_rate": 4.015391405596334e-06,
+      "loss": 0.9954,
+      "step": 19420
+    },
+    {
+      "epoch": 0.5852409638554217,
+      "grad_norm": 4.611302971350435,
+      "learning_rate": 4.010509082853629e-06,
+      "loss": 1.0505,
+      "step": 19430
+    },
+    {
+      "epoch": 0.5855421686746988,
+      "grad_norm": 5.544955522529904,
+      "learning_rate": 4.0056277418320675e-06,
+      "loss": 0.9018,
+      "step": 19440
+    },
+    {
+      "epoch": 0.5858433734939759,
+      "grad_norm": 2.0730805611868024,
+      "learning_rate": 4.000747387374655e-06,
+      "loss": 0.8096,
+      "step": 19450
+    },
+    {
+      "epoch": 0.5861445783132531,
+      "grad_norm": 3.2711034213875223,
+      "learning_rate": 3.9958680243234296e-06,
+      "loss": 0.9079,
+      "step": 19460
+    },
+    {
+      "epoch": 0.5864457831325302,
+      "grad_norm": 2.0374152947334303,
+      "learning_rate": 3.990989657519438e-06,
+      "loss": 0.9764,
+      "step": 19470
+    },
+    {
+      "epoch": 0.5867469879518072,
+      "grad_norm": 1.8549051826050869,
+      "learning_rate": 3.986112291802742e-06,
+      "loss": 0.9459,
+      "step": 19480
+    },
+    {
+      "epoch": 0.5870481927710843,
+      "grad_norm": 9.582038381432506,
+      "learning_rate": 3.9812359320124035e-06,
+      "loss": 0.8867,
+      "step": 19490
+    },
+    {
+      "epoch": 0.5873493975903614,
+      "grad_norm": 4.144057842139621,
+      "learning_rate": 3.976360582986498e-06,
+      "loss": 0.9371,
+      "step": 19500
+    },
+    {
+      "epoch": 0.5876506024096385,
+      "grad_norm": 4.375544782740267,
+      "learning_rate": 3.97148624956209e-06,
+      "loss": 0.8329,
+      "step": 19510
+    },
+    {
+      "epoch": 0.5879518072289157,
+      "grad_norm": 4.780092041284829,
+      "learning_rate": 3.966612936575235e-06,
+      "loss": 0.8714,
+      "step": 19520
+    },
+    {
+      "epoch": 0.5882530120481928,
+      "grad_norm": 4.38604775170126,
+      "learning_rate": 3.961740648860984e-06,
+      "loss": 0.9381,
+      "step": 19530
+    },
+    {
+      "epoch": 0.5885542168674699,
+      "grad_norm": 4.404034979998394,
+      "learning_rate": 3.9568693912533634e-06,
+      "loss": 0.8795,
+      "step": 19540
+    },
+    {
+      "epoch": 0.588855421686747,
+      "grad_norm": 3.8054401865690033,
+      "learning_rate": 3.951999168585382e-06,
+      "loss": 1.0025,
+      "step": 19550
+    },
+    {
+      "epoch": 0.5891566265060241,
+      "grad_norm": 4.564835777626075,
+      "learning_rate": 3.947129985689015e-06,
+      "loss": 0.8826,
+      "step": 19560
+    },
+    {
+      "epoch": 0.5894578313253012,
+      "grad_norm": 4.05307633524432,
+      "learning_rate": 3.942261847395219e-06,
+      "loss": 0.9938,
+      "step": 19570
+    },
+    {
+      "epoch": 0.5897590361445784,
+      "grad_norm": 3.9612524260890796,
+      "learning_rate": 3.9373947585339006e-06,
+      "loss": 0.8985,
+      "step": 19580
+    },
+    {
+      "epoch": 0.5900602409638555,
+      "grad_norm": 4.269101592949929,
+      "learning_rate": 3.932528723933931e-06,
+      "loss": 0.9395,
+      "step": 19590
+    },
+    {
+      "epoch": 0.5903614457831325,
+      "grad_norm": 2.091604360713433,
+      "learning_rate": 3.927663748423139e-06,
+      "loss": 0.9688,
+      "step": 19600
+    },
+    {
+      "epoch": 0.5906626506024096,
+      "grad_norm": 3.947046091444418,
+      "learning_rate": 3.922799836828295e-06,
+      "loss": 0.8924,
+      "step": 19610
+    },
+    {
+      "epoch": 0.5909638554216867,
+      "grad_norm": 2.0258762535317145,
+      "learning_rate": 3.917936993975119e-06,
+      "loss": 0.9684,
+      "step": 19620
+    },
+    {
+      "epoch": 0.5912650602409638,
+      "grad_norm": 6.170474460728932,
+      "learning_rate": 3.913075224688268e-06,
+      "loss": 1.0269,
+      "step": 19630
+    },
+    {
+      "epoch": 0.591566265060241,
+      "grad_norm": 3.4139801657324136,
+      "learning_rate": 3.90821453379134e-06,
+      "loss": 0.9085,
+      "step": 19640
+    },
+    {
+      "epoch": 0.5918674698795181,
+      "grad_norm": 2.0513524319317438,
+      "learning_rate": 3.903354926106855e-06,
+      "loss": 0.8842,
+      "step": 19650
+    },
+    {
+      "epoch": 0.5921686746987952,
+      "grad_norm": 3.4406615044779416,
+      "learning_rate": 3.8984964064562584e-06,
+      "loss": 0.8554,
+      "step": 19660
+    },
+    {
+      "epoch": 0.5924698795180723,
+      "grad_norm": 2.129923436331325,
+      "learning_rate": 3.893638979659925e-06,
+      "loss": 0.8639,
+      "step": 19670
+    },
+    {
+      "epoch": 0.5927710843373494,
+      "grad_norm": 4.864839862012951,
+      "learning_rate": 3.888782650537137e-06,
+      "loss": 1.0265,
+      "step": 19680
+    },
+    {
+      "epoch": 0.5930722891566265,
+      "grad_norm": 2.046001811984436,
+      "learning_rate": 3.883927423906091e-06,
+      "loss": 0.9386,
+      "step": 19690
+    },
+    {
+      "epoch": 0.5933734939759037,
+      "grad_norm": 4.2082964948951,
+      "learning_rate": 3.8790733045838865e-06,
+      "loss": 0.8413,
+      "step": 19700
+    },
+    {
+      "epoch": 0.5936746987951808,
+      "grad_norm": 1.875959934560138,
+      "learning_rate": 3.87422029738653e-06,
+      "loss": 0.835,
+      "step": 19710
+    },
+    {
+      "epoch": 0.5939759036144578,
+      "grad_norm": 4.572085365663675,
+      "learning_rate": 3.869368407128922e-06,
+      "loss": 0.9211,
+      "step": 19720
+    },
+    {
+      "epoch": 0.5942771084337349,
+      "grad_norm": 3.7668194982300194,
+      "learning_rate": 3.86451763862485e-06,
+      "loss": 1.0513,
+      "step": 19730
+    },
+    {
+      "epoch": 0.594578313253012,
+      "grad_norm": 4.738725122704846,
+      "learning_rate": 3.8596679966869946e-06,
+      "loss": 0.8918,
+      "step": 19740
+    },
+    {
+      "epoch": 0.5948795180722891,
+      "grad_norm": 2.096804706431497,
+      "learning_rate": 3.854819486126919e-06,
+      "loss": 0.9515,
+      "step": 19750
+    },
+    {
+      "epoch": 0.5951807228915663,
+      "grad_norm": 2.02920731379469,
+      "learning_rate": 3.84997211175506e-06,
+      "loss": 0.9241,
+      "step": 19760
+    },
+    {
+      "epoch": 0.5954819277108434,
+      "grad_norm": 1.9833211657330019,
+      "learning_rate": 3.8451258783807285e-06,
+      "loss": 0.8581,
+      "step": 19770
+    },
+    {
+      "epoch": 0.5957831325301205,
+      "grad_norm": 14.198194598438121,
+      "learning_rate": 3.840280790812107e-06,
+      "loss": 0.9752,
+      "step": 19780
+    },
+    {
+      "epoch": 0.5960843373493976,
+      "grad_norm": 5.687403900320029,
+      "learning_rate": 3.835436853856234e-06,
+      "loss": 0.9123,
+      "step": 19790
+    },
+    {
+      "epoch": 0.5963855421686747,
+      "grad_norm": 5.157563081151716,
+      "learning_rate": 3.830594072319014e-06,
+      "loss": 0.8172,
+      "step": 19800
+    },
+    {
+      "epoch": 0.5966867469879518,
+      "grad_norm": 5.654074952590619,
+      "learning_rate": 3.825752451005196e-06,
+      "loss": 0.9453,
+      "step": 19810
+    },
+    {
+      "epoch": 0.596987951807229,
+      "grad_norm": 3.956724339976459,
+      "learning_rate": 3.82091199471839e-06,
+      "loss": 0.8837,
+      "step": 19820
+    },
+    {
+      "epoch": 0.5972891566265061,
+      "grad_norm": 4.089607905283346,
+      "learning_rate": 3.816072708261041e-06,
+      "loss": 0.8458,
+      "step": 19830
+    },
+    {
+      "epoch": 0.5975903614457831,
+      "grad_norm": 6.045610661537695,
+      "learning_rate": 3.811234596434433e-06,
+      "loss": 0.9642,
+      "step": 19840
+    },
+    {
+      "epoch": 0.5978915662650602,
+      "grad_norm": 2.066210435504891,
+      "learning_rate": 3.8063976640386926e-06,
+      "loss": 0.9222,
+      "step": 19850
+    },
+    {
+      "epoch": 0.5981927710843373,
+      "grad_norm": 6.38183896078686,
+      "learning_rate": 3.801561915872768e-06,
+      "loss": 1.0151,
+      "step": 19860
+    },
+    {
+      "epoch": 0.5984939759036144,
+      "grad_norm": 4.832405016095391,
+      "learning_rate": 3.7967273567344374e-06,
+      "loss": 0.9098,
+      "step": 19870
+    },
+    {
+      "epoch": 0.5987951807228916,
+      "grad_norm": 3.857923639667753,
+      "learning_rate": 3.791893991420294e-06,
+      "loss": 0.9968,
+      "step": 19880
+    },
+    {
+      "epoch": 0.5990963855421687,
+      "grad_norm": 11.921100115865212,
+      "learning_rate": 3.7870618247257558e-06,
+      "loss": 0.9933,
+      "step": 19890
+    },
+    {
+      "epoch": 0.5993975903614458,
+      "grad_norm": 3.493248593589962,
+      "learning_rate": 3.782230861445041e-06,
+      "loss": 0.9846,
+      "step": 19900
+    },
+    {
+      "epoch": 0.5996987951807229,
+      "grad_norm": 2.015379592753621,
+      "learning_rate": 3.77740110637118e-06,
+      "loss": 0.8482,
+      "step": 19910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.000139425927677,
+      "learning_rate": 3.7725725642960047e-06,
+      "loss": 0.8741,
+      "step": 19920
+    },
+    {
+      "epoch": 0.6003012048192771,
+      "grad_norm": 2.172752870145956,
+      "learning_rate": 3.7677452400101413e-06,
+      "loss": 0.9144,
+      "step": 19930
+    },
+    {
+      "epoch": 0.6006024096385543,
+      "grad_norm": 4.668322278976479,
+      "learning_rate": 3.762919138303008e-06,
+      "loss": 0.9732,
+      "step": 19940
+    },
+    {
+      "epoch": 0.6009036144578314,
+      "grad_norm": 5.782784828200442,
+      "learning_rate": 3.7580942639628093e-06,
+      "loss": 0.9736,
+      "step": 19950
+    },
+    {
+      "epoch": 0.6012048192771084,
+      "grad_norm": 5.371808547098476,
+      "learning_rate": 3.753270621776536e-06,
+      "loss": 0.8671,
+      "step": 19960
+    },
+    {
+      "epoch": 0.6015060240963855,
+      "grad_norm": 5.290640387085528,
+      "learning_rate": 3.748448216529953e-06,
+      "loss": 0.8577,
+      "step": 19970
+    },
+    {
+      "epoch": 0.6018072289156626,
+      "grad_norm": 3.353020582675196,
+      "learning_rate": 3.7436270530075973e-06,
+      "loss": 0.9605,
+      "step": 19980
+    },
+    {
+      "epoch": 0.6021084337349397,
+      "grad_norm": 1.9049160341811593,
+      "learning_rate": 3.7388071359927745e-06,
+      "loss": 0.9228,
+      "step": 19990
+    },
+    {
+      "epoch": 0.6024096385542169,
+      "grad_norm": 2.0013597522662403,
+      "learning_rate": 3.7339884702675566e-06,
+      "loss": 0.9478,
+      "step": 20000
+    },
+    {
+      "epoch": 0.602710843373494,
+      "grad_norm": 4.185587185278027,
+      "learning_rate": 3.7291710606127706e-06,
+      "loss": 1.0358,
+      "step": 20010
+    },
+    {
+      "epoch": 0.6030120481927711,
+      "grad_norm": 5.135797612131306,
+      "learning_rate": 3.7243549118079973e-06,
+      "loss": 0.8667,
+      "step": 20020
+    },
+    {
+      "epoch": 0.6033132530120482,
+      "grad_norm": 4.792552611251369,
+      "learning_rate": 3.71954002863157e-06,
+      "loss": 0.862,
+      "step": 20030
+    },
+    {
+      "epoch": 0.6036144578313253,
+      "grad_norm": 13.706315325819853,
+      "learning_rate": 3.714726415860564e-06,
+      "loss": 0.8559,
+      "step": 20040
+    },
+    {
+      "epoch": 0.6039156626506024,
+      "grad_norm": 1.9506257104378237,
+      "learning_rate": 3.7099140782707916e-06,
+      "loss": 0.7836,
+      "step": 20050
+    },
+    {
+      "epoch": 0.6042168674698796,
+      "grad_norm": 5.40062128101387,
+      "learning_rate": 3.7051030206368026e-06,
+      "loss": 0.9976,
+      "step": 20060
+    },
+    {
+      "epoch": 0.6045180722891567,
+      "grad_norm": 4.688678898598641,
+      "learning_rate": 3.70029324773188e-06,
+      "loss": 0.9433,
+      "step": 20070
+    },
+    {
+      "epoch": 0.6048192771084338,
+      "grad_norm": 4.427468895008978,
+      "learning_rate": 3.6954847643280256e-06,
+      "loss": 0.9371,
+      "step": 20080
+    },
+    {
+      "epoch": 0.6051204819277108,
+      "grad_norm": 4.190971723261509,
+      "learning_rate": 3.6906775751959667e-06,
+      "loss": 0.9586,
+      "step": 20090
+    },
+    {
+      "epoch": 0.6054216867469879,
+      "grad_norm": 4.474464589994147,
+      "learning_rate": 3.6858716851051446e-06,
+      "loss": 0.8448,
+      "step": 20100
+    },
+    {
+      "epoch": 0.605722891566265,
+      "grad_norm": 4.959125383185958,
+      "learning_rate": 3.6810670988237134e-06,
+      "loss": 1.0073,
+      "step": 20110
+    },
+    {
+      "epoch": 0.6060240963855422,
+      "grad_norm": 1.9839334817570982,
+      "learning_rate": 3.676263821118532e-06,
+      "loss": 0.9316,
+      "step": 20120
+    },
+    {
+      "epoch": 0.6063253012048193,
+      "grad_norm": 4.1999695178417955,
+      "learning_rate": 3.6714618567551574e-06,
+      "loss": 1.0083,
+      "step": 20130
+    },
+    {
+      "epoch": 0.6066265060240964,
+      "grad_norm": 4.495331797883539,
+      "learning_rate": 3.666661210497854e-06,
+      "loss": 0.94,
+      "step": 20140
+    },
+    {
+      "epoch": 0.6069277108433735,
+      "grad_norm": 1.9935464448559828,
+      "learning_rate": 3.6618618871095675e-06,
+      "loss": 0.9016,
+      "step": 20150
+    },
+    {
+      "epoch": 0.6072289156626506,
+      "grad_norm": 2.02391953136053,
+      "learning_rate": 3.657063891351935e-06,
+      "loss": 0.9214,
+      "step": 20160
+    },
+    {
+      "epoch": 0.6075301204819277,
+      "grad_norm": 2.0226333520760065,
+      "learning_rate": 3.652267227985281e-06,
+      "loss": 0.9001,
+      "step": 20170
+    },
+    {
+      "epoch": 0.6078313253012049,
+      "grad_norm": 5.376248438595198,
+      "learning_rate": 3.647471901768601e-06,
+      "loss": 0.894,
+      "step": 20180
+    },
+    {
+      "epoch": 0.608132530120482,
+      "grad_norm": 4.628581474920254,
+      "learning_rate": 3.6426779174595693e-06,
+      "loss": 1.0381,
+      "step": 20190
+    },
+    {
+      "epoch": 0.608433734939759,
+      "grad_norm": 13.161008381782125,
+      "learning_rate": 3.637885279814521e-06,
+      "loss": 0.7851,
+      "step": 20200
+    },
+    {
+      "epoch": 0.6087349397590361,
+      "grad_norm": 4.633153878072175,
+      "learning_rate": 3.633093993588468e-06,
+      "loss": 0.9597,
+      "step": 20210
+    },
+    {
+      "epoch": 0.6090361445783132,
+      "grad_norm": 5.207531496874266,
+      "learning_rate": 3.6283040635350687e-06,
+      "loss": 0.9143,
+      "step": 20220
+    },
+    {
+      "epoch": 0.6093373493975903,
+      "grad_norm": 4.992257214257305,
+      "learning_rate": 3.623515494406643e-06,
+      "loss": 0.9635,
+      "step": 20230
+    },
+    {
+      "epoch": 0.6096385542168675,
+      "grad_norm": 4.397151380303392,
+      "learning_rate": 3.6187282909541577e-06,
+      "loss": 0.9109,
+      "step": 20240
+    },
+    {
+      "epoch": 0.6099397590361446,
+      "grad_norm": 2.082415468363504,
+      "learning_rate": 3.613942457927227e-06,
+      "loss": 0.9128,
+      "step": 20250
+    },
+    {
+      "epoch": 0.6102409638554217,
+      "grad_norm": 5.463364776790733,
+      "learning_rate": 3.6091580000741037e-06,
+      "loss": 0.8842,
+      "step": 20260
+    },
+    {
+      "epoch": 0.6105421686746988,
+      "grad_norm": 1.9078278870053709,
+      "learning_rate": 3.6043749221416765e-06,
+      "loss": 0.8429,
+      "step": 20270
+    },
+    {
+      "epoch": 0.6108433734939759,
+      "grad_norm": 4.108902519580929,
+      "learning_rate": 3.5995932288754655e-06,
+      "loss": 0.9702,
+      "step": 20280
+    },
+    {
+      "epoch": 0.611144578313253,
+      "grad_norm": 5.556046897969918,
+      "learning_rate": 3.5948129250196178e-06,
+      "loss": 0.8819,
+      "step": 20290
+    },
+    {
+      "epoch": 0.6114457831325302,
+      "grad_norm": 5.095835017905952,
+      "learning_rate": 3.590034015316899e-06,
+      "loss": 0.9301,
+      "step": 20300
+    },
+    {
+      "epoch": 0.6117469879518073,
+      "grad_norm": 1.8649830024983534,
+      "learning_rate": 3.585256504508692e-06,
+      "loss": 0.9225,
+      "step": 20310
+    },
+    {
+      "epoch": 0.6120481927710844,
+      "grad_norm": 4.849726724922682,
+      "learning_rate": 3.5804803973349967e-06,
+      "loss": 0.9656,
+      "step": 20320
+    },
+    {
+      "epoch": 0.6123493975903614,
+      "grad_norm": 1.7924568659544633,
+      "learning_rate": 3.5757056985344156e-06,
+      "loss": 0.9468,
+      "step": 20330
+    },
+    {
+      "epoch": 0.6126506024096385,
+      "grad_norm": 3.975614826615879,
+      "learning_rate": 3.5709324128441537e-06,
+      "loss": 0.8459,
+      "step": 20340
+    },
+    {
+      "epoch": 0.6129518072289156,
+      "grad_norm": 7.2789878240228685,
+      "learning_rate": 3.5661605450000175e-06,
+      "loss": 0.9311,
+      "step": 20350
+    },
+    {
+      "epoch": 0.6132530120481928,
+      "grad_norm": 3.9529557637741175,
+      "learning_rate": 3.561390099736404e-06,
+      "loss": 0.907,
+      "step": 20360
+    },
+    {
+      "epoch": 0.6135542168674699,
+      "grad_norm": 4.465699181195703,
+      "learning_rate": 3.5566210817863008e-06,
+      "loss": 0.9014,
+      "step": 20370
+    },
+    {
+      "epoch": 0.613855421686747,
+      "grad_norm": 5.32725086027786,
+      "learning_rate": 3.5518534958812728e-06,
+      "loss": 0.9585,
+      "step": 20380
+    },
+    {
+      "epoch": 0.6141566265060241,
+      "grad_norm": 4.276732014918137,
+      "learning_rate": 3.5470873467514766e-06,
+      "loss": 0.9224,
+      "step": 20390
+    },
+    {
+      "epoch": 0.6144578313253012,
+      "grad_norm": 1.9450433772783566,
+      "learning_rate": 3.542322639125632e-06,
+      "loss": 0.8167,
+      "step": 20400
+    },
+    {
+      "epoch": 0.6147590361445783,
+      "grad_norm": 5.263741054792196,
+      "learning_rate": 3.5375593777310324e-06,
+      "loss": 0.9317,
+      "step": 20410
+    },
+    {
+      "epoch": 0.6150602409638555,
+      "grad_norm": 5.15699066847789,
+      "learning_rate": 3.532797567293539e-06,
+      "loss": 0.9665,
+      "step": 20420
+    },
+    {
+      "epoch": 0.6153614457831326,
+      "grad_norm": 4.8665988430664875,
+      "learning_rate": 3.5280372125375693e-06,
+      "loss": 0.8968,
+      "step": 20430
+    },
+    {
+      "epoch": 0.6156626506024097,
+      "grad_norm": 5.321997110097974,
+      "learning_rate": 3.523278318186101e-06,
+      "loss": 0.9159,
+      "step": 20440
+    },
+    {
+      "epoch": 0.6159638554216867,
+      "grad_norm": 4.8472300516266715,
+      "learning_rate": 3.5185208889606538e-06,
+      "loss": 1.0397,
+      "step": 20450
+    },
+    {
+      "epoch": 0.6162650602409638,
+      "grad_norm": 4.693231960247368,
+      "learning_rate": 3.513764929581308e-06,
+      "loss": 1.0035,
+      "step": 20460
+    },
+    {
+      "epoch": 0.6165662650602409,
+      "grad_norm": 5.936565705194099,
+      "learning_rate": 3.509010444766674e-06,
+      "loss": 0.9271,
+      "step": 20470
+    },
+    {
+      "epoch": 0.6168674698795181,
+      "grad_norm": 7.129494217818609,
+      "learning_rate": 3.5042574392339026e-06,
+      "loss": 0.9824,
+      "step": 20480
+    },
+    {
+      "epoch": 0.6171686746987952,
+      "grad_norm": 5.480968897885322,
+      "learning_rate": 3.4995059176986767e-06,
+      "loss": 1.0018,
+      "step": 20490
+    },
+    {
+      "epoch": 0.6174698795180723,
+      "grad_norm": 9.132938514513636,
+      "learning_rate": 3.4947558848752104e-06,
+      "loss": 1.0223,
+      "step": 20500
+    },
+    {
+      "epoch": 0.6177710843373494,
+      "grad_norm": 4.1959733837300055,
+      "learning_rate": 3.4900073454762357e-06,
+      "loss": 0.9116,
+      "step": 20510
+    },
+    {
+      "epoch": 0.6180722891566265,
+      "grad_norm": 5.362536928090458,
+      "learning_rate": 3.485260304213005e-06,
+      "loss": 0.8883,
+      "step": 20520
+    },
+    {
+      "epoch": 0.6183734939759036,
+      "grad_norm": 1.9062269351441563,
+      "learning_rate": 3.4805147657952852e-06,
+      "loss": 0.9426,
+      "step": 20530
+    },
+    {
+      "epoch": 0.6186746987951808,
+      "grad_norm": 4.679132122109383,
+      "learning_rate": 3.475770734931353e-06,
+      "loss": 0.9292,
+      "step": 20540
+    },
+    {
+      "epoch": 0.6189759036144579,
+      "grad_norm": 1.9220586476908237,
+      "learning_rate": 3.4710282163279845e-06,
+      "loss": 0.9043,
+      "step": 20550
+    },
+    {
+      "epoch": 0.619277108433735,
+      "grad_norm": 4.785419168173132,
+      "learning_rate": 3.466287214690459e-06,
+      "loss": 0.9772,
+      "step": 20560
+    },
+    {
+      "epoch": 0.619578313253012,
+      "grad_norm": 2.0911223943850508,
+      "learning_rate": 3.4615477347225525e-06,
+      "loss": 0.9037,
+      "step": 20570
+    },
+    {
+      "epoch": 0.6198795180722891,
+      "grad_norm": 4.823163971264218,
+      "learning_rate": 3.4568097811265276e-06,
+      "loss": 0.986,
+      "step": 20580
+    },
+    {
+      "epoch": 0.6201807228915662,
+      "grad_norm": 3.910411253177122,
+      "learning_rate": 3.452073358603134e-06,
+      "loss": 0.9997,
+      "step": 20590
+    },
+    {
+      "epoch": 0.6204819277108434,
+      "grad_norm": 7.497753390009346,
+      "learning_rate": 3.4473384718516034e-06,
+      "loss": 0.8518,
+      "step": 20600
+    },
+    {
+      "epoch": 0.6207831325301205,
+      "grad_norm": 5.469582698139495,
+      "learning_rate": 3.4426051255696434e-06,
+      "loss": 0.9357,
+      "step": 20610
+    },
+    {
+      "epoch": 0.6210843373493976,
+      "grad_norm": 4.964654287958929,
+      "learning_rate": 3.437873324453431e-06,
+      "loss": 0.9093,
+      "step": 20620
+    },
+    {
+      "epoch": 0.6213855421686747,
+      "grad_norm": 4.095211442335393,
+      "learning_rate": 3.4331430731976116e-06,
+      "loss": 1.0008,
+      "step": 20630
+    },
+    {
+      "epoch": 0.6216867469879518,
+      "grad_norm": 1.955445042840258,
+      "learning_rate": 3.428414376495295e-06,
+      "loss": 0.8561,
+      "step": 20640
+    },
+    {
+      "epoch": 0.6219879518072289,
+      "grad_norm": 4.372613284683938,
+      "learning_rate": 3.423687239038045e-06,
+      "loss": 0.896,
+      "step": 20650
+    },
+    {
+      "epoch": 0.6222891566265061,
+      "grad_norm": 5.321712441142346,
+      "learning_rate": 3.4189616655158803e-06,
+      "loss": 0.8856,
+      "step": 20660
+    },
+    {
+      "epoch": 0.6225903614457832,
+      "grad_norm": 4.741666056448641,
+      "learning_rate": 3.4142376606172694e-06,
+      "loss": 0.9079,
+      "step": 20670
+    },
+    {
+      "epoch": 0.6228915662650603,
+      "grad_norm": 2.084327553394434,
+      "learning_rate": 3.4095152290291216e-06,
+      "loss": 0.8534,
+      "step": 20680
+    },
+    {
+      "epoch": 0.6231927710843373,
+      "grad_norm": 5.713163657925427,
+      "learning_rate": 3.4047943754367887e-06,
+      "loss": 0.928,
+      "step": 20690
+    },
+    {
+      "epoch": 0.6234939759036144,
+      "grad_norm": 6.769462101164566,
+      "learning_rate": 3.400075104524049e-06,
+      "loss": 0.9881,
+      "step": 20700
+    },
+    {
+      "epoch": 0.6237951807228915,
+      "grad_norm": 4.803985239731172,
+      "learning_rate": 3.3953574209731234e-06,
+      "loss": 1.0066,
+      "step": 20710
+    },
+    {
+      "epoch": 0.6240963855421687,
+      "grad_norm": 1.9421694425476366,
+      "learning_rate": 3.3906413294646468e-06,
+      "loss": 0.885,
+      "step": 20720
+    },
+    {
+      "epoch": 0.6243975903614458,
+      "grad_norm": 4.404078424357399,
+      "learning_rate": 3.3859268346776795e-06,
+      "loss": 0.9244,
+      "step": 20730
+    },
+    {
+      "epoch": 0.6246987951807229,
+      "grad_norm": 2.092217539247877,
+      "learning_rate": 3.381213941289695e-06,
+      "loss": 0.9255,
+      "step": 20740
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 5.637280669702821,
+      "learning_rate": 3.3765026539765832e-06,
+      "loss": 1.0102,
+      "step": 20750
+    },
+    {
+      "epoch": 0.6253012048192771,
+      "grad_norm": 4.940501786030459,
+      "learning_rate": 3.371792977412635e-06,
+      "loss": 0.8817,
+      "step": 20760
+    },
+    {
+      "epoch": 0.6256024096385542,
+      "grad_norm": 2.011743660124654,
+      "learning_rate": 3.3670849162705426e-06,
+      "loss": 0.9622,
+      "step": 20770
+    },
+    {
+      "epoch": 0.6259036144578313,
+      "grad_norm": 4.069794571218934,
+      "learning_rate": 3.362378475221404e-06,
+      "loss": 0.9492,
+      "step": 20780
+    },
+    {
+      "epoch": 0.6262048192771085,
+      "grad_norm": 5.497143811656474,
+      "learning_rate": 3.357673658934699e-06,
+      "loss": 0.9171,
+      "step": 20790
+    },
+    {
+      "epoch": 0.6265060240963856,
+      "grad_norm": 6.532696567953873,
+      "learning_rate": 3.352970472078302e-06,
+      "loss": 0.8539,
+      "step": 20800
+    },
+    {
+      "epoch": 0.6268072289156627,
+      "grad_norm": 6.8112701715789274,
+      "learning_rate": 3.3482689193184677e-06,
+      "loss": 0.894,
+      "step": 20810
+    },
+    {
+      "epoch": 0.6271084337349397,
+      "grad_norm": 6.67718122150574,
+      "learning_rate": 3.3435690053198335e-06,
+      "loss": 0.9361,
+      "step": 20820
+    },
+    {
+      "epoch": 0.6274096385542168,
+      "grad_norm": 9.138516553035258,
+      "learning_rate": 3.3388707347454058e-06,
+      "loss": 0.9582,
+      "step": 20830
+    },
+    {
+      "epoch": 0.6277108433734939,
+      "grad_norm": 3.9872707008689683,
+      "learning_rate": 3.3341741122565636e-06,
+      "loss": 0.9401,
+      "step": 20840
+    },
+    {
+      "epoch": 0.6280120481927711,
+      "grad_norm": 13.163429872336879,
+      "learning_rate": 3.3294791425130512e-06,
+      "loss": 0.9015,
+      "step": 20850
+    },
+    {
+      "epoch": 0.6283132530120482,
+      "grad_norm": 5.050196117407373,
+      "learning_rate": 3.3247858301729726e-06,
+      "loss": 0.997,
+      "step": 20860
+    },
+    {
+      "epoch": 0.6286144578313253,
+      "grad_norm": 6.989204899709363,
+      "learning_rate": 3.3200941798927845e-06,
+      "loss": 1.0165,
+      "step": 20870
+    },
+    {
+      "epoch": 0.6289156626506024,
+      "grad_norm": 5.340432081133796,
+      "learning_rate": 3.3154041963272965e-06,
+      "loss": 0.9836,
+      "step": 20880
+    },
+    {
+      "epoch": 0.6292168674698795,
+      "grad_norm": 4.391546715755655,
+      "learning_rate": 3.3107158841296684e-06,
+      "loss": 1.0053,
+      "step": 20890
+    },
+    {
+      "epoch": 0.6295180722891566,
+      "grad_norm": 5.811563634386749,
+      "learning_rate": 3.3060292479513955e-06,
+      "loss": 0.9144,
+      "step": 20900
+    },
+    {
+      "epoch": 0.6298192771084338,
+      "grad_norm": 4.471575062661988,
+      "learning_rate": 3.301344292442314e-06,
+      "loss": 0.9673,
+      "step": 20910
+    },
+    {
+      "epoch": 0.6301204819277109,
+      "grad_norm": 10.841630783650825,
+      "learning_rate": 3.2966610222505926e-06,
+      "loss": 0.8931,
+      "step": 20920
+    },
+    {
+      "epoch": 0.630421686746988,
+      "grad_norm": 2.026385924513245,
+      "learning_rate": 3.2919794420227273e-06,
+      "loss": 0.8836,
+      "step": 20930
+    },
+    {
+      "epoch": 0.630722891566265,
+      "grad_norm": 1.96031208345594,
+      "learning_rate": 3.2872995564035353e-06,
+      "loss": 0.8269,
+      "step": 20940
+    },
+    {
+      "epoch": 0.6310240963855421,
+      "grad_norm": 4.189081050732709,
+      "learning_rate": 3.2826213700361543e-06,
+      "loss": 0.9789,
+      "step": 20950
+    },
+    {
+      "epoch": 0.6313253012048192,
+      "grad_norm": 7.1258907113227385,
+      "learning_rate": 3.2779448875620384e-06,
+      "loss": 1.0044,
+      "step": 20960
+    },
+    {
+      "epoch": 0.6316265060240964,
+      "grad_norm": 5.258855136928314,
+      "learning_rate": 3.2732701136209475e-06,
+      "loss": 1.0181,
+      "step": 20970
+    },
+    {
+      "epoch": 0.6319277108433735,
+      "grad_norm": 1.997481998766399,
+      "learning_rate": 3.268597052850948e-06,
+      "loss": 0.9278,
+      "step": 20980
+    },
+    {
+      "epoch": 0.6322289156626506,
+      "grad_norm": 4.314780294049713,
+      "learning_rate": 3.263925709888406e-06,
+      "loss": 0.9017,
+      "step": 20990
+    },
+    {
+      "epoch": 0.6325301204819277,
+      "grad_norm": 2.0480808621873416,
+      "learning_rate": 3.2592560893679843e-06,
+      "loss": 1.0029,
+      "step": 21000
+    },
+    {
+      "epoch": 0.6328313253012048,
+      "grad_norm": 2.180648854142401,
+      "learning_rate": 3.254588195922637e-06,
+      "loss": 0.8996,
+      "step": 21010
+    },
+    {
+      "epoch": 0.6331325301204819,
+      "grad_norm": 3.762274851734408,
+      "learning_rate": 3.2499220341836e-06,
+      "loss": 0.9475,
+      "step": 21020
+    },
+    {
+      "epoch": 0.6334337349397591,
+      "grad_norm": 2.0875260721228552,
+      "learning_rate": 3.2452576087804016e-06,
+      "loss": 0.9259,
+      "step": 21030
+    },
+    {
+      "epoch": 0.6337349397590362,
+      "grad_norm": 5.2835387813964765,
+      "learning_rate": 3.240594924340835e-06,
+      "loss": 0.9248,
+      "step": 21040
+    },
+    {
+      "epoch": 0.6340361445783133,
+      "grad_norm": 5.87641539779945,
+      "learning_rate": 3.2359339854909743e-06,
+      "loss": 0.9547,
+      "step": 21050
+    },
+    {
+      "epoch": 0.6343373493975903,
+      "grad_norm": 1.9379416125510331,
+      "learning_rate": 3.2312747968551585e-06,
+      "loss": 0.9719,
+      "step": 21060
+    },
+    {
+      "epoch": 0.6346385542168674,
+      "grad_norm": 2.0093708815095805,
+      "learning_rate": 3.2266173630559916e-06,
+      "loss": 0.906,
+      "step": 21070
+    },
+    {
+      "epoch": 0.6349397590361445,
+      "grad_norm": 1.9094212823134054,
+      "learning_rate": 3.221961688714336e-06,
+      "loss": 0.9814,
+      "step": 21080
+    },
+    {
+      "epoch": 0.6352409638554217,
+      "grad_norm": 4.4177387430879484,
+      "learning_rate": 3.217307778449306e-06,
+      "loss": 0.9457,
+      "step": 21090
+    },
+    {
+      "epoch": 0.6355421686746988,
+      "grad_norm": 3.8700286203190046,
+      "learning_rate": 3.2126556368782713e-06,
+      "loss": 0.8604,
+      "step": 21100
+    },
+    {
+      "epoch": 0.6358433734939759,
+      "grad_norm": 4.233246673498202,
+      "learning_rate": 3.2080052686168417e-06,
+      "loss": 0.9447,
+      "step": 21110
+    },
+    {
+      "epoch": 0.636144578313253,
+      "grad_norm": 4.449524022982166,
+      "learning_rate": 3.2033566782788694e-06,
+      "loss": 1.0633,
+      "step": 21120
+    },
+    {
+      "epoch": 0.6364457831325301,
+      "grad_norm": 1.8467476537699954,
+      "learning_rate": 3.1987098704764425e-06,
+      "loss": 0.8408,
+      "step": 21130
+    },
+    {
+      "epoch": 0.6367469879518072,
+      "grad_norm": 4.668357602821081,
+      "learning_rate": 3.1940648498198812e-06,
+      "loss": 1.0066,
+      "step": 21140
+    },
+    {
+      "epoch": 0.6370481927710844,
+      "grad_norm": 6.243675596567151,
+      "learning_rate": 3.1894216209177327e-06,
+      "loss": 1.0347,
+      "step": 21150
+    },
+    {
+      "epoch": 0.6373493975903615,
+      "grad_norm": 5.874779925360037,
+      "learning_rate": 3.1847801883767638e-06,
+      "loss": 1.0558,
+      "step": 21160
+    },
+    {
+      "epoch": 0.6376506024096386,
+      "grad_norm": 3.719219897415577,
+      "learning_rate": 3.180140556801963e-06,
+      "loss": 0.8836,
+      "step": 21170
+    },
+    {
+      "epoch": 0.6379518072289156,
+      "grad_norm": 5.218115830473539,
+      "learning_rate": 3.1755027307965314e-06,
+      "loss": 0.8676,
+      "step": 21180
+    },
+    {
+      "epoch": 0.6382530120481927,
+      "grad_norm": 5.262196822833833,
+      "learning_rate": 3.1708667149618743e-06,
+      "loss": 0.9798,
+      "step": 21190
+    },
+    {
+      "epoch": 0.6385542168674698,
+      "grad_norm": 4.635588946933596,
+      "learning_rate": 3.166232513897604e-06,
+      "loss": 1.0011,
+      "step": 21200
+    },
+    {
+      "epoch": 0.638855421686747,
+      "grad_norm": 4.0853151451849286,
+      "learning_rate": 3.1616001322015354e-06,
+      "loss": 0.9333,
+      "step": 21210
+    },
+    {
+      "epoch": 0.6391566265060241,
+      "grad_norm": 4.519658605776038,
+      "learning_rate": 3.156969574469674e-06,
+      "loss": 0.889,
+      "step": 21220
+    },
+    {
+      "epoch": 0.6394578313253012,
+      "grad_norm": 3.5410025112383807,
+      "learning_rate": 3.1523408452962156e-06,
+      "loss": 0.9852,
+      "step": 21230
+    },
+    {
+      "epoch": 0.6397590361445783,
+      "grad_norm": 6.185458822317108,
+      "learning_rate": 3.1477139492735455e-06,
+      "loss": 0.9616,
+      "step": 21240
+    },
+    {
+      "epoch": 0.6400602409638554,
+      "grad_norm": 8.127777230816926,
+      "learning_rate": 3.1430888909922273e-06,
+      "loss": 0.9867,
+      "step": 21250
+    },
+    {
+      "epoch": 0.6403614457831325,
+      "grad_norm": 6.7524889501059056,
+      "learning_rate": 3.1384656750410013e-06,
+      "loss": 1.0125,
+      "step": 21260
+    },
+    {
+      "epoch": 0.6406626506024097,
+      "grad_norm": 4.575187436704799,
+      "learning_rate": 3.13384430600678e-06,
+      "loss": 0.828,
+      "step": 21270
+    },
+    {
+      "epoch": 0.6409638554216868,
+      "grad_norm": 4.58634340419795,
+      "learning_rate": 3.129224788474646e-06,
+      "loss": 0.8931,
+      "step": 21280
+    },
+    {
+      "epoch": 0.6412650602409639,
+      "grad_norm": 6.398077486688091,
+      "learning_rate": 3.1246071270278422e-06,
+      "loss": 0.8867,
+      "step": 21290
+    },
+    {
+      "epoch": 0.641566265060241,
+      "grad_norm": 8.562850814957184,
+      "learning_rate": 3.1199913262477697e-06,
+      "loss": 0.891,
+      "step": 21300
+    },
+    {
+      "epoch": 0.641867469879518,
+      "grad_norm": 7.3977415208563855,
+      "learning_rate": 3.1153773907139855e-06,
+      "loss": 0.827,
+      "step": 21310
+    },
+    {
+      "epoch": 0.6421686746987951,
+      "grad_norm": 3.906194519864792,
+      "learning_rate": 3.1107653250041956e-06,
+      "loss": 0.9751,
+      "step": 21320
+    },
+    {
+      "epoch": 0.6424698795180723,
+      "grad_norm": 4.563963427294925,
+      "learning_rate": 3.1061551336942513e-06,
+      "loss": 0.9647,
+      "step": 21330
+    },
+    {
+      "epoch": 0.6427710843373494,
+      "grad_norm": 4.5838313654218785,
+      "learning_rate": 3.101546821358139e-06,
+      "loss": 0.9122,
+      "step": 21340
+    },
+    {
+      "epoch": 0.6430722891566265,
+      "grad_norm": 24.549443758048707,
+      "learning_rate": 3.096940392567991e-06,
+      "loss": 0.9398,
+      "step": 21350
+    },
+    {
+      "epoch": 0.6433734939759036,
+      "grad_norm": 3.2979766870583487,
+      "learning_rate": 3.0923358518940616e-06,
+      "loss": 1.0312,
+      "step": 21360
+    },
+    {
+      "epoch": 0.6436746987951807,
+      "grad_norm": 4.862204403795849,
+      "learning_rate": 3.0877332039047357e-06,
+      "loss": 1.0492,
+      "step": 21370
+    },
+    {
+      "epoch": 0.6439759036144578,
+      "grad_norm": 5.941711738298755,
+      "learning_rate": 3.083132453166519e-06,
+      "loss": 0.8433,
+      "step": 21380
+    },
+    {
+      "epoch": 0.644277108433735,
+      "grad_norm": 2.065313804808686,
+      "learning_rate": 3.0785336042440384e-06,
+      "loss": 0.9444,
+      "step": 21390
+    },
+    {
+      "epoch": 0.6445783132530121,
+      "grad_norm": 4.452525298541014,
+      "learning_rate": 3.073936661700029e-06,
+      "loss": 0.8828,
+      "step": 21400
+    },
+    {
+      "epoch": 0.6448795180722892,
+      "grad_norm": 2.2144776002648268,
+      "learning_rate": 3.069341630095337e-06,
+      "loss": 0.8824,
+      "step": 21410
+    },
+    {
+      "epoch": 0.6451807228915662,
+      "grad_norm": 5.615872580136008,
+      "learning_rate": 3.0647485139889145e-06,
+      "loss": 0.9733,
+      "step": 21420
+    },
+    {
+      "epoch": 0.6454819277108433,
+      "grad_norm": 1.9154614170494852,
+      "learning_rate": 3.0601573179378095e-06,
+      "loss": 0.834,
+      "step": 21430
+    },
+    {
+      "epoch": 0.6457831325301204,
+      "grad_norm": 4.325685577824708,
+      "learning_rate": 3.055568046497167e-06,
+      "loss": 0.8614,
+      "step": 21440
+    },
+    {
+      "epoch": 0.6460843373493976,
+      "grad_norm": 4.667964013476934,
+      "learning_rate": 3.0509807042202205e-06,
+      "loss": 0.9739,
+      "step": 21450
+    },
+    {
+      "epoch": 0.6463855421686747,
+      "grad_norm": 5.049570869182073,
+      "learning_rate": 3.046395295658293e-06,
+      "loss": 1.0064,
+      "step": 21460
+    },
+    {
+      "epoch": 0.6466867469879518,
+      "grad_norm": 4.754221646873815,
+      "learning_rate": 3.0418118253607876e-06,
+      "loss": 1.089,
+      "step": 21470
+    },
+    {
+      "epoch": 0.6469879518072289,
+      "grad_norm": 5.263135051321002,
+      "learning_rate": 3.037230297875181e-06,
+      "loss": 0.9869,
+      "step": 21480
+    },
+    {
+      "epoch": 0.647289156626506,
+      "grad_norm": 6.299277849284255,
+      "learning_rate": 3.032650717747028e-06,
+      "loss": 0.879,
+      "step": 21490
+    },
+    {
+      "epoch": 0.6475903614457831,
+      "grad_norm": 6.14985684922634,
+      "learning_rate": 3.0280730895199494e-06,
+      "loss": 0.9518,
+      "step": 21500
+    },
+    {
+      "epoch": 0.6478915662650603,
+      "grad_norm": 5.010786809719041,
+      "learning_rate": 3.0234974177356246e-06,
+      "loss": 0.9193,
+      "step": 21510
+    },
+    {
+      "epoch": 0.6481927710843374,
+      "grad_norm": 3.884190230743868,
+      "learning_rate": 3.0189237069337973e-06,
+      "loss": 0.986,
+      "step": 21520
+    },
+    {
+      "epoch": 0.6484939759036145,
+      "grad_norm": 1.9760447241771413,
+      "learning_rate": 3.0143519616522654e-06,
+      "loss": 0.779,
+      "step": 21530
+    },
+    {
+      "epoch": 0.6487951807228916,
+      "grad_norm": 4.191624279760544,
+      "learning_rate": 3.009782186426875e-06,
+      "loss": 0.9806,
+      "step": 21540
+    },
+    {
+      "epoch": 0.6490963855421686,
+      "grad_norm": 6.759338320255287,
+      "learning_rate": 3.0052143857915184e-06,
+      "loss": 0.8706,
+      "step": 21550
+    },
+    {
+      "epoch": 0.6493975903614457,
+      "grad_norm": 4.264958450597003,
+      "learning_rate": 3.0006485642781268e-06,
+      "loss": 0.9832,
+      "step": 21560
+    },
+    {
+      "epoch": 0.6496987951807229,
+      "grad_norm": 10.751978846120968,
+      "learning_rate": 2.9960847264166715e-06,
+      "loss": 0.9013,
+      "step": 21570
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0557452139786863,
+      "learning_rate": 2.991522876735154e-06,
+      "loss": 0.9567,
+      "step": 21580
+    },
+    {
+      "epoch": 0.6503012048192771,
+      "grad_norm": 5.519619416231213,
+      "learning_rate": 2.986963019759599e-06,
+      "loss": 0.9118,
+      "step": 21590
+    },
+    {
+      "epoch": 0.6506024096385542,
+      "grad_norm": 1.759742155520633,
+      "learning_rate": 2.9824051600140647e-06,
+      "loss": 0.9272,
+      "step": 21600
+    },
+    {
+      "epoch": 0.6509036144578313,
+      "grad_norm": 10.756528558162993,
+      "learning_rate": 2.9778493020206155e-06,
+      "loss": 0.9989,
+      "step": 21610
+    },
+    {
+      "epoch": 0.6512048192771084,
+      "grad_norm": 4.89241199410044,
+      "learning_rate": 2.9732954502993383e-06,
+      "loss": 0.9212,
+      "step": 21620
+    },
+    {
+      "epoch": 0.6515060240963856,
+      "grad_norm": 6.12006754836966,
+      "learning_rate": 2.9687436093683252e-06,
+      "loss": 0.8998,
+      "step": 21630
+    },
+    {
+      "epoch": 0.6518072289156627,
+      "grad_norm": 1.8227140458962348,
+      "learning_rate": 2.9641937837436762e-06,
+      "loss": 0.9195,
+      "step": 21640
+    },
+    {
+      "epoch": 0.6521084337349398,
+      "grad_norm": 8.056246962058308,
+      "learning_rate": 2.9596459779394902e-06,
+      "loss": 0.9162,
+      "step": 21650
+    },
+    {
+      "epoch": 0.6524096385542169,
+      "grad_norm": 1.8153106365669807,
+      "learning_rate": 2.955100196467859e-06,
+      "loss": 0.9189,
+      "step": 21660
+    },
+    {
+      "epoch": 0.6527108433734939,
+      "grad_norm": 4.19096140512222,
+      "learning_rate": 2.9505564438388746e-06,
+      "loss": 0.9901,
+      "step": 21670
+    },
+    {
+      "epoch": 0.653012048192771,
+      "grad_norm": 4.40998587864527,
+      "learning_rate": 2.946014724560607e-06,
+      "loss": 0.9241,
+      "step": 21680
+    },
+    {
+      "epoch": 0.6533132530120482,
+      "grad_norm": 2.08107272327245,
+      "learning_rate": 2.9414750431391136e-06,
+      "loss": 0.9554,
+      "step": 21690
+    },
+    {
+      "epoch": 0.6536144578313253,
+      "grad_norm": 1.9749393817026102,
+      "learning_rate": 2.936937404078428e-06,
+      "loss": 0.8746,
+      "step": 21700
+    },
+    {
+      "epoch": 0.6539156626506024,
+      "grad_norm": 6.080370488751193,
+      "learning_rate": 2.93240181188056e-06,
+      "loss": 0.8827,
+      "step": 21710
+    },
+    {
+      "epoch": 0.6542168674698795,
+      "grad_norm": 4.235770083389657,
+      "learning_rate": 2.9278682710454863e-06,
+      "loss": 0.9714,
+      "step": 21720
+    },
+    {
+      "epoch": 0.6545180722891566,
+      "grad_norm": 4.691061814383612,
+      "learning_rate": 2.9233367860711485e-06,
+      "loss": 0.9328,
+      "step": 21730
+    },
+    {
+      "epoch": 0.6548192771084337,
+      "grad_norm": 3.777238716083142,
+      "learning_rate": 2.91880736145345e-06,
+      "loss": 1.0006,
+      "step": 21740
+    },
+    {
+      "epoch": 0.6551204819277109,
+      "grad_norm": 4.1143117256133435,
+      "learning_rate": 2.9142800016862487e-06,
+      "loss": 0.9042,
+      "step": 21750
+    },
+    {
+      "epoch": 0.655421686746988,
+      "grad_norm": 4.58042100265641,
+      "learning_rate": 2.9097547112613556e-06,
+      "loss": 0.9748,
+      "step": 21760
+    },
+    {
+      "epoch": 0.6557228915662651,
+      "grad_norm": 4.354875232053938,
+      "learning_rate": 2.9052314946685223e-06,
+      "loss": 0.9474,
+      "step": 21770
+    },
+    {
+      "epoch": 0.6560240963855422,
+      "grad_norm": 5.6314880479577285,
+      "learning_rate": 2.9007103563954525e-06,
+      "loss": 1.0376,
+      "step": 21780
+    },
+    {
+      "epoch": 0.6563253012048192,
+      "grad_norm": 1.9681158907671024,
+      "learning_rate": 2.896191300927782e-06,
+      "loss": 0.9036,
+      "step": 21790
+    },
+    {
+      "epoch": 0.6566265060240963,
+      "grad_norm": 2.1799316261529778,
+      "learning_rate": 2.89167433274908e-06,
+      "loss": 0.8854,
+      "step": 21800
+    },
+    {
+      "epoch": 0.6569277108433735,
+      "grad_norm": 3.6531469138663084,
+      "learning_rate": 2.887159456340849e-06,
+      "loss": 0.961,
+      "step": 21810
+    },
+    {
+      "epoch": 0.6572289156626506,
+      "grad_norm": 4.7381749172876075,
+      "learning_rate": 2.8826466761825102e-06,
+      "loss": 0.7997,
+      "step": 21820
+    },
+    {
+      "epoch": 0.6575301204819277,
+      "grad_norm": 10.760848820952543,
+      "learning_rate": 2.8781359967514094e-06,
+      "loss": 0.874,
+      "step": 21830
+    },
+    {
+      "epoch": 0.6578313253012048,
+      "grad_norm": 3.110054614216242,
+      "learning_rate": 2.8736274225228045e-06,
+      "loss": 0.9682,
+      "step": 21840
+    },
+    {
+      "epoch": 0.6581325301204819,
+      "grad_norm": 5.364596429239417,
+      "learning_rate": 2.869120957969872e-06,
+      "loss": 0.9245,
+      "step": 21850
+    },
+    {
+      "epoch": 0.658433734939759,
+      "grad_norm": 7.324574404699955,
+      "learning_rate": 2.864616607563689e-06,
+      "loss": 0.9278,
+      "step": 21860
+    },
+    {
+      "epoch": 0.6587349397590362,
+      "grad_norm": 4.062528090718364,
+      "learning_rate": 2.860114375773233e-06,
+      "loss": 0.8504,
+      "step": 21870
+    },
+    {
+      "epoch": 0.6590361445783133,
+      "grad_norm": 4.1432392143125885,
+      "learning_rate": 2.855614267065383e-06,
+      "loss": 0.9679,
+      "step": 21880
+    },
+    {
+      "epoch": 0.6593373493975904,
+      "grad_norm": 4.627951143207221,
+      "learning_rate": 2.8511162859049156e-06,
+      "loss": 0.8838,
+      "step": 21890
+    },
+    {
+      "epoch": 0.6596385542168675,
+      "grad_norm": 1.9882933520820298,
+      "learning_rate": 2.8466204367544904e-06,
+      "loss": 0.9857,
+      "step": 21900
+    },
+    {
+      "epoch": 0.6599397590361445,
+      "grad_norm": 13.916313659291012,
+      "learning_rate": 2.8421267240746535e-06,
+      "loss": 0.916,
+      "step": 21910
+    },
+    {
+      "epoch": 0.6602409638554216,
+      "grad_norm": 1.8548867782047582,
+      "learning_rate": 2.8376351523238323e-06,
+      "loss": 0.902,
+      "step": 21920
+    },
+    {
+      "epoch": 0.6605421686746988,
+      "grad_norm": 2.1290771136000566,
+      "learning_rate": 2.833145725958328e-06,
+      "loss": 0.9487,
+      "step": 21930
+    },
+    {
+      "epoch": 0.6608433734939759,
+      "grad_norm": 4.476812288039868,
+      "learning_rate": 2.828658449432317e-06,
+      "loss": 0.8224,
+      "step": 21940
+    },
+    {
+      "epoch": 0.661144578313253,
+      "grad_norm": 5.157041111022135,
+      "learning_rate": 2.8241733271978367e-06,
+      "loss": 0.9983,
+      "step": 21950
+    },
+    {
+      "epoch": 0.6614457831325301,
+      "grad_norm": 3.906482197609336,
+      "learning_rate": 2.819690363704797e-06,
+      "loss": 1.0521,
+      "step": 21960
+    },
+    {
+      "epoch": 0.6617469879518072,
+      "grad_norm": 5.18254525615854,
+      "learning_rate": 2.8152095634009534e-06,
+      "loss": 0.9504,
+      "step": 21970
+    },
+    {
+      "epoch": 0.6620481927710843,
+      "grad_norm": 3.485683403366198,
+      "learning_rate": 2.810730930731923e-06,
+      "loss": 0.8491,
+      "step": 21980
+    },
+    {
+      "epoch": 0.6623493975903615,
+      "grad_norm": 2.013843803590297,
+      "learning_rate": 2.806254470141174e-06,
+      "loss": 0.8204,
+      "step": 21990
+    },
+    {
+      "epoch": 0.6626506024096386,
+      "grad_norm": 4.72464974501394,
+      "learning_rate": 2.8017801860700144e-06,
+      "loss": 1.0248,
+      "step": 22000
+    },
+    {
+      "epoch": 0.6629518072289157,
+      "grad_norm": 4.382429131044873,
+      "learning_rate": 2.797308082957597e-06,
+      "loss": 0.9791,
+      "step": 22010
+    },
+    {
+      "epoch": 0.6632530120481928,
+      "grad_norm": 4.119000658522985,
+      "learning_rate": 2.792838165240903e-06,
+      "loss": 0.955,
+      "step": 22020
+    },
+    {
+      "epoch": 0.6635542168674698,
+      "grad_norm": 4.186191863720116,
+      "learning_rate": 2.788370437354757e-06,
+      "loss": 0.9253,
+      "step": 22030
+    },
+    {
+      "epoch": 0.6638554216867469,
+      "grad_norm": 2.129445625917607,
+      "learning_rate": 2.783904903731803e-06,
+      "loss": 0.9486,
+      "step": 22040
+    },
+    {
+      "epoch": 0.6641566265060241,
+      "grad_norm": 3.633194431817707,
+      "learning_rate": 2.779441568802509e-06,
+      "loss": 0.9825,
+      "step": 22050
+    },
+    {
+      "epoch": 0.6644578313253012,
+      "grad_norm": 1.9479286988967859,
+      "learning_rate": 2.7749804369951628e-06,
+      "loss": 0.9261,
+      "step": 22060
+    },
+    {
+      "epoch": 0.6647590361445783,
+      "grad_norm": 2.0032089183615884,
+      "learning_rate": 2.770521512735867e-06,
+      "loss": 0.8994,
+      "step": 22070
+    },
+    {
+      "epoch": 0.6650602409638554,
+      "grad_norm": 5.779242393722818,
+      "learning_rate": 2.7660648004485326e-06,
+      "loss": 0.8372,
+      "step": 22080
+    },
+    {
+      "epoch": 0.6653614457831325,
+      "grad_norm": 1.8492622598829354,
+      "learning_rate": 2.7616103045548736e-06,
+      "loss": 0.7804,
+      "step": 22090
+    },
+    {
+      "epoch": 0.6656626506024096,
+      "grad_norm": 2.101195866679531,
+      "learning_rate": 2.757158029474414e-06,
+      "loss": 0.9847,
+      "step": 22100
+    },
+    {
+      "epoch": 0.6659638554216868,
+      "grad_norm": 3.7557309817886506,
+      "learning_rate": 2.7527079796244655e-06,
+      "loss": 0.8384,
+      "step": 22110
+    },
+    {
+      "epoch": 0.6662650602409639,
+      "grad_norm": 2.035807673876726,
+      "learning_rate": 2.748260159420133e-06,
+      "loss": 0.87,
+      "step": 22120
+    },
+    {
+      "epoch": 0.666566265060241,
+      "grad_norm": 4.009242037465268,
+      "learning_rate": 2.743814573274309e-06,
+      "loss": 0.8758,
+      "step": 22130
+    },
+    {
+      "epoch": 0.6668674698795181,
+      "grad_norm": 1.8449185982373857,
+      "learning_rate": 2.739371225597679e-06,
+      "loss": 0.9592,
+      "step": 22140
+    },
+    {
+      "epoch": 0.6671686746987951,
+      "grad_norm": 5.088642472988054,
+      "learning_rate": 2.734930120798695e-06,
+      "loss": 0.9281,
+      "step": 22150
+    },
+    {
+      "epoch": 0.6674698795180722,
+      "grad_norm": 3.9790770984468335,
+      "learning_rate": 2.73049126328359e-06,
+      "loss": 0.973,
+      "step": 22160
+    },
+    {
+      "epoch": 0.6677710843373494,
+      "grad_norm": 3.747401373086219,
+      "learning_rate": 2.7260546574563667e-06,
+      "loss": 1.0107,
+      "step": 22170
+    },
+    {
+      "epoch": 0.6680722891566265,
+      "grad_norm": 4.065737050583372,
+      "learning_rate": 2.721620307718793e-06,
+      "loss": 0.9498,
+      "step": 22180
+    },
+    {
+      "epoch": 0.6683734939759036,
+      "grad_norm": 1.9647076713611802,
+      "learning_rate": 2.7171882184703983e-06,
+      "loss": 0.8652,
+      "step": 22190
+    },
+    {
+      "epoch": 0.6686746987951807,
+      "grad_norm": 3.8641790642423692,
+      "learning_rate": 2.712758394108471e-06,
+      "loss": 0.961,
+      "step": 22200
+    },
+    {
+      "epoch": 0.6689759036144578,
+      "grad_norm": 5.289873751951875,
+      "learning_rate": 2.7083308390280505e-06,
+      "loss": 0.9675,
+      "step": 22210
+    },
+    {
+      "epoch": 0.6692771084337349,
+      "grad_norm": 2.178191652381719,
+      "learning_rate": 2.703905557621925e-06,
+      "loss": 1.0099,
+      "step": 22220
+    },
+    {
+      "epoch": 0.6695783132530121,
+      "grad_norm": 3.4198526034186187,
+      "learning_rate": 2.6994825542806256e-06,
+      "loss": 0.9222,
+      "step": 22230
+    },
+    {
+      "epoch": 0.6698795180722892,
+      "grad_norm": 4.575566807041047,
+      "learning_rate": 2.6950618333924293e-06,
+      "loss": 0.9811,
+      "step": 22240
+    },
+    {
+      "epoch": 0.6701807228915663,
+      "grad_norm": 4.2939082694984,
+      "learning_rate": 2.690643399343341e-06,
+      "loss": 0.9601,
+      "step": 22250
+    },
+    {
+      "epoch": 0.6704819277108434,
+      "grad_norm": 2.108921527222617,
+      "learning_rate": 2.686227256517104e-06,
+      "loss": 0.9413,
+      "step": 22260
+    },
+    {
+      "epoch": 0.6707831325301205,
+      "grad_norm": 2.0112532897038213,
+      "learning_rate": 2.6818134092951763e-06,
+      "loss": 0.9312,
+      "step": 22270
+    },
+    {
+      "epoch": 0.6710843373493975,
+      "grad_norm": 4.732203064573461,
+      "learning_rate": 2.677401862056753e-06,
+      "loss": 0.8101,
+      "step": 22280
+    },
+    {
+      "epoch": 0.6713855421686747,
+      "grad_norm": 4.006396254824637,
+      "learning_rate": 2.6729926191787393e-06,
+      "loss": 0.9421,
+      "step": 22290
+    },
+    {
+      "epoch": 0.6716867469879518,
+      "grad_norm": 3.878257768154987,
+      "learning_rate": 2.668585685035754e-06,
+      "loss": 0.9401,
+      "step": 22300
+    },
+    {
+      "epoch": 0.6719879518072289,
+      "grad_norm": 5.608380025610696,
+      "learning_rate": 2.664181064000127e-06,
+      "loss": 0.867,
+      "step": 22310
+    },
+    {
+      "epoch": 0.672289156626506,
+      "grad_norm": 5.272795359162881,
+      "learning_rate": 2.6597787604418932e-06,
+      "loss": 1.0358,
+      "step": 22320
+    },
+    {
+      "epoch": 0.6725903614457831,
+      "grad_norm": 4.007619309517507,
+      "learning_rate": 2.6553787787287887e-06,
+      "loss": 0.9925,
+      "step": 22330
+    },
+    {
+      "epoch": 0.6728915662650602,
+      "grad_norm": 1.901190551834578,
+      "learning_rate": 2.650981123226243e-06,
+      "loss": 1.0218,
+      "step": 22340
+    },
+    {
+      "epoch": 0.6731927710843374,
+      "grad_norm": 1.9356280137402926,
+      "learning_rate": 2.646585798297384e-06,
+      "loss": 0.9435,
+      "step": 22350
+    },
+    {
+      "epoch": 0.6734939759036145,
+      "grad_norm": 8.077916312214645,
+      "learning_rate": 2.64219280830302e-06,
+      "loss": 0.9134,
+      "step": 22360
+    },
+    {
+      "epoch": 0.6737951807228916,
+      "grad_norm": 4.015693029221075,
+      "learning_rate": 2.6378021576016467e-06,
+      "loss": 0.8836,
+      "step": 22370
+    },
+    {
+      "epoch": 0.6740963855421687,
+      "grad_norm": 4.858974586780675,
+      "learning_rate": 2.6334138505494357e-06,
+      "loss": 0.8455,
+      "step": 22380
+    },
+    {
+      "epoch": 0.6743975903614458,
+      "grad_norm": 4.446218214935879,
+      "learning_rate": 2.6290278915002397e-06,
+      "loss": 0.9219,
+      "step": 22390
+    },
+    {
+      "epoch": 0.6746987951807228,
+      "grad_norm": 2.024909669609107,
+      "learning_rate": 2.6246442848055774e-06,
+      "loss": 0.8926,
+      "step": 22400
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 4.101509972320014,
+      "learning_rate": 2.6202630348146323e-06,
+      "loss": 0.9962,
+      "step": 22410
+    },
+    {
+      "epoch": 0.6753012048192771,
+      "grad_norm": 6.324377242940589,
+      "learning_rate": 2.615884145874253e-06,
+      "loss": 0.8651,
+      "step": 22420
+    },
+    {
+      "epoch": 0.6756024096385542,
+      "grad_norm": 1.9698858673636552,
+      "learning_rate": 2.6115076223289425e-06,
+      "loss": 0.9822,
+      "step": 22430
+    },
+    {
+      "epoch": 0.6759036144578313,
+      "grad_norm": 4.18087979642135,
+      "learning_rate": 2.6071334685208603e-06,
+      "loss": 0.9398,
+      "step": 22440
+    },
+    {
+      "epoch": 0.6762048192771084,
+      "grad_norm": 4.399047417508821,
+      "learning_rate": 2.6027616887898115e-06,
+      "loss": 0.9018,
+      "step": 22450
+    },
+    {
+      "epoch": 0.6765060240963855,
+      "grad_norm": 4.4678499320129355,
+      "learning_rate": 2.598392287473248e-06,
+      "loss": 1.0286,
+      "step": 22460
+    },
+    {
+      "epoch": 0.6768072289156627,
+      "grad_norm": 4.733618726120682,
+      "learning_rate": 2.5940252689062624e-06,
+      "loss": 0.8244,
+      "step": 22470
+    },
+    {
+      "epoch": 0.6771084337349398,
+      "grad_norm": 3.536548697342829,
+      "learning_rate": 2.5896606374215782e-06,
+      "loss": 0.9445,
+      "step": 22480
+    },
+    {
+      "epoch": 0.6774096385542169,
+      "grad_norm": 5.336909603721039,
+      "learning_rate": 2.5852983973495595e-06,
+      "loss": 0.924,
+      "step": 22490
+    },
+    {
+      "epoch": 0.677710843373494,
+      "grad_norm": 10.108833754127733,
+      "learning_rate": 2.580938553018193e-06,
+      "loss": 0.8747,
+      "step": 22500
+    },
+    {
+      "epoch": 0.678012048192771,
+      "grad_norm": 5.081763839435943,
+      "learning_rate": 2.5765811087530836e-06,
+      "loss": 0.9274,
+      "step": 22510
+    },
+    {
+      "epoch": 0.6783132530120481,
+      "grad_norm": 4.37994774438734,
+      "learning_rate": 2.5722260688774603e-06,
+      "loss": 0.9863,
+      "step": 22520
+    },
+    {
+      "epoch": 0.6786144578313253,
+      "grad_norm": 1.9454870534852542,
+      "learning_rate": 2.567873437712169e-06,
+      "loss": 0.8846,
+      "step": 22530
+    },
+    {
+      "epoch": 0.6789156626506024,
+      "grad_norm": 5.497447563037934,
+      "learning_rate": 2.563523219575661e-06,
+      "loss": 0.9687,
+      "step": 22540
+    },
+    {
+      "epoch": 0.6792168674698795,
+      "grad_norm": 4.765911217095411,
+      "learning_rate": 2.5591754187839963e-06,
+      "loss": 1.0395,
+      "step": 22550
+    },
+    {
+      "epoch": 0.6795180722891566,
+      "grad_norm": 5.84346943692407,
+      "learning_rate": 2.554830039650834e-06,
+      "loss": 0.9722,
+      "step": 22560
+    },
+    {
+      "epoch": 0.6798192771084337,
+      "grad_norm": 4.435380094093738,
+      "learning_rate": 2.550487086487433e-06,
+      "loss": 0.9452,
+      "step": 22570
+    },
+    {
+      "epoch": 0.6801204819277108,
+      "grad_norm": 1.9225631411474826,
+      "learning_rate": 2.5461465636026437e-06,
+      "loss": 0.893,
+      "step": 22580
+    },
+    {
+      "epoch": 0.680421686746988,
+      "grad_norm": 3.8590481171700857,
+      "learning_rate": 2.541808475302905e-06,
+      "loss": 0.836,
+      "step": 22590
+    },
+    {
+      "epoch": 0.6807228915662651,
+      "grad_norm": 3.360641772594274,
+      "learning_rate": 2.5374728258922453e-06,
+      "loss": 0.9933,
+      "step": 22600
+    },
+    {
+      "epoch": 0.6810240963855422,
+      "grad_norm": 6.8797070632404385,
+      "learning_rate": 2.5331396196722657e-06,
+      "loss": 0.9276,
+      "step": 22610
+    },
+    {
+      "epoch": 0.6813253012048193,
+      "grad_norm": 2.06110467254301,
+      "learning_rate": 2.5288088609421465e-06,
+      "loss": 0.8647,
+      "step": 22620
+    },
+    {
+      "epoch": 0.6816265060240964,
+      "grad_norm": 7.585795875612082,
+      "learning_rate": 2.5244805539986394e-06,
+      "loss": 0.9727,
+      "step": 22630
+    },
+    {
+      "epoch": 0.6819277108433734,
+      "grad_norm": 4.816359568779887,
+      "learning_rate": 2.5201547031360675e-06,
+      "loss": 0.9718,
+      "step": 22640
+    },
+    {
+      "epoch": 0.6822289156626506,
+      "grad_norm": 6.680003480617296,
+      "learning_rate": 2.5158313126463143e-06,
+      "loss": 0.9246,
+      "step": 22650
+    },
+    {
+      "epoch": 0.6825301204819277,
+      "grad_norm": 4.603594325945119,
+      "learning_rate": 2.511510386818815e-06,
+      "loss": 0.992,
+      "step": 22660
+    },
+    {
+      "epoch": 0.6828313253012048,
+      "grad_norm": 7.142321913370482,
+      "learning_rate": 2.5071919299405727e-06,
+      "loss": 0.9281,
+      "step": 22670
+    },
+    {
+      "epoch": 0.6831325301204819,
+      "grad_norm": 7.449322790098273,
+      "learning_rate": 2.502875946296133e-06,
+      "loss": 0.9674,
+      "step": 22680
+    },
+    {
+      "epoch": 0.683433734939759,
+      "grad_norm": 3.7241719379183857,
+      "learning_rate": 2.4985624401675882e-06,
+      "loss": 0.9732,
+      "step": 22690
+    },
+    {
+      "epoch": 0.6837349397590361,
+      "grad_norm": 6.4546878872351465,
+      "learning_rate": 2.494251415834575e-06,
+      "loss": 0.9987,
+      "step": 22700
+    },
+    {
+      "epoch": 0.6840361445783133,
+      "grad_norm": 3.806961100254641,
+      "learning_rate": 2.4899428775742647e-06,
+      "loss": 0.9286,
+      "step": 22710
+    },
+    {
+      "epoch": 0.6843373493975904,
+      "grad_norm": 1.8950758895547435,
+      "learning_rate": 2.4856368296613642e-06,
+      "loss": 0.9196,
+      "step": 22720
+    },
+    {
+      "epoch": 0.6846385542168675,
+      "grad_norm": 4.920706839616375,
+      "learning_rate": 2.481333276368107e-06,
+      "loss": 0.8401,
+      "step": 22730
+    },
+    {
+      "epoch": 0.6849397590361446,
+      "grad_norm": 2.137533867682102,
+      "learning_rate": 2.4770322219642577e-06,
+      "loss": 0.9533,
+      "step": 22740
+    },
+    {
+      "epoch": 0.6852409638554217,
+      "grad_norm": 3.5879285046816864,
+      "learning_rate": 2.4727336707170973e-06,
+      "loss": 0.9342,
+      "step": 22750
+    },
+    {
+      "epoch": 0.6855421686746987,
+      "grad_norm": 4.667930268627476,
+      "learning_rate": 2.4684376268914203e-06,
+      "loss": 0.9658,
+      "step": 22760
+    },
+    {
+      "epoch": 0.685843373493976,
+      "grad_norm": 5.974689819002652,
+      "learning_rate": 2.464144094749536e-06,
+      "loss": 1.0093,
+      "step": 22770
+    },
+    {
+      "epoch": 0.686144578313253,
+      "grad_norm": 3.794590653212425,
+      "learning_rate": 2.4598530785512666e-06,
+      "loss": 0.9778,
+      "step": 22780
+    },
+    {
+      "epoch": 0.6864457831325301,
+      "grad_norm": 3.727741198701043,
+      "learning_rate": 2.455564582553932e-06,
+      "loss": 0.9811,
+      "step": 22790
+    },
+    {
+      "epoch": 0.6867469879518072,
+      "grad_norm": 2.065284952343428,
+      "learning_rate": 2.4512786110123537e-06,
+      "loss": 0.8216,
+      "step": 22800
+    },
+    {
+      "epoch": 0.6870481927710843,
+      "grad_norm": 3.6275863401766553,
+      "learning_rate": 2.4469951681788483e-06,
+      "loss": 0.9555,
+      "step": 22810
+    },
+    {
+      "epoch": 0.6873493975903614,
+      "grad_norm": 7.553084797736424,
+      "learning_rate": 2.4427142583032244e-06,
+      "loss": 1.0507,
+      "step": 22820
+    },
+    {
+      "epoch": 0.6876506024096386,
+      "grad_norm": 4.238734301726904,
+      "learning_rate": 2.4384358856327762e-06,
+      "loss": 0.9991,
+      "step": 22830
+    },
+    {
+      "epoch": 0.6879518072289157,
+      "grad_norm": 3.7476143896509155,
+      "learning_rate": 2.434160054412282e-06,
+      "loss": 0.9174,
+      "step": 22840
+    },
+    {
+      "epoch": 0.6882530120481928,
+      "grad_norm": 11.350813759651311,
+      "learning_rate": 2.4298867688839967e-06,
+      "loss": 0.9428,
+      "step": 22850
+    },
+    {
+      "epoch": 0.6885542168674699,
+      "grad_norm": 6.179310566217496,
+      "learning_rate": 2.425616033287652e-06,
+      "loss": 0.8853,
+      "step": 22860
+    },
+    {
+      "epoch": 0.688855421686747,
+      "grad_norm": 2.030623479241787,
+      "learning_rate": 2.4213478518604483e-06,
+      "loss": 0.9189,
+      "step": 22870
+    },
+    {
+      "epoch": 0.689156626506024,
+      "grad_norm": 4.5070303401318705,
+      "learning_rate": 2.4170822288370494e-06,
+      "loss": 1.0062,
+      "step": 22880
+    },
+    {
+      "epoch": 0.6894578313253013,
+      "grad_norm": 4.86138478304233,
+      "learning_rate": 2.412819168449588e-06,
+      "loss": 0.8871,
+      "step": 22890
+    },
+    {
+      "epoch": 0.6897590361445783,
+      "grad_norm": 2.049620763160923,
+      "learning_rate": 2.4085586749276484e-06,
+      "loss": 0.8572,
+      "step": 22900
+    },
+    {
+      "epoch": 0.6900602409638554,
+      "grad_norm": 4.422205159863266,
+      "learning_rate": 2.4043007524982648e-06,
+      "loss": 0.9579,
+      "step": 22910
+    },
+    {
+      "epoch": 0.6903614457831325,
+      "grad_norm": 118.33740839028648,
+      "learning_rate": 2.4000454053859312e-06,
+      "loss": 0.8765,
+      "step": 22920
+    },
+    {
+      "epoch": 0.6906626506024096,
+      "grad_norm": 5.252863704536765,
+      "learning_rate": 2.395792637812577e-06,
+      "loss": 0.9471,
+      "step": 22930
+    },
+    {
+      "epoch": 0.6909638554216867,
+      "grad_norm": 4.163202570653524,
+      "learning_rate": 2.391542453997578e-06,
+      "loss": 0.9042,
+      "step": 22940
+    },
+    {
+      "epoch": 0.6912650602409639,
+      "grad_norm": 4.409153591249788,
+      "learning_rate": 2.3872948581577427e-06,
+      "loss": 0.9913,
+      "step": 22950
+    },
+    {
+      "epoch": 0.691566265060241,
+      "grad_norm": 4.796527835663719,
+      "learning_rate": 2.383049854507314e-06,
+      "loss": 1.0036,
+      "step": 22960
+    },
+    {
+      "epoch": 0.6918674698795181,
+      "grad_norm": 1.9767573256592306,
+      "learning_rate": 2.3788074472579618e-06,
+      "loss": 0.9704,
+      "step": 22970
+    },
+    {
+      "epoch": 0.6921686746987952,
+      "grad_norm": 3.6699769998441423,
+      "learning_rate": 2.37456764061878e-06,
+      "loss": 0.8877,
+      "step": 22980
+    },
+    {
+      "epoch": 0.6924698795180723,
+      "grad_norm": 2.0295044424929753,
+      "learning_rate": 2.370330438796288e-06,
+      "loss": 0.9752,
+      "step": 22990
+    },
+    {
+      "epoch": 0.6927710843373494,
+      "grad_norm": 2.06139463612227,
+      "learning_rate": 2.366095845994411e-06,
+      "loss": 0.9895,
+      "step": 23000
+    },
+    {
+      "epoch": 0.6930722891566266,
+      "grad_norm": 6.780188527433399,
+      "learning_rate": 2.361863866414491e-06,
+      "loss": 0.8674,
+      "step": 23010
+    },
+    {
+      "epoch": 0.6933734939759036,
+      "grad_norm": 3.97929879716764,
+      "learning_rate": 2.357634504255276e-06,
+      "loss": 0.971,
+      "step": 23020
+    },
+    {
+      "epoch": 0.6936746987951807,
+      "grad_norm": 4.50396029068378,
+      "learning_rate": 2.3534077637129223e-06,
+      "loss": 1.0326,
+      "step": 23030
+    },
+    {
+      "epoch": 0.6939759036144578,
+      "grad_norm": 5.325511959961041,
+      "learning_rate": 2.349183648980977e-06,
+      "loss": 0.9477,
+      "step": 23040
+    },
+    {
+      "epoch": 0.6942771084337349,
+      "grad_norm": 5.396004785947532,
+      "learning_rate": 2.344962164250387e-06,
+      "loss": 0.9319,
+      "step": 23050
+    },
+    {
+      "epoch": 0.694578313253012,
+      "grad_norm": 2.1516068683771143,
+      "learning_rate": 2.3407433137094894e-06,
+      "loss": 0.9495,
+      "step": 23060
+    },
+    {
+      "epoch": 0.6948795180722892,
+      "grad_norm": 6.29960294901912,
+      "learning_rate": 2.3365271015440056e-06,
+      "loss": 0.9331,
+      "step": 23070
+    },
+    {
+      "epoch": 0.6951807228915663,
+      "grad_norm": 7.340386043870026,
+      "learning_rate": 2.3323135319370414e-06,
+      "loss": 0.9396,
+      "step": 23080
+    },
+    {
+      "epoch": 0.6954819277108434,
+      "grad_norm": 5.744500811222904,
+      "learning_rate": 2.3281026090690804e-06,
+      "loss": 0.8906,
+      "step": 23090
+    },
+    {
+      "epoch": 0.6957831325301205,
+      "grad_norm": 5.180073843839168,
+      "learning_rate": 2.3238943371179807e-06,
+      "loss": 0.8364,
+      "step": 23100
+    },
+    {
+      "epoch": 0.6960843373493976,
+      "grad_norm": 6.6853397673151695,
+      "learning_rate": 2.3196887202589686e-06,
+      "loss": 1.015,
+      "step": 23110
+    },
+    {
+      "epoch": 0.6963855421686747,
+      "grad_norm": 4.10735060104868,
+      "learning_rate": 2.315485762664637e-06,
+      "loss": 0.934,
+      "step": 23120
+    },
+    {
+      "epoch": 0.6966867469879519,
+      "grad_norm": 7.867614441488752,
+      "learning_rate": 2.3112854685049397e-06,
+      "loss": 0.8877,
+      "step": 23130
+    },
+    {
+      "epoch": 0.696987951807229,
+      "grad_norm": 7.277405452113129,
+      "learning_rate": 2.307087841947192e-06,
+      "loss": 1.0054,
+      "step": 23140
+    },
+    {
+      "epoch": 0.697289156626506,
+      "grad_norm": 4.4566872438858205,
+      "learning_rate": 2.3028928871560596e-06,
+      "loss": 0.9394,
+      "step": 23150
+    },
+    {
+      "epoch": 0.6975903614457831,
+      "grad_norm": 5.153560737661477,
+      "learning_rate": 2.298700608293552e-06,
+      "loss": 0.8738,
+      "step": 23160
+    },
+    {
+      "epoch": 0.6978915662650602,
+      "grad_norm": 1.8817787680846003,
+      "learning_rate": 2.294511009519034e-06,
+      "loss": 0.9579,
+      "step": 23170
+    },
+    {
+      "epoch": 0.6981927710843373,
+      "grad_norm": 1.827764606945967,
+      "learning_rate": 2.290324094989205e-06,
+      "loss": 0.9579,
+      "step": 23180
+    },
+    {
+      "epoch": 0.6984939759036145,
+      "grad_norm": 5.123900870063656,
+      "learning_rate": 2.286139868858103e-06,
+      "loss": 0.9587,
+      "step": 23190
+    },
+    {
+      "epoch": 0.6987951807228916,
+      "grad_norm": 5.61876482017539,
+      "learning_rate": 2.281958335277098e-06,
+      "loss": 0.9038,
+      "step": 23200
+    },
+    {
+      "epoch": 0.6990963855421687,
+      "grad_norm": 8.317484335056166,
+      "learning_rate": 2.277779498394889e-06,
+      "loss": 1.075,
+      "step": 23210
+    },
+    {
+      "epoch": 0.6993975903614458,
+      "grad_norm": 3.896827643658223,
+      "learning_rate": 2.273603362357498e-06,
+      "loss": 1.0394,
+      "step": 23220
+    },
+    {
+      "epoch": 0.6996987951807229,
+      "grad_norm": 1.8037348674155738,
+      "learning_rate": 2.2694299313082693e-06,
+      "loss": 0.8951,
+      "step": 23230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.3173602814544605,
+      "learning_rate": 2.265259209387867e-06,
+      "loss": 0.9407,
+      "step": 23240
+    },
+    {
+      "epoch": 0.7003012048192772,
+      "grad_norm": 3.9829492743839823,
+      "learning_rate": 2.261091200734258e-06,
+      "loss": 0.9511,
+      "step": 23250
+    },
+    {
+      "epoch": 0.7006024096385542,
+      "grad_norm": 4.057457437562215,
+      "learning_rate": 2.2569259094827243e-06,
+      "loss": 0.9642,
+      "step": 23260
+    },
+    {
+      "epoch": 0.7009036144578313,
+      "grad_norm": 2.0068270999768654,
+      "learning_rate": 2.252763339765849e-06,
+      "loss": 0.9944,
+      "step": 23270
+    },
+    {
+      "epoch": 0.7012048192771084,
+      "grad_norm": 4.6468511646938815,
+      "learning_rate": 2.248603495713518e-06,
+      "loss": 0.9326,
+      "step": 23280
+    },
+    {
+      "epoch": 0.7015060240963855,
+      "grad_norm": 5.193256640454499,
+      "learning_rate": 2.2444463814529115e-06,
+      "loss": 1.0054,
+      "step": 23290
+    },
+    {
+      "epoch": 0.7018072289156626,
+      "grad_norm": 5.146488303300006,
+      "learning_rate": 2.2402920011084995e-06,
+      "loss": 0.8454,
+      "step": 23300
+    },
+    {
+      "epoch": 0.7021084337349398,
+      "grad_norm": 8.582656638559628,
+      "learning_rate": 2.2361403588020404e-06,
+      "loss": 0.9063,
+      "step": 23310
+    },
+    {
+      "epoch": 0.7024096385542169,
+      "grad_norm": 6.552475722104356,
+      "learning_rate": 2.2319914586525776e-06,
+      "loss": 0.9258,
+      "step": 23320
+    },
+    {
+      "epoch": 0.702710843373494,
+      "grad_norm": 4.684625547165652,
+      "learning_rate": 2.2278453047764333e-06,
+      "loss": 0.8923,
+      "step": 23330
+    },
+    {
+      "epoch": 0.7030120481927711,
+      "grad_norm": 4.419180083014369,
+      "learning_rate": 2.223701901287203e-06,
+      "loss": 0.9415,
+      "step": 23340
+    },
+    {
+      "epoch": 0.7033132530120482,
+      "grad_norm": 19.834028959074853,
+      "learning_rate": 2.2195612522957555e-06,
+      "loss": 0.9937,
+      "step": 23350
+    },
+    {
+      "epoch": 0.7036144578313253,
+      "grad_norm": 4.22237959378623,
+      "learning_rate": 2.2154233619102266e-06,
+      "loss": 1.0087,
+      "step": 23360
+    },
+    {
+      "epoch": 0.7039156626506025,
+      "grad_norm": 2.1830823978849496,
+      "learning_rate": 2.2112882342360154e-06,
+      "loss": 0.8925,
+      "step": 23370
+    },
+    {
+      "epoch": 0.7042168674698795,
+      "grad_norm": 1.9840030888992062,
+      "learning_rate": 2.2071558733757757e-06,
+      "loss": 0.8569,
+      "step": 23380
+    },
+    {
+      "epoch": 0.7045180722891566,
+      "grad_norm": 3.9195650077889987,
+      "learning_rate": 2.2030262834294267e-06,
+      "loss": 0.9396,
+      "step": 23390
+    },
+    {
+      "epoch": 0.7048192771084337,
+      "grad_norm": 4.2134851341482795,
+      "learning_rate": 2.1988994684941256e-06,
+      "loss": 0.9038,
+      "step": 23400
+    },
+    {
+      "epoch": 0.7051204819277108,
+      "grad_norm": 6.404384021273373,
+      "learning_rate": 2.194775432664282e-06,
+      "loss": 0.8647,
+      "step": 23410
+    },
+    {
+      "epoch": 0.7054216867469879,
+      "grad_norm": 3.9554394561372503,
+      "learning_rate": 2.1906541800315535e-06,
+      "loss": 0.9661,
+      "step": 23420
+    },
+    {
+      "epoch": 0.7057228915662651,
+      "grad_norm": 5.5619589086390055,
+      "learning_rate": 2.1865357146848286e-06,
+      "loss": 0.9434,
+      "step": 23430
+    },
+    {
+      "epoch": 0.7060240963855422,
+      "grad_norm": 1.8748269889229263,
+      "learning_rate": 2.1824200407102332e-06,
+      "loss": 0.9844,
+      "step": 23440
+    },
+    {
+      "epoch": 0.7063253012048193,
+      "grad_norm": 5.4605865906090685,
+      "learning_rate": 2.178307162191124e-06,
+      "loss": 0.8661,
+      "step": 23450
+    },
+    {
+      "epoch": 0.7066265060240964,
+      "grad_norm": 3.641901518181943,
+      "learning_rate": 2.1741970832080846e-06,
+      "loss": 0.9045,
+      "step": 23460
+    },
+    {
+      "epoch": 0.7069277108433735,
+      "grad_norm": 1.9161795197656988,
+      "learning_rate": 2.170089807838919e-06,
+      "loss": 0.8912,
+      "step": 23470
+    },
+    {
+      "epoch": 0.7072289156626506,
+      "grad_norm": 5.085086927965741,
+      "learning_rate": 2.16598534015865e-06,
+      "loss": 0.9936,
+      "step": 23480
+    },
+    {
+      "epoch": 0.7075301204819278,
+      "grad_norm": 6.773889369343873,
+      "learning_rate": 2.161883684239522e-06,
+      "loss": 0.9811,
+      "step": 23490
+    },
+    {
+      "epoch": 0.7078313253012049,
+      "grad_norm": 3.6246561338536094,
+      "learning_rate": 2.1577848441509764e-06,
+      "loss": 0.9314,
+      "step": 23500
+    },
+    {
+      "epoch": 0.7081325301204819,
+      "grad_norm": 7.518998824039743,
+      "learning_rate": 2.1536888239596714e-06,
+      "loss": 0.8642,
+      "step": 23510
+    },
+    {
+      "epoch": 0.708433734939759,
+      "grad_norm": 1.8952967771917464,
+      "learning_rate": 2.149595627729461e-06,
+      "loss": 0.8739,
+      "step": 23520
+    },
+    {
+      "epoch": 0.7087349397590361,
+      "grad_norm": 5.191648142393608,
+      "learning_rate": 2.145505259521405e-06,
+      "loss": 0.9417,
+      "step": 23530
+    },
+    {
+      "epoch": 0.7090361445783132,
+      "grad_norm": 3.640175408075149,
+      "learning_rate": 2.141417723393752e-06,
+      "loss": 1.0436,
+      "step": 23540
+    },
+    {
+      "epoch": 0.7093373493975904,
+      "grad_norm": 5.168422152662087,
+      "learning_rate": 2.1373330234019374e-06,
+      "loss": 0.9235,
+      "step": 23550
+    },
+    {
+      "epoch": 0.7096385542168675,
+      "grad_norm": 5.069873338886798,
+      "learning_rate": 2.133251163598591e-06,
+      "loss": 0.9272,
+      "step": 23560
+    },
+    {
+      "epoch": 0.7099397590361446,
+      "grad_norm": 3.787466833212854,
+      "learning_rate": 2.129172148033519e-06,
+      "loss": 0.936,
+      "step": 23570
+    },
+    {
+      "epoch": 0.7102409638554217,
+      "grad_norm": 2.207250153121542,
+      "learning_rate": 2.125095980753708e-06,
+      "loss": 0.8954,
+      "step": 23580
+    },
+    {
+      "epoch": 0.7105421686746988,
+      "grad_norm": 3.9108531345414344,
+      "learning_rate": 2.121022665803317e-06,
+      "loss": 0.9488,
+      "step": 23590
+    },
+    {
+      "epoch": 0.7108433734939759,
+      "grad_norm": 5.025478982342228,
+      "learning_rate": 2.1169522072236774e-06,
+      "loss": 0.8871,
+      "step": 23600
+    },
+    {
+      "epoch": 0.7111445783132531,
+      "grad_norm": 3.8721950596206653,
+      "learning_rate": 2.1128846090532852e-06,
+      "loss": 1.0189,
+      "step": 23610
+    },
+    {
+      "epoch": 0.7114457831325302,
+      "grad_norm": 4.262500698458053,
+      "learning_rate": 2.1088198753277973e-06,
+      "loss": 0.9333,
+      "step": 23620
+    },
+    {
+      "epoch": 0.7117469879518072,
+      "grad_norm": 3.796946909369294,
+      "learning_rate": 2.1047580100800286e-06,
+      "loss": 0.8375,
+      "step": 23630
+    },
+    {
+      "epoch": 0.7120481927710843,
+      "grad_norm": 3.8388330766887484,
+      "learning_rate": 2.1006990173399556e-06,
+      "loss": 0.8744,
+      "step": 23640
+    },
+    {
+      "epoch": 0.7123493975903614,
+      "grad_norm": 1.9531315204735107,
+      "learning_rate": 2.0966429011346923e-06,
+      "loss": 0.855,
+      "step": 23650
+    },
+    {
+      "epoch": 0.7126506024096385,
+      "grad_norm": 5.284646124851012,
+      "learning_rate": 2.0925896654885057e-06,
+      "loss": 1.0146,
+      "step": 23660
+    },
+    {
+      "epoch": 0.7129518072289157,
+      "grad_norm": 1.9618156904959965,
+      "learning_rate": 2.0885393144228076e-06,
+      "loss": 0.8855,
+      "step": 23670
+    },
+    {
+      "epoch": 0.7132530120481928,
+      "grad_norm": 4.726929580568592,
+      "learning_rate": 2.0844918519561425e-06,
+      "loss": 0.9278,
+      "step": 23680
+    },
+    {
+      "epoch": 0.7135542168674699,
+      "grad_norm": 4.034876126551746,
+      "learning_rate": 2.0804472821041933e-06,
+      "loss": 1.0064,
+      "step": 23690
+    },
+    {
+      "epoch": 0.713855421686747,
+      "grad_norm": 4.2548237651244545,
+      "learning_rate": 2.0764056088797646e-06,
+      "loss": 0.9422,
+      "step": 23700
+    },
+    {
+      "epoch": 0.7141566265060241,
+      "grad_norm": 4.6760236051599735,
+      "learning_rate": 2.0723668362927996e-06,
+      "loss": 0.9429,
+      "step": 23710
+    },
+    {
+      "epoch": 0.7144578313253012,
+      "grad_norm": 5.734079540702262,
+      "learning_rate": 2.0683309683503543e-06,
+      "loss": 0.8584,
+      "step": 23720
+    },
+    {
+      "epoch": 0.7147590361445784,
+      "grad_norm": 4.416688870319067,
+      "learning_rate": 2.0642980090566065e-06,
+      "loss": 1.0068,
+      "step": 23730
+    },
+    {
+      "epoch": 0.7150602409638555,
+      "grad_norm": 4.93105304401678,
+      "learning_rate": 2.0602679624128473e-06,
+      "loss": 1.0497,
+      "step": 23740
+    },
+    {
+      "epoch": 0.7153614457831325,
+      "grad_norm": 6.316933040636142,
+      "learning_rate": 2.0562408324174786e-06,
+      "loss": 0.935,
+      "step": 23750
+    },
+    {
+      "epoch": 0.7156626506024096,
+      "grad_norm": 4.090206799769014,
+      "learning_rate": 2.0522166230660074e-06,
+      "loss": 0.9591,
+      "step": 23760
+    },
+    {
+      "epoch": 0.7159638554216867,
+      "grad_norm": 3.768301742741461,
+      "learning_rate": 2.0481953383510426e-06,
+      "loss": 0.8223,
+      "step": 23770
+    },
+    {
+      "epoch": 0.7162650602409638,
+      "grad_norm": 4.6501395199172,
+      "learning_rate": 2.0441769822622964e-06,
+      "loss": 0.8744,
+      "step": 23780
+    },
+    {
+      "epoch": 0.716566265060241,
+      "grad_norm": 4.157839214519675,
+      "learning_rate": 2.040161558786571e-06,
+      "loss": 0.9164,
+      "step": 23790
+    },
+    {
+      "epoch": 0.7168674698795181,
+      "grad_norm": 6.003499162384984,
+      "learning_rate": 2.036149071907754e-06,
+      "loss": 0.964,
+      "step": 23800
+    },
+    {
+      "epoch": 0.7171686746987952,
+      "grad_norm": 5.252391559160354,
+      "learning_rate": 2.032139525606831e-06,
+      "loss": 1.0058,
+      "step": 23810
+    },
+    {
+      "epoch": 0.7174698795180723,
+      "grad_norm": 5.43251284612108,
+      "learning_rate": 2.0281329238618617e-06,
+      "loss": 0.8611,
+      "step": 23820
+    },
+    {
+      "epoch": 0.7177710843373494,
+      "grad_norm": 5.133765137376551,
+      "learning_rate": 2.024129270647988e-06,
+      "loss": 1.0552,
+      "step": 23830
+    },
+    {
+      "epoch": 0.7180722891566265,
+      "grad_norm": 4.602515463311826,
+      "learning_rate": 2.020128569937424e-06,
+      "loss": 0.9639,
+      "step": 23840
+    },
+    {
+      "epoch": 0.7183734939759037,
+      "grad_norm": 5.153338287802595,
+      "learning_rate": 2.0161308256994565e-06,
+      "loss": 0.9653,
+      "step": 23850
+    },
+    {
+      "epoch": 0.7186746987951808,
+      "grad_norm": 4.945162207016891,
+      "learning_rate": 2.0121360419004383e-06,
+      "loss": 0.9646,
+      "step": 23860
+    },
+    {
+      "epoch": 0.7189759036144578,
+      "grad_norm": 5.3412116960756375,
+      "learning_rate": 2.0081442225037847e-06,
+      "loss": 0.9442,
+      "step": 23870
+    },
+    {
+      "epoch": 0.7192771084337349,
+      "grad_norm": 3.438474027398295,
+      "learning_rate": 2.0041553714699697e-06,
+      "loss": 0.8855,
+      "step": 23880
+    },
+    {
+      "epoch": 0.719578313253012,
+      "grad_norm": 4.528565600324454,
+      "learning_rate": 2.000169492756523e-06,
+      "loss": 0.8812,
+      "step": 23890
+    },
+    {
+      "epoch": 0.7198795180722891,
+      "grad_norm": 5.096291955102272,
+      "learning_rate": 1.9961865903180256e-06,
+      "loss": 0.8498,
+      "step": 23900
+    },
+    {
+      "epoch": 0.7201807228915663,
+      "grad_norm": 4.798050354142986,
+      "learning_rate": 1.9922066681061025e-06,
+      "loss": 1.003,
+      "step": 23910
+    },
+    {
+      "epoch": 0.7204819277108434,
+      "grad_norm": 2.0863242738648755,
+      "learning_rate": 1.9882297300694283e-06,
+      "loss": 0.9505,
+      "step": 23920
+    },
+    {
+      "epoch": 0.7207831325301205,
+      "grad_norm": 2.012477642800008,
+      "learning_rate": 1.984255780153711e-06,
+      "loss": 0.9697,
+      "step": 23930
+    },
+    {
+      "epoch": 0.7210843373493976,
+      "grad_norm": 4.348988317452686,
+      "learning_rate": 1.9802848223016976e-06,
+      "loss": 0.9739,
+      "step": 23940
+    },
+    {
+      "epoch": 0.7213855421686747,
+      "grad_norm": 4.641332928099159,
+      "learning_rate": 1.9763168604531594e-06,
+      "loss": 0.9557,
+      "step": 23950
+    },
+    {
+      "epoch": 0.7216867469879518,
+      "grad_norm": 1.886539282374377,
+      "learning_rate": 1.972351898544907e-06,
+      "loss": 0.9418,
+      "step": 23960
+    },
+    {
+      "epoch": 0.721987951807229,
+      "grad_norm": 2.065348782109405,
+      "learning_rate": 1.9683899405107647e-06,
+      "loss": 0.8994,
+      "step": 23970
+    },
+    {
+      "epoch": 0.7222891566265061,
+      "grad_norm": 8.660343030699208,
+      "learning_rate": 1.964430990281582e-06,
+      "loss": 0.9721,
+      "step": 23980
+    },
+    {
+      "epoch": 0.7225903614457831,
+      "grad_norm": 9.122400281515935,
+      "learning_rate": 1.9604750517852206e-06,
+      "loss": 0.8195,
+      "step": 23990
+    },
+    {
+      "epoch": 0.7228915662650602,
+      "grad_norm": 1.9745064009367765,
+      "learning_rate": 1.956522128946557e-06,
+      "loss": 0.811,
+      "step": 24000
+    },
+    {
+      "epoch": 0.7231927710843373,
+      "grad_norm": 4.9300901012920235,
+      "learning_rate": 1.9525722256874736e-06,
+      "loss": 0.9386,
+      "step": 24010
+    },
+    {
+      "epoch": 0.7234939759036144,
+      "grad_norm": 2.05432062396776,
+      "learning_rate": 1.9486253459268563e-06,
+      "loss": 0.8124,
+      "step": 24020
+    },
+    {
+      "epoch": 0.7237951807228916,
+      "grad_norm": 4.115157880050282,
+      "learning_rate": 1.9446814935805986e-06,
+      "loss": 1.0251,
+      "step": 24030
+    },
+    {
+      "epoch": 0.7240963855421687,
+      "grad_norm": 5.215746241768255,
+      "learning_rate": 1.940740672561579e-06,
+      "loss": 1.0002,
+      "step": 24040
+    },
+    {
+      "epoch": 0.7243975903614458,
+      "grad_norm": 1.8961637352206742,
+      "learning_rate": 1.936802886779674e-06,
+      "loss": 0.8946,
+      "step": 24050
+    },
+    {
+      "epoch": 0.7246987951807229,
+      "grad_norm": 4.0003651505033355,
+      "learning_rate": 1.9328681401417516e-06,
+      "loss": 0.947,
+      "step": 24060
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 2.071378430624273,
+      "learning_rate": 1.928936436551661e-06,
+      "loss": 0.9388,
+      "step": 24070
+    },
+    {
+      "epoch": 0.7253012048192771,
+      "grad_norm": 1.9878794103311674,
+      "learning_rate": 1.9250077799102323e-06,
+      "loss": 0.8782,
+      "step": 24080
+    },
+    {
+      "epoch": 0.7256024096385543,
+      "grad_norm": 4.456132427939708,
+      "learning_rate": 1.921082174115273e-06,
+      "loss": 0.8974,
+      "step": 24090
+    },
+    {
+      "epoch": 0.7259036144578314,
+      "grad_norm": 5.232572618042077,
+      "learning_rate": 1.9171596230615647e-06,
+      "loss": 0.9636,
+      "step": 24100
+    },
+    {
+      "epoch": 0.7262048192771084,
+      "grad_norm": 3.935718031287158,
+      "learning_rate": 1.913240130640856e-06,
+      "loss": 1.0512,
+      "step": 24110
+    },
+    {
+      "epoch": 0.7265060240963855,
+      "grad_norm": 3.922542714074301,
+      "learning_rate": 1.909323700741862e-06,
+      "loss": 0.9966,
+      "step": 24120
+    },
+    {
+      "epoch": 0.7268072289156626,
+      "grad_norm": 5.855878513334449,
+      "learning_rate": 1.905410337250264e-06,
+      "loss": 0.8188,
+      "step": 24130
+    },
+    {
+      "epoch": 0.7271084337349397,
+      "grad_norm": 5.932471616130165,
+      "learning_rate": 1.901500044048692e-06,
+      "loss": 0.9092,
+      "step": 24140
+    },
+    {
+      "epoch": 0.7274096385542169,
+      "grad_norm": 1.848641242811829,
+      "learning_rate": 1.8975928250167352e-06,
+      "loss": 0.9093,
+      "step": 24150
+    },
+    {
+      "epoch": 0.727710843373494,
+      "grad_norm": 5.0570936729903115,
+      "learning_rate": 1.8936886840309309e-06,
+      "loss": 0.8805,
+      "step": 24160
+    },
+    {
+      "epoch": 0.7280120481927711,
+      "grad_norm": 1.8956793833929124,
+      "learning_rate": 1.8897876249647672e-06,
+      "loss": 0.9153,
+      "step": 24170
+    },
+    {
+      "epoch": 0.7283132530120482,
+      "grad_norm": 5.456779690160975,
+      "learning_rate": 1.885889651688671e-06,
+      "loss": 0.952,
+      "step": 24180
+    },
+    {
+      "epoch": 0.7286144578313253,
+      "grad_norm": 8.698279686646263,
+      "learning_rate": 1.8819947680700035e-06,
+      "loss": 1.0162,
+      "step": 24190
+    },
+    {
+      "epoch": 0.7289156626506024,
+      "grad_norm": 4.846038386260658,
+      "learning_rate": 1.8781029779730646e-06,
+      "loss": 0.915,
+      "step": 24200
+    },
+    {
+      "epoch": 0.7292168674698796,
+      "grad_norm": 4.973815362553223,
+      "learning_rate": 1.874214285259089e-06,
+      "loss": 0.9525,
+      "step": 24210
+    },
+    {
+      "epoch": 0.7295180722891567,
+      "grad_norm": 5.31951673220082,
+      "learning_rate": 1.870328693786232e-06,
+      "loss": 0.9442,
+      "step": 24220
+    },
+    {
+      "epoch": 0.7298192771084338,
+      "grad_norm": 4.182075503567281,
+      "learning_rate": 1.8664462074095746e-06,
+      "loss": 0.9526,
+      "step": 24230
+    },
+    {
+      "epoch": 0.7301204819277108,
+      "grad_norm": 1.9328226925727665,
+      "learning_rate": 1.8625668299811162e-06,
+      "loss": 0.8531,
+      "step": 24240
+    },
+    {
+      "epoch": 0.7304216867469879,
+      "grad_norm": 2.1175901768808028,
+      "learning_rate": 1.8586905653497722e-06,
+      "loss": 0.8759,
+      "step": 24250
+    },
+    {
+      "epoch": 0.730722891566265,
+      "grad_norm": 4.939943011416337,
+      "learning_rate": 1.8548174173613703e-06,
+      "loss": 1.0196,
+      "step": 24260
+    },
+    {
+      "epoch": 0.7310240963855422,
+      "grad_norm": 4.539405539248213,
+      "learning_rate": 1.8509473898586432e-06,
+      "loss": 0.9586,
+      "step": 24270
+    },
+    {
+      "epoch": 0.7313253012048193,
+      "grad_norm": 6.273212072370957,
+      "learning_rate": 1.8470804866812354e-06,
+      "loss": 0.8979,
+      "step": 24280
+    },
+    {
+      "epoch": 0.7316265060240964,
+      "grad_norm": 5.963814795774179,
+      "learning_rate": 1.8432167116656802e-06,
+      "loss": 0.9229,
+      "step": 24290
+    },
+    {
+      "epoch": 0.7319277108433735,
+      "grad_norm": 2.0443692619301235,
+      "learning_rate": 1.8393560686454137e-06,
+      "loss": 0.9121,
+      "step": 24300
+    },
+    {
+      "epoch": 0.7322289156626506,
+      "grad_norm": 2.008271172152894,
+      "learning_rate": 1.835498561450767e-06,
+      "loss": 0.9958,
+      "step": 24310
+    },
+    {
+      "epoch": 0.7325301204819277,
+      "grad_norm": 5.858191431963845,
+      "learning_rate": 1.8316441939089552e-06,
+      "loss": 0.9358,
+      "step": 24320
+    },
+    {
+      "epoch": 0.7328313253012049,
+      "grad_norm": 2.0139009342278364,
+      "learning_rate": 1.8277929698440812e-06,
+      "loss": 0.9076,
+      "step": 24330
+    },
+    {
+      "epoch": 0.733132530120482,
+      "grad_norm": 4.292322829070784,
+      "learning_rate": 1.823944893077127e-06,
+      "loss": 0.8848,
+      "step": 24340
+    },
+    {
+      "epoch": 0.733433734939759,
+      "grad_norm": 21.60801004264943,
+      "learning_rate": 1.8200999674259533e-06,
+      "loss": 0.984,
+      "step": 24350
+    },
+    {
+      "epoch": 0.7337349397590361,
+      "grad_norm": 4.7668084802540935,
+      "learning_rate": 1.8162581967052945e-06,
+      "loss": 0.9263,
+      "step": 24360
+    },
+    {
+      "epoch": 0.7340361445783132,
+      "grad_norm": 7.522471792672586,
+      "learning_rate": 1.8124195847267533e-06,
+      "loss": 0.9809,
+      "step": 24370
+    },
+    {
+      "epoch": 0.7343373493975903,
+      "grad_norm": 5.633887550884112,
+      "learning_rate": 1.8085841352988043e-06,
+      "loss": 0.8985,
+      "step": 24380
+    },
+    {
+      "epoch": 0.7346385542168675,
+      "grad_norm": 3.317601174174788,
+      "learning_rate": 1.8047518522267748e-06,
+      "loss": 0.8532,
+      "step": 24390
+    },
+    {
+      "epoch": 0.7349397590361446,
+      "grad_norm": 3.786751412267653,
+      "learning_rate": 1.800922739312857e-06,
+      "loss": 0.9554,
+      "step": 24400
+    },
+    {
+      "epoch": 0.7352409638554217,
+      "grad_norm": 4.1541404138077445,
+      "learning_rate": 1.7970968003560958e-06,
+      "loss": 0.9615,
+      "step": 24410
+    },
+    {
+      "epoch": 0.7355421686746988,
+      "grad_norm": 6.0467211461924775,
+      "learning_rate": 1.7932740391523901e-06,
+      "loss": 0.9148,
+      "step": 24420
+    },
+    {
+      "epoch": 0.7358433734939759,
+      "grad_norm": 2.051997275765705,
+      "learning_rate": 1.7894544594944845e-06,
+      "loss": 0.8672,
+      "step": 24430
+    },
+    {
+      "epoch": 0.736144578313253,
+      "grad_norm": 6.513789781450807,
+      "learning_rate": 1.7856380651719623e-06,
+      "loss": 0.9379,
+      "step": 24440
+    },
+    {
+      "epoch": 0.7364457831325302,
+      "grad_norm": 4.272217606125275,
+      "learning_rate": 1.78182485997125e-06,
+      "loss": 1.0536,
+      "step": 24450
+    },
+    {
+      "epoch": 0.7367469879518073,
+      "grad_norm": 4.4105061160893015,
+      "learning_rate": 1.7780148476756148e-06,
+      "loss": 0.8337,
+      "step": 24460
+    },
+    {
+      "epoch": 0.7370481927710844,
+      "grad_norm": 2.100337684473239,
+      "learning_rate": 1.7742080320651496e-06,
+      "loss": 0.9263,
+      "step": 24470
+    },
+    {
+      "epoch": 0.7373493975903614,
+      "grad_norm": 25.125198507261835,
+      "learning_rate": 1.7704044169167772e-06,
+      "loss": 1.0516,
+      "step": 24480
+    },
+    {
+      "epoch": 0.7376506024096385,
+      "grad_norm": 4.191105317121492,
+      "learning_rate": 1.7666040060042456e-06,
+      "loss": 0.8798,
+      "step": 24490
+    },
+    {
+      "epoch": 0.7379518072289156,
+      "grad_norm": 2.2092416042783145,
+      "learning_rate": 1.7628068030981244e-06,
+      "loss": 0.9417,
+      "step": 24500
+    },
+    {
+      "epoch": 0.7382530120481928,
+      "grad_norm": 2.0755191122301753,
+      "learning_rate": 1.7590128119657995e-06,
+      "loss": 0.8016,
+      "step": 24510
+    },
+    {
+      "epoch": 0.7385542168674699,
+      "grad_norm": 7.201579467178426,
+      "learning_rate": 1.7552220363714689e-06,
+      "loss": 1.0353,
+      "step": 24520
+    },
+    {
+      "epoch": 0.738855421686747,
+      "grad_norm": 1.8705647557631861,
+      "learning_rate": 1.7514344800761474e-06,
+      "loss": 0.9524,
+      "step": 24530
+    },
+    {
+      "epoch": 0.7391566265060241,
+      "grad_norm": 4.779697056317637,
+      "learning_rate": 1.7476501468376438e-06,
+      "loss": 0.781,
+      "step": 24540
+    },
+    {
+      "epoch": 0.7394578313253012,
+      "grad_norm": 4.847883618628411,
+      "learning_rate": 1.7438690404105769e-06,
+      "loss": 0.9694,
+      "step": 24550
+    },
+    {
+      "epoch": 0.7397590361445783,
+      "grad_norm": 5.072446795267843,
+      "learning_rate": 1.740091164546366e-06,
+      "loss": 0.9741,
+      "step": 24560
+    },
+    {
+      "epoch": 0.7400602409638555,
+      "grad_norm": 6.690733857850204,
+      "learning_rate": 1.7363165229932194e-06,
+      "loss": 0.8466,
+      "step": 24570
+    },
+    {
+      "epoch": 0.7403614457831326,
+      "grad_norm": 4.6258476710645065,
+      "learning_rate": 1.7325451194961413e-06,
+      "loss": 0.8819,
+      "step": 24580
+    },
+    {
+      "epoch": 0.7406626506024097,
+      "grad_norm": 1.9028741288540398,
+      "learning_rate": 1.7287769577969166e-06,
+      "loss": 0.8972,
+      "step": 24590
+    },
+    {
+      "epoch": 0.7409638554216867,
+      "grad_norm": 5.927244604032882,
+      "learning_rate": 1.7250120416341227e-06,
+      "loss": 1.0344,
+      "step": 24600
+    },
+    {
+      "epoch": 0.7412650602409638,
+      "grad_norm": 3.7833930409397754,
+      "learning_rate": 1.7212503747431104e-06,
+      "loss": 0.9957,
+      "step": 24610
+    },
+    {
+      "epoch": 0.7415662650602409,
+      "grad_norm": 5.744623205380776,
+      "learning_rate": 1.7174919608560097e-06,
+      "loss": 0.9835,
+      "step": 24620
+    },
+    {
+      "epoch": 0.7418674698795181,
+      "grad_norm": 4.4215341591308865,
+      "learning_rate": 1.7137368037017215e-06,
+      "loss": 1.0573,
+      "step": 24630
+    },
+    {
+      "epoch": 0.7421686746987952,
+      "grad_norm": 7.040927591742778,
+      "learning_rate": 1.7099849070059177e-06,
+      "loss": 0.9191,
+      "step": 24640
+    },
+    {
+      "epoch": 0.7424698795180723,
+      "grad_norm": 2.1104550722668503,
+      "learning_rate": 1.7062362744910321e-06,
+      "loss": 0.9583,
+      "step": 24650
+    },
+    {
+      "epoch": 0.7427710843373494,
+      "grad_norm": 4.692163455292239,
+      "learning_rate": 1.702490909876262e-06,
+      "loss": 0.9226,
+      "step": 24660
+    },
+    {
+      "epoch": 0.7430722891566265,
+      "grad_norm": 1.8573874755026485,
+      "learning_rate": 1.6987488168775644e-06,
+      "loss": 0.7367,
+      "step": 24670
+    },
+    {
+      "epoch": 0.7433734939759036,
+      "grad_norm": 2.0192474808514405,
+      "learning_rate": 1.6950099992076485e-06,
+      "loss": 0.8619,
+      "step": 24680
+    },
+    {
+      "epoch": 0.7436746987951808,
+      "grad_norm": 7.776734395100246,
+      "learning_rate": 1.6912744605759707e-06,
+      "loss": 0.9502,
+      "step": 24690
+    },
+    {
+      "epoch": 0.7439759036144579,
+      "grad_norm": 1.9551977449250517,
+      "learning_rate": 1.6875422046887368e-06,
+      "loss": 0.7732,
+      "step": 24700
+    },
+    {
+      "epoch": 0.744277108433735,
+      "grad_norm": 10.390101692580956,
+      "learning_rate": 1.6838132352488995e-06,
+      "loss": 0.8319,
+      "step": 24710
+    },
+    {
+      "epoch": 0.744578313253012,
+      "grad_norm": 4.101030765300637,
+      "learning_rate": 1.680087555956146e-06,
+      "loss": 0.9334,
+      "step": 24720
+    },
+    {
+      "epoch": 0.7448795180722891,
+      "grad_norm": 6.581037482993133,
+      "learning_rate": 1.676365170506899e-06,
+      "loss": 0.8971,
+      "step": 24730
+    },
+    {
+      "epoch": 0.7451807228915662,
+      "grad_norm": 4.655724688667062,
+      "learning_rate": 1.6726460825943158e-06,
+      "loss": 0.9744,
+      "step": 24740
+    },
+    {
+      "epoch": 0.7454819277108434,
+      "grad_norm": 4.121847130103446,
+      "learning_rate": 1.6689302959082803e-06,
+      "loss": 0.988,
+      "step": 24750
+    },
+    {
+      "epoch": 0.7457831325301205,
+      "grad_norm": 3.818761445007322,
+      "learning_rate": 1.665217814135403e-06,
+      "loss": 0.9183,
+      "step": 24760
+    },
+    {
+      "epoch": 0.7460843373493976,
+      "grad_norm": 1.9836031457443304,
+      "learning_rate": 1.6615086409590125e-06,
+      "loss": 0.9151,
+      "step": 24770
+    },
+    {
+      "epoch": 0.7463855421686747,
+      "grad_norm": 5.138352253816458,
+      "learning_rate": 1.6578027800591578e-06,
+      "loss": 0.9541,
+      "step": 24780
+    },
+    {
+      "epoch": 0.7466867469879518,
+      "grad_norm": 3.8045675529228618,
+      "learning_rate": 1.6541002351125996e-06,
+      "loss": 0.8808,
+      "step": 24790
+    },
+    {
+      "epoch": 0.7469879518072289,
+      "grad_norm": 1.9958088955738258,
+      "learning_rate": 1.6504010097928074e-06,
+      "loss": 0.9771,
+      "step": 24800
+    },
+    {
+      "epoch": 0.7472891566265061,
+      "grad_norm": 4.234970270474972,
+      "learning_rate": 1.6467051077699632e-06,
+      "loss": 0.9119,
+      "step": 24810
+    },
+    {
+      "epoch": 0.7475903614457832,
+      "grad_norm": 2.0204691596181665,
+      "learning_rate": 1.6430125327109458e-06,
+      "loss": 0.9406,
+      "step": 24820
+    },
+    {
+      "epoch": 0.7478915662650603,
+      "grad_norm": 4.04979919903914,
+      "learning_rate": 1.6393232882793363e-06,
+      "loss": 0.9047,
+      "step": 24830
+    },
+    {
+      "epoch": 0.7481927710843373,
+      "grad_norm": 5.428702237379441,
+      "learning_rate": 1.6356373781354058e-06,
+      "loss": 0.9933,
+      "step": 24840
+    },
+    {
+      "epoch": 0.7484939759036144,
+      "grad_norm": 1.981495260166855,
+      "learning_rate": 1.6319548059361256e-06,
+      "loss": 0.8876,
+      "step": 24850
+    },
+    {
+      "epoch": 0.7487951807228915,
+      "grad_norm": 6.288176359296529,
+      "learning_rate": 1.6282755753351504e-06,
+      "loss": 0.8966,
+      "step": 24860
+    },
+    {
+      "epoch": 0.7490963855421687,
+      "grad_norm": 4.264526207466045,
+      "learning_rate": 1.6245996899828203e-06,
+      "loss": 0.8744,
+      "step": 24870
+    },
+    {
+      "epoch": 0.7493975903614458,
+      "grad_norm": 4.085038290569103,
+      "learning_rate": 1.6209271535261567e-06,
+      "loss": 0.9458,
+      "step": 24880
+    },
+    {
+      "epoch": 0.7496987951807229,
+      "grad_norm": 3.91845735237114,
+      "learning_rate": 1.6172579696088575e-06,
+      "loss": 0.9562,
+      "step": 24890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.83514146567771,
+      "learning_rate": 1.6135921418712959e-06,
+      "loss": 0.9126,
+      "step": 24900
+    },
+    {
+      "epoch": 0.7503012048192771,
+      "grad_norm": 3.788367410339539,
+      "learning_rate": 1.6099296739505116e-06,
+      "loss": 0.9895,
+      "step": 24910
+    },
+    {
+      "epoch": 0.7506024096385542,
+      "grad_norm": 5.607228355530344,
+      "learning_rate": 1.6062705694802199e-06,
+      "loss": 0.9194,
+      "step": 24920
+    },
+    {
+      "epoch": 0.7509036144578313,
+      "grad_norm": 1.9758269214360387,
+      "learning_rate": 1.602614832090788e-06,
+      "loss": 0.9182,
+      "step": 24930
+    },
+    {
+      "epoch": 0.7512048192771085,
+      "grad_norm": 1.9166091401241117,
+      "learning_rate": 1.5989624654092477e-06,
+      "loss": 0.9763,
+      "step": 24940
+    },
+    {
+      "epoch": 0.7515060240963856,
+      "grad_norm": 4.8752466540305655,
+      "learning_rate": 1.5953134730592862e-06,
+      "loss": 0.9461,
+      "step": 24950
+    },
+    {
+      "epoch": 0.7518072289156627,
+      "grad_norm": 5.198409199290026,
+      "learning_rate": 1.591667858661245e-06,
+      "loss": 0.9829,
+      "step": 24960
+    },
+    {
+      "epoch": 0.7521084337349397,
+      "grad_norm": 4.5765780169884005,
+      "learning_rate": 1.5880256258321107e-06,
+      "loss": 0.936,
+      "step": 24970
+    },
+    {
+      "epoch": 0.7524096385542168,
+      "grad_norm": 3.933693088792404,
+      "learning_rate": 1.584386778185517e-06,
+      "loss": 0.9314,
+      "step": 24980
+    },
+    {
+      "epoch": 0.7527108433734939,
+      "grad_norm": 4.366482613012169,
+      "learning_rate": 1.580751319331737e-06,
+      "loss": 0.878,
+      "step": 24990
+    },
+    {
+      "epoch": 0.7530120481927711,
+      "grad_norm": 1.9704952849185235,
+      "learning_rate": 1.5771192528776835e-06,
+      "loss": 0.911,
+      "step": 25000
+    },
+    {
+      "epoch": 0.7533132530120482,
+      "grad_norm": 4.153164024254792,
+      "learning_rate": 1.5734905824269031e-06,
+      "loss": 0.8928,
+      "step": 25010
+    },
+    {
+      "epoch": 0.7536144578313253,
+      "grad_norm": 2.011400930365946,
+      "learning_rate": 1.569865311579572e-06,
+      "loss": 0.832,
+      "step": 25020
+    },
+    {
+      "epoch": 0.7539156626506024,
+      "grad_norm": 6.482812341396677,
+      "learning_rate": 1.566243443932496e-06,
+      "loss": 1.0043,
+      "step": 25030
+    },
+    {
+      "epoch": 0.7542168674698795,
+      "grad_norm": 4.972915906899383,
+      "learning_rate": 1.5626249830791008e-06,
+      "loss": 0.886,
+      "step": 25040
+    },
+    {
+      "epoch": 0.7545180722891566,
+      "grad_norm": 4.700232307689609,
+      "learning_rate": 1.5590099326094333e-06,
+      "loss": 1.011,
+      "step": 25050
+    },
+    {
+      "epoch": 0.7548192771084338,
+      "grad_norm": 5.37168699961588,
+      "learning_rate": 1.555398296110161e-06,
+      "loss": 1.0653,
+      "step": 25060
+    },
+    {
+      "epoch": 0.7551204819277109,
+      "grad_norm": 1.9595477990699068,
+      "learning_rate": 1.5517900771645605e-06,
+      "loss": 0.855,
+      "step": 25070
+    },
+    {
+      "epoch": 0.755421686746988,
+      "grad_norm": 3.7930480459430465,
+      "learning_rate": 1.5481852793525143e-06,
+      "loss": 0.8334,
+      "step": 25080
+    },
+    {
+      "epoch": 0.755722891566265,
+      "grad_norm": 4.34052251223654,
+      "learning_rate": 1.5445839062505151e-06,
+      "loss": 0.8419,
+      "step": 25090
+    },
+    {
+      "epoch": 0.7560240963855421,
+      "grad_norm": 13.245431403369835,
+      "learning_rate": 1.5409859614316596e-06,
+      "loss": 0.8426,
+      "step": 25100
+    },
+    {
+      "epoch": 0.7563253012048192,
+      "grad_norm": 5.004260104855831,
+      "learning_rate": 1.5373914484656394e-06,
+      "loss": 0.8687,
+      "step": 25110
+    },
+    {
+      "epoch": 0.7566265060240964,
+      "grad_norm": 2.1527268242741924,
+      "learning_rate": 1.5338003709187416e-06,
+      "loss": 0.8651,
+      "step": 25120
+    },
+    {
+      "epoch": 0.7569277108433735,
+      "grad_norm": 3.7745336458304855,
+      "learning_rate": 1.5302127323538463e-06,
+      "loss": 0.9284,
+      "step": 25130
+    },
+    {
+      "epoch": 0.7572289156626506,
+      "grad_norm": 6.693207735348509,
+      "learning_rate": 1.52662853633042e-06,
+      "loss": 0.8987,
+      "step": 25140
+    },
+    {
+      "epoch": 0.7575301204819277,
+      "grad_norm": 6.261899342116549,
+      "learning_rate": 1.5230477864045152e-06,
+      "loss": 0.9968,
+      "step": 25150
+    },
+    {
+      "epoch": 0.7578313253012048,
+      "grad_norm": 2.022033652286476,
+      "learning_rate": 1.5194704861287629e-06,
+      "loss": 0.8978,
+      "step": 25160
+    },
+    {
+      "epoch": 0.7581325301204819,
+      "grad_norm": 2.136605032190541,
+      "learning_rate": 1.5158966390523783e-06,
+      "loss": 0.9367,
+      "step": 25170
+    },
+    {
+      "epoch": 0.7584337349397591,
+      "grad_norm": 5.391587866635553,
+      "learning_rate": 1.5123262487211404e-06,
+      "loss": 0.954,
+      "step": 25180
+    },
+    {
+      "epoch": 0.7587349397590362,
+      "grad_norm": 1.9421043730029601,
+      "learning_rate": 1.5087593186774052e-06,
+      "loss": 0.88,
+      "step": 25190
+    },
+    {
+      "epoch": 0.7590361445783133,
+      "grad_norm": 4.471435294641672,
+      "learning_rate": 1.5051958524600935e-06,
+      "loss": 0.9018,
+      "step": 25200
+    },
+    {
+      "epoch": 0.7593373493975903,
+      "grad_norm": 5.533919834547196,
+      "learning_rate": 1.5016358536046927e-06,
+      "loss": 0.9512,
+      "step": 25210
+    },
+    {
+      "epoch": 0.7596385542168674,
+      "grad_norm": 3.849065864864674,
+      "learning_rate": 1.4980793256432474e-06,
+      "loss": 0.8885,
+      "step": 25220
+    },
+    {
+      "epoch": 0.7599397590361445,
+      "grad_norm": 6.609822498760474,
+      "learning_rate": 1.4945262721043547e-06,
+      "loss": 0.8584,
+      "step": 25230
+    },
+    {
+      "epoch": 0.7602409638554217,
+      "grad_norm": 3.490521918344758,
+      "learning_rate": 1.4909766965131717e-06,
+      "loss": 0.9345,
+      "step": 25240
+    },
+    {
+      "epoch": 0.7605421686746988,
+      "grad_norm": 5.146801124625636,
+      "learning_rate": 1.4874306023914014e-06,
+      "loss": 0.9167,
+      "step": 25250
+    },
+    {
+      "epoch": 0.7608433734939759,
+      "grad_norm": 10.898026703823326,
+      "learning_rate": 1.4838879932572925e-06,
+      "loss": 0.9373,
+      "step": 25260
+    },
+    {
+      "epoch": 0.761144578313253,
+      "grad_norm": 4.133079212933276,
+      "learning_rate": 1.4803488726256366e-06,
+      "loss": 0.9397,
+      "step": 25270
+    },
+    {
+      "epoch": 0.7614457831325301,
+      "grad_norm": 6.394389690149214,
+      "learning_rate": 1.4768132440077643e-06,
+      "loss": 0.9996,
+      "step": 25280
+    },
+    {
+      "epoch": 0.7617469879518072,
+      "grad_norm": 6.617472460746821,
+      "learning_rate": 1.4732811109115409e-06,
+      "loss": 1.0381,
+      "step": 25290
+    },
+    {
+      "epoch": 0.7620481927710844,
+      "grad_norm": 7.540817796003784,
+      "learning_rate": 1.469752476841363e-06,
+      "loss": 0.9503,
+      "step": 25300
+    },
+    {
+      "epoch": 0.7623493975903615,
+      "grad_norm": 4.566984221797826,
+      "learning_rate": 1.4662273452981596e-06,
+      "loss": 0.9915,
+      "step": 25310
+    },
+    {
+      "epoch": 0.7626506024096386,
+      "grad_norm": 4.309362832475164,
+      "learning_rate": 1.462705719779382e-06,
+      "loss": 0.8111,
+      "step": 25320
+    },
+    {
+      "epoch": 0.7629518072289156,
+      "grad_norm": 2.075270544694661,
+      "learning_rate": 1.4591876037789999e-06,
+      "loss": 0.9168,
+      "step": 25330
+    },
+    {
+      "epoch": 0.7632530120481927,
+      "grad_norm": 4.078201471715807,
+      "learning_rate": 1.4556730007875037e-06,
+      "loss": 0.9053,
+      "step": 25340
+    },
+    {
+      "epoch": 0.7635542168674698,
+      "grad_norm": 5.59988233509037,
+      "learning_rate": 1.4521619142919025e-06,
+      "loss": 0.8567,
+      "step": 25350
+    },
+    {
+      "epoch": 0.763855421686747,
+      "grad_norm": 5.27547099207277,
+      "learning_rate": 1.4486543477757097e-06,
+      "loss": 0.9761,
+      "step": 25360
+    },
+    {
+      "epoch": 0.7641566265060241,
+      "grad_norm": 4.081857627172848,
+      "learning_rate": 1.4451503047189513e-06,
+      "loss": 0.8441,
+      "step": 25370
+    },
+    {
+      "epoch": 0.7644578313253012,
+      "grad_norm": 1.821742254320572,
+      "learning_rate": 1.441649788598154e-06,
+      "loss": 0.9515,
+      "step": 25380
+    },
+    {
+      "epoch": 0.7647590361445783,
+      "grad_norm": 4.395702922868068,
+      "learning_rate": 1.438152802886348e-06,
+      "loss": 0.9061,
+      "step": 25390
+    },
+    {
+      "epoch": 0.7650602409638554,
+      "grad_norm": 4.236608318380971,
+      "learning_rate": 1.4346593510530582e-06,
+      "loss": 0.9274,
+      "step": 25400
+    },
+    {
+      "epoch": 0.7653614457831325,
+      "grad_norm": 5.825112972513538,
+      "learning_rate": 1.4311694365643048e-06,
+      "loss": 0.9536,
+      "step": 25410
+    },
+    {
+      "epoch": 0.7656626506024097,
+      "grad_norm": 1.9536025996663187,
+      "learning_rate": 1.4276830628826023e-06,
+      "loss": 0.9233,
+      "step": 25420
+    },
+    {
+      "epoch": 0.7659638554216868,
+      "grad_norm": 3.983628338422071,
+      "learning_rate": 1.4242002334669436e-06,
+      "loss": 0.9305,
+      "step": 25430
+    },
+    {
+      "epoch": 0.7662650602409639,
+      "grad_norm": 3.5290911915940693,
+      "learning_rate": 1.4207209517728099e-06,
+      "loss": 0.9015,
+      "step": 25440
+    },
+    {
+      "epoch": 0.766566265060241,
+      "grad_norm": 4.276417649329165,
+      "learning_rate": 1.4172452212521664e-06,
+      "loss": 0.9532,
+      "step": 25450
+    },
+    {
+      "epoch": 0.766867469879518,
+      "grad_norm": 1.9858692635874682,
+      "learning_rate": 1.4137730453534487e-06,
+      "loss": 0.9327,
+      "step": 25460
+    },
+    {
+      "epoch": 0.7671686746987951,
+      "grad_norm": 4.63601285056426,
+      "learning_rate": 1.41030442752157e-06,
+      "loss": 0.9592,
+      "step": 25470
+    },
+    {
+      "epoch": 0.7674698795180723,
+      "grad_norm": 6.333195246507176,
+      "learning_rate": 1.4068393711979074e-06,
+      "loss": 0.878,
+      "step": 25480
+    },
+    {
+      "epoch": 0.7677710843373494,
+      "grad_norm": 5.1602068957904486,
+      "learning_rate": 1.403377879820312e-06,
+      "loss": 0.9782,
+      "step": 25490
+    },
+    {
+      "epoch": 0.7680722891566265,
+      "grad_norm": 5.632738739078735,
+      "learning_rate": 1.3999199568230942e-06,
+      "loss": 0.8566,
+      "step": 25500
+    },
+    {
+      "epoch": 0.7683734939759036,
+      "grad_norm": 6.291188487112138,
+      "learning_rate": 1.396465605637024e-06,
+      "loss": 1.0043,
+      "step": 25510
+    },
+    {
+      "epoch": 0.7686746987951807,
+      "grad_norm": 3.9444677940487876,
+      "learning_rate": 1.3930148296893276e-06,
+      "loss": 1.0077,
+      "step": 25520
+    },
+    {
+      "epoch": 0.7689759036144578,
+      "grad_norm": 4.184600759754945,
+      "learning_rate": 1.3895676324036844e-06,
+      "loss": 0.9654,
+      "step": 25530
+    },
+    {
+      "epoch": 0.769277108433735,
+      "grad_norm": 2.1370891402903993,
+      "learning_rate": 1.3861240172002243e-06,
+      "loss": 0.8956,
+      "step": 25540
+    },
+    {
+      "epoch": 0.7695783132530121,
+      "grad_norm": 3.8578609209664854,
+      "learning_rate": 1.3826839874955195e-06,
+      "loss": 0.8371,
+      "step": 25550
+    },
+    {
+      "epoch": 0.7698795180722892,
+      "grad_norm": 4.3944620783098465,
+      "learning_rate": 1.3792475467025911e-06,
+      "loss": 0.8816,
+      "step": 25560
+    },
+    {
+      "epoch": 0.7701807228915662,
+      "grad_norm": 6.249051779071241,
+      "learning_rate": 1.3758146982308963e-06,
+      "loss": 0.9595,
+      "step": 25570
+    },
+    {
+      "epoch": 0.7704819277108433,
+      "grad_norm": 1.7535816912474849,
+      "learning_rate": 1.372385445486324e-06,
+      "loss": 0.9491,
+      "step": 25580
+    },
+    {
+      "epoch": 0.7707831325301204,
+      "grad_norm": 5.212520780688571,
+      "learning_rate": 1.3689597918712e-06,
+      "loss": 0.8841,
+      "step": 25590
+    },
+    {
+      "epoch": 0.7710843373493976,
+      "grad_norm": 1.9867911549451127,
+      "learning_rate": 1.3655377407842813e-06,
+      "loss": 0.8906,
+      "step": 25600
+    },
+    {
+      "epoch": 0.7713855421686747,
+      "grad_norm": 4.224056730000401,
+      "learning_rate": 1.3621192956207473e-06,
+      "loss": 0.9832,
+      "step": 25610
+    },
+    {
+      "epoch": 0.7716867469879518,
+      "grad_norm": 2.0984283815411566,
+      "learning_rate": 1.3587044597721993e-06,
+      "loss": 0.9103,
+      "step": 25620
+    },
+    {
+      "epoch": 0.7719879518072289,
+      "grad_norm": 4.466657918298324,
+      "learning_rate": 1.3552932366266602e-06,
+      "loss": 0.9708,
+      "step": 25630
+    },
+    {
+      "epoch": 0.772289156626506,
+      "grad_norm": 4.419096139977014,
+      "learning_rate": 1.3518856295685662e-06,
+      "loss": 0.901,
+      "step": 25640
+    },
+    {
+      "epoch": 0.7725903614457831,
+      "grad_norm": 6.365450128341391,
+      "learning_rate": 1.3484816419787672e-06,
+      "loss": 0.9741,
+      "step": 25650
+    },
+    {
+      "epoch": 0.7728915662650603,
+      "grad_norm": 4.394197331045439,
+      "learning_rate": 1.3450812772345218e-06,
+      "loss": 0.887,
+      "step": 25660
+    },
+    {
+      "epoch": 0.7731927710843374,
+      "grad_norm": 2.023354498009975,
+      "learning_rate": 1.3416845387094935e-06,
+      "loss": 0.9366,
+      "step": 25670
+    },
+    {
+      "epoch": 0.7734939759036145,
+      "grad_norm": 4.904535265001817,
+      "learning_rate": 1.3382914297737492e-06,
+      "loss": 0.8131,
+      "step": 25680
+    },
+    {
+      "epoch": 0.7737951807228916,
+      "grad_norm": 7.310231445392361,
+      "learning_rate": 1.3349019537937524e-06,
+      "loss": 0.957,
+      "step": 25690
+    },
+    {
+      "epoch": 0.7740963855421686,
+      "grad_norm": 6.724978090106085,
+      "learning_rate": 1.3315161141323668e-06,
+      "loss": 0.909,
+      "step": 25700
+    },
+    {
+      "epoch": 0.7743975903614457,
+      "grad_norm": 2.0153041638234757,
+      "learning_rate": 1.3281339141488447e-06,
+      "loss": 0.9193,
+      "step": 25710
+    },
+    {
+      "epoch": 0.7746987951807229,
+      "grad_norm": 1.9314462357714723,
+      "learning_rate": 1.3247553571988287e-06,
+      "loss": 0.9334,
+      "step": 25720
+    },
+    {
+      "epoch": 0.775,
+      "grad_norm": 4.223220828327241,
+      "learning_rate": 1.321380446634342e-06,
+      "loss": 0.9743,
+      "step": 25730
+    },
+    {
+      "epoch": 0.7753012048192771,
+      "grad_norm": 3.9906748037488002,
+      "learning_rate": 1.3180091858037992e-06,
+      "loss": 0.8912,
+      "step": 25740
+    },
+    {
+      "epoch": 0.7756024096385542,
+      "grad_norm": 2.0000721620101864,
+      "learning_rate": 1.3146415780519866e-06,
+      "loss": 0.8386,
+      "step": 25750
+    },
+    {
+      "epoch": 0.7759036144578313,
+      "grad_norm": 4.821778037362633,
+      "learning_rate": 1.3112776267200695e-06,
+      "loss": 0.9574,
+      "step": 25760
+    },
+    {
+      "epoch": 0.7762048192771084,
+      "grad_norm": 4.030402565410141,
+      "learning_rate": 1.3079173351455843e-06,
+      "loss": 0.947,
+      "step": 25770
+    },
+    {
+      "epoch": 0.7765060240963856,
+      "grad_norm": 5.159710010514664,
+      "learning_rate": 1.304560706662436e-06,
+      "loss": 0.8636,
+      "step": 25780
+    },
+    {
+      "epoch": 0.7768072289156627,
+      "grad_norm": 5.878278774009832,
+      "learning_rate": 1.3012077446008969e-06,
+      "loss": 0.9339,
+      "step": 25790
+    },
+    {
+      "epoch": 0.7771084337349398,
+      "grad_norm": 4.866088555096551,
+      "learning_rate": 1.2978584522875981e-06,
+      "loss": 0.925,
+      "step": 25800
+    },
+    {
+      "epoch": 0.7774096385542169,
+      "grad_norm": 6.796986463528478,
+      "learning_rate": 1.294512833045537e-06,
+      "loss": 0.9362,
+      "step": 25810
+    },
+    {
+      "epoch": 0.7777108433734939,
+      "grad_norm": 6.357639347649964,
+      "learning_rate": 1.2911708901940572e-06,
+      "loss": 0.9411,
+      "step": 25820
+    },
+    {
+      "epoch": 0.778012048192771,
+      "grad_norm": 5.6679978434498155,
+      "learning_rate": 1.2878326270488622e-06,
+      "loss": 0.8914,
+      "step": 25830
+    },
+    {
+      "epoch": 0.7783132530120482,
+      "grad_norm": 2.0276944468996456,
+      "learning_rate": 1.2844980469220003e-06,
+      "loss": 0.9039,
+      "step": 25840
+    },
+    {
+      "epoch": 0.7786144578313253,
+      "grad_norm": 4.731582787629073,
+      "learning_rate": 1.28116715312187e-06,
+      "loss": 0.8664,
+      "step": 25850
+    },
+    {
+      "epoch": 0.7789156626506024,
+      "grad_norm": 4.425328161933482,
+      "learning_rate": 1.2778399489532085e-06,
+      "loss": 0.9914,
+      "step": 25860
+    },
+    {
+      "epoch": 0.7792168674698795,
+      "grad_norm": 2.01502210466282,
+      "learning_rate": 1.2745164377170937e-06,
+      "loss": 0.9123,
+      "step": 25870
+    },
+    {
+      "epoch": 0.7795180722891566,
+      "grad_norm": 3.658112612814461,
+      "learning_rate": 1.2711966227109402e-06,
+      "loss": 0.9839,
+      "step": 25880
+    },
+    {
+      "epoch": 0.7798192771084337,
+      "grad_norm": 1.9038679834427659,
+      "learning_rate": 1.2678805072284944e-06,
+      "loss": 0.8812,
+      "step": 25890
+    },
+    {
+      "epoch": 0.7801204819277109,
+      "grad_norm": 6.253973569990244,
+      "learning_rate": 1.2645680945598327e-06,
+      "loss": 0.9659,
+      "step": 25900
+    },
+    {
+      "epoch": 0.780421686746988,
+      "grad_norm": 3.9352510119647124,
+      "learning_rate": 1.261259387991358e-06,
+      "loss": 1.0112,
+      "step": 25910
+    },
+    {
+      "epoch": 0.7807228915662651,
+      "grad_norm": 3.68180517325047,
+      "learning_rate": 1.2579543908057956e-06,
+      "loss": 0.7992,
+      "step": 25920
+    },
+    {
+      "epoch": 0.7810240963855422,
+      "grad_norm": 4.6948677189212455,
+      "learning_rate": 1.254653106282191e-06,
+      "loss": 0.8736,
+      "step": 25930
+    },
+    {
+      "epoch": 0.7813253012048192,
+      "grad_norm": 4.42954429704751,
+      "learning_rate": 1.2513555376959047e-06,
+      "loss": 0.9852,
+      "step": 25940
+    },
+    {
+      "epoch": 0.7816265060240963,
+      "grad_norm": 2.137342924937682,
+      "learning_rate": 1.248061688318614e-06,
+      "loss": 0.8774,
+      "step": 25950
+    },
+    {
+      "epoch": 0.7819277108433735,
+      "grad_norm": 1.975982299553347,
+      "learning_rate": 1.2447715614183053e-06,
+      "loss": 0.9876,
+      "step": 25960
+    },
+    {
+      "epoch": 0.7822289156626506,
+      "grad_norm": 6.550414400072866,
+      "learning_rate": 1.2414851602592665e-06,
+      "loss": 0.9902,
+      "step": 25970
+    },
+    {
+      "epoch": 0.7825301204819277,
+      "grad_norm": 6.126236084567978,
+      "learning_rate": 1.2382024881020937e-06,
+      "loss": 0.988,
+      "step": 25980
+    },
+    {
+      "epoch": 0.7828313253012048,
+      "grad_norm": 2.105531679291249,
+      "learning_rate": 1.2349235482036849e-06,
+      "loss": 0.9077,
+      "step": 25990
+    },
+    {
+      "epoch": 0.7831325301204819,
+      "grad_norm": 2.021805743567685,
+      "learning_rate": 1.231648343817231e-06,
+      "loss": 0.898,
+      "step": 26000
+    },
+    {
+      "epoch": 0.783433734939759,
+      "grad_norm": 4.846061086067915,
+      "learning_rate": 1.2283768781922195e-06,
+      "loss": 0.9338,
+      "step": 26010
+    },
+    {
+      "epoch": 0.7837349397590362,
+      "grad_norm": 4.12075668655565,
+      "learning_rate": 1.2251091545744265e-06,
+      "loss": 0.9795,
+      "step": 26020
+    },
+    {
+      "epoch": 0.7840361445783133,
+      "grad_norm": 4.2233381745493785,
+      "learning_rate": 1.2218451762059174e-06,
+      "loss": 0.8972,
+      "step": 26030
+    },
+    {
+      "epoch": 0.7843373493975904,
+      "grad_norm": 4.236742207240406,
+      "learning_rate": 1.2185849463250405e-06,
+      "loss": 0.9595,
+      "step": 26040
+    },
+    {
+      "epoch": 0.7846385542168675,
+      "grad_norm": 4.45407343943746,
+      "learning_rate": 1.2153284681664235e-06,
+      "loss": 1.0062,
+      "step": 26050
+    },
+    {
+      "epoch": 0.7849397590361445,
+      "grad_norm": 6.0861661086256085,
+      "learning_rate": 1.212075744960979e-06,
+      "loss": 1.0037,
+      "step": 26060
+    },
+    {
+      "epoch": 0.7852409638554216,
+      "grad_norm": 3.70085945934955,
+      "learning_rate": 1.2088267799358833e-06,
+      "loss": 0.9244,
+      "step": 26070
+    },
+    {
+      "epoch": 0.7855421686746988,
+      "grad_norm": 3.9099942360199713,
+      "learning_rate": 1.2055815763145923e-06,
+      "loss": 0.9985,
+      "step": 26080
+    },
+    {
+      "epoch": 0.7858433734939759,
+      "grad_norm": 1.7590069586623232,
+      "learning_rate": 1.2023401373168248e-06,
+      "loss": 0.8263,
+      "step": 26090
+    },
+    {
+      "epoch": 0.786144578313253,
+      "grad_norm": 4.4828406707090735,
+      "learning_rate": 1.1991024661585704e-06,
+      "loss": 0.9566,
+      "step": 26100
+    },
+    {
+      "epoch": 0.7864457831325301,
+      "grad_norm": 1.8357938637751958,
+      "learning_rate": 1.1958685660520775e-06,
+      "loss": 0.8148,
+      "step": 26110
+    },
+    {
+      "epoch": 0.7867469879518072,
+      "grad_norm": 3.6533553251729853,
+      "learning_rate": 1.1926384402058478e-06,
+      "loss": 0.9665,
+      "step": 26120
+    },
+    {
+      "epoch": 0.7870481927710843,
+      "grad_norm": 4.057183995762749,
+      "learning_rate": 1.189412091824647e-06,
+      "loss": 0.972,
+      "step": 26130
+    },
+    {
+      "epoch": 0.7873493975903615,
+      "grad_norm": 4.130984563490733,
+      "learning_rate": 1.1861895241094895e-06,
+      "loss": 0.9189,
+      "step": 26140
+    },
+    {
+      "epoch": 0.7876506024096386,
+      "grad_norm": 4.488646157668425,
+      "learning_rate": 1.1829707402576375e-06,
+      "loss": 0.9066,
+      "step": 26150
+    },
+    {
+      "epoch": 0.7879518072289157,
+      "grad_norm": 8.370888803521982,
+      "learning_rate": 1.1797557434625994e-06,
+      "loss": 0.9105,
+      "step": 26160
+    },
+    {
+      "epoch": 0.7882530120481928,
+      "grad_norm": 2.1094400794436625,
+      "learning_rate": 1.1765445369141276e-06,
+      "loss": 0.7758,
+      "step": 26170
+    },
+    {
+      "epoch": 0.7885542168674698,
+      "grad_norm": 2.0947795704711227,
+      "learning_rate": 1.1733371237982133e-06,
+      "loss": 0.915,
+      "step": 26180
+    },
+    {
+      "epoch": 0.7888554216867469,
+      "grad_norm": 2.0468458622487566,
+      "learning_rate": 1.1701335072970825e-06,
+      "loss": 0.9879,
+      "step": 26190
+    },
+    {
+      "epoch": 0.7891566265060241,
+      "grad_norm": 2.0626807496444,
+      "learning_rate": 1.1669336905891987e-06,
+      "loss": 0.9199,
+      "step": 26200
+    },
+    {
+      "epoch": 0.7894578313253012,
+      "grad_norm": 4.506999692296973,
+      "learning_rate": 1.1637376768492524e-06,
+      "loss": 0.9,
+      "step": 26210
+    },
+    {
+      "epoch": 0.7897590361445783,
+      "grad_norm": 5.957970252927702,
+      "learning_rate": 1.16054546924816e-06,
+      "loss": 0.9151,
+      "step": 26220
+    },
+    {
+      "epoch": 0.7900602409638554,
+      "grad_norm": 5.405482708816962,
+      "learning_rate": 1.1573570709530608e-06,
+      "loss": 0.946,
+      "step": 26230
+    },
+    {
+      "epoch": 0.7903614457831325,
+      "grad_norm": 5.447065296117892,
+      "learning_rate": 1.1541724851273213e-06,
+      "loss": 0.9401,
+      "step": 26240
+    },
+    {
+      "epoch": 0.7906626506024096,
+      "grad_norm": 5.157807215195378,
+      "learning_rate": 1.1509917149305183e-06,
+      "loss": 0.9993,
+      "step": 26250
+    },
+    {
+      "epoch": 0.7909638554216868,
+      "grad_norm": 1.8861021646220957,
+      "learning_rate": 1.1478147635184482e-06,
+      "loss": 0.9656,
+      "step": 26260
+    },
+    {
+      "epoch": 0.7912650602409639,
+      "grad_norm": 5.129408790893935,
+      "learning_rate": 1.144641634043111e-06,
+      "loss": 1.0337,
+      "step": 26270
+    },
+    {
+      "epoch": 0.791566265060241,
+      "grad_norm": 4.1282872652533,
+      "learning_rate": 1.1414723296527247e-06,
+      "loss": 0.8983,
+      "step": 26280
+    },
+    {
+      "epoch": 0.7918674698795181,
+      "grad_norm": 1.8705992631642177,
+      "learning_rate": 1.1383068534917057e-06,
+      "loss": 0.913,
+      "step": 26290
+    },
+    {
+      "epoch": 0.7921686746987951,
+      "grad_norm": 4.052602527318772,
+      "learning_rate": 1.1351452087006737e-06,
+      "loss": 0.9466,
+      "step": 26300
+    },
+    {
+      "epoch": 0.7924698795180722,
+      "grad_norm": 1.822214358987177,
+      "learning_rate": 1.1319873984164475e-06,
+      "loss": 0.8877,
+      "step": 26310
+    },
+    {
+      "epoch": 0.7927710843373494,
+      "grad_norm": 4.221300445407804,
+      "learning_rate": 1.1288334257720412e-06,
+      "loss": 0.9885,
+      "step": 26320
+    },
+    {
+      "epoch": 0.7930722891566265,
+      "grad_norm": 2.197146428680317,
+      "learning_rate": 1.1256832938966622e-06,
+      "loss": 1.0005,
+      "step": 26330
+    },
+    {
+      "epoch": 0.7933734939759036,
+      "grad_norm": 4.5410982316680695,
+      "learning_rate": 1.1225370059157042e-06,
+      "loss": 0.8014,
+      "step": 26340
+    },
+    {
+      "epoch": 0.7936746987951807,
+      "grad_norm": 5.395947773614742,
+      "learning_rate": 1.1193945649507526e-06,
+      "loss": 0.9843,
+      "step": 26350
+    },
+    {
+      "epoch": 0.7939759036144578,
+      "grad_norm": 4.461932971950932,
+      "learning_rate": 1.1162559741195733e-06,
+      "loss": 0.8463,
+      "step": 26360
+    },
+    {
+      "epoch": 0.7942771084337349,
+      "grad_norm": 6.6869036841805025,
+      "learning_rate": 1.1131212365361083e-06,
+      "loss": 0.8218,
+      "step": 26370
+    },
+    {
+      "epoch": 0.7945783132530121,
+      "grad_norm": 4.854899512075135,
+      "learning_rate": 1.1099903553104835e-06,
+      "loss": 0.9564,
+      "step": 26380
+    },
+    {
+      "epoch": 0.7948795180722892,
+      "grad_norm": 3.688571325764524,
+      "learning_rate": 1.1068633335489943e-06,
+      "loss": 0.886,
+      "step": 26390
+    },
+    {
+      "epoch": 0.7951807228915663,
+      "grad_norm": 5.898503386801963,
+      "learning_rate": 1.1037401743541082e-06,
+      "loss": 0.9585,
+      "step": 26400
+    },
+    {
+      "epoch": 0.7954819277108434,
+      "grad_norm": 5.451504270955381,
+      "learning_rate": 1.1006208808244612e-06,
+      "loss": 0.9333,
+      "step": 26410
+    },
+    {
+      "epoch": 0.7957831325301205,
+      "grad_norm": 4.581186674401167,
+      "learning_rate": 1.0975054560548521e-06,
+      "loss": 0.9347,
+      "step": 26420
+    },
+    {
+      "epoch": 0.7960843373493975,
+      "grad_norm": 4.238600543352366,
+      "learning_rate": 1.0943939031362432e-06,
+      "loss": 0.9473,
+      "step": 26430
+    },
+    {
+      "epoch": 0.7963855421686747,
+      "grad_norm": 4.2327094364943365,
+      "learning_rate": 1.0912862251557532e-06,
+      "loss": 0.8378,
+      "step": 26440
+    },
+    {
+      "epoch": 0.7966867469879518,
+      "grad_norm": 2.094152982256059,
+      "learning_rate": 1.0881824251966593e-06,
+      "loss": 0.8707,
+      "step": 26450
+    },
+    {
+      "epoch": 0.7969879518072289,
+      "grad_norm": 12.966820603992703,
+      "learning_rate": 1.0850825063383912e-06,
+      "loss": 0.8267,
+      "step": 26460
+    },
+    {
+      "epoch": 0.797289156626506,
+      "grad_norm": 5.9430399670813285,
+      "learning_rate": 1.0819864716565227e-06,
+      "loss": 0.8869,
+      "step": 26470
+    },
+    {
+      "epoch": 0.7975903614457831,
+      "grad_norm": 6.82874328676258,
+      "learning_rate": 1.0788943242227768e-06,
+      "loss": 0.8684,
+      "step": 26480
+    },
+    {
+      "epoch": 0.7978915662650602,
+      "grad_norm": 1.9055393103703155,
+      "learning_rate": 1.0758060671050247e-06,
+      "loss": 0.8856,
+      "step": 26490
+    },
+    {
+      "epoch": 0.7981927710843374,
+      "grad_norm": 4.630105577091288,
+      "learning_rate": 1.072721703367271e-06,
+      "loss": 1.0358,
+      "step": 26500
+    },
+    {
+      "epoch": 0.7984939759036145,
+      "grad_norm": 4.965017021441382,
+      "learning_rate": 1.0696412360696617e-06,
+      "loss": 1.0116,
+      "step": 26510
+    },
+    {
+      "epoch": 0.7987951807228916,
+      "grad_norm": 1.9918649889001563,
+      "learning_rate": 1.0665646682684722e-06,
+      "loss": 0.8275,
+      "step": 26520
+    },
+    {
+      "epoch": 0.7990963855421687,
+      "grad_norm": 2.012178059055671,
+      "learning_rate": 1.063492003016115e-06,
+      "loss": 0.8861,
+      "step": 26530
+    },
+    {
+      "epoch": 0.7993975903614458,
+      "grad_norm": 6.9462012180711215,
+      "learning_rate": 1.0604232433611272e-06,
+      "loss": 0.9874,
+      "step": 26540
+    },
+    {
+      "epoch": 0.7996987951807228,
+      "grad_norm": 4.515389197051747,
+      "learning_rate": 1.057358392348171e-06,
+      "loss": 1.0043,
+      "step": 26550
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.028675444764515,
+      "learning_rate": 1.0542974530180327e-06,
+      "loss": 0.8569,
+      "step": 26560
+    },
+    {
+      "epoch": 0.8003012048192771,
+      "grad_norm": 2.098557752811368,
+      "learning_rate": 1.051240428407615e-06,
+      "loss": 0.8994,
+      "step": 26570
+    },
+    {
+      "epoch": 0.8006024096385542,
+      "grad_norm": 4.527629327335188,
+      "learning_rate": 1.0481873215499377e-06,
+      "loss": 0.9069,
+      "step": 26580
+    },
+    {
+      "epoch": 0.8009036144578313,
+      "grad_norm": 4.538578866914151,
+      "learning_rate": 1.0451381354741336e-06,
+      "loss": 0.9278,
+      "step": 26590
+    },
+    {
+      "epoch": 0.8012048192771084,
+      "grad_norm": 3.337709672653328,
+      "learning_rate": 1.042092873205447e-06,
+      "loss": 0.8074,
+      "step": 26600
+    },
+    {
+      "epoch": 0.8015060240963855,
+      "grad_norm": 10.619111251584942,
+      "learning_rate": 1.0390515377652288e-06,
+      "loss": 0.9584,
+      "step": 26610
+    },
+    {
+      "epoch": 0.8018072289156627,
+      "grad_norm": 4.81985411250438,
+      "learning_rate": 1.0360141321709288e-06,
+      "loss": 0.9831,
+      "step": 26620
+    },
+    {
+      "epoch": 0.8021084337349398,
+      "grad_norm": 3.3686071971015323,
+      "learning_rate": 1.0329806594361059e-06,
+      "loss": 0.8553,
+      "step": 26630
+    },
+    {
+      "epoch": 0.8024096385542169,
+      "grad_norm": 4.826253370488143,
+      "learning_rate": 1.0299511225704113e-06,
+      "loss": 0.9601,
+      "step": 26640
+    },
+    {
+      "epoch": 0.802710843373494,
+      "grad_norm": 4.7909237230489214,
+      "learning_rate": 1.0269255245795938e-06,
+      "loss": 0.9377,
+      "step": 26650
+    },
+    {
+      "epoch": 0.803012048192771,
+      "grad_norm": 3.985511976147391,
+      "learning_rate": 1.0239038684654928e-06,
+      "loss": 1.0171,
+      "step": 26660
+    },
+    {
+      "epoch": 0.8033132530120481,
+      "grad_norm": 4.668711471278927,
+      "learning_rate": 1.020886157226038e-06,
+      "loss": 0.9964,
+      "step": 26670
+    },
+    {
+      "epoch": 0.8036144578313253,
+      "grad_norm": 1.923426713562242,
+      "learning_rate": 1.0178723938552449e-06,
+      "loss": 0.9291,
+      "step": 26680
+    },
+    {
+      "epoch": 0.8039156626506024,
+      "grad_norm": 4.4701170492883815,
+      "learning_rate": 1.01486258134321e-06,
+      "loss": 0.8607,
+      "step": 26690
+    },
+    {
+      "epoch": 0.8042168674698795,
+      "grad_norm": 4.72559958319008,
+      "learning_rate": 1.011856722676116e-06,
+      "loss": 0.9948,
+      "step": 26700
+    },
+    {
+      "epoch": 0.8045180722891566,
+      "grad_norm": 6.178153048399851,
+      "learning_rate": 1.0088548208362149e-06,
+      "loss": 0.978,
+      "step": 26710
+    },
+    {
+      "epoch": 0.8048192771084337,
+      "grad_norm": 3.9557340854033525,
+      "learning_rate": 1.0058568788018374e-06,
+      "loss": 0.9143,
+      "step": 26720
+    },
+    {
+      "epoch": 0.8051204819277108,
+      "grad_norm": 6.397975028572285,
+      "learning_rate": 1.0028628995473844e-06,
+      "loss": 0.955,
+      "step": 26730
+    },
+    {
+      "epoch": 0.805421686746988,
+      "grad_norm": 3.949178383445514,
+      "learning_rate": 9.998728860433277e-07,
+      "loss": 0.9726,
+      "step": 26740
+    },
+    {
+      "epoch": 0.8057228915662651,
+      "grad_norm": 5.8382587758648,
+      "learning_rate": 9.968868412562012e-07,
+      "loss": 0.9155,
+      "step": 26750
+    },
+    {
+      "epoch": 0.8060240963855422,
+      "grad_norm": 1.7851946741763713,
+      "learning_rate": 9.939047681486018e-07,
+      "loss": 0.905,
+      "step": 26760
+    },
+    {
+      "epoch": 0.8063253012048193,
+      "grad_norm": 3.9588179541706343,
+      "learning_rate": 9.90926669679187e-07,
+      "loss": 0.8882,
+      "step": 26770
+    },
+    {
+      "epoch": 0.8066265060240964,
+      "grad_norm": 5.512485201013972,
+      "learning_rate": 9.879525488026704e-07,
+      "loss": 0.9073,
+      "step": 26780
+    },
+    {
+      "epoch": 0.8069277108433734,
+      "grad_norm": 4.473173363507019,
+      "learning_rate": 9.84982408469819e-07,
+      "loss": 0.8525,
+      "step": 26790
+    },
+    {
+      "epoch": 0.8072289156626506,
+      "grad_norm": 4.623229976299066,
+      "learning_rate": 9.820162516274512e-07,
+      "loss": 0.9826,
+      "step": 26800
+    },
+    {
+      "epoch": 0.8075301204819277,
+      "grad_norm": 2.0461130174625115,
+      "learning_rate": 9.790540812184319e-07,
+      "loss": 0.908,
+      "step": 26810
+    },
+    {
+      "epoch": 0.8078313253012048,
+      "grad_norm": 1.821179105196857,
+      "learning_rate": 9.760959001816723e-07,
+      "loss": 0.8484,
+      "step": 26820
+    },
+    {
+      "epoch": 0.8081325301204819,
+      "grad_norm": 2.122677229149528,
+      "learning_rate": 9.731417114521246e-07,
+      "loss": 0.8311,
+      "step": 26830
+    },
+    {
+      "epoch": 0.808433734939759,
+      "grad_norm": 1.9949072899409344,
+      "learning_rate": 9.70191517960779e-07,
+      "loss": 0.9034,
+      "step": 26840
+    },
+    {
+      "epoch": 0.8087349397590361,
+      "grad_norm": 4.051532263239917,
+      "learning_rate": 9.672453226346673e-07,
+      "loss": 0.8529,
+      "step": 26850
+    },
+    {
+      "epoch": 0.8090361445783133,
+      "grad_norm": 5.001389372697416,
+      "learning_rate": 9.643031283968467e-07,
+      "loss": 0.9398,
+      "step": 26860
+    },
+    {
+      "epoch": 0.8093373493975904,
+      "grad_norm": 1.995055674959236,
+      "learning_rate": 9.61364938166408e-07,
+      "loss": 0.9159,
+      "step": 26870
+    },
+    {
+      "epoch": 0.8096385542168675,
+      "grad_norm": 7.349946295979108,
+      "learning_rate": 9.584307548584725e-07,
+      "loss": 0.876,
+      "step": 26880
+    },
+    {
+      "epoch": 0.8099397590361446,
+      "grad_norm": 1.9244608147555216,
+      "learning_rate": 9.555005813841827e-07,
+      "loss": 0.8988,
+      "step": 26890
+    },
+    {
+      "epoch": 0.8102409638554217,
+      "grad_norm": 7.079759318224367,
+      "learning_rate": 9.525744206507032e-07,
+      "loss": 0.8887,
+      "step": 26900
+    },
+    {
+      "epoch": 0.8105421686746987,
+      "grad_norm": 4.764109926375551,
+      "learning_rate": 9.496522755612175e-07,
+      "loss": 0.9922,
+      "step": 26910
+    },
+    {
+      "epoch": 0.810843373493976,
+      "grad_norm": 3.475587394695104,
+      "learning_rate": 9.467341490149251e-07,
+      "loss": 0.9398,
+      "step": 26920
+    },
+    {
+      "epoch": 0.811144578313253,
+      "grad_norm": 3.433999525350799,
+      "learning_rate": 9.438200439070388e-07,
+      "loss": 0.8413,
+      "step": 26930
+    },
+    {
+      "epoch": 0.8114457831325301,
+      "grad_norm": 6.004915672093086,
+      "learning_rate": 9.409099631287799e-07,
+      "loss": 0.9537,
+      "step": 26940
+    },
+    {
+      "epoch": 0.8117469879518072,
+      "grad_norm": 5.08164683715821,
+      "learning_rate": 9.380039095673821e-07,
+      "loss": 0.8516,
+      "step": 26950
+    },
+    {
+      "epoch": 0.8120481927710843,
+      "grad_norm": 1.9924714282201934,
+      "learning_rate": 9.351018861060762e-07,
+      "loss": 0.8413,
+      "step": 26960
+    },
+    {
+      "epoch": 0.8123493975903614,
+      "grad_norm": 3.730570459866916,
+      "learning_rate": 9.322038956240992e-07,
+      "loss": 0.8713,
+      "step": 26970
+    },
+    {
+      "epoch": 0.8126506024096386,
+      "grad_norm": 4.125942399348648,
+      "learning_rate": 9.293099409966844e-07,
+      "loss": 0.9627,
+      "step": 26980
+    },
+    {
+      "epoch": 0.8129518072289157,
+      "grad_norm": 4.636791732008932,
+      "learning_rate": 9.264200250950645e-07,
+      "loss": 0.841,
+      "step": 26990
+    },
+    {
+      "epoch": 0.8132530120481928,
+      "grad_norm": 5.350309927339916,
+      "learning_rate": 9.235341507864637e-07,
+      "loss": 1.0298,
+      "step": 27000
+    },
+    {
+      "epoch": 0.8135542168674699,
+      "grad_norm": 4.087683638110185,
+      "learning_rate": 9.206523209340912e-07,
+      "loss": 0.9649,
+      "step": 27010
+    },
+    {
+      "epoch": 0.813855421686747,
+      "grad_norm": 4.769646376143996,
+      "learning_rate": 9.177745383971515e-07,
+      "loss": 0.9672,
+      "step": 27020
+    },
+    {
+      "epoch": 0.814156626506024,
+      "grad_norm": 5.657734836139984,
+      "learning_rate": 9.14900806030829e-07,
+      "loss": 0.898,
+      "step": 27030
+    },
+    {
+      "epoch": 0.8144578313253013,
+      "grad_norm": 4.37602630907073,
+      "learning_rate": 9.120311266862908e-07,
+      "loss": 0.9639,
+      "step": 27040
+    },
+    {
+      "epoch": 0.8147590361445783,
+      "grad_norm": 3.7446863659705256,
+      "learning_rate": 9.091655032106822e-07,
+      "loss": 1.0256,
+      "step": 27050
+    },
+    {
+      "epoch": 0.8150602409638554,
+      "grad_norm": 5.369127240316652,
+      "learning_rate": 9.063039384471256e-07,
+      "loss": 1.0176,
+      "step": 27060
+    },
+    {
+      "epoch": 0.8153614457831325,
+      "grad_norm": 4.056586251695432,
+      "learning_rate": 9.034464352347156e-07,
+      "loss": 0.8408,
+      "step": 27070
+    },
+    {
+      "epoch": 0.8156626506024096,
+      "grad_norm": 6.484866556665476,
+      "learning_rate": 9.005929964085169e-07,
+      "loss": 0.9512,
+      "step": 27080
+    },
+    {
+      "epoch": 0.8159638554216867,
+      "grad_norm": 2.0939348844131223,
+      "learning_rate": 8.977436247995619e-07,
+      "loss": 0.935,
+      "step": 27090
+    },
+    {
+      "epoch": 0.8162650602409639,
+      "grad_norm": 4.1015451369900395,
+      "learning_rate": 8.948983232348518e-07,
+      "loss": 0.9712,
+      "step": 27100
+    },
+    {
+      "epoch": 0.816566265060241,
+      "grad_norm": 12.325101995524495,
+      "learning_rate": 8.920570945373414e-07,
+      "loss": 0.8931,
+      "step": 27110
+    },
+    {
+      "epoch": 0.8168674698795181,
+      "grad_norm": 7.337973386375015,
+      "learning_rate": 8.892199415259501e-07,
+      "loss": 0.8685,
+      "step": 27120
+    },
+    {
+      "epoch": 0.8171686746987952,
+      "grad_norm": 1.9946831973745822,
+      "learning_rate": 8.863868670155545e-07,
+      "loss": 0.8989,
+      "step": 27130
+    },
+    {
+      "epoch": 0.8174698795180723,
+      "grad_norm": 4.4044825487754045,
+      "learning_rate": 8.835578738169825e-07,
+      "loss": 0.9525,
+      "step": 27140
+    },
+    {
+      "epoch": 0.8177710843373494,
+      "grad_norm": 5.400466321171892,
+      "learning_rate": 8.807329647370139e-07,
+      "loss": 0.8243,
+      "step": 27150
+    },
+    {
+      "epoch": 0.8180722891566266,
+      "grad_norm": 5.935745729566731,
+      "learning_rate": 8.779121425783721e-07,
+      "loss": 0.9381,
+      "step": 27160
+    },
+    {
+      "epoch": 0.8183734939759036,
+      "grad_norm": 4.614854388393017,
+      "learning_rate": 8.750954101397335e-07,
+      "loss": 1.0315,
+      "step": 27170
+    },
+    {
+      "epoch": 0.8186746987951807,
+      "grad_norm": 1.9215296013732723,
+      "learning_rate": 8.722827702157105e-07,
+      "loss": 1.0177,
+      "step": 27180
+    },
+    {
+      "epoch": 0.8189759036144578,
+      "grad_norm": 1.9749572471131391,
+      "learning_rate": 8.694742255968586e-07,
+      "loss": 0.9112,
+      "step": 27190
+    },
+    {
+      "epoch": 0.8192771084337349,
+      "grad_norm": 2.0190811152014208,
+      "learning_rate": 8.666697790696682e-07,
+      "loss": 0.8902,
+      "step": 27200
+    },
+    {
+      "epoch": 0.819578313253012,
+      "grad_norm": 4.08313914158281,
+      "learning_rate": 8.638694334165642e-07,
+      "loss": 0.9238,
+      "step": 27210
+    },
+    {
+      "epoch": 0.8198795180722892,
+      "grad_norm": 4.244208090872529,
+      "learning_rate": 8.610731914159037e-07,
+      "loss": 0.9479,
+      "step": 27220
+    },
+    {
+      "epoch": 0.8201807228915663,
+      "grad_norm": 4.531661928586746,
+      "learning_rate": 8.582810558419708e-07,
+      "loss": 0.9322,
+      "step": 27230
+    },
+    {
+      "epoch": 0.8204819277108434,
+      "grad_norm": 4.523393454805247,
+      "learning_rate": 8.554930294649777e-07,
+      "loss": 0.9203,
+      "step": 27240
+    },
+    {
+      "epoch": 0.8207831325301205,
+      "grad_norm": 1.8254250854960814,
+      "learning_rate": 8.527091150510592e-07,
+      "loss": 0.9018,
+      "step": 27250
+    },
+    {
+      "epoch": 0.8210843373493976,
+      "grad_norm": 2.010002656092585,
+      "learning_rate": 8.499293153622657e-07,
+      "loss": 0.8595,
+      "step": 27260
+    },
+    {
+      "epoch": 0.8213855421686747,
+      "grad_norm": 2.0340982050325973,
+      "learning_rate": 8.471536331565733e-07,
+      "loss": 0.9179,
+      "step": 27270
+    },
+    {
+      "epoch": 0.8216867469879519,
+      "grad_norm": 1.9881953867347788,
+      "learning_rate": 8.443820711878664e-07,
+      "loss": 0.8816,
+      "step": 27280
+    },
+    {
+      "epoch": 0.821987951807229,
+      "grad_norm": 4.854072714959214,
+      "learning_rate": 8.416146322059449e-07,
+      "loss": 0.8966,
+      "step": 27290
+    },
+    {
+      "epoch": 0.822289156626506,
+      "grad_norm": 22.399579502033067,
+      "learning_rate": 8.388513189565161e-07,
+      "loss": 0.9374,
+      "step": 27300
+    },
+    {
+      "epoch": 0.8225903614457831,
+      "grad_norm": 14.495352721079653,
+      "learning_rate": 8.360921341811956e-07,
+      "loss": 0.9245,
+      "step": 27310
+    },
+    {
+      "epoch": 0.8228915662650602,
+      "grad_norm": 7.079842567759796,
+      "learning_rate": 8.333370806175017e-07,
+      "loss": 0.8785,
+      "step": 27320
+    },
+    {
+      "epoch": 0.8231927710843373,
+      "grad_norm": 6.054767404580933,
+      "learning_rate": 8.305861609988553e-07,
+      "loss": 0.8306,
+      "step": 27330
+    },
+    {
+      "epoch": 0.8234939759036145,
+      "grad_norm": 5.258622117514524,
+      "learning_rate": 8.278393780545746e-07,
+      "loss": 0.9699,
+      "step": 27340
+    },
+    {
+      "epoch": 0.8237951807228916,
+      "grad_norm": 2.036278786840511,
+      "learning_rate": 8.250967345098731e-07,
+      "loss": 0.9187,
+      "step": 27350
+    },
+    {
+      "epoch": 0.8240963855421687,
+      "grad_norm": 2.0687201110763773,
+      "learning_rate": 8.223582330858599e-07,
+      "loss": 0.9041,
+      "step": 27360
+    },
+    {
+      "epoch": 0.8243975903614458,
+      "grad_norm": 6.540868711120688,
+      "learning_rate": 8.196238764995301e-07,
+      "loss": 0.9113,
+      "step": 27370
+    },
+    {
+      "epoch": 0.8246987951807229,
+      "grad_norm": 4.674074221348497,
+      "learning_rate": 8.168936674637728e-07,
+      "loss": 0.9354,
+      "step": 27380
+    },
+    {
+      "epoch": 0.825,
+      "grad_norm": 5.2923946551416226,
+      "learning_rate": 8.141676086873574e-07,
+      "loss": 0.8405,
+      "step": 27390
+    },
+    {
+      "epoch": 0.8253012048192772,
+      "grad_norm": 17.16648235120451,
+      "learning_rate": 8.114457028749373e-07,
+      "loss": 0.8569,
+      "step": 27400
+    },
+    {
+      "epoch": 0.8256024096385542,
+      "grad_norm": 5.5318367812979465,
+      "learning_rate": 8.08727952727042e-07,
+      "loss": 0.9101,
+      "step": 27410
+    },
+    {
+      "epoch": 0.8259036144578313,
+      "grad_norm": 11.042453676325039,
+      "learning_rate": 8.060143609400845e-07,
+      "loss": 0.9282,
+      "step": 27420
+    },
+    {
+      "epoch": 0.8262048192771084,
+      "grad_norm": 6.320479980702559,
+      "learning_rate": 8.033049302063478e-07,
+      "loss": 0.9145,
+      "step": 27430
+    },
+    {
+      "epoch": 0.8265060240963855,
+      "grad_norm": 5.461256521021767,
+      "learning_rate": 8.005996632139867e-07,
+      "loss": 0.9507,
+      "step": 27440
+    },
+    {
+      "epoch": 0.8268072289156626,
+      "grad_norm": 5.4490153905479,
+      "learning_rate": 7.978985626470264e-07,
+      "loss": 0.932,
+      "step": 27450
+    },
+    {
+      "epoch": 0.8271084337349398,
+      "grad_norm": 33.65867301668931,
+      "learning_rate": 7.952016311853572e-07,
+      "loss": 0.922,
+      "step": 27460
+    },
+    {
+      "epoch": 0.8274096385542169,
+      "grad_norm": 9.655256469529428,
+      "learning_rate": 7.925088715047341e-07,
+      "loss": 0.8934,
+      "step": 27470
+    },
+    {
+      "epoch": 0.827710843373494,
+      "grad_norm": 5.215517515809109,
+      "learning_rate": 7.898202862767701e-07,
+      "loss": 0.9111,
+      "step": 27480
+    },
+    {
+      "epoch": 0.8280120481927711,
+      "grad_norm": 3.9926902043041985,
+      "learning_rate": 7.871358781689431e-07,
+      "loss": 0.869,
+      "step": 27490
+    },
+    {
+      "epoch": 0.8283132530120482,
+      "grad_norm": 6.770181487323086,
+      "learning_rate": 7.844556498445788e-07,
+      "loss": 1.0588,
+      "step": 27500
+    },
+    {
+      "epoch": 0.8286144578313253,
+      "grad_norm": 4.003342302838501,
+      "learning_rate": 7.8177960396286e-07,
+      "loss": 0.9382,
+      "step": 27510
+    },
+    {
+      "epoch": 0.8289156626506025,
+      "grad_norm": 6.312607703702879,
+      "learning_rate": 7.791077431788207e-07,
+      "loss": 1.025,
+      "step": 27520
+    },
+    {
+      "epoch": 0.8292168674698795,
+      "grad_norm": 3.9398134718276556,
+      "learning_rate": 7.764400701433417e-07,
+      "loss": 0.9401,
+      "step": 27530
+    },
+    {
+      "epoch": 0.8295180722891566,
+      "grad_norm": 5.226230967539029,
+      "learning_rate": 7.737765875031477e-07,
+      "loss": 0.9035,
+      "step": 27540
+    },
+    {
+      "epoch": 0.8298192771084337,
+      "grad_norm": 5.003291406080699,
+      "learning_rate": 7.711172979008069e-07,
+      "loss": 0.8738,
+      "step": 27550
+    },
+    {
+      "epoch": 0.8301204819277108,
+      "grad_norm": 6.972945541490178,
+      "learning_rate": 7.684622039747281e-07,
+      "loss": 0.9869,
+      "step": 27560
+    },
+    {
+      "epoch": 0.8304216867469879,
+      "grad_norm": 4.674914719889957,
+      "learning_rate": 7.658113083591556e-07,
+      "loss": 0.946,
+      "step": 27570
+    },
+    {
+      "epoch": 0.8307228915662651,
+      "grad_norm": 4.091640041552626,
+      "learning_rate": 7.631646136841702e-07,
+      "loss": 0.8646,
+      "step": 27580
+    },
+    {
+      "epoch": 0.8310240963855422,
+      "grad_norm": 6.11366045652225,
+      "learning_rate": 7.605221225756837e-07,
+      "loss": 0.8583,
+      "step": 27590
+    },
+    {
+      "epoch": 0.8313253012048193,
+      "grad_norm": 4.702228869067837,
+      "learning_rate": 7.578838376554376e-07,
+      "loss": 0.9626,
+      "step": 27600
+    },
+    {
+      "epoch": 0.8316265060240964,
+      "grad_norm": 2.0034731218658015,
+      "learning_rate": 7.552497615410003e-07,
+      "loss": 0.8524,
+      "step": 27610
+    },
+    {
+      "epoch": 0.8319277108433735,
+      "grad_norm": 4.105676082020187,
+      "learning_rate": 7.526198968457632e-07,
+      "loss": 0.8579,
+      "step": 27620
+    },
+    {
+      "epoch": 0.8322289156626506,
+      "grad_norm": 8.181732148696257,
+      "learning_rate": 7.499942461789423e-07,
+      "loss": 0.9798,
+      "step": 27630
+    },
+    {
+      "epoch": 0.8325301204819278,
+      "grad_norm": 5.061684993884733,
+      "learning_rate": 7.473728121455703e-07,
+      "loss": 0.9056,
+      "step": 27640
+    },
+    {
+      "epoch": 0.8328313253012049,
+      "grad_norm": 1.90854313938708,
+      "learning_rate": 7.447555973464988e-07,
+      "loss": 0.8684,
+      "step": 27650
+    },
+    {
+      "epoch": 0.8331325301204819,
+      "grad_norm": 4.554426097884475,
+      "learning_rate": 7.421426043783864e-07,
+      "loss": 0.9668,
+      "step": 27660
+    },
+    {
+      "epoch": 0.833433734939759,
+      "grad_norm": 5.653339492385188,
+      "learning_rate": 7.395338358337123e-07,
+      "loss": 0.92,
+      "step": 27670
+    },
+    {
+      "epoch": 0.8337349397590361,
+      "grad_norm": 1.9014041837363123,
+      "learning_rate": 7.369292943007589e-07,
+      "loss": 0.9226,
+      "step": 27680
+    },
+    {
+      "epoch": 0.8340361445783132,
+      "grad_norm": 4.548615361967034,
+      "learning_rate": 7.343289823636168e-07,
+      "loss": 0.8829,
+      "step": 27690
+    },
+    {
+      "epoch": 0.8343373493975904,
+      "grad_norm": 3.809847836225357,
+      "learning_rate": 7.317329026021791e-07,
+      "loss": 0.8267,
+      "step": 27700
+    },
+    {
+      "epoch": 0.8346385542168675,
+      "grad_norm": 2.142259832202635,
+      "learning_rate": 7.291410575921399e-07,
+      "loss": 0.8336,
+      "step": 27710
+    },
+    {
+      "epoch": 0.8349397590361446,
+      "grad_norm": 6.038088531105821,
+      "learning_rate": 7.265534499049925e-07,
+      "loss": 1.0077,
+      "step": 27720
+    },
+    {
+      "epoch": 0.8352409638554217,
+      "grad_norm": 10.756075274716798,
+      "learning_rate": 7.239700821080254e-07,
+      "loss": 0.913,
+      "step": 27730
+    },
+    {
+      "epoch": 0.8355421686746988,
+      "grad_norm": 11.948788612904204,
+      "learning_rate": 7.213909567643235e-07,
+      "loss": 0.9393,
+      "step": 27740
+    },
+    {
+      "epoch": 0.8358433734939759,
+      "grad_norm": 5.679665390212351,
+      "learning_rate": 7.188160764327573e-07,
+      "loss": 0.926,
+      "step": 27750
+    },
+    {
+      "epoch": 0.8361445783132531,
+      "grad_norm": 3.898848576546342,
+      "learning_rate": 7.162454436679883e-07,
+      "loss": 0.8605,
+      "step": 27760
+    },
+    {
+      "epoch": 0.8364457831325302,
+      "grad_norm": 1.938386863051616,
+      "learning_rate": 7.136790610204663e-07,
+      "loss": 0.941,
+      "step": 27770
+    },
+    {
+      "epoch": 0.8367469879518072,
+      "grad_norm": 3.751209077060178,
+      "learning_rate": 7.111169310364203e-07,
+      "loss": 1.0116,
+      "step": 27780
+    },
+    {
+      "epoch": 0.8370481927710843,
+      "grad_norm": 9.978767089668626,
+      "learning_rate": 7.08559056257862e-07,
+      "loss": 0.9203,
+      "step": 27790
+    },
+    {
+      "epoch": 0.8373493975903614,
+      "grad_norm": 2.106694785336448,
+      "learning_rate": 7.060054392225812e-07,
+      "loss": 0.7004,
+      "step": 27800
+    },
+    {
+      "epoch": 0.8376506024096385,
+      "grad_norm": 2.0191945671021276,
+      "learning_rate": 7.034560824641434e-07,
+      "loss": 0.9192,
+      "step": 27810
+    },
+    {
+      "epoch": 0.8379518072289157,
+      "grad_norm": 2.0488468271818965,
+      "learning_rate": 7.009109885118859e-07,
+      "loss": 0.7888,
+      "step": 27820
+    },
+    {
+      "epoch": 0.8382530120481928,
+      "grad_norm": 4.571375666175543,
+      "learning_rate": 6.983701598909192e-07,
+      "loss": 0.9304,
+      "step": 27830
+    },
+    {
+      "epoch": 0.8385542168674699,
+      "grad_norm": 1.8653745309472625,
+      "learning_rate": 6.958335991221194e-07,
+      "loss": 0.857,
+      "step": 27840
+    },
+    {
+      "epoch": 0.838855421686747,
+      "grad_norm": 4.377028856709774,
+      "learning_rate": 6.933013087221302e-07,
+      "loss": 0.9011,
+      "step": 27850
+    },
+    {
+      "epoch": 0.8391566265060241,
+      "grad_norm": 4.817173377502368,
+      "learning_rate": 6.907732912033577e-07,
+      "loss": 0.8631,
+      "step": 27860
+    },
+    {
+      "epoch": 0.8394578313253012,
+      "grad_norm": 1.9625969406495993,
+      "learning_rate": 6.882495490739671e-07,
+      "loss": 0.9104,
+      "step": 27870
+    },
+    {
+      "epoch": 0.8397590361445784,
+      "grad_norm": 9.30273923683444,
+      "learning_rate": 6.857300848378857e-07,
+      "loss": 0.9997,
+      "step": 27880
+    },
+    {
+      "epoch": 0.8400602409638555,
+      "grad_norm": 6.362259796832259,
+      "learning_rate": 6.832149009947942e-07,
+      "loss": 0.8664,
+      "step": 27890
+    },
+    {
+      "epoch": 0.8403614457831325,
+      "grad_norm": 2.0409815446761757,
+      "learning_rate": 6.807040000401244e-07,
+      "loss": 0.8709,
+      "step": 27900
+    },
+    {
+      "epoch": 0.8406626506024096,
+      "grad_norm": 2.18058333288868,
+      "learning_rate": 6.781973844650614e-07,
+      "loss": 0.8885,
+      "step": 27910
+    },
+    {
+      "epoch": 0.8409638554216867,
+      "grad_norm": 4.384988353873599,
+      "learning_rate": 6.75695056756539e-07,
+      "loss": 0.993,
+      "step": 27920
+    },
+    {
+      "epoch": 0.8412650602409638,
+      "grad_norm": 4.6712190495818815,
+      "learning_rate": 6.73197019397236e-07,
+      "loss": 0.9277,
+      "step": 27930
+    },
+    {
+      "epoch": 0.841566265060241,
+      "grad_norm": 5.043727698460975,
+      "learning_rate": 6.70703274865574e-07,
+      "loss": 0.9338,
+      "step": 27940
+    },
+    {
+      "epoch": 0.8418674698795181,
+      "grad_norm": 6.531805156073384,
+      "learning_rate": 6.682138256357157e-07,
+      "loss": 0.8899,
+      "step": 27950
+    },
+    {
+      "epoch": 0.8421686746987952,
+      "grad_norm": 2.1279970719393453,
+      "learning_rate": 6.657286741775626e-07,
+      "loss": 0.8106,
+      "step": 27960
+    },
+    {
+      "epoch": 0.8424698795180723,
+      "grad_norm": 3.9752962599390713,
+      "learning_rate": 6.632478229567524e-07,
+      "loss": 0.8501,
+      "step": 27970
+    },
+    {
+      "epoch": 0.8427710843373494,
+      "grad_norm": 1.9133696843401249,
+      "learning_rate": 6.60771274434654e-07,
+      "loss": 0.8012,
+      "step": 27980
+    },
+    {
+      "epoch": 0.8430722891566265,
+      "grad_norm": 2.0254618726877114,
+      "learning_rate": 6.582990310683729e-07,
+      "loss": 0.8161,
+      "step": 27990
+    },
+    {
+      "epoch": 0.8433734939759037,
+      "grad_norm": 4.752619948600181,
+      "learning_rate": 6.558310953107372e-07,
+      "loss": 0.9181,
+      "step": 28000
+    },
+    {
+      "epoch": 0.8436746987951808,
+      "grad_norm": 9.54033400244362,
+      "learning_rate": 6.533674696103026e-07,
+      "loss": 0.9582,
+      "step": 28010
+    },
+    {
+      "epoch": 0.8439759036144578,
+      "grad_norm": 8.0533671144653,
+      "learning_rate": 6.509081564113529e-07,
+      "loss": 0.9613,
+      "step": 28020
+    },
+    {
+      "epoch": 0.8442771084337349,
+      "grad_norm": 11.676701993064908,
+      "learning_rate": 6.484531581538878e-07,
+      "loss": 0.9921,
+      "step": 28030
+    },
+    {
+      "epoch": 0.844578313253012,
+      "grad_norm": 6.1489331001793275,
+      "learning_rate": 6.460024772736301e-07,
+      "loss": 0.9244,
+      "step": 28040
+    },
+    {
+      "epoch": 0.8448795180722891,
+      "grad_norm": 6.338948370998788,
+      "learning_rate": 6.435561162020138e-07,
+      "loss": 0.9592,
+      "step": 28050
+    },
+    {
+      "epoch": 0.8451807228915663,
+      "grad_norm": 5.08435345343287,
+      "learning_rate": 6.411140773661933e-07,
+      "loss": 0.8618,
+      "step": 28060
+    },
+    {
+      "epoch": 0.8454819277108434,
+      "grad_norm": 4.90263337957172,
+      "learning_rate": 6.386763631890313e-07,
+      "loss": 0.856,
+      "step": 28070
+    },
+    {
+      "epoch": 0.8457831325301205,
+      "grad_norm": 7.501212199113993,
+      "learning_rate": 6.362429760890998e-07,
+      "loss": 0.9388,
+      "step": 28080
+    },
+    {
+      "epoch": 0.8460843373493976,
+      "grad_norm": 4.989775120116956,
+      "learning_rate": 6.338139184806791e-07,
+      "loss": 0.972,
+      "step": 28090
+    },
+    {
+      "epoch": 0.8463855421686747,
+      "grad_norm": 4.313002046257058,
+      "learning_rate": 6.31389192773752e-07,
+      "loss": 0.957,
+      "step": 28100
+    },
+    {
+      "epoch": 0.8466867469879518,
+      "grad_norm": 13.806894022089965,
+      "learning_rate": 6.289688013740047e-07,
+      "loss": 0.9736,
+      "step": 28110
+    },
+    {
+      "epoch": 0.846987951807229,
+      "grad_norm": 5.688792724738393,
+      "learning_rate": 6.265527466828231e-07,
+      "loss": 0.9773,
+      "step": 28120
+    },
+    {
+      "epoch": 0.8472891566265061,
+      "grad_norm": 5.432975160320383,
+      "learning_rate": 6.241410310972906e-07,
+      "loss": 0.8773,
+      "step": 28130
+    },
+    {
+      "epoch": 0.8475903614457831,
+      "grad_norm": 4.854691723593697,
+      "learning_rate": 6.217336570101873e-07,
+      "loss": 0.9537,
+      "step": 28140
+    },
+    {
+      "epoch": 0.8478915662650602,
+      "grad_norm": 8.78639374796019,
+      "learning_rate": 6.193306268099802e-07,
+      "loss": 0.8473,
+      "step": 28150
+    },
+    {
+      "epoch": 0.8481927710843373,
+      "grad_norm": 9.266631210306816,
+      "learning_rate": 6.169319428808307e-07,
+      "loss": 0.9589,
+      "step": 28160
+    },
+    {
+      "epoch": 0.8484939759036144,
+      "grad_norm": 4.322136497921134,
+      "learning_rate": 6.145376076025895e-07,
+      "loss": 0.9879,
+      "step": 28170
+    },
+    {
+      "epoch": 0.8487951807228916,
+      "grad_norm": 1.8729514199890531,
+      "learning_rate": 6.1214762335079e-07,
+      "loss": 0.7766,
+      "step": 28180
+    },
+    {
+      "epoch": 0.8490963855421687,
+      "grad_norm": 1.9564038691269208,
+      "learning_rate": 6.097619924966497e-07,
+      "loss": 0.9058,
+      "step": 28190
+    },
+    {
+      "epoch": 0.8493975903614458,
+      "grad_norm": 5.975231215790924,
+      "learning_rate": 6.073807174070667e-07,
+      "loss": 0.9175,
+      "step": 28200
+    },
+    {
+      "epoch": 0.8496987951807229,
+      "grad_norm": 4.1131203058007975,
+      "learning_rate": 6.050038004446179e-07,
+      "loss": 0.8922,
+      "step": 28210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.169024954256811,
+      "learning_rate": 6.026312439675553e-07,
+      "loss": 1.0015,
+      "step": 28220
+    },
+    {
+      "epoch": 0.8503012048192771,
+      "grad_norm": 5.934489393567222,
+      "learning_rate": 6.002630503298057e-07,
+      "loss": 0.9796,
+      "step": 28230
+    },
+    {
+      "epoch": 0.8506024096385543,
+      "grad_norm": 2.0530468074302832,
+      "learning_rate": 5.978992218809665e-07,
+      "loss": 0.7166,
+      "step": 28240
+    },
+    {
+      "epoch": 0.8509036144578314,
+      "grad_norm": 5.537437951695425,
+      "learning_rate": 5.955397609663061e-07,
+      "loss": 0.755,
+      "step": 28250
+    },
+    {
+      "epoch": 0.8512048192771084,
+      "grad_norm": 5.2229317713025845,
+      "learning_rate": 5.931846699267558e-07,
+      "loss": 0.9379,
+      "step": 28260
+    },
+    {
+      "epoch": 0.8515060240963855,
+      "grad_norm": 4.330571942184469,
+      "learning_rate": 5.908339510989158e-07,
+      "loss": 0.9476,
+      "step": 28270
+    },
+    {
+      "epoch": 0.8518072289156626,
+      "grad_norm": 4.870343631628832,
+      "learning_rate": 5.884876068150463e-07,
+      "loss": 0.958,
+      "step": 28280
+    },
+    {
+      "epoch": 0.8521084337349397,
+      "grad_norm": 6.372409493185241,
+      "learning_rate": 5.861456394030679e-07,
+      "loss": 0.9806,
+      "step": 28290
+    },
+    {
+      "epoch": 0.8524096385542169,
+      "grad_norm": 5.715537889441745,
+      "learning_rate": 5.838080511865557e-07,
+      "loss": 0.9923,
+      "step": 28300
+    },
+    {
+      "epoch": 0.852710843373494,
+      "grad_norm": 1.9619118766202412,
+      "learning_rate": 5.81474844484744e-07,
+      "loss": 0.83,
+      "step": 28310
+    },
+    {
+      "epoch": 0.8530120481927711,
+      "grad_norm": 4.227665926617872,
+      "learning_rate": 5.791460216125194e-07,
+      "loss": 0.9914,
+      "step": 28320
+    },
+    {
+      "epoch": 0.8533132530120482,
+      "grad_norm": 4.853282412993819,
+      "learning_rate": 5.768215848804165e-07,
+      "loss": 0.9985,
+      "step": 28330
+    },
+    {
+      "epoch": 0.8536144578313253,
+      "grad_norm": 5.066157065378188,
+      "learning_rate": 5.745015365946211e-07,
+      "loss": 0.7837,
+      "step": 28340
+    },
+    {
+      "epoch": 0.8539156626506024,
+      "grad_norm": 5.156033438228298,
+      "learning_rate": 5.721858790569634e-07,
+      "loss": 0.9316,
+      "step": 28350
+    },
+    {
+      "epoch": 0.8542168674698796,
+      "grad_norm": 4.299751966843597,
+      "learning_rate": 5.69874614564917e-07,
+      "loss": 0.9433,
+      "step": 28360
+    },
+    {
+      "epoch": 0.8545180722891567,
+      "grad_norm": 1.981874435981955,
+      "learning_rate": 5.675677454115979e-07,
+      "loss": 0.9611,
+      "step": 28370
+    },
+    {
+      "epoch": 0.8548192771084338,
+      "grad_norm": 4.902718733660948,
+      "learning_rate": 5.652652738857628e-07,
+      "loss": 0.9413,
+      "step": 28380
+    },
+    {
+      "epoch": 0.8551204819277108,
+      "grad_norm": 2.1582210019316275,
+      "learning_rate": 5.629672022718013e-07,
+      "loss": 0.9425,
+      "step": 28390
+    },
+    {
+      "epoch": 0.8554216867469879,
+      "grad_norm": 5.880947894711182,
+      "learning_rate": 5.606735328497409e-07,
+      "loss": 0.9013,
+      "step": 28400
+    },
+    {
+      "epoch": 0.855722891566265,
+      "grad_norm": 5.2299504726403026,
+      "learning_rate": 5.583842678952389e-07,
+      "loss": 0.9574,
+      "step": 28410
+    },
+    {
+      "epoch": 0.8560240963855422,
+      "grad_norm": 10.1635832698442,
+      "learning_rate": 5.560994096795869e-07,
+      "loss": 0.9176,
+      "step": 28420
+    },
+    {
+      "epoch": 0.8563253012048193,
+      "grad_norm": 4.693638440759502,
+      "learning_rate": 5.538189604697003e-07,
+      "loss": 0.9092,
+      "step": 28430
+    },
+    {
+      "epoch": 0.8566265060240964,
+      "grad_norm": 2.0933814621392255,
+      "learning_rate": 5.515429225281222e-07,
+      "loss": 0.9523,
+      "step": 28440
+    },
+    {
+      "epoch": 0.8569277108433735,
+      "grad_norm": 7.34226923750672,
+      "learning_rate": 5.492712981130171e-07,
+      "loss": 0.9428,
+      "step": 28450
+    },
+    {
+      "epoch": 0.8572289156626506,
+      "grad_norm": 5.167497352716724,
+      "learning_rate": 5.470040894781731e-07,
+      "loss": 0.9119,
+      "step": 28460
+    },
+    {
+      "epoch": 0.8575301204819277,
+      "grad_norm": 4.948847085140676,
+      "learning_rate": 5.447412988729961e-07,
+      "loss": 0.8916,
+      "step": 28470
+    },
+    {
+      "epoch": 0.8578313253012049,
+      "grad_norm": 8.995865512218028,
+      "learning_rate": 5.424829285425077e-07,
+      "loss": 0.9015,
+      "step": 28480
+    },
+    {
+      "epoch": 0.858132530120482,
+      "grad_norm": 2.1015400470155883,
+      "learning_rate": 5.402289807273453e-07,
+      "loss": 0.8613,
+      "step": 28490
+    },
+    {
+      "epoch": 0.858433734939759,
+      "grad_norm": 1.9559923309080423,
+      "learning_rate": 5.379794576637581e-07,
+      "loss": 0.8829,
+      "step": 28500
+    },
+    {
+      "epoch": 0.8587349397590361,
+      "grad_norm": 4.912852581162225,
+      "learning_rate": 5.357343615836041e-07,
+      "loss": 0.9039,
+      "step": 28510
+    },
+    {
+      "epoch": 0.8590361445783132,
+      "grad_norm": 4.7333957609298,
+      "learning_rate": 5.334936947143526e-07,
+      "loss": 0.9241,
+      "step": 28520
+    },
+    {
+      "epoch": 0.8593373493975903,
+      "grad_norm": 5.678929972574859,
+      "learning_rate": 5.312574592790759e-07,
+      "loss": 0.9969,
+      "step": 28530
+    },
+    {
+      "epoch": 0.8596385542168675,
+      "grad_norm": 2.150067626238408,
+      "learning_rate": 5.290256574964481e-07,
+      "loss": 0.7996,
+      "step": 28540
+    },
+    {
+      "epoch": 0.8599397590361446,
+      "grad_norm": 1.9698321735428013,
+      "learning_rate": 5.267982915807451e-07,
+      "loss": 0.9553,
+      "step": 28550
+    },
+    {
+      "epoch": 0.8602409638554217,
+      "grad_norm": 2.024910226438677,
+      "learning_rate": 5.245753637418461e-07,
+      "loss": 0.8779,
+      "step": 28560
+    },
+    {
+      "epoch": 0.8605421686746988,
+      "grad_norm": 9.891051431340566,
+      "learning_rate": 5.223568761852227e-07,
+      "loss": 0.9883,
+      "step": 28570
+    },
+    {
+      "epoch": 0.8608433734939759,
+      "grad_norm": 6.249959715005973,
+      "learning_rate": 5.201428311119416e-07,
+      "loss": 0.9391,
+      "step": 28580
+    },
+    {
+      "epoch": 0.861144578313253,
+      "grad_norm": 1.9420678070551143,
+      "learning_rate": 5.17933230718663e-07,
+      "loss": 0.8541,
+      "step": 28590
+    },
+    {
+      "epoch": 0.8614457831325302,
+      "grad_norm": 5.106089369429489,
+      "learning_rate": 5.157280771976364e-07,
+      "loss": 0.8666,
+      "step": 28600
+    },
+    {
+      "epoch": 0.8617469879518073,
+      "grad_norm": 4.352358368164967,
+      "learning_rate": 5.135273727367007e-07,
+      "loss": 0.922,
+      "step": 28610
+    },
+    {
+      "epoch": 0.8620481927710844,
+      "grad_norm": 14.17160551898297,
+      "learning_rate": 5.113311195192777e-07,
+      "loss": 0.9217,
+      "step": 28620
+    },
+    {
+      "epoch": 0.8623493975903614,
+      "grad_norm": 4.038349320253583,
+      "learning_rate": 5.091393197243788e-07,
+      "loss": 0.9002,
+      "step": 28630
+    },
+    {
+      "epoch": 0.8626506024096385,
+      "grad_norm": 5.02470867896239,
+      "learning_rate": 5.0695197552659e-07,
+      "loss": 0.9273,
+      "step": 28640
+    },
+    {
+      "epoch": 0.8629518072289156,
+      "grad_norm": 5.043995569523077,
+      "learning_rate": 5.047690890960804e-07,
+      "loss": 0.8193,
+      "step": 28650
+    },
+    {
+      "epoch": 0.8632530120481928,
+      "grad_norm": 8.273084337120943,
+      "learning_rate": 5.025906625985949e-07,
+      "loss": 0.8695,
+      "step": 28660
+    },
+    {
+      "epoch": 0.8635542168674699,
+      "grad_norm": 1.8102228845810138,
+      "learning_rate": 5.004166981954567e-07,
+      "loss": 0.849,
+      "step": 28670
+    },
+    {
+      "epoch": 0.863855421686747,
+      "grad_norm": 4.412327049290363,
+      "learning_rate": 4.982471980435594e-07,
+      "loss": 0.9532,
+      "step": 28680
+    },
+    {
+      "epoch": 0.8641566265060241,
+      "grad_norm": 5.895678021778057,
+      "learning_rate": 4.96082164295365e-07,
+      "loss": 0.8382,
+      "step": 28690
+    },
+    {
+      "epoch": 0.8644578313253012,
+      "grad_norm": 1.701427289067495,
+      "learning_rate": 4.939215990989088e-07,
+      "loss": 0.8398,
+      "step": 28700
+    },
+    {
+      "epoch": 0.8647590361445783,
+      "grad_norm": 6.346153380798089,
+      "learning_rate": 4.917655045977898e-07,
+      "loss": 0.9549,
+      "step": 28710
+    },
+    {
+      "epoch": 0.8650602409638555,
+      "grad_norm": 4.536059544543669,
+      "learning_rate": 4.896138829311731e-07,
+      "loss": 0.9226,
+      "step": 28720
+    },
+    {
+      "epoch": 0.8653614457831326,
+      "grad_norm": 5.853451445807957,
+      "learning_rate": 4.874667362337843e-07,
+      "loss": 0.9585,
+      "step": 28730
+    },
+    {
+      "epoch": 0.8656626506024097,
+      "grad_norm": 7.489524281315221,
+      "learning_rate": 4.853240666359105e-07,
+      "loss": 0.906,
+      "step": 28740
+    },
+    {
+      "epoch": 0.8659638554216867,
+      "grad_norm": 6.235676165574543,
+      "learning_rate": 4.831858762633967e-07,
+      "loss": 0.7827,
+      "step": 28750
+    },
+    {
+      "epoch": 0.8662650602409638,
+      "grad_norm": 4.931312234611484,
+      "learning_rate": 4.810521672376417e-07,
+      "loss": 0.895,
+      "step": 28760
+    },
+    {
+      "epoch": 0.8665662650602409,
+      "grad_norm": 5.895538762613074,
+      "learning_rate": 4.789229416756025e-07,
+      "loss": 0.9286,
+      "step": 28770
+    },
+    {
+      "epoch": 0.8668674698795181,
+      "grad_norm": 2.128250819595505,
+      "learning_rate": 4.7679820168978543e-07,
+      "loss": 0.8672,
+      "step": 28780
+    },
+    {
+      "epoch": 0.8671686746987952,
+      "grad_norm": 2.1421179135103907,
+      "learning_rate": 4.746779493882442e-07,
+      "loss": 0.9731,
+      "step": 28790
+    },
+    {
+      "epoch": 0.8674698795180723,
+      "grad_norm": 2.056383986798822,
+      "learning_rate": 4.725621868745828e-07,
+      "loss": 0.8649,
+      "step": 28800
+    },
+    {
+      "epoch": 0.8677710843373494,
+      "grad_norm": 4.537467977290869,
+      "learning_rate": 4.7045091624795145e-07,
+      "loss": 0.9488,
+      "step": 28810
+    },
+    {
+      "epoch": 0.8680722891566265,
+      "grad_norm": 1.8406715038440107,
+      "learning_rate": 4.6834413960304135e-07,
+      "loss": 0.8982,
+      "step": 28820
+    },
+    {
+      "epoch": 0.8683734939759036,
+      "grad_norm": 6.047113860764686,
+      "learning_rate": 4.6624185903008713e-07,
+      "loss": 0.8888,
+      "step": 28830
+    },
+    {
+      "epoch": 0.8686746987951808,
+      "grad_norm": 2.0991278994213616,
+      "learning_rate": 4.6414407661486025e-07,
+      "loss": 0.8784,
+      "step": 28840
+    },
+    {
+      "epoch": 0.8689759036144579,
+      "grad_norm": 11.595911442501958,
+      "learning_rate": 4.6205079443867207e-07,
+      "loss": 0.956,
+      "step": 28850
+    },
+    {
+      "epoch": 0.869277108433735,
+      "grad_norm": 4.874018227772749,
+      "learning_rate": 4.5996201457836675e-07,
+      "loss": 0.9858,
+      "step": 28860
+    },
+    {
+      "epoch": 0.869578313253012,
+      "grad_norm": 4.402909274266867,
+      "learning_rate": 4.5787773910632214e-07,
+      "loss": 0.9021,
+      "step": 28870
+    },
+    {
+      "epoch": 0.8698795180722891,
+      "grad_norm": 2.0683943509907445,
+      "learning_rate": 4.5579797009044913e-07,
+      "loss": 0.8625,
+      "step": 28880
+    },
+    {
+      "epoch": 0.8701807228915662,
+      "grad_norm": 4.791022091052091,
+      "learning_rate": 4.537227095941843e-07,
+      "loss": 0.9933,
+      "step": 28890
+    },
+    {
+      "epoch": 0.8704819277108434,
+      "grad_norm": 4.39227937163775,
+      "learning_rate": 4.516519596764929e-07,
+      "loss": 0.9145,
+      "step": 28900
+    },
+    {
+      "epoch": 0.8707831325301205,
+      "grad_norm": 5.743297542284759,
+      "learning_rate": 4.4958572239186374e-07,
+      "loss": 0.9188,
+      "step": 28910
+    },
+    {
+      "epoch": 0.8710843373493976,
+      "grad_norm": 5.555509186765791,
+      "learning_rate": 4.475239997903108e-07,
+      "loss": 0.903,
+      "step": 28920
+    },
+    {
+      "epoch": 0.8713855421686747,
+      "grad_norm": 4.98994266768898,
+      "learning_rate": 4.454667939173685e-07,
+      "loss": 0.8964,
+      "step": 28930
+    },
+    {
+      "epoch": 0.8716867469879518,
+      "grad_norm": 6.261471622766507,
+      "learning_rate": 4.4341410681408516e-07,
+      "loss": 0.9495,
+      "step": 28940
+    },
+    {
+      "epoch": 0.8719879518072289,
+      "grad_norm": 3.62450816266926,
+      "learning_rate": 4.4136594051703286e-07,
+      "loss": 0.979,
+      "step": 28950
+    },
+    {
+      "epoch": 0.8722891566265061,
+      "grad_norm": 2.067644154191991,
+      "learning_rate": 4.393222970582939e-07,
+      "loss": 0.9022,
+      "step": 28960
+    },
+    {
+      "epoch": 0.8725903614457832,
+      "grad_norm": 2.1342799963472414,
+      "learning_rate": 4.372831784654652e-07,
+      "loss": 0.9252,
+      "step": 28970
+    },
+    {
+      "epoch": 0.8728915662650603,
+      "grad_norm": 14.530165084636979,
+      "learning_rate": 4.352485867616524e-07,
+      "loss": 0.8677,
+      "step": 28980
+    },
+    {
+      "epoch": 0.8731927710843373,
+      "grad_norm": 2.0120063513613693,
+      "learning_rate": 4.332185239654718e-07,
+      "loss": 0.8509,
+      "step": 28990
+    },
+    {
+      "epoch": 0.8734939759036144,
+      "grad_norm": 4.865687883027641,
+      "learning_rate": 4.311929920910451e-07,
+      "loss": 0.935,
+      "step": 29000
+    },
+    {
+      "epoch": 0.8737951807228915,
+      "grad_norm": 2.0245036414742166,
+      "learning_rate": 4.2917199314799873e-07,
+      "loss": 0.9141,
+      "step": 29010
+    },
+    {
+      "epoch": 0.8740963855421687,
+      "grad_norm": 5.76091144579355,
+      "learning_rate": 4.271555291414636e-07,
+      "loss": 0.9055,
+      "step": 29020
+    },
+    {
+      "epoch": 0.8743975903614458,
+      "grad_norm": 4.315560309100055,
+      "learning_rate": 4.2514360207206953e-07,
+      "loss": 0.8906,
+      "step": 29030
+    },
+    {
+      "epoch": 0.8746987951807229,
+      "grad_norm": 8.076654083839747,
+      "learning_rate": 4.231362139359441e-07,
+      "loss": 0.832,
+      "step": 29040
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 4.456687958244754,
+      "learning_rate": 4.211333667247125e-07,
+      "loss": 0.9492,
+      "step": 29050
+    },
+    {
+      "epoch": 0.8753012048192771,
+      "grad_norm": 3.612573791932306,
+      "learning_rate": 4.191350624254975e-07,
+      "loss": 0.877,
+      "step": 29060
+    },
+    {
+      "epoch": 0.8756024096385542,
+      "grad_norm": 8.883093819274588,
+      "learning_rate": 4.171413030209098e-07,
+      "loss": 0.9089,
+      "step": 29070
+    },
+    {
+      "epoch": 0.8759036144578313,
+      "grad_norm": 6.558894267229141,
+      "learning_rate": 4.1515209048905434e-07,
+      "loss": 0.9195,
+      "step": 29080
+    },
+    {
+      "epoch": 0.8762048192771085,
+      "grad_norm": 9.114461342653444,
+      "learning_rate": 4.1316742680352406e-07,
+      "loss": 0.9902,
+      "step": 29090
+    },
+    {
+      "epoch": 0.8765060240963856,
+      "grad_norm": 4.721154027582496,
+      "learning_rate": 4.111873139333972e-07,
+      "loss": 0.9006,
+      "step": 29100
+    },
+    {
+      "epoch": 0.8768072289156627,
+      "grad_norm": 4.562759120003542,
+      "learning_rate": 4.092117538432394e-07,
+      "loss": 0.9693,
+      "step": 29110
+    },
+    {
+      "epoch": 0.8771084337349397,
+      "grad_norm": 1.9824518715510289,
+      "learning_rate": 4.072407484930968e-07,
+      "loss": 0.7591,
+      "step": 29120
+    },
+    {
+      "epoch": 0.8774096385542168,
+      "grad_norm": 3.6681972154747706,
+      "learning_rate": 4.052742998384995e-07,
+      "loss": 0.9263,
+      "step": 29130
+    },
+    {
+      "epoch": 0.8777108433734939,
+      "grad_norm": 5.787985110484834,
+      "learning_rate": 4.0331240983045363e-07,
+      "loss": 0.8554,
+      "step": 29140
+    },
+    {
+      "epoch": 0.8780120481927711,
+      "grad_norm": 4.563737633826767,
+      "learning_rate": 4.0135508041544423e-07,
+      "loss": 0.9257,
+      "step": 29150
+    },
+    {
+      "epoch": 0.8783132530120482,
+      "grad_norm": 4.423986763627893,
+      "learning_rate": 3.9940231353543077e-07,
+      "loss": 0.8615,
+      "step": 29160
+    },
+    {
+      "epoch": 0.8786144578313253,
+      "grad_norm": 4.18139800162496,
+      "learning_rate": 3.974541111278474e-07,
+      "loss": 0.9298,
+      "step": 29170
+    },
+    {
+      "epoch": 0.8789156626506024,
+      "grad_norm": 5.813227606398191,
+      "learning_rate": 3.955104751255995e-07,
+      "loss": 0.9526,
+      "step": 29180
+    },
+    {
+      "epoch": 0.8792168674698795,
+      "grad_norm": 5.430429125831835,
+      "learning_rate": 3.935714074570579e-07,
+      "loss": 0.85,
+      "step": 29190
+    },
+    {
+      "epoch": 0.8795180722891566,
+      "grad_norm": 4.656406443588307,
+      "learning_rate": 3.916369100460665e-07,
+      "loss": 0.8522,
+      "step": 29200
+    },
+    {
+      "epoch": 0.8798192771084338,
+      "grad_norm": 5.18714013040448,
+      "learning_rate": 3.8970698481193225e-07,
+      "loss": 0.8949,
+      "step": 29210
+    },
+    {
+      "epoch": 0.8801204819277109,
+      "grad_norm": 1.9378784598863399,
+      "learning_rate": 3.877816336694262e-07,
+      "loss": 0.8809,
+      "step": 29220
+    },
+    {
+      "epoch": 0.880421686746988,
+      "grad_norm": 4.198225000624405,
+      "learning_rate": 3.858608585287799e-07,
+      "loss": 1.0119,
+      "step": 29230
+    },
+    {
+      "epoch": 0.880722891566265,
+      "grad_norm": 4.546057128649085,
+      "learning_rate": 3.839446612956871e-07,
+      "loss": 0.9377,
+      "step": 29240
+    },
+    {
+      "epoch": 0.8810240963855421,
+      "grad_norm": 1.9557703770719894,
+      "learning_rate": 3.8203304387129823e-07,
+      "loss": 0.8512,
+      "step": 29250
+    },
+    {
+      "epoch": 0.8813253012048192,
+      "grad_norm": 2.1599891644315354,
+      "learning_rate": 3.80126008152219e-07,
+      "loss": 0.9077,
+      "step": 29260
+    },
+    {
+      "epoch": 0.8816265060240964,
+      "grad_norm": 2.1197042372122463,
+      "learning_rate": 3.782235560305131e-07,
+      "loss": 0.934,
+      "step": 29270
+    },
+    {
+      "epoch": 0.8819277108433735,
+      "grad_norm": 4.0269587095371735,
+      "learning_rate": 3.763256893936917e-07,
+      "loss": 0.9689,
+      "step": 29280
+    },
+    {
+      "epoch": 0.8822289156626506,
+      "grad_norm": 3.7807803120359527,
+      "learning_rate": 3.744324101247199e-07,
+      "loss": 1.0031,
+      "step": 29290
+    },
+    {
+      "epoch": 0.8825301204819277,
+      "grad_norm": 1.9451313314176517,
+      "learning_rate": 3.725437201020088e-07,
+      "loss": 0.9282,
+      "step": 29300
+    },
+    {
+      "epoch": 0.8828313253012048,
+      "grad_norm": 11.240100733399817,
+      "learning_rate": 3.706596211994201e-07,
+      "loss": 0.9891,
+      "step": 29310
+    },
+    {
+      "epoch": 0.8831325301204819,
+      "grad_norm": 7.8575070793060675,
+      "learning_rate": 3.687801152862569e-07,
+      "loss": 0.8873,
+      "step": 29320
+    },
+    {
+      "epoch": 0.8834337349397591,
+      "grad_norm": 4.446012880321386,
+      "learning_rate": 3.669052042272675e-07,
+      "loss": 0.8067,
+      "step": 29330
+    },
+    {
+      "epoch": 0.8837349397590362,
+      "grad_norm": 4.7464167208914825,
+      "learning_rate": 3.6503488988264e-07,
+      "loss": 0.8927,
+      "step": 29340
+    },
+    {
+      "epoch": 0.8840361445783133,
+      "grad_norm": 2.1024173526693457,
+      "learning_rate": 3.631691741080023e-07,
+      "loss": 0.8519,
+      "step": 29350
+    },
+    {
+      "epoch": 0.8843373493975903,
+      "grad_norm": 7.569368125402966,
+      "learning_rate": 3.61308058754421e-07,
+      "loss": 1.0047,
+      "step": 29360
+    },
+    {
+      "epoch": 0.8846385542168674,
+      "grad_norm": 5.215962429635607,
+      "learning_rate": 3.5945154566839613e-07,
+      "loss": 0.8994,
+      "step": 29370
+    },
+    {
+      "epoch": 0.8849397590361445,
+      "grad_norm": 5.006863733803054,
+      "learning_rate": 3.575996366918638e-07,
+      "loss": 0.9277,
+      "step": 29380
+    },
+    {
+      "epoch": 0.8852409638554217,
+      "grad_norm": 5.158016403608478,
+      "learning_rate": 3.5575233366219044e-07,
+      "loss": 0.893,
+      "step": 29390
+    },
+    {
+      "epoch": 0.8855421686746988,
+      "grad_norm": 5.014009912667711,
+      "learning_rate": 3.539096384121743e-07,
+      "loss": 0.9196,
+      "step": 29400
+    },
+    {
+      "epoch": 0.8858433734939759,
+      "grad_norm": 4.33250483163687,
+      "learning_rate": 3.520715527700402e-07,
+      "loss": 0.97,
+      "step": 29410
+    },
+    {
+      "epoch": 0.886144578313253,
+      "grad_norm": 6.012638414468784,
+      "learning_rate": 3.502380785594428e-07,
+      "loss": 0.9374,
+      "step": 29420
+    },
+    {
+      "epoch": 0.8864457831325301,
+      "grad_norm": 4.251775072235473,
+      "learning_rate": 3.4840921759945745e-07,
+      "loss": 0.9737,
+      "step": 29430
+    },
+    {
+      "epoch": 0.8867469879518072,
+      "grad_norm": 4.518795440404182,
+      "learning_rate": 3.465849717045838e-07,
+      "loss": 0.9384,
+      "step": 29440
+    },
+    {
+      "epoch": 0.8870481927710844,
+      "grad_norm": 2.081210093857427,
+      "learning_rate": 3.447653426847447e-07,
+      "loss": 0.9253,
+      "step": 29450
+    },
+    {
+      "epoch": 0.8873493975903615,
+      "grad_norm": 5.565653379563694,
+      "learning_rate": 3.42950332345281e-07,
+      "loss": 0.9628,
+      "step": 29460
+    },
+    {
+      "epoch": 0.8876506024096386,
+      "grad_norm": 4.017165584844056,
+      "learning_rate": 3.411399424869499e-07,
+      "loss": 0.9577,
+      "step": 29470
+    },
+    {
+      "epoch": 0.8879518072289156,
+      "grad_norm": 3.984888910738686,
+      "learning_rate": 3.393341749059259e-07,
+      "loss": 0.9496,
+      "step": 29480
+    },
+    {
+      "epoch": 0.8882530120481927,
+      "grad_norm": 2.073737194733733,
+      "learning_rate": 3.37533031393798e-07,
+      "loss": 0.8238,
+      "step": 29490
+    },
+    {
+      "epoch": 0.8885542168674698,
+      "grad_norm": 10.737941249040539,
+      "learning_rate": 3.3573651373756513e-07,
+      "loss": 0.8978,
+      "step": 29500
+    },
+    {
+      "epoch": 0.888855421686747,
+      "grad_norm": 2.022991667976808,
+      "learning_rate": 3.3394462371963864e-07,
+      "loss": 0.928,
+      "step": 29510
+    },
+    {
+      "epoch": 0.8891566265060241,
+      "grad_norm": 7.719472513286403,
+      "learning_rate": 3.3215736311784044e-07,
+      "loss": 1.0267,
+      "step": 29520
+    },
+    {
+      "epoch": 0.8894578313253012,
+      "grad_norm": 5.256950506393739,
+      "learning_rate": 3.303747337053936e-07,
+      "loss": 0.7333,
+      "step": 29530
+    },
+    {
+      "epoch": 0.8897590361445783,
+      "grad_norm": 2.0995705984504593,
+      "learning_rate": 3.2859673725093133e-07,
+      "loss": 0.8656,
+      "step": 29540
+    },
+    {
+      "epoch": 0.8900602409638554,
+      "grad_norm": 3.7102400130303046,
+      "learning_rate": 3.2682337551848795e-07,
+      "loss": 0.9209,
+      "step": 29550
+    },
+    {
+      "epoch": 0.8903614457831325,
+      "grad_norm": 4.1883773578173855,
+      "learning_rate": 3.2505465026750115e-07,
+      "loss": 0.8555,
+      "step": 29560
+    },
+    {
+      "epoch": 0.8906626506024097,
+      "grad_norm": 1.9198249481048593,
+      "learning_rate": 3.2329056325280885e-07,
+      "loss": 0.8714,
+      "step": 29570
+    },
+    {
+      "epoch": 0.8909638554216868,
+      "grad_norm": 5.923383194381662,
+      "learning_rate": 3.2153111622464263e-07,
+      "loss": 0.9234,
+      "step": 29580
+    },
+    {
+      "epoch": 0.8912650602409639,
+      "grad_norm": 5.567215823246787,
+      "learning_rate": 3.1977631092863613e-07,
+      "loss": 0.9195,
+      "step": 29590
+    },
+    {
+      "epoch": 0.891566265060241,
+      "grad_norm": 7.9148940355828605,
+      "learning_rate": 3.1802614910581506e-07,
+      "loss": 0.8443,
+      "step": 29600
+    },
+    {
+      "epoch": 0.891867469879518,
+      "grad_norm": 2.050219279673027,
+      "learning_rate": 3.162806324925988e-07,
+      "loss": 0.728,
+      "step": 29610
+    },
+    {
+      "epoch": 0.8921686746987951,
+      "grad_norm": 6.271753398148368,
+      "learning_rate": 3.1453976282079636e-07,
+      "loss": 0.9619,
+      "step": 29620
+    },
+    {
+      "epoch": 0.8924698795180723,
+      "grad_norm": 4.354315334600484,
+      "learning_rate": 3.1280354181760896e-07,
+      "loss": 0.9477,
+      "step": 29630
+    },
+    {
+      "epoch": 0.8927710843373494,
+      "grad_norm": 5.223722917723963,
+      "learning_rate": 3.1107197120562317e-07,
+      "loss": 0.9797,
+      "step": 29640
+    },
+    {
+      "epoch": 0.8930722891566265,
+      "grad_norm": 4.558827839557073,
+      "learning_rate": 3.09345052702813e-07,
+      "loss": 0.8947,
+      "step": 29650
+    },
+    {
+      "epoch": 0.8933734939759036,
+      "grad_norm": 5.37366913024313,
+      "learning_rate": 3.0762278802253743e-07,
+      "loss": 0.8967,
+      "step": 29660
+    },
+    {
+      "epoch": 0.8936746987951807,
+      "grad_norm": 3.9853627599869226,
+      "learning_rate": 3.059051788735379e-07,
+      "loss": 0.9019,
+      "step": 29670
+    },
+    {
+      "epoch": 0.8939759036144578,
+      "grad_norm": 5.107432456134144,
+      "learning_rate": 3.0419222695993455e-07,
+      "loss": 0.9377,
+      "step": 29680
+    },
+    {
+      "epoch": 0.894277108433735,
+      "grad_norm": 2.1163388036344823,
+      "learning_rate": 3.024839339812291e-07,
+      "loss": 0.8239,
+      "step": 29690
+    },
+    {
+      "epoch": 0.8945783132530121,
+      "grad_norm": 1.822929408467778,
+      "learning_rate": 3.007803016323013e-07,
+      "loss": 0.9157,
+      "step": 29700
+    },
+    {
+      "epoch": 0.8948795180722892,
+      "grad_norm": 4.682403873330276,
+      "learning_rate": 2.990813316034064e-07,
+      "loss": 0.918,
+      "step": 29710
+    },
+    {
+      "epoch": 0.8951807228915662,
+      "grad_norm": 5.200497158377645,
+      "learning_rate": 2.9738702558017305e-07,
+      "loss": 0.8555,
+      "step": 29720
+    },
+    {
+      "epoch": 0.8954819277108433,
+      "grad_norm": 6.504588539674624,
+      "learning_rate": 2.9569738524360236e-07,
+      "loss": 1.0323,
+      "step": 29730
+    },
+    {
+      "epoch": 0.8957831325301204,
+      "grad_norm": 4.495593669503393,
+      "learning_rate": 2.940124122700677e-07,
+      "loss": 1.0411,
+      "step": 29740
+    },
+    {
+      "epoch": 0.8960843373493976,
+      "grad_norm": 4.287366824328198,
+      "learning_rate": 2.923321083313113e-07,
+      "loss": 0.8035,
+      "step": 29750
+    },
+    {
+      "epoch": 0.8963855421686747,
+      "grad_norm": 5.086546250823018,
+      "learning_rate": 2.9065647509444296e-07,
+      "loss": 0.9774,
+      "step": 29760
+    },
+    {
+      "epoch": 0.8966867469879518,
+      "grad_norm": 2.155168225757024,
+      "learning_rate": 2.8898551422193833e-07,
+      "loss": 0.9169,
+      "step": 29770
+    },
+    {
+      "epoch": 0.8969879518072289,
+      "grad_norm": 16.435698828928746,
+      "learning_rate": 2.873192273716369e-07,
+      "loss": 0.8862,
+      "step": 29780
+    },
+    {
+      "epoch": 0.897289156626506,
+      "grad_norm": 4.641324427557074,
+      "learning_rate": 2.856576161967417e-07,
+      "loss": 0.9304,
+      "step": 29790
+    },
+    {
+      "epoch": 0.8975903614457831,
+      "grad_norm": 1.940484722335563,
+      "learning_rate": 2.840006823458158e-07,
+      "loss": 0.8366,
+      "step": 29800
+    },
+    {
+      "epoch": 0.8978915662650603,
+      "grad_norm": 7.506862616845636,
+      "learning_rate": 2.823484274627836e-07,
+      "loss": 0.8632,
+      "step": 29810
+    },
+    {
+      "epoch": 0.8981927710843374,
+      "grad_norm": 4.803067142316807,
+      "learning_rate": 2.8070085318692597e-07,
+      "loss": 0.9566,
+      "step": 29820
+    },
+    {
+      "epoch": 0.8984939759036145,
+      "grad_norm": 4.802931629665585,
+      "learning_rate": 2.7905796115287767e-07,
+      "loss": 0.9485,
+      "step": 29830
+    },
+    {
+      "epoch": 0.8987951807228916,
+      "grad_norm": 6.143256075314575,
+      "learning_rate": 2.774197529906325e-07,
+      "loss": 0.7594,
+      "step": 29840
+    },
+    {
+      "epoch": 0.8990963855421686,
+      "grad_norm": 5.319450338673451,
+      "learning_rate": 2.7578623032553487e-07,
+      "loss": 0.8765,
+      "step": 29850
+    },
+    {
+      "epoch": 0.8993975903614457,
+      "grad_norm": 5.465696219193865,
+      "learning_rate": 2.741573947782794e-07,
+      "loss": 0.9911,
+      "step": 29860
+    },
+    {
+      "epoch": 0.8996987951807229,
+      "grad_norm": 2.0878824258379707,
+      "learning_rate": 2.725332479649123e-07,
+      "loss": 0.8584,
+      "step": 29870
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.149453599378429,
+      "learning_rate": 2.7091379149682683e-07,
+      "loss": 0.9967,
+      "step": 29880
+    },
+    {
+      "epoch": 0.9003012048192771,
+      "grad_norm": 1.9416211530643281,
+      "learning_rate": 2.6929902698076327e-07,
+      "loss": 0.9359,
+      "step": 29890
+    },
+    {
+      "epoch": 0.9006024096385542,
+      "grad_norm": 4.35994852925526,
+      "learning_rate": 2.676889560188051e-07,
+      "loss": 0.9473,
+      "step": 29900
+    },
+    {
+      "epoch": 0.9009036144578313,
+      "grad_norm": 4.730471412138855,
+      "learning_rate": 2.660835802083833e-07,
+      "loss": 0.855,
+      "step": 29910
+    },
+    {
+      "epoch": 0.9012048192771084,
+      "grad_norm": 4.648347910220079,
+      "learning_rate": 2.6448290114226616e-07,
+      "loss": 0.9593,
+      "step": 29920
+    },
+    {
+      "epoch": 0.9015060240963856,
+      "grad_norm": 1.9949640203058063,
+      "learning_rate": 2.628869204085638e-07,
+      "loss": 0.7891,
+      "step": 29930
+    },
+    {
+      "epoch": 0.9018072289156627,
+      "grad_norm": 7.036764321044006,
+      "learning_rate": 2.612956395907235e-07,
+      "loss": 0.9608,
+      "step": 29940
+    },
+    {
+      "epoch": 0.9021084337349398,
+      "grad_norm": 10.185960314989945,
+      "learning_rate": 2.597090602675334e-07,
+      "loss": 0.9936,
+      "step": 29950
+    },
+    {
+      "epoch": 0.9024096385542169,
+      "grad_norm": 1.957574016901448,
+      "learning_rate": 2.5812718401311334e-07,
+      "loss": 0.9282,
+      "step": 29960
+    },
+    {
+      "epoch": 0.9027108433734939,
+      "grad_norm": 5.031416696909351,
+      "learning_rate": 2.5655001239691836e-07,
+      "loss": 0.9147,
+      "step": 29970
+    },
+    {
+      "epoch": 0.903012048192771,
+      "grad_norm": 4.117105535052751,
+      "learning_rate": 2.5497754698373455e-07,
+      "loss": 0.8457,
+      "step": 29980
+    },
+    {
+      "epoch": 0.9033132530120482,
+      "grad_norm": 1.9777963266760097,
+      "learning_rate": 2.5340978933368177e-07,
+      "loss": 0.8008,
+      "step": 29990
+    },
+    {
+      "epoch": 0.9036144578313253,
+      "grad_norm": 4.484332589137819,
+      "learning_rate": 2.518467410022063e-07,
+      "loss": 0.9254,
+      "step": 30000
+    },
+    {
+      "epoch": 0.9039156626506024,
+      "grad_norm": 4.427670215416986,
+      "learning_rate": 2.5028840354008264e-07,
+      "loss": 0.8791,
+      "step": 30010
+    },
+    {
+      "epoch": 0.9042168674698795,
+      "grad_norm": 8.377019267464451,
+      "learning_rate": 2.487347784934119e-07,
+      "loss": 0.8925,
+      "step": 30020
+    },
+    {
+      "epoch": 0.9045180722891566,
+      "grad_norm": 7.2365708481771485,
+      "learning_rate": 2.4718586740361985e-07,
+      "loss": 0.8452,
+      "step": 30030
+    },
+    {
+      "epoch": 0.9048192771084337,
+      "grad_norm": 2.0534488121332792,
+      "learning_rate": 2.45641671807455e-07,
+      "loss": 0.8459,
+      "step": 30040
+    },
+    {
+      "epoch": 0.9051204819277109,
+      "grad_norm": 4.301532552562159,
+      "learning_rate": 2.4410219323698636e-07,
+      "loss": 0.9414,
+      "step": 30050
+    },
+    {
+      "epoch": 0.905421686746988,
+      "grad_norm": 2.038146180435159,
+      "learning_rate": 2.42567433219606e-07,
+      "loss": 0.8977,
+      "step": 30060
+    },
+    {
+      "epoch": 0.9057228915662651,
+      "grad_norm": 6.0234597980613485,
+      "learning_rate": 2.410373932780219e-07,
+      "loss": 1.0074,
+      "step": 30070
+    },
+    {
+      "epoch": 0.9060240963855422,
+      "grad_norm": 4.470062653453608,
+      "learning_rate": 2.3951207493025876e-07,
+      "loss": 0.9697,
+      "step": 30080
+    },
+    {
+      "epoch": 0.9063253012048192,
+      "grad_norm": 4.061143895409584,
+      "learning_rate": 2.379914796896582e-07,
+      "loss": 0.9099,
+      "step": 30090
+    },
+    {
+      "epoch": 0.9066265060240963,
+      "grad_norm": 4.789426982661688,
+      "learning_rate": 2.364756090648762e-07,
+      "loss": 0.9098,
+      "step": 30100
+    },
+    {
+      "epoch": 0.9069277108433735,
+      "grad_norm": 2.078045479960634,
+      "learning_rate": 2.349644645598792e-07,
+      "loss": 0.9128,
+      "step": 30110
+    },
+    {
+      "epoch": 0.9072289156626506,
+      "grad_norm": 4.79462034368817,
+      "learning_rate": 2.3345804767394564e-07,
+      "loss": 0.8735,
+      "step": 30120
+    },
+    {
+      "epoch": 0.9075301204819277,
+      "grad_norm": 10.25970322701426,
+      "learning_rate": 2.3195635990166442e-07,
+      "loss": 0.9362,
+      "step": 30130
+    },
+    {
+      "epoch": 0.9078313253012048,
+      "grad_norm": 51.63692180692184,
+      "learning_rate": 2.3045940273293154e-07,
+      "loss": 0.9505,
+      "step": 30140
+    },
+    {
+      "epoch": 0.9081325301204819,
+      "grad_norm": 6.726779326874728,
+      "learning_rate": 2.2896717765294784e-07,
+      "loss": 0.8549,
+      "step": 30150
+    },
+    {
+      "epoch": 0.908433734939759,
+      "grad_norm": 11.326967021423837,
+      "learning_rate": 2.274796861422246e-07,
+      "loss": 0.8707,
+      "step": 30160
+    },
+    {
+      "epoch": 0.9087349397590362,
+      "grad_norm": 5.071628059826979,
+      "learning_rate": 2.2599692967656962e-07,
+      "loss": 0.9701,
+      "step": 30170
+    },
+    {
+      "epoch": 0.9090361445783133,
+      "grad_norm": 6.955451787569282,
+      "learning_rate": 2.2451890972709778e-07,
+      "loss": 0.9932,
+      "step": 30180
+    },
+    {
+      "epoch": 0.9093373493975904,
+      "grad_norm": 2.2274587409892757,
+      "learning_rate": 2.2304562776022276e-07,
+      "loss": 0.8634,
+      "step": 30190
+    },
+    {
+      "epoch": 0.9096385542168675,
+      "grad_norm": 1.998660494421673,
+      "learning_rate": 2.2157708523765864e-07,
+      "loss": 0.8311,
+      "step": 30200
+    },
+    {
+      "epoch": 0.9099397590361445,
+      "grad_norm": 60.562075316733505,
+      "learning_rate": 2.2011328361641605e-07,
+      "loss": 0.8249,
+      "step": 30210
+    },
+    {
+      "epoch": 0.9102409638554216,
+      "grad_norm": 1.8539763405018266,
+      "learning_rate": 2.1865422434880267e-07,
+      "loss": 0.9395,
+      "step": 30220
+    },
+    {
+      "epoch": 0.9105421686746988,
+      "grad_norm": 5.549849941885362,
+      "learning_rate": 2.171999088824195e-07,
+      "loss": 1.0061,
+      "step": 30230
+    },
+    {
+      "epoch": 0.9108433734939759,
+      "grad_norm": 5.079218708837629,
+      "learning_rate": 2.1575033866016447e-07,
+      "loss": 0.9713,
+      "step": 30240
+    },
+    {
+      "epoch": 0.911144578313253,
+      "grad_norm": 1.9301915959028308,
+      "learning_rate": 2.1430551512022336e-07,
+      "loss": 0.9012,
+      "step": 30250
+    },
+    {
+      "epoch": 0.9114457831325301,
+      "grad_norm": 4.992085712734793,
+      "learning_rate": 2.1286543969607608e-07,
+      "loss": 0.9912,
+      "step": 30260
+    },
+    {
+      "epoch": 0.9117469879518072,
+      "grad_norm": 4.811049837802725,
+      "learning_rate": 2.1143011381648925e-07,
+      "loss": 0.819,
+      "step": 30270
+    },
+    {
+      "epoch": 0.9120481927710843,
+      "grad_norm": 5.748533215147977,
+      "learning_rate": 2.0999953890551816e-07,
+      "loss": 0.9567,
+      "step": 30280
+    },
+    {
+      "epoch": 0.9123493975903615,
+      "grad_norm": 4.694246260071003,
+      "learning_rate": 2.0857371638250467e-07,
+      "loss": 1.0229,
+      "step": 30290
+    },
+    {
+      "epoch": 0.9126506024096386,
+      "grad_norm": 5.091136449551038,
+      "learning_rate": 2.0715264766207443e-07,
+      "loss": 0.9766,
+      "step": 30300
+    },
+    {
+      "epoch": 0.9129518072289157,
+      "grad_norm": 5.033870702189704,
+      "learning_rate": 2.0573633415413962e-07,
+      "loss": 0.9008,
+      "step": 30310
+    },
+    {
+      "epoch": 0.9132530120481928,
+      "grad_norm": 7.73187832623893,
+      "learning_rate": 2.0432477726389122e-07,
+      "loss": 0.8716,
+      "step": 30320
+    },
+    {
+      "epoch": 0.9135542168674698,
+      "grad_norm": 4.657926851190474,
+      "learning_rate": 2.029179783918006e-07,
+      "loss": 0.9224,
+      "step": 30330
+    },
+    {
+      "epoch": 0.9138554216867469,
+      "grad_norm": 3.823697448591747,
+      "learning_rate": 2.0151593893362243e-07,
+      "loss": 0.9542,
+      "step": 30340
+    },
+    {
+      "epoch": 0.9141566265060241,
+      "grad_norm": 2.043246204959972,
+      "learning_rate": 2.0011866028038617e-07,
+      "loss": 0.8999,
+      "step": 30350
+    },
+    {
+      "epoch": 0.9144578313253012,
+      "grad_norm": 4.866599488871076,
+      "learning_rate": 1.987261438183985e-07,
+      "loss": 0.9034,
+      "step": 30360
+    },
+    {
+      "epoch": 0.9147590361445783,
+      "grad_norm": 6.132504023016106,
+      "learning_rate": 1.9733839092924088e-07,
+      "loss": 0.9367,
+      "step": 30370
+    },
+    {
+      "epoch": 0.9150602409638554,
+      "grad_norm": 6.151309864962023,
+      "learning_rate": 1.9595540298977035e-07,
+      "loss": 0.9902,
+      "step": 30380
+    },
+    {
+      "epoch": 0.9153614457831325,
+      "grad_norm": 12.372261174562896,
+      "learning_rate": 1.9457718137211423e-07,
+      "loss": 0.9355,
+      "step": 30390
+    },
+    {
+      "epoch": 0.9156626506024096,
+      "grad_norm": 4.628814514104804,
+      "learning_rate": 1.9320372744367155e-07,
+      "loss": 0.986,
+      "step": 30400
+    },
+    {
+      "epoch": 0.9159638554216868,
+      "grad_norm": 4.252012260236093,
+      "learning_rate": 1.9183504256711393e-07,
+      "loss": 0.8832,
+      "step": 30410
+    },
+    {
+      "epoch": 0.9162650602409639,
+      "grad_norm": 1.9118522641793776,
+      "learning_rate": 1.9047112810037626e-07,
+      "loss": 0.9056,
+      "step": 30420
+    },
+    {
+      "epoch": 0.916566265060241,
+      "grad_norm": 8.01935915873801,
+      "learning_rate": 1.8911198539666387e-07,
+      "loss": 0.8082,
+      "step": 30430
+    },
+    {
+      "epoch": 0.9168674698795181,
+      "grad_norm": 4.406163984505128,
+      "learning_rate": 1.877576158044464e-07,
+      "loss": 0.9766,
+      "step": 30440
+    },
+    {
+      "epoch": 0.9171686746987951,
+      "grad_norm": 1.882210522818602,
+      "learning_rate": 1.8640802066746067e-07,
+      "loss": 0.9741,
+      "step": 30450
+    },
+    {
+      "epoch": 0.9174698795180722,
+      "grad_norm": 6.823712374912623,
+      "learning_rate": 1.8506320132470334e-07,
+      "loss": 0.9299,
+      "step": 30460
+    },
+    {
+      "epoch": 0.9177710843373494,
+      "grad_norm": 6.084795298150425,
+      "learning_rate": 1.8372315911043325e-07,
+      "loss": 0.8878,
+      "step": 30470
+    },
+    {
+      "epoch": 0.9180722891566265,
+      "grad_norm": 4.503662702585053,
+      "learning_rate": 1.8238789535416967e-07,
+      "loss": 0.9787,
+      "step": 30480
+    },
+    {
+      "epoch": 0.9183734939759036,
+      "grad_norm": 1.9354436821973109,
+      "learning_rate": 1.8105741138069288e-07,
+      "loss": 0.901,
+      "step": 30490
+    },
+    {
+      "epoch": 0.9186746987951807,
+      "grad_norm": 3.836766585626372,
+      "learning_rate": 1.7973170851003918e-07,
+      "loss": 0.8324,
+      "step": 30500
+    },
+    {
+      "epoch": 0.9189759036144578,
+      "grad_norm": 4.612246788895821,
+      "learning_rate": 1.7841078805750146e-07,
+      "loss": 0.9591,
+      "step": 30510
+    },
+    {
+      "epoch": 0.9192771084337349,
+      "grad_norm": 5.449970799952609,
+      "learning_rate": 1.7709465133362803e-07,
+      "loss": 0.8628,
+      "step": 30520
+    },
+    {
+      "epoch": 0.9195783132530121,
+      "grad_norm": 6.4991130263966985,
+      "learning_rate": 1.7578329964422047e-07,
+      "loss": 0.9341,
+      "step": 30530
+    },
+    {
+      "epoch": 0.9198795180722892,
+      "grad_norm": 5.400668146662582,
+      "learning_rate": 1.7447673429033361e-07,
+      "loss": 0.7959,
+      "step": 30540
+    },
+    {
+      "epoch": 0.9201807228915663,
+      "grad_norm": 1.9159164496951613,
+      "learning_rate": 1.7317495656827278e-07,
+      "loss": 0.845,
+      "step": 30550
+    },
+    {
+      "epoch": 0.9204819277108434,
+      "grad_norm": 4.772949695590476,
+      "learning_rate": 1.7187796776959586e-07,
+      "loss": 0.9239,
+      "step": 30560
+    },
+    {
+      "epoch": 0.9207831325301205,
+      "grad_norm": 5.5642641650127995,
+      "learning_rate": 1.705857691811047e-07,
+      "loss": 0.8805,
+      "step": 30570
+    },
+    {
+      "epoch": 0.9210843373493975,
+      "grad_norm": 5.006196830374669,
+      "learning_rate": 1.6929836208485206e-07,
+      "loss": 0.881,
+      "step": 30580
+    },
+    {
+      "epoch": 0.9213855421686747,
+      "grad_norm": 4.193056881037043,
+      "learning_rate": 1.6801574775813623e-07,
+      "loss": 0.9169,
+      "step": 30590
+    },
+    {
+      "epoch": 0.9216867469879518,
+      "grad_norm": 4.17810379679463,
+      "learning_rate": 1.6673792747349983e-07,
+      "loss": 0.9344,
+      "step": 30600
+    },
+    {
+      "epoch": 0.9219879518072289,
+      "grad_norm": 1.9398758124726878,
+      "learning_rate": 1.6546490249873037e-07,
+      "loss": 0.8758,
+      "step": 30610
+    },
+    {
+      "epoch": 0.922289156626506,
+      "grad_norm": 5.275194518414215,
+      "learning_rate": 1.6419667409685425e-07,
+      "loss": 0.8999,
+      "step": 30620
+    },
+    {
+      "epoch": 0.9225903614457831,
+      "grad_norm": 5.263900088497955,
+      "learning_rate": 1.6293324352614325e-07,
+      "loss": 0.873,
+      "step": 30630
+    },
+    {
+      "epoch": 0.9228915662650602,
+      "grad_norm": 2.0493681426677814,
+      "learning_rate": 1.6167461204010692e-07,
+      "loss": 0.9109,
+      "step": 30640
+    },
+    {
+      "epoch": 0.9231927710843374,
+      "grad_norm": 4.9495360231559955,
+      "learning_rate": 1.6042078088749357e-07,
+      "loss": 0.8935,
+      "step": 30650
+    },
+    {
+      "epoch": 0.9234939759036145,
+      "grad_norm": 4.5307800067679365,
+      "learning_rate": 1.5917175131228812e-07,
+      "loss": 0.8849,
+      "step": 30660
+    },
+    {
+      "epoch": 0.9237951807228916,
+      "grad_norm": 5.417541286798659,
+      "learning_rate": 1.579275245537132e-07,
+      "loss": 0.9285,
+      "step": 30670
+    },
+    {
+      "epoch": 0.9240963855421687,
+      "grad_norm": 4.290994336416196,
+      "learning_rate": 1.5668810184622517e-07,
+      "loss": 0.9537,
+      "step": 30680
+    },
+    {
+      "epoch": 0.9243975903614458,
+      "grad_norm": 6.093207043891591,
+      "learning_rate": 1.5545348441951437e-07,
+      "loss": 0.9438,
+      "step": 30690
+    },
+    {
+      "epoch": 0.9246987951807228,
+      "grad_norm": 2.210913932320992,
+      "learning_rate": 1.5422367349850364e-07,
+      "loss": 0.9114,
+      "step": 30700
+    },
+    {
+      "epoch": 0.925,
+      "grad_norm": 5.106178602573226,
+      "learning_rate": 1.5299867030334815e-07,
+      "loss": 0.8904,
+      "step": 30710
+    },
+    {
+      "epoch": 0.9253012048192771,
+      "grad_norm": 2.162176062789831,
+      "learning_rate": 1.5177847604943062e-07,
+      "loss": 0.9336,
+      "step": 30720
+    },
+    {
+      "epoch": 0.9256024096385542,
+      "grad_norm": 3.841243784591798,
+      "learning_rate": 1.5056309194736385e-07,
+      "loss": 0.8639,
+      "step": 30730
+    },
+    {
+      "epoch": 0.9259036144578313,
+      "grad_norm": 7.206173562157358,
+      "learning_rate": 1.493525192029899e-07,
+      "loss": 0.9642,
+      "step": 30740
+    },
+    {
+      "epoch": 0.9262048192771084,
+      "grad_norm": 4.856945621486708,
+      "learning_rate": 1.4814675901737408e-07,
+      "loss": 1.0129,
+      "step": 30750
+    },
+    {
+      "epoch": 0.9265060240963855,
+      "grad_norm": 2.023495186035137,
+      "learning_rate": 1.469458125868095e-07,
+      "loss": 0.949,
+      "step": 30760
+    },
+    {
+      "epoch": 0.9268072289156627,
+      "grad_norm": 7.040211268226052,
+      "learning_rate": 1.4574968110281251e-07,
+      "loss": 0.812,
+      "step": 30770
+    },
+    {
+      "epoch": 0.9271084337349398,
+      "grad_norm": 1.973717105095577,
+      "learning_rate": 1.4455836575212157e-07,
+      "loss": 0.8822,
+      "step": 30780
+    },
+    {
+      "epoch": 0.9274096385542169,
+      "grad_norm": 6.261733297612659,
+      "learning_rate": 1.4337186771669841e-07,
+      "loss": 0.8421,
+      "step": 30790
+    },
+    {
+      "epoch": 0.927710843373494,
+      "grad_norm": 5.860160504921429,
+      "learning_rate": 1.421901881737231e-07,
+      "loss": 0.8834,
+      "step": 30800
+    },
+    {
+      "epoch": 0.928012048192771,
+      "grad_norm": 5.592990358537241,
+      "learning_rate": 1.410133282955972e-07,
+      "loss": 0.9817,
+      "step": 30810
+    },
+    {
+      "epoch": 0.9283132530120481,
+      "grad_norm": 6.692170712749857,
+      "learning_rate": 1.3984128924993955e-07,
+      "loss": 0.9008,
+      "step": 30820
+    },
+    {
+      "epoch": 0.9286144578313253,
+      "grad_norm": 4.387551000030284,
+      "learning_rate": 1.3867407219958496e-07,
+      "loss": 0.8981,
+      "step": 30830
+    },
+    {
+      "epoch": 0.9289156626506024,
+      "grad_norm": 5.003253856961021,
+      "learning_rate": 1.3751167830258594e-07,
+      "loss": 0.9477,
+      "step": 30840
+    },
+    {
+      "epoch": 0.9292168674698795,
+      "grad_norm": 4.538494758761143,
+      "learning_rate": 1.3635410871220944e-07,
+      "loss": 0.8519,
+      "step": 30850
+    },
+    {
+      "epoch": 0.9295180722891566,
+      "grad_norm": 7.638884039798874,
+      "learning_rate": 1.3520136457693512e-07,
+      "loss": 0.9269,
+      "step": 30860
+    },
+    {
+      "epoch": 0.9298192771084337,
+      "grad_norm": 5.134368648792981,
+      "learning_rate": 1.3405344704045477e-07,
+      "loss": 0.9676,
+      "step": 30870
+    },
+    {
+      "epoch": 0.9301204819277108,
+      "grad_norm": 5.596770449556262,
+      "learning_rate": 1.3291035724167233e-07,
+      "loss": 0.9257,
+      "step": 30880
+    },
+    {
+      "epoch": 0.930421686746988,
+      "grad_norm": 4.989679447843252,
+      "learning_rate": 1.3177209631470233e-07,
+      "loss": 0.9817,
+      "step": 30890
+    },
+    {
+      "epoch": 0.9307228915662651,
+      "grad_norm": 3.975918294128282,
+      "learning_rate": 1.306386653888675e-07,
+      "loss": 0.8391,
+      "step": 30900
+    },
+    {
+      "epoch": 0.9310240963855422,
+      "grad_norm": 5.902556027724682,
+      "learning_rate": 1.2951006558869883e-07,
+      "loss": 0.9168,
+      "step": 30910
+    },
+    {
+      "epoch": 0.9313253012048193,
+      "grad_norm": 2.0539439601061664,
+      "learning_rate": 1.2838629803393343e-07,
+      "loss": 0.9179,
+      "step": 30920
+    },
+    {
+      "epoch": 0.9316265060240964,
+      "grad_norm": 4.9912392983740554,
+      "learning_rate": 1.27267363839515e-07,
+      "loss": 0.8725,
+      "step": 30930
+    },
+    {
+      "epoch": 0.9319277108433734,
+      "grad_norm": 4.393343174533207,
+      "learning_rate": 1.2615326411559102e-07,
+      "loss": 0.9116,
+      "step": 30940
+    },
+    {
+      "epoch": 0.9322289156626506,
+      "grad_norm": 5.64237266061262,
+      "learning_rate": 1.2504399996751458e-07,
+      "loss": 0.9591,
+      "step": 30950
+    },
+    {
+      "epoch": 0.9325301204819277,
+      "grad_norm": 4.303841741723094,
+      "learning_rate": 1.2393957249583865e-07,
+      "loss": 0.9469,
+      "step": 30960
+    },
+    {
+      "epoch": 0.9328313253012048,
+      "grad_norm": 4.9202804625223715,
+      "learning_rate": 1.2283998279631782e-07,
+      "loss": 0.8744,
+      "step": 30970
+    },
+    {
+      "epoch": 0.9331325301204819,
+      "grad_norm": 4.117656483874047,
+      "learning_rate": 1.2174523195990774e-07,
+      "loss": 0.9997,
+      "step": 30980
+    },
+    {
+      "epoch": 0.933433734939759,
+      "grad_norm": 5.926551823705494,
+      "learning_rate": 1.2065532107276346e-07,
+      "loss": 0.8809,
+      "step": 30990
+    },
+    {
+      "epoch": 0.9337349397590361,
+      "grad_norm": 2.1586242117054866,
+      "learning_rate": 1.1957025121623723e-07,
+      "loss": 0.8989,
+      "step": 31000
+    },
+    {
+      "epoch": 0.9340361445783133,
+      "grad_norm": 2.0099367131032566,
+      "learning_rate": 1.1849002346687843e-07,
+      "loss": 0.8628,
+      "step": 31010
+    },
+    {
+      "epoch": 0.9343373493975904,
+      "grad_norm": 6.407564840941291,
+      "learning_rate": 1.1741463889643312e-07,
+      "loss": 0.8549,
+      "step": 31020
+    },
+    {
+      "epoch": 0.9346385542168675,
+      "grad_norm": 2.203204610823805,
+      "learning_rate": 1.1634409857184115e-07,
+      "loss": 0.9774,
+      "step": 31030
+    },
+    {
+      "epoch": 0.9349397590361446,
+      "grad_norm": 2.0892029281619084,
+      "learning_rate": 1.1527840355523678e-07,
+      "loss": 0.8511,
+      "step": 31040
+    },
+    {
+      "epoch": 0.9352409638554217,
+      "grad_norm": 6.670492149767532,
+      "learning_rate": 1.1421755490394703e-07,
+      "loss": 0.9171,
+      "step": 31050
+    },
+    {
+      "epoch": 0.9355421686746987,
+      "grad_norm": 3.9126140592894263,
+      "learning_rate": 1.1316155367048997e-07,
+      "loss": 0.9736,
+      "step": 31060
+    },
+    {
+      "epoch": 0.935843373493976,
+      "grad_norm": 5.508606439584242,
+      "learning_rate": 1.121104009025753e-07,
+      "loss": 0.9986,
+      "step": 31070
+    },
+    {
+      "epoch": 0.936144578313253,
+      "grad_norm": 4.513596381840643,
+      "learning_rate": 1.1106409764310099e-07,
+      "loss": 0.9629,
+      "step": 31080
+    },
+    {
+      "epoch": 0.9364457831325301,
+      "grad_norm": 4.458743380851925,
+      "learning_rate": 1.1002264493015557e-07,
+      "loss": 0.9075,
+      "step": 31090
+    },
+    {
+      "epoch": 0.9367469879518072,
+      "grad_norm": 2.0394683884665663,
+      "learning_rate": 1.089860437970136e-07,
+      "loss": 0.9574,
+      "step": 31100
+    },
+    {
+      "epoch": 0.9370481927710843,
+      "grad_norm": 4.490584890742716,
+      "learning_rate": 1.0795429527213685e-07,
+      "loss": 0.8513,
+      "step": 31110
+    },
+    {
+      "epoch": 0.9373493975903614,
+      "grad_norm": 4.560890693225779,
+      "learning_rate": 1.0692740037917038e-07,
+      "loss": 0.8868,
+      "step": 31120
+    },
+    {
+      "epoch": 0.9376506024096386,
+      "grad_norm": 5.304866129428448,
+      "learning_rate": 1.0590536013694808e-07,
+      "loss": 0.951,
+      "step": 31130
+    },
+    {
+      "epoch": 0.9379518072289157,
+      "grad_norm": 4.785267266198701,
+      "learning_rate": 1.0488817555948328e-07,
+      "loss": 0.9663,
+      "step": 31140
+    },
+    {
+      "epoch": 0.9382530120481928,
+      "grad_norm": 4.757059489339679,
+      "learning_rate": 1.0387584765597425e-07,
+      "loss": 0.9207,
+      "step": 31150
+    },
+    {
+      "epoch": 0.9385542168674699,
+      "grad_norm": 6.21566281231201,
+      "learning_rate": 1.0286837743079869e-07,
+      "loss": 0.878,
+      "step": 31160
+    },
+    {
+      "epoch": 0.938855421686747,
+      "grad_norm": 7.000515722109719,
+      "learning_rate": 1.0186576588351704e-07,
+      "loss": 0.9788,
+      "step": 31170
+    },
+    {
+      "epoch": 0.939156626506024,
+      "grad_norm": 4.50859646721263,
+      "learning_rate": 1.0086801400886748e-07,
+      "loss": 0.9873,
+      "step": 31180
+    },
+    {
+      "epoch": 0.9394578313253013,
+      "grad_norm": 4.815024079717224,
+      "learning_rate": 9.987512279676648e-08,
+      "loss": 0.9023,
+      "step": 31190
+    },
+    {
+      "epoch": 0.9397590361445783,
+      "grad_norm": 5.786073200387487,
+      "learning_rate": 9.88870932323105e-08,
+      "loss": 0.9268,
+      "step": 31200
+    },
+    {
+      "epoch": 0.9400602409638554,
+      "grad_norm": 5.513886971916418,
+      "learning_rate": 9.790392629576928e-08,
+      "loss": 0.9748,
+      "step": 31210
+    },
+    {
+      "epoch": 0.9403614457831325,
+      "grad_norm": 6.165981363990004,
+      "learning_rate": 9.692562296259034e-08,
+      "loss": 0.9337,
+      "step": 31220
+    },
+    {
+      "epoch": 0.9406626506024096,
+      "grad_norm": 5.622474764916678,
+      "learning_rate": 9.595218420339502e-08,
+      "loss": 0.9343,
+      "step": 31230
+    },
+    {
+      "epoch": 0.9409638554216867,
+      "grad_norm": 4.29689025272055,
+      "learning_rate": 9.498361098397857e-08,
+      "loss": 0.8573,
+      "step": 31240
+    },
+    {
+      "epoch": 0.9412650602409639,
+      "grad_norm": 2.1228911427173878,
+      "learning_rate": 9.401990426530783e-08,
+      "loss": 0.8563,
+      "step": 31250
+    },
+    {
+      "epoch": 0.941566265060241,
+      "grad_norm": 5.556232352565769,
+      "learning_rate": 9.306106500352352e-08,
+      "loss": 0.8452,
+      "step": 31260
+    },
+    {
+      "epoch": 0.9418674698795181,
+      "grad_norm": 9.77660883553847,
+      "learning_rate": 9.210709414993468e-08,
+      "loss": 0.9925,
+      "step": 31270
+    },
+    {
+      "epoch": 0.9421686746987952,
+      "grad_norm": 7.768326396804313,
+      "learning_rate": 9.115799265102144e-08,
+      "loss": 0.9125,
+      "step": 31280
+    },
+    {
+      "epoch": 0.9424698795180723,
+      "grad_norm": 6.1472972464912425,
+      "learning_rate": 9.021376144843331e-08,
+      "loss": 0.944,
+      "step": 31290
+    },
+    {
+      "epoch": 0.9427710843373494,
+      "grad_norm": 4.65595788334406,
+      "learning_rate": 8.927440147898703e-08,
+      "loss": 0.9154,
+      "step": 31300
+    },
+    {
+      "epoch": 0.9430722891566266,
+      "grad_norm": 7.772291502764226,
+      "learning_rate": 8.833991367466543e-08,
+      "loss": 0.9027,
+      "step": 31310
+    },
+    {
+      "epoch": 0.9433734939759036,
+      "grad_norm": 6.089237439721673,
+      "learning_rate": 8.741029896261965e-08,
+      "loss": 0.9101,
+      "step": 31320
+    },
+    {
+      "epoch": 0.9436746987951807,
+      "grad_norm": 1.9917472325210748,
+      "learning_rate": 8.648555826516358e-08,
+      "loss": 0.8349,
+      "step": 31330
+    },
+    {
+      "epoch": 0.9439759036144578,
+      "grad_norm": 1.8828427949970825,
+      "learning_rate": 8.556569249977776e-08,
+      "loss": 0.8208,
+      "step": 31340
+    },
+    {
+      "epoch": 0.9442771084337349,
+      "grad_norm": 1.9236935981975551,
+      "learning_rate": 8.465070257910545e-08,
+      "loss": 0.9346,
+      "step": 31350
+    },
+    {
+      "epoch": 0.944578313253012,
+      "grad_norm": 6.512249637191026,
+      "learning_rate": 8.374058941094943e-08,
+      "loss": 0.9793,
+      "step": 31360
+    },
+    {
+      "epoch": 0.9448795180722892,
+      "grad_norm": 5.268319202207796,
+      "learning_rate": 8.283535389827791e-08,
+      "loss": 0.8733,
+      "step": 31370
+    },
+    {
+      "epoch": 0.9451807228915663,
+      "grad_norm": 4.2062384269533,
+      "learning_rate": 8.193499693921858e-08,
+      "loss": 0.9278,
+      "step": 31380
+    },
+    {
+      "epoch": 0.9454819277108434,
+      "grad_norm": 2.2600880356395043,
+      "learning_rate": 8.103951942705746e-08,
+      "loss": 0.7669,
+      "step": 31390
+    },
+    {
+      "epoch": 0.9457831325301205,
+      "grad_norm": 2.011035092391403,
+      "learning_rate": 8.014892225024162e-08,
+      "loss": 0.9416,
+      "step": 31400
+    },
+    {
+      "epoch": 0.9460843373493976,
+      "grad_norm": 4.914858948626751,
+      "learning_rate": 7.926320629237372e-08,
+      "loss": 0.8178,
+      "step": 31410
+    },
+    {
+      "epoch": 0.9463855421686747,
+      "grad_norm": 4.724209174896651,
+      "learning_rate": 7.838237243221636e-08,
+      "loss": 0.8016,
+      "step": 31420
+    },
+    {
+      "epoch": 0.9466867469879519,
+      "grad_norm": 5.234814754925279,
+      "learning_rate": 7.750642154368549e-08,
+      "loss": 0.8757,
+      "step": 31430
+    },
+    {
+      "epoch": 0.946987951807229,
+      "grad_norm": 2.009043731039072,
+      "learning_rate": 7.663535449585368e-08,
+      "loss": 0.8948,
+      "step": 31440
+    },
+    {
+      "epoch": 0.947289156626506,
+      "grad_norm": 5.637252015383081,
+      "learning_rate": 7.57691721529502e-08,
+      "loss": 0.9159,
+      "step": 31450
+    },
+    {
+      "epoch": 0.9475903614457831,
+      "grad_norm": 7.47274754525009,
+      "learning_rate": 7.490787537435318e-08,
+      "loss": 1.01,
+      "step": 31460
+    },
+    {
+      "epoch": 0.9478915662650602,
+      "grad_norm": 5.614873453119678,
+      "learning_rate": 7.405146501459737e-08,
+      "loss": 0.8489,
+      "step": 31470
+    },
+    {
+      "epoch": 0.9481927710843373,
+      "grad_norm": 5.174851348793695,
+      "learning_rate": 7.319994192336922e-08,
+      "loss": 0.8795,
+      "step": 31480
+    },
+    {
+      "epoch": 0.9484939759036145,
+      "grad_norm": 1.9722186899695042,
+      "learning_rate": 7.235330694550402e-08,
+      "loss": 0.9215,
+      "step": 31490
+    },
+    {
+      "epoch": 0.9487951807228916,
+      "grad_norm": 4.67144163926662,
+      "learning_rate": 7.151156092098987e-08,
+      "loss": 0.9562,
+      "step": 31500
+    },
+    {
+      "epoch": 0.9490963855421687,
+      "grad_norm": 6.409503643827968,
+      "learning_rate": 7.067470468496207e-08,
+      "loss": 0.8968,
+      "step": 31510
+    },
+    {
+      "epoch": 0.9493975903614458,
+      "grad_norm": 5.000686619576055,
+      "learning_rate": 6.984273906770645e-08,
+      "loss": 0.7606,
+      "step": 31520
+    },
+    {
+      "epoch": 0.9496987951807229,
+      "grad_norm": 5.353730688646645,
+      "learning_rate": 6.901566489465494e-08,
+      "loss": 0.9252,
+      "step": 31530
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.48665505738073,
+      "learning_rate": 6.819348298638839e-08,
+      "loss": 0.9965,
+      "step": 31540
+    },
+    {
+      "epoch": 0.9503012048192772,
+      "grad_norm": 5.496642102580685,
+      "learning_rate": 6.73761941586315e-08,
+      "loss": 0.9488,
+      "step": 31550
+    },
+    {
+      "epoch": 0.9506024096385542,
+      "grad_norm": 1.8154120551670836,
+      "learning_rate": 6.656379922225676e-08,
+      "loss": 0.8733,
+      "step": 31560
+    },
+    {
+      "epoch": 0.9509036144578313,
+      "grad_norm": 21.46219883385068,
+      "learning_rate": 6.575629898327884e-08,
+      "loss": 0.979,
+      "step": 31570
+    },
+    {
+      "epoch": 0.9512048192771084,
+      "grad_norm": 2.0278059526542354,
+      "learning_rate": 6.495369424285746e-08,
+      "loss": 0.9338,
+      "step": 31580
+    },
+    {
+      "epoch": 0.9515060240963855,
+      "grad_norm": 20.890020765453254,
+      "learning_rate": 6.415598579729566e-08,
+      "loss": 0.8763,
+      "step": 31590
+    },
+    {
+      "epoch": 0.9518072289156626,
+      "grad_norm": 5.253818125619825,
+      "learning_rate": 6.336317443803808e-08,
+      "loss": 0.9124,
+      "step": 31600
+    },
+    {
+      "epoch": 0.9521084337349398,
+      "grad_norm": 2.0196564665635317,
+      "learning_rate": 6.257526095166999e-08,
+      "loss": 0.7798,
+      "step": 31610
+    },
+    {
+      "epoch": 0.9524096385542169,
+      "grad_norm": 4.460652751428492,
+      "learning_rate": 6.179224611991774e-08,
+      "loss": 0.873,
+      "step": 31620
+    },
+    {
+      "epoch": 0.952710843373494,
+      "grad_norm": 4.5934344920094015,
+      "learning_rate": 6.101413071964879e-08,
+      "loss": 0.8855,
+      "step": 31630
+    },
+    {
+      "epoch": 0.9530120481927711,
+      "grad_norm": 5.692243288757897,
+      "learning_rate": 6.024091552286781e-08,
+      "loss": 0.9729,
+      "step": 31640
+    },
+    {
+      "epoch": 0.9533132530120482,
+      "grad_norm": 4.2118553333912745,
+      "learning_rate": 5.9472601296718945e-08,
+      "loss": 0.9201,
+      "step": 31650
+    },
+    {
+      "epoch": 0.9536144578313253,
+      "grad_norm": 5.00798606357242,
+      "learning_rate": 5.8709188803482973e-08,
+      "loss": 0.9454,
+      "step": 31660
+    },
+    {
+      "epoch": 0.9539156626506025,
+      "grad_norm": 5.478380218887263,
+      "learning_rate": 5.795067880057792e-08,
+      "loss": 0.8821,
+      "step": 31670
+    },
+    {
+      "epoch": 0.9542168674698795,
+      "grad_norm": 5.5916796675069484,
+      "learning_rate": 5.7197072040557356e-08,
+      "loss": 1.0141,
+      "step": 31680
+    },
+    {
+      "epoch": 0.9545180722891566,
+      "grad_norm": 2.082708147343465,
+      "learning_rate": 5.644836927111152e-08,
+      "loss": 0.8511,
+      "step": 31690
+    },
+    {
+      "epoch": 0.9548192771084337,
+      "grad_norm": 1.9356641257286196,
+      "learning_rate": 5.57045712350629e-08,
+      "loss": 0.9663,
+      "step": 31700
+    },
+    {
+      "epoch": 0.9551204819277108,
+      "grad_norm": 10.966520500300694,
+      "learning_rate": 5.496567867037006e-08,
+      "loss": 0.9417,
+      "step": 31710
+    },
+    {
+      "epoch": 0.9554216867469879,
+      "grad_norm": 4.067956182307647,
+      "learning_rate": 5.423169231012271e-08,
+      "loss": 0.9089,
+      "step": 31720
+    },
+    {
+      "epoch": 0.9557228915662651,
+      "grad_norm": 4.452756075880683,
+      "learning_rate": 5.350261288254499e-08,
+      "loss": 0.9967,
+      "step": 31730
+    },
+    {
+      "epoch": 0.9560240963855422,
+      "grad_norm": 5.582902630953903,
+      "learning_rate": 5.277844111098995e-08,
+      "loss": 0.9021,
+      "step": 31740
+    },
+    {
+      "epoch": 0.9563253012048193,
+      "grad_norm": 6.445885955031362,
+      "learning_rate": 5.205917771394398e-08,
+      "loss": 0.9789,
+      "step": 31750
+    },
+    {
+      "epoch": 0.9566265060240964,
+      "grad_norm": 4.251170884605689,
+      "learning_rate": 5.134482340502178e-08,
+      "loss": 1.0117,
+      "step": 31760
+    },
+    {
+      "epoch": 0.9569277108433735,
+      "grad_norm": 1.9909931612164995,
+      "learning_rate": 5.063537889296921e-08,
+      "loss": 0.9339,
+      "step": 31770
+    },
+    {
+      "epoch": 0.9572289156626506,
+      "grad_norm": 4.586901197703373,
+      "learning_rate": 4.993084488165933e-08,
+      "loss": 0.9644,
+      "step": 31780
+    },
+    {
+      "epoch": 0.9575301204819278,
+      "grad_norm": 4.2002467135643835,
+      "learning_rate": 4.923122207009412e-08,
+      "loss": 0.983,
+      "step": 31790
+    },
+    {
+      "epoch": 0.9578313253012049,
+      "grad_norm": 1.9544588383018127,
+      "learning_rate": 4.8536511152402766e-08,
+      "loss": 0.9102,
+      "step": 31800
+    },
+    {
+      "epoch": 0.9581325301204819,
+      "grad_norm": 4.585237811146019,
+      "learning_rate": 4.784671281784115e-08,
+      "loss": 0.7348,
+      "step": 31810
+    },
+    {
+      "epoch": 0.958433734939759,
+      "grad_norm": 2.065119496477985,
+      "learning_rate": 4.7161827750791254e-08,
+      "loss": 0.9061,
+      "step": 31820
+    },
+    {
+      "epoch": 0.9587349397590361,
+      "grad_norm": 4.734611038207768,
+      "learning_rate": 4.648185663075955e-08,
+      "loss": 0.9375,
+      "step": 31830
+    },
+    {
+      "epoch": 0.9590361445783132,
+      "grad_norm": 4.172196493271511,
+      "learning_rate": 4.580680013237915e-08,
+      "loss": 0.8531,
+      "step": 31840
+    },
+    {
+      "epoch": 0.9593373493975904,
+      "grad_norm": 2.1088456606552666,
+      "learning_rate": 4.513665892540486e-08,
+      "loss": 0.8898,
+      "step": 31850
+    },
+    {
+      "epoch": 0.9596385542168675,
+      "grad_norm": 2.1762155217981154,
+      "learning_rate": 4.44714336747154e-08,
+      "loss": 0.9557,
+      "step": 31860
+    },
+    {
+      "epoch": 0.9599397590361446,
+      "grad_norm": 4.08967158877884,
+      "learning_rate": 4.381112504031337e-08,
+      "loss": 0.8883,
+      "step": 31870
+    },
+    {
+      "epoch": 0.9602409638554217,
+      "grad_norm": 4.736864621834937,
+      "learning_rate": 4.3155733677321397e-08,
+      "loss": 0.9858,
+      "step": 31880
+    },
+    {
+      "epoch": 0.9605421686746988,
+      "grad_norm": 5.065952788450292,
+      "learning_rate": 4.2505260235986e-08,
+      "loss": 0.8857,
+      "step": 31890
+    },
+    {
+      "epoch": 0.9608433734939759,
+      "grad_norm": 4.572014718952944,
+      "learning_rate": 4.18597053616715e-08,
+      "loss": 0.9817,
+      "step": 31900
+    },
+    {
+      "epoch": 0.9611445783132531,
+      "grad_norm": 6.998044500990425,
+      "learning_rate": 4.121906969486444e-08,
+      "loss": 0.9341,
+      "step": 31910
+    },
+    {
+      "epoch": 0.9614457831325302,
+      "grad_norm": 5.199759990994423,
+      "learning_rate": 4.058335387117029e-08,
+      "loss": 0.8573,
+      "step": 31920
+    },
+    {
+      "epoch": 0.9617469879518072,
+      "grad_norm": 1.9173809977988288,
+      "learning_rate": 3.9952558521311746e-08,
+      "loss": 0.8957,
+      "step": 31930
+    },
+    {
+      "epoch": 0.9620481927710843,
+      "grad_norm": 1.9320919093829663,
+      "learning_rate": 3.932668427113262e-08,
+      "loss": 0.9187,
+      "step": 31940
+    },
+    {
+      "epoch": 0.9623493975903614,
+      "grad_norm": 8.590303173871973,
+      "learning_rate": 3.87057317415912e-08,
+      "loss": 0.8842,
+      "step": 31950
+    },
+    {
+      "epoch": 0.9626506024096385,
+      "grad_norm": 6.829054113107583,
+      "learning_rate": 3.8089701548765237e-08,
+      "loss": 0.9649,
+      "step": 31960
+    },
+    {
+      "epoch": 0.9629518072289157,
+      "grad_norm": 5.002661137745544,
+      "learning_rate": 3.747859430384637e-08,
+      "loss": 0.9588,
+      "step": 31970
+    },
+    {
+      "epoch": 0.9632530120481928,
+      "grad_norm": 1.833743228665942,
+      "learning_rate": 3.687241061314462e-08,
+      "loss": 0.8163,
+      "step": 31980
+    },
+    {
+      "epoch": 0.9635542168674699,
+      "grad_norm": 3.858606939722687,
+      "learning_rate": 3.627115107808332e-08,
+      "loss": 0.9026,
+      "step": 31990
+    },
+    {
+      "epoch": 0.963855421686747,
+      "grad_norm": 5.733580986982268,
+      "learning_rate": 3.567481629520086e-08,
+      "loss": 0.9104,
+      "step": 32000
+    },
+    {
+      "epoch": 0.9641566265060241,
+      "grad_norm": 1.916862277391112,
+      "learning_rate": 3.508340685614842e-08,
+      "loss": 0.8529,
+      "step": 32010
+    },
+    {
+      "epoch": 0.9644578313253012,
+      "grad_norm": 2.081678156309162,
+      "learning_rate": 3.449692334769217e-08,
+      "loss": 0.9556,
+      "step": 32020
+    },
+    {
+      "epoch": 0.9647590361445784,
+      "grad_norm": 6.252452685789378,
+      "learning_rate": 3.391536635171055e-08,
+      "loss": 0.8811,
+      "step": 32030
+    },
+    {
+      "epoch": 0.9650602409638555,
+      "grad_norm": 14.46103305347846,
+      "learning_rate": 3.333873644519425e-08,
+      "loss": 0.9338,
+      "step": 32040
+    },
+    {
+      "epoch": 0.9653614457831325,
+      "grad_norm": 5.190856685469573,
+      "learning_rate": 3.2767034200245076e-08,
+      "loss": 0.9446,
+      "step": 32050
+    },
+    {
+      "epoch": 0.9656626506024096,
+      "grad_norm": 4.403507020006852,
+      "learning_rate": 3.220026018407541e-08,
+      "loss": 0.9359,
+      "step": 32060
+    },
+    {
+      "epoch": 0.9659638554216867,
+      "grad_norm": 4.787761078065262,
+      "learning_rate": 3.163841495900988e-08,
+      "loss": 0.9338,
+      "step": 32070
+    },
+    {
+      "epoch": 0.9662650602409638,
+      "grad_norm": 1.9184515601386845,
+      "learning_rate": 3.108149908248093e-08,
+      "loss": 0.9012,
+      "step": 32080
+    },
+    {
+      "epoch": 0.966566265060241,
+      "grad_norm": 2.0031446249235,
+      "learning_rate": 3.052951310703212e-08,
+      "loss": 0.8868,
+      "step": 32090
+    },
+    {
+      "epoch": 0.9668674698795181,
+      "grad_norm": 4.95976830218123,
+      "learning_rate": 2.998245758031482e-08,
+      "loss": 0.9571,
+      "step": 32100
+    },
+    {
+      "epoch": 0.9671686746987952,
+      "grad_norm": 5.087747737811243,
+      "learning_rate": 2.9440333045088753e-08,
+      "loss": 0.9274,
+      "step": 32110
+    },
+    {
+      "epoch": 0.9674698795180723,
+      "grad_norm": 1.9009154478365193,
+      "learning_rate": 2.8903140039222012e-08,
+      "loss": 0.8484,
+      "step": 32120
+    },
+    {
+      "epoch": 0.9677710843373494,
+      "grad_norm": 8.640822327018284,
+      "learning_rate": 2.8370879095688807e-08,
+      "loss": 0.9061,
+      "step": 32130
+    },
+    {
+      "epoch": 0.9680722891566265,
+      "grad_norm": 4.714580331834582,
+      "learning_rate": 2.7843550742570614e-08,
+      "loss": 0.9258,
+      "step": 32140
+    },
+    {
+      "epoch": 0.9683734939759037,
+      "grad_norm": 5.033035821592232,
+      "learning_rate": 2.7321155503055584e-08,
+      "loss": 0.9246,
+      "step": 32150
+    },
+    {
+      "epoch": 0.9686746987951808,
+      "grad_norm": 5.628529182506936,
+      "learning_rate": 2.680369389543691e-08,
+      "loss": 0.9865,
+      "step": 32160
+    },
+    {
+      "epoch": 0.9689759036144578,
+      "grad_norm": 10.175220114494271,
+      "learning_rate": 2.629116643311225e-08,
+      "loss": 0.9374,
+      "step": 32170
+    },
+    {
+      "epoch": 0.9692771084337349,
+      "grad_norm": 6.520144800887773,
+      "learning_rate": 2.578357362458539e-08,
+      "loss": 0.8914,
+      "step": 32180
+    },
+    {
+      "epoch": 0.969578313253012,
+      "grad_norm": 1.9414490245596985,
+      "learning_rate": 2.5280915973462385e-08,
+      "loss": 0.8121,
+      "step": 32190
+    },
+    {
+      "epoch": 0.9698795180722891,
+      "grad_norm": 5.199655962184366,
+      "learning_rate": 2.4783193978454857e-08,
+      "loss": 0.9901,
+      "step": 32200
+    },
+    {
+      "epoch": 0.9701807228915663,
+      "grad_norm": 7.304053814674291,
+      "learning_rate": 2.429040813337613e-08,
+      "loss": 0.947,
+      "step": 32210
+    },
+    {
+      "epoch": 0.9704819277108434,
+      "grad_norm": 4.89995129035935,
+      "learning_rate": 2.380255892714234e-08,
+      "loss": 0.8451,
+      "step": 32220
+    },
+    {
+      "epoch": 0.9707831325301205,
+      "grad_norm": 6.366477750083791,
+      "learning_rate": 2.3319646843772416e-08,
+      "loss": 0.9329,
+      "step": 32230
+    },
+    {
+      "epoch": 0.9710843373493976,
+      "grad_norm": 7.3058390489962886,
+      "learning_rate": 2.2841672362386435e-08,
+      "loss": 0.9899,
+      "step": 32240
+    },
+    {
+      "epoch": 0.9713855421686747,
+      "grad_norm": 7.049171259139614,
+      "learning_rate": 2.236863595720562e-08,
+      "loss": 0.9625,
+      "step": 32250
+    },
+    {
+      "epoch": 0.9716867469879518,
+      "grad_norm": 7.964210915652266,
+      "learning_rate": 2.1900538097551217e-08,
+      "loss": 0.9222,
+      "step": 32260
+    },
+    {
+      "epoch": 0.971987951807229,
+      "grad_norm": 6.530625776104598,
+      "learning_rate": 2.143737924784617e-08,
+      "loss": 0.913,
+      "step": 32270
+    },
+    {
+      "epoch": 0.9722891566265061,
+      "grad_norm": 5.324411150574649,
+      "learning_rate": 2.097915986761234e-08,
+      "loss": 0.91,
+      "step": 32280
+    },
+    {
+      "epoch": 0.9725903614457831,
+      "grad_norm": 9.723645218471404,
+      "learning_rate": 2.052588041147108e-08,
+      "loss": 0.9928,
+      "step": 32290
+    },
+    {
+      "epoch": 0.9728915662650602,
+      "grad_norm": 5.943995608444509,
+      "learning_rate": 2.0077541329142636e-08,
+      "loss": 0.9234,
+      "step": 32300
+    },
+    {
+      "epoch": 0.9731927710843373,
+      "grad_norm": 1.851216045599393,
+      "learning_rate": 1.9634143065445645e-08,
+      "loss": 0.9536,
+      "step": 32310
+    },
+    {
+      "epoch": 0.9734939759036144,
+      "grad_norm": 5.6125492081688,
+      "learning_rate": 1.9195686060295982e-08,
+      "loss": 0.9597,
+      "step": 32320
+    },
+    {
+      "epoch": 0.9737951807228916,
+      "grad_norm": 4.996823292904083,
+      "learning_rate": 1.8762170748708454e-08,
+      "loss": 0.8892,
+      "step": 32330
+    },
+    {
+      "epoch": 0.9740963855421687,
+      "grad_norm": 5.104708268626902,
+      "learning_rate": 1.833359756079456e-08,
+      "loss": 1.0204,
+      "step": 32340
+    },
+    {
+      "epoch": 0.9743975903614458,
+      "grad_norm": 7.326232584172267,
+      "learning_rate": 1.790996692176139e-08,
+      "loss": 0.8854,
+      "step": 32350
+    },
+    {
+      "epoch": 0.9746987951807229,
+      "grad_norm": 1.8982627522393387,
+      "learning_rate": 1.7491279251913297e-08,
+      "loss": 0.926,
+      "step": 32360
+    },
+    {
+      "epoch": 0.975,
+      "grad_norm": 1.961582971142859,
+      "learning_rate": 1.7077534966650767e-08,
+      "loss": 0.9338,
+      "step": 32370
+    },
+    {
+      "epoch": 0.9753012048192771,
+      "grad_norm": 7.204296466118696,
+      "learning_rate": 1.6668734476468775e-08,
+      "loss": 0.9141,
+      "step": 32380
+    },
+    {
+      "epoch": 0.9756024096385543,
+      "grad_norm": 4.445754732635552,
+      "learning_rate": 1.6264878186958434e-08,
+      "loss": 0.9337,
+      "step": 32390
+    },
+    {
+      "epoch": 0.9759036144578314,
+      "grad_norm": 4.664311706562546,
+      "learning_rate": 1.5865966498803674e-08,
+      "loss": 0.9858,
+      "step": 32400
+    },
+    {
+      "epoch": 0.9762048192771084,
+      "grad_norm": 5.723175316978166,
+      "learning_rate": 1.5471999807784578e-08,
+      "loss": 0.8587,
+      "step": 32410
+    },
+    {
+      "epoch": 0.9765060240963855,
+      "grad_norm": 6.969543067650618,
+      "learning_rate": 1.508297850477458e-08,
+      "loss": 0.9664,
+      "step": 32420
+    },
+    {
+      "epoch": 0.9768072289156626,
+      "grad_norm": 1.7843877358981983,
+      "learning_rate": 1.4698902975739948e-08,
+      "loss": 0.9528,
+      "step": 32430
+    },
+    {
+      "epoch": 0.9771084337349397,
+      "grad_norm": 1.9878703513714073,
+      "learning_rate": 1.431977360173975e-08,
+      "loss": 0.8372,
+      "step": 32440
+    },
+    {
+      "epoch": 0.9774096385542169,
+      "grad_norm": 5.725179353496761,
+      "learning_rate": 1.3945590758927541e-08,
+      "loss": 0.888,
+      "step": 32450
+    },
+    {
+      "epoch": 0.977710843373494,
+      "grad_norm": 5.246160467460025,
+      "learning_rate": 1.357635481854691e-08,
+      "loss": 0.9842,
+      "step": 32460
+    },
+    {
+      "epoch": 0.9780120481927711,
+      "grad_norm": 6.26476584244363,
+      "learning_rate": 1.3212066146934267e-08,
+      "loss": 0.9102,
+      "step": 32470
+    },
+    {
+      "epoch": 0.9783132530120482,
+      "grad_norm": 3.957553293059344,
+      "learning_rate": 1.2852725105519382e-08,
+      "loss": 0.9382,
+      "step": 32480
+    },
+    {
+      "epoch": 0.9786144578313253,
+      "grad_norm": 2.018288924985573,
+      "learning_rate": 1.2498332050820406e-08,
+      "loss": 0.9859,
+      "step": 32490
+    },
+    {
+      "epoch": 0.9789156626506024,
+      "grad_norm": 5.731505873250158,
+      "learning_rate": 1.214888733444719e-08,
+      "loss": 0.9666,
+      "step": 32500
+    },
+    {
+      "epoch": 0.9792168674698796,
+      "grad_norm": 4.205672805287298,
+      "learning_rate": 1.1804391303101292e-08,
+      "loss": 0.9027,
+      "step": 32510
+    },
+    {
+      "epoch": 0.9795180722891567,
+      "grad_norm": 1.9724752924746713,
+      "learning_rate": 1.1464844298573752e-08,
+      "loss": 0.8686,
+      "step": 32520
+    },
+    {
+      "epoch": 0.9798192771084338,
+      "grad_norm": 2.04575420580311,
+      "learning_rate": 1.1130246657745092e-08,
+      "loss": 0.7719,
+      "step": 32530
+    },
+    {
+      "epoch": 0.9801204819277108,
+      "grad_norm": 6.5790252772524465,
+      "learning_rate": 1.080059871258532e-08,
+      "loss": 0.8899,
+      "step": 32540
+    },
+    {
+      "epoch": 0.9804216867469879,
+      "grad_norm": 1.9387316441227798,
+      "learning_rate": 1.047590079015448e-08,
+      "loss": 0.8163,
+      "step": 32550
+    },
+    {
+      "epoch": 0.980722891566265,
+      "grad_norm": 1.9019712434002312,
+      "learning_rate": 1.015615321260044e-08,
+      "loss": 0.9109,
+      "step": 32560
+    },
+    {
+      "epoch": 0.9810240963855422,
+      "grad_norm": 5.366533735264594,
+      "learning_rate": 9.841356297159988e-09,
+      "loss": 0.8855,
+      "step": 32570
+    },
+    {
+      "epoch": 0.9813253012048193,
+      "grad_norm": 6.614995026563534,
+      "learning_rate": 9.531510356158291e-09,
+      "loss": 0.8818,
+      "step": 32580
+    },
+    {
+      "epoch": 0.9816265060240964,
+      "grad_norm": 5.51945953883837,
+      "learning_rate": 9.226615697008335e-09,
+      "loss": 0.9751,
+      "step": 32590
+    },
+    {
+      "epoch": 0.9819277108433735,
+      "grad_norm": 6.390542965586899,
+      "learning_rate": 8.926672622210365e-09,
+      "loss": 0.9374,
+      "step": 32600
+    },
+    {
+      "epoch": 0.9822289156626506,
+      "grad_norm": 7.353054114327854,
+      "learning_rate": 8.631681429353e-09,
+      "loss": 0.8522,
+      "step": 32610
+    },
+    {
+      "epoch": 0.9825301204819277,
+      "grad_norm": 16.68028728382817,
+      "learning_rate": 8.341642411109907e-09,
+      "loss": 0.878,
+      "step": 32620
+    },
+    {
+      "epoch": 0.9828313253012049,
+      "grad_norm": 7.066433488752781,
+      "learning_rate": 8.056555855243675e-09,
+      "loss": 0.9412,
+      "step": 32630
+    },
+    {
+      "epoch": 0.983132530120482,
+      "grad_norm": 4.135364704257834,
+      "learning_rate": 7.77642204460194e-09,
+      "loss": 0.9071,
+      "step": 32640
+    },
+    {
+      "epoch": 0.983433734939759,
+      "grad_norm": 2.1166164025756977,
+      "learning_rate": 7.501241257118485e-09,
+      "loss": 0.9534,
+      "step": 32650
+    },
+    {
+      "epoch": 0.9837349397590361,
+      "grad_norm": 1.8835170239698753,
+      "learning_rate": 7.2310137658138104e-09,
+      "loss": 0.8746,
+      "step": 32660
+    },
+    {
+      "epoch": 0.9840361445783132,
+      "grad_norm": 5.390526178714944,
+      "learning_rate": 6.9657398387928954e-09,
+      "loss": 0.7723,
+      "step": 32670
+    },
+    {
+      "epoch": 0.9843373493975903,
+      "grad_norm": 4.587666920735826,
+      "learning_rate": 6.705419739246876e-09,
+      "loss": 0.9205,
+      "step": 32680
+    },
+    {
+      "epoch": 0.9846385542168675,
+      "grad_norm": 2.038842973204118,
+      "learning_rate": 6.450053725452488e-09,
+      "loss": 0.8732,
+      "step": 32690
+    },
+    {
+      "epoch": 0.9849397590361446,
+      "grad_norm": 1.9878965511900648,
+      "learning_rate": 6.199642050769283e-09,
+      "loss": 0.9325,
+      "step": 32700
+    },
+    {
+      "epoch": 0.9852409638554217,
+      "grad_norm": 17.11659900685862,
+      "learning_rate": 5.954184963644083e-09,
+      "loss": 0.7917,
+      "step": 32710
+    },
+    {
+      "epoch": 0.9855421686746988,
+      "grad_norm": 8.021908437912327,
+      "learning_rate": 5.713682707604862e-09,
+      "loss": 0.9266,
+      "step": 32720
+    },
+    {
+      "epoch": 0.9858433734939759,
+      "grad_norm": 6.410494503733885,
+      "learning_rate": 5.478135521266859e-09,
+      "loss": 1.022,
+      "step": 32730
+    },
+    {
+      "epoch": 0.986144578313253,
+      "grad_norm": 2.0533458036334253,
+      "learning_rate": 5.247543638327024e-09,
+      "loss": 0.8769,
+      "step": 32740
+    },
+    {
+      "epoch": 0.9864457831325302,
+      "grad_norm": 6.70943462401911,
+      "learning_rate": 5.021907287566241e-09,
+      "loss": 0.9252,
+      "step": 32750
+    },
+    {
+      "epoch": 0.9867469879518073,
+      "grad_norm": 4.3351846787495845,
+      "learning_rate": 4.801226692849881e-09,
+      "loss": 0.9755,
+      "step": 32760
+    },
+    {
+      "epoch": 0.9870481927710844,
+      "grad_norm": 5.899850504536801,
+      "learning_rate": 4.585502073125026e-09,
+      "loss": 0.8919,
+      "step": 32770
+    },
+    {
+      "epoch": 0.9873493975903614,
+      "grad_norm": 4.026783699486491,
+      "learning_rate": 4.37473364242269e-09,
+      "loss": 0.8622,
+      "step": 32780
+    },
+    {
+      "epoch": 0.9876506024096385,
+      "grad_norm": 2.059837098189508,
+      "learning_rate": 4.168921609856158e-09,
+      "loss": 0.922,
+      "step": 32790
+    },
+    {
+      "epoch": 0.9879518072289156,
+      "grad_norm": 5.903112831821014,
+      "learning_rate": 3.968066179621533e-09,
+      "loss": 0.9151,
+      "step": 32800
+    },
+    {
+      "epoch": 0.9882530120481928,
+      "grad_norm": 6.380133537258821,
+      "learning_rate": 3.7721675509971855e-09,
+      "loss": 0.9358,
+      "step": 32810
+    },
+    {
+      "epoch": 0.9885542168674699,
+      "grad_norm": 4.691649478423261,
+      "learning_rate": 3.5812259183426457e-09,
+      "loss": 0.8861,
+      "step": 32820
+    },
+    {
+      "epoch": 0.988855421686747,
+      "grad_norm": 2.020453275973914,
+      "learning_rate": 3.3952414711013737e-09,
+      "loss": 0.9482,
+      "step": 32830
+    },
+    {
+      "epoch": 0.9891566265060241,
+      "grad_norm": 4.887013192306234,
+      "learning_rate": 3.2142143937968774e-09,
+      "loss": 1.0159,
+      "step": 32840
+    },
+    {
+      "epoch": 0.9894578313253012,
+      "grad_norm": 4.478210861833627,
+      "learning_rate": 3.0381448660343758e-09,
+      "loss": 0.9274,
+      "step": 32850
+    },
+    {
+      "epoch": 0.9897590361445783,
+      "grad_norm": 5.917669075335433,
+      "learning_rate": 2.8670330625019117e-09,
+      "loss": 0.7962,
+      "step": 32860
+    },
+    {
+      "epoch": 0.9900602409638555,
+      "grad_norm": 6.769969376664644,
+      "learning_rate": 2.7008791529664624e-09,
+      "loss": 0.9538,
+      "step": 32870
+    },
+    {
+      "epoch": 0.9903614457831326,
+      "grad_norm": 5.406044559926571,
+      "learning_rate": 2.539683302277829e-09,
+      "loss": 0.9612,
+      "step": 32880
+    },
+    {
+      "epoch": 0.9906626506024097,
+      "grad_norm": 1.8584573720910964,
+      "learning_rate": 2.3834456703658583e-09,
+      "loss": 0.8362,
+      "step": 32890
+    },
+    {
+      "epoch": 0.9909638554216867,
+      "grad_norm": 8.331550664394987,
+      "learning_rate": 2.232166412241554e-09,
+      "loss": 0.8863,
+      "step": 32900
+    },
+    {
+      "epoch": 0.9912650602409638,
+      "grad_norm": 4.381243677666166,
+      "learning_rate": 2.085845677996523e-09,
+      "loss": 0.9229,
+      "step": 32910
+    },
+    {
+      "epoch": 0.9915662650602409,
+      "grad_norm": 7.120700168343828,
+      "learning_rate": 1.944483612802417e-09,
+      "loss": 0.9992,
+      "step": 32920
+    },
+    {
+      "epoch": 0.9918674698795181,
+      "grad_norm": 2.1293013436985726,
+      "learning_rate": 1.8080803569109351e-09,
+      "loss": 0.8772,
+      "step": 32930
+    },
+    {
+      "epoch": 0.9921686746987952,
+      "grad_norm": 2.0645965470437555,
+      "learning_rate": 1.6766360456549335e-09,
+      "loss": 0.9249,
+      "step": 32940
+    },
+    {
+      "epoch": 0.9924698795180723,
+      "grad_norm": 2.0943255183398923,
+      "learning_rate": 1.5501508094456496e-09,
+      "loss": 0.8664,
+      "step": 32950
+    },
+    {
+      "epoch": 0.9927710843373494,
+      "grad_norm": 5.175432571590235,
+      "learning_rate": 1.4286247737754776e-09,
+      "loss": 0.8166,
+      "step": 32960
+    },
+    {
+      "epoch": 0.9930722891566265,
+      "grad_norm": 6.522363733828929,
+      "learning_rate": 1.3120580592168586e-09,
+      "loss": 0.897,
+      "step": 32970
+    },
+    {
+      "epoch": 0.9933734939759036,
+      "grad_norm": 1.9844911119178683,
+      "learning_rate": 1.2004507814200595e-09,
+      "loss": 0.945,
+      "step": 32980
+    },
+    {
+      "epoch": 0.9936746987951808,
+      "grad_norm": 5.505876329498268,
+      "learning_rate": 1.0938030511170595e-09,
+      "loss": 0.8917,
+      "step": 32990
+    },
+    {
+      "epoch": 0.9939759036144579,
+      "grad_norm": 24.282183839792587,
+      "learning_rate": 9.92114974117664e-10,
+      "loss": 0.9364,
+      "step": 33000
+    },
+    {
+      "epoch": 0.994277108433735,
+      "grad_norm": 9.23266439400529,
+      "learning_rate": 8.953866513111698e-10,
+      "loss": 0.9321,
+      "step": 33010
+    },
+    {
+      "epoch": 0.994578313253012,
+      "grad_norm": 4.571343950740616,
+      "learning_rate": 8.036181786669206e-10,
+      "loss": 0.916,
+      "step": 33020
+    },
+    {
+      "epoch": 0.9948795180722891,
+      "grad_norm": 9.543958338609684,
+      "learning_rate": 7.168096472326413e-10,
+      "loss": 0.9559,
+      "step": 33030
+    },
+    {
+      "epoch": 0.9951807228915662,
+      "grad_norm": 3.926562177985658,
+      "learning_rate": 6.349611431349934e-10,
+      "loss": 0.9291,
+      "step": 33040
+    },
+    {
+      "epoch": 0.9954819277108434,
+      "grad_norm": 5.010326811827543,
+      "learning_rate": 5.580727475795744e-10,
+      "loss": 1.0484,
+      "step": 33050
+    },
+    {
+      "epoch": 0.9957831325301205,
+      "grad_norm": 7.221217099561627,
+      "learning_rate": 4.861445368514739e-10,
+      "loss": 0.9055,
+      "step": 33060
+    },
+    {
+      "epoch": 0.9960843373493976,
+      "grad_norm": 5.002598418152769,
+      "learning_rate": 4.1917658231416247e-10,
+      "loss": 0.9872,
+      "step": 33070
+    },
+    {
+      "epoch": 0.9963855421686747,
+      "grad_norm": 6.469774961883145,
+      "learning_rate": 3.571689504100473e-10,
+      "loss": 0.8994,
+      "step": 33080
+    },
+    {
+      "epoch": 0.9966867469879518,
+      "grad_norm": 2.218287374798166,
+      "learning_rate": 3.0012170265880656e-10,
+      "loss": 0.8879,
+      "step": 33090
+    },
+    {
+      "epoch": 0.9969879518072289,
+      "grad_norm": 3.8477052272546644,
+      "learning_rate": 2.4803489566016526e-10,
+      "loss": 0.8326,
+      "step": 33100
+    },
+    {
+      "epoch": 0.9972891566265061,
+      "grad_norm": 27.923816696036397,
+      "learning_rate": 2.0090858109278465e-10,
+      "loss": 0.9737,
+      "step": 33110
+    },
+    {
+      "epoch": 0.9975903614457832,
+      "grad_norm": 4.693803475151824,
+      "learning_rate": 1.5874280571148704e-10,
+      "loss": 0.9478,
+      "step": 33120
+    },
+    {
+      "epoch": 0.9978915662650603,
+      "grad_norm": 4.523075656820095,
+      "learning_rate": 1.215376113522515e-10,
+      "loss": 1.0148,
+      "step": 33130
+    },
+    {
+      "epoch": 0.9981927710843373,
+      "grad_norm": 17.090115553209188,
+      "learning_rate": 8.929303492721808e-11,
+      "loss": 0.8852,
+      "step": 33140
+    },
+    {
+      "epoch": 0.9984939759036144,
+      "grad_norm": 8.636003886826249,
+      "learning_rate": 6.200910842801833e-11,
+      "loss": 0.8452,
+      "step": 33150
+    },
+    {
+      "epoch": 0.9987951807228915,
+      "grad_norm": 4.263752612525351,
+      "learning_rate": 3.968585892466514e-11,
+      "loss": 0.7981,
+      "step": 33160
+    },
+    {
+      "epoch": 0.9990963855421687,
+      "grad_norm": 5.731074372657009,
+      "learning_rate": 2.232330856499765e-11,
+      "loss": 0.9671,
+      "step": 33170
+    },
+    {
+      "epoch": 0.9993975903614458,
+      "grad_norm": 2.1559076434496554,
+      "learning_rate": 9.921474574681179e-12,
+      "loss": 0.9056,
+      "step": 33180
+    },
+    {
+      "epoch": 0.9996987951807229,
+      "grad_norm": 6.4165937373810955,
+      "learning_rate": 2.480369258872628e-12,
+      "loss": 0.9025,
+      "step": 33190
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.918614350791673,
+      "learning_rate": 0.0,
+      "loss": 1.0485,
+      "step": 33200
+    },
+    {
+      "epoch": 1.0,
+      "step": 33200,
+      "total_flos": 1.6067564005752832e+16,
+      "train_loss": 0.0,
+      "train_runtime": 5.1234,
+      "train_samples_per_second": 829462.23,
+      "train_steps_per_second": 6480.076
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 33200,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 3320,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.6067564005752832e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}