diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,53815 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 23049,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003904724716907458,
+      "grad_norm": 22.37471580505371,
+      "learning_rate": 6.000000000000001e-08,
+      "loss": 1.5848,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007809449433814916,
+      "grad_norm": 24.825096130371094,
+      "learning_rate": 1.2000000000000002e-07,
+      "loss": 1.4111,
+      "step": 6
+    },
+    {
+      "epoch": 0.0011714174150722373,
+      "grad_norm": 22.30377769470215,
+      "learning_rate": 1.8e-07,
+      "loss": 1.4883,
+      "step": 9
+    },
+    {
+      "epoch": 0.0015618898867629833,
+      "grad_norm": 22.412364959716797,
+      "learning_rate": 2.4000000000000003e-07,
+      "loss": 1.5305,
+      "step": 12
+    },
+    {
+      "epoch": 0.001952362358453729,
+      "grad_norm": 22.382173538208008,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 1.4476,
+      "step": 15
+    },
+    {
+      "epoch": 0.0023428348301444747,
+      "grad_norm": 18.971206665039062,
+      "learning_rate": 3.6e-07,
+      "loss": 1.3828,
+      "step": 18
+    },
+    {
+      "epoch": 0.002733307301835221,
+      "grad_norm": 17.42239761352539,
+      "learning_rate": 4.2000000000000006e-07,
+      "loss": 1.4491,
+      "step": 21
+    },
+    {
+      "epoch": 0.0031237797735259665,
+      "grad_norm": 16.123905181884766,
+      "learning_rate": 4.800000000000001e-07,
+      "loss": 1.418,
+      "step": 24
+    },
+    {
+      "epoch": 0.003514252245216712,
+      "grad_norm": 15.892184257507324,
+      "learning_rate": 5.4e-07,
+      "loss": 1.4415,
+      "step": 27
+    },
+    {
+      "epoch": 0.003904724716907458,
+      "grad_norm": 12.987159729003906,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.3907,
+      "step": 30
+    },
+    {
+      "epoch": 0.004295197188598204,
+      "grad_norm": 11.818194389343262,
+      "learning_rate": 6.6e-07,
+      "loss": 1.421,
+      "step": 33
+    },
+    {
+      "epoch": 0.004685669660288949,
+      "grad_norm": 11.758996963500977,
+      "learning_rate": 7.2e-07,
+      "loss": 1.2863,
+      "step": 36
+    },
+    {
+      "epoch": 0.005076142131979695,
+      "grad_norm": 11.128711700439453,
+      "learning_rate": 7.8e-07,
+      "loss": 1.2689,
+      "step": 39
+    },
+    {
+      "epoch": 0.005466614603670442,
+      "grad_norm": 10.793634414672852,
+      "learning_rate": 8.400000000000001e-07,
+      "loss": 1.3767,
+      "step": 42
+    },
+    {
+      "epoch": 0.005857087075361187,
+      "grad_norm": 10.585275650024414,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 1.2601,
+      "step": 45
+    },
+    {
+      "epoch": 0.006247559547051933,
+      "grad_norm": 9.969756126403809,
+      "learning_rate": 9.600000000000001e-07,
+      "loss": 1.1707,
+      "step": 48
+    },
+    {
+      "epoch": 0.006638032018742679,
+      "grad_norm": 9.342447280883789,
+      "learning_rate": 1.02e-06,
+      "loss": 1.1647,
+      "step": 51
+    },
+    {
+      "epoch": 0.007028504490433424,
+      "grad_norm": 9.255729675292969,
+      "learning_rate": 1.08e-06,
+      "loss": 1.1294,
+      "step": 54
+    },
+    {
+      "epoch": 0.00741897696212417,
+      "grad_norm": 8.989304542541504,
+      "learning_rate": 1.14e-06,
+      "loss": 1.1128,
+      "step": 57
+    },
+    {
+      "epoch": 0.007809449433814916,
+      "grad_norm": 8.631978034973145,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.2147,
+      "step": 60
+    },
+    {
+      "epoch": 0.008199921905505662,
+      "grad_norm": 9.135863304138184,
+      "learning_rate": 1.26e-06,
+      "loss": 1.052,
+      "step": 63
+    },
+    {
+      "epoch": 0.008590394377196407,
+      "grad_norm": 7.280301570892334,
+      "learning_rate": 1.32e-06,
+      "loss": 0.9579,
+      "step": 66
+    },
+    {
+      "epoch": 0.008980866848887154,
+      "grad_norm": 6.385473728179932,
+      "learning_rate": 1.3800000000000001e-06,
+      "loss": 1.1031,
+      "step": 69
+    },
+    {
+      "epoch": 0.009371339320577899,
+      "grad_norm": 4.986073017120361,
+      "learning_rate": 1.44e-06,
+      "loss": 0.9746,
+      "step": 72
+    },
+    {
+      "epoch": 0.009761811792268645,
+      "grad_norm": 4.2078423500061035,
+      "learning_rate": 1.5e-06,
+      "loss": 1.0148,
+      "step": 75
+    },
+    {
+      "epoch": 0.01015228426395939,
+      "grad_norm": 4.854979038238525,
+      "learning_rate": 1.56e-06,
+      "loss": 1.005,
+      "step": 78
+    },
+    {
+      "epoch": 0.010542756735650137,
+      "grad_norm": 3.9257068634033203,
+      "learning_rate": 1.6200000000000002e-06,
+      "loss": 0.8129,
+      "step": 81
+    },
+    {
+      "epoch": 0.010933229207340883,
+      "grad_norm": 4.077581882476807,
+      "learning_rate": 1.6800000000000002e-06,
+      "loss": 1.0132,
+      "step": 84
+    },
+    {
+      "epoch": 0.011323701679031628,
+      "grad_norm": 4.484386444091797,
+      "learning_rate": 1.74e-06,
+      "loss": 0.9022,
+      "step": 87
+    },
+    {
+      "epoch": 0.011714174150722375,
+      "grad_norm": 3.997361898422241,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.8801,
+      "step": 90
+    },
+    {
+      "epoch": 0.01210464662241312,
+      "grad_norm": 3.9567580223083496,
+      "learning_rate": 1.8600000000000002e-06,
+      "loss": 0.9124,
+      "step": 93
+    },
+    {
+      "epoch": 0.012495119094103866,
+      "grad_norm": 4.3611297607421875,
+      "learning_rate": 1.9200000000000003e-06,
+      "loss": 0.939,
+      "step": 96
+    },
+    {
+      "epoch": 0.012885591565794611,
+      "grad_norm": 7.97072172164917,
+      "learning_rate": 1.98e-06,
+      "loss": 0.9517,
+      "step": 99
+    },
+    {
+      "epoch": 0.013276064037485357,
+      "grad_norm": 3.360591173171997,
+      "learning_rate": 2.04e-06,
+      "loss": 0.7842,
+      "step": 102
+    },
+    {
+      "epoch": 0.013666536509176102,
+      "grad_norm": 8.188989639282227,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.9067,
+      "step": 105
+    },
+    {
+      "epoch": 0.014057008980866849,
+      "grad_norm": 5.126707077026367,
+      "learning_rate": 2.16e-06,
+      "loss": 0.8402,
+      "step": 108
+    },
+    {
+      "epoch": 0.014447481452557595,
+      "grad_norm": 3.1572999954223633,
+      "learning_rate": 2.2200000000000003e-06,
+      "loss": 0.7993,
+      "step": 111
+    },
+    {
+      "epoch": 0.01483795392424834,
+      "grad_norm": 3.5879690647125244,
+      "learning_rate": 2.28e-06,
+      "loss": 0.8593,
+      "step": 114
+    },
+    {
+      "epoch": 0.015228426395939087,
+      "grad_norm": 3.5023837089538574,
+      "learning_rate": 2.3400000000000005e-06,
+      "loss": 0.8434,
+      "step": 117
+    },
+    {
+      "epoch": 0.015618898867629832,
+      "grad_norm": 3.2916157245635986,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.8904,
+      "step": 120
+    },
+    {
+      "epoch": 0.016009371339320577,
+      "grad_norm": 4.094881534576416,
+      "learning_rate": 2.46e-06,
+      "loss": 0.8106,
+      "step": 123
+    },
+    {
+      "epoch": 0.016399843811011325,
+      "grad_norm": 7.149028301239014,
+      "learning_rate": 2.52e-06,
+      "loss": 0.8551,
+      "step": 126
+    },
+    {
+      "epoch": 0.01679031628270207,
+      "grad_norm": 3.3029215335845947,
+      "learning_rate": 2.5800000000000003e-06,
+      "loss": 0.8387,
+      "step": 129
+    },
+    {
+      "epoch": 0.017180788754392814,
+      "grad_norm": 3.166794538497925,
+      "learning_rate": 2.64e-06,
+      "loss": 0.819,
+      "step": 132
+    },
+    {
+      "epoch": 0.017571261226083563,
+      "grad_norm": 4.197257995605469,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.7449,
+      "step": 135
+    },
+    {
+      "epoch": 0.017961733697774308,
+      "grad_norm": 3.171112060546875,
+      "learning_rate": 2.7600000000000003e-06,
+      "loss": 0.7152,
+      "step": 138
+    },
+    {
+      "epoch": 0.018352206169465052,
+      "grad_norm": 3.2386510372161865,
+      "learning_rate": 2.82e-06,
+      "loss": 0.7647,
+      "step": 141
+    },
+    {
+      "epoch": 0.018742678641155797,
+      "grad_norm": 3.2060446739196777,
+      "learning_rate": 2.88e-06,
+      "loss": 0.737,
+      "step": 144
+    },
+    {
+      "epoch": 0.019133151112846546,
+      "grad_norm": 3.846658706665039,
+      "learning_rate": 2.9400000000000002e-06,
+      "loss": 0.7577,
+      "step": 147
+    },
+    {
+      "epoch": 0.01952362358453729,
+      "grad_norm": 3.3511500358581543,
+      "learning_rate": 3e-06,
+      "loss": 0.7303,
+      "step": 150
+    },
+    {
+      "epoch": 0.019914096056228035,
+      "grad_norm": 3.5826847553253174,
+      "learning_rate": 3.0600000000000003e-06,
+      "loss": 0.8347,
+      "step": 153
+    },
+    {
+      "epoch": 0.02030456852791878,
+      "grad_norm": 3.3204636573791504,
+      "learning_rate": 3.12e-06,
+      "loss": 0.7353,
+      "step": 156
+    },
+    {
+      "epoch": 0.02069504099960953,
+      "grad_norm": 3.548901319503784,
+      "learning_rate": 3.1800000000000005e-06,
+      "loss": 0.8361,
+      "step": 159
+    },
+    {
+      "epoch": 0.021085513471300273,
+      "grad_norm": 3.251668691635132,
+      "learning_rate": 3.2400000000000003e-06,
+      "loss": 0.7772,
+      "step": 162
+    },
+    {
+      "epoch": 0.021475985942991018,
+      "grad_norm": 3.0531423091888428,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.724,
+      "step": 165
+    },
+    {
+      "epoch": 0.021866458414681766,
+      "grad_norm": 3.8247272968292236,
+      "learning_rate": 3.3600000000000004e-06,
+      "loss": 0.6697,
+      "step": 168
+    },
+    {
+      "epoch": 0.02225693088637251,
+      "grad_norm": 6.110504627227783,
+      "learning_rate": 3.4200000000000007e-06,
+      "loss": 0.8024,
+      "step": 171
+    },
+    {
+      "epoch": 0.022647403358063256,
+      "grad_norm": 3.497457504272461,
+      "learning_rate": 3.48e-06,
+      "loss": 0.7605,
+      "step": 174
+    },
+    {
+      "epoch": 0.023037875829754,
+      "grad_norm": 3.874572515487671,
+      "learning_rate": 3.54e-06,
+      "loss": 0.8033,
+      "step": 177
+    },
+    {
+      "epoch": 0.02342834830144475,
+      "grad_norm": 3.286510467529297,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.7136,
+      "step": 180
+    },
+    {
+      "epoch": 0.023818820773135494,
+      "grad_norm": 3.194321870803833,
+      "learning_rate": 3.66e-06,
+      "loss": 0.804,
+      "step": 183
+    },
+    {
+      "epoch": 0.02420929324482624,
+      "grad_norm": 3.3385202884674072,
+      "learning_rate": 3.7200000000000004e-06,
+      "loss": 0.7735,
+      "step": 186
+    },
+    {
+      "epoch": 0.024599765716516987,
+      "grad_norm": 3.5652225017547607,
+      "learning_rate": 3.7800000000000002e-06,
+      "loss": 0.8019,
+      "step": 189
+    },
+    {
+      "epoch": 0.024990238188207732,
+      "grad_norm": 3.887373447418213,
+      "learning_rate": 3.8400000000000005e-06,
+      "loss": 0.732,
+      "step": 192
+    },
+    {
+      "epoch": 0.025380710659898477,
+      "grad_norm": 3.134536027908325,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.7466,
+      "step": 195
+    },
+    {
+      "epoch": 0.025771183131589222,
+      "grad_norm": 3.26898455619812,
+      "learning_rate": 3.96e-06,
+      "loss": 0.7262,
+      "step": 198
+    },
+    {
+      "epoch": 0.02616165560327997,
+      "grad_norm": 3.5477030277252197,
+      "learning_rate": 4.0200000000000005e-06,
+      "loss": 0.8538,
+      "step": 201
+    },
+    {
+      "epoch": 0.026552128074970715,
+      "grad_norm": 3.780203104019165,
+      "learning_rate": 4.08e-06,
+      "loss": 0.6764,
+      "step": 204
+    },
+    {
+      "epoch": 0.02694260054666146,
+      "grad_norm": 3.250094413757324,
+      "learning_rate": 4.14e-06,
+      "loss": 0.8611,
+      "step": 207
+    },
+    {
+      "epoch": 0.027333073018352205,
+      "grad_norm": 5.036427021026611,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.8276,
+      "step": 210
+    },
+    {
+      "epoch": 0.027723545490042953,
+      "grad_norm": 2.8822755813598633,
+      "learning_rate": 4.26e-06,
+      "loss": 0.8168,
+      "step": 213
+    },
+    {
+      "epoch": 0.028114017961733698,
+      "grad_norm": 3.374403953552246,
+      "learning_rate": 4.32e-06,
+      "loss": 0.8104,
+      "step": 216
+    },
+    {
+      "epoch": 0.028504490433424443,
+      "grad_norm": 4.382564544677734,
+      "learning_rate": 4.38e-06,
+      "loss": 0.769,
+      "step": 219
+    },
+    {
+      "epoch": 0.02889496290511519,
+      "grad_norm": 4.98586893081665,
+      "learning_rate": 4.440000000000001e-06,
+      "loss": 0.8022,
+      "step": 222
+    },
+    {
+      "epoch": 0.029285435376805936,
+      "grad_norm": 3.1688528060913086,
+      "learning_rate": 4.5e-06,
+      "loss": 0.7269,
+      "step": 225
+    },
+    {
+      "epoch": 0.02967590784849668,
+      "grad_norm": 3.384666681289673,
+      "learning_rate": 4.56e-06,
+      "loss": 0.7402,
+      "step": 228
+    },
+    {
+      "epoch": 0.030066380320187425,
+      "grad_norm": 3.3539681434631348,
+      "learning_rate": 4.620000000000001e-06,
+      "loss": 0.8786,
+      "step": 231
+    },
+    {
+      "epoch": 0.030456852791878174,
+      "grad_norm": 3.280231237411499,
+      "learning_rate": 4.680000000000001e-06,
+      "loss": 0.6507,
+      "step": 234
+    },
+    {
+      "epoch": 0.03084732526356892,
+      "grad_norm": 2.9844179153442383,
+      "learning_rate": 4.74e-06,
+      "loss": 0.7314,
+      "step": 237
+    },
+    {
+      "epoch": 0.031237797735259663,
+      "grad_norm": 5.182901382446289,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.7851,
+      "step": 240
+    },
+    {
+      "epoch": 0.03162827020695041,
+      "grad_norm": 4.9192023277282715,
+      "learning_rate": 4.86e-06,
+      "loss": 0.7062,
+      "step": 243
+    },
+    {
+      "epoch": 0.03201874267864115,
+      "grad_norm": 3.2586448192596436,
+      "learning_rate": 4.92e-06,
+      "loss": 0.778,
+      "step": 246
+    },
+    {
+      "epoch": 0.032409215150331905,
+      "grad_norm": 3.3434247970581055,
+      "learning_rate": 4.980000000000001e-06,
+      "loss": 0.7214,
+      "step": 249
+    },
+    {
+      "epoch": 0.03279968762202265,
+      "grad_norm": 3.2206249237060547,
+      "learning_rate": 5.04e-06,
+      "loss": 0.75,
+      "step": 252
+    },
+    {
+      "epoch": 0.033190160093713394,
+      "grad_norm": 4.057199954986572,
+      "learning_rate": 5.1e-06,
+      "loss": 0.8155,
+      "step": 255
+    },
+    {
+      "epoch": 0.03358063256540414,
+      "grad_norm": 3.0813217163085938,
+      "learning_rate": 5.1600000000000006e-06,
+      "loss": 0.7603,
+      "step": 258
+    },
+    {
+      "epoch": 0.033971105037094884,
+      "grad_norm": 3.3523905277252197,
+      "learning_rate": 5.220000000000001e-06,
+      "loss": 0.738,
+      "step": 261
+    },
+    {
+      "epoch": 0.03436157750878563,
+      "grad_norm": 5.289391994476318,
+      "learning_rate": 5.28e-06,
+      "loss": 0.6952,
+      "step": 264
+    },
+    {
+      "epoch": 0.034752049980476374,
+      "grad_norm": 3.419605016708374,
+      "learning_rate": 5.3400000000000005e-06,
+      "loss": 0.7415,
+      "step": 267
+    },
+    {
+      "epoch": 0.035142522452167126,
+      "grad_norm": 2.9998910427093506,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 0.813,
+      "step": 270
+    },
+    {
+      "epoch": 0.03553299492385787,
+      "grad_norm": 3.2090444564819336,
+      "learning_rate": 5.460000000000001e-06,
+      "loss": 0.7232,
+      "step": 273
+    },
+    {
+      "epoch": 0.035923467395548615,
+      "grad_norm": 4.806464672088623,
+      "learning_rate": 5.5200000000000005e-06,
+      "loss": 0.865,
+      "step": 276
+    },
+    {
+      "epoch": 0.03631393986723936,
+      "grad_norm": 3.2608954906463623,
+      "learning_rate": 5.580000000000001e-06,
+      "loss": 0.7011,
+      "step": 279
+    },
+    {
+      "epoch": 0.036704412338930105,
+      "grad_norm": 3.387782096862793,
+      "learning_rate": 5.64e-06,
+      "loss": 0.7094,
+      "step": 282
+    },
+    {
+      "epoch": 0.03709488481062085,
+      "grad_norm": 3.7122697830200195,
+      "learning_rate": 5.7e-06,
+      "loss": 0.7027,
+      "step": 285
+    },
+    {
+      "epoch": 0.037485357282311595,
+      "grad_norm": 2.901911497116089,
+      "learning_rate": 5.76e-06,
+      "loss": 0.6686,
+      "step": 288
+    },
+    {
+      "epoch": 0.03787582975400234,
+      "grad_norm": 4.1798996925354,
+      "learning_rate": 5.82e-06,
+      "loss": 0.7313,
+      "step": 291
+    },
+    {
+      "epoch": 0.03826630222569309,
+      "grad_norm": 3.4261515140533447,
+      "learning_rate": 5.8800000000000005e-06,
+      "loss": 0.7415,
+      "step": 294
+    },
+    {
+      "epoch": 0.038656774697383836,
+      "grad_norm": 3.0005760192871094,
+      "learning_rate": 5.94e-06,
+      "loss": 0.7753,
+      "step": 297
+    },
+    {
+      "epoch": 0.03904724716907458,
+      "grad_norm": 3.1258742809295654,
+      "learning_rate": 6e-06,
+      "loss": 0.6959,
+      "step": 300
+    },
+    {
+      "epoch": 0.039437719640765326,
+      "grad_norm": 4.299694538116455,
+      "learning_rate": 6.0600000000000004e-06,
+      "loss": 0.6857,
+      "step": 303
+    },
+    {
+      "epoch": 0.03982819211245607,
+      "grad_norm": 3.4530158042907715,
+      "learning_rate": 6.120000000000001e-06,
+      "loss": 0.7072,
+      "step": 306
+    },
+    {
+      "epoch": 0.040218664584146815,
+      "grad_norm": 3.939462900161743,
+      "learning_rate": 6.18e-06,
+      "loss": 0.7783,
+      "step": 309
+    },
+    {
+      "epoch": 0.04060913705583756,
+      "grad_norm": 3.1602437496185303,
+      "learning_rate": 6.24e-06,
+      "loss": 0.7953,
+      "step": 312
+    },
+    {
+      "epoch": 0.04099960952752831,
+      "grad_norm": 2.8542044162750244,
+      "learning_rate": 6.300000000000001e-06,
+      "loss": 0.779,
+      "step": 315
+    },
+    {
+      "epoch": 0.04139008199921906,
+      "grad_norm": 3.5136380195617676,
+      "learning_rate": 6.360000000000001e-06,
+      "loss": 0.755,
+      "step": 318
+    },
+    {
+      "epoch": 0.0417805544709098,
+      "grad_norm": 3.060673713684082,
+      "learning_rate": 6.42e-06,
+      "loss": 0.7327,
+      "step": 321
+    },
+    {
+      "epoch": 0.04217102694260055,
+      "grad_norm": 3.174912691116333,
+      "learning_rate": 6.480000000000001e-06,
+      "loss": 0.7553,
+      "step": 324
+    },
+    {
+      "epoch": 0.04256149941429129,
+      "grad_norm": 3.189807653427124,
+      "learning_rate": 6.540000000000001e-06,
+      "loss": 0.815,
+      "step": 327
+    },
+    {
+      "epoch": 0.042951971885982036,
+      "grad_norm": 2.7848079204559326,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 0.6674,
+      "step": 330
+    },
+    {
+      "epoch": 0.04334244435767278,
+      "grad_norm": 3.377772331237793,
+      "learning_rate": 6.660000000000001e-06,
+      "loss": 0.6562,
+      "step": 333
+    },
+    {
+      "epoch": 0.04373291682936353,
+      "grad_norm": 3.255612373352051,
+      "learning_rate": 6.720000000000001e-06,
+      "loss": 0.71,
+      "step": 336
+    },
+    {
+      "epoch": 0.04412338930105428,
+      "grad_norm": 3.997131824493408,
+      "learning_rate": 6.780000000000001e-06,
+      "loss": 0.7364,
+      "step": 339
+    },
+    {
+      "epoch": 0.04451386177274502,
+      "grad_norm": 3.373277425765991,
+      "learning_rate": 6.8400000000000014e-06,
+      "loss": 0.7041,
+      "step": 342
+    },
+    {
+      "epoch": 0.04490433424443577,
+      "grad_norm": 3.901888608932495,
+      "learning_rate": 6.9e-06,
+      "loss": 0.6684,
+      "step": 345
+    },
+    {
+      "epoch": 0.04529480671612651,
+      "grad_norm": 3.7340526580810547,
+      "learning_rate": 6.96e-06,
+      "loss": 0.7079,
+      "step": 348
+    },
+    {
+      "epoch": 0.04568527918781726,
+      "grad_norm": 3.2724201679229736,
+      "learning_rate": 7.0200000000000006e-06,
+      "loss": 0.7308,
+      "step": 351
+    },
+    {
+      "epoch": 0.046075751659508,
+      "grad_norm": 2.8971149921417236,
+      "learning_rate": 7.08e-06,
+      "loss": 0.7238,
+      "step": 354
+    },
+    {
+      "epoch": 0.046466224131198754,
+      "grad_norm": 2.782917022705078,
+      "learning_rate": 7.14e-06,
+      "loss": 0.637,
+      "step": 357
+    },
+    {
+      "epoch": 0.0468566966028895,
+      "grad_norm": 4.860075950622559,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 0.6668,
+      "step": 360
+    },
+    {
+      "epoch": 0.04724716907458024,
+      "grad_norm": 3.7025034427642822,
+      "learning_rate": 7.260000000000001e-06,
+      "loss": 0.7802,
+      "step": 363
+    },
+    {
+      "epoch": 0.04763764154627099,
+      "grad_norm": 3.697002410888672,
+      "learning_rate": 7.32e-06,
+      "loss": 0.7448,
+      "step": 366
+    },
+    {
+      "epoch": 0.04802811401796173,
+      "grad_norm": 4.358737468719482,
+      "learning_rate": 7.3800000000000005e-06,
+      "loss": 0.7448,
+      "step": 369
+    },
+    {
+      "epoch": 0.04841858648965248,
+      "grad_norm": 3.1445984840393066,
+      "learning_rate": 7.440000000000001e-06,
+      "loss": 0.7037,
+      "step": 372
+    },
+    {
+      "epoch": 0.04880905896134322,
+      "grad_norm": 5.41956090927124,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.7322,
+      "step": 375
+    },
+    {
+      "epoch": 0.049199531433033974,
+      "grad_norm": 3.0740654468536377,
+      "learning_rate": 7.5600000000000005e-06,
+      "loss": 0.773,
+      "step": 378
+    },
+    {
+      "epoch": 0.04959000390472472,
+      "grad_norm": 3.2865071296691895,
+      "learning_rate": 7.620000000000001e-06,
+      "loss": 0.7685,
+      "step": 381
+    },
+    {
+      "epoch": 0.049980476376415464,
+      "grad_norm": 3.2605721950531006,
+      "learning_rate": 7.680000000000001e-06,
+      "loss": 0.7097,
+      "step": 384
+    },
+    {
+      "epoch": 0.05037094884810621,
+      "grad_norm": 2.969805955886841,
+      "learning_rate": 7.74e-06,
+      "loss": 0.5873,
+      "step": 387
+    },
+    {
+      "epoch": 0.050761421319796954,
+      "grad_norm": 3.413188934326172,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 0.7874,
+      "step": 390
+    },
+    {
+      "epoch": 0.0511518937914877,
+      "grad_norm": 4.274326324462891,
+      "learning_rate": 7.860000000000001e-06,
+      "loss": 0.7592,
+      "step": 393
+    },
+    {
+      "epoch": 0.051542366263178443,
+      "grad_norm": 2.9528450965881348,
+      "learning_rate": 7.92e-06,
+      "loss": 0.6821,
+      "step": 396
+    },
+    {
+      "epoch": 0.05193283873486919,
+      "grad_norm": 2.8550336360931396,
+      "learning_rate": 7.980000000000002e-06,
+      "loss": 0.6741,
+      "step": 399
+    },
+    {
+      "epoch": 0.05232331120655994,
+      "grad_norm": 2.7226080894470215,
+      "learning_rate": 8.040000000000001e-06,
+      "loss": 0.5926,
+      "step": 402
+    },
+    {
+      "epoch": 0.052713783678250685,
+      "grad_norm": 2.704489231109619,
+      "learning_rate": 8.1e-06,
+      "loss": 0.7503,
+      "step": 405
+    },
+    {
+      "epoch": 0.05310425614994143,
+      "grad_norm": 3.636996269226074,
+      "learning_rate": 8.16e-06,
+      "loss": 0.7294,
+      "step": 408
+    },
+    {
+      "epoch": 0.053494728621632175,
+      "grad_norm": 3.1197781562805176,
+      "learning_rate": 8.220000000000001e-06,
+      "loss": 0.7368,
+      "step": 411
+    },
+    {
+      "epoch": 0.05388520109332292,
+      "grad_norm": 3.0387392044067383,
+      "learning_rate": 8.28e-06,
+      "loss": 0.752,
+      "step": 414
+    },
+    {
+      "epoch": 0.054275673565013664,
+      "grad_norm": 3.60607647895813,
+      "learning_rate": 8.34e-06,
+      "loss": 0.7767,
+      "step": 417
+    },
+    {
+      "epoch": 0.05466614603670441,
+      "grad_norm": 3.62844181060791,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.6498,
+      "step": 420
+    },
+    {
+      "epoch": 0.05505661850839516,
+      "grad_norm": 3.531270980834961,
+      "learning_rate": 8.46e-06,
+      "loss": 0.7256,
+      "step": 423
+    },
+    {
+      "epoch": 0.055447090980085906,
+      "grad_norm": 3.4062726497650146,
+      "learning_rate": 8.52e-06,
+      "loss": 0.7756,
+      "step": 426
+    },
+    {
+      "epoch": 0.05583756345177665,
+      "grad_norm": 3.633138656616211,
+      "learning_rate": 8.580000000000001e-06,
+      "loss": 0.7159,
+      "step": 429
+    },
+    {
+      "epoch": 0.056228035923467395,
+      "grad_norm": 4.716714859008789,
+      "learning_rate": 8.64e-06,
+      "loss": 0.7503,
+      "step": 432
+    },
+    {
+      "epoch": 0.05661850839515814,
+      "grad_norm": 3.8363606929779053,
+      "learning_rate": 8.700000000000001e-06,
+      "loss": 0.708,
+      "step": 435
+    },
+    {
+      "epoch": 0.057008980866848885,
+      "grad_norm": 3.476318597793579,
+      "learning_rate": 8.76e-06,
+      "loss": 0.6928,
+      "step": 438
+    },
+    {
+      "epoch": 0.05739945333853963,
+      "grad_norm": 4.243283748626709,
+      "learning_rate": 8.82e-06,
+      "loss": 0.5729,
+      "step": 441
+    },
+    {
+      "epoch": 0.05778992581023038,
+      "grad_norm": 18.63896942138672,
+      "learning_rate": 8.880000000000001e-06,
+      "loss": 0.893,
+      "step": 444
+    },
+    {
+      "epoch": 0.05818039828192113,
+      "grad_norm": 3.8758761882781982,
+      "learning_rate": 8.94e-06,
+      "loss": 0.655,
+      "step": 447
+    },
+    {
+      "epoch": 0.05857087075361187,
+      "grad_norm": 2.645306348800659,
+      "learning_rate": 9e-06,
+      "loss": 0.6287,
+      "step": 450
+    },
+    {
+      "epoch": 0.058961343225302616,
+      "grad_norm": 2.911987066268921,
+      "learning_rate": 9.060000000000001e-06,
+      "loss": 0.7451,
+      "step": 453
+    },
+    {
+      "epoch": 0.05935181569699336,
+      "grad_norm": 3.4977638721466064,
+      "learning_rate": 9.12e-06,
+      "loss": 0.6821,
+      "step": 456
+    },
+    {
+      "epoch": 0.059742288168684106,
+      "grad_norm": 3.464571714401245,
+      "learning_rate": 9.180000000000002e-06,
+      "loss": 0.6449,
+      "step": 459
+    },
+    {
+      "epoch": 0.06013276064037485,
+      "grad_norm": 3.0574846267700195,
+      "learning_rate": 9.240000000000001e-06,
+      "loss": 0.7833,
+      "step": 462
+    },
+    {
+      "epoch": 0.0605232331120656,
+      "grad_norm": 3.181323766708374,
+      "learning_rate": 9.3e-06,
+      "loss": 0.759,
+      "step": 465
+    },
+    {
+      "epoch": 0.06091370558375635,
+      "grad_norm": 2.9368088245391846,
+      "learning_rate": 9.360000000000002e-06,
+      "loss": 0.681,
+      "step": 468
+    },
+    {
+      "epoch": 0.06130417805544709,
+      "grad_norm": 2.9919021129608154,
+      "learning_rate": 9.42e-06,
+      "loss": 0.6649,
+      "step": 471
+    },
+    {
+      "epoch": 0.06169465052713784,
+      "grad_norm": 3.4639129638671875,
+      "learning_rate": 9.48e-06,
+      "loss": 0.6816,
+      "step": 474
+    },
+    {
+      "epoch": 0.06208512299882858,
+      "grad_norm": 4.285403251647949,
+      "learning_rate": 9.54e-06,
+      "loss": 0.8209,
+      "step": 477
+    },
+    {
+      "epoch": 0.06247559547051933,
+      "grad_norm": 2.8827459812164307,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.7042,
+      "step": 480
+    },
+    {
+      "epoch": 0.06286606794221007,
+      "grad_norm": 3.176536798477173,
+      "learning_rate": 9.66e-06,
+      "loss": 0.7312,
+      "step": 483
+    },
+    {
+      "epoch": 0.06325654041390082,
+      "grad_norm": 4.091725826263428,
+      "learning_rate": 9.72e-06,
+      "loss": 0.6797,
+      "step": 486
+    },
+    {
+      "epoch": 0.06364701288559156,
+      "grad_norm": 3.53378963470459,
+      "learning_rate": 9.780000000000001e-06,
+      "loss": 0.6557,
+      "step": 489
+    },
+    {
+      "epoch": 0.0640374853572823,
+      "grad_norm": 3.440068006515503,
+      "learning_rate": 9.84e-06,
+      "loss": 0.6972,
+      "step": 492
+    },
+    {
+      "epoch": 0.06442795782897305,
+      "grad_norm": 3.1737592220306396,
+      "learning_rate": 9.9e-06,
+      "loss": 0.6832,
+      "step": 495
+    },
+    {
+      "epoch": 0.06481843030066381,
+      "grad_norm": 2.9271981716156006,
+      "learning_rate": 9.960000000000001e-06,
+      "loss": 0.7136,
+      "step": 498
+    },
+    {
+      "epoch": 0.06520890277235455,
+      "grad_norm": 4.281307697296143,
+      "learning_rate": 9.999999951472807e-06,
+      "loss": 0.7126,
+      "step": 501
+    },
+    {
+      "epoch": 0.0655993752440453,
+      "grad_norm": 4.331837177276611,
+      "learning_rate": 9.999999223564915e-06,
+      "loss": 0.6736,
+      "step": 504
+    },
+    {
+      "epoch": 0.06598984771573604,
+      "grad_norm": 3.0359110832214355,
+      "learning_rate": 9.999997622167676e-06,
+      "loss": 0.7754,
+      "step": 507
+    },
+    {
+      "epoch": 0.06638032018742679,
+      "grad_norm": 2.992680072784424,
+      "learning_rate": 9.999995147281374e-06,
+      "loss": 0.8101,
+      "step": 510
+    },
+    {
+      "epoch": 0.06677079265911753,
+      "grad_norm": 2.822829484939575,
+      "learning_rate": 9.999991798906435e-06,
+      "loss": 0.6705,
+      "step": 513
+    },
+    {
+      "epoch": 0.06716126513080828,
+      "grad_norm": 3.342939615249634,
+      "learning_rate": 9.999987577043449e-06,
+      "loss": 0.7045,
+      "step": 516
+    },
+    {
+      "epoch": 0.06755173760249902,
+      "grad_norm": 4.064047813415527,
+      "learning_rate": 9.999982481693151e-06,
+      "loss": 0.7007,
+      "step": 519
+    },
+    {
+      "epoch": 0.06794221007418977,
+      "grad_norm": 2.783782720565796,
+      "learning_rate": 9.999976512856434e-06,
+      "loss": 0.5655,
+      "step": 522
+    },
+    {
+      "epoch": 0.06833268254588051,
+      "grad_norm": 2.931049346923828,
+      "learning_rate": 9.999969670534335e-06,
+      "loss": 0.677,
+      "step": 525
+    },
+    {
+      "epoch": 0.06872315501757126,
+      "grad_norm": 3.2958552837371826,
+      "learning_rate": 9.999961954728054e-06,
+      "loss": 0.7741,
+      "step": 528
+    },
+    {
+      "epoch": 0.069113627489262,
+      "grad_norm": 2.8830935955047607,
+      "learning_rate": 9.999953365438939e-06,
+      "loss": 0.6607,
+      "step": 531
+    },
+    {
+      "epoch": 0.06950409996095275,
+      "grad_norm": 3.2814905643463135,
+      "learning_rate": 9.99994390266849e-06,
+      "loss": 0.7587,
+      "step": 534
+    },
+    {
+      "epoch": 0.06989457243264349,
+      "grad_norm": 4.126666069030762,
+      "learning_rate": 9.999933566418358e-06,
+      "loss": 0.7806,
+      "step": 537
+    },
+    {
+      "epoch": 0.07028504490433425,
+      "grad_norm": 3.1992428302764893,
+      "learning_rate": 9.99992235669035e-06,
+      "loss": 0.7045,
+      "step": 540
+    },
+    {
+      "epoch": 0.070675517376025,
+      "grad_norm": 3.3059449195861816,
+      "learning_rate": 9.999910273486427e-06,
+      "loss": 0.8061,
+      "step": 543
+    },
+    {
+      "epoch": 0.07106598984771574,
+      "grad_norm": 5.540506839752197,
+      "learning_rate": 9.999897316808695e-06,
+      "loss": 0.7123,
+      "step": 546
+    },
+    {
+      "epoch": 0.07145646231940649,
+      "grad_norm": 2.840941905975342,
+      "learning_rate": 9.999883486659421e-06,
+      "loss": 0.6211,
+      "step": 549
+    },
+    {
+      "epoch": 0.07184693479109723,
+      "grad_norm": 2.806445837020874,
+      "learning_rate": 9.99986878304102e-06,
+      "loss": 0.7114,
+      "step": 552
+    },
+    {
+      "epoch": 0.07223740726278798,
+      "grad_norm": 4.31725549697876,
+      "learning_rate": 9.99985320595606e-06,
+      "loss": 0.6363,
+      "step": 555
+    },
+    {
+      "epoch": 0.07262787973447872,
+      "grad_norm": 3.1201353073120117,
+      "learning_rate": 9.999836755407264e-06,
+      "loss": 0.7572,
+      "step": 558
+    },
+    {
+      "epoch": 0.07301835220616947,
+      "grad_norm": 4.234785079956055,
+      "learning_rate": 9.999819431397506e-06,
+      "loss": 0.6565,
+      "step": 561
+    },
+    {
+      "epoch": 0.07340882467786021,
+      "grad_norm": 3.2616043090820312,
+      "learning_rate": 9.999801233929808e-06,
+      "loss": 0.7319,
+      "step": 564
+    },
+    {
+      "epoch": 0.07379929714955095,
+      "grad_norm": 2.796783924102783,
+      "learning_rate": 9.999782163007357e-06,
+      "loss": 0.7293,
+      "step": 567
+    },
+    {
+      "epoch": 0.0741897696212417,
+      "grad_norm": 3.4508984088897705,
+      "learning_rate": 9.999762218633478e-06,
+      "loss": 0.8077,
+      "step": 570
+    },
+    {
+      "epoch": 0.07458024209293244,
+      "grad_norm": 2.6913444995880127,
+      "learning_rate": 9.999741400811656e-06,
+      "loss": 0.6777,
+      "step": 573
+    },
+    {
+      "epoch": 0.07497071456462319,
+      "grad_norm": 3.599764108657837,
+      "learning_rate": 9.99971970954553e-06,
+      "loss": 0.7533,
+      "step": 576
+    },
+    {
+      "epoch": 0.07536118703631393,
+      "grad_norm": 3.3088502883911133,
+      "learning_rate": 9.999697144838889e-06,
+      "loss": 0.605,
+      "step": 579
+    },
+    {
+      "epoch": 0.07575165950800468,
+      "grad_norm": 3.051250457763672,
+      "learning_rate": 9.999673706695676e-06,
+      "loss": 0.736,
+      "step": 582
+    },
+    {
+      "epoch": 0.07614213197969544,
+      "grad_norm": 2.775729179382324,
+      "learning_rate": 9.999649395119983e-06,
+      "loss": 0.7246,
+      "step": 585
+    },
+    {
+      "epoch": 0.07653260445138618,
+      "grad_norm": 3.2636523246765137,
+      "learning_rate": 9.999624210116057e-06,
+      "loss": 0.8448,
+      "step": 588
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 2.841562032699585,
+      "learning_rate": 9.999598151688301e-06,
+      "loss": 0.6813,
+      "step": 591
+    },
+    {
+      "epoch": 0.07731354939476767,
+      "grad_norm": 3.2505574226379395,
+      "learning_rate": 9.999571219841264e-06,
+      "loss": 0.6421,
+      "step": 594
+    },
+    {
+      "epoch": 0.07770402186645842,
+      "grad_norm": 4.958602428436279,
+      "learning_rate": 9.999543414579655e-06,
+      "loss": 0.7551,
+      "step": 597
+    },
+    {
+      "epoch": 0.07809449433814916,
+      "grad_norm": 3.9999706745147705,
+      "learning_rate": 9.999514735908326e-06,
+      "loss": 0.6695,
+      "step": 600
+    },
+    {
+      "epoch": 0.0784849668098399,
+      "grad_norm": 4.219988822937012,
+      "learning_rate": 9.999485183832291e-06,
+      "loss": 0.7231,
+      "step": 603
+    },
+    {
+      "epoch": 0.07887543928153065,
+      "grad_norm": 2.954352855682373,
+      "learning_rate": 9.999454758356713e-06,
+      "loss": 0.6714,
+      "step": 606
+    },
+    {
+      "epoch": 0.0792659117532214,
+      "grad_norm": 4.073672294616699,
+      "learning_rate": 9.999423459486906e-06,
+      "loss": 0.7234,
+      "step": 609
+    },
+    {
+      "epoch": 0.07965638422491214,
+      "grad_norm": 3.4436042308807373,
+      "learning_rate": 9.999391287228337e-06,
+      "loss": 0.8006,
+      "step": 612
+    },
+    {
+      "epoch": 0.08004685669660289,
+      "grad_norm": 3.66202974319458,
+      "learning_rate": 9.999358241586627e-06,
+      "loss": 0.6682,
+      "step": 615
+    },
+    {
+      "epoch": 0.08043732916829363,
+      "grad_norm": 4.19435977935791,
+      "learning_rate": 9.99932432256755e-06,
+      "loss": 0.7799,
+      "step": 618
+    },
+    {
+      "epoch": 0.08082780163998438,
+      "grad_norm": 3.3048112392425537,
+      "learning_rate": 9.99928953017703e-06,
+      "loss": 0.7087,
+      "step": 621
+    },
+    {
+      "epoch": 0.08121827411167512,
+      "grad_norm": 2.822140693664551,
+      "learning_rate": 9.999253864421147e-06,
+      "loss": 0.6894,
+      "step": 624
+    },
+    {
+      "epoch": 0.08160874658336588,
+      "grad_norm": 2.9324841499328613,
+      "learning_rate": 9.99921732530613e-06,
+      "loss": 0.6938,
+      "step": 627
+    },
+    {
+      "epoch": 0.08199921905505662,
+      "grad_norm": 3.84236478805542,
+      "learning_rate": 9.999179912838364e-06,
+      "loss": 0.783,
+      "step": 630
+    },
+    {
+      "epoch": 0.08238969152674737,
+      "grad_norm": 3.0340070724487305,
+      "learning_rate": 9.999141627024384e-06,
+      "loss": 0.7381,
+      "step": 633
+    },
+    {
+      "epoch": 0.08278016399843811,
+      "grad_norm": 4.351922512054443,
+      "learning_rate": 9.999102467870877e-06,
+      "loss": 0.7803,
+      "step": 636
+    },
+    {
+      "epoch": 0.08317063647012886,
+      "grad_norm": 2.6362740993499756,
+      "learning_rate": 9.999062435384686e-06,
+      "loss": 0.598,
+      "step": 639
+    },
+    {
+      "epoch": 0.0835611089418196,
+      "grad_norm": 3.186671495437622,
+      "learning_rate": 9.999021529572806e-06,
+      "loss": 0.7747,
+      "step": 642
+    },
+    {
+      "epoch": 0.08395158141351035,
+      "grad_norm": 3.305826187133789,
+      "learning_rate": 9.99897975044238e-06,
+      "loss": 0.7638,
+      "step": 645
+    },
+    {
+      "epoch": 0.0843420538852011,
+      "grad_norm": 3.8169655799865723,
+      "learning_rate": 9.998937098000705e-06,
+      "loss": 0.7015,
+      "step": 648
+    },
+    {
+      "epoch": 0.08473252635689184,
+      "grad_norm": 3.1275665760040283,
+      "learning_rate": 9.99889357225524e-06,
+      "loss": 0.6977,
+      "step": 651
+    },
+    {
+      "epoch": 0.08512299882858258,
+      "grad_norm": 3.6872782707214355,
+      "learning_rate": 9.998849173213581e-06,
+      "loss": 0.7307,
+      "step": 654
+    },
+    {
+      "epoch": 0.08551347130027333,
+      "grad_norm": 3.6949827671051025,
+      "learning_rate": 9.998803900883487e-06,
+      "loss": 0.6681,
+      "step": 657
+    },
+    {
+      "epoch": 0.08590394377196407,
+      "grad_norm": 3.2044758796691895,
+      "learning_rate": 9.99875775527287e-06,
+      "loss": 0.6699,
+      "step": 660
+    },
+    {
+      "epoch": 0.08629441624365482,
+      "grad_norm": 3.4927258491516113,
+      "learning_rate": 9.998710736389787e-06,
+      "loss": 0.6393,
+      "step": 663
+    },
+    {
+      "epoch": 0.08668488871534556,
+      "grad_norm": 3.450648784637451,
+      "learning_rate": 9.998662844242456e-06,
+      "loss": 0.6751,
+      "step": 666
+    },
+    {
+      "epoch": 0.08707536118703631,
+      "grad_norm": 3.036738157272339,
+      "learning_rate": 9.99861407883924e-06,
+      "loss": 0.7606,
+      "step": 669
+    },
+    {
+      "epoch": 0.08746583365872707,
+      "grad_norm": 3.3343772888183594,
+      "learning_rate": 9.998564440188661e-06,
+      "loss": 0.7669,
+      "step": 672
+    },
+    {
+      "epoch": 0.08785630613041781,
+      "grad_norm": 3.513052225112915,
+      "learning_rate": 9.998513928299389e-06,
+      "loss": 0.6073,
+      "step": 675
+    },
+    {
+      "epoch": 0.08824677860210856,
+      "grad_norm": 2.9392402172088623,
+      "learning_rate": 9.998462543180249e-06,
+      "loss": 0.6474,
+      "step": 678
+    },
+    {
+      "epoch": 0.0886372510737993,
+      "grad_norm": 3.289640188217163,
+      "learning_rate": 9.998410284840217e-06,
+      "loss": 0.6522,
+      "step": 681
+    },
+    {
+      "epoch": 0.08902772354549005,
+      "grad_norm": 5.727564811706543,
+      "learning_rate": 9.998357153288425e-06,
+      "loss": 0.6354,
+      "step": 684
+    },
+    {
+      "epoch": 0.08941819601718079,
+      "grad_norm": 2.87939715385437,
+      "learning_rate": 9.998303148534153e-06,
+      "loss": 0.6943,
+      "step": 687
+    },
+    {
+      "epoch": 0.08980866848887153,
+      "grad_norm": 5.636757850646973,
+      "learning_rate": 9.998248270586837e-06,
+      "loss": 0.651,
+      "step": 690
+    },
+    {
+      "epoch": 0.09019914096056228,
+      "grad_norm": 4.383195400238037,
+      "learning_rate": 9.99819251945606e-06,
+      "loss": 0.7694,
+      "step": 693
+    },
+    {
+      "epoch": 0.09058961343225302,
+      "grad_norm": 3.0773353576660156,
+      "learning_rate": 9.998135895151567e-06,
+      "loss": 0.6747,
+      "step": 696
+    },
+    {
+      "epoch": 0.09098008590394377,
+      "grad_norm": 3.2680556774139404,
+      "learning_rate": 9.998078397683246e-06,
+      "loss": 0.7254,
+      "step": 699
+    },
+    {
+      "epoch": 0.09137055837563451,
+      "grad_norm": 7.307571887969971,
+      "learning_rate": 9.998020027061145e-06,
+      "loss": 0.671,
+      "step": 702
+    },
+    {
+      "epoch": 0.09176103084732526,
+      "grad_norm": 2.9415576457977295,
+      "learning_rate": 9.99796078329546e-06,
+      "loss": 0.7709,
+      "step": 705
+    },
+    {
+      "epoch": 0.092151503319016,
+      "grad_norm": 2.814495086669922,
+      "learning_rate": 9.997900666396539e-06,
+      "loss": 0.6947,
+      "step": 708
+    },
+    {
+      "epoch": 0.09254197579070675,
+      "grad_norm": 5.5829081535339355,
+      "learning_rate": 9.997839676374885e-06,
+      "loss": 0.73,
+      "step": 711
+    },
+    {
+      "epoch": 0.09293244826239751,
+      "grad_norm": 5.697535037994385,
+      "learning_rate": 9.997777813241154e-06,
+      "loss": 0.6227,
+      "step": 714
+    },
+    {
+      "epoch": 0.09332292073408825,
+      "grad_norm": 3.0062732696533203,
+      "learning_rate": 9.997715077006152e-06,
+      "loss": 0.6035,
+      "step": 717
+    },
+    {
+      "epoch": 0.093713393205779,
+      "grad_norm": 2.8236165046691895,
+      "learning_rate": 9.997651467680843e-06,
+      "loss": 0.7009,
+      "step": 720
+    },
+    {
+      "epoch": 0.09410386567746974,
+      "grad_norm": 2.7727997303009033,
+      "learning_rate": 9.997586985276333e-06,
+      "loss": 0.6508,
+      "step": 723
+    },
+    {
+      "epoch": 0.09449433814916049,
+      "grad_norm": 3.65282940864563,
+      "learning_rate": 9.99752162980389e-06,
+      "loss": 0.7025,
+      "step": 726
+    },
+    {
+      "epoch": 0.09488481062085123,
+      "grad_norm": 3.1343119144439697,
+      "learning_rate": 9.997455401274932e-06,
+      "loss": 0.7112,
+      "step": 729
+    },
+    {
+      "epoch": 0.09527528309254198,
+      "grad_norm": 3.504101037979126,
+      "learning_rate": 9.99738829970103e-06,
+      "loss": 0.7705,
+      "step": 732
+    },
+    {
+      "epoch": 0.09566575556423272,
+      "grad_norm": 2.7946465015411377,
+      "learning_rate": 9.997320325093903e-06,
+      "loss": 0.5736,
+      "step": 735
+    },
+    {
+      "epoch": 0.09605622803592347,
+      "grad_norm": 3.221226215362549,
+      "learning_rate": 9.99725147746543e-06,
+      "loss": 0.647,
+      "step": 738
+    },
+    {
+      "epoch": 0.09644670050761421,
+      "grad_norm": 5.345529079437256,
+      "learning_rate": 9.997181756827634e-06,
+      "loss": 0.5896,
+      "step": 741
+    },
+    {
+      "epoch": 0.09683717297930496,
+      "grad_norm": 3.8715436458587646,
+      "learning_rate": 9.9971111631927e-06,
+      "loss": 0.7666,
+      "step": 744
+    },
+    {
+      "epoch": 0.0972276454509957,
+      "grad_norm": 2.9973673820495605,
+      "learning_rate": 9.997039696572956e-06,
+      "loss": 0.6106,
+      "step": 747
+    },
+    {
+      "epoch": 0.09761811792268645,
+      "grad_norm": 5.8889851570129395,
+      "learning_rate": 9.996967356980891e-06,
+      "loss": 0.6416,
+      "step": 750
+    },
+    {
+      "epoch": 0.09800859039437719,
+      "grad_norm": 3.953415632247925,
+      "learning_rate": 9.99689414442914e-06,
+      "loss": 0.744,
+      "step": 753
+    },
+    {
+      "epoch": 0.09839906286606795,
+      "grad_norm": 4.9584760665893555,
+      "learning_rate": 9.996820058930495e-06,
+      "loss": 0.6435,
+      "step": 756
+    },
+    {
+      "epoch": 0.0987895353377587,
+      "grad_norm": 2.774200916290283,
+      "learning_rate": 9.996745100497898e-06,
+      "loss": 0.6054,
+      "step": 759
+    },
+    {
+      "epoch": 0.09918000780944944,
+      "grad_norm": 4.4339118003845215,
+      "learning_rate": 9.996669269144442e-06,
+      "loss": 0.6135,
+      "step": 762
+    },
+    {
+      "epoch": 0.09957048028114018,
+      "grad_norm": 3.3275766372680664,
+      "learning_rate": 9.996592564883376e-06,
+      "loss": 0.7489,
+      "step": 765
+    },
+    {
+      "epoch": 0.09996095275283093,
+      "grad_norm": 3.1275625228881836,
+      "learning_rate": 9.996514987728101e-06,
+      "loss": 0.825,
+      "step": 768
+    },
+    {
+      "epoch": 0.10035142522452167,
+      "grad_norm": 3.67669415473938,
+      "learning_rate": 9.99643653769217e-06,
+      "loss": 0.8093,
+      "step": 771
+    },
+    {
+      "epoch": 0.10074189769621242,
+      "grad_norm": 3.200958013534546,
+      "learning_rate": 9.996357214789284e-06,
+      "loss": 0.7373,
+      "step": 774
+    },
+    {
+      "epoch": 0.10113237016790316,
+      "grad_norm": 2.847730875015259,
+      "learning_rate": 9.996277019033305e-06,
+      "loss": 0.7335,
+      "step": 777
+    },
+    {
+      "epoch": 0.10152284263959391,
+      "grad_norm": 4.33748722076416,
+      "learning_rate": 9.99619595043824e-06,
+      "loss": 0.6688,
+      "step": 780
+    },
+    {
+      "epoch": 0.10191331511128465,
+      "grad_norm": 3.9746432304382324,
+      "learning_rate": 9.996114009018254e-06,
+      "loss": 0.7776,
+      "step": 783
+    },
+    {
+      "epoch": 0.1023037875829754,
+      "grad_norm": 3.600278854370117,
+      "learning_rate": 9.996031194787661e-06,
+      "loss": 0.7403,
+      "step": 786
+    },
+    {
+      "epoch": 0.10269426005466614,
+      "grad_norm": 2.957000494003296,
+      "learning_rate": 9.995947507760928e-06,
+      "loss": 0.7485,
+      "step": 789
+    },
+    {
+      "epoch": 0.10308473252635689,
+      "grad_norm": 4.908221244812012,
+      "learning_rate": 9.995862947952676e-06,
+      "loss": 0.748,
+      "step": 792
+    },
+    {
+      "epoch": 0.10347520499804763,
+      "grad_norm": 3.816086769104004,
+      "learning_rate": 9.995777515377677e-06,
+      "loss": 0.6787,
+      "step": 795
+    },
+    {
+      "epoch": 0.10386567746973838,
+      "grad_norm": 2.948234796524048,
+      "learning_rate": 9.995691210050854e-06,
+      "loss": 0.7267,
+      "step": 798
+    },
+    {
+      "epoch": 0.10425614994142914,
+      "grad_norm": 2.818983793258667,
+      "learning_rate": 9.995604031987287e-06,
+      "loss": 0.6914,
+      "step": 801
+    },
+    {
+      "epoch": 0.10464662241311988,
+      "grad_norm": 3.0057449340820312,
+      "learning_rate": 9.995515981202206e-06,
+      "loss": 0.7294,
+      "step": 804
+    },
+    {
+      "epoch": 0.10503709488481063,
+      "grad_norm": 3.676535129547119,
+      "learning_rate": 9.99542705771099e-06,
+      "loss": 0.6236,
+      "step": 807
+    },
+    {
+      "epoch": 0.10542756735650137,
+      "grad_norm": 3.7774033546447754,
+      "learning_rate": 9.995337261529176e-06,
+      "loss": 0.6749,
+      "step": 810
+    },
+    {
+      "epoch": 0.10581803982819211,
+      "grad_norm": 3.2406165599823,
+      "learning_rate": 9.995246592672451e-06,
+      "loss": 0.729,
+      "step": 813
+    },
+    {
+      "epoch": 0.10620851229988286,
+      "grad_norm": 5.026638984680176,
+      "learning_rate": 9.995155051156657e-06,
+      "loss": 0.7358,
+      "step": 816
+    },
+    {
+      "epoch": 0.1065989847715736,
+      "grad_norm": 3.2842941284179688,
+      "learning_rate": 9.995062636997783e-06,
+      "loss": 0.6659,
+      "step": 819
+    },
+    {
+      "epoch": 0.10698945724326435,
+      "grad_norm": 2.9089906215667725,
+      "learning_rate": 9.994969350211974e-06,
+      "loss": 0.6672,
+      "step": 822
+    },
+    {
+      "epoch": 0.1073799297149551,
+      "grad_norm": 2.6798436641693115,
+      "learning_rate": 9.994875190815527e-06,
+      "loss": 0.6423,
+      "step": 825
+    },
+    {
+      "epoch": 0.10777040218664584,
+      "grad_norm": 2.918283462524414,
+      "learning_rate": 9.994780158824891e-06,
+      "loss": 0.7637,
+      "step": 828
+    },
+    {
+      "epoch": 0.10816087465833658,
+      "grad_norm": 3.470376491546631,
+      "learning_rate": 9.99468425425667e-06,
+      "loss": 0.7711,
+      "step": 831
+    },
+    {
+      "epoch": 0.10855134713002733,
+      "grad_norm": 3.4783475399017334,
+      "learning_rate": 9.994587477127617e-06,
+      "loss": 0.8156,
+      "step": 834
+    },
+    {
+      "epoch": 0.10894181960171807,
+      "grad_norm": 2.7268190383911133,
+      "learning_rate": 9.994489827454638e-06,
+      "loss": 0.6253,
+      "step": 837
+    },
+    {
+      "epoch": 0.10933229207340882,
+      "grad_norm": 3.936289072036743,
+      "learning_rate": 9.994391305254795e-06,
+      "loss": 0.6787,
+      "step": 840
+    },
+    {
+      "epoch": 0.10972276454509958,
+      "grad_norm": 3.084895610809326,
+      "learning_rate": 9.994291910545296e-06,
+      "loss": 0.6393,
+      "step": 843
+    },
+    {
+      "epoch": 0.11011323701679032,
+      "grad_norm": 2.9369547367095947,
+      "learning_rate": 9.994191643343508e-06,
+      "loss": 0.6388,
+      "step": 846
+    },
+    {
+      "epoch": 0.11050370948848107,
+      "grad_norm": 2.8913955688476562,
+      "learning_rate": 9.994090503666945e-06,
+      "loss": 0.6447,
+      "step": 849
+    },
+    {
+      "epoch": 0.11089418196017181,
+      "grad_norm": 2.990058183670044,
+      "learning_rate": 9.993988491533277e-06,
+      "loss": 0.6725,
+      "step": 852
+    },
+    {
+      "epoch": 0.11128465443186256,
+      "grad_norm": 3.0597927570343018,
+      "learning_rate": 9.993885606960325e-06,
+      "loss": 0.7052,
+      "step": 855
+    },
+    {
+      "epoch": 0.1116751269035533,
+      "grad_norm": 4.286864280700684,
+      "learning_rate": 9.993781849966064e-06,
+      "loss": 0.6557,
+      "step": 858
+    },
+    {
+      "epoch": 0.11206559937524405,
+      "grad_norm": 2.70944881439209,
+      "learning_rate": 9.993677220568619e-06,
+      "loss": 0.7614,
+      "step": 861
+    },
+    {
+      "epoch": 0.11245607184693479,
+      "grad_norm": 2.6077942848205566,
+      "learning_rate": 9.993571718786268e-06,
+      "loss": 0.713,
+      "step": 864
+    },
+    {
+      "epoch": 0.11284654431862554,
+      "grad_norm": 2.833029270172119,
+      "learning_rate": 9.993465344637443e-06,
+      "loss": 0.6649,
+      "step": 867
+    },
+    {
+      "epoch": 0.11323701679031628,
+      "grad_norm": 2.6848433017730713,
+      "learning_rate": 9.993358098140727e-06,
+      "loss": 0.7228,
+      "step": 870
+    },
+    {
+      "epoch": 0.11362748926200703,
+      "grad_norm": 3.404714822769165,
+      "learning_rate": 9.993249979314857e-06,
+      "loss": 0.6024,
+      "step": 873
+    },
+    {
+      "epoch": 0.11401796173369777,
+      "grad_norm": 4.099708557128906,
+      "learning_rate": 9.993140988178718e-06,
+      "loss": 0.664,
+      "step": 876
+    },
+    {
+      "epoch": 0.11440843420538852,
+      "grad_norm": 2.908158302307129,
+      "learning_rate": 9.993031124751353e-06,
+      "loss": 0.7663,
+      "step": 879
+    },
+    {
+      "epoch": 0.11479890667707926,
+      "grad_norm": 2.5075185298919678,
+      "learning_rate": 9.992920389051955e-06,
+      "loss": 0.7454,
+      "step": 882
+    },
+    {
+      "epoch": 0.11518937914877002,
+      "grad_norm": 3.0805585384368896,
+      "learning_rate": 9.992808781099868e-06,
+      "loss": 0.7226,
+      "step": 885
+    },
+    {
+      "epoch": 0.11557985162046076,
+      "grad_norm": 3.1145894527435303,
+      "learning_rate": 9.992696300914591e-06,
+      "loss": 0.732,
+      "step": 888
+    },
+    {
+      "epoch": 0.11597032409215151,
+      "grad_norm": 3.146247386932373,
+      "learning_rate": 9.992582948515772e-06,
+      "loss": 0.785,
+      "step": 891
+    },
+    {
+      "epoch": 0.11636079656384225,
+      "grad_norm": 3.7035303115844727,
+      "learning_rate": 9.992468723923216e-06,
+      "loss": 0.7226,
+      "step": 894
+    },
+    {
+      "epoch": 0.116751269035533,
+      "grad_norm": 3.59032940864563,
+      "learning_rate": 9.992353627156876e-06,
+      "loss": 0.6496,
+      "step": 897
+    },
+    {
+      "epoch": 0.11714174150722374,
+      "grad_norm": 3.1689419746398926,
+      "learning_rate": 9.992237658236859e-06,
+      "loss": 0.6467,
+      "step": 900
+    },
+    {
+      "epoch": 0.11753221397891449,
+      "grad_norm": 2.9618992805480957,
+      "learning_rate": 9.992120817183427e-06,
+      "loss": 0.6797,
+      "step": 903
+    },
+    {
+      "epoch": 0.11792268645060523,
+      "grad_norm": 2.657771587371826,
+      "learning_rate": 9.992003104016988e-06,
+      "loss": 0.6924,
+      "step": 906
+    },
+    {
+      "epoch": 0.11831315892229598,
+      "grad_norm": 2.938518524169922,
+      "learning_rate": 9.99188451875811e-06,
+      "loss": 0.722,
+      "step": 909
+    },
+    {
+      "epoch": 0.11870363139398672,
+      "grad_norm": 2.878551959991455,
+      "learning_rate": 9.991765061427508e-06,
+      "loss": 0.6749,
+      "step": 912
+    },
+    {
+      "epoch": 0.11909410386567747,
+      "grad_norm": 3.0917224884033203,
+      "learning_rate": 9.99164473204605e-06,
+      "loss": 0.634,
+      "step": 915
+    },
+    {
+      "epoch": 0.11948457633736821,
+      "grad_norm": 3.893941879272461,
+      "learning_rate": 9.991523530634758e-06,
+      "loss": 0.7202,
+      "step": 918
+    },
+    {
+      "epoch": 0.11987504880905896,
+      "grad_norm": 12.49660873413086,
+      "learning_rate": 9.991401457214807e-06,
+      "loss": 0.7348,
+      "step": 921
+    },
+    {
+      "epoch": 0.1202655212807497,
+      "grad_norm": 2.8306703567504883,
+      "learning_rate": 9.99127851180752e-06,
+      "loss": 0.7267,
+      "step": 924
+    },
+    {
+      "epoch": 0.12065599375244045,
+      "grad_norm": 3.2142062187194824,
+      "learning_rate": 9.99115469443438e-06,
+      "loss": 0.6601,
+      "step": 927
+    },
+    {
+      "epoch": 0.1210464662241312,
+      "grad_norm": 2.776210308074951,
+      "learning_rate": 9.991030005117013e-06,
+      "loss": 0.671,
+      "step": 930
+    },
+    {
+      "epoch": 0.12143693869582195,
+      "grad_norm": 3.657951593399048,
+      "learning_rate": 9.990904443877203e-06,
+      "loss": 0.6504,
+      "step": 933
+    },
+    {
+      "epoch": 0.1218274111675127,
+      "grad_norm": 2.87703800201416,
+      "learning_rate": 9.990778010736885e-06,
+      "loss": 0.7212,
+      "step": 936
+    },
+    {
+      "epoch": 0.12221788363920344,
+      "grad_norm": 4.6955718994140625,
+      "learning_rate": 9.99065070571815e-06,
+      "loss": 0.6704,
+      "step": 939
+    },
+    {
+      "epoch": 0.12260835611089418,
+      "grad_norm": 4.890464782714844,
+      "learning_rate": 9.990522528843236e-06,
+      "loss": 0.587,
+      "step": 942
+    },
+    {
+      "epoch": 0.12299882858258493,
+      "grad_norm": 3.7932112216949463,
+      "learning_rate": 9.990393480134532e-06,
+      "loss": 0.6647,
+      "step": 945
+    },
+    {
+      "epoch": 0.12338930105427567,
+      "grad_norm": 2.9928741455078125,
+      "learning_rate": 9.990263559614589e-06,
+      "loss": 0.7145,
+      "step": 948
+    },
+    {
+      "epoch": 0.12377977352596642,
+      "grad_norm": 2.880676031112671,
+      "learning_rate": 9.990132767306097e-06,
+      "loss": 0.6792,
+      "step": 951
+    },
+    {
+      "epoch": 0.12417024599765716,
+      "grad_norm": 2.795924186706543,
+      "learning_rate": 9.990001103231909e-06,
+      "loss": 0.6022,
+      "step": 954
+    },
+    {
+      "epoch": 0.12456071846934791,
+      "grad_norm": 2.9928035736083984,
+      "learning_rate": 9.989868567415027e-06,
+      "loss": 0.6714,
+      "step": 957
+    },
+    {
+      "epoch": 0.12495119094103865,
+      "grad_norm": 2.6535706520080566,
+      "learning_rate": 9.989735159878601e-06,
+      "loss": 0.6269,
+      "step": 960
+    },
+    {
+      "epoch": 0.1253416634127294,
+      "grad_norm": 2.8510921001434326,
+      "learning_rate": 9.98960088064594e-06,
+      "loss": 0.7881,
+      "step": 963
+    },
+    {
+      "epoch": 0.12573213588442014,
+      "grad_norm": 2.8703200817108154,
+      "learning_rate": 9.989465729740504e-06,
+      "loss": 0.641,
+      "step": 966
+    },
+    {
+      "epoch": 0.1261226083561109,
+      "grad_norm": 2.8146934509277344,
+      "learning_rate": 9.989329707185899e-06,
+      "loss": 0.5499,
+      "step": 969
+    },
+    {
+      "epoch": 0.12651308082780163,
+      "grad_norm": 2.848714590072632,
+      "learning_rate": 9.989192813005891e-06,
+      "loss": 0.6357,
+      "step": 972
+    },
+    {
+      "epoch": 0.12690355329949238,
+      "grad_norm": 3.125911235809326,
+      "learning_rate": 9.989055047224393e-06,
+      "loss": 0.8084,
+      "step": 975
+    },
+    {
+      "epoch": 0.12729402577118312,
+      "grad_norm": 3.852508306503296,
+      "learning_rate": 9.988916409865476e-06,
+      "loss": 0.6652,
+      "step": 978
+    },
+    {
+      "epoch": 0.12768449824287387,
+      "grad_norm": 3.4313900470733643,
+      "learning_rate": 9.988776900953356e-06,
+      "loss": 0.6909,
+      "step": 981
+    },
+    {
+      "epoch": 0.1280749707145646,
+      "grad_norm": 2.9528424739837646,
+      "learning_rate": 9.988636520512407e-06,
+      "loss": 0.6335,
+      "step": 984
+    },
+    {
+      "epoch": 0.12846544318625536,
+      "grad_norm": 3.1923022270202637,
+      "learning_rate": 9.988495268567152e-06,
+      "loss": 0.7599,
+      "step": 987
+    },
+    {
+      "epoch": 0.1288559156579461,
+      "grad_norm": 2.8430709838867188,
+      "learning_rate": 9.988353145142267e-06,
+      "loss": 0.659,
+      "step": 990
+    },
+    {
+      "epoch": 0.12924638812963687,
+      "grad_norm": 4.609011173248291,
+      "learning_rate": 9.988210150262582e-06,
+      "loss": 0.7288,
+      "step": 993
+    },
+    {
+      "epoch": 0.12963686060132762,
+      "grad_norm": 3.1884047985076904,
+      "learning_rate": 9.98806628395308e-06,
+      "loss": 0.8857,
+      "step": 996
+    },
+    {
+      "epoch": 0.13002733307301836,
+      "grad_norm": 4.605678081512451,
+      "learning_rate": 9.987921546238888e-06,
+      "loss": 0.6314,
+      "step": 999
+    },
+    {
+      "epoch": 0.1304178055447091,
+      "grad_norm": 2.8281877040863037,
+      "learning_rate": 9.987775937145297e-06,
+      "loss": 0.6079,
+      "step": 1002
+    },
+    {
+      "epoch": 0.13080827801639985,
+      "grad_norm": 7.240689277648926,
+      "learning_rate": 9.987629456697741e-06,
+      "loss": 0.6683,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1311987504880906,
+      "grad_norm": 3.1431357860565186,
+      "learning_rate": 9.987482104921813e-06,
+      "loss": 0.7741,
+      "step": 1008
+    },
+    {
+      "epoch": 0.13158922295978134,
+      "grad_norm": 3.569730043411255,
+      "learning_rate": 9.98733388184325e-06,
+      "loss": 0.83,
+      "step": 1011
+    },
+    {
+      "epoch": 0.1319796954314721,
+      "grad_norm": 2.689365863800049,
+      "learning_rate": 9.987184787487953e-06,
+      "loss": 0.7279,
+      "step": 1014
+    },
+    {
+      "epoch": 0.13237016790316283,
+      "grad_norm": 2.8707399368286133,
+      "learning_rate": 9.987034821881965e-06,
+      "loss": 0.7021,
+      "step": 1017
+    },
+    {
+      "epoch": 0.13276064037485358,
+      "grad_norm": 2.8549368381500244,
+      "learning_rate": 9.986883985051485e-06,
+      "loss": 0.6305,
+      "step": 1020
+    },
+    {
+      "epoch": 0.13315111284654432,
+      "grad_norm": 2.8607091903686523,
+      "learning_rate": 9.986732277022862e-06,
+      "loss": 0.647,
+      "step": 1023
+    },
+    {
+      "epoch": 0.13354158531823507,
+      "grad_norm": 3.3489599227905273,
+      "learning_rate": 9.986579697822601e-06,
+      "loss": 0.624,
+      "step": 1026
+    },
+    {
+      "epoch": 0.1339320577899258,
+      "grad_norm": 3.7060370445251465,
+      "learning_rate": 9.986426247477358e-06,
+      "loss": 0.7278,
+      "step": 1029
+    },
+    {
+      "epoch": 0.13432253026161656,
+      "grad_norm": 3.295768976211548,
+      "learning_rate": 9.98627192601394e-06,
+      "loss": 0.6298,
+      "step": 1032
+    },
+    {
+      "epoch": 0.1347130027333073,
+      "grad_norm": 4.761738300323486,
+      "learning_rate": 9.986116733459303e-06,
+      "loss": 0.8255,
+      "step": 1035
+    },
+    {
+      "epoch": 0.13510347520499805,
+      "grad_norm": 2.9362332820892334,
+      "learning_rate": 9.985960669840564e-06,
+      "loss": 0.6209,
+      "step": 1038
+    },
+    {
+      "epoch": 0.1354939476766888,
+      "grad_norm": 2.879631519317627,
+      "learning_rate": 9.985803735184986e-06,
+      "loss": 0.6995,
+      "step": 1041
+    },
+    {
+      "epoch": 0.13588442014837954,
+      "grad_norm": 2.8256990909576416,
+      "learning_rate": 9.985645929519983e-06,
+      "loss": 0.6906,
+      "step": 1044
+    },
+    {
+      "epoch": 0.13627489262007028,
+      "grad_norm": 2.7403573989868164,
+      "learning_rate": 9.985487252873125e-06,
+      "loss": 0.5904,
+      "step": 1047
+    },
+    {
+      "epoch": 0.13666536509176103,
+      "grad_norm": 2.701652765274048,
+      "learning_rate": 9.98532770527213e-06,
+      "loss": 0.6527,
+      "step": 1050
+    },
+    {
+      "epoch": 0.13705583756345177,
+      "grad_norm": 2.5606539249420166,
+      "learning_rate": 9.985167286744875e-06,
+      "loss": 0.7061,
+      "step": 1053
+    },
+    {
+      "epoch": 0.13744631003514252,
+      "grad_norm": 2.821444511413574,
+      "learning_rate": 9.98500599731938e-06,
+      "loss": 0.671,
+      "step": 1056
+    },
+    {
+      "epoch": 0.13783678250683326,
+      "grad_norm": 3.4483652114868164,
+      "learning_rate": 9.984843837023826e-06,
+      "loss": 0.5973,
+      "step": 1059
+    },
+    {
+      "epoch": 0.138227254978524,
+      "grad_norm": 2.7346315383911133,
+      "learning_rate": 9.984680805886538e-06,
+      "loss": 0.6508,
+      "step": 1062
+    },
+    {
+      "epoch": 0.13861772745021475,
+      "grad_norm": 3.0941390991210938,
+      "learning_rate": 9.984516903936002e-06,
+      "loss": 0.6686,
+      "step": 1065
+    },
+    {
+      "epoch": 0.1390081999219055,
+      "grad_norm": 2.95706844329834,
+      "learning_rate": 9.984352131200847e-06,
+      "loss": 0.7592,
+      "step": 1068
+    },
+    {
+      "epoch": 0.13939867239359624,
+      "grad_norm": 2.627673864364624,
+      "learning_rate": 9.984186487709862e-06,
+      "loss": 0.6032,
+      "step": 1071
+    },
+    {
+      "epoch": 0.13978914486528699,
+      "grad_norm": 3.3096377849578857,
+      "learning_rate": 9.984019973491981e-06,
+      "loss": 0.7385,
+      "step": 1074
+    },
+    {
+      "epoch": 0.14017961733697773,
+      "grad_norm": 2.916146993637085,
+      "learning_rate": 9.983852588576296e-06,
+      "loss": 0.6303,
+      "step": 1077
+    },
+    {
+      "epoch": 0.1405700898086685,
+      "grad_norm": 6.1797614097595215,
+      "learning_rate": 9.983684332992049e-06,
+      "loss": 0.6442,
+      "step": 1080
+    },
+    {
+      "epoch": 0.14096056228035925,
+      "grad_norm": 4.810859203338623,
+      "learning_rate": 9.983515206768633e-06,
+      "loss": 0.7992,
+      "step": 1083
+    },
+    {
+      "epoch": 0.14135103475205,
+      "grad_norm": 2.918363094329834,
+      "learning_rate": 9.983345209935593e-06,
+      "loss": 0.6568,
+      "step": 1086
+    },
+    {
+      "epoch": 0.14174150722374074,
+      "grad_norm": 2.725175380706787,
+      "learning_rate": 9.983174342522628e-06,
+      "loss": 0.6201,
+      "step": 1089
+    },
+    {
+      "epoch": 0.14213197969543148,
+      "grad_norm": 3.9198555946350098,
+      "learning_rate": 9.983002604559591e-06,
+      "loss": 0.6546,
+      "step": 1092
+    },
+    {
+      "epoch": 0.14252245216712223,
+      "grad_norm": 2.7483971118927,
+      "learning_rate": 9.98282999607648e-06,
+      "loss": 0.7721,
+      "step": 1095
+    },
+    {
+      "epoch": 0.14291292463881297,
+      "grad_norm": 3.3657147884368896,
+      "learning_rate": 9.982656517103451e-06,
+      "loss": 0.7828,
+      "step": 1098
+    },
+    {
+      "epoch": 0.14330339711050372,
+      "grad_norm": 2.881145715713501,
+      "learning_rate": 9.982482167670811e-06,
+      "loss": 0.6587,
+      "step": 1101
+    },
+    {
+      "epoch": 0.14369386958219446,
+      "grad_norm": 3.116957187652588,
+      "learning_rate": 9.982306947809016e-06,
+      "loss": 0.8168,
+      "step": 1104
+    },
+    {
+      "epoch": 0.1440843420538852,
+      "grad_norm": 3.0553340911865234,
+      "learning_rate": 9.98213085754868e-06,
+      "loss": 0.6254,
+      "step": 1107
+    },
+    {
+      "epoch": 0.14447481452557595,
+      "grad_norm": 2.685007095336914,
+      "learning_rate": 9.981953896920564e-06,
+      "loss": 0.6368,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1448652869972667,
+      "grad_norm": 3.4758572578430176,
+      "learning_rate": 9.981776065955583e-06,
+      "loss": 0.6683,
+      "step": 1113
+    },
+    {
+      "epoch": 0.14525575946895744,
+      "grad_norm": 3.288086414337158,
+      "learning_rate": 9.981597364684804e-06,
+      "loss": 0.6544,
+      "step": 1116
+    },
+    {
+      "epoch": 0.14564623194064819,
+      "grad_norm": 2.9728589057922363,
+      "learning_rate": 9.981417793139443e-06,
+      "loss": 0.7256,
+      "step": 1119
+    },
+    {
+      "epoch": 0.14603670441233893,
+      "grad_norm": 4.481566429138184,
+      "learning_rate": 9.981237351350874e-06,
+      "loss": 0.6933,
+      "step": 1122
+    },
+    {
+      "epoch": 0.14642717688402968,
+      "grad_norm": 3.00492000579834,
+      "learning_rate": 9.98105603935062e-06,
+      "loss": 0.7278,
+      "step": 1125
+    },
+    {
+      "epoch": 0.14681764935572042,
+      "grad_norm": 3.2771825790405273,
+      "learning_rate": 9.980873857170352e-06,
+      "loss": 0.7472,
+      "step": 1128
+    },
+    {
+      "epoch": 0.14720812182741116,
+      "grad_norm": 4.5184550285339355,
+      "learning_rate": 9.980690804841901e-06,
+      "loss": 0.6744,
+      "step": 1131
+    },
+    {
+      "epoch": 0.1475985942991019,
+      "grad_norm": 3.881673574447632,
+      "learning_rate": 9.980506882397246e-06,
+      "loss": 0.5923,
+      "step": 1134
+    },
+    {
+      "epoch": 0.14798906677079265,
+      "grad_norm": 3.5104172229766846,
+      "learning_rate": 9.980322089868512e-06,
+      "loss": 0.6073,
+      "step": 1137
+    },
+    {
+      "epoch": 0.1483795392424834,
+      "grad_norm": 4.261860370635986,
+      "learning_rate": 9.980136427287989e-06,
+      "loss": 0.6716,
+      "step": 1140
+    },
+    {
+      "epoch": 0.14877001171417414,
+      "grad_norm": 3.353560209274292,
+      "learning_rate": 9.979949894688108e-06,
+      "loss": 0.7559,
+      "step": 1143
+    },
+    {
+      "epoch": 0.1491604841858649,
+      "grad_norm": 3.719203233718872,
+      "learning_rate": 9.979762492101456e-06,
+      "loss": 0.7475,
+      "step": 1146
+    },
+    {
+      "epoch": 0.14955095665755563,
+      "grad_norm": 3.833672523498535,
+      "learning_rate": 9.979574219560773e-06,
+      "loss": 0.7024,
+      "step": 1149
+    },
+    {
+      "epoch": 0.14994142912924638,
+      "grad_norm": 3.349752426147461,
+      "learning_rate": 9.97938507709895e-06,
+      "loss": 0.681,
+      "step": 1152
+    },
+    {
+      "epoch": 0.15033190160093712,
+      "grad_norm": 3.8509340286254883,
+      "learning_rate": 9.979195064749029e-06,
+      "loss": 0.7446,
+      "step": 1155
+    },
+    {
+      "epoch": 0.15072237407262787,
+      "grad_norm": 3.9146859645843506,
+      "learning_rate": 9.979004182544204e-06,
+      "loss": 0.6898,
+      "step": 1158
+    },
+    {
+      "epoch": 0.1511128465443186,
+      "grad_norm": 3.239340305328369,
+      "learning_rate": 9.978812430517824e-06,
+      "loss": 0.7282,
+      "step": 1161
+    },
+    {
+      "epoch": 0.15150331901600936,
+      "grad_norm": 4.0539727210998535,
+      "learning_rate": 9.978619808703385e-06,
+      "loss": 0.6493,
+      "step": 1164
+    },
+    {
+      "epoch": 0.15189379148770013,
+      "grad_norm": 3.1297805309295654,
+      "learning_rate": 9.978426317134538e-06,
+      "loss": 0.7316,
+      "step": 1167
+    },
+    {
+      "epoch": 0.15228426395939088,
+      "grad_norm": 3.2664248943328857,
+      "learning_rate": 9.978231955845089e-06,
+      "loss": 0.6631,
+      "step": 1170
+    },
+    {
+      "epoch": 0.15267473643108162,
+      "grad_norm": 3.597419023513794,
+      "learning_rate": 9.978036724868989e-06,
+      "loss": 0.661,
+      "step": 1173
+    },
+    {
+      "epoch": 0.15306520890277237,
+      "grad_norm": 2.828190565109253,
+      "learning_rate": 9.977840624240345e-06,
+      "loss": 0.6447,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1534556813744631,
+      "grad_norm": 2.887118339538574,
+      "learning_rate": 9.977643653993415e-06,
+      "loss": 0.7544,
+      "step": 1179
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 4.281754016876221,
+      "learning_rate": 9.977445814162612e-06,
+      "loss": 0.7907,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1542366263178446,
+      "grad_norm": 2.87988018989563,
+      "learning_rate": 9.977247104782496e-06,
+      "loss": 0.6872,
+      "step": 1185
+    },
+    {
+      "epoch": 0.15462709878953534,
+      "grad_norm": 3.0348126888275146,
+      "learning_rate": 9.97704752588778e-06,
+      "loss": 0.6695,
+      "step": 1188
+    },
+    {
+      "epoch": 0.1550175712612261,
+      "grad_norm": 2.9127049446105957,
+      "learning_rate": 9.976847077513331e-06,
+      "loss": 0.6395,
+      "step": 1191
+    },
+    {
+      "epoch": 0.15540804373291683,
+      "grad_norm": 2.811537981033325,
+      "learning_rate": 9.976645759694167e-06,
+      "loss": 0.6382,
+      "step": 1194
+    },
+    {
+      "epoch": 0.15579851620460758,
+      "grad_norm": 3.1080002784729004,
+      "learning_rate": 9.976443572465462e-06,
+      "loss": 0.6556,
+      "step": 1197
+    },
+    {
+      "epoch": 0.15618898867629832,
+      "grad_norm": 3.4639227390289307,
+      "learning_rate": 9.97624051586253e-06,
+      "loss": 0.6893,
+      "step": 1200
+    },
+    {
+      "epoch": 0.15657946114798907,
+      "grad_norm": 3.166987895965576,
+      "learning_rate": 9.97603658992085e-06,
+      "loss": 0.669,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1569699336196798,
+      "grad_norm": 3.154684543609619,
+      "learning_rate": 9.975831794676045e-06,
+      "loss": 0.7224,
+      "step": 1206
+    },
+    {
+      "epoch": 0.15736040609137056,
+      "grad_norm": 4.444057464599609,
+      "learning_rate": 9.975626130163893e-06,
+      "loss": 0.6632,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1577508785630613,
+      "grad_norm": 2.8680520057678223,
+      "learning_rate": 9.975419596420325e-06,
+      "loss": 0.7838,
+      "step": 1212
+    },
+    {
+      "epoch": 0.15814135103475205,
+      "grad_norm": 2.656010866165161,
+      "learning_rate": 9.975212193481419e-06,
+      "loss": 0.6376,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1585318235064428,
+      "grad_norm": 2.9363276958465576,
+      "learning_rate": 9.975003921383409e-06,
+      "loss": 0.7675,
+      "step": 1218
+    },
+    {
+      "epoch": 0.15892229597813354,
+      "grad_norm": 2.935682535171509,
+      "learning_rate": 9.97479478016268e-06,
+      "loss": 0.749,
+      "step": 1221
+    },
+    {
+      "epoch": 0.15931276844982428,
+      "grad_norm": 4.171043395996094,
+      "learning_rate": 9.974584769855768e-06,
+      "loss": 0.64,
+      "step": 1224
+    },
+    {
+      "epoch": 0.15970324092151503,
+      "grad_norm": 3.761404037475586,
+      "learning_rate": 9.974373890499363e-06,
+      "loss": 0.6779,
+      "step": 1227
+    },
+    {
+      "epoch": 0.16009371339320577,
+      "grad_norm": 3.7097084522247314,
+      "learning_rate": 9.974162142130302e-06,
+      "loss": 0.6745,
+      "step": 1230
+    },
+    {
+      "epoch": 0.16048418586489652,
+      "grad_norm": 2.918391704559326,
+      "learning_rate": 9.97394952478558e-06,
+      "loss": 0.6632,
+      "step": 1233
+    },
+    {
+      "epoch": 0.16087465833658726,
+      "grad_norm": 2.916435956954956,
+      "learning_rate": 9.97373603850234e-06,
+      "loss": 0.7137,
+      "step": 1236
+    },
+    {
+      "epoch": 0.161265130808278,
+      "grad_norm": 4.109777450561523,
+      "learning_rate": 9.973521683317877e-06,
+      "loss": 0.6859,
+      "step": 1239
+    },
+    {
+      "epoch": 0.16165560327996875,
+      "grad_norm": 3.2378978729248047,
+      "learning_rate": 9.973306459269639e-06,
+      "loss": 0.6296,
+      "step": 1242
+    },
+    {
+      "epoch": 0.1620460757516595,
+      "grad_norm": 3.034580707550049,
+      "learning_rate": 9.973090366395223e-06,
+      "loss": 0.6648,
+      "step": 1245
+    },
+    {
+      "epoch": 0.16243654822335024,
+      "grad_norm": 2.901724100112915,
+      "learning_rate": 9.972873404732383e-06,
+      "loss": 0.6546,
+      "step": 1248
+    },
+    {
+      "epoch": 0.16282702069504099,
+      "grad_norm": 4.608283519744873,
+      "learning_rate": 9.972655574319022e-06,
+      "loss": 0.6832,
+      "step": 1251
+    },
+    {
+      "epoch": 0.16321749316673176,
+      "grad_norm": 2.890130043029785,
+      "learning_rate": 9.972436875193191e-06,
+      "loss": 0.7566,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1636079656384225,
+      "grad_norm": 2.8199384212493896,
+      "learning_rate": 9.972217307393099e-06,
+      "loss": 0.654,
+      "step": 1257
+    },
+    {
+      "epoch": 0.16399843811011325,
+      "grad_norm": 3.453725576400757,
+      "learning_rate": 9.971996870957104e-06,
+      "loss": 0.6728,
+      "step": 1260
+    },
+    {
+      "epoch": 0.164388910581804,
+      "grad_norm": 2.9947872161865234,
+      "learning_rate": 9.971775565923715e-06,
+      "loss": 0.8324,
+      "step": 1263
+    },
+    {
+      "epoch": 0.16477938305349474,
+      "grad_norm": 4.272884368896484,
+      "learning_rate": 9.971553392331593e-06,
+      "loss": 0.6809,
+      "step": 1266
+    },
+    {
+      "epoch": 0.16516985552518548,
+      "grad_norm": 3.4897780418395996,
+      "learning_rate": 9.971330350219553e-06,
+      "loss": 0.6167,
+      "step": 1269
+    },
+    {
+      "epoch": 0.16556032799687623,
+      "grad_norm": 3.0097784996032715,
+      "learning_rate": 9.971106439626559e-06,
+      "loss": 0.5921,
+      "step": 1272
+    },
+    {
+      "epoch": 0.16595080046856697,
+      "grad_norm": 5.600336074829102,
+      "learning_rate": 9.970881660591727e-06,
+      "loss": 0.6879,
+      "step": 1275
+    },
+    {
+      "epoch": 0.16634127294025772,
+      "grad_norm": 2.6105566024780273,
+      "learning_rate": 9.970656013154326e-06,
+      "loss": 0.651,
+      "step": 1278
+    },
+    {
+      "epoch": 0.16673174541194846,
+      "grad_norm": 2.573819875717163,
+      "learning_rate": 9.970429497353777e-06,
+      "loss": 0.6906,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1671222178836392,
+      "grad_norm": 2.971766233444214,
+      "learning_rate": 9.97020211322965e-06,
+      "loss": 0.6863,
+      "step": 1284
+    },
+    {
+      "epoch": 0.16751269035532995,
+      "grad_norm": 2.8079757690429688,
+      "learning_rate": 9.96997386082167e-06,
+      "loss": 0.6304,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1679031628270207,
+      "grad_norm": 3.1326887607574463,
+      "learning_rate": 9.969744740169713e-06,
+      "loss": 0.6407,
+      "step": 1290
+    },
+    {
+      "epoch": 0.16829363529871144,
+      "grad_norm": 2.8672449588775635,
+      "learning_rate": 9.969514751313803e-06,
+      "loss": 0.764,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1686841077704022,
+      "grad_norm": 2.8906450271606445,
+      "learning_rate": 9.969283894294121e-06,
+      "loss": 0.8143,
+      "step": 1296
+    },
+    {
+      "epoch": 0.16907458024209293,
+      "grad_norm": 2.724889039993286,
+      "learning_rate": 9.969052169150997e-06,
+      "loss": 0.6327,
+      "step": 1299
+    },
+    {
+      "epoch": 0.16946505271378368,
+      "grad_norm": 2.980803966522217,
+      "learning_rate": 9.968819575924911e-06,
+      "loss": 0.6832,
+      "step": 1302
+    },
+    {
+      "epoch": 0.16985552518547442,
+      "grad_norm": 2.6396119594573975,
+      "learning_rate": 9.9685861146565e-06,
+      "loss": 0.6276,
+      "step": 1305
+    },
+    {
+      "epoch": 0.17024599765716517,
+      "grad_norm": 5.2581257820129395,
+      "learning_rate": 9.968351785386545e-06,
+      "loss": 0.6789,
+      "step": 1308
+    },
+    {
+      "epoch": 0.1706364701288559,
+      "grad_norm": 2.702031373977661,
+      "learning_rate": 9.968116588155986e-06,
+      "loss": 0.7108,
+      "step": 1311
+    },
+    {
+      "epoch": 0.17102694260054666,
+      "grad_norm": 2.8001067638397217,
+      "learning_rate": 9.967880523005911e-06,
+      "loss": 0.6184,
+      "step": 1314
+    },
+    {
+      "epoch": 0.1714174150722374,
+      "grad_norm": 2.5163309574127197,
+      "learning_rate": 9.967643589977559e-06,
+      "loss": 0.6862,
+      "step": 1317
+    },
+    {
+      "epoch": 0.17180788754392814,
+      "grad_norm": 3.189988374710083,
+      "learning_rate": 9.967405789112322e-06,
+      "loss": 0.6618,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1721983600156189,
+      "grad_norm": 3.1581954956054688,
+      "learning_rate": 9.967167120451744e-06,
+      "loss": 0.6224,
+      "step": 1323
+    },
+    {
+      "epoch": 0.17258883248730963,
+      "grad_norm": 4.420681953430176,
+      "learning_rate": 9.966927584037518e-06,
+      "loss": 0.7317,
+      "step": 1326
+    },
+    {
+      "epoch": 0.17297930495900038,
+      "grad_norm": 2.627751350402832,
+      "learning_rate": 9.966687179911494e-06,
+      "loss": 0.706,
+      "step": 1329
+    },
+    {
+      "epoch": 0.17336977743069112,
+      "grad_norm": 2.905385971069336,
+      "learning_rate": 9.966445908115668e-06,
+      "loss": 0.6165,
+      "step": 1332
+    },
+    {
+      "epoch": 0.17376024990238187,
+      "grad_norm": 2.9616243839263916,
+      "learning_rate": 9.966203768692189e-06,
+      "loss": 0.619,
+      "step": 1335
+    },
+    {
+      "epoch": 0.17415072237407261,
+      "grad_norm": 3.9424779415130615,
+      "learning_rate": 9.965960761683358e-06,
+      "loss": 0.6366,
+      "step": 1338
+    },
+    {
+      "epoch": 0.1745411948457634,
+      "grad_norm": 2.880329132080078,
+      "learning_rate": 9.965716887131628e-06,
+      "loss": 0.7437,
+      "step": 1341
+    },
+    {
+      "epoch": 0.17493166731745413,
+      "grad_norm": 2.994891405105591,
+      "learning_rate": 9.965472145079606e-06,
+      "loss": 0.7427,
+      "step": 1344
+    },
+    {
+      "epoch": 0.17532213978914488,
+      "grad_norm": 3.0199975967407227,
+      "learning_rate": 9.965226535570047e-06,
+      "loss": 0.6787,
+      "step": 1347
+    },
+    {
+      "epoch": 0.17571261226083562,
+      "grad_norm": 4.0724873542785645,
+      "learning_rate": 9.964980058645856e-06,
+      "loss": 0.6798,
+      "step": 1350
+    },
+    {
+      "epoch": 0.17610308473252637,
+      "grad_norm": 3.5381436347961426,
+      "learning_rate": 9.964732714350093e-06,
+      "loss": 0.6456,
+      "step": 1353
+    },
+    {
+      "epoch": 0.1764935572042171,
+      "grad_norm": 2.9115025997161865,
+      "learning_rate": 9.964484502725972e-06,
+      "loss": 0.6812,
+      "step": 1356
+    },
+    {
+      "epoch": 0.17688402967590786,
+      "grad_norm": 2.9412219524383545,
+      "learning_rate": 9.964235423816851e-06,
+      "loss": 0.6854,
+      "step": 1359
+    },
+    {
+      "epoch": 0.1772745021475986,
+      "grad_norm": 3.002401351928711,
+      "learning_rate": 9.963985477666242e-06,
+      "loss": 0.6684,
+      "step": 1362
+    },
+    {
+      "epoch": 0.17766497461928935,
+      "grad_norm": 5.3304877281188965,
+      "learning_rate": 9.963734664317816e-06,
+      "loss": 0.7724,
+      "step": 1365
+    },
+    {
+      "epoch": 0.1780554470909801,
+      "grad_norm": 3.2181010246276855,
+      "learning_rate": 9.963482983815385e-06,
+      "loss": 0.6604,
+      "step": 1368
+    },
+    {
+      "epoch": 0.17844591956267083,
+      "grad_norm": 3.477856397628784,
+      "learning_rate": 9.963230436202918e-06,
+      "loss": 0.6447,
+      "step": 1371
+    },
+    {
+      "epoch": 0.17883639203436158,
+      "grad_norm": 3.450572967529297,
+      "learning_rate": 9.962977021524535e-06,
+      "loss": 0.5988,
+      "step": 1374
+    },
+    {
+      "epoch": 0.17922686450605232,
+      "grad_norm": 4.131798267364502,
+      "learning_rate": 9.962722739824506e-06,
+      "loss": 0.6665,
+      "step": 1377
+    },
+    {
+      "epoch": 0.17961733697774307,
+      "grad_norm": 3.188093423843384,
+      "learning_rate": 9.962467591147256e-06,
+      "loss": 0.6613,
+      "step": 1380
+    },
+    {
+      "epoch": 0.18000780944943381,
+      "grad_norm": 3.239835023880005,
+      "learning_rate": 9.962211575537357e-06,
+      "loss": 0.6192,
+      "step": 1383
+    },
+    {
+      "epoch": 0.18039828192112456,
+      "grad_norm": 3.1334681510925293,
+      "learning_rate": 9.961954693039535e-06,
+      "loss": 0.6414,
+      "step": 1386
+    },
+    {
+      "epoch": 0.1807887543928153,
+      "grad_norm": 2.906036376953125,
+      "learning_rate": 9.961696943698667e-06,
+      "loss": 0.7304,
+      "step": 1389
+    },
+    {
+      "epoch": 0.18117922686450605,
+      "grad_norm": 4.601681709289551,
+      "learning_rate": 9.961438327559778e-06,
+      "loss": 0.6493,
+      "step": 1392
+    },
+    {
+      "epoch": 0.1815696993361968,
+      "grad_norm": 3.0899908542633057,
+      "learning_rate": 9.961178844668054e-06,
+      "loss": 0.6764,
+      "step": 1395
+    },
+    {
+      "epoch": 0.18196017180788754,
+      "grad_norm": 3.128690242767334,
+      "learning_rate": 9.96091849506882e-06,
+      "loss": 0.7436,
+      "step": 1398
+    },
+    {
+      "epoch": 0.18235064427957828,
+      "grad_norm": 4.097222805023193,
+      "learning_rate": 9.960657278807562e-06,
+      "loss": 0.6236,
+      "step": 1401
+    },
+    {
+      "epoch": 0.18274111675126903,
+      "grad_norm": 2.8826756477355957,
+      "learning_rate": 9.960395195929915e-06,
+      "loss": 0.7283,
+      "step": 1404
+    },
+    {
+      "epoch": 0.18313158922295977,
+      "grad_norm": 2.9129467010498047,
+      "learning_rate": 9.96013224648166e-06,
+      "loss": 0.7447,
+      "step": 1407
+    },
+    {
+      "epoch": 0.18352206169465052,
+      "grad_norm": 3.429777145385742,
+      "learning_rate": 9.959868430508737e-06,
+      "loss": 0.5577,
+      "step": 1410
+    },
+    {
+      "epoch": 0.18391253416634126,
+      "grad_norm": 6.315993785858154,
+      "learning_rate": 9.959603748057234e-06,
+      "loss": 0.5291,
+      "step": 1413
+    },
+    {
+      "epoch": 0.184303006638032,
+      "grad_norm": 2.7502191066741943,
+      "learning_rate": 9.959338199173387e-06,
+      "loss": 0.7087,
+      "step": 1416
+    },
+    {
+      "epoch": 0.18469347910972275,
+      "grad_norm": 3.25874662399292,
+      "learning_rate": 9.959071783903592e-06,
+      "loss": 0.6868,
+      "step": 1419
+    },
+    {
+      "epoch": 0.1850839515814135,
+      "grad_norm": 2.6682872772216797,
+      "learning_rate": 9.958804502294388e-06,
+      "loss": 0.598,
+      "step": 1422
+    },
+    {
+      "epoch": 0.18547442405310427,
+      "grad_norm": 4.8309221267700195,
+      "learning_rate": 9.95853635439247e-06,
+      "loss": 0.7511,
+      "step": 1425
+    },
+    {
+      "epoch": 0.18586489652479501,
+      "grad_norm": 5.207489013671875,
+      "learning_rate": 9.95826734024468e-06,
+      "loss": 0.7661,
+      "step": 1428
+    },
+    {
+      "epoch": 0.18625536899648576,
+      "grad_norm": 2.6510419845581055,
+      "learning_rate": 9.95799745989802e-06,
+      "loss": 0.6397,
+      "step": 1431
+    },
+    {
+      "epoch": 0.1866458414681765,
+      "grad_norm": 3.0544328689575195,
+      "learning_rate": 9.957726713399631e-06,
+      "loss": 0.6224,
+      "step": 1434
+    },
+    {
+      "epoch": 0.18703631393986725,
+      "grad_norm": 2.7472503185272217,
+      "learning_rate": 9.957455100796815e-06,
+      "loss": 0.6854,
+      "step": 1437
+    },
+    {
+      "epoch": 0.187426786411558,
+      "grad_norm": 3.071814775466919,
+      "learning_rate": 9.957182622137022e-06,
+      "loss": 0.6508,
+      "step": 1440
+    },
+    {
+      "epoch": 0.18781725888324874,
+      "grad_norm": 2.8780770301818848,
+      "learning_rate": 9.956909277467854e-06,
+      "loss": 0.6678,
+      "step": 1443
+    },
+    {
+      "epoch": 0.18820773135493948,
+      "grad_norm": 3.1620678901672363,
+      "learning_rate": 9.956635066837062e-06,
+      "loss": 0.6834,
+      "step": 1446
+    },
+    {
+      "epoch": 0.18859820382663023,
+      "grad_norm": 2.79729962348938,
+      "learning_rate": 9.956359990292552e-06,
+      "loss": 0.6528,
+      "step": 1449
+    },
+    {
+      "epoch": 0.18898867629832097,
+      "grad_norm": 2.7644762992858887,
+      "learning_rate": 9.956084047882377e-06,
+      "loss": 0.7051,
+      "step": 1452
+    },
+    {
+      "epoch": 0.18937914877001172,
+      "grad_norm": 3.1195924282073975,
+      "learning_rate": 9.955807239654746e-06,
+      "loss": 0.7568,
+      "step": 1455
+    },
+    {
+      "epoch": 0.18976962124170246,
+      "grad_norm": 2.780245065689087,
+      "learning_rate": 9.955529565658017e-06,
+      "loss": 0.55,
+      "step": 1458
+    },
+    {
+      "epoch": 0.1901600937133932,
+      "grad_norm": 5.350193977355957,
+      "learning_rate": 9.955251025940696e-06,
+      "loss": 0.6525,
+      "step": 1461
+    },
+    {
+      "epoch": 0.19055056618508395,
+      "grad_norm": 2.754183530807495,
+      "learning_rate": 9.954971620551446e-06,
+      "loss": 0.6989,
+      "step": 1464
+    },
+    {
+      "epoch": 0.1909410386567747,
+      "grad_norm": 2.848510980606079,
+      "learning_rate": 9.954691349539076e-06,
+      "loss": 0.7376,
+      "step": 1467
+    },
+    {
+      "epoch": 0.19133151112846544,
+      "grad_norm": 4.006906509399414,
+      "learning_rate": 9.954410212952551e-06,
+      "loss": 0.68,
+      "step": 1470
+    },
+    {
+      "epoch": 0.1917219836001562,
+      "grad_norm": 2.6759417057037354,
+      "learning_rate": 9.954128210840985e-06,
+      "loss": 0.6434,
+      "step": 1473
+    },
+    {
+      "epoch": 0.19211245607184693,
+      "grad_norm": 2.9079430103302,
+      "learning_rate": 9.953845343253643e-06,
+      "loss": 0.6194,
+      "step": 1476
+    },
+    {
+      "epoch": 0.19250292854353768,
+      "grad_norm": 2.6229822635650635,
+      "learning_rate": 9.953561610239941e-06,
+      "loss": 0.6018,
+      "step": 1479
+    },
+    {
+      "epoch": 0.19289340101522842,
+      "grad_norm": 2.8164374828338623,
+      "learning_rate": 9.953277011849444e-06,
+      "loss": 0.6276,
+      "step": 1482
+    },
+    {
+      "epoch": 0.19328387348691917,
+      "grad_norm": 3.2668967247009277,
+      "learning_rate": 9.952991548131876e-06,
+      "loss": 0.722,
+      "step": 1485
+    },
+    {
+      "epoch": 0.1936743459586099,
+      "grad_norm": 2.878713846206665,
+      "learning_rate": 9.952705219137102e-06,
+      "loss": 0.7137,
+      "step": 1488
+    },
+    {
+      "epoch": 0.19406481843030066,
+      "grad_norm": 2.783433437347412,
+      "learning_rate": 9.952418024915146e-06,
+      "loss": 0.6569,
+      "step": 1491
+    },
+    {
+      "epoch": 0.1944552909019914,
+      "grad_norm": 3.105311870574951,
+      "learning_rate": 9.95212996551618e-06,
+      "loss": 0.6587,
+      "step": 1494
+    },
+    {
+      "epoch": 0.19484576337368215,
+      "grad_norm": 6.641456604003906,
+      "learning_rate": 9.951841040990527e-06,
+      "loss": 0.6604,
+      "step": 1497
+    },
+    {
+      "epoch": 0.1952362358453729,
+      "grad_norm": 3.2183666229248047,
+      "learning_rate": 9.951551251388661e-06,
+      "loss": 0.74,
+      "step": 1500
+    },
+    {
+      "epoch": 0.19562670831706364,
+      "grad_norm": 2.9036731719970703,
+      "learning_rate": 9.951260596761208e-06,
+      "loss": 0.6216,
+      "step": 1503
+    },
+    {
+      "epoch": 0.19601718078875438,
+      "grad_norm": 3.435842752456665,
+      "learning_rate": 9.950969077158944e-06,
+      "loss": 0.7316,
+      "step": 1506
+    },
+    {
+      "epoch": 0.19640765326044513,
+      "grad_norm": 2.714198350906372,
+      "learning_rate": 9.950676692632797e-06,
+      "loss": 0.566,
+      "step": 1509
+    },
+    {
+      "epoch": 0.1967981257321359,
+      "grad_norm": 3.2397942543029785,
+      "learning_rate": 9.950383443233848e-06,
+      "loss": 0.702,
+      "step": 1512
+    },
+    {
+      "epoch": 0.19718859820382664,
+      "grad_norm": 2.9357926845550537,
+      "learning_rate": 9.950089329013324e-06,
+      "loss": 0.7299,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1975790706755174,
+      "grad_norm": 4.893215179443359,
+      "learning_rate": 9.949794350022609e-06,
+      "loss": 0.7201,
+      "step": 1518
+    },
+    {
+      "epoch": 0.19796954314720813,
+      "grad_norm": 2.9395291805267334,
+      "learning_rate": 9.949498506313232e-06,
+      "loss": 0.7369,
+      "step": 1521
+    },
+    {
+      "epoch": 0.19836001561889888,
+      "grad_norm": 2.985732078552246,
+      "learning_rate": 9.949201797936882e-06,
+      "loss": 0.7187,
+      "step": 1524
+    },
+    {
+      "epoch": 0.19875048809058962,
+      "grad_norm": 3.205756425857544,
+      "learning_rate": 9.948904224945386e-06,
+      "loss": 0.6903,
+      "step": 1527
+    },
+    {
+      "epoch": 0.19914096056228037,
+      "grad_norm": 3.1247010231018066,
+      "learning_rate": 9.948605787390735e-06,
+      "loss": 0.6934,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1995314330339711,
+      "grad_norm": 2.7591116428375244,
+      "learning_rate": 9.948306485325061e-06,
+      "loss": 0.6708,
+      "step": 1533
+    },
+    {
+      "epoch": 0.19992190550566186,
+      "grad_norm": 3.138434648513794,
+      "learning_rate": 9.948006318800657e-06,
+      "loss": 0.7368,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2003123779773526,
+      "grad_norm": 6.248907089233398,
+      "learning_rate": 9.947705287869956e-06,
+      "loss": 0.8298,
+      "step": 1539
+    },
+    {
+      "epoch": 0.20070285044904335,
+      "grad_norm": 4.321200847625732,
+      "learning_rate": 9.947403392585548e-06,
+      "loss": 0.6983,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2010933229207341,
+      "grad_norm": 2.919874429702759,
+      "learning_rate": 9.947100633000178e-06,
+      "loss": 0.6522,
+      "step": 1545
+    },
+    {
+      "epoch": 0.20148379539242484,
+      "grad_norm": 4.117648601531982,
+      "learning_rate": 9.946797009166732e-06,
+      "loss": 0.5987,
+      "step": 1548
+    },
+    {
+      "epoch": 0.20187426786411558,
+      "grad_norm": 3.4389774799346924,
+      "learning_rate": 9.946492521138258e-06,
+      "loss": 0.7094,
+      "step": 1551
+    },
+    {
+      "epoch": 0.20226474033580633,
+      "grad_norm": 2.9114327430725098,
+      "learning_rate": 9.946187168967944e-06,
+      "loss": 0.7512,
+      "step": 1554
+    },
+    {
+      "epoch": 0.20265521280749707,
+      "grad_norm": 2.9621176719665527,
+      "learning_rate": 9.94588095270914e-06,
+      "loss": 0.6346,
+      "step": 1557
+    },
+    {
+      "epoch": 0.20304568527918782,
+      "grad_norm": 4.908132553100586,
+      "learning_rate": 9.945573872415334e-06,
+      "loss": 0.7007,
+      "step": 1560
+    },
+    {
+      "epoch": 0.20343615775087856,
+      "grad_norm": 3.1218419075012207,
+      "learning_rate": 9.94526592814018e-06,
+      "loss": 0.6411,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2038266302225693,
+      "grad_norm": 2.8320937156677246,
+      "learning_rate": 9.944957119937471e-06,
+      "loss": 0.576,
+      "step": 1566
+    },
+    {
+      "epoch": 0.20421710269426005,
+      "grad_norm": 2.7374327182769775,
+      "learning_rate": 9.944647447861154e-06,
+      "loss": 0.5977,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2046075751659508,
+      "grad_norm": 3.095136880874634,
+      "learning_rate": 9.944336911965332e-06,
+      "loss": 0.6583,
+      "step": 1572
+    },
+    {
+      "epoch": 0.20499804763764154,
+      "grad_norm": 3.028735399246216,
+      "learning_rate": 9.944025512304251e-06,
+      "loss": 0.6727,
+      "step": 1575
+    },
+    {
+      "epoch": 0.20538852010933228,
+      "grad_norm": 4.234666347503662,
+      "learning_rate": 9.943713248932314e-06,
+      "loss": 0.6762,
+      "step": 1578
+    },
+    {
+      "epoch": 0.20577899258102303,
+      "grad_norm": 2.7858617305755615,
+      "learning_rate": 9.943400121904074e-06,
+      "loss": 0.7175,
+      "step": 1581
+    },
+    {
+      "epoch": 0.20616946505271377,
+      "grad_norm": 2.894981861114502,
+      "learning_rate": 9.943086131274231e-06,
+      "loss": 0.6666,
+      "step": 1584
+    },
+    {
+      "epoch": 0.20655993752440452,
+      "grad_norm": 2.6675071716308594,
+      "learning_rate": 9.94277127709764e-06,
+      "loss": 0.6473,
+      "step": 1587
+    },
+    {
+      "epoch": 0.20695040999609526,
+      "grad_norm": 2.9758174419403076,
+      "learning_rate": 9.942455559429304e-06,
+      "loss": 0.6711,
+      "step": 1590
+    },
+    {
+      "epoch": 0.207340882467786,
+      "grad_norm": 4.3356781005859375,
+      "learning_rate": 9.94213897832438e-06,
+      "loss": 0.6205,
+      "step": 1593
+    },
+    {
+      "epoch": 0.20773135493947675,
+      "grad_norm": 3.1128265857696533,
+      "learning_rate": 9.941821533838172e-06,
+      "loss": 0.677,
+      "step": 1596
+    },
+    {
+      "epoch": 0.20812182741116753,
+      "grad_norm": 3.0443339347839355,
+      "learning_rate": 9.941503226026139e-06,
+      "loss": 0.7659,
+      "step": 1599
+    },
+    {
+      "epoch": 0.20851229988285827,
+      "grad_norm": 3.6503114700317383,
+      "learning_rate": 9.941184054943888e-06,
+      "loss": 0.6364,
+      "step": 1602
+    },
+    {
+      "epoch": 0.20890277235454902,
+      "grad_norm": 3.119065761566162,
+      "learning_rate": 9.940864020647178e-06,
+      "loss": 0.6696,
+      "step": 1605
+    },
+    {
+      "epoch": 0.20929324482623976,
+      "grad_norm": 2.870157241821289,
+      "learning_rate": 9.940543123191916e-06,
+      "loss": 0.7786,
+      "step": 1608
+    },
+    {
+      "epoch": 0.2096837172979305,
+      "grad_norm": 2.8346638679504395,
+      "learning_rate": 9.940221362634165e-06,
+      "loss": 0.7752,
+      "step": 1611
+    },
+    {
+      "epoch": 0.21007418976962125,
+      "grad_norm": 2.9750826358795166,
+      "learning_rate": 9.939898739030135e-06,
+      "loss": 0.6307,
+      "step": 1614
+    },
+    {
+      "epoch": 0.210464662241312,
+      "grad_norm": 3.138920307159424,
+      "learning_rate": 9.939575252436186e-06,
+      "loss": 0.8262,
+      "step": 1617
+    },
+    {
+      "epoch": 0.21085513471300274,
+      "grad_norm": 2.8036742210388184,
+      "learning_rate": 9.939250902908832e-06,
+      "loss": 0.801,
+      "step": 1620
+    },
+    {
+      "epoch": 0.21124560718469348,
+      "grad_norm": 2.724499225616455,
+      "learning_rate": 9.938925690504737e-06,
+      "loss": 0.65,
+      "step": 1623
+    },
+    {
+      "epoch": 0.21163607965638423,
+      "grad_norm": 3.0661230087280273,
+      "learning_rate": 9.938599615280713e-06,
+      "loss": 0.7127,
+      "step": 1626
+    },
+    {
+      "epoch": 0.21202655212807497,
+      "grad_norm": 2.5042243003845215,
+      "learning_rate": 9.938272677293727e-06,
+      "loss": 0.6147,
+      "step": 1629
+    },
+    {
+      "epoch": 0.21241702459976572,
+      "grad_norm": 3.2207114696502686,
+      "learning_rate": 9.937944876600891e-06,
+      "loss": 0.7055,
+      "step": 1632
+    },
+    {
+      "epoch": 0.21280749707145646,
+      "grad_norm": 3.2957639694213867,
+      "learning_rate": 9.937616213259474e-06,
+      "loss": 0.7794,
+      "step": 1635
+    },
+    {
+      "epoch": 0.2131979695431472,
+      "grad_norm": 3.5852959156036377,
+      "learning_rate": 9.93728668732689e-06,
+      "loss": 0.6852,
+      "step": 1638
+    },
+    {
+      "epoch": 0.21358844201483795,
+      "grad_norm": 3.0781290531158447,
+      "learning_rate": 9.936956298860711e-06,
+      "loss": 0.7384,
+      "step": 1641
+    },
+    {
+      "epoch": 0.2139789144865287,
+      "grad_norm": 2.6453282833099365,
+      "learning_rate": 9.93662504791865e-06,
+      "loss": 0.6695,
+      "step": 1644
+    },
+    {
+      "epoch": 0.21436938695821944,
+      "grad_norm": 3.183838367462158,
+      "learning_rate": 9.93629293455858e-06,
+      "loss": 0.7207,
+      "step": 1647
+    },
+    {
+      "epoch": 0.2147598594299102,
+      "grad_norm": 2.969679832458496,
+      "learning_rate": 9.935959958838519e-06,
+      "loss": 0.6558,
+      "step": 1650
+    },
+    {
+      "epoch": 0.21515033190160093,
+      "grad_norm": 2.812744617462158,
+      "learning_rate": 9.935626120816636e-06,
+      "loss": 0.6737,
+      "step": 1653
+    },
+    {
+      "epoch": 0.21554080437329168,
+      "grad_norm": 2.789893627166748,
+      "learning_rate": 9.935291420551252e-06,
+      "loss": 0.6972,
+      "step": 1656
+    },
+    {
+      "epoch": 0.21593127684498242,
+      "grad_norm": 2.601032018661499,
+      "learning_rate": 9.934955858100838e-06,
+      "loss": 0.6458,
+      "step": 1659
+    },
+    {
+      "epoch": 0.21632174931667317,
+      "grad_norm": 3.639371395111084,
+      "learning_rate": 9.93461943352402e-06,
+      "loss": 0.7118,
+      "step": 1662
+    },
+    {
+      "epoch": 0.2167122217883639,
+      "grad_norm": 3.122035264968872,
+      "learning_rate": 9.934282146879568e-06,
+      "loss": 0.7283,
+      "step": 1665
+    },
+    {
+      "epoch": 0.21710269426005466,
+      "grad_norm": 2.9482860565185547,
+      "learning_rate": 9.933943998226403e-06,
+      "loss": 0.6477,
+      "step": 1668
+    },
+    {
+      "epoch": 0.2174931667317454,
+      "grad_norm": 2.530093193054199,
+      "learning_rate": 9.933604987623603e-06,
+      "loss": 0.6032,
+      "step": 1671
+    },
+    {
+      "epoch": 0.21788363920343615,
+      "grad_norm": 2.5820152759552,
+      "learning_rate": 9.93326511513039e-06,
+      "loss": 0.6156,
+      "step": 1674
+    },
+    {
+      "epoch": 0.2182741116751269,
+      "grad_norm": 3.059326410293579,
+      "learning_rate": 9.93292438080614e-06,
+      "loss": 0.6979,
+      "step": 1677
+    },
+    {
+      "epoch": 0.21866458414681764,
+      "grad_norm": 3.3234145641326904,
+      "learning_rate": 9.932582784710377e-06,
+      "loss": 0.6619,
+      "step": 1680
+    },
+    {
+      "epoch": 0.21905505661850838,
+      "grad_norm": 2.865112066268921,
+      "learning_rate": 9.932240326902777e-06,
+      "loss": 0.6367,
+      "step": 1683
+    },
+    {
+      "epoch": 0.21944552909019915,
+      "grad_norm": 2.950484037399292,
+      "learning_rate": 9.93189700744317e-06,
+      "loss": 0.7528,
+      "step": 1686
+    },
+    {
+      "epoch": 0.2198360015618899,
+      "grad_norm": 2.690485715866089,
+      "learning_rate": 9.931552826391529e-06,
+      "loss": 0.6232,
+      "step": 1689
+    },
+    {
+      "epoch": 0.22022647403358064,
+      "grad_norm": 2.7388739585876465,
+      "learning_rate": 9.931207783807984e-06,
+      "loss": 0.7912,
+      "step": 1692
+    },
+    {
+      "epoch": 0.2206169465052714,
+      "grad_norm": 2.996173620223999,
+      "learning_rate": 9.930861879752814e-06,
+      "loss": 0.7042,
+      "step": 1695
+    },
+    {
+      "epoch": 0.22100741897696213,
+      "grad_norm": 3.6289913654327393,
+      "learning_rate": 9.930515114286446e-06,
+      "loss": 0.713,
+      "step": 1698
+    },
+    {
+      "epoch": 0.22139789144865288,
+      "grad_norm": 2.484739065170288,
+      "learning_rate": 9.93016748746946e-06,
+      "loss": 0.7148,
+      "step": 1701
+    },
+    {
+      "epoch": 0.22178836392034362,
+      "grad_norm": 3.885227918624878,
+      "learning_rate": 9.929818999362585e-06,
+      "loss": 0.5977,
+      "step": 1704
+    },
+    {
+      "epoch": 0.22217883639203437,
+      "grad_norm": 3.2628114223480225,
+      "learning_rate": 9.929469650026705e-06,
+      "loss": 0.6511,
+      "step": 1707
+    },
+    {
+      "epoch": 0.2225693088637251,
+      "grad_norm": 2.676222085952759,
+      "learning_rate": 9.929119439522843e-06,
+      "loss": 0.7086,
+      "step": 1710
+    },
+    {
+      "epoch": 0.22295978133541586,
+      "grad_norm": 2.7704615592956543,
+      "learning_rate": 9.928768367912186e-06,
+      "loss": 0.6214,
+      "step": 1713
+    },
+    {
+      "epoch": 0.2233502538071066,
+      "grad_norm": 2.829817295074463,
+      "learning_rate": 9.928416435256062e-06,
+      "loss": 0.7286,
+      "step": 1716
+    },
+    {
+      "epoch": 0.22374072627879735,
+      "grad_norm": 4.687877178192139,
+      "learning_rate": 9.928063641615958e-06,
+      "loss": 0.6466,
+      "step": 1719
+    },
+    {
+      "epoch": 0.2241311987504881,
+      "grad_norm": 2.777306079864502,
+      "learning_rate": 9.9277099870535e-06,
+      "loss": 0.7183,
+      "step": 1722
+    },
+    {
+      "epoch": 0.22452167122217884,
+      "grad_norm": 3.198352575302124,
+      "learning_rate": 9.927355471630475e-06,
+      "loss": 0.6335,
+      "step": 1725
+    },
+    {
+      "epoch": 0.22491214369386958,
+      "grad_norm": 3.063441276550293,
+      "learning_rate": 9.927000095408814e-06,
+      "loss": 0.7046,
+      "step": 1728
+    },
+    {
+      "epoch": 0.22530261616556033,
+      "grad_norm": 2.7282402515411377,
+      "learning_rate": 9.926643858450602e-06,
+      "loss": 0.6636,
+      "step": 1731
+    },
+    {
+      "epoch": 0.22569308863725107,
+      "grad_norm": 3.712812900543213,
+      "learning_rate": 9.926286760818072e-06,
+      "loss": 0.763,
+      "step": 1734
+    },
+    {
+      "epoch": 0.22608356110894182,
+      "grad_norm": 2.7162492275238037,
+      "learning_rate": 9.925928802573608e-06,
+      "loss": 0.7452,
+      "step": 1737
+    },
+    {
+      "epoch": 0.22647403358063256,
+      "grad_norm": 2.689542770385742,
+      "learning_rate": 9.925569983779744e-06,
+      "loss": 0.6457,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2268645060523233,
+      "grad_norm": 3.449857234954834,
+      "learning_rate": 9.925210304499168e-06,
+      "loss": 0.6226,
+      "step": 1743
+    },
+    {
+      "epoch": 0.22725497852401405,
+      "grad_norm": 3.3526997566223145,
+      "learning_rate": 9.92484976479471e-06,
+      "loss": 0.6801,
+      "step": 1746
+    },
+    {
+      "epoch": 0.2276454509957048,
+      "grad_norm": 3.0590097904205322,
+      "learning_rate": 9.924488364729362e-06,
+      "loss": 0.7149,
+      "step": 1749
+    },
+    {
+      "epoch": 0.22803592346739554,
+      "grad_norm": 4.266837120056152,
+      "learning_rate": 9.924126104366255e-06,
+      "loss": 0.6816,
+      "step": 1752
+    },
+    {
+      "epoch": 0.22842639593908629,
+      "grad_norm": 2.6770737171173096,
+      "learning_rate": 9.923762983768674e-06,
+      "loss": 0.7303,
+      "step": 1755
+    },
+    {
+      "epoch": 0.22881686841077703,
+      "grad_norm": 2.922563314437866,
+      "learning_rate": 9.92339900300006e-06,
+      "loss": 0.6389,
+      "step": 1758
+    },
+    {
+      "epoch": 0.22920734088246777,
+      "grad_norm": 3.8620433807373047,
+      "learning_rate": 9.923034162123996e-06,
+      "loss": 0.6251,
+      "step": 1761
+    },
+    {
+      "epoch": 0.22959781335415852,
+      "grad_norm": 3.186556339263916,
+      "learning_rate": 9.922668461204222e-06,
+      "loss": 0.8476,
+      "step": 1764
+    },
+    {
+      "epoch": 0.22998828582584926,
+      "grad_norm": 3.0054855346679688,
+      "learning_rate": 9.922301900304622e-06,
+      "loss": 0.6719,
+      "step": 1767
+    },
+    {
+      "epoch": 0.23037875829754004,
+      "grad_norm": 3.260510206222534,
+      "learning_rate": 9.921934479489236e-06,
+      "loss": 0.6796,
+      "step": 1770
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 3.021167755126953,
+      "learning_rate": 9.921566198822252e-06,
+      "loss": 0.667,
+      "step": 1773
+    },
+    {
+      "epoch": 0.23115970324092153,
+      "grad_norm": 3.6485886573791504,
+      "learning_rate": 9.921197058368005e-06,
+      "loss": 0.7357,
+      "step": 1776
+    },
+    {
+      "epoch": 0.23155017571261227,
+      "grad_norm": 2.8014872074127197,
+      "learning_rate": 9.920827058190984e-06,
+      "loss": 0.8078,
+      "step": 1779
+    },
+    {
+      "epoch": 0.23194064818430302,
+      "grad_norm": 2.7139265537261963,
+      "learning_rate": 9.92045619835583e-06,
+      "loss": 0.7207,
+      "step": 1782
+    },
+    {
+      "epoch": 0.23233112065599376,
+      "grad_norm": 2.5168399810791016,
+      "learning_rate": 9.920084478927327e-06,
+      "loss": 0.5873,
+      "step": 1785
+    },
+    {
+      "epoch": 0.2327215931276845,
+      "grad_norm": 3.7160627841949463,
+      "learning_rate": 9.919711899970417e-06,
+      "loss": 0.7382,
+      "step": 1788
+    },
+    {
+      "epoch": 0.23311206559937525,
+      "grad_norm": 2.766188383102417,
+      "learning_rate": 9.919338461550188e-06,
+      "loss": 0.6981,
+      "step": 1791
+    },
+    {
+      "epoch": 0.233502538071066,
+      "grad_norm": 3.4832241535186768,
+      "learning_rate": 9.918964163731878e-06,
+      "loss": 0.6844,
+      "step": 1794
+    },
+    {
+      "epoch": 0.23389301054275674,
+      "grad_norm": 2.998603582382202,
+      "learning_rate": 9.918589006580877e-06,
+      "loss": 0.6986,
+      "step": 1797
+    },
+    {
+      "epoch": 0.23428348301444749,
+      "grad_norm": 2.618968963623047,
+      "learning_rate": 9.918212990162724e-06,
+      "loss": 0.6967,
+      "step": 1800
+    },
+    {
+      "epoch": 0.23467395548613823,
+      "grad_norm": 3.121068000793457,
+      "learning_rate": 9.917836114543105e-06,
+      "loss": 0.5803,
+      "step": 1803
+    },
+    {
+      "epoch": 0.23506442795782898,
+      "grad_norm": 2.8870410919189453,
+      "learning_rate": 9.917458379787865e-06,
+      "loss": 0.66,
+      "step": 1806
+    },
+    {
+      "epoch": 0.23545490042951972,
+      "grad_norm": 3.4096627235412598,
+      "learning_rate": 9.917079785962991e-06,
+      "loss": 0.644,
+      "step": 1809
+    },
+    {
+      "epoch": 0.23584537290121046,
+      "grad_norm": 3.3763647079467773,
+      "learning_rate": 9.916700333134622e-06,
+      "loss": 0.7743,
+      "step": 1812
+    },
+    {
+      "epoch": 0.2362358453729012,
+      "grad_norm": 3.2882561683654785,
+      "learning_rate": 9.916320021369049e-06,
+      "loss": 0.7217,
+      "step": 1815
+    },
+    {
+      "epoch": 0.23662631784459195,
+      "grad_norm": 2.849679946899414,
+      "learning_rate": 9.91593885073271e-06,
+      "loss": 0.6473,
+      "step": 1818
+    },
+    {
+      "epoch": 0.2370167903162827,
+      "grad_norm": 2.5509214401245117,
+      "learning_rate": 9.915556821292194e-06,
+      "loss": 0.6598,
+      "step": 1821
+    },
+    {
+      "epoch": 0.23740726278797344,
+      "grad_norm": 4.192689895629883,
+      "learning_rate": 9.915173933114243e-06,
+      "loss": 0.737,
+      "step": 1824
+    },
+    {
+      "epoch": 0.2377977352596642,
+      "grad_norm": 3.805640459060669,
+      "learning_rate": 9.914790186265747e-06,
+      "loss": 0.6514,
+      "step": 1827
+    },
+    {
+      "epoch": 0.23818820773135493,
+      "grad_norm": 2.7506203651428223,
+      "learning_rate": 9.914405580813744e-06,
+      "loss": 0.6088,
+      "step": 1830
+    },
+    {
+      "epoch": 0.23857868020304568,
+      "grad_norm": 2.9429638385772705,
+      "learning_rate": 9.914020116825425e-06,
+      "loss": 0.6348,
+      "step": 1833
+    },
+    {
+      "epoch": 0.23896915267473642,
+      "grad_norm": 2.8457984924316406,
+      "learning_rate": 9.913633794368128e-06,
+      "loss": 0.6043,
+      "step": 1836
+    },
+    {
+      "epoch": 0.23935962514642717,
+      "grad_norm": 3.119004011154175,
+      "learning_rate": 9.913246613509344e-06,
+      "loss": 0.6326,
+      "step": 1839
+    },
+    {
+      "epoch": 0.2397500976181179,
+      "grad_norm": 2.674212694168091,
+      "learning_rate": 9.912858574316714e-06,
+      "loss": 0.6505,
+      "step": 1842
+    },
+    {
+      "epoch": 0.24014057008980866,
+      "grad_norm": 4.301478862762451,
+      "learning_rate": 9.912469676858025e-06,
+      "loss": 0.6864,
+      "step": 1845
+    },
+    {
+      "epoch": 0.2405310425614994,
+      "grad_norm": 3.199939727783203,
+      "learning_rate": 9.912079921201216e-06,
+      "loss": 0.716,
+      "step": 1848
+    },
+    {
+      "epoch": 0.24092151503319015,
+      "grad_norm": 3.2009308338165283,
+      "learning_rate": 9.911689307414381e-06,
+      "loss": 0.7442,
+      "step": 1851
+    },
+    {
+      "epoch": 0.2413119875048809,
+      "grad_norm": 4.463657855987549,
+      "learning_rate": 9.911297835565755e-06,
+      "loss": 0.6488,
+      "step": 1854
+    },
+    {
+      "epoch": 0.24170245997657167,
+      "grad_norm": 2.5389511585235596,
+      "learning_rate": 9.91090550572373e-06,
+      "loss": 0.6598,
+      "step": 1857
+    },
+    {
+      "epoch": 0.2420929324482624,
+      "grad_norm": 2.7800350189208984,
+      "learning_rate": 9.910512317956845e-06,
+      "loss": 0.6538,
+      "step": 1860
+    },
+    {
+      "epoch": 0.24248340491995315,
+      "grad_norm": 4.617105007171631,
+      "learning_rate": 9.910118272333787e-06,
+      "loss": 0.6534,
+      "step": 1863
+    },
+    {
+      "epoch": 0.2428738773916439,
+      "grad_norm": 4.014925479888916,
+      "learning_rate": 9.909723368923397e-06,
+      "loss": 0.7108,
+      "step": 1866
+    },
+    {
+      "epoch": 0.24326434986333464,
+      "grad_norm": 3.7771573066711426,
+      "learning_rate": 9.909327607794663e-06,
+      "loss": 0.7403,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2436548223350254,
+      "grad_norm": 4.131522178649902,
+      "learning_rate": 9.908930989016723e-06,
+      "loss": 0.6846,
+      "step": 1872
+    },
+    {
+      "epoch": 0.24404529480671613,
+      "grad_norm": 2.5078864097595215,
+      "learning_rate": 9.908533512658867e-06,
+      "loss": 0.7045,
+      "step": 1875
+    },
+    {
+      "epoch": 0.24443576727840688,
+      "grad_norm": 3.6496095657348633,
+      "learning_rate": 9.90813517879053e-06,
+      "loss": 0.6804,
+      "step": 1878
+    },
+    {
+      "epoch": 0.24482623975009762,
+      "grad_norm": 3.153731107711792,
+      "learning_rate": 9.907735987481306e-06,
+      "loss": 0.77,
+      "step": 1881
+    },
+    {
+      "epoch": 0.24521671222178837,
+      "grad_norm": 3.234714984893799,
+      "learning_rate": 9.907335938800925e-06,
+      "loss": 0.6396,
+      "step": 1884
+    },
+    {
+      "epoch": 0.2456071846934791,
+      "grad_norm": 4.209356784820557,
+      "learning_rate": 9.906935032819283e-06,
+      "loss": 0.6937,
+      "step": 1887
+    },
+    {
+      "epoch": 0.24599765716516986,
+      "grad_norm": 2.8862154483795166,
+      "learning_rate": 9.906533269606412e-06,
+      "loss": 0.7209,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2463881296368606,
+      "grad_norm": 3.8933069705963135,
+      "learning_rate": 9.9061306492325e-06,
+      "loss": 0.7016,
+      "step": 1893
+    },
+    {
+      "epoch": 0.24677860210855135,
+      "grad_norm": 3.536663770675659,
+      "learning_rate": 9.905727171767885e-06,
+      "loss": 0.6479,
+      "step": 1896
+    },
+    {
+      "epoch": 0.2471690745802421,
+      "grad_norm": 2.742396354675293,
+      "learning_rate": 9.905322837283054e-06,
+      "loss": 0.7213,
+      "step": 1899
+    },
+    {
+      "epoch": 0.24755954705193284,
+      "grad_norm": 3.2880959510803223,
+      "learning_rate": 9.904917645848642e-06,
+      "loss": 0.7038,
+      "step": 1902
+    },
+    {
+      "epoch": 0.24795001952362358,
+      "grad_norm": 2.86783504486084,
+      "learning_rate": 9.904511597535435e-06,
+      "loss": 0.7333,
+      "step": 1905
+    },
+    {
+      "epoch": 0.24834049199531433,
+      "grad_norm": 2.66743803024292,
+      "learning_rate": 9.904104692414372e-06,
+      "loss": 0.6619,
+      "step": 1908
+    },
+    {
+      "epoch": 0.24873096446700507,
+      "grad_norm": 2.8109350204467773,
+      "learning_rate": 9.903696930556534e-06,
+      "loss": 0.6984,
+      "step": 1911
+    },
+    {
+      "epoch": 0.24912143693869582,
+      "grad_norm": 4.756025791168213,
+      "learning_rate": 9.903288312033158e-06,
+      "loss": 0.6782,
+      "step": 1914
+    },
+    {
+      "epoch": 0.24951190941038656,
+      "grad_norm": 3.219858169555664,
+      "learning_rate": 9.902878836915628e-06,
+      "loss": 0.7427,
+      "step": 1917
+    },
+    {
+      "epoch": 0.2499023818820773,
+      "grad_norm": 2.6937522888183594,
+      "learning_rate": 9.902468505275481e-06,
+      "loss": 0.6722,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2502928543537681,
+      "grad_norm": 3.728074073791504,
+      "learning_rate": 9.9020573171844e-06,
+      "loss": 0.667,
+      "step": 1923
+    },
+    {
+      "epoch": 0.2506833268254588,
+      "grad_norm": 2.520673990249634,
+      "learning_rate": 9.901645272714216e-06,
+      "loss": 0.6062,
+      "step": 1926
+    },
+    {
+      "epoch": 0.25107379929714957,
+      "grad_norm": 3.0606751441955566,
+      "learning_rate": 9.901232371936916e-06,
+      "loss": 0.6204,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2514642717688403,
+      "grad_norm": 3.0006275177001953,
+      "learning_rate": 9.90081861492463e-06,
+      "loss": 0.7612,
+      "step": 1932
+    },
+    {
+      "epoch": 0.25185474424053106,
+      "grad_norm": 3.9713451862335205,
+      "learning_rate": 9.900404001749643e-06,
+      "loss": 0.7894,
+      "step": 1935
+    },
+    {
+      "epoch": 0.2522452167122218,
+      "grad_norm": 2.761209726333618,
+      "learning_rate": 9.899988532484386e-06,
+      "loss": 0.6158,
+      "step": 1938
+    },
+    {
+      "epoch": 0.25263568918391255,
+      "grad_norm": 2.8086750507354736,
+      "learning_rate": 9.89957220720144e-06,
+      "loss": 0.6952,
+      "step": 1941
+    },
+    {
+      "epoch": 0.25302616165560327,
+      "grad_norm": 5.129540920257568,
+      "learning_rate": 9.899155025973535e-06,
+      "loss": 0.6314,
+      "step": 1944
+    },
+    {
+      "epoch": 0.25341663412729404,
+      "grad_norm": 2.605057954788208,
+      "learning_rate": 9.898736988873555e-06,
+      "loss": 0.7729,
+      "step": 1947
+    },
+    {
+      "epoch": 0.25380710659898476,
+      "grad_norm": 4.0138044357299805,
+      "learning_rate": 9.898318095974529e-06,
+      "loss": 0.7074,
+      "step": 1950
+    },
+    {
+      "epoch": 0.25419757907067553,
+      "grad_norm": 2.733633041381836,
+      "learning_rate": 9.897898347349635e-06,
+      "loss": 0.6754,
+      "step": 1953
+    },
+    {
+      "epoch": 0.25458805154236624,
+      "grad_norm": 2.873657703399658,
+      "learning_rate": 9.897477743072203e-06,
+      "loss": 0.6609,
+      "step": 1956
+    },
+    {
+      "epoch": 0.254978524014057,
+      "grad_norm": 3.4984524250030518,
+      "learning_rate": 9.897056283215713e-06,
+      "loss": 0.6305,
+      "step": 1959
+    },
+    {
+      "epoch": 0.25536899648574773,
+      "grad_norm": 2.7125892639160156,
+      "learning_rate": 9.896633967853793e-06,
+      "loss": 0.6463,
+      "step": 1962
+    },
+    {
+      "epoch": 0.2557594689574385,
+      "grad_norm": 2.670562505722046,
+      "learning_rate": 9.896210797060218e-06,
+      "loss": 0.6669,
+      "step": 1965
+    },
+    {
+      "epoch": 0.2561499414291292,
+      "grad_norm": 3.608865261077881,
+      "learning_rate": 9.895786770908918e-06,
+      "loss": 0.6898,
+      "step": 1968
+    },
+    {
+      "epoch": 0.25654041390082,
+      "grad_norm": 3.5786221027374268,
+      "learning_rate": 9.895361889473969e-06,
+      "loss": 0.7525,
+      "step": 1971
+    },
+    {
+      "epoch": 0.2569308863725107,
+      "grad_norm": 3.3023688793182373,
+      "learning_rate": 9.894936152829595e-06,
+      "loss": 0.7672,
+      "step": 1974
+    },
+    {
+      "epoch": 0.2573213588442015,
+      "grad_norm": 2.93623948097229,
+      "learning_rate": 9.894509561050173e-06,
+      "loss": 0.7114,
+      "step": 1977
+    },
+    {
+      "epoch": 0.2577118313158922,
+      "grad_norm": 3.6028199195861816,
+      "learning_rate": 9.894082114210226e-06,
+      "loss": 0.6342,
+      "step": 1980
+    },
+    {
+      "epoch": 0.258102303787583,
+      "grad_norm": 3.522853374481201,
+      "learning_rate": 9.893653812384432e-06,
+      "loss": 0.6205,
+      "step": 1983
+    },
+    {
+      "epoch": 0.25849277625927375,
+      "grad_norm": 2.625506639480591,
+      "learning_rate": 9.893224655647609e-06,
+      "loss": 0.6737,
+      "step": 1986
+    },
+    {
+      "epoch": 0.25888324873096447,
+      "grad_norm": 3.548536777496338,
+      "learning_rate": 9.892794644074735e-06,
+      "loss": 0.6586,
+      "step": 1989
+    },
+    {
+      "epoch": 0.25927372120265524,
+      "grad_norm": 4.54545783996582,
+      "learning_rate": 9.892363777740928e-06,
+      "loss": 0.7415,
+      "step": 1992
+    },
+    {
+      "epoch": 0.25966419367434596,
+      "grad_norm": 2.687030792236328,
+      "learning_rate": 9.89193205672146e-06,
+      "loss": 0.5988,
+      "step": 1995
+    },
+    {
+      "epoch": 0.26005466614603673,
+      "grad_norm": 4.913947582244873,
+      "learning_rate": 9.891499481091755e-06,
+      "loss": 0.707,
+      "step": 1998
+    },
+    {
+      "epoch": 0.26044513861772745,
+      "grad_norm": 2.828240394592285,
+      "learning_rate": 9.891066050927381e-06,
+      "loss": 0.6659,
+      "step": 2001
+    },
+    {
+      "epoch": 0.2608356110894182,
+      "grad_norm": 2.456120014190674,
+      "learning_rate": 9.890631766304054e-06,
+      "loss": 0.743,
+      "step": 2004
+    },
+    {
+      "epoch": 0.26122608356110893,
+      "grad_norm": 3.4322896003723145,
+      "learning_rate": 9.890196627297649e-06,
+      "loss": 0.6586,
+      "step": 2007
+    },
+    {
+      "epoch": 0.2616165560327997,
+      "grad_norm": 2.4019174575805664,
+      "learning_rate": 9.88976063398418e-06,
+      "loss": 0.6277,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2620070285044904,
+      "grad_norm": 2.6000640392303467,
+      "learning_rate": 9.889323786439815e-06,
+      "loss": 0.6874,
+      "step": 2013
+    },
+    {
+      "epoch": 0.2623975009761812,
+      "grad_norm": 2.7879741191864014,
+      "learning_rate": 9.88888608474087e-06,
+      "loss": 0.6844,
+      "step": 2016
+    },
+    {
+      "epoch": 0.2627879734478719,
+      "grad_norm": 2.620607852935791,
+      "learning_rate": 9.888447528963809e-06,
+      "loss": 0.6139,
+      "step": 2019
+    },
+    {
+      "epoch": 0.2631784459195627,
+      "grad_norm": 3.1650354862213135,
+      "learning_rate": 9.88800811918525e-06,
+      "loss": 0.6203,
+      "step": 2022
+    },
+    {
+      "epoch": 0.2635689183912534,
+      "grad_norm": 2.8393919467926025,
+      "learning_rate": 9.887567855481955e-06,
+      "loss": 0.6931,
+      "step": 2025
+    },
+    {
+      "epoch": 0.2639593908629442,
+      "grad_norm": 3.6173481941223145,
+      "learning_rate": 9.88712673793084e-06,
+      "loss": 0.6675,
+      "step": 2028
+    },
+    {
+      "epoch": 0.2643498633346349,
+      "grad_norm": 3.9958341121673584,
+      "learning_rate": 9.88668476660896e-06,
+      "loss": 0.6857,
+      "step": 2031
+    },
+    {
+      "epoch": 0.26474033580632567,
+      "grad_norm": 2.9784352779388428,
+      "learning_rate": 9.886241941593535e-06,
+      "loss": 0.7343,
+      "step": 2034
+    },
+    {
+      "epoch": 0.2651308082780164,
+      "grad_norm": 3.043720006942749,
+      "learning_rate": 9.885798262961921e-06,
+      "loss": 0.6138,
+      "step": 2037
+    },
+    {
+      "epoch": 0.26552128074970716,
+      "grad_norm": 2.8800082206726074,
+      "learning_rate": 9.88535373079163e-06,
+      "loss": 0.7174,
+      "step": 2040
+    },
+    {
+      "epoch": 0.2659117532213979,
+      "grad_norm": 2.96820330619812,
+      "learning_rate": 9.884908345160318e-06,
+      "loss": 0.6844,
+      "step": 2043
+    },
+    {
+      "epoch": 0.26630222569308865,
+      "grad_norm": 4.006218910217285,
+      "learning_rate": 9.884462106145794e-06,
+      "loss": 0.5799,
+      "step": 2046
+    },
+    {
+      "epoch": 0.26669269816477936,
+      "grad_norm": 4.129579544067383,
+      "learning_rate": 9.884015013826015e-06,
+      "loss": 0.7025,
+      "step": 2049
+    },
+    {
+      "epoch": 0.26708317063647014,
+      "grad_norm": 6.496399879455566,
+      "learning_rate": 9.88356706827909e-06,
+      "loss": 0.6598,
+      "step": 2052
+    },
+    {
+      "epoch": 0.26747364310816085,
+      "grad_norm": 10.179603576660156,
+      "learning_rate": 9.88311826958327e-06,
+      "loss": 0.7294,
+      "step": 2055
+    },
+    {
+      "epoch": 0.2678641155798516,
+      "grad_norm": 4.363855838775635,
+      "learning_rate": 9.882668617816962e-06,
+      "loss": 0.6529,
+      "step": 2058
+    },
+    {
+      "epoch": 0.26825458805154234,
+      "grad_norm": 2.7566702365875244,
+      "learning_rate": 9.882218113058716e-06,
+      "loss": 0.6574,
+      "step": 2061
+    },
+    {
+      "epoch": 0.2686450605232331,
+      "grad_norm": 2.922041416168213,
+      "learning_rate": 9.88176675538724e-06,
+      "loss": 0.6732,
+      "step": 2064
+    },
+    {
+      "epoch": 0.26903553299492383,
+      "grad_norm": 4.118600368499756,
+      "learning_rate": 9.881314544881377e-06,
+      "loss": 0.6912,
+      "step": 2067
+    },
+    {
+      "epoch": 0.2694260054666146,
+      "grad_norm": 2.772783041000366,
+      "learning_rate": 9.880861481620134e-06,
+      "loss": 0.7012,
+      "step": 2070
+    },
+    {
+      "epoch": 0.2698164779383054,
+      "grad_norm": 2.855739116668701,
+      "learning_rate": 9.88040756568266e-06,
+      "loss": 0.7134,
+      "step": 2073
+    },
+    {
+      "epoch": 0.2702069504099961,
+      "grad_norm": 2.6456918716430664,
+      "learning_rate": 9.879952797148249e-06,
+      "loss": 0.6548,
+      "step": 2076
+    },
+    {
+      "epoch": 0.27059742288168687,
+      "grad_norm": 2.943904399871826,
+      "learning_rate": 9.87949717609635e-06,
+      "loss": 0.6378,
+      "step": 2079
+    },
+    {
+      "epoch": 0.2709878953533776,
+      "grad_norm": 2.9040253162384033,
+      "learning_rate": 9.87904070260656e-06,
+      "loss": 0.649,
+      "step": 2082
+    },
+    {
+      "epoch": 0.27137836782506836,
+      "grad_norm": 2.9723544120788574,
+      "learning_rate": 9.878583376758623e-06,
+      "loss": 0.705,
+      "step": 2085
+    },
+    {
+      "epoch": 0.2717688402967591,
+      "grad_norm": 3.080580949783325,
+      "learning_rate": 9.878125198632433e-06,
+      "loss": 0.7454,
+      "step": 2088
+    },
+    {
+      "epoch": 0.27215931276844985,
+      "grad_norm": 2.676499843597412,
+      "learning_rate": 9.877666168308034e-06,
+      "loss": 0.7742,
+      "step": 2091
+    },
+    {
+      "epoch": 0.27254978524014056,
+      "grad_norm": 2.981276035308838,
+      "learning_rate": 9.877206285865614e-06,
+      "loss": 0.7173,
+      "step": 2094
+    },
+    {
+      "epoch": 0.27294025771183134,
+      "grad_norm": 2.9143402576446533,
+      "learning_rate": 9.876745551385519e-06,
+      "loss": 0.6747,
+      "step": 2097
+    },
+    {
+      "epoch": 0.27333073018352205,
+      "grad_norm": 2.7891063690185547,
+      "learning_rate": 9.876283964948232e-06,
+      "loss": 0.5657,
+      "step": 2100
+    },
+    {
+      "epoch": 0.2737212026552128,
+      "grad_norm": 2.83236026763916,
+      "learning_rate": 9.875821526634397e-06,
+      "loss": 0.7013,
+      "step": 2103
+    },
+    {
+      "epoch": 0.27411167512690354,
+      "grad_norm": 2.9020628929138184,
+      "learning_rate": 9.875358236524798e-06,
+      "loss": 0.6061,
+      "step": 2106
+    },
+    {
+      "epoch": 0.2745021475985943,
+      "grad_norm": 2.637514352798462,
+      "learning_rate": 9.874894094700372e-06,
+      "loss": 0.6832,
+      "step": 2109
+    },
+    {
+      "epoch": 0.27489262007028503,
+      "grad_norm": 2.8690152168273926,
+      "learning_rate": 9.874429101242202e-06,
+      "loss": 0.7544,
+      "step": 2112
+    },
+    {
+      "epoch": 0.2752830925419758,
+      "grad_norm": 3.029694080352783,
+      "learning_rate": 9.873963256231522e-06,
+      "loss": 0.7404,
+      "step": 2115
+    },
+    {
+      "epoch": 0.2756735650136665,
+      "grad_norm": 3.4346494674682617,
+      "learning_rate": 9.873496559749716e-06,
+      "loss": 0.6089,
+      "step": 2118
+    },
+    {
+      "epoch": 0.2760640374853573,
+      "grad_norm": 2.6336498260498047,
+      "learning_rate": 9.873029011878312e-06,
+      "loss": 0.7007,
+      "step": 2121
+    },
+    {
+      "epoch": 0.276454509957048,
+      "grad_norm": 2.4632461071014404,
+      "learning_rate": 9.872560612698992e-06,
+      "loss": 0.6416,
+      "step": 2124
+    },
+    {
+      "epoch": 0.2768449824287388,
+      "grad_norm": 2.5596401691436768,
+      "learning_rate": 9.872091362293581e-06,
+      "loss": 0.6107,
+      "step": 2127
+    },
+    {
+      "epoch": 0.2772354549004295,
+      "grad_norm": 3.3999154567718506,
+      "learning_rate": 9.87162126074406e-06,
+      "loss": 0.6935,
+      "step": 2130
+    },
+    {
+      "epoch": 0.2776259273721203,
+      "grad_norm": 3.2630927562713623,
+      "learning_rate": 9.871150308132554e-06,
+      "loss": 0.6553,
+      "step": 2133
+    },
+    {
+      "epoch": 0.278016399843811,
+      "grad_norm": 3.5304367542266846,
+      "learning_rate": 9.870678504541336e-06,
+      "loss": 0.7273,
+      "step": 2136
+    },
+    {
+      "epoch": 0.27840687231550176,
+      "grad_norm": 3.2613844871520996,
+      "learning_rate": 9.87020585005283e-06,
+      "loss": 0.6989,
+      "step": 2139
+    },
+    {
+      "epoch": 0.2787973447871925,
+      "grad_norm": 2.5427675247192383,
+      "learning_rate": 9.869732344749605e-06,
+      "loss": 0.7172,
+      "step": 2142
+    },
+    {
+      "epoch": 0.27918781725888325,
+      "grad_norm": 2.583522081375122,
+      "learning_rate": 9.869257988714386e-06,
+      "loss": 0.688,
+      "step": 2145
+    },
+    {
+      "epoch": 0.27957828973057397,
+      "grad_norm": 3.7895023822784424,
+      "learning_rate": 9.86878278203004e-06,
+      "loss": 0.6906,
+      "step": 2148
+    },
+    {
+      "epoch": 0.27996876220226474,
+      "grad_norm": 2.4927175045013428,
+      "learning_rate": 9.868306724779584e-06,
+      "loss": 0.661,
+      "step": 2151
+    },
+    {
+      "epoch": 0.28035923467395546,
+      "grad_norm": 2.5533857345581055,
+      "learning_rate": 9.867829817046183e-06,
+      "loss": 0.6139,
+      "step": 2154
+    },
+    {
+      "epoch": 0.28074970714564623,
+      "grad_norm": 3.4451935291290283,
+      "learning_rate": 9.867352058913156e-06,
+      "loss": 0.78,
+      "step": 2157
+    },
+    {
+      "epoch": 0.281140179617337,
+      "grad_norm": 2.7176883220672607,
+      "learning_rate": 9.866873450463963e-06,
+      "loss": 0.7304,
+      "step": 2160
+    },
+    {
+      "epoch": 0.2815306520890277,
+      "grad_norm": 2.8105108737945557,
+      "learning_rate": 9.866393991782215e-06,
+      "loss": 0.6057,
+      "step": 2163
+    },
+    {
+      "epoch": 0.2819211245607185,
+      "grad_norm": 4.555068492889404,
+      "learning_rate": 9.865913682951675e-06,
+      "loss": 0.7168,
+      "step": 2166
+    },
+    {
+      "epoch": 0.2823115970324092,
+      "grad_norm": 4.762774467468262,
+      "learning_rate": 9.865432524056252e-06,
+      "loss": 0.6895,
+      "step": 2169
+    },
+    {
+      "epoch": 0.2827020695041,
+      "grad_norm": 2.9473941326141357,
+      "learning_rate": 9.86495051518e-06,
+      "loss": 0.6986,
+      "step": 2172
+    },
+    {
+      "epoch": 0.2830925419757907,
+      "grad_norm": 3.883134603500366,
+      "learning_rate": 9.86446765640713e-06,
+      "loss": 0.6505,
+      "step": 2175
+    },
+    {
+      "epoch": 0.2834830144474815,
+      "grad_norm": 2.467970132827759,
+      "learning_rate": 9.863983947821993e-06,
+      "loss": 0.7058,
+      "step": 2178
+    },
+    {
+      "epoch": 0.2838734869191722,
+      "grad_norm": 2.5856029987335205,
+      "learning_rate": 9.863499389509092e-06,
+      "loss": 0.6564,
+      "step": 2181
+    },
+    {
+      "epoch": 0.28426395939086296,
+      "grad_norm": 3.4759273529052734,
+      "learning_rate": 9.86301398155308e-06,
+      "loss": 0.644,
+      "step": 2184
+    },
+    {
+      "epoch": 0.2846544318625537,
+      "grad_norm": 3.2896201610565186,
+      "learning_rate": 9.862527724038755e-06,
+      "loss": 0.6862,
+      "step": 2187
+    },
+    {
+      "epoch": 0.28504490433424445,
+      "grad_norm": 3.064748764038086,
+      "learning_rate": 9.862040617051065e-06,
+      "loss": 0.6644,
+      "step": 2190
+    },
+    {
+      "epoch": 0.28543537680593517,
+      "grad_norm": 2.8241119384765625,
+      "learning_rate": 9.861552660675109e-06,
+      "loss": 0.6167,
+      "step": 2193
+    },
+    {
+      "epoch": 0.28582584927762594,
+      "grad_norm": 2.6539626121520996,
+      "learning_rate": 9.86106385499613e-06,
+      "loss": 0.6695,
+      "step": 2196
+    },
+    {
+      "epoch": 0.28621632174931666,
+      "grad_norm": 3.7717602252960205,
+      "learning_rate": 9.86057420009952e-06,
+      "loss": 0.7306,
+      "step": 2199
+    },
+    {
+      "epoch": 0.28660679422100743,
+      "grad_norm": 3.5839555263519287,
+      "learning_rate": 9.860083696070825e-06,
+      "loss": 0.6344,
+      "step": 2202
+    },
+    {
+      "epoch": 0.28699726669269815,
+      "grad_norm": 2.718700408935547,
+      "learning_rate": 9.85959234299573e-06,
+      "loss": 0.6781,
+      "step": 2205
+    },
+    {
+      "epoch": 0.2873877391643889,
+      "grad_norm": 3.010596990585327,
+      "learning_rate": 9.859100140960079e-06,
+      "loss": 0.7763,
+      "step": 2208
+    },
+    {
+      "epoch": 0.28777821163607964,
+      "grad_norm": 2.9677681922912598,
+      "learning_rate": 9.858607090049851e-06,
+      "loss": 0.6675,
+      "step": 2211
+    },
+    {
+      "epoch": 0.2881686841077704,
+      "grad_norm": 2.874814510345459,
+      "learning_rate": 9.858113190351189e-06,
+      "loss": 0.8113,
+      "step": 2214
+    },
+    {
+      "epoch": 0.28855915657946113,
+      "grad_norm": 2.915309190750122,
+      "learning_rate": 9.857618441950372e-06,
+      "loss": 0.7219,
+      "step": 2217
+    },
+    {
+      "epoch": 0.2889496290511519,
+      "grad_norm": 3.3547255992889404,
+      "learning_rate": 9.857122844933831e-06,
+      "loss": 0.689,
+      "step": 2220
+    },
+    {
+      "epoch": 0.2893401015228426,
+      "grad_norm": 2.769690752029419,
+      "learning_rate": 9.856626399388146e-06,
+      "loss": 0.6282,
+      "step": 2223
+    },
+    {
+      "epoch": 0.2897305739945334,
+      "grad_norm": 2.8011720180511475,
+      "learning_rate": 9.856129105400048e-06,
+      "loss": 0.6347,
+      "step": 2226
+    },
+    {
+      "epoch": 0.2901210464662241,
+      "grad_norm": 4.004727363586426,
+      "learning_rate": 9.855630963056411e-06,
+      "loss": 0.6332,
+      "step": 2229
+    },
+    {
+      "epoch": 0.2905115189379149,
+      "grad_norm": 3.3377795219421387,
+      "learning_rate": 9.855131972444259e-06,
+      "loss": 0.7273,
+      "step": 2232
+    },
+    {
+      "epoch": 0.2909019914096056,
+      "grad_norm": 4.1533379554748535,
+      "learning_rate": 9.854632133650765e-06,
+      "loss": 0.6631,
+      "step": 2235
+    },
+    {
+      "epoch": 0.29129246388129637,
+      "grad_norm": 3.070554256439209,
+      "learning_rate": 9.85413144676325e-06,
+      "loss": 0.6178,
+      "step": 2238
+    },
+    {
+      "epoch": 0.2916829363529871,
+      "grad_norm": 4.225495338439941,
+      "learning_rate": 9.853629911869182e-06,
+      "loss": 0.6509,
+      "step": 2241
+    },
+    {
+      "epoch": 0.29207340882467786,
+      "grad_norm": 3.233717441558838,
+      "learning_rate": 9.853127529056182e-06,
+      "loss": 0.7106,
+      "step": 2244
+    },
+    {
+      "epoch": 0.29246388129636863,
+      "grad_norm": 2.8367974758148193,
+      "learning_rate": 9.852624298412008e-06,
+      "loss": 0.6732,
+      "step": 2247
+    },
+    {
+      "epoch": 0.29285435376805935,
+      "grad_norm": 3.421760320663452,
+      "learning_rate": 9.852120220024579e-06,
+      "loss": 0.7217,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2932448262397501,
+      "grad_norm": 3.042901039123535,
+      "learning_rate": 9.851615293981956e-06,
+      "loss": 0.7075,
+      "step": 2253
+    },
+    {
+      "epoch": 0.29363529871144084,
+      "grad_norm": 2.7977664470672607,
+      "learning_rate": 9.851109520372346e-06,
+      "loss": 0.6137,
+      "step": 2256
+    },
+    {
+      "epoch": 0.2940257711831316,
+      "grad_norm": 2.4614906311035156,
+      "learning_rate": 9.85060289928411e-06,
+      "loss": 0.621,
+      "step": 2259
+    },
+    {
+      "epoch": 0.29441624365482233,
+      "grad_norm": 2.9462318420410156,
+      "learning_rate": 9.85009543080575e-06,
+      "loss": 0.7196,
+      "step": 2262
+    },
+    {
+      "epoch": 0.2948067161265131,
+      "grad_norm": 2.9534754753112793,
+      "learning_rate": 9.849587115025923e-06,
+      "loss": 0.6909,
+      "step": 2265
+    },
+    {
+      "epoch": 0.2951971885982038,
+      "grad_norm": 2.725046157836914,
+      "learning_rate": 9.849077952033427e-06,
+      "loss": 0.676,
+      "step": 2268
+    },
+    {
+      "epoch": 0.2955876610698946,
+      "grad_norm": 2.515141487121582,
+      "learning_rate": 9.848567941917216e-06,
+      "loss": 0.6987,
+      "step": 2271
+    },
+    {
+      "epoch": 0.2959781335415853,
+      "grad_norm": 2.8634941577911377,
+      "learning_rate": 9.848057084766382e-06,
+      "loss": 0.6337,
+      "step": 2274
+    },
+    {
+      "epoch": 0.2963686060132761,
+      "grad_norm": 3.2547008991241455,
+      "learning_rate": 9.847545380670176e-06,
+      "loss": 0.6618,
+      "step": 2277
+    },
+    {
+      "epoch": 0.2967590784849668,
+      "grad_norm": 3.159667730331421,
+      "learning_rate": 9.847032829717991e-06,
+      "loss": 0.6848,
+      "step": 2280
+    },
+    {
+      "epoch": 0.29714955095665757,
+      "grad_norm": 3.1204230785369873,
+      "learning_rate": 9.846519431999366e-06,
+      "loss": 0.6681,
+      "step": 2283
+    },
+    {
+      "epoch": 0.2975400234283483,
+      "grad_norm": 2.5976922512054443,
+      "learning_rate": 9.846005187603992e-06,
+      "loss": 0.8029,
+      "step": 2286
+    },
+    {
+      "epoch": 0.29793049590003906,
+      "grad_norm": 3.485644578933716,
+      "learning_rate": 9.845490096621706e-06,
+      "loss": 0.6945,
+      "step": 2289
+    },
+    {
+      "epoch": 0.2983209683717298,
+      "grad_norm": 3.9062230587005615,
+      "learning_rate": 9.844974159142494e-06,
+      "loss": 0.6611,
+      "step": 2292
+    },
+    {
+      "epoch": 0.29871144084342055,
+      "grad_norm": 2.7590322494506836,
+      "learning_rate": 9.844457375256488e-06,
+      "loss": 0.6508,
+      "step": 2295
+    },
+    {
+      "epoch": 0.29910191331511127,
+      "grad_norm": 2.883934736251831,
+      "learning_rate": 9.84393974505397e-06,
+      "loss": 0.7574,
+      "step": 2298
+    },
+    {
+      "epoch": 0.29949238578680204,
+      "grad_norm": 2.7723381519317627,
+      "learning_rate": 9.843421268625371e-06,
+      "loss": 0.604,
+      "step": 2301
+    },
+    {
+      "epoch": 0.29988285825849276,
+      "grad_norm": 2.5612845420837402,
+      "learning_rate": 9.842901946061263e-06,
+      "loss": 0.6271,
+      "step": 2304
+    },
+    {
+      "epoch": 0.30027333073018353,
+      "grad_norm": 2.4496188163757324,
+      "learning_rate": 9.842381777452373e-06,
+      "loss": 0.6317,
+      "step": 2307
+    },
+    {
+      "epoch": 0.30066380320187425,
+      "grad_norm": 2.936002731323242,
+      "learning_rate": 9.841860762889574e-06,
+      "loss": 0.6229,
+      "step": 2310
+    },
+    {
+      "epoch": 0.301054275673565,
+      "grad_norm": 2.9964542388916016,
+      "learning_rate": 9.841338902463885e-06,
+      "loss": 0.667,
+      "step": 2313
+    },
+    {
+      "epoch": 0.30144474814525574,
+      "grad_norm": 3.358015775680542,
+      "learning_rate": 9.840816196266475e-06,
+      "loss": 0.6421,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3018352206169465,
+      "grad_norm": 3.8633599281311035,
+      "learning_rate": 9.840292644388659e-06,
+      "loss": 0.6912,
+      "step": 2319
+    },
+    {
+      "epoch": 0.3022256930886372,
+      "grad_norm": 2.9916861057281494,
+      "learning_rate": 9.839768246921901e-06,
+      "loss": 0.6766,
+      "step": 2322
+    },
+    {
+      "epoch": 0.302616165560328,
+      "grad_norm": 3.3866844177246094,
+      "learning_rate": 9.83924300395781e-06,
+      "loss": 0.6397,
+      "step": 2325
+    },
+    {
+      "epoch": 0.3030066380320187,
+      "grad_norm": 2.5590834617614746,
+      "learning_rate": 9.838716915588148e-06,
+      "loss": 0.7354,
+      "step": 2328
+    },
+    {
+      "epoch": 0.3033971105037095,
+      "grad_norm": 2.6998791694641113,
+      "learning_rate": 9.83818998190482e-06,
+      "loss": 0.7159,
+      "step": 2331
+    },
+    {
+      "epoch": 0.30378758297540026,
+      "grad_norm": 2.2209694385528564,
+      "learning_rate": 9.837662202999879e-06,
+      "loss": 0.677,
+      "step": 2334
+    },
+    {
+      "epoch": 0.304178055447091,
+      "grad_norm": 3.1843392848968506,
+      "learning_rate": 9.83713357896553e-06,
+      "loss": 0.6385,
+      "step": 2337
+    },
+    {
+      "epoch": 0.30456852791878175,
+      "grad_norm": 2.818430185317993,
+      "learning_rate": 9.836604109894118e-06,
+      "loss": 0.643,
+      "step": 2340
+    },
+    {
+      "epoch": 0.30495900039047247,
+      "grad_norm": 2.380825996398926,
+      "learning_rate": 9.836073795878144e-06,
+      "loss": 0.6909,
+      "step": 2343
+    },
+    {
+      "epoch": 0.30534947286216324,
+      "grad_norm": 2.392238140106201,
+      "learning_rate": 9.835542637010253e-06,
+      "loss": 0.6184,
+      "step": 2346
+    },
+    {
+      "epoch": 0.30573994533385396,
+      "grad_norm": 3.895524263381958,
+      "learning_rate": 9.835010633383234e-06,
+      "loss": 0.735,
+      "step": 2349
+    },
+    {
+      "epoch": 0.30613041780554473,
+      "grad_norm": 2.8263940811157227,
+      "learning_rate": 9.834477785090028e-06,
+      "loss": 0.6762,
+      "step": 2352
+    },
+    {
+      "epoch": 0.30652089027723545,
+      "grad_norm": 3.5129168033599854,
+      "learning_rate": 9.833944092223725e-06,
+      "loss": 0.746,
+      "step": 2355
+    },
+    {
+      "epoch": 0.3069113627489262,
+      "grad_norm": 2.720536947250366,
+      "learning_rate": 9.833409554877558e-06,
+      "loss": 0.6833,
+      "step": 2358
+    },
+    {
+      "epoch": 0.30730183522061694,
+      "grad_norm": 2.6948177814483643,
+      "learning_rate": 9.83287417314491e-06,
+      "loss": 0.6223,
+      "step": 2361
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 2.750875234603882,
+      "learning_rate": 9.832337947119311e-06,
+      "loss": 0.7024,
+      "step": 2364
+    },
+    {
+      "epoch": 0.3080827801639984,
+      "grad_norm": 3.7236342430114746,
+      "learning_rate": 9.831800876894436e-06,
+      "loss": 0.7258,
+      "step": 2367
+    },
+    {
+      "epoch": 0.3084732526356892,
+      "grad_norm": 2.5590686798095703,
+      "learning_rate": 9.831262962564114e-06,
+      "loss": 0.6718,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3088637251073799,
+      "grad_norm": 3.692631483078003,
+      "learning_rate": 9.830724204222316e-06,
+      "loss": 0.6383,
+      "step": 2373
+    },
+    {
+      "epoch": 0.3092541975790707,
+      "grad_norm": 2.6844470500946045,
+      "learning_rate": 9.830184601963162e-06,
+      "loss": 0.6747,
+      "step": 2376
+    },
+    {
+      "epoch": 0.3096446700507614,
+      "grad_norm": 3.532243490219116,
+      "learning_rate": 9.82964415588092e-06,
+      "loss": 0.72,
+      "step": 2379
+    },
+    {
+      "epoch": 0.3100351425224522,
+      "grad_norm": 2.9188520908355713,
+      "learning_rate": 9.829102866070002e-06,
+      "loss": 0.6895,
+      "step": 2382
+    },
+    {
+      "epoch": 0.3104256149941429,
+      "grad_norm": 2.847670316696167,
+      "learning_rate": 9.828560732624974e-06,
+      "loss": 0.7294,
+      "step": 2385
+    },
+    {
+      "epoch": 0.31081608746583367,
+      "grad_norm": 2.8164069652557373,
+      "learning_rate": 9.828017755640543e-06,
+      "loss": 0.7019,
+      "step": 2388
+    },
+    {
+      "epoch": 0.3112065599375244,
+      "grad_norm": 2.716837167739868,
+      "learning_rate": 9.827473935211567e-06,
+      "loss": 0.7291,
+      "step": 2391
+    },
+    {
+      "epoch": 0.31159703240921516,
+      "grad_norm": 2.867050886154175,
+      "learning_rate": 9.82692927143305e-06,
+      "loss": 0.6845,
+      "step": 2394
+    },
+    {
+      "epoch": 0.3119875048809059,
+      "grad_norm": 2.4712677001953125,
+      "learning_rate": 9.826383764400143e-06,
+      "loss": 0.5568,
+      "step": 2397
+    },
+    {
+      "epoch": 0.31237797735259665,
+      "grad_norm": 2.6688764095306396,
+      "learning_rate": 9.825837414208147e-06,
+      "loss": 0.6441,
+      "step": 2400
+    },
+    {
+      "epoch": 0.31276844982428736,
+      "grad_norm": 2.8412606716156006,
+      "learning_rate": 9.825290220952507e-06,
+      "loss": 0.7189,
+      "step": 2403
+    },
+    {
+      "epoch": 0.31315892229597814,
+      "grad_norm": 4.064339637756348,
+      "learning_rate": 9.824742184728815e-06,
+      "loss": 0.595,
+      "step": 2406
+    },
+    {
+      "epoch": 0.31354939476766885,
+      "grad_norm": 2.6639394760131836,
+      "learning_rate": 9.824193305632814e-06,
+      "loss": 0.6398,
+      "step": 2409
+    },
+    {
+      "epoch": 0.3139398672393596,
+      "grad_norm": 3.3248283863067627,
+      "learning_rate": 9.823643583760389e-06,
+      "loss": 0.6377,
+      "step": 2412
+    },
+    {
+      "epoch": 0.31433033971105034,
+      "grad_norm": 4.568389415740967,
+      "learning_rate": 9.823093019207578e-06,
+      "loss": 0.687,
+      "step": 2415
+    },
+    {
+      "epoch": 0.3147208121827411,
+      "grad_norm": 3.337682008743286,
+      "learning_rate": 9.822541612070566e-06,
+      "loss": 0.6771,
+      "step": 2418
+    },
+    {
+      "epoch": 0.3151112846544319,
+      "grad_norm": 2.871934652328491,
+      "learning_rate": 9.821989362445676e-06,
+      "loss": 0.7428,
+      "step": 2421
+    },
+    {
+      "epoch": 0.3155017571261226,
+      "grad_norm": 3.3546926975250244,
+      "learning_rate": 9.82143627042939e-06,
+      "loss": 0.6353,
+      "step": 2424
+    },
+    {
+      "epoch": 0.3158922295978134,
+      "grad_norm": 2.4146947860717773,
+      "learning_rate": 9.820882336118332e-06,
+      "loss": 0.6301,
+      "step": 2427
+    },
+    {
+      "epoch": 0.3162827020695041,
+      "grad_norm": 2.7984747886657715,
+      "learning_rate": 9.820327559609268e-06,
+      "loss": 0.5877,
+      "step": 2430
+    },
+    {
+      "epoch": 0.31667317454119487,
+      "grad_norm": 4.265294551849365,
+      "learning_rate": 9.819771940999123e-06,
+      "loss": 0.6227,
+      "step": 2433
+    },
+    {
+      "epoch": 0.3170636470128856,
+      "grad_norm": 2.921192169189453,
+      "learning_rate": 9.819215480384956e-06,
+      "loss": 0.6851,
+      "step": 2436
+    },
+    {
+      "epoch": 0.31745411948457636,
+      "grad_norm": 2.67498517036438,
+      "learning_rate": 9.818658177863985e-06,
+      "loss": 0.5766,
+      "step": 2439
+    },
+    {
+      "epoch": 0.3178445919562671,
+      "grad_norm": 2.93373703956604,
+      "learning_rate": 9.818100033533567e-06,
+      "loss": 0.7525,
+      "step": 2442
+    },
+    {
+      "epoch": 0.31823506442795785,
+      "grad_norm": 2.9454703330993652,
+      "learning_rate": 9.817541047491209e-06,
+      "loss": 0.7152,
+      "step": 2445
+    },
+    {
+      "epoch": 0.31862553689964856,
+      "grad_norm": 2.5796053409576416,
+      "learning_rate": 9.816981219834565e-06,
+      "loss": 0.6848,
+      "step": 2448
+    },
+    {
+      "epoch": 0.31901600937133934,
+      "grad_norm": 3.1265947818756104,
+      "learning_rate": 9.816420550661434e-06,
+      "loss": 0.7028,
+      "step": 2451
+    },
+    {
+      "epoch": 0.31940648184303005,
+      "grad_norm": 2.7802746295928955,
+      "learning_rate": 9.815859040069766e-06,
+      "loss": 0.6456,
+      "step": 2454
+    },
+    {
+      "epoch": 0.3197969543147208,
+      "grad_norm": 3.0092685222625732,
+      "learning_rate": 9.815296688157654e-06,
+      "loss": 0.6739,
+      "step": 2457
+    },
+    {
+      "epoch": 0.32018742678641154,
+      "grad_norm": 3.065697193145752,
+      "learning_rate": 9.81473349502334e-06,
+      "loss": 0.6604,
+      "step": 2460
+    },
+    {
+      "epoch": 0.3205778992581023,
+      "grad_norm": 2.455078601837158,
+      "learning_rate": 9.814169460765215e-06,
+      "loss": 0.6433,
+      "step": 2463
+    },
+    {
+      "epoch": 0.32096837172979303,
+      "grad_norm": 2.841989040374756,
+      "learning_rate": 9.813604585481812e-06,
+      "loss": 0.6915,
+      "step": 2466
+    },
+    {
+      "epoch": 0.3213588442014838,
+      "grad_norm": 2.2834136486053467,
+      "learning_rate": 9.813038869271814e-06,
+      "loss": 0.634,
+      "step": 2469
+    },
+    {
+      "epoch": 0.3217493166731745,
+      "grad_norm": 3.262573719024658,
+      "learning_rate": 9.81247231223405e-06,
+      "loss": 0.6987,
+      "step": 2472
+    },
+    {
+      "epoch": 0.3221397891448653,
+      "grad_norm": 2.7389028072357178,
+      "learning_rate": 9.811904914467498e-06,
+      "loss": 0.6553,
+      "step": 2475
+    },
+    {
+      "epoch": 0.322530261616556,
+      "grad_norm": 2.779263973236084,
+      "learning_rate": 9.81133667607128e-06,
+      "loss": 0.7449,
+      "step": 2478
+    },
+    {
+      "epoch": 0.3229207340882468,
+      "grad_norm": 2.815535306930542,
+      "learning_rate": 9.810767597144668e-06,
+      "loss": 0.5643,
+      "step": 2481
+    },
+    {
+      "epoch": 0.3233112065599375,
+      "grad_norm": 2.794158458709717,
+      "learning_rate": 9.810197677787074e-06,
+      "loss": 0.7033,
+      "step": 2484
+    },
+    {
+      "epoch": 0.3237016790316283,
+      "grad_norm": 3.1413052082061768,
+      "learning_rate": 9.809626918098066e-06,
+      "loss": 0.6482,
+      "step": 2487
+    },
+    {
+      "epoch": 0.324092151503319,
+      "grad_norm": 2.798581838607788,
+      "learning_rate": 9.809055318177353e-06,
+      "loss": 0.6147,
+      "step": 2490
+    },
+    {
+      "epoch": 0.32448262397500977,
+      "grad_norm": 3.2870383262634277,
+      "learning_rate": 9.808482878124794e-06,
+      "loss": 0.6716,
+      "step": 2493
+    },
+    {
+      "epoch": 0.3248730964467005,
+      "grad_norm": 2.5646088123321533,
+      "learning_rate": 9.807909598040392e-06,
+      "loss": 0.6022,
+      "step": 2496
+    },
+    {
+      "epoch": 0.32526356891839125,
+      "grad_norm": 3.979936361312866,
+      "learning_rate": 9.807335478024297e-06,
+      "loss": 0.6804,
+      "step": 2499
+    },
+    {
+      "epoch": 0.32565404139008197,
+      "grad_norm": 3.786860704421997,
+      "learning_rate": 9.806760518176806e-06,
+      "loss": 0.6807,
+      "step": 2502
+    },
+    {
+      "epoch": 0.32604451386177274,
+      "grad_norm": 3.001896619796753,
+      "learning_rate": 9.806184718598365e-06,
+      "loss": 0.6799,
+      "step": 2505
+    },
+    {
+      "epoch": 0.3264349863334635,
+      "grad_norm": 2.965797185897827,
+      "learning_rate": 9.805608079389566e-06,
+      "loss": 0.7123,
+      "step": 2508
+    },
+    {
+      "epoch": 0.32682545880515423,
+      "grad_norm": 5.5680460929870605,
+      "learning_rate": 9.805030600651143e-06,
+      "loss": 0.6111,
+      "step": 2511
+    },
+    {
+      "epoch": 0.327215931276845,
+      "grad_norm": 3.2845263481140137,
+      "learning_rate": 9.804452282483983e-06,
+      "loss": 0.614,
+      "step": 2514
+    },
+    {
+      "epoch": 0.3276064037485357,
+      "grad_norm": 2.4378509521484375,
+      "learning_rate": 9.803873124989116e-06,
+      "loss": 0.712,
+      "step": 2517
+    },
+    {
+      "epoch": 0.3279968762202265,
+      "grad_norm": 2.590458393096924,
+      "learning_rate": 9.803293128267721e-06,
+      "loss": 0.62,
+      "step": 2520
+    },
+    {
+      "epoch": 0.3283873486919172,
+      "grad_norm": 2.5005409717559814,
+      "learning_rate": 9.802712292421121e-06,
+      "loss": 0.7454,
+      "step": 2523
+    },
+    {
+      "epoch": 0.328777821163608,
+      "grad_norm": 3.139564037322998,
+      "learning_rate": 9.802130617550788e-06,
+      "loss": 0.7651,
+      "step": 2526
+    },
+    {
+      "epoch": 0.3291682936352987,
+      "grad_norm": 3.2621004581451416,
+      "learning_rate": 9.801548103758335e-06,
+      "loss": 0.7262,
+      "step": 2529
+    },
+    {
+      "epoch": 0.3295587661069895,
+      "grad_norm": 2.5339832305908203,
+      "learning_rate": 9.800964751145533e-06,
+      "loss": 0.6914,
+      "step": 2532
+    },
+    {
+      "epoch": 0.3299492385786802,
+      "grad_norm": 3.3442342281341553,
+      "learning_rate": 9.800380559814284e-06,
+      "loss": 0.6403,
+      "step": 2535
+    },
+    {
+      "epoch": 0.33033971105037097,
+      "grad_norm": 2.522374153137207,
+      "learning_rate": 9.799795529866654e-06,
+      "loss": 0.6177,
+      "step": 2538
+    },
+    {
+      "epoch": 0.3307301835220617,
+      "grad_norm": 2.8395447731018066,
+      "learning_rate": 9.79920966140484e-06,
+      "loss": 0.6954,
+      "step": 2541
+    },
+    {
+      "epoch": 0.33112065599375246,
+      "grad_norm": 2.558065891265869,
+      "learning_rate": 9.798622954531194e-06,
+      "loss": 0.6864,
+      "step": 2544
+    },
+    {
+      "epoch": 0.3315111284654432,
+      "grad_norm": 3.8972573280334473,
+      "learning_rate": 9.798035409348214e-06,
+      "loss": 0.6951,
+      "step": 2547
+    },
+    {
+      "epoch": 0.33190160093713394,
+      "grad_norm": 3.3120248317718506,
+      "learning_rate": 9.797447025958542e-06,
+      "loss": 0.5917,
+      "step": 2550
+    },
+    {
+      "epoch": 0.33229207340882466,
+      "grad_norm": 2.999908685684204,
+      "learning_rate": 9.796857804464966e-06,
+      "loss": 0.6607,
+      "step": 2553
+    },
+    {
+      "epoch": 0.33268254588051543,
+      "grad_norm": 2.7225866317749023,
+      "learning_rate": 9.796267744970423e-06,
+      "loss": 0.7139,
+      "step": 2556
+    },
+    {
+      "epoch": 0.33307301835220615,
+      "grad_norm": 2.655848741531372,
+      "learning_rate": 9.795676847577995e-06,
+      "loss": 0.7146,
+      "step": 2559
+    },
+    {
+      "epoch": 0.3334634908238969,
+      "grad_norm": 3.1451730728149414,
+      "learning_rate": 9.795085112390909e-06,
+      "loss": 0.5869,
+      "step": 2562
+    },
+    {
+      "epoch": 0.33385396329558764,
+      "grad_norm": 3.549569845199585,
+      "learning_rate": 9.794492539512544e-06,
+      "loss": 0.6792,
+      "step": 2565
+    },
+    {
+      "epoch": 0.3342444357672784,
+      "grad_norm": 2.678586959838867,
+      "learning_rate": 9.793899129046417e-06,
+      "loss": 0.5984,
+      "step": 2568
+    },
+    {
+      "epoch": 0.33463490823896913,
+      "grad_norm": 2.842653274536133,
+      "learning_rate": 9.793304881096195e-06,
+      "loss": 0.6587,
+      "step": 2571
+    },
+    {
+      "epoch": 0.3350253807106599,
+      "grad_norm": 2.715420722961426,
+      "learning_rate": 9.792709795765695e-06,
+      "loss": 0.6969,
+      "step": 2574
+    },
+    {
+      "epoch": 0.3354158531823506,
+      "grad_norm": 3.2462103366851807,
+      "learning_rate": 9.792113873158877e-06,
+      "loss": 0.7084,
+      "step": 2577
+    },
+    {
+      "epoch": 0.3358063256540414,
+      "grad_norm": 3.362494468688965,
+      "learning_rate": 9.791517113379847e-06,
+      "loss": 0.7104,
+      "step": 2580
+    },
+    {
+      "epoch": 0.3361967981257321,
+      "grad_norm": 2.433021306991577,
+      "learning_rate": 9.790919516532856e-06,
+      "loss": 0.6128,
+      "step": 2583
+    },
+    {
+      "epoch": 0.3365872705974229,
+      "grad_norm": 2.534708261489868,
+      "learning_rate": 9.790321082722306e-06,
+      "loss": 0.6569,
+      "step": 2586
+    },
+    {
+      "epoch": 0.3369777430691136,
+      "grad_norm": 2.7385549545288086,
+      "learning_rate": 9.789721812052738e-06,
+      "loss": 0.6337,
+      "step": 2589
+    },
+    {
+      "epoch": 0.3373682155408044,
+      "grad_norm": 2.9946787357330322,
+      "learning_rate": 9.789121704628845e-06,
+      "loss": 0.6794,
+      "step": 2592
+    },
+    {
+      "epoch": 0.33775868801249515,
+      "grad_norm": 2.673137664794922,
+      "learning_rate": 9.788520760555467e-06,
+      "loss": 0.6952,
+      "step": 2595
+    },
+    {
+      "epoch": 0.33814916048418586,
+      "grad_norm": 2.8383102416992188,
+      "learning_rate": 9.787918979937584e-06,
+      "loss": 0.7175,
+      "step": 2598
+    },
+    {
+      "epoch": 0.33853963295587663,
+      "grad_norm": 2.792264223098755,
+      "learning_rate": 9.78731636288033e-06,
+      "loss": 0.6472,
+      "step": 2601
+    },
+    {
+      "epoch": 0.33893010542756735,
+      "grad_norm": 3.444775342941284,
+      "learning_rate": 9.786712909488976e-06,
+      "loss": 0.6164,
+      "step": 2604
+    },
+    {
+      "epoch": 0.3393205778992581,
+      "grad_norm": 2.6704118251800537,
+      "learning_rate": 9.786108619868948e-06,
+      "loss": 0.6913,
+      "step": 2607
+    },
+    {
+      "epoch": 0.33971105037094884,
+      "grad_norm": 3.128438949584961,
+      "learning_rate": 9.785503494125812e-06,
+      "loss": 0.6493,
+      "step": 2610
+    },
+    {
+      "epoch": 0.3401015228426396,
+      "grad_norm": 3.342224359512329,
+      "learning_rate": 9.784897532365283e-06,
+      "loss": 0.7241,
+      "step": 2613
+    },
+    {
+      "epoch": 0.34049199531433033,
+      "grad_norm": 3.057779550552368,
+      "learning_rate": 9.78429073469322e-06,
+      "loss": 0.6189,
+      "step": 2616
+    },
+    {
+      "epoch": 0.3408824677860211,
+      "grad_norm": 2.809861898422241,
+      "learning_rate": 9.783683101215632e-06,
+      "loss": 0.6926,
+      "step": 2619
+    },
+    {
+      "epoch": 0.3412729402577118,
+      "grad_norm": 2.8063440322875977,
+      "learning_rate": 9.783074632038669e-06,
+      "loss": 0.7242,
+      "step": 2622
+    },
+    {
+      "epoch": 0.3416634127294026,
+      "grad_norm": 2.8597209453582764,
+      "learning_rate": 9.78246532726863e-06,
+      "loss": 0.6391,
+      "step": 2625
+    },
+    {
+      "epoch": 0.3420538852010933,
+      "grad_norm": 2.6011080741882324,
+      "learning_rate": 9.78185518701196e-06,
+      "loss": 0.6819,
+      "step": 2628
+    },
+    {
+      "epoch": 0.3424443576727841,
+      "grad_norm": 3.5617270469665527,
+      "learning_rate": 9.781244211375247e-06,
+      "loss": 0.733,
+      "step": 2631
+    },
+    {
+      "epoch": 0.3428348301444748,
+      "grad_norm": 3.6274831295013428,
+      "learning_rate": 9.78063240046523e-06,
+      "loss": 0.7171,
+      "step": 2634
+    },
+    {
+      "epoch": 0.3432253026161656,
+      "grad_norm": 2.6554691791534424,
+      "learning_rate": 9.780019754388786e-06,
+      "loss": 0.7321,
+      "step": 2637
+    },
+    {
+      "epoch": 0.3436157750878563,
+      "grad_norm": 2.574505090713501,
+      "learning_rate": 9.779406273252949e-06,
+      "loss": 0.7734,
+      "step": 2640
+    },
+    {
+      "epoch": 0.34400624755954706,
+      "grad_norm": 4.732491970062256,
+      "learning_rate": 9.77879195716489e-06,
+      "loss": 0.7968,
+      "step": 2643
+    },
+    {
+      "epoch": 0.3443967200312378,
+      "grad_norm": 3.508035182952881,
+      "learning_rate": 9.778176806231931e-06,
+      "loss": 0.6802,
+      "step": 2646
+    },
+    {
+      "epoch": 0.34478719250292855,
+      "grad_norm": 2.745506763458252,
+      "learning_rate": 9.777560820561533e-06,
+      "loss": 0.6712,
+      "step": 2649
+    },
+    {
+      "epoch": 0.34517766497461927,
+      "grad_norm": 2.6209089756011963,
+      "learning_rate": 9.776944000261313e-06,
+      "loss": 0.6477,
+      "step": 2652
+    },
+    {
+      "epoch": 0.34556813744631004,
+      "grad_norm": 3.0945277214050293,
+      "learning_rate": 9.776326345439024e-06,
+      "loss": 0.6644,
+      "step": 2655
+    },
+    {
+      "epoch": 0.34595860991800076,
+      "grad_norm": 3.156196117401123,
+      "learning_rate": 9.77570785620257e-06,
+      "loss": 0.6713,
+      "step": 2658
+    },
+    {
+      "epoch": 0.34634908238969153,
+      "grad_norm": 3.839165210723877,
+      "learning_rate": 9.77508853266e-06,
+      "loss": 0.7557,
+      "step": 2661
+    },
+    {
+      "epoch": 0.34673955486138225,
+      "grad_norm": 3.0956363677978516,
+      "learning_rate": 9.77446837491951e-06,
+      "loss": 0.7111,
+      "step": 2664
+    },
+    {
+      "epoch": 0.347130027333073,
+      "grad_norm": 2.516638994216919,
+      "learning_rate": 9.773847383089439e-06,
+      "loss": 0.6385,
+      "step": 2667
+    },
+    {
+      "epoch": 0.34752049980476374,
+      "grad_norm": 2.5300917625427246,
+      "learning_rate": 9.773225557278272e-06,
+      "loss": 0.6205,
+      "step": 2670
+    },
+    {
+      "epoch": 0.3479109722764545,
+      "grad_norm": 2.5988070964813232,
+      "learning_rate": 9.77260289759464e-06,
+      "loss": 0.6469,
+      "step": 2673
+    },
+    {
+      "epoch": 0.34830144474814523,
+      "grad_norm": 2.493129014968872,
+      "learning_rate": 9.771979404147324e-06,
+      "loss": 0.6266,
+      "step": 2676
+    },
+    {
+      "epoch": 0.348691917219836,
+      "grad_norm": 2.7826688289642334,
+      "learning_rate": 9.771355077045244e-06,
+      "loss": 0.6741,
+      "step": 2679
+    },
+    {
+      "epoch": 0.3490823896915268,
+      "grad_norm": 2.9177565574645996,
+      "learning_rate": 9.77072991639747e-06,
+      "loss": 0.764,
+      "step": 2682
+    },
+    {
+      "epoch": 0.3494728621632175,
+      "grad_norm": 3.730968475341797,
+      "learning_rate": 9.770103922313215e-06,
+      "loss": 0.5789,
+      "step": 2685
+    },
+    {
+      "epoch": 0.34986333463490826,
+      "grad_norm": 2.848114490509033,
+      "learning_rate": 9.76947709490184e-06,
+      "loss": 0.6656,
+      "step": 2688
+    },
+    {
+      "epoch": 0.350253807106599,
+      "grad_norm": 3.0486176013946533,
+      "learning_rate": 9.768849434272851e-06,
+      "loss": 0.6857,
+      "step": 2691
+    },
+    {
+      "epoch": 0.35064427957828975,
+      "grad_norm": 2.6924357414245605,
+      "learning_rate": 9.768220940535897e-06,
+      "loss": 0.6939,
+      "step": 2694
+    },
+    {
+      "epoch": 0.35103475204998047,
+      "grad_norm": 3.603853225708008,
+      "learning_rate": 9.767591613800775e-06,
+      "loss": 0.6395,
+      "step": 2697
+    },
+    {
+      "epoch": 0.35142522452167124,
+      "grad_norm": 2.700711250305176,
+      "learning_rate": 9.76696145417743e-06,
+      "loss": 0.6732,
+      "step": 2700
+    },
+    {
+      "epoch": 0.35181569699336196,
+      "grad_norm": 2.3906078338623047,
+      "learning_rate": 9.766330461775944e-06,
+      "loss": 0.5556,
+      "step": 2703
+    },
+    {
+      "epoch": 0.35220616946505273,
+      "grad_norm": 2.9107656478881836,
+      "learning_rate": 9.765698636706555e-06,
+      "loss": 0.7257,
+      "step": 2706
+    },
+    {
+      "epoch": 0.35259664193674345,
+      "grad_norm": 2.9227166175842285,
+      "learning_rate": 9.765065979079639e-06,
+      "loss": 0.5716,
+      "step": 2709
+    },
+    {
+      "epoch": 0.3529871144084342,
+      "grad_norm": 3.1230244636535645,
+      "learning_rate": 9.764432489005722e-06,
+      "loss": 0.6969,
+      "step": 2712
+    },
+    {
+      "epoch": 0.35337758688012494,
+      "grad_norm": 2.984856128692627,
+      "learning_rate": 9.763798166595473e-06,
+      "loss": 0.6435,
+      "step": 2715
+    },
+    {
+      "epoch": 0.3537680593518157,
+      "grad_norm": 2.4585700035095215,
+      "learning_rate": 9.763163011959702e-06,
+      "loss": 0.7096,
+      "step": 2718
+    },
+    {
+      "epoch": 0.35415853182350643,
+      "grad_norm": 2.7415413856506348,
+      "learning_rate": 9.762527025209377e-06,
+      "loss": 0.7252,
+      "step": 2721
+    },
+    {
+      "epoch": 0.3545490042951972,
+      "grad_norm": 3.0851871967315674,
+      "learning_rate": 9.761890206455597e-06,
+      "loss": 0.6221,
+      "step": 2724
+    },
+    {
+      "epoch": 0.3549394767668879,
+      "grad_norm": 4.111907482147217,
+      "learning_rate": 9.761252555809616e-06,
+      "loss": 0.6155,
+      "step": 2727
+    },
+    {
+      "epoch": 0.3553299492385787,
+      "grad_norm": 3.664571762084961,
+      "learning_rate": 9.76061407338283e-06,
+      "loss": 0.6705,
+      "step": 2730
+    },
+    {
+      "epoch": 0.3557204217102694,
+      "grad_norm": 2.795311689376831,
+      "learning_rate": 9.75997475928678e-06,
+      "loss": 0.6785,
+      "step": 2733
+    },
+    {
+      "epoch": 0.3561108941819602,
+      "grad_norm": 3.0778064727783203,
+      "learning_rate": 9.759334613633154e-06,
+      "loss": 0.7244,
+      "step": 2736
+    },
+    {
+      "epoch": 0.3565013666536509,
+      "grad_norm": 3.36773943901062,
+      "learning_rate": 9.758693636533782e-06,
+      "loss": 0.6062,
+      "step": 2739
+    },
+    {
+      "epoch": 0.35689183912534167,
+      "grad_norm": 2.6564176082611084,
+      "learning_rate": 9.758051828100643e-06,
+      "loss": 0.7016,
+      "step": 2742
+    },
+    {
+      "epoch": 0.3572823115970324,
+      "grad_norm": 2.544980049133301,
+      "learning_rate": 9.75740918844586e-06,
+      "loss": 0.6046,
+      "step": 2745
+    },
+    {
+      "epoch": 0.35767278406872316,
+      "grad_norm": 3.575075626373291,
+      "learning_rate": 9.756765717681698e-06,
+      "loss": 0.663,
+      "step": 2748
+    },
+    {
+      "epoch": 0.3580632565404139,
+      "grad_norm": 2.658778429031372,
+      "learning_rate": 9.756121415920572e-06,
+      "loss": 0.6775,
+      "step": 2751
+    },
+    {
+      "epoch": 0.35845372901210465,
+      "grad_norm": 2.7959978580474854,
+      "learning_rate": 9.755476283275042e-06,
+      "loss": 0.7469,
+      "step": 2754
+    },
+    {
+      "epoch": 0.35884420148379537,
+      "grad_norm": 2.5027709007263184,
+      "learning_rate": 9.754830319857809e-06,
+      "loss": 0.6239,
+      "step": 2757
+    },
+    {
+      "epoch": 0.35923467395548614,
+      "grad_norm": 2.6545188426971436,
+      "learning_rate": 9.75418352578172e-06,
+      "loss": 0.6756,
+      "step": 2760
+    },
+    {
+      "epoch": 0.3596251464271769,
+      "grad_norm": 2.5996038913726807,
+      "learning_rate": 9.753535901159772e-06,
+      "loss": 0.7895,
+      "step": 2763
+    },
+    {
+      "epoch": 0.36001561889886763,
+      "grad_norm": 2.747310161590576,
+      "learning_rate": 9.752887446105101e-06,
+      "loss": 0.6933,
+      "step": 2766
+    },
+    {
+      "epoch": 0.3604060913705584,
+      "grad_norm": 2.4019601345062256,
+      "learning_rate": 9.752238160730994e-06,
+      "loss": 0.5939,
+      "step": 2769
+    },
+    {
+      "epoch": 0.3607965638422491,
+      "grad_norm": 3.0222368240356445,
+      "learning_rate": 9.751588045150875e-06,
+      "loss": 0.7407,
+      "step": 2772
+    },
+    {
+      "epoch": 0.3611870363139399,
+      "grad_norm": 3.187483310699463,
+      "learning_rate": 9.750937099478322e-06,
+      "loss": 0.6874,
+      "step": 2775
+    },
+    {
+      "epoch": 0.3615775087856306,
+      "grad_norm": 2.897294282913208,
+      "learning_rate": 9.750285323827051e-06,
+      "loss": 0.7485,
+      "step": 2778
+    },
+    {
+      "epoch": 0.3619679812573214,
+      "grad_norm": 2.5609171390533447,
+      "learning_rate": 9.749632718310927e-06,
+      "loss": 0.6426,
+      "step": 2781
+    },
+    {
+      "epoch": 0.3623584537290121,
+      "grad_norm": 2.4874250888824463,
+      "learning_rate": 9.74897928304396e-06,
+      "loss": 0.6134,
+      "step": 2784
+    },
+    {
+      "epoch": 0.36274892620070287,
+      "grad_norm": 3.737605333328247,
+      "learning_rate": 9.748325018140301e-06,
+      "loss": 0.661,
+      "step": 2787
+    },
+    {
+      "epoch": 0.3631393986723936,
+      "grad_norm": 2.8321633338928223,
+      "learning_rate": 9.747669923714252e-06,
+      "loss": 0.7641,
+      "step": 2790
+    },
+    {
+      "epoch": 0.36352987114408436,
+      "grad_norm": 2.5597903728485107,
+      "learning_rate": 9.747013999880255e-06,
+      "loss": 0.6625,
+      "step": 2793
+    },
+    {
+      "epoch": 0.3639203436157751,
+      "grad_norm": 2.5689401626586914,
+      "learning_rate": 9.746357246752898e-06,
+      "loss": 0.6729,
+      "step": 2796
+    },
+    {
+      "epoch": 0.36431081608746585,
+      "grad_norm": 2.8747706413269043,
+      "learning_rate": 9.745699664446914e-06,
+      "loss": 0.6406,
+      "step": 2799
+    },
+    {
+      "epoch": 0.36470128855915657,
+      "grad_norm": 2.724287271499634,
+      "learning_rate": 9.745041253077183e-06,
+      "loss": 0.6439,
+      "step": 2802
+    },
+    {
+      "epoch": 0.36509176103084734,
+      "grad_norm": 3.9166243076324463,
+      "learning_rate": 9.744382012758727e-06,
+      "loss": 0.6298,
+      "step": 2805
+    },
+    {
+      "epoch": 0.36548223350253806,
+      "grad_norm": 3.197875499725342,
+      "learning_rate": 9.743721943606715e-06,
+      "loss": 0.6032,
+      "step": 2808
+    },
+    {
+      "epoch": 0.36587270597422883,
+      "grad_norm": 2.6559009552001953,
+      "learning_rate": 9.743061045736457e-06,
+      "loss": 0.6065,
+      "step": 2811
+    },
+    {
+      "epoch": 0.36626317844591955,
+      "grad_norm": 2.4632251262664795,
+      "learning_rate": 9.742399319263414e-06,
+      "loss": 0.6398,
+      "step": 2814
+    },
+    {
+      "epoch": 0.3666536509176103,
+      "grad_norm": 3.159674882888794,
+      "learning_rate": 9.741736764303185e-06,
+      "loss": 0.5769,
+      "step": 2817
+    },
+    {
+      "epoch": 0.36704412338930104,
+      "grad_norm": 2.6667957305908203,
+      "learning_rate": 9.74107338097152e-06,
+      "loss": 0.7085,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3674345958609918,
+      "grad_norm": 2.641826629638672,
+      "learning_rate": 9.740409169384308e-06,
+      "loss": 0.6944,
+      "step": 2823
+    },
+    {
+      "epoch": 0.3678250683326825,
+      "grad_norm": 3.9662795066833496,
+      "learning_rate": 9.739744129657586e-06,
+      "loss": 0.6217,
+      "step": 2826
+    },
+    {
+      "epoch": 0.3682155408043733,
+      "grad_norm": 3.6376278400421143,
+      "learning_rate": 9.739078261907537e-06,
+      "loss": 0.7337,
+      "step": 2829
+    },
+    {
+      "epoch": 0.368606013276064,
+      "grad_norm": 3.675191640853882,
+      "learning_rate": 9.738411566250485e-06,
+      "loss": 0.6678,
+      "step": 2832
+    },
+    {
+      "epoch": 0.3689964857477548,
+      "grad_norm": 2.5543577671051025,
+      "learning_rate": 9.7377440428029e-06,
+      "loss": 0.577,
+      "step": 2835
+    },
+    {
+      "epoch": 0.3693869582194455,
+      "grad_norm": 3.843170166015625,
+      "learning_rate": 9.737075691681398e-06,
+      "loss": 0.7626,
+      "step": 2838
+    },
+    {
+      "epoch": 0.3697774306911363,
+      "grad_norm": 2.62917423248291,
+      "learning_rate": 9.736406513002737e-06,
+      "loss": 0.6669,
+      "step": 2841
+    },
+    {
+      "epoch": 0.370167903162827,
+      "grad_norm": 2.891164541244507,
+      "learning_rate": 9.735736506883822e-06,
+      "loss": 0.6709,
+      "step": 2844
+    },
+    {
+      "epoch": 0.37055837563451777,
+      "grad_norm": 2.4671757221221924,
+      "learning_rate": 9.735065673441702e-06,
+      "loss": 0.5747,
+      "step": 2847
+    },
+    {
+      "epoch": 0.37094884810620854,
+      "grad_norm": 3.6108286380767822,
+      "learning_rate": 9.73439401279357e-06,
+      "loss": 0.6485,
+      "step": 2850
+    },
+    {
+      "epoch": 0.37133932057789926,
+      "grad_norm": 2.4093971252441406,
+      "learning_rate": 9.733721525056764e-06,
+      "loss": 0.6907,
+      "step": 2853
+    },
+    {
+      "epoch": 0.37172979304959003,
+      "grad_norm": 4.289933681488037,
+      "learning_rate": 9.733048210348767e-06,
+      "loss": 0.6965,
+      "step": 2856
+    },
+    {
+      "epoch": 0.37212026552128075,
+      "grad_norm": 2.8142073154449463,
+      "learning_rate": 9.732374068787202e-06,
+      "loss": 0.7055,
+      "step": 2859
+    },
+    {
+      "epoch": 0.3725107379929715,
+      "grad_norm": 2.733341693878174,
+      "learning_rate": 9.731699100489845e-06,
+      "loss": 0.7459,
+      "step": 2862
+    },
+    {
+      "epoch": 0.37290121046466224,
+      "grad_norm": 3.4828221797943115,
+      "learning_rate": 9.731023305574608e-06,
+      "loss": 0.6431,
+      "step": 2865
+    },
+    {
+      "epoch": 0.373291682936353,
+      "grad_norm": 2.436321258544922,
+      "learning_rate": 9.730346684159553e-06,
+      "loss": 0.6435,
+      "step": 2868
+    },
+    {
+      "epoch": 0.3736821554080437,
+      "grad_norm": 2.3857946395874023,
+      "learning_rate": 9.729669236362882e-06,
+      "loss": 0.6081,
+      "step": 2871
+    },
+    {
+      "epoch": 0.3740726278797345,
+      "grad_norm": 3.848266839981079,
+      "learning_rate": 9.728990962302946e-06,
+      "loss": 0.6367,
+      "step": 2874
+    },
+    {
+      "epoch": 0.3744631003514252,
+      "grad_norm": 2.9143803119659424,
+      "learning_rate": 9.72831186209824e-06,
+      "loss": 0.6555,
+      "step": 2877
+    },
+    {
+      "epoch": 0.374853572823116,
+      "grad_norm": 2.645496129989624,
+      "learning_rate": 9.727631935867394e-06,
+      "loss": 0.6999,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3752440452948067,
+      "grad_norm": 2.874675750732422,
+      "learning_rate": 9.726951183729196e-06,
+      "loss": 0.6976,
+      "step": 2883
+    },
+    {
+      "epoch": 0.3756345177664975,
+      "grad_norm": 2.555006742477417,
+      "learning_rate": 9.726269605802569e-06,
+      "loss": 0.6693,
+      "step": 2886
+    },
+    {
+      "epoch": 0.3760249902381882,
+      "grad_norm": 2.870241641998291,
+      "learning_rate": 9.725587202206588e-06,
+      "loss": 0.6434,
+      "step": 2889
+    },
+    {
+      "epoch": 0.37641546270987897,
+      "grad_norm": 4.20858097076416,
+      "learning_rate": 9.724903973060461e-06,
+      "loss": 0.7726,
+      "step": 2892
+    },
+    {
+      "epoch": 0.3768059351815697,
+      "grad_norm": 2.6733803749084473,
+      "learning_rate": 9.72421991848355e-06,
+      "loss": 0.7146,
+      "step": 2895
+    },
+    {
+      "epoch": 0.37719640765326046,
+      "grad_norm": 2.6210622787475586,
+      "learning_rate": 9.723535038595358e-06,
+      "loss": 0.7285,
+      "step": 2898
+    },
+    {
+      "epoch": 0.3775868801249512,
+      "grad_norm": 3.7322444915771484,
+      "learning_rate": 9.722849333515532e-06,
+      "loss": 0.7671,
+      "step": 2901
+    },
+    {
+      "epoch": 0.37797735259664195,
+      "grad_norm": 3.78657603263855,
+      "learning_rate": 9.722162803363863e-06,
+      "loss": 0.6344,
+      "step": 2904
+    },
+    {
+      "epoch": 0.37836782506833266,
+      "grad_norm": 3.663477659225464,
+      "learning_rate": 9.721475448260286e-06,
+      "loss": 0.6887,
+      "step": 2907
+    },
+    {
+      "epoch": 0.37875829754002344,
+      "grad_norm": 2.8419249057769775,
+      "learning_rate": 9.72078726832488e-06,
+      "loss": 0.7098,
+      "step": 2910
+    },
+    {
+      "epoch": 0.37914877001171415,
+      "grad_norm": 5.145716190338135,
+      "learning_rate": 9.72009826367787e-06,
+      "loss": 0.6945,
+      "step": 2913
+    },
+    {
+      "epoch": 0.3795392424834049,
+      "grad_norm": 3.3549532890319824,
+      "learning_rate": 9.719408434439623e-06,
+      "loss": 0.6757,
+      "step": 2916
+    },
+    {
+      "epoch": 0.37992971495509564,
+      "grad_norm": 2.720118284225464,
+      "learning_rate": 9.71871778073065e-06,
+      "loss": 0.7617,
+      "step": 2919
+    },
+    {
+      "epoch": 0.3803201874267864,
+      "grad_norm": 2.3909714221954346,
+      "learning_rate": 9.718026302671608e-06,
+      "loss": 0.6084,
+      "step": 2922
+    },
+    {
+      "epoch": 0.38071065989847713,
+      "grad_norm": 2.6086878776550293,
+      "learning_rate": 9.717334000383297e-06,
+      "loss": 0.6376,
+      "step": 2925
+    },
+    {
+      "epoch": 0.3811011323701679,
+      "grad_norm": 2.780679702758789,
+      "learning_rate": 9.716640873986658e-06,
+      "loss": 0.712,
+      "step": 2928
+    },
+    {
+      "epoch": 0.3814916048418586,
+      "grad_norm": 2.720557451248169,
+      "learning_rate": 9.715946923602781e-06,
+      "loss": 0.643,
+      "step": 2931
+    },
+    {
+      "epoch": 0.3818820773135494,
+      "grad_norm": 5.16661262512207,
+      "learning_rate": 9.715252149352898e-06,
+      "loss": 0.705,
+      "step": 2934
+    },
+    {
+      "epoch": 0.38227254978524017,
+      "grad_norm": 3.2445664405822754,
+      "learning_rate": 9.714556551358385e-06,
+      "loss": 0.6784,
+      "step": 2937
+    },
+    {
+      "epoch": 0.3826630222569309,
+      "grad_norm": 3.182467460632324,
+      "learning_rate": 9.713860129740759e-06,
+      "loss": 0.714,
+      "step": 2940
+    },
+    {
+      "epoch": 0.38305349472862166,
+      "grad_norm": 2.596245288848877,
+      "learning_rate": 9.713162884621686e-06,
+      "loss": 0.5894,
+      "step": 2943
+    },
+    {
+      "epoch": 0.3834439672003124,
+      "grad_norm": 2.317925453186035,
+      "learning_rate": 9.71246481612297e-06,
+      "loss": 0.5575,
+      "step": 2946
+    },
+    {
+      "epoch": 0.38383443967200315,
+      "grad_norm": 2.5028085708618164,
+      "learning_rate": 9.711765924366567e-06,
+      "loss": 0.6765,
+      "step": 2949
+    },
+    {
+      "epoch": 0.38422491214369386,
+      "grad_norm": 2.46987247467041,
+      "learning_rate": 9.711066209474568e-06,
+      "loss": 0.7498,
+      "step": 2952
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 2.608255386352539,
+      "learning_rate": 9.710365671569214e-06,
+      "loss": 0.718,
+      "step": 2955
+    },
+    {
+      "epoch": 0.38500585708707535,
+      "grad_norm": 3.6582038402557373,
+      "learning_rate": 9.709664310772884e-06,
+      "loss": 0.7416,
+      "step": 2958
+    },
+    {
+      "epoch": 0.3853963295587661,
+      "grad_norm": 3.149799108505249,
+      "learning_rate": 9.708962127208105e-06,
+      "loss": 0.746,
+      "step": 2961
+    },
+    {
+      "epoch": 0.38578680203045684,
+      "grad_norm": 2.3268182277679443,
+      "learning_rate": 9.708259120997551e-06,
+      "loss": 0.5427,
+      "step": 2964
+    },
+    {
+      "epoch": 0.3861772745021476,
+      "grad_norm": 2.4898290634155273,
+      "learning_rate": 9.707555292264035e-06,
+      "loss": 0.6849,
+      "step": 2967
+    },
+    {
+      "epoch": 0.38656774697383833,
+      "grad_norm": 2.5410478115081787,
+      "learning_rate": 9.706850641130512e-06,
+      "loss": 0.6692,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3869582194455291,
+      "grad_norm": 2.798302173614502,
+      "learning_rate": 9.706145167720082e-06,
+      "loss": 0.5719,
+      "step": 2973
+    },
+    {
+      "epoch": 0.3873486919172198,
+      "grad_norm": 3.7535929679870605,
+      "learning_rate": 9.705438872155993e-06,
+      "loss": 0.7151,
+      "step": 2976
+    },
+    {
+      "epoch": 0.3877391643889106,
+      "grad_norm": 2.450899124145508,
+      "learning_rate": 9.70473175456163e-06,
+      "loss": 0.7171,
+      "step": 2979
+    },
+    {
+      "epoch": 0.3881296368606013,
+      "grad_norm": 3.063359260559082,
+      "learning_rate": 9.704023815060528e-06,
+      "loss": 0.7145,
+      "step": 2982
+    },
+    {
+      "epoch": 0.3885201093322921,
+      "grad_norm": 2.760772228240967,
+      "learning_rate": 9.703315053776362e-06,
+      "loss": 0.7893,
+      "step": 2985
+    },
+    {
+      "epoch": 0.3889105818039828,
+      "grad_norm": 2.4443037509918213,
+      "learning_rate": 9.702605470832948e-06,
+      "loss": 0.6993,
+      "step": 2988
+    },
+    {
+      "epoch": 0.3893010542756736,
+      "grad_norm": 2.651404619216919,
+      "learning_rate": 9.701895066354255e-06,
+      "loss": 0.6193,
+      "step": 2991
+    },
+    {
+      "epoch": 0.3896915267473643,
+      "grad_norm": 3.2075984477996826,
+      "learning_rate": 9.701183840464383e-06,
+      "loss": 0.6428,
+      "step": 2994
+    },
+    {
+      "epoch": 0.39008199921905506,
+      "grad_norm": 2.8586585521698,
+      "learning_rate": 9.700471793287582e-06,
+      "loss": 0.7017,
+      "step": 2997
+    },
+    {
+      "epoch": 0.3904724716907458,
+      "grad_norm": 2.386969804763794,
+      "learning_rate": 9.69975892494825e-06,
+      "loss": 0.5862,
+      "step": 3000
+    },
+    {
+      "epoch": 0.39086294416243655,
+      "grad_norm": 2.74922776222229,
+      "learning_rate": 9.69904523557092e-06,
+      "loss": 0.6879,
+      "step": 3003
+    },
+    {
+      "epoch": 0.39125341663412727,
+      "grad_norm": 2.8284764289855957,
+      "learning_rate": 9.698330725280271e-06,
+      "loss": 0.6019,
+      "step": 3006
+    },
+    {
+      "epoch": 0.39164388910581804,
+      "grad_norm": 4.191535472869873,
+      "learning_rate": 9.69761539420113e-06,
+      "loss": 0.6605,
+      "step": 3009
+    },
+    {
+      "epoch": 0.39203436157750876,
+      "grad_norm": 2.883096218109131,
+      "learning_rate": 9.696899242458462e-06,
+      "loss": 0.7188,
+      "step": 3012
+    },
+    {
+      "epoch": 0.39242483404919953,
+      "grad_norm": 4.879593372344971,
+      "learning_rate": 9.696182270177377e-06,
+      "loss": 0.6858,
+      "step": 3015
+    },
+    {
+      "epoch": 0.39281530652089025,
+      "grad_norm": 3.190108060836792,
+      "learning_rate": 9.695464477483127e-06,
+      "loss": 0.7377,
+      "step": 3018
+    },
+    {
+      "epoch": 0.393205778992581,
+      "grad_norm": 2.7659029960632324,
+      "learning_rate": 9.694745864501113e-06,
+      "loss": 0.6033,
+      "step": 3021
+    },
+    {
+      "epoch": 0.3935962514642718,
+      "grad_norm": 2.5670714378356934,
+      "learning_rate": 9.694026431356872e-06,
+      "loss": 0.6231,
+      "step": 3024
+    },
+    {
+      "epoch": 0.3939867239359625,
+      "grad_norm": 2.614166498184204,
+      "learning_rate": 9.693306178176086e-06,
+      "loss": 0.696,
+      "step": 3027
+    },
+    {
+      "epoch": 0.3943771964076533,
+      "grad_norm": 2.583684206008911,
+      "learning_rate": 9.692585105084588e-06,
+      "loss": 0.6598,
+      "step": 3030
+    },
+    {
+      "epoch": 0.394767668879344,
+      "grad_norm": 2.6008903980255127,
+      "learning_rate": 9.691863212208342e-06,
+      "loss": 0.6207,
+      "step": 3033
+    },
+    {
+      "epoch": 0.3951581413510348,
+      "grad_norm": 2.974407196044922,
+      "learning_rate": 9.691140499673462e-06,
+      "loss": 0.6982,
+      "step": 3036
+    },
+    {
+      "epoch": 0.3955486138227255,
+      "grad_norm": 3.109865665435791,
+      "learning_rate": 9.690416967606207e-06,
+      "loss": 0.8394,
+      "step": 3039
+    },
+    {
+      "epoch": 0.39593908629441626,
+      "grad_norm": 2.6200222969055176,
+      "learning_rate": 9.689692616132975e-06,
+      "loss": 0.6467,
+      "step": 3042
+    },
+    {
+      "epoch": 0.396329558766107,
+      "grad_norm": 2.555178642272949,
+      "learning_rate": 9.688967445380306e-06,
+      "loss": 0.7058,
+      "step": 3045
+    },
+    {
+      "epoch": 0.39672003123779775,
+      "grad_norm": 2.755859375,
+      "learning_rate": 9.688241455474892e-06,
+      "loss": 0.637,
+      "step": 3048
+    },
+    {
+      "epoch": 0.39711050370948847,
+      "grad_norm": 2.5034401416778564,
+      "learning_rate": 9.687514646543557e-06,
+      "loss": 0.7212,
+      "step": 3051
+    },
+    {
+      "epoch": 0.39750097618117924,
+      "grad_norm": 3.507744550704956,
+      "learning_rate": 9.686787018713273e-06,
+      "loss": 0.6072,
+      "step": 3054
+    },
+    {
+      "epoch": 0.39789144865286996,
+      "grad_norm": 3.1820497512817383,
+      "learning_rate": 9.686058572111157e-06,
+      "loss": 0.7151,
+      "step": 3057
+    },
+    {
+      "epoch": 0.39828192112456073,
+      "grad_norm": 2.979086399078369,
+      "learning_rate": 9.685329306864468e-06,
+      "loss": 0.8117,
+      "step": 3060
+    },
+    {
+      "epoch": 0.39867239359625145,
+      "grad_norm": 2.7010393142700195,
+      "learning_rate": 9.684599223100604e-06,
+      "loss": 0.6791,
+      "step": 3063
+    },
+    {
+      "epoch": 0.3990628660679422,
+      "grad_norm": 2.8512115478515625,
+      "learning_rate": 9.68386832094711e-06,
+      "loss": 0.6964,
+      "step": 3066
+    },
+    {
+      "epoch": 0.39945333853963294,
+      "grad_norm": 2.7704756259918213,
+      "learning_rate": 9.683136600531674e-06,
+      "loss": 0.7149,
+      "step": 3069
+    },
+    {
+      "epoch": 0.3998438110113237,
+      "grad_norm": 2.513730764389038,
+      "learning_rate": 9.682404061982129e-06,
+      "loss": 0.6607,
+      "step": 3072
+    },
+    {
+      "epoch": 0.40023428348301443,
+      "grad_norm": 2.799468517303467,
+      "learning_rate": 9.681670705426442e-06,
+      "loss": 0.6806,
+      "step": 3075
+    },
+    {
+      "epoch": 0.4006247559547052,
+      "grad_norm": 4.188741683959961,
+      "learning_rate": 9.680936530992731e-06,
+      "loss": 0.5816,
+      "step": 3078
+    },
+    {
+      "epoch": 0.4010152284263959,
+      "grad_norm": 2.830998659133911,
+      "learning_rate": 9.680201538809257e-06,
+      "loss": 0.6774,
+      "step": 3081
+    },
+    {
+      "epoch": 0.4014057008980867,
+      "grad_norm": 2.5481221675872803,
+      "learning_rate": 9.679465729004419e-06,
+      "loss": 0.5939,
+      "step": 3084
+    },
+    {
+      "epoch": 0.4017961733697774,
+      "grad_norm": 2.98976469039917,
+      "learning_rate": 9.678729101706763e-06,
+      "loss": 0.6127,
+      "step": 3087
+    },
+    {
+      "epoch": 0.4021866458414682,
+      "grad_norm": 3.0015671253204346,
+      "learning_rate": 9.677991657044973e-06,
+      "loss": 0.6858,
+      "step": 3090
+    },
+    {
+      "epoch": 0.4025771183131589,
+      "grad_norm": 3.737126350402832,
+      "learning_rate": 9.677253395147886e-06,
+      "loss": 0.7177,
+      "step": 3093
+    },
+    {
+      "epoch": 0.40296759078484967,
+      "grad_norm": 2.3965842723846436,
+      "learning_rate": 9.676514316144468e-06,
+      "loss": 0.6822,
+      "step": 3096
+    },
+    {
+      "epoch": 0.4033580632565404,
+      "grad_norm": 2.6666600704193115,
+      "learning_rate": 9.675774420163835e-06,
+      "loss": 0.6958,
+      "step": 3099
+    },
+    {
+      "epoch": 0.40374853572823116,
+      "grad_norm": 3.0011026859283447,
+      "learning_rate": 9.675033707335249e-06,
+      "loss": 0.7197,
+      "step": 3102
+    },
+    {
+      "epoch": 0.4041390081999219,
+      "grad_norm": 2.5443384647369385,
+      "learning_rate": 9.674292177788109e-06,
+      "loss": 0.6865,
+      "step": 3105
+    },
+    {
+      "epoch": 0.40452948067161265,
+      "grad_norm": 2.5568745136260986,
+      "learning_rate": 9.67354983165196e-06,
+      "loss": 0.624,
+      "step": 3108
+    },
+    {
+      "epoch": 0.4049199531433034,
+      "grad_norm": 4.514736652374268,
+      "learning_rate": 9.672806669056486e-06,
+      "loss": 0.6752,
+      "step": 3111
+    },
+    {
+      "epoch": 0.40531042561499414,
+      "grad_norm": 2.690507650375366,
+      "learning_rate": 9.672062690131516e-06,
+      "loss": 0.7393,
+      "step": 3114
+    },
+    {
+      "epoch": 0.4057008980866849,
+      "grad_norm": 3.8276305198669434,
+      "learning_rate": 9.671317895007025e-06,
+      "loss": 0.6249,
+      "step": 3117
+    },
+    {
+      "epoch": 0.40609137055837563,
+      "grad_norm": 2.517906427383423,
+      "learning_rate": 9.670572283813123e-06,
+      "loss": 0.5731,
+      "step": 3120
+    },
+    {
+      "epoch": 0.4064818430300664,
+      "grad_norm": 2.6286120414733887,
+      "learning_rate": 9.669825856680068e-06,
+      "loss": 0.6554,
+      "step": 3123
+    },
+    {
+      "epoch": 0.4068723155017571,
+      "grad_norm": 2.6419596672058105,
+      "learning_rate": 9.669078613738263e-06,
+      "loss": 0.6567,
+      "step": 3126
+    },
+    {
+      "epoch": 0.4072627879734479,
+      "grad_norm": 2.7058355808258057,
+      "learning_rate": 9.668330555118243e-06,
+      "loss": 0.8195,
+      "step": 3129
+    },
+    {
+      "epoch": 0.4076532604451386,
+      "grad_norm": 2.805713176727295,
+      "learning_rate": 9.667581680950698e-06,
+      "loss": 0.7018,
+      "step": 3132
+    },
+    {
+      "epoch": 0.4080437329168294,
+      "grad_norm": 2.345257520675659,
+      "learning_rate": 9.66683199136645e-06,
+      "loss": 0.6237,
+      "step": 3135
+    },
+    {
+      "epoch": 0.4084342053885201,
+      "grad_norm": 2.4524874687194824,
+      "learning_rate": 9.666081486496472e-06,
+      "loss": 0.6889,
+      "step": 3138
+    },
+    {
+      "epoch": 0.40882467786021087,
+      "grad_norm": 2.8738114833831787,
+      "learning_rate": 9.665330166471875e-06,
+      "loss": 0.6038,
+      "step": 3141
+    },
+    {
+      "epoch": 0.4092151503319016,
+      "grad_norm": 3.0100152492523193,
+      "learning_rate": 9.664578031423913e-06,
+      "loss": 0.6482,
+      "step": 3144
+    },
+    {
+      "epoch": 0.40960562280359236,
+      "grad_norm": 2.748427391052246,
+      "learning_rate": 9.663825081483979e-06,
+      "loss": 0.6699,
+      "step": 3147
+    },
+    {
+      "epoch": 0.4099960952752831,
+      "grad_norm": 2.4630000591278076,
+      "learning_rate": 9.663071316783617e-06,
+      "loss": 0.6064,
+      "step": 3150
+    },
+    {
+      "epoch": 0.41038656774697385,
+      "grad_norm": 3.131740093231201,
+      "learning_rate": 9.662316737454505e-06,
+      "loss": 0.6362,
+      "step": 3153
+    },
+    {
+      "epoch": 0.41077704021866457,
+      "grad_norm": 2.545337200164795,
+      "learning_rate": 9.661561343628465e-06,
+      "loss": 0.6804,
+      "step": 3156
+    },
+    {
+      "epoch": 0.41116751269035534,
+      "grad_norm": 2.7401156425476074,
+      "learning_rate": 9.660805135437468e-06,
+      "loss": 0.679,
+      "step": 3159
+    },
+    {
+      "epoch": 0.41155798516204606,
+      "grad_norm": 2.6069376468658447,
+      "learning_rate": 9.660048113013616e-06,
+      "loss": 0.6729,
+      "step": 3162
+    },
+    {
+      "epoch": 0.41194845763373683,
+      "grad_norm": 2.470309257507324,
+      "learning_rate": 9.659290276489164e-06,
+      "loss": 0.6852,
+      "step": 3165
+    },
+    {
+      "epoch": 0.41233893010542755,
+      "grad_norm": 2.4761130809783936,
+      "learning_rate": 9.658531625996502e-06,
+      "loss": 0.6221,
+      "step": 3168
+    },
+    {
+      "epoch": 0.4127294025771183,
+      "grad_norm": 2.603297472000122,
+      "learning_rate": 9.657772161668164e-06,
+      "loss": 0.645,
+      "step": 3171
+    },
+    {
+      "epoch": 0.41311987504880904,
+      "grad_norm": 5.437543869018555,
+      "learning_rate": 9.657011883636828e-06,
+      "loss": 0.7422,
+      "step": 3174
+    },
+    {
+      "epoch": 0.4135103475204998,
+      "grad_norm": 2.724668025970459,
+      "learning_rate": 9.656250792035314e-06,
+      "loss": 0.6882,
+      "step": 3177
+    },
+    {
+      "epoch": 0.4139008199921905,
+      "grad_norm": 2.5859644412994385,
+      "learning_rate": 9.655488886996582e-06,
+      "loss": 0.5831,
+      "step": 3180
+    },
+    {
+      "epoch": 0.4142912924638813,
+      "grad_norm": 2.3966543674468994,
+      "learning_rate": 9.654726168653733e-06,
+      "loss": 0.5881,
+      "step": 3183
+    },
+    {
+      "epoch": 0.414681764935572,
+      "grad_norm": 2.52030611038208,
+      "learning_rate": 9.653962637140016e-06,
+      "loss": 0.6998,
+      "step": 3186
+    },
+    {
+      "epoch": 0.4150722374072628,
+      "grad_norm": 2.5317649841308594,
+      "learning_rate": 9.653198292588816e-06,
+      "loss": 0.7423,
+      "step": 3189
+    },
+    {
+      "epoch": 0.4154627098789535,
+      "grad_norm": 2.806925058364868,
+      "learning_rate": 9.652433135133666e-06,
+      "loss": 0.7737,
+      "step": 3192
+    },
+    {
+      "epoch": 0.4158531823506443,
+      "grad_norm": 2.7214372158050537,
+      "learning_rate": 9.651667164908232e-06,
+      "loss": 0.6412,
+      "step": 3195
+    },
+    {
+      "epoch": 0.41624365482233505,
+      "grad_norm": 3.378222942352295,
+      "learning_rate": 9.65090038204633e-06,
+      "loss": 0.5841,
+      "step": 3198
+    },
+    {
+      "epoch": 0.41663412729402577,
+      "grad_norm": 5.052935600280762,
+      "learning_rate": 9.650132786681916e-06,
+      "loss": 0.6651,
+      "step": 3201
+    },
+    {
+      "epoch": 0.41702459976571654,
+      "grad_norm": 3.4297244548797607,
+      "learning_rate": 9.649364378949087e-06,
+      "loss": 0.6839,
+      "step": 3204
+    },
+    {
+      "epoch": 0.41741507223740726,
+      "grad_norm": 3.513136148452759,
+      "learning_rate": 9.64859515898208e-06,
+      "loss": 0.727,
+      "step": 3207
+    },
+    {
+      "epoch": 0.41780554470909803,
+      "grad_norm": 3.3228838443756104,
+      "learning_rate": 9.64782512691528e-06,
+      "loss": 0.7235,
+      "step": 3210
+    },
+    {
+      "epoch": 0.41819601718078875,
+      "grad_norm": 3.5547192096710205,
+      "learning_rate": 9.647054282883207e-06,
+      "loss": 0.7343,
+      "step": 3213
+    },
+    {
+      "epoch": 0.4185864896524795,
+      "grad_norm": 2.559657096862793,
+      "learning_rate": 9.646282627020527e-06,
+      "loss": 0.5984,
+      "step": 3216
+    },
+    {
+      "epoch": 0.41897696212417024,
+      "grad_norm": 2.8096330165863037,
+      "learning_rate": 9.645510159462047e-06,
+      "loss": 0.7274,
+      "step": 3219
+    },
+    {
+      "epoch": 0.419367434595861,
+      "grad_norm": 2.950134038925171,
+      "learning_rate": 9.644736880342714e-06,
+      "loss": 0.6025,
+      "step": 3222
+    },
+    {
+      "epoch": 0.4197579070675517,
+      "grad_norm": 2.600390911102295,
+      "learning_rate": 9.643962789797619e-06,
+      "loss": 0.7445,
+      "step": 3225
+    },
+    {
+      "epoch": 0.4201483795392425,
+      "grad_norm": 2.848917245864868,
+      "learning_rate": 9.643187887961993e-06,
+      "loss": 0.6197,
+      "step": 3228
+    },
+    {
+      "epoch": 0.4205388520109332,
+      "grad_norm": 2.5057928562164307,
+      "learning_rate": 9.642412174971214e-06,
+      "loss": 0.7009,
+      "step": 3231
+    },
+    {
+      "epoch": 0.420929324482624,
+      "grad_norm": 3.8111495971679688,
+      "learning_rate": 9.641635650960792e-06,
+      "loss": 0.589,
+      "step": 3234
+    },
+    {
+      "epoch": 0.4213197969543147,
+      "grad_norm": 4.075084686279297,
+      "learning_rate": 9.640858316066387e-06,
+      "loss": 0.6813,
+      "step": 3237
+    },
+    {
+      "epoch": 0.4217102694260055,
+      "grad_norm": 3.2589192390441895,
+      "learning_rate": 9.640080170423796e-06,
+      "loss": 0.6352,
+      "step": 3240
+    },
+    {
+      "epoch": 0.4221007418976962,
+      "grad_norm": 3.0924248695373535,
+      "learning_rate": 9.63930121416896e-06,
+      "loss": 0.581,
+      "step": 3243
+    },
+    {
+      "epoch": 0.42249121436938697,
+      "grad_norm": 3.0321390628814697,
+      "learning_rate": 9.638521447437965e-06,
+      "loss": 0.7448,
+      "step": 3246
+    },
+    {
+      "epoch": 0.4228816868410777,
+      "grad_norm": 2.8482680320739746,
+      "learning_rate": 9.637740870367028e-06,
+      "loss": 0.7065,
+      "step": 3249
+    },
+    {
+      "epoch": 0.42327215931276846,
+      "grad_norm": 2.815291404724121,
+      "learning_rate": 9.636959483092518e-06,
+      "loss": 0.5734,
+      "step": 3252
+    },
+    {
+      "epoch": 0.4236626317844592,
+      "grad_norm": 2.5750129222869873,
+      "learning_rate": 9.636177285750942e-06,
+      "loss": 0.6645,
+      "step": 3255
+    },
+    {
+      "epoch": 0.42405310425614995,
+      "grad_norm": 2.154196262359619,
+      "learning_rate": 9.635394278478947e-06,
+      "loss": 0.6641,
+      "step": 3258
+    },
+    {
+      "epoch": 0.42444357672784067,
+      "grad_norm": 2.7113664150238037,
+      "learning_rate": 9.634610461413322e-06,
+      "loss": 0.6695,
+      "step": 3261
+    },
+    {
+      "epoch": 0.42483404919953144,
+      "grad_norm": 2.7851827144622803,
+      "learning_rate": 9.633825834691e-06,
+      "loss": 0.6974,
+      "step": 3264
+    },
+    {
+      "epoch": 0.42522452167122216,
+      "grad_norm": 2.6589601039886475,
+      "learning_rate": 9.633040398449052e-06,
+      "loss": 0.7072,
+      "step": 3267
+    },
+    {
+      "epoch": 0.42561499414291293,
+      "grad_norm": 2.8155033588409424,
+      "learning_rate": 9.632254152824693e-06,
+      "loss": 0.652,
+      "step": 3270
+    },
+    {
+      "epoch": 0.42600546661460365,
+      "grad_norm": 3.5873613357543945,
+      "learning_rate": 9.63146709795528e-06,
+      "loss": 0.6695,
+      "step": 3273
+    },
+    {
+      "epoch": 0.4263959390862944,
+      "grad_norm": 2.7894959449768066,
+      "learning_rate": 9.630679233978308e-06,
+      "loss": 0.6908,
+      "step": 3276
+    },
+    {
+      "epoch": 0.42678641155798513,
+      "grad_norm": 2.545112371444702,
+      "learning_rate": 9.629890561031414e-06,
+      "loss": 0.6758,
+      "step": 3279
+    },
+    {
+      "epoch": 0.4271768840296759,
+      "grad_norm": 2.591367483139038,
+      "learning_rate": 9.629101079252379e-06,
+      "loss": 0.6303,
+      "step": 3282
+    },
+    {
+      "epoch": 0.4275673565013667,
+      "grad_norm": 2.6541688442230225,
+      "learning_rate": 9.628310788779125e-06,
+      "loss": 0.728,
+      "step": 3285
+    },
+    {
+      "epoch": 0.4279578289730574,
+      "grad_norm": 2.391240358352661,
+      "learning_rate": 9.627519689749711e-06,
+      "loss": 0.6469,
+      "step": 3288
+    },
+    {
+      "epoch": 0.42834830144474817,
+      "grad_norm": 2.4447021484375,
+      "learning_rate": 9.626727782302343e-06,
+      "loss": 0.6447,
+      "step": 3291
+    },
+    {
+      "epoch": 0.4287387739164389,
+      "grad_norm": 2.9437851905822754,
+      "learning_rate": 9.625935066575364e-06,
+      "loss": 0.7597,
+      "step": 3294
+    },
+    {
+      "epoch": 0.42912924638812966,
+      "grad_norm": 2.9596848487854004,
+      "learning_rate": 9.625141542707261e-06,
+      "loss": 0.7102,
+      "step": 3297
+    },
+    {
+      "epoch": 0.4295197188598204,
+      "grad_norm": 3.0737996101379395,
+      "learning_rate": 9.62434721083666e-06,
+      "loss": 0.597,
+      "step": 3300
+    },
+    {
+      "epoch": 0.42991019133151115,
+      "grad_norm": 2.6365230083465576,
+      "learning_rate": 9.62355207110233e-06,
+      "loss": 0.6877,
+      "step": 3303
+    },
+    {
+      "epoch": 0.43030066380320187,
+      "grad_norm": 3.259915351867676,
+      "learning_rate": 9.62275612364318e-06,
+      "loss": 0.6127,
+      "step": 3306
+    },
+    {
+      "epoch": 0.43069113627489264,
+      "grad_norm": 2.5275659561157227,
+      "learning_rate": 9.621959368598259e-06,
+      "loss": 0.5959,
+      "step": 3309
+    },
+    {
+      "epoch": 0.43108160874658336,
+      "grad_norm": 2.8045878410339355,
+      "learning_rate": 9.62116180610676e-06,
+      "loss": 0.582,
+      "step": 3312
+    },
+    {
+      "epoch": 0.43147208121827413,
+      "grad_norm": 2.7796387672424316,
+      "learning_rate": 9.620363436308015e-06,
+      "loss": 0.714,
+      "step": 3315
+    },
+    {
+      "epoch": 0.43186255368996485,
+      "grad_norm": 2.583375930786133,
+      "learning_rate": 9.619564259341497e-06,
+      "loss": 0.7435,
+      "step": 3318
+    },
+    {
+      "epoch": 0.4322530261616556,
+      "grad_norm": 2.662533760070801,
+      "learning_rate": 9.618764275346821e-06,
+      "loss": 0.6005,
+      "step": 3321
+    },
+    {
+      "epoch": 0.43264349863334634,
+      "grad_norm": 2.9329593181610107,
+      "learning_rate": 9.617963484463744e-06,
+      "loss": 0.6643,
+      "step": 3324
+    },
+    {
+      "epoch": 0.4330339711050371,
+      "grad_norm": 2.6769914627075195,
+      "learning_rate": 9.61716188683216e-06,
+      "loss": 0.745,
+      "step": 3327
+    },
+    {
+      "epoch": 0.4334244435767278,
+      "grad_norm": 2.9171302318573,
+      "learning_rate": 9.616359482592108e-06,
+      "loss": 0.7305,
+      "step": 3330
+    },
+    {
+      "epoch": 0.4338149160484186,
+      "grad_norm": 3.290548086166382,
+      "learning_rate": 9.615556271883766e-06,
+      "loss": 0.7274,
+      "step": 3333
+    },
+    {
+      "epoch": 0.4342053885201093,
+      "grad_norm": 2.999380588531494,
+      "learning_rate": 9.61475225484745e-06,
+      "loss": 0.628,
+      "step": 3336
+    },
+    {
+      "epoch": 0.4345958609918001,
+      "grad_norm": 2.652371644973755,
+      "learning_rate": 9.613947431623627e-06,
+      "loss": 0.6668,
+      "step": 3339
+    },
+    {
+      "epoch": 0.4349863334634908,
+      "grad_norm": 2.5840189456939697,
+      "learning_rate": 9.613141802352893e-06,
+      "loss": 0.664,
+      "step": 3342
+    },
+    {
+      "epoch": 0.4353768059351816,
+      "grad_norm": 2.5839924812316895,
+      "learning_rate": 9.61233536717599e-06,
+      "loss": 0.7491,
+      "step": 3345
+    },
+    {
+      "epoch": 0.4357672784068723,
+      "grad_norm": 2.6329665184020996,
+      "learning_rate": 9.6115281262338e-06,
+      "loss": 0.6435,
+      "step": 3348
+    },
+    {
+      "epoch": 0.43615775087856307,
+      "grad_norm": 2.31581974029541,
+      "learning_rate": 9.61072007966735e-06,
+      "loss": 0.6642,
+      "step": 3351
+    },
+    {
+      "epoch": 0.4365482233502538,
+      "grad_norm": 3.6810152530670166,
+      "learning_rate": 9.609911227617802e-06,
+      "loss": 0.677,
+      "step": 3354
+    },
+    {
+      "epoch": 0.43693869582194456,
+      "grad_norm": 2.4327340126037598,
+      "learning_rate": 9.609101570226458e-06,
+      "loss": 0.6408,
+      "step": 3357
+    },
+    {
+      "epoch": 0.4373291682936353,
+      "grad_norm": 2.6379623413085938,
+      "learning_rate": 9.608291107634767e-06,
+      "loss": 0.6083,
+      "step": 3360
+    },
+    {
+      "epoch": 0.43771964076532605,
+      "grad_norm": 2.7450592517852783,
+      "learning_rate": 9.607479839984313e-06,
+      "loss": 0.6663,
+      "step": 3363
+    },
+    {
+      "epoch": 0.43811011323701676,
+      "grad_norm": 2.3480405807495117,
+      "learning_rate": 9.606667767416824e-06,
+      "loss": 0.5952,
+      "step": 3366
+    },
+    {
+      "epoch": 0.43850058570870754,
+      "grad_norm": 2.8158650398254395,
+      "learning_rate": 9.605854890074169e-06,
+      "loss": 0.745,
+      "step": 3369
+    },
+    {
+      "epoch": 0.4388910581803983,
+      "grad_norm": 2.310330390930176,
+      "learning_rate": 9.60504120809835e-06,
+      "loss": 0.5277,
+      "step": 3372
+    },
+    {
+      "epoch": 0.439281530652089,
+      "grad_norm": 2.6338181495666504,
+      "learning_rate": 9.604226721631525e-06,
+      "loss": 0.6792,
+      "step": 3375
+    },
+    {
+      "epoch": 0.4396720031237798,
+      "grad_norm": 2.5001606941223145,
+      "learning_rate": 9.603411430815974e-06,
+      "loss": 0.61,
+      "step": 3378
+    },
+    {
+      "epoch": 0.4400624755954705,
+      "grad_norm": 2.8085408210754395,
+      "learning_rate": 9.60259533579413e-06,
+      "loss": 0.6936,
+      "step": 3381
+    },
+    {
+      "epoch": 0.4404529480671613,
+      "grad_norm": 2.701272964477539,
+      "learning_rate": 9.601778436708564e-06,
+      "loss": 0.584,
+      "step": 3384
+    },
+    {
+      "epoch": 0.440843420538852,
+      "grad_norm": 2.5402166843414307,
+      "learning_rate": 9.600960733701988e-06,
+      "loss": 0.6306,
+      "step": 3387
+    },
+    {
+      "epoch": 0.4412338930105428,
+      "grad_norm": 2.4810187816619873,
+      "learning_rate": 9.600142226917248e-06,
+      "loss": 0.6259,
+      "step": 3390
+    },
+    {
+      "epoch": 0.4416243654822335,
+      "grad_norm": 2.939995765686035,
+      "learning_rate": 9.599322916497338e-06,
+      "loss": 0.6119,
+      "step": 3393
+    },
+    {
+      "epoch": 0.44201483795392427,
+      "grad_norm": 2.663740873336792,
+      "learning_rate": 9.598502802585392e-06,
+      "loss": 0.6582,
+      "step": 3396
+    },
+    {
+      "epoch": 0.442405310425615,
+      "grad_norm": 2.5204548835754395,
+      "learning_rate": 9.597681885324679e-06,
+      "loss": 0.7289,
+      "step": 3399
+    },
+    {
+      "epoch": 0.44279578289730576,
+      "grad_norm": 2.44008207321167,
+      "learning_rate": 9.596860164858612e-06,
+      "loss": 0.5731,
+      "step": 3402
+    },
+    {
+      "epoch": 0.4431862553689965,
+      "grad_norm": 2.584868907928467,
+      "learning_rate": 9.596037641330746e-06,
+      "loss": 0.6289,
+      "step": 3405
+    },
+    {
+      "epoch": 0.44357672784068725,
+      "grad_norm": 3.060908079147339,
+      "learning_rate": 9.595214314884773e-06,
+      "loss": 0.7144,
+      "step": 3408
+    },
+    {
+      "epoch": 0.44396720031237796,
+      "grad_norm": 2.6084563732147217,
+      "learning_rate": 9.594390185664526e-06,
+      "loss": 0.6185,
+      "step": 3411
+    },
+    {
+      "epoch": 0.44435767278406874,
+      "grad_norm": 2.7041709423065186,
+      "learning_rate": 9.593565253813978e-06,
+      "loss": 0.6793,
+      "step": 3414
+    },
+    {
+      "epoch": 0.44474814525575945,
+      "grad_norm": 2.9432315826416016,
+      "learning_rate": 9.592739519477243e-06,
+      "loss": 0.6493,
+      "step": 3417
+    },
+    {
+      "epoch": 0.4451386177274502,
+      "grad_norm": 2.6161787509918213,
+      "learning_rate": 9.591912982798576e-06,
+      "loss": 0.6717,
+      "step": 3420
+    },
+    {
+      "epoch": 0.44552909019914094,
+      "grad_norm": 3.7241930961608887,
+      "learning_rate": 9.591085643922372e-06,
+      "loss": 0.7395,
+      "step": 3423
+    },
+    {
+      "epoch": 0.4459195626708317,
+      "grad_norm": 2.6377596855163574,
+      "learning_rate": 9.590257502993164e-06,
+      "loss": 0.733,
+      "step": 3426
+    },
+    {
+      "epoch": 0.44631003514252243,
+      "grad_norm": 2.6171798706054688,
+      "learning_rate": 9.589428560155627e-06,
+      "loss": 0.6539,
+      "step": 3429
+    },
+    {
+      "epoch": 0.4467005076142132,
+      "grad_norm": 3.0598702430725098,
+      "learning_rate": 9.588598815554573e-06,
+      "loss": 0.7271,
+      "step": 3432
+    },
+    {
+      "epoch": 0.4470909800859039,
+      "grad_norm": 2.866503953933716,
+      "learning_rate": 9.587768269334962e-06,
+      "loss": 0.6864,
+      "step": 3435
+    },
+    {
+      "epoch": 0.4474814525575947,
+      "grad_norm": 2.591628074645996,
+      "learning_rate": 9.586936921641884e-06,
+      "loss": 0.6335,
+      "step": 3438
+    },
+    {
+      "epoch": 0.4478719250292854,
+      "grad_norm": 2.506284236907959,
+      "learning_rate": 9.586104772620575e-06,
+      "loss": 0.6376,
+      "step": 3441
+    },
+    {
+      "epoch": 0.4482623975009762,
+      "grad_norm": 2.7341339588165283,
+      "learning_rate": 9.585271822416412e-06,
+      "loss": 0.7072,
+      "step": 3444
+    },
+    {
+      "epoch": 0.4486528699726669,
+      "grad_norm": 2.406806230545044,
+      "learning_rate": 9.584438071174904e-06,
+      "loss": 0.5989,
+      "step": 3447
+    },
+    {
+      "epoch": 0.4490433424443577,
+      "grad_norm": 2.4429965019226074,
+      "learning_rate": 9.583603519041713e-06,
+      "loss": 0.6158,
+      "step": 3450
+    },
+    {
+      "epoch": 0.4494338149160484,
+      "grad_norm": 2.7472376823425293,
+      "learning_rate": 9.582768166162628e-06,
+      "loss": 0.6195,
+      "step": 3453
+    },
+    {
+      "epoch": 0.44982428738773916,
+      "grad_norm": 2.816849946975708,
+      "learning_rate": 9.581932012683583e-06,
+      "loss": 0.7533,
+      "step": 3456
+    },
+    {
+      "epoch": 0.45021475985942994,
+      "grad_norm": 2.6621181964874268,
+      "learning_rate": 9.581095058750658e-06,
+      "loss": 0.6995,
+      "step": 3459
+    },
+    {
+      "epoch": 0.45060523233112065,
+      "grad_norm": 2.8741328716278076,
+      "learning_rate": 9.580257304510062e-06,
+      "loss": 0.6234,
+      "step": 3462
+    },
+    {
+      "epoch": 0.4509957048028114,
+      "grad_norm": 2.814319372177124,
+      "learning_rate": 9.579418750108149e-06,
+      "loss": 0.7944,
+      "step": 3465
+    },
+    {
+      "epoch": 0.45138617727450214,
+      "grad_norm": 2.745290517807007,
+      "learning_rate": 9.578579395691417e-06,
+      "loss": 0.6016,
+      "step": 3468
+    },
+    {
+      "epoch": 0.4517766497461929,
+      "grad_norm": 4.0121541023254395,
+      "learning_rate": 9.577739241406494e-06,
+      "loss": 0.6709,
+      "step": 3471
+    },
+    {
+      "epoch": 0.45216712221788363,
+      "grad_norm": 2.6668434143066406,
+      "learning_rate": 9.576898287400155e-06,
+      "loss": 0.6304,
+      "step": 3474
+    },
+    {
+      "epoch": 0.4525575946895744,
+      "grad_norm": 2.569227933883667,
+      "learning_rate": 9.576056533819316e-06,
+      "loss": 0.6098,
+      "step": 3477
+    },
+    {
+      "epoch": 0.4529480671612651,
+      "grad_norm": 2.43481183052063,
+      "learning_rate": 9.575213980811027e-06,
+      "loss": 0.6843,
+      "step": 3480
+    },
+    {
+      "epoch": 0.4533385396329559,
+      "grad_norm": 2.772892713546753,
+      "learning_rate": 9.57437062852248e-06,
+      "loss": 0.6885,
+      "step": 3483
+    },
+    {
+      "epoch": 0.4537290121046466,
+      "grad_norm": 2.392486333847046,
+      "learning_rate": 9.573526477101006e-06,
+      "loss": 0.6022,
+      "step": 3486
+    },
+    {
+      "epoch": 0.4541194845763374,
+      "grad_norm": 2.679281234741211,
+      "learning_rate": 9.572681526694079e-06,
+      "loss": 0.6724,
+      "step": 3489
+    },
+    {
+      "epoch": 0.4545099570480281,
+      "grad_norm": 2.970829725265503,
+      "learning_rate": 9.571835777449307e-06,
+      "loss": 0.6936,
+      "step": 3492
+    },
+    {
+      "epoch": 0.4549004295197189,
+      "grad_norm": 2.3695054054260254,
+      "learning_rate": 9.570989229514445e-06,
+      "loss": 0.5728,
+      "step": 3495
+    },
+    {
+      "epoch": 0.4552909019914096,
+      "grad_norm": 2.7799527645111084,
+      "learning_rate": 9.57014188303738e-06,
+      "loss": 0.6046,
+      "step": 3498
+    },
+    {
+      "epoch": 0.45568137446310036,
+      "grad_norm": 2.7889013290405273,
+      "learning_rate": 9.569293738166141e-06,
+      "loss": 0.7424,
+      "step": 3501
+    },
+    {
+      "epoch": 0.4560718469347911,
+      "grad_norm": 4.006762504577637,
+      "learning_rate": 9.568444795048899e-06,
+      "loss": 0.62,
+      "step": 3504
+    },
+    {
+      "epoch": 0.45646231940648185,
+      "grad_norm": 2.510829448699951,
+      "learning_rate": 9.567595053833963e-06,
+      "loss": 0.6677,
+      "step": 3507
+    },
+    {
+      "epoch": 0.45685279187817257,
+      "grad_norm": 2.6628668308258057,
+      "learning_rate": 9.566744514669777e-06,
+      "loss": 0.6483,
+      "step": 3510
+    },
+    {
+      "epoch": 0.45724326434986334,
+      "grad_norm": 2.8696441650390625,
+      "learning_rate": 9.565893177704934e-06,
+      "loss": 0.6337,
+      "step": 3513
+    },
+    {
+      "epoch": 0.45763373682155406,
+      "grad_norm": 2.641186237335205,
+      "learning_rate": 9.565041043088157e-06,
+      "loss": 0.6788,
+      "step": 3516
+    },
+    {
+      "epoch": 0.45802420929324483,
+      "grad_norm": 3.316789150238037,
+      "learning_rate": 9.564188110968314e-06,
+      "loss": 0.6791,
+      "step": 3519
+    },
+    {
+      "epoch": 0.45841468176493555,
+      "grad_norm": 2.8417608737945557,
+      "learning_rate": 9.563334381494409e-06,
+      "loss": 0.6841,
+      "step": 3522
+    },
+    {
+      "epoch": 0.4588051542366263,
+      "grad_norm": 3.1960368156433105,
+      "learning_rate": 9.562479854815587e-06,
+      "loss": 0.6391,
+      "step": 3525
+    },
+    {
+      "epoch": 0.45919562670831704,
+      "grad_norm": 2.585097074508667,
+      "learning_rate": 9.561624531081132e-06,
+      "loss": 0.7299,
+      "step": 3528
+    },
+    {
+      "epoch": 0.4595860991800078,
+      "grad_norm": 2.7368311882019043,
+      "learning_rate": 9.56076841044047e-06,
+      "loss": 0.6428,
+      "step": 3531
+    },
+    {
+      "epoch": 0.45997657165169853,
+      "grad_norm": 2.4325826168060303,
+      "learning_rate": 9.559911493043159e-06,
+      "loss": 0.617,
+      "step": 3534
+    },
+    {
+      "epoch": 0.4603670441233893,
+      "grad_norm": 2.53668212890625,
+      "learning_rate": 9.559053779038903e-06,
+      "loss": 0.6855,
+      "step": 3537
+    },
+    {
+      "epoch": 0.4607575165950801,
+      "grad_norm": 3.207590341567993,
+      "learning_rate": 9.558195268577543e-06,
+      "loss": 0.6649,
+      "step": 3540
+    },
+    {
+      "epoch": 0.4611479890667708,
+      "grad_norm": 2.7838118076324463,
+      "learning_rate": 9.557335961809059e-06,
+      "loss": 0.6823,
+      "step": 3543
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 2.8149774074554443,
+      "learning_rate": 9.556475858883569e-06,
+      "loss": 0.6601,
+      "step": 3546
+    },
+    {
+      "epoch": 0.4619289340101523,
+      "grad_norm": 2.708364725112915,
+      "learning_rate": 9.555614959951333e-06,
+      "loss": 0.7265,
+      "step": 3549
+    },
+    {
+      "epoch": 0.46231940648184305,
+      "grad_norm": 3.0776867866516113,
+      "learning_rate": 9.554753265162746e-06,
+      "loss": 0.573,
+      "step": 3552
+    },
+    {
+      "epoch": 0.46270987895353377,
+      "grad_norm": 2.5604937076568604,
+      "learning_rate": 9.553890774668347e-06,
+      "loss": 0.6187,
+      "step": 3555
+    },
+    {
+      "epoch": 0.46310035142522454,
+      "grad_norm": 3.412839651107788,
+      "learning_rate": 9.553027488618806e-06,
+      "loss": 0.6175,
+      "step": 3558
+    },
+    {
+      "epoch": 0.46349082389691526,
+      "grad_norm": 2.88525128364563,
+      "learning_rate": 9.552163407164945e-06,
+      "loss": 0.7192,
+      "step": 3561
+    },
+    {
+      "epoch": 0.46388129636860603,
+      "grad_norm": 3.187467098236084,
+      "learning_rate": 9.551298530457711e-06,
+      "loss": 0.6381,
+      "step": 3564
+    },
+    {
+      "epoch": 0.46427176884029675,
+      "grad_norm": 2.5330684185028076,
+      "learning_rate": 9.5504328586482e-06,
+      "loss": 0.7419,
+      "step": 3567
+    },
+    {
+      "epoch": 0.4646622413119875,
+      "grad_norm": 3.115802049636841,
+      "learning_rate": 9.54956639188764e-06,
+      "loss": 0.6623,
+      "step": 3570
+    },
+    {
+      "epoch": 0.46505271378367824,
+      "grad_norm": 2.369229555130005,
+      "learning_rate": 9.548699130327401e-06,
+      "loss": 0.6891,
+      "step": 3573
+    },
+    {
+      "epoch": 0.465443186255369,
+      "grad_norm": 3.0128095149993896,
+      "learning_rate": 9.547831074118995e-06,
+      "loss": 0.6068,
+      "step": 3576
+    },
+    {
+      "epoch": 0.46583365872705973,
+      "grad_norm": 2.500648021697998,
+      "learning_rate": 9.546962223414067e-06,
+      "loss": 0.7091,
+      "step": 3579
+    },
+    {
+      "epoch": 0.4662241311987505,
+      "grad_norm": 2.3269920349121094,
+      "learning_rate": 9.546092578364403e-06,
+      "loss": 0.6648,
+      "step": 3582
+    },
+    {
+      "epoch": 0.4666146036704412,
+      "grad_norm": 2.717773675918579,
+      "learning_rate": 9.545222139121931e-06,
+      "loss": 0.5747,
+      "step": 3585
+    },
+    {
+      "epoch": 0.467005076142132,
+      "grad_norm": 3.3133246898651123,
+      "learning_rate": 9.544350905838712e-06,
+      "loss": 0.6512,
+      "step": 3588
+    },
+    {
+      "epoch": 0.4673955486138227,
+      "grad_norm": 2.5333547592163086,
+      "learning_rate": 9.54347887866695e-06,
+      "loss": 0.5804,
+      "step": 3591
+    },
+    {
+      "epoch": 0.4677860210855135,
+      "grad_norm": 2.5862746238708496,
+      "learning_rate": 9.542606057758986e-06,
+      "loss": 0.6361,
+      "step": 3594
+    },
+    {
+      "epoch": 0.4681764935572042,
+      "grad_norm": 2.544423818588257,
+      "learning_rate": 9.5417324432673e-06,
+      "loss": 0.6235,
+      "step": 3597
+    },
+    {
+      "epoch": 0.46856696602889497,
+      "grad_norm": 2.701850652694702,
+      "learning_rate": 9.540858035344509e-06,
+      "loss": 0.725,
+      "step": 3600
+    },
+    {
+      "epoch": 0.4689574385005857,
+      "grad_norm": 2.778111457824707,
+      "learning_rate": 9.539982834143373e-06,
+      "loss": 0.5713,
+      "step": 3603
+    },
+    {
+      "epoch": 0.46934791097227646,
+      "grad_norm": 4.258044719696045,
+      "learning_rate": 9.539106839816787e-06,
+      "loss": 0.7481,
+      "step": 3606
+    },
+    {
+      "epoch": 0.4697383834439672,
+      "grad_norm": 4.864741325378418,
+      "learning_rate": 9.538230052517786e-06,
+      "loss": 0.7349,
+      "step": 3609
+    },
+    {
+      "epoch": 0.47012885591565795,
+      "grad_norm": 2.560973882675171,
+      "learning_rate": 9.53735247239954e-06,
+      "loss": 0.5515,
+      "step": 3612
+    },
+    {
+      "epoch": 0.47051932838734867,
+      "grad_norm": 2.455519676208496,
+      "learning_rate": 9.536474099615362e-06,
+      "loss": 0.6881,
+      "step": 3615
+    },
+    {
+      "epoch": 0.47090980085903944,
+      "grad_norm": 2.9638755321502686,
+      "learning_rate": 9.535594934318702e-06,
+      "loss": 0.6421,
+      "step": 3618
+    },
+    {
+      "epoch": 0.47130027333073016,
+      "grad_norm": 3.095149517059326,
+      "learning_rate": 9.53471497666315e-06,
+      "loss": 0.6678,
+      "step": 3621
+    },
+    {
+      "epoch": 0.47169074580242093,
+      "grad_norm": 3.0638108253479004,
+      "learning_rate": 9.533834226802431e-06,
+      "loss": 0.6058,
+      "step": 3624
+    },
+    {
+      "epoch": 0.4720812182741117,
+      "grad_norm": 2.333308696746826,
+      "learning_rate": 9.53295268489041e-06,
+      "loss": 0.5788,
+      "step": 3627
+    },
+    {
+      "epoch": 0.4724716907458024,
+      "grad_norm": 3.2323508262634277,
+      "learning_rate": 9.53207035108109e-06,
+      "loss": 0.7214,
+      "step": 3630
+    },
+    {
+      "epoch": 0.4728621632174932,
+      "grad_norm": 2.5078964233398438,
+      "learning_rate": 9.531187225528615e-06,
+      "loss": 0.7241,
+      "step": 3633
+    },
+    {
+      "epoch": 0.4732526356891839,
+      "grad_norm": 2.5901105403900146,
+      "learning_rate": 9.530303308387263e-06,
+      "loss": 0.707,
+      "step": 3636
+    },
+    {
+      "epoch": 0.4736431081608747,
+      "grad_norm": 2.5510504245758057,
+      "learning_rate": 9.529418599811455e-06,
+      "loss": 0.7274,
+      "step": 3639
+    },
+    {
+      "epoch": 0.4740335806325654,
+      "grad_norm": 2.424675703048706,
+      "learning_rate": 9.528533099955745e-06,
+      "loss": 0.6885,
+      "step": 3642
+    },
+    {
+      "epoch": 0.47442405310425617,
+      "grad_norm": 3.0174453258514404,
+      "learning_rate": 9.527646808974828e-06,
+      "loss": 0.6603,
+      "step": 3645
+    },
+    {
+      "epoch": 0.4748145255759469,
+      "grad_norm": 2.4409093856811523,
+      "learning_rate": 9.52675972702354e-06,
+      "loss": 0.684,
+      "step": 3648
+    },
+    {
+      "epoch": 0.47520499804763766,
+      "grad_norm": 2.731062889099121,
+      "learning_rate": 9.52587185425685e-06,
+      "loss": 0.7147,
+      "step": 3651
+    },
+    {
+      "epoch": 0.4755954705193284,
+      "grad_norm": 2.9121267795562744,
+      "learning_rate": 9.524983190829868e-06,
+      "loss": 0.7414,
+      "step": 3654
+    },
+    {
+      "epoch": 0.47598594299101915,
+      "grad_norm": 2.500781297683716,
+      "learning_rate": 9.52409373689784e-06,
+      "loss": 0.6857,
+      "step": 3657
+    },
+    {
+      "epoch": 0.47637641546270987,
+      "grad_norm": 2.709920644760132,
+      "learning_rate": 9.523203492616158e-06,
+      "loss": 0.6614,
+      "step": 3660
+    },
+    {
+      "epoch": 0.47676688793440064,
+      "grad_norm": 2.8489487171173096,
+      "learning_rate": 9.522312458140338e-06,
+      "loss": 0.5616,
+      "step": 3663
+    },
+    {
+      "epoch": 0.47715736040609136,
+      "grad_norm": 3.2674407958984375,
+      "learning_rate": 9.521420633626045e-06,
+      "loss": 0.5813,
+      "step": 3666
+    },
+    {
+      "epoch": 0.47754783287778213,
+      "grad_norm": 2.6726903915405273,
+      "learning_rate": 9.52052801922908e-06,
+      "loss": 0.6633,
+      "step": 3669
+    },
+    {
+      "epoch": 0.47793830534947285,
+      "grad_norm": 2.6989340782165527,
+      "learning_rate": 9.51963461510538e-06,
+      "loss": 0.6985,
+      "step": 3672
+    },
+    {
+      "epoch": 0.4783287778211636,
+      "grad_norm": 2.936947822570801,
+      "learning_rate": 9.51874042141102e-06,
+      "loss": 0.6472,
+      "step": 3675
+    },
+    {
+      "epoch": 0.47871925029285434,
+      "grad_norm": 2.9338979721069336,
+      "learning_rate": 9.517845438302213e-06,
+      "loss": 0.6153,
+      "step": 3678
+    },
+    {
+      "epoch": 0.4791097227645451,
+      "grad_norm": 2.5126919746398926,
+      "learning_rate": 9.516949665935314e-06,
+      "loss": 0.6328,
+      "step": 3681
+    },
+    {
+      "epoch": 0.4795001952362358,
+      "grad_norm": 2.4342703819274902,
+      "learning_rate": 9.516053104466811e-06,
+      "loss": 0.6888,
+      "step": 3684
+    },
+    {
+      "epoch": 0.4798906677079266,
+      "grad_norm": 3.6857855319976807,
+      "learning_rate": 9.51515575405333e-06,
+      "loss": 0.614,
+      "step": 3687
+    },
+    {
+      "epoch": 0.4802811401796173,
+      "grad_norm": 2.602360963821411,
+      "learning_rate": 9.514257614851639e-06,
+      "loss": 0.7064,
+      "step": 3690
+    },
+    {
+      "epoch": 0.4806716126513081,
+      "grad_norm": 3.5609123706817627,
+      "learning_rate": 9.513358687018636e-06,
+      "loss": 0.78,
+      "step": 3693
+    },
+    {
+      "epoch": 0.4810620851229988,
+      "grad_norm": 2.5246832370758057,
+      "learning_rate": 9.512458970711366e-06,
+      "loss": 0.6885,
+      "step": 3696
+    },
+    {
+      "epoch": 0.4814525575946896,
+      "grad_norm": 2.706555128097534,
+      "learning_rate": 9.51155846608701e-06,
+      "loss": 0.659,
+      "step": 3699
+    },
+    {
+      "epoch": 0.4818430300663803,
+      "grad_norm": 3.1037871837615967,
+      "learning_rate": 9.510657173302878e-06,
+      "loss": 0.6932,
+      "step": 3702
+    },
+    {
+      "epoch": 0.48223350253807107,
+      "grad_norm": 2.7866525650024414,
+      "learning_rate": 9.509755092516427e-06,
+      "loss": 0.6329,
+      "step": 3705
+    },
+    {
+      "epoch": 0.4826239750097618,
+      "grad_norm": 2.528336524963379,
+      "learning_rate": 9.508852223885248e-06,
+      "loss": 0.5991,
+      "step": 3708
+    },
+    {
+      "epoch": 0.48301444748145256,
+      "grad_norm": 2.683349609375,
+      "learning_rate": 9.507948567567073e-06,
+      "loss": 0.6348,
+      "step": 3711
+    },
+    {
+      "epoch": 0.48340491995314333,
+      "grad_norm": 3.169064998626709,
+      "learning_rate": 9.507044123719764e-06,
+      "loss": 0.758,
+      "step": 3714
+    },
+    {
+      "epoch": 0.48379539242483405,
+      "grad_norm": 2.669492721557617,
+      "learning_rate": 9.50613889250133e-06,
+      "loss": 0.6458,
+      "step": 3717
+    },
+    {
+      "epoch": 0.4841858648965248,
+      "grad_norm": 4.006582736968994,
+      "learning_rate": 9.50523287406991e-06,
+      "loss": 0.6128,
+      "step": 3720
+    },
+    {
+      "epoch": 0.48457633736821554,
+      "grad_norm": 2.645681858062744,
+      "learning_rate": 9.504326068583784e-06,
+      "loss": 0.651,
+      "step": 3723
+    },
+    {
+      "epoch": 0.4849668098399063,
+      "grad_norm": 2.4632298946380615,
+      "learning_rate": 9.503418476201371e-06,
+      "loss": 0.5938,
+      "step": 3726
+    },
+    {
+      "epoch": 0.485357282311597,
+      "grad_norm": 2.407679557800293,
+      "learning_rate": 9.502510097081223e-06,
+      "loss": 0.7057,
+      "step": 3729
+    },
+    {
+      "epoch": 0.4857477547832878,
+      "grad_norm": 2.6753084659576416,
+      "learning_rate": 9.501600931382034e-06,
+      "loss": 0.7167,
+      "step": 3732
+    },
+    {
+      "epoch": 0.4861382272549785,
+      "grad_norm": 2.707786798477173,
+      "learning_rate": 9.500690979262632e-06,
+      "loss": 0.7135,
+      "step": 3735
+    },
+    {
+      "epoch": 0.4865286997266693,
+      "grad_norm": 2.5371201038360596,
+      "learning_rate": 9.499780240881981e-06,
+      "loss": 0.7335,
+      "step": 3738
+    },
+    {
+      "epoch": 0.48691917219836,
+      "grad_norm": 2.661376953125,
+      "learning_rate": 9.498868716399192e-06,
+      "loss": 0.6258,
+      "step": 3741
+    },
+    {
+      "epoch": 0.4873096446700508,
+      "grad_norm": 2.647146463394165,
+      "learning_rate": 9.4979564059735e-06,
+      "loss": 0.7245,
+      "step": 3744
+    },
+    {
+      "epoch": 0.4877001171417415,
+      "grad_norm": 3.4181723594665527,
+      "learning_rate": 9.497043309764289e-06,
+      "loss": 0.7563,
+      "step": 3747
+    },
+    {
+      "epoch": 0.48809058961343227,
+      "grad_norm": 3.5437850952148438,
+      "learning_rate": 9.496129427931069e-06,
+      "loss": 0.7247,
+      "step": 3750
+    },
+    {
+      "epoch": 0.488481062085123,
+      "grad_norm": 3.7329277992248535,
+      "learning_rate": 9.495214760633498e-06,
+      "loss": 0.6169,
+      "step": 3753
+    },
+    {
+      "epoch": 0.48887153455681376,
+      "grad_norm": 2.8779706954956055,
+      "learning_rate": 9.494299308031365e-06,
+      "loss": 0.6749,
+      "step": 3756
+    },
+    {
+      "epoch": 0.4892620070285045,
+      "grad_norm": 2.7706687450408936,
+      "learning_rate": 9.493383070284597e-06,
+      "loss": 0.6352,
+      "step": 3759
+    },
+    {
+      "epoch": 0.48965247950019525,
+      "grad_norm": 2.6291885375976562,
+      "learning_rate": 9.49246604755326e-06,
+      "loss": 0.5958,
+      "step": 3762
+    },
+    {
+      "epoch": 0.49004295197188597,
+      "grad_norm": 2.48679256439209,
+      "learning_rate": 9.491548239997555e-06,
+      "loss": 0.6365,
+      "step": 3765
+    },
+    {
+      "epoch": 0.49043342444357674,
+      "grad_norm": 2.522580862045288,
+      "learning_rate": 9.490629647777821e-06,
+      "loss": 0.6723,
+      "step": 3768
+    },
+    {
+      "epoch": 0.49082389691526745,
+      "grad_norm": 2.4899022579193115,
+      "learning_rate": 9.489710271054533e-06,
+      "loss": 0.651,
+      "step": 3771
+    },
+    {
+      "epoch": 0.4912143693869582,
+      "grad_norm": 3.645275592803955,
+      "learning_rate": 9.488790109988308e-06,
+      "loss": 0.6087,
+      "step": 3774
+    },
+    {
+      "epoch": 0.49160484185864894,
+      "grad_norm": 2.5229358673095703,
+      "learning_rate": 9.487869164739893e-06,
+      "loss": 0.6372,
+      "step": 3777
+    },
+    {
+      "epoch": 0.4919953143303397,
+      "grad_norm": 2.5674540996551514,
+      "learning_rate": 9.486947435470176e-06,
+      "loss": 0.585,
+      "step": 3780
+    },
+    {
+      "epoch": 0.49238578680203043,
+      "grad_norm": 2.749277353286743,
+      "learning_rate": 9.486024922340181e-06,
+      "loss": 0.6834,
+      "step": 3783
+    },
+    {
+      "epoch": 0.4927762592737212,
+      "grad_norm": 2.4357573986053467,
+      "learning_rate": 9.485101625511071e-06,
+      "loss": 0.6562,
+      "step": 3786
+    },
+    {
+      "epoch": 0.4931667317454119,
+      "grad_norm": 2.85579776763916,
+      "learning_rate": 9.48417754514414e-06,
+      "loss": 0.6401,
+      "step": 3789
+    },
+    {
+      "epoch": 0.4935572042171027,
+      "grad_norm": 2.402381658554077,
+      "learning_rate": 9.483252681400825e-06,
+      "loss": 0.7112,
+      "step": 3792
+    },
+    {
+      "epoch": 0.4939476766887934,
+      "grad_norm": 3.6302082538604736,
+      "learning_rate": 9.4823270344427e-06,
+      "loss": 0.71,
+      "step": 3795
+    },
+    {
+      "epoch": 0.4943381491604842,
+      "grad_norm": 2.943791627883911,
+      "learning_rate": 9.48140060443147e-06,
+      "loss": 0.657,
+      "step": 3798
+    },
+    {
+      "epoch": 0.49472862163217496,
+      "grad_norm": 3.03796648979187,
+      "learning_rate": 9.480473391528982e-06,
+      "loss": 0.7657,
+      "step": 3801
+    },
+    {
+      "epoch": 0.4951190941038657,
+      "grad_norm": 3.2505815029144287,
+      "learning_rate": 9.479545395897219e-06,
+      "loss": 0.6439,
+      "step": 3804
+    },
+    {
+      "epoch": 0.49550956657555645,
+      "grad_norm": 2.497419834136963,
+      "learning_rate": 9.478616617698297e-06,
+      "loss": 0.7109,
+      "step": 3807
+    },
+    {
+      "epoch": 0.49590003904724717,
+      "grad_norm": 2.562572956085205,
+      "learning_rate": 9.477687057094476e-06,
+      "loss": 0.7191,
+      "step": 3810
+    },
+    {
+      "epoch": 0.49629051151893794,
+      "grad_norm": 2.6476757526397705,
+      "learning_rate": 9.476756714248142e-06,
+      "loss": 0.6237,
+      "step": 3813
+    },
+    {
+      "epoch": 0.49668098399062866,
+      "grad_norm": 2.27634859085083,
+      "learning_rate": 9.475825589321831e-06,
+      "loss": 0.635,
+      "step": 3816
+    },
+    {
+      "epoch": 0.4970714564623194,
+      "grad_norm": 2.4250454902648926,
+      "learning_rate": 9.474893682478205e-06,
+      "loss": 0.6073,
+      "step": 3819
+    },
+    {
+      "epoch": 0.49746192893401014,
+      "grad_norm": 2.7182376384735107,
+      "learning_rate": 9.473960993880068e-06,
+      "loss": 0.6665,
+      "step": 3822
+    },
+    {
+      "epoch": 0.4978524014057009,
+      "grad_norm": 2.8492777347564697,
+      "learning_rate": 9.473027523690355e-06,
+      "loss": 0.5861,
+      "step": 3825
+    },
+    {
+      "epoch": 0.49824287387739163,
+      "grad_norm": 2.5110251903533936,
+      "learning_rate": 9.472093272072142e-06,
+      "loss": 0.5882,
+      "step": 3828
+    },
+    {
+      "epoch": 0.4986333463490824,
+      "grad_norm": 2.5196032524108887,
+      "learning_rate": 9.471158239188645e-06,
+      "loss": 0.7131,
+      "step": 3831
+    },
+    {
+      "epoch": 0.4990238188207731,
+      "grad_norm": 2.7564871311187744,
+      "learning_rate": 9.470222425203209e-06,
+      "loss": 0.7012,
+      "step": 3834
+    },
+    {
+      "epoch": 0.4994142912924639,
+      "grad_norm": 2.5324935913085938,
+      "learning_rate": 9.469285830279318e-06,
+      "loss": 0.6357,
+      "step": 3837
+    },
+    {
+      "epoch": 0.4998047637641546,
+      "grad_norm": 2.3627541065216064,
+      "learning_rate": 9.468348454580596e-06,
+      "loss": 0.6458,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5001952362358454,
+      "grad_norm": 3.514322519302368,
+      "learning_rate": 9.467410298270798e-06,
+      "loss": 0.5593,
+      "step": 3843
+    },
+    {
+      "epoch": 0.5005857087075362,
+      "grad_norm": 2.6262550354003906,
+      "learning_rate": 9.46647136151382e-06,
+      "loss": 0.6549,
+      "step": 3846
+    },
+    {
+      "epoch": 0.5009761811792268,
+      "grad_norm": 3.208850860595703,
+      "learning_rate": 9.46553164447369e-06,
+      "loss": 0.6028,
+      "step": 3849
+    },
+    {
+      "epoch": 0.5013666536509176,
+      "grad_norm": 2.4970576763153076,
+      "learning_rate": 9.464591147314577e-06,
+      "loss": 0.712,
+      "step": 3852
+    },
+    {
+      "epoch": 0.5017571261226084,
+      "grad_norm": 3.3862645626068115,
+      "learning_rate": 9.463649870200782e-06,
+      "loss": 0.5893,
+      "step": 3855
+    },
+    {
+      "epoch": 0.5021475985942991,
+      "grad_norm": 5.486196041107178,
+      "learning_rate": 9.462707813296746e-06,
+      "loss": 0.7407,
+      "step": 3858
+    },
+    {
+      "epoch": 0.5025380710659898,
+      "grad_norm": 2.571362018585205,
+      "learning_rate": 9.46176497676704e-06,
+      "loss": 0.5659,
+      "step": 3861
+    },
+    {
+      "epoch": 0.5029285435376806,
+      "grad_norm": 2.665663957595825,
+      "learning_rate": 9.460821360776382e-06,
+      "loss": 0.7724,
+      "step": 3864
+    },
+    {
+      "epoch": 0.5033190160093713,
+      "grad_norm": 3.4792850017547607,
+      "learning_rate": 9.459876965489614e-06,
+      "loss": 0.6739,
+      "step": 3867
+    },
+    {
+      "epoch": 0.5037094884810621,
+      "grad_norm": 3.114149332046509,
+      "learning_rate": 9.458931791071723e-06,
+      "loss": 0.6073,
+      "step": 3870
+    },
+    {
+      "epoch": 0.5040999609527528,
+      "grad_norm": 4.276490211486816,
+      "learning_rate": 9.457985837687832e-06,
+      "loss": 0.7193,
+      "step": 3873
+    },
+    {
+      "epoch": 0.5044904334244436,
+      "grad_norm": 2.727604389190674,
+      "learning_rate": 9.457039105503188e-06,
+      "loss": 0.6752,
+      "step": 3876
+    },
+    {
+      "epoch": 0.5048809058961343,
+      "grad_norm": 2.9666244983673096,
+      "learning_rate": 9.456091594683192e-06,
+      "loss": 0.7623,
+      "step": 3879
+    },
+    {
+      "epoch": 0.5052713783678251,
+      "grad_norm": 2.6261138916015625,
+      "learning_rate": 9.455143305393367e-06,
+      "loss": 0.7193,
+      "step": 3882
+    },
+    {
+      "epoch": 0.5056618508395159,
+      "grad_norm": 3.4070770740509033,
+      "learning_rate": 9.454194237799379e-06,
+      "loss": 0.6537,
+      "step": 3885
+    },
+    {
+      "epoch": 0.5060523233112065,
+      "grad_norm": 2.4243500232696533,
+      "learning_rate": 9.453244392067028e-06,
+      "loss": 0.6781,
+      "step": 3888
+    },
+    {
+      "epoch": 0.5064427957828973,
+      "grad_norm": 3.1091771125793457,
+      "learning_rate": 9.45229376836225e-06,
+      "loss": 0.6486,
+      "step": 3891
+    },
+    {
+      "epoch": 0.5068332682545881,
+      "grad_norm": 2.519580841064453,
+      "learning_rate": 9.451342366851117e-06,
+      "loss": 0.6303,
+      "step": 3894
+    },
+    {
+      "epoch": 0.5072237407262788,
+      "grad_norm": 2.6116557121276855,
+      "learning_rate": 9.450390187699838e-06,
+      "loss": 0.6344,
+      "step": 3897
+    },
+    {
+      "epoch": 0.5076142131979695,
+      "grad_norm": 2.69616961479187,
+      "learning_rate": 9.449437231074755e-06,
+      "loss": 0.6245,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5080046856696603,
+      "grad_norm": 2.675161600112915,
+      "learning_rate": 9.448483497142347e-06,
+      "loss": 0.6087,
+      "step": 3903
+    },
+    {
+      "epoch": 0.5083951581413511,
+      "grad_norm": 2.4945101737976074,
+      "learning_rate": 9.44752898606923e-06,
+      "loss": 0.6146,
+      "step": 3906
+    },
+    {
+      "epoch": 0.5087856306130418,
+      "grad_norm": 2.615118980407715,
+      "learning_rate": 9.446573698022155e-06,
+      "loss": 0.8205,
+      "step": 3909
+    },
+    {
+      "epoch": 0.5091761030847325,
+      "grad_norm": 4.189955234527588,
+      "learning_rate": 9.445617633168012e-06,
+      "loss": 0.6343,
+      "step": 3912
+    },
+    {
+      "epoch": 0.5095665755564233,
+      "grad_norm": 2.6967740058898926,
+      "learning_rate": 9.444660791673818e-06,
+      "loss": 0.58,
+      "step": 3915
+    },
+    {
+      "epoch": 0.509957048028114,
+      "grad_norm": 2.989888906478882,
+      "learning_rate": 9.443703173706734e-06,
+      "loss": 0.6818,
+      "step": 3918
+    },
+    {
+      "epoch": 0.5103475204998048,
+      "grad_norm": 3.485555410385132,
+      "learning_rate": 9.442744779434054e-06,
+      "loss": 0.612,
+      "step": 3921
+    },
+    {
+      "epoch": 0.5107379929714955,
+      "grad_norm": 2.7845234870910645,
+      "learning_rate": 9.441785609023208e-06,
+      "loss": 0.6826,
+      "step": 3924
+    },
+    {
+      "epoch": 0.5111284654431862,
+      "grad_norm": 2.6713454723358154,
+      "learning_rate": 9.44082566264176e-06,
+      "loss": 0.6219,
+      "step": 3927
+    },
+    {
+      "epoch": 0.511518937914877,
+      "grad_norm": 3.361051321029663,
+      "learning_rate": 9.43986494045741e-06,
+      "loss": 0.669,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5119094103865678,
+      "grad_norm": 3.308380365371704,
+      "learning_rate": 9.438903442637997e-06,
+      "loss": 0.6721,
+      "step": 3933
+    },
+    {
+      "epoch": 0.5122998828582584,
+      "grad_norm": 2.7254140377044678,
+      "learning_rate": 9.43794116935149e-06,
+      "loss": 0.6372,
+      "step": 3936
+    },
+    {
+      "epoch": 0.5126903553299492,
+      "grad_norm": 2.9412028789520264,
+      "learning_rate": 9.436978120765996e-06,
+      "loss": 0.7481,
+      "step": 3939
+    },
+    {
+      "epoch": 0.51308082780164,
+      "grad_norm": 2.711512804031372,
+      "learning_rate": 9.436014297049758e-06,
+      "loss": 0.7198,
+      "step": 3942
+    },
+    {
+      "epoch": 0.5134713002733308,
+      "grad_norm": 2.667794942855835,
+      "learning_rate": 9.435049698371155e-06,
+      "loss": 0.681,
+      "step": 3945
+    },
+    {
+      "epoch": 0.5138617727450214,
+      "grad_norm": 3.5074334144592285,
+      "learning_rate": 9.4340843248987e-06,
+      "loss": 0.6905,
+      "step": 3948
+    },
+    {
+      "epoch": 0.5142522452167122,
+      "grad_norm": 3.1205289363861084,
+      "learning_rate": 9.43311817680104e-06,
+      "loss": 0.7021,
+      "step": 3951
+    },
+    {
+      "epoch": 0.514642717688403,
+      "grad_norm": 2.667708396911621,
+      "learning_rate": 9.432151254246961e-06,
+      "loss": 0.7106,
+      "step": 3954
+    },
+    {
+      "epoch": 0.5150331901600937,
+      "grad_norm": 2.348116874694824,
+      "learning_rate": 9.431183557405383e-06,
+      "loss": 0.5503,
+      "step": 3957
+    },
+    {
+      "epoch": 0.5154236626317844,
+      "grad_norm": 3.321404457092285,
+      "learning_rate": 9.430215086445358e-06,
+      "loss": 0.6023,
+      "step": 3960
+    },
+    {
+      "epoch": 0.5158141351034752,
+      "grad_norm": 2.7746543884277344,
+      "learning_rate": 9.429245841536079e-06,
+      "loss": 0.678,
+      "step": 3963
+    },
+    {
+      "epoch": 0.516204607575166,
+      "grad_norm": 2.634472131729126,
+      "learning_rate": 9.428275822846868e-06,
+      "loss": 0.6451,
+      "step": 3966
+    },
+    {
+      "epoch": 0.5165950800468567,
+      "grad_norm": 2.5320868492126465,
+      "learning_rate": 9.427305030547185e-06,
+      "loss": 0.6206,
+      "step": 3969
+    },
+    {
+      "epoch": 0.5169855525185475,
+      "grad_norm": 2.5869338512420654,
+      "learning_rate": 9.42633346480663e-06,
+      "loss": 0.6766,
+      "step": 3972
+    },
+    {
+      "epoch": 0.5173760249902382,
+      "grad_norm": 2.589397430419922,
+      "learning_rate": 9.425361125794928e-06,
+      "loss": 0.7054,
+      "step": 3975
+    },
+    {
+      "epoch": 0.5177664974619289,
+      "grad_norm": 2.9746224880218506,
+      "learning_rate": 9.424388013681947e-06,
+      "loss": 0.7086,
+      "step": 3978
+    },
+    {
+      "epoch": 0.5181569699336197,
+      "grad_norm": 2.565460205078125,
+      "learning_rate": 9.42341412863769e-06,
+      "loss": 0.6858,
+      "step": 3981
+    },
+    {
+      "epoch": 0.5185474424053105,
+      "grad_norm": 2.5512335300445557,
+      "learning_rate": 9.422439470832288e-06,
+      "loss": 0.6332,
+      "step": 3984
+    },
+    {
+      "epoch": 0.5189379148770011,
+      "grad_norm": 3.2780306339263916,
+      "learning_rate": 9.421464040436012e-06,
+      "loss": 0.664,
+      "step": 3987
+    },
+    {
+      "epoch": 0.5193283873486919,
+      "grad_norm": 2.374025821685791,
+      "learning_rate": 9.420487837619272e-06,
+      "loss": 0.5543,
+      "step": 3990
+    },
+    {
+      "epoch": 0.5197188598203827,
+      "grad_norm": 2.818821907043457,
+      "learning_rate": 9.419510862552608e-06,
+      "loss": 0.6181,
+      "step": 3993
+    },
+    {
+      "epoch": 0.5201093322920735,
+      "grad_norm": 2.3839192390441895,
+      "learning_rate": 9.41853311540669e-06,
+      "loss": 0.6783,
+      "step": 3996
+    },
+    {
+      "epoch": 0.5204998047637641,
+      "grad_norm": 2.5750620365142822,
+      "learning_rate": 9.417554596352334e-06,
+      "loss": 0.6101,
+      "step": 3999
+    },
+    {
+      "epoch": 0.5208902772354549,
+      "grad_norm": 2.7210500240325928,
+      "learning_rate": 9.416575305560482e-06,
+      "loss": 0.749,
+      "step": 4002
+    },
+    {
+      "epoch": 0.5212807497071457,
+      "grad_norm": 2.856265068054199,
+      "learning_rate": 9.415595243202217e-06,
+      "loss": 0.6108,
+      "step": 4005
+    },
+    {
+      "epoch": 0.5216712221788364,
+      "grad_norm": 2.2772696018218994,
+      "learning_rate": 9.41461440944875e-06,
+      "loss": 0.7071,
+      "step": 4008
+    },
+    {
+      "epoch": 0.5220616946505271,
+      "grad_norm": 2.629115104675293,
+      "learning_rate": 9.413632804471434e-06,
+      "loss": 0.6378,
+      "step": 4011
+    },
+    {
+      "epoch": 0.5224521671222179,
+      "grad_norm": 2.853085517883301,
+      "learning_rate": 9.412650428441752e-06,
+      "loss": 0.6871,
+      "step": 4014
+    },
+    {
+      "epoch": 0.5228426395939086,
+      "grad_norm": 2.5856900215148926,
+      "learning_rate": 9.411667281531322e-06,
+      "loss": 0.6206,
+      "step": 4017
+    },
+    {
+      "epoch": 0.5232331120655994,
+      "grad_norm": 2.705491065979004,
+      "learning_rate": 9.4106833639119e-06,
+      "loss": 0.5966,
+      "step": 4020
+    },
+    {
+      "epoch": 0.5236235845372901,
+      "grad_norm": 2.5959479808807373,
+      "learning_rate": 9.409698675755372e-06,
+      "loss": 0.6265,
+      "step": 4023
+    },
+    {
+      "epoch": 0.5240140570089808,
+      "grad_norm": 2.661630630493164,
+      "learning_rate": 9.408713217233762e-06,
+      "loss": 0.6688,
+      "step": 4026
+    },
+    {
+      "epoch": 0.5244045294806716,
+      "grad_norm": 2.437685966491699,
+      "learning_rate": 9.40772698851923e-06,
+      "loss": 0.6566,
+      "step": 4029
+    },
+    {
+      "epoch": 0.5247950019523624,
+      "grad_norm": 2.658416986465454,
+      "learning_rate": 9.406739989784061e-06,
+      "loss": 0.679,
+      "step": 4032
+    },
+    {
+      "epoch": 0.5251854744240531,
+      "grad_norm": 3.007324695587158,
+      "learning_rate": 9.40575222120069e-06,
+      "loss": 0.6227,
+      "step": 4035
+    },
+    {
+      "epoch": 0.5255759468957438,
+      "grad_norm": 2.6084835529327393,
+      "learning_rate": 9.404763682941673e-06,
+      "loss": 0.6149,
+      "step": 4038
+    },
+    {
+      "epoch": 0.5259664193674346,
+      "grad_norm": 3.3151960372924805,
+      "learning_rate": 9.403774375179707e-06,
+      "loss": 0.6682,
+      "step": 4041
+    },
+    {
+      "epoch": 0.5263568918391254,
+      "grad_norm": 2.847337007522583,
+      "learning_rate": 9.402784298087622e-06,
+      "loss": 0.596,
+      "step": 4044
+    },
+    {
+      "epoch": 0.526747364310816,
+      "grad_norm": 3.080592393875122,
+      "learning_rate": 9.401793451838382e-06,
+      "loss": 0.7028,
+      "step": 4047
+    },
+    {
+      "epoch": 0.5271378367825068,
+      "grad_norm": 2.5581657886505127,
+      "learning_rate": 9.400801836605085e-06,
+      "loss": 0.6566,
+      "step": 4050
+    },
+    {
+      "epoch": 0.5275283092541976,
+      "grad_norm": 2.55700421333313,
+      "learning_rate": 9.399809452560966e-06,
+      "loss": 0.6734,
+      "step": 4053
+    },
+    {
+      "epoch": 0.5279187817258884,
+      "grad_norm": 2.44401478767395,
+      "learning_rate": 9.398816299879392e-06,
+      "loss": 0.5712,
+      "step": 4056
+    },
+    {
+      "epoch": 0.5283092541975791,
+      "grad_norm": 2.4714529514312744,
+      "learning_rate": 9.397822378733864e-06,
+      "loss": 0.6925,
+      "step": 4059
+    },
+    {
+      "epoch": 0.5286997266692698,
+      "grad_norm": 3.4595723152160645,
+      "learning_rate": 9.396827689298018e-06,
+      "loss": 0.6653,
+      "step": 4062
+    },
+    {
+      "epoch": 0.5290901991409606,
+      "grad_norm": 2.528628349304199,
+      "learning_rate": 9.395832231745624e-06,
+      "loss": 0.6362,
+      "step": 4065
+    },
+    {
+      "epoch": 0.5294806716126513,
+      "grad_norm": 2.4336354732513428,
+      "learning_rate": 9.394836006250587e-06,
+      "loss": 0.6348,
+      "step": 4068
+    },
+    {
+      "epoch": 0.5298711440843421,
+      "grad_norm": 2.7131993770599365,
+      "learning_rate": 9.393839012986944e-06,
+      "loss": 0.675,
+      "step": 4071
+    },
+    {
+      "epoch": 0.5302616165560328,
+      "grad_norm": 2.5495808124542236,
+      "learning_rate": 9.39284125212887e-06,
+      "loss": 0.6013,
+      "step": 4074
+    },
+    {
+      "epoch": 0.5306520890277235,
+      "grad_norm": 2.3943541049957275,
+      "learning_rate": 9.39184272385067e-06,
+      "loss": 0.653,
+      "step": 4077
+    },
+    {
+      "epoch": 0.5310425614994143,
+      "grad_norm": 2.781402111053467,
+      "learning_rate": 9.390843428326785e-06,
+      "loss": 0.7808,
+      "step": 4080
+    },
+    {
+      "epoch": 0.5314330339711051,
+      "grad_norm": 3.896851062774658,
+      "learning_rate": 9.38984336573179e-06,
+      "loss": 0.6671,
+      "step": 4083
+    },
+    {
+      "epoch": 0.5318235064427957,
+      "grad_norm": 2.4791414737701416,
+      "learning_rate": 9.388842536240395e-06,
+      "loss": 0.7199,
+      "step": 4086
+    },
+    {
+      "epoch": 0.5322139789144865,
+      "grad_norm": 2.551300048828125,
+      "learning_rate": 9.387840940027439e-06,
+      "loss": 0.5786,
+      "step": 4089
+    },
+    {
+      "epoch": 0.5326044513861773,
+      "grad_norm": 2.9271490573883057,
+      "learning_rate": 9.386838577267906e-06,
+      "loss": 0.755,
+      "step": 4092
+    },
+    {
+      "epoch": 0.5329949238578681,
+      "grad_norm": 2.4005637168884277,
+      "learning_rate": 9.385835448136897e-06,
+      "loss": 0.662,
+      "step": 4095
+    },
+    {
+      "epoch": 0.5333853963295587,
+      "grad_norm": 2.360596179962158,
+      "learning_rate": 9.384831552809665e-06,
+      "loss": 0.6733,
+      "step": 4098
+    },
+    {
+      "epoch": 0.5337758688012495,
+      "grad_norm": 2.809666395187378,
+      "learning_rate": 9.383826891461583e-06,
+      "loss": 0.5828,
+      "step": 4101
+    },
+    {
+      "epoch": 0.5341663412729403,
+      "grad_norm": 2.80281138420105,
+      "learning_rate": 9.382821464268166e-06,
+      "loss": 0.6344,
+      "step": 4104
+    },
+    {
+      "epoch": 0.534556813744631,
+      "grad_norm": 2.5210893154144287,
+      "learning_rate": 9.38181527140506e-06,
+      "loss": 0.6828,
+      "step": 4107
+    },
+    {
+      "epoch": 0.5349472862163217,
+      "grad_norm": 3.6144542694091797,
+      "learning_rate": 9.380808313048045e-06,
+      "loss": 0.7309,
+      "step": 4110
+    },
+    {
+      "epoch": 0.5353377586880125,
+      "grad_norm": 2.6584270000457764,
+      "learning_rate": 9.379800589373032e-06,
+      "loss": 0.6842,
+      "step": 4113
+    },
+    {
+      "epoch": 0.5357282311597032,
+      "grad_norm": 2.685014247894287,
+      "learning_rate": 9.378792100556069e-06,
+      "loss": 0.64,
+      "step": 4116
+    },
+    {
+      "epoch": 0.536118703631394,
+      "grad_norm": 2.667910099029541,
+      "learning_rate": 9.37778284677334e-06,
+      "loss": 0.7603,
+      "step": 4119
+    },
+    {
+      "epoch": 0.5365091761030847,
+      "grad_norm": 2.820080041885376,
+      "learning_rate": 9.376772828201155e-06,
+      "loss": 0.6083,
+      "step": 4122
+    },
+    {
+      "epoch": 0.5368996485747755,
+      "grad_norm": 2.422607898712158,
+      "learning_rate": 9.375762045015966e-06,
+      "loss": 0.6028,
+      "step": 4125
+    },
+    {
+      "epoch": 0.5372901210464662,
+      "grad_norm": 3.0713932514190674,
+      "learning_rate": 9.374750497394352e-06,
+      "loss": 0.7543,
+      "step": 4128
+    },
+    {
+      "epoch": 0.537680593518157,
+      "grad_norm": 2.311880111694336,
+      "learning_rate": 9.373738185513028e-06,
+      "loss": 0.6959,
+      "step": 4131
+    },
+    {
+      "epoch": 0.5380710659898477,
+      "grad_norm": 2.619349241256714,
+      "learning_rate": 9.372725109548846e-06,
+      "loss": 0.7416,
+      "step": 4134
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 2.9388935565948486,
+      "learning_rate": 9.371711269678786e-06,
+      "loss": 0.737,
+      "step": 4137
+    },
+    {
+      "epoch": 0.5388520109332292,
+      "grad_norm": 2.5478711128234863,
+      "learning_rate": 9.370696666079964e-06,
+      "loss": 0.7358,
+      "step": 4140
+    },
+    {
+      "epoch": 0.53924248340492,
+      "grad_norm": 4.10902214050293,
+      "learning_rate": 9.369681298929629e-06,
+      "loss": 0.6521,
+      "step": 4143
+    },
+    {
+      "epoch": 0.5396329558766108,
+      "grad_norm": 2.6128830909729004,
+      "learning_rate": 9.368665168405163e-06,
+      "loss": 0.663,
+      "step": 4146
+    },
+    {
+      "epoch": 0.5400234283483014,
+      "grad_norm": 3.5728492736816406,
+      "learning_rate": 9.367648274684082e-06,
+      "loss": 0.6235,
+      "step": 4149
+    },
+    {
+      "epoch": 0.5404139008199922,
+      "grad_norm": 3.0014986991882324,
+      "learning_rate": 9.366630617944037e-06,
+      "loss": 0.607,
+      "step": 4152
+    },
+    {
+      "epoch": 0.540804373291683,
+      "grad_norm": 2.6722731590270996,
+      "learning_rate": 9.365612198362807e-06,
+      "loss": 0.6592,
+      "step": 4155
+    },
+    {
+      "epoch": 0.5411948457633737,
+      "grad_norm": 3.075171947479248,
+      "learning_rate": 9.364593016118311e-06,
+      "loss": 0.7503,
+      "step": 4158
+    },
+    {
+      "epoch": 0.5415853182350644,
+      "grad_norm": 2.6275832653045654,
+      "learning_rate": 9.363573071388598e-06,
+      "loss": 0.6354,
+      "step": 4161
+    },
+    {
+      "epoch": 0.5419757907067552,
+      "grad_norm": 2.640997886657715,
+      "learning_rate": 9.362552364351849e-06,
+      "loss": 0.6437,
+      "step": 4164
+    },
+    {
+      "epoch": 0.5423662631784459,
+      "grad_norm": 4.610156059265137,
+      "learning_rate": 9.361530895186378e-06,
+      "loss": 0.7229,
+      "step": 4167
+    },
+    {
+      "epoch": 0.5427567356501367,
+      "grad_norm": 2.4628376960754395,
+      "learning_rate": 9.360508664070634e-06,
+      "loss": 0.7439,
+      "step": 4170
+    },
+    {
+      "epoch": 0.5431472081218274,
+      "grad_norm": 3.297962188720703,
+      "learning_rate": 9.359485671183202e-06,
+      "loss": 0.7502,
+      "step": 4173
+    },
+    {
+      "epoch": 0.5435376805935181,
+      "grad_norm": 2.7493808269500732,
+      "learning_rate": 9.358461916702793e-06,
+      "loss": 0.5738,
+      "step": 4176
+    },
+    {
+      "epoch": 0.5439281530652089,
+      "grad_norm": 2.37622332572937,
+      "learning_rate": 9.357437400808256e-06,
+      "loss": 0.6252,
+      "step": 4179
+    },
+    {
+      "epoch": 0.5443186255368997,
+      "grad_norm": 2.4607114791870117,
+      "learning_rate": 9.356412123678572e-06,
+      "loss": 0.6813,
+      "step": 4182
+    },
+    {
+      "epoch": 0.5447090980085904,
+      "grad_norm": 3.2492547035217285,
+      "learning_rate": 9.355386085492855e-06,
+      "loss": 0.6283,
+      "step": 4185
+    },
+    {
+      "epoch": 0.5450995704802811,
+      "grad_norm": 2.3244526386260986,
+      "learning_rate": 9.35435928643035e-06,
+      "loss": 0.5893,
+      "step": 4188
+    },
+    {
+      "epoch": 0.5454900429519719,
+      "grad_norm": 3.75888729095459,
+      "learning_rate": 9.353331726670438e-06,
+      "loss": 0.7102,
+      "step": 4191
+    },
+    {
+      "epoch": 0.5458805154236627,
+      "grad_norm": 2.8710289001464844,
+      "learning_rate": 9.352303406392634e-06,
+      "loss": 0.6786,
+      "step": 4194
+    },
+    {
+      "epoch": 0.5462709878953533,
+      "grad_norm": 2.649484634399414,
+      "learning_rate": 9.351274325776578e-06,
+      "loss": 0.6359,
+      "step": 4197
+    },
+    {
+      "epoch": 0.5466614603670441,
+      "grad_norm": 2.539466381072998,
+      "learning_rate": 9.350244485002051e-06,
+      "loss": 0.7931,
+      "step": 4200
+    },
+    {
+      "epoch": 0.5470519328387349,
+      "grad_norm": 2.964155912399292,
+      "learning_rate": 9.349213884248967e-06,
+      "loss": 0.6971,
+      "step": 4203
+    },
+    {
+      "epoch": 0.5474424053104257,
+      "grad_norm": 2.4749128818511963,
+      "learning_rate": 9.348182523697365e-06,
+      "loss": 0.6917,
+      "step": 4206
+    },
+    {
+      "epoch": 0.5478328777821163,
+      "grad_norm": 2.6950347423553467,
+      "learning_rate": 9.347150403527422e-06,
+      "loss": 0.6089,
+      "step": 4209
+    },
+    {
+      "epoch": 0.5482233502538071,
+      "grad_norm": 3.4852073192596436,
+      "learning_rate": 9.34611752391945e-06,
+      "loss": 0.6896,
+      "step": 4212
+    },
+    {
+      "epoch": 0.5486138227254979,
+      "grad_norm": 2.6837730407714844,
+      "learning_rate": 9.34508388505389e-06,
+      "loss": 0.6311,
+      "step": 4215
+    },
+    {
+      "epoch": 0.5490042951971886,
+      "grad_norm": 3.1561684608459473,
+      "learning_rate": 9.344049487111316e-06,
+      "loss": 0.6248,
+      "step": 4218
+    },
+    {
+      "epoch": 0.5493947676688793,
+      "grad_norm": 3.7130353450775146,
+      "learning_rate": 9.343014330272432e-06,
+      "loss": 0.6496,
+      "step": 4221
+    },
+    {
+      "epoch": 0.5497852401405701,
+      "grad_norm": 3.125425338745117,
+      "learning_rate": 9.341978414718084e-06,
+      "loss": 0.7285,
+      "step": 4224
+    },
+    {
+      "epoch": 0.5501757126122608,
+      "grad_norm": 2.448007583618164,
+      "learning_rate": 9.34094174062924e-06,
+      "loss": 0.7151,
+      "step": 4227
+    },
+    {
+      "epoch": 0.5505661850839516,
+      "grad_norm": 2.704099655151367,
+      "learning_rate": 9.339904308187006e-06,
+      "loss": 0.7632,
+      "step": 4230
+    },
+    {
+      "epoch": 0.5509566575556424,
+      "grad_norm": 2.6874077320098877,
+      "learning_rate": 9.33886611757262e-06,
+      "loss": 0.6707,
+      "step": 4233
+    },
+    {
+      "epoch": 0.551347130027333,
+      "grad_norm": 2.497955799102783,
+      "learning_rate": 9.33782716896745e-06,
+      "loss": 0.6996,
+      "step": 4236
+    },
+    {
+      "epoch": 0.5517376024990238,
+      "grad_norm": 2.5674867630004883,
+      "learning_rate": 9.336787462553001e-06,
+      "loss": 0.6761,
+      "step": 4239
+    },
+    {
+      "epoch": 0.5521280749707146,
+      "grad_norm": 2.8246846199035645,
+      "learning_rate": 9.335746998510902e-06,
+      "loss": 0.7313,
+      "step": 4242
+    },
+    {
+      "epoch": 0.5525185474424054,
+      "grad_norm": 3.6085305213928223,
+      "learning_rate": 9.334705777022926e-06,
+      "loss": 0.7069,
+      "step": 4245
+    },
+    {
+      "epoch": 0.552909019914096,
+      "grad_norm": 2.782156229019165,
+      "learning_rate": 9.333663798270969e-06,
+      "loss": 0.6735,
+      "step": 4248
+    },
+    {
+      "epoch": 0.5532994923857868,
+      "grad_norm": 2.7994630336761475,
+      "learning_rate": 9.332621062437064e-06,
+      "loss": 0.6201,
+      "step": 4251
+    },
+    {
+      "epoch": 0.5536899648574776,
+      "grad_norm": 2.498595952987671,
+      "learning_rate": 9.331577569703374e-06,
+      "loss": 0.7614,
+      "step": 4254
+    },
+    {
+      "epoch": 0.5540804373291683,
+      "grad_norm": 2.729330062866211,
+      "learning_rate": 9.330533320252193e-06,
+      "loss": 0.6992,
+      "step": 4257
+    },
+    {
+      "epoch": 0.554470909800859,
+      "grad_norm": 2.681333303451538,
+      "learning_rate": 9.32948831426595e-06,
+      "loss": 0.7586,
+      "step": 4260
+    },
+    {
+      "epoch": 0.5548613822725498,
+      "grad_norm": 2.6096370220184326,
+      "learning_rate": 9.32844255192721e-06,
+      "loss": 0.6834,
+      "step": 4263
+    },
+    {
+      "epoch": 0.5552518547442405,
+      "grad_norm": 2.6331887245178223,
+      "learning_rate": 9.32739603341866e-06,
+      "loss": 0.6116,
+      "step": 4266
+    },
+    {
+      "epoch": 0.5556423272159313,
+      "grad_norm": 2.9159796237945557,
+      "learning_rate": 9.326348758923127e-06,
+      "loss": 0.6944,
+      "step": 4269
+    },
+    {
+      "epoch": 0.556032799687622,
+      "grad_norm": 2.7244224548339844,
+      "learning_rate": 9.325300728623567e-06,
+      "loss": 0.7559,
+      "step": 4272
+    },
+    {
+      "epoch": 0.5564232721593128,
+      "grad_norm": 2.5639004707336426,
+      "learning_rate": 9.324251942703068e-06,
+      "loss": 0.7335,
+      "step": 4275
+    },
+    {
+      "epoch": 0.5568137446310035,
+      "grad_norm": 2.4377241134643555,
+      "learning_rate": 9.323202401344852e-06,
+      "loss": 0.667,
+      "step": 4278
+    },
+    {
+      "epoch": 0.5572042171026943,
+      "grad_norm": 2.6609203815460205,
+      "learning_rate": 9.322152104732272e-06,
+      "loss": 0.6463,
+      "step": 4281
+    },
+    {
+      "epoch": 0.557594689574385,
+      "grad_norm": 3.7862682342529297,
+      "learning_rate": 9.321101053048812e-06,
+      "loss": 0.6735,
+      "step": 4284
+    },
+    {
+      "epoch": 0.5579851620460757,
+      "grad_norm": 3.0916101932525635,
+      "learning_rate": 9.320049246478086e-06,
+      "loss": 0.6867,
+      "step": 4287
+    },
+    {
+      "epoch": 0.5583756345177665,
+      "grad_norm": 2.4427428245544434,
+      "learning_rate": 9.318996685203848e-06,
+      "loss": 0.6473,
+      "step": 4290
+    },
+    {
+      "epoch": 0.5587661069894573,
+      "grad_norm": 2.9090192317962646,
+      "learning_rate": 9.317943369409973e-06,
+      "loss": 0.6461,
+      "step": 4293
+    },
+    {
+      "epoch": 0.5591565794611479,
+      "grad_norm": 5.3703742027282715,
+      "learning_rate": 9.316889299280475e-06,
+      "loss": 0.716,
+      "step": 4296
+    },
+    {
+      "epoch": 0.5595470519328387,
+      "grad_norm": 4.018325328826904,
+      "learning_rate": 9.315834474999498e-06,
+      "loss": 0.716,
+      "step": 4299
+    },
+    {
+      "epoch": 0.5599375244045295,
+      "grad_norm": 2.85209059715271,
+      "learning_rate": 9.314778896751317e-06,
+      "loss": 0.6133,
+      "step": 4302
+    },
+    {
+      "epoch": 0.5603279968762203,
+      "grad_norm": 2.487459182739258,
+      "learning_rate": 9.31372256472034e-06,
+      "loss": 0.6144,
+      "step": 4305
+    },
+    {
+      "epoch": 0.5607184693479109,
+      "grad_norm": 2.43287992477417,
+      "learning_rate": 9.312665479091106e-06,
+      "loss": 0.5989,
+      "step": 4308
+    },
+    {
+      "epoch": 0.5611089418196017,
+      "grad_norm": 3.1222102642059326,
+      "learning_rate": 9.311607640048286e-06,
+      "loss": 0.5524,
+      "step": 4311
+    },
+    {
+      "epoch": 0.5614994142912925,
+      "grad_norm": 2.515286445617676,
+      "learning_rate": 9.31054904777668e-06,
+      "loss": 0.6501,
+      "step": 4314
+    },
+    {
+      "epoch": 0.5618898867629832,
+      "grad_norm": 2.938828706741333,
+      "learning_rate": 9.309489702461223e-06,
+      "loss": 0.6829,
+      "step": 4317
+    },
+    {
+      "epoch": 0.562280359234674,
+      "grad_norm": 2.4656574726104736,
+      "learning_rate": 9.308429604286982e-06,
+      "loss": 0.6517,
+      "step": 4320
+    },
+    {
+      "epoch": 0.5626708317063647,
+      "grad_norm": 3.7046689987182617,
+      "learning_rate": 9.307368753439153e-06,
+      "loss": 0.6415,
+      "step": 4323
+    },
+    {
+      "epoch": 0.5630613041780554,
+      "grad_norm": 3.0883255004882812,
+      "learning_rate": 9.306307150103064e-06,
+      "loss": 0.7135,
+      "step": 4326
+    },
+    {
+      "epoch": 0.5634517766497462,
+      "grad_norm": 2.469595432281494,
+      "learning_rate": 9.305244794464174e-06,
+      "loss": 0.5753,
+      "step": 4329
+    },
+    {
+      "epoch": 0.563842249121437,
+      "grad_norm": 2.6837077140808105,
+      "learning_rate": 9.304181686708077e-06,
+      "loss": 0.6382,
+      "step": 4332
+    },
+    {
+      "epoch": 0.5642327215931277,
+      "grad_norm": 2.8714263439178467,
+      "learning_rate": 9.303117827020493e-06,
+      "loss": 0.7308,
+      "step": 4335
+    },
+    {
+      "epoch": 0.5646231940648184,
+      "grad_norm": 2.2070488929748535,
+      "learning_rate": 9.302053215587276e-06,
+      "loss": 0.7811,
+      "step": 4338
+    },
+    {
+      "epoch": 0.5650136665365092,
+      "grad_norm": 2.573568105697632,
+      "learning_rate": 9.300987852594414e-06,
+      "loss": 0.7276,
+      "step": 4341
+    },
+    {
+      "epoch": 0.5654041390082,
+      "grad_norm": 2.5360209941864014,
+      "learning_rate": 9.299921738228023e-06,
+      "loss": 0.6433,
+      "step": 4344
+    },
+    {
+      "epoch": 0.5657946114798906,
+      "grad_norm": 3.1989572048187256,
+      "learning_rate": 9.298854872674348e-06,
+      "loss": 0.6491,
+      "step": 4347
+    },
+    {
+      "epoch": 0.5661850839515814,
+      "grad_norm": 2.791049003601074,
+      "learning_rate": 9.297787256119772e-06,
+      "loss": 0.6953,
+      "step": 4350
+    },
+    {
+      "epoch": 0.5665755564232722,
+      "grad_norm": 3.882077693939209,
+      "learning_rate": 9.296718888750802e-06,
+      "loss": 0.697,
+      "step": 4353
+    },
+    {
+      "epoch": 0.566966028894963,
+      "grad_norm": 2.5977842807769775,
+      "learning_rate": 9.295649770754082e-06,
+      "loss": 0.6448,
+      "step": 4356
+    },
+    {
+      "epoch": 0.5673565013666536,
+      "grad_norm": 2.584838628768921,
+      "learning_rate": 9.294579902316382e-06,
+      "loss": 0.6445,
+      "step": 4359
+    },
+    {
+      "epoch": 0.5677469738383444,
+      "grad_norm": 3.0195114612579346,
+      "learning_rate": 9.293509283624611e-06,
+      "loss": 0.7209,
+      "step": 4362
+    },
+    {
+      "epoch": 0.5681374463100352,
+      "grad_norm": 3.4873313903808594,
+      "learning_rate": 9.292437914865798e-06,
+      "loss": 0.7565,
+      "step": 4365
+    },
+    {
+      "epoch": 0.5685279187817259,
+      "grad_norm": 2.704925298690796,
+      "learning_rate": 9.291365796227111e-06,
+      "loss": 0.8317,
+      "step": 4368
+    },
+    {
+      "epoch": 0.5689183912534166,
+      "grad_norm": 2.3734493255615234,
+      "learning_rate": 9.290292927895848e-06,
+      "loss": 0.6708,
+      "step": 4371
+    },
+    {
+      "epoch": 0.5693088637251074,
+      "grad_norm": 2.2318577766418457,
+      "learning_rate": 9.289219310059437e-06,
+      "loss": 0.6154,
+      "step": 4374
+    },
+    {
+      "epoch": 0.5696993361967981,
+      "grad_norm": 3.829906702041626,
+      "learning_rate": 9.288144942905432e-06,
+      "loss": 0.7834,
+      "step": 4377
+    },
+    {
+      "epoch": 0.5700898086684889,
+      "grad_norm": 2.0775136947631836,
+      "learning_rate": 9.28706982662153e-06,
+      "loss": 0.5858,
+      "step": 4380
+    },
+    {
+      "epoch": 0.5704802811401796,
+      "grad_norm": 2.4510486125946045,
+      "learning_rate": 9.285993961395548e-06,
+      "loss": 0.5897,
+      "step": 4383
+    },
+    {
+      "epoch": 0.5708707536118703,
+      "grad_norm": 2.438112735748291,
+      "learning_rate": 9.284917347415435e-06,
+      "loss": 0.6553,
+      "step": 4386
+    },
+    {
+      "epoch": 0.5712612260835611,
+      "grad_norm": 3.0546696186065674,
+      "learning_rate": 9.283839984869276e-06,
+      "loss": 0.6921,
+      "step": 4389
+    },
+    {
+      "epoch": 0.5716516985552519,
+      "grad_norm": 2.234543561935425,
+      "learning_rate": 9.282761873945285e-06,
+      "loss": 0.5942,
+      "step": 4392
+    },
+    {
+      "epoch": 0.5720421710269425,
+      "grad_norm": 2.697658061981201,
+      "learning_rate": 9.281683014831804e-06,
+      "loss": 0.6324,
+      "step": 4395
+    },
+    {
+      "epoch": 0.5724326434986333,
+      "grad_norm": 2.6841681003570557,
+      "learning_rate": 9.280603407717306e-06,
+      "loss": 0.5907,
+      "step": 4398
+    },
+    {
+      "epoch": 0.5728231159703241,
+      "grad_norm": 2.358556032180786,
+      "learning_rate": 9.2795230527904e-06,
+      "loss": 0.6638,
+      "step": 4401
+    },
+    {
+      "epoch": 0.5732135884420149,
+      "grad_norm": 4.04594087600708,
+      "learning_rate": 9.278441950239819e-06,
+      "loss": 0.6402,
+      "step": 4404
+    },
+    {
+      "epoch": 0.5736040609137056,
+      "grad_norm": 2.5940234661102295,
+      "learning_rate": 9.277360100254428e-06,
+      "loss": 0.6327,
+      "step": 4407
+    },
+    {
+      "epoch": 0.5739945333853963,
+      "grad_norm": 2.4827663898468018,
+      "learning_rate": 9.276277503023226e-06,
+      "loss": 0.6645,
+      "step": 4410
+    },
+    {
+      "epoch": 0.5743850058570871,
+      "grad_norm": 2.8493151664733887,
+      "learning_rate": 9.27519415873534e-06,
+      "loss": 0.6459,
+      "step": 4413
+    },
+    {
+      "epoch": 0.5747754783287778,
+      "grad_norm": 2.4425408840179443,
+      "learning_rate": 9.274110067580031e-06,
+      "loss": 0.5905,
+      "step": 4416
+    },
+    {
+      "epoch": 0.5751659508004686,
+      "grad_norm": 2.5381839275360107,
+      "learning_rate": 9.273025229746683e-06,
+      "loss": 0.7284,
+      "step": 4419
+    },
+    {
+      "epoch": 0.5755564232721593,
+      "grad_norm": 3.9269378185272217,
+      "learning_rate": 9.271939645424816e-06,
+      "loss": 0.6981,
+      "step": 4422
+    },
+    {
+      "epoch": 0.57594689574385,
+      "grad_norm": 2.5297110080718994,
+      "learning_rate": 9.27085331480408e-06,
+      "loss": 0.7223,
+      "step": 4425
+    },
+    {
+      "epoch": 0.5763373682155408,
+      "grad_norm": 2.864983320236206,
+      "learning_rate": 9.269766238074255e-06,
+      "loss": 0.7924,
+      "step": 4428
+    },
+    {
+      "epoch": 0.5767278406872316,
+      "grad_norm": 2.4377658367156982,
+      "learning_rate": 9.26867841542525e-06,
+      "loss": 0.6703,
+      "step": 4431
+    },
+    {
+      "epoch": 0.5771183131589223,
+      "grad_norm": 2.9342570304870605,
+      "learning_rate": 9.267589847047105e-06,
+      "loss": 0.6622,
+      "step": 4434
+    },
+    {
+      "epoch": 0.577508785630613,
+      "grad_norm": 2.2394070625305176,
+      "learning_rate": 9.266500533129994e-06,
+      "loss": 0.7219,
+      "step": 4437
+    },
+    {
+      "epoch": 0.5778992581023038,
+      "grad_norm": 3.120809316635132,
+      "learning_rate": 9.265410473864214e-06,
+      "loss": 0.6276,
+      "step": 4440
+    },
+    {
+      "epoch": 0.5782897305739946,
+      "grad_norm": 4.084946155548096,
+      "learning_rate": 9.264319669440197e-06,
+      "loss": 0.6136,
+      "step": 4443
+    },
+    {
+      "epoch": 0.5786802030456852,
+      "grad_norm": 3.8687705993652344,
+      "learning_rate": 9.263228120048504e-06,
+      "loss": 0.698,
+      "step": 4446
+    },
+    {
+      "epoch": 0.579070675517376,
+      "grad_norm": 2.582995891571045,
+      "learning_rate": 9.262135825879827e-06,
+      "loss": 0.6726,
+      "step": 4449
+    },
+    {
+      "epoch": 0.5794611479890668,
+      "grad_norm": 3.0313069820404053,
+      "learning_rate": 9.261042787124987e-06,
+      "loss": 0.6088,
+      "step": 4452
+    },
+    {
+      "epoch": 0.5798516204607576,
+      "grad_norm": 2.4391238689422607,
+      "learning_rate": 9.259949003974938e-06,
+      "loss": 0.6239,
+      "step": 4455
+    },
+    {
+      "epoch": 0.5802420929324482,
+      "grad_norm": 2.8740012645721436,
+      "learning_rate": 9.258854476620758e-06,
+      "loss": 0.6486,
+      "step": 4458
+    },
+    {
+      "epoch": 0.580632565404139,
+      "grad_norm": 2.8367395401000977,
+      "learning_rate": 9.257759205253662e-06,
+      "loss": 0.7247,
+      "step": 4461
+    },
+    {
+      "epoch": 0.5810230378758298,
+      "grad_norm": 2.558208703994751,
+      "learning_rate": 9.256663190064987e-06,
+      "loss": 0.6522,
+      "step": 4464
+    },
+    {
+      "epoch": 0.5814135103475205,
+      "grad_norm": 2.159445285797119,
+      "learning_rate": 9.25556643124621e-06,
+      "loss": 0.5987,
+      "step": 4467
+    },
+    {
+      "epoch": 0.5818039828192112,
+      "grad_norm": 2.406489849090576,
+      "learning_rate": 9.25446892898893e-06,
+      "loss": 0.7873,
+      "step": 4470
+    },
+    {
+      "epoch": 0.582194455290902,
+      "grad_norm": 2.4776813983917236,
+      "learning_rate": 9.253370683484876e-06,
+      "loss": 0.6753,
+      "step": 4473
+    },
+    {
+      "epoch": 0.5825849277625927,
+      "grad_norm": 2.6619131565093994,
+      "learning_rate": 9.252271694925913e-06,
+      "loss": 0.7201,
+      "step": 4476
+    },
+    {
+      "epoch": 0.5829754002342835,
+      "grad_norm": 2.598264455795288,
+      "learning_rate": 9.25117196350403e-06,
+      "loss": 0.679,
+      "step": 4479
+    },
+    {
+      "epoch": 0.5833658727059742,
+      "grad_norm": 3.2932846546173096,
+      "learning_rate": 9.250071489411348e-06,
+      "loss": 0.6224,
+      "step": 4482
+    },
+    {
+      "epoch": 0.583756345177665,
+      "grad_norm": 2.7969069480895996,
+      "learning_rate": 9.248970272840116e-06,
+      "loss": 0.5868,
+      "step": 4485
+    },
+    {
+      "epoch": 0.5841468176493557,
+      "grad_norm": 2.904038190841675,
+      "learning_rate": 9.247868313982719e-06,
+      "loss": 0.6435,
+      "step": 4488
+    },
+    {
+      "epoch": 0.5845372901210465,
+      "grad_norm": 2.6889734268188477,
+      "learning_rate": 9.246765613031661e-06,
+      "loss": 0.6655,
+      "step": 4491
+    },
+    {
+      "epoch": 0.5849277625927373,
+      "grad_norm": 2.590397834777832,
+      "learning_rate": 9.245662170179586e-06,
+      "loss": 0.6866,
+      "step": 4494
+    },
+    {
+      "epoch": 0.5853182350644279,
+      "grad_norm": 2.5878326892852783,
+      "learning_rate": 9.24455798561926e-06,
+      "loss": 0.5923,
+      "step": 4497
+    },
+    {
+      "epoch": 0.5857087075361187,
+      "grad_norm": 3.6354682445526123,
+      "learning_rate": 9.243453059543586e-06,
+      "loss": 0.7215,
+      "step": 4500
+    },
+    {
+      "epoch": 0.5860991800078095,
+      "grad_norm": 2.499769687652588,
+      "learning_rate": 9.242347392145587e-06,
+      "loss": 0.7189,
+      "step": 4503
+    },
+    {
+      "epoch": 0.5864896524795002,
+      "grad_norm": 2.806671619415283,
+      "learning_rate": 9.241240983618423e-06,
+      "loss": 0.8219,
+      "step": 4506
+    },
+    {
+      "epoch": 0.5868801249511909,
+      "grad_norm": 2.4375195503234863,
+      "learning_rate": 9.240133834155382e-06,
+      "loss": 0.7025,
+      "step": 4509
+    },
+    {
+      "epoch": 0.5872705974228817,
+      "grad_norm": 3.987379789352417,
+      "learning_rate": 9.239025943949882e-06,
+      "loss": 0.6628,
+      "step": 4512
+    },
+    {
+      "epoch": 0.5876610698945725,
+      "grad_norm": 2.5904123783111572,
+      "learning_rate": 9.237917313195465e-06,
+      "loss": 0.6761,
+      "step": 4515
+    },
+    {
+      "epoch": 0.5880515423662632,
+      "grad_norm": 2.4467389583587646,
+      "learning_rate": 9.236807942085809e-06,
+      "loss": 0.5364,
+      "step": 4518
+    },
+    {
+      "epoch": 0.5884420148379539,
+      "grad_norm": 3.339595317840576,
+      "learning_rate": 9.235697830814718e-06,
+      "loss": 0.6519,
+      "step": 4521
+    },
+    {
+      "epoch": 0.5888324873096447,
+      "grad_norm": 4.496737003326416,
+      "learning_rate": 9.234586979576127e-06,
+      "loss": 0.6588,
+      "step": 4524
+    },
+    {
+      "epoch": 0.5892229597813354,
+      "grad_norm": 2.4723331928253174,
+      "learning_rate": 9.2334753885641e-06,
+      "loss": 0.6806,
+      "step": 4527
+    },
+    {
+      "epoch": 0.5896134322530262,
+      "grad_norm": 2.589547872543335,
+      "learning_rate": 9.232363057972828e-06,
+      "loss": 0.5903,
+      "step": 4530
+    },
+    {
+      "epoch": 0.5900039047247169,
+      "grad_norm": 3.4298384189605713,
+      "learning_rate": 9.231249987996632e-06,
+      "loss": 0.6546,
+      "step": 4533
+    },
+    {
+      "epoch": 0.5903943771964076,
+      "grad_norm": 2.480419158935547,
+      "learning_rate": 9.230136178829967e-06,
+      "loss": 0.7137,
+      "step": 4536
+    },
+    {
+      "epoch": 0.5907848496680984,
+      "grad_norm": 2.6856725215911865,
+      "learning_rate": 9.229021630667407e-06,
+      "loss": 0.6591,
+      "step": 4539
+    },
+    {
+      "epoch": 0.5911753221397892,
+      "grad_norm": 4.8847975730896,
+      "learning_rate": 9.227906343703668e-06,
+      "loss": 0.756,
+      "step": 4542
+    },
+    {
+      "epoch": 0.5915657946114798,
+      "grad_norm": 2.5259408950805664,
+      "learning_rate": 9.226790318133583e-06,
+      "loss": 0.592,
+      "step": 4545
+    },
+    {
+      "epoch": 0.5919562670831706,
+      "grad_norm": 3.381965160369873,
+      "learning_rate": 9.225673554152122e-06,
+      "loss": 0.7512,
+      "step": 4548
+    },
+    {
+      "epoch": 0.5923467395548614,
+      "grad_norm": 2.3359241485595703,
+      "learning_rate": 9.224556051954381e-06,
+      "loss": 0.6277,
+      "step": 4551
+    },
+    {
+      "epoch": 0.5927372120265522,
+      "grad_norm": 2.8028383255004883,
+      "learning_rate": 9.223437811735583e-06,
+      "loss": 0.7311,
+      "step": 4554
+    },
+    {
+      "epoch": 0.5931276844982428,
+      "grad_norm": 2.1525771617889404,
+      "learning_rate": 9.222318833691085e-06,
+      "loss": 0.6426,
+      "step": 4557
+    },
+    {
+      "epoch": 0.5935181569699336,
+      "grad_norm": 2.6036198139190674,
+      "learning_rate": 9.22119911801637e-06,
+      "loss": 0.6085,
+      "step": 4560
+    },
+    {
+      "epoch": 0.5939086294416244,
+      "grad_norm": 2.456698417663574,
+      "learning_rate": 9.220078664907048e-06,
+      "loss": 0.6745,
+      "step": 4563
+    },
+    {
+      "epoch": 0.5942991019133151,
+      "grad_norm": 3.952479600906372,
+      "learning_rate": 9.218957474558862e-06,
+      "loss": 0.6955,
+      "step": 4566
+    },
+    {
+      "epoch": 0.5946895743850058,
+      "grad_norm": 3.0322844982147217,
+      "learning_rate": 9.217835547167682e-06,
+      "loss": 0.7331,
+      "step": 4569
+    },
+    {
+      "epoch": 0.5950800468566966,
+      "grad_norm": 2.9084672927856445,
+      "learning_rate": 9.216712882929503e-06,
+      "loss": 0.6295,
+      "step": 4572
+    },
+    {
+      "epoch": 0.5954705193283873,
+      "grad_norm": 2.524294376373291,
+      "learning_rate": 9.215589482040455e-06,
+      "loss": 0.6615,
+      "step": 4575
+    },
+    {
+      "epoch": 0.5958609918000781,
+      "grad_norm": 2.533430576324463,
+      "learning_rate": 9.21446534469679e-06,
+      "loss": 0.6768,
+      "step": 4578
+    },
+    {
+      "epoch": 0.5962514642717689,
+      "grad_norm": 2.4197540283203125,
+      "learning_rate": 9.213340471094899e-06,
+      "loss": 0.688,
+      "step": 4581
+    },
+    {
+      "epoch": 0.5966419367434596,
+      "grad_norm": 2.344994068145752,
+      "learning_rate": 9.212214861431289e-06,
+      "loss": 0.7125,
+      "step": 4584
+    },
+    {
+      "epoch": 0.5970324092151503,
+      "grad_norm": 2.436912775039673,
+      "learning_rate": 9.211088515902604e-06,
+      "loss": 0.6781,
+      "step": 4587
+    },
+    {
+      "epoch": 0.5974228816868411,
+      "grad_norm": 2.6664328575134277,
+      "learning_rate": 9.209961434705614e-06,
+      "loss": 0.6564,
+      "step": 4590
+    },
+    {
+      "epoch": 0.5978133541585319,
+      "grad_norm": 3.217984437942505,
+      "learning_rate": 9.20883361803722e-06,
+      "loss": 0.6577,
+      "step": 4593
+    },
+    {
+      "epoch": 0.5982038266302225,
+      "grad_norm": 2.4378035068511963,
+      "learning_rate": 9.207705066094445e-06,
+      "loss": 0.5963,
+      "step": 4596
+    },
+    {
+      "epoch": 0.5985942991019133,
+      "grad_norm": 2.4753284454345703,
+      "learning_rate": 9.206575779074448e-06,
+      "loss": 0.6725,
+      "step": 4599
+    },
+    {
+      "epoch": 0.5989847715736041,
+      "grad_norm": 2.7560603618621826,
+      "learning_rate": 9.20544575717451e-06,
+      "loss": 0.6058,
+      "step": 4602
+    },
+    {
+      "epoch": 0.5993752440452949,
+      "grad_norm": 2.4613637924194336,
+      "learning_rate": 9.204315000592046e-06,
+      "loss": 0.7212,
+      "step": 4605
+    },
+    {
+      "epoch": 0.5997657165169855,
+      "grad_norm": 2.4374942779541016,
+      "learning_rate": 9.203183509524596e-06,
+      "loss": 0.5989,
+      "step": 4608
+    },
+    {
+      "epoch": 0.6001561889886763,
+      "grad_norm": 2.783355236053467,
+      "learning_rate": 9.202051284169829e-06,
+      "loss": 0.6998,
+      "step": 4611
+    },
+    {
+      "epoch": 0.6005466614603671,
+      "grad_norm": 2.803696870803833,
+      "learning_rate": 9.200918324725543e-06,
+      "loss": 0.6278,
+      "step": 4614
+    },
+    {
+      "epoch": 0.6009371339320578,
+      "grad_norm": 3.1450366973876953,
+      "learning_rate": 9.199784631389663e-06,
+      "loss": 0.7003,
+      "step": 4617
+    },
+    {
+      "epoch": 0.6013276064037485,
+      "grad_norm": 3.2626290321350098,
+      "learning_rate": 9.198650204360241e-06,
+      "loss": 0.5592,
+      "step": 4620
+    },
+    {
+      "epoch": 0.6017180788754393,
+      "grad_norm": 2.694031238555908,
+      "learning_rate": 9.197515043835463e-06,
+      "loss": 0.6437,
+      "step": 4623
+    },
+    {
+      "epoch": 0.60210855134713,
+      "grad_norm": 3.2066242694854736,
+      "learning_rate": 9.196379150013638e-06,
+      "loss": 0.6952,
+      "step": 4626
+    },
+    {
+      "epoch": 0.6024990238188208,
+      "grad_norm": 3.181091547012329,
+      "learning_rate": 9.195242523093202e-06,
+      "loss": 0.6107,
+      "step": 4629
+    },
+    {
+      "epoch": 0.6028894962905115,
+      "grad_norm": 3.754946231842041,
+      "learning_rate": 9.194105163272722e-06,
+      "loss": 0.7689,
+      "step": 4632
+    },
+    {
+      "epoch": 0.6032799687622022,
+      "grad_norm": 3.5265393257141113,
+      "learning_rate": 9.192967070750895e-06,
+      "loss": 0.6393,
+      "step": 4635
+    },
+    {
+      "epoch": 0.603670441233893,
+      "grad_norm": 3.310450315475464,
+      "learning_rate": 9.191828245726539e-06,
+      "loss": 0.718,
+      "step": 4638
+    },
+    {
+      "epoch": 0.6040609137055838,
+      "grad_norm": 2.7861340045928955,
+      "learning_rate": 9.19068868839861e-06,
+      "loss": 0.6306,
+      "step": 4641
+    },
+    {
+      "epoch": 0.6044513861772745,
+      "grad_norm": 2.8099071979522705,
+      "learning_rate": 9.189548398966181e-06,
+      "loss": 0.7781,
+      "step": 4644
+    },
+    {
+      "epoch": 0.6048418586489652,
+      "grad_norm": 2.284885883331299,
+      "learning_rate": 9.18840737762846e-06,
+      "loss": 0.5758,
+      "step": 4647
+    },
+    {
+      "epoch": 0.605232331120656,
+      "grad_norm": 2.39620304107666,
+      "learning_rate": 9.187265624584782e-06,
+      "loss": 0.5665,
+      "step": 4650
+    },
+    {
+      "epoch": 0.6056228035923468,
+      "grad_norm": 2.680431604385376,
+      "learning_rate": 9.186123140034607e-06,
+      "loss": 0.7102,
+      "step": 4653
+    },
+    {
+      "epoch": 0.6060132760640374,
+      "grad_norm": 2.5489349365234375,
+      "learning_rate": 9.184979924177527e-06,
+      "loss": 0.61,
+      "step": 4656
+    },
+    {
+      "epoch": 0.6064037485357282,
+      "grad_norm": 2.9585001468658447,
+      "learning_rate": 9.183835977213257e-06,
+      "loss": 0.6092,
+      "step": 4659
+    },
+    {
+      "epoch": 0.606794221007419,
+      "grad_norm": 2.7119574546813965,
+      "learning_rate": 9.182691299341643e-06,
+      "loss": 0.6834,
+      "step": 4662
+    },
+    {
+      "epoch": 0.6071846934791097,
+      "grad_norm": 2.6506659984588623,
+      "learning_rate": 9.181545890762661e-06,
+      "loss": 0.6116,
+      "step": 4665
+    },
+    {
+      "epoch": 0.6075751659508005,
+      "grad_norm": 2.5166819095611572,
+      "learning_rate": 9.180399751676407e-06,
+      "loss": 0.6865,
+      "step": 4668
+    },
+    {
+      "epoch": 0.6079656384224912,
+      "grad_norm": 3.43406343460083,
+      "learning_rate": 9.17925288228311e-06,
+      "loss": 0.6894,
+      "step": 4671
+    },
+    {
+      "epoch": 0.608356110894182,
+      "grad_norm": 2.9682881832122803,
+      "learning_rate": 9.178105282783127e-06,
+      "loss": 0.6195,
+      "step": 4674
+    },
+    {
+      "epoch": 0.6087465833658727,
+      "grad_norm": 4.136783123016357,
+      "learning_rate": 9.17695695337694e-06,
+      "loss": 0.5971,
+      "step": 4677
+    },
+    {
+      "epoch": 0.6091370558375635,
+      "grad_norm": 2.4573512077331543,
+      "learning_rate": 9.175807894265161e-06,
+      "loss": 0.7037,
+      "step": 4680
+    },
+    {
+      "epoch": 0.6095275283092542,
+      "grad_norm": 2.2731971740722656,
+      "learning_rate": 9.174658105648526e-06,
+      "loss": 0.6454,
+      "step": 4683
+    },
+    {
+      "epoch": 0.6099180007809449,
+      "grad_norm": 3.480694532394409,
+      "learning_rate": 9.173507587727904e-06,
+      "loss": 0.6724,
+      "step": 4686
+    },
+    {
+      "epoch": 0.6103084732526357,
+      "grad_norm": 2.483057975769043,
+      "learning_rate": 9.172356340704285e-06,
+      "loss": 0.7026,
+      "step": 4689
+    },
+    {
+      "epoch": 0.6106989457243265,
+      "grad_norm": 2.4843204021453857,
+      "learning_rate": 9.171204364778791e-06,
+      "loss": 0.6642,
+      "step": 4692
+    },
+    {
+      "epoch": 0.6110894181960171,
+      "grad_norm": 2.6703476905822754,
+      "learning_rate": 9.17005166015267e-06,
+      "loss": 0.5992,
+      "step": 4695
+    },
+    {
+      "epoch": 0.6114798906677079,
+      "grad_norm": 2.2671613693237305,
+      "learning_rate": 9.168898227027296e-06,
+      "loss": 0.6149,
+      "step": 4698
+    },
+    {
+      "epoch": 0.6118703631393987,
+      "grad_norm": 2.83113169670105,
+      "learning_rate": 9.167744065604171e-06,
+      "loss": 0.7512,
+      "step": 4701
+    },
+    {
+      "epoch": 0.6122608356110895,
+      "grad_norm": 2.2886712551116943,
+      "learning_rate": 9.166589176084925e-06,
+      "loss": 0.603,
+      "step": 4704
+    },
+    {
+      "epoch": 0.6126513080827801,
+      "grad_norm": 2.4751598834991455,
+      "learning_rate": 9.165433558671318e-06,
+      "loss": 0.6055,
+      "step": 4707
+    },
+    {
+      "epoch": 0.6130417805544709,
+      "grad_norm": 2.5859594345092773,
+      "learning_rate": 9.164277213565228e-06,
+      "loss": 0.6932,
+      "step": 4710
+    },
+    {
+      "epoch": 0.6134322530261617,
+      "grad_norm": 2.6246237754821777,
+      "learning_rate": 9.163120140968671e-06,
+      "loss": 0.5861,
+      "step": 4713
+    },
+    {
+      "epoch": 0.6138227254978524,
+      "grad_norm": 2.5987634658813477,
+      "learning_rate": 9.161962341083784e-06,
+      "loss": 0.729,
+      "step": 4716
+    },
+    {
+      "epoch": 0.6142131979695431,
+      "grad_norm": 2.6038060188293457,
+      "learning_rate": 9.160803814112829e-06,
+      "loss": 0.6701,
+      "step": 4719
+    },
+    {
+      "epoch": 0.6146036704412339,
+      "grad_norm": 3.065664768218994,
+      "learning_rate": 9.159644560258201e-06,
+      "loss": 0.7233,
+      "step": 4722
+    },
+    {
+      "epoch": 0.6149941429129246,
+      "grad_norm": 2.4268383979797363,
+      "learning_rate": 9.15848457972242e-06,
+      "loss": 0.6326,
+      "step": 4725
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 2.269376277923584,
+      "learning_rate": 9.157323872708131e-06,
+      "loss": 0.6076,
+      "step": 4728
+    },
+    {
+      "epoch": 0.6157750878563061,
+      "grad_norm": 2.622344970703125,
+      "learning_rate": 9.156162439418108e-06,
+      "loss": 0.6793,
+      "step": 4731
+    },
+    {
+      "epoch": 0.6161655603279969,
+      "grad_norm": 3.469761371612549,
+      "learning_rate": 9.15500028005525e-06,
+      "loss": 0.7301,
+      "step": 4734
+    },
+    {
+      "epoch": 0.6165560327996876,
+      "grad_norm": 2.781172752380371,
+      "learning_rate": 9.153837394822582e-06,
+      "loss": 0.7908,
+      "step": 4737
+    },
+    {
+      "epoch": 0.6169465052713784,
+      "grad_norm": 2.506148099899292,
+      "learning_rate": 9.15267378392326e-06,
+      "loss": 0.7075,
+      "step": 4740
+    },
+    {
+      "epoch": 0.6173369777430691,
+      "grad_norm": 3.1186656951904297,
+      "learning_rate": 9.151509447560566e-06,
+      "loss": 0.5936,
+      "step": 4743
+    },
+    {
+      "epoch": 0.6177274502147598,
+      "grad_norm": 2.5772218704223633,
+      "learning_rate": 9.150344385937904e-06,
+      "loss": 0.6029,
+      "step": 4746
+    },
+    {
+      "epoch": 0.6181179226864506,
+      "grad_norm": 2.423085927963257,
+      "learning_rate": 9.149178599258807e-06,
+      "loss": 0.7071,
+      "step": 4749
+    },
+    {
+      "epoch": 0.6185083951581414,
+      "grad_norm": 3.1963913440704346,
+      "learning_rate": 9.14801208772694e-06,
+      "loss": 0.6297,
+      "step": 4752
+    },
+    {
+      "epoch": 0.6188988676298322,
+      "grad_norm": 2.7055928707122803,
+      "learning_rate": 9.146844851546087e-06,
+      "loss": 0.7194,
+      "step": 4755
+    },
+    {
+      "epoch": 0.6192893401015228,
+      "grad_norm": 2.607105016708374,
+      "learning_rate": 9.145676890920161e-06,
+      "loss": 0.6897,
+      "step": 4758
+    },
+    {
+      "epoch": 0.6196798125732136,
+      "grad_norm": 2.645681381225586,
+      "learning_rate": 9.144508206053203e-06,
+      "loss": 0.636,
+      "step": 4761
+    },
+    {
+      "epoch": 0.6200702850449044,
+      "grad_norm": 2.521937131881714,
+      "learning_rate": 9.143338797149381e-06,
+      "loss": 0.6608,
+      "step": 4764
+    },
+    {
+      "epoch": 0.6204607575165951,
+      "grad_norm": 2.7200021743774414,
+      "learning_rate": 9.142168664412987e-06,
+      "loss": 0.6228,
+      "step": 4767
+    },
+    {
+      "epoch": 0.6208512299882858,
+      "grad_norm": 3.465608835220337,
+      "learning_rate": 9.140997808048442e-06,
+      "loss": 0.6801,
+      "step": 4770
+    },
+    {
+      "epoch": 0.6212417024599766,
+      "grad_norm": 2.4828827381134033,
+      "learning_rate": 9.139826228260292e-06,
+      "loss": 0.6134,
+      "step": 4773
+    },
+    {
+      "epoch": 0.6216321749316673,
+      "grad_norm": 2.365074634552002,
+      "learning_rate": 9.138653925253207e-06,
+      "loss": 0.5044,
+      "step": 4776
+    },
+    {
+      "epoch": 0.6220226474033581,
+      "grad_norm": 2.630094051361084,
+      "learning_rate": 9.137480899231987e-06,
+      "loss": 0.5772,
+      "step": 4779
+    },
+    {
+      "epoch": 0.6224131198750488,
+      "grad_norm": 2.5148115158081055,
+      "learning_rate": 9.13630715040156e-06,
+      "loss": 0.6968,
+      "step": 4782
+    },
+    {
+      "epoch": 0.6228035923467395,
+      "grad_norm": 2.7790770530700684,
+      "learning_rate": 9.135132678966975e-06,
+      "loss": 0.5712,
+      "step": 4785
+    },
+    {
+      "epoch": 0.6231940648184303,
+      "grad_norm": 2.780951738357544,
+      "learning_rate": 9.13395748513341e-06,
+      "loss": 0.6046,
+      "step": 4788
+    },
+    {
+      "epoch": 0.6235845372901211,
+      "grad_norm": 2.728588819503784,
+      "learning_rate": 9.132781569106168e-06,
+      "loss": 0.734,
+      "step": 4791
+    },
+    {
+      "epoch": 0.6239750097618117,
+      "grad_norm": 2.4009926319122314,
+      "learning_rate": 9.131604931090681e-06,
+      "loss": 0.62,
+      "step": 4794
+    },
+    {
+      "epoch": 0.6243654822335025,
+      "grad_norm": 3.470907211303711,
+      "learning_rate": 9.130427571292503e-06,
+      "loss": 0.7862,
+      "step": 4797
+    },
+    {
+      "epoch": 0.6247559547051933,
+      "grad_norm": 3.43316912651062,
+      "learning_rate": 9.129249489917317e-06,
+      "loss": 0.651,
+      "step": 4800
+    },
+    {
+      "epoch": 0.6251464271768841,
+      "grad_norm": 2.8522374629974365,
+      "learning_rate": 9.128070687170933e-06,
+      "loss": 0.5899,
+      "step": 4803
+    },
+    {
+      "epoch": 0.6255368996485747,
+      "grad_norm": 2.6155104637145996,
+      "learning_rate": 9.126891163259286e-06,
+      "loss": 0.6208,
+      "step": 4806
+    },
+    {
+      "epoch": 0.6259273721202655,
+      "grad_norm": 2.6993188858032227,
+      "learning_rate": 9.125710918388431e-06,
+      "loss": 0.6091,
+      "step": 4809
+    },
+    {
+      "epoch": 0.6263178445919563,
+      "grad_norm": 3.6405208110809326,
+      "learning_rate": 9.12452995276456e-06,
+      "loss": 0.6169,
+      "step": 4812
+    },
+    {
+      "epoch": 0.626708317063647,
+      "grad_norm": 2.4731392860412598,
+      "learning_rate": 9.123348266593983e-06,
+      "loss": 0.6637,
+      "step": 4815
+    },
+    {
+      "epoch": 0.6270987895353377,
+      "grad_norm": 2.377974271774292,
+      "learning_rate": 9.122165860083136e-06,
+      "loss": 0.6365,
+      "step": 4818
+    },
+    {
+      "epoch": 0.6274892620070285,
+      "grad_norm": 2.286585569381714,
+      "learning_rate": 9.120982733438587e-06,
+      "loss": 0.6843,
+      "step": 4821
+    },
+    {
+      "epoch": 0.6278797344787193,
+      "grad_norm": 2.7373149394989014,
+      "learning_rate": 9.119798886867025e-06,
+      "loss": 0.6896,
+      "step": 4824
+    },
+    {
+      "epoch": 0.62827020695041,
+      "grad_norm": 2.6947765350341797,
+      "learning_rate": 9.118614320575261e-06,
+      "loss": 0.6867,
+      "step": 4827
+    },
+    {
+      "epoch": 0.6286606794221007,
+      "grad_norm": 2.763946771621704,
+      "learning_rate": 9.117429034770241e-06,
+      "loss": 0.6705,
+      "step": 4830
+    },
+    {
+      "epoch": 0.6290511518937915,
+      "grad_norm": 2.622236728668213,
+      "learning_rate": 9.116243029659032e-06,
+      "loss": 0.7195,
+      "step": 4833
+    },
+    {
+      "epoch": 0.6294416243654822,
+      "grad_norm": 2.6941754817962646,
+      "learning_rate": 9.115056305448822e-06,
+      "loss": 0.7228,
+      "step": 4836
+    },
+    {
+      "epoch": 0.629832096837173,
+      "grad_norm": 3.360145092010498,
+      "learning_rate": 9.113868862346934e-06,
+      "loss": 0.679,
+      "step": 4839
+    },
+    {
+      "epoch": 0.6302225693088638,
+      "grad_norm": 2.4699013233184814,
+      "learning_rate": 9.11268070056081e-06,
+      "loss": 0.6504,
+      "step": 4842
+    },
+    {
+      "epoch": 0.6306130417805544,
+      "grad_norm": 2.265643835067749,
+      "learning_rate": 9.111491820298019e-06,
+      "loss": 0.6572,
+      "step": 4845
+    },
+    {
+      "epoch": 0.6310035142522452,
+      "grad_norm": 2.609482765197754,
+      "learning_rate": 9.110302221766257e-06,
+      "loss": 0.643,
+      "step": 4848
+    },
+    {
+      "epoch": 0.631393986723936,
+      "grad_norm": 2.8599019050598145,
+      "learning_rate": 9.109111905173342e-06,
+      "loss": 0.7884,
+      "step": 4851
+    },
+    {
+      "epoch": 0.6317844591956268,
+      "grad_norm": 2.703747272491455,
+      "learning_rate": 9.107920870727223e-06,
+      "loss": 0.6817,
+      "step": 4854
+    },
+    {
+      "epoch": 0.6321749316673174,
+      "grad_norm": 2.2183678150177,
+      "learning_rate": 9.106729118635968e-06,
+      "loss": 0.6101,
+      "step": 4857
+    },
+    {
+      "epoch": 0.6325654041390082,
+      "grad_norm": 2.5604469776153564,
+      "learning_rate": 9.105536649107778e-06,
+      "loss": 0.7131,
+      "step": 4860
+    },
+    {
+      "epoch": 0.632955876610699,
+      "grad_norm": 2.3223118782043457,
+      "learning_rate": 9.10434346235097e-06,
+      "loss": 0.6237,
+      "step": 4863
+    },
+    {
+      "epoch": 0.6333463490823897,
+      "grad_norm": 2.4612014293670654,
+      "learning_rate": 9.103149558573995e-06,
+      "loss": 0.695,
+      "step": 4866
+    },
+    {
+      "epoch": 0.6337368215540804,
+      "grad_norm": 3.208245277404785,
+      "learning_rate": 9.101954937985422e-06,
+      "loss": 0.6685,
+      "step": 4869
+    },
+    {
+      "epoch": 0.6341272940257712,
+      "grad_norm": 3.3025028705596924,
+      "learning_rate": 9.100759600793951e-06,
+      "loss": 0.6657,
+      "step": 4872
+    },
+    {
+      "epoch": 0.6345177664974619,
+      "grad_norm": 2.8037331104278564,
+      "learning_rate": 9.099563547208406e-06,
+      "loss": 0.573,
+      "step": 4875
+    },
+    {
+      "epoch": 0.6349082389691527,
+      "grad_norm": 2.62872052192688,
+      "learning_rate": 9.098366777437733e-06,
+      "loss": 0.7397,
+      "step": 4878
+    },
+    {
+      "epoch": 0.6352987114408434,
+      "grad_norm": 2.564282178878784,
+      "learning_rate": 9.097169291691007e-06,
+      "loss": 0.626,
+      "step": 4881
+    },
+    {
+      "epoch": 0.6356891839125342,
+      "grad_norm": 2.9302821159362793,
+      "learning_rate": 9.095971090177421e-06,
+      "loss": 0.7665,
+      "step": 4884
+    },
+    {
+      "epoch": 0.6360796563842249,
+      "grad_norm": 2.7549326419830322,
+      "learning_rate": 9.094772173106305e-06,
+      "loss": 0.7412,
+      "step": 4887
+    },
+    {
+      "epoch": 0.6364701288559157,
+      "grad_norm": 2.602814197540283,
+      "learning_rate": 9.093572540687104e-06,
+      "loss": 0.6038,
+      "step": 4890
+    },
+    {
+      "epoch": 0.6368606013276064,
+      "grad_norm": 2.4310333728790283,
+      "learning_rate": 9.09237219312939e-06,
+      "loss": 0.6858,
+      "step": 4893
+    },
+    {
+      "epoch": 0.6372510737992971,
+      "grad_norm": 2.8889222145080566,
+      "learning_rate": 9.091171130642866e-06,
+      "loss": 0.6245,
+      "step": 4896
+    },
+    {
+      "epoch": 0.6376415462709879,
+      "grad_norm": 2.956254005432129,
+      "learning_rate": 9.089969353437351e-06,
+      "loss": 0.7212,
+      "step": 4899
+    },
+    {
+      "epoch": 0.6380320187426787,
+      "grad_norm": 2.366194248199463,
+      "learning_rate": 9.088766861722793e-06,
+      "loss": 0.6385,
+      "step": 4902
+    },
+    {
+      "epoch": 0.6384224912143693,
+      "grad_norm": 2.7603938579559326,
+      "learning_rate": 9.087563655709266e-06,
+      "loss": 0.7277,
+      "step": 4905
+    },
+    {
+      "epoch": 0.6388129636860601,
+      "grad_norm": 3.687945604324341,
+      "learning_rate": 9.086359735606969e-06,
+      "loss": 0.6555,
+      "step": 4908
+    },
+    {
+      "epoch": 0.6392034361577509,
+      "grad_norm": 3.259951591491699,
+      "learning_rate": 9.085155101626221e-06,
+      "loss": 0.7351,
+      "step": 4911
+    },
+    {
+      "epoch": 0.6395939086294417,
+      "grad_norm": 2.4649972915649414,
+      "learning_rate": 9.083949753977471e-06,
+      "loss": 0.5857,
+      "step": 4914
+    },
+    {
+      "epoch": 0.6399843811011323,
+      "grad_norm": 2.663567304611206,
+      "learning_rate": 9.08274369287129e-06,
+      "loss": 0.648,
+      "step": 4917
+    },
+    {
+      "epoch": 0.6403748535728231,
+      "grad_norm": 2.2157647609710693,
+      "learning_rate": 9.081536918518377e-06,
+      "loss": 0.6431,
+      "step": 4920
+    },
+    {
+      "epoch": 0.6407653260445139,
+      "grad_norm": 2.5896902084350586,
+      "learning_rate": 9.080329431129548e-06,
+      "loss": 0.6592,
+      "step": 4923
+    },
+    {
+      "epoch": 0.6411557985162046,
+      "grad_norm": 3.887195348739624,
+      "learning_rate": 9.07912123091575e-06,
+      "loss": 0.7259,
+      "step": 4926
+    },
+    {
+      "epoch": 0.6415462709878954,
+      "grad_norm": 2.6314144134521484,
+      "learning_rate": 9.077912318088057e-06,
+      "loss": 0.5699,
+      "step": 4929
+    },
+    {
+      "epoch": 0.6419367434595861,
+      "grad_norm": 2.7441248893737793,
+      "learning_rate": 9.076702692857661e-06,
+      "loss": 0.6653,
+      "step": 4932
+    },
+    {
+      "epoch": 0.6423272159312768,
+      "grad_norm": 3.4282479286193848,
+      "learning_rate": 9.075492355435878e-06,
+      "loss": 0.634,
+      "step": 4935
+    },
+    {
+      "epoch": 0.6427176884029676,
+      "grad_norm": 2.2707557678222656,
+      "learning_rate": 9.074281306034156e-06,
+      "loss": 0.6404,
+      "step": 4938
+    },
+    {
+      "epoch": 0.6431081608746584,
+      "grad_norm": 2.613741397857666,
+      "learning_rate": 9.073069544864062e-06,
+      "loss": 0.7175,
+      "step": 4941
+    },
+    {
+      "epoch": 0.643498633346349,
+      "grad_norm": 3.953160285949707,
+      "learning_rate": 9.071857072137284e-06,
+      "loss": 0.6577,
+      "step": 4944
+    },
+    {
+      "epoch": 0.6438891058180398,
+      "grad_norm": 2.4851486682891846,
+      "learning_rate": 9.070643888065642e-06,
+      "loss": 0.6288,
+      "step": 4947
+    },
+    {
+      "epoch": 0.6442795782897306,
+      "grad_norm": 2.4990899562835693,
+      "learning_rate": 9.069429992861078e-06,
+      "loss": 0.7342,
+      "step": 4950
+    },
+    {
+      "epoch": 0.6446700507614214,
+      "grad_norm": 2.3601980209350586,
+      "learning_rate": 9.068215386735655e-06,
+      "loss": 0.568,
+      "step": 4953
+    },
+    {
+      "epoch": 0.645060523233112,
+      "grad_norm": 2.9806814193725586,
+      "learning_rate": 9.06700006990156e-06,
+      "loss": 0.6322,
+      "step": 4956
+    },
+    {
+      "epoch": 0.6454509957048028,
+      "grad_norm": 3.1821587085723877,
+      "learning_rate": 9.06578404257111e-06,
+      "loss": 0.7119,
+      "step": 4959
+    },
+    {
+      "epoch": 0.6458414681764936,
+      "grad_norm": 2.242640972137451,
+      "learning_rate": 9.064567304956741e-06,
+      "loss": 0.7047,
+      "step": 4962
+    },
+    {
+      "epoch": 0.6462319406481843,
+      "grad_norm": 2.693079710006714,
+      "learning_rate": 9.063349857271015e-06,
+      "loss": 0.6865,
+      "step": 4965
+    },
+    {
+      "epoch": 0.646622413119875,
+      "grad_norm": 3.0986106395721436,
+      "learning_rate": 9.062131699726615e-06,
+      "loss": 0.6241,
+      "step": 4968
+    },
+    {
+      "epoch": 0.6470128855915658,
+      "grad_norm": 2.6668357849121094,
+      "learning_rate": 9.060912832536354e-06,
+      "loss": 0.7457,
+      "step": 4971
+    },
+    {
+      "epoch": 0.6474033580632566,
+      "grad_norm": 2.3594837188720703,
+      "learning_rate": 9.059693255913165e-06,
+      "loss": 0.6704,
+      "step": 4974
+    },
+    {
+      "epoch": 0.6477938305349473,
+      "grad_norm": 3.7347023487091064,
+      "learning_rate": 9.058472970070102e-06,
+      "loss": 0.6353,
+      "step": 4977
+    },
+    {
+      "epoch": 0.648184303006638,
+      "grad_norm": 2.6183836460113525,
+      "learning_rate": 9.05725197522035e-06,
+      "loss": 0.6586,
+      "step": 4980
+    },
+    {
+      "epoch": 0.6485747754783288,
+      "grad_norm": 2.691641330718994,
+      "learning_rate": 9.056030271577213e-06,
+      "loss": 0.6442,
+      "step": 4983
+    },
+    {
+      "epoch": 0.6489652479500195,
+      "grad_norm": 2.6118125915527344,
+      "learning_rate": 9.054807859354122e-06,
+      "loss": 0.6896,
+      "step": 4986
+    },
+    {
+      "epoch": 0.6493557204217103,
+      "grad_norm": 2.557551383972168,
+      "learning_rate": 9.053584738764625e-06,
+      "loss": 0.7078,
+      "step": 4989
+    },
+    {
+      "epoch": 0.649746192893401,
+      "grad_norm": 2.4774162769317627,
+      "learning_rate": 9.052360910022404e-06,
+      "loss": 0.7401,
+      "step": 4992
+    },
+    {
+      "epoch": 0.6501366653650917,
+      "grad_norm": 2.230543851852417,
+      "learning_rate": 9.051136373341256e-06,
+      "loss": 0.6496,
+      "step": 4995
+    },
+    {
+      "epoch": 0.6505271378367825,
+      "grad_norm": 3.3752076625823975,
+      "learning_rate": 9.049911128935104e-06,
+      "loss": 0.5981,
+      "step": 4998
+    },
+    {
+      "epoch": 0.6509176103084733,
+      "grad_norm": 2.407163619995117,
+      "learning_rate": 9.048685177018001e-06,
+      "loss": 0.6235,
+      "step": 5001
+    },
+    {
+      "epoch": 0.6513080827801639,
+      "grad_norm": 2.324744939804077,
+      "learning_rate": 9.047458517804113e-06,
+      "loss": 0.6085,
+      "step": 5004
+    },
+    {
+      "epoch": 0.6516985552518547,
+      "grad_norm": 2.3298988342285156,
+      "learning_rate": 9.046231151507738e-06,
+      "loss": 0.5624,
+      "step": 5007
+    },
+    {
+      "epoch": 0.6520890277235455,
+      "grad_norm": 3.9053573608398438,
+      "learning_rate": 9.04500307834329e-06,
+      "loss": 0.687,
+      "step": 5010
+    },
+    {
+      "epoch": 0.6524795001952363,
+      "grad_norm": 2.682884931564331,
+      "learning_rate": 9.043774298525316e-06,
+      "loss": 0.6969,
+      "step": 5013
+    },
+    {
+      "epoch": 0.652869972666927,
+      "grad_norm": 2.7601308822631836,
+      "learning_rate": 9.04254481226848e-06,
+      "loss": 0.6236,
+      "step": 5016
+    },
+    {
+      "epoch": 0.6532604451386177,
+      "grad_norm": 2.39310622215271,
+      "learning_rate": 9.041314619787566e-06,
+      "loss": 0.6217,
+      "step": 5019
+    },
+    {
+      "epoch": 0.6536509176103085,
+      "grad_norm": 2.353529453277588,
+      "learning_rate": 9.040083721297493e-06,
+      "loss": 0.6112,
+      "step": 5022
+    },
+    {
+      "epoch": 0.6540413900819992,
+      "grad_norm": 2.530123472213745,
+      "learning_rate": 9.03885211701329e-06,
+      "loss": 0.6053,
+      "step": 5025
+    },
+    {
+      "epoch": 0.65443186255369,
+      "grad_norm": 2.888646125793457,
+      "learning_rate": 9.037619807150121e-06,
+      "loss": 0.6883,
+      "step": 5028
+    },
+    {
+      "epoch": 0.6548223350253807,
+      "grad_norm": 2.598736524581909,
+      "learning_rate": 9.036386791923265e-06,
+      "loss": 0.6944,
+      "step": 5031
+    },
+    {
+      "epoch": 0.6552128074970714,
+      "grad_norm": 2.7701828479766846,
+      "learning_rate": 9.035153071548127e-06,
+      "loss": 0.6272,
+      "step": 5034
+    },
+    {
+      "epoch": 0.6556032799687622,
+      "grad_norm": 2.3290834426879883,
+      "learning_rate": 9.033918646240236e-06,
+      "loss": 0.6815,
+      "step": 5037
+    },
+    {
+      "epoch": 0.655993752440453,
+      "grad_norm": 2.8352549076080322,
+      "learning_rate": 9.032683516215245e-06,
+      "loss": 0.662,
+      "step": 5040
+    },
+    {
+      "epoch": 0.6563842249121437,
+      "grad_norm": 4.3119659423828125,
+      "learning_rate": 9.031447681688926e-06,
+      "loss": 0.6342,
+      "step": 5043
+    },
+    {
+      "epoch": 0.6567746973838344,
+      "grad_norm": 2.828580379486084,
+      "learning_rate": 9.03021114287718e-06,
+      "loss": 0.7723,
+      "step": 5046
+    },
+    {
+      "epoch": 0.6571651698555252,
+      "grad_norm": 2.502751350402832,
+      "learning_rate": 9.028973899996022e-06,
+      "loss": 0.6425,
+      "step": 5049
+    },
+    {
+      "epoch": 0.657555642327216,
+      "grad_norm": 2.7101762294769287,
+      "learning_rate": 9.027735953261603e-06,
+      "loss": 0.6139,
+      "step": 5052
+    },
+    {
+      "epoch": 0.6579461147989066,
+      "grad_norm": 2.368980646133423,
+      "learning_rate": 9.026497302890184e-06,
+      "loss": 0.6776,
+      "step": 5055
+    },
+    {
+      "epoch": 0.6583365872705974,
+      "grad_norm": 2.5557427406311035,
+      "learning_rate": 9.025257949098158e-06,
+      "loss": 0.6162,
+      "step": 5058
+    },
+    {
+      "epoch": 0.6587270597422882,
+      "grad_norm": 2.6219143867492676,
+      "learning_rate": 9.024017892102036e-06,
+      "loss": 0.6914,
+      "step": 5061
+    },
+    {
+      "epoch": 0.659117532213979,
+      "grad_norm": 6.491698741912842,
+      "learning_rate": 9.022777132118452e-06,
+      "loss": 0.7507,
+      "step": 5064
+    },
+    {
+      "epoch": 0.6595080046856696,
+      "grad_norm": 2.6775612831115723,
+      "learning_rate": 9.021535669364167e-06,
+      "loss": 0.7287,
+      "step": 5067
+    },
+    {
+      "epoch": 0.6598984771573604,
+      "grad_norm": 3.4174671173095703,
+      "learning_rate": 9.020293504056061e-06,
+      "loss": 0.7409,
+      "step": 5070
+    },
+    {
+      "epoch": 0.6602889496290512,
+      "grad_norm": 2.259105682373047,
+      "learning_rate": 9.019050636411135e-06,
+      "loss": 0.5771,
+      "step": 5073
+    },
+    {
+      "epoch": 0.6606794221007419,
+      "grad_norm": 2.384159803390503,
+      "learning_rate": 9.01780706664652e-06,
+      "loss": 0.6882,
+      "step": 5076
+    },
+    {
+      "epoch": 0.6610698945724326,
+      "grad_norm": 2.471353054046631,
+      "learning_rate": 9.016562794979463e-06,
+      "loss": 0.6606,
+      "step": 5079
+    },
+    {
+      "epoch": 0.6614603670441234,
+      "grad_norm": 2.462852716445923,
+      "learning_rate": 9.015317821627332e-06,
+      "loss": 0.6092,
+      "step": 5082
+    },
+    {
+      "epoch": 0.6618508395158141,
+      "grad_norm": 2.5678861141204834,
+      "learning_rate": 9.01407214680763e-06,
+      "loss": 0.7437,
+      "step": 5085
+    },
+    {
+      "epoch": 0.6622413119875049,
+      "grad_norm": 2.7121429443359375,
+      "learning_rate": 9.012825770737963e-06,
+      "loss": 0.6702,
+      "step": 5088
+    },
+    {
+      "epoch": 0.6626317844591956,
+      "grad_norm": 2.5541470050811768,
+      "learning_rate": 9.011578693636078e-06,
+      "loss": 0.6909,
+      "step": 5091
+    },
+    {
+      "epoch": 0.6630222569308863,
+      "grad_norm": 3.0044548511505127,
+      "learning_rate": 9.010330915719834e-06,
+      "loss": 0.6704,
+      "step": 5094
+    },
+    {
+      "epoch": 0.6634127294025771,
+      "grad_norm": 2.6041769981384277,
+      "learning_rate": 9.009082437207215e-06,
+      "loss": 0.6867,
+      "step": 5097
+    },
+    {
+      "epoch": 0.6638032018742679,
+      "grad_norm": 2.428577423095703,
+      "learning_rate": 9.00783325831633e-06,
+      "loss": 0.6128,
+      "step": 5100
+    },
+    {
+      "epoch": 0.6641936743459587,
+      "grad_norm": 2.460480213165283,
+      "learning_rate": 9.006583379265405e-06,
+      "loss": 0.6392,
+      "step": 5103
+    },
+    {
+      "epoch": 0.6645841468176493,
+      "grad_norm": 2.7471256256103516,
+      "learning_rate": 9.00533280027279e-06,
+      "loss": 0.6698,
+      "step": 5106
+    },
+    {
+      "epoch": 0.6649746192893401,
+      "grad_norm": 2.6716361045837402,
+      "learning_rate": 9.004081521556965e-06,
+      "loss": 0.7285,
+      "step": 5109
+    },
+    {
+      "epoch": 0.6653650917610309,
+      "grad_norm": 2.468386173248291,
+      "learning_rate": 9.002829543336518e-06,
+      "loss": 0.6738,
+      "step": 5112
+    },
+    {
+      "epoch": 0.6657555642327216,
+      "grad_norm": 2.630680561065674,
+      "learning_rate": 9.001576865830173e-06,
+      "loss": 0.7629,
+      "step": 5115
+    },
+    {
+      "epoch": 0.6661460367044123,
+      "grad_norm": 2.4180047512054443,
+      "learning_rate": 9.000323489256766e-06,
+      "loss": 0.6747,
+      "step": 5118
+    },
+    {
+      "epoch": 0.6665365091761031,
+      "grad_norm": 4.522125720977783,
+      "learning_rate": 8.999069413835262e-06,
+      "loss": 0.6331,
+      "step": 5121
+    },
+    {
+      "epoch": 0.6669269816477938,
+      "grad_norm": 2.297905445098877,
+      "learning_rate": 8.997814639784743e-06,
+      "loss": 0.5958,
+      "step": 5124
+    },
+    {
+      "epoch": 0.6673174541194846,
+      "grad_norm": 2.3950231075286865,
+      "learning_rate": 8.996559167324417e-06,
+      "loss": 0.6209,
+      "step": 5127
+    },
+    {
+      "epoch": 0.6677079265911753,
+      "grad_norm": 2.4264347553253174,
+      "learning_rate": 8.995302996673613e-06,
+      "loss": 0.6492,
+      "step": 5130
+    },
+    {
+      "epoch": 0.668098399062866,
+      "grad_norm": 3.127281427383423,
+      "learning_rate": 8.99404612805178e-06,
+      "loss": 0.6959,
+      "step": 5133
+    },
+    {
+      "epoch": 0.6684888715345568,
+      "grad_norm": 2.8744657039642334,
+      "learning_rate": 8.99278856167849e-06,
+      "loss": 0.6416,
+      "step": 5136
+    },
+    {
+      "epoch": 0.6688793440062476,
+      "grad_norm": 2.994250535964966,
+      "learning_rate": 8.991530297773437e-06,
+      "loss": 0.6651,
+      "step": 5139
+    },
+    {
+      "epoch": 0.6692698164779383,
+      "grad_norm": 2.4786083698272705,
+      "learning_rate": 8.99027133655644e-06,
+      "loss": 0.6271,
+      "step": 5142
+    },
+    {
+      "epoch": 0.669660288949629,
+      "grad_norm": 2.418321371078491,
+      "learning_rate": 8.989011678247433e-06,
+      "loss": 0.6906,
+      "step": 5145
+    },
+    {
+      "epoch": 0.6700507614213198,
+      "grad_norm": 2.5830347537994385,
+      "learning_rate": 8.987751323066479e-06,
+      "loss": 0.7363,
+      "step": 5148
+    },
+    {
+      "epoch": 0.6704412338930106,
+      "grad_norm": 2.7174127101898193,
+      "learning_rate": 8.986490271233757e-06,
+      "loss": 0.7439,
+      "step": 5151
+    },
+    {
+      "epoch": 0.6708317063647012,
+      "grad_norm": 2.870361804962158,
+      "learning_rate": 8.985228522969571e-06,
+      "loss": 0.6398,
+      "step": 5154
+    },
+    {
+      "epoch": 0.671222178836392,
+      "grad_norm": 2.797898292541504,
+      "learning_rate": 8.983966078494346e-06,
+      "loss": 0.6451,
+      "step": 5157
+    },
+    {
+      "epoch": 0.6716126513080828,
+      "grad_norm": 2.4643352031707764,
+      "learning_rate": 8.98270293802863e-06,
+      "loss": 0.6963,
+      "step": 5160
+    },
+    {
+      "epoch": 0.6720031237797736,
+      "grad_norm": 2.715022325515747,
+      "learning_rate": 8.981439101793086e-06,
+      "loss": 0.6041,
+      "step": 5163
+    },
+    {
+      "epoch": 0.6723935962514642,
+      "grad_norm": 2.511801242828369,
+      "learning_rate": 8.980174570008506e-06,
+      "loss": 0.6198,
+      "step": 5166
+    },
+    {
+      "epoch": 0.672784068723155,
+      "grad_norm": 2.53893780708313,
+      "learning_rate": 8.978909342895806e-06,
+      "loss": 0.6393,
+      "step": 5169
+    },
+    {
+      "epoch": 0.6731745411948458,
+      "grad_norm": 2.6536126136779785,
+      "learning_rate": 8.977643420676009e-06,
+      "loss": 0.7114,
+      "step": 5172
+    },
+    {
+      "epoch": 0.6735650136665365,
+      "grad_norm": 2.3529465198516846,
+      "learning_rate": 8.976376803570278e-06,
+      "loss": 0.6438,
+      "step": 5175
+    },
+    {
+      "epoch": 0.6739554861382272,
+      "grad_norm": 2.5893778800964355,
+      "learning_rate": 8.975109491799883e-06,
+      "loss": 0.661,
+      "step": 5178
+    },
+    {
+      "epoch": 0.674345958609918,
+      "grad_norm": 2.3905746936798096,
+      "learning_rate": 8.973841485586224e-06,
+      "loss": 0.5583,
+      "step": 5181
+    },
+    {
+      "epoch": 0.6747364310816087,
+      "grad_norm": 2.6573519706726074,
+      "learning_rate": 8.972572785150815e-06,
+      "loss": 0.6492,
+      "step": 5184
+    },
+    {
+      "epoch": 0.6751269035532995,
+      "grad_norm": 2.810039520263672,
+      "learning_rate": 8.971303390715299e-06,
+      "loss": 0.6346,
+      "step": 5187
+    },
+    {
+      "epoch": 0.6755173760249903,
+      "grad_norm": 3.9205002784729004,
+      "learning_rate": 8.970033302501433e-06,
+      "loss": 0.7605,
+      "step": 5190
+    },
+    {
+      "epoch": 0.675907848496681,
+      "grad_norm": 2.3390591144561768,
+      "learning_rate": 8.968762520731103e-06,
+      "loss": 0.6231,
+      "step": 5193
+    },
+    {
+      "epoch": 0.6762983209683717,
+      "grad_norm": 2.8558874130249023,
+      "learning_rate": 8.96749104562631e-06,
+      "loss": 0.6479,
+      "step": 5196
+    },
+    {
+      "epoch": 0.6766887934400625,
+      "grad_norm": 2.8038036823272705,
+      "learning_rate": 8.966218877409173e-06,
+      "loss": 0.7772,
+      "step": 5199
+    },
+    {
+      "epoch": 0.6770792659117533,
+      "grad_norm": 2.364079475402832,
+      "learning_rate": 8.964946016301946e-06,
+      "loss": 0.5847,
+      "step": 5202
+    },
+    {
+      "epoch": 0.6774697383834439,
+      "grad_norm": 2.699650764465332,
+      "learning_rate": 8.963672462526991e-06,
+      "loss": 0.667,
+      "step": 5205
+    },
+    {
+      "epoch": 0.6778602108551347,
+      "grad_norm": 2.889913320541382,
+      "learning_rate": 8.962398216306794e-06,
+      "loss": 0.5679,
+      "step": 5208
+    },
+    {
+      "epoch": 0.6782506833268255,
+      "grad_norm": 3.220335006713867,
+      "learning_rate": 8.961123277863965e-06,
+      "loss": 0.7115,
+      "step": 5211
+    },
+    {
+      "epoch": 0.6786411557985162,
+      "grad_norm": 2.591503143310547,
+      "learning_rate": 8.959847647421231e-06,
+      "loss": 0.6948,
+      "step": 5214
+    },
+    {
+      "epoch": 0.6790316282702069,
+      "grad_norm": 2.682509183883667,
+      "learning_rate": 8.958571325201446e-06,
+      "loss": 0.5903,
+      "step": 5217
+    },
+    {
+      "epoch": 0.6794221007418977,
+      "grad_norm": 3.3060030937194824,
+      "learning_rate": 8.957294311427575e-06,
+      "loss": 0.6707,
+      "step": 5220
+    },
+    {
+      "epoch": 0.6798125732135885,
+      "grad_norm": 2.3943710327148438,
+      "learning_rate": 8.956016606322715e-06,
+      "loss": 0.6522,
+      "step": 5223
+    },
+    {
+      "epoch": 0.6802030456852792,
+      "grad_norm": 2.6341614723205566,
+      "learning_rate": 8.954738210110075e-06,
+      "loss": 0.5324,
+      "step": 5226
+    },
+    {
+      "epoch": 0.6805935181569699,
+      "grad_norm": 3.0761325359344482,
+      "learning_rate": 8.953459123012988e-06,
+      "loss": 0.6379,
+      "step": 5229
+    },
+    {
+      "epoch": 0.6809839906286607,
+      "grad_norm": 2.637105703353882,
+      "learning_rate": 8.952179345254912e-06,
+      "loss": 0.7148,
+      "step": 5232
+    },
+    {
+      "epoch": 0.6813744631003514,
+      "grad_norm": 2.6795005798339844,
+      "learning_rate": 8.950898877059417e-06,
+      "loss": 0.6578,
+      "step": 5235
+    },
+    {
+      "epoch": 0.6817649355720422,
+      "grad_norm": 2.5985288619995117,
+      "learning_rate": 8.9496177186502e-06,
+      "loss": 0.6964,
+      "step": 5238
+    },
+    {
+      "epoch": 0.6821554080437329,
+      "grad_norm": 3.457732677459717,
+      "learning_rate": 8.948335870251075e-06,
+      "loss": 0.7052,
+      "step": 5241
+    },
+    {
+      "epoch": 0.6825458805154236,
+      "grad_norm": 2.462507724761963,
+      "learning_rate": 8.947053332085981e-06,
+      "loss": 0.6498,
+      "step": 5244
+    },
+    {
+      "epoch": 0.6829363529871144,
+      "grad_norm": 2.4393460750579834,
+      "learning_rate": 8.945770104378973e-06,
+      "loss": 0.6244,
+      "step": 5247
+    },
+    {
+      "epoch": 0.6833268254588052,
+      "grad_norm": 2.7880349159240723,
+      "learning_rate": 8.944486187354229e-06,
+      "loss": 0.7032,
+      "step": 5250
+    },
+    {
+      "epoch": 0.6837172979304958,
+      "grad_norm": 2.685492753982544,
+      "learning_rate": 8.943201581236045e-06,
+      "loss": 0.6221,
+      "step": 5253
+    },
+    {
+      "epoch": 0.6841077704021866,
+      "grad_norm": 3.0528624057769775,
+      "learning_rate": 8.94191628624884e-06,
+      "loss": 0.6866,
+      "step": 5256
+    },
+    {
+      "epoch": 0.6844982428738774,
+      "grad_norm": 2.485928773880005,
+      "learning_rate": 8.940630302617153e-06,
+      "loss": 0.6147,
+      "step": 5259
+    },
+    {
+      "epoch": 0.6848887153455682,
+      "grad_norm": 2.2878990173339844,
+      "learning_rate": 8.939343630565643e-06,
+      "loss": 0.5663,
+      "step": 5262
+    },
+    {
+      "epoch": 0.6852791878172588,
+      "grad_norm": 3.4641735553741455,
+      "learning_rate": 8.938056270319086e-06,
+      "loss": 0.6713,
+      "step": 5265
+    },
+    {
+      "epoch": 0.6856696602889496,
+      "grad_norm": 2.376103162765503,
+      "learning_rate": 8.936768222102382e-06,
+      "loss": 0.6678,
+      "step": 5268
+    },
+    {
+      "epoch": 0.6860601327606404,
+      "grad_norm": 2.2345285415649414,
+      "learning_rate": 8.935479486140556e-06,
+      "loss": 0.5991,
+      "step": 5271
+    },
+    {
+      "epoch": 0.6864506052323311,
+      "grad_norm": 2.701521873474121,
+      "learning_rate": 8.934190062658738e-06,
+      "loss": 0.687,
+      "step": 5274
+    },
+    {
+      "epoch": 0.6868410777040219,
+      "grad_norm": 2.6253838539123535,
+      "learning_rate": 8.932899951882195e-06,
+      "loss": 0.6385,
+      "step": 5277
+    },
+    {
+      "epoch": 0.6872315501757126,
+      "grad_norm": 2.780701160430908,
+      "learning_rate": 8.931609154036303e-06,
+      "loss": 0.6819,
+      "step": 5280
+    },
+    {
+      "epoch": 0.6876220226474034,
+      "grad_norm": 2.612517833709717,
+      "learning_rate": 8.930317669346565e-06,
+      "loss": 0.5871,
+      "step": 5283
+    },
+    {
+      "epoch": 0.6880124951190941,
+      "grad_norm": 2.6523735523223877,
+      "learning_rate": 8.929025498038595e-06,
+      "loss": 0.6296,
+      "step": 5286
+    },
+    {
+      "epoch": 0.6884029675907849,
+      "grad_norm": 3.8099377155303955,
+      "learning_rate": 8.927732640338138e-06,
+      "loss": 0.7596,
+      "step": 5289
+    },
+    {
+      "epoch": 0.6887934400624756,
+      "grad_norm": 2.856313705444336,
+      "learning_rate": 8.92643909647105e-06,
+      "loss": 0.7508,
+      "step": 5292
+    },
+    {
+      "epoch": 0.6891839125341663,
+      "grad_norm": 3.0642218589782715,
+      "learning_rate": 8.925144866663313e-06,
+      "loss": 0.6367,
+      "step": 5295
+    },
+    {
+      "epoch": 0.6895743850058571,
+      "grad_norm": 2.504683017730713,
+      "learning_rate": 8.923849951141025e-06,
+      "loss": 0.5706,
+      "step": 5298
+    },
+    {
+      "epoch": 0.6899648574775479,
+      "grad_norm": 2.745274543762207,
+      "learning_rate": 8.922554350130404e-06,
+      "loss": 0.6845,
+      "step": 5301
+    },
+    {
+      "epoch": 0.6903553299492385,
+      "grad_norm": 2.6270129680633545,
+      "learning_rate": 8.921258063857792e-06,
+      "loss": 0.7274,
+      "step": 5304
+    },
+    {
+      "epoch": 0.6907458024209293,
+      "grad_norm": 2.979092597961426,
+      "learning_rate": 8.919961092549643e-06,
+      "loss": 0.6765,
+      "step": 5307
+    },
+    {
+      "epoch": 0.6911362748926201,
+      "grad_norm": 3.506718873977661,
+      "learning_rate": 8.91866343643254e-06,
+      "loss": 0.6857,
+      "step": 5310
+    },
+    {
+      "epoch": 0.6915267473643109,
+      "grad_norm": 3.513763189315796,
+      "learning_rate": 8.917365095733176e-06,
+      "loss": 0.6291,
+      "step": 5313
+    },
+    {
+      "epoch": 0.6919172198360015,
+      "grad_norm": 2.3708810806274414,
+      "learning_rate": 8.91606607067837e-06,
+      "loss": 0.5938,
+      "step": 5316
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 2.5921242237091064,
+      "learning_rate": 8.914766361495063e-06,
+      "loss": 0.6986,
+      "step": 5319
+    },
+    {
+      "epoch": 0.6926981647793831,
+      "grad_norm": 2.9242255687713623,
+      "learning_rate": 8.913465968410307e-06,
+      "loss": 0.7117,
+      "step": 5322
+    },
+    {
+      "epoch": 0.6930886372510738,
+      "grad_norm": 2.42401123046875,
+      "learning_rate": 8.912164891651277e-06,
+      "loss": 0.6327,
+      "step": 5325
+    },
+    {
+      "epoch": 0.6934791097227645,
+      "grad_norm": 2.31184458732605,
+      "learning_rate": 8.910863131445273e-06,
+      "loss": 0.6479,
+      "step": 5328
+    },
+    {
+      "epoch": 0.6938695821944553,
+      "grad_norm": 2.514890193939209,
+      "learning_rate": 8.909560688019705e-06,
+      "loss": 0.7545,
+      "step": 5331
+    },
+    {
+      "epoch": 0.694260054666146,
+      "grad_norm": 2.294757127761841,
+      "learning_rate": 8.908257561602112e-06,
+      "loss": 0.6594,
+      "step": 5334
+    },
+    {
+      "epoch": 0.6946505271378368,
+      "grad_norm": 2.3671836853027344,
+      "learning_rate": 8.906953752420142e-06,
+      "loss": 0.6085,
+      "step": 5337
+    },
+    {
+      "epoch": 0.6950409996095275,
+      "grad_norm": 2.5829970836639404,
+      "learning_rate": 8.905649260701571e-06,
+      "loss": 0.7096,
+      "step": 5340
+    },
+    {
+      "epoch": 0.6954314720812182,
+      "grad_norm": 2.5177972316741943,
+      "learning_rate": 8.904344086674292e-06,
+      "loss": 0.6784,
+      "step": 5343
+    },
+    {
+      "epoch": 0.695821944552909,
+      "grad_norm": 2.783193349838257,
+      "learning_rate": 8.903038230566314e-06,
+      "loss": 0.7778,
+      "step": 5346
+    },
+    {
+      "epoch": 0.6962124170245998,
+      "grad_norm": 2.4683210849761963,
+      "learning_rate": 8.901731692605767e-06,
+      "loss": 0.6198,
+      "step": 5349
+    },
+    {
+      "epoch": 0.6966028894962905,
+      "grad_norm": 2.2853734493255615,
+      "learning_rate": 8.900424473020904e-06,
+      "loss": 0.6472,
+      "step": 5352
+    },
+    {
+      "epoch": 0.6969933619679812,
+      "grad_norm": 3.5069949626922607,
+      "learning_rate": 8.899116572040087e-06,
+      "loss": 0.6187,
+      "step": 5355
+    },
+    {
+      "epoch": 0.697383834439672,
+      "grad_norm": 2.6812729835510254,
+      "learning_rate": 8.897807989891809e-06,
+      "loss": 0.6091,
+      "step": 5358
+    },
+    {
+      "epoch": 0.6977743069113628,
+      "grad_norm": 2.5712661743164062,
+      "learning_rate": 8.896498726804677e-06,
+      "loss": 0.6686,
+      "step": 5361
+    },
+    {
+      "epoch": 0.6981647793830535,
+      "grad_norm": 3.0275492668151855,
+      "learning_rate": 8.895188783007412e-06,
+      "loss": 0.6555,
+      "step": 5364
+    },
+    {
+      "epoch": 0.6985552518547442,
+      "grad_norm": 2.5897321701049805,
+      "learning_rate": 8.893878158728861e-06,
+      "loss": 0.6643,
+      "step": 5367
+    },
+    {
+      "epoch": 0.698945724326435,
+      "grad_norm": 2.7661631107330322,
+      "learning_rate": 8.892566854197988e-06,
+      "loss": 0.6861,
+      "step": 5370
+    },
+    {
+      "epoch": 0.6993361967981258,
+      "grad_norm": 2.3814573287963867,
+      "learning_rate": 8.891254869643873e-06,
+      "loss": 0.6581,
+      "step": 5373
+    },
+    {
+      "epoch": 0.6997266692698165,
+      "grad_norm": 2.452768564224243,
+      "learning_rate": 8.88994220529572e-06,
+      "loss": 0.6112,
+      "step": 5376
+    },
+    {
+      "epoch": 0.7001171417415072,
+      "grad_norm": 2.68501615524292,
+      "learning_rate": 8.888628861382846e-06,
+      "loss": 0.604,
+      "step": 5379
+    },
+    {
+      "epoch": 0.700507614213198,
+      "grad_norm": 2.675839424133301,
+      "learning_rate": 8.88731483813469e-06,
+      "loss": 0.7262,
+      "step": 5382
+    },
+    {
+      "epoch": 0.7008980866848887,
+      "grad_norm": 4.175928592681885,
+      "learning_rate": 8.88600013578081e-06,
+      "loss": 0.6624,
+      "step": 5385
+    },
+    {
+      "epoch": 0.7012885591565795,
+      "grad_norm": 2.8697991371154785,
+      "learning_rate": 8.884684754550882e-06,
+      "loss": 0.6547,
+      "step": 5388
+    },
+    {
+      "epoch": 0.7016790316282702,
+      "grad_norm": 4.71966552734375,
+      "learning_rate": 8.8833686946747e-06,
+      "loss": 0.591,
+      "step": 5391
+    },
+    {
+      "epoch": 0.7020695040999609,
+      "grad_norm": 2.5646259784698486,
+      "learning_rate": 8.882051956382175e-06,
+      "loss": 0.7158,
+      "step": 5394
+    },
+    {
+      "epoch": 0.7024599765716517,
+      "grad_norm": 2.482074737548828,
+      "learning_rate": 8.88073453990334e-06,
+      "loss": 0.6647,
+      "step": 5397
+    },
+    {
+      "epoch": 0.7028504490433425,
+      "grad_norm": 2.8584282398223877,
+      "learning_rate": 8.879416445468344e-06,
+      "loss": 0.5627,
+      "step": 5400
+    },
+    {
+      "epoch": 0.7032409215150331,
+      "grad_norm": 2.51839280128479,
+      "learning_rate": 8.878097673307458e-06,
+      "loss": 0.7744,
+      "step": 5403
+    },
+    {
+      "epoch": 0.7036313939867239,
+      "grad_norm": 2.4203269481658936,
+      "learning_rate": 8.876778223651067e-06,
+      "loss": 0.5973,
+      "step": 5406
+    },
+    {
+      "epoch": 0.7040218664584147,
+      "grad_norm": 2.3395378589630127,
+      "learning_rate": 8.875458096729674e-06,
+      "loss": 0.5872,
+      "step": 5409
+    },
+    {
+      "epoch": 0.7044123389301055,
+      "grad_norm": 2.6769402027130127,
+      "learning_rate": 8.874137292773906e-06,
+      "loss": 0.6168,
+      "step": 5412
+    },
+    {
+      "epoch": 0.7048028114017961,
+      "grad_norm": 2.5752854347229004,
+      "learning_rate": 8.872815812014501e-06,
+      "loss": 0.7544,
+      "step": 5415
+    },
+    {
+      "epoch": 0.7051932838734869,
+      "grad_norm": 2.6731069087982178,
+      "learning_rate": 8.871493654682321e-06,
+      "loss": 0.6455,
+      "step": 5418
+    },
+    {
+      "epoch": 0.7055837563451777,
+      "grad_norm": 2.6608500480651855,
+      "learning_rate": 8.870170821008346e-06,
+      "loss": 0.6383,
+      "step": 5421
+    },
+    {
+      "epoch": 0.7059742288168684,
+      "grad_norm": 3.0593783855438232,
+      "learning_rate": 8.86884731122367e-06,
+      "loss": 0.6723,
+      "step": 5424
+    },
+    {
+      "epoch": 0.7063647012885591,
+      "grad_norm": 2.809638738632202,
+      "learning_rate": 8.867523125559504e-06,
+      "loss": 0.6422,
+      "step": 5427
+    },
+    {
+      "epoch": 0.7067551737602499,
+      "grad_norm": 2.5025877952575684,
+      "learning_rate": 8.866198264247187e-06,
+      "loss": 0.687,
+      "step": 5430
+    },
+    {
+      "epoch": 0.7071456462319406,
+      "grad_norm": 3.1199793815612793,
+      "learning_rate": 8.864872727518168e-06,
+      "loss": 0.5947,
+      "step": 5433
+    },
+    {
+      "epoch": 0.7075361187036314,
+      "grad_norm": 2.5214881896972656,
+      "learning_rate": 8.863546515604012e-06,
+      "loss": 0.6269,
+      "step": 5436
+    },
+    {
+      "epoch": 0.7079265911753222,
+      "grad_norm": 2.481205940246582,
+      "learning_rate": 8.862219628736409e-06,
+      "loss": 0.7641,
+      "step": 5439
+    },
+    {
+      "epoch": 0.7083170636470129,
+      "grad_norm": 2.585766553878784,
+      "learning_rate": 8.86089206714716e-06,
+      "loss": 0.698,
+      "step": 5442
+    },
+    {
+      "epoch": 0.7087075361187036,
+      "grad_norm": 3.276780843734741,
+      "learning_rate": 8.859563831068188e-06,
+      "loss": 0.6059,
+      "step": 5445
+    },
+    {
+      "epoch": 0.7090980085903944,
+      "grad_norm": 2.3243212699890137,
+      "learning_rate": 8.858234920731536e-06,
+      "loss": 0.6083,
+      "step": 5448
+    },
+    {
+      "epoch": 0.7094884810620852,
+      "grad_norm": 2.7607672214508057,
+      "learning_rate": 8.856905336369359e-06,
+      "loss": 0.7022,
+      "step": 5451
+    },
+    {
+      "epoch": 0.7098789535337758,
+      "grad_norm": 2.5045323371887207,
+      "learning_rate": 8.855575078213933e-06,
+      "loss": 0.5898,
+      "step": 5454
+    },
+    {
+      "epoch": 0.7102694260054666,
+      "grad_norm": 2.378309488296509,
+      "learning_rate": 8.854244146497654e-06,
+      "loss": 0.6493,
+      "step": 5457
+    },
+    {
+      "epoch": 0.7106598984771574,
+      "grad_norm": 3.019075870513916,
+      "learning_rate": 8.852912541453029e-06,
+      "loss": 0.5836,
+      "step": 5460
+    },
+    {
+      "epoch": 0.7110503709488482,
+      "grad_norm": 2.7453722953796387,
+      "learning_rate": 8.85158026331269e-06,
+      "loss": 0.59,
+      "step": 5463
+    },
+    {
+      "epoch": 0.7114408434205388,
+      "grad_norm": 2.3514020442962646,
+      "learning_rate": 8.85024731230938e-06,
+      "loss": 0.5568,
+      "step": 5466
+    },
+    {
+      "epoch": 0.7118313158922296,
+      "grad_norm": 2.4078609943389893,
+      "learning_rate": 8.848913688675963e-06,
+      "loss": 0.694,
+      "step": 5469
+    },
+    {
+      "epoch": 0.7122217883639204,
+      "grad_norm": 2.3529157638549805,
+      "learning_rate": 8.847579392645425e-06,
+      "loss": 0.6763,
+      "step": 5472
+    },
+    {
+      "epoch": 0.7126122608356111,
+      "grad_norm": 2.9866087436676025,
+      "learning_rate": 8.846244424450858e-06,
+      "loss": 0.6423,
+      "step": 5475
+    },
+    {
+      "epoch": 0.7130027333073018,
+      "grad_norm": 2.784130334854126,
+      "learning_rate": 8.844908784325483e-06,
+      "loss": 0.6326,
+      "step": 5478
+    },
+    {
+      "epoch": 0.7133932057789926,
+      "grad_norm": 2.18365740776062,
+      "learning_rate": 8.84357247250263e-06,
+      "loss": 0.5958,
+      "step": 5481
+    },
+    {
+      "epoch": 0.7137836782506833,
+      "grad_norm": 2.3948042392730713,
+      "learning_rate": 8.842235489215755e-06,
+      "loss": 0.7437,
+      "step": 5484
+    },
+    {
+      "epoch": 0.7141741507223741,
+      "grad_norm": 2.9066920280456543,
+      "learning_rate": 8.84089783469842e-06,
+      "loss": 0.6481,
+      "step": 5487
+    },
+    {
+      "epoch": 0.7145646231940648,
+      "grad_norm": 2.770456075668335,
+      "learning_rate": 8.839559509184317e-06,
+      "loss": 0.6289,
+      "step": 5490
+    },
+    {
+      "epoch": 0.7149550956657555,
+      "grad_norm": 2.8254899978637695,
+      "learning_rate": 8.838220512907241e-06,
+      "loss": 0.6341,
+      "step": 5493
+    },
+    {
+      "epoch": 0.7153455681374463,
+      "grad_norm": 2.9626049995422363,
+      "learning_rate": 8.836880846101118e-06,
+      "loss": 0.6394,
+      "step": 5496
+    },
+    {
+      "epoch": 0.7157360406091371,
+      "grad_norm": 2.709442377090454,
+      "learning_rate": 8.835540508999982e-06,
+      "loss": 0.7887,
+      "step": 5499
+    },
+    {
+      "epoch": 0.7161265130808278,
+      "grad_norm": 2.8093841075897217,
+      "learning_rate": 8.834199501837988e-06,
+      "loss": 0.5902,
+      "step": 5502
+    },
+    {
+      "epoch": 0.7165169855525185,
+      "grad_norm": 2.7774314880371094,
+      "learning_rate": 8.832857824849407e-06,
+      "loss": 0.6338,
+      "step": 5505
+    },
+    {
+      "epoch": 0.7169074580242093,
+      "grad_norm": 4.112159729003906,
+      "learning_rate": 8.831515478268627e-06,
+      "loss": 0.6495,
+      "step": 5508
+    },
+    {
+      "epoch": 0.7172979304959001,
+      "grad_norm": 3.2125117778778076,
+      "learning_rate": 8.830172462330155e-06,
+      "loss": 0.639,
+      "step": 5511
+    },
+    {
+      "epoch": 0.7176884029675907,
+      "grad_norm": 2.521543264389038,
+      "learning_rate": 8.828828777268609e-06,
+      "loss": 0.6834,
+      "step": 5514
+    },
+    {
+      "epoch": 0.7180788754392815,
+      "grad_norm": 2.344866991043091,
+      "learning_rate": 8.827484423318731e-06,
+      "loss": 0.7033,
+      "step": 5517
+    },
+    {
+      "epoch": 0.7184693479109723,
+      "grad_norm": 3.047065496444702,
+      "learning_rate": 8.826139400715377e-06,
+      "loss": 0.6674,
+      "step": 5520
+    },
+    {
+      "epoch": 0.718859820382663,
+      "grad_norm": 2.916560411453247,
+      "learning_rate": 8.824793709693517e-06,
+      "loss": 0.6385,
+      "step": 5523
+    },
+    {
+      "epoch": 0.7192502928543538,
+      "grad_norm": 2.3283605575561523,
+      "learning_rate": 8.823447350488243e-06,
+      "loss": 0.6017,
+      "step": 5526
+    },
+    {
+      "epoch": 0.7196407653260445,
+      "grad_norm": 3.2105703353881836,
+      "learning_rate": 8.822100323334761e-06,
+      "loss": 0.635,
+      "step": 5529
+    },
+    {
+      "epoch": 0.7200312377977353,
+      "grad_norm": 2.4558160305023193,
+      "learning_rate": 8.820752628468391e-06,
+      "loss": 0.6822,
+      "step": 5532
+    },
+    {
+      "epoch": 0.720421710269426,
+      "grad_norm": 2.4689717292785645,
+      "learning_rate": 8.819404266124575e-06,
+      "loss": 0.7866,
+      "step": 5535
+    },
+    {
+      "epoch": 0.7208121827411168,
+      "grad_norm": 2.5155982971191406,
+      "learning_rate": 8.818055236538872e-06,
+      "loss": 0.6903,
+      "step": 5538
+    },
+    {
+      "epoch": 0.7212026552128075,
+      "grad_norm": 3.5505385398864746,
+      "learning_rate": 8.816705539946948e-06,
+      "loss": 0.5982,
+      "step": 5541
+    },
+    {
+      "epoch": 0.7215931276844982,
+      "grad_norm": 2.4933745861053467,
+      "learning_rate": 8.815355176584595e-06,
+      "loss": 0.6077,
+      "step": 5544
+    },
+    {
+      "epoch": 0.721983600156189,
+      "grad_norm": 3.3649566173553467,
+      "learning_rate": 8.81400414668772e-06,
+      "loss": 0.7257,
+      "step": 5547
+    },
+    {
+      "epoch": 0.7223740726278798,
+      "grad_norm": 2.446016550064087,
+      "learning_rate": 8.812652450492345e-06,
+      "loss": 0.5711,
+      "step": 5550
+    },
+    {
+      "epoch": 0.7227645450995704,
+      "grad_norm": 2.715980052947998,
+      "learning_rate": 8.811300088234607e-06,
+      "loss": 0.6866,
+      "step": 5553
+    },
+    {
+      "epoch": 0.7231550175712612,
+      "grad_norm": 2.724425792694092,
+      "learning_rate": 8.80994706015076e-06,
+      "loss": 0.6597,
+      "step": 5556
+    },
+    {
+      "epoch": 0.723545490042952,
+      "grad_norm": 3.384270429611206,
+      "learning_rate": 8.808593366477177e-06,
+      "loss": 0.6568,
+      "step": 5559
+    },
+    {
+      "epoch": 0.7239359625146428,
+      "grad_norm": 2.415548086166382,
+      "learning_rate": 8.807239007450345e-06,
+      "loss": 0.6081,
+      "step": 5562
+    },
+    {
+      "epoch": 0.7243264349863334,
+      "grad_norm": 2.3857581615448,
+      "learning_rate": 8.805883983306869e-06,
+      "loss": 0.7436,
+      "step": 5565
+    },
+    {
+      "epoch": 0.7247169074580242,
+      "grad_norm": 2.494887351989746,
+      "learning_rate": 8.804528294283466e-06,
+      "loss": 0.6451,
+      "step": 5568
+    },
+    {
+      "epoch": 0.725107379929715,
+      "grad_norm": 2.5483858585357666,
+      "learning_rate": 8.803171940616974e-06,
+      "loss": 0.6226,
+      "step": 5571
+    },
+    {
+      "epoch": 0.7254978524014057,
+      "grad_norm": 2.5574679374694824,
+      "learning_rate": 8.801814922544345e-06,
+      "loss": 0.6289,
+      "step": 5574
+    },
+    {
+      "epoch": 0.7258883248730964,
+      "grad_norm": 2.4144797325134277,
+      "learning_rate": 8.800457240302646e-06,
+      "loss": 0.6128,
+      "step": 5577
+    },
+    {
+      "epoch": 0.7262787973447872,
+      "grad_norm": 2.4427857398986816,
+      "learning_rate": 8.799098894129063e-06,
+      "loss": 0.6263,
+      "step": 5580
+    },
+    {
+      "epoch": 0.726669269816478,
+      "grad_norm": 2.785493850708008,
+      "learning_rate": 8.797739884260896e-06,
+      "loss": 0.7186,
+      "step": 5583
+    },
+    {
+      "epoch": 0.7270597422881687,
+      "grad_norm": 2.493211030960083,
+      "learning_rate": 8.79638021093556e-06,
+      "loss": 0.6333,
+      "step": 5586
+    },
+    {
+      "epoch": 0.7274502147598594,
+      "grad_norm": 2.7701425552368164,
+      "learning_rate": 8.795019874390587e-06,
+      "loss": 0.7146,
+      "step": 5589
+    },
+    {
+      "epoch": 0.7278406872315502,
+      "grad_norm": 2.20003604888916,
+      "learning_rate": 8.793658874863626e-06,
+      "loss": 0.5846,
+      "step": 5592
+    },
+    {
+      "epoch": 0.7282311597032409,
+      "grad_norm": 3.9232680797576904,
+      "learning_rate": 8.79229721259244e-06,
+      "loss": 0.6657,
+      "step": 5595
+    },
+    {
+      "epoch": 0.7286216321749317,
+      "grad_norm": 2.5265204906463623,
+      "learning_rate": 8.790934887814908e-06,
+      "loss": 0.6148,
+      "step": 5598
+    },
+    {
+      "epoch": 0.7290121046466224,
+      "grad_norm": 2.7512199878692627,
+      "learning_rate": 8.789571900769028e-06,
+      "loss": 0.6684,
+      "step": 5601
+    },
+    {
+      "epoch": 0.7294025771183131,
+      "grad_norm": 2.3781023025512695,
+      "learning_rate": 8.788208251692908e-06,
+      "loss": 0.7028,
+      "step": 5604
+    },
+    {
+      "epoch": 0.7297930495900039,
+      "grad_norm": 2.4492669105529785,
+      "learning_rate": 8.786843940824775e-06,
+      "loss": 0.5972,
+      "step": 5607
+    },
+    {
+      "epoch": 0.7301835220616947,
+      "grad_norm": 2.4138023853302,
+      "learning_rate": 8.785478968402972e-06,
+      "loss": 0.5792,
+      "step": 5610
+    },
+    {
+      "epoch": 0.7305739945333855,
+      "grad_norm": 3.772378921508789,
+      "learning_rate": 8.784113334665958e-06,
+      "loss": 0.7113,
+      "step": 5613
+    },
+    {
+      "epoch": 0.7309644670050761,
+      "grad_norm": 2.438852071762085,
+      "learning_rate": 8.782747039852304e-06,
+      "loss": 0.6087,
+      "step": 5616
+    },
+    {
+      "epoch": 0.7313549394767669,
+      "grad_norm": 2.447324514389038,
+      "learning_rate": 8.7813800842007e-06,
+      "loss": 0.6334,
+      "step": 5619
+    },
+    {
+      "epoch": 0.7317454119484577,
+      "grad_norm": 2.2790205478668213,
+      "learning_rate": 8.78001246794995e-06,
+      "loss": 0.6197,
+      "step": 5622
+    },
+    {
+      "epoch": 0.7321358844201484,
+      "grad_norm": 2.459559917449951,
+      "learning_rate": 8.778644191338974e-06,
+      "loss": 0.6726,
+      "step": 5625
+    },
+    {
+      "epoch": 0.7325263568918391,
+      "grad_norm": 2.6263234615325928,
+      "learning_rate": 8.777275254606808e-06,
+      "loss": 0.6843,
+      "step": 5628
+    },
+    {
+      "epoch": 0.7329168293635299,
+      "grad_norm": 2.590893507003784,
+      "learning_rate": 8.775905657992599e-06,
+      "loss": 0.7584,
+      "step": 5631
+    },
+    {
+      "epoch": 0.7333073018352206,
+      "grad_norm": 2.9429869651794434,
+      "learning_rate": 8.774535401735616e-06,
+      "loss": 0.687,
+      "step": 5634
+    },
+    {
+      "epoch": 0.7336977743069114,
+      "grad_norm": 2.8409743309020996,
+      "learning_rate": 8.773164486075238e-06,
+      "loss": 0.7379,
+      "step": 5637
+    },
+    {
+      "epoch": 0.7340882467786021,
+      "grad_norm": 2.462329864501953,
+      "learning_rate": 8.771792911250963e-06,
+      "loss": 0.6646,
+      "step": 5640
+    },
+    {
+      "epoch": 0.7344787192502928,
+      "grad_norm": 2.554640769958496,
+      "learning_rate": 8.770420677502401e-06,
+      "loss": 0.7051,
+      "step": 5643
+    },
+    {
+      "epoch": 0.7348691917219836,
+      "grad_norm": 2.4346702098846436,
+      "learning_rate": 8.769047785069277e-06,
+      "loss": 0.6723,
+      "step": 5646
+    },
+    {
+      "epoch": 0.7352596641936744,
+      "grad_norm": 2.7797698974609375,
+      "learning_rate": 8.767674234191436e-06,
+      "loss": 0.6458,
+      "step": 5649
+    },
+    {
+      "epoch": 0.735650136665365,
+      "grad_norm": 2.8324296474456787,
+      "learning_rate": 8.76630002510883e-06,
+      "loss": 0.7206,
+      "step": 5652
+    },
+    {
+      "epoch": 0.7360406091370558,
+      "grad_norm": 2.280325412750244,
+      "learning_rate": 8.764925158061537e-06,
+      "loss": 0.6426,
+      "step": 5655
+    },
+    {
+      "epoch": 0.7364310816087466,
+      "grad_norm": 3.168109655380249,
+      "learning_rate": 8.763549633289737e-06,
+      "loss": 0.6371,
+      "step": 5658
+    },
+    {
+      "epoch": 0.7368215540804374,
+      "grad_norm": 3.2661867141723633,
+      "learning_rate": 8.762173451033731e-06,
+      "loss": 0.6164,
+      "step": 5661
+    },
+    {
+      "epoch": 0.737212026552128,
+      "grad_norm": 4.669281959533691,
+      "learning_rate": 8.760796611533939e-06,
+      "loss": 0.6704,
+      "step": 5664
+    },
+    {
+      "epoch": 0.7376024990238188,
+      "grad_norm": 2.702472686767578,
+      "learning_rate": 8.75941911503089e-06,
+      "loss": 0.6053,
+      "step": 5667
+    },
+    {
+      "epoch": 0.7379929714955096,
+      "grad_norm": 2.3921306133270264,
+      "learning_rate": 8.758040961765233e-06,
+      "loss": 0.6229,
+      "step": 5670
+    },
+    {
+      "epoch": 0.7383834439672003,
+      "grad_norm": 2.605787515640259,
+      "learning_rate": 8.756662151977724e-06,
+      "loss": 0.7242,
+      "step": 5673
+    },
+    {
+      "epoch": 0.738773916438891,
+      "grad_norm": 2.5881075859069824,
+      "learning_rate": 8.755282685909239e-06,
+      "loss": 0.6878,
+      "step": 5676
+    },
+    {
+      "epoch": 0.7391643889105818,
+      "grad_norm": 2.3970558643341064,
+      "learning_rate": 8.753902563800769e-06,
+      "loss": 0.6808,
+      "step": 5679
+    },
+    {
+      "epoch": 0.7395548613822726,
+      "grad_norm": 2.6305997371673584,
+      "learning_rate": 8.752521785893418e-06,
+      "loss": 0.6692,
+      "step": 5682
+    },
+    {
+      "epoch": 0.7399453338539633,
+      "grad_norm": 3.6381564140319824,
+      "learning_rate": 8.751140352428406e-06,
+      "loss": 0.6898,
+      "step": 5685
+    },
+    {
+      "epoch": 0.740335806325654,
+      "grad_norm": 2.4845845699310303,
+      "learning_rate": 8.749758263647066e-06,
+      "loss": 0.6876,
+      "step": 5688
+    },
+    {
+      "epoch": 0.7407262787973448,
+      "grad_norm": 2.307579517364502,
+      "learning_rate": 8.748375519790846e-06,
+      "loss": 0.5223,
+      "step": 5691
+    },
+    {
+      "epoch": 0.7411167512690355,
+      "grad_norm": 2.4577105045318604,
+      "learning_rate": 8.746992121101306e-06,
+      "loss": 0.6736,
+      "step": 5694
+    },
+    {
+      "epoch": 0.7415072237407263,
+      "grad_norm": 2.539513111114502,
+      "learning_rate": 8.745608067820127e-06,
+      "loss": 0.6554,
+      "step": 5697
+    },
+    {
+      "epoch": 0.7418976962124171,
+      "grad_norm": 2.9040579795837402,
+      "learning_rate": 8.744223360189097e-06,
+      "loss": 0.6961,
+      "step": 5700
+    },
+    {
+      "epoch": 0.7422881686841077,
+      "grad_norm": 3.254425525665283,
+      "learning_rate": 8.742837998450122e-06,
+      "loss": 0.6879,
+      "step": 5703
+    },
+    {
+      "epoch": 0.7426786411557985,
+      "grad_norm": 3.117102861404419,
+      "learning_rate": 8.741451982845224e-06,
+      "loss": 0.6168,
+      "step": 5706
+    },
+    {
+      "epoch": 0.7430691136274893,
+      "grad_norm": 2.4279301166534424,
+      "learning_rate": 8.740065313616536e-06,
+      "loss": 0.6354,
+      "step": 5709
+    },
+    {
+      "epoch": 0.7434595860991801,
+      "grad_norm": 2.2769625186920166,
+      "learning_rate": 8.738677991006304e-06,
+      "loss": 0.6411,
+      "step": 5712
+    },
+    {
+      "epoch": 0.7438500585708707,
+      "grad_norm": 2.597215175628662,
+      "learning_rate": 8.737290015256892e-06,
+      "loss": 0.6494,
+      "step": 5715
+    },
+    {
+      "epoch": 0.7442405310425615,
+      "grad_norm": 3.5001959800720215,
+      "learning_rate": 8.735901386610777e-06,
+      "loss": 0.6644,
+      "step": 5718
+    },
+    {
+      "epoch": 0.7446310035142523,
+      "grad_norm": 2.5007054805755615,
+      "learning_rate": 8.73451210531055e-06,
+      "loss": 0.643,
+      "step": 5721
+    },
+    {
+      "epoch": 0.745021475985943,
+      "grad_norm": 2.652787923812866,
+      "learning_rate": 8.733122171598914e-06,
+      "loss": 0.7299,
+      "step": 5724
+    },
+    {
+      "epoch": 0.7454119484576337,
+      "grad_norm": 2.6559813022613525,
+      "learning_rate": 8.731731585718687e-06,
+      "loss": 0.6918,
+      "step": 5727
+    },
+    {
+      "epoch": 0.7458024209293245,
+      "grad_norm": 2.646437644958496,
+      "learning_rate": 8.730340347912803e-06,
+      "loss": 0.6691,
+      "step": 5730
+    },
+    {
+      "epoch": 0.7461928934010152,
+      "grad_norm": 2.6184146404266357,
+      "learning_rate": 8.728948458424307e-06,
+      "loss": 0.6599,
+      "step": 5733
+    },
+    {
+      "epoch": 0.746583365872706,
+      "grad_norm": 2.406266927719116,
+      "learning_rate": 8.72755591749636e-06,
+      "loss": 0.6671,
+      "step": 5736
+    },
+    {
+      "epoch": 0.7469738383443967,
+      "grad_norm": 3.493818759918213,
+      "learning_rate": 8.726162725372237e-06,
+      "loss": 0.6482,
+      "step": 5739
+    },
+    {
+      "epoch": 0.7473643108160875,
+      "grad_norm": 2.1511340141296387,
+      "learning_rate": 8.724768882295324e-06,
+      "loss": 0.6,
+      "step": 5742
+    },
+    {
+      "epoch": 0.7477547832877782,
+      "grad_norm": 2.556877851486206,
+      "learning_rate": 8.723374388509123e-06,
+      "loss": 0.6613,
+      "step": 5745
+    },
+    {
+      "epoch": 0.748145255759469,
+      "grad_norm": 2.497506856918335,
+      "learning_rate": 8.721979244257247e-06,
+      "loss": 0.6198,
+      "step": 5748
+    },
+    {
+      "epoch": 0.7485357282311597,
+      "grad_norm": 2.378761053085327,
+      "learning_rate": 8.72058344978343e-06,
+      "loss": 0.5909,
+      "step": 5751
+    },
+    {
+      "epoch": 0.7489262007028504,
+      "grad_norm": 3.043630838394165,
+      "learning_rate": 8.71918700533151e-06,
+      "loss": 0.6554,
+      "step": 5754
+    },
+    {
+      "epoch": 0.7493166731745412,
+      "grad_norm": 2.5283005237579346,
+      "learning_rate": 8.717789911145445e-06,
+      "loss": 0.6827,
+      "step": 5757
+    },
+    {
+      "epoch": 0.749707145646232,
+      "grad_norm": 2.631117105484009,
+      "learning_rate": 8.716392167469303e-06,
+      "loss": 0.6304,
+      "step": 5760
+    },
+    {
+      "epoch": 0.7500976181179226,
+      "grad_norm": 2.4751923084259033,
+      "learning_rate": 8.714993774547267e-06,
+      "loss": 0.7159,
+      "step": 5763
+    },
+    {
+      "epoch": 0.7504880905896134,
+      "grad_norm": 3.3558480739593506,
+      "learning_rate": 8.713594732623635e-06,
+      "loss": 0.7029,
+      "step": 5766
+    },
+    {
+      "epoch": 0.7508785630613042,
+      "grad_norm": 2.266209125518799,
+      "learning_rate": 8.712195041942814e-06,
+      "loss": 0.7196,
+      "step": 5769
+    },
+    {
+      "epoch": 0.751269035532995,
+      "grad_norm": 2.9685750007629395,
+      "learning_rate": 8.71079470274933e-06,
+      "loss": 0.6283,
+      "step": 5772
+    },
+    {
+      "epoch": 0.7516595080046856,
+      "grad_norm": 2.5255727767944336,
+      "learning_rate": 8.709393715287817e-06,
+      "loss": 0.5652,
+      "step": 5775
+    },
+    {
+      "epoch": 0.7520499804763764,
+      "grad_norm": 2.282809257507324,
+      "learning_rate": 8.707992079803025e-06,
+      "loss": 0.5898,
+      "step": 5778
+    },
+    {
+      "epoch": 0.7524404529480672,
+      "grad_norm": 2.4005398750305176,
+      "learning_rate": 8.706589796539818e-06,
+      "loss": 0.7173,
+      "step": 5781
+    },
+    {
+      "epoch": 0.7528309254197579,
+      "grad_norm": 3.0054335594177246,
+      "learning_rate": 8.70518686574317e-06,
+      "loss": 0.6631,
+      "step": 5784
+    },
+    {
+      "epoch": 0.7532213978914487,
+      "grad_norm": 2.390732526779175,
+      "learning_rate": 8.703783287658172e-06,
+      "loss": 0.7253,
+      "step": 5787
+    },
+    {
+      "epoch": 0.7536118703631394,
+      "grad_norm": 2.417182445526123,
+      "learning_rate": 8.702379062530026e-06,
+      "loss": 0.6376,
+      "step": 5790
+    },
+    {
+      "epoch": 0.7540023428348301,
+      "grad_norm": 2.577763795852661,
+      "learning_rate": 8.700974190604045e-06,
+      "loss": 0.621,
+      "step": 5793
+    },
+    {
+      "epoch": 0.7543928153065209,
+      "grad_norm": 2.8259828090667725,
+      "learning_rate": 8.69956867212566e-06,
+      "loss": 0.624,
+      "step": 5796
+    },
+    {
+      "epoch": 0.7547832877782117,
+      "grad_norm": 3.2486746311187744,
+      "learning_rate": 8.698162507340408e-06,
+      "loss": 0.6333,
+      "step": 5799
+    },
+    {
+      "epoch": 0.7551737602499023,
+      "grad_norm": 2.8928191661834717,
+      "learning_rate": 8.696755696493949e-06,
+      "loss": 0.6203,
+      "step": 5802
+    },
+    {
+      "epoch": 0.7555642327215931,
+      "grad_norm": 3.078284502029419,
+      "learning_rate": 8.695348239832045e-06,
+      "loss": 0.6175,
+      "step": 5805
+    },
+    {
+      "epoch": 0.7559547051932839,
+      "grad_norm": 2.5001373291015625,
+      "learning_rate": 8.693940137600578e-06,
+      "loss": 0.639,
+      "step": 5808
+    },
+    {
+      "epoch": 0.7563451776649747,
+      "grad_norm": 2.3842241764068604,
+      "learning_rate": 8.692531390045538e-06,
+      "loss": 0.6531,
+      "step": 5811
+    },
+    {
+      "epoch": 0.7567356501366653,
+      "grad_norm": 2.515578269958496,
+      "learning_rate": 8.691121997413035e-06,
+      "loss": 0.6603,
+      "step": 5814
+    },
+    {
+      "epoch": 0.7571261226083561,
+      "grad_norm": 2.7876977920532227,
+      "learning_rate": 8.689711959949282e-06,
+      "loss": 0.6454,
+      "step": 5817
+    },
+    {
+      "epoch": 0.7575165950800469,
+      "grad_norm": 3.4954264163970947,
+      "learning_rate": 8.688301277900613e-06,
+      "loss": 0.7285,
+      "step": 5820
+    },
+    {
+      "epoch": 0.7579070675517376,
+      "grad_norm": 2.9758989810943604,
+      "learning_rate": 8.686889951513468e-06,
+      "loss": 0.6519,
+      "step": 5823
+    },
+    {
+      "epoch": 0.7582975400234283,
+      "grad_norm": 2.79794979095459,
+      "learning_rate": 8.685477981034407e-06,
+      "loss": 0.6698,
+      "step": 5826
+    },
+    {
+      "epoch": 0.7586880124951191,
+      "grad_norm": 2.7669429779052734,
+      "learning_rate": 8.684065366710093e-06,
+      "loss": 0.7602,
+      "step": 5829
+    },
+    {
+      "epoch": 0.7590784849668099,
+      "grad_norm": 2.231877326965332,
+      "learning_rate": 8.682652108787312e-06,
+      "loss": 0.6112,
+      "step": 5832
+    },
+    {
+      "epoch": 0.7594689574385006,
+      "grad_norm": 2.530808210372925,
+      "learning_rate": 8.681238207512955e-06,
+      "loss": 0.5664,
+      "step": 5835
+    },
+    {
+      "epoch": 0.7598594299101913,
+      "grad_norm": 3.670186758041382,
+      "learning_rate": 8.679823663134025e-06,
+      "loss": 0.5716,
+      "step": 5838
+    },
+    {
+      "epoch": 0.7602499023818821,
+      "grad_norm": 2.7073922157287598,
+      "learning_rate": 8.678408475897643e-06,
+      "loss": 0.556,
+      "step": 5841
+    },
+    {
+      "epoch": 0.7606403748535728,
+      "grad_norm": 2.874410390853882,
+      "learning_rate": 8.67699264605104e-06,
+      "loss": 0.6784,
+      "step": 5844
+    },
+    {
+      "epoch": 0.7610308473252636,
+      "grad_norm": 2.438572883605957,
+      "learning_rate": 8.675576173841555e-06,
+      "loss": 0.7022,
+      "step": 5847
+    },
+    {
+      "epoch": 0.7614213197969543,
+      "grad_norm": 2.563101053237915,
+      "learning_rate": 8.674159059516645e-06,
+      "loss": 0.7107,
+      "step": 5850
+    },
+    {
+      "epoch": 0.761811792268645,
+      "grad_norm": 2.388939380645752,
+      "learning_rate": 8.672741303323877e-06,
+      "loss": 0.6116,
+      "step": 5853
+    },
+    {
+      "epoch": 0.7622022647403358,
+      "grad_norm": 2.337643623352051,
+      "learning_rate": 8.671322905510931e-06,
+      "loss": 0.6408,
+      "step": 5856
+    },
+    {
+      "epoch": 0.7625927372120266,
+      "grad_norm": 2.49255633354187,
+      "learning_rate": 8.669903866325594e-06,
+      "loss": 0.6909,
+      "step": 5859
+    },
+    {
+      "epoch": 0.7629832096837172,
+      "grad_norm": 2.3588013648986816,
+      "learning_rate": 8.668484186015775e-06,
+      "loss": 0.6309,
+      "step": 5862
+    },
+    {
+      "epoch": 0.763373682155408,
+      "grad_norm": 2.3805925846099854,
+      "learning_rate": 8.667063864829483e-06,
+      "loss": 0.6424,
+      "step": 5865
+    },
+    {
+      "epoch": 0.7637641546270988,
+      "grad_norm": 3.148681879043579,
+      "learning_rate": 8.665642903014851e-06,
+      "loss": 0.6926,
+      "step": 5868
+    },
+    {
+      "epoch": 0.7641546270987896,
+      "grad_norm": 2.75685977935791,
+      "learning_rate": 8.664221300820114e-06,
+      "loss": 0.6597,
+      "step": 5871
+    },
+    {
+      "epoch": 0.7645450995704803,
+      "grad_norm": 2.3916969299316406,
+      "learning_rate": 8.662799058493625e-06,
+      "loss": 0.5482,
+      "step": 5874
+    },
+    {
+      "epoch": 0.764935572042171,
+      "grad_norm": 2.5081396102905273,
+      "learning_rate": 8.661376176283844e-06,
+      "loss": 0.6723,
+      "step": 5877
+    },
+    {
+      "epoch": 0.7653260445138618,
+      "grad_norm": 2.1820735931396484,
+      "learning_rate": 8.659952654439348e-06,
+      "loss": 0.6866,
+      "step": 5880
+    },
+    {
+      "epoch": 0.7657165169855525,
+      "grad_norm": 2.611650228500366,
+      "learning_rate": 8.658528493208825e-06,
+      "loss": 0.7078,
+      "step": 5883
+    },
+    {
+      "epoch": 0.7661069894572433,
+      "grad_norm": 2.291048765182495,
+      "learning_rate": 8.657103692841067e-06,
+      "loss": 0.6261,
+      "step": 5886
+    },
+    {
+      "epoch": 0.766497461928934,
+      "grad_norm": 2.417051076889038,
+      "learning_rate": 8.655678253584989e-06,
+      "loss": 0.6115,
+      "step": 5889
+    },
+    {
+      "epoch": 0.7668879344006247,
+      "grad_norm": 2.867833137512207,
+      "learning_rate": 8.65425217568961e-06,
+      "loss": 0.7631,
+      "step": 5892
+    },
+    {
+      "epoch": 0.7672784068723155,
+      "grad_norm": 2.566713571548462,
+      "learning_rate": 8.652825459404065e-06,
+      "loss": 0.5086,
+      "step": 5895
+    },
+    {
+      "epoch": 0.7676688793440063,
+      "grad_norm": 2.6429190635681152,
+      "learning_rate": 8.651398104977595e-06,
+      "loss": 0.5853,
+      "step": 5898
+    },
+    {
+      "epoch": 0.768059351815697,
+      "grad_norm": 2.9396679401397705,
+      "learning_rate": 8.649970112659558e-06,
+      "loss": 0.5673,
+      "step": 5901
+    },
+    {
+      "epoch": 0.7684498242873877,
+      "grad_norm": 2.601750612258911,
+      "learning_rate": 8.648541482699422e-06,
+      "loss": 0.7371,
+      "step": 5904
+    },
+    {
+      "epoch": 0.7688402967590785,
+      "grad_norm": 2.7551345825195312,
+      "learning_rate": 8.647112215346763e-06,
+      "loss": 0.6192,
+      "step": 5907
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 2.5565147399902344,
+      "learning_rate": 8.645682310851273e-06,
+      "loss": 0.6002,
+      "step": 5910
+    },
+    {
+      "epoch": 0.7696212417024599,
+      "grad_norm": 3.5521817207336426,
+      "learning_rate": 8.644251769462751e-06,
+      "loss": 0.6178,
+      "step": 5913
+    },
+    {
+      "epoch": 0.7700117141741507,
+      "grad_norm": 2.3843579292297363,
+      "learning_rate": 8.642820591431114e-06,
+      "loss": 0.5701,
+      "step": 5916
+    },
+    {
+      "epoch": 0.7704021866458415,
+      "grad_norm": 2.349097490310669,
+      "learning_rate": 8.641388777006381e-06,
+      "loss": 0.6391,
+      "step": 5919
+    },
+    {
+      "epoch": 0.7707926591175323,
+      "grad_norm": 2.2657618522644043,
+      "learning_rate": 8.639956326438688e-06,
+      "loss": 0.6661,
+      "step": 5922
+    },
+    {
+      "epoch": 0.7711831315892229,
+      "grad_norm": 3.2351880073547363,
+      "learning_rate": 8.638523239978285e-06,
+      "loss": 0.6367,
+      "step": 5925
+    },
+    {
+      "epoch": 0.7715736040609137,
+      "grad_norm": 2.358475923538208,
+      "learning_rate": 8.637089517875524e-06,
+      "loss": 0.6163,
+      "step": 5928
+    },
+    {
+      "epoch": 0.7719640765326045,
+      "grad_norm": 2.8347368240356445,
+      "learning_rate": 8.635655160380879e-06,
+      "loss": 0.7084,
+      "step": 5931
+    },
+    {
+      "epoch": 0.7723545490042952,
+      "grad_norm": 2.7897112369537354,
+      "learning_rate": 8.634220167744922e-06,
+      "loss": 0.612,
+      "step": 5934
+    },
+    {
+      "epoch": 0.7727450214759859,
+      "grad_norm": 2.506141424179077,
+      "learning_rate": 8.632784540218348e-06,
+      "loss": 0.6386,
+      "step": 5937
+    },
+    {
+      "epoch": 0.7731354939476767,
+      "grad_norm": 2.703279495239258,
+      "learning_rate": 8.631348278051956e-06,
+      "loss": 0.6336,
+      "step": 5940
+    },
+    {
+      "epoch": 0.7735259664193674,
+      "grad_norm": 2.3759536743164062,
+      "learning_rate": 8.62991138149666e-06,
+      "loss": 0.6074,
+      "step": 5943
+    },
+    {
+      "epoch": 0.7739164388910582,
+      "grad_norm": 3.115846872329712,
+      "learning_rate": 8.628473850803484e-06,
+      "loss": 0.6421,
+      "step": 5946
+    },
+    {
+      "epoch": 0.7743069113627489,
+      "grad_norm": 3.065495729446411,
+      "learning_rate": 8.627035686223557e-06,
+      "loss": 0.6935,
+      "step": 5949
+    },
+    {
+      "epoch": 0.7746973838344396,
+      "grad_norm": 2.664783477783203,
+      "learning_rate": 8.625596888008127e-06,
+      "loss": 0.6864,
+      "step": 5952
+    },
+    {
+      "epoch": 0.7750878563061304,
+      "grad_norm": 2.6792876720428467,
+      "learning_rate": 8.624157456408547e-06,
+      "loss": 0.7312,
+      "step": 5955
+    },
+    {
+      "epoch": 0.7754783287778212,
+      "grad_norm": 2.283553123474121,
+      "learning_rate": 8.622717391676284e-06,
+      "loss": 0.6556,
+      "step": 5958
+    },
+    {
+      "epoch": 0.775868801249512,
+      "grad_norm": 3.2782957553863525,
+      "learning_rate": 8.621276694062915e-06,
+      "loss": 0.6916,
+      "step": 5961
+    },
+    {
+      "epoch": 0.7762592737212026,
+      "grad_norm": 2.4025683403015137,
+      "learning_rate": 8.619835363820123e-06,
+      "loss": 0.8082,
+      "step": 5964
+    },
+    {
+      "epoch": 0.7766497461928934,
+      "grad_norm": 2.7000598907470703,
+      "learning_rate": 8.61839340119971e-06,
+      "loss": 0.6322,
+      "step": 5967
+    },
+    {
+      "epoch": 0.7770402186645842,
+      "grad_norm": 3.0906264781951904,
+      "learning_rate": 8.61695080645358e-06,
+      "loss": 0.6954,
+      "step": 5970
+    },
+    {
+      "epoch": 0.7774306911362749,
+      "grad_norm": 2.45308256149292,
+      "learning_rate": 8.615507579833754e-06,
+      "loss": 0.7327,
+      "step": 5973
+    },
+    {
+      "epoch": 0.7778211636079656,
+      "grad_norm": 2.349186420440674,
+      "learning_rate": 8.61406372159236e-06,
+      "loss": 0.6414,
+      "step": 5976
+    },
+    {
+      "epoch": 0.7782116360796564,
+      "grad_norm": 2.472166061401367,
+      "learning_rate": 8.612619231981636e-06,
+      "loss": 0.6639,
+      "step": 5979
+    },
+    {
+      "epoch": 0.7786021085513471,
+      "grad_norm": 2.506831645965576,
+      "learning_rate": 8.611174111253932e-06,
+      "loss": 0.5965,
+      "step": 5982
+    },
+    {
+      "epoch": 0.7789925810230379,
+      "grad_norm": 3.1906239986419678,
+      "learning_rate": 8.609728359661709e-06,
+      "loss": 0.6799,
+      "step": 5985
+    },
+    {
+      "epoch": 0.7793830534947286,
+      "grad_norm": 3.2461135387420654,
+      "learning_rate": 8.608281977457534e-06,
+      "loss": 0.7414,
+      "step": 5988
+    },
+    {
+      "epoch": 0.7797735259664194,
+      "grad_norm": 2.6268630027770996,
+      "learning_rate": 8.606834964894089e-06,
+      "loss": 0.655,
+      "step": 5991
+    },
+    {
+      "epoch": 0.7801639984381101,
+      "grad_norm": 2.3125433921813965,
+      "learning_rate": 8.605387322224162e-06,
+      "loss": 0.6857,
+      "step": 5994
+    },
+    {
+      "epoch": 0.7805544709098009,
+      "grad_norm": 3.1327126026153564,
+      "learning_rate": 8.603939049700655e-06,
+      "loss": 0.7118,
+      "step": 5997
+    },
+    {
+      "epoch": 0.7809449433814916,
+      "grad_norm": 3.7016682624816895,
+      "learning_rate": 8.602490147576579e-06,
+      "loss": 0.7118,
+      "step": 6000
+    },
+    {
+      "epoch": 0.7813354158531823,
+      "grad_norm": 2.4639270305633545,
+      "learning_rate": 8.601040616105053e-06,
+      "loss": 0.7133,
+      "step": 6003
+    },
+    {
+      "epoch": 0.7817258883248731,
+      "grad_norm": 2.384366989135742,
+      "learning_rate": 8.599590455539304e-06,
+      "loss": 0.6218,
+      "step": 6006
+    },
+    {
+      "epoch": 0.7821163607965639,
+      "grad_norm": 3.026371717453003,
+      "learning_rate": 8.598139666132676e-06,
+      "loss": 0.6333,
+      "step": 6009
+    },
+    {
+      "epoch": 0.7825068332682545,
+      "grad_norm": 2.3918018341064453,
+      "learning_rate": 8.596688248138618e-06,
+      "loss": 0.5672,
+      "step": 6012
+    },
+    {
+      "epoch": 0.7828973057399453,
+      "grad_norm": 2.7614986896514893,
+      "learning_rate": 8.59523620181069e-06,
+      "loss": 0.68,
+      "step": 6015
+    },
+    {
+      "epoch": 0.7832877782116361,
+      "grad_norm": 2.838771343231201,
+      "learning_rate": 8.59378352740256e-06,
+      "loss": 0.7742,
+      "step": 6018
+    },
+    {
+      "epoch": 0.7836782506833269,
+      "grad_norm": 2.317967176437378,
+      "learning_rate": 8.592330225168008e-06,
+      "loss": 0.6576,
+      "step": 6021
+    },
+    {
+      "epoch": 0.7840687231550175,
+      "grad_norm": 2.4622199535369873,
+      "learning_rate": 8.590876295360924e-06,
+      "loss": 0.706,
+      "step": 6024
+    },
+    {
+      "epoch": 0.7844591956267083,
+      "grad_norm": 2.2674942016601562,
+      "learning_rate": 8.589421738235304e-06,
+      "loss": 0.5825,
+      "step": 6027
+    },
+    {
+      "epoch": 0.7848496680983991,
+      "grad_norm": 2.866095781326294,
+      "learning_rate": 8.587966554045259e-06,
+      "loss": 0.6909,
+      "step": 6030
+    },
+    {
+      "epoch": 0.7852401405700898,
+      "grad_norm": 2.193502902984619,
+      "learning_rate": 8.586510743045002e-06,
+      "loss": 0.7236,
+      "step": 6033
+    },
+    {
+      "epoch": 0.7856306130417805,
+      "grad_norm": 2.5779576301574707,
+      "learning_rate": 8.585054305488866e-06,
+      "loss": 0.6158,
+      "step": 6036
+    },
+    {
+      "epoch": 0.7860210855134713,
+      "grad_norm": 2.6559667587280273,
+      "learning_rate": 8.583597241631283e-06,
+      "loss": 0.5794,
+      "step": 6039
+    },
+    {
+      "epoch": 0.786411557985162,
+      "grad_norm": 2.5779333114624023,
+      "learning_rate": 8.582139551726801e-06,
+      "loss": 0.6935,
+      "step": 6042
+    },
+    {
+      "epoch": 0.7868020304568528,
+      "grad_norm": 2.5629148483276367,
+      "learning_rate": 8.580681236030075e-06,
+      "loss": 0.615,
+      "step": 6045
+    },
+    {
+      "epoch": 0.7871925029285436,
+      "grad_norm": 2.830688238143921,
+      "learning_rate": 8.579222294795871e-06,
+      "loss": 0.6147,
+      "step": 6048
+    },
+    {
+      "epoch": 0.7875829754002343,
+      "grad_norm": 2.249699115753174,
+      "learning_rate": 8.57776272827906e-06,
+      "loss": 0.6282,
+      "step": 6051
+    },
+    {
+      "epoch": 0.787973447871925,
+      "grad_norm": 2.603372097015381,
+      "learning_rate": 8.576302536734628e-06,
+      "loss": 0.8085,
+      "step": 6054
+    },
+    {
+      "epoch": 0.7883639203436158,
+      "grad_norm": 2.5415706634521484,
+      "learning_rate": 8.574841720417666e-06,
+      "loss": 0.733,
+      "step": 6057
+    },
+    {
+      "epoch": 0.7887543928153066,
+      "grad_norm": 2.7278621196746826,
+      "learning_rate": 8.573380279583374e-06,
+      "loss": 0.6182,
+      "step": 6060
+    },
+    {
+      "epoch": 0.7891448652869972,
+      "grad_norm": 2.673105001449585,
+      "learning_rate": 8.571918214487068e-06,
+      "loss": 0.702,
+      "step": 6063
+    },
+    {
+      "epoch": 0.789535337758688,
+      "grad_norm": 2.781696081161499,
+      "learning_rate": 8.570455525384161e-06,
+      "loss": 0.6388,
+      "step": 6066
+    },
+    {
+      "epoch": 0.7899258102303788,
+      "grad_norm": 2.392241954803467,
+      "learning_rate": 8.568992212530187e-06,
+      "loss": 0.5784,
+      "step": 6069
+    },
+    {
+      "epoch": 0.7903162827020696,
+      "grad_norm": 3.162004232406616,
+      "learning_rate": 8.56752827618078e-06,
+      "loss": 0.7068,
+      "step": 6072
+    },
+    {
+      "epoch": 0.7907067551737602,
+      "grad_norm": 2.509340763092041,
+      "learning_rate": 8.566063716591689e-06,
+      "loss": 0.6832,
+      "step": 6075
+    },
+    {
+      "epoch": 0.791097227645451,
+      "grad_norm": 3.036722421646118,
+      "learning_rate": 8.56459853401877e-06,
+      "loss": 0.7436,
+      "step": 6078
+    },
+    {
+      "epoch": 0.7914877001171418,
+      "grad_norm": 2.521820306777954,
+      "learning_rate": 8.563132728717983e-06,
+      "loss": 0.6142,
+      "step": 6081
+    },
+    {
+      "epoch": 0.7918781725888325,
+      "grad_norm": 2.626577377319336,
+      "learning_rate": 8.561666300945406e-06,
+      "loss": 0.6366,
+      "step": 6084
+    },
+    {
+      "epoch": 0.7922686450605232,
+      "grad_norm": 2.6800007820129395,
+      "learning_rate": 8.560199250957218e-06,
+      "loss": 0.6728,
+      "step": 6087
+    },
+    {
+      "epoch": 0.792659117532214,
+      "grad_norm": 2.4256107807159424,
+      "learning_rate": 8.55873157900971e-06,
+      "loss": 0.6083,
+      "step": 6090
+    },
+    {
+      "epoch": 0.7930495900039047,
+      "grad_norm": 2.7973265647888184,
+      "learning_rate": 8.557263285359282e-06,
+      "loss": 0.6473,
+      "step": 6093
+    },
+    {
+      "epoch": 0.7934400624755955,
+      "grad_norm": 4.817023277282715,
+      "learning_rate": 8.55579437026244e-06,
+      "loss": 0.6608,
+      "step": 6096
+    },
+    {
+      "epoch": 0.7938305349472862,
+      "grad_norm": 2.4599454402923584,
+      "learning_rate": 8.554324833975805e-06,
+      "loss": 0.5633,
+      "step": 6099
+    },
+    {
+      "epoch": 0.7942210074189769,
+      "grad_norm": 2.911656618118286,
+      "learning_rate": 8.552854676756097e-06,
+      "loss": 0.683,
+      "step": 6102
+    },
+    {
+      "epoch": 0.7946114798906677,
+      "grad_norm": 2.656649589538574,
+      "learning_rate": 8.551383898860152e-06,
+      "loss": 0.7645,
+      "step": 6105
+    },
+    {
+      "epoch": 0.7950019523623585,
+      "grad_norm": 2.6597232818603516,
+      "learning_rate": 8.54991250054491e-06,
+      "loss": 0.6028,
+      "step": 6108
+    },
+    {
+      "epoch": 0.7953924248340491,
+      "grad_norm": 2.368178367614746,
+      "learning_rate": 8.548440482067422e-06,
+      "loss": 0.6649,
+      "step": 6111
+    },
+    {
+      "epoch": 0.7957828973057399,
+      "grad_norm": 2.8005642890930176,
+      "learning_rate": 8.546967843684846e-06,
+      "loss": 0.6525,
+      "step": 6114
+    },
+    {
+      "epoch": 0.7961733697774307,
+      "grad_norm": 2.426015853881836,
+      "learning_rate": 8.54549458565445e-06,
+      "loss": 0.5669,
+      "step": 6117
+    },
+    {
+      "epoch": 0.7965638422491215,
+      "grad_norm": 2.4316413402557373,
+      "learning_rate": 8.544020708233608e-06,
+      "loss": 0.6233,
+      "step": 6120
+    },
+    {
+      "epoch": 0.7969543147208121,
+      "grad_norm": 2.4489283561706543,
+      "learning_rate": 8.542546211679806e-06,
+      "loss": 0.7188,
+      "step": 6123
+    },
+    {
+      "epoch": 0.7973447871925029,
+      "grad_norm": 3.5121216773986816,
+      "learning_rate": 8.541071096250633e-06,
+      "loss": 0.6304,
+      "step": 6126
+    },
+    {
+      "epoch": 0.7977352596641937,
+      "grad_norm": 2.336344003677368,
+      "learning_rate": 8.539595362203787e-06,
+      "loss": 0.6539,
+      "step": 6129
+    },
+    {
+      "epoch": 0.7981257321358844,
+      "grad_norm": 2.4872500896453857,
+      "learning_rate": 8.538119009797079e-06,
+      "loss": 0.6592,
+      "step": 6132
+    },
+    {
+      "epoch": 0.7985162046075752,
+      "grad_norm": 2.25089955329895,
+      "learning_rate": 8.536642039288421e-06,
+      "loss": 0.6421,
+      "step": 6135
+    },
+    {
+      "epoch": 0.7989066770792659,
+      "grad_norm": 4.216074466705322,
+      "learning_rate": 8.53516445093584e-06,
+      "loss": 0.6712,
+      "step": 6138
+    },
+    {
+      "epoch": 0.7992971495509567,
+      "grad_norm": 3.146594285964966,
+      "learning_rate": 8.533686244997466e-06,
+      "loss": 0.7159,
+      "step": 6141
+    },
+    {
+      "epoch": 0.7996876220226474,
+      "grad_norm": 2.209770441055298,
+      "learning_rate": 8.53220742173154e-06,
+      "loss": 0.6019,
+      "step": 6144
+    },
+    {
+      "epoch": 0.8000780944943382,
+      "grad_norm": 3.0704033374786377,
+      "learning_rate": 8.530727981396406e-06,
+      "loss": 0.7287,
+      "step": 6147
+    },
+    {
+      "epoch": 0.8004685669660289,
+      "grad_norm": 2.567683219909668,
+      "learning_rate": 8.529247924250524e-06,
+      "loss": 0.6043,
+      "step": 6150
+    },
+    {
+      "epoch": 0.8008590394377196,
+      "grad_norm": 3.0786571502685547,
+      "learning_rate": 8.527767250552452e-06,
+      "loss": 0.6735,
+      "step": 6153
+    },
+    {
+      "epoch": 0.8012495119094104,
+      "grad_norm": 2.3544533252716064,
+      "learning_rate": 8.526285960560864e-06,
+      "loss": 0.6422,
+      "step": 6156
+    },
+    {
+      "epoch": 0.8016399843811012,
+      "grad_norm": 2.1804707050323486,
+      "learning_rate": 8.524804054534535e-06,
+      "loss": 0.6256,
+      "step": 6159
+    },
+    {
+      "epoch": 0.8020304568527918,
+      "grad_norm": 2.7374136447906494,
+      "learning_rate": 8.523321532732354e-06,
+      "loss": 0.5924,
+      "step": 6162
+    },
+    {
+      "epoch": 0.8024209293244826,
+      "grad_norm": 2.6108615398406982,
+      "learning_rate": 8.521838395413312e-06,
+      "loss": 0.7202,
+      "step": 6165
+    },
+    {
+      "epoch": 0.8028114017961734,
+      "grad_norm": 2.354628086090088,
+      "learning_rate": 8.520354642836512e-06,
+      "loss": 0.6135,
+      "step": 6168
+    },
+    {
+      "epoch": 0.8032018742678642,
+      "grad_norm": 2.340169668197632,
+      "learning_rate": 8.518870275261161e-06,
+      "loss": 0.6044,
+      "step": 6171
+    },
+    {
+      "epoch": 0.8035923467395548,
+      "grad_norm": 2.519382953643799,
+      "learning_rate": 8.517385292946578e-06,
+      "loss": 0.6896,
+      "step": 6174
+    },
+    {
+      "epoch": 0.8039828192112456,
+      "grad_norm": 2.6927080154418945,
+      "learning_rate": 8.515899696152183e-06,
+      "loss": 0.5729,
+      "step": 6177
+    },
+    {
+      "epoch": 0.8043732916829364,
+      "grad_norm": 2.767089605331421,
+      "learning_rate": 8.514413485137505e-06,
+      "loss": 0.5569,
+      "step": 6180
+    },
+    {
+      "epoch": 0.8047637641546271,
+      "grad_norm": 3.1286277770996094,
+      "learning_rate": 8.512926660162186e-06,
+      "loss": 0.6846,
+      "step": 6183
+    },
+    {
+      "epoch": 0.8051542366263178,
+      "grad_norm": 3.36635684967041,
+      "learning_rate": 8.511439221485971e-06,
+      "loss": 0.6434,
+      "step": 6186
+    },
+    {
+      "epoch": 0.8055447090980086,
+      "grad_norm": 2.8603038787841797,
+      "learning_rate": 8.50995116936871e-06,
+      "loss": 0.7117,
+      "step": 6189
+    },
+    {
+      "epoch": 0.8059351815696993,
+      "grad_norm": 2.3195858001708984,
+      "learning_rate": 8.508462504070363e-06,
+      "loss": 0.6657,
+      "step": 6192
+    },
+    {
+      "epoch": 0.8063256540413901,
+      "grad_norm": 2.69417142868042,
+      "learning_rate": 8.506973225850996e-06,
+      "loss": 0.6849,
+      "step": 6195
+    },
+    {
+      "epoch": 0.8067161265130808,
+      "grad_norm": 2.8202831745147705,
+      "learning_rate": 8.505483334970787e-06,
+      "loss": 0.7059,
+      "step": 6198
+    },
+    {
+      "epoch": 0.8071065989847716,
+      "grad_norm": 2.3420493602752686,
+      "learning_rate": 8.503992831690011e-06,
+      "loss": 0.6732,
+      "step": 6201
+    },
+    {
+      "epoch": 0.8074970714564623,
+      "grad_norm": 2.349193811416626,
+      "learning_rate": 8.502501716269061e-06,
+      "loss": 0.6318,
+      "step": 6204
+    },
+    {
+      "epoch": 0.8078875439281531,
+      "grad_norm": 2.51961350440979,
+      "learning_rate": 8.501009988968427e-06,
+      "loss": 0.6682,
+      "step": 6207
+    },
+    {
+      "epoch": 0.8082780163998438,
+      "grad_norm": 2.5512888431549072,
+      "learning_rate": 8.499517650048715e-06,
+      "loss": 0.6768,
+      "step": 6210
+    },
+    {
+      "epoch": 0.8086684888715345,
+      "grad_norm": 2.58780837059021,
+      "learning_rate": 8.498024699770631e-06,
+      "loss": 0.6196,
+      "step": 6213
+    },
+    {
+      "epoch": 0.8090589613432253,
+      "grad_norm": 3.9888927936553955,
+      "learning_rate": 8.49653113839499e-06,
+      "loss": 0.7719,
+      "step": 6216
+    },
+    {
+      "epoch": 0.8094494338149161,
+      "grad_norm": 3.109391927719116,
+      "learning_rate": 8.495036966182716e-06,
+      "loss": 0.7418,
+      "step": 6219
+    },
+    {
+      "epoch": 0.8098399062866068,
+      "grad_norm": 2.7313342094421387,
+      "learning_rate": 8.493542183394835e-06,
+      "loss": 0.7156,
+      "step": 6222
+    },
+    {
+      "epoch": 0.8102303787582975,
+      "grad_norm": 2.2368297576904297,
+      "learning_rate": 8.492046790292485e-06,
+      "loss": 0.673,
+      "step": 6225
+    },
+    {
+      "epoch": 0.8106208512299883,
+      "grad_norm": 2.565585136413574,
+      "learning_rate": 8.490550787136906e-06,
+      "loss": 0.6379,
+      "step": 6228
+    },
+    {
+      "epoch": 0.811011323701679,
+      "grad_norm": 2.4983272552490234,
+      "learning_rate": 8.489054174189448e-06,
+      "loss": 0.6459,
+      "step": 6231
+    },
+    {
+      "epoch": 0.8114017961733698,
+      "grad_norm": 2.1908957958221436,
+      "learning_rate": 8.487556951711567e-06,
+      "loss": 0.6178,
+      "step": 6234
+    },
+    {
+      "epoch": 0.8117922686450605,
+      "grad_norm": 2.3688669204711914,
+      "learning_rate": 8.486059119964822e-06,
+      "loss": 0.7055,
+      "step": 6237
+    },
+    {
+      "epoch": 0.8121827411167513,
+      "grad_norm": 2.3758630752563477,
+      "learning_rate": 8.484560679210883e-06,
+      "loss": 0.5999,
+      "step": 6240
+    },
+    {
+      "epoch": 0.812573213588442,
+      "grad_norm": 2.6127567291259766,
+      "learning_rate": 8.483061629711522e-06,
+      "loss": 0.7506,
+      "step": 6243
+    },
+    {
+      "epoch": 0.8129636860601328,
+      "grad_norm": 3.369610548019409,
+      "learning_rate": 8.481561971728622e-06,
+      "loss": 0.6454,
+      "step": 6246
+    },
+    {
+      "epoch": 0.8133541585318235,
+      "grad_norm": 2.357160806655884,
+      "learning_rate": 8.480061705524173e-06,
+      "loss": 0.621,
+      "step": 6249
+    },
+    {
+      "epoch": 0.8137446310035142,
+      "grad_norm": 2.6496059894561768,
+      "learning_rate": 8.47856083136026e-06,
+      "loss": 0.7154,
+      "step": 6252
+    },
+    {
+      "epoch": 0.814135103475205,
+      "grad_norm": 2.5359816551208496,
+      "learning_rate": 8.47705934949909e-06,
+      "loss": 0.7051,
+      "step": 6255
+    },
+    {
+      "epoch": 0.8145255759468958,
+      "grad_norm": 2.283336877822876,
+      "learning_rate": 8.475557260202966e-06,
+      "loss": 0.646,
+      "step": 6258
+    },
+    {
+      "epoch": 0.8149160484185864,
+      "grad_norm": 2.6512975692749023,
+      "learning_rate": 8.474054563734303e-06,
+      "loss": 0.6226,
+      "step": 6261
+    },
+    {
+      "epoch": 0.8153065208902772,
+      "grad_norm": 2.991300582885742,
+      "learning_rate": 8.472551260355612e-06,
+      "loss": 0.7599,
+      "step": 6264
+    },
+    {
+      "epoch": 0.815696993361968,
+      "grad_norm": 2.309941530227661,
+      "learning_rate": 8.471047350329523e-06,
+      "loss": 0.5883,
+      "step": 6267
+    },
+    {
+      "epoch": 0.8160874658336588,
+      "grad_norm": 2.3625473976135254,
+      "learning_rate": 8.469542833918762e-06,
+      "loss": 0.7385,
+      "step": 6270
+    },
+    {
+      "epoch": 0.8164779383053494,
+      "grad_norm": 2.453748941421509,
+      "learning_rate": 8.46803771138617e-06,
+      "loss": 0.5488,
+      "step": 6273
+    },
+    {
+      "epoch": 0.8168684107770402,
+      "grad_norm": 3.088710069656372,
+      "learning_rate": 8.466531982994684e-06,
+      "loss": 0.6956,
+      "step": 6276
+    },
+    {
+      "epoch": 0.817258883248731,
+      "grad_norm": 2.4059526920318604,
+      "learning_rate": 8.465025649007352e-06,
+      "loss": 0.5862,
+      "step": 6279
+    },
+    {
+      "epoch": 0.8176493557204217,
+      "grad_norm": 3.4859941005706787,
+      "learning_rate": 8.463518709687328e-06,
+      "loss": 0.5537,
+      "step": 6282
+    },
+    {
+      "epoch": 0.8180398281921124,
+      "grad_norm": 3.3922863006591797,
+      "learning_rate": 8.462011165297873e-06,
+      "loss": 0.6667,
+      "step": 6285
+    },
+    {
+      "epoch": 0.8184303006638032,
+      "grad_norm": 3.5943069458007812,
+      "learning_rate": 8.46050301610235e-06,
+      "loss": 0.6736,
+      "step": 6288
+    },
+    {
+      "epoch": 0.818820773135494,
+      "grad_norm": 2.6332778930664062,
+      "learning_rate": 8.45899426236423e-06,
+      "loss": 0.7386,
+      "step": 6291
+    },
+    {
+      "epoch": 0.8192112456071847,
+      "grad_norm": 3.304346799850464,
+      "learning_rate": 8.45748490434709e-06,
+      "loss": 0.7207,
+      "step": 6294
+    },
+    {
+      "epoch": 0.8196017180788754,
+      "grad_norm": 2.711082935333252,
+      "learning_rate": 8.455974942314612e-06,
+      "loss": 0.6716,
+      "step": 6297
+    },
+    {
+      "epoch": 0.8199921905505662,
+      "grad_norm": 2.3270423412323,
+      "learning_rate": 8.454464376530579e-06,
+      "loss": 0.634,
+      "step": 6300
+    },
+    {
+      "epoch": 0.8203826630222569,
+      "grad_norm": 2.4965059757232666,
+      "learning_rate": 8.452953207258888e-06,
+      "loss": 0.6086,
+      "step": 6303
+    },
+    {
+      "epoch": 0.8207731354939477,
+      "grad_norm": 2.57068133354187,
+      "learning_rate": 8.451441434763534e-06,
+      "loss": 0.6383,
+      "step": 6306
+    },
+    {
+      "epoch": 0.8211636079656385,
+      "grad_norm": 3.073572874069214,
+      "learning_rate": 8.449929059308623e-06,
+      "loss": 0.6085,
+      "step": 6309
+    },
+    {
+      "epoch": 0.8215540804373291,
+      "grad_norm": 2.8972861766815186,
+      "learning_rate": 8.448416081158363e-06,
+      "loss": 0.76,
+      "step": 6312
+    },
+    {
+      "epoch": 0.8219445529090199,
+      "grad_norm": 2.6822032928466797,
+      "learning_rate": 8.446902500577067e-06,
+      "loss": 0.6039,
+      "step": 6315
+    },
+    {
+      "epoch": 0.8223350253807107,
+      "grad_norm": 2.42512845993042,
+      "learning_rate": 8.445388317829157e-06,
+      "loss": 0.6074,
+      "step": 6318
+    },
+    {
+      "epoch": 0.8227254978524015,
+      "grad_norm": 2.6248953342437744,
+      "learning_rate": 8.443873533179156e-06,
+      "loss": 0.7003,
+      "step": 6321
+    },
+    {
+      "epoch": 0.8231159703240921,
+      "grad_norm": 3.51706862449646,
+      "learning_rate": 8.442358146891692e-06,
+      "loss": 0.6427,
+      "step": 6324
+    },
+    {
+      "epoch": 0.8235064427957829,
+      "grad_norm": 2.4549152851104736,
+      "learning_rate": 8.440842159231503e-06,
+      "loss": 0.662,
+      "step": 6327
+    },
+    {
+      "epoch": 0.8238969152674737,
+      "grad_norm": 2.4990108013153076,
+      "learning_rate": 8.439325570463426e-06,
+      "loss": 0.7352,
+      "step": 6330
+    },
+    {
+      "epoch": 0.8242873877391644,
+      "grad_norm": 3.4981563091278076,
+      "learning_rate": 8.437808380852408e-06,
+      "loss": 0.7033,
+      "step": 6333
+    },
+    {
+      "epoch": 0.8246778602108551,
+      "grad_norm": 3.689436674118042,
+      "learning_rate": 8.436290590663498e-06,
+      "loss": 0.6423,
+      "step": 6336
+    },
+    {
+      "epoch": 0.8250683326825459,
+      "grad_norm": 2.182603359222412,
+      "learning_rate": 8.43477220016185e-06,
+      "loss": 0.6086,
+      "step": 6339
+    },
+    {
+      "epoch": 0.8254588051542366,
+      "grad_norm": 2.4376380443573,
+      "learning_rate": 8.433253209612727e-06,
+      "loss": 0.6321,
+      "step": 6342
+    },
+    {
+      "epoch": 0.8258492776259274,
+      "grad_norm": 2.2977375984191895,
+      "learning_rate": 8.431733619281486e-06,
+      "loss": 0.6188,
+      "step": 6345
+    },
+    {
+      "epoch": 0.8262397500976181,
+      "grad_norm": 2.3864758014678955,
+      "learning_rate": 8.430213429433605e-06,
+      "loss": 0.653,
+      "step": 6348
+    },
+    {
+      "epoch": 0.8266302225693088,
+      "grad_norm": 3.494652509689331,
+      "learning_rate": 8.42869264033465e-06,
+      "loss": 0.6787,
+      "step": 6351
+    },
+    {
+      "epoch": 0.8270206950409996,
+      "grad_norm": 2.394007682800293,
+      "learning_rate": 8.427171252250308e-06,
+      "loss": 0.5673,
+      "step": 6354
+    },
+    {
+      "epoch": 0.8274111675126904,
+      "grad_norm": 2.576982259750366,
+      "learning_rate": 8.425649265446356e-06,
+      "loss": 0.7278,
+      "step": 6357
+    },
+    {
+      "epoch": 0.827801639984381,
+      "grad_norm": 2.5857975482940674,
+      "learning_rate": 8.424126680188684e-06,
+      "loss": 0.7381,
+      "step": 6360
+    },
+    {
+      "epoch": 0.8281921124560718,
+      "grad_norm": 2.553446054458618,
+      "learning_rate": 8.422603496743285e-06,
+      "loss": 0.5989,
+      "step": 6363
+    },
+    {
+      "epoch": 0.8285825849277626,
+      "grad_norm": 2.7090554237365723,
+      "learning_rate": 8.421079715376255e-06,
+      "loss": 0.6736,
+      "step": 6366
+    },
+    {
+      "epoch": 0.8289730573994534,
+      "grad_norm": 2.7122576236724854,
+      "learning_rate": 8.419555336353793e-06,
+      "loss": 0.6736,
+      "step": 6369
+    },
+    {
+      "epoch": 0.829363529871144,
+      "grad_norm": 2.37516713142395,
+      "learning_rate": 8.418030359942211e-06,
+      "loss": 0.6084,
+      "step": 6372
+    },
+    {
+      "epoch": 0.8297540023428348,
+      "grad_norm": 2.34924054145813,
+      "learning_rate": 8.416504786407913e-06,
+      "loss": 0.7132,
+      "step": 6375
+    },
+    {
+      "epoch": 0.8301444748145256,
+      "grad_norm": 2.3876399993896484,
+      "learning_rate": 8.414978616017418e-06,
+      "loss": 0.7109,
+      "step": 6378
+    },
+    {
+      "epoch": 0.8305349472862164,
+      "grad_norm": 2.995637893676758,
+      "learning_rate": 8.413451849037342e-06,
+      "loss": 0.6192,
+      "step": 6381
+    },
+    {
+      "epoch": 0.830925419757907,
+      "grad_norm": 2.5374836921691895,
+      "learning_rate": 8.41192448573441e-06,
+      "loss": 0.6034,
+      "step": 6384
+    },
+    {
+      "epoch": 0.8313158922295978,
+      "grad_norm": 2.576464891433716,
+      "learning_rate": 8.410396526375446e-06,
+      "loss": 0.6396,
+      "step": 6387
+    },
+    {
+      "epoch": 0.8317063647012886,
+      "grad_norm": 2.4232656955718994,
+      "learning_rate": 8.408867971227384e-06,
+      "loss": 0.6386,
+      "step": 6390
+    },
+    {
+      "epoch": 0.8320968371729793,
+      "grad_norm": 2.4770760536193848,
+      "learning_rate": 8.40733882055726e-06,
+      "loss": 0.6802,
+      "step": 6393
+    },
+    {
+      "epoch": 0.8324873096446701,
+      "grad_norm": 2.2709827423095703,
+      "learning_rate": 8.40580907463221e-06,
+      "loss": 0.6144,
+      "step": 6396
+    },
+    {
+      "epoch": 0.8328777821163608,
+      "grad_norm": 2.4368174076080322,
+      "learning_rate": 8.40427873371948e-06,
+      "loss": 0.6847,
+      "step": 6399
+    },
+    {
+      "epoch": 0.8332682545880515,
+      "grad_norm": 2.510909080505371,
+      "learning_rate": 8.402747798086417e-06,
+      "loss": 0.7264,
+      "step": 6402
+    },
+    {
+      "epoch": 0.8336587270597423,
+      "grad_norm": 3.945446491241455,
+      "learning_rate": 8.401216268000473e-06,
+      "loss": 0.7692,
+      "step": 6405
+    },
+    {
+      "epoch": 0.8340491995314331,
+      "grad_norm": 2.4416685104370117,
+      "learning_rate": 8.3996841437292e-06,
+      "loss": 0.6565,
+      "step": 6408
+    },
+    {
+      "epoch": 0.8344396720031237,
+      "grad_norm": 2.5184824466705322,
+      "learning_rate": 8.39815142554026e-06,
+      "loss": 0.6929,
+      "step": 6411
+    },
+    {
+      "epoch": 0.8348301444748145,
+      "grad_norm": 2.375833511352539,
+      "learning_rate": 8.396618113701416e-06,
+      "loss": 0.6166,
+      "step": 6414
+    },
+    {
+      "epoch": 0.8352206169465053,
+      "grad_norm": 2.406198024749756,
+      "learning_rate": 8.395084208480531e-06,
+      "loss": 0.5858,
+      "step": 6417
+    },
+    {
+      "epoch": 0.8356110894181961,
+      "grad_norm": 3.6379899978637695,
+      "learning_rate": 8.393549710145578e-06,
+      "loss": 0.5477,
+      "step": 6420
+    },
+    {
+      "epoch": 0.8360015618898867,
+      "grad_norm": 2.8601226806640625,
+      "learning_rate": 8.39201461896463e-06,
+      "loss": 0.7262,
+      "step": 6423
+    },
+    {
+      "epoch": 0.8363920343615775,
+      "grad_norm": 2.3782405853271484,
+      "learning_rate": 8.390478935205864e-06,
+      "loss": 0.672,
+      "step": 6426
+    },
+    {
+      "epoch": 0.8367825068332683,
+      "grad_norm": 2.471163749694824,
+      "learning_rate": 8.388942659137558e-06,
+      "loss": 0.6127,
+      "step": 6429
+    },
+    {
+      "epoch": 0.837172979304959,
+      "grad_norm": 2.592690944671631,
+      "learning_rate": 8.3874057910281e-06,
+      "loss": 0.7282,
+      "step": 6432
+    },
+    {
+      "epoch": 0.8375634517766497,
+      "grad_norm": 2.696824073791504,
+      "learning_rate": 8.385868331145977e-06,
+      "loss": 0.6982,
+      "step": 6435
+    },
+    {
+      "epoch": 0.8379539242483405,
+      "grad_norm": 2.5590860843658447,
+      "learning_rate": 8.38433027975978e-06,
+      "loss": 0.6906,
+      "step": 6438
+    },
+    {
+      "epoch": 0.8383443967200312,
+      "grad_norm": 2.547417163848877,
+      "learning_rate": 8.3827916371382e-06,
+      "loss": 0.6392,
+      "step": 6441
+    },
+    {
+      "epoch": 0.838734869191722,
+      "grad_norm": 2.264625310897827,
+      "learning_rate": 8.381252403550043e-06,
+      "loss": 0.6396,
+      "step": 6444
+    },
+    {
+      "epoch": 0.8391253416634127,
+      "grad_norm": 2.935336112976074,
+      "learning_rate": 8.3797125792642e-06,
+      "loss": 0.6588,
+      "step": 6447
+    },
+    {
+      "epoch": 0.8395158141351035,
+      "grad_norm": 2.308274745941162,
+      "learning_rate": 8.378172164549678e-06,
+      "loss": 0.6231,
+      "step": 6450
+    },
+    {
+      "epoch": 0.8399062866067942,
+      "grad_norm": 2.976280927658081,
+      "learning_rate": 8.376631159675587e-06,
+      "loss": 0.6117,
+      "step": 6453
+    },
+    {
+      "epoch": 0.840296759078485,
+      "grad_norm": 2.33328914642334,
+      "learning_rate": 8.375089564911137e-06,
+      "loss": 0.7065,
+      "step": 6456
+    },
+    {
+      "epoch": 0.8406872315501757,
+      "grad_norm": 2.5370125770568848,
+      "learning_rate": 8.373547380525639e-06,
+      "loss": 0.671,
+      "step": 6459
+    },
+    {
+      "epoch": 0.8410777040218664,
+      "grad_norm": 3.066671133041382,
+      "learning_rate": 8.372004606788511e-06,
+      "loss": 0.5947,
+      "step": 6462
+    },
+    {
+      "epoch": 0.8414681764935572,
+      "grad_norm": 2.562814474105835,
+      "learning_rate": 8.370461243969272e-06,
+      "loss": 0.5773,
+      "step": 6465
+    },
+    {
+      "epoch": 0.841858648965248,
+      "grad_norm": 2.8281662464141846,
+      "learning_rate": 8.368917292337544e-06,
+      "loss": 0.6202,
+      "step": 6468
+    },
+    {
+      "epoch": 0.8422491214369386,
+      "grad_norm": 3.3414337635040283,
+      "learning_rate": 8.36737275216305e-06,
+      "loss": 0.6927,
+      "step": 6471
+    },
+    {
+      "epoch": 0.8426395939086294,
+      "grad_norm": 2.5855307579040527,
+      "learning_rate": 8.365827623715624e-06,
+      "loss": 0.6142,
+      "step": 6474
+    },
+    {
+      "epoch": 0.8430300663803202,
+      "grad_norm": 2.5391762256622314,
+      "learning_rate": 8.36428190726519e-06,
+      "loss": 0.6912,
+      "step": 6477
+    },
+    {
+      "epoch": 0.843420538852011,
+      "grad_norm": 2.4301860332489014,
+      "learning_rate": 8.362735603081784e-06,
+      "loss": 0.7405,
+      "step": 6480
+    },
+    {
+      "epoch": 0.8438110113237017,
+      "grad_norm": 2.3643648624420166,
+      "learning_rate": 8.361188711435543e-06,
+      "loss": 0.5884,
+      "step": 6483
+    },
+    {
+      "epoch": 0.8442014837953924,
+      "grad_norm": 2.4440126419067383,
+      "learning_rate": 8.359641232596707e-06,
+      "loss": 0.6901,
+      "step": 6486
+    },
+    {
+      "epoch": 0.8445919562670832,
+      "grad_norm": 2.4889235496520996,
+      "learning_rate": 8.358093166835614e-06,
+      "loss": 0.6792,
+      "step": 6489
+    },
+    {
+      "epoch": 0.8449824287387739,
+      "grad_norm": 2.9333462715148926,
+      "learning_rate": 8.356544514422708e-06,
+      "loss": 0.535,
+      "step": 6492
+    },
+    {
+      "epoch": 0.8453729012104647,
+      "grad_norm": 3.2094194889068604,
+      "learning_rate": 8.354995275628536e-06,
+      "loss": 0.713,
+      "step": 6495
+    },
+    {
+      "epoch": 0.8457633736821554,
+      "grad_norm": 2.7420010566711426,
+      "learning_rate": 8.35344545072375e-06,
+      "loss": 0.6427,
+      "step": 6498
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 2.5620057582855225,
+      "learning_rate": 8.351895039979096e-06,
+      "loss": 0.7041,
+      "step": 6501
+    },
+    {
+      "epoch": 0.8465443186255369,
+      "grad_norm": 3.0430588722229004,
+      "learning_rate": 8.350344043665432e-06,
+      "loss": 0.6237,
+      "step": 6504
+    },
+    {
+      "epoch": 0.8469347910972277,
+      "grad_norm": 3.560917377471924,
+      "learning_rate": 8.34879246205371e-06,
+      "loss": 0.6937,
+      "step": 6507
+    },
+    {
+      "epoch": 0.8473252635689184,
+      "grad_norm": 2.730560064315796,
+      "learning_rate": 8.34724029541499e-06,
+      "loss": 0.6022,
+      "step": 6510
+    },
+    {
+      "epoch": 0.8477157360406091,
+      "grad_norm": 2.495201587677002,
+      "learning_rate": 8.345687544020432e-06,
+      "loss": 0.6019,
+      "step": 6513
+    },
+    {
+      "epoch": 0.8481062085122999,
+      "grad_norm": 2.6979150772094727,
+      "learning_rate": 8.344134208141298e-06,
+      "loss": 0.696,
+      "step": 6516
+    },
+    {
+      "epoch": 0.8484966809839907,
+      "grad_norm": 3.2949979305267334,
+      "learning_rate": 8.342580288048953e-06,
+      "loss": 0.6723,
+      "step": 6519
+    },
+    {
+      "epoch": 0.8488871534556813,
+      "grad_norm": 2.2897496223449707,
+      "learning_rate": 8.341025784014865e-06,
+      "loss": 0.6004,
+      "step": 6522
+    },
+    {
+      "epoch": 0.8492776259273721,
+      "grad_norm": 2.2679920196533203,
+      "learning_rate": 8.3394706963106e-06,
+      "loss": 0.6487,
+      "step": 6525
+    },
+    {
+      "epoch": 0.8496680983990629,
+      "grad_norm": 2.6986122131347656,
+      "learning_rate": 8.337915025207829e-06,
+      "loss": 0.6389,
+      "step": 6528
+    },
+    {
+      "epoch": 0.8500585708707536,
+      "grad_norm": 2.4359288215637207,
+      "learning_rate": 8.336358770978325e-06,
+      "loss": 0.5729,
+      "step": 6531
+    },
+    {
+      "epoch": 0.8504490433424443,
+      "grad_norm": 2.355321168899536,
+      "learning_rate": 8.334801933893963e-06,
+      "loss": 0.6444,
+      "step": 6534
+    },
+    {
+      "epoch": 0.8508395158141351,
+      "grad_norm": 2.818239212036133,
+      "learning_rate": 8.333244514226718e-06,
+      "loss": 0.6435,
+      "step": 6537
+    },
+    {
+      "epoch": 0.8512299882858259,
+      "grad_norm": 2.3462061882019043,
+      "learning_rate": 8.331686512248669e-06,
+      "loss": 0.7376,
+      "step": 6540
+    },
+    {
+      "epoch": 0.8516204607575166,
+      "grad_norm": 2.6086630821228027,
+      "learning_rate": 8.330127928231994e-06,
+      "loss": 0.6756,
+      "step": 6543
+    },
+    {
+      "epoch": 0.8520109332292073,
+      "grad_norm": 2.5126383304595947,
+      "learning_rate": 8.328568762448978e-06,
+      "loss": 0.6713,
+      "step": 6546
+    },
+    {
+      "epoch": 0.8524014057008981,
+      "grad_norm": 3.5347037315368652,
+      "learning_rate": 8.327009015172e-06,
+      "loss": 0.6268,
+      "step": 6549
+    },
+    {
+      "epoch": 0.8527918781725888,
+      "grad_norm": 2.4854538440704346,
+      "learning_rate": 8.325448686673545e-06,
+      "loss": 0.6253,
+      "step": 6552
+    },
+    {
+      "epoch": 0.8531823506442796,
+      "grad_norm": 2.2126731872558594,
+      "learning_rate": 8.323887777226204e-06,
+      "loss": 0.5835,
+      "step": 6555
+    },
+    {
+      "epoch": 0.8535728231159703,
+      "grad_norm": 3.192509889602661,
+      "learning_rate": 8.322326287102655e-06,
+      "loss": 0.5728,
+      "step": 6558
+    },
+    {
+      "epoch": 0.853963295587661,
+      "grad_norm": 2.4523346424102783,
+      "learning_rate": 8.320764216575696e-06,
+      "loss": 0.5734,
+      "step": 6561
+    },
+    {
+      "epoch": 0.8543537680593518,
+      "grad_norm": 2.1568102836608887,
+      "learning_rate": 8.319201565918214e-06,
+      "loss": 0.7255,
+      "step": 6564
+    },
+    {
+      "epoch": 0.8547442405310426,
+      "grad_norm": 2.56260347366333,
+      "learning_rate": 8.317638335403203e-06,
+      "loss": 0.6623,
+      "step": 6567
+    },
+    {
+      "epoch": 0.8551347130027334,
+      "grad_norm": 2.450410842895508,
+      "learning_rate": 8.31607452530375e-06,
+      "loss": 0.6081,
+      "step": 6570
+    },
+    {
+      "epoch": 0.855525185474424,
+      "grad_norm": 2.3223392963409424,
+      "learning_rate": 8.314510135893057e-06,
+      "loss": 0.6811,
+      "step": 6573
+    },
+    {
+      "epoch": 0.8559156579461148,
+      "grad_norm": 2.9705634117126465,
+      "learning_rate": 8.312945167444413e-06,
+      "loss": 0.7399,
+      "step": 6576
+    },
+    {
+      "epoch": 0.8563061304178056,
+      "grad_norm": 3.151143789291382,
+      "learning_rate": 8.31137962023122e-06,
+      "loss": 0.6165,
+      "step": 6579
+    },
+    {
+      "epoch": 0.8566966028894963,
+      "grad_norm": 2.2744345664978027,
+      "learning_rate": 8.309813494526973e-06,
+      "loss": 0.5956,
+      "step": 6582
+    },
+    {
+      "epoch": 0.857087075361187,
+      "grad_norm": 2.3356449604034424,
+      "learning_rate": 8.30824679060527e-06,
+      "loss": 0.6373,
+      "step": 6585
+    },
+    {
+      "epoch": 0.8574775478328778,
+      "grad_norm": 2.322843551635742,
+      "learning_rate": 8.306679508739813e-06,
+      "loss": 0.6431,
+      "step": 6588
+    },
+    {
+      "epoch": 0.8578680203045685,
+      "grad_norm": 2.799908399581909,
+      "learning_rate": 8.305111649204402e-06,
+      "loss": 0.6925,
+      "step": 6591
+    },
+    {
+      "epoch": 0.8582584927762593,
+      "grad_norm": 3.1793437004089355,
+      "learning_rate": 8.30354321227294e-06,
+      "loss": 0.6127,
+      "step": 6594
+    },
+    {
+      "epoch": 0.85864896524795,
+      "grad_norm": 2.3757338523864746,
+      "learning_rate": 8.301974198219427e-06,
+      "loss": 0.6473,
+      "step": 6597
+    },
+    {
+      "epoch": 0.8590394377196408,
+      "grad_norm": 2.6627843379974365,
+      "learning_rate": 8.300404607317968e-06,
+      "loss": 0.7966,
+      "step": 6600
+    },
+    {
+      "epoch": 0.8594299101913315,
+      "grad_norm": 2.2631454467773438,
+      "learning_rate": 8.298834439842768e-06,
+      "loss": 0.5955,
+      "step": 6603
+    },
+    {
+      "epoch": 0.8598203826630223,
+      "grad_norm": 3.50620174407959,
+      "learning_rate": 8.29726369606813e-06,
+      "loss": 0.6787,
+      "step": 6606
+    },
+    {
+      "epoch": 0.860210855134713,
+      "grad_norm": 2.180896520614624,
+      "learning_rate": 8.295692376268462e-06,
+      "loss": 0.561,
+      "step": 6609
+    },
+    {
+      "epoch": 0.8606013276064037,
+      "grad_norm": 2.7363123893737793,
+      "learning_rate": 8.29412048071827e-06,
+      "loss": 0.6337,
+      "step": 6612
+    },
+    {
+      "epoch": 0.8609918000780945,
+      "grad_norm": 2.419677495956421,
+      "learning_rate": 8.292548009692156e-06,
+      "loss": 0.6391,
+      "step": 6615
+    },
+    {
+      "epoch": 0.8613822725497853,
+      "grad_norm": 2.2987353801727295,
+      "learning_rate": 8.290974963464835e-06,
+      "loss": 0.5479,
+      "step": 6618
+    },
+    {
+      "epoch": 0.8617727450214759,
+      "grad_norm": 2.5347087383270264,
+      "learning_rate": 8.289401342311108e-06,
+      "loss": 0.6663,
+      "step": 6621
+    },
+    {
+      "epoch": 0.8621632174931667,
+      "grad_norm": 2.4396309852600098,
+      "learning_rate": 8.287827146505888e-06,
+      "loss": 0.5425,
+      "step": 6624
+    },
+    {
+      "epoch": 0.8625536899648575,
+      "grad_norm": 2.2343242168426514,
+      "learning_rate": 8.286252376324181e-06,
+      "loss": 0.5953,
+      "step": 6627
+    },
+    {
+      "epoch": 0.8629441624365483,
+      "grad_norm": 2.7455759048461914,
+      "learning_rate": 8.284677032041099e-06,
+      "loss": 0.6759,
+      "step": 6630
+    },
+    {
+      "epoch": 0.8633346349082389,
+      "grad_norm": 2.5142693519592285,
+      "learning_rate": 8.283101113931849e-06,
+      "loss": 0.6229,
+      "step": 6633
+    },
+    {
+      "epoch": 0.8637251073799297,
+      "grad_norm": 2.3203072547912598,
+      "learning_rate": 8.281524622271741e-06,
+      "loss": 0.6268,
+      "step": 6636
+    },
+    {
+      "epoch": 0.8641155798516205,
+      "grad_norm": 2.863619565963745,
+      "learning_rate": 8.279947557336184e-06,
+      "loss": 0.6373,
+      "step": 6639
+    },
+    {
+      "epoch": 0.8645060523233112,
+      "grad_norm": 2.654784679412842,
+      "learning_rate": 8.278369919400688e-06,
+      "loss": 0.6699,
+      "step": 6642
+    },
+    {
+      "epoch": 0.8648965247950019,
+      "grad_norm": 2.5628163814544678,
+      "learning_rate": 8.276791708740865e-06,
+      "loss": 0.7191,
+      "step": 6645
+    },
+    {
+      "epoch": 0.8652869972666927,
+      "grad_norm": 2.7298977375030518,
+      "learning_rate": 8.275212925632424e-06,
+      "loss": 0.6759,
+      "step": 6648
+    },
+    {
+      "epoch": 0.8656774697383834,
+      "grad_norm": 2.9911298751831055,
+      "learning_rate": 8.273633570351175e-06,
+      "loss": 0.6965,
+      "step": 6651
+    },
+    {
+      "epoch": 0.8660679422100742,
+      "grad_norm": 2.741440534591675,
+      "learning_rate": 8.272053643173028e-06,
+      "loss": 0.5874,
+      "step": 6654
+    },
+    {
+      "epoch": 0.866458414681765,
+      "grad_norm": 2.3754491806030273,
+      "learning_rate": 8.270473144373992e-06,
+      "loss": 0.7394,
+      "step": 6657
+    },
+    {
+      "epoch": 0.8668488871534556,
+      "grad_norm": 2.6968066692352295,
+      "learning_rate": 8.268892074230179e-06,
+      "loss": 0.6542,
+      "step": 6660
+    },
+    {
+      "epoch": 0.8672393596251464,
+      "grad_norm": 2.947608709335327,
+      "learning_rate": 8.267310433017795e-06,
+      "loss": 0.6674,
+      "step": 6663
+    },
+    {
+      "epoch": 0.8676298320968372,
+      "grad_norm": 2.4192142486572266,
+      "learning_rate": 8.265728221013154e-06,
+      "loss": 0.6469,
+      "step": 6666
+    },
+    {
+      "epoch": 0.868020304568528,
+      "grad_norm": 2.530045986175537,
+      "learning_rate": 8.264145438492664e-06,
+      "loss": 0.735,
+      "step": 6669
+    },
+    {
+      "epoch": 0.8684107770402186,
+      "grad_norm": 3.38845157623291,
+      "learning_rate": 8.26256208573283e-06,
+      "loss": 0.6779,
+      "step": 6672
+    },
+    {
+      "epoch": 0.8688012495119094,
+      "grad_norm": 2.9682278633117676,
+      "learning_rate": 8.260978163010265e-06,
+      "loss": 0.6009,
+      "step": 6675
+    },
+    {
+      "epoch": 0.8691917219836002,
+      "grad_norm": 2.4228317737579346,
+      "learning_rate": 8.259393670601673e-06,
+      "loss": 0.6627,
+      "step": 6678
+    },
+    {
+      "epoch": 0.869582194455291,
+      "grad_norm": 2.3679723739624023,
+      "learning_rate": 8.257808608783864e-06,
+      "loss": 0.6578,
+      "step": 6681
+    },
+    {
+      "epoch": 0.8699726669269816,
+      "grad_norm": 2.9238123893737793,
+      "learning_rate": 8.256222977833746e-06,
+      "loss": 0.6094,
+      "step": 6684
+    },
+    {
+      "epoch": 0.8703631393986724,
+      "grad_norm": 3.277010440826416,
+      "learning_rate": 8.254636778028321e-06,
+      "loss": 0.5665,
+      "step": 6687
+    },
+    {
+      "epoch": 0.8707536118703632,
+      "grad_norm": 2.598862648010254,
+      "learning_rate": 8.2530500096447e-06,
+      "loss": 0.6638,
+      "step": 6690
+    },
+    {
+      "epoch": 0.8711440843420539,
+      "grad_norm": 2.5538406372070312,
+      "learning_rate": 8.251462672960087e-06,
+      "loss": 0.7227,
+      "step": 6693
+    },
+    {
+      "epoch": 0.8715345568137446,
+      "grad_norm": 2.3932902812957764,
+      "learning_rate": 8.249874768251783e-06,
+      "loss": 0.6756,
+      "step": 6696
+    },
+    {
+      "epoch": 0.8719250292854354,
+      "grad_norm": 2.3504838943481445,
+      "learning_rate": 8.248286295797194e-06,
+      "loss": 0.671,
+      "step": 6699
+    },
+    {
+      "epoch": 0.8723155017571261,
+      "grad_norm": 3.473229169845581,
+      "learning_rate": 8.246697255873822e-06,
+      "loss": 0.6147,
+      "step": 6702
+    },
+    {
+      "epoch": 0.8727059742288169,
+      "grad_norm": 2.742168426513672,
+      "learning_rate": 8.24510764875927e-06,
+      "loss": 0.6286,
+      "step": 6705
+    },
+    {
+      "epoch": 0.8730964467005076,
+      "grad_norm": 2.26391863822937,
+      "learning_rate": 8.243517474731238e-06,
+      "loss": 0.5674,
+      "step": 6708
+    },
+    {
+      "epoch": 0.8734869191721983,
+      "grad_norm": 2.6377511024475098,
+      "learning_rate": 8.241926734067528e-06,
+      "loss": 0.731,
+      "step": 6711
+    },
+    {
+      "epoch": 0.8738773916438891,
+      "grad_norm": 2.8868167400360107,
+      "learning_rate": 8.240335427046037e-06,
+      "loss": 0.7232,
+      "step": 6714
+    },
+    {
+      "epoch": 0.8742678641155799,
+      "grad_norm": 2.5035762786865234,
+      "learning_rate": 8.238743553944762e-06,
+      "loss": 0.7436,
+      "step": 6717
+    },
+    {
+      "epoch": 0.8746583365872705,
+      "grad_norm": 4.231232643127441,
+      "learning_rate": 8.237151115041803e-06,
+      "loss": 0.6408,
+      "step": 6720
+    },
+    {
+      "epoch": 0.8750488090589613,
+      "grad_norm": 3.021878719329834,
+      "learning_rate": 8.235558110615354e-06,
+      "loss": 0.6811,
+      "step": 6723
+    },
+    {
+      "epoch": 0.8754392815306521,
+      "grad_norm": 3.348742961883545,
+      "learning_rate": 8.233964540943708e-06,
+      "loss": 0.7516,
+      "step": 6726
+    },
+    {
+      "epoch": 0.8758297540023429,
+      "grad_norm": 2.2513198852539062,
+      "learning_rate": 8.232370406305263e-06,
+      "loss": 0.6115,
+      "step": 6729
+    },
+    {
+      "epoch": 0.8762202264740335,
+      "grad_norm": 2.705314874649048,
+      "learning_rate": 8.230775706978507e-06,
+      "loss": 0.6867,
+      "step": 6732
+    },
+    {
+      "epoch": 0.8766106989457243,
+      "grad_norm": 3.001579523086548,
+      "learning_rate": 8.22918044324203e-06,
+      "loss": 0.6006,
+      "step": 6735
+    },
+    {
+      "epoch": 0.8770011714174151,
+      "grad_norm": 2.306036949157715,
+      "learning_rate": 8.227584615374524e-06,
+      "loss": 0.5314,
+      "step": 6738
+    },
+    {
+      "epoch": 0.8773916438891058,
+      "grad_norm": 2.6267998218536377,
+      "learning_rate": 8.225988223654775e-06,
+      "loss": 0.7156,
+      "step": 6741
+    },
+    {
+      "epoch": 0.8777821163607966,
+      "grad_norm": 2.350863456726074,
+      "learning_rate": 8.224391268361672e-06,
+      "loss": 0.5479,
+      "step": 6744
+    },
+    {
+      "epoch": 0.8781725888324873,
+      "grad_norm": 2.4865407943725586,
+      "learning_rate": 8.222793749774194e-06,
+      "loss": 0.5811,
+      "step": 6747
+    },
+    {
+      "epoch": 0.878563061304178,
+      "grad_norm": 3.3585386276245117,
+      "learning_rate": 8.221195668171429e-06,
+      "loss": 0.6603,
+      "step": 6750
+    },
+    {
+      "epoch": 0.8789535337758688,
+      "grad_norm": 4.48010778427124,
+      "learning_rate": 8.219597023832558e-06,
+      "loss": 0.6756,
+      "step": 6753
+    },
+    {
+      "epoch": 0.8793440062475596,
+      "grad_norm": 2.635272979736328,
+      "learning_rate": 8.21799781703686e-06,
+      "loss": 0.7552,
+      "step": 6756
+    },
+    {
+      "epoch": 0.8797344787192503,
+      "grad_norm": 2.7308220863342285,
+      "learning_rate": 8.216398048063712e-06,
+      "loss": 0.6473,
+      "step": 6759
+    },
+    {
+      "epoch": 0.880124951190941,
+      "grad_norm": 3.390993118286133,
+      "learning_rate": 8.214797717192591e-06,
+      "loss": 0.7474,
+      "step": 6762
+    },
+    {
+      "epoch": 0.8805154236626318,
+      "grad_norm": 2.1488442420959473,
+      "learning_rate": 8.213196824703074e-06,
+      "loss": 0.5481,
+      "step": 6765
+    },
+    {
+      "epoch": 0.8809058961343226,
+      "grad_norm": 2.6741676330566406,
+      "learning_rate": 8.21159537087483e-06,
+      "loss": 0.6344,
+      "step": 6768
+    },
+    {
+      "epoch": 0.8812963686060132,
+      "grad_norm": 2.426910877227783,
+      "learning_rate": 8.20999335598763e-06,
+      "loss": 0.6935,
+      "step": 6771
+    },
+    {
+      "epoch": 0.881686841077704,
+      "grad_norm": 2.539985179901123,
+      "learning_rate": 8.208390780321344e-06,
+      "loss": 0.6417,
+      "step": 6774
+    },
+    {
+      "epoch": 0.8820773135493948,
+      "grad_norm": 3.312417507171631,
+      "learning_rate": 8.20678764415594e-06,
+      "loss": 0.6397,
+      "step": 6777
+    },
+    {
+      "epoch": 0.8824677860210856,
+      "grad_norm": 2.3949477672576904,
+      "learning_rate": 8.205183947771478e-06,
+      "loss": 0.6372,
+      "step": 6780
+    },
+    {
+      "epoch": 0.8828582584927762,
+      "grad_norm": 2.6882550716400146,
+      "learning_rate": 8.203579691448124e-06,
+      "loss": 0.7021,
+      "step": 6783
+    },
+    {
+      "epoch": 0.883248730964467,
+      "grad_norm": 2.244438886642456,
+      "learning_rate": 8.201974875466138e-06,
+      "loss": 0.6369,
+      "step": 6786
+    },
+    {
+      "epoch": 0.8836392034361578,
+      "grad_norm": 2.579512357711792,
+      "learning_rate": 8.200369500105876e-06,
+      "loss": 0.6198,
+      "step": 6789
+    },
+    {
+      "epoch": 0.8840296759078485,
+      "grad_norm": 2.45475172996521,
+      "learning_rate": 8.198763565647796e-06,
+      "loss": 0.6765,
+      "step": 6792
+    },
+    {
+      "epoch": 0.8844201483795392,
+      "grad_norm": 2.1796491146087646,
+      "learning_rate": 8.19715707237245e-06,
+      "loss": 0.5684,
+      "step": 6795
+    },
+    {
+      "epoch": 0.88481062085123,
+      "grad_norm": 3.6202688217163086,
+      "learning_rate": 8.195550020560488e-06,
+      "loss": 0.6061,
+      "step": 6798
+    },
+    {
+      "epoch": 0.8852010933229207,
+      "grad_norm": 2.273062229156494,
+      "learning_rate": 8.193942410492662e-06,
+      "loss": 0.5358,
+      "step": 6801
+    },
+    {
+      "epoch": 0.8855915657946115,
+      "grad_norm": 2.308729410171509,
+      "learning_rate": 8.192334242449816e-06,
+      "loss": 0.5674,
+      "step": 6804
+    },
+    {
+      "epoch": 0.8859820382663022,
+      "grad_norm": 2.949068784713745,
+      "learning_rate": 8.190725516712893e-06,
+      "loss": 0.6682,
+      "step": 6807
+    },
+    {
+      "epoch": 0.886372510737993,
+      "grad_norm": 2.661355972290039,
+      "learning_rate": 8.189116233562933e-06,
+      "loss": 0.7175,
+      "step": 6810
+    },
+    {
+      "epoch": 0.8867629832096837,
+      "grad_norm": 3.1999619007110596,
+      "learning_rate": 8.187506393281076e-06,
+      "loss": 0.6816,
+      "step": 6813
+    },
+    {
+      "epoch": 0.8871534556813745,
+      "grad_norm": 3.3736467361450195,
+      "learning_rate": 8.185895996148558e-06,
+      "loss": 0.6235,
+      "step": 6816
+    },
+    {
+      "epoch": 0.8875439281530652,
+      "grad_norm": 2.3668558597564697,
+      "learning_rate": 8.184285042446713e-06,
+      "loss": 0.7111,
+      "step": 6819
+    },
+    {
+      "epoch": 0.8879344006247559,
+      "grad_norm": 3.4571516513824463,
+      "learning_rate": 8.18267353245697e-06,
+      "loss": 0.6578,
+      "step": 6822
+    },
+    {
+      "epoch": 0.8883248730964467,
+      "grad_norm": 2.8542165756225586,
+      "learning_rate": 8.181061466460856e-06,
+      "loss": 0.6717,
+      "step": 6825
+    },
+    {
+      "epoch": 0.8887153455681375,
+      "grad_norm": 2.5274338722229004,
+      "learning_rate": 8.179448844739995e-06,
+      "loss": 0.7508,
+      "step": 6828
+    },
+    {
+      "epoch": 0.8891058180398282,
+      "grad_norm": 2.284775733947754,
+      "learning_rate": 8.177835667576108e-06,
+      "loss": 0.6448,
+      "step": 6831
+    },
+    {
+      "epoch": 0.8894962905115189,
+      "grad_norm": 2.375109910964966,
+      "learning_rate": 8.176221935251016e-06,
+      "loss": 0.6777,
+      "step": 6834
+    },
+    {
+      "epoch": 0.8898867629832097,
+      "grad_norm": 2.5060408115386963,
+      "learning_rate": 8.174607648046635e-06,
+      "loss": 0.7086,
+      "step": 6837
+    },
+    {
+      "epoch": 0.8902772354549005,
+      "grad_norm": 3.7444376945495605,
+      "learning_rate": 8.172992806244976e-06,
+      "loss": 0.6429,
+      "step": 6840
+    },
+    {
+      "epoch": 0.8906677079265912,
+      "grad_norm": 2.7690014839172363,
+      "learning_rate": 8.171377410128149e-06,
+      "loss": 0.6279,
+      "step": 6843
+    },
+    {
+      "epoch": 0.8910581803982819,
+      "grad_norm": 3.0757744312286377,
+      "learning_rate": 8.169761459978358e-06,
+      "loss": 0.7566,
+      "step": 6846
+    },
+    {
+      "epoch": 0.8914486528699727,
+      "grad_norm": 2.608480930328369,
+      "learning_rate": 8.16814495607791e-06,
+      "loss": 0.6398,
+      "step": 6849
+    },
+    {
+      "epoch": 0.8918391253416634,
+      "grad_norm": 2.6468796730041504,
+      "learning_rate": 8.166527898709202e-06,
+      "loss": 0.6788,
+      "step": 6852
+    },
+    {
+      "epoch": 0.8922295978133542,
+      "grad_norm": 2.874540328979492,
+      "learning_rate": 8.164910288154733e-06,
+      "loss": 0.6425,
+      "step": 6855
+    },
+    {
+      "epoch": 0.8926200702850449,
+      "grad_norm": 2.4776840209960938,
+      "learning_rate": 8.163292124697094e-06,
+      "loss": 0.7579,
+      "step": 6858
+    },
+    {
+      "epoch": 0.8930105427567356,
+      "grad_norm": 2.2084155082702637,
+      "learning_rate": 8.161673408618975e-06,
+      "loss": 0.6798,
+      "step": 6861
+    },
+    {
+      "epoch": 0.8934010152284264,
+      "grad_norm": 2.4386301040649414,
+      "learning_rate": 8.160054140203163e-06,
+      "loss": 0.6953,
+      "step": 6864
+    },
+    {
+      "epoch": 0.8937914877001172,
+      "grad_norm": 2.9873063564300537,
+      "learning_rate": 8.15843431973254e-06,
+      "loss": 0.6849,
+      "step": 6867
+    },
+    {
+      "epoch": 0.8941819601718078,
+      "grad_norm": 2.3077900409698486,
+      "learning_rate": 8.156813947490086e-06,
+      "loss": 0.5833,
+      "step": 6870
+    },
+    {
+      "epoch": 0.8945724326434986,
+      "grad_norm": 3.516432046890259,
+      "learning_rate": 8.155193023758876e-06,
+      "loss": 0.6973,
+      "step": 6873
+    },
+    {
+      "epoch": 0.8949629051151894,
+      "grad_norm": 2.4575483798980713,
+      "learning_rate": 8.153571548822083e-06,
+      "loss": 0.6842,
+      "step": 6876
+    },
+    {
+      "epoch": 0.8953533775868802,
+      "grad_norm": 2.8039913177490234,
+      "learning_rate": 8.151949522962975e-06,
+      "loss": 0.6416,
+      "step": 6879
+    },
+    {
+      "epoch": 0.8957438500585708,
+      "grad_norm": 3.1583502292633057,
+      "learning_rate": 8.150326946464913e-06,
+      "loss": 0.6764,
+      "step": 6882
+    },
+    {
+      "epoch": 0.8961343225302616,
+      "grad_norm": 2.762176752090454,
+      "learning_rate": 8.148703819611364e-06,
+      "loss": 0.6455,
+      "step": 6885
+    },
+    {
+      "epoch": 0.8965247950019524,
+      "grad_norm": 3.457850694656372,
+      "learning_rate": 8.147080142685882e-06,
+      "loss": 0.6589,
+      "step": 6888
+    },
+    {
+      "epoch": 0.8969152674736431,
+      "grad_norm": 3.8411946296691895,
+      "learning_rate": 8.145455915972117e-06,
+      "loss": 0.7218,
+      "step": 6891
+    },
+    {
+      "epoch": 0.8973057399453338,
+      "grad_norm": 2.2801125049591064,
+      "learning_rate": 8.143831139753822e-06,
+      "loss": 0.6221,
+      "step": 6894
+    },
+    {
+      "epoch": 0.8976962124170246,
+      "grad_norm": 2.8186426162719727,
+      "learning_rate": 8.14220581431484e-06,
+      "loss": 0.7055,
+      "step": 6897
+    },
+    {
+      "epoch": 0.8980866848887153,
+      "grad_norm": 2.3961057662963867,
+      "learning_rate": 8.140579939939113e-06,
+      "loss": 0.6705,
+      "step": 6900
+    },
+    {
+      "epoch": 0.8984771573604061,
+      "grad_norm": 2.616764783859253,
+      "learning_rate": 8.138953516910676e-06,
+      "loss": 0.5349,
+      "step": 6903
+    },
+    {
+      "epoch": 0.8988676298320968,
+      "grad_norm": 2.015042543411255,
+      "learning_rate": 8.137326545513664e-06,
+      "loss": 0.535,
+      "step": 6906
+    },
+    {
+      "epoch": 0.8992581023037876,
+      "grad_norm": 2.598292112350464,
+      "learning_rate": 8.135699026032305e-06,
+      "loss": 0.7247,
+      "step": 6909
+    },
+    {
+      "epoch": 0.8996485747754783,
+      "grad_norm": 2.3289687633514404,
+      "learning_rate": 8.134070958750923e-06,
+      "loss": 0.6876,
+      "step": 6912
+    },
+    {
+      "epoch": 0.9000390472471691,
+      "grad_norm": 2.219310760498047,
+      "learning_rate": 8.132442343953937e-06,
+      "loss": 0.6288,
+      "step": 6915
+    },
+    {
+      "epoch": 0.9004295197188599,
+      "grad_norm": 2.5258898735046387,
+      "learning_rate": 8.130813181925862e-06,
+      "loss": 0.6711,
+      "step": 6918
+    },
+    {
+      "epoch": 0.9008199921905505,
+      "grad_norm": 2.4190750122070312,
+      "learning_rate": 8.129183472951312e-06,
+      "loss": 0.7272,
+      "step": 6921
+    },
+    {
+      "epoch": 0.9012104646622413,
+      "grad_norm": 2.566995620727539,
+      "learning_rate": 8.127553217314991e-06,
+      "loss": 0.735,
+      "step": 6924
+    },
+    {
+      "epoch": 0.9016009371339321,
+      "grad_norm": 2.446164608001709,
+      "learning_rate": 8.125922415301704e-06,
+      "loss": 0.6453,
+      "step": 6927
+    },
+    {
+      "epoch": 0.9019914096056229,
+      "grad_norm": 2.655388593673706,
+      "learning_rate": 8.124291067196347e-06,
+      "loss": 0.5551,
+      "step": 6930
+    },
+    {
+      "epoch": 0.9023818820773135,
+      "grad_norm": 2.4122257232666016,
+      "learning_rate": 8.12265917328391e-06,
+      "loss": 0.6901,
+      "step": 6933
+    },
+    {
+      "epoch": 0.9027723545490043,
+      "grad_norm": 2.494718551635742,
+      "learning_rate": 8.121026733849486e-06,
+      "loss": 0.6336,
+      "step": 6936
+    },
+    {
+      "epoch": 0.9031628270206951,
+      "grad_norm": 2.373518466949463,
+      "learning_rate": 8.119393749178258e-06,
+      "loss": 0.6479,
+      "step": 6939
+    },
+    {
+      "epoch": 0.9035532994923858,
+      "grad_norm": 2.2196645736694336,
+      "learning_rate": 8.117760219555505e-06,
+      "loss": 0.5689,
+      "step": 6942
+    },
+    {
+      "epoch": 0.9039437719640765,
+      "grad_norm": 3.160625457763672,
+      "learning_rate": 8.116126145266599e-06,
+      "loss": 0.627,
+      "step": 6945
+    },
+    {
+      "epoch": 0.9043342444357673,
+      "grad_norm": 2.553009271621704,
+      "learning_rate": 8.114491526597012e-06,
+      "loss": 0.5639,
+      "step": 6948
+    },
+    {
+      "epoch": 0.904724716907458,
+      "grad_norm": 2.1808202266693115,
+      "learning_rate": 8.112856363832307e-06,
+      "loss": 0.5542,
+      "step": 6951
+    },
+    {
+      "epoch": 0.9051151893791488,
+      "grad_norm": 2.4375030994415283,
+      "learning_rate": 8.111220657258144e-06,
+      "loss": 0.5841,
+      "step": 6954
+    },
+    {
+      "epoch": 0.9055056618508395,
+      "grad_norm": 2.3346710205078125,
+      "learning_rate": 8.109584407160277e-06,
+      "loss": 0.6731,
+      "step": 6957
+    },
+    {
+      "epoch": 0.9058961343225302,
+      "grad_norm": 2.4284756183624268,
+      "learning_rate": 8.107947613824554e-06,
+      "loss": 0.6966,
+      "step": 6960
+    },
+    {
+      "epoch": 0.906286606794221,
+      "grad_norm": 3.4861011505126953,
+      "learning_rate": 8.106310277536921e-06,
+      "loss": 0.6403,
+      "step": 6963
+    },
+    {
+      "epoch": 0.9066770792659118,
+      "grad_norm": 3.8682639598846436,
+      "learning_rate": 8.104672398583419e-06,
+      "loss": 0.6438,
+      "step": 6966
+    },
+    {
+      "epoch": 0.9070675517376025,
+      "grad_norm": 2.232022523880005,
+      "learning_rate": 8.10303397725018e-06,
+      "loss": 0.545,
+      "step": 6969
+    },
+    {
+      "epoch": 0.9074580242092932,
+      "grad_norm": 3.7557497024536133,
+      "learning_rate": 8.101395013823433e-06,
+      "loss": 0.7008,
+      "step": 6972
+    },
+    {
+      "epoch": 0.907848496680984,
+      "grad_norm": 2.4933996200561523,
+      "learning_rate": 8.099755508589502e-06,
+      "loss": 0.6323,
+      "step": 6975
+    },
+    {
+      "epoch": 0.9082389691526748,
+      "grad_norm": 2.4882712364196777,
+      "learning_rate": 8.098115461834803e-06,
+      "loss": 0.6243,
+      "step": 6978
+    },
+    {
+      "epoch": 0.9086294416243654,
+      "grad_norm": 2.3607735633850098,
+      "learning_rate": 8.096474873845851e-06,
+      "loss": 0.6152,
+      "step": 6981
+    },
+    {
+      "epoch": 0.9090199140960562,
+      "grad_norm": 3.5572595596313477,
+      "learning_rate": 8.094833744909252e-06,
+      "loss": 0.7645,
+      "step": 6984
+    },
+    {
+      "epoch": 0.909410386567747,
+      "grad_norm": 3.0269618034362793,
+      "learning_rate": 8.09319207531171e-06,
+      "loss": 0.6273,
+      "step": 6987
+    },
+    {
+      "epoch": 0.9098008590394377,
+      "grad_norm": 3.1948349475860596,
+      "learning_rate": 8.091549865340019e-06,
+      "loss": 0.6008,
+      "step": 6990
+    },
+    {
+      "epoch": 0.9101913315111285,
+      "grad_norm": 2.419280529022217,
+      "learning_rate": 8.08990711528107e-06,
+      "loss": 0.6304,
+      "step": 6993
+    },
+    {
+      "epoch": 0.9105818039828192,
+      "grad_norm": 2.2587809562683105,
+      "learning_rate": 8.088263825421847e-06,
+      "loss": 0.6211,
+      "step": 6996
+    },
+    {
+      "epoch": 0.91097227645451,
+      "grad_norm": 2.709536552429199,
+      "learning_rate": 8.086619996049431e-06,
+      "loss": 0.6343,
+      "step": 6999
+    },
+    {
+      "epoch": 0.9113627489262007,
+      "grad_norm": 5.089230060577393,
+      "learning_rate": 8.084975627450995e-06,
+      "loss": 0.6053,
+      "step": 7002
+    },
+    {
+      "epoch": 0.9117532213978915,
+      "grad_norm": 2.689100742340088,
+      "learning_rate": 8.083330719913808e-06,
+      "loss": 0.668,
+      "step": 7005
+    },
+    {
+      "epoch": 0.9121436938695822,
+      "grad_norm": 2.3222506046295166,
+      "learning_rate": 8.08168527372523e-06,
+      "loss": 0.5972,
+      "step": 7008
+    },
+    {
+      "epoch": 0.9125341663412729,
+      "grad_norm": 2.3201560974121094,
+      "learning_rate": 8.080039289172717e-06,
+      "loss": 0.6853,
+      "step": 7011
+    },
+    {
+      "epoch": 0.9129246388129637,
+      "grad_norm": 2.8677849769592285,
+      "learning_rate": 8.078392766543821e-06,
+      "loss": 0.5319,
+      "step": 7014
+    },
+    {
+      "epoch": 0.9133151112846545,
+      "grad_norm": 2.3285553455352783,
+      "learning_rate": 8.076745706126184e-06,
+      "loss": 0.6937,
+      "step": 7017
+    },
+    {
+      "epoch": 0.9137055837563451,
+      "grad_norm": 2.447075366973877,
+      "learning_rate": 8.075098108207544e-06,
+      "loss": 0.7326,
+      "step": 7020
+    },
+    {
+      "epoch": 0.9140960562280359,
+      "grad_norm": 3.435882091522217,
+      "learning_rate": 8.073449973075733e-06,
+      "loss": 0.6119,
+      "step": 7023
+    },
+    {
+      "epoch": 0.9144865286997267,
+      "grad_norm": 2.431121587753296,
+      "learning_rate": 8.071801301018678e-06,
+      "loss": 0.6601,
+      "step": 7026
+    },
+    {
+      "epoch": 0.9148770011714175,
+      "grad_norm": 3.276074171066284,
+      "learning_rate": 8.070152092324399e-06,
+      "loss": 0.6196,
+      "step": 7029
+    },
+    {
+      "epoch": 0.9152674736431081,
+      "grad_norm": 2.349234104156494,
+      "learning_rate": 8.068502347281006e-06,
+      "loss": 0.5846,
+      "step": 7032
+    },
+    {
+      "epoch": 0.9156579461147989,
+      "grad_norm": 2.1926095485687256,
+      "learning_rate": 8.06685206617671e-06,
+      "loss": 0.7039,
+      "step": 7035
+    },
+    {
+      "epoch": 0.9160484185864897,
+      "grad_norm": 2.8905913829803467,
+      "learning_rate": 8.06520124929981e-06,
+      "loss": 0.6995,
+      "step": 7038
+    },
+    {
+      "epoch": 0.9164388910581804,
+      "grad_norm": 2.275421380996704,
+      "learning_rate": 8.063549896938698e-06,
+      "loss": 0.5875,
+      "step": 7041
+    },
+    {
+      "epoch": 0.9168293635298711,
+      "grad_norm": 2.2735865116119385,
+      "learning_rate": 8.061898009381865e-06,
+      "loss": 0.6204,
+      "step": 7044
+    },
+    {
+      "epoch": 0.9172198360015619,
+      "grad_norm": 2.518444299697876,
+      "learning_rate": 8.06024558691789e-06,
+      "loss": 0.6785,
+      "step": 7047
+    },
+    {
+      "epoch": 0.9176103084732526,
+      "grad_norm": 2.3565497398376465,
+      "learning_rate": 8.05859262983545e-06,
+      "loss": 0.6093,
+      "step": 7050
+    },
+    {
+      "epoch": 0.9180007809449434,
+      "grad_norm": 2.3102149963378906,
+      "learning_rate": 8.056939138423313e-06,
+      "loss": 0.701,
+      "step": 7053
+    },
+    {
+      "epoch": 0.9183912534166341,
+      "grad_norm": 3.3671510219573975,
+      "learning_rate": 8.055285112970337e-06,
+      "loss": 0.6835,
+      "step": 7056
+    },
+    {
+      "epoch": 0.9187817258883249,
+      "grad_norm": 2.4449825286865234,
+      "learning_rate": 8.05363055376548e-06,
+      "loss": 0.6697,
+      "step": 7059
+    },
+    {
+      "epoch": 0.9191721983600156,
+      "grad_norm": 2.696711540222168,
+      "learning_rate": 8.051975461097789e-06,
+      "loss": 0.6039,
+      "step": 7062
+    },
+    {
+      "epoch": 0.9195626708317064,
+      "grad_norm": 2.4236817359924316,
+      "learning_rate": 8.050319835256406e-06,
+      "loss": 0.6511,
+      "step": 7065
+    },
+    {
+      "epoch": 0.9199531433033971,
+      "grad_norm": 2.9492392539978027,
+      "learning_rate": 8.048663676530563e-06,
+      "loss": 0.6339,
+      "step": 7068
+    },
+    {
+      "epoch": 0.9203436157750878,
+      "grad_norm": 2.4083123207092285,
+      "learning_rate": 8.04700698520959e-06,
+      "loss": 0.6088,
+      "step": 7071
+    },
+    {
+      "epoch": 0.9207340882467786,
+      "grad_norm": 2.4848201274871826,
+      "learning_rate": 8.045349761582908e-06,
+      "loss": 0.6274,
+      "step": 7074
+    },
+    {
+      "epoch": 0.9211245607184694,
+      "grad_norm": 6.317074298858643,
+      "learning_rate": 8.043692005940029e-06,
+      "loss": 0.6443,
+      "step": 7077
+    },
+    {
+      "epoch": 0.9215150331901601,
+      "grad_norm": 2.408294916152954,
+      "learning_rate": 8.042033718570559e-06,
+      "loss": 0.694,
+      "step": 7080
+    },
+    {
+      "epoch": 0.9219055056618508,
+      "grad_norm": 2.3996033668518066,
+      "learning_rate": 8.040374899764198e-06,
+      "loss": 0.6197,
+      "step": 7083
+    },
+    {
+      "epoch": 0.9222959781335416,
+      "grad_norm": 2.790170907974243,
+      "learning_rate": 8.038715549810737e-06,
+      "loss": 0.575,
+      "step": 7086
+    },
+    {
+      "epoch": 0.9226864506052324,
+      "grad_norm": 2.455018997192383,
+      "learning_rate": 8.037055669000062e-06,
+      "loss": 0.6055,
+      "step": 7089
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 2.327902317047119,
+      "learning_rate": 8.035395257622151e-06,
+      "loss": 0.5594,
+      "step": 7092
+    },
+    {
+      "epoch": 0.9234673955486138,
+      "grad_norm": 3.6916446685791016,
+      "learning_rate": 8.033734315967074e-06,
+      "loss": 0.6818,
+      "step": 7095
+    },
+    {
+      "epoch": 0.9238578680203046,
+      "grad_norm": 2.5879135131835938,
+      "learning_rate": 8.032072844324995e-06,
+      "loss": 0.66,
+      "step": 7098
+    },
+    {
+      "epoch": 0.9242483404919953,
+      "grad_norm": 2.4714932441711426,
+      "learning_rate": 8.030410842986169e-06,
+      "loss": 0.6118,
+      "step": 7101
+    },
+    {
+      "epoch": 0.9246388129636861,
+      "grad_norm": 2.5215554237365723,
+      "learning_rate": 8.028748312240942e-06,
+      "loss": 0.7066,
+      "step": 7104
+    },
+    {
+      "epoch": 0.9250292854353768,
+      "grad_norm": 2.2146830558776855,
+      "learning_rate": 8.027085252379755e-06,
+      "loss": 0.5609,
+      "step": 7107
+    },
+    {
+      "epoch": 0.9254197579070675,
+      "grad_norm": 2.9621617794036865,
+      "learning_rate": 8.025421663693147e-06,
+      "loss": 0.6719,
+      "step": 7110
+    },
+    {
+      "epoch": 0.9258102303787583,
+      "grad_norm": 2.4687397480010986,
+      "learning_rate": 8.023757546471737e-06,
+      "loss": 0.634,
+      "step": 7113
+    },
+    {
+      "epoch": 0.9262007028504491,
+      "grad_norm": 2.19496488571167,
+      "learning_rate": 8.022092901006245e-06,
+      "loss": 0.6325,
+      "step": 7116
+    },
+    {
+      "epoch": 0.9265911753221397,
+      "grad_norm": 3.014942169189453,
+      "learning_rate": 8.020427727587479e-06,
+      "loss": 0.6796,
+      "step": 7119
+    },
+    {
+      "epoch": 0.9269816477938305,
+      "grad_norm": 2.605684995651245,
+      "learning_rate": 8.018762026506344e-06,
+      "loss": 0.6616,
+      "step": 7122
+    },
+    {
+      "epoch": 0.9273721202655213,
+      "grad_norm": 2.285712957382202,
+      "learning_rate": 8.017095798053834e-06,
+      "loss": 0.5857,
+      "step": 7125
+    },
+    {
+      "epoch": 0.9277625927372121,
+      "grad_norm": 2.385016918182373,
+      "learning_rate": 8.015429042521034e-06,
+      "loss": 0.6805,
+      "step": 7128
+    },
+    {
+      "epoch": 0.9281530652089027,
+      "grad_norm": 2.6234757900238037,
+      "learning_rate": 8.013761760199125e-06,
+      "loss": 0.6721,
+      "step": 7131
+    },
+    {
+      "epoch": 0.9285435376805935,
+      "grad_norm": 2.4633302688598633,
+      "learning_rate": 8.012093951379376e-06,
+      "loss": 0.6305,
+      "step": 7134
+    },
+    {
+      "epoch": 0.9289340101522843,
+      "grad_norm": 2.8097805976867676,
+      "learning_rate": 8.01042561635315e-06,
+      "loss": 0.613,
+      "step": 7137
+    },
+    {
+      "epoch": 0.929324482623975,
+      "grad_norm": 2.5998694896698,
+      "learning_rate": 8.008756755411902e-06,
+      "loss": 0.619,
+      "step": 7140
+    },
+    {
+      "epoch": 0.9297149550956657,
+      "grad_norm": 2.280726194381714,
+      "learning_rate": 8.007087368847178e-06,
+      "loss": 0.6404,
+      "step": 7143
+    },
+    {
+      "epoch": 0.9301054275673565,
+      "grad_norm": 2.546746015548706,
+      "learning_rate": 8.005417456950617e-06,
+      "loss": 0.6563,
+      "step": 7146
+    },
+    {
+      "epoch": 0.9304959000390473,
+      "grad_norm": 3.274017572402954,
+      "learning_rate": 8.003747020013948e-06,
+      "loss": 0.6148,
+      "step": 7149
+    },
+    {
+      "epoch": 0.930886372510738,
+      "grad_norm": 2.3492894172668457,
+      "learning_rate": 8.002076058328996e-06,
+      "loss": 0.5916,
+      "step": 7152
+    },
+    {
+      "epoch": 0.9312768449824287,
+      "grad_norm": 2.176417112350464,
+      "learning_rate": 8.00040457218767e-06,
+      "loss": 0.5756,
+      "step": 7155
+    },
+    {
+      "epoch": 0.9316673174541195,
+      "grad_norm": 2.4730896949768066,
+      "learning_rate": 7.998732561881976e-06,
+      "loss": 0.5959,
+      "step": 7158
+    },
+    {
+      "epoch": 0.9320577899258102,
+      "grad_norm": 2.558849573135376,
+      "learning_rate": 7.997060027704016e-06,
+      "loss": 0.6642,
+      "step": 7161
+    },
+    {
+      "epoch": 0.932448262397501,
+      "grad_norm": 2.7973668575286865,
+      "learning_rate": 7.99538696994597e-06,
+      "loss": 0.6965,
+      "step": 7164
+    },
+    {
+      "epoch": 0.9328387348691918,
+      "grad_norm": 2.7359659671783447,
+      "learning_rate": 7.993713388900124e-06,
+      "loss": 0.671,
+      "step": 7167
+    },
+    {
+      "epoch": 0.9332292073408824,
+      "grad_norm": 3.094665765762329,
+      "learning_rate": 7.992039284858846e-06,
+      "loss": 0.6593,
+      "step": 7170
+    },
+    {
+      "epoch": 0.9336196798125732,
+      "grad_norm": 2.6220576763153076,
+      "learning_rate": 7.990364658114599e-06,
+      "loss": 0.6067,
+      "step": 7173
+    },
+    {
+      "epoch": 0.934010152284264,
+      "grad_norm": 2.2661070823669434,
+      "learning_rate": 7.988689508959936e-06,
+      "loss": 0.6265,
+      "step": 7176
+    },
+    {
+      "epoch": 0.9344006247559548,
+      "grad_norm": 3.1826601028442383,
+      "learning_rate": 7.987013837687505e-06,
+      "loss": 0.6483,
+      "step": 7179
+    },
+    {
+      "epoch": 0.9347910972276454,
+      "grad_norm": 2.6422383785247803,
+      "learning_rate": 7.98533764459004e-06,
+      "loss": 0.6612,
+      "step": 7182
+    },
+    {
+      "epoch": 0.9351815696993362,
+      "grad_norm": 2.5566980838775635,
+      "learning_rate": 7.983660929960368e-06,
+      "loss": 0.6701,
+      "step": 7185
+    },
+    {
+      "epoch": 0.935572042171027,
+      "grad_norm": 2.446725368499756,
+      "learning_rate": 7.98198369409141e-06,
+      "loss": 0.62,
+      "step": 7188
+    },
+    {
+      "epoch": 0.9359625146427177,
+      "grad_norm": 2.3360819816589355,
+      "learning_rate": 7.980305937276172e-06,
+      "loss": 0.598,
+      "step": 7191
+    },
+    {
+      "epoch": 0.9363529871144084,
+      "grad_norm": 2.8677642345428467,
+      "learning_rate": 7.978627659807757e-06,
+      "loss": 0.6929,
+      "step": 7194
+    },
+    {
+      "epoch": 0.9367434595860992,
+      "grad_norm": 2.7148616313934326,
+      "learning_rate": 7.976948861979356e-06,
+      "loss": 0.626,
+      "step": 7197
+    },
+    {
+      "epoch": 0.9371339320577899,
+      "grad_norm": 2.4548072814941406,
+      "learning_rate": 7.975269544084251e-06,
+      "loss": 0.6934,
+      "step": 7200
+    },
+    {
+      "epoch": 0.9375244045294807,
+      "grad_norm": 3.422706365585327,
+      "learning_rate": 7.973589706415816e-06,
+      "loss": 0.666,
+      "step": 7203
+    },
+    {
+      "epoch": 0.9379148770011714,
+      "grad_norm": 2.493299722671509,
+      "learning_rate": 7.971909349267514e-06,
+      "loss": 0.7444,
+      "step": 7206
+    },
+    {
+      "epoch": 0.9383053494728621,
+      "grad_norm": 2.5340187549591064,
+      "learning_rate": 7.970228472932901e-06,
+      "loss": 0.6665,
+      "step": 7209
+    },
+    {
+      "epoch": 0.9386958219445529,
+      "grad_norm": 2.3306686878204346,
+      "learning_rate": 7.968547077705624e-06,
+      "loss": 0.6389,
+      "step": 7212
+    },
+    {
+      "epoch": 0.9390862944162437,
+      "grad_norm": 2.3347361087799072,
+      "learning_rate": 7.966865163879416e-06,
+      "loss": 0.6495,
+      "step": 7215
+    },
+    {
+      "epoch": 0.9394767668879344,
+      "grad_norm": 2.574287176132202,
+      "learning_rate": 7.965182731748104e-06,
+      "loss": 0.6285,
+      "step": 7218
+    },
+    {
+      "epoch": 0.9398672393596251,
+      "grad_norm": 2.4919655323028564,
+      "learning_rate": 7.96349978160561e-06,
+      "loss": 0.6683,
+      "step": 7221
+    },
+    {
+      "epoch": 0.9402577118313159,
+      "grad_norm": 2.245663642883301,
+      "learning_rate": 7.961816313745936e-06,
+      "loss": 0.5926,
+      "step": 7224
+    },
+    {
+      "epoch": 0.9406481843030067,
+      "grad_norm": 2.493183135986328,
+      "learning_rate": 7.960132328463184e-06,
+      "loss": 0.6197,
+      "step": 7227
+    },
+    {
+      "epoch": 0.9410386567746973,
+      "grad_norm": 2.5422475337982178,
+      "learning_rate": 7.958447826051538e-06,
+      "loss": 0.6305,
+      "step": 7230
+    },
+    {
+      "epoch": 0.9414291292463881,
+      "grad_norm": 2.432318925857544,
+      "learning_rate": 7.956762806805285e-06,
+      "loss": 0.7437,
+      "step": 7233
+    },
+    {
+      "epoch": 0.9418196017180789,
+      "grad_norm": 2.516645669937134,
+      "learning_rate": 7.955077271018788e-06,
+      "loss": 0.6369,
+      "step": 7236
+    },
+    {
+      "epoch": 0.9422100741897697,
+      "grad_norm": 2.9639275074005127,
+      "learning_rate": 7.953391218986507e-06,
+      "loss": 0.5996,
+      "step": 7239
+    },
+    {
+      "epoch": 0.9426005466614603,
+      "grad_norm": 2.503748655319214,
+      "learning_rate": 7.951704651002994e-06,
+      "loss": 0.6876,
+      "step": 7242
+    },
+    {
+      "epoch": 0.9429910191331511,
+      "grad_norm": 2.3590073585510254,
+      "learning_rate": 7.950017567362888e-06,
+      "loss": 0.6315,
+      "step": 7245
+    },
+    {
+      "epoch": 0.9433814916048419,
+      "grad_norm": 2.718970775604248,
+      "learning_rate": 7.948329968360919e-06,
+      "loss": 0.7917,
+      "step": 7248
+    },
+    {
+      "epoch": 0.9437719640765326,
+      "grad_norm": 3.0263495445251465,
+      "learning_rate": 7.946641854291908e-06,
+      "loss": 0.62,
+      "step": 7251
+    },
+    {
+      "epoch": 0.9441624365482234,
+      "grad_norm": 2.2088472843170166,
+      "learning_rate": 7.944953225450762e-06,
+      "loss": 0.5835,
+      "step": 7254
+    },
+    {
+      "epoch": 0.9445529090199141,
+      "grad_norm": 2.943924903869629,
+      "learning_rate": 7.943264082132484e-06,
+      "loss": 0.6698,
+      "step": 7257
+    },
+    {
+      "epoch": 0.9449433814916048,
+      "grad_norm": 3.520203113555908,
+      "learning_rate": 7.941574424632161e-06,
+      "loss": 0.6166,
+      "step": 7260
+    },
+    {
+      "epoch": 0.9453338539632956,
+      "grad_norm": 2.5051796436309814,
+      "learning_rate": 7.939884253244977e-06,
+      "loss": 0.5959,
+      "step": 7263
+    },
+    {
+      "epoch": 0.9457243264349864,
+      "grad_norm": 2.657858371734619,
+      "learning_rate": 7.938193568266195e-06,
+      "loss": 0.6384,
+      "step": 7266
+    },
+    {
+      "epoch": 0.946114798906677,
+      "grad_norm": 2.747825860977173,
+      "learning_rate": 7.93650236999118e-06,
+      "loss": 0.6422,
+      "step": 7269
+    },
+    {
+      "epoch": 0.9465052713783678,
+      "grad_norm": 2.390871524810791,
+      "learning_rate": 7.93481065871538e-06,
+      "loss": 0.6069,
+      "step": 7272
+    },
+    {
+      "epoch": 0.9468957438500586,
+      "grad_norm": 4.321902275085449,
+      "learning_rate": 7.933118434734329e-06,
+      "loss": 0.6149,
+      "step": 7275
+    },
+    {
+      "epoch": 0.9472862163217494,
+      "grad_norm": 2.5380969047546387,
+      "learning_rate": 7.931425698343657e-06,
+      "loss": 0.6597,
+      "step": 7278
+    },
+    {
+      "epoch": 0.94767668879344,
+      "grad_norm": 2.3168387413024902,
+      "learning_rate": 7.929732449839085e-06,
+      "loss": 0.6158,
+      "step": 7281
+    },
+    {
+      "epoch": 0.9480671612651308,
+      "grad_norm": 2.2340736389160156,
+      "learning_rate": 7.928038689516417e-06,
+      "loss": 0.546,
+      "step": 7284
+    },
+    {
+      "epoch": 0.9484576337368216,
+      "grad_norm": 2.3683393001556396,
+      "learning_rate": 7.92634441767155e-06,
+      "loss": 0.5225,
+      "step": 7287
+    },
+    {
+      "epoch": 0.9488481062085123,
+      "grad_norm": 3.3821890354156494,
+      "learning_rate": 7.924649634600468e-06,
+      "loss": 0.7159,
+      "step": 7290
+    },
+    {
+      "epoch": 0.949238578680203,
+      "grad_norm": 2.016606569290161,
+      "learning_rate": 7.922954340599247e-06,
+      "loss": 0.578,
+      "step": 7293
+    },
+    {
+      "epoch": 0.9496290511518938,
+      "grad_norm": 3.1906471252441406,
+      "learning_rate": 7.921258535964051e-06,
+      "loss": 0.7206,
+      "step": 7296
+    },
+    {
+      "epoch": 0.9500195236235845,
+      "grad_norm": 2.839524507522583,
+      "learning_rate": 7.919562220991137e-06,
+      "loss": 0.6458,
+      "step": 7299
+    },
+    {
+      "epoch": 0.9504099960952753,
+      "grad_norm": 2.620429039001465,
+      "learning_rate": 7.917865395976844e-06,
+      "loss": 0.6736,
+      "step": 7302
+    },
+    {
+      "epoch": 0.950800468566966,
+      "grad_norm": 2.331636667251587,
+      "learning_rate": 7.916168061217603e-06,
+      "loss": 0.6511,
+      "step": 7305
+    },
+    {
+      "epoch": 0.9511909410386568,
+      "grad_norm": 3.166801929473877,
+      "learning_rate": 7.914470217009937e-06,
+      "loss": 0.7378,
+      "step": 7308
+    },
+    {
+      "epoch": 0.9515814135103475,
+      "grad_norm": 2.2052841186523438,
+      "learning_rate": 7.912771863650457e-06,
+      "loss": 0.6025,
+      "step": 7311
+    },
+    {
+      "epoch": 0.9519718859820383,
+      "grad_norm": 2.397726058959961,
+      "learning_rate": 7.911073001435859e-06,
+      "loss": 0.6423,
+      "step": 7314
+    },
+    {
+      "epoch": 0.952362358453729,
+      "grad_norm": 2.694310426712036,
+      "learning_rate": 7.909373630662931e-06,
+      "loss": 0.6696,
+      "step": 7317
+    },
+    {
+      "epoch": 0.9527528309254197,
+      "grad_norm": 2.3516201972961426,
+      "learning_rate": 7.90767375162855e-06,
+      "loss": 0.6661,
+      "step": 7320
+    },
+    {
+      "epoch": 0.9531433033971105,
+      "grad_norm": 3.105769395828247,
+      "learning_rate": 7.905973364629682e-06,
+      "loss": 0.643,
+      "step": 7323
+    },
+    {
+      "epoch": 0.9535337758688013,
+      "grad_norm": 2.4792470932006836,
+      "learning_rate": 7.904272469963381e-06,
+      "loss": 0.6866,
+      "step": 7326
+    },
+    {
+      "epoch": 0.9539242483404919,
+      "grad_norm": 2.354459285736084,
+      "learning_rate": 7.90257106792679e-06,
+      "loss": 0.6899,
+      "step": 7329
+    },
+    {
+      "epoch": 0.9543147208121827,
+      "grad_norm": 2.265612840652466,
+      "learning_rate": 7.900869158817137e-06,
+      "loss": 0.6038,
+      "step": 7332
+    },
+    {
+      "epoch": 0.9547051932838735,
+      "grad_norm": 2.663803815841675,
+      "learning_rate": 7.899166742931745e-06,
+      "loss": 0.6569,
+      "step": 7335
+    },
+    {
+      "epoch": 0.9550956657555643,
+      "grad_norm": 2.5542824268341064,
+      "learning_rate": 7.897463820568024e-06,
+      "loss": 0.6466,
+      "step": 7338
+    },
+    {
+      "epoch": 0.955486138227255,
+      "grad_norm": 2.5738067626953125,
+      "learning_rate": 7.895760392023467e-06,
+      "loss": 0.6287,
+      "step": 7341
+    },
+    {
+      "epoch": 0.9558766106989457,
+      "grad_norm": 2.7856462001800537,
+      "learning_rate": 7.894056457595661e-06,
+      "loss": 0.6974,
+      "step": 7344
+    },
+    {
+      "epoch": 0.9562670831706365,
+      "grad_norm": 2.3538174629211426,
+      "learning_rate": 7.892352017582281e-06,
+      "loss": 0.5836,
+      "step": 7347
+    },
+    {
+      "epoch": 0.9566575556423272,
+      "grad_norm": 3.389510154724121,
+      "learning_rate": 7.89064707228109e-06,
+      "loss": 0.6142,
+      "step": 7350
+    },
+    {
+      "epoch": 0.957048028114018,
+      "grad_norm": 2.4188413619995117,
+      "learning_rate": 7.888941621989934e-06,
+      "loss": 0.5828,
+      "step": 7353
+    },
+    {
+      "epoch": 0.9574385005857087,
+      "grad_norm": 2.8886420726776123,
+      "learning_rate": 7.887235667006754e-06,
+      "loss": 0.5977,
+      "step": 7356
+    },
+    {
+      "epoch": 0.9578289730573994,
+      "grad_norm": 3.743659019470215,
+      "learning_rate": 7.885529207629578e-06,
+      "loss": 0.6807,
+      "step": 7359
+    },
+    {
+      "epoch": 0.9582194455290902,
+      "grad_norm": 3.0627243518829346,
+      "learning_rate": 7.883822244156518e-06,
+      "loss": 0.6187,
+      "step": 7362
+    },
+    {
+      "epoch": 0.958609918000781,
+      "grad_norm": 2.3821604251861572,
+      "learning_rate": 7.88211477688578e-06,
+      "loss": 0.6304,
+      "step": 7365
+    },
+    {
+      "epoch": 0.9590003904724717,
+      "grad_norm": 2.6071219444274902,
+      "learning_rate": 7.880406806115655e-06,
+      "loss": 0.6801,
+      "step": 7368
+    },
+    {
+      "epoch": 0.9593908629441624,
+      "grad_norm": 2.3850324153900146,
+      "learning_rate": 7.878698332144518e-06,
+      "loss": 0.7271,
+      "step": 7371
+    },
+    {
+      "epoch": 0.9597813354158532,
+      "grad_norm": 2.5399577617645264,
+      "learning_rate": 7.87698935527084e-06,
+      "loss": 0.6031,
+      "step": 7374
+    },
+    {
+      "epoch": 0.960171807887544,
+      "grad_norm": 2.3061318397521973,
+      "learning_rate": 7.875279875793173e-06,
+      "loss": 0.6618,
+      "step": 7377
+    },
+    {
+      "epoch": 0.9605622803592346,
+      "grad_norm": 2.355522632598877,
+      "learning_rate": 7.87356989401016e-06,
+      "loss": 0.5788,
+      "step": 7380
+    },
+    {
+      "epoch": 0.9609527528309254,
+      "grad_norm": 2.948932409286499,
+      "learning_rate": 7.871859410220531e-06,
+      "loss": 0.7404,
+      "step": 7383
+    },
+    {
+      "epoch": 0.9613432253026162,
+      "grad_norm": 2.565260648727417,
+      "learning_rate": 7.870148424723107e-06,
+      "loss": 0.6905,
+      "step": 7386
+    },
+    {
+      "epoch": 0.961733697774307,
+      "grad_norm": 3.4541046619415283,
+      "learning_rate": 7.86843693781679e-06,
+      "loss": 0.6341,
+      "step": 7389
+    },
+    {
+      "epoch": 0.9621241702459976,
+      "grad_norm": 2.6894516944885254,
+      "learning_rate": 7.866724949800574e-06,
+      "loss": 0.6463,
+      "step": 7392
+    },
+    {
+      "epoch": 0.9625146427176884,
+      "grad_norm": 3.081102132797241,
+      "learning_rate": 7.86501246097354e-06,
+      "loss": 0.6437,
+      "step": 7395
+    },
+    {
+      "epoch": 0.9629051151893792,
+      "grad_norm": 2.0764200687408447,
+      "learning_rate": 7.863299471634855e-06,
+      "loss": 0.5674,
+      "step": 7398
+    },
+    {
+      "epoch": 0.9632955876610699,
+      "grad_norm": 2.589940309524536,
+      "learning_rate": 7.861585982083777e-06,
+      "loss": 0.6612,
+      "step": 7401
+    },
+    {
+      "epoch": 0.9636860601327606,
+      "grad_norm": 2.2398934364318848,
+      "learning_rate": 7.859871992619647e-06,
+      "loss": 0.5479,
+      "step": 7404
+    },
+    {
+      "epoch": 0.9640765326044514,
+      "grad_norm": 2.3958423137664795,
+      "learning_rate": 7.8581575035419e-06,
+      "loss": 0.6572,
+      "step": 7407
+    },
+    {
+      "epoch": 0.9644670050761421,
+      "grad_norm": 2.667583465576172,
+      "learning_rate": 7.856442515150044e-06,
+      "loss": 0.6274,
+      "step": 7410
+    },
+    {
+      "epoch": 0.9648574775478329,
+      "grad_norm": 2.350374460220337,
+      "learning_rate": 7.854727027743693e-06,
+      "loss": 0.6201,
+      "step": 7413
+    },
+    {
+      "epoch": 0.9652479500195236,
+      "grad_norm": 2.8348028659820557,
+      "learning_rate": 7.853011041622536e-06,
+      "loss": 0.6498,
+      "step": 7416
+    },
+    {
+      "epoch": 0.9656384224912143,
+      "grad_norm": 2.5437428951263428,
+      "learning_rate": 7.85129455708635e-06,
+      "loss": 0.6588,
+      "step": 7419
+    },
+    {
+      "epoch": 0.9660288949629051,
+      "grad_norm": 3.723634958267212,
+      "learning_rate": 7.849577574435004e-06,
+      "loss": 0.6834,
+      "step": 7422
+    },
+    {
+      "epoch": 0.9664193674345959,
+      "grad_norm": 2.6529476642608643,
+      "learning_rate": 7.847860093968452e-06,
+      "loss": 0.6071,
+      "step": 7425
+    },
+    {
+      "epoch": 0.9668098399062867,
+      "grad_norm": 2.431433916091919,
+      "learning_rate": 7.84614211598673e-06,
+      "loss": 0.6335,
+      "step": 7428
+    },
+    {
+      "epoch": 0.9672003123779773,
+      "grad_norm": 2.597510576248169,
+      "learning_rate": 7.84442364078997e-06,
+      "loss": 0.6908,
+      "step": 7431
+    },
+    {
+      "epoch": 0.9675907848496681,
+      "grad_norm": 3.476114273071289,
+      "learning_rate": 7.842704668678383e-06,
+      "loss": 0.6247,
+      "step": 7434
+    },
+    {
+      "epoch": 0.9679812573213589,
+      "grad_norm": 4.752762794494629,
+      "learning_rate": 7.84098519995227e-06,
+      "loss": 0.5914,
+      "step": 7437
+    },
+    {
+      "epoch": 0.9683717297930496,
+      "grad_norm": 2.079385995864868,
+      "learning_rate": 7.839265234912019e-06,
+      "loss": 0.6201,
+      "step": 7440
+    },
+    {
+      "epoch": 0.9687622022647403,
+      "grad_norm": 2.5645883083343506,
+      "learning_rate": 7.837544773858104e-06,
+      "loss": 0.6475,
+      "step": 7443
+    },
+    {
+      "epoch": 0.9691526747364311,
+      "grad_norm": 2.471785068511963,
+      "learning_rate": 7.835823817091088e-06,
+      "loss": 0.5883,
+      "step": 7446
+    },
+    {
+      "epoch": 0.9695431472081218,
+      "grad_norm": 2.9823968410491943,
+      "learning_rate": 7.834102364911615e-06,
+      "loss": 0.6943,
+      "step": 7449
+    },
+    {
+      "epoch": 0.9699336196798126,
+      "grad_norm": 3.442391872406006,
+      "learning_rate": 7.832380417620421e-06,
+      "loss": 0.7353,
+      "step": 7452
+    },
+    {
+      "epoch": 0.9703240921515033,
+      "grad_norm": 2.499734401702881,
+      "learning_rate": 7.83065797551833e-06,
+      "loss": 0.6728,
+      "step": 7455
+    },
+    {
+      "epoch": 0.970714564623194,
+      "grad_norm": 2.1826016902923584,
+      "learning_rate": 7.828935038906242e-06,
+      "loss": 0.5552,
+      "step": 7458
+    },
+    {
+      "epoch": 0.9711050370948848,
+      "grad_norm": 3.049651861190796,
+      "learning_rate": 7.827211608085156e-06,
+      "loss": 0.5896,
+      "step": 7461
+    },
+    {
+      "epoch": 0.9714955095665756,
+      "grad_norm": 3.0418035984039307,
+      "learning_rate": 7.825487683356147e-06,
+      "loss": 0.6457,
+      "step": 7464
+    },
+    {
+      "epoch": 0.9718859820382663,
+      "grad_norm": 2.458627700805664,
+      "learning_rate": 7.823763265020385e-06,
+      "loss": 0.5146,
+      "step": 7467
+    },
+    {
+      "epoch": 0.972276454509957,
+      "grad_norm": 2.405871629714966,
+      "learning_rate": 7.822038353379123e-06,
+      "loss": 0.6785,
+      "step": 7470
+    },
+    {
+      "epoch": 0.9726669269816478,
+      "grad_norm": 3.09692120552063,
+      "learning_rate": 7.820312948733694e-06,
+      "loss": 0.6946,
+      "step": 7473
+    },
+    {
+      "epoch": 0.9730573994533386,
+      "grad_norm": 2.938538074493408,
+      "learning_rate": 7.818587051385528e-06,
+      "loss": 0.5546,
+      "step": 7476
+    },
+    {
+      "epoch": 0.9734478719250292,
+      "grad_norm": 2.3744659423828125,
+      "learning_rate": 7.816860661636133e-06,
+      "loss": 0.6079,
+      "step": 7479
+    },
+    {
+      "epoch": 0.97383834439672,
+      "grad_norm": 2.385146379470825,
+      "learning_rate": 7.815133779787106e-06,
+      "loss": 0.6877,
+      "step": 7482
+    },
+    {
+      "epoch": 0.9742288168684108,
+      "grad_norm": 2.245079278945923,
+      "learning_rate": 7.81340640614013e-06,
+      "loss": 0.6019,
+      "step": 7485
+    },
+    {
+      "epoch": 0.9746192893401016,
+      "grad_norm": 2.719306707382202,
+      "learning_rate": 7.811678540996974e-06,
+      "loss": 0.7438,
+      "step": 7488
+    },
+    {
+      "epoch": 0.9750097618117922,
+      "grad_norm": 2.31829833984375,
+      "learning_rate": 7.80995018465949e-06,
+      "loss": 0.6019,
+      "step": 7491
+    },
+    {
+      "epoch": 0.975400234283483,
+      "grad_norm": 2.4517407417297363,
+      "learning_rate": 7.808221337429622e-06,
+      "loss": 0.706,
+      "step": 7494
+    },
+    {
+      "epoch": 0.9757907067551738,
+      "grad_norm": 2.333136558532715,
+      "learning_rate": 7.806491999609393e-06,
+      "loss": 0.5492,
+      "step": 7497
+    },
+    {
+      "epoch": 0.9761811792268645,
+      "grad_norm": 2.3452956676483154,
+      "learning_rate": 7.804762171500915e-06,
+      "loss": 0.6827,
+      "step": 7500
+    },
+    {
+      "epoch": 0.9765716516985552,
+      "grad_norm": 2.5150270462036133,
+      "learning_rate": 7.803031853406389e-06,
+      "loss": 0.6804,
+      "step": 7503
+    },
+    {
+      "epoch": 0.976962124170246,
+      "grad_norm": 2.3206984996795654,
+      "learning_rate": 7.801301045628091e-06,
+      "loss": 0.6799,
+      "step": 7506
+    },
+    {
+      "epoch": 0.9773525966419367,
+      "grad_norm": 2.463517904281616,
+      "learning_rate": 7.799569748468395e-06,
+      "loss": 0.6221,
+      "step": 7509
+    },
+    {
+      "epoch": 0.9777430691136275,
+      "grad_norm": 2.432781457901001,
+      "learning_rate": 7.797837962229755e-06,
+      "loss": 0.744,
+      "step": 7512
+    },
+    {
+      "epoch": 0.9781335415853183,
+      "grad_norm": 2.4999756813049316,
+      "learning_rate": 7.796105687214705e-06,
+      "loss": 0.6272,
+      "step": 7515
+    },
+    {
+      "epoch": 0.978524014057009,
+      "grad_norm": 2.6092498302459717,
+      "learning_rate": 7.794372923725876e-06,
+      "loss": 0.6399,
+      "step": 7518
+    },
+    {
+      "epoch": 0.9789144865286997,
+      "grad_norm": 2.057215452194214,
+      "learning_rate": 7.792639672065978e-06,
+      "loss": 0.5434,
+      "step": 7521
+    },
+    {
+      "epoch": 0.9793049590003905,
+      "grad_norm": 2.6579465866088867,
+      "learning_rate": 7.790905932537802e-06,
+      "loss": 0.6238,
+      "step": 7524
+    },
+    {
+      "epoch": 0.9796954314720813,
+      "grad_norm": 2.4319543838500977,
+      "learning_rate": 7.78917170544423e-06,
+      "loss": 0.6155,
+      "step": 7527
+    },
+    {
+      "epoch": 0.9800859039437719,
+      "grad_norm": 2.190973997116089,
+      "learning_rate": 7.787436991088228e-06,
+      "loss": 0.605,
+      "step": 7530
+    },
+    {
+      "epoch": 0.9804763764154627,
+      "grad_norm": 3.2778749465942383,
+      "learning_rate": 7.78570178977285e-06,
+      "loss": 0.6937,
+      "step": 7533
+    },
+    {
+      "epoch": 0.9808668488871535,
+      "grad_norm": 3.431509256362915,
+      "learning_rate": 7.783966101801228e-06,
+      "loss": 0.6452,
+      "step": 7536
+    },
+    {
+      "epoch": 0.9812573213588442,
+      "grad_norm": 2.6262214183807373,
+      "learning_rate": 7.782229927476585e-06,
+      "loss": 0.649,
+      "step": 7539
+    },
+    {
+      "epoch": 0.9816477938305349,
+      "grad_norm": 2.3871302604675293,
+      "learning_rate": 7.780493267102226e-06,
+      "loss": 0.6498,
+      "step": 7542
+    },
+    {
+      "epoch": 0.9820382663022257,
+      "grad_norm": 2.3250765800476074,
+      "learning_rate": 7.778756120981544e-06,
+      "loss": 0.608,
+      "step": 7545
+    },
+    {
+      "epoch": 0.9824287387739165,
+      "grad_norm": 2.921252489089966,
+      "learning_rate": 7.777018489418011e-06,
+      "loss": 0.6643,
+      "step": 7548
+    },
+    {
+      "epoch": 0.9828192112456072,
+      "grad_norm": 3.2305235862731934,
+      "learning_rate": 7.775280372715193e-06,
+      "loss": 0.6425,
+      "step": 7551
+    },
+    {
+      "epoch": 0.9832096837172979,
+      "grad_norm": 2.3904078006744385,
+      "learning_rate": 7.77354177117673e-06,
+      "loss": 0.6353,
+      "step": 7554
+    },
+    {
+      "epoch": 0.9836001561889887,
+      "grad_norm": 3.6194827556610107,
+      "learning_rate": 7.771802685106356e-06,
+      "loss": 0.6311,
+      "step": 7557
+    },
+    {
+      "epoch": 0.9839906286606794,
+      "grad_norm": 2.5549159049987793,
+      "learning_rate": 7.770063114807882e-06,
+      "loss": 0.6819,
+      "step": 7560
+    },
+    {
+      "epoch": 0.9843811011323702,
+      "grad_norm": 2.8810348510742188,
+      "learning_rate": 7.76832306058521e-06,
+      "loss": 0.6459,
+      "step": 7563
+    },
+    {
+      "epoch": 0.9847715736040609,
+      "grad_norm": 2.3030362129211426,
+      "learning_rate": 7.766582522742323e-06,
+      "loss": 0.6363,
+      "step": 7566
+    },
+    {
+      "epoch": 0.9851620460757516,
+      "grad_norm": 2.76395583152771,
+      "learning_rate": 7.764841501583288e-06,
+      "loss": 0.572,
+      "step": 7569
+    },
+    {
+      "epoch": 0.9855525185474424,
+      "grad_norm": 2.2686634063720703,
+      "learning_rate": 7.76309999741226e-06,
+      "loss": 0.5821,
+      "step": 7572
+    },
+    {
+      "epoch": 0.9859429910191332,
+      "grad_norm": 2.786400079727173,
+      "learning_rate": 7.761358010533478e-06,
+      "loss": 0.6529,
+      "step": 7575
+    },
+    {
+      "epoch": 0.9863334634908238,
+      "grad_norm": 2.2984442710876465,
+      "learning_rate": 7.759615541251257e-06,
+      "loss": 0.6813,
+      "step": 7578
+    },
+    {
+      "epoch": 0.9867239359625146,
+      "grad_norm": 2.821648359298706,
+      "learning_rate": 7.757872589870008e-06,
+      "loss": 0.6864,
+      "step": 7581
+    },
+    {
+      "epoch": 0.9871144084342054,
+      "grad_norm": 2.334958791732788,
+      "learning_rate": 7.756129156694219e-06,
+      "loss": 0.6827,
+      "step": 7584
+    },
+    {
+      "epoch": 0.9875048809058962,
+      "grad_norm": 2.1806464195251465,
+      "learning_rate": 7.754385242028464e-06,
+      "loss": 0.6831,
+      "step": 7587
+    },
+    {
+      "epoch": 0.9878953533775868,
+      "grad_norm": 2.2671940326690674,
+      "learning_rate": 7.7526408461774e-06,
+      "loss": 0.6335,
+      "step": 7590
+    },
+    {
+      "epoch": 0.9882858258492776,
+      "grad_norm": 2.351231098175049,
+      "learning_rate": 7.750895969445773e-06,
+      "loss": 0.6129,
+      "step": 7593
+    },
+    {
+      "epoch": 0.9886762983209684,
+      "grad_norm": 2.558773994445801,
+      "learning_rate": 7.749150612138407e-06,
+      "loss": 0.771,
+      "step": 7596
+    },
+    {
+      "epoch": 0.9890667707926591,
+      "grad_norm": 4.013977527618408,
+      "learning_rate": 7.747404774560213e-06,
+      "loss": 0.7569,
+      "step": 7599
+    },
+    {
+      "epoch": 0.9894572432643499,
+      "grad_norm": 2.6245102882385254,
+      "learning_rate": 7.745658457016182e-06,
+      "loss": 0.6937,
+      "step": 7602
+    },
+    {
+      "epoch": 0.9898477157360406,
+      "grad_norm": 2.6696815490722656,
+      "learning_rate": 7.743911659811399e-06,
+      "loss": 0.5789,
+      "step": 7605
+    },
+    {
+      "epoch": 0.9902381882077314,
+      "grad_norm": 2.585771322250366,
+      "learning_rate": 7.74216438325102e-06,
+      "loss": 0.7235,
+      "step": 7608
+    },
+    {
+      "epoch": 0.9906286606794221,
+      "grad_norm": 2.760922431945801,
+      "learning_rate": 7.740416627640287e-06,
+      "loss": 0.6648,
+      "step": 7611
+    },
+    {
+      "epoch": 0.9910191331511129,
+      "grad_norm": 2.4276010990142822,
+      "learning_rate": 7.738668393284538e-06,
+      "loss": 0.5827,
+      "step": 7614
+    },
+    {
+      "epoch": 0.9914096056228036,
+      "grad_norm": 2.4627444744110107,
+      "learning_rate": 7.736919680489183e-06,
+      "loss": 0.565,
+      "step": 7617
+    },
+    {
+      "epoch": 0.9918000780944943,
+      "grad_norm": 2.5422356128692627,
+      "learning_rate": 7.735170489559715e-06,
+      "loss": 0.6392,
+      "step": 7620
+    },
+    {
+      "epoch": 0.9921905505661851,
+      "grad_norm": 3.172285318374634,
+      "learning_rate": 7.733420820801718e-06,
+      "loss": 0.7205,
+      "step": 7623
+    },
+    {
+      "epoch": 0.9925810230378759,
+      "grad_norm": 2.1353375911712646,
+      "learning_rate": 7.731670674520854e-06,
+      "loss": 0.5064,
+      "step": 7626
+    },
+    {
+      "epoch": 0.9929714955095665,
+      "grad_norm": 2.693587303161621,
+      "learning_rate": 7.72992005102287e-06,
+      "loss": 0.6689,
+      "step": 7629
+    },
+    {
+      "epoch": 0.9933619679812573,
+      "grad_norm": 2.3884193897247314,
+      "learning_rate": 7.728168950613595e-06,
+      "loss": 0.5864,
+      "step": 7632
+    },
+    {
+      "epoch": 0.9937524404529481,
+      "grad_norm": 2.1766364574432373,
+      "learning_rate": 7.726417373598944e-06,
+      "loss": 0.726,
+      "step": 7635
+    },
+    {
+      "epoch": 0.9941429129246389,
+      "grad_norm": 3.2286782264709473,
+      "learning_rate": 7.724665320284913e-06,
+      "loss": 0.6333,
+      "step": 7638
+    },
+    {
+      "epoch": 0.9945333853963295,
+      "grad_norm": 2.1552257537841797,
+      "learning_rate": 7.722912790977582e-06,
+      "loss": 0.5859,
+      "step": 7641
+    },
+    {
+      "epoch": 0.9949238578680203,
+      "grad_norm": 2.586132049560547,
+      "learning_rate": 7.721159785983116e-06,
+      "loss": 0.5727,
+      "step": 7644
+    },
+    {
+      "epoch": 0.9953143303397111,
+      "grad_norm": 2.324312686920166,
+      "learning_rate": 7.71940630560776e-06,
+      "loss": 0.6321,
+      "step": 7647
+    },
+    {
+      "epoch": 0.9957048028114018,
+      "grad_norm": 2.7428348064422607,
+      "learning_rate": 7.717652350157843e-06,
+      "loss": 0.7647,
+      "step": 7650
+    },
+    {
+      "epoch": 0.9960952752830925,
+      "grad_norm": 2.904000759124756,
+      "learning_rate": 7.715897919939775e-06,
+      "loss": 0.6042,
+      "step": 7653
+    },
+    {
+      "epoch": 0.9964857477547833,
+      "grad_norm": 2.2930192947387695,
+      "learning_rate": 7.714143015260056e-06,
+      "loss": 0.6164,
+      "step": 7656
+    },
+    {
+      "epoch": 0.996876220226474,
+      "grad_norm": 2.6954126358032227,
+      "learning_rate": 7.712387636425261e-06,
+      "loss": 0.6288,
+      "step": 7659
+    },
+    {
+      "epoch": 0.9972666926981648,
+      "grad_norm": 2.165433168411255,
+      "learning_rate": 7.710631783742053e-06,
+      "loss": 0.6372,
+      "step": 7662
+    },
+    {
+      "epoch": 0.9976571651698555,
+      "grad_norm": 2.2576329708099365,
+      "learning_rate": 7.708875457517176e-06,
+      "loss": 0.7483,
+      "step": 7665
+    },
+    {
+      "epoch": 0.9980476376415462,
+      "grad_norm": 2.5120785236358643,
+      "learning_rate": 7.707118658057453e-06,
+      "loss": 0.6769,
+      "step": 7668
+    },
+    {
+      "epoch": 0.998438110113237,
+      "grad_norm": 2.3859639167785645,
+      "learning_rate": 7.705361385669795e-06,
+      "loss": 0.6798,
+      "step": 7671
+    },
+    {
+      "epoch": 0.9988285825849278,
+      "grad_norm": 2.457106828689575,
+      "learning_rate": 7.703603640661195e-06,
+      "loss": 0.6851,
+      "step": 7674
+    },
+    {
+      "epoch": 0.9992190550566185,
+      "grad_norm": 2.6297943592071533,
+      "learning_rate": 7.701845423338727e-06,
+      "loss": 0.6308,
+      "step": 7677
+    },
+    {
+      "epoch": 0.9996095275283092,
+      "grad_norm": 2.5548126697540283,
+      "learning_rate": 7.700086734009546e-06,
+      "loss": 0.5697,
+      "step": 7680
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.604892730712891,
+      "learning_rate": 7.698327572980893e-06,
+      "loss": 0.5724,
+      "step": 7683
+    },
+    {
+      "epoch": 1.0003904724716908,
+      "grad_norm": 2.4292237758636475,
+      "learning_rate": 7.69656794056009e-06,
+      "loss": 0.5014,
+      "step": 7686
+    },
+    {
+      "epoch": 1.0007809449433815,
+      "grad_norm": 2.254239320755005,
+      "learning_rate": 7.694807837054542e-06,
+      "loss": 0.5559,
+      "step": 7689
+    },
+    {
+      "epoch": 1.0011714174150723,
+      "grad_norm": 2.3204002380371094,
+      "learning_rate": 7.693047262771734e-06,
+      "loss": 0.4617,
+      "step": 7692
+    },
+    {
+      "epoch": 1.001561889886763,
+      "grad_norm": 2.1934375762939453,
+      "learning_rate": 7.691286218019232e-06,
+      "loss": 0.5337,
+      "step": 7695
+    },
+    {
+      "epoch": 1.0019523623584536,
+      "grad_norm": 3.9040029048919678,
+      "learning_rate": 7.689524703104691e-06,
+      "loss": 0.5133,
+      "step": 7698
+    },
+    {
+      "epoch": 1.0023428348301444,
+      "grad_norm": 2.4869489669799805,
+      "learning_rate": 7.687762718335841e-06,
+      "loss": 0.478,
+      "step": 7701
+    },
+    {
+      "epoch": 1.0027333073018352,
+      "grad_norm": 2.8475353717803955,
+      "learning_rate": 7.686000264020499e-06,
+      "loss": 0.5887,
+      "step": 7704
+    },
+    {
+      "epoch": 1.003123779773526,
+      "grad_norm": 2.571333408355713,
+      "learning_rate": 7.684237340466563e-06,
+      "loss": 0.5476,
+      "step": 7707
+    },
+    {
+      "epoch": 1.0035142522452167,
+      "grad_norm": 3.4672353267669678,
+      "learning_rate": 7.682473947982008e-06,
+      "loss": 0.5168,
+      "step": 7710
+    },
+    {
+      "epoch": 1.0039047247169075,
+      "grad_norm": 2.6517016887664795,
+      "learning_rate": 7.680710086874899e-06,
+      "loss": 0.5201,
+      "step": 7713
+    },
+    {
+      "epoch": 1.0042951971885983,
+      "grad_norm": 3.378917694091797,
+      "learning_rate": 7.678945757453375e-06,
+      "loss": 0.5947,
+      "step": 7716
+    },
+    {
+      "epoch": 1.004685669660289,
+      "grad_norm": 3.3272864818573,
+      "learning_rate": 7.677180960025665e-06,
+      "loss": 0.5434,
+      "step": 7719
+    },
+    {
+      "epoch": 1.0050761421319796,
+      "grad_norm": 2.3880414962768555,
+      "learning_rate": 7.675415694900072e-06,
+      "loss": 0.489,
+      "step": 7722
+    },
+    {
+      "epoch": 1.0054666146036704,
+      "grad_norm": 2.6195037364959717,
+      "learning_rate": 7.673649962384985e-06,
+      "loss": 0.5526,
+      "step": 7725
+    },
+    {
+      "epoch": 1.0058570870753611,
+      "grad_norm": 2.506817102432251,
+      "learning_rate": 7.671883762788877e-06,
+      "loss": 0.4631,
+      "step": 7728
+    },
+    {
+      "epoch": 1.006247559547052,
+      "grad_norm": 3.4505410194396973,
+      "learning_rate": 7.670117096420294e-06,
+      "loss": 0.4856,
+      "step": 7731
+    },
+    {
+      "epoch": 1.0066380320187427,
+      "grad_norm": 2.148554801940918,
+      "learning_rate": 7.668349963587872e-06,
+      "loss": 0.4429,
+      "step": 7734
+    },
+    {
+      "epoch": 1.0070285044904335,
+      "grad_norm": 2.645296812057495,
+      "learning_rate": 7.666582364600324e-06,
+      "loss": 0.5216,
+      "step": 7737
+    },
+    {
+      "epoch": 1.0074189769621242,
+      "grad_norm": 3.129601001739502,
+      "learning_rate": 7.664814299766447e-06,
+      "loss": 0.5421,
+      "step": 7740
+    },
+    {
+      "epoch": 1.007809449433815,
+      "grad_norm": 2.280381917953491,
+      "learning_rate": 7.66304576939512e-06,
+      "loss": 0.5322,
+      "step": 7743
+    },
+    {
+      "epoch": 1.0081999219055056,
+      "grad_norm": 2.548676013946533,
+      "learning_rate": 7.661276773795297e-06,
+      "loss": 0.513,
+      "step": 7746
+    },
+    {
+      "epoch": 1.0085903943771963,
+      "grad_norm": 3.303908109664917,
+      "learning_rate": 7.65950731327602e-06,
+      "loss": 0.5974,
+      "step": 7749
+    },
+    {
+      "epoch": 1.008980866848887,
+      "grad_norm": 2.7848570346832275,
+      "learning_rate": 7.657737388146411e-06,
+      "loss": 0.5288,
+      "step": 7752
+    },
+    {
+      "epoch": 1.0093713393205779,
+      "grad_norm": 2.4875972270965576,
+      "learning_rate": 7.65596699871567e-06,
+      "loss": 0.5485,
+      "step": 7755
+    },
+    {
+      "epoch": 1.0097618117922686,
+      "grad_norm": 2.5876026153564453,
+      "learning_rate": 7.654196145293082e-06,
+      "loss": 0.4749,
+      "step": 7758
+    },
+    {
+      "epoch": 1.0101522842639594,
+      "grad_norm": 2.8364417552948,
+      "learning_rate": 7.652424828188011e-06,
+      "loss": 0.533,
+      "step": 7761
+    },
+    {
+      "epoch": 1.0105427567356502,
+      "grad_norm": 2.3310322761535645,
+      "learning_rate": 7.650653047709906e-06,
+      "loss": 0.6102,
+      "step": 7764
+    },
+    {
+      "epoch": 1.010933229207341,
+      "grad_norm": 4.061567306518555,
+      "learning_rate": 7.648880804168287e-06,
+      "loss": 0.4694,
+      "step": 7767
+    },
+    {
+      "epoch": 1.0113237016790317,
+      "grad_norm": 2.307548999786377,
+      "learning_rate": 7.647108097872763e-06,
+      "loss": 0.5576,
+      "step": 7770
+    },
+    {
+      "epoch": 1.0117141741507223,
+      "grad_norm": 2.360511541366577,
+      "learning_rate": 7.645334929133024e-06,
+      "loss": 0.6458,
+      "step": 7773
+    },
+    {
+      "epoch": 1.012104646622413,
+      "grad_norm": 2.422124147415161,
+      "learning_rate": 7.643561298258836e-06,
+      "loss": 0.439,
+      "step": 7776
+    },
+    {
+      "epoch": 1.0124951190941038,
+      "grad_norm": 2.6831116676330566,
+      "learning_rate": 7.641787205560051e-06,
+      "loss": 0.5588,
+      "step": 7779
+    },
+    {
+      "epoch": 1.0128855915657946,
+      "grad_norm": 2.6215929985046387,
+      "learning_rate": 7.6400126513466e-06,
+      "loss": 0.5985,
+      "step": 7782
+    },
+    {
+      "epoch": 1.0132760640374854,
+      "grad_norm": 3.8551838397979736,
+      "learning_rate": 7.638237635928493e-06,
+      "loss": 0.5659,
+      "step": 7785
+    },
+    {
+      "epoch": 1.0136665365091762,
+      "grad_norm": 2.633751392364502,
+      "learning_rate": 7.63646215961582e-06,
+      "loss": 0.5141,
+      "step": 7788
+    },
+    {
+      "epoch": 1.014057008980867,
+      "grad_norm": 2.5351755619049072,
+      "learning_rate": 7.634686222718757e-06,
+      "loss": 0.473,
+      "step": 7791
+    },
+    {
+      "epoch": 1.0144474814525577,
+      "grad_norm": 2.423412799835205,
+      "learning_rate": 7.632909825547549e-06,
+      "loss": 0.4506,
+      "step": 7794
+    },
+    {
+      "epoch": 1.0148379539242482,
+      "grad_norm": 2.8047244548797607,
+      "learning_rate": 7.631132968412536e-06,
+      "loss": 0.4848,
+      "step": 7797
+    },
+    {
+      "epoch": 1.015228426395939,
+      "grad_norm": 2.771131753921509,
+      "learning_rate": 7.629355651624126e-06,
+      "loss": 0.4393,
+      "step": 7800
+    },
+    {
+      "epoch": 1.0156188988676298,
+      "grad_norm": 3.0369348526000977,
+      "learning_rate": 7.627577875492817e-06,
+      "loss": 0.4429,
+      "step": 7803
+    },
+    {
+      "epoch": 1.0160093713393206,
+      "grad_norm": 2.7859838008880615,
+      "learning_rate": 7.625799640329181e-06,
+      "loss": 0.4625,
+      "step": 7806
+    },
+    {
+      "epoch": 1.0163998438110113,
+      "grad_norm": 2.7366514205932617,
+      "learning_rate": 7.62402094644387e-06,
+      "loss": 0.5831,
+      "step": 7809
+    },
+    {
+      "epoch": 1.0167903162827021,
+      "grad_norm": 2.6900017261505127,
+      "learning_rate": 7.622241794147622e-06,
+      "loss": 0.461,
+      "step": 7812
+    },
+    {
+      "epoch": 1.0171807887543929,
+      "grad_norm": 2.5590507984161377,
+      "learning_rate": 7.6204621837512495e-06,
+      "loss": 0.5923,
+      "step": 7815
+    },
+    {
+      "epoch": 1.0175712612260837,
+      "grad_norm": 2.411809206008911,
+      "learning_rate": 7.6186821155656435e-06,
+      "loss": 0.4458,
+      "step": 7818
+    },
+    {
+      "epoch": 1.0179617336977742,
+      "grad_norm": 3.017052412033081,
+      "learning_rate": 7.616901589901781e-06,
+      "loss": 0.5078,
+      "step": 7821
+    },
+    {
+      "epoch": 1.018352206169465,
+      "grad_norm": 2.686730146408081,
+      "learning_rate": 7.615120607070717e-06,
+      "loss": 0.5053,
+      "step": 7824
+    },
+    {
+      "epoch": 1.0187426786411558,
+      "grad_norm": 2.5899784564971924,
+      "learning_rate": 7.613339167383585e-06,
+      "loss": 0.5209,
+      "step": 7827
+    },
+    {
+      "epoch": 1.0191331511128465,
+      "grad_norm": 2.8320977687835693,
+      "learning_rate": 7.6115572711515975e-06,
+      "loss": 0.588,
+      "step": 7830
+    },
+    {
+      "epoch": 1.0195236235845373,
+      "grad_norm": 2.990689516067505,
+      "learning_rate": 7.609774918686048e-06,
+      "loss": 0.5531,
+      "step": 7833
+    },
+    {
+      "epoch": 1.019914096056228,
+      "grad_norm": 2.3511035442352295,
+      "learning_rate": 7.60799211029831e-06,
+      "loss": 0.4349,
+      "step": 7836
+    },
+    {
+      "epoch": 1.0203045685279188,
+      "grad_norm": 4.445713520050049,
+      "learning_rate": 7.606208846299839e-06,
+      "loss": 0.4621,
+      "step": 7839
+    },
+    {
+      "epoch": 1.0206950409996096,
+      "grad_norm": 2.555044651031494,
+      "learning_rate": 7.604425127002162e-06,
+      "loss": 0.5091,
+      "step": 7842
+    },
+    {
+      "epoch": 1.0210855134713002,
+      "grad_norm": 2.6413726806640625,
+      "learning_rate": 7.602640952716897e-06,
+      "loss": 0.5894,
+      "step": 7845
+    },
+    {
+      "epoch": 1.021475985942991,
+      "grad_norm": 3.004298210144043,
+      "learning_rate": 7.600856323755732e-06,
+      "loss": 0.467,
+      "step": 7848
+    },
+    {
+      "epoch": 1.0218664584146817,
+      "grad_norm": 3.5717835426330566,
+      "learning_rate": 7.599071240430438e-06,
+      "loss": 0.5908,
+      "step": 7851
+    },
+    {
+      "epoch": 1.0222569308863725,
+      "grad_norm": 4.032369613647461,
+      "learning_rate": 7.5972857030528654e-06,
+      "loss": 0.5001,
+      "step": 7854
+    },
+    {
+      "epoch": 1.0226474033580633,
+      "grad_norm": 2.679647922515869,
+      "learning_rate": 7.595499711934946e-06,
+      "loss": 0.549,
+      "step": 7857
+    },
+    {
+      "epoch": 1.023037875829754,
+      "grad_norm": 2.6583425998687744,
+      "learning_rate": 7.593713267388686e-06,
+      "loss": 0.4949,
+      "step": 7860
+    },
+    {
+      "epoch": 1.0234283483014448,
+      "grad_norm": 2.699441432952881,
+      "learning_rate": 7.591926369726174e-06,
+      "loss": 0.5405,
+      "step": 7863
+    },
+    {
+      "epoch": 1.0238188207731356,
+      "grad_norm": 2.4705007076263428,
+      "learning_rate": 7.590139019259579e-06,
+      "loss": 0.6412,
+      "step": 7866
+    },
+    {
+      "epoch": 1.0242092932448263,
+      "grad_norm": 2.6664178371429443,
+      "learning_rate": 7.588351216301147e-06,
+      "loss": 0.4991,
+      "step": 7869
+    },
+    {
+      "epoch": 1.024599765716517,
+      "grad_norm": 2.7439897060394287,
+      "learning_rate": 7.5865629611632005e-06,
+      "loss": 0.5448,
+      "step": 7872
+    },
+    {
+      "epoch": 1.0249902381882077,
+      "grad_norm": 2.5113844871520996,
+      "learning_rate": 7.584774254158147e-06,
+      "loss": 0.5012,
+      "step": 7875
+    },
+    {
+      "epoch": 1.0253807106598984,
+      "grad_norm": 2.640521287918091,
+      "learning_rate": 7.582985095598469e-06,
+      "loss": 0.5282,
+      "step": 7878
+    },
+    {
+      "epoch": 1.0257711831315892,
+      "grad_norm": 2.277463912963867,
+      "learning_rate": 7.5811954857967285e-06,
+      "loss": 0.4448,
+      "step": 7881
+    },
+    {
+      "epoch": 1.02616165560328,
+      "grad_norm": 2.85617733001709,
+      "learning_rate": 7.579405425065567e-06,
+      "loss": 0.4533,
+      "step": 7884
+    },
+    {
+      "epoch": 1.0265521280749708,
+      "grad_norm": 2.714686155319214,
+      "learning_rate": 7.577614913717703e-06,
+      "loss": 0.4766,
+      "step": 7887
+    },
+    {
+      "epoch": 1.0269426005466615,
+      "grad_norm": 4.346346378326416,
+      "learning_rate": 7.575823952065936e-06,
+      "loss": 0.4783,
+      "step": 7890
+    },
+    {
+      "epoch": 1.0273330730183523,
+      "grad_norm": 2.2822585105895996,
+      "learning_rate": 7.574032540423145e-06,
+      "loss": 0.4686,
+      "step": 7893
+    },
+    {
+      "epoch": 1.0277235454900429,
+      "grad_norm": 2.7824783325195312,
+      "learning_rate": 7.572240679102283e-06,
+      "loss": 0.5298,
+      "step": 7896
+    },
+    {
+      "epoch": 1.0281140179617336,
+      "grad_norm": 2.2325971126556396,
+      "learning_rate": 7.570448368416387e-06,
+      "loss": 0.4116,
+      "step": 7899
+    },
+    {
+      "epoch": 1.0285044904334244,
+      "grad_norm": 2.75234317779541,
+      "learning_rate": 7.568655608678566e-06,
+      "loss": 0.5276,
+      "step": 7902
+    },
+    {
+      "epoch": 1.0288949629051152,
+      "grad_norm": 2.7082951068878174,
+      "learning_rate": 7.566862400202015e-06,
+      "loss": 0.5465,
+      "step": 7905
+    },
+    {
+      "epoch": 1.029285435376806,
+      "grad_norm": 2.646604299545288,
+      "learning_rate": 7.5650687433000026e-06,
+      "loss": 0.5085,
+      "step": 7908
+    },
+    {
+      "epoch": 1.0296759078484967,
+      "grad_norm": 2.7950475215911865,
+      "learning_rate": 7.563274638285876e-06,
+      "loss": 0.4742,
+      "step": 7911
+    },
+    {
+      "epoch": 1.0300663803201875,
+      "grad_norm": 2.585192918777466,
+      "learning_rate": 7.5614800854730645e-06,
+      "loss": 0.4821,
+      "step": 7914
+    },
+    {
+      "epoch": 1.0304568527918783,
+      "grad_norm": 2.4464495182037354,
+      "learning_rate": 7.559685085175069e-06,
+      "loss": 0.5618,
+      "step": 7917
+    },
+    {
+      "epoch": 1.0308473252635688,
+      "grad_norm": 2.5040671825408936,
+      "learning_rate": 7.557889637705473e-06,
+      "loss": 0.4845,
+      "step": 7920
+    },
+    {
+      "epoch": 1.0312377977352596,
+      "grad_norm": 2.605333089828491,
+      "learning_rate": 7.556093743377941e-06,
+      "loss": 0.5349,
+      "step": 7923
+    },
+    {
+      "epoch": 1.0316282702069504,
+      "grad_norm": 2.8396682739257812,
+      "learning_rate": 7.5542974025062076e-06,
+      "loss": 0.5243,
+      "step": 7926
+    },
+    {
+      "epoch": 1.0320187426786411,
+      "grad_norm": 2.373145818710327,
+      "learning_rate": 7.552500615404093e-06,
+      "loss": 0.4422,
+      "step": 7929
+    },
+    {
+      "epoch": 1.032409215150332,
+      "grad_norm": 2.7025516033172607,
+      "learning_rate": 7.550703382385488e-06,
+      "loss": 0.5432,
+      "step": 7932
+    },
+    {
+      "epoch": 1.0327996876220227,
+      "grad_norm": 3.0393564701080322,
+      "learning_rate": 7.548905703764371e-06,
+      "loss": 0.4643,
+      "step": 7935
+    },
+    {
+      "epoch": 1.0331901600937134,
+      "grad_norm": 2.475292921066284,
+      "learning_rate": 7.5471075798547865e-06,
+      "loss": 0.4984,
+      "step": 7938
+    },
+    {
+      "epoch": 1.0335806325654042,
+      "grad_norm": 2.519763708114624,
+      "learning_rate": 7.545309010970867e-06,
+      "loss": 0.4986,
+      "step": 7941
+    },
+    {
+      "epoch": 1.0339711050370948,
+      "grad_norm": 2.4332497119903564,
+      "learning_rate": 7.54350999742682e-06,
+      "loss": 0.4935,
+      "step": 7944
+    },
+    {
+      "epoch": 1.0343615775087855,
+      "grad_norm": 3.0140507221221924,
+      "learning_rate": 7.5417105395369235e-06,
+      "loss": 0.4621,
+      "step": 7947
+    },
+    {
+      "epoch": 1.0347520499804763,
+      "grad_norm": 2.577070713043213,
+      "learning_rate": 7.539910637615546e-06,
+      "loss": 0.4794,
+      "step": 7950
+    },
+    {
+      "epoch": 1.035142522452167,
+      "grad_norm": 2.6139345169067383,
+      "learning_rate": 7.538110291977123e-06,
+      "loss": 0.476,
+      "step": 7953
+    },
+    {
+      "epoch": 1.0355329949238579,
+      "grad_norm": 2.6360697746276855,
+      "learning_rate": 7.53630950293617e-06,
+      "loss": 0.4843,
+      "step": 7956
+    },
+    {
+      "epoch": 1.0359234673955486,
+      "grad_norm": 2.5535480976104736,
+      "learning_rate": 7.5345082708072836e-06,
+      "loss": 0.4639,
+      "step": 7959
+    },
+    {
+      "epoch": 1.0363139398672394,
+      "grad_norm": 2.9951012134552,
+      "learning_rate": 7.532706595905133e-06,
+      "loss": 0.5363,
+      "step": 7962
+    },
+    {
+      "epoch": 1.0367044123389302,
+      "grad_norm": 2.2908568382263184,
+      "learning_rate": 7.53090447854447e-06,
+      "loss": 0.4262,
+      "step": 7965
+    },
+    {
+      "epoch": 1.037094884810621,
+      "grad_norm": 2.7684085369110107,
+      "learning_rate": 7.529101919040116e-06,
+      "loss": 0.5007,
+      "step": 7968
+    },
+    {
+      "epoch": 1.0374853572823115,
+      "grad_norm": 2.3997421264648438,
+      "learning_rate": 7.5272989177069795e-06,
+      "loss": 0.5509,
+      "step": 7971
+    },
+    {
+      "epoch": 1.0378758297540023,
+      "grad_norm": 2.433640241622925,
+      "learning_rate": 7.525495474860037e-06,
+      "loss": 0.5496,
+      "step": 7974
+    },
+    {
+      "epoch": 1.038266302225693,
+      "grad_norm": 3.6235389709472656,
+      "learning_rate": 7.52369159081435e-06,
+      "loss": 0.5144,
+      "step": 7977
+    },
+    {
+      "epoch": 1.0386567746973838,
+      "grad_norm": 2.715113401412964,
+      "learning_rate": 7.521887265885049e-06,
+      "loss": 0.4914,
+      "step": 7980
+    },
+    {
+      "epoch": 1.0390472471690746,
+      "grad_norm": 2.352916955947876,
+      "learning_rate": 7.520082500387349e-06,
+      "loss": 0.4605,
+      "step": 7983
+    },
+    {
+      "epoch": 1.0394377196407654,
+      "grad_norm": 2.825352668762207,
+      "learning_rate": 7.518277294636538e-06,
+      "loss": 0.5833,
+      "step": 7986
+    },
+    {
+      "epoch": 1.0398281921124561,
+      "grad_norm": 2.906801462173462,
+      "learning_rate": 7.51647164894798e-06,
+      "loss": 0.5857,
+      "step": 7989
+    },
+    {
+      "epoch": 1.040218664584147,
+      "grad_norm": 2.4375948905944824,
+      "learning_rate": 7.51466556363712e-06,
+      "loss": 0.4738,
+      "step": 7992
+    },
+    {
+      "epoch": 1.0406091370558375,
+      "grad_norm": 2.6101300716400146,
+      "learning_rate": 7.512859039019476e-06,
+      "loss": 0.4943,
+      "step": 7995
+    },
+    {
+      "epoch": 1.0409996095275282,
+      "grad_norm": 7.20686149597168,
+      "learning_rate": 7.511052075410644e-06,
+      "loss": 0.5433,
+      "step": 7998
+    },
+    {
+      "epoch": 1.041390081999219,
+      "grad_norm": 2.5502688884735107,
+      "learning_rate": 7.509244673126298e-06,
+      "loss": 0.5695,
+      "step": 8001
+    },
+    {
+      "epoch": 1.0417805544709098,
+      "grad_norm": 3.865182399749756,
+      "learning_rate": 7.507436832482185e-06,
+      "loss": 0.4969,
+      "step": 8004
+    },
+    {
+      "epoch": 1.0421710269426006,
+      "grad_norm": 2.659177541732788,
+      "learning_rate": 7.5056285537941335e-06,
+      "loss": 0.6332,
+      "step": 8007
+    },
+    {
+      "epoch": 1.0425614994142913,
+      "grad_norm": 2.6241230964660645,
+      "learning_rate": 7.503819837378042e-06,
+      "loss": 0.4546,
+      "step": 8010
+    },
+    {
+      "epoch": 1.042951971885982,
+      "grad_norm": 3.0326120853424072,
+      "learning_rate": 7.502010683549894e-06,
+      "loss": 0.5339,
+      "step": 8013
+    },
+    {
+      "epoch": 1.0433424443576729,
+      "grad_norm": 2.694370985031128,
+      "learning_rate": 7.500201092625743e-06,
+      "loss": 0.4685,
+      "step": 8016
+    },
+    {
+      "epoch": 1.0437329168293634,
+      "grad_norm": 2.683987855911255,
+      "learning_rate": 7.498391064921721e-06,
+      "loss": 0.5303,
+      "step": 8019
+    },
+    {
+      "epoch": 1.0441233893010542,
+      "grad_norm": 3.791745901107788,
+      "learning_rate": 7.496580600754036e-06,
+      "loss": 0.4798,
+      "step": 8022
+    },
+    {
+      "epoch": 1.044513861772745,
+      "grad_norm": 3.1122212409973145,
+      "learning_rate": 7.494769700438971e-06,
+      "loss": 0.5105,
+      "step": 8025
+    },
+    {
+      "epoch": 1.0449043342444357,
+      "grad_norm": 2.536529302597046,
+      "learning_rate": 7.492958364292888e-06,
+      "loss": 0.4569,
+      "step": 8028
+    },
+    {
+      "epoch": 1.0452948067161265,
+      "grad_norm": 2.741360902786255,
+      "learning_rate": 7.491146592632223e-06,
+      "loss": 0.5162,
+      "step": 8031
+    },
+    {
+      "epoch": 1.0456852791878173,
+      "grad_norm": 2.675175905227661,
+      "learning_rate": 7.48933438577349e-06,
+      "loss": 0.4978,
+      "step": 8034
+    },
+    {
+      "epoch": 1.046075751659508,
+      "grad_norm": 2.6189332008361816,
+      "learning_rate": 7.487521744033275e-06,
+      "loss": 0.4782,
+      "step": 8037
+    },
+    {
+      "epoch": 1.0464662241311988,
+      "grad_norm": 3.0063090324401855,
+      "learning_rate": 7.485708667728245e-06,
+      "loss": 0.4818,
+      "step": 8040
+    },
+    {
+      "epoch": 1.0468566966028896,
+      "grad_norm": 3.2774417400360107,
+      "learning_rate": 7.483895157175141e-06,
+      "loss": 0.4994,
+      "step": 8043
+    },
+    {
+      "epoch": 1.0472471690745802,
+      "grad_norm": 2.4329867362976074,
+      "learning_rate": 7.482081212690777e-06,
+      "loss": 0.478,
+      "step": 8046
+    },
+    {
+      "epoch": 1.047637641546271,
+      "grad_norm": 2.885873556137085,
+      "learning_rate": 7.480266834592047e-06,
+      "loss": 0.5012,
+      "step": 8049
+    },
+    {
+      "epoch": 1.0480281140179617,
+      "grad_norm": 2.5598886013031006,
+      "learning_rate": 7.478452023195918e-06,
+      "loss": 0.4845,
+      "step": 8052
+    },
+    {
+      "epoch": 1.0484185864896525,
+      "grad_norm": 3.7768354415893555,
+      "learning_rate": 7.476636778819435e-06,
+      "loss": 0.4945,
+      "step": 8055
+    },
+    {
+      "epoch": 1.0488090589613432,
+      "grad_norm": 3.060943365097046,
+      "learning_rate": 7.474821101779718e-06,
+      "loss": 0.5462,
+      "step": 8058
+    },
+    {
+      "epoch": 1.049199531433034,
+      "grad_norm": 2.9565443992614746,
+      "learning_rate": 7.47300499239396e-06,
+      "loss": 0.5928,
+      "step": 8061
+    },
+    {
+      "epoch": 1.0495900039047248,
+      "grad_norm": 2.9404168128967285,
+      "learning_rate": 7.471188450979432e-06,
+      "loss": 0.6085,
+      "step": 8064
+    },
+    {
+      "epoch": 1.0499804763764156,
+      "grad_norm": 2.50423264503479,
+      "learning_rate": 7.4693714778534795e-06,
+      "loss": 0.5282,
+      "step": 8067
+    },
+    {
+      "epoch": 1.0503709488481061,
+      "grad_norm": 2.8531126976013184,
+      "learning_rate": 7.467554073333525e-06,
+      "loss": 0.5469,
+      "step": 8070
+    },
+    {
+      "epoch": 1.0507614213197969,
+      "grad_norm": 2.6135265827178955,
+      "learning_rate": 7.465736237737066e-06,
+      "loss": 0.5384,
+      "step": 8073
+    },
+    {
+      "epoch": 1.0511518937914877,
+      "grad_norm": 2.651714324951172,
+      "learning_rate": 7.463917971381672e-06,
+      "loss": 0.4965,
+      "step": 8076
+    },
+    {
+      "epoch": 1.0515423662631784,
+      "grad_norm": 2.368147611618042,
+      "learning_rate": 7.462099274584993e-06,
+      "loss": 0.5733,
+      "step": 8079
+    },
+    {
+      "epoch": 1.0519328387348692,
+      "grad_norm": 3.0812952518463135,
+      "learning_rate": 7.460280147664749e-06,
+      "loss": 0.4543,
+      "step": 8082
+    },
+    {
+      "epoch": 1.05232331120656,
+      "grad_norm": 2.6388537883758545,
+      "learning_rate": 7.458460590938741e-06,
+      "loss": 0.5554,
+      "step": 8085
+    },
+    {
+      "epoch": 1.0527137836782507,
+      "grad_norm": 2.5167076587677,
+      "learning_rate": 7.4566406047248385e-06,
+      "loss": 0.5311,
+      "step": 8088
+    },
+    {
+      "epoch": 1.0531042561499415,
+      "grad_norm": 3.7936503887176514,
+      "learning_rate": 7.454820189340989e-06,
+      "loss": 0.5205,
+      "step": 8091
+    },
+    {
+      "epoch": 1.053494728621632,
+      "grad_norm": 3.482969045639038,
+      "learning_rate": 7.452999345105218e-06,
+      "loss": 0.4703,
+      "step": 8094
+    },
+    {
+      "epoch": 1.0538852010933228,
+      "grad_norm": 3.16825270652771,
+      "learning_rate": 7.451178072335621e-06,
+      "loss": 0.4716,
+      "step": 8097
+    },
+    {
+      "epoch": 1.0542756735650136,
+      "grad_norm": 2.6491572856903076,
+      "learning_rate": 7.449356371350371e-06,
+      "loss": 0.5886,
+      "step": 8100
+    },
+    {
+      "epoch": 1.0546661460367044,
+      "grad_norm": 2.7450640201568604,
+      "learning_rate": 7.447534242467718e-06,
+      "loss": 0.4394,
+      "step": 8103
+    },
+    {
+      "epoch": 1.0550566185083952,
+      "grad_norm": 2.392484664916992,
+      "learning_rate": 7.445711686005978e-06,
+      "loss": 0.4185,
+      "step": 8106
+    },
+    {
+      "epoch": 1.055447090980086,
+      "grad_norm": 2.569303274154663,
+      "learning_rate": 7.443888702283555e-06,
+      "loss": 0.4916,
+      "step": 8109
+    },
+    {
+      "epoch": 1.0558375634517767,
+      "grad_norm": 2.565568447113037,
+      "learning_rate": 7.442065291618915e-06,
+      "loss": 0.4677,
+      "step": 8112
+    },
+    {
+      "epoch": 1.0562280359234675,
+      "grad_norm": 2.54575252532959,
+      "learning_rate": 7.440241454330606e-06,
+      "loss": 0.4087,
+      "step": 8115
+    },
+    {
+      "epoch": 1.0566185083951583,
+      "grad_norm": 3.055168390274048,
+      "learning_rate": 7.438417190737248e-06,
+      "loss": 0.442,
+      "step": 8118
+    },
+    {
+      "epoch": 1.0570089808668488,
+      "grad_norm": 2.5168895721435547,
+      "learning_rate": 7.436592501157538e-06,
+      "loss": 0.4346,
+      "step": 8121
+    },
+    {
+      "epoch": 1.0573994533385396,
+      "grad_norm": 2.5084915161132812,
+      "learning_rate": 7.434767385910243e-06,
+      "loss": 0.4508,
+      "step": 8124
+    },
+    {
+      "epoch": 1.0577899258102303,
+      "grad_norm": 3.041224479675293,
+      "learning_rate": 7.432941845314207e-06,
+      "loss": 0.4823,
+      "step": 8127
+    },
+    {
+      "epoch": 1.0581803982819211,
+      "grad_norm": 3.2237794399261475,
+      "learning_rate": 7.431115879688351e-06,
+      "loss": 0.5292,
+      "step": 8130
+    },
+    {
+      "epoch": 1.058570870753612,
+      "grad_norm": 3.038773775100708,
+      "learning_rate": 7.429289489351663e-06,
+      "loss": 0.5,
+      "step": 8133
+    },
+    {
+      "epoch": 1.0589613432253027,
+      "grad_norm": 3.018162488937378,
+      "learning_rate": 7.4274626746232125e-06,
+      "loss": 0.4816,
+      "step": 8136
+    },
+    {
+      "epoch": 1.0593518156969934,
+      "grad_norm": 2.773552179336548,
+      "learning_rate": 7.425635435822139e-06,
+      "loss": 0.4643,
+      "step": 8139
+    },
+    {
+      "epoch": 1.0597422881686842,
+      "grad_norm": 2.652228593826294,
+      "learning_rate": 7.423807773267659e-06,
+      "loss": 0.5507,
+      "step": 8142
+    },
+    {
+      "epoch": 1.0601327606403748,
+      "grad_norm": 2.473726511001587,
+      "learning_rate": 7.421979687279058e-06,
+      "loss": 0.5871,
+      "step": 8145
+    },
+    {
+      "epoch": 1.0605232331120655,
+      "grad_norm": 2.8862738609313965,
+      "learning_rate": 7.420151178175702e-06,
+      "loss": 0.5979,
+      "step": 8148
+    },
+    {
+      "epoch": 1.0609137055837563,
+      "grad_norm": 3.1102709770202637,
+      "learning_rate": 7.4183222462770266e-06,
+      "loss": 0.4983,
+      "step": 8151
+    },
+    {
+      "epoch": 1.061304178055447,
+      "grad_norm": 2.4361696243286133,
+      "learning_rate": 7.416492891902541e-06,
+      "loss": 0.4916,
+      "step": 8154
+    },
+    {
+      "epoch": 1.0616946505271379,
+      "grad_norm": 2.7870354652404785,
+      "learning_rate": 7.414663115371832e-06,
+      "loss": 0.4924,
+      "step": 8157
+    },
+    {
+      "epoch": 1.0620851229988286,
+      "grad_norm": 2.7369415760040283,
+      "learning_rate": 7.412832917004556e-06,
+      "loss": 0.5325,
+      "step": 8160
+    },
+    {
+      "epoch": 1.0624755954705194,
+      "grad_norm": 2.71121883392334,
+      "learning_rate": 7.411002297120444e-06,
+      "loss": 0.5131,
+      "step": 8163
+    },
+    {
+      "epoch": 1.0628660679422102,
+      "grad_norm": 2.594827890396118,
+      "learning_rate": 7.409171256039305e-06,
+      "loss": 0.468,
+      "step": 8166
+    },
+    {
+      "epoch": 1.0632565404139007,
+      "grad_norm": 2.7998273372650146,
+      "learning_rate": 7.407339794081013e-06,
+      "loss": 0.4999,
+      "step": 8169
+    },
+    {
+      "epoch": 1.0636470128855915,
+      "grad_norm": 2.753424882888794,
+      "learning_rate": 7.405507911565526e-06,
+      "loss": 0.517,
+      "step": 8172
+    },
+    {
+      "epoch": 1.0640374853572823,
+      "grad_norm": 2.8335673809051514,
+      "learning_rate": 7.403675608812866e-06,
+      "loss": 0.4969,
+      "step": 8175
+    },
+    {
+      "epoch": 1.064427957828973,
+      "grad_norm": 2.668989419937134,
+      "learning_rate": 7.401842886143133e-06,
+      "loss": 0.4657,
+      "step": 8178
+    },
+    {
+      "epoch": 1.0648184303006638,
+      "grad_norm": 3.0732874870300293,
+      "learning_rate": 7.400009743876502e-06,
+      "loss": 0.4661,
+      "step": 8181
+    },
+    {
+      "epoch": 1.0652089027723546,
+      "grad_norm": 2.5212926864624023,
+      "learning_rate": 7.398176182333217e-06,
+      "loss": 0.4812,
+      "step": 8184
+    },
+    {
+      "epoch": 1.0655993752440454,
+      "grad_norm": 2.622633457183838,
+      "learning_rate": 7.396342201833597e-06,
+      "loss": 0.5319,
+      "step": 8187
+    },
+    {
+      "epoch": 1.0659898477157361,
+      "grad_norm": 3.5963311195373535,
+      "learning_rate": 7.394507802698037e-06,
+      "loss": 0.52,
+      "step": 8190
+    },
+    {
+      "epoch": 1.066380320187427,
+      "grad_norm": 2.81231951713562,
+      "learning_rate": 7.392672985247002e-06,
+      "loss": 0.5517,
+      "step": 8193
+    },
+    {
+      "epoch": 1.0667707926591175,
+      "grad_norm": 2.674590826034546,
+      "learning_rate": 7.390837749801027e-06,
+      "loss": 0.5542,
+      "step": 8196
+    },
+    {
+      "epoch": 1.0671612651308082,
+      "grad_norm": 2.3117876052856445,
+      "learning_rate": 7.389002096680729e-06,
+      "loss": 0.4823,
+      "step": 8199
+    },
+    {
+      "epoch": 1.067551737602499,
+      "grad_norm": 2.682879686355591,
+      "learning_rate": 7.387166026206789e-06,
+      "loss": 0.4202,
+      "step": 8202
+    },
+    {
+      "epoch": 1.0679422100741898,
+      "grad_norm": 3.0507800579071045,
+      "learning_rate": 7.3853295386999665e-06,
+      "loss": 0.5087,
+      "step": 8205
+    },
+    {
+      "epoch": 1.0683326825458805,
+      "grad_norm": 3.6044976711273193,
+      "learning_rate": 7.383492634481093e-06,
+      "loss": 0.5968,
+      "step": 8208
+    },
+    {
+      "epoch": 1.0687231550175713,
+      "grad_norm": 2.9628076553344727,
+      "learning_rate": 7.381655313871069e-06,
+      "loss": 0.5076,
+      "step": 8211
+    },
+    {
+      "epoch": 1.069113627489262,
+      "grad_norm": 2.616960287094116,
+      "learning_rate": 7.379817577190873e-06,
+      "loss": 0.5251,
+      "step": 8214
+    },
+    {
+      "epoch": 1.0695040999609526,
+      "grad_norm": 3.7109944820404053,
+      "learning_rate": 7.377979424761551e-06,
+      "loss": 0.5161,
+      "step": 8217
+    },
+    {
+      "epoch": 1.0698945724326434,
+      "grad_norm": 2.695524215698242,
+      "learning_rate": 7.376140856904227e-06,
+      "loss": 0.525,
+      "step": 8220
+    },
+    {
+      "epoch": 1.0702850449043342,
+      "grad_norm": 2.8099405765533447,
+      "learning_rate": 7.374301873940093e-06,
+      "loss": 0.5413,
+      "step": 8223
+    },
+    {
+      "epoch": 1.070675517376025,
+      "grad_norm": 2.4617536067962646,
+      "learning_rate": 7.372462476190417e-06,
+      "loss": 0.5188,
+      "step": 8226
+    },
+    {
+      "epoch": 1.0710659898477157,
+      "grad_norm": 2.8862340450286865,
+      "learning_rate": 7.370622663976539e-06,
+      "loss": 0.499,
+      "step": 8229
+    },
+    {
+      "epoch": 1.0714564623194065,
+      "grad_norm": 2.755232334136963,
+      "learning_rate": 7.3687824376198665e-06,
+      "loss": 0.4981,
+      "step": 8232
+    },
+    {
+      "epoch": 1.0718469347910973,
+      "grad_norm": 2.4629595279693604,
+      "learning_rate": 7.3669417974418865e-06,
+      "loss": 0.501,
+      "step": 8235
+    },
+    {
+      "epoch": 1.072237407262788,
+      "grad_norm": 2.6113266944885254,
+      "learning_rate": 7.365100743764153e-06,
+      "loss": 0.4459,
+      "step": 8238
+    },
+    {
+      "epoch": 1.0726278797344788,
+      "grad_norm": 2.6274919509887695,
+      "learning_rate": 7.363259276908294e-06,
+      "loss": 0.4977,
+      "step": 8241
+    },
+    {
+      "epoch": 1.0730183522061694,
+      "grad_norm": 2.575099468231201,
+      "learning_rate": 7.3614173971960134e-06,
+      "loss": 0.5597,
+      "step": 8244
+    },
+    {
+      "epoch": 1.0734088246778601,
+      "grad_norm": 2.7904398441314697,
+      "learning_rate": 7.35957510494908e-06,
+      "loss": 0.5409,
+      "step": 8247
+    },
+    {
+      "epoch": 1.073799297149551,
+      "grad_norm": 2.6250545978546143,
+      "learning_rate": 7.357732400489342e-06,
+      "loss": 0.4675,
+      "step": 8250
+    },
+    {
+      "epoch": 1.0741897696212417,
+      "grad_norm": 2.454293727874756,
+      "learning_rate": 7.35588928413871e-06,
+      "loss": 0.5445,
+      "step": 8253
+    },
+    {
+      "epoch": 1.0745802420929325,
+      "grad_norm": 2.9731876850128174,
+      "learning_rate": 7.354045756219177e-06,
+      "loss": 0.5563,
+      "step": 8256
+    },
+    {
+      "epoch": 1.0749707145646232,
+      "grad_norm": 2.4799342155456543,
+      "learning_rate": 7.352201817052804e-06,
+      "loss": 0.5015,
+      "step": 8259
+    },
+    {
+      "epoch": 1.075361187036314,
+      "grad_norm": 2.2865350246429443,
+      "learning_rate": 7.350357466961719e-06,
+      "loss": 0.4247,
+      "step": 8262
+    },
+    {
+      "epoch": 1.0757516595080048,
+      "grad_norm": 2.5727767944335938,
+      "learning_rate": 7.348512706268132e-06,
+      "loss": 0.5026,
+      "step": 8265
+    },
+    {
+      "epoch": 1.0761421319796955,
+      "grad_norm": 2.5520944595336914,
+      "learning_rate": 7.346667535294314e-06,
+      "loss": 0.4901,
+      "step": 8268
+    },
+    {
+      "epoch": 1.076532604451386,
+      "grad_norm": 2.7078912258148193,
+      "learning_rate": 7.344821954362615e-06,
+      "loss": 0.5435,
+      "step": 8271
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 2.717442512512207,
+      "learning_rate": 7.342975963795454e-06,
+      "loss": 0.4969,
+      "step": 8274
+    },
+    {
+      "epoch": 1.0773135493947676,
+      "grad_norm": 2.646892547607422,
+      "learning_rate": 7.341129563915319e-06,
+      "loss": 0.6138,
+      "step": 8277
+    },
+    {
+      "epoch": 1.0777040218664584,
+      "grad_norm": 3.4911608695983887,
+      "learning_rate": 7.339282755044776e-06,
+      "loss": 0.4708,
+      "step": 8280
+    },
+    {
+      "epoch": 1.0780944943381492,
+      "grad_norm": 2.682307243347168,
+      "learning_rate": 7.337435537506456e-06,
+      "loss": 0.4995,
+      "step": 8283
+    },
+    {
+      "epoch": 1.07848496680984,
+      "grad_norm": 2.9293227195739746,
+      "learning_rate": 7.335587911623065e-06,
+      "loss": 0.4823,
+      "step": 8286
+    },
+    {
+      "epoch": 1.0788754392815307,
+      "grad_norm": 2.717480182647705,
+      "learning_rate": 7.33373987771738e-06,
+      "loss": 0.4743,
+      "step": 8289
+    },
+    {
+      "epoch": 1.0792659117532213,
+      "grad_norm": 2.61503005027771,
+      "learning_rate": 7.331891436112246e-06,
+      "loss": 0.4466,
+      "step": 8292
+    },
+    {
+      "epoch": 1.079656384224912,
+      "grad_norm": 2.5344207286834717,
+      "learning_rate": 7.330042587130586e-06,
+      "loss": 0.4884,
+      "step": 8295
+    },
+    {
+      "epoch": 1.0800468566966028,
+      "grad_norm": 2.708646774291992,
+      "learning_rate": 7.328193331095387e-06,
+      "loss": 0.4836,
+      "step": 8298
+    },
+    {
+      "epoch": 1.0804373291682936,
+      "grad_norm": 2.689146041870117,
+      "learning_rate": 7.326343668329711e-06,
+      "loss": 0.5197,
+      "step": 8301
+    },
+    {
+      "epoch": 1.0808278016399844,
+      "grad_norm": 2.5231661796569824,
+      "learning_rate": 7.324493599156688e-06,
+      "loss": 0.5055,
+      "step": 8304
+    },
+    {
+      "epoch": 1.0812182741116751,
+      "grad_norm": 2.644646167755127,
+      "learning_rate": 7.322643123899525e-06,
+      "loss": 0.4866,
+      "step": 8307
+    },
+    {
+      "epoch": 1.081608746583366,
+      "grad_norm": 3.1305809020996094,
+      "learning_rate": 7.3207922428814935e-06,
+      "loss": 0.5115,
+      "step": 8310
+    },
+    {
+      "epoch": 1.0819992190550567,
+      "grad_norm": 2.661428928375244,
+      "learning_rate": 7.318940956425941e-06,
+      "loss": 0.5316,
+      "step": 8313
+    },
+    {
+      "epoch": 1.0823896915267475,
+      "grad_norm": 2.929919958114624,
+      "learning_rate": 7.317089264856281e-06,
+      "loss": 0.5167,
+      "step": 8316
+    },
+    {
+      "epoch": 1.082780163998438,
+      "grad_norm": 2.6820735931396484,
+      "learning_rate": 7.315237168496e-06,
+      "loss": 0.5193,
+      "step": 8319
+    },
+    {
+      "epoch": 1.0831706364701288,
+      "grad_norm": 2.9934213161468506,
+      "learning_rate": 7.313384667668657e-06,
+      "loss": 0.4755,
+      "step": 8322
+    },
+    {
+      "epoch": 1.0835611089418196,
+      "grad_norm": 2.795145273208618,
+      "learning_rate": 7.311531762697879e-06,
+      "loss": 0.4628,
+      "step": 8325
+    },
+    {
+      "epoch": 1.0839515814135103,
+      "grad_norm": 2.4267120361328125,
+      "learning_rate": 7.309678453907365e-06,
+      "loss": 0.4883,
+      "step": 8328
+    },
+    {
+      "epoch": 1.084342053885201,
+      "grad_norm": 2.76674485206604,
+      "learning_rate": 7.307824741620883e-06,
+      "loss": 0.4929,
+      "step": 8331
+    },
+    {
+      "epoch": 1.0847325263568919,
+      "grad_norm": 2.8393499851226807,
+      "learning_rate": 7.3059706261622745e-06,
+      "loss": 0.5926,
+      "step": 8334
+    },
+    {
+      "epoch": 1.0851229988285827,
+      "grad_norm": 2.8024489879608154,
+      "learning_rate": 7.304116107855449e-06,
+      "loss": 0.5291,
+      "step": 8337
+    },
+    {
+      "epoch": 1.0855134713002734,
+      "grad_norm": 2.2711691856384277,
+      "learning_rate": 7.302261187024386e-06,
+      "loss": 0.5723,
+      "step": 8340
+    },
+    {
+      "epoch": 1.085903943771964,
+      "grad_norm": 2.544989585876465,
+      "learning_rate": 7.300405863993136e-06,
+      "loss": 0.4604,
+      "step": 8343
+    },
+    {
+      "epoch": 1.0862944162436547,
+      "grad_norm": 4.123943328857422,
+      "learning_rate": 7.298550139085823e-06,
+      "loss": 0.4671,
+      "step": 8346
+    },
+    {
+      "epoch": 1.0866848887153455,
+      "grad_norm": 2.6443090438842773,
+      "learning_rate": 7.296694012626635e-06,
+      "loss": 0.4958,
+      "step": 8349
+    },
+    {
+      "epoch": 1.0870753611870363,
+      "grad_norm": 2.406564474105835,
+      "learning_rate": 7.294837484939835e-06,
+      "loss": 0.4587,
+      "step": 8352
+    },
+    {
+      "epoch": 1.087465833658727,
+      "grad_norm": 2.6871235370635986,
+      "learning_rate": 7.292980556349754e-06,
+      "loss": 0.5546,
+      "step": 8355
+    },
+    {
+      "epoch": 1.0878563061304178,
+      "grad_norm": 2.515164375305176,
+      "learning_rate": 7.291123227180793e-06,
+      "loss": 0.46,
+      "step": 8358
+    },
+    {
+      "epoch": 1.0882467786021086,
+      "grad_norm": 2.631863832473755,
+      "learning_rate": 7.289265497757424e-06,
+      "loss": 0.4499,
+      "step": 8361
+    },
+    {
+      "epoch": 1.0886372510737994,
+      "grad_norm": 3.6877024173736572,
+      "learning_rate": 7.287407368404189e-06,
+      "loss": 0.519,
+      "step": 8364
+    },
+    {
+      "epoch": 1.08902772354549,
+      "grad_norm": 2.77193021774292,
+      "learning_rate": 7.2855488394456975e-06,
+      "loss": 0.5345,
+      "step": 8367
+    },
+    {
+      "epoch": 1.0894181960171807,
+      "grad_norm": 2.9497268199920654,
+      "learning_rate": 7.283689911206633e-06,
+      "loss": 0.4815,
+      "step": 8370
+    },
+    {
+      "epoch": 1.0898086684888715,
+      "grad_norm": 2.750422477722168,
+      "learning_rate": 7.281830584011745e-06,
+      "loss": 0.4844,
+      "step": 8373
+    },
+    {
+      "epoch": 1.0901991409605623,
+      "grad_norm": 2.9934771060943604,
+      "learning_rate": 7.279970858185854e-06,
+      "loss": 0.4681,
+      "step": 8376
+    },
+    {
+      "epoch": 1.090589613432253,
+      "grad_norm": 2.8141846656799316,
+      "learning_rate": 7.278110734053852e-06,
+      "loss": 0.5763,
+      "step": 8379
+    },
+    {
+      "epoch": 1.0909800859039438,
+      "grad_norm": 2.7781667709350586,
+      "learning_rate": 7.276250211940695e-06,
+      "loss": 0.5492,
+      "step": 8382
+    },
+    {
+      "epoch": 1.0913705583756346,
+      "grad_norm": 2.797043800354004,
+      "learning_rate": 7.274389292171416e-06,
+      "loss": 0.4497,
+      "step": 8385
+    },
+    {
+      "epoch": 1.0917610308473253,
+      "grad_norm": 2.90187406539917,
+      "learning_rate": 7.272527975071111e-06,
+      "loss": 0.4726,
+      "step": 8388
+    },
+    {
+      "epoch": 1.0921515033190161,
+      "grad_norm": 2.9917635917663574,
+      "learning_rate": 7.270666260964949e-06,
+      "loss": 0.5161,
+      "step": 8391
+    },
+    {
+      "epoch": 1.0925419757907067,
+      "grad_norm": 2.316232919692993,
+      "learning_rate": 7.268804150178171e-06,
+      "loss": 0.4295,
+      "step": 8394
+    },
+    {
+      "epoch": 1.0929324482623974,
+      "grad_norm": 2.60370135307312,
+      "learning_rate": 7.266941643036077e-06,
+      "loss": 0.4582,
+      "step": 8397
+    },
+    {
+      "epoch": 1.0933229207340882,
+      "grad_norm": 2.546489953994751,
+      "learning_rate": 7.26507873986405e-06,
+      "loss": 0.4716,
+      "step": 8400
+    },
+    {
+      "epoch": 1.093713393205779,
+      "grad_norm": 2.750613212585449,
+      "learning_rate": 7.26321544098753e-06,
+      "loss": 0.5778,
+      "step": 8403
+    },
+    {
+      "epoch": 1.0941038656774698,
+      "grad_norm": 2.4992191791534424,
+      "learning_rate": 7.261351746732035e-06,
+      "loss": 0.4807,
+      "step": 8406
+    },
+    {
+      "epoch": 1.0944943381491605,
+      "grad_norm": 2.8816723823547363,
+      "learning_rate": 7.2594876574231465e-06,
+      "loss": 0.5364,
+      "step": 8409
+    },
+    {
+      "epoch": 1.0948848106208513,
+      "grad_norm": 2.80865478515625,
+      "learning_rate": 7.257623173386516e-06,
+      "loss": 0.546,
+      "step": 8412
+    },
+    {
+      "epoch": 1.095275283092542,
+      "grad_norm": 2.4753668308258057,
+      "learning_rate": 7.25575829494787e-06,
+      "loss": 0.426,
+      "step": 8415
+    },
+    {
+      "epoch": 1.0956657555642326,
+      "grad_norm": 2.413982629776001,
+      "learning_rate": 7.253893022432993e-06,
+      "loss": 0.4686,
+      "step": 8418
+    },
+    {
+      "epoch": 1.0960562280359234,
+      "grad_norm": 2.514122247695923,
+      "learning_rate": 7.2520273561677455e-06,
+      "loss": 0.5196,
+      "step": 8421
+    },
+    {
+      "epoch": 1.0964467005076142,
+      "grad_norm": 2.6076467037200928,
+      "learning_rate": 7.25016129647806e-06,
+      "loss": 0.5595,
+      "step": 8424
+    },
+    {
+      "epoch": 1.096837172979305,
+      "grad_norm": 3.611821174621582,
+      "learning_rate": 7.248294843689927e-06,
+      "loss": 0.5155,
+      "step": 8427
+    },
+    {
+      "epoch": 1.0972276454509957,
+      "grad_norm": 3.5970561504364014,
+      "learning_rate": 7.246427998129414e-06,
+      "loss": 0.5353,
+      "step": 8430
+    },
+    {
+      "epoch": 1.0976181179226865,
+      "grad_norm": 2.886139392852783,
+      "learning_rate": 7.244560760122656e-06,
+      "loss": 0.5134,
+      "step": 8433
+    },
+    {
+      "epoch": 1.0980085903943773,
+      "grad_norm": 2.6416497230529785,
+      "learning_rate": 7.242693129995857e-06,
+      "loss": 0.4561,
+      "step": 8436
+    },
+    {
+      "epoch": 1.098399062866068,
+      "grad_norm": 2.9228434562683105,
+      "learning_rate": 7.2408251080752845e-06,
+      "loss": 0.4602,
+      "step": 8439
+    },
+    {
+      "epoch": 1.0987895353377586,
+      "grad_norm": 2.7232162952423096,
+      "learning_rate": 7.2389566946872795e-06,
+      "loss": 0.4795,
+      "step": 8442
+    },
+    {
+      "epoch": 1.0991800078094494,
+      "grad_norm": 2.603848934173584,
+      "learning_rate": 7.23708789015825e-06,
+      "loss": 0.5097,
+      "step": 8445
+    },
+    {
+      "epoch": 1.0995704802811401,
+      "grad_norm": 3.2897467613220215,
+      "learning_rate": 7.235218694814673e-06,
+      "loss": 0.5706,
+      "step": 8448
+    },
+    {
+      "epoch": 1.099960952752831,
+      "grad_norm": 2.6159188747406006,
+      "learning_rate": 7.233349108983091e-06,
+      "loss": 0.5107,
+      "step": 8451
+    },
+    {
+      "epoch": 1.1003514252245217,
+      "grad_norm": 2.4405086040496826,
+      "learning_rate": 7.231479132990118e-06,
+      "loss": 0.5245,
+      "step": 8454
+    },
+    {
+      "epoch": 1.1007418976962124,
+      "grad_norm": 2.8570423126220703,
+      "learning_rate": 7.229608767162437e-06,
+      "loss": 0.4532,
+      "step": 8457
+    },
+    {
+      "epoch": 1.1011323701679032,
+      "grad_norm": 3.2515525817871094,
+      "learning_rate": 7.2277380118267924e-06,
+      "loss": 0.5231,
+      "step": 8460
+    },
+    {
+      "epoch": 1.101522842639594,
+      "grad_norm": 2.619053602218628,
+      "learning_rate": 7.2258668673100055e-06,
+      "loss": 0.4558,
+      "step": 8463
+    },
+    {
+      "epoch": 1.1019133151112848,
+      "grad_norm": 3.4879095554351807,
+      "learning_rate": 7.223995333938958e-06,
+      "loss": 0.5,
+      "step": 8466
+    },
+    {
+      "epoch": 1.1023037875829753,
+      "grad_norm": 2.678898334503174,
+      "learning_rate": 7.222123412040605e-06,
+      "loss": 0.5018,
+      "step": 8469
+    },
+    {
+      "epoch": 1.102694260054666,
+      "grad_norm": 2.831430196762085,
+      "learning_rate": 7.220251101941966e-06,
+      "loss": 0.5194,
+      "step": 8472
+    },
+    {
+      "epoch": 1.1030847325263569,
+      "grad_norm": 2.5068278312683105,
+      "learning_rate": 7.21837840397013e-06,
+      "loss": 0.5053,
+      "step": 8475
+    },
+    {
+      "epoch": 1.1034752049980476,
+      "grad_norm": 4.6476898193359375,
+      "learning_rate": 7.216505318452254e-06,
+      "loss": 0.4749,
+      "step": 8478
+    },
+    {
+      "epoch": 1.1038656774697384,
+      "grad_norm": 2.4245364665985107,
+      "learning_rate": 7.214631845715563e-06,
+      "loss": 0.6265,
+      "step": 8481
+    },
+    {
+      "epoch": 1.1042561499414292,
+      "grad_norm": 2.79682993888855,
+      "learning_rate": 7.212757986087346e-06,
+      "loss": 0.5395,
+      "step": 8484
+    },
+    {
+      "epoch": 1.10464662241312,
+      "grad_norm": 4.285904407501221,
+      "learning_rate": 7.210883739894963e-06,
+      "loss": 0.4866,
+      "step": 8487
+    },
+    {
+      "epoch": 1.1050370948848107,
+      "grad_norm": 2.6022517681121826,
+      "learning_rate": 7.209009107465843e-06,
+      "loss": 0.4493,
+      "step": 8490
+    },
+    {
+      "epoch": 1.1054275673565013,
+      "grad_norm": 2.655932664871216,
+      "learning_rate": 7.207134089127479e-06,
+      "loss": 0.4609,
+      "step": 8493
+    },
+    {
+      "epoch": 1.105818039828192,
+      "grad_norm": 2.3036811351776123,
+      "learning_rate": 7.205258685207433e-06,
+      "loss": 0.5362,
+      "step": 8496
+    },
+    {
+      "epoch": 1.1062085122998828,
+      "grad_norm": 3.0361623764038086,
+      "learning_rate": 7.203382896033332e-06,
+      "loss": 0.5391,
+      "step": 8499
+    },
+    {
+      "epoch": 1.1065989847715736,
+      "grad_norm": 3.3767175674438477,
+      "learning_rate": 7.201506721932876e-06,
+      "loss": 0.5089,
+      "step": 8502
+    },
+    {
+      "epoch": 1.1069894572432644,
+      "grad_norm": 2.5257315635681152,
+      "learning_rate": 7.199630163233828e-06,
+      "loss": 0.475,
+      "step": 8505
+    },
+    {
+      "epoch": 1.1073799297149551,
+      "grad_norm": 2.7609949111938477,
+      "learning_rate": 7.197753220264017e-06,
+      "loss": 0.6277,
+      "step": 8508
+    },
+    {
+      "epoch": 1.107770402186646,
+      "grad_norm": 2.56369686126709,
+      "learning_rate": 7.1958758933513405e-06,
+      "loss": 0.4942,
+      "step": 8511
+    },
+    {
+      "epoch": 1.1081608746583367,
+      "grad_norm": 2.836712121963501,
+      "learning_rate": 7.1939981828237646e-06,
+      "loss": 0.4874,
+      "step": 8514
+    },
+    {
+      "epoch": 1.1085513471300272,
+      "grad_norm": 2.9525368213653564,
+      "learning_rate": 7.192120089009322e-06,
+      "loss": 0.5098,
+      "step": 8517
+    },
+    {
+      "epoch": 1.108941819601718,
+      "grad_norm": 2.7775232791900635,
+      "learning_rate": 7.190241612236113e-06,
+      "loss": 0.5331,
+      "step": 8520
+    },
+    {
+      "epoch": 1.1093322920734088,
+      "grad_norm": 4.186012268066406,
+      "learning_rate": 7.188362752832302e-06,
+      "loss": 0.5345,
+      "step": 8523
+    },
+    {
+      "epoch": 1.1097227645450995,
+      "grad_norm": 2.920480728149414,
+      "learning_rate": 7.18648351112612e-06,
+      "loss": 0.5221,
+      "step": 8526
+    },
+    {
+      "epoch": 1.1101132370167903,
+      "grad_norm": 2.599139928817749,
+      "learning_rate": 7.184603887445869e-06,
+      "loss": 0.5322,
+      "step": 8529
+    },
+    {
+      "epoch": 1.110503709488481,
+      "grad_norm": 2.4106132984161377,
+      "learning_rate": 7.182723882119915e-06,
+      "loss": 0.4708,
+      "step": 8532
+    },
+    {
+      "epoch": 1.1108941819601719,
+      "grad_norm": 2.7855916023254395,
+      "learning_rate": 7.18084349547669e-06,
+      "loss": 0.5707,
+      "step": 8535
+    },
+    {
+      "epoch": 1.1112846544318626,
+      "grad_norm": 2.641810655593872,
+      "learning_rate": 7.178962727844694e-06,
+      "loss": 0.5432,
+      "step": 8538
+    },
+    {
+      "epoch": 1.1116751269035534,
+      "grad_norm": 2.4662587642669678,
+      "learning_rate": 7.177081579552494e-06,
+      "loss": 0.5,
+      "step": 8541
+    },
+    {
+      "epoch": 1.112065599375244,
+      "grad_norm": 3.3329203128814697,
+      "learning_rate": 7.175200050928722e-06,
+      "loss": 0.4517,
+      "step": 8544
+    },
+    {
+      "epoch": 1.1124560718469347,
+      "grad_norm": 2.6928961277008057,
+      "learning_rate": 7.173318142302077e-06,
+      "loss": 0.4792,
+      "step": 8547
+    },
+    {
+      "epoch": 1.1128465443186255,
+      "grad_norm": 3.24125075340271,
+      "learning_rate": 7.171435854001324e-06,
+      "loss": 0.5087,
+      "step": 8550
+    },
+    {
+      "epoch": 1.1132370167903163,
+      "grad_norm": 2.6062655448913574,
+      "learning_rate": 7.169553186355296e-06,
+      "loss": 0.5109,
+      "step": 8553
+    },
+    {
+      "epoch": 1.113627489262007,
+      "grad_norm": 3.467337131500244,
+      "learning_rate": 7.167670139692888e-06,
+      "loss": 0.5299,
+      "step": 8556
+    },
+    {
+      "epoch": 1.1140179617336978,
+      "grad_norm": 2.869731903076172,
+      "learning_rate": 7.16578671434307e-06,
+      "loss": 0.4565,
+      "step": 8559
+    },
+    {
+      "epoch": 1.1144084342053886,
+      "grad_norm": 2.4524435997009277,
+      "learning_rate": 7.163902910634868e-06,
+      "loss": 0.4392,
+      "step": 8562
+    },
+    {
+      "epoch": 1.1147989066770791,
+      "grad_norm": 4.4590067863464355,
+      "learning_rate": 7.16201872889738e-06,
+      "loss": 0.413,
+      "step": 8565
+    },
+    {
+      "epoch": 1.11518937914877,
+      "grad_norm": 2.6730213165283203,
+      "learning_rate": 7.160134169459768e-06,
+      "loss": 0.4561,
+      "step": 8568
+    },
+    {
+      "epoch": 1.1155798516204607,
+      "grad_norm": 2.4888358116149902,
+      "learning_rate": 7.15824923265126e-06,
+      "loss": 0.5105,
+      "step": 8571
+    },
+    {
+      "epoch": 1.1159703240921515,
+      "grad_norm": 2.672909736633301,
+      "learning_rate": 7.156363918801152e-06,
+      "loss": 0.5565,
+      "step": 8574
+    },
+    {
+      "epoch": 1.1163607965638422,
+      "grad_norm": 2.708988666534424,
+      "learning_rate": 7.154478228238804e-06,
+      "loss": 0.556,
+      "step": 8577
+    },
+    {
+      "epoch": 1.116751269035533,
+      "grad_norm": 2.6112136840820312,
+      "learning_rate": 7.152592161293642e-06,
+      "loss": 0.5554,
+      "step": 8580
+    },
+    {
+      "epoch": 1.1171417415072238,
+      "grad_norm": 2.5211217403411865,
+      "learning_rate": 7.150705718295157e-06,
+      "loss": 0.4915,
+      "step": 8583
+    },
+    {
+      "epoch": 1.1175322139789146,
+      "grad_norm": 3.3289098739624023,
+      "learning_rate": 7.1488188995729095e-06,
+      "loss": 0.4083,
+      "step": 8586
+    },
+    {
+      "epoch": 1.1179226864506053,
+      "grad_norm": 2.5261623859405518,
+      "learning_rate": 7.146931705456518e-06,
+      "loss": 0.4479,
+      "step": 8589
+    },
+    {
+      "epoch": 1.1183131589222959,
+      "grad_norm": 2.5002522468566895,
+      "learning_rate": 7.145044136275675e-06,
+      "loss": 0.5424,
+      "step": 8592
+    },
+    {
+      "epoch": 1.1187036313939867,
+      "grad_norm": 2.2631165981292725,
+      "learning_rate": 7.143156192360135e-06,
+      "loss": 0.4406,
+      "step": 8595
+    },
+    {
+      "epoch": 1.1190941038656774,
+      "grad_norm": 2.8084895610809326,
+      "learning_rate": 7.141267874039715e-06,
+      "loss": 0.4845,
+      "step": 8598
+    },
+    {
+      "epoch": 1.1194845763373682,
+      "grad_norm": 2.751574993133545,
+      "learning_rate": 7.139379181644304e-06,
+      "loss": 0.4932,
+      "step": 8601
+    },
+    {
+      "epoch": 1.119875048809059,
+      "grad_norm": 2.6821987628936768,
+      "learning_rate": 7.137490115503848e-06,
+      "loss": 0.4911,
+      "step": 8604
+    },
+    {
+      "epoch": 1.1202655212807497,
+      "grad_norm": 2.521465301513672,
+      "learning_rate": 7.135600675948367e-06,
+      "loss": 0.5492,
+      "step": 8607
+    },
+    {
+      "epoch": 1.1206559937524405,
+      "grad_norm": 2.7268893718719482,
+      "learning_rate": 7.133710863307941e-06,
+      "loss": 0.5966,
+      "step": 8610
+    },
+    {
+      "epoch": 1.1210464662241313,
+      "grad_norm": 3.1640634536743164,
+      "learning_rate": 7.131820677912715e-06,
+      "loss": 0.5761,
+      "step": 8613
+    },
+    {
+      "epoch": 1.121436938695822,
+      "grad_norm": 2.8102428913116455,
+      "learning_rate": 7.129930120092902e-06,
+      "loss": 0.4964,
+      "step": 8616
+    },
+    {
+      "epoch": 1.1218274111675126,
+      "grad_norm": 2.8150181770324707,
+      "learning_rate": 7.128039190178776e-06,
+      "loss": 0.5629,
+      "step": 8619
+    },
+    {
+      "epoch": 1.1222178836392034,
+      "grad_norm": 2.691941976547241,
+      "learning_rate": 7.1261478885006815e-06,
+      "loss": 0.4933,
+      "step": 8622
+    },
+    {
+      "epoch": 1.1226083561108942,
+      "grad_norm": 3.071762800216675,
+      "learning_rate": 7.124256215389023e-06,
+      "loss": 0.5538,
+      "step": 8625
+    },
+    {
+      "epoch": 1.122998828582585,
+      "grad_norm": 3.8000946044921875,
+      "learning_rate": 7.122364171174273e-06,
+      "loss": 0.5261,
+      "step": 8628
+    },
+    {
+      "epoch": 1.1233893010542757,
+      "grad_norm": 2.708369493484497,
+      "learning_rate": 7.1204717561869684e-06,
+      "loss": 0.537,
+      "step": 8631
+    },
+    {
+      "epoch": 1.1237797735259665,
+      "grad_norm": 2.911895275115967,
+      "learning_rate": 7.118578970757707e-06,
+      "loss": 0.489,
+      "step": 8634
+    },
+    {
+      "epoch": 1.1241702459976572,
+      "grad_norm": 2.5384697914123535,
+      "learning_rate": 7.116685815217157e-06,
+      "loss": 0.4744,
+      "step": 8637
+    },
+    {
+      "epoch": 1.1245607184693478,
+      "grad_norm": 2.503662586212158,
+      "learning_rate": 7.114792289896046e-06,
+      "loss": 0.5045,
+      "step": 8640
+    },
+    {
+      "epoch": 1.1249511909410386,
+      "grad_norm": 2.787290334701538,
+      "learning_rate": 7.112898395125174e-06,
+      "loss": 0.471,
+      "step": 8643
+    },
+    {
+      "epoch": 1.1253416634127293,
+      "grad_norm": 2.368600368499756,
+      "learning_rate": 7.111004131235396e-06,
+      "loss": 0.4786,
+      "step": 8646
+    },
+    {
+      "epoch": 1.1257321358844201,
+      "grad_norm": 2.707949638366699,
+      "learning_rate": 7.109109498557636e-06,
+      "loss": 0.5496,
+      "step": 8649
+    },
+    {
+      "epoch": 1.1261226083561109,
+      "grad_norm": 2.765071392059326,
+      "learning_rate": 7.107214497422885e-06,
+      "loss": 0.5355,
+      "step": 8652
+    },
+    {
+      "epoch": 1.1265130808278017,
+      "grad_norm": 3.1850650310516357,
+      "learning_rate": 7.105319128162194e-06,
+      "loss": 0.4455,
+      "step": 8655
+    },
+    {
+      "epoch": 1.1269035532994924,
+      "grad_norm": 2.8872485160827637,
+      "learning_rate": 7.10342339110668e-06,
+      "loss": 0.5017,
+      "step": 8658
+    },
+    {
+      "epoch": 1.1272940257711832,
+      "grad_norm": 3.188868761062622,
+      "learning_rate": 7.101527286587524e-06,
+      "loss": 0.52,
+      "step": 8661
+    },
+    {
+      "epoch": 1.127684498242874,
+      "grad_norm": 2.5735905170440674,
+      "learning_rate": 7.099630814935973e-06,
+      "loss": 0.513,
+      "step": 8664
+    },
+    {
+      "epoch": 1.1280749707145645,
+      "grad_norm": 2.725409507751465,
+      "learning_rate": 7.097733976483335e-06,
+      "loss": 0.5538,
+      "step": 8667
+    },
+    {
+      "epoch": 1.1284654431862553,
+      "grad_norm": 2.974015235900879,
+      "learning_rate": 7.095836771560984e-06,
+      "loss": 0.5158,
+      "step": 8670
+    },
+    {
+      "epoch": 1.128855915657946,
+      "grad_norm": 2.86106276512146,
+      "learning_rate": 7.093939200500359e-06,
+      "loss": 0.5367,
+      "step": 8673
+    },
+    {
+      "epoch": 1.1292463881296368,
+      "grad_norm": 3.0025503635406494,
+      "learning_rate": 7.092041263632961e-06,
+      "loss": 0.5411,
+      "step": 8676
+    },
+    {
+      "epoch": 1.1296368606013276,
+      "grad_norm": 3.5087170600891113,
+      "learning_rate": 7.090142961290354e-06,
+      "loss": 0.564,
+      "step": 8679
+    },
+    {
+      "epoch": 1.1300273330730184,
+      "grad_norm": 2.5758073329925537,
+      "learning_rate": 7.088244293804169e-06,
+      "loss": 0.4342,
+      "step": 8682
+    },
+    {
+      "epoch": 1.1304178055447092,
+      "grad_norm": 2.531656503677368,
+      "learning_rate": 7.086345261506098e-06,
+      "loss": 0.5625,
+      "step": 8685
+    },
+    {
+      "epoch": 1.1308082780164,
+      "grad_norm": 2.7261338233947754,
+      "learning_rate": 7.0844458647279e-06,
+      "loss": 0.4574,
+      "step": 8688
+    },
+    {
+      "epoch": 1.1311987504880907,
+      "grad_norm": 2.582547664642334,
+      "learning_rate": 7.082546103801394e-06,
+      "loss": 0.5195,
+      "step": 8691
+    },
+    {
+      "epoch": 1.1315892229597813,
+      "grad_norm": 2.5053184032440186,
+      "learning_rate": 7.080645979058466e-06,
+      "loss": 0.4558,
+      "step": 8694
+    },
+    {
+      "epoch": 1.131979695431472,
+      "grad_norm": 2.540330410003662,
+      "learning_rate": 7.0787454908310614e-06,
+      "loss": 0.5509,
+      "step": 8697
+    },
+    {
+      "epoch": 1.1323701679031628,
+      "grad_norm": 2.711557149887085,
+      "learning_rate": 7.076844639451193e-06,
+      "loss": 0.5265,
+      "step": 8700
+    },
+    {
+      "epoch": 1.1327606403748536,
+      "grad_norm": 2.436702013015747,
+      "learning_rate": 7.074943425250933e-06,
+      "loss": 0.4933,
+      "step": 8703
+    },
+    {
+      "epoch": 1.1331511128465444,
+      "grad_norm": 2.6545193195343018,
+      "learning_rate": 7.073041848562424e-06,
+      "loss": 0.4736,
+      "step": 8706
+    },
+    {
+      "epoch": 1.1335415853182351,
+      "grad_norm": 2.6401283740997314,
+      "learning_rate": 7.071139909717865e-06,
+      "loss": 0.5213,
+      "step": 8709
+    },
+    {
+      "epoch": 1.133932057789926,
+      "grad_norm": 2.585341453552246,
+      "learning_rate": 7.06923760904952e-06,
+      "loss": 0.5262,
+      "step": 8712
+    },
+    {
+      "epoch": 1.1343225302616164,
+      "grad_norm": 2.6922566890716553,
+      "learning_rate": 7.067334946889718e-06,
+      "loss": 0.5317,
+      "step": 8715
+    },
+    {
+      "epoch": 1.1347130027333072,
+      "grad_norm": 2.583522081375122,
+      "learning_rate": 7.06543192357085e-06,
+      "loss": 0.4766,
+      "step": 8718
+    },
+    {
+      "epoch": 1.135103475204998,
+      "grad_norm": 2.4560515880584717,
+      "learning_rate": 7.06352853942537e-06,
+      "loss": 0.5026,
+      "step": 8721
+    },
+    {
+      "epoch": 1.1354939476766888,
+      "grad_norm": 2.9161624908447266,
+      "learning_rate": 7.061624794785795e-06,
+      "loss": 0.5901,
+      "step": 8724
+    },
+    {
+      "epoch": 1.1358844201483795,
+      "grad_norm": 2.5634748935699463,
+      "learning_rate": 7.059720689984705e-06,
+      "loss": 0.496,
+      "step": 8727
+    },
+    {
+      "epoch": 1.1362748926200703,
+      "grad_norm": 2.5747694969177246,
+      "learning_rate": 7.0578162253547445e-06,
+      "loss": 0.5148,
+      "step": 8730
+    },
+    {
+      "epoch": 1.136665365091761,
+      "grad_norm": 2.320279121398926,
+      "learning_rate": 7.055911401228618e-06,
+      "loss": 0.4465,
+      "step": 8733
+    },
+    {
+      "epoch": 1.1370558375634519,
+      "grad_norm": 2.7718331813812256,
+      "learning_rate": 7.054006217939093e-06,
+      "loss": 0.6083,
+      "step": 8736
+    },
+    {
+      "epoch": 1.1374463100351426,
+      "grad_norm": 2.2558786869049072,
+      "learning_rate": 7.052100675819006e-06,
+      "loss": 0.4439,
+      "step": 8739
+    },
+    {
+      "epoch": 1.1378367825068332,
+      "grad_norm": 2.8862807750701904,
+      "learning_rate": 7.050194775201246e-06,
+      "loss": 0.5045,
+      "step": 8742
+    },
+    {
+      "epoch": 1.138227254978524,
+      "grad_norm": 2.867145538330078,
+      "learning_rate": 7.048288516418772e-06,
+      "loss": 0.5259,
+      "step": 8745
+    },
+    {
+      "epoch": 1.1386177274502147,
+      "grad_norm": 2.575711965560913,
+      "learning_rate": 7.046381899804602e-06,
+      "loss": 0.562,
+      "step": 8748
+    },
+    {
+      "epoch": 1.1390081999219055,
+      "grad_norm": 2.9559128284454346,
+      "learning_rate": 7.044474925691821e-06,
+      "loss": 0.5471,
+      "step": 8751
+    },
+    {
+      "epoch": 1.1393986723935963,
+      "grad_norm": 2.526271343231201,
+      "learning_rate": 7.042567594413571e-06,
+      "loss": 0.5216,
+      "step": 8754
+    },
+    {
+      "epoch": 1.139789144865287,
+      "grad_norm": 2.697111129760742,
+      "learning_rate": 7.040659906303058e-06,
+      "loss": 0.4498,
+      "step": 8757
+    },
+    {
+      "epoch": 1.1401796173369778,
+      "grad_norm": 2.5852532386779785,
+      "learning_rate": 7.038751861693553e-06,
+      "loss": 0.4273,
+      "step": 8760
+    },
+    {
+      "epoch": 1.1405700898086686,
+      "grad_norm": 2.4026541709899902,
+      "learning_rate": 7.036843460918388e-06,
+      "loss": 0.4601,
+      "step": 8763
+    },
+    {
+      "epoch": 1.1409605622803594,
+      "grad_norm": 3.018265962600708,
+      "learning_rate": 7.034934704310954e-06,
+      "loss": 0.537,
+      "step": 8766
+    },
+    {
+      "epoch": 1.14135103475205,
+      "grad_norm": 2.6514596939086914,
+      "learning_rate": 7.0330255922047075e-06,
+      "loss": 0.4801,
+      "step": 8769
+    },
+    {
+      "epoch": 1.1417415072237407,
+      "grad_norm": 2.793213129043579,
+      "learning_rate": 7.031116124933167e-06,
+      "loss": 0.5271,
+      "step": 8772
+    },
+    {
+      "epoch": 1.1421319796954315,
+      "grad_norm": 3.0178632736206055,
+      "learning_rate": 7.029206302829914e-06,
+      "loss": 0.4884,
+      "step": 8775
+    },
+    {
+      "epoch": 1.1425224521671222,
+      "grad_norm": 2.857133388519287,
+      "learning_rate": 7.027296126228586e-06,
+      "loss": 0.5116,
+      "step": 8778
+    },
+    {
+      "epoch": 1.142912924638813,
+      "grad_norm": 2.7881019115448,
+      "learning_rate": 7.0253855954628925e-06,
+      "loss": 0.4901,
+      "step": 8781
+    },
+    {
+      "epoch": 1.1433033971105038,
+      "grad_norm": 2.7480220794677734,
+      "learning_rate": 7.023474710866595e-06,
+      "loss": 0.4977,
+      "step": 8784
+    },
+    {
+      "epoch": 1.1436938695821945,
+      "grad_norm": 2.6863813400268555,
+      "learning_rate": 7.021563472773522e-06,
+      "loss": 0.476,
+      "step": 8787
+    },
+    {
+      "epoch": 1.144084342053885,
+      "grad_norm": 2.9496448040008545,
+      "learning_rate": 7.019651881517562e-06,
+      "loss": 0.4938,
+      "step": 8790
+    },
+    {
+      "epoch": 1.1444748145255759,
+      "grad_norm": 2.909560203552246,
+      "learning_rate": 7.017739937432668e-06,
+      "loss": 0.4599,
+      "step": 8793
+    },
+    {
+      "epoch": 1.1448652869972666,
+      "grad_norm": 3.5830419063568115,
+      "learning_rate": 7.015827640852852e-06,
+      "loss": 0.5422,
+      "step": 8796
+    },
+    {
+      "epoch": 1.1452557594689574,
+      "grad_norm": 3.0168848037719727,
+      "learning_rate": 7.013914992112187e-06,
+      "loss": 0.4376,
+      "step": 8799
+    },
+    {
+      "epoch": 1.1456462319406482,
+      "grad_norm": 2.4287290573120117,
+      "learning_rate": 7.0120019915448125e-06,
+      "loss": 0.4951,
+      "step": 8802
+    },
+    {
+      "epoch": 1.146036704412339,
+      "grad_norm": 2.5160810947418213,
+      "learning_rate": 7.01008863948492e-06,
+      "loss": 0.5767,
+      "step": 8805
+    },
+    {
+      "epoch": 1.1464271768840297,
+      "grad_norm": 2.722810745239258,
+      "learning_rate": 7.00817493626677e-06,
+      "loss": 0.5479,
+      "step": 8808
+    },
+    {
+      "epoch": 1.1468176493557205,
+      "grad_norm": 3.0669102668762207,
+      "learning_rate": 7.006260882224684e-06,
+      "loss": 0.6298,
+      "step": 8811
+    },
+    {
+      "epoch": 1.1472081218274113,
+      "grad_norm": 2.84295392036438,
+      "learning_rate": 7.004346477693042e-06,
+      "loss": 0.4986,
+      "step": 8814
+    },
+    {
+      "epoch": 1.1475985942991018,
+      "grad_norm": 2.5158627033233643,
+      "learning_rate": 7.0024317230062884e-06,
+      "loss": 0.5512,
+      "step": 8817
+    },
+    {
+      "epoch": 1.1479890667707926,
+      "grad_norm": 2.525624990463257,
+      "learning_rate": 7.0005166184989245e-06,
+      "loss": 0.4828,
+      "step": 8820
+    },
+    {
+      "epoch": 1.1483795392424834,
+      "grad_norm": 3.02347993850708,
+      "learning_rate": 6.9986011645055175e-06,
+      "loss": 0.483,
+      "step": 8823
+    },
+    {
+      "epoch": 1.1487700117141741,
+      "grad_norm": 3.2486398220062256,
+      "learning_rate": 6.99668536136069e-06,
+      "loss": 0.6375,
+      "step": 8826
+    },
+    {
+      "epoch": 1.149160484185865,
+      "grad_norm": 3.728907823562622,
+      "learning_rate": 6.9947692093991295e-06,
+      "loss": 0.5128,
+      "step": 8829
+    },
+    {
+      "epoch": 1.1495509566575557,
+      "grad_norm": 3.2733185291290283,
+      "learning_rate": 6.992852708955586e-06,
+      "loss": 0.5249,
+      "step": 8832
+    },
+    {
+      "epoch": 1.1499414291292465,
+      "grad_norm": 2.754006862640381,
+      "learning_rate": 6.990935860364865e-06,
+      "loss": 0.5313,
+      "step": 8835
+    },
+    {
+      "epoch": 1.150331901600937,
+      "grad_norm": 2.4702494144439697,
+      "learning_rate": 6.989018663961838e-06,
+      "loss": 0.4722,
+      "step": 8838
+    },
+    {
+      "epoch": 1.1507223740726278,
+      "grad_norm": 2.463707208633423,
+      "learning_rate": 6.987101120081436e-06,
+      "loss": 0.4612,
+      "step": 8841
+    },
+    {
+      "epoch": 1.1511128465443186,
+      "grad_norm": 2.601174831390381,
+      "learning_rate": 6.9851832290586465e-06,
+      "loss": 0.4494,
+      "step": 8844
+    },
+    {
+      "epoch": 1.1515033190160093,
+      "grad_norm": 2.667945623397827,
+      "learning_rate": 6.983264991228525e-06,
+      "loss": 0.4577,
+      "step": 8847
+    },
+    {
+      "epoch": 1.1518937914877,
+      "grad_norm": 2.5175373554229736,
+      "learning_rate": 6.981346406926179e-06,
+      "loss": 0.5165,
+      "step": 8850
+    },
+    {
+      "epoch": 1.1522842639593909,
+      "grad_norm": 2.911750078201294,
+      "learning_rate": 6.979427476486786e-06,
+      "loss": 0.5315,
+      "step": 8853
+    },
+    {
+      "epoch": 1.1526747364310816,
+      "grad_norm": 2.7814128398895264,
+      "learning_rate": 6.9775082002455775e-06,
+      "loss": 0.4972,
+      "step": 8856
+    },
+    {
+      "epoch": 1.1530652089027724,
+      "grad_norm": 2.6590986251831055,
+      "learning_rate": 6.975588578537846e-06,
+      "loss": 0.553,
+      "step": 8859
+    },
+    {
+      "epoch": 1.1534556813744632,
+      "grad_norm": 2.5675268173217773,
+      "learning_rate": 6.973668611698945e-06,
+      "loss": 0.4934,
+      "step": 8862
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 2.761390447616577,
+      "learning_rate": 6.971748300064291e-06,
+      "loss": 0.6005,
+      "step": 8865
+    },
+    {
+      "epoch": 1.1542366263178445,
+      "grad_norm": 2.5827791690826416,
+      "learning_rate": 6.969827643969356e-06,
+      "loss": 0.4941,
+      "step": 8868
+    },
+    {
+      "epoch": 1.1546270987895353,
+      "grad_norm": 2.479886293411255,
+      "learning_rate": 6.9679066437496744e-06,
+      "loss": 0.4605,
+      "step": 8871
+    },
+    {
+      "epoch": 1.155017571261226,
+      "grad_norm": 2.5713136196136475,
+      "learning_rate": 6.965985299740844e-06,
+      "loss": 0.567,
+      "step": 8874
+    },
+    {
+      "epoch": 1.1554080437329168,
+      "grad_norm": 2.7975733280181885,
+      "learning_rate": 6.964063612278517e-06,
+      "loss": 0.5209,
+      "step": 8877
+    },
+    {
+      "epoch": 1.1557985162046076,
+      "grad_norm": 2.6016461849212646,
+      "learning_rate": 6.962141581698408e-06,
+      "loss": 0.4932,
+      "step": 8880
+    },
+    {
+      "epoch": 1.1561889886762984,
+      "grad_norm": 3.229020357131958,
+      "learning_rate": 6.960219208336293e-06,
+      "loss": 0.5739,
+      "step": 8883
+    },
+    {
+      "epoch": 1.1565794611479892,
+      "grad_norm": 2.7121598720550537,
+      "learning_rate": 6.958296492528005e-06,
+      "loss": 0.5604,
+      "step": 8886
+    },
+    {
+      "epoch": 1.15696993361968,
+      "grad_norm": 3.215475082397461,
+      "learning_rate": 6.956373434609441e-06,
+      "loss": 0.5234,
+      "step": 8889
+    },
+    {
+      "epoch": 1.1573604060913705,
+      "grad_norm": 2.619798421859741,
+      "learning_rate": 6.954450034916552e-06,
+      "loss": 0.4628,
+      "step": 8892
+    },
+    {
+      "epoch": 1.1577508785630612,
+      "grad_norm": 2.3444454669952393,
+      "learning_rate": 6.952526293785356e-06,
+      "loss": 0.4755,
+      "step": 8895
+    },
+    {
+      "epoch": 1.158141351034752,
+      "grad_norm": 2.6880340576171875,
+      "learning_rate": 6.950602211551921e-06,
+      "loss": 0.4455,
+      "step": 8898
+    },
+    {
+      "epoch": 1.1585318235064428,
+      "grad_norm": 2.406294345855713,
+      "learning_rate": 6.948677788552384e-06,
+      "loss": 0.495,
+      "step": 8901
+    },
+    {
+      "epoch": 1.1589222959781336,
+      "grad_norm": 3.347994327545166,
+      "learning_rate": 6.946753025122938e-06,
+      "loss": 0.4894,
+      "step": 8904
+    },
+    {
+      "epoch": 1.1593127684498243,
+      "grad_norm": 2.817830801010132,
+      "learning_rate": 6.944827921599832e-06,
+      "loss": 0.5655,
+      "step": 8907
+    },
+    {
+      "epoch": 1.159703240921515,
+      "grad_norm": 3.159911870956421,
+      "learning_rate": 6.942902478319382e-06,
+      "loss": 0.5178,
+      "step": 8910
+    },
+    {
+      "epoch": 1.1600937133932057,
+      "grad_norm": 2.5966691970825195,
+      "learning_rate": 6.940976695617954e-06,
+      "loss": 0.5147,
+      "step": 8913
+    },
+    {
+      "epoch": 1.1604841858648964,
+      "grad_norm": 2.7088546752929688,
+      "learning_rate": 6.9390505738319815e-06,
+      "loss": 0.472,
+      "step": 8916
+    },
+    {
+      "epoch": 1.1608746583365872,
+      "grad_norm": 2.583479404449463,
+      "learning_rate": 6.937124113297953e-06,
+      "loss": 0.4644,
+      "step": 8919
+    },
+    {
+      "epoch": 1.161265130808278,
+      "grad_norm": 3.7508015632629395,
+      "learning_rate": 6.935197314352415e-06,
+      "loss": 0.502,
+      "step": 8922
+    },
+    {
+      "epoch": 1.1616556032799688,
+      "grad_norm": 2.6737987995147705,
+      "learning_rate": 6.93327017733198e-06,
+      "loss": 0.5079,
+      "step": 8925
+    },
+    {
+      "epoch": 1.1620460757516595,
+      "grad_norm": 2.8131909370422363,
+      "learning_rate": 6.93134270257331e-06,
+      "loss": 0.5143,
+      "step": 8928
+    },
+    {
+      "epoch": 1.1624365482233503,
+      "grad_norm": 3.131113290786743,
+      "learning_rate": 6.929414890413135e-06,
+      "loss": 0.4909,
+      "step": 8931
+    },
+    {
+      "epoch": 1.162827020695041,
+      "grad_norm": 2.632091522216797,
+      "learning_rate": 6.9274867411882355e-06,
+      "loss": 0.4803,
+      "step": 8934
+    },
+    {
+      "epoch": 1.1632174931667318,
+      "grad_norm": 3.4867727756500244,
+      "learning_rate": 6.925558255235458e-06,
+      "loss": 0.5363,
+      "step": 8937
+    },
+    {
+      "epoch": 1.1636079656384224,
+      "grad_norm": 2.5351741313934326,
+      "learning_rate": 6.923629432891704e-06,
+      "loss": 0.465,
+      "step": 8940
+    },
+    {
+      "epoch": 1.1639984381101132,
+      "grad_norm": 2.7857537269592285,
+      "learning_rate": 6.921700274493935e-06,
+      "loss": 0.5723,
+      "step": 8943
+    },
+    {
+      "epoch": 1.164388910581804,
+      "grad_norm": 2.703540802001953,
+      "learning_rate": 6.9197707803791714e-06,
+      "loss": 0.5018,
+      "step": 8946
+    },
+    {
+      "epoch": 1.1647793830534947,
+      "grad_norm": 2.7754833698272705,
+      "learning_rate": 6.917840950884489e-06,
+      "loss": 0.4933,
+      "step": 8949
+    },
+    {
+      "epoch": 1.1651698555251855,
+      "grad_norm": 2.64084792137146,
+      "learning_rate": 6.915910786347029e-06,
+      "loss": 0.5401,
+      "step": 8952
+    },
+    {
+      "epoch": 1.1655603279968763,
+      "grad_norm": 2.382725715637207,
+      "learning_rate": 6.913980287103984e-06,
+      "loss": 0.4424,
+      "step": 8955
+    },
+    {
+      "epoch": 1.165950800468567,
+      "grad_norm": 2.396775245666504,
+      "learning_rate": 6.912049453492609e-06,
+      "loss": 0.4835,
+      "step": 8958
+    },
+    {
+      "epoch": 1.1663412729402578,
+      "grad_norm": 2.7328484058380127,
+      "learning_rate": 6.910118285850218e-06,
+      "loss": 0.5132,
+      "step": 8961
+    },
+    {
+      "epoch": 1.1667317454119486,
+      "grad_norm": 2.620654821395874,
+      "learning_rate": 6.90818678451418e-06,
+      "loss": 0.4648,
+      "step": 8964
+    },
+    {
+      "epoch": 1.1671222178836391,
+      "grad_norm": 3.5922820568084717,
+      "learning_rate": 6.906254949821926e-06,
+      "loss": 0.5845,
+      "step": 8967
+    },
+    {
+      "epoch": 1.16751269035533,
+      "grad_norm": 2.563399314880371,
+      "learning_rate": 6.904322782110942e-06,
+      "loss": 0.595,
+      "step": 8970
+    },
+    {
+      "epoch": 1.1679031628270207,
+      "grad_norm": 3.2699246406555176,
+      "learning_rate": 6.9023902817187735e-06,
+      "loss": 0.4712,
+      "step": 8973
+    },
+    {
+      "epoch": 1.1682936352987114,
+      "grad_norm": 2.6587018966674805,
+      "learning_rate": 6.900457448983024e-06,
+      "loss": 0.5394,
+      "step": 8976
+    },
+    {
+      "epoch": 1.1686841077704022,
+      "grad_norm": 3.6553778648376465,
+      "learning_rate": 6.898524284241357e-06,
+      "loss": 0.5078,
+      "step": 8979
+    },
+    {
+      "epoch": 1.169074580242093,
+      "grad_norm": 2.561161994934082,
+      "learning_rate": 6.896590787831493e-06,
+      "loss": 0.5351,
+      "step": 8982
+    },
+    {
+      "epoch": 1.1694650527137838,
+      "grad_norm": 3.0311195850372314,
+      "learning_rate": 6.894656960091206e-06,
+      "loss": 0.5395,
+      "step": 8985
+    },
+    {
+      "epoch": 1.1698555251854743,
+      "grad_norm": 2.600670337677002,
+      "learning_rate": 6.892722801358336e-06,
+      "loss": 0.4744,
+      "step": 8988
+    },
+    {
+      "epoch": 1.170245997657165,
+      "grad_norm": 2.8165881633758545,
+      "learning_rate": 6.890788311970773e-06,
+      "loss": 0.5708,
+      "step": 8991
+    },
+    {
+      "epoch": 1.1706364701288559,
+      "grad_norm": 3.0961506366729736,
+      "learning_rate": 6.888853492266469e-06,
+      "loss": 0.4359,
+      "step": 8994
+    },
+    {
+      "epoch": 1.1710269426005466,
+      "grad_norm": 2.740213394165039,
+      "learning_rate": 6.886918342583433e-06,
+      "loss": 0.4675,
+      "step": 8997
+    },
+    {
+      "epoch": 1.1714174150722374,
+      "grad_norm": 2.547696828842163,
+      "learning_rate": 6.884982863259734e-06,
+      "loss": 0.544,
+      "step": 9000
+    },
+    {
+      "epoch": 1.1718078875439282,
+      "grad_norm": 2.924173355102539,
+      "learning_rate": 6.883047054633494e-06,
+      "loss": 0.4527,
+      "step": 9003
+    },
+    {
+      "epoch": 1.172198360015619,
+      "grad_norm": 2.757521629333496,
+      "learning_rate": 6.8811109170428935e-06,
+      "loss": 0.4779,
+      "step": 9006
+    },
+    {
+      "epoch": 1.1725888324873097,
+      "grad_norm": 2.3954832553863525,
+      "learning_rate": 6.8791744508261735e-06,
+      "loss": 0.4461,
+      "step": 9009
+    },
+    {
+      "epoch": 1.1729793049590005,
+      "grad_norm": 2.38798189163208,
+      "learning_rate": 6.877237656321631e-06,
+      "loss": 0.5077,
+      "step": 9012
+    },
+    {
+      "epoch": 1.173369777430691,
+      "grad_norm": 2.7606191635131836,
+      "learning_rate": 6.875300533867619e-06,
+      "loss": 0.5737,
+      "step": 9015
+    },
+    {
+      "epoch": 1.1737602499023818,
+      "grad_norm": 2.400142192840576,
+      "learning_rate": 6.873363083802547e-06,
+      "loss": 0.5388,
+      "step": 9018
+    },
+    {
+      "epoch": 1.1741507223740726,
+      "grad_norm": 3.098991632461548,
+      "learning_rate": 6.8714253064648865e-06,
+      "loss": 0.5075,
+      "step": 9021
+    },
+    {
+      "epoch": 1.1745411948457634,
+      "grad_norm": 2.514508008956909,
+      "learning_rate": 6.8694872021931625e-06,
+      "loss": 0.5503,
+      "step": 9024
+    },
+    {
+      "epoch": 1.1749316673174541,
+      "grad_norm": 2.7697696685791016,
+      "learning_rate": 6.867548771325956e-06,
+      "loss": 0.5224,
+      "step": 9027
+    },
+    {
+      "epoch": 1.175322139789145,
+      "grad_norm": 2.679353952407837,
+      "learning_rate": 6.865610014201909e-06,
+      "loss": 0.5027,
+      "step": 9030
+    },
+    {
+      "epoch": 1.1757126122608357,
+      "grad_norm": 2.770353317260742,
+      "learning_rate": 6.863670931159716e-06,
+      "loss": 0.4884,
+      "step": 9033
+    },
+    {
+      "epoch": 1.1761030847325264,
+      "grad_norm": 2.543590784072876,
+      "learning_rate": 6.861731522538133e-06,
+      "loss": 0.4744,
+      "step": 9036
+    },
+    {
+      "epoch": 1.1764935572042172,
+      "grad_norm": 3.0317022800445557,
+      "learning_rate": 6.859791788675969e-06,
+      "loss": 0.5167,
+      "step": 9039
+    },
+    {
+      "epoch": 1.1768840296759078,
+      "grad_norm": 2.7118606567382812,
+      "learning_rate": 6.8578517299120916e-06,
+      "loss": 0.5217,
+      "step": 9042
+    },
+    {
+      "epoch": 1.1772745021475985,
+      "grad_norm": 2.709425449371338,
+      "learning_rate": 6.855911346585427e-06,
+      "loss": 0.5527,
+      "step": 9045
+    },
+    {
+      "epoch": 1.1776649746192893,
+      "grad_norm": 2.604459762573242,
+      "learning_rate": 6.853970639034953e-06,
+      "loss": 0.4463,
+      "step": 9048
+    },
+    {
+      "epoch": 1.17805544709098,
+      "grad_norm": 2.796820878982544,
+      "learning_rate": 6.852029607599707e-06,
+      "loss": 0.5053,
+      "step": 9051
+    },
+    {
+      "epoch": 1.1784459195626709,
+      "grad_norm": 3.0297458171844482,
+      "learning_rate": 6.850088252618787e-06,
+      "loss": 0.5075,
+      "step": 9054
+    },
+    {
+      "epoch": 1.1788363920343616,
+      "grad_norm": 2.483165979385376,
+      "learning_rate": 6.84814657443134e-06,
+      "loss": 0.4528,
+      "step": 9057
+    },
+    {
+      "epoch": 1.1792268645060524,
+      "grad_norm": 2.606029987335205,
+      "learning_rate": 6.846204573376576e-06,
+      "loss": 0.4552,
+      "step": 9060
+    },
+    {
+      "epoch": 1.179617336977743,
+      "grad_norm": 2.4798717498779297,
+      "learning_rate": 6.844262249793755e-06,
+      "loss": 0.514,
+      "step": 9063
+    },
+    {
+      "epoch": 1.1800078094494337,
+      "grad_norm": 2.535163164138794,
+      "learning_rate": 6.842319604022201e-06,
+      "loss": 0.5319,
+      "step": 9066
+    },
+    {
+      "epoch": 1.1803982819211245,
+      "grad_norm": 2.676571846008301,
+      "learning_rate": 6.840376636401285e-06,
+      "loss": 0.4704,
+      "step": 9069
+    },
+    {
+      "epoch": 1.1807887543928153,
+      "grad_norm": 2.842242956161499,
+      "learning_rate": 6.838433347270444e-06,
+      "loss": 0.496,
+      "step": 9072
+    },
+    {
+      "epoch": 1.181179226864506,
+      "grad_norm": 2.5695507526397705,
+      "learning_rate": 6.8364897369691655e-06,
+      "loss": 0.5432,
+      "step": 9075
+    },
+    {
+      "epoch": 1.1815696993361968,
+      "grad_norm": 2.565473794937134,
+      "learning_rate": 6.834545805836992e-06,
+      "loss": 0.4648,
+      "step": 9078
+    },
+    {
+      "epoch": 1.1819601718078876,
+      "grad_norm": 2.2036924362182617,
+      "learning_rate": 6.832601554213525e-06,
+      "loss": 0.4596,
+      "step": 9081
+    },
+    {
+      "epoch": 1.1823506442795784,
+      "grad_norm": 2.521524429321289,
+      "learning_rate": 6.830656982438421e-06,
+      "loss": 0.4932,
+      "step": 9084
+    },
+    {
+      "epoch": 1.1827411167512691,
+      "grad_norm": 2.6289567947387695,
+      "learning_rate": 6.828712090851395e-06,
+      "loss": 0.5188,
+      "step": 9087
+    },
+    {
+      "epoch": 1.1831315892229597,
+      "grad_norm": 2.556784152984619,
+      "learning_rate": 6.826766879792215e-06,
+      "loss": 0.5411,
+      "step": 9090
+    },
+    {
+      "epoch": 1.1835220616946505,
+      "grad_norm": 3.470583915710449,
+      "learning_rate": 6.824821349600702e-06,
+      "loss": 0.5407,
+      "step": 9093
+    },
+    {
+      "epoch": 1.1839125341663412,
+      "grad_norm": 2.5393197536468506,
+      "learning_rate": 6.822875500616739e-06,
+      "loss": 0.4989,
+      "step": 9096
+    },
+    {
+      "epoch": 1.184303006638032,
+      "grad_norm": 2.5908868312835693,
+      "learning_rate": 6.82092933318026e-06,
+      "loss": 0.4926,
+      "step": 9099
+    },
+    {
+      "epoch": 1.1846934791097228,
+      "grad_norm": 3.1211929321289062,
+      "learning_rate": 6.818982847631258e-06,
+      "loss": 0.517,
+      "step": 9102
+    },
+    {
+      "epoch": 1.1850839515814136,
+      "grad_norm": 2.8722972869873047,
+      "learning_rate": 6.8170360443097794e-06,
+      "loss": 0.5064,
+      "step": 9105
+    },
+    {
+      "epoch": 1.1854744240531043,
+      "grad_norm": 2.619140386581421,
+      "learning_rate": 6.815088923555925e-06,
+      "loss": 0.4872,
+      "step": 9108
+    },
+    {
+      "epoch": 1.185864896524795,
+      "grad_norm": 2.4975740909576416,
+      "learning_rate": 6.813141485709856e-06,
+      "loss": 0.4782,
+      "step": 9111
+    },
+    {
+      "epoch": 1.1862553689964859,
+      "grad_norm": 2.6161115169525146,
+      "learning_rate": 6.811193731111782e-06,
+      "loss": 0.4789,
+      "step": 9114
+    },
+    {
+      "epoch": 1.1866458414681764,
+      "grad_norm": 3.224074602127075,
+      "learning_rate": 6.809245660101974e-06,
+      "loss": 0.5793,
+      "step": 9117
+    },
+    {
+      "epoch": 1.1870363139398672,
+      "grad_norm": 2.5327675342559814,
+      "learning_rate": 6.8072972730207555e-06,
+      "loss": 0.4145,
+      "step": 9120
+    },
+    {
+      "epoch": 1.187426786411558,
+      "grad_norm": 2.6852951049804688,
+      "learning_rate": 6.8053485702085045e-06,
+      "loss": 0.4999,
+      "step": 9123
+    },
+    {
+      "epoch": 1.1878172588832487,
+      "grad_norm": 2.3816871643066406,
+      "learning_rate": 6.8033995520056565e-06,
+      "loss": 0.5238,
+      "step": 9126
+    },
+    {
+      "epoch": 1.1882077313549395,
+      "grad_norm": 2.494142770767212,
+      "learning_rate": 6.801450218752701e-06,
+      "loss": 0.4763,
+      "step": 9129
+    },
+    {
+      "epoch": 1.1885982038266303,
+      "grad_norm": 2.643019676208496,
+      "learning_rate": 6.799500570790182e-06,
+      "loss": 0.5655,
+      "step": 9132
+    },
+    {
+      "epoch": 1.188988676298321,
+      "grad_norm": 2.7931339740753174,
+      "learning_rate": 6.797550608458698e-06,
+      "loss": 0.5176,
+      "step": 9135
+    },
+    {
+      "epoch": 1.1893791487700116,
+      "grad_norm": 2.6419882774353027,
+      "learning_rate": 6.795600332098905e-06,
+      "loss": 0.4891,
+      "step": 9138
+    },
+    {
+      "epoch": 1.1897696212417024,
+      "grad_norm": 2.681978940963745,
+      "learning_rate": 6.793649742051511e-06,
+      "loss": 0.4745,
+      "step": 9141
+    },
+    {
+      "epoch": 1.1901600937133932,
+      "grad_norm": 2.605980396270752,
+      "learning_rate": 6.7916988386572806e-06,
+      "loss": 0.4831,
+      "step": 9144
+    },
+    {
+      "epoch": 1.190550566185084,
+      "grad_norm": 2.5457866191864014,
+      "learning_rate": 6.789747622257033e-06,
+      "loss": 0.4513,
+      "step": 9147
+    },
+    {
+      "epoch": 1.1909410386567747,
+      "grad_norm": 2.7378089427948,
+      "learning_rate": 6.787796093191638e-06,
+      "loss": 0.5176,
+      "step": 9150
+    },
+    {
+      "epoch": 1.1913315111284655,
+      "grad_norm": 2.543762683868408,
+      "learning_rate": 6.785844251802031e-06,
+      "loss": 0.5133,
+      "step": 9153
+    },
+    {
+      "epoch": 1.1917219836001562,
+      "grad_norm": 2.86997652053833,
+      "learning_rate": 6.783892098429187e-06,
+      "loss": 0.5483,
+      "step": 9156
+    },
+    {
+      "epoch": 1.192112456071847,
+      "grad_norm": 3.302907943725586,
+      "learning_rate": 6.781939633414146e-06,
+      "loss": 0.5072,
+      "step": 9159
+    },
+    {
+      "epoch": 1.1925029285435378,
+      "grad_norm": 2.7741358280181885,
+      "learning_rate": 6.779986857098002e-06,
+      "loss": 0.4742,
+      "step": 9162
+    },
+    {
+      "epoch": 1.1928934010152283,
+      "grad_norm": 2.5811760425567627,
+      "learning_rate": 6.778033769821896e-06,
+      "loss": 0.5102,
+      "step": 9165
+    },
+    {
+      "epoch": 1.1932838734869191,
+      "grad_norm": 4.686474800109863,
+      "learning_rate": 6.776080371927033e-06,
+      "loss": 0.4987,
+      "step": 9168
+    },
+    {
+      "epoch": 1.1936743459586099,
+      "grad_norm": 3.189173460006714,
+      "learning_rate": 6.774126663754663e-06,
+      "loss": 0.464,
+      "step": 9171
+    },
+    {
+      "epoch": 1.1940648184303007,
+      "grad_norm": 2.439980983734131,
+      "learning_rate": 6.7721726456461e-06,
+      "loss": 0.447,
+      "step": 9174
+    },
+    {
+      "epoch": 1.1944552909019914,
+      "grad_norm": 2.823490858078003,
+      "learning_rate": 6.770218317942701e-06,
+      "loss": 0.4427,
+      "step": 9177
+    },
+    {
+      "epoch": 1.1948457633736822,
+      "grad_norm": 2.5944924354553223,
+      "learning_rate": 6.768263680985888e-06,
+      "loss": 0.4201,
+      "step": 9180
+    },
+    {
+      "epoch": 1.195236235845373,
+      "grad_norm": 3.171487808227539,
+      "learning_rate": 6.766308735117129e-06,
+      "loss": 0.615,
+      "step": 9183
+    },
+    {
+      "epoch": 1.1956267083170635,
+      "grad_norm": 2.696826219558716,
+      "learning_rate": 6.764353480677949e-06,
+      "loss": 0.5554,
+      "step": 9186
+    },
+    {
+      "epoch": 1.1960171807887543,
+      "grad_norm": 2.919696569442749,
+      "learning_rate": 6.762397918009929e-06,
+      "loss": 0.5412,
+      "step": 9189
+    },
+    {
+      "epoch": 1.196407653260445,
+      "grad_norm": 2.540598154067993,
+      "learning_rate": 6.760442047454699e-06,
+      "loss": 0.5434,
+      "step": 9192
+    },
+    {
+      "epoch": 1.1967981257321358,
+      "grad_norm": 2.6347830295562744,
+      "learning_rate": 6.758485869353948e-06,
+      "loss": 0.4581,
+      "step": 9195
+    },
+    {
+      "epoch": 1.1971885982038266,
+      "grad_norm": 2.45042085647583,
+      "learning_rate": 6.756529384049415e-06,
+      "loss": 0.5198,
+      "step": 9198
+    },
+    {
+      "epoch": 1.1975790706755174,
+      "grad_norm": 3.13641357421875,
+      "learning_rate": 6.754572591882892e-06,
+      "loss": 0.46,
+      "step": 9201
+    },
+    {
+      "epoch": 1.1979695431472082,
+      "grad_norm": 3.144970417022705,
+      "learning_rate": 6.752615493196231e-06,
+      "loss": 0.4713,
+      "step": 9204
+    },
+    {
+      "epoch": 1.198360015618899,
+      "grad_norm": 2.6827385425567627,
+      "learning_rate": 6.750658088331326e-06,
+      "loss": 0.5368,
+      "step": 9207
+    },
+    {
+      "epoch": 1.1987504880905897,
+      "grad_norm": 2.579334259033203,
+      "learning_rate": 6.7487003776301394e-06,
+      "loss": 0.5117,
+      "step": 9210
+    },
+    {
+      "epoch": 1.1991409605622803,
+      "grad_norm": 2.422724962234497,
+      "learning_rate": 6.746742361434675e-06,
+      "loss": 0.3792,
+      "step": 9213
+    },
+    {
+      "epoch": 1.199531433033971,
+      "grad_norm": 2.511230230331421,
+      "learning_rate": 6.744784040086994e-06,
+      "loss": 0.5195,
+      "step": 9216
+    },
+    {
+      "epoch": 1.1999219055056618,
+      "grad_norm": 2.661550283432007,
+      "learning_rate": 6.742825413929213e-06,
+      "loss": 0.4506,
+      "step": 9219
+    },
+    {
+      "epoch": 1.2003123779773526,
+      "grad_norm": 2.8288450241088867,
+      "learning_rate": 6.740866483303497e-06,
+      "loss": 0.5301,
+      "step": 9222
+    },
+    {
+      "epoch": 1.2007028504490433,
+      "grad_norm": 2.494079351425171,
+      "learning_rate": 6.73890724855207e-06,
+      "loss": 0.4576,
+      "step": 9225
+    },
+    {
+      "epoch": 1.2010933229207341,
+      "grad_norm": 2.7465291023254395,
+      "learning_rate": 6.736947710017202e-06,
+      "loss": 0.4389,
+      "step": 9228
+    },
+    {
+      "epoch": 1.201483795392425,
+      "grad_norm": 3.119746208190918,
+      "learning_rate": 6.734987868041226e-06,
+      "loss": 0.519,
+      "step": 9231
+    },
+    {
+      "epoch": 1.2018742678641157,
+      "grad_norm": 2.939565896987915,
+      "learning_rate": 6.733027722966519e-06,
+      "loss": 0.5981,
+      "step": 9234
+    },
+    {
+      "epoch": 1.2022647403358064,
+      "grad_norm": 2.5819921493530273,
+      "learning_rate": 6.731067275135512e-06,
+      "loss": 0.5423,
+      "step": 9237
+    },
+    {
+      "epoch": 1.202655212807497,
+      "grad_norm": 2.769885301589966,
+      "learning_rate": 6.7291065248906975e-06,
+      "loss": 0.481,
+      "step": 9240
+    },
+    {
+      "epoch": 1.2030456852791878,
+      "grad_norm": 2.8047449588775635,
+      "learning_rate": 6.727145472574608e-06,
+      "loss": 0.4647,
+      "step": 9243
+    },
+    {
+      "epoch": 1.2034361577508785,
+      "grad_norm": 2.8334429264068604,
+      "learning_rate": 6.725184118529839e-06,
+      "loss": 0.525,
+      "step": 9246
+    },
+    {
+      "epoch": 1.2038266302225693,
+      "grad_norm": 2.9303195476531982,
+      "learning_rate": 6.723222463099033e-06,
+      "loss": 0.4845,
+      "step": 9249
+    },
+    {
+      "epoch": 1.20421710269426,
+      "grad_norm": 2.522890567779541,
+      "learning_rate": 6.721260506624888e-06,
+      "loss": 0.423,
+      "step": 9252
+    },
+    {
+      "epoch": 1.2046075751659509,
+      "grad_norm": 3.1758248805999756,
+      "learning_rate": 6.719298249450153e-06,
+      "loss": 0.4912,
+      "step": 9255
+    },
+    {
+      "epoch": 1.2049980476376416,
+      "grad_norm": 2.49548602104187,
+      "learning_rate": 6.7173356919176315e-06,
+      "loss": 0.4624,
+      "step": 9258
+    },
+    {
+      "epoch": 1.2053885201093322,
+      "grad_norm": 2.884762763977051,
+      "learning_rate": 6.7153728343701776e-06,
+      "loss": 0.4791,
+      "step": 9261
+    },
+    {
+      "epoch": 1.205778992581023,
+      "grad_norm": 2.4931530952453613,
+      "learning_rate": 6.7134096771506976e-06,
+      "loss": 0.542,
+      "step": 9264
+    },
+    {
+      "epoch": 1.2061694650527137,
+      "grad_norm": 2.408496618270874,
+      "learning_rate": 6.711446220602152e-06,
+      "loss": 0.4824,
+      "step": 9267
+    },
+    {
+      "epoch": 1.2065599375244045,
+      "grad_norm": 2.6510279178619385,
+      "learning_rate": 6.70948246506755e-06,
+      "loss": 0.5171,
+      "step": 9270
+    },
+    {
+      "epoch": 1.2069504099960953,
+      "grad_norm": 2.6602025032043457,
+      "learning_rate": 6.707518410889959e-06,
+      "loss": 0.5202,
+      "step": 9273
+    },
+    {
+      "epoch": 1.207340882467786,
+      "grad_norm": 3.2938411235809326,
+      "learning_rate": 6.7055540584124955e-06,
+      "loss": 0.4416,
+      "step": 9276
+    },
+    {
+      "epoch": 1.2077313549394768,
+      "grad_norm": 2.6444809436798096,
+      "learning_rate": 6.703589407978324e-06,
+      "loss": 0.5285,
+      "step": 9279
+    },
+    {
+      "epoch": 1.2081218274111676,
+      "grad_norm": 2.7918782234191895,
+      "learning_rate": 6.7016244599306675e-06,
+      "loss": 0.5213,
+      "step": 9282
+    },
+    {
+      "epoch": 1.2085122998828584,
+      "grad_norm": 2.3297805786132812,
+      "learning_rate": 6.699659214612797e-06,
+      "loss": 0.4901,
+      "step": 9285
+    },
+    {
+      "epoch": 1.208902772354549,
+      "grad_norm": 2.6607468128204346,
+      "learning_rate": 6.697693672368038e-06,
+      "loss": 0.518,
+      "step": 9288
+    },
+    {
+      "epoch": 1.2092932448262397,
+      "grad_norm": 2.5616703033447266,
+      "learning_rate": 6.695727833539765e-06,
+      "loss": 0.47,
+      "step": 9291
+    },
+    {
+      "epoch": 1.2096837172979304,
+      "grad_norm": 3.2793567180633545,
+      "learning_rate": 6.693761698471406e-06,
+      "loss": 0.552,
+      "step": 9294
+    },
+    {
+      "epoch": 1.2100741897696212,
+      "grad_norm": 2.4998676776885986,
+      "learning_rate": 6.6917952675064435e-06,
+      "loss": 0.5091,
+      "step": 9297
+    },
+    {
+      "epoch": 1.210464662241312,
+      "grad_norm": 2.6549830436706543,
+      "learning_rate": 6.689828540988406e-06,
+      "loss": 0.5316,
+      "step": 9300
+    },
+    {
+      "epoch": 1.2108551347130028,
+      "grad_norm": 2.714661121368408,
+      "learning_rate": 6.687861519260877e-06,
+      "loss": 0.5968,
+      "step": 9303
+    },
+    {
+      "epoch": 1.2112456071846935,
+      "grad_norm": 4.035742282867432,
+      "learning_rate": 6.685894202667491e-06,
+      "loss": 0.4923,
+      "step": 9306
+    },
+    {
+      "epoch": 1.2116360796563843,
+      "grad_norm": 2.6630351543426514,
+      "learning_rate": 6.683926591551934e-06,
+      "loss": 0.5069,
+      "step": 9309
+    },
+    {
+      "epoch": 1.212026552128075,
+      "grad_norm": 2.704087257385254,
+      "learning_rate": 6.681958686257945e-06,
+      "loss": 0.5276,
+      "step": 9312
+    },
+    {
+      "epoch": 1.2124170245997656,
+      "grad_norm": 4.812976837158203,
+      "learning_rate": 6.679990487129311e-06,
+      "loss": 0.5277,
+      "step": 9315
+    },
+    {
+      "epoch": 1.2128074970714564,
+      "grad_norm": 2.845241069793701,
+      "learning_rate": 6.678021994509874e-06,
+      "loss": 0.5935,
+      "step": 9318
+    },
+    {
+      "epoch": 1.2131979695431472,
+      "grad_norm": 2.6971890926361084,
+      "learning_rate": 6.676053208743525e-06,
+      "loss": 0.5589,
+      "step": 9321
+    },
+    {
+      "epoch": 1.213588442014838,
+      "grad_norm": 2.571746349334717,
+      "learning_rate": 6.674084130174204e-06,
+      "loss": 0.4472,
+      "step": 9324
+    },
+    {
+      "epoch": 1.2139789144865287,
+      "grad_norm": 2.5498881340026855,
+      "learning_rate": 6.67211475914591e-06,
+      "loss": 0.496,
+      "step": 9327
+    },
+    {
+      "epoch": 1.2143693869582195,
+      "grad_norm": 2.6578900814056396,
+      "learning_rate": 6.670145096002683e-06,
+      "loss": 0.5228,
+      "step": 9330
+    },
+    {
+      "epoch": 1.2147598594299103,
+      "grad_norm": 2.7154641151428223,
+      "learning_rate": 6.668175141088622e-06,
+      "loss": 0.5132,
+      "step": 9333
+    },
+    {
+      "epoch": 1.2151503319016008,
+      "grad_norm": 2.644035577774048,
+      "learning_rate": 6.666204894747874e-06,
+      "loss": 0.5248,
+      "step": 9336
+    },
+    {
+      "epoch": 1.2155408043732916,
+      "grad_norm": 2.640990734100342,
+      "learning_rate": 6.664234357324636e-06,
+      "loss": 0.4759,
+      "step": 9339
+    },
+    {
+      "epoch": 1.2159312768449824,
+      "grad_norm": 2.595244884490967,
+      "learning_rate": 6.662263529163155e-06,
+      "loss": 0.5262,
+      "step": 9342
+    },
+    {
+      "epoch": 1.2163217493166731,
+      "grad_norm": 2.70097279548645,
+      "learning_rate": 6.660292410607734e-06,
+      "loss": 0.5814,
+      "step": 9345
+    },
+    {
+      "epoch": 1.216712221788364,
+      "grad_norm": 3.5157313346862793,
+      "learning_rate": 6.658321002002722e-06,
+      "loss": 0.508,
+      "step": 9348
+    },
+    {
+      "epoch": 1.2171026942600547,
+      "grad_norm": 2.607250213623047,
+      "learning_rate": 6.656349303692519e-06,
+      "loss": 0.501,
+      "step": 9351
+    },
+    {
+      "epoch": 1.2174931667317455,
+      "grad_norm": 3.2209982872009277,
+      "learning_rate": 6.654377316021576e-06,
+      "loss": 0.4115,
+      "step": 9354
+    },
+    {
+      "epoch": 1.2178836392034362,
+      "grad_norm": 2.4936866760253906,
+      "learning_rate": 6.652405039334396e-06,
+      "loss": 0.4835,
+      "step": 9357
+    },
+    {
+      "epoch": 1.218274111675127,
+      "grad_norm": 2.4629316329956055,
+      "learning_rate": 6.650432473975534e-06,
+      "loss": 0.4565,
+      "step": 9360
+    },
+    {
+      "epoch": 1.2186645841468176,
+      "grad_norm": 3.328362464904785,
+      "learning_rate": 6.648459620289589e-06,
+      "loss": 0.5637,
+      "step": 9363
+    },
+    {
+      "epoch": 1.2190550566185083,
+      "grad_norm": 2.8752493858337402,
+      "learning_rate": 6.646486478621217e-06,
+      "loss": 0.4952,
+      "step": 9366
+    },
+    {
+      "epoch": 1.219445529090199,
+      "grad_norm": 2.651597738265991,
+      "learning_rate": 6.644513049315121e-06,
+      "loss": 0.4809,
+      "step": 9369
+    },
+    {
+      "epoch": 1.2198360015618899,
+      "grad_norm": 2.642346143722534,
+      "learning_rate": 6.642539332716055e-06,
+      "loss": 0.4713,
+      "step": 9372
+    },
+    {
+      "epoch": 1.2202264740335806,
+      "grad_norm": 2.764996290206909,
+      "learning_rate": 6.6405653291688225e-06,
+      "loss": 0.5554,
+      "step": 9375
+    },
+    {
+      "epoch": 1.2206169465052714,
+      "grad_norm": 2.660748243331909,
+      "learning_rate": 6.638591039018277e-06,
+      "loss": 0.475,
+      "step": 9378
+    },
+    {
+      "epoch": 1.2210074189769622,
+      "grad_norm": 2.4974265098571777,
+      "learning_rate": 6.636616462609324e-06,
+      "loss": 0.5118,
+      "step": 9381
+    },
+    {
+      "epoch": 1.221397891448653,
+      "grad_norm": 3.5743143558502197,
+      "learning_rate": 6.634641600286921e-06,
+      "loss": 0.5091,
+      "step": 9384
+    },
+    {
+      "epoch": 1.2217883639203437,
+      "grad_norm": 2.744140625,
+      "learning_rate": 6.632666452396067e-06,
+      "loss": 0.4696,
+      "step": 9387
+    },
+    {
+      "epoch": 1.2221788363920343,
+      "grad_norm": 2.7472984790802,
+      "learning_rate": 6.630691019281819e-06,
+      "loss": 0.4901,
+      "step": 9390
+    },
+    {
+      "epoch": 1.222569308863725,
+      "grad_norm": 3.2367637157440186,
+      "learning_rate": 6.6287153012892805e-06,
+      "loss": 0.4851,
+      "step": 9393
+    },
+    {
+      "epoch": 1.2229597813354158,
+      "grad_norm": 2.966864824295044,
+      "learning_rate": 6.626739298763605e-06,
+      "loss": 0.4722,
+      "step": 9396
+    },
+    {
+      "epoch": 1.2233502538071066,
+      "grad_norm": 2.585268020629883,
+      "learning_rate": 6.624763012049995e-06,
+      "loss": 0.4988,
+      "step": 9399
+    },
+    {
+      "epoch": 1.2237407262787974,
+      "grad_norm": 2.772038221359253,
+      "learning_rate": 6.622786441493706e-06,
+      "loss": 0.5594,
+      "step": 9402
+    },
+    {
+      "epoch": 1.2241311987504881,
+      "grad_norm": 3.0204148292541504,
+      "learning_rate": 6.62080958744004e-06,
+      "loss": 0.4736,
+      "step": 9405
+    },
+    {
+      "epoch": 1.224521671222179,
+      "grad_norm": 2.651444673538208,
+      "learning_rate": 6.618832450234348e-06,
+      "loss": 0.473,
+      "step": 9408
+    },
+    {
+      "epoch": 1.2249121436938695,
+      "grad_norm": 3.069899559020996,
+      "learning_rate": 6.6168550302220334e-06,
+      "loss": 0.618,
+      "step": 9411
+    },
+    {
+      "epoch": 1.2253026161655602,
+      "grad_norm": 2.877786874771118,
+      "learning_rate": 6.6148773277485455e-06,
+      "loss": 0.5673,
+      "step": 9414
+    },
+    {
+      "epoch": 1.225693088637251,
+      "grad_norm": 2.4998676776885986,
+      "learning_rate": 6.612899343159385e-06,
+      "loss": 0.4381,
+      "step": 9417
+    },
+    {
+      "epoch": 1.2260835611089418,
+      "grad_norm": 2.4126298427581787,
+      "learning_rate": 6.610921076800103e-06,
+      "loss": 0.5247,
+      "step": 9420
+    },
+    {
+      "epoch": 1.2264740335806326,
+      "grad_norm": 2.9004220962524414,
+      "learning_rate": 6.608942529016298e-06,
+      "loss": 0.5114,
+      "step": 9423
+    },
+    {
+      "epoch": 1.2268645060523233,
+      "grad_norm": 3.1982264518737793,
+      "learning_rate": 6.606963700153618e-06,
+      "loss": 0.5308,
+      "step": 9426
+    },
+    {
+      "epoch": 1.227254978524014,
+      "grad_norm": 2.521232843399048,
+      "learning_rate": 6.604984590557759e-06,
+      "loss": 0.4787,
+      "step": 9429
+    },
+    {
+      "epoch": 1.2276454509957049,
+      "grad_norm": 3.711055040359497,
+      "learning_rate": 6.603005200574471e-06,
+      "loss": 0.4829,
+      "step": 9432
+    },
+    {
+      "epoch": 1.2280359234673957,
+      "grad_norm": 2.418956756591797,
+      "learning_rate": 6.601025530549544e-06,
+      "loss": 0.4596,
+      "step": 9435
+    },
+    {
+      "epoch": 1.2284263959390862,
+      "grad_norm": 2.617719888687134,
+      "learning_rate": 6.5990455808288256e-06,
+      "loss": 0.5762,
+      "step": 9438
+    },
+    {
+      "epoch": 1.228816868410777,
+      "grad_norm": 3.037282705307007,
+      "learning_rate": 6.597065351758207e-06,
+      "loss": 0.4857,
+      "step": 9441
+    },
+    {
+      "epoch": 1.2292073408824677,
+      "grad_norm": 2.377511739730835,
+      "learning_rate": 6.5950848436836335e-06,
+      "loss": 0.4933,
+      "step": 9444
+    },
+    {
+      "epoch": 1.2295978133541585,
+      "grad_norm": 2.6414594650268555,
+      "learning_rate": 6.5931040569510926e-06,
+      "loss": 0.5575,
+      "step": 9447
+    },
+    {
+      "epoch": 1.2299882858258493,
+      "grad_norm": 3.2469959259033203,
+      "learning_rate": 6.591122991906625e-06,
+      "loss": 0.537,
+      "step": 9450
+    },
+    {
+      "epoch": 1.23037875829754,
+      "grad_norm": 2.7081236839294434,
+      "learning_rate": 6.5891416488963155e-06,
+      "loss": 0.5333,
+      "step": 9453
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 3.2291836738586426,
+      "learning_rate": 6.587160028266306e-06,
+      "loss": 0.5007,
+      "step": 9456
+    },
+    {
+      "epoch": 1.2311597032409216,
+      "grad_norm": 3.501340389251709,
+      "learning_rate": 6.585178130362776e-06,
+      "loss": 0.4621,
+      "step": 9459
+    },
+    {
+      "epoch": 1.2315501757126124,
+      "grad_norm": 2.6834263801574707,
+      "learning_rate": 6.583195955531963e-06,
+      "loss": 0.472,
+      "step": 9462
+    },
+    {
+      "epoch": 1.231940648184303,
+      "grad_norm": 3.268996238708496,
+      "learning_rate": 6.581213504120146e-06,
+      "loss": 0.5295,
+      "step": 9465
+    },
+    {
+      "epoch": 1.2323311206559937,
+      "grad_norm": 2.7084834575653076,
+      "learning_rate": 6.579230776473658e-06,
+      "loss": 0.543,
+      "step": 9468
+    },
+    {
+      "epoch": 1.2327215931276845,
+      "grad_norm": 2.728600263595581,
+      "learning_rate": 6.577247772938874e-06,
+      "loss": 0.5473,
+      "step": 9471
+    },
+    {
+      "epoch": 1.2331120655993753,
+      "grad_norm": 3.1851062774658203,
+      "learning_rate": 6.575264493862221e-06,
+      "loss": 0.5181,
+      "step": 9474
+    },
+    {
+      "epoch": 1.233502538071066,
+      "grad_norm": 2.5818753242492676,
+      "learning_rate": 6.573280939590178e-06,
+      "loss": 0.433,
+      "step": 9477
+    },
+    {
+      "epoch": 1.2338930105427568,
+      "grad_norm": 2.6541337966918945,
+      "learning_rate": 6.571297110469261e-06,
+      "loss": 0.503,
+      "step": 9480
+    },
+    {
+      "epoch": 1.2342834830144476,
+      "grad_norm": 2.538254499435425,
+      "learning_rate": 6.569313006846048e-06,
+      "loss": 0.5565,
+      "step": 9483
+    },
+    {
+      "epoch": 1.2346739554861381,
+      "grad_norm": 2.7005538940429688,
+      "learning_rate": 6.567328629067151e-06,
+      "loss": 0.5625,
+      "step": 9486
+    },
+    {
+      "epoch": 1.235064427957829,
+      "grad_norm": 2.8296282291412354,
+      "learning_rate": 6.565343977479241e-06,
+      "loss": 0.5815,
+      "step": 9489
+    },
+    {
+      "epoch": 1.2354549004295197,
+      "grad_norm": 2.405651092529297,
+      "learning_rate": 6.563359052429031e-06,
+      "loss": 0.4166,
+      "step": 9492
+    },
+    {
+      "epoch": 1.2358453729012104,
+      "grad_norm": 2.3672516345977783,
+      "learning_rate": 6.561373854263283e-06,
+      "loss": 0.5841,
+      "step": 9495
+    },
+    {
+      "epoch": 1.2362358453729012,
+      "grad_norm": 2.7860894203186035,
+      "learning_rate": 6.559388383328808e-06,
+      "loss": 0.4879,
+      "step": 9498
+    },
+    {
+      "epoch": 1.236626317844592,
+      "grad_norm": 2.6302614212036133,
+      "learning_rate": 6.5574026399724625e-06,
+      "loss": 0.5354,
+      "step": 9501
+    },
+    {
+      "epoch": 1.2370167903162828,
+      "grad_norm": 2.8287155628204346,
+      "learning_rate": 6.5554166245411525e-06,
+      "loss": 0.4181,
+      "step": 9504
+    },
+    {
+      "epoch": 1.2374072627879735,
+      "grad_norm": 2.8489389419555664,
+      "learning_rate": 6.55343033738183e-06,
+      "loss": 0.4872,
+      "step": 9507
+    },
+    {
+      "epoch": 1.2377977352596643,
+      "grad_norm": 3.101821184158325,
+      "learning_rate": 6.551443778841495e-06,
+      "loss": 0.5089,
+      "step": 9510
+    },
+    {
+      "epoch": 1.2381882077313549,
+      "grad_norm": 3.262808084487915,
+      "learning_rate": 6.549456949267197e-06,
+      "loss": 0.5069,
+      "step": 9513
+    },
+    {
+      "epoch": 1.2385786802030456,
+      "grad_norm": 2.664828300476074,
+      "learning_rate": 6.547469849006027e-06,
+      "loss": 0.4743,
+      "step": 9516
+    },
+    {
+      "epoch": 1.2389691526747364,
+      "grad_norm": 2.8644232749938965,
+      "learning_rate": 6.545482478405133e-06,
+      "loss": 0.4491,
+      "step": 9519
+    },
+    {
+      "epoch": 1.2393596251464272,
+      "grad_norm": 2.658860921859741,
+      "learning_rate": 6.543494837811698e-06,
+      "loss": 0.5217,
+      "step": 9522
+    },
+    {
+      "epoch": 1.239750097618118,
+      "grad_norm": 2.8123228549957275,
+      "learning_rate": 6.541506927572965e-06,
+      "loss": 0.52,
+      "step": 9525
+    },
+    {
+      "epoch": 1.2401405700898087,
+      "grad_norm": 2.8045687675476074,
+      "learning_rate": 6.539518748036212e-06,
+      "loss": 0.463,
+      "step": 9528
+    },
+    {
+      "epoch": 1.2405310425614995,
+      "grad_norm": 2.54270076751709,
+      "learning_rate": 6.537530299548774e-06,
+      "loss": 0.5799,
+      "step": 9531
+    },
+    {
+      "epoch": 1.24092151503319,
+      "grad_norm": 2.9174578189849854,
+      "learning_rate": 6.535541582458027e-06,
+      "loss": 0.4923,
+      "step": 9534
+    },
+    {
+      "epoch": 1.2413119875048808,
+      "grad_norm": 3.1313931941986084,
+      "learning_rate": 6.533552597111395e-06,
+      "loss": 0.5614,
+      "step": 9537
+    },
+    {
+      "epoch": 1.2417024599765716,
+      "grad_norm": 2.6302151679992676,
+      "learning_rate": 6.531563343856352e-06,
+      "loss": 0.5368,
+      "step": 9540
+    },
+    {
+      "epoch": 1.2420929324482624,
+      "grad_norm": 2.6476080417633057,
+      "learning_rate": 6.5295738230404125e-06,
+      "loss": 0.5409,
+      "step": 9543
+    },
+    {
+      "epoch": 1.2424834049199531,
+      "grad_norm": 2.8190321922302246,
+      "learning_rate": 6.527584035011145e-06,
+      "loss": 0.4559,
+      "step": 9546
+    },
+    {
+      "epoch": 1.242873877391644,
+      "grad_norm": 3.6176981925964355,
+      "learning_rate": 6.525593980116161e-06,
+      "loss": 0.5597,
+      "step": 9549
+    },
+    {
+      "epoch": 1.2432643498633347,
+      "grad_norm": 2.844064474105835,
+      "learning_rate": 6.523603658703117e-06,
+      "loss": 0.575,
+      "step": 9552
+    },
+    {
+      "epoch": 1.2436548223350254,
+      "grad_norm": 3.310737371444702,
+      "learning_rate": 6.52161307111972e-06,
+      "loss": 0.5425,
+      "step": 9555
+    },
+    {
+      "epoch": 1.2440452948067162,
+      "grad_norm": 2.583648204803467,
+      "learning_rate": 6.519622217713719e-06,
+      "loss": 0.4643,
+      "step": 9558
+    },
+    {
+      "epoch": 1.2444357672784068,
+      "grad_norm": 2.5113677978515625,
+      "learning_rate": 6.517631098832914e-06,
+      "loss": 0.4568,
+      "step": 9561
+    },
+    {
+      "epoch": 1.2448262397500975,
+      "grad_norm": 4.185740947723389,
+      "learning_rate": 6.515639714825148e-06,
+      "loss": 0.4515,
+      "step": 9564
+    },
+    {
+      "epoch": 1.2452167122217883,
+      "grad_norm": 2.941195011138916,
+      "learning_rate": 6.513648066038314e-06,
+      "loss": 0.5402,
+      "step": 9567
+    },
+    {
+      "epoch": 1.245607184693479,
+      "grad_norm": 2.2286741733551025,
+      "learning_rate": 6.511656152820347e-06,
+      "loss": 0.4221,
+      "step": 9570
+    },
+    {
+      "epoch": 1.2459976571651699,
+      "grad_norm": 2.8112363815307617,
+      "learning_rate": 6.509663975519228e-06,
+      "loss": 0.5133,
+      "step": 9573
+    },
+    {
+      "epoch": 1.2463881296368606,
+      "grad_norm": 2.6967437267303467,
+      "learning_rate": 6.507671534482991e-06,
+      "loss": 0.5319,
+      "step": 9576
+    },
+    {
+      "epoch": 1.2467786021085514,
+      "grad_norm": 2.501628875732422,
+      "learning_rate": 6.505678830059707e-06,
+      "loss": 0.4809,
+      "step": 9579
+    },
+    {
+      "epoch": 1.2471690745802422,
+      "grad_norm": 2.820998191833496,
+      "learning_rate": 6.5036858625974986e-06,
+      "loss": 0.5363,
+      "step": 9582
+    },
+    {
+      "epoch": 1.247559547051933,
+      "grad_norm": 2.792283296585083,
+      "learning_rate": 6.501692632444534e-06,
+      "loss": 0.4598,
+      "step": 9585
+    },
+    {
+      "epoch": 1.2479500195236235,
+      "grad_norm": 2.6301913261413574,
+      "learning_rate": 6.499699139949025e-06,
+      "loss": 0.4655,
+      "step": 9588
+    },
+    {
+      "epoch": 1.2483404919953143,
+      "grad_norm": 2.497735023498535,
+      "learning_rate": 6.497705385459232e-06,
+      "loss": 0.5135,
+      "step": 9591
+    },
+    {
+      "epoch": 1.248730964467005,
+      "grad_norm": 2.354994535446167,
+      "learning_rate": 6.4957113693234586e-06,
+      "loss": 0.4236,
+      "step": 9594
+    },
+    {
+      "epoch": 1.2491214369386958,
+      "grad_norm": 3.677664041519165,
+      "learning_rate": 6.493717091890056e-06,
+      "loss": 0.4943,
+      "step": 9597
+    },
+    {
+      "epoch": 1.2495119094103866,
+      "grad_norm": 2.9446117877960205,
+      "learning_rate": 6.491722553507419e-06,
+      "loss": 0.5157,
+      "step": 9600
+    },
+    {
+      "epoch": 1.2499023818820774,
+      "grad_norm": 3.580937623977661,
+      "learning_rate": 6.48972775452399e-06,
+      "loss": 0.5131,
+      "step": 9603
+    },
+    {
+      "epoch": 1.2502928543537681,
+      "grad_norm": 2.604628562927246,
+      "learning_rate": 6.487732695288256e-06,
+      "loss": 0.4847,
+      "step": 9606
+    },
+    {
+      "epoch": 1.2506833268254587,
+      "grad_norm": 2.9998080730438232,
+      "learning_rate": 6.4857373761487505e-06,
+      "loss": 0.5538,
+      "step": 9609
+    },
+    {
+      "epoch": 1.2510737992971497,
+      "grad_norm": 2.7411253452301025,
+      "learning_rate": 6.4837417974540505e-06,
+      "loss": 0.4721,
+      "step": 9612
+    },
+    {
+      "epoch": 1.2514642717688402,
+      "grad_norm": 2.709904432296753,
+      "learning_rate": 6.481745959552781e-06,
+      "loss": 0.446,
+      "step": 9615
+    },
+    {
+      "epoch": 1.251854744240531,
+      "grad_norm": 2.6614980697631836,
+      "learning_rate": 6.479749862793609e-06,
+      "loss": 0.5096,
+      "step": 9618
+    },
+    {
+      "epoch": 1.2522452167122218,
+      "grad_norm": 2.518862724304199,
+      "learning_rate": 6.477753507525249e-06,
+      "loss": 0.5016,
+      "step": 9621
+    },
+    {
+      "epoch": 1.2526356891839125,
+      "grad_norm": 2.293060302734375,
+      "learning_rate": 6.475756894096458e-06,
+      "loss": 0.5011,
+      "step": 9624
+    },
+    {
+      "epoch": 1.2530261616556033,
+      "grad_norm": 2.695930004119873,
+      "learning_rate": 6.4737600228560435e-06,
+      "loss": 0.5012,
+      "step": 9627
+    },
+    {
+      "epoch": 1.253416634127294,
+      "grad_norm": 2.469388246536255,
+      "learning_rate": 6.471762894152853e-06,
+      "loss": 0.489,
+      "step": 9630
+    },
+    {
+      "epoch": 1.2538071065989849,
+      "grad_norm": 3.3694796562194824,
+      "learning_rate": 6.469765508335783e-06,
+      "loss": 0.435,
+      "step": 9633
+    },
+    {
+      "epoch": 1.2541975790706754,
+      "grad_norm": 2.474310874938965,
+      "learning_rate": 6.467767865753768e-06,
+      "loss": 0.4652,
+      "step": 9636
+    },
+    {
+      "epoch": 1.2545880515423662,
+      "grad_norm": 3.2254626750946045,
+      "learning_rate": 6.465769966755795e-06,
+      "loss": 0.5496,
+      "step": 9639
+    },
+    {
+      "epoch": 1.254978524014057,
+      "grad_norm": 2.502992630004883,
+      "learning_rate": 6.4637718116908945e-06,
+      "loss": 0.4993,
+      "step": 9642
+    },
+    {
+      "epoch": 1.2553689964857477,
+      "grad_norm": 2.6724181175231934,
+      "learning_rate": 6.461773400908136e-06,
+      "loss": 0.5079,
+      "step": 9645
+    },
+    {
+      "epoch": 1.2557594689574385,
+      "grad_norm": 2.582564353942871,
+      "learning_rate": 6.459774734756639e-06,
+      "loss": 0.546,
+      "step": 9648
+    },
+    {
+      "epoch": 1.2561499414291293,
+      "grad_norm": 2.8137545585632324,
+      "learning_rate": 6.457775813585567e-06,
+      "loss": 0.5178,
+      "step": 9651
+    },
+    {
+      "epoch": 1.25654041390082,
+      "grad_norm": 3.133225679397583,
+      "learning_rate": 6.4557766377441285e-06,
+      "loss": 0.5144,
+      "step": 9654
+    },
+    {
+      "epoch": 1.2569308863725106,
+      "grad_norm": 2.940943717956543,
+      "learning_rate": 6.453777207581573e-06,
+      "loss": 0.4967,
+      "step": 9657
+    },
+    {
+      "epoch": 1.2573213588442016,
+      "grad_norm": 2.1819236278533936,
+      "learning_rate": 6.451777523447197e-06,
+      "loss": 0.439,
+      "step": 9660
+    },
+    {
+      "epoch": 1.2577118313158921,
+      "grad_norm": 2.8423125743865967,
+      "learning_rate": 6.449777585690344e-06,
+      "loss": 0.5978,
+      "step": 9663
+    },
+    {
+      "epoch": 1.258102303787583,
+      "grad_norm": 2.4884462356567383,
+      "learning_rate": 6.447777394660394e-06,
+      "loss": 0.5501,
+      "step": 9666
+    },
+    {
+      "epoch": 1.2584927762592737,
+      "grad_norm": 2.5996439456939697,
+      "learning_rate": 6.445776950706779e-06,
+      "loss": 0.5206,
+      "step": 9669
+    },
+    {
+      "epoch": 1.2588832487309645,
+      "grad_norm": 2.7258780002593994,
+      "learning_rate": 6.4437762541789735e-06,
+      "loss": 0.4822,
+      "step": 9672
+    },
+    {
+      "epoch": 1.2592737212026552,
+      "grad_norm": 2.3117198944091797,
+      "learning_rate": 6.441775305426494e-06,
+      "loss": 0.4431,
+      "step": 9675
+    },
+    {
+      "epoch": 1.259664193674346,
+      "grad_norm": 2.780622959136963,
+      "learning_rate": 6.4397741047989e-06,
+      "loss": 0.382,
+      "step": 9678
+    },
+    {
+      "epoch": 1.2600546661460368,
+      "grad_norm": 2.5801424980163574,
+      "learning_rate": 6.4377726526458e-06,
+      "loss": 0.4101,
+      "step": 9681
+    },
+    {
+      "epoch": 1.2604451386177273,
+      "grad_norm": 2.5400123596191406,
+      "learning_rate": 6.435770949316843e-06,
+      "loss": 0.4303,
+      "step": 9684
+    },
+    {
+      "epoch": 1.2608356110894183,
+      "grad_norm": 2.6822738647460938,
+      "learning_rate": 6.43376899516172e-06,
+      "loss": 0.4843,
+      "step": 9687
+    },
+    {
+      "epoch": 1.2612260835611089,
+      "grad_norm": 2.613121747970581,
+      "learning_rate": 6.43176679053017e-06,
+      "loss": 0.4722,
+      "step": 9690
+    },
+    {
+      "epoch": 1.2616165560327997,
+      "grad_norm": 3.226043224334717,
+      "learning_rate": 6.429764335771973e-06,
+      "loss": 0.4816,
+      "step": 9693
+    },
+    {
+      "epoch": 1.2620070285044904,
+      "grad_norm": 2.7644383907318115,
+      "learning_rate": 6.427761631236955e-06,
+      "loss": 0.4896,
+      "step": 9696
+    },
+    {
+      "epoch": 1.2623975009761812,
+      "grad_norm": 2.632202625274658,
+      "learning_rate": 6.4257586772749845e-06,
+      "loss": 0.5031,
+      "step": 9699
+    },
+    {
+      "epoch": 1.262787973447872,
+      "grad_norm": 3.154134750366211,
+      "learning_rate": 6.423755474235972e-06,
+      "loss": 0.5501,
+      "step": 9702
+    },
+    {
+      "epoch": 1.2631784459195627,
+      "grad_norm": 2.597161054611206,
+      "learning_rate": 6.421752022469874e-06,
+      "loss": 0.592,
+      "step": 9705
+    },
+    {
+      "epoch": 1.2635689183912535,
+      "grad_norm": 2.492748975753784,
+      "learning_rate": 6.4197483223266865e-06,
+      "loss": 0.5851,
+      "step": 9708
+    },
+    {
+      "epoch": 1.263959390862944,
+      "grad_norm": 2.564765453338623,
+      "learning_rate": 6.417744374156455e-06,
+      "loss": 0.5529,
+      "step": 9711
+    },
+    {
+      "epoch": 1.2643498633346348,
+      "grad_norm": 2.9283761978149414,
+      "learning_rate": 6.4157401783092645e-06,
+      "loss": 0.5119,
+      "step": 9714
+    },
+    {
+      "epoch": 1.2647403358063256,
+      "grad_norm": 2.577751874923706,
+      "learning_rate": 6.413735735135241e-06,
+      "loss": 0.4277,
+      "step": 9717
+    },
+    {
+      "epoch": 1.2651308082780164,
+      "grad_norm": 2.47666072845459,
+      "learning_rate": 6.411731044984562e-06,
+      "loss": 0.5617,
+      "step": 9720
+    },
+    {
+      "epoch": 1.2655212807497072,
+      "grad_norm": 4.063993453979492,
+      "learning_rate": 6.409726108207436e-06,
+      "loss": 0.4548,
+      "step": 9723
+    },
+    {
+      "epoch": 1.265911753221398,
+      "grad_norm": 2.558497428894043,
+      "learning_rate": 6.407720925154126e-06,
+      "loss": 0.5154,
+      "step": 9726
+    },
+    {
+      "epoch": 1.2663022256930887,
+      "grad_norm": 2.8390650749206543,
+      "learning_rate": 6.4057154961749324e-06,
+      "loss": 0.5305,
+      "step": 9729
+    },
+    {
+      "epoch": 1.2666926981647793,
+      "grad_norm": 2.5285894870758057,
+      "learning_rate": 6.403709821620198e-06,
+      "loss": 0.4643,
+      "step": 9732
+    },
+    {
+      "epoch": 1.2670831706364702,
+      "grad_norm": 2.475567579269409,
+      "learning_rate": 6.401703901840311e-06,
+      "loss": 0.52,
+      "step": 9735
+    },
+    {
+      "epoch": 1.2674736431081608,
+      "grad_norm": 2.560523271560669,
+      "learning_rate": 6.3996977371857e-06,
+      "loss": 0.5761,
+      "step": 9738
+    },
+    {
+      "epoch": 1.2678641155798516,
+      "grad_norm": 2.965449810028076,
+      "learning_rate": 6.397691328006839e-06,
+      "loss": 0.481,
+      "step": 9741
+    },
+    {
+      "epoch": 1.2682545880515423,
+      "grad_norm": 2.673515558242798,
+      "learning_rate": 6.395684674654245e-06,
+      "loss": 0.5183,
+      "step": 9744
+    },
+    {
+      "epoch": 1.2686450605232331,
+      "grad_norm": 2.865999221801758,
+      "learning_rate": 6.393677777478473e-06,
+      "loss": 0.5216,
+      "step": 9747
+    },
+    {
+      "epoch": 1.2690355329949239,
+      "grad_norm": 2.5293450355529785,
+      "learning_rate": 6.391670636830126e-06,
+      "loss": 0.496,
+      "step": 9750
+    },
+    {
+      "epoch": 1.2694260054666147,
+      "grad_norm": 2.695828914642334,
+      "learning_rate": 6.389663253059846e-06,
+      "loss": 0.5161,
+      "step": 9753
+    },
+    {
+      "epoch": 1.2698164779383054,
+      "grad_norm": 3.5010106563568115,
+      "learning_rate": 6.3876556265183185e-06,
+      "loss": 0.4878,
+      "step": 9756
+    },
+    {
+      "epoch": 1.270206950409996,
+      "grad_norm": 3.126572847366333,
+      "learning_rate": 6.3856477575562735e-06,
+      "loss": 0.5395,
+      "step": 9759
+    },
+    {
+      "epoch": 1.270597422881687,
+      "grad_norm": 3.661149024963379,
+      "learning_rate": 6.38363964652448e-06,
+      "loss": 0.5382,
+      "step": 9762
+    },
+    {
+      "epoch": 1.2709878953533775,
+      "grad_norm": 2.6320817470550537,
+      "learning_rate": 6.381631293773751e-06,
+      "loss": 0.5113,
+      "step": 9765
+    },
+    {
+      "epoch": 1.2713783678250683,
+      "grad_norm": 2.5496699810028076,
+      "learning_rate": 6.3796226996549404e-06,
+      "loss": 0.5602,
+      "step": 9768
+    },
+    {
+      "epoch": 1.271768840296759,
+      "grad_norm": 2.5300614833831787,
+      "learning_rate": 6.3776138645189475e-06,
+      "loss": 0.5367,
+      "step": 9771
+    },
+    {
+      "epoch": 1.2721593127684498,
+      "grad_norm": 2.5111279487609863,
+      "learning_rate": 6.37560478871671e-06,
+      "loss": 0.4835,
+      "step": 9774
+    },
+    {
+      "epoch": 1.2725497852401406,
+      "grad_norm": 2.6758015155792236,
+      "learning_rate": 6.37359547259921e-06,
+      "loss": 0.5407,
+      "step": 9777
+    },
+    {
+      "epoch": 1.2729402577118314,
+      "grad_norm": 2.8629322052001953,
+      "learning_rate": 6.371585916517471e-06,
+      "loss": 0.4738,
+      "step": 9780
+    },
+    {
+      "epoch": 1.2733307301835222,
+      "grad_norm": 2.585297107696533,
+      "learning_rate": 6.3695761208225585e-06,
+      "loss": 0.496,
+      "step": 9783
+    },
+    {
+      "epoch": 1.2737212026552127,
+      "grad_norm": 3.2529196739196777,
+      "learning_rate": 6.3675660858655765e-06,
+      "loss": 0.4647,
+      "step": 9786
+    },
+    {
+      "epoch": 1.2741116751269035,
+      "grad_norm": 2.405118703842163,
+      "learning_rate": 6.3655558119976765e-06,
+      "loss": 0.5462,
+      "step": 9789
+    },
+    {
+      "epoch": 1.2745021475985943,
+      "grad_norm": 2.579508066177368,
+      "learning_rate": 6.363545299570051e-06,
+      "loss": 0.4761,
+      "step": 9792
+    },
+    {
+      "epoch": 1.274892620070285,
+      "grad_norm": 2.9745607376098633,
+      "learning_rate": 6.361534548933928e-06,
+      "loss": 0.5342,
+      "step": 9795
+    },
+    {
+      "epoch": 1.2752830925419758,
+      "grad_norm": 2.931020975112915,
+      "learning_rate": 6.359523560440585e-06,
+      "loss": 0.4915,
+      "step": 9798
+    },
+    {
+      "epoch": 1.2756735650136666,
+      "grad_norm": 2.9338982105255127,
+      "learning_rate": 6.357512334441336e-06,
+      "loss": 0.6454,
+      "step": 9801
+    },
+    {
+      "epoch": 1.2760640374853573,
+      "grad_norm": 2.669160842895508,
+      "learning_rate": 6.355500871287538e-06,
+      "loss": 0.4672,
+      "step": 9804
+    },
+    {
+      "epoch": 1.276454509957048,
+      "grad_norm": 2.584463119506836,
+      "learning_rate": 6.353489171330588e-06,
+      "loss": 0.5323,
+      "step": 9807
+    },
+    {
+      "epoch": 1.276844982428739,
+      "grad_norm": 2.854555606842041,
+      "learning_rate": 6.351477234921928e-06,
+      "loss": 0.4001,
+      "step": 9810
+    },
+    {
+      "epoch": 1.2772354549004294,
+      "grad_norm": 3.679852247238159,
+      "learning_rate": 6.349465062413038e-06,
+      "loss": 0.5706,
+      "step": 9813
+    },
+    {
+      "epoch": 1.2776259273721202,
+      "grad_norm": 2.8256568908691406,
+      "learning_rate": 6.34745265415544e-06,
+      "loss": 0.4925,
+      "step": 9816
+    },
+    {
+      "epoch": 1.278016399843811,
+      "grad_norm": 3.215463876724243,
+      "learning_rate": 6.3454400105006985e-06,
+      "loss": 0.5522,
+      "step": 9819
+    },
+    {
+      "epoch": 1.2784068723155018,
+      "grad_norm": 2.2167320251464844,
+      "learning_rate": 6.343427131800417e-06,
+      "loss": 0.4712,
+      "step": 9822
+    },
+    {
+      "epoch": 1.2787973447871925,
+      "grad_norm": 2.5765607357025146,
+      "learning_rate": 6.341414018406242e-06,
+      "loss": 0.5017,
+      "step": 9825
+    },
+    {
+      "epoch": 1.2791878172588833,
+      "grad_norm": 2.5398025512695312,
+      "learning_rate": 6.3394006706698615e-06,
+      "loss": 0.4501,
+      "step": 9828
+    },
+    {
+      "epoch": 1.279578289730574,
+      "grad_norm": 2.628095865249634,
+      "learning_rate": 6.337387088943e-06,
+      "loss": 0.5367,
+      "step": 9831
+    },
+    {
+      "epoch": 1.2799687622022646,
+      "grad_norm": 2.486548662185669,
+      "learning_rate": 6.335373273577429e-06,
+      "loss": 0.5093,
+      "step": 9834
+    },
+    {
+      "epoch": 1.2803592346739554,
+      "grad_norm": 2.8743481636047363,
+      "learning_rate": 6.333359224924955e-06,
+      "loss": 0.4291,
+      "step": 9837
+    },
+    {
+      "epoch": 1.2807497071456462,
+      "grad_norm": 2.8362040519714355,
+      "learning_rate": 6.331344943337428e-06,
+      "loss": 0.4317,
+      "step": 9840
+    },
+    {
+      "epoch": 1.281140179617337,
+      "grad_norm": 2.6773505210876465,
+      "learning_rate": 6.329330429166741e-06,
+      "loss": 0.4869,
+      "step": 9843
+    },
+    {
+      "epoch": 1.2815306520890277,
+      "grad_norm": 2.4184865951538086,
+      "learning_rate": 6.327315682764825e-06,
+      "loss": 0.5656,
+      "step": 9846
+    },
+    {
+      "epoch": 1.2819211245607185,
+      "grad_norm": 2.5900416374206543,
+      "learning_rate": 6.325300704483653e-06,
+      "loss": 0.4823,
+      "step": 9849
+    },
+    {
+      "epoch": 1.2823115970324093,
+      "grad_norm": 2.582472085952759,
+      "learning_rate": 6.3232854946752345e-06,
+      "loss": 0.4939,
+      "step": 9852
+    },
+    {
+      "epoch": 1.2827020695041,
+      "grad_norm": 2.6832892894744873,
+      "learning_rate": 6.321270053691624e-06,
+      "loss": 0.499,
+      "step": 9855
+    },
+    {
+      "epoch": 1.2830925419757908,
+      "grad_norm": 2.630359649658203,
+      "learning_rate": 6.319254381884914e-06,
+      "loss": 0.4966,
+      "step": 9858
+    },
+    {
+      "epoch": 1.2834830144474814,
+      "grad_norm": 2.7337570190429688,
+      "learning_rate": 6.317238479607239e-06,
+      "loss": 0.4762,
+      "step": 9861
+    },
+    {
+      "epoch": 1.2838734869191721,
+      "grad_norm": 2.502153158187866,
+      "learning_rate": 6.315222347210773e-06,
+      "loss": 0.5532,
+      "step": 9864
+    },
+    {
+      "epoch": 1.284263959390863,
+      "grad_norm": 2.532254934310913,
+      "learning_rate": 6.31320598504773e-06,
+      "loss": 0.5093,
+      "step": 9867
+    },
+    {
+      "epoch": 1.2846544318625537,
+      "grad_norm": 2.7918901443481445,
+      "learning_rate": 6.311189393470364e-06,
+      "loss": 0.5242,
+      "step": 9870
+    },
+    {
+      "epoch": 1.2850449043342445,
+      "grad_norm": 2.472677230834961,
+      "learning_rate": 6.309172572830969e-06,
+      "loss": 0.4492,
+      "step": 9873
+    },
+    {
+      "epoch": 1.2854353768059352,
+      "grad_norm": 2.5617218017578125,
+      "learning_rate": 6.30715552348188e-06,
+      "loss": 0.4832,
+      "step": 9876
+    },
+    {
+      "epoch": 1.285825849277626,
+      "grad_norm": 2.8872265815734863,
+      "learning_rate": 6.30513824577547e-06,
+      "loss": 0.5347,
+      "step": 9879
+    },
+    {
+      "epoch": 1.2862163217493165,
+      "grad_norm": 2.543060302734375,
+      "learning_rate": 6.3031207400641535e-06,
+      "loss": 0.553,
+      "step": 9882
+    },
+    {
+      "epoch": 1.2866067942210075,
+      "grad_norm": 2.5379858016967773,
+      "learning_rate": 6.301103006700388e-06,
+      "loss": 0.4914,
+      "step": 9885
+    },
+    {
+      "epoch": 1.286997266692698,
+      "grad_norm": 2.7542781829833984,
+      "learning_rate": 6.299085046036662e-06,
+      "loss": 0.4767,
+      "step": 9888
+    },
+    {
+      "epoch": 1.2873877391643889,
+      "grad_norm": 2.86118745803833,
+      "learning_rate": 6.297066858425512e-06,
+      "loss": 0.4647,
+      "step": 9891
+    },
+    {
+      "epoch": 1.2877782116360796,
+      "grad_norm": 4.252162456512451,
+      "learning_rate": 6.29504844421951e-06,
+      "loss": 0.4904,
+      "step": 9894
+    },
+    {
+      "epoch": 1.2881686841077704,
+      "grad_norm": 3.537123441696167,
+      "learning_rate": 6.2930298037712704e-06,
+      "loss": 0.5455,
+      "step": 9897
+    },
+    {
+      "epoch": 1.2885591565794612,
+      "grad_norm": 3.0261664390563965,
+      "learning_rate": 6.2910109374334434e-06,
+      "loss": 0.5058,
+      "step": 9900
+    },
+    {
+      "epoch": 1.288949629051152,
+      "grad_norm": 2.73046612739563,
+      "learning_rate": 6.288991845558721e-06,
+      "loss": 0.5526,
+      "step": 9903
+    },
+    {
+      "epoch": 1.2893401015228427,
+      "grad_norm": 2.695645332336426,
+      "learning_rate": 6.286972528499835e-06,
+      "loss": 0.6358,
+      "step": 9906
+    },
+    {
+      "epoch": 1.2897305739945333,
+      "grad_norm": 2.4671976566314697,
+      "learning_rate": 6.284952986609556e-06,
+      "loss": 0.5519,
+      "step": 9909
+    },
+    {
+      "epoch": 1.290121046466224,
+      "grad_norm": 2.7098228931427,
+      "learning_rate": 6.282933220240695e-06,
+      "loss": 0.4461,
+      "step": 9912
+    },
+    {
+      "epoch": 1.2905115189379148,
+      "grad_norm": 2.5738272666931152,
+      "learning_rate": 6.280913229746096e-06,
+      "loss": 0.4398,
+      "step": 9915
+    },
+    {
+      "epoch": 1.2909019914096056,
+      "grad_norm": 2.765180826187134,
+      "learning_rate": 6.278893015478652e-06,
+      "loss": 0.6561,
+      "step": 9918
+    },
+    {
+      "epoch": 1.2912924638812964,
+      "grad_norm": 2.4514658451080322,
+      "learning_rate": 6.27687257779129e-06,
+      "loss": 0.4876,
+      "step": 9921
+    },
+    {
+      "epoch": 1.2916829363529871,
+      "grad_norm": 2.56715726852417,
+      "learning_rate": 6.274851917036971e-06,
+      "loss": 0.459,
+      "step": 9924
+    },
+    {
+      "epoch": 1.292073408824678,
+      "grad_norm": 2.630765438079834,
+      "learning_rate": 6.272831033568708e-06,
+      "loss": 0.4325,
+      "step": 9927
+    },
+    {
+      "epoch": 1.2924638812963687,
+      "grad_norm": 2.567124843597412,
+      "learning_rate": 6.27080992773954e-06,
+      "loss": 0.4577,
+      "step": 9930
+    },
+    {
+      "epoch": 1.2928543537680595,
+      "grad_norm": 2.8632030487060547,
+      "learning_rate": 6.26878859990255e-06,
+      "loss": 0.5061,
+      "step": 9933
+    },
+    {
+      "epoch": 1.29324482623975,
+      "grad_norm": 3.047332525253296,
+      "learning_rate": 6.266767050410862e-06,
+      "loss": 0.4394,
+      "step": 9936
+    },
+    {
+      "epoch": 1.2936352987114408,
+      "grad_norm": 2.710806131362915,
+      "learning_rate": 6.264745279617634e-06,
+      "loss": 0.4752,
+      "step": 9939
+    },
+    {
+      "epoch": 1.2940257711831316,
+      "grad_norm": 2.510077714920044,
+      "learning_rate": 6.262723287876068e-06,
+      "loss": 0.4945,
+      "step": 9942
+    },
+    {
+      "epoch": 1.2944162436548223,
+      "grad_norm": 3.1183907985687256,
+      "learning_rate": 6.260701075539397e-06,
+      "loss": 0.4816,
+      "step": 9945
+    },
+    {
+      "epoch": 1.294806716126513,
+      "grad_norm": 2.8329877853393555,
+      "learning_rate": 6.258678642960902e-06,
+      "loss": 0.4994,
+      "step": 9948
+    },
+    {
+      "epoch": 1.2951971885982039,
+      "grad_norm": 2.953737735748291,
+      "learning_rate": 6.256655990493896e-06,
+      "loss": 0.4451,
+      "step": 9951
+    },
+    {
+      "epoch": 1.2955876610698946,
+      "grad_norm": 2.770921468734741,
+      "learning_rate": 6.254633118491732e-06,
+      "loss": 0.5525,
+      "step": 9954
+    },
+    {
+      "epoch": 1.2959781335415852,
+      "grad_norm": 2.6168859004974365,
+      "learning_rate": 6.252610027307803e-06,
+      "loss": 0.4346,
+      "step": 9957
+    },
+    {
+      "epoch": 1.2963686060132762,
+      "grad_norm": 2.5253684520721436,
+      "learning_rate": 6.250586717295535e-06,
+      "loss": 0.4968,
+      "step": 9960
+    },
+    {
+      "epoch": 1.2967590784849667,
+      "grad_norm": 2.5443472862243652,
+      "learning_rate": 6.248563188808401e-06,
+      "loss": 0.4767,
+      "step": 9963
+    },
+    {
+      "epoch": 1.2971495509566575,
+      "grad_norm": 2.66750431060791,
+      "learning_rate": 6.246539442199901e-06,
+      "loss": 0.5305,
+      "step": 9966
+    },
+    {
+      "epoch": 1.2975400234283483,
+      "grad_norm": 3.3093185424804688,
+      "learning_rate": 6.244515477823585e-06,
+      "loss": 0.5501,
+      "step": 9969
+    },
+    {
+      "epoch": 1.297930495900039,
+      "grad_norm": 2.5666654109954834,
+      "learning_rate": 6.242491296033033e-06,
+      "loss": 0.5458,
+      "step": 9972
+    },
+    {
+      "epoch": 1.2983209683717298,
+      "grad_norm": 2.224759340286255,
+      "learning_rate": 6.240466897181865e-06,
+      "loss": 0.4821,
+      "step": 9975
+    },
+    {
+      "epoch": 1.2987114408434206,
+      "grad_norm": 2.936877727508545,
+      "learning_rate": 6.23844228162374e-06,
+      "loss": 0.4621,
+      "step": 9978
+    },
+    {
+      "epoch": 1.2991019133151114,
+      "grad_norm": 2.833364248275757,
+      "learning_rate": 6.236417449712353e-06,
+      "loss": 0.4913,
+      "step": 9981
+    },
+    {
+      "epoch": 1.299492385786802,
+      "grad_norm": 2.554704189300537,
+      "learning_rate": 6.23439240180144e-06,
+      "loss": 0.4368,
+      "step": 9984
+    },
+    {
+      "epoch": 1.2998828582584927,
+      "grad_norm": 2.432718276977539,
+      "learning_rate": 6.232367138244768e-06,
+      "loss": 0.5568,
+      "step": 9987
+    },
+    {
+      "epoch": 1.3002733307301835,
+      "grad_norm": 2.5396578311920166,
+      "learning_rate": 6.230341659396152e-06,
+      "loss": 0.5137,
+      "step": 9990
+    },
+    {
+      "epoch": 1.3006638032018742,
+      "grad_norm": 2.720970392227173,
+      "learning_rate": 6.228315965609437e-06,
+      "loss": 0.474,
+      "step": 9993
+    },
+    {
+      "epoch": 1.301054275673565,
+      "grad_norm": 3.310605525970459,
+      "learning_rate": 6.226290057238506e-06,
+      "loss": 0.4991,
+      "step": 9996
+    },
+    {
+      "epoch": 1.3014447481452558,
+      "grad_norm": 2.8268284797668457,
+      "learning_rate": 6.224263934637281e-06,
+      "loss": 0.3968,
+      "step": 9999
+    },
+    {
+      "epoch": 1.3018352206169466,
+      "grad_norm": 2.8904576301574707,
+      "learning_rate": 6.222237598159723e-06,
+      "loss": 0.5225,
+      "step": 10002
+    },
+    {
+      "epoch": 1.3022256930886371,
+      "grad_norm": 2.540018081665039,
+      "learning_rate": 6.220211048159826e-06,
+      "loss": 0.4189,
+      "step": 10005
+    },
+    {
+      "epoch": 1.302616165560328,
+      "grad_norm": 2.706920862197876,
+      "learning_rate": 6.2181842849916284e-06,
+      "loss": 0.5126,
+      "step": 10008
+    },
+    {
+      "epoch": 1.3030066380320187,
+      "grad_norm": 3.605525016784668,
+      "learning_rate": 6.216157309009198e-06,
+      "loss": 0.4911,
+      "step": 10011
+    },
+    {
+      "epoch": 1.3033971105037094,
+      "grad_norm": 2.4646642208099365,
+      "learning_rate": 6.214130120566643e-06,
+      "loss": 0.4261,
+      "step": 10014
+    },
+    {
+      "epoch": 1.3037875829754002,
+      "grad_norm": 2.7980430126190186,
+      "learning_rate": 6.212102720018112e-06,
+      "loss": 0.483,
+      "step": 10017
+    },
+    {
+      "epoch": 1.304178055447091,
+      "grad_norm": 2.9838016033172607,
+      "learning_rate": 6.210075107717785e-06,
+      "loss": 0.4596,
+      "step": 10020
+    },
+    {
+      "epoch": 1.3045685279187818,
+      "grad_norm": 2.551286220550537,
+      "learning_rate": 6.208047284019881e-06,
+      "loss": 0.4226,
+      "step": 10023
+    },
+    {
+      "epoch": 1.3049590003904725,
+      "grad_norm": 2.640700340270996,
+      "learning_rate": 6.20601924927866e-06,
+      "loss": 0.4905,
+      "step": 10026
+    },
+    {
+      "epoch": 1.3053494728621633,
+      "grad_norm": 2.759453296661377,
+      "learning_rate": 6.203991003848411e-06,
+      "loss": 0.467,
+      "step": 10029
+    },
+    {
+      "epoch": 1.3057399453338538,
+      "grad_norm": 2.907013177871704,
+      "learning_rate": 6.201962548083468e-06,
+      "loss": 0.487,
+      "step": 10032
+    },
+    {
+      "epoch": 1.3061304178055448,
+      "grad_norm": 2.470888376235962,
+      "learning_rate": 6.199933882338196e-06,
+      "loss": 0.4687,
+      "step": 10035
+    },
+    {
+      "epoch": 1.3065208902772354,
+      "grad_norm": 2.576772451400757,
+      "learning_rate": 6.197905006966999e-06,
+      "loss": 0.425,
+      "step": 10038
+    },
+    {
+      "epoch": 1.3069113627489262,
+      "grad_norm": 2.713982582092285,
+      "learning_rate": 6.195875922324318e-06,
+      "loss": 0.4787,
+      "step": 10041
+    },
+    {
+      "epoch": 1.307301835220617,
+      "grad_norm": 2.516087532043457,
+      "learning_rate": 6.1938466287646285e-06,
+      "loss": 0.4707,
+      "step": 10044
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 2.569885492324829,
+      "learning_rate": 6.191817126642444e-06,
+      "loss": 0.458,
+      "step": 10047
+    },
+    {
+      "epoch": 1.3080827801639985,
+      "grad_norm": 3.064462184906006,
+      "learning_rate": 6.189787416312315e-06,
+      "loss": 0.5624,
+      "step": 10050
+    },
+    {
+      "epoch": 1.3084732526356893,
+      "grad_norm": 2.9090516567230225,
+      "learning_rate": 6.187757498128827e-06,
+      "loss": 0.5414,
+      "step": 10053
+    },
+    {
+      "epoch": 1.30886372510738,
+      "grad_norm": 2.532334566116333,
+      "learning_rate": 6.185727372446604e-06,
+      "loss": 0.5469,
+      "step": 10056
+    },
+    {
+      "epoch": 1.3092541975790706,
+      "grad_norm": 2.665684223175049,
+      "learning_rate": 6.183697039620302e-06,
+      "loss": 0.5266,
+      "step": 10059
+    },
+    {
+      "epoch": 1.3096446700507614,
+      "grad_norm": 2.6369564533233643,
+      "learning_rate": 6.181666500004617e-06,
+      "loss": 0.4774,
+      "step": 10062
+    },
+    {
+      "epoch": 1.3100351425224521,
+      "grad_norm": 2.548985481262207,
+      "learning_rate": 6.179635753954283e-06,
+      "loss": 0.4342,
+      "step": 10065
+    },
+    {
+      "epoch": 1.310425614994143,
+      "grad_norm": 2.7811813354492188,
+      "learning_rate": 6.177604801824062e-06,
+      "loss": 0.4539,
+      "step": 10068
+    },
+    {
+      "epoch": 1.3108160874658337,
+      "grad_norm": 2.4228222370147705,
+      "learning_rate": 6.17557364396876e-06,
+      "loss": 0.446,
+      "step": 10071
+    },
+    {
+      "epoch": 1.3112065599375244,
+      "grad_norm": 2.639486789703369,
+      "learning_rate": 6.173542280743214e-06,
+      "loss": 0.5583,
+      "step": 10074
+    },
+    {
+      "epoch": 1.3115970324092152,
+      "grad_norm": 3.5657029151916504,
+      "learning_rate": 6.171510712502303e-06,
+      "loss": 0.5388,
+      "step": 10077
+    },
+    {
+      "epoch": 1.3119875048809058,
+      "grad_norm": 2.815169095993042,
+      "learning_rate": 6.169478939600933e-06,
+      "loss": 0.5293,
+      "step": 10080
+    },
+    {
+      "epoch": 1.3123779773525968,
+      "grad_norm": 2.9462995529174805,
+      "learning_rate": 6.167446962394054e-06,
+      "loss": 0.5047,
+      "step": 10083
+    },
+    {
+      "epoch": 1.3127684498242873,
+      "grad_norm": 2.7291252613067627,
+      "learning_rate": 6.165414781236647e-06,
+      "loss": 0.4032,
+      "step": 10086
+    },
+    {
+      "epoch": 1.313158922295978,
+      "grad_norm": 2.6978507041931152,
+      "learning_rate": 6.163382396483728e-06,
+      "loss": 0.5027,
+      "step": 10089
+    },
+    {
+      "epoch": 1.3135493947676689,
+      "grad_norm": 2.74324369430542,
+      "learning_rate": 6.161349808490351e-06,
+      "loss": 0.4742,
+      "step": 10092
+    },
+    {
+      "epoch": 1.3139398672393596,
+      "grad_norm": 2.520897388458252,
+      "learning_rate": 6.159317017611607e-06,
+      "loss": 0.4695,
+      "step": 10095
+    },
+    {
+      "epoch": 1.3143303397110504,
+      "grad_norm": 2.367314338684082,
+      "learning_rate": 6.157284024202619e-06,
+      "loss": 0.4521,
+      "step": 10098
+    },
+    {
+      "epoch": 1.3147208121827412,
+      "grad_norm": 2.571027994155884,
+      "learning_rate": 6.155250828618547e-06,
+      "loss": 0.4327,
+      "step": 10101
+    },
+    {
+      "epoch": 1.315111284654432,
+      "grad_norm": 2.6807379722595215,
+      "learning_rate": 6.153217431214583e-06,
+      "loss": 0.5297,
+      "step": 10104
+    },
+    {
+      "epoch": 1.3155017571261225,
+      "grad_norm": 2.974998950958252,
+      "learning_rate": 6.1511838323459624e-06,
+      "loss": 0.5418,
+      "step": 10107
+    },
+    {
+      "epoch": 1.3158922295978135,
+      "grad_norm": 2.9916090965270996,
+      "learning_rate": 6.149150032367946e-06,
+      "loss": 0.4774,
+      "step": 10110
+    },
+    {
+      "epoch": 1.316282702069504,
+      "grad_norm": 3.1814825534820557,
+      "learning_rate": 6.147116031635838e-06,
+      "loss": 0.5896,
+      "step": 10113
+    },
+    {
+      "epoch": 1.3166731745411948,
+      "grad_norm": 2.669076681137085,
+      "learning_rate": 6.145081830504971e-06,
+      "loss": 0.4869,
+      "step": 10116
+    },
+    {
+      "epoch": 1.3170636470128856,
+      "grad_norm": 2.4291157722473145,
+      "learning_rate": 6.1430474293307175e-06,
+      "loss": 0.4442,
+      "step": 10119
+    },
+    {
+      "epoch": 1.3174541194845764,
+      "grad_norm": 2.7454960346221924,
+      "learning_rate": 6.141012828468484e-06,
+      "loss": 0.506,
+      "step": 10122
+    },
+    {
+      "epoch": 1.3178445919562671,
+      "grad_norm": 2.744112730026245,
+      "learning_rate": 6.138978028273709e-06,
+      "loss": 0.5368,
+      "step": 10125
+    },
+    {
+      "epoch": 1.318235064427958,
+      "grad_norm": 2.635687828063965,
+      "learning_rate": 6.13694302910187e-06,
+      "loss": 0.4753,
+      "step": 10128
+    },
+    {
+      "epoch": 1.3186255368996487,
+      "grad_norm": 2.746645212173462,
+      "learning_rate": 6.134907831308473e-06,
+      "loss": 0.4922,
+      "step": 10131
+    },
+    {
+      "epoch": 1.3190160093713392,
+      "grad_norm": 2.9933664798736572,
+      "learning_rate": 6.132872435249067e-06,
+      "loss": 0.5043,
+      "step": 10134
+    },
+    {
+      "epoch": 1.31940648184303,
+      "grad_norm": 2.8414406776428223,
+      "learning_rate": 6.130836841279228e-06,
+      "loss": 0.4989,
+      "step": 10137
+    },
+    {
+      "epoch": 1.3197969543147208,
+      "grad_norm": 2.6284008026123047,
+      "learning_rate": 6.128801049754572e-06,
+      "loss": 0.5045,
+      "step": 10140
+    },
+    {
+      "epoch": 1.3201874267864115,
+      "grad_norm": 2.5800864696502686,
+      "learning_rate": 6.1267650610307496e-06,
+      "loss": 0.4773,
+      "step": 10143
+    },
+    {
+      "epoch": 1.3205778992581023,
+      "grad_norm": 3.2710795402526855,
+      "learning_rate": 6.1247288754634395e-06,
+      "loss": 0.4755,
+      "step": 10146
+    },
+    {
+      "epoch": 1.320968371729793,
+      "grad_norm": 2.9163713455200195,
+      "learning_rate": 6.122692493408362e-06,
+      "loss": 0.5315,
+      "step": 10149
+    },
+    {
+      "epoch": 1.3213588442014839,
+      "grad_norm": 3.6550145149230957,
+      "learning_rate": 6.120655915221268e-06,
+      "loss": 0.5406,
+      "step": 10152
+    },
+    {
+      "epoch": 1.3217493166731744,
+      "grad_norm": 2.446315050125122,
+      "learning_rate": 6.118619141257941e-06,
+      "loss": 0.4987,
+      "step": 10155
+    },
+    {
+      "epoch": 1.3221397891448654,
+      "grad_norm": 2.639047622680664,
+      "learning_rate": 6.116582171874204e-06,
+      "loss": 0.497,
+      "step": 10158
+    },
+    {
+      "epoch": 1.322530261616556,
+      "grad_norm": 3.0936222076416016,
+      "learning_rate": 6.11454500742591e-06,
+      "loss": 0.451,
+      "step": 10161
+    },
+    {
+      "epoch": 1.3229207340882467,
+      "grad_norm": 2.499833106994629,
+      "learning_rate": 6.112507648268951e-06,
+      "loss": 0.5365,
+      "step": 10164
+    },
+    {
+      "epoch": 1.3233112065599375,
+      "grad_norm": 2.466047763824463,
+      "learning_rate": 6.110470094759243e-06,
+      "loss": 0.4519,
+      "step": 10167
+    },
+    {
+      "epoch": 1.3237016790316283,
+      "grad_norm": 2.9250972270965576,
+      "learning_rate": 6.1084323472527465e-06,
+      "loss": 0.4602,
+      "step": 10170
+    },
+    {
+      "epoch": 1.324092151503319,
+      "grad_norm": 2.504556894302368,
+      "learning_rate": 6.106394406105451e-06,
+      "loss": 0.4688,
+      "step": 10173
+    },
+    {
+      "epoch": 1.3244826239750098,
+      "grad_norm": 3.0200345516204834,
+      "learning_rate": 6.104356271673379e-06,
+      "loss": 0.4318,
+      "step": 10176
+    },
+    {
+      "epoch": 1.3248730964467006,
+      "grad_norm": 2.429672956466675,
+      "learning_rate": 6.102317944312592e-06,
+      "loss": 0.4749,
+      "step": 10179
+    },
+    {
+      "epoch": 1.3252635689183911,
+      "grad_norm": 2.6296794414520264,
+      "learning_rate": 6.1002794243791774e-06,
+      "loss": 0.4971,
+      "step": 10182
+    },
+    {
+      "epoch": 1.325654041390082,
+      "grad_norm": 2.859323024749756,
+      "learning_rate": 6.098240712229263e-06,
+      "loss": 0.553,
+      "step": 10185
+    },
+    {
+      "epoch": 1.3260445138617727,
+      "grad_norm": 2.4780797958374023,
+      "learning_rate": 6.096201808219005e-06,
+      "loss": 0.4585,
+      "step": 10188
+    },
+    {
+      "epoch": 1.3264349863334635,
+      "grad_norm": 2.843698263168335,
+      "learning_rate": 6.094162712704599e-06,
+      "loss": 0.5146,
+      "step": 10191
+    },
+    {
+      "epoch": 1.3268254588051542,
+      "grad_norm": 2.588787317276001,
+      "learning_rate": 6.0921234260422675e-06,
+      "loss": 0.5512,
+      "step": 10194
+    },
+    {
+      "epoch": 1.327215931276845,
+      "grad_norm": 2.7030463218688965,
+      "learning_rate": 6.090083948588271e-06,
+      "loss": 0.4705,
+      "step": 10197
+    },
+    {
+      "epoch": 1.3276064037485358,
+      "grad_norm": 2.659209966659546,
+      "learning_rate": 6.088044280698903e-06,
+      "loss": 0.5182,
+      "step": 10200
+    },
+    {
+      "epoch": 1.3279968762202266,
+      "grad_norm": 2.3089091777801514,
+      "learning_rate": 6.086004422730487e-06,
+      "loss": 0.5118,
+      "step": 10203
+    },
+    {
+      "epoch": 1.3283873486919173,
+      "grad_norm": 3.2967066764831543,
+      "learning_rate": 6.083964375039384e-06,
+      "loss": 0.4932,
+      "step": 10206
+    },
+    {
+      "epoch": 1.3287778211636079,
+      "grad_norm": 3.2022552490234375,
+      "learning_rate": 6.081924137981984e-06,
+      "loss": 0.5659,
+      "step": 10209
+    },
+    {
+      "epoch": 1.3291682936352986,
+      "grad_norm": 2.5667223930358887,
+      "learning_rate": 6.079883711914713e-06,
+      "loss": 0.4576,
+      "step": 10212
+    },
+    {
+      "epoch": 1.3295587661069894,
+      "grad_norm": 2.711458444595337,
+      "learning_rate": 6.07784309719403e-06,
+      "loss": 0.4464,
+      "step": 10215
+    },
+    {
+      "epoch": 1.3299492385786802,
+      "grad_norm": 2.9583401679992676,
+      "learning_rate": 6.075802294176425e-06,
+      "loss": 0.5373,
+      "step": 10218
+    },
+    {
+      "epoch": 1.330339711050371,
+      "grad_norm": 4.091579914093018,
+      "learning_rate": 6.073761303218423e-06,
+      "loss": 0.5137,
+      "step": 10221
+    },
+    {
+      "epoch": 1.3307301835220617,
+      "grad_norm": 2.6844708919525146,
+      "learning_rate": 6.071720124676579e-06,
+      "loss": 0.6416,
+      "step": 10224
+    },
+    {
+      "epoch": 1.3311206559937525,
+      "grad_norm": 2.5595126152038574,
+      "learning_rate": 6.069678758907486e-06,
+      "loss": 0.55,
+      "step": 10227
+    },
+    {
+      "epoch": 1.331511128465443,
+      "grad_norm": 2.9795103073120117,
+      "learning_rate": 6.067637206267761e-06,
+      "loss": 0.4621,
+      "step": 10230
+    },
+    {
+      "epoch": 1.331901600937134,
+      "grad_norm": 2.5117387771606445,
+      "learning_rate": 6.065595467114064e-06,
+      "loss": 0.4778,
+      "step": 10233
+    },
+    {
+      "epoch": 1.3322920734088246,
+      "grad_norm": 2.6409218311309814,
+      "learning_rate": 6.063553541803081e-06,
+      "loss": 0.4728,
+      "step": 10236
+    },
+    {
+      "epoch": 1.3326825458805154,
+      "grad_norm": 2.658996105194092,
+      "learning_rate": 6.061511430691529e-06,
+      "loss": 0.4789,
+      "step": 10239
+    },
+    {
+      "epoch": 1.3330730183522062,
+      "grad_norm": 2.801865339279175,
+      "learning_rate": 6.059469134136167e-06,
+      "loss": 0.4593,
+      "step": 10242
+    },
+    {
+      "epoch": 1.333463490823897,
+      "grad_norm": 2.8045008182525635,
+      "learning_rate": 6.057426652493773e-06,
+      "loss": 0.5459,
+      "step": 10245
+    },
+    {
+      "epoch": 1.3338539632955877,
+      "grad_norm": 2.777100086212158,
+      "learning_rate": 6.055383986121169e-06,
+      "loss": 0.578,
+      "step": 10248
+    },
+    {
+      "epoch": 1.3342444357672785,
+      "grad_norm": 2.5679752826690674,
+      "learning_rate": 6.053341135375202e-06,
+      "loss": 0.4892,
+      "step": 10251
+    },
+    {
+      "epoch": 1.3346349082389692,
+      "grad_norm": 2.5535635948181152,
+      "learning_rate": 6.051298100612755e-06,
+      "loss": 0.5193,
+      "step": 10254
+    },
+    {
+      "epoch": 1.3350253807106598,
+      "grad_norm": 2.5816335678100586,
+      "learning_rate": 6.049254882190742e-06,
+      "loss": 0.5621,
+      "step": 10257
+    },
+    {
+      "epoch": 1.3354158531823506,
+      "grad_norm": 2.74711275100708,
+      "learning_rate": 6.047211480466105e-06,
+      "loss": 0.5361,
+      "step": 10260
+    },
+    {
+      "epoch": 1.3358063256540413,
+      "grad_norm": 2.556450605392456,
+      "learning_rate": 6.045167895795829e-06,
+      "loss": 0.4354,
+      "step": 10263
+    },
+    {
+      "epoch": 1.336196798125732,
+      "grad_norm": 2.905630588531494,
+      "learning_rate": 6.043124128536919e-06,
+      "loss": 0.4784,
+      "step": 10266
+    },
+    {
+      "epoch": 1.3365872705974229,
+      "grad_norm": 3.139017343521118,
+      "learning_rate": 6.041080179046418e-06,
+      "loss": 0.5801,
+      "step": 10269
+    },
+    {
+      "epoch": 1.3369777430691137,
+      "grad_norm": 2.844486713409424,
+      "learning_rate": 6.039036047681402e-06,
+      "loss": 0.5141,
+      "step": 10272
+    },
+    {
+      "epoch": 1.3373682155408044,
+      "grad_norm": 2.940715789794922,
+      "learning_rate": 6.036991734798971e-06,
+      "loss": 0.5246,
+      "step": 10275
+    },
+    {
+      "epoch": 1.3377586880124952,
+      "grad_norm": 2.516834020614624,
+      "learning_rate": 6.034947240756268e-06,
+      "loss": 0.4542,
+      "step": 10278
+    },
+    {
+      "epoch": 1.338149160484186,
+      "grad_norm": 2.7946300506591797,
+      "learning_rate": 6.032902565910456e-06,
+      "loss": 0.5008,
+      "step": 10281
+    },
+    {
+      "epoch": 1.3385396329558765,
+      "grad_norm": 2.5001931190490723,
+      "learning_rate": 6.030857710618743e-06,
+      "loss": 0.4844,
+      "step": 10284
+    },
+    {
+      "epoch": 1.3389301054275673,
+      "grad_norm": 2.5661776065826416,
+      "learning_rate": 6.0288126752383535e-06,
+      "loss": 0.4938,
+      "step": 10287
+    },
+    {
+      "epoch": 1.339320577899258,
+      "grad_norm": 2.576228618621826,
+      "learning_rate": 6.026767460126555e-06,
+      "loss": 0.5228,
+      "step": 10290
+    },
+    {
+      "epoch": 1.3397110503709488,
+      "grad_norm": 2.4661436080932617,
+      "learning_rate": 6.0247220656406415e-06,
+      "loss": 0.4381,
+      "step": 10293
+    },
+    {
+      "epoch": 1.3401015228426396,
+      "grad_norm": 2.5111136436462402,
+      "learning_rate": 6.022676492137939e-06,
+      "loss": 0.5015,
+      "step": 10296
+    },
+    {
+      "epoch": 1.3404919953143304,
+      "grad_norm": 2.6229360103607178,
+      "learning_rate": 6.020630739975803e-06,
+      "loss": 0.4764,
+      "step": 10299
+    },
+    {
+      "epoch": 1.3408824677860212,
+      "grad_norm": 2.8185629844665527,
+      "learning_rate": 6.018584809511625e-06,
+      "loss": 0.5273,
+      "step": 10302
+    },
+    {
+      "epoch": 1.3412729402577117,
+      "grad_norm": 2.5472960472106934,
+      "learning_rate": 6.0165387011028235e-06,
+      "loss": 0.4626,
+      "step": 10305
+    },
+    {
+      "epoch": 1.3416634127294027,
+      "grad_norm": 2.914748430252075,
+      "learning_rate": 6.01449241510685e-06,
+      "loss": 0.5434,
+      "step": 10308
+    },
+    {
+      "epoch": 1.3420538852010933,
+      "grad_norm": 2.704376220703125,
+      "learning_rate": 6.012445951881185e-06,
+      "loss": 0.5164,
+      "step": 10311
+    },
+    {
+      "epoch": 1.342444357672784,
+      "grad_norm": 2.3576114177703857,
+      "learning_rate": 6.010399311783343e-06,
+      "loss": 0.4096,
+      "step": 10314
+    },
+    {
+      "epoch": 1.3428348301444748,
+      "grad_norm": 2.456395387649536,
+      "learning_rate": 6.008352495170866e-06,
+      "loss": 0.4971,
+      "step": 10317
+    },
+    {
+      "epoch": 1.3432253026161656,
+      "grad_norm": 2.8089687824249268,
+      "learning_rate": 6.006305502401329e-06,
+      "loss": 0.4943,
+      "step": 10320
+    },
+    {
+      "epoch": 1.3436157750878563,
+      "grad_norm": 2.7782235145568848,
+      "learning_rate": 6.0042583338323376e-06,
+      "loss": 0.5548,
+      "step": 10323
+    },
+    {
+      "epoch": 1.3440062475595471,
+      "grad_norm": 2.7142183780670166,
+      "learning_rate": 6.002210989821528e-06,
+      "loss": 0.5179,
+      "step": 10326
+    },
+    {
+      "epoch": 1.344396720031238,
+      "grad_norm": 2.6159982681274414,
+      "learning_rate": 6.0001634707265675e-06,
+      "loss": 0.4984,
+      "step": 10329
+    },
+    {
+      "epoch": 1.3447871925029284,
+      "grad_norm": 2.9001638889312744,
+      "learning_rate": 5.998115776905152e-06,
+      "loss": 0.5373,
+      "step": 10332
+    },
+    {
+      "epoch": 1.3451776649746192,
+      "grad_norm": 2.7461464405059814,
+      "learning_rate": 5.996067908715012e-06,
+      "loss": 0.5562,
+      "step": 10335
+    },
+    {
+      "epoch": 1.34556813744631,
+      "grad_norm": 2.5092601776123047,
+      "learning_rate": 5.994019866513901e-06,
+      "loss": 0.5171,
+      "step": 10338
+    },
+    {
+      "epoch": 1.3459586099180008,
+      "grad_norm": 2.727618455886841,
+      "learning_rate": 5.991971650659612e-06,
+      "loss": 0.4846,
+      "step": 10341
+    },
+    {
+      "epoch": 1.3463490823896915,
+      "grad_norm": 2.589345932006836,
+      "learning_rate": 5.9899232615099626e-06,
+      "loss": 0.5663,
+      "step": 10344
+    },
+    {
+      "epoch": 1.3467395548613823,
+      "grad_norm": 3.144259452819824,
+      "learning_rate": 5.9878746994227996e-06,
+      "loss": 0.6201,
+      "step": 10347
+    },
+    {
+      "epoch": 1.347130027333073,
+      "grad_norm": 2.4682085514068604,
+      "learning_rate": 5.985825964756008e-06,
+      "loss": 0.4205,
+      "step": 10350
+    },
+    {
+      "epoch": 1.3475204998047636,
+      "grad_norm": 3.090812921524048,
+      "learning_rate": 5.9837770578674925e-06,
+      "loss": 0.4667,
+      "step": 10353
+    },
+    {
+      "epoch": 1.3479109722764546,
+      "grad_norm": 2.575993299484253,
+      "learning_rate": 5.981727979115195e-06,
+      "loss": 0.4956,
+      "step": 10356
+    },
+    {
+      "epoch": 1.3483014447481452,
+      "grad_norm": 2.3932414054870605,
+      "learning_rate": 5.979678728857086e-06,
+      "loss": 0.4554,
+      "step": 10359
+    },
+    {
+      "epoch": 1.348691917219836,
+      "grad_norm": 2.7043232917785645,
+      "learning_rate": 5.977629307451162e-06,
+      "loss": 0.5633,
+      "step": 10362
+    },
+    {
+      "epoch": 1.3490823896915267,
+      "grad_norm": 2.7044363021850586,
+      "learning_rate": 5.975579715255455e-06,
+      "loss": 0.431,
+      "step": 10365
+    },
+    {
+      "epoch": 1.3494728621632175,
+      "grad_norm": 2.512634038925171,
+      "learning_rate": 5.973529952628023e-06,
+      "loss": 0.4995,
+      "step": 10368
+    },
+    {
+      "epoch": 1.3498633346349083,
+      "grad_norm": 3.501483917236328,
+      "learning_rate": 5.97148001992696e-06,
+      "loss": 0.4722,
+      "step": 10371
+    },
+    {
+      "epoch": 1.350253807106599,
+      "grad_norm": 2.4441893100738525,
+      "learning_rate": 5.969429917510378e-06,
+      "loss": 0.5325,
+      "step": 10374
+    },
+    {
+      "epoch": 1.3506442795782898,
+      "grad_norm": 2.3490185737609863,
+      "learning_rate": 5.9673796457364295e-06,
+      "loss": 0.456,
+      "step": 10377
+    },
+    {
+      "epoch": 1.3510347520499804,
+      "grad_norm": 3.762827157974243,
+      "learning_rate": 5.965329204963292e-06,
+      "loss": 0.5797,
+      "step": 10380
+    },
+    {
+      "epoch": 1.3514252245216714,
+      "grad_norm": 3.0791232585906982,
+      "learning_rate": 5.9632785955491735e-06,
+      "loss": 0.4798,
+      "step": 10383
+    },
+    {
+      "epoch": 1.351815696993362,
+      "grad_norm": 2.7821664810180664,
+      "learning_rate": 5.961227817852311e-06,
+      "loss": 0.5121,
+      "step": 10386
+    },
+    {
+      "epoch": 1.3522061694650527,
+      "grad_norm": 2.4701087474823,
+      "learning_rate": 5.959176872230969e-06,
+      "loss": 0.4682,
+      "step": 10389
+    },
+    {
+      "epoch": 1.3525966419367434,
+      "grad_norm": 2.5216877460479736,
+      "learning_rate": 5.957125759043449e-06,
+      "loss": 0.4486,
+      "step": 10392
+    },
+    {
+      "epoch": 1.3529871144084342,
+      "grad_norm": 2.7570204734802246,
+      "learning_rate": 5.955074478648068e-06,
+      "loss": 0.5457,
+      "step": 10395
+    },
+    {
+      "epoch": 1.353377586880125,
+      "grad_norm": 2.721740484237671,
+      "learning_rate": 5.9530230314031875e-06,
+      "loss": 0.4864,
+      "step": 10398
+    },
+    {
+      "epoch": 1.3537680593518158,
+      "grad_norm": 2.7100112438201904,
+      "learning_rate": 5.950971417667189e-06,
+      "loss": 0.4506,
+      "step": 10401
+    },
+    {
+      "epoch": 1.3541585318235065,
+      "grad_norm": 3.179159641265869,
+      "learning_rate": 5.948919637798482e-06,
+      "loss": 0.5391,
+      "step": 10404
+    },
+    {
+      "epoch": 1.354549004295197,
+      "grad_norm": 2.8454229831695557,
+      "learning_rate": 5.946867692155511e-06,
+      "loss": 0.507,
+      "step": 10407
+    },
+    {
+      "epoch": 1.3549394767668879,
+      "grad_norm": 2.766425609588623,
+      "learning_rate": 5.944815581096746e-06,
+      "loss": 0.5161,
+      "step": 10410
+    },
+    {
+      "epoch": 1.3553299492385786,
+      "grad_norm": 2.6787941455841064,
+      "learning_rate": 5.942763304980689e-06,
+      "loss": 0.4643,
+      "step": 10413
+    },
+    {
+      "epoch": 1.3557204217102694,
+      "grad_norm": 2.679385185241699,
+      "learning_rate": 5.940710864165863e-06,
+      "loss": 0.5447,
+      "step": 10416
+    },
+    {
+      "epoch": 1.3561108941819602,
+      "grad_norm": 2.693063497543335,
+      "learning_rate": 5.938658259010829e-06,
+      "loss": 0.5279,
+      "step": 10419
+    },
+    {
+      "epoch": 1.356501366653651,
+      "grad_norm": 2.5846731662750244,
+      "learning_rate": 5.936605489874172e-06,
+      "loss": 0.5074,
+      "step": 10422
+    },
+    {
+      "epoch": 1.3568918391253417,
+      "grad_norm": 2.7756001949310303,
+      "learning_rate": 5.934552557114507e-06,
+      "loss": 0.4502,
+      "step": 10425
+    },
+    {
+      "epoch": 1.3572823115970323,
+      "grad_norm": 2.7776410579681396,
+      "learning_rate": 5.932499461090475e-06,
+      "loss": 0.5334,
+      "step": 10428
+    },
+    {
+      "epoch": 1.3576727840687233,
+      "grad_norm": 2.835615396499634,
+      "learning_rate": 5.930446202160749e-06,
+      "loss": 0.5619,
+      "step": 10431
+    },
+    {
+      "epoch": 1.3580632565404138,
+      "grad_norm": 2.890580177307129,
+      "learning_rate": 5.928392780684028e-06,
+      "loss": 0.561,
+      "step": 10434
+    },
+    {
+      "epoch": 1.3584537290121046,
+      "grad_norm": 4.082695960998535,
+      "learning_rate": 5.926339197019043e-06,
+      "loss": 0.5714,
+      "step": 10437
+    },
+    {
+      "epoch": 1.3588442014837954,
+      "grad_norm": 2.4185142517089844,
+      "learning_rate": 5.924285451524549e-06,
+      "loss": 0.5067,
+      "step": 10440
+    },
+    {
+      "epoch": 1.3592346739554861,
+      "grad_norm": 2.5827300548553467,
+      "learning_rate": 5.922231544559331e-06,
+      "loss": 0.4823,
+      "step": 10443
+    },
+    {
+      "epoch": 1.359625146427177,
+      "grad_norm": 2.5972402095794678,
+      "learning_rate": 5.9201774764822e-06,
+      "loss": 0.5133,
+      "step": 10446
+    },
+    {
+      "epoch": 1.3600156188988677,
+      "grad_norm": 2.655304431915283,
+      "learning_rate": 5.918123247652002e-06,
+      "loss": 0.4895,
+      "step": 10449
+    },
+    {
+      "epoch": 1.3604060913705585,
+      "grad_norm": 3.1130921840667725,
+      "learning_rate": 5.916068858427604e-06,
+      "loss": 0.5255,
+      "step": 10452
+    },
+    {
+      "epoch": 1.360796563842249,
+      "grad_norm": 2.6465952396392822,
+      "learning_rate": 5.914014309167901e-06,
+      "loss": 0.5161,
+      "step": 10455
+    },
+    {
+      "epoch": 1.36118703631394,
+      "grad_norm": 2.8651537895202637,
+      "learning_rate": 5.911959600231825e-06,
+      "loss": 0.4893,
+      "step": 10458
+    },
+    {
+      "epoch": 1.3615775087856306,
+      "grad_norm": 2.6356594562530518,
+      "learning_rate": 5.909904731978323e-06,
+      "loss": 0.4806,
+      "step": 10461
+    },
+    {
+      "epoch": 1.3619679812573213,
+      "grad_norm": 2.7509796619415283,
+      "learning_rate": 5.90784970476638e-06,
+      "loss": 0.4904,
+      "step": 10464
+    },
+    {
+      "epoch": 1.362358453729012,
+      "grad_norm": 2.8054521083831787,
+      "learning_rate": 5.905794518955002e-06,
+      "loss": 0.5197,
+      "step": 10467
+    },
+    {
+      "epoch": 1.3627489262007029,
+      "grad_norm": 2.970839262008667,
+      "learning_rate": 5.903739174903226e-06,
+      "loss": 0.4888,
+      "step": 10470
+    },
+    {
+      "epoch": 1.3631393986723936,
+      "grad_norm": 2.5401625633239746,
+      "learning_rate": 5.901683672970118e-06,
+      "loss": 0.6049,
+      "step": 10473
+    },
+    {
+      "epoch": 1.3635298711440844,
+      "grad_norm": 2.5715386867523193,
+      "learning_rate": 5.89962801351477e-06,
+      "loss": 0.4562,
+      "step": 10476
+    },
+    {
+      "epoch": 1.3639203436157752,
+      "grad_norm": 2.8467347621917725,
+      "learning_rate": 5.897572196896301e-06,
+      "loss": 0.4964,
+      "step": 10479
+    },
+    {
+      "epoch": 1.3643108160874657,
+      "grad_norm": 3.585697650909424,
+      "learning_rate": 5.895516223473856e-06,
+      "loss": 0.5151,
+      "step": 10482
+    },
+    {
+      "epoch": 1.3647012885591565,
+      "grad_norm": 2.776897668838501,
+      "learning_rate": 5.8934600936066115e-06,
+      "loss": 0.5529,
+      "step": 10485
+    },
+    {
+      "epoch": 1.3650917610308473,
+      "grad_norm": 2.829629421234131,
+      "learning_rate": 5.891403807653768e-06,
+      "loss": 0.5388,
+      "step": 10488
+    },
+    {
+      "epoch": 1.365482233502538,
+      "grad_norm": 2.7404351234436035,
+      "learning_rate": 5.889347365974554e-06,
+      "loss": 0.4966,
+      "step": 10491
+    },
+    {
+      "epoch": 1.3658727059742288,
+      "grad_norm": 2.4462528228759766,
+      "learning_rate": 5.887290768928228e-06,
+      "loss": 0.5372,
+      "step": 10494
+    },
+    {
+      "epoch": 1.3662631784459196,
+      "grad_norm": 2.7154576778411865,
+      "learning_rate": 5.88523401687407e-06,
+      "loss": 0.586,
+      "step": 10497
+    },
+    {
+      "epoch": 1.3666536509176104,
+      "grad_norm": 3.031604766845703,
+      "learning_rate": 5.883177110171392e-06,
+      "loss": 0.5067,
+      "step": 10500
+    },
+    {
+      "epoch": 1.367044123389301,
+      "grad_norm": 4.211544513702393,
+      "learning_rate": 5.881120049179529e-06,
+      "loss": 0.5154,
+      "step": 10503
+    },
+    {
+      "epoch": 1.367434595860992,
+      "grad_norm": 2.9091553688049316,
+      "learning_rate": 5.8790628342578485e-06,
+      "loss": 0.4859,
+      "step": 10506
+    },
+    {
+      "epoch": 1.3678250683326825,
+      "grad_norm": 2.4611270427703857,
+      "learning_rate": 5.87700546576574e-06,
+      "loss": 0.4909,
+      "step": 10509
+    },
+    {
+      "epoch": 1.3682155408043732,
+      "grad_norm": 2.56895112991333,
+      "learning_rate": 5.874947944062621e-06,
+      "loss": 0.491,
+      "step": 10512
+    },
+    {
+      "epoch": 1.368606013276064,
+      "grad_norm": 2.8200464248657227,
+      "learning_rate": 5.872890269507938e-06,
+      "loss": 0.4582,
+      "step": 10515
+    },
+    {
+      "epoch": 1.3689964857477548,
+      "grad_norm": 2.5796611309051514,
+      "learning_rate": 5.870832442461161e-06,
+      "loss": 0.432,
+      "step": 10518
+    },
+    {
+      "epoch": 1.3693869582194456,
+      "grad_norm": 2.872462034225464,
+      "learning_rate": 5.868774463281788e-06,
+      "loss": 0.5382,
+      "step": 10521
+    },
+    {
+      "epoch": 1.3697774306911363,
+      "grad_norm": 2.5397450923919678,
+      "learning_rate": 5.866716332329343e-06,
+      "loss": 0.4294,
+      "step": 10524
+    },
+    {
+      "epoch": 1.370167903162827,
+      "grad_norm": 2.746539831161499,
+      "learning_rate": 5.8646580499633786e-06,
+      "loss": 0.5046,
+      "step": 10527
+    },
+    {
+      "epoch": 1.3705583756345177,
+      "grad_norm": 2.5908896923065186,
+      "learning_rate": 5.862599616543473e-06,
+      "loss": 0.5438,
+      "step": 10530
+    },
+    {
+      "epoch": 1.3709488481062087,
+      "grad_norm": 3.0889134407043457,
+      "learning_rate": 5.860541032429227e-06,
+      "loss": 0.5459,
+      "step": 10533
+    },
+    {
+      "epoch": 1.3713393205778992,
+      "grad_norm": 2.655442953109741,
+      "learning_rate": 5.858482297980275e-06,
+      "loss": 0.4929,
+      "step": 10536
+    },
+    {
+      "epoch": 1.37172979304959,
+      "grad_norm": 2.756154775619507,
+      "learning_rate": 5.856423413556269e-06,
+      "loss": 0.4623,
+      "step": 10539
+    },
+    {
+      "epoch": 1.3721202655212807,
+      "grad_norm": 2.576265573501587,
+      "learning_rate": 5.854364379516896e-06,
+      "loss": 0.5525,
+      "step": 10542
+    },
+    {
+      "epoch": 1.3725107379929715,
+      "grad_norm": 2.8419313430786133,
+      "learning_rate": 5.852305196221864e-06,
+      "loss": 0.5179,
+      "step": 10545
+    },
+    {
+      "epoch": 1.3729012104646623,
+      "grad_norm": 2.7825493812561035,
+      "learning_rate": 5.8502458640309055e-06,
+      "loss": 0.4992,
+      "step": 10548
+    },
+    {
+      "epoch": 1.373291682936353,
+      "grad_norm": 2.590987205505371,
+      "learning_rate": 5.8481863833037846e-06,
+      "loss": 0.489,
+      "step": 10551
+    },
+    {
+      "epoch": 1.3736821554080438,
+      "grad_norm": 2.5935583114624023,
+      "learning_rate": 5.846126754400285e-06,
+      "loss": 0.5009,
+      "step": 10554
+    },
+    {
+      "epoch": 1.3740726278797344,
+      "grad_norm": 3.053290605545044,
+      "learning_rate": 5.844066977680223e-06,
+      "loss": 0.4684,
+      "step": 10557
+    },
+    {
+      "epoch": 1.3744631003514252,
+      "grad_norm": 3.1412622928619385,
+      "learning_rate": 5.842007053503436e-06,
+      "loss": 0.5036,
+      "step": 10560
+    },
+    {
+      "epoch": 1.374853572823116,
+      "grad_norm": 2.869248390197754,
+      "learning_rate": 5.839946982229786e-06,
+      "loss": 0.432,
+      "step": 10563
+    },
+    {
+      "epoch": 1.3752440452948067,
+      "grad_norm": 2.681100606918335,
+      "learning_rate": 5.8378867642191675e-06,
+      "loss": 0.5079,
+      "step": 10566
+    },
+    {
+      "epoch": 1.3756345177664975,
+      "grad_norm": 2.4988043308258057,
+      "learning_rate": 5.835826399831492e-06,
+      "loss": 0.4901,
+      "step": 10569
+    },
+    {
+      "epoch": 1.3760249902381883,
+      "grad_norm": 2.498563766479492,
+      "learning_rate": 5.833765889426706e-06,
+      "loss": 0.4638,
+      "step": 10572
+    },
+    {
+      "epoch": 1.376415462709879,
+      "grad_norm": 2.5112462043762207,
+      "learning_rate": 5.831705233364768e-06,
+      "loss": 0.4716,
+      "step": 10575
+    },
+    {
+      "epoch": 1.3768059351815696,
+      "grad_norm": 2.4168717861175537,
+      "learning_rate": 5.82964443200568e-06,
+      "loss": 0.476,
+      "step": 10578
+    },
+    {
+      "epoch": 1.3771964076532606,
+      "grad_norm": 2.5891149044036865,
+      "learning_rate": 5.827583485709453e-06,
+      "loss": 0.5093,
+      "step": 10581
+    },
+    {
+      "epoch": 1.3775868801249511,
+      "grad_norm": 2.5921523571014404,
+      "learning_rate": 5.825522394836132e-06,
+      "loss": 0.5532,
+      "step": 10584
+    },
+    {
+      "epoch": 1.377977352596642,
+      "grad_norm": 2.529782295227051,
+      "learning_rate": 5.823461159745786e-06,
+      "loss": 0.4955,
+      "step": 10587
+    },
+    {
+      "epoch": 1.3783678250683327,
+      "grad_norm": 2.8079707622528076,
+      "learning_rate": 5.821399780798507e-06,
+      "loss": 0.5489,
+      "step": 10590
+    },
+    {
+      "epoch": 1.3787582975400234,
+      "grad_norm": 3.1426773071289062,
+      "learning_rate": 5.8193382583544155e-06,
+      "loss": 0.478,
+      "step": 10593
+    },
+    {
+      "epoch": 1.3791487700117142,
+      "grad_norm": 2.480963706970215,
+      "learning_rate": 5.817276592773651e-06,
+      "loss": 0.4766,
+      "step": 10596
+    },
+    {
+      "epoch": 1.379539242483405,
+      "grad_norm": 2.4679062366485596,
+      "learning_rate": 5.815214784416386e-06,
+      "loss": 0.5077,
+      "step": 10599
+    },
+    {
+      "epoch": 1.3799297149550958,
+      "grad_norm": 2.7252042293548584,
+      "learning_rate": 5.813152833642816e-06,
+      "loss": 0.518,
+      "step": 10602
+    },
+    {
+      "epoch": 1.3803201874267863,
+      "grad_norm": 2.4545648097991943,
+      "learning_rate": 5.811090740813154e-06,
+      "loss": 0.4882,
+      "step": 10605
+    },
+    {
+      "epoch": 1.380710659898477,
+      "grad_norm": 2.6716935634613037,
+      "learning_rate": 5.809028506287647e-06,
+      "loss": 0.5061,
+      "step": 10608
+    },
+    {
+      "epoch": 1.3811011323701678,
+      "grad_norm": 2.5040860176086426,
+      "learning_rate": 5.806966130426561e-06,
+      "loss": 0.4739,
+      "step": 10611
+    },
+    {
+      "epoch": 1.3814916048418586,
+      "grad_norm": 2.5503172874450684,
+      "learning_rate": 5.80490361359019e-06,
+      "loss": 0.5359,
+      "step": 10614
+    },
+    {
+      "epoch": 1.3818820773135494,
+      "grad_norm": 2.5785112380981445,
+      "learning_rate": 5.802840956138851e-06,
+      "loss": 0.4595,
+      "step": 10617
+    },
+    {
+      "epoch": 1.3822725497852402,
+      "grad_norm": 3.1451265811920166,
+      "learning_rate": 5.800778158432886e-06,
+      "loss": 0.4806,
+      "step": 10620
+    },
+    {
+      "epoch": 1.382663022256931,
+      "grad_norm": 2.8773903846740723,
+      "learning_rate": 5.798715220832661e-06,
+      "loss": 0.5106,
+      "step": 10623
+    },
+    {
+      "epoch": 1.3830534947286217,
+      "grad_norm": 2.5292470455169678,
+      "learning_rate": 5.796652143698568e-06,
+      "loss": 0.4749,
+      "step": 10626
+    },
+    {
+      "epoch": 1.3834439672003125,
+      "grad_norm": 2.4957058429718018,
+      "learning_rate": 5.7945889273910215e-06,
+      "loss": 0.5352,
+      "step": 10629
+    },
+    {
+      "epoch": 1.383834439672003,
+      "grad_norm": 2.4294910430908203,
+      "learning_rate": 5.79252557227046e-06,
+      "loss": 0.5267,
+      "step": 10632
+    },
+    {
+      "epoch": 1.3842249121436938,
+      "grad_norm": 3.0004422664642334,
+      "learning_rate": 5.7904620786973476e-06,
+      "loss": 0.4788,
+      "step": 10635
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 2.777956962585449,
+      "learning_rate": 5.788398447032174e-06,
+      "loss": 0.5147,
+      "step": 10638
+    },
+    {
+      "epoch": 1.3850058570870754,
+      "grad_norm": 3.005110263824463,
+      "learning_rate": 5.78633467763545e-06,
+      "loss": 0.5815,
+      "step": 10641
+    },
+    {
+      "epoch": 1.3853963295587661,
+      "grad_norm": 2.7683937549591064,
+      "learning_rate": 5.784270770867713e-06,
+      "loss": 0.5067,
+      "step": 10644
+    },
+    {
+      "epoch": 1.385786802030457,
+      "grad_norm": 2.3799400329589844,
+      "learning_rate": 5.782206727089521e-06,
+      "loss": 0.4352,
+      "step": 10647
+    },
+    {
+      "epoch": 1.3861772745021477,
+      "grad_norm": 2.8313100337982178,
+      "learning_rate": 5.780142546661461e-06,
+      "loss": 0.4987,
+      "step": 10650
+    },
+    {
+      "epoch": 1.3865677469738382,
+      "grad_norm": 2.4416072368621826,
+      "learning_rate": 5.778078229944137e-06,
+      "loss": 0.5242,
+      "step": 10653
+    },
+    {
+      "epoch": 1.3869582194455292,
+      "grad_norm": 4.677280426025391,
+      "learning_rate": 5.776013777298184e-06,
+      "loss": 0.5669,
+      "step": 10656
+    },
+    {
+      "epoch": 1.3873486919172198,
+      "grad_norm": 2.549098014831543,
+      "learning_rate": 5.7739491890842556e-06,
+      "loss": 0.4694,
+      "step": 10659
+    },
+    {
+      "epoch": 1.3877391643889105,
+      "grad_norm": 2.7043087482452393,
+      "learning_rate": 5.771884465663033e-06,
+      "loss": 0.5064,
+      "step": 10662
+    },
+    {
+      "epoch": 1.3881296368606013,
+      "grad_norm": 2.652681350708008,
+      "learning_rate": 5.7698196073952195e-06,
+      "loss": 0.4974,
+      "step": 10665
+    },
+    {
+      "epoch": 1.388520109332292,
+      "grad_norm": 2.560001850128174,
+      "learning_rate": 5.767754614641538e-06,
+      "loss": 0.5313,
+      "step": 10668
+    },
+    {
+      "epoch": 1.3889105818039829,
+      "grad_norm": 2.911827325820923,
+      "learning_rate": 5.76568948776274e-06,
+      "loss": 0.5815,
+      "step": 10671
+    },
+    {
+      "epoch": 1.3893010542756736,
+      "grad_norm": 3.264103889465332,
+      "learning_rate": 5.763624227119602e-06,
+      "loss": 0.5138,
+      "step": 10674
+    },
+    {
+      "epoch": 1.3896915267473644,
+      "grad_norm": 2.7543458938598633,
+      "learning_rate": 5.761558833072915e-06,
+      "loss": 0.5643,
+      "step": 10677
+    },
+    {
+      "epoch": 1.390081999219055,
+      "grad_norm": 2.639909505844116,
+      "learning_rate": 5.759493305983504e-06,
+      "loss": 0.5443,
+      "step": 10680
+    },
+    {
+      "epoch": 1.3904724716907457,
+      "grad_norm": 3.0217959880828857,
+      "learning_rate": 5.757427646212208e-06,
+      "loss": 0.5419,
+      "step": 10683
+    },
+    {
+      "epoch": 1.3908629441624365,
+      "grad_norm": 2.6086668968200684,
+      "learning_rate": 5.755361854119898e-06,
+      "loss": 0.5186,
+      "step": 10686
+    },
+    {
+      "epoch": 1.3912534166341273,
+      "grad_norm": 2.5692412853240967,
+      "learning_rate": 5.753295930067461e-06,
+      "loss": 0.439,
+      "step": 10689
+    },
+    {
+      "epoch": 1.391643889105818,
+      "grad_norm": 2.9494736194610596,
+      "learning_rate": 5.751229874415808e-06,
+      "loss": 0.4788,
+      "step": 10692
+    },
+    {
+      "epoch": 1.3920343615775088,
+      "grad_norm": 4.0618896484375,
+      "learning_rate": 5.749163687525878e-06,
+      "loss": 0.4985,
+      "step": 10695
+    },
+    {
+      "epoch": 1.3924248340491996,
+      "grad_norm": 2.8040335178375244,
+      "learning_rate": 5.747097369758626e-06,
+      "loss": 0.5112,
+      "step": 10698
+    },
+    {
+      "epoch": 1.3928153065208901,
+      "grad_norm": 3.7752537727355957,
+      "learning_rate": 5.745030921475036e-06,
+      "loss": 0.5142,
+      "step": 10701
+    },
+    {
+      "epoch": 1.3932057789925811,
+      "grad_norm": 2.545715093612671,
+      "learning_rate": 5.742964343036111e-06,
+      "loss": 0.4986,
+      "step": 10704
+    },
+    {
+      "epoch": 1.3935962514642717,
+      "grad_norm": 3.0850815773010254,
+      "learning_rate": 5.74089763480288e-06,
+      "loss": 0.4749,
+      "step": 10707
+    },
+    {
+      "epoch": 1.3939867239359625,
+      "grad_norm": 2.574460029602051,
+      "learning_rate": 5.738830797136389e-06,
+      "loss": 0.5006,
+      "step": 10710
+    },
+    {
+      "epoch": 1.3943771964076532,
+      "grad_norm": 2.6434524059295654,
+      "learning_rate": 5.736763830397713e-06,
+      "loss": 0.4904,
+      "step": 10713
+    },
+    {
+      "epoch": 1.394767668879344,
+      "grad_norm": 2.6245415210723877,
+      "learning_rate": 5.734696734947946e-06,
+      "loss": 0.4879,
+      "step": 10716
+    },
+    {
+      "epoch": 1.3951581413510348,
+      "grad_norm": 2.7214295864105225,
+      "learning_rate": 5.732629511148204e-06,
+      "loss": 0.5052,
+      "step": 10719
+    },
+    {
+      "epoch": 1.3955486138227255,
+      "grad_norm": 2.8227062225341797,
+      "learning_rate": 5.730562159359628e-06,
+      "loss": 0.5524,
+      "step": 10722
+    },
+    {
+      "epoch": 1.3959390862944163,
+      "grad_norm": 4.676625728607178,
+      "learning_rate": 5.728494679943378e-06,
+      "loss": 0.5505,
+      "step": 10725
+    },
+    {
+      "epoch": 1.3963295587661069,
+      "grad_norm": 3.0134570598602295,
+      "learning_rate": 5.726427073260641e-06,
+      "loss": 0.4246,
+      "step": 10728
+    },
+    {
+      "epoch": 1.3967200312377979,
+      "grad_norm": 2.913252353668213,
+      "learning_rate": 5.7243593396726235e-06,
+      "loss": 0.5162,
+      "step": 10731
+    },
+    {
+      "epoch": 1.3971105037094884,
+      "grad_norm": 2.4458465576171875,
+      "learning_rate": 5.722291479540552e-06,
+      "loss": 0.496,
+      "step": 10734
+    },
+    {
+      "epoch": 1.3975009761811792,
+      "grad_norm": 2.5969252586364746,
+      "learning_rate": 5.72022349322568e-06,
+      "loss": 0.5335,
+      "step": 10737
+    },
+    {
+      "epoch": 1.39789144865287,
+      "grad_norm": 2.6486926078796387,
+      "learning_rate": 5.7181553810892785e-06,
+      "loss": 0.4731,
+      "step": 10740
+    },
+    {
+      "epoch": 1.3982819211245607,
+      "grad_norm": 2.3636739253997803,
+      "learning_rate": 5.716087143492643e-06,
+      "loss": 0.4643,
+      "step": 10743
+    },
+    {
+      "epoch": 1.3986723935962515,
+      "grad_norm": 2.5083768367767334,
+      "learning_rate": 5.714018780797088e-06,
+      "loss": 0.4983,
+      "step": 10746
+    },
+    {
+      "epoch": 1.3990628660679423,
+      "grad_norm": 2.5721659660339355,
+      "learning_rate": 5.7119502933639545e-06,
+      "loss": 0.5938,
+      "step": 10749
+    },
+    {
+      "epoch": 1.399453338539633,
+      "grad_norm": 2.918208360671997,
+      "learning_rate": 5.709881681554604e-06,
+      "loss": 0.4612,
+      "step": 10752
+    },
+    {
+      "epoch": 1.3998438110113236,
+      "grad_norm": 2.538756847381592,
+      "learning_rate": 5.7078129457304165e-06,
+      "loss": 0.4997,
+      "step": 10755
+    },
+    {
+      "epoch": 1.4002342834830144,
+      "grad_norm": 2.635209798812866,
+      "learning_rate": 5.7057440862527965e-06,
+      "loss": 0.4774,
+      "step": 10758
+    },
+    {
+      "epoch": 1.4006247559547051,
+      "grad_norm": 2.62276554107666,
+      "learning_rate": 5.70367510348317e-06,
+      "loss": 0.4467,
+      "step": 10761
+    },
+    {
+      "epoch": 1.401015228426396,
+      "grad_norm": 2.6300337314605713,
+      "learning_rate": 5.7016059977829805e-06,
+      "loss": 0.4499,
+      "step": 10764
+    },
+    {
+      "epoch": 1.4014057008980867,
+      "grad_norm": 2.4392457008361816,
+      "learning_rate": 5.6995367695137e-06,
+      "loss": 0.5159,
+      "step": 10767
+    },
+    {
+      "epoch": 1.4017961733697775,
+      "grad_norm": 2.7532896995544434,
+      "learning_rate": 5.697467419036819e-06,
+      "loss": 0.5695,
+      "step": 10770
+    },
+    {
+      "epoch": 1.4021866458414682,
+      "grad_norm": 3.3758609294891357,
+      "learning_rate": 5.695397946713848e-06,
+      "loss": 0.5062,
+      "step": 10773
+    },
+    {
+      "epoch": 1.4025771183131588,
+      "grad_norm": 2.9240944385528564,
+      "learning_rate": 5.6933283529063165e-06,
+      "loss": 0.4811,
+      "step": 10776
+    },
+    {
+      "epoch": 1.4029675907848498,
+      "grad_norm": 3.1382761001586914,
+      "learning_rate": 5.691258637975781e-06,
+      "loss": 0.506,
+      "step": 10779
+    },
+    {
+      "epoch": 1.4033580632565403,
+      "grad_norm": 2.694122314453125,
+      "learning_rate": 5.689188802283816e-06,
+      "loss": 0.5352,
+      "step": 10782
+    },
+    {
+      "epoch": 1.403748535728231,
+      "grad_norm": 2.632483720779419,
+      "learning_rate": 5.687118846192015e-06,
+      "loss": 0.4797,
+      "step": 10785
+    },
+    {
+      "epoch": 1.4041390081999219,
+      "grad_norm": 2.6875948905944824,
+      "learning_rate": 5.685048770062e-06,
+      "loss": 0.5539,
+      "step": 10788
+    },
+    {
+      "epoch": 1.4045294806716127,
+      "grad_norm": 2.9431047439575195,
+      "learning_rate": 5.682978574255404e-06,
+      "loss": 0.5159,
+      "step": 10791
+    },
+    {
+      "epoch": 1.4049199531433034,
+      "grad_norm": 2.7504560947418213,
+      "learning_rate": 5.680908259133889e-06,
+      "loss": 0.4741,
+      "step": 10794
+    },
+    {
+      "epoch": 1.4053104256149942,
+      "grad_norm": 2.588517189025879,
+      "learning_rate": 5.678837825059134e-06,
+      "loss": 0.5162,
+      "step": 10797
+    },
+    {
+      "epoch": 1.405700898086685,
+      "grad_norm": 2.9769070148468018,
+      "learning_rate": 5.676767272392837e-06,
+      "loss": 0.5516,
+      "step": 10800
+    },
+    {
+      "epoch": 1.4060913705583755,
+      "grad_norm": 3.6146578788757324,
+      "learning_rate": 5.674696601496723e-06,
+      "loss": 0.4143,
+      "step": 10803
+    },
+    {
+      "epoch": 1.4064818430300665,
+      "grad_norm": 2.6829850673675537,
+      "learning_rate": 5.672625812732531e-06,
+      "loss": 0.4618,
+      "step": 10806
+    },
+    {
+      "epoch": 1.406872315501757,
+      "grad_norm": 2.5348923206329346,
+      "learning_rate": 5.670554906462024e-06,
+      "loss": 0.4398,
+      "step": 10809
+    },
+    {
+      "epoch": 1.4072627879734478,
+      "grad_norm": 2.8809609413146973,
+      "learning_rate": 5.668483883046987e-06,
+      "loss": 0.438,
+      "step": 10812
+    },
+    {
+      "epoch": 1.4076532604451386,
+      "grad_norm": 2.6564998626708984,
+      "learning_rate": 5.666412742849222e-06,
+      "loss": 0.4728,
+      "step": 10815
+    },
+    {
+      "epoch": 1.4080437329168294,
+      "grad_norm": 2.5094847679138184,
+      "learning_rate": 5.664341486230552e-06,
+      "loss": 0.5797,
+      "step": 10818
+    },
+    {
+      "epoch": 1.4084342053885202,
+      "grad_norm": 4.3344268798828125,
+      "learning_rate": 5.662270113552822e-06,
+      "loss": 0.4779,
+      "step": 10821
+    },
+    {
+      "epoch": 1.408824677860211,
+      "grad_norm": 2.705709218978882,
+      "learning_rate": 5.660198625177897e-06,
+      "loss": 0.5035,
+      "step": 10824
+    },
+    {
+      "epoch": 1.4092151503319017,
+      "grad_norm": 2.7563042640686035,
+      "learning_rate": 5.65812702146766e-06,
+      "loss": 0.4518,
+      "step": 10827
+    },
+    {
+      "epoch": 1.4096056228035923,
+      "grad_norm": 2.515812635421753,
+      "learning_rate": 5.656055302784017e-06,
+      "loss": 0.4553,
+      "step": 10830
+    },
+    {
+      "epoch": 1.409996095275283,
+      "grad_norm": 2.5328338146209717,
+      "learning_rate": 5.653983469488893e-06,
+      "loss": 0.467,
+      "step": 10833
+    },
+    {
+      "epoch": 1.4103865677469738,
+      "grad_norm": 2.820492744445801,
+      "learning_rate": 5.651911521944233e-06,
+      "loss": 0.5257,
+      "step": 10836
+    },
+    {
+      "epoch": 1.4107770402186646,
+      "grad_norm": 3.2355074882507324,
+      "learning_rate": 5.649839460512001e-06,
+      "loss": 0.4985,
+      "step": 10839
+    },
+    {
+      "epoch": 1.4111675126903553,
+      "grad_norm": 3.2596559524536133,
+      "learning_rate": 5.647767285554183e-06,
+      "loss": 0.5677,
+      "step": 10842
+    },
+    {
+      "epoch": 1.4115579851620461,
+      "grad_norm": 2.754706382751465,
+      "learning_rate": 5.645694997432783e-06,
+      "loss": 0.4839,
+      "step": 10845
+    },
+    {
+      "epoch": 1.4119484576337369,
+      "grad_norm": 2.3893485069274902,
+      "learning_rate": 5.643622596509823e-06,
+      "loss": 0.4494,
+      "step": 10848
+    },
+    {
+      "epoch": 1.4123389301054274,
+      "grad_norm": 2.934878349304199,
+      "learning_rate": 5.641550083147352e-06,
+      "loss": 0.4586,
+      "step": 10851
+    },
+    {
+      "epoch": 1.4127294025771184,
+      "grad_norm": 2.4555704593658447,
+      "learning_rate": 5.639477457707431e-06,
+      "loss": 0.5323,
+      "step": 10854
+    },
+    {
+      "epoch": 1.413119875048809,
+      "grad_norm": 2.666161298751831,
+      "learning_rate": 5.6374047205521424e-06,
+      "loss": 0.4899,
+      "step": 10857
+    },
+    {
+      "epoch": 1.4135103475204998,
+      "grad_norm": 2.8357090950012207,
+      "learning_rate": 5.6353318720435905e-06,
+      "loss": 0.5637,
+      "step": 10860
+    },
+    {
+      "epoch": 1.4139008199921905,
+      "grad_norm": 2.9021785259246826,
+      "learning_rate": 5.633258912543897e-06,
+      "loss": 0.5518,
+      "step": 10863
+    },
+    {
+      "epoch": 1.4142912924638813,
+      "grad_norm": 2.5638427734375,
+      "learning_rate": 5.631185842415203e-06,
+      "loss": 0.448,
+      "step": 10866
+    },
+    {
+      "epoch": 1.414681764935572,
+      "grad_norm": 2.585672378540039,
+      "learning_rate": 5.62911266201967e-06,
+      "loss": 0.5891,
+      "step": 10869
+    },
+    {
+      "epoch": 1.4150722374072628,
+      "grad_norm": 2.641378164291382,
+      "learning_rate": 5.62703937171948e-06,
+      "loss": 0.4652,
+      "step": 10872
+    },
+    {
+      "epoch": 1.4154627098789536,
+      "grad_norm": 2.626474618911743,
+      "learning_rate": 5.62496597187683e-06,
+      "loss": 0.5193,
+      "step": 10875
+    },
+    {
+      "epoch": 1.4158531823506442,
+      "grad_norm": 2.8795952796936035,
+      "learning_rate": 5.622892462853938e-06,
+      "loss": 0.4486,
+      "step": 10878
+    },
+    {
+      "epoch": 1.4162436548223352,
+      "grad_norm": 3.0229554176330566,
+      "learning_rate": 5.620818845013046e-06,
+      "loss": 0.501,
+      "step": 10881
+    },
+    {
+      "epoch": 1.4166341272940257,
+      "grad_norm": 2.5767412185668945,
+      "learning_rate": 5.618745118716406e-06,
+      "loss": 0.462,
+      "step": 10884
+    },
+    {
+      "epoch": 1.4170245997657165,
+      "grad_norm": 2.8493659496307373,
+      "learning_rate": 5.616671284326296e-06,
+      "loss": 0.5408,
+      "step": 10887
+    },
+    {
+      "epoch": 1.4174150722374073,
+      "grad_norm": 2.6280994415283203,
+      "learning_rate": 5.6145973422050085e-06,
+      "loss": 0.5258,
+      "step": 10890
+    },
+    {
+      "epoch": 1.417805544709098,
+      "grad_norm": 2.548902988433838,
+      "learning_rate": 5.61252329271486e-06,
+      "loss": 0.4983,
+      "step": 10893
+    },
+    {
+      "epoch": 1.4181960171807888,
+      "grad_norm": 2.673133373260498,
+      "learning_rate": 5.6104491362181805e-06,
+      "loss": 0.6353,
+      "step": 10896
+    },
+    {
+      "epoch": 1.4185864896524796,
+      "grad_norm": 2.571382999420166,
+      "learning_rate": 5.60837487307732e-06,
+      "loss": 0.4516,
+      "step": 10899
+    },
+    {
+      "epoch": 1.4189769621241703,
+      "grad_norm": 3.19085693359375,
+      "learning_rate": 5.60630050365465e-06,
+      "loss": 0.5573,
+      "step": 10902
+    },
+    {
+      "epoch": 1.419367434595861,
+      "grad_norm": 2.6122026443481445,
+      "learning_rate": 5.604226028312558e-06,
+      "loss": 0.534,
+      "step": 10905
+    },
+    {
+      "epoch": 1.4197579070675517,
+      "grad_norm": 4.08894157409668,
+      "learning_rate": 5.602151447413449e-06,
+      "loss": 0.5676,
+      "step": 10908
+    },
+    {
+      "epoch": 1.4201483795392424,
+      "grad_norm": 2.857332229614258,
+      "learning_rate": 5.600076761319748e-06,
+      "loss": 0.5727,
+      "step": 10911
+    },
+    {
+      "epoch": 1.4205388520109332,
+      "grad_norm": 5.386052131652832,
+      "learning_rate": 5.5980019703939006e-06,
+      "loss": 0.4778,
+      "step": 10914
+    },
+    {
+      "epoch": 1.420929324482624,
+      "grad_norm": 2.5507993698120117,
+      "learning_rate": 5.595927074998367e-06,
+      "loss": 0.5343,
+      "step": 10917
+    },
+    {
+      "epoch": 1.4213197969543148,
+      "grad_norm": 3.1667394638061523,
+      "learning_rate": 5.593852075495627e-06,
+      "loss": 0.4644,
+      "step": 10920
+    },
+    {
+      "epoch": 1.4217102694260055,
+      "grad_norm": 2.6564674377441406,
+      "learning_rate": 5.591776972248178e-06,
+      "loss": 0.465,
+      "step": 10923
+    },
+    {
+      "epoch": 1.422100741897696,
+      "grad_norm": 2.5531013011932373,
+      "learning_rate": 5.589701765618539e-06,
+      "loss": 0.4859,
+      "step": 10926
+    },
+    {
+      "epoch": 1.422491214369387,
+      "grad_norm": 2.634110689163208,
+      "learning_rate": 5.587626455969238e-06,
+      "loss": 0.5084,
+      "step": 10929
+    },
+    {
+      "epoch": 1.4228816868410776,
+      "grad_norm": 2.458155632019043,
+      "learning_rate": 5.5855510436628345e-06,
+      "loss": 0.4223,
+      "step": 10932
+    },
+    {
+      "epoch": 1.4232721593127684,
+      "grad_norm": 2.6002464294433594,
+      "learning_rate": 5.583475529061895e-06,
+      "loss": 0.607,
+      "step": 10935
+    },
+    {
+      "epoch": 1.4236626317844592,
+      "grad_norm": 2.449547290802002,
+      "learning_rate": 5.581399912529008e-06,
+      "loss": 0.4332,
+      "step": 10938
+    },
+    {
+      "epoch": 1.42405310425615,
+      "grad_norm": 2.828279972076416,
+      "learning_rate": 5.579324194426779e-06,
+      "loss": 0.5483,
+      "step": 10941
+    },
+    {
+      "epoch": 1.4244435767278407,
+      "grad_norm": 2.5671331882476807,
+      "learning_rate": 5.577248375117832e-06,
+      "loss": 0.5026,
+      "step": 10944
+    },
+    {
+      "epoch": 1.4248340491995315,
+      "grad_norm": 2.9858782291412354,
+      "learning_rate": 5.575172454964808e-06,
+      "loss": 0.5081,
+      "step": 10947
+    },
+    {
+      "epoch": 1.4252245216712223,
+      "grad_norm": 2.493577003479004,
+      "learning_rate": 5.573096434330366e-06,
+      "loss": 0.4517,
+      "step": 10950
+    },
+    {
+      "epoch": 1.4256149941429128,
+      "grad_norm": 2.619349718093872,
+      "learning_rate": 5.571020313577183e-06,
+      "loss": 0.5067,
+      "step": 10953
+    },
+    {
+      "epoch": 1.4260054666146036,
+      "grad_norm": 2.6694607734680176,
+      "learning_rate": 5.5689440930679514e-06,
+      "loss": 0.5137,
+      "step": 10956
+    },
+    {
+      "epoch": 1.4263959390862944,
+      "grad_norm": 2.9713809490203857,
+      "learning_rate": 5.566867773165386e-06,
+      "loss": 0.5501,
+      "step": 10959
+    },
+    {
+      "epoch": 1.4267864115579851,
+      "grad_norm": 2.9507107734680176,
+      "learning_rate": 5.564791354232211e-06,
+      "loss": 0.5056,
+      "step": 10962
+    },
+    {
+      "epoch": 1.427176884029676,
+      "grad_norm": 3.161423683166504,
+      "learning_rate": 5.562714836631175e-06,
+      "loss": 0.461,
+      "step": 10965
+    },
+    {
+      "epoch": 1.4275673565013667,
+      "grad_norm": 2.630687713623047,
+      "learning_rate": 5.560638220725042e-06,
+      "loss": 0.4621,
+      "step": 10968
+    },
+    {
+      "epoch": 1.4279578289730575,
+      "grad_norm": 2.7884042263031006,
+      "learning_rate": 5.55856150687659e-06,
+      "loss": 0.5301,
+      "step": 10971
+    },
+    {
+      "epoch": 1.4283483014447482,
+      "grad_norm": 2.516737222671509,
+      "learning_rate": 5.5564846954486184e-06,
+      "loss": 0.4985,
+      "step": 10974
+    },
+    {
+      "epoch": 1.428738773916439,
+      "grad_norm": 2.5410304069519043,
+      "learning_rate": 5.55440778680394e-06,
+      "loss": 0.3917,
+      "step": 10977
+    },
+    {
+      "epoch": 1.4291292463881295,
+      "grad_norm": 2.4755969047546387,
+      "learning_rate": 5.552330781305389e-06,
+      "loss": 0.4808,
+      "step": 10980
+    },
+    {
+      "epoch": 1.4295197188598203,
+      "grad_norm": 3.0844016075134277,
+      "learning_rate": 5.550253679315812e-06,
+      "loss": 0.4758,
+      "step": 10983
+    },
+    {
+      "epoch": 1.429910191331511,
+      "grad_norm": 2.555314779281616,
+      "learning_rate": 5.548176481198075e-06,
+      "loss": 0.4533,
+      "step": 10986
+    },
+    {
+      "epoch": 1.4303006638032019,
+      "grad_norm": 3.199594736099243,
+      "learning_rate": 5.5460991873150605e-06,
+      "loss": 0.5575,
+      "step": 10989
+    },
+    {
+      "epoch": 1.4306911362748926,
+      "grad_norm": 2.5074615478515625,
+      "learning_rate": 5.544021798029665e-06,
+      "loss": 0.4556,
+      "step": 10992
+    },
+    {
+      "epoch": 1.4310816087465834,
+      "grad_norm": 2.9451918601989746,
+      "learning_rate": 5.541944313704807e-06,
+      "loss": 0.5797,
+      "step": 10995
+    },
+    {
+      "epoch": 1.4314720812182742,
+      "grad_norm": 2.5799622535705566,
+      "learning_rate": 5.539866734703416e-06,
+      "loss": 0.4921,
+      "step": 10998
+    },
+    {
+      "epoch": 1.4318625536899647,
+      "grad_norm": 2.503661870956421,
+      "learning_rate": 5.537789061388445e-06,
+      "loss": 0.5223,
+      "step": 11001
+    },
+    {
+      "epoch": 1.4322530261616557,
+      "grad_norm": 2.756873846054077,
+      "learning_rate": 5.535711294122854e-06,
+      "loss": 0.5098,
+      "step": 11004
+    },
+    {
+      "epoch": 1.4326434986333463,
+      "grad_norm": 2.3765101432800293,
+      "learning_rate": 5.533633433269627e-06,
+      "loss": 0.5223,
+      "step": 11007
+    },
+    {
+      "epoch": 1.433033971105037,
+      "grad_norm": 2.8661093711853027,
+      "learning_rate": 5.531555479191764e-06,
+      "loss": 0.5067,
+      "step": 11010
+    },
+    {
+      "epoch": 1.4334244435767278,
+      "grad_norm": 2.547419309616089,
+      "learning_rate": 5.529477432252275e-06,
+      "loss": 0.4815,
+      "step": 11013
+    },
+    {
+      "epoch": 1.4338149160484186,
+      "grad_norm": 2.4830684661865234,
+      "learning_rate": 5.527399292814193e-06,
+      "loss": 0.501,
+      "step": 11016
+    },
+    {
+      "epoch": 1.4342053885201094,
+      "grad_norm": 2.8555281162261963,
+      "learning_rate": 5.525321061240563e-06,
+      "loss": 0.4946,
+      "step": 11019
+    },
+    {
+      "epoch": 1.4345958609918001,
+      "grad_norm": 2.6124064922332764,
+      "learning_rate": 5.523242737894451e-06,
+      "loss": 0.4766,
+      "step": 11022
+    },
+    {
+      "epoch": 1.434986333463491,
+      "grad_norm": 2.4766440391540527,
+      "learning_rate": 5.521164323138931e-06,
+      "loss": 0.4742,
+      "step": 11025
+    },
+    {
+      "epoch": 1.4353768059351815,
+      "grad_norm": 2.639648199081421,
+      "learning_rate": 5.519085817337101e-06,
+      "loss": 0.4725,
+      "step": 11028
+    },
+    {
+      "epoch": 1.4357672784068722,
+      "grad_norm": 2.593116521835327,
+      "learning_rate": 5.517007220852072e-06,
+      "loss": 0.4893,
+      "step": 11031
+    },
+    {
+      "epoch": 1.436157750878563,
+      "grad_norm": 2.823061466217041,
+      "learning_rate": 5.514928534046968e-06,
+      "loss": 0.5143,
+      "step": 11034
+    },
+    {
+      "epoch": 1.4365482233502538,
+      "grad_norm": 2.7042064666748047,
+      "learning_rate": 5.512849757284932e-06,
+      "loss": 0.5257,
+      "step": 11037
+    },
+    {
+      "epoch": 1.4369386958219446,
+      "grad_norm": 2.685746431350708,
+      "learning_rate": 5.510770890929122e-06,
+      "loss": 0.5436,
+      "step": 11040
+    },
+    {
+      "epoch": 1.4373291682936353,
+      "grad_norm": 2.78102970123291,
+      "learning_rate": 5.5086919353427124e-06,
+      "loss": 0.5199,
+      "step": 11043
+    },
+    {
+      "epoch": 1.437719640765326,
+      "grad_norm": 2.8201215267181396,
+      "learning_rate": 5.506612890888892e-06,
+      "loss": 0.6273,
+      "step": 11046
+    },
+    {
+      "epoch": 1.4381101132370167,
+      "grad_norm": 2.8764889240264893,
+      "learning_rate": 5.5045337579308654e-06,
+      "loss": 0.4641,
+      "step": 11049
+    },
+    {
+      "epoch": 1.4385005857087076,
+      "grad_norm": 4.958584308624268,
+      "learning_rate": 5.502454536831854e-06,
+      "loss": 0.484,
+      "step": 11052
+    },
+    {
+      "epoch": 1.4388910581803982,
+      "grad_norm": 2.74894380569458,
+      "learning_rate": 5.5003752279550905e-06,
+      "loss": 0.5001,
+      "step": 11055
+    },
+    {
+      "epoch": 1.439281530652089,
+      "grad_norm": 3.1531951427459717,
+      "learning_rate": 5.498295831663827e-06,
+      "loss": 0.4416,
+      "step": 11058
+    },
+    {
+      "epoch": 1.4396720031237797,
+      "grad_norm": 2.578901767730713,
+      "learning_rate": 5.496216348321329e-06,
+      "loss": 0.4324,
+      "step": 11061
+    },
+    {
+      "epoch": 1.4400624755954705,
+      "grad_norm": 2.5892860889434814,
+      "learning_rate": 5.49413677829088e-06,
+      "loss": 0.4925,
+      "step": 11064
+    },
+    {
+      "epoch": 1.4404529480671613,
+      "grad_norm": 2.979255199432373,
+      "learning_rate": 5.492057121935777e-06,
+      "loss": 0.5162,
+      "step": 11067
+    },
+    {
+      "epoch": 1.440843420538852,
+      "grad_norm": 2.781018018722534,
+      "learning_rate": 5.489977379619328e-06,
+      "loss": 0.5983,
+      "step": 11070
+    },
+    {
+      "epoch": 1.4412338930105428,
+      "grad_norm": 2.6435060501098633,
+      "learning_rate": 5.487897551704862e-06,
+      "loss": 0.4696,
+      "step": 11073
+    },
+    {
+      "epoch": 1.4416243654822334,
+      "grad_norm": 2.8207993507385254,
+      "learning_rate": 5.48581763855572e-06,
+      "loss": 0.6021,
+      "step": 11076
+    },
+    {
+      "epoch": 1.4420148379539244,
+      "grad_norm": 2.7664260864257812,
+      "learning_rate": 5.4837376405352595e-06,
+      "loss": 0.5321,
+      "step": 11079
+    },
+    {
+      "epoch": 1.442405310425615,
+      "grad_norm": 2.488525390625,
+      "learning_rate": 5.481657558006849e-06,
+      "loss": 0.4147,
+      "step": 11082
+    },
+    {
+      "epoch": 1.4427957828973057,
+      "grad_norm": 2.4613826274871826,
+      "learning_rate": 5.4795773913338765e-06,
+      "loss": 0.5012,
+      "step": 11085
+    },
+    {
+      "epoch": 1.4431862553689965,
+      "grad_norm": 3.2609658241271973,
+      "learning_rate": 5.477497140879745e-06,
+      "loss": 0.5489,
+      "step": 11088
+    },
+    {
+      "epoch": 1.4435767278406872,
+      "grad_norm": 2.521787405014038,
+      "learning_rate": 5.475416807007866e-06,
+      "loss": 0.5193,
+      "step": 11091
+    },
+    {
+      "epoch": 1.443967200312378,
+      "grad_norm": 3.109292984008789,
+      "learning_rate": 5.473336390081671e-06,
+      "loss": 0.515,
+      "step": 11094
+    },
+    {
+      "epoch": 1.4443576727840688,
+      "grad_norm": 3.495357036590576,
+      "learning_rate": 5.471255890464604e-06,
+      "loss": 0.5132,
+      "step": 11097
+    },
+    {
+      "epoch": 1.4447481452557596,
+      "grad_norm": 3.0955452919006348,
+      "learning_rate": 5.469175308520124e-06,
+      "loss": 0.4092,
+      "step": 11100
+    },
+    {
+      "epoch": 1.4451386177274501,
+      "grad_norm": 2.3907532691955566,
+      "learning_rate": 5.467094644611705e-06,
+      "loss": 0.4515,
+      "step": 11103
+    },
+    {
+      "epoch": 1.4455290901991409,
+      "grad_norm": 2.64176869392395,
+      "learning_rate": 5.465013899102836e-06,
+      "loss": 0.5404,
+      "step": 11106
+    },
+    {
+      "epoch": 1.4459195626708317,
+      "grad_norm": 2.4482946395874023,
+      "learning_rate": 5.4629330723570154e-06,
+      "loss": 0.4778,
+      "step": 11109
+    },
+    {
+      "epoch": 1.4463100351425224,
+      "grad_norm": 2.8428847789764404,
+      "learning_rate": 5.460852164737761e-06,
+      "loss": 0.5936,
+      "step": 11112
+    },
+    {
+      "epoch": 1.4467005076142132,
+      "grad_norm": 2.7340033054351807,
+      "learning_rate": 5.458771176608602e-06,
+      "loss": 0.497,
+      "step": 11115
+    },
+    {
+      "epoch": 1.447090980085904,
+      "grad_norm": 2.5455846786499023,
+      "learning_rate": 5.456690108333086e-06,
+      "loss": 0.452,
+      "step": 11118
+    },
+    {
+      "epoch": 1.4474814525575947,
+      "grad_norm": 2.627605676651001,
+      "learning_rate": 5.454608960274765e-06,
+      "loss": 0.5079,
+      "step": 11121
+    },
+    {
+      "epoch": 1.4478719250292853,
+      "grad_norm": 2.4849627017974854,
+      "learning_rate": 5.452527732797219e-06,
+      "loss": 0.46,
+      "step": 11124
+    },
+    {
+      "epoch": 1.4482623975009763,
+      "grad_norm": 3.0936851501464844,
+      "learning_rate": 5.450446426264028e-06,
+      "loss": 0.5866,
+      "step": 11127
+    },
+    {
+      "epoch": 1.4486528699726668,
+      "grad_norm": 2.7293624877929688,
+      "learning_rate": 5.448365041038796e-06,
+      "loss": 0.5474,
+      "step": 11130
+    },
+    {
+      "epoch": 1.4490433424443576,
+      "grad_norm": 2.670182704925537,
+      "learning_rate": 5.446283577485132e-06,
+      "loss": 0.5431,
+      "step": 11133
+    },
+    {
+      "epoch": 1.4494338149160484,
+      "grad_norm": 2.502770185470581,
+      "learning_rate": 5.4442020359666655e-06,
+      "loss": 0.4375,
+      "step": 11136
+    },
+    {
+      "epoch": 1.4498242873877392,
+      "grad_norm": 2.5619125366210938,
+      "learning_rate": 5.442120416847041e-06,
+      "loss": 0.533,
+      "step": 11139
+    },
+    {
+      "epoch": 1.45021475985943,
+      "grad_norm": 2.700932025909424,
+      "learning_rate": 5.440038720489906e-06,
+      "loss": 0.4754,
+      "step": 11142
+    },
+    {
+      "epoch": 1.4506052323311207,
+      "grad_norm": 2.834325075149536,
+      "learning_rate": 5.437956947258935e-06,
+      "loss": 0.4661,
+      "step": 11145
+    },
+    {
+      "epoch": 1.4509957048028115,
+      "grad_norm": 2.6361968517303467,
+      "learning_rate": 5.435875097517805e-06,
+      "loss": 0.5139,
+      "step": 11148
+    },
+    {
+      "epoch": 1.451386177274502,
+      "grad_norm": 2.8682637214660645,
+      "learning_rate": 5.433793171630213e-06,
+      "loss": 0.4919,
+      "step": 11151
+    },
+    {
+      "epoch": 1.451776649746193,
+      "grad_norm": 2.7310025691986084,
+      "learning_rate": 5.431711169959866e-06,
+      "loss": 0.4384,
+      "step": 11154
+    },
+    {
+      "epoch": 1.4521671222178836,
+      "grad_norm": 3.184248208999634,
+      "learning_rate": 5.429629092870488e-06,
+      "loss": 0.4457,
+      "step": 11157
+    },
+    {
+      "epoch": 1.4525575946895743,
+      "grad_norm": 2.8572938442230225,
+      "learning_rate": 5.4275469407258096e-06,
+      "loss": 0.5435,
+      "step": 11160
+    },
+    {
+      "epoch": 1.4529480671612651,
+      "grad_norm": 2.845240354537964,
+      "learning_rate": 5.425464713889579e-06,
+      "loss": 0.5269,
+      "step": 11163
+    },
+    {
+      "epoch": 1.453338539632956,
+      "grad_norm": 2.744442939758301,
+      "learning_rate": 5.42338241272556e-06,
+      "loss": 0.5117,
+      "step": 11166
+    },
+    {
+      "epoch": 1.4537290121046467,
+      "grad_norm": 3.74177622795105,
+      "learning_rate": 5.4213000375975226e-06,
+      "loss": 0.4822,
+      "step": 11169
+    },
+    {
+      "epoch": 1.4541194845763374,
+      "grad_norm": 2.628488779067993,
+      "learning_rate": 5.419217588869255e-06,
+      "loss": 0.4378,
+      "step": 11172
+    },
+    {
+      "epoch": 1.4545099570480282,
+      "grad_norm": 2.955030918121338,
+      "learning_rate": 5.4171350669045585e-06,
+      "loss": 0.536,
+      "step": 11175
+    },
+    {
+      "epoch": 1.4549004295197188,
+      "grad_norm": 2.7912299633026123,
+      "learning_rate": 5.415052472067241e-06,
+      "loss": 0.5115,
+      "step": 11178
+    },
+    {
+      "epoch": 1.4552909019914095,
+      "grad_norm": 2.868971824645996,
+      "learning_rate": 5.412969804721132e-06,
+      "loss": 0.4951,
+      "step": 11181
+    },
+    {
+      "epoch": 1.4556813744631003,
+      "grad_norm": 2.6621592044830322,
+      "learning_rate": 5.410887065230064e-06,
+      "loss": 0.5373,
+      "step": 11184
+    },
+    {
+      "epoch": 1.456071846934791,
+      "grad_norm": 2.808011770248413,
+      "learning_rate": 5.4088042539578925e-06,
+      "loss": 0.5717,
+      "step": 11187
+    },
+    {
+      "epoch": 1.4564623194064819,
+      "grad_norm": 2.3780086040496826,
+      "learning_rate": 5.406721371268476e-06,
+      "loss": 0.4945,
+      "step": 11190
+    },
+    {
+      "epoch": 1.4568527918781726,
+      "grad_norm": 2.6770284175872803,
+      "learning_rate": 5.404638417525693e-06,
+      "loss": 0.4986,
+      "step": 11193
+    },
+    {
+      "epoch": 1.4572432643498634,
+      "grad_norm": 2.5734894275665283,
+      "learning_rate": 5.4025553930934295e-06,
+      "loss": 0.4697,
+      "step": 11196
+    },
+    {
+      "epoch": 1.457633736821554,
+      "grad_norm": 2.9276740550994873,
+      "learning_rate": 5.4004722983355854e-06,
+      "loss": 0.5275,
+      "step": 11199
+    },
+    {
+      "epoch": 1.458024209293245,
+      "grad_norm": 2.488164186477661,
+      "learning_rate": 5.398389133616074e-06,
+      "loss": 0.4691,
+      "step": 11202
+    },
+    {
+      "epoch": 1.4584146817649355,
+      "grad_norm": 2.68459153175354,
+      "learning_rate": 5.396305899298817e-06,
+      "loss": 0.545,
+      "step": 11205
+    },
+    {
+      "epoch": 1.4588051542366263,
+      "grad_norm": 2.5225677490234375,
+      "learning_rate": 5.394222595747755e-06,
+      "loss": 0.4992,
+      "step": 11208
+    },
+    {
+      "epoch": 1.459195626708317,
+      "grad_norm": 2.5488972663879395,
+      "learning_rate": 5.3921392233268345e-06,
+      "loss": 0.3998,
+      "step": 11211
+    },
+    {
+      "epoch": 1.4595860991800078,
+      "grad_norm": 2.9433960914611816,
+      "learning_rate": 5.390055782400016e-06,
+      "loss": 0.5108,
+      "step": 11214
+    },
+    {
+      "epoch": 1.4599765716516986,
+      "grad_norm": 2.6426749229431152,
+      "learning_rate": 5.387972273331273e-06,
+      "loss": 0.4687,
+      "step": 11217
+    },
+    {
+      "epoch": 1.4603670441233894,
+      "grad_norm": 2.4266717433929443,
+      "learning_rate": 5.38588869648459e-06,
+      "loss": 0.4355,
+      "step": 11220
+    },
+    {
+      "epoch": 1.4607575165950801,
+      "grad_norm": 2.7653090953826904,
+      "learning_rate": 5.383805052223964e-06,
+      "loss": 0.5161,
+      "step": 11223
+    },
+    {
+      "epoch": 1.4611479890667707,
+      "grad_norm": 2.5222866535186768,
+      "learning_rate": 5.381721340913403e-06,
+      "loss": 0.5055,
+      "step": 11226
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 2.504310369491577,
+      "learning_rate": 5.379637562916925e-06,
+      "loss": 0.5179,
+      "step": 11229
+    },
+    {
+      "epoch": 1.4619289340101522,
+      "grad_norm": 3.081963300704956,
+      "learning_rate": 5.377553718598566e-06,
+      "loss": 0.5027,
+      "step": 11232
+    },
+    {
+      "epoch": 1.462319406481843,
+      "grad_norm": 2.8216118812561035,
+      "learning_rate": 5.375469808322364e-06,
+      "loss": 0.5212,
+      "step": 11235
+    },
+    {
+      "epoch": 1.4627098789535338,
+      "grad_norm": 2.547189474105835,
+      "learning_rate": 5.3733858324523795e-06,
+      "loss": 0.4809,
+      "step": 11238
+    },
+    {
+      "epoch": 1.4631003514252245,
+      "grad_norm": 2.744575023651123,
+      "learning_rate": 5.371301791352673e-06,
+      "loss": 0.4945,
+      "step": 11241
+    },
+    {
+      "epoch": 1.4634908238969153,
+      "grad_norm": 2.8713624477386475,
+      "learning_rate": 5.369217685387326e-06,
+      "loss": 0.5239,
+      "step": 11244
+    },
+    {
+      "epoch": 1.463881296368606,
+      "grad_norm": 2.840270519256592,
+      "learning_rate": 5.367133514920425e-06,
+      "loss": 0.4806,
+      "step": 11247
+    },
+    {
+      "epoch": 1.4642717688402969,
+      "grad_norm": 3.1532773971557617,
+      "learning_rate": 5.3650492803160715e-06,
+      "loss": 0.5324,
+      "step": 11250
+    },
+    {
+      "epoch": 1.4646622413119874,
+      "grad_norm": 2.547405958175659,
+      "learning_rate": 5.362964981938378e-06,
+      "loss": 0.5092,
+      "step": 11253
+    },
+    {
+      "epoch": 1.4650527137836782,
+      "grad_norm": 2.8435122966766357,
+      "learning_rate": 5.360880620151464e-06,
+      "loss": 0.5244,
+      "step": 11256
+    },
+    {
+      "epoch": 1.465443186255369,
+      "grad_norm": 2.7665226459503174,
+      "learning_rate": 5.358796195319467e-06,
+      "loss": 0.5135,
+      "step": 11259
+    },
+    {
+      "epoch": 1.4658336587270597,
+      "grad_norm": 2.7496416568756104,
+      "learning_rate": 5.356711707806527e-06,
+      "loss": 0.5243,
+      "step": 11262
+    },
+    {
+      "epoch": 1.4662241311987505,
+      "grad_norm": 2.4985382556915283,
+      "learning_rate": 5.354627157976803e-06,
+      "loss": 0.4796,
+      "step": 11265
+    },
+    {
+      "epoch": 1.4666146036704413,
+      "grad_norm": 3.4736976623535156,
+      "learning_rate": 5.35254254619446e-06,
+      "loss": 0.4909,
+      "step": 11268
+    },
+    {
+      "epoch": 1.467005076142132,
+      "grad_norm": 2.8436739444732666,
+      "learning_rate": 5.3504578728236755e-06,
+      "loss": 0.5168,
+      "step": 11271
+    },
+    {
+      "epoch": 1.4673955486138226,
+      "grad_norm": 3.0882272720336914,
+      "learning_rate": 5.34837313822864e-06,
+      "loss": 0.5582,
+      "step": 11274
+    },
+    {
+      "epoch": 1.4677860210855136,
+      "grad_norm": 5.404419898986816,
+      "learning_rate": 5.346288342773549e-06,
+      "loss": 0.511,
+      "step": 11277
+    },
+    {
+      "epoch": 1.4681764935572041,
+      "grad_norm": 2.911956548690796,
+      "learning_rate": 5.344203486822612e-06,
+      "loss": 0.4474,
+      "step": 11280
+    },
+    {
+      "epoch": 1.468566966028895,
+      "grad_norm": 3.251417398452759,
+      "learning_rate": 5.342118570740052e-06,
+      "loss": 0.565,
+      "step": 11283
+    },
+    {
+      "epoch": 1.4689574385005857,
+      "grad_norm": 2.6077330112457275,
+      "learning_rate": 5.340033594890096e-06,
+      "loss": 0.5114,
+      "step": 11286
+    },
+    {
+      "epoch": 1.4693479109722765,
+      "grad_norm": 2.613679885864258,
+      "learning_rate": 5.337948559636986e-06,
+      "loss": 0.4709,
+      "step": 11289
+    },
+    {
+      "epoch": 1.4697383834439672,
+      "grad_norm": 3.1710855960845947,
+      "learning_rate": 5.335863465344974e-06,
+      "loss": 0.4774,
+      "step": 11292
+    },
+    {
+      "epoch": 1.470128855915658,
+      "grad_norm": 2.876929521560669,
+      "learning_rate": 5.333778312378323e-06,
+      "loss": 0.5028,
+      "step": 11295
+    },
+    {
+      "epoch": 1.4705193283873488,
+      "grad_norm": 2.627826690673828,
+      "learning_rate": 5.3316931011013005e-06,
+      "loss": 0.4586,
+      "step": 11298
+    },
+    {
+      "epoch": 1.4709098008590393,
+      "grad_norm": 3.5629796981811523,
+      "learning_rate": 5.329607831878191e-06,
+      "loss": 0.5185,
+      "step": 11301
+    },
+    {
+      "epoch": 1.47130027333073,
+      "grad_norm": 2.592860221862793,
+      "learning_rate": 5.327522505073288e-06,
+      "loss": 0.508,
+      "step": 11304
+    },
+    {
+      "epoch": 1.4716907458024209,
+      "grad_norm": 2.701939344406128,
+      "learning_rate": 5.325437121050892e-06,
+      "loss": 0.5413,
+      "step": 11307
+    },
+    {
+      "epoch": 1.4720812182741116,
+      "grad_norm": 2.25862717628479,
+      "learning_rate": 5.323351680175315e-06,
+      "loss": 0.4955,
+      "step": 11310
+    },
+    {
+      "epoch": 1.4724716907458024,
+      "grad_norm": 2.581425666809082,
+      "learning_rate": 5.3212661828108804e-06,
+      "loss": 0.5118,
+      "step": 11313
+    },
+    {
+      "epoch": 1.4728621632174932,
+      "grad_norm": 2.7208588123321533,
+      "learning_rate": 5.31918062932192e-06,
+      "loss": 0.5058,
+      "step": 11316
+    },
+    {
+      "epoch": 1.473252635689184,
+      "grad_norm": 2.554607391357422,
+      "learning_rate": 5.317095020072773e-06,
+      "loss": 0.4651,
+      "step": 11319
+    },
+    {
+      "epoch": 1.4736431081608747,
+      "grad_norm": 2.7195706367492676,
+      "learning_rate": 5.315009355427795e-06,
+      "loss": 0.5694,
+      "step": 11322
+    },
+    {
+      "epoch": 1.4740335806325655,
+      "grad_norm": 2.790142774581909,
+      "learning_rate": 5.312923635751344e-06,
+      "loss": 0.4539,
+      "step": 11325
+    },
+    {
+      "epoch": 1.474424053104256,
+      "grad_norm": 2.372218132019043,
+      "learning_rate": 5.310837861407794e-06,
+      "loss": 0.4228,
+      "step": 11328
+    },
+    {
+      "epoch": 1.4748145255759468,
+      "grad_norm": 2.9723665714263916,
+      "learning_rate": 5.308752032761522e-06,
+      "loss": 0.469,
+      "step": 11331
+    },
+    {
+      "epoch": 1.4752049980476376,
+      "grad_norm": 2.6167614459991455,
+      "learning_rate": 5.306666150176919e-06,
+      "loss": 0.5017,
+      "step": 11334
+    },
+    {
+      "epoch": 1.4755954705193284,
+      "grad_norm": 2.730600357055664,
+      "learning_rate": 5.304580214018385e-06,
+      "loss": 0.5289,
+      "step": 11337
+    },
+    {
+      "epoch": 1.4759859429910192,
+      "grad_norm": 2.659367799758911,
+      "learning_rate": 5.30249422465033e-06,
+      "loss": 0.425,
+      "step": 11340
+    },
+    {
+      "epoch": 1.47637641546271,
+      "grad_norm": 2.761075019836426,
+      "learning_rate": 5.300408182437169e-06,
+      "loss": 0.5373,
+      "step": 11343
+    },
+    {
+      "epoch": 1.4767668879344007,
+      "grad_norm": 2.6184074878692627,
+      "learning_rate": 5.298322087743331e-06,
+      "loss": 0.5185,
+      "step": 11346
+    },
+    {
+      "epoch": 1.4771573604060912,
+      "grad_norm": 2.6529762744903564,
+      "learning_rate": 5.296235940933251e-06,
+      "loss": 0.4562,
+      "step": 11349
+    },
+    {
+      "epoch": 1.4775478328777822,
+      "grad_norm": 2.837994337081909,
+      "learning_rate": 5.294149742371375e-06,
+      "loss": 0.535,
+      "step": 11352
+    },
+    {
+      "epoch": 1.4779383053494728,
+      "grad_norm": 2.4222748279571533,
+      "learning_rate": 5.292063492422159e-06,
+      "loss": 0.4687,
+      "step": 11355
+    },
+    {
+      "epoch": 1.4783287778211636,
+      "grad_norm": 2.741462230682373,
+      "learning_rate": 5.289977191450064e-06,
+      "loss": 0.5317,
+      "step": 11358
+    },
+    {
+      "epoch": 1.4787192502928543,
+      "grad_norm": 2.7143518924713135,
+      "learning_rate": 5.287890839819566e-06,
+      "loss": 0.5798,
+      "step": 11361
+    },
+    {
+      "epoch": 1.479109722764545,
+      "grad_norm": 2.4520294666290283,
+      "learning_rate": 5.285804437895141e-06,
+      "loss": 0.4746,
+      "step": 11364
+    },
+    {
+      "epoch": 1.4795001952362359,
+      "grad_norm": 3.155306577682495,
+      "learning_rate": 5.283717986041285e-06,
+      "loss": 0.5228,
+      "step": 11367
+    },
+    {
+      "epoch": 1.4798906677079267,
+      "grad_norm": 2.5893876552581787,
+      "learning_rate": 5.281631484622491e-06,
+      "loss": 0.4359,
+      "step": 11370
+    },
+    {
+      "epoch": 1.4802811401796174,
+      "grad_norm": 2.689603090286255,
+      "learning_rate": 5.27954493400327e-06,
+      "loss": 0.5414,
+      "step": 11373
+    },
+    {
+      "epoch": 1.480671612651308,
+      "grad_norm": 2.8122377395629883,
+      "learning_rate": 5.277458334548138e-06,
+      "loss": 0.4752,
+      "step": 11376
+    },
+    {
+      "epoch": 1.4810620851229988,
+      "grad_norm": 2.8746378421783447,
+      "learning_rate": 5.2753716866216175e-06,
+      "loss": 0.5519,
+      "step": 11379
+    },
+    {
+      "epoch": 1.4814525575946895,
+      "grad_norm": 2.670448064804077,
+      "learning_rate": 5.273284990588243e-06,
+      "loss": 0.4566,
+      "step": 11382
+    },
+    {
+      "epoch": 1.4818430300663803,
+      "grad_norm": 2.758208751678467,
+      "learning_rate": 5.2711982468125556e-06,
+      "loss": 0.5005,
+      "step": 11385
+    },
+    {
+      "epoch": 1.482233502538071,
+      "grad_norm": 2.7960431575775146,
+      "learning_rate": 5.269111455659105e-06,
+      "loss": 0.4609,
+      "step": 11388
+    },
+    {
+      "epoch": 1.4826239750097618,
+      "grad_norm": 3.621656656265259,
+      "learning_rate": 5.26702461749245e-06,
+      "loss": 0.5131,
+      "step": 11391
+    },
+    {
+      "epoch": 1.4830144474814526,
+      "grad_norm": 3.8719372749328613,
+      "learning_rate": 5.264937732677153e-06,
+      "loss": 0.4925,
+      "step": 11394
+    },
+    {
+      "epoch": 1.4834049199531434,
+      "grad_norm": 2.7081613540649414,
+      "learning_rate": 5.262850801577796e-06,
+      "loss": 0.5013,
+      "step": 11397
+    },
+    {
+      "epoch": 1.4837953924248342,
+      "grad_norm": 2.990955114364624,
+      "learning_rate": 5.260763824558954e-06,
+      "loss": 0.5611,
+      "step": 11400
+    },
+    {
+      "epoch": 1.4841858648965247,
+      "grad_norm": 2.5260677337646484,
+      "learning_rate": 5.258676801985222e-06,
+      "loss": 0.5057,
+      "step": 11403
+    },
+    {
+      "epoch": 1.4845763373682155,
+      "grad_norm": 2.441223382949829,
+      "learning_rate": 5.256589734221196e-06,
+      "loss": 0.4837,
+      "step": 11406
+    },
+    {
+      "epoch": 1.4849668098399063,
+      "grad_norm": 2.5809214115142822,
+      "learning_rate": 5.254502621631482e-06,
+      "loss": 0.4651,
+      "step": 11409
+    },
+    {
+      "epoch": 1.485357282311597,
+      "grad_norm": 2.6332476139068604,
+      "learning_rate": 5.252415464580698e-06,
+      "loss": 0.5485,
+      "step": 11412
+    },
+    {
+      "epoch": 1.4857477547832878,
+      "grad_norm": 2.675795316696167,
+      "learning_rate": 5.25032826343346e-06,
+      "loss": 0.5142,
+      "step": 11415
+    },
+    {
+      "epoch": 1.4861382272549786,
+      "grad_norm": 2.489439010620117,
+      "learning_rate": 5.248241018554404e-06,
+      "loss": 0.4575,
+      "step": 11418
+    },
+    {
+      "epoch": 1.4865286997266693,
+      "grad_norm": 2.5916011333465576,
+      "learning_rate": 5.246153730308162e-06,
+      "loss": 0.51,
+      "step": 11421
+    },
+    {
+      "epoch": 1.48691917219836,
+      "grad_norm": 2.8045613765716553,
+      "learning_rate": 5.244066399059385e-06,
+      "loss": 0.4623,
+      "step": 11424
+    },
+    {
+      "epoch": 1.487309644670051,
+      "grad_norm": 2.699486255645752,
+      "learning_rate": 5.241979025172717e-06,
+      "loss": 0.5239,
+      "step": 11427
+    },
+    {
+      "epoch": 1.4877001171417414,
+      "grad_norm": 2.300581693649292,
+      "learning_rate": 5.239891609012824e-06,
+      "loss": 0.4958,
+      "step": 11430
+    },
+    {
+      "epoch": 1.4880905896134322,
+      "grad_norm": 2.496739625930786,
+      "learning_rate": 5.237804150944373e-06,
+      "loss": 0.5159,
+      "step": 11433
+    },
+    {
+      "epoch": 1.488481062085123,
+      "grad_norm": 2.7403767108917236,
+      "learning_rate": 5.2357166513320344e-06,
+      "loss": 0.5758,
+      "step": 11436
+    },
+    {
+      "epoch": 1.4888715345568138,
+      "grad_norm": 3.07275128364563,
+      "learning_rate": 5.233629110540494e-06,
+      "loss": 0.4913,
+      "step": 11439
+    },
+    {
+      "epoch": 1.4892620070285045,
+      "grad_norm": 2.733052968978882,
+      "learning_rate": 5.2315415289344405e-06,
+      "loss": 0.5081,
+      "step": 11442
+    },
+    {
+      "epoch": 1.4896524795001953,
+      "grad_norm": 2.944462299346924,
+      "learning_rate": 5.229453906878569e-06,
+      "loss": 0.5173,
+      "step": 11445
+    },
+    {
+      "epoch": 1.490042951971886,
+      "grad_norm": 2.523071527481079,
+      "learning_rate": 5.227366244737582e-06,
+      "loss": 0.5012,
+      "step": 11448
+    },
+    {
+      "epoch": 1.4904334244435766,
+      "grad_norm": 2.42546010017395,
+      "learning_rate": 5.225278542876189e-06,
+      "loss": 0.5246,
+      "step": 11451
+    },
+    {
+      "epoch": 1.4908238969152674,
+      "grad_norm": 2.5009751319885254,
+      "learning_rate": 5.2231908016591104e-06,
+      "loss": 0.4843,
+      "step": 11454
+    },
+    {
+      "epoch": 1.4912143693869582,
+      "grad_norm": 3.1024861335754395,
+      "learning_rate": 5.221103021451066e-06,
+      "loss": 0.4763,
+      "step": 11457
+    },
+    {
+      "epoch": 1.491604841858649,
+      "grad_norm": 2.727797269821167,
+      "learning_rate": 5.219015202616792e-06,
+      "loss": 0.5251,
+      "step": 11460
+    },
+    {
+      "epoch": 1.4919953143303397,
+      "grad_norm": 2.642549753189087,
+      "learning_rate": 5.2169273455210205e-06,
+      "loss": 0.4429,
+      "step": 11463
+    },
+    {
+      "epoch": 1.4923857868020305,
+      "grad_norm": 2.5647194385528564,
+      "learning_rate": 5.214839450528498e-06,
+      "loss": 0.5479,
+      "step": 11466
+    },
+    {
+      "epoch": 1.4927762592737213,
+      "grad_norm": 2.818277597427368,
+      "learning_rate": 5.212751518003977e-06,
+      "loss": 0.4881,
+      "step": 11469
+    },
+    {
+      "epoch": 1.4931667317454118,
+      "grad_norm": 2.8775992393493652,
+      "learning_rate": 5.210663548312212e-06,
+      "loss": 0.536,
+      "step": 11472
+    },
+    {
+      "epoch": 1.4935572042171028,
+      "grad_norm": 3.7240641117095947,
+      "learning_rate": 5.208575541817971e-06,
+      "loss": 0.4804,
+      "step": 11475
+    },
+    {
+      "epoch": 1.4939476766887934,
+      "grad_norm": 2.5948472023010254,
+      "learning_rate": 5.206487498886017e-06,
+      "loss": 0.6184,
+      "step": 11478
+    },
+    {
+      "epoch": 1.4943381491604841,
+      "grad_norm": 3.2808594703674316,
+      "learning_rate": 5.2043994198811356e-06,
+      "loss": 0.6083,
+      "step": 11481
+    },
+    {
+      "epoch": 1.494728621632175,
+      "grad_norm": 2.4906272888183594,
+      "learning_rate": 5.202311305168103e-06,
+      "loss": 0.4931,
+      "step": 11484
+    },
+    {
+      "epoch": 1.4951190941038657,
+      "grad_norm": 2.869908571243286,
+      "learning_rate": 5.200223155111711e-06,
+      "loss": 0.5207,
+      "step": 11487
+    },
+    {
+      "epoch": 1.4955095665755564,
+      "grad_norm": 2.8915436267852783,
+      "learning_rate": 5.198134970076757e-06,
+      "loss": 0.5586,
+      "step": 11490
+    },
+    {
+      "epoch": 1.4959000390472472,
+      "grad_norm": 3.477677345275879,
+      "learning_rate": 5.196046750428039e-06,
+      "loss": 0.5211,
+      "step": 11493
+    },
+    {
+      "epoch": 1.496290511518938,
+      "grad_norm": 2.8467986583709717,
+      "learning_rate": 5.193958496530367e-06,
+      "loss": 0.4828,
+      "step": 11496
+    },
+    {
+      "epoch": 1.4966809839906285,
+      "grad_norm": 2.7901194095611572,
+      "learning_rate": 5.1918702087485515e-06,
+      "loss": 0.5162,
+      "step": 11499
+    },
+    {
+      "epoch": 1.4970714564623195,
+      "grad_norm": 2.9377663135528564,
+      "learning_rate": 5.189781887447417e-06,
+      "loss": 0.4016,
+      "step": 11502
+    },
+    {
+      "epoch": 1.49746192893401,
+      "grad_norm": 2.7478699684143066,
+      "learning_rate": 5.187693532991784e-06,
+      "loss": 0.4597,
+      "step": 11505
+    },
+    {
+      "epoch": 1.4978524014057009,
+      "grad_norm": 2.764273166656494,
+      "learning_rate": 5.185605145746487e-06,
+      "loss": 0.4994,
+      "step": 11508
+    },
+    {
+      "epoch": 1.4982428738773916,
+      "grad_norm": 2.593223810195923,
+      "learning_rate": 5.183516726076362e-06,
+      "loss": 0.5124,
+      "step": 11511
+    },
+    {
+      "epoch": 1.4986333463490824,
+      "grad_norm": 2.9273481369018555,
+      "learning_rate": 5.181428274346249e-06,
+      "loss": 0.4994,
+      "step": 11514
+    },
+    {
+      "epoch": 1.4990238188207732,
+      "grad_norm": 2.4240915775299072,
+      "learning_rate": 5.179339790920999e-06,
+      "loss": 0.485,
+      "step": 11517
+    },
+    {
+      "epoch": 1.499414291292464,
+      "grad_norm": 2.7207272052764893,
+      "learning_rate": 5.177251276165465e-06,
+      "loss": 0.538,
+      "step": 11520
+    },
+    {
+      "epoch": 1.4998047637641547,
+      "grad_norm": 2.5581884384155273,
+      "learning_rate": 5.175162730444505e-06,
+      "loss": 0.4982,
+      "step": 11523
+    },
+    {
+      "epoch": 1.5001952362358453,
+      "grad_norm": 2.6691441535949707,
+      "learning_rate": 5.173074154122986e-06,
+      "loss": 0.5418,
+      "step": 11526
+    },
+    {
+      "epoch": 1.5005857087075363,
+      "grad_norm": 2.560971736907959,
+      "learning_rate": 5.170985547565775e-06,
+      "loss": 0.473,
+      "step": 11529
+    },
+    {
+      "epoch": 1.5009761811792268,
+      "grad_norm": 2.5895869731903076,
+      "learning_rate": 5.1688969111377505e-06,
+      "loss": 0.4541,
+      "step": 11532
+    },
+    {
+      "epoch": 1.5013666536509176,
+      "grad_norm": 2.396502733230591,
+      "learning_rate": 5.166808245203789e-06,
+      "loss": 0.51,
+      "step": 11535
+    },
+    {
+      "epoch": 1.5017571261226084,
+      "grad_norm": 2.532872200012207,
+      "learning_rate": 5.164719550128779e-06,
+      "loss": 0.5406,
+      "step": 11538
+    },
+    {
+      "epoch": 1.5021475985942991,
+      "grad_norm": 2.419482946395874,
+      "learning_rate": 5.162630826277609e-06,
+      "loss": 0.4598,
+      "step": 11541
+    },
+    {
+      "epoch": 1.50253807106599,
+      "grad_norm": 2.596123218536377,
+      "learning_rate": 5.160542074015177e-06,
+      "loss": 0.4867,
+      "step": 11544
+    },
+    {
+      "epoch": 1.5029285435376805,
+      "grad_norm": 2.9170663356781006,
+      "learning_rate": 5.158453293706383e-06,
+      "loss": 0.4896,
+      "step": 11547
+    },
+    {
+      "epoch": 1.5033190160093715,
+      "grad_norm": 2.6187119483947754,
+      "learning_rate": 5.15636448571613e-06,
+      "loss": 0.4578,
+      "step": 11550
+    },
+    {
+      "epoch": 1.503709488481062,
+      "grad_norm": 2.4606587886810303,
+      "learning_rate": 5.1542756504093315e-06,
+      "loss": 0.4461,
+      "step": 11553
+    },
+    {
+      "epoch": 1.5040999609527528,
+      "grad_norm": 2.533982038497925,
+      "learning_rate": 5.152186788150901e-06,
+      "loss": 0.499,
+      "step": 11556
+    },
+    {
+      "epoch": 1.5044904334244436,
+      "grad_norm": 2.852947473526001,
+      "learning_rate": 5.1500978993057596e-06,
+      "loss": 0.5007,
+      "step": 11559
+    },
+    {
+      "epoch": 1.5048809058961343,
+      "grad_norm": 3.0088095664978027,
+      "learning_rate": 5.1480089842388295e-06,
+      "loss": 0.5055,
+      "step": 11562
+    },
+    {
+      "epoch": 1.505271378367825,
+      "grad_norm": 2.5373263359069824,
+      "learning_rate": 5.145920043315041e-06,
+      "loss": 0.4603,
+      "step": 11565
+    },
+    {
+      "epoch": 1.5056618508395159,
+      "grad_norm": 2.940673828125,
+      "learning_rate": 5.143831076899329e-06,
+      "loss": 0.6183,
+      "step": 11568
+    },
+    {
+      "epoch": 1.5060523233112066,
+      "grad_norm": 2.945150375366211,
+      "learning_rate": 5.14174208535663e-06,
+      "loss": 0.55,
+      "step": 11571
+    },
+    {
+      "epoch": 1.5064427957828972,
+      "grad_norm": 2.7196662425994873,
+      "learning_rate": 5.1396530690518876e-06,
+      "loss": 0.5235,
+      "step": 11574
+    },
+    {
+      "epoch": 1.5068332682545882,
+      "grad_norm": 2.892042875289917,
+      "learning_rate": 5.137564028350048e-06,
+      "loss": 0.5291,
+      "step": 11577
+    },
+    {
+      "epoch": 1.5072237407262787,
+      "grad_norm": 2.4638333320617676,
+      "learning_rate": 5.135474963616062e-06,
+      "loss": 0.4451,
+      "step": 11580
+    },
+    {
+      "epoch": 1.5076142131979695,
+      "grad_norm": 2.493852138519287,
+      "learning_rate": 5.133385875214883e-06,
+      "loss": 0.5536,
+      "step": 11583
+    },
+    {
+      "epoch": 1.5080046856696603,
+      "grad_norm": 2.891443967819214,
+      "learning_rate": 5.131296763511473e-06,
+      "loss": 0.4479,
+      "step": 11586
+    },
+    {
+      "epoch": 1.508395158141351,
+      "grad_norm": 2.568324327468872,
+      "learning_rate": 5.129207628870796e-06,
+      "loss": 0.4524,
+      "step": 11589
+    },
+    {
+      "epoch": 1.5087856306130418,
+      "grad_norm": 3.3236899375915527,
+      "learning_rate": 5.127118471657817e-06,
+      "loss": 0.4885,
+      "step": 11592
+    },
+    {
+      "epoch": 1.5091761030847324,
+      "grad_norm": 2.7332510948181152,
+      "learning_rate": 5.125029292237508e-06,
+      "loss": 0.5649,
+      "step": 11595
+    },
+    {
+      "epoch": 1.5095665755564234,
+      "grad_norm": 2.6280899047851562,
+      "learning_rate": 5.122940090974846e-06,
+      "loss": 0.4571,
+      "step": 11598
+    },
+    {
+      "epoch": 1.509957048028114,
+      "grad_norm": 2.642153024673462,
+      "learning_rate": 5.120850868234808e-06,
+      "loss": 0.5028,
+      "step": 11601
+    },
+    {
+      "epoch": 1.510347520499805,
+      "grad_norm": 2.4956655502319336,
+      "learning_rate": 5.118761624382377e-06,
+      "loss": 0.5421,
+      "step": 11604
+    },
+    {
+      "epoch": 1.5107379929714955,
+      "grad_norm": 3.0363759994506836,
+      "learning_rate": 5.11667235978254e-06,
+      "loss": 0.5359,
+      "step": 11607
+    },
+    {
+      "epoch": 1.5111284654431862,
+      "grad_norm": 3.446927070617676,
+      "learning_rate": 5.114583074800289e-06,
+      "loss": 0.5317,
+      "step": 11610
+    },
+    {
+      "epoch": 1.511518937914877,
+      "grad_norm": 2.6204493045806885,
+      "learning_rate": 5.112493769800614e-06,
+      "loss": 0.4863,
+      "step": 11613
+    },
+    {
+      "epoch": 1.5119094103865678,
+      "grad_norm": 2.5534026622772217,
+      "learning_rate": 5.110404445148515e-06,
+      "loss": 0.5407,
+      "step": 11616
+    },
+    {
+      "epoch": 1.5122998828582586,
+      "grad_norm": 2.5729386806488037,
+      "learning_rate": 5.108315101208991e-06,
+      "loss": 0.538,
+      "step": 11619
+    },
+    {
+      "epoch": 1.512690355329949,
+      "grad_norm": 2.6536967754364014,
+      "learning_rate": 5.106225738347047e-06,
+      "loss": 0.4916,
+      "step": 11622
+    },
+    {
+      "epoch": 1.51308082780164,
+      "grad_norm": 2.6462488174438477,
+      "learning_rate": 5.10413635692769e-06,
+      "loss": 0.4578,
+      "step": 11625
+    },
+    {
+      "epoch": 1.5134713002733307,
+      "grad_norm": 2.5776121616363525,
+      "learning_rate": 5.10204695731593e-06,
+      "loss": 0.5028,
+      "step": 11628
+    },
+    {
+      "epoch": 1.5138617727450214,
+      "grad_norm": 2.8399088382720947,
+      "learning_rate": 5.099957539876783e-06,
+      "loss": 0.5134,
+      "step": 11631
+    },
+    {
+      "epoch": 1.5142522452167122,
+      "grad_norm": 2.53810453414917,
+      "learning_rate": 5.097868104975262e-06,
+      "loss": 0.5759,
+      "step": 11634
+    },
+    {
+      "epoch": 1.514642717688403,
+      "grad_norm": 2.6173861026763916,
+      "learning_rate": 5.09577865297639e-06,
+      "loss": 0.522,
+      "step": 11637
+    },
+    {
+      "epoch": 1.5150331901600937,
+      "grad_norm": 2.800180673599243,
+      "learning_rate": 5.0936891842451895e-06,
+      "loss": 0.6223,
+      "step": 11640
+    },
+    {
+      "epoch": 1.5154236626317843,
+      "grad_norm": 2.824063539505005,
+      "learning_rate": 5.0915996991466845e-06,
+      "loss": 0.4638,
+      "step": 11643
+    },
+    {
+      "epoch": 1.5158141351034753,
+      "grad_norm": 2.6472010612487793,
+      "learning_rate": 5.089510198045904e-06,
+      "loss": 0.4483,
+      "step": 11646
+    },
+    {
+      "epoch": 1.5162046075751658,
+      "grad_norm": 2.6510941982269287,
+      "learning_rate": 5.087420681307881e-06,
+      "loss": 0.5283,
+      "step": 11649
+    },
+    {
+      "epoch": 1.5165950800468568,
+      "grad_norm": 2.5269935131073,
+      "learning_rate": 5.085331149297649e-06,
+      "loss": 0.4922,
+      "step": 11652
+    },
+    {
+      "epoch": 1.5169855525185474,
+      "grad_norm": 2.427276134490967,
+      "learning_rate": 5.083241602380246e-06,
+      "loss": 0.4429,
+      "step": 11655
+    },
+    {
+      "epoch": 1.5173760249902382,
+      "grad_norm": 2.5796139240264893,
+      "learning_rate": 5.081152040920708e-06,
+      "loss": 0.5555,
+      "step": 11658
+    },
+    {
+      "epoch": 1.517766497461929,
+      "grad_norm": 3.097916603088379,
+      "learning_rate": 5.079062465284081e-06,
+      "loss": 0.5131,
+      "step": 11661
+    },
+    {
+      "epoch": 1.5181569699336197,
+      "grad_norm": 2.6202802658081055,
+      "learning_rate": 5.076972875835406e-06,
+      "loss": 0.5081,
+      "step": 11664
+    },
+    {
+      "epoch": 1.5185474424053105,
+      "grad_norm": 2.4596011638641357,
+      "learning_rate": 5.074883272939732e-06,
+      "loss": 0.4617,
+      "step": 11667
+    },
+    {
+      "epoch": 1.518937914877001,
+      "grad_norm": 3.6132545471191406,
+      "learning_rate": 5.072793656962108e-06,
+      "loss": 0.4359,
+      "step": 11670
+    },
+    {
+      "epoch": 1.519328387348692,
+      "grad_norm": 2.3989944458007812,
+      "learning_rate": 5.0707040282675855e-06,
+      "loss": 0.4565,
+      "step": 11673
+    },
+    {
+      "epoch": 1.5197188598203826,
+      "grad_norm": 3.203303813934326,
+      "learning_rate": 5.068614387221218e-06,
+      "loss": 0.4667,
+      "step": 11676
+    },
+    {
+      "epoch": 1.5201093322920736,
+      "grad_norm": 2.8090553283691406,
+      "learning_rate": 5.066524734188061e-06,
+      "loss": 0.5105,
+      "step": 11679
+    },
+    {
+      "epoch": 1.5204998047637641,
+      "grad_norm": 2.759253740310669,
+      "learning_rate": 5.064435069533174e-06,
+      "loss": 0.5128,
+      "step": 11682
+    },
+    {
+      "epoch": 1.520890277235455,
+      "grad_norm": 2.5702476501464844,
+      "learning_rate": 5.062345393621615e-06,
+      "loss": 0.4793,
+      "step": 11685
+    },
+    {
+      "epoch": 1.5212807497071457,
+      "grad_norm": 2.932741403579712,
+      "learning_rate": 5.060255706818447e-06,
+      "loss": 0.4439,
+      "step": 11688
+    },
+    {
+      "epoch": 1.5216712221788364,
+      "grad_norm": 2.6461048126220703,
+      "learning_rate": 5.058166009488733e-06,
+      "loss": 0.5125,
+      "step": 11691
+    },
+    {
+      "epoch": 1.5220616946505272,
+      "grad_norm": 2.705612897872925,
+      "learning_rate": 5.05607630199754e-06,
+      "loss": 0.533,
+      "step": 11694
+    },
+    {
+      "epoch": 1.5224521671222178,
+      "grad_norm": 2.7387025356292725,
+      "learning_rate": 5.0539865847099354e-06,
+      "loss": 0.4833,
+      "step": 11697
+    },
+    {
+      "epoch": 1.5228426395939088,
+      "grad_norm": 2.522993803024292,
+      "learning_rate": 5.051896857990988e-06,
+      "loss": 0.5356,
+      "step": 11700
+    },
+    {
+      "epoch": 1.5232331120655993,
+      "grad_norm": 2.885016441345215,
+      "learning_rate": 5.049807122205768e-06,
+      "loss": 0.5451,
+      "step": 11703
+    },
+    {
+      "epoch": 1.52362358453729,
+      "grad_norm": 2.7136266231536865,
+      "learning_rate": 5.04771737771935e-06,
+      "loss": 0.4738,
+      "step": 11706
+    },
+    {
+      "epoch": 1.5240140570089808,
+      "grad_norm": 2.6322624683380127,
+      "learning_rate": 5.045627624896804e-06,
+      "loss": 0.5244,
+      "step": 11709
+    },
+    {
+      "epoch": 1.5244045294806716,
+      "grad_norm": 3.288853406906128,
+      "learning_rate": 5.0435378641032095e-06,
+      "loss": 0.5564,
+      "step": 11712
+    },
+    {
+      "epoch": 1.5247950019523624,
+      "grad_norm": 2.5992119312286377,
+      "learning_rate": 5.0414480957036415e-06,
+      "loss": 0.5101,
+      "step": 11715
+    },
+    {
+      "epoch": 1.525185474424053,
+      "grad_norm": 2.7537200450897217,
+      "learning_rate": 5.039358320063179e-06,
+      "loss": 0.5499,
+      "step": 11718
+    },
+    {
+      "epoch": 1.525575946895744,
+      "grad_norm": 2.5986032485961914,
+      "learning_rate": 5.037268537546901e-06,
+      "loss": 0.4637,
+      "step": 11721
+    },
+    {
+      "epoch": 1.5259664193674345,
+      "grad_norm": 2.795783042907715,
+      "learning_rate": 5.035178748519887e-06,
+      "loss": 0.5092,
+      "step": 11724
+    },
+    {
+      "epoch": 1.5263568918391255,
+      "grad_norm": 2.8021063804626465,
+      "learning_rate": 5.03308895334722e-06,
+      "loss": 0.5189,
+      "step": 11727
+    },
+    {
+      "epoch": 1.526747364310816,
+      "grad_norm": 2.8685810565948486,
+      "learning_rate": 5.0309991523939805e-06,
+      "loss": 0.5911,
+      "step": 11730
+    },
+    {
+      "epoch": 1.5271378367825068,
+      "grad_norm": 2.5764682292938232,
+      "learning_rate": 5.028909346025257e-06,
+      "loss": 0.4561,
+      "step": 11733
+    },
+    {
+      "epoch": 1.5275283092541976,
+      "grad_norm": 2.9899494647979736,
+      "learning_rate": 5.026819534606131e-06,
+      "loss": 0.4998,
+      "step": 11736
+    },
+    {
+      "epoch": 1.5279187817258884,
+      "grad_norm": 2.4830875396728516,
+      "learning_rate": 5.024729718501688e-06,
+      "loss": 0.4199,
+      "step": 11739
+    },
+    {
+      "epoch": 1.5283092541975791,
+      "grad_norm": 2.9966747760772705,
+      "learning_rate": 5.022639898077016e-06,
+      "loss": 0.4864,
+      "step": 11742
+    },
+    {
+      "epoch": 1.5286997266692697,
+      "grad_norm": 2.7699429988861084,
+      "learning_rate": 5.020550073697202e-06,
+      "loss": 0.5197,
+      "step": 11745
+    },
+    {
+      "epoch": 1.5290901991409607,
+      "grad_norm": 3.0958468914031982,
+      "learning_rate": 5.018460245727333e-06,
+      "loss": 0.4696,
+      "step": 11748
+    },
+    {
+      "epoch": 1.5294806716126512,
+      "grad_norm": 2.7975351810455322,
+      "learning_rate": 5.016370414532495e-06,
+      "loss": 0.563,
+      "step": 11751
+    },
+    {
+      "epoch": 1.5298711440843422,
+      "grad_norm": 3.140310525894165,
+      "learning_rate": 5.014280580477782e-06,
+      "loss": 0.5284,
+      "step": 11754
+    },
+    {
+      "epoch": 1.5302616165560328,
+      "grad_norm": 2.853102684020996,
+      "learning_rate": 5.012190743928282e-06,
+      "loss": 0.4683,
+      "step": 11757
+    },
+    {
+      "epoch": 1.5306520890277235,
+      "grad_norm": 2.716357946395874,
+      "learning_rate": 5.010100905249084e-06,
+      "loss": 0.4811,
+      "step": 11760
+    },
+    {
+      "epoch": 1.5310425614994143,
+      "grad_norm": 3.0148074626922607,
+      "learning_rate": 5.0080110648052815e-06,
+      "loss": 0.5135,
+      "step": 11763
+    },
+    {
+      "epoch": 1.531433033971105,
+      "grad_norm": 2.7584571838378906,
+      "learning_rate": 5.00592122296196e-06,
+      "loss": 0.4751,
+      "step": 11766
+    },
+    {
+      "epoch": 1.5318235064427959,
+      "grad_norm": 2.487861156463623,
+      "learning_rate": 5.003831380084216e-06,
+      "loss": 0.5302,
+      "step": 11769
+    },
+    {
+      "epoch": 1.5322139789144864,
+      "grad_norm": 2.4335525035858154,
+      "learning_rate": 5.001741536537135e-06,
+      "loss": 0.4591,
+      "step": 11772
+    },
+    {
+      "epoch": 1.5326044513861774,
+      "grad_norm": 2.387814521789551,
+      "learning_rate": 4.999651692685813e-06,
+      "loss": 0.4683,
+      "step": 11775
+    },
+    {
+      "epoch": 1.532994923857868,
+      "grad_norm": 2.30008602142334,
+      "learning_rate": 4.997561848895338e-06,
+      "loss": 0.4718,
+      "step": 11778
+    },
+    {
+      "epoch": 1.5333853963295587,
+      "grad_norm": 2.351167678833008,
+      "learning_rate": 4.995472005530804e-06,
+      "loss": 0.4853,
+      "step": 11781
+    },
+    {
+      "epoch": 1.5337758688012495,
+      "grad_norm": 2.77471923828125,
+      "learning_rate": 4.993382162957302e-06,
+      "loss": 0.4924,
+      "step": 11784
+    },
+    {
+      "epoch": 1.5341663412729403,
+      "grad_norm": 2.6090359687805176,
+      "learning_rate": 4.991292321539921e-06,
+      "loss": 0.5111,
+      "step": 11787
+    },
+    {
+      "epoch": 1.534556813744631,
+      "grad_norm": 2.52298903465271,
+      "learning_rate": 4.989202481643755e-06,
+      "loss": 0.4663,
+      "step": 11790
+    },
+    {
+      "epoch": 1.5349472862163216,
+      "grad_norm": 3.639498472213745,
+      "learning_rate": 4.98711264363389e-06,
+      "loss": 0.4876,
+      "step": 11793
+    },
+    {
+      "epoch": 1.5353377586880126,
+      "grad_norm": 2.813267946243286,
+      "learning_rate": 4.98502280787542e-06,
+      "loss": 0.5949,
+      "step": 11796
+    },
+    {
+      "epoch": 1.5357282311597031,
+      "grad_norm": 2.9353835582733154,
+      "learning_rate": 4.9829329747334345e-06,
+      "loss": 0.4813,
+      "step": 11799
+    },
+    {
+      "epoch": 1.5361187036313941,
+      "grad_norm": 2.37113618850708,
+      "learning_rate": 4.9808431445730225e-06,
+      "loss": 0.4632,
+      "step": 11802
+    },
+    {
+      "epoch": 1.5365091761030847,
+      "grad_norm": 3.059115409851074,
+      "learning_rate": 4.978753317759271e-06,
+      "loss": 0.5102,
+      "step": 11805
+    },
+    {
+      "epoch": 1.5368996485747755,
+      "grad_norm": 2.4613428115844727,
+      "learning_rate": 4.976663494657271e-06,
+      "loss": 0.4513,
+      "step": 11808
+    },
+    {
+      "epoch": 1.5372901210464662,
+      "grad_norm": 2.785877227783203,
+      "learning_rate": 4.97457367563211e-06,
+      "loss": 0.5541,
+      "step": 11811
+    },
+    {
+      "epoch": 1.537680593518157,
+      "grad_norm": 2.452495574951172,
+      "learning_rate": 4.972483861048875e-06,
+      "loss": 0.4388,
+      "step": 11814
+    },
+    {
+      "epoch": 1.5380710659898478,
+      "grad_norm": 2.4876022338867188,
+      "learning_rate": 4.970394051272651e-06,
+      "loss": 0.5395,
+      "step": 11817
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 2.76699161529541,
+      "learning_rate": 4.968304246668524e-06,
+      "loss": 0.5129,
+      "step": 11820
+    },
+    {
+      "epoch": 1.5388520109332293,
+      "grad_norm": 2.546515703201294,
+      "learning_rate": 4.9662144476015785e-06,
+      "loss": 0.4904,
+      "step": 11823
+    },
+    {
+      "epoch": 1.5392424834049199,
+      "grad_norm": 2.5809149742126465,
+      "learning_rate": 4.964124654436898e-06,
+      "loss": 0.4912,
+      "step": 11826
+    },
+    {
+      "epoch": 1.5396329558766109,
+      "grad_norm": 3.259404182434082,
+      "learning_rate": 4.962034867539565e-06,
+      "loss": 0.5205,
+      "step": 11829
+    },
+    {
+      "epoch": 1.5400234283483014,
+      "grad_norm": 2.743436098098755,
+      "learning_rate": 4.959945087274659e-06,
+      "loss": 0.4914,
+      "step": 11832
+    },
+    {
+      "epoch": 1.5404139008199922,
+      "grad_norm": 2.653174638748169,
+      "learning_rate": 4.957855314007265e-06,
+      "loss": 0.5208,
+      "step": 11835
+    },
+    {
+      "epoch": 1.540804373291683,
+      "grad_norm": 2.54440975189209,
+      "learning_rate": 4.955765548102457e-06,
+      "loss": 0.4773,
+      "step": 11838
+    },
+    {
+      "epoch": 1.5411948457633737,
+      "grad_norm": 3.731177806854248,
+      "learning_rate": 4.9536757899253156e-06,
+      "loss": 0.4968,
+      "step": 11841
+    },
+    {
+      "epoch": 1.5415853182350645,
+      "grad_norm": 2.7818851470947266,
+      "learning_rate": 4.951586039840916e-06,
+      "loss": 0.4352,
+      "step": 11844
+    },
+    {
+      "epoch": 1.541975790706755,
+      "grad_norm": 2.795779228210449,
+      "learning_rate": 4.949496298214331e-06,
+      "loss": 0.4873,
+      "step": 11847
+    },
+    {
+      "epoch": 1.542366263178446,
+      "grad_norm": 2.714663028717041,
+      "learning_rate": 4.9474065654106384e-06,
+      "loss": 0.6013,
+      "step": 11850
+    },
+    {
+      "epoch": 1.5427567356501366,
+      "grad_norm": 2.7950053215026855,
+      "learning_rate": 4.945316841794909e-06,
+      "loss": 0.4848,
+      "step": 11853
+    },
+    {
+      "epoch": 1.5431472081218274,
+      "grad_norm": 2.954528570175171,
+      "learning_rate": 4.943227127732212e-06,
+      "loss": 0.6347,
+      "step": 11856
+    },
+    {
+      "epoch": 1.5435376805935181,
+      "grad_norm": 2.76151442527771,
+      "learning_rate": 4.9411374235876155e-06,
+      "loss": 0.5226,
+      "step": 11859
+    },
+    {
+      "epoch": 1.543928153065209,
+      "grad_norm": 4.05312967300415,
+      "learning_rate": 4.939047729726189e-06,
+      "loss": 0.5568,
+      "step": 11862
+    },
+    {
+      "epoch": 1.5443186255368997,
+      "grad_norm": 2.8274898529052734,
+      "learning_rate": 4.936958046512994e-06,
+      "loss": 0.5498,
+      "step": 11865
+    },
+    {
+      "epoch": 1.5447090980085902,
+      "grad_norm": 2.768404483795166,
+      "learning_rate": 4.934868374313097e-06,
+      "loss": 0.5259,
+      "step": 11868
+    },
+    {
+      "epoch": 1.5450995704802812,
+      "grad_norm": 2.711665153503418,
+      "learning_rate": 4.9327787134915576e-06,
+      "loss": 0.5094,
+      "step": 11871
+    },
+    {
+      "epoch": 1.5454900429519718,
+      "grad_norm": 2.3921022415161133,
+      "learning_rate": 4.9306890644134345e-06,
+      "loss": 0.4367,
+      "step": 11874
+    },
+    {
+      "epoch": 1.5458805154236628,
+      "grad_norm": 2.8708338737487793,
+      "learning_rate": 4.928599427443788e-06,
+      "loss": 0.5638,
+      "step": 11877
+    },
+    {
+      "epoch": 1.5462709878953533,
+      "grad_norm": 2.8968679904937744,
+      "learning_rate": 4.92650980294767e-06,
+      "loss": 0.524,
+      "step": 11880
+    },
+    {
+      "epoch": 1.546661460367044,
+      "grad_norm": 3.2613213062286377,
+      "learning_rate": 4.924420191290134e-06,
+      "loss": 0.4995,
+      "step": 11883
+    },
+    {
+      "epoch": 1.5470519328387349,
+      "grad_norm": 2.6707489490509033,
+      "learning_rate": 4.922330592836234e-06,
+      "loss": 0.4213,
+      "step": 11886
+    },
+    {
+      "epoch": 1.5474424053104257,
+      "grad_norm": 2.552860736846924,
+      "learning_rate": 4.920241007951012e-06,
+      "loss": 0.4674,
+      "step": 11889
+    },
+    {
+      "epoch": 1.5478328777821164,
+      "grad_norm": 3.741816282272339,
+      "learning_rate": 4.918151436999521e-06,
+      "loss": 0.4466,
+      "step": 11892
+    },
+    {
+      "epoch": 1.548223350253807,
+      "grad_norm": 3.1363577842712402,
+      "learning_rate": 4.916061880346802e-06,
+      "loss": 0.4513,
+      "step": 11895
+    },
+    {
+      "epoch": 1.548613822725498,
+      "grad_norm": 3.1754071712493896,
+      "learning_rate": 4.913972338357895e-06,
+      "loss": 0.4901,
+      "step": 11898
+    },
+    {
+      "epoch": 1.5490042951971885,
+      "grad_norm": 2.3715884685516357,
+      "learning_rate": 4.91188281139784e-06,
+      "loss": 0.5472,
+      "step": 11901
+    },
+    {
+      "epoch": 1.5493947676688793,
+      "grad_norm": 2.7456414699554443,
+      "learning_rate": 4.909793299831673e-06,
+      "loss": 0.5158,
+      "step": 11904
+    },
+    {
+      "epoch": 1.54978524014057,
+      "grad_norm": 2.7107367515563965,
+      "learning_rate": 4.907703804024425e-06,
+      "loss": 0.4408,
+      "step": 11907
+    },
+    {
+      "epoch": 1.5501757126122608,
+      "grad_norm": 2.649536609649658,
+      "learning_rate": 4.90561432434113e-06,
+      "loss": 0.534,
+      "step": 11910
+    },
+    {
+      "epoch": 1.5505661850839516,
+      "grad_norm": 2.4240305423736572,
+      "learning_rate": 4.903524861146814e-06,
+      "loss": 0.49,
+      "step": 11913
+    },
+    {
+      "epoch": 1.5509566575556424,
+      "grad_norm": 2.6543915271759033,
+      "learning_rate": 4.901435414806502e-06,
+      "loss": 0.4731,
+      "step": 11916
+    },
+    {
+      "epoch": 1.5513471300273332,
+      "grad_norm": 2.658057928085327,
+      "learning_rate": 4.899345985685215e-06,
+      "loss": 0.4885,
+      "step": 11919
+    },
+    {
+      "epoch": 1.5517376024990237,
+      "grad_norm": 2.6774919033050537,
+      "learning_rate": 4.8972565741479715e-06,
+      "loss": 0.4726,
+      "step": 11922
+    },
+    {
+      "epoch": 1.5521280749707147,
+      "grad_norm": 2.612046718597412,
+      "learning_rate": 4.89516718055979e-06,
+      "loss": 0.4846,
+      "step": 11925
+    },
+    {
+      "epoch": 1.5525185474424053,
+      "grad_norm": 2.682743549346924,
+      "learning_rate": 4.893077805285681e-06,
+      "loss": 0.4907,
+      "step": 11928
+    },
+    {
+      "epoch": 1.552909019914096,
+      "grad_norm": 2.6726317405700684,
+      "learning_rate": 4.890988448690653e-06,
+      "loss": 0.4273,
+      "step": 11931
+    },
+    {
+      "epoch": 1.5532994923857868,
+      "grad_norm": 3.1792590618133545,
+      "learning_rate": 4.8888991111397144e-06,
+      "loss": 0.5008,
+      "step": 11934
+    },
+    {
+      "epoch": 1.5536899648574776,
+      "grad_norm": 2.972460985183716,
+      "learning_rate": 4.886809792997868e-06,
+      "loss": 0.534,
+      "step": 11937
+    },
+    {
+      "epoch": 1.5540804373291683,
+      "grad_norm": 2.578676223754883,
+      "learning_rate": 4.884720494630113e-06,
+      "loss": 0.5175,
+      "step": 11940
+    },
+    {
+      "epoch": 1.554470909800859,
+      "grad_norm": 2.6860499382019043,
+      "learning_rate": 4.882631216401445e-06,
+      "loss": 0.4532,
+      "step": 11943
+    },
+    {
+      "epoch": 1.5548613822725499,
+      "grad_norm": 2.5615968704223633,
+      "learning_rate": 4.880541958676856e-06,
+      "loss": 0.4637,
+      "step": 11946
+    },
+    {
+      "epoch": 1.5552518547442404,
+      "grad_norm": 2.9076266288757324,
+      "learning_rate": 4.878452721821336e-06,
+      "loss": 0.5384,
+      "step": 11949
+    },
+    {
+      "epoch": 1.5556423272159314,
+      "grad_norm": 2.601215124130249,
+      "learning_rate": 4.876363506199869e-06,
+      "loss": 0.4748,
+      "step": 11952
+    },
+    {
+      "epoch": 1.556032799687622,
+      "grad_norm": 2.8393771648406982,
+      "learning_rate": 4.874274312177439e-06,
+      "loss": 0.5763,
+      "step": 11955
+    },
+    {
+      "epoch": 1.5564232721593128,
+      "grad_norm": 2.3092920780181885,
+      "learning_rate": 4.87218514011902e-06,
+      "loss": 0.4322,
+      "step": 11958
+    },
+    {
+      "epoch": 1.5568137446310035,
+      "grad_norm": 2.9448838233947754,
+      "learning_rate": 4.87009599038959e-06,
+      "loss": 0.5118,
+      "step": 11961
+    },
+    {
+      "epoch": 1.5572042171026943,
+      "grad_norm": 2.6871285438537598,
+      "learning_rate": 4.868006863354117e-06,
+      "loss": 0.5206,
+      "step": 11964
+    },
+    {
+      "epoch": 1.557594689574385,
+      "grad_norm": 2.516818046569824,
+      "learning_rate": 4.865917759377567e-06,
+      "loss": 0.4144,
+      "step": 11967
+    },
+    {
+      "epoch": 1.5579851620460756,
+      "grad_norm": 3.410189151763916,
+      "learning_rate": 4.8638286788249025e-06,
+      "loss": 0.4504,
+      "step": 11970
+    },
+    {
+      "epoch": 1.5583756345177666,
+      "grad_norm": 3.0182759761810303,
+      "learning_rate": 4.861739622061078e-06,
+      "loss": 0.4726,
+      "step": 11973
+    },
+    {
+      "epoch": 1.5587661069894572,
+      "grad_norm": 2.5583255290985107,
+      "learning_rate": 4.859650589451055e-06,
+      "loss": 0.457,
+      "step": 11976
+    },
+    {
+      "epoch": 1.559156579461148,
+      "grad_norm": 2.401313304901123,
+      "learning_rate": 4.857561581359777e-06,
+      "loss": 0.5042,
+      "step": 11979
+    },
+    {
+      "epoch": 1.5595470519328387,
+      "grad_norm": 2.8148505687713623,
+      "learning_rate": 4.855472598152193e-06,
+      "loss": 0.4324,
+      "step": 11982
+    },
+    {
+      "epoch": 1.5599375244045295,
+      "grad_norm": 3.2659788131713867,
+      "learning_rate": 4.8533836401932395e-06,
+      "loss": 0.4854,
+      "step": 11985
+    },
+    {
+      "epoch": 1.5603279968762203,
+      "grad_norm": 2.5836613178253174,
+      "learning_rate": 4.851294707847856e-06,
+      "loss": 0.5318,
+      "step": 11988
+    },
+    {
+      "epoch": 1.5607184693479108,
+      "grad_norm": 2.268071413040161,
+      "learning_rate": 4.849205801480976e-06,
+      "loss": 0.4897,
+      "step": 11991
+    },
+    {
+      "epoch": 1.5611089418196018,
+      "grad_norm": 2.9751930236816406,
+      "learning_rate": 4.847116921457524e-06,
+      "loss": 0.4713,
+      "step": 11994
+    },
+    {
+      "epoch": 1.5614994142912924,
+      "grad_norm": 3.3204164505004883,
+      "learning_rate": 4.8450280681424235e-06,
+      "loss": 0.4727,
+      "step": 11997
+    },
+    {
+      "epoch": 1.5618898867629833,
+      "grad_norm": 2.7896077632904053,
+      "learning_rate": 4.842939241900595e-06,
+      "loss": 0.5829,
+      "step": 12000
+    },
+    {
+      "epoch": 1.562280359234674,
+      "grad_norm": 2.7530436515808105,
+      "learning_rate": 4.84085044309695e-06,
+      "loss": 0.4501,
+      "step": 12003
+    },
+    {
+      "epoch": 1.5626708317063647,
+      "grad_norm": 2.5019564628601074,
+      "learning_rate": 4.838761672096398e-06,
+      "loss": 0.4916,
+      "step": 12006
+    },
+    {
+      "epoch": 1.5630613041780554,
+      "grad_norm": 2.6398847103118896,
+      "learning_rate": 4.8366729292638425e-06,
+      "loss": 0.5605,
+      "step": 12009
+    },
+    {
+      "epoch": 1.5634517766497462,
+      "grad_norm": 3.4240760803222656,
+      "learning_rate": 4.834584214964182e-06,
+      "loss": 0.5437,
+      "step": 12012
+    },
+    {
+      "epoch": 1.563842249121437,
+      "grad_norm": 2.8203554153442383,
+      "learning_rate": 4.8324955295623105e-06,
+      "loss": 0.5711,
+      "step": 12015
+    },
+    {
+      "epoch": 1.5642327215931275,
+      "grad_norm": 2.9343690872192383,
+      "learning_rate": 4.8304068734231194e-06,
+      "loss": 0.4177,
+      "step": 12018
+    },
+    {
+      "epoch": 1.5646231940648185,
+      "grad_norm": 2.7858777046203613,
+      "learning_rate": 4.82831824691149e-06,
+      "loss": 0.4601,
+      "step": 12021
+    },
+    {
+      "epoch": 1.565013666536509,
+      "grad_norm": 2.6615116596221924,
+      "learning_rate": 4.826229650392301e-06,
+      "loss": 0.5292,
+      "step": 12024
+    },
+    {
+      "epoch": 1.5654041390082,
+      "grad_norm": 2.6517207622528076,
+      "learning_rate": 4.824141084230429e-06,
+      "loss": 0.5592,
+      "step": 12027
+    },
+    {
+      "epoch": 1.5657946114798906,
+      "grad_norm": 2.7702314853668213,
+      "learning_rate": 4.822052548790737e-06,
+      "loss": 0.5166,
+      "step": 12030
+    },
+    {
+      "epoch": 1.5661850839515814,
+      "grad_norm": 2.714916706085205,
+      "learning_rate": 4.819964044438092e-06,
+      "loss": 0.4259,
+      "step": 12033
+    },
+    {
+      "epoch": 1.5665755564232722,
+      "grad_norm": 2.4574501514434814,
+      "learning_rate": 4.81787557153735e-06,
+      "loss": 0.4761,
+      "step": 12036
+    },
+    {
+      "epoch": 1.566966028894963,
+      "grad_norm": 2.5279035568237305,
+      "learning_rate": 4.81578713045336e-06,
+      "loss": 0.4604,
+      "step": 12039
+    },
+    {
+      "epoch": 1.5673565013666537,
+      "grad_norm": 2.7598211765289307,
+      "learning_rate": 4.813698721550973e-06,
+      "loss": 0.4823,
+      "step": 12042
+    },
+    {
+      "epoch": 1.5677469738383443,
+      "grad_norm": 2.556950330734253,
+      "learning_rate": 4.811610345195027e-06,
+      "loss": 0.4942,
+      "step": 12045
+    },
+    {
+      "epoch": 1.5681374463100353,
+      "grad_norm": 3.494272470474243,
+      "learning_rate": 4.809522001750358e-06,
+      "loss": 0.505,
+      "step": 12048
+    },
+    {
+      "epoch": 1.5685279187817258,
+      "grad_norm": 2.75669002532959,
+      "learning_rate": 4.807433691581793e-06,
+      "loss": 0.486,
+      "step": 12051
+    },
+    {
+      "epoch": 1.5689183912534166,
+      "grad_norm": 2.629504680633545,
+      "learning_rate": 4.805345415054158e-06,
+      "loss": 0.4452,
+      "step": 12054
+    },
+    {
+      "epoch": 1.5693088637251074,
+      "grad_norm": 2.816953659057617,
+      "learning_rate": 4.803257172532267e-06,
+      "loss": 0.4767,
+      "step": 12057
+    },
+    {
+      "epoch": 1.5696993361967981,
+      "grad_norm": 2.689758777618408,
+      "learning_rate": 4.801168964380938e-06,
+      "loss": 0.5014,
+      "step": 12060
+    },
+    {
+      "epoch": 1.570089808668489,
+      "grad_norm": 2.534708261489868,
+      "learning_rate": 4.79908079096497e-06,
+      "loss": 0.4778,
+      "step": 12063
+    },
+    {
+      "epoch": 1.5704802811401795,
+      "grad_norm": 2.8629817962646484,
+      "learning_rate": 4.796992652649166e-06,
+      "loss": 0.5156,
+      "step": 12066
+    },
+    {
+      "epoch": 1.5708707536118705,
+      "grad_norm": 3.1996259689331055,
+      "learning_rate": 4.794904549798319e-06,
+      "loss": 0.4859,
+      "step": 12069
+    },
+    {
+      "epoch": 1.571261226083561,
+      "grad_norm": 2.5046546459198,
+      "learning_rate": 4.792816482777216e-06,
+      "loss": 0.5298,
+      "step": 12072
+    },
+    {
+      "epoch": 1.571651698555252,
+      "grad_norm": 2.7203421592712402,
+      "learning_rate": 4.790728451950636e-06,
+      "loss": 0.5006,
+      "step": 12075
+    },
+    {
+      "epoch": 1.5720421710269425,
+      "grad_norm": 2.6201863288879395,
+      "learning_rate": 4.7886404576833564e-06,
+      "loss": 0.5196,
+      "step": 12078
+    },
+    {
+      "epoch": 1.5724326434986333,
+      "grad_norm": 2.636479377746582,
+      "learning_rate": 4.786552500340144e-06,
+      "loss": 0.4408,
+      "step": 12081
+    },
+    {
+      "epoch": 1.572823115970324,
+      "grad_norm": 2.937016487121582,
+      "learning_rate": 4.784464580285761e-06,
+      "loss": 0.5735,
+      "step": 12084
+    },
+    {
+      "epoch": 1.5732135884420149,
+      "grad_norm": 2.6083309650421143,
+      "learning_rate": 4.782376697884962e-06,
+      "loss": 0.5457,
+      "step": 12087
+    },
+    {
+      "epoch": 1.5736040609137056,
+      "grad_norm": 2.875025987625122,
+      "learning_rate": 4.780288853502496e-06,
+      "loss": 0.5335,
+      "step": 12090
+    },
+    {
+      "epoch": 1.5739945333853962,
+      "grad_norm": 2.6229593753814697,
+      "learning_rate": 4.778201047503106e-06,
+      "loss": 0.5258,
+      "step": 12093
+    },
+    {
+      "epoch": 1.5743850058570872,
+      "grad_norm": 2.5524308681488037,
+      "learning_rate": 4.776113280251525e-06,
+      "loss": 0.4433,
+      "step": 12096
+    },
+    {
+      "epoch": 1.5747754783287777,
+      "grad_norm": 2.6454172134399414,
+      "learning_rate": 4.7740255521124825e-06,
+      "loss": 0.5138,
+      "step": 12099
+    },
+    {
+      "epoch": 1.5751659508004687,
+      "grad_norm": 2.7520358562469482,
+      "learning_rate": 4.771937863450701e-06,
+      "loss": 0.5128,
+      "step": 12102
+    },
+    {
+      "epoch": 1.5755564232721593,
+      "grad_norm": 2.627397060394287,
+      "learning_rate": 4.769850214630897e-06,
+      "loss": 0.5503,
+      "step": 12105
+    },
+    {
+      "epoch": 1.57594689574385,
+      "grad_norm": 2.6487503051757812,
+      "learning_rate": 4.767762606017775e-06,
+      "loss": 0.4488,
+      "step": 12108
+    },
+    {
+      "epoch": 1.5763373682155408,
+      "grad_norm": 2.3210136890411377,
+      "learning_rate": 4.765675037976038e-06,
+      "loss": 0.4593,
+      "step": 12111
+    },
+    {
+      "epoch": 1.5767278406872316,
+      "grad_norm": 2.7287368774414062,
+      "learning_rate": 4.763587510870377e-06,
+      "loss": 0.5627,
+      "step": 12114
+    },
+    {
+      "epoch": 1.5771183131589224,
+      "grad_norm": 2.7396085262298584,
+      "learning_rate": 4.761500025065482e-06,
+      "loss": 0.4773,
+      "step": 12117
+    },
+    {
+      "epoch": 1.577508785630613,
+      "grad_norm": 2.732445478439331,
+      "learning_rate": 4.7594125809260315e-06,
+      "loss": 0.4791,
+      "step": 12120
+    },
+    {
+      "epoch": 1.577899258102304,
+      "grad_norm": 2.82453989982605,
+      "learning_rate": 4.7573251788166954e-06,
+      "loss": 0.5459,
+      "step": 12123
+    },
+    {
+      "epoch": 1.5782897305739945,
+      "grad_norm": 2.6506190299987793,
+      "learning_rate": 4.755237819102141e-06,
+      "loss": 0.4923,
+      "step": 12126
+    },
+    {
+      "epoch": 1.5786802030456852,
+      "grad_norm": 2.771843910217285,
+      "learning_rate": 4.7531505021470245e-06,
+      "loss": 0.5828,
+      "step": 12129
+    },
+    {
+      "epoch": 1.579070675517376,
+      "grad_norm": 2.5433382987976074,
+      "learning_rate": 4.751063228315996e-06,
+      "loss": 0.5142,
+      "step": 12132
+    },
+    {
+      "epoch": 1.5794611479890668,
+      "grad_norm": 2.8748183250427246,
+      "learning_rate": 4.748975997973698e-06,
+      "loss": 0.5367,
+      "step": 12135
+    },
+    {
+      "epoch": 1.5798516204607576,
+      "grad_norm": 2.996832847595215,
+      "learning_rate": 4.746888811484765e-06,
+      "loss": 0.4731,
+      "step": 12138
+    },
+    {
+      "epoch": 1.580242092932448,
+      "grad_norm": 2.554917335510254,
+      "learning_rate": 4.744801669213822e-06,
+      "loss": 0.4605,
+      "step": 12141
+    },
+    {
+      "epoch": 1.580632565404139,
+      "grad_norm": 2.709306240081787,
+      "learning_rate": 4.742714571525492e-06,
+      "loss": 0.5183,
+      "step": 12144
+    },
+    {
+      "epoch": 1.5810230378758297,
+      "grad_norm": 3.325532913208008,
+      "learning_rate": 4.740627518784387e-06,
+      "loss": 0.4707,
+      "step": 12147
+    },
+    {
+      "epoch": 1.5814135103475206,
+      "grad_norm": 2.7227187156677246,
+      "learning_rate": 4.738540511355107e-06,
+      "loss": 0.4803,
+      "step": 12150
+    },
+    {
+      "epoch": 1.5818039828192112,
+      "grad_norm": 2.8616886138916016,
+      "learning_rate": 4.736453549602249e-06,
+      "loss": 0.5472,
+      "step": 12153
+    },
+    {
+      "epoch": 1.582194455290902,
+      "grad_norm": 2.477766513824463,
+      "learning_rate": 4.734366633890404e-06,
+      "loss": 0.5774,
+      "step": 12156
+    },
+    {
+      "epoch": 1.5825849277625927,
+      "grad_norm": 2.6389472484588623,
+      "learning_rate": 4.732279764584148e-06,
+      "loss": 0.4967,
+      "step": 12159
+    },
+    {
+      "epoch": 1.5829754002342835,
+      "grad_norm": 3.282442569732666,
+      "learning_rate": 4.730192942048054e-06,
+      "loss": 0.4934,
+      "step": 12162
+    },
+    {
+      "epoch": 1.5833658727059743,
+      "grad_norm": 2.8340752124786377,
+      "learning_rate": 4.7281061666466845e-06,
+      "loss": 0.5129,
+      "step": 12165
+    },
+    {
+      "epoch": 1.5837563451776648,
+      "grad_norm": 2.659324884414673,
+      "learning_rate": 4.726019438744596e-06,
+      "loss": 0.4756,
+      "step": 12168
+    },
+    {
+      "epoch": 1.5841468176493558,
+      "grad_norm": 2.673013925552368,
+      "learning_rate": 4.723932758706337e-06,
+      "loss": 0.4839,
+      "step": 12171
+    },
+    {
+      "epoch": 1.5845372901210464,
+      "grad_norm": 2.854623556137085,
+      "learning_rate": 4.721846126896442e-06,
+      "loss": 0.5232,
+      "step": 12174
+    },
+    {
+      "epoch": 1.5849277625927374,
+      "grad_norm": 2.7628884315490723,
+      "learning_rate": 4.7197595436794445e-06,
+      "loss": 0.4909,
+      "step": 12177
+    },
+    {
+      "epoch": 1.585318235064428,
+      "grad_norm": 2.850160598754883,
+      "learning_rate": 4.717673009419865e-06,
+      "loss": 0.4582,
+      "step": 12180
+    },
+    {
+      "epoch": 1.5857087075361187,
+      "grad_norm": 3.7766642570495605,
+      "learning_rate": 4.715586524482216e-06,
+      "loss": 0.4465,
+      "step": 12183
+    },
+    {
+      "epoch": 1.5860991800078095,
+      "grad_norm": 2.8595075607299805,
+      "learning_rate": 4.7135000892310025e-06,
+      "loss": 0.4837,
+      "step": 12186
+    },
+    {
+      "epoch": 1.5864896524795002,
+      "grad_norm": 2.8177318572998047,
+      "learning_rate": 4.711413704030722e-06,
+      "loss": 0.4759,
+      "step": 12189
+    },
+    {
+      "epoch": 1.586880124951191,
+      "grad_norm": 2.634204149246216,
+      "learning_rate": 4.709327369245861e-06,
+      "loss": 0.4633,
+      "step": 12192
+    },
+    {
+      "epoch": 1.5872705974228816,
+      "grad_norm": 2.40295147895813,
+      "learning_rate": 4.7072410852408965e-06,
+      "loss": 0.499,
+      "step": 12195
+    },
+    {
+      "epoch": 1.5876610698945726,
+      "grad_norm": 3.038196325302124,
+      "learning_rate": 4.705154852380299e-06,
+      "loss": 0.4878,
+      "step": 12198
+    },
+    {
+      "epoch": 1.5880515423662631,
+      "grad_norm": 2.4882776737213135,
+      "learning_rate": 4.7030686710285275e-06,
+      "loss": 0.4522,
+      "step": 12201
+    },
+    {
+      "epoch": 1.5884420148379539,
+      "grad_norm": 3.963737964630127,
+      "learning_rate": 4.700982541550034e-06,
+      "loss": 0.5776,
+      "step": 12204
+    },
+    {
+      "epoch": 1.5888324873096447,
+      "grad_norm": 2.4971764087677,
+      "learning_rate": 4.6988964643092635e-06,
+      "loss": 0.4437,
+      "step": 12207
+    },
+    {
+      "epoch": 1.5892229597813354,
+      "grad_norm": 2.5991008281707764,
+      "learning_rate": 4.696810439670645e-06,
+      "loss": 0.4736,
+      "step": 12210
+    },
+    {
+      "epoch": 1.5896134322530262,
+      "grad_norm": 3.21665620803833,
+      "learning_rate": 4.694724467998607e-06,
+      "loss": 0.4886,
+      "step": 12213
+    },
+    {
+      "epoch": 1.5900039047247168,
+      "grad_norm": 2.903533458709717,
+      "learning_rate": 4.692638549657561e-06,
+      "loss": 0.4829,
+      "step": 12216
+    },
+    {
+      "epoch": 1.5903943771964077,
+      "grad_norm": 2.9863076210021973,
+      "learning_rate": 4.690552685011913e-06,
+      "loss": 0.5482,
+      "step": 12219
+    },
+    {
+      "epoch": 1.5907848496680983,
+      "grad_norm": 2.6374809741973877,
+      "learning_rate": 4.688466874426062e-06,
+      "loss": 0.5568,
+      "step": 12222
+    },
+    {
+      "epoch": 1.5911753221397893,
+      "grad_norm": 3.212864398956299,
+      "learning_rate": 4.68638111826439e-06,
+      "loss": 0.4724,
+      "step": 12225
+    },
+    {
+      "epoch": 1.5915657946114798,
+      "grad_norm": 2.7329304218292236,
+      "learning_rate": 4.684295416891278e-06,
+      "loss": 0.5167,
+      "step": 12228
+    },
+    {
+      "epoch": 1.5919562670831706,
+      "grad_norm": 2.6516621112823486,
+      "learning_rate": 4.6822097706710935e-06,
+      "loss": 0.519,
+      "step": 12231
+    },
+    {
+      "epoch": 1.5923467395548614,
+      "grad_norm": 2.7666425704956055,
+      "learning_rate": 4.680124179968193e-06,
+      "loss": 0.5548,
+      "step": 12234
+    },
+    {
+      "epoch": 1.5927372120265522,
+      "grad_norm": 2.8573389053344727,
+      "learning_rate": 4.678038645146926e-06,
+      "loss": 0.5459,
+      "step": 12237
+    },
+    {
+      "epoch": 1.593127684498243,
+      "grad_norm": 2.5972604751586914,
+      "learning_rate": 4.67595316657163e-06,
+      "loss": 0.4768,
+      "step": 12240
+    },
+    {
+      "epoch": 1.5935181569699335,
+      "grad_norm": 3.3133761882781982,
+      "learning_rate": 4.673867744606633e-06,
+      "loss": 0.4992,
+      "step": 12243
+    },
+    {
+      "epoch": 1.5939086294416245,
+      "grad_norm": 2.886195659637451,
+      "learning_rate": 4.671782379616256e-06,
+      "loss": 0.5065,
+      "step": 12246
+    },
+    {
+      "epoch": 1.594299101913315,
+      "grad_norm": 2.5426437854766846,
+      "learning_rate": 4.669697071964807e-06,
+      "loss": 0.4161,
+      "step": 12249
+    },
+    {
+      "epoch": 1.5946895743850058,
+      "grad_norm": 3.0730109214782715,
+      "learning_rate": 4.667611822016584e-06,
+      "loss": 0.6377,
+      "step": 12252
+    },
+    {
+      "epoch": 1.5950800468566966,
+      "grad_norm": 2.649171829223633,
+      "learning_rate": 4.665526630135877e-06,
+      "loss": 0.5299,
+      "step": 12255
+    },
+    {
+      "epoch": 1.5954705193283873,
+      "grad_norm": 2.3111774921417236,
+      "learning_rate": 4.663441496686964e-06,
+      "loss": 0.4358,
+      "step": 12258
+    },
+    {
+      "epoch": 1.5958609918000781,
+      "grad_norm": 3.4514715671539307,
+      "learning_rate": 4.661356422034113e-06,
+      "loss": 0.572,
+      "step": 12261
+    },
+    {
+      "epoch": 1.596251464271769,
+      "grad_norm": 2.392629623413086,
+      "learning_rate": 4.659271406541584e-06,
+      "loss": 0.5273,
+      "step": 12264
+    },
+    {
+      "epoch": 1.5966419367434597,
+      "grad_norm": 2.7930562496185303,
+      "learning_rate": 4.65718645057362e-06,
+      "loss": 0.483,
+      "step": 12267
+    },
+    {
+      "epoch": 1.5970324092151502,
+      "grad_norm": 2.5676727294921875,
+      "learning_rate": 4.655101554494465e-06,
+      "loss": 0.4614,
+      "step": 12270
+    },
+    {
+      "epoch": 1.5974228816868412,
+      "grad_norm": 2.318683385848999,
+      "learning_rate": 4.653016718668342e-06,
+      "loss": 0.4482,
+      "step": 12273
+    },
+    {
+      "epoch": 1.5978133541585318,
+      "grad_norm": 2.654034376144409,
+      "learning_rate": 4.650931943459469e-06,
+      "loss": 0.4685,
+      "step": 12276
+    },
+    {
+      "epoch": 1.5982038266302225,
+      "grad_norm": 2.6656546592712402,
+      "learning_rate": 4.64884722923205e-06,
+      "loss": 0.5102,
+      "step": 12279
+    },
+    {
+      "epoch": 1.5985942991019133,
+      "grad_norm": 2.5751311779022217,
+      "learning_rate": 4.646762576350282e-06,
+      "loss": 0.4802,
+      "step": 12282
+    },
+    {
+      "epoch": 1.598984771573604,
+      "grad_norm": 2.3519175052642822,
+      "learning_rate": 4.644677985178349e-06,
+      "loss": 0.468,
+      "step": 12285
+    },
+    {
+      "epoch": 1.5993752440452949,
+      "grad_norm": 2.790675401687622,
+      "learning_rate": 4.642593456080425e-06,
+      "loss": 0.6174,
+      "step": 12288
+    },
+    {
+      "epoch": 1.5997657165169854,
+      "grad_norm": 2.661125898361206,
+      "learning_rate": 4.640508989420672e-06,
+      "loss": 0.4254,
+      "step": 12291
+    },
+    {
+      "epoch": 1.6001561889886764,
+      "grad_norm": 2.642672300338745,
+      "learning_rate": 4.638424585563241e-06,
+      "loss": 0.4429,
+      "step": 12294
+    },
+    {
+      "epoch": 1.600546661460367,
+      "grad_norm": 2.7480437755584717,
+      "learning_rate": 4.636340244872275e-06,
+      "loss": 0.5078,
+      "step": 12297
+    },
+    {
+      "epoch": 1.600937133932058,
+      "grad_norm": 3.16561222076416,
+      "learning_rate": 4.634255967711905e-06,
+      "loss": 0.5383,
+      "step": 12300
+    },
+    {
+      "epoch": 1.6013276064037485,
+      "grad_norm": 2.9352433681488037,
+      "learning_rate": 4.632171754446246e-06,
+      "loss": 0.4734,
+      "step": 12303
+    },
+    {
+      "epoch": 1.6017180788754393,
+      "grad_norm": 2.6060163974761963,
+      "learning_rate": 4.630087605439407e-06,
+      "loss": 0.466,
+      "step": 12306
+    },
+    {
+      "epoch": 1.60210855134713,
+      "grad_norm": 2.827497720718384,
+      "learning_rate": 4.628003521055486e-06,
+      "loss": 0.4221,
+      "step": 12309
+    },
+    {
+      "epoch": 1.6024990238188208,
+      "grad_norm": 2.8147799968719482,
+      "learning_rate": 4.625919501658568e-06,
+      "loss": 0.4873,
+      "step": 12312
+    },
+    {
+      "epoch": 1.6028894962905116,
+      "grad_norm": 3.2619431018829346,
+      "learning_rate": 4.623835547612726e-06,
+      "loss": 0.4946,
+      "step": 12315
+    },
+    {
+      "epoch": 1.6032799687622021,
+      "grad_norm": 3.4461402893066406,
+      "learning_rate": 4.621751659282021e-06,
+      "loss": 0.5536,
+      "step": 12318
+    },
+    {
+      "epoch": 1.6036704412338931,
+      "grad_norm": 2.9235973358154297,
+      "learning_rate": 4.619667837030508e-06,
+      "loss": 0.5901,
+      "step": 12321
+    },
+    {
+      "epoch": 1.6040609137055837,
+      "grad_norm": 2.7743449211120605,
+      "learning_rate": 4.6175840812222214e-06,
+      "loss": 0.4504,
+      "step": 12324
+    },
+    {
+      "epoch": 1.6044513861772745,
+      "grad_norm": 3.0675461292266846,
+      "learning_rate": 4.615500392221193e-06,
+      "loss": 0.5658,
+      "step": 12327
+    },
+    {
+      "epoch": 1.6048418586489652,
+      "grad_norm": 2.25282883644104,
+      "learning_rate": 4.613416770391437e-06,
+      "loss": 0.4366,
+      "step": 12330
+    },
+    {
+      "epoch": 1.605232331120656,
+      "grad_norm": 2.5219714641571045,
+      "learning_rate": 4.611333216096957e-06,
+      "loss": 0.574,
+      "step": 12333
+    },
+    {
+      "epoch": 1.6056228035923468,
+      "grad_norm": 2.765669345855713,
+      "learning_rate": 4.6092497297017475e-06,
+      "loss": 0.5322,
+      "step": 12336
+    },
+    {
+      "epoch": 1.6060132760640373,
+      "grad_norm": 2.6073596477508545,
+      "learning_rate": 4.607166311569787e-06,
+      "loss": 0.4499,
+      "step": 12339
+    },
+    {
+      "epoch": 1.6064037485357283,
+      "grad_norm": 2.747248411178589,
+      "learning_rate": 4.605082962065047e-06,
+      "loss": 0.4734,
+      "step": 12342
+    },
+    {
+      "epoch": 1.6067942210074189,
+      "grad_norm": 3.6643686294555664,
+      "learning_rate": 4.602999681551482e-06,
+      "loss": 0.4507,
+      "step": 12345
+    },
+    {
+      "epoch": 1.6071846934791099,
+      "grad_norm": 2.712887763977051,
+      "learning_rate": 4.600916470393037e-06,
+      "loss": 0.5074,
+      "step": 12348
+    },
+    {
+      "epoch": 1.6075751659508004,
+      "grad_norm": 2.606999635696411,
+      "learning_rate": 4.5988333289536444e-06,
+      "loss": 0.4703,
+      "step": 12351
+    },
+    {
+      "epoch": 1.6079656384224912,
+      "grad_norm": 3.723257064819336,
+      "learning_rate": 4.596750257597227e-06,
+      "loss": 0.5086,
+      "step": 12354
+    },
+    {
+      "epoch": 1.608356110894182,
+      "grad_norm": 3.431612253189087,
+      "learning_rate": 4.59466725668769e-06,
+      "loss": 0.5268,
+      "step": 12357
+    },
+    {
+      "epoch": 1.6087465833658727,
+      "grad_norm": 2.8018877506256104,
+      "learning_rate": 4.592584326588931e-06,
+      "loss": 0.5083,
+      "step": 12360
+    },
+    {
+      "epoch": 1.6091370558375635,
+      "grad_norm": 2.578345775604248,
+      "learning_rate": 4.590501467664834e-06,
+      "loss": 0.4861,
+      "step": 12363
+    },
+    {
+      "epoch": 1.609527528309254,
+      "grad_norm": 2.7130467891693115,
+      "learning_rate": 4.588418680279268e-06,
+      "loss": 0.5426,
+      "step": 12366
+    },
+    {
+      "epoch": 1.609918000780945,
+      "grad_norm": 2.725135564804077,
+      "learning_rate": 4.5863359647960924e-06,
+      "loss": 0.4927,
+      "step": 12369
+    },
+    {
+      "epoch": 1.6103084732526356,
+      "grad_norm": 2.6991970539093018,
+      "learning_rate": 4.584253321579155e-06,
+      "loss": 0.5302,
+      "step": 12372
+    },
+    {
+      "epoch": 1.6106989457243266,
+      "grad_norm": 2.4623939990997314,
+      "learning_rate": 4.582170750992287e-06,
+      "loss": 0.3889,
+      "step": 12375
+    },
+    {
+      "epoch": 1.6110894181960171,
+      "grad_norm": 2.5120255947113037,
+      "learning_rate": 4.580088253399311e-06,
+      "loss": 0.427,
+      "step": 12378
+    },
+    {
+      "epoch": 1.611479890667708,
+      "grad_norm": 2.3307392597198486,
+      "learning_rate": 4.578005829164032e-06,
+      "loss": 0.5037,
+      "step": 12381
+    },
+    {
+      "epoch": 1.6118703631393987,
+      "grad_norm": 2.7387290000915527,
+      "learning_rate": 4.575923478650246e-06,
+      "loss": 0.4973,
+      "step": 12384
+    },
+    {
+      "epoch": 1.6122608356110895,
+      "grad_norm": 2.8055717945098877,
+      "learning_rate": 4.573841202221739e-06,
+      "loss": 0.5121,
+      "step": 12387
+    },
+    {
+      "epoch": 1.6126513080827802,
+      "grad_norm": 3.864004373550415,
+      "learning_rate": 4.5717590002422755e-06,
+      "loss": 0.5525,
+      "step": 12390
+    },
+    {
+      "epoch": 1.6130417805544708,
+      "grad_norm": 2.713977336883545,
+      "learning_rate": 4.569676873075613e-06,
+      "loss": 0.5923,
+      "step": 12393
+    },
+    {
+      "epoch": 1.6134322530261618,
+      "grad_norm": 2.7594332695007324,
+      "learning_rate": 4.567594821085497e-06,
+      "loss": 0.5122,
+      "step": 12396
+    },
+    {
+      "epoch": 1.6138227254978523,
+      "grad_norm": 2.3967032432556152,
+      "learning_rate": 4.565512844635657e-06,
+      "loss": 0.4932,
+      "step": 12399
+    },
+    {
+      "epoch": 1.614213197969543,
+      "grad_norm": 2.6698012351989746,
+      "learning_rate": 4.563430944089807e-06,
+      "loss": 0.457,
+      "step": 12402
+    },
+    {
+      "epoch": 1.6146036704412339,
+      "grad_norm": 2.7440128326416016,
+      "learning_rate": 4.561349119811655e-06,
+      "loss": 0.4945,
+      "step": 12405
+    },
+    {
+      "epoch": 1.6149941429129246,
+      "grad_norm": 2.666889190673828,
+      "learning_rate": 4.559267372164886e-06,
+      "loss": 0.5029,
+      "step": 12408
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 2.826246738433838,
+      "learning_rate": 4.557185701513182e-06,
+      "loss": 0.4937,
+      "step": 12411
+    },
+    {
+      "epoch": 1.615775087856306,
+      "grad_norm": 2.5194458961486816,
+      "learning_rate": 4.555104108220205e-06,
+      "loss": 0.4776,
+      "step": 12414
+    },
+    {
+      "epoch": 1.616165560327997,
+      "grad_norm": 2.725764274597168,
+      "learning_rate": 4.553022592649602e-06,
+      "loss": 0.5434,
+      "step": 12417
+    },
+    {
+      "epoch": 1.6165560327996875,
+      "grad_norm": 2.692974805831909,
+      "learning_rate": 4.550941155165015e-06,
+      "loss": 0.518,
+      "step": 12420
+    },
+    {
+      "epoch": 1.6169465052713785,
+      "grad_norm": 2.79667329788208,
+      "learning_rate": 4.548859796130061e-06,
+      "loss": 0.5053,
+      "step": 12423
+    },
+    {
+      "epoch": 1.617336977743069,
+      "grad_norm": 2.6578612327575684,
+      "learning_rate": 4.546778515908352e-06,
+      "loss": 0.4535,
+      "step": 12426
+    },
+    {
+      "epoch": 1.6177274502147598,
+      "grad_norm": 2.6429173946380615,
+      "learning_rate": 4.544697314863483e-06,
+      "loss": 0.4714,
+      "step": 12429
+    },
+    {
+      "epoch": 1.6181179226864506,
+      "grad_norm": 2.6964876651763916,
+      "learning_rate": 4.542616193359035e-06,
+      "loss": 0.5593,
+      "step": 12432
+    },
+    {
+      "epoch": 1.6185083951581414,
+      "grad_norm": 2.86681866645813,
+      "learning_rate": 4.540535151758575e-06,
+      "loss": 0.5782,
+      "step": 12435
+    },
+    {
+      "epoch": 1.6188988676298322,
+      "grad_norm": 4.218873977661133,
+      "learning_rate": 4.538454190425658e-06,
+      "loss": 0.5248,
+      "step": 12438
+    },
+    {
+      "epoch": 1.6192893401015227,
+      "grad_norm": 4.104538917541504,
+      "learning_rate": 4.5363733097238235e-06,
+      "loss": 0.5671,
+      "step": 12441
+    },
+    {
+      "epoch": 1.6196798125732137,
+      "grad_norm": 2.703684091567993,
+      "learning_rate": 4.534292510016597e-06,
+      "loss": 0.522,
+      "step": 12444
+    },
+    {
+      "epoch": 1.6200702850449042,
+      "grad_norm": 3.4295432567596436,
+      "learning_rate": 4.532211791667488e-06,
+      "loss": 0.5239,
+      "step": 12447
+    },
+    {
+      "epoch": 1.6204607575165952,
+      "grad_norm": 2.69512939453125,
+      "learning_rate": 4.5301311550399966e-06,
+      "loss": 0.4816,
+      "step": 12450
+    },
+    {
+      "epoch": 1.6208512299882858,
+      "grad_norm": 2.700019121170044,
+      "learning_rate": 4.5280506004976025e-06,
+      "loss": 0.4808,
+      "step": 12453
+    },
+    {
+      "epoch": 1.6212417024599766,
+      "grad_norm": 2.682190179824829,
+      "learning_rate": 4.525970128403777e-06,
+      "loss": 0.5554,
+      "step": 12456
+    },
+    {
+      "epoch": 1.6216321749316673,
+      "grad_norm": 2.4642701148986816,
+      "learning_rate": 4.523889739121971e-06,
+      "loss": 0.4808,
+      "step": 12459
+    },
+    {
+      "epoch": 1.622022647403358,
+      "grad_norm": 2.6819307804107666,
+      "learning_rate": 4.521809433015627e-06,
+      "loss": 0.5932,
+      "step": 12462
+    },
+    {
+      "epoch": 1.6224131198750489,
+      "grad_norm": 2.5704028606414795,
+      "learning_rate": 4.51972921044817e-06,
+      "loss": 0.5527,
+      "step": 12465
+    },
+    {
+      "epoch": 1.6228035923467394,
+      "grad_norm": 2.388164758682251,
+      "learning_rate": 4.517649071783008e-06,
+      "loss": 0.5643,
+      "step": 12468
+    },
+    {
+      "epoch": 1.6231940648184304,
+      "grad_norm": 2.3138227462768555,
+      "learning_rate": 4.51556901738354e-06,
+      "loss": 0.4487,
+      "step": 12471
+    },
+    {
+      "epoch": 1.623584537290121,
+      "grad_norm": 2.4271912574768066,
+      "learning_rate": 4.513489047613144e-06,
+      "loss": 0.465,
+      "step": 12474
+    },
+    {
+      "epoch": 1.6239750097618117,
+      "grad_norm": 2.532554864883423,
+      "learning_rate": 4.511409162835188e-06,
+      "loss": 0.5073,
+      "step": 12477
+    },
+    {
+      "epoch": 1.6243654822335025,
+      "grad_norm": 2.848936080932617,
+      "learning_rate": 4.509329363413023e-06,
+      "loss": 0.4336,
+      "step": 12480
+    },
+    {
+      "epoch": 1.6247559547051933,
+      "grad_norm": 2.899810791015625,
+      "learning_rate": 4.507249649709988e-06,
+      "loss": 0.5284,
+      "step": 12483
+    },
+    {
+      "epoch": 1.625146427176884,
+      "grad_norm": 2.649650812149048,
+      "learning_rate": 4.505170022089401e-06,
+      "loss": 0.506,
+      "step": 12486
+    },
+    {
+      "epoch": 1.6255368996485746,
+      "grad_norm": 2.40224552154541,
+      "learning_rate": 4.503090480914571e-06,
+      "loss": 0.4498,
+      "step": 12489
+    },
+    {
+      "epoch": 1.6259273721202656,
+      "grad_norm": 2.5746147632598877,
+      "learning_rate": 4.501011026548789e-06,
+      "loss": 0.5017,
+      "step": 12492
+    },
+    {
+      "epoch": 1.6263178445919562,
+      "grad_norm": 2.6533167362213135,
+      "learning_rate": 4.49893165935533e-06,
+      "loss": 0.5273,
+      "step": 12495
+    },
+    {
+      "epoch": 1.6267083170636472,
+      "grad_norm": 4.123754501342773,
+      "learning_rate": 4.496852379697456e-06,
+      "loss": 0.5213,
+      "step": 12498
+    },
+    {
+      "epoch": 1.6270987895353377,
+      "grad_norm": 3.875533103942871,
+      "learning_rate": 4.494773187938413e-06,
+      "loss": 0.5422,
+      "step": 12501
+    },
+    {
+      "epoch": 1.6274892620070285,
+      "grad_norm": 3.554651975631714,
+      "learning_rate": 4.4926940844414316e-06,
+      "loss": 0.4757,
+      "step": 12504
+    },
+    {
+      "epoch": 1.6278797344787193,
+      "grad_norm": 2.468736171722412,
+      "learning_rate": 4.490615069569727e-06,
+      "loss": 0.5176,
+      "step": 12507
+    },
+    {
+      "epoch": 1.62827020695041,
+      "grad_norm": 3.26057767868042,
+      "learning_rate": 4.488536143686497e-06,
+      "loss": 0.522,
+      "step": 12510
+    },
+    {
+      "epoch": 1.6286606794221008,
+      "grad_norm": 2.696108102798462,
+      "learning_rate": 4.486457307154927e-06,
+      "loss": 0.5466,
+      "step": 12513
+    },
+    {
+      "epoch": 1.6290511518937913,
+      "grad_norm": 2.7262580394744873,
+      "learning_rate": 4.4843785603381855e-06,
+      "loss": 0.5753,
+      "step": 12516
+    },
+    {
+      "epoch": 1.6294416243654823,
+      "grad_norm": 2.864877462387085,
+      "learning_rate": 4.482299903599424e-06,
+      "loss": 0.457,
+      "step": 12519
+    },
+    {
+      "epoch": 1.629832096837173,
+      "grad_norm": 2.8265585899353027,
+      "learning_rate": 4.48022133730178e-06,
+      "loss": 0.5091,
+      "step": 12522
+    },
+    {
+      "epoch": 1.630222569308864,
+      "grad_norm": 3.3130998611450195,
+      "learning_rate": 4.478142861808375e-06,
+      "loss": 0.4634,
+      "step": 12525
+    },
+    {
+      "epoch": 1.6306130417805544,
+      "grad_norm": 2.6155717372894287,
+      "learning_rate": 4.476064477482316e-06,
+      "loss": 0.4637,
+      "step": 12528
+    },
+    {
+      "epoch": 1.6310035142522452,
+      "grad_norm": 2.8541154861450195,
+      "learning_rate": 4.4739861846866885e-06,
+      "loss": 0.5217,
+      "step": 12531
+    },
+    {
+      "epoch": 1.631393986723936,
+      "grad_norm": 2.5578043460845947,
+      "learning_rate": 4.47190798378457e-06,
+      "loss": 0.5005,
+      "step": 12534
+    },
+    {
+      "epoch": 1.6317844591956268,
+      "grad_norm": 3.508741617202759,
+      "learning_rate": 4.469829875139014e-06,
+      "loss": 0.512,
+      "step": 12537
+    },
+    {
+      "epoch": 1.6321749316673175,
+      "grad_norm": 2.5680432319641113,
+      "learning_rate": 4.467751859113064e-06,
+      "loss": 0.5627,
+      "step": 12540
+    },
+    {
+      "epoch": 1.632565404139008,
+      "grad_norm": 2.6496567726135254,
+      "learning_rate": 4.465673936069746e-06,
+      "loss": 0.4628,
+      "step": 12543
+    },
+    {
+      "epoch": 1.632955876610699,
+      "grad_norm": 2.5644404888153076,
+      "learning_rate": 4.463596106372066e-06,
+      "loss": 0.5009,
+      "step": 12546
+    },
+    {
+      "epoch": 1.6333463490823896,
+      "grad_norm": 2.2369980812072754,
+      "learning_rate": 4.461518370383017e-06,
+      "loss": 0.4671,
+      "step": 12549
+    },
+    {
+      "epoch": 1.6337368215540804,
+      "grad_norm": 2.7351465225219727,
+      "learning_rate": 4.459440728465578e-06,
+      "loss": 0.5614,
+      "step": 12552
+    },
+    {
+      "epoch": 1.6341272940257712,
+      "grad_norm": 2.928203821182251,
+      "learning_rate": 4.4573631809827045e-06,
+      "loss": 0.5391,
+      "step": 12555
+    },
+    {
+      "epoch": 1.634517766497462,
+      "grad_norm": 2.912785291671753,
+      "learning_rate": 4.4552857282973435e-06,
+      "loss": 0.4928,
+      "step": 12558
+    },
+    {
+      "epoch": 1.6349082389691527,
+      "grad_norm": 3.1030282974243164,
+      "learning_rate": 4.453208370772417e-06,
+      "loss": 0.5712,
+      "step": 12561
+    },
+    {
+      "epoch": 1.6352987114408433,
+      "grad_norm": 2.6347246170043945,
+      "learning_rate": 4.45113110877084e-06,
+      "loss": 0.5575,
+      "step": 12564
+    },
+    {
+      "epoch": 1.6356891839125343,
+      "grad_norm": 2.7265217304229736,
+      "learning_rate": 4.449053942655503e-06,
+      "loss": 0.5683,
+      "step": 12567
+    },
+    {
+      "epoch": 1.6360796563842248,
+      "grad_norm": 2.5423009395599365,
+      "learning_rate": 4.446976872789284e-06,
+      "loss": 0.4451,
+      "step": 12570
+    },
+    {
+      "epoch": 1.6364701288559158,
+      "grad_norm": 2.4595041275024414,
+      "learning_rate": 4.444899899535042e-06,
+      "loss": 0.4713,
+      "step": 12573
+    },
+    {
+      "epoch": 1.6368606013276064,
+      "grad_norm": 2.64188814163208,
+      "learning_rate": 4.442823023255619e-06,
+      "loss": 0.4481,
+      "step": 12576
+    },
+    {
+      "epoch": 1.6372510737992971,
+      "grad_norm": 2.5329110622406006,
+      "learning_rate": 4.440746244313842e-06,
+      "loss": 0.5019,
+      "step": 12579
+    },
+    {
+      "epoch": 1.637641546270988,
+      "grad_norm": 2.4973838329315186,
+      "learning_rate": 4.43866956307252e-06,
+      "loss": 0.4662,
+      "step": 12582
+    },
+    {
+      "epoch": 1.6380320187426787,
+      "grad_norm": 2.651660203933716,
+      "learning_rate": 4.436592979894445e-06,
+      "loss": 0.4375,
+      "step": 12585
+    },
+    {
+      "epoch": 1.6384224912143694,
+      "grad_norm": 2.4576289653778076,
+      "learning_rate": 4.4345164951423895e-06,
+      "loss": 0.4207,
+      "step": 12588
+    },
+    {
+      "epoch": 1.63881296368606,
+      "grad_norm": 3.1014697551727295,
+      "learning_rate": 4.432440109179113e-06,
+      "loss": 0.5241,
+      "step": 12591
+    },
+    {
+      "epoch": 1.639203436157751,
+      "grad_norm": 2.6351943016052246,
+      "learning_rate": 4.430363822367357e-06,
+      "loss": 0.5933,
+      "step": 12594
+    },
+    {
+      "epoch": 1.6395939086294415,
+      "grad_norm": 2.748256206512451,
+      "learning_rate": 4.428287635069841e-06,
+      "loss": 0.5146,
+      "step": 12597
+    },
+    {
+      "epoch": 1.6399843811011323,
+      "grad_norm": 2.712522506713867,
+      "learning_rate": 4.426211547649274e-06,
+      "loss": 0.5062,
+      "step": 12600
+    },
+    {
+      "epoch": 1.640374853572823,
+      "grad_norm": 2.814810276031494,
+      "learning_rate": 4.42413556046834e-06,
+      "loss": 0.5229,
+      "step": 12603
+    },
+    {
+      "epoch": 1.6407653260445139,
+      "grad_norm": 2.7721574306488037,
+      "learning_rate": 4.422059673889714e-06,
+      "loss": 0.4604,
+      "step": 12606
+    },
+    {
+      "epoch": 1.6411557985162046,
+      "grad_norm": 2.700059175491333,
+      "learning_rate": 4.419983888276047e-06,
+      "loss": 0.5371,
+      "step": 12609
+    },
+    {
+      "epoch": 1.6415462709878954,
+      "grad_norm": 2.785081624984741,
+      "learning_rate": 4.417908203989975e-06,
+      "loss": 0.5373,
+      "step": 12612
+    },
+    {
+      "epoch": 1.6419367434595862,
+      "grad_norm": 2.8077304363250732,
+      "learning_rate": 4.415832621394116e-06,
+      "loss": 0.5162,
+      "step": 12615
+    },
+    {
+      "epoch": 1.6423272159312767,
+      "grad_norm": 2.6027612686157227,
+      "learning_rate": 4.413757140851067e-06,
+      "loss": 0.5698,
+      "step": 12618
+    },
+    {
+      "epoch": 1.6427176884029677,
+      "grad_norm": 2.704460620880127,
+      "learning_rate": 4.411681762723415e-06,
+      "loss": 0.476,
+      "step": 12621
+    },
+    {
+      "epoch": 1.6431081608746583,
+      "grad_norm": 2.906745672225952,
+      "learning_rate": 4.409606487373718e-06,
+      "loss": 0.4695,
+      "step": 12624
+    },
+    {
+      "epoch": 1.643498633346349,
+      "grad_norm": 2.890692710876465,
+      "learning_rate": 4.407531315164527e-06,
+      "loss": 0.5786,
+      "step": 12627
+    },
+    {
+      "epoch": 1.6438891058180398,
+      "grad_norm": 2.488058567047119,
+      "learning_rate": 4.4054562464583705e-06,
+      "loss": 0.4332,
+      "step": 12630
+    },
+    {
+      "epoch": 1.6442795782897306,
+      "grad_norm": 2.7957355976104736,
+      "learning_rate": 4.403381281617755e-06,
+      "loss": 0.4859,
+      "step": 12633
+    },
+    {
+      "epoch": 1.6446700507614214,
+      "grad_norm": 2.960820436477661,
+      "learning_rate": 4.401306421005176e-06,
+      "loss": 0.4931,
+      "step": 12636
+    },
+    {
+      "epoch": 1.645060523233112,
+      "grad_norm": 2.780045986175537,
+      "learning_rate": 4.399231664983104e-06,
+      "loss": 0.5052,
+      "step": 12639
+    },
+    {
+      "epoch": 1.645450995704803,
+      "grad_norm": 2.800703525543213,
+      "learning_rate": 4.3971570139139975e-06,
+      "loss": 0.57,
+      "step": 12642
+    },
+    {
+      "epoch": 1.6458414681764935,
+      "grad_norm": 3.8522229194641113,
+      "learning_rate": 4.395082468160291e-06,
+      "loss": 0.5197,
+      "step": 12645
+    },
+    {
+      "epoch": 1.6462319406481845,
+      "grad_norm": 2.728604793548584,
+      "learning_rate": 4.393008028084407e-06,
+      "loss": 0.4438,
+      "step": 12648
+    },
+    {
+      "epoch": 1.646622413119875,
+      "grad_norm": 2.7166476249694824,
+      "learning_rate": 4.390933694048742e-06,
+      "loss": 0.5547,
+      "step": 12651
+    },
+    {
+      "epoch": 1.6470128855915658,
+      "grad_norm": 2.5482025146484375,
+      "learning_rate": 4.3888594664156795e-06,
+      "loss": 0.5274,
+      "step": 12654
+    },
+    {
+      "epoch": 1.6474033580632566,
+      "grad_norm": 3.0686612129211426,
+      "learning_rate": 4.386785345547584e-06,
+      "loss": 0.5536,
+      "step": 12657
+    },
+    {
+      "epoch": 1.6477938305349473,
+      "grad_norm": 2.8095195293426514,
+      "learning_rate": 4.384711331806797e-06,
+      "loss": 0.4901,
+      "step": 12660
+    },
+    {
+      "epoch": 1.648184303006638,
+      "grad_norm": 2.81028151512146,
+      "learning_rate": 4.3826374255556476e-06,
+      "loss": 0.515,
+      "step": 12663
+    },
+    {
+      "epoch": 1.6485747754783286,
+      "grad_norm": 2.4701220989227295,
+      "learning_rate": 4.38056362715644e-06,
+      "loss": 0.5399,
+      "step": 12666
+    },
+    {
+      "epoch": 1.6489652479500196,
+      "grad_norm": 2.5714948177337646,
+      "learning_rate": 4.378489936971463e-06,
+      "loss": 0.5045,
+      "step": 12669
+    },
+    {
+      "epoch": 1.6493557204217102,
+      "grad_norm": 2.7451059818267822,
+      "learning_rate": 4.376416355362989e-06,
+      "loss": 0.4809,
+      "step": 12672
+    },
+    {
+      "epoch": 1.649746192893401,
+      "grad_norm": 2.625004529953003,
+      "learning_rate": 4.3743428826932635e-06,
+      "loss": 0.5552,
+      "step": 12675
+    },
+    {
+      "epoch": 1.6501366653650917,
+      "grad_norm": 2.387716770172119,
+      "learning_rate": 4.37226951932452e-06,
+      "loss": 0.4466,
+      "step": 12678
+    },
+    {
+      "epoch": 1.6505271378367825,
+      "grad_norm": 2.606682777404785,
+      "learning_rate": 4.370196265618973e-06,
+      "loss": 0.5303,
+      "step": 12681
+    },
+    {
+      "epoch": 1.6509176103084733,
+      "grad_norm": 2.3871326446533203,
+      "learning_rate": 4.368123121938812e-06,
+      "loss": 0.4816,
+      "step": 12684
+    },
+    {
+      "epoch": 1.6513080827801638,
+      "grad_norm": 3.2255170345306396,
+      "learning_rate": 4.3660500886462105e-06,
+      "loss": 0.5146,
+      "step": 12687
+    },
+    {
+      "epoch": 1.6516985552518548,
+      "grad_norm": 2.7184829711914062,
+      "learning_rate": 4.3639771661033275e-06,
+      "loss": 0.5417,
+      "step": 12690
+    },
+    {
+      "epoch": 1.6520890277235454,
+      "grad_norm": 3.5075788497924805,
+      "learning_rate": 4.361904354672296e-06,
+      "loss": 0.4598,
+      "step": 12693
+    },
+    {
+      "epoch": 1.6524795001952364,
+      "grad_norm": 2.910029649734497,
+      "learning_rate": 4.3598316547152295e-06,
+      "loss": 0.5129,
+      "step": 12696
+    },
+    {
+      "epoch": 1.652869972666927,
+      "grad_norm": 2.9071853160858154,
+      "learning_rate": 4.357759066594228e-06,
+      "loss": 0.5485,
+      "step": 12699
+    },
+    {
+      "epoch": 1.6532604451386177,
+      "grad_norm": 2.824410915374756,
+      "learning_rate": 4.3556865906713654e-06,
+      "loss": 0.5358,
+      "step": 12702
+    },
+    {
+      "epoch": 1.6536509176103085,
+      "grad_norm": 2.680872678756714,
+      "learning_rate": 4.3536142273087005e-06,
+      "loss": 0.5531,
+      "step": 12705
+    },
+    {
+      "epoch": 1.6540413900819992,
+      "grad_norm": 2.904576063156128,
+      "learning_rate": 4.351541976868271e-06,
+      "loss": 0.523,
+      "step": 12708
+    },
+    {
+      "epoch": 1.65443186255369,
+      "grad_norm": 2.6715736389160156,
+      "learning_rate": 4.349469839712093e-06,
+      "loss": 0.4883,
+      "step": 12711
+    },
+    {
+      "epoch": 1.6548223350253806,
+      "grad_norm": 2.5224480628967285,
+      "learning_rate": 4.347397816202165e-06,
+      "loss": 0.509,
+      "step": 12714
+    },
+    {
+      "epoch": 1.6552128074970716,
+      "grad_norm": 3.416038990020752,
+      "learning_rate": 4.345325906700467e-06,
+      "loss": 0.4453,
+      "step": 12717
+    },
+    {
+      "epoch": 1.655603279968762,
+      "grad_norm": 2.9145963191986084,
+      "learning_rate": 4.343254111568954e-06,
+      "loss": 0.5879,
+      "step": 12720
+    },
+    {
+      "epoch": 1.655993752440453,
+      "grad_norm": 3.9935879707336426,
+      "learning_rate": 4.341182431169568e-06,
+      "loss": 0.5139,
+      "step": 12723
+    },
+    {
+      "epoch": 1.6563842249121437,
+      "grad_norm": 4.118599891662598,
+      "learning_rate": 4.339110865864225e-06,
+      "loss": 0.5033,
+      "step": 12726
+    },
+    {
+      "epoch": 1.6567746973838344,
+      "grad_norm": 2.687406301498413,
+      "learning_rate": 4.337039416014821e-06,
+      "loss": 0.5637,
+      "step": 12729
+    },
+    {
+      "epoch": 1.6571651698555252,
+      "grad_norm": 2.5384199619293213,
+      "learning_rate": 4.334968081983238e-06,
+      "loss": 0.5492,
+      "step": 12732
+    },
+    {
+      "epoch": 1.657555642327216,
+      "grad_norm": 2.816277265548706,
+      "learning_rate": 4.3328968641313326e-06,
+      "loss": 0.4586,
+      "step": 12735
+    },
+    {
+      "epoch": 1.6579461147989067,
+      "grad_norm": 2.6040713787078857,
+      "learning_rate": 4.330825762820942e-06,
+      "loss": 0.4576,
+      "step": 12738
+    },
+    {
+      "epoch": 1.6583365872705973,
+      "grad_norm": 2.7248170375823975,
+      "learning_rate": 4.3287547784138815e-06,
+      "loss": 0.4663,
+      "step": 12741
+    },
+    {
+      "epoch": 1.6587270597422883,
+      "grad_norm": 2.8028006553649902,
+      "learning_rate": 4.32668391127195e-06,
+      "loss": 0.5485,
+      "step": 12744
+    },
+    {
+      "epoch": 1.6591175322139788,
+      "grad_norm": 2.470500946044922,
+      "learning_rate": 4.324613161756923e-06,
+      "loss": 0.4919,
+      "step": 12747
+    },
+    {
+      "epoch": 1.6595080046856696,
+      "grad_norm": 2.61730694770813,
+      "learning_rate": 4.322542530230556e-06,
+      "loss": 0.4849,
+      "step": 12750
+    },
+    {
+      "epoch": 1.6598984771573604,
+      "grad_norm": 2.686258316040039,
+      "learning_rate": 4.320472017054584e-06,
+      "loss": 0.4206,
+      "step": 12753
+    },
+    {
+      "epoch": 1.6602889496290512,
+      "grad_norm": 2.6718552112579346,
+      "learning_rate": 4.318401622590719e-06,
+      "loss": 0.5302,
+      "step": 12756
+    },
+    {
+      "epoch": 1.660679422100742,
+      "grad_norm": 2.5929765701293945,
+      "learning_rate": 4.316331347200659e-06,
+      "loss": 0.4858,
+      "step": 12759
+    },
+    {
+      "epoch": 1.6610698945724325,
+      "grad_norm": 2.6052663326263428,
+      "learning_rate": 4.314261191246073e-06,
+      "loss": 0.4703,
+      "step": 12762
+    },
+    {
+      "epoch": 1.6614603670441235,
+      "grad_norm": 2.5814874172210693,
+      "learning_rate": 4.312191155088616e-06,
+      "loss": 0.4815,
+      "step": 12765
+    },
+    {
+      "epoch": 1.661850839515814,
+      "grad_norm": 2.9272522926330566,
+      "learning_rate": 4.310121239089915e-06,
+      "loss": 0.5128,
+      "step": 12768
+    },
+    {
+      "epoch": 1.662241311987505,
+      "grad_norm": 3.3689401149749756,
+      "learning_rate": 4.308051443611582e-06,
+      "loss": 0.4717,
+      "step": 12771
+    },
+    {
+      "epoch": 1.6626317844591956,
+      "grad_norm": 2.8045191764831543,
+      "learning_rate": 4.305981769015207e-06,
+      "loss": 0.4954,
+      "step": 12774
+    },
+    {
+      "epoch": 1.6630222569308863,
+      "grad_norm": 2.869739294052124,
+      "learning_rate": 4.303912215662359e-06,
+      "loss": 0.4864,
+      "step": 12777
+    },
+    {
+      "epoch": 1.6634127294025771,
+      "grad_norm": 2.9662373065948486,
+      "learning_rate": 4.30184278391458e-06,
+      "loss": 0.5554,
+      "step": 12780
+    },
+    {
+      "epoch": 1.663803201874268,
+      "grad_norm": 2.603210926055908,
+      "learning_rate": 4.299773474133398e-06,
+      "loss": 0.467,
+      "step": 12783
+    },
+    {
+      "epoch": 1.6641936743459587,
+      "grad_norm": 2.6853482723236084,
+      "learning_rate": 4.297704286680319e-06,
+      "loss": 0.5312,
+      "step": 12786
+    },
+    {
+      "epoch": 1.6645841468176492,
+      "grad_norm": 2.606921672821045,
+      "learning_rate": 4.295635221916823e-06,
+      "loss": 0.5665,
+      "step": 12789
+    },
+    {
+      "epoch": 1.6649746192893402,
+      "grad_norm": 2.5418643951416016,
+      "learning_rate": 4.293566280204371e-06,
+      "loss": 0.4628,
+      "step": 12792
+    },
+    {
+      "epoch": 1.6653650917610308,
+      "grad_norm": 3.40142822265625,
+      "learning_rate": 4.2914974619044045e-06,
+      "loss": 0.487,
+      "step": 12795
+    },
+    {
+      "epoch": 1.6657555642327218,
+      "grad_norm": 2.476050853729248,
+      "learning_rate": 4.289428767378341e-06,
+      "loss": 0.5231,
+      "step": 12798
+    },
+    {
+      "epoch": 1.6661460367044123,
+      "grad_norm": 2.9447269439697266,
+      "learning_rate": 4.287360196987578e-06,
+      "loss": 0.5296,
+      "step": 12801
+    },
+    {
+      "epoch": 1.666536509176103,
+      "grad_norm": 3.3223366737365723,
+      "learning_rate": 4.2852917510934876e-06,
+      "loss": 0.4443,
+      "step": 12804
+    },
+    {
+      "epoch": 1.6669269816477938,
+      "grad_norm": 2.9464173316955566,
+      "learning_rate": 4.283223430057425e-06,
+      "loss": 0.5343,
+      "step": 12807
+    },
+    {
+      "epoch": 1.6673174541194846,
+      "grad_norm": 3.260638475418091,
+      "learning_rate": 4.281155234240722e-06,
+      "loss": 0.4831,
+      "step": 12810
+    },
+    {
+      "epoch": 1.6677079265911754,
+      "grad_norm": 2.604071617126465,
+      "learning_rate": 4.279087164004686e-06,
+      "loss": 0.5149,
+      "step": 12813
+    },
+    {
+      "epoch": 1.668098399062866,
+      "grad_norm": 2.497521162033081,
+      "learning_rate": 4.277019219710607e-06,
+      "loss": 0.4481,
+      "step": 12816
+    },
+    {
+      "epoch": 1.668488871534557,
+      "grad_norm": 2.61600399017334,
+      "learning_rate": 4.274951401719748e-06,
+      "loss": 0.5209,
+      "step": 12819
+    },
+    {
+      "epoch": 1.6688793440062475,
+      "grad_norm": 2.627828359603882,
+      "learning_rate": 4.272883710393356e-06,
+      "loss": 0.538,
+      "step": 12822
+    },
+    {
+      "epoch": 1.6692698164779383,
+      "grad_norm": 2.7958621978759766,
+      "learning_rate": 4.270816146092649e-06,
+      "loss": 0.4669,
+      "step": 12825
+    },
+    {
+      "epoch": 1.669660288949629,
+      "grad_norm": 2.8905887603759766,
+      "learning_rate": 4.268748709178828e-06,
+      "loss": 0.554,
+      "step": 12828
+    },
+    {
+      "epoch": 1.6700507614213198,
+      "grad_norm": 2.683310031890869,
+      "learning_rate": 4.2666814000130685e-06,
+      "loss": 0.518,
+      "step": 12831
+    },
+    {
+      "epoch": 1.6704412338930106,
+      "grad_norm": 3.2548582553863525,
+      "learning_rate": 4.264614218956525e-06,
+      "loss": 0.514,
+      "step": 12834
+    },
+    {
+      "epoch": 1.6708317063647011,
+      "grad_norm": 2.7682833671569824,
+      "learning_rate": 4.262547166370333e-06,
+      "loss": 0.5013,
+      "step": 12837
+    },
+    {
+      "epoch": 1.6712221788363921,
+      "grad_norm": 2.7718143463134766,
+      "learning_rate": 4.2604802426155975e-06,
+      "loss": 0.5049,
+      "step": 12840
+    },
+    {
+      "epoch": 1.6716126513080827,
+      "grad_norm": 2.5910987854003906,
+      "learning_rate": 4.258413448053409e-06,
+      "loss": 0.5122,
+      "step": 12843
+    },
+    {
+      "epoch": 1.6720031237797737,
+      "grad_norm": 2.675654172897339,
+      "learning_rate": 4.25634678304483e-06,
+      "loss": 0.5298,
+      "step": 12846
+    },
+    {
+      "epoch": 1.6723935962514642,
+      "grad_norm": 2.951941728591919,
+      "learning_rate": 4.254280247950904e-06,
+      "loss": 0.5185,
+      "step": 12849
+    },
+    {
+      "epoch": 1.672784068723155,
+      "grad_norm": 3.080920696258545,
+      "learning_rate": 4.252213843132651e-06,
+      "loss": 0.4948,
+      "step": 12852
+    },
+    {
+      "epoch": 1.6731745411948458,
+      "grad_norm": 2.5052084922790527,
+      "learning_rate": 4.250147568951062e-06,
+      "loss": 0.4856,
+      "step": 12855
+    },
+    {
+      "epoch": 1.6735650136665365,
+      "grad_norm": 2.8432629108428955,
+      "learning_rate": 4.2480814257671195e-06,
+      "loss": 0.4367,
+      "step": 12858
+    },
+    {
+      "epoch": 1.6739554861382273,
+      "grad_norm": 2.4311513900756836,
+      "learning_rate": 4.24601541394177e-06,
+      "loss": 0.5279,
+      "step": 12861
+    },
+    {
+      "epoch": 1.6743459586099179,
+      "grad_norm": 2.772369861602783,
+      "learning_rate": 4.243949533835941e-06,
+      "loss": 0.5058,
+      "step": 12864
+    },
+    {
+      "epoch": 1.6747364310816089,
+      "grad_norm": 2.680986166000366,
+      "learning_rate": 4.241883785810538e-06,
+      "loss": 0.5034,
+      "step": 12867
+    },
+    {
+      "epoch": 1.6751269035532994,
+      "grad_norm": 2.6189749240875244,
+      "learning_rate": 4.239818170226442e-06,
+      "loss": 0.4952,
+      "step": 12870
+    },
+    {
+      "epoch": 1.6755173760249904,
+      "grad_norm": 2.844604253768921,
+      "learning_rate": 4.237752687444514e-06,
+      "loss": 0.5086,
+      "step": 12873
+    },
+    {
+      "epoch": 1.675907848496681,
+      "grad_norm": 2.81457781791687,
+      "learning_rate": 4.235687337825586e-06,
+      "loss": 0.4829,
+      "step": 12876
+    },
+    {
+      "epoch": 1.6762983209683717,
+      "grad_norm": 3.1727747917175293,
+      "learning_rate": 4.233622121730474e-06,
+      "loss": 0.5165,
+      "step": 12879
+    },
+    {
+      "epoch": 1.6766887934400625,
+      "grad_norm": 2.7962770462036133,
+      "learning_rate": 4.231557039519965e-06,
+      "loss": 0.4056,
+      "step": 12882
+    },
+    {
+      "epoch": 1.6770792659117533,
+      "grad_norm": 2.9042179584503174,
+      "learning_rate": 4.229492091554823e-06,
+      "loss": 0.6229,
+      "step": 12885
+    },
+    {
+      "epoch": 1.677469738383444,
+      "grad_norm": 2.709332227706909,
+      "learning_rate": 4.227427278195794e-06,
+      "loss": 0.4915,
+      "step": 12888
+    },
+    {
+      "epoch": 1.6778602108551346,
+      "grad_norm": 2.407127857208252,
+      "learning_rate": 4.225362599803592e-06,
+      "loss": 0.445,
+      "step": 12891
+    },
+    {
+      "epoch": 1.6782506833268256,
+      "grad_norm": 2.5972049236297607,
+      "learning_rate": 4.2232980567389156e-06,
+      "loss": 0.4591,
+      "step": 12894
+    },
+    {
+      "epoch": 1.6786411557985161,
+      "grad_norm": 3.5093605518341064,
+      "learning_rate": 4.221233649362432e-06,
+      "loss": 0.5018,
+      "step": 12897
+    },
+    {
+      "epoch": 1.679031628270207,
+      "grad_norm": 2.601149320602417,
+      "learning_rate": 4.219169378034795e-06,
+      "loss": 0.4992,
+      "step": 12900
+    },
+    {
+      "epoch": 1.6794221007418977,
+      "grad_norm": 2.5034475326538086,
+      "learning_rate": 4.217105243116623e-06,
+      "loss": 0.4433,
+      "step": 12903
+    },
+    {
+      "epoch": 1.6798125732135885,
+      "grad_norm": 2.8355612754821777,
+      "learning_rate": 4.21504124496852e-06,
+      "loss": 0.4554,
+      "step": 12906
+    },
+    {
+      "epoch": 1.6802030456852792,
+      "grad_norm": 2.9951822757720947,
+      "learning_rate": 4.212977383951059e-06,
+      "loss": 0.5543,
+      "step": 12909
+    },
+    {
+      "epoch": 1.6805935181569698,
+      "grad_norm": 3.2354860305786133,
+      "learning_rate": 4.210913660424793e-06,
+      "loss": 0.4783,
+      "step": 12912
+    },
+    {
+      "epoch": 1.6809839906286608,
+      "grad_norm": 2.7228434085845947,
+      "learning_rate": 4.208850074750251e-06,
+      "loss": 0.4661,
+      "step": 12915
+    },
+    {
+      "epoch": 1.6813744631003513,
+      "grad_norm": 3.545387029647827,
+      "learning_rate": 4.206786627287936e-06,
+      "loss": 0.5935,
+      "step": 12918
+    },
+    {
+      "epoch": 1.6817649355720423,
+      "grad_norm": 2.7536470890045166,
+      "learning_rate": 4.20472331839833e-06,
+      "loss": 0.4677,
+      "step": 12921
+    },
+    {
+      "epoch": 1.6821554080437329,
+      "grad_norm": 2.6740031242370605,
+      "learning_rate": 4.202660148441886e-06,
+      "loss": 0.5313,
+      "step": 12924
+    },
+    {
+      "epoch": 1.6825458805154236,
+      "grad_norm": 2.6446104049682617,
+      "learning_rate": 4.200597117779038e-06,
+      "loss": 0.4866,
+      "step": 12927
+    },
+    {
+      "epoch": 1.6829363529871144,
+      "grad_norm": 2.780635118484497,
+      "learning_rate": 4.198534226770191e-06,
+      "loss": 0.5019,
+      "step": 12930
+    },
+    {
+      "epoch": 1.6833268254588052,
+      "grad_norm": 3.038558006286621,
+      "learning_rate": 4.196471475775728e-06,
+      "loss": 0.4894,
+      "step": 12933
+    },
+    {
+      "epoch": 1.683717297930496,
+      "grad_norm": 3.2223899364471436,
+      "learning_rate": 4.1944088651560085e-06,
+      "loss": 0.5071,
+      "step": 12936
+    },
+    {
+      "epoch": 1.6841077704021865,
+      "grad_norm": 3.295980930328369,
+      "learning_rate": 4.192346395271364e-06,
+      "loss": 0.538,
+      "step": 12939
+    },
+    {
+      "epoch": 1.6844982428738775,
+      "grad_norm": 2.9584147930145264,
+      "learning_rate": 4.190284066482107e-06,
+      "loss": 0.551,
+      "step": 12942
+    },
+    {
+      "epoch": 1.684888715345568,
+      "grad_norm": 2.542959451675415,
+      "learning_rate": 4.18822187914852e-06,
+      "loss": 0.4231,
+      "step": 12945
+    },
+    {
+      "epoch": 1.6852791878172588,
+      "grad_norm": 2.58151912689209,
+      "learning_rate": 4.186159833630862e-06,
+      "loss": 0.5433,
+      "step": 12948
+    },
+    {
+      "epoch": 1.6856696602889496,
+      "grad_norm": 2.462104082107544,
+      "learning_rate": 4.18409793028937e-06,
+      "loss": 0.4551,
+      "step": 12951
+    },
+    {
+      "epoch": 1.6860601327606404,
+      "grad_norm": 2.931121587753296,
+      "learning_rate": 4.182036169484252e-06,
+      "loss": 0.5095,
+      "step": 12954
+    },
+    {
+      "epoch": 1.6864506052323311,
+      "grad_norm": 2.6130144596099854,
+      "learning_rate": 4.1799745515756964e-06,
+      "loss": 0.4584,
+      "step": 12957
+    },
+    {
+      "epoch": 1.686841077704022,
+      "grad_norm": 2.993987798690796,
+      "learning_rate": 4.17791307692386e-06,
+      "loss": 0.4698,
+      "step": 12960
+    },
+    {
+      "epoch": 1.6872315501757127,
+      "grad_norm": 2.883889675140381,
+      "learning_rate": 4.1758517458888805e-06,
+      "loss": 0.5164,
+      "step": 12963
+    },
+    {
+      "epoch": 1.6876220226474032,
+      "grad_norm": 2.5255353450775146,
+      "learning_rate": 4.173790558830868e-06,
+      "loss": 0.4782,
+      "step": 12966
+    },
+    {
+      "epoch": 1.6880124951190942,
+      "grad_norm": 2.975459575653076,
+      "learning_rate": 4.171729516109904e-06,
+      "loss": 0.5081,
+      "step": 12969
+    },
+    {
+      "epoch": 1.6884029675907848,
+      "grad_norm": 2.7381391525268555,
+      "learning_rate": 4.169668618086054e-06,
+      "loss": 0.5112,
+      "step": 12972
+    },
+    {
+      "epoch": 1.6887934400624756,
+      "grad_norm": 2.576137065887451,
+      "learning_rate": 4.167607865119348e-06,
+      "loss": 0.5115,
+      "step": 12975
+    },
+    {
+      "epoch": 1.6891839125341663,
+      "grad_norm": 2.776411533355713,
+      "learning_rate": 4.165547257569797e-06,
+      "loss": 0.4519,
+      "step": 12978
+    },
+    {
+      "epoch": 1.689574385005857,
+      "grad_norm": 2.5057878494262695,
+      "learning_rate": 4.163486795797384e-06,
+      "loss": 0.4683,
+      "step": 12981
+    },
+    {
+      "epoch": 1.6899648574775479,
+      "grad_norm": 2.751605272293091,
+      "learning_rate": 4.161426480162069e-06,
+      "loss": 0.4207,
+      "step": 12984
+    },
+    {
+      "epoch": 1.6903553299492384,
+      "grad_norm": 2.7747738361358643,
+      "learning_rate": 4.1593663110237845e-06,
+      "loss": 0.5169,
+      "step": 12987
+    },
+    {
+      "epoch": 1.6907458024209294,
+      "grad_norm": 2.7248165607452393,
+      "learning_rate": 4.157306288742435e-06,
+      "loss": 0.4886,
+      "step": 12990
+    },
+    {
+      "epoch": 1.69113627489262,
+      "grad_norm": 4.524834156036377,
+      "learning_rate": 4.155246413677907e-06,
+      "loss": 0.4705,
+      "step": 12993
+    },
+    {
+      "epoch": 1.691526747364311,
+      "grad_norm": 3.088681936264038,
+      "learning_rate": 4.153186686190051e-06,
+      "loss": 0.4504,
+      "step": 12996
+    },
+    {
+      "epoch": 1.6919172198360015,
+      "grad_norm": 2.302208662033081,
+      "learning_rate": 4.151127106638701e-06,
+      "loss": 0.4188,
+      "step": 12999
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 2.612056255340576,
+      "learning_rate": 4.149067675383659e-06,
+      "loss": 0.5164,
+      "step": 13002
+    },
+    {
+      "epoch": 1.692698164779383,
+      "grad_norm": 2.487496852874756,
+      "learning_rate": 4.147008392784703e-06,
+      "loss": 0.4577,
+      "step": 13005
+    },
+    {
+      "epoch": 1.6930886372510738,
+      "grad_norm": 2.817664861679077,
+      "learning_rate": 4.144949259201589e-06,
+      "loss": 0.4666,
+      "step": 13008
+    },
+    {
+      "epoch": 1.6934791097227646,
+      "grad_norm": 3.112154006958008,
+      "learning_rate": 4.142890274994038e-06,
+      "loss": 0.5034,
+      "step": 13011
+    },
+    {
+      "epoch": 1.6938695821944552,
+      "grad_norm": 2.818704605102539,
+      "learning_rate": 4.1408314405217544e-06,
+      "loss": 0.5956,
+      "step": 13014
+    },
+    {
+      "epoch": 1.6942600546661462,
+      "grad_norm": 2.6162893772125244,
+      "learning_rate": 4.138772756144411e-06,
+      "loss": 0.5144,
+      "step": 13017
+    },
+    {
+      "epoch": 1.6946505271378367,
+      "grad_norm": 3.4802184104919434,
+      "learning_rate": 4.136714222221654e-06,
+      "loss": 0.375,
+      "step": 13020
+    },
+    {
+      "epoch": 1.6950409996095275,
+      "grad_norm": 2.7583959102630615,
+      "learning_rate": 4.134655839113105e-06,
+      "loss": 0.4669,
+      "step": 13023
+    },
+    {
+      "epoch": 1.6954314720812182,
+      "grad_norm": 3.098271369934082,
+      "learning_rate": 4.132597607178362e-06,
+      "loss": 0.5267,
+      "step": 13026
+    },
+    {
+      "epoch": 1.695821944552909,
+      "grad_norm": 2.7036032676696777,
+      "learning_rate": 4.130539526776994e-06,
+      "loss": 0.5162,
+      "step": 13029
+    },
+    {
+      "epoch": 1.6962124170245998,
+      "grad_norm": 2.6259026527404785,
+      "learning_rate": 4.12848159826854e-06,
+      "loss": 0.4926,
+      "step": 13032
+    },
+    {
+      "epoch": 1.6966028894962903,
+      "grad_norm": 3.8111186027526855,
+      "learning_rate": 4.126423822012516e-06,
+      "loss": 0.4105,
+      "step": 13035
+    },
+    {
+      "epoch": 1.6969933619679813,
+      "grad_norm": 2.448268175125122,
+      "learning_rate": 4.1243661983684155e-06,
+      "loss": 0.5241,
+      "step": 13038
+    },
+    {
+      "epoch": 1.697383834439672,
+      "grad_norm": 2.864579200744629,
+      "learning_rate": 4.1223087276956964e-06,
+      "loss": 0.5952,
+      "step": 13041
+    },
+    {
+      "epoch": 1.6977743069113629,
+      "grad_norm": 2.495445728302002,
+      "learning_rate": 4.120251410353797e-06,
+      "loss": 0.4932,
+      "step": 13044
+    },
+    {
+      "epoch": 1.6981647793830534,
+      "grad_norm": 2.7435507774353027,
+      "learning_rate": 4.1181942467021246e-06,
+      "loss": 0.4825,
+      "step": 13047
+    },
+    {
+      "epoch": 1.6985552518547442,
+      "grad_norm": 2.677366018295288,
+      "learning_rate": 4.116137237100062e-06,
+      "loss": 0.4762,
+      "step": 13050
+    },
+    {
+      "epoch": 1.698945724326435,
+      "grad_norm": 2.592958688735962,
+      "learning_rate": 4.1140803819069665e-06,
+      "loss": 0.4575,
+      "step": 13053
+    },
+    {
+      "epoch": 1.6993361967981258,
+      "grad_norm": 2.5551681518554688,
+      "learning_rate": 4.112023681482163e-06,
+      "loss": 0.5381,
+      "step": 13056
+    },
+    {
+      "epoch": 1.6997266692698165,
+      "grad_norm": 2.487131357192993,
+      "learning_rate": 4.109967136184955e-06,
+      "loss": 0.4652,
+      "step": 13059
+    },
+    {
+      "epoch": 1.700117141741507,
+      "grad_norm": 2.729330539703369,
+      "learning_rate": 4.107910746374616e-06,
+      "loss": 0.4711,
+      "step": 13062
+    },
+    {
+      "epoch": 1.700507614213198,
+      "grad_norm": 2.5914466381073,
+      "learning_rate": 4.105854512410391e-06,
+      "loss": 0.4262,
+      "step": 13065
+    },
+    {
+      "epoch": 1.7008980866848886,
+      "grad_norm": 2.8775665760040283,
+      "learning_rate": 4.1037984346515035e-06,
+      "loss": 0.4693,
+      "step": 13068
+    },
+    {
+      "epoch": 1.7012885591565796,
+      "grad_norm": 4.019662380218506,
+      "learning_rate": 4.101742513457144e-06,
+      "loss": 0.5601,
+      "step": 13071
+    },
+    {
+      "epoch": 1.7016790316282702,
+      "grad_norm": 2.315446615219116,
+      "learning_rate": 4.099686749186478e-06,
+      "loss": 0.4316,
+      "step": 13074
+    },
+    {
+      "epoch": 1.702069504099961,
+      "grad_norm": 2.7482759952545166,
+      "learning_rate": 4.097631142198641e-06,
+      "loss": 0.4958,
+      "step": 13077
+    },
+    {
+      "epoch": 1.7024599765716517,
+      "grad_norm": 2.536959409713745,
+      "learning_rate": 4.0955756928527484e-06,
+      "loss": 0.4941,
+      "step": 13080
+    },
+    {
+      "epoch": 1.7028504490433425,
+      "grad_norm": 2.7001805305480957,
+      "learning_rate": 4.093520401507878e-06,
+      "loss": 0.4224,
+      "step": 13083
+    },
+    {
+      "epoch": 1.7032409215150333,
+      "grad_norm": 2.7134954929351807,
+      "learning_rate": 4.091465268523086e-06,
+      "loss": 0.5029,
+      "step": 13086
+    },
+    {
+      "epoch": 1.7036313939867238,
+      "grad_norm": 2.511518716812134,
+      "learning_rate": 4.089410294257401e-06,
+      "loss": 0.4307,
+      "step": 13089
+    },
+    {
+      "epoch": 1.7040218664584148,
+      "grad_norm": 2.8347418308258057,
+      "learning_rate": 4.087355479069822e-06,
+      "loss": 0.4964,
+      "step": 13092
+    },
+    {
+      "epoch": 1.7044123389301054,
+      "grad_norm": 2.555575132369995,
+      "learning_rate": 4.085300823319321e-06,
+      "loss": 0.4809,
+      "step": 13095
+    },
+    {
+      "epoch": 1.7048028114017961,
+      "grad_norm": 2.7679803371429443,
+      "learning_rate": 4.083246327364842e-06,
+      "loss": 0.4417,
+      "step": 13098
+    },
+    {
+      "epoch": 1.705193283873487,
+      "grad_norm": 2.7733895778656006,
+      "learning_rate": 4.0811919915653e-06,
+      "loss": 0.5306,
+      "step": 13101
+    },
+    {
+      "epoch": 1.7055837563451777,
+      "grad_norm": 2.7801830768585205,
+      "learning_rate": 4.079137816279586e-06,
+      "loss": 0.5075,
+      "step": 13104
+    },
+    {
+      "epoch": 1.7059742288168684,
+      "grad_norm": 3.2246150970458984,
+      "learning_rate": 4.077083801866555e-06,
+      "loss": 0.5001,
+      "step": 13107
+    },
+    {
+      "epoch": 1.706364701288559,
+      "grad_norm": 2.7091383934020996,
+      "learning_rate": 4.0750299486850436e-06,
+      "loss": 0.4837,
+      "step": 13110
+    },
+    {
+      "epoch": 1.70675517376025,
+      "grad_norm": 3.151174783706665,
+      "learning_rate": 4.072976257093855e-06,
+      "loss": 0.5339,
+      "step": 13113
+    },
+    {
+      "epoch": 1.7071456462319405,
+      "grad_norm": 2.550407886505127,
+      "learning_rate": 4.070922727451765e-06,
+      "loss": 0.4892,
+      "step": 13116
+    },
+    {
+      "epoch": 1.7075361187036315,
+      "grad_norm": 2.9066054821014404,
+      "learning_rate": 4.068869360117519e-06,
+      "loss": 0.4448,
+      "step": 13119
+    },
+    {
+      "epoch": 1.707926591175322,
+      "grad_norm": 2.571066379547119,
+      "learning_rate": 4.066816155449837e-06,
+      "loss": 0.6094,
+      "step": 13122
+    },
+    {
+      "epoch": 1.7083170636470129,
+      "grad_norm": 2.580421209335327,
+      "learning_rate": 4.06476311380741e-06,
+      "loss": 0.4401,
+      "step": 13125
+    },
+    {
+      "epoch": 1.7087075361187036,
+      "grad_norm": 2.7733590602874756,
+      "learning_rate": 4.0627102355488986e-06,
+      "loss": 0.5215,
+      "step": 13128
+    },
+    {
+      "epoch": 1.7090980085903944,
+      "grad_norm": 2.7099995613098145,
+      "learning_rate": 4.060657521032939e-06,
+      "loss": 0.5325,
+      "step": 13131
+    },
+    {
+      "epoch": 1.7094884810620852,
+      "grad_norm": 2.444837808609009,
+      "learning_rate": 4.058604970618133e-06,
+      "loss": 0.4751,
+      "step": 13134
+    },
+    {
+      "epoch": 1.7098789535337757,
+      "grad_norm": 2.5356292724609375,
+      "learning_rate": 4.056552584663059e-06,
+      "loss": 0.4456,
+      "step": 13137
+    },
+    {
+      "epoch": 1.7102694260054667,
+      "grad_norm": 2.789571762084961,
+      "learning_rate": 4.054500363526264e-06,
+      "loss": 0.4897,
+      "step": 13140
+    },
+    {
+      "epoch": 1.7106598984771573,
+      "grad_norm": 2.5640852451324463,
+      "learning_rate": 4.052448307566265e-06,
+      "loss": 0.4146,
+      "step": 13143
+    },
+    {
+      "epoch": 1.7110503709488483,
+      "grad_norm": 3.146996259689331,
+      "learning_rate": 4.050396417141555e-06,
+      "loss": 0.4841,
+      "step": 13146
+    },
+    {
+      "epoch": 1.7114408434205388,
+      "grad_norm": 2.4805715084075928,
+      "learning_rate": 4.048344692610591e-06,
+      "loss": 0.4652,
+      "step": 13149
+    },
+    {
+      "epoch": 1.7118313158922296,
+      "grad_norm": 2.782636880874634,
+      "learning_rate": 4.046293134331808e-06,
+      "loss": 0.5592,
+      "step": 13152
+    },
+    {
+      "epoch": 1.7122217883639204,
+      "grad_norm": 2.8912787437438965,
+      "learning_rate": 4.044241742663608e-06,
+      "loss": 0.5252,
+      "step": 13155
+    },
+    {
+      "epoch": 1.7126122608356111,
+      "grad_norm": 2.741149425506592,
+      "learning_rate": 4.042190517964366e-06,
+      "loss": 0.5303,
+      "step": 13158
+    },
+    {
+      "epoch": 1.713002733307302,
+      "grad_norm": 2.541614294052124,
+      "learning_rate": 4.040139460592425e-06,
+      "loss": 0.4626,
+      "step": 13161
+    },
+    {
+      "epoch": 1.7133932057789925,
+      "grad_norm": 2.6299006938934326,
+      "learning_rate": 4.038088570906101e-06,
+      "loss": 0.485,
+      "step": 13164
+    },
+    {
+      "epoch": 1.7137836782506835,
+      "grad_norm": 2.4726014137268066,
+      "learning_rate": 4.036037849263681e-06,
+      "loss": 0.4611,
+      "step": 13167
+    },
+    {
+      "epoch": 1.714174150722374,
+      "grad_norm": 2.893159866333008,
+      "learning_rate": 4.03398729602342e-06,
+      "loss": 0.5536,
+      "step": 13170
+    },
+    {
+      "epoch": 1.7145646231940648,
+      "grad_norm": 2.6248669624328613,
+      "learning_rate": 4.031936911543547e-06,
+      "loss": 0.461,
+      "step": 13173
+    },
+    {
+      "epoch": 1.7149550956657555,
+      "grad_norm": 2.4585137367248535,
+      "learning_rate": 4.029886696182258e-06,
+      "loss": 0.486,
+      "step": 13176
+    },
+    {
+      "epoch": 1.7153455681374463,
+      "grad_norm": 3.081102132797241,
+      "learning_rate": 4.027836650297722e-06,
+      "loss": 0.4455,
+      "step": 13179
+    },
+    {
+      "epoch": 1.715736040609137,
+      "grad_norm": 2.7504281997680664,
+      "learning_rate": 4.025786774248079e-06,
+      "loss": 0.4965,
+      "step": 13182
+    },
+    {
+      "epoch": 1.7161265130808276,
+      "grad_norm": 2.750417709350586,
+      "learning_rate": 4.023737068391437e-06,
+      "loss": 0.4974,
+      "step": 13185
+    },
+    {
+      "epoch": 1.7165169855525186,
+      "grad_norm": 2.768533229827881,
+      "learning_rate": 4.021687533085876e-06,
+      "loss": 0.5187,
+      "step": 13188
+    },
+    {
+      "epoch": 1.7169074580242092,
+      "grad_norm": 2.4056451320648193,
+      "learning_rate": 4.019638168689442e-06,
+      "loss": 0.4074,
+      "step": 13191
+    },
+    {
+      "epoch": 1.7172979304959002,
+      "grad_norm": 2.582237482070923,
+      "learning_rate": 4.0175889755601605e-06,
+      "loss": 0.5469,
+      "step": 13194
+    },
+    {
+      "epoch": 1.7176884029675907,
+      "grad_norm": 2.816178560256958,
+      "learning_rate": 4.015539954056017e-06,
+      "loss": 0.4812,
+      "step": 13197
+    },
+    {
+      "epoch": 1.7180788754392815,
+      "grad_norm": 2.4552249908447266,
+      "learning_rate": 4.013491104534973e-06,
+      "loss": 0.4056,
+      "step": 13200
+    },
+    {
+      "epoch": 1.7184693479109723,
+      "grad_norm": 2.762195110321045,
+      "learning_rate": 4.011442427354958e-06,
+      "loss": 0.5348,
+      "step": 13203
+    },
+    {
+      "epoch": 1.718859820382663,
+      "grad_norm": 2.9049007892608643,
+      "learning_rate": 4.009393922873871e-06,
+      "loss": 0.4404,
+      "step": 13206
+    },
+    {
+      "epoch": 1.7192502928543538,
+      "grad_norm": 2.658714532852173,
+      "learning_rate": 4.007345591449583e-06,
+      "loss": 0.5044,
+      "step": 13209
+    },
+    {
+      "epoch": 1.7196407653260444,
+      "grad_norm": 3.3005869388580322,
+      "learning_rate": 4.005297433439929e-06,
+      "loss": 0.4398,
+      "step": 13212
+    },
+    {
+      "epoch": 1.7200312377977354,
+      "grad_norm": 2.9398393630981445,
+      "learning_rate": 4.003249449202723e-06,
+      "loss": 0.4689,
+      "step": 13215
+    },
+    {
+      "epoch": 1.720421710269426,
+      "grad_norm": 2.541372060775757,
+      "learning_rate": 4.0012016390957414e-06,
+      "loss": 0.5121,
+      "step": 13218
+    },
+    {
+      "epoch": 1.720812182741117,
+      "grad_norm": 2.7919490337371826,
+      "learning_rate": 3.999154003476732e-06,
+      "loss": 0.4993,
+      "step": 13221
+    },
+    {
+      "epoch": 1.7212026552128075,
+      "grad_norm": 3.1717112064361572,
+      "learning_rate": 3.997106542703413e-06,
+      "loss": 0.5892,
+      "step": 13224
+    },
+    {
+      "epoch": 1.7215931276844982,
+      "grad_norm": 2.6251416206359863,
+      "learning_rate": 3.99505925713347e-06,
+      "loss": 0.47,
+      "step": 13227
+    },
+    {
+      "epoch": 1.721983600156189,
+      "grad_norm": 3.013826608657837,
+      "learning_rate": 3.993012147124561e-06,
+      "loss": 0.5313,
+      "step": 13230
+    },
+    {
+      "epoch": 1.7223740726278798,
+      "grad_norm": 2.7360174655914307,
+      "learning_rate": 3.990965213034311e-06,
+      "loss": 0.6029,
+      "step": 13233
+    },
+    {
+      "epoch": 1.7227645450995706,
+      "grad_norm": 2.6772170066833496,
+      "learning_rate": 3.988918455220317e-06,
+      "loss": 0.4684,
+      "step": 13236
+    },
+    {
+      "epoch": 1.723155017571261,
+      "grad_norm": 2.4262611865997314,
+      "learning_rate": 3.986871874040141e-06,
+      "loss": 0.5268,
+      "step": 13239
+    },
+    {
+      "epoch": 1.723545490042952,
+      "grad_norm": 2.6841237545013428,
+      "learning_rate": 3.9848254698513176e-06,
+      "loss": 0.5358,
+      "step": 13242
+    },
+    {
+      "epoch": 1.7239359625146427,
+      "grad_norm": 2.6044881343841553,
+      "learning_rate": 3.98277924301135e-06,
+      "loss": 0.6248,
+      "step": 13245
+    },
+    {
+      "epoch": 1.7243264349863334,
+      "grad_norm": 3.0996434688568115,
+      "learning_rate": 3.980733193877707e-06,
+      "loss": 0.504,
+      "step": 13248
+    },
+    {
+      "epoch": 1.7247169074580242,
+      "grad_norm": 2.843388795852661,
+      "learning_rate": 3.978687322807832e-06,
+      "loss": 0.4786,
+      "step": 13251
+    },
+    {
+      "epoch": 1.725107379929715,
+      "grad_norm": 3.045637607574463,
+      "learning_rate": 3.9766416301591336e-06,
+      "loss": 0.4879,
+      "step": 13254
+    },
+    {
+      "epoch": 1.7254978524014057,
+      "grad_norm": 2.5478973388671875,
+      "learning_rate": 3.974596116288988e-06,
+      "loss": 0.4843,
+      "step": 13257
+    },
+    {
+      "epoch": 1.7258883248730963,
+      "grad_norm": 2.8354203701019287,
+      "learning_rate": 3.972550781554745e-06,
+      "loss": 0.5909,
+      "step": 13260
+    },
+    {
+      "epoch": 1.7262787973447873,
+      "grad_norm": 2.7125988006591797,
+      "learning_rate": 3.970505626313718e-06,
+      "loss": 0.4752,
+      "step": 13263
+    },
+    {
+      "epoch": 1.7266692698164778,
+      "grad_norm": 2.5882513523101807,
+      "learning_rate": 3.9684606509231935e-06,
+      "loss": 0.4578,
+      "step": 13266
+    },
+    {
+      "epoch": 1.7270597422881688,
+      "grad_norm": 2.7400310039520264,
+      "learning_rate": 3.966415855740423e-06,
+      "loss": 0.5583,
+      "step": 13269
+    },
+    {
+      "epoch": 1.7274502147598594,
+      "grad_norm": 2.917891025543213,
+      "learning_rate": 3.964371241122627e-06,
+      "loss": 0.4712,
+      "step": 13272
+    },
+    {
+      "epoch": 1.7278406872315502,
+      "grad_norm": 2.7035434246063232,
+      "learning_rate": 3.962326807426996e-06,
+      "loss": 0.5,
+      "step": 13275
+    },
+    {
+      "epoch": 1.728231159703241,
+      "grad_norm": 2.7468101978302,
+      "learning_rate": 3.960282555010691e-06,
+      "loss": 0.5003,
+      "step": 13278
+    },
+    {
+      "epoch": 1.7286216321749317,
+      "grad_norm": 2.587204694747925,
+      "learning_rate": 3.958238484230835e-06,
+      "loss": 0.5208,
+      "step": 13281
+    },
+    {
+      "epoch": 1.7290121046466225,
+      "grad_norm": 2.676217555999756,
+      "learning_rate": 3.956194595444525e-06,
+      "loss": 0.5368,
+      "step": 13284
+    },
+    {
+      "epoch": 1.729402577118313,
+      "grad_norm": 2.5884201526641846,
+      "learning_rate": 3.954150889008823e-06,
+      "loss": 0.5131,
+      "step": 13287
+    },
+    {
+      "epoch": 1.729793049590004,
+      "grad_norm": 2.455409049987793,
+      "learning_rate": 3.95210736528076e-06,
+      "loss": 0.4353,
+      "step": 13290
+    },
+    {
+      "epoch": 1.7301835220616946,
+      "grad_norm": 2.7997970581054688,
+      "learning_rate": 3.9500640246173376e-06,
+      "loss": 0.5197,
+      "step": 13293
+    },
+    {
+      "epoch": 1.7305739945333856,
+      "grad_norm": 2.450458288192749,
+      "learning_rate": 3.948020867375521e-06,
+      "loss": 0.5324,
+      "step": 13296
+    },
+    {
+      "epoch": 1.7309644670050761,
+      "grad_norm": 2.428149461746216,
+      "learning_rate": 3.945977893912244e-06,
+      "loss": 0.508,
+      "step": 13299
+    },
+    {
+      "epoch": 1.7313549394767669,
+      "grad_norm": 4.35640811920166,
+      "learning_rate": 3.943935104584413e-06,
+      "loss": 0.525,
+      "step": 13302
+    },
+    {
+      "epoch": 1.7317454119484577,
+      "grad_norm": 2.549283266067505,
+      "learning_rate": 3.941892499748897e-06,
+      "loss": 0.4751,
+      "step": 13305
+    },
+    {
+      "epoch": 1.7321358844201484,
+      "grad_norm": 3.114838123321533,
+      "learning_rate": 3.9398500797625355e-06,
+      "loss": 0.5122,
+      "step": 13308
+    },
+    {
+      "epoch": 1.7325263568918392,
+      "grad_norm": 3.187574863433838,
+      "learning_rate": 3.937807844982136e-06,
+      "loss": 0.4961,
+      "step": 13311
+    },
+    {
+      "epoch": 1.7329168293635298,
+      "grad_norm": 2.5373189449310303,
+      "learning_rate": 3.93576579576447e-06,
+      "loss": 0.4976,
+      "step": 13314
+    },
+    {
+      "epoch": 1.7333073018352207,
+      "grad_norm": 2.5374085903167725,
+      "learning_rate": 3.93372393246628e-06,
+      "loss": 0.4775,
+      "step": 13317
+    },
+    {
+      "epoch": 1.7336977743069113,
+      "grad_norm": 2.4994728565216064,
+      "learning_rate": 3.931682255444276e-06,
+      "loss": 0.4706,
+      "step": 13320
+    },
+    {
+      "epoch": 1.734088246778602,
+      "grad_norm": 4.567479610443115,
+      "learning_rate": 3.929640765055137e-06,
+      "loss": 0.5715,
+      "step": 13323
+    },
+    {
+      "epoch": 1.7344787192502928,
+      "grad_norm": 2.689119815826416,
+      "learning_rate": 3.927599461655503e-06,
+      "loss": 0.4771,
+      "step": 13326
+    },
+    {
+      "epoch": 1.7348691917219836,
+      "grad_norm": 2.8153414726257324,
+      "learning_rate": 3.925558345601987e-06,
+      "loss": 0.4448,
+      "step": 13329
+    },
+    {
+      "epoch": 1.7352596641936744,
+      "grad_norm": 2.966712236404419,
+      "learning_rate": 3.923517417251168e-06,
+      "loss": 0.5059,
+      "step": 13332
+    },
+    {
+      "epoch": 1.735650136665365,
+      "grad_norm": 2.778205156326294,
+      "learning_rate": 3.921476676959591e-06,
+      "loss": 0.4661,
+      "step": 13335
+    },
+    {
+      "epoch": 1.736040609137056,
+      "grad_norm": 3.179136276245117,
+      "learning_rate": 3.919436125083771e-06,
+      "loss": 0.5206,
+      "step": 13338
+    },
+    {
+      "epoch": 1.7364310816087465,
+      "grad_norm": 2.8156237602233887,
+      "learning_rate": 3.917395761980186e-06,
+      "loss": 0.5265,
+      "step": 13341
+    },
+    {
+      "epoch": 1.7368215540804375,
+      "grad_norm": 2.721435546875,
+      "learning_rate": 3.915355588005283e-06,
+      "loss": 0.5218,
+      "step": 13344
+    },
+    {
+      "epoch": 1.737212026552128,
+      "grad_norm": 3.0874862670898438,
+      "learning_rate": 3.913315603515479e-06,
+      "loss": 0.4019,
+      "step": 13347
+    },
+    {
+      "epoch": 1.7376024990238188,
+      "grad_norm": 3.176990032196045,
+      "learning_rate": 3.911275808867151e-06,
+      "loss": 0.5453,
+      "step": 13350
+    },
+    {
+      "epoch": 1.7379929714955096,
+      "grad_norm": 2.735309600830078,
+      "learning_rate": 3.909236204416651e-06,
+      "loss": 0.4105,
+      "step": 13353
+    },
+    {
+      "epoch": 1.7383834439672003,
+      "grad_norm": 2.8676302433013916,
+      "learning_rate": 3.90719679052029e-06,
+      "loss": 0.4664,
+      "step": 13356
+    },
+    {
+      "epoch": 1.7387739164388911,
+      "grad_norm": 3.0311317443847656,
+      "learning_rate": 3.905157567534349e-06,
+      "loss": 0.509,
+      "step": 13359
+    },
+    {
+      "epoch": 1.7391643889105817,
+      "grad_norm": 2.751351833343506,
+      "learning_rate": 3.9031185358150794e-06,
+      "loss": 0.5447,
+      "step": 13362
+    },
+    {
+      "epoch": 1.7395548613822727,
+      "grad_norm": 2.6076507568359375,
+      "learning_rate": 3.901079695718696e-06,
+      "loss": 0.4675,
+      "step": 13365
+    },
+    {
+      "epoch": 1.7399453338539632,
+      "grad_norm": 2.406752109527588,
+      "learning_rate": 3.899041047601375e-06,
+      "loss": 0.5027,
+      "step": 13368
+    },
+    {
+      "epoch": 1.740335806325654,
+      "grad_norm": 2.558464527130127,
+      "learning_rate": 3.897002591819269e-06,
+      "loss": 0.4831,
+      "step": 13371
+    },
+    {
+      "epoch": 1.7407262787973448,
+      "grad_norm": 2.5584099292755127,
+      "learning_rate": 3.894964328728489e-06,
+      "loss": 0.4087,
+      "step": 13374
+    },
+    {
+      "epoch": 1.7411167512690355,
+      "grad_norm": 2.5314037799835205,
+      "learning_rate": 3.8929262586851164e-06,
+      "loss": 0.4534,
+      "step": 13377
+    },
+    {
+      "epoch": 1.7415072237407263,
+      "grad_norm": 2.593812942504883,
+      "learning_rate": 3.890888382045198e-06,
+      "loss": 0.454,
+      "step": 13380
+    },
+    {
+      "epoch": 1.741897696212417,
+      "grad_norm": 2.6239094734191895,
+      "learning_rate": 3.8888506991647455e-06,
+      "loss": 0.4661,
+      "step": 13383
+    },
+    {
+      "epoch": 1.7422881686841079,
+      "grad_norm": 2.606416940689087,
+      "learning_rate": 3.886813210399738e-06,
+      "loss": 0.5233,
+      "step": 13386
+    },
+    {
+      "epoch": 1.7426786411557984,
+      "grad_norm": 2.7739343643188477,
+      "learning_rate": 3.884775916106121e-06,
+      "loss": 0.4525,
+      "step": 13389
+    },
+    {
+      "epoch": 1.7430691136274894,
+      "grad_norm": 2.44478440284729,
+      "learning_rate": 3.882738816639806e-06,
+      "loss": 0.452,
+      "step": 13392
+    },
+    {
+      "epoch": 1.74345958609918,
+      "grad_norm": 2.652439832687378,
+      "learning_rate": 3.880701912356668e-06,
+      "loss": 0.4519,
+      "step": 13395
+    },
+    {
+      "epoch": 1.7438500585708707,
+      "grad_norm": 2.414402484893799,
+      "learning_rate": 3.878665203612553e-06,
+      "loss": 0.4507,
+      "step": 13398
+    },
+    {
+      "epoch": 1.7442405310425615,
+      "grad_norm": 2.590601682662964,
+      "learning_rate": 3.876628690763265e-06,
+      "loss": 0.4609,
+      "step": 13401
+    },
+    {
+      "epoch": 1.7446310035142523,
+      "grad_norm": 2.637747049331665,
+      "learning_rate": 3.874592374164583e-06,
+      "loss": 0.6,
+      "step": 13404
+    },
+    {
+      "epoch": 1.745021475985943,
+      "grad_norm": 2.618107795715332,
+      "learning_rate": 3.872556254172246e-06,
+      "loss": 0.5137,
+      "step": 13407
+    },
+    {
+      "epoch": 1.7454119484576336,
+      "grad_norm": 2.5348377227783203,
+      "learning_rate": 3.870520331141961e-06,
+      "loss": 0.5094,
+      "step": 13410
+    },
+    {
+      "epoch": 1.7458024209293246,
+      "grad_norm": 2.787950277328491,
+      "learning_rate": 3.868484605429396e-06,
+      "loss": 0.5294,
+      "step": 13413
+    },
+    {
+      "epoch": 1.7461928934010151,
+      "grad_norm": 2.4089295864105225,
+      "learning_rate": 3.866449077390192e-06,
+      "loss": 0.439,
+      "step": 13416
+    },
+    {
+      "epoch": 1.7465833658727061,
+      "grad_norm": 2.6536457538604736,
+      "learning_rate": 3.864413747379948e-06,
+      "loss": 0.4402,
+      "step": 13419
+    },
+    {
+      "epoch": 1.7469738383443967,
+      "grad_norm": 2.460862398147583,
+      "learning_rate": 3.862378615754233e-06,
+      "loss": 0.4751,
+      "step": 13422
+    },
+    {
+      "epoch": 1.7473643108160875,
+      "grad_norm": 2.9238407611846924,
+      "learning_rate": 3.860343682868583e-06,
+      "loss": 0.4904,
+      "step": 13425
+    },
+    {
+      "epoch": 1.7477547832877782,
+      "grad_norm": 2.762789249420166,
+      "learning_rate": 3.858308949078492e-06,
+      "loss": 0.5113,
+      "step": 13428
+    },
+    {
+      "epoch": 1.748145255759469,
+      "grad_norm": 2.810913324356079,
+      "learning_rate": 3.856274414739428e-06,
+      "loss": 0.4988,
+      "step": 13431
+    },
+    {
+      "epoch": 1.7485357282311598,
+      "grad_norm": 2.568662643432617,
+      "learning_rate": 3.854240080206815e-06,
+      "loss": 0.5601,
+      "step": 13434
+    },
+    {
+      "epoch": 1.7489262007028503,
+      "grad_norm": 2.7099461555480957,
+      "learning_rate": 3.852205945836051e-06,
+      "loss": 0.4212,
+      "step": 13437
+    },
+    {
+      "epoch": 1.7493166731745413,
+      "grad_norm": 2.6403145790100098,
+      "learning_rate": 3.850172011982494e-06,
+      "loss": 0.4719,
+      "step": 13440
+    },
+    {
+      "epoch": 1.7497071456462319,
+      "grad_norm": 2.7137389183044434,
+      "learning_rate": 3.848138279001466e-06,
+      "loss": 0.4694,
+      "step": 13443
+    },
+    {
+      "epoch": 1.7500976181179226,
+      "grad_norm": 2.654871702194214,
+      "learning_rate": 3.8461047472482584e-06,
+      "loss": 0.4603,
+      "step": 13446
+    },
+    {
+      "epoch": 1.7504880905896134,
+      "grad_norm": 2.468869209289551,
+      "learning_rate": 3.844071417078124e-06,
+      "loss": 0.4592,
+      "step": 13449
+    },
+    {
+      "epoch": 1.7508785630613042,
+      "grad_norm": 3.2677621841430664,
+      "learning_rate": 3.842038288846282e-06,
+      "loss": 0.5041,
+      "step": 13452
+    },
+    {
+      "epoch": 1.751269035532995,
+      "grad_norm": 2.530336380004883,
+      "learning_rate": 3.8400053629079145e-06,
+      "loss": 0.5458,
+      "step": 13455
+    },
+    {
+      "epoch": 1.7516595080046855,
+      "grad_norm": 2.7045438289642334,
+      "learning_rate": 3.8379726396181705e-06,
+      "loss": 0.4496,
+      "step": 13458
+    },
+    {
+      "epoch": 1.7520499804763765,
+      "grad_norm": 2.640897035598755,
+      "learning_rate": 3.83594011933216e-06,
+      "loss": 0.4873,
+      "step": 13461
+    },
+    {
+      "epoch": 1.752440452948067,
+      "grad_norm": 2.674314022064209,
+      "learning_rate": 3.833907802404963e-06,
+      "loss": 0.5766,
+      "step": 13464
+    },
+    {
+      "epoch": 1.752830925419758,
+      "grad_norm": 3.62727689743042,
+      "learning_rate": 3.83187568919162e-06,
+      "loss": 0.4801,
+      "step": 13467
+    },
+    {
+      "epoch": 1.7532213978914486,
+      "grad_norm": 2.658531904220581,
+      "learning_rate": 3.829843780047137e-06,
+      "loss": 0.5359,
+      "step": 13470
+    },
+    {
+      "epoch": 1.7536118703631394,
+      "grad_norm": 3.2108259201049805,
+      "learning_rate": 3.827812075326483e-06,
+      "loss": 0.542,
+      "step": 13473
+    },
+    {
+      "epoch": 1.7540023428348301,
+      "grad_norm": 2.5640816688537598,
+      "learning_rate": 3.825780575384595e-06,
+      "loss": 0.5172,
+      "step": 13476
+    },
+    {
+      "epoch": 1.754392815306521,
+      "grad_norm": 4.314723491668701,
+      "learning_rate": 3.823749280576369e-06,
+      "loss": 0.4687,
+      "step": 13479
+    },
+    {
+      "epoch": 1.7547832877782117,
+      "grad_norm": 2.8595173358917236,
+      "learning_rate": 3.821718191256669e-06,
+      "loss": 0.5252,
+      "step": 13482
+    },
+    {
+      "epoch": 1.7551737602499022,
+      "grad_norm": 2.5360357761383057,
+      "learning_rate": 3.819687307780321e-06,
+      "loss": 0.4925,
+      "step": 13485
+    },
+    {
+      "epoch": 1.7555642327215932,
+      "grad_norm": 2.3102357387542725,
+      "learning_rate": 3.81765663050212e-06,
+      "loss": 0.516,
+      "step": 13488
+    },
+    {
+      "epoch": 1.7559547051932838,
+      "grad_norm": 2.8565187454223633,
+      "learning_rate": 3.8156261597768165e-06,
+      "loss": 0.5065,
+      "step": 13491
+    },
+    {
+      "epoch": 1.7563451776649748,
+      "grad_norm": 3.1690919399261475,
+      "learning_rate": 3.8135958959591334e-06,
+      "loss": 0.506,
+      "step": 13494
+    },
+    {
+      "epoch": 1.7567356501366653,
+      "grad_norm": 2.7095978260040283,
+      "learning_rate": 3.8115658394037496e-06,
+      "loss": 0.5125,
+      "step": 13497
+    },
+    {
+      "epoch": 1.757126122608356,
+      "grad_norm": 2.614197254180908,
+      "learning_rate": 3.809535990465314e-06,
+      "loss": 0.5541,
+      "step": 13500
+    },
+    {
+      "epoch": 1.7575165950800469,
+      "grad_norm": 3.0147502422332764,
+      "learning_rate": 3.807506349498438e-06,
+      "loss": 0.4678,
+      "step": 13503
+    },
+    {
+      "epoch": 1.7579070675517376,
+      "grad_norm": 2.663944721221924,
+      "learning_rate": 3.8054769168576924e-06,
+      "loss": 0.4942,
+      "step": 13506
+    },
+    {
+      "epoch": 1.7582975400234284,
+      "grad_norm": 2.467390775680542,
+      "learning_rate": 3.803447692897617e-06,
+      "loss": 0.4927,
+      "step": 13509
+    },
+    {
+      "epoch": 1.758688012495119,
+      "grad_norm": 3.1490232944488525,
+      "learning_rate": 3.8014186779727123e-06,
+      "loss": 0.5778,
+      "step": 13512
+    },
+    {
+      "epoch": 1.75907848496681,
+      "grad_norm": 2.497331142425537,
+      "learning_rate": 3.7993898724374435e-06,
+      "loss": 0.4939,
+      "step": 13515
+    },
+    {
+      "epoch": 1.7594689574385005,
+      "grad_norm": 2.6486501693725586,
+      "learning_rate": 3.7973612766462387e-06,
+      "loss": 0.4405,
+      "step": 13518
+    },
+    {
+      "epoch": 1.7598594299101913,
+      "grad_norm": 3.311718702316284,
+      "learning_rate": 3.7953328909534876e-06,
+      "loss": 0.5304,
+      "step": 13521
+    },
+    {
+      "epoch": 1.760249902381882,
+      "grad_norm": 3.001080274581909,
+      "learning_rate": 3.7933047157135465e-06,
+      "loss": 0.6345,
+      "step": 13524
+    },
+    {
+      "epoch": 1.7606403748535728,
+      "grad_norm": 2.920088529586792,
+      "learning_rate": 3.7912767512807318e-06,
+      "loss": 0.5114,
+      "step": 13527
+    },
+    {
+      "epoch": 1.7610308473252636,
+      "grad_norm": 2.604182720184326,
+      "learning_rate": 3.7892489980093285e-06,
+      "loss": 0.5102,
+      "step": 13530
+    },
+    {
+      "epoch": 1.7614213197969542,
+      "grad_norm": 3.248330593109131,
+      "learning_rate": 3.7872214562535765e-06,
+      "loss": 0.4517,
+      "step": 13533
+    },
+    {
+      "epoch": 1.7618117922686451,
+      "grad_norm": 3.4100341796875,
+      "learning_rate": 3.785194126367685e-06,
+      "loss": 0.4723,
+      "step": 13536
+    },
+    {
+      "epoch": 1.7622022647403357,
+      "grad_norm": 3.491152286529541,
+      "learning_rate": 3.783167008705825e-06,
+      "loss": 0.5118,
+      "step": 13539
+    },
+    {
+      "epoch": 1.7625927372120267,
+      "grad_norm": 2.765254259109497,
+      "learning_rate": 3.7811401036221283e-06,
+      "loss": 0.4562,
+      "step": 13542
+    },
+    {
+      "epoch": 1.7629832096837172,
+      "grad_norm": 2.513669967651367,
+      "learning_rate": 3.779113411470692e-06,
+      "loss": 0.5244,
+      "step": 13545
+    },
+    {
+      "epoch": 1.763373682155408,
+      "grad_norm": 2.982708215713501,
+      "learning_rate": 3.7770869326055733e-06,
+      "loss": 0.5286,
+      "step": 13548
+    },
+    {
+      "epoch": 1.7637641546270988,
+      "grad_norm": 3.210015058517456,
+      "learning_rate": 3.7750606673807945e-06,
+      "loss": 0.5115,
+      "step": 13551
+    },
+    {
+      "epoch": 1.7641546270987896,
+      "grad_norm": 3.011054277420044,
+      "learning_rate": 3.773034616150342e-06,
+      "loss": 0.5402,
+      "step": 13554
+    },
+    {
+      "epoch": 1.7645450995704803,
+      "grad_norm": 4.221112251281738,
+      "learning_rate": 3.7710087792681594e-06,
+      "loss": 0.6668,
+      "step": 13557
+    },
+    {
+      "epoch": 1.7649355720421709,
+      "grad_norm": 2.5336077213287354,
+      "learning_rate": 3.7689831570881584e-06,
+      "loss": 0.4922,
+      "step": 13560
+    },
+    {
+      "epoch": 1.7653260445138619,
+      "grad_norm": 2.933840751647949,
+      "learning_rate": 3.7669577499642094e-06,
+      "loss": 0.5391,
+      "step": 13563
+    },
+    {
+      "epoch": 1.7657165169855524,
+      "grad_norm": 2.6583011150360107,
+      "learning_rate": 3.7649325582501478e-06,
+      "loss": 0.5105,
+      "step": 13566
+    },
+    {
+      "epoch": 1.7661069894572434,
+      "grad_norm": 2.7706074714660645,
+      "learning_rate": 3.7629075822997685e-06,
+      "loss": 0.5597,
+      "step": 13569
+    },
+    {
+      "epoch": 1.766497461928934,
+      "grad_norm": 2.6521201133728027,
+      "learning_rate": 3.7608828224668346e-06,
+      "loss": 0.4072,
+      "step": 13572
+    },
+    {
+      "epoch": 1.7668879344006247,
+      "grad_norm": 2.8983657360076904,
+      "learning_rate": 3.7588582791050644e-06,
+      "loss": 0.5915,
+      "step": 13575
+    },
+    {
+      "epoch": 1.7672784068723155,
+      "grad_norm": 2.609997510910034,
+      "learning_rate": 3.7568339525681407e-06,
+      "loss": 0.4157,
+      "step": 13578
+    },
+    {
+      "epoch": 1.7676688793440063,
+      "grad_norm": 3.424751043319702,
+      "learning_rate": 3.754809843209712e-06,
+      "loss": 0.5333,
+      "step": 13581
+    },
+    {
+      "epoch": 1.768059351815697,
+      "grad_norm": 4.254384994506836,
+      "learning_rate": 3.752785951383383e-06,
+      "loss": 0.5174,
+      "step": 13584
+    },
+    {
+      "epoch": 1.7684498242873876,
+      "grad_norm": 3.230611562728882,
+      "learning_rate": 3.7507622774427242e-06,
+      "loss": 0.602,
+      "step": 13587
+    },
+    {
+      "epoch": 1.7688402967590786,
+      "grad_norm": 3.1927285194396973,
+      "learning_rate": 3.748738821741269e-06,
+      "loss": 0.523,
+      "step": 13590
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 3.054600238800049,
+      "learning_rate": 3.7467155846325086e-06,
+      "loss": 0.4559,
+      "step": 13593
+    },
+    {
+      "epoch": 1.76962124170246,
+      "grad_norm": 2.9370806217193604,
+      "learning_rate": 3.7446925664699e-06,
+      "loss": 0.5063,
+      "step": 13596
+    },
+    {
+      "epoch": 1.7700117141741507,
+      "grad_norm": 2.851138114929199,
+      "learning_rate": 3.7426697676068575e-06,
+      "loss": 0.4684,
+      "step": 13599
+    },
+    {
+      "epoch": 1.7704021866458415,
+      "grad_norm": 2.6864585876464844,
+      "learning_rate": 3.740647188396762e-06,
+      "loss": 0.4323,
+      "step": 13602
+    },
+    {
+      "epoch": 1.7707926591175323,
+      "grad_norm": 3.145974636077881,
+      "learning_rate": 3.7386248291929544e-06,
+      "loss": 0.4725,
+      "step": 13605
+    },
+    {
+      "epoch": 1.7711831315892228,
+      "grad_norm": 4.056112766265869,
+      "learning_rate": 3.7366026903487346e-06,
+      "loss": 0.6875,
+      "step": 13608
+    },
+    {
+      "epoch": 1.7715736040609138,
+      "grad_norm": 2.8808858394622803,
+      "learning_rate": 3.7345807722173655e-06,
+      "loss": 0.6169,
+      "step": 13611
+    },
+    {
+      "epoch": 1.7719640765326043,
+      "grad_norm": 3.3072636127471924,
+      "learning_rate": 3.732559075152075e-06,
+      "loss": 0.5757,
+      "step": 13614
+    },
+    {
+      "epoch": 1.7723545490042953,
+      "grad_norm": 2.6923253536224365,
+      "learning_rate": 3.730537599506049e-06,
+      "loss": 0.5674,
+      "step": 13617
+    },
+    {
+      "epoch": 1.772745021475986,
+      "grad_norm": 2.857133150100708,
+      "learning_rate": 3.7285163456324323e-06,
+      "loss": 0.5753,
+      "step": 13620
+    },
+    {
+      "epoch": 1.7731354939476767,
+      "grad_norm": 2.6387181282043457,
+      "learning_rate": 3.7264953138843363e-06,
+      "loss": 0.525,
+      "step": 13623
+    },
+    {
+      "epoch": 1.7735259664193674,
+      "grad_norm": 2.5029585361480713,
+      "learning_rate": 3.724474504614829e-06,
+      "loss": 0.4137,
+      "step": 13626
+    },
+    {
+      "epoch": 1.7739164388910582,
+      "grad_norm": 2.632925510406494,
+      "learning_rate": 3.7224539181769425e-06,
+      "loss": 0.5328,
+      "step": 13629
+    },
+    {
+      "epoch": 1.774306911362749,
+      "grad_norm": 2.7073307037353516,
+      "learning_rate": 3.7204335549236703e-06,
+      "loss": 0.4625,
+      "step": 13632
+    },
+    {
+      "epoch": 1.7746973838344395,
+      "grad_norm": 2.770639181137085,
+      "learning_rate": 3.718413415207962e-06,
+      "loss": 0.4865,
+      "step": 13635
+    },
+    {
+      "epoch": 1.7750878563061305,
+      "grad_norm": 3.0762181282043457,
+      "learning_rate": 3.7163934993827364e-06,
+      "loss": 0.522,
+      "step": 13638
+    },
+    {
+      "epoch": 1.775478328777821,
+      "grad_norm": 2.7162609100341797,
+      "learning_rate": 3.714373807800864e-06,
+      "loss": 0.4633,
+      "step": 13641
+    },
+    {
+      "epoch": 1.775868801249512,
+      "grad_norm": 2.4745163917541504,
+      "learning_rate": 3.7123543408151843e-06,
+      "loss": 0.4783,
+      "step": 13644
+    },
+    {
+      "epoch": 1.7762592737212026,
+      "grad_norm": 2.7672512531280518,
+      "learning_rate": 3.710335098778492e-06,
+      "loss": 0.548,
+      "step": 13647
+    },
+    {
+      "epoch": 1.7766497461928934,
+      "grad_norm": 2.50407075881958,
+      "learning_rate": 3.7083160820435445e-06,
+      "loss": 0.4029,
+      "step": 13650
+    },
+    {
+      "epoch": 1.7770402186645842,
+      "grad_norm": 2.7135939598083496,
+      "learning_rate": 3.706297290963059e-06,
+      "loss": 0.4438,
+      "step": 13653
+    },
+    {
+      "epoch": 1.777430691136275,
+      "grad_norm": 2.827840805053711,
+      "learning_rate": 3.7042787258897163e-06,
+      "loss": 0.5588,
+      "step": 13656
+    },
+    {
+      "epoch": 1.7778211636079657,
+      "grad_norm": 2.7864699363708496,
+      "learning_rate": 3.7022603871761554e-06,
+      "loss": 0.4841,
+      "step": 13659
+    },
+    {
+      "epoch": 1.7782116360796563,
+      "grad_norm": 3.1784870624542236,
+      "learning_rate": 3.700242275174973e-06,
+      "loss": 0.5815,
+      "step": 13662
+    },
+    {
+      "epoch": 1.7786021085513473,
+      "grad_norm": 2.666109323501587,
+      "learning_rate": 3.698224390238732e-06,
+      "loss": 0.5251,
+      "step": 13665
+    },
+    {
+      "epoch": 1.7789925810230378,
+      "grad_norm": 2.4882564544677734,
+      "learning_rate": 3.6962067327199523e-06,
+      "loss": 0.4551,
+      "step": 13668
+    },
+    {
+      "epoch": 1.7793830534947286,
+      "grad_norm": 2.7891669273376465,
+      "learning_rate": 3.6941893029711123e-06,
+      "loss": 0.4774,
+      "step": 13671
+    },
+    {
+      "epoch": 1.7797735259664194,
+      "grad_norm": 2.7892396450042725,
+      "learning_rate": 3.6921721013446555e-06,
+      "loss": 0.512,
+      "step": 13674
+    },
+    {
+      "epoch": 1.7801639984381101,
+      "grad_norm": 2.7493927478790283,
+      "learning_rate": 3.690155128192979e-06,
+      "loss": 0.3913,
+      "step": 13677
+    },
+    {
+      "epoch": 1.780554470909801,
+      "grad_norm": 2.922779083251953,
+      "learning_rate": 3.6881383838684475e-06,
+      "loss": 0.5359,
+      "step": 13680
+    },
+    {
+      "epoch": 1.7809449433814915,
+      "grad_norm": 3.476430892944336,
+      "learning_rate": 3.6861218687233813e-06,
+      "loss": 0.4305,
+      "step": 13683
+    },
+    {
+      "epoch": 1.7813354158531824,
+      "grad_norm": 2.7361629009246826,
+      "learning_rate": 3.6841055831100593e-06,
+      "loss": 0.5422,
+      "step": 13686
+    },
+    {
+      "epoch": 1.781725888324873,
+      "grad_norm": 2.4145750999450684,
+      "learning_rate": 3.6820895273807257e-06,
+      "loss": 0.5088,
+      "step": 13689
+    },
+    {
+      "epoch": 1.782116360796564,
+      "grad_norm": 2.698687791824341,
+      "learning_rate": 3.6800737018875765e-06,
+      "loss": 0.5185,
+      "step": 13692
+    },
+    {
+      "epoch": 1.7825068332682545,
+      "grad_norm": 3.8540420532226562,
+      "learning_rate": 3.678058106982775e-06,
+      "loss": 0.4911,
+      "step": 13695
+    },
+    {
+      "epoch": 1.7828973057399453,
+      "grad_norm": 2.5574920177459717,
+      "learning_rate": 3.676042743018442e-06,
+      "loss": 0.5211,
+      "step": 13698
+    },
+    {
+      "epoch": 1.783287778211636,
+      "grad_norm": 2.5759153366088867,
+      "learning_rate": 3.674027610346658e-06,
+      "loss": 0.455,
+      "step": 13701
+    },
+    {
+      "epoch": 1.7836782506833269,
+      "grad_norm": 2.5695457458496094,
+      "learning_rate": 3.672012709319459e-06,
+      "loss": 0.4909,
+      "step": 13704
+    },
+    {
+      "epoch": 1.7840687231550176,
+      "grad_norm": 2.9643442630767822,
+      "learning_rate": 3.669998040288847e-06,
+      "loss": 0.4506,
+      "step": 13707
+    },
+    {
+      "epoch": 1.7844591956267082,
+      "grad_norm": 2.8210248947143555,
+      "learning_rate": 3.66798360360678e-06,
+      "loss": 0.4668,
+      "step": 13710
+    },
+    {
+      "epoch": 1.7848496680983992,
+      "grad_norm": 2.568918228149414,
+      "learning_rate": 3.6659693996251745e-06,
+      "loss": 0.4581,
+      "step": 13713
+    },
+    {
+      "epoch": 1.7852401405700897,
+      "grad_norm": 2.8247551918029785,
+      "learning_rate": 3.663955428695908e-06,
+      "loss": 0.4848,
+      "step": 13716
+    },
+    {
+      "epoch": 1.7856306130417805,
+      "grad_norm": 2.960132598876953,
+      "learning_rate": 3.6619416911708196e-06,
+      "loss": 0.497,
+      "step": 13719
+    },
+    {
+      "epoch": 1.7860210855134713,
+      "grad_norm": 2.265368938446045,
+      "learning_rate": 3.6599281874017005e-06,
+      "loss": 0.4379,
+      "step": 13722
+    },
+    {
+      "epoch": 1.786411557985162,
+      "grad_norm": 2.88246750831604,
+      "learning_rate": 3.6579149177403093e-06,
+      "loss": 0.61,
+      "step": 13725
+    },
+    {
+      "epoch": 1.7868020304568528,
+      "grad_norm": 2.980820655822754,
+      "learning_rate": 3.6559018825383587e-06,
+      "loss": 0.5359,
+      "step": 13728
+    },
+    {
+      "epoch": 1.7871925029285436,
+      "grad_norm": 2.5079283714294434,
+      "learning_rate": 3.6538890821475204e-06,
+      "loss": 0.5089,
+      "step": 13731
+    },
+    {
+      "epoch": 1.7875829754002344,
+      "grad_norm": 2.517543077468872,
+      "learning_rate": 3.6518765169194294e-06,
+      "loss": 0.4798,
+      "step": 13734
+    },
+    {
+      "epoch": 1.787973447871925,
+      "grad_norm": 3.10083270072937,
+      "learning_rate": 3.649864187205672e-06,
+      "loss": 0.4349,
+      "step": 13737
+    },
+    {
+      "epoch": 1.788363920343616,
+      "grad_norm": 2.687077522277832,
+      "learning_rate": 3.647852093357803e-06,
+      "loss": 0.4688,
+      "step": 13740
+    },
+    {
+      "epoch": 1.7887543928153065,
+      "grad_norm": 2.802783966064453,
+      "learning_rate": 3.645840235727328e-06,
+      "loss": 0.4867,
+      "step": 13743
+    },
+    {
+      "epoch": 1.7891448652869972,
+      "grad_norm": 2.6640493869781494,
+      "learning_rate": 3.6438286146657166e-06,
+      "loss": 0.5281,
+      "step": 13746
+    },
+    {
+      "epoch": 1.789535337758688,
+      "grad_norm": 2.8820464611053467,
+      "learning_rate": 3.6418172305243914e-06,
+      "loss": 0.4845,
+      "step": 13749
+    },
+    {
+      "epoch": 1.7899258102303788,
+      "grad_norm": 2.6128082275390625,
+      "learning_rate": 3.6398060836547404e-06,
+      "loss": 0.4531,
+      "step": 13752
+    },
+    {
+      "epoch": 1.7903162827020696,
+      "grad_norm": 2.659701108932495,
+      "learning_rate": 3.637795174408104e-06,
+      "loss": 0.5323,
+      "step": 13755
+    },
+    {
+      "epoch": 1.79070675517376,
+      "grad_norm": 2.875169277191162,
+      "learning_rate": 3.635784503135785e-06,
+      "loss": 0.5001,
+      "step": 13758
+    },
+    {
+      "epoch": 1.791097227645451,
+      "grad_norm": 3.0239455699920654,
+      "learning_rate": 3.6337740701890446e-06,
+      "loss": 0.4817,
+      "step": 13761
+    },
+    {
+      "epoch": 1.7914877001171416,
+      "grad_norm": 2.8980631828308105,
+      "learning_rate": 3.6317638759190985e-06,
+      "loss": 0.4957,
+      "step": 13764
+    },
+    {
+      "epoch": 1.7918781725888326,
+      "grad_norm": 2.978820562362671,
+      "learning_rate": 3.629753920677126e-06,
+      "loss": 0.5658,
+      "step": 13767
+    },
+    {
+      "epoch": 1.7922686450605232,
+      "grad_norm": 2.6764168739318848,
+      "learning_rate": 3.6277442048142615e-06,
+      "loss": 0.5507,
+      "step": 13770
+    },
+    {
+      "epoch": 1.792659117532214,
+      "grad_norm": 2.452681064605713,
+      "learning_rate": 3.6257347286815956e-06,
+      "loss": 0.4449,
+      "step": 13773
+    },
+    {
+      "epoch": 1.7930495900039047,
+      "grad_norm": 2.4262232780456543,
+      "learning_rate": 3.623725492630184e-06,
+      "loss": 0.4299,
+      "step": 13776
+    },
+    {
+      "epoch": 1.7934400624755955,
+      "grad_norm": 2.8017055988311768,
+      "learning_rate": 3.6217164970110296e-06,
+      "loss": 0.4115,
+      "step": 13779
+    },
+    {
+      "epoch": 1.7938305349472863,
+      "grad_norm": 2.725071430206299,
+      "learning_rate": 3.6197077421751077e-06,
+      "loss": 0.5487,
+      "step": 13782
+    },
+    {
+      "epoch": 1.7942210074189768,
+      "grad_norm": 3.4036777019500732,
+      "learning_rate": 3.6176992284733375e-06,
+      "loss": 0.4692,
+      "step": 13785
+    },
+    {
+      "epoch": 1.7946114798906678,
+      "grad_norm": 2.6664416790008545,
+      "learning_rate": 3.6156909562566054e-06,
+      "loss": 0.5513,
+      "step": 13788
+    },
+    {
+      "epoch": 1.7950019523623584,
+      "grad_norm": 3.166633129119873,
+      "learning_rate": 3.6136829258757503e-06,
+      "loss": 0.4979,
+      "step": 13791
+    },
+    {
+      "epoch": 1.7953924248340491,
+      "grad_norm": 2.793639898300171,
+      "learning_rate": 3.611675137681572e-06,
+      "loss": 0.5232,
+      "step": 13794
+    },
+    {
+      "epoch": 1.79578289730574,
+      "grad_norm": 2.91501784324646,
+      "learning_rate": 3.609667592024827e-06,
+      "loss": 0.516,
+      "step": 13797
+    },
+    {
+      "epoch": 1.7961733697774307,
+      "grad_norm": 2.355177164077759,
+      "learning_rate": 3.607660289256228e-06,
+      "loss": 0.5682,
+      "step": 13800
+    },
+    {
+      "epoch": 1.7965638422491215,
+      "grad_norm": 2.7519314289093018,
+      "learning_rate": 3.6056532297264486e-06,
+      "loss": 0.5224,
+      "step": 13803
+    },
+    {
+      "epoch": 1.796954314720812,
+      "grad_norm": 3.09912109375,
+      "learning_rate": 3.603646413786115e-06,
+      "loss": 0.502,
+      "step": 13806
+    },
+    {
+      "epoch": 1.797344787192503,
+      "grad_norm": 2.3987009525299072,
+      "learning_rate": 3.601639841785816e-06,
+      "loss": 0.4405,
+      "step": 13809
+    },
+    {
+      "epoch": 1.7977352596641936,
+      "grad_norm": 2.6418380737304688,
+      "learning_rate": 3.599633514076096e-06,
+      "loss": 0.4318,
+      "step": 13812
+    },
+    {
+      "epoch": 1.7981257321358846,
+      "grad_norm": 2.3674392700195312,
+      "learning_rate": 3.5976274310074536e-06,
+      "loss": 0.453,
+      "step": 13815
+    },
+    {
+      "epoch": 1.798516204607575,
+      "grad_norm": 2.514341354370117,
+      "learning_rate": 3.595621592930351e-06,
+      "loss": 0.4702,
+      "step": 13818
+    },
+    {
+      "epoch": 1.7989066770792659,
+      "grad_norm": 2.4524552822113037,
+      "learning_rate": 3.5936160001951977e-06,
+      "loss": 0.4447,
+      "step": 13821
+    },
+    {
+      "epoch": 1.7992971495509567,
+      "grad_norm": 2.8817389011383057,
+      "learning_rate": 3.5916106531523737e-06,
+      "loss": 0.5262,
+      "step": 13824
+    },
+    {
+      "epoch": 1.7996876220226474,
+      "grad_norm": 2.756680965423584,
+      "learning_rate": 3.5896055521522043e-06,
+      "loss": 0.5041,
+      "step": 13827
+    },
+    {
+      "epoch": 1.8000780944943382,
+      "grad_norm": 2.7653608322143555,
+      "learning_rate": 3.587600697544979e-06,
+      "loss": 0.5412,
+      "step": 13830
+    },
+    {
+      "epoch": 1.8004685669660287,
+      "grad_norm": 2.692890167236328,
+      "learning_rate": 3.585596089680941e-06,
+      "loss": 0.5134,
+      "step": 13833
+    },
+    {
+      "epoch": 1.8008590394377197,
+      "grad_norm": 2.555551767349243,
+      "learning_rate": 3.5835917289102893e-06,
+      "loss": 0.4865,
+      "step": 13836
+    },
+    {
+      "epoch": 1.8012495119094103,
+      "grad_norm": 2.5611484050750732,
+      "learning_rate": 3.5815876155831845e-06,
+      "loss": 0.4808,
+      "step": 13839
+    },
+    {
+      "epoch": 1.8016399843811013,
+      "grad_norm": 3.0679988861083984,
+      "learning_rate": 3.5795837500497388e-06,
+      "loss": 0.5976,
+      "step": 13842
+    },
+    {
+      "epoch": 1.8020304568527918,
+      "grad_norm": 3.132232427597046,
+      "learning_rate": 3.577580132660023e-06,
+      "loss": 0.4778,
+      "step": 13845
+    },
+    {
+      "epoch": 1.8024209293244826,
+      "grad_norm": 3.60653018951416,
+      "learning_rate": 3.575576763764067e-06,
+      "loss": 0.4976,
+      "step": 13848
+    },
+    {
+      "epoch": 1.8028114017961734,
+      "grad_norm": 2.553593873977661,
+      "learning_rate": 3.573573643711852e-06,
+      "loss": 0.4926,
+      "step": 13851
+    },
+    {
+      "epoch": 1.8032018742678642,
+      "grad_norm": 2.7795355319976807,
+      "learning_rate": 3.5715707728533227e-06,
+      "loss": 0.4787,
+      "step": 13854
+    },
+    {
+      "epoch": 1.803592346739555,
+      "grad_norm": 2.8556997776031494,
+      "learning_rate": 3.5695681515383727e-06,
+      "loss": 0.4926,
+      "step": 13857
+    },
+    {
+      "epoch": 1.8039828192112455,
+      "grad_norm": 2.585430383682251,
+      "learning_rate": 3.5675657801168583e-06,
+      "loss": 0.4774,
+      "step": 13860
+    },
+    {
+      "epoch": 1.8043732916829365,
+      "grad_norm": 3.1514101028442383,
+      "learning_rate": 3.5655636589385874e-06,
+      "loss": 0.6497,
+      "step": 13863
+    },
+    {
+      "epoch": 1.804763764154627,
+      "grad_norm": 2.7750942707061768,
+      "learning_rate": 3.563561788353329e-06,
+      "loss": 0.5409,
+      "step": 13866
+    },
+    {
+      "epoch": 1.8051542366263178,
+      "grad_norm": 2.840725898742676,
+      "learning_rate": 3.561560168710804e-06,
+      "loss": 0.6219,
+      "step": 13869
+    },
+    {
+      "epoch": 1.8055447090980086,
+      "grad_norm": 2.4959402084350586,
+      "learning_rate": 3.559558800360692e-06,
+      "loss": 0.475,
+      "step": 13872
+    },
+    {
+      "epoch": 1.8059351815696993,
+      "grad_norm": 2.7684502601623535,
+      "learning_rate": 3.557557683652627e-06,
+      "loss": 0.4395,
+      "step": 13875
+    },
+    {
+      "epoch": 1.8063256540413901,
+      "grad_norm": 3.1551716327667236,
+      "learning_rate": 3.5555568189362e-06,
+      "loss": 0.4665,
+      "step": 13878
+    },
+    {
+      "epoch": 1.8067161265130807,
+      "grad_norm": 3.5174217224121094,
+      "learning_rate": 3.5535562065609598e-06,
+      "loss": 0.4556,
+      "step": 13881
+    },
+    {
+      "epoch": 1.8071065989847717,
+      "grad_norm": 2.683762311935425,
+      "learning_rate": 3.551555846876405e-06,
+      "loss": 0.4735,
+      "step": 13884
+    },
+    {
+      "epoch": 1.8074970714564622,
+      "grad_norm": 2.77344012260437,
+      "learning_rate": 3.5495557402319975e-06,
+      "loss": 0.5262,
+      "step": 13887
+    },
+    {
+      "epoch": 1.8078875439281532,
+      "grad_norm": 3.604360342025757,
+      "learning_rate": 3.5475558869771516e-06,
+      "loss": 0.5683,
+      "step": 13890
+    },
+    {
+      "epoch": 1.8082780163998438,
+      "grad_norm": 2.6248981952667236,
+      "learning_rate": 3.545556287461236e-06,
+      "loss": 0.4836,
+      "step": 13893
+    },
+    {
+      "epoch": 1.8086684888715345,
+      "grad_norm": 3.0263054370880127,
+      "learning_rate": 3.543556942033577e-06,
+      "loss": 0.4698,
+      "step": 13896
+    },
+    {
+      "epoch": 1.8090589613432253,
+      "grad_norm": 2.9796009063720703,
+      "learning_rate": 3.5415578510434572e-06,
+      "loss": 0.4816,
+      "step": 13899
+    },
+    {
+      "epoch": 1.809449433814916,
+      "grad_norm": 2.5007407665252686,
+      "learning_rate": 3.539559014840112e-06,
+      "loss": 0.4536,
+      "step": 13902
+    },
+    {
+      "epoch": 1.8098399062866068,
+      "grad_norm": 4.0354323387146,
+      "learning_rate": 3.537560433772733e-06,
+      "loss": 0.5184,
+      "step": 13905
+    },
+    {
+      "epoch": 1.8102303787582974,
+      "grad_norm": 2.4867868423461914,
+      "learning_rate": 3.5355621081904717e-06,
+      "loss": 0.4836,
+      "step": 13908
+    },
+    {
+      "epoch": 1.8106208512299884,
+      "grad_norm": 2.469088315963745,
+      "learning_rate": 3.5335640384424296e-06,
+      "loss": 0.5388,
+      "step": 13911
+    },
+    {
+      "epoch": 1.811011323701679,
+      "grad_norm": 2.7055606842041016,
+      "learning_rate": 3.531566224877665e-06,
+      "loss": 0.4555,
+      "step": 13914
+    },
+    {
+      "epoch": 1.81140179617337,
+      "grad_norm": 2.7885193824768066,
+      "learning_rate": 3.529568667845192e-06,
+      "loss": 0.5323,
+      "step": 13917
+    },
+    {
+      "epoch": 1.8117922686450605,
+      "grad_norm": 2.801211357116699,
+      "learning_rate": 3.5275713676939782e-06,
+      "loss": 0.4274,
+      "step": 13920
+    },
+    {
+      "epoch": 1.8121827411167513,
+      "grad_norm": 2.6563968658447266,
+      "learning_rate": 3.525574324772949e-06,
+      "loss": 0.4206,
+      "step": 13923
+    },
+    {
+      "epoch": 1.812573213588442,
+      "grad_norm": 3.055948495864868,
+      "learning_rate": 3.523577539430985e-06,
+      "loss": 0.4404,
+      "step": 13926
+    },
+    {
+      "epoch": 1.8129636860601328,
+      "grad_norm": 2.6117143630981445,
+      "learning_rate": 3.5215810120169182e-06,
+      "loss": 0.4857,
+      "step": 13929
+    },
+    {
+      "epoch": 1.8133541585318236,
+      "grad_norm": 2.4713497161865234,
+      "learning_rate": 3.5195847428795388e-06,
+      "loss": 0.4886,
+      "step": 13932
+    },
+    {
+      "epoch": 1.8137446310035141,
+      "grad_norm": 2.4422478675842285,
+      "learning_rate": 3.5175887323675896e-06,
+      "loss": 0.5087,
+      "step": 13935
+    },
+    {
+      "epoch": 1.8141351034752051,
+      "grad_norm": 2.721740484237671,
+      "learning_rate": 3.5155929808297706e-06,
+      "loss": 0.4973,
+      "step": 13938
+    },
+    {
+      "epoch": 1.8145255759468957,
+      "grad_norm": 2.5684258937835693,
+      "learning_rate": 3.5135974886147358e-06,
+      "loss": 0.4052,
+      "step": 13941
+    },
+    {
+      "epoch": 1.8149160484185864,
+      "grad_norm": 2.683708667755127,
+      "learning_rate": 3.5116022560710916e-06,
+      "loss": 0.4971,
+      "step": 13944
+    },
+    {
+      "epoch": 1.8153065208902772,
+      "grad_norm": 2.9392552375793457,
+      "learning_rate": 3.5096072835474015e-06,
+      "loss": 0.4841,
+      "step": 13947
+    },
+    {
+      "epoch": 1.815696993361968,
+      "grad_norm": 2.6948726177215576,
+      "learning_rate": 3.5076125713921844e-06,
+      "loss": 0.5777,
+      "step": 13950
+    },
+    {
+      "epoch": 1.8160874658336588,
+      "grad_norm": 2.9752037525177,
+      "learning_rate": 3.505618119953913e-06,
+      "loss": 0.4687,
+      "step": 13953
+    },
+    {
+      "epoch": 1.8164779383053493,
+      "grad_norm": 2.543264865875244,
+      "learning_rate": 3.5036239295810105e-06,
+      "loss": 0.4773,
+      "step": 13956
+    },
+    {
+      "epoch": 1.8168684107770403,
+      "grad_norm": 2.5448291301727295,
+      "learning_rate": 3.5016300006218607e-06,
+      "loss": 0.5217,
+      "step": 13959
+    },
+    {
+      "epoch": 1.8172588832487309,
+      "grad_norm": 2.9891393184661865,
+      "learning_rate": 3.4996363334247975e-06,
+      "loss": 0.4791,
+      "step": 13962
+    },
+    {
+      "epoch": 1.8176493557204219,
+      "grad_norm": 2.651190757751465,
+      "learning_rate": 3.49764292833811e-06,
+      "loss": 0.5361,
+      "step": 13965
+    },
+    {
+      "epoch": 1.8180398281921124,
+      "grad_norm": 2.7708990573883057,
+      "learning_rate": 3.4956497857100437e-06,
+      "loss": 0.4645,
+      "step": 13968
+    },
+    {
+      "epoch": 1.8184303006638032,
+      "grad_norm": 2.510695457458496,
+      "learning_rate": 3.493656905888794e-06,
+      "loss": 0.4663,
+      "step": 13971
+    },
+    {
+      "epoch": 1.818820773135494,
+      "grad_norm": 3.075878381729126,
+      "learning_rate": 3.4916642892225138e-06,
+      "loss": 0.5056,
+      "step": 13974
+    },
+    {
+      "epoch": 1.8192112456071847,
+      "grad_norm": 2.6439952850341797,
+      "learning_rate": 3.4896719360593106e-06,
+      "loss": 0.533,
+      "step": 13977
+    },
+    {
+      "epoch": 1.8196017180788755,
+      "grad_norm": 2.594844102859497,
+      "learning_rate": 3.4876798467472415e-06,
+      "loss": 0.4981,
+      "step": 13980
+    },
+    {
+      "epoch": 1.819992190550566,
+      "grad_norm": 3.148606538772583,
+      "learning_rate": 3.4856880216343235e-06,
+      "loss": 0.5199,
+      "step": 13983
+    },
+    {
+      "epoch": 1.820382663022257,
+      "grad_norm": 2.5869855880737305,
+      "learning_rate": 3.4836964610685207e-06,
+      "loss": 0.437,
+      "step": 13986
+    },
+    {
+      "epoch": 1.8207731354939476,
+      "grad_norm": 2.792637586593628,
+      "learning_rate": 3.4817051653977553e-06,
+      "loss": 0.5815,
+      "step": 13989
+    },
+    {
+      "epoch": 1.8211636079656386,
+      "grad_norm": 2.534087896347046,
+      "learning_rate": 3.479714134969905e-06,
+      "loss": 0.4854,
+      "step": 13992
+    },
+    {
+      "epoch": 1.8215540804373291,
+      "grad_norm": 4.054605960845947,
+      "learning_rate": 3.4777233701327974e-06,
+      "loss": 0.5168,
+      "step": 13995
+    },
+    {
+      "epoch": 1.82194455290902,
+      "grad_norm": 2.612586498260498,
+      "learning_rate": 3.4757328712342143e-06,
+      "loss": 0.4172,
+      "step": 13998
+    },
+    {
+      "epoch": 1.8223350253807107,
+      "grad_norm": 2.6771693229675293,
+      "learning_rate": 3.4737426386218913e-06,
+      "loss": 0.4638,
+      "step": 14001
+    },
+    {
+      "epoch": 1.8227254978524015,
+      "grad_norm": 2.4215550422668457,
+      "learning_rate": 3.4717526726435204e-06,
+      "loss": 0.5155,
+      "step": 14004
+    },
+    {
+      "epoch": 1.8231159703240922,
+      "grad_norm": 2.9423484802246094,
+      "learning_rate": 3.469762973646741e-06,
+      "loss": 0.4784,
+      "step": 14007
+    },
+    {
+      "epoch": 1.8235064427957828,
+      "grad_norm": 3.212175130844116,
+      "learning_rate": 3.4677735419791507e-06,
+      "loss": 0.4752,
+      "step": 14010
+    },
+    {
+      "epoch": 1.8238969152674738,
+      "grad_norm": 2.640934944152832,
+      "learning_rate": 3.465784377988301e-06,
+      "loss": 0.5054,
+      "step": 14013
+    },
+    {
+      "epoch": 1.8242873877391643,
+      "grad_norm": 2.4397079944610596,
+      "learning_rate": 3.4637954820216914e-06,
+      "loss": 0.4716,
+      "step": 14016
+    },
+    {
+      "epoch": 1.824677860210855,
+      "grad_norm": 2.737032413482666,
+      "learning_rate": 3.4618068544267806e-06,
+      "loss": 0.547,
+      "step": 14019
+    },
+    {
+      "epoch": 1.8250683326825459,
+      "grad_norm": 2.630606174468994,
+      "learning_rate": 3.459818495550976e-06,
+      "loss": 0.5541,
+      "step": 14022
+    },
+    {
+      "epoch": 1.8254588051542366,
+      "grad_norm": 2.5901596546173096,
+      "learning_rate": 3.4578304057416394e-06,
+      "loss": 0.4804,
+      "step": 14025
+    },
+    {
+      "epoch": 1.8258492776259274,
+      "grad_norm": 2.6611945629119873,
+      "learning_rate": 3.455842585346088e-06,
+      "loss": 0.5694,
+      "step": 14028
+    },
+    {
+      "epoch": 1.826239750097618,
+      "grad_norm": 2.6162936687469482,
+      "learning_rate": 3.4538550347115863e-06,
+      "loss": 0.5321,
+      "step": 14031
+    },
+    {
+      "epoch": 1.826630222569309,
+      "grad_norm": 2.7024919986724854,
+      "learning_rate": 3.4518677541853584e-06,
+      "loss": 0.4661,
+      "step": 14034
+    },
+    {
+      "epoch": 1.8270206950409995,
+      "grad_norm": 2.514699697494507,
+      "learning_rate": 3.4498807441145775e-06,
+      "loss": 0.5,
+      "step": 14037
+    },
+    {
+      "epoch": 1.8274111675126905,
+      "grad_norm": 2.608577013015747,
+      "learning_rate": 3.4478940048463705e-06,
+      "loss": 0.4691,
+      "step": 14040
+    },
+    {
+      "epoch": 1.827801639984381,
+      "grad_norm": 2.8621628284454346,
+      "learning_rate": 3.445907536727814e-06,
+      "loss": 0.4189,
+      "step": 14043
+    },
+    {
+      "epoch": 1.8281921124560718,
+      "grad_norm": 2.61087703704834,
+      "learning_rate": 3.4439213401059436e-06,
+      "loss": 0.4572,
+      "step": 14046
+    },
+    {
+      "epoch": 1.8285825849277626,
+      "grad_norm": 2.482490062713623,
+      "learning_rate": 3.4419354153277398e-06,
+      "loss": 0.4796,
+      "step": 14049
+    },
+    {
+      "epoch": 1.8289730573994534,
+      "grad_norm": 2.7227578163146973,
+      "learning_rate": 3.4399497627401414e-06,
+      "loss": 0.4849,
+      "step": 14052
+    },
+    {
+      "epoch": 1.8293635298711441,
+      "grad_norm": 2.800520181655884,
+      "learning_rate": 3.437964382690039e-06,
+      "loss": 0.4679,
+      "step": 14055
+    },
+    {
+      "epoch": 1.8297540023428347,
+      "grad_norm": 2.624976396560669,
+      "learning_rate": 3.4359792755242716e-06,
+      "loss": 0.5038,
+      "step": 14058
+    },
+    {
+      "epoch": 1.8301444748145257,
+      "grad_norm": 2.559096097946167,
+      "learning_rate": 3.4339944415896354e-06,
+      "loss": 0.4533,
+      "step": 14061
+    },
+    {
+      "epoch": 1.8305349472862162,
+      "grad_norm": 2.5498814582824707,
+      "learning_rate": 3.432009881232875e-06,
+      "loss": 0.4878,
+      "step": 14064
+    },
+    {
+      "epoch": 1.830925419757907,
+      "grad_norm": 2.93689227104187,
+      "learning_rate": 3.4300255948006893e-06,
+      "loss": 0.4776,
+      "step": 14067
+    },
+    {
+      "epoch": 1.8313158922295978,
+      "grad_norm": 2.3573851585388184,
+      "learning_rate": 3.4280415826397304e-06,
+      "loss": 0.4692,
+      "step": 14070
+    },
+    {
+      "epoch": 1.8317063647012886,
+      "grad_norm": 2.303317070007324,
+      "learning_rate": 3.426057845096598e-06,
+      "loss": 0.4765,
+      "step": 14073
+    },
+    {
+      "epoch": 1.8320968371729793,
+      "grad_norm": 2.6520111560821533,
+      "learning_rate": 3.4240743825178514e-06,
+      "loss": 0.4867,
+      "step": 14076
+    },
+    {
+      "epoch": 1.83248730964467,
+      "grad_norm": 2.8947081565856934,
+      "learning_rate": 3.4220911952499943e-06,
+      "loss": 0.5239,
+      "step": 14079
+    },
+    {
+      "epoch": 1.8328777821163609,
+      "grad_norm": 2.4253430366516113,
+      "learning_rate": 3.4201082836394868e-06,
+      "loss": 0.5101,
+      "step": 14082
+    },
+    {
+      "epoch": 1.8332682545880514,
+      "grad_norm": 3.378649950027466,
+      "learning_rate": 3.418125648032737e-06,
+      "loss": 0.4679,
+      "step": 14085
+    },
+    {
+      "epoch": 1.8336587270597424,
+      "grad_norm": 2.7288544178009033,
+      "learning_rate": 3.4161432887761093e-06,
+      "loss": 0.4794,
+      "step": 14088
+    },
+    {
+      "epoch": 1.834049199531433,
+      "grad_norm": 2.832486152648926,
+      "learning_rate": 3.414161206215918e-06,
+      "loss": 0.5284,
+      "step": 14091
+    },
+    {
+      "epoch": 1.8344396720031237,
+      "grad_norm": 2.27494215965271,
+      "learning_rate": 3.4121794006984265e-06,
+      "loss": 0.4875,
+      "step": 14094
+    },
+    {
+      "epoch": 1.8348301444748145,
+      "grad_norm": 2.7032313346862793,
+      "learning_rate": 3.4101978725698553e-06,
+      "loss": 0.4654,
+      "step": 14097
+    },
+    {
+      "epoch": 1.8352206169465053,
+      "grad_norm": 2.672342300415039,
+      "learning_rate": 3.40821662217637e-06,
+      "loss": 0.4787,
+      "step": 14100
+    },
+    {
+      "epoch": 1.835611089418196,
+      "grad_norm": 2.9763712882995605,
+      "learning_rate": 3.4062356498640915e-06,
+      "loss": 0.4966,
+      "step": 14103
+    },
+    {
+      "epoch": 1.8360015618898866,
+      "grad_norm": 2.8209915161132812,
+      "learning_rate": 3.4042549559790938e-06,
+      "loss": 0.5311,
+      "step": 14106
+    },
+    {
+      "epoch": 1.8363920343615776,
+      "grad_norm": 3.304166078567505,
+      "learning_rate": 3.4022745408673973e-06,
+      "loss": 0.4676,
+      "step": 14109
+    },
+    {
+      "epoch": 1.8367825068332682,
+      "grad_norm": 2.846513271331787,
+      "learning_rate": 3.400294404874978e-06,
+      "loss": 0.4849,
+      "step": 14112
+    },
+    {
+      "epoch": 1.8371729793049592,
+      "grad_norm": 2.843594789505005,
+      "learning_rate": 3.3983145483477582e-06,
+      "loss": 0.5584,
+      "step": 14115
+    },
+    {
+      "epoch": 1.8375634517766497,
+      "grad_norm": 2.5627312660217285,
+      "learning_rate": 3.39633497163162e-06,
+      "loss": 0.4594,
+      "step": 14118
+    },
+    {
+      "epoch": 1.8379539242483405,
+      "grad_norm": 2.5003201961517334,
+      "learning_rate": 3.394355675072388e-06,
+      "loss": 0.425,
+      "step": 14121
+    },
+    {
+      "epoch": 1.8383443967200312,
+      "grad_norm": 2.563779354095459,
+      "learning_rate": 3.3923766590158425e-06,
+      "loss": 0.5306,
+      "step": 14124
+    },
+    {
+      "epoch": 1.838734869191722,
+      "grad_norm": 2.529606819152832,
+      "learning_rate": 3.3903979238077124e-06,
+      "loss": 0.5286,
+      "step": 14127
+    },
+    {
+      "epoch": 1.8391253416634128,
+      "grad_norm": 2.8574764728546143,
+      "learning_rate": 3.3884194697936777e-06,
+      "loss": 0.5582,
+      "step": 14130
+    },
+    {
+      "epoch": 1.8395158141351033,
+      "grad_norm": 2.485995292663574,
+      "learning_rate": 3.3864412973193734e-06,
+      "loss": 0.4713,
+      "step": 14133
+    },
+    {
+      "epoch": 1.8399062866067943,
+      "grad_norm": 3.2460153102874756,
+      "learning_rate": 3.3844634067303783e-06,
+      "loss": 0.4945,
+      "step": 14136
+    },
+    {
+      "epoch": 1.840296759078485,
+      "grad_norm": 3.2644894123077393,
+      "learning_rate": 3.382485798372228e-06,
+      "loss": 0.524,
+      "step": 14139
+    },
+    {
+      "epoch": 1.8406872315501757,
+      "grad_norm": 2.7396457195281982,
+      "learning_rate": 3.380508472590407e-06,
+      "loss": 0.4508,
+      "step": 14142
+    },
+    {
+      "epoch": 1.8410777040218664,
+      "grad_norm": 2.678882598876953,
+      "learning_rate": 3.3785314297303477e-06,
+      "loss": 0.4521,
+      "step": 14145
+    },
+    {
+      "epoch": 1.8414681764935572,
+      "grad_norm": 2.4532623291015625,
+      "learning_rate": 3.3765546701374375e-06,
+      "loss": 0.4508,
+      "step": 14148
+    },
+    {
+      "epoch": 1.841858648965248,
+      "grad_norm": 2.8880367279052734,
+      "learning_rate": 3.3745781941570104e-06,
+      "loss": 0.4395,
+      "step": 14151
+    },
+    {
+      "epoch": 1.8422491214369385,
+      "grad_norm": 2.6723203659057617,
+      "learning_rate": 3.372602002134353e-06,
+      "loss": 0.4853,
+      "step": 14154
+    },
+    {
+      "epoch": 1.8426395939086295,
+      "grad_norm": 2.651268243789673,
+      "learning_rate": 3.370626094414702e-06,
+      "loss": 0.4853,
+      "step": 14157
+    },
+    {
+      "epoch": 1.84303006638032,
+      "grad_norm": 2.4674911499023438,
+      "learning_rate": 3.368650471343246e-06,
+      "loss": 0.5133,
+      "step": 14160
+    },
+    {
+      "epoch": 1.843420538852011,
+      "grad_norm": 2.827803373336792,
+      "learning_rate": 3.36667513326512e-06,
+      "loss": 0.4601,
+      "step": 14163
+    },
+    {
+      "epoch": 1.8438110113237016,
+      "grad_norm": 3.614109754562378,
+      "learning_rate": 3.364700080525412e-06,
+      "loss": 0.5692,
+      "step": 14166
+    },
+    {
+      "epoch": 1.8442014837953924,
+      "grad_norm": 3.3453476428985596,
+      "learning_rate": 3.362725313469161e-06,
+      "loss": 0.512,
+      "step": 14169
+    },
+    {
+      "epoch": 1.8445919562670832,
+      "grad_norm": 2.600367546081543,
+      "learning_rate": 3.3607508324413525e-06,
+      "loss": 0.5574,
+      "step": 14172
+    },
+    {
+      "epoch": 1.844982428738774,
+      "grad_norm": 2.761702299118042,
+      "learning_rate": 3.3587766377869256e-06,
+      "loss": 0.4362,
+      "step": 14175
+    },
+    {
+      "epoch": 1.8453729012104647,
+      "grad_norm": 2.6490187644958496,
+      "learning_rate": 3.3568027298507673e-06,
+      "loss": 0.5518,
+      "step": 14178
+    },
+    {
+      "epoch": 1.8457633736821553,
+      "grad_norm": 2.837568521499634,
+      "learning_rate": 3.3548291089777146e-06,
+      "loss": 0.4928,
+      "step": 14181
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 2.8118865489959717,
+      "learning_rate": 3.352855775512557e-06,
+      "loss": 0.5401,
+      "step": 14184
+    },
+    {
+      "epoch": 1.8465443186255368,
+      "grad_norm": 2.6508400440216064,
+      "learning_rate": 3.350882729800029e-06,
+      "loss": 0.5607,
+      "step": 14187
+    },
+    {
+      "epoch": 1.8469347910972278,
+      "grad_norm": 3.0374929904937744,
+      "learning_rate": 3.3489099721848188e-06,
+      "loss": 0.4713,
+      "step": 14190
+    },
+    {
+      "epoch": 1.8473252635689184,
+      "grad_norm": 2.809018135070801,
+      "learning_rate": 3.346937503011565e-06,
+      "loss": 0.5143,
+      "step": 14193
+    },
+    {
+      "epoch": 1.8477157360406091,
+      "grad_norm": 2.843602418899536,
+      "learning_rate": 3.34496532262485e-06,
+      "loss": 0.4967,
+      "step": 14196
+    },
+    {
+      "epoch": 1.8481062085123,
+      "grad_norm": 2.7359254360198975,
+      "learning_rate": 3.3429934313692102e-06,
+      "loss": 0.527,
+      "step": 14199
+    },
+    {
+      "epoch": 1.8484966809839907,
+      "grad_norm": 3.020136594772339,
+      "learning_rate": 3.341021829589134e-06,
+      "loss": 0.5108,
+      "step": 14202
+    },
+    {
+      "epoch": 1.8488871534556814,
+      "grad_norm": 2.610989809036255,
+      "learning_rate": 3.3390505176290544e-06,
+      "loss": 0.56,
+      "step": 14205
+    },
+    {
+      "epoch": 1.849277625927372,
+      "grad_norm": 3.2436866760253906,
+      "learning_rate": 3.3370794958333547e-06,
+      "loss": 0.4896,
+      "step": 14208
+    },
+    {
+      "epoch": 1.849668098399063,
+      "grad_norm": 2.493400812149048,
+      "learning_rate": 3.33510876454637e-06,
+      "loss": 0.463,
+      "step": 14211
+    },
+    {
+      "epoch": 1.8500585708707535,
+      "grad_norm": 2.6635570526123047,
+      "learning_rate": 3.33313832411238e-06,
+      "loss": 0.4602,
+      "step": 14214
+    },
+    {
+      "epoch": 1.8504490433424443,
+      "grad_norm": 2.576613664627075,
+      "learning_rate": 3.3311681748756185e-06,
+      "loss": 0.4683,
+      "step": 14217
+    },
+    {
+      "epoch": 1.850839515814135,
+      "grad_norm": 2.853515148162842,
+      "learning_rate": 3.3291983171802676e-06,
+      "loss": 0.5186,
+      "step": 14220
+    },
+    {
+      "epoch": 1.8512299882858259,
+      "grad_norm": 2.5207786560058594,
+      "learning_rate": 3.3272287513704544e-06,
+      "loss": 0.4683,
+      "step": 14223
+    },
+    {
+      "epoch": 1.8516204607575166,
+      "grad_norm": 2.8509247303009033,
+      "learning_rate": 3.32525947779026e-06,
+      "loss": 0.4833,
+      "step": 14226
+    },
+    {
+      "epoch": 1.8520109332292072,
+      "grad_norm": 4.261683464050293,
+      "learning_rate": 3.3232904967837116e-06,
+      "loss": 0.5181,
+      "step": 14229
+    },
+    {
+      "epoch": 1.8524014057008982,
+      "grad_norm": 2.9350645542144775,
+      "learning_rate": 3.3213218086947857e-06,
+      "loss": 0.4807,
+      "step": 14232
+    },
+    {
+      "epoch": 1.8527918781725887,
+      "grad_norm": 2.8778061866760254,
+      "learning_rate": 3.3193534138674094e-06,
+      "loss": 0.4726,
+      "step": 14235
+    },
+    {
+      "epoch": 1.8531823506442797,
+      "grad_norm": 3.9157052040100098,
+      "learning_rate": 3.3173853126454546e-06,
+      "loss": 0.5698,
+      "step": 14238
+    },
+    {
+      "epoch": 1.8535728231159703,
+      "grad_norm": 2.5443649291992188,
+      "learning_rate": 3.3154175053727478e-06,
+      "loss": 0.5034,
+      "step": 14241
+    },
+    {
+      "epoch": 1.853963295587661,
+      "grad_norm": 2.9121346473693848,
+      "learning_rate": 3.313449992393055e-06,
+      "loss": 0.5932,
+      "step": 14244
+    },
+    {
+      "epoch": 1.8543537680593518,
+      "grad_norm": 2.4990174770355225,
+      "learning_rate": 3.311482774050105e-06,
+      "loss": 0.4869,
+      "step": 14247
+    },
+    {
+      "epoch": 1.8547442405310426,
+      "grad_norm": 3.1738932132720947,
+      "learning_rate": 3.3095158506875603e-06,
+      "loss": 0.6015,
+      "step": 14250
+    },
+    {
+      "epoch": 1.8551347130027334,
+      "grad_norm": 2.808356285095215,
+      "learning_rate": 3.3075492226490404e-06,
+      "loss": 0.4492,
+      "step": 14253
+    },
+    {
+      "epoch": 1.855525185474424,
+      "grad_norm": 2.824268102645874,
+      "learning_rate": 3.3055828902781116e-06,
+      "loss": 0.5146,
+      "step": 14256
+    },
+    {
+      "epoch": 1.855915657946115,
+      "grad_norm": 2.653419017791748,
+      "learning_rate": 3.3036168539182867e-06,
+      "loss": 0.5002,
+      "step": 14259
+    },
+    {
+      "epoch": 1.8563061304178055,
+      "grad_norm": 2.6380562782287598,
+      "learning_rate": 3.3016511139130296e-06,
+      "loss": 0.4467,
+      "step": 14262
+    },
+    {
+      "epoch": 1.8566966028894965,
+      "grad_norm": 3.018332004547119,
+      "learning_rate": 3.299685670605749e-06,
+      "loss": 0.5476,
+      "step": 14265
+    },
+    {
+      "epoch": 1.857087075361187,
+      "grad_norm": 2.474071979522705,
+      "learning_rate": 3.297720524339805e-06,
+      "loss": 0.5168,
+      "step": 14268
+    },
+    {
+      "epoch": 1.8574775478328778,
+      "grad_norm": 2.407060384750366,
+      "learning_rate": 3.295755675458505e-06,
+      "loss": 0.44,
+      "step": 14271
+    },
+    {
+      "epoch": 1.8578680203045685,
+      "grad_norm": 2.6494884490966797,
+      "learning_rate": 3.2937911243051035e-06,
+      "loss": 0.56,
+      "step": 14274
+    },
+    {
+      "epoch": 1.8582584927762593,
+      "grad_norm": 2.5988290309906006,
+      "learning_rate": 3.291826871222803e-06,
+      "loss": 0.458,
+      "step": 14277
+    },
+    {
+      "epoch": 1.85864896524795,
+      "grad_norm": 3.0355000495910645,
+      "learning_rate": 3.2898629165547534e-06,
+      "loss": 0.497,
+      "step": 14280
+    },
+    {
+      "epoch": 1.8590394377196406,
+      "grad_norm": 2.6298789978027344,
+      "learning_rate": 3.2878992606440557e-06,
+      "loss": 0.537,
+      "step": 14283
+    },
+    {
+      "epoch": 1.8594299101913316,
+      "grad_norm": 3.61897349357605,
+      "learning_rate": 3.2859359038337537e-06,
+      "loss": 0.5495,
+      "step": 14286
+    },
+    {
+      "epoch": 1.8598203826630222,
+      "grad_norm": 2.8695971965789795,
+      "learning_rate": 3.283972846466846e-06,
+      "loss": 0.4939,
+      "step": 14289
+    },
+    {
+      "epoch": 1.860210855134713,
+      "grad_norm": 2.84236741065979,
+      "learning_rate": 3.2820100888862703e-06,
+      "loss": 0.4501,
+      "step": 14292
+    },
+    {
+      "epoch": 1.8606013276064037,
+      "grad_norm": 2.9210715293884277,
+      "learning_rate": 3.2800476314349184e-06,
+      "loss": 0.5261,
+      "step": 14295
+    },
+    {
+      "epoch": 1.8609918000780945,
+      "grad_norm": 2.3659050464630127,
+      "learning_rate": 3.2780854744556284e-06,
+      "loss": 0.4849,
+      "step": 14298
+    },
+    {
+      "epoch": 1.8613822725497853,
+      "grad_norm": 2.6904006004333496,
+      "learning_rate": 3.276123618291182e-06,
+      "loss": 0.4635,
+      "step": 14301
+    },
+    {
+      "epoch": 1.8617727450214758,
+      "grad_norm": 3.254288911819458,
+      "learning_rate": 3.274162063284314e-06,
+      "loss": 0.4929,
+      "step": 14304
+    },
+    {
+      "epoch": 1.8621632174931668,
+      "grad_norm": 2.973808526992798,
+      "learning_rate": 3.2722008097777025e-06,
+      "loss": 0.4474,
+      "step": 14307
+    },
+    {
+      "epoch": 1.8625536899648574,
+      "grad_norm": 2.9174914360046387,
+      "learning_rate": 3.2702398581139742e-06,
+      "loss": 0.5224,
+      "step": 14310
+    },
+    {
+      "epoch": 1.8629441624365484,
+      "grad_norm": 2.607135057449341,
+      "learning_rate": 3.268279208635705e-06,
+      "loss": 0.4667,
+      "step": 14313
+    },
+    {
+      "epoch": 1.863334634908239,
+      "grad_norm": 2.792543649673462,
+      "learning_rate": 3.266318861685414e-06,
+      "loss": 0.4632,
+      "step": 14316
+    },
+    {
+      "epoch": 1.8637251073799297,
+      "grad_norm": 2.533592462539673,
+      "learning_rate": 3.2643588176055706e-06,
+      "loss": 0.4745,
+      "step": 14319
+    },
+    {
+      "epoch": 1.8641155798516205,
+      "grad_norm": 2.8035097122192383,
+      "learning_rate": 3.2623990767385923e-06,
+      "loss": 0.4732,
+      "step": 14322
+    },
+    {
+      "epoch": 1.8645060523233112,
+      "grad_norm": 2.933763265609741,
+      "learning_rate": 3.2604396394268377e-06,
+      "loss": 0.567,
+      "step": 14325
+    },
+    {
+      "epoch": 1.864896524795002,
+      "grad_norm": 2.4261467456817627,
+      "learning_rate": 3.2584805060126183e-06,
+      "loss": 0.429,
+      "step": 14328
+    },
+    {
+      "epoch": 1.8652869972666926,
+      "grad_norm": 2.9292244911193848,
+      "learning_rate": 3.2565216768381924e-06,
+      "loss": 0.4519,
+      "step": 14331
+    },
+    {
+      "epoch": 1.8656774697383836,
+      "grad_norm": 2.7582828998565674,
+      "learning_rate": 3.2545631522457623e-06,
+      "loss": 0.5999,
+      "step": 14334
+    },
+    {
+      "epoch": 1.866067942210074,
+      "grad_norm": 2.7902820110321045,
+      "learning_rate": 3.2526049325774762e-06,
+      "loss": 0.4989,
+      "step": 14337
+    },
+    {
+      "epoch": 1.866458414681765,
+      "grad_norm": 2.306330442428589,
+      "learning_rate": 3.2506470181754336e-06,
+      "loss": 0.529,
+      "step": 14340
+    },
+    {
+      "epoch": 1.8668488871534556,
+      "grad_norm": 2.968900442123413,
+      "learning_rate": 3.2486894093816755e-06,
+      "loss": 0.5525,
+      "step": 14343
+    },
+    {
+      "epoch": 1.8672393596251464,
+      "grad_norm": 2.6159310340881348,
+      "learning_rate": 3.246732106538194e-06,
+      "loss": 0.4139,
+      "step": 14346
+    },
+    {
+      "epoch": 1.8676298320968372,
+      "grad_norm": 2.9599854946136475,
+      "learning_rate": 3.2447751099869264e-06,
+      "loss": 0.5086,
+      "step": 14349
+    },
+    {
+      "epoch": 1.868020304568528,
+      "grad_norm": 2.811408519744873,
+      "learning_rate": 3.242818420069753e-06,
+      "loss": 0.4619,
+      "step": 14352
+    },
+    {
+      "epoch": 1.8684107770402187,
+      "grad_norm": 3.143258810043335,
+      "learning_rate": 3.240862037128506e-06,
+      "loss": 0.5827,
+      "step": 14355
+    },
+    {
+      "epoch": 1.8688012495119093,
+      "grad_norm": 2.813084840774536,
+      "learning_rate": 3.23890596150496e-06,
+      "loss": 0.497,
+      "step": 14358
+    },
+    {
+      "epoch": 1.8691917219836003,
+      "grad_norm": 2.9719557762145996,
+      "learning_rate": 3.2369501935408375e-06,
+      "loss": 0.5179,
+      "step": 14361
+    },
+    {
+      "epoch": 1.8695821944552908,
+      "grad_norm": 3.1469063758850098,
+      "learning_rate": 3.234994733577808e-06,
+      "loss": 0.4145,
+      "step": 14364
+    },
+    {
+      "epoch": 1.8699726669269816,
+      "grad_norm": 2.816202402114868,
+      "learning_rate": 3.2330395819574845e-06,
+      "loss": 0.4828,
+      "step": 14367
+    },
+    {
+      "epoch": 1.8703631393986724,
+      "grad_norm": 2.6076595783233643,
+      "learning_rate": 3.2310847390214283e-06,
+      "loss": 0.5034,
+      "step": 14370
+    },
+    {
+      "epoch": 1.8707536118703632,
+      "grad_norm": 3.1273179054260254,
+      "learning_rate": 3.229130205111147e-06,
+      "loss": 0.4748,
+      "step": 14373
+    },
+    {
+      "epoch": 1.871144084342054,
+      "grad_norm": 2.559382438659668,
+      "learning_rate": 3.2271759805680956e-06,
+      "loss": 0.4771,
+      "step": 14376
+    },
+    {
+      "epoch": 1.8715345568137445,
+      "grad_norm": 2.720404624938965,
+      "learning_rate": 3.22522206573367e-06,
+      "loss": 0.4631,
+      "step": 14379
+    },
+    {
+      "epoch": 1.8719250292854355,
+      "grad_norm": 3.042336940765381,
+      "learning_rate": 3.223268460949215e-06,
+      "loss": 0.5467,
+      "step": 14382
+    },
+    {
+      "epoch": 1.872315501757126,
+      "grad_norm": 2.677138090133667,
+      "learning_rate": 3.221315166556024e-06,
+      "loss": 0.4852,
+      "step": 14385
+    },
+    {
+      "epoch": 1.872705974228817,
+      "grad_norm": 3.1034047603607178,
+      "learning_rate": 3.21936218289533e-06,
+      "loss": 0.5535,
+      "step": 14388
+    },
+    {
+      "epoch": 1.8730964467005076,
+      "grad_norm": 3.5317320823669434,
+      "learning_rate": 3.217409510308318e-06,
+      "loss": 0.4424,
+      "step": 14391
+    },
+    {
+      "epoch": 1.8734869191721983,
+      "grad_norm": 3.066141366958618,
+      "learning_rate": 3.215457149136114e-06,
+      "loss": 0.454,
+      "step": 14394
+    },
+    {
+      "epoch": 1.8738773916438891,
+      "grad_norm": 2.683347702026367,
+      "learning_rate": 3.213505099719791e-06,
+      "loss": 0.4301,
+      "step": 14397
+    },
+    {
+      "epoch": 1.8742678641155799,
+      "grad_norm": 2.6969237327575684,
+      "learning_rate": 3.2115533624003703e-06,
+      "loss": 0.5306,
+      "step": 14400
+    },
+    {
+      "epoch": 1.8746583365872707,
+      "grad_norm": 2.678342580795288,
+      "learning_rate": 3.2096019375188135e-06,
+      "loss": 0.4622,
+      "step": 14403
+    },
+    {
+      "epoch": 1.8750488090589612,
+      "grad_norm": 2.750361919403076,
+      "learning_rate": 3.2076508254160334e-06,
+      "loss": 0.4639,
+      "step": 14406
+    },
+    {
+      "epoch": 1.8754392815306522,
+      "grad_norm": 2.4361095428466797,
+      "learning_rate": 3.205700026432882e-06,
+      "loss": 0.4648,
+      "step": 14409
+    },
+    {
+      "epoch": 1.8758297540023428,
+      "grad_norm": 2.7799482345581055,
+      "learning_rate": 3.2037495409101603e-06,
+      "loss": 0.5366,
+      "step": 14412
+    },
+    {
+      "epoch": 1.8762202264740335,
+      "grad_norm": 2.5228137969970703,
+      "learning_rate": 3.201799369188616e-06,
+      "loss": 0.5324,
+      "step": 14415
+    },
+    {
+      "epoch": 1.8766106989457243,
+      "grad_norm": 2.955665349960327,
+      "learning_rate": 3.1998495116089413e-06,
+      "loss": 0.528,
+      "step": 14418
+    },
+    {
+      "epoch": 1.877001171417415,
+      "grad_norm": 3.2559332847595215,
+      "learning_rate": 3.1978999685117674e-06,
+      "loss": 0.4378,
+      "step": 14421
+    },
+    {
+      "epoch": 1.8773916438891058,
+      "grad_norm": 3.0776727199554443,
+      "learning_rate": 3.1959507402376787e-06,
+      "loss": 0.4957,
+      "step": 14424
+    },
+    {
+      "epoch": 1.8777821163607966,
+      "grad_norm": 2.5709543228149414,
+      "learning_rate": 3.1940018271272013e-06,
+      "loss": 0.5957,
+      "step": 14427
+    },
+    {
+      "epoch": 1.8781725888324874,
+      "grad_norm": 2.643850564956665,
+      "learning_rate": 3.1920532295208045e-06,
+      "loss": 0.5714,
+      "step": 14430
+    },
+    {
+      "epoch": 1.878563061304178,
+      "grad_norm": 2.7562592029571533,
+      "learning_rate": 3.190104947758905e-06,
+      "loss": 0.527,
+      "step": 14433
+    },
+    {
+      "epoch": 1.878953533775869,
+      "grad_norm": 2.548182964324951,
+      "learning_rate": 3.1881569821818646e-06,
+      "loss": 0.4998,
+      "step": 14436
+    },
+    {
+      "epoch": 1.8793440062475595,
+      "grad_norm": 2.421619176864624,
+      "learning_rate": 3.186209333129987e-06,
+      "loss": 0.4503,
+      "step": 14439
+    },
+    {
+      "epoch": 1.8797344787192503,
+      "grad_norm": 2.976463794708252,
+      "learning_rate": 3.1842620009435244e-06,
+      "loss": 0.5083,
+      "step": 14442
+    },
+    {
+      "epoch": 1.880124951190941,
+      "grad_norm": 3.011476516723633,
+      "learning_rate": 3.1823149859626695e-06,
+      "loss": 0.4516,
+      "step": 14445
+    },
+    {
+      "epoch": 1.8805154236626318,
+      "grad_norm": 2.777480125427246,
+      "learning_rate": 3.180368288527563e-06,
+      "loss": 0.5437,
+      "step": 14448
+    },
+    {
+      "epoch": 1.8809058961343226,
+      "grad_norm": 2.6614065170288086,
+      "learning_rate": 3.1784219089782885e-06,
+      "loss": 0.532,
+      "step": 14451
+    },
+    {
+      "epoch": 1.8812963686060131,
+      "grad_norm": 3.332582473754883,
+      "learning_rate": 3.176475847654873e-06,
+      "loss": 0.5112,
+      "step": 14454
+    },
+    {
+      "epoch": 1.8816868410777041,
+      "grad_norm": 3.011148452758789,
+      "learning_rate": 3.1745301048972923e-06,
+      "loss": 0.5642,
+      "step": 14457
+    },
+    {
+      "epoch": 1.8820773135493947,
+      "grad_norm": 3.288445472717285,
+      "learning_rate": 3.1725846810454612e-06,
+      "loss": 0.4546,
+      "step": 14460
+    },
+    {
+      "epoch": 1.8824677860210857,
+      "grad_norm": 2.9788262844085693,
+      "learning_rate": 3.170639576439244e-06,
+      "loss": 0.5007,
+      "step": 14463
+    },
+    {
+      "epoch": 1.8828582584927762,
+      "grad_norm": 2.3742592334747314,
+      "learning_rate": 3.1686947914184424e-06,
+      "loss": 0.4228,
+      "step": 14466
+    },
+    {
+      "epoch": 1.883248730964467,
+      "grad_norm": 2.842820405960083,
+      "learning_rate": 3.1667503263228093e-06,
+      "loss": 0.5175,
+      "step": 14469
+    },
+    {
+      "epoch": 1.8836392034361578,
+      "grad_norm": 3.063159942626953,
+      "learning_rate": 3.1648061814920372e-06,
+      "loss": 0.476,
+      "step": 14472
+    },
+    {
+      "epoch": 1.8840296759078485,
+      "grad_norm": 2.8648762702941895,
+      "learning_rate": 3.1628623572657646e-06,
+      "loss": 0.4767,
+      "step": 14475
+    },
+    {
+      "epoch": 1.8844201483795393,
+      "grad_norm": 2.5164406299591064,
+      "learning_rate": 3.160918853983574e-06,
+      "loss": 0.5384,
+      "step": 14478
+    },
+    {
+      "epoch": 1.8848106208512299,
+      "grad_norm": 2.923412322998047,
+      "learning_rate": 3.1589756719849897e-06,
+      "loss": 0.4499,
+      "step": 14481
+    },
+    {
+      "epoch": 1.8852010933229209,
+      "grad_norm": 2.5238845348358154,
+      "learning_rate": 3.1570328116094835e-06,
+      "loss": 0.4332,
+      "step": 14484
+    },
+    {
+      "epoch": 1.8855915657946114,
+      "grad_norm": 2.5288538932800293,
+      "learning_rate": 3.155090273196467e-06,
+      "loss": 0.4952,
+      "step": 14487
+    },
+    {
+      "epoch": 1.8859820382663022,
+      "grad_norm": 2.2970383167266846,
+      "learning_rate": 3.153148057085299e-06,
+      "loss": 0.4388,
+      "step": 14490
+    },
+    {
+      "epoch": 1.886372510737993,
+      "grad_norm": 4.423757553100586,
+      "learning_rate": 3.1512061636152814e-06,
+      "loss": 0.4552,
+      "step": 14493
+    },
+    {
+      "epoch": 1.8867629832096837,
+      "grad_norm": 3.0830297470092773,
+      "learning_rate": 3.149264593125655e-06,
+      "loss": 0.5075,
+      "step": 14496
+    },
+    {
+      "epoch": 1.8871534556813745,
+      "grad_norm": 2.8348443508148193,
+      "learning_rate": 3.147323345955612e-06,
+      "loss": 0.5471,
+      "step": 14499
+    },
+    {
+      "epoch": 1.887543928153065,
+      "grad_norm": 2.6750619411468506,
+      "learning_rate": 3.1453824224442836e-06,
+      "loss": 0.458,
+      "step": 14502
+    },
+    {
+      "epoch": 1.887934400624756,
+      "grad_norm": 2.684641122817993,
+      "learning_rate": 3.143441822930745e-06,
+      "loss": 0.4285,
+      "step": 14505
+    },
+    {
+      "epoch": 1.8883248730964466,
+      "grad_norm": 2.603320598602295,
+      "learning_rate": 3.1415015477540136e-06,
+      "loss": 0.5014,
+      "step": 14508
+    },
+    {
+      "epoch": 1.8887153455681376,
+      "grad_norm": 2.6286399364471436,
+      "learning_rate": 3.1395615972530514e-06,
+      "loss": 0.4378,
+      "step": 14511
+    },
+    {
+      "epoch": 1.8891058180398281,
+      "grad_norm": 2.7393951416015625,
+      "learning_rate": 3.137621971766766e-06,
+      "loss": 0.5872,
+      "step": 14514
+    },
+    {
+      "epoch": 1.889496290511519,
+      "grad_norm": 2.693389415740967,
+      "learning_rate": 3.1356826716340027e-06,
+      "loss": 0.5285,
+      "step": 14517
+    },
+    {
+      "epoch": 1.8898867629832097,
+      "grad_norm": 2.694274663925171,
+      "learning_rate": 3.1337436971935565e-06,
+      "loss": 0.4483,
+      "step": 14520
+    },
+    {
+      "epoch": 1.8902772354549005,
+      "grad_norm": 2.6547138690948486,
+      "learning_rate": 3.131805048784159e-06,
+      "loss": 0.4728,
+      "step": 14523
+    },
+    {
+      "epoch": 1.8906677079265912,
+      "grad_norm": 2.5658202171325684,
+      "learning_rate": 3.129866726744489e-06,
+      "loss": 0.5587,
+      "step": 14526
+    },
+    {
+      "epoch": 1.8910581803982818,
+      "grad_norm": 2.6531288623809814,
+      "learning_rate": 3.1279287314131694e-06,
+      "loss": 0.5347,
+      "step": 14529
+    },
+    {
+      "epoch": 1.8914486528699728,
+      "grad_norm": 2.8203883171081543,
+      "learning_rate": 3.1259910631287605e-06,
+      "loss": 0.4708,
+      "step": 14532
+    },
+    {
+      "epoch": 1.8918391253416633,
+      "grad_norm": 2.5969855785369873,
+      "learning_rate": 3.1240537222297716e-06,
+      "loss": 0.5771,
+      "step": 14535
+    },
+    {
+      "epoch": 1.8922295978133543,
+      "grad_norm": 3.149728775024414,
+      "learning_rate": 3.122116709054649e-06,
+      "loss": 0.5279,
+      "step": 14538
+    },
+    {
+      "epoch": 1.8926200702850449,
+      "grad_norm": 2.6601550579071045,
+      "learning_rate": 3.120180023941789e-06,
+      "loss": 0.4675,
+      "step": 14541
+    },
+    {
+      "epoch": 1.8930105427567356,
+      "grad_norm": 2.72967529296875,
+      "learning_rate": 3.118243667229523e-06,
+      "loss": 0.4924,
+      "step": 14544
+    },
+    {
+      "epoch": 1.8934010152284264,
+      "grad_norm": 3.1012916564941406,
+      "learning_rate": 3.116307639256131e-06,
+      "loss": 0.4926,
+      "step": 14547
+    },
+    {
+      "epoch": 1.8937914877001172,
+      "grad_norm": 2.993875026702881,
+      "learning_rate": 3.1143719403598307e-06,
+      "loss": 0.4754,
+      "step": 14550
+    },
+    {
+      "epoch": 1.894181960171808,
+      "grad_norm": 4.121201515197754,
+      "learning_rate": 3.1124365708787856e-06,
+      "loss": 0.5015,
+      "step": 14553
+    },
+    {
+      "epoch": 1.8945724326434985,
+      "grad_norm": 2.526926040649414,
+      "learning_rate": 3.110501531151102e-06,
+      "loss": 0.4477,
+      "step": 14556
+    },
+    {
+      "epoch": 1.8949629051151895,
+      "grad_norm": 3.1489815711975098,
+      "learning_rate": 3.1085668215148245e-06,
+      "loss": 0.444,
+      "step": 14559
+    },
+    {
+      "epoch": 1.89535337758688,
+      "grad_norm": 2.5161519050598145,
+      "learning_rate": 3.1066324423079445e-06,
+      "loss": 0.5054,
+      "step": 14562
+    },
+    {
+      "epoch": 1.8957438500585708,
+      "grad_norm": 2.526221990585327,
+      "learning_rate": 3.104698393868395e-06,
+      "loss": 0.489,
+      "step": 14565
+    },
+    {
+      "epoch": 1.8961343225302616,
+      "grad_norm": 2.54593825340271,
+      "learning_rate": 3.102764676534048e-06,
+      "loss": 0.49,
+      "step": 14568
+    },
+    {
+      "epoch": 1.8965247950019524,
+      "grad_norm": 3.2360286712646484,
+      "learning_rate": 3.1008312906427212e-06,
+      "loss": 0.4462,
+      "step": 14571
+    },
+    {
+      "epoch": 1.8969152674736431,
+      "grad_norm": 3.160997152328491,
+      "learning_rate": 3.098898236532172e-06,
+      "loss": 0.4875,
+      "step": 14574
+    },
+    {
+      "epoch": 1.8973057399453337,
+      "grad_norm": 4.4345808029174805,
+      "learning_rate": 3.096965514540102e-06,
+      "loss": 0.5209,
+      "step": 14577
+    },
+    {
+      "epoch": 1.8976962124170247,
+      "grad_norm": 2.559445858001709,
+      "learning_rate": 3.0950331250041515e-06,
+      "loss": 0.5191,
+      "step": 14580
+    },
+    {
+      "epoch": 1.8980866848887152,
+      "grad_norm": 2.8593218326568604,
+      "learning_rate": 3.093101068261909e-06,
+      "loss": 0.4575,
+      "step": 14583
+    },
+    {
+      "epoch": 1.8984771573604062,
+      "grad_norm": 2.4053094387054443,
+      "learning_rate": 3.0911693446508973e-06,
+      "loss": 0.4613,
+      "step": 14586
+    },
+    {
+      "epoch": 1.8988676298320968,
+      "grad_norm": 2.4780166149139404,
+      "learning_rate": 3.089237954508585e-06,
+      "loss": 0.4018,
+      "step": 14589
+    },
+    {
+      "epoch": 1.8992581023037876,
+      "grad_norm": 2.995244264602661,
+      "learning_rate": 3.0873068981723842e-06,
+      "loss": 0.5322,
+      "step": 14592
+    },
+    {
+      "epoch": 1.8996485747754783,
+      "grad_norm": 2.625032663345337,
+      "learning_rate": 3.085376175979643e-06,
+      "loss": 0.4439,
+      "step": 14595
+    },
+    {
+      "epoch": 1.900039047247169,
+      "grad_norm": 2.8063700199127197,
+      "learning_rate": 3.083445788267657e-06,
+      "loss": 0.4898,
+      "step": 14598
+    },
+    {
+      "epoch": 1.9004295197188599,
+      "grad_norm": 2.738100051879883,
+      "learning_rate": 3.081515735373659e-06,
+      "loss": 0.5017,
+      "step": 14601
+    },
+    {
+      "epoch": 1.9008199921905504,
+      "grad_norm": 2.68695330619812,
+      "learning_rate": 3.0795860176348267e-06,
+      "loss": 0.4812,
+      "step": 14604
+    },
+    {
+      "epoch": 1.9012104646622414,
+      "grad_norm": 3.1662650108337402,
+      "learning_rate": 3.0776566353882775e-06,
+      "loss": 0.4975,
+      "step": 14607
+    },
+    {
+      "epoch": 1.901600937133932,
+      "grad_norm": 2.6604361534118652,
+      "learning_rate": 3.0757275889710697e-06,
+      "loss": 0.4858,
+      "step": 14610
+    },
+    {
+      "epoch": 1.901991409605623,
+      "grad_norm": 2.82057523727417,
+      "learning_rate": 3.0737988787202034e-06,
+      "loss": 0.4714,
+      "step": 14613
+    },
+    {
+      "epoch": 1.9023818820773135,
+      "grad_norm": 2.7449402809143066,
+      "learning_rate": 3.071870504972623e-06,
+      "loss": 0.5916,
+      "step": 14616
+    },
+    {
+      "epoch": 1.9027723545490043,
+      "grad_norm": 2.767760753631592,
+      "learning_rate": 3.0699424680652066e-06,
+      "loss": 0.5381,
+      "step": 14619
+    },
+    {
+      "epoch": 1.903162827020695,
+      "grad_norm": 2.4877452850341797,
+      "learning_rate": 3.0680147683347803e-06,
+      "loss": 0.4768,
+      "step": 14622
+    },
+    {
+      "epoch": 1.9035532994923858,
+      "grad_norm": 2.712571382522583,
+      "learning_rate": 3.066087406118111e-06,
+      "loss": 0.4785,
+      "step": 14625
+    },
+    {
+      "epoch": 1.9039437719640766,
+      "grad_norm": 2.5379676818847656,
+      "learning_rate": 3.064160381751905e-06,
+      "loss": 0.4833,
+      "step": 14628
+    },
+    {
+      "epoch": 1.9043342444357672,
+      "grad_norm": 2.888591766357422,
+      "learning_rate": 3.062233695572806e-06,
+      "loss": 0.5205,
+      "step": 14631
+    },
+    {
+      "epoch": 1.9047247169074581,
+      "grad_norm": 2.4471936225891113,
+      "learning_rate": 3.060307347917405e-06,
+      "loss": 0.4541,
+      "step": 14634
+    },
+    {
+      "epoch": 1.9051151893791487,
+      "grad_norm": 2.5203559398651123,
+      "learning_rate": 3.0583813391222294e-06,
+      "loss": 0.4671,
+      "step": 14637
+    },
+    {
+      "epoch": 1.9055056618508395,
+      "grad_norm": 2.5161707401275635,
+      "learning_rate": 3.056455669523749e-06,
+      "loss": 0.4245,
+      "step": 14640
+    },
+    {
+      "epoch": 1.9058961343225302,
+      "grad_norm": 2.660309076309204,
+      "learning_rate": 3.0545303394583755e-06,
+      "loss": 0.4664,
+      "step": 14643
+    },
+    {
+      "epoch": 1.906286606794221,
+      "grad_norm": 2.8389623165130615,
+      "learning_rate": 3.0526053492624574e-06,
+      "loss": 0.5387,
+      "step": 14646
+    },
+    {
+      "epoch": 1.9066770792659118,
+      "grad_norm": 2.4667160511016846,
+      "learning_rate": 3.05068069927229e-06,
+      "loss": 0.4764,
+      "step": 14649
+    },
+    {
+      "epoch": 1.9070675517376023,
+      "grad_norm": 3.343127727508545,
+      "learning_rate": 3.0487563898241025e-06,
+      "loss": 0.5026,
+      "step": 14652
+    },
+    {
+      "epoch": 1.9074580242092933,
+      "grad_norm": 2.870053291320801,
+      "learning_rate": 3.046832421254068e-06,
+      "loss": 0.4699,
+      "step": 14655
+    },
+    {
+      "epoch": 1.9078484966809839,
+      "grad_norm": 2.796257495880127,
+      "learning_rate": 3.0449087938983025e-06,
+      "loss": 0.5665,
+      "step": 14658
+    },
+    {
+      "epoch": 1.9082389691526749,
+      "grad_norm": 2.6486213207244873,
+      "learning_rate": 3.0429855080928567e-06,
+      "loss": 0.4896,
+      "step": 14661
+    },
+    {
+      "epoch": 1.9086294416243654,
+      "grad_norm": 2.7578749656677246,
+      "learning_rate": 3.0410625641737245e-06,
+      "loss": 0.4799,
+      "step": 14664
+    },
+    {
+      "epoch": 1.9090199140960562,
+      "grad_norm": 2.627326726913452,
+      "learning_rate": 3.0391399624768424e-06,
+      "loss": 0.4241,
+      "step": 14667
+    },
+    {
+      "epoch": 1.909410386567747,
+      "grad_norm": 2.7783567905426025,
+      "learning_rate": 3.0372177033380846e-06,
+      "loss": 0.4997,
+      "step": 14670
+    },
+    {
+      "epoch": 1.9098008590394377,
+      "grad_norm": 2.5225064754486084,
+      "learning_rate": 3.0352957870932643e-06,
+      "loss": 0.4955,
+      "step": 14673
+    },
+    {
+      "epoch": 1.9101913315111285,
+      "grad_norm": 4.349731922149658,
+      "learning_rate": 3.0333742140781374e-06,
+      "loss": 0.4517,
+      "step": 14676
+    },
+    {
+      "epoch": 1.910581803982819,
+      "grad_norm": 2.4601027965545654,
+      "learning_rate": 3.031452984628398e-06,
+      "loss": 0.5083,
+      "step": 14679
+    },
+    {
+      "epoch": 1.91097227645451,
+      "grad_norm": 2.8652219772338867,
+      "learning_rate": 3.02953209907968e-06,
+      "loss": 0.4887,
+      "step": 14682
+    },
+    {
+      "epoch": 1.9113627489262006,
+      "grad_norm": 2.7341558933258057,
+      "learning_rate": 3.0276115577675604e-06,
+      "loss": 0.5125,
+      "step": 14685
+    },
+    {
+      "epoch": 1.9117532213978916,
+      "grad_norm": 3.915335178375244,
+      "learning_rate": 3.025691361027552e-06,
+      "loss": 0.531,
+      "step": 14688
+    },
+    {
+      "epoch": 1.9121436938695822,
+      "grad_norm": 3.0976195335388184,
+      "learning_rate": 3.023771509195108e-06,
+      "loss": 0.4759,
+      "step": 14691
+    },
+    {
+      "epoch": 1.912534166341273,
+      "grad_norm": 2.6159005165100098,
+      "learning_rate": 3.0218520026056264e-06,
+      "loss": 0.4972,
+      "step": 14694
+    },
+    {
+      "epoch": 1.9129246388129637,
+      "grad_norm": 2.7680933475494385,
+      "learning_rate": 3.019932841594437e-06,
+      "loss": 0.5076,
+      "step": 14697
+    },
+    {
+      "epoch": 1.9133151112846545,
+      "grad_norm": 3.1167070865631104,
+      "learning_rate": 3.0180140264968153e-06,
+      "loss": 0.6106,
+      "step": 14700
+    },
+    {
+      "epoch": 1.9137055837563453,
+      "grad_norm": 2.826345682144165,
+      "learning_rate": 3.0160955576479735e-06,
+      "loss": 0.4774,
+      "step": 14703
+    },
+    {
+      "epoch": 1.9140960562280358,
+      "grad_norm": 2.6846492290496826,
+      "learning_rate": 3.014177435383063e-06,
+      "loss": 0.4761,
+      "step": 14706
+    },
+    {
+      "epoch": 1.9144865286997268,
+      "grad_norm": 2.6631195545196533,
+      "learning_rate": 3.0122596600371777e-06,
+      "loss": 0.4419,
+      "step": 14709
+    },
+    {
+      "epoch": 1.9148770011714173,
+      "grad_norm": 2.869974374771118,
+      "learning_rate": 3.010342231945349e-06,
+      "loss": 0.4628,
+      "step": 14712
+    },
+    {
+      "epoch": 1.9152674736431081,
+      "grad_norm": 2.954019546508789,
+      "learning_rate": 3.008425151442546e-06,
+      "loss": 0.4112,
+      "step": 14715
+    },
+    {
+      "epoch": 1.915657946114799,
+      "grad_norm": 3.6011927127838135,
+      "learning_rate": 3.0065084188636794e-06,
+      "loss": 0.5629,
+      "step": 14718
+    },
+    {
+      "epoch": 1.9160484185864897,
+      "grad_norm": 2.377660036087036,
+      "learning_rate": 3.0045920345435996e-06,
+      "loss": 0.4984,
+      "step": 14721
+    },
+    {
+      "epoch": 1.9164388910581804,
+      "grad_norm": 2.4976119995117188,
+      "learning_rate": 3.002675998817093e-06,
+      "loss": 0.5637,
+      "step": 14724
+    },
+    {
+      "epoch": 1.916829363529871,
+      "grad_norm": 2.8942108154296875,
+      "learning_rate": 3.0007603120188877e-06,
+      "loss": 0.4453,
+      "step": 14727
+    },
+    {
+      "epoch": 1.917219836001562,
+      "grad_norm": 2.675811290740967,
+      "learning_rate": 2.998844974483649e-06,
+      "loss": 0.4768,
+      "step": 14730
+    },
+    {
+      "epoch": 1.9176103084732525,
+      "grad_norm": 2.6159439086914062,
+      "learning_rate": 2.9969299865459845e-06,
+      "loss": 0.4118,
+      "step": 14733
+    },
+    {
+      "epoch": 1.9180007809449435,
+      "grad_norm": 2.5294883251190186,
+      "learning_rate": 2.995015348540438e-06,
+      "loss": 0.4178,
+      "step": 14736
+    },
+    {
+      "epoch": 1.918391253416634,
+      "grad_norm": 2.5951461791992188,
+      "learning_rate": 2.993101060801491e-06,
+      "loss": 0.5095,
+      "step": 14739
+    },
+    {
+      "epoch": 1.9187817258883249,
+      "grad_norm": 2.7234127521514893,
+      "learning_rate": 2.991187123663567e-06,
+      "loss": 0.4648,
+      "step": 14742
+    },
+    {
+      "epoch": 1.9191721983600156,
+      "grad_norm": 2.6454546451568604,
+      "learning_rate": 2.9892735374610273e-06,
+      "loss": 0.4701,
+      "step": 14745
+    },
+    {
+      "epoch": 1.9195626708317064,
+      "grad_norm": 2.923607349395752,
+      "learning_rate": 2.987360302528169e-06,
+      "loss": 0.4573,
+      "step": 14748
+    },
+    {
+      "epoch": 1.9199531433033972,
+      "grad_norm": 2.9072718620300293,
+      "learning_rate": 2.9854474191992323e-06,
+      "loss": 0.457,
+      "step": 14751
+    },
+    {
+      "epoch": 1.9203436157750877,
+      "grad_norm": 2.4134531021118164,
+      "learning_rate": 2.983534887808394e-06,
+      "loss": 0.4027,
+      "step": 14754
+    },
+    {
+      "epoch": 1.9207340882467787,
+      "grad_norm": 2.700012683868408,
+      "learning_rate": 2.9816227086897696e-06,
+      "loss": 0.4745,
+      "step": 14757
+    },
+    {
+      "epoch": 1.9211245607184693,
+      "grad_norm": 2.7360305786132812,
+      "learning_rate": 2.9797108821774114e-06,
+      "loss": 0.4525,
+      "step": 14760
+    },
+    {
+      "epoch": 1.9215150331901603,
+      "grad_norm": 2.5662286281585693,
+      "learning_rate": 2.9777994086053123e-06,
+      "loss": 0.5103,
+      "step": 14763
+    },
+    {
+      "epoch": 1.9219055056618508,
+      "grad_norm": 2.873728036880493,
+      "learning_rate": 2.975888288307402e-06,
+      "loss": 0.5333,
+      "step": 14766
+    },
+    {
+      "epoch": 1.9222959781335416,
+      "grad_norm": 3.446164846420288,
+      "learning_rate": 2.973977521617549e-06,
+      "loss": 0.4311,
+      "step": 14769
+    },
+    {
+      "epoch": 1.9226864506052324,
+      "grad_norm": 2.5216710567474365,
+      "learning_rate": 2.9720671088695628e-06,
+      "loss": 0.5313,
+      "step": 14772
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 2.530691385269165,
+      "learning_rate": 2.9701570503971846e-06,
+      "loss": 0.497,
+      "step": 14775
+    },
+    {
+      "epoch": 1.923467395548614,
+      "grad_norm": 2.462676763534546,
+      "learning_rate": 2.968247346534101e-06,
+      "loss": 0.4939,
+      "step": 14778
+    },
+    {
+      "epoch": 1.9238578680203045,
+      "grad_norm": 3.943150043487549,
+      "learning_rate": 2.9663379976139307e-06,
+      "loss": 0.5034,
+      "step": 14781
+    },
+    {
+      "epoch": 1.9242483404919954,
+      "grad_norm": 2.7674999237060547,
+      "learning_rate": 2.964429003970234e-06,
+      "loss": 0.4747,
+      "step": 14784
+    },
+    {
+      "epoch": 1.924638812963686,
+      "grad_norm": 2.529186725616455,
+      "learning_rate": 2.9625203659365094e-06,
+      "loss": 0.4283,
+      "step": 14787
+    },
+    {
+      "epoch": 1.9250292854353768,
+      "grad_norm": 2.8938591480255127,
+      "learning_rate": 2.9606120838461884e-06,
+      "loss": 0.4594,
+      "step": 14790
+    },
+    {
+      "epoch": 1.9254197579070675,
+      "grad_norm": 2.950148820877075,
+      "learning_rate": 2.958704158032647e-06,
+      "loss": 0.5019,
+      "step": 14793
+    },
+    {
+      "epoch": 1.9258102303787583,
+      "grad_norm": 3.631539821624756,
+      "learning_rate": 2.956796588829195e-06,
+      "loss": 0.4972,
+      "step": 14796
+    },
+    {
+      "epoch": 1.926200702850449,
+      "grad_norm": 2.9207494258880615,
+      "learning_rate": 2.954889376569081e-06,
+      "loss": 0.5222,
+      "step": 14799
+    },
+    {
+      "epoch": 1.9265911753221396,
+      "grad_norm": 2.5649592876434326,
+      "learning_rate": 2.9529825215854907e-06,
+      "loss": 0.5079,
+      "step": 14802
+    },
+    {
+      "epoch": 1.9269816477938306,
+      "grad_norm": 2.7472033500671387,
+      "learning_rate": 2.951076024211547e-06,
+      "loss": 0.4459,
+      "step": 14805
+    },
+    {
+      "epoch": 1.9273721202655212,
+      "grad_norm": 2.761784315109253,
+      "learning_rate": 2.949169884780313e-06,
+      "loss": 0.4845,
+      "step": 14808
+    },
+    {
+      "epoch": 1.9277625927372122,
+      "grad_norm": 2.7410855293273926,
+      "learning_rate": 2.947264103624784e-06,
+      "loss": 0.5347,
+      "step": 14811
+    },
+    {
+      "epoch": 1.9281530652089027,
+      "grad_norm": 3.519693613052368,
+      "learning_rate": 2.9453586810778996e-06,
+      "loss": 0.512,
+      "step": 14814
+    },
+    {
+      "epoch": 1.9285435376805935,
+      "grad_norm": 2.6707022190093994,
+      "learning_rate": 2.9434536174725305e-06,
+      "loss": 0.4977,
+      "step": 14817
+    },
+    {
+      "epoch": 1.9289340101522843,
+      "grad_norm": 3.4989326000213623,
+      "learning_rate": 2.941548913141487e-06,
+      "loss": 0.4988,
+      "step": 14820
+    },
+    {
+      "epoch": 1.929324482623975,
+      "grad_norm": 2.744438886642456,
+      "learning_rate": 2.9396445684175196e-06,
+      "loss": 0.4738,
+      "step": 14823
+    },
+    {
+      "epoch": 1.9297149550956658,
+      "grad_norm": 3.4008469581604004,
+      "learning_rate": 2.9377405836333106e-06,
+      "loss": 0.4601,
+      "step": 14826
+    },
+    {
+      "epoch": 1.9301054275673564,
+      "grad_norm": 3.585939407348633,
+      "learning_rate": 2.9358369591214847e-06,
+      "loss": 0.4585,
+      "step": 14829
+    },
+    {
+      "epoch": 1.9304959000390474,
+      "grad_norm": 2.683812379837036,
+      "learning_rate": 2.9339336952145957e-06,
+      "loss": 0.4817,
+      "step": 14832
+    },
+    {
+      "epoch": 1.930886372510738,
+      "grad_norm": 2.6055679321289062,
+      "learning_rate": 2.932030792245148e-06,
+      "loss": 0.5232,
+      "step": 14835
+    },
+    {
+      "epoch": 1.9312768449824287,
+      "grad_norm": 2.5260632038116455,
+      "learning_rate": 2.9301282505455687e-06,
+      "loss": 0.4211,
+      "step": 14838
+    },
+    {
+      "epoch": 1.9316673174541195,
+      "grad_norm": 2.553103446960449,
+      "learning_rate": 2.9282260704482313e-06,
+      "loss": 0.4662,
+      "step": 14841
+    },
+    {
+      "epoch": 1.9320577899258102,
+      "grad_norm": 2.7395856380462646,
+      "learning_rate": 2.9263242522854397e-06,
+      "loss": 0.4706,
+      "step": 14844
+    },
+    {
+      "epoch": 1.932448262397501,
+      "grad_norm": 2.5782995223999023,
+      "learning_rate": 2.924422796389439e-06,
+      "loss": 0.4471,
+      "step": 14847
+    },
+    {
+      "epoch": 1.9328387348691918,
+      "grad_norm": 2.5767011642456055,
+      "learning_rate": 2.92252170309241e-06,
+      "loss": 0.4797,
+      "step": 14850
+    },
+    {
+      "epoch": 1.9332292073408825,
+      "grad_norm": 2.6651394367218018,
+      "learning_rate": 2.920620972726468e-06,
+      "loss": 0.5292,
+      "step": 14853
+    },
+    {
+      "epoch": 1.933619679812573,
+      "grad_norm": 3.3890509605407715,
+      "learning_rate": 2.9187206056236693e-06,
+      "loss": 0.4272,
+      "step": 14856
+    },
+    {
+      "epoch": 1.934010152284264,
+      "grad_norm": 3.1942598819732666,
+      "learning_rate": 2.916820602116e-06,
+      "loss": 0.5652,
+      "step": 14859
+    },
+    {
+      "epoch": 1.9344006247559546,
+      "grad_norm": 2.847402572631836,
+      "learning_rate": 2.914920962535391e-06,
+      "loss": 0.467,
+      "step": 14862
+    },
+    {
+      "epoch": 1.9347910972276454,
+      "grad_norm": 2.6904537677764893,
+      "learning_rate": 2.913021687213704e-06,
+      "loss": 0.5246,
+      "step": 14865
+    },
+    {
+      "epoch": 1.9351815696993362,
+      "grad_norm": 2.631735324859619,
+      "learning_rate": 2.9111227764827376e-06,
+      "loss": 0.4657,
+      "step": 14868
+    },
+    {
+      "epoch": 1.935572042171027,
+      "grad_norm": 2.6257803440093994,
+      "learning_rate": 2.9092242306742256e-06,
+      "loss": 0.5195,
+      "step": 14871
+    },
+    {
+      "epoch": 1.9359625146427177,
+      "grad_norm": 2.7459394931793213,
+      "learning_rate": 2.9073260501198424e-06,
+      "loss": 0.4976,
+      "step": 14874
+    },
+    {
+      "epoch": 1.9363529871144083,
+      "grad_norm": 2.597118854522705,
+      "learning_rate": 2.905428235151198e-06,
+      "loss": 0.4381,
+      "step": 14877
+    },
+    {
+      "epoch": 1.9367434595860993,
+      "grad_norm": 2.7893166542053223,
+      "learning_rate": 2.9035307860998346e-06,
+      "loss": 0.4861,
+      "step": 14880
+    },
+    {
+      "epoch": 1.9371339320577898,
+      "grad_norm": 2.5637564659118652,
+      "learning_rate": 2.90163370329723e-06,
+      "loss": 0.5199,
+      "step": 14883
+    },
+    {
+      "epoch": 1.9375244045294808,
+      "grad_norm": 2.49005389213562,
+      "learning_rate": 2.899736987074806e-06,
+      "loss": 0.5072,
+      "step": 14886
+    },
+    {
+      "epoch": 1.9379148770011714,
+      "grad_norm": 2.6753792762756348,
+      "learning_rate": 2.8978406377639114e-06,
+      "loss": 0.483,
+      "step": 14889
+    },
+    {
+      "epoch": 1.9383053494728621,
+      "grad_norm": 2.4277470111846924,
+      "learning_rate": 2.8959446556958333e-06,
+      "loss": 0.5358,
+      "step": 14892
+    },
+    {
+      "epoch": 1.938695821944553,
+      "grad_norm": 2.332995653152466,
+      "learning_rate": 2.8940490412017997e-06,
+      "loss": 0.4988,
+      "step": 14895
+    },
+    {
+      "epoch": 1.9390862944162437,
+      "grad_norm": 2.639230728149414,
+      "learning_rate": 2.892153794612968e-06,
+      "loss": 0.4237,
+      "step": 14898
+    },
+    {
+      "epoch": 1.9394767668879345,
+      "grad_norm": 2.7841386795043945,
+      "learning_rate": 2.8902589162604323e-06,
+      "loss": 0.4862,
+      "step": 14901
+    },
+    {
+      "epoch": 1.939867239359625,
+      "grad_norm": 2.618056535720825,
+      "learning_rate": 2.8883644064752274e-06,
+      "loss": 0.4738,
+      "step": 14904
+    },
+    {
+      "epoch": 1.940257711831316,
+      "grad_norm": 2.2609994411468506,
+      "learning_rate": 2.886470265588319e-06,
+      "loss": 0.4599,
+      "step": 14907
+    },
+    {
+      "epoch": 1.9406481843030066,
+      "grad_norm": 2.6283135414123535,
+      "learning_rate": 2.8845764939306063e-06,
+      "loss": 0.5095,
+      "step": 14910
+    },
+    {
+      "epoch": 1.9410386567746973,
+      "grad_norm": 2.641413688659668,
+      "learning_rate": 2.8826830918329325e-06,
+      "loss": 0.4114,
+      "step": 14913
+    },
+    {
+      "epoch": 1.941429129246388,
+      "grad_norm": 2.5801405906677246,
+      "learning_rate": 2.8807900596260663e-06,
+      "loss": 0.5321,
+      "step": 14916
+    },
+    {
+      "epoch": 1.9418196017180789,
+      "grad_norm": 2.513852834701538,
+      "learning_rate": 2.87889739764072e-06,
+      "loss": 0.5193,
+      "step": 14919
+    },
+    {
+      "epoch": 1.9422100741897697,
+      "grad_norm": 2.852112054824829,
+      "learning_rate": 2.8770051062075343e-06,
+      "loss": 0.4681,
+      "step": 14922
+    },
+    {
+      "epoch": 1.9426005466614602,
+      "grad_norm": 2.5019607543945312,
+      "learning_rate": 2.8751131856570935e-06,
+      "loss": 0.5315,
+      "step": 14925
+    },
+    {
+      "epoch": 1.9429910191331512,
+      "grad_norm": 2.95793080329895,
+      "learning_rate": 2.873221636319908e-06,
+      "loss": 0.5047,
+      "step": 14928
+    },
+    {
+      "epoch": 1.9433814916048417,
+      "grad_norm": 2.9171652793884277,
+      "learning_rate": 2.871330458526429e-06,
+      "loss": 0.5912,
+      "step": 14931
+    },
+    {
+      "epoch": 1.9437719640765327,
+      "grad_norm": 3.1407294273376465,
+      "learning_rate": 2.8694396526070383e-06,
+      "loss": 0.4123,
+      "step": 14934
+    },
+    {
+      "epoch": 1.9441624365482233,
+      "grad_norm": 2.9281039237976074,
+      "learning_rate": 2.8675492188920605e-06,
+      "loss": 0.6034,
+      "step": 14937
+    },
+    {
+      "epoch": 1.944552909019914,
+      "grad_norm": 3.462409257888794,
+      "learning_rate": 2.865659157711748e-06,
+      "loss": 0.4745,
+      "step": 14940
+    },
+    {
+      "epoch": 1.9449433814916048,
+      "grad_norm": 3.067357301712036,
+      "learning_rate": 2.863769469396289e-06,
+      "loss": 0.4869,
+      "step": 14943
+    },
+    {
+      "epoch": 1.9453338539632956,
+      "grad_norm": 2.8917782306671143,
+      "learning_rate": 2.8618801542758116e-06,
+      "loss": 0.513,
+      "step": 14946
+    },
+    {
+      "epoch": 1.9457243264349864,
+      "grad_norm": 2.6178336143493652,
+      "learning_rate": 2.859991212680373e-06,
+      "loss": 0.5033,
+      "step": 14949
+    },
+    {
+      "epoch": 1.946114798906677,
+      "grad_norm": 2.847200870513916,
+      "learning_rate": 2.858102644939966e-06,
+      "loss": 0.4524,
+      "step": 14952
+    },
+    {
+      "epoch": 1.946505271378368,
+      "grad_norm": 2.400238275527954,
+      "learning_rate": 2.8562144513845236e-06,
+      "loss": 0.412,
+      "step": 14955
+    },
+    {
+      "epoch": 1.9468957438500585,
+      "grad_norm": 2.936591386795044,
+      "learning_rate": 2.8543266323439034e-06,
+      "loss": 0.4973,
+      "step": 14958
+    },
+    {
+      "epoch": 1.9472862163217495,
+      "grad_norm": 2.7296829223632812,
+      "learning_rate": 2.8524391881479096e-06,
+      "loss": 0.4715,
+      "step": 14961
+    },
+    {
+      "epoch": 1.94767668879344,
+      "grad_norm": 2.435722589492798,
+      "learning_rate": 2.8505521191262697e-06,
+      "loss": 0.4847,
+      "step": 14964
+    },
+    {
+      "epoch": 1.9480671612651308,
+      "grad_norm": 2.6124179363250732,
+      "learning_rate": 2.8486654256086543e-06,
+      "loss": 0.5464,
+      "step": 14967
+    },
+    {
+      "epoch": 1.9484576337368216,
+      "grad_norm": 3.493014335632324,
+      "learning_rate": 2.8467791079246636e-06,
+      "loss": 0.5372,
+      "step": 14970
+    },
+    {
+      "epoch": 1.9488481062085123,
+      "grad_norm": 3.0467758178710938,
+      "learning_rate": 2.8448931664038315e-06,
+      "loss": 0.551,
+      "step": 14973
+    },
+    {
+      "epoch": 1.9492385786802031,
+      "grad_norm": 2.6954057216644287,
+      "learning_rate": 2.843007601375631e-06,
+      "loss": 0.5891,
+      "step": 14976
+    },
+    {
+      "epoch": 1.9496290511518937,
+      "grad_norm": 3.6716325283050537,
+      "learning_rate": 2.8411224131694647e-06,
+      "loss": 0.4633,
+      "step": 14979
+    },
+    {
+      "epoch": 1.9500195236235847,
+      "grad_norm": 2.4117796421051025,
+      "learning_rate": 2.839237602114672e-06,
+      "loss": 0.5026,
+      "step": 14982
+    },
+    {
+      "epoch": 1.9504099960952752,
+      "grad_norm": 3.0987799167633057,
+      "learning_rate": 2.837353168540522e-06,
+      "loss": 0.5195,
+      "step": 14985
+    },
+    {
+      "epoch": 1.950800468566966,
+      "grad_norm": 2.7744829654693604,
+      "learning_rate": 2.8354691127762256e-06,
+      "loss": 0.4605,
+      "step": 14988
+    },
+    {
+      "epoch": 1.9511909410386568,
+      "grad_norm": 2.757351875305176,
+      "learning_rate": 2.8335854351509223e-06,
+      "loss": 0.506,
+      "step": 14991
+    },
+    {
+      "epoch": 1.9515814135103475,
+      "grad_norm": 2.551328659057617,
+      "learning_rate": 2.8317021359936837e-06,
+      "loss": 0.5723,
+      "step": 14994
+    },
+    {
+      "epoch": 1.9519718859820383,
+      "grad_norm": 2.6989192962646484,
+      "learning_rate": 2.829819215633523e-06,
+      "loss": 0.4556,
+      "step": 14997
+    },
+    {
+      "epoch": 1.9523623584537289,
+      "grad_norm": 2.8726086616516113,
+      "learning_rate": 2.8279366743993776e-06,
+      "loss": 0.4711,
+      "step": 15000
+    },
+    {
+      "epoch": 1.9527528309254198,
+      "grad_norm": 2.742509603500366,
+      "learning_rate": 2.8260545126201277e-06,
+      "loss": 0.4994,
+      "step": 15003
+    },
+    {
+      "epoch": 1.9531433033971104,
+      "grad_norm": 2.504814624786377,
+      "learning_rate": 2.824172730624579e-06,
+      "loss": 0.5377,
+      "step": 15006
+    },
+    {
+      "epoch": 1.9535337758688014,
+      "grad_norm": 2.9309048652648926,
+      "learning_rate": 2.82229132874148e-06,
+      "loss": 0.5133,
+      "step": 15009
+    },
+    {
+      "epoch": 1.953924248340492,
+      "grad_norm": 4.551779270172119,
+      "learning_rate": 2.8204103072995036e-06,
+      "loss": 0.4612,
+      "step": 15012
+    },
+    {
+      "epoch": 1.9543147208121827,
+      "grad_norm": 3.107567548751831,
+      "learning_rate": 2.81852966662726e-06,
+      "loss": 0.5182,
+      "step": 15015
+    },
+    {
+      "epoch": 1.9547051932838735,
+      "grad_norm": 2.6413309574127197,
+      "learning_rate": 2.8166494070532958e-06,
+      "loss": 0.4866,
+      "step": 15018
+    },
+    {
+      "epoch": 1.9550956657555643,
+      "grad_norm": 2.4750940799713135,
+      "learning_rate": 2.8147695289060874e-06,
+      "loss": 0.4426,
+      "step": 15021
+    },
+    {
+      "epoch": 1.955486138227255,
+      "grad_norm": 3.1785085201263428,
+      "learning_rate": 2.8128900325140427e-06,
+      "loss": 0.473,
+      "step": 15024
+    },
+    {
+      "epoch": 1.9558766106989456,
+      "grad_norm": 2.49540114402771,
+      "learning_rate": 2.8110109182055112e-06,
+      "loss": 0.4889,
+      "step": 15027
+    },
+    {
+      "epoch": 1.9562670831706366,
+      "grad_norm": 2.723663330078125,
+      "learning_rate": 2.8091321863087672e-06,
+      "loss": 0.541,
+      "step": 15030
+    },
+    {
+      "epoch": 1.9566575556423271,
+      "grad_norm": 3.729959011077881,
+      "learning_rate": 2.8072538371520208e-06,
+      "loss": 0.4935,
+      "step": 15033
+    },
+    {
+      "epoch": 1.9570480281140181,
+      "grad_norm": 2.766430139541626,
+      "learning_rate": 2.805375871063415e-06,
+      "loss": 0.5315,
+      "step": 15036
+    },
+    {
+      "epoch": 1.9574385005857087,
+      "grad_norm": 2.1455156803131104,
+      "learning_rate": 2.8034982883710293e-06,
+      "loss": 0.4341,
+      "step": 15039
+    },
+    {
+      "epoch": 1.9578289730573994,
+      "grad_norm": 2.6431660652160645,
+      "learning_rate": 2.8016210894028694e-06,
+      "loss": 0.4833,
+      "step": 15042
+    },
+    {
+      "epoch": 1.9582194455290902,
+      "grad_norm": 2.4709603786468506,
+      "learning_rate": 2.799744274486883e-06,
+      "loss": 0.4393,
+      "step": 15045
+    },
+    {
+      "epoch": 1.958609918000781,
+      "grad_norm": 2.707580804824829,
+      "learning_rate": 2.797867843950941e-06,
+      "loss": 0.4678,
+      "step": 15048
+    },
+    {
+      "epoch": 1.9590003904724718,
+      "grad_norm": 2.598074197769165,
+      "learning_rate": 2.795991798122856e-06,
+      "loss": 0.5941,
+      "step": 15051
+    },
+    {
+      "epoch": 1.9593908629441623,
+      "grad_norm": 2.8426082134246826,
+      "learning_rate": 2.794116137330367e-06,
+      "loss": 0.5295,
+      "step": 15054
+    },
+    {
+      "epoch": 1.9597813354158533,
+      "grad_norm": 2.351702928543091,
+      "learning_rate": 2.792240861901147e-06,
+      "loss": 0.4504,
+      "step": 15057
+    },
+    {
+      "epoch": 1.9601718078875439,
+      "grad_norm": 2.866417646408081,
+      "learning_rate": 2.7903659721628063e-06,
+      "loss": 0.4783,
+      "step": 15060
+    },
+    {
+      "epoch": 1.9605622803592346,
+      "grad_norm": 2.6855812072753906,
+      "learning_rate": 2.788491468442881e-06,
+      "loss": 0.4801,
+      "step": 15063
+    },
+    {
+      "epoch": 1.9609527528309254,
+      "grad_norm": 2.5768182277679443,
+      "learning_rate": 2.7866173510688423e-06,
+      "loss": 0.4532,
+      "step": 15066
+    },
+    {
+      "epoch": 1.9613432253026162,
+      "grad_norm": 2.8649215698242188,
+      "learning_rate": 2.7847436203680977e-06,
+      "loss": 0.5368,
+      "step": 15069
+    },
+    {
+      "epoch": 1.961733697774307,
+      "grad_norm": 3.515437602996826,
+      "learning_rate": 2.7828702766679827e-06,
+      "loss": 0.5493,
+      "step": 15072
+    },
+    {
+      "epoch": 1.9621241702459975,
+      "grad_norm": 2.8081016540527344,
+      "learning_rate": 2.780997320295764e-06,
+      "loss": 0.5247,
+      "step": 15075
+    },
+    {
+      "epoch": 1.9625146427176885,
+      "grad_norm": 2.8509280681610107,
+      "learning_rate": 2.7791247515786475e-06,
+      "loss": 0.5239,
+      "step": 15078
+    },
+    {
+      "epoch": 1.962905115189379,
+      "grad_norm": 2.9370276927948,
+      "learning_rate": 2.777252570843765e-06,
+      "loss": 0.4763,
+      "step": 15081
+    },
+    {
+      "epoch": 1.96329558766107,
+      "grad_norm": 3.092346429824829,
+      "learning_rate": 2.77538077841818e-06,
+      "loss": 0.4743,
+      "step": 15084
+    },
+    {
+      "epoch": 1.9636860601327606,
+      "grad_norm": 2.7640228271484375,
+      "learning_rate": 2.7735093746288933e-06,
+      "loss": 0.5355,
+      "step": 15087
+    },
+    {
+      "epoch": 1.9640765326044514,
+      "grad_norm": 2.5823822021484375,
+      "learning_rate": 2.7716383598028367e-06,
+      "loss": 0.4613,
+      "step": 15090
+    },
+    {
+      "epoch": 1.9644670050761421,
+      "grad_norm": 2.547419786453247,
+      "learning_rate": 2.769767734266871e-06,
+      "loss": 0.5324,
+      "step": 15093
+    },
+    {
+      "epoch": 1.964857477547833,
+      "grad_norm": 2.7568259239196777,
+      "learning_rate": 2.7678974983477907e-06,
+      "loss": 0.5248,
+      "step": 15096
+    },
+    {
+      "epoch": 1.9652479500195237,
+      "grad_norm": 2.7360951900482178,
+      "learning_rate": 2.7660276523723195e-06,
+      "loss": 0.4737,
+      "step": 15099
+    },
+    {
+      "epoch": 1.9656384224912142,
+      "grad_norm": 2.4766454696655273,
+      "learning_rate": 2.7641581966671203e-06,
+      "loss": 0.5236,
+      "step": 15102
+    },
+    {
+      "epoch": 1.9660288949629052,
+      "grad_norm": 3.0755808353424072,
+      "learning_rate": 2.7622891315587803e-06,
+      "loss": 0.4267,
+      "step": 15105
+    },
+    {
+      "epoch": 1.9664193674345958,
+      "grad_norm": 2.7649264335632324,
+      "learning_rate": 2.760420457373819e-06,
+      "loss": 0.4645,
+      "step": 15108
+    },
+    {
+      "epoch": 1.9668098399062868,
+      "grad_norm": 2.694270372390747,
+      "learning_rate": 2.7585521744386954e-06,
+      "loss": 0.5818,
+      "step": 15111
+    },
+    {
+      "epoch": 1.9672003123779773,
+      "grad_norm": 2.4713313579559326,
+      "learning_rate": 2.7566842830797914e-06,
+      "loss": 0.4033,
+      "step": 15114
+    },
+    {
+      "epoch": 1.967590784849668,
+      "grad_norm": 2.431579351425171,
+      "learning_rate": 2.754816783623421e-06,
+      "loss": 0.3948,
+      "step": 15117
+    },
+    {
+      "epoch": 1.9679812573213589,
+      "grad_norm": 3.099750518798828,
+      "learning_rate": 2.7529496763958385e-06,
+      "loss": 0.4857,
+      "step": 15120
+    },
+    {
+      "epoch": 1.9683717297930496,
+      "grad_norm": 2.6645519733428955,
+      "learning_rate": 2.7510829617232197e-06,
+      "loss": 0.5021,
+      "step": 15123
+    },
+    {
+      "epoch": 1.9687622022647404,
+      "grad_norm": 3.6451914310455322,
+      "learning_rate": 2.7492166399316746e-06,
+      "loss": 0.495,
+      "step": 15126
+    },
+    {
+      "epoch": 1.969152674736431,
+      "grad_norm": 2.5093352794647217,
+      "learning_rate": 2.7473507113472477e-06,
+      "loss": 0.4996,
+      "step": 15129
+    },
+    {
+      "epoch": 1.969543147208122,
+      "grad_norm": 2.9083104133605957,
+      "learning_rate": 2.7454851762959146e-06,
+      "loss": 0.5105,
+      "step": 15132
+    },
+    {
+      "epoch": 1.9699336196798125,
+      "grad_norm": 2.977738618850708,
+      "learning_rate": 2.7436200351035784e-06,
+      "loss": 0.5711,
+      "step": 15135
+    },
+    {
+      "epoch": 1.9703240921515033,
+      "grad_norm": 2.227489948272705,
+      "learning_rate": 2.7417552880960736e-06,
+      "loss": 0.4285,
+      "step": 15138
+    },
+    {
+      "epoch": 1.970714564623194,
+      "grad_norm": 2.565983295440674,
+      "learning_rate": 2.739890935599171e-06,
+      "loss": 0.4802,
+      "step": 15141
+    },
+    {
+      "epoch": 1.9711050370948848,
+      "grad_norm": 2.6615631580352783,
+      "learning_rate": 2.738026977938567e-06,
+      "loss": 0.4581,
+      "step": 15144
+    },
+    {
+      "epoch": 1.9714955095665756,
+      "grad_norm": 2.5697174072265625,
+      "learning_rate": 2.736163415439892e-06,
+      "loss": 0.4529,
+      "step": 15147
+    },
+    {
+      "epoch": 1.9718859820382661,
+      "grad_norm": 2.6330325603485107,
+      "learning_rate": 2.734300248428704e-06,
+      "loss": 0.5237,
+      "step": 15150
+    },
+    {
+      "epoch": 1.9722764545099571,
+      "grad_norm": 2.766741991043091,
+      "learning_rate": 2.7324374772304978e-06,
+      "loss": 0.4946,
+      "step": 15153
+    },
+    {
+      "epoch": 1.9726669269816477,
+      "grad_norm": 2.515528917312622,
+      "learning_rate": 2.7305751021706943e-06,
+      "loss": 0.5103,
+      "step": 15156
+    },
+    {
+      "epoch": 1.9730573994533387,
+      "grad_norm": 2.4971110820770264,
+      "learning_rate": 2.7287131235746446e-06,
+      "loss": 0.4888,
+      "step": 15159
+    },
+    {
+      "epoch": 1.9734478719250292,
+      "grad_norm": 2.7696313858032227,
+      "learning_rate": 2.7268515417676354e-06,
+      "loss": 0.4618,
+      "step": 15162
+    },
+    {
+      "epoch": 1.97383834439672,
+      "grad_norm": 2.6422767639160156,
+      "learning_rate": 2.7249903570748805e-06,
+      "loss": 0.5001,
+      "step": 15165
+    },
+    {
+      "epoch": 1.9742288168684108,
+      "grad_norm": 2.5587644577026367,
+      "learning_rate": 2.7231295698215223e-06,
+      "loss": 0.4142,
+      "step": 15168
+    },
+    {
+      "epoch": 1.9746192893401016,
+      "grad_norm": 2.5920650959014893,
+      "learning_rate": 2.721269180332638e-06,
+      "loss": 0.4981,
+      "step": 15171
+    },
+    {
+      "epoch": 1.9750097618117923,
+      "grad_norm": 2.7538814544677734,
+      "learning_rate": 2.7194091889332364e-06,
+      "loss": 0.5542,
+      "step": 15174
+    },
+    {
+      "epoch": 1.9754002342834829,
+      "grad_norm": 2.732511281967163,
+      "learning_rate": 2.717549595948251e-06,
+      "loss": 0.4629,
+      "step": 15177
+    },
+    {
+      "epoch": 1.9757907067551739,
+      "grad_norm": 3.137880325317383,
+      "learning_rate": 2.7156904017025485e-06,
+      "loss": 0.4911,
+      "step": 15180
+    },
+    {
+      "epoch": 1.9761811792268644,
+      "grad_norm": 3.1518633365631104,
+      "learning_rate": 2.7138316065209298e-06,
+      "loss": 0.5212,
+      "step": 15183
+    },
+    {
+      "epoch": 1.9765716516985552,
+      "grad_norm": 2.9920685291290283,
+      "learning_rate": 2.7119732107281193e-06,
+      "loss": 0.4845,
+      "step": 15186
+    },
+    {
+      "epoch": 1.976962124170246,
+      "grad_norm": 2.8355932235717773,
+      "learning_rate": 2.710115214648775e-06,
+      "loss": 0.5633,
+      "step": 15189
+    },
+    {
+      "epoch": 1.9773525966419367,
+      "grad_norm": 3.028092861175537,
+      "learning_rate": 2.708257618607485e-06,
+      "loss": 0.5472,
+      "step": 15192
+    },
+    {
+      "epoch": 1.9777430691136275,
+      "grad_norm": 2.849092483520508,
+      "learning_rate": 2.7064004229287688e-06,
+      "loss": 0.5292,
+      "step": 15195
+    },
+    {
+      "epoch": 1.9781335415853183,
+      "grad_norm": 2.641608238220215,
+      "learning_rate": 2.704543627937074e-06,
+      "loss": 0.5607,
+      "step": 15198
+    },
+    {
+      "epoch": 1.978524014057009,
+      "grad_norm": 2.6991071701049805,
+      "learning_rate": 2.702687233956777e-06,
+      "loss": 0.4641,
+      "step": 15201
+    },
+    {
+      "epoch": 1.9789144865286996,
+      "grad_norm": 2.600306272506714,
+      "learning_rate": 2.7008312413121886e-06,
+      "loss": 0.5011,
+      "step": 15204
+    },
+    {
+      "epoch": 1.9793049590003906,
+      "grad_norm": 2.6777749061584473,
+      "learning_rate": 2.6989756503275454e-06,
+      "loss": 0.488,
+      "step": 15207
+    },
+    {
+      "epoch": 1.9796954314720812,
+      "grad_norm": 2.550184965133667,
+      "learning_rate": 2.697120461327014e-06,
+      "loss": 0.5401,
+      "step": 15210
+    },
+    {
+      "epoch": 1.980085903943772,
+      "grad_norm": 2.9522573947906494,
+      "learning_rate": 2.6952656746346937e-06,
+      "loss": 0.5706,
+      "step": 15213
+    },
+    {
+      "epoch": 1.9804763764154627,
+      "grad_norm": 3.306687355041504,
+      "learning_rate": 2.6934112905746136e-06,
+      "loss": 0.5365,
+      "step": 15216
+    },
+    {
+      "epoch": 1.9808668488871535,
+      "grad_norm": 3.0320816040039062,
+      "learning_rate": 2.6915573094707282e-06,
+      "loss": 0.4609,
+      "step": 15219
+    },
+    {
+      "epoch": 1.9812573213588442,
+      "grad_norm": 2.547999620437622,
+      "learning_rate": 2.689703731646922e-06,
+      "loss": 0.5136,
+      "step": 15222
+    },
+    {
+      "epoch": 1.9816477938305348,
+      "grad_norm": 3.0180954933166504,
+      "learning_rate": 2.687850557427017e-06,
+      "loss": 0.4992,
+      "step": 15225
+    },
+    {
+      "epoch": 1.9820382663022258,
+      "grad_norm": 3.133898973464966,
+      "learning_rate": 2.685997787134755e-06,
+      "loss": 0.51,
+      "step": 15228
+    },
+    {
+      "epoch": 1.9824287387739163,
+      "grad_norm": 2.631911039352417,
+      "learning_rate": 2.6841454210938095e-06,
+      "loss": 0.4903,
+      "step": 15231
+    },
+    {
+      "epoch": 1.9828192112456073,
+      "grad_norm": 4.11314582824707,
+      "learning_rate": 2.6822934596277893e-06,
+      "loss": 0.5155,
+      "step": 15234
+    },
+    {
+      "epoch": 1.9832096837172979,
+      "grad_norm": 2.362086772918701,
+      "learning_rate": 2.6804419030602256e-06,
+      "loss": 0.461,
+      "step": 15237
+    },
+    {
+      "epoch": 1.9836001561889887,
+      "grad_norm": 2.590529203414917,
+      "learning_rate": 2.6785907517145825e-06,
+      "loss": 0.4995,
+      "step": 15240
+    },
+    {
+      "epoch": 1.9839906286606794,
+      "grad_norm": 2.6236050128936768,
+      "learning_rate": 2.676740005914249e-06,
+      "loss": 0.5005,
+      "step": 15243
+    },
+    {
+      "epoch": 1.9843811011323702,
+      "grad_norm": 2.9158525466918945,
+      "learning_rate": 2.6748896659825507e-06,
+      "loss": 0.5008,
+      "step": 15246
+    },
+    {
+      "epoch": 1.984771573604061,
+      "grad_norm": 2.9477591514587402,
+      "learning_rate": 2.673039732242737e-06,
+      "loss": 0.4673,
+      "step": 15249
+    },
+    {
+      "epoch": 1.9851620460757515,
+      "grad_norm": 4.2003397941589355,
+      "learning_rate": 2.671190205017985e-06,
+      "loss": 0.4733,
+      "step": 15252
+    },
+    {
+      "epoch": 1.9855525185474425,
+      "grad_norm": 2.9860401153564453,
+      "learning_rate": 2.669341084631405e-06,
+      "loss": 0.4566,
+      "step": 15255
+    },
+    {
+      "epoch": 1.985942991019133,
+      "grad_norm": 2.4780070781707764,
+      "learning_rate": 2.6674923714060365e-06,
+      "loss": 0.476,
+      "step": 15258
+    },
+    {
+      "epoch": 1.9863334634908238,
+      "grad_norm": 2.6294336318969727,
+      "learning_rate": 2.6656440656648434e-06,
+      "loss": 0.4586,
+      "step": 15261
+    },
+    {
+      "epoch": 1.9867239359625146,
+      "grad_norm": 2.643993616104126,
+      "learning_rate": 2.66379616773072e-06,
+      "loss": 0.4864,
+      "step": 15264
+    },
+    {
+      "epoch": 1.9871144084342054,
+      "grad_norm": 2.825512647628784,
+      "learning_rate": 2.6619486779264924e-06,
+      "loss": 0.5533,
+      "step": 15267
+    },
+    {
+      "epoch": 1.9875048809058962,
+      "grad_norm": 2.878120183944702,
+      "learning_rate": 2.6601015965749135e-06,
+      "loss": 0.4189,
+      "step": 15270
+    },
+    {
+      "epoch": 1.9878953533775867,
+      "grad_norm": 2.5639851093292236,
+      "learning_rate": 2.65825492399866e-06,
+      "loss": 0.5051,
+      "step": 15273
+    },
+    {
+      "epoch": 1.9882858258492777,
+      "grad_norm": 2.7018072605133057,
+      "learning_rate": 2.6564086605203478e-06,
+      "loss": 0.4696,
+      "step": 15276
+    },
+    {
+      "epoch": 1.9886762983209683,
+      "grad_norm": 2.7019267082214355,
+      "learning_rate": 2.654562806462512e-06,
+      "loss": 0.4619,
+      "step": 15279
+    },
+    {
+      "epoch": 1.9890667707926593,
+      "grad_norm": 2.6082382202148438,
+      "learning_rate": 2.652717362147618e-06,
+      "loss": 0.5409,
+      "step": 15282
+    },
+    {
+      "epoch": 1.9894572432643498,
+      "grad_norm": 2.8006742000579834,
+      "learning_rate": 2.6508723278980654e-06,
+      "loss": 0.4565,
+      "step": 15285
+    },
+    {
+      "epoch": 1.9898477157360406,
+      "grad_norm": 2.544560194015503,
+      "learning_rate": 2.6490277040361743e-06,
+      "loss": 0.3934,
+      "step": 15288
+    },
+    {
+      "epoch": 1.9902381882077314,
+      "grad_norm": 2.632777214050293,
+      "learning_rate": 2.647183490884198e-06,
+      "loss": 0.5214,
+      "step": 15291
+    },
+    {
+      "epoch": 1.9906286606794221,
+      "grad_norm": 2.4380311965942383,
+      "learning_rate": 2.6453396887643124e-06,
+      "loss": 0.4517,
+      "step": 15294
+    },
+    {
+      "epoch": 1.991019133151113,
+      "grad_norm": 2.867617607116699,
+      "learning_rate": 2.6434962979986334e-06,
+      "loss": 0.4878,
+      "step": 15297
+    },
+    {
+      "epoch": 1.9914096056228034,
+      "grad_norm": 2.7441964149475098,
+      "learning_rate": 2.641653318909194e-06,
+      "loss": 0.5228,
+      "step": 15300
+    },
+    {
+      "epoch": 1.9918000780944944,
+      "grad_norm": 2.700723171234131,
+      "learning_rate": 2.6398107518179584e-06,
+      "loss": 0.4856,
+      "step": 15303
+    },
+    {
+      "epoch": 1.992190550566185,
+      "grad_norm": 2.6741185188293457,
+      "learning_rate": 2.637968597046818e-06,
+      "loss": 0.5286,
+      "step": 15306
+    },
+    {
+      "epoch": 1.992581023037876,
+      "grad_norm": 2.861276388168335,
+      "learning_rate": 2.6361268549175957e-06,
+      "loss": 0.5157,
+      "step": 15309
+    },
+    {
+      "epoch": 1.9929714955095665,
+      "grad_norm": 2.8468337059020996,
+      "learning_rate": 2.6342855257520393e-06,
+      "loss": 0.4486,
+      "step": 15312
+    },
+    {
+      "epoch": 1.9933619679812573,
+      "grad_norm": 2.4954781532287598,
+      "learning_rate": 2.632444609871824e-06,
+      "loss": 0.4334,
+      "step": 15315
+    },
+    {
+      "epoch": 1.993752440452948,
+      "grad_norm": 2.777869939804077,
+      "learning_rate": 2.630604107598555e-06,
+      "loss": 0.4058,
+      "step": 15318
+    },
+    {
+      "epoch": 1.9941429129246389,
+      "grad_norm": 3.0267093181610107,
+      "learning_rate": 2.6287640192537645e-06,
+      "loss": 0.5133,
+      "step": 15321
+    },
+    {
+      "epoch": 1.9945333853963296,
+      "grad_norm": 2.5115318298339844,
+      "learning_rate": 2.62692434515891e-06,
+      "loss": 0.4361,
+      "step": 15324
+    },
+    {
+      "epoch": 1.9949238578680202,
+      "grad_norm": 2.824373245239258,
+      "learning_rate": 2.6250850856353815e-06,
+      "loss": 0.4602,
+      "step": 15327
+    },
+    {
+      "epoch": 1.9953143303397112,
+      "grad_norm": 2.420016050338745,
+      "learning_rate": 2.6232462410044927e-06,
+      "loss": 0.4136,
+      "step": 15330
+    },
+    {
+      "epoch": 1.9957048028114017,
+      "grad_norm": 2.629142999649048,
+      "learning_rate": 2.6214078115874843e-06,
+      "loss": 0.5071,
+      "step": 15333
+    },
+    {
+      "epoch": 1.9960952752830925,
+      "grad_norm": 2.6193673610687256,
+      "learning_rate": 2.6195697977055262e-06,
+      "loss": 0.4901,
+      "step": 15336
+    },
+    {
+      "epoch": 1.9964857477547833,
+      "grad_norm": 3.3041958808898926,
+      "learning_rate": 2.6177321996797193e-06,
+      "loss": 0.4341,
+      "step": 15339
+    },
+    {
+      "epoch": 1.996876220226474,
+      "grad_norm": 2.496450424194336,
+      "learning_rate": 2.615895017831086e-06,
+      "loss": 0.4542,
+      "step": 15342
+    },
+    {
+      "epoch": 1.9972666926981648,
+      "grad_norm": 2.4021811485290527,
+      "learning_rate": 2.6140582524805746e-06,
+      "loss": 0.4821,
+      "step": 15345
+    },
+    {
+      "epoch": 1.9976571651698554,
+      "grad_norm": 2.5634875297546387,
+      "learning_rate": 2.6122219039490704e-06,
+      "loss": 0.4816,
+      "step": 15348
+    },
+    {
+      "epoch": 1.9980476376415464,
+      "grad_norm": 2.7742042541503906,
+      "learning_rate": 2.6103859725573756e-06,
+      "loss": 0.5172,
+      "step": 15351
+    },
+    {
+      "epoch": 1.998438110113237,
+      "grad_norm": 2.705730676651001,
+      "learning_rate": 2.6085504586262245e-06,
+      "loss": 0.4998,
+      "step": 15354
+    },
+    {
+      "epoch": 1.998828582584928,
+      "grad_norm": 2.538831949234009,
+      "learning_rate": 2.606715362476275e-06,
+      "loss": 0.474,
+      "step": 15357
+    },
+    {
+      "epoch": 1.9992190550566185,
+      "grad_norm": 2.5744338035583496,
+      "learning_rate": 2.6048806844281206e-06,
+      "loss": 0.5268,
+      "step": 15360
+    },
+    {
+      "epoch": 1.9996095275283092,
+      "grad_norm": 2.8102900981903076,
+      "learning_rate": 2.6030464248022704e-06,
+      "loss": 0.5102,
+      "step": 15363
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.22767162322998,
+      "learning_rate": 2.601212583919166e-06,
+      "loss": 0.5386,
+      "step": 15366
+    },
+    {
+      "epoch": 2.0003904724716906,
+      "grad_norm": 2.527303695678711,
+      "learning_rate": 2.5993791620991783e-06,
+      "loss": 0.3502,
+      "step": 15369
+    },
+    {
+      "epoch": 2.0007809449433815,
+      "grad_norm": 2.931882858276367,
+      "learning_rate": 2.5975461596626016e-06,
+      "loss": 0.3747,
+      "step": 15372
+    },
+    {
+      "epoch": 2.001171417415072,
+      "grad_norm": 2.3010025024414062,
+      "learning_rate": 2.5957135769296543e-06,
+      "loss": 0.3607,
+      "step": 15375
+    },
+    {
+      "epoch": 2.001561889886763,
+      "grad_norm": 2.2463386058807373,
+      "learning_rate": 2.5938814142204873e-06,
+      "loss": 0.3475,
+      "step": 15378
+    },
+    {
+      "epoch": 2.0019523623584536,
+      "grad_norm": 2.4900364875793457,
+      "learning_rate": 2.592049671855178e-06,
+      "loss": 0.3501,
+      "step": 15381
+    },
+    {
+      "epoch": 2.0023428348301446,
+      "grad_norm": 2.3296518325805664,
+      "learning_rate": 2.5902183501537247e-06,
+      "loss": 0.4047,
+      "step": 15384
+    },
+    {
+      "epoch": 2.002733307301835,
+      "grad_norm": 2.306422710418701,
+      "learning_rate": 2.5883874494360544e-06,
+      "loss": 0.372,
+      "step": 15387
+    },
+    {
+      "epoch": 2.003123779773526,
+      "grad_norm": 2.479424238204956,
+      "learning_rate": 2.5865569700220257e-06,
+      "loss": 0.3311,
+      "step": 15390
+    },
+    {
+      "epoch": 2.0035142522452167,
+      "grad_norm": 2.6242549419403076,
+      "learning_rate": 2.584726912231417e-06,
+      "loss": 0.3669,
+      "step": 15393
+    },
+    {
+      "epoch": 2.0039047247169073,
+      "grad_norm": 2.8145828247070312,
+      "learning_rate": 2.582897276383933e-06,
+      "loss": 0.411,
+      "step": 15396
+    },
+    {
+      "epoch": 2.0042951971885983,
+      "grad_norm": 2.4189159870147705,
+      "learning_rate": 2.5810680627992134e-06,
+      "loss": 0.3667,
+      "step": 15399
+    },
+    {
+      "epoch": 2.004685669660289,
+      "grad_norm": 2.6705586910247803,
+      "learning_rate": 2.579239271796814e-06,
+      "loss": 0.3744,
+      "step": 15402
+    },
+    {
+      "epoch": 2.00507614213198,
+      "grad_norm": 2.6565775871276855,
+      "learning_rate": 2.5774109036962208e-06,
+      "loss": 0.3976,
+      "step": 15405
+    },
+    {
+      "epoch": 2.0054666146036704,
+      "grad_norm": 2.923213005065918,
+      "learning_rate": 2.5755829588168444e-06,
+      "loss": 0.3592,
+      "step": 15408
+    },
+    {
+      "epoch": 2.0058570870753614,
+      "grad_norm": 2.6897075176239014,
+      "learning_rate": 2.573755437478027e-06,
+      "loss": 0.3244,
+      "step": 15411
+    },
+    {
+      "epoch": 2.006247559547052,
+      "grad_norm": 2.722161293029785,
+      "learning_rate": 2.571928339999031e-06,
+      "loss": 0.2853,
+      "step": 15414
+    },
+    {
+      "epoch": 2.0066380320187425,
+      "grad_norm": 2.8695123195648193,
+      "learning_rate": 2.570101666699044e-06,
+      "loss": 0.3987,
+      "step": 15417
+    },
+    {
+      "epoch": 2.0070285044904335,
+      "grad_norm": 2.4492764472961426,
+      "learning_rate": 2.5682754178971838e-06,
+      "loss": 0.3593,
+      "step": 15420
+    },
+    {
+      "epoch": 2.007418976962124,
+      "grad_norm": 2.7777905464172363,
+      "learning_rate": 2.5664495939124945e-06,
+      "loss": 0.3572,
+      "step": 15423
+    },
+    {
+      "epoch": 2.007809449433815,
+      "grad_norm": 2.52078914642334,
+      "learning_rate": 2.564624195063942e-06,
+      "loss": 0.346,
+      "step": 15426
+    },
+    {
+      "epoch": 2.0081999219055056,
+      "grad_norm": 2.5213940143585205,
+      "learning_rate": 2.5627992216704167e-06,
+      "loss": 0.3278,
+      "step": 15429
+    },
+    {
+      "epoch": 2.0085903943771966,
+      "grad_norm": 2.6475985050201416,
+      "learning_rate": 2.560974674050743e-06,
+      "loss": 0.3902,
+      "step": 15432
+    },
+    {
+      "epoch": 2.008980866848887,
+      "grad_norm": 2.8680548667907715,
+      "learning_rate": 2.5591505525236626e-06,
+      "loss": 0.3529,
+      "step": 15435
+    },
+    {
+      "epoch": 2.009371339320578,
+      "grad_norm": 2.852402687072754,
+      "learning_rate": 2.557326857407844e-06,
+      "loss": 0.3346,
+      "step": 15438
+    },
+    {
+      "epoch": 2.0097618117922686,
+      "grad_norm": 2.724487066268921,
+      "learning_rate": 2.555503589021886e-06,
+      "loss": 0.3578,
+      "step": 15441
+    },
+    {
+      "epoch": 2.010152284263959,
+      "grad_norm": 3.298837900161743,
+      "learning_rate": 2.553680747684309e-06,
+      "loss": 0.3784,
+      "step": 15444
+    },
+    {
+      "epoch": 2.01054275673565,
+      "grad_norm": 2.7460174560546875,
+      "learning_rate": 2.551858333713557e-06,
+      "loss": 0.3613,
+      "step": 15447
+    },
+    {
+      "epoch": 2.0109332292073407,
+      "grad_norm": 2.6980228424072266,
+      "learning_rate": 2.5500363474280066e-06,
+      "loss": 0.3708,
+      "step": 15450
+    },
+    {
+      "epoch": 2.0113237016790317,
+      "grad_norm": 2.433493137359619,
+      "learning_rate": 2.548214789145951e-06,
+      "loss": 0.3527,
+      "step": 15453
+    },
+    {
+      "epoch": 2.0117141741507223,
+      "grad_norm": 2.826143741607666,
+      "learning_rate": 2.5463936591856153e-06,
+      "loss": 0.3262,
+      "step": 15456
+    },
+    {
+      "epoch": 2.0121046466224133,
+      "grad_norm": 2.4797494411468506,
+      "learning_rate": 2.5445729578651427e-06,
+      "loss": 0.3859,
+      "step": 15459
+    },
+    {
+      "epoch": 2.012495119094104,
+      "grad_norm": 2.5826714038848877,
+      "learning_rate": 2.5427526855026097e-06,
+      "loss": 0.3374,
+      "step": 15462
+    },
+    {
+      "epoch": 2.012885591565795,
+      "grad_norm": 2.6993298530578613,
+      "learning_rate": 2.540932842416015e-06,
+      "loss": 0.3859,
+      "step": 15465
+    },
+    {
+      "epoch": 2.0132760640374854,
+      "grad_norm": 2.468688488006592,
+      "learning_rate": 2.5391134289232794e-06,
+      "loss": 0.3522,
+      "step": 15468
+    },
+    {
+      "epoch": 2.013666536509176,
+      "grad_norm": 3.0440926551818848,
+      "learning_rate": 2.5372944453422486e-06,
+      "loss": 0.3925,
+      "step": 15471
+    },
+    {
+      "epoch": 2.014057008980867,
+      "grad_norm": 2.1184885501861572,
+      "learning_rate": 2.5354758919906995e-06,
+      "loss": 0.2867,
+      "step": 15474
+    },
+    {
+      "epoch": 2.0144474814525575,
+      "grad_norm": 2.9124364852905273,
+      "learning_rate": 2.5336577691863286e-06,
+      "loss": 0.3881,
+      "step": 15477
+    },
+    {
+      "epoch": 2.0148379539242485,
+      "grad_norm": 2.6721181869506836,
+      "learning_rate": 2.531840077246754e-06,
+      "loss": 0.3463,
+      "step": 15480
+    },
+    {
+      "epoch": 2.015228426395939,
+      "grad_norm": 2.512578248977661,
+      "learning_rate": 2.5300228164895275e-06,
+      "loss": 0.3355,
+      "step": 15483
+    },
+    {
+      "epoch": 2.01561889886763,
+      "grad_norm": 2.8622634410858154,
+      "learning_rate": 2.5282059872321192e-06,
+      "loss": 0.4173,
+      "step": 15486
+    },
+    {
+      "epoch": 2.0160093713393206,
+      "grad_norm": 2.445178985595703,
+      "learning_rate": 2.526389589791923e-06,
+      "loss": 0.3202,
+      "step": 15489
+    },
+    {
+      "epoch": 2.016399843811011,
+      "grad_norm": 2.7001988887786865,
+      "learning_rate": 2.524573624486264e-06,
+      "loss": 0.3657,
+      "step": 15492
+    },
+    {
+      "epoch": 2.016790316282702,
+      "grad_norm": 2.5869991779327393,
+      "learning_rate": 2.5227580916323846e-06,
+      "loss": 0.4093,
+      "step": 15495
+    },
+    {
+      "epoch": 2.0171807887543927,
+      "grad_norm": 2.6643974781036377,
+      "learning_rate": 2.5209429915474536e-06,
+      "loss": 0.3307,
+      "step": 15498
+    },
+    {
+      "epoch": 2.0175712612260837,
+      "grad_norm": 2.8480355739593506,
+      "learning_rate": 2.5191283245485686e-06,
+      "loss": 0.3815,
+      "step": 15501
+    },
+    {
+      "epoch": 2.017961733697774,
+      "grad_norm": 2.674461603164673,
+      "learning_rate": 2.517314090952745e-06,
+      "loss": 0.3721,
+      "step": 15504
+    },
+    {
+      "epoch": 2.018352206169465,
+      "grad_norm": 2.5194647312164307,
+      "learning_rate": 2.515500291076928e-06,
+      "loss": 0.303,
+      "step": 15507
+    },
+    {
+      "epoch": 2.0187426786411558,
+      "grad_norm": 2.625236749649048,
+      "learning_rate": 2.5136869252379825e-06,
+      "loss": 0.4136,
+      "step": 15510
+    },
+    {
+      "epoch": 2.0191331511128467,
+      "grad_norm": 2.7869434356689453,
+      "learning_rate": 2.511873993752702e-06,
+      "loss": 0.3184,
+      "step": 15513
+    },
+    {
+      "epoch": 2.0195236235845373,
+      "grad_norm": 2.7091245651245117,
+      "learning_rate": 2.5100614969378006e-06,
+      "loss": 0.3589,
+      "step": 15516
+    },
+    {
+      "epoch": 2.019914096056228,
+      "grad_norm": 3.2992379665374756,
+      "learning_rate": 2.508249435109918e-06,
+      "loss": 0.3377,
+      "step": 15519
+    },
+    {
+      "epoch": 2.020304568527919,
+      "grad_norm": 2.686725616455078,
+      "learning_rate": 2.5064378085856146e-06,
+      "loss": 0.3125,
+      "step": 15522
+    },
+    {
+      "epoch": 2.0206950409996094,
+      "grad_norm": 2.7225749492645264,
+      "learning_rate": 2.5046266176813825e-06,
+      "loss": 0.3388,
+      "step": 15525
+    },
+    {
+      "epoch": 2.0210855134713004,
+      "grad_norm": 2.5445516109466553,
+      "learning_rate": 2.5028158627136313e-06,
+      "loss": 0.3304,
+      "step": 15528
+    },
+    {
+      "epoch": 2.021475985942991,
+      "grad_norm": 2.5298032760620117,
+      "learning_rate": 2.5010055439986935e-06,
+      "loss": 0.359,
+      "step": 15531
+    },
+    {
+      "epoch": 2.021866458414682,
+      "grad_norm": 2.7906692028045654,
+      "learning_rate": 2.4991956618528317e-06,
+      "loss": 0.3837,
+      "step": 15534
+    },
+    {
+      "epoch": 2.0222569308863725,
+      "grad_norm": 2.5515687465667725,
+      "learning_rate": 2.4973862165922268e-06,
+      "loss": 0.3741,
+      "step": 15537
+    },
+    {
+      "epoch": 2.0226474033580635,
+      "grad_norm": 2.7983710765838623,
+      "learning_rate": 2.495577208532984e-06,
+      "loss": 0.4018,
+      "step": 15540
+    },
+    {
+      "epoch": 2.023037875829754,
+      "grad_norm": 2.931863784790039,
+      "learning_rate": 2.493768637991135e-06,
+      "loss": 0.3165,
+      "step": 15543
+    },
+    {
+      "epoch": 2.0234283483014446,
+      "grad_norm": 2.761305809020996,
+      "learning_rate": 2.491960505282632e-06,
+      "loss": 0.3567,
+      "step": 15546
+    },
+    {
+      "epoch": 2.0238188207731356,
+      "grad_norm": 2.8153398036956787,
+      "learning_rate": 2.4901528107233535e-06,
+      "loss": 0.3898,
+      "step": 15549
+    },
+    {
+      "epoch": 2.024209293244826,
+      "grad_norm": 2.77380633354187,
+      "learning_rate": 2.4883455546290975e-06,
+      "loss": 0.3286,
+      "step": 15552
+    },
+    {
+      "epoch": 2.024599765716517,
+      "grad_norm": 2.8505361080169678,
+      "learning_rate": 2.486538737315591e-06,
+      "loss": 0.3429,
+      "step": 15555
+    },
+    {
+      "epoch": 2.0249902381882077,
+      "grad_norm": 2.722792863845825,
+      "learning_rate": 2.4847323590984797e-06,
+      "loss": 0.4518,
+      "step": 15558
+    },
+    {
+      "epoch": 2.0253807106598987,
+      "grad_norm": 2.7249341011047363,
+      "learning_rate": 2.482926420293332e-06,
+      "loss": 0.3427,
+      "step": 15561
+    },
+    {
+      "epoch": 2.025771183131589,
+      "grad_norm": 2.7627315521240234,
+      "learning_rate": 2.4811209212156455e-06,
+      "loss": 0.3595,
+      "step": 15564
+    },
+    {
+      "epoch": 2.0261616556032798,
+      "grad_norm": 2.6366686820983887,
+      "learning_rate": 2.479315862180835e-06,
+      "loss": 0.3352,
+      "step": 15567
+    },
+    {
+      "epoch": 2.0265521280749708,
+      "grad_norm": 2.850191354751587,
+      "learning_rate": 2.477511243504241e-06,
+      "loss": 0.3398,
+      "step": 15570
+    },
+    {
+      "epoch": 2.0269426005466613,
+      "grad_norm": 3.085858106613159,
+      "learning_rate": 2.475707065501124e-06,
+      "loss": 0.3551,
+      "step": 15573
+    },
+    {
+      "epoch": 2.0273330730183523,
+      "grad_norm": 2.575150728225708,
+      "learning_rate": 2.473903328486674e-06,
+      "loss": 0.3393,
+      "step": 15576
+    },
+    {
+      "epoch": 2.027723545490043,
+      "grad_norm": 2.692265748977661,
+      "learning_rate": 2.4721000327759988e-06,
+      "loss": 0.4056,
+      "step": 15579
+    },
+    {
+      "epoch": 2.028114017961734,
+      "grad_norm": 2.6966402530670166,
+      "learning_rate": 2.4702971786841278e-06,
+      "loss": 0.3393,
+      "step": 15582
+    },
+    {
+      "epoch": 2.0285044904334244,
+      "grad_norm": 2.7873032093048096,
+      "learning_rate": 2.46849476652602e-06,
+      "loss": 0.3656,
+      "step": 15585
+    },
+    {
+      "epoch": 2.0288949629051154,
+      "grad_norm": 2.6634371280670166,
+      "learning_rate": 2.4666927966165487e-06,
+      "loss": 0.3544,
+      "step": 15588
+    },
+    {
+      "epoch": 2.029285435376806,
+      "grad_norm": 2.635133981704712,
+      "learning_rate": 2.464891269270519e-06,
+      "loss": 0.4251,
+      "step": 15591
+    },
+    {
+      "epoch": 2.0296759078484965,
+      "grad_norm": 2.715653657913208,
+      "learning_rate": 2.4630901848026494e-06,
+      "loss": 0.3665,
+      "step": 15594
+    },
+    {
+      "epoch": 2.0300663803201875,
+      "grad_norm": 4.321242332458496,
+      "learning_rate": 2.4612895435275896e-06,
+      "loss": 0.3857,
+      "step": 15597
+    },
+    {
+      "epoch": 2.030456852791878,
+      "grad_norm": 2.4996755123138428,
+      "learning_rate": 2.4594893457599056e-06,
+      "loss": 0.3667,
+      "step": 15600
+    },
+    {
+      "epoch": 2.030847325263569,
+      "grad_norm": 2.7729170322418213,
+      "learning_rate": 2.4576895918140866e-06,
+      "loss": 0.3656,
+      "step": 15603
+    },
+    {
+      "epoch": 2.0312377977352596,
+      "grad_norm": 2.7078745365142822,
+      "learning_rate": 2.45589028200455e-06,
+      "loss": 0.3683,
+      "step": 15606
+    },
+    {
+      "epoch": 2.0316282702069506,
+      "grad_norm": 2.8833117485046387,
+      "learning_rate": 2.4540914166456286e-06,
+      "loss": 0.353,
+      "step": 15609
+    },
+    {
+      "epoch": 2.032018742678641,
+      "grad_norm": 3.260715961456299,
+      "learning_rate": 2.452292996051581e-06,
+      "loss": 0.4125,
+      "step": 15612
+    },
+    {
+      "epoch": 2.0324092151503317,
+      "grad_norm": 2.7390503883361816,
+      "learning_rate": 2.450495020536586e-06,
+      "loss": 0.4213,
+      "step": 15615
+    },
+    {
+      "epoch": 2.0327996876220227,
+      "grad_norm": 2.833608865737915,
+      "learning_rate": 2.4486974904147488e-06,
+      "loss": 0.3744,
+      "step": 15618
+    },
+    {
+      "epoch": 2.0331901600937132,
+      "grad_norm": 3.231580972671509,
+      "learning_rate": 2.446900406000093e-06,
+      "loss": 0.3103,
+      "step": 15621
+    },
+    {
+      "epoch": 2.0335806325654042,
+      "grad_norm": 2.8441002368927,
+      "learning_rate": 2.445103767606563e-06,
+      "loss": 0.3917,
+      "step": 15624
+    },
+    {
+      "epoch": 2.0339711050370948,
+      "grad_norm": 2.8229074478149414,
+      "learning_rate": 2.443307575548033e-06,
+      "loss": 0.345,
+      "step": 15627
+    },
+    {
+      "epoch": 2.0343615775087858,
+      "grad_norm": 2.594104766845703,
+      "learning_rate": 2.4415118301382885e-06,
+      "loss": 0.3214,
+      "step": 15630
+    },
+    {
+      "epoch": 2.0347520499804763,
+      "grad_norm": 3.496530532836914,
+      "learning_rate": 2.4397165316910472e-06,
+      "loss": 0.4241,
+      "step": 15633
+    },
+    {
+      "epoch": 2.0351425224521673,
+      "grad_norm": 2.7391481399536133,
+      "learning_rate": 2.4379216805199396e-06,
+      "loss": 0.3217,
+      "step": 15636
+    },
+    {
+      "epoch": 2.035532994923858,
+      "grad_norm": 2.622236728668213,
+      "learning_rate": 2.436127276938526e-06,
+      "loss": 0.4088,
+      "step": 15639
+    },
+    {
+      "epoch": 2.0359234673955484,
+      "grad_norm": 2.8432962894439697,
+      "learning_rate": 2.434333321260285e-06,
+      "loss": 0.3536,
+      "step": 15642
+    },
+    {
+      "epoch": 2.0363139398672394,
+      "grad_norm": 2.6063525676727295,
+      "learning_rate": 2.432539813798612e-06,
+      "loss": 0.378,
+      "step": 15645
+    },
+    {
+      "epoch": 2.03670441233893,
+      "grad_norm": 2.529179334640503,
+      "learning_rate": 2.430746754866835e-06,
+      "loss": 0.3053,
+      "step": 15648
+    },
+    {
+      "epoch": 2.037094884810621,
+      "grad_norm": 2.843583583831787,
+      "learning_rate": 2.428954144778195e-06,
+      "loss": 0.4077,
+      "step": 15651
+    },
+    {
+      "epoch": 2.0374853572823115,
+      "grad_norm": 2.8022096157073975,
+      "learning_rate": 2.4271619838458552e-06,
+      "loss": 0.3926,
+      "step": 15654
+    },
+    {
+      "epoch": 2.0378758297540025,
+      "grad_norm": 2.6679294109344482,
+      "learning_rate": 2.4253702723829066e-06,
+      "loss": 0.3386,
+      "step": 15657
+    },
+    {
+      "epoch": 2.038266302225693,
+      "grad_norm": 2.6503212451934814,
+      "learning_rate": 2.423579010702355e-06,
+      "loss": 0.3481,
+      "step": 15660
+    },
+    {
+      "epoch": 2.038656774697384,
+      "grad_norm": 2.8369200229644775,
+      "learning_rate": 2.4217881991171297e-06,
+      "loss": 0.4236,
+      "step": 15663
+    },
+    {
+      "epoch": 2.0390472471690746,
+      "grad_norm": 2.750505208969116,
+      "learning_rate": 2.4199978379400806e-06,
+      "loss": 0.4346,
+      "step": 15666
+    },
+    {
+      "epoch": 2.039437719640765,
+      "grad_norm": 2.6629180908203125,
+      "learning_rate": 2.418207927483984e-06,
+      "loss": 0.3667,
+      "step": 15669
+    },
+    {
+      "epoch": 2.039828192112456,
+      "grad_norm": 2.9319400787353516,
+      "learning_rate": 2.416418468061529e-06,
+      "loss": 0.4105,
+      "step": 15672
+    },
+    {
+      "epoch": 2.0402186645841467,
+      "grad_norm": 3.028381109237671,
+      "learning_rate": 2.4146294599853348e-06,
+      "loss": 0.3763,
+      "step": 15675
+    },
+    {
+      "epoch": 2.0406091370558377,
+      "grad_norm": 2.613823890686035,
+      "learning_rate": 2.412840903567933e-06,
+      "loss": 0.3848,
+      "step": 15678
+    },
+    {
+      "epoch": 2.0409996095275282,
+      "grad_norm": 2.842607021331787,
+      "learning_rate": 2.411052799121784e-06,
+      "loss": 0.3756,
+      "step": 15681
+    },
+    {
+      "epoch": 2.0413900819992192,
+      "grad_norm": 2.8494303226470947,
+      "learning_rate": 2.409265146959265e-06,
+      "loss": 0.382,
+      "step": 15684
+    },
+    {
+      "epoch": 2.04178055447091,
+      "grad_norm": 3.151146650314331,
+      "learning_rate": 2.4074779473926734e-06,
+      "loss": 0.349,
+      "step": 15687
+    },
+    {
+      "epoch": 2.0421710269426003,
+      "grad_norm": 2.663558006286621,
+      "learning_rate": 2.405691200734232e-06,
+      "loss": 0.4037,
+      "step": 15690
+    },
+    {
+      "epoch": 2.0425614994142913,
+      "grad_norm": 2.7715821266174316,
+      "learning_rate": 2.40390490729608e-06,
+      "loss": 0.333,
+      "step": 15693
+    },
+    {
+      "epoch": 2.042951971885982,
+      "grad_norm": 2.496654748916626,
+      "learning_rate": 2.4021190673902777e-06,
+      "loss": 0.3529,
+      "step": 15696
+    },
+    {
+      "epoch": 2.043342444357673,
+      "grad_norm": 2.606013059616089,
+      "learning_rate": 2.4003336813288112e-06,
+      "loss": 0.3328,
+      "step": 15699
+    },
+    {
+      "epoch": 2.0437329168293634,
+      "grad_norm": 2.736743211746216,
+      "learning_rate": 2.3985487494235814e-06,
+      "loss": 0.3804,
+      "step": 15702
+    },
+    {
+      "epoch": 2.0441233893010544,
+      "grad_norm": 2.691873550415039,
+      "learning_rate": 2.396764271986411e-06,
+      "loss": 0.3932,
+      "step": 15705
+    },
+    {
+      "epoch": 2.044513861772745,
+      "grad_norm": 3.0015580654144287,
+      "learning_rate": 2.3949802493290475e-06,
+      "loss": 0.3425,
+      "step": 15708
+    },
+    {
+      "epoch": 2.044904334244436,
+      "grad_norm": 3.084564208984375,
+      "learning_rate": 2.393196681763154e-06,
+      "loss": 0.3792,
+      "step": 15711
+    },
+    {
+      "epoch": 2.0452948067161265,
+      "grad_norm": 2.7556662559509277,
+      "learning_rate": 2.3914135696003144e-06,
+      "loss": 0.3621,
+      "step": 15714
+    },
+    {
+      "epoch": 2.045685279187817,
+      "grad_norm": 2.8170785903930664,
+      "learning_rate": 2.3896309131520367e-06,
+      "loss": 0.336,
+      "step": 15717
+    },
+    {
+      "epoch": 2.046075751659508,
+      "grad_norm": 2.6950416564941406,
+      "learning_rate": 2.387848712729749e-06,
+      "loss": 0.3685,
+      "step": 15720
+    },
+    {
+      "epoch": 2.0464662241311986,
+      "grad_norm": 3.019321918487549,
+      "learning_rate": 2.386066968644796e-06,
+      "loss": 0.3972,
+      "step": 15723
+    },
+    {
+      "epoch": 2.0468566966028896,
+      "grad_norm": 2.6395628452301025,
+      "learning_rate": 2.384285681208445e-06,
+      "loss": 0.3128,
+      "step": 15726
+    },
+    {
+      "epoch": 2.04724716907458,
+      "grad_norm": 2.728309392929077,
+      "learning_rate": 2.3825048507318806e-06,
+      "loss": 0.3894,
+      "step": 15729
+    },
+    {
+      "epoch": 2.047637641546271,
+      "grad_norm": 2.9258289337158203,
+      "learning_rate": 2.380724477526214e-06,
+      "loss": 0.4404,
+      "step": 15732
+    },
+    {
+      "epoch": 2.0480281140179617,
+      "grad_norm": 3.706683397293091,
+      "learning_rate": 2.3789445619024716e-06,
+      "loss": 0.3687,
+      "step": 15735
+    },
+    {
+      "epoch": 2.0484185864896527,
+      "grad_norm": 2.782982349395752,
+      "learning_rate": 2.3771651041715978e-06,
+      "loss": 0.3007,
+      "step": 15738
+    },
+    {
+      "epoch": 2.0488090589613432,
+      "grad_norm": 2.7452354431152344,
+      "learning_rate": 2.3753861046444647e-06,
+      "loss": 0.3442,
+      "step": 15741
+    },
+    {
+      "epoch": 2.049199531433034,
+      "grad_norm": 2.8988213539123535,
+      "learning_rate": 2.373607563631858e-06,
+      "loss": 0.3732,
+      "step": 15744
+    },
+    {
+      "epoch": 2.049590003904725,
+      "grad_norm": 3.2119078636169434,
+      "learning_rate": 2.371829481444483e-06,
+      "loss": 0.4084,
+      "step": 15747
+    },
+    {
+      "epoch": 2.0499804763764153,
+      "grad_norm": 2.665807008743286,
+      "learning_rate": 2.3700518583929704e-06,
+      "loss": 0.3438,
+      "step": 15750
+    },
+    {
+      "epoch": 2.0503709488481063,
+      "grad_norm": 3.3220880031585693,
+      "learning_rate": 2.3682746947878653e-06,
+      "loss": 0.3147,
+      "step": 15753
+    },
+    {
+      "epoch": 2.050761421319797,
+      "grad_norm": 2.919126510620117,
+      "learning_rate": 2.3664979909396334e-06,
+      "loss": 0.3434,
+      "step": 15756
+    },
+    {
+      "epoch": 2.051151893791488,
+      "grad_norm": 2.8159265518188477,
+      "learning_rate": 2.364721747158662e-06,
+      "loss": 0.37,
+      "step": 15759
+    },
+    {
+      "epoch": 2.0515423662631784,
+      "grad_norm": 2.8563246726989746,
+      "learning_rate": 2.3629459637552593e-06,
+      "loss": 0.3405,
+      "step": 15762
+    },
+    {
+      "epoch": 2.051932838734869,
+      "grad_norm": 2.744643449783325,
+      "learning_rate": 2.3611706410396497e-06,
+      "loss": 0.4583,
+      "step": 15765
+    },
+    {
+      "epoch": 2.05232331120656,
+      "grad_norm": 2.9006803035736084,
+      "learning_rate": 2.3593957793219757e-06,
+      "loss": 0.4063,
+      "step": 15768
+    },
+    {
+      "epoch": 2.0527137836782505,
+      "grad_norm": 2.7164108753204346,
+      "learning_rate": 2.357621378912306e-06,
+      "loss": 0.3646,
+      "step": 15771
+    },
+    {
+      "epoch": 2.0531042561499415,
+      "grad_norm": 2.580453395843506,
+      "learning_rate": 2.3558474401206222e-06,
+      "loss": 0.3469,
+      "step": 15774
+    },
+    {
+      "epoch": 2.053494728621632,
+      "grad_norm": 3.763662099838257,
+      "learning_rate": 2.354073963256829e-06,
+      "loss": 0.3854,
+      "step": 15777
+    },
+    {
+      "epoch": 2.053885201093323,
+      "grad_norm": 2.7697255611419678,
+      "learning_rate": 2.352300948630745e-06,
+      "loss": 0.294,
+      "step": 15780
+    },
+    {
+      "epoch": 2.0542756735650136,
+      "grad_norm": 2.634119749069214,
+      "learning_rate": 2.350528396552118e-06,
+      "loss": 0.3524,
+      "step": 15783
+    },
+    {
+      "epoch": 2.0546661460367046,
+      "grad_norm": 2.6129114627838135,
+      "learning_rate": 2.348756307330607e-06,
+      "loss": 0.3661,
+      "step": 15786
+    },
+    {
+      "epoch": 2.055056618508395,
+      "grad_norm": 3.03295636177063,
+      "learning_rate": 2.3469846812757892e-06,
+      "loss": 0.3947,
+      "step": 15789
+    },
+    {
+      "epoch": 2.0554470909800857,
+      "grad_norm": 2.5799500942230225,
+      "learning_rate": 2.345213518697168e-06,
+      "loss": 0.3157,
+      "step": 15792
+    },
+    {
+      "epoch": 2.0558375634517767,
+      "grad_norm": 2.8251993656158447,
+      "learning_rate": 2.343442819904161e-06,
+      "loss": 0.3364,
+      "step": 15795
+    },
+    {
+      "epoch": 2.0562280359234673,
+      "grad_norm": 3.4497342109680176,
+      "learning_rate": 2.341672585206102e-06,
+      "loss": 0.3281,
+      "step": 15798
+    },
+    {
+      "epoch": 2.0566185083951583,
+      "grad_norm": 2.605712652206421,
+      "learning_rate": 2.339902814912251e-06,
+      "loss": 0.4153,
+      "step": 15801
+    },
+    {
+      "epoch": 2.057008980866849,
+      "grad_norm": 2.881829261779785,
+      "learning_rate": 2.3381335093317837e-06,
+      "loss": 0.3554,
+      "step": 15804
+    },
+    {
+      "epoch": 2.05739945333854,
+      "grad_norm": 2.6509134769439697,
+      "learning_rate": 2.3363646687737925e-06,
+      "loss": 0.3419,
+      "step": 15807
+    },
+    {
+      "epoch": 2.0577899258102303,
+      "grad_norm": 2.6324145793914795,
+      "learning_rate": 2.3345962935472884e-06,
+      "loss": 0.3845,
+      "step": 15810
+    },
+    {
+      "epoch": 2.0581803982819213,
+      "grad_norm": 3.0626988410949707,
+      "learning_rate": 2.3328283839612063e-06,
+      "loss": 0.3581,
+      "step": 15813
+    },
+    {
+      "epoch": 2.058570870753612,
+      "grad_norm": 2.609157085418701,
+      "learning_rate": 2.331060940324395e-06,
+      "loss": 0.3613,
+      "step": 15816
+    },
+    {
+      "epoch": 2.0589613432253024,
+      "grad_norm": 2.7834722995758057,
+      "learning_rate": 2.3292939629456206e-06,
+      "loss": 0.418,
+      "step": 15819
+    },
+    {
+      "epoch": 2.0593518156969934,
+      "grad_norm": 2.971318006515503,
+      "learning_rate": 2.3275274521335743e-06,
+      "loss": 0.3108,
+      "step": 15822
+    },
+    {
+      "epoch": 2.059742288168684,
+      "grad_norm": 3.0066676139831543,
+      "learning_rate": 2.3257614081968606e-06,
+      "loss": 0.3441,
+      "step": 15825
+    },
+    {
+      "epoch": 2.060132760640375,
+      "grad_norm": 2.56901216506958,
+      "learning_rate": 2.3239958314440027e-06,
+      "loss": 0.3672,
+      "step": 15828
+    },
+    {
+      "epoch": 2.0605232331120655,
+      "grad_norm": 2.7101621627807617,
+      "learning_rate": 2.3222307221834417e-06,
+      "loss": 0.3722,
+      "step": 15831
+    },
+    {
+      "epoch": 2.0609137055837565,
+      "grad_norm": 2.768842935562134,
+      "learning_rate": 2.3204660807235426e-06,
+      "loss": 0.4064,
+      "step": 15834
+    },
+    {
+      "epoch": 2.061304178055447,
+      "grad_norm": 2.6292800903320312,
+      "learning_rate": 2.3187019073725816e-06,
+      "loss": 0.3984,
+      "step": 15837
+    },
+    {
+      "epoch": 2.0616946505271376,
+      "grad_norm": 2.5181849002838135,
+      "learning_rate": 2.3169382024387547e-06,
+      "loss": 0.3555,
+      "step": 15840
+    },
+    {
+      "epoch": 2.0620851229988286,
+      "grad_norm": 2.9010989665985107,
+      "learning_rate": 2.3151749662301803e-06,
+      "loss": 0.3347,
+      "step": 15843
+    },
+    {
+      "epoch": 2.062475595470519,
+      "grad_norm": 2.7419636249542236,
+      "learning_rate": 2.313412199054893e-06,
+      "loss": 0.3382,
+      "step": 15846
+    },
+    {
+      "epoch": 2.06286606794221,
+      "grad_norm": 2.6093106269836426,
+      "learning_rate": 2.3116499012208428e-06,
+      "loss": 0.3602,
+      "step": 15849
+    },
+    {
+      "epoch": 2.0632565404139007,
+      "grad_norm": 2.8060364723205566,
+      "learning_rate": 2.3098880730358968e-06,
+      "loss": 0.3394,
+      "step": 15852
+    },
+    {
+      "epoch": 2.0636470128855917,
+      "grad_norm": 2.528196334838867,
+      "learning_rate": 2.308126714807848e-06,
+      "loss": 0.3496,
+      "step": 15855
+    },
+    {
+      "epoch": 2.0640374853572823,
+      "grad_norm": 2.6221911907196045,
+      "learning_rate": 2.306365826844399e-06,
+      "loss": 0.3142,
+      "step": 15858
+    },
+    {
+      "epoch": 2.0644279578289733,
+      "grad_norm": 2.5942671298980713,
+      "learning_rate": 2.3046054094531715e-06,
+      "loss": 0.3163,
+      "step": 15861
+    },
+    {
+      "epoch": 2.064818430300664,
+      "grad_norm": 2.879770040512085,
+      "learning_rate": 2.3028454629417106e-06,
+      "loss": 0.3983,
+      "step": 15864
+    },
+    {
+      "epoch": 2.0652089027723544,
+      "grad_norm": 3.0213136672973633,
+      "learning_rate": 2.3010859876174734e-06,
+      "loss": 0.3028,
+      "step": 15867
+    },
+    {
+      "epoch": 2.0655993752440454,
+      "grad_norm": 2.5107553005218506,
+      "learning_rate": 2.2993269837878346e-06,
+      "loss": 0.355,
+      "step": 15870
+    },
+    {
+      "epoch": 2.065989847715736,
+      "grad_norm": 2.631068229675293,
+      "learning_rate": 2.297568451760092e-06,
+      "loss": 0.387,
+      "step": 15873
+    },
+    {
+      "epoch": 2.066380320187427,
+      "grad_norm": 3.205779552459717,
+      "learning_rate": 2.295810391841456e-06,
+      "loss": 0.3629,
+      "step": 15876
+    },
+    {
+      "epoch": 2.0667707926591175,
+      "grad_norm": 2.532114028930664,
+      "learning_rate": 2.294052804339056e-06,
+      "loss": 0.4009,
+      "step": 15879
+    },
+    {
+      "epoch": 2.0671612651308084,
+      "grad_norm": 2.673632860183716,
+      "learning_rate": 2.292295689559934e-06,
+      "loss": 0.3439,
+      "step": 15882
+    },
+    {
+      "epoch": 2.067551737602499,
+      "grad_norm": 2.8151535987854004,
+      "learning_rate": 2.2905390478110635e-06,
+      "loss": 0.3562,
+      "step": 15885
+    },
+    {
+      "epoch": 2.0679422100741895,
+      "grad_norm": 2.834294557571411,
+      "learning_rate": 2.2887828793993212e-06,
+      "loss": 0.3683,
+      "step": 15888
+    },
+    {
+      "epoch": 2.0683326825458805,
+      "grad_norm": 2.7955880165100098,
+      "learning_rate": 2.287027184631506e-06,
+      "loss": 0.4139,
+      "step": 15891
+    },
+    {
+      "epoch": 2.068723155017571,
+      "grad_norm": 3.368584394454956,
+      "learning_rate": 2.285271963814333e-06,
+      "loss": 0.3226,
+      "step": 15894
+    },
+    {
+      "epoch": 2.069113627489262,
+      "grad_norm": 2.835724353790283,
+      "learning_rate": 2.2835172172544384e-06,
+      "loss": 0.3797,
+      "step": 15897
+    },
+    {
+      "epoch": 2.0695040999609526,
+      "grad_norm": 3.315255880355835,
+      "learning_rate": 2.281762945258372e-06,
+      "loss": 0.379,
+      "step": 15900
+    },
+    {
+      "epoch": 2.0698945724326436,
+      "grad_norm": 2.753709316253662,
+      "learning_rate": 2.2800091481325983e-06,
+      "loss": 0.3426,
+      "step": 15903
+    },
+    {
+      "epoch": 2.070285044904334,
+      "grad_norm": 3.0424439907073975,
+      "learning_rate": 2.278255826183506e-06,
+      "loss": 0.4311,
+      "step": 15906
+    },
+    {
+      "epoch": 2.070675517376025,
+      "grad_norm": 2.6219685077667236,
+      "learning_rate": 2.2765029797173954e-06,
+      "loss": 0.3532,
+      "step": 15909
+    },
+    {
+      "epoch": 2.0710659898477157,
+      "grad_norm": 3.344362735748291,
+      "learning_rate": 2.274750609040483e-06,
+      "loss": 0.3858,
+      "step": 15912
+    },
+    {
+      "epoch": 2.0714564623194063,
+      "grad_norm": 3.3275110721588135,
+      "learning_rate": 2.2729987144589083e-06,
+      "loss": 0.3672,
+      "step": 15915
+    },
+    {
+      "epoch": 2.0718469347910973,
+      "grad_norm": 2.963124990463257,
+      "learning_rate": 2.271247296278721e-06,
+      "loss": 0.3988,
+      "step": 15918
+    },
+    {
+      "epoch": 2.072237407262788,
+      "grad_norm": 2.8209590911865234,
+      "learning_rate": 2.2694963548058885e-06,
+      "loss": 0.3627,
+      "step": 15921
+    },
+    {
+      "epoch": 2.072627879734479,
+      "grad_norm": 2.65281081199646,
+      "learning_rate": 2.2677458903462994e-06,
+      "loss": 0.4011,
+      "step": 15924
+    },
+    {
+      "epoch": 2.0730183522061694,
+      "grad_norm": 2.652015209197998,
+      "learning_rate": 2.2659959032057566e-06,
+      "loss": 0.3114,
+      "step": 15927
+    },
+    {
+      "epoch": 2.0734088246778604,
+      "grad_norm": 2.65907883644104,
+      "learning_rate": 2.2642463936899785e-06,
+      "loss": 0.3698,
+      "step": 15930
+    },
+    {
+      "epoch": 2.073799297149551,
+      "grad_norm": 2.8742198944091797,
+      "learning_rate": 2.2624973621045983e-06,
+      "loss": 0.3786,
+      "step": 15933
+    },
+    {
+      "epoch": 2.074189769621242,
+      "grad_norm": 2.9052553176879883,
+      "learning_rate": 2.2607488087551716e-06,
+      "loss": 0.3868,
+      "step": 15936
+    },
+    {
+      "epoch": 2.0745802420929325,
+      "grad_norm": 2.704512596130371,
+      "learning_rate": 2.2590007339471657e-06,
+      "loss": 0.3433,
+      "step": 15939
+    },
+    {
+      "epoch": 2.074970714564623,
+      "grad_norm": 2.6483607292175293,
+      "learning_rate": 2.257253137985966e-06,
+      "loss": 0.3306,
+      "step": 15942
+    },
+    {
+      "epoch": 2.075361187036314,
+      "grad_norm": 2.474104642868042,
+      "learning_rate": 2.25550602117687e-06,
+      "loss": 0.3005,
+      "step": 15945
+    },
+    {
+      "epoch": 2.0757516595080046,
+      "grad_norm": 2.596907138824463,
+      "learning_rate": 2.2537593838251016e-06,
+      "loss": 0.3277,
+      "step": 15948
+    },
+    {
+      "epoch": 2.0761421319796955,
+      "grad_norm": 2.7480621337890625,
+      "learning_rate": 2.252013226235791e-06,
+      "loss": 0.4159,
+      "step": 15951
+    },
+    {
+      "epoch": 2.076532604451386,
+      "grad_norm": 2.6667375564575195,
+      "learning_rate": 2.250267548713987e-06,
+      "loss": 0.3156,
+      "step": 15954
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 3.392796516418457,
+      "learning_rate": 2.2485223515646597e-06,
+      "loss": 0.3844,
+      "step": 15957
+    },
+    {
+      "epoch": 2.0773135493947676,
+      "grad_norm": 2.91853404045105,
+      "learning_rate": 2.246777635092689e-06,
+      "loss": 0.4067,
+      "step": 15960
+    },
+    {
+      "epoch": 2.0777040218664586,
+      "grad_norm": 2.7192370891571045,
+      "learning_rate": 2.245033399602872e-06,
+      "loss": 0.3416,
+      "step": 15963
+    },
+    {
+      "epoch": 2.078094494338149,
+      "grad_norm": 2.7866413593292236,
+      "learning_rate": 2.2432896453999243e-06,
+      "loss": 0.3529,
+      "step": 15966
+    },
+    {
+      "epoch": 2.0784849668098397,
+      "grad_norm": 2.8930845260620117,
+      "learning_rate": 2.2415463727884785e-06,
+      "loss": 0.3483,
+      "step": 15969
+    },
+    {
+      "epoch": 2.0788754392815307,
+      "grad_norm": 2.707551956176758,
+      "learning_rate": 2.239803582073078e-06,
+      "loss": 0.3398,
+      "step": 15972
+    },
+    {
+      "epoch": 2.0792659117532213,
+      "grad_norm": 2.8100128173828125,
+      "learning_rate": 2.2380612735581835e-06,
+      "loss": 0.3944,
+      "step": 15975
+    },
+    {
+      "epoch": 2.0796563842249123,
+      "grad_norm": 2.5131068229675293,
+      "learning_rate": 2.236319447548176e-06,
+      "loss": 0.3561,
+      "step": 15978
+    },
+    {
+      "epoch": 2.080046856696603,
+      "grad_norm": 2.6995739936828613,
+      "learning_rate": 2.234578104347347e-06,
+      "loss": 0.3422,
+      "step": 15981
+    },
+    {
+      "epoch": 2.080437329168294,
+      "grad_norm": 4.326542854309082,
+      "learning_rate": 2.2328372442599057e-06,
+      "loss": 0.3936,
+      "step": 15984
+    },
+    {
+      "epoch": 2.0808278016399844,
+      "grad_norm": 2.6090240478515625,
+      "learning_rate": 2.231096867589975e-06,
+      "loss": 0.3217,
+      "step": 15987
+    },
+    {
+      "epoch": 2.081218274111675,
+      "grad_norm": 3.206176519393921,
+      "learning_rate": 2.2293569746415976e-06,
+      "loss": 0.3001,
+      "step": 15990
+    },
+    {
+      "epoch": 2.081608746583366,
+      "grad_norm": 2.488536834716797,
+      "learning_rate": 2.2276175657187288e-06,
+      "loss": 0.2904,
+      "step": 15993
+    },
+    {
+      "epoch": 2.0819992190550565,
+      "grad_norm": 2.5979604721069336,
+      "learning_rate": 2.225878641125237e-06,
+      "loss": 0.3798,
+      "step": 15996
+    },
+    {
+      "epoch": 2.0823896915267475,
+      "grad_norm": 2.6461970806121826,
+      "learning_rate": 2.2241402011649127e-06,
+      "loss": 0.3123,
+      "step": 15999
+    },
+    {
+      "epoch": 2.082780163998438,
+      "grad_norm": 2.82248592376709,
+      "learning_rate": 2.2224022461414553e-06,
+      "loss": 0.374,
+      "step": 16002
+    },
+    {
+      "epoch": 2.083170636470129,
+      "grad_norm": 2.730311632156372,
+      "learning_rate": 2.220664776358481e-06,
+      "loss": 0.44,
+      "step": 16005
+    },
+    {
+      "epoch": 2.0835611089418196,
+      "grad_norm": 2.5103518962860107,
+      "learning_rate": 2.2189277921195228e-06,
+      "loss": 0.3888,
+      "step": 16008
+    },
+    {
+      "epoch": 2.0839515814135106,
+      "grad_norm": 2.6272377967834473,
+      "learning_rate": 2.217191293728031e-06,
+      "loss": 0.3516,
+      "step": 16011
+    },
+    {
+      "epoch": 2.084342053885201,
+      "grad_norm": 2.5898125171661377,
+      "learning_rate": 2.2154552814873663e-06,
+      "loss": 0.3519,
+      "step": 16014
+    },
+    {
+      "epoch": 2.0847325263568917,
+      "grad_norm": 2.7031917572021484,
+      "learning_rate": 2.213719755700804e-06,
+      "loss": 0.3611,
+      "step": 16017
+    },
+    {
+      "epoch": 2.0851229988285827,
+      "grad_norm": 2.8753886222839355,
+      "learning_rate": 2.21198471667154e-06,
+      "loss": 0.4155,
+      "step": 16020
+    },
+    {
+      "epoch": 2.085513471300273,
+      "grad_norm": 2.8421316146850586,
+      "learning_rate": 2.210250164702682e-06,
+      "loss": 0.3379,
+      "step": 16023
+    },
+    {
+      "epoch": 2.085903943771964,
+      "grad_norm": 2.7893972396850586,
+      "learning_rate": 2.208516100097249e-06,
+      "loss": 0.3434,
+      "step": 16026
+    },
+    {
+      "epoch": 2.0862944162436547,
+      "grad_norm": 2.8135573863983154,
+      "learning_rate": 2.206782523158183e-06,
+      "loss": 0.3668,
+      "step": 16029
+    },
+    {
+      "epoch": 2.0866848887153457,
+      "grad_norm": 2.7129807472229004,
+      "learning_rate": 2.2050494341883344e-06,
+      "loss": 0.3845,
+      "step": 16032
+    },
+    {
+      "epoch": 2.0870753611870363,
+      "grad_norm": 3.1642825603485107,
+      "learning_rate": 2.203316833490469e-06,
+      "loss": 0.3545,
+      "step": 16035
+    },
+    {
+      "epoch": 2.087465833658727,
+      "grad_norm": 2.634582757949829,
+      "learning_rate": 2.2015847213672686e-06,
+      "loss": 0.3136,
+      "step": 16038
+    },
+    {
+      "epoch": 2.087856306130418,
+      "grad_norm": 3.1089882850646973,
+      "learning_rate": 2.1998530981213318e-06,
+      "loss": 0.3507,
+      "step": 16041
+    },
+    {
+      "epoch": 2.0882467786021084,
+      "grad_norm": 3.0786855220794678,
+      "learning_rate": 2.1981219640551683e-06,
+      "loss": 0.3322,
+      "step": 16044
+    },
+    {
+      "epoch": 2.0886372510737994,
+      "grad_norm": 2.9572317600250244,
+      "learning_rate": 2.1963913194712013e-06,
+      "loss": 0.4171,
+      "step": 16047
+    },
+    {
+      "epoch": 2.08902772354549,
+      "grad_norm": 3.005173921585083,
+      "learning_rate": 2.1946611646717726e-06,
+      "loss": 0.3304,
+      "step": 16050
+    },
+    {
+      "epoch": 2.089418196017181,
+      "grad_norm": 2.7299466133117676,
+      "learning_rate": 2.192931499959139e-06,
+      "loss": 0.3869,
+      "step": 16053
+    },
+    {
+      "epoch": 2.0898086684888715,
+      "grad_norm": 3.0412538051605225,
+      "learning_rate": 2.191202325635467e-06,
+      "loss": 0.3458,
+      "step": 16056
+    },
+    {
+      "epoch": 2.0901991409605625,
+      "grad_norm": 3.1715352535247803,
+      "learning_rate": 2.1894736420028383e-06,
+      "loss": 0.3498,
+      "step": 16059
+    },
+    {
+      "epoch": 2.090589613432253,
+      "grad_norm": 2.733077049255371,
+      "learning_rate": 2.1877454493632533e-06,
+      "loss": 0.3784,
+      "step": 16062
+    },
+    {
+      "epoch": 2.0909800859039436,
+      "grad_norm": 2.767333745956421,
+      "learning_rate": 2.1860177480186224e-06,
+      "loss": 0.3984,
+      "step": 16065
+    },
+    {
+      "epoch": 2.0913705583756346,
+      "grad_norm": 2.9046363830566406,
+      "learning_rate": 2.1842905382707695e-06,
+      "loss": 0.3301,
+      "step": 16068
+    },
+    {
+      "epoch": 2.091761030847325,
+      "grad_norm": 2.561779737472534,
+      "learning_rate": 2.182563820421438e-06,
+      "loss": 0.3179,
+      "step": 16071
+    },
+    {
+      "epoch": 2.092151503319016,
+      "grad_norm": 2.805349111557007,
+      "learning_rate": 2.18083759477228e-06,
+      "loss": 0.4154,
+      "step": 16074
+    },
+    {
+      "epoch": 2.0925419757907067,
+      "grad_norm": 2.7377660274505615,
+      "learning_rate": 2.1791118616248615e-06,
+      "loss": 0.3782,
+      "step": 16077
+    },
+    {
+      "epoch": 2.0929324482623977,
+      "grad_norm": 2.9340991973876953,
+      "learning_rate": 2.1773866212806684e-06,
+      "loss": 0.3573,
+      "step": 16080
+    },
+    {
+      "epoch": 2.093322920734088,
+      "grad_norm": 3.0035879611968994,
+      "learning_rate": 2.1756618740410944e-06,
+      "loss": 0.3522,
+      "step": 16083
+    },
+    {
+      "epoch": 2.093713393205779,
+      "grad_norm": 2.919433832168579,
+      "learning_rate": 2.1739376202074504e-06,
+      "loss": 0.4147,
+      "step": 16086
+    },
+    {
+      "epoch": 2.0941038656774698,
+      "grad_norm": 2.8312833309173584,
+      "learning_rate": 2.172213860080956e-06,
+      "loss": 0.3854,
+      "step": 16089
+    },
+    {
+      "epoch": 2.0944943381491603,
+      "grad_norm": 2.7150704860687256,
+      "learning_rate": 2.1704905939627523e-06,
+      "loss": 0.3951,
+      "step": 16092
+    },
+    {
+      "epoch": 2.0948848106208513,
+      "grad_norm": 2.916109323501587,
+      "learning_rate": 2.168767822153891e-06,
+      "loss": 0.3698,
+      "step": 16095
+    },
+    {
+      "epoch": 2.095275283092542,
+      "grad_norm": 2.8838951587677,
+      "learning_rate": 2.1670455449553352e-06,
+      "loss": 0.3452,
+      "step": 16098
+    },
+    {
+      "epoch": 2.095665755564233,
+      "grad_norm": 2.8266618251800537,
+      "learning_rate": 2.1653237626679607e-06,
+      "loss": 0.3314,
+      "step": 16101
+    },
+    {
+      "epoch": 2.0960562280359234,
+      "grad_norm": 2.667581796646118,
+      "learning_rate": 2.163602475592564e-06,
+      "loss": 0.3685,
+      "step": 16104
+    },
+    {
+      "epoch": 2.0964467005076144,
+      "grad_norm": 2.537261724472046,
+      "learning_rate": 2.1618816840298474e-06,
+      "loss": 0.2771,
+      "step": 16107
+    },
+    {
+      "epoch": 2.096837172979305,
+      "grad_norm": 2.6051278114318848,
+      "learning_rate": 2.1601613882804283e-06,
+      "loss": 0.346,
+      "step": 16110
+    },
+    {
+      "epoch": 2.0972276454509955,
+      "grad_norm": 3.2211477756500244,
+      "learning_rate": 2.158441588644843e-06,
+      "loss": 0.3435,
+      "step": 16113
+    },
+    {
+      "epoch": 2.0976181179226865,
+      "grad_norm": 2.916541576385498,
+      "learning_rate": 2.1567222854235337e-06,
+      "loss": 0.3735,
+      "step": 16116
+    },
+    {
+      "epoch": 2.098008590394377,
+      "grad_norm": 2.6063573360443115,
+      "learning_rate": 2.1550034789168584e-06,
+      "loss": 0.3075,
+      "step": 16119
+    },
+    {
+      "epoch": 2.098399062866068,
+      "grad_norm": 2.746978759765625,
+      "learning_rate": 2.1532851694250916e-06,
+      "loss": 0.3371,
+      "step": 16122
+    },
+    {
+      "epoch": 2.0987895353377586,
+      "grad_norm": 2.748178243637085,
+      "learning_rate": 2.1515673572484173e-06,
+      "loss": 0.3474,
+      "step": 16125
+    },
+    {
+      "epoch": 2.0991800078094496,
+      "grad_norm": 2.8788228034973145,
+      "learning_rate": 2.1498500426869325e-06,
+      "loss": 0.3808,
+      "step": 16128
+    },
+    {
+      "epoch": 2.09957048028114,
+      "grad_norm": 3.1142728328704834,
+      "learning_rate": 2.1481332260406502e-06,
+      "loss": 0.4338,
+      "step": 16131
+    },
+    {
+      "epoch": 2.099960952752831,
+      "grad_norm": 2.57977032661438,
+      "learning_rate": 2.1464169076094922e-06,
+      "loss": 0.3976,
+      "step": 16134
+    },
+    {
+      "epoch": 2.1003514252245217,
+      "grad_norm": 2.79011869430542,
+      "learning_rate": 2.1447010876932992e-06,
+      "loss": 0.3907,
+      "step": 16137
+    },
+    {
+      "epoch": 2.1007418976962122,
+      "grad_norm": 2.7332963943481445,
+      "learning_rate": 2.142985766591818e-06,
+      "loss": 0.3246,
+      "step": 16140
+    },
+    {
+      "epoch": 2.101132370167903,
+      "grad_norm": 3.274784803390503,
+      "learning_rate": 2.141270944604715e-06,
+      "loss": 0.3906,
+      "step": 16143
+    },
+    {
+      "epoch": 2.1015228426395938,
+      "grad_norm": 2.857478618621826,
+      "learning_rate": 2.139556622031564e-06,
+      "loss": 0.5127,
+      "step": 16146
+    },
+    {
+      "epoch": 2.1019133151112848,
+      "grad_norm": 2.596496105194092,
+      "learning_rate": 2.1378427991718533e-06,
+      "loss": 0.3371,
+      "step": 16149
+    },
+    {
+      "epoch": 2.1023037875829753,
+      "grad_norm": 2.747124671936035,
+      "learning_rate": 2.1361294763249828e-06,
+      "loss": 0.3566,
+      "step": 16152
+    },
+    {
+      "epoch": 2.1026942600546663,
+      "grad_norm": 2.936936140060425,
+      "learning_rate": 2.13441665379027e-06,
+      "loss": 0.3858,
+      "step": 16155
+    },
+    {
+      "epoch": 2.103084732526357,
+      "grad_norm": 2.8474388122558594,
+      "learning_rate": 2.1327043318669396e-06,
+      "loss": 0.3285,
+      "step": 16158
+    },
+    {
+      "epoch": 2.1034752049980474,
+      "grad_norm": 2.6837317943573,
+      "learning_rate": 2.130992510854128e-06,
+      "loss": 0.3799,
+      "step": 16161
+    },
+    {
+      "epoch": 2.1038656774697384,
+      "grad_norm": 3.10606050491333,
+      "learning_rate": 2.1292811910508916e-06,
+      "loss": 0.3795,
+      "step": 16164
+    },
+    {
+      "epoch": 2.104256149941429,
+      "grad_norm": 2.5549697875976562,
+      "learning_rate": 2.127570372756192e-06,
+      "loss": 0.3574,
+      "step": 16167
+    },
+    {
+      "epoch": 2.10464662241312,
+      "grad_norm": 2.6415367126464844,
+      "learning_rate": 2.1258600562689035e-06,
+      "loss": 0.3661,
+      "step": 16170
+    },
+    {
+      "epoch": 2.1050370948848105,
+      "grad_norm": 2.8995654582977295,
+      "learning_rate": 2.124150241887819e-06,
+      "loss": 0.3972,
+      "step": 16173
+    },
+    {
+      "epoch": 2.1054275673565015,
+      "grad_norm": 2.8920884132385254,
+      "learning_rate": 2.1224409299116356e-06,
+      "loss": 0.3544,
+      "step": 16176
+    },
+    {
+      "epoch": 2.105818039828192,
+      "grad_norm": 2.7235703468322754,
+      "learning_rate": 2.1207321206389702e-06,
+      "loss": 0.3626,
+      "step": 16179
+    },
+    {
+      "epoch": 2.106208512299883,
+      "grad_norm": 2.957944869995117,
+      "learning_rate": 2.119023814368344e-06,
+      "loss": 0.3354,
+      "step": 16182
+    },
+    {
+      "epoch": 2.1065989847715736,
+      "grad_norm": 2.9017412662506104,
+      "learning_rate": 2.117316011398199e-06,
+      "loss": 0.4035,
+      "step": 16185
+    },
+    {
+      "epoch": 2.106989457243264,
+      "grad_norm": 2.5684914588928223,
+      "learning_rate": 2.115608712026882e-06,
+      "loss": 0.313,
+      "step": 16188
+    },
+    {
+      "epoch": 2.107379929714955,
+      "grad_norm": 2.6348636150360107,
+      "learning_rate": 2.113901916552653e-06,
+      "loss": 0.3506,
+      "step": 16191
+    },
+    {
+      "epoch": 2.1077704021866457,
+      "grad_norm": 2.8889946937561035,
+      "learning_rate": 2.1121956252736903e-06,
+      "loss": 0.3517,
+      "step": 16194
+    },
+    {
+      "epoch": 2.1081608746583367,
+      "grad_norm": 3.1232848167419434,
+      "learning_rate": 2.1104898384880766e-06,
+      "loss": 0.3806,
+      "step": 16197
+    },
+    {
+      "epoch": 2.1085513471300272,
+      "grad_norm": 2.5369369983673096,
+      "learning_rate": 2.10878455649381e-06,
+      "loss": 0.3079,
+      "step": 16200
+    },
+    {
+      "epoch": 2.1089418196017182,
+      "grad_norm": 2.765028953552246,
+      "learning_rate": 2.1070797795887965e-06,
+      "loss": 0.3596,
+      "step": 16203
+    },
+    {
+      "epoch": 2.1093322920734088,
+      "grad_norm": 2.585745096206665,
+      "learning_rate": 2.1053755080708614e-06,
+      "loss": 0.357,
+      "step": 16206
+    },
+    {
+      "epoch": 2.1097227645450998,
+      "grad_norm": 3.108854055404663,
+      "learning_rate": 2.1036717422377364e-06,
+      "loss": 0.3649,
+      "step": 16209
+    },
+    {
+      "epoch": 2.1101132370167903,
+      "grad_norm": 2.787053108215332,
+      "learning_rate": 2.101968482387063e-06,
+      "loss": 0.378,
+      "step": 16212
+    },
+    {
+      "epoch": 2.110503709488481,
+      "grad_norm": 2.599722146987915,
+      "learning_rate": 2.1002657288164002e-06,
+      "loss": 0.3513,
+      "step": 16215
+    },
+    {
+      "epoch": 2.110894181960172,
+      "grad_norm": 2.803863763809204,
+      "learning_rate": 2.0985634818232136e-06,
+      "loss": 0.3057,
+      "step": 16218
+    },
+    {
+      "epoch": 2.1112846544318624,
+      "grad_norm": 2.749025344848633,
+      "learning_rate": 2.096861741704884e-06,
+      "loss": 0.4025,
+      "step": 16221
+    },
+    {
+      "epoch": 2.1116751269035534,
+      "grad_norm": 2.9662301540374756,
+      "learning_rate": 2.0951605087586994e-06,
+      "loss": 0.4046,
+      "step": 16224
+    },
+    {
+      "epoch": 2.112065599375244,
+      "grad_norm": 2.85591721534729,
+      "learning_rate": 2.0934597832818653e-06,
+      "loss": 0.3853,
+      "step": 16227
+    },
+    {
+      "epoch": 2.112456071846935,
+      "grad_norm": 3.045888900756836,
+      "learning_rate": 2.0917595655714925e-06,
+      "loss": 0.3448,
+      "step": 16230
+    },
+    {
+      "epoch": 2.1128465443186255,
+      "grad_norm": 2.8176419734954834,
+      "learning_rate": 2.0900598559246032e-06,
+      "loss": 0.312,
+      "step": 16233
+    },
+    {
+      "epoch": 2.1132370167903165,
+      "grad_norm": 2.6675214767456055,
+      "learning_rate": 2.0883606546381372e-06,
+      "loss": 0.3868,
+      "step": 16236
+    },
+    {
+      "epoch": 2.113627489262007,
+      "grad_norm": 2.7344720363616943,
+      "learning_rate": 2.08666196200894e-06,
+      "loss": 0.3555,
+      "step": 16239
+    },
+    {
+      "epoch": 2.1140179617336976,
+      "grad_norm": 2.40110445022583,
+      "learning_rate": 2.084963778333768e-06,
+      "loss": 0.3102,
+      "step": 16242
+    },
+    {
+      "epoch": 2.1144084342053886,
+      "grad_norm": 2.716338634490967,
+      "learning_rate": 2.083266103909292e-06,
+      "loss": 0.3574,
+      "step": 16245
+    },
+    {
+      "epoch": 2.114798906677079,
+      "grad_norm": 2.801283121109009,
+      "learning_rate": 2.081568939032093e-06,
+      "loss": 0.3879,
+      "step": 16248
+    },
+    {
+      "epoch": 2.11518937914877,
+      "grad_norm": 2.7228639125823975,
+      "learning_rate": 2.07987228399866e-06,
+      "loss": 0.3327,
+      "step": 16251
+    },
+    {
+      "epoch": 2.1155798516204607,
+      "grad_norm": 3.10695743560791,
+      "learning_rate": 2.0781761391053944e-06,
+      "loss": 0.4271,
+      "step": 16254
+    },
+    {
+      "epoch": 2.1159703240921517,
+      "grad_norm": 2.5058023929595947,
+      "learning_rate": 2.076480504648613e-06,
+      "loss": 0.3314,
+      "step": 16257
+    },
+    {
+      "epoch": 2.1163607965638422,
+      "grad_norm": 2.8873941898345947,
+      "learning_rate": 2.074785380924535e-06,
+      "loss": 0.4765,
+      "step": 16260
+    },
+    {
+      "epoch": 2.116751269035533,
+      "grad_norm": 2.7840540409088135,
+      "learning_rate": 2.073090768229299e-06,
+      "loss": 0.4192,
+      "step": 16263
+    },
+    {
+      "epoch": 2.117141741507224,
+      "grad_norm": 2.253389596939087,
+      "learning_rate": 2.071396666858947e-06,
+      "loss": 0.3191,
+      "step": 16266
+    },
+    {
+      "epoch": 2.1175322139789143,
+      "grad_norm": 2.8525044918060303,
+      "learning_rate": 2.069703077109438e-06,
+      "loss": 0.3817,
+      "step": 16269
+    },
+    {
+      "epoch": 2.1179226864506053,
+      "grad_norm": 2.5376346111297607,
+      "learning_rate": 2.0680099992766366e-06,
+      "loss": 0.276,
+      "step": 16272
+    },
+    {
+      "epoch": 2.118313158922296,
+      "grad_norm": 2.743023633956909,
+      "learning_rate": 2.0663174336563193e-06,
+      "loss": 0.3222,
+      "step": 16275
+    },
+    {
+      "epoch": 2.118703631393987,
+      "grad_norm": 2.61458158493042,
+      "learning_rate": 2.0646253805441757e-06,
+      "loss": 0.2982,
+      "step": 16278
+    },
+    {
+      "epoch": 2.1190941038656774,
+      "grad_norm": 3.0593173503875732,
+      "learning_rate": 2.0629338402358035e-06,
+      "loss": 0.3798,
+      "step": 16281
+    },
+    {
+      "epoch": 2.1194845763373684,
+      "grad_norm": 2.7648446559906006,
+      "learning_rate": 2.0612428130267087e-06,
+      "loss": 0.4024,
+      "step": 16284
+    },
+    {
+      "epoch": 2.119875048809059,
+      "grad_norm": 3.026467800140381,
+      "learning_rate": 2.0595522992123148e-06,
+      "loss": 0.3453,
+      "step": 16287
+    },
+    {
+      "epoch": 2.1202655212807495,
+      "grad_norm": 2.49055814743042,
+      "learning_rate": 2.057862299087947e-06,
+      "loss": 0.3484,
+      "step": 16290
+    },
+    {
+      "epoch": 2.1206559937524405,
+      "grad_norm": 2.641812324523926,
+      "learning_rate": 2.056172812948846e-06,
+      "loss": 0.351,
+      "step": 16293
+    },
+    {
+      "epoch": 2.121046466224131,
+      "grad_norm": 2.7587594985961914,
+      "learning_rate": 2.0544838410901625e-06,
+      "loss": 0.3781,
+      "step": 16296
+    },
+    {
+      "epoch": 2.121436938695822,
+      "grad_norm": 2.873347759246826,
+      "learning_rate": 2.052795383806955e-06,
+      "loss": 0.3831,
+      "step": 16299
+    },
+    {
+      "epoch": 2.1218274111675126,
+      "grad_norm": 2.9109420776367188,
+      "learning_rate": 2.0511074413941934e-06,
+      "loss": 0.4032,
+      "step": 16302
+    },
+    {
+      "epoch": 2.1222178836392036,
+      "grad_norm": 2.9024267196655273,
+      "learning_rate": 2.0494200141467576e-06,
+      "loss": 0.3648,
+      "step": 16305
+    },
+    {
+      "epoch": 2.122608356110894,
+      "grad_norm": 2.8061764240264893,
+      "learning_rate": 2.04773310235944e-06,
+      "loss": 0.3407,
+      "step": 16308
+    },
+    {
+      "epoch": 2.1229988285825847,
+      "grad_norm": 2.516742706298828,
+      "learning_rate": 2.0460467063269384e-06,
+      "loss": 0.3815,
+      "step": 16311
+    },
+    {
+      "epoch": 2.1233893010542757,
+      "grad_norm": 2.9556503295898438,
+      "learning_rate": 2.0443608263438635e-06,
+      "loss": 0.3183,
+      "step": 16314
+    },
+    {
+      "epoch": 2.1237797735259663,
+      "grad_norm": 2.734612464904785,
+      "learning_rate": 2.0426754627047328e-06,
+      "loss": 0.3251,
+      "step": 16317
+    },
+    {
+      "epoch": 2.1241702459976572,
+      "grad_norm": 2.9752840995788574,
+      "learning_rate": 2.040990615703979e-06,
+      "loss": 0.3447,
+      "step": 16320
+    },
+    {
+      "epoch": 2.124560718469348,
+      "grad_norm": 2.810737371444702,
+      "learning_rate": 2.0393062856359396e-06,
+      "loss": 0.3616,
+      "step": 16323
+    },
+    {
+      "epoch": 2.124951190941039,
+      "grad_norm": 2.919602870941162,
+      "learning_rate": 2.0376224727948625e-06,
+      "loss": 0.3873,
+      "step": 16326
+    },
+    {
+      "epoch": 2.1253416634127293,
+      "grad_norm": 2.8167080879211426,
+      "learning_rate": 2.035939177474909e-06,
+      "loss": 0.347,
+      "step": 16329
+    },
+    {
+      "epoch": 2.1257321358844203,
+      "grad_norm": 3.348229169845581,
+      "learning_rate": 2.0342563999701454e-06,
+      "loss": 0.3292,
+      "step": 16332
+    },
+    {
+      "epoch": 2.126122608356111,
+      "grad_norm": 2.9174301624298096,
+      "learning_rate": 2.032574140574548e-06,
+      "loss": 0.338,
+      "step": 16335
+    },
+    {
+      "epoch": 2.1265130808278014,
+      "grad_norm": 3.1886136531829834,
+      "learning_rate": 2.0308923995820077e-06,
+      "loss": 0.3348,
+      "step": 16338
+    },
+    {
+      "epoch": 2.1269035532994924,
+      "grad_norm": 2.831287145614624,
+      "learning_rate": 2.0292111772863193e-06,
+      "loss": 0.419,
+      "step": 16341
+    },
+    {
+      "epoch": 2.127294025771183,
+      "grad_norm": 2.922455310821533,
+      "learning_rate": 2.0275304739811864e-06,
+      "loss": 0.3707,
+      "step": 16344
+    },
+    {
+      "epoch": 2.127684498242874,
+      "grad_norm": 2.8592848777770996,
+      "learning_rate": 2.0258502899602266e-06,
+      "loss": 0.3573,
+      "step": 16347
+    },
+    {
+      "epoch": 2.1280749707145645,
+      "grad_norm": 3.3974833488464355,
+      "learning_rate": 2.0241706255169663e-06,
+      "loss": 0.3741,
+      "step": 16350
+    },
+    {
+      "epoch": 2.1284654431862555,
+      "grad_norm": 2.918564558029175,
+      "learning_rate": 2.0224914809448374e-06,
+      "loss": 0.3557,
+      "step": 16353
+    },
+    {
+      "epoch": 2.128855915657946,
+      "grad_norm": 2.8875043392181396,
+      "learning_rate": 2.0208128565371813e-06,
+      "loss": 0.4248,
+      "step": 16356
+    },
+    {
+      "epoch": 2.129246388129637,
+      "grad_norm": 2.7835309505462646,
+      "learning_rate": 2.019134752587254e-06,
+      "loss": 0.4035,
+      "step": 16359
+    },
+    {
+      "epoch": 2.1296368606013276,
+      "grad_norm": 2.815490245819092,
+      "learning_rate": 2.017457169388214e-06,
+      "loss": 0.3947,
+      "step": 16362
+    },
+    {
+      "epoch": 2.130027333073018,
+      "grad_norm": 3.1463191509246826,
+      "learning_rate": 2.0157801072331325e-06,
+      "loss": 0.3536,
+      "step": 16365
+    },
+    {
+      "epoch": 2.130417805544709,
+      "grad_norm": 2.900385618209839,
+      "learning_rate": 2.0141035664149868e-06,
+      "loss": 0.3911,
+      "step": 16368
+    },
+    {
+      "epoch": 2.1308082780163997,
+      "grad_norm": 3.139719009399414,
+      "learning_rate": 2.0124275472266678e-06,
+      "loss": 0.3523,
+      "step": 16371
+    },
+    {
+      "epoch": 2.1311987504880907,
+      "grad_norm": 2.8128786087036133,
+      "learning_rate": 2.010752049960972e-06,
+      "loss": 0.4084,
+      "step": 16374
+    },
+    {
+      "epoch": 2.1315892229597813,
+      "grad_norm": 2.783865213394165,
+      "learning_rate": 2.0090770749106024e-06,
+      "loss": 0.3048,
+      "step": 16377
+    },
+    {
+      "epoch": 2.1319796954314723,
+      "grad_norm": 2.7417304515838623,
+      "learning_rate": 2.007402622368178e-06,
+      "loss": 0.3362,
+      "step": 16380
+    },
+    {
+      "epoch": 2.132370167903163,
+      "grad_norm": 2.974012613296509,
+      "learning_rate": 2.00572869262622e-06,
+      "loss": 0.3618,
+      "step": 16383
+    },
+    {
+      "epoch": 2.132760640374854,
+      "grad_norm": 2.9564473628997803,
+      "learning_rate": 2.004055285977158e-06,
+      "loss": 0.3369,
+      "step": 16386
+    },
+    {
+      "epoch": 2.1331511128465444,
+      "grad_norm": 2.8104441165924072,
+      "learning_rate": 2.0023824027133356e-06,
+      "loss": 0.3538,
+      "step": 16389
+    },
+    {
+      "epoch": 2.133541585318235,
+      "grad_norm": 3.0070455074310303,
+      "learning_rate": 2.0007100431270027e-06,
+      "loss": 0.3255,
+      "step": 16392
+    },
+    {
+      "epoch": 2.133932057789926,
+      "grad_norm": 2.8861727714538574,
+      "learning_rate": 1.999038207510316e-06,
+      "loss": 0.3306,
+      "step": 16395
+    },
+    {
+      "epoch": 2.1343225302616164,
+      "grad_norm": 2.9251303672790527,
+      "learning_rate": 1.9973668961553394e-06,
+      "loss": 0.3543,
+      "step": 16398
+    },
+    {
+      "epoch": 2.1347130027333074,
+      "grad_norm": 2.2964024543762207,
+      "learning_rate": 1.9956961093540513e-06,
+      "loss": 0.3386,
+      "step": 16401
+    },
+    {
+      "epoch": 2.135103475204998,
+      "grad_norm": 2.716637134552002,
+      "learning_rate": 1.9940258473983326e-06,
+      "loss": 0.3383,
+      "step": 16404
+    },
+    {
+      "epoch": 2.135493947676689,
+      "grad_norm": 2.994713068008423,
+      "learning_rate": 1.992356110579975e-06,
+      "loss": 0.3323,
+      "step": 16407
+    },
+    {
+      "epoch": 2.1358844201483795,
+      "grad_norm": 2.87947940826416,
+      "learning_rate": 1.9906868991906754e-06,
+      "loss": 0.4167,
+      "step": 16410
+    },
+    {
+      "epoch": 2.13627489262007,
+      "grad_norm": 2.95758318901062,
+      "learning_rate": 1.989018213522046e-06,
+      "loss": 0.3449,
+      "step": 16413
+    },
+    {
+      "epoch": 2.136665365091761,
+      "grad_norm": 2.653963565826416,
+      "learning_rate": 1.9873500538656005e-06,
+      "loss": 0.3985,
+      "step": 16416
+    },
+    {
+      "epoch": 2.1370558375634516,
+      "grad_norm": 2.6448793411254883,
+      "learning_rate": 1.985682420512761e-06,
+      "loss": 0.3981,
+      "step": 16419
+    },
+    {
+      "epoch": 2.1374463100351426,
+      "grad_norm": 2.5381240844726562,
+      "learning_rate": 1.9840153137548634e-06,
+      "loss": 0.2987,
+      "step": 16422
+    },
+    {
+      "epoch": 2.137836782506833,
+      "grad_norm": 2.6718051433563232,
+      "learning_rate": 1.982348733883146e-06,
+      "loss": 0.3729,
+      "step": 16425
+    },
+    {
+      "epoch": 2.138227254978524,
+      "grad_norm": 3.0153956413269043,
+      "learning_rate": 1.980682681188754e-06,
+      "loss": 0.416,
+      "step": 16428
+    },
+    {
+      "epoch": 2.1386177274502147,
+      "grad_norm": 3.0083117485046387,
+      "learning_rate": 1.979017155962747e-06,
+      "loss": 0.382,
+      "step": 16431
+    },
+    {
+      "epoch": 2.1390081999219053,
+      "grad_norm": 2.849533796310425,
+      "learning_rate": 1.9773521584960888e-06,
+      "loss": 0.3712,
+      "step": 16434
+    },
+    {
+      "epoch": 2.1393986723935963,
+      "grad_norm": 2.806645154953003,
+      "learning_rate": 1.9756876890796496e-06,
+      "loss": 0.3822,
+      "step": 16437
+    },
+    {
+      "epoch": 2.139789144865287,
+      "grad_norm": 2.8463988304138184,
+      "learning_rate": 1.9740237480042075e-06,
+      "loss": 0.357,
+      "step": 16440
+    },
+    {
+      "epoch": 2.140179617336978,
+      "grad_norm": 3.0570833683013916,
+      "learning_rate": 1.9723603355604526e-06,
+      "loss": 0.3629,
+      "step": 16443
+    },
+    {
+      "epoch": 2.1405700898086684,
+      "grad_norm": 2.503617286682129,
+      "learning_rate": 1.9706974520389776e-06,
+      "loss": 0.3127,
+      "step": 16446
+    },
+    {
+      "epoch": 2.1409605622803594,
+      "grad_norm": 3.2526721954345703,
+      "learning_rate": 1.9690350977302837e-06,
+      "loss": 0.3756,
+      "step": 16449
+    },
+    {
+      "epoch": 2.14135103475205,
+      "grad_norm": 2.4304678440093994,
+      "learning_rate": 1.967373272924783e-06,
+      "loss": 0.2854,
+      "step": 16452
+    },
+    {
+      "epoch": 2.141741507223741,
+      "grad_norm": 2.5894525051116943,
+      "learning_rate": 1.9657119779127926e-06,
+      "loss": 0.3817,
+      "step": 16455
+    },
+    {
+      "epoch": 2.1421319796954315,
+      "grad_norm": 2.8062849044799805,
+      "learning_rate": 1.9640512129845365e-06,
+      "loss": 0.3606,
+      "step": 16458
+    },
+    {
+      "epoch": 2.142522452167122,
+      "grad_norm": 2.921401262283325,
+      "learning_rate": 1.9623909784301442e-06,
+      "loss": 0.4018,
+      "step": 16461
+    },
+    {
+      "epoch": 2.142912924638813,
+      "grad_norm": 2.643878698348999,
+      "learning_rate": 1.9607312745396602e-06,
+      "loss": 0.3517,
+      "step": 16464
+    },
+    {
+      "epoch": 2.1433033971105035,
+      "grad_norm": 2.543466806411743,
+      "learning_rate": 1.9590721016030285e-06,
+      "loss": 0.3569,
+      "step": 16467
+    },
+    {
+      "epoch": 2.1436938695821945,
+      "grad_norm": 2.8661959171295166,
+      "learning_rate": 1.957413459910102e-06,
+      "loss": 0.3746,
+      "step": 16470
+    },
+    {
+      "epoch": 2.144084342053885,
+      "grad_norm": 2.730772018432617,
+      "learning_rate": 1.9557553497506432e-06,
+      "loss": 0.2907,
+      "step": 16473
+    },
+    {
+      "epoch": 2.144474814525576,
+      "grad_norm": 2.7955386638641357,
+      "learning_rate": 1.954097771414322e-06,
+      "loss": 0.3574,
+      "step": 16476
+    },
+    {
+      "epoch": 2.1448652869972666,
+      "grad_norm": 2.8942129611968994,
+      "learning_rate": 1.952440725190713e-06,
+      "loss": 0.3566,
+      "step": 16479
+    },
+    {
+      "epoch": 2.1452557594689576,
+      "grad_norm": 2.645179510116577,
+      "learning_rate": 1.9507842113692967e-06,
+      "loss": 0.3757,
+      "step": 16482
+    },
+    {
+      "epoch": 2.145646231940648,
+      "grad_norm": 2.6048803329467773,
+      "learning_rate": 1.9491282302394653e-06,
+      "loss": 0.3349,
+      "step": 16485
+    },
+    {
+      "epoch": 2.1460367044123387,
+      "grad_norm": 3.326573371887207,
+      "learning_rate": 1.947472782090514e-06,
+      "loss": 0.3798,
+      "step": 16488
+    },
+    {
+      "epoch": 2.1464271768840297,
+      "grad_norm": 2.8624017238616943,
+      "learning_rate": 1.9458178672116445e-06,
+      "loss": 0.3991,
+      "step": 16491
+    },
+    {
+      "epoch": 2.1468176493557203,
+      "grad_norm": 2.8629748821258545,
+      "learning_rate": 1.9441634858919705e-06,
+      "loss": 0.3984,
+      "step": 16494
+    },
+    {
+      "epoch": 2.1472081218274113,
+      "grad_norm": 2.9364376068115234,
+      "learning_rate": 1.9425096384205066e-06,
+      "loss": 0.3873,
+      "step": 16497
+    },
+    {
+      "epoch": 2.147598594299102,
+      "grad_norm": 2.7493503093719482,
+      "learning_rate": 1.9408563250861756e-06,
+      "loss": 0.3023,
+      "step": 16500
+    },
+    {
+      "epoch": 2.147989066770793,
+      "grad_norm": 2.7298777103424072,
+      "learning_rate": 1.9392035461778104e-06,
+      "loss": 0.3675,
+      "step": 16503
+    },
+    {
+      "epoch": 2.1483795392424834,
+      "grad_norm": 3.0014255046844482,
+      "learning_rate": 1.937551301984147e-06,
+      "loss": 0.3866,
+      "step": 16506
+    },
+    {
+      "epoch": 2.1487700117141744,
+      "grad_norm": 2.657407522201538,
+      "learning_rate": 1.9358995927938284e-06,
+      "loss": 0.4154,
+      "step": 16509
+    },
+    {
+      "epoch": 2.149160484185865,
+      "grad_norm": 2.729097843170166,
+      "learning_rate": 1.934248418895401e-06,
+      "loss": 0.3707,
+      "step": 16512
+    },
+    {
+      "epoch": 2.1495509566575555,
+      "grad_norm": 3.432372808456421,
+      "learning_rate": 1.93259778057733e-06,
+      "loss": 0.3662,
+      "step": 16515
+    },
+    {
+      "epoch": 2.1499414291292465,
+      "grad_norm": 2.9015185832977295,
+      "learning_rate": 1.9309476781279735e-06,
+      "loss": 0.3547,
+      "step": 16518
+    },
+    {
+      "epoch": 2.150331901600937,
+      "grad_norm": 2.588021993637085,
+      "learning_rate": 1.9292981118356013e-06,
+      "loss": 0.3424,
+      "step": 16521
+    },
+    {
+      "epoch": 2.150722374072628,
+      "grad_norm": 2.5547730922698975,
+      "learning_rate": 1.927649081988387e-06,
+      "loss": 0.357,
+      "step": 16524
+    },
+    {
+      "epoch": 2.1511128465443186,
+      "grad_norm": 2.5534119606018066,
+      "learning_rate": 1.926000588874417e-06,
+      "loss": 0.3456,
+      "step": 16527
+    },
+    {
+      "epoch": 2.1515033190160096,
+      "grad_norm": 3.0660245418548584,
+      "learning_rate": 1.924352632781677e-06,
+      "loss": 0.3396,
+      "step": 16530
+    },
+    {
+      "epoch": 2.1518937914877,
+      "grad_norm": 2.4878122806549072,
+      "learning_rate": 1.9227052139980606e-06,
+      "loss": 0.3132,
+      "step": 16533
+    },
+    {
+      "epoch": 2.152284263959391,
+      "grad_norm": 2.8623673915863037,
+      "learning_rate": 1.921058332811371e-06,
+      "loss": 0.3601,
+      "step": 16536
+    },
+    {
+      "epoch": 2.1526747364310816,
+      "grad_norm": 2.5879554748535156,
+      "learning_rate": 1.9194119895093137e-06,
+      "loss": 0.3428,
+      "step": 16539
+    },
+    {
+      "epoch": 2.153065208902772,
+      "grad_norm": 2.7629878520965576,
+      "learning_rate": 1.9177661843794994e-06,
+      "loss": 0.367,
+      "step": 16542
+    },
+    {
+      "epoch": 2.153455681374463,
+      "grad_norm": 2.715153694152832,
+      "learning_rate": 1.9161209177094504e-06,
+      "loss": 0.3238,
+      "step": 16545
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 3.1860060691833496,
+      "learning_rate": 1.9144761897865895e-06,
+      "loss": 0.3534,
+      "step": 16548
+    },
+    {
+      "epoch": 2.1542366263178447,
+      "grad_norm": 2.8940200805664062,
+      "learning_rate": 1.9128320008982456e-06,
+      "loss": 0.3401,
+      "step": 16551
+    },
+    {
+      "epoch": 2.1546270987895353,
+      "grad_norm": 2.862811326980591,
+      "learning_rate": 1.9111883513316564e-06,
+      "loss": 0.3994,
+      "step": 16554
+    },
+    {
+      "epoch": 2.1550175712612263,
+      "grad_norm": 2.693480968475342,
+      "learning_rate": 1.909545241373966e-06,
+      "loss": 0.3594,
+      "step": 16557
+    },
+    {
+      "epoch": 2.155408043732917,
+      "grad_norm": 2.9128196239471436,
+      "learning_rate": 1.9079026713122206e-06,
+      "loss": 0.3923,
+      "step": 16560
+    },
+    {
+      "epoch": 2.1557985162046074,
+      "grad_norm": 2.5476996898651123,
+      "learning_rate": 1.9062606414333723e-06,
+      "loss": 0.3357,
+      "step": 16563
+    },
+    {
+      "epoch": 2.1561889886762984,
+      "grad_norm": 2.8649590015411377,
+      "learning_rate": 1.9046191520242835e-06,
+      "loss": 0.3318,
+      "step": 16566
+    },
+    {
+      "epoch": 2.156579461147989,
+      "grad_norm": 3.019973039627075,
+      "learning_rate": 1.902978203371717e-06,
+      "loss": 0.3477,
+      "step": 16569
+    },
+    {
+      "epoch": 2.15696993361968,
+      "grad_norm": 2.824190855026245,
+      "learning_rate": 1.901337795762343e-06,
+      "loss": 0.3407,
+      "step": 16572
+    },
+    {
+      "epoch": 2.1573604060913705,
+      "grad_norm": 2.4870200157165527,
+      "learning_rate": 1.8996979294827355e-06,
+      "loss": 0.3377,
+      "step": 16575
+    },
+    {
+      "epoch": 2.1577508785630615,
+      "grad_norm": 2.9998185634613037,
+      "learning_rate": 1.89805860481938e-06,
+      "loss": 0.3986,
+      "step": 16578
+    },
+    {
+      "epoch": 2.158141351034752,
+      "grad_norm": 2.575596809387207,
+      "learning_rate": 1.8964198220586599e-06,
+      "loss": 0.3309,
+      "step": 16581
+    },
+    {
+      "epoch": 2.1585318235064426,
+      "grad_norm": 3.0732362270355225,
+      "learning_rate": 1.894781581486867e-06,
+      "loss": 0.3105,
+      "step": 16584
+    },
+    {
+      "epoch": 2.1589222959781336,
+      "grad_norm": 2.591250419616699,
+      "learning_rate": 1.8931438833902005e-06,
+      "loss": 0.3741,
+      "step": 16587
+    },
+    {
+      "epoch": 2.159312768449824,
+      "grad_norm": 2.6407053470611572,
+      "learning_rate": 1.8915067280547622e-06,
+      "loss": 0.3275,
+      "step": 16590
+    },
+    {
+      "epoch": 2.159703240921515,
+      "grad_norm": 3.031137704849243,
+      "learning_rate": 1.8898701157665572e-06,
+      "loss": 0.36,
+      "step": 16593
+    },
+    {
+      "epoch": 2.1600937133932057,
+      "grad_norm": 2.810857057571411,
+      "learning_rate": 1.8882340468115002e-06,
+      "loss": 0.3636,
+      "step": 16596
+    },
+    {
+      "epoch": 2.1604841858648967,
+      "grad_norm": 3.037799119949341,
+      "learning_rate": 1.8865985214754107e-06,
+      "loss": 0.3521,
+      "step": 16599
+    },
+    {
+      "epoch": 2.160874658336587,
+      "grad_norm": 3.4607412815093994,
+      "learning_rate": 1.8849635400440098e-06,
+      "loss": 0.3723,
+      "step": 16602
+    },
+    {
+      "epoch": 2.161265130808278,
+      "grad_norm": 2.7808799743652344,
+      "learning_rate": 1.8833291028029239e-06,
+      "loss": 0.3518,
+      "step": 16605
+    },
+    {
+      "epoch": 2.1616556032799688,
+      "grad_norm": 2.7576420307159424,
+      "learning_rate": 1.881695210037689e-06,
+      "loss": 0.3491,
+      "step": 16608
+    },
+    {
+      "epoch": 2.1620460757516593,
+      "grad_norm": 2.637984037399292,
+      "learning_rate": 1.8800618620337407e-06,
+      "loss": 0.3452,
+      "step": 16611
+    },
+    {
+      "epoch": 2.1624365482233503,
+      "grad_norm": 2.8887691497802734,
+      "learning_rate": 1.8784290590764199e-06,
+      "loss": 0.3166,
+      "step": 16614
+    },
+    {
+      "epoch": 2.162827020695041,
+      "grad_norm": 2.620800495147705,
+      "learning_rate": 1.8767968014509774e-06,
+      "loss": 0.3097,
+      "step": 16617
+    },
+    {
+      "epoch": 2.163217493166732,
+      "grad_norm": 2.542872190475464,
+      "learning_rate": 1.8751650894425639e-06,
+      "loss": 0.326,
+      "step": 16620
+    },
+    {
+      "epoch": 2.1636079656384224,
+      "grad_norm": 3.511993169784546,
+      "learning_rate": 1.8735339233362355e-06,
+      "loss": 0.39,
+      "step": 16623
+    },
+    {
+      "epoch": 2.1639984381101134,
+      "grad_norm": 2.7915451526641846,
+      "learning_rate": 1.8719033034169514e-06,
+      "loss": 0.4285,
+      "step": 16626
+    },
+    {
+      "epoch": 2.164388910581804,
+      "grad_norm": 2.7394583225250244,
+      "learning_rate": 1.8702732299695813e-06,
+      "loss": 0.3401,
+      "step": 16629
+    },
+    {
+      "epoch": 2.164779383053495,
+      "grad_norm": 2.7059719562530518,
+      "learning_rate": 1.8686437032788945e-06,
+      "loss": 0.3685,
+      "step": 16632
+    },
+    {
+      "epoch": 2.1651698555251855,
+      "grad_norm": 2.651163339614868,
+      "learning_rate": 1.8670147236295632e-06,
+      "loss": 0.3437,
+      "step": 16635
+    },
+    {
+      "epoch": 2.165560327996876,
+      "grad_norm": 2.559772253036499,
+      "learning_rate": 1.8653862913061693e-06,
+      "loss": 0.2954,
+      "step": 16638
+    },
+    {
+      "epoch": 2.165950800468567,
+      "grad_norm": 2.9166457653045654,
+      "learning_rate": 1.8637584065931974e-06,
+      "loss": 0.2884,
+      "step": 16641
+    },
+    {
+      "epoch": 2.1663412729402576,
+      "grad_norm": 2.951847791671753,
+      "learning_rate": 1.862131069775034e-06,
+      "loss": 0.4385,
+      "step": 16644
+    },
+    {
+      "epoch": 2.1667317454119486,
+      "grad_norm": 2.7294256687164307,
+      "learning_rate": 1.8605042811359702e-06,
+      "loss": 0.3622,
+      "step": 16647
+    },
+    {
+      "epoch": 2.167122217883639,
+      "grad_norm": 2.931398630142212,
+      "learning_rate": 1.8588780409602053e-06,
+      "loss": 0.3915,
+      "step": 16650
+    },
+    {
+      "epoch": 2.16751269035533,
+      "grad_norm": 3.169494152069092,
+      "learning_rate": 1.8572523495318389e-06,
+      "loss": 0.3643,
+      "step": 16653
+    },
+    {
+      "epoch": 2.1679031628270207,
+      "grad_norm": 3.00234055519104,
+      "learning_rate": 1.855627207134874e-06,
+      "loss": 0.3221,
+      "step": 16656
+    },
+    {
+      "epoch": 2.1682936352987117,
+      "grad_norm": 2.8568825721740723,
+      "learning_rate": 1.854002614053223e-06,
+      "loss": 0.4029,
+      "step": 16659
+    },
+    {
+      "epoch": 2.168684107770402,
+      "grad_norm": 2.6517200469970703,
+      "learning_rate": 1.852378570570697e-06,
+      "loss": 0.3926,
+      "step": 16662
+    },
+    {
+      "epoch": 2.1690745802420928,
+      "grad_norm": 2.9220314025878906,
+      "learning_rate": 1.8507550769710115e-06,
+      "loss": 0.3665,
+      "step": 16665
+    },
+    {
+      "epoch": 2.1694650527137838,
+      "grad_norm": 2.624634265899658,
+      "learning_rate": 1.849132133537791e-06,
+      "loss": 0.3212,
+      "step": 16668
+    },
+    {
+      "epoch": 2.1698555251854743,
+      "grad_norm": 2.7517733573913574,
+      "learning_rate": 1.8475097405545578e-06,
+      "loss": 0.3628,
+      "step": 16671
+    },
+    {
+      "epoch": 2.1702459976571653,
+      "grad_norm": 2.7567718029022217,
+      "learning_rate": 1.8458878983047412e-06,
+      "loss": 0.3733,
+      "step": 16674
+    },
+    {
+      "epoch": 2.170636470128856,
+      "grad_norm": 2.902865171432495,
+      "learning_rate": 1.8442666070716719e-06,
+      "loss": 0.3259,
+      "step": 16677
+    },
+    {
+      "epoch": 2.171026942600547,
+      "grad_norm": 3.230361223220825,
+      "learning_rate": 1.842645867138587e-06,
+      "loss": 0.3797,
+      "step": 16680
+    },
+    {
+      "epoch": 2.1714174150722374,
+      "grad_norm": 3.068331003189087,
+      "learning_rate": 1.8410256787886298e-06,
+      "loss": 0.4157,
+      "step": 16683
+    },
+    {
+      "epoch": 2.171807887543928,
+      "grad_norm": 3.102295160293579,
+      "learning_rate": 1.8394060423048404e-06,
+      "loss": 0.3912,
+      "step": 16686
+    },
+    {
+      "epoch": 2.172198360015619,
+      "grad_norm": 2.4631145000457764,
+      "learning_rate": 1.8377869579701647e-06,
+      "loss": 0.3233,
+      "step": 16689
+    },
+    {
+      "epoch": 2.1725888324873095,
+      "grad_norm": 3.171217679977417,
+      "learning_rate": 1.8361684260674572e-06,
+      "loss": 0.3941,
+      "step": 16692
+    },
+    {
+      "epoch": 2.1729793049590005,
+      "grad_norm": 3.546006679534912,
+      "learning_rate": 1.8345504468794694e-06,
+      "loss": 0.3908,
+      "step": 16695
+    },
+    {
+      "epoch": 2.173369777430691,
+      "grad_norm": 2.745622158050537,
+      "learning_rate": 1.832933020688858e-06,
+      "loss": 0.3229,
+      "step": 16698
+    },
+    {
+      "epoch": 2.173760249902382,
+      "grad_norm": 2.7853684425354004,
+      "learning_rate": 1.8313161477781871e-06,
+      "loss": 0.3425,
+      "step": 16701
+    },
+    {
+      "epoch": 2.1741507223740726,
+      "grad_norm": 2.7178430557250977,
+      "learning_rate": 1.8296998284299195e-06,
+      "loss": 0.4037,
+      "step": 16704
+    },
+    {
+      "epoch": 2.1745411948457636,
+      "grad_norm": 2.7298033237457275,
+      "learning_rate": 1.8280840629264202e-06,
+      "loss": 0.3505,
+      "step": 16707
+    },
+    {
+      "epoch": 2.174931667317454,
+      "grad_norm": 2.750066041946411,
+      "learning_rate": 1.8264688515499645e-06,
+      "loss": 0.3359,
+      "step": 16710
+    },
+    {
+      "epoch": 2.1753221397891447,
+      "grad_norm": 3.090998888015747,
+      "learning_rate": 1.824854194582724e-06,
+      "loss": 0.3278,
+      "step": 16713
+    },
+    {
+      "epoch": 2.1757126122608357,
+      "grad_norm": 2.6128833293914795,
+      "learning_rate": 1.823240092306775e-06,
+      "loss": 0.3463,
+      "step": 16716
+    },
+    {
+      "epoch": 2.1761030847325262,
+      "grad_norm": 2.647390604019165,
+      "learning_rate": 1.8216265450041004e-06,
+      "loss": 0.3686,
+      "step": 16719
+    },
+    {
+      "epoch": 2.176493557204217,
+      "grad_norm": 2.6248843669891357,
+      "learning_rate": 1.8200135529565805e-06,
+      "loss": 0.3933,
+      "step": 16722
+    },
+    {
+      "epoch": 2.1768840296759078,
+      "grad_norm": 2.8739254474639893,
+      "learning_rate": 1.8184011164460046e-06,
+      "loss": 0.394,
+      "step": 16725
+    },
+    {
+      "epoch": 2.1772745021475988,
+      "grad_norm": 2.8517467975616455,
+      "learning_rate": 1.816789235754059e-06,
+      "loss": 0.3486,
+      "step": 16728
+    },
+    {
+      "epoch": 2.1776649746192893,
+      "grad_norm": 2.5461232662200928,
+      "learning_rate": 1.8151779111623392e-06,
+      "loss": 0.3518,
+      "step": 16731
+    },
+    {
+      "epoch": 2.17805544709098,
+      "grad_norm": 2.520355463027954,
+      "learning_rate": 1.8135671429523377e-06,
+      "loss": 0.3742,
+      "step": 16734
+    },
+    {
+      "epoch": 2.178445919562671,
+      "grad_norm": 2.9696972370147705,
+      "learning_rate": 1.811956931405454e-06,
+      "loss": 0.3798,
+      "step": 16737
+    },
+    {
+      "epoch": 2.1788363920343614,
+      "grad_norm": 2.75524640083313,
+      "learning_rate": 1.8103472768029856e-06,
+      "loss": 0.329,
+      "step": 16740
+    },
+    {
+      "epoch": 2.1792268645060524,
+      "grad_norm": 2.857969284057617,
+      "learning_rate": 1.8087381794261394e-06,
+      "loss": 0.3514,
+      "step": 16743
+    },
+    {
+      "epoch": 2.179617336977743,
+      "grad_norm": 3.1860451698303223,
+      "learning_rate": 1.80712963955602e-06,
+      "loss": 0.3085,
+      "step": 16746
+    },
+    {
+      "epoch": 2.180007809449434,
+      "grad_norm": 3.36942458152771,
+      "learning_rate": 1.8055216574736346e-06,
+      "loss": 0.3323,
+      "step": 16749
+    },
+    {
+      "epoch": 2.1803982819211245,
+      "grad_norm": 3.2723255157470703,
+      "learning_rate": 1.8039142334598964e-06,
+      "loss": 0.4076,
+      "step": 16752
+    },
+    {
+      "epoch": 2.1807887543928155,
+      "grad_norm": 2.8995213508605957,
+      "learning_rate": 1.8023073677956183e-06,
+      "loss": 0.3193,
+      "step": 16755
+    },
+    {
+      "epoch": 2.181179226864506,
+      "grad_norm": 3.000784397125244,
+      "learning_rate": 1.8007010607615144e-06,
+      "loss": 0.4389,
+      "step": 16758
+    },
+    {
+      "epoch": 2.1815696993361966,
+      "grad_norm": 2.861037492752075,
+      "learning_rate": 1.7990953126382065e-06,
+      "loss": 0.3881,
+      "step": 16761
+    },
+    {
+      "epoch": 2.1819601718078876,
+      "grad_norm": 2.6603360176086426,
+      "learning_rate": 1.797490123706212e-06,
+      "loss": 0.3744,
+      "step": 16764
+    },
+    {
+      "epoch": 2.182350644279578,
+      "grad_norm": 2.7340359687805176,
+      "learning_rate": 1.795885494245958e-06,
+      "loss": 0.3644,
+      "step": 16767
+    },
+    {
+      "epoch": 2.182741116751269,
+      "grad_norm": 2.9198856353759766,
+      "learning_rate": 1.7942814245377654e-06,
+      "loss": 0.3499,
+      "step": 16770
+    },
+    {
+      "epoch": 2.1831315892229597,
+      "grad_norm": 3.0503599643707275,
+      "learning_rate": 1.7926779148618661e-06,
+      "loss": 0.3224,
+      "step": 16773
+    },
+    {
+      "epoch": 2.1835220616946507,
+      "grad_norm": 2.8313817977905273,
+      "learning_rate": 1.7910749654983879e-06,
+      "loss": 0.3579,
+      "step": 16776
+    },
+    {
+      "epoch": 2.1839125341663412,
+      "grad_norm": 2.788543224334717,
+      "learning_rate": 1.7894725767273601e-06,
+      "loss": 0.3443,
+      "step": 16779
+    },
+    {
+      "epoch": 2.1843030066380322,
+      "grad_norm": 2.825108289718628,
+      "learning_rate": 1.7878707488287216e-06,
+      "loss": 0.3615,
+      "step": 16782
+    },
+    {
+      "epoch": 2.184693479109723,
+      "grad_norm": 2.4772067070007324,
+      "learning_rate": 1.7862694820823062e-06,
+      "loss": 0.3303,
+      "step": 16785
+    },
+    {
+      "epoch": 2.1850839515814133,
+      "grad_norm": 2.8775320053100586,
+      "learning_rate": 1.784668776767851e-06,
+      "loss": 0.3626,
+      "step": 16788
+    },
+    {
+      "epoch": 2.1854744240531043,
+      "grad_norm": 2.8523988723754883,
+      "learning_rate": 1.783068633164995e-06,
+      "loss": 0.2922,
+      "step": 16791
+    },
+    {
+      "epoch": 2.185864896524795,
+      "grad_norm": 2.919550657272339,
+      "learning_rate": 1.7814690515532828e-06,
+      "loss": 0.3725,
+      "step": 16794
+    },
+    {
+      "epoch": 2.186255368996486,
+      "grad_norm": 2.9476890563964844,
+      "learning_rate": 1.779870032212157e-06,
+      "loss": 0.3546,
+      "step": 16797
+    },
+    {
+      "epoch": 2.1866458414681764,
+      "grad_norm": 2.7363481521606445,
+      "learning_rate": 1.7782715754209607e-06,
+      "loss": 0.4004,
+      "step": 16800
+    },
+    {
+      "epoch": 2.1870363139398674,
+      "grad_norm": 2.9156899452209473,
+      "learning_rate": 1.776673681458944e-06,
+      "loss": 0.3733,
+      "step": 16803
+    },
+    {
+      "epoch": 2.187426786411558,
+      "grad_norm": 2.7484629154205322,
+      "learning_rate": 1.7750763506052526e-06,
+      "loss": 0.3773,
+      "step": 16806
+    },
+    {
+      "epoch": 2.187817258883249,
+      "grad_norm": 2.7555909156799316,
+      "learning_rate": 1.77347958313894e-06,
+      "loss": 0.3394,
+      "step": 16809
+    },
+    {
+      "epoch": 2.1882077313549395,
+      "grad_norm": 2.8585703372955322,
+      "learning_rate": 1.7718833793389556e-06,
+      "loss": 0.349,
+      "step": 16812
+    },
+    {
+      "epoch": 2.18859820382663,
+      "grad_norm": 2.7602007389068604,
+      "learning_rate": 1.770287739484155e-06,
+      "loss": 0.3731,
+      "step": 16815
+    },
+    {
+      "epoch": 2.188988676298321,
+      "grad_norm": 2.4553885459899902,
+      "learning_rate": 1.768692663853292e-06,
+      "loss": 0.3372,
+      "step": 16818
+    },
+    {
+      "epoch": 2.1893791487700116,
+      "grad_norm": 2.777724504470825,
+      "learning_rate": 1.7670981527250213e-06,
+      "loss": 0.4093,
+      "step": 16821
+    },
+    {
+      "epoch": 2.1897696212417026,
+      "grad_norm": 2.8050687313079834,
+      "learning_rate": 1.7655042063779043e-06,
+      "loss": 0.3752,
+      "step": 16824
+    },
+    {
+      "epoch": 2.190160093713393,
+      "grad_norm": 3.285545587539673,
+      "learning_rate": 1.7639108250903974e-06,
+      "loss": 0.4672,
+      "step": 16827
+    },
+    {
+      "epoch": 2.190550566185084,
+      "grad_norm": 3.2588181495666504,
+      "learning_rate": 1.762318009140862e-06,
+      "loss": 0.4154,
+      "step": 16830
+    },
+    {
+      "epoch": 2.1909410386567747,
+      "grad_norm": 2.734658718109131,
+      "learning_rate": 1.7607257588075582e-06,
+      "loss": 0.4109,
+      "step": 16833
+    },
+    {
+      "epoch": 2.1913315111284652,
+      "grad_norm": 2.7743442058563232,
+      "learning_rate": 1.7591340743686507e-06,
+      "loss": 0.3764,
+      "step": 16836
+    },
+    {
+      "epoch": 2.1917219836001562,
+      "grad_norm": 2.652029514312744,
+      "learning_rate": 1.7575429561022029e-06,
+      "loss": 0.4011,
+      "step": 16839
+    },
+    {
+      "epoch": 2.192112456071847,
+      "grad_norm": 2.7200498580932617,
+      "learning_rate": 1.755952404286178e-06,
+      "loss": 0.3842,
+      "step": 16842
+    },
+    {
+      "epoch": 2.192502928543538,
+      "grad_norm": 2.506895065307617,
+      "learning_rate": 1.7543624191984455e-06,
+      "loss": 0.3566,
+      "step": 16845
+    },
+    {
+      "epoch": 2.1928934010152283,
+      "grad_norm": 3.026963233947754,
+      "learning_rate": 1.7527730011167681e-06,
+      "loss": 0.3718,
+      "step": 16848
+    },
+    {
+      "epoch": 2.1932838734869193,
+      "grad_norm": 2.6555018424987793,
+      "learning_rate": 1.7511841503188187e-06,
+      "loss": 0.3382,
+      "step": 16851
+    },
+    {
+      "epoch": 2.19367434595861,
+      "grad_norm": 2.6906399726867676,
+      "learning_rate": 1.7495958670821617e-06,
+      "loss": 0.3304,
+      "step": 16854
+    },
+    {
+      "epoch": 2.1940648184303004,
+      "grad_norm": 3.381065607070923,
+      "learning_rate": 1.7480081516842705e-06,
+      "loss": 0.3892,
+      "step": 16857
+    },
+    {
+      "epoch": 2.1944552909019914,
+      "grad_norm": 2.733503580093384,
+      "learning_rate": 1.7464210044025144e-06,
+      "loss": 0.3378,
+      "step": 16860
+    },
+    {
+      "epoch": 2.194845763373682,
+      "grad_norm": 2.755075693130493,
+      "learning_rate": 1.744834425514162e-06,
+      "loss": 0.3439,
+      "step": 16863
+    },
+    {
+      "epoch": 2.195236235845373,
+      "grad_norm": 2.853973388671875,
+      "learning_rate": 1.7432484152963896e-06,
+      "loss": 0.3667,
+      "step": 16866
+    },
+    {
+      "epoch": 2.1956267083170635,
+      "grad_norm": 2.4795007705688477,
+      "learning_rate": 1.7416629740262681e-06,
+      "loss": 0.3341,
+      "step": 16869
+    },
+    {
+      "epoch": 2.1960171807887545,
+      "grad_norm": 2.5564217567443848,
+      "learning_rate": 1.7400781019807684e-06,
+      "loss": 0.3706,
+      "step": 16872
+    },
+    {
+      "epoch": 2.196407653260445,
+      "grad_norm": 2.830836057662964,
+      "learning_rate": 1.738493799436768e-06,
+      "loss": 0.354,
+      "step": 16875
+    },
+    {
+      "epoch": 2.196798125732136,
+      "grad_norm": 2.753403425216675,
+      "learning_rate": 1.7369100666710398e-06,
+      "loss": 0.3636,
+      "step": 16878
+    },
+    {
+      "epoch": 2.1971885982038266,
+      "grad_norm": 2.559079170227051,
+      "learning_rate": 1.7353269039602588e-06,
+      "loss": 0.3772,
+      "step": 16881
+    },
+    {
+      "epoch": 2.197579070675517,
+      "grad_norm": 2.9409003257751465,
+      "learning_rate": 1.7337443115809976e-06,
+      "loss": 0.3509,
+      "step": 16884
+    },
+    {
+      "epoch": 2.197969543147208,
+      "grad_norm": 2.7231342792510986,
+      "learning_rate": 1.7321622898097362e-06,
+      "loss": 0.4125,
+      "step": 16887
+    },
+    {
+      "epoch": 2.1983600156188987,
+      "grad_norm": 2.7584779262542725,
+      "learning_rate": 1.7305808389228462e-06,
+      "loss": 0.3734,
+      "step": 16890
+    },
+    {
+      "epoch": 2.1987504880905897,
+      "grad_norm": 2.9606661796569824,
+      "learning_rate": 1.7289999591966072e-06,
+      "loss": 0.3863,
+      "step": 16893
+    },
+    {
+      "epoch": 2.1991409605622803,
+      "grad_norm": 2.588604688644409,
+      "learning_rate": 1.7274196509071927e-06,
+      "loss": 0.3034,
+      "step": 16896
+    },
+    {
+      "epoch": 2.1995314330339713,
+      "grad_norm": 2.7741434574127197,
+      "learning_rate": 1.7258399143306825e-06,
+      "loss": 0.3722,
+      "step": 16899
+    },
+    {
+      "epoch": 2.199921905505662,
+      "grad_norm": 2.741807699203491,
+      "learning_rate": 1.7242607497430514e-06,
+      "loss": 0.3266,
+      "step": 16902
+    },
+    {
+      "epoch": 2.200312377977353,
+      "grad_norm": 2.6828136444091797,
+      "learning_rate": 1.7226821574201747e-06,
+      "loss": 0.4204,
+      "step": 16905
+    },
+    {
+      "epoch": 2.2007028504490433,
+      "grad_norm": 2.492523431777954,
+      "learning_rate": 1.721104137637832e-06,
+      "loss": 0.3388,
+      "step": 16908
+    },
+    {
+      "epoch": 2.201093322920734,
+      "grad_norm": 2.79616117477417,
+      "learning_rate": 1.7195266906716985e-06,
+      "loss": 0.3348,
+      "step": 16911
+    },
+    {
+      "epoch": 2.201483795392425,
+      "grad_norm": 2.682907819747925,
+      "learning_rate": 1.7179498167973496e-06,
+      "loss": 0.3421,
+      "step": 16914
+    },
+    {
+      "epoch": 2.2018742678641154,
+      "grad_norm": 3.588071823120117,
+      "learning_rate": 1.7163735162902651e-06,
+      "loss": 0.374,
+      "step": 16917
+    },
+    {
+      "epoch": 2.2022647403358064,
+      "grad_norm": 3.160140037536621,
+      "learning_rate": 1.7147977894258193e-06,
+      "loss": 0.4139,
+      "step": 16920
+    },
+    {
+      "epoch": 2.202655212807497,
+      "grad_norm": 2.9492440223693848,
+      "learning_rate": 1.713222636479287e-06,
+      "loss": 0.4267,
+      "step": 16923
+    },
+    {
+      "epoch": 2.203045685279188,
+      "grad_norm": 2.716421365737915,
+      "learning_rate": 1.7116480577258477e-06,
+      "loss": 0.3267,
+      "step": 16926
+    },
+    {
+      "epoch": 2.2034361577508785,
+      "grad_norm": 3.1146531105041504,
+      "learning_rate": 1.7100740534405746e-06,
+      "loss": 0.3581,
+      "step": 16929
+    },
+    {
+      "epoch": 2.2038266302225695,
+      "grad_norm": 3.101052761077881,
+      "learning_rate": 1.708500623898442e-06,
+      "loss": 0.3646,
+      "step": 16932
+    },
+    {
+      "epoch": 2.20421710269426,
+      "grad_norm": 2.7972285747528076,
+      "learning_rate": 1.7069277693743258e-06,
+      "loss": 0.3556,
+      "step": 16935
+    },
+    {
+      "epoch": 2.2046075751659506,
+      "grad_norm": 3.1059136390686035,
+      "learning_rate": 1.705355490143003e-06,
+      "loss": 0.3617,
+      "step": 16938
+    },
+    {
+      "epoch": 2.2049980476376416,
+      "grad_norm": 2.703691244125366,
+      "learning_rate": 1.7037837864791445e-06,
+      "loss": 0.3557,
+      "step": 16941
+    },
+    {
+      "epoch": 2.205388520109332,
+      "grad_norm": 2.6937639713287354,
+      "learning_rate": 1.7022126586573246e-06,
+      "loss": 0.3241,
+      "step": 16944
+    },
+    {
+      "epoch": 2.205778992581023,
+      "grad_norm": 3.200873613357544,
+      "learning_rate": 1.7006421069520141e-06,
+      "loss": 0.3686,
+      "step": 16947
+    },
+    {
+      "epoch": 2.2061694650527137,
+      "grad_norm": 2.610318183898926,
+      "learning_rate": 1.699072131637588e-06,
+      "loss": 0.339,
+      "step": 16950
+    },
+    {
+      "epoch": 2.2065599375244047,
+      "grad_norm": 3.0442748069763184,
+      "learning_rate": 1.6975027329883166e-06,
+      "loss": 0.4444,
+      "step": 16953
+    },
+    {
+      "epoch": 2.2069504099960953,
+      "grad_norm": 2.716014862060547,
+      "learning_rate": 1.6959339112783685e-06,
+      "loss": 0.3636,
+      "step": 16956
+    },
+    {
+      "epoch": 2.207340882467786,
+      "grad_norm": 2.61544132232666,
+      "learning_rate": 1.694365666781817e-06,
+      "loss": 0.3378,
+      "step": 16959
+    },
+    {
+      "epoch": 2.207731354939477,
+      "grad_norm": 2.998279094696045,
+      "learning_rate": 1.6927979997726295e-06,
+      "loss": 0.2909,
+      "step": 16962
+    },
+    {
+      "epoch": 2.2081218274111674,
+      "grad_norm": 2.755953311920166,
+      "learning_rate": 1.6912309105246726e-06,
+      "loss": 0.3346,
+      "step": 16965
+    },
+    {
+      "epoch": 2.2085122998828584,
+      "grad_norm": 2.780045986175537,
+      "learning_rate": 1.6896643993117168e-06,
+      "loss": 0.3338,
+      "step": 16968
+    },
+    {
+      "epoch": 2.208902772354549,
+      "grad_norm": 2.9842076301574707,
+      "learning_rate": 1.6880984664074262e-06,
+      "loss": 0.3819,
+      "step": 16971
+    },
+    {
+      "epoch": 2.20929324482624,
+      "grad_norm": 2.786750078201294,
+      "learning_rate": 1.6865331120853645e-06,
+      "loss": 0.3506,
+      "step": 16974
+    },
+    {
+      "epoch": 2.2096837172979304,
+      "grad_norm": 2.8136911392211914,
+      "learning_rate": 1.6849683366189978e-06,
+      "loss": 0.3695,
+      "step": 16977
+    },
+    {
+      "epoch": 2.2100741897696214,
+      "grad_norm": 2.6133673191070557,
+      "learning_rate": 1.6834041402816908e-06,
+      "loss": 0.3035,
+      "step": 16980
+    },
+    {
+      "epoch": 2.210464662241312,
+      "grad_norm": 2.6975417137145996,
+      "learning_rate": 1.6818405233467034e-06,
+      "loss": 0.314,
+      "step": 16983
+    },
+    {
+      "epoch": 2.2108551347130025,
+      "grad_norm": 2.7998573780059814,
+      "learning_rate": 1.6802774860871939e-06,
+      "loss": 0.3707,
+      "step": 16986
+    },
+    {
+      "epoch": 2.2112456071846935,
+      "grad_norm": 2.9110021591186523,
+      "learning_rate": 1.678715028776226e-06,
+      "loss": 0.4747,
+      "step": 16989
+    },
+    {
+      "epoch": 2.211636079656384,
+      "grad_norm": 2.649843215942383,
+      "learning_rate": 1.6771531516867557e-06,
+      "loss": 0.3751,
+      "step": 16992
+    },
+    {
+      "epoch": 2.212026552128075,
+      "grad_norm": 3.525862693786621,
+      "learning_rate": 1.6755918550916395e-06,
+      "loss": 0.3078,
+      "step": 16995
+    },
+    {
+      "epoch": 2.2124170245997656,
+      "grad_norm": 2.670433759689331,
+      "learning_rate": 1.6740311392636311e-06,
+      "loss": 0.3935,
+      "step": 16998
+    },
+    {
+      "epoch": 2.2128074970714566,
+      "grad_norm": 2.793560266494751,
+      "learning_rate": 1.6724710044753872e-06,
+      "loss": 0.3365,
+      "step": 17001
+    },
+    {
+      "epoch": 2.213197969543147,
+      "grad_norm": 3.4935295581817627,
+      "learning_rate": 1.6709114509994588e-06,
+      "loss": 0.3708,
+      "step": 17004
+    },
+    {
+      "epoch": 2.2135884420148377,
+      "grad_norm": 2.874403238296509,
+      "learning_rate": 1.6693524791082948e-06,
+      "loss": 0.3428,
+      "step": 17007
+    },
+    {
+      "epoch": 2.2139789144865287,
+      "grad_norm": 2.7574665546417236,
+      "learning_rate": 1.6677940890742484e-06,
+      "loss": 0.3819,
+      "step": 17010
+    },
+    {
+      "epoch": 2.2143693869582193,
+      "grad_norm": 2.5127577781677246,
+      "learning_rate": 1.6662362811695637e-06,
+      "loss": 0.3112,
+      "step": 17013
+    },
+    {
+      "epoch": 2.2147598594299103,
+      "grad_norm": 2.6434106826782227,
+      "learning_rate": 1.6646790556663867e-06,
+      "loss": 0.3468,
+      "step": 17016
+    },
+    {
+      "epoch": 2.215150331901601,
+      "grad_norm": 2.921630859375,
+      "learning_rate": 1.6631224128367612e-06,
+      "loss": 0.3818,
+      "step": 17019
+    },
+    {
+      "epoch": 2.215540804373292,
+      "grad_norm": 3.2600367069244385,
+      "learning_rate": 1.6615663529526328e-06,
+      "loss": 0.407,
+      "step": 17022
+    },
+    {
+      "epoch": 2.2159312768449824,
+      "grad_norm": 2.5993316173553467,
+      "learning_rate": 1.6600108762858392e-06,
+      "loss": 0.3508,
+      "step": 17025
+    },
+    {
+      "epoch": 2.2163217493166734,
+      "grad_norm": 3.402862310409546,
+      "learning_rate": 1.6584559831081176e-06,
+      "loss": 0.3205,
+      "step": 17028
+    },
+    {
+      "epoch": 2.216712221788364,
+      "grad_norm": 2.991856813430786,
+      "learning_rate": 1.6569016736911082e-06,
+      "loss": 0.3868,
+      "step": 17031
+    },
+    {
+      "epoch": 2.2171026942600545,
+      "grad_norm": 2.9160122871398926,
+      "learning_rate": 1.6553479483063434e-06,
+      "loss": 0.3622,
+      "step": 17034
+    },
+    {
+      "epoch": 2.2174931667317455,
+      "grad_norm": 2.687974691390991,
+      "learning_rate": 1.653794807225254e-06,
+      "loss": 0.3137,
+      "step": 17037
+    },
+    {
+      "epoch": 2.217883639203436,
+      "grad_norm": 2.7909791469573975,
+      "learning_rate": 1.6522422507191744e-06,
+      "loss": 0.3004,
+      "step": 17040
+    },
+    {
+      "epoch": 2.218274111675127,
+      "grad_norm": 2.7353012561798096,
+      "learning_rate": 1.6506902790593303e-06,
+      "loss": 0.3487,
+      "step": 17043
+    },
+    {
+      "epoch": 2.2186645841468176,
+      "grad_norm": 2.6116976737976074,
+      "learning_rate": 1.649138892516849e-06,
+      "loss": 0.3245,
+      "step": 17046
+    },
+    {
+      "epoch": 2.2190550566185085,
+      "grad_norm": 2.7857844829559326,
+      "learning_rate": 1.6475880913627522e-06,
+      "loss": 0.3987,
+      "step": 17049
+    },
+    {
+      "epoch": 2.219445529090199,
+      "grad_norm": 2.628438711166382,
+      "learning_rate": 1.646037875867965e-06,
+      "loss": 0.3248,
+      "step": 17052
+    },
+    {
+      "epoch": 2.21983600156189,
+      "grad_norm": 2.738862991333008,
+      "learning_rate": 1.6444882463033058e-06,
+      "loss": 0.3295,
+      "step": 17055
+    },
+    {
+      "epoch": 2.2202264740335806,
+      "grad_norm": 2.46667742729187,
+      "learning_rate": 1.6429392029394886e-06,
+      "loss": 0.3429,
+      "step": 17058
+    },
+    {
+      "epoch": 2.220616946505271,
+      "grad_norm": 2.5244455337524414,
+      "learning_rate": 1.6413907460471306e-06,
+      "loss": 0.3478,
+      "step": 17061
+    },
+    {
+      "epoch": 2.221007418976962,
+      "grad_norm": 2.5417511463165283,
+      "learning_rate": 1.6398428758967455e-06,
+      "loss": 0.3285,
+      "step": 17064
+    },
+    {
+      "epoch": 2.2213978914486527,
+      "grad_norm": 2.653949499130249,
+      "learning_rate": 1.6382955927587414e-06,
+      "loss": 0.3263,
+      "step": 17067
+    },
+    {
+      "epoch": 2.2217883639203437,
+      "grad_norm": 2.884857416152954,
+      "learning_rate": 1.6367488969034234e-06,
+      "loss": 0.3631,
+      "step": 17070
+    },
+    {
+      "epoch": 2.2221788363920343,
+      "grad_norm": 2.860215663909912,
+      "learning_rate": 1.6352027886009997e-06,
+      "loss": 0.3883,
+      "step": 17073
+    },
+    {
+      "epoch": 2.2225693088637253,
+      "grad_norm": 2.657209873199463,
+      "learning_rate": 1.6336572681215701e-06,
+      "loss": 0.3437,
+      "step": 17076
+    },
+    {
+      "epoch": 2.222959781335416,
+      "grad_norm": 2.648395538330078,
+      "learning_rate": 1.6321123357351327e-06,
+      "loss": 0.3321,
+      "step": 17079
+    },
+    {
+      "epoch": 2.223350253807107,
+      "grad_norm": 2.7004330158233643,
+      "learning_rate": 1.6305679917115864e-06,
+      "loss": 0.3254,
+      "step": 17082
+    },
+    {
+      "epoch": 2.2237407262787974,
+      "grad_norm": 2.858809232711792,
+      "learning_rate": 1.6290242363207238e-06,
+      "loss": 0.3445,
+      "step": 17085
+    },
+    {
+      "epoch": 2.224131198750488,
+      "grad_norm": 2.9426462650299072,
+      "learning_rate": 1.6274810698322341e-06,
+      "loss": 0.3974,
+      "step": 17088
+    },
+    {
+      "epoch": 2.224521671222179,
+      "grad_norm": 2.777961015701294,
+      "learning_rate": 1.6259384925157084e-06,
+      "loss": 0.4052,
+      "step": 17091
+    },
+    {
+      "epoch": 2.2249121436938695,
+      "grad_norm": 2.6763064861297607,
+      "learning_rate": 1.6243965046406302e-06,
+      "loss": 0.3834,
+      "step": 17094
+    },
+    {
+      "epoch": 2.2253026161655605,
+      "grad_norm": 2.739316463470459,
+      "learning_rate": 1.6228551064763814e-06,
+      "loss": 0.3819,
+      "step": 17097
+    },
+    {
+      "epoch": 2.225693088637251,
+      "grad_norm": 2.7301228046417236,
+      "learning_rate": 1.6213142982922376e-06,
+      "loss": 0.2822,
+      "step": 17100
+    },
+    {
+      "epoch": 2.226083561108942,
+      "grad_norm": 3.0507218837738037,
+      "learning_rate": 1.6197740803573813e-06,
+      "loss": 0.3252,
+      "step": 17103
+    },
+    {
+      "epoch": 2.2264740335806326,
+      "grad_norm": 2.8858330249786377,
+      "learning_rate": 1.6182344529408828e-06,
+      "loss": 0.3756,
+      "step": 17106
+    },
+    {
+      "epoch": 2.226864506052323,
+      "grad_norm": 2.5173003673553467,
+      "learning_rate": 1.6166954163117105e-06,
+      "loss": 0.3249,
+      "step": 17109
+    },
+    {
+      "epoch": 2.227254978524014,
+      "grad_norm": 2.6058590412139893,
+      "learning_rate": 1.6151569707387305e-06,
+      "loss": 0.3337,
+      "step": 17112
+    },
+    {
+      "epoch": 2.2276454509957047,
+      "grad_norm": 3.0819878578186035,
+      "learning_rate": 1.6136191164907084e-06,
+      "loss": 0.4114,
+      "step": 17115
+    },
+    {
+      "epoch": 2.2280359234673957,
+      "grad_norm": 3.0594425201416016,
+      "learning_rate": 1.6120818538363026e-06,
+      "loss": 0.4065,
+      "step": 17118
+    },
+    {
+      "epoch": 2.228426395939086,
+      "grad_norm": 3.019721746444702,
+      "learning_rate": 1.6105451830440683e-06,
+      "loss": 0.333,
+      "step": 17121
+    },
+    {
+      "epoch": 2.228816868410777,
+      "grad_norm": 2.7887985706329346,
+      "learning_rate": 1.6090091043824618e-06,
+      "loss": 0.37,
+      "step": 17124
+    },
+    {
+      "epoch": 2.2292073408824677,
+      "grad_norm": 2.4664344787597656,
+      "learning_rate": 1.6074736181198309e-06,
+      "loss": 0.3093,
+      "step": 17127
+    },
+    {
+      "epoch": 2.2295978133541583,
+      "grad_norm": 3.16357684135437,
+      "learning_rate": 1.6059387245244208e-06,
+      "loss": 0.4148,
+      "step": 17130
+    },
+    {
+      "epoch": 2.2299882858258493,
+      "grad_norm": 2.7439839839935303,
+      "learning_rate": 1.604404423864377e-06,
+      "loss": 0.3784,
+      "step": 17133
+    },
+    {
+      "epoch": 2.23037875829754,
+      "grad_norm": 3.130523443222046,
+      "learning_rate": 1.6028707164077367e-06,
+      "loss": 0.4011,
+      "step": 17136
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 2.8755338191986084,
+      "learning_rate": 1.6013376024224363e-06,
+      "loss": 0.3792,
+      "step": 17139
+    },
+    {
+      "epoch": 2.2311597032409214,
+      "grad_norm": 2.6047468185424805,
+      "learning_rate": 1.5998050821763033e-06,
+      "loss": 0.3807,
+      "step": 17142
+    },
+    {
+      "epoch": 2.2315501757126124,
+      "grad_norm": 2.492357015609741,
+      "learning_rate": 1.598273155937073e-06,
+      "loss": 0.3403,
+      "step": 17145
+    },
+    {
+      "epoch": 2.231940648184303,
+      "grad_norm": 3.0767757892608643,
+      "learning_rate": 1.5967418239723664e-06,
+      "loss": 0.3441,
+      "step": 17148
+    },
+    {
+      "epoch": 2.232331120655994,
+      "grad_norm": 2.8667490482330322,
+      "learning_rate": 1.5952110865497017e-06,
+      "loss": 0.3956,
+      "step": 17151
+    },
+    {
+      "epoch": 2.2327215931276845,
+      "grad_norm": 2.9534928798675537,
+      "learning_rate": 1.5936809439364992e-06,
+      "loss": 0.3098,
+      "step": 17154
+    },
+    {
+      "epoch": 2.233112065599375,
+      "grad_norm": 2.540276288986206,
+      "learning_rate": 1.59215139640007e-06,
+      "loss": 0.3845,
+      "step": 17157
+    },
+    {
+      "epoch": 2.233502538071066,
+      "grad_norm": 2.741945743560791,
+      "learning_rate": 1.590622444207623e-06,
+      "loss": 0.3631,
+      "step": 17160
+    },
+    {
+      "epoch": 2.2338930105427566,
+      "grad_norm": 2.531273603439331,
+      "learning_rate": 1.5890940876262612e-06,
+      "loss": 0.3323,
+      "step": 17163
+    },
+    {
+      "epoch": 2.2342834830144476,
+      "grad_norm": 3.036818027496338,
+      "learning_rate": 1.587566326922988e-06,
+      "loss": 0.3745,
+      "step": 17166
+    },
+    {
+      "epoch": 2.234673955486138,
+      "grad_norm": 2.696261167526245,
+      "learning_rate": 1.5860391623646986e-06,
+      "loss": 0.3351,
+      "step": 17169
+    },
+    {
+      "epoch": 2.235064427957829,
+      "grad_norm": 2.7615208625793457,
+      "learning_rate": 1.5845125942181844e-06,
+      "loss": 0.3886,
+      "step": 17172
+    },
+    {
+      "epoch": 2.2354549004295197,
+      "grad_norm": 3.1380362510681152,
+      "learning_rate": 1.5829866227501367e-06,
+      "loss": 0.3734,
+      "step": 17175
+    },
+    {
+      "epoch": 2.2358453729012107,
+      "grad_norm": 2.6224167346954346,
+      "learning_rate": 1.5814612482271368e-06,
+      "loss": 0.3422,
+      "step": 17178
+    },
+    {
+      "epoch": 2.236235845372901,
+      "grad_norm": 2.8184878826141357,
+      "learning_rate": 1.5799364709156645e-06,
+      "loss": 0.3194,
+      "step": 17181
+    },
+    {
+      "epoch": 2.2366263178445918,
+      "grad_norm": 2.749946355819702,
+      "learning_rate": 1.5784122910820976e-06,
+      "loss": 0.3587,
+      "step": 17184
+    },
+    {
+      "epoch": 2.2370167903162828,
+      "grad_norm": 3.1600887775421143,
+      "learning_rate": 1.5768887089927031e-06,
+      "loss": 0.3815,
+      "step": 17187
+    },
+    {
+      "epoch": 2.2374072627879733,
+      "grad_norm": 2.7639830112457275,
+      "learning_rate": 1.5753657249136523e-06,
+      "loss": 0.3757,
+      "step": 17190
+    },
+    {
+      "epoch": 2.2377977352596643,
+      "grad_norm": 2.792595624923706,
+      "learning_rate": 1.573843339111003e-06,
+      "loss": 0.3124,
+      "step": 17193
+    },
+    {
+      "epoch": 2.238188207731355,
+      "grad_norm": 3.1691551208496094,
+      "learning_rate": 1.5723215518507168e-06,
+      "loss": 0.3515,
+      "step": 17196
+    },
+    {
+      "epoch": 2.238578680203046,
+      "grad_norm": 2.5455732345581055,
+      "learning_rate": 1.570800363398644e-06,
+      "loss": 0.3376,
+      "step": 17199
+    },
+    {
+      "epoch": 2.2389691526747364,
+      "grad_norm": 2.700528383255005,
+      "learning_rate": 1.5692797740205345e-06,
+      "loss": 0.3563,
+      "step": 17202
+    },
+    {
+      "epoch": 2.2393596251464274,
+      "grad_norm": 3.191154718399048,
+      "learning_rate": 1.5677597839820292e-06,
+      "loss": 0.402,
+      "step": 17205
+    },
+    {
+      "epoch": 2.239750097618118,
+      "grad_norm": 2.647813320159912,
+      "learning_rate": 1.566240393548671e-06,
+      "loss": 0.3639,
+      "step": 17208
+    },
+    {
+      "epoch": 2.2401405700898085,
+      "grad_norm": 2.85390043258667,
+      "learning_rate": 1.5647216029858924e-06,
+      "loss": 0.3551,
+      "step": 17211
+    },
+    {
+      "epoch": 2.2405310425614995,
+      "grad_norm": 2.7630879878997803,
+      "learning_rate": 1.5632034125590212e-06,
+      "loss": 0.3158,
+      "step": 17214
+    },
+    {
+      "epoch": 2.24092151503319,
+      "grad_norm": 2.9064273834228516,
+      "learning_rate": 1.5616858225332858e-06,
+      "loss": 0.3789,
+      "step": 17217
+    },
+    {
+      "epoch": 2.241311987504881,
+      "grad_norm": 2.9090540409088135,
+      "learning_rate": 1.560168833173804e-06,
+      "loss": 0.3402,
+      "step": 17220
+    },
+    {
+      "epoch": 2.2417024599765716,
+      "grad_norm": 3.207202911376953,
+      "learning_rate": 1.5586524447455892e-06,
+      "loss": 0.4046,
+      "step": 17223
+    },
+    {
+      "epoch": 2.2420929324482626,
+      "grad_norm": 2.5313596725463867,
+      "learning_rate": 1.5571366575135544e-06,
+      "loss": 0.3498,
+      "step": 17226
+    },
+    {
+      "epoch": 2.242483404919953,
+      "grad_norm": 3.0708365440368652,
+      "learning_rate": 1.555621471742501e-06,
+      "loss": 0.3484,
+      "step": 17229
+    },
+    {
+      "epoch": 2.242873877391644,
+      "grad_norm": 3.1129088401794434,
+      "learning_rate": 1.5541068876971322e-06,
+      "loss": 0.4346,
+      "step": 17232
+    },
+    {
+      "epoch": 2.2432643498633347,
+      "grad_norm": 2.6781063079833984,
+      "learning_rate": 1.552592905642039e-06,
+      "loss": 0.3724,
+      "step": 17235
+    },
+    {
+      "epoch": 2.2436548223350252,
+      "grad_norm": 3.136171817779541,
+      "learning_rate": 1.5510795258417149e-06,
+      "loss": 0.4596,
+      "step": 17238
+    },
+    {
+      "epoch": 2.244045294806716,
+      "grad_norm": 2.706580877304077,
+      "learning_rate": 1.5495667485605425e-06,
+      "loss": 0.3578,
+      "step": 17241
+    },
+    {
+      "epoch": 2.2444357672784068,
+      "grad_norm": 2.8641068935394287,
+      "learning_rate": 1.5480545740627984e-06,
+      "loss": 0.3521,
+      "step": 17244
+    },
+    {
+      "epoch": 2.2448262397500978,
+      "grad_norm": 2.81913161277771,
+      "learning_rate": 1.5465430026126605e-06,
+      "loss": 0.385,
+      "step": 17247
+    },
+    {
+      "epoch": 2.2452167122217883,
+      "grad_norm": 2.8231277465820312,
+      "learning_rate": 1.5450320344741942e-06,
+      "loss": 0.3675,
+      "step": 17250
+    },
+    {
+      "epoch": 2.2456071846934793,
+      "grad_norm": 2.623539686203003,
+      "learning_rate": 1.5435216699113641e-06,
+      "loss": 0.3081,
+      "step": 17253
+    },
+    {
+      "epoch": 2.24599765716517,
+      "grad_norm": 3.059145927429199,
+      "learning_rate": 1.5420119091880247e-06,
+      "loss": 0.4299,
+      "step": 17256
+    },
+    {
+      "epoch": 2.2463881296368604,
+      "grad_norm": 2.7407338619232178,
+      "learning_rate": 1.5405027525679323e-06,
+      "loss": 0.329,
+      "step": 17259
+    },
+    {
+      "epoch": 2.2467786021085514,
+      "grad_norm": 3.146299123764038,
+      "learning_rate": 1.5389942003147313e-06,
+      "loss": 0.3063,
+      "step": 17262
+    },
+    {
+      "epoch": 2.247169074580242,
+      "grad_norm": 2.4751954078674316,
+      "learning_rate": 1.5374862526919605e-06,
+      "loss": 0.3154,
+      "step": 17265
+    },
+    {
+      "epoch": 2.247559547051933,
+      "grad_norm": 2.857436180114746,
+      "learning_rate": 1.5359789099630596e-06,
+      "loss": 0.4555,
+      "step": 17268
+    },
+    {
+      "epoch": 2.2479500195236235,
+      "grad_norm": 3.6831679344177246,
+      "learning_rate": 1.5344721723913535e-06,
+      "loss": 0.3972,
+      "step": 17271
+    },
+    {
+      "epoch": 2.2483404919953145,
+      "grad_norm": 2.737274646759033,
+      "learning_rate": 1.5329660402400703e-06,
+      "loss": 0.3751,
+      "step": 17274
+    },
+    {
+      "epoch": 2.248730964467005,
+      "grad_norm": 3.0916683673858643,
+      "learning_rate": 1.5314605137723254e-06,
+      "loss": 0.4125,
+      "step": 17277
+    },
+    {
+      "epoch": 2.2491214369386956,
+      "grad_norm": 2.5740551948547363,
+      "learning_rate": 1.529955593251133e-06,
+      "loss": 0.3866,
+      "step": 17280
+    },
+    {
+      "epoch": 2.2495119094103866,
+      "grad_norm": 2.7472755908966064,
+      "learning_rate": 1.5284512789393984e-06,
+      "loss": 0.3552,
+      "step": 17283
+    },
+    {
+      "epoch": 2.249902381882077,
+      "grad_norm": 2.672215461730957,
+      "learning_rate": 1.526947571099921e-06,
+      "loss": 0.3339,
+      "step": 17286
+    },
+    {
+      "epoch": 2.250292854353768,
+      "grad_norm": 2.8021116256713867,
+      "learning_rate": 1.525444469995398e-06,
+      "loss": 0.3135,
+      "step": 17289
+    },
+    {
+      "epoch": 2.2506833268254587,
+      "grad_norm": 2.7901899814605713,
+      "learning_rate": 1.5239419758884171e-06,
+      "loss": 0.2812,
+      "step": 17292
+    },
+    {
+      "epoch": 2.2510737992971497,
+      "grad_norm": 2.8856074810028076,
+      "learning_rate": 1.5224400890414587e-06,
+      "loss": 0.4185,
+      "step": 17295
+    },
+    {
+      "epoch": 2.2514642717688402,
+      "grad_norm": 2.6049904823303223,
+      "learning_rate": 1.5209388097169026e-06,
+      "loss": 0.3618,
+      "step": 17298
+    },
+    {
+      "epoch": 2.2518547442405312,
+      "grad_norm": 2.7991714477539062,
+      "learning_rate": 1.5194381381770173e-06,
+      "loss": 0.3726,
+      "step": 17301
+    },
+    {
+      "epoch": 2.2522452167122218,
+      "grad_norm": 3.187734603881836,
+      "learning_rate": 1.5179380746839678e-06,
+      "loss": 0.4094,
+      "step": 17304
+    },
+    {
+      "epoch": 2.2526356891839123,
+      "grad_norm": 2.6936240196228027,
+      "learning_rate": 1.5164386194998094e-06,
+      "loss": 0.3345,
+      "step": 17307
+    },
+    {
+      "epoch": 2.2530261616556033,
+      "grad_norm": 2.7314226627349854,
+      "learning_rate": 1.5149397728864979e-06,
+      "loss": 0.3792,
+      "step": 17310
+    },
+    {
+      "epoch": 2.253416634127294,
+      "grad_norm": 2.7893855571746826,
+      "learning_rate": 1.5134415351058744e-06,
+      "loss": 0.3403,
+      "step": 17313
+    },
+    {
+      "epoch": 2.253807106598985,
+      "grad_norm": 2.6449899673461914,
+      "learning_rate": 1.5119439064196823e-06,
+      "loss": 0.3472,
+      "step": 17316
+    },
+    {
+      "epoch": 2.2541975790706754,
+      "grad_norm": 3.2970783710479736,
+      "learning_rate": 1.5104468870895495e-06,
+      "loss": 0.4036,
+      "step": 17319
+    },
+    {
+      "epoch": 2.2545880515423664,
+      "grad_norm": 3.0442137718200684,
+      "learning_rate": 1.5089504773770069e-06,
+      "loss": 0.324,
+      "step": 17322
+    },
+    {
+      "epoch": 2.254978524014057,
+      "grad_norm": 2.8000681400299072,
+      "learning_rate": 1.5074546775434718e-06,
+      "loss": 0.3458,
+      "step": 17325
+    },
+    {
+      "epoch": 2.255368996485748,
+      "grad_norm": 2.9774391651153564,
+      "learning_rate": 1.5059594878502554e-06,
+      "loss": 0.3931,
+      "step": 17328
+    },
+    {
+      "epoch": 2.2557594689574385,
+      "grad_norm": 2.576894521713257,
+      "learning_rate": 1.5044649085585678e-06,
+      "loss": 0.3472,
+      "step": 17331
+    },
+    {
+      "epoch": 2.256149941429129,
+      "grad_norm": 2.661025285720825,
+      "learning_rate": 1.5029709399295066e-06,
+      "loss": 0.3077,
+      "step": 17334
+    },
+    {
+      "epoch": 2.25654041390082,
+      "grad_norm": 2.5842506885528564,
+      "learning_rate": 1.5014775822240645e-06,
+      "loss": 0.4093,
+      "step": 17337
+    },
+    {
+      "epoch": 2.2569308863725106,
+      "grad_norm": 3.0119876861572266,
+      "learning_rate": 1.4999848357031305e-06,
+      "loss": 0.3561,
+      "step": 17340
+    },
+    {
+      "epoch": 2.2573213588442016,
+      "grad_norm": 2.854466438293457,
+      "learning_rate": 1.498492700627483e-06,
+      "loss": 0.3519,
+      "step": 17343
+    },
+    {
+      "epoch": 2.257711831315892,
+      "grad_norm": 3.8666045665740967,
+      "learning_rate": 1.4970011772577925e-06,
+      "loss": 0.3475,
+      "step": 17346
+    },
+    {
+      "epoch": 2.258102303787583,
+      "grad_norm": 2.4900221824645996,
+      "learning_rate": 1.495510265854629e-06,
+      "loss": 0.3278,
+      "step": 17349
+    },
+    {
+      "epoch": 2.2584927762592737,
+      "grad_norm": 2.904358386993408,
+      "learning_rate": 1.4940199666784495e-06,
+      "loss": 0.3085,
+      "step": 17352
+    },
+    {
+      "epoch": 2.2588832487309647,
+      "grad_norm": 2.5294814109802246,
+      "learning_rate": 1.4925302799896053e-06,
+      "loss": 0.3618,
+      "step": 17355
+    },
+    {
+      "epoch": 2.2592737212026552,
+      "grad_norm": 2.6248788833618164,
+      "learning_rate": 1.491041206048342e-06,
+      "loss": 0.355,
+      "step": 17358
+    },
+    {
+      "epoch": 2.259664193674346,
+      "grad_norm": 2.460315465927124,
+      "learning_rate": 1.4895527451147995e-06,
+      "loss": 0.3503,
+      "step": 17361
+    },
+    {
+      "epoch": 2.260054666146037,
+      "grad_norm": 2.8616855144500732,
+      "learning_rate": 1.488064897449008e-06,
+      "loss": 0.4144,
+      "step": 17364
+    },
+    {
+      "epoch": 2.2604451386177273,
+      "grad_norm": 2.8192522525787354,
+      "learning_rate": 1.4865776633108908e-06,
+      "loss": 0.372,
+      "step": 17367
+    },
+    {
+      "epoch": 2.2608356110894183,
+      "grad_norm": 2.8044068813323975,
+      "learning_rate": 1.4850910429602633e-06,
+      "loss": 0.3388,
+      "step": 17370
+    },
+    {
+      "epoch": 2.261226083561109,
+      "grad_norm": 2.620159864425659,
+      "learning_rate": 1.4836050366568378e-06,
+      "loss": 0.356,
+      "step": 17373
+    },
+    {
+      "epoch": 2.2616165560328,
+      "grad_norm": 2.9459471702575684,
+      "learning_rate": 1.4821196446602148e-06,
+      "loss": 0.4086,
+      "step": 17376
+    },
+    {
+      "epoch": 2.2620070285044904,
+      "grad_norm": 2.8624160289764404,
+      "learning_rate": 1.4806348672298875e-06,
+      "loss": 0.4079,
+      "step": 17379
+    },
+    {
+      "epoch": 2.2623975009761814,
+      "grad_norm": 2.7950611114501953,
+      "learning_rate": 1.4791507046252474e-06,
+      "loss": 0.3334,
+      "step": 17382
+    },
+    {
+      "epoch": 2.262787973447872,
+      "grad_norm": 2.842926502227783,
+      "learning_rate": 1.4776671571055723e-06,
+      "loss": 0.3641,
+      "step": 17385
+    },
+    {
+      "epoch": 2.2631784459195625,
+      "grad_norm": 2.3689680099487305,
+      "learning_rate": 1.476184224930033e-06,
+      "loss": 0.3775,
+      "step": 17388
+    },
+    {
+      "epoch": 2.2635689183912535,
+      "grad_norm": 2.761807918548584,
+      "learning_rate": 1.4747019083576986e-06,
+      "loss": 0.3717,
+      "step": 17391
+    },
+    {
+      "epoch": 2.263959390862944,
+      "grad_norm": 2.9553542137145996,
+      "learning_rate": 1.4732202076475244e-06,
+      "loss": 0.3061,
+      "step": 17394
+    },
+    {
+      "epoch": 2.264349863334635,
+      "grad_norm": 2.7715632915496826,
+      "learning_rate": 1.4717391230583595e-06,
+      "loss": 0.3667,
+      "step": 17397
+    },
+    {
+      "epoch": 2.2647403358063256,
+      "grad_norm": 2.5581488609313965,
+      "learning_rate": 1.4702586548489467e-06,
+      "loss": 0.3156,
+      "step": 17400
+    },
+    {
+      "epoch": 2.265130808278016,
+      "grad_norm": 2.83203387260437,
+      "learning_rate": 1.4687788032779233e-06,
+      "loss": 0.3902,
+      "step": 17403
+    },
+    {
+      "epoch": 2.265521280749707,
+      "grad_norm": 3.0823655128479004,
+      "learning_rate": 1.4672995686038145e-06,
+      "loss": 0.4572,
+      "step": 17406
+    },
+    {
+      "epoch": 2.2659117532213977,
+      "grad_norm": 2.8829867839813232,
+      "learning_rate": 1.4658209510850373e-06,
+      "loss": 0.3649,
+      "step": 17409
+    },
+    {
+      "epoch": 2.2663022256930887,
+      "grad_norm": 2.9928228855133057,
+      "learning_rate": 1.4643429509799073e-06,
+      "loss": 0.3896,
+      "step": 17412
+    },
+    {
+      "epoch": 2.2666926981647793,
+      "grad_norm": 2.664480686187744,
+      "learning_rate": 1.4628655685466258e-06,
+      "loss": 0.3732,
+      "step": 17415
+    },
+    {
+      "epoch": 2.2670831706364702,
+      "grad_norm": 2.658304452896118,
+      "learning_rate": 1.4613888040432884e-06,
+      "loss": 0.3118,
+      "step": 17418
+    },
+    {
+      "epoch": 2.267473643108161,
+      "grad_norm": 2.6738052368164062,
+      "learning_rate": 1.459912657727881e-06,
+      "loss": 0.3854,
+      "step": 17421
+    },
+    {
+      "epoch": 2.267864115579852,
+      "grad_norm": 3.1468684673309326,
+      "learning_rate": 1.4584371298582866e-06,
+      "loss": 0.3526,
+      "step": 17424
+    },
+    {
+      "epoch": 2.2682545880515423,
+      "grad_norm": 2.8303110599517822,
+      "learning_rate": 1.456962220692275e-06,
+      "loss": 0.3028,
+      "step": 17427
+    },
+    {
+      "epoch": 2.268645060523233,
+      "grad_norm": 2.5083484649658203,
+      "learning_rate": 1.455487930487509e-06,
+      "loss": 0.3114,
+      "step": 17430
+    },
+    {
+      "epoch": 2.269035532994924,
+      "grad_norm": 2.9310407638549805,
+      "learning_rate": 1.4540142595015461e-06,
+      "loss": 0.3595,
+      "step": 17433
+    },
+    {
+      "epoch": 2.2694260054666144,
+      "grad_norm": 3.0042102336883545,
+      "learning_rate": 1.4525412079918327e-06,
+      "loss": 0.3652,
+      "step": 17436
+    },
+    {
+      "epoch": 2.2698164779383054,
+      "grad_norm": 2.961367130279541,
+      "learning_rate": 1.451068776215706e-06,
+      "loss": 0.4294,
+      "step": 17439
+    },
+    {
+      "epoch": 2.270206950409996,
+      "grad_norm": 2.6684181690216064,
+      "learning_rate": 1.449596964430398e-06,
+      "loss": 0.3593,
+      "step": 17442
+    },
+    {
+      "epoch": 2.270597422881687,
+      "grad_norm": 3.18241548538208,
+      "learning_rate": 1.448125772893033e-06,
+      "loss": 0.3256,
+      "step": 17445
+    },
+    {
+      "epoch": 2.2709878953533775,
+      "grad_norm": 2.6874358654022217,
+      "learning_rate": 1.4466552018606235e-06,
+      "loss": 0.3528,
+      "step": 17448
+    },
+    {
+      "epoch": 2.2713783678250685,
+      "grad_norm": 2.966919422149658,
+      "learning_rate": 1.4451852515900733e-06,
+      "loss": 0.4124,
+      "step": 17451
+    },
+    {
+      "epoch": 2.271768840296759,
+      "grad_norm": 2.7124547958374023,
+      "learning_rate": 1.443715922338183e-06,
+      "loss": 0.3237,
+      "step": 17454
+    },
+    {
+      "epoch": 2.2721593127684496,
+      "grad_norm": 2.423863172531128,
+      "learning_rate": 1.44224721436164e-06,
+      "loss": 0.3078,
+      "step": 17457
+    },
+    {
+      "epoch": 2.2725497852401406,
+      "grad_norm": 2.627487897872925,
+      "learning_rate": 1.4407791279170225e-06,
+      "loss": 0.3764,
+      "step": 17460
+    },
+    {
+      "epoch": 2.272940257711831,
+      "grad_norm": 2.7886862754821777,
+      "learning_rate": 1.439311663260805e-06,
+      "loss": 0.3625,
+      "step": 17463
+    },
+    {
+      "epoch": 2.273330730183522,
+      "grad_norm": 2.791762113571167,
+      "learning_rate": 1.43784482064935e-06,
+      "loss": 0.3917,
+      "step": 17466
+    },
+    {
+      "epoch": 2.2737212026552127,
+      "grad_norm": 3.137033700942993,
+      "learning_rate": 1.4363786003389108e-06,
+      "loss": 0.3928,
+      "step": 17469
+    },
+    {
+      "epoch": 2.2741116751269037,
+      "grad_norm": 2.5043063163757324,
+      "learning_rate": 1.4349130025856322e-06,
+      "loss": 0.3395,
+      "step": 17472
+    },
+    {
+      "epoch": 2.2745021475985943,
+      "grad_norm": 2.988461494445801,
+      "learning_rate": 1.4334480276455532e-06,
+      "loss": 0.3501,
+      "step": 17475
+    },
+    {
+      "epoch": 2.2748926200702853,
+      "grad_norm": 3.358941078186035,
+      "learning_rate": 1.4319836757746014e-06,
+      "loss": 0.3746,
+      "step": 17478
+    },
+    {
+      "epoch": 2.275283092541976,
+      "grad_norm": 2.6883740425109863,
+      "learning_rate": 1.4305199472285936e-06,
+      "loss": 0.3567,
+      "step": 17481
+    },
+    {
+      "epoch": 2.2756735650136664,
+      "grad_norm": 2.756044864654541,
+      "learning_rate": 1.4290568422632417e-06,
+      "loss": 0.3877,
+      "step": 17484
+    },
+    {
+      "epoch": 2.2760640374853573,
+      "grad_norm": 2.9537763595581055,
+      "learning_rate": 1.4275943611341491e-06,
+      "loss": 0.3604,
+      "step": 17487
+    },
+    {
+      "epoch": 2.276454509957048,
+      "grad_norm": 3.069265365600586,
+      "learning_rate": 1.4261325040968065e-06,
+      "loss": 0.3909,
+      "step": 17490
+    },
+    {
+      "epoch": 2.276844982428739,
+      "grad_norm": 2.7966959476470947,
+      "learning_rate": 1.4246712714065953e-06,
+      "loss": 0.3825,
+      "step": 17493
+    },
+    {
+      "epoch": 2.2772354549004294,
+      "grad_norm": 2.7966370582580566,
+      "learning_rate": 1.4232106633187932e-06,
+      "loss": 0.3359,
+      "step": 17496
+    },
+    {
+      "epoch": 2.2776259273721204,
+      "grad_norm": 2.9485177993774414,
+      "learning_rate": 1.4217506800885638e-06,
+      "loss": 0.3706,
+      "step": 17499
+    },
+    {
+      "epoch": 2.278016399843811,
+      "grad_norm": 2.909208297729492,
+      "learning_rate": 1.4202913219709612e-06,
+      "loss": 0.3484,
+      "step": 17502
+    },
+    {
+      "epoch": 2.278406872315502,
+      "grad_norm": 2.8797898292541504,
+      "learning_rate": 1.4188325892209359e-06,
+      "loss": 0.4045,
+      "step": 17505
+    },
+    {
+      "epoch": 2.2787973447871925,
+      "grad_norm": 2.849977970123291,
+      "learning_rate": 1.417374482093324e-06,
+      "loss": 0.3927,
+      "step": 17508
+    },
+    {
+      "epoch": 2.279187817258883,
+      "grad_norm": 2.766188383102417,
+      "learning_rate": 1.4159170008428513e-06,
+      "loss": 0.3263,
+      "step": 17511
+    },
+    {
+      "epoch": 2.279578289730574,
+      "grad_norm": 2.4690568447113037,
+      "learning_rate": 1.4144601457241402e-06,
+      "loss": 0.3001,
+      "step": 17514
+    },
+    {
+      "epoch": 2.2799687622022646,
+      "grad_norm": 2.894524335861206,
+      "learning_rate": 1.4130039169916986e-06,
+      "loss": 0.3283,
+      "step": 17517
+    },
+    {
+      "epoch": 2.2803592346739556,
+      "grad_norm": 2.6616241931915283,
+      "learning_rate": 1.4115483148999277e-06,
+      "loss": 0.3245,
+      "step": 17520
+    },
+    {
+      "epoch": 2.280749707145646,
+      "grad_norm": 2.9152584075927734,
+      "learning_rate": 1.410093339703113e-06,
+      "loss": 0.3555,
+      "step": 17523
+    },
+    {
+      "epoch": 2.281140179617337,
+      "grad_norm": 2.7940874099731445,
+      "learning_rate": 1.4086389916554439e-06,
+      "loss": 0.3349,
+      "step": 17526
+    },
+    {
+      "epoch": 2.2815306520890277,
+      "grad_norm": 2.692840814590454,
+      "learning_rate": 1.4071852710109867e-06,
+      "loss": 0.3424,
+      "step": 17529
+    },
+    {
+      "epoch": 2.2819211245607187,
+      "grad_norm": 2.8541390895843506,
+      "learning_rate": 1.4057321780237055e-06,
+      "loss": 0.4011,
+      "step": 17532
+    },
+    {
+      "epoch": 2.2823115970324093,
+      "grad_norm": 2.5435984134674072,
+      "learning_rate": 1.4042797129474495e-06,
+      "loss": 0.2974,
+      "step": 17535
+    },
+    {
+      "epoch": 2.2827020695041,
+      "grad_norm": 2.908129930496216,
+      "learning_rate": 1.4028278760359649e-06,
+      "loss": 0.3302,
+      "step": 17538
+    },
+    {
+      "epoch": 2.283092541975791,
+      "grad_norm": 2.641526222229004,
+      "learning_rate": 1.4013766675428831e-06,
+      "loss": 0.332,
+      "step": 17541
+    },
+    {
+      "epoch": 2.2834830144474814,
+      "grad_norm": 2.8526651859283447,
+      "learning_rate": 1.3999260877217259e-06,
+      "loss": 0.3843,
+      "step": 17544
+    },
+    {
+      "epoch": 2.2838734869191724,
+      "grad_norm": 2.7316031455993652,
+      "learning_rate": 1.3984761368259087e-06,
+      "loss": 0.3498,
+      "step": 17547
+    },
+    {
+      "epoch": 2.284263959390863,
+      "grad_norm": 2.5377614498138428,
+      "learning_rate": 1.3970268151087341e-06,
+      "loss": 0.3165,
+      "step": 17550
+    },
+    {
+      "epoch": 2.2846544318625535,
+      "grad_norm": 2.78267502784729,
+      "learning_rate": 1.3955781228233938e-06,
+      "loss": 0.3007,
+      "step": 17553
+    },
+    {
+      "epoch": 2.2850449043342445,
+      "grad_norm": 2.560595989227295,
+      "learning_rate": 1.3941300602229746e-06,
+      "loss": 0.3325,
+      "step": 17556
+    },
+    {
+      "epoch": 2.285435376805935,
+      "grad_norm": 2.923177480697632,
+      "learning_rate": 1.3926826275604476e-06,
+      "loss": 0.4159,
+      "step": 17559
+    },
+    {
+      "epoch": 2.285825849277626,
+      "grad_norm": 2.99003267288208,
+      "learning_rate": 1.3912358250886775e-06,
+      "loss": 0.3853,
+      "step": 17562
+    },
+    {
+      "epoch": 2.2862163217493165,
+      "grad_norm": 2.788792610168457,
+      "learning_rate": 1.3897896530604138e-06,
+      "loss": 0.3513,
+      "step": 17565
+    },
+    {
+      "epoch": 2.2866067942210075,
+      "grad_norm": 3.1519463062286377,
+      "learning_rate": 1.3883441117283058e-06,
+      "loss": 0.4351,
+      "step": 17568
+    },
+    {
+      "epoch": 2.286997266692698,
+      "grad_norm": 2.7040798664093018,
+      "learning_rate": 1.386899201344884e-06,
+      "loss": 0.3028,
+      "step": 17571
+    },
+    {
+      "epoch": 2.287387739164389,
+      "grad_norm": 3.0929603576660156,
+      "learning_rate": 1.3854549221625696e-06,
+      "loss": 0.3445,
+      "step": 17574
+    },
+    {
+      "epoch": 2.2877782116360796,
+      "grad_norm": 2.985767364501953,
+      "learning_rate": 1.3840112744336775e-06,
+      "loss": 0.3272,
+      "step": 17577
+    },
+    {
+      "epoch": 2.28816868410777,
+      "grad_norm": 2.829904794692993,
+      "learning_rate": 1.3825682584104088e-06,
+      "loss": 0.3655,
+      "step": 17580
+    },
+    {
+      "epoch": 2.288559156579461,
+      "grad_norm": 3.0158586502075195,
+      "learning_rate": 1.3811258743448553e-06,
+      "loss": 0.3236,
+      "step": 17583
+    },
+    {
+      "epoch": 2.2889496290511517,
+      "grad_norm": 2.844355344772339,
+      "learning_rate": 1.3796841224889973e-06,
+      "loss": 0.3448,
+      "step": 17586
+    },
+    {
+      "epoch": 2.2893401015228427,
+      "grad_norm": 2.8782236576080322,
+      "learning_rate": 1.3782430030947087e-06,
+      "loss": 0.4191,
+      "step": 17589
+    },
+    {
+      "epoch": 2.2897305739945333,
+      "grad_norm": 2.368623733520508,
+      "learning_rate": 1.3768025164137478e-06,
+      "loss": 0.3288,
+      "step": 17592
+    },
+    {
+      "epoch": 2.2901210464662243,
+      "grad_norm": 2.9353768825531006,
+      "learning_rate": 1.375362662697764e-06,
+      "loss": 0.3488,
+      "step": 17595
+    },
+    {
+      "epoch": 2.290511518937915,
+      "grad_norm": 2.769230604171753,
+      "learning_rate": 1.3739234421982995e-06,
+      "loss": 0.3142,
+      "step": 17598
+    },
+    {
+      "epoch": 2.290901991409606,
+      "grad_norm": 2.760913133621216,
+      "learning_rate": 1.3724848551667812e-06,
+      "loss": 0.3319,
+      "step": 17601
+    },
+    {
+      "epoch": 2.2912924638812964,
+      "grad_norm": 2.777470350265503,
+      "learning_rate": 1.3710469018545263e-06,
+      "loss": 0.3467,
+      "step": 17604
+    },
+    {
+      "epoch": 2.291682936352987,
+      "grad_norm": 2.680386543273926,
+      "learning_rate": 1.3696095825127436e-06,
+      "loss": 0.3023,
+      "step": 17607
+    },
+    {
+      "epoch": 2.292073408824678,
+      "grad_norm": 2.7917463779449463,
+      "learning_rate": 1.3681728973925313e-06,
+      "loss": 0.393,
+      "step": 17610
+    },
+    {
+      "epoch": 2.2924638812963685,
+      "grad_norm": 3.1995327472686768,
+      "learning_rate": 1.3667368467448734e-06,
+      "loss": 0.3997,
+      "step": 17613
+    },
+    {
+      "epoch": 2.2928543537680595,
+      "grad_norm": 2.46683406829834,
+      "learning_rate": 1.365301430820643e-06,
+      "loss": 0.3356,
+      "step": 17616
+    },
+    {
+      "epoch": 2.29324482623975,
+      "grad_norm": 3.129554271697998,
+      "learning_rate": 1.3638666498706082e-06,
+      "loss": 0.3385,
+      "step": 17619
+    },
+    {
+      "epoch": 2.293635298711441,
+      "grad_norm": 2.8934786319732666,
+      "learning_rate": 1.3624325041454206e-06,
+      "loss": 0.3568,
+      "step": 17622
+    },
+    {
+      "epoch": 2.2940257711831316,
+      "grad_norm": 2.703233242034912,
+      "learning_rate": 1.360998993895622e-06,
+      "loss": 0.3981,
+      "step": 17625
+    },
+    {
+      "epoch": 2.2944162436548226,
+      "grad_norm": 2.704498529434204,
+      "learning_rate": 1.3595661193716426e-06,
+      "loss": 0.3571,
+      "step": 17628
+    },
+    {
+      "epoch": 2.294806716126513,
+      "grad_norm": 2.624086856842041,
+      "learning_rate": 1.3581338808238048e-06,
+      "loss": 0.3518,
+      "step": 17631
+    },
+    {
+      "epoch": 2.2951971885982037,
+      "grad_norm": 2.516305685043335,
+      "learning_rate": 1.356702278502317e-06,
+      "loss": 0.3229,
+      "step": 17634
+    },
+    {
+      "epoch": 2.2955876610698946,
+      "grad_norm": 2.7615859508514404,
+      "learning_rate": 1.3552713126572752e-06,
+      "loss": 0.3107,
+      "step": 17637
+    },
+    {
+      "epoch": 2.295978133541585,
+      "grad_norm": 2.727630138397217,
+      "learning_rate": 1.353840983538669e-06,
+      "loss": 0.367,
+      "step": 17640
+    },
+    {
+      "epoch": 2.296368606013276,
+      "grad_norm": 2.636464834213257,
+      "learning_rate": 1.3524112913963728e-06,
+      "loss": 0.3203,
+      "step": 17643
+    },
+    {
+      "epoch": 2.2967590784849667,
+      "grad_norm": 2.5415730476379395,
+      "learning_rate": 1.3509822364801489e-06,
+      "loss": 0.3375,
+      "step": 17646
+    },
+    {
+      "epoch": 2.2971495509566577,
+      "grad_norm": 3.3716161251068115,
+      "learning_rate": 1.3495538190396524e-06,
+      "loss": 0.3362,
+      "step": 17649
+    },
+    {
+      "epoch": 2.2975400234283483,
+      "grad_norm": 2.9915719032287598,
+      "learning_rate": 1.348126039324425e-06,
+      "loss": 0.3083,
+      "step": 17652
+    },
+    {
+      "epoch": 2.2979304959000393,
+      "grad_norm": 3.0488333702087402,
+      "learning_rate": 1.3466988975838967e-06,
+      "loss": 0.3817,
+      "step": 17655
+    },
+    {
+      "epoch": 2.29832096837173,
+      "grad_norm": 3.0410192012786865,
+      "learning_rate": 1.3452723940673839e-06,
+      "loss": 0.367,
+      "step": 17658
+    },
+    {
+      "epoch": 2.2987114408434204,
+      "grad_norm": 3.0488669872283936,
+      "learning_rate": 1.343846529024097e-06,
+      "loss": 0.3775,
+      "step": 17661
+    },
+    {
+      "epoch": 2.2991019133151114,
+      "grad_norm": 2.820769786834717,
+      "learning_rate": 1.3424213027031297e-06,
+      "loss": 0.3809,
+      "step": 17664
+    },
+    {
+      "epoch": 2.299492385786802,
+      "grad_norm": 2.7791459560394287,
+      "learning_rate": 1.3409967153534654e-06,
+      "loss": 0.3571,
+      "step": 17667
+    },
+    {
+      "epoch": 2.299882858258493,
+      "grad_norm": 3.0136008262634277,
+      "learning_rate": 1.3395727672239789e-06,
+      "loss": 0.3502,
+      "step": 17670
+    },
+    {
+      "epoch": 2.3002733307301835,
+      "grad_norm": 2.868488073348999,
+      "learning_rate": 1.3381494585634292e-06,
+      "loss": 0.3603,
+      "step": 17673
+    },
+    {
+      "epoch": 2.300663803201874,
+      "grad_norm": 2.7699248790740967,
+      "learning_rate": 1.3367267896204662e-06,
+      "loss": 0.3638,
+      "step": 17676
+    },
+    {
+      "epoch": 2.301054275673565,
+      "grad_norm": 2.7452714443206787,
+      "learning_rate": 1.3353047606436248e-06,
+      "loss": 0.3532,
+      "step": 17679
+    },
+    {
+      "epoch": 2.3014447481452556,
+      "grad_norm": 2.6756961345672607,
+      "learning_rate": 1.3338833718813337e-06,
+      "loss": 0.3326,
+      "step": 17682
+    },
+    {
+      "epoch": 2.3018352206169466,
+      "grad_norm": 2.506359100341797,
+      "learning_rate": 1.3324626235819055e-06,
+      "loss": 0.3736,
+      "step": 17685
+    },
+    {
+      "epoch": 2.302225693088637,
+      "grad_norm": 2.6358907222747803,
+      "learning_rate": 1.3310425159935398e-06,
+      "loss": 0.3283,
+      "step": 17688
+    },
+    {
+      "epoch": 2.302616165560328,
+      "grad_norm": 2.714524745941162,
+      "learning_rate": 1.3296230493643282e-06,
+      "loss": 0.3764,
+      "step": 17691
+    },
+    {
+      "epoch": 2.3030066380320187,
+      "grad_norm": 2.8931455612182617,
+      "learning_rate": 1.3282042239422505e-06,
+      "loss": 0.3603,
+      "step": 17694
+    },
+    {
+      "epoch": 2.3033971105037097,
+      "grad_norm": 3.029452323913574,
+      "learning_rate": 1.3267860399751698e-06,
+      "loss": 0.3331,
+      "step": 17697
+    },
+    {
+      "epoch": 2.3037875829754,
+      "grad_norm": 2.932361125946045,
+      "learning_rate": 1.3253684977108394e-06,
+      "loss": 0.3672,
+      "step": 17700
+    },
+    {
+      "epoch": 2.3041780554470908,
+      "grad_norm": 3.243929624557495,
+      "learning_rate": 1.3239515973969042e-06,
+      "loss": 0.3559,
+      "step": 17703
+    },
+    {
+      "epoch": 2.3045685279187818,
+      "grad_norm": 2.824859142303467,
+      "learning_rate": 1.322535339280891e-06,
+      "loss": 0.4494,
+      "step": 17706
+    },
+    {
+      "epoch": 2.3049590003904723,
+      "grad_norm": 2.740846633911133,
+      "learning_rate": 1.3211197236102163e-06,
+      "loss": 0.3501,
+      "step": 17709
+    },
+    {
+      "epoch": 2.3053494728621633,
+      "grad_norm": 2.9181246757507324,
+      "learning_rate": 1.3197047506321887e-06,
+      "loss": 0.3736,
+      "step": 17712
+    },
+    {
+      "epoch": 2.305739945333854,
+      "grad_norm": 2.6927707195281982,
+      "learning_rate": 1.3182904205939983e-06,
+      "loss": 0.348,
+      "step": 17715
+    },
+    {
+      "epoch": 2.306130417805545,
+      "grad_norm": 2.88547945022583,
+      "learning_rate": 1.3168767337427251e-06,
+      "loss": 0.4355,
+      "step": 17718
+    },
+    {
+      "epoch": 2.3065208902772354,
+      "grad_norm": 2.848472833633423,
+      "learning_rate": 1.3154636903253398e-06,
+      "loss": 0.3914,
+      "step": 17721
+    },
+    {
+      "epoch": 2.3069113627489264,
+      "grad_norm": 2.7487449645996094,
+      "learning_rate": 1.3140512905886965e-06,
+      "loss": 0.3643,
+      "step": 17724
+    },
+    {
+      "epoch": 2.307301835220617,
+      "grad_norm": 2.6262028217315674,
+      "learning_rate": 1.312639534779539e-06,
+      "loss": 0.3678,
+      "step": 17727
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 2.6408939361572266,
+      "learning_rate": 1.3112284231444961e-06,
+      "loss": 0.3232,
+      "step": 17730
+    },
+    {
+      "epoch": 2.3080827801639985,
+      "grad_norm": 2.451420307159424,
+      "learning_rate": 1.3098179559300877e-06,
+      "loss": 0.3868,
+      "step": 17733
+    },
+    {
+      "epoch": 2.308473252635689,
+      "grad_norm": 2.4723563194274902,
+      "learning_rate": 1.3084081333827204e-06,
+      "loss": 0.3023,
+      "step": 17736
+    },
+    {
+      "epoch": 2.30886372510738,
+      "grad_norm": 2.9598135948181152,
+      "learning_rate": 1.3069989557486868e-06,
+      "loss": 0.3302,
+      "step": 17739
+    },
+    {
+      "epoch": 2.3092541975790706,
+      "grad_norm": 2.8125874996185303,
+      "learning_rate": 1.305590423274165e-06,
+      "loss": 0.3879,
+      "step": 17742
+    },
+    {
+      "epoch": 2.3096446700507616,
+      "grad_norm": 2.6237545013427734,
+      "learning_rate": 1.3041825362052258e-06,
+      "loss": 0.4228,
+      "step": 17745
+    },
+    {
+      "epoch": 2.310035142522452,
+      "grad_norm": 2.6856138706207275,
+      "learning_rate": 1.3027752947878226e-06,
+      "loss": 0.3903,
+      "step": 17748
+    },
+    {
+      "epoch": 2.310425614994143,
+      "grad_norm": 2.5263748168945312,
+      "learning_rate": 1.301368699267796e-06,
+      "loss": 0.328,
+      "step": 17751
+    },
+    {
+      "epoch": 2.3108160874658337,
+      "grad_norm": 2.7629663944244385,
+      "learning_rate": 1.2999627498908785e-06,
+      "loss": 0.345,
+      "step": 17754
+    },
+    {
+      "epoch": 2.311206559937524,
+      "grad_norm": 2.620549201965332,
+      "learning_rate": 1.2985574469026847e-06,
+      "loss": 0.3989,
+      "step": 17757
+    },
+    {
+      "epoch": 2.311597032409215,
+      "grad_norm": 3.176088333129883,
+      "learning_rate": 1.297152790548717e-06,
+      "loss": 0.3533,
+      "step": 17760
+    },
+    {
+      "epoch": 2.3119875048809058,
+      "grad_norm": 2.64750075340271,
+      "learning_rate": 1.2957487810743686e-06,
+      "loss": 0.3045,
+      "step": 17763
+    },
+    {
+      "epoch": 2.3123779773525968,
+      "grad_norm": 2.6987838745117188,
+      "learning_rate": 1.2943454187249154e-06,
+      "loss": 0.3716,
+      "step": 17766
+    },
+    {
+      "epoch": 2.3127684498242873,
+      "grad_norm": 2.43686580657959,
+      "learning_rate": 1.2929427037455206e-06,
+      "loss": 0.328,
+      "step": 17769
+    },
+    {
+      "epoch": 2.3131589222959783,
+      "grad_norm": 2.7529258728027344,
+      "learning_rate": 1.2915406363812384e-06,
+      "loss": 0.3639,
+      "step": 17772
+    },
+    {
+      "epoch": 2.313549394767669,
+      "grad_norm": 2.7594821453094482,
+      "learning_rate": 1.2901392168770038e-06,
+      "loss": 0.3543,
+      "step": 17775
+    },
+    {
+      "epoch": 2.31393986723936,
+      "grad_norm": 2.7785696983337402,
+      "learning_rate": 1.2887384454776453e-06,
+      "loss": 0.3617,
+      "step": 17778
+    },
+    {
+      "epoch": 2.3143303397110504,
+      "grad_norm": 2.8949971199035645,
+      "learning_rate": 1.2873383224278717e-06,
+      "loss": 0.3065,
+      "step": 17781
+    },
+    {
+      "epoch": 2.314720812182741,
+      "grad_norm": 2.9128878116607666,
+      "learning_rate": 1.2859388479722846e-06,
+      "loss": 0.3951,
+      "step": 17784
+    },
+    {
+      "epoch": 2.315111284654432,
+      "grad_norm": 3.0318446159362793,
+      "learning_rate": 1.2845400223553666e-06,
+      "loss": 0.4716,
+      "step": 17787
+    },
+    {
+      "epoch": 2.3155017571261225,
+      "grad_norm": 3.099668264389038,
+      "learning_rate": 1.2831418458214912e-06,
+      "loss": 0.3485,
+      "step": 17790
+    },
+    {
+      "epoch": 2.3158922295978135,
+      "grad_norm": 2.8386688232421875,
+      "learning_rate": 1.2817443186149148e-06,
+      "loss": 0.3256,
+      "step": 17793
+    },
+    {
+      "epoch": 2.316282702069504,
+      "grad_norm": 2.8773865699768066,
+      "learning_rate": 1.280347440979785e-06,
+      "loss": 0.377,
+      "step": 17796
+    },
+    {
+      "epoch": 2.316673174541195,
+      "grad_norm": 2.7028415203094482,
+      "learning_rate": 1.2789512131601323e-06,
+      "loss": 0.3315,
+      "step": 17799
+    },
+    {
+      "epoch": 2.3170636470128856,
+      "grad_norm": 2.7776808738708496,
+      "learning_rate": 1.2775556353998736e-06,
+      "loss": 0.3006,
+      "step": 17802
+    },
+    {
+      "epoch": 2.3174541194845766,
+      "grad_norm": 3.288285732269287,
+      "learning_rate": 1.2761607079428157e-06,
+      "loss": 0.4505,
+      "step": 17805
+    },
+    {
+      "epoch": 2.317844591956267,
+      "grad_norm": 2.7252891063690186,
+      "learning_rate": 1.2747664310326486e-06,
+      "loss": 0.3552,
+      "step": 17808
+    },
+    {
+      "epoch": 2.3182350644279577,
+      "grad_norm": 2.534891128540039,
+      "learning_rate": 1.2733728049129473e-06,
+      "loss": 0.3113,
+      "step": 17811
+    },
+    {
+      "epoch": 2.3186255368996487,
+      "grad_norm": 2.530785083770752,
+      "learning_rate": 1.271979829827179e-06,
+      "loss": 0.3388,
+      "step": 17814
+    },
+    {
+      "epoch": 2.3190160093713392,
+      "grad_norm": 2.7788338661193848,
+      "learning_rate": 1.2705875060186902e-06,
+      "loss": 0.3419,
+      "step": 17817
+    },
+    {
+      "epoch": 2.31940648184303,
+      "grad_norm": 2.8521158695220947,
+      "learning_rate": 1.2691958337307204e-06,
+      "loss": 0.3486,
+      "step": 17820
+    },
+    {
+      "epoch": 2.3197969543147208,
+      "grad_norm": 3.0471034049987793,
+      "learning_rate": 1.267804813206388e-06,
+      "loss": 0.3333,
+      "step": 17823
+    },
+    {
+      "epoch": 2.3201874267864113,
+      "grad_norm": 2.7590341567993164,
+      "learning_rate": 1.266414444688705e-06,
+      "loss": 0.3046,
+      "step": 17826
+    },
+    {
+      "epoch": 2.3205778992581023,
+      "grad_norm": 2.6539900302886963,
+      "learning_rate": 1.2650247284205646e-06,
+      "loss": 0.3385,
+      "step": 17829
+    },
+    {
+      "epoch": 2.320968371729793,
+      "grad_norm": 2.5913002490997314,
+      "learning_rate": 1.263635664644745e-06,
+      "loss": 0.3055,
+      "step": 17832
+    },
+    {
+      "epoch": 2.321358844201484,
+      "grad_norm": 2.7210516929626465,
+      "learning_rate": 1.2622472536039164e-06,
+      "loss": 0.3651,
+      "step": 17835
+    },
+    {
+      "epoch": 2.3217493166731744,
+      "grad_norm": 2.820171594619751,
+      "learning_rate": 1.2608594955406296e-06,
+      "loss": 0.3353,
+      "step": 17838
+    },
+    {
+      "epoch": 2.3221397891448654,
+      "grad_norm": 2.9673337936401367,
+      "learning_rate": 1.2594723906973222e-06,
+      "loss": 0.401,
+      "step": 17841
+    },
+    {
+      "epoch": 2.322530261616556,
+      "grad_norm": 2.832606792449951,
+      "learning_rate": 1.2580859393163185e-06,
+      "loss": 0.4028,
+      "step": 17844
+    },
+    {
+      "epoch": 2.322920734088247,
+      "grad_norm": 2.623936653137207,
+      "learning_rate": 1.2567001416398306e-06,
+      "loss": 0.3175,
+      "step": 17847
+    },
+    {
+      "epoch": 2.3233112065599375,
+      "grad_norm": 2.713925361633301,
+      "learning_rate": 1.2553149979099533e-06,
+      "loss": 0.3578,
+      "step": 17850
+    },
+    {
+      "epoch": 2.323701679031628,
+      "grad_norm": 2.729015827178955,
+      "learning_rate": 1.2539305083686665e-06,
+      "loss": 0.3648,
+      "step": 17853
+    },
+    {
+      "epoch": 2.324092151503319,
+      "grad_norm": 2.9586987495422363,
+      "learning_rate": 1.2525466732578407e-06,
+      "loss": 0.3634,
+      "step": 17856
+    },
+    {
+      "epoch": 2.3244826239750096,
+      "grad_norm": 2.6526246070861816,
+      "learning_rate": 1.2511634928192262e-06,
+      "loss": 0.3579,
+      "step": 17859
+    },
+    {
+      "epoch": 2.3248730964467006,
+      "grad_norm": 2.8939921855926514,
+      "learning_rate": 1.249780967294465e-06,
+      "loss": 0.3976,
+      "step": 17862
+    },
+    {
+      "epoch": 2.325263568918391,
+      "grad_norm": 2.95613169670105,
+      "learning_rate": 1.2483990969250776e-06,
+      "loss": 0.3352,
+      "step": 17865
+    },
+    {
+      "epoch": 2.325654041390082,
+      "grad_norm": 2.9669291973114014,
+      "learning_rate": 1.247017881952477e-06,
+      "loss": 0.3695,
+      "step": 17868
+    },
+    {
+      "epoch": 2.3260445138617727,
+      "grad_norm": 2.924299478530884,
+      "learning_rate": 1.245637322617958e-06,
+      "loss": 0.4432,
+      "step": 17871
+    },
+    {
+      "epoch": 2.3264349863334637,
+      "grad_norm": 2.776327610015869,
+      "learning_rate": 1.2442574191626988e-06,
+      "loss": 0.4004,
+      "step": 17874
+    },
+    {
+      "epoch": 2.3268254588051542,
+      "grad_norm": 2.6248810291290283,
+      "learning_rate": 1.2428781718277688e-06,
+      "loss": 0.3616,
+      "step": 17877
+    },
+    {
+      "epoch": 2.327215931276845,
+      "grad_norm": 2.729130983352661,
+      "learning_rate": 1.2414995808541186e-06,
+      "loss": 0.3167,
+      "step": 17880
+    },
+    {
+      "epoch": 2.327606403748536,
+      "grad_norm": 2.6671786308288574,
+      "learning_rate": 1.2401216464825838e-06,
+      "loss": 0.3463,
+      "step": 17883
+    },
+    {
+      "epoch": 2.3279968762202263,
+      "grad_norm": 2.7406859397888184,
+      "learning_rate": 1.2387443689538886e-06,
+      "loss": 0.4256,
+      "step": 17886
+    },
+    {
+      "epoch": 2.3283873486919173,
+      "grad_norm": 2.8111987113952637,
+      "learning_rate": 1.23736774850864e-06,
+      "loss": 0.3992,
+      "step": 17889
+    },
+    {
+      "epoch": 2.328777821163608,
+      "grad_norm": 2.890775680541992,
+      "learning_rate": 1.2359917853873304e-06,
+      "loss": 0.3903,
+      "step": 17892
+    },
+    {
+      "epoch": 2.329168293635299,
+      "grad_norm": 2.70546817779541,
+      "learning_rate": 1.2346164798303356e-06,
+      "loss": 0.3601,
+      "step": 17895
+    },
+    {
+      "epoch": 2.3295587661069894,
+      "grad_norm": 2.588521957397461,
+      "learning_rate": 1.2332418320779226e-06,
+      "loss": 0.3838,
+      "step": 17898
+    },
+    {
+      "epoch": 2.3299492385786804,
+      "grad_norm": 2.766190767288208,
+      "learning_rate": 1.2318678423702358e-06,
+      "loss": 0.362,
+      "step": 17901
+    },
+    {
+      "epoch": 2.330339711050371,
+      "grad_norm": 2.680659055709839,
+      "learning_rate": 1.230494510947311e-06,
+      "loss": 0.4384,
+      "step": 17904
+    },
+    {
+      "epoch": 2.3307301835220615,
+      "grad_norm": 2.6411266326904297,
+      "learning_rate": 1.2291218380490644e-06,
+      "loss": 0.3412,
+      "step": 17907
+    },
+    {
+      "epoch": 2.3311206559937525,
+      "grad_norm": 2.884023666381836,
+      "learning_rate": 1.2277498239153007e-06,
+      "loss": 0.3798,
+      "step": 17910
+    },
+    {
+      "epoch": 2.331511128465443,
+      "grad_norm": 2.8152637481689453,
+      "learning_rate": 1.2263784687857078e-06,
+      "loss": 0.4039,
+      "step": 17913
+    },
+    {
+      "epoch": 2.331901600937134,
+      "grad_norm": 3.315389394760132,
+      "learning_rate": 1.225007772899856e-06,
+      "loss": 0.3791,
+      "step": 17916
+    },
+    {
+      "epoch": 2.3322920734088246,
+      "grad_norm": 3.0322203636169434,
+      "learning_rate": 1.2236377364972063e-06,
+      "loss": 0.3572,
+      "step": 17919
+    },
+    {
+      "epoch": 2.3326825458805156,
+      "grad_norm": 2.6254141330718994,
+      "learning_rate": 1.2222683598171003e-06,
+      "loss": 0.3524,
+      "step": 17922
+    },
+    {
+      "epoch": 2.333073018352206,
+      "grad_norm": 2.877248525619507,
+      "learning_rate": 1.2208996430987625e-06,
+      "loss": 0.3472,
+      "step": 17925
+    },
+    {
+      "epoch": 2.333463490823897,
+      "grad_norm": 2.834516763687134,
+      "learning_rate": 1.2195315865813085e-06,
+      "loss": 0.4036,
+      "step": 17928
+    },
+    {
+      "epoch": 2.3338539632955877,
+      "grad_norm": 2.6819605827331543,
+      "learning_rate": 1.2181641905037338e-06,
+      "loss": 0.3461,
+      "step": 17931
+    },
+    {
+      "epoch": 2.3342444357672782,
+      "grad_norm": 2.962890148162842,
+      "learning_rate": 1.2167974551049177e-06,
+      "loss": 0.403,
+      "step": 17934
+    },
+    {
+      "epoch": 2.3346349082389692,
+      "grad_norm": 3.020080089569092,
+      "learning_rate": 1.2154313806236284e-06,
+      "loss": 0.4079,
+      "step": 17937
+    },
+    {
+      "epoch": 2.33502538071066,
+      "grad_norm": 2.56555438041687,
+      "learning_rate": 1.214065967298516e-06,
+      "loss": 0.3579,
+      "step": 17940
+    },
+    {
+      "epoch": 2.335415853182351,
+      "grad_norm": 3.1184959411621094,
+      "learning_rate": 1.2127012153681128e-06,
+      "loss": 0.4087,
+      "step": 17943
+    },
+    {
+      "epoch": 2.3358063256540413,
+      "grad_norm": 2.3602640628814697,
+      "learning_rate": 1.21133712507084e-06,
+      "loss": 0.3074,
+      "step": 17946
+    },
+    {
+      "epoch": 2.336196798125732,
+      "grad_norm": 3.22487735748291,
+      "learning_rate": 1.2099736966450026e-06,
+      "loss": 0.3818,
+      "step": 17949
+    },
+    {
+      "epoch": 2.336587270597423,
+      "grad_norm": 2.820575475692749,
+      "learning_rate": 1.208610930328788e-06,
+      "loss": 0.3451,
+      "step": 17952
+    },
+    {
+      "epoch": 2.3369777430691134,
+      "grad_norm": 2.8845815658569336,
+      "learning_rate": 1.2072488263602672e-06,
+      "loss": 0.3954,
+      "step": 17955
+    },
+    {
+      "epoch": 2.3373682155408044,
+      "grad_norm": 2.554292917251587,
+      "learning_rate": 1.2058873849773966e-06,
+      "loss": 0.3808,
+      "step": 17958
+    },
+    {
+      "epoch": 2.337758688012495,
+      "grad_norm": 2.7079639434814453,
+      "learning_rate": 1.2045266064180195e-06,
+      "loss": 0.4351,
+      "step": 17961
+    },
+    {
+      "epoch": 2.338149160484186,
+      "grad_norm": 2.7546820640563965,
+      "learning_rate": 1.2031664909198597e-06,
+      "loss": 0.4072,
+      "step": 17964
+    },
+    {
+      "epoch": 2.3385396329558765,
+      "grad_norm": 2.8127317428588867,
+      "learning_rate": 1.2018070387205256e-06,
+      "loss": 0.3719,
+      "step": 17967
+    },
+    {
+      "epoch": 2.3389301054275675,
+      "grad_norm": 2.782156229019165,
+      "learning_rate": 1.2004482500575126e-06,
+      "loss": 0.3663,
+      "step": 17970
+    },
+    {
+      "epoch": 2.339320577899258,
+      "grad_norm": 3.034802198410034,
+      "learning_rate": 1.1990901251681974e-06,
+      "loss": 0.3969,
+      "step": 17973
+    },
+    {
+      "epoch": 2.3397110503709486,
+      "grad_norm": 2.7988011837005615,
+      "learning_rate": 1.1977326642898395e-06,
+      "loss": 0.3998,
+      "step": 17976
+    },
+    {
+      "epoch": 2.3401015228426396,
+      "grad_norm": 2.6385715007781982,
+      "learning_rate": 1.1963758676595883e-06,
+      "loss": 0.3839,
+      "step": 17979
+    },
+    {
+      "epoch": 2.34049199531433,
+      "grad_norm": 2.976228713989258,
+      "learning_rate": 1.195019735514471e-06,
+      "loss": 0.3575,
+      "step": 17982
+    },
+    {
+      "epoch": 2.340882467786021,
+      "grad_norm": 2.7459535598754883,
+      "learning_rate": 1.1936642680914007e-06,
+      "loss": 0.3541,
+      "step": 17985
+    },
+    {
+      "epoch": 2.3412729402577117,
+      "grad_norm": 2.916914939880371,
+      "learning_rate": 1.1923094656271745e-06,
+      "loss": 0.395,
+      "step": 17988
+    },
+    {
+      "epoch": 2.3416634127294027,
+      "grad_norm": 2.652353286743164,
+      "learning_rate": 1.1909553283584763e-06,
+      "loss": 0.3298,
+      "step": 17991
+    },
+    {
+      "epoch": 2.3420538852010933,
+      "grad_norm": 2.817486524581909,
+      "learning_rate": 1.1896018565218692e-06,
+      "loss": 0.3436,
+      "step": 17994
+    },
+    {
+      "epoch": 2.3424443576727842,
+      "grad_norm": 2.9051315784454346,
+      "learning_rate": 1.1882490503538003e-06,
+      "loss": 0.3559,
+      "step": 17997
+    },
+    {
+      "epoch": 2.342834830144475,
+      "grad_norm": 2.805586099624634,
+      "learning_rate": 1.1868969100906052e-06,
+      "loss": 0.3098,
+      "step": 18000
+    },
+    {
+      "epoch": 2.3432253026161654,
+      "grad_norm": 2.8245179653167725,
+      "learning_rate": 1.1855454359684982e-06,
+      "loss": 0.3487,
+      "step": 18003
+    },
+    {
+      "epoch": 2.3436157750878563,
+      "grad_norm": 2.939572811126709,
+      "learning_rate": 1.1841946282235788e-06,
+      "loss": 0.3594,
+      "step": 18006
+    },
+    {
+      "epoch": 2.344006247559547,
+      "grad_norm": 2.8865296840667725,
+      "learning_rate": 1.1828444870918292e-06,
+      "loss": 0.3613,
+      "step": 18009
+    },
+    {
+      "epoch": 2.344396720031238,
+      "grad_norm": 2.665978193283081,
+      "learning_rate": 1.1814950128091197e-06,
+      "loss": 0.3966,
+      "step": 18012
+    },
+    {
+      "epoch": 2.3447871925029284,
+      "grad_norm": 2.8467841148376465,
+      "learning_rate": 1.180146205611198e-06,
+      "loss": 0.354,
+      "step": 18015
+    },
+    {
+      "epoch": 2.3451776649746194,
+      "grad_norm": 2.763833999633789,
+      "learning_rate": 1.1787980657336967e-06,
+      "loss": 0.3064,
+      "step": 18018
+    },
+    {
+      "epoch": 2.34556813744631,
+      "grad_norm": 2.5303876399993896,
+      "learning_rate": 1.1774505934121361e-06,
+      "loss": 0.3062,
+      "step": 18021
+    },
+    {
+      "epoch": 2.345958609918001,
+      "grad_norm": 2.5517313480377197,
+      "learning_rate": 1.176103788881916e-06,
+      "loss": 0.3531,
+      "step": 18024
+    },
+    {
+      "epoch": 2.3463490823896915,
+      "grad_norm": 2.689030885696411,
+      "learning_rate": 1.1747576523783178e-06,
+      "loss": 0.3974,
+      "step": 18027
+    },
+    {
+      "epoch": 2.346739554861382,
+      "grad_norm": 2.534513235092163,
+      "learning_rate": 1.1734121841365104e-06,
+      "loss": 0.335,
+      "step": 18030
+    },
+    {
+      "epoch": 2.347130027333073,
+      "grad_norm": 2.8086178302764893,
+      "learning_rate": 1.1720673843915465e-06,
+      "loss": 0.3382,
+      "step": 18033
+    },
+    {
+      "epoch": 2.3475204998047636,
+      "grad_norm": 2.9617831707000732,
+      "learning_rate": 1.1707232533783574e-06,
+      "loss": 0.3986,
+      "step": 18036
+    },
+    {
+      "epoch": 2.3479109722764546,
+      "grad_norm": 2.74955415725708,
+      "learning_rate": 1.169379791331759e-06,
+      "loss": 0.3935,
+      "step": 18039
+    },
+    {
+      "epoch": 2.348301444748145,
+      "grad_norm": 2.898005962371826,
+      "learning_rate": 1.1680369984864536e-06,
+      "loss": 0.5264,
+      "step": 18042
+    },
+    {
+      "epoch": 2.348691917219836,
+      "grad_norm": 3.2025671005249023,
+      "learning_rate": 1.1666948750770236e-06,
+      "loss": 0.3295,
+      "step": 18045
+    },
+    {
+      "epoch": 2.3490823896915267,
+      "grad_norm": 2.760274887084961,
+      "learning_rate": 1.1653534213379348e-06,
+      "loss": 0.4275,
+      "step": 18048
+    },
+    {
+      "epoch": 2.3494728621632177,
+      "grad_norm": 2.571065664291382,
+      "learning_rate": 1.1640126375035348e-06,
+      "loss": 0.4015,
+      "step": 18051
+    },
+    {
+      "epoch": 2.3498633346349083,
+      "grad_norm": 2.9767799377441406,
+      "learning_rate": 1.1626725238080593e-06,
+      "loss": 0.3862,
+      "step": 18054
+    },
+    {
+      "epoch": 2.350253807106599,
+      "grad_norm": 2.5151243209838867,
+      "learning_rate": 1.161333080485621e-06,
+      "loss": 0.3542,
+      "step": 18057
+    },
+    {
+      "epoch": 2.35064427957829,
+      "grad_norm": 3.4934723377227783,
+      "learning_rate": 1.1599943077702163e-06,
+      "loss": 0.3933,
+      "step": 18060
+    },
+    {
+      "epoch": 2.3510347520499804,
+      "grad_norm": 2.7796919345855713,
+      "learning_rate": 1.1586562058957302e-06,
+      "loss": 0.3271,
+      "step": 18063
+    },
+    {
+      "epoch": 2.3514252245216714,
+      "grad_norm": 2.8203134536743164,
+      "learning_rate": 1.157318775095923e-06,
+      "loss": 0.3522,
+      "step": 18066
+    },
+    {
+      "epoch": 2.351815696993362,
+      "grad_norm": 3.154359817504883,
+      "learning_rate": 1.155982015604441e-06,
+      "loss": 0.2913,
+      "step": 18069
+    },
+    {
+      "epoch": 2.352206169465053,
+      "grad_norm": 2.685695171356201,
+      "learning_rate": 1.1546459276548145e-06,
+      "loss": 0.3054,
+      "step": 18072
+    },
+    {
+      "epoch": 2.3525966419367434,
+      "grad_norm": 3.0085608959198,
+      "learning_rate": 1.153310511480456e-06,
+      "loss": 0.4078,
+      "step": 18075
+    },
+    {
+      "epoch": 2.3529871144084344,
+      "grad_norm": 3.074324607849121,
+      "learning_rate": 1.1519757673146586e-06,
+      "loss": 0.3455,
+      "step": 18078
+    },
+    {
+      "epoch": 2.353377586880125,
+      "grad_norm": 3.065185546875,
+      "learning_rate": 1.1506416953905986e-06,
+      "loss": 0.3692,
+      "step": 18081
+    },
+    {
+      "epoch": 2.3537680593518155,
+      "grad_norm": 2.8767306804656982,
+      "learning_rate": 1.1493082959413383e-06,
+      "loss": 0.3868,
+      "step": 18084
+    },
+    {
+      "epoch": 2.3541585318235065,
+      "grad_norm": 3.119493007659912,
+      "learning_rate": 1.1479755691998172e-06,
+      "loss": 0.3715,
+      "step": 18087
+    },
+    {
+      "epoch": 2.354549004295197,
+      "grad_norm": 2.8516952991485596,
+      "learning_rate": 1.1466435153988597e-06,
+      "loss": 0.3386,
+      "step": 18090
+    },
+    {
+      "epoch": 2.354939476766888,
+      "grad_norm": 2.712786912918091,
+      "learning_rate": 1.1453121347711755e-06,
+      "loss": 0.3865,
+      "step": 18093
+    },
+    {
+      "epoch": 2.3553299492385786,
+      "grad_norm": 3.0909574031829834,
+      "learning_rate": 1.1439814275493522e-06,
+      "loss": 0.3825,
+      "step": 18096
+    },
+    {
+      "epoch": 2.355720421710269,
+      "grad_norm": 2.726844549179077,
+      "learning_rate": 1.1426513939658611e-06,
+      "loss": 0.3344,
+      "step": 18099
+    },
+    {
+      "epoch": 2.35611089418196,
+      "grad_norm": 2.7704012393951416,
+      "learning_rate": 1.1413220342530556e-06,
+      "loss": 0.4041,
+      "step": 18102
+    },
+    {
+      "epoch": 2.3565013666536507,
+      "grad_norm": 2.8426220417022705,
+      "learning_rate": 1.1399933486431747e-06,
+      "loss": 0.3427,
+      "step": 18105
+    },
+    {
+      "epoch": 2.3568918391253417,
+      "grad_norm": 2.8885271549224854,
+      "learning_rate": 1.138665337368336e-06,
+      "loss": 0.4052,
+      "step": 18108
+    },
+    {
+      "epoch": 2.3572823115970323,
+      "grad_norm": 2.9842803478240967,
+      "learning_rate": 1.1373380006605378e-06,
+      "loss": 0.3765,
+      "step": 18111
+    },
+    {
+      "epoch": 2.3576727840687233,
+      "grad_norm": 2.573024272918701,
+      "learning_rate": 1.1360113387516654e-06,
+      "loss": 0.2962,
+      "step": 18114
+    },
+    {
+      "epoch": 2.358063256540414,
+      "grad_norm": 2.5947093963623047,
+      "learning_rate": 1.1346853518734845e-06,
+      "loss": 0.322,
+      "step": 18117
+    },
+    {
+      "epoch": 2.358453729012105,
+      "grad_norm": 3.0403120517730713,
+      "learning_rate": 1.133360040257641e-06,
+      "loss": 0.4033,
+      "step": 18120
+    },
+    {
+      "epoch": 2.3588442014837954,
+      "grad_norm": 2.958102226257324,
+      "learning_rate": 1.1320354041356636e-06,
+      "loss": 0.3175,
+      "step": 18123
+    },
+    {
+      "epoch": 2.359234673955486,
+      "grad_norm": 2.5290329456329346,
+      "learning_rate": 1.1307114437389648e-06,
+      "loss": 0.3399,
+      "step": 18126
+    },
+    {
+      "epoch": 2.359625146427177,
+      "grad_norm": 2.7605209350585938,
+      "learning_rate": 1.1293881592988375e-06,
+      "loss": 0.3571,
+      "step": 18129
+    },
+    {
+      "epoch": 2.3600156188988675,
+      "grad_norm": 3.629070997238159,
+      "learning_rate": 1.128065551046455e-06,
+      "loss": 0.3737,
+      "step": 18132
+    },
+    {
+      "epoch": 2.3604060913705585,
+      "grad_norm": 3.1546218395233154,
+      "learning_rate": 1.1267436192128762e-06,
+      "loss": 0.3364,
+      "step": 18135
+    },
+    {
+      "epoch": 2.360796563842249,
+      "grad_norm": 2.7438509464263916,
+      "learning_rate": 1.1254223640290391e-06,
+      "loss": 0.322,
+      "step": 18138
+    },
+    {
+      "epoch": 2.36118703631394,
+      "grad_norm": 2.711299419403076,
+      "learning_rate": 1.1241017857257624e-06,
+      "loss": 0.3045,
+      "step": 18141
+    },
+    {
+      "epoch": 2.3615775087856306,
+      "grad_norm": 3.051964282989502,
+      "learning_rate": 1.1227818845337523e-06,
+      "loss": 0.3652,
+      "step": 18144
+    },
+    {
+      "epoch": 2.3619679812573215,
+      "grad_norm": 2.6929006576538086,
+      "learning_rate": 1.12146266068359e-06,
+      "loss": 0.3793,
+      "step": 18147
+    },
+    {
+      "epoch": 2.362358453729012,
+      "grad_norm": 2.7894413471221924,
+      "learning_rate": 1.1201441144057413e-06,
+      "loss": 0.3304,
+      "step": 18150
+    },
+    {
+      "epoch": 2.3627489262007026,
+      "grad_norm": 2.5840792655944824,
+      "learning_rate": 1.1188262459305515e-06,
+      "loss": 0.3482,
+      "step": 18153
+    },
+    {
+      "epoch": 2.3631393986723936,
+      "grad_norm": 3.0223450660705566,
+      "learning_rate": 1.117509055488254e-06,
+      "loss": 0.4019,
+      "step": 18156
+    },
+    {
+      "epoch": 2.363529871144084,
+      "grad_norm": 2.9108870029449463,
+      "learning_rate": 1.1161925433089578e-06,
+      "loss": 0.3457,
+      "step": 18159
+    },
+    {
+      "epoch": 2.363920343615775,
+      "grad_norm": 2.797194480895996,
+      "learning_rate": 1.114876709622653e-06,
+      "loss": 0.3378,
+      "step": 18162
+    },
+    {
+      "epoch": 2.3643108160874657,
+      "grad_norm": 2.627267599105835,
+      "learning_rate": 1.1135615546592132e-06,
+      "loss": 0.3834,
+      "step": 18165
+    },
+    {
+      "epoch": 2.3647012885591567,
+      "grad_norm": 2.9231019020080566,
+      "learning_rate": 1.1122470786483946e-06,
+      "loss": 0.3934,
+      "step": 18168
+    },
+    {
+      "epoch": 2.3650917610308473,
+      "grad_norm": 2.57883358001709,
+      "learning_rate": 1.1109332818198338e-06,
+      "loss": 0.4163,
+      "step": 18171
+    },
+    {
+      "epoch": 2.3654822335025383,
+      "grad_norm": 3.0240585803985596,
+      "learning_rate": 1.1096201644030446e-06,
+      "loss": 0.371,
+      "step": 18174
+    },
+    {
+      "epoch": 2.365872705974229,
+      "grad_norm": 2.887322425842285,
+      "learning_rate": 1.108307726627431e-06,
+      "loss": 0.3658,
+      "step": 18177
+    },
+    {
+      "epoch": 2.3662631784459194,
+      "grad_norm": 2.607095956802368,
+      "learning_rate": 1.1069959687222704e-06,
+      "loss": 0.3982,
+      "step": 18180
+    },
+    {
+      "epoch": 2.3666536509176104,
+      "grad_norm": 3.5808451175689697,
+      "learning_rate": 1.1056848909167223e-06,
+      "loss": 0.3628,
+      "step": 18183
+    },
+    {
+      "epoch": 2.367044123389301,
+      "grad_norm": 3.329209566116333,
+      "learning_rate": 1.1043744934398332e-06,
+      "loss": 0.3591,
+      "step": 18186
+    },
+    {
+      "epoch": 2.367434595860992,
+      "grad_norm": 2.737271308898926,
+      "learning_rate": 1.1030647765205248e-06,
+      "loss": 0.3535,
+      "step": 18189
+    },
+    {
+      "epoch": 2.3678250683326825,
+      "grad_norm": 2.989947557449341,
+      "learning_rate": 1.1017557403876012e-06,
+      "loss": 0.3633,
+      "step": 18192
+    },
+    {
+      "epoch": 2.3682155408043735,
+      "grad_norm": 2.9825870990753174,
+      "learning_rate": 1.1004473852697484e-06,
+      "loss": 0.3497,
+      "step": 18195
+    },
+    {
+      "epoch": 2.368606013276064,
+      "grad_norm": 3.08147931098938,
+      "learning_rate": 1.0991397113955355e-06,
+      "loss": 0.371,
+      "step": 18198
+    },
+    {
+      "epoch": 2.368996485747755,
+      "grad_norm": 2.5311295986175537,
+      "learning_rate": 1.0978327189934085e-06,
+      "loss": 0.2979,
+      "step": 18201
+    },
+    {
+      "epoch": 2.3693869582194456,
+      "grad_norm": 2.984403371810913,
+      "learning_rate": 1.0965264082916954e-06,
+      "loss": 0.388,
+      "step": 18204
+    },
+    {
+      "epoch": 2.369777430691136,
+      "grad_norm": 2.8916494846343994,
+      "learning_rate": 1.0952207795186086e-06,
+      "loss": 0.3466,
+      "step": 18207
+    },
+    {
+      "epoch": 2.370167903162827,
+      "grad_norm": 2.8561973571777344,
+      "learning_rate": 1.0939158329022371e-06,
+      "loss": 0.3492,
+      "step": 18210
+    },
+    {
+      "epoch": 2.3705583756345177,
+      "grad_norm": 2.66711163520813,
+      "learning_rate": 1.0926115686705523e-06,
+      "loss": 0.3246,
+      "step": 18213
+    },
+    {
+      "epoch": 2.3709488481062087,
+      "grad_norm": 2.8468332290649414,
+      "learning_rate": 1.0913079870514055e-06,
+      "loss": 0.4033,
+      "step": 18216
+    },
+    {
+      "epoch": 2.371339320577899,
+      "grad_norm": 2.7105188369750977,
+      "learning_rate": 1.0900050882725316e-06,
+      "loss": 0.3426,
+      "step": 18219
+    },
+    {
+      "epoch": 2.37172979304959,
+      "grad_norm": 3.1098861694335938,
+      "learning_rate": 1.0887028725615433e-06,
+      "loss": 0.3738,
+      "step": 18222
+    },
+    {
+      "epoch": 2.3721202655212807,
+      "grad_norm": 2.924067258834839,
+      "learning_rate": 1.0874013401459338e-06,
+      "loss": 0.3517,
+      "step": 18225
+    },
+    {
+      "epoch": 2.3725107379929717,
+      "grad_norm": 2.9183595180511475,
+      "learning_rate": 1.0861004912530804e-06,
+      "loss": 0.3759,
+      "step": 18228
+    },
+    {
+      "epoch": 2.3729012104646623,
+      "grad_norm": 2.9503722190856934,
+      "learning_rate": 1.084800326110238e-06,
+      "loss": 0.3523,
+      "step": 18231
+    },
+    {
+      "epoch": 2.373291682936353,
+      "grad_norm": 2.875462770462036,
+      "learning_rate": 1.0835008449445406e-06,
+      "loss": 0.33,
+      "step": 18234
+    },
+    {
+      "epoch": 2.373682155408044,
+      "grad_norm": 2.600781202316284,
+      "learning_rate": 1.0822020479830064e-06,
+      "loss": 0.3591,
+      "step": 18237
+    },
+    {
+      "epoch": 2.3740726278797344,
+      "grad_norm": 2.702268123626709,
+      "learning_rate": 1.0809039354525342e-06,
+      "loss": 0.3282,
+      "step": 18240
+    },
+    {
+      "epoch": 2.3744631003514254,
+      "grad_norm": 2.848515748977661,
+      "learning_rate": 1.0796065075798995e-06,
+      "loss": 0.2933,
+      "step": 18243
+    },
+    {
+      "epoch": 2.374853572823116,
+      "grad_norm": 2.4967212677001953,
+      "learning_rate": 1.0783097645917594e-06,
+      "loss": 0.3887,
+      "step": 18246
+    },
+    {
+      "epoch": 2.3752440452948065,
+      "grad_norm": 2.6020994186401367,
+      "learning_rate": 1.0770137067146552e-06,
+      "loss": 0.3011,
+      "step": 18249
+    },
+    {
+      "epoch": 2.3756345177664975,
+      "grad_norm": 2.908611297607422,
+      "learning_rate": 1.0757183341750033e-06,
+      "loss": 0.3948,
+      "step": 18252
+    },
+    {
+      "epoch": 2.376024990238188,
+      "grad_norm": 3.1716883182525635,
+      "learning_rate": 1.0744236471991016e-06,
+      "loss": 0.3868,
+      "step": 18255
+    },
+    {
+      "epoch": 2.376415462709879,
+      "grad_norm": 2.6089026927948,
+      "learning_rate": 1.0731296460131319e-06,
+      "loss": 0.3903,
+      "step": 18258
+    },
+    {
+      "epoch": 2.3768059351815696,
+      "grad_norm": 2.7589609622955322,
+      "learning_rate": 1.0718363308431524e-06,
+      "loss": 0.3843,
+      "step": 18261
+    },
+    {
+      "epoch": 2.3771964076532606,
+      "grad_norm": 2.489070415496826,
+      "learning_rate": 1.0705437019151016e-06,
+      "loss": 0.3683,
+      "step": 18264
+    },
+    {
+      "epoch": 2.377586880124951,
+      "grad_norm": 2.6437742710113525,
+      "learning_rate": 1.069251759454799e-06,
+      "loss": 0.4111,
+      "step": 18267
+    },
+    {
+      "epoch": 2.377977352596642,
+      "grad_norm": 2.870964288711548,
+      "learning_rate": 1.067960503687946e-06,
+      "loss": 0.3937,
+      "step": 18270
+    },
+    {
+      "epoch": 2.3783678250683327,
+      "grad_norm": 2.932481527328491,
+      "learning_rate": 1.066669934840121e-06,
+      "loss": 0.3452,
+      "step": 18273
+    },
+    {
+      "epoch": 2.378758297540023,
+      "grad_norm": 2.969200611114502,
+      "learning_rate": 1.065380053136783e-06,
+      "loss": 0.3703,
+      "step": 18276
+    },
+    {
+      "epoch": 2.379148770011714,
+      "grad_norm": 3.3855032920837402,
+      "learning_rate": 1.0640908588032722e-06,
+      "loss": 0.3405,
+      "step": 18279
+    },
+    {
+      "epoch": 2.3795392424834048,
+      "grad_norm": 2.7585716247558594,
+      "learning_rate": 1.0628023520648102e-06,
+      "loss": 0.3448,
+      "step": 18282
+    },
+    {
+      "epoch": 2.3799297149550958,
+      "grad_norm": 2.9133167266845703,
+      "learning_rate": 1.0615145331464937e-06,
+      "loss": 0.3375,
+      "step": 18285
+    },
+    {
+      "epoch": 2.3803201874267863,
+      "grad_norm": 2.9258482456207275,
+      "learning_rate": 1.0602274022733023e-06,
+      "loss": 0.4111,
+      "step": 18288
+    },
+    {
+      "epoch": 2.3807106598984773,
+      "grad_norm": 2.8225510120391846,
+      "learning_rate": 1.0589409596700966e-06,
+      "loss": 0.3206,
+      "step": 18291
+    },
+    {
+      "epoch": 2.381101132370168,
+      "grad_norm": 2.6298437118530273,
+      "learning_rate": 1.0576552055616151e-06,
+      "loss": 0.3135,
+      "step": 18294
+    },
+    {
+      "epoch": 2.381491604841859,
+      "grad_norm": 2.665774345397949,
+      "learning_rate": 1.0563701401724735e-06,
+      "loss": 0.4122,
+      "step": 18297
+    },
+    {
+      "epoch": 2.3818820773135494,
+      "grad_norm": 2.916292190551758,
+      "learning_rate": 1.0550857637271744e-06,
+      "loss": 0.4741,
+      "step": 18300
+    },
+    {
+      "epoch": 2.38227254978524,
+      "grad_norm": 2.903151512145996,
+      "learning_rate": 1.0538020764500929e-06,
+      "loss": 0.3216,
+      "step": 18303
+    },
+    {
+      "epoch": 2.382663022256931,
+      "grad_norm": 2.8354744911193848,
+      "learning_rate": 1.052519078565486e-06,
+      "loss": 0.3428,
+      "step": 18306
+    },
+    {
+      "epoch": 2.3830534947286215,
+      "grad_norm": 2.986859083175659,
+      "learning_rate": 1.051236770297493e-06,
+      "loss": 0.3477,
+      "step": 18309
+    },
+    {
+      "epoch": 2.3834439672003125,
+      "grad_norm": 2.7683563232421875,
+      "learning_rate": 1.0499551518701296e-06,
+      "loss": 0.3927,
+      "step": 18312
+    },
+    {
+      "epoch": 2.383834439672003,
+      "grad_norm": 2.669161319732666,
+      "learning_rate": 1.048674223507291e-06,
+      "loss": 0.3973,
+      "step": 18315
+    },
+    {
+      "epoch": 2.384224912143694,
+      "grad_norm": 2.9369685649871826,
+      "learning_rate": 1.047393985432752e-06,
+      "loss": 0.3683,
+      "step": 18318
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 2.722029209136963,
+      "learning_rate": 1.0461144378701688e-06,
+      "loss": 0.2915,
+      "step": 18321
+    },
+    {
+      "epoch": 2.3850058570870756,
+      "grad_norm": 2.844775438308716,
+      "learning_rate": 1.0448355810430766e-06,
+      "loss": 0.3954,
+      "step": 18324
+    },
+    {
+      "epoch": 2.385396329558766,
+      "grad_norm": 2.715589761734009,
+      "learning_rate": 1.0435574151748878e-06,
+      "loss": 0.3586,
+      "step": 18327
+    },
+    {
+      "epoch": 2.3857868020304567,
+      "grad_norm": 3.002556562423706,
+      "learning_rate": 1.0422799404888945e-06,
+      "loss": 0.4168,
+      "step": 18330
+    },
+    {
+      "epoch": 2.3861772745021477,
+      "grad_norm": 2.758995532989502,
+      "learning_rate": 1.0410031572082712e-06,
+      "loss": 0.3743,
+      "step": 18333
+    },
+    {
+      "epoch": 2.3865677469738382,
+      "grad_norm": 2.8067500591278076,
+      "learning_rate": 1.0397270655560676e-06,
+      "loss": 0.3647,
+      "step": 18336
+    },
+    {
+      "epoch": 2.386958219445529,
+      "grad_norm": 2.654280185699463,
+      "learning_rate": 1.0384516657552129e-06,
+      "loss": 0.3467,
+      "step": 18339
+    },
+    {
+      "epoch": 2.3873486919172198,
+      "grad_norm": 2.651869535446167,
+      "learning_rate": 1.03717695802852e-06,
+      "loss": 0.3492,
+      "step": 18342
+    },
+    {
+      "epoch": 2.3877391643889108,
+      "grad_norm": 2.8447647094726562,
+      "learning_rate": 1.035902942598676e-06,
+      "loss": 0.3398,
+      "step": 18345
+    },
+    {
+      "epoch": 2.3881296368606013,
+      "grad_norm": 2.987757444381714,
+      "learning_rate": 1.034629619688247e-06,
+      "loss": 0.441,
+      "step": 18348
+    },
+    {
+      "epoch": 2.3885201093322923,
+      "grad_norm": 2.6804354190826416,
+      "learning_rate": 1.0333569895196832e-06,
+      "loss": 0.3859,
+      "step": 18351
+    },
+    {
+      "epoch": 2.388910581803983,
+      "grad_norm": 2.683786153793335,
+      "learning_rate": 1.0320850523153087e-06,
+      "loss": 0.3567,
+      "step": 18354
+    },
+    {
+      "epoch": 2.3893010542756734,
+      "grad_norm": 2.5274264812469482,
+      "learning_rate": 1.0308138082973285e-06,
+      "loss": 0.2944,
+      "step": 18357
+    },
+    {
+      "epoch": 2.3896915267473644,
+      "grad_norm": 2.8091444969177246,
+      "learning_rate": 1.0295432576878246e-06,
+      "loss": 0.343,
+      "step": 18360
+    },
+    {
+      "epoch": 2.390081999219055,
+      "grad_norm": 2.602357864379883,
+      "learning_rate": 1.0282734007087601e-06,
+      "loss": 0.312,
+      "step": 18363
+    },
+    {
+      "epoch": 2.390472471690746,
+      "grad_norm": 2.9505271911621094,
+      "learning_rate": 1.0270042375819795e-06,
+      "loss": 0.3419,
+      "step": 18366
+    },
+    {
+      "epoch": 2.3908629441624365,
+      "grad_norm": 2.739809989929199,
+      "learning_rate": 1.025735768529199e-06,
+      "loss": 0.3484,
+      "step": 18369
+    },
+    {
+      "epoch": 2.391253416634127,
+      "grad_norm": 2.7395405769348145,
+      "learning_rate": 1.0244679937720203e-06,
+      "loss": 0.3406,
+      "step": 18372
+    },
+    {
+      "epoch": 2.391643889105818,
+      "grad_norm": 3.019829034805298,
+      "learning_rate": 1.0232009135319198e-06,
+      "loss": 0.4535,
+      "step": 18375
+    },
+    {
+      "epoch": 2.3920343615775086,
+      "grad_norm": 2.979292631149292,
+      "learning_rate": 1.021934528030254e-06,
+      "loss": 0.4343,
+      "step": 18378
+    },
+    {
+      "epoch": 2.3924248340491996,
+      "grad_norm": 2.9323439598083496,
+      "learning_rate": 1.0206688374882562e-06,
+      "loss": 0.3784,
+      "step": 18381
+    },
+    {
+      "epoch": 2.39281530652089,
+      "grad_norm": 2.577038049697876,
+      "learning_rate": 1.0194038421270426e-06,
+      "loss": 0.3109,
+      "step": 18384
+    },
+    {
+      "epoch": 2.393205778992581,
+      "grad_norm": 3.026864767074585,
+      "learning_rate": 1.0181395421676038e-06,
+      "loss": 0.3493,
+      "step": 18387
+    },
+    {
+      "epoch": 2.3935962514642717,
+      "grad_norm": 2.9030680656433105,
+      "learning_rate": 1.0168759378308085e-06,
+      "loss": 0.348,
+      "step": 18390
+    },
+    {
+      "epoch": 2.3939867239359627,
+      "grad_norm": 2.8374221324920654,
+      "learning_rate": 1.0156130293374094e-06,
+      "loss": 0.3319,
+      "step": 18393
+    },
+    {
+      "epoch": 2.3943771964076532,
+      "grad_norm": 2.8880085945129395,
+      "learning_rate": 1.0143508169080323e-06,
+      "loss": 0.3814,
+      "step": 18396
+    },
+    {
+      "epoch": 2.394767668879344,
+      "grad_norm": 2.656985282897949,
+      "learning_rate": 1.013089300763181e-06,
+      "loss": 0.371,
+      "step": 18399
+    },
+    {
+      "epoch": 2.3951581413510348,
+      "grad_norm": 2.770444631576538,
+      "learning_rate": 1.0118284811232432e-06,
+      "loss": 0.3254,
+      "step": 18402
+    },
+    {
+      "epoch": 2.3955486138227253,
+      "grad_norm": 2.928893804550171,
+      "learning_rate": 1.010568358208479e-06,
+      "loss": 0.3686,
+      "step": 18405
+    },
+    {
+      "epoch": 2.3959390862944163,
+      "grad_norm": 3.908521890640259,
+      "learning_rate": 1.009308932239031e-06,
+      "loss": 0.3932,
+      "step": 18408
+    },
+    {
+      "epoch": 2.396329558766107,
+      "grad_norm": 2.9738271236419678,
+      "learning_rate": 1.008050203434916e-06,
+      "loss": 0.3626,
+      "step": 18411
+    },
+    {
+      "epoch": 2.396720031237798,
+      "grad_norm": 3.0634217262268066,
+      "learning_rate": 1.006792172016034e-06,
+      "loss": 0.3842,
+      "step": 18414
+    },
+    {
+      "epoch": 2.3971105037094884,
+      "grad_norm": 2.783146858215332,
+      "learning_rate": 1.005534838202159e-06,
+      "loss": 0.3737,
+      "step": 18417
+    },
+    {
+      "epoch": 2.3975009761811794,
+      "grad_norm": 2.868107795715332,
+      "learning_rate": 1.004278202212945e-06,
+      "loss": 0.3692,
+      "step": 18420
+    },
+    {
+      "epoch": 2.39789144865287,
+      "grad_norm": 3.5386221408843994,
+      "learning_rate": 1.0030222642679217e-06,
+      "loss": 0.3371,
+      "step": 18423
+    },
+    {
+      "epoch": 2.3982819211245605,
+      "grad_norm": 2.8862059116363525,
+      "learning_rate": 1.0017670245865014e-06,
+      "loss": 0.3265,
+      "step": 18426
+    },
+    {
+      "epoch": 2.3986723935962515,
+      "grad_norm": 2.635298490524292,
+      "learning_rate": 1.0005124833879714e-06,
+      "loss": 0.329,
+      "step": 18429
+    },
+    {
+      "epoch": 2.399062866067942,
+      "grad_norm": 2.8082430362701416,
+      "learning_rate": 9.99258640891495e-07,
+      "loss": 0.3584,
+      "step": 18432
+    },
+    {
+      "epoch": 2.399453338539633,
+      "grad_norm": 3.525629997253418,
+      "learning_rate": 9.980054973161196e-07,
+      "loss": 0.381,
+      "step": 18435
+    },
+    {
+      "epoch": 2.3998438110113236,
+      "grad_norm": 3.2471086978912354,
+      "learning_rate": 9.967530528807644e-07,
+      "loss": 0.3784,
+      "step": 18438
+    },
+    {
+      "epoch": 2.4002342834830146,
+      "grad_norm": 3.815002202987671,
+      "learning_rate": 9.95501307804228e-07,
+      "loss": 0.3329,
+      "step": 18441
+    },
+    {
+      "epoch": 2.400624755954705,
+      "grad_norm": 3.0479953289031982,
+      "learning_rate": 9.942502623051908e-07,
+      "loss": 0.3522,
+      "step": 18444
+    },
+    {
+      "epoch": 2.401015228426396,
+      "grad_norm": 2.807485818862915,
+      "learning_rate": 9.929999166022042e-07,
+      "loss": 0.3269,
+      "step": 18447
+    },
+    {
+      "epoch": 2.4014057008980867,
+      "grad_norm": 2.907914400100708,
+      "learning_rate": 9.91750270913704e-07,
+      "loss": 0.3308,
+      "step": 18450
+    },
+    {
+      "epoch": 2.4017961733697772,
+      "grad_norm": 2.5052411556243896,
+      "learning_rate": 9.905013254579976e-07,
+      "loss": 0.3099,
+      "step": 18453
+    },
+    {
+      "epoch": 2.4021866458414682,
+      "grad_norm": 2.7181320190429688,
+      "learning_rate": 9.892530804532768e-07,
+      "loss": 0.3439,
+      "step": 18456
+    },
+    {
+      "epoch": 2.402577118313159,
+      "grad_norm": 3.0974040031433105,
+      "learning_rate": 9.880055361176049e-07,
+      "loss": 0.4185,
+      "step": 18459
+    },
+    {
+      "epoch": 2.40296759078485,
+      "grad_norm": 3.3100745677948,
+      "learning_rate": 9.867586926689249e-07,
+      "loss": 0.3995,
+      "step": 18462
+    },
+    {
+      "epoch": 2.4033580632565403,
+      "grad_norm": 2.790778160095215,
+      "learning_rate": 9.8551255032506e-07,
+      "loss": 0.3481,
+      "step": 18465
+    },
+    {
+      "epoch": 2.4037485357282313,
+      "grad_norm": 2.46254563331604,
+      "learning_rate": 9.842671093037075e-07,
+      "loss": 0.3661,
+      "step": 18468
+    },
+    {
+      "epoch": 2.404139008199922,
+      "grad_norm": 2.7810397148132324,
+      "learning_rate": 9.830223698224428e-07,
+      "loss": 0.365,
+      "step": 18471
+    },
+    {
+      "epoch": 2.404529480671613,
+      "grad_norm": 2.764301300048828,
+      "learning_rate": 9.817783320987183e-07,
+      "loss": 0.3454,
+      "step": 18474
+    },
+    {
+      "epoch": 2.4049199531433034,
+      "grad_norm": 3.241234302520752,
+      "learning_rate": 9.805349963498672e-07,
+      "loss": 0.3636,
+      "step": 18477
+    },
+    {
+      "epoch": 2.405310425614994,
+      "grad_norm": 2.733487844467163,
+      "learning_rate": 9.792923627930972e-07,
+      "loss": 0.3371,
+      "step": 18480
+    },
+    {
+      "epoch": 2.405700898086685,
+      "grad_norm": 2.4645349979400635,
+      "learning_rate": 9.780504316454915e-07,
+      "loss": 0.351,
+      "step": 18483
+    },
+    {
+      "epoch": 2.4060913705583755,
+      "grad_norm": 2.7485122680664062,
+      "learning_rate": 9.768092031240155e-07,
+      "loss": 0.399,
+      "step": 18486
+    },
+    {
+      "epoch": 2.4064818430300665,
+      "grad_norm": 2.626877784729004,
+      "learning_rate": 9.75568677445507e-07,
+      "loss": 0.3811,
+      "step": 18489
+    },
+    {
+      "epoch": 2.406872315501757,
+      "grad_norm": 2.7127082347869873,
+      "learning_rate": 9.743288548266855e-07,
+      "loss": 0.3375,
+      "step": 18492
+    },
+    {
+      "epoch": 2.407262787973448,
+      "grad_norm": 2.722588300704956,
+      "learning_rate": 9.730897354841435e-07,
+      "loss": 0.3514,
+      "step": 18495
+    },
+    {
+      "epoch": 2.4076532604451386,
+      "grad_norm": 3.016986846923828,
+      "learning_rate": 9.718513196343539e-07,
+      "loss": 0.3149,
+      "step": 18498
+    },
+    {
+      "epoch": 2.4080437329168296,
+      "grad_norm": 2.604357957839966,
+      "learning_rate": 9.70613607493665e-07,
+      "loss": 0.3478,
+      "step": 18501
+    },
+    {
+      "epoch": 2.40843420538852,
+      "grad_norm": 2.603797435760498,
+      "learning_rate": 9.693765992783017e-07,
+      "loss": 0.309,
+      "step": 18504
+    },
+    {
+      "epoch": 2.4088246778602107,
+      "grad_norm": 2.9746592044830322,
+      "learning_rate": 9.681402952043677e-07,
+      "loss": 0.3852,
+      "step": 18507
+    },
+    {
+      "epoch": 2.4092151503319017,
+      "grad_norm": 3.175025224685669,
+      "learning_rate": 9.669046954878425e-07,
+      "loss": 0.3769,
+      "step": 18510
+    },
+    {
+      "epoch": 2.4096056228035923,
+      "grad_norm": 2.841076374053955,
+      "learning_rate": 9.65669800344582e-07,
+      "loss": 0.3367,
+      "step": 18513
+    },
+    {
+      "epoch": 2.4099960952752832,
+      "grad_norm": 2.7390999794006348,
+      "learning_rate": 9.644356099903208e-07,
+      "loss": 0.3437,
+      "step": 18516
+    },
+    {
+      "epoch": 2.410386567746974,
+      "grad_norm": 2.828277349472046,
+      "learning_rate": 9.632021246406693e-07,
+      "loss": 0.3318,
+      "step": 18519
+    },
+    {
+      "epoch": 2.4107770402186643,
+      "grad_norm": 2.4388415813446045,
+      "learning_rate": 9.619693445111145e-07,
+      "loss": 0.344,
+      "step": 18522
+    },
+    {
+      "epoch": 2.4111675126903553,
+      "grad_norm": 2.6832144260406494,
+      "learning_rate": 9.607372698170191e-07,
+      "loss": 0.3834,
+      "step": 18525
+    },
+    {
+      "epoch": 2.411557985162046,
+      "grad_norm": 2.858412742614746,
+      "learning_rate": 9.595059007736268e-07,
+      "loss": 0.3879,
+      "step": 18528
+    },
+    {
+      "epoch": 2.411948457633737,
+      "grad_norm": 2.7546768188476562,
+      "learning_rate": 9.582752375960519e-07,
+      "loss": 0.3568,
+      "step": 18531
+    },
+    {
+      "epoch": 2.4123389301054274,
+      "grad_norm": 2.633880853652954,
+      "learning_rate": 9.570452804992925e-07,
+      "loss": 0.3444,
+      "step": 18534
+    },
+    {
+      "epoch": 2.4127294025771184,
+      "grad_norm": 2.5289742946624756,
+      "learning_rate": 9.558160296982154e-07,
+      "loss": 0.346,
+      "step": 18537
+    },
+    {
+      "epoch": 2.413119875048809,
+      "grad_norm": 2.8704309463500977,
+      "learning_rate": 9.54587485407572e-07,
+      "loss": 0.3572,
+      "step": 18540
+    },
+    {
+      "epoch": 2.4135103475205,
+      "grad_norm": 2.7778470516204834,
+      "learning_rate": 9.533596478419843e-07,
+      "loss": 0.417,
+      "step": 18543
+    },
+    {
+      "epoch": 2.4139008199921905,
+      "grad_norm": 2.997497320175171,
+      "learning_rate": 9.521325172159518e-07,
+      "loss": 0.3553,
+      "step": 18546
+    },
+    {
+      "epoch": 2.414291292463881,
+      "grad_norm": 3.075047016143799,
+      "learning_rate": 9.509060937438546e-07,
+      "loss": 0.3914,
+      "step": 18549
+    },
+    {
+      "epoch": 2.414681764935572,
+      "grad_norm": 2.607013463973999,
+      "learning_rate": 9.496803776399449e-07,
+      "loss": 0.3676,
+      "step": 18552
+    },
+    {
+      "epoch": 2.4150722374072626,
+      "grad_norm": 2.543388843536377,
+      "learning_rate": 9.484553691183512e-07,
+      "loss": 0.3634,
+      "step": 18555
+    },
+    {
+      "epoch": 2.4154627098789536,
+      "grad_norm": 2.790480852127075,
+      "learning_rate": 9.472310683930824e-07,
+      "loss": 0.4357,
+      "step": 18558
+    },
+    {
+      "epoch": 2.415853182350644,
+      "grad_norm": 2.679603338241577,
+      "learning_rate": 9.460074756780202e-07,
+      "loss": 0.3298,
+      "step": 18561
+    },
+    {
+      "epoch": 2.416243654822335,
+      "grad_norm": 2.8844940662384033,
+      "learning_rate": 9.44784591186923e-07,
+      "loss": 0.3547,
+      "step": 18564
+    },
+    {
+      "epoch": 2.4166341272940257,
+      "grad_norm": 2.9198670387268066,
+      "learning_rate": 9.435624151334272e-07,
+      "loss": 0.3595,
+      "step": 18567
+    },
+    {
+      "epoch": 2.4170245997657167,
+      "grad_norm": 2.825845956802368,
+      "learning_rate": 9.423409477310446e-07,
+      "loss": 0.3814,
+      "step": 18570
+    },
+    {
+      "epoch": 2.4174150722374073,
+      "grad_norm": 2.549248456954956,
+      "learning_rate": 9.411201891931609e-07,
+      "loss": 0.3238,
+      "step": 18573
+    },
+    {
+      "epoch": 2.417805544709098,
+      "grad_norm": 3.1343023777008057,
+      "learning_rate": 9.399001397330415e-07,
+      "loss": 0.4215,
+      "step": 18576
+    },
+    {
+      "epoch": 2.418196017180789,
+      "grad_norm": 2.932382822036743,
+      "learning_rate": 9.386807995638275e-07,
+      "loss": 0.3348,
+      "step": 18579
+    },
+    {
+      "epoch": 2.4185864896524794,
+      "grad_norm": 2.6527950763702393,
+      "learning_rate": 9.374621688985341e-07,
+      "loss": 0.3421,
+      "step": 18582
+    },
+    {
+      "epoch": 2.4189769621241703,
+      "grad_norm": 2.6357645988464355,
+      "learning_rate": 9.362442479500539e-07,
+      "loss": 0.33,
+      "step": 18585
+    },
+    {
+      "epoch": 2.419367434595861,
+      "grad_norm": 2.9315764904022217,
+      "learning_rate": 9.350270369311531e-07,
+      "loss": 0.3976,
+      "step": 18588
+    },
+    {
+      "epoch": 2.419757907067552,
+      "grad_norm": 3.037123441696167,
+      "learning_rate": 9.338105360544786e-07,
+      "loss": 0.3024,
+      "step": 18591
+    },
+    {
+      "epoch": 2.4201483795392424,
+      "grad_norm": 2.859992742538452,
+      "learning_rate": 9.325947455325496e-07,
+      "loss": 0.4043,
+      "step": 18594
+    },
+    {
+      "epoch": 2.4205388520109334,
+      "grad_norm": 2.6656060218811035,
+      "learning_rate": 9.313796655777613e-07,
+      "loss": 0.3422,
+      "step": 18597
+    },
+    {
+      "epoch": 2.420929324482624,
+      "grad_norm": 2.7755658626556396,
+      "learning_rate": 9.301652964023866e-07,
+      "loss": 0.3565,
+      "step": 18600
+    },
+    {
+      "epoch": 2.4213197969543145,
+      "grad_norm": 2.9836435317993164,
+      "learning_rate": 9.289516382185737e-07,
+      "loss": 0.3223,
+      "step": 18603
+    },
+    {
+      "epoch": 2.4217102694260055,
+      "grad_norm": 2.6815292835235596,
+      "learning_rate": 9.277386912383435e-07,
+      "loss": 0.403,
+      "step": 18606
+    },
+    {
+      "epoch": 2.422100741897696,
+      "grad_norm": 3.0170223712921143,
+      "learning_rate": 9.265264556735987e-07,
+      "loss": 0.2842,
+      "step": 18609
+    },
+    {
+      "epoch": 2.422491214369387,
+      "grad_norm": 2.64164662361145,
+      "learning_rate": 9.253149317361126e-07,
+      "loss": 0.3211,
+      "step": 18612
+    },
+    {
+      "epoch": 2.4228816868410776,
+      "grad_norm": 2.9962964057922363,
+      "learning_rate": 9.24104119637535e-07,
+      "loss": 0.3325,
+      "step": 18615
+    },
+    {
+      "epoch": 2.4232721593127686,
+      "grad_norm": 2.4860761165618896,
+      "learning_rate": 9.228940195893932e-07,
+      "loss": 0.3338,
+      "step": 18618
+    },
+    {
+      "epoch": 2.423662631784459,
+      "grad_norm": 2.7918808460235596,
+      "learning_rate": 9.216846318030908e-07,
+      "loss": 0.3377,
+      "step": 18621
+    },
+    {
+      "epoch": 2.42405310425615,
+      "grad_norm": 3.0834054946899414,
+      "learning_rate": 9.204759564899029e-07,
+      "loss": 0.3111,
+      "step": 18624
+    },
+    {
+      "epoch": 2.4244435767278407,
+      "grad_norm": 3.3791403770446777,
+      "learning_rate": 9.192679938609827e-07,
+      "loss": 0.3618,
+      "step": 18627
+    },
+    {
+      "epoch": 2.4248340491995313,
+      "grad_norm": 2.551917314529419,
+      "learning_rate": 9.180607441273604e-07,
+      "loss": 0.368,
+      "step": 18630
+    },
+    {
+      "epoch": 2.4252245216712223,
+      "grad_norm": 2.7034595012664795,
+      "learning_rate": 9.168542074999392e-07,
+      "loss": 0.3756,
+      "step": 18633
+    },
+    {
+      "epoch": 2.425614994142913,
+      "grad_norm": 2.543304443359375,
+      "learning_rate": 9.15648384189498e-07,
+      "loss": 0.3233,
+      "step": 18636
+    },
+    {
+      "epoch": 2.426005466614604,
+      "grad_norm": 3.1651837825775146,
+      "learning_rate": 9.144432744066905e-07,
+      "loss": 0.3206,
+      "step": 18639
+    },
+    {
+      "epoch": 2.4263959390862944,
+      "grad_norm": 2.559799909591675,
+      "learning_rate": 9.132388783620499e-07,
+      "loss": 0.3498,
+      "step": 18642
+    },
+    {
+      "epoch": 2.426786411557985,
+      "grad_norm": 2.8475115299224854,
+      "learning_rate": 9.120351962659796e-07,
+      "loss": 0.346,
+      "step": 18645
+    },
+    {
+      "epoch": 2.427176884029676,
+      "grad_norm": 2.8560378551483154,
+      "learning_rate": 9.108322283287596e-07,
+      "loss": 0.3304,
+      "step": 18648
+    },
+    {
+      "epoch": 2.427567356501367,
+      "grad_norm": 2.9788451194763184,
+      "learning_rate": 9.096299747605481e-07,
+      "loss": 0.3449,
+      "step": 18651
+    },
+    {
+      "epoch": 2.4279578289730575,
+      "grad_norm": 2.8146660327911377,
+      "learning_rate": 9.084284357713752e-07,
+      "loss": 0.4016,
+      "step": 18654
+    },
+    {
+      "epoch": 2.428348301444748,
+      "grad_norm": 2.616931200027466,
+      "learning_rate": 9.072276115711459e-07,
+      "loss": 0.3531,
+      "step": 18657
+    },
+    {
+      "epoch": 2.428738773916439,
+      "grad_norm": 2.9515371322631836,
+      "learning_rate": 9.06027502369643e-07,
+      "loss": 0.3594,
+      "step": 18660
+    },
+    {
+      "epoch": 2.4291292463881295,
+      "grad_norm": 2.725242853164673,
+      "learning_rate": 9.048281083765243e-07,
+      "loss": 0.3437,
+      "step": 18663
+    },
+    {
+      "epoch": 2.4295197188598205,
+      "grad_norm": 3.099229574203491,
+      "learning_rate": 9.036294298013199e-07,
+      "loss": 0.4032,
+      "step": 18666
+    },
+    {
+      "epoch": 2.429910191331511,
+      "grad_norm": 2.691160202026367,
+      "learning_rate": 9.024314668534356e-07,
+      "loss": 0.3516,
+      "step": 18669
+    },
+    {
+      "epoch": 2.4303006638032016,
+      "grad_norm": 2.8280410766601562,
+      "learning_rate": 9.012342197421548e-07,
+      "loss": 0.3906,
+      "step": 18672
+    },
+    {
+      "epoch": 2.4306911362748926,
+      "grad_norm": 2.8767683506011963,
+      "learning_rate": 9.000376886766337e-07,
+      "loss": 0.3625,
+      "step": 18675
+    },
+    {
+      "epoch": 2.431081608746583,
+      "grad_norm": 2.8653829097747803,
+      "learning_rate": 8.988418738659016e-07,
+      "loss": 0.3302,
+      "step": 18678
+    },
+    {
+      "epoch": 2.431472081218274,
+      "grad_norm": 2.558035135269165,
+      "learning_rate": 8.976467755188684e-07,
+      "loss": 0.3544,
+      "step": 18681
+    },
+    {
+      "epoch": 2.4318625536899647,
+      "grad_norm": 2.6624937057495117,
+      "learning_rate": 8.964523938443131e-07,
+      "loss": 0.3324,
+      "step": 18684
+    },
+    {
+      "epoch": 2.4322530261616557,
+      "grad_norm": 2.9672672748565674,
+      "learning_rate": 8.952587290508919e-07,
+      "loss": 0.3928,
+      "step": 18687
+    },
+    {
+      "epoch": 2.4326434986333463,
+      "grad_norm": 2.827082395553589,
+      "learning_rate": 8.940657813471349e-07,
+      "loss": 0.3936,
+      "step": 18690
+    },
+    {
+      "epoch": 2.4330339711050373,
+      "grad_norm": 2.829801321029663,
+      "learning_rate": 8.928735509414488e-07,
+      "loss": 0.3735,
+      "step": 18693
+    },
+    {
+      "epoch": 2.433424443576728,
+      "grad_norm": 2.6571168899536133,
+      "learning_rate": 8.916820380421138e-07,
+      "loss": 0.3167,
+      "step": 18696
+    },
+    {
+      "epoch": 2.4338149160484184,
+      "grad_norm": 3.000652551651001,
+      "learning_rate": 8.904912428572827e-07,
+      "loss": 0.3927,
+      "step": 18699
+    },
+    {
+      "epoch": 2.4342053885201094,
+      "grad_norm": 3.0371198654174805,
+      "learning_rate": 8.893011655949862e-07,
+      "loss": 0.4359,
+      "step": 18702
+    },
+    {
+      "epoch": 2.4345958609918,
+      "grad_norm": 2.5225956439971924,
+      "learning_rate": 8.881118064631294e-07,
+      "loss": 0.3617,
+      "step": 18705
+    },
+    {
+      "epoch": 2.434986333463491,
+      "grad_norm": 3.0411794185638428,
+      "learning_rate": 8.869231656694904e-07,
+      "loss": 0.4085,
+      "step": 18708
+    },
+    {
+      "epoch": 2.4353768059351815,
+      "grad_norm": 2.582329750061035,
+      "learning_rate": 8.857352434217203e-07,
+      "loss": 0.3344,
+      "step": 18711
+    },
+    {
+      "epoch": 2.4357672784068725,
+      "grad_norm": 2.818873643875122,
+      "learning_rate": 8.845480399273493e-07,
+      "loss": 0.3445,
+      "step": 18714
+    },
+    {
+      "epoch": 2.436157750878563,
+      "grad_norm": 2.471068859100342,
+      "learning_rate": 8.83361555393778e-07,
+      "loss": 0.3724,
+      "step": 18717
+    },
+    {
+      "epoch": 2.436548223350254,
+      "grad_norm": 2.8934266567230225,
+      "learning_rate": 8.821757900282812e-07,
+      "loss": 0.3259,
+      "step": 18720
+    },
+    {
+      "epoch": 2.4369386958219446,
+      "grad_norm": 3.3927173614501953,
+      "learning_rate": 8.809907440380134e-07,
+      "loss": 0.4121,
+      "step": 18723
+    },
+    {
+      "epoch": 2.437329168293635,
+      "grad_norm": 2.8776323795318604,
+      "learning_rate": 8.798064176299964e-07,
+      "loss": 0.3497,
+      "step": 18726
+    },
+    {
+      "epoch": 2.437719640765326,
+      "grad_norm": 2.814983606338501,
+      "learning_rate": 8.7862281101113e-07,
+      "loss": 0.3731,
+      "step": 18729
+    },
+    {
+      "epoch": 2.4381101132370167,
+      "grad_norm": 2.808781862258911,
+      "learning_rate": 8.774399243881898e-07,
+      "loss": 0.403,
+      "step": 18732
+    },
+    {
+      "epoch": 2.4385005857087076,
+      "grad_norm": 3.211045980453491,
+      "learning_rate": 8.762577579678222e-07,
+      "loss": 0.3331,
+      "step": 18735
+    },
+    {
+      "epoch": 2.438891058180398,
+      "grad_norm": 2.9599382877349854,
+      "learning_rate": 8.750763119565486e-07,
+      "loss": 0.3298,
+      "step": 18738
+    },
+    {
+      "epoch": 2.439281530652089,
+      "grad_norm": 2.901627779006958,
+      "learning_rate": 8.73895586560764e-07,
+      "loss": 0.4148,
+      "step": 18741
+    },
+    {
+      "epoch": 2.4396720031237797,
+      "grad_norm": 2.719869375228882,
+      "learning_rate": 8.727155819867423e-07,
+      "loss": 0.3681,
+      "step": 18744
+    },
+    {
+      "epoch": 2.4400624755954707,
+      "grad_norm": 2.781661033630371,
+      "learning_rate": 8.715362984406261e-07,
+      "loss": 0.3967,
+      "step": 18747
+    },
+    {
+      "epoch": 2.4404529480671613,
+      "grad_norm": 3.1544594764709473,
+      "learning_rate": 8.703577361284338e-07,
+      "loss": 0.3487,
+      "step": 18750
+    },
+    {
+      "epoch": 2.440843420538852,
+      "grad_norm": 3.0680270195007324,
+      "learning_rate": 8.691798952560559e-07,
+      "loss": 0.3621,
+      "step": 18753
+    },
+    {
+      "epoch": 2.441233893010543,
+      "grad_norm": 2.580583333969116,
+      "learning_rate": 8.680027760292614e-07,
+      "loss": 0.3696,
+      "step": 18756
+    },
+    {
+      "epoch": 2.4416243654822334,
+      "grad_norm": 2.498162269592285,
+      "learning_rate": 8.668263786536896e-07,
+      "loss": 0.3002,
+      "step": 18759
+    },
+    {
+      "epoch": 2.4420148379539244,
+      "grad_norm": 2.785461902618408,
+      "learning_rate": 8.656507033348538e-07,
+      "loss": 0.3497,
+      "step": 18762
+    },
+    {
+      "epoch": 2.442405310425615,
+      "grad_norm": 2.962099313735962,
+      "learning_rate": 8.644757502781437e-07,
+      "loss": 0.4551,
+      "step": 18765
+    },
+    {
+      "epoch": 2.442795782897306,
+      "grad_norm": 2.782252550125122,
+      "learning_rate": 8.633015196888201e-07,
+      "loss": 0.3868,
+      "step": 18768
+    },
+    {
+      "epoch": 2.4431862553689965,
+      "grad_norm": 2.9131968021392822,
+      "learning_rate": 8.621280117720171e-07,
+      "loss": 0.3787,
+      "step": 18771
+    },
+    {
+      "epoch": 2.4435767278406875,
+      "grad_norm": 2.526874303817749,
+      "learning_rate": 8.60955226732747e-07,
+      "loss": 0.306,
+      "step": 18774
+    },
+    {
+      "epoch": 2.443967200312378,
+      "grad_norm": 2.654900312423706,
+      "learning_rate": 8.59783164775892e-07,
+      "loss": 0.3161,
+      "step": 18777
+    },
+    {
+      "epoch": 2.4443576727840686,
+      "grad_norm": 2.776625633239746,
+      "learning_rate": 8.586118261062076e-07,
+      "loss": 0.3483,
+      "step": 18780
+    },
+    {
+      "epoch": 2.4447481452557596,
+      "grad_norm": 2.6517176628112793,
+      "learning_rate": 8.574412109283232e-07,
+      "loss": 0.3079,
+      "step": 18783
+    },
+    {
+      "epoch": 2.44513861772745,
+      "grad_norm": 2.642512321472168,
+      "learning_rate": 8.562713194467465e-07,
+      "loss": 0.3765,
+      "step": 18786
+    },
+    {
+      "epoch": 2.445529090199141,
+      "grad_norm": 2.8423049449920654,
+      "learning_rate": 8.551021518658536e-07,
+      "loss": 0.3594,
+      "step": 18789
+    },
+    {
+      "epoch": 2.4459195626708317,
+      "grad_norm": 3.162525177001953,
+      "learning_rate": 8.539337083898936e-07,
+      "loss": 0.3961,
+      "step": 18792
+    },
+    {
+      "epoch": 2.446310035142522,
+      "grad_norm": 2.591289520263672,
+      "learning_rate": 8.527659892229944e-07,
+      "loss": 0.4013,
+      "step": 18795
+    },
+    {
+      "epoch": 2.446700507614213,
+      "grad_norm": 2.7982068061828613,
+      "learning_rate": 8.515989945691522e-07,
+      "loss": 0.3741,
+      "step": 18798
+    },
+    {
+      "epoch": 2.4470909800859038,
+      "grad_norm": 2.731457471847534,
+      "learning_rate": 8.504327246322386e-07,
+      "loss": 0.3658,
+      "step": 18801
+    },
+    {
+      "epoch": 2.4474814525575947,
+      "grad_norm": 2.960292100906372,
+      "learning_rate": 8.492671796159968e-07,
+      "loss": 0.4037,
+      "step": 18804
+    },
+    {
+      "epoch": 2.4478719250292853,
+      "grad_norm": 2.7071995735168457,
+      "learning_rate": 8.48102359724049e-07,
+      "loss": 0.392,
+      "step": 18807
+    },
+    {
+      "epoch": 2.4482623975009763,
+      "grad_norm": 2.889734983444214,
+      "learning_rate": 8.46938265159884e-07,
+      "loss": 0.3443,
+      "step": 18810
+    },
+    {
+      "epoch": 2.448652869972667,
+      "grad_norm": 2.8918561935424805,
+      "learning_rate": 8.457748961268664e-07,
+      "loss": 0.3643,
+      "step": 18813
+    },
+    {
+      "epoch": 2.449043342444358,
+      "grad_norm": 3.017657518386841,
+      "learning_rate": 8.446122528282363e-07,
+      "loss": 0.3877,
+      "step": 18816
+    },
+    {
+      "epoch": 2.4494338149160484,
+      "grad_norm": 2.657172679901123,
+      "learning_rate": 8.434503354671042e-07,
+      "loss": 0.3669,
+      "step": 18819
+    },
+    {
+      "epoch": 2.449824287387739,
+      "grad_norm": 2.905705451965332,
+      "learning_rate": 8.422891442464531e-07,
+      "loss": 0.3495,
+      "step": 18822
+    },
+    {
+      "epoch": 2.45021475985943,
+      "grad_norm": 2.8718855381011963,
+      "learning_rate": 8.41128679369142e-07,
+      "loss": 0.4314,
+      "step": 18825
+    },
+    {
+      "epoch": 2.4506052323311205,
+      "grad_norm": 2.7882487773895264,
+      "learning_rate": 8.399689410379025e-07,
+      "loss": 0.3464,
+      "step": 18828
+    },
+    {
+      "epoch": 2.4509957048028115,
+      "grad_norm": 2.7789225578308105,
+      "learning_rate": 8.388099294553382e-07,
+      "loss": 0.3243,
+      "step": 18831
+    },
+    {
+      "epoch": 2.451386177274502,
+      "grad_norm": 2.7858150005340576,
+      "learning_rate": 8.376516448239236e-07,
+      "loss": 0.3557,
+      "step": 18834
+    },
+    {
+      "epoch": 2.451776649746193,
+      "grad_norm": 2.6816978454589844,
+      "learning_rate": 8.364940873460115e-07,
+      "loss": 0.382,
+      "step": 18837
+    },
+    {
+      "epoch": 2.4521671222178836,
+      "grad_norm": 3.012741804122925,
+      "learning_rate": 8.353372572238238e-07,
+      "loss": 0.3826,
+      "step": 18840
+    },
+    {
+      "epoch": 2.4525575946895746,
+      "grad_norm": 2.689615488052368,
+      "learning_rate": 8.341811546594564e-07,
+      "loss": 0.3332,
+      "step": 18843
+    },
+    {
+      "epoch": 2.452948067161265,
+      "grad_norm": 2.9826838970184326,
+      "learning_rate": 8.330257798548763e-07,
+      "loss": 0.3838,
+      "step": 18846
+    },
+    {
+      "epoch": 2.4533385396329557,
+      "grad_norm": 2.5051567554473877,
+      "learning_rate": 8.318711330119272e-07,
+      "loss": 0.3054,
+      "step": 18849
+    },
+    {
+      "epoch": 2.4537290121046467,
+      "grad_norm": 2.6066651344299316,
+      "learning_rate": 8.307172143323233e-07,
+      "loss": 0.3481,
+      "step": 18852
+    },
+    {
+      "epoch": 2.454119484576337,
+      "grad_norm": 2.808889865875244,
+      "learning_rate": 8.295640240176494e-07,
+      "loss": 0.3487,
+      "step": 18855
+    },
+    {
+      "epoch": 2.454509957048028,
+      "grad_norm": 2.8130900859832764,
+      "learning_rate": 8.28411562269369e-07,
+      "loss": 0.3552,
+      "step": 18858
+    },
+    {
+      "epoch": 2.4549004295197188,
+      "grad_norm": 2.8734047412872314,
+      "learning_rate": 8.272598292888124e-07,
+      "loss": 0.3521,
+      "step": 18861
+    },
+    {
+      "epoch": 2.4552909019914098,
+      "grad_norm": 2.4436731338500977,
+      "learning_rate": 8.261088252771848e-07,
+      "loss": 0.3438,
+      "step": 18864
+    },
+    {
+      "epoch": 2.4556813744631003,
+      "grad_norm": 2.721705675125122,
+      "learning_rate": 8.249585504355645e-07,
+      "loss": 0.4103,
+      "step": 18867
+    },
+    {
+      "epoch": 2.4560718469347913,
+      "grad_norm": 2.883448839187622,
+      "learning_rate": 8.238090049649033e-07,
+      "loss": 0.368,
+      "step": 18870
+    },
+    {
+      "epoch": 2.456462319406482,
+      "grad_norm": 2.8921525478363037,
+      "learning_rate": 8.226601890660241e-07,
+      "loss": 0.3569,
+      "step": 18873
+    },
+    {
+      "epoch": 2.4568527918781724,
+      "grad_norm": 2.69545316696167,
+      "learning_rate": 8.215121029396206e-07,
+      "loss": 0.3431,
+      "step": 18876
+    },
+    {
+      "epoch": 2.4572432643498634,
+      "grad_norm": 3.1997456550598145,
+      "learning_rate": 8.203647467862636e-07,
+      "loss": 0.286,
+      "step": 18879
+    },
+    {
+      "epoch": 2.457633736821554,
+      "grad_norm": 3.0184547901153564,
+      "learning_rate": 8.192181208063926e-07,
+      "loss": 0.3426,
+      "step": 18882
+    },
+    {
+      "epoch": 2.458024209293245,
+      "grad_norm": 2.5744235515594482,
+      "learning_rate": 8.180722252003198e-07,
+      "loss": 0.3675,
+      "step": 18885
+    },
+    {
+      "epoch": 2.4584146817649355,
+      "grad_norm": 2.549051523208618,
+      "learning_rate": 8.169270601682328e-07,
+      "loss": 0.39,
+      "step": 18888
+    },
+    {
+      "epoch": 2.4588051542366265,
+      "grad_norm": 2.9480321407318115,
+      "learning_rate": 8.157826259101886e-07,
+      "loss": 0.3437,
+      "step": 18891
+    },
+    {
+      "epoch": 2.459195626708317,
+      "grad_norm": 2.6580848693847656,
+      "learning_rate": 8.146389226261176e-07,
+      "loss": 0.2907,
+      "step": 18894
+    },
+    {
+      "epoch": 2.459586099180008,
+      "grad_norm": 2.8073930740356445,
+      "learning_rate": 8.134959505158208e-07,
+      "loss": 0.346,
+      "step": 18897
+    },
+    {
+      "epoch": 2.4599765716516986,
+      "grad_norm": 2.8346378803253174,
+      "learning_rate": 8.123537097789752e-07,
+      "loss": 0.3428,
+      "step": 18900
+    },
+    {
+      "epoch": 2.460367044123389,
+      "grad_norm": 2.9591760635375977,
+      "learning_rate": 8.112122006151268e-07,
+      "loss": 0.3972,
+      "step": 18903
+    },
+    {
+      "epoch": 2.46075751659508,
+      "grad_norm": 3.013507843017578,
+      "learning_rate": 8.100714232236945e-07,
+      "loss": 0.3298,
+      "step": 18906
+    },
+    {
+      "epoch": 2.4611479890667707,
+      "grad_norm": 2.7471742630004883,
+      "learning_rate": 8.089313778039698e-07,
+      "loss": 0.309,
+      "step": 18909
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 3.1585850715637207,
+      "learning_rate": 8.077920645551179e-07,
+      "loss": 0.361,
+      "step": 18912
+    },
+    {
+      "epoch": 2.4619289340101522,
+      "grad_norm": 2.589958906173706,
+      "learning_rate": 8.066534836761736e-07,
+      "loss": 0.3519,
+      "step": 18915
+    },
+    {
+      "epoch": 2.462319406481843,
+      "grad_norm": 3.804331064224243,
+      "learning_rate": 8.055156353660426e-07,
+      "loss": 0.3417,
+      "step": 18918
+    },
+    {
+      "epoch": 2.4627098789535338,
+      "grad_norm": 3.136734962463379,
+      "learning_rate": 8.043785198235076e-07,
+      "loss": 0.3533,
+      "step": 18921
+    },
+    {
+      "epoch": 2.4631003514252248,
+      "grad_norm": 2.8037362098693848,
+      "learning_rate": 8.032421372472188e-07,
+      "loss": 0.3386,
+      "step": 18924
+    },
+    {
+      "epoch": 2.4634908238969153,
+      "grad_norm": 2.6273999214172363,
+      "learning_rate": 8.021064878356987e-07,
+      "loss": 0.3358,
+      "step": 18927
+    },
+    {
+      "epoch": 2.463881296368606,
+      "grad_norm": 2.6162285804748535,
+      "learning_rate": 8.009715717873451e-07,
+      "loss": 0.366,
+      "step": 18930
+    },
+    {
+      "epoch": 2.464271768840297,
+      "grad_norm": 2.9784958362579346,
+      "learning_rate": 7.998373893004246e-07,
+      "loss": 0.3863,
+      "step": 18933
+    },
+    {
+      "epoch": 2.4646622413119874,
+      "grad_norm": 3.3731791973114014,
+      "learning_rate": 7.987039405730757e-07,
+      "loss": 0.3216,
+      "step": 18936
+    },
+    {
+      "epoch": 2.4650527137836784,
+      "grad_norm": 2.652818441390991,
+      "learning_rate": 7.975712258033108e-07,
+      "loss": 0.3587,
+      "step": 18939
+    },
+    {
+      "epoch": 2.465443186255369,
+      "grad_norm": 2.968871831893921,
+      "learning_rate": 7.964392451890119e-07,
+      "loss": 0.3334,
+      "step": 18942
+    },
+    {
+      "epoch": 2.4658336587270595,
+      "grad_norm": 2.5186870098114014,
+      "learning_rate": 7.953079989279344e-07,
+      "loss": 0.3153,
+      "step": 18945
+    },
+    {
+      "epoch": 2.4662241311987505,
+      "grad_norm": 2.8169445991516113,
+      "learning_rate": 7.941774872177027e-07,
+      "loss": 0.3013,
+      "step": 18948
+    },
+    {
+      "epoch": 2.466614603670441,
+      "grad_norm": 2.731722831726074,
+      "learning_rate": 7.930477102558159e-07,
+      "loss": 0.3281,
+      "step": 18951
+    },
+    {
+      "epoch": 2.467005076142132,
+      "grad_norm": 2.587888240814209,
+      "learning_rate": 7.919186682396457e-07,
+      "loss": 0.3223,
+      "step": 18954
+    },
+    {
+      "epoch": 2.4673955486138226,
+      "grad_norm": 2.6688907146453857,
+      "learning_rate": 7.907903613664314e-07,
+      "loss": 0.3775,
+      "step": 18957
+    },
+    {
+      "epoch": 2.4677860210855136,
+      "grad_norm": 2.8114142417907715,
+      "learning_rate": 7.896627898332848e-07,
+      "loss": 0.3064,
+      "step": 18960
+    },
+    {
+      "epoch": 2.468176493557204,
+      "grad_norm": 2.9139301776885986,
+      "learning_rate": 7.885359538371929e-07,
+      "loss": 0.4012,
+      "step": 18963
+    },
+    {
+      "epoch": 2.468566966028895,
+      "grad_norm": 2.7291030883789062,
+      "learning_rate": 7.874098535750103e-07,
+      "loss": 0.3384,
+      "step": 18966
+    },
+    {
+      "epoch": 2.4689574385005857,
+      "grad_norm": 2.8256447315216064,
+      "learning_rate": 7.862844892434629e-07,
+      "loss": 0.3605,
+      "step": 18969
+    },
+    {
+      "epoch": 2.4693479109722762,
+      "grad_norm": 2.81532621383667,
+      "learning_rate": 7.85159861039152e-07,
+      "loss": 0.3608,
+      "step": 18972
+    },
+    {
+      "epoch": 2.4697383834439672,
+      "grad_norm": 2.861802816390991,
+      "learning_rate": 7.84035969158547e-07,
+      "loss": 0.3409,
+      "step": 18975
+    },
+    {
+      "epoch": 2.470128855915658,
+      "grad_norm": 2.916384696960449,
+      "learning_rate": 7.829128137979875e-07,
+      "loss": 0.3839,
+      "step": 18978
+    },
+    {
+      "epoch": 2.470519328387349,
+      "grad_norm": 2.9499900341033936,
+      "learning_rate": 7.817903951536892e-07,
+      "loss": 0.3667,
+      "step": 18981
+    },
+    {
+      "epoch": 2.4709098008590393,
+      "grad_norm": 2.9499289989471436,
+      "learning_rate": 7.806687134217356e-07,
+      "loss": 0.3399,
+      "step": 18984
+    },
+    {
+      "epoch": 2.4713002733307303,
+      "grad_norm": 2.9317288398742676,
+      "learning_rate": 7.795477687980801e-07,
+      "loss": 0.3157,
+      "step": 18987
+    },
+    {
+      "epoch": 2.471690745802421,
+      "grad_norm": 3.2856903076171875,
+      "learning_rate": 7.784275614785519e-07,
+      "loss": 0.386,
+      "step": 18990
+    },
+    {
+      "epoch": 2.472081218274112,
+      "grad_norm": 3.034611225128174,
+      "learning_rate": 7.773080916588466e-07,
+      "loss": 0.3312,
+      "step": 18993
+    },
+    {
+      "epoch": 2.4724716907458024,
+      "grad_norm": 2.9121174812316895,
+      "learning_rate": 7.761893595345354e-07,
+      "loss": 0.4084,
+      "step": 18996
+    },
+    {
+      "epoch": 2.472862163217493,
+      "grad_norm": 2.7845377922058105,
+      "learning_rate": 7.750713653010567e-07,
+      "loss": 0.3619,
+      "step": 18999
+    },
+    {
+      "epoch": 2.473252635689184,
+      "grad_norm": 2.673917531967163,
+      "learning_rate": 7.73954109153724e-07,
+      "loss": 0.4123,
+      "step": 19002
+    },
+    {
+      "epoch": 2.4736431081608745,
+      "grad_norm": 2.777308225631714,
+      "learning_rate": 7.728375912877178e-07,
+      "loss": 0.3202,
+      "step": 19005
+    },
+    {
+      "epoch": 2.4740335806325655,
+      "grad_norm": 3.0712320804595947,
+      "learning_rate": 7.717218118980918e-07,
+      "loss": 0.3864,
+      "step": 19008
+    },
+    {
+      "epoch": 2.474424053104256,
+      "grad_norm": 3.155332088470459,
+      "learning_rate": 7.706067711797687e-07,
+      "loss": 0.3849,
+      "step": 19011
+    },
+    {
+      "epoch": 2.474814525575947,
+      "grad_norm": 2.9384818077087402,
+      "learning_rate": 7.694924693275468e-07,
+      "loss": 0.3452,
+      "step": 19014
+    },
+    {
+      "epoch": 2.4752049980476376,
+      "grad_norm": 2.8019471168518066,
+      "learning_rate": 7.683789065360908e-07,
+      "loss": 0.3621,
+      "step": 19017
+    },
+    {
+      "epoch": 2.4755954705193286,
+      "grad_norm": 3.059307098388672,
+      "learning_rate": 7.672660829999367e-07,
+      "loss": 0.3504,
+      "step": 19020
+    },
+    {
+      "epoch": 2.475985942991019,
+      "grad_norm": 3.4540011882781982,
+      "learning_rate": 7.661539989134947e-07,
+      "loss": 0.3184,
+      "step": 19023
+    },
+    {
+      "epoch": 2.4763764154627097,
+      "grad_norm": 2.706732988357544,
+      "learning_rate": 7.650426544710427e-07,
+      "loss": 0.3627,
+      "step": 19026
+    },
+    {
+      "epoch": 2.4767668879344007,
+      "grad_norm": 2.591989040374756,
+      "learning_rate": 7.63932049866728e-07,
+      "loss": 0.4024,
+      "step": 19029
+    },
+    {
+      "epoch": 2.4771573604060912,
+      "grad_norm": 2.8087947368621826,
+      "learning_rate": 7.628221852945744e-07,
+      "loss": 0.3468,
+      "step": 19032
+    },
+    {
+      "epoch": 2.4775478328777822,
+      "grad_norm": 2.681309700012207,
+      "learning_rate": 7.617130609484702e-07,
+      "loss": 0.3382,
+      "step": 19035
+    },
+    {
+      "epoch": 2.477938305349473,
+      "grad_norm": 3.0248937606811523,
+      "learning_rate": 7.606046770221792e-07,
+      "loss": 0.417,
+      "step": 19038
+    },
+    {
+      "epoch": 2.478328777821164,
+      "grad_norm": 2.839232921600342,
+      "learning_rate": 7.594970337093316e-07,
+      "loss": 0.3179,
+      "step": 19041
+    },
+    {
+      "epoch": 2.4787192502928543,
+      "grad_norm": 2.8499197959899902,
+      "learning_rate": 7.583901312034331e-07,
+      "loss": 0.367,
+      "step": 19044
+    },
+    {
+      "epoch": 2.4791097227645453,
+      "grad_norm": 2.664505958557129,
+      "learning_rate": 7.572839696978551e-07,
+      "loss": 0.3077,
+      "step": 19047
+    },
+    {
+      "epoch": 2.479500195236236,
+      "grad_norm": 2.8773906230926514,
+      "learning_rate": 7.561785493858409e-07,
+      "loss": 0.3265,
+      "step": 19050
+    },
+    {
+      "epoch": 2.4798906677079264,
+      "grad_norm": 3.14871883392334,
+      "learning_rate": 7.550738704605082e-07,
+      "loss": 0.3526,
+      "step": 19053
+    },
+    {
+      "epoch": 2.4802811401796174,
+      "grad_norm": 2.6195621490478516,
+      "learning_rate": 7.539699331148398e-07,
+      "loss": 0.3692,
+      "step": 19056
+    },
+    {
+      "epoch": 2.480671612651308,
+      "grad_norm": 2.551382541656494,
+      "learning_rate": 7.52866737541692e-07,
+      "loss": 0.2936,
+      "step": 19059
+    },
+    {
+      "epoch": 2.481062085122999,
+      "grad_norm": 2.6689510345458984,
+      "learning_rate": 7.517642839337896e-07,
+      "loss": 0.3484,
+      "step": 19062
+    },
+    {
+      "epoch": 2.4814525575946895,
+      "grad_norm": 2.526609420776367,
+      "learning_rate": 7.506625724837302e-07,
+      "loss": 0.403,
+      "step": 19065
+    },
+    {
+      "epoch": 2.48184303006638,
+      "grad_norm": 2.6223344802856445,
+      "learning_rate": 7.495616033839808e-07,
+      "loss": 0.3567,
+      "step": 19068
+    },
+    {
+      "epoch": 2.482233502538071,
+      "grad_norm": 2.91428804397583,
+      "learning_rate": 7.484613768268762e-07,
+      "loss": 0.3981,
+      "step": 19071
+    },
+    {
+      "epoch": 2.4826239750097616,
+      "grad_norm": 2.6756317615509033,
+      "learning_rate": 7.473618930046267e-07,
+      "loss": 0.3424,
+      "step": 19074
+    },
+    {
+      "epoch": 2.4830144474814526,
+      "grad_norm": 2.6484274864196777,
+      "learning_rate": 7.462631521093066e-07,
+      "loss": 0.302,
+      "step": 19077
+    },
+    {
+      "epoch": 2.483404919953143,
+      "grad_norm": 2.8635783195495605,
+      "learning_rate": 7.451651543328664e-07,
+      "loss": 0.3657,
+      "step": 19080
+    },
+    {
+      "epoch": 2.483795392424834,
+      "grad_norm": 2.7862300872802734,
+      "learning_rate": 7.440678998671219e-07,
+      "loss": 0.3287,
+      "step": 19083
+    },
+    {
+      "epoch": 2.4841858648965247,
+      "grad_norm": 2.5042202472686768,
+      "learning_rate": 7.429713889037632e-07,
+      "loss": 0.3273,
+      "step": 19086
+    },
+    {
+      "epoch": 2.4845763373682157,
+      "grad_norm": 2.9982264041900635,
+      "learning_rate": 7.418756216343475e-07,
+      "loss": 0.4191,
+      "step": 19089
+    },
+    {
+      "epoch": 2.4849668098399063,
+      "grad_norm": 2.7572524547576904,
+      "learning_rate": 7.407805982503019e-07,
+      "loss": 0.3332,
+      "step": 19092
+    },
+    {
+      "epoch": 2.485357282311597,
+      "grad_norm": 2.9605908393859863,
+      "learning_rate": 7.396863189429265e-07,
+      "loss": 0.3352,
+      "step": 19095
+    },
+    {
+      "epoch": 2.485747754783288,
+      "grad_norm": 2.8530290126800537,
+      "learning_rate": 7.385927839033891e-07,
+      "loss": 0.3067,
+      "step": 19098
+    },
+    {
+      "epoch": 2.4861382272549784,
+      "grad_norm": 2.7384355068206787,
+      "learning_rate": 7.374999933227261e-07,
+      "loss": 0.365,
+      "step": 19101
+    },
+    {
+      "epoch": 2.4865286997266693,
+      "grad_norm": 2.806218385696411,
+      "learning_rate": 7.36407947391849e-07,
+      "loss": 0.3437,
+      "step": 19104
+    },
+    {
+      "epoch": 2.48691917219836,
+      "grad_norm": 2.798231363296509,
+      "learning_rate": 7.353166463015338e-07,
+      "loss": 0.3763,
+      "step": 19107
+    },
+    {
+      "epoch": 2.487309644670051,
+      "grad_norm": 2.5547235012054443,
+      "learning_rate": 7.342260902424292e-07,
+      "loss": 0.3022,
+      "step": 19110
+    },
+    {
+      "epoch": 2.4877001171417414,
+      "grad_norm": 2.6578516960144043,
+      "learning_rate": 7.331362794050512e-07,
+      "loss": 0.3696,
+      "step": 19113
+    },
+    {
+      "epoch": 2.4880905896134324,
+      "grad_norm": 2.6646409034729004,
+      "learning_rate": 7.320472139797902e-07,
+      "loss": 0.3517,
+      "step": 19116
+    },
+    {
+      "epoch": 2.488481062085123,
+      "grad_norm": 2.4392144680023193,
+      "learning_rate": 7.309588941569018e-07,
+      "loss": 0.308,
+      "step": 19119
+    },
+    {
+      "epoch": 2.4888715345568135,
+      "grad_norm": 2.8057920932769775,
+      "learning_rate": 7.298713201265145e-07,
+      "loss": 0.4005,
+      "step": 19122
+    },
+    {
+      "epoch": 2.4892620070285045,
+      "grad_norm": 2.7063517570495605,
+      "learning_rate": 7.287844920786236e-07,
+      "loss": 0.3574,
+      "step": 19125
+    },
+    {
+      "epoch": 2.489652479500195,
+      "grad_norm": 2.536086082458496,
+      "learning_rate": 7.276984102030976e-07,
+      "loss": 0.3572,
+      "step": 19128
+    },
+    {
+      "epoch": 2.490042951971886,
+      "grad_norm": 2.7980449199676514,
+      "learning_rate": 7.266130746896722e-07,
+      "loss": 0.4331,
+      "step": 19131
+    },
+    {
+      "epoch": 2.4904334244435766,
+      "grad_norm": 2.7580485343933105,
+      "learning_rate": 7.25528485727951e-07,
+      "loss": 0.3337,
+      "step": 19134
+    },
+    {
+      "epoch": 2.4908238969152676,
+      "grad_norm": 2.91790771484375,
+      "learning_rate": 7.244446435074126e-07,
+      "loss": 0.3567,
+      "step": 19137
+    },
+    {
+      "epoch": 2.491214369386958,
+      "grad_norm": 2.8744025230407715,
+      "learning_rate": 7.233615482174005e-07,
+      "loss": 0.3997,
+      "step": 19140
+    },
+    {
+      "epoch": 2.491604841858649,
+      "grad_norm": 2.7797534465789795,
+      "learning_rate": 7.22279200047128e-07,
+      "loss": 0.3623,
+      "step": 19143
+    },
+    {
+      "epoch": 2.4919953143303397,
+      "grad_norm": 2.9023027420043945,
+      "learning_rate": 7.211975991856812e-07,
+      "loss": 0.3873,
+      "step": 19146
+    },
+    {
+      "epoch": 2.4923857868020303,
+      "grad_norm": 2.7867956161499023,
+      "learning_rate": 7.201167458220131e-07,
+      "loss": 0.4055,
+      "step": 19149
+    },
+    {
+      "epoch": 2.4927762592737213,
+      "grad_norm": 2.779937505722046,
+      "learning_rate": 7.190366401449444e-07,
+      "loss": 0.3783,
+      "step": 19152
+    },
+    {
+      "epoch": 2.493166731745412,
+      "grad_norm": 2.892500400543213,
+      "learning_rate": 7.179572823431702e-07,
+      "loss": 0.3472,
+      "step": 19155
+    },
+    {
+      "epoch": 2.493557204217103,
+      "grad_norm": 2.9717111587524414,
+      "learning_rate": 7.168786726052501e-07,
+      "loss": 0.3959,
+      "step": 19158
+    },
+    {
+      "epoch": 2.4939476766887934,
+      "grad_norm": 3.3151369094848633,
+      "learning_rate": 7.158008111196152e-07,
+      "loss": 0.3453,
+      "step": 19161
+    },
+    {
+      "epoch": 2.4943381491604844,
+      "grad_norm": 2.6525325775146484,
+      "learning_rate": 7.147236980745653e-07,
+      "loss": 0.2747,
+      "step": 19164
+    },
+    {
+      "epoch": 2.494728621632175,
+      "grad_norm": 2.505096197128296,
+      "learning_rate": 7.136473336582722e-07,
+      "loss": 0.3938,
+      "step": 19167
+    },
+    {
+      "epoch": 2.495119094103866,
+      "grad_norm": 3.1451778411865234,
+      "learning_rate": 7.125717180587721e-07,
+      "loss": 0.3658,
+      "step": 19170
+    },
+    {
+      "epoch": 2.4955095665755564,
+      "grad_norm": 3.026200294494629,
+      "learning_rate": 7.114968514639736e-07,
+      "loss": 0.3788,
+      "step": 19173
+    },
+    {
+      "epoch": 2.495900039047247,
+      "grad_norm": 2.9026551246643066,
+      "learning_rate": 7.104227340616527e-07,
+      "loss": 0.3708,
+      "step": 19176
+    },
+    {
+      "epoch": 2.496290511518938,
+      "grad_norm": 2.9777603149414062,
+      "learning_rate": 7.093493660394568e-07,
+      "loss": 0.3116,
+      "step": 19179
+    },
+    {
+      "epoch": 2.4966809839906285,
+      "grad_norm": 2.5679221153259277,
+      "learning_rate": 7.082767475849011e-07,
+      "loss": 0.3198,
+      "step": 19182
+    },
+    {
+      "epoch": 2.4970714564623195,
+      "grad_norm": 2.768268585205078,
+      "learning_rate": 7.072048788853675e-07,
+      "loss": 0.4067,
+      "step": 19185
+    },
+    {
+      "epoch": 2.49746192893401,
+      "grad_norm": 2.612443447113037,
+      "learning_rate": 7.061337601281121e-07,
+      "loss": 0.3287,
+      "step": 19188
+    },
+    {
+      "epoch": 2.497852401405701,
+      "grad_norm": 2.776230573654175,
+      "learning_rate": 7.050633915002559e-07,
+      "loss": 0.3972,
+      "step": 19191
+    },
+    {
+      "epoch": 2.4982428738773916,
+      "grad_norm": 2.519664764404297,
+      "learning_rate": 7.039937731887886e-07,
+      "loss": 0.3023,
+      "step": 19194
+    },
+    {
+      "epoch": 2.4986333463490826,
+      "grad_norm": 3.676971197128296,
+      "learning_rate": 7.029249053805731e-07,
+      "loss": 0.398,
+      "step": 19197
+    },
+    {
+      "epoch": 2.499023818820773,
+      "grad_norm": 2.587867021560669,
+      "learning_rate": 7.018567882623368e-07,
+      "loss": 0.3471,
+      "step": 19200
+    },
+    {
+      "epoch": 2.4994142912924637,
+      "grad_norm": 3.008563280105591,
+      "learning_rate": 7.007894220206763e-07,
+      "loss": 0.3319,
+      "step": 19203
+    },
+    {
+      "epoch": 2.4998047637641547,
+      "grad_norm": 3.0095622539520264,
+      "learning_rate": 6.997228068420597e-07,
+      "loss": 0.299,
+      "step": 19206
+    },
+    {
+      "epoch": 2.5001952362358453,
+      "grad_norm": 3.0652036666870117,
+      "learning_rate": 6.986569429128238e-07,
+      "loss": 0.4295,
+      "step": 19209
+    },
+    {
+      "epoch": 2.5005857087075363,
+      "grad_norm": 2.4899280071258545,
+      "learning_rate": 6.975918304191709e-07,
+      "loss": 0.4072,
+      "step": 19212
+    },
+    {
+      "epoch": 2.500976181179227,
+      "grad_norm": 2.737656593322754,
+      "learning_rate": 6.965274695471729e-07,
+      "loss": 0.392,
+      "step": 19215
+    },
+    {
+      "epoch": 2.5013666536509174,
+      "grad_norm": 2.7664875984191895,
+      "learning_rate": 6.954638604827741e-07,
+      "loss": 0.2942,
+      "step": 19218
+    },
+    {
+      "epoch": 2.5017571261226084,
+      "grad_norm": 2.665586233139038,
+      "learning_rate": 6.944010034117837e-07,
+      "loss": 0.3066,
+      "step": 19221
+    },
+    {
+      "epoch": 2.5021475985942994,
+      "grad_norm": 2.6148107051849365,
+      "learning_rate": 6.933388985198802e-07,
+      "loss": 0.3325,
+      "step": 19224
+    },
+    {
+      "epoch": 2.50253807106599,
+      "grad_norm": 2.9085710048675537,
+      "learning_rate": 6.922775459926101e-07,
+      "loss": 0.3846,
+      "step": 19227
+    },
+    {
+      "epoch": 2.5029285435376805,
+      "grad_norm": 2.714949369430542,
+      "learning_rate": 6.912169460153917e-07,
+      "loss": 0.3562,
+      "step": 19230
+    },
+    {
+      "epoch": 2.5033190160093715,
+      "grad_norm": 2.8644955158233643,
+      "learning_rate": 6.901570987735085e-07,
+      "loss": 0.2756,
+      "step": 19233
+    },
+    {
+      "epoch": 2.503709488481062,
+      "grad_norm": 2.9508283138275146,
+      "learning_rate": 6.890980044521123e-07,
+      "loss": 0.4106,
+      "step": 19236
+    },
+    {
+      "epoch": 2.504099960952753,
+      "grad_norm": 2.899592161178589,
+      "learning_rate": 6.880396632362268e-07,
+      "loss": 0.3553,
+      "step": 19239
+    },
+    {
+      "epoch": 2.5044904334244436,
+      "grad_norm": 2.7175285816192627,
+      "learning_rate": 6.869820753107415e-07,
+      "loss": 0.3397,
+      "step": 19242
+    },
+    {
+      "epoch": 2.504880905896134,
+      "grad_norm": 2.8348207473754883,
+      "learning_rate": 6.859252408604134e-07,
+      "loss": 0.4374,
+      "step": 19245
+    },
+    {
+      "epoch": 2.505271378367825,
+      "grad_norm": 2.8067331314086914,
+      "learning_rate": 6.848691600698698e-07,
+      "loss": 0.3312,
+      "step": 19248
+    },
+    {
+      "epoch": 2.505661850839516,
+      "grad_norm": 3.3372321128845215,
+      "learning_rate": 6.838138331236077e-07,
+      "loss": 0.3608,
+      "step": 19251
+    },
+    {
+      "epoch": 2.5060523233112066,
+      "grad_norm": 2.9257891178131104,
+      "learning_rate": 6.827592602059891e-07,
+      "loss": 0.3447,
+      "step": 19254
+    },
+    {
+      "epoch": 2.506442795782897,
+      "grad_norm": 3.0877225399017334,
+      "learning_rate": 6.817054415012441e-07,
+      "loss": 0.3567,
+      "step": 19257
+    },
+    {
+      "epoch": 2.506833268254588,
+      "grad_norm": 3.7706382274627686,
+      "learning_rate": 6.806523771934753e-07,
+      "loss": 0.3516,
+      "step": 19260
+    },
+    {
+      "epoch": 2.5072237407262787,
+      "grad_norm": 2.557373285293579,
+      "learning_rate": 6.796000674666498e-07,
+      "loss": 0.3081,
+      "step": 19263
+    },
+    {
+      "epoch": 2.5076142131979697,
+      "grad_norm": 2.9034101963043213,
+      "learning_rate": 6.785485125046037e-07,
+      "loss": 0.3248,
+      "step": 19266
+    },
+    {
+      "epoch": 2.5080046856696603,
+      "grad_norm": 2.5061423778533936,
+      "learning_rate": 6.774977124910398e-07,
+      "loss": 0.328,
+      "step": 19269
+    },
+    {
+      "epoch": 2.508395158141351,
+      "grad_norm": 2.783360481262207,
+      "learning_rate": 6.764476676095344e-07,
+      "loss": 0.3268,
+      "step": 19272
+    },
+    {
+      "epoch": 2.508785630613042,
+      "grad_norm": 2.8032476902008057,
+      "learning_rate": 6.753983780435253e-07,
+      "loss": 0.3779,
+      "step": 19275
+    },
+    {
+      "epoch": 2.5091761030847324,
+      "grad_norm": 3.029937267303467,
+      "learning_rate": 6.743498439763213e-07,
+      "loss": 0.356,
+      "step": 19278
+    },
+    {
+      "epoch": 2.5095665755564234,
+      "grad_norm": 2.802793264389038,
+      "learning_rate": 6.733020655911004e-07,
+      "loss": 0.3366,
+      "step": 19281
+    },
+    {
+      "epoch": 2.509957048028114,
+      "grad_norm": 2.936530590057373,
+      "learning_rate": 6.722550430709068e-07,
+      "loss": 0.3973,
+      "step": 19284
+    },
+    {
+      "epoch": 2.510347520499805,
+      "grad_norm": 2.8553144931793213,
+      "learning_rate": 6.71208776598652e-07,
+      "loss": 0.3852,
+      "step": 19287
+    },
+    {
+      "epoch": 2.5107379929714955,
+      "grad_norm": 2.726348400115967,
+      "learning_rate": 6.701632663571172e-07,
+      "loss": 0.3928,
+      "step": 19290
+    },
+    {
+      "epoch": 2.5111284654431865,
+      "grad_norm": 3.0208041667938232,
+      "learning_rate": 6.691185125289523e-07,
+      "loss": 0.4231,
+      "step": 19293
+    },
+    {
+      "epoch": 2.511518937914877,
+      "grad_norm": 2.8433890342712402,
+      "learning_rate": 6.680745152966722e-07,
+      "loss": 0.3642,
+      "step": 19296
+    },
+    {
+      "epoch": 2.5119094103865676,
+      "grad_norm": 2.9164557456970215,
+      "learning_rate": 6.670312748426605e-07,
+      "loss": 0.3276,
+      "step": 19299
+    },
+    {
+      "epoch": 2.5122998828582586,
+      "grad_norm": 2.7807259559631348,
+      "learning_rate": 6.659887913491709e-07,
+      "loss": 0.4351,
+      "step": 19302
+    },
+    {
+      "epoch": 2.512690355329949,
+      "grad_norm": 2.888089179992676,
+      "learning_rate": 6.64947064998322e-07,
+      "loss": 0.354,
+      "step": 19305
+    },
+    {
+      "epoch": 2.51308082780164,
+      "grad_norm": 2.8115479946136475,
+      "learning_rate": 6.639060959721e-07,
+      "loss": 0.3543,
+      "step": 19308
+    },
+    {
+      "epoch": 2.5134713002733307,
+      "grad_norm": 2.8642683029174805,
+      "learning_rate": 6.628658844523622e-07,
+      "loss": 0.303,
+      "step": 19311
+    },
+    {
+      "epoch": 2.513861772745021,
+      "grad_norm": 2.651221752166748,
+      "learning_rate": 6.618264306208305e-07,
+      "loss": 0.3848,
+      "step": 19314
+    },
+    {
+      "epoch": 2.514252245216712,
+      "grad_norm": 3.1669631004333496,
+      "learning_rate": 6.607877346590958e-07,
+      "loss": 0.3538,
+      "step": 19317
+    },
+    {
+      "epoch": 2.514642717688403,
+      "grad_norm": 2.709857225418091,
+      "learning_rate": 6.597497967486139e-07,
+      "loss": 0.4179,
+      "step": 19320
+    },
+    {
+      "epoch": 2.5150331901600937,
+      "grad_norm": 2.7595081329345703,
+      "learning_rate": 6.587126170707125e-07,
+      "loss": 0.3483,
+      "step": 19323
+    },
+    {
+      "epoch": 2.5154236626317843,
+      "grad_norm": 2.8310909271240234,
+      "learning_rate": 6.576761958065847e-07,
+      "loss": 0.3987,
+      "step": 19326
+    },
+    {
+      "epoch": 2.5158141351034753,
+      "grad_norm": 2.7588865756988525,
+      "learning_rate": 6.566405331372899e-07,
+      "loss": 0.3472,
+      "step": 19329
+    },
+    {
+      "epoch": 2.516204607575166,
+      "grad_norm": 2.9078924655914307,
+      "learning_rate": 6.556056292437563e-07,
+      "loss": 0.4267,
+      "step": 19332
+    },
+    {
+      "epoch": 2.516595080046857,
+      "grad_norm": 2.8975186347961426,
+      "learning_rate": 6.545714843067813e-07,
+      "loss": 0.392,
+      "step": 19335
+    },
+    {
+      "epoch": 2.5169855525185474,
+      "grad_norm": 2.952500581741333,
+      "learning_rate": 6.535380985070261e-07,
+      "loss": 0.3441,
+      "step": 19338
+    },
+    {
+      "epoch": 2.517376024990238,
+      "grad_norm": 2.9215312004089355,
+      "learning_rate": 6.525054720250207e-07,
+      "loss": 0.4316,
+      "step": 19341
+    },
+    {
+      "epoch": 2.517766497461929,
+      "grad_norm": 2.8318424224853516,
+      "learning_rate": 6.514736050411652e-07,
+      "loss": 0.3531,
+      "step": 19344
+    },
+    {
+      "epoch": 2.51815696993362,
+      "grad_norm": 3.1043033599853516,
+      "learning_rate": 6.504424977357221e-07,
+      "loss": 0.4214,
+      "step": 19347
+    },
+    {
+      "epoch": 2.5185474424053105,
+      "grad_norm": 2.7383644580841064,
+      "learning_rate": 6.49412150288824e-07,
+      "loss": 0.3557,
+      "step": 19350
+    },
+    {
+      "epoch": 2.518937914877001,
+      "grad_norm": 2.9225306510925293,
+      "learning_rate": 6.483825628804719e-07,
+      "loss": 0.3435,
+      "step": 19353
+    },
+    {
+      "epoch": 2.519328387348692,
+      "grad_norm": 2.5163514614105225,
+      "learning_rate": 6.473537356905313e-07,
+      "loss": 0.3363,
+      "step": 19356
+    },
+    {
+      "epoch": 2.5197188598203826,
+      "grad_norm": 2.8150081634521484,
+      "learning_rate": 6.463256688987357e-07,
+      "loss": 0.3409,
+      "step": 19359
+    },
+    {
+      "epoch": 2.5201093322920736,
+      "grad_norm": 2.6332247257232666,
+      "learning_rate": 6.452983626846876e-07,
+      "loss": 0.3493,
+      "step": 19362
+    },
+    {
+      "epoch": 2.520499804763764,
+      "grad_norm": 2.6498818397521973,
+      "learning_rate": 6.44271817227855e-07,
+      "loss": 0.3274,
+      "step": 19365
+    },
+    {
+      "epoch": 2.5208902772354547,
+      "grad_norm": 3.2936317920684814,
+      "learning_rate": 6.432460327075723e-07,
+      "loss": 0.3614,
+      "step": 19368
+    },
+    {
+      "epoch": 2.5212807497071457,
+      "grad_norm": 2.7875030040740967,
+      "learning_rate": 6.422210093030407e-07,
+      "loss": 0.4025,
+      "step": 19371
+    },
+    {
+      "epoch": 2.5216712221788367,
+      "grad_norm": 2.455900192260742,
+      "learning_rate": 6.41196747193334e-07,
+      "loss": 0.3627,
+      "step": 19374
+    },
+    {
+      "epoch": 2.522061694650527,
+      "grad_norm": 2.7226195335388184,
+      "learning_rate": 6.401732465573851e-07,
+      "loss": 0.335,
+      "step": 19377
+    },
+    {
+      "epoch": 2.5224521671222178,
+      "grad_norm": 2.6656506061553955,
+      "learning_rate": 6.391505075739984e-07,
+      "loss": 0.3006,
+      "step": 19380
+    },
+    {
+      "epoch": 2.5228426395939088,
+      "grad_norm": 2.610862970352173,
+      "learning_rate": 6.381285304218437e-07,
+      "loss": 0.3197,
+      "step": 19383
+    },
+    {
+      "epoch": 2.5232331120655993,
+      "grad_norm": 2.577313184738159,
+      "learning_rate": 6.371073152794593e-07,
+      "loss": 0.3171,
+      "step": 19386
+    },
+    {
+      "epoch": 2.5236235845372903,
+      "grad_norm": 2.949810028076172,
+      "learning_rate": 6.360868623252486e-07,
+      "loss": 0.3888,
+      "step": 19389
+    },
+    {
+      "epoch": 2.524014057008981,
+      "grad_norm": 3.948436975479126,
+      "learning_rate": 6.350671717374818e-07,
+      "loss": 0.3426,
+      "step": 19392
+    },
+    {
+      "epoch": 2.5244045294806714,
+      "grad_norm": 2.535118341445923,
+      "learning_rate": 6.340482436942991e-07,
+      "loss": 0.3528,
+      "step": 19395
+    },
+    {
+      "epoch": 2.5247950019523624,
+      "grad_norm": 2.7280428409576416,
+      "learning_rate": 6.330300783737031e-07,
+      "loss": 0.4199,
+      "step": 19398
+    },
+    {
+      "epoch": 2.525185474424053,
+      "grad_norm": 2.778590440750122,
+      "learning_rate": 6.320126759535645e-07,
+      "loss": 0.3757,
+      "step": 19401
+    },
+    {
+      "epoch": 2.525575946895744,
+      "grad_norm": 2.919978141784668,
+      "learning_rate": 6.309960366116242e-07,
+      "loss": 0.3648,
+      "step": 19404
+    },
+    {
+      "epoch": 2.5259664193674345,
+      "grad_norm": 2.7053189277648926,
+      "learning_rate": 6.299801605254846e-07,
+      "loss": 0.3384,
+      "step": 19407
+    },
+    {
+      "epoch": 2.5263568918391255,
+      "grad_norm": 2.8377797603607178,
+      "learning_rate": 6.289650478726167e-07,
+      "loss": 0.2978,
+      "step": 19410
+    },
+    {
+      "epoch": 2.526747364310816,
+      "grad_norm": 2.6428401470184326,
+      "learning_rate": 6.279506988303602e-07,
+      "loss": 0.3359,
+      "step": 19413
+    },
+    {
+      "epoch": 2.527137836782507,
+      "grad_norm": 2.710667610168457,
+      "learning_rate": 6.2693711357592e-07,
+      "loss": 0.3566,
+      "step": 19416
+    },
+    {
+      "epoch": 2.5275283092541976,
+      "grad_norm": 2.882359504699707,
+      "learning_rate": 6.259242922863662e-07,
+      "loss": 0.3227,
+      "step": 19419
+    },
+    {
+      "epoch": 2.527918781725888,
+      "grad_norm": 2.9712507724761963,
+      "learning_rate": 6.249122351386361e-07,
+      "loss": 0.4212,
+      "step": 19422
+    },
+    {
+      "epoch": 2.528309254197579,
+      "grad_norm": 2.5499699115753174,
+      "learning_rate": 6.239009423095355e-07,
+      "loss": 0.3462,
+      "step": 19425
+    },
+    {
+      "epoch": 2.5286997266692697,
+      "grad_norm": 3.2019617557525635,
+      "learning_rate": 6.228904139757347e-07,
+      "loss": 0.3946,
+      "step": 19428
+    },
+    {
+      "epoch": 2.5290901991409607,
+      "grad_norm": 2.812934398651123,
+      "learning_rate": 6.218806503137697e-07,
+      "loss": 0.3762,
+      "step": 19431
+    },
+    {
+      "epoch": 2.529480671612651,
+      "grad_norm": 2.723848581314087,
+      "learning_rate": 6.208716515000446e-07,
+      "loss": 0.377,
+      "step": 19434
+    },
+    {
+      "epoch": 2.529871144084342,
+      "grad_norm": 2.7258222103118896,
+      "learning_rate": 6.198634177108303e-07,
+      "loss": 0.3686,
+      "step": 19437
+    },
+    {
+      "epoch": 2.5302616165560328,
+      "grad_norm": 2.727672815322876,
+      "learning_rate": 6.188559491222628e-07,
+      "loss": 0.4371,
+      "step": 19440
+    },
+    {
+      "epoch": 2.5306520890277238,
+      "grad_norm": 2.907092809677124,
+      "learning_rate": 6.17849245910343e-07,
+      "loss": 0.3584,
+      "step": 19443
+    },
+    {
+      "epoch": 2.5310425614994143,
+      "grad_norm": 2.331328868865967,
+      "learning_rate": 6.168433082509423e-07,
+      "loss": 0.3078,
+      "step": 19446
+    },
+    {
+      "epoch": 2.531433033971105,
+      "grad_norm": 3.0786802768707275,
+      "learning_rate": 6.158381363197951e-07,
+      "loss": 0.384,
+      "step": 19449
+    },
+    {
+      "epoch": 2.531823506442796,
+      "grad_norm": 2.793458938598633,
+      "learning_rate": 6.148337302925011e-07,
+      "loss": 0.3512,
+      "step": 19452
+    },
+    {
+      "epoch": 2.5322139789144864,
+      "grad_norm": 2.8592000007629395,
+      "learning_rate": 6.138300903445299e-07,
+      "loss": 0.3117,
+      "step": 19455
+    },
+    {
+      "epoch": 2.5326044513861774,
+      "grad_norm": 2.7495779991149902,
+      "learning_rate": 6.128272166512156e-07,
+      "loss": 0.3858,
+      "step": 19458
+    },
+    {
+      "epoch": 2.532994923857868,
+      "grad_norm": 2.65899920463562,
+      "learning_rate": 6.11825109387757e-07,
+      "loss": 0.3549,
+      "step": 19461
+    },
+    {
+      "epoch": 2.5333853963295585,
+      "grad_norm": 2.677633285522461,
+      "learning_rate": 6.108237687292202e-07,
+      "loss": 0.3314,
+      "step": 19464
+    },
+    {
+      "epoch": 2.5337758688012495,
+      "grad_norm": 2.7956135272979736,
+      "learning_rate": 6.098231948505379e-07,
+      "loss": 0.4367,
+      "step": 19467
+    },
+    {
+      "epoch": 2.5341663412729405,
+      "grad_norm": 2.5864150524139404,
+      "learning_rate": 6.088233879265076e-07,
+      "loss": 0.3262,
+      "step": 19470
+    },
+    {
+      "epoch": 2.534556813744631,
+      "grad_norm": 2.944986343383789,
+      "learning_rate": 6.078243481317931e-07,
+      "loss": 0.4178,
+      "step": 19473
+    },
+    {
+      "epoch": 2.5349472862163216,
+      "grad_norm": 2.8859469890594482,
+      "learning_rate": 6.068260756409261e-07,
+      "loss": 0.443,
+      "step": 19476
+    },
+    {
+      "epoch": 2.5353377586880126,
+      "grad_norm": 2.6689913272857666,
+      "learning_rate": 6.058285706283023e-07,
+      "loss": 0.331,
+      "step": 19479
+    },
+    {
+      "epoch": 2.535728231159703,
+      "grad_norm": 2.97873592376709,
+      "learning_rate": 6.048318332681824e-07,
+      "loss": 0.3991,
+      "step": 19482
+    },
+    {
+      "epoch": 2.536118703631394,
+      "grad_norm": 2.8189284801483154,
+      "learning_rate": 6.038358637346947e-07,
+      "loss": 0.3456,
+      "step": 19485
+    },
+    {
+      "epoch": 2.5365091761030847,
+      "grad_norm": 2.695420026779175,
+      "learning_rate": 6.028406622018346e-07,
+      "loss": 0.3596,
+      "step": 19488
+    },
+    {
+      "epoch": 2.5368996485747752,
+      "grad_norm": 2.9843835830688477,
+      "learning_rate": 6.018462288434601e-07,
+      "loss": 0.3328,
+      "step": 19491
+    },
+    {
+      "epoch": 2.5372901210464662,
+      "grad_norm": 2.7242064476013184,
+      "learning_rate": 6.008525638332963e-07,
+      "loss": 0.4317,
+      "step": 19494
+    },
+    {
+      "epoch": 2.5376805935181572,
+      "grad_norm": 2.6902647018432617,
+      "learning_rate": 5.998596673449348e-07,
+      "loss": 0.3795,
+      "step": 19497
+    },
+    {
+      "epoch": 2.5380710659898478,
+      "grad_norm": 3.0032718181610107,
+      "learning_rate": 5.98867539551834e-07,
+      "loss": 0.3836,
+      "step": 19500
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 3.2121152877807617,
+      "learning_rate": 5.978761806273159e-07,
+      "loss": 0.3354,
+      "step": 19503
+    },
+    {
+      "epoch": 2.5388520109332293,
+      "grad_norm": 3.1183557510375977,
+      "learning_rate": 5.968855907445669e-07,
+      "loss": 0.2782,
+      "step": 19506
+    },
+    {
+      "epoch": 2.53924248340492,
+      "grad_norm": 2.6611621379852295,
+      "learning_rate": 5.958957700766432e-07,
+      "loss": 0.3565,
+      "step": 19509
+    },
+    {
+      "epoch": 2.539632955876611,
+      "grad_norm": 3.2292160987854004,
+      "learning_rate": 5.949067187964642e-07,
+      "loss": 0.3928,
+      "step": 19512
+    },
+    {
+      "epoch": 2.5400234283483014,
+      "grad_norm": 2.88360595703125,
+      "learning_rate": 5.939184370768131e-07,
+      "loss": 0.3521,
+      "step": 19515
+    },
+    {
+      "epoch": 2.540413900819992,
+      "grad_norm": 2.851783037185669,
+      "learning_rate": 5.929309250903425e-07,
+      "loss": 0.3978,
+      "step": 19518
+    },
+    {
+      "epoch": 2.540804373291683,
+      "grad_norm": 2.560197114944458,
+      "learning_rate": 5.91944183009569e-07,
+      "loss": 0.3088,
+      "step": 19521
+    },
+    {
+      "epoch": 2.541194845763374,
+      "grad_norm": 3.081719398498535,
+      "learning_rate": 5.909582110068724e-07,
+      "loss": 0.3519,
+      "step": 19524
+    },
+    {
+      "epoch": 2.5415853182350645,
+      "grad_norm": 2.6352970600128174,
+      "learning_rate": 5.899730092545014e-07,
+      "loss": 0.3859,
+      "step": 19527
+    },
+    {
+      "epoch": 2.541975790706755,
+      "grad_norm": 2.6711513996124268,
+      "learning_rate": 5.889885779245691e-07,
+      "loss": 0.3733,
+      "step": 19530
+    },
+    {
+      "epoch": 2.542366263178446,
+      "grad_norm": 2.8868939876556396,
+      "learning_rate": 5.880049171890523e-07,
+      "loss": 0.3391,
+      "step": 19533
+    },
+    {
+      "epoch": 2.5427567356501366,
+      "grad_norm": 2.6922616958618164,
+      "learning_rate": 5.870220272197941e-07,
+      "loss": 0.3535,
+      "step": 19536
+    },
+    {
+      "epoch": 2.5431472081218276,
+      "grad_norm": 2.6204869747161865,
+      "learning_rate": 5.860399081885043e-07,
+      "loss": 0.3502,
+      "step": 19539
+    },
+    {
+      "epoch": 2.543537680593518,
+      "grad_norm": 2.710123062133789,
+      "learning_rate": 5.85058560266758e-07,
+      "loss": 0.3269,
+      "step": 19542
+    },
+    {
+      "epoch": 2.5439281530652087,
+      "grad_norm": 2.685530424118042,
+      "learning_rate": 5.840779836259936e-07,
+      "loss": 0.3666,
+      "step": 19545
+    },
+    {
+      "epoch": 2.5443186255368997,
+      "grad_norm": 3.140651226043701,
+      "learning_rate": 5.830981784375145e-07,
+      "loss": 0.3366,
+      "step": 19548
+    },
+    {
+      "epoch": 2.5447090980085902,
+      "grad_norm": 2.9641339778900146,
+      "learning_rate": 5.821191448724934e-07,
+      "loss": 0.3882,
+      "step": 19551
+    },
+    {
+      "epoch": 2.5450995704802812,
+      "grad_norm": 2.5823094844818115,
+      "learning_rate": 5.811408831019633e-07,
+      "loss": 0.3802,
+      "step": 19554
+    },
+    {
+      "epoch": 2.545490042951972,
+      "grad_norm": 3.241447925567627,
+      "learning_rate": 5.801633932968237e-07,
+      "loss": 0.4017,
+      "step": 19557
+    },
+    {
+      "epoch": 2.545880515423663,
+      "grad_norm": 2.6996021270751953,
+      "learning_rate": 5.79186675627843e-07,
+      "loss": 0.3679,
+      "step": 19560
+    },
+    {
+      "epoch": 2.5462709878953533,
+      "grad_norm": 3.129182815551758,
+      "learning_rate": 5.782107302656497e-07,
+      "loss": 0.4384,
+      "step": 19563
+    },
+    {
+      "epoch": 2.5466614603670443,
+      "grad_norm": 2.9826695919036865,
+      "learning_rate": 5.772355573807386e-07,
+      "loss": 0.392,
+      "step": 19566
+    },
+    {
+      "epoch": 2.547051932838735,
+      "grad_norm": 2.7523341178894043,
+      "learning_rate": 5.762611571434729e-07,
+      "loss": 0.337,
+      "step": 19569
+    },
+    {
+      "epoch": 2.5474424053104254,
+      "grad_norm": 2.7614572048187256,
+      "learning_rate": 5.752875297240762e-07,
+      "loss": 0.3406,
+      "step": 19572
+    },
+    {
+      "epoch": 2.5478328777821164,
+      "grad_norm": 3.031256675720215,
+      "learning_rate": 5.743146752926404e-07,
+      "loss": 0.3894,
+      "step": 19575
+    },
+    {
+      "epoch": 2.548223350253807,
+      "grad_norm": 2.6525719165802,
+      "learning_rate": 5.733425940191196e-07,
+      "loss": 0.3594,
+      "step": 19578
+    },
+    {
+      "epoch": 2.548613822725498,
+      "grad_norm": 3.2900519371032715,
+      "learning_rate": 5.723712860733349e-07,
+      "loss": 0.4027,
+      "step": 19581
+    },
+    {
+      "epoch": 2.5490042951971885,
+      "grad_norm": 2.974825620651245,
+      "learning_rate": 5.714007516249731e-07,
+      "loss": 0.3328,
+      "step": 19584
+    },
+    {
+      "epoch": 2.549394767668879,
+      "grad_norm": 3.1361944675445557,
+      "learning_rate": 5.704309908435829e-07,
+      "loss": 0.3499,
+      "step": 19587
+    },
+    {
+      "epoch": 2.54978524014057,
+      "grad_norm": 2.7374889850616455,
+      "learning_rate": 5.69462003898581e-07,
+      "loss": 0.3019,
+      "step": 19590
+    },
+    {
+      "epoch": 2.550175712612261,
+      "grad_norm": 2.7288341522216797,
+      "learning_rate": 5.684937909592464e-07,
+      "loss": 0.2799,
+      "step": 19593
+    },
+    {
+      "epoch": 2.5505661850839516,
+      "grad_norm": 2.551090717315674,
+      "learning_rate": 5.675263521947244e-07,
+      "loss": 0.3312,
+      "step": 19596
+    },
+    {
+      "epoch": 2.550956657555642,
+      "grad_norm": 2.9384982585906982,
+      "learning_rate": 5.665596877740226e-07,
+      "loss": 0.346,
+      "step": 19599
+    },
+    {
+      "epoch": 2.551347130027333,
+      "grad_norm": 2.4525489807128906,
+      "learning_rate": 5.655937978660181e-07,
+      "loss": 0.267,
+      "step": 19602
+    },
+    {
+      "epoch": 2.5517376024990237,
+      "grad_norm": 3.117671251296997,
+      "learning_rate": 5.646286826394487e-07,
+      "loss": 0.393,
+      "step": 19605
+    },
+    {
+      "epoch": 2.5521280749707147,
+      "grad_norm": 2.629549264907837,
+      "learning_rate": 5.636643422629162e-07,
+      "loss": 0.341,
+      "step": 19608
+    },
+    {
+      "epoch": 2.5525185474424053,
+      "grad_norm": 2.5423882007598877,
+      "learning_rate": 5.62700776904892e-07,
+      "loss": 0.3133,
+      "step": 19611
+    },
+    {
+      "epoch": 2.552909019914096,
+      "grad_norm": 2.8479764461517334,
+      "learning_rate": 5.617379867337069e-07,
+      "loss": 0.3234,
+      "step": 19614
+    },
+    {
+      "epoch": 2.553299492385787,
+      "grad_norm": 2.5962421894073486,
+      "learning_rate": 5.607759719175581e-07,
+      "loss": 0.3422,
+      "step": 19617
+    },
+    {
+      "epoch": 2.553689964857478,
+      "grad_norm": 2.9569382667541504,
+      "learning_rate": 5.59814732624509e-07,
+      "loss": 0.3549,
+      "step": 19620
+    },
+    {
+      "epoch": 2.5540804373291683,
+      "grad_norm": 2.6492676734924316,
+      "learning_rate": 5.588542690224847e-07,
+      "loss": 0.3782,
+      "step": 19623
+    },
+    {
+      "epoch": 2.554470909800859,
+      "grad_norm": 2.628129482269287,
+      "learning_rate": 5.578945812792774e-07,
+      "loss": 0.2951,
+      "step": 19626
+    },
+    {
+      "epoch": 2.55486138227255,
+      "grad_norm": 2.4740099906921387,
+      "learning_rate": 5.569356695625411e-07,
+      "loss": 0.2915,
+      "step": 19629
+    },
+    {
+      "epoch": 2.5552518547442404,
+      "grad_norm": 2.6669182777404785,
+      "learning_rate": 5.559775340397972e-07,
+      "loss": 0.353,
+      "step": 19632
+    },
+    {
+      "epoch": 2.5556423272159314,
+      "grad_norm": 3.07163667678833,
+      "learning_rate": 5.550201748784295e-07,
+      "loss": 0.3871,
+      "step": 19635
+    },
+    {
+      "epoch": 2.556032799687622,
+      "grad_norm": 3.0605452060699463,
+      "learning_rate": 5.54063592245686e-07,
+      "loss": 0.3861,
+      "step": 19638
+    },
+    {
+      "epoch": 2.5564232721593125,
+      "grad_norm": 3.2628283500671387,
+      "learning_rate": 5.531077863086798e-07,
+      "loss": 0.3629,
+      "step": 19641
+    },
+    {
+      "epoch": 2.5568137446310035,
+      "grad_norm": 2.872985601425171,
+      "learning_rate": 5.521527572343888e-07,
+      "loss": 0.3232,
+      "step": 19644
+    },
+    {
+      "epoch": 2.5572042171026945,
+      "grad_norm": 2.832007884979248,
+      "learning_rate": 5.511985051896546e-07,
+      "loss": 0.3507,
+      "step": 19647
+    },
+    {
+      "epoch": 2.557594689574385,
+      "grad_norm": 3.344686985015869,
+      "learning_rate": 5.502450303411816e-07,
+      "loss": 0.2877,
+      "step": 19650
+    },
+    {
+      "epoch": 2.5579851620460756,
+      "grad_norm": 2.880563974380493,
+      "learning_rate": 5.492923328555416e-07,
+      "loss": 0.3184,
+      "step": 19653
+    },
+    {
+      "epoch": 2.5583756345177666,
+      "grad_norm": 2.620983839035034,
+      "learning_rate": 5.483404128991682e-07,
+      "loss": 0.3796,
+      "step": 19656
+    },
+    {
+      "epoch": 2.558766106989457,
+      "grad_norm": 2.6286096572875977,
+      "learning_rate": 5.473892706383587e-07,
+      "loss": 0.3424,
+      "step": 19659
+    },
+    {
+      "epoch": 2.559156579461148,
+      "grad_norm": 2.9008336067199707,
+      "learning_rate": 5.464389062392783e-07,
+      "loss": 0.3349,
+      "step": 19662
+    },
+    {
+      "epoch": 2.5595470519328387,
+      "grad_norm": 2.709076404571533,
+      "learning_rate": 5.454893198679507e-07,
+      "loss": 0.3707,
+      "step": 19665
+    },
+    {
+      "epoch": 2.5599375244045293,
+      "grad_norm": 3.218621253967285,
+      "learning_rate": 5.445405116902686e-07,
+      "loss": 0.3714,
+      "step": 19668
+    },
+    {
+      "epoch": 2.5603279968762203,
+      "grad_norm": 2.944977283477783,
+      "learning_rate": 5.435924818719857e-07,
+      "loss": 0.3407,
+      "step": 19671
+    },
+    {
+      "epoch": 2.560718469347911,
+      "grad_norm": 2.735933303833008,
+      "learning_rate": 5.426452305787222e-07,
+      "loss": 0.3715,
+      "step": 19674
+    },
+    {
+      "epoch": 2.561108941819602,
+      "grad_norm": 2.785769462585449,
+      "learning_rate": 5.416987579759597e-07,
+      "loss": 0.3397,
+      "step": 19677
+    },
+    {
+      "epoch": 2.5614994142912924,
+      "grad_norm": 3.0471670627593994,
+      "learning_rate": 5.407530642290442e-07,
+      "loss": 0.3995,
+      "step": 19680
+    },
+    {
+      "epoch": 2.5618898867629833,
+      "grad_norm": 2.528057336807251,
+      "learning_rate": 5.398081495031893e-07,
+      "loss": 0.3123,
+      "step": 19683
+    },
+    {
+      "epoch": 2.562280359234674,
+      "grad_norm": 2.9553000926971436,
+      "learning_rate": 5.388640139634671e-07,
+      "loss": 0.3747,
+      "step": 19686
+    },
+    {
+      "epoch": 2.562670831706365,
+      "grad_norm": 2.6886420249938965,
+      "learning_rate": 5.37920657774817e-07,
+      "loss": 0.3329,
+      "step": 19689
+    },
+    {
+      "epoch": 2.5630613041780554,
+      "grad_norm": 2.6885476112365723,
+      "learning_rate": 5.369780811020403e-07,
+      "loss": 0.318,
+      "step": 19692
+    },
+    {
+      "epoch": 2.563451776649746,
+      "grad_norm": 2.7424001693725586,
+      "learning_rate": 5.360362841098043e-07,
+      "loss": 0.3761,
+      "step": 19695
+    },
+    {
+      "epoch": 2.563842249121437,
+      "grad_norm": 2.803973913192749,
+      "learning_rate": 5.350952669626397e-07,
+      "loss": 0.3694,
+      "step": 19698
+    },
+    {
+      "epoch": 2.5642327215931275,
+      "grad_norm": 2.916654109954834,
+      "learning_rate": 5.341550298249376e-07,
+      "loss": 0.4022,
+      "step": 19701
+    },
+    {
+      "epoch": 2.5646231940648185,
+      "grad_norm": 3.0839455127716064,
+      "learning_rate": 5.332155728609578e-07,
+      "loss": 0.3461,
+      "step": 19704
+    },
+    {
+      "epoch": 2.565013666536509,
+      "grad_norm": 2.8296546936035156,
+      "learning_rate": 5.322768962348201e-07,
+      "loss": 0.3642,
+      "step": 19707
+    },
+    {
+      "epoch": 2.5654041390082,
+      "grad_norm": 2.6688482761383057,
+      "learning_rate": 5.313390001105106e-07,
+      "loss": 0.3447,
+      "step": 19710
+    },
+    {
+      "epoch": 2.5657946114798906,
+      "grad_norm": 2.6866798400878906,
+      "learning_rate": 5.304018846518765e-07,
+      "loss": 0.3273,
+      "step": 19713
+    },
+    {
+      "epoch": 2.5661850839515816,
+      "grad_norm": 2.8025333881378174,
+      "learning_rate": 5.294655500226315e-07,
+      "loss": 0.35,
+      "step": 19716
+    },
+    {
+      "epoch": 2.566575556423272,
+      "grad_norm": 2.8305671215057373,
+      "learning_rate": 5.285299963863499e-07,
+      "loss": 0.3589,
+      "step": 19719
+    },
+    {
+      "epoch": 2.5669660288949627,
+      "grad_norm": 2.790302276611328,
+      "learning_rate": 5.275952239064708e-07,
+      "loss": 0.2947,
+      "step": 19722
+    },
+    {
+      "epoch": 2.5673565013666537,
+      "grad_norm": 2.739863872528076,
+      "learning_rate": 5.266612327462978e-07,
+      "loss": 0.3707,
+      "step": 19725
+    },
+    {
+      "epoch": 2.5677469738383443,
+      "grad_norm": 2.8916966915130615,
+      "learning_rate": 5.257280230689976e-07,
+      "loss": 0.373,
+      "step": 19728
+    },
+    {
+      "epoch": 2.5681374463100353,
+      "grad_norm": 2.925980567932129,
+      "learning_rate": 5.247955950375977e-07,
+      "loss": 0.3426,
+      "step": 19731
+    },
+    {
+      "epoch": 2.568527918781726,
+      "grad_norm": 2.7908196449279785,
+      "learning_rate": 5.238639488149944e-07,
+      "loss": 0.3189,
+      "step": 19734
+    },
+    {
+      "epoch": 2.5689183912534164,
+      "grad_norm": 2.8355512619018555,
+      "learning_rate": 5.229330845639424e-07,
+      "loss": 0.3587,
+      "step": 19737
+    },
+    {
+      "epoch": 2.5693088637251074,
+      "grad_norm": 2.771878480911255,
+      "learning_rate": 5.220030024470623e-07,
+      "loss": 0.3447,
+      "step": 19740
+    },
+    {
+      "epoch": 2.5696993361967984,
+      "grad_norm": 2.8163154125213623,
+      "learning_rate": 5.210737026268364e-07,
+      "loss": 0.3142,
+      "step": 19743
+    },
+    {
+      "epoch": 2.570089808668489,
+      "grad_norm": 2.9766600131988525,
+      "learning_rate": 5.201451852656137e-07,
+      "loss": 0.396,
+      "step": 19746
+    },
+    {
+      "epoch": 2.5704802811401795,
+      "grad_norm": 2.753567934036255,
+      "learning_rate": 5.192174505256014e-07,
+      "loss": 0.3244,
+      "step": 19749
+    },
+    {
+      "epoch": 2.5708707536118705,
+      "grad_norm": 2.696614980697632,
+      "learning_rate": 5.182904985688758e-07,
+      "loss": 0.3593,
+      "step": 19752
+    },
+    {
+      "epoch": 2.571261226083561,
+      "grad_norm": 3.015610456466675,
+      "learning_rate": 5.173643295573704e-07,
+      "loss": 0.3158,
+      "step": 19755
+    },
+    {
+      "epoch": 2.571651698555252,
+      "grad_norm": 2.938066244125366,
+      "learning_rate": 5.164389436528877e-07,
+      "loss": 0.3686,
+      "step": 19758
+    },
+    {
+      "epoch": 2.5720421710269425,
+      "grad_norm": 2.7356908321380615,
+      "learning_rate": 5.155143410170899e-07,
+      "loss": 0.3215,
+      "step": 19761
+    },
+    {
+      "epoch": 2.572432643498633,
+      "grad_norm": 2.872227668762207,
+      "learning_rate": 5.14590521811501e-07,
+      "loss": 0.4187,
+      "step": 19764
+    },
+    {
+      "epoch": 2.572823115970324,
+      "grad_norm": 2.6312549114227295,
+      "learning_rate": 5.136674861975138e-07,
+      "loss": 0.3705,
+      "step": 19767
+    },
+    {
+      "epoch": 2.573213588442015,
+      "grad_norm": 3.510446310043335,
+      "learning_rate": 5.127452343363787e-07,
+      "loss": 0.3801,
+      "step": 19770
+    },
+    {
+      "epoch": 2.5736040609137056,
+      "grad_norm": 2.8721954822540283,
+      "learning_rate": 5.118237663892101e-07,
+      "loss": 0.2959,
+      "step": 19773
+    },
+    {
+      "epoch": 2.573994533385396,
+      "grad_norm": 2.6207385063171387,
+      "learning_rate": 5.109030825169886e-07,
+      "loss": 0.3646,
+      "step": 19776
+    },
+    {
+      "epoch": 2.574385005857087,
+      "grad_norm": 2.5428993701934814,
+      "learning_rate": 5.099831828805552e-07,
+      "loss": 0.329,
+      "step": 19779
+    },
+    {
+      "epoch": 2.5747754783287777,
+      "grad_norm": 2.899235725402832,
+      "learning_rate": 5.090640676406134e-07,
+      "loss": 0.3592,
+      "step": 19782
+    },
+    {
+      "epoch": 2.5751659508004687,
+      "grad_norm": 2.9908111095428467,
+      "learning_rate": 5.081457369577319e-07,
+      "loss": 0.4394,
+      "step": 19785
+    },
+    {
+      "epoch": 2.5755564232721593,
+      "grad_norm": 2.572739601135254,
+      "learning_rate": 5.07228190992341e-07,
+      "loss": 0.3404,
+      "step": 19788
+    },
+    {
+      "epoch": 2.57594689574385,
+      "grad_norm": 2.801175117492676,
+      "learning_rate": 5.063114299047328e-07,
+      "loss": 0.3375,
+      "step": 19791
+    },
+    {
+      "epoch": 2.576337368215541,
+      "grad_norm": 2.8451387882232666,
+      "learning_rate": 5.053954538550643e-07,
+      "loss": 0.3875,
+      "step": 19794
+    },
+    {
+      "epoch": 2.576727840687232,
+      "grad_norm": 2.8498640060424805,
+      "learning_rate": 5.044802630033557e-07,
+      "loss": 0.4177,
+      "step": 19797
+    },
+    {
+      "epoch": 2.5771183131589224,
+      "grad_norm": 3.0200397968292236,
+      "learning_rate": 5.035658575094882e-07,
+      "loss": 0.3213,
+      "step": 19800
+    },
+    {
+      "epoch": 2.577508785630613,
+      "grad_norm": 3.0799405574798584,
+      "learning_rate": 5.02652237533206e-07,
+      "loss": 0.4103,
+      "step": 19803
+    },
+    {
+      "epoch": 2.577899258102304,
+      "grad_norm": 2.6357455253601074,
+      "learning_rate": 5.017394032341161e-07,
+      "loss": 0.2962,
+      "step": 19806
+    },
+    {
+      "epoch": 2.5782897305739945,
+      "grad_norm": 2.754474401473999,
+      "learning_rate": 5.008273547716902e-07,
+      "loss": 0.3561,
+      "step": 19809
+    },
+    {
+      "epoch": 2.5786802030456855,
+      "grad_norm": 2.776348829269409,
+      "learning_rate": 4.999160923052604e-07,
+      "loss": 0.3432,
+      "step": 19812
+    },
+    {
+      "epoch": 2.579070675517376,
+      "grad_norm": 2.6089048385620117,
+      "learning_rate": 4.990056159940221e-07,
+      "loss": 0.3669,
+      "step": 19815
+    },
+    {
+      "epoch": 2.5794611479890666,
+      "grad_norm": 2.7450947761535645,
+      "learning_rate": 4.980959259970347e-07,
+      "loss": 0.3924,
+      "step": 19818
+    },
+    {
+      "epoch": 2.5798516204607576,
+      "grad_norm": 2.4832842350006104,
+      "learning_rate": 4.971870224732184e-07,
+      "loss": 0.3488,
+      "step": 19821
+    },
+    {
+      "epoch": 2.580242092932448,
+      "grad_norm": 3.0447306632995605,
+      "learning_rate": 4.962789055813555e-07,
+      "loss": 0.3573,
+      "step": 19824
+    },
+    {
+      "epoch": 2.580632565404139,
+      "grad_norm": 2.752413272857666,
+      "learning_rate": 4.953715754800947e-07,
+      "loss": 0.3524,
+      "step": 19827
+    },
+    {
+      "epoch": 2.5810230378758297,
+      "grad_norm": 2.7045111656188965,
+      "learning_rate": 4.944650323279432e-07,
+      "loss": 0.3342,
+      "step": 19830
+    },
+    {
+      "epoch": 2.5814135103475206,
+      "grad_norm": 2.5593090057373047,
+      "learning_rate": 4.935592762832714e-07,
+      "loss": 0.3024,
+      "step": 19833
+    },
+    {
+      "epoch": 2.581803982819211,
+      "grad_norm": 2.776071786880493,
+      "learning_rate": 4.926543075043133e-07,
+      "loss": 0.3213,
+      "step": 19836
+    },
+    {
+      "epoch": 2.582194455290902,
+      "grad_norm": 2.6728298664093018,
+      "learning_rate": 4.917501261491675e-07,
+      "loss": 0.3532,
+      "step": 19839
+    },
+    {
+      "epoch": 2.5825849277625927,
+      "grad_norm": 2.9024364948272705,
+      "learning_rate": 4.908467323757898e-07,
+      "loss": 0.3839,
+      "step": 19842
+    },
+    {
+      "epoch": 2.5829754002342833,
+      "grad_norm": 2.7096471786499023,
+      "learning_rate": 4.899441263420019e-07,
+      "loss": 0.3769,
+      "step": 19845
+    },
+    {
+      "epoch": 2.5833658727059743,
+      "grad_norm": 2.882371425628662,
+      "learning_rate": 4.890423082054879e-07,
+      "loss": 0.3751,
+      "step": 19848
+    },
+    {
+      "epoch": 2.583756345177665,
+      "grad_norm": 2.866162061691284,
+      "learning_rate": 4.881412781237927e-07,
+      "loss": 0.3513,
+      "step": 19851
+    },
+    {
+      "epoch": 2.584146817649356,
+      "grad_norm": 2.852534770965576,
+      "learning_rate": 4.872410362543251e-07,
+      "loss": 0.3674,
+      "step": 19854
+    },
+    {
+      "epoch": 2.5845372901210464,
+      "grad_norm": 2.7616591453552246,
+      "learning_rate": 4.863415827543539e-07,
+      "loss": 0.3503,
+      "step": 19857
+    },
+    {
+      "epoch": 2.5849277625927374,
+      "grad_norm": 2.753648519515991,
+      "learning_rate": 4.854429177810138e-07,
+      "loss": 0.3549,
+      "step": 19860
+    },
+    {
+      "epoch": 2.585318235064428,
+      "grad_norm": 2.8216161727905273,
+      "learning_rate": 4.845450414912989e-07,
+      "loss": 0.3334,
+      "step": 19863
+    },
+    {
+      "epoch": 2.585708707536119,
+      "grad_norm": 2.91745662689209,
+      "learning_rate": 4.836479540420653e-07,
+      "loss": 0.41,
+      "step": 19866
+    },
+    {
+      "epoch": 2.5860991800078095,
+      "grad_norm": 3.008896589279175,
+      "learning_rate": 4.827516555900335e-07,
+      "loss": 0.3873,
+      "step": 19869
+    },
+    {
+      "epoch": 2.5864896524795,
+      "grad_norm": 2.9034063816070557,
+      "learning_rate": 4.818561462917848e-07,
+      "loss": 0.3723,
+      "step": 19872
+    },
+    {
+      "epoch": 2.586880124951191,
+      "grad_norm": 3.033580780029297,
+      "learning_rate": 4.809614263037621e-07,
+      "loss": 0.4597,
+      "step": 19875
+    },
+    {
+      "epoch": 2.5872705974228816,
+      "grad_norm": 2.697132110595703,
+      "learning_rate": 4.800674957822709e-07,
+      "loss": 0.3399,
+      "step": 19878
+    },
+    {
+      "epoch": 2.5876610698945726,
+      "grad_norm": 2.8508565425872803,
+      "learning_rate": 4.791743548834809e-07,
+      "loss": 0.4714,
+      "step": 19881
+    },
+    {
+      "epoch": 2.588051542366263,
+      "grad_norm": 2.757885694503784,
+      "learning_rate": 4.782820037634206e-07,
+      "loss": 0.4015,
+      "step": 19884
+    },
+    {
+      "epoch": 2.5884420148379537,
+      "grad_norm": 2.6383426189422607,
+      "learning_rate": 4.773904425779807e-07,
+      "loss": 0.3587,
+      "step": 19887
+    },
+    {
+      "epoch": 2.5888324873096447,
+      "grad_norm": 2.7818305492401123,
+      "learning_rate": 4.7649967148291787e-07,
+      "loss": 0.3431,
+      "step": 19890
+    },
+    {
+      "epoch": 2.5892229597813357,
+      "grad_norm": 2.781233549118042,
+      "learning_rate": 4.7560969063384587e-07,
+      "loss": 0.3586,
+      "step": 19893
+    },
+    {
+      "epoch": 2.589613432253026,
+      "grad_norm": 2.4520440101623535,
+      "learning_rate": 4.747205001862421e-07,
+      "loss": 0.3147,
+      "step": 19896
+    },
+    {
+      "epoch": 2.5900039047247168,
+      "grad_norm": 2.895876407623291,
+      "learning_rate": 4.7383210029544826e-07,
+      "loss": 0.4083,
+      "step": 19899
+    },
+    {
+      "epoch": 2.5903943771964077,
+      "grad_norm": 2.7189314365386963,
+      "learning_rate": 4.7294449111666474e-07,
+      "loss": 0.3006,
+      "step": 19902
+    },
+    {
+      "epoch": 2.5907848496680983,
+      "grad_norm": 2.5522122383117676,
+      "learning_rate": 4.720576728049553e-07,
+      "loss": 0.3353,
+      "step": 19905
+    },
+    {
+      "epoch": 2.5911753221397893,
+      "grad_norm": 2.8779947757720947,
+      "learning_rate": 4.711716455152437e-07,
+      "loss": 0.3466,
+      "step": 19908
+    },
+    {
+      "epoch": 2.59156579461148,
+      "grad_norm": 3.0736544132232666,
+      "learning_rate": 4.7028640940231964e-07,
+      "loss": 0.3948,
+      "step": 19911
+    },
+    {
+      "epoch": 2.5919562670831704,
+      "grad_norm": 2.9793858528137207,
+      "learning_rate": 4.694019646208309e-07,
+      "loss": 0.4169,
+      "step": 19914
+    },
+    {
+      "epoch": 2.5923467395548614,
+      "grad_norm": 2.964721441268921,
+      "learning_rate": 4.685183113252867e-07,
+      "loss": 0.3437,
+      "step": 19917
+    },
+    {
+      "epoch": 2.5927372120265524,
+      "grad_norm": 2.8925373554229736,
+      "learning_rate": 4.6763544967006027e-07,
+      "loss": 0.3583,
+      "step": 19920
+    },
+    {
+      "epoch": 2.593127684498243,
+      "grad_norm": 2.848095178604126,
+      "learning_rate": 4.667533798093876e-07,
+      "loss": 0.3217,
+      "step": 19923
+    },
+    {
+      "epoch": 2.5935181569699335,
+      "grad_norm": 2.6200132369995117,
+      "learning_rate": 4.6587210189736277e-07,
+      "loss": 0.4197,
+      "step": 19926
+    },
+    {
+      "epoch": 2.5939086294416245,
+      "grad_norm": 3.714834451675415,
+      "learning_rate": 4.6499161608794253e-07,
+      "loss": 0.3621,
+      "step": 19929
+    },
+    {
+      "epoch": 2.594299101913315,
+      "grad_norm": 2.842902898788452,
+      "learning_rate": 4.641119225349472e-07,
+      "loss": 0.4068,
+      "step": 19932
+    },
+    {
+      "epoch": 2.594689574385006,
+      "grad_norm": 2.6865339279174805,
+      "learning_rate": 4.632330213920572e-07,
+      "loss": 0.3629,
+      "step": 19935
+    },
+    {
+      "epoch": 2.5950800468566966,
+      "grad_norm": 2.490020751953125,
+      "learning_rate": 4.6235491281281354e-07,
+      "loss": 0.3555,
+      "step": 19938
+    },
+    {
+      "epoch": 2.595470519328387,
+      "grad_norm": 2.818798780441284,
+      "learning_rate": 4.614775969506219e-07,
+      "loss": 0.3819,
+      "step": 19941
+    },
+    {
+      "epoch": 2.595860991800078,
+      "grad_norm": 2.8724567890167236,
+      "learning_rate": 4.6060107395874575e-07,
+      "loss": 0.3519,
+      "step": 19944
+    },
+    {
+      "epoch": 2.596251464271769,
+      "grad_norm": 2.898418664932251,
+      "learning_rate": 4.597253439903121e-07,
+      "loss": 0.3509,
+      "step": 19947
+    },
+    {
+      "epoch": 2.5966419367434597,
+      "grad_norm": 2.461851119995117,
+      "learning_rate": 4.588504071983102e-07,
+      "loss": 0.3343,
+      "step": 19950
+    },
+    {
+      "epoch": 2.59703240921515,
+      "grad_norm": 3.4729630947113037,
+      "learning_rate": 4.579762637355889e-07,
+      "loss": 0.3653,
+      "step": 19953
+    },
+    {
+      "epoch": 2.597422881686841,
+      "grad_norm": 2.9097652435302734,
+      "learning_rate": 4.5710291375485995e-07,
+      "loss": 0.3598,
+      "step": 19956
+    },
+    {
+      "epoch": 2.5978133541585318,
+      "grad_norm": 2.9537763595581055,
+      "learning_rate": 4.5623035740869237e-07,
+      "loss": 0.3626,
+      "step": 19959
+    },
+    {
+      "epoch": 2.5982038266302228,
+      "grad_norm": 2.8022215366363525,
+      "learning_rate": 4.553585948495254e-07,
+      "loss": 0.3239,
+      "step": 19962
+    },
+    {
+      "epoch": 2.5985942991019133,
+      "grad_norm": 3.070570945739746,
+      "learning_rate": 4.544876262296505e-07,
+      "loss": 0.397,
+      "step": 19965
+    },
+    {
+      "epoch": 2.598984771573604,
+      "grad_norm": 2.5882253646850586,
+      "learning_rate": 4.536174517012254e-07,
+      "loss": 0.3869,
+      "step": 19968
+    },
+    {
+      "epoch": 2.599375244045295,
+      "grad_norm": 2.780194044113159,
+      "learning_rate": 4.527480714162663e-07,
+      "loss": 0.3459,
+      "step": 19971
+    },
+    {
+      "epoch": 2.5997657165169854,
+      "grad_norm": 2.8268415927886963,
+      "learning_rate": 4.5187948552665394e-07,
+      "loss": 0.3352,
+      "step": 19974
+    },
+    {
+      "epoch": 2.6001561889886764,
+      "grad_norm": 2.7509450912475586,
+      "learning_rate": 4.5101169418412804e-07,
+      "loss": 0.3166,
+      "step": 19977
+    },
+    {
+      "epoch": 2.600546661460367,
+      "grad_norm": 2.9784107208251953,
+      "learning_rate": 4.501446975402879e-07,
+      "loss": 0.3974,
+      "step": 19980
+    },
+    {
+      "epoch": 2.600937133932058,
+      "grad_norm": 2.658228874206543,
+      "learning_rate": 4.49278495746599e-07,
+      "loss": 0.2955,
+      "step": 19983
+    },
+    {
+      "epoch": 2.6013276064037485,
+      "grad_norm": 2.9209887981414795,
+      "learning_rate": 4.4841308895438363e-07,
+      "loss": 0.4038,
+      "step": 19986
+    },
+    {
+      "epoch": 2.6017180788754395,
+      "grad_norm": 3.115154981613159,
+      "learning_rate": 4.475484773148253e-07,
+      "loss": 0.4056,
+      "step": 19989
+    },
+    {
+      "epoch": 2.60210855134713,
+      "grad_norm": 3.18581223487854,
+      "learning_rate": 4.4668466097897214e-07,
+      "loss": 0.3335,
+      "step": 19992
+    },
+    {
+      "epoch": 2.6024990238188206,
+      "grad_norm": 2.865859270095825,
+      "learning_rate": 4.458216400977294e-07,
+      "loss": 0.3325,
+      "step": 19995
+    },
+    {
+      "epoch": 2.6028894962905116,
+      "grad_norm": 2.806417465209961,
+      "learning_rate": 4.449594148218661e-07,
+      "loss": 0.3596,
+      "step": 19998
+    },
+    {
+      "epoch": 2.603279968762202,
+      "grad_norm": 2.626926898956299,
+      "learning_rate": 4.4409798530200887e-07,
+      "loss": 0.3196,
+      "step": 20001
+    },
+    {
+      "epoch": 2.603670441233893,
+      "grad_norm": 2.9506349563598633,
+      "learning_rate": 4.4323735168865067e-07,
+      "loss": 0.3033,
+      "step": 20004
+    },
+    {
+      "epoch": 2.6040609137055837,
+      "grad_norm": 2.665252685546875,
+      "learning_rate": 4.423775141321418e-07,
+      "loss": 0.3409,
+      "step": 20007
+    },
+    {
+      "epoch": 2.6044513861772742,
+      "grad_norm": 2.7964937686920166,
+      "learning_rate": 4.4151847278269213e-07,
+      "loss": 0.4056,
+      "step": 20010
+    },
+    {
+      "epoch": 2.6048418586489652,
+      "grad_norm": 2.7492387294769287,
+      "learning_rate": 4.40660227790376e-07,
+      "loss": 0.3769,
+      "step": 20013
+    },
+    {
+      "epoch": 2.605232331120656,
+      "grad_norm": 2.866215944290161,
+      "learning_rate": 4.398027793051274e-07,
+      "loss": 0.3084,
+      "step": 20016
+    },
+    {
+      "epoch": 2.6056228035923468,
+      "grad_norm": 2.538553237915039,
+      "learning_rate": 4.389461274767398e-07,
+      "loss": 0.3312,
+      "step": 20019
+    },
+    {
+      "epoch": 2.6060132760640373,
+      "grad_norm": 2.998384475708008,
+      "learning_rate": 4.3809027245486745e-07,
+      "loss": 0.3473,
+      "step": 20022
+    },
+    {
+      "epoch": 2.6064037485357283,
+      "grad_norm": 3.2550208568573,
+      "learning_rate": 4.3723521438902907e-07,
+      "loss": 0.3076,
+      "step": 20025
+    },
+    {
+      "epoch": 2.606794221007419,
+      "grad_norm": 2.7280972003936768,
+      "learning_rate": 4.3638095342859953e-07,
+      "loss": 0.393,
+      "step": 20028
+    },
+    {
+      "epoch": 2.60718469347911,
+      "grad_norm": 2.8212859630584717,
+      "learning_rate": 4.3552748972281623e-07,
+      "loss": 0.3391,
+      "step": 20031
+    },
+    {
+      "epoch": 2.6075751659508004,
+      "grad_norm": 2.6648576259613037,
+      "learning_rate": 4.3467482342077927e-07,
+      "loss": 0.2934,
+      "step": 20034
+    },
+    {
+      "epoch": 2.607965638422491,
+      "grad_norm": 2.6012842655181885,
+      "learning_rate": 4.3382295467144675e-07,
+      "loss": 0.3,
+      "step": 20037
+    },
+    {
+      "epoch": 2.608356110894182,
+      "grad_norm": 2.910468578338623,
+      "learning_rate": 4.329718836236374e-07,
+      "loss": 0.3812,
+      "step": 20040
+    },
+    {
+      "epoch": 2.608746583365873,
+      "grad_norm": 3.135700225830078,
+      "learning_rate": 4.3212161042603174e-07,
+      "loss": 0.3906,
+      "step": 20043
+    },
+    {
+      "epoch": 2.6091370558375635,
+      "grad_norm": 2.9458134174346924,
+      "learning_rate": 4.312721352271726e-07,
+      "loss": 0.3705,
+      "step": 20046
+    },
+    {
+      "epoch": 2.609527528309254,
+      "grad_norm": 2.7890570163726807,
+      "learning_rate": 4.304234581754602e-07,
+      "loss": 0.3879,
+      "step": 20049
+    },
+    {
+      "epoch": 2.609918000780945,
+      "grad_norm": 2.8146069049835205,
+      "learning_rate": 4.2957557941915586e-07,
+      "loss": 0.3378,
+      "step": 20052
+    },
+    {
+      "epoch": 2.6103084732526356,
+      "grad_norm": 2.8290700912475586,
+      "learning_rate": 4.287284991063839e-07,
+      "loss": 0.3459,
+      "step": 20055
+    },
+    {
+      "epoch": 2.6106989457243266,
+      "grad_norm": 2.8539445400238037,
+      "learning_rate": 4.278822173851266e-07,
+      "loss": 0.3173,
+      "step": 20058
+    },
+    {
+      "epoch": 2.611089418196017,
+      "grad_norm": 2.866018056869507,
+      "learning_rate": 4.270367344032278e-07,
+      "loss": 0.3547,
+      "step": 20061
+    },
+    {
+      "epoch": 2.6114798906677077,
+      "grad_norm": 2.7992513179779053,
+      "learning_rate": 4.2619205030838993e-07,
+      "loss": 0.2944,
+      "step": 20064
+    },
+    {
+      "epoch": 2.6118703631393987,
+      "grad_norm": 2.9336869716644287,
+      "learning_rate": 4.2534816524818054e-07,
+      "loss": 0.327,
+      "step": 20067
+    },
+    {
+      "epoch": 2.6122608356110897,
+      "grad_norm": 2.6383185386657715,
+      "learning_rate": 4.245050793700228e-07,
+      "loss": 0.3384,
+      "step": 20070
+    },
+    {
+      "epoch": 2.6126513080827802,
+      "grad_norm": 2.8900146484375,
+      "learning_rate": 4.2366279282120113e-07,
+      "loss": 0.3804,
+      "step": 20073
+    },
+    {
+      "epoch": 2.613041780554471,
+      "grad_norm": 2.861504316329956,
+      "learning_rate": 4.2282130574886336e-07,
+      "loss": 0.3865,
+      "step": 20076
+    },
+    {
+      "epoch": 2.6134322530261618,
+      "grad_norm": 3.0964694023132324,
+      "learning_rate": 4.2198061830001467e-07,
+      "loss": 0.3979,
+      "step": 20079
+    },
+    {
+      "epoch": 2.6138227254978523,
+      "grad_norm": 2.622819423675537,
+      "learning_rate": 4.211407306215198e-07,
+      "loss": 0.3653,
+      "step": 20082
+    },
+    {
+      "epoch": 2.6142131979695433,
+      "grad_norm": 2.8246631622314453,
+      "learning_rate": 4.20301642860107e-07,
+      "loss": 0.3648,
+      "step": 20085
+    },
+    {
+      "epoch": 2.614603670441234,
+      "grad_norm": 2.9626944065093994,
+      "learning_rate": 4.194633551623645e-07,
+      "loss": 0.3551,
+      "step": 20088
+    },
+    {
+      "epoch": 2.6149941429129244,
+      "grad_norm": 3.4414031505584717,
+      "learning_rate": 4.186258676747368e-07,
+      "loss": 0.3116,
+      "step": 20091
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 2.694920539855957,
+      "learning_rate": 4.177891805435319e-07,
+      "loss": 0.3415,
+      "step": 20094
+    },
+    {
+      "epoch": 2.615775087856306,
+      "grad_norm": 2.726844549179077,
+      "learning_rate": 4.169532939149185e-07,
+      "loss": 0.349,
+      "step": 20097
+    },
+    {
+      "epoch": 2.616165560327997,
+      "grad_norm": 2.7306230068206787,
+      "learning_rate": 4.1611820793492294e-07,
+      "loss": 0.3459,
+      "step": 20100
+    },
+    {
+      "epoch": 2.6165560327996875,
+      "grad_norm": 2.7479240894317627,
+      "learning_rate": 4.152839227494332e-07,
+      "loss": 0.3669,
+      "step": 20103
+    },
+    {
+      "epoch": 2.6169465052713785,
+      "grad_norm": 2.936683416366577,
+      "learning_rate": 4.1445043850419754e-07,
+      "loss": 0.3581,
+      "step": 20106
+    },
+    {
+      "epoch": 2.617336977743069,
+      "grad_norm": 2.6035518646240234,
+      "learning_rate": 4.1361775534482397e-07,
+      "loss": 0.3312,
+      "step": 20109
+    },
+    {
+      "epoch": 2.61772745021476,
+      "grad_norm": 2.8025290966033936,
+      "learning_rate": 4.1278587341678e-07,
+      "loss": 0.3405,
+      "step": 20112
+    },
+    {
+      "epoch": 2.6181179226864506,
+      "grad_norm": 2.8133556842803955,
+      "learning_rate": 4.119547928653933e-07,
+      "loss": 0.3747,
+      "step": 20115
+    },
+    {
+      "epoch": 2.618508395158141,
+      "grad_norm": 2.7612462043762207,
+      "learning_rate": 4.111245138358527e-07,
+      "loss": 0.3516,
+      "step": 20118
+    },
+    {
+      "epoch": 2.618898867629832,
+      "grad_norm": 2.857440710067749,
+      "learning_rate": 4.1029503647320666e-07,
+      "loss": 0.3266,
+      "step": 20121
+    },
+    {
+      "epoch": 2.6192893401015227,
+      "grad_norm": 2.9601211547851562,
+      "learning_rate": 4.094663609223615e-07,
+      "loss": 0.3357,
+      "step": 20124
+    },
+    {
+      "epoch": 2.6196798125732137,
+      "grad_norm": 2.95589017868042,
+      "learning_rate": 4.086384873280863e-07,
+      "loss": 0.3328,
+      "step": 20127
+    },
+    {
+      "epoch": 2.6200702850449042,
+      "grad_norm": 3.003638744354248,
+      "learning_rate": 4.0781141583500825e-07,
+      "loss": 0.3574,
+      "step": 20130
+    },
+    {
+      "epoch": 2.6204607575165952,
+      "grad_norm": 2.928009033203125,
+      "learning_rate": 4.069851465876157e-07,
+      "loss": 0.4451,
+      "step": 20133
+    },
+    {
+      "epoch": 2.620851229988286,
+      "grad_norm": 3.0818943977355957,
+      "learning_rate": 4.061596797302553e-07,
+      "loss": 0.3478,
+      "step": 20136
+    },
+    {
+      "epoch": 2.621241702459977,
+      "grad_norm": 3.1439309120178223,
+      "learning_rate": 4.053350154071356e-07,
+      "loss": 0.3554,
+      "step": 20139
+    },
+    {
+      "epoch": 2.6216321749316673,
+      "grad_norm": 2.61098313331604,
+      "learning_rate": 4.045111537623231e-07,
+      "loss": 0.3411,
+      "step": 20142
+    },
+    {
+      "epoch": 2.622022647403358,
+      "grad_norm": 2.8817806243896484,
+      "learning_rate": 4.0368809493974314e-07,
+      "loss": 0.4183,
+      "step": 20145
+    },
+    {
+      "epoch": 2.622413119875049,
+      "grad_norm": 2.906568765640259,
+      "learning_rate": 4.028658390831852e-07,
+      "loss": 0.3663,
+      "step": 20148
+    },
+    {
+      "epoch": 2.6228035923467394,
+      "grad_norm": 2.7763681411743164,
+      "learning_rate": 4.0204438633629383e-07,
+      "loss": 0.395,
+      "step": 20151
+    },
+    {
+      "epoch": 2.6231940648184304,
+      "grad_norm": 2.8315131664276123,
+      "learning_rate": 4.0122373684257474e-07,
+      "loss": 0.3649,
+      "step": 20154
+    },
+    {
+      "epoch": 2.623584537290121,
+      "grad_norm": 3.1497440338134766,
+      "learning_rate": 4.0040389074539555e-07,
+      "loss": 0.3771,
+      "step": 20157
+    },
+    {
+      "epoch": 2.6239750097618115,
+      "grad_norm": 2.719172477722168,
+      "learning_rate": 3.9958484818797995e-07,
+      "loss": 0.3331,
+      "step": 20160
+    },
+    {
+      "epoch": 2.6243654822335025,
+      "grad_norm": 3.178786277770996,
+      "learning_rate": 3.9876660931341405e-07,
+      "loss": 0.3777,
+      "step": 20163
+    },
+    {
+      "epoch": 2.6247559547051935,
+      "grad_norm": 2.772921085357666,
+      "learning_rate": 3.9794917426464074e-07,
+      "loss": 0.3759,
+      "step": 20166
+    },
+    {
+      "epoch": 2.625146427176884,
+      "grad_norm": 2.7855470180511475,
+      "learning_rate": 3.971325431844664e-07,
+      "loss": 0.3807,
+      "step": 20169
+    },
+    {
+      "epoch": 2.6255368996485746,
+      "grad_norm": 2.967480182647705,
+      "learning_rate": 3.963167162155529e-07,
+      "loss": 0.3764,
+      "step": 20172
+    },
+    {
+      "epoch": 2.6259273721202656,
+      "grad_norm": 2.7995481491088867,
+      "learning_rate": 3.9550169350042464e-07,
+      "loss": 0.3978,
+      "step": 20175
+    },
+    {
+      "epoch": 2.626317844591956,
+      "grad_norm": 3.274163246154785,
+      "learning_rate": 3.946874751814639e-07,
+      "loss": 0.3355,
+      "step": 20178
+    },
+    {
+      "epoch": 2.626708317063647,
+      "grad_norm": 2.6372921466827393,
+      "learning_rate": 3.93874061400914e-07,
+      "loss": 0.3227,
+      "step": 20181
+    },
+    {
+      "epoch": 2.6270987895353377,
+      "grad_norm": 2.845742702484131,
+      "learning_rate": 3.9306145230087524e-07,
+      "loss": 0.3701,
+      "step": 20184
+    },
+    {
+      "epoch": 2.6274892620070283,
+      "grad_norm": 2.5727438926696777,
+      "learning_rate": 3.9224964802330847e-07,
+      "loss": 0.3382,
+      "step": 20187
+    },
+    {
+      "epoch": 2.6278797344787193,
+      "grad_norm": 3.014756679534912,
+      "learning_rate": 3.914386487100358e-07,
+      "loss": 0.3586,
+      "step": 20190
+    },
+    {
+      "epoch": 2.6282702069504102,
+      "grad_norm": 2.7406253814697266,
+      "learning_rate": 3.9062845450273613e-07,
+      "loss": 0.3508,
+      "step": 20193
+    },
+    {
+      "epoch": 2.628660679422101,
+      "grad_norm": 2.826164960861206,
+      "learning_rate": 3.898190655429479e-07,
+      "loss": 0.4258,
+      "step": 20196
+    },
+    {
+      "epoch": 2.6290511518937913,
+      "grad_norm": 3.219733238220215,
+      "learning_rate": 3.890104819720719e-07,
+      "loss": 0.3397,
+      "step": 20199
+    },
+    {
+      "epoch": 2.6294416243654823,
+      "grad_norm": 2.6495490074157715,
+      "learning_rate": 3.8820270393136403e-07,
+      "loss": 0.3296,
+      "step": 20202
+    },
+    {
+      "epoch": 2.629832096837173,
+      "grad_norm": 3.0028276443481445,
+      "learning_rate": 3.873957315619414e-07,
+      "loss": 0.3653,
+      "step": 20205
+    },
+    {
+      "epoch": 2.630222569308864,
+      "grad_norm": 2.973289728164673,
+      "learning_rate": 3.8658956500478127e-07,
+      "loss": 0.3419,
+      "step": 20208
+    },
+    {
+      "epoch": 2.6306130417805544,
+      "grad_norm": 3.0267601013183594,
+      "learning_rate": 3.857842044007193e-07,
+      "loss": 0.3816,
+      "step": 20211
+    },
+    {
+      "epoch": 2.631003514252245,
+      "grad_norm": 2.8618972301483154,
+      "learning_rate": 3.849796498904496e-07,
+      "loss": 0.3335,
+      "step": 20214
+    },
+    {
+      "epoch": 2.631393986723936,
+      "grad_norm": 3.1597700119018555,
+      "learning_rate": 3.841759016145258e-07,
+      "loss": 0.4617,
+      "step": 20217
+    },
+    {
+      "epoch": 2.631784459195627,
+      "grad_norm": 3.0783212184906006,
+      "learning_rate": 3.8337295971336285e-07,
+      "loss": 0.4516,
+      "step": 20220
+    },
+    {
+      "epoch": 2.6321749316673175,
+      "grad_norm": 2.588809013366699,
+      "learning_rate": 3.825708243272319e-07,
+      "loss": 0.3026,
+      "step": 20223
+    },
+    {
+      "epoch": 2.632565404139008,
+      "grad_norm": 2.936833381652832,
+      "learning_rate": 3.817694955962642e-07,
+      "loss": 0.3481,
+      "step": 20226
+    },
+    {
+      "epoch": 2.632955876610699,
+      "grad_norm": 2.49509596824646,
+      "learning_rate": 3.8096897366044936e-07,
+      "loss": 0.3365,
+      "step": 20229
+    },
+    {
+      "epoch": 2.6333463490823896,
+      "grad_norm": 2.5137100219726562,
+      "learning_rate": 3.801692586596384e-07,
+      "loss": 0.3061,
+      "step": 20232
+    },
+    {
+      "epoch": 2.6337368215540806,
+      "grad_norm": 3.77799916267395,
+      "learning_rate": 3.7937035073353947e-07,
+      "loss": 0.3653,
+      "step": 20235
+    },
+    {
+      "epoch": 2.634127294025771,
+      "grad_norm": 3.221705198287964,
+      "learning_rate": 3.785722500217193e-07,
+      "loss": 0.354,
+      "step": 20238
+    },
+    {
+      "epoch": 2.6345177664974617,
+      "grad_norm": 2.6136856079101562,
+      "learning_rate": 3.7777495666360574e-07,
+      "loss": 0.3911,
+      "step": 20241
+    },
+    {
+      "epoch": 2.6349082389691527,
+      "grad_norm": 2.5668489933013916,
+      "learning_rate": 3.76978470798483e-07,
+      "loss": 0.2889,
+      "step": 20244
+    },
+    {
+      "epoch": 2.6352987114408433,
+      "grad_norm": 2.760801076889038,
+      "learning_rate": 3.7618279256549597e-07,
+      "loss": 0.3479,
+      "step": 20247
+    },
+    {
+      "epoch": 2.6356891839125343,
+      "grad_norm": 2.920722723007202,
+      "learning_rate": 3.7538792210364825e-07,
+      "loss": 0.3856,
+      "step": 20250
+    },
+    {
+      "epoch": 2.636079656384225,
+      "grad_norm": 3.082993268966675,
+      "learning_rate": 3.745938595518017e-07,
+      "loss": 0.3859,
+      "step": 20253
+    },
+    {
+      "epoch": 2.636470128855916,
+      "grad_norm": 3.254218101501465,
+      "learning_rate": 3.7380060504867697e-07,
+      "loss": 0.3902,
+      "step": 20256
+    },
+    {
+      "epoch": 2.6368606013276064,
+      "grad_norm": 2.6879429817199707,
+      "learning_rate": 3.7300815873285435e-07,
+      "loss": 0.3567,
+      "step": 20259
+    },
+    {
+      "epoch": 2.6372510737992974,
+      "grad_norm": 2.662904977798462,
+      "learning_rate": 3.722165207427736e-07,
+      "loss": 0.3279,
+      "step": 20262
+    },
+    {
+      "epoch": 2.637641546270988,
+      "grad_norm": 2.596517562866211,
+      "learning_rate": 3.7142569121673135e-07,
+      "loss": 0.3851,
+      "step": 20265
+    },
+    {
+      "epoch": 2.6380320187426785,
+      "grad_norm": 2.8582632541656494,
+      "learning_rate": 3.706356702928826e-07,
+      "loss": 0.364,
+      "step": 20268
+    },
+    {
+      "epoch": 2.6384224912143694,
+      "grad_norm": 2.894286632537842,
+      "learning_rate": 3.6984645810924423e-07,
+      "loss": 0.3438,
+      "step": 20271
+    },
+    {
+      "epoch": 2.63881296368606,
+      "grad_norm": 2.674154758453369,
+      "learning_rate": 3.690580548036893e-07,
+      "loss": 0.3641,
+      "step": 20274
+    },
+    {
+      "epoch": 2.639203436157751,
+      "grad_norm": 2.6992299556732178,
+      "learning_rate": 3.6827046051395035e-07,
+      "loss": 0.3502,
+      "step": 20277
+    },
+    {
+      "epoch": 2.6395939086294415,
+      "grad_norm": 2.4692561626434326,
+      "learning_rate": 3.674836753776173e-07,
+      "loss": 0.3447,
+      "step": 20280
+    },
+    {
+      "epoch": 2.639984381101132,
+      "grad_norm": 2.7542686462402344,
+      "learning_rate": 3.666976995321414e-07,
+      "loss": 0.3577,
+      "step": 20283
+    },
+    {
+      "epoch": 2.640374853572823,
+      "grad_norm": 3.003816843032837,
+      "learning_rate": 3.6591253311483056e-07,
+      "loss": 0.2958,
+      "step": 20286
+    },
+    {
+      "epoch": 2.640765326044514,
+      "grad_norm": 2.7368361949920654,
+      "learning_rate": 3.6512817626285056e-07,
+      "loss": 0.3826,
+      "step": 20289
+    },
+    {
+      "epoch": 2.6411557985162046,
+      "grad_norm": 2.821338176727295,
+      "learning_rate": 3.6434462911322856e-07,
+      "loss": 0.3569,
+      "step": 20292
+    },
+    {
+      "epoch": 2.641546270987895,
+      "grad_norm": 2.715587615966797,
+      "learning_rate": 3.635618918028477e-07,
+      "loss": 0.369,
+      "step": 20295
+    },
+    {
+      "epoch": 2.641936743459586,
+      "grad_norm": 2.71569561958313,
+      "learning_rate": 3.627799644684504e-07,
+      "loss": 0.3227,
+      "step": 20298
+    },
+    {
+      "epoch": 2.6423272159312767,
+      "grad_norm": 3.0673129558563232,
+      "learning_rate": 3.6199884724663734e-07,
+      "loss": 0.3651,
+      "step": 20301
+    },
+    {
+      "epoch": 2.6427176884029677,
+      "grad_norm": 3.4038453102111816,
+      "learning_rate": 3.612185402738705e-07,
+      "loss": 0.3598,
+      "step": 20304
+    },
+    {
+      "epoch": 2.6431081608746583,
+      "grad_norm": 2.976976156234741,
+      "learning_rate": 3.604390436864652e-07,
+      "loss": 0.3942,
+      "step": 20307
+    },
+    {
+      "epoch": 2.643498633346349,
+      "grad_norm": 3.015279531478882,
+      "learning_rate": 3.596603576205987e-07,
+      "loss": 0.413,
+      "step": 20310
+    },
+    {
+      "epoch": 2.64388910581804,
+      "grad_norm": 2.8536202907562256,
+      "learning_rate": 3.5888248221230605e-07,
+      "loss": 0.3735,
+      "step": 20313
+    },
+    {
+      "epoch": 2.644279578289731,
+      "grad_norm": 2.8353049755096436,
+      "learning_rate": 3.5810541759748076e-07,
+      "loss": 0.3604,
+      "step": 20316
+    },
+    {
+      "epoch": 2.6446700507614214,
+      "grad_norm": 2.661177158355713,
+      "learning_rate": 3.5732916391187254e-07,
+      "loss": 0.3191,
+      "step": 20319
+    },
+    {
+      "epoch": 2.645060523233112,
+      "grad_norm": 2.5510129928588867,
+      "learning_rate": 3.5655372129109356e-07,
+      "loss": 0.3296,
+      "step": 20322
+    },
+    {
+      "epoch": 2.645450995704803,
+      "grad_norm": 2.8251776695251465,
+      "learning_rate": 3.557790898706115e-07,
+      "loss": 0.3705,
+      "step": 20325
+    },
+    {
+      "epoch": 2.6458414681764935,
+      "grad_norm": 2.7708895206451416,
+      "learning_rate": 3.55005269785752e-07,
+      "loss": 0.3361,
+      "step": 20328
+    },
+    {
+      "epoch": 2.6462319406481845,
+      "grad_norm": 2.593419313430786,
+      "learning_rate": 3.5423226117169973e-07,
+      "loss": 0.3692,
+      "step": 20331
+    },
+    {
+      "epoch": 2.646622413119875,
+      "grad_norm": 3.5748655796051025,
+      "learning_rate": 3.5346006416349886e-07,
+      "loss": 0.3352,
+      "step": 20334
+    },
+    {
+      "epoch": 2.6470128855915656,
+      "grad_norm": 3.5419743061065674,
+      "learning_rate": 3.5268867889604983e-07,
+      "loss": 0.3423,
+      "step": 20337
+    },
+    {
+      "epoch": 2.6474033580632566,
+      "grad_norm": 2.934070587158203,
+      "learning_rate": 3.5191810550411155e-07,
+      "loss": 0.3658,
+      "step": 20340
+    },
+    {
+      "epoch": 2.6477938305349475,
+      "grad_norm": 2.5605757236480713,
+      "learning_rate": 3.511483441223018e-07,
+      "loss": 0.3037,
+      "step": 20343
+    },
+    {
+      "epoch": 2.648184303006638,
+      "grad_norm": 2.988960027694702,
+      "learning_rate": 3.503793948850975e-07,
+      "loss": 0.3674,
+      "step": 20346
+    },
+    {
+      "epoch": 2.6485747754783286,
+      "grad_norm": 2.7395620346069336,
+      "learning_rate": 3.4961125792683184e-07,
+      "loss": 0.3509,
+      "step": 20349
+    },
+    {
+      "epoch": 2.6489652479500196,
+      "grad_norm": 2.5244340896606445,
+      "learning_rate": 3.488439333816951e-07,
+      "loss": 0.3272,
+      "step": 20352
+    },
+    {
+      "epoch": 2.64935572042171,
+      "grad_norm": 2.6971912384033203,
+      "learning_rate": 3.480774213837396e-07,
+      "loss": 0.3596,
+      "step": 20355
+    },
+    {
+      "epoch": 2.649746192893401,
+      "grad_norm": 3.134498119354248,
+      "learning_rate": 3.4731172206687257e-07,
+      "loss": 0.3619,
+      "step": 20358
+    },
+    {
+      "epoch": 2.6501366653650917,
+      "grad_norm": 2.9849419593811035,
+      "learning_rate": 3.465468355648588e-07,
+      "loss": 0.3809,
+      "step": 20361
+    },
+    {
+      "epoch": 2.6505271378367823,
+      "grad_norm": 2.864431858062744,
+      "learning_rate": 3.457827620113241e-07,
+      "loss": 0.3084,
+      "step": 20364
+    },
+    {
+      "epoch": 2.6509176103084733,
+      "grad_norm": 3.1214239597320557,
+      "learning_rate": 3.4501950153975003e-07,
+      "loss": 0.3852,
+      "step": 20367
+    },
+    {
+      "epoch": 2.651308082780164,
+      "grad_norm": 2.6842384338378906,
+      "learning_rate": 3.4425705428347556e-07,
+      "loss": 0.3501,
+      "step": 20370
+    },
+    {
+      "epoch": 2.651698555251855,
+      "grad_norm": 3.266386032104492,
+      "learning_rate": 3.434954203757007e-07,
+      "loss": 0.3648,
+      "step": 20373
+    },
+    {
+      "epoch": 2.6520890277235454,
+      "grad_norm": 2.81931471824646,
+      "learning_rate": 3.427345999494797e-07,
+      "loss": 0.3676,
+      "step": 20376
+    },
+    {
+      "epoch": 2.6524795001952364,
+      "grad_norm": 2.9153671264648438,
+      "learning_rate": 3.419745931377261e-07,
+      "loss": 0.3825,
+      "step": 20379
+    },
+    {
+      "epoch": 2.652869972666927,
+      "grad_norm": 2.6502585411071777,
+      "learning_rate": 3.41215400073211e-07,
+      "loss": 0.3146,
+      "step": 20382
+    },
+    {
+      "epoch": 2.653260445138618,
+      "grad_norm": 2.8781330585479736,
+      "learning_rate": 3.404570208885666e-07,
+      "loss": 0.4162,
+      "step": 20385
+    },
+    {
+      "epoch": 2.6536509176103085,
+      "grad_norm": 3.011082410812378,
+      "learning_rate": 3.3969945571627805e-07,
+      "loss": 0.3475,
+      "step": 20388
+    },
+    {
+      "epoch": 2.654041390081999,
+      "grad_norm": 2.7348506450653076,
+      "learning_rate": 3.389427046886906e-07,
+      "loss": 0.3648,
+      "step": 20391
+    },
+    {
+      "epoch": 2.65443186255369,
+      "grad_norm": 2.8573429584503174,
+      "learning_rate": 3.381867679380069e-07,
+      "loss": 0.3945,
+      "step": 20394
+    },
+    {
+      "epoch": 2.6548223350253806,
+      "grad_norm": 2.513740301132202,
+      "learning_rate": 3.374316455962884e-07,
+      "loss": 0.302,
+      "step": 20397
+    },
+    {
+      "epoch": 2.6552128074970716,
+      "grad_norm": 2.592041015625,
+      "learning_rate": 3.36677337795453e-07,
+      "loss": 0.3118,
+      "step": 20400
+    },
+    {
+      "epoch": 2.655603279968762,
+      "grad_norm": 2.622539520263672,
+      "learning_rate": 3.359238446672752e-07,
+      "loss": 0.346,
+      "step": 20403
+    },
+    {
+      "epoch": 2.655993752440453,
+      "grad_norm": 2.9038712978363037,
+      "learning_rate": 3.3517116634339097e-07,
+      "loss": 0.3932,
+      "step": 20406
+    },
+    {
+      "epoch": 2.6563842249121437,
+      "grad_norm": 2.3799030780792236,
+      "learning_rate": 3.344193029552911e-07,
+      "loss": 0.3106,
+      "step": 20409
+    },
+    {
+      "epoch": 2.6567746973838346,
+      "grad_norm": 2.7198572158813477,
+      "learning_rate": 3.336682546343228e-07,
+      "loss": 0.3262,
+      "step": 20412
+    },
+    {
+      "epoch": 2.657165169855525,
+      "grad_norm": 2.7029428482055664,
+      "learning_rate": 3.329180215116945e-07,
+      "loss": 0.3479,
+      "step": 20415
+    },
+    {
+      "epoch": 2.6575556423272158,
+      "grad_norm": 2.6045374870300293,
+      "learning_rate": 3.3216860371847013e-07,
+      "loss": 0.3007,
+      "step": 20418
+    },
+    {
+      "epoch": 2.6579461147989067,
+      "grad_norm": 2.965404748916626,
+      "learning_rate": 3.314200013855706e-07,
+      "loss": 0.3138,
+      "step": 20421
+    },
+    {
+      "epoch": 2.6583365872705973,
+      "grad_norm": 2.752427577972412,
+      "learning_rate": 3.3067221464377407e-07,
+      "loss": 0.3779,
+      "step": 20424
+    },
+    {
+      "epoch": 2.6587270597422883,
+      "grad_norm": 2.7018942832946777,
+      "learning_rate": 3.29925243623721e-07,
+      "loss": 0.3085,
+      "step": 20427
+    },
+    {
+      "epoch": 2.659117532213979,
+      "grad_norm": 2.695830821990967,
+      "learning_rate": 3.2917908845590263e-07,
+      "loss": 0.3486,
+      "step": 20430
+    },
+    {
+      "epoch": 2.6595080046856694,
+      "grad_norm": 2.9110679626464844,
+      "learning_rate": 3.2843374927067126e-07,
+      "loss": 0.3119,
+      "step": 20433
+    },
+    {
+      "epoch": 2.6598984771573604,
+      "grad_norm": 2.8462071418762207,
+      "learning_rate": 3.276892261982373e-07,
+      "loss": 0.3364,
+      "step": 20436
+    },
+    {
+      "epoch": 2.6602889496290514,
+      "grad_norm": 2.907721757888794,
+      "learning_rate": 3.26945519368666e-07,
+      "loss": 0.4112,
+      "step": 20439
+    },
+    {
+      "epoch": 2.660679422100742,
+      "grad_norm": 2.566709041595459,
+      "learning_rate": 3.2620262891188195e-07,
+      "loss": 0.3851,
+      "step": 20442
+    },
+    {
+      "epoch": 2.6610698945724325,
+      "grad_norm": 2.8268721103668213,
+      "learning_rate": 3.254605549576656e-07,
+      "loss": 0.3545,
+      "step": 20445
+    },
+    {
+      "epoch": 2.6614603670441235,
+      "grad_norm": 2.767282485961914,
+      "learning_rate": 3.2471929763565725e-07,
+      "loss": 0.3658,
+      "step": 20448
+    },
+    {
+      "epoch": 2.661850839515814,
+      "grad_norm": 3.0643694400787354,
+      "learning_rate": 3.2397885707535216e-07,
+      "loss": 0.3307,
+      "step": 20451
+    },
+    {
+      "epoch": 2.662241311987505,
+      "grad_norm": 3.210627555847168,
+      "learning_rate": 3.2323923340610296e-07,
+      "loss": 0.3483,
+      "step": 20454
+    },
+    {
+      "epoch": 2.6626317844591956,
+      "grad_norm": 2.5691981315612793,
+      "learning_rate": 3.2250042675712246e-07,
+      "loss": 0.2977,
+      "step": 20457
+    },
+    {
+      "epoch": 2.663022256930886,
+      "grad_norm": 2.906088352203369,
+      "learning_rate": 3.2176243725747736e-07,
+      "loss": 0.3708,
+      "step": 20460
+    },
+    {
+      "epoch": 2.663412729402577,
+      "grad_norm": 2.960641622543335,
+      "learning_rate": 3.210252650360918e-07,
+      "loss": 0.3184,
+      "step": 20463
+    },
+    {
+      "epoch": 2.663803201874268,
+      "grad_norm": 2.767688274383545,
+      "learning_rate": 3.2028891022174934e-07,
+      "loss": 0.364,
+      "step": 20466
+    },
+    {
+      "epoch": 2.6641936743459587,
+      "grad_norm": 2.6504275798797607,
+      "learning_rate": 3.1955337294309054e-07,
+      "loss": 0.3184,
+      "step": 20469
+    },
+    {
+      "epoch": 2.664584146817649,
+      "grad_norm": 2.9848546981811523,
+      "learning_rate": 3.1881865332861086e-07,
+      "loss": 0.3828,
+      "step": 20472
+    },
+    {
+      "epoch": 2.66497461928934,
+      "grad_norm": 2.9928441047668457,
+      "learning_rate": 3.180847515066643e-07,
+      "loss": 0.3589,
+      "step": 20475
+    },
+    {
+      "epoch": 2.6653650917610308,
+      "grad_norm": 3.2330639362335205,
+      "learning_rate": 3.173516676054628e-07,
+      "loss": 0.3848,
+      "step": 20478
+    },
+    {
+      "epoch": 2.6657555642327218,
+      "grad_norm": 2.6625139713287354,
+      "learning_rate": 3.1661940175307437e-07,
+      "loss": 0.365,
+      "step": 20481
+    },
+    {
+      "epoch": 2.6661460367044123,
+      "grad_norm": 2.730323553085327,
+      "learning_rate": 3.15887954077424e-07,
+      "loss": 0.328,
+      "step": 20484
+    },
+    {
+      "epoch": 2.666536509176103,
+      "grad_norm": 2.785428524017334,
+      "learning_rate": 3.1515732470629335e-07,
+      "loss": 0.3185,
+      "step": 20487
+    },
+    {
+      "epoch": 2.666926981647794,
+      "grad_norm": 2.6901137828826904,
+      "learning_rate": 3.144275137673236e-07,
+      "loss": 0.298,
+      "step": 20490
+    },
+    {
+      "epoch": 2.667317454119485,
+      "grad_norm": 2.662198066711426,
+      "learning_rate": 3.1369852138801006e-07,
+      "loss": 0.3452,
+      "step": 20493
+    },
+    {
+      "epoch": 2.6677079265911754,
+      "grad_norm": 2.7666854858398438,
+      "learning_rate": 3.1297034769570523e-07,
+      "loss": 0.333,
+      "step": 20496
+    },
+    {
+      "epoch": 2.668098399062866,
+      "grad_norm": 2.821671962738037,
+      "learning_rate": 3.1224299281762184e-07,
+      "loss": 0.3391,
+      "step": 20499
+    },
+    {
+      "epoch": 2.668488871534557,
+      "grad_norm": 2.7123184204101562,
+      "learning_rate": 3.115164568808254e-07,
+      "loss": 0.3117,
+      "step": 20502
+    },
+    {
+      "epoch": 2.6688793440062475,
+      "grad_norm": 2.6388189792633057,
+      "learning_rate": 3.107907400122406e-07,
+      "loss": 0.3992,
+      "step": 20505
+    },
+    {
+      "epoch": 2.6692698164779385,
+      "grad_norm": 2.8583977222442627,
+      "learning_rate": 3.100658423386488e-07,
+      "loss": 0.3633,
+      "step": 20508
+    },
+    {
+      "epoch": 2.669660288949629,
+      "grad_norm": 2.892751932144165,
+      "learning_rate": 3.093417639866886e-07,
+      "loss": 0.3982,
+      "step": 20511
+    },
+    {
+      "epoch": 2.6700507614213196,
+      "grad_norm": 2.8561367988586426,
+      "learning_rate": 3.0861850508285496e-07,
+      "loss": 0.4017,
+      "step": 20514
+    },
+    {
+      "epoch": 2.6704412338930106,
+      "grad_norm": 3.0648305416107178,
+      "learning_rate": 3.07896065753498e-07,
+      "loss": 0.4046,
+      "step": 20517
+    },
+    {
+      "epoch": 2.670831706364701,
+      "grad_norm": 2.555460214614868,
+      "learning_rate": 3.0717444612482883e-07,
+      "loss": 0.3425,
+      "step": 20520
+    },
+    {
+      "epoch": 2.671222178836392,
+      "grad_norm": 2.744659185409546,
+      "learning_rate": 3.064536463229112e-07,
+      "loss": 0.4306,
+      "step": 20523
+    },
+    {
+      "epoch": 2.6716126513080827,
+      "grad_norm": 2.578082323074341,
+      "learning_rate": 3.0573366647366764e-07,
+      "loss": 0.3021,
+      "step": 20526
+    },
+    {
+      "epoch": 2.6720031237797737,
+      "grad_norm": 2.590057849884033,
+      "learning_rate": 3.0501450670287756e-07,
+      "loss": 0.3247,
+      "step": 20529
+    },
+    {
+      "epoch": 2.672393596251464,
+      "grad_norm": 2.5789554119110107,
+      "learning_rate": 3.0429616713617607e-07,
+      "loss": 0.3732,
+      "step": 20532
+    },
+    {
+      "epoch": 2.672784068723155,
+      "grad_norm": 2.6770694255828857,
+      "learning_rate": 3.0357864789905653e-07,
+      "loss": 0.3459,
+      "step": 20535
+    },
+    {
+      "epoch": 2.6731745411948458,
+      "grad_norm": 2.7113239765167236,
+      "learning_rate": 3.0286194911686606e-07,
+      "loss": 0.3877,
+      "step": 20538
+    },
+    {
+      "epoch": 2.6735650136665363,
+      "grad_norm": 2.6579864025115967,
+      "learning_rate": 3.0214607091481276e-07,
+      "loss": 0.3727,
+      "step": 20541
+    },
+    {
+      "epoch": 2.6739554861382273,
+      "grad_norm": 2.62597918510437,
+      "learning_rate": 3.0143101341795823e-07,
+      "loss": 0.3935,
+      "step": 20544
+    },
+    {
+      "epoch": 2.674345958609918,
+      "grad_norm": 2.7485511302948,
+      "learning_rate": 3.0071677675122035e-07,
+      "loss": 0.4071,
+      "step": 20547
+    },
+    {
+      "epoch": 2.674736431081609,
+      "grad_norm": 2.8608615398406982,
+      "learning_rate": 3.0000336103937597e-07,
+      "loss": 0.3377,
+      "step": 20550
+    },
+    {
+      "epoch": 2.6751269035532994,
+      "grad_norm": 2.4845223426818848,
+      "learning_rate": 2.9929076640705714e-07,
+      "loss": 0.324,
+      "step": 20553
+    },
+    {
+      "epoch": 2.6755173760249904,
+      "grad_norm": 2.7939839363098145,
+      "learning_rate": 2.9857899297875304e-07,
+      "loss": 0.4306,
+      "step": 20556
+    },
+    {
+      "epoch": 2.675907848496681,
+      "grad_norm": 2.790100574493408,
+      "learning_rate": 2.9786804087880816e-07,
+      "loss": 0.4033,
+      "step": 20559
+    },
+    {
+      "epoch": 2.676298320968372,
+      "grad_norm": 2.7309770584106445,
+      "learning_rate": 2.9715791023142484e-07,
+      "loss": 0.3931,
+      "step": 20562
+    },
+    {
+      "epoch": 2.6766887934400625,
+      "grad_norm": 2.698127031326294,
+      "learning_rate": 2.9644860116066155e-07,
+      "loss": 0.3301,
+      "step": 20565
+    },
+    {
+      "epoch": 2.677079265911753,
+      "grad_norm": 2.612318754196167,
+      "learning_rate": 2.95740113790432e-07,
+      "loss": 0.3219,
+      "step": 20568
+    },
+    {
+      "epoch": 2.677469738383444,
+      "grad_norm": 2.7074458599090576,
+      "learning_rate": 2.950324482445088e-07,
+      "loss": 0.3249,
+      "step": 20571
+    },
+    {
+      "epoch": 2.6778602108551346,
+      "grad_norm": 2.573791265487671,
+      "learning_rate": 2.943256046465193e-07,
+      "loss": 0.3419,
+      "step": 20574
+    },
+    {
+      "epoch": 2.6782506833268256,
+      "grad_norm": 3.1822309494018555,
+      "learning_rate": 2.936195831199468e-07,
+      "loss": 0.3264,
+      "step": 20577
+    },
+    {
+      "epoch": 2.678641155798516,
+      "grad_norm": 2.9114439487457275,
+      "learning_rate": 2.929143837881326e-07,
+      "loss": 0.4111,
+      "step": 20580
+    },
+    {
+      "epoch": 2.6790316282702067,
+      "grad_norm": 2.9312427043914795,
+      "learning_rate": 2.922100067742739e-07,
+      "loss": 0.4224,
+      "step": 20583
+    },
+    {
+      "epoch": 2.6794221007418977,
+      "grad_norm": 3.390676259994507,
+      "learning_rate": 2.9150645220142273e-07,
+      "loss": 0.3307,
+      "step": 20586
+    },
+    {
+      "epoch": 2.6798125732135887,
+      "grad_norm": 2.750668525695801,
+      "learning_rate": 2.908037201924885e-07,
+      "loss": 0.4167,
+      "step": 20589
+    },
+    {
+      "epoch": 2.6802030456852792,
+      "grad_norm": 2.667370319366455,
+      "learning_rate": 2.9010181087023804e-07,
+      "loss": 0.3587,
+      "step": 20592
+    },
+    {
+      "epoch": 2.68059351815697,
+      "grad_norm": 3.1470043659210205,
+      "learning_rate": 2.894007243572933e-07,
+      "loss": 0.5031,
+      "step": 20595
+    },
+    {
+      "epoch": 2.6809839906286608,
+      "grad_norm": 2.8547494411468506,
+      "learning_rate": 2.887004607761329e-07,
+      "loss": 0.3249,
+      "step": 20598
+    },
+    {
+      "epoch": 2.6813744631003513,
+      "grad_norm": 2.8937506675720215,
+      "learning_rate": 2.880010202490896e-07,
+      "loss": 0.4048,
+      "step": 20601
+    },
+    {
+      "epoch": 2.6817649355720423,
+      "grad_norm": 2.470564365386963,
+      "learning_rate": 2.873024028983562e-07,
+      "loss": 0.2885,
+      "step": 20604
+    },
+    {
+      "epoch": 2.682155408043733,
+      "grad_norm": 2.804993152618408,
+      "learning_rate": 2.8660460884597953e-07,
+      "loss": 0.3877,
+      "step": 20607
+    },
+    {
+      "epoch": 2.6825458805154234,
+      "grad_norm": 2.796926259994507,
+      "learning_rate": 2.859076382138609e-07,
+      "loss": 0.3878,
+      "step": 20610
+    },
+    {
+      "epoch": 2.6829363529871144,
+      "grad_norm": 2.4991455078125,
+      "learning_rate": 2.852114911237619e-07,
+      "loss": 0.2931,
+      "step": 20613
+    },
+    {
+      "epoch": 2.6833268254588054,
+      "grad_norm": 2.7926175594329834,
+      "learning_rate": 2.845161676972968e-07,
+      "loss": 0.3678,
+      "step": 20616
+    },
+    {
+      "epoch": 2.683717297930496,
+      "grad_norm": 3.0255000591278076,
+      "learning_rate": 2.838216680559364e-07,
+      "loss": 0.3649,
+      "step": 20619
+    },
+    {
+      "epoch": 2.6841077704021865,
+      "grad_norm": 2.6406142711639404,
+      "learning_rate": 2.8312799232101007e-07,
+      "loss": 0.3252,
+      "step": 20622
+    },
+    {
+      "epoch": 2.6844982428738775,
+      "grad_norm": 2.9375603199005127,
+      "learning_rate": 2.824351406137005e-07,
+      "loss": 0.3596,
+      "step": 20625
+    },
+    {
+      "epoch": 2.684888715345568,
+      "grad_norm": 2.7059028148651123,
+      "learning_rate": 2.817431130550474e-07,
+      "loss": 0.4015,
+      "step": 20628
+    },
+    {
+      "epoch": 2.685279187817259,
+      "grad_norm": 2.975783586502075,
+      "learning_rate": 2.81051909765947e-07,
+      "loss": 0.345,
+      "step": 20631
+    },
+    {
+      "epoch": 2.6856696602889496,
+      "grad_norm": 2.8388760089874268,
+      "learning_rate": 2.8036153086714976e-07,
+      "loss": 0.3673,
+      "step": 20634
+    },
+    {
+      "epoch": 2.68606013276064,
+      "grad_norm": 2.8770854473114014,
+      "learning_rate": 2.7967197647926547e-07,
+      "loss": 0.3382,
+      "step": 20637
+    },
+    {
+      "epoch": 2.686450605232331,
+      "grad_norm": 2.836329460144043,
+      "learning_rate": 2.78983246722756e-07,
+      "loss": 0.4078,
+      "step": 20640
+    },
+    {
+      "epoch": 2.686841077704022,
+      "grad_norm": 2.931439161300659,
+      "learning_rate": 2.7829534171794236e-07,
+      "loss": 0.3589,
+      "step": 20643
+    },
+    {
+      "epoch": 2.6872315501757127,
+      "grad_norm": 2.7307281494140625,
+      "learning_rate": 2.7760826158499955e-07,
+      "loss": 0.3384,
+      "step": 20646
+    },
+    {
+      "epoch": 2.6876220226474032,
+      "grad_norm": 3.2210898399353027,
+      "learning_rate": 2.769220064439593e-07,
+      "loss": 0.4175,
+      "step": 20649
+    },
+    {
+      "epoch": 2.6880124951190942,
+      "grad_norm": 2.8683855533599854,
+      "learning_rate": 2.7623657641470734e-07,
+      "loss": 0.3799,
+      "step": 20652
+    },
+    {
+      "epoch": 2.688402967590785,
+      "grad_norm": 2.820208787918091,
+      "learning_rate": 2.755519716169891e-07,
+      "loss": 0.4057,
+      "step": 20655
+    },
+    {
+      "epoch": 2.688793440062476,
+      "grad_norm": 2.7478857040405273,
+      "learning_rate": 2.7486819217040273e-07,
+      "loss": 0.3818,
+      "step": 20658
+    },
+    {
+      "epoch": 2.6891839125341663,
+      "grad_norm": 2.4639649391174316,
+      "learning_rate": 2.7418523819440214e-07,
+      "loss": 0.2983,
+      "step": 20661
+    },
+    {
+      "epoch": 2.689574385005857,
+      "grad_norm": 2.7031710147857666,
+      "learning_rate": 2.735031098082996e-07,
+      "loss": 0.3065,
+      "step": 20664
+    },
+    {
+      "epoch": 2.689964857477548,
+      "grad_norm": 3.303699254989624,
+      "learning_rate": 2.728218071312605e-07,
+      "loss": 0.3628,
+      "step": 20667
+    },
+    {
+      "epoch": 2.6903553299492384,
+      "grad_norm": 2.7157487869262695,
+      "learning_rate": 2.721413302823067e-07,
+      "loss": 0.3855,
+      "step": 20670
+    },
+    {
+      "epoch": 2.6907458024209294,
+      "grad_norm": 2.9699556827545166,
+      "learning_rate": 2.714616793803171e-07,
+      "loss": 0.3907,
+      "step": 20673
+    },
+    {
+      "epoch": 2.69113627489262,
+      "grad_norm": 2.9308390617370605,
+      "learning_rate": 2.707828545440239e-07,
+      "loss": 0.3205,
+      "step": 20676
+    },
+    {
+      "epoch": 2.691526747364311,
+      "grad_norm": 2.65004563331604,
+      "learning_rate": 2.701048558920183e-07,
+      "loss": 0.2922,
+      "step": 20679
+    },
+    {
+      "epoch": 2.6919172198360015,
+      "grad_norm": 3.5007753372192383,
+      "learning_rate": 2.6942768354274283e-07,
+      "loss": 0.354,
+      "step": 20682
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 2.951852560043335,
+      "learning_rate": 2.687513376145007e-07,
+      "loss": 0.3592,
+      "step": 20685
+    },
+    {
+      "epoch": 2.692698164779383,
+      "grad_norm": 2.6974475383758545,
+      "learning_rate": 2.6807581822544616e-07,
+      "loss": 0.3565,
+      "step": 20688
+    },
+    {
+      "epoch": 2.6930886372510736,
+      "grad_norm": 3.6439194679260254,
+      "learning_rate": 2.6740112549359154e-07,
+      "loss": 0.3602,
+      "step": 20691
+    },
+    {
+      "epoch": 2.6934791097227646,
+      "grad_norm": 2.9142510890960693,
+      "learning_rate": 2.6672725953680476e-07,
+      "loss": 0.4112,
+      "step": 20694
+    },
+    {
+      "epoch": 2.693869582194455,
+      "grad_norm": 2.7999353408813477,
+      "learning_rate": 2.660542204728084e-07,
+      "loss": 0.3124,
+      "step": 20697
+    },
+    {
+      "epoch": 2.694260054666146,
+      "grad_norm": 2.537796974182129,
+      "learning_rate": 2.6538200841918103e-07,
+      "loss": 0.3301,
+      "step": 20700
+    },
+    {
+      "epoch": 2.6946505271378367,
+      "grad_norm": 3.0205585956573486,
+      "learning_rate": 2.6471062349335606e-07,
+      "loss": 0.3173,
+      "step": 20703
+    },
+    {
+      "epoch": 2.6950409996095273,
+      "grad_norm": 2.9223034381866455,
+      "learning_rate": 2.640400658126241e-07,
+      "loss": 0.4226,
+      "step": 20706
+    },
+    {
+      "epoch": 2.6954314720812182,
+      "grad_norm": 2.9517760276794434,
+      "learning_rate": 2.633703354941297e-07,
+      "loss": 0.3851,
+      "step": 20709
+    },
+    {
+      "epoch": 2.6958219445529092,
+      "grad_norm": 3.43226957321167,
+      "learning_rate": 2.6270143265487214e-07,
+      "loss": 0.4494,
+      "step": 20712
+    },
+    {
+      "epoch": 2.6962124170246,
+      "grad_norm": 2.9508416652679443,
+      "learning_rate": 2.6203335741170955e-07,
+      "loss": 0.4241,
+      "step": 20715
+    },
+    {
+      "epoch": 2.6966028894962903,
+      "grad_norm": 2.6791441440582275,
+      "learning_rate": 2.613661098813519e-07,
+      "loss": 0.3442,
+      "step": 20718
+    },
+    {
+      "epoch": 2.6969933619679813,
+      "grad_norm": 2.6088688373565674,
+      "learning_rate": 2.6069969018036655e-07,
+      "loss": 0.3336,
+      "step": 20721
+    },
+    {
+      "epoch": 2.697383834439672,
+      "grad_norm": 2.9644179344177246,
+      "learning_rate": 2.6003409842517425e-07,
+      "loss": 0.3761,
+      "step": 20724
+    },
+    {
+      "epoch": 2.697774306911363,
+      "grad_norm": 2.931863307952881,
+      "learning_rate": 2.5936933473205473e-07,
+      "loss": 0.3512,
+      "step": 20727
+    },
+    {
+      "epoch": 2.6981647793830534,
+      "grad_norm": 3.0752289295196533,
+      "learning_rate": 2.5870539921713955e-07,
+      "loss": 0.3662,
+      "step": 20730
+    },
+    {
+      "epoch": 2.698555251854744,
+      "grad_norm": 2.8674252033233643,
+      "learning_rate": 2.5804229199641594e-07,
+      "loss": 0.3455,
+      "step": 20733
+    },
+    {
+      "epoch": 2.698945724326435,
+      "grad_norm": 2.8869755268096924,
+      "learning_rate": 2.57380013185729e-07,
+      "loss": 0.3543,
+      "step": 20736
+    },
+    {
+      "epoch": 2.699336196798126,
+      "grad_norm": 2.9068305492401123,
+      "learning_rate": 2.567185629007768e-07,
+      "loss": 0.4262,
+      "step": 20739
+    },
+    {
+      "epoch": 2.6997266692698165,
+      "grad_norm": 2.6254804134368896,
+      "learning_rate": 2.560579412571129e-07,
+      "loss": 0.4404,
+      "step": 20742
+    },
+    {
+      "epoch": 2.700117141741507,
+      "grad_norm": 3.4937686920166016,
+      "learning_rate": 2.5539814837014734e-07,
+      "loss": 0.3142,
+      "step": 20745
+    },
+    {
+      "epoch": 2.700507614213198,
+      "grad_norm": 2.548431158065796,
+      "learning_rate": 2.547391843551439e-07,
+      "loss": 0.3637,
+      "step": 20748
+    },
+    {
+      "epoch": 2.7008980866848886,
+      "grad_norm": 3.108660936355591,
+      "learning_rate": 2.540810493272222e-07,
+      "loss": 0.3424,
+      "step": 20751
+    },
+    {
+      "epoch": 2.7012885591565796,
+      "grad_norm": 2.6465303897857666,
+      "learning_rate": 2.53423743401357e-07,
+      "loss": 0.3347,
+      "step": 20754
+    },
+    {
+      "epoch": 2.70167903162827,
+      "grad_norm": 2.571742057800293,
+      "learning_rate": 2.5276726669237917e-07,
+      "loss": 0.2957,
+      "step": 20757
+    },
+    {
+      "epoch": 2.7020695040999607,
+      "grad_norm": 2.738924741744995,
+      "learning_rate": 2.5211161931497195e-07,
+      "loss": 0.3669,
+      "step": 20760
+    },
+    {
+      "epoch": 2.7024599765716517,
+      "grad_norm": 2.8582653999328613,
+      "learning_rate": 2.5145680138367823e-07,
+      "loss": 0.4023,
+      "step": 20763
+    },
+    {
+      "epoch": 2.7028504490433427,
+      "grad_norm": 2.8408925533294678,
+      "learning_rate": 2.5080281301289034e-07,
+      "loss": 0.424,
+      "step": 20766
+    },
+    {
+      "epoch": 2.7032409215150333,
+      "grad_norm": 2.618551015853882,
+      "learning_rate": 2.5014965431686133e-07,
+      "loss": 0.2904,
+      "step": 20769
+    },
+    {
+      "epoch": 2.703631393986724,
+      "grad_norm": 2.550246000289917,
+      "learning_rate": 2.4949732540969553e-07,
+      "loss": 0.3352,
+      "step": 20772
+    },
+    {
+      "epoch": 2.704021866458415,
+      "grad_norm": 2.683685541152954,
+      "learning_rate": 2.488458264053523e-07,
+      "loss": 0.3334,
+      "step": 20775
+    },
+    {
+      "epoch": 2.7044123389301054,
+      "grad_norm": 2.9266393184661865,
+      "learning_rate": 2.481951574176494e-07,
+      "loss": 0.3804,
+      "step": 20778
+    },
+    {
+      "epoch": 2.7048028114017963,
+      "grad_norm": 2.7701256275177,
+      "learning_rate": 2.4754531856025557e-07,
+      "loss": 0.364,
+      "step": 20781
+    },
+    {
+      "epoch": 2.705193283873487,
+      "grad_norm": 2.6166975498199463,
+      "learning_rate": 2.4689630994669646e-07,
+      "loss": 0.3529,
+      "step": 20784
+    },
+    {
+      "epoch": 2.7055837563451774,
+      "grad_norm": 2.662623643875122,
+      "learning_rate": 2.462481316903537e-07,
+      "loss": 0.296,
+      "step": 20787
+    },
+    {
+      "epoch": 2.7059742288168684,
+      "grad_norm": 3.0897302627563477,
+      "learning_rate": 2.4560078390446216e-07,
+      "loss": 0.4447,
+      "step": 20790
+    },
+    {
+      "epoch": 2.706364701288559,
+      "grad_norm": 2.9640982151031494,
+      "learning_rate": 2.4495426670211154e-07,
+      "loss": 0.3694,
+      "step": 20793
+    },
+    {
+      "epoch": 2.70675517376025,
+      "grad_norm": 2.899473190307617,
+      "learning_rate": 2.443085801962469e-07,
+      "loss": 0.3705,
+      "step": 20796
+    },
+    {
+      "epoch": 2.7071456462319405,
+      "grad_norm": 2.732571601867676,
+      "learning_rate": 2.4366372449966924e-07,
+      "loss": 0.3467,
+      "step": 20799
+    },
+    {
+      "epoch": 2.7075361187036315,
+      "grad_norm": 4.155490398406982,
+      "learning_rate": 2.4301969972503223e-07,
+      "loss": 0.3475,
+      "step": 20802
+    },
+    {
+      "epoch": 2.707926591175322,
+      "grad_norm": 2.9572913646698,
+      "learning_rate": 2.4237650598484707e-07,
+      "loss": 0.3588,
+      "step": 20805
+    },
+    {
+      "epoch": 2.708317063647013,
+      "grad_norm": 2.9816484451293945,
+      "learning_rate": 2.417341433914777e-07,
+      "loss": 0.3269,
+      "step": 20808
+    },
+    {
+      "epoch": 2.7087075361187036,
+      "grad_norm": 2.436990261077881,
+      "learning_rate": 2.4109261205714386e-07,
+      "loss": 0.3758,
+      "step": 20811
+    },
+    {
+      "epoch": 2.709098008590394,
+      "grad_norm": 2.87477707862854,
+      "learning_rate": 2.404519120939197e-07,
+      "loss": 0.3635,
+      "step": 20814
+    },
+    {
+      "epoch": 2.709488481062085,
+      "grad_norm": 2.7576093673706055,
+      "learning_rate": 2.3981204361373247e-07,
+      "loss": 0.3162,
+      "step": 20817
+    },
+    {
+      "epoch": 2.7098789535337757,
+      "grad_norm": 3.003218650817871,
+      "learning_rate": 2.3917300672836876e-07,
+      "loss": 0.344,
+      "step": 20820
+    },
+    {
+      "epoch": 2.7102694260054667,
+      "grad_norm": 2.829570770263672,
+      "learning_rate": 2.385348015494648e-07,
+      "loss": 0.399,
+      "step": 20823
+    },
+    {
+      "epoch": 2.7106598984771573,
+      "grad_norm": 2.7657134532928467,
+      "learning_rate": 2.378974281885138e-07,
+      "loss": 0.3849,
+      "step": 20826
+    },
+    {
+      "epoch": 2.7110503709488483,
+      "grad_norm": 3.6430046558380127,
+      "learning_rate": 2.3726088675686542e-07,
+      "loss": 0.3418,
+      "step": 20829
+    },
+    {
+      "epoch": 2.711440843420539,
+      "grad_norm": 3.0621814727783203,
+      "learning_rate": 2.366251773657202e-07,
+      "loss": 0.3826,
+      "step": 20832
+    },
+    {
+      "epoch": 2.71183131589223,
+      "grad_norm": 2.840770721435547,
+      "learning_rate": 2.359903001261349e-07,
+      "loss": 0.3095,
+      "step": 20835
+    },
+    {
+      "epoch": 2.7122217883639204,
+      "grad_norm": 2.6842424869537354,
+      "learning_rate": 2.353562551490235e-07,
+      "loss": 0.3479,
+      "step": 20838
+    },
+    {
+      "epoch": 2.712612260835611,
+      "grad_norm": 2.5753138065338135,
+      "learning_rate": 2.3472304254515022e-07,
+      "loss": 0.3303,
+      "step": 20841
+    },
+    {
+      "epoch": 2.713002733307302,
+      "grad_norm": 4.057579040527344,
+      "learning_rate": 2.3409066242513655e-07,
+      "loss": 0.3138,
+      "step": 20844
+    },
+    {
+      "epoch": 2.7133932057789925,
+      "grad_norm": 2.7594196796417236,
+      "learning_rate": 2.3345911489945806e-07,
+      "loss": 0.35,
+      "step": 20847
+    },
+    {
+      "epoch": 2.7137836782506835,
+      "grad_norm": 2.6751604080200195,
+      "learning_rate": 2.3282840007844586e-07,
+      "loss": 0.3223,
+      "step": 20850
+    },
+    {
+      "epoch": 2.714174150722374,
+      "grad_norm": 3.2291057109832764,
+      "learning_rate": 2.3219851807228298e-07,
+      "loss": 0.4123,
+      "step": 20853
+    },
+    {
+      "epoch": 2.7145646231940646,
+      "grad_norm": 2.8833391666412354,
+      "learning_rate": 2.3156946899100918e-07,
+      "loss": 0.3697,
+      "step": 20856
+    },
+    {
+      "epoch": 2.7149550956657555,
+      "grad_norm": 2.8411707878112793,
+      "learning_rate": 2.3094125294451709e-07,
+      "loss": 0.3905,
+      "step": 20859
+    },
+    {
+      "epoch": 2.7153455681374465,
+      "grad_norm": 2.843876361846924,
+      "learning_rate": 2.3031387004255667e-07,
+      "loss": 0.3087,
+      "step": 20862
+    },
+    {
+      "epoch": 2.715736040609137,
+      "grad_norm": 2.7246623039245605,
+      "learning_rate": 2.2968732039472864e-07,
+      "loss": 0.3058,
+      "step": 20865
+    },
+    {
+      "epoch": 2.7161265130808276,
+      "grad_norm": 3.1246981620788574,
+      "learning_rate": 2.2906160411048982e-07,
+      "loss": 0.4129,
+      "step": 20868
+    },
+    {
+      "epoch": 2.7165169855525186,
+      "grad_norm": 2.8314807415008545,
+      "learning_rate": 2.2843672129915284e-07,
+      "loss": 0.3662,
+      "step": 20871
+    },
+    {
+      "epoch": 2.716907458024209,
+      "grad_norm": 2.765364408493042,
+      "learning_rate": 2.278126720698831e-07,
+      "loss": 0.3987,
+      "step": 20874
+    },
+    {
+      "epoch": 2.7172979304959,
+      "grad_norm": 2.515321969985962,
+      "learning_rate": 2.2718945653169954e-07,
+      "loss": 0.3245,
+      "step": 20877
+    },
+    {
+      "epoch": 2.7176884029675907,
+      "grad_norm": 3.015044689178467,
+      "learning_rate": 2.2656707479347783e-07,
+      "loss": 0.3235,
+      "step": 20880
+    },
+    {
+      "epoch": 2.7180788754392813,
+      "grad_norm": 2.9350194931030273,
+      "learning_rate": 2.2594552696394655e-07,
+      "loss": 0.3248,
+      "step": 20883
+    },
+    {
+      "epoch": 2.7184693479109723,
+      "grad_norm": 2.5761871337890625,
+      "learning_rate": 2.2532481315168774e-07,
+      "loss": 0.4054,
+      "step": 20886
+    },
+    {
+      "epoch": 2.7188598203826633,
+      "grad_norm": 2.6208813190460205,
+      "learning_rate": 2.2470493346513967e-07,
+      "loss": 0.3896,
+      "step": 20889
+    },
+    {
+      "epoch": 2.719250292854354,
+      "grad_norm": 2.8215200901031494,
+      "learning_rate": 2.2408588801259456e-07,
+      "loss": 0.3916,
+      "step": 20892
+    },
+    {
+      "epoch": 2.7196407653260444,
+      "grad_norm": 2.862316608428955,
+      "learning_rate": 2.2346767690219762e-07,
+      "loss": 0.3927,
+      "step": 20895
+    },
+    {
+      "epoch": 2.7200312377977354,
+      "grad_norm": 2.7122604846954346,
+      "learning_rate": 2.228503002419491e-07,
+      "loss": 0.3755,
+      "step": 20898
+    },
+    {
+      "epoch": 2.720421710269426,
+      "grad_norm": 2.7024004459381104,
+      "learning_rate": 2.2223375813970382e-07,
+      "loss": 0.382,
+      "step": 20901
+    },
+    {
+      "epoch": 2.720812182741117,
+      "grad_norm": 3.0531258583068848,
+      "learning_rate": 2.2161805070316955e-07,
+      "loss": 0.427,
+      "step": 20904
+    },
+    {
+      "epoch": 2.7212026552128075,
+      "grad_norm": 2.6628942489624023,
+      "learning_rate": 2.2100317803991023e-07,
+      "loss": 0.3531,
+      "step": 20907
+    },
+    {
+      "epoch": 2.721593127684498,
+      "grad_norm": 2.617644786834717,
+      "learning_rate": 2.203891402573416e-07,
+      "loss": 0.3378,
+      "step": 20910
+    },
+    {
+      "epoch": 2.721983600156189,
+      "grad_norm": 3.2814929485321045,
+      "learning_rate": 2.197759374627356e-07,
+      "loss": 0.4266,
+      "step": 20913
+    },
+    {
+      "epoch": 2.72237407262788,
+      "grad_norm": 2.880181074142456,
+      "learning_rate": 2.1916356976321717e-07,
+      "loss": 0.4139,
+      "step": 20916
+    },
+    {
+      "epoch": 2.7227645450995706,
+      "grad_norm": 3.127866506576538,
+      "learning_rate": 2.1855203726576512e-07,
+      "loss": 0.4289,
+      "step": 20919
+    },
+    {
+      "epoch": 2.723155017571261,
+      "grad_norm": 2.892545461654663,
+      "learning_rate": 2.17941340077214e-07,
+      "loss": 0.3724,
+      "step": 20922
+    },
+    {
+      "epoch": 2.723545490042952,
+      "grad_norm": 3.038330078125,
+      "learning_rate": 2.1733147830425129e-07,
+      "loss": 0.3305,
+      "step": 20925
+    },
+    {
+      "epoch": 2.7239359625146427,
+      "grad_norm": 2.841825246810913,
+      "learning_rate": 2.1672245205341668e-07,
+      "loss": 0.355,
+      "step": 20928
+    },
+    {
+      "epoch": 2.7243264349863336,
+      "grad_norm": 2.6401593685150146,
+      "learning_rate": 2.1611426143110792e-07,
+      "loss": 0.3588,
+      "step": 20931
+    },
+    {
+      "epoch": 2.724716907458024,
+      "grad_norm": 2.7327427864074707,
+      "learning_rate": 2.1550690654357387e-07,
+      "loss": 0.3542,
+      "step": 20934
+    },
+    {
+      "epoch": 2.7251073799297147,
+      "grad_norm": 2.9311492443084717,
+      "learning_rate": 2.1490038749691855e-07,
+      "loss": 0.3667,
+      "step": 20937
+    },
+    {
+      "epoch": 2.7254978524014057,
+      "grad_norm": 2.581843614578247,
+      "learning_rate": 2.1429470439709832e-07,
+      "loss": 0.3361,
+      "step": 20940
+    },
+    {
+      "epoch": 2.7258883248730963,
+      "grad_norm": 2.531803607940674,
+      "learning_rate": 2.1368985734992632e-07,
+      "loss": 0.3147,
+      "step": 20943
+    },
+    {
+      "epoch": 2.7262787973447873,
+      "grad_norm": 2.8539505004882812,
+      "learning_rate": 2.130858464610669e-07,
+      "loss": 0.3391,
+      "step": 20946
+    },
+    {
+      "epoch": 2.726669269816478,
+      "grad_norm": 2.8276593685150146,
+      "learning_rate": 2.124826718360401e-07,
+      "loss": 0.3272,
+      "step": 20949
+    },
+    {
+      "epoch": 2.727059742288169,
+      "grad_norm": 3.1101760864257812,
+      "learning_rate": 2.1188033358021887e-07,
+      "loss": 0.4169,
+      "step": 20952
+    },
+    {
+      "epoch": 2.7274502147598594,
+      "grad_norm": 2.8859715461730957,
+      "learning_rate": 2.1127883179883123e-07,
+      "loss": 0.343,
+      "step": 20955
+    },
+    {
+      "epoch": 2.7278406872315504,
+      "grad_norm": 2.626885414123535,
+      "learning_rate": 2.1067816659695705e-07,
+      "loss": 0.354,
+      "step": 20958
+    },
+    {
+      "epoch": 2.728231159703241,
+      "grad_norm": 3.709789276123047,
+      "learning_rate": 2.100783380795318e-07,
+      "loss": 0.4314,
+      "step": 20961
+    },
+    {
+      "epoch": 2.7286216321749315,
+      "grad_norm": 3.0039596557617188,
+      "learning_rate": 2.0947934635134504e-07,
+      "loss": 0.3791,
+      "step": 20964
+    },
+    {
+      "epoch": 2.7290121046466225,
+      "grad_norm": 2.7142927646636963,
+      "learning_rate": 2.0888119151703855e-07,
+      "loss": 0.4363,
+      "step": 20967
+    },
+    {
+      "epoch": 2.729402577118313,
+      "grad_norm": 2.700593948364258,
+      "learning_rate": 2.0828387368110825e-07,
+      "loss": 0.2971,
+      "step": 20970
+    },
+    {
+      "epoch": 2.729793049590004,
+      "grad_norm": 2.7692973613739014,
+      "learning_rate": 2.0768739294790453e-07,
+      "loss": 0.3486,
+      "step": 20973
+    },
+    {
+      "epoch": 2.7301835220616946,
+      "grad_norm": 2.7702739238739014,
+      "learning_rate": 2.070917494216329e-07,
+      "loss": 0.3118,
+      "step": 20976
+    },
+    {
+      "epoch": 2.7305739945333856,
+      "grad_norm": 2.789076805114746,
+      "learning_rate": 2.0649694320634962e-07,
+      "loss": 0.4196,
+      "step": 20979
+    },
+    {
+      "epoch": 2.730964467005076,
+      "grad_norm": 2.585819721221924,
+      "learning_rate": 2.059029744059654e-07,
+      "loss": 0.3413,
+      "step": 20982
+    },
+    {
+      "epoch": 2.731354939476767,
+      "grad_norm": 2.9140193462371826,
+      "learning_rate": 2.0530984312424728e-07,
+      "loss": 0.3709,
+      "step": 20985
+    },
+    {
+      "epoch": 2.7317454119484577,
+      "grad_norm": 2.968364715576172,
+      "learning_rate": 2.0471754946481293e-07,
+      "loss": 0.3156,
+      "step": 20988
+    },
+    {
+      "epoch": 2.732135884420148,
+      "grad_norm": 2.853219985961914,
+      "learning_rate": 2.04126093531134e-07,
+      "loss": 0.3475,
+      "step": 20991
+    },
+    {
+      "epoch": 2.732526356891839,
+      "grad_norm": 2.715740919113159,
+      "learning_rate": 2.035354754265384e-07,
+      "loss": 0.337,
+      "step": 20994
+    },
+    {
+      "epoch": 2.7329168293635298,
+      "grad_norm": 2.861161947250366,
+      "learning_rate": 2.029456952542047e-07,
+      "loss": 0.3736,
+      "step": 20997
+    },
+    {
+      "epoch": 2.7333073018352207,
+      "grad_norm": 3.0572056770324707,
+      "learning_rate": 2.0235675311716606e-07,
+      "loss": 0.3697,
+      "step": 21000
+    },
+    {
+      "epoch": 2.7336977743069113,
+      "grad_norm": 3.1050283908843994,
+      "learning_rate": 2.0176864911831074e-07,
+      "loss": 0.3451,
+      "step": 21003
+    },
+    {
+      "epoch": 2.734088246778602,
+      "grad_norm": 2.7847084999084473,
+      "learning_rate": 2.0118138336037818e-07,
+      "loss": 0.3791,
+      "step": 21006
+    },
+    {
+      "epoch": 2.734478719250293,
+      "grad_norm": 2.646023988723755,
+      "learning_rate": 2.0059495594596245e-07,
+      "loss": 0.3152,
+      "step": 21009
+    },
+    {
+      "epoch": 2.734869191721984,
+      "grad_norm": 2.6113431453704834,
+      "learning_rate": 2.000093669775105e-07,
+      "loss": 0.3829,
+      "step": 21012
+    },
+    {
+      "epoch": 2.7352596641936744,
+      "grad_norm": 3.1198410987854004,
+      "learning_rate": 1.9942461655732604e-07,
+      "loss": 0.3661,
+      "step": 21015
+    },
+    {
+      "epoch": 2.735650136665365,
+      "grad_norm": 3.019252061843872,
+      "learning_rate": 1.9884070478756124e-07,
+      "loss": 0.3273,
+      "step": 21018
+    },
+    {
+      "epoch": 2.736040609137056,
+      "grad_norm": 2.6802523136138916,
+      "learning_rate": 1.982576317702256e-07,
+      "loss": 0.3518,
+      "step": 21021
+    },
+    {
+      "epoch": 2.7364310816087465,
+      "grad_norm": 2.885608196258545,
+      "learning_rate": 1.976753976071799e-07,
+      "loss": 0.3363,
+      "step": 21024
+    },
+    {
+      "epoch": 2.7368215540804375,
+      "grad_norm": 3.0259361267089844,
+      "learning_rate": 1.9709400240014e-07,
+      "loss": 0.4027,
+      "step": 21027
+    },
+    {
+      "epoch": 2.737212026552128,
+      "grad_norm": 2.8382019996643066,
+      "learning_rate": 1.9651344625067404e-07,
+      "loss": 0.3445,
+      "step": 21030
+    },
+    {
+      "epoch": 2.7376024990238186,
+      "grad_norm": 2.760059356689453,
+      "learning_rate": 1.959337292602037e-07,
+      "loss": 0.341,
+      "step": 21033
+    },
+    {
+      "epoch": 2.7379929714955096,
+      "grad_norm": 2.7008633613586426,
+      "learning_rate": 1.9535485153000467e-07,
+      "loss": 0.3574,
+      "step": 21036
+    },
+    {
+      "epoch": 2.7383834439672006,
+      "grad_norm": 2.705885410308838,
+      "learning_rate": 1.94776813161206e-07,
+      "loss": 0.3451,
+      "step": 21039
+    },
+    {
+      "epoch": 2.738773916438891,
+      "grad_norm": 2.571812629699707,
+      "learning_rate": 1.941996142547886e-07,
+      "loss": 0.3598,
+      "step": 21042
+    },
+    {
+      "epoch": 2.7391643889105817,
+      "grad_norm": 2.7097597122192383,
+      "learning_rate": 1.9362325491158907e-07,
+      "loss": 0.4084,
+      "step": 21045
+    },
+    {
+      "epoch": 2.7395548613822727,
+      "grad_norm": 2.7397844791412354,
+      "learning_rate": 1.9304773523229626e-07,
+      "loss": 0.3158,
+      "step": 21048
+    },
+    {
+      "epoch": 2.739945333853963,
+      "grad_norm": 2.917585849761963,
+      "learning_rate": 1.924730553174503e-07,
+      "loss": 0.3487,
+      "step": 21051
+    },
+    {
+      "epoch": 2.740335806325654,
+      "grad_norm": 3.6193621158599854,
+      "learning_rate": 1.918992152674487e-07,
+      "loss": 0.3871,
+      "step": 21054
+    },
+    {
+      "epoch": 2.7407262787973448,
+      "grad_norm": 2.6469061374664307,
+      "learning_rate": 1.9132621518254014e-07,
+      "loss": 0.3751,
+      "step": 21057
+    },
+    {
+      "epoch": 2.7411167512690353,
+      "grad_norm": 2.455784797668457,
+      "learning_rate": 1.9075405516282562e-07,
+      "loss": 0.3128,
+      "step": 21060
+    },
+    {
+      "epoch": 2.7415072237407263,
+      "grad_norm": 2.900134325027466,
+      "learning_rate": 1.9018273530825972e-07,
+      "loss": 0.3579,
+      "step": 21063
+    },
+    {
+      "epoch": 2.7418976962124173,
+      "grad_norm": 2.7763450145721436,
+      "learning_rate": 1.8961225571865194e-07,
+      "loss": 0.3071,
+      "step": 21066
+    },
+    {
+      "epoch": 2.742288168684108,
+      "grad_norm": 3.2267343997955322,
+      "learning_rate": 1.890426164936643e-07,
+      "loss": 0.3962,
+      "step": 21069
+    },
+    {
+      "epoch": 2.7426786411557984,
+      "grad_norm": 2.72977876663208,
+      "learning_rate": 1.8847381773280991e-07,
+      "loss": 0.3407,
+      "step": 21072
+    },
+    {
+      "epoch": 2.7430691136274894,
+      "grad_norm": 3.236475944519043,
+      "learning_rate": 1.879058595354577e-07,
+      "loss": 0.4347,
+      "step": 21075
+    },
+    {
+      "epoch": 2.74345958609918,
+      "grad_norm": 2.5034587383270264,
+      "learning_rate": 1.873387420008288e-07,
+      "loss": 0.3305,
+      "step": 21078
+    },
+    {
+      "epoch": 2.743850058570871,
+      "grad_norm": 2.9593002796173096,
+      "learning_rate": 1.867724652279973e-07,
+      "loss": 0.3307,
+      "step": 21081
+    },
+    {
+      "epoch": 2.7442405310425615,
+      "grad_norm": 2.766495943069458,
+      "learning_rate": 1.8620702931589018e-07,
+      "loss": 0.3772,
+      "step": 21084
+    },
+    {
+      "epoch": 2.744631003514252,
+      "grad_norm": 2.7512776851654053,
+      "learning_rate": 1.85642434363289e-07,
+      "loss": 0.3637,
+      "step": 21087
+    },
+    {
+      "epoch": 2.745021475985943,
+      "grad_norm": 2.5375959873199463,
+      "learning_rate": 1.8507868046882648e-07,
+      "loss": 0.3328,
+      "step": 21090
+    },
+    {
+      "epoch": 2.7454119484576336,
+      "grad_norm": 3.020958662033081,
+      "learning_rate": 1.845157677309889e-07,
+      "loss": 0.3264,
+      "step": 21093
+    },
+    {
+      "epoch": 2.7458024209293246,
+      "grad_norm": 2.4167263507843018,
+      "learning_rate": 1.8395369624811643e-07,
+      "loss": 0.2931,
+      "step": 21096
+    },
+    {
+      "epoch": 2.746192893401015,
+      "grad_norm": 2.777470827102661,
+      "learning_rate": 1.833924661184022e-07,
+      "loss": 0.4431,
+      "step": 21099
+    },
+    {
+      "epoch": 2.746583365872706,
+      "grad_norm": 2.5936362743377686,
+      "learning_rate": 1.8283207743989118e-07,
+      "loss": 0.3583,
+      "step": 21102
+    },
+    {
+      "epoch": 2.7469738383443967,
+      "grad_norm": 2.837986946105957,
+      "learning_rate": 1.822725303104822e-07,
+      "loss": 0.3669,
+      "step": 21105
+    },
+    {
+      "epoch": 2.7473643108160877,
+      "grad_norm": 2.606499671936035,
+      "learning_rate": 1.8171382482792765e-07,
+      "loss": 0.3378,
+      "step": 21108
+    },
+    {
+      "epoch": 2.7477547832877782,
+      "grad_norm": 3.059852123260498,
+      "learning_rate": 1.8115596108983168e-07,
+      "loss": 0.3491,
+      "step": 21111
+    },
+    {
+      "epoch": 2.7481452557594688,
+      "grad_norm": 2.9495184421539307,
+      "learning_rate": 1.8059893919365135e-07,
+      "loss": 0.4117,
+      "step": 21114
+    },
+    {
+      "epoch": 2.7485357282311598,
+      "grad_norm": 2.9733073711395264,
+      "learning_rate": 1.8004275923669824e-07,
+      "loss": 0.4087,
+      "step": 21117
+    },
+    {
+      "epoch": 2.7489262007028503,
+      "grad_norm": 2.503350019454956,
+      "learning_rate": 1.7948742131613571e-07,
+      "loss": 0.3365,
+      "step": 21120
+    },
+    {
+      "epoch": 2.7493166731745413,
+      "grad_norm": 2.9972898960113525,
+      "learning_rate": 1.7893292552897956e-07,
+      "loss": 0.3652,
+      "step": 21123
+    },
+    {
+      "epoch": 2.749707145646232,
+      "grad_norm": 2.9477357864379883,
+      "learning_rate": 1.783792719720989e-07,
+      "loss": 0.3609,
+      "step": 21126
+    },
+    {
+      "epoch": 2.7500976181179224,
+      "grad_norm": 2.971484422683716,
+      "learning_rate": 1.7782646074221643e-07,
+      "loss": 0.3444,
+      "step": 21129
+    },
+    {
+      "epoch": 2.7504880905896134,
+      "grad_norm": 2.396226644515991,
+      "learning_rate": 1.7727449193590707e-07,
+      "loss": 0.3327,
+      "step": 21132
+    },
+    {
+      "epoch": 2.7508785630613044,
+      "grad_norm": 2.7505605220794678,
+      "learning_rate": 1.7672336564959813e-07,
+      "loss": 0.3816,
+      "step": 21135
+    },
+    {
+      "epoch": 2.751269035532995,
+      "grad_norm": 2.9691169261932373,
+      "learning_rate": 1.761730819795704e-07,
+      "loss": 0.3654,
+      "step": 21138
+    },
+    {
+      "epoch": 2.7516595080046855,
+      "grad_norm": 3.171504259109497,
+      "learning_rate": 1.7562364102195806e-07,
+      "loss": 0.3361,
+      "step": 21141
+    },
+    {
+      "epoch": 2.7520499804763765,
+      "grad_norm": 2.5362813472747803,
+      "learning_rate": 1.7507504287274603e-07,
+      "loss": 0.2974,
+      "step": 21144
+    },
+    {
+      "epoch": 2.752440452948067,
+      "grad_norm": 2.5266830921173096,
+      "learning_rate": 1.7452728762777372e-07,
+      "loss": 0.4153,
+      "step": 21147
+    },
+    {
+      "epoch": 2.752830925419758,
+      "grad_norm": 3.282205820083618,
+      "learning_rate": 1.739803753827335e-07,
+      "loss": 0.2816,
+      "step": 21150
+    },
+    {
+      "epoch": 2.7532213978914486,
+      "grad_norm": 2.7888989448547363,
+      "learning_rate": 1.7343430623316947e-07,
+      "loss": 0.3416,
+      "step": 21153
+    },
+    {
+      "epoch": 2.753611870363139,
+      "grad_norm": 2.560568332672119,
+      "learning_rate": 1.728890802744776e-07,
+      "loss": 0.3341,
+      "step": 21156
+    },
+    {
+      "epoch": 2.75400234283483,
+      "grad_norm": 3.4520578384399414,
+      "learning_rate": 1.723446976019094e-07,
+      "loss": 0.3717,
+      "step": 21159
+    },
+    {
+      "epoch": 2.754392815306521,
+      "grad_norm": 2.7265379428863525,
+      "learning_rate": 1.7180115831056665e-07,
+      "loss": 0.3872,
+      "step": 21162
+    },
+    {
+      "epoch": 2.7547832877782117,
+      "grad_norm": 2.842808246612549,
+      "learning_rate": 1.712584624954039e-07,
+      "loss": 0.3927,
+      "step": 21165
+    },
+    {
+      "epoch": 2.7551737602499022,
+      "grad_norm": 2.4746949672698975,
+      "learning_rate": 1.707166102512303e-07,
+      "loss": 0.3752,
+      "step": 21168
+    },
+    {
+      "epoch": 2.7555642327215932,
+      "grad_norm": 2.7992329597473145,
+      "learning_rate": 1.7017560167270519e-07,
+      "loss": 0.3562,
+      "step": 21171
+    },
+    {
+      "epoch": 2.755954705193284,
+      "grad_norm": 3.015634059906006,
+      "learning_rate": 1.6963543685434236e-07,
+      "loss": 0.4082,
+      "step": 21174
+    },
+    {
+      "epoch": 2.7563451776649748,
+      "grad_norm": 2.884288787841797,
+      "learning_rate": 1.6909611589050635e-07,
+      "loss": 0.3104,
+      "step": 21177
+    },
+    {
+      "epoch": 2.7567356501366653,
+      "grad_norm": 3.6291792392730713,
+      "learning_rate": 1.6855763887541565e-07,
+      "loss": 0.3426,
+      "step": 21180
+    },
+    {
+      "epoch": 2.757126122608356,
+      "grad_norm": 2.6148040294647217,
+      "learning_rate": 1.6802000590314283e-07,
+      "loss": 0.3069,
+      "step": 21183
+    },
+    {
+      "epoch": 2.757516595080047,
+      "grad_norm": 3.024834156036377,
+      "learning_rate": 1.6748321706760994e-07,
+      "loss": 0.3358,
+      "step": 21186
+    },
+    {
+      "epoch": 2.757907067551738,
+      "grad_norm": 2.5863471031188965,
+      "learning_rate": 1.66947272462592e-07,
+      "loss": 0.3238,
+      "step": 21189
+    },
+    {
+      "epoch": 2.7582975400234284,
+      "grad_norm": 3.284332036972046,
+      "learning_rate": 1.6641217218171912e-07,
+      "loss": 0.4084,
+      "step": 21192
+    },
+    {
+      "epoch": 2.758688012495119,
+      "grad_norm": 3.1912519931793213,
+      "learning_rate": 1.658779163184715e-07,
+      "loss": 0.353,
+      "step": 21195
+    },
+    {
+      "epoch": 2.75907848496681,
+      "grad_norm": 2.728577136993408,
+      "learning_rate": 1.6534450496618171e-07,
+      "loss": 0.408,
+      "step": 21198
+    },
+    {
+      "epoch": 2.7594689574385005,
+      "grad_norm": 2.7622365951538086,
+      "learning_rate": 1.648119382180363e-07,
+      "loss": 0.3854,
+      "step": 21201
+    },
+    {
+      "epoch": 2.7598594299101915,
+      "grad_norm": 3.1888158321380615,
+      "learning_rate": 1.6428021616707423e-07,
+      "loss": 0.4322,
+      "step": 21204
+    },
+    {
+      "epoch": 2.760249902381882,
+      "grad_norm": 2.6636276245117188,
+      "learning_rate": 1.6374933890618504e-07,
+      "loss": 0.3482,
+      "step": 21207
+    },
+    {
+      "epoch": 2.7606403748535726,
+      "grad_norm": 3.0101702213287354,
+      "learning_rate": 1.6321930652811236e-07,
+      "loss": 0.3703,
+      "step": 21210
+    },
+    {
+      "epoch": 2.7610308473252636,
+      "grad_norm": 3.2155416011810303,
+      "learning_rate": 1.6269011912545208e-07,
+      "loss": 0.384,
+      "step": 21213
+    },
+    {
+      "epoch": 2.761421319796954,
+      "grad_norm": 2.743858814239502,
+      "learning_rate": 1.6216177679065136e-07,
+      "loss": 0.3099,
+      "step": 21216
+    },
+    {
+      "epoch": 2.761811792268645,
+      "grad_norm": 2.7741172313690186,
+      "learning_rate": 1.6163427961601086e-07,
+      "loss": 0.3717,
+      "step": 21219
+    },
+    {
+      "epoch": 2.7622022647403357,
+      "grad_norm": 3.037571907043457,
+      "learning_rate": 1.6110762769368294e-07,
+      "loss": 0.3058,
+      "step": 21222
+    },
+    {
+      "epoch": 2.7625927372120267,
+      "grad_norm": 2.6147408485412598,
+      "learning_rate": 1.60581821115674e-07,
+      "loss": 0.3178,
+      "step": 21225
+    },
+    {
+      "epoch": 2.7629832096837172,
+      "grad_norm": 2.956163167953491,
+      "learning_rate": 1.6005685997383945e-07,
+      "loss": 0.3459,
+      "step": 21228
+    },
+    {
+      "epoch": 2.7633736821554082,
+      "grad_norm": 3.1551594734191895,
+      "learning_rate": 1.5953274435988985e-07,
+      "loss": 0.3189,
+      "step": 21231
+    },
+    {
+      "epoch": 2.763764154627099,
+      "grad_norm": 2.789609670639038,
+      "learning_rate": 1.590094743653875e-07,
+      "loss": 0.3035,
+      "step": 21234
+    },
+    {
+      "epoch": 2.7641546270987893,
+      "grad_norm": 3.474595308303833,
+      "learning_rate": 1.5848705008174535e-07,
+      "loss": 0.434,
+      "step": 21237
+    },
+    {
+      "epoch": 2.7645450995704803,
+      "grad_norm": 2.9483280181884766,
+      "learning_rate": 1.5796547160023046e-07,
+      "loss": 0.3975,
+      "step": 21240
+    },
+    {
+      "epoch": 2.764935572042171,
+      "grad_norm": 3.087123155593872,
+      "learning_rate": 1.5744473901196211e-07,
+      "loss": 0.306,
+      "step": 21243
+    },
+    {
+      "epoch": 2.765326044513862,
+      "grad_norm": 2.7233099937438965,
+      "learning_rate": 1.5692485240791034e-07,
+      "loss": 0.3202,
+      "step": 21246
+    },
+    {
+      "epoch": 2.7657165169855524,
+      "grad_norm": 3.1020421981811523,
+      "learning_rate": 1.5640581187889857e-07,
+      "loss": 0.3874,
+      "step": 21249
+    },
+    {
+      "epoch": 2.7661069894572434,
+      "grad_norm": 3.102666139602661,
+      "learning_rate": 1.558876175156021e-07,
+      "loss": 0.3741,
+      "step": 21252
+    },
+    {
+      "epoch": 2.766497461928934,
+      "grad_norm": 2.6903672218322754,
+      "learning_rate": 1.5537026940854794e-07,
+      "loss": 0.3412,
+      "step": 21255
+    },
+    {
+      "epoch": 2.766887934400625,
+      "grad_norm": 2.9768612384796143,
+      "learning_rate": 1.54853767648116e-07,
+      "loss": 0.3017,
+      "step": 21258
+    },
+    {
+      "epoch": 2.7672784068723155,
+      "grad_norm": 3.4429922103881836,
+      "learning_rate": 1.543381123245391e-07,
+      "loss": 0.321,
+      "step": 21261
+    },
+    {
+      "epoch": 2.767668879344006,
+      "grad_norm": 2.8304970264434814,
+      "learning_rate": 1.538233035278991e-07,
+      "loss": 0.328,
+      "step": 21264
+    },
+    {
+      "epoch": 2.768059351815697,
+      "grad_norm": 2.5452582836151123,
+      "learning_rate": 1.5330934134813346e-07,
+      "loss": 0.2859,
+      "step": 21267
+    },
+    {
+      "epoch": 2.7684498242873876,
+      "grad_norm": 2.8561484813690186,
+      "learning_rate": 1.5279622587502986e-07,
+      "loss": 0.4397,
+      "step": 21270
+    },
+    {
+      "epoch": 2.7688402967590786,
+      "grad_norm": 3.0606303215026855,
+      "learning_rate": 1.5228395719822876e-07,
+      "loss": 0.4187,
+      "step": 21273
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 2.421595335006714,
+      "learning_rate": 1.5177253540722247e-07,
+      "loss": 0.2697,
+      "step": 21276
+    },
+    {
+      "epoch": 2.7696212417024597,
+      "grad_norm": 3.015394449234009,
+      "learning_rate": 1.512619605913551e-07,
+      "loss": 0.3576,
+      "step": 21279
+    },
+    {
+      "epoch": 2.7700117141741507,
+      "grad_norm": 2.740720510482788,
+      "learning_rate": 1.5075223283982255e-07,
+      "loss": 0.3246,
+      "step": 21282
+    },
+    {
+      "epoch": 2.7704021866458417,
+      "grad_norm": 2.849069833755493,
+      "learning_rate": 1.5024335224167407e-07,
+      "loss": 0.342,
+      "step": 21285
+    },
+    {
+      "epoch": 2.7707926591175323,
+      "grad_norm": 2.9401018619537354,
+      "learning_rate": 1.4973531888580916e-07,
+      "loss": 0.3456,
+      "step": 21288
+    },
+    {
+      "epoch": 2.771183131589223,
+      "grad_norm": 2.6727066040039062,
+      "learning_rate": 1.4922813286098016e-07,
+      "loss": 0.3463,
+      "step": 21291
+    },
+    {
+      "epoch": 2.771573604060914,
+      "grad_norm": 3.0198166370391846,
+      "learning_rate": 1.487217942557928e-07,
+      "loss": 0.3842,
+      "step": 21294
+    },
+    {
+      "epoch": 2.7719640765326043,
+      "grad_norm": 2.9366579055786133,
+      "learning_rate": 1.4821630315870194e-07,
+      "loss": 0.367,
+      "step": 21297
+    },
+    {
+      "epoch": 2.7723545490042953,
+      "grad_norm": 3.7701351642608643,
+      "learning_rate": 1.4771165965801582e-07,
+      "loss": 0.4056,
+      "step": 21300
+    },
+    {
+      "epoch": 2.772745021475986,
+      "grad_norm": 2.716285467147827,
+      "learning_rate": 1.4720786384189557e-07,
+      "loss": 0.3333,
+      "step": 21303
+    },
+    {
+      "epoch": 2.7731354939476764,
+      "grad_norm": 2.849595546722412,
+      "learning_rate": 1.4670491579835245e-07,
+      "loss": 0.3918,
+      "step": 21306
+    },
+    {
+      "epoch": 2.7735259664193674,
+      "grad_norm": 2.8046321868896484,
+      "learning_rate": 1.462028156152512e-07,
+      "loss": 0.3327,
+      "step": 21309
+    },
+    {
+      "epoch": 2.7739164388910584,
+      "grad_norm": 3.198570966720581,
+      "learning_rate": 1.4570156338030606e-07,
+      "loss": 0.3182,
+      "step": 21312
+    },
+    {
+      "epoch": 2.774306911362749,
+      "grad_norm": 2.8275673389434814,
+      "learning_rate": 1.4520115918108701e-07,
+      "loss": 0.3296,
+      "step": 21315
+    },
+    {
+      "epoch": 2.7746973838344395,
+      "grad_norm": 2.7337069511413574,
+      "learning_rate": 1.447016031050119e-07,
+      "loss": 0.3556,
+      "step": 21318
+    },
+    {
+      "epoch": 2.7750878563061305,
+      "grad_norm": 2.702484369277954,
+      "learning_rate": 1.442028952393526e-07,
+      "loss": 0.4011,
+      "step": 21321
+    },
+    {
+      "epoch": 2.775478328777821,
+      "grad_norm": 2.6298041343688965,
+      "learning_rate": 1.4370503567123274e-07,
+      "loss": 0.3267,
+      "step": 21324
+    },
+    {
+      "epoch": 2.775868801249512,
+      "grad_norm": 2.7082111835479736,
+      "learning_rate": 1.4320802448762716e-07,
+      "loss": 0.3419,
+      "step": 21327
+    },
+    {
+      "epoch": 2.7762592737212026,
+      "grad_norm": 3.18163800239563,
+      "learning_rate": 1.4271186177536256e-07,
+      "loss": 0.3421,
+      "step": 21330
+    },
+    {
+      "epoch": 2.776649746192893,
+      "grad_norm": 2.9443161487579346,
+      "learning_rate": 1.4221654762111624e-07,
+      "loss": 0.3675,
+      "step": 21333
+    },
+    {
+      "epoch": 2.777040218664584,
+      "grad_norm": 3.03935170173645,
+      "learning_rate": 1.4172208211142124e-07,
+      "loss": 0.3989,
+      "step": 21336
+    },
+    {
+      "epoch": 2.777430691136275,
+      "grad_norm": 2.619476795196533,
+      "learning_rate": 1.4122846533265733e-07,
+      "loss": 0.3682,
+      "step": 21339
+    },
+    {
+      "epoch": 2.7778211636079657,
+      "grad_norm": 2.9960954189300537,
+      "learning_rate": 1.4073569737105942e-07,
+      "loss": 0.3476,
+      "step": 21342
+    },
+    {
+      "epoch": 2.7782116360796563,
+      "grad_norm": 3.048250675201416,
+      "learning_rate": 1.4024377831271253e-07,
+      "loss": 0.38,
+      "step": 21345
+    },
+    {
+      "epoch": 2.7786021085513473,
+      "grad_norm": 3.039592981338501,
+      "learning_rate": 1.3975270824355402e-07,
+      "loss": 0.2967,
+      "step": 21348
+    },
+    {
+      "epoch": 2.778992581023038,
+      "grad_norm": 3.194598913192749,
+      "learning_rate": 1.3926248724937363e-07,
+      "loss": 0.4696,
+      "step": 21351
+    },
+    {
+      "epoch": 2.779383053494729,
+      "grad_norm": 2.963008165359497,
+      "learning_rate": 1.3877311541581063e-07,
+      "loss": 0.3861,
+      "step": 21354
+    },
+    {
+      "epoch": 2.7797735259664194,
+      "grad_norm": 2.9709367752075195,
+      "learning_rate": 1.3828459282835828e-07,
+      "loss": 0.4144,
+      "step": 21357
+    },
+    {
+      "epoch": 2.78016399843811,
+      "grad_norm": 2.8352601528167725,
+      "learning_rate": 1.3779691957235996e-07,
+      "loss": 0.3747,
+      "step": 21360
+    },
+    {
+      "epoch": 2.780554470909801,
+      "grad_norm": 2.657566547393799,
+      "learning_rate": 1.3731009573301035e-07,
+      "loss": 0.3459,
+      "step": 21363
+    },
+    {
+      "epoch": 2.7809449433814915,
+      "grad_norm": 2.850067615509033,
+      "learning_rate": 1.368241213953586e-07,
+      "loss": 0.4104,
+      "step": 21366
+    },
+    {
+      "epoch": 2.7813354158531824,
+      "grad_norm": 3.0880982875823975,
+      "learning_rate": 1.3633899664430183e-07,
+      "loss": 0.36,
+      "step": 21369
+    },
+    {
+      "epoch": 2.781725888324873,
+      "grad_norm": 3.7259879112243652,
+      "learning_rate": 1.3585472156459e-07,
+      "loss": 0.3831,
+      "step": 21372
+    },
+    {
+      "epoch": 2.782116360796564,
+      "grad_norm": 3.025913953781128,
+      "learning_rate": 1.3537129624082657e-07,
+      "loss": 0.3829,
+      "step": 21375
+    },
+    {
+      "epoch": 2.7825068332682545,
+      "grad_norm": 3.209459066390991,
+      "learning_rate": 1.3488872075746395e-07,
+      "loss": 0.3427,
+      "step": 21378
+    },
+    {
+      "epoch": 2.7828973057399455,
+      "grad_norm": 2.893686294555664,
+      "learning_rate": 1.3440699519880695e-07,
+      "loss": 0.3317,
+      "step": 21381
+    },
+    {
+      "epoch": 2.783287778211636,
+      "grad_norm": 2.8265738487243652,
+      "learning_rate": 1.3392611964901159e-07,
+      "loss": 0.3864,
+      "step": 21384
+    },
+    {
+      "epoch": 2.7836782506833266,
+      "grad_norm": 2.7982900142669678,
+      "learning_rate": 1.334460941920873e-07,
+      "loss": 0.3053,
+      "step": 21387
+    },
+    {
+      "epoch": 2.7840687231550176,
+      "grad_norm": 2.68939471244812,
+      "learning_rate": 1.329669189118915e-07,
+      "loss": 0.4247,
+      "step": 21390
+    },
+    {
+      "epoch": 2.784459195626708,
+      "grad_norm": 2.8545985221862793,
+      "learning_rate": 1.324885938921372e-07,
+      "loss": 0.3723,
+      "step": 21393
+    },
+    {
+      "epoch": 2.784849668098399,
+      "grad_norm": 2.9172253608703613,
+      "learning_rate": 1.3201111921638532e-07,
+      "loss": 0.346,
+      "step": 21396
+    },
+    {
+      "epoch": 2.7852401405700897,
+      "grad_norm": 2.608978271484375,
+      "learning_rate": 1.3153449496805028e-07,
+      "loss": 0.3201,
+      "step": 21399
+    },
+    {
+      "epoch": 2.7856306130417803,
+      "grad_norm": 2.9829037189483643,
+      "learning_rate": 1.3105872123039765e-07,
+      "loss": 0.4295,
+      "step": 21402
+    },
+    {
+      "epoch": 2.7860210855134713,
+      "grad_norm": 2.7616302967071533,
+      "learning_rate": 1.305837980865432e-07,
+      "loss": 0.4219,
+      "step": 21405
+    },
+    {
+      "epoch": 2.7864115579851623,
+      "grad_norm": 3.1220195293426514,
+      "learning_rate": 1.3010972561945555e-07,
+      "loss": 0.3856,
+      "step": 21408
+    },
+    {
+      "epoch": 2.786802030456853,
+      "grad_norm": 2.6154117584228516,
+      "learning_rate": 1.2963650391195403e-07,
+      "loss": 0.3607,
+      "step": 21411
+    },
+    {
+      "epoch": 2.7871925029285434,
+      "grad_norm": 2.7787892818450928,
+      "learning_rate": 1.2916413304670972e-07,
+      "loss": 0.3516,
+      "step": 21414
+    },
+    {
+      "epoch": 2.7875829754002344,
+      "grad_norm": 3.176905632019043,
+      "learning_rate": 1.2869261310624437e-07,
+      "loss": 0.3978,
+      "step": 21417
+    },
+    {
+      "epoch": 2.787973447871925,
+      "grad_norm": 2.9123852252960205,
+      "learning_rate": 1.282219441729321e-07,
+      "loss": 0.3417,
+      "step": 21420
+    },
+    {
+      "epoch": 2.788363920343616,
+      "grad_norm": 2.5749077796936035,
+      "learning_rate": 1.2775212632899715e-07,
+      "loss": 0.345,
+      "step": 21423
+    },
+    {
+      "epoch": 2.7887543928153065,
+      "grad_norm": 2.712224006652832,
+      "learning_rate": 1.2728315965651606e-07,
+      "loss": 0.4133,
+      "step": 21426
+    },
+    {
+      "epoch": 2.789144865286997,
+      "grad_norm": 2.8069446086883545,
+      "learning_rate": 1.2681504423741665e-07,
+      "loss": 0.3276,
+      "step": 21429
+    },
+    {
+      "epoch": 2.789535337758688,
+      "grad_norm": 2.740924596786499,
+      "learning_rate": 1.2634778015347682e-07,
+      "loss": 0.3468,
+      "step": 21432
+    },
+    {
+      "epoch": 2.789925810230379,
+      "grad_norm": 2.735494613647461,
+      "learning_rate": 1.2588136748632685e-07,
+      "loss": 0.3151,
+      "step": 21435
+    },
+    {
+      "epoch": 2.7903162827020696,
+      "grad_norm": 2.7889745235443115,
+      "learning_rate": 1.2541580631744931e-07,
+      "loss": 0.338,
+      "step": 21438
+    },
+    {
+      "epoch": 2.79070675517376,
+      "grad_norm": 2.608374834060669,
+      "learning_rate": 1.249510967281753e-07,
+      "loss": 0.3711,
+      "step": 21441
+    },
+    {
+      "epoch": 2.791097227645451,
+      "grad_norm": 2.477095365524292,
+      "learning_rate": 1.2448723879968927e-07,
+      "loss": 0.3962,
+      "step": 21444
+    },
+    {
+      "epoch": 2.7914877001171416,
+      "grad_norm": 3.061278820037842,
+      "learning_rate": 1.2402423261302532e-07,
+      "loss": 0.3421,
+      "step": 21447
+    },
+    {
+      "epoch": 2.7918781725888326,
+      "grad_norm": 2.966050386428833,
+      "learning_rate": 1.2356207824907152e-07,
+      "loss": 0.3855,
+      "step": 21450
+    },
+    {
+      "epoch": 2.792268645060523,
+      "grad_norm": 2.758970022201538,
+      "learning_rate": 1.2310077578856328e-07,
+      "loss": 0.3178,
+      "step": 21453
+    },
+    {
+      "epoch": 2.7926591175322137,
+      "grad_norm": 2.657818555831909,
+      "learning_rate": 1.2264032531209004e-07,
+      "loss": 0.3399,
+      "step": 21456
+    },
+    {
+      "epoch": 2.7930495900039047,
+      "grad_norm": 2.7079572677612305,
+      "learning_rate": 1.2218072690009187e-07,
+      "loss": 0.3295,
+      "step": 21459
+    },
+    {
+      "epoch": 2.7934400624755957,
+      "grad_norm": 2.721400737762451,
+      "learning_rate": 1.2172198063285957e-07,
+      "loss": 0.3228,
+      "step": 21462
+    },
+    {
+      "epoch": 2.7938305349472863,
+      "grad_norm": 2.8544654846191406,
+      "learning_rate": 1.2126408659053402e-07,
+      "loss": 0.3513,
+      "step": 21465
+    },
+    {
+      "epoch": 2.794221007418977,
+      "grad_norm": 3.039076566696167,
+      "learning_rate": 1.2080704485310957e-07,
+      "loss": 0.4097,
+      "step": 21468
+    },
+    {
+      "epoch": 2.794611479890668,
+      "grad_norm": 3.0144577026367188,
+      "learning_rate": 1.2035085550043013e-07,
+      "loss": 0.3648,
+      "step": 21471
+    },
+    {
+      "epoch": 2.7950019523623584,
+      "grad_norm": 2.8478612899780273,
+      "learning_rate": 1.1989551861219084e-07,
+      "loss": 0.3556,
+      "step": 21474
+    },
+    {
+      "epoch": 2.7953924248340494,
+      "grad_norm": 3.181854724884033,
+      "learning_rate": 1.1944103426793808e-07,
+      "loss": 0.2727,
+      "step": 21477
+    },
+    {
+      "epoch": 2.79578289730574,
+      "grad_norm": 2.986677885055542,
+      "learning_rate": 1.1898740254706942e-07,
+      "loss": 0.3345,
+      "step": 21480
+    },
+    {
+      "epoch": 2.7961733697774305,
+      "grad_norm": 2.7655344009399414,
+      "learning_rate": 1.1853462352883371e-07,
+      "loss": 0.4025,
+      "step": 21483
+    },
+    {
+      "epoch": 2.7965638422491215,
+      "grad_norm": 2.599412202835083,
+      "learning_rate": 1.180826972923299e-07,
+      "loss": 0.3651,
+      "step": 21486
+    },
+    {
+      "epoch": 2.796954314720812,
+      "grad_norm": 3.225376844406128,
+      "learning_rate": 1.1763162391650929e-07,
+      "loss": 0.3836,
+      "step": 21489
+    },
+    {
+      "epoch": 2.797344787192503,
+      "grad_norm": 2.734292507171631,
+      "learning_rate": 1.1718140348017271e-07,
+      "loss": 0.3683,
+      "step": 21492
+    },
+    {
+      "epoch": 2.7977352596641936,
+      "grad_norm": 2.984931230545044,
+      "learning_rate": 1.1673203606197336e-07,
+      "loss": 0.3585,
+      "step": 21495
+    },
+    {
+      "epoch": 2.7981257321358846,
+      "grad_norm": 2.9327147006988525,
+      "learning_rate": 1.1628352174041346e-07,
+      "loss": 0.4024,
+      "step": 21498
+    },
+    {
+      "epoch": 2.798516204607575,
+      "grad_norm": 2.466592788696289,
+      "learning_rate": 1.1583586059384921e-07,
+      "loss": 0.33,
+      "step": 21501
+    },
+    {
+      "epoch": 2.798906677079266,
+      "grad_norm": 2.7316441535949707,
+      "learning_rate": 1.1538905270048528e-07,
+      "loss": 0.312,
+      "step": 21504
+    },
+    {
+      "epoch": 2.7992971495509567,
+      "grad_norm": 2.5499911308288574,
+      "learning_rate": 1.1494309813837756e-07,
+      "loss": 0.3187,
+      "step": 21507
+    },
+    {
+      "epoch": 2.799687622022647,
+      "grad_norm": 2.7340035438537598,
+      "learning_rate": 1.1449799698543429e-07,
+      "loss": 0.3893,
+      "step": 21510
+    },
+    {
+      "epoch": 2.800078094494338,
+      "grad_norm": 2.626269578933716,
+      "learning_rate": 1.1405374931941382e-07,
+      "loss": 0.3604,
+      "step": 21513
+    },
+    {
+      "epoch": 2.8004685669660287,
+      "grad_norm": 2.762885808944702,
+      "learning_rate": 1.1361035521792407e-07,
+      "loss": 0.3397,
+      "step": 21516
+    },
+    {
+      "epoch": 2.8008590394377197,
+      "grad_norm": 2.9258060455322266,
+      "learning_rate": 1.1316781475842586e-07,
+      "loss": 0.3689,
+      "step": 21519
+    },
+    {
+      "epoch": 2.8012495119094103,
+      "grad_norm": 2.73075008392334,
+      "learning_rate": 1.1272612801823069e-07,
+      "loss": 0.3302,
+      "step": 21522
+    },
+    {
+      "epoch": 2.8016399843811013,
+      "grad_norm": 2.7661848068237305,
+      "learning_rate": 1.1228529507449904e-07,
+      "loss": 0.4068,
+      "step": 21525
+    },
+    {
+      "epoch": 2.802030456852792,
+      "grad_norm": 2.8157827854156494,
+      "learning_rate": 1.1184531600424431e-07,
+      "loss": 0.3384,
+      "step": 21528
+    },
+    {
+      "epoch": 2.802420929324483,
+      "grad_norm": 2.6738975048065186,
+      "learning_rate": 1.1140619088432946e-07,
+      "loss": 0.3366,
+      "step": 21531
+    },
+    {
+      "epoch": 2.8028114017961734,
+      "grad_norm": 3.107644557952881,
+      "learning_rate": 1.1096791979146981e-07,
+      "loss": 0.3749,
+      "step": 21534
+    },
+    {
+      "epoch": 2.803201874267864,
+      "grad_norm": 2.7247238159179688,
+      "learning_rate": 1.1053050280222855e-07,
+      "loss": 0.3524,
+      "step": 21537
+    },
+    {
+      "epoch": 2.803592346739555,
+      "grad_norm": 2.4612529277801514,
+      "learning_rate": 1.1009393999302287e-07,
+      "loss": 0.3645,
+      "step": 21540
+    },
+    {
+      "epoch": 2.8039828192112455,
+      "grad_norm": 2.981336832046509,
+      "learning_rate": 1.09658231440119e-07,
+      "loss": 0.4023,
+      "step": 21543
+    },
+    {
+      "epoch": 2.8043732916829365,
+      "grad_norm": 2.6686904430389404,
+      "learning_rate": 1.0922337721963494e-07,
+      "loss": 0.3697,
+      "step": 21546
+    },
+    {
+      "epoch": 2.804763764154627,
+      "grad_norm": 2.5626204013824463,
+      "learning_rate": 1.0878937740753714e-07,
+      "loss": 0.3182,
+      "step": 21549
+    },
+    {
+      "epoch": 2.8051542366263176,
+      "grad_norm": 3.065051317214966,
+      "learning_rate": 1.0835623207964607e-07,
+      "loss": 0.3935,
+      "step": 21552
+    },
+    {
+      "epoch": 2.8055447090980086,
+      "grad_norm": 4.100696086883545,
+      "learning_rate": 1.0792394131163064e-07,
+      "loss": 0.3165,
+      "step": 21555
+    },
+    {
+      "epoch": 2.8059351815696996,
+      "grad_norm": 2.8829846382141113,
+      "learning_rate": 1.07492505179011e-07,
+      "loss": 0.3021,
+      "step": 21558
+    },
+    {
+      "epoch": 2.80632565404139,
+      "grad_norm": 2.999645471572876,
+      "learning_rate": 1.0706192375715851e-07,
+      "loss": 0.3379,
+      "step": 21561
+    },
+    {
+      "epoch": 2.8067161265130807,
+      "grad_norm": 2.903561592102051,
+      "learning_rate": 1.0663219712129469e-07,
+      "loss": 0.4568,
+      "step": 21564
+    },
+    {
+      "epoch": 2.8071065989847717,
+      "grad_norm": 2.6506104469299316,
+      "learning_rate": 1.0620332534649225e-07,
+      "loss": 0.3272,
+      "step": 21567
+    },
+    {
+      "epoch": 2.807497071456462,
+      "grad_norm": 2.9835636615753174,
+      "learning_rate": 1.0577530850767348e-07,
+      "loss": 0.3592,
+      "step": 21570
+    },
+    {
+      "epoch": 2.807887543928153,
+      "grad_norm": 2.599337100982666,
+      "learning_rate": 1.0534814667961246e-07,
+      "loss": 0.3549,
+      "step": 21573
+    },
+    {
+      "epoch": 2.8082780163998438,
+      "grad_norm": 2.8641128540039062,
+      "learning_rate": 1.0492183993693394e-07,
+      "loss": 0.3638,
+      "step": 21576
+    },
+    {
+      "epoch": 2.8086684888715343,
+      "grad_norm": 2.740431547164917,
+      "learning_rate": 1.0449638835411114e-07,
+      "loss": 0.3511,
+      "step": 21579
+    },
+    {
+      "epoch": 2.8090589613432253,
+      "grad_norm": 2.9625089168548584,
+      "learning_rate": 1.0407179200547124e-07,
+      "loss": 0.4805,
+      "step": 21582
+    },
+    {
+      "epoch": 2.8094494338149163,
+      "grad_norm": 2.8378477096557617,
+      "learning_rate": 1.0364805096518993e-07,
+      "loss": 0.349,
+      "step": 21585
+    },
+    {
+      "epoch": 2.809839906286607,
+      "grad_norm": 2.665264368057251,
+      "learning_rate": 1.0322516530729298e-07,
+      "loss": 0.3023,
+      "step": 21588
+    },
+    {
+      "epoch": 2.8102303787582974,
+      "grad_norm": 2.7869749069213867,
+      "learning_rate": 1.0280313510565909e-07,
+      "loss": 0.3308,
+      "step": 21591
+    },
+    {
+      "epoch": 2.8106208512299884,
+      "grad_norm": 3.251950740814209,
+      "learning_rate": 1.0238196043401538e-07,
+      "loss": 0.2892,
+      "step": 21594
+    },
+    {
+      "epoch": 2.811011323701679,
+      "grad_norm": 2.933887004852295,
+      "learning_rate": 1.0196164136594022e-07,
+      "loss": 0.366,
+      "step": 21597
+    },
+    {
+      "epoch": 2.81140179617337,
+      "grad_norm": 2.547628164291382,
+      "learning_rate": 1.0154217797486098e-07,
+      "loss": 0.3066,
+      "step": 21600
+    },
+    {
+      "epoch": 2.8117922686450605,
+      "grad_norm": 2.8008322715759277,
+      "learning_rate": 1.0112357033405962e-07,
+      "loss": 0.3796,
+      "step": 21603
+    },
+    {
+      "epoch": 2.812182741116751,
+      "grad_norm": 2.5808677673339844,
+      "learning_rate": 1.0070581851666428e-07,
+      "loss": 0.3241,
+      "step": 21606
+    },
+    {
+      "epoch": 2.812573213588442,
+      "grad_norm": 2.5986881256103516,
+      "learning_rate": 1.0028892259565658e-07,
+      "loss": 0.3361,
+      "step": 21609
+    },
+    {
+      "epoch": 2.812963686060133,
+      "grad_norm": 2.557717800140381,
+      "learning_rate": 9.987288264386552e-08,
+      "loss": 0.346,
+      "step": 21612
+    },
+    {
+      "epoch": 2.8133541585318236,
+      "grad_norm": 2.836681842803955,
+      "learning_rate": 9.945769873397404e-08,
+      "loss": 0.3852,
+      "step": 21615
+    },
+    {
+      "epoch": 2.813744631003514,
+      "grad_norm": 2.740900754928589,
+      "learning_rate": 9.904337093851358e-08,
+      "loss": 0.402,
+      "step": 21618
+    },
+    {
+      "epoch": 2.814135103475205,
+      "grad_norm": 2.7481210231781006,
+      "learning_rate": 9.862989932986566e-08,
+      "loss": 0.347,
+      "step": 21621
+    },
+    {
+      "epoch": 2.8145255759468957,
+      "grad_norm": 2.67153000831604,
+      "learning_rate": 9.82172839802642e-08,
+      "loss": 0.3092,
+      "step": 21624
+    },
+    {
+      "epoch": 2.8149160484185867,
+      "grad_norm": 2.9304611682891846,
+      "learning_rate": 9.780552496179096e-08,
+      "loss": 0.3068,
+      "step": 21627
+    },
+    {
+      "epoch": 2.815306520890277,
+      "grad_norm": 2.806035280227661,
+      "learning_rate": 9.739462234637898e-08,
+      "loss": 0.3729,
+      "step": 21630
+    },
+    {
+      "epoch": 2.8156969933619678,
+      "grad_norm": 3.032376766204834,
+      "learning_rate": 9.698457620581359e-08,
+      "loss": 0.3683,
+      "step": 21633
+    },
+    {
+      "epoch": 2.8160874658336588,
+      "grad_norm": 3.0666725635528564,
+      "learning_rate": 9.657538661172861e-08,
+      "loss": 0.3887,
+      "step": 21636
+    },
+    {
+      "epoch": 2.8164779383053493,
+      "grad_norm": 2.8096840381622314,
+      "learning_rate": 9.616705363560796e-08,
+      "loss": 0.3416,
+      "step": 21639
+    },
+    {
+      "epoch": 2.8168684107770403,
+      "grad_norm": 2.5482988357543945,
+      "learning_rate": 9.575957734878627e-08,
+      "loss": 0.3262,
+      "step": 21642
+    },
+    {
+      "epoch": 2.817258883248731,
+      "grad_norm": 2.7446038722991943,
+      "learning_rate": 9.535295782245046e-08,
+      "loss": 0.3893,
+      "step": 21645
+    },
+    {
+      "epoch": 2.817649355720422,
+      "grad_norm": 2.8740649223327637,
+      "learning_rate": 9.494719512763429e-08,
+      "loss": 0.3946,
+      "step": 21648
+    },
+    {
+      "epoch": 2.8180398281921124,
+      "grad_norm": 2.647369623184204,
+      "learning_rate": 9.454228933522491e-08,
+      "loss": 0.3482,
+      "step": 21651
+    },
+    {
+      "epoch": 2.8184303006638034,
+      "grad_norm": 2.5511598587036133,
+      "learning_rate": 9.413824051595799e-08,
+      "loss": 0.315,
+      "step": 21654
+    },
+    {
+      "epoch": 2.818820773135494,
+      "grad_norm": 2.809553623199463,
+      "learning_rate": 9.373504874041984e-08,
+      "loss": 0.4283,
+      "step": 21657
+    },
+    {
+      "epoch": 2.8192112456071845,
+      "grad_norm": 2.821474075317383,
+      "learning_rate": 9.333271407904743e-08,
+      "loss": 0.3556,
+      "step": 21660
+    },
+    {
+      "epoch": 2.8196017180788755,
+      "grad_norm": 2.612962484359741,
+      "learning_rate": 9.293123660212733e-08,
+      "loss": 0.3268,
+      "step": 21663
+    },
+    {
+      "epoch": 2.819992190550566,
+      "grad_norm": 3.18330454826355,
+      "learning_rate": 9.253061637979788e-08,
+      "loss": 0.3513,
+      "step": 21666
+    },
+    {
+      "epoch": 2.820382663022257,
+      "grad_norm": 2.551689624786377,
+      "learning_rate": 9.213085348204587e-08,
+      "loss": 0.3637,
+      "step": 21669
+    },
+    {
+      "epoch": 2.8207731354939476,
+      "grad_norm": 2.9806084632873535,
+      "learning_rate": 9.173194797870877e-08,
+      "loss": 0.4023,
+      "step": 21672
+    },
+    {
+      "epoch": 2.8211636079656386,
+      "grad_norm": 2.7304067611694336,
+      "learning_rate": 9.133389993947528e-08,
+      "loss": 0.3305,
+      "step": 21675
+    },
+    {
+      "epoch": 2.821554080437329,
+      "grad_norm": 3.107208013534546,
+      "learning_rate": 9.09367094338831e-08,
+      "loss": 0.3507,
+      "step": 21678
+    },
+    {
+      "epoch": 2.82194455290902,
+      "grad_norm": 2.694958448410034,
+      "learning_rate": 9.054037653132008e-08,
+      "loss": 0.3679,
+      "step": 21681
+    },
+    {
+      "epoch": 2.8223350253807107,
+      "grad_norm": 2.6622893810272217,
+      "learning_rate": 9.014490130102527e-08,
+      "loss": 0.3641,
+      "step": 21684
+    },
+    {
+      "epoch": 2.8227254978524012,
+      "grad_norm": 2.8670809268951416,
+      "learning_rate": 8.975028381208784e-08,
+      "loss": 0.3707,
+      "step": 21687
+    },
+    {
+      "epoch": 2.8231159703240922,
+      "grad_norm": 2.79952335357666,
+      "learning_rate": 8.935652413344598e-08,
+      "loss": 0.4008,
+      "step": 21690
+    },
+    {
+      "epoch": 2.823506442795783,
+      "grad_norm": 3.0517783164978027,
+      "learning_rate": 8.89636223338891e-08,
+      "loss": 0.4025,
+      "step": 21693
+    },
+    {
+      "epoch": 2.8238969152674738,
+      "grad_norm": 2.722447395324707,
+      "learning_rate": 8.857157848205566e-08,
+      "loss": 0.344,
+      "step": 21696
+    },
+    {
+      "epoch": 2.8242873877391643,
+      "grad_norm": 2.9243016242980957,
+      "learning_rate": 8.818039264643586e-08,
+      "loss": 0.3724,
+      "step": 21699
+    },
+    {
+      "epoch": 2.824677860210855,
+      "grad_norm": 3.2354788780212402,
+      "learning_rate": 8.779006489536834e-08,
+      "loss": 0.3583,
+      "step": 21702
+    },
+    {
+      "epoch": 2.825068332682546,
+      "grad_norm": 2.571131706237793,
+      "learning_rate": 8.740059529704248e-08,
+      "loss": 0.3751,
+      "step": 21705
+    },
+    {
+      "epoch": 2.825458805154237,
+      "grad_norm": 3.589383363723755,
+      "learning_rate": 8.701198391949827e-08,
+      "loss": 0.3349,
+      "step": 21708
+    },
+    {
+      "epoch": 2.8258492776259274,
+      "grad_norm": 2.8357677459716797,
+      "learning_rate": 8.662423083062532e-08,
+      "loss": 0.3465,
+      "step": 21711
+    },
+    {
+      "epoch": 2.826239750097618,
+      "grad_norm": 3.248215913772583,
+      "learning_rate": 8.62373360981622e-08,
+      "loss": 0.3655,
+      "step": 21714
+    },
+    {
+      "epoch": 2.826630222569309,
+      "grad_norm": 2.9728636741638184,
+      "learning_rate": 8.585129978969986e-08,
+      "loss": 0.3857,
+      "step": 21717
+    },
+    {
+      "epoch": 2.8270206950409995,
+      "grad_norm": 3.6474063396453857,
+      "learning_rate": 8.546612197267768e-08,
+      "loss": 0.3819,
+      "step": 21720
+    },
+    {
+      "epoch": 2.8274111675126905,
+      "grad_norm": 2.7672176361083984,
+      "learning_rate": 8.508180271438516e-08,
+      "loss": 0.4161,
+      "step": 21723
+    },
+    {
+      "epoch": 2.827801639984381,
+      "grad_norm": 2.943279504776001,
+      "learning_rate": 8.469834208196193e-08,
+      "loss": 0.3341,
+      "step": 21726
+    },
+    {
+      "epoch": 2.8281921124560716,
+      "grad_norm": 3.383356809616089,
+      "learning_rate": 8.431574014239885e-08,
+      "loss": 0.3586,
+      "step": 21729
+    },
+    {
+      "epoch": 2.8285825849277626,
+      "grad_norm": 2.866734266281128,
+      "learning_rate": 8.393399696253412e-08,
+      "loss": 0.3224,
+      "step": 21732
+    },
+    {
+      "epoch": 2.8289730573994536,
+      "grad_norm": 2.8714816570281982,
+      "learning_rate": 8.355311260905829e-08,
+      "loss": 0.416,
+      "step": 21735
+    },
+    {
+      "epoch": 2.829363529871144,
+      "grad_norm": 2.4566404819488525,
+      "learning_rate": 8.317308714851146e-08,
+      "loss": 0.3209,
+      "step": 21738
+    },
+    {
+      "epoch": 2.8297540023428347,
+      "grad_norm": 3.835968255996704,
+      "learning_rate": 8.279392064728276e-08,
+      "loss": 0.3415,
+      "step": 21741
+    },
+    {
+      "epoch": 2.8301444748145257,
+      "grad_norm": 2.9507713317871094,
+      "learning_rate": 8.24156131716114e-08,
+      "loss": 0.3388,
+      "step": 21744
+    },
+    {
+      "epoch": 2.8305349472862162,
+      "grad_norm": 2.818152904510498,
+      "learning_rate": 8.203816478758785e-08,
+      "loss": 0.3826,
+      "step": 21747
+    },
+    {
+      "epoch": 2.8309254197579072,
+      "grad_norm": 2.5711591243743896,
+      "learning_rate": 8.166157556115107e-08,
+      "loss": 0.3195,
+      "step": 21750
+    },
+    {
+      "epoch": 2.831315892229598,
+      "grad_norm": 3.0602824687957764,
+      "learning_rate": 8.128584555809005e-08,
+      "loss": 0.333,
+      "step": 21753
+    },
+    {
+      "epoch": 2.8317063647012883,
+      "grad_norm": 2.9426445960998535,
+      "learning_rate": 8.091097484404454e-08,
+      "loss": 0.3469,
+      "step": 21756
+    },
+    {
+      "epoch": 2.8320968371729793,
+      "grad_norm": 2.615579843521118,
+      "learning_rate": 8.053696348450324e-08,
+      "loss": 0.3241,
+      "step": 21759
+    },
+    {
+      "epoch": 2.8324873096446703,
+      "grad_norm": 2.8304200172424316,
+      "learning_rate": 8.016381154480557e-08,
+      "loss": 0.3492,
+      "step": 21762
+    },
+    {
+      "epoch": 2.832877782116361,
+      "grad_norm": 2.8419861793518066,
+      "learning_rate": 7.979151909013993e-08,
+      "loss": 0.3959,
+      "step": 21765
+    },
+    {
+      "epoch": 2.8332682545880514,
+      "grad_norm": 3.1419153213500977,
+      "learning_rate": 7.942008618554543e-08,
+      "loss": 0.3467,
+      "step": 21768
+    },
+    {
+      "epoch": 2.8336587270597424,
+      "grad_norm": 2.6105754375457764,
+      "learning_rate": 7.904951289591068e-08,
+      "loss": 0.3219,
+      "step": 21771
+    },
+    {
+      "epoch": 2.834049199531433,
+      "grad_norm": 2.874232530593872,
+      "learning_rate": 7.867979928597336e-08,
+      "loss": 0.4414,
+      "step": 21774
+    },
+    {
+      "epoch": 2.834439672003124,
+      "grad_norm": 2.67608642578125,
+      "learning_rate": 7.831094542032236e-08,
+      "loss": 0.3289,
+      "step": 21777
+    },
+    {
+      "epoch": 2.8348301444748145,
+      "grad_norm": 2.974731922149658,
+      "learning_rate": 7.794295136339613e-08,
+      "loss": 0.3602,
+      "step": 21780
+    },
+    {
+      "epoch": 2.835220616946505,
+      "grad_norm": 2.8034987449645996,
+      "learning_rate": 7.757581717948104e-08,
+      "loss": 0.4196,
+      "step": 21783
+    },
+    {
+      "epoch": 2.835611089418196,
+      "grad_norm": 2.63897442817688,
+      "learning_rate": 7.720954293271576e-08,
+      "loss": 0.3953,
+      "step": 21786
+    },
+    {
+      "epoch": 2.8360015618898866,
+      "grad_norm": 2.5943167209625244,
+      "learning_rate": 7.68441286870869e-08,
+      "loss": 0.2894,
+      "step": 21789
+    },
+    {
+      "epoch": 2.8363920343615776,
+      "grad_norm": 2.8868579864501953,
+      "learning_rate": 7.647957450643284e-08,
+      "loss": 0.4103,
+      "step": 21792
+    },
+    {
+      "epoch": 2.836782506833268,
+      "grad_norm": 2.7333850860595703,
+      "learning_rate": 7.611588045443874e-08,
+      "loss": 0.3394,
+      "step": 21795
+    },
+    {
+      "epoch": 2.837172979304959,
+      "grad_norm": 2.8235933780670166,
+      "learning_rate": 7.575304659464211e-08,
+      "loss": 0.3461,
+      "step": 21798
+    },
+    {
+      "epoch": 2.8375634517766497,
+      "grad_norm": 2.699286699295044,
+      "learning_rate": 7.539107299042947e-08,
+      "loss": 0.3186,
+      "step": 21801
+    },
+    {
+      "epoch": 2.8379539242483407,
+      "grad_norm": 2.984445095062256,
+      "learning_rate": 7.502995970503634e-08,
+      "loss": 0.3464,
+      "step": 21804
+    },
+    {
+      "epoch": 2.8383443967200312,
+      "grad_norm": 2.8218915462493896,
+      "learning_rate": 7.466970680154839e-08,
+      "loss": 0.3115,
+      "step": 21807
+    },
+    {
+      "epoch": 2.838734869191722,
+      "grad_norm": 2.748157262802124,
+      "learning_rate": 7.431031434290137e-08,
+      "loss": 0.3772,
+      "step": 21810
+    },
+    {
+      "epoch": 2.839125341663413,
+      "grad_norm": 2.9761040210723877,
+      "learning_rate": 7.395178239188006e-08,
+      "loss": 0.4132,
+      "step": 21813
+    },
+    {
+      "epoch": 2.8395158141351033,
+      "grad_norm": 2.9955594539642334,
+      "learning_rate": 7.359411101111991e-08,
+      "loss": 0.3501,
+      "step": 21816
+    },
+    {
+      "epoch": 2.8399062866067943,
+      "grad_norm": 2.834073305130005,
+      "learning_rate": 7.323730026310483e-08,
+      "loss": 0.3444,
+      "step": 21819
+    },
+    {
+      "epoch": 2.840296759078485,
+      "grad_norm": 2.5343804359436035,
+      "learning_rate": 7.28813502101694e-08,
+      "loss": 0.3856,
+      "step": 21822
+    },
+    {
+      "epoch": 2.8406872315501754,
+      "grad_norm": 2.5103886127471924,
+      "learning_rate": 7.252626091449666e-08,
+      "loss": 0.3501,
+      "step": 21825
+    },
+    {
+      "epoch": 2.8410777040218664,
+      "grad_norm": 2.7703588008880615,
+      "learning_rate": 7.217203243811976e-08,
+      "loss": 0.3207,
+      "step": 21828
+    },
+    {
+      "epoch": 2.8414681764935574,
+      "grad_norm": 2.8706281185150146,
+      "learning_rate": 7.181866484292255e-08,
+      "loss": 0.36,
+      "step": 21831
+    },
+    {
+      "epoch": 2.841858648965248,
+      "grad_norm": 2.7826225757598877,
+      "learning_rate": 7.146615819063729e-08,
+      "loss": 0.3871,
+      "step": 21834
+    },
+    {
+      "epoch": 2.8422491214369385,
+      "grad_norm": 2.82706880569458,
+      "learning_rate": 7.111451254284584e-08,
+      "loss": 0.3323,
+      "step": 21837
+    },
+    {
+      "epoch": 2.8426395939086295,
+      "grad_norm": 2.9153218269348145,
+      "learning_rate": 7.076372796098019e-08,
+      "loss": 0.3573,
+      "step": 21840
+    },
+    {
+      "epoch": 2.84303006638032,
+      "grad_norm": 2.676373243331909,
+      "learning_rate": 7.041380450632185e-08,
+      "loss": 0.3178,
+      "step": 21843
+    },
+    {
+      "epoch": 2.843420538852011,
+      "grad_norm": 2.866745948791504,
+      "learning_rate": 7.006474224000138e-08,
+      "loss": 0.3973,
+      "step": 21846
+    },
+    {
+      "epoch": 2.8438110113237016,
+      "grad_norm": 2.364248514175415,
+      "learning_rate": 6.971654122299943e-08,
+      "loss": 0.3069,
+      "step": 21849
+    },
+    {
+      "epoch": 2.844201483795392,
+      "grad_norm": 2.777941942214966,
+      "learning_rate": 6.936920151614567e-08,
+      "loss": 0.2974,
+      "step": 21852
+    },
+    {
+      "epoch": 2.844591956267083,
+      "grad_norm": 2.982527017593384,
+      "learning_rate": 6.90227231801205e-08,
+      "loss": 0.3763,
+      "step": 21855
+    },
+    {
+      "epoch": 2.844982428738774,
+      "grad_norm": 2.7366418838500977,
+      "learning_rate": 6.867710627545154e-08,
+      "loss": 0.3435,
+      "step": 21858
+    },
+    {
+      "epoch": 2.8453729012104647,
+      "grad_norm": 3.094679832458496,
+      "learning_rate": 6.833235086251889e-08,
+      "loss": 0.3818,
+      "step": 21861
+    },
+    {
+      "epoch": 2.8457633736821553,
+      "grad_norm": 2.775191068649292,
+      "learning_rate": 6.798845700154987e-08,
+      "loss": 0.3325,
+      "step": 21864
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 2.8869082927703857,
+      "learning_rate": 6.764542475262147e-08,
+      "loss": 0.3708,
+      "step": 21867
+    },
+    {
+      "epoch": 2.846544318625537,
+      "grad_norm": 3.177225112915039,
+      "learning_rate": 6.73032541756613e-08,
+      "loss": 0.3381,
+      "step": 21870
+    },
+    {
+      "epoch": 2.846934791097228,
+      "grad_norm": 2.946268081665039,
+      "learning_rate": 6.696194533044598e-08,
+      "loss": 0.3494,
+      "step": 21873
+    },
+    {
+      "epoch": 2.8473252635689184,
+      "grad_norm": 3.021143674850464,
+      "learning_rate": 6.662149827660114e-08,
+      "loss": 0.3219,
+      "step": 21876
+    },
+    {
+      "epoch": 2.847715736040609,
+      "grad_norm": 2.826537847518921,
+      "learning_rate": 6.628191307360199e-08,
+      "loss": 0.341,
+      "step": 21879
+    },
+    {
+      "epoch": 2.8481062085123,
+      "grad_norm": 2.8287034034729004,
+      "learning_rate": 6.594318978077386e-08,
+      "loss": 0.3956,
+      "step": 21882
+    },
+    {
+      "epoch": 2.848496680983991,
+      "grad_norm": 3.175198793411255,
+      "learning_rate": 6.56053284572905e-08,
+      "loss": 0.2685,
+      "step": 21885
+    },
+    {
+      "epoch": 2.8488871534556814,
+      "grad_norm": 2.6262009143829346,
+      "learning_rate": 6.526832916217584e-08,
+      "loss": 0.374,
+      "step": 21888
+    },
+    {
+      "epoch": 2.849277625927372,
+      "grad_norm": 2.9307057857513428,
+      "learning_rate": 6.493219195430334e-08,
+      "loss": 0.4002,
+      "step": 21891
+    },
+    {
+      "epoch": 2.849668098399063,
+      "grad_norm": 2.8721837997436523,
+      "learning_rate": 6.459691689239433e-08,
+      "loss": 0.3986,
+      "step": 21894
+    },
+    {
+      "epoch": 2.8500585708707535,
+      "grad_norm": 2.7488579750061035,
+      "learning_rate": 6.426250403502199e-08,
+      "loss": 0.333,
+      "step": 21897
+    },
+    {
+      "epoch": 2.8504490433424445,
+      "grad_norm": 2.9429168701171875,
+      "learning_rate": 6.392895344060623e-08,
+      "loss": 0.3574,
+      "step": 21900
+    },
+    {
+      "epoch": 2.850839515814135,
+      "grad_norm": 2.7157819271087646,
+      "learning_rate": 6.359626516741935e-08,
+      "loss": 0.3414,
+      "step": 21903
+    },
+    {
+      "epoch": 2.8512299882858256,
+      "grad_norm": 2.7233452796936035,
+      "learning_rate": 6.326443927357983e-08,
+      "loss": 0.3299,
+      "step": 21906
+    },
+    {
+      "epoch": 2.8516204607575166,
+      "grad_norm": 2.926488161087036,
+      "learning_rate": 6.293347581705689e-08,
+      "loss": 0.3944,
+      "step": 21909
+    },
+    {
+      "epoch": 2.852010933229207,
+      "grad_norm": 2.6980791091918945,
+      "learning_rate": 6.260337485567037e-08,
+      "loss": 0.3758,
+      "step": 21912
+    },
+    {
+      "epoch": 2.852401405700898,
+      "grad_norm": 2.777859926223755,
+      "learning_rate": 6.227413644708691e-08,
+      "loss": 0.2841,
+      "step": 21915
+    },
+    {
+      "epoch": 2.8527918781725887,
+      "grad_norm": 3.123918056488037,
+      "learning_rate": 6.194576064882496e-08,
+      "loss": 0.365,
+      "step": 21918
+    },
+    {
+      "epoch": 2.8531823506442797,
+      "grad_norm": 2.804738998413086,
+      "learning_rate": 6.161824751824974e-08,
+      "loss": 0.3655,
+      "step": 21921
+    },
+    {
+      "epoch": 2.8535728231159703,
+      "grad_norm": 2.6704933643341064,
+      "learning_rate": 6.129159711257826e-08,
+      "loss": 0.2882,
+      "step": 21924
+    },
+    {
+      "epoch": 2.8539632955876613,
+      "grad_norm": 2.9849040508270264,
+      "learning_rate": 6.096580948887543e-08,
+      "loss": 0.3722,
+      "step": 21927
+    },
+    {
+      "epoch": 2.854353768059352,
+      "grad_norm": 2.9017884731292725,
+      "learning_rate": 6.064088470405516e-08,
+      "loss": 0.3921,
+      "step": 21930
+    },
+    {
+      "epoch": 2.8547442405310424,
+      "grad_norm": 2.6805455684661865,
+      "learning_rate": 6.03168228148815e-08,
+      "loss": 0.3453,
+      "step": 21933
+    },
+    {
+      "epoch": 2.8551347130027334,
+      "grad_norm": 2.8592615127563477,
+      "learning_rate": 5.999362387796747e-08,
+      "loss": 0.3311,
+      "step": 21936
+    },
+    {
+      "epoch": 2.855525185474424,
+      "grad_norm": 3.1579458713531494,
+      "learning_rate": 5.967128794977462e-08,
+      "loss": 0.3666,
+      "step": 21939
+    },
+    {
+      "epoch": 2.855915657946115,
+      "grad_norm": 2.6894073486328125,
+      "learning_rate": 5.9349815086615084e-08,
+      "loss": 0.3648,
+      "step": 21942
+    },
+    {
+      "epoch": 2.8563061304178055,
+      "grad_norm": 2.765660047531128,
+      "learning_rate": 5.902920534464951e-08,
+      "loss": 0.3761,
+      "step": 21945
+    },
+    {
+      "epoch": 2.8566966028894965,
+      "grad_norm": 2.739934206008911,
+      "learning_rate": 5.870945877988754e-08,
+      "loss": 0.293,
+      "step": 21948
+    },
+    {
+      "epoch": 2.857087075361187,
+      "grad_norm": 3.662998914718628,
+      "learning_rate": 5.839057544818783e-08,
+      "loss": 0.3738,
+      "step": 21951
+    },
+    {
+      "epoch": 2.857477547832878,
+      "grad_norm": 2.789340019226074,
+      "learning_rate": 5.8072555405259135e-08,
+      "loss": 0.3837,
+      "step": 21954
+    },
+    {
+      "epoch": 2.8578680203045685,
+      "grad_norm": 2.4492013454437256,
+      "learning_rate": 5.7755398706658694e-08,
+      "loss": 0.3126,
+      "step": 21957
+    },
+    {
+      "epoch": 2.858258492776259,
+      "grad_norm": 2.5558364391326904,
+      "learning_rate": 5.7439105407792736e-08,
+      "loss": 0.31,
+      "step": 21960
+    },
+    {
+      "epoch": 2.85864896524795,
+      "grad_norm": 2.8237719535827637,
+      "learning_rate": 5.712367556391818e-08,
+      "loss": 0.3935,
+      "step": 21963
+    },
+    {
+      "epoch": 2.8590394377196406,
+      "grad_norm": 2.918532371520996,
+      "learning_rate": 5.6809109230138714e-08,
+      "loss": 0.3618,
+      "step": 21966
+    },
+    {
+      "epoch": 2.8594299101913316,
+      "grad_norm": 3.382922410964966,
+      "learning_rate": 5.6495406461409274e-08,
+      "loss": 0.3443,
+      "step": 21969
+    },
+    {
+      "epoch": 2.859820382663022,
+      "grad_norm": 2.8826119899749756,
+      "learning_rate": 5.6182567312532134e-08,
+      "loss": 0.4316,
+      "step": 21972
+    },
+    {
+      "epoch": 2.8602108551347127,
+      "grad_norm": 2.8180668354034424,
+      "learning_rate": 5.587059183816079e-08,
+      "loss": 0.3589,
+      "step": 21975
+    },
+    {
+      "epoch": 2.8606013276064037,
+      "grad_norm": 2.677135705947876,
+      "learning_rate": 5.5559480092795545e-08,
+      "loss": 0.3745,
+      "step": 21978
+    },
+    {
+      "epoch": 2.8609918000780947,
+      "grad_norm": 2.9708173274993896,
+      "learning_rate": 5.5249232130787924e-08,
+      "loss": 0.3381,
+      "step": 21981
+    },
+    {
+      "epoch": 2.8613822725497853,
+      "grad_norm": 3.1857826709747314,
+      "learning_rate": 5.493984800633734e-08,
+      "loss": 0.3931,
+      "step": 21984
+    },
+    {
+      "epoch": 2.861772745021476,
+      "grad_norm": 3.0874578952789307,
+      "learning_rate": 5.4631327773492226e-08,
+      "loss": 0.3146,
+      "step": 21987
+    },
+    {
+      "epoch": 2.862163217493167,
+      "grad_norm": 2.952526807785034,
+      "learning_rate": 5.4323671486150585e-08,
+      "loss": 0.3709,
+      "step": 21990
+    },
+    {
+      "epoch": 2.8625536899648574,
+      "grad_norm": 2.7421607971191406,
+      "learning_rate": 5.401687919805942e-08,
+      "loss": 0.3305,
+      "step": 21993
+    },
+    {
+      "epoch": 2.8629441624365484,
+      "grad_norm": 3.3687474727630615,
+      "learning_rate": 5.3710950962814755e-08,
+      "loss": 0.4446,
+      "step": 21996
+    },
+    {
+      "epoch": 2.863334634908239,
+      "grad_norm": 3.39333176612854,
+      "learning_rate": 5.3405886833861054e-08,
+      "loss": 0.3432,
+      "step": 21999
+    },
+    {
+      "epoch": 2.8637251073799295,
+      "grad_norm": 2.7644834518432617,
+      "learning_rate": 5.3101686864492904e-08,
+      "loss": 0.3486,
+      "step": 22002
+    },
+    {
+      "epoch": 2.8641155798516205,
+      "grad_norm": 2.7170677185058594,
+      "learning_rate": 5.279835110785392e-08,
+      "loss": 0.3633,
+      "step": 22005
+    },
+    {
+      "epoch": 2.8645060523233115,
+      "grad_norm": 3.4381911754608154,
+      "learning_rate": 5.249587961693503e-08,
+      "loss": 0.3431,
+      "step": 22008
+    },
+    {
+      "epoch": 2.864896524795002,
+      "grad_norm": 2.583317756652832,
+      "learning_rate": 5.219427244457842e-08,
+      "loss": 0.3621,
+      "step": 22011
+    },
+    {
+      "epoch": 2.8652869972666926,
+      "grad_norm": 2.8179497718811035,
+      "learning_rate": 5.189352964347305e-08,
+      "loss": 0.3758,
+      "step": 22014
+    },
+    {
+      "epoch": 2.8656774697383836,
+      "grad_norm": 2.632049560546875,
+      "learning_rate": 5.159365126615967e-08,
+      "loss": 0.3725,
+      "step": 22017
+    },
+    {
+      "epoch": 2.866067942210074,
+      "grad_norm": 3.091775894165039,
+      "learning_rate": 5.129463736502471e-08,
+      "loss": 0.3826,
+      "step": 22020
+    },
+    {
+      "epoch": 2.866458414681765,
+      "grad_norm": 2.7316322326660156,
+      "learning_rate": 5.099648799230583e-08,
+      "loss": 0.3998,
+      "step": 22023
+    },
+    {
+      "epoch": 2.8668488871534556,
+      "grad_norm": 2.750880002975464,
+      "learning_rate": 5.0699203200089694e-08,
+      "loss": 0.4074,
+      "step": 22026
+    },
+    {
+      "epoch": 2.867239359625146,
+      "grad_norm": 3.477369785308838,
+      "learning_rate": 5.0402783040311435e-08,
+      "loss": 0.4091,
+      "step": 22029
+    },
+    {
+      "epoch": 2.867629832096837,
+      "grad_norm": 2.8512279987335205,
+      "learning_rate": 5.010722756475406e-08,
+      "loss": 0.3391,
+      "step": 22032
+    },
+    {
+      "epoch": 2.868020304568528,
+      "grad_norm": 2.6762948036193848,
+      "learning_rate": 4.981253682505127e-08,
+      "loss": 0.2929,
+      "step": 22035
+    },
+    {
+      "epoch": 2.8684107770402187,
+      "grad_norm": 3.086232900619507,
+      "learning_rate": 4.951871087268412e-08,
+      "loss": 0.3973,
+      "step": 22038
+    },
+    {
+      "epoch": 2.8688012495119093,
+      "grad_norm": 2.8311057090759277,
+      "learning_rate": 4.922574975898431e-08,
+      "loss": 0.3275,
+      "step": 22041
+    },
+    {
+      "epoch": 2.8691917219836003,
+      "grad_norm": 2.6848394870758057,
+      "learning_rate": 4.893365353513091e-08,
+      "loss": 0.3176,
+      "step": 22044
+    },
+    {
+      "epoch": 2.869582194455291,
+      "grad_norm": 2.7619194984436035,
+      "learning_rate": 4.86424222521531e-08,
+      "loss": 0.3016,
+      "step": 22047
+    },
+    {
+      "epoch": 2.869972666926982,
+      "grad_norm": 2.7256760597229004,
+      "learning_rate": 4.835205596092796e-08,
+      "loss": 0.3447,
+      "step": 22050
+    },
+    {
+      "epoch": 2.8703631393986724,
+      "grad_norm": 2.8061459064483643,
+      "learning_rate": 4.806255471218102e-08,
+      "loss": 0.3466,
+      "step": 22053
+    },
+    {
+      "epoch": 2.870753611870363,
+      "grad_norm": 2.5725607872009277,
+      "learning_rate": 4.7773918556489054e-08,
+      "loss": 0.3098,
+      "step": 22056
+    },
+    {
+      "epoch": 2.871144084342054,
+      "grad_norm": 2.703907012939453,
+      "learning_rate": 4.748614754427561e-08,
+      "loss": 0.3863,
+      "step": 22059
+    },
+    {
+      "epoch": 2.8715345568137445,
+      "grad_norm": 2.798107385635376,
+      "learning_rate": 4.71992417258138e-08,
+      "loss": 0.3247,
+      "step": 22062
+    },
+    {
+      "epoch": 2.8719250292854355,
+      "grad_norm": 2.4142208099365234,
+      "learning_rate": 4.691320115122466e-08,
+      "loss": 0.3145,
+      "step": 22065
+    },
+    {
+      "epoch": 2.872315501757126,
+      "grad_norm": 2.5614852905273438,
+      "learning_rate": 4.6628025870479875e-08,
+      "loss": 0.3519,
+      "step": 22068
+    },
+    {
+      "epoch": 2.872705974228817,
+      "grad_norm": 2.7241270542144775,
+      "learning_rate": 4.6343715933399034e-08,
+      "loss": 0.3275,
+      "step": 22071
+    },
+    {
+      "epoch": 2.8730964467005076,
+      "grad_norm": 2.801772356033325,
+      "learning_rate": 4.6060271389649635e-08,
+      "loss": 0.3661,
+      "step": 22074
+    },
+    {
+      "epoch": 2.8734869191721986,
+      "grad_norm": 2.6916000843048096,
+      "learning_rate": 4.577769228874873e-08,
+      "loss": 0.3467,
+      "step": 22077
+    },
+    {
+      "epoch": 2.873877391643889,
+      "grad_norm": 2.7385640144348145,
+      "learning_rate": 4.54959786800635e-08,
+      "loss": 0.3489,
+      "step": 22080
+    },
+    {
+      "epoch": 2.8742678641155797,
+      "grad_norm": 2.9576168060302734,
+      "learning_rate": 4.521513061280791e-08,
+      "loss": 0.333,
+      "step": 22083
+    },
+    {
+      "epoch": 2.8746583365872707,
+      "grad_norm": 2.8429853916168213,
+      "learning_rate": 4.4935148136045495e-08,
+      "loss": 0.3436,
+      "step": 22086
+    },
+    {
+      "epoch": 2.875048809058961,
+      "grad_norm": 2.8129966259002686,
+      "learning_rate": 4.465603129868934e-08,
+      "loss": 0.3853,
+      "step": 22089
+    },
+    {
+      "epoch": 2.875439281530652,
+      "grad_norm": 2.891261100769043,
+      "learning_rate": 4.437778014949934e-08,
+      "loss": 0.3138,
+      "step": 22092
+    },
+    {
+      "epoch": 2.8758297540023428,
+      "grad_norm": 2.6374096870422363,
+      "learning_rate": 4.410039473708605e-08,
+      "loss": 0.2911,
+      "step": 22095
+    },
+    {
+      "epoch": 2.8762202264740333,
+      "grad_norm": 2.888641119003296,
+      "learning_rate": 4.3823875109908486e-08,
+      "loss": 0.4262,
+      "step": 22098
+    },
+    {
+      "epoch": 2.8766106989457243,
+      "grad_norm": 3.0171844959259033,
+      "learning_rate": 4.354822131627357e-08,
+      "loss": 0.427,
+      "step": 22101
+    },
+    {
+      "epoch": 2.8770011714174153,
+      "grad_norm": 2.790205717086792,
+      "learning_rate": 4.327343340433721e-08,
+      "loss": 0.3196,
+      "step": 22104
+    },
+    {
+      "epoch": 2.877391643889106,
+      "grad_norm": 2.8684253692626953,
+      "learning_rate": 4.29995114221049e-08,
+      "loss": 0.3728,
+      "step": 22107
+    },
+    {
+      "epoch": 2.8777821163607964,
+      "grad_norm": 2.8882906436920166,
+      "learning_rate": 4.2726455417430037e-08,
+      "loss": 0.3488,
+      "step": 22110
+    },
+    {
+      "epoch": 2.8781725888324874,
+      "grad_norm": 2.429396152496338,
+      "learning_rate": 4.2454265438014454e-08,
+      "loss": 0.3179,
+      "step": 22113
+    },
+    {
+      "epoch": 2.878563061304178,
+      "grad_norm": 3.0140392780303955,
+      "learning_rate": 4.2182941531410113e-08,
+      "loss": 0.3496,
+      "step": 22116
+    },
+    {
+      "epoch": 2.878953533775869,
+      "grad_norm": 3.1088218688964844,
+      "learning_rate": 4.191248374501577e-08,
+      "loss": 0.3918,
+      "step": 22119
+    },
+    {
+      "epoch": 2.8793440062475595,
+      "grad_norm": 2.7448649406433105,
+      "learning_rate": 4.16428921260803e-08,
+      "loss": 0.4052,
+      "step": 22122
+    },
+    {
+      "epoch": 2.87973447871925,
+      "grad_norm": 2.797527551651001,
+      "learning_rate": 4.1374166721701026e-08,
+      "loss": 0.3408,
+      "step": 22125
+    },
+    {
+      "epoch": 2.880124951190941,
+      "grad_norm": 2.954578399658203,
+      "learning_rate": 4.110630757882261e-08,
+      "loss": 0.3704,
+      "step": 22128
+    },
+    {
+      "epoch": 2.880515423662632,
+      "grad_norm": 2.8430421352386475,
+      "learning_rate": 4.0839314744240966e-08,
+      "loss": 0.4094,
+      "step": 22131
+    },
+    {
+      "epoch": 2.8809058961343226,
+      "grad_norm": 2.5656564235687256,
+      "learning_rate": 4.057318826459822e-08,
+      "loss": 0.3717,
+      "step": 22134
+    },
+    {
+      "epoch": 2.881296368606013,
+      "grad_norm": 2.829988956451416,
+      "learning_rate": 4.0307928186386625e-08,
+      "loss": 0.3751,
+      "step": 22137
+    },
+    {
+      "epoch": 2.881686841077704,
+      "grad_norm": 2.584980010986328,
+      "learning_rate": 4.004353455594578e-08,
+      "loss": 0.3584,
+      "step": 22140
+    },
+    {
+      "epoch": 2.8820773135493947,
+      "grad_norm": 3.4724276065826416,
+      "learning_rate": 3.978000741946597e-08,
+      "loss": 0.3441,
+      "step": 22143
+    },
+    {
+      "epoch": 2.8824677860210857,
+      "grad_norm": 2.891096830368042,
+      "learning_rate": 3.951734682298314e-08,
+      "loss": 0.3937,
+      "step": 22146
+    },
+    {
+      "epoch": 2.882858258492776,
+      "grad_norm": 3.165163993835449,
+      "learning_rate": 3.9255552812385025e-08,
+      "loss": 0.3763,
+      "step": 22149
+    },
+    {
+      "epoch": 2.8832487309644668,
+      "grad_norm": 2.5909931659698486,
+      "learning_rate": 3.8994625433406155e-08,
+      "loss": 0.3113,
+      "step": 22152
+    },
+    {
+      "epoch": 2.8836392034361578,
+      "grad_norm": 2.9258785247802734,
+      "learning_rate": 3.873456473162951e-08,
+      "loss": 0.3515,
+      "step": 22155
+    },
+    {
+      "epoch": 2.8840296759078488,
+      "grad_norm": 2.447117805480957,
+      "learning_rate": 3.847537075248764e-08,
+      "loss": 0.333,
+      "step": 22158
+    },
+    {
+      "epoch": 2.8844201483795393,
+      "grad_norm": 2.608942985534668,
+      "learning_rate": 3.821704354126099e-08,
+      "loss": 0.3675,
+      "step": 22161
+    },
+    {
+      "epoch": 2.88481062085123,
+      "grad_norm": 2.856931447982788,
+      "learning_rate": 3.795958314307846e-08,
+      "loss": 0.3207,
+      "step": 22164
+    },
+    {
+      "epoch": 2.885201093322921,
+      "grad_norm": 3.8806073665618896,
+      "learning_rate": 3.770298960291796e-08,
+      "loss": 0.3051,
+      "step": 22167
+    },
+    {
+      "epoch": 2.8855915657946114,
+      "grad_norm": 2.8698956966400146,
+      "learning_rate": 3.744726296560641e-08,
+      "loss": 0.3686,
+      "step": 22170
+    },
+    {
+      "epoch": 2.8859820382663024,
+      "grad_norm": 2.8077917098999023,
+      "learning_rate": 3.7192403275818636e-08,
+      "loss": 0.3602,
+      "step": 22173
+    },
+    {
+      "epoch": 2.886372510737993,
+      "grad_norm": 2.8903465270996094,
+      "learning_rate": 3.693841057807734e-08,
+      "loss": 0.3563,
+      "step": 22176
+    },
+    {
+      "epoch": 2.8867629832096835,
+      "grad_norm": 2.8778324127197266,
+      "learning_rate": 3.6685284916755384e-08,
+      "loss": 0.3589,
+      "step": 22179
+    },
+    {
+      "epoch": 2.8871534556813745,
+      "grad_norm": 2.758140802383423,
+      "learning_rate": 3.643302633607237e-08,
+      "loss": 0.3862,
+      "step": 22182
+    },
+    {
+      "epoch": 2.887543928153065,
+      "grad_norm": 2.449650526046753,
+      "learning_rate": 3.618163488009807e-08,
+      "loss": 0.3352,
+      "step": 22185
+    },
+    {
+      "epoch": 2.887934400624756,
+      "grad_norm": 2.983010768890381,
+      "learning_rate": 3.593111059274956e-08,
+      "loss": 0.3191,
+      "step": 22188
+    },
+    {
+      "epoch": 2.8883248730964466,
+      "grad_norm": 2.961902618408203,
+      "learning_rate": 3.5681453517793506e-08,
+      "loss": 0.3256,
+      "step": 22191
+    },
+    {
+      "epoch": 2.8887153455681376,
+      "grad_norm": 2.506606340408325,
+      "learning_rate": 3.54326636988439e-08,
+      "loss": 0.3348,
+      "step": 22194
+    },
+    {
+      "epoch": 2.889105818039828,
+      "grad_norm": 2.9733147621154785,
+      "learning_rate": 3.518474117936432e-08,
+      "loss": 0.3812,
+      "step": 22197
+    },
+    {
+      "epoch": 2.889496290511519,
+      "grad_norm": 2.4794435501098633,
+      "learning_rate": 3.493768600266567e-08,
+      "loss": 0.3011,
+      "step": 22200
+    },
+    {
+      "epoch": 2.8898867629832097,
+      "grad_norm": 2.974116563796997,
+      "learning_rate": 3.469149821190842e-08,
+      "loss": 0.337,
+      "step": 22203
+    },
+    {
+      "epoch": 2.8902772354549002,
+      "grad_norm": 3.0011672973632812,
+      "learning_rate": 3.4446177850100957e-08,
+      "loss": 0.4151,
+      "step": 22206
+    },
+    {
+      "epoch": 2.8906677079265912,
+      "grad_norm": 2.728896141052246,
+      "learning_rate": 3.42017249601001e-08,
+      "loss": 0.3306,
+      "step": 22209
+    },
+    {
+      "epoch": 2.8910581803982818,
+      "grad_norm": 2.9429938793182373,
+      "learning_rate": 3.395813958461169e-08,
+      "loss": 0.3284,
+      "step": 22212
+    },
+    {
+      "epoch": 2.8914486528699728,
+      "grad_norm": 2.7792723178863525,
+      "learning_rate": 3.371542176618891e-08,
+      "loss": 0.3404,
+      "step": 22215
+    },
+    {
+      "epoch": 2.8918391253416633,
+      "grad_norm": 2.7647454738616943,
+      "learning_rate": 3.347357154723452e-08,
+      "loss": 0.3339,
+      "step": 22218
+    },
+    {
+      "epoch": 2.8922295978133543,
+      "grad_norm": 2.6491119861602783,
+      "learning_rate": 3.323258896999915e-08,
+      "loss": 0.3466,
+      "step": 22221
+    },
+    {
+      "epoch": 2.892620070285045,
+      "grad_norm": 2.659365177154541,
+      "learning_rate": 3.2992474076581904e-08,
+      "loss": 0.3819,
+      "step": 22224
+    },
+    {
+      "epoch": 2.893010542756736,
+      "grad_norm": 3.4432740211486816,
+      "learning_rate": 3.27532269089309e-08,
+      "loss": 0.3364,
+      "step": 22227
+    },
+    {
+      "epoch": 2.8934010152284264,
+      "grad_norm": 2.9346230030059814,
+      "learning_rate": 3.251484750884048e-08,
+      "loss": 0.3694,
+      "step": 22230
+    },
+    {
+      "epoch": 2.893791487700117,
+      "grad_norm": 3.1397318840026855,
+      "learning_rate": 3.227733591795734e-08,
+      "loss": 0.3735,
+      "step": 22233
+    },
+    {
+      "epoch": 2.894181960171808,
+      "grad_norm": 2.8210697174072266,
+      "learning_rate": 3.204069217777217e-08,
+      "loss": 0.3545,
+      "step": 22236
+    },
+    {
+      "epoch": 2.8945724326434985,
+      "grad_norm": 2.865011692047119,
+      "learning_rate": 3.1804916329627456e-08,
+      "loss": 0.3738,
+      "step": 22239
+    },
+    {
+      "epoch": 2.8949629051151895,
+      "grad_norm": 2.7513821125030518,
+      "learning_rate": 3.157000841471247e-08,
+      "loss": 0.3312,
+      "step": 22242
+    },
+    {
+      "epoch": 2.89535337758688,
+      "grad_norm": 2.894726037979126,
+      "learning_rate": 3.1335968474064395e-08,
+      "loss": 0.3482,
+      "step": 22245
+    },
+    {
+      "epoch": 2.8957438500585706,
+      "grad_norm": 2.9314725399017334,
+      "learning_rate": 3.110279654857051e-08,
+      "loss": 0.3727,
+      "step": 22248
+    },
+    {
+      "epoch": 2.8961343225302616,
+      "grad_norm": 2.801659107208252,
+      "learning_rate": 3.087049267896492e-08,
+      "loss": 0.3829,
+      "step": 22251
+    },
+    {
+      "epoch": 2.8965247950019526,
+      "grad_norm": 2.8862311840057373,
+      "learning_rate": 3.063905690583069e-08,
+      "loss": 0.3442,
+      "step": 22254
+    },
+    {
+      "epoch": 2.896915267473643,
+      "grad_norm": 2.668565273284912,
+      "learning_rate": 3.040848926959938e-08,
+      "loss": 0.3629,
+      "step": 22257
+    },
+    {
+      "epoch": 2.8973057399453337,
+      "grad_norm": 2.784465789794922,
+      "learning_rate": 3.017878981055045e-08,
+      "loss": 0.3723,
+      "step": 22260
+    },
+    {
+      "epoch": 2.8976962124170247,
+      "grad_norm": 2.8275580406188965,
+      "learning_rate": 2.9949958568811774e-08,
+      "loss": 0.4457,
+      "step": 22263
+    },
+    {
+      "epoch": 2.8980866848887152,
+      "grad_norm": 2.947969913482666,
+      "learning_rate": 2.9721995584360286e-08,
+      "loss": 0.3491,
+      "step": 22266
+    },
+    {
+      "epoch": 2.8984771573604062,
+      "grad_norm": 3.009458303451538,
+      "learning_rate": 2.9494900897019675e-08,
+      "loss": 0.3924,
+      "step": 22269
+    },
+    {
+      "epoch": 2.898867629832097,
+      "grad_norm": 2.875168561935425,
+      "learning_rate": 2.926867454646376e-08,
+      "loss": 0.3703,
+      "step": 22272
+    },
+    {
+      "epoch": 2.8992581023037873,
+      "grad_norm": 2.945216417312622,
+      "learning_rate": 2.904331657221371e-08,
+      "loss": 0.3615,
+      "step": 22275
+    },
+    {
+      "epoch": 2.8996485747754783,
+      "grad_norm": 3.359584331512451,
+      "learning_rate": 2.8818827013638583e-08,
+      "loss": 0.3504,
+      "step": 22278
+    },
+    {
+      "epoch": 2.9000390472471693,
+      "grad_norm": 2.8335955142974854,
+      "learning_rate": 2.859520590995646e-08,
+      "loss": 0.3456,
+      "step": 22281
+    },
+    {
+      "epoch": 2.90042951971886,
+      "grad_norm": 2.685680627822876,
+      "learning_rate": 2.837245330023386e-08,
+      "loss": 0.3402,
+      "step": 22284
+    },
+    {
+      "epoch": 2.9008199921905504,
+      "grad_norm": 2.90457820892334,
+      "learning_rate": 2.815056922338466e-08,
+      "loss": 0.3305,
+      "step": 22287
+    },
+    {
+      "epoch": 2.9012104646622414,
+      "grad_norm": 2.739206314086914,
+      "learning_rate": 2.79295537181723e-08,
+      "loss": 0.338,
+      "step": 22290
+    },
+    {
+      "epoch": 2.901600937133932,
+      "grad_norm": 3.032250165939331,
+      "learning_rate": 2.770940682320644e-08,
+      "loss": 0.3378,
+      "step": 22293
+    },
+    {
+      "epoch": 2.901991409605623,
+      "grad_norm": 2.631166696548462,
+      "learning_rate": 2.749012857694744e-08,
+      "loss": 0.3898,
+      "step": 22296
+    },
+    {
+      "epoch": 2.9023818820773135,
+      "grad_norm": 2.825129747390747,
+      "learning_rate": 2.7271719017702424e-08,
+      "loss": 0.3608,
+      "step": 22299
+    },
+    {
+      "epoch": 2.902772354549004,
+      "grad_norm": 2.757324457168579,
+      "learning_rate": 2.7054178183626988e-08,
+      "loss": 0.4074,
+      "step": 22302
+    },
+    {
+      "epoch": 2.903162827020695,
+      "grad_norm": 2.9708240032196045,
+      "learning_rate": 2.683750611272462e-08,
+      "loss": 0.314,
+      "step": 22305
+    },
+    {
+      "epoch": 2.903553299492386,
+      "grad_norm": 3.2644076347351074,
+      "learning_rate": 2.6621702842848372e-08,
+      "loss": 0.3526,
+      "step": 22308
+    },
+    {
+      "epoch": 2.9039437719640766,
+      "grad_norm": 2.9245588779449463,
+      "learning_rate": 2.6406768411698093e-08,
+      "loss": 0.3377,
+      "step": 22311
+    },
+    {
+      "epoch": 2.904334244435767,
+      "grad_norm": 2.6325671672821045,
+      "learning_rate": 2.6192702856822073e-08,
+      "loss": 0.3118,
+      "step": 22314
+    },
+    {
+      "epoch": 2.904724716907458,
+      "grad_norm": 3.0001933574676514,
+      "learning_rate": 2.597950621561818e-08,
+      "loss": 0.3108,
+      "step": 22317
+    },
+    {
+      "epoch": 2.9051151893791487,
+      "grad_norm": 3.227618932723999,
+      "learning_rate": 2.5767178525330504e-08,
+      "loss": 0.39,
+      "step": 22320
+    },
+    {
+      "epoch": 2.9055056618508397,
+      "grad_norm": 2.91813325881958,
+      "learning_rate": 2.555571982305216e-08,
+      "loss": 0.3227,
+      "step": 22323
+    },
+    {
+      "epoch": 2.9058961343225302,
+      "grad_norm": 2.577854633331299,
+      "learning_rate": 2.534513014572526e-08,
+      "loss": 0.2979,
+      "step": 22326
+    },
+    {
+      "epoch": 2.906286606794221,
+      "grad_norm": 2.738893985748291,
+      "learning_rate": 2.513540953013871e-08,
+      "loss": 0.377,
+      "step": 22329
+    },
+    {
+      "epoch": 2.906677079265912,
+      "grad_norm": 2.668468475341797,
+      "learning_rate": 2.4926558012930425e-08,
+      "loss": 0.3525,
+      "step": 22332
+    },
+    {
+      "epoch": 2.9070675517376023,
+      "grad_norm": 2.666583299636841,
+      "learning_rate": 2.471857563058677e-08,
+      "loss": 0.3823,
+      "step": 22335
+    },
+    {
+      "epoch": 2.9074580242092933,
+      "grad_norm": 3.1053268909454346,
+      "learning_rate": 2.4511462419441466e-08,
+      "loss": 0.3824,
+      "step": 22338
+    },
+    {
+      "epoch": 2.907848496680984,
+      "grad_norm": 2.7575018405914307,
+      "learning_rate": 2.4305218415677235e-08,
+      "loss": 0.3984,
+      "step": 22341
+    },
+    {
+      "epoch": 2.908238969152675,
+      "grad_norm": 3.4204914569854736,
+      "learning_rate": 2.409984365532303e-08,
+      "loss": 0.3681,
+      "step": 22344
+    },
+    {
+      "epoch": 2.9086294416243654,
+      "grad_norm": 2.5891804695129395,
+      "learning_rate": 2.389533817425904e-08,
+      "loss": 0.318,
+      "step": 22347
+    },
+    {
+      "epoch": 2.9090199140960564,
+      "grad_norm": 2.9645187854766846,
+      "learning_rate": 2.3691702008211136e-08,
+      "loss": 0.3515,
+      "step": 22350
+    },
+    {
+      "epoch": 2.909410386567747,
+      "grad_norm": 3.2020936012268066,
+      "learning_rate": 2.348893519275364e-08,
+      "loss": 0.3788,
+      "step": 22353
+    },
+    {
+      "epoch": 2.9098008590394375,
+      "grad_norm": 2.665403127670288,
+      "learning_rate": 2.3287037763310984e-08,
+      "loss": 0.3548,
+      "step": 22356
+    },
+    {
+      "epoch": 2.9101913315111285,
+      "grad_norm": 2.6909213066101074,
+      "learning_rate": 2.3086009755152738e-08,
+      "loss": 0.3125,
+      "step": 22359
+    },
+    {
+      "epoch": 2.910581803982819,
+      "grad_norm": 2.763498067855835,
+      "learning_rate": 2.2885851203399146e-08,
+      "loss": 0.3864,
+      "step": 22362
+    },
+    {
+      "epoch": 2.91097227645451,
+      "grad_norm": 2.798532485961914,
+      "learning_rate": 2.268656214301668e-08,
+      "loss": 0.3635,
+      "step": 22365
+    },
+    {
+      "epoch": 2.9113627489262006,
+      "grad_norm": 2.9937210083007812,
+      "learning_rate": 2.2488142608821373e-08,
+      "loss": 0.4099,
+      "step": 22368
+    },
+    {
+      "epoch": 2.9117532213978916,
+      "grad_norm": 2.972405195236206,
+      "learning_rate": 2.2290592635476615e-08,
+      "loss": 0.4077,
+      "step": 22371
+    },
+    {
+      "epoch": 2.912143693869582,
+      "grad_norm": 2.8520312309265137,
+      "learning_rate": 2.2093912257493133e-08,
+      "loss": 0.3675,
+      "step": 22374
+    },
+    {
+      "epoch": 2.912534166341273,
+      "grad_norm": 2.894411087036133,
+      "learning_rate": 2.1898101509231772e-08,
+      "loss": 0.3879,
+      "step": 22377
+    },
+    {
+      "epoch": 2.9129246388129637,
+      "grad_norm": 2.9919564723968506,
+      "learning_rate": 2.170316042489906e-08,
+      "loss": 0.4072,
+      "step": 22380
+    },
+    {
+      "epoch": 2.9133151112846543,
+      "grad_norm": 2.7866508960723877,
+      "learning_rate": 2.1509089038551645e-08,
+      "loss": 0.3156,
+      "step": 22383
+    },
+    {
+      "epoch": 2.9137055837563453,
+      "grad_norm": 3.07585072517395,
+      "learning_rate": 2.1315887384093513e-08,
+      "loss": 0.3686,
+      "step": 22386
+    },
+    {
+      "epoch": 2.914096056228036,
+      "grad_norm": 3.257378339767456,
+      "learning_rate": 2.1123555495276005e-08,
+      "loss": 0.3546,
+      "step": 22389
+    },
+    {
+      "epoch": 2.914486528699727,
+      "grad_norm": 3.2451460361480713,
+      "learning_rate": 2.093209340569946e-08,
+      "loss": 0.346,
+      "step": 22392
+    },
+    {
+      "epoch": 2.9148770011714173,
+      "grad_norm": 2.622567653656006,
+      "learning_rate": 2.074150114881157e-08,
+      "loss": 0.3506,
+      "step": 22395
+    },
+    {
+      "epoch": 2.915267473643108,
+      "grad_norm": 2.6424460411071777,
+      "learning_rate": 2.055177875790848e-08,
+      "loss": 0.3617,
+      "step": 22398
+    },
+    {
+      "epoch": 2.915657946114799,
+      "grad_norm": 2.814903974533081,
+      "learning_rate": 2.036292626613423e-08,
+      "loss": 0.3418,
+      "step": 22401
+    },
+    {
+      "epoch": 2.91604841858649,
+      "grad_norm": 2.869673728942871,
+      "learning_rate": 2.0174943706481874e-08,
+      "loss": 0.3141,
+      "step": 22404
+    },
+    {
+      "epoch": 2.9164388910581804,
+      "grad_norm": 2.6564950942993164,
+      "learning_rate": 1.9987831111790147e-08,
+      "loss": 0.3139,
+      "step": 22407
+    },
+    {
+      "epoch": 2.916829363529871,
+      "grad_norm": 2.9242870807647705,
+      "learning_rate": 1.980158851474845e-08,
+      "loss": 0.4097,
+      "step": 22410
+    },
+    {
+      "epoch": 2.917219836001562,
+      "grad_norm": 2.8616063594818115,
+      "learning_rate": 1.9616215947892427e-08,
+      "loss": 0.3452,
+      "step": 22413
+    },
+    {
+      "epoch": 2.9176103084732525,
+      "grad_norm": 2.781572103500366,
+      "learning_rate": 1.9431713443605616e-08,
+      "loss": 0.384,
+      "step": 22416
+    },
+    {
+      "epoch": 2.9180007809449435,
+      "grad_norm": 3.0392205715179443,
+      "learning_rate": 1.924808103412168e-08,
+      "loss": 0.4144,
+      "step": 22419
+    },
+    {
+      "epoch": 2.918391253416634,
+      "grad_norm": 2.8837602138519287,
+      "learning_rate": 1.9065318751519402e-08,
+      "loss": 0.3395,
+      "step": 22422
+    },
+    {
+      "epoch": 2.9187817258883246,
+      "grad_norm": 2.5549280643463135,
+      "learning_rate": 1.8883426627727684e-08,
+      "loss": 0.3312,
+      "step": 22425
+    },
+    {
+      "epoch": 2.9191721983600156,
+      "grad_norm": 3.014940023422241,
+      "learning_rate": 1.870240469452278e-08,
+      "loss": 0.3757,
+      "step": 22428
+    },
+    {
+      "epoch": 2.9195626708317066,
+      "grad_norm": 2.8069894313812256,
+      "learning_rate": 1.8522252983528832e-08,
+      "loss": 0.3599,
+      "step": 22431
+    },
+    {
+      "epoch": 2.919953143303397,
+      "grad_norm": 4.53384256362915,
+      "learning_rate": 1.8342971526217334e-08,
+      "loss": 0.3301,
+      "step": 22434
+    },
+    {
+      "epoch": 2.9203436157750877,
+      "grad_norm": 2.5881564617156982,
+      "learning_rate": 1.8164560353909344e-08,
+      "loss": 0.3605,
+      "step": 22437
+    },
+    {
+      "epoch": 2.9207340882467787,
+      "grad_norm": 2.660621166229248,
+      "learning_rate": 1.798701949777215e-08,
+      "loss": 0.3414,
+      "step": 22440
+    },
+    {
+      "epoch": 2.9211245607184693,
+      "grad_norm": 2.7579736709594727,
+      "learning_rate": 1.7810348988822058e-08,
+      "loss": 0.3739,
+      "step": 22443
+    },
+    {
+      "epoch": 2.9215150331901603,
+      "grad_norm": 3.0707061290740967,
+      "learning_rate": 1.7634548857922707e-08,
+      "loss": 0.3433,
+      "step": 22446
+    },
+    {
+      "epoch": 2.921905505661851,
+      "grad_norm": 2.9784891605377197,
+      "learning_rate": 1.7459619135786753e-08,
+      "loss": 0.3297,
+      "step": 22449
+    },
+    {
+      "epoch": 2.9222959781335414,
+      "grad_norm": 2.9023778438568115,
+      "learning_rate": 1.728555985297309e-08,
+      "loss": 0.3122,
+      "step": 22452
+    },
+    {
+      "epoch": 2.9226864506052324,
+      "grad_norm": 2.9779181480407715,
+      "learning_rate": 1.711237103989072e-08,
+      "loss": 0.3818,
+      "step": 22455
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 2.76466703414917,
+      "learning_rate": 1.6940052726793776e-08,
+      "loss": 0.3163,
+      "step": 22458
+    },
+    {
+      "epoch": 2.923467395548614,
+      "grad_norm": 2.9506185054779053,
+      "learning_rate": 1.6768604943787624e-08,
+      "loss": 0.2745,
+      "step": 22461
+    },
+    {
+      "epoch": 2.9238578680203045,
+      "grad_norm": 2.865941047668457,
+      "learning_rate": 1.659802772082275e-08,
+      "loss": 0.3797,
+      "step": 22464
+    },
+    {
+      "epoch": 2.9242483404919954,
+      "grad_norm": 2.825014591217041,
+      "learning_rate": 1.642832108769865e-08,
+      "loss": 0.4346,
+      "step": 22467
+    },
+    {
+      "epoch": 2.924638812963686,
+      "grad_norm": 2.9251394271850586,
+      "learning_rate": 1.6259485074063276e-08,
+      "loss": 0.3382,
+      "step": 22470
+    },
+    {
+      "epoch": 2.925029285435377,
+      "grad_norm": 2.5438222885131836,
+      "learning_rate": 1.6091519709411363e-08,
+      "loss": 0.3026,
+      "step": 22473
+    },
+    {
+      "epoch": 2.9254197579070675,
+      "grad_norm": 2.7614340782165527,
+      "learning_rate": 1.5924425023086665e-08,
+      "loss": 0.3606,
+      "step": 22476
+    },
+    {
+      "epoch": 2.925810230378758,
+      "grad_norm": 2.884199619293213,
+      "learning_rate": 1.575820104427972e-08,
+      "loss": 0.3452,
+      "step": 22479
+    },
+    {
+      "epoch": 2.926200702850449,
+      "grad_norm": 2.60504150390625,
+      "learning_rate": 1.559284780202952e-08,
+      "loss": 0.3219,
+      "step": 22482
+    },
+    {
+      "epoch": 2.9265911753221396,
+      "grad_norm": 3.0407423973083496,
+      "learning_rate": 1.5428365325223516e-08,
+      "loss": 0.3445,
+      "step": 22485
+    },
+    {
+      "epoch": 2.9269816477938306,
+      "grad_norm": 2.683939218521118,
+      "learning_rate": 1.5264753642595387e-08,
+      "loss": 0.3529,
+      "step": 22488
+    },
+    {
+      "epoch": 2.927372120265521,
+      "grad_norm": 2.9559898376464844,
+      "learning_rate": 1.510201278272949e-08,
+      "loss": 0.3729,
+      "step": 22491
+    },
+    {
+      "epoch": 2.927762592737212,
+      "grad_norm": 2.676271438598633,
+      "learning_rate": 1.4940142774054755e-08,
+      "loss": 0.4197,
+      "step": 22494
+    },
+    {
+      "epoch": 2.9281530652089027,
+      "grad_norm": 3.0455739498138428,
+      "learning_rate": 1.4779143644850225e-08,
+      "loss": 0.4026,
+      "step": 22497
+    },
+    {
+      "epoch": 2.9285435376805937,
+      "grad_norm": 3.0256705284118652,
+      "learning_rate": 1.4619015423241178e-08,
+      "loss": 0.4349,
+      "step": 22500
+    },
+    {
+      "epoch": 2.9289340101522843,
+      "grad_norm": 2.6953423023223877,
+      "learning_rate": 1.4459758137203572e-08,
+      "loss": 0.324,
+      "step": 22503
+    },
+    {
+      "epoch": 2.929324482623975,
+      "grad_norm": 3.021960735321045,
+      "learning_rate": 1.4301371814557374e-08,
+      "loss": 0.3427,
+      "step": 22506
+    },
+    {
+      "epoch": 2.929714955095666,
+      "grad_norm": 2.9862983226776123,
+      "learning_rate": 1.4143856482973228e-08,
+      "loss": 0.3093,
+      "step": 22509
+    },
+    {
+      "epoch": 2.9301054275673564,
+      "grad_norm": 2.836103916168213,
+      "learning_rate": 1.3987212169969121e-08,
+      "loss": 0.3272,
+      "step": 22512
+    },
+    {
+      "epoch": 2.9304959000390474,
+      "grad_norm": 2.806823492050171,
+      "learning_rate": 1.3831438902909834e-08,
+      "loss": 0.3633,
+      "step": 22515
+    },
+    {
+      "epoch": 2.930886372510738,
+      "grad_norm": 3.1762378215789795,
+      "learning_rate": 1.3676536709008593e-08,
+      "loss": 0.3304,
+      "step": 22518
+    },
+    {
+      "epoch": 2.9312768449824285,
+      "grad_norm": 2.651766300201416,
+      "learning_rate": 1.3522505615326531e-08,
+      "loss": 0.3698,
+      "step": 22521
+    },
+    {
+      "epoch": 2.9316673174541195,
+      "grad_norm": 2.5449414253234863,
+      "learning_rate": 1.336934564877268e-08,
+      "loss": 0.4071,
+      "step": 22524
+    },
+    {
+      "epoch": 2.9320577899258105,
+      "grad_norm": 2.726243495941162,
+      "learning_rate": 1.3217056836104525e-08,
+      "loss": 0.4334,
+      "step": 22527
+    },
+    {
+      "epoch": 2.932448262397501,
+      "grad_norm": 2.689444065093994,
+      "learning_rate": 1.3065639203925229e-08,
+      "loss": 0.3296,
+      "step": 22530
+    },
+    {
+      "epoch": 2.9328387348691916,
+      "grad_norm": 2.9110662937164307,
+      "learning_rate": 1.2915092778688077e-08,
+      "loss": 0.381,
+      "step": 22533
+    },
+    {
+      "epoch": 2.9332292073408825,
+      "grad_norm": 3.017112970352173,
+      "learning_rate": 1.2765417586692586e-08,
+      "loss": 0.366,
+      "step": 22536
+    },
+    {
+      "epoch": 2.933619679812573,
+      "grad_norm": 2.5785021781921387,
+      "learning_rate": 1.2616613654087285e-08,
+      "loss": 0.3984,
+      "step": 22539
+    },
+    {
+      "epoch": 2.934010152284264,
+      "grad_norm": 2.9283220767974854,
+      "learning_rate": 1.2468681006868044e-08,
+      "loss": 0.3302,
+      "step": 22542
+    },
+    {
+      "epoch": 2.9344006247559546,
+      "grad_norm": 2.7230112552642822,
+      "learning_rate": 1.2321619670877527e-08,
+      "loss": 0.3774,
+      "step": 22545
+    },
+    {
+      "epoch": 2.934791097227645,
+      "grad_norm": 2.7927682399749756,
+      "learning_rate": 1.2175429671807405e-08,
+      "loss": 0.421,
+      "step": 22548
+    },
+    {
+      "epoch": 2.935181569699336,
+      "grad_norm": 2.6731719970703125,
+      "learning_rate": 1.2030111035197245e-08,
+      "loss": 0.3555,
+      "step": 22551
+    },
+    {
+      "epoch": 2.935572042171027,
+      "grad_norm": 2.5603408813476562,
+      "learning_rate": 1.1885663786433411e-08,
+      "loss": 0.3375,
+      "step": 22554
+    },
+    {
+      "epoch": 2.9359625146427177,
+      "grad_norm": 2.7101588249206543,
+      "learning_rate": 1.1742087950750714e-08,
+      "loss": 0.3384,
+      "step": 22557
+    },
+    {
+      "epoch": 2.9363529871144083,
+      "grad_norm": 2.6867103576660156,
+      "learning_rate": 1.1599383553231314e-08,
+      "loss": 0.3236,
+      "step": 22560
+    },
+    {
+      "epoch": 2.9367434595860993,
+      "grad_norm": 3.0580341815948486,
+      "learning_rate": 1.1457550618805824e-08,
+      "loss": 0.3854,
+      "step": 22563
+    },
+    {
+      "epoch": 2.93713393205779,
+      "grad_norm": 2.929643154144287,
+      "learning_rate": 1.1316589172251091e-08,
+      "loss": 0.344,
+      "step": 22566
+    },
+    {
+      "epoch": 2.937524404529481,
+      "grad_norm": 2.806877613067627,
+      "learning_rate": 1.1176499238194639e-08,
+      "loss": 0.3385,
+      "step": 22569
+    },
+    {
+      "epoch": 2.9379148770011714,
+      "grad_norm": 2.79831600189209,
+      "learning_rate": 1.1037280841108e-08,
+      "loss": 0.3901,
+      "step": 22572
+    },
+    {
+      "epoch": 2.938305349472862,
+      "grad_norm": 2.647631883621216,
+      "learning_rate": 1.0898934005313389e-08,
+      "loss": 0.3789,
+      "step": 22575
+    },
+    {
+      "epoch": 2.938695821944553,
+      "grad_norm": 2.671846866607666,
+      "learning_rate": 1.0761458754979804e-08,
+      "loss": 0.2923,
+      "step": 22578
+    },
+    {
+      "epoch": 2.939086294416244,
+      "grad_norm": 2.559683084487915,
+      "learning_rate": 1.0624855114123035e-08,
+      "loss": 0.3096,
+      "step": 22581
+    },
+    {
+      "epoch": 2.9394767668879345,
+      "grad_norm": 2.7834372520446777,
+      "learning_rate": 1.0489123106608434e-08,
+      "loss": 0.3449,
+      "step": 22584
+    },
+    {
+      "epoch": 2.939867239359625,
+      "grad_norm": 2.7174694538116455,
+      "learning_rate": 1.0354262756147593e-08,
+      "loss": 0.3508,
+      "step": 22587
+    },
+    {
+      "epoch": 2.940257711831316,
+      "grad_norm": 2.7829127311706543,
+      "learning_rate": 1.0220274086299998e-08,
+      "loss": 0.397,
+      "step": 22590
+    },
+    {
+      "epoch": 2.9406481843030066,
+      "grad_norm": 2.9258644580841064,
+      "learning_rate": 1.0087157120474145e-08,
+      "loss": 0.3757,
+      "step": 22593
+    },
+    {
+      "epoch": 2.9410386567746976,
+      "grad_norm": 2.8890163898468018,
+      "learning_rate": 9.954911881924212e-09,
+      "loss": 0.3301,
+      "step": 22596
+    },
+    {
+      "epoch": 2.941429129246388,
+      "grad_norm": 2.4715800285339355,
+      "learning_rate": 9.82353839375394e-09,
+      "loss": 0.3031,
+      "step": 22599
+    },
+    {
+      "epoch": 2.9418196017180787,
+      "grad_norm": 2.8599045276641846,
+      "learning_rate": 9.693036678913303e-09,
+      "loss": 0.3638,
+      "step": 22602
+    },
+    {
+      "epoch": 2.9422100741897697,
+      "grad_norm": 2.724221706390381,
+      "learning_rate": 9.563406760201288e-09,
+      "loss": 0.4034,
+      "step": 22605
+    },
+    {
+      "epoch": 2.94260054666146,
+      "grad_norm": 2.7690560817718506,
+      "learning_rate": 9.434648660263668e-09,
+      "loss": 0.3274,
+      "step": 22608
+    },
+    {
+      "epoch": 2.942991019133151,
+      "grad_norm": 2.534320592880249,
+      "learning_rate": 9.30676240159467e-09,
+      "loss": 0.3847,
+      "step": 22611
+    },
+    {
+      "epoch": 2.9433814916048417,
+      "grad_norm": 2.9107460975646973,
+      "learning_rate": 9.179748006535317e-09,
+      "loss": 0.3455,
+      "step": 22614
+    },
+    {
+      "epoch": 2.9437719640765327,
+      "grad_norm": 2.507080554962158,
+      "learning_rate": 9.05360549727452e-09,
+      "loss": 0.3132,
+      "step": 22617
+    },
+    {
+      "epoch": 2.9441624365482233,
+      "grad_norm": 2.537360191345215,
+      "learning_rate": 8.928334895849656e-09,
+      "loss": 0.3198,
+      "step": 22620
+    },
+    {
+      "epoch": 2.9445529090199143,
+      "grad_norm": 2.297013282775879,
+      "learning_rate": 8.803936224144883e-09,
+      "loss": 0.2882,
+      "step": 22623
+    },
+    {
+      "epoch": 2.944943381491605,
+      "grad_norm": 2.7976059913635254,
+      "learning_rate": 8.680409503892817e-09,
+      "loss": 0.4199,
+      "step": 22626
+    },
+    {
+      "epoch": 2.9453338539632954,
+      "grad_norm": 2.64867901802063,
+      "learning_rate": 8.557754756672864e-09,
+      "loss": 0.3469,
+      "step": 22629
+    },
+    {
+      "epoch": 2.9457243264349864,
+      "grad_norm": 2.9491302967071533,
+      "learning_rate": 8.435972003912329e-09,
+      "loss": 0.3483,
+      "step": 22632
+    },
+    {
+      "epoch": 2.946114798906677,
+      "grad_norm": 2.6923861503601074,
+      "learning_rate": 8.31506126688697e-09,
+      "loss": 0.3447,
+      "step": 22635
+    },
+    {
+      "epoch": 2.946505271378368,
+      "grad_norm": 3.363018035888672,
+      "learning_rate": 8.19502256671989e-09,
+      "loss": 0.3398,
+      "step": 22638
+    },
+    {
+      "epoch": 2.9468957438500585,
+      "grad_norm": 2.8834447860717773,
+      "learning_rate": 8.075855924380427e-09,
+      "loss": 0.4022,
+      "step": 22641
+    },
+    {
+      "epoch": 2.9472862163217495,
+      "grad_norm": 2.857146978378296,
+      "learning_rate": 7.957561360688038e-09,
+      "loss": 0.3495,
+      "step": 22644
+    },
+    {
+      "epoch": 2.94767668879344,
+      "grad_norm": 2.984123468399048,
+      "learning_rate": 7.840138896307303e-09,
+      "loss": 0.308,
+      "step": 22647
+    },
+    {
+      "epoch": 2.948067161265131,
+      "grad_norm": 3.035522222518921,
+      "learning_rate": 7.723588551752925e-09,
+      "loss": 0.3807,
+      "step": 22650
+    },
+    {
+      "epoch": 2.9484576337368216,
+      "grad_norm": 2.7213213443756104,
+      "learning_rate": 7.607910347385283e-09,
+      "loss": 0.3534,
+      "step": 22653
+    },
+    {
+      "epoch": 2.948848106208512,
+      "grad_norm": 2.8660976886749268,
+      "learning_rate": 7.493104303413212e-09,
+      "loss": 0.3385,
+      "step": 22656
+    },
+    {
+      "epoch": 2.949238578680203,
+      "grad_norm": 2.926711320877075,
+      "learning_rate": 7.379170439892891e-09,
+      "loss": 0.3535,
+      "step": 22659
+    },
+    {
+      "epoch": 2.9496290511518937,
+      "grad_norm": 2.7307538986206055,
+      "learning_rate": 7.266108776728953e-09,
+      "loss": 0.3584,
+      "step": 22662
+    },
+    {
+      "epoch": 2.9500195236235847,
+      "grad_norm": 2.8166987895965576,
+      "learning_rate": 7.153919333672266e-09,
+      "loss": 0.322,
+      "step": 22665
+    },
+    {
+      "epoch": 2.950409996095275,
+      "grad_norm": 2.866868734359741,
+      "learning_rate": 7.042602130322707e-09,
+      "loss": 0.3712,
+      "step": 22668
+    },
+    {
+      "epoch": 2.9508004685669658,
+      "grad_norm": 2.841460704803467,
+      "learning_rate": 6.932157186126942e-09,
+      "loss": 0.4131,
+      "step": 22671
+    },
+    {
+      "epoch": 2.9511909410386568,
+      "grad_norm": 2.5666918754577637,
+      "learning_rate": 6.822584520379538e-09,
+      "loss": 0.3417,
+      "step": 22674
+    },
+    {
+      "epoch": 2.9515814135103478,
+      "grad_norm": 3.38871431350708,
+      "learning_rate": 6.713884152222405e-09,
+      "loss": 0.3104,
+      "step": 22677
+    },
+    {
+      "epoch": 2.9519718859820383,
+      "grad_norm": 3.0104832649230957,
+      "learning_rate": 6.6060561006453525e-09,
+      "loss": 0.4312,
+      "step": 22680
+    },
+    {
+      "epoch": 2.952362358453729,
+      "grad_norm": 3.21549129486084,
+      "learning_rate": 6.499100384485535e-09,
+      "loss": 0.3817,
+      "step": 22683
+    },
+    {
+      "epoch": 2.95275283092542,
+      "grad_norm": 3.07248592376709,
+      "learning_rate": 6.393017022428005e-09,
+      "loss": 0.3889,
+      "step": 22686
+    },
+    {
+      "epoch": 2.9531433033971104,
+      "grad_norm": 2.6400251388549805,
+      "learning_rate": 6.287806033005717e-09,
+      "loss": 0.3869,
+      "step": 22689
+    },
+    {
+      "epoch": 2.9535337758688014,
+      "grad_norm": 2.670640707015991,
+      "learning_rate": 6.1834674345984115e-09,
+      "loss": 0.2964,
+      "step": 22692
+    },
+    {
+      "epoch": 2.953924248340492,
+      "grad_norm": 2.874105215072632,
+      "learning_rate": 6.080001245433731e-09,
+      "loss": 0.3724,
+      "step": 22695
+    },
+    {
+      "epoch": 2.9543147208121825,
+      "grad_norm": 2.8548583984375,
+      "learning_rate": 5.977407483587217e-09,
+      "loss": 0.3864,
+      "step": 22698
+    },
+    {
+      "epoch": 2.9547051932838735,
+      "grad_norm": 2.812763214111328,
+      "learning_rate": 5.875686166981753e-09,
+      "loss": 0.3374,
+      "step": 22701
+    },
+    {
+      "epoch": 2.9550956657555645,
+      "grad_norm": 2.711881637573242,
+      "learning_rate": 5.7748373133875715e-09,
+      "loss": 0.3739,
+      "step": 22704
+    },
+    {
+      "epoch": 2.955486138227255,
+      "grad_norm": 2.6988415718078613,
+      "learning_rate": 5.674860940423354e-09,
+      "loss": 0.334,
+      "step": 22707
+    },
+    {
+      "epoch": 2.9558766106989456,
+      "grad_norm": 2.473297357559204,
+      "learning_rate": 5.5757570655545765e-09,
+      "loss": 0.3704,
+      "step": 22710
+    },
+    {
+      "epoch": 2.9562670831706366,
+      "grad_norm": 2.6674001216888428,
+      "learning_rate": 5.477525706094056e-09,
+      "loss": 0.364,
+      "step": 22713
+    },
+    {
+      "epoch": 2.956657555642327,
+      "grad_norm": 2.651705741882324,
+      "learning_rate": 5.380166879202508e-09,
+      "loss": 0.3361,
+      "step": 22716
+    },
+    {
+      "epoch": 2.957048028114018,
+      "grad_norm": 2.93682599067688,
+      "learning_rate": 5.283680601889107e-09,
+      "loss": 0.4136,
+      "step": 22719
+    },
+    {
+      "epoch": 2.9574385005857087,
+      "grad_norm": 2.921872854232788,
+      "learning_rate": 5.188066891009258e-09,
+      "loss": 0.3346,
+      "step": 22722
+    },
+    {
+      "epoch": 2.9578289730573992,
+      "grad_norm": 2.8283815383911133,
+      "learning_rate": 5.093325763266821e-09,
+      "loss": 0.2995,
+      "step": 22725
+    },
+    {
+      "epoch": 2.95821944552909,
+      "grad_norm": 2.4506702423095703,
+      "learning_rate": 4.999457235212446e-09,
+      "loss": 0.3262,
+      "step": 22728
+    },
+    {
+      "epoch": 2.958609918000781,
+      "grad_norm": 2.559972047805786,
+      "learning_rate": 4.906461323244683e-09,
+      "loss": 0.3491,
+      "step": 22731
+    },
+    {
+      "epoch": 2.9590003904724718,
+      "grad_norm": 2.777678966522217,
+      "learning_rate": 4.814338043609979e-09,
+      "loss": 0.3715,
+      "step": 22734
+    },
+    {
+      "epoch": 2.9593908629441623,
+      "grad_norm": 2.797595977783203,
+      "learning_rate": 4.7230874124026825e-09,
+      "loss": 0.3736,
+      "step": 22737
+    },
+    {
+      "epoch": 2.9597813354158533,
+      "grad_norm": 2.576273202896118,
+      "learning_rate": 4.632709445562822e-09,
+      "loss": 0.3177,
+      "step": 22740
+    },
+    {
+      "epoch": 2.960171807887544,
+      "grad_norm": 2.8237977027893066,
+      "learning_rate": 4.543204158879988e-09,
+      "loss": 0.3287,
+      "step": 22743
+    },
+    {
+      "epoch": 2.960562280359235,
+      "grad_norm": 2.5393621921539307,
+      "learning_rate": 4.454571567991117e-09,
+      "loss": 0.3608,
+      "step": 22746
+    },
+    {
+      "epoch": 2.9609527528309254,
+      "grad_norm": 2.9079854488372803,
+      "learning_rate": 4.3668116883788245e-09,
+      "loss": 0.3497,
+      "step": 22749
+    },
+    {
+      "epoch": 2.961343225302616,
+      "grad_norm": 2.9478375911712646,
+      "learning_rate": 4.2799245353752905e-09,
+      "loss": 0.3858,
+      "step": 22752
+    },
+    {
+      "epoch": 2.961733697774307,
+      "grad_norm": 2.845107316970825,
+      "learning_rate": 4.193910124160039e-09,
+      "loss": 0.3556,
+      "step": 22755
+    },
+    {
+      "epoch": 2.9621241702459975,
+      "grad_norm": 2.638409376144409,
+      "learning_rate": 4.108768469758273e-09,
+      "loss": 0.2997,
+      "step": 22758
+    },
+    {
+      "epoch": 2.9625146427176885,
+      "grad_norm": 3.1192007064819336,
+      "learning_rate": 4.0244995870453166e-09,
+      "loss": 0.3626,
+      "step": 22761
+    },
+    {
+      "epoch": 2.962905115189379,
+      "grad_norm": 2.773621082305908,
+      "learning_rate": 3.941103490742171e-09,
+      "loss": 0.2836,
+      "step": 22764
+    },
+    {
+      "epoch": 2.96329558766107,
+      "grad_norm": 2.718259811401367,
+      "learning_rate": 3.858580195418293e-09,
+      "loss": 0.3435,
+      "step": 22767
+    },
+    {
+      "epoch": 2.9636860601327606,
+      "grad_norm": 2.8922083377838135,
+      "learning_rate": 3.77692971548993e-09,
+      "loss": 0.3966,
+      "step": 22770
+    },
+    {
+      "epoch": 2.9640765326044516,
+      "grad_norm": 2.6764767169952393,
+      "learning_rate": 3.6961520652212256e-09,
+      "loss": 0.3352,
+      "step": 22773
+    },
+    {
+      "epoch": 2.964467005076142,
+      "grad_norm": 2.891240119934082,
+      "learning_rate": 3.6162472587242257e-09,
+      "loss": 0.3341,
+      "step": 22776
+    },
+    {
+      "epoch": 2.9648574775478327,
+      "grad_norm": 2.7254581451416016,
+      "learning_rate": 3.5372153099583195e-09,
+      "loss": 0.3189,
+      "step": 22779
+    },
+    {
+      "epoch": 2.9652479500195237,
+      "grad_norm": 3.07497501373291,
+      "learning_rate": 3.459056232729685e-09,
+      "loss": 0.3985,
+      "step": 22782
+    },
+    {
+      "epoch": 2.9656384224912142,
+      "grad_norm": 2.8364953994750977,
+      "learning_rate": 3.3817700406924004e-09,
+      "loss": 0.3572,
+      "step": 22785
+    },
+    {
+      "epoch": 2.9660288949629052,
+      "grad_norm": 2.677969217300415,
+      "learning_rate": 3.305356747348998e-09,
+      "loss": 0.3667,
+      "step": 22788
+    },
+    {
+      "epoch": 2.9664193674345958,
+      "grad_norm": 2.8867709636688232,
+      "learning_rate": 3.229816366047689e-09,
+      "loss": 0.3836,
+      "step": 22791
+    },
+    {
+      "epoch": 2.9668098399062868,
+      "grad_norm": 2.7615604400634766,
+      "learning_rate": 3.15514890998625e-09,
+      "loss": 0.3543,
+      "step": 22794
+    },
+    {
+      "epoch": 2.9672003123779773,
+      "grad_norm": 2.8149735927581787,
+      "learning_rate": 3.0813543922081357e-09,
+      "loss": 0.3806,
+      "step": 22797
+    },
+    {
+      "epoch": 2.9675907848496683,
+      "grad_norm": 2.686567544937134,
+      "learning_rate": 3.0084328256058116e-09,
+      "loss": 0.3664,
+      "step": 22800
+    },
+    {
+      "epoch": 2.967981257321359,
+      "grad_norm": 2.6036901473999023,
+      "learning_rate": 2.9363842229179763e-09,
+      "loss": 0.3127,
+      "step": 22803
+    },
+    {
+      "epoch": 2.9683717297930494,
+      "grad_norm": 2.843297004699707,
+      "learning_rate": 2.865208596731783e-09,
+      "loss": 0.3899,
+      "step": 22806
+    },
+    {
+      "epoch": 2.9687622022647404,
+      "grad_norm": 2.4631242752075195,
+      "learning_rate": 2.7949059594806203e-09,
+      "loss": 0.3895,
+      "step": 22809
+    },
+    {
+      "epoch": 2.969152674736431,
+      "grad_norm": 2.619086980819702,
+      "learning_rate": 2.7254763234474404e-09,
+      "loss": 0.277,
+      "step": 22812
+    },
+    {
+      "epoch": 2.969543147208122,
+      "grad_norm": 2.584728717803955,
+      "learning_rate": 2.6569197007603186e-09,
+      "loss": 0.3257,
+      "step": 22815
+    },
+    {
+      "epoch": 2.9699336196798125,
+      "grad_norm": 3.1949074268341064,
+      "learning_rate": 2.589236103396897e-09,
+      "loss": 0.408,
+      "step": 22818
+    },
+    {
+      "epoch": 2.970324092151503,
+      "grad_norm": 2.800149440765381,
+      "learning_rate": 2.5224255431804957e-09,
+      "loss": 0.3305,
+      "step": 22821
+    },
+    {
+      "epoch": 2.970714564623194,
+      "grad_norm": 2.811697006225586,
+      "learning_rate": 2.4564880317834438e-09,
+      "loss": 0.3886,
+      "step": 22824
+    },
+    {
+      "epoch": 2.971105037094885,
+      "grad_norm": 2.7836661338806152,
+      "learning_rate": 2.391423580724861e-09,
+      "loss": 0.3741,
+      "step": 22827
+    },
+    {
+      "epoch": 2.9714955095665756,
+      "grad_norm": 2.9304723739624023,
+      "learning_rate": 2.327232201370655e-09,
+      "loss": 0.3688,
+      "step": 22830
+    },
+    {
+      "epoch": 2.971885982038266,
+      "grad_norm": 2.7736597061157227,
+      "learning_rate": 2.263913904935744e-09,
+      "loss": 0.3645,
+      "step": 22833
+    },
+    {
+      "epoch": 2.972276454509957,
+      "grad_norm": 2.527412176132202,
+      "learning_rate": 2.20146870248128e-09,
+      "loss": 0.3668,
+      "step": 22836
+    },
+    {
+      "epoch": 2.9726669269816477,
+      "grad_norm": 2.9895877838134766,
+      "learning_rate": 2.139896604916314e-09,
+      "loss": 0.4066,
+      "step": 22839
+    },
+    {
+      "epoch": 2.9730573994533387,
+      "grad_norm": 2.8589468002319336,
+      "learning_rate": 2.079197622997242e-09,
+      "loss": 0.3825,
+      "step": 22842
+    },
+    {
+      "epoch": 2.9734478719250292,
+      "grad_norm": 2.533005475997925,
+      "learning_rate": 2.0193717673283597e-09,
+      "loss": 0.3998,
+      "step": 22845
+    },
+    {
+      "epoch": 2.97383834439672,
+      "grad_norm": 2.7130699157714844,
+      "learning_rate": 1.9604190483613062e-09,
+      "loss": 0.3201,
+      "step": 22848
+    },
+    {
+      "epoch": 2.974228816868411,
+      "grad_norm": 2.8447370529174805,
+      "learning_rate": 1.9023394763945104e-09,
+      "loss": 0.3743,
+      "step": 22851
+    },
+    {
+      "epoch": 2.974619289340102,
+      "grad_norm": 3.0323686599731445,
+      "learning_rate": 1.8451330615748553e-09,
+      "loss": 0.3989,
+      "step": 22854
+    },
+    {
+      "epoch": 2.9750097618117923,
+      "grad_norm": 3.1013906002044678,
+      "learning_rate": 1.7887998138954587e-09,
+      "loss": 0.3767,
+      "step": 22857
+    },
+    {
+      "epoch": 2.975400234283483,
+      "grad_norm": 2.792752504348755,
+      "learning_rate": 1.7333397431984477e-09,
+      "loss": 0.3696,
+      "step": 22860
+    },
+    {
+      "epoch": 2.975790706755174,
+      "grad_norm": 2.7394256591796875,
+      "learning_rate": 1.6787528591716284e-09,
+      "loss": 0.3893,
+      "step": 22863
+    },
+    {
+      "epoch": 2.9761811792268644,
+      "grad_norm": 3.603832244873047,
+      "learning_rate": 1.6250391713523717e-09,
+      "loss": 0.3438,
+      "step": 22866
+    },
+    {
+      "epoch": 2.9765716516985554,
+      "grad_norm": 2.743468761444092,
+      "learning_rate": 1.5721986891231722e-09,
+      "loss": 0.3395,
+      "step": 22869
+    },
+    {
+      "epoch": 2.976962124170246,
+      "grad_norm": 2.7976534366607666,
+      "learning_rate": 1.5202314217160895e-09,
+      "loss": 0.3345,
+      "step": 22872
+    },
+    {
+      "epoch": 2.9773525966419365,
+      "grad_norm": 2.8233566284179688,
+      "learning_rate": 1.4691373782088625e-09,
+      "loss": 0.4015,
+      "step": 22875
+    },
+    {
+      "epoch": 2.9777430691136275,
+      "grad_norm": 2.91632080078125,
+      "learning_rate": 1.418916567528239e-09,
+      "loss": 0.3493,
+      "step": 22878
+    },
+    {
+      "epoch": 2.9781335415853185,
+      "grad_norm": 2.812035322189331,
+      "learning_rate": 1.3695689984477566e-09,
+      "loss": 0.3718,
+      "step": 22881
+    },
+    {
+      "epoch": 2.978524014057009,
+      "grad_norm": 2.8257293701171875,
+      "learning_rate": 1.3210946795877423e-09,
+      "loss": 0.3545,
+      "step": 22884
+    },
+    {
+      "epoch": 2.9789144865286996,
+      "grad_norm": 2.968236207962036,
+      "learning_rate": 1.2734936194164215e-09,
+      "loss": 0.3936,
+      "step": 22887
+    },
+    {
+      "epoch": 2.9793049590003906,
+      "grad_norm": 2.8840749263763428,
+      "learning_rate": 1.2267658262504756e-09,
+      "loss": 0.4015,
+      "step": 22890
+    },
+    {
+      "epoch": 2.979695431472081,
+      "grad_norm": 2.828843355178833,
+      "learning_rate": 1.1809113082522638e-09,
+      "loss": 0.2979,
+      "step": 22893
+    },
+    {
+      "epoch": 2.980085903943772,
+      "grad_norm": 2.849477767944336,
+      "learning_rate": 1.1359300734331558e-09,
+      "loss": 0.3837,
+      "step": 22896
+    },
+    {
+      "epoch": 2.9804763764154627,
+      "grad_norm": 2.721007823944092,
+      "learning_rate": 1.0918221296507547e-09,
+      "loss": 0.4118,
+      "step": 22899
+    },
+    {
+      "epoch": 2.9808668488871533,
+      "grad_norm": 2.5880935192108154,
+      "learning_rate": 1.0485874846111189e-09,
+      "loss": 0.3802,
+      "step": 22902
+    },
+    {
+      "epoch": 2.9812573213588442,
+      "grad_norm": 2.763317108154297,
+      "learning_rate": 1.0062261458670952e-09,
+      "loss": 0.3266,
+      "step": 22905
+    },
+    {
+      "epoch": 2.981647793830535,
+      "grad_norm": 2.6199758052825928,
+      "learning_rate": 9.647381208188755e-10,
+      "loss": 0.3894,
+      "step": 22908
+    },
+    {
+      "epoch": 2.982038266302226,
+      "grad_norm": 3.725672960281372,
+      "learning_rate": 9.241234167139956e-10,
+      "loss": 0.4053,
+      "step": 22911
+    },
+    {
+      "epoch": 2.9824287387739163,
+      "grad_norm": 2.939629077911377,
+      "learning_rate": 8.843820406490011e-10,
+      "loss": 0.3776,
+      "step": 22914
+    },
+    {
+      "epoch": 2.9828192112456073,
+      "grad_norm": 2.6625006198883057,
+      "learning_rate": 8.455139995655615e-10,
+      "loss": 0.3032,
+      "step": 22917
+    },
+    {
+      "epoch": 2.983209683717298,
+      "grad_norm": 2.914444923400879,
+      "learning_rate": 8.075193002538006e-10,
+      "loss": 0.3967,
+      "step": 22920
+    },
+    {
+      "epoch": 2.983600156188989,
+      "grad_norm": 2.957470655441284,
+      "learning_rate": 7.703979493522973e-10,
+      "loss": 0.3776,
+      "step": 22923
+    },
+    {
+      "epoch": 2.9839906286606794,
+      "grad_norm": 3.152146100997925,
+      "learning_rate": 7.341499533447538e-10,
+      "loss": 0.3585,
+      "step": 22926
+    },
+    {
+      "epoch": 2.98438110113237,
+      "grad_norm": 2.8697428703308105,
+      "learning_rate": 6.987753185649926e-10,
+      "loss": 0.4167,
+      "step": 22929
+    },
+    {
+      "epoch": 2.984771573604061,
+      "grad_norm": 3.1371231079101562,
+      "learning_rate": 6.642740511914047e-10,
+      "loss": 0.3816,
+      "step": 22932
+    },
+    {
+      "epoch": 2.9851620460757515,
+      "grad_norm": 3.021503210067749,
+      "learning_rate": 6.306461572525013e-10,
+      "loss": 0.3535,
+      "step": 22935
+    },
+    {
+      "epoch": 2.9855525185474425,
+      "grad_norm": 2.5989933013916016,
+      "learning_rate": 5.978916426230274e-10,
+      "loss": 0.3127,
+      "step": 22938
+    },
+    {
+      "epoch": 2.985942991019133,
+      "grad_norm": 2.7822329998016357,
+      "learning_rate": 5.660105130239624e-10,
+      "loss": 0.354,
+      "step": 22941
+    },
+    {
+      "epoch": 2.9863334634908236,
+      "grad_norm": 2.7127301692962646,
+      "learning_rate": 5.350027740258501e-10,
+      "loss": 0.3868,
+      "step": 22944
+    },
+    {
+      "epoch": 2.9867239359625146,
+      "grad_norm": 3.0589182376861572,
+      "learning_rate": 5.048684310454688e-10,
+      "loss": 0.3267,
+      "step": 22947
+    },
+    {
+      "epoch": 2.9871144084342056,
+      "grad_norm": 2.661449670791626,
+      "learning_rate": 4.756074893469409e-10,
+      "loss": 0.3292,
+      "step": 22950
+    },
+    {
+      "epoch": 2.987504880905896,
+      "grad_norm": 2.9641380310058594,
+      "learning_rate": 4.4721995404284345e-10,
+      "loss": 0.4186,
+      "step": 22953
+    },
+    {
+      "epoch": 2.9878953533775867,
+      "grad_norm": 2.9655439853668213,
+      "learning_rate": 4.197058300914325e-10,
+      "loss": 0.3803,
+      "step": 22956
+    },
+    {
+      "epoch": 2.9882858258492777,
+      "grad_norm": 3.2909016609191895,
+      "learning_rate": 3.930651222999737e-10,
+      "loss": 0.391,
+      "step": 22959
+    },
+    {
+      "epoch": 2.9886762983209683,
+      "grad_norm": 2.9940195083618164,
+      "learning_rate": 3.6729783532196696e-10,
+      "loss": 0.3484,
+      "step": 22962
+    },
+    {
+      "epoch": 2.9890667707926593,
+      "grad_norm": 3.0727415084838867,
+      "learning_rate": 3.424039736599216e-10,
+      "loss": 0.3758,
+      "step": 22965
+    },
+    {
+      "epoch": 2.98945724326435,
+      "grad_norm": 2.5275540351867676,
+      "learning_rate": 3.1838354166202623e-10,
+      "loss": 0.3679,
+      "step": 22968
+    },
+    {
+      "epoch": 2.9898477157360404,
+      "grad_norm": 2.776193380355835,
+      "learning_rate": 2.9523654352492384e-10,
+      "loss": 0.354,
+      "step": 22971
+    },
+    {
+      "epoch": 2.9902381882077314,
+      "grad_norm": 2.7540743350982666,
+      "learning_rate": 2.7296298329204664e-10,
+      "loss": 0.3332,
+      "step": 22974
+    },
+    {
+      "epoch": 2.9906286606794223,
+      "grad_norm": 2.7912518978118896,
+      "learning_rate": 2.5156286485417126e-10,
+      "loss": 0.327,
+      "step": 22977
+    },
+    {
+      "epoch": 2.991019133151113,
+      "grad_norm": 3.1262850761413574,
+      "learning_rate": 2.3103619195052884e-10,
+      "loss": 0.3989,
+      "step": 22980
+    },
+    {
+      "epoch": 2.9914096056228034,
+      "grad_norm": 3.24723744392395,
+      "learning_rate": 2.1138296816713976e-10,
+      "loss": 0.3658,
+      "step": 22983
+    },
+    {
+      "epoch": 2.9918000780944944,
+      "grad_norm": 2.843860149383545,
+      "learning_rate": 1.9260319693736874e-10,
+      "loss": 0.3578,
+      "step": 22986
+    },
+    {
+      "epoch": 2.992190550566185,
+      "grad_norm": 3.0490710735321045,
+      "learning_rate": 1.7469688154136966e-10,
+      "loss": 0.3589,
+      "step": 22989
+    },
+    {
+      "epoch": 2.992581023037876,
+      "grad_norm": 2.6627109050750732,
+      "learning_rate": 1.5766402510775102e-10,
+      "loss": 0.3676,
+      "step": 22992
+    },
+    {
+      "epoch": 2.9929714955095665,
+      "grad_norm": 3.4538979530334473,
+      "learning_rate": 1.4150463061191055e-10,
+      "loss": 0.3702,
+      "step": 22995
+    },
+    {
+      "epoch": 2.993361967981257,
+      "grad_norm": 2.638334274291992,
+      "learning_rate": 1.2621870087714538e-10,
+      "loss": 0.3135,
+      "step": 22998
+    },
+    {
+      "epoch": 2.993752440452948,
+      "grad_norm": 2.811047077178955,
+      "learning_rate": 1.11806238574097e-10,
+      "loss": 0.3619,
+      "step": 23001
+    },
+    {
+      "epoch": 2.994142912924639,
+      "grad_norm": 2.9179980754852295,
+      "learning_rate": 9.826724622019613e-11,
+      "loss": 0.3805,
+      "step": 23004
+    },
+    {
+      "epoch": 2.9945333853963296,
+      "grad_norm": 2.6326956748962402,
+      "learning_rate": 8.56017261807729e-11,
+      "loss": 0.3641,
+      "step": 23007
+    },
+    {
+      "epoch": 2.99492385786802,
+      "grad_norm": 2.799358367919922,
+      "learning_rate": 7.380968066794669e-11,
+      "loss": 0.3647,
+      "step": 23010
+    },
+    {
+      "epoch": 2.995314330339711,
+      "grad_norm": 2.7768449783325195,
+      "learning_rate": 6.289111174284657e-11,
+      "loss": 0.3445,
+      "step": 23013
+    },
+    {
+      "epoch": 2.9957048028114017,
+      "grad_norm": 2.603933095932007,
+      "learning_rate": 5.284602131228056e-11,
+      "loss": 0.3638,
+      "step": 23016
+    },
+    {
+      "epoch": 2.9960952752830927,
+      "grad_norm": 2.9793689250946045,
+      "learning_rate": 4.367441113095616e-11,
+      "loss": 0.393,
+      "step": 23019
+    },
+    {
+      "epoch": 2.9964857477547833,
+      "grad_norm": 3.043372631072998,
+      "learning_rate": 3.53762828009252e-11,
+      "loss": 0.4288,
+      "step": 23022
+    },
+    {
+      "epoch": 2.996876220226474,
+      "grad_norm": 3.007805109024048,
+      "learning_rate": 2.795163777269405e-11,
+      "loss": 0.3576,
+      "step": 23025
+    },
+    {
+      "epoch": 2.997266692698165,
+      "grad_norm": 2.5332658290863037,
+      "learning_rate": 2.1400477343003213e-11,
+      "loss": 0.3063,
+      "step": 23028
+    },
+    {
+      "epoch": 2.9976571651698554,
+      "grad_norm": 2.6876015663146973,
+      "learning_rate": 1.572280265649262e-11,
+      "loss": 0.3617,
+      "step": 23031
+    },
+    {
+      "epoch": 2.9980476376415464,
+      "grad_norm": 3.169299840927124,
+      "learning_rate": 1.0918614704036323e-11,
+      "loss": 0.4214,
+      "step": 23034
+    },
+    {
+      "epoch": 2.998438110113237,
+      "grad_norm": 2.8437774181365967,
+      "learning_rate": 6.987914326073153e-12,
+      "loss": 0.3391,
+      "step": 23037
+    },
+    {
+      "epoch": 2.998828582584928,
+      "grad_norm": 3.0276384353637695,
+      "learning_rate": 3.93070220927605e-12,
+      "loss": 0.3807,
+      "step": 23040
+    },
+    {
+      "epoch": 2.9992190550566185,
+      "grad_norm": 2.6813414096832275,
+      "learning_rate": 1.7469788865520642e-12,
+      "loss": 0.3787,
+      "step": 23043
+    },
+    {
+      "epoch": 2.9996095275283094,
+      "grad_norm": 2.8834426403045654,
+      "learning_rate": 4.367447409281411e-13,
+      "loss": 0.3336,
+      "step": 23046
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.5650634765625,
+      "learning_rate": 0.0,
+      "loss": 0.2652,
+      "step": 23049
+    }
+  ],
+  "logging_steps": 3,
+  "max_steps": 23049,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.421608510493819e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}