diff --git "a/RAG-Tuning-S/trainer_state.json" "b/RAG-Tuning-S/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/RAG-Tuning-S/trainer_state.json"
@@ -0,0 +1,34636 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 24710,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00020234722784297855,
+      "grad_norm": 2.9933409690856934,
+      "learning_rate": 1.0101010101010103e-06,
+      "loss": 1.5596,
+      "step": 5
+    },
+    {
+      "epoch": 0.0004046944556859571,
+      "grad_norm": 2.9861791133880615,
+      "learning_rate": 2.0202020202020206e-06,
+      "loss": 1.5265,
+      "step": 10
+    },
+    {
+      "epoch": 0.0006070416835289356,
+      "grad_norm": 2.7128653526306152,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 1.5348,
+      "step": 15
+    },
+    {
+      "epoch": 0.0008093889113719142,
+      "grad_norm": 2.5973939895629883,
+      "learning_rate": 4.040404040404041e-06,
+      "loss": 1.5126,
+      "step": 20
+    },
+    {
+      "epoch": 0.0010117361392148927,
+      "grad_norm": 2.5141825675964355,
+      "learning_rate": 5.050505050505051e-06,
+      "loss": 1.4586,
+      "step": 25
+    },
+    {
+      "epoch": 0.0012140833670578712,
+      "grad_norm": 2.514392137527466,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 1.4285,
+      "step": 30
+    },
+    {
+      "epoch": 0.00141643059490085,
+      "grad_norm": 2.238926410675049,
+      "learning_rate": 7.0707070707070704e-06,
+      "loss": 1.4558,
+      "step": 35
+    },
+    {
+      "epoch": 0.0016187778227438284,
+      "grad_norm": 2.2784197330474854,
+      "learning_rate": 8.080808080808082e-06,
+      "loss": 1.3484,
+      "step": 40
+    },
+    {
+      "epoch": 0.001821125050586807,
+      "grad_norm": 1.9064857959747314,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 1.3921,
+      "step": 45
+    },
+    {
+      "epoch": 0.0020234722784297854,
+      "grad_norm": 1.7825236320495605,
+      "learning_rate": 1.0101010101010101e-05,
+      "loss": 1.3447,
+      "step": 50
+    },
+    {
+      "epoch": 0.002225819506272764,
+      "grad_norm": 1.9339872598648071,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 1.3706,
+      "step": 55
+    },
+    {
+      "epoch": 0.0024281667341157424,
+      "grad_norm": 1.8017120361328125,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 1.2476,
+      "step": 60
+    },
+    {
+      "epoch": 0.002630513961958721,
+      "grad_norm": 2.308884620666504,
+      "learning_rate": 1.3131313131313134e-05,
+      "loss": 1.2883,
+      "step": 65
+    },
+    {
+      "epoch": 0.0028328611898017,
+      "grad_norm": 2.2627971172332764,
+      "learning_rate": 1.4141414141414141e-05,
+      "loss": 1.2553,
+      "step": 70
+    },
+    {
+      "epoch": 0.003035208417644678,
+      "grad_norm": 1.89839768409729,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 1.1874,
+      "step": 75
+    },
+    {
+      "epoch": 0.003237555645487657,
+      "grad_norm": 2.02109956741333,
+      "learning_rate": 1.6161616161616165e-05,
+      "loss": 1.1467,
+      "step": 80
+    },
+    {
+      "epoch": 0.0034399028733306356,
+      "grad_norm": 2.418301820755005,
+      "learning_rate": 1.7171717171717173e-05,
+      "loss": 1.1796,
+      "step": 85
+    },
+    {
+      "epoch": 0.003642250101173614,
+      "grad_norm": 1.8888823986053467,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 1.1421,
+      "step": 90
+    },
+    {
+      "epoch": 0.0038445973290165926,
+      "grad_norm": 1.6717690229415894,
+      "learning_rate": 1.919191919191919e-05,
+      "loss": 1.1976,
+      "step": 95
+    },
+    {
+      "epoch": 0.004046944556859571,
+      "grad_norm": 1.9028016328811646,
+      "learning_rate": 2.0202020202020203e-05,
+      "loss": 1.1088,
+      "step": 100
+    },
+    {
+      "epoch": 0.00424929178470255,
+      "grad_norm": 1.8961031436920166,
+      "learning_rate": 2.1212121212121215e-05,
+      "loss": 1.1656,
+      "step": 105
+    },
+    {
+      "epoch": 0.004451639012545528,
+      "grad_norm": 2.270857095718384,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 1.1573,
+      "step": 110
+    },
+    {
+      "epoch": 0.0046539862403885066,
+      "grad_norm": 1.9814001321792603,
+      "learning_rate": 2.3232323232323232e-05,
+      "loss": 1.1281,
+      "step": 115
+    },
+    {
+      "epoch": 0.004856333468231485,
+      "grad_norm": 2.3769867420196533,
+      "learning_rate": 2.4242424242424244e-05,
+      "loss": 1.1401,
+      "step": 120
+    },
+    {
+      "epoch": 0.005058680696074464,
+      "grad_norm": 3.6437642574310303,
+      "learning_rate": 2.5252525252525256e-05,
+      "loss": 1.1035,
+      "step": 125
+    },
+    {
+      "epoch": 0.005261027923917442,
+      "grad_norm": 2.351130962371826,
+      "learning_rate": 2.6262626262626268e-05,
+      "loss": 1.1744,
+      "step": 130
+    },
+    {
+      "epoch": 0.0054633751517604206,
+      "grad_norm": 2.160093069076538,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 1.1117,
+      "step": 135
+    },
+    {
+      "epoch": 0.0056657223796034,
+      "grad_norm": 2.536659002304077,
+      "learning_rate": 2.8282828282828282e-05,
+      "loss": 1.1664,
+      "step": 140
+    },
+    {
+      "epoch": 0.005868069607446378,
+      "grad_norm": 2.9314124584198,
+      "learning_rate": 2.9292929292929294e-05,
+      "loss": 1.0955,
+      "step": 145
+    },
+    {
+      "epoch": 0.006070416835289356,
+      "grad_norm": 2.7155613899230957,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 1.1312,
+      "step": 150
+    },
+    {
+      "epoch": 0.006272764063132335,
+      "grad_norm": 2.258028268814087,
+      "learning_rate": 3.131313131313132e-05,
+      "loss": 1.088,
+      "step": 155
+    },
+    {
+      "epoch": 0.006475111290975314,
+      "grad_norm": 2.6700029373168945,
+      "learning_rate": 3.232323232323233e-05,
+      "loss": 1.183,
+      "step": 160
+    },
+    {
+      "epoch": 0.006677458518818292,
+      "grad_norm": 2.71748948097229,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.1442,
+      "step": 165
+    },
+    {
+      "epoch": 0.006879805746661271,
+      "grad_norm": 2.9919252395629883,
+      "learning_rate": 3.434343434343435e-05,
+      "loss": 1.0959,
+      "step": 170
+    },
+    {
+      "epoch": 0.007082152974504249,
+      "grad_norm": 2.545902967453003,
+      "learning_rate": 3.535353535353535e-05,
+      "loss": 1.1813,
+      "step": 175
+    },
+    {
+      "epoch": 0.007284500202347228,
+      "grad_norm": 2.7845704555511475,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 1.0744,
+      "step": 180
+    },
+    {
+      "epoch": 0.007486847430190206,
+      "grad_norm": 3.071317672729492,
+      "learning_rate": 3.7373737373737376e-05,
+      "loss": 1.1003,
+      "step": 185
+    },
+    {
+      "epoch": 0.007689194658033185,
+      "grad_norm": 2.471811294555664,
+      "learning_rate": 3.838383838383838e-05,
+      "loss": 1.1205,
+      "step": 190
+    },
+    {
+      "epoch": 0.007891541885876164,
+      "grad_norm": 2.479722738265991,
+      "learning_rate": 3.939393939393939e-05,
+      "loss": 1.1572,
+      "step": 195
+    },
+    {
+      "epoch": 0.008093889113719142,
+      "grad_norm": 2.5352907180786133,
+      "learning_rate": 4.0404040404040405e-05,
+      "loss": 1.0776,
+      "step": 200
+    },
+    {
+      "epoch": 0.00829623634156212,
+      "grad_norm": 2.292621612548828,
+      "learning_rate": 4.141414141414142e-05,
+      "loss": 1.1347,
+      "step": 205
+    },
+    {
+      "epoch": 0.0084985835694051,
+      "grad_norm": 2.502856731414795,
+      "learning_rate": 4.242424242424243e-05,
+      "loss": 1.1138,
+      "step": 210
+    },
+    {
+      "epoch": 0.008700930797248077,
+      "grad_norm": 2.574815034866333,
+      "learning_rate": 4.343434343434344e-05,
+      "loss": 1.095,
+      "step": 215
+    },
+    {
+      "epoch": 0.008903278025091057,
+      "grad_norm": 2.8264522552490234,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 1.1218,
+      "step": 220
+    },
+    {
+      "epoch": 0.009105625252934034,
+      "grad_norm": 2.918670892715454,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 1.1477,
+      "step": 225
+    },
+    {
+      "epoch": 0.009307972480777013,
+      "grad_norm": 2.5850000381469727,
+      "learning_rate": 4.6464646464646464e-05,
+      "loss": 1.06,
+      "step": 230
+    },
+    {
+      "epoch": 0.009510319708619992,
+      "grad_norm": 2.5582289695739746,
+      "learning_rate": 4.7474747474747476e-05,
+      "loss": 1.1202,
+      "step": 235
+    },
+    {
+      "epoch": 0.00971266693646297,
+      "grad_norm": 2.556922197341919,
+      "learning_rate": 4.848484848484849e-05,
+      "loss": 1.0806,
+      "step": 240
+    },
+    {
+      "epoch": 0.009915014164305949,
+      "grad_norm": 2.5984082221984863,
+      "learning_rate": 4.94949494949495e-05,
+      "loss": 1.1722,
+      "step": 245
+    },
+    {
+      "epoch": 0.010117361392148928,
+      "grad_norm": 2.8163859844207764,
+      "learning_rate": 5.050505050505051e-05,
+      "loss": 1.121,
+      "step": 250
+    },
+    {
+      "epoch": 0.010319708619991905,
+      "grad_norm": 2.8502392768859863,
+      "learning_rate": 5.151515151515152e-05,
+      "loss": 1.0742,
+      "step": 255
+    },
+    {
+      "epoch": 0.010522055847834885,
+      "grad_norm": 2.6049704551696777,
+      "learning_rate": 5.2525252525252536e-05,
+      "loss": 1.1487,
+      "step": 260
+    },
+    {
+      "epoch": 0.010724403075677864,
+      "grad_norm": 2.836660861968994,
+      "learning_rate": 5.353535353535354e-05,
+      "loss": 1.0866,
+      "step": 265
+    },
+    {
+      "epoch": 0.010926750303520841,
+      "grad_norm": 2.7358970642089844,
+      "learning_rate": 5.4545454545454546e-05,
+      "loss": 1.1158,
+      "step": 270
+    },
+    {
+      "epoch": 0.01112909753136382,
+      "grad_norm": 2.190136194229126,
+      "learning_rate": 5.555555555555556e-05,
+      "loss": 1.0547,
+      "step": 275
+    },
+    {
+      "epoch": 0.0113314447592068,
+      "grad_norm": 3.045732021331787,
+      "learning_rate": 5.6565656565656563e-05,
+      "loss": 1.0845,
+      "step": 280
+    },
+    {
+      "epoch": 0.011533791987049777,
+      "grad_norm": 2.5552291870117188,
+      "learning_rate": 5.757575757575758e-05,
+      "loss": 1.1007,
+      "step": 285
+    },
+    {
+      "epoch": 0.011736139214892756,
+      "grad_norm": 2.478412628173828,
+      "learning_rate": 5.858585858585859e-05,
+      "loss": 1.2019,
+      "step": 290
+    },
+    {
+      "epoch": 0.011938486442735735,
+      "grad_norm": 2.5224175453186035,
+      "learning_rate": 5.959595959595959e-05,
+      "loss": 1.031,
+      "step": 295
+    },
+    {
+      "epoch": 0.012140833670578713,
+      "grad_norm": 2.516164779663086,
+      "learning_rate": 6.060606060606061e-05,
+      "loss": 1.052,
+      "step": 300
+    },
+    {
+      "epoch": 0.012343180898421692,
+      "grad_norm": 2.3297924995422363,
+      "learning_rate": 6.161616161616162e-05,
+      "loss": 1.1779,
+      "step": 305
+    },
+    {
+      "epoch": 0.01254552812626467,
+      "grad_norm": 2.5777242183685303,
+      "learning_rate": 6.262626262626264e-05,
+      "loss": 1.0896,
+      "step": 310
+    },
+    {
+      "epoch": 0.012747875354107648,
+      "grad_norm": 2.1272308826446533,
+      "learning_rate": 6.363636363636364e-05,
+      "loss": 1.1065,
+      "step": 315
+    },
+    {
+      "epoch": 0.012950222581950627,
+      "grad_norm": 2.5600132942199707,
+      "learning_rate": 6.464646464646466e-05,
+      "loss": 1.1035,
+      "step": 320
+    },
+    {
+      "epoch": 0.013152569809793607,
+      "grad_norm": 2.5389156341552734,
+      "learning_rate": 6.565656565656566e-05,
+      "loss": 1.0641,
+      "step": 325
+    },
+    {
+      "epoch": 0.013354917037636584,
+      "grad_norm": 1.9946951866149902,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 1.0919,
+      "step": 330
+    },
+    {
+      "epoch": 0.013557264265479563,
+      "grad_norm": 2.1555545330047607,
+      "learning_rate": 6.767676767676769e-05,
+      "loss": 1.0443,
+      "step": 335
+    },
+    {
+      "epoch": 0.013759611493322542,
+      "grad_norm": 2.399268865585327,
+      "learning_rate": 6.86868686868687e-05,
+      "loss": 1.0891,
+      "step": 340
+    },
+    {
+      "epoch": 0.01396195872116552,
+      "grad_norm": 2.261282205581665,
+      "learning_rate": 6.96969696969697e-05,
+      "loss": 1.0574,
+      "step": 345
+    },
+    {
+      "epoch": 0.014164305949008499,
+      "grad_norm": 2.1969683170318604,
+      "learning_rate": 7.07070707070707e-05,
+      "loss": 1.1118,
+      "step": 350
+    },
+    {
+      "epoch": 0.014366653176851478,
+      "grad_norm": 2.359020471572876,
+      "learning_rate": 7.171717171717171e-05,
+      "loss": 1.1636,
+      "step": 355
+    },
+    {
+      "epoch": 0.014569000404694455,
+      "grad_norm": 2.1169421672821045,
+      "learning_rate": 7.272727272727273e-05,
+      "loss": 1.0405,
+      "step": 360
+    },
+    {
+      "epoch": 0.014771347632537435,
+      "grad_norm": 2.5211663246154785,
+      "learning_rate": 7.373737373737373e-05,
+      "loss": 1.1449,
+      "step": 365
+    },
+    {
+      "epoch": 0.014973694860380412,
+      "grad_norm": 2.165107250213623,
+      "learning_rate": 7.474747474747475e-05,
+      "loss": 1.102,
+      "step": 370
+    },
+    {
+      "epoch": 0.015176042088223391,
+      "grad_norm": 2.2055845260620117,
+      "learning_rate": 7.575757575757576e-05,
+      "loss": 1.1009,
+      "step": 375
+    },
+    {
+      "epoch": 0.01537838931606637,
+      "grad_norm": 2.101855516433716,
+      "learning_rate": 7.676767676767676e-05,
+      "loss": 1.0385,
+      "step": 380
+    },
+    {
+      "epoch": 0.015580736543909348,
+      "grad_norm": 2.1183292865753174,
+      "learning_rate": 7.777777777777778e-05,
+      "loss": 1.0664,
+      "step": 385
+    },
+    {
+      "epoch": 0.01578308377175233,
+      "grad_norm": 2.391688823699951,
+      "learning_rate": 7.878787878787879e-05,
+      "loss": 1.095,
+      "step": 390
+    },
+    {
+      "epoch": 0.015985430999595304,
+      "grad_norm": 2.121948719024658,
+      "learning_rate": 7.97979797979798e-05,
+      "loss": 1.0647,
+      "step": 395
+    },
+    {
+      "epoch": 0.016187778227438283,
+      "grad_norm": 2.142220973968506,
+      "learning_rate": 8.080808080808081e-05,
+      "loss": 1.1395,
+      "step": 400
+    },
+    {
+      "epoch": 0.016390125455281263,
+      "grad_norm": 2.2406890392303467,
+      "learning_rate": 8.181818181818183e-05,
+      "loss": 1.0615,
+      "step": 405
+    },
+    {
+      "epoch": 0.01659247268312424,
+      "grad_norm": 2.036979913711548,
+      "learning_rate": 8.282828282828283e-05,
+      "loss": 1.0771,
+      "step": 410
+    },
+    {
+      "epoch": 0.01679481991096722,
+      "grad_norm": 1.961674690246582,
+      "learning_rate": 8.383838383838384e-05,
+      "loss": 1.0595,
+      "step": 415
+    },
+    {
+      "epoch": 0.0169971671388102,
+      "grad_norm": 1.9730347394943237,
+      "learning_rate": 8.484848484848486e-05,
+      "loss": 0.9874,
+      "step": 420
+    },
+    {
+      "epoch": 0.017199514366653176,
+      "grad_norm": 1.7028435468673706,
+      "learning_rate": 8.585858585858586e-05,
+      "loss": 1.071,
+      "step": 425
+    },
+    {
+      "epoch": 0.017401861594496155,
+      "grad_norm": 1.8906868696212769,
+      "learning_rate": 8.686868686868688e-05,
+      "loss": 1.0726,
+      "step": 430
+    },
+    {
+      "epoch": 0.017604208822339134,
+      "grad_norm": 1.80305016040802,
+      "learning_rate": 8.787878787878789e-05,
+      "loss": 1.0412,
+      "step": 435
+    },
+    {
+      "epoch": 0.017806556050182113,
+      "grad_norm": 2.0835936069488525,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 1.1184,
+      "step": 440
+    },
+    {
+      "epoch": 0.018008903278025092,
+      "grad_norm": 1.9061412811279297,
+      "learning_rate": 8.98989898989899e-05,
+      "loss": 1.0537,
+      "step": 445
+    },
+    {
+      "epoch": 0.018211250505868068,
+      "grad_norm": 2.6074249744415283,
+      "learning_rate": 9.090909090909092e-05,
+      "loss": 1.088,
+      "step": 450
+    },
+    {
+      "epoch": 0.018413597733711047,
+      "grad_norm": 1.6072639226913452,
+      "learning_rate": 9.191919191919192e-05,
+      "loss": 1.1465,
+      "step": 455
+    },
+    {
+      "epoch": 0.018615944961554026,
+      "grad_norm": 1.7711330652236938,
+      "learning_rate": 9.292929292929293e-05,
+      "loss": 1.0707,
+      "step": 460
+    },
+    {
+      "epoch": 0.018818292189397005,
+      "grad_norm": 2.105698823928833,
+      "learning_rate": 9.393939393939395e-05,
+      "loss": 1.0771,
+      "step": 465
+    },
+    {
+      "epoch": 0.019020639417239985,
+      "grad_norm": 2.5414555072784424,
+      "learning_rate": 9.494949494949495e-05,
+      "loss": 1.1593,
+      "step": 470
+    },
+    {
+      "epoch": 0.019222986645082964,
+      "grad_norm": 1.7689402103424072,
+      "learning_rate": 9.595959595959596e-05,
+      "loss": 1.0731,
+      "step": 475
+    },
+    {
+      "epoch": 0.01942533387292594,
+      "grad_norm": 1.7413325309753418,
+      "learning_rate": 9.696969696969698e-05,
+      "loss": 1.048,
+      "step": 480
+    },
+    {
+      "epoch": 0.01962768110076892,
+      "grad_norm": 1.7363505363464355,
+      "learning_rate": 9.797979797979798e-05,
+      "loss": 1.0845,
+      "step": 485
+    },
+    {
+      "epoch": 0.019830028328611898,
+      "grad_norm": 1.6728994846343994,
+      "learning_rate": 9.8989898989899e-05,
+      "loss": 1.1154,
+      "step": 490
+    },
+    {
+      "epoch": 0.020032375556454877,
+      "grad_norm": 2.205254077911377,
+      "learning_rate": 0.0001,
+      "loss": 1.0316,
+      "step": 495
+    },
+    {
+      "epoch": 0.020234722784297856,
+      "grad_norm": 1.5651170015335083,
+      "learning_rate": 9.997935164154451e-05,
+      "loss": 1.1225,
+      "step": 500
+    },
+    {
+      "epoch": 0.020437070012140835,
+      "grad_norm": 1.6764373779296875,
+      "learning_rate": 9.9958703283089e-05,
+      "loss": 1.0609,
+      "step": 505
+    },
+    {
+      "epoch": 0.02063941723998381,
+      "grad_norm": 1.4497535228729248,
+      "learning_rate": 9.993805492463349e-05,
+      "loss": 1.0826,
+      "step": 510
+    },
+    {
+      "epoch": 0.02084176446782679,
+      "grad_norm": 1.9358776807785034,
+      "learning_rate": 9.9917406566178e-05,
+      "loss": 1.076,
+      "step": 515
+    },
+    {
+      "epoch": 0.02104411169566977,
+      "grad_norm": 1.9405758380889893,
+      "learning_rate": 9.98967582077225e-05,
+      "loss": 1.0944,
+      "step": 520
+    },
+    {
+      "epoch": 0.021246458923512748,
+      "grad_norm": 1.8175926208496094,
+      "learning_rate": 9.987610984926699e-05,
+      "loss": 1.0325,
+      "step": 525
+    },
+    {
+      "epoch": 0.021448806151355727,
+      "grad_norm": 1.5051298141479492,
+      "learning_rate": 9.985546149081149e-05,
+      "loss": 1.087,
+      "step": 530
+    },
+    {
+      "epoch": 0.021651153379198707,
+      "grad_norm": 1.5366894006729126,
+      "learning_rate": 9.983481313235598e-05,
+      "loss": 1.0605,
+      "step": 535
+    },
+    {
+      "epoch": 0.021853500607041682,
+      "grad_norm": 1.655651330947876,
+      "learning_rate": 9.981416477390047e-05,
+      "loss": 1.0321,
+      "step": 540
+    },
+    {
+      "epoch": 0.02205584783488466,
+      "grad_norm": 1.7619483470916748,
+      "learning_rate": 9.979351641544498e-05,
+      "loss": 1.0845,
+      "step": 545
+    },
+    {
+      "epoch": 0.02225819506272764,
+      "grad_norm": 1.6305811405181885,
+      "learning_rate": 9.977286805698948e-05,
+      "loss": 1.1212,
+      "step": 550
+    },
+    {
+      "epoch": 0.02246054229057062,
+      "grad_norm": 1.5049748420715332,
+      "learning_rate": 9.975221969853397e-05,
+      "loss": 1.0087,
+      "step": 555
+    },
+    {
+      "epoch": 0.0226628895184136,
+      "grad_norm": 1.5912582874298096,
+      "learning_rate": 9.973157134007846e-05,
+      "loss": 1.046,
+      "step": 560
+    },
+    {
+      "epoch": 0.022865236746256578,
+      "grad_norm": 1.4049822092056274,
+      "learning_rate": 9.971092298162297e-05,
+      "loss": 1.0532,
+      "step": 565
+    },
+    {
+      "epoch": 0.023067583974099554,
+      "grad_norm": 1.6603472232818604,
+      "learning_rate": 9.969027462316747e-05,
+      "loss": 1.081,
+      "step": 570
+    },
+    {
+      "epoch": 0.023269931201942533,
+      "grad_norm": 1.5358834266662598,
+      "learning_rate": 9.966962626471196e-05,
+      "loss": 1.0783,
+      "step": 575
+    },
+    {
+      "epoch": 0.023472278429785512,
+      "grad_norm": 1.5615278482437134,
+      "learning_rate": 9.964897790625645e-05,
+      "loss": 1.0672,
+      "step": 580
+    },
+    {
+      "epoch": 0.02367462565762849,
+      "grad_norm": 1.6925407648086548,
+      "learning_rate": 9.962832954780096e-05,
+      "loss": 1.0514,
+      "step": 585
+    },
+    {
+      "epoch": 0.02387697288547147,
+      "grad_norm": 1.834684133529663,
+      "learning_rate": 9.960768118934545e-05,
+      "loss": 1.0672,
+      "step": 590
+    },
+    {
+      "epoch": 0.024079320113314446,
+      "grad_norm": 1.6813185214996338,
+      "learning_rate": 9.958703283088995e-05,
+      "loss": 1.0448,
+      "step": 595
+    },
+    {
+      "epoch": 0.024281667341157425,
+      "grad_norm": 1.5269089937210083,
+      "learning_rate": 9.956638447243445e-05,
+      "loss": 1.1439,
+      "step": 600
+    },
+    {
+      "epoch": 0.024484014569000404,
+      "grad_norm": 1.4650517702102661,
+      "learning_rate": 9.954573611397893e-05,
+      "loss": 1.0276,
+      "step": 605
+    },
+    {
+      "epoch": 0.024686361796843383,
+      "grad_norm": 1.6329649686813354,
+      "learning_rate": 9.952508775552344e-05,
+      "loss": 1.0589,
+      "step": 610
+    },
+    {
+      "epoch": 0.024888709024686363,
+      "grad_norm": 1.4608739614486694,
+      "learning_rate": 9.950443939706794e-05,
+      "loss": 1.1105,
+      "step": 615
+    },
+    {
+      "epoch": 0.02509105625252934,
+      "grad_norm": 1.4417763948440552,
+      "learning_rate": 9.948379103861244e-05,
+      "loss": 1.0282,
+      "step": 620
+    },
+    {
+      "epoch": 0.025293403480372317,
+      "grad_norm": 1.7151459455490112,
+      "learning_rate": 9.946314268015693e-05,
+      "loss": 1.0601,
+      "step": 625
+    },
+    {
+      "epoch": 0.025495750708215296,
+      "grad_norm": 1.565805435180664,
+      "learning_rate": 9.944249432170143e-05,
+      "loss": 1.0962,
+      "step": 630
+    },
+    {
+      "epoch": 0.025698097936058276,
+      "grad_norm": 1.6773918867111206,
+      "learning_rate": 9.942184596324593e-05,
+      "loss": 1.1163,
+      "step": 635
+    },
+    {
+      "epoch": 0.025900445163901255,
+      "grad_norm": 1.8091540336608887,
+      "learning_rate": 9.940119760479042e-05,
+      "loss": 1.0751,
+      "step": 640
+    },
+    {
+      "epoch": 0.026102792391744234,
+      "grad_norm": 1.5862232446670532,
+      "learning_rate": 9.938054924633492e-05,
+      "loss": 1.0864,
+      "step": 645
+    },
+    {
+      "epoch": 0.026305139619587213,
+      "grad_norm": 1.5611926317214966,
+      "learning_rate": 9.935990088787941e-05,
+      "loss": 1.0644,
+      "step": 650
+    },
+    {
+      "epoch": 0.02650748684743019,
+      "grad_norm": 1.6312520503997803,
+      "learning_rate": 9.933925252942392e-05,
+      "loss": 1.096,
+      "step": 655
+    },
+    {
+      "epoch": 0.026709834075273168,
+      "grad_norm": 1.7054439783096313,
+      "learning_rate": 9.931860417096841e-05,
+      "loss": 1.0949,
+      "step": 660
+    },
+    {
+      "epoch": 0.026912181303116147,
+      "grad_norm": 1.5077580213546753,
+      "learning_rate": 9.929795581251291e-05,
+      "loss": 1.0756,
+      "step": 665
+    },
+    {
+      "epoch": 0.027114528530959126,
+      "grad_norm": 1.8034014701843262,
+      "learning_rate": 9.927730745405742e-05,
+      "loss": 1.0091,
+      "step": 670
+    },
+    {
+      "epoch": 0.027316875758802105,
+      "grad_norm": 1.5382342338562012,
+      "learning_rate": 9.92566590956019e-05,
+      "loss": 1.1142,
+      "step": 675
+    },
+    {
+      "epoch": 0.027519222986645085,
+      "grad_norm": 1.607927680015564,
+      "learning_rate": 9.92360107371464e-05,
+      "loss": 1.0273,
+      "step": 680
+    },
+    {
+      "epoch": 0.02772157021448806,
+      "grad_norm": 1.561005711555481,
+      "learning_rate": 9.92153623786909e-05,
+      "loss": 1.0857,
+      "step": 685
+    },
+    {
+      "epoch": 0.02792391744233104,
+      "grad_norm": 1.7696928977966309,
+      "learning_rate": 9.919471402023539e-05,
+      "loss": 1.0289,
+      "step": 690
+    },
+    {
+      "epoch": 0.02812626467017402,
+      "grad_norm": 1.628246545791626,
+      "learning_rate": 9.91740656617799e-05,
+      "loss": 1.0224,
+      "step": 695
+    },
+    {
+      "epoch": 0.028328611898016998,
+      "grad_norm": 1.578949213027954,
+      "learning_rate": 9.915341730332439e-05,
+      "loss": 1.0747,
+      "step": 700
+    },
+    {
+      "epoch": 0.028530959125859977,
+      "grad_norm": 1.3843584060668945,
+      "learning_rate": 9.913276894486889e-05,
+      "loss": 1.0602,
+      "step": 705
+    },
+    {
+      "epoch": 0.028733306353702956,
+      "grad_norm": 1.7278279066085815,
+      "learning_rate": 9.911212058641338e-05,
+      "loss": 1.0388,
+      "step": 710
+    },
+    {
+      "epoch": 0.02893565358154593,
+      "grad_norm": 1.4063900709152222,
+      "learning_rate": 9.909147222795789e-05,
+      "loss": 1.0551,
+      "step": 715
+    },
+    {
+      "epoch": 0.02913800080938891,
+      "grad_norm": 1.477317214012146,
+      "learning_rate": 9.907082386950238e-05,
+      "loss": 1.0041,
+      "step": 720
+    },
+    {
+      "epoch": 0.02934034803723189,
+      "grad_norm": 1.8149793148040771,
+      "learning_rate": 9.905017551104687e-05,
+      "loss": 1.0325,
+      "step": 725
+    },
+    {
+      "epoch": 0.02954269526507487,
+      "grad_norm": 1.5079026222229004,
+      "learning_rate": 9.902952715259137e-05,
+      "loss": 1.0383,
+      "step": 730
+    },
+    {
+      "epoch": 0.029745042492917848,
+      "grad_norm": 1.6261570453643799,
+      "learning_rate": 9.900887879413588e-05,
+      "loss": 1.0451,
+      "step": 735
+    },
+    {
+      "epoch": 0.029947389720760824,
+      "grad_norm": 1.351682186126709,
+      "learning_rate": 9.898823043568038e-05,
+      "loss": 1.1057,
+      "step": 740
+    },
+    {
+      "epoch": 0.030149736948603803,
+      "grad_norm": 1.3424296379089355,
+      "learning_rate": 9.896758207722487e-05,
+      "loss": 1.0734,
+      "step": 745
+    },
+    {
+      "epoch": 0.030352084176446782,
+      "grad_norm": 1.5590558052062988,
+      "learning_rate": 9.894693371876936e-05,
+      "loss": 1.0689,
+      "step": 750
+    },
+    {
+      "epoch": 0.03055443140428976,
+      "grad_norm": 1.5399047136306763,
+      "learning_rate": 9.892628536031386e-05,
+      "loss": 1.0878,
+      "step": 755
+    },
+    {
+      "epoch": 0.03075677863213274,
+      "grad_norm": 1.5072790384292603,
+      "learning_rate": 9.890563700185835e-05,
+      "loss": 1.0718,
+      "step": 760
+    },
+    {
+      "epoch": 0.03095912585997572,
+      "grad_norm": 1.5990835428237915,
+      "learning_rate": 9.888498864340286e-05,
+      "loss": 1.0803,
+      "step": 765
+    },
+    {
+      "epoch": 0.031161473087818695,
+      "grad_norm": 1.5335228443145752,
+      "learning_rate": 9.886434028494735e-05,
+      "loss": 1.0789,
+      "step": 770
+    },
+    {
+      "epoch": 0.03136382031566168,
+      "grad_norm": 1.611234188079834,
+      "learning_rate": 9.884369192649184e-05,
+      "loss": 1.0253,
+      "step": 775
+    },
+    {
+      "epoch": 0.03156616754350466,
+      "grad_norm": 1.4838621616363525,
+      "learning_rate": 9.882304356803634e-05,
+      "loss": 1.0687,
+      "step": 780
+    },
+    {
+      "epoch": 0.03176851477134763,
+      "grad_norm": 1.3513612747192383,
+      "learning_rate": 9.880239520958085e-05,
+      "loss": 1.061,
+      "step": 785
+    },
+    {
+      "epoch": 0.03197086199919061,
+      "grad_norm": 1.3502790927886963,
+      "learning_rate": 9.878174685112534e-05,
+      "loss": 1.0791,
+      "step": 790
+    },
+    {
+      "epoch": 0.03217320922703359,
+      "grad_norm": 1.496467113494873,
+      "learning_rate": 9.876109849266983e-05,
+      "loss": 1.0942,
+      "step": 795
+    },
+    {
+      "epoch": 0.03237555645487657,
+      "grad_norm": 1.894826889038086,
+      "learning_rate": 9.874045013421433e-05,
+      "loss": 1.0523,
+      "step": 800
+    },
+    {
+      "epoch": 0.032577903682719546,
+      "grad_norm": 1.5444755554199219,
+      "learning_rate": 9.871980177575884e-05,
+      "loss": 1.044,
+      "step": 805
+    },
+    {
+      "epoch": 0.032780250910562525,
+      "grad_norm": 1.4778664112091064,
+      "learning_rate": 9.869915341730333e-05,
+      "loss": 1.0029,
+      "step": 810
+    },
+    {
+      "epoch": 0.032982598138405504,
+      "grad_norm": 1.549651861190796,
+      "learning_rate": 9.867850505884783e-05,
+      "loss": 1.0215,
+      "step": 815
+    },
+    {
+      "epoch": 0.03318494536624848,
+      "grad_norm": 1.4651092290878296,
+      "learning_rate": 9.865785670039232e-05,
+      "loss": 0.9801,
+      "step": 820
+    },
+    {
+      "epoch": 0.03338729259409146,
+      "grad_norm": 1.2966331243515015,
+      "learning_rate": 9.863720834193681e-05,
+      "loss": 1.0499,
+      "step": 825
+    },
+    {
+      "epoch": 0.03358963982193444,
+      "grad_norm": 1.3640475273132324,
+      "learning_rate": 9.861655998348132e-05,
+      "loss": 1.0223,
+      "step": 830
+    },
+    {
+      "epoch": 0.03379198704977742,
+      "grad_norm": 1.25409734249115,
+      "learning_rate": 9.859591162502582e-05,
+      "loss": 1.1069,
+      "step": 835
+    },
+    {
+      "epoch": 0.0339943342776204,
+      "grad_norm": 1.547042965888977,
+      "learning_rate": 9.857526326657031e-05,
+      "loss": 1.0552,
+      "step": 840
+    },
+    {
+      "epoch": 0.03419668150546337,
+      "grad_norm": 1.4446684122085571,
+      "learning_rate": 9.85546149081148e-05,
+      "loss": 1.0126,
+      "step": 845
+    },
+    {
+      "epoch": 0.03439902873330635,
+      "grad_norm": 1.2919784784317017,
+      "learning_rate": 9.85339665496593e-05,
+      "loss": 1.0333,
+      "step": 850
+    },
+    {
+      "epoch": 0.03460137596114933,
+      "grad_norm": 1.2963666915893555,
+      "learning_rate": 9.851331819120381e-05,
+      "loss": 1.04,
+      "step": 855
+    },
+    {
+      "epoch": 0.03480372318899231,
+      "grad_norm": 1.4333245754241943,
+      "learning_rate": 9.84926698327483e-05,
+      "loss": 1.0674,
+      "step": 860
+    },
+    {
+      "epoch": 0.03500607041683529,
+      "grad_norm": 1.4501776695251465,
+      "learning_rate": 9.847202147429279e-05,
+      "loss": 1.1066,
+      "step": 865
+    },
+    {
+      "epoch": 0.03520841764467827,
+      "grad_norm": 1.4631706476211548,
+      "learning_rate": 9.84513731158373e-05,
+      "loss": 1.0648,
+      "step": 870
+    },
+    {
+      "epoch": 0.03541076487252125,
+      "grad_norm": 1.4564629793167114,
+      "learning_rate": 9.84307247573818e-05,
+      "loss": 1.0179,
+      "step": 875
+    },
+    {
+      "epoch": 0.035613112100364226,
+      "grad_norm": 1.3551617860794067,
+      "learning_rate": 9.841007639892629e-05,
+      "loss": 1.0684,
+      "step": 880
+    },
+    {
+      "epoch": 0.035815459328207205,
+      "grad_norm": 1.4886606931686401,
+      "learning_rate": 9.83894280404708e-05,
+      "loss": 1.0865,
+      "step": 885
+    },
+    {
+      "epoch": 0.036017806556050184,
+      "grad_norm": 1.4882872104644775,
+      "learning_rate": 9.836877968201528e-05,
+      "loss": 1.0389,
+      "step": 890
+    },
+    {
+      "epoch": 0.036220153783893164,
+      "grad_norm": 1.275161862373352,
+      "learning_rate": 9.834813132355977e-05,
+      "loss": 1.1254,
+      "step": 895
+    },
+    {
+      "epoch": 0.036422501011736136,
+      "grad_norm": 1.3433767557144165,
+      "learning_rate": 9.832748296510428e-05,
+      "loss": 1.0277,
+      "step": 900
+    },
+    {
+      "epoch": 0.036624848239579115,
+      "grad_norm": 1.473062515258789,
+      "learning_rate": 9.830683460664878e-05,
+      "loss": 1.0902,
+      "step": 905
+    },
+    {
+      "epoch": 0.036827195467422094,
+      "grad_norm": 1.5629346370697021,
+      "learning_rate": 9.828618624819327e-05,
+      "loss": 1.1125,
+      "step": 910
+    },
+    {
+      "epoch": 0.03702954269526507,
+      "grad_norm": 1.3655176162719727,
+      "learning_rate": 9.826553788973776e-05,
+      "loss": 1.0535,
+      "step": 915
+    },
+    {
+      "epoch": 0.03723188992310805,
+      "grad_norm": 1.3610424995422363,
+      "learning_rate": 9.824488953128227e-05,
+      "loss": 1.0586,
+      "step": 920
+    },
+    {
+      "epoch": 0.03743423715095103,
+      "grad_norm": 1.3422636985778809,
+      "learning_rate": 9.822424117282677e-05,
+      "loss": 1.0694,
+      "step": 925
+    },
+    {
+      "epoch": 0.03763658437879401,
+      "grad_norm": 1.4051406383514404,
+      "learning_rate": 9.820359281437126e-05,
+      "loss": 1.0393,
+      "step": 930
+    },
+    {
+      "epoch": 0.03783893160663699,
+      "grad_norm": 1.2342345714569092,
+      "learning_rate": 9.818294445591575e-05,
+      "loss": 1.0426,
+      "step": 935
+    },
+    {
+      "epoch": 0.03804127883447997,
+      "grad_norm": 1.2904876470565796,
+      "learning_rate": 9.816229609746026e-05,
+      "loss": 1.048,
+      "step": 940
+    },
+    {
+      "epoch": 0.03824362606232295,
+      "grad_norm": 1.5387827157974243,
+      "learning_rate": 9.814164773900475e-05,
+      "loss": 1.0752,
+      "step": 945
+    },
+    {
+      "epoch": 0.03844597329016593,
+      "grad_norm": 1.2083275318145752,
+      "learning_rate": 9.812099938054925e-05,
+      "loss": 1.0497,
+      "step": 950
+    },
+    {
+      "epoch": 0.038648320518008906,
+      "grad_norm": 1.3084746599197388,
+      "learning_rate": 9.810035102209376e-05,
+      "loss": 1.0597,
+      "step": 955
+    },
+    {
+      "epoch": 0.03885066774585188,
+      "grad_norm": 1.4155186414718628,
+      "learning_rate": 9.807970266363825e-05,
+      "loss": 1.0142,
+      "step": 960
+    },
+    {
+      "epoch": 0.03905301497369486,
+      "grad_norm": 1.272055983543396,
+      "learning_rate": 9.805905430518274e-05,
+      "loss": 1.0295,
+      "step": 965
+    },
+    {
+      "epoch": 0.03925536220153784,
+      "grad_norm": 1.25702702999115,
+      "learning_rate": 9.803840594672724e-05,
+      "loss": 1.0694,
+      "step": 970
+    },
+    {
+      "epoch": 0.039457709429380816,
+      "grad_norm": 1.3168611526489258,
+      "learning_rate": 9.801775758827175e-05,
+      "loss": 1.004,
+      "step": 975
+    },
+    {
+      "epoch": 0.039660056657223795,
+      "grad_norm": 1.3408727645874023,
+      "learning_rate": 9.799710922981624e-05,
+      "loss": 1.051,
+      "step": 980
+    },
+    {
+      "epoch": 0.039862403885066774,
+      "grad_norm": 1.5531495809555054,
+      "learning_rate": 9.797646087136073e-05,
+      "loss": 1.0327,
+      "step": 985
+    },
+    {
+      "epoch": 0.040064751112909754,
+      "grad_norm": 1.474417805671692,
+      "learning_rate": 9.795581251290523e-05,
+      "loss": 1.0806,
+      "step": 990
+    },
+    {
+      "epoch": 0.04026709834075273,
+      "grad_norm": 1.354904055595398,
+      "learning_rate": 9.793516415444972e-05,
+      "loss": 1.0267,
+      "step": 995
+    },
+    {
+      "epoch": 0.04046944556859571,
+      "grad_norm": 1.3857539892196655,
+      "learning_rate": 9.791451579599422e-05,
+      "loss": 1.0049,
+      "step": 1000
+    },
+    {
+      "epoch": 0.04067179279643869,
+      "grad_norm": 1.3299477100372314,
+      "learning_rate": 9.789386743753872e-05,
+      "loss": 1.0257,
+      "step": 1005
+    },
+    {
+      "epoch": 0.04087414002428167,
+      "grad_norm": 1.3018697500228882,
+      "learning_rate": 9.787321907908322e-05,
+      "loss": 1.0152,
+      "step": 1010
+    },
+    {
+      "epoch": 0.04107648725212465,
+      "grad_norm": 1.437354564666748,
+      "learning_rate": 9.785257072062771e-05,
+      "loss": 1.0539,
+      "step": 1015
+    },
+    {
+      "epoch": 0.04127883447996762,
+      "grad_norm": 1.4543030261993408,
+      "learning_rate": 9.783192236217221e-05,
+      "loss": 1.0573,
+      "step": 1020
+    },
+    {
+      "epoch": 0.0414811817078106,
+      "grad_norm": 1.41005539894104,
+      "learning_rate": 9.781127400371672e-05,
+      "loss": 1.0377,
+      "step": 1025
+    },
+    {
+      "epoch": 0.04168352893565358,
+      "grad_norm": 1.3793244361877441,
+      "learning_rate": 9.779062564526121e-05,
+      "loss": 1.0468,
+      "step": 1030
+    },
+    {
+      "epoch": 0.04188587616349656,
+      "grad_norm": 1.3902945518493652,
+      "learning_rate": 9.77699772868057e-05,
+      "loss": 1.061,
+      "step": 1035
+    },
+    {
+      "epoch": 0.04208822339133954,
+      "grad_norm": 1.3566033840179443,
+      "learning_rate": 9.77493289283502e-05,
+      "loss": 1.0839,
+      "step": 1040
+    },
+    {
+      "epoch": 0.04229057061918252,
+      "grad_norm": 1.3005142211914062,
+      "learning_rate": 9.77286805698947e-05,
+      "loss": 1.0235,
+      "step": 1045
+    },
+    {
+      "epoch": 0.042492917847025496,
+      "grad_norm": 1.4049321413040161,
+      "learning_rate": 9.77080322114392e-05,
+      "loss": 1.063,
+      "step": 1050
+    },
+    {
+      "epoch": 0.042695265074868476,
+      "grad_norm": 1.2821658849716187,
+      "learning_rate": 9.768738385298369e-05,
+      "loss": 1.0399,
+      "step": 1055
+    },
+    {
+      "epoch": 0.042897612302711455,
+      "grad_norm": 1.2726246118545532,
+      "learning_rate": 9.766673549452819e-05,
+      "loss": 1.0331,
+      "step": 1060
+    },
+    {
+      "epoch": 0.043099959530554434,
+      "grad_norm": 1.390950322151184,
+      "learning_rate": 9.764608713607268e-05,
+      "loss": 1.0974,
+      "step": 1065
+    },
+    {
+      "epoch": 0.04330230675839741,
+      "grad_norm": 1.430403709411621,
+      "learning_rate": 9.762543877761719e-05,
+      "loss": 0.979,
+      "step": 1070
+    },
+    {
+      "epoch": 0.043504653986240385,
+      "grad_norm": 1.2843495607376099,
+      "learning_rate": 9.760479041916169e-05,
+      "loss": 1.0334,
+      "step": 1075
+    },
+    {
+      "epoch": 0.043707001214083364,
+      "grad_norm": 1.4190033674240112,
+      "learning_rate": 9.758414206070617e-05,
+      "loss": 1.0625,
+      "step": 1080
+    },
+    {
+      "epoch": 0.043909348441926344,
+      "grad_norm": 1.4139447212219238,
+      "learning_rate": 9.756349370225067e-05,
+      "loss": 1.0092,
+      "step": 1085
+    },
+    {
+      "epoch": 0.04411169566976932,
+      "grad_norm": 1.532297968864441,
+      "learning_rate": 9.754284534379518e-05,
+      "loss": 1.0471,
+      "step": 1090
+    },
+    {
+      "epoch": 0.0443140428976123,
+      "grad_norm": 1.3429151773452759,
+      "learning_rate": 9.752219698533968e-05,
+      "loss": 1.0501,
+      "step": 1095
+    },
+    {
+      "epoch": 0.04451639012545528,
+      "grad_norm": 1.6312729120254517,
+      "learning_rate": 9.750154862688417e-05,
+      "loss": 0.9615,
+      "step": 1100
+    },
+    {
+      "epoch": 0.04471873735329826,
+      "grad_norm": 1.2032707929611206,
+      "learning_rate": 9.748090026842866e-05,
+      "loss": 1.0629,
+      "step": 1105
+    },
+    {
+      "epoch": 0.04492108458114124,
+      "grad_norm": 1.2905160188674927,
+      "learning_rate": 9.746025190997317e-05,
+      "loss": 1.0609,
+      "step": 1110
+    },
+    {
+      "epoch": 0.04512343180898422,
+      "grad_norm": 1.4337005615234375,
+      "learning_rate": 9.743960355151766e-05,
+      "loss": 1.0235,
+      "step": 1115
+    },
+    {
+      "epoch": 0.0453257790368272,
+      "grad_norm": 1.4135839939117432,
+      "learning_rate": 9.741895519306216e-05,
+      "loss": 1.0707,
+      "step": 1120
+    },
+    {
+      "epoch": 0.04552812626467018,
+      "grad_norm": 1.3754996061325073,
+      "learning_rate": 9.739830683460665e-05,
+      "loss": 1.0873,
+      "step": 1125
+    },
+    {
+      "epoch": 0.045730473492513156,
+      "grad_norm": 1.2317836284637451,
+      "learning_rate": 9.737765847615114e-05,
+      "loss": 1.0743,
+      "step": 1130
+    },
+    {
+      "epoch": 0.04593282072035613,
+      "grad_norm": 1.5067507028579712,
+      "learning_rate": 9.735701011769565e-05,
+      "loss": 1.0218,
+      "step": 1135
+    },
+    {
+      "epoch": 0.04613516794819911,
+      "grad_norm": 1.308156132698059,
+      "learning_rate": 9.733636175924015e-05,
+      "loss": 1.0798,
+      "step": 1140
+    },
+    {
+      "epoch": 0.046337515176042086,
+      "grad_norm": 1.4515304565429688,
+      "learning_rate": 9.731571340078465e-05,
+      "loss": 1.028,
+      "step": 1145
+    },
+    {
+      "epoch": 0.046539862403885066,
+      "grad_norm": 1.418377161026001,
+      "learning_rate": 9.729506504232913e-05,
+      "loss": 1.0614,
+      "step": 1150
+    },
+    {
+      "epoch": 0.046742209631728045,
+      "grad_norm": 1.3889667987823486,
+      "learning_rate": 9.727441668387363e-05,
+      "loss": 1.0,
+      "step": 1155
+    },
+    {
+      "epoch": 0.046944556859571024,
+      "grad_norm": 1.2457728385925293,
+      "learning_rate": 9.725376832541814e-05,
+      "loss": 1.0673,
+      "step": 1160
+    },
+    {
+      "epoch": 0.047146904087414,
+      "grad_norm": 1.36286199092865,
+      "learning_rate": 9.723311996696263e-05,
+      "loss": 0.9971,
+      "step": 1165
+    },
+    {
+      "epoch": 0.04734925131525698,
+      "grad_norm": 1.5431827306747437,
+      "learning_rate": 9.721247160850713e-05,
+      "loss": 1.05,
+      "step": 1170
+    },
+    {
+      "epoch": 0.04755159854309996,
+      "grad_norm": 1.2391327619552612,
+      "learning_rate": 9.719182325005162e-05,
+      "loss": 1.0493,
+      "step": 1175
+    },
+    {
+      "epoch": 0.04775394577094294,
+      "grad_norm": 1.3675509691238403,
+      "learning_rate": 9.717117489159611e-05,
+      "loss": 1.067,
+      "step": 1180
+    },
+    {
+      "epoch": 0.04795629299878592,
+      "grad_norm": 1.4171303510665894,
+      "learning_rate": 9.715052653314062e-05,
+      "loss": 1.0593,
+      "step": 1185
+    },
+    {
+      "epoch": 0.04815864022662889,
+      "grad_norm": 1.3410508632659912,
+      "learning_rate": 9.712987817468512e-05,
+      "loss": 0.9839,
+      "step": 1190
+    },
+    {
+      "epoch": 0.04836098745447187,
+      "grad_norm": 1.3725942373275757,
+      "learning_rate": 9.710922981622961e-05,
+      "loss": 1.0113,
+      "step": 1195
+    },
+    {
+      "epoch": 0.04856333468231485,
+      "grad_norm": 1.2524265050888062,
+      "learning_rate": 9.70885814577741e-05,
+      "loss": 1.0413,
+      "step": 1200
+    },
+    {
+      "epoch": 0.04876568191015783,
+      "grad_norm": 1.3189339637756348,
+      "learning_rate": 9.706793309931861e-05,
+      "loss": 1.0651,
+      "step": 1205
+    },
+    {
+      "epoch": 0.04896802913800081,
+      "grad_norm": 1.4519147872924805,
+      "learning_rate": 9.704728474086311e-05,
+      "loss": 1.0089,
+      "step": 1210
+    },
+    {
+      "epoch": 0.04917037636584379,
+      "grad_norm": 1.250978708267212,
+      "learning_rate": 9.70266363824076e-05,
+      "loss": 1.0253,
+      "step": 1215
+    },
+    {
+      "epoch": 0.04937272359368677,
+      "grad_norm": 1.262344479560852,
+      "learning_rate": 9.700598802395209e-05,
+      "loss": 1.0247,
+      "step": 1220
+    },
+    {
+      "epoch": 0.049575070821529746,
+      "grad_norm": 1.3184866905212402,
+      "learning_rate": 9.69853396654966e-05,
+      "loss": 1.0358,
+      "step": 1225
+    },
+    {
+      "epoch": 0.049777418049372725,
+      "grad_norm": 1.4247705936431885,
+      "learning_rate": 9.69646913070411e-05,
+      "loss": 1.0147,
+      "step": 1230
+    },
+    {
+      "epoch": 0.049979765277215704,
+      "grad_norm": 1.520044207572937,
+      "learning_rate": 9.694404294858559e-05,
+      "loss": 1.0491,
+      "step": 1235
+    },
+    {
+      "epoch": 0.05018211250505868,
+      "grad_norm": 1.2870755195617676,
+      "learning_rate": 9.69233945901301e-05,
+      "loss": 1.0032,
+      "step": 1240
+    },
+    {
+      "epoch": 0.05038445973290166,
+      "grad_norm": 1.2937649488449097,
+      "learning_rate": 9.690274623167459e-05,
+      "loss": 0.9758,
+      "step": 1245
+    },
+    {
+      "epoch": 0.050586806960744635,
+      "grad_norm": 1.4870978593826294,
+      "learning_rate": 9.688209787321908e-05,
+      "loss": 1.0998,
+      "step": 1250
+    },
+    {
+      "epoch": 0.050789154188587614,
+      "grad_norm": 1.2912105321884155,
+      "learning_rate": 9.686144951476358e-05,
+      "loss": 0.9734,
+      "step": 1255
+    },
+    {
+      "epoch": 0.05099150141643059,
+      "grad_norm": 1.3370693922042847,
+      "learning_rate": 9.684080115630808e-05,
+      "loss": 1.0799,
+      "step": 1260
+    },
+    {
+      "epoch": 0.05119384864427357,
+      "grad_norm": 1.1921980381011963,
+      "learning_rate": 9.682015279785257e-05,
+      "loss": 0.9679,
+      "step": 1265
+    },
+    {
+      "epoch": 0.05139619587211655,
+      "grad_norm": 1.3673900365829468,
+      "learning_rate": 9.679950443939707e-05,
+      "loss": 1.0979,
+      "step": 1270
+    },
+    {
+      "epoch": 0.05159854309995953,
+      "grad_norm": 1.3121528625488281,
+      "learning_rate": 9.677885608094157e-05,
+      "loss": 1.0508,
+      "step": 1275
+    },
+    {
+      "epoch": 0.05180089032780251,
+      "grad_norm": 1.416839361190796,
+      "learning_rate": 9.675820772248607e-05,
+      "loss": 1.0685,
+      "step": 1280
+    },
+    {
+      "epoch": 0.05200323755564549,
+      "grad_norm": 1.3744992017745972,
+      "learning_rate": 9.673755936403056e-05,
+      "loss": 1.0447,
+      "step": 1285
+    },
+    {
+      "epoch": 0.05220558478348847,
+      "grad_norm": 1.2653158903121948,
+      "learning_rate": 9.671691100557507e-05,
+      "loss": 1.0382,
+      "step": 1290
+    },
+    {
+      "epoch": 0.05240793201133145,
+      "grad_norm": 1.4127179384231567,
+      "learning_rate": 9.669626264711956e-05,
+      "loss": 1.0803,
+      "step": 1295
+    },
+    {
+      "epoch": 0.052610279239174426,
+      "grad_norm": 1.256456971168518,
+      "learning_rate": 9.667561428866405e-05,
+      "loss": 1.0679,
+      "step": 1300
+    },
+    {
+      "epoch": 0.052812626467017405,
+      "grad_norm": 1.3415658473968506,
+      "learning_rate": 9.665496593020855e-05,
+      "loss": 1.0205,
+      "step": 1305
+    },
+    {
+      "epoch": 0.05301497369486038,
+      "grad_norm": 1.4126713275909424,
+      "learning_rate": 9.663431757175306e-05,
+      "loss": 1.0116,
+      "step": 1310
+    },
+    {
+      "epoch": 0.05321732092270336,
+      "grad_norm": 1.3043524026870728,
+      "learning_rate": 9.661366921329755e-05,
+      "loss": 0.997,
+      "step": 1315
+    },
+    {
+      "epoch": 0.053419668150546336,
+      "grad_norm": 1.3901809453964233,
+      "learning_rate": 9.659302085484204e-05,
+      "loss": 1.0033,
+      "step": 1320
+    },
+    {
+      "epoch": 0.053622015378389315,
+      "grad_norm": 1.386953592300415,
+      "learning_rate": 9.657237249638654e-05,
+      "loss": 1.0546,
+      "step": 1325
+    },
+    {
+      "epoch": 0.053824362606232294,
+      "grad_norm": 1.2973097562789917,
+      "learning_rate": 9.655172413793105e-05,
+      "loss": 1.0523,
+      "step": 1330
+    },
+    {
+      "epoch": 0.05402670983407527,
+      "grad_norm": 1.3961467742919922,
+      "learning_rate": 9.653107577947554e-05,
+      "loss": 1.0213,
+      "step": 1335
+    },
+    {
+      "epoch": 0.05422905706191825,
+      "grad_norm": 1.2452396154403687,
+      "learning_rate": 9.651042742102003e-05,
+      "loss": 1.0588,
+      "step": 1340
+    },
+    {
+      "epoch": 0.05443140428976123,
+      "grad_norm": 1.25496244430542,
+      "learning_rate": 9.648977906256453e-05,
+      "loss": 1.0773,
+      "step": 1345
+    },
+    {
+      "epoch": 0.05463375151760421,
+      "grad_norm": 1.4122798442840576,
+      "learning_rate": 9.646913070410902e-05,
+      "loss": 1.0801,
+      "step": 1350
+    },
+    {
+      "epoch": 0.05483609874544719,
+      "grad_norm": 1.3670319318771362,
+      "learning_rate": 9.644848234565353e-05,
+      "loss": 1.0736,
+      "step": 1355
+    },
+    {
+      "epoch": 0.05503844597329017,
+      "grad_norm": 1.346893548965454,
+      "learning_rate": 9.642783398719803e-05,
+      "loss": 0.974,
+      "step": 1360
+    },
+    {
+      "epoch": 0.05524079320113314,
+      "grad_norm": 1.3234906196594238,
+      "learning_rate": 9.640718562874252e-05,
+      "loss": 1.0739,
+      "step": 1365
+    },
+    {
+      "epoch": 0.05544314042897612,
+      "grad_norm": 1.2294628620147705,
+      "learning_rate": 9.638653727028701e-05,
+      "loss": 1.0281,
+      "step": 1370
+    },
+    {
+      "epoch": 0.0556454876568191,
+      "grad_norm": 1.2931239604949951,
+      "learning_rate": 9.636588891183152e-05,
+      "loss": 1.0329,
+      "step": 1375
+    },
+    {
+      "epoch": 0.05584783488466208,
+      "grad_norm": 1.3048388957977295,
+      "learning_rate": 9.634524055337602e-05,
+      "loss": 1.0759,
+      "step": 1380
+    },
+    {
+      "epoch": 0.05605018211250506,
+      "grad_norm": 1.412604570388794,
+      "learning_rate": 9.632459219492051e-05,
+      "loss": 1.0506,
+      "step": 1385
+    },
+    {
+      "epoch": 0.05625252934034804,
+      "grad_norm": 1.3981273174285889,
+      "learning_rate": 9.6303943836465e-05,
+      "loss": 1.0566,
+      "step": 1390
+    },
+    {
+      "epoch": 0.056454876568191016,
+      "grad_norm": 1.2319412231445312,
+      "learning_rate": 9.62832954780095e-05,
+      "loss": 1.0059,
+      "step": 1395
+    },
+    {
+      "epoch": 0.056657223796033995,
+      "grad_norm": 1.2856113910675049,
+      "learning_rate": 9.6262647119554e-05,
+      "loss": 1.0594,
+      "step": 1400
+    },
+    {
+      "epoch": 0.056859571023876974,
+      "grad_norm": 1.5007933378219604,
+      "learning_rate": 9.62419987610985e-05,
+      "loss": 1.0191,
+      "step": 1405
+    },
+    {
+      "epoch": 0.057061918251719954,
+      "grad_norm": 1.3999196290969849,
+      "learning_rate": 9.622135040264299e-05,
+      "loss": 1.0383,
+      "step": 1410
+    },
+    {
+      "epoch": 0.05726426547956293,
+      "grad_norm": 1.3111178874969482,
+      "learning_rate": 9.62007020441875e-05,
+      "loss": 1.0648,
+      "step": 1415
+    },
+    {
+      "epoch": 0.05746661270740591,
+      "grad_norm": 1.2582710981369019,
+      "learning_rate": 9.618005368573198e-05,
+      "loss": 1.1045,
+      "step": 1420
+    },
+    {
+      "epoch": 0.057668959935248884,
+      "grad_norm": 1.445177674293518,
+      "learning_rate": 9.615940532727649e-05,
+      "loss": 1.0277,
+      "step": 1425
+    },
+    {
+      "epoch": 0.05787130716309186,
+      "grad_norm": 1.3125499486923218,
+      "learning_rate": 9.613875696882099e-05,
+      "loss": 1.0525,
+      "step": 1430
+    },
+    {
+      "epoch": 0.05807365439093484,
+      "grad_norm": 1.3522056341171265,
+      "learning_rate": 9.611810861036547e-05,
+      "loss": 1.0191,
+      "step": 1435
+    },
+    {
+      "epoch": 0.05827600161877782,
+      "grad_norm": 1.19707453250885,
+      "learning_rate": 9.609746025190997e-05,
+      "loss": 1.0182,
+      "step": 1440
+    },
+    {
+      "epoch": 0.0584783488466208,
+      "grad_norm": 1.3498811721801758,
+      "learning_rate": 9.607681189345448e-05,
+      "loss": 1.0606,
+      "step": 1445
+    },
+    {
+      "epoch": 0.05868069607446378,
+      "grad_norm": 1.1649166345596313,
+      "learning_rate": 9.605616353499898e-05,
+      "loss": 1.0361,
+      "step": 1450
+    },
+    {
+      "epoch": 0.05888304330230676,
+      "grad_norm": 1.2549792528152466,
+      "learning_rate": 9.603551517654347e-05,
+      "loss": 1.0384,
+      "step": 1455
+    },
+    {
+      "epoch": 0.05908539053014974,
+      "grad_norm": 1.2901394367218018,
+      "learning_rate": 9.601486681808796e-05,
+      "loss": 0.9723,
+      "step": 1460
+    },
+    {
+      "epoch": 0.05928773775799272,
+      "grad_norm": 1.189284324645996,
+      "learning_rate": 9.599421845963247e-05,
+      "loss": 1.0405,
+      "step": 1465
+    },
+    {
+      "epoch": 0.059490084985835696,
+      "grad_norm": 1.2396763563156128,
+      "learning_rate": 9.597357010117696e-05,
+      "loss": 1.0173,
+      "step": 1470
+    },
+    {
+      "epoch": 0.059692432213678676,
+      "grad_norm": 1.291982650756836,
+      "learning_rate": 9.595292174272146e-05,
+      "loss": 1.07,
+      "step": 1475
+    },
+    {
+      "epoch": 0.05989477944152165,
+      "grad_norm": 1.4714529514312744,
+      "learning_rate": 9.593227338426595e-05,
+      "loss": 1.0584,
+      "step": 1480
+    },
+    {
+      "epoch": 0.06009712666936463,
+      "grad_norm": 1.2746899127960205,
+      "learning_rate": 9.591162502581044e-05,
+      "loss": 1.0436,
+      "step": 1485
+    },
+    {
+      "epoch": 0.060299473897207606,
+      "grad_norm": 1.3938498497009277,
+      "learning_rate": 9.589097666735495e-05,
+      "loss": 1.0457,
+      "step": 1490
+    },
+    {
+      "epoch": 0.060501821125050585,
+      "grad_norm": 1.505146861076355,
+      "learning_rate": 9.587032830889945e-05,
+      "loss": 1.0649,
+      "step": 1495
+    },
+    {
+      "epoch": 0.060704168352893564,
+      "grad_norm": 1.349221110343933,
+      "learning_rate": 9.584967995044395e-05,
+      "loss": 1.0201,
+      "step": 1500
+    },
+    {
+      "epoch": 0.060906515580736544,
+      "grad_norm": 1.2473959922790527,
+      "learning_rate": 9.582903159198845e-05,
+      "loss": 1.0345,
+      "step": 1505
+    },
+    {
+      "epoch": 0.06110886280857952,
+      "grad_norm": 1.2991083860397339,
+      "learning_rate": 9.580838323353294e-05,
+      "loss": 1.0163,
+      "step": 1510
+    },
+    {
+      "epoch": 0.0613112100364225,
+      "grad_norm": 1.4911009073257446,
+      "learning_rate": 9.578773487507744e-05,
+      "loss": 1.0206,
+      "step": 1515
+    },
+    {
+      "epoch": 0.06151355726426548,
+      "grad_norm": 1.3233214616775513,
+      "learning_rate": 9.576708651662193e-05,
+      "loss": 1.0825,
+      "step": 1520
+    },
+    {
+      "epoch": 0.06171590449210846,
+      "grad_norm": 1.4564628601074219,
+      "learning_rate": 9.574643815816643e-05,
+      "loss": 1.0681,
+      "step": 1525
+    },
+    {
+      "epoch": 0.06191825171995144,
+      "grad_norm": 1.3113048076629639,
+      "learning_rate": 9.572578979971092e-05,
+      "loss": 1.0998,
+      "step": 1530
+    },
+    {
+      "epoch": 0.06212059894779442,
+      "grad_norm": 1.437929630279541,
+      "learning_rate": 9.570514144125542e-05,
+      "loss": 0.9953,
+      "step": 1535
+    },
+    {
+      "epoch": 0.06232294617563739,
+      "grad_norm": 1.33054518699646,
+      "learning_rate": 9.568449308279992e-05,
+      "loss": 1.0731,
+      "step": 1540
+    },
+    {
+      "epoch": 0.06252529340348037,
+      "grad_norm": 1.2281547784805298,
+      "learning_rate": 9.566384472434442e-05,
+      "loss": 1.046,
+      "step": 1545
+    },
+    {
+      "epoch": 0.06272764063132336,
+      "grad_norm": 1.413252830505371,
+      "learning_rate": 9.564319636588891e-05,
+      "loss": 1.0632,
+      "step": 1550
+    },
+    {
+      "epoch": 0.06292998785916633,
+      "grad_norm": 1.4634878635406494,
+      "learning_rate": 9.56225480074334e-05,
+      "loss": 0.9653,
+      "step": 1555
+    },
+    {
+      "epoch": 0.06313233508700931,
+      "grad_norm": 1.1909383535385132,
+      "learning_rate": 9.560189964897791e-05,
+      "loss": 0.9952,
+      "step": 1560
+    },
+    {
+      "epoch": 0.06333468231485229,
+      "grad_norm": 1.3073731660842896,
+      "learning_rate": 9.558125129052241e-05,
+      "loss": 1.0672,
+      "step": 1565
+    },
+    {
+      "epoch": 0.06353702954269526,
+      "grad_norm": 1.221425175666809,
+      "learning_rate": 9.55606029320669e-05,
+      "loss": 1.0212,
+      "step": 1570
+    },
+    {
+      "epoch": 0.06373937677053824,
+      "grad_norm": 1.3182533979415894,
+      "learning_rate": 9.553995457361141e-05,
+      "loss": 1.0355,
+      "step": 1575
+    },
+    {
+      "epoch": 0.06394172399838122,
+      "grad_norm": 1.2224425077438354,
+      "learning_rate": 9.55193062151559e-05,
+      "loss": 0.9842,
+      "step": 1580
+    },
+    {
+      "epoch": 0.0641440712262242,
+      "grad_norm": 1.3230282068252563,
+      "learning_rate": 9.54986578567004e-05,
+      "loss": 0.9847,
+      "step": 1585
+    },
+    {
+      "epoch": 0.06434641845406718,
+      "grad_norm": 1.6240323781967163,
+      "learning_rate": 9.547800949824489e-05,
+      "loss": 1.0517,
+      "step": 1590
+    },
+    {
+      "epoch": 0.06454876568191016,
+      "grad_norm": 1.2214101552963257,
+      "learning_rate": 9.54573611397894e-05,
+      "loss": 1.0031,
+      "step": 1595
+    },
+    {
+      "epoch": 0.06475111290975313,
+      "grad_norm": 1.3238086700439453,
+      "learning_rate": 9.543671278133389e-05,
+      "loss": 0.9992,
+      "step": 1600
+    },
+    {
+      "epoch": 0.06495346013759612,
+      "grad_norm": 1.3082062005996704,
+      "learning_rate": 9.541606442287838e-05,
+      "loss": 1.0782,
+      "step": 1605
+    },
+    {
+      "epoch": 0.06515580736543909,
+      "grad_norm": 1.2175467014312744,
+      "learning_rate": 9.539541606442288e-05,
+      "loss": 1.0405,
+      "step": 1610
+    },
+    {
+      "epoch": 0.06535815459328208,
+      "grad_norm": 1.2372517585754395,
+      "learning_rate": 9.537476770596739e-05,
+      "loss": 1.0847,
+      "step": 1615
+    },
+    {
+      "epoch": 0.06556050182112505,
+      "grad_norm": 1.3659371137619019,
+      "learning_rate": 9.535411934751188e-05,
+      "loss": 1.0472,
+      "step": 1620
+    },
+    {
+      "epoch": 0.06576284904896802,
+      "grad_norm": 1.4121381044387817,
+      "learning_rate": 9.533347098905637e-05,
+      "loss": 1.0403,
+      "step": 1625
+    },
+    {
+      "epoch": 0.06596519627681101,
+      "grad_norm": 1.247285008430481,
+      "learning_rate": 9.531282263060087e-05,
+      "loss": 1.0667,
+      "step": 1630
+    },
+    {
+      "epoch": 0.06616754350465398,
+      "grad_norm": 1.2073017358779907,
+      "learning_rate": 9.529217427214537e-05,
+      "loss": 1.0118,
+      "step": 1635
+    },
+    {
+      "epoch": 0.06636989073249697,
+      "grad_norm": 1.5580484867095947,
+      "learning_rate": 9.527152591368987e-05,
+      "loss": 1.1244,
+      "step": 1640
+    },
+    {
+      "epoch": 0.06657223796033994,
+      "grad_norm": 1.1624681949615479,
+      "learning_rate": 9.525087755523437e-05,
+      "loss": 1.0397,
+      "step": 1645
+    },
+    {
+      "epoch": 0.06677458518818292,
+      "grad_norm": 1.4433684349060059,
+      "learning_rate": 9.523022919677886e-05,
+      "loss": 1.0386,
+      "step": 1650
+    },
+    {
+      "epoch": 0.0669769324160259,
+      "grad_norm": 1.217288613319397,
+      "learning_rate": 9.520958083832335e-05,
+      "loss": 1.0078,
+      "step": 1655
+    },
+    {
+      "epoch": 0.06717927964386888,
+      "grad_norm": 1.3176137208938599,
+      "learning_rate": 9.518893247986785e-05,
+      "loss": 1.0367,
+      "step": 1660
+    },
+    {
+      "epoch": 0.06738162687171186,
+      "grad_norm": 1.3557639122009277,
+      "learning_rate": 9.516828412141236e-05,
+      "loss": 1.0161,
+      "step": 1665
+    },
+    {
+      "epoch": 0.06758397409955484,
+      "grad_norm": 1.1859376430511475,
+      "learning_rate": 9.514763576295685e-05,
+      "loss": 1.0092,
+      "step": 1670
+    },
+    {
+      "epoch": 0.06778632132739781,
+      "grad_norm": 1.2193354368209839,
+      "learning_rate": 9.512698740450134e-05,
+      "loss": 1.0393,
+      "step": 1675
+    },
+    {
+      "epoch": 0.0679886685552408,
+      "grad_norm": 1.1776807308197021,
+      "learning_rate": 9.510633904604584e-05,
+      "loss": 1.0088,
+      "step": 1680
+    },
+    {
+      "epoch": 0.06819101578308377,
+      "grad_norm": 1.2251741886138916,
+      "learning_rate": 9.508569068759035e-05,
+      "loss": 1.0542,
+      "step": 1685
+    },
+    {
+      "epoch": 0.06839336301092674,
+      "grad_norm": 1.4066046476364136,
+      "learning_rate": 9.506504232913484e-05,
+      "loss": 1.0778,
+      "step": 1690
+    },
+    {
+      "epoch": 0.06859571023876973,
+      "grad_norm": 1.1775314807891846,
+      "learning_rate": 9.504439397067933e-05,
+      "loss": 0.9923,
+      "step": 1695
+    },
+    {
+      "epoch": 0.0687980574666127,
+      "grad_norm": 1.2167171239852905,
+      "learning_rate": 9.502374561222383e-05,
+      "loss": 0.9782,
+      "step": 1700
+    },
+    {
+      "epoch": 0.06900040469445569,
+      "grad_norm": 1.356432557106018,
+      "learning_rate": 9.500309725376832e-05,
+      "loss": 1.0112,
+      "step": 1705
+    },
+    {
+      "epoch": 0.06920275192229866,
+      "grad_norm": 1.240261435508728,
+      "learning_rate": 9.498244889531283e-05,
+      "loss": 1.0609,
+      "step": 1710
+    },
+    {
+      "epoch": 0.06940509915014165,
+      "grad_norm": 1.2335928678512573,
+      "learning_rate": 9.496180053685733e-05,
+      "loss": 1.0303,
+      "step": 1715
+    },
+    {
+      "epoch": 0.06960744637798462,
+      "grad_norm": 1.190518856048584,
+      "learning_rate": 9.494115217840182e-05,
+      "loss": 1.0079,
+      "step": 1720
+    },
+    {
+      "epoch": 0.0698097936058276,
+      "grad_norm": 1.315698504447937,
+      "learning_rate": 9.492050381994631e-05,
+      "loss": 0.9892,
+      "step": 1725
+    },
+    {
+      "epoch": 0.07001214083367058,
+      "grad_norm": 1.4281184673309326,
+      "learning_rate": 9.489985546149082e-05,
+      "loss": 1.0088,
+      "step": 1730
+    },
+    {
+      "epoch": 0.07021448806151356,
+      "grad_norm": 1.3519691228866577,
+      "learning_rate": 9.487920710303532e-05,
+      "loss": 1.0213,
+      "step": 1735
+    },
+    {
+      "epoch": 0.07041683528935654,
+      "grad_norm": 1.1940042972564697,
+      "learning_rate": 9.485855874457981e-05,
+      "loss": 1.0422,
+      "step": 1740
+    },
+    {
+      "epoch": 0.07061918251719951,
+      "grad_norm": 1.2125798463821411,
+      "learning_rate": 9.48379103861243e-05,
+      "loss": 1.0255,
+      "step": 1745
+    },
+    {
+      "epoch": 0.0708215297450425,
+      "grad_norm": 1.251612901687622,
+      "learning_rate": 9.48172620276688e-05,
+      "loss": 1.0212,
+      "step": 1750
+    },
+    {
+      "epoch": 0.07102387697288547,
+      "grad_norm": 1.4712817668914795,
+      "learning_rate": 9.47966136692133e-05,
+      "loss": 1.0403,
+      "step": 1755
+    },
+    {
+      "epoch": 0.07122622420072845,
+      "grad_norm": 1.2654145956039429,
+      "learning_rate": 9.47759653107578e-05,
+      "loss": 1.0262,
+      "step": 1760
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 1.3461631536483765,
+      "learning_rate": 9.475531695230229e-05,
+      "loss": 1.0176,
+      "step": 1765
+    },
+    {
+      "epoch": 0.07163091865641441,
+      "grad_norm": 1.3915088176727295,
+      "learning_rate": 9.47346685938468e-05,
+      "loss": 1.0428,
+      "step": 1770
+    },
+    {
+      "epoch": 0.07183326588425738,
+      "grad_norm": 1.377109169960022,
+      "learning_rate": 9.471402023539129e-05,
+      "loss": 1.0332,
+      "step": 1775
+    },
+    {
+      "epoch": 0.07203561311210037,
+      "grad_norm": 1.136925220489502,
+      "learning_rate": 9.469337187693579e-05,
+      "loss": 1.0588,
+      "step": 1780
+    },
+    {
+      "epoch": 0.07223796033994334,
+      "grad_norm": 1.3466987609863281,
+      "learning_rate": 9.46727235184803e-05,
+      "loss": 1.0595,
+      "step": 1785
+    },
+    {
+      "epoch": 0.07244030756778633,
+      "grad_norm": 1.435737133026123,
+      "learning_rate": 9.465207516002478e-05,
+      "loss": 1.0148,
+      "step": 1790
+    },
+    {
+      "epoch": 0.0726426547956293,
+      "grad_norm": 1.4260509014129639,
+      "learning_rate": 9.463142680156927e-05,
+      "loss": 1.1411,
+      "step": 1795
+    },
+    {
+      "epoch": 0.07284500202347227,
+      "grad_norm": 1.1842601299285889,
+      "learning_rate": 9.461077844311378e-05,
+      "loss": 1.043,
+      "step": 1800
+    },
+    {
+      "epoch": 0.07304734925131526,
+      "grad_norm": 1.2776232957839966,
+      "learning_rate": 9.459013008465828e-05,
+      "loss": 1.0548,
+      "step": 1805
+    },
+    {
+      "epoch": 0.07324969647915823,
+      "grad_norm": 1.310470700263977,
+      "learning_rate": 9.456948172620277e-05,
+      "loss": 1.0171,
+      "step": 1810
+    },
+    {
+      "epoch": 0.07345204370700122,
+      "grad_norm": 1.2996270656585693,
+      "learning_rate": 9.454883336774726e-05,
+      "loss": 1.0352,
+      "step": 1815
+    },
+    {
+      "epoch": 0.07365439093484419,
+      "grad_norm": 1.381764531135559,
+      "learning_rate": 9.452818500929177e-05,
+      "loss": 1.0938,
+      "step": 1820
+    },
+    {
+      "epoch": 0.07385673816268717,
+      "grad_norm": 1.3565207719802856,
+      "learning_rate": 9.450753665083626e-05,
+      "loss": 1.0069,
+      "step": 1825
+    },
+    {
+      "epoch": 0.07405908539053015,
+      "grad_norm": 1.2954041957855225,
+      "learning_rate": 9.448688829238076e-05,
+      "loss": 0.9616,
+      "step": 1830
+    },
+    {
+      "epoch": 0.07426143261837313,
+      "grad_norm": 1.3097357749938965,
+      "learning_rate": 9.446623993392527e-05,
+      "loss": 1.0161,
+      "step": 1835
+    },
+    {
+      "epoch": 0.0744637798462161,
+      "grad_norm": 1.1322047710418701,
+      "learning_rate": 9.444559157546974e-05,
+      "loss": 0.9654,
+      "step": 1840
+    },
+    {
+      "epoch": 0.07466612707405909,
+      "grad_norm": 1.252774715423584,
+      "learning_rate": 9.442494321701425e-05,
+      "loss": 1.0625,
+      "step": 1845
+    },
+    {
+      "epoch": 0.07486847430190206,
+      "grad_norm": 1.0868663787841797,
+      "learning_rate": 9.440429485855875e-05,
+      "loss": 1.0023,
+      "step": 1850
+    },
+    {
+      "epoch": 0.07507082152974505,
+      "grad_norm": 1.290256381034851,
+      "learning_rate": 9.438364650010326e-05,
+      "loss": 1.018,
+      "step": 1855
+    },
+    {
+      "epoch": 0.07527316875758802,
+      "grad_norm": 1.159004807472229,
+      "learning_rate": 9.436299814164775e-05,
+      "loss": 1.0381,
+      "step": 1860
+    },
+    {
+      "epoch": 0.075475515985431,
+      "grad_norm": 1.345625400543213,
+      "learning_rate": 9.434234978319224e-05,
+      "loss": 1.0238,
+      "step": 1865
+    },
+    {
+      "epoch": 0.07567786321327398,
+      "grad_norm": 1.199846625328064,
+      "learning_rate": 9.432170142473674e-05,
+      "loss": 1.0145,
+      "step": 1870
+    },
+    {
+      "epoch": 0.07588021044111695,
+      "grad_norm": 1.122860074043274,
+      "learning_rate": 9.430105306628123e-05,
+      "loss": 0.9912,
+      "step": 1875
+    },
+    {
+      "epoch": 0.07608255766895994,
+      "grad_norm": 1.387333631515503,
+      "learning_rate": 9.428040470782574e-05,
+      "loss": 1.0813,
+      "step": 1880
+    },
+    {
+      "epoch": 0.07628490489680291,
+      "grad_norm": 1.2576044797897339,
+      "learning_rate": 9.425975634937023e-05,
+      "loss": 1.1047,
+      "step": 1885
+    },
+    {
+      "epoch": 0.0764872521246459,
+      "grad_norm": 1.2864686250686646,
+      "learning_rate": 9.423910799091473e-05,
+      "loss": 1.0696,
+      "step": 1890
+    },
+    {
+      "epoch": 0.07668959935248887,
+      "grad_norm": 1.2306206226348877,
+      "learning_rate": 9.421845963245922e-05,
+      "loss": 0.9927,
+      "step": 1895
+    },
+    {
+      "epoch": 0.07689194658033185,
+      "grad_norm": 1.1592239141464233,
+      "learning_rate": 9.419781127400372e-05,
+      "loss": 1.0237,
+      "step": 1900
+    },
+    {
+      "epoch": 0.07709429380817483,
+      "grad_norm": 1.3854925632476807,
+      "learning_rate": 9.417716291554823e-05,
+      "loss": 1.0869,
+      "step": 1905
+    },
+    {
+      "epoch": 0.07729664103601781,
+      "grad_norm": 1.3255068063735962,
+      "learning_rate": 9.41565145570927e-05,
+      "loss": 1.0857,
+      "step": 1910
+    },
+    {
+      "epoch": 0.07749898826386079,
+      "grad_norm": 1.4847171306610107,
+      "learning_rate": 9.413586619863721e-05,
+      "loss": 1.0406,
+      "step": 1915
+    },
+    {
+      "epoch": 0.07770133549170376,
+      "grad_norm": 1.382543921470642,
+      "learning_rate": 9.411521784018171e-05,
+      "loss": 1.0376,
+      "step": 1920
+    },
+    {
+      "epoch": 0.07790368271954674,
+      "grad_norm": 1.2871674299240112,
+      "learning_rate": 9.40945694817262e-05,
+      "loss": 1.0433,
+      "step": 1925
+    },
+    {
+      "epoch": 0.07810602994738972,
+      "grad_norm": 1.2749968767166138,
+      "learning_rate": 9.407392112327071e-05,
+      "loss": 1.0692,
+      "step": 1930
+    },
+    {
+      "epoch": 0.0783083771752327,
+      "grad_norm": 1.2841780185699463,
+      "learning_rate": 9.40532727648152e-05,
+      "loss": 1.0352,
+      "step": 1935
+    },
+    {
+      "epoch": 0.07851072440307567,
+      "grad_norm": 1.3862274885177612,
+      "learning_rate": 9.40326244063597e-05,
+      "loss": 1.0197,
+      "step": 1940
+    },
+    {
+      "epoch": 0.07871307163091866,
+      "grad_norm": 1.2190536260604858,
+      "learning_rate": 9.40119760479042e-05,
+      "loss": 1.0357,
+      "step": 1945
+    },
+    {
+      "epoch": 0.07891541885876163,
+      "grad_norm": 1.2616676092147827,
+      "learning_rate": 9.39913276894487e-05,
+      "loss": 1.03,
+      "step": 1950
+    },
+    {
+      "epoch": 0.07911776608660462,
+      "grad_norm": 1.3833273649215698,
+      "learning_rate": 9.397067933099319e-05,
+      "loss": 1.0219,
+      "step": 1955
+    },
+    {
+      "epoch": 0.07932011331444759,
+      "grad_norm": 1.2849769592285156,
+      "learning_rate": 9.395003097253768e-05,
+      "loss": 1.002,
+      "step": 1960
+    },
+    {
+      "epoch": 0.07952246054229058,
+      "grad_norm": 1.333940863609314,
+      "learning_rate": 9.392938261408218e-05,
+      "loss": 0.9834,
+      "step": 1965
+    },
+    {
+      "epoch": 0.07972480777013355,
+      "grad_norm": 1.343896508216858,
+      "learning_rate": 9.390873425562669e-05,
+      "loss": 1.0425,
+      "step": 1970
+    },
+    {
+      "epoch": 0.07992715499797652,
+      "grad_norm": 1.1878740787506104,
+      "learning_rate": 9.388808589717118e-05,
+      "loss": 1.0398,
+      "step": 1975
+    },
+    {
+      "epoch": 0.08012950222581951,
+      "grad_norm": 1.4231536388397217,
+      "learning_rate": 9.386743753871567e-05,
+      "loss": 0.982,
+      "step": 1980
+    },
+    {
+      "epoch": 0.08033184945366248,
+      "grad_norm": 1.3932554721832275,
+      "learning_rate": 9.384678918026017e-05,
+      "loss": 1.0108,
+      "step": 1985
+    },
+    {
+      "epoch": 0.08053419668150547,
+      "grad_norm": 1.2101713418960571,
+      "learning_rate": 9.382614082180468e-05,
+      "loss": 1.002,
+      "step": 1990
+    },
+    {
+      "epoch": 0.08073654390934844,
+      "grad_norm": 1.2729885578155518,
+      "learning_rate": 9.380549246334917e-05,
+      "loss": 1.0472,
+      "step": 1995
+    },
+    {
+      "epoch": 0.08093889113719142,
+      "grad_norm": 1.488101601600647,
+      "learning_rate": 9.378484410489367e-05,
+      "loss": 1.0282,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0811412383650344,
+      "grad_norm": 1.326236605644226,
+      "learning_rate": 9.376419574643816e-05,
+      "loss": 1.029,
+      "step": 2005
+    },
+    {
+      "epoch": 0.08134358559287738,
+      "grad_norm": 1.3277944326400757,
+      "learning_rate": 9.374354738798265e-05,
+      "loss": 1.041,
+      "step": 2010
+    },
+    {
+      "epoch": 0.08154593282072035,
+      "grad_norm": 1.202103614807129,
+      "learning_rate": 9.372289902952716e-05,
+      "loss": 1.0915,
+      "step": 2015
+    },
+    {
+      "epoch": 0.08174828004856334,
+      "grad_norm": 1.2832789421081543,
+      "learning_rate": 9.370225067107166e-05,
+      "loss": 0.9902,
+      "step": 2020
+    },
+    {
+      "epoch": 0.08195062727640631,
+      "grad_norm": 1.2172725200653076,
+      "learning_rate": 9.368160231261615e-05,
+      "loss": 1.0475,
+      "step": 2025
+    },
+    {
+      "epoch": 0.0821529745042493,
+      "grad_norm": 1.1746991872787476,
+      "learning_rate": 9.366095395416064e-05,
+      "loss": 1.0354,
+      "step": 2030
+    },
+    {
+      "epoch": 0.08235532173209227,
+      "grad_norm": 1.2759616374969482,
+      "learning_rate": 9.364030559570515e-05,
+      "loss": 0.9939,
+      "step": 2035
+    },
+    {
+      "epoch": 0.08255766895993524,
+      "grad_norm": 1.2615772485733032,
+      "learning_rate": 9.361965723724965e-05,
+      "loss": 1.0286,
+      "step": 2040
+    },
+    {
+      "epoch": 0.08276001618777823,
+      "grad_norm": 1.3714519739151,
+      "learning_rate": 9.359900887879414e-05,
+      "loss": 1.0196,
+      "step": 2045
+    },
+    {
+      "epoch": 0.0829623634156212,
+      "grad_norm": 1.1940999031066895,
+      "learning_rate": 9.357836052033864e-05,
+      "loss": 1.0553,
+      "step": 2050
+    },
+    {
+      "epoch": 0.08316471064346419,
+      "grad_norm": 1.2421550750732422,
+      "learning_rate": 9.355771216188313e-05,
+      "loss": 1.0067,
+      "step": 2055
+    },
+    {
+      "epoch": 0.08336705787130716,
+      "grad_norm": 1.223291277885437,
+      "learning_rate": 9.353706380342762e-05,
+      "loss": 1.0034,
+      "step": 2060
+    },
+    {
+      "epoch": 0.08356940509915015,
+      "grad_norm": 1.191751480102539,
+      "learning_rate": 9.351641544497213e-05,
+      "loss": 1.0465,
+      "step": 2065
+    },
+    {
+      "epoch": 0.08377175232699312,
+      "grad_norm": 1.1390283107757568,
+      "learning_rate": 9.349576708651663e-05,
+      "loss": 1.0437,
+      "step": 2070
+    },
+    {
+      "epoch": 0.0839740995548361,
+      "grad_norm": 1.1788053512573242,
+      "learning_rate": 9.347511872806112e-05,
+      "loss": 0.9663,
+      "step": 2075
+    },
+    {
+      "epoch": 0.08417644678267908,
+      "grad_norm": 1.1897859573364258,
+      "learning_rate": 9.345447036960561e-05,
+      "loss": 1.0421,
+      "step": 2080
+    },
+    {
+      "epoch": 0.08437879401052206,
+      "grad_norm": 1.1853989362716675,
+      "learning_rate": 9.343382201115012e-05,
+      "loss": 1.0293,
+      "step": 2085
+    },
+    {
+      "epoch": 0.08458114123836503,
+      "grad_norm": 1.0884404182434082,
+      "learning_rate": 9.341317365269462e-05,
+      "loss": 0.9338,
+      "step": 2090
+    },
+    {
+      "epoch": 0.084783488466208,
+      "grad_norm": 1.1858789920806885,
+      "learning_rate": 9.339252529423911e-05,
+      "loss": 1.0303,
+      "step": 2095
+    },
+    {
+      "epoch": 0.08498583569405099,
+      "grad_norm": 1.2867400646209717,
+      "learning_rate": 9.33718769357836e-05,
+      "loss": 1.0058,
+      "step": 2100
+    },
+    {
+      "epoch": 0.08518818292189397,
+      "grad_norm": 1.1759954690933228,
+      "learning_rate": 9.335122857732811e-05,
+      "loss": 1.0497,
+      "step": 2105
+    },
+    {
+      "epoch": 0.08539053014973695,
+      "grad_norm": 1.2815693616867065,
+      "learning_rate": 9.33305802188726e-05,
+      "loss": 1.0262,
+      "step": 2110
+    },
+    {
+      "epoch": 0.08559287737757992,
+      "grad_norm": 1.2004886865615845,
+      "learning_rate": 9.33099318604171e-05,
+      "loss": 1.0051,
+      "step": 2115
+    },
+    {
+      "epoch": 0.08579522460542291,
+      "grad_norm": 1.178978681564331,
+      "learning_rate": 9.32892835019616e-05,
+      "loss": 1.0415,
+      "step": 2120
+    },
+    {
+      "epoch": 0.08599757183326588,
+      "grad_norm": 1.2252072095870972,
+      "learning_rate": 9.32686351435061e-05,
+      "loss": 0.999,
+      "step": 2125
+    },
+    {
+      "epoch": 0.08619991906110887,
+      "grad_norm": 1.2782108783721924,
+      "learning_rate": 9.324798678505059e-05,
+      "loss": 1.0602,
+      "step": 2130
+    },
+    {
+      "epoch": 0.08640226628895184,
+      "grad_norm": 1.1181281805038452,
+      "learning_rate": 9.322733842659509e-05,
+      "loss": 0.9785,
+      "step": 2135
+    },
+    {
+      "epoch": 0.08660461351679483,
+      "grad_norm": 1.2799198627471924,
+      "learning_rate": 9.32066900681396e-05,
+      "loss": 0.9867,
+      "step": 2140
+    },
+    {
+      "epoch": 0.0868069607446378,
+      "grad_norm": 1.1740738153457642,
+      "learning_rate": 9.318604170968409e-05,
+      "loss": 0.9925,
+      "step": 2145
+    },
+    {
+      "epoch": 0.08700930797248077,
+      "grad_norm": 1.1738531589508057,
+      "learning_rate": 9.316539335122858e-05,
+      "loss": 1.0328,
+      "step": 2150
+    },
+    {
+      "epoch": 0.08721165520032376,
+      "grad_norm": 1.2151775360107422,
+      "learning_rate": 9.314474499277308e-05,
+      "loss": 1.0374,
+      "step": 2155
+    },
+    {
+      "epoch": 0.08741400242816673,
+      "grad_norm": 1.2134398221969604,
+      "learning_rate": 9.312409663431758e-05,
+      "loss": 1.0062,
+      "step": 2160
+    },
+    {
+      "epoch": 0.08761634965600971,
+      "grad_norm": 1.2791131734848022,
+      "learning_rate": 9.310344827586207e-05,
+      "loss": 1.1111,
+      "step": 2165
+    },
+    {
+      "epoch": 0.08781869688385269,
+      "grad_norm": 1.198521614074707,
+      "learning_rate": 9.308279991740657e-05,
+      "loss": 1.0132,
+      "step": 2170
+    },
+    {
+      "epoch": 0.08802104411169567,
+      "grad_norm": 1.2800092697143555,
+      "learning_rate": 9.306215155895107e-05,
+      "loss": 1.051,
+      "step": 2175
+    },
+    {
+      "epoch": 0.08822339133953865,
+      "grad_norm": 1.3057935237884521,
+      "learning_rate": 9.304150320049556e-05,
+      "loss": 1.1417,
+      "step": 2180
+    },
+    {
+      "epoch": 0.08842573856738163,
+      "grad_norm": 1.2939696311950684,
+      "learning_rate": 9.302085484204006e-05,
+      "loss": 0.9713,
+      "step": 2185
+    },
+    {
+      "epoch": 0.0886280857952246,
+      "grad_norm": 1.4257221221923828,
+      "learning_rate": 9.300020648358457e-05,
+      "loss": 1.0949,
+      "step": 2190
+    },
+    {
+      "epoch": 0.08883043302306759,
+      "grad_norm": 1.3082605600357056,
+      "learning_rate": 9.297955812512905e-05,
+      "loss": 1.0281,
+      "step": 2195
+    },
+    {
+      "epoch": 0.08903278025091056,
+      "grad_norm": 1.3272719383239746,
+      "learning_rate": 9.295890976667355e-05,
+      "loss": 1.0791,
+      "step": 2200
+    },
+    {
+      "epoch": 0.08923512747875353,
+      "grad_norm": 1.2881433963775635,
+      "learning_rate": 9.293826140821805e-05,
+      "loss": 1.0072,
+      "step": 2205
+    },
+    {
+      "epoch": 0.08943747470659652,
+      "grad_norm": 1.3805770874023438,
+      "learning_rate": 9.291761304976256e-05,
+      "loss": 1.0069,
+      "step": 2210
+    },
+    {
+      "epoch": 0.08963982193443949,
+      "grad_norm": 1.2969526052474976,
+      "learning_rate": 9.289696469130705e-05,
+      "loss": 1.0512,
+      "step": 2215
+    },
+    {
+      "epoch": 0.08984216916228248,
+      "grad_norm": 1.3401156663894653,
+      "learning_rate": 9.287631633285154e-05,
+      "loss": 1.0234,
+      "step": 2220
+    },
+    {
+      "epoch": 0.09004451639012545,
+      "grad_norm": 1.2723054885864258,
+      "learning_rate": 9.285566797439604e-05,
+      "loss": 1.0485,
+      "step": 2225
+    },
+    {
+      "epoch": 0.09024686361796844,
+      "grad_norm": 1.2819281816482544,
+      "learning_rate": 9.283501961594053e-05,
+      "loss": 1.0066,
+      "step": 2230
+    },
+    {
+      "epoch": 0.09044921084581141,
+      "grad_norm": 1.3629616498947144,
+      "learning_rate": 9.281437125748504e-05,
+      "loss": 1.0007,
+      "step": 2235
+    },
+    {
+      "epoch": 0.0906515580736544,
+      "grad_norm": 1.213684320449829,
+      "learning_rate": 9.279372289902953e-05,
+      "loss": 0.9829,
+      "step": 2240
+    },
+    {
+      "epoch": 0.09085390530149737,
+      "grad_norm": 1.2831264734268188,
+      "learning_rate": 9.277307454057403e-05,
+      "loss": 1.0386,
+      "step": 2245
+    },
+    {
+      "epoch": 0.09105625252934035,
+      "grad_norm": 1.2570645809173584,
+      "learning_rate": 9.275242618211852e-05,
+      "loss": 1.0436,
+      "step": 2250
+    },
+    {
+      "epoch": 0.09125859975718333,
+      "grad_norm": 1.1203771829605103,
+      "learning_rate": 9.273177782366303e-05,
+      "loss": 0.9944,
+      "step": 2255
+    },
+    {
+      "epoch": 0.09146094698502631,
+      "grad_norm": 1.2696106433868408,
+      "learning_rate": 9.271112946520753e-05,
+      "loss": 1.0134,
+      "step": 2260
+    },
+    {
+      "epoch": 0.09166329421286928,
+      "grad_norm": 1.2304413318634033,
+      "learning_rate": 9.269048110675202e-05,
+      "loss": 1.0011,
+      "step": 2265
+    },
+    {
+      "epoch": 0.09186564144071226,
+      "grad_norm": 1.2513751983642578,
+      "learning_rate": 9.266983274829651e-05,
+      "loss": 1.0538,
+      "step": 2270
+    },
+    {
+      "epoch": 0.09206798866855524,
+      "grad_norm": 1.2047661542892456,
+      "learning_rate": 9.264918438984102e-05,
+      "loss": 0.9995,
+      "step": 2275
+    },
+    {
+      "epoch": 0.09227033589639821,
+      "grad_norm": 1.3607407808303833,
+      "learning_rate": 9.26285360313855e-05,
+      "loss": 1.0474,
+      "step": 2280
+    },
+    {
+      "epoch": 0.0924726831242412,
+      "grad_norm": 1.2867265939712524,
+      "learning_rate": 9.260788767293001e-05,
+      "loss": 1.0599,
+      "step": 2285
+    },
+    {
+      "epoch": 0.09267503035208417,
+      "grad_norm": 1.1794477701187134,
+      "learning_rate": 9.25872393144745e-05,
+      "loss": 1.0008,
+      "step": 2290
+    },
+    {
+      "epoch": 0.09287737757992716,
+      "grad_norm": 1.3165411949157715,
+      "learning_rate": 9.2566590956019e-05,
+      "loss": 1.0157,
+      "step": 2295
+    },
+    {
+      "epoch": 0.09307972480777013,
+      "grad_norm": 1.1853994131088257,
+      "learning_rate": 9.25459425975635e-05,
+      "loss": 1.079,
+      "step": 2300
+    },
+    {
+      "epoch": 0.09328207203561312,
+      "grad_norm": 1.2266639471054077,
+      "learning_rate": 9.2525294239108e-05,
+      "loss": 0.9848,
+      "step": 2305
+    },
+    {
+      "epoch": 0.09348441926345609,
+      "grad_norm": 1.2124624252319336,
+      "learning_rate": 9.250464588065249e-05,
+      "loss": 1.046,
+      "step": 2310
+    },
+    {
+      "epoch": 0.09368676649129908,
+      "grad_norm": 1.186888337135315,
+      "learning_rate": 9.248399752219698e-05,
+      "loss": 1.0597,
+      "step": 2315
+    },
+    {
+      "epoch": 0.09388911371914205,
+      "grad_norm": 1.2497700452804565,
+      "learning_rate": 9.246334916374148e-05,
+      "loss": 1.0414,
+      "step": 2320
+    },
+    {
+      "epoch": 0.09409146094698502,
+      "grad_norm": 1.2072032690048218,
+      "learning_rate": 9.244270080528599e-05,
+      "loss": 1.0096,
+      "step": 2325
+    },
+    {
+      "epoch": 0.094293808174828,
+      "grad_norm": 1.2232177257537842,
+      "learning_rate": 9.242205244683048e-05,
+      "loss": 1.0635,
+      "step": 2330
+    },
+    {
+      "epoch": 0.09449615540267098,
+      "grad_norm": 1.243944764137268,
+      "learning_rate": 9.240140408837498e-05,
+      "loss": 1.0287,
+      "step": 2335
+    },
+    {
+      "epoch": 0.09469850263051396,
+      "grad_norm": 1.3213430643081665,
+      "learning_rate": 9.238075572991947e-05,
+      "loss": 1.0428,
+      "step": 2340
+    },
+    {
+      "epoch": 0.09490084985835694,
+      "grad_norm": 1.518526315689087,
+      "learning_rate": 9.236010737146398e-05,
+      "loss": 0.9719,
+      "step": 2345
+    },
+    {
+      "epoch": 0.09510319708619992,
+      "grad_norm": 1.4371111392974854,
+      "learning_rate": 9.233945901300847e-05,
+      "loss": 1.0664,
+      "step": 2350
+    },
+    {
+      "epoch": 0.0953055443140429,
+      "grad_norm": 1.2702082395553589,
+      "learning_rate": 9.231881065455297e-05,
+      "loss": 1.0641,
+      "step": 2355
+    },
+    {
+      "epoch": 0.09550789154188588,
+      "grad_norm": 1.4100208282470703,
+      "learning_rate": 9.229816229609746e-05,
+      "loss": 0.9899,
+      "step": 2360
+    },
+    {
+      "epoch": 0.09571023876972885,
+      "grad_norm": 1.2306245565414429,
+      "learning_rate": 9.227751393764195e-05,
+      "loss": 0.9753,
+      "step": 2365
+    },
+    {
+      "epoch": 0.09591258599757184,
+      "grad_norm": 1.3112229108810425,
+      "learning_rate": 9.225686557918646e-05,
+      "loss": 1.0417,
+      "step": 2370
+    },
+    {
+      "epoch": 0.09611493322541481,
+      "grad_norm": 1.2760628461837769,
+      "learning_rate": 9.223621722073096e-05,
+      "loss": 1.01,
+      "step": 2375
+    },
+    {
+      "epoch": 0.09631728045325778,
+      "grad_norm": 1.2782249450683594,
+      "learning_rate": 9.221556886227547e-05,
+      "loss": 1.0407,
+      "step": 2380
+    },
+    {
+      "epoch": 0.09651962768110077,
+      "grad_norm": 1.4450081586837769,
+      "learning_rate": 9.219492050381994e-05,
+      "loss": 1.0293,
+      "step": 2385
+    },
+    {
+      "epoch": 0.09672197490894374,
+      "grad_norm": 1.0772294998168945,
+      "learning_rate": 9.217427214536445e-05,
+      "loss": 1.025,
+      "step": 2390
+    },
+    {
+      "epoch": 0.09692432213678673,
+      "grad_norm": 1.2389899492263794,
+      "learning_rate": 9.215362378690895e-05,
+      "loss": 1.0022,
+      "step": 2395
+    },
+    {
+      "epoch": 0.0971266693646297,
+      "grad_norm": 1.2315919399261475,
+      "learning_rate": 9.213297542845344e-05,
+      "loss": 1.0203,
+      "step": 2400
+    },
+    {
+      "epoch": 0.09732901659247269,
+      "grad_norm": 1.1966350078582764,
+      "learning_rate": 9.211232706999795e-05,
+      "loss": 0.9981,
+      "step": 2405
+    },
+    {
+      "epoch": 0.09753136382031566,
+      "grad_norm": 1.2289438247680664,
+      "learning_rate": 9.209167871154244e-05,
+      "loss": 1.0131,
+      "step": 2410
+    },
+    {
+      "epoch": 0.09773371104815864,
+      "grad_norm": 1.2669563293457031,
+      "learning_rate": 9.207103035308693e-05,
+      "loss": 1.0016,
+      "step": 2415
+    },
+    {
+      "epoch": 0.09793605827600162,
+      "grad_norm": 1.2027100324630737,
+      "learning_rate": 9.205038199463143e-05,
+      "loss": 1.0285,
+      "step": 2420
+    },
+    {
+      "epoch": 0.0981384055038446,
+      "grad_norm": 1.133360743522644,
+      "learning_rate": 9.202973363617593e-05,
+      "loss": 0.9646,
+      "step": 2425
+    },
+    {
+      "epoch": 0.09834075273168758,
+      "grad_norm": 1.3844754695892334,
+      "learning_rate": 9.200908527772042e-05,
+      "loss": 1.0183,
+      "step": 2430
+    },
+    {
+      "epoch": 0.09854309995953056,
+      "grad_norm": 1.2737170457839966,
+      "learning_rate": 9.198843691926492e-05,
+      "loss": 0.9493,
+      "step": 2435
+    },
+    {
+      "epoch": 0.09874544718737353,
+      "grad_norm": 1.3229756355285645,
+      "learning_rate": 9.196778856080942e-05,
+      "loss": 1.0441,
+      "step": 2440
+    },
+    {
+      "epoch": 0.0989477944152165,
+      "grad_norm": 1.3351441621780396,
+      "learning_rate": 9.194714020235392e-05,
+      "loss": 0.9643,
+      "step": 2445
+    },
+    {
+      "epoch": 0.09915014164305949,
+      "grad_norm": 1.4130533933639526,
+      "learning_rate": 9.192649184389841e-05,
+      "loss": 1.0617,
+      "step": 2450
+    },
+    {
+      "epoch": 0.09935248887090246,
+      "grad_norm": 1.3605046272277832,
+      "learning_rate": 9.19058434854429e-05,
+      "loss": 1.0063,
+      "step": 2455
+    },
+    {
+      "epoch": 0.09955483609874545,
+      "grad_norm": 1.1895331144332886,
+      "learning_rate": 9.188519512698741e-05,
+      "loss": 1.0004,
+      "step": 2460
+    },
+    {
+      "epoch": 0.09975718332658842,
+      "grad_norm": 1.2517513036727905,
+      "learning_rate": 9.186454676853191e-05,
+      "loss": 1.037,
+      "step": 2465
+    },
+    {
+      "epoch": 0.09995953055443141,
+      "grad_norm": 1.2222354412078857,
+      "learning_rate": 9.18438984100764e-05,
+      "loss": 1.0397,
+      "step": 2470
+    },
+    {
+      "epoch": 0.10016187778227438,
+      "grad_norm": 1.2808754444122314,
+      "learning_rate": 9.182325005162091e-05,
+      "loss": 1.016,
+      "step": 2475
+    },
+    {
+      "epoch": 0.10036422501011737,
+      "grad_norm": 1.2805005311965942,
+      "learning_rate": 9.18026016931654e-05,
+      "loss": 0.9915,
+      "step": 2480
+    },
+    {
+      "epoch": 0.10056657223796034,
+      "grad_norm": 1.2117308378219604,
+      "learning_rate": 9.178195333470989e-05,
+      "loss": 1.0035,
+      "step": 2485
+    },
+    {
+      "epoch": 0.10076891946580332,
+      "grad_norm": 1.232007622718811,
+      "learning_rate": 9.176130497625439e-05,
+      "loss": 1.0447,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1009712666936463,
+      "grad_norm": 1.306788444519043,
+      "learning_rate": 9.17406566177989e-05,
+      "loss": 1.0473,
+      "step": 2495
+    },
+    {
+      "epoch": 0.10117361392148927,
+      "grad_norm": 1.1974440813064575,
+      "learning_rate": 9.172000825934339e-05,
+      "loss": 1.0293,
+      "step": 2500
+    },
+    {
+      "epoch": 0.10137596114933226,
+      "grad_norm": 1.2958585023880005,
+      "learning_rate": 9.169935990088788e-05,
+      "loss": 1.0503,
+      "step": 2505
+    },
+    {
+      "epoch": 0.10157830837717523,
+      "grad_norm": 1.1704072952270508,
+      "learning_rate": 9.167871154243238e-05,
+      "loss": 1.0912,
+      "step": 2510
+    },
+    {
+      "epoch": 0.10178065560501821,
+      "grad_norm": 1.2891044616699219,
+      "learning_rate": 9.165806318397689e-05,
+      "loss": 1.0266,
+      "step": 2515
+    },
+    {
+      "epoch": 0.10198300283286119,
+      "grad_norm": 1.2312427759170532,
+      "learning_rate": 9.163741482552138e-05,
+      "loss": 0.9993,
+      "step": 2520
+    },
+    {
+      "epoch": 0.10218535006070417,
+      "grad_norm": 1.2950193881988525,
+      "learning_rate": 9.161676646706587e-05,
+      "loss": 1.0399,
+      "step": 2525
+    },
+    {
+      "epoch": 0.10238769728854714,
+      "grad_norm": 1.2646780014038086,
+      "learning_rate": 9.159611810861037e-05,
+      "loss": 1.0725,
+      "step": 2530
+    },
+    {
+      "epoch": 0.10259004451639013,
+      "grad_norm": 1.1012483835220337,
+      "learning_rate": 9.157546975015486e-05,
+      "loss": 1.0173,
+      "step": 2535
+    },
+    {
+      "epoch": 0.1027923917442331,
+      "grad_norm": 1.278733491897583,
+      "learning_rate": 9.155482139169937e-05,
+      "loss": 1.0597,
+      "step": 2540
+    },
+    {
+      "epoch": 0.10299473897207609,
+      "grad_norm": 1.0903396606445312,
+      "learning_rate": 9.153417303324387e-05,
+      "loss": 0.9878,
+      "step": 2545
+    },
+    {
+      "epoch": 0.10319708619991906,
+      "grad_norm": 1.290555477142334,
+      "learning_rate": 9.151352467478836e-05,
+      "loss": 1.0107,
+      "step": 2550
+    },
+    {
+      "epoch": 0.10339943342776203,
+      "grad_norm": 1.3766034841537476,
+      "learning_rate": 9.149287631633285e-05,
+      "loss": 1.0373,
+      "step": 2555
+    },
+    {
+      "epoch": 0.10360178065560502,
+      "grad_norm": 1.2376962900161743,
+      "learning_rate": 9.147222795787735e-05,
+      "loss": 1.1126,
+      "step": 2560
+    },
+    {
+      "epoch": 0.10380412788344799,
+      "grad_norm": 1.1498472690582275,
+      "learning_rate": 9.145157959942186e-05,
+      "loss": 1.0453,
+      "step": 2565
+    },
+    {
+      "epoch": 0.10400647511129098,
+      "grad_norm": 1.2858022451400757,
+      "learning_rate": 9.143093124096635e-05,
+      "loss": 1.0428,
+      "step": 2570
+    },
+    {
+      "epoch": 0.10420882233913395,
+      "grad_norm": 1.219605803489685,
+      "learning_rate": 9.141028288251084e-05,
+      "loss": 1.0017,
+      "step": 2575
+    },
+    {
+      "epoch": 0.10441116956697694,
+      "grad_norm": 1.31840181350708,
+      "learning_rate": 9.138963452405534e-05,
+      "loss": 1.0432,
+      "step": 2580
+    },
+    {
+      "epoch": 0.10461351679481991,
+      "grad_norm": 1.096554160118103,
+      "learning_rate": 9.136898616559983e-05,
+      "loss": 0.9801,
+      "step": 2585
+    },
+    {
+      "epoch": 0.1048158640226629,
+      "grad_norm": 1.1494611501693726,
+      "learning_rate": 9.134833780714434e-05,
+      "loss": 1.0558,
+      "step": 2590
+    },
+    {
+      "epoch": 0.10501821125050587,
+      "grad_norm": 1.178409218788147,
+      "learning_rate": 9.132768944868884e-05,
+      "loss": 1.0043,
+      "step": 2595
+    },
+    {
+      "epoch": 0.10522055847834885,
+      "grad_norm": 1.216878056526184,
+      "learning_rate": 9.130704109023333e-05,
+      "loss": 1.0066,
+      "step": 2600
+    },
+    {
+      "epoch": 0.10542290570619182,
+      "grad_norm": 1.221721887588501,
+      "learning_rate": 9.128639273177782e-05,
+      "loss": 1.0711,
+      "step": 2605
+    },
+    {
+      "epoch": 0.10562525293403481,
+      "grad_norm": 1.1642242670059204,
+      "learning_rate": 9.126574437332233e-05,
+      "loss": 1.0198,
+      "step": 2610
+    },
+    {
+      "epoch": 0.10582760016187778,
+      "grad_norm": 1.2565926313400269,
+      "learning_rate": 9.124509601486683e-05,
+      "loss": 0.9287,
+      "step": 2615
+    },
+    {
+      "epoch": 0.10602994738972076,
+      "grad_norm": 1.2995041608810425,
+      "learning_rate": 9.122444765641132e-05,
+      "loss": 1.051,
+      "step": 2620
+    },
+    {
+      "epoch": 0.10623229461756374,
+      "grad_norm": 1.163234829902649,
+      "learning_rate": 9.120379929795581e-05,
+      "loss": 1.0064,
+      "step": 2625
+    },
+    {
+      "epoch": 0.10643464184540671,
+      "grad_norm": 1.3579676151275635,
+      "learning_rate": 9.118315093950032e-05,
+      "loss": 1.0226,
+      "step": 2630
+    },
+    {
+      "epoch": 0.1066369890732497,
+      "grad_norm": 1.2112040519714355,
+      "learning_rate": 9.116250258104481e-05,
+      "loss": 1.0841,
+      "step": 2635
+    },
+    {
+      "epoch": 0.10683933630109267,
+      "grad_norm": 1.1701689958572388,
+      "learning_rate": 9.114185422258931e-05,
+      "loss": 1.0144,
+      "step": 2640
+    },
+    {
+      "epoch": 0.10704168352893566,
+      "grad_norm": 1.2789571285247803,
+      "learning_rate": 9.11212058641338e-05,
+      "loss": 1.0044,
+      "step": 2645
+    },
+    {
+      "epoch": 0.10724403075677863,
+      "grad_norm": 1.1566429138183594,
+      "learning_rate": 9.11005575056783e-05,
+      "loss": 1.0511,
+      "step": 2650
+    },
+    {
+      "epoch": 0.10744637798462162,
+      "grad_norm": 1.3453134298324585,
+      "learning_rate": 9.10799091472228e-05,
+      "loss": 1.0144,
+      "step": 2655
+    },
+    {
+      "epoch": 0.10764872521246459,
+      "grad_norm": 1.2248982191085815,
+      "learning_rate": 9.10592607887673e-05,
+      "loss": 1.0253,
+      "step": 2660
+    },
+    {
+      "epoch": 0.10785107244030757,
+      "grad_norm": 1.1890755891799927,
+      "learning_rate": 9.10386124303118e-05,
+      "loss": 1.0566,
+      "step": 2665
+    },
+    {
+      "epoch": 0.10805341966815055,
+      "grad_norm": 1.3527295589447021,
+      "learning_rate": 9.101796407185628e-05,
+      "loss": 1.0318,
+      "step": 2670
+    },
+    {
+      "epoch": 0.10825576689599352,
+      "grad_norm": 1.111272931098938,
+      "learning_rate": 9.099731571340079e-05,
+      "loss": 1.0269,
+      "step": 2675
+    },
+    {
+      "epoch": 0.1084581141238365,
+      "grad_norm": 1.19521963596344,
+      "learning_rate": 9.097666735494529e-05,
+      "loss": 0.9893,
+      "step": 2680
+    },
+    {
+      "epoch": 0.10866046135167948,
+      "grad_norm": 1.269953966140747,
+      "learning_rate": 9.095601899648978e-05,
+      "loss": 0.9787,
+      "step": 2685
+    },
+    {
+      "epoch": 0.10886280857952246,
+      "grad_norm": 1.2876157760620117,
+      "learning_rate": 9.093537063803428e-05,
+      "loss": 0.981,
+      "step": 2690
+    },
+    {
+      "epoch": 0.10906515580736544,
+      "grad_norm": 1.2349538803100586,
+      "learning_rate": 9.091472227957877e-05,
+      "loss": 0.9918,
+      "step": 2695
+    },
+    {
+      "epoch": 0.10926750303520842,
+      "grad_norm": 1.266349196434021,
+      "learning_rate": 9.089407392112328e-05,
+      "loss": 0.9416,
+      "step": 2700
+    },
+    {
+      "epoch": 0.1094698502630514,
+      "grad_norm": 1.300087571144104,
+      "learning_rate": 9.087342556266777e-05,
+      "loss": 1.0499,
+      "step": 2705
+    },
+    {
+      "epoch": 0.10967219749089438,
+      "grad_norm": 1.2513761520385742,
+      "learning_rate": 9.085277720421227e-05,
+      "loss": 1.0398,
+      "step": 2710
+    },
+    {
+      "epoch": 0.10987454471873735,
+      "grad_norm": 1.3558810949325562,
+      "learning_rate": 9.083212884575676e-05,
+      "loss": 1.0786,
+      "step": 2715
+    },
+    {
+      "epoch": 0.11007689194658034,
+      "grad_norm": 1.3106088638305664,
+      "learning_rate": 9.081148048730125e-05,
+      "loss": 0.982,
+      "step": 2720
+    },
+    {
+      "epoch": 0.11027923917442331,
+      "grad_norm": 1.3098465204238892,
+      "learning_rate": 9.079083212884576e-05,
+      "loss": 0.9478,
+      "step": 2725
+    },
+    {
+      "epoch": 0.11048158640226628,
+      "grad_norm": 1.139407992362976,
+      "learning_rate": 9.077018377039026e-05,
+      "loss": 1.0434,
+      "step": 2730
+    },
+    {
+      "epoch": 0.11068393363010927,
+      "grad_norm": 1.3353989124298096,
+      "learning_rate": 9.074953541193477e-05,
+      "loss": 1.0305,
+      "step": 2735
+    },
+    {
+      "epoch": 0.11088628085795224,
+      "grad_norm": 1.0765186548233032,
+      "learning_rate": 9.072888705347924e-05,
+      "loss": 1.0019,
+      "step": 2740
+    },
+    {
+      "epoch": 0.11108862808579523,
+      "grad_norm": 1.3412421941757202,
+      "learning_rate": 9.070823869502375e-05,
+      "loss": 1.0322,
+      "step": 2745
+    },
+    {
+      "epoch": 0.1112909753136382,
+      "grad_norm": 1.3108158111572266,
+      "learning_rate": 9.068759033656825e-05,
+      "loss": 1.0105,
+      "step": 2750
+    },
+    {
+      "epoch": 0.11149332254148119,
+      "grad_norm": 1.2250312566757202,
+      "learning_rate": 9.066694197811274e-05,
+      "loss": 1.03,
+      "step": 2755
+    },
+    {
+      "epoch": 0.11169566976932416,
+      "grad_norm": 1.0880059003829956,
+      "learning_rate": 9.064629361965725e-05,
+      "loss": 1.0655,
+      "step": 2760
+    },
+    {
+      "epoch": 0.11189801699716714,
+      "grad_norm": 1.1405460834503174,
+      "learning_rate": 9.062564526120174e-05,
+      "loss": 1.0332,
+      "step": 2765
+    },
+    {
+      "epoch": 0.11210036422501012,
+      "grad_norm": 1.14668607711792,
+      "learning_rate": 9.060499690274623e-05,
+      "loss": 1.0427,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1123027114528531,
+      "grad_norm": 1.274117112159729,
+      "learning_rate": 9.058434854429073e-05,
+      "loss": 1.0524,
+      "step": 2775
+    },
+    {
+      "epoch": 0.11250505868069607,
+      "grad_norm": 1.285851240158081,
+      "learning_rate": 9.056370018583524e-05,
+      "loss": 1.0323,
+      "step": 2780
+    },
+    {
+      "epoch": 0.11270740590853905,
+      "grad_norm": 1.1409989595413208,
+      "learning_rate": 9.054305182737973e-05,
+      "loss": 0.9721,
+      "step": 2785
+    },
+    {
+      "epoch": 0.11290975313638203,
+      "grad_norm": 1.226494550704956,
+      "learning_rate": 9.052240346892422e-05,
+      "loss": 0.9696,
+      "step": 2790
+    },
+    {
+      "epoch": 0.113112100364225,
+      "grad_norm": 1.3489365577697754,
+      "learning_rate": 9.050175511046872e-05,
+      "loss": 1.0354,
+      "step": 2795
+    },
+    {
+      "epoch": 0.11331444759206799,
+      "grad_norm": 1.182013988494873,
+      "learning_rate": 9.048110675201322e-05,
+      "loss": 1.0206,
+      "step": 2800
+    },
+    {
+      "epoch": 0.11351679481991096,
+      "grad_norm": 1.232530951499939,
+      "learning_rate": 9.046045839355772e-05,
+      "loss": 1.0469,
+      "step": 2805
+    },
+    {
+      "epoch": 0.11371914204775395,
+      "grad_norm": 1.1662116050720215,
+      "learning_rate": 9.043981003510222e-05,
+      "loss": 1.1351,
+      "step": 2810
+    },
+    {
+      "epoch": 0.11392148927559692,
+      "grad_norm": 1.294116735458374,
+      "learning_rate": 9.041916167664671e-05,
+      "loss": 0.9955,
+      "step": 2815
+    },
+    {
+      "epoch": 0.11412383650343991,
+      "grad_norm": 1.209054708480835,
+      "learning_rate": 9.039851331819121e-05,
+      "loss": 1.1027,
+      "step": 2820
+    },
+    {
+      "epoch": 0.11432618373128288,
+      "grad_norm": 1.2408312559127808,
+      "learning_rate": 9.03778649597357e-05,
+      "loss": 0.9704,
+      "step": 2825
+    },
+    {
+      "epoch": 0.11452853095912587,
+      "grad_norm": 1.1527987718582153,
+      "learning_rate": 9.035721660128021e-05,
+      "loss": 1.04,
+      "step": 2830
+    },
+    {
+      "epoch": 0.11473087818696884,
+      "grad_norm": 1.2887940406799316,
+      "learning_rate": 9.03365682428247e-05,
+      "loss": 0.9587,
+      "step": 2835
+    },
+    {
+      "epoch": 0.11493322541481182,
+      "grad_norm": 1.1732901334762573,
+      "learning_rate": 9.031591988436919e-05,
+      "loss": 1.0022,
+      "step": 2840
+    },
+    {
+      "epoch": 0.1151355726426548,
+      "grad_norm": 1.3233847618103027,
+      "learning_rate": 9.02952715259137e-05,
+      "loss": 1.0523,
+      "step": 2845
+    },
+    {
+      "epoch": 0.11533791987049777,
+      "grad_norm": 1.1359490156173706,
+      "learning_rate": 9.02746231674582e-05,
+      "loss": 1.022,
+      "step": 2850
+    },
+    {
+      "epoch": 0.11554026709834075,
+      "grad_norm": 1.1741466522216797,
+      "learning_rate": 9.025397480900269e-05,
+      "loss": 1.0289,
+      "step": 2855
+    },
+    {
+      "epoch": 0.11574261432618373,
+      "grad_norm": 1.1990710496902466,
+      "learning_rate": 9.023332645054718e-05,
+      "loss": 1.0145,
+      "step": 2860
+    },
+    {
+      "epoch": 0.11594496155402671,
+      "grad_norm": 1.251400351524353,
+      "learning_rate": 9.021267809209168e-05,
+      "loss": 1.019,
+      "step": 2865
+    },
+    {
+      "epoch": 0.11614730878186968,
+      "grad_norm": 1.3180302381515503,
+      "learning_rate": 9.019202973363619e-05,
+      "loss": 1.0802,
+      "step": 2870
+    },
+    {
+      "epoch": 0.11634965600971267,
+      "grad_norm": 1.184276819229126,
+      "learning_rate": 9.017138137518068e-05,
+      "loss": 0.9496,
+      "step": 2875
+    },
+    {
+      "epoch": 0.11655200323755564,
+      "grad_norm": 1.1231951713562012,
+      "learning_rate": 9.015073301672518e-05,
+      "loss": 1.0757,
+      "step": 2880
+    },
+    {
+      "epoch": 0.11675435046539863,
+      "grad_norm": 1.166365385055542,
+      "learning_rate": 9.013008465826967e-05,
+      "loss": 1.0515,
+      "step": 2885
+    },
+    {
+      "epoch": 0.1169566976932416,
+      "grad_norm": 1.0476434230804443,
+      "learning_rate": 9.010943629981416e-05,
+      "loss": 0.9998,
+      "step": 2890
+    },
+    {
+      "epoch": 0.11715904492108459,
+      "grad_norm": 1.2853447198867798,
+      "learning_rate": 9.008878794135867e-05,
+      "loss": 1.0104,
+      "step": 2895
+    },
+    {
+      "epoch": 0.11736139214892756,
+      "grad_norm": 1.2302528619766235,
+      "learning_rate": 9.006813958290317e-05,
+      "loss": 0.9935,
+      "step": 2900
+    },
+    {
+      "epoch": 0.11756373937677053,
+      "grad_norm": 1.1947234869003296,
+      "learning_rate": 9.004749122444766e-05,
+      "loss": 1.0557,
+      "step": 2905
+    },
+    {
+      "epoch": 0.11776608660461352,
+      "grad_norm": 1.4698901176452637,
+      "learning_rate": 9.002684286599215e-05,
+      "loss": 1.0283,
+      "step": 2910
+    },
+    {
+      "epoch": 0.11796843383245649,
+      "grad_norm": 1.3562864065170288,
+      "learning_rate": 9.000619450753666e-05,
+      "loss": 1.022,
+      "step": 2915
+    },
+    {
+      "epoch": 0.11817078106029948,
+      "grad_norm": 1.2436528205871582,
+      "learning_rate": 8.998554614908116e-05,
+      "loss": 1.0331,
+      "step": 2920
+    },
+    {
+      "epoch": 0.11837312828814245,
+      "grad_norm": 1.2349079847335815,
+      "learning_rate": 8.996489779062565e-05,
+      "loss": 1.0529,
+      "step": 2925
+    },
+    {
+      "epoch": 0.11857547551598543,
+      "grad_norm": 1.2597922086715698,
+      "learning_rate": 8.994424943217014e-05,
+      "loss": 1.0482,
+      "step": 2930
+    },
+    {
+      "epoch": 0.1187778227438284,
+      "grad_norm": 1.2937184572219849,
+      "learning_rate": 8.992360107371464e-05,
+      "loss": 1.0562,
+      "step": 2935
+    },
+    {
+      "epoch": 0.11898016997167139,
+      "grad_norm": 1.1591882705688477,
+      "learning_rate": 8.990295271525914e-05,
+      "loss": 1.0604,
+      "step": 2940
+    },
+    {
+      "epoch": 0.11918251719951437,
+      "grad_norm": 1.2187837362289429,
+      "learning_rate": 8.988230435680364e-05,
+      "loss": 1.0112,
+      "step": 2945
+    },
+    {
+      "epoch": 0.11938486442735735,
+      "grad_norm": 1.216113567352295,
+      "learning_rate": 8.986165599834814e-05,
+      "loss": 1.0057,
+      "step": 2950
+    },
+    {
+      "epoch": 0.11958721165520032,
+      "grad_norm": 1.253690242767334,
+      "learning_rate": 8.984100763989263e-05,
+      "loss": 0.9962,
+      "step": 2955
+    },
+    {
+      "epoch": 0.1197895588830433,
+      "grad_norm": 1.1752190589904785,
+      "learning_rate": 8.982035928143712e-05,
+      "loss": 0.9657,
+      "step": 2960
+    },
+    {
+      "epoch": 0.11999190611088628,
+      "grad_norm": 1.2840155363082886,
+      "learning_rate": 8.979971092298163e-05,
+      "loss": 1.0075,
+      "step": 2965
+    },
+    {
+      "epoch": 0.12019425333872925,
+      "grad_norm": 1.1470037698745728,
+      "learning_rate": 8.977906256452613e-05,
+      "loss": 1.0055,
+      "step": 2970
+    },
+    {
+      "epoch": 0.12039660056657224,
+      "grad_norm": 1.2354692220687866,
+      "learning_rate": 8.975841420607062e-05,
+      "loss": 1.0202,
+      "step": 2975
+    },
+    {
+      "epoch": 0.12059894779441521,
+      "grad_norm": 1.3677310943603516,
+      "learning_rate": 8.973776584761511e-05,
+      "loss": 1.0447,
+      "step": 2980
+    },
+    {
+      "epoch": 0.1208012950222582,
+      "grad_norm": 1.186793327331543,
+      "learning_rate": 8.971711748915962e-05,
+      "loss": 1.0494,
+      "step": 2985
+    },
+    {
+      "epoch": 0.12100364225010117,
+      "grad_norm": 1.5150768756866455,
+      "learning_rate": 8.969646913070411e-05,
+      "loss": 1.0144,
+      "step": 2990
+    },
+    {
+      "epoch": 0.12120598947794416,
+      "grad_norm": 1.2600922584533691,
+      "learning_rate": 8.967582077224861e-05,
+      "loss": 1.0463,
+      "step": 2995
+    },
+    {
+      "epoch": 0.12140833670578713,
+      "grad_norm": 1.2720588445663452,
+      "learning_rate": 8.96551724137931e-05,
+      "loss": 0.9943,
+      "step": 3000
+    },
+    {
+      "epoch": 0.12161068393363011,
+      "grad_norm": 1.140571117401123,
+      "learning_rate": 8.963452405533761e-05,
+      "loss": 0.9777,
+      "step": 3005
+    },
+    {
+      "epoch": 0.12181303116147309,
+      "grad_norm": 1.1251741647720337,
+      "learning_rate": 8.96138756968821e-05,
+      "loss": 1.0613,
+      "step": 3010
+    },
+    {
+      "epoch": 0.12201537838931607,
+      "grad_norm": 1.306720495223999,
+      "learning_rate": 8.95932273384266e-05,
+      "loss": 0.9563,
+      "step": 3015
+    },
+    {
+      "epoch": 0.12221772561715905,
+      "grad_norm": 1.2379626035690308,
+      "learning_rate": 8.95725789799711e-05,
+      "loss": 1.0389,
+      "step": 3020
+    },
+    {
+      "epoch": 0.12242007284500202,
+      "grad_norm": 1.3128517866134644,
+      "learning_rate": 8.955193062151558e-05,
+      "loss": 0.985,
+      "step": 3025
+    },
+    {
+      "epoch": 0.122622420072845,
+      "grad_norm": 1.198457956314087,
+      "learning_rate": 8.953128226306009e-05,
+      "loss": 1.0527,
+      "step": 3030
+    },
+    {
+      "epoch": 0.12282476730068798,
+      "grad_norm": 1.2627030611038208,
+      "learning_rate": 8.951063390460459e-05,
+      "loss": 1.0217,
+      "step": 3035
+    },
+    {
+      "epoch": 0.12302711452853096,
+      "grad_norm": 1.3503879308700562,
+      "learning_rate": 8.948998554614908e-05,
+      "loss": 0.9815,
+      "step": 3040
+    },
+    {
+      "epoch": 0.12322946175637393,
+      "grad_norm": 1.3443350791931152,
+      "learning_rate": 8.946933718769359e-05,
+      "loss": 1.049,
+      "step": 3045
+    },
+    {
+      "epoch": 0.12343180898421692,
+      "grad_norm": 1.224956750869751,
+      "learning_rate": 8.944868882923808e-05,
+      "loss": 1.0303,
+      "step": 3050
+    },
+    {
+      "epoch": 0.12363415621205989,
+      "grad_norm": 1.2062424421310425,
+      "learning_rate": 8.942804047078258e-05,
+      "loss": 1.0169,
+      "step": 3055
+    },
+    {
+      "epoch": 0.12383650343990288,
+      "grad_norm": 1.3424972295761108,
+      "learning_rate": 8.940739211232707e-05,
+      "loss": 1.0114,
+      "step": 3060
+    },
+    {
+      "epoch": 0.12403885066774585,
+      "grad_norm": 1.1913210153579712,
+      "learning_rate": 8.938674375387157e-05,
+      "loss": 0.9841,
+      "step": 3065
+    },
+    {
+      "epoch": 0.12424119789558884,
+      "grad_norm": 1.1994571685791016,
+      "learning_rate": 8.936609539541607e-05,
+      "loss": 1.016,
+      "step": 3070
+    },
+    {
+      "epoch": 0.12444354512343181,
+      "grad_norm": 1.1911218166351318,
+      "learning_rate": 8.934544703696056e-05,
+      "loss": 1.0134,
+      "step": 3075
+    },
+    {
+      "epoch": 0.12464589235127478,
+      "grad_norm": 1.2363308668136597,
+      "learning_rate": 8.932479867850506e-05,
+      "loss": 1.0685,
+      "step": 3080
+    },
+    {
+      "epoch": 0.12484823957911777,
+      "grad_norm": 1.1953235864639282,
+      "learning_rate": 8.930415032004956e-05,
+      "loss": 1.0589,
+      "step": 3085
+    },
+    {
+      "epoch": 0.12505058680696074,
+      "grad_norm": 1.1757105588912964,
+      "learning_rate": 8.928350196159407e-05,
+      "loss": 1.0334,
+      "step": 3090
+    },
+    {
+      "epoch": 0.12525293403480373,
+      "grad_norm": 1.1546622514724731,
+      "learning_rate": 8.926285360313856e-05,
+      "loss": 1.0462,
+      "step": 3095
+    },
+    {
+      "epoch": 0.1254552812626467,
+      "grad_norm": 1.316259741783142,
+      "learning_rate": 8.924220524468305e-05,
+      "loss": 1.0553,
+      "step": 3100
+    },
+    {
+      "epoch": 0.12565762849048967,
+      "grad_norm": 1.2377395629882812,
+      "learning_rate": 8.922155688622755e-05,
+      "loss": 0.9546,
+      "step": 3105
+    },
+    {
+      "epoch": 0.12585997571833266,
+      "grad_norm": 1.2293140888214111,
+      "learning_rate": 8.920090852777204e-05,
+      "loss": 1.0169,
+      "step": 3110
+    },
+    {
+      "epoch": 0.12606232294617564,
+      "grad_norm": 1.112518548965454,
+      "learning_rate": 8.918026016931655e-05,
+      "loss": 1.0356,
+      "step": 3115
+    },
+    {
+      "epoch": 0.12626467017401863,
+      "grad_norm": 1.2322146892547607,
+      "learning_rate": 8.915961181086104e-05,
+      "loss": 1.0166,
+      "step": 3120
+    },
+    {
+      "epoch": 0.1264670174018616,
+      "grad_norm": 1.1821508407592773,
+      "learning_rate": 8.913896345240553e-05,
+      "loss": 1.0381,
+      "step": 3125
+    },
+    {
+      "epoch": 0.12666936462970457,
+      "grad_norm": 1.210894227027893,
+      "learning_rate": 8.911831509395003e-05,
+      "loss": 0.9886,
+      "step": 3130
+    },
+    {
+      "epoch": 0.12687171185754756,
+      "grad_norm": 1.4033355712890625,
+      "learning_rate": 8.909766673549454e-05,
+      "loss": 1.0438,
+      "step": 3135
+    },
+    {
+      "epoch": 0.12707405908539052,
+      "grad_norm": 1.3342198133468628,
+      "learning_rate": 8.907701837703904e-05,
+      "loss": 1.0696,
+      "step": 3140
+    },
+    {
+      "epoch": 0.1272764063132335,
+      "grad_norm": 1.3505562543869019,
+      "learning_rate": 8.905637001858352e-05,
+      "loss": 1.0322,
+      "step": 3145
+    },
+    {
+      "epoch": 0.1274787535410765,
+      "grad_norm": 1.1969664096832275,
+      "learning_rate": 8.903572166012802e-05,
+      "loss": 0.9917,
+      "step": 3150
+    },
+    {
+      "epoch": 0.12768110076891948,
+      "grad_norm": 1.1248817443847656,
+      "learning_rate": 8.901507330167253e-05,
+      "loss": 1.022,
+      "step": 3155
+    },
+    {
+      "epoch": 0.12788344799676243,
+      "grad_norm": 1.4438211917877197,
+      "learning_rate": 8.899442494321702e-05,
+      "loss": 1.021,
+      "step": 3160
+    },
+    {
+      "epoch": 0.12808579522460542,
+      "grad_norm": 1.2094289064407349,
+      "learning_rate": 8.897377658476152e-05,
+      "loss": 1.0498,
+      "step": 3165
+    },
+    {
+      "epoch": 0.1282881424524484,
+      "grad_norm": 1.34991455078125,
+      "learning_rate": 8.895312822630601e-05,
+      "loss": 1.0577,
+      "step": 3170
+    },
+    {
+      "epoch": 0.1284904896802914,
+      "grad_norm": 1.1748855113983154,
+      "learning_rate": 8.893247986785052e-05,
+      "loss": 1.0063,
+      "step": 3175
+    },
+    {
+      "epoch": 0.12869283690813435,
+      "grad_norm": 1.327193021774292,
+      "learning_rate": 8.8911831509395e-05,
+      "loss": 1.036,
+      "step": 3180
+    },
+    {
+      "epoch": 0.12889518413597734,
+      "grad_norm": 1.2270658016204834,
+      "learning_rate": 8.889118315093951e-05,
+      "loss": 1.0837,
+      "step": 3185
+    },
+    {
+      "epoch": 0.12909753136382032,
+      "grad_norm": 1.312050223350525,
+      "learning_rate": 8.8870534792484e-05,
+      "loss": 0.9861,
+      "step": 3190
+    },
+    {
+      "epoch": 0.12929987859166328,
+      "grad_norm": 1.2333570718765259,
+      "learning_rate": 8.884988643402849e-05,
+      "loss": 1.0555,
+      "step": 3195
+    },
+    {
+      "epoch": 0.12950222581950627,
+      "grad_norm": 1.2160171270370483,
+      "learning_rate": 8.8829238075573e-05,
+      "loss": 1.0088,
+      "step": 3200
+    },
+    {
+      "epoch": 0.12970457304734925,
+      "grad_norm": 1.2750040292739868,
+      "learning_rate": 8.88085897171175e-05,
+      "loss": 1.0411,
+      "step": 3205
+    },
+    {
+      "epoch": 0.12990692027519224,
+      "grad_norm": 1.3623186349868774,
+      "learning_rate": 8.878794135866199e-05,
+      "loss": 1.0382,
+      "step": 3210
+    },
+    {
+      "epoch": 0.1301092675030352,
+      "grad_norm": 1.3808802366256714,
+      "learning_rate": 8.876729300020648e-05,
+      "loss": 1.0019,
+      "step": 3215
+    },
+    {
+      "epoch": 0.13031161473087818,
+      "grad_norm": 1.3014731407165527,
+      "learning_rate": 8.874664464175098e-05,
+      "loss": 1.0205,
+      "step": 3220
+    },
+    {
+      "epoch": 0.13051396195872117,
+      "grad_norm": 1.1322846412658691,
+      "learning_rate": 8.872599628329549e-05,
+      "loss": 1.0346,
+      "step": 3225
+    },
+    {
+      "epoch": 0.13071630918656416,
+      "grad_norm": 1.2544022798538208,
+      "learning_rate": 8.870534792483998e-05,
+      "loss": 1.0122,
+      "step": 3230
+    },
+    {
+      "epoch": 0.13091865641440711,
+      "grad_norm": 1.1611053943634033,
+      "learning_rate": 8.868469956638448e-05,
+      "loss": 1.0121,
+      "step": 3235
+    },
+    {
+      "epoch": 0.1311210036422501,
+      "grad_norm": 1.2711516618728638,
+      "learning_rate": 8.866405120792897e-05,
+      "loss": 1.0305,
+      "step": 3240
+    },
+    {
+      "epoch": 0.1313233508700931,
+      "grad_norm": 1.1643027067184448,
+      "learning_rate": 8.864340284947346e-05,
+      "loss": 1.0167,
+      "step": 3245
+    },
+    {
+      "epoch": 0.13152569809793604,
+      "grad_norm": 1.1797773838043213,
+      "learning_rate": 8.862275449101797e-05,
+      "loss": 1.0055,
+      "step": 3250
+    },
+    {
+      "epoch": 0.13172804532577903,
+      "grad_norm": 1.0747370719909668,
+      "learning_rate": 8.860210613256247e-05,
+      "loss": 0.9986,
+      "step": 3255
+    },
+    {
+      "epoch": 0.13193039255362202,
+      "grad_norm": 1.1961429119110107,
+      "learning_rate": 8.858145777410696e-05,
+      "loss": 1.059,
+      "step": 3260
+    },
+    {
+      "epoch": 0.132132739781465,
+      "grad_norm": 1.2358653545379639,
+      "learning_rate": 8.856080941565145e-05,
+      "loss": 0.9815,
+      "step": 3265
+    },
+    {
+      "epoch": 0.13233508700930796,
+      "grad_norm": 1.2921229600906372,
+      "learning_rate": 8.854016105719596e-05,
+      "loss": 1.0244,
+      "step": 3270
+    },
+    {
+      "epoch": 0.13253743423715095,
+      "grad_norm": 1.1417112350463867,
+      "learning_rate": 8.851951269874046e-05,
+      "loss": 0.9919,
+      "step": 3275
+    },
+    {
+      "epoch": 0.13273978146499393,
+      "grad_norm": 1.225769281387329,
+      "learning_rate": 8.849886434028495e-05,
+      "loss": 0.9792,
+      "step": 3280
+    },
+    {
+      "epoch": 0.13294212869283692,
+      "grad_norm": 1.1507500410079956,
+      "learning_rate": 8.847821598182944e-05,
+      "loss": 0.9951,
+      "step": 3285
+    },
+    {
+      "epoch": 0.13314447592067988,
+      "grad_norm": 1.2668052911758423,
+      "learning_rate": 8.845756762337395e-05,
+      "loss": 1.0371,
+      "step": 3290
+    },
+    {
+      "epoch": 0.13334682314852286,
+      "grad_norm": 1.1190897226333618,
+      "learning_rate": 8.843691926491844e-05,
+      "loss": 1.0822,
+      "step": 3295
+    },
+    {
+      "epoch": 0.13354917037636585,
+      "grad_norm": 1.2662688493728638,
+      "learning_rate": 8.841627090646294e-05,
+      "loss": 0.9802,
+      "step": 3300
+    },
+    {
+      "epoch": 0.13375151760420884,
+      "grad_norm": 1.2569993734359741,
+      "learning_rate": 8.839562254800744e-05,
+      "loss": 1.0317,
+      "step": 3305
+    },
+    {
+      "epoch": 0.1339538648320518,
+      "grad_norm": 1.0859438180923462,
+      "learning_rate": 8.837497418955194e-05,
+      "loss": 1.0241,
+      "step": 3310
+    },
+    {
+      "epoch": 0.13415621205989478,
+      "grad_norm": 1.2171080112457275,
+      "learning_rate": 8.835432583109643e-05,
+      "loss": 0.9646,
+      "step": 3315
+    },
+    {
+      "epoch": 0.13435855928773777,
+      "grad_norm": 1.2457976341247559,
+      "learning_rate": 8.833367747264093e-05,
+      "loss": 1.0491,
+      "step": 3320
+    },
+    {
+      "epoch": 0.13456090651558072,
+      "grad_norm": 1.2020723819732666,
+      "learning_rate": 8.831302911418543e-05,
+      "loss": 1.0559,
+      "step": 3325
+    },
+    {
+      "epoch": 0.1347632537434237,
+      "grad_norm": 1.2359280586242676,
+      "learning_rate": 8.829238075572992e-05,
+      "loss": 1.0673,
+      "step": 3330
+    },
+    {
+      "epoch": 0.1349656009712667,
+      "grad_norm": 1.119128942489624,
+      "learning_rate": 8.827173239727442e-05,
+      "loss": 1.0245,
+      "step": 3335
+    },
+    {
+      "epoch": 0.13516794819910968,
+      "grad_norm": 1.1393615007400513,
+      "learning_rate": 8.825108403881892e-05,
+      "loss": 1.0338,
+      "step": 3340
+    },
+    {
+      "epoch": 0.13537029542695264,
+      "grad_norm": 1.202042579650879,
+      "learning_rate": 8.823043568036341e-05,
+      "loss": 1.0262,
+      "step": 3345
+    },
+    {
+      "epoch": 0.13557264265479563,
+      "grad_norm": 1.2606916427612305,
+      "learning_rate": 8.820978732190791e-05,
+      "loss": 0.9757,
+      "step": 3350
+    },
+    {
+      "epoch": 0.1357749898826386,
+      "grad_norm": 1.1343713998794556,
+      "learning_rate": 8.818913896345242e-05,
+      "loss": 1.0005,
+      "step": 3355
+    },
+    {
+      "epoch": 0.1359773371104816,
+      "grad_norm": 1.2434736490249634,
+      "learning_rate": 8.816849060499691e-05,
+      "loss": 1.0394,
+      "step": 3360
+    },
+    {
+      "epoch": 0.13617968433832456,
+      "grad_norm": 1.1181955337524414,
+      "learning_rate": 8.81478422465414e-05,
+      "loss": 1.0572,
+      "step": 3365
+    },
+    {
+      "epoch": 0.13638203156616754,
+      "grad_norm": 1.2072449922561646,
+      "learning_rate": 8.81271938880859e-05,
+      "loss": 1.0625,
+      "step": 3370
+    },
+    {
+      "epoch": 0.13658437879401053,
+      "grad_norm": 1.2195651531219482,
+      "learning_rate": 8.810654552963041e-05,
+      "loss": 1.0399,
+      "step": 3375
+    },
+    {
+      "epoch": 0.1367867260218535,
+      "grad_norm": 1.175723671913147,
+      "learning_rate": 8.80858971711749e-05,
+      "loss": 0.9996,
+      "step": 3380
+    },
+    {
+      "epoch": 0.13698907324969647,
+      "grad_norm": 1.1045277118682861,
+      "learning_rate": 8.806524881271939e-05,
+      "loss": 1.0098,
+      "step": 3385
+    },
+    {
+      "epoch": 0.13719142047753946,
+      "grad_norm": 1.2739930152893066,
+      "learning_rate": 8.804460045426389e-05,
+      "loss": 0.9593,
+      "step": 3390
+    },
+    {
+      "epoch": 0.13739376770538245,
+      "grad_norm": 1.0590320825576782,
+      "learning_rate": 8.80239520958084e-05,
+      "loss": 1.0003,
+      "step": 3395
+    },
+    {
+      "epoch": 0.1375961149332254,
+      "grad_norm": 1.220510482788086,
+      "learning_rate": 8.800330373735289e-05,
+      "loss": 1.0289,
+      "step": 3400
+    },
+    {
+      "epoch": 0.1377984621610684,
+      "grad_norm": 1.1820683479309082,
+      "learning_rate": 8.798265537889738e-05,
+      "loss": 0.9708,
+      "step": 3405
+    },
+    {
+      "epoch": 0.13800080938891138,
+      "grad_norm": 1.138117790222168,
+      "learning_rate": 8.796200702044188e-05,
+      "loss": 1.0258,
+      "step": 3410
+    },
+    {
+      "epoch": 0.13820315661675436,
+      "grad_norm": 1.1486241817474365,
+      "learning_rate": 8.794135866198637e-05,
+      "loss": 1.0041,
+      "step": 3415
+    },
+    {
+      "epoch": 0.13840550384459732,
+      "grad_norm": 1.3090317249298096,
+      "learning_rate": 8.792071030353088e-05,
+      "loss": 0.9687,
+      "step": 3420
+    },
+    {
+      "epoch": 0.1386078510724403,
+      "grad_norm": 1.2961742877960205,
+      "learning_rate": 8.790006194507538e-05,
+      "loss": 1.0142,
+      "step": 3425
+    },
+    {
+      "epoch": 0.1388101983002833,
+      "grad_norm": 1.1060842275619507,
+      "learning_rate": 8.787941358661986e-05,
+      "loss": 0.988,
+      "step": 3430
+    },
+    {
+      "epoch": 0.13901254552812625,
+      "grad_norm": 1.1456315517425537,
+      "learning_rate": 8.785876522816436e-05,
+      "loss": 0.9785,
+      "step": 3435
+    },
+    {
+      "epoch": 0.13921489275596924,
+      "grad_norm": 1.2191643714904785,
+      "learning_rate": 8.783811686970887e-05,
+      "loss": 1.0835,
+      "step": 3440
+    },
+    {
+      "epoch": 0.13941723998381222,
+      "grad_norm": 1.2775850296020508,
+      "learning_rate": 8.781746851125337e-05,
+      "loss": 0.98,
+      "step": 3445
+    },
+    {
+      "epoch": 0.1396195872116552,
+      "grad_norm": 1.239769458770752,
+      "learning_rate": 8.779682015279786e-05,
+      "loss": 1.0547,
+      "step": 3450
+    },
+    {
+      "epoch": 0.13982193443949817,
+      "grad_norm": 1.193727970123291,
+      "learning_rate": 8.777617179434235e-05,
+      "loss": 0.976,
+      "step": 3455
+    },
+    {
+      "epoch": 0.14002428166734115,
+      "grad_norm": 1.2514077425003052,
+      "learning_rate": 8.775552343588685e-05,
+      "loss": 1.0088,
+      "step": 3460
+    },
+    {
+      "epoch": 0.14022662889518414,
+      "grad_norm": 1.3171583414077759,
+      "learning_rate": 8.773487507743134e-05,
+      "loss": 1.0476,
+      "step": 3465
+    },
+    {
+      "epoch": 0.14042897612302713,
+      "grad_norm": 1.1793632507324219,
+      "learning_rate": 8.771422671897585e-05,
+      "loss": 1.052,
+      "step": 3470
+    },
+    {
+      "epoch": 0.14063132335087009,
+      "grad_norm": 1.096036672592163,
+      "learning_rate": 8.769357836052034e-05,
+      "loss": 1.0444,
+      "step": 3475
+    },
+    {
+      "epoch": 0.14083367057871307,
+      "grad_norm": 1.3066802024841309,
+      "learning_rate": 8.767293000206483e-05,
+      "loss": 1.0321,
+      "step": 3480
+    },
+    {
+      "epoch": 0.14103601780655606,
+      "grad_norm": 1.175502896308899,
+      "learning_rate": 8.765228164360933e-05,
+      "loss": 1.046,
+      "step": 3485
+    },
+    {
+      "epoch": 0.14123836503439902,
+      "grad_norm": 1.2827718257904053,
+      "learning_rate": 8.763163328515384e-05,
+      "loss": 1.0152,
+      "step": 3490
+    },
+    {
+      "epoch": 0.141440712262242,
+      "grad_norm": 1.280197262763977,
+      "learning_rate": 8.761098492669834e-05,
+      "loss": 1.0324,
+      "step": 3495
+    },
+    {
+      "epoch": 0.141643059490085,
+      "grad_norm": 1.1645952463150024,
+      "learning_rate": 8.759033656824282e-05,
+      "loss": 1.0167,
+      "step": 3500
+    },
+    {
+      "epoch": 0.14184540671792797,
+      "grad_norm": 1.2604427337646484,
+      "learning_rate": 8.756968820978732e-05,
+      "loss": 1.028,
+      "step": 3505
+    },
+    {
+      "epoch": 0.14204775394577093,
+      "grad_norm": 1.3135136365890503,
+      "learning_rate": 8.754903985133183e-05,
+      "loss": 1.0121,
+      "step": 3510
+    },
+    {
+      "epoch": 0.14225010117361392,
+      "grad_norm": 1.1825295686721802,
+      "learning_rate": 8.752839149287632e-05,
+      "loss": 1.023,
+      "step": 3515
+    },
+    {
+      "epoch": 0.1424524484014569,
+      "grad_norm": 1.134616732597351,
+      "learning_rate": 8.750774313442082e-05,
+      "loss": 0.9759,
+      "step": 3520
+    },
+    {
+      "epoch": 0.1426547956292999,
+      "grad_norm": 1.1520540714263916,
+      "learning_rate": 8.748709477596531e-05,
+      "loss": 1.0127,
+      "step": 3525
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 1.1450047492980957,
+      "learning_rate": 8.746644641750982e-05,
+      "loss": 1.0345,
+      "step": 3530
+    },
+    {
+      "epoch": 0.14305949008498584,
+      "grad_norm": 1.087993860244751,
+      "learning_rate": 8.744579805905431e-05,
+      "loss": 0.968,
+      "step": 3535
+    },
+    {
+      "epoch": 0.14326183731282882,
+      "grad_norm": 1.3990801572799683,
+      "learning_rate": 8.742514970059881e-05,
+      "loss": 0.989,
+      "step": 3540
+    },
+    {
+      "epoch": 0.14346418454067178,
+      "grad_norm": 1.4547699689865112,
+      "learning_rate": 8.74045013421433e-05,
+      "loss": 1.0015,
+      "step": 3545
+    },
+    {
+      "epoch": 0.14366653176851477,
+      "grad_norm": 1.1456403732299805,
+      "learning_rate": 8.738385298368779e-05,
+      "loss": 0.9822,
+      "step": 3550
+    },
+    {
+      "epoch": 0.14386887899635775,
+      "grad_norm": 1.2532793283462524,
+      "learning_rate": 8.73632046252323e-05,
+      "loss": 1.0256,
+      "step": 3555
+    },
+    {
+      "epoch": 0.14407122622420074,
+      "grad_norm": 1.1669167280197144,
+      "learning_rate": 8.73425562667768e-05,
+      "loss": 0.9475,
+      "step": 3560
+    },
+    {
+      "epoch": 0.1442735734520437,
+      "grad_norm": 1.130604863166809,
+      "learning_rate": 8.732190790832129e-05,
+      "loss": 1.0035,
+      "step": 3565
+    },
+    {
+      "epoch": 0.14447592067988668,
+      "grad_norm": 1.2054868936538696,
+      "learning_rate": 8.73012595498658e-05,
+      "loss": 1.0052,
+      "step": 3570
+    },
+    {
+      "epoch": 0.14467826790772967,
+      "grad_norm": 1.2163805961608887,
+      "learning_rate": 8.728061119141029e-05,
+      "loss": 0.9924,
+      "step": 3575
+    },
+    {
+      "epoch": 0.14488061513557265,
+      "grad_norm": 1.3135406970977783,
+      "learning_rate": 8.725996283295479e-05,
+      "loss": 0.9828,
+      "step": 3580
+    },
+    {
+      "epoch": 0.1450829623634156,
+      "grad_norm": 1.1675292253494263,
+      "learning_rate": 8.723931447449928e-05,
+      "loss": 1.0074,
+      "step": 3585
+    },
+    {
+      "epoch": 0.1452853095912586,
+      "grad_norm": 1.3223679065704346,
+      "learning_rate": 8.721866611604378e-05,
+      "loss": 1.0411,
+      "step": 3590
+    },
+    {
+      "epoch": 0.14548765681910159,
+      "grad_norm": 1.2408627271652222,
+      "learning_rate": 8.719801775758827e-05,
+      "loss": 0.9985,
+      "step": 3595
+    },
+    {
+      "epoch": 0.14569000404694454,
+      "grad_norm": 1.055103063583374,
+      "learning_rate": 8.717736939913277e-05,
+      "loss": 1.0143,
+      "step": 3600
+    },
+    {
+      "epoch": 0.14589235127478753,
+      "grad_norm": 1.174357295036316,
+      "learning_rate": 8.715672104067727e-05,
+      "loss": 1.0361,
+      "step": 3605
+    },
+    {
+      "epoch": 0.14609469850263052,
+      "grad_norm": 1.195693850517273,
+      "learning_rate": 8.713607268222177e-05,
+      "loss": 1.0385,
+      "step": 3610
+    },
+    {
+      "epoch": 0.1462970457304735,
+      "grad_norm": 1.2959665060043335,
+      "learning_rate": 8.711542432376626e-05,
+      "loss": 1.0121,
+      "step": 3615
+    },
+    {
+      "epoch": 0.14649939295831646,
+      "grad_norm": 1.25341796875,
+      "learning_rate": 8.709477596531075e-05,
+      "loss": 1.0562,
+      "step": 3620
+    },
+    {
+      "epoch": 0.14670174018615945,
+      "grad_norm": 1.2359027862548828,
+      "learning_rate": 8.707412760685526e-05,
+      "loss": 1.0372,
+      "step": 3625
+    },
+    {
+      "epoch": 0.14690408741400243,
+      "grad_norm": 1.233233094215393,
+      "learning_rate": 8.705347924839976e-05,
+      "loss": 1.02,
+      "step": 3630
+    },
+    {
+      "epoch": 0.14710643464184542,
+      "grad_norm": 1.246022343635559,
+      "learning_rate": 8.703283088994425e-05,
+      "loss": 1.0463,
+      "step": 3635
+    },
+    {
+      "epoch": 0.14730878186968838,
+      "grad_norm": 1.2579573392868042,
+      "learning_rate": 8.701218253148876e-05,
+      "loss": 0.9662,
+      "step": 3640
+    },
+    {
+      "epoch": 0.14751112909753136,
+      "grad_norm": 1.2932707071304321,
+      "learning_rate": 8.699153417303325e-05,
+      "loss": 1.024,
+      "step": 3645
+    },
+    {
+      "epoch": 0.14771347632537435,
+      "grad_norm": 1.1545662879943848,
+      "learning_rate": 8.697088581457774e-05,
+      "loss": 1.0321,
+      "step": 3650
+    },
+    {
+      "epoch": 0.1479158235532173,
+      "grad_norm": 1.1988214254379272,
+      "learning_rate": 8.695023745612224e-05,
+      "loss": 1.0274,
+      "step": 3655
+    },
+    {
+      "epoch": 0.1481181707810603,
+      "grad_norm": 1.1396006345748901,
+      "learning_rate": 8.692958909766675e-05,
+      "loss": 1.094,
+      "step": 3660
+    },
+    {
+      "epoch": 0.14832051800890328,
+      "grad_norm": 1.1652321815490723,
+      "learning_rate": 8.690894073921124e-05,
+      "loss": 1.0602,
+      "step": 3665
+    },
+    {
+      "epoch": 0.14852286523674627,
+      "grad_norm": 1.1104350090026855,
+      "learning_rate": 8.688829238075573e-05,
+      "loss": 1.0074,
+      "step": 3670
+    },
+    {
+      "epoch": 0.14872521246458922,
+      "grad_norm": 1.164095163345337,
+      "learning_rate": 8.686764402230023e-05,
+      "loss": 0.9879,
+      "step": 3675
+    },
+    {
+      "epoch": 0.1489275596924322,
+      "grad_norm": 1.276949167251587,
+      "learning_rate": 8.684699566384474e-05,
+      "loss": 1.0931,
+      "step": 3680
+    },
+    {
+      "epoch": 0.1491299069202752,
+      "grad_norm": 1.211052417755127,
+      "learning_rate": 8.682634730538923e-05,
+      "loss": 1.0339,
+      "step": 3685
+    },
+    {
+      "epoch": 0.14933225414811818,
+      "grad_norm": 1.135284662246704,
+      "learning_rate": 8.680569894693372e-05,
+      "loss": 1.0324,
+      "step": 3690
+    },
+    {
+      "epoch": 0.14953460137596114,
+      "grad_norm": 1.1767339706420898,
+      "learning_rate": 8.678505058847822e-05,
+      "loss": 1.0568,
+      "step": 3695
+    },
+    {
+      "epoch": 0.14973694860380413,
+      "grad_norm": 1.199989676475525,
+      "learning_rate": 8.676440223002271e-05,
+      "loss": 1.0421,
+      "step": 3700
+    },
+    {
+      "epoch": 0.1499392958316471,
+      "grad_norm": 1.4178532361984253,
+      "learning_rate": 8.674375387156722e-05,
+      "loss": 1.0056,
+      "step": 3705
+    },
+    {
+      "epoch": 0.1501416430594901,
+      "grad_norm": 1.312332034111023,
+      "learning_rate": 8.672310551311172e-05,
+      "loss": 0.9942,
+      "step": 3710
+    },
+    {
+      "epoch": 0.15034399028733306,
+      "grad_norm": 1.1598105430603027,
+      "learning_rate": 8.670245715465621e-05,
+      "loss": 1.0479,
+      "step": 3715
+    },
+    {
+      "epoch": 0.15054633751517604,
+      "grad_norm": 1.192805290222168,
+      "learning_rate": 8.66818087962007e-05,
+      "loss": 1.0352,
+      "step": 3720
+    },
+    {
+      "epoch": 0.15074868474301903,
+      "grad_norm": 1.024114727973938,
+      "learning_rate": 8.66611604377452e-05,
+      "loss": 1.0282,
+      "step": 3725
+    },
+    {
+      "epoch": 0.150951031970862,
+      "grad_norm": 1.0919108390808105,
+      "learning_rate": 8.664051207928971e-05,
+      "loss": 1.0674,
+      "step": 3730
+    },
+    {
+      "epoch": 0.15115337919870497,
+      "grad_norm": 1.2635074853897095,
+      "learning_rate": 8.66198637208342e-05,
+      "loss": 1.0537,
+      "step": 3735
+    },
+    {
+      "epoch": 0.15135572642654796,
+      "grad_norm": 1.2155044078826904,
+      "learning_rate": 8.659921536237869e-05,
+      "loss": 1.0337,
+      "step": 3740
+    },
+    {
+      "epoch": 0.15155807365439095,
+      "grad_norm": 1.3911570310592651,
+      "learning_rate": 8.65785670039232e-05,
+      "loss": 0.9941,
+      "step": 3745
+    },
+    {
+      "epoch": 0.1517604208822339,
+      "grad_norm": 1.2916566133499146,
+      "learning_rate": 8.65579186454677e-05,
+      "loss": 1.0786,
+      "step": 3750
+    },
+    {
+      "epoch": 0.1519627681100769,
+      "grad_norm": 1.1407253742218018,
+      "learning_rate": 8.653727028701219e-05,
+      "loss": 0.9985,
+      "step": 3755
+    },
+    {
+      "epoch": 0.15216511533791988,
+      "grad_norm": 1.1862139701843262,
+      "learning_rate": 8.651662192855668e-05,
+      "loss": 0.9468,
+      "step": 3760
+    },
+    {
+      "epoch": 0.15236746256576286,
+      "grad_norm": 1.4419888257980347,
+      "learning_rate": 8.649597357010118e-05,
+      "loss": 1.0126,
+      "step": 3765
+    },
+    {
+      "epoch": 0.15256980979360582,
+      "grad_norm": 1.222913384437561,
+      "learning_rate": 8.647532521164567e-05,
+      "loss": 0.9966,
+      "step": 3770
+    },
+    {
+      "epoch": 0.1527721570214488,
+      "grad_norm": 1.0772240161895752,
+      "learning_rate": 8.645467685319018e-05,
+      "loss": 1.035,
+      "step": 3775
+    },
+    {
+      "epoch": 0.1529745042492918,
+      "grad_norm": 1.1999720335006714,
+      "learning_rate": 8.643402849473468e-05,
+      "loss": 0.9401,
+      "step": 3780
+    },
+    {
+      "epoch": 0.15317685147713475,
+      "grad_norm": 1.3183565139770508,
+      "learning_rate": 8.641338013627916e-05,
+      "loss": 1.0557,
+      "step": 3785
+    },
+    {
+      "epoch": 0.15337919870497774,
+      "grad_norm": 1.2679200172424316,
+      "learning_rate": 8.639273177782366e-05,
+      "loss": 0.969,
+      "step": 3790
+    },
+    {
+      "epoch": 0.15358154593282072,
+      "grad_norm": 1.2706146240234375,
+      "learning_rate": 8.637208341936817e-05,
+      "loss": 0.938,
+      "step": 3795
+    },
+    {
+      "epoch": 0.1537838931606637,
+      "grad_norm": 1.203769564628601,
+      "learning_rate": 8.635143506091267e-05,
+      "loss": 1.0065,
+      "step": 3800
+    },
+    {
+      "epoch": 0.15398624038850667,
+      "grad_norm": 1.2441059350967407,
+      "learning_rate": 8.633078670245716e-05,
+      "loss": 1.0512,
+      "step": 3805
+    },
+    {
+      "epoch": 0.15418858761634965,
+      "grad_norm": 1.2276763916015625,
+      "learning_rate": 8.631013834400165e-05,
+      "loss": 0.9929,
+      "step": 3810
+    },
+    {
+      "epoch": 0.15439093484419264,
+      "grad_norm": 1.284480094909668,
+      "learning_rate": 8.628948998554616e-05,
+      "loss": 1.0114,
+      "step": 3815
+    },
+    {
+      "epoch": 0.15459328207203563,
+      "grad_norm": 1.270969033241272,
+      "learning_rate": 8.626884162709065e-05,
+      "loss": 1.0202,
+      "step": 3820
+    },
+    {
+      "epoch": 0.15479562929987858,
+      "grad_norm": 1.3259066343307495,
+      "learning_rate": 8.624819326863515e-05,
+      "loss": 1.0032,
+      "step": 3825
+    },
+    {
+      "epoch": 0.15499797652772157,
+      "grad_norm": 1.2419929504394531,
+      "learning_rate": 8.622754491017964e-05,
+      "loss": 0.9993,
+      "step": 3830
+    },
+    {
+      "epoch": 0.15520032375556456,
+      "grad_norm": 1.2978901863098145,
+      "learning_rate": 8.620689655172413e-05,
+      "loss": 0.9806,
+      "step": 3835
+    },
+    {
+      "epoch": 0.15540267098340751,
+      "grad_norm": 1.2446649074554443,
+      "learning_rate": 8.618624819326864e-05,
+      "loss": 0.9643,
+      "step": 3840
+    },
+    {
+      "epoch": 0.1556050182112505,
+      "grad_norm": 1.1495150327682495,
+      "learning_rate": 8.616559983481314e-05,
+      "loss": 1.0339,
+      "step": 3845
+    },
+    {
+      "epoch": 0.1558073654390935,
+      "grad_norm": 1.1303049325942993,
+      "learning_rate": 8.614495147635764e-05,
+      "loss": 1.015,
+      "step": 3850
+    },
+    {
+      "epoch": 0.15600971266693647,
+      "grad_norm": 1.3211051225662231,
+      "learning_rate": 8.612430311790213e-05,
+      "loss": 1.0436,
+      "step": 3855
+    },
+    {
+      "epoch": 0.15621205989477943,
+      "grad_norm": 1.1358425617218018,
+      "learning_rate": 8.610365475944662e-05,
+      "loss": 0.9691,
+      "step": 3860
+    },
+    {
+      "epoch": 0.15641440712262242,
+      "grad_norm": 1.2107272148132324,
+      "learning_rate": 8.608300640099113e-05,
+      "loss": 0.9924,
+      "step": 3865
+    },
+    {
+      "epoch": 0.1566167543504654,
+      "grad_norm": 1.2978575229644775,
+      "learning_rate": 8.606235804253562e-05,
+      "loss": 1.0476,
+      "step": 3870
+    },
+    {
+      "epoch": 0.1568191015783084,
+      "grad_norm": 1.103603482246399,
+      "learning_rate": 8.604170968408012e-05,
+      "loss": 1.0207,
+      "step": 3875
+    },
+    {
+      "epoch": 0.15702144880615135,
+      "grad_norm": 1.2778555154800415,
+      "learning_rate": 8.602106132562461e-05,
+      "loss": 0.9794,
+      "step": 3880
+    },
+    {
+      "epoch": 0.15722379603399433,
+      "grad_norm": 1.2931251525878906,
+      "learning_rate": 8.600041296716912e-05,
+      "loss": 1.0217,
+      "step": 3885
+    },
+    {
+      "epoch": 0.15742614326183732,
+      "grad_norm": 1.2048687934875488,
+      "learning_rate": 8.597976460871361e-05,
+      "loss": 0.9122,
+      "step": 3890
+    },
+    {
+      "epoch": 0.15762849048968028,
+      "grad_norm": 1.1445924043655396,
+      "learning_rate": 8.595911625025811e-05,
+      "loss": 0.9711,
+      "step": 3895
+    },
+    {
+      "epoch": 0.15783083771752326,
+      "grad_norm": 1.1007800102233887,
+      "learning_rate": 8.593846789180262e-05,
+      "loss": 1.0013,
+      "step": 3900
+    },
+    {
+      "epoch": 0.15803318494536625,
+      "grad_norm": 1.1982959508895874,
+      "learning_rate": 8.59178195333471e-05,
+      "loss": 1.0429,
+      "step": 3905
+    },
+    {
+      "epoch": 0.15823553217320924,
+      "grad_norm": 1.206376552581787,
+      "learning_rate": 8.58971711748916e-05,
+      "loss": 1.0588,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1584378794010522,
+      "grad_norm": 1.2315871715545654,
+      "learning_rate": 8.58765228164361e-05,
+      "loss": 1.052,
+      "step": 3915
+    },
+    {
+      "epoch": 0.15864022662889518,
+      "grad_norm": 1.1181589365005493,
+      "learning_rate": 8.585587445798059e-05,
+      "loss": 1.0116,
+      "step": 3920
+    },
+    {
+      "epoch": 0.15884257385673817,
+      "grad_norm": 1.2370235919952393,
+      "learning_rate": 8.58352260995251e-05,
+      "loss": 1.0724,
+      "step": 3925
+    },
+    {
+      "epoch": 0.15904492108458115,
+      "grad_norm": 1.3594273328781128,
+      "learning_rate": 8.581457774106959e-05,
+      "loss": 1.0249,
+      "step": 3930
+    },
+    {
+      "epoch": 0.1592472683124241,
+      "grad_norm": 1.2228608131408691,
+      "learning_rate": 8.579392938261409e-05,
+      "loss": 0.9731,
+      "step": 3935
+    },
+    {
+      "epoch": 0.1594496155402671,
+      "grad_norm": 1.1914616823196411,
+      "learning_rate": 8.577328102415858e-05,
+      "loss": 1.0062,
+      "step": 3940
+    },
+    {
+      "epoch": 0.15965196276811008,
+      "grad_norm": 1.0850211381912231,
+      "learning_rate": 8.575263266570309e-05,
+      "loss": 1.007,
+      "step": 3945
+    },
+    {
+      "epoch": 0.15985430999595304,
+      "grad_norm": 1.281343698501587,
+      "learning_rate": 8.573198430724758e-05,
+      "loss": 1.0535,
+      "step": 3950
+    },
+    {
+      "epoch": 0.16005665722379603,
+      "grad_norm": 1.243461012840271,
+      "learning_rate": 8.571133594879207e-05,
+      "loss": 1.0938,
+      "step": 3955
+    },
+    {
+      "epoch": 0.16025900445163901,
+      "grad_norm": 1.1410428285598755,
+      "learning_rate": 8.569068759033657e-05,
+      "loss": 1.0383,
+      "step": 3960
+    },
+    {
+      "epoch": 0.160461351679482,
+      "grad_norm": 1.1548258066177368,
+      "learning_rate": 8.567003923188107e-05,
+      "loss": 1.065,
+      "step": 3965
+    },
+    {
+      "epoch": 0.16066369890732496,
+      "grad_norm": 1.187805414199829,
+      "learning_rate": 8.564939087342557e-05,
+      "loss": 0.9705,
+      "step": 3970
+    },
+    {
+      "epoch": 0.16086604613516794,
+      "grad_norm": 1.3056522607803345,
+      "learning_rate": 8.562874251497006e-05,
+      "loss": 1.0489,
+      "step": 3975
+    },
+    {
+      "epoch": 0.16106839336301093,
+      "grad_norm": 1.1718940734863281,
+      "learning_rate": 8.560809415651456e-05,
+      "loss": 0.9866,
+      "step": 3980
+    },
+    {
+      "epoch": 0.16127074059085392,
+      "grad_norm": 1.1825013160705566,
+      "learning_rate": 8.558744579805906e-05,
+      "loss": 1.0026,
+      "step": 3985
+    },
+    {
+      "epoch": 0.16147308781869688,
+      "grad_norm": 1.1425038576126099,
+      "learning_rate": 8.556679743960355e-05,
+      "loss": 0.9875,
+      "step": 3990
+    },
+    {
+      "epoch": 0.16167543504653986,
+      "grad_norm": 1.316321849822998,
+      "learning_rate": 8.554614908114806e-05,
+      "loss": 1.0179,
+      "step": 3995
+    },
+    {
+      "epoch": 0.16187778227438285,
+      "grad_norm": 1.2819961309432983,
+      "learning_rate": 8.552550072269255e-05,
+      "loss": 1.0207,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1620801295022258,
+      "grad_norm": 1.1831226348876953,
+      "learning_rate": 8.550485236423704e-05,
+      "loss": 1.0433,
+      "step": 4005
+    },
+    {
+      "epoch": 0.1622824767300688,
+      "grad_norm": 1.1329317092895508,
+      "learning_rate": 8.548420400578154e-05,
+      "loss": 1.0272,
+      "step": 4010
+    },
+    {
+      "epoch": 0.16248482395791178,
+      "grad_norm": 1.1299233436584473,
+      "learning_rate": 8.546355564732605e-05,
+      "loss": 1.005,
+      "step": 4015
+    },
+    {
+      "epoch": 0.16268717118575476,
+      "grad_norm": 1.5367989540100098,
+      "learning_rate": 8.544290728887054e-05,
+      "loss": 1.026,
+      "step": 4020
+    },
+    {
+      "epoch": 0.16288951841359772,
+      "grad_norm": 1.4270910024642944,
+      "learning_rate": 8.542225893041503e-05,
+      "loss": 1.0135,
+      "step": 4025
+    },
+    {
+      "epoch": 0.1630918656414407,
+      "grad_norm": 1.3334009647369385,
+      "learning_rate": 8.540161057195953e-05,
+      "loss": 1.0924,
+      "step": 4030
+    },
+    {
+      "epoch": 0.1632942128692837,
+      "grad_norm": 1.2332525253295898,
+      "learning_rate": 8.538096221350404e-05,
+      "loss": 1.0167,
+      "step": 4035
+    },
+    {
+      "epoch": 0.16349656009712668,
+      "grad_norm": 1.2154927253723145,
+      "learning_rate": 8.536031385504853e-05,
+      "loss": 0.9935,
+      "step": 4040
+    },
+    {
+      "epoch": 0.16369890732496964,
+      "grad_norm": 1.1705427169799805,
+      "learning_rate": 8.533966549659302e-05,
+      "loss": 1.0358,
+      "step": 4045
+    },
+    {
+      "epoch": 0.16390125455281263,
+      "grad_norm": 1.1485010385513306,
+      "learning_rate": 8.531901713813752e-05,
+      "loss": 0.9742,
+      "step": 4050
+    },
+    {
+      "epoch": 0.1641036017806556,
+      "grad_norm": 1.3040324449539185,
+      "learning_rate": 8.529836877968201e-05,
+      "loss": 1.0109,
+      "step": 4055
+    },
+    {
+      "epoch": 0.1643059490084986,
+      "grad_norm": 1.1587469577789307,
+      "learning_rate": 8.527772042122652e-05,
+      "loss": 1.005,
+      "step": 4060
+    },
+    {
+      "epoch": 0.16450829623634156,
+      "grad_norm": 1.1227656602859497,
+      "learning_rate": 8.525707206277102e-05,
+      "loss": 0.9587,
+      "step": 4065
+    },
+    {
+      "epoch": 0.16471064346418454,
+      "grad_norm": 1.2229607105255127,
+      "learning_rate": 8.523642370431551e-05,
+      "loss": 0.9741,
+      "step": 4070
+    },
+    {
+      "epoch": 0.16491299069202753,
+      "grad_norm": 1.2067303657531738,
+      "learning_rate": 8.521577534586e-05,
+      "loss": 0.9957,
+      "step": 4075
+    },
+    {
+      "epoch": 0.1651153379198705,
+      "grad_norm": 1.1254018545150757,
+      "learning_rate": 8.51951269874045e-05,
+      "loss": 1.0125,
+      "step": 4080
+    },
+    {
+      "epoch": 0.16531768514771347,
+      "grad_norm": 1.163508653640747,
+      "learning_rate": 8.517447862894901e-05,
+      "loss": 1.0425,
+      "step": 4085
+    },
+    {
+      "epoch": 0.16552003237555646,
+      "grad_norm": 1.2304508686065674,
+      "learning_rate": 8.51538302704935e-05,
+      "loss": 0.9819,
+      "step": 4090
+    },
+    {
+      "epoch": 0.16572237960339944,
+      "grad_norm": 1.19584059715271,
+      "learning_rate": 8.513318191203799e-05,
+      "loss": 1.008,
+      "step": 4095
+    },
+    {
+      "epoch": 0.1659247268312424,
+      "grad_norm": 1.2220029830932617,
+      "learning_rate": 8.51125335535825e-05,
+      "loss": 0.982,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1661270740590854,
+      "grad_norm": 1.150023102760315,
+      "learning_rate": 8.5091885195127e-05,
+      "loss": 1.0726,
+      "step": 4105
+    },
+    {
+      "epoch": 0.16632942128692838,
+      "grad_norm": 1.1686456203460693,
+      "learning_rate": 8.507123683667149e-05,
+      "loss": 0.9949,
+      "step": 4110
+    },
+    {
+      "epoch": 0.16653176851477136,
+      "grad_norm": 1.0913928747177124,
+      "learning_rate": 8.5050588478216e-05,
+      "loss": 1.0185,
+      "step": 4115
+    },
+    {
+      "epoch": 0.16673411574261432,
+      "grad_norm": 1.1012707948684692,
+      "learning_rate": 8.502994011976048e-05,
+      "loss": 1.0289,
+      "step": 4120
+    },
+    {
+      "epoch": 0.1669364629704573,
+      "grad_norm": 1.2062115669250488,
+      "learning_rate": 8.500929176130497e-05,
+      "loss": 1.0099,
+      "step": 4125
+    },
+    {
+      "epoch": 0.1671388101983003,
+      "grad_norm": 1.0764849185943604,
+      "learning_rate": 8.498864340284948e-05,
+      "loss": 0.9775,
+      "step": 4130
+    },
+    {
+      "epoch": 0.16734115742614325,
+      "grad_norm": 1.1648088693618774,
+      "learning_rate": 8.496799504439398e-05,
+      "loss": 1.0453,
+      "step": 4135
+    },
+    {
+      "epoch": 0.16754350465398624,
+      "grad_norm": 1.1501338481903076,
+      "learning_rate": 8.494734668593847e-05,
+      "loss": 0.9998,
+      "step": 4140
+    },
+    {
+      "epoch": 0.16774585188182922,
+      "grad_norm": 1.1651593446731567,
+      "learning_rate": 8.492669832748296e-05,
+      "loss": 1.068,
+      "step": 4145
+    },
+    {
+      "epoch": 0.1679481991096722,
+      "grad_norm": 1.299416422843933,
+      "learning_rate": 8.490604996902747e-05,
+      "loss": 1.0172,
+      "step": 4150
+    },
+    {
+      "epoch": 0.16815054633751517,
+      "grad_norm": 1.2932381629943848,
+      "learning_rate": 8.488540161057197e-05,
+      "loss": 0.9496,
+      "step": 4155
+    },
+    {
+      "epoch": 0.16835289356535815,
+      "grad_norm": 1.2833499908447266,
+      "learning_rate": 8.486475325211646e-05,
+      "loss": 1.0473,
+      "step": 4160
+    },
+    {
+      "epoch": 0.16855524079320114,
+      "grad_norm": 1.069675087928772,
+      "learning_rate": 8.484410489366095e-05,
+      "loss": 1.0372,
+      "step": 4165
+    },
+    {
+      "epoch": 0.16875758802104412,
+      "grad_norm": 1.2872824668884277,
+      "learning_rate": 8.482345653520546e-05,
+      "loss": 1.028,
+      "step": 4170
+    },
+    {
+      "epoch": 0.16895993524888708,
+      "grad_norm": 1.1362758874893188,
+      "learning_rate": 8.480280817674995e-05,
+      "loss": 1.0358,
+      "step": 4175
+    },
+    {
+      "epoch": 0.16916228247673007,
+      "grad_norm": 1.3370100259780884,
+      "learning_rate": 8.478215981829445e-05,
+      "loss": 1.002,
+      "step": 4180
+    },
+    {
+      "epoch": 0.16936462970457306,
+      "grad_norm": 1.0713083744049072,
+      "learning_rate": 8.476151145983896e-05,
+      "loss": 1.017,
+      "step": 4185
+    },
+    {
+      "epoch": 0.169566976932416,
+      "grad_norm": 1.1837053298950195,
+      "learning_rate": 8.474086310138343e-05,
+      "loss": 1.0054,
+      "step": 4190
+    },
+    {
+      "epoch": 0.169769324160259,
+      "grad_norm": 1.2177187204360962,
+      "learning_rate": 8.472021474292794e-05,
+      "loss": 0.9947,
+      "step": 4195
+    },
+    {
+      "epoch": 0.16997167138810199,
+      "grad_norm": 1.2140326499938965,
+      "learning_rate": 8.469956638447244e-05,
+      "loss": 1.1113,
+      "step": 4200
+    },
+    {
+      "epoch": 0.17017401861594497,
+      "grad_norm": 1.2045141458511353,
+      "learning_rate": 8.467891802601694e-05,
+      "loss": 1.081,
+      "step": 4205
+    },
+    {
+      "epoch": 0.17037636584378793,
+      "grad_norm": 1.1738569736480713,
+      "learning_rate": 8.465826966756144e-05,
+      "loss": 1.0536,
+      "step": 4210
+    },
+    {
+      "epoch": 0.17057871307163092,
+      "grad_norm": 1.1751322746276855,
+      "learning_rate": 8.463762130910593e-05,
+      "loss": 0.9973,
+      "step": 4215
+    },
+    {
+      "epoch": 0.1707810602994739,
+      "grad_norm": 1.2755372524261475,
+      "learning_rate": 8.461697295065043e-05,
+      "loss": 1.0359,
+      "step": 4220
+    },
+    {
+      "epoch": 0.1709834075273169,
+      "grad_norm": 1.3054158687591553,
+      "learning_rate": 8.459632459219492e-05,
+      "loss": 1.0842,
+      "step": 4225
+    },
+    {
+      "epoch": 0.17118575475515985,
+      "grad_norm": 1.2696008682250977,
+      "learning_rate": 8.457567623373942e-05,
+      "loss": 1.0586,
+      "step": 4230
+    },
+    {
+      "epoch": 0.17138810198300283,
+      "grad_norm": 1.0441927909851074,
+      "learning_rate": 8.455502787528392e-05,
+      "loss": 0.973,
+      "step": 4235
+    },
+    {
+      "epoch": 0.17159044921084582,
+      "grad_norm": 1.0617234706878662,
+      "learning_rate": 8.453437951682842e-05,
+      "loss": 1.0067,
+      "step": 4240
+    },
+    {
+      "epoch": 0.17179279643868878,
+      "grad_norm": 1.1689379215240479,
+      "learning_rate": 8.451373115837291e-05,
+      "loss": 1.0333,
+      "step": 4245
+    },
+    {
+      "epoch": 0.17199514366653176,
+      "grad_norm": 1.168121337890625,
+      "learning_rate": 8.449308279991741e-05,
+      "loss": 0.9874,
+      "step": 4250
+    },
+    {
+      "epoch": 0.17219749089437475,
+      "grad_norm": 1.1850627660751343,
+      "learning_rate": 8.447243444146192e-05,
+      "loss": 1.056,
+      "step": 4255
+    },
+    {
+      "epoch": 0.17239983812221774,
+      "grad_norm": 1.170566439628601,
+      "learning_rate": 8.44517860830064e-05,
+      "loss": 1.0082,
+      "step": 4260
+    },
+    {
+      "epoch": 0.1726021853500607,
+      "grad_norm": 1.25258207321167,
+      "learning_rate": 8.44311377245509e-05,
+      "loss": 1.0581,
+      "step": 4265
+    },
+    {
+      "epoch": 0.17280453257790368,
+      "grad_norm": 1.2613723278045654,
+      "learning_rate": 8.44104893660954e-05,
+      "loss": 0.9972,
+      "step": 4270
+    },
+    {
+      "epoch": 0.17300687980574667,
+      "grad_norm": 1.3771018981933594,
+      "learning_rate": 8.43898410076399e-05,
+      "loss": 0.9891,
+      "step": 4275
+    },
+    {
+      "epoch": 0.17320922703358965,
+      "grad_norm": 1.164170265197754,
+      "learning_rate": 8.43691926491844e-05,
+      "loss": 1.0615,
+      "step": 4280
+    },
+    {
+      "epoch": 0.1734115742614326,
+      "grad_norm": 1.2523274421691895,
+      "learning_rate": 8.434854429072889e-05,
+      "loss": 1.0216,
+      "step": 4285
+    },
+    {
+      "epoch": 0.1736139214892756,
+      "grad_norm": 1.168658971786499,
+      "learning_rate": 8.432789593227339e-05,
+      "loss": 1.0541,
+      "step": 4290
+    },
+    {
+      "epoch": 0.17381626871711858,
+      "grad_norm": 1.125502586364746,
+      "learning_rate": 8.430724757381788e-05,
+      "loss": 1.0072,
+      "step": 4295
+    },
+    {
+      "epoch": 0.17401861594496154,
+      "grad_norm": 1.0859943628311157,
+      "learning_rate": 8.428659921536239e-05,
+      "loss": 1.0341,
+      "step": 4300
+    },
+    {
+      "epoch": 0.17422096317280453,
+      "grad_norm": 1.2627450227737427,
+      "learning_rate": 8.426595085690688e-05,
+      "loss": 1.0175,
+      "step": 4305
+    },
+    {
+      "epoch": 0.1744233104006475,
+      "grad_norm": 1.1298612356185913,
+      "learning_rate": 8.424530249845137e-05,
+      "loss": 1.0278,
+      "step": 4310
+    },
+    {
+      "epoch": 0.1746256576284905,
+      "grad_norm": 1.203171968460083,
+      "learning_rate": 8.422465413999587e-05,
+      "loss": 1.0459,
+      "step": 4315
+    },
+    {
+      "epoch": 0.17482800485633346,
+      "grad_norm": 1.181216835975647,
+      "learning_rate": 8.420400578154038e-05,
+      "loss": 1.0512,
+      "step": 4320
+    },
+    {
+      "epoch": 0.17503035208417644,
+      "grad_norm": 1.2767380475997925,
+      "learning_rate": 8.418335742308488e-05,
+      "loss": 0.9839,
+      "step": 4325
+    },
+    {
+      "epoch": 0.17523269931201943,
+      "grad_norm": 1.1616220474243164,
+      "learning_rate": 8.416270906462936e-05,
+      "loss": 0.9819,
+      "step": 4330
+    },
+    {
+      "epoch": 0.17543504653986242,
+      "grad_norm": 1.036577582359314,
+      "learning_rate": 8.414206070617386e-05,
+      "loss": 1.0161,
+      "step": 4335
+    },
+    {
+      "epoch": 0.17563739376770537,
+      "grad_norm": 1.205965280532837,
+      "learning_rate": 8.412141234771837e-05,
+      "loss": 1.0092,
+      "step": 4340
+    },
+    {
+      "epoch": 0.17583974099554836,
+      "grad_norm": 1.1267623901367188,
+      "learning_rate": 8.410076398926286e-05,
+      "loss": 1.0086,
+      "step": 4345
+    },
+    {
+      "epoch": 0.17604208822339135,
+      "grad_norm": 1.0909661054611206,
+      "learning_rate": 8.408011563080736e-05,
+      "loss": 1.0171,
+      "step": 4350
+    },
+    {
+      "epoch": 0.1762444354512343,
+      "grad_norm": 1.1458765268325806,
+      "learning_rate": 8.405946727235185e-05,
+      "loss": 1.0693,
+      "step": 4355
+    },
+    {
+      "epoch": 0.1764467826790773,
+      "grad_norm": 1.093388557434082,
+      "learning_rate": 8.403881891389634e-05,
+      "loss": 0.9917,
+      "step": 4360
+    },
+    {
+      "epoch": 0.17664912990692028,
+      "grad_norm": 1.2402604818344116,
+      "learning_rate": 8.401817055544084e-05,
+      "loss": 1.0196,
+      "step": 4365
+    },
+    {
+      "epoch": 0.17685147713476326,
+      "grad_norm": 1.253222942352295,
+      "learning_rate": 8.399752219698535e-05,
+      "loss": 1.0269,
+      "step": 4370
+    },
+    {
+      "epoch": 0.17705382436260622,
+      "grad_norm": 1.1422299146652222,
+      "learning_rate": 8.397687383852984e-05,
+      "loss": 0.9778,
+      "step": 4375
+    },
+    {
+      "epoch": 0.1772561715904492,
+      "grad_norm": 1.1944977045059204,
+      "learning_rate": 8.395622548007433e-05,
+      "loss": 1.0795,
+      "step": 4380
+    },
+    {
+      "epoch": 0.1774585188182922,
+      "grad_norm": 1.1311194896697998,
+      "learning_rate": 8.393557712161883e-05,
+      "loss": 1.051,
+      "step": 4385
+    },
+    {
+      "epoch": 0.17766086604613518,
+      "grad_norm": 1.1790812015533447,
+      "learning_rate": 8.391492876316334e-05,
+      "loss": 1.0372,
+      "step": 4390
+    },
+    {
+      "epoch": 0.17786321327397814,
+      "grad_norm": 1.2022043466567993,
+      "learning_rate": 8.389428040470783e-05,
+      "loss": 0.9778,
+      "step": 4395
+    },
+    {
+      "epoch": 0.17806556050182112,
+      "grad_norm": 1.250864863395691,
+      "learning_rate": 8.387363204625233e-05,
+      "loss": 1.0344,
+      "step": 4400
+    },
+    {
+      "epoch": 0.1782679077296641,
+      "grad_norm": 1.234934687614441,
+      "learning_rate": 8.385298368779682e-05,
+      "loss": 1.0017,
+      "step": 4405
+    },
+    {
+      "epoch": 0.17847025495750707,
+      "grad_norm": 1.1351219415664673,
+      "learning_rate": 8.383233532934131e-05,
+      "loss": 1.0024,
+      "step": 4410
+    },
+    {
+      "epoch": 0.17867260218535005,
+      "grad_norm": 1.1732226610183716,
+      "learning_rate": 8.381168697088582e-05,
+      "loss": 1.0497,
+      "step": 4415
+    },
+    {
+      "epoch": 0.17887494941319304,
+      "grad_norm": 1.1598520278930664,
+      "learning_rate": 8.379103861243032e-05,
+      "loss": 1.0466,
+      "step": 4420
+    },
+    {
+      "epoch": 0.17907729664103603,
+      "grad_norm": 1.2276618480682373,
+      "learning_rate": 8.377039025397481e-05,
+      "loss": 0.9633,
+      "step": 4425
+    },
+    {
+      "epoch": 0.17927964386887899,
+      "grad_norm": 1.175768494606018,
+      "learning_rate": 8.37497418955193e-05,
+      "loss": 1.0127,
+      "step": 4430
+    },
+    {
+      "epoch": 0.17948199109672197,
+      "grad_norm": 1.171652913093567,
+      "learning_rate": 8.372909353706381e-05,
+      "loss": 0.9925,
+      "step": 4435
+    },
+    {
+      "epoch": 0.17968433832456496,
+      "grad_norm": 1.2681893110275269,
+      "learning_rate": 8.370844517860831e-05,
+      "loss": 1.0505,
+      "step": 4440
+    },
+    {
+      "epoch": 0.17988668555240794,
+      "grad_norm": 1.1942038536071777,
+      "learning_rate": 8.36877968201528e-05,
+      "loss": 0.9726,
+      "step": 4445
+    },
+    {
+      "epoch": 0.1800890327802509,
+      "grad_norm": 1.1117100715637207,
+      "learning_rate": 8.366714846169729e-05,
+      "loss": 1.0066,
+      "step": 4450
+    },
+    {
+      "epoch": 0.1802913800080939,
+      "grad_norm": 1.1762025356292725,
+      "learning_rate": 8.36465001032418e-05,
+      "loss": 1.0156,
+      "step": 4455
+    },
+    {
+      "epoch": 0.18049372723593687,
+      "grad_norm": 1.2027854919433594,
+      "learning_rate": 8.36258517447863e-05,
+      "loss": 1.011,
+      "step": 4460
+    },
+    {
+      "epoch": 0.18069607446377986,
+      "grad_norm": 1.1443296670913696,
+      "learning_rate": 8.360520338633079e-05,
+      "loss": 1.0126,
+      "step": 4465
+    },
+    {
+      "epoch": 0.18089842169162282,
+      "grad_norm": 1.1379833221435547,
+      "learning_rate": 8.35845550278753e-05,
+      "loss": 1.0172,
+      "step": 4470
+    },
+    {
+      "epoch": 0.1811007689194658,
+      "grad_norm": 1.2371028661727905,
+      "learning_rate": 8.356390666941979e-05,
+      "loss": 1.0788,
+      "step": 4475
+    },
+    {
+      "epoch": 0.1813031161473088,
+      "grad_norm": 1.101577639579773,
+      "learning_rate": 8.354325831096428e-05,
+      "loss": 1.042,
+      "step": 4480
+    },
+    {
+      "epoch": 0.18150546337515175,
+      "grad_norm": 1.0971204042434692,
+      "learning_rate": 8.352260995250878e-05,
+      "loss": 0.9927,
+      "step": 4485
+    },
+    {
+      "epoch": 0.18170781060299473,
+      "grad_norm": 1.16663658618927,
+      "learning_rate": 8.350196159405328e-05,
+      "loss": 1.0371,
+      "step": 4490
+    },
+    {
+      "epoch": 0.18191015783083772,
+      "grad_norm": 1.2800010442733765,
+      "learning_rate": 8.348131323559777e-05,
+      "loss": 1.0155,
+      "step": 4495
+    },
+    {
+      "epoch": 0.1821125050586807,
+      "grad_norm": 1.0925393104553223,
+      "learning_rate": 8.346066487714227e-05,
+      "loss": 1.013,
+      "step": 4500
+    },
+    {
+      "epoch": 0.18231485228652367,
+      "grad_norm": 1.227115273475647,
+      "learning_rate": 8.344001651868677e-05,
+      "loss": 0.9777,
+      "step": 4505
+    },
+    {
+      "epoch": 0.18251719951436665,
+      "grad_norm": 1.0858711004257202,
+      "learning_rate": 8.341936816023127e-05,
+      "loss": 1.0006,
+      "step": 4510
+    },
+    {
+      "epoch": 0.18271954674220964,
+      "grad_norm": 1.1313248872756958,
+      "learning_rate": 8.339871980177576e-05,
+      "loss": 1.0027,
+      "step": 4515
+    },
+    {
+      "epoch": 0.18292189397005262,
+      "grad_norm": 1.0812978744506836,
+      "learning_rate": 8.337807144332025e-05,
+      "loss": 0.9642,
+      "step": 4520
+    },
+    {
+      "epoch": 0.18312424119789558,
+      "grad_norm": 1.1415776014328003,
+      "learning_rate": 8.335742308486476e-05,
+      "loss": 0.9529,
+      "step": 4525
+    },
+    {
+      "epoch": 0.18332658842573857,
+      "grad_norm": 1.1783350706100464,
+      "learning_rate": 8.333677472640925e-05,
+      "loss": 1.0325,
+      "step": 4530
+    },
+    {
+      "epoch": 0.18352893565358155,
+      "grad_norm": 1.1650296449661255,
+      "learning_rate": 8.331612636795375e-05,
+      "loss": 0.9809,
+      "step": 4535
+    },
+    {
+      "epoch": 0.1837312828814245,
+      "grad_norm": 1.091321587562561,
+      "learning_rate": 8.329547800949826e-05,
+      "loss": 1.0616,
+      "step": 4540
+    },
+    {
+      "epoch": 0.1839336301092675,
+      "grad_norm": 1.1316403150558472,
+      "learning_rate": 8.327482965104273e-05,
+      "loss": 1.0276,
+      "step": 4545
+    },
+    {
+      "epoch": 0.18413597733711048,
+      "grad_norm": 1.3371331691741943,
+      "learning_rate": 8.325418129258724e-05,
+      "loss": 0.9285,
+      "step": 4550
+    },
+    {
+      "epoch": 0.18433832456495347,
+      "grad_norm": 1.232073187828064,
+      "learning_rate": 8.323353293413174e-05,
+      "loss": 1.0179,
+      "step": 4555
+    },
+    {
+      "epoch": 0.18454067179279643,
+      "grad_norm": 1.1317367553710938,
+      "learning_rate": 8.321288457567625e-05,
+      "loss": 1.0738,
+      "step": 4560
+    },
+    {
+      "epoch": 0.18474301902063942,
+      "grad_norm": 1.2579426765441895,
+      "learning_rate": 8.319223621722074e-05,
+      "loss": 0.9997,
+      "step": 4565
+    },
+    {
+      "epoch": 0.1849453662484824,
+      "grad_norm": 1.0826157331466675,
+      "learning_rate": 8.317158785876523e-05,
+      "loss": 1.034,
+      "step": 4570
+    },
+    {
+      "epoch": 0.1851477134763254,
+      "grad_norm": 1.2121678590774536,
+      "learning_rate": 8.315093950030973e-05,
+      "loss": 0.9689,
+      "step": 4575
+    },
+    {
+      "epoch": 0.18535006070416835,
+      "grad_norm": 1.1447327136993408,
+      "learning_rate": 8.313029114185422e-05,
+      "loss": 0.9275,
+      "step": 4580
+    },
+    {
+      "epoch": 0.18555240793201133,
+      "grad_norm": 1.2216501235961914,
+      "learning_rate": 8.310964278339873e-05,
+      "loss": 1.0189,
+      "step": 4585
+    },
+    {
+      "epoch": 0.18575475515985432,
+      "grad_norm": 1.1991691589355469,
+      "learning_rate": 8.308899442494322e-05,
+      "loss": 0.9912,
+      "step": 4590
+    },
+    {
+      "epoch": 0.18595710238769728,
+      "grad_norm": 1.0804492235183716,
+      "learning_rate": 8.306834606648772e-05,
+      "loss": 0.9893,
+      "step": 4595
+    },
+    {
+      "epoch": 0.18615944961554026,
+      "grad_norm": 1.140123963356018,
+      "learning_rate": 8.304769770803221e-05,
+      "loss": 1.015,
+      "step": 4600
+    },
+    {
+      "epoch": 0.18636179684338325,
+      "grad_norm": 1.0930538177490234,
+      "learning_rate": 8.302704934957672e-05,
+      "loss": 1.0898,
+      "step": 4605
+    },
+    {
+      "epoch": 0.18656414407122623,
+      "grad_norm": 1.1015198230743408,
+      "learning_rate": 8.300640099112122e-05,
+      "loss": 1.012,
+      "step": 4610
+    },
+    {
+      "epoch": 0.1867664912990692,
+      "grad_norm": 1.2269937992095947,
+      "learning_rate": 8.298575263266571e-05,
+      "loss": 1.007,
+      "step": 4615
+    },
+    {
+      "epoch": 0.18696883852691218,
+      "grad_norm": 1.1857010126113892,
+      "learning_rate": 8.29651042742102e-05,
+      "loss": 0.9999,
+      "step": 4620
+    },
+    {
+      "epoch": 0.18717118575475516,
+      "grad_norm": 1.1667230129241943,
+      "learning_rate": 8.29444559157547e-05,
+      "loss": 1.0035,
+      "step": 4625
+    },
+    {
+      "epoch": 0.18737353298259815,
+      "grad_norm": 1.1285626888275146,
+      "learning_rate": 8.29238075572992e-05,
+      "loss": 1.0846,
+      "step": 4630
+    },
+    {
+      "epoch": 0.1875758802104411,
+      "grad_norm": 1.2603074312210083,
+      "learning_rate": 8.29031591988437e-05,
+      "loss": 1.0128,
+      "step": 4635
+    },
+    {
+      "epoch": 0.1877782274382841,
+      "grad_norm": 1.1492124795913696,
+      "learning_rate": 8.288251084038819e-05,
+      "loss": 1.0123,
+      "step": 4640
+    },
+    {
+      "epoch": 0.18798057466612708,
+      "grad_norm": 1.0780872106552124,
+      "learning_rate": 8.28618624819327e-05,
+      "loss": 1.0135,
+      "step": 4645
+    },
+    {
+      "epoch": 0.18818292189397004,
+      "grad_norm": 1.1884671449661255,
+      "learning_rate": 8.284121412347718e-05,
+      "loss": 1.0077,
+      "step": 4650
+    },
+    {
+      "epoch": 0.18838526912181303,
+      "grad_norm": 1.22124445438385,
+      "learning_rate": 8.282056576502169e-05,
+      "loss": 1.0195,
+      "step": 4655
+    },
+    {
+      "epoch": 0.188587616349656,
+      "grad_norm": 1.2182954549789429,
+      "learning_rate": 8.279991740656619e-05,
+      "loss": 0.9951,
+      "step": 4660
+    },
+    {
+      "epoch": 0.188789963577499,
+      "grad_norm": 1.06657075881958,
+      "learning_rate": 8.277926904811067e-05,
+      "loss": 1.013,
+      "step": 4665
+    },
+    {
+      "epoch": 0.18899231080534196,
+      "grad_norm": 1.447752833366394,
+      "learning_rate": 8.275862068965517e-05,
+      "loss": 1.0266,
+      "step": 4670
+    },
+    {
+      "epoch": 0.18919465803318494,
+      "grad_norm": 1.1149766445159912,
+      "learning_rate": 8.273797233119968e-05,
+      "loss": 1.0416,
+      "step": 4675
+    },
+    {
+      "epoch": 0.18939700526102793,
+      "grad_norm": 1.1205918788909912,
+      "learning_rate": 8.271732397274418e-05,
+      "loss": 1.0024,
+      "step": 4680
+    },
+    {
+      "epoch": 0.18959935248887091,
+      "grad_norm": 1.1822267770767212,
+      "learning_rate": 8.269667561428867e-05,
+      "loss": 1.035,
+      "step": 4685
+    },
+    {
+      "epoch": 0.18980169971671387,
+      "grad_norm": 1.1343849897384644,
+      "learning_rate": 8.267602725583316e-05,
+      "loss": 1.0002,
+      "step": 4690
+    },
+    {
+      "epoch": 0.19000404694455686,
+      "grad_norm": 1.2238893508911133,
+      "learning_rate": 8.265537889737767e-05,
+      "loss": 1.0472,
+      "step": 4695
+    },
+    {
+      "epoch": 0.19020639417239985,
+      "grad_norm": 1.2157957553863525,
+      "learning_rate": 8.263473053892216e-05,
+      "loss": 1.0133,
+      "step": 4700
+    },
+    {
+      "epoch": 0.1904087414002428,
+      "grad_norm": 1.1864662170410156,
+      "learning_rate": 8.261408218046666e-05,
+      "loss": 1.0209,
+      "step": 4705
+    },
+    {
+      "epoch": 0.1906110886280858,
+      "grad_norm": 1.296589970588684,
+      "learning_rate": 8.259343382201115e-05,
+      "loss": 1.0054,
+      "step": 4710
+    },
+    {
+      "epoch": 0.19081343585592878,
+      "grad_norm": 1.1743710041046143,
+      "learning_rate": 8.257278546355564e-05,
+      "loss": 0.9936,
+      "step": 4715
+    },
+    {
+      "epoch": 0.19101578308377176,
+      "grad_norm": 1.2557445764541626,
+      "learning_rate": 8.255213710510015e-05,
+      "loss": 0.9909,
+      "step": 4720
+    },
+    {
+      "epoch": 0.19121813031161472,
+      "grad_norm": 1.2726306915283203,
+      "learning_rate": 8.253148874664465e-05,
+      "loss": 1.0016,
+      "step": 4725
+    },
+    {
+      "epoch": 0.1914204775394577,
+      "grad_norm": 1.1929681301116943,
+      "learning_rate": 8.251084038818915e-05,
+      "loss": 0.9949,
+      "step": 4730
+    },
+    {
+      "epoch": 0.1916228247673007,
+      "grad_norm": 1.2225488424301147,
+      "learning_rate": 8.249019202973363e-05,
+      "loss": 1.0163,
+      "step": 4735
+    },
+    {
+      "epoch": 0.19182517199514368,
+      "grad_norm": 1.4360350370407104,
+      "learning_rate": 8.246954367127814e-05,
+      "loss": 1.0541,
+      "step": 4740
+    },
+    {
+      "epoch": 0.19202751922298664,
+      "grad_norm": 1.2667735815048218,
+      "learning_rate": 8.244889531282264e-05,
+      "loss": 1.0523,
+      "step": 4745
+    },
+    {
+      "epoch": 0.19222986645082962,
+      "grad_norm": 1.2565869092941284,
+      "learning_rate": 8.242824695436713e-05,
+      "loss": 1.0108,
+      "step": 4750
+    },
+    {
+      "epoch": 0.1924322136786726,
+      "grad_norm": 1.1470930576324463,
+      "learning_rate": 8.240759859591163e-05,
+      "loss": 0.98,
+      "step": 4755
+    },
+    {
+      "epoch": 0.19263456090651557,
+      "grad_norm": 1.1337534189224243,
+      "learning_rate": 8.238695023745612e-05,
+      "loss": 0.9693,
+      "step": 4760
+    },
+    {
+      "epoch": 0.19283690813435855,
+      "grad_norm": 1.1028400659561157,
+      "learning_rate": 8.236630187900061e-05,
+      "loss": 1.0606,
+      "step": 4765
+    },
+    {
+      "epoch": 0.19303925536220154,
+      "grad_norm": 1.245178461074829,
+      "learning_rate": 8.234565352054512e-05,
+      "loss": 1.0099,
+      "step": 4770
+    },
+    {
+      "epoch": 0.19324160259004453,
+      "grad_norm": 1.2336320877075195,
+      "learning_rate": 8.232500516208962e-05,
+      "loss": 1.0134,
+      "step": 4775
+    },
+    {
+      "epoch": 0.19344394981788748,
+      "grad_norm": 1.170240879058838,
+      "learning_rate": 8.230435680363411e-05,
+      "loss": 1.0229,
+      "step": 4780
+    },
+    {
+      "epoch": 0.19364629704573047,
+      "grad_norm": 1.2225584983825684,
+      "learning_rate": 8.22837084451786e-05,
+      "loss": 1.0356,
+      "step": 4785
+    },
+    {
+      "epoch": 0.19384864427357346,
+      "grad_norm": 1.2629694938659668,
+      "learning_rate": 8.226306008672311e-05,
+      "loss": 1.0283,
+      "step": 4790
+    },
+    {
+      "epoch": 0.19405099150141644,
+      "grad_norm": 1.2679898738861084,
+      "learning_rate": 8.224241172826761e-05,
+      "loss": 1.0095,
+      "step": 4795
+    },
+    {
+      "epoch": 0.1942533387292594,
+      "grad_norm": 1.2377355098724365,
+      "learning_rate": 8.22217633698121e-05,
+      "loss": 1.0122,
+      "step": 4800
+    },
+    {
+      "epoch": 0.1944556859571024,
+      "grad_norm": 1.1721128225326538,
+      "learning_rate": 8.22011150113566e-05,
+      "loss": 1.0383,
+      "step": 4805
+    },
+    {
+      "epoch": 0.19465803318494537,
+      "grad_norm": 1.0998475551605225,
+      "learning_rate": 8.21804666529011e-05,
+      "loss": 0.9939,
+      "step": 4810
+    },
+    {
+      "epoch": 0.19486038041278833,
+      "grad_norm": 1.3133405447006226,
+      "learning_rate": 8.21598182944456e-05,
+      "loss": 0.9786,
+      "step": 4815
+    },
+    {
+      "epoch": 0.19506272764063132,
+      "grad_norm": 1.1301445960998535,
+      "learning_rate": 8.213916993599009e-05,
+      "loss": 1.0173,
+      "step": 4820
+    },
+    {
+      "epoch": 0.1952650748684743,
+      "grad_norm": 1.3158178329467773,
+      "learning_rate": 8.21185215775346e-05,
+      "loss": 1.016,
+      "step": 4825
+    },
+    {
+      "epoch": 0.1954674220963173,
+      "grad_norm": 1.1632481813430786,
+      "learning_rate": 8.209787321907909e-05,
+      "loss": 1.0331,
+      "step": 4830
+    },
+    {
+      "epoch": 0.19566976932416025,
+      "grad_norm": 1.1861474514007568,
+      "learning_rate": 8.207722486062358e-05,
+      "loss": 1.0242,
+      "step": 4835
+    },
+    {
+      "epoch": 0.19587211655200323,
+      "grad_norm": 1.1996656656265259,
+      "learning_rate": 8.205657650216808e-05,
+      "loss": 1.0216,
+      "step": 4840
+    },
+    {
+      "epoch": 0.19607446377984622,
+      "grad_norm": 1.0401570796966553,
+      "learning_rate": 8.203592814371259e-05,
+      "loss": 1.0539,
+      "step": 4845
+    },
+    {
+      "epoch": 0.1962768110076892,
+      "grad_norm": 1.1545593738555908,
+      "learning_rate": 8.201527978525708e-05,
+      "loss": 1.0752,
+      "step": 4850
+    },
+    {
+      "epoch": 0.19647915823553216,
+      "grad_norm": 1.1579617261886597,
+      "learning_rate": 8.199463142680157e-05,
+      "loss": 0.9972,
+      "step": 4855
+    },
+    {
+      "epoch": 0.19668150546337515,
+      "grad_norm": 1.154383897781372,
+      "learning_rate": 8.197398306834607e-05,
+      "loss": 1.0125,
+      "step": 4860
+    },
+    {
+      "epoch": 0.19688385269121814,
+      "grad_norm": 1.2943758964538574,
+      "learning_rate": 8.195333470989057e-05,
+      "loss": 1.0772,
+      "step": 4865
+    },
+    {
+      "epoch": 0.19708619991906112,
+      "grad_norm": 1.2109222412109375,
+      "learning_rate": 8.193268635143506e-05,
+      "loss": 0.9919,
+      "step": 4870
+    },
+    {
+      "epoch": 0.19728854714690408,
+      "grad_norm": 1.1765272617340088,
+      "learning_rate": 8.191203799297956e-05,
+      "loss": 1.0669,
+      "step": 4875
+    },
+    {
+      "epoch": 0.19749089437474707,
+      "grad_norm": 1.085466980934143,
+      "learning_rate": 8.189138963452406e-05,
+      "loss": 1.0174,
+      "step": 4880
+    },
+    {
+      "epoch": 0.19769324160259005,
+      "grad_norm": 1.313828468322754,
+      "learning_rate": 8.187074127606855e-05,
+      "loss": 0.9742,
+      "step": 4885
+    },
+    {
+      "epoch": 0.197895588830433,
+      "grad_norm": 1.1793757677078247,
+      "learning_rate": 8.185009291761305e-05,
+      "loss": 1.0921,
+      "step": 4890
+    },
+    {
+      "epoch": 0.198097936058276,
+      "grad_norm": 1.1285810470581055,
+      "learning_rate": 8.182944455915756e-05,
+      "loss": 0.9753,
+      "step": 4895
+    },
+    {
+      "epoch": 0.19830028328611898,
+      "grad_norm": 1.2235735654830933,
+      "learning_rate": 8.180879620070205e-05,
+      "loss": 1.0708,
+      "step": 4900
+    },
+    {
+      "epoch": 0.19850263051396197,
+      "grad_norm": 1.3059691190719604,
+      "learning_rate": 8.178814784224654e-05,
+      "loss": 0.9947,
+      "step": 4905
+    },
+    {
+      "epoch": 0.19870497774180493,
+      "grad_norm": 1.2152923345565796,
+      "learning_rate": 8.176749948379104e-05,
+      "loss": 0.9984,
+      "step": 4910
+    },
+    {
+      "epoch": 0.19890732496964791,
+      "grad_norm": 1.2941012382507324,
+      "learning_rate": 8.174685112533555e-05,
+      "loss": 1.0011,
+      "step": 4915
+    },
+    {
+      "epoch": 0.1991096721974909,
+      "grad_norm": 1.0703668594360352,
+      "learning_rate": 8.172620276688004e-05,
+      "loss": 1.0087,
+      "step": 4920
+    },
+    {
+      "epoch": 0.1993120194253339,
+      "grad_norm": 1.228332757949829,
+      "learning_rate": 8.170555440842453e-05,
+      "loss": 0.9731,
+      "step": 4925
+    },
+    {
+      "epoch": 0.19951436665317684,
+      "grad_norm": 1.259106993675232,
+      "learning_rate": 8.168490604996903e-05,
+      "loss": 0.9851,
+      "step": 4930
+    },
+    {
+      "epoch": 0.19971671388101983,
+      "grad_norm": 1.2263469696044922,
+      "learning_rate": 8.166425769151352e-05,
+      "loss": 0.9997,
+      "step": 4935
+    },
+    {
+      "epoch": 0.19991906110886282,
+      "grad_norm": 1.289626121520996,
+      "learning_rate": 8.164360933305803e-05,
+      "loss": 1.0157,
+      "step": 4940
+    },
+    {
+      "epoch": 0.20012140833670577,
+      "grad_norm": 1.1667588949203491,
+      "learning_rate": 8.162296097460253e-05,
+      "loss": 1.0841,
+      "step": 4945
+    },
+    {
+      "epoch": 0.20032375556454876,
+      "grad_norm": 1.2754533290863037,
+      "learning_rate": 8.160231261614702e-05,
+      "loss": 1.042,
+      "step": 4950
+    },
+    {
+      "epoch": 0.20052610279239175,
+      "grad_norm": 1.2538858652114868,
+      "learning_rate": 8.158166425769151e-05,
+      "loss": 0.9863,
+      "step": 4955
+    },
+    {
+      "epoch": 0.20072845002023473,
+      "grad_norm": 1.2489347457885742,
+      "learning_rate": 8.156101589923602e-05,
+      "loss": 0.99,
+      "step": 4960
+    },
+    {
+      "epoch": 0.2009307972480777,
+      "grad_norm": 1.3711369037628174,
+      "learning_rate": 8.154036754078052e-05,
+      "loss": 0.9796,
+      "step": 4965
+    },
+    {
+      "epoch": 0.20113314447592068,
+      "grad_norm": 1.1020487546920776,
+      "learning_rate": 8.151971918232501e-05,
+      "loss": 1.0118,
+      "step": 4970
+    },
+    {
+      "epoch": 0.20133549170376366,
+      "grad_norm": 1.0652621984481812,
+      "learning_rate": 8.14990708238695e-05,
+      "loss": 0.9861,
+      "step": 4975
+    },
+    {
+      "epoch": 0.20153783893160665,
+      "grad_norm": 1.2487317323684692,
+      "learning_rate": 8.1478422465414e-05,
+      "loss": 1.0211,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2017401861594496,
+      "grad_norm": 1.0528713464736938,
+      "learning_rate": 8.14577741069585e-05,
+      "loss": 1.0425,
+      "step": 4985
+    },
+    {
+      "epoch": 0.2019425333872926,
+      "grad_norm": 1.261579990386963,
+      "learning_rate": 8.1437125748503e-05,
+      "loss": 1.0487,
+      "step": 4990
+    },
+    {
+      "epoch": 0.20214488061513558,
+      "grad_norm": 1.2511799335479736,
+      "learning_rate": 8.141647739004749e-05,
+      "loss": 1.0015,
+      "step": 4995
+    },
+    {
+      "epoch": 0.20234722784297854,
+      "grad_norm": 1.2175990343093872,
+      "learning_rate": 8.1395829031592e-05,
+      "loss": 1.0144,
+      "step": 5000
+    },
+    {
+      "epoch": 0.20254957507082152,
+      "grad_norm": 1.232932448387146,
+      "learning_rate": 8.137518067313649e-05,
+      "loss": 1.0212,
+      "step": 5005
+    },
+    {
+      "epoch": 0.2027519222986645,
+      "grad_norm": 1.2393242120742798,
+      "learning_rate": 8.135453231468099e-05,
+      "loss": 1.0109,
+      "step": 5010
+    },
+    {
+      "epoch": 0.2029542695265075,
+      "grad_norm": 1.2997404336929321,
+      "learning_rate": 8.13338839562255e-05,
+      "loss": 1.0225,
+      "step": 5015
+    },
+    {
+      "epoch": 0.20315661675435046,
+      "grad_norm": 1.0594301223754883,
+      "learning_rate": 8.131323559776997e-05,
+      "loss": 1.0358,
+      "step": 5020
+    },
+    {
+      "epoch": 0.20335896398219344,
+      "grad_norm": 1.193315863609314,
+      "learning_rate": 8.129258723931447e-05,
+      "loss": 0.9909,
+      "step": 5025
+    },
+    {
+      "epoch": 0.20356131121003643,
+      "grad_norm": 1.2019113302230835,
+      "learning_rate": 8.127193888085898e-05,
+      "loss": 1.0085,
+      "step": 5030
+    },
+    {
+      "epoch": 0.2037636584378794,
+      "grad_norm": 1.2004979848861694,
+      "learning_rate": 8.125129052240348e-05,
+      "loss": 1.0471,
+      "step": 5035
+    },
+    {
+      "epoch": 0.20396600566572237,
+      "grad_norm": 1.210090160369873,
+      "learning_rate": 8.123064216394797e-05,
+      "loss": 0.9805,
+      "step": 5040
+    },
+    {
+      "epoch": 0.20416835289356536,
+      "grad_norm": 1.1681770086288452,
+      "learning_rate": 8.120999380549246e-05,
+      "loss": 1.0305,
+      "step": 5045
+    },
+    {
+      "epoch": 0.20437070012140834,
+      "grad_norm": 1.2289959192276,
+      "learning_rate": 8.118934544703697e-05,
+      "loss": 1.037,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2045730473492513,
+      "grad_norm": 1.1978446245193481,
+      "learning_rate": 8.116869708858146e-05,
+      "loss": 0.994,
+      "step": 5055
+    },
+    {
+      "epoch": 0.2047753945770943,
+      "grad_norm": 1.1448619365692139,
+      "learning_rate": 8.114804873012596e-05,
+      "loss": 1.0227,
+      "step": 5060
+    },
+    {
+      "epoch": 0.20497774180493727,
+      "grad_norm": 1.124419927597046,
+      "learning_rate": 8.112740037167045e-05,
+      "loss": 0.9965,
+      "step": 5065
+    },
+    {
+      "epoch": 0.20518008903278026,
+      "grad_norm": 1.0857465267181396,
+      "learning_rate": 8.110675201321494e-05,
+      "loss": 1.0547,
+      "step": 5070
+    },
+    {
+      "epoch": 0.20538243626062322,
+      "grad_norm": 1.1845394372940063,
+      "learning_rate": 8.108610365475945e-05,
+      "loss": 1.1202,
+      "step": 5075
+    },
+    {
+      "epoch": 0.2055847834884662,
+      "grad_norm": 1.098524570465088,
+      "learning_rate": 8.106545529630395e-05,
+      "loss": 1.0128,
+      "step": 5080
+    },
+    {
+      "epoch": 0.2057871307163092,
+      "grad_norm": 1.2847548723220825,
+      "learning_rate": 8.104480693784846e-05,
+      "loss": 1.0487,
+      "step": 5085
+    },
+    {
+      "epoch": 0.20598947794415218,
+      "grad_norm": 1.1287879943847656,
+      "learning_rate": 8.102415857939293e-05,
+      "loss": 0.9686,
+      "step": 5090
+    },
+    {
+      "epoch": 0.20619182517199514,
+      "grad_norm": 1.1327563524246216,
+      "learning_rate": 8.100351022093744e-05,
+      "loss": 1.0362,
+      "step": 5095
+    },
+    {
+      "epoch": 0.20639417239983812,
+      "grad_norm": 1.1193857192993164,
+      "learning_rate": 8.098286186248194e-05,
+      "loss": 1.0032,
+      "step": 5100
+    },
+    {
+      "epoch": 0.2065965196276811,
+      "grad_norm": 1.1401958465576172,
+      "learning_rate": 8.096221350402643e-05,
+      "loss": 1.0223,
+      "step": 5105
+    },
+    {
+      "epoch": 0.20679886685552407,
+      "grad_norm": 1.1872715950012207,
+      "learning_rate": 8.094156514557094e-05,
+      "loss": 0.9417,
+      "step": 5110
+    },
+    {
+      "epoch": 0.20700121408336705,
+      "grad_norm": 1.2122455835342407,
+      "learning_rate": 8.092091678711543e-05,
+      "loss": 1.0725,
+      "step": 5115
+    },
+    {
+      "epoch": 0.20720356131121004,
+      "grad_norm": 1.1419376134872437,
+      "learning_rate": 8.090026842865992e-05,
+      "loss": 1.0154,
+      "step": 5120
+    },
+    {
+      "epoch": 0.20740590853905302,
+      "grad_norm": 1.2034239768981934,
+      "learning_rate": 8.087962007020442e-05,
+      "loss": 1.0571,
+      "step": 5125
+    },
+    {
+      "epoch": 0.20760825576689598,
+      "grad_norm": 1.117256760597229,
+      "learning_rate": 8.085897171174892e-05,
+      "loss": 0.9774,
+      "step": 5130
+    },
+    {
+      "epoch": 0.20781060299473897,
+      "grad_norm": 1.1485257148742676,
+      "learning_rate": 8.083832335329341e-05,
+      "loss": 1.0187,
+      "step": 5135
+    },
+    {
+      "epoch": 0.20801295022258195,
+      "grad_norm": 1.197513222694397,
+      "learning_rate": 8.08176749948379e-05,
+      "loss": 1.049,
+      "step": 5140
+    },
+    {
+      "epoch": 0.20821529745042494,
+      "grad_norm": 1.0740694999694824,
+      "learning_rate": 8.079702663638241e-05,
+      "loss": 1.0412,
+      "step": 5145
+    },
+    {
+      "epoch": 0.2084176446782679,
+      "grad_norm": 1.2320221662521362,
+      "learning_rate": 8.077637827792691e-05,
+      "loss": 1.0407,
+      "step": 5150
+    },
+    {
+      "epoch": 0.20861999190611089,
+      "grad_norm": 1.1767700910568237,
+      "learning_rate": 8.07557299194714e-05,
+      "loss": 0.9273,
+      "step": 5155
+    },
+    {
+      "epoch": 0.20882233913395387,
+      "grad_norm": 1.3007943630218506,
+      "learning_rate": 8.073508156101591e-05,
+      "loss": 1.0011,
+      "step": 5160
+    },
+    {
+      "epoch": 0.20902468636179683,
+      "grad_norm": 1.441924810409546,
+      "learning_rate": 8.07144332025604e-05,
+      "loss": 0.9932,
+      "step": 5165
+    },
+    {
+      "epoch": 0.20922703358963982,
+      "grad_norm": 1.1403335332870483,
+      "learning_rate": 8.06937848441049e-05,
+      "loss": 1.0037,
+      "step": 5170
+    },
+    {
+      "epoch": 0.2094293808174828,
+      "grad_norm": 1.2318395376205444,
+      "learning_rate": 8.06731364856494e-05,
+      "loss": 0.9821,
+      "step": 5175
+    },
+    {
+      "epoch": 0.2096317280453258,
+      "grad_norm": 1.3537083864212036,
+      "learning_rate": 8.06524881271939e-05,
+      "loss": 0.9923,
+      "step": 5180
+    },
+    {
+      "epoch": 0.20983407527316875,
+      "grad_norm": 1.2805323600769043,
+      "learning_rate": 8.063183976873839e-05,
+      "loss": 1.0084,
+      "step": 5185
+    },
+    {
+      "epoch": 0.21003642250101173,
+      "grad_norm": 1.1598711013793945,
+      "learning_rate": 8.061119141028288e-05,
+      "loss": 0.9564,
+      "step": 5190
+    },
+    {
+      "epoch": 0.21023876972885472,
+      "grad_norm": 1.1318501234054565,
+      "learning_rate": 8.059054305182738e-05,
+      "loss": 0.9757,
+      "step": 5195
+    },
+    {
+      "epoch": 0.2104411169566977,
+      "grad_norm": 1.1981204748153687,
+      "learning_rate": 8.056989469337189e-05,
+      "loss": 1.0163,
+      "step": 5200
+    },
+    {
+      "epoch": 0.21064346418454066,
+      "grad_norm": 1.2053571939468384,
+      "learning_rate": 8.054924633491638e-05,
+      "loss": 0.9828,
+      "step": 5205
+    },
+    {
+      "epoch": 0.21084581141238365,
+      "grad_norm": 1.0915008783340454,
+      "learning_rate": 8.052859797646087e-05,
+      "loss": 1.0114,
+      "step": 5210
+    },
+    {
+      "epoch": 0.21104815864022664,
+      "grad_norm": 1.2757831811904907,
+      "learning_rate": 8.050794961800537e-05,
+      "loss": 1.0101,
+      "step": 5215
+    },
+    {
+      "epoch": 0.21125050586806962,
+      "grad_norm": 1.116852879524231,
+      "learning_rate": 8.048730125954988e-05,
+      "loss": 1.0005,
+      "step": 5220
+    },
+    {
+      "epoch": 0.21145285309591258,
+      "grad_norm": 1.1757320165634155,
+      "learning_rate": 8.046665290109437e-05,
+      "loss": 1.0358,
+      "step": 5225
+    },
+    {
+      "epoch": 0.21165520032375557,
+      "grad_norm": 1.096636414527893,
+      "learning_rate": 8.044600454263887e-05,
+      "loss": 1.026,
+      "step": 5230
+    },
+    {
+      "epoch": 0.21185754755159855,
+      "grad_norm": 1.0272443294525146,
+      "learning_rate": 8.042535618418336e-05,
+      "loss": 1.0651,
+      "step": 5235
+    },
+    {
+      "epoch": 0.2120598947794415,
+      "grad_norm": 1.132805585861206,
+      "learning_rate": 8.040470782572785e-05,
+      "loss": 1.0081,
+      "step": 5240
+    },
+    {
+      "epoch": 0.2122622420072845,
+      "grad_norm": 1.165757656097412,
+      "learning_rate": 8.038405946727236e-05,
+      "loss": 1.0167,
+      "step": 5245
+    },
+    {
+      "epoch": 0.21246458923512748,
+      "grad_norm": 1.3743444681167603,
+      "learning_rate": 8.036341110881686e-05,
+      "loss": 1.0315,
+      "step": 5250
+    },
+    {
+      "epoch": 0.21266693646297047,
+      "grad_norm": 1.0902739763259888,
+      "learning_rate": 8.034276275036135e-05,
+      "loss": 1.0458,
+      "step": 5255
+    },
+    {
+      "epoch": 0.21286928369081343,
+      "grad_norm": 1.3256388902664185,
+      "learning_rate": 8.032211439190584e-05,
+      "loss": 1.0317,
+      "step": 5260
+    },
+    {
+      "epoch": 0.2130716309186564,
+      "grad_norm": 1.2284111976623535,
+      "learning_rate": 8.030146603345034e-05,
+      "loss": 1.0429,
+      "step": 5265
+    },
+    {
+      "epoch": 0.2132739781464994,
+      "grad_norm": 1.137561559677124,
+      "learning_rate": 8.028081767499485e-05,
+      "loss": 0.9972,
+      "step": 5270
+    },
+    {
+      "epoch": 0.21347632537434238,
+      "grad_norm": 1.2012466192245483,
+      "learning_rate": 8.026016931653934e-05,
+      "loss": 1.0394,
+      "step": 5275
+    },
+    {
+      "epoch": 0.21367867260218534,
+      "grad_norm": 1.210174798965454,
+      "learning_rate": 8.023952095808383e-05,
+      "loss": 1.0,
+      "step": 5280
+    },
+    {
+      "epoch": 0.21388101983002833,
+      "grad_norm": 1.2262771129608154,
+      "learning_rate": 8.021887259962833e-05,
+      "loss": 1.0235,
+      "step": 5285
+    },
+    {
+      "epoch": 0.21408336705787132,
+      "grad_norm": 1.1709131002426147,
+      "learning_rate": 8.019822424117282e-05,
+      "loss": 0.984,
+      "step": 5290
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 1.3027403354644775,
+      "learning_rate": 8.017757588271733e-05,
+      "loss": 1.0056,
+      "step": 5295
+    },
+    {
+      "epoch": 0.21448806151355726,
+      "grad_norm": 1.119411826133728,
+      "learning_rate": 8.015692752426183e-05,
+      "loss": 1.0415,
+      "step": 5300
+    },
+    {
+      "epoch": 0.21469040874140025,
+      "grad_norm": 1.345045804977417,
+      "learning_rate": 8.013627916580632e-05,
+      "loss": 1.0551,
+      "step": 5305
+    },
+    {
+      "epoch": 0.21489275596924323,
+      "grad_norm": 1.1461879014968872,
+      "learning_rate": 8.011563080735081e-05,
+      "loss": 1.0299,
+      "step": 5310
+    },
+    {
+      "epoch": 0.2150951031970862,
+      "grad_norm": 1.107438564300537,
+      "learning_rate": 8.009498244889532e-05,
+      "loss": 1.0377,
+      "step": 5315
+    },
+    {
+      "epoch": 0.21529745042492918,
+      "grad_norm": 1.1863116025924683,
+      "learning_rate": 8.007433409043982e-05,
+      "loss": 0.9396,
+      "step": 5320
+    },
+    {
+      "epoch": 0.21549979765277216,
+      "grad_norm": 1.131195306777954,
+      "learning_rate": 8.005368573198431e-05,
+      "loss": 1.0107,
+      "step": 5325
+    },
+    {
+      "epoch": 0.21570214488061515,
+      "grad_norm": 1.1013654470443726,
+      "learning_rate": 8.00330373735288e-05,
+      "loss": 1.0834,
+      "step": 5330
+    },
+    {
+      "epoch": 0.2159044921084581,
+      "grad_norm": 1.3237591981887817,
+      "learning_rate": 8.001238901507331e-05,
+      "loss": 0.9525,
+      "step": 5335
+    },
+    {
+      "epoch": 0.2161068393363011,
+      "grad_norm": 1.237547755241394,
+      "learning_rate": 7.99917406566178e-05,
+      "loss": 1.0534,
+      "step": 5340
+    },
+    {
+      "epoch": 0.21630918656414408,
+      "grad_norm": 1.21968674659729,
+      "learning_rate": 7.99710922981623e-05,
+      "loss": 1.0109,
+      "step": 5345
+    },
+    {
+      "epoch": 0.21651153379198704,
+      "grad_norm": 1.2342300415039062,
+      "learning_rate": 7.995044393970679e-05,
+      "loss": 1.009,
+      "step": 5350
+    },
+    {
+      "epoch": 0.21671388101983002,
+      "grad_norm": 1.2325963973999023,
+      "learning_rate": 7.99297955812513e-05,
+      "loss": 1.0736,
+      "step": 5355
+    },
+    {
+      "epoch": 0.216916228247673,
+      "grad_norm": 1.1565101146697998,
+      "learning_rate": 7.990914722279579e-05,
+      "loss": 0.9942,
+      "step": 5360
+    },
+    {
+      "epoch": 0.217118575475516,
+      "grad_norm": 1.2055383920669556,
+      "learning_rate": 7.988849886434029e-05,
+      "loss": 1.0574,
+      "step": 5365
+    },
+    {
+      "epoch": 0.21732092270335895,
+      "grad_norm": 1.2551612854003906,
+      "learning_rate": 7.98678505058848e-05,
+      "loss": 1.0091,
+      "step": 5370
+    },
+    {
+      "epoch": 0.21752326993120194,
+      "grad_norm": 1.1769312620162964,
+      "learning_rate": 7.984720214742929e-05,
+      "loss": 1.0012,
+      "step": 5375
+    },
+    {
+      "epoch": 0.21772561715904493,
+      "grad_norm": 1.2143961191177368,
+      "learning_rate": 7.982655378897378e-05,
+      "loss": 1.057,
+      "step": 5380
+    },
+    {
+      "epoch": 0.2179279643868879,
+      "grad_norm": 1.179459571838379,
+      "learning_rate": 7.980590543051828e-05,
+      "loss": 0.9842,
+      "step": 5385
+    },
+    {
+      "epoch": 0.21813031161473087,
+      "grad_norm": 1.2663486003875732,
+      "learning_rate": 7.978525707206278e-05,
+      "loss": 0.9715,
+      "step": 5390
+    },
+    {
+      "epoch": 0.21833265884257386,
+      "grad_norm": 1.3901774883270264,
+      "learning_rate": 7.976460871360727e-05,
+      "loss": 1.0165,
+      "step": 5395
+    },
+    {
+      "epoch": 0.21853500607041684,
+      "grad_norm": 1.1746774911880493,
+      "learning_rate": 7.974396035515176e-05,
+      "loss": 1.0111,
+      "step": 5400
+    },
+    {
+      "epoch": 0.2187373532982598,
+      "grad_norm": 1.140589952468872,
+      "learning_rate": 7.972331199669627e-05,
+      "loss": 1.0013,
+      "step": 5405
+    },
+    {
+      "epoch": 0.2189397005261028,
+      "grad_norm": 1.3139210939407349,
+      "learning_rate": 7.970266363824076e-05,
+      "loss": 1.0556,
+      "step": 5410
+    },
+    {
+      "epoch": 0.21914204775394577,
+      "grad_norm": 0.9810004830360413,
+      "learning_rate": 7.968201527978526e-05,
+      "loss": 1.0059,
+      "step": 5415
+    },
+    {
+      "epoch": 0.21934439498178876,
+      "grad_norm": 1.1734237670898438,
+      "learning_rate": 7.966136692132977e-05,
+      "loss": 1.048,
+      "step": 5420
+    },
+    {
+      "epoch": 0.21954674220963172,
+      "grad_norm": 1.1313258409500122,
+      "learning_rate": 7.964071856287424e-05,
+      "loss": 0.93,
+      "step": 5425
+    },
+    {
+      "epoch": 0.2197490894374747,
+      "grad_norm": 1.2841383218765259,
+      "learning_rate": 7.962007020441875e-05,
+      "loss": 1.0334,
+      "step": 5430
+    },
+    {
+      "epoch": 0.2199514366653177,
+      "grad_norm": 1.191046118736267,
+      "learning_rate": 7.959942184596325e-05,
+      "loss": 1.0424,
+      "step": 5435
+    },
+    {
+      "epoch": 0.22015378389316068,
+      "grad_norm": 1.2136199474334717,
+      "learning_rate": 7.957877348750776e-05,
+      "loss": 1.0314,
+      "step": 5440
+    },
+    {
+      "epoch": 0.22035613112100363,
+      "grad_norm": 1.2416642904281616,
+      "learning_rate": 7.955812512905225e-05,
+      "loss": 1.0247,
+      "step": 5445
+    },
+    {
+      "epoch": 0.22055847834884662,
+      "grad_norm": 1.175480604171753,
+      "learning_rate": 7.953747677059674e-05,
+      "loss": 1.0402,
+      "step": 5450
+    },
+    {
+      "epoch": 0.2207608255766896,
+      "grad_norm": 1.2230721712112427,
+      "learning_rate": 7.951682841214124e-05,
+      "loss": 1.0268,
+      "step": 5455
+    },
+    {
+      "epoch": 0.22096317280453256,
+      "grad_norm": 1.1856865882873535,
+      "learning_rate": 7.949618005368573e-05,
+      "loss": 0.9919,
+      "step": 5460
+    },
+    {
+      "epoch": 0.22116552003237555,
+      "grad_norm": 1.1516177654266357,
+      "learning_rate": 7.947553169523024e-05,
+      "loss": 1.0664,
+      "step": 5465
+    },
+    {
+      "epoch": 0.22136786726021854,
+      "grad_norm": 1.0597288608551025,
+      "learning_rate": 7.945488333677473e-05,
+      "loss": 0.9861,
+      "step": 5470
+    },
+    {
+      "epoch": 0.22157021448806152,
+      "grad_norm": 1.228318691253662,
+      "learning_rate": 7.943423497831922e-05,
+      "loss": 0.9624,
+      "step": 5475
+    },
+    {
+      "epoch": 0.22177256171590448,
+      "grad_norm": 1.11944580078125,
+      "learning_rate": 7.941358661986372e-05,
+      "loss": 1.0034,
+      "step": 5480
+    },
+    {
+      "epoch": 0.22197490894374747,
+      "grad_norm": 1.184719443321228,
+      "learning_rate": 7.939293826140823e-05,
+      "loss": 0.9928,
+      "step": 5485
+    },
+    {
+      "epoch": 0.22217725617159045,
+      "grad_norm": 1.2186683416366577,
+      "learning_rate": 7.937228990295273e-05,
+      "loss": 1.0219,
+      "step": 5490
+    },
+    {
+      "epoch": 0.22237960339943344,
+      "grad_norm": 1.0990355014801025,
+      "learning_rate": 7.93516415444972e-05,
+      "loss": 1.0182,
+      "step": 5495
+    },
+    {
+      "epoch": 0.2225819506272764,
+      "grad_norm": 1.1261568069458008,
+      "learning_rate": 7.933099318604171e-05,
+      "loss": 0.9818,
+      "step": 5500
+    },
+    {
+      "epoch": 0.22278429785511938,
+      "grad_norm": 1.1164358854293823,
+      "learning_rate": 7.931034482758621e-05,
+      "loss": 1.0395,
+      "step": 5505
+    },
+    {
+      "epoch": 0.22298664508296237,
+      "grad_norm": 1.0908489227294922,
+      "learning_rate": 7.92896964691307e-05,
+      "loss": 0.9929,
+      "step": 5510
+    },
+    {
+      "epoch": 0.22318899231080533,
+      "grad_norm": 1.1659622192382812,
+      "learning_rate": 7.926904811067521e-05,
+      "loss": 0.9867,
+      "step": 5515
+    },
+    {
+      "epoch": 0.22339133953864831,
+      "grad_norm": 1.1731120347976685,
+      "learning_rate": 7.92483997522197e-05,
+      "loss": 1.0232,
+      "step": 5520
+    },
+    {
+      "epoch": 0.2235936867664913,
+      "grad_norm": 1.2476763725280762,
+      "learning_rate": 7.92277513937642e-05,
+      "loss": 1.0217,
+      "step": 5525
+    },
+    {
+      "epoch": 0.2237960339943343,
+      "grad_norm": 1.1859256029129028,
+      "learning_rate": 7.92071030353087e-05,
+      "loss": 1.0695,
+      "step": 5530
+    },
+    {
+      "epoch": 0.22399838122217725,
+      "grad_norm": 1.3493760824203491,
+      "learning_rate": 7.91864546768532e-05,
+      "loss": 0.9726,
+      "step": 5535
+    },
+    {
+      "epoch": 0.22420072845002023,
+      "grad_norm": 1.1764709949493408,
+      "learning_rate": 7.916580631839769e-05,
+      "loss": 1.0557,
+      "step": 5540
+    },
+    {
+      "epoch": 0.22440307567786322,
+      "grad_norm": 1.2197325229644775,
+      "learning_rate": 7.914515795994218e-05,
+      "loss": 0.937,
+      "step": 5545
+    },
+    {
+      "epoch": 0.2246054229057062,
+      "grad_norm": 1.1879414319992065,
+      "learning_rate": 7.912450960148668e-05,
+      "loss": 1.0306,
+      "step": 5550
+    },
+    {
+      "epoch": 0.22480777013354916,
+      "grad_norm": 1.2065080404281616,
+      "learning_rate": 7.910386124303119e-05,
+      "loss": 1.0551,
+      "step": 5555
+    },
+    {
+      "epoch": 0.22501011736139215,
+      "grad_norm": 1.0465185642242432,
+      "learning_rate": 7.908321288457568e-05,
+      "loss": 1.0172,
+      "step": 5560
+    },
+    {
+      "epoch": 0.22521246458923513,
+      "grad_norm": 1.1151576042175293,
+      "learning_rate": 7.906256452612017e-05,
+      "loss": 1.0684,
+      "step": 5565
+    },
+    {
+      "epoch": 0.2254148118170781,
+      "grad_norm": 1.2189502716064453,
+      "learning_rate": 7.904191616766467e-05,
+      "loss": 1.0053,
+      "step": 5570
+    },
+    {
+      "epoch": 0.22561715904492108,
+      "grad_norm": 1.2066047191619873,
+      "learning_rate": 7.902126780920918e-05,
+      "loss": 0.9839,
+      "step": 5575
+    },
+    {
+      "epoch": 0.22581950627276406,
+      "grad_norm": 1.1031054258346558,
+      "learning_rate": 7.900061945075367e-05,
+      "loss": 0.9268,
+      "step": 5580
+    },
+    {
+      "epoch": 0.22602185350060705,
+      "grad_norm": 1.3175009489059448,
+      "learning_rate": 7.897997109229817e-05,
+      "loss": 1.0683,
+      "step": 5585
+    },
+    {
+      "epoch": 0.22622420072845,
+      "grad_norm": 1.1946015357971191,
+      "learning_rate": 7.895932273384266e-05,
+      "loss": 1.0687,
+      "step": 5590
+    },
+    {
+      "epoch": 0.226426547956293,
+      "grad_norm": 1.0944335460662842,
+      "learning_rate": 7.893867437538715e-05,
+      "loss": 0.9761,
+      "step": 5595
+    },
+    {
+      "epoch": 0.22662889518413598,
+      "grad_norm": 1.2366076707839966,
+      "learning_rate": 7.891802601693166e-05,
+      "loss": 1.0127,
+      "step": 5600
+    },
+    {
+      "epoch": 0.22683124241197897,
+      "grad_norm": 1.172518014907837,
+      "learning_rate": 7.889737765847616e-05,
+      "loss": 1.0268,
+      "step": 5605
+    },
+    {
+      "epoch": 0.22703358963982193,
+      "grad_norm": 1.0506185293197632,
+      "learning_rate": 7.887672930002065e-05,
+      "loss": 1.0219,
+      "step": 5610
+    },
+    {
+      "epoch": 0.2272359368676649,
+      "grad_norm": 1.2196459770202637,
+      "learning_rate": 7.885608094156514e-05,
+      "loss": 1.0136,
+      "step": 5615
+    },
+    {
+      "epoch": 0.2274382840955079,
+      "grad_norm": 1.1773611307144165,
+      "learning_rate": 7.883543258310965e-05,
+      "loss": 0.9439,
+      "step": 5620
+    },
+    {
+      "epoch": 0.22764063132335088,
+      "grad_norm": 1.2770379781723022,
+      "learning_rate": 7.881478422465415e-05,
+      "loss": 1.0332,
+      "step": 5625
+    },
+    {
+      "epoch": 0.22784297855119384,
+      "grad_norm": 1.1902120113372803,
+      "learning_rate": 7.879413586619864e-05,
+      "loss": 1.0306,
+      "step": 5630
+    },
+    {
+      "epoch": 0.22804532577903683,
+      "grad_norm": 1.0333483219146729,
+      "learning_rate": 7.877348750774313e-05,
+      "loss": 1.0273,
+      "step": 5635
+    },
+    {
+      "epoch": 0.22824767300687981,
+      "grad_norm": 1.2635235786437988,
+      "learning_rate": 7.875283914928764e-05,
+      "loss": 0.991,
+      "step": 5640
+    },
+    {
+      "epoch": 0.22845002023472277,
+      "grad_norm": 1.229201078414917,
+      "learning_rate": 7.873219079083213e-05,
+      "loss": 1.0122,
+      "step": 5645
+    },
+    {
+      "epoch": 0.22865236746256576,
+      "grad_norm": 1.1765732765197754,
+      "learning_rate": 7.871154243237663e-05,
+      "loss": 1.0348,
+      "step": 5650
+    },
+    {
+      "epoch": 0.22885471469040874,
+      "grad_norm": 1.1195917129516602,
+      "learning_rate": 7.869089407392113e-05,
+      "loss": 0.962,
+      "step": 5655
+    },
+    {
+      "epoch": 0.22905706191825173,
+      "grad_norm": 1.4746836423873901,
+      "learning_rate": 7.867024571546562e-05,
+      "loss": 1.0242,
+      "step": 5660
+    },
+    {
+      "epoch": 0.2292594091460947,
+      "grad_norm": 1.0648586750030518,
+      "learning_rate": 7.864959735701011e-05,
+      "loss": 1.0264,
+      "step": 5665
+    },
+    {
+      "epoch": 0.22946175637393768,
+      "grad_norm": 1.2590371370315552,
+      "learning_rate": 7.862894899855462e-05,
+      "loss": 0.9696,
+      "step": 5670
+    },
+    {
+      "epoch": 0.22966410360178066,
+      "grad_norm": 1.1653438806533813,
+      "learning_rate": 7.860830064009912e-05,
+      "loss": 0.9849,
+      "step": 5675
+    },
+    {
+      "epoch": 0.22986645082962365,
+      "grad_norm": 1.2064003944396973,
+      "learning_rate": 7.858765228164361e-05,
+      "loss": 1.0079,
+      "step": 5680
+    },
+    {
+      "epoch": 0.2300687980574666,
+      "grad_norm": 1.1991125345230103,
+      "learning_rate": 7.85670039231881e-05,
+      "loss": 1.0827,
+      "step": 5685
+    },
+    {
+      "epoch": 0.2302711452853096,
+      "grad_norm": 1.1990333795547485,
+      "learning_rate": 7.854635556473261e-05,
+      "loss": 1.039,
+      "step": 5690
+    },
+    {
+      "epoch": 0.23047349251315258,
+      "grad_norm": 1.106918215751648,
+      "learning_rate": 7.85257072062771e-05,
+      "loss": 0.9845,
+      "step": 5695
+    },
+    {
+      "epoch": 0.23067583974099554,
+      "grad_norm": 1.1864373683929443,
+      "learning_rate": 7.85050588478216e-05,
+      "loss": 0.967,
+      "step": 5700
+    },
+    {
+      "epoch": 0.23087818696883852,
+      "grad_norm": 1.2981541156768799,
+      "learning_rate": 7.848441048936611e-05,
+      "loss": 1.0868,
+      "step": 5705
+    },
+    {
+      "epoch": 0.2310805341966815,
+      "grad_norm": 1.1637886762619019,
+      "learning_rate": 7.84637621309106e-05,
+      "loss": 1.006,
+      "step": 5710
+    },
+    {
+      "epoch": 0.2312828814245245,
+      "grad_norm": 1.1366996765136719,
+      "learning_rate": 7.844311377245509e-05,
+      "loss": 1.0588,
+      "step": 5715
+    },
+    {
+      "epoch": 0.23148522865236745,
+      "grad_norm": 1.1532763242721558,
+      "learning_rate": 7.842246541399959e-05,
+      "loss": 1.0586,
+      "step": 5720
+    },
+    {
+      "epoch": 0.23168757588021044,
+      "grad_norm": 1.2979742288589478,
+      "learning_rate": 7.84018170555441e-05,
+      "loss": 1.0527,
+      "step": 5725
+    },
+    {
+      "epoch": 0.23188992310805343,
+      "grad_norm": 1.1109341382980347,
+      "learning_rate": 7.838116869708859e-05,
+      "loss": 1.0451,
+      "step": 5730
+    },
+    {
+      "epoch": 0.2320922703358964,
+      "grad_norm": 1.1665804386138916,
+      "learning_rate": 7.836052033863308e-05,
+      "loss": 1.0379,
+      "step": 5735
+    },
+    {
+      "epoch": 0.23229461756373937,
+      "grad_norm": 1.217777967453003,
+      "learning_rate": 7.833987198017758e-05,
+      "loss": 0.9944,
+      "step": 5740
+    },
+    {
+      "epoch": 0.23249696479158236,
+      "grad_norm": 1.1025943756103516,
+      "learning_rate": 7.831922362172209e-05,
+      "loss": 0.9897,
+      "step": 5745
+    },
+    {
+      "epoch": 0.23269931201942534,
+      "grad_norm": 1.0555897951126099,
+      "learning_rate": 7.829857526326658e-05,
+      "loss": 1.0565,
+      "step": 5750
+    },
+    {
+      "epoch": 0.2329016592472683,
+      "grad_norm": 1.1598858833312988,
+      "learning_rate": 7.827792690481107e-05,
+      "loss": 0.977,
+      "step": 5755
+    },
+    {
+      "epoch": 0.2331040064751113,
+      "grad_norm": 1.0308842658996582,
+      "learning_rate": 7.825727854635557e-05,
+      "loss": 1.048,
+      "step": 5760
+    },
+    {
+      "epoch": 0.23330635370295427,
+      "grad_norm": 1.1150622367858887,
+      "learning_rate": 7.823663018790006e-05,
+      "loss": 0.9863,
+      "step": 5765
+    },
+    {
+      "epoch": 0.23350870093079726,
+      "grad_norm": 1.189045786857605,
+      "learning_rate": 7.821598182944456e-05,
+      "loss": 1.0165,
+      "step": 5770
+    },
+    {
+      "epoch": 0.23371104815864022,
+      "grad_norm": 1.2735435962677002,
+      "learning_rate": 7.819533347098907e-05,
+      "loss": 1.038,
+      "step": 5775
+    },
+    {
+      "epoch": 0.2339133953864832,
+      "grad_norm": 1.2038140296936035,
+      "learning_rate": 7.817468511253355e-05,
+      "loss": 1.0072,
+      "step": 5780
+    },
+    {
+      "epoch": 0.2341157426143262,
+      "grad_norm": 1.0952842235565186,
+      "learning_rate": 7.815403675407805e-05,
+      "loss": 0.9773,
+      "step": 5785
+    },
+    {
+      "epoch": 0.23431808984216917,
+      "grad_norm": 1.3182307481765747,
+      "learning_rate": 7.813338839562255e-05,
+      "loss": 0.9853,
+      "step": 5790
+    },
+    {
+      "epoch": 0.23452043707001213,
+      "grad_norm": 1.1275579929351807,
+      "learning_rate": 7.811274003716706e-05,
+      "loss": 0.9792,
+      "step": 5795
+    },
+    {
+      "epoch": 0.23472278429785512,
+      "grad_norm": 1.2322643995285034,
+      "learning_rate": 7.809209167871155e-05,
+      "loss": 0.9901,
+      "step": 5800
+    },
+    {
+      "epoch": 0.2349251315256981,
+      "grad_norm": 1.1326335668563843,
+      "learning_rate": 7.807144332025604e-05,
+      "loss": 0.977,
+      "step": 5805
+    },
+    {
+      "epoch": 0.23512747875354106,
+      "grad_norm": 1.1416250467300415,
+      "learning_rate": 7.805079496180054e-05,
+      "loss": 0.9662,
+      "step": 5810
+    },
+    {
+      "epoch": 0.23532982598138405,
+      "grad_norm": 1.2562077045440674,
+      "learning_rate": 7.803014660334503e-05,
+      "loss": 0.9903,
+      "step": 5815
+    },
+    {
+      "epoch": 0.23553217320922704,
+      "grad_norm": 1.15487539768219,
+      "learning_rate": 7.800949824488954e-05,
+      "loss": 1.0253,
+      "step": 5820
+    },
+    {
+      "epoch": 0.23573452043707002,
+      "grad_norm": 1.0592039823532104,
+      "learning_rate": 7.798884988643403e-05,
+      "loss": 0.944,
+      "step": 5825
+    },
+    {
+      "epoch": 0.23593686766491298,
+      "grad_norm": 1.2863718271255493,
+      "learning_rate": 7.796820152797853e-05,
+      "loss": 1.0292,
+      "step": 5830
+    },
+    {
+      "epoch": 0.23613921489275597,
+      "grad_norm": 1.145825743675232,
+      "learning_rate": 7.794755316952302e-05,
+      "loss": 1.0172,
+      "step": 5835
+    },
+    {
+      "epoch": 0.23634156212059895,
+      "grad_norm": 1.0824973583221436,
+      "learning_rate": 7.792690481106753e-05,
+      "loss": 0.9922,
+      "step": 5840
+    },
+    {
+      "epoch": 0.23654390934844194,
+      "grad_norm": 1.1497865915298462,
+      "learning_rate": 7.790625645261203e-05,
+      "loss": 1.0203,
+      "step": 5845
+    },
+    {
+      "epoch": 0.2367462565762849,
+      "grad_norm": 1.1841051578521729,
+      "learning_rate": 7.788560809415651e-05,
+      "loss": 1.0286,
+      "step": 5850
+    },
+    {
+      "epoch": 0.23694860380412788,
+      "grad_norm": 1.1089333295822144,
+      "learning_rate": 7.786495973570101e-05,
+      "loss": 1.0517,
+      "step": 5855
+    },
+    {
+      "epoch": 0.23715095103197087,
+      "grad_norm": 1.3060795068740845,
+      "learning_rate": 7.784431137724552e-05,
+      "loss": 1.0358,
+      "step": 5860
+    },
+    {
+      "epoch": 0.23735329825981383,
+      "grad_norm": 1.2065348625183105,
+      "learning_rate": 7.782366301879e-05,
+      "loss": 1.0508,
+      "step": 5865
+    },
+    {
+      "epoch": 0.2375556454876568,
+      "grad_norm": 1.0978989601135254,
+      "learning_rate": 7.780301466033451e-05,
+      "loss": 0.9901,
+      "step": 5870
+    },
+    {
+      "epoch": 0.2377579927154998,
+      "grad_norm": 1.1329617500305176,
+      "learning_rate": 7.7782366301879e-05,
+      "loss": 1.0113,
+      "step": 5875
+    },
+    {
+      "epoch": 0.23796033994334279,
+      "grad_norm": 1.1056050062179565,
+      "learning_rate": 7.77617179434235e-05,
+      "loss": 1.0738,
+      "step": 5880
+    },
+    {
+      "epoch": 0.23816268717118574,
+      "grad_norm": 1.3044637441635132,
+      "learning_rate": 7.7741069584968e-05,
+      "loss": 1.0375,
+      "step": 5885
+    },
+    {
+      "epoch": 0.23836503439902873,
+      "grad_norm": 1.2118210792541504,
+      "learning_rate": 7.77204212265125e-05,
+      "loss": 1.0014,
+      "step": 5890
+    },
+    {
+      "epoch": 0.23856738162687172,
+      "grad_norm": 1.3539366722106934,
+      "learning_rate": 7.769977286805699e-05,
+      "loss": 0.978,
+      "step": 5895
+    },
+    {
+      "epoch": 0.2387697288547147,
+      "grad_norm": 1.1183351278305054,
+      "learning_rate": 7.767912450960148e-05,
+      "loss": 0.9477,
+      "step": 5900
+    },
+    {
+      "epoch": 0.23897207608255766,
+      "grad_norm": 1.2246525287628174,
+      "learning_rate": 7.765847615114599e-05,
+      "loss": 1.0384,
+      "step": 5905
+    },
+    {
+      "epoch": 0.23917442331040065,
+      "grad_norm": 1.2153912782669067,
+      "learning_rate": 7.763782779269049e-05,
+      "loss": 0.9998,
+      "step": 5910
+    },
+    {
+      "epoch": 0.23937677053824363,
+      "grad_norm": 1.2209187746047974,
+      "learning_rate": 7.761717943423498e-05,
+      "loss": 1.0025,
+      "step": 5915
+    },
+    {
+      "epoch": 0.2395791177660866,
+      "grad_norm": 1.1944117546081543,
+      "learning_rate": 7.759653107577948e-05,
+      "loss": 1.0176,
+      "step": 5920
+    },
+    {
+      "epoch": 0.23978146499392958,
+      "grad_norm": 1.1967753171920776,
+      "learning_rate": 7.757588271732397e-05,
+      "loss": 1.0214,
+      "step": 5925
+    },
+    {
+      "epoch": 0.23998381222177256,
+      "grad_norm": 1.0911521911621094,
+      "learning_rate": 7.755523435886848e-05,
+      "loss": 0.9598,
+      "step": 5930
+    },
+    {
+      "epoch": 0.24018615944961555,
+      "grad_norm": 1.2010823488235474,
+      "learning_rate": 7.753458600041297e-05,
+      "loss": 0.9866,
+      "step": 5935
+    },
+    {
+      "epoch": 0.2403885066774585,
+      "grad_norm": 1.1828703880310059,
+      "learning_rate": 7.751393764195747e-05,
+      "loss": 0.9887,
+      "step": 5940
+    },
+    {
+      "epoch": 0.2405908539053015,
+      "grad_norm": 1.1219539642333984,
+      "learning_rate": 7.749328928350196e-05,
+      "loss": 1.0612,
+      "step": 5945
+    },
+    {
+      "epoch": 0.24079320113314448,
+      "grad_norm": 1.1368293762207031,
+      "learning_rate": 7.747264092504645e-05,
+      "loss": 0.9991,
+      "step": 5950
+    },
+    {
+      "epoch": 0.24099554836098747,
+      "grad_norm": 1.1059019565582275,
+      "learning_rate": 7.745199256659096e-05,
+      "loss": 1.0682,
+      "step": 5955
+    },
+    {
+      "epoch": 0.24119789558883042,
+      "grad_norm": 1.0998921394348145,
+      "learning_rate": 7.743134420813546e-05,
+      "loss": 1.0012,
+      "step": 5960
+    },
+    {
+      "epoch": 0.2414002428166734,
+      "grad_norm": 1.2579197883605957,
+      "learning_rate": 7.741069584967997e-05,
+      "loss": 1.0193,
+      "step": 5965
+    },
+    {
+      "epoch": 0.2416025900445164,
+      "grad_norm": 1.1530227661132812,
+      "learning_rate": 7.739004749122444e-05,
+      "loss": 1.024,
+      "step": 5970
+    },
+    {
+      "epoch": 0.24180493727235935,
+      "grad_norm": 1.229365587234497,
+      "learning_rate": 7.736939913276895e-05,
+      "loss": 1.0177,
+      "step": 5975
+    },
+    {
+      "epoch": 0.24200728450020234,
+      "grad_norm": 1.150947093963623,
+      "learning_rate": 7.734875077431345e-05,
+      "loss": 1.0463,
+      "step": 5980
+    },
+    {
+      "epoch": 0.24220963172804533,
+      "grad_norm": 1.0764719247817993,
+      "learning_rate": 7.732810241585794e-05,
+      "loss": 1.0137,
+      "step": 5985
+    },
+    {
+      "epoch": 0.2424119789558883,
+      "grad_norm": 1.0965443849563599,
+      "learning_rate": 7.730745405740245e-05,
+      "loss": 1.0019,
+      "step": 5990
+    },
+    {
+      "epoch": 0.24261432618373127,
+      "grad_norm": 1.1556476354599,
+      "learning_rate": 7.728680569894694e-05,
+      "loss": 0.9645,
+      "step": 5995
+    },
+    {
+      "epoch": 0.24281667341157426,
+      "grad_norm": 1.1066598892211914,
+      "learning_rate": 7.726615734049143e-05,
+      "loss": 1.0096,
+      "step": 6000
+    },
+    {
+      "epoch": 0.24301902063941724,
+      "grad_norm": 1.0250757932662964,
+      "learning_rate": 7.724550898203593e-05,
+      "loss": 1.0202,
+      "step": 6005
+    },
+    {
+      "epoch": 0.24322136786726023,
+      "grad_norm": 1.1670464277267456,
+      "learning_rate": 7.722486062358044e-05,
+      "loss": 1.004,
+      "step": 6010
+    },
+    {
+      "epoch": 0.2434237150951032,
+      "grad_norm": 1.0662355422973633,
+      "learning_rate": 7.720421226512493e-05,
+      "loss": 1.0175,
+      "step": 6015
+    },
+    {
+      "epoch": 0.24362606232294617,
+      "grad_norm": 1.116700291633606,
+      "learning_rate": 7.718356390666942e-05,
+      "loss": 1.0203,
+      "step": 6020
+    },
+    {
+      "epoch": 0.24382840955078916,
+      "grad_norm": 1.2482693195343018,
+      "learning_rate": 7.716291554821392e-05,
+      "loss": 1.0136,
+      "step": 6025
+    },
+    {
+      "epoch": 0.24403075677863215,
+      "grad_norm": 1.1996208429336548,
+      "learning_rate": 7.714226718975842e-05,
+      "loss": 0.9716,
+      "step": 6030
+    },
+    {
+      "epoch": 0.2442331040064751,
+      "grad_norm": 1.2952680587768555,
+      "learning_rate": 7.712161883130291e-05,
+      "loss": 1.0123,
+      "step": 6035
+    },
+    {
+      "epoch": 0.2444354512343181,
+      "grad_norm": 1.1774368286132812,
+      "learning_rate": 7.71009704728474e-05,
+      "loss": 1.0258,
+      "step": 6040
+    },
+    {
+      "epoch": 0.24463779846216108,
+      "grad_norm": 1.0907679796218872,
+      "learning_rate": 7.708032211439191e-05,
+      "loss": 1.0059,
+      "step": 6045
+    },
+    {
+      "epoch": 0.24484014569000404,
+      "grad_norm": 1.2474931478500366,
+      "learning_rate": 7.70596737559364e-05,
+      "loss": 1.0642,
+      "step": 6050
+    },
+    {
+      "epoch": 0.24504249291784702,
+      "grad_norm": 1.1548250913619995,
+      "learning_rate": 7.70390253974809e-05,
+      "loss": 1.0278,
+      "step": 6055
+    },
+    {
+      "epoch": 0.24524484014569,
+      "grad_norm": 1.126945972442627,
+      "learning_rate": 7.701837703902541e-05,
+      "loss": 0.9554,
+      "step": 6060
+    },
+    {
+      "epoch": 0.245447187373533,
+      "grad_norm": 1.265137791633606,
+      "learning_rate": 7.69977286805699e-05,
+      "loss": 1.0365,
+      "step": 6065
+    },
+    {
+      "epoch": 0.24564953460137595,
+      "grad_norm": 1.3145203590393066,
+      "learning_rate": 7.697708032211439e-05,
+      "loss": 1.0036,
+      "step": 6070
+    },
+    {
+      "epoch": 0.24585188182921894,
+      "grad_norm": 1.2152060270309448,
+      "learning_rate": 7.695643196365889e-05,
+      "loss": 0.9918,
+      "step": 6075
+    },
+    {
+      "epoch": 0.24605422905706192,
+      "grad_norm": 1.1506688594818115,
+      "learning_rate": 7.69357836052034e-05,
+      "loss": 1.0007,
+      "step": 6080
+    },
+    {
+      "epoch": 0.2462565762849049,
+      "grad_norm": 1.2358943223953247,
+      "learning_rate": 7.691513524674789e-05,
+      "loss": 1.001,
+      "step": 6085
+    },
+    {
+      "epoch": 0.24645892351274787,
+      "grad_norm": 1.3536697626113892,
+      "learning_rate": 7.689448688829238e-05,
+      "loss": 1.0041,
+      "step": 6090
+    },
+    {
+      "epoch": 0.24666127074059085,
+      "grad_norm": 1.1014422178268433,
+      "learning_rate": 7.687383852983688e-05,
+      "loss": 1.0577,
+      "step": 6095
+    },
+    {
+      "epoch": 0.24686361796843384,
+      "grad_norm": 1.2396435737609863,
+      "learning_rate": 7.685319017138139e-05,
+      "loss": 1.0578,
+      "step": 6100
+    },
+    {
+      "epoch": 0.2470659651962768,
+      "grad_norm": 1.0494437217712402,
+      "learning_rate": 7.683254181292588e-05,
+      "loss": 1.0146,
+      "step": 6105
+    },
+    {
+      "epoch": 0.24726831242411978,
+      "grad_norm": 1.1579269170761108,
+      "learning_rate": 7.681189345447037e-05,
+      "loss": 1.0344,
+      "step": 6110
+    },
+    {
+      "epoch": 0.24747065965196277,
+      "grad_norm": 1.0838518142700195,
+      "learning_rate": 7.679124509601487e-05,
+      "loss": 1.0192,
+      "step": 6115
+    },
+    {
+      "epoch": 0.24767300687980576,
+      "grad_norm": 1.1838576793670654,
+      "learning_rate": 7.677059673755936e-05,
+      "loss": 0.9935,
+      "step": 6120
+    },
+    {
+      "epoch": 0.24787535410764872,
+      "grad_norm": 1.1222121715545654,
+      "learning_rate": 7.674994837910387e-05,
+      "loss": 0.9936,
+      "step": 6125
+    },
+    {
+      "epoch": 0.2480777013354917,
+      "grad_norm": 1.2552130222320557,
+      "learning_rate": 7.672930002064837e-05,
+      "loss": 1.0226,
+      "step": 6130
+    },
+    {
+      "epoch": 0.2482800485633347,
+      "grad_norm": 1.1589610576629639,
+      "learning_rate": 7.670865166219286e-05,
+      "loss": 1.0136,
+      "step": 6135
+    },
+    {
+      "epoch": 0.24848239579117767,
+      "grad_norm": 1.0672346353530884,
+      "learning_rate": 7.668800330373735e-05,
+      "loss": 0.9752,
+      "step": 6140
+    },
+    {
+      "epoch": 0.24868474301902063,
+      "grad_norm": 1.1981768608093262,
+      "learning_rate": 7.666735494528186e-05,
+      "loss": 1.0218,
+      "step": 6145
+    },
+    {
+      "epoch": 0.24888709024686362,
+      "grad_norm": 1.2581173181533813,
+      "learning_rate": 7.664670658682636e-05,
+      "loss": 0.9783,
+      "step": 6150
+    },
+    {
+      "epoch": 0.2490894374747066,
+      "grad_norm": 1.217789649963379,
+      "learning_rate": 7.662605822837085e-05,
+      "loss": 1.0005,
+      "step": 6155
+    },
+    {
+      "epoch": 0.24929178470254956,
+      "grad_norm": 1.1852567195892334,
+      "learning_rate": 7.660540986991534e-05,
+      "loss": 1.0515,
+      "step": 6160
+    },
+    {
+      "epoch": 0.24949413193039255,
+      "grad_norm": 1.1580768823623657,
+      "learning_rate": 7.658476151145984e-05,
+      "loss": 1.0018,
+      "step": 6165
+    },
+    {
+      "epoch": 0.24969647915823553,
+      "grad_norm": 1.112929344177246,
+      "learning_rate": 7.656411315300434e-05,
+      "loss": 0.9975,
+      "step": 6170
+    },
+    {
+      "epoch": 0.24989882638607852,
+      "grad_norm": 1.1734755039215088,
+      "learning_rate": 7.654346479454884e-05,
+      "loss": 0.9948,
+      "step": 6175
+    },
+    {
+      "epoch": 0.2501011736139215,
+      "grad_norm": 1.334947943687439,
+      "learning_rate": 7.652281643609333e-05,
+      "loss": 0.9802,
+      "step": 6180
+    },
+    {
+      "epoch": 0.2503035208417645,
+      "grad_norm": 1.2167521715164185,
+      "learning_rate": 7.650216807763783e-05,
+      "loss": 0.9401,
+      "step": 6185
+    },
+    {
+      "epoch": 0.25050586806960745,
+      "grad_norm": 1.1352683305740356,
+      "learning_rate": 7.648151971918232e-05,
+      "loss": 0.952,
+      "step": 6190
+    },
+    {
+      "epoch": 0.2507082152974504,
+      "grad_norm": 1.062712550163269,
+      "learning_rate": 7.646087136072683e-05,
+      "loss": 1.0991,
+      "step": 6195
+    },
+    {
+      "epoch": 0.2509105625252934,
+      "grad_norm": 1.1996139287948608,
+      "learning_rate": 7.644022300227133e-05,
+      "loss": 1.0518,
+      "step": 6200
+    },
+    {
+      "epoch": 0.2511129097531364,
+      "grad_norm": 1.1053889989852905,
+      "learning_rate": 7.641957464381582e-05,
+      "loss": 1.0319,
+      "step": 6205
+    },
+    {
+      "epoch": 0.25131525698097934,
+      "grad_norm": 1.2459156513214111,
+      "learning_rate": 7.639892628536031e-05,
+      "loss": 1.0318,
+      "step": 6210
+    },
+    {
+      "epoch": 0.25151760420882235,
+      "grad_norm": 1.159220576286316,
+      "learning_rate": 7.637827792690482e-05,
+      "loss": 1.0361,
+      "step": 6215
+    },
+    {
+      "epoch": 0.2517199514366653,
+      "grad_norm": 1.0500812530517578,
+      "learning_rate": 7.635762956844931e-05,
+      "loss": 1.0122,
+      "step": 6220
+    },
+    {
+      "epoch": 0.25192229866450827,
+      "grad_norm": 1.1624324321746826,
+      "learning_rate": 7.633698120999381e-05,
+      "loss": 1.0061,
+      "step": 6225
+    },
+    {
+      "epoch": 0.2521246458923513,
+      "grad_norm": 1.254845380783081,
+      "learning_rate": 7.63163328515383e-05,
+      "loss": 1.0205,
+      "step": 6230
+    },
+    {
+      "epoch": 0.25232699312019424,
+      "grad_norm": 1.1832902431488037,
+      "learning_rate": 7.62956844930828e-05,
+      "loss": 1.0651,
+      "step": 6235
+    },
+    {
+      "epoch": 0.25252934034803726,
+      "grad_norm": 1.1697665452957153,
+      "learning_rate": 7.62750361346273e-05,
+      "loss": 1.0184,
+      "step": 6240
+    },
+    {
+      "epoch": 0.2527316875758802,
+      "grad_norm": 1.0969288349151611,
+      "learning_rate": 7.62543877761718e-05,
+      "loss": 0.9464,
+      "step": 6245
+    },
+    {
+      "epoch": 0.2529340348037232,
+      "grad_norm": 1.3374019861221313,
+      "learning_rate": 7.62337394177163e-05,
+      "loss": 1.0608,
+      "step": 6250
+    },
+    {
+      "epoch": 0.2531363820315662,
+      "grad_norm": 1.2375236749649048,
+      "learning_rate": 7.621309105926078e-05,
+      "loss": 0.9866,
+      "step": 6255
+    },
+    {
+      "epoch": 0.25333872925940915,
+      "grad_norm": 1.2215721607208252,
+      "learning_rate": 7.619244270080529e-05,
+      "loss": 1.0422,
+      "step": 6260
+    },
+    {
+      "epoch": 0.2535410764872521,
+      "grad_norm": 1.124558687210083,
+      "learning_rate": 7.617179434234979e-05,
+      "loss": 0.9831,
+      "step": 6265
+    },
+    {
+      "epoch": 0.2537434237150951,
+      "grad_norm": 1.1476385593414307,
+      "learning_rate": 7.615114598389428e-05,
+      "loss": 1.0262,
+      "step": 6270
+    },
+    {
+      "epoch": 0.2539457709429381,
+      "grad_norm": 1.2365763187408447,
+      "learning_rate": 7.613049762543879e-05,
+      "loss": 1.0476,
+      "step": 6275
+    },
+    {
+      "epoch": 0.25414811817078103,
+      "grad_norm": 1.1760541200637817,
+      "learning_rate": 7.610984926698328e-05,
+      "loss": 1.0112,
+      "step": 6280
+    },
+    {
+      "epoch": 0.25435046539862405,
+      "grad_norm": 1.1309133768081665,
+      "learning_rate": 7.608920090852778e-05,
+      "loss": 1.0201,
+      "step": 6285
+    },
+    {
+      "epoch": 0.254552812626467,
+      "grad_norm": 1.133035659790039,
+      "learning_rate": 7.606855255007227e-05,
+      "loss": 1.0136,
+      "step": 6290
+    },
+    {
+      "epoch": 0.25475515985431,
+      "grad_norm": 1.1882902383804321,
+      "learning_rate": 7.604790419161677e-05,
+      "loss": 0.9954,
+      "step": 6295
+    },
+    {
+      "epoch": 0.254957507082153,
+      "grad_norm": 1.1148749589920044,
+      "learning_rate": 7.602725583316126e-05,
+      "loss": 1.0052,
+      "step": 6300
+    },
+    {
+      "epoch": 0.25515985430999594,
+      "grad_norm": 1.1846961975097656,
+      "learning_rate": 7.600660747470576e-05,
+      "loss": 0.9672,
+      "step": 6305
+    },
+    {
+      "epoch": 0.25536220153783895,
+      "grad_norm": 1.0866841077804565,
+      "learning_rate": 7.598595911625026e-05,
+      "loss": 1.0197,
+      "step": 6310
+    },
+    {
+      "epoch": 0.2555645487656819,
+      "grad_norm": 1.1378873586654663,
+      "learning_rate": 7.596531075779476e-05,
+      "loss": 1.0278,
+      "step": 6315
+    },
+    {
+      "epoch": 0.25576689599352487,
+      "grad_norm": 1.1977087259292603,
+      "learning_rate": 7.594466239933927e-05,
+      "loss": 0.988,
+      "step": 6320
+    },
+    {
+      "epoch": 0.2559692432213679,
+      "grad_norm": 1.1284993886947632,
+      "learning_rate": 7.592401404088374e-05,
+      "loss": 1.0356,
+      "step": 6325
+    },
+    {
+      "epoch": 0.25617159044921084,
+      "grad_norm": 1.0692222118377686,
+      "learning_rate": 7.590336568242825e-05,
+      "loss": 0.9913,
+      "step": 6330
+    },
+    {
+      "epoch": 0.2563739376770538,
+      "grad_norm": 1.2074707746505737,
+      "learning_rate": 7.588271732397275e-05,
+      "loss": 1.0101,
+      "step": 6335
+    },
+    {
+      "epoch": 0.2565762849048968,
+      "grad_norm": 1.0940556526184082,
+      "learning_rate": 7.586206896551724e-05,
+      "loss": 1.0192,
+      "step": 6340
+    },
+    {
+      "epoch": 0.25677863213273977,
+      "grad_norm": 1.2523458003997803,
+      "learning_rate": 7.584142060706175e-05,
+      "loss": 1.0089,
+      "step": 6345
+    },
+    {
+      "epoch": 0.2569809793605828,
+      "grad_norm": 1.1117687225341797,
+      "learning_rate": 7.582077224860624e-05,
+      "loss": 1.0206,
+      "step": 6350
+    },
+    {
+      "epoch": 0.25718332658842574,
+      "grad_norm": 1.1576602458953857,
+      "learning_rate": 7.580012389015073e-05,
+      "loss": 0.9893,
+      "step": 6355
+    },
+    {
+      "epoch": 0.2573856738162687,
+      "grad_norm": 1.1742234230041504,
+      "learning_rate": 7.577947553169523e-05,
+      "loss": 1.0281,
+      "step": 6360
+    },
+    {
+      "epoch": 0.2575880210441117,
+      "grad_norm": 1.2967197895050049,
+      "learning_rate": 7.575882717323974e-05,
+      "loss": 1.0162,
+      "step": 6365
+    },
+    {
+      "epoch": 0.2577903682719547,
+      "grad_norm": 1.1597435474395752,
+      "learning_rate": 7.573817881478423e-05,
+      "loss": 0.9695,
+      "step": 6370
+    },
+    {
+      "epoch": 0.25799271549979763,
+      "grad_norm": 1.1363296508789062,
+      "learning_rate": 7.571753045632872e-05,
+      "loss": 1.0629,
+      "step": 6375
+    },
+    {
+      "epoch": 0.25819506272764065,
+      "grad_norm": 1.2326802015304565,
+      "learning_rate": 7.569688209787322e-05,
+      "loss": 0.972,
+      "step": 6380
+    },
+    {
+      "epoch": 0.2583974099554836,
+      "grad_norm": 1.0850800275802612,
+      "learning_rate": 7.567623373941773e-05,
+      "loss": 0.9562,
+      "step": 6385
+    },
+    {
+      "epoch": 0.25859975718332656,
+      "grad_norm": 1.3475916385650635,
+      "learning_rate": 7.565558538096222e-05,
+      "loss": 0.96,
+      "step": 6390
+    },
+    {
+      "epoch": 0.2588021044111696,
+      "grad_norm": 1.1954843997955322,
+      "learning_rate": 7.56349370225067e-05,
+      "loss": 1.0534,
+      "step": 6395
+    },
+    {
+      "epoch": 0.25900445163901253,
+      "grad_norm": 1.2429571151733398,
+      "learning_rate": 7.561428866405121e-05,
+      "loss": 1.0081,
+      "step": 6400
+    },
+    {
+      "epoch": 0.25920679886685555,
+      "grad_norm": 1.2253036499023438,
+      "learning_rate": 7.559364030559571e-05,
+      "loss": 0.9653,
+      "step": 6405
+    },
+    {
+      "epoch": 0.2594091460946985,
+      "grad_norm": 1.1782009601593018,
+      "learning_rate": 7.55729919471402e-05,
+      "loss": 1.0345,
+      "step": 6410
+    },
+    {
+      "epoch": 0.25961149332254146,
+      "grad_norm": 1.2082030773162842,
+      "learning_rate": 7.555234358868471e-05,
+      "loss": 1.0469,
+      "step": 6415
+    },
+    {
+      "epoch": 0.2598138405503845,
+      "grad_norm": 1.2173150777816772,
+      "learning_rate": 7.55316952302292e-05,
+      "loss": 0.991,
+      "step": 6420
+    },
+    {
+      "epoch": 0.26001618777822744,
+      "grad_norm": 1.286616325378418,
+      "learning_rate": 7.551104687177369e-05,
+      "loss": 0.9516,
+      "step": 6425
+    },
+    {
+      "epoch": 0.2602185350060704,
+      "grad_norm": 1.1527777910232544,
+      "learning_rate": 7.54903985133182e-05,
+      "loss": 1.0253,
+      "step": 6430
+    },
+    {
+      "epoch": 0.2604208822339134,
+      "grad_norm": 1.1592366695404053,
+      "learning_rate": 7.54697501548627e-05,
+      "loss": 1.0062,
+      "step": 6435
+    },
+    {
+      "epoch": 0.26062322946175637,
+      "grad_norm": 1.2501929998397827,
+      "learning_rate": 7.544910179640719e-05,
+      "loss": 1.0231,
+      "step": 6440
+    },
+    {
+      "epoch": 0.2608255766895993,
+      "grad_norm": 1.3574258089065552,
+      "learning_rate": 7.542845343795168e-05,
+      "loss": 0.9926,
+      "step": 6445
+    },
+    {
+      "epoch": 0.26102792391744234,
+      "grad_norm": 1.1337406635284424,
+      "learning_rate": 7.540780507949618e-05,
+      "loss": 0.9602,
+      "step": 6450
+    },
+    {
+      "epoch": 0.2612302711452853,
+      "grad_norm": 1.170088291168213,
+      "learning_rate": 7.538715672104069e-05,
+      "loss": 1.0621,
+      "step": 6455
+    },
+    {
+      "epoch": 0.2614326183731283,
+      "grad_norm": 1.0618562698364258,
+      "learning_rate": 7.536650836258518e-05,
+      "loss": 1.016,
+      "step": 6460
+    },
+    {
+      "epoch": 0.26163496560097127,
+      "grad_norm": 1.1312986612319946,
+      "learning_rate": 7.534586000412968e-05,
+      "loss": 1.029,
+      "step": 6465
+    },
+    {
+      "epoch": 0.26183731282881423,
+      "grad_norm": 1.052512764930725,
+      "learning_rate": 7.532521164567417e-05,
+      "loss": 1.0404,
+      "step": 6470
+    },
+    {
+      "epoch": 0.26203966005665724,
+      "grad_norm": 1.1444694995880127,
+      "learning_rate": 7.530456328721866e-05,
+      "loss": 0.9828,
+      "step": 6475
+    },
+    {
+      "epoch": 0.2622420072845002,
+      "grad_norm": 1.016725778579712,
+      "learning_rate": 7.528391492876317e-05,
+      "loss": 1.0533,
+      "step": 6480
+    },
+    {
+      "epoch": 0.26244435451234316,
+      "grad_norm": 1.1582973003387451,
+      "learning_rate": 7.526326657030767e-05,
+      "loss": 0.9797,
+      "step": 6485
+    },
+    {
+      "epoch": 0.2626467017401862,
+      "grad_norm": 1.1264119148254395,
+      "learning_rate": 7.524261821185216e-05,
+      "loss": 0.9768,
+      "step": 6490
+    },
+    {
+      "epoch": 0.26284904896802913,
+      "grad_norm": 1.236449956893921,
+      "learning_rate": 7.522196985339665e-05,
+      "loss": 0.9688,
+      "step": 6495
+    },
+    {
+      "epoch": 0.2630513961958721,
+      "grad_norm": 1.2319833040237427,
+      "learning_rate": 7.520132149494116e-05,
+      "loss": 1.0407,
+      "step": 6500
+    },
+    {
+      "epoch": 0.2632537434237151,
+      "grad_norm": 1.259884238243103,
+      "learning_rate": 7.518067313648566e-05,
+      "loss": 1.0292,
+      "step": 6505
+    },
+    {
+      "epoch": 0.26345609065155806,
+      "grad_norm": 1.0968300104141235,
+      "learning_rate": 7.516002477803015e-05,
+      "loss": 1.0098,
+      "step": 6510
+    },
+    {
+      "epoch": 0.2636584378794011,
+      "grad_norm": 1.209838628768921,
+      "learning_rate": 7.513937641957464e-05,
+      "loss": 1.0292,
+      "step": 6515
+    },
+    {
+      "epoch": 0.26386078510724403,
+      "grad_norm": 1.2530168294906616,
+      "learning_rate": 7.511872806111915e-05,
+      "loss": 1.0444,
+      "step": 6520
+    },
+    {
+      "epoch": 0.264063132335087,
+      "grad_norm": 1.2566606998443604,
+      "learning_rate": 7.509807970266364e-05,
+      "loss": 1.0087,
+      "step": 6525
+    },
+    {
+      "epoch": 0.26426547956293,
+      "grad_norm": 1.1261377334594727,
+      "learning_rate": 7.507743134420814e-05,
+      "loss": 1.0128,
+      "step": 6530
+    },
+    {
+      "epoch": 0.26446782679077296,
+      "grad_norm": 1.2107958793640137,
+      "learning_rate": 7.505678298575264e-05,
+      "loss": 1.0155,
+      "step": 6535
+    },
+    {
+      "epoch": 0.2646701740186159,
+      "grad_norm": 1.0442723035812378,
+      "learning_rate": 7.503613462729714e-05,
+      "loss": 0.9896,
+      "step": 6540
+    },
+    {
+      "epoch": 0.26487252124645894,
+      "grad_norm": 1.1261143684387207,
+      "learning_rate": 7.501548626884163e-05,
+      "loss": 0.9892,
+      "step": 6545
+    },
+    {
+      "epoch": 0.2650748684743019,
+      "grad_norm": 1.144325613975525,
+      "learning_rate": 7.499483791038613e-05,
+      "loss": 1.0115,
+      "step": 6550
+    },
+    {
+      "epoch": 0.26527721570214485,
+      "grad_norm": 1.214985966682434,
+      "learning_rate": 7.497418955193063e-05,
+      "loss": 1.0264,
+      "step": 6555
+    },
+    {
+      "epoch": 0.26547956292998787,
+      "grad_norm": 1.2811050415039062,
+      "learning_rate": 7.495354119347512e-05,
+      "loss": 0.9441,
+      "step": 6560
+    },
+    {
+      "epoch": 0.2656819101578308,
+      "grad_norm": 1.319085955619812,
+      "learning_rate": 7.493289283501961e-05,
+      "loss": 1.0454,
+      "step": 6565
+    },
+    {
+      "epoch": 0.26588425738567384,
+      "grad_norm": 1.091052770614624,
+      "learning_rate": 7.491224447656412e-05,
+      "loss": 1.1001,
+      "step": 6570
+    },
+    {
+      "epoch": 0.2660866046135168,
+      "grad_norm": 1.2728432416915894,
+      "learning_rate": 7.489159611810861e-05,
+      "loss": 1.0169,
+      "step": 6575
+    },
+    {
+      "epoch": 0.26628895184135976,
+      "grad_norm": 1.3494826555252075,
+      "learning_rate": 7.487094775965311e-05,
+      "loss": 1.0424,
+      "step": 6580
+    },
+    {
+      "epoch": 0.26649129906920277,
+      "grad_norm": 1.1133043766021729,
+      "learning_rate": 7.48502994011976e-05,
+      "loss": 1.0255,
+      "step": 6585
+    },
+    {
+      "epoch": 0.26669364629704573,
+      "grad_norm": 1.2886162996292114,
+      "learning_rate": 7.482965104274211e-05,
+      "loss": 1.0015,
+      "step": 6590
+    },
+    {
+      "epoch": 0.2668959935248887,
+      "grad_norm": 1.2574056386947632,
+      "learning_rate": 7.48090026842866e-05,
+      "loss": 1.0546,
+      "step": 6595
+    },
+    {
+      "epoch": 0.2670983407527317,
+      "grad_norm": 1.2051347494125366,
+      "learning_rate": 7.47883543258311e-05,
+      "loss": 0.9476,
+      "step": 6600
+    },
+    {
+      "epoch": 0.26730068798057466,
+      "grad_norm": 1.0692707300186157,
+      "learning_rate": 7.47677059673756e-05,
+      "loss": 0.9864,
+      "step": 6605
+    },
+    {
+      "epoch": 0.26750303520841767,
+      "grad_norm": 1.1030018329620361,
+      "learning_rate": 7.474705760892008e-05,
+      "loss": 1.026,
+      "step": 6610
+    },
+    {
+      "epoch": 0.26770538243626063,
+      "grad_norm": 1.1978520154953003,
+      "learning_rate": 7.472640925046459e-05,
+      "loss": 0.9951,
+      "step": 6615
+    },
+    {
+      "epoch": 0.2679077296641036,
+      "grad_norm": 1.2217199802398682,
+      "learning_rate": 7.470576089200909e-05,
+      "loss": 1.0317,
+      "step": 6620
+    },
+    {
+      "epoch": 0.2681100768919466,
+      "grad_norm": 1.103721261024475,
+      "learning_rate": 7.468511253355358e-05,
+      "loss": 1.0316,
+      "step": 6625
+    },
+    {
+      "epoch": 0.26831242411978956,
+      "grad_norm": 1.1881301403045654,
+      "learning_rate": 7.466446417509809e-05,
+      "loss": 0.9826,
+      "step": 6630
+    },
+    {
+      "epoch": 0.2685147713476325,
+      "grad_norm": 1.2023499011993408,
+      "learning_rate": 7.464381581664258e-05,
+      "loss": 1.0445,
+      "step": 6635
+    },
+    {
+      "epoch": 0.26871711857547553,
+      "grad_norm": 1.1933207511901855,
+      "learning_rate": 7.462316745818708e-05,
+      "loss": 0.9749,
+      "step": 6640
+    },
+    {
+      "epoch": 0.2689194658033185,
+      "grad_norm": 1.2708888053894043,
+      "learning_rate": 7.460251909973157e-05,
+      "loss": 1.0065,
+      "step": 6645
+    },
+    {
+      "epoch": 0.26912181303116145,
+      "grad_norm": 1.2242518663406372,
+      "learning_rate": 7.458187074127608e-05,
+      "loss": 1.0361,
+      "step": 6650
+    },
+    {
+      "epoch": 0.26932416025900446,
+      "grad_norm": 1.0936435461044312,
+      "learning_rate": 7.456122238282057e-05,
+      "loss": 1.0353,
+      "step": 6655
+    },
+    {
+      "epoch": 0.2695265074868474,
+      "grad_norm": 1.0418567657470703,
+      "learning_rate": 7.454057402436506e-05,
+      "loss": 1.0422,
+      "step": 6660
+    },
+    {
+      "epoch": 0.26972885471469044,
+      "grad_norm": 1.101579189300537,
+      "learning_rate": 7.451992566590956e-05,
+      "loss": 1.0011,
+      "step": 6665
+    },
+    {
+      "epoch": 0.2699312019425334,
+      "grad_norm": 1.1149858236312866,
+      "learning_rate": 7.449927730745406e-05,
+      "loss": 1.0366,
+      "step": 6670
+    },
+    {
+      "epoch": 0.27013354917037635,
+      "grad_norm": 1.158974528312683,
+      "learning_rate": 7.447862894899857e-05,
+      "loss": 1.0378,
+      "step": 6675
+    },
+    {
+      "epoch": 0.27033589639821937,
+      "grad_norm": 1.1632798910140991,
+      "learning_rate": 7.445798059054306e-05,
+      "loss": 1.01,
+      "step": 6680
+    },
+    {
+      "epoch": 0.2705382436260623,
+      "grad_norm": 1.1123712062835693,
+      "learning_rate": 7.443733223208755e-05,
+      "loss": 0.9903,
+      "step": 6685
+    },
+    {
+      "epoch": 0.2707405908539053,
+      "grad_norm": 1.2239755392074585,
+      "learning_rate": 7.441668387363205e-05,
+      "loss": 0.9864,
+      "step": 6690
+    },
+    {
+      "epoch": 0.2709429380817483,
+      "grad_norm": 1.1682302951812744,
+      "learning_rate": 7.439603551517654e-05,
+      "loss": 1.0237,
+      "step": 6695
+    },
+    {
+      "epoch": 0.27114528530959126,
+      "grad_norm": 1.0680562257766724,
+      "learning_rate": 7.437538715672105e-05,
+      "loss": 1.022,
+      "step": 6700
+    },
+    {
+      "epoch": 0.2713476325374342,
+      "grad_norm": 1.2862532138824463,
+      "learning_rate": 7.435473879826554e-05,
+      "loss": 1.0518,
+      "step": 6705
+    },
+    {
+      "epoch": 0.2715499797652772,
+      "grad_norm": 1.133471965789795,
+      "learning_rate": 7.433409043981003e-05,
+      "loss": 0.9525,
+      "step": 6710
+    },
+    {
+      "epoch": 0.2717523269931202,
+      "grad_norm": 1.1541469097137451,
+      "learning_rate": 7.431344208135453e-05,
+      "loss": 1.0238,
+      "step": 6715
+    },
+    {
+      "epoch": 0.2719546742209632,
+      "grad_norm": 1.2197669744491577,
+      "learning_rate": 7.429279372289904e-05,
+      "loss": 1.0381,
+      "step": 6720
+    },
+    {
+      "epoch": 0.27215702144880616,
+      "grad_norm": 1.1882134675979614,
+      "learning_rate": 7.427214536444353e-05,
+      "loss": 1.0635,
+      "step": 6725
+    },
+    {
+      "epoch": 0.2723593686766491,
+      "grad_norm": 1.17057204246521,
+      "learning_rate": 7.425149700598802e-05,
+      "loss": 1.0651,
+      "step": 6730
+    },
+    {
+      "epoch": 0.27256171590449213,
+      "grad_norm": 1.131810188293457,
+      "learning_rate": 7.423084864753252e-05,
+      "loss": 0.9965,
+      "step": 6735
+    },
+    {
+      "epoch": 0.2727640631323351,
+      "grad_norm": 1.1254119873046875,
+      "learning_rate": 7.421020028907703e-05,
+      "loss": 1.0012,
+      "step": 6740
+    },
+    {
+      "epoch": 0.27296641036017805,
+      "grad_norm": 1.146243929862976,
+      "learning_rate": 7.418955193062152e-05,
+      "loss": 1.0263,
+      "step": 6745
+    },
+    {
+      "epoch": 0.27316875758802106,
+      "grad_norm": 1.1290664672851562,
+      "learning_rate": 7.416890357216602e-05,
+      "loss": 1.0291,
+      "step": 6750
+    },
+    {
+      "epoch": 0.273371104815864,
+      "grad_norm": 1.1978673934936523,
+      "learning_rate": 7.414825521371051e-05,
+      "loss": 1.0562,
+      "step": 6755
+    },
+    {
+      "epoch": 0.273573452043707,
+      "grad_norm": 1.1925079822540283,
+      "learning_rate": 7.412760685525502e-05,
+      "loss": 1.0621,
+      "step": 6760
+    },
+    {
+      "epoch": 0.27377579927155,
+      "grad_norm": 1.1550565958023071,
+      "learning_rate": 7.41069584967995e-05,
+      "loss": 0.9679,
+      "step": 6765
+    },
+    {
+      "epoch": 0.27397814649939295,
+      "grad_norm": 1.2220369577407837,
+      "learning_rate": 7.408631013834401e-05,
+      "loss": 1.0226,
+      "step": 6770
+    },
+    {
+      "epoch": 0.27418049372723596,
+      "grad_norm": 1.0354042053222656,
+      "learning_rate": 7.40656617798885e-05,
+      "loss": 1.0115,
+      "step": 6775
+    },
+    {
+      "epoch": 0.2743828409550789,
+      "grad_norm": 1.2225804328918457,
+      "learning_rate": 7.404501342143299e-05,
+      "loss": 1.0814,
+      "step": 6780
+    },
+    {
+      "epoch": 0.2745851881829219,
+      "grad_norm": 1.1748751401901245,
+      "learning_rate": 7.40243650629775e-05,
+      "loss": 1.0224,
+      "step": 6785
+    },
+    {
+      "epoch": 0.2747875354107649,
+      "grad_norm": 1.0723620653152466,
+      "learning_rate": 7.4003716704522e-05,
+      "loss": 1.0631,
+      "step": 6790
+    },
+    {
+      "epoch": 0.27498988263860785,
+      "grad_norm": 1.219942569732666,
+      "learning_rate": 7.398306834606649e-05,
+      "loss": 1.0023,
+      "step": 6795
+    },
+    {
+      "epoch": 0.2751922298664508,
+      "grad_norm": 1.3363230228424072,
+      "learning_rate": 7.396241998761098e-05,
+      "loss": 0.9559,
+      "step": 6800
+    },
+    {
+      "epoch": 0.2753945770942938,
+      "grad_norm": 1.2138142585754395,
+      "learning_rate": 7.394177162915548e-05,
+      "loss": 1.0219,
+      "step": 6805
+    },
+    {
+      "epoch": 0.2755969243221368,
+      "grad_norm": 1.2388733625411987,
+      "learning_rate": 7.392112327069999e-05,
+      "loss": 0.9824,
+      "step": 6810
+    },
+    {
+      "epoch": 0.27579927154997974,
+      "grad_norm": 1.1982988119125366,
+      "learning_rate": 7.390047491224448e-05,
+      "loss": 0.9888,
+      "step": 6815
+    },
+    {
+      "epoch": 0.27600161877782275,
+      "grad_norm": 1.2767200469970703,
+      "learning_rate": 7.387982655378898e-05,
+      "loss": 0.9389,
+      "step": 6820
+    },
+    {
+      "epoch": 0.2762039660056657,
+      "grad_norm": 1.1729559898376465,
+      "learning_rate": 7.385917819533347e-05,
+      "loss": 1.0463,
+      "step": 6825
+    },
+    {
+      "epoch": 0.2764063132335087,
+      "grad_norm": 1.0749800205230713,
+      "learning_rate": 7.383852983687796e-05,
+      "loss": 0.9682,
+      "step": 6830
+    },
+    {
+      "epoch": 0.2766086604613517,
+      "grad_norm": 1.3590329885482788,
+      "learning_rate": 7.381788147842247e-05,
+      "loss": 1.0713,
+      "step": 6835
+    },
+    {
+      "epoch": 0.27681100768919464,
+      "grad_norm": 1.1369819641113281,
+      "learning_rate": 7.379723311996697e-05,
+      "loss": 1.0568,
+      "step": 6840
+    },
+    {
+      "epoch": 0.27701335491703766,
+      "grad_norm": 1.242820143699646,
+      "learning_rate": 7.377658476151146e-05,
+      "loss": 0.9609,
+      "step": 6845
+    },
+    {
+      "epoch": 0.2772157021448806,
+      "grad_norm": 1.1668909788131714,
+      "learning_rate": 7.375593640305595e-05,
+      "loss": 1.0397,
+      "step": 6850
+    },
+    {
+      "epoch": 0.2774180493727236,
+      "grad_norm": 1.2021713256835938,
+      "learning_rate": 7.373528804460046e-05,
+      "loss": 0.9497,
+      "step": 6855
+    },
+    {
+      "epoch": 0.2776203966005666,
+      "grad_norm": 1.191312313079834,
+      "learning_rate": 7.371463968614496e-05,
+      "loss": 0.9175,
+      "step": 6860
+    },
+    {
+      "epoch": 0.27782274382840955,
+      "grad_norm": 1.2159568071365356,
+      "learning_rate": 7.369399132768945e-05,
+      "loss": 1.038,
+      "step": 6865
+    },
+    {
+      "epoch": 0.2780250910562525,
+      "grad_norm": 1.1642060279846191,
+      "learning_rate": 7.367334296923394e-05,
+      "loss": 0.9647,
+      "step": 6870
+    },
+    {
+      "epoch": 0.2782274382840955,
+      "grad_norm": 1.1729024648666382,
+      "learning_rate": 7.365269461077845e-05,
+      "loss": 0.9741,
+      "step": 6875
+    },
+    {
+      "epoch": 0.2784297855119385,
+      "grad_norm": 1.2645635604858398,
+      "learning_rate": 7.363204625232294e-05,
+      "loss": 1.0328,
+      "step": 6880
+    },
+    {
+      "epoch": 0.2786321327397815,
+      "grad_norm": 1.1870768070220947,
+      "learning_rate": 7.361139789386744e-05,
+      "loss": 1.0114,
+      "step": 6885
+    },
+    {
+      "epoch": 0.27883447996762445,
+      "grad_norm": 1.1066211462020874,
+      "learning_rate": 7.359074953541195e-05,
+      "loss": 0.9975,
+      "step": 6890
+    },
+    {
+      "epoch": 0.2790368271954674,
+      "grad_norm": 1.148180603981018,
+      "learning_rate": 7.357010117695644e-05,
+      "loss": 0.9911,
+      "step": 6895
+    },
+    {
+      "epoch": 0.2792391744233104,
+      "grad_norm": 1.1885652542114258,
+      "learning_rate": 7.354945281850093e-05,
+      "loss": 1.0457,
+      "step": 6900
+    },
+    {
+      "epoch": 0.2794415216511534,
+      "grad_norm": 1.3599821329116821,
+      "learning_rate": 7.352880446004543e-05,
+      "loss": 1.0676,
+      "step": 6905
+    },
+    {
+      "epoch": 0.27964386887899634,
+      "grad_norm": 1.051513433456421,
+      "learning_rate": 7.350815610158993e-05,
+      "loss": 0.9989,
+      "step": 6910
+    },
+    {
+      "epoch": 0.27984621610683935,
+      "grad_norm": 1.2398626804351807,
+      "learning_rate": 7.348750774313443e-05,
+      "loss": 1.0297,
+      "step": 6915
+    },
+    {
+      "epoch": 0.2800485633346823,
+      "grad_norm": 1.2511979341506958,
+      "learning_rate": 7.346685938467892e-05,
+      "loss": 0.9924,
+      "step": 6920
+    },
+    {
+      "epoch": 0.28025091056252527,
+      "grad_norm": 1.1494407653808594,
+      "learning_rate": 7.344621102622342e-05,
+      "loss": 0.9201,
+      "step": 6925
+    },
+    {
+      "epoch": 0.2804532577903683,
+      "grad_norm": 1.199025273323059,
+      "learning_rate": 7.342556266776791e-05,
+      "loss": 0.9594,
+      "step": 6930
+    },
+    {
+      "epoch": 0.28065560501821124,
+      "grad_norm": 1.1938326358795166,
+      "learning_rate": 7.340491430931241e-05,
+      "loss": 0.9798,
+      "step": 6935
+    },
+    {
+      "epoch": 0.28085795224605425,
+      "grad_norm": 1.061036229133606,
+      "learning_rate": 7.33842659508569e-05,
+      "loss": 0.981,
+      "step": 6940
+    },
+    {
+      "epoch": 0.2810602994738972,
+      "grad_norm": 1.1499427556991577,
+      "learning_rate": 7.336361759240141e-05,
+      "loss": 0.9694,
+      "step": 6945
+    },
+    {
+      "epoch": 0.28126264670174017,
+      "grad_norm": 1.351099967956543,
+      "learning_rate": 7.33429692339459e-05,
+      "loss": 1.0097,
+      "step": 6950
+    },
+    {
+      "epoch": 0.2814649939295832,
+      "grad_norm": 1.0928959846496582,
+      "learning_rate": 7.33223208754904e-05,
+      "loss": 1.0144,
+      "step": 6955
+    },
+    {
+      "epoch": 0.28166734115742614,
+      "grad_norm": 1.2172058820724487,
+      "learning_rate": 7.330167251703491e-05,
+      "loss": 1.0061,
+      "step": 6960
+    },
+    {
+      "epoch": 0.2818696883852691,
+      "grad_norm": 1.1754196882247925,
+      "learning_rate": 7.32810241585794e-05,
+      "loss": 1.0083,
+      "step": 6965
+    },
+    {
+      "epoch": 0.2820720356131121,
+      "grad_norm": 1.2395962476730347,
+      "learning_rate": 7.326037580012389e-05,
+      "loss": 1.0243,
+      "step": 6970
+    },
+    {
+      "epoch": 0.2822743828409551,
+      "grad_norm": 1.160964012145996,
+      "learning_rate": 7.323972744166839e-05,
+      "loss": 1.003,
+      "step": 6975
+    },
+    {
+      "epoch": 0.28247673006879803,
+      "grad_norm": 1.1365638971328735,
+      "learning_rate": 7.321907908321288e-05,
+      "loss": 1.0273,
+      "step": 6980
+    },
+    {
+      "epoch": 0.28267907729664105,
+      "grad_norm": 1.196136713027954,
+      "learning_rate": 7.319843072475739e-05,
+      "loss": 1.0482,
+      "step": 6985
+    },
+    {
+      "epoch": 0.282881424524484,
+      "grad_norm": 1.2548751831054688,
+      "learning_rate": 7.317778236630188e-05,
+      "loss": 1.0182,
+      "step": 6990
+    },
+    {
+      "epoch": 0.283083771752327,
+      "grad_norm": 1.1191009283065796,
+      "learning_rate": 7.315713400784638e-05,
+      "loss": 1.0106,
+      "step": 6995
+    },
+    {
+      "epoch": 0.28328611898017,
+      "grad_norm": 1.2741000652313232,
+      "learning_rate": 7.313648564939087e-05,
+      "loss": 1.037,
+      "step": 7000
+    },
+    {
+      "epoch": 0.28348846620801293,
+      "grad_norm": 1.0600517988204956,
+      "learning_rate": 7.311583729093538e-05,
+      "loss": 0.9773,
+      "step": 7005
+    },
+    {
+      "epoch": 0.28369081343585595,
+      "grad_norm": 1.126713514328003,
+      "learning_rate": 7.309518893247988e-05,
+      "loss": 1.066,
+      "step": 7010
+    },
+    {
+      "epoch": 0.2838931606636989,
+      "grad_norm": 1.0954577922821045,
+      "learning_rate": 7.307454057402436e-05,
+      "loss": 1.019,
+      "step": 7015
+    },
+    {
+      "epoch": 0.28409550789154187,
+      "grad_norm": 1.3487249612808228,
+      "learning_rate": 7.305389221556886e-05,
+      "loss": 1.0658,
+      "step": 7020
+    },
+    {
+      "epoch": 0.2842978551193849,
+      "grad_norm": 1.0767533779144287,
+      "learning_rate": 7.303324385711337e-05,
+      "loss": 0.9997,
+      "step": 7025
+    },
+    {
+      "epoch": 0.28450020234722784,
+      "grad_norm": 1.2380719184875488,
+      "learning_rate": 7.301259549865787e-05,
+      "loss": 1.0391,
+      "step": 7030
+    },
+    {
+      "epoch": 0.2847025495750708,
+      "grad_norm": 1.1037429571151733,
+      "learning_rate": 7.299194714020236e-05,
+      "loss": 0.9583,
+      "step": 7035
+    },
+    {
+      "epoch": 0.2849048968029138,
+      "grad_norm": 1.2484043836593628,
+      "learning_rate": 7.297129878174685e-05,
+      "loss": 1.0238,
+      "step": 7040
+    },
+    {
+      "epoch": 0.28510724403075677,
+      "grad_norm": 1.1408405303955078,
+      "learning_rate": 7.295065042329136e-05,
+      "loss": 1.06,
+      "step": 7045
+    },
+    {
+      "epoch": 0.2853095912585998,
+      "grad_norm": 1.1027109622955322,
+      "learning_rate": 7.293000206483585e-05,
+      "loss": 0.977,
+      "step": 7050
+    },
+    {
+      "epoch": 0.28551193848644274,
+      "grad_norm": 1.1841386556625366,
+      "learning_rate": 7.290935370638035e-05,
+      "loss": 1.0146,
+      "step": 7055
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 1.19268000125885,
+      "learning_rate": 7.288870534792484e-05,
+      "loss": 1.0176,
+      "step": 7060
+    },
+    {
+      "epoch": 0.2859166329421287,
+      "grad_norm": 1.1941889524459839,
+      "learning_rate": 7.286805698946933e-05,
+      "loss": 1.0325,
+      "step": 7065
+    },
+    {
+      "epoch": 0.28611898016997167,
+      "grad_norm": 1.0459282398223877,
+      "learning_rate": 7.284740863101383e-05,
+      "loss": 0.9613,
+      "step": 7070
+    },
+    {
+      "epoch": 0.28632132739781463,
+      "grad_norm": 1.2329634428024292,
+      "learning_rate": 7.282676027255834e-05,
+      "loss": 1.0587,
+      "step": 7075
+    },
+    {
+      "epoch": 0.28652367462565764,
+      "grad_norm": 1.180199384689331,
+      "learning_rate": 7.280611191410284e-05,
+      "loss": 0.9983,
+      "step": 7080
+    },
+    {
+      "epoch": 0.2867260218535006,
+      "grad_norm": 1.201682448387146,
+      "learning_rate": 7.278546355564732e-05,
+      "loss": 0.9952,
+      "step": 7085
+    },
+    {
+      "epoch": 0.28692836908134356,
+      "grad_norm": 1.1668422222137451,
+      "learning_rate": 7.276481519719182e-05,
+      "loss": 0.9602,
+      "step": 7090
+    },
+    {
+      "epoch": 0.2871307163091866,
+      "grad_norm": 1.1423628330230713,
+      "learning_rate": 7.274416683873633e-05,
+      "loss": 1.0044,
+      "step": 7095
+    },
+    {
+      "epoch": 0.28733306353702953,
+      "grad_norm": 1.217802882194519,
+      "learning_rate": 7.272351848028082e-05,
+      "loss": 1.0701,
+      "step": 7100
+    },
+    {
+      "epoch": 0.28753541076487255,
+      "grad_norm": 0.9813905954360962,
+      "learning_rate": 7.270287012182532e-05,
+      "loss": 1.03,
+      "step": 7105
+    },
+    {
+      "epoch": 0.2877377579927155,
+      "grad_norm": 1.2254445552825928,
+      "learning_rate": 7.268222176336981e-05,
+      "loss": 1.0226,
+      "step": 7110
+    },
+    {
+      "epoch": 0.28794010522055846,
+      "grad_norm": 1.2287665605545044,
+      "learning_rate": 7.266157340491432e-05,
+      "loss": 1.037,
+      "step": 7115
+    },
+    {
+      "epoch": 0.2881424524484015,
+      "grad_norm": 1.2598094940185547,
+      "learning_rate": 7.264092504645881e-05,
+      "loss": 0.968,
+      "step": 7120
+    },
+    {
+      "epoch": 0.28834479967624443,
+      "grad_norm": 1.1876108646392822,
+      "learning_rate": 7.262027668800331e-05,
+      "loss": 1.0464,
+      "step": 7125
+    },
+    {
+      "epoch": 0.2885471469040874,
+      "grad_norm": 1.1915405988693237,
+      "learning_rate": 7.25996283295478e-05,
+      "loss": 0.98,
+      "step": 7130
+    },
+    {
+      "epoch": 0.2887494941319304,
+      "grad_norm": 1.2311309576034546,
+      "learning_rate": 7.257897997109229e-05,
+      "loss": 1.0208,
+      "step": 7135
+    },
+    {
+      "epoch": 0.28895184135977336,
+      "grad_norm": 1.3213067054748535,
+      "learning_rate": 7.25583316126368e-05,
+      "loss": 0.9752,
+      "step": 7140
+    },
+    {
+      "epoch": 0.2891541885876163,
+      "grad_norm": 1.1361618041992188,
+      "learning_rate": 7.25376832541813e-05,
+      "loss": 0.9961,
+      "step": 7145
+    },
+    {
+      "epoch": 0.28935653581545934,
+      "grad_norm": 1.1876906156539917,
+      "learning_rate": 7.251703489572579e-05,
+      "loss": 0.9563,
+      "step": 7150
+    },
+    {
+      "epoch": 0.2895588830433023,
+      "grad_norm": 1.2165839672088623,
+      "learning_rate": 7.249638653727028e-05,
+      "loss": 1.0085,
+      "step": 7155
+    },
+    {
+      "epoch": 0.2897612302711453,
+      "grad_norm": 1.0741130113601685,
+      "learning_rate": 7.247573817881479e-05,
+      "loss": 1.0152,
+      "step": 7160
+    },
+    {
+      "epoch": 0.28996357749898827,
+      "grad_norm": 1.1732629537582397,
+      "learning_rate": 7.245508982035929e-05,
+      "loss": 1.0184,
+      "step": 7165
+    },
+    {
+      "epoch": 0.2901659247268312,
+      "grad_norm": 1.1553676128387451,
+      "learning_rate": 7.243444146190378e-05,
+      "loss": 0.9966,
+      "step": 7170
+    },
+    {
+      "epoch": 0.29036827195467424,
+      "grad_norm": 1.230701208114624,
+      "learning_rate": 7.241379310344828e-05,
+      "loss": 1.0381,
+      "step": 7175
+    },
+    {
+      "epoch": 0.2905706191825172,
+      "grad_norm": 1.1343389749526978,
+      "learning_rate": 7.239314474499278e-05,
+      "loss": 1.0462,
+      "step": 7180
+    },
+    {
+      "epoch": 0.29077296641036016,
+      "grad_norm": 1.1028715372085571,
+      "learning_rate": 7.237249638653727e-05,
+      "loss": 1.0403,
+      "step": 7185
+    },
+    {
+      "epoch": 0.29097531363820317,
+      "grad_norm": 1.2318840026855469,
+      "learning_rate": 7.235184802808177e-05,
+      "loss": 0.9667,
+      "step": 7190
+    },
+    {
+      "epoch": 0.29117766086604613,
+      "grad_norm": 1.1649657487869263,
+      "learning_rate": 7.233119966962627e-05,
+      "loss": 1.0312,
+      "step": 7195
+    },
+    {
+      "epoch": 0.2913800080938891,
+      "grad_norm": 1.1640312671661377,
+      "learning_rate": 7.231055131117076e-05,
+      "loss": 1.0257,
+      "step": 7200
+    },
+    {
+      "epoch": 0.2915823553217321,
+      "grad_norm": 1.3711130619049072,
+      "learning_rate": 7.228990295271526e-05,
+      "loss": 1.0607,
+      "step": 7205
+    },
+    {
+      "epoch": 0.29178470254957506,
+      "grad_norm": 1.1209375858306885,
+      "learning_rate": 7.226925459425976e-05,
+      "loss": 0.9962,
+      "step": 7210
+    },
+    {
+      "epoch": 0.2919870497774181,
+      "grad_norm": 1.1353542804718018,
+      "learning_rate": 7.224860623580426e-05,
+      "loss": 0.9753,
+      "step": 7215
+    },
+    {
+      "epoch": 0.29218939700526103,
+      "grad_norm": 1.2244285345077515,
+      "learning_rate": 7.222795787734875e-05,
+      "loss": 1.0051,
+      "step": 7220
+    },
+    {
+      "epoch": 0.292391744233104,
+      "grad_norm": 1.119053602218628,
+      "learning_rate": 7.220730951889326e-05,
+      "loss": 1.0178,
+      "step": 7225
+    },
+    {
+      "epoch": 0.292594091460947,
+      "grad_norm": 1.1514949798583984,
+      "learning_rate": 7.218666116043775e-05,
+      "loss": 1.0123,
+      "step": 7230
+    },
+    {
+      "epoch": 0.29279643868878996,
+      "grad_norm": 1.1903537511825562,
+      "learning_rate": 7.216601280198224e-05,
+      "loss": 0.9997,
+      "step": 7235
+    },
+    {
+      "epoch": 0.2929987859166329,
+      "grad_norm": 1.1025440692901611,
+      "learning_rate": 7.214536444352674e-05,
+      "loss": 1.0114,
+      "step": 7240
+    },
+    {
+      "epoch": 0.29320113314447593,
+      "grad_norm": 1.1389793157577515,
+      "learning_rate": 7.212471608507125e-05,
+      "loss": 0.9734,
+      "step": 7245
+    },
+    {
+      "epoch": 0.2934034803723189,
+      "grad_norm": 1.1365653276443481,
+      "learning_rate": 7.210406772661574e-05,
+      "loss": 0.9598,
+      "step": 7250
+    },
+    {
+      "epoch": 0.29360582760016185,
+      "grad_norm": 1.265367031097412,
+      "learning_rate": 7.208341936816023e-05,
+      "loss": 1.0125,
+      "step": 7255
+    },
+    {
+      "epoch": 0.29380817482800486,
+      "grad_norm": 1.2603001594543457,
+      "learning_rate": 7.206277100970473e-05,
+      "loss": 1.028,
+      "step": 7260
+    },
+    {
+      "epoch": 0.2940105220558478,
+      "grad_norm": 1.2002099752426147,
+      "learning_rate": 7.204212265124924e-05,
+      "loss": 0.9804,
+      "step": 7265
+    },
+    {
+      "epoch": 0.29421286928369084,
+      "grad_norm": 1.0949093103408813,
+      "learning_rate": 7.202147429279373e-05,
+      "loss": 0.9647,
+      "step": 7270
+    },
+    {
+      "epoch": 0.2944152165115338,
+      "grad_norm": 1.195500373840332,
+      "learning_rate": 7.200082593433822e-05,
+      "loss": 1.012,
+      "step": 7275
+    },
+    {
+      "epoch": 0.29461756373937675,
+      "grad_norm": 1.0756292343139648,
+      "learning_rate": 7.198017757588272e-05,
+      "loss": 1.0446,
+      "step": 7280
+    },
+    {
+      "epoch": 0.29481991096721977,
+      "grad_norm": 1.1095898151397705,
+      "learning_rate": 7.195952921742721e-05,
+      "loss": 1.055,
+      "step": 7285
+    },
+    {
+      "epoch": 0.2950222581950627,
+      "grad_norm": 1.0555248260498047,
+      "learning_rate": 7.193888085897172e-05,
+      "loss": 0.9849,
+      "step": 7290
+    },
+    {
+      "epoch": 0.2952246054229057,
+      "grad_norm": 1.2022385597229004,
+      "learning_rate": 7.191823250051622e-05,
+      "loss": 1.0044,
+      "step": 7295
+    },
+    {
+      "epoch": 0.2954269526507487,
+      "grad_norm": 1.196317195892334,
+      "learning_rate": 7.189758414206071e-05,
+      "loss": 1.0223,
+      "step": 7300
+    },
+    {
+      "epoch": 0.29562929987859166,
+      "grad_norm": 1.2295843362808228,
+      "learning_rate": 7.18769357836052e-05,
+      "loss": 1.0401,
+      "step": 7305
+    },
+    {
+      "epoch": 0.2958316471064346,
+      "grad_norm": 1.1430981159210205,
+      "learning_rate": 7.18562874251497e-05,
+      "loss": 1.0177,
+      "step": 7310
+    },
+    {
+      "epoch": 0.29603399433427763,
+      "grad_norm": 1.2077337503433228,
+      "learning_rate": 7.183563906669421e-05,
+      "loss": 1.0704,
+      "step": 7315
+    },
+    {
+      "epoch": 0.2962363415621206,
+      "grad_norm": 1.1060575246810913,
+      "learning_rate": 7.18149907082387e-05,
+      "loss": 1.0082,
+      "step": 7320
+    },
+    {
+      "epoch": 0.2964386887899636,
+      "grad_norm": 1.1972790956497192,
+      "learning_rate": 7.179434234978319e-05,
+      "loss": 0.9948,
+      "step": 7325
+    },
+    {
+      "epoch": 0.29664103601780656,
+      "grad_norm": 1.2054885625839233,
+      "learning_rate": 7.17736939913277e-05,
+      "loss": 0.9546,
+      "step": 7330
+    },
+    {
+      "epoch": 0.2968433832456495,
+      "grad_norm": 1.1200127601623535,
+      "learning_rate": 7.17530456328722e-05,
+      "loss": 1.041,
+      "step": 7335
+    },
+    {
+      "epoch": 0.29704573047349253,
+      "grad_norm": 1.0624228715896606,
+      "learning_rate": 7.173239727441669e-05,
+      "loss": 1.039,
+      "step": 7340
+    },
+    {
+      "epoch": 0.2972480777013355,
+      "grad_norm": 1.0303196907043457,
+      "learning_rate": 7.171174891596118e-05,
+      "loss": 1.1089,
+      "step": 7345
+    },
+    {
+      "epoch": 0.29745042492917845,
+      "grad_norm": 1.1626769304275513,
+      "learning_rate": 7.169110055750568e-05,
+      "loss": 1.0513,
+      "step": 7350
+    },
+    {
+      "epoch": 0.29765277215702146,
+      "grad_norm": 1.2582601308822632,
+      "learning_rate": 7.167045219905017e-05,
+      "loss": 0.9749,
+      "step": 7355
+    },
+    {
+      "epoch": 0.2978551193848644,
+      "grad_norm": 1.1853293180465698,
+      "learning_rate": 7.164980384059468e-05,
+      "loss": 0.9623,
+      "step": 7360
+    },
+    {
+      "epoch": 0.29805746661270743,
+      "grad_norm": 1.2648826837539673,
+      "learning_rate": 7.162915548213918e-05,
+      "loss": 1.088,
+      "step": 7365
+    },
+    {
+      "epoch": 0.2982598138405504,
+      "grad_norm": 1.1166695356369019,
+      "learning_rate": 7.160850712368366e-05,
+      "loss": 1.0096,
+      "step": 7370
+    },
+    {
+      "epoch": 0.29846216106839335,
+      "grad_norm": 1.2206895351409912,
+      "learning_rate": 7.158785876522816e-05,
+      "loss": 1.0238,
+      "step": 7375
+    },
+    {
+      "epoch": 0.29866450829623636,
+      "grad_norm": 1.195966362953186,
+      "learning_rate": 7.156721040677267e-05,
+      "loss": 1.0151,
+      "step": 7380
+    },
+    {
+      "epoch": 0.2988668555240793,
+      "grad_norm": 1.0893754959106445,
+      "learning_rate": 7.154656204831717e-05,
+      "loss": 0.9902,
+      "step": 7385
+    },
+    {
+      "epoch": 0.2990692027519223,
+      "grad_norm": 1.1973437070846558,
+      "learning_rate": 7.152591368986166e-05,
+      "loss": 0.9808,
+      "step": 7390
+    },
+    {
+      "epoch": 0.2992715499797653,
+      "grad_norm": 1.2002286911010742,
+      "learning_rate": 7.150526533140615e-05,
+      "loss": 0.9959,
+      "step": 7395
+    },
+    {
+      "epoch": 0.29947389720760825,
+      "grad_norm": 1.122012972831726,
+      "learning_rate": 7.148461697295066e-05,
+      "loss": 0.9995,
+      "step": 7400
+    },
+    {
+      "epoch": 0.2996762444354512,
+      "grad_norm": 1.1145695447921753,
+      "learning_rate": 7.146396861449515e-05,
+      "loss": 1.0395,
+      "step": 7405
+    },
+    {
+      "epoch": 0.2998785916632942,
+      "grad_norm": 1.151845097541809,
+      "learning_rate": 7.144332025603965e-05,
+      "loss": 1.0706,
+      "step": 7410
+    },
+    {
+      "epoch": 0.3000809388911372,
+      "grad_norm": 1.1342051029205322,
+      "learning_rate": 7.142267189758414e-05,
+      "loss": 1.0394,
+      "step": 7415
+    },
+    {
+      "epoch": 0.3002832861189802,
+      "grad_norm": 1.1038486957550049,
+      "learning_rate": 7.140202353912863e-05,
+      "loss": 0.9879,
+      "step": 7420
+    },
+    {
+      "epoch": 0.30048563334682316,
+      "grad_norm": 1.143272876739502,
+      "learning_rate": 7.138137518067314e-05,
+      "loss": 1.0117,
+      "step": 7425
+    },
+    {
+      "epoch": 0.3006879805746661,
+      "grad_norm": 1.2253665924072266,
+      "learning_rate": 7.136072682221764e-05,
+      "loss": 1.0272,
+      "step": 7430
+    },
+    {
+      "epoch": 0.3008903278025091,
+      "grad_norm": 1.1236751079559326,
+      "learning_rate": 7.134007846376214e-05,
+      "loss": 1.0089,
+      "step": 7435
+    },
+    {
+      "epoch": 0.3010926750303521,
+      "grad_norm": 1.4188357591629028,
+      "learning_rate": 7.131943010530663e-05,
+      "loss": 1.0131,
+      "step": 7440
+    },
+    {
+      "epoch": 0.30129502225819504,
+      "grad_norm": 1.224983811378479,
+      "learning_rate": 7.129878174685113e-05,
+      "loss": 1.0096,
+      "step": 7445
+    },
+    {
+      "epoch": 0.30149736948603806,
+      "grad_norm": 1.0498284101486206,
+      "learning_rate": 7.127813338839563e-05,
+      "loss": 1.0623,
+      "step": 7450
+    },
+    {
+      "epoch": 0.301699716713881,
+      "grad_norm": 1.3433605432510376,
+      "learning_rate": 7.125748502994012e-05,
+      "loss": 1.0363,
+      "step": 7455
+    },
+    {
+      "epoch": 0.301902063941724,
+      "grad_norm": 1.3511546850204468,
+      "learning_rate": 7.123683667148462e-05,
+      "loss": 1.023,
+      "step": 7460
+    },
+    {
+      "epoch": 0.302104411169567,
+      "grad_norm": 1.1709057092666626,
+      "learning_rate": 7.121618831302911e-05,
+      "loss": 1.0136,
+      "step": 7465
+    },
+    {
+      "epoch": 0.30230675839740995,
+      "grad_norm": 1.1816767454147339,
+      "learning_rate": 7.119553995457362e-05,
+      "loss": 0.9868,
+      "step": 7470
+    },
+    {
+      "epoch": 0.30250910562525296,
+      "grad_norm": 1.1204698085784912,
+      "learning_rate": 7.117489159611811e-05,
+      "loss": 1.0717,
+      "step": 7475
+    },
+    {
+      "epoch": 0.3027114528530959,
+      "grad_norm": 1.2430353164672852,
+      "learning_rate": 7.115424323766261e-05,
+      "loss": 1.0317,
+      "step": 7480
+    },
+    {
+      "epoch": 0.3029138000809389,
+      "grad_norm": 1.0575361251831055,
+      "learning_rate": 7.11335948792071e-05,
+      "loss": 0.9902,
+      "step": 7485
+    },
+    {
+      "epoch": 0.3031161473087819,
+      "grad_norm": 1.0389426946640015,
+      "learning_rate": 7.11129465207516e-05,
+      "loss": 1.0214,
+      "step": 7490
+    },
+    {
+      "epoch": 0.30331849453662485,
+      "grad_norm": 1.0878791809082031,
+      "learning_rate": 7.10922981622961e-05,
+      "loss": 0.9587,
+      "step": 7495
+    },
+    {
+      "epoch": 0.3035208417644678,
+      "grad_norm": 1.1660854816436768,
+      "learning_rate": 7.10716498038406e-05,
+      "loss": 0.9606,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3037231889923108,
+      "grad_norm": 1.1651402711868286,
+      "learning_rate": 7.105100144538509e-05,
+      "loss": 1.0076,
+      "step": 7505
+    },
+    {
+      "epoch": 0.3039255362201538,
+      "grad_norm": 1.248314380645752,
+      "learning_rate": 7.10303530869296e-05,
+      "loss": 1.0847,
+      "step": 7510
+    },
+    {
+      "epoch": 0.30412788344799674,
+      "grad_norm": 1.1461608409881592,
+      "learning_rate": 7.100970472847409e-05,
+      "loss": 1.0003,
+      "step": 7515
+    },
+    {
+      "epoch": 0.30433023067583975,
+      "grad_norm": 1.2227619886398315,
+      "learning_rate": 7.098905637001859e-05,
+      "loss": 1.0442,
+      "step": 7520
+    },
+    {
+      "epoch": 0.3045325779036827,
+      "grad_norm": 1.2554718255996704,
+      "learning_rate": 7.096840801156308e-05,
+      "loss": 1.0341,
+      "step": 7525
+    },
+    {
+      "epoch": 0.3047349251315257,
+      "grad_norm": 1.1214122772216797,
+      "learning_rate": 7.094775965310759e-05,
+      "loss": 0.9841,
+      "step": 7530
+    },
+    {
+      "epoch": 0.3049372723593687,
+      "grad_norm": 1.2532953023910522,
+      "learning_rate": 7.092711129465208e-05,
+      "loss": 1.0041,
+      "step": 7535
+    },
+    {
+      "epoch": 0.30513961958721164,
+      "grad_norm": 1.1885756254196167,
+      "learning_rate": 7.090646293619657e-05,
+      "loss": 1.0439,
+      "step": 7540
+    },
+    {
+      "epoch": 0.30534196681505466,
+      "grad_norm": 1.074864149093628,
+      "learning_rate": 7.088581457774107e-05,
+      "loss": 1.0343,
+      "step": 7545
+    },
+    {
+      "epoch": 0.3055443140428976,
+      "grad_norm": 1.1654303073883057,
+      "learning_rate": 7.086516621928558e-05,
+      "loss": 0.9469,
+      "step": 7550
+    },
+    {
+      "epoch": 0.30574666127074057,
+      "grad_norm": 1.1716253757476807,
+      "learning_rate": 7.084451786083007e-05,
+      "loss": 1.0065,
+      "step": 7555
+    },
+    {
+      "epoch": 0.3059490084985836,
+      "grad_norm": 1.120162844657898,
+      "learning_rate": 7.082386950237456e-05,
+      "loss": 1.0365,
+      "step": 7560
+    },
+    {
+      "epoch": 0.30615135572642654,
+      "grad_norm": 1.263826847076416,
+      "learning_rate": 7.080322114391906e-05,
+      "loss": 1.0265,
+      "step": 7565
+    },
+    {
+      "epoch": 0.3063537029542695,
+      "grad_norm": 1.2363839149475098,
+      "learning_rate": 7.078257278546356e-05,
+      "loss": 0.9972,
+      "step": 7570
+    },
+    {
+      "epoch": 0.3065560501821125,
+      "grad_norm": 1.1157922744750977,
+      "learning_rate": 7.076192442700806e-05,
+      "loss": 0.9479,
+      "step": 7575
+    },
+    {
+      "epoch": 0.3067583974099555,
+      "grad_norm": 1.2315844297409058,
+      "learning_rate": 7.074127606855256e-05,
+      "loss": 0.995,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3069607446377985,
+      "grad_norm": 1.3027738332748413,
+      "learning_rate": 7.072062771009705e-05,
+      "loss": 0.9842,
+      "step": 7585
+    },
+    {
+      "epoch": 0.30716309186564145,
+      "grad_norm": 1.1640541553497314,
+      "learning_rate": 7.069997935164154e-05,
+      "loss": 1.0227,
+      "step": 7590
+    },
+    {
+      "epoch": 0.3073654390934844,
+      "grad_norm": 1.2683963775634766,
+      "learning_rate": 7.067933099318604e-05,
+      "loss": 1.0546,
+      "step": 7595
+    },
+    {
+      "epoch": 0.3075677863213274,
+      "grad_norm": 1.0934149026870728,
+      "learning_rate": 7.065868263473055e-05,
+      "loss": 0.959,
+      "step": 7600
+    },
+    {
+      "epoch": 0.3077701335491704,
+      "grad_norm": 1.579567551612854,
+      "learning_rate": 7.063803427627504e-05,
+      "loss": 0.981,
+      "step": 7605
+    },
+    {
+      "epoch": 0.30797248077701334,
+      "grad_norm": 1.078355073928833,
+      "learning_rate": 7.061738591781953e-05,
+      "loss": 1.0538,
+      "step": 7610
+    },
+    {
+      "epoch": 0.30817482800485635,
+      "grad_norm": 1.025793433189392,
+      "learning_rate": 7.059673755936403e-05,
+      "loss": 1.0032,
+      "step": 7615
+    },
+    {
+      "epoch": 0.3083771752326993,
+      "grad_norm": 1.0965713262557983,
+      "learning_rate": 7.057608920090854e-05,
+      "loss": 1.0416,
+      "step": 7620
+    },
+    {
+      "epoch": 0.30857952246054227,
+      "grad_norm": 1.094805121421814,
+      "learning_rate": 7.055544084245303e-05,
+      "loss": 1.0167,
+      "step": 7625
+    },
+    {
+      "epoch": 0.3087818696883853,
+      "grad_norm": 1.0859425067901611,
+      "learning_rate": 7.053479248399752e-05,
+      "loss": 1.031,
+      "step": 7630
+    },
+    {
+      "epoch": 0.30898421691622824,
+      "grad_norm": 1.1122045516967773,
+      "learning_rate": 7.051414412554202e-05,
+      "loss": 0.9729,
+      "step": 7635
+    },
+    {
+      "epoch": 0.30918656414407125,
+      "grad_norm": 1.1862298250198364,
+      "learning_rate": 7.049349576708651e-05,
+      "loss": 0.9737,
+      "step": 7640
+    },
+    {
+      "epoch": 0.3093889113719142,
+      "grad_norm": 1.2225682735443115,
+      "learning_rate": 7.047284740863102e-05,
+      "loss": 0.983,
+      "step": 7645
+    },
+    {
+      "epoch": 0.30959125859975717,
+      "grad_norm": 1.1471396684646606,
+      "learning_rate": 7.045219905017552e-05,
+      "loss": 1.0125,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3097936058276002,
+      "grad_norm": 1.17530357837677,
+      "learning_rate": 7.043155069172001e-05,
+      "loss": 0.9278,
+      "step": 7655
+    },
+    {
+      "epoch": 0.30999595305544314,
+      "grad_norm": 1.224523901939392,
+      "learning_rate": 7.04109023332645e-05,
+      "loss": 1.0306,
+      "step": 7660
+    },
+    {
+      "epoch": 0.3101983002832861,
+      "grad_norm": 1.2725141048431396,
+      "learning_rate": 7.0390253974809e-05,
+      "loss": 0.9934,
+      "step": 7665
+    },
+    {
+      "epoch": 0.3104006475111291,
+      "grad_norm": 1.2264436483383179,
+      "learning_rate": 7.036960561635351e-05,
+      "loss": 1.0184,
+      "step": 7670
+    },
+    {
+      "epoch": 0.31060299473897207,
+      "grad_norm": 1.1794722080230713,
+      "learning_rate": 7.0348957257898e-05,
+      "loss": 0.9811,
+      "step": 7675
+    },
+    {
+      "epoch": 0.31080534196681503,
+      "grad_norm": 1.1239395141601562,
+      "learning_rate": 7.032830889944249e-05,
+      "loss": 0.9433,
+      "step": 7680
+    },
+    {
+      "epoch": 0.31100768919465804,
+      "grad_norm": 1.1605916023254395,
+      "learning_rate": 7.0307660540987e-05,
+      "loss": 0.9934,
+      "step": 7685
+    },
+    {
+      "epoch": 0.311210036422501,
+      "grad_norm": 1.1802419424057007,
+      "learning_rate": 7.02870121825315e-05,
+      "loss": 1.03,
+      "step": 7690
+    },
+    {
+      "epoch": 0.311412383650344,
+      "grad_norm": 1.2433397769927979,
+      "learning_rate": 7.026636382407599e-05,
+      "loss": 0.9642,
+      "step": 7695
+    },
+    {
+      "epoch": 0.311614730878187,
+      "grad_norm": 1.1377280950546265,
+      "learning_rate": 7.024571546562048e-05,
+      "loss": 1.0044,
+      "step": 7700
+    },
+    {
+      "epoch": 0.31181707810602993,
+      "grad_norm": 1.1464645862579346,
+      "learning_rate": 7.022506710716498e-05,
+      "loss": 0.9674,
+      "step": 7705
+    },
+    {
+      "epoch": 0.31201942533387295,
+      "grad_norm": 1.0886270999908447,
+      "learning_rate": 7.020441874870948e-05,
+      "loss": 0.9622,
+      "step": 7710
+    },
+    {
+      "epoch": 0.3122217725617159,
+      "grad_norm": 1.2346837520599365,
+      "learning_rate": 7.018377039025398e-05,
+      "loss": 0.9845,
+      "step": 7715
+    },
+    {
+      "epoch": 0.31242411978955886,
+      "grad_norm": 1.0920614004135132,
+      "learning_rate": 7.016312203179848e-05,
+      "loss": 1.0288,
+      "step": 7720
+    },
+    {
+      "epoch": 0.3126264670174019,
+      "grad_norm": 1.0794581174850464,
+      "learning_rate": 7.014247367334297e-05,
+      "loss": 1.0598,
+      "step": 7725
+    },
+    {
+      "epoch": 0.31282881424524484,
+      "grad_norm": 1.2226083278656006,
+      "learning_rate": 7.012182531488746e-05,
+      "loss": 0.9955,
+      "step": 7730
+    },
+    {
+      "epoch": 0.3130311614730878,
+      "grad_norm": 1.1473082304000854,
+      "learning_rate": 7.010117695643197e-05,
+      "loss": 1.0104,
+      "step": 7735
+    },
+    {
+      "epoch": 0.3132335087009308,
+      "grad_norm": 1.152406930923462,
+      "learning_rate": 7.008052859797647e-05,
+      "loss": 0.99,
+      "step": 7740
+    },
+    {
+      "epoch": 0.31343585592877377,
+      "grad_norm": 1.3039500713348389,
+      "learning_rate": 7.005988023952096e-05,
+      "loss": 1.0358,
+      "step": 7745
+    },
+    {
+      "epoch": 0.3136382031566168,
+      "grad_norm": 1.223159670829773,
+      "learning_rate": 7.003923188106545e-05,
+      "loss": 1.0109,
+      "step": 7750
+    },
+    {
+      "epoch": 0.31384055038445974,
+      "grad_norm": 1.1193894147872925,
+      "learning_rate": 7.001858352260996e-05,
+      "loss": 1.0297,
+      "step": 7755
+    },
+    {
+      "epoch": 0.3140428976123027,
+      "grad_norm": 1.1684590578079224,
+      "learning_rate": 6.999793516415445e-05,
+      "loss": 1.0384,
+      "step": 7760
+    },
+    {
+      "epoch": 0.3142452448401457,
+      "grad_norm": 1.1572351455688477,
+      "learning_rate": 6.997728680569895e-05,
+      "loss": 0.9851,
+      "step": 7765
+    },
+    {
+      "epoch": 0.31444759206798867,
+      "grad_norm": 1.0718976259231567,
+      "learning_rate": 6.995663844724346e-05,
+      "loss": 0.9977,
+      "step": 7770
+    },
+    {
+      "epoch": 0.3146499392958316,
+      "grad_norm": 1.1924231052398682,
+      "learning_rate": 6.993599008878793e-05,
+      "loss": 1.0081,
+      "step": 7775
+    },
+    {
+      "epoch": 0.31485228652367464,
+      "grad_norm": 1.25642991065979,
+      "learning_rate": 6.991534173033244e-05,
+      "loss": 0.9735,
+      "step": 7780
+    },
+    {
+      "epoch": 0.3150546337515176,
+      "grad_norm": 1.0597145557403564,
+      "learning_rate": 6.989469337187694e-05,
+      "loss": 0.9961,
+      "step": 7785
+    },
+    {
+      "epoch": 0.31525698097936056,
+      "grad_norm": 1.1548058986663818,
+      "learning_rate": 6.987404501342145e-05,
+      "loss": 1.0639,
+      "step": 7790
+    },
+    {
+      "epoch": 0.31545932820720357,
+      "grad_norm": 1.20090913772583,
+      "learning_rate": 6.985339665496594e-05,
+      "loss": 1.0209,
+      "step": 7795
+    },
+    {
+      "epoch": 0.31566167543504653,
+      "grad_norm": 1.1690526008605957,
+      "learning_rate": 6.983274829651043e-05,
+      "loss": 1.0425,
+      "step": 7800
+    },
+    {
+      "epoch": 0.31586402266288954,
+      "grad_norm": 1.0979046821594238,
+      "learning_rate": 6.981209993805493e-05,
+      "loss": 1.0157,
+      "step": 7805
+    },
+    {
+      "epoch": 0.3160663698907325,
+      "grad_norm": 1.1317203044891357,
+      "learning_rate": 6.979145157959942e-05,
+      "loss": 1.0308,
+      "step": 7810
+    },
+    {
+      "epoch": 0.31626871711857546,
+      "grad_norm": 1.0178236961364746,
+      "learning_rate": 6.977080322114393e-05,
+      "loss": 0.9917,
+      "step": 7815
+    },
+    {
+      "epoch": 0.3164710643464185,
+      "grad_norm": 1.1497842073440552,
+      "learning_rate": 6.975015486268842e-05,
+      "loss": 0.9876,
+      "step": 7820
+    },
+    {
+      "epoch": 0.31667341157426143,
+      "grad_norm": 1.106931209564209,
+      "learning_rate": 6.972950650423292e-05,
+      "loss": 1.0162,
+      "step": 7825
+    },
+    {
+      "epoch": 0.3168757588021044,
+      "grad_norm": 1.0901240110397339,
+      "learning_rate": 6.970885814577741e-05,
+      "loss": 1.0236,
+      "step": 7830
+    },
+    {
+      "epoch": 0.3170781060299474,
+      "grad_norm": 1.0298885107040405,
+      "learning_rate": 6.968820978732191e-05,
+      "loss": 1.0623,
+      "step": 7835
+    },
+    {
+      "epoch": 0.31728045325779036,
+      "grad_norm": 1.3310502767562866,
+      "learning_rate": 6.966756142886642e-05,
+      "loss": 1.0257,
+      "step": 7840
+    },
+    {
+      "epoch": 0.3174828004856333,
+      "grad_norm": 1.2882601022720337,
+      "learning_rate": 6.96469130704109e-05,
+      "loss": 1.0523,
+      "step": 7845
+    },
+    {
+      "epoch": 0.31768514771347633,
+      "grad_norm": 1.3289625644683838,
+      "learning_rate": 6.96262647119554e-05,
+      "loss": 1.0001,
+      "step": 7850
+    },
+    {
+      "epoch": 0.3178874949413193,
+      "grad_norm": 1.1948840618133545,
+      "learning_rate": 6.96056163534999e-05,
+      "loss": 1.0072,
+      "step": 7855
+    },
+    {
+      "epoch": 0.3180898421691623,
+      "grad_norm": 1.1636812686920166,
+      "learning_rate": 6.95849679950444e-05,
+      "loss": 0.9555,
+      "step": 7860
+    },
+    {
+      "epoch": 0.31829218939700527,
+      "grad_norm": 1.1317676305770874,
+      "learning_rate": 6.95643196365889e-05,
+      "loss": 0.9662,
+      "step": 7865
+    },
+    {
+      "epoch": 0.3184945366248482,
+      "grad_norm": 1.180823564529419,
+      "learning_rate": 6.954367127813339e-05,
+      "loss": 0.984,
+      "step": 7870
+    },
+    {
+      "epoch": 0.31869688385269124,
+      "grad_norm": 1.0699626207351685,
+      "learning_rate": 6.952302291967789e-05,
+      "loss": 0.9885,
+      "step": 7875
+    },
+    {
+      "epoch": 0.3188992310805342,
+      "grad_norm": 1.0747015476226807,
+      "learning_rate": 6.950237456122238e-05,
+      "loss": 1.0191,
+      "step": 7880
+    },
+    {
+      "epoch": 0.31910157830837715,
+      "grad_norm": 1.1225368976593018,
+      "learning_rate": 6.948172620276689e-05,
+      "loss": 1.0072,
+      "step": 7885
+    },
+    {
+      "epoch": 0.31930392553622017,
+      "grad_norm": 1.081352710723877,
+      "learning_rate": 6.946107784431138e-05,
+      "loss": 1.0131,
+      "step": 7890
+    },
+    {
+      "epoch": 0.3195062727640631,
+      "grad_norm": 1.1860682964324951,
+      "learning_rate": 6.944042948585587e-05,
+      "loss": 1.0169,
+      "step": 7895
+    },
+    {
+      "epoch": 0.3197086199919061,
+      "grad_norm": 1.1549993753433228,
+      "learning_rate": 6.941978112740037e-05,
+      "loss": 1.0003,
+      "step": 7900
+    },
+    {
+      "epoch": 0.3199109672197491,
+      "grad_norm": 1.1799170970916748,
+      "learning_rate": 6.939913276894488e-05,
+      "loss": 0.9382,
+      "step": 7905
+    },
+    {
+      "epoch": 0.32011331444759206,
+      "grad_norm": 1.1656957864761353,
+      "learning_rate": 6.937848441048937e-05,
+      "loss": 1.0021,
+      "step": 7910
+    },
+    {
+      "epoch": 0.32031566167543507,
+      "grad_norm": 1.3107892274856567,
+      "learning_rate": 6.935783605203386e-05,
+      "loss": 0.9928,
+      "step": 7915
+    },
+    {
+      "epoch": 0.32051800890327803,
+      "grad_norm": 1.1690044403076172,
+      "learning_rate": 6.933718769357836e-05,
+      "loss": 0.9826,
+      "step": 7920
+    },
+    {
+      "epoch": 0.320720356131121,
+      "grad_norm": 1.1157152652740479,
+      "learning_rate": 6.931653933512287e-05,
+      "loss": 1.0577,
+      "step": 7925
+    },
+    {
+      "epoch": 0.320922703358964,
+      "grad_norm": 1.170356273651123,
+      "learning_rate": 6.929589097666736e-05,
+      "loss": 0.9903,
+      "step": 7930
+    },
+    {
+      "epoch": 0.32112505058680696,
+      "grad_norm": 1.1648330688476562,
+      "learning_rate": 6.927524261821186e-05,
+      "loss": 1.0133,
+      "step": 7935
+    },
+    {
+      "epoch": 0.3213273978146499,
+      "grad_norm": 1.267077922821045,
+      "learning_rate": 6.925459425975635e-05,
+      "loss": 1.0096,
+      "step": 7940
+    },
+    {
+      "epoch": 0.32152974504249293,
+      "grad_norm": 1.2341781854629517,
+      "learning_rate": 6.923394590130084e-05,
+      "loss": 0.9657,
+      "step": 7945
+    },
+    {
+      "epoch": 0.3217320922703359,
+      "grad_norm": 1.1833523511886597,
+      "learning_rate": 6.921329754284535e-05,
+      "loss": 1.011,
+      "step": 7950
+    },
+    {
+      "epoch": 0.32193443949817885,
+      "grad_norm": 1.206613540649414,
+      "learning_rate": 6.919264918438985e-05,
+      "loss": 1.0141,
+      "step": 7955
+    },
+    {
+      "epoch": 0.32213678672602186,
+      "grad_norm": 1.2183783054351807,
+      "learning_rate": 6.917200082593434e-05,
+      "loss": 0.9733,
+      "step": 7960
+    },
+    {
+      "epoch": 0.3223391339538648,
+      "grad_norm": 1.0917104482650757,
+      "learning_rate": 6.915135246747883e-05,
+      "loss": 1.0339,
+      "step": 7965
+    },
+    {
+      "epoch": 0.32254148118170783,
+      "grad_norm": 1.1361770629882812,
+      "learning_rate": 6.913070410902333e-05,
+      "loss": 0.9885,
+      "step": 7970
+    },
+    {
+      "epoch": 0.3227438284095508,
+      "grad_norm": 1.0975098609924316,
+      "learning_rate": 6.911005575056784e-05,
+      "loss": 1.0272,
+      "step": 7975
+    },
+    {
+      "epoch": 0.32294617563739375,
+      "grad_norm": 1.1798059940338135,
+      "learning_rate": 6.908940739211233e-05,
+      "loss": 1.0756,
+      "step": 7980
+    },
+    {
+      "epoch": 0.32314852286523676,
+      "grad_norm": 1.1872261762619019,
+      "learning_rate": 6.906875903365683e-05,
+      "loss": 0.962,
+      "step": 7985
+    },
+    {
+      "epoch": 0.3233508700930797,
+      "grad_norm": 1.101155400276184,
+      "learning_rate": 6.904811067520132e-05,
+      "loss": 1.0203,
+      "step": 7990
+    },
+    {
+      "epoch": 0.3235532173209227,
+      "grad_norm": 1.2731853723526,
+      "learning_rate": 6.902746231674581e-05,
+      "loss": 0.9985,
+      "step": 7995
+    },
+    {
+      "epoch": 0.3237555645487657,
+      "grad_norm": 1.1103601455688477,
+      "learning_rate": 6.900681395829032e-05,
+      "loss": 1.0107,
+      "step": 8000
+    },
+    {
+      "epoch": 0.32395791177660865,
+      "grad_norm": 1.0946158170700073,
+      "learning_rate": 6.898616559983482e-05,
+      "loss": 1.0405,
+      "step": 8005
+    },
+    {
+      "epoch": 0.3241602590044516,
+      "grad_norm": 1.145131230354309,
+      "learning_rate": 6.896551724137931e-05,
+      "loss": 1.0078,
+      "step": 8010
+    },
+    {
+      "epoch": 0.3243626062322946,
+      "grad_norm": 1.2215243577957153,
+      "learning_rate": 6.89448688829238e-05,
+      "loss": 1.0298,
+      "step": 8015
+    },
+    {
+      "epoch": 0.3245649534601376,
+      "grad_norm": 1.146912693977356,
+      "learning_rate": 6.892422052446831e-05,
+      "loss": 1.0104,
+      "step": 8020
+    },
+    {
+      "epoch": 0.3247673006879806,
+      "grad_norm": 1.0603580474853516,
+      "learning_rate": 6.890357216601281e-05,
+      "loss": 1.0011,
+      "step": 8025
+    },
+    {
+      "epoch": 0.32496964791582356,
+      "grad_norm": 1.2084134817123413,
+      "learning_rate": 6.88829238075573e-05,
+      "loss": 1.031,
+      "step": 8030
+    },
+    {
+      "epoch": 0.3251719951436665,
+      "grad_norm": 1.2100253105163574,
+      "learning_rate": 6.886227544910179e-05,
+      "loss": 1.017,
+      "step": 8035
+    },
+    {
+      "epoch": 0.32537434237150953,
+      "grad_norm": 1.1398605108261108,
+      "learning_rate": 6.88416270906463e-05,
+      "loss": 1.0072,
+      "step": 8040
+    },
+    {
+      "epoch": 0.3255766895993525,
+      "grad_norm": 1.1360002756118774,
+      "learning_rate": 6.88209787321908e-05,
+      "loss": 1.0873,
+      "step": 8045
+    },
+    {
+      "epoch": 0.32577903682719545,
+      "grad_norm": 1.2177627086639404,
+      "learning_rate": 6.880033037373529e-05,
+      "loss": 1.0156,
+      "step": 8050
+    },
+    {
+      "epoch": 0.32598138405503846,
+      "grad_norm": 1.3088794946670532,
+      "learning_rate": 6.87796820152798e-05,
+      "loss": 0.9743,
+      "step": 8055
+    },
+    {
+      "epoch": 0.3261837312828814,
+      "grad_norm": 1.2426222562789917,
+      "learning_rate": 6.875903365682429e-05,
+      "loss": 0.9841,
+      "step": 8060
+    },
+    {
+      "epoch": 0.3263860785107244,
+      "grad_norm": 1.1549899578094482,
+      "learning_rate": 6.873838529836878e-05,
+      "loss": 1.0433,
+      "step": 8065
+    },
+    {
+      "epoch": 0.3265884257385674,
+      "grad_norm": 1.0796194076538086,
+      "learning_rate": 6.871773693991328e-05,
+      "loss": 1.0458,
+      "step": 8070
+    },
+    {
+      "epoch": 0.32679077296641035,
+      "grad_norm": 1.2137360572814941,
+      "learning_rate": 6.869708858145778e-05,
+      "loss": 0.9939,
+      "step": 8075
+    },
+    {
+      "epoch": 0.32699312019425336,
+      "grad_norm": 1.224730134010315,
+      "learning_rate": 6.867644022300228e-05,
+      "loss": 1.0541,
+      "step": 8080
+    },
+    {
+      "epoch": 0.3271954674220963,
+      "grad_norm": 1.3345255851745605,
+      "learning_rate": 6.865579186454677e-05,
+      "loss": 1.0071,
+      "step": 8085
+    },
+    {
+      "epoch": 0.3273978146499393,
+      "grad_norm": 1.2496358156204224,
+      "learning_rate": 6.863514350609127e-05,
+      "loss": 0.9642,
+      "step": 8090
+    },
+    {
+      "epoch": 0.3276001618777823,
+      "grad_norm": 1.1511274576187134,
+      "learning_rate": 6.861449514763577e-05,
+      "loss": 1.0755,
+      "step": 8095
+    },
+    {
+      "epoch": 0.32780250910562525,
+      "grad_norm": 1.2001534700393677,
+      "learning_rate": 6.859384678918026e-05,
+      "loss": 1.0396,
+      "step": 8100
+    },
+    {
+      "epoch": 0.3280048563334682,
+      "grad_norm": 1.0557612180709839,
+      "learning_rate": 6.857319843072476e-05,
+      "loss": 0.9412,
+      "step": 8105
+    },
+    {
+      "epoch": 0.3282072035613112,
+      "grad_norm": 1.2284162044525146,
+      "learning_rate": 6.855255007226926e-05,
+      "loss": 1.0042,
+      "step": 8110
+    },
+    {
+      "epoch": 0.3284095507891542,
+      "grad_norm": 1.176843523979187,
+      "learning_rate": 6.853190171381375e-05,
+      "loss": 1.0567,
+      "step": 8115
+    },
+    {
+      "epoch": 0.3286118980169972,
+      "grad_norm": 1.2524343729019165,
+      "learning_rate": 6.851125335535825e-05,
+      "loss": 1.0043,
+      "step": 8120
+    },
+    {
+      "epoch": 0.32881424524484015,
+      "grad_norm": 1.2177811861038208,
+      "learning_rate": 6.849060499690276e-05,
+      "loss": 0.9998,
+      "step": 8125
+    },
+    {
+      "epoch": 0.3290165924726831,
+      "grad_norm": 1.309518814086914,
+      "learning_rate": 6.846995663844723e-05,
+      "loss": 1.0058,
+      "step": 8130
+    },
+    {
+      "epoch": 0.3292189397005261,
+      "grad_norm": 1.1757453680038452,
+      "learning_rate": 6.844930827999174e-05,
+      "loss": 1.039,
+      "step": 8135
+    },
+    {
+      "epoch": 0.3294212869283691,
+      "grad_norm": 1.3004614114761353,
+      "learning_rate": 6.842865992153624e-05,
+      "loss": 0.9395,
+      "step": 8140
+    },
+    {
+      "epoch": 0.32962363415621204,
+      "grad_norm": 1.1624009609222412,
+      "learning_rate": 6.840801156308075e-05,
+      "loss": 1.0253,
+      "step": 8145
+    },
+    {
+      "epoch": 0.32982598138405506,
+      "grad_norm": 1.1674039363861084,
+      "learning_rate": 6.838736320462524e-05,
+      "loss": 0.9377,
+      "step": 8150
+    },
+    {
+      "epoch": 0.330028328611898,
+      "grad_norm": 1.3319944143295288,
+      "learning_rate": 6.836671484616973e-05,
+      "loss": 1.0289,
+      "step": 8155
+    },
+    {
+      "epoch": 0.330230675839741,
+      "grad_norm": 1.2189615964889526,
+      "learning_rate": 6.834606648771423e-05,
+      "loss": 1.0323,
+      "step": 8160
+    },
+    {
+      "epoch": 0.330433023067584,
+      "grad_norm": 1.179700255393982,
+      "learning_rate": 6.832541812925872e-05,
+      "loss": 0.98,
+      "step": 8165
+    },
+    {
+      "epoch": 0.33063537029542694,
+      "grad_norm": 1.1240464448928833,
+      "learning_rate": 6.830476977080323e-05,
+      "loss": 0.9815,
+      "step": 8170
+    },
+    {
+      "epoch": 0.33083771752326996,
+      "grad_norm": 1.1898940801620483,
+      "learning_rate": 6.828412141234772e-05,
+      "loss": 1.005,
+      "step": 8175
+    },
+    {
+      "epoch": 0.3310400647511129,
+      "grad_norm": 1.1352357864379883,
+      "learning_rate": 6.826347305389222e-05,
+      "loss": 1.0635,
+      "step": 8180
+    },
+    {
+      "epoch": 0.3312424119789559,
+      "grad_norm": 1.1874902248382568,
+      "learning_rate": 6.824282469543671e-05,
+      "loss": 1.0506,
+      "step": 8185
+    },
+    {
+      "epoch": 0.3314447592067989,
+      "grad_norm": 1.250718116760254,
+      "learning_rate": 6.822217633698122e-05,
+      "loss": 1.0776,
+      "step": 8190
+    },
+    {
+      "epoch": 0.33164710643464185,
+      "grad_norm": 1.1884347200393677,
+      "learning_rate": 6.820152797852572e-05,
+      "loss": 1.0312,
+      "step": 8195
+    },
+    {
+      "epoch": 0.3318494536624848,
+      "grad_norm": 1.047613501548767,
+      "learning_rate": 6.818087962007021e-05,
+      "loss": 0.98,
+      "step": 8200
+    },
+    {
+      "epoch": 0.3320518008903278,
+      "grad_norm": 1.2003908157348633,
+      "learning_rate": 6.81602312616147e-05,
+      "loss": 1.0186,
+      "step": 8205
+    },
+    {
+      "epoch": 0.3322541481181708,
+      "grad_norm": 1.148883581161499,
+      "learning_rate": 6.81395829031592e-05,
+      "loss": 0.9383,
+      "step": 8210
+    },
+    {
+      "epoch": 0.33245649534601374,
+      "grad_norm": 1.173259973526001,
+      "learning_rate": 6.81189345447037e-05,
+      "loss": 1.0041,
+      "step": 8215
+    },
+    {
+      "epoch": 0.33265884257385675,
+      "grad_norm": 1.140166997909546,
+      "learning_rate": 6.80982861862482e-05,
+      "loss": 1.0139,
+      "step": 8220
+    },
+    {
+      "epoch": 0.3328611898016997,
+      "grad_norm": 1.3193848133087158,
+      "learning_rate": 6.807763782779269e-05,
+      "loss": 1.0187,
+      "step": 8225
+    },
+    {
+      "epoch": 0.3330635370295427,
+      "grad_norm": 1.1724793910980225,
+      "learning_rate": 6.80569894693372e-05,
+      "loss": 1.0155,
+      "step": 8230
+    },
+    {
+      "epoch": 0.3332658842573857,
+      "grad_norm": 1.2499336004257202,
+      "learning_rate": 6.803634111088168e-05,
+      "loss": 1.0641,
+      "step": 8235
+    },
+    {
+      "epoch": 0.33346823148522864,
+      "grad_norm": 1.2024049758911133,
+      "learning_rate": 6.801569275242619e-05,
+      "loss": 1.0274,
+      "step": 8240
+    },
+    {
+      "epoch": 0.33367057871307165,
+      "grad_norm": 1.1916636228561401,
+      "learning_rate": 6.799504439397068e-05,
+      "loss": 0.9823,
+      "step": 8245
+    },
+    {
+      "epoch": 0.3338729259409146,
+      "grad_norm": 1.1534245014190674,
+      "learning_rate": 6.797439603551517e-05,
+      "loss": 0.9894,
+      "step": 8250
+    },
+    {
+      "epoch": 0.33407527316875757,
+      "grad_norm": 1.3492305278778076,
+      "learning_rate": 6.795374767705967e-05,
+      "loss": 1.0102,
+      "step": 8255
+    },
+    {
+      "epoch": 0.3342776203966006,
+      "grad_norm": 1.2801117897033691,
+      "learning_rate": 6.793309931860418e-05,
+      "loss": 0.9809,
+      "step": 8260
+    },
+    {
+      "epoch": 0.33447996762444354,
+      "grad_norm": 1.2604655027389526,
+      "learning_rate": 6.791245096014868e-05,
+      "loss": 1.0098,
+      "step": 8265
+    },
+    {
+      "epoch": 0.3346823148522865,
+      "grad_norm": 1.1137096881866455,
+      "learning_rate": 6.789180260169317e-05,
+      "loss": 0.9381,
+      "step": 8270
+    },
+    {
+      "epoch": 0.3348846620801295,
+      "grad_norm": 1.128345012664795,
+      "learning_rate": 6.787115424323766e-05,
+      "loss": 0.9622,
+      "step": 8275
+    },
+    {
+      "epoch": 0.33508700930797247,
+      "grad_norm": 1.153509259223938,
+      "learning_rate": 6.785050588478217e-05,
+      "loss": 0.9851,
+      "step": 8280
+    },
+    {
+      "epoch": 0.3352893565358155,
+      "grad_norm": 1.18508780002594,
+      "learning_rate": 6.782985752632666e-05,
+      "loss": 0.9857,
+      "step": 8285
+    },
+    {
+      "epoch": 0.33549170376365844,
+      "grad_norm": 1.2997406721115112,
+      "learning_rate": 6.780920916787116e-05,
+      "loss": 1.0219,
+      "step": 8290
+    },
+    {
+      "epoch": 0.3356940509915014,
+      "grad_norm": 1.0437331199645996,
+      "learning_rate": 6.778856080941565e-05,
+      "loss": 0.9976,
+      "step": 8295
+    },
+    {
+      "epoch": 0.3358963982193444,
+      "grad_norm": 1.1383854150772095,
+      "learning_rate": 6.776791245096014e-05,
+      "loss": 0.9957,
+      "step": 8300
+    },
+    {
+      "epoch": 0.3360987454471874,
+      "grad_norm": 1.0369354486465454,
+      "learning_rate": 6.774726409250465e-05,
+      "loss": 0.9836,
+      "step": 8305
+    },
+    {
+      "epoch": 0.33630109267503033,
+      "grad_norm": 0.9924631118774414,
+      "learning_rate": 6.772661573404915e-05,
+      "loss": 1.0246,
+      "step": 8310
+    },
+    {
+      "epoch": 0.33650343990287335,
+      "grad_norm": 1.0379579067230225,
+      "learning_rate": 6.770596737559366e-05,
+      "loss": 0.9578,
+      "step": 8315
+    },
+    {
+      "epoch": 0.3367057871307163,
+      "grad_norm": 1.2099262475967407,
+      "learning_rate": 6.768531901713813e-05,
+      "loss": 1.0898,
+      "step": 8320
+    },
+    {
+      "epoch": 0.33690813435855926,
+      "grad_norm": 1.0912680625915527,
+      "learning_rate": 6.766467065868264e-05,
+      "loss": 0.9977,
+      "step": 8325
+    },
+    {
+      "epoch": 0.3371104815864023,
+      "grad_norm": 1.2483148574829102,
+      "learning_rate": 6.764402230022714e-05,
+      "loss": 1.0372,
+      "step": 8330
+    },
+    {
+      "epoch": 0.33731282881424524,
+      "grad_norm": 1.259277105331421,
+      "learning_rate": 6.762337394177163e-05,
+      "loss": 0.998,
+      "step": 8335
+    },
+    {
+      "epoch": 0.33751517604208825,
+      "grad_norm": 1.105028510093689,
+      "learning_rate": 6.760272558331613e-05,
+      "loss": 1.0244,
+      "step": 8340
+    },
+    {
+      "epoch": 0.3377175232699312,
+      "grad_norm": 1.0760465860366821,
+      "learning_rate": 6.758207722486063e-05,
+      "loss": 1.0465,
+      "step": 8345
+    },
+    {
+      "epoch": 0.33791987049777417,
+      "grad_norm": 1.388336420059204,
+      "learning_rate": 6.756142886640512e-05,
+      "loss": 0.9558,
+      "step": 8350
+    },
+    {
+      "epoch": 0.3381222177256172,
+      "grad_norm": 1.2203596830368042,
+      "learning_rate": 6.754078050794962e-05,
+      "loss": 0.9896,
+      "step": 8355
+    },
+    {
+      "epoch": 0.33832456495346014,
+      "grad_norm": 1.1341876983642578,
+      "learning_rate": 6.752013214949412e-05,
+      "loss": 0.9645,
+      "step": 8360
+    },
+    {
+      "epoch": 0.3385269121813031,
+      "grad_norm": 1.2270278930664062,
+      "learning_rate": 6.749948379103861e-05,
+      "loss": 1.034,
+      "step": 8365
+    },
+    {
+      "epoch": 0.3387292594091461,
+      "grad_norm": 1.097016453742981,
+      "learning_rate": 6.74788354325831e-05,
+      "loss": 0.9782,
+      "step": 8370
+    },
+    {
+      "epoch": 0.33893160663698907,
+      "grad_norm": 1.2199102640151978,
+      "learning_rate": 6.745818707412761e-05,
+      "loss": 0.942,
+      "step": 8375
+    },
+    {
+      "epoch": 0.339133953864832,
+      "grad_norm": 1.2162120342254639,
+      "learning_rate": 6.743753871567211e-05,
+      "loss": 0.9886,
+      "step": 8380
+    },
+    {
+      "epoch": 0.33933630109267504,
+      "grad_norm": 1.1713508367538452,
+      "learning_rate": 6.74168903572166e-05,
+      "loss": 0.9767,
+      "step": 8385
+    },
+    {
+      "epoch": 0.339538648320518,
+      "grad_norm": 1.2054520845413208,
+      "learning_rate": 6.73962419987611e-05,
+      "loss": 0.965,
+      "step": 8390
+    },
+    {
+      "epoch": 0.339740995548361,
+      "grad_norm": 1.1206656694412231,
+      "learning_rate": 6.73755936403056e-05,
+      "loss": 0.98,
+      "step": 8395
+    },
+    {
+      "epoch": 0.33994334277620397,
+      "grad_norm": 1.171546459197998,
+      "learning_rate": 6.73549452818501e-05,
+      "loss": 1.0471,
+      "step": 8400
+    },
+    {
+      "epoch": 0.34014569000404693,
+      "grad_norm": 1.2924500703811646,
+      "learning_rate": 6.733429692339459e-05,
+      "loss": 0.9631,
+      "step": 8405
+    },
+    {
+      "epoch": 0.34034803723188994,
+      "grad_norm": 1.4253216981887817,
+      "learning_rate": 6.73136485649391e-05,
+      "loss": 0.9822,
+      "step": 8410
+    },
+    {
+      "epoch": 0.3405503844597329,
+      "grad_norm": 1.2813270092010498,
+      "learning_rate": 6.729300020648359e-05,
+      "loss": 1.0079,
+      "step": 8415
+    },
+    {
+      "epoch": 0.34075273168757586,
+      "grad_norm": 1.2764500379562378,
+      "learning_rate": 6.727235184802808e-05,
+      "loss": 1.0137,
+      "step": 8420
+    },
+    {
+      "epoch": 0.3409550789154189,
+      "grad_norm": 1.2863959074020386,
+      "learning_rate": 6.725170348957258e-05,
+      "loss": 1.0109,
+      "step": 8425
+    },
+    {
+      "epoch": 0.34115742614326183,
+      "grad_norm": 1.150322675704956,
+      "learning_rate": 6.723105513111709e-05,
+      "loss": 0.9935,
+      "step": 8430
+    },
+    {
+      "epoch": 0.3413597733711048,
+      "grad_norm": 1.238601803779602,
+      "learning_rate": 6.721040677266158e-05,
+      "loss": 0.9995,
+      "step": 8435
+    },
+    {
+      "epoch": 0.3415621205989478,
+      "grad_norm": 1.0317527055740356,
+      "learning_rate": 6.718975841420607e-05,
+      "loss": 1.0332,
+      "step": 8440
+    },
+    {
+      "epoch": 0.34176446782679076,
+      "grad_norm": 1.1871312856674194,
+      "learning_rate": 6.716911005575057e-05,
+      "loss": 0.9883,
+      "step": 8445
+    },
+    {
+      "epoch": 0.3419668150546338,
+      "grad_norm": 1.317663550376892,
+      "learning_rate": 6.714846169729508e-05,
+      "loss": 1.0207,
+      "step": 8450
+    },
+    {
+      "epoch": 0.34216916228247674,
+      "grad_norm": 1.2404143810272217,
+      "learning_rate": 6.712781333883957e-05,
+      "loss": 0.977,
+      "step": 8455
+    },
+    {
+      "epoch": 0.3423715095103197,
+      "grad_norm": 1.241338849067688,
+      "learning_rate": 6.710716498038406e-05,
+      "loss": 0.9577,
+      "step": 8460
+    },
+    {
+      "epoch": 0.3425738567381627,
+      "grad_norm": 1.1391888856887817,
+      "learning_rate": 6.708651662192856e-05,
+      "loss": 0.9494,
+      "step": 8465
+    },
+    {
+      "epoch": 0.34277620396600567,
+      "grad_norm": 1.0304710865020752,
+      "learning_rate": 6.706586826347305e-05,
+      "loss": 1.0009,
+      "step": 8470
+    },
+    {
+      "epoch": 0.3429785511938486,
+      "grad_norm": 1.1427103281021118,
+      "learning_rate": 6.704521990501756e-05,
+      "loss": 0.9629,
+      "step": 8475
+    },
+    {
+      "epoch": 0.34318089842169164,
+      "grad_norm": 1.000894546508789,
+      "learning_rate": 6.702457154656206e-05,
+      "loss": 0.9834,
+      "step": 8480
+    },
+    {
+      "epoch": 0.3433832456495346,
+      "grad_norm": 1.1159125566482544,
+      "learning_rate": 6.700392318810655e-05,
+      "loss": 1.0035,
+      "step": 8485
+    },
+    {
+      "epoch": 0.34358559287737755,
+      "grad_norm": 1.2044910192489624,
+      "learning_rate": 6.698327482965104e-05,
+      "loss": 1.0604,
+      "step": 8490
+    },
+    {
+      "epoch": 0.34378794010522057,
+      "grad_norm": 1.2182488441467285,
+      "learning_rate": 6.696262647119554e-05,
+      "loss": 1.0079,
+      "step": 8495
+    },
+    {
+      "epoch": 0.3439902873330635,
+      "grad_norm": 1.2006793022155762,
+      "learning_rate": 6.694197811274005e-05,
+      "loss": 1.0278,
+      "step": 8500
+    },
+    {
+      "epoch": 0.34419263456090654,
+      "grad_norm": 1.1410280466079712,
+      "learning_rate": 6.692132975428454e-05,
+      "loss": 1.0009,
+      "step": 8505
+    },
+    {
+      "epoch": 0.3443949817887495,
+      "grad_norm": 1.0319397449493408,
+      "learning_rate": 6.690068139582903e-05,
+      "loss": 1.0348,
+      "step": 8510
+    },
+    {
+      "epoch": 0.34459732901659246,
+      "grad_norm": 1.0982345342636108,
+      "learning_rate": 6.688003303737353e-05,
+      "loss": 1.0038,
+      "step": 8515
+    },
+    {
+      "epoch": 0.34479967624443547,
+      "grad_norm": 1.2336574792861938,
+      "learning_rate": 6.685938467891802e-05,
+      "loss": 1.0311,
+      "step": 8520
+    },
+    {
+      "epoch": 0.34500202347227843,
+      "grad_norm": 1.215314269065857,
+      "learning_rate": 6.683873632046253e-05,
+      "loss": 1.0142,
+      "step": 8525
+    },
+    {
+      "epoch": 0.3452043707001214,
+      "grad_norm": 1.2042006254196167,
+      "learning_rate": 6.681808796200703e-05,
+      "loss": 1.0162,
+      "step": 8530
+    },
+    {
+      "epoch": 0.3454067179279644,
+      "grad_norm": 1.1784260272979736,
+      "learning_rate": 6.679743960355152e-05,
+      "loss": 1.0074,
+      "step": 8535
+    },
+    {
+      "epoch": 0.34560906515580736,
+      "grad_norm": 1.1501988172531128,
+      "learning_rate": 6.677679124509601e-05,
+      "loss": 0.9841,
+      "step": 8540
+    },
+    {
+      "epoch": 0.3458114123836503,
+      "grad_norm": 1.0892448425292969,
+      "learning_rate": 6.675614288664052e-05,
+      "loss": 1.022,
+      "step": 8545
+    },
+    {
+      "epoch": 0.34601375961149333,
+      "grad_norm": 1.1556189060211182,
+      "learning_rate": 6.673549452818502e-05,
+      "loss": 0.9731,
+      "step": 8550
+    },
+    {
+      "epoch": 0.3462161068393363,
+      "grad_norm": 1.184709072113037,
+      "learning_rate": 6.671484616972951e-05,
+      "loss": 0.9789,
+      "step": 8555
+    },
+    {
+      "epoch": 0.3464184540671793,
+      "grad_norm": 1.1903433799743652,
+      "learning_rate": 6.6694197811274e-05,
+      "loss": 0.9733,
+      "step": 8560
+    },
+    {
+      "epoch": 0.34662080129502226,
+      "grad_norm": 1.0671143531799316,
+      "learning_rate": 6.66735494528185e-05,
+      "loss": 0.9794,
+      "step": 8565
+    },
+    {
+      "epoch": 0.3468231485228652,
+      "grad_norm": 1.2823303937911987,
+      "learning_rate": 6.6652901094363e-05,
+      "loss": 1.0007,
+      "step": 8570
+    },
+    {
+      "epoch": 0.34702549575070823,
+      "grad_norm": 1.497361660003662,
+      "learning_rate": 6.66322527359075e-05,
+      "loss": 1.0172,
+      "step": 8575
+    },
+    {
+      "epoch": 0.3472278429785512,
+      "grad_norm": 1.0234427452087402,
+      "learning_rate": 6.661160437745199e-05,
+      "loss": 0.9712,
+      "step": 8580
+    },
+    {
+      "epoch": 0.34743019020639415,
+      "grad_norm": 1.2735185623168945,
+      "learning_rate": 6.65909560189965e-05,
+      "loss": 1.0148,
+      "step": 8585
+    },
+    {
+      "epoch": 0.34763253743423717,
+      "grad_norm": 1.2002781629562378,
+      "learning_rate": 6.657030766054099e-05,
+      "loss": 1.0283,
+      "step": 8590
+    },
+    {
+      "epoch": 0.3478348846620801,
+      "grad_norm": 1.1865485906600952,
+      "learning_rate": 6.654965930208549e-05,
+      "loss": 1.0345,
+      "step": 8595
+    },
+    {
+      "epoch": 0.3480372318899231,
+      "grad_norm": 1.1625314950942993,
+      "learning_rate": 6.652901094363e-05,
+      "loss": 0.9528,
+      "step": 8600
+    },
+    {
+      "epoch": 0.3482395791177661,
+      "grad_norm": 1.134570598602295,
+      "learning_rate": 6.650836258517447e-05,
+      "loss": 0.9629,
+      "step": 8605
+    },
+    {
+      "epoch": 0.34844192634560905,
+      "grad_norm": 1.257711410522461,
+      "learning_rate": 6.648771422671898e-05,
+      "loss": 1.0164,
+      "step": 8610
+    },
+    {
+      "epoch": 0.34864427357345207,
+      "grad_norm": 1.0914216041564941,
+      "learning_rate": 6.646706586826348e-05,
+      "loss": 1.0292,
+      "step": 8615
+    },
+    {
+      "epoch": 0.348846620801295,
+      "grad_norm": 1.0632694959640503,
+      "learning_rate": 6.644641750980798e-05,
+      "loss": 0.9774,
+      "step": 8620
+    },
+    {
+      "epoch": 0.349048968029138,
+      "grad_norm": 1.3002113103866577,
+      "learning_rate": 6.642576915135247e-05,
+      "loss": 1.1101,
+      "step": 8625
+    },
+    {
+      "epoch": 0.349251315256981,
+      "grad_norm": 1.139308214187622,
+      "learning_rate": 6.640512079289696e-05,
+      "loss": 1.041,
+      "step": 8630
+    },
+    {
+      "epoch": 0.34945366248482396,
+      "grad_norm": 1.0839093923568726,
+      "learning_rate": 6.638447243444147e-05,
+      "loss": 0.9781,
+      "step": 8635
+    },
+    {
+      "epoch": 0.3496560097126669,
+      "grad_norm": 1.075213074684143,
+      "learning_rate": 6.636382407598596e-05,
+      "loss": 0.9966,
+      "step": 8640
+    },
+    {
+      "epoch": 0.34985835694050993,
+      "grad_norm": 1.1610937118530273,
+      "learning_rate": 6.634317571753046e-05,
+      "loss": 1.0111,
+      "step": 8645
+    },
+    {
+      "epoch": 0.3500607041683529,
+      "grad_norm": 1.100316047668457,
+      "learning_rate": 6.632252735907495e-05,
+      "loss": 0.9945,
+      "step": 8650
+    },
+    {
+      "epoch": 0.35026305139619585,
+      "grad_norm": 1.2150291204452515,
+      "learning_rate": 6.630187900061944e-05,
+      "loss": 0.9342,
+      "step": 8655
+    },
+    {
+      "epoch": 0.35046539862403886,
+      "grad_norm": 1.0899677276611328,
+      "learning_rate": 6.628123064216395e-05,
+      "loss": 1.0108,
+      "step": 8660
+    },
+    {
+      "epoch": 0.3506677458518818,
+      "grad_norm": 1.3460205793380737,
+      "learning_rate": 6.626058228370845e-05,
+      "loss": 1.0266,
+      "step": 8665
+    },
+    {
+      "epoch": 0.35087009307972483,
+      "grad_norm": 1.128051519393921,
+      "learning_rate": 6.623993392525296e-05,
+      "loss": 0.9024,
+      "step": 8670
+    },
+    {
+      "epoch": 0.3510724403075678,
+      "grad_norm": 1.1437114477157593,
+      "learning_rate": 6.621928556679743e-05,
+      "loss": 1.0248,
+      "step": 8675
+    },
+    {
+      "epoch": 0.35127478753541075,
+      "grad_norm": 1.159614086151123,
+      "learning_rate": 6.619863720834194e-05,
+      "loss": 1.0539,
+      "step": 8680
+    },
+    {
+      "epoch": 0.35147713476325376,
+      "grad_norm": 1.203351378440857,
+      "learning_rate": 6.617798884988644e-05,
+      "loss": 0.9775,
+      "step": 8685
+    },
+    {
+      "epoch": 0.3516794819910967,
+      "grad_norm": 1.1921762228012085,
+      "learning_rate": 6.615734049143093e-05,
+      "loss": 1.017,
+      "step": 8690
+    },
+    {
+      "epoch": 0.3518818292189397,
+      "grad_norm": 1.1406958103179932,
+      "learning_rate": 6.613669213297544e-05,
+      "loss": 0.9743,
+      "step": 8695
+    },
+    {
+      "epoch": 0.3520841764467827,
+      "grad_norm": 1.2871545553207397,
+      "learning_rate": 6.611604377451993e-05,
+      "loss": 1.0171,
+      "step": 8700
+    },
+    {
+      "epoch": 0.35228652367462565,
+      "grad_norm": 1.2367531061172485,
+      "learning_rate": 6.609539541606442e-05,
+      "loss": 1.0321,
+      "step": 8705
+    },
+    {
+      "epoch": 0.3524888709024686,
+      "grad_norm": 1.133935570716858,
+      "learning_rate": 6.607474705760892e-05,
+      "loss": 0.9645,
+      "step": 8710
+    },
+    {
+      "epoch": 0.3526912181303116,
+      "grad_norm": 1.1796834468841553,
+      "learning_rate": 6.605409869915343e-05,
+      "loss": 0.9797,
+      "step": 8715
+    },
+    {
+      "epoch": 0.3528935653581546,
+      "grad_norm": 1.1752643585205078,
+      "learning_rate": 6.603345034069792e-05,
+      "loss": 1.019,
+      "step": 8720
+    },
+    {
+      "epoch": 0.3530959125859976,
+      "grad_norm": 1.2460026741027832,
+      "learning_rate": 6.60128019822424e-05,
+      "loss": 1.0578,
+      "step": 8725
+    },
+    {
+      "epoch": 0.35329825981384055,
+      "grad_norm": 1.1869683265686035,
+      "learning_rate": 6.599215362378691e-05,
+      "loss": 1.0183,
+      "step": 8730
+    },
+    {
+      "epoch": 0.3535006070416835,
+      "grad_norm": 1.1391346454620361,
+      "learning_rate": 6.597150526533141e-05,
+      "loss": 1.0393,
+      "step": 8735
+    },
+    {
+      "epoch": 0.3537029542695265,
+      "grad_norm": 1.1753348112106323,
+      "learning_rate": 6.59508569068759e-05,
+      "loss": 1.0113,
+      "step": 8740
+    },
+    {
+      "epoch": 0.3539053014973695,
+      "grad_norm": 1.1217572689056396,
+      "learning_rate": 6.593020854842041e-05,
+      "loss": 0.9776,
+      "step": 8745
+    },
+    {
+      "epoch": 0.35410764872521244,
+      "grad_norm": 1.3061097860336304,
+      "learning_rate": 6.59095601899649e-05,
+      "loss": 1.0277,
+      "step": 8750
+    },
+    {
+      "epoch": 0.35430999595305546,
+      "grad_norm": 1.143172264099121,
+      "learning_rate": 6.58889118315094e-05,
+      "loss": 0.9993,
+      "step": 8755
+    },
+    {
+      "epoch": 0.3545123431808984,
+      "grad_norm": 1.1326736211776733,
+      "learning_rate": 6.58682634730539e-05,
+      "loss": 1.015,
+      "step": 8760
+    },
+    {
+      "epoch": 0.3547146904087414,
+      "grad_norm": 1.2135322093963623,
+      "learning_rate": 6.58476151145984e-05,
+      "loss": 1.0018,
+      "step": 8765
+    },
+    {
+      "epoch": 0.3549170376365844,
+      "grad_norm": 1.2909351587295532,
+      "learning_rate": 6.582696675614289e-05,
+      "loss": 1.0505,
+      "step": 8770
+    },
+    {
+      "epoch": 0.35511938486442735,
+      "grad_norm": 1.08174729347229,
+      "learning_rate": 6.580631839768738e-05,
+      "loss": 1.0007,
+      "step": 8775
+    },
+    {
+      "epoch": 0.35532173209227036,
+      "grad_norm": 1.168229579925537,
+      "learning_rate": 6.578567003923188e-05,
+      "loss": 0.9724,
+      "step": 8780
+    },
+    {
+      "epoch": 0.3555240793201133,
+      "grad_norm": 1.1908806562423706,
+      "learning_rate": 6.576502168077639e-05,
+      "loss": 1.0154,
+      "step": 8785
+    },
+    {
+      "epoch": 0.3557264265479563,
+      "grad_norm": 1.086943507194519,
+      "learning_rate": 6.574437332232088e-05,
+      "loss": 1.0814,
+      "step": 8790
+    },
+    {
+      "epoch": 0.3559287737757993,
+      "grad_norm": 1.2127574682235718,
+      "learning_rate": 6.572372496386537e-05,
+      "loss": 1.0067,
+      "step": 8795
+    },
+    {
+      "epoch": 0.35613112100364225,
+      "grad_norm": 1.2844805717468262,
+      "learning_rate": 6.570307660540987e-05,
+      "loss": 1.0409,
+      "step": 8800
+    },
+    {
+      "epoch": 0.3563334682314852,
+      "grad_norm": 1.143509030342102,
+      "learning_rate": 6.568242824695438e-05,
+      "loss": 1.0219,
+      "step": 8805
+    },
+    {
+      "epoch": 0.3565358154593282,
+      "grad_norm": 1.1596078872680664,
+      "learning_rate": 6.566177988849887e-05,
+      "loss": 1.0286,
+      "step": 8810
+    },
+    {
+      "epoch": 0.3567381626871712,
+      "grad_norm": 1.260378122329712,
+      "learning_rate": 6.564113153004337e-05,
+      "loss": 0.9993,
+      "step": 8815
+    },
+    {
+      "epoch": 0.35694050991501414,
+      "grad_norm": 1.2232650518417358,
+      "learning_rate": 6.562048317158786e-05,
+      "loss": 0.917,
+      "step": 8820
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 1.0145820379257202,
+      "learning_rate": 6.559983481313235e-05,
+      "loss": 0.9615,
+      "step": 8825
+    },
+    {
+      "epoch": 0.3573452043707001,
+      "grad_norm": 1.060854434967041,
+      "learning_rate": 6.557918645467686e-05,
+      "loss": 0.9444,
+      "step": 8830
+    },
+    {
+      "epoch": 0.3575475515985431,
+      "grad_norm": 1.2130520343780518,
+      "learning_rate": 6.555853809622136e-05,
+      "loss": 1.0045,
+      "step": 8835
+    },
+    {
+      "epoch": 0.3577498988263861,
+      "grad_norm": 1.0780822038650513,
+      "learning_rate": 6.553788973776585e-05,
+      "loss": 1.0154,
+      "step": 8840
+    },
+    {
+      "epoch": 0.35795224605422904,
+      "grad_norm": 1.2764308452606201,
+      "learning_rate": 6.551724137931034e-05,
+      "loss": 0.9512,
+      "step": 8845
+    },
+    {
+      "epoch": 0.35815459328207205,
+      "grad_norm": 1.157224416732788,
+      "learning_rate": 6.549659302085485e-05,
+      "loss": 0.9735,
+      "step": 8850
+    },
+    {
+      "epoch": 0.358356940509915,
+      "grad_norm": 1.2173582315444946,
+      "learning_rate": 6.547594466239935e-05,
+      "loss": 0.9817,
+      "step": 8855
+    },
+    {
+      "epoch": 0.35855928773775797,
+      "grad_norm": 1.1990514993667603,
+      "learning_rate": 6.545529630394384e-05,
+      "loss": 1.0039,
+      "step": 8860
+    },
+    {
+      "epoch": 0.358761634965601,
+      "grad_norm": 1.1545562744140625,
+      "learning_rate": 6.543464794548833e-05,
+      "loss": 0.9734,
+      "step": 8865
+    },
+    {
+      "epoch": 0.35896398219344394,
+      "grad_norm": 1.3034543991088867,
+      "learning_rate": 6.541399958703283e-05,
+      "loss": 1.047,
+      "step": 8870
+    },
+    {
+      "epoch": 0.3591663294212869,
+      "grad_norm": 1.164923071861267,
+      "learning_rate": 6.539335122857733e-05,
+      "loss": 1.0514,
+      "step": 8875
+    },
+    {
+      "epoch": 0.3593686766491299,
+      "grad_norm": 1.2409639358520508,
+      "learning_rate": 6.537270287012183e-05,
+      "loss": 1.0195,
+      "step": 8880
+    },
+    {
+      "epoch": 0.3595710238769729,
+      "grad_norm": 1.2295268774032593,
+      "learning_rate": 6.535205451166633e-05,
+      "loss": 1.0683,
+      "step": 8885
+    },
+    {
+      "epoch": 0.3597733711048159,
+      "grad_norm": 1.1249150037765503,
+      "learning_rate": 6.533140615321082e-05,
+      "loss": 1.0313,
+      "step": 8890
+    },
+    {
+      "epoch": 0.35997571833265885,
+      "grad_norm": 1.0729941129684448,
+      "learning_rate": 6.531075779475531e-05,
+      "loss": 1.0461,
+      "step": 8895
+    },
+    {
+      "epoch": 0.3601780655605018,
+      "grad_norm": 1.1724107265472412,
+      "learning_rate": 6.529010943629982e-05,
+      "loss": 1.002,
+      "step": 8900
+    },
+    {
+      "epoch": 0.3603804127883448,
+      "grad_norm": 1.2224723100662231,
+      "learning_rate": 6.526946107784432e-05,
+      "loss": 1.0158,
+      "step": 8905
+    },
+    {
+      "epoch": 0.3605827600161878,
+      "grad_norm": 1.2359263896942139,
+      "learning_rate": 6.524881271938881e-05,
+      "loss": 0.9996,
+      "step": 8910
+    },
+    {
+      "epoch": 0.36078510724403073,
+      "grad_norm": 1.0514304637908936,
+      "learning_rate": 6.52281643609333e-05,
+      "loss": 1.0003,
+      "step": 8915
+    },
+    {
+      "epoch": 0.36098745447187375,
+      "grad_norm": 1.1479510068893433,
+      "learning_rate": 6.520751600247781e-05,
+      "loss": 1.0162,
+      "step": 8920
+    },
+    {
+      "epoch": 0.3611898016997167,
+      "grad_norm": 1.1739133596420288,
+      "learning_rate": 6.51868676440223e-05,
+      "loss": 0.9948,
+      "step": 8925
+    },
+    {
+      "epoch": 0.3613921489275597,
+      "grad_norm": 1.0608174800872803,
+      "learning_rate": 6.51662192855668e-05,
+      "loss": 0.9774,
+      "step": 8930
+    },
+    {
+      "epoch": 0.3615944961554027,
+      "grad_norm": 1.2887754440307617,
+      "learning_rate": 6.514557092711129e-05,
+      "loss": 0.9972,
+      "step": 8935
+    },
+    {
+      "epoch": 0.36179684338324564,
+      "grad_norm": 1.2014949321746826,
+      "learning_rate": 6.51249225686558e-05,
+      "loss": 0.9897,
+      "step": 8940
+    },
+    {
+      "epoch": 0.36199919061108865,
+      "grad_norm": 1.1276029348373413,
+      "learning_rate": 6.510427421020029e-05,
+      "loss": 1.022,
+      "step": 8945
+    },
+    {
+      "epoch": 0.3622015378389316,
+      "grad_norm": 1.2513147592544556,
+      "learning_rate": 6.508362585174479e-05,
+      "loss": 1.103,
+      "step": 8950
+    },
+    {
+      "epoch": 0.36240388506677457,
+      "grad_norm": 1.149227261543274,
+      "learning_rate": 6.50629774932893e-05,
+      "loss": 0.9761,
+      "step": 8955
+    },
+    {
+      "epoch": 0.3626062322946176,
+      "grad_norm": 1.2752100229263306,
+      "learning_rate": 6.504232913483379e-05,
+      "loss": 0.9947,
+      "step": 8960
+    },
+    {
+      "epoch": 0.36280857952246054,
+      "grad_norm": 0.9958732724189758,
+      "learning_rate": 6.502168077637828e-05,
+      "loss": 0.9444,
+      "step": 8965
+    },
+    {
+      "epoch": 0.3630109267503035,
+      "grad_norm": 1.1153346300125122,
+      "learning_rate": 6.500103241792278e-05,
+      "loss": 0.9872,
+      "step": 8970
+    },
+    {
+      "epoch": 0.3632132739781465,
+      "grad_norm": 1.1249080896377563,
+      "learning_rate": 6.498038405946728e-05,
+      "loss": 1.0455,
+      "step": 8975
+    },
+    {
+      "epoch": 0.36341562120598947,
+      "grad_norm": 1.1198047399520874,
+      "learning_rate": 6.495973570101178e-05,
+      "loss": 1.0316,
+      "step": 8980
+    },
+    {
+      "epoch": 0.3636179684338325,
+      "grad_norm": 1.1231813430786133,
+      "learning_rate": 6.493908734255627e-05,
+      "loss": 1.0009,
+      "step": 8985
+    },
+    {
+      "epoch": 0.36382031566167544,
+      "grad_norm": 1.1274144649505615,
+      "learning_rate": 6.491843898410077e-05,
+      "loss": 0.9737,
+      "step": 8990
+    },
+    {
+      "epoch": 0.3640226628895184,
+      "grad_norm": 1.113556981086731,
+      "learning_rate": 6.489779062564526e-05,
+      "loss": 0.9659,
+      "step": 8995
+    },
+    {
+      "epoch": 0.3642250101173614,
+      "grad_norm": 1.1571226119995117,
+      "learning_rate": 6.487714226718976e-05,
+      "loss": 1.0494,
+      "step": 9000
+    },
+    {
+      "epoch": 0.3644273573452044,
+      "grad_norm": 1.1607192754745483,
+      "learning_rate": 6.485649390873425e-05,
+      "loss": 0.9423,
+      "step": 9005
+    },
+    {
+      "epoch": 0.36462970457304733,
+      "grad_norm": 1.0619757175445557,
+      "learning_rate": 6.483584555027875e-05,
+      "loss": 1.0382,
+      "step": 9010
+    },
+    {
+      "epoch": 0.36483205180089034,
+      "grad_norm": 1.1314587593078613,
+      "learning_rate": 6.481519719182325e-05,
+      "loss": 0.9555,
+      "step": 9015
+    },
+    {
+      "epoch": 0.3650343990287333,
+      "grad_norm": 1.0620434284210205,
+      "learning_rate": 6.479454883336775e-05,
+      "loss": 0.9703,
+      "step": 9020
+    },
+    {
+      "epoch": 0.36523674625657626,
+      "grad_norm": 1.18526291847229,
+      "learning_rate": 6.477390047491226e-05,
+      "loss": 0.9698,
+      "step": 9025
+    },
+    {
+      "epoch": 0.3654390934844193,
+      "grad_norm": 1.2484488487243652,
+      "learning_rate": 6.475325211645675e-05,
+      "loss": 1.0249,
+      "step": 9030
+    },
+    {
+      "epoch": 0.36564144071226223,
+      "grad_norm": 1.2012985944747925,
+      "learning_rate": 6.473260375800124e-05,
+      "loss": 0.9949,
+      "step": 9035
+    },
+    {
+      "epoch": 0.36584378794010525,
+      "grad_norm": 1.0843980312347412,
+      "learning_rate": 6.471195539954574e-05,
+      "loss": 1.0073,
+      "step": 9040
+    },
+    {
+      "epoch": 0.3660461351679482,
+      "grad_norm": 1.2630796432495117,
+      "learning_rate": 6.469130704109023e-05,
+      "loss": 1.0256,
+      "step": 9045
+    },
+    {
+      "epoch": 0.36624848239579116,
+      "grad_norm": 1.2447381019592285,
+      "learning_rate": 6.467065868263474e-05,
+      "loss": 1.0428,
+      "step": 9050
+    },
+    {
+      "epoch": 0.3664508296236342,
+      "grad_norm": 1.0718586444854736,
+      "learning_rate": 6.465001032417923e-05,
+      "loss": 1.0333,
+      "step": 9055
+    },
+    {
+      "epoch": 0.36665317685147714,
+      "grad_norm": 1.1159979104995728,
+      "learning_rate": 6.462936196572372e-05,
+      "loss": 1.0247,
+      "step": 9060
+    },
+    {
+      "epoch": 0.3668555240793201,
+      "grad_norm": 1.1500697135925293,
+      "learning_rate": 6.460871360726822e-05,
+      "loss": 1.0224,
+      "step": 9065
+    },
+    {
+      "epoch": 0.3670578713071631,
+      "grad_norm": 1.1497061252593994,
+      "learning_rate": 6.458806524881273e-05,
+      "loss": 1.0292,
+      "step": 9070
+    },
+    {
+      "epoch": 0.36726021853500607,
+      "grad_norm": 1.248008131980896,
+      "learning_rate": 6.456741689035723e-05,
+      "loss": 1.0007,
+      "step": 9075
+    },
+    {
+      "epoch": 0.367462565762849,
+      "grad_norm": 1.2401740550994873,
+      "learning_rate": 6.454676853190171e-05,
+      "loss": 0.9876,
+      "step": 9080
+    },
+    {
+      "epoch": 0.36766491299069204,
+      "grad_norm": 1.2434844970703125,
+      "learning_rate": 6.452612017344621e-05,
+      "loss": 1.0152,
+      "step": 9085
+    },
+    {
+      "epoch": 0.367867260218535,
+      "grad_norm": 1.1539874076843262,
+      "learning_rate": 6.450547181499072e-05,
+      "loss": 1.0284,
+      "step": 9090
+    },
+    {
+      "epoch": 0.368069607446378,
+      "grad_norm": 1.2069408893585205,
+      "learning_rate": 6.44848234565352e-05,
+      "loss": 0.9982,
+      "step": 9095
+    },
+    {
+      "epoch": 0.36827195467422097,
+      "grad_norm": 1.275341510772705,
+      "learning_rate": 6.446417509807971e-05,
+      "loss": 0.9699,
+      "step": 9100
+    },
+    {
+      "epoch": 0.3684743019020639,
+      "grad_norm": 1.1785728931427002,
+      "learning_rate": 6.44435267396242e-05,
+      "loss": 0.9883,
+      "step": 9105
+    },
+    {
+      "epoch": 0.36867664912990694,
+      "grad_norm": 1.1641168594360352,
+      "learning_rate": 6.44228783811687e-05,
+      "loss": 1.059,
+      "step": 9110
+    },
+    {
+      "epoch": 0.3688789963577499,
+      "grad_norm": 1.095104694366455,
+      "learning_rate": 6.44022300227132e-05,
+      "loss": 1.0262,
+      "step": 9115
+    },
+    {
+      "epoch": 0.36908134358559286,
+      "grad_norm": 1.2540876865386963,
+      "learning_rate": 6.43815816642577e-05,
+      "loss": 0.9935,
+      "step": 9120
+    },
+    {
+      "epoch": 0.36928369081343587,
+      "grad_norm": 1.1737287044525146,
+      "learning_rate": 6.436093330580219e-05,
+      "loss": 1.0222,
+      "step": 9125
+    },
+    {
+      "epoch": 0.36948603804127883,
+      "grad_norm": 1.1528546810150146,
+      "learning_rate": 6.434028494734668e-05,
+      "loss": 0.9998,
+      "step": 9130
+    },
+    {
+      "epoch": 0.3696883852691218,
+      "grad_norm": 1.2154673337936401,
+      "learning_rate": 6.431963658889118e-05,
+      "loss": 1.0409,
+      "step": 9135
+    },
+    {
+      "epoch": 0.3698907324969648,
+      "grad_norm": 1.1713812351226807,
+      "learning_rate": 6.429898823043569e-05,
+      "loss": 1.0076,
+      "step": 9140
+    },
+    {
+      "epoch": 0.37009307972480776,
+      "grad_norm": 1.2802129983901978,
+      "learning_rate": 6.427833987198018e-05,
+      "loss": 0.9767,
+      "step": 9145
+    },
+    {
+      "epoch": 0.3702954269526508,
+      "grad_norm": 1.17072594165802,
+      "learning_rate": 6.425769151352467e-05,
+      "loss": 0.9871,
+      "step": 9150
+    },
+    {
+      "epoch": 0.37049777418049373,
+      "grad_norm": 1.6219251155853271,
+      "learning_rate": 6.423704315506917e-05,
+      "loss": 1.0339,
+      "step": 9155
+    },
+    {
+      "epoch": 0.3707001214083367,
+      "grad_norm": 1.1033552885055542,
+      "learning_rate": 6.421639479661368e-05,
+      "loss": 1.0638,
+      "step": 9160
+    },
+    {
+      "epoch": 0.3709024686361797,
+      "grad_norm": 1.3425971269607544,
+      "learning_rate": 6.419574643815817e-05,
+      "loss": 1.0028,
+      "step": 9165
+    },
+    {
+      "epoch": 0.37110481586402266,
+      "grad_norm": 1.1628460884094238,
+      "learning_rate": 6.417509807970267e-05,
+      "loss": 0.9876,
+      "step": 9170
+    },
+    {
+      "epoch": 0.3713071630918656,
+      "grad_norm": 1.2166439294815063,
+      "learning_rate": 6.415444972124716e-05,
+      "loss": 1.0155,
+      "step": 9175
+    },
+    {
+      "epoch": 0.37150951031970864,
+      "grad_norm": 1.1808024644851685,
+      "learning_rate": 6.413380136279165e-05,
+      "loss": 0.9711,
+      "step": 9180
+    },
+    {
+      "epoch": 0.3717118575475516,
+      "grad_norm": 1.1127512454986572,
+      "learning_rate": 6.411315300433616e-05,
+      "loss": 0.9662,
+      "step": 9185
+    },
+    {
+      "epoch": 0.37191420477539455,
+      "grad_norm": 1.2843769788742065,
+      "learning_rate": 6.409250464588066e-05,
+      "loss": 1.021,
+      "step": 9190
+    },
+    {
+      "epoch": 0.37211655200323757,
+      "grad_norm": 1.1903021335601807,
+      "learning_rate": 6.407185628742515e-05,
+      "loss": 0.9993,
+      "step": 9195
+    },
+    {
+      "epoch": 0.3723188992310805,
+      "grad_norm": 1.1956493854522705,
+      "learning_rate": 6.405120792896964e-05,
+      "loss": 1.0009,
+      "step": 9200
+    },
+    {
+      "epoch": 0.37252124645892354,
+      "grad_norm": 1.1848713159561157,
+      "learning_rate": 6.403055957051415e-05,
+      "loss": 1.0089,
+      "step": 9205
+    },
+    {
+      "epoch": 0.3727235936867665,
+      "grad_norm": 1.0949673652648926,
+      "learning_rate": 6.400991121205865e-05,
+      "loss": 1.03,
+      "step": 9210
+    },
+    {
+      "epoch": 0.37292594091460946,
+      "grad_norm": 1.3276467323303223,
+      "learning_rate": 6.398926285360314e-05,
+      "loss": 1.0071,
+      "step": 9215
+    },
+    {
+      "epoch": 0.37312828814245247,
+      "grad_norm": 1.1459453105926514,
+      "learning_rate": 6.396861449514763e-05,
+      "loss": 1.0276,
+      "step": 9220
+    },
+    {
+      "epoch": 0.3733306353702954,
+      "grad_norm": 1.1494052410125732,
+      "learning_rate": 6.394796613669214e-05,
+      "loss": 1.0309,
+      "step": 9225
+    },
+    {
+      "epoch": 0.3735329825981384,
+      "grad_norm": 1.1526970863342285,
+      "learning_rate": 6.392731777823663e-05,
+      "loss": 1.0385,
+      "step": 9230
+    },
+    {
+      "epoch": 0.3737353298259814,
+      "grad_norm": 1.1462080478668213,
+      "learning_rate": 6.390666941978113e-05,
+      "loss": 1.0035,
+      "step": 9235
+    },
+    {
+      "epoch": 0.37393767705382436,
+      "grad_norm": 1.1997939348220825,
+      "learning_rate": 6.388602106132563e-05,
+      "loss": 0.9418,
+      "step": 9240
+    },
+    {
+      "epoch": 0.3741400242816673,
+      "grad_norm": 1.1789708137512207,
+      "learning_rate": 6.386537270287013e-05,
+      "loss": 1.0534,
+      "step": 9245
+    },
+    {
+      "epoch": 0.37434237150951033,
+      "grad_norm": 1.092958688735962,
+      "learning_rate": 6.384472434441462e-05,
+      "loss": 1.0371,
+      "step": 9250
+    },
+    {
+      "epoch": 0.3745447187373533,
+      "grad_norm": 1.2328760623931885,
+      "learning_rate": 6.382407598595912e-05,
+      "loss": 1.0039,
+      "step": 9255
+    },
+    {
+      "epoch": 0.3747470659651963,
+      "grad_norm": 1.1537787914276123,
+      "learning_rate": 6.380342762750362e-05,
+      "loss": 1.0679,
+      "step": 9260
+    },
+    {
+      "epoch": 0.37494941319303926,
+      "grad_norm": 1.1393468379974365,
+      "learning_rate": 6.378277926904811e-05,
+      "loss": 1.0122,
+      "step": 9265
+    },
+    {
+      "epoch": 0.3751517604208822,
+      "grad_norm": 1.287347435951233,
+      "learning_rate": 6.37621309105926e-05,
+      "loss": 0.9618,
+      "step": 9270
+    },
+    {
+      "epoch": 0.37535410764872523,
+      "grad_norm": 1.0002747774124146,
+      "learning_rate": 6.374148255213711e-05,
+      "loss": 0.9842,
+      "step": 9275
+    },
+    {
+      "epoch": 0.3755564548765682,
+      "grad_norm": 1.2078219652175903,
+      "learning_rate": 6.37208341936816e-05,
+      "loss": 1.0083,
+      "step": 9280
+    },
+    {
+      "epoch": 0.37575880210441115,
+      "grad_norm": 1.1521834135055542,
+      "learning_rate": 6.37001858352261e-05,
+      "loss": 0.9759,
+      "step": 9285
+    },
+    {
+      "epoch": 0.37596114933225416,
+      "grad_norm": 1.1871899366378784,
+      "learning_rate": 6.367953747677061e-05,
+      "loss": 1.0046,
+      "step": 9290
+    },
+    {
+      "epoch": 0.3761634965600971,
+      "grad_norm": 1.1746609210968018,
+      "learning_rate": 6.36588891183151e-05,
+      "loss": 1.0287,
+      "step": 9295
+    },
+    {
+      "epoch": 0.3763658437879401,
+      "grad_norm": 1.1792657375335693,
+      "learning_rate": 6.363824075985959e-05,
+      "loss": 1.0478,
+      "step": 9300
+    },
+    {
+      "epoch": 0.3765681910157831,
+      "grad_norm": 1.1547367572784424,
+      "learning_rate": 6.361759240140409e-05,
+      "loss": 1.0958,
+      "step": 9305
+    },
+    {
+      "epoch": 0.37677053824362605,
+      "grad_norm": 1.2225704193115234,
+      "learning_rate": 6.35969440429486e-05,
+      "loss": 1.0126,
+      "step": 9310
+    },
+    {
+      "epoch": 0.37697288547146907,
+      "grad_norm": 1.0655434131622314,
+      "learning_rate": 6.357629568449309e-05,
+      "loss": 1.0005,
+      "step": 9315
+    },
+    {
+      "epoch": 0.377175232699312,
+      "grad_norm": 1.0822560787200928,
+      "learning_rate": 6.355564732603758e-05,
+      "loss": 1.0338,
+      "step": 9320
+    },
+    {
+      "epoch": 0.377377579927155,
+      "grad_norm": 1.159787893295288,
+      "learning_rate": 6.353499896758208e-05,
+      "loss": 1.034,
+      "step": 9325
+    },
+    {
+      "epoch": 0.377579927154998,
+      "grad_norm": 1.0686500072479248,
+      "learning_rate": 6.351435060912659e-05,
+      "loss": 1.0069,
+      "step": 9330
+    },
+    {
+      "epoch": 0.37778227438284095,
+      "grad_norm": 1.304579734802246,
+      "learning_rate": 6.349370225067108e-05,
+      "loss": 0.9354,
+      "step": 9335
+    },
+    {
+      "epoch": 0.3779846216106839,
+      "grad_norm": 1.1627440452575684,
+      "learning_rate": 6.347305389221557e-05,
+      "loss": 0.9977,
+      "step": 9340
+    },
+    {
+      "epoch": 0.3781869688385269,
+      "grad_norm": 1.22258460521698,
+      "learning_rate": 6.345240553376007e-05,
+      "loss": 1.009,
+      "step": 9345
+    },
+    {
+      "epoch": 0.3783893160663699,
+      "grad_norm": 1.247749924659729,
+      "learning_rate": 6.343175717530456e-05,
+      "loss": 0.9786,
+      "step": 9350
+    },
+    {
+      "epoch": 0.37859166329421284,
+      "grad_norm": 1.127065896987915,
+      "learning_rate": 6.341110881684907e-05,
+      "loss": 0.9699,
+      "step": 9355
+    },
+    {
+      "epoch": 0.37879401052205586,
+      "grad_norm": 1.182802677154541,
+      "learning_rate": 6.339046045839357e-05,
+      "loss": 0.9907,
+      "step": 9360
+    },
+    {
+      "epoch": 0.3789963577498988,
+      "grad_norm": 1.0922465324401855,
+      "learning_rate": 6.336981209993805e-05,
+      "loss": 0.9663,
+      "step": 9365
+    },
+    {
+      "epoch": 0.37919870497774183,
+      "grad_norm": 1.1843430995941162,
+      "learning_rate": 6.334916374148255e-05,
+      "loss": 1.0436,
+      "step": 9370
+    },
+    {
+      "epoch": 0.3794010522055848,
+      "grad_norm": 1.1872503757476807,
+      "learning_rate": 6.332851538302705e-05,
+      "loss": 1.0583,
+      "step": 9375
+    },
+    {
+      "epoch": 0.37960339943342775,
+      "grad_norm": 1.1438274383544922,
+      "learning_rate": 6.330786702457156e-05,
+      "loss": 1.0213,
+      "step": 9380
+    },
+    {
+      "epoch": 0.37980574666127076,
+      "grad_norm": 1.1314841508865356,
+      "learning_rate": 6.328721866611605e-05,
+      "loss": 1.0038,
+      "step": 9385
+    },
+    {
+      "epoch": 0.3800080938891137,
+      "grad_norm": 1.3166422843933105,
+      "learning_rate": 6.326657030766054e-05,
+      "loss": 1.0751,
+      "step": 9390
+    },
+    {
+      "epoch": 0.3802104411169567,
+      "grad_norm": 1.115728735923767,
+      "learning_rate": 6.324592194920504e-05,
+      "loss": 1.033,
+      "step": 9395
+    },
+    {
+      "epoch": 0.3804127883447997,
+      "grad_norm": 1.0546056032180786,
+      "learning_rate": 6.322527359074953e-05,
+      "loss": 0.931,
+      "step": 9400
+    },
+    {
+      "epoch": 0.38061513557264265,
+      "grad_norm": 1.1586906909942627,
+      "learning_rate": 6.320462523229404e-05,
+      "loss": 1.0064,
+      "step": 9405
+    },
+    {
+      "epoch": 0.3808174828004856,
+      "grad_norm": 1.2858645915985107,
+      "learning_rate": 6.318397687383853e-05,
+      "loss": 1.0041,
+      "step": 9410
+    },
+    {
+      "epoch": 0.3810198300283286,
+      "grad_norm": 1.1011475324630737,
+      "learning_rate": 6.316332851538302e-05,
+      "loss": 1.0199,
+      "step": 9415
+    },
+    {
+      "epoch": 0.3812221772561716,
+      "grad_norm": 1.1812069416046143,
+      "learning_rate": 6.314268015692752e-05,
+      "loss": 0.998,
+      "step": 9420
+    },
+    {
+      "epoch": 0.3814245244840146,
+      "grad_norm": 1.2132670879364014,
+      "learning_rate": 6.312203179847203e-05,
+      "loss": 1.0099,
+      "step": 9425
+    },
+    {
+      "epoch": 0.38162687171185755,
+      "grad_norm": 1.3264063596725464,
+      "learning_rate": 6.310138344001653e-05,
+      "loss": 1.0115,
+      "step": 9430
+    },
+    {
+      "epoch": 0.3818292189397005,
+      "grad_norm": 1.084452509880066,
+      "learning_rate": 6.308073508156101e-05,
+      "loss": 0.9762,
+      "step": 9435
+    },
+    {
+      "epoch": 0.3820315661675435,
+      "grad_norm": 1.258212685585022,
+      "learning_rate": 6.306008672310551e-05,
+      "loss": 0.9482,
+      "step": 9440
+    },
+    {
+      "epoch": 0.3822339133953865,
+      "grad_norm": 1.2183173894882202,
+      "learning_rate": 6.303943836465002e-05,
+      "loss": 0.9731,
+      "step": 9445
+    },
+    {
+      "epoch": 0.38243626062322944,
+      "grad_norm": 1.2455830574035645,
+      "learning_rate": 6.301879000619451e-05,
+      "loss": 1.0306,
+      "step": 9450
+    },
+    {
+      "epoch": 0.38263860785107245,
+      "grad_norm": 1.2332731485366821,
+      "learning_rate": 6.299814164773901e-05,
+      "loss": 1.0703,
+      "step": 9455
+    },
+    {
+      "epoch": 0.3828409550789154,
+      "grad_norm": 1.212498664855957,
+      "learning_rate": 6.29774932892835e-05,
+      "loss": 0.9481,
+      "step": 9460
+    },
+    {
+      "epoch": 0.38304330230675837,
+      "grad_norm": 1.132043719291687,
+      "learning_rate": 6.2956844930828e-05,
+      "loss": 0.9952,
+      "step": 9465
+    },
+    {
+      "epoch": 0.3832456495346014,
+      "grad_norm": 1.193706750869751,
+      "learning_rate": 6.29361965723725e-05,
+      "loss": 0.9954,
+      "step": 9470
+    },
+    {
+      "epoch": 0.38344799676244434,
+      "grad_norm": 1.206764817237854,
+      "learning_rate": 6.2915548213917e-05,
+      "loss": 1.0108,
+      "step": 9475
+    },
+    {
+      "epoch": 0.38365034399028736,
+      "grad_norm": 1.2077068090438843,
+      "learning_rate": 6.289489985546149e-05,
+      "loss": 1.0416,
+      "step": 9480
+    },
+    {
+      "epoch": 0.3838526912181303,
+      "grad_norm": 1.2133302688598633,
+      "learning_rate": 6.287425149700598e-05,
+      "loss": 1.0439,
+      "step": 9485
+    },
+    {
+      "epoch": 0.3840550384459733,
+      "grad_norm": 1.1433510780334473,
+      "learning_rate": 6.285360313855049e-05,
+      "loss": 0.995,
+      "step": 9490
+    },
+    {
+      "epoch": 0.3842573856738163,
+      "grad_norm": 1.2141852378845215,
+      "learning_rate": 6.283295478009499e-05,
+      "loss": 0.9989,
+      "step": 9495
+    },
+    {
+      "epoch": 0.38445973290165925,
+      "grad_norm": 1.1626108884811401,
+      "learning_rate": 6.281230642163948e-05,
+      "loss": 0.9894,
+      "step": 9500
+    },
+    {
+      "epoch": 0.3846620801295022,
+      "grad_norm": 1.2495766878128052,
+      "learning_rate": 6.279165806318398e-05,
+      "loss": 0.9594,
+      "step": 9505
+    },
+    {
+      "epoch": 0.3848644273573452,
+      "grad_norm": 1.1957738399505615,
+      "learning_rate": 6.277100970472848e-05,
+      "loss": 1.0635,
+      "step": 9510
+    },
+    {
+      "epoch": 0.3850667745851882,
+      "grad_norm": 1.1707223653793335,
+      "learning_rate": 6.275036134627298e-05,
+      "loss": 1.0226,
+      "step": 9515
+    },
+    {
+      "epoch": 0.38526912181303113,
+      "grad_norm": 1.2480435371398926,
+      "learning_rate": 6.272971298781747e-05,
+      "loss": 1.0637,
+      "step": 9520
+    },
+    {
+      "epoch": 0.38547146904087415,
+      "grad_norm": 1.229467749595642,
+      "learning_rate": 6.270906462936197e-05,
+      "loss": 0.9747,
+      "step": 9525
+    },
+    {
+      "epoch": 0.3856738162687171,
+      "grad_norm": 1.2177376747131348,
+      "learning_rate": 6.268841627090646e-05,
+      "loss": 0.9665,
+      "step": 9530
+    },
+    {
+      "epoch": 0.3858761634965601,
+      "grad_norm": 1.1545181274414062,
+      "learning_rate": 6.266776791245095e-05,
+      "loss": 0.9642,
+      "step": 9535
+    },
+    {
+      "epoch": 0.3860785107244031,
+      "grad_norm": 1.0587862730026245,
+      "learning_rate": 6.264711955399546e-05,
+      "loss": 0.9979,
+      "step": 9540
+    },
+    {
+      "epoch": 0.38628085795224604,
+      "grad_norm": 1.0684212446212769,
+      "learning_rate": 6.262647119553996e-05,
+      "loss": 1.0164,
+      "step": 9545
+    },
+    {
+      "epoch": 0.38648320518008905,
+      "grad_norm": 1.3630390167236328,
+      "learning_rate": 6.260582283708445e-05,
+      "loss": 0.9687,
+      "step": 9550
+    },
+    {
+      "epoch": 0.386685552407932,
+      "grad_norm": 1.1170276403427124,
+      "learning_rate": 6.258517447862894e-05,
+      "loss": 1.0348,
+      "step": 9555
+    },
+    {
+      "epoch": 0.38688789963577497,
+      "grad_norm": 1.0596883296966553,
+      "learning_rate": 6.256452612017345e-05,
+      "loss": 1.0136,
+      "step": 9560
+    },
+    {
+      "epoch": 0.387090246863618,
+      "grad_norm": 1.0961753129959106,
+      "learning_rate": 6.254387776171795e-05,
+      "loss": 0.9488,
+      "step": 9565
+    },
+    {
+      "epoch": 0.38729259409146094,
+      "grad_norm": 1.198262333869934,
+      "learning_rate": 6.252322940326244e-05,
+      "loss": 1.0494,
+      "step": 9570
+    },
+    {
+      "epoch": 0.3874949413193039,
+      "grad_norm": 1.2458857297897339,
+      "learning_rate": 6.250258104480695e-05,
+      "loss": 1.0483,
+      "step": 9575
+    },
+    {
+      "epoch": 0.3876972885471469,
+      "grad_norm": 1.0762958526611328,
+      "learning_rate": 6.248193268635144e-05,
+      "loss": 1.0529,
+      "step": 9580
+    },
+    {
+      "epoch": 0.38789963577498987,
+      "grad_norm": 1.2601096630096436,
+      "learning_rate": 6.246128432789593e-05,
+      "loss": 0.9977,
+      "step": 9585
+    },
+    {
+      "epoch": 0.3881019830028329,
+      "grad_norm": 1.2507660388946533,
+      "learning_rate": 6.244063596944043e-05,
+      "loss": 1.0319,
+      "step": 9590
+    },
+    {
+      "epoch": 0.38830433023067584,
+      "grad_norm": 1.0800540447235107,
+      "learning_rate": 6.241998761098494e-05,
+      "loss": 1.0686,
+      "step": 9595
+    },
+    {
+      "epoch": 0.3885066774585188,
+      "grad_norm": 1.1687123775482178,
+      "learning_rate": 6.239933925252943e-05,
+      "loss": 1.0005,
+      "step": 9600
+    },
+    {
+      "epoch": 0.3887090246863618,
+      "grad_norm": 1.1139582395553589,
+      "learning_rate": 6.237869089407392e-05,
+      "loss": 1.0277,
+      "step": 9605
+    },
+    {
+      "epoch": 0.3889113719142048,
+      "grad_norm": 1.252946376800537,
+      "learning_rate": 6.235804253561842e-05,
+      "loss": 1.0128,
+      "step": 9610
+    },
+    {
+      "epoch": 0.38911371914204773,
+      "grad_norm": 1.088563084602356,
+      "learning_rate": 6.233739417716293e-05,
+      "loss": 0.9953,
+      "step": 9615
+    },
+    {
+      "epoch": 0.38931606636989075,
+      "grad_norm": 1.0879402160644531,
+      "learning_rate": 6.231674581870742e-05,
+      "loss": 0.9698,
+      "step": 9620
+    },
+    {
+      "epoch": 0.3895184135977337,
+      "grad_norm": 1.1478641033172607,
+      "learning_rate": 6.22960974602519e-05,
+      "loss": 1.0346,
+      "step": 9625
+    },
+    {
+      "epoch": 0.38972076082557666,
+      "grad_norm": 1.1672147512435913,
+      "learning_rate": 6.227544910179641e-05,
+      "loss": 0.9942,
+      "step": 9630
+    },
+    {
+      "epoch": 0.3899231080534197,
+      "grad_norm": 1.1004835367202759,
+      "learning_rate": 6.22548007433409e-05,
+      "loss": 1.0549,
+      "step": 9635
+    },
+    {
+      "epoch": 0.39012545528126263,
+      "grad_norm": 1.100008487701416,
+      "learning_rate": 6.22341523848854e-05,
+      "loss": 1.0119,
+      "step": 9640
+    },
+    {
+      "epoch": 0.39032780250910565,
+      "grad_norm": 1.2148017883300781,
+      "learning_rate": 6.221350402642991e-05,
+      "loss": 1.0729,
+      "step": 9645
+    },
+    {
+      "epoch": 0.3905301497369486,
+      "grad_norm": 1.2151126861572266,
+      "learning_rate": 6.21928556679744e-05,
+      "loss": 1.0249,
+      "step": 9650
+    },
+    {
+      "epoch": 0.39073249696479156,
+      "grad_norm": 1.074285864830017,
+      "learning_rate": 6.217220730951889e-05,
+      "loss": 1.0042,
+      "step": 9655
+    },
+    {
+      "epoch": 0.3909348441926346,
+      "grad_norm": 1.1544793844223022,
+      "learning_rate": 6.21515589510634e-05,
+      "loss": 1.0442,
+      "step": 9660
+    },
+    {
+      "epoch": 0.39113719142047754,
+      "grad_norm": 1.2228337526321411,
+      "learning_rate": 6.21309105926079e-05,
+      "loss": 1.0684,
+      "step": 9665
+    },
+    {
+      "epoch": 0.3913395386483205,
+      "grad_norm": 1.201342225074768,
+      "learning_rate": 6.211026223415239e-05,
+      "loss": 1.0026,
+      "step": 9670
+    },
+    {
+      "epoch": 0.3915418858761635,
+      "grad_norm": 1.127369999885559,
+      "learning_rate": 6.208961387569688e-05,
+      "loss": 0.999,
+      "step": 9675
+    },
+    {
+      "epoch": 0.39174423310400647,
+      "grad_norm": 1.2213554382324219,
+      "learning_rate": 6.206896551724138e-05,
+      "loss": 0.9542,
+      "step": 9680
+    },
+    {
+      "epoch": 0.3919465803318495,
+      "grad_norm": 1.2262746095657349,
+      "learning_rate": 6.204831715878589e-05,
+      "loss": 1.0872,
+      "step": 9685
+    },
+    {
+      "epoch": 0.39214892755969244,
+      "grad_norm": 1.2475377321243286,
+      "learning_rate": 6.202766880033038e-05,
+      "loss": 1.0248,
+      "step": 9690
+    },
+    {
+      "epoch": 0.3923512747875354,
+      "grad_norm": 1.161120057106018,
+      "learning_rate": 6.200702044187487e-05,
+      "loss": 1.0235,
+      "step": 9695
+    },
+    {
+      "epoch": 0.3925536220153784,
+      "grad_norm": 1.3133748769760132,
+      "learning_rate": 6.198637208341937e-05,
+      "loss": 1.0224,
+      "step": 9700
+    },
+    {
+      "epoch": 0.39275596924322137,
+      "grad_norm": 1.2146892547607422,
+      "learning_rate": 6.196572372496386e-05,
+      "loss": 1.0056,
+      "step": 9705
+    },
+    {
+      "epoch": 0.39295831647106433,
+      "grad_norm": 1.1127619743347168,
+      "learning_rate": 6.194507536650837e-05,
+      "loss": 1.0124,
+      "step": 9710
+    },
+    {
+      "epoch": 0.39316066369890734,
+      "grad_norm": 1.1972129344940186,
+      "learning_rate": 6.192442700805287e-05,
+      "loss": 1.0149,
+      "step": 9715
+    },
+    {
+      "epoch": 0.3933630109267503,
+      "grad_norm": 1.1928164958953857,
+      "learning_rate": 6.190377864959736e-05,
+      "loss": 1.018,
+      "step": 9720
+    },
+    {
+      "epoch": 0.39356535815459326,
+      "grad_norm": 1.1728615760803223,
+      "learning_rate": 6.188313029114185e-05,
+      "loss": 0.9853,
+      "step": 9725
+    },
+    {
+      "epoch": 0.3937677053824363,
+      "grad_norm": 1.1502978801727295,
+      "learning_rate": 6.186248193268636e-05,
+      "loss": 0.9828,
+      "step": 9730
+    },
+    {
+      "epoch": 0.39397005261027923,
+      "grad_norm": 1.169958233833313,
+      "learning_rate": 6.184183357423086e-05,
+      "loss": 0.9781,
+      "step": 9735
+    },
+    {
+      "epoch": 0.39417239983812224,
+      "grad_norm": 1.134246826171875,
+      "learning_rate": 6.182118521577535e-05,
+      "loss": 1.0098,
+      "step": 9740
+    },
+    {
+      "epoch": 0.3943747470659652,
+      "grad_norm": 1.4439961910247803,
+      "learning_rate": 6.180053685731984e-05,
+      "loss": 1.0077,
+      "step": 9745
+    },
+    {
+      "epoch": 0.39457709429380816,
+      "grad_norm": 1.2301324605941772,
+      "learning_rate": 6.177988849886435e-05,
+      "loss": 0.9961,
+      "step": 9750
+    },
+    {
+      "epoch": 0.3947794415216512,
+      "grad_norm": 1.174716591835022,
+      "learning_rate": 6.175924014040884e-05,
+      "loss": 1.0414,
+      "step": 9755
+    },
+    {
+      "epoch": 0.39498178874949413,
+      "grad_norm": 1.16427743434906,
+      "learning_rate": 6.173859178195334e-05,
+      "loss": 1.0298,
+      "step": 9760
+    },
+    {
+      "epoch": 0.3951841359773371,
+      "grad_norm": 1.3493810892105103,
+      "learning_rate": 6.171794342349783e-05,
+      "loss": 1.0464,
+      "step": 9765
+    },
+    {
+      "epoch": 0.3953864832051801,
+      "grad_norm": 1.3521122932434082,
+      "learning_rate": 6.169729506504233e-05,
+      "loss": 0.9717,
+      "step": 9770
+    },
+    {
+      "epoch": 0.39558883043302306,
+      "grad_norm": 1.2052327394485474,
+      "learning_rate": 6.167664670658683e-05,
+      "loss": 0.9965,
+      "step": 9775
+    },
+    {
+      "epoch": 0.395791177660866,
+      "grad_norm": 1.150696873664856,
+      "learning_rate": 6.165599834813133e-05,
+      "loss": 0.9972,
+      "step": 9780
+    },
+    {
+      "epoch": 0.39599352488870904,
+      "grad_norm": 1.1349776983261108,
+      "learning_rate": 6.163534998967583e-05,
+      "loss": 1.0575,
+      "step": 9785
+    },
+    {
+      "epoch": 0.396195872116552,
+      "grad_norm": 1.1687307357788086,
+      "learning_rate": 6.161470163122032e-05,
+      "loss": 1.0335,
+      "step": 9790
+    },
+    {
+      "epoch": 0.396398219344395,
+      "grad_norm": 1.2072405815124512,
+      "learning_rate": 6.159405327276481e-05,
+      "loss": 1.0044,
+      "step": 9795
+    },
+    {
+      "epoch": 0.39660056657223797,
+      "grad_norm": 1.060883641242981,
+      "learning_rate": 6.157340491430932e-05,
+      "loss": 1.0538,
+      "step": 9800
+    },
+    {
+      "epoch": 0.3968029138000809,
+      "grad_norm": 1.0977377891540527,
+      "learning_rate": 6.155275655585381e-05,
+      "loss": 0.9912,
+      "step": 9805
+    },
+    {
+      "epoch": 0.39700526102792394,
+      "grad_norm": 1.2743343114852905,
+      "learning_rate": 6.153210819739831e-05,
+      "loss": 1.0414,
+      "step": 9810
+    },
+    {
+      "epoch": 0.3972076082557669,
+      "grad_norm": 1.1622213125228882,
+      "learning_rate": 6.15114598389428e-05,
+      "loss": 0.9953,
+      "step": 9815
+    },
+    {
+      "epoch": 0.39740995548360986,
+      "grad_norm": 1.1404904127120972,
+      "learning_rate": 6.149081148048731e-05,
+      "loss": 1.035,
+      "step": 9820
+    },
+    {
+      "epoch": 0.39761230271145287,
+      "grad_norm": 1.1930956840515137,
+      "learning_rate": 6.14701631220318e-05,
+      "loss": 1.0397,
+      "step": 9825
+    },
+    {
+      "epoch": 0.39781464993929583,
+      "grad_norm": 1.158272385597229,
+      "learning_rate": 6.14495147635763e-05,
+      "loss": 1.0771,
+      "step": 9830
+    },
+    {
+      "epoch": 0.3980169971671388,
+      "grad_norm": 1.0452218055725098,
+      "learning_rate": 6.14288664051208e-05,
+      "loss": 1.029,
+      "step": 9835
+    },
+    {
+      "epoch": 0.3982193443949818,
+      "grad_norm": 1.1018149852752686,
+      "learning_rate": 6.140821804666528e-05,
+      "loss": 0.981,
+      "step": 9840
+    },
+    {
+      "epoch": 0.39842169162282476,
+      "grad_norm": 1.0617828369140625,
+      "learning_rate": 6.138756968820979e-05,
+      "loss": 1.0095,
+      "step": 9845
+    },
+    {
+      "epoch": 0.3986240388506678,
+      "grad_norm": 1.1889824867248535,
+      "learning_rate": 6.136692132975429e-05,
+      "loss": 0.9951,
+      "step": 9850
+    },
+    {
+      "epoch": 0.39882638607851073,
+      "grad_norm": 1.192192792892456,
+      "learning_rate": 6.134627297129878e-05,
+      "loss": 1.0154,
+      "step": 9855
+    },
+    {
+      "epoch": 0.3990287333063537,
+      "grad_norm": 1.2727915048599243,
+      "learning_rate": 6.132562461284329e-05,
+      "loss": 0.9683,
+      "step": 9860
+    },
+    {
+      "epoch": 0.3992310805341967,
+      "grad_norm": 1.1856448650360107,
+      "learning_rate": 6.130497625438778e-05,
+      "loss": 0.9876,
+      "step": 9865
+    },
+    {
+      "epoch": 0.39943342776203966,
+      "grad_norm": 1.104466438293457,
+      "learning_rate": 6.128432789593228e-05,
+      "loss": 0.9436,
+      "step": 9870
+    },
+    {
+      "epoch": 0.3996357749898826,
+      "grad_norm": 1.1249479055404663,
+      "learning_rate": 6.126367953747677e-05,
+      "loss": 1.0464,
+      "step": 9875
+    },
+    {
+      "epoch": 0.39983812221772563,
+      "grad_norm": 1.209954857826233,
+      "learning_rate": 6.124303117902128e-05,
+      "loss": 1.0734,
+      "step": 9880
+    },
+    {
+      "epoch": 0.4000404694455686,
+      "grad_norm": 1.080848217010498,
+      "learning_rate": 6.122238282056577e-05,
+      "loss": 0.9687,
+      "step": 9885
+    },
+    {
+      "epoch": 0.40024281667341155,
+      "grad_norm": 1.124177098274231,
+      "learning_rate": 6.120173446211026e-05,
+      "loss": 1.0057,
+      "step": 9890
+    },
+    {
+      "epoch": 0.40044516390125456,
+      "grad_norm": 1.0730397701263428,
+      "learning_rate": 6.118108610365476e-05,
+      "loss": 0.9908,
+      "step": 9895
+    },
+    {
+      "epoch": 0.4006475111290975,
+      "grad_norm": 1.192895531654358,
+      "learning_rate": 6.116043774519926e-05,
+      "loss": 1.0482,
+      "step": 9900
+    },
+    {
+      "epoch": 0.40084985835694054,
+      "grad_norm": 1.1992424726486206,
+      "learning_rate": 6.113978938674377e-05,
+      "loss": 0.9487,
+      "step": 9905
+    },
+    {
+      "epoch": 0.4010522055847835,
+      "grad_norm": 1.164494276046753,
+      "learning_rate": 6.111914102828825e-05,
+      "loss": 0.9974,
+      "step": 9910
+    },
+    {
+      "epoch": 0.40125455281262645,
+      "grad_norm": 0.9880785942077637,
+      "learning_rate": 6.109849266983275e-05,
+      "loss": 1.0034,
+      "step": 9915
+    },
+    {
+      "epoch": 0.40145690004046947,
+      "grad_norm": 1.2381869554519653,
+      "learning_rate": 6.107784431137725e-05,
+      "loss": 1.0313,
+      "step": 9920
+    },
+    {
+      "epoch": 0.4016592472683124,
+      "grad_norm": 1.1616636514663696,
+      "learning_rate": 6.105719595292174e-05,
+      "loss": 1.0086,
+      "step": 9925
+    },
+    {
+      "epoch": 0.4018615944961554,
+      "grad_norm": 1.2813246250152588,
+      "learning_rate": 6.103654759446625e-05,
+      "loss": 0.9827,
+      "step": 9930
+    },
+    {
+      "epoch": 0.4020639417239984,
+      "grad_norm": 1.1164380311965942,
+      "learning_rate": 6.1015899236010745e-05,
+      "loss": 1.0219,
+      "step": 9935
+    },
+    {
+      "epoch": 0.40226628895184136,
+      "grad_norm": 1.4000154733657837,
+      "learning_rate": 6.099525087755523e-05,
+      "loss": 0.9641,
+      "step": 9940
+    },
+    {
+      "epoch": 0.4024686361796843,
+      "grad_norm": 1.2360674142837524,
+      "learning_rate": 6.097460251909973e-05,
+      "loss": 0.9762,
+      "step": 9945
+    },
+    {
+      "epoch": 0.4026709834075273,
+      "grad_norm": 1.1454941034317017,
+      "learning_rate": 6.095395416064423e-05,
+      "loss": 1.0313,
+      "step": 9950
+    },
+    {
+      "epoch": 0.4028733306353703,
+      "grad_norm": 1.0879707336425781,
+      "learning_rate": 6.0933305802188735e-05,
+      "loss": 1.003,
+      "step": 9955
+    },
+    {
+      "epoch": 0.4030756778632133,
+      "grad_norm": 1.1175845861434937,
+      "learning_rate": 6.0912657443733225e-05,
+      "loss": 0.9766,
+      "step": 9960
+    },
+    {
+      "epoch": 0.40327802509105626,
+      "grad_norm": 1.144715428352356,
+      "learning_rate": 6.089200908527772e-05,
+      "loss": 0.9904,
+      "step": 9965
+    },
+    {
+      "epoch": 0.4034803723188992,
+      "grad_norm": 1.1816964149475098,
+      "learning_rate": 6.0871360726822226e-05,
+      "loss": 1.0159,
+      "step": 9970
+    },
+    {
+      "epoch": 0.40368271954674223,
+      "grad_norm": 1.2622935771942139,
+      "learning_rate": 6.085071236836671e-05,
+      "loss": 0.9864,
+      "step": 9975
+    },
+    {
+      "epoch": 0.4038850667745852,
+      "grad_norm": 1.110995888710022,
+      "learning_rate": 6.0830064009911214e-05,
+      "loss": 0.9845,
+      "step": 9980
+    },
+    {
+      "epoch": 0.40408741400242815,
+      "grad_norm": 1.1176053285598755,
+      "learning_rate": 6.080941565145571e-05,
+      "loss": 0.9725,
+      "step": 9985
+    },
+    {
+      "epoch": 0.40428976123027116,
+      "grad_norm": 1.1164429187774658,
+      "learning_rate": 6.07887672930002e-05,
+      "loss": 0.9887,
+      "step": 9990
+    },
+    {
+      "epoch": 0.4044921084581141,
+      "grad_norm": 1.1734952926635742,
+      "learning_rate": 6.0768118934544706e-05,
+      "loss": 0.9763,
+      "step": 9995
+    },
+    {
+      "epoch": 0.4046944556859571,
+      "grad_norm": 1.2598659992218018,
+      "learning_rate": 6.0747470576089203e-05,
+      "loss": 0.9681,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4048968029138001,
+      "grad_norm": 1.0857003927230835,
+      "learning_rate": 6.072682221763371e-05,
+      "loss": 0.9386,
+      "step": 10005
+    },
+    {
+      "epoch": 0.40509915014164305,
+      "grad_norm": 1.3082906007766724,
+      "learning_rate": 6.07061738591782e-05,
+      "loss": 1.0256,
+      "step": 10010
+    },
+    {
+      "epoch": 0.40530149736948606,
+      "grad_norm": 1.2251588106155396,
+      "learning_rate": 6.0685525500722695e-05,
+      "loss": 0.9768,
+      "step": 10015
+    },
+    {
+      "epoch": 0.405503844597329,
+      "grad_norm": 1.1803746223449707,
+      "learning_rate": 6.066487714226719e-05,
+      "loss": 1.0686,
+      "step": 10020
+    },
+    {
+      "epoch": 0.405706191825172,
+      "grad_norm": 1.1959571838378906,
+      "learning_rate": 6.064422878381168e-05,
+      "loss": 1.0325,
+      "step": 10025
+    },
+    {
+      "epoch": 0.405908539053015,
+      "grad_norm": 1.0180895328521729,
+      "learning_rate": 6.062358042535619e-05,
+      "loss": 1.0369,
+      "step": 10030
+    },
+    {
+      "epoch": 0.40611088628085795,
+      "grad_norm": 1.203304409980774,
+      "learning_rate": 6.0602932066900685e-05,
+      "loss": 0.9924,
+      "step": 10035
+    },
+    {
+      "epoch": 0.4063132335087009,
+      "grad_norm": 1.200363278388977,
+      "learning_rate": 6.058228370844519e-05,
+      "loss": 1.0219,
+      "step": 10040
+    },
+    {
+      "epoch": 0.4065155807365439,
+      "grad_norm": 1.0810167789459229,
+      "learning_rate": 6.056163534998968e-05,
+      "loss": 0.9429,
+      "step": 10045
+    },
+    {
+      "epoch": 0.4067179279643869,
+      "grad_norm": 1.1412184238433838,
+      "learning_rate": 6.0540986991534176e-05,
+      "loss": 0.9684,
+      "step": 10050
+    },
+    {
+      "epoch": 0.40692027519222984,
+      "grad_norm": 1.1554936170578003,
+      "learning_rate": 6.0520338633078674e-05,
+      "loss": 1.0392,
+      "step": 10055
+    },
+    {
+      "epoch": 0.40712262242007285,
+      "grad_norm": 1.2978181838989258,
+      "learning_rate": 6.0499690274623164e-05,
+      "loss": 1.0067,
+      "step": 10060
+    },
+    {
+      "epoch": 0.4073249696479158,
+      "grad_norm": 1.1897743940353394,
+      "learning_rate": 6.047904191616767e-05,
+      "loss": 1.0091,
+      "step": 10065
+    },
+    {
+      "epoch": 0.4075273168757588,
+      "grad_norm": 1.1835557222366333,
+      "learning_rate": 6.0458393557712166e-05,
+      "loss": 1.0197,
+      "step": 10070
+    },
+    {
+      "epoch": 0.4077296641036018,
+      "grad_norm": 1.0526257753372192,
+      "learning_rate": 6.0437745199256656e-05,
+      "loss": 1.0808,
+      "step": 10075
+    },
+    {
+      "epoch": 0.40793201133144474,
+      "grad_norm": 1.0810807943344116,
+      "learning_rate": 6.041709684080116e-05,
+      "loss": 0.9582,
+      "step": 10080
+    },
+    {
+      "epoch": 0.40813435855928776,
+      "grad_norm": 1.1587885618209839,
+      "learning_rate": 6.039644848234566e-05,
+      "loss": 0.969,
+      "step": 10085
+    },
+    {
+      "epoch": 0.4083367057871307,
+      "grad_norm": 1.2569379806518555,
+      "learning_rate": 6.037580012389016e-05,
+      "loss": 1.0162,
+      "step": 10090
+    },
+    {
+      "epoch": 0.4085390530149737,
+      "grad_norm": 1.0843753814697266,
+      "learning_rate": 6.0355151765434645e-05,
+      "loss": 0.9554,
+      "step": 10095
+    },
+    {
+      "epoch": 0.4087414002428167,
+      "grad_norm": 1.225560188293457,
+      "learning_rate": 6.033450340697915e-05,
+      "loss": 1.0209,
+      "step": 10100
+    },
+    {
+      "epoch": 0.40894374747065965,
+      "grad_norm": 1.1393972635269165,
+      "learning_rate": 6.031385504852365e-05,
+      "loss": 1.0273,
+      "step": 10105
+    },
+    {
+      "epoch": 0.4091460946985026,
+      "grad_norm": 1.196390151977539,
+      "learning_rate": 6.029320669006814e-05,
+      "loss": 1.0688,
+      "step": 10110
+    },
+    {
+      "epoch": 0.4093484419263456,
+      "grad_norm": 1.1796727180480957,
+      "learning_rate": 6.027255833161264e-05,
+      "loss": 0.9621,
+      "step": 10115
+    },
+    {
+      "epoch": 0.4095507891541886,
+      "grad_norm": 1.1920093297958374,
+      "learning_rate": 6.025190997315714e-05,
+      "loss": 1.0199,
+      "step": 10120
+    },
+    {
+      "epoch": 0.4097531363820316,
+      "grad_norm": 1.0450389385223389,
+      "learning_rate": 6.023126161470164e-05,
+      "loss": 0.9026,
+      "step": 10125
+    },
+    {
+      "epoch": 0.40995548360987455,
+      "grad_norm": 1.3306684494018555,
+      "learning_rate": 6.0210613256246126e-05,
+      "loss": 1.0129,
+      "step": 10130
+    },
+    {
+      "epoch": 0.4101578308377175,
+      "grad_norm": 1.085116982460022,
+      "learning_rate": 6.018996489779063e-05,
+      "loss": 1.0175,
+      "step": 10135
+    },
+    {
+      "epoch": 0.4103601780655605,
+      "grad_norm": 1.1638381481170654,
+      "learning_rate": 6.016931653933513e-05,
+      "loss": 0.9514,
+      "step": 10140
+    },
+    {
+      "epoch": 0.4105625252934035,
+      "grad_norm": 1.2508279085159302,
+      "learning_rate": 6.014866818087962e-05,
+      "loss": 0.996,
+      "step": 10145
+    },
+    {
+      "epoch": 0.41076487252124644,
+      "grad_norm": 1.301274061203003,
+      "learning_rate": 6.012801982242412e-05,
+      "loss": 0.9411,
+      "step": 10150
+    },
+    {
+      "epoch": 0.41096721974908945,
+      "grad_norm": 1.1069871187210083,
+      "learning_rate": 6.010737146396862e-05,
+      "loss": 1.0319,
+      "step": 10155
+    },
+    {
+      "epoch": 0.4111695669769324,
+      "grad_norm": 1.1158190965652466,
+      "learning_rate": 6.008672310551311e-05,
+      "loss": 0.9787,
+      "step": 10160
+    },
+    {
+      "epoch": 0.41137191420477537,
+      "grad_norm": 1.184366226196289,
+      "learning_rate": 6.006607474705761e-05,
+      "loss": 1.0631,
+      "step": 10165
+    },
+    {
+      "epoch": 0.4115742614326184,
+      "grad_norm": 1.2500015497207642,
+      "learning_rate": 6.004542638860211e-05,
+      "loss": 0.9607,
+      "step": 10170
+    },
+    {
+      "epoch": 0.41177660866046134,
+      "grad_norm": 1.1481871604919434,
+      "learning_rate": 6.002477803014661e-05,
+      "loss": 0.9862,
+      "step": 10175
+    },
+    {
+      "epoch": 0.41197895588830435,
+      "grad_norm": 1.2750754356384277,
+      "learning_rate": 6.00041296716911e-05,
+      "loss": 0.9573,
+      "step": 10180
+    },
+    {
+      "epoch": 0.4121813031161473,
+      "grad_norm": 1.166333794593811,
+      "learning_rate": 5.9983481313235603e-05,
+      "loss": 1.013,
+      "step": 10185
+    },
+    {
+      "epoch": 0.41238365034399027,
+      "grad_norm": 1.0812666416168213,
+      "learning_rate": 5.99628329547801e-05,
+      "loss": 1.013,
+      "step": 10190
+    },
+    {
+      "epoch": 0.4125859975718333,
+      "grad_norm": 1.2474825382232666,
+      "learning_rate": 5.994218459632459e-05,
+      "loss": 0.9837,
+      "step": 10195
+    },
+    {
+      "epoch": 0.41278834479967624,
+      "grad_norm": 1.1481378078460693,
+      "learning_rate": 5.992153623786909e-05,
+      "loss": 1.0531,
+      "step": 10200
+    },
+    {
+      "epoch": 0.4129906920275192,
+      "grad_norm": 1.0672534704208374,
+      "learning_rate": 5.990088787941359e-05,
+      "loss": 1.0336,
+      "step": 10205
+    },
+    {
+      "epoch": 0.4131930392553622,
+      "grad_norm": 1.1827770471572876,
+      "learning_rate": 5.988023952095808e-05,
+      "loss": 1.0338,
+      "step": 10210
+    },
+    {
+      "epoch": 0.4133953864832052,
+      "grad_norm": 1.1003824472427368,
+      "learning_rate": 5.985959116250258e-05,
+      "loss": 1.0321,
+      "step": 10215
+    },
+    {
+      "epoch": 0.41359773371104813,
+      "grad_norm": 1.1213117837905884,
+      "learning_rate": 5.9838942804047085e-05,
+      "loss": 0.9712,
+      "step": 10220
+    },
+    {
+      "epoch": 0.41380008093889115,
+      "grad_norm": 1.1966711282730103,
+      "learning_rate": 5.981829444559158e-05,
+      "loss": 0.9826,
+      "step": 10225
+    },
+    {
+      "epoch": 0.4140024281667341,
+      "grad_norm": 1.1761008501052856,
+      "learning_rate": 5.979764608713607e-05,
+      "loss": 1.0222,
+      "step": 10230
+    },
+    {
+      "epoch": 0.4142047753945771,
+      "grad_norm": 1.2960001230239868,
+      "learning_rate": 5.977699772868057e-05,
+      "loss": 0.9271,
+      "step": 10235
+    },
+    {
+      "epoch": 0.4144071226224201,
+      "grad_norm": 1.204140067100525,
+      "learning_rate": 5.9756349370225074e-05,
+      "loss": 1.0036,
+      "step": 10240
+    },
+    {
+      "epoch": 0.41460946985026303,
+      "grad_norm": 1.1577142477035522,
+      "learning_rate": 5.9735701011769564e-05,
+      "loss": 0.986,
+      "step": 10245
+    },
+    {
+      "epoch": 0.41481181707810605,
+      "grad_norm": 1.135983943939209,
+      "learning_rate": 5.971505265331406e-05,
+      "loss": 1.0024,
+      "step": 10250
+    },
+    {
+      "epoch": 0.415014164305949,
+      "grad_norm": 1.3453749418258667,
+      "learning_rate": 5.9694404294858566e-05,
+      "loss": 0.9585,
+      "step": 10255
+    },
+    {
+      "epoch": 0.41521651153379197,
+      "grad_norm": 1.1529349088668823,
+      "learning_rate": 5.967375593640306e-05,
+      "loss": 1.0154,
+      "step": 10260
+    },
+    {
+      "epoch": 0.415418858761635,
+      "grad_norm": 1.1858090162277222,
+      "learning_rate": 5.9653107577947553e-05,
+      "loss": 0.9613,
+      "step": 10265
+    },
+    {
+      "epoch": 0.41562120598947794,
+      "grad_norm": 1.1481332778930664,
+      "learning_rate": 5.963245921949205e-05,
+      "loss": 0.9912,
+      "step": 10270
+    },
+    {
+      "epoch": 0.4158235532173209,
+      "grad_norm": 1.128801703453064,
+      "learning_rate": 5.9611810861036555e-05,
+      "loss": 1.0349,
+      "step": 10275
+    },
+    {
+      "epoch": 0.4160259004451639,
+      "grad_norm": 1.2165671586990356,
+      "learning_rate": 5.9591162502581045e-05,
+      "loss": 1.0492,
+      "step": 10280
+    },
+    {
+      "epoch": 0.41622824767300687,
+      "grad_norm": 1.1541047096252441,
+      "learning_rate": 5.957051414412554e-05,
+      "loss": 1.0514,
+      "step": 10285
+    },
+    {
+      "epoch": 0.4164305949008499,
+      "grad_norm": 1.4071108102798462,
+      "learning_rate": 5.954986578567005e-05,
+      "loss": 1.0151,
+      "step": 10290
+    },
+    {
+      "epoch": 0.41663294212869284,
+      "grad_norm": 1.2491838932037354,
+      "learning_rate": 5.952921742721454e-05,
+      "loss": 0.9952,
+      "step": 10295
+    },
+    {
+      "epoch": 0.4168352893565358,
+      "grad_norm": 1.1746826171875,
+      "learning_rate": 5.9508569068759034e-05,
+      "loss": 1.0023,
+      "step": 10300
+    },
+    {
+      "epoch": 0.4170376365843788,
+      "grad_norm": 1.1057684421539307,
+      "learning_rate": 5.948792071030354e-05,
+      "loss": 0.9433,
+      "step": 10305
+    },
+    {
+      "epoch": 0.41723998381222177,
+      "grad_norm": 1.190808892250061,
+      "learning_rate": 5.9467272351848036e-05,
+      "loss": 0.9958,
+      "step": 10310
+    },
+    {
+      "epoch": 0.41744233104006473,
+      "grad_norm": 1.114880084991455,
+      "learning_rate": 5.9446623993392526e-05,
+      "loss": 1.0478,
+      "step": 10315
+    },
+    {
+      "epoch": 0.41764467826790774,
+      "grad_norm": 1.089569091796875,
+      "learning_rate": 5.9425975634937024e-05,
+      "loss": 0.9614,
+      "step": 10320
+    },
+    {
+      "epoch": 0.4178470254957507,
+      "grad_norm": 1.2067276239395142,
+      "learning_rate": 5.940532727648153e-05,
+      "loss": 0.9489,
+      "step": 10325
+    },
+    {
+      "epoch": 0.41804937272359366,
+      "grad_norm": 1.0179661512374878,
+      "learning_rate": 5.938467891802602e-05,
+      "loss": 1.0104,
+      "step": 10330
+    },
+    {
+      "epoch": 0.4182517199514367,
+      "grad_norm": 1.0859757661819458,
+      "learning_rate": 5.9364030559570516e-05,
+      "loss": 1.0502,
+      "step": 10335
+    },
+    {
+      "epoch": 0.41845406717927963,
+      "grad_norm": 1.1184546947479248,
+      "learning_rate": 5.934338220111502e-05,
+      "loss": 1.0075,
+      "step": 10340
+    },
+    {
+      "epoch": 0.41865641440712265,
+      "grad_norm": 1.2337294816970825,
+      "learning_rate": 5.932273384265952e-05,
+      "loss": 1.0723,
+      "step": 10345
+    },
+    {
+      "epoch": 0.4188587616349656,
+      "grad_norm": 1.2576165199279785,
+      "learning_rate": 5.930208548420401e-05,
+      "loss": 0.8998,
+      "step": 10350
+    },
+    {
+      "epoch": 0.41906110886280856,
+      "grad_norm": 1.1309709548950195,
+      "learning_rate": 5.9281437125748505e-05,
+      "loss": 0.9974,
+      "step": 10355
+    },
+    {
+      "epoch": 0.4192634560906516,
+      "grad_norm": 1.310569167137146,
+      "learning_rate": 5.926078876729301e-05,
+      "loss": 0.9868,
+      "step": 10360
+    },
+    {
+      "epoch": 0.41946580331849453,
+      "grad_norm": 1.265221357345581,
+      "learning_rate": 5.92401404088375e-05,
+      "loss": 1.0637,
+      "step": 10365
+    },
+    {
+      "epoch": 0.4196681505463375,
+      "grad_norm": 1.1957981586456299,
+      "learning_rate": 5.9219492050382e-05,
+      "loss": 1.0066,
+      "step": 10370
+    },
+    {
+      "epoch": 0.4198704977741805,
+      "grad_norm": 1.2208542823791504,
+      "learning_rate": 5.91988436919265e-05,
+      "loss": 0.9734,
+      "step": 10375
+    },
+    {
+      "epoch": 0.42007284500202346,
+      "grad_norm": 1.2308872938156128,
+      "learning_rate": 5.9178195333470984e-05,
+      "loss": 1.0411,
+      "step": 10380
+    },
+    {
+      "epoch": 0.4202751922298664,
+      "grad_norm": 1.1346920728683472,
+      "learning_rate": 5.915754697501549e-05,
+      "loss": 1.0573,
+      "step": 10385
+    },
+    {
+      "epoch": 0.42047753945770944,
+      "grad_norm": 1.2495075464248657,
+      "learning_rate": 5.9136898616559986e-05,
+      "loss": 1.0205,
+      "step": 10390
+    },
+    {
+      "epoch": 0.4206798866855524,
+      "grad_norm": 1.196473479270935,
+      "learning_rate": 5.911625025810449e-05,
+      "loss": 1.0096,
+      "step": 10395
+    },
+    {
+      "epoch": 0.4208822339133954,
+      "grad_norm": 1.1124287843704224,
+      "learning_rate": 5.909560189964898e-05,
+      "loss": 0.9755,
+      "step": 10400
+    },
+    {
+      "epoch": 0.42108458114123837,
+      "grad_norm": 1.1813316345214844,
+      "learning_rate": 5.907495354119348e-05,
+      "loss": 1.0343,
+      "step": 10405
+    },
+    {
+      "epoch": 0.4212869283690813,
+      "grad_norm": 1.1222776174545288,
+      "learning_rate": 5.905430518273798e-05,
+      "loss": 1.0145,
+      "step": 10410
+    },
+    {
+      "epoch": 0.42148927559692434,
+      "grad_norm": 1.0855779647827148,
+      "learning_rate": 5.9033656824282466e-05,
+      "loss": 0.9858,
+      "step": 10415
+    },
+    {
+      "epoch": 0.4216916228247673,
+      "grad_norm": 1.155497670173645,
+      "learning_rate": 5.901300846582697e-05,
+      "loss": 1.0356,
+      "step": 10420
+    },
+    {
+      "epoch": 0.42189397005261026,
+      "grad_norm": 1.403380274772644,
+      "learning_rate": 5.899236010737147e-05,
+      "loss": 0.9808,
+      "step": 10425
+    },
+    {
+      "epoch": 0.42209631728045327,
+      "grad_norm": 1.2330858707427979,
+      "learning_rate": 5.897171174891596e-05,
+      "loss": 0.9357,
+      "step": 10430
+    },
+    {
+      "epoch": 0.42229866450829623,
+      "grad_norm": 1.1334590911865234,
+      "learning_rate": 5.895106339046046e-05,
+      "loss": 1.0152,
+      "step": 10435
+    },
+    {
+      "epoch": 0.42250101173613924,
+      "grad_norm": 1.1154812574386597,
+      "learning_rate": 5.893041503200496e-05,
+      "loss": 0.9547,
+      "step": 10440
+    },
+    {
+      "epoch": 0.4227033589639822,
+      "grad_norm": 1.1414520740509033,
+      "learning_rate": 5.890976667354946e-05,
+      "loss": 1.056,
+      "step": 10445
+    },
+    {
+      "epoch": 0.42290570619182516,
+      "grad_norm": 1.1283037662506104,
+      "learning_rate": 5.8889118315093947e-05,
+      "loss": 1.0158,
+      "step": 10450
+    },
+    {
+      "epoch": 0.4231080534196682,
+      "grad_norm": 1.2176074981689453,
+      "learning_rate": 5.886846995663845e-05,
+      "loss": 0.9993,
+      "step": 10455
+    },
+    {
+      "epoch": 0.42331040064751113,
+      "grad_norm": 1.1889210939407349,
+      "learning_rate": 5.884782159818295e-05,
+      "loss": 0.9266,
+      "step": 10460
+    },
+    {
+      "epoch": 0.4235127478753541,
+      "grad_norm": 1.1583353281021118,
+      "learning_rate": 5.882717323972744e-05,
+      "loss": 1.0195,
+      "step": 10465
+    },
+    {
+      "epoch": 0.4237150951031971,
+      "grad_norm": 1.1713775396347046,
+      "learning_rate": 5.880652488127194e-05,
+      "loss": 0.9766,
+      "step": 10470
+    },
+    {
+      "epoch": 0.42391744233104006,
+      "grad_norm": 1.212542176246643,
+      "learning_rate": 5.878587652281644e-05,
+      "loss": 0.9821,
+      "step": 10475
+    },
+    {
+      "epoch": 0.424119789558883,
+      "grad_norm": 1.2066254615783691,
+      "learning_rate": 5.8765228164360944e-05,
+      "loss": 1.0263,
+      "step": 10480
+    },
+    {
+      "epoch": 0.42432213678672603,
+      "grad_norm": 1.1281640529632568,
+      "learning_rate": 5.874457980590543e-05,
+      "loss": 0.9598,
+      "step": 10485
+    },
+    {
+      "epoch": 0.424524484014569,
+      "grad_norm": 1.1599007844924927,
+      "learning_rate": 5.872393144744993e-05,
+      "loss": 0.9789,
+      "step": 10490
+    },
+    {
+      "epoch": 0.424726831242412,
+      "grad_norm": 1.2747465372085571,
+      "learning_rate": 5.870328308899443e-05,
+      "loss": 0.9692,
+      "step": 10495
+    },
+    {
+      "epoch": 0.42492917847025496,
+      "grad_norm": 1.1403895616531372,
+      "learning_rate": 5.868263473053892e-05,
+      "loss": 0.9715,
+      "step": 10500
+    },
+    {
+      "epoch": 0.4251315256980979,
+      "grad_norm": 1.0249996185302734,
+      "learning_rate": 5.8661986372083424e-05,
+      "loss": 0.9788,
+      "step": 10505
+    },
+    {
+      "epoch": 0.42533387292594094,
+      "grad_norm": 1.1349996328353882,
+      "learning_rate": 5.864133801362792e-05,
+      "loss": 0.9761,
+      "step": 10510
+    },
+    {
+      "epoch": 0.4255362201537839,
+      "grad_norm": 1.0648891925811768,
+      "learning_rate": 5.862068965517241e-05,
+      "loss": 0.9968,
+      "step": 10515
+    },
+    {
+      "epoch": 0.42573856738162685,
+      "grad_norm": 1.2156158685684204,
+      "learning_rate": 5.860004129671691e-05,
+      "loss": 0.9648,
+      "step": 10520
+    },
+    {
+      "epoch": 0.42594091460946987,
+      "grad_norm": 1.1436049938201904,
+      "learning_rate": 5.857939293826141e-05,
+      "loss": 1.0048,
+      "step": 10525
+    },
+    {
+      "epoch": 0.4261432618373128,
+      "grad_norm": 1.1715660095214844,
+      "learning_rate": 5.855874457980591e-05,
+      "loss": 1.0497,
+      "step": 10530
+    },
+    {
+      "epoch": 0.4263456090651558,
+      "grad_norm": 1.1499030590057373,
+      "learning_rate": 5.85380962213504e-05,
+      "loss": 1.0035,
+      "step": 10535
+    },
+    {
+      "epoch": 0.4265479562929988,
+      "grad_norm": 1.1474660634994507,
+      "learning_rate": 5.8517447862894905e-05,
+      "loss": 1.0365,
+      "step": 10540
+    },
+    {
+      "epoch": 0.42675030352084176,
+      "grad_norm": 1.1669492721557617,
+      "learning_rate": 5.84967995044394e-05,
+      "loss": 0.9716,
+      "step": 10545
+    },
+    {
+      "epoch": 0.42695265074868477,
+      "grad_norm": 1.2073500156402588,
+      "learning_rate": 5.847615114598389e-05,
+      "loss": 0.9615,
+      "step": 10550
+    },
+    {
+      "epoch": 0.42715499797652773,
+      "grad_norm": 1.3077940940856934,
+      "learning_rate": 5.84555027875284e-05,
+      "loss": 1.0056,
+      "step": 10555
+    },
+    {
+      "epoch": 0.4273573452043707,
+      "grad_norm": 1.0633203983306885,
+      "learning_rate": 5.8434854429072894e-05,
+      "loss": 0.9749,
+      "step": 10560
+    },
+    {
+      "epoch": 0.4275596924322137,
+      "grad_norm": 1.124089241027832,
+      "learning_rate": 5.8414206070617384e-05,
+      "loss": 0.9614,
+      "step": 10565
+    },
+    {
+      "epoch": 0.42776203966005666,
+      "grad_norm": 1.088313102722168,
+      "learning_rate": 5.839355771216188e-05,
+      "loss": 0.9373,
+      "step": 10570
+    },
+    {
+      "epoch": 0.4279643868878996,
+      "grad_norm": 1.002710223197937,
+      "learning_rate": 5.8372909353706386e-05,
+      "loss": 1.0047,
+      "step": 10575
+    },
+    {
+      "epoch": 0.42816673411574263,
+      "grad_norm": 1.0879201889038086,
+      "learning_rate": 5.835226099525088e-05,
+      "loss": 0.9382,
+      "step": 10580
+    },
+    {
+      "epoch": 0.4283690813435856,
+      "grad_norm": 1.248063087463379,
+      "learning_rate": 5.8331612636795374e-05,
+      "loss": 1.0214,
+      "step": 10585
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 1.1118930578231812,
+      "learning_rate": 5.831096427833988e-05,
+      "loss": 1.0565,
+      "step": 10590
+    },
+    {
+      "epoch": 0.42877377579927156,
+      "grad_norm": 1.1189217567443848,
+      "learning_rate": 5.8290315919884375e-05,
+      "loss": 1.0402,
+      "step": 10595
+    },
+    {
+      "epoch": 0.4289761230271145,
+      "grad_norm": 1.099325180053711,
+      "learning_rate": 5.8269667561428866e-05,
+      "loss": 0.9871,
+      "step": 10600
+    },
+    {
+      "epoch": 0.42917847025495753,
+      "grad_norm": 1.2033805847167969,
+      "learning_rate": 5.824901920297336e-05,
+      "loss": 0.9529,
+      "step": 10605
+    },
+    {
+      "epoch": 0.4293808174828005,
+      "grad_norm": 0.9986255764961243,
+      "learning_rate": 5.822837084451787e-05,
+      "loss": 0.9639,
+      "step": 10610
+    },
+    {
+      "epoch": 0.42958316471064345,
+      "grad_norm": 1.23876953125,
+      "learning_rate": 5.8207722486062364e-05,
+      "loss": 1.0431,
+      "step": 10615
+    },
+    {
+      "epoch": 0.42978551193848646,
+      "grad_norm": 1.1878960132598877,
+      "learning_rate": 5.8187074127606855e-05,
+      "loss": 0.9407,
+      "step": 10620
+    },
+    {
+      "epoch": 0.4299878591663294,
+      "grad_norm": 1.0324996709823608,
+      "learning_rate": 5.816642576915136e-05,
+      "loss": 1.0141,
+      "step": 10625
+    },
+    {
+      "epoch": 0.4301902063941724,
+      "grad_norm": 1.0734326839447021,
+      "learning_rate": 5.8145777410695856e-05,
+      "loss": 1.0244,
+      "step": 10630
+    },
+    {
+      "epoch": 0.4303925536220154,
+      "grad_norm": 1.0938656330108643,
+      "learning_rate": 5.8125129052240347e-05,
+      "loss": 0.9731,
+      "step": 10635
+    },
+    {
+      "epoch": 0.43059490084985835,
+      "grad_norm": 1.1891802549362183,
+      "learning_rate": 5.8104480693784844e-05,
+      "loss": 0.9565,
+      "step": 10640
+    },
+    {
+      "epoch": 0.4307972480777013,
+      "grad_norm": 1.6427687406539917,
+      "learning_rate": 5.808383233532935e-05,
+      "loss": 1.015,
+      "step": 10645
+    },
+    {
+      "epoch": 0.4309995953055443,
+      "grad_norm": 1.1486482620239258,
+      "learning_rate": 5.806318397687384e-05,
+      "loss": 1.0105,
+      "step": 10650
+    },
+    {
+      "epoch": 0.4312019425333873,
+      "grad_norm": 1.1644834280014038,
+      "learning_rate": 5.8042535618418336e-05,
+      "loss": 1.0327,
+      "step": 10655
+    },
+    {
+      "epoch": 0.4314042897612303,
+      "grad_norm": 1.1761866807937622,
+      "learning_rate": 5.802188725996284e-05,
+      "loss": 0.9772,
+      "step": 10660
+    },
+    {
+      "epoch": 0.43160663698907326,
+      "grad_norm": 1.1902142763137817,
+      "learning_rate": 5.800123890150734e-05,
+      "loss": 0.9875,
+      "step": 10665
+    },
+    {
+      "epoch": 0.4318089842169162,
+      "grad_norm": 1.199822187423706,
+      "learning_rate": 5.798059054305183e-05,
+      "loss": 1.0068,
+      "step": 10670
+    },
+    {
+      "epoch": 0.43201133144475923,
+      "grad_norm": 1.155624508857727,
+      "learning_rate": 5.7959942184596325e-05,
+      "loss": 0.9842,
+      "step": 10675
+    },
+    {
+      "epoch": 0.4322136786726022,
+      "grad_norm": 1.1096001863479614,
+      "learning_rate": 5.793929382614083e-05,
+      "loss": 0.9737,
+      "step": 10680
+    },
+    {
+      "epoch": 0.43241602590044514,
+      "grad_norm": 1.1587976217269897,
+      "learning_rate": 5.791864546768532e-05,
+      "loss": 1.0342,
+      "step": 10685
+    },
+    {
+      "epoch": 0.43261837312828816,
+      "grad_norm": 1.1066583395004272,
+      "learning_rate": 5.789799710922982e-05,
+      "loss": 0.9886,
+      "step": 10690
+    },
+    {
+      "epoch": 0.4328207203561311,
+      "grad_norm": 1.148389458656311,
+      "learning_rate": 5.787734875077432e-05,
+      "loss": 0.986,
+      "step": 10695
+    },
+    {
+      "epoch": 0.4330230675839741,
+      "grad_norm": 1.2748706340789795,
+      "learning_rate": 5.785670039231882e-05,
+      "loss": 0.9912,
+      "step": 10700
+    },
+    {
+      "epoch": 0.4332254148118171,
+      "grad_norm": 1.3484047651290894,
+      "learning_rate": 5.783605203386331e-05,
+      "loss": 0.9831,
+      "step": 10705
+    },
+    {
+      "epoch": 0.43342776203966005,
+      "grad_norm": 1.1345332860946655,
+      "learning_rate": 5.7815403675407806e-05,
+      "loss": 0.9494,
+      "step": 10710
+    },
+    {
+      "epoch": 0.43363010926750306,
+      "grad_norm": 1.3133196830749512,
+      "learning_rate": 5.779475531695231e-05,
+      "loss": 1.0192,
+      "step": 10715
+    },
+    {
+      "epoch": 0.433832456495346,
+      "grad_norm": 1.258872389793396,
+      "learning_rate": 5.77741069584968e-05,
+      "loss": 1.0557,
+      "step": 10720
+    },
+    {
+      "epoch": 0.434034803723189,
+      "grad_norm": 1.1701213121414185,
+      "learning_rate": 5.77534586000413e-05,
+      "loss": 1.0009,
+      "step": 10725
+    },
+    {
+      "epoch": 0.434237150951032,
+      "grad_norm": 1.240444302558899,
+      "learning_rate": 5.77328102415858e-05,
+      "loss": 0.9558,
+      "step": 10730
+    },
+    {
+      "epoch": 0.43443949817887495,
+      "grad_norm": 1.263913631439209,
+      "learning_rate": 5.7712161883130286e-05,
+      "loss": 1.0595,
+      "step": 10735
+    },
+    {
+      "epoch": 0.4346418454067179,
+      "grad_norm": 1.1159262657165527,
+      "learning_rate": 5.769151352467479e-05,
+      "loss": 0.9617,
+      "step": 10740
+    },
+    {
+      "epoch": 0.4348441926345609,
+      "grad_norm": 1.154449701309204,
+      "learning_rate": 5.767086516621929e-05,
+      "loss": 1.0043,
+      "step": 10745
+    },
+    {
+      "epoch": 0.4350465398624039,
+      "grad_norm": 1.0637264251708984,
+      "learning_rate": 5.765021680776379e-05,
+      "loss": 1.0084,
+      "step": 10750
+    },
+    {
+      "epoch": 0.43524888709024684,
+      "grad_norm": 1.221408724784851,
+      "learning_rate": 5.762956844930828e-05,
+      "loss": 0.9945,
+      "step": 10755
+    },
+    {
+      "epoch": 0.43545123431808985,
+      "grad_norm": 1.239949345588684,
+      "learning_rate": 5.760892009085278e-05,
+      "loss": 0.964,
+      "step": 10760
+    },
+    {
+      "epoch": 0.4356535815459328,
+      "grad_norm": 1.1827601194381714,
+      "learning_rate": 5.758827173239728e-05,
+      "loss": 0.967,
+      "step": 10765
+    },
+    {
+      "epoch": 0.4358559287737758,
+      "grad_norm": 1.1999480724334717,
+      "learning_rate": 5.7567623373941774e-05,
+      "loss": 0.9989,
+      "step": 10770
+    },
+    {
+      "epoch": 0.4360582760016188,
+      "grad_norm": 1.1385078430175781,
+      "learning_rate": 5.754697501548627e-05,
+      "loss": 0.9738,
+      "step": 10775
+    },
+    {
+      "epoch": 0.43626062322946174,
+      "grad_norm": 1.2215017080307007,
+      "learning_rate": 5.752632665703077e-05,
+      "loss": 1.0081,
+      "step": 10780
+    },
+    {
+      "epoch": 0.43646297045730476,
+      "grad_norm": 1.2245571613311768,
+      "learning_rate": 5.750567829857526e-05,
+      "loss": 0.9952,
+      "step": 10785
+    },
+    {
+      "epoch": 0.4366653176851477,
+      "grad_norm": 1.1160204410552979,
+      "learning_rate": 5.748502994011976e-05,
+      "loss": 0.9938,
+      "step": 10790
+    },
+    {
+      "epoch": 0.43686766491299067,
+      "grad_norm": 1.2266266345977783,
+      "learning_rate": 5.746438158166426e-05,
+      "loss": 0.9662,
+      "step": 10795
+    },
+    {
+      "epoch": 0.4370700121408337,
+      "grad_norm": 1.1597180366516113,
+      "learning_rate": 5.7443733223208764e-05,
+      "loss": 0.9727,
+      "step": 10800
+    },
+    {
+      "epoch": 0.43727235936867664,
+      "grad_norm": 1.157707691192627,
+      "learning_rate": 5.7423084864753255e-05,
+      "loss": 0.9458,
+      "step": 10805
+    },
+    {
+      "epoch": 0.4374747065965196,
+      "grad_norm": 1.1504061222076416,
+      "learning_rate": 5.740243650629775e-05,
+      "loss": 0.9792,
+      "step": 10810
+    },
+    {
+      "epoch": 0.4376770538243626,
+      "grad_norm": 1.170724630355835,
+      "learning_rate": 5.738178814784225e-05,
+      "loss": 0.9569,
+      "step": 10815
+    },
+    {
+      "epoch": 0.4378794010522056,
+      "grad_norm": 1.3073015213012695,
+      "learning_rate": 5.736113978938674e-05,
+      "loss": 1.042,
+      "step": 10820
+    },
+    {
+      "epoch": 0.4380817482800486,
+      "grad_norm": 1.232712745666504,
+      "learning_rate": 5.7340491430931244e-05,
+      "loss": 0.9973,
+      "step": 10825
+    },
+    {
+      "epoch": 0.43828409550789155,
+      "grad_norm": 1.0662283897399902,
+      "learning_rate": 5.731984307247574e-05,
+      "loss": 0.9902,
+      "step": 10830
+    },
+    {
+      "epoch": 0.4384864427357345,
+      "grad_norm": 1.2456985712051392,
+      "learning_rate": 5.7299194714020245e-05,
+      "loss": 1.0268,
+      "step": 10835
+    },
+    {
+      "epoch": 0.4386887899635775,
+      "grad_norm": 1.0141834020614624,
+      "learning_rate": 5.7278546355564736e-05,
+      "loss": 1.0123,
+      "step": 10840
+    },
+    {
+      "epoch": 0.4388911371914205,
+      "grad_norm": 1.1344342231750488,
+      "learning_rate": 5.725789799710923e-05,
+      "loss": 1.0521,
+      "step": 10845
+    },
+    {
+      "epoch": 0.43909348441926344,
+      "grad_norm": 1.1500192880630493,
+      "learning_rate": 5.723724963865374e-05,
+      "loss": 1.001,
+      "step": 10850
+    },
+    {
+      "epoch": 0.43929583164710645,
+      "grad_norm": 1.23470139503479,
+      "learning_rate": 5.721660128019822e-05,
+      "loss": 1.0242,
+      "step": 10855
+    },
+    {
+      "epoch": 0.4394981788749494,
+      "grad_norm": 1.1283819675445557,
+      "learning_rate": 5.7195952921742725e-05,
+      "loss": 1.0173,
+      "step": 10860
+    },
+    {
+      "epoch": 0.43970052610279237,
+      "grad_norm": 1.0971150398254395,
+      "learning_rate": 5.717530456328722e-05,
+      "loss": 0.9823,
+      "step": 10865
+    },
+    {
+      "epoch": 0.4399028733306354,
+      "grad_norm": 1.1335411071777344,
+      "learning_rate": 5.715465620483171e-05,
+      "loss": 1.0396,
+      "step": 10870
+    },
+    {
+      "epoch": 0.44010522055847834,
+      "grad_norm": 1.16185462474823,
+      "learning_rate": 5.713400784637622e-05,
+      "loss": 1.0164,
+      "step": 10875
+    },
+    {
+      "epoch": 0.44030756778632135,
+      "grad_norm": 1.1602494716644287,
+      "learning_rate": 5.7113359487920714e-05,
+      "loss": 0.9838,
+      "step": 10880
+    },
+    {
+      "epoch": 0.4405099150141643,
+      "grad_norm": 1.1723899841308594,
+      "learning_rate": 5.709271112946522e-05,
+      "loss": 0.9555,
+      "step": 10885
+    },
+    {
+      "epoch": 0.44071226224200727,
+      "grad_norm": 1.0596345663070679,
+      "learning_rate": 5.70720627710097e-05,
+      "loss": 0.9953,
+      "step": 10890
+    },
+    {
+      "epoch": 0.4409146094698503,
+      "grad_norm": 1.4138137102127075,
+      "learning_rate": 5.7051414412554206e-05,
+      "loss": 1.0426,
+      "step": 10895
+    },
+    {
+      "epoch": 0.44111695669769324,
+      "grad_norm": 1.1126396656036377,
+      "learning_rate": 5.70307660540987e-05,
+      "loss": 1.0198,
+      "step": 10900
+    },
+    {
+      "epoch": 0.4413193039255362,
+      "grad_norm": 1.155994176864624,
+      "learning_rate": 5.7010117695643194e-05,
+      "loss": 1.002,
+      "step": 10905
+    },
+    {
+      "epoch": 0.4415216511533792,
+      "grad_norm": 1.209667682647705,
+      "learning_rate": 5.69894693371877e-05,
+      "loss": 0.978,
+      "step": 10910
+    },
+    {
+      "epoch": 0.44172399838122217,
+      "grad_norm": 1.2612943649291992,
+      "learning_rate": 5.6968820978732195e-05,
+      "loss": 0.9871,
+      "step": 10915
+    },
+    {
+      "epoch": 0.44192634560906513,
+      "grad_norm": 1.131173849105835,
+      "learning_rate": 5.6948172620276686e-05,
+      "loss": 0.9714,
+      "step": 10920
+    },
+    {
+      "epoch": 0.44212869283690814,
+      "grad_norm": 1.2085996866226196,
+      "learning_rate": 5.692752426182118e-05,
+      "loss": 1.0168,
+      "step": 10925
+    },
+    {
+      "epoch": 0.4423310400647511,
+      "grad_norm": 1.3031142950057983,
+      "learning_rate": 5.690687590336569e-05,
+      "loss": 0.9735,
+      "step": 10930
+    },
+    {
+      "epoch": 0.4425333872925941,
+      "grad_norm": 1.1572239398956299,
+      "learning_rate": 5.6886227544910184e-05,
+      "loss": 1.02,
+      "step": 10935
+    },
+    {
+      "epoch": 0.4427357345204371,
+      "grad_norm": 1.1638578176498413,
+      "learning_rate": 5.6865579186454675e-05,
+      "loss": 0.9327,
+      "step": 10940
+    },
+    {
+      "epoch": 0.44293808174828003,
+      "grad_norm": 1.2543002367019653,
+      "learning_rate": 5.684493082799918e-05,
+      "loss": 1.006,
+      "step": 10945
+    },
+    {
+      "epoch": 0.44314042897612305,
+      "grad_norm": 1.1135634183883667,
+      "learning_rate": 5.6824282469543676e-05,
+      "loss": 0.9771,
+      "step": 10950
+    },
+    {
+      "epoch": 0.443342776203966,
+      "grad_norm": 1.211946725845337,
+      "learning_rate": 5.680363411108817e-05,
+      "loss": 0.9941,
+      "step": 10955
+    },
+    {
+      "epoch": 0.44354512343180896,
+      "grad_norm": 1.198978304862976,
+      "learning_rate": 5.6782985752632664e-05,
+      "loss": 1.0382,
+      "step": 10960
+    },
+    {
+      "epoch": 0.443747470659652,
+      "grad_norm": 1.2260339260101318,
+      "learning_rate": 5.676233739417717e-05,
+      "loss": 1.0234,
+      "step": 10965
+    },
+    {
+      "epoch": 0.44394981788749494,
+      "grad_norm": 1.1562124490737915,
+      "learning_rate": 5.6741689035721665e-05,
+      "loss": 1.0022,
+      "step": 10970
+    },
+    {
+      "epoch": 0.4441521651153379,
+      "grad_norm": 1.0694836378097534,
+      "learning_rate": 5.6721040677266156e-05,
+      "loss": 0.9675,
+      "step": 10975
+    },
+    {
+      "epoch": 0.4443545123431809,
+      "grad_norm": 1.1761274337768555,
+      "learning_rate": 5.670039231881066e-05,
+      "loss": 1.0342,
+      "step": 10980
+    },
+    {
+      "epoch": 0.44455685957102387,
+      "grad_norm": 1.164414882659912,
+      "learning_rate": 5.667974396035516e-05,
+      "loss": 1.004,
+      "step": 10985
+    },
+    {
+      "epoch": 0.4447592067988669,
+      "grad_norm": 1.249759554862976,
+      "learning_rate": 5.665909560189965e-05,
+      "loss": 1.0289,
+      "step": 10990
+    },
+    {
+      "epoch": 0.44496155402670984,
+      "grad_norm": 1.251045823097229,
+      "learning_rate": 5.6638447243444145e-05,
+      "loss": 0.9858,
+      "step": 10995
+    },
+    {
+      "epoch": 0.4451639012545528,
+      "grad_norm": 1.205181360244751,
+      "learning_rate": 5.661779888498865e-05,
+      "loss": 0.9696,
+      "step": 11000
+    },
+    {
+      "epoch": 0.4453662484823958,
+      "grad_norm": 1.1822527647018433,
+      "learning_rate": 5.659715052653314e-05,
+      "loss": 0.9929,
+      "step": 11005
+    },
+    {
+      "epoch": 0.44556859571023877,
+      "grad_norm": 1.169823169708252,
+      "learning_rate": 5.657650216807764e-05,
+      "loss": 0.9968,
+      "step": 11010
+    },
+    {
+      "epoch": 0.4457709429380817,
+      "grad_norm": 1.0847971439361572,
+      "learning_rate": 5.655585380962214e-05,
+      "loss": 1.0832,
+      "step": 11015
+    },
+    {
+      "epoch": 0.44597329016592474,
+      "grad_norm": 1.2362971305847168,
+      "learning_rate": 5.653520545116664e-05,
+      "loss": 1.039,
+      "step": 11020
+    },
+    {
+      "epoch": 0.4461756373937677,
+      "grad_norm": 1.1471666097640991,
+      "learning_rate": 5.651455709271113e-05,
+      "loss": 1.0255,
+      "step": 11025
+    },
+    {
+      "epoch": 0.44637798462161066,
+      "grad_norm": 1.1853643655776978,
+      "learning_rate": 5.6493908734255626e-05,
+      "loss": 0.9831,
+      "step": 11030
+    },
+    {
+      "epoch": 0.44658033184945367,
+      "grad_norm": 1.054250955581665,
+      "learning_rate": 5.647326037580013e-05,
+      "loss": 1.0651,
+      "step": 11035
+    },
+    {
+      "epoch": 0.44678267907729663,
+      "grad_norm": 1.2345876693725586,
+      "learning_rate": 5.645261201734462e-05,
+      "loss": 0.9421,
+      "step": 11040
+    },
+    {
+      "epoch": 0.44698502630513964,
+      "grad_norm": 1.2633880376815796,
+      "learning_rate": 5.643196365888912e-05,
+      "loss": 0.9787,
+      "step": 11045
+    },
+    {
+      "epoch": 0.4471873735329826,
+      "grad_norm": 1.1371312141418457,
+      "learning_rate": 5.641131530043362e-05,
+      "loss": 0.942,
+      "step": 11050
+    },
+    {
+      "epoch": 0.44738972076082556,
+      "grad_norm": 1.1530300378799438,
+      "learning_rate": 5.639066694197812e-05,
+      "loss": 0.9845,
+      "step": 11055
+    },
+    {
+      "epoch": 0.4475920679886686,
+      "grad_norm": 1.1365545988082886,
+      "learning_rate": 5.637001858352261e-05,
+      "loss": 0.9124,
+      "step": 11060
+    },
+    {
+      "epoch": 0.44779441521651153,
+      "grad_norm": 1.3358913660049438,
+      "learning_rate": 5.634937022506711e-05,
+      "loss": 1.014,
+      "step": 11065
+    },
+    {
+      "epoch": 0.4479967624443545,
+      "grad_norm": 1.2181421518325806,
+      "learning_rate": 5.632872186661161e-05,
+      "loss": 1.0082,
+      "step": 11070
+    },
+    {
+      "epoch": 0.4481991096721975,
+      "grad_norm": 1.3012564182281494,
+      "learning_rate": 5.63080735081561e-05,
+      "loss": 0.9234,
+      "step": 11075
+    },
+    {
+      "epoch": 0.44840145690004046,
+      "grad_norm": 1.2210414409637451,
+      "learning_rate": 5.62874251497006e-05,
+      "loss": 1.0027,
+      "step": 11080
+    },
+    {
+      "epoch": 0.4486038041278834,
+      "grad_norm": 1.1770230531692505,
+      "learning_rate": 5.62667767912451e-05,
+      "loss": 0.9833,
+      "step": 11085
+    },
+    {
+      "epoch": 0.44880615135572643,
+      "grad_norm": 1.2124720811843872,
+      "learning_rate": 5.6246128432789594e-05,
+      "loss": 1.0064,
+      "step": 11090
+    },
+    {
+      "epoch": 0.4490084985835694,
+      "grad_norm": 1.2008227109909058,
+      "learning_rate": 5.622548007433409e-05,
+      "loss": 0.9921,
+      "step": 11095
+    },
+    {
+      "epoch": 0.4492108458114124,
+      "grad_norm": 1.0997034311294556,
+      "learning_rate": 5.6204831715878595e-05,
+      "loss": 1.0189,
+      "step": 11100
+    },
+    {
+      "epoch": 0.44941319303925537,
+      "grad_norm": 1.2778221368789673,
+      "learning_rate": 5.618418335742309e-05,
+      "loss": 0.9971,
+      "step": 11105
+    },
+    {
+      "epoch": 0.4496155402670983,
+      "grad_norm": 1.1807756423950195,
+      "learning_rate": 5.616353499896758e-05,
+      "loss": 0.9356,
+      "step": 11110
+    },
+    {
+      "epoch": 0.44981788749494134,
+      "grad_norm": 1.2341501712799072,
+      "learning_rate": 5.614288664051208e-05,
+      "loss": 1.0366,
+      "step": 11115
+    },
+    {
+      "epoch": 0.4500202347227843,
+      "grad_norm": 1.3213276863098145,
+      "learning_rate": 5.6122238282056584e-05,
+      "loss": 1.0088,
+      "step": 11120
+    },
+    {
+      "epoch": 0.45022258195062725,
+      "grad_norm": 1.3128165006637573,
+      "learning_rate": 5.6101589923601075e-05,
+      "loss": 1.04,
+      "step": 11125
+    },
+    {
+      "epoch": 0.45042492917847027,
+      "grad_norm": 1.2558141946792603,
+      "learning_rate": 5.608094156514557e-05,
+      "loss": 0.9987,
+      "step": 11130
+    },
+    {
+      "epoch": 0.4506272764063132,
+      "grad_norm": 1.163042426109314,
+      "learning_rate": 5.6060293206690076e-05,
+      "loss": 0.9675,
+      "step": 11135
+    },
+    {
+      "epoch": 0.4508296236341562,
+      "grad_norm": 1.1548547744750977,
+      "learning_rate": 5.603964484823456e-05,
+      "loss": 0.9765,
+      "step": 11140
+    },
+    {
+      "epoch": 0.4510319708619992,
+      "grad_norm": 1.1608883142471313,
+      "learning_rate": 5.6018996489779064e-05,
+      "loss": 0.9643,
+      "step": 11145
+    },
+    {
+      "epoch": 0.45123431808984216,
+      "grad_norm": 1.1438285112380981,
+      "learning_rate": 5.599834813132356e-05,
+      "loss": 0.9694,
+      "step": 11150
+    },
+    {
+      "epoch": 0.45143666531768517,
+      "grad_norm": 1.365047812461853,
+      "learning_rate": 5.5977699772868065e-05,
+      "loss": 1.0216,
+      "step": 11155
+    },
+    {
+      "epoch": 0.45163901254552813,
+      "grad_norm": 1.269562005996704,
+      "learning_rate": 5.5957051414412556e-05,
+      "loss": 1.0335,
+      "step": 11160
+    },
+    {
+      "epoch": 0.4518413597733711,
+      "grad_norm": 1.1969783306121826,
+      "learning_rate": 5.593640305595705e-05,
+      "loss": 1.0172,
+      "step": 11165
+    },
+    {
+      "epoch": 0.4520437070012141,
+      "grad_norm": 1.3603763580322266,
+      "learning_rate": 5.591575469750156e-05,
+      "loss": 0.9595,
+      "step": 11170
+    },
+    {
+      "epoch": 0.45224605422905706,
+      "grad_norm": 1.3282946348190308,
+      "learning_rate": 5.589510633904604e-05,
+      "loss": 0.9697,
+      "step": 11175
+    },
+    {
+      "epoch": 0.4524484014569,
+      "grad_norm": 0.9992858171463013,
+      "learning_rate": 5.5874457980590545e-05,
+      "loss": 0.9657,
+      "step": 11180
+    },
+    {
+      "epoch": 0.45265074868474303,
+      "grad_norm": 1.1385672092437744,
+      "learning_rate": 5.585380962213504e-05,
+      "loss": 0.9614,
+      "step": 11185
+    },
+    {
+      "epoch": 0.452853095912586,
+      "grad_norm": 1.3753652572631836,
+      "learning_rate": 5.5833161263679547e-05,
+      "loss": 0.8998,
+      "step": 11190
+    },
+    {
+      "epoch": 0.45305544314042895,
+      "grad_norm": 1.194413423538208,
+      "learning_rate": 5.581251290522404e-05,
+      "loss": 0.986,
+      "step": 11195
+    },
+    {
+      "epoch": 0.45325779036827196,
+      "grad_norm": 1.1120665073394775,
+      "learning_rate": 5.5791864546768534e-05,
+      "loss": 0.999,
+      "step": 11200
+    },
+    {
+      "epoch": 0.4534601375961149,
+      "grad_norm": 1.3044672012329102,
+      "learning_rate": 5.577121618831304e-05,
+      "loss": 0.989,
+      "step": 11205
+    },
+    {
+      "epoch": 0.45366248482395793,
+      "grad_norm": 1.0678913593292236,
+      "learning_rate": 5.575056782985752e-05,
+      "loss": 0.9563,
+      "step": 11210
+    },
+    {
+      "epoch": 0.4538648320518009,
+      "grad_norm": 1.3210352659225464,
+      "learning_rate": 5.5729919471402026e-05,
+      "loss": 0.9694,
+      "step": 11215
+    },
+    {
+      "epoch": 0.45406717927964385,
+      "grad_norm": 1.2733067274093628,
+      "learning_rate": 5.5709271112946524e-05,
+      "loss": 1.0371,
+      "step": 11220
+    },
+    {
+      "epoch": 0.45426952650748686,
+      "grad_norm": 1.194913387298584,
+      "learning_rate": 5.5688622754491014e-05,
+      "loss": 1.0445,
+      "step": 11225
+    },
+    {
+      "epoch": 0.4544718737353298,
+      "grad_norm": 1.1859469413757324,
+      "learning_rate": 5.566797439603552e-05,
+      "loss": 1.0561,
+      "step": 11230
+    },
+    {
+      "epoch": 0.4546742209631728,
+      "grad_norm": 1.1227186918258667,
+      "learning_rate": 5.5647326037580015e-05,
+      "loss": 1.0166,
+      "step": 11235
+    },
+    {
+      "epoch": 0.4548765681910158,
+      "grad_norm": 1.1340081691741943,
+      "learning_rate": 5.562667767912452e-05,
+      "loss": 0.952,
+      "step": 11240
+    },
+    {
+      "epoch": 0.45507891541885875,
+      "grad_norm": 1.1835229396820068,
+      "learning_rate": 5.5606029320669e-05,
+      "loss": 1.005,
+      "step": 11245
+    },
+    {
+      "epoch": 0.45528126264670177,
+      "grad_norm": 1.2764209508895874,
+      "learning_rate": 5.558538096221351e-05,
+      "loss": 0.9819,
+      "step": 11250
+    },
+    {
+      "epoch": 0.4554836098745447,
+      "grad_norm": 1.1682292222976685,
+      "learning_rate": 5.5564732603758005e-05,
+      "loss": 0.8616,
+      "step": 11255
+    },
+    {
+      "epoch": 0.4556859571023877,
+      "grad_norm": 1.1520472764968872,
+      "learning_rate": 5.5544084245302495e-05,
+      "loss": 1.045,
+      "step": 11260
+    },
+    {
+      "epoch": 0.4558883043302307,
+      "grad_norm": 1.1778866052627563,
+      "learning_rate": 5.5523435886847e-05,
+      "loss": 0.9874,
+      "step": 11265
+    },
+    {
+      "epoch": 0.45609065155807366,
+      "grad_norm": 1.0828711986541748,
+      "learning_rate": 5.5502787528391497e-05,
+      "loss": 0.9994,
+      "step": 11270
+    },
+    {
+      "epoch": 0.4562929987859166,
+      "grad_norm": 1.2597899436950684,
+      "learning_rate": 5.5482139169936e-05,
+      "loss": 1.0214,
+      "step": 11275
+    },
+    {
+      "epoch": 0.45649534601375963,
+      "grad_norm": 1.1565966606140137,
+      "learning_rate": 5.5461490811480484e-05,
+      "loss": 1.0158,
+      "step": 11280
+    },
+    {
+      "epoch": 0.4566976932416026,
+      "grad_norm": 1.20098876953125,
+      "learning_rate": 5.544084245302499e-05,
+      "loss": 0.9437,
+      "step": 11285
+    },
+    {
+      "epoch": 0.45690004046944555,
+      "grad_norm": 1.16834557056427,
+      "learning_rate": 5.5420194094569486e-05,
+      "loss": 0.9824,
+      "step": 11290
+    },
+    {
+      "epoch": 0.45710238769728856,
+      "grad_norm": 1.1236181259155273,
+      "learning_rate": 5.5399545736113976e-05,
+      "loss": 0.9707,
+      "step": 11295
+    },
+    {
+      "epoch": 0.4573047349251315,
+      "grad_norm": 1.1530249118804932,
+      "learning_rate": 5.537889737765848e-05,
+      "loss": 1.0455,
+      "step": 11300
+    },
+    {
+      "epoch": 0.45750708215297453,
+      "grad_norm": 1.2669841051101685,
+      "learning_rate": 5.535824901920298e-05,
+      "loss": 1.018,
+      "step": 11305
+    },
+    {
+      "epoch": 0.4577094293808175,
+      "grad_norm": 1.2789355516433716,
+      "learning_rate": 5.533760066074747e-05,
+      "loss": 0.9761,
+      "step": 11310
+    },
+    {
+      "epoch": 0.45791177660866045,
+      "grad_norm": 1.2220227718353271,
+      "learning_rate": 5.531695230229197e-05,
+      "loss": 1.0403,
+      "step": 11315
+    },
+    {
+      "epoch": 0.45811412383650346,
+      "grad_norm": 1.1159273386001587,
+      "learning_rate": 5.529630394383647e-05,
+      "loss": 0.9661,
+      "step": 11320
+    },
+    {
+      "epoch": 0.4583164710643464,
+      "grad_norm": 1.0662544965744019,
+      "learning_rate": 5.527565558538097e-05,
+      "loss": 1.0306,
+      "step": 11325
+    },
+    {
+      "epoch": 0.4585188182921894,
+      "grad_norm": 1.3621026277542114,
+      "learning_rate": 5.525500722692546e-05,
+      "loss": 1.0545,
+      "step": 11330
+    },
+    {
+      "epoch": 0.4587211655200324,
+      "grad_norm": 1.242497205734253,
+      "learning_rate": 5.523435886846996e-05,
+      "loss": 0.998,
+      "step": 11335
+    },
+    {
+      "epoch": 0.45892351274787535,
+      "grad_norm": 1.1795145273208618,
+      "learning_rate": 5.521371051001446e-05,
+      "loss": 1.023,
+      "step": 11340
+    },
+    {
+      "epoch": 0.4591258599757183,
+      "grad_norm": 1.172516942024231,
+      "learning_rate": 5.519306215155895e-05,
+      "loss": 1.0403,
+      "step": 11345
+    },
+    {
+      "epoch": 0.4593282072035613,
+      "grad_norm": 1.035287857055664,
+      "learning_rate": 5.517241379310345e-05,
+      "loss": 1.0572,
+      "step": 11350
+    },
+    {
+      "epoch": 0.4595305544314043,
+      "grad_norm": 1.2234196662902832,
+      "learning_rate": 5.515176543464795e-05,
+      "loss": 0.9477,
+      "step": 11355
+    },
+    {
+      "epoch": 0.4597329016592473,
+      "grad_norm": 1.054702639579773,
+      "learning_rate": 5.513111707619244e-05,
+      "loss": 0.9754,
+      "step": 11360
+    },
+    {
+      "epoch": 0.45993524888709025,
+      "grad_norm": 1.1466641426086426,
+      "learning_rate": 5.511046871773694e-05,
+      "loss": 1.0138,
+      "step": 11365
+    },
+    {
+      "epoch": 0.4601375961149332,
+      "grad_norm": 1.1666395664215088,
+      "learning_rate": 5.508982035928144e-05,
+      "loss": 0.9376,
+      "step": 11370
+    },
+    {
+      "epoch": 0.4603399433427762,
+      "grad_norm": 1.1848865747451782,
+      "learning_rate": 5.506917200082594e-05,
+      "loss": 1.0463,
+      "step": 11375
+    },
+    {
+      "epoch": 0.4605422905706192,
+      "grad_norm": 1.2855501174926758,
+      "learning_rate": 5.504852364237043e-05,
+      "loss": 0.9538,
+      "step": 11380
+    },
+    {
+      "epoch": 0.46074463779846214,
+      "grad_norm": 1.3183473348617554,
+      "learning_rate": 5.5027875283914934e-05,
+      "loss": 0.9932,
+      "step": 11385
+    },
+    {
+      "epoch": 0.46094698502630516,
+      "grad_norm": 1.1759259700775146,
+      "learning_rate": 5.500722692545943e-05,
+      "loss": 1.0433,
+      "step": 11390
+    },
+    {
+      "epoch": 0.4611493322541481,
+      "grad_norm": 1.193129539489746,
+      "learning_rate": 5.498657856700392e-05,
+      "loss": 0.9335,
+      "step": 11395
+    },
+    {
+      "epoch": 0.4613516794819911,
+      "grad_norm": 1.177869439125061,
+      "learning_rate": 5.496593020854842e-05,
+      "loss": 1.0172,
+      "step": 11400
+    },
+    {
+      "epoch": 0.4615540267098341,
+      "grad_norm": 1.1561225652694702,
+      "learning_rate": 5.4945281850092924e-05,
+      "loss": 1.0128,
+      "step": 11405
+    },
+    {
+      "epoch": 0.46175637393767704,
+      "grad_norm": 1.1183866262435913,
+      "learning_rate": 5.492463349163742e-05,
+      "loss": 1.0491,
+      "step": 11410
+    },
+    {
+      "epoch": 0.46195872116552006,
+      "grad_norm": 1.114930510520935,
+      "learning_rate": 5.490398513318191e-05,
+      "loss": 0.9707,
+      "step": 11415
+    },
+    {
+      "epoch": 0.462161068393363,
+      "grad_norm": 1.1722952127456665,
+      "learning_rate": 5.4883336774726415e-05,
+      "loss": 0.9784,
+      "step": 11420
+    },
+    {
+      "epoch": 0.462363415621206,
+      "grad_norm": 1.2154377698898315,
+      "learning_rate": 5.486268841627091e-05,
+      "loss": 1.0743,
+      "step": 11425
+    },
+    {
+      "epoch": 0.462565762849049,
+      "grad_norm": 1.1334062814712524,
+      "learning_rate": 5.48420400578154e-05,
+      "loss": 0.9652,
+      "step": 11430
+    },
+    {
+      "epoch": 0.46276811007689195,
+      "grad_norm": 1.1201444864273071,
+      "learning_rate": 5.48213916993599e-05,
+      "loss": 0.9014,
+      "step": 11435
+    },
+    {
+      "epoch": 0.4629704573047349,
+      "grad_norm": 1.1657004356384277,
+      "learning_rate": 5.4800743340904405e-05,
+      "loss": 1.0671,
+      "step": 11440
+    },
+    {
+      "epoch": 0.4631728045325779,
+      "grad_norm": 1.132319450378418,
+      "learning_rate": 5.4780094982448895e-05,
+      "loss": 0.9902,
+      "step": 11445
+    },
+    {
+      "epoch": 0.4633751517604209,
+      "grad_norm": 1.185171127319336,
+      "learning_rate": 5.475944662399339e-05,
+      "loss": 1.0329,
+      "step": 11450
+    },
+    {
+      "epoch": 0.46357749898826384,
+      "grad_norm": 1.1581697463989258,
+      "learning_rate": 5.4738798265537896e-05,
+      "loss": 1.0316,
+      "step": 11455
+    },
+    {
+      "epoch": 0.46377984621610685,
+      "grad_norm": 1.1859978437423706,
+      "learning_rate": 5.4718149907082394e-05,
+      "loss": 1.0021,
+      "step": 11460
+    },
+    {
+      "epoch": 0.4639821934439498,
+      "grad_norm": 3.962794780731201,
+      "learning_rate": 5.4697501548626884e-05,
+      "loss": 0.9965,
+      "step": 11465
+    },
+    {
+      "epoch": 0.4641845406717928,
+      "grad_norm": 1.3404104709625244,
+      "learning_rate": 5.467685319017138e-05,
+      "loss": 0.9506,
+      "step": 11470
+    },
+    {
+      "epoch": 0.4643868878996358,
+      "grad_norm": 1.3747602701187134,
+      "learning_rate": 5.4656204831715886e-05,
+      "loss": 0.9788,
+      "step": 11475
+    },
+    {
+      "epoch": 0.46458923512747874,
+      "grad_norm": 1.3079259395599365,
+      "learning_rate": 5.4635556473260376e-05,
+      "loss": 1.0819,
+      "step": 11480
+    },
+    {
+      "epoch": 0.46479158235532175,
+      "grad_norm": 1.1761444807052612,
+      "learning_rate": 5.4614908114804873e-05,
+      "loss": 0.9534,
+      "step": 11485
+    },
+    {
+      "epoch": 0.4649939295831647,
+      "grad_norm": 1.1462078094482422,
+      "learning_rate": 5.459425975634938e-05,
+      "loss": 1.0265,
+      "step": 11490
+    },
+    {
+      "epoch": 0.46519627681100767,
+      "grad_norm": 1.1891603469848633,
+      "learning_rate": 5.457361139789386e-05,
+      "loss": 1.0158,
+      "step": 11495
+    },
+    {
+      "epoch": 0.4653986240388507,
+      "grad_norm": 1.1987360715866089,
+      "learning_rate": 5.4552963039438365e-05,
+      "loss": 0.9456,
+      "step": 11500
+    },
+    {
+      "epoch": 0.46560097126669364,
+      "grad_norm": 1.1482681035995483,
+      "learning_rate": 5.453231468098286e-05,
+      "loss": 0.9909,
+      "step": 11505
+    },
+    {
+      "epoch": 0.4658033184945366,
+      "grad_norm": 1.2518885135650635,
+      "learning_rate": 5.451166632252737e-05,
+      "loss": 1.0183,
+      "step": 11510
+    },
+    {
+      "epoch": 0.4660056657223796,
+      "grad_norm": 1.112511157989502,
+      "learning_rate": 5.449101796407186e-05,
+      "loss": 1.0381,
+      "step": 11515
+    },
+    {
+      "epoch": 0.4662080129502226,
+      "grad_norm": 1.1407140493392944,
+      "learning_rate": 5.4470369605616355e-05,
+      "loss": 0.9752,
+      "step": 11520
+    },
+    {
+      "epoch": 0.4664103601780656,
+      "grad_norm": 1.2545710802078247,
+      "learning_rate": 5.444972124716086e-05,
+      "loss": 1.0329,
+      "step": 11525
+    },
+    {
+      "epoch": 0.46661270740590854,
+      "grad_norm": 1.160461187362671,
+      "learning_rate": 5.442907288870535e-05,
+      "loss": 0.9949,
+      "step": 11530
+    },
+    {
+      "epoch": 0.4668150546337515,
+      "grad_norm": 1.1197930574417114,
+      "learning_rate": 5.4408424530249846e-05,
+      "loss": 1.017,
+      "step": 11535
+    },
+    {
+      "epoch": 0.4670174018615945,
+      "grad_norm": 1.1824170351028442,
+      "learning_rate": 5.4387776171794344e-05,
+      "loss": 0.9759,
+      "step": 11540
+    },
+    {
+      "epoch": 0.4672197490894375,
+      "grad_norm": 1.122239589691162,
+      "learning_rate": 5.436712781333885e-05,
+      "loss": 1.0082,
+      "step": 11545
+    },
+    {
+      "epoch": 0.46742209631728043,
+      "grad_norm": 1.1205852031707764,
+      "learning_rate": 5.434647945488334e-05,
+      "loss": 1.0055,
+      "step": 11550
+    },
+    {
+      "epoch": 0.46762444354512345,
+      "grad_norm": 1.1242936849594116,
+      "learning_rate": 5.4325831096427836e-05,
+      "loss": 0.9787,
+      "step": 11555
+    },
+    {
+      "epoch": 0.4678267907729664,
+      "grad_norm": 1.4059381484985352,
+      "learning_rate": 5.430518273797234e-05,
+      "loss": 0.9984,
+      "step": 11560
+    },
+    {
+      "epoch": 0.46802913800080936,
+      "grad_norm": 1.2341316938400269,
+      "learning_rate": 5.428453437951683e-05,
+      "loss": 1.0389,
+      "step": 11565
+    },
+    {
+      "epoch": 0.4682314852286524,
+      "grad_norm": 1.228027105331421,
+      "learning_rate": 5.426388602106133e-05,
+      "loss": 1.0147,
+      "step": 11570
+    },
+    {
+      "epoch": 0.46843383245649534,
+      "grad_norm": 1.076759934425354,
+      "learning_rate": 5.4243237662605825e-05,
+      "loss": 1.0026,
+      "step": 11575
+    },
+    {
+      "epoch": 0.46863617968433835,
+      "grad_norm": 1.3640527725219727,
+      "learning_rate": 5.4222589304150315e-05,
+      "loss": 1.0227,
+      "step": 11580
+    },
+    {
+      "epoch": 0.4688385269121813,
+      "grad_norm": 1.1773855686187744,
+      "learning_rate": 5.420194094569482e-05,
+      "loss": 0.991,
+      "step": 11585
+    },
+    {
+      "epoch": 0.46904087414002427,
+      "grad_norm": 1.178702473640442,
+      "learning_rate": 5.418129258723932e-05,
+      "loss": 0.9894,
+      "step": 11590
+    },
+    {
+      "epoch": 0.4692432213678673,
+      "grad_norm": 1.263631820678711,
+      "learning_rate": 5.416064422878382e-05,
+      "loss": 1.0234,
+      "step": 11595
+    },
+    {
+      "epoch": 0.46944556859571024,
+      "grad_norm": 1.1302398443222046,
+      "learning_rate": 5.413999587032831e-05,
+      "loss": 1.0389,
+      "step": 11600
+    },
+    {
+      "epoch": 0.4696479158235532,
+      "grad_norm": 1.1499837636947632,
+      "learning_rate": 5.411934751187281e-05,
+      "loss": 1.0284,
+      "step": 11605
+    },
+    {
+      "epoch": 0.4698502630513962,
+      "grad_norm": 1.2414751052856445,
+      "learning_rate": 5.409869915341731e-05,
+      "loss": 0.9275,
+      "step": 11610
+    },
+    {
+      "epoch": 0.47005261027923917,
+      "grad_norm": 1.13211190700531,
+      "learning_rate": 5.4078050794961796e-05,
+      "loss": 0.9419,
+      "step": 11615
+    },
+    {
+      "epoch": 0.4702549575070821,
+      "grad_norm": 1.1615794897079468,
+      "learning_rate": 5.40574024365063e-05,
+      "loss": 0.9729,
+      "step": 11620
+    },
+    {
+      "epoch": 0.47045730473492514,
+      "grad_norm": 1.194742202758789,
+      "learning_rate": 5.40367540780508e-05,
+      "loss": 1.0409,
+      "step": 11625
+    },
+    {
+      "epoch": 0.4706596519627681,
+      "grad_norm": 1.2788242101669312,
+      "learning_rate": 5.40161057195953e-05,
+      "loss": 1.0719,
+      "step": 11630
+    },
+    {
+      "epoch": 0.4708619991906111,
+      "grad_norm": 1.361441731452942,
+      "learning_rate": 5.399545736113979e-05,
+      "loss": 0.9966,
+      "step": 11635
+    },
+    {
+      "epoch": 0.47106434641845407,
+      "grad_norm": 1.1751059293746948,
+      "learning_rate": 5.397480900268429e-05,
+      "loss": 1.053,
+      "step": 11640
+    },
+    {
+      "epoch": 0.47126669364629703,
+      "grad_norm": 1.1899535655975342,
+      "learning_rate": 5.3954160644228794e-05,
+      "loss": 1.0003,
+      "step": 11645
+    },
+    {
+      "epoch": 0.47146904087414004,
+      "grad_norm": 1.1658504009246826,
+      "learning_rate": 5.393351228577328e-05,
+      "loss": 1.0052,
+      "step": 11650
+    },
+    {
+      "epoch": 0.471671388101983,
+      "grad_norm": 1.126866102218628,
+      "learning_rate": 5.391286392731778e-05,
+      "loss": 0.9929,
+      "step": 11655
+    },
+    {
+      "epoch": 0.47187373532982596,
+      "grad_norm": 1.2745146751403809,
+      "learning_rate": 5.389221556886228e-05,
+      "loss": 1.0301,
+      "step": 11660
+    },
+    {
+      "epoch": 0.472076082557669,
+      "grad_norm": 1.0762321949005127,
+      "learning_rate": 5.387156721040677e-05,
+      "loss": 1.0129,
+      "step": 11665
+    },
+    {
+      "epoch": 0.47227842978551193,
+      "grad_norm": 1.1167628765106201,
+      "learning_rate": 5.3850918851951273e-05,
+      "loss": 0.9681,
+      "step": 11670
+    },
+    {
+      "epoch": 0.4724807770133549,
+      "grad_norm": 1.1747218370437622,
+      "learning_rate": 5.383027049349577e-05,
+      "loss": 1.0471,
+      "step": 11675
+    },
+    {
+      "epoch": 0.4726831242411979,
+      "grad_norm": 1.3087221384048462,
+      "learning_rate": 5.3809622135040275e-05,
+      "loss": 1.0013,
+      "step": 11680
+    },
+    {
+      "epoch": 0.47288547146904086,
+      "grad_norm": 1.1940945386886597,
+      "learning_rate": 5.378897377658476e-05,
+      "loss": 0.9774,
+      "step": 11685
+    },
+    {
+      "epoch": 0.4730878186968839,
+      "grad_norm": 1.164388656616211,
+      "learning_rate": 5.376832541812926e-05,
+      "loss": 1.0504,
+      "step": 11690
+    },
+    {
+      "epoch": 0.47329016592472684,
+      "grad_norm": 1.1465179920196533,
+      "learning_rate": 5.374767705967376e-05,
+      "loss": 1.0383,
+      "step": 11695
+    },
+    {
+      "epoch": 0.4734925131525698,
+      "grad_norm": 1.2491196393966675,
+      "learning_rate": 5.372702870121825e-05,
+      "loss": 0.9987,
+      "step": 11700
+    },
+    {
+      "epoch": 0.4736948603804128,
+      "grad_norm": 1.29067862033844,
+      "learning_rate": 5.3706380342762755e-05,
+      "loss": 1.0046,
+      "step": 11705
+    },
+    {
+      "epoch": 0.47389720760825577,
+      "grad_norm": 1.1442447900772095,
+      "learning_rate": 5.368573198430725e-05,
+      "loss": 0.9822,
+      "step": 11710
+    },
+    {
+      "epoch": 0.4740995548360987,
+      "grad_norm": 1.1371489763259888,
+      "learning_rate": 5.366508362585174e-05,
+      "loss": 0.9596,
+      "step": 11715
+    },
+    {
+      "epoch": 0.47430190206394174,
+      "grad_norm": 1.167847752571106,
+      "learning_rate": 5.364443526739624e-05,
+      "loss": 1.0093,
+      "step": 11720
+    },
+    {
+      "epoch": 0.4745042492917847,
+      "grad_norm": 1.2384512424468994,
+      "learning_rate": 5.3623786908940744e-05,
+      "loss": 1.0428,
+      "step": 11725
+    },
+    {
+      "epoch": 0.47470659651962765,
+      "grad_norm": 1.1138396263122559,
+      "learning_rate": 5.360313855048524e-05,
+      "loss": 0.9754,
+      "step": 11730
+    },
+    {
+      "epoch": 0.47490894374747067,
+      "grad_norm": 1.1507554054260254,
+      "learning_rate": 5.358249019202973e-05,
+      "loss": 0.9751,
+      "step": 11735
+    },
+    {
+      "epoch": 0.4751112909753136,
+      "grad_norm": 1.2503911256790161,
+      "learning_rate": 5.3561841833574236e-05,
+      "loss": 0.9987,
+      "step": 11740
+    },
+    {
+      "epoch": 0.47531363820315664,
+      "grad_norm": 1.0976605415344238,
+      "learning_rate": 5.354119347511873e-05,
+      "loss": 1.0059,
+      "step": 11745
+    },
+    {
+      "epoch": 0.4755159854309996,
+      "grad_norm": 1.3331193923950195,
+      "learning_rate": 5.3520545116663223e-05,
+      "loss": 0.985,
+      "step": 11750
+    },
+    {
+      "epoch": 0.47571833265884256,
+      "grad_norm": 1.1028928756713867,
+      "learning_rate": 5.349989675820772e-05,
+      "loss": 1.0103,
+      "step": 11755
+    },
+    {
+      "epoch": 0.47592067988668557,
+      "grad_norm": 1.3108789920806885,
+      "learning_rate": 5.3479248399752225e-05,
+      "loss": 1.0209,
+      "step": 11760
+    },
+    {
+      "epoch": 0.47612302711452853,
+      "grad_norm": 1.1938402652740479,
+      "learning_rate": 5.345860004129672e-05,
+      "loss": 0.9593,
+      "step": 11765
+    },
+    {
+      "epoch": 0.4763253743423715,
+      "grad_norm": 1.2749603986740112,
+      "learning_rate": 5.343795168284121e-05,
+      "loss": 1.0241,
+      "step": 11770
+    },
+    {
+      "epoch": 0.4765277215702145,
+      "grad_norm": 1.2287843227386475,
+      "learning_rate": 5.341730332438572e-05,
+      "loss": 1.013,
+      "step": 11775
+    },
+    {
+      "epoch": 0.47673006879805746,
+      "grad_norm": 1.2256628274917603,
+      "learning_rate": 5.3396654965930214e-05,
+      "loss": 1.0203,
+      "step": 11780
+    },
+    {
+      "epoch": 0.4769324160259004,
+      "grad_norm": 1.266689658164978,
+      "learning_rate": 5.3376006607474705e-05,
+      "loss": 0.988,
+      "step": 11785
+    },
+    {
+      "epoch": 0.47713476325374343,
+      "grad_norm": 1.2061651945114136,
+      "learning_rate": 5.33553582490192e-05,
+      "loss": 1.0481,
+      "step": 11790
+    },
+    {
+      "epoch": 0.4773371104815864,
+      "grad_norm": 1.203644037246704,
+      "learning_rate": 5.3334709890563706e-05,
+      "loss": 1.0139,
+      "step": 11795
+    },
+    {
+      "epoch": 0.4775394577094294,
+      "grad_norm": 1.2693864107131958,
+      "learning_rate": 5.3314061532108196e-05,
+      "loss": 1.0385,
+      "step": 11800
+    },
+    {
+      "epoch": 0.47774180493727236,
+      "grad_norm": 1.257041096687317,
+      "learning_rate": 5.3293413173652694e-05,
+      "loss": 0.9903,
+      "step": 11805
+    },
+    {
+      "epoch": 0.4779441521651153,
+      "grad_norm": 1.1452276706695557,
+      "learning_rate": 5.32727648151972e-05,
+      "loss": 1.0217,
+      "step": 11810
+    },
+    {
+      "epoch": 0.47814649939295834,
+      "grad_norm": 1.167068362236023,
+      "learning_rate": 5.3252116456741695e-05,
+      "loss": 1.0112,
+      "step": 11815
+    },
+    {
+      "epoch": 0.4783488466208013,
+      "grad_norm": 1.1084812879562378,
+      "learning_rate": 5.3231468098286186e-05,
+      "loss": 0.9981,
+      "step": 11820
+    },
+    {
+      "epoch": 0.47855119384864425,
+      "grad_norm": 1.2631816864013672,
+      "learning_rate": 5.321081973983068e-05,
+      "loss": 0.9906,
+      "step": 11825
+    },
+    {
+      "epoch": 0.47875354107648727,
+      "grad_norm": 1.2514671087265015,
+      "learning_rate": 5.319017138137519e-05,
+      "loss": 1.0225,
+      "step": 11830
+    },
+    {
+      "epoch": 0.4789558883043302,
+      "grad_norm": 1.3078927993774414,
+      "learning_rate": 5.316952302291968e-05,
+      "loss": 0.9914,
+      "step": 11835
+    },
+    {
+      "epoch": 0.4791582355321732,
+      "grad_norm": 1.336737036705017,
+      "learning_rate": 5.3148874664464175e-05,
+      "loss": 1.0396,
+      "step": 11840
+    },
+    {
+      "epoch": 0.4793605827600162,
+      "grad_norm": 1.0106703042984009,
+      "learning_rate": 5.312822630600868e-05,
+      "loss": 0.9494,
+      "step": 11845
+    },
+    {
+      "epoch": 0.47956292998785915,
+      "grad_norm": 1.2406935691833496,
+      "learning_rate": 5.3107577947553176e-05,
+      "loss": 0.96,
+      "step": 11850
+    },
+    {
+      "epoch": 0.47976527721570217,
+      "grad_norm": 1.166337251663208,
+      "learning_rate": 5.308692958909767e-05,
+      "loss": 0.9897,
+      "step": 11855
+    },
+    {
+      "epoch": 0.4799676244435451,
+      "grad_norm": 1.230905532836914,
+      "learning_rate": 5.306628123064217e-05,
+      "loss": 1.0616,
+      "step": 11860
+    },
+    {
+      "epoch": 0.4801699716713881,
+      "grad_norm": 1.2330104112625122,
+      "learning_rate": 5.304563287218667e-05,
+      "loss": 1.0153,
+      "step": 11865
+    },
+    {
+      "epoch": 0.4803723188992311,
+      "grad_norm": 1.6057419776916504,
+      "learning_rate": 5.302498451373116e-05,
+      "loss": 0.9794,
+      "step": 11870
+    },
+    {
+      "epoch": 0.48057466612707406,
+      "grad_norm": 1.1932746171951294,
+      "learning_rate": 5.3004336155275656e-05,
+      "loss": 0.9847,
+      "step": 11875
+    },
+    {
+      "epoch": 0.480777013354917,
+      "grad_norm": 1.1735295057296753,
+      "learning_rate": 5.298368779682016e-05,
+      "loss": 1.0061,
+      "step": 11880
+    },
+    {
+      "epoch": 0.48097936058276003,
+      "grad_norm": 1.2277863025665283,
+      "learning_rate": 5.296303943836465e-05,
+      "loss": 1.0974,
+      "step": 11885
+    },
+    {
+      "epoch": 0.481181707810603,
+      "grad_norm": 1.1564351320266724,
+      "learning_rate": 5.294239107990915e-05,
+      "loss": 1.0016,
+      "step": 11890
+    },
+    {
+      "epoch": 0.48138405503844595,
+      "grad_norm": 1.0572887659072876,
+      "learning_rate": 5.292174272145365e-05,
+      "loss": 0.9457,
+      "step": 11895
+    },
+    {
+      "epoch": 0.48158640226628896,
+      "grad_norm": 1.0993452072143555,
+      "learning_rate": 5.290109436299815e-05,
+      "loss": 0.9996,
+      "step": 11900
+    },
+    {
+      "epoch": 0.4817887494941319,
+      "grad_norm": 1.2279669046401978,
+      "learning_rate": 5.288044600454264e-05,
+      "loss": 1.0324,
+      "step": 11905
+    },
+    {
+      "epoch": 0.48199109672197493,
+      "grad_norm": 1.054007649421692,
+      "learning_rate": 5.285979764608714e-05,
+      "loss": 1.0432,
+      "step": 11910
+    },
+    {
+      "epoch": 0.4821934439498179,
+      "grad_norm": 1.1128709316253662,
+      "learning_rate": 5.283914928763164e-05,
+      "loss": 1.0279,
+      "step": 11915
+    },
+    {
+      "epoch": 0.48239579117766085,
+      "grad_norm": 1.351121425628662,
+      "learning_rate": 5.281850092917613e-05,
+      "loss": 0.9776,
+      "step": 11920
+    },
+    {
+      "epoch": 0.48259813840550386,
+      "grad_norm": 1.1498552560806274,
+      "learning_rate": 5.279785257072063e-05,
+      "loss": 0.9965,
+      "step": 11925
+    },
+    {
+      "epoch": 0.4828004856333468,
+      "grad_norm": 1.1827574968338013,
+      "learning_rate": 5.277720421226513e-05,
+      "loss": 0.9636,
+      "step": 11930
+    },
+    {
+      "epoch": 0.4830028328611898,
+      "grad_norm": 1.0850706100463867,
+      "learning_rate": 5.275655585380962e-05,
+      "loss": 1.0092,
+      "step": 11935
+    },
+    {
+      "epoch": 0.4832051800890328,
+      "grad_norm": 1.1755199432373047,
+      "learning_rate": 5.273590749535412e-05,
+      "loss": 1.0064,
+      "step": 11940
+    },
+    {
+      "epoch": 0.48340752731687575,
+      "grad_norm": 1.2772828340530396,
+      "learning_rate": 5.271525913689862e-05,
+      "loss": 1.0346,
+      "step": 11945
+    },
+    {
+      "epoch": 0.4836098745447187,
+      "grad_norm": 1.2006957530975342,
+      "learning_rate": 5.269461077844312e-05,
+      "loss": 0.9633,
+      "step": 11950
+    },
+    {
+      "epoch": 0.4838122217725617,
+      "grad_norm": 1.2521257400512695,
+      "learning_rate": 5.267396241998761e-05,
+      "loss": 1.0277,
+      "step": 11955
+    },
+    {
+      "epoch": 0.4840145690004047,
+      "grad_norm": 1.2037335634231567,
+      "learning_rate": 5.265331406153211e-05,
+      "loss": 1.0424,
+      "step": 11960
+    },
+    {
+      "epoch": 0.4842169162282477,
+      "grad_norm": 1.098489761352539,
+      "learning_rate": 5.2632665703076614e-05,
+      "loss": 1.0258,
+      "step": 11965
+    },
+    {
+      "epoch": 0.48441926345609065,
+      "grad_norm": 1.0685847997665405,
+      "learning_rate": 5.26120173446211e-05,
+      "loss": 1.0309,
+      "step": 11970
+    },
+    {
+      "epoch": 0.4846216106839336,
+      "grad_norm": 1.1604846715927124,
+      "learning_rate": 5.25913689861656e-05,
+      "loss": 1.0864,
+      "step": 11975
+    },
+    {
+      "epoch": 0.4848239579117766,
+      "grad_norm": 1.1499050855636597,
+      "learning_rate": 5.25707206277101e-05,
+      "loss": 0.9741,
+      "step": 11980
+    },
+    {
+      "epoch": 0.4850263051396196,
+      "grad_norm": 1.188353180885315,
+      "learning_rate": 5.25500722692546e-05,
+      "loss": 0.9864,
+      "step": 11985
+    },
+    {
+      "epoch": 0.48522865236746254,
+      "grad_norm": 1.1696805953979492,
+      "learning_rate": 5.2529423910799094e-05,
+      "loss": 0.9617,
+      "step": 11990
+    },
+    {
+      "epoch": 0.48543099959530556,
+      "grad_norm": 1.3206312656402588,
+      "learning_rate": 5.250877555234359e-05,
+      "loss": 1.0467,
+      "step": 11995
+    },
+    {
+      "epoch": 0.4856333468231485,
+      "grad_norm": 1.13299560546875,
+      "learning_rate": 5.2488127193888095e-05,
+      "loss": 0.973,
+      "step": 12000
+    },
+    {
+      "epoch": 0.48583569405099153,
+      "grad_norm": 1.416884422302246,
+      "learning_rate": 5.246747883543258e-05,
+      "loss": 1.0054,
+      "step": 12005
+    },
+    {
+      "epoch": 0.4860380412788345,
+      "grad_norm": 1.2453945875167847,
+      "learning_rate": 5.244683047697708e-05,
+      "loss": 1.0447,
+      "step": 12010
+    },
+    {
+      "epoch": 0.48624038850667745,
+      "grad_norm": 1.1166871786117554,
+      "learning_rate": 5.242618211852158e-05,
+      "loss": 0.9752,
+      "step": 12015
+    },
+    {
+      "epoch": 0.48644273573452046,
+      "grad_norm": 1.2610543966293335,
+      "learning_rate": 5.240553376006607e-05,
+      "loss": 0.9689,
+      "step": 12020
+    },
+    {
+      "epoch": 0.4866450829623634,
+      "grad_norm": 1.2604610919952393,
+      "learning_rate": 5.2384885401610575e-05,
+      "loss": 0.9667,
+      "step": 12025
+    },
+    {
+      "epoch": 0.4868474301902064,
+      "grad_norm": 1.3531112670898438,
+      "learning_rate": 5.236423704315507e-05,
+      "loss": 1.0058,
+      "step": 12030
+    },
+    {
+      "epoch": 0.4870497774180494,
+      "grad_norm": 1.0576692819595337,
+      "learning_rate": 5.2343588684699576e-05,
+      "loss": 1.0281,
+      "step": 12035
+    },
+    {
+      "epoch": 0.48725212464589235,
+      "grad_norm": 1.122986912727356,
+      "learning_rate": 5.232294032624406e-05,
+      "loss": 1.0254,
+      "step": 12040
+    },
+    {
+      "epoch": 0.4874544718737353,
+      "grad_norm": 1.157700538635254,
+      "learning_rate": 5.2302291967788564e-05,
+      "loss": 0.9843,
+      "step": 12045
+    },
+    {
+      "epoch": 0.4876568191015783,
+      "grad_norm": 1.2433433532714844,
+      "learning_rate": 5.228164360933306e-05,
+      "loss": 0.9999,
+      "step": 12050
+    },
+    {
+      "epoch": 0.4878591663294213,
+      "grad_norm": 1.188632845878601,
+      "learning_rate": 5.226099525087755e-05,
+      "loss": 0.9848,
+      "step": 12055
+    },
+    {
+      "epoch": 0.4880615135572643,
+      "grad_norm": 1.1356099843978882,
+      "learning_rate": 5.2240346892422056e-05,
+      "loss": 0.936,
+      "step": 12060
+    },
+    {
+      "epoch": 0.48826386078510725,
+      "grad_norm": 1.1671898365020752,
+      "learning_rate": 5.221969853396655e-05,
+      "loss": 0.9922,
+      "step": 12065
+    },
+    {
+      "epoch": 0.4884662080129502,
+      "grad_norm": 1.1089935302734375,
+      "learning_rate": 5.2199050175511044e-05,
+      "loss": 1.0443,
+      "step": 12070
+    },
+    {
+      "epoch": 0.4886685552407932,
+      "grad_norm": 1.323712706565857,
+      "learning_rate": 5.217840181705555e-05,
+      "loss": 1.0233,
+      "step": 12075
+    },
+    {
+      "epoch": 0.4888709024686362,
+      "grad_norm": 1.0796269178390503,
+      "learning_rate": 5.2157753458600045e-05,
+      "loss": 1.0662,
+      "step": 12080
+    },
+    {
+      "epoch": 0.48907324969647914,
+      "grad_norm": 1.1861119270324707,
+      "learning_rate": 5.213710510014454e-05,
+      "loss": 0.9814,
+      "step": 12085
+    },
+    {
+      "epoch": 0.48927559692432215,
+      "grad_norm": 1.104269027709961,
+      "learning_rate": 5.211645674168903e-05,
+      "loss": 0.9476,
+      "step": 12090
+    },
+    {
+      "epoch": 0.4894779441521651,
+      "grad_norm": 1.081963300704956,
+      "learning_rate": 5.209580838323354e-05,
+      "loss": 1.0231,
+      "step": 12095
+    },
+    {
+      "epoch": 0.48968029138000807,
+      "grad_norm": 1.1605753898620605,
+      "learning_rate": 5.2075160024778034e-05,
+      "loss": 1.0055,
+      "step": 12100
+    },
+    {
+      "epoch": 0.4898826386078511,
+      "grad_norm": 1.227674961090088,
+      "learning_rate": 5.2054511666322525e-05,
+      "loss": 0.9811,
+      "step": 12105
+    },
+    {
+      "epoch": 0.49008498583569404,
+      "grad_norm": 1.1121258735656738,
+      "learning_rate": 5.203386330786703e-05,
+      "loss": 1.0026,
+      "step": 12110
+    },
+    {
+      "epoch": 0.49028733306353706,
+      "grad_norm": 1.1806455850601196,
+      "learning_rate": 5.2013214949411526e-05,
+      "loss": 0.9891,
+      "step": 12115
+    },
+    {
+      "epoch": 0.49048968029138,
+      "grad_norm": 1.2708841562271118,
+      "learning_rate": 5.1992566590956023e-05,
+      "loss": 0.9987,
+      "step": 12120
+    },
+    {
+      "epoch": 0.490692027519223,
+      "grad_norm": 1.3260775804519653,
+      "learning_rate": 5.1971918232500514e-05,
+      "loss": 0.9689,
+      "step": 12125
+    },
+    {
+      "epoch": 0.490894374747066,
+      "grad_norm": 1.106037974357605,
+      "learning_rate": 5.195126987404502e-05,
+      "loss": 1.0022,
+      "step": 12130
+    },
+    {
+      "epoch": 0.49109672197490895,
+      "grad_norm": 1.2738261222839355,
+      "learning_rate": 5.1930621515589515e-05,
+      "loss": 1.0052,
+      "step": 12135
+    },
+    {
+      "epoch": 0.4912990692027519,
+      "grad_norm": 1.1397292613983154,
+      "learning_rate": 5.1909973157134006e-05,
+      "loss": 0.9537,
+      "step": 12140
+    },
+    {
+      "epoch": 0.4915014164305949,
+      "grad_norm": 1.1503263711929321,
+      "learning_rate": 5.188932479867851e-05,
+      "loss": 1.0132,
+      "step": 12145
+    },
+    {
+      "epoch": 0.4917037636584379,
+      "grad_norm": 1.117822289466858,
+      "learning_rate": 5.186867644022301e-05,
+      "loss": 1.0289,
+      "step": 12150
+    },
+    {
+      "epoch": 0.49190611088628083,
+      "grad_norm": 1.0901275873184204,
+      "learning_rate": 5.18480280817675e-05,
+      "loss": 0.993,
+      "step": 12155
+    },
+    {
+      "epoch": 0.49210845811412385,
+      "grad_norm": 1.1580654382705688,
+      "learning_rate": 5.1827379723311995e-05,
+      "loss": 0.9897,
+      "step": 12160
+    },
+    {
+      "epoch": 0.4923108053419668,
+      "grad_norm": 1.170100212097168,
+      "learning_rate": 5.18067313648565e-05,
+      "loss": 1.0015,
+      "step": 12165
+    },
+    {
+      "epoch": 0.4925131525698098,
+      "grad_norm": 1.1789393424987793,
+      "learning_rate": 5.1786083006400996e-05,
+      "loss": 0.9359,
+      "step": 12170
+    },
+    {
+      "epoch": 0.4927154997976528,
+      "grad_norm": 1.11495840549469,
+      "learning_rate": 5.176543464794549e-05,
+      "loss": 1.0539,
+      "step": 12175
+    },
+    {
+      "epoch": 0.49291784702549574,
+      "grad_norm": 1.2007088661193848,
+      "learning_rate": 5.174478628948999e-05,
+      "loss": 1.0957,
+      "step": 12180
+    },
+    {
+      "epoch": 0.49312019425333875,
+      "grad_norm": 1.1438637971878052,
+      "learning_rate": 5.172413793103449e-05,
+      "loss": 1.0123,
+      "step": 12185
+    },
+    {
+      "epoch": 0.4933225414811817,
+      "grad_norm": 1.118512511253357,
+      "learning_rate": 5.170348957257898e-05,
+      "loss": 0.9884,
+      "step": 12190
+    },
+    {
+      "epoch": 0.49352488870902467,
+      "grad_norm": 1.3337222337722778,
+      "learning_rate": 5.1682841214123476e-05,
+      "loss": 0.9369,
+      "step": 12195
+    },
+    {
+      "epoch": 0.4937272359368677,
+      "grad_norm": 1.2440013885498047,
+      "learning_rate": 5.166219285566798e-05,
+      "loss": 1.0743,
+      "step": 12200
+    },
+    {
+      "epoch": 0.49392958316471064,
+      "grad_norm": 1.0439364910125732,
+      "learning_rate": 5.164154449721248e-05,
+      "loss": 1.0591,
+      "step": 12205
+    },
+    {
+      "epoch": 0.4941319303925536,
+      "grad_norm": 1.130218505859375,
+      "learning_rate": 5.162089613875697e-05,
+      "loss": 1.0178,
+      "step": 12210
+    },
+    {
+      "epoch": 0.4943342776203966,
+      "grad_norm": 1.1202657222747803,
+      "learning_rate": 5.160024778030147e-05,
+      "loss": 0.9759,
+      "step": 12215
+    },
+    {
+      "epoch": 0.49453662484823957,
+      "grad_norm": 1.234213948249817,
+      "learning_rate": 5.157959942184597e-05,
+      "loss": 0.9899,
+      "step": 12220
+    },
+    {
+      "epoch": 0.4947389720760826,
+      "grad_norm": 1.201436161994934,
+      "learning_rate": 5.155895106339046e-05,
+      "loss": 1.0586,
+      "step": 12225
+    },
+    {
+      "epoch": 0.49494131930392554,
+      "grad_norm": 1.166717767715454,
+      "learning_rate": 5.153830270493496e-05,
+      "loss": 0.9978,
+      "step": 12230
+    },
+    {
+      "epoch": 0.4951436665317685,
+      "grad_norm": 1.1749318838119507,
+      "learning_rate": 5.151765434647946e-05,
+      "loss": 0.967,
+      "step": 12235
+    },
+    {
+      "epoch": 0.4953460137596115,
+      "grad_norm": 1.2312084436416626,
+      "learning_rate": 5.149700598802395e-05,
+      "loss": 0.9739,
+      "step": 12240
+    },
+    {
+      "epoch": 0.4955483609874545,
+      "grad_norm": 1.212628960609436,
+      "learning_rate": 5.147635762956845e-05,
+      "loss": 0.9882,
+      "step": 12245
+    },
+    {
+      "epoch": 0.49575070821529743,
+      "grad_norm": 1.3387917280197144,
+      "learning_rate": 5.145570927111295e-05,
+      "loss": 1.0215,
+      "step": 12250
+    },
+    {
+      "epoch": 0.49595305544314044,
+      "grad_norm": 1.0717612504959106,
+      "learning_rate": 5.143506091265745e-05,
+      "loss": 1.054,
+      "step": 12255
+    },
+    {
+      "epoch": 0.4961554026709834,
+      "grad_norm": 1.0845282077789307,
+      "learning_rate": 5.141441255420194e-05,
+      "loss": 0.972,
+      "step": 12260
+    },
+    {
+      "epoch": 0.49635774989882636,
+      "grad_norm": 1.11933171749115,
+      "learning_rate": 5.139376419574644e-05,
+      "loss": 0.9062,
+      "step": 12265
+    },
+    {
+      "epoch": 0.4965600971266694,
+      "grad_norm": 1.1734405755996704,
+      "learning_rate": 5.137311583729094e-05,
+      "loss": 0.9905,
+      "step": 12270
+    },
+    {
+      "epoch": 0.49676244435451233,
+      "grad_norm": 1.300087332725525,
+      "learning_rate": 5.135246747883543e-05,
+      "loss": 0.9919,
+      "step": 12275
+    },
+    {
+      "epoch": 0.49696479158235535,
+      "grad_norm": 1.2729748487472534,
+      "learning_rate": 5.133181912037993e-05,
+      "loss": 0.9914,
+      "step": 12280
+    },
+    {
+      "epoch": 0.4971671388101983,
+      "grad_norm": 1.3504347801208496,
+      "learning_rate": 5.1311170761924434e-05,
+      "loss": 1.0429,
+      "step": 12285
+    },
+    {
+      "epoch": 0.49736948603804126,
+      "grad_norm": 1.2052627801895142,
+      "learning_rate": 5.1290522403468925e-05,
+      "loss": 1.0078,
+      "step": 12290
+    },
+    {
+      "epoch": 0.4975718332658843,
+      "grad_norm": 1.3131917715072632,
+      "learning_rate": 5.126987404501342e-05,
+      "loss": 0.9245,
+      "step": 12295
+    },
+    {
+      "epoch": 0.49777418049372724,
+      "grad_norm": 1.3341904878616333,
+      "learning_rate": 5.124922568655792e-05,
+      "loss": 0.9872,
+      "step": 12300
+    },
+    {
+      "epoch": 0.4979765277215702,
+      "grad_norm": 1.1806317567825317,
+      "learning_rate": 5.1228577328102423e-05,
+      "loss": 0.9939,
+      "step": 12305
+    },
+    {
+      "epoch": 0.4981788749494132,
+      "grad_norm": 1.2073943614959717,
+      "learning_rate": 5.1207928969646914e-05,
+      "loss": 1.0261,
+      "step": 12310
+    },
+    {
+      "epoch": 0.49838122217725617,
+      "grad_norm": 1.3189400434494019,
+      "learning_rate": 5.118728061119141e-05,
+      "loss": 1.0034,
+      "step": 12315
+    },
+    {
+      "epoch": 0.4985835694050991,
+      "grad_norm": 1.1953120231628418,
+      "learning_rate": 5.1166632252735915e-05,
+      "loss": 0.9618,
+      "step": 12320
+    },
+    {
+      "epoch": 0.49878591663294214,
+      "grad_norm": 1.1699455976486206,
+      "learning_rate": 5.1145983894280406e-05,
+      "loss": 1.0228,
+      "step": 12325
+    },
+    {
+      "epoch": 0.4989882638607851,
+      "grad_norm": 1.1556991338729858,
+      "learning_rate": 5.11253355358249e-05,
+      "loss": 0.9953,
+      "step": 12330
+    },
+    {
+      "epoch": 0.4991906110886281,
+      "grad_norm": 1.2345956563949585,
+      "learning_rate": 5.11046871773694e-05,
+      "loss": 0.9339,
+      "step": 12335
+    },
+    {
+      "epoch": 0.49939295831647107,
+      "grad_norm": 1.6040476560592651,
+      "learning_rate": 5.1084038818913904e-05,
+      "loss": 0.9312,
+      "step": 12340
+    },
+    {
+      "epoch": 0.49959530554431403,
+      "grad_norm": 1.1234400272369385,
+      "learning_rate": 5.1063390460458395e-05,
+      "loss": 1.0077,
+      "step": 12345
+    },
+    {
+      "epoch": 0.49979765277215704,
+      "grad_norm": 1.2733681201934814,
+      "learning_rate": 5.104274210200289e-05,
+      "loss": 0.9654,
+      "step": 12350
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.193357229232788,
+      "learning_rate": 5.1022093743547396e-05,
+      "loss": 0.9975,
+      "step": 12355
+    },
+    {
+      "epoch": 0.500202347227843,
+      "grad_norm": 1.1366767883300781,
+      "learning_rate": 5.100144538509189e-05,
+      "loss": 0.9933,
+      "step": 12360
+    },
+    {
+      "epoch": 0.5004046944556859,
+      "grad_norm": 1.148776888847351,
+      "learning_rate": 5.0980797026636384e-05,
+      "loss": 1.0654,
+      "step": 12365
+    },
+    {
+      "epoch": 0.500607041683529,
+      "grad_norm": 1.2743234634399414,
+      "learning_rate": 5.096014866818088e-05,
+      "loss": 0.9324,
+      "step": 12370
+    },
+    {
+      "epoch": 0.5008093889113719,
+      "grad_norm": 1.2196996212005615,
+      "learning_rate": 5.093950030972537e-05,
+      "loss": 1.0129,
+      "step": 12375
+    },
+    {
+      "epoch": 0.5010117361392149,
+      "grad_norm": 1.080466389656067,
+      "learning_rate": 5.0918851951269876e-05,
+      "loss": 0.967,
+      "step": 12380
+    },
+    {
+      "epoch": 0.5012140833670579,
+      "grad_norm": 1.15315580368042,
+      "learning_rate": 5.089820359281437e-05,
+      "loss": 1.0039,
+      "step": 12385
+    },
+    {
+      "epoch": 0.5014164305949008,
+      "grad_norm": 1.1998361349105835,
+      "learning_rate": 5.087755523435888e-05,
+      "loss": 1.0225,
+      "step": 12390
+    },
+    {
+      "epoch": 0.5016187778227438,
+      "grad_norm": 1.108877420425415,
+      "learning_rate": 5.085690687590337e-05,
+      "loss": 0.9265,
+      "step": 12395
+    },
+    {
+      "epoch": 0.5018211250505868,
+      "grad_norm": 1.1324632167816162,
+      "learning_rate": 5.0836258517447865e-05,
+      "loss": 0.9246,
+      "step": 12400
+    },
+    {
+      "epoch": 0.5020234722784298,
+      "grad_norm": 1.2282084226608276,
+      "learning_rate": 5.081561015899237e-05,
+      "loss": 1.0447,
+      "step": 12405
+    },
+    {
+      "epoch": 0.5022258195062728,
+      "grad_norm": 1.1794309616088867,
+      "learning_rate": 5.079496180053685e-05,
+      "loss": 1.0202,
+      "step": 12410
+    },
+    {
+      "epoch": 0.5024281667341157,
+      "grad_norm": 1.2083922624588013,
+      "learning_rate": 5.077431344208136e-05,
+      "loss": 1.006,
+      "step": 12415
+    },
+    {
+      "epoch": 0.5026305139619587,
+      "grad_norm": 1.2532967329025269,
+      "learning_rate": 5.0753665083625854e-05,
+      "loss": 1.0159,
+      "step": 12420
+    },
+    {
+      "epoch": 0.5028328611898017,
+      "grad_norm": 1.284624457359314,
+      "learning_rate": 5.0733016725170345e-05,
+      "loss": 1.0005,
+      "step": 12425
+    },
+    {
+      "epoch": 0.5030352084176447,
+      "grad_norm": 1.196325659751892,
+      "learning_rate": 5.071236836671485e-05,
+      "loss": 1.0457,
+      "step": 12430
+    },
+    {
+      "epoch": 0.5032375556454877,
+      "grad_norm": 1.2377886772155762,
+      "learning_rate": 5.0691720008259346e-05,
+      "loss": 1.0229,
+      "step": 12435
+    },
+    {
+      "epoch": 0.5034399028733306,
+      "grad_norm": 1.2728221416473389,
+      "learning_rate": 5.067107164980385e-05,
+      "loss": 1.0108,
+      "step": 12440
+    },
+    {
+      "epoch": 0.5036422501011736,
+      "grad_norm": 1.28044855594635,
+      "learning_rate": 5.0650423291348334e-05,
+      "loss": 1.0369,
+      "step": 12445
+    },
+    {
+      "epoch": 0.5038445973290165,
+      "grad_norm": 1.1525059938430786,
+      "learning_rate": 5.062977493289284e-05,
+      "loss": 0.9861,
+      "step": 12450
+    },
+    {
+      "epoch": 0.5040469445568596,
+      "grad_norm": 1.2731013298034668,
+      "learning_rate": 5.0609126574437336e-05,
+      "loss": 1.0362,
+      "step": 12455
+    },
+    {
+      "epoch": 0.5042492917847026,
+      "grad_norm": 1.2046812772750854,
+      "learning_rate": 5.0588478215981826e-05,
+      "loss": 1.0207,
+      "step": 12460
+    },
+    {
+      "epoch": 0.5044516390125455,
+      "grad_norm": 1.1956570148468018,
+      "learning_rate": 5.056782985752633e-05,
+      "loss": 0.9847,
+      "step": 12465
+    },
+    {
+      "epoch": 0.5046539862403885,
+      "grad_norm": 1.0756455659866333,
+      "learning_rate": 5.054718149907083e-05,
+      "loss": 0.9889,
+      "step": 12470
+    },
+    {
+      "epoch": 0.5048563334682314,
+      "grad_norm": 1.2245359420776367,
+      "learning_rate": 5.052653314061533e-05,
+      "loss": 1.043,
+      "step": 12475
+    },
+    {
+      "epoch": 0.5050586806960745,
+      "grad_norm": 1.3225243091583252,
+      "learning_rate": 5.0505884782159815e-05,
+      "loss": 0.9622,
+      "step": 12480
+    },
+    {
+      "epoch": 0.5052610279239175,
+      "grad_norm": 1.2724545001983643,
+      "learning_rate": 5.048523642370432e-05,
+      "loss": 1.0257,
+      "step": 12485
+    },
+    {
+      "epoch": 0.5054633751517604,
+      "grad_norm": 1.2298985719680786,
+      "learning_rate": 5.0464588065248817e-05,
+      "loss": 0.9616,
+      "step": 12490
+    },
+    {
+      "epoch": 0.5056657223796034,
+      "grad_norm": 1.1233696937561035,
+      "learning_rate": 5.044393970679331e-05,
+      "loss": 0.9892,
+      "step": 12495
+    },
+    {
+      "epoch": 0.5058680696074463,
+      "grad_norm": 1.1848814487457275,
+      "learning_rate": 5.042329134833781e-05,
+      "loss": 1.0535,
+      "step": 12500
+    },
+    {
+      "epoch": 0.5060704168352893,
+      "grad_norm": 1.1247293949127197,
+      "learning_rate": 5.040264298988231e-05,
+      "loss": 0.9868,
+      "step": 12505
+    },
+    {
+      "epoch": 0.5062727640631324,
+      "grad_norm": 1.1931838989257812,
+      "learning_rate": 5.03819946314268e-05,
+      "loss": 0.9507,
+      "step": 12510
+    },
+    {
+      "epoch": 0.5064751112909753,
+      "grad_norm": 1.1550499200820923,
+      "learning_rate": 5.0361346272971296e-05,
+      "loss": 1.051,
+      "step": 12515
+    },
+    {
+      "epoch": 0.5066774585188183,
+      "grad_norm": 1.3351200819015503,
+      "learning_rate": 5.03406979145158e-05,
+      "loss": 0.9849,
+      "step": 12520
+    },
+    {
+      "epoch": 0.5068798057466612,
+      "grad_norm": 1.156421184539795,
+      "learning_rate": 5.03200495560603e-05,
+      "loss": 0.9706,
+      "step": 12525
+    },
+    {
+      "epoch": 0.5070821529745042,
+      "grad_norm": 1.1729159355163574,
+      "learning_rate": 5.029940119760479e-05,
+      "loss": 1.0254,
+      "step": 12530
+    },
+    {
+      "epoch": 0.5072845002023473,
+      "grad_norm": 1.1957979202270508,
+      "learning_rate": 5.027875283914929e-05,
+      "loss": 0.985,
+      "step": 12535
+    },
+    {
+      "epoch": 0.5074868474301902,
+      "grad_norm": 1.2675886154174805,
+      "learning_rate": 5.025810448069379e-05,
+      "loss": 1.0182,
+      "step": 12540
+    },
+    {
+      "epoch": 0.5076891946580332,
+      "grad_norm": 1.2329570055007935,
+      "learning_rate": 5.023745612223828e-05,
+      "loss": 0.9963,
+      "step": 12545
+    },
+    {
+      "epoch": 0.5078915418858762,
+      "grad_norm": 1.115394115447998,
+      "learning_rate": 5.021680776378278e-05,
+      "loss": 1.0099,
+      "step": 12550
+    },
+    {
+      "epoch": 0.5080938891137191,
+      "grad_norm": 1.1686569452285767,
+      "learning_rate": 5.019615940532728e-05,
+      "loss": 0.9875,
+      "step": 12555
+    },
+    {
+      "epoch": 0.5082962363415621,
+      "grad_norm": 1.1489678621292114,
+      "learning_rate": 5.017551104687178e-05,
+      "loss": 0.9574,
+      "step": 12560
+    },
+    {
+      "epoch": 0.5084985835694051,
+      "grad_norm": 1.2116782665252686,
+      "learning_rate": 5.015486268841627e-05,
+      "loss": 1.0453,
+      "step": 12565
+    },
+    {
+      "epoch": 0.5087009307972481,
+      "grad_norm": 1.6617372035980225,
+      "learning_rate": 5.013421432996077e-05,
+      "loss": 1.0092,
+      "step": 12570
+    },
+    {
+      "epoch": 0.508903278025091,
+      "grad_norm": 1.1859623193740845,
+      "learning_rate": 5.011356597150527e-05,
+      "loss": 0.9867,
+      "step": 12575
+    },
+    {
+      "epoch": 0.509105625252934,
+      "grad_norm": 1.0471400022506714,
+      "learning_rate": 5.009291761304976e-05,
+      "loss": 1.0128,
+      "step": 12580
+    },
+    {
+      "epoch": 0.509307972480777,
+      "grad_norm": 1.1846133470535278,
+      "learning_rate": 5.007226925459426e-05,
+      "loss": 1.0121,
+      "step": 12585
+    },
+    {
+      "epoch": 0.50951031970862,
+      "grad_norm": 1.2684739828109741,
+      "learning_rate": 5.005162089613876e-05,
+      "loss": 0.9923,
+      "step": 12590
+    },
+    {
+      "epoch": 0.509712666936463,
+      "grad_norm": 1.230245590209961,
+      "learning_rate": 5.003097253768325e-05,
+      "loss": 1.0056,
+      "step": 12595
+    },
+    {
+      "epoch": 0.509915014164306,
+      "grad_norm": 1.1914970874786377,
+      "learning_rate": 5.001032417922775e-05,
+      "loss": 0.9449,
+      "step": 12600
+    },
+    {
+      "epoch": 0.5101173613921489,
+      "grad_norm": 1.1601983308792114,
+      "learning_rate": 4.9989675820772254e-05,
+      "loss": 0.988,
+      "step": 12605
+    },
+    {
+      "epoch": 0.5103197086199919,
+      "grad_norm": 1.1290122270584106,
+      "learning_rate": 4.9969027462316745e-05,
+      "loss": 1.0195,
+      "step": 12610
+    },
+    {
+      "epoch": 0.5105220558478348,
+      "grad_norm": 1.2394769191741943,
+      "learning_rate": 4.994837910386125e-05,
+      "loss": 0.9847,
+      "step": 12615
+    },
+    {
+      "epoch": 0.5107244030756779,
+      "grad_norm": 1.1106399297714233,
+      "learning_rate": 4.9927730745405746e-05,
+      "loss": 0.9432,
+      "step": 12620
+    },
+    {
+      "epoch": 0.5109267503035209,
+      "grad_norm": 1.2086856365203857,
+      "learning_rate": 4.990708238695024e-05,
+      "loss": 1.0175,
+      "step": 12625
+    },
+    {
+      "epoch": 0.5111290975313638,
+      "grad_norm": 1.1757665872573853,
+      "learning_rate": 4.988643402849474e-05,
+      "loss": 1.0162,
+      "step": 12630
+    },
+    {
+      "epoch": 0.5113314447592068,
+      "grad_norm": 1.10592782497406,
+      "learning_rate": 4.986578567003923e-05,
+      "loss": 1.0415,
+      "step": 12635
+    },
+    {
+      "epoch": 0.5115337919870497,
+      "grad_norm": 1.193536400794983,
+      "learning_rate": 4.9845137311583736e-05,
+      "loss": 0.9638,
+      "step": 12640
+    },
+    {
+      "epoch": 0.5117361392148928,
+      "grad_norm": 1.2107716798782349,
+      "learning_rate": 4.9824488953128226e-05,
+      "loss": 1.0404,
+      "step": 12645
+    },
+    {
+      "epoch": 0.5119384864427358,
+      "grad_norm": 1.1552340984344482,
+      "learning_rate": 4.980384059467272e-05,
+      "loss": 1.0049,
+      "step": 12650
+    },
+    {
+      "epoch": 0.5121408336705787,
+      "grad_norm": 1.0847322940826416,
+      "learning_rate": 4.978319223621723e-05,
+      "loss": 0.9955,
+      "step": 12655
+    },
+    {
+      "epoch": 0.5123431808984217,
+      "grad_norm": 1.1630275249481201,
+      "learning_rate": 4.976254387776172e-05,
+      "loss": 0.9893,
+      "step": 12660
+    },
+    {
+      "epoch": 0.5125455281262646,
+      "grad_norm": 1.1975666284561157,
+      "learning_rate": 4.974189551930622e-05,
+      "loss": 1.0614,
+      "step": 12665
+    },
+    {
+      "epoch": 0.5127478753541076,
+      "grad_norm": 1.1546673774719238,
+      "learning_rate": 4.972124716085071e-05,
+      "loss": 1.0466,
+      "step": 12670
+    },
+    {
+      "epoch": 0.5129502225819507,
+      "grad_norm": 1.2232493162155151,
+      "learning_rate": 4.970059880239521e-05,
+      "loss": 0.9794,
+      "step": 12675
+    },
+    {
+      "epoch": 0.5131525698097936,
+      "grad_norm": 1.173681378364563,
+      "learning_rate": 4.967995044393971e-05,
+      "loss": 1.0032,
+      "step": 12680
+    },
+    {
+      "epoch": 0.5133549170376366,
+      "grad_norm": 1.1275198459625244,
+      "learning_rate": 4.9659302085484204e-05,
+      "loss": 1.0221,
+      "step": 12685
+    },
+    {
+      "epoch": 0.5135572642654795,
+      "grad_norm": 1.188533067703247,
+      "learning_rate": 4.963865372702871e-05,
+      "loss": 0.941,
+      "step": 12690
+    },
+    {
+      "epoch": 0.5137596114933225,
+      "grad_norm": 1.095301628112793,
+      "learning_rate": 4.96180053685732e-05,
+      "loss": 0.9973,
+      "step": 12695
+    },
+    {
+      "epoch": 0.5139619587211656,
+      "grad_norm": 1.2311757802963257,
+      "learning_rate": 4.9597357010117696e-05,
+      "loss": 1.0309,
+      "step": 12700
+    },
+    {
+      "epoch": 0.5141643059490085,
+      "grad_norm": 1.1870546340942383,
+      "learning_rate": 4.9576708651662194e-05,
+      "loss": 1.0679,
+      "step": 12705
+    },
+    {
+      "epoch": 0.5143666531768515,
+      "grad_norm": 1.185866355895996,
+      "learning_rate": 4.955606029320669e-05,
+      "loss": 1.0102,
+      "step": 12710
+    },
+    {
+      "epoch": 0.5145690004046944,
+      "grad_norm": 1.1721301078796387,
+      "learning_rate": 4.953541193475119e-05,
+      "loss": 1.0154,
+      "step": 12715
+    },
+    {
+      "epoch": 0.5147713476325374,
+      "grad_norm": 1.2105289697647095,
+      "learning_rate": 4.9514763576295685e-05,
+      "loss": 1.0298,
+      "step": 12720
+    },
+    {
+      "epoch": 0.5149736948603804,
+      "grad_norm": 1.054449200630188,
+      "learning_rate": 4.949411521784019e-05,
+      "loss": 0.9709,
+      "step": 12725
+    },
+    {
+      "epoch": 0.5151760420882234,
+      "grad_norm": 1.1701349020004272,
+      "learning_rate": 4.947346685938468e-05,
+      "loss": 0.974,
+      "step": 12730
+    },
+    {
+      "epoch": 0.5153783893160664,
+      "grad_norm": 1.2717715501785278,
+      "learning_rate": 4.945281850092918e-05,
+      "loss": 1.0077,
+      "step": 12735
+    },
+    {
+      "epoch": 0.5155807365439093,
+      "grad_norm": 1.1190024614334106,
+      "learning_rate": 4.9432170142473675e-05,
+      "loss": 0.9512,
+      "step": 12740
+    },
+    {
+      "epoch": 0.5157830837717523,
+      "grad_norm": 1.200972557067871,
+      "learning_rate": 4.941152178401817e-05,
+      "loss": 1.0014,
+      "step": 12745
+    },
+    {
+      "epoch": 0.5159854309995953,
+      "grad_norm": 1.2721880674362183,
+      "learning_rate": 4.939087342556267e-05,
+      "loss": 0.9819,
+      "step": 12750
+    },
+    {
+      "epoch": 0.5161877782274383,
+      "grad_norm": 1.1532000303268433,
+      "learning_rate": 4.9370225067107167e-05,
+      "loss": 0.9611,
+      "step": 12755
+    },
+    {
+      "epoch": 0.5163901254552813,
+      "grad_norm": 1.1171449422836304,
+      "learning_rate": 4.9349576708651664e-05,
+      "loss": 0.9495,
+      "step": 12760
+    },
+    {
+      "epoch": 0.5165924726831242,
+      "grad_norm": 1.7924264669418335,
+      "learning_rate": 4.932892835019616e-05,
+      "loss": 1.0215,
+      "step": 12765
+    },
+    {
+      "epoch": 0.5167948199109672,
+      "grad_norm": 1.16851806640625,
+      "learning_rate": 4.930827999174066e-05,
+      "loss": 1.0273,
+      "step": 12770
+    },
+    {
+      "epoch": 0.5169971671388102,
+      "grad_norm": 1.1129615306854248,
+      "learning_rate": 4.9287631633285156e-05,
+      "loss": 1.0412,
+      "step": 12775
+    },
+    {
+      "epoch": 0.5171995143666531,
+      "grad_norm": 1.0944582223892212,
+      "learning_rate": 4.926698327482965e-05,
+      "loss": 1.0335,
+      "step": 12780
+    },
+    {
+      "epoch": 0.5174018615944962,
+      "grad_norm": 1.0807654857635498,
+      "learning_rate": 4.924633491637415e-05,
+      "loss": 0.9239,
+      "step": 12785
+    },
+    {
+      "epoch": 0.5176042088223392,
+      "grad_norm": 1.2499336004257202,
+      "learning_rate": 4.922568655791865e-05,
+      "loss": 0.9914,
+      "step": 12790
+    },
+    {
+      "epoch": 0.5178065560501821,
+      "grad_norm": 1.159722089767456,
+      "learning_rate": 4.9205038199463145e-05,
+      "loss": 0.9644,
+      "step": 12795
+    },
+    {
+      "epoch": 0.5180089032780251,
+      "grad_norm": 1.2099483013153076,
+      "learning_rate": 4.918438984100764e-05,
+      "loss": 0.9875,
+      "step": 12800
+    },
+    {
+      "epoch": 0.518211250505868,
+      "grad_norm": 1.1559176445007324,
+      "learning_rate": 4.916374148255214e-05,
+      "loss": 1.0176,
+      "step": 12805
+    },
+    {
+      "epoch": 0.5184135977337111,
+      "grad_norm": 1.1882922649383545,
+      "learning_rate": 4.914309312409664e-05,
+      "loss": 0.9865,
+      "step": 12810
+    },
+    {
+      "epoch": 0.518615944961554,
+      "grad_norm": 1.1380903720855713,
+      "learning_rate": 4.9122444765641134e-05,
+      "loss": 0.9426,
+      "step": 12815
+    },
+    {
+      "epoch": 0.518818292189397,
+      "grad_norm": 1.1586993932724,
+      "learning_rate": 4.910179640718563e-05,
+      "loss": 1.057,
+      "step": 12820
+    },
+    {
+      "epoch": 0.51902063941724,
+      "grad_norm": 1.2551969289779663,
+      "learning_rate": 4.908114804873013e-05,
+      "loss": 1.0243,
+      "step": 12825
+    },
+    {
+      "epoch": 0.5192229866450829,
+      "grad_norm": 1.0013608932495117,
+      "learning_rate": 4.9060499690274626e-05,
+      "loss": 0.9527,
+      "step": 12830
+    },
+    {
+      "epoch": 0.5194253338729259,
+      "grad_norm": 1.2260642051696777,
+      "learning_rate": 4.903985133181912e-05,
+      "loss": 0.9735,
+      "step": 12835
+    },
+    {
+      "epoch": 0.519627681100769,
+      "grad_norm": 1.1197712421417236,
+      "learning_rate": 4.901920297336362e-05,
+      "loss": 0.99,
+      "step": 12840
+    },
+    {
+      "epoch": 0.5198300283286119,
+      "grad_norm": 1.198989748954773,
+      "learning_rate": 4.899855461490812e-05,
+      "loss": 0.9567,
+      "step": 12845
+    },
+    {
+      "epoch": 0.5200323755564549,
+      "grad_norm": 1.0800516605377197,
+      "learning_rate": 4.8977906256452615e-05,
+      "loss": 0.9871,
+      "step": 12850
+    },
+    {
+      "epoch": 0.5202347227842978,
+      "grad_norm": 1.0519342422485352,
+      "learning_rate": 4.895725789799711e-05,
+      "loss": 1.0305,
+      "step": 12855
+    },
+    {
+      "epoch": 0.5204370700121408,
+      "grad_norm": 1.2230823040008545,
+      "learning_rate": 4.893660953954161e-05,
+      "loss": 0.9869,
+      "step": 12860
+    },
+    {
+      "epoch": 0.5206394172399839,
+      "grad_norm": 1.1805706024169922,
+      "learning_rate": 4.891596118108611e-05,
+      "loss": 1.0341,
+      "step": 12865
+    },
+    {
+      "epoch": 0.5208417644678268,
+      "grad_norm": 1.1650760173797607,
+      "learning_rate": 4.8895312822630604e-05,
+      "loss": 0.9969,
+      "step": 12870
+    },
+    {
+      "epoch": 0.5210441116956698,
+      "grad_norm": 1.0749975442886353,
+      "learning_rate": 4.88746644641751e-05,
+      "loss": 1.0302,
+      "step": 12875
+    },
+    {
+      "epoch": 0.5212464589235127,
+      "grad_norm": 1.0182640552520752,
+      "learning_rate": 4.88540161057196e-05,
+      "loss": 0.9953,
+      "step": 12880
+    },
+    {
+      "epoch": 0.5214488061513557,
+      "grad_norm": 1.1881667375564575,
+      "learning_rate": 4.8833367747264096e-05,
+      "loss": 0.9945,
+      "step": 12885
+    },
+    {
+      "epoch": 0.5216511533791987,
+      "grad_norm": 1.1521013975143433,
+      "learning_rate": 4.8812719388808594e-05,
+      "loss": 1.0438,
+      "step": 12890
+    },
+    {
+      "epoch": 0.5218535006070417,
+      "grad_norm": 1.1074672937393188,
+      "learning_rate": 4.8792071030353084e-05,
+      "loss": 1.0017,
+      "step": 12895
+    },
+    {
+      "epoch": 0.5220558478348847,
+      "grad_norm": 1.1211540699005127,
+      "learning_rate": 4.877142267189759e-05,
+      "loss": 1.0537,
+      "step": 12900
+    },
+    {
+      "epoch": 0.5222581950627276,
+      "grad_norm": 1.1508073806762695,
+      "learning_rate": 4.8750774313442085e-05,
+      "loss": 0.978,
+      "step": 12905
+    },
+    {
+      "epoch": 0.5224605422905706,
+      "grad_norm": 1.1583364009857178,
+      "learning_rate": 4.873012595498658e-05,
+      "loss": 1.0321,
+      "step": 12910
+    },
+    {
+      "epoch": 0.5226628895184136,
+      "grad_norm": 1.2299565076828003,
+      "learning_rate": 4.870947759653108e-05,
+      "loss": 0.9972,
+      "step": 12915
+    },
+    {
+      "epoch": 0.5228652367462566,
+      "grad_norm": 1.1356695890426636,
+      "learning_rate": 4.868882923807557e-05,
+      "loss": 1.0134,
+      "step": 12920
+    },
+    {
+      "epoch": 0.5230675839740996,
+      "grad_norm": 1.2544360160827637,
+      "learning_rate": 4.8668180879620075e-05,
+      "loss": 1.0036,
+      "step": 12925
+    },
+    {
+      "epoch": 0.5232699312019425,
+      "grad_norm": 1.1607639789581299,
+      "learning_rate": 4.8647532521164565e-05,
+      "loss": 0.9904,
+      "step": 12930
+    },
+    {
+      "epoch": 0.5234722784297855,
+      "grad_norm": 1.2394320964813232,
+      "learning_rate": 4.862688416270907e-05,
+      "loss": 0.9547,
+      "step": 12935
+    },
+    {
+      "epoch": 0.5236746256576285,
+      "grad_norm": 1.1878379583358765,
+      "learning_rate": 4.8606235804253567e-05,
+      "loss": 0.9685,
+      "step": 12940
+    },
+    {
+      "epoch": 0.5238769728854714,
+      "grad_norm": 1.2129117250442505,
+      "learning_rate": 4.858558744579806e-05,
+      "loss": 1.0465,
+      "step": 12945
+    },
+    {
+      "epoch": 0.5240793201133145,
+      "grad_norm": 1.0647077560424805,
+      "learning_rate": 4.856493908734256e-05,
+      "loss": 0.9771,
+      "step": 12950
+    },
+    {
+      "epoch": 0.5242816673411574,
+      "grad_norm": 1.1161081790924072,
+      "learning_rate": 4.854429072888705e-05,
+      "loss": 1.0098,
+      "step": 12955
+    },
+    {
+      "epoch": 0.5244840145690004,
+      "grad_norm": 1.1606014966964722,
+      "learning_rate": 4.8523642370431556e-05,
+      "loss": 1.032,
+      "step": 12960
+    },
+    {
+      "epoch": 0.5246863617968434,
+      "grad_norm": 1.1828463077545166,
+      "learning_rate": 4.8502994011976046e-05,
+      "loss": 0.999,
+      "step": 12965
+    },
+    {
+      "epoch": 0.5248887090246863,
+      "grad_norm": 1.135551929473877,
+      "learning_rate": 4.848234565352055e-05,
+      "loss": 0.9996,
+      "step": 12970
+    },
+    {
+      "epoch": 0.5250910562525294,
+      "grad_norm": 1.2061209678649902,
+      "learning_rate": 4.846169729506505e-05,
+      "loss": 0.9967,
+      "step": 12975
+    },
+    {
+      "epoch": 0.5252934034803723,
+      "grad_norm": 1.1298943758010864,
+      "learning_rate": 4.844104893660954e-05,
+      "loss": 0.9926,
+      "step": 12980
+    },
+    {
+      "epoch": 0.5254957507082153,
+      "grad_norm": 1.1890193223953247,
+      "learning_rate": 4.842040057815404e-05,
+      "loss": 1.0114,
+      "step": 12985
+    },
+    {
+      "epoch": 0.5256980979360583,
+      "grad_norm": 1.0580012798309326,
+      "learning_rate": 4.839975221969853e-05,
+      "loss": 0.9898,
+      "step": 12990
+    },
+    {
+      "epoch": 0.5259004451639012,
+      "grad_norm": 1.1724084615707397,
+      "learning_rate": 4.837910386124304e-05,
+      "loss": 1.0239,
+      "step": 12995
+    },
+    {
+      "epoch": 0.5261027923917442,
+      "grad_norm": 1.0939544439315796,
+      "learning_rate": 4.8358455502787534e-05,
+      "loss": 1.073,
+      "step": 13000
+    },
+    {
+      "epoch": 0.5263051396195872,
+      "grad_norm": 1.1149251461029053,
+      "learning_rate": 4.8337807144332025e-05,
+      "loss": 0.9969,
+      "step": 13005
+    },
+    {
+      "epoch": 0.5265074868474302,
+      "grad_norm": 1.157451868057251,
+      "learning_rate": 4.831715878587653e-05,
+      "loss": 0.9862,
+      "step": 13010
+    },
+    {
+      "epoch": 0.5267098340752732,
+      "grad_norm": 1.1021413803100586,
+      "learning_rate": 4.829651042742102e-05,
+      "loss": 0.9423,
+      "step": 13015
+    },
+    {
+      "epoch": 0.5269121813031161,
+      "grad_norm": 1.1386486291885376,
+      "learning_rate": 4.827586206896552e-05,
+      "loss": 0.95,
+      "step": 13020
+    },
+    {
+      "epoch": 0.5271145285309591,
+      "grad_norm": 1.3019928932189941,
+      "learning_rate": 4.8255213710510014e-05,
+      "loss": 0.9973,
+      "step": 13025
+    },
+    {
+      "epoch": 0.5273168757588022,
+      "grad_norm": 1.1542195081710815,
+      "learning_rate": 4.823456535205451e-05,
+      "loss": 0.973,
+      "step": 13030
+    },
+    {
+      "epoch": 0.5275192229866451,
+      "grad_norm": 1.1593953371047974,
+      "learning_rate": 4.8213916993599015e-05,
+      "loss": 0.9795,
+      "step": 13035
+    },
+    {
+      "epoch": 0.5277215702144881,
+      "grad_norm": 1.2002090215682983,
+      "learning_rate": 4.8193268635143506e-05,
+      "loss": 1.0046,
+      "step": 13040
+    },
+    {
+      "epoch": 0.527923917442331,
+      "grad_norm": 1.2281185388565063,
+      "learning_rate": 4.817262027668801e-05,
+      "loss": 1.0708,
+      "step": 13045
+    },
+    {
+      "epoch": 0.528126264670174,
+      "grad_norm": 1.063997507095337,
+      "learning_rate": 4.81519719182325e-05,
+      "loss": 1.0381,
+      "step": 13050
+    },
+    {
+      "epoch": 0.5283286118980169,
+      "grad_norm": 1.196027159690857,
+      "learning_rate": 4.8131323559777e-05,
+      "loss": 0.9978,
+      "step": 13055
+    },
+    {
+      "epoch": 0.52853095912586,
+      "grad_norm": 1.3122018575668335,
+      "learning_rate": 4.8110675201321495e-05,
+      "loss": 0.9686,
+      "step": 13060
+    },
+    {
+      "epoch": 0.528733306353703,
+      "grad_norm": 1.0822436809539795,
+      "learning_rate": 4.809002684286599e-05,
+      "loss": 1.0321,
+      "step": 13065
+    },
+    {
+      "epoch": 0.5289356535815459,
+      "grad_norm": 1.1360067129135132,
+      "learning_rate": 4.8069378484410496e-05,
+      "loss": 0.9818,
+      "step": 13070
+    },
+    {
+      "epoch": 0.5291380008093889,
+      "grad_norm": 1.1602964401245117,
+      "learning_rate": 4.804873012595499e-05,
+      "loss": 0.9691,
+      "step": 13075
+    },
+    {
+      "epoch": 0.5293403480372318,
+      "grad_norm": 1.0667049884796143,
+      "learning_rate": 4.802808176749949e-05,
+      "loss": 0.9802,
+      "step": 13080
+    },
+    {
+      "epoch": 0.5295426952650749,
+      "grad_norm": 1.2503582239151,
+      "learning_rate": 4.800743340904398e-05,
+      "loss": 1.0567,
+      "step": 13085
+    },
+    {
+      "epoch": 0.5297450424929179,
+      "grad_norm": 1.201912522315979,
+      "learning_rate": 4.798678505058848e-05,
+      "loss": 1.0163,
+      "step": 13090
+    },
+    {
+      "epoch": 0.5299473897207608,
+      "grad_norm": 1.2826204299926758,
+      "learning_rate": 4.7966136692132976e-05,
+      "loss": 0.9855,
+      "step": 13095
+    },
+    {
+      "epoch": 0.5301497369486038,
+      "grad_norm": 1.1695324182510376,
+      "learning_rate": 4.794548833367747e-05,
+      "loss": 0.9906,
+      "step": 13100
+    },
+    {
+      "epoch": 0.5303520841764467,
+      "grad_norm": 1.195469617843628,
+      "learning_rate": 4.792483997522198e-05,
+      "loss": 1.0072,
+      "step": 13105
+    },
+    {
+      "epoch": 0.5305544314042897,
+      "grad_norm": 1.1526228189468384,
+      "learning_rate": 4.790419161676647e-05,
+      "loss": 0.9963,
+      "step": 13110
+    },
+    {
+      "epoch": 0.5307567786321328,
+      "grad_norm": 1.158942699432373,
+      "learning_rate": 4.7883543258310965e-05,
+      "loss": 1.0502,
+      "step": 13115
+    },
+    {
+      "epoch": 0.5309591258599757,
+      "grad_norm": 1.2206699848175049,
+      "learning_rate": 4.786289489985546e-05,
+      "loss": 1.0469,
+      "step": 13120
+    },
+    {
+      "epoch": 0.5311614730878187,
+      "grad_norm": 1.1328481435775757,
+      "learning_rate": 4.784224654139996e-05,
+      "loss": 0.976,
+      "step": 13125
+    },
+    {
+      "epoch": 0.5313638203156617,
+      "grad_norm": 1.1425429582595825,
+      "learning_rate": 4.782159818294446e-05,
+      "loss": 1.0228,
+      "step": 13130
+    },
+    {
+      "epoch": 0.5315661675435046,
+      "grad_norm": 1.1963837146759033,
+      "learning_rate": 4.7800949824488954e-05,
+      "loss": 0.9743,
+      "step": 13135
+    },
+    {
+      "epoch": 0.5317685147713477,
+      "grad_norm": 1.151746153831482,
+      "learning_rate": 4.778030146603345e-05,
+      "loss": 1.0042,
+      "step": 13140
+    },
+    {
+      "epoch": 0.5319708619991906,
+      "grad_norm": 1.2935644388198853,
+      "learning_rate": 4.775965310757795e-05,
+      "loss": 1.0134,
+      "step": 13145
+    },
+    {
+      "epoch": 0.5321732092270336,
+      "grad_norm": 1.2662339210510254,
+      "learning_rate": 4.7739004749122446e-05,
+      "loss": 1.0167,
+      "step": 13150
+    },
+    {
+      "epoch": 0.5323755564548766,
+      "grad_norm": 1.1560618877410889,
+      "learning_rate": 4.7718356390666944e-05,
+      "loss": 0.9559,
+      "step": 13155
+    },
+    {
+      "epoch": 0.5325779036827195,
+      "grad_norm": 1.0124341249465942,
+      "learning_rate": 4.769770803221144e-05,
+      "loss": 1.009,
+      "step": 13160
+    },
+    {
+      "epoch": 0.5327802509105626,
+      "grad_norm": 1.2502763271331787,
+      "learning_rate": 4.767705967375594e-05,
+      "loss": 0.9563,
+      "step": 13165
+    },
+    {
+      "epoch": 0.5329825981384055,
+      "grad_norm": 1.1667861938476562,
+      "learning_rate": 4.7656411315300435e-05,
+      "loss": 1.0503,
+      "step": 13170
+    },
+    {
+      "epoch": 0.5331849453662485,
+      "grad_norm": 1.2142013311386108,
+      "learning_rate": 4.763576295684493e-05,
+      "loss": 0.9992,
+      "step": 13175
+    },
+    {
+      "epoch": 0.5333872925940915,
+      "grad_norm": 1.1926441192626953,
+      "learning_rate": 4.761511459838943e-05,
+      "loss": 0.9737,
+      "step": 13180
+    },
+    {
+      "epoch": 0.5335896398219344,
+      "grad_norm": 1.1604886054992676,
+      "learning_rate": 4.759446623993393e-05,
+      "loss": 1.0423,
+      "step": 13185
+    },
+    {
+      "epoch": 0.5337919870497774,
+      "grad_norm": 1.208500623703003,
+      "learning_rate": 4.7573817881478425e-05,
+      "loss": 0.9808,
+      "step": 13190
+    },
+    {
+      "epoch": 0.5339943342776204,
+      "grad_norm": 1.0896100997924805,
+      "learning_rate": 4.755316952302292e-05,
+      "loss": 0.9782,
+      "step": 13195
+    },
+    {
+      "epoch": 0.5341966815054634,
+      "grad_norm": 1.2205355167388916,
+      "learning_rate": 4.753252116456742e-05,
+      "loss": 1.0097,
+      "step": 13200
+    },
+    {
+      "epoch": 0.5343990287333064,
+      "grad_norm": 1.2296879291534424,
+      "learning_rate": 4.7511872806111917e-05,
+      "loss": 1.0112,
+      "step": 13205
+    },
+    {
+      "epoch": 0.5346013759611493,
+      "grad_norm": 1.1248998641967773,
+      "learning_rate": 4.7491224447656414e-05,
+      "loss": 0.9598,
+      "step": 13210
+    },
+    {
+      "epoch": 0.5348037231889923,
+      "grad_norm": 1.1658109426498413,
+      "learning_rate": 4.747057608920091e-05,
+      "loss": 1.0401,
+      "step": 13215
+    },
+    {
+      "epoch": 0.5350060704168353,
+      "grad_norm": 1.14699387550354,
+      "learning_rate": 4.744992773074541e-05,
+      "loss": 1.0084,
+      "step": 13220
+    },
+    {
+      "epoch": 0.5352084176446783,
+      "grad_norm": 1.141981601715088,
+      "learning_rate": 4.7429279372289906e-05,
+      "loss": 0.964,
+      "step": 13225
+    },
+    {
+      "epoch": 0.5354107648725213,
+      "grad_norm": 1.2816085815429688,
+      "learning_rate": 4.74086310138344e-05,
+      "loss": 1.0027,
+      "step": 13230
+    },
+    {
+      "epoch": 0.5356131121003642,
+      "grad_norm": 1.151843547821045,
+      "learning_rate": 4.73879826553789e-05,
+      "loss": 0.988,
+      "step": 13235
+    },
+    {
+      "epoch": 0.5358154593282072,
+      "grad_norm": 1.249070405960083,
+      "learning_rate": 4.73673342969234e-05,
+      "loss": 0.971,
+      "step": 13240
+    },
+    {
+      "epoch": 0.5360178065560501,
+      "grad_norm": 1.0401089191436768,
+      "learning_rate": 4.7346685938467895e-05,
+      "loss": 0.9856,
+      "step": 13245
+    },
+    {
+      "epoch": 0.5362201537838932,
+      "grad_norm": 1.2624151706695557,
+      "learning_rate": 4.732603758001239e-05,
+      "loss": 0.9866,
+      "step": 13250
+    },
+    {
+      "epoch": 0.5364225010117362,
+      "grad_norm": 1.4663299322128296,
+      "learning_rate": 4.730538922155689e-05,
+      "loss": 1.0089,
+      "step": 13255
+    },
+    {
+      "epoch": 0.5366248482395791,
+      "grad_norm": 1.167694330215454,
+      "learning_rate": 4.728474086310139e-05,
+      "loss": 0.9083,
+      "step": 13260
+    },
+    {
+      "epoch": 0.5368271954674221,
+      "grad_norm": 1.2311296463012695,
+      "learning_rate": 4.7264092504645884e-05,
+      "loss": 0.9911,
+      "step": 13265
+    },
+    {
+      "epoch": 0.537029542695265,
+      "grad_norm": 1.1577832698822021,
+      "learning_rate": 4.724344414619038e-05,
+      "loss": 0.9533,
+      "step": 13270
+    },
+    {
+      "epoch": 0.5372318899231081,
+      "grad_norm": 1.2128522396087646,
+      "learning_rate": 4.722279578773487e-05,
+      "loss": 0.997,
+      "step": 13275
+    },
+    {
+      "epoch": 0.5374342371509511,
+      "grad_norm": 1.1526645421981812,
+      "learning_rate": 4.7202147429279376e-05,
+      "loss": 0.9862,
+      "step": 13280
+    },
+    {
+      "epoch": 0.537636584378794,
+      "grad_norm": 1.2666778564453125,
+      "learning_rate": 4.718149907082387e-05,
+      "loss": 0.9863,
+      "step": 13285
+    },
+    {
+      "epoch": 0.537838931606637,
+      "grad_norm": 1.4724574089050293,
+      "learning_rate": 4.716085071236837e-05,
+      "loss": 0.982,
+      "step": 13290
+    },
+    {
+      "epoch": 0.5380412788344799,
+      "grad_norm": 1.155213713645935,
+      "learning_rate": 4.714020235391287e-05,
+      "loss": 0.9791,
+      "step": 13295
+    },
+    {
+      "epoch": 0.5382436260623229,
+      "grad_norm": 1.1476895809173584,
+      "learning_rate": 4.7119553995457365e-05,
+      "loss": 0.9988,
+      "step": 13300
+    },
+    {
+      "epoch": 0.538445973290166,
+      "grad_norm": 1.1798290014266968,
+      "learning_rate": 4.709890563700186e-05,
+      "loss": 1.0468,
+      "step": 13305
+    },
+    {
+      "epoch": 0.5386483205180089,
+      "grad_norm": 1.0469969511032104,
+      "learning_rate": 4.707825727854635e-05,
+      "loss": 0.9303,
+      "step": 13310
+    },
+    {
+      "epoch": 0.5388506677458519,
+      "grad_norm": 1.2639477252960205,
+      "learning_rate": 4.705760892009086e-05,
+      "loss": 0.9489,
+      "step": 13315
+    },
+    {
+      "epoch": 0.5390530149736948,
+      "grad_norm": 1.20005464553833,
+      "learning_rate": 4.7036960561635354e-05,
+      "loss": 1.0058,
+      "step": 13320
+    },
+    {
+      "epoch": 0.5392553622015378,
+      "grad_norm": 1.1812065839767456,
+      "learning_rate": 4.701631220317985e-05,
+      "loss": 0.9417,
+      "step": 13325
+    },
+    {
+      "epoch": 0.5394577094293809,
+      "grad_norm": 1.2369160652160645,
+      "learning_rate": 4.699566384472435e-05,
+      "loss": 1.0068,
+      "step": 13330
+    },
+    {
+      "epoch": 0.5396600566572238,
+      "grad_norm": 1.1857260465621948,
+      "learning_rate": 4.697501548626884e-05,
+      "loss": 0.9527,
+      "step": 13335
+    },
+    {
+      "epoch": 0.5398624038850668,
+      "grad_norm": 1.1948858499526978,
+      "learning_rate": 4.6954367127813344e-05,
+      "loss": 1.0241,
+      "step": 13340
+    },
+    {
+      "epoch": 0.5400647511129097,
+      "grad_norm": 1.1948617696762085,
+      "learning_rate": 4.6933718769357834e-05,
+      "loss": 0.9988,
+      "step": 13345
+    },
+    {
+      "epoch": 0.5402670983407527,
+      "grad_norm": 1.1652532815933228,
+      "learning_rate": 4.691307041090234e-05,
+      "loss": 1.051,
+      "step": 13350
+    },
+    {
+      "epoch": 0.5404694455685957,
+      "grad_norm": 1.1360259056091309,
+      "learning_rate": 4.6892422052446835e-05,
+      "loss": 0.965,
+      "step": 13355
+    },
+    {
+      "epoch": 0.5406717927964387,
+      "grad_norm": 1.151434302330017,
+      "learning_rate": 4.6871773693991326e-05,
+      "loss": 1.0157,
+      "step": 13360
+    },
+    {
+      "epoch": 0.5408741400242817,
+      "grad_norm": 1.175532579421997,
+      "learning_rate": 4.685112533553583e-05,
+      "loss": 0.9648,
+      "step": 13365
+    },
+    {
+      "epoch": 0.5410764872521246,
+      "grad_norm": 1.2540494203567505,
+      "learning_rate": 4.683047697708032e-05,
+      "loss": 1.0434,
+      "step": 13370
+    },
+    {
+      "epoch": 0.5412788344799676,
+      "grad_norm": 1.1384494304656982,
+      "learning_rate": 4.6809828618624825e-05,
+      "loss": 1.0127,
+      "step": 13375
+    },
+    {
+      "epoch": 0.5414811817078106,
+      "grad_norm": 1.1299713850021362,
+      "learning_rate": 4.678918026016932e-05,
+      "loss": 0.982,
+      "step": 13380
+    },
+    {
+      "epoch": 0.5416835289356536,
+      "grad_norm": 1.1605921983718872,
+      "learning_rate": 4.676853190171381e-05,
+      "loss": 1.0333,
+      "step": 13385
+    },
+    {
+      "epoch": 0.5418858761634966,
+      "grad_norm": 1.2055753469467163,
+      "learning_rate": 4.6747883543258316e-05,
+      "loss": 1.0047,
+      "step": 13390
+    },
+    {
+      "epoch": 0.5420882233913396,
+      "grad_norm": 1.1855299472808838,
+      "learning_rate": 4.672723518480281e-05,
+      "loss": 0.9953,
+      "step": 13395
+    },
+    {
+      "epoch": 0.5422905706191825,
+      "grad_norm": 1.2290223836898804,
+      "learning_rate": 4.670658682634731e-05,
+      "loss": 0.9913,
+      "step": 13400
+    },
+    {
+      "epoch": 0.5424929178470255,
+      "grad_norm": 1.1520344018936157,
+      "learning_rate": 4.66859384678918e-05,
+      "loss": 1.0972,
+      "step": 13405
+    },
+    {
+      "epoch": 0.5426952650748684,
+      "grad_norm": 1.0018000602722168,
+      "learning_rate": 4.66652901094363e-05,
+      "loss": 0.9873,
+      "step": 13410
+    },
+    {
+      "epoch": 0.5428976123027115,
+      "grad_norm": 1.1549125909805298,
+      "learning_rate": 4.66446417509808e-05,
+      "loss": 1.0019,
+      "step": 13415
+    },
+    {
+      "epoch": 0.5430999595305545,
+      "grad_norm": 1.1142404079437256,
+      "learning_rate": 4.6623993392525293e-05,
+      "loss": 0.9615,
+      "step": 13420
+    },
+    {
+      "epoch": 0.5433023067583974,
+      "grad_norm": 1.4011167287826538,
+      "learning_rate": 4.66033450340698e-05,
+      "loss": 1.0121,
+      "step": 13425
+    },
+    {
+      "epoch": 0.5435046539862404,
+      "grad_norm": 1.1823188066482544,
+      "learning_rate": 4.658269667561429e-05,
+      "loss": 0.9744,
+      "step": 13430
+    },
+    {
+      "epoch": 0.5437070012140833,
+      "grad_norm": 1.189070463180542,
+      "learning_rate": 4.656204831715879e-05,
+      "loss": 1.0399,
+      "step": 13435
+    },
+    {
+      "epoch": 0.5439093484419264,
+      "grad_norm": 1.09059476852417,
+      "learning_rate": 4.654139995870328e-05,
+      "loss": 1.0146,
+      "step": 13440
+    },
+    {
+      "epoch": 0.5441116956697694,
+      "grad_norm": 1.1695646047592163,
+      "learning_rate": 4.652075160024778e-05,
+      "loss": 0.9673,
+      "step": 13445
+    },
+    {
+      "epoch": 0.5443140428976123,
+      "grad_norm": 1.3173099756240845,
+      "learning_rate": 4.6500103241792284e-05,
+      "loss": 1.0139,
+      "step": 13450
+    },
+    {
+      "epoch": 0.5445163901254553,
+      "grad_norm": 1.3592511415481567,
+      "learning_rate": 4.6479454883336775e-05,
+      "loss": 0.9623,
+      "step": 13455
+    },
+    {
+      "epoch": 0.5447187373532982,
+      "grad_norm": 1.2108453512191772,
+      "learning_rate": 4.645880652488128e-05,
+      "loss": 1.0226,
+      "step": 13460
+    },
+    {
+      "epoch": 0.5449210845811412,
+      "grad_norm": 1.280344843864441,
+      "learning_rate": 4.643815816642577e-05,
+      "loss": 0.9857,
+      "step": 13465
+    },
+    {
+      "epoch": 0.5451234318089843,
+      "grad_norm": 1.1430131196975708,
+      "learning_rate": 4.6417509807970266e-05,
+      "loss": 0.9474,
+      "step": 13470
+    },
+    {
+      "epoch": 0.5453257790368272,
+      "grad_norm": 1.2163547277450562,
+      "learning_rate": 4.6396861449514764e-05,
+      "loss": 0.9842,
+      "step": 13475
+    },
+    {
+      "epoch": 0.5455281262646702,
+      "grad_norm": 1.1540495157241821,
+      "learning_rate": 4.637621309105926e-05,
+      "loss": 0.9969,
+      "step": 13480
+    },
+    {
+      "epoch": 0.5457304734925131,
+      "grad_norm": 1.2115650177001953,
+      "learning_rate": 4.6355564732603765e-05,
+      "loss": 0.981,
+      "step": 13485
+    },
+    {
+      "epoch": 0.5459328207203561,
+      "grad_norm": 1.1425132751464844,
+      "learning_rate": 4.6334916374148256e-05,
+      "loss": 0.9518,
+      "step": 13490
+    },
+    {
+      "epoch": 0.5461351679481992,
+      "grad_norm": 1.1303656101226807,
+      "learning_rate": 4.631426801569275e-05,
+      "loss": 1.0307,
+      "step": 13495
+    },
+    {
+      "epoch": 0.5463375151760421,
+      "grad_norm": 1.317814826965332,
+      "learning_rate": 4.629361965723725e-05,
+      "loss": 1.0072,
+      "step": 13500
+    },
+    {
+      "epoch": 0.5465398624038851,
+      "grad_norm": 1.2411702871322632,
+      "learning_rate": 4.627297129878175e-05,
+      "loss": 0.9987,
+      "step": 13505
+    },
+    {
+      "epoch": 0.546742209631728,
+      "grad_norm": 1.2443106174468994,
+      "learning_rate": 4.6252322940326245e-05,
+      "loss": 0.9274,
+      "step": 13510
+    },
+    {
+      "epoch": 0.546944556859571,
+      "grad_norm": 1.24956214427948,
+      "learning_rate": 4.623167458187074e-05,
+      "loss": 1.0074,
+      "step": 13515
+    },
+    {
+      "epoch": 0.547146904087414,
+      "grad_norm": 1.0481141805648804,
+      "learning_rate": 4.621102622341524e-05,
+      "loss": 0.9905,
+      "step": 13520
+    },
+    {
+      "epoch": 0.547349251315257,
+      "grad_norm": 1.0809906721115112,
+      "learning_rate": 4.619037786495974e-05,
+      "loss": 1.0201,
+      "step": 13525
+    },
+    {
+      "epoch": 0.5475515985431,
+      "grad_norm": 1.2640081644058228,
+      "learning_rate": 4.6169729506504234e-05,
+      "loss": 0.9462,
+      "step": 13530
+    },
+    {
+      "epoch": 0.5477539457709429,
+      "grad_norm": 1.1775264739990234,
+      "learning_rate": 4.614908114804873e-05,
+      "loss": 0.9548,
+      "step": 13535
+    },
+    {
+      "epoch": 0.5479562929987859,
+      "grad_norm": 1.2151538133621216,
+      "learning_rate": 4.612843278959323e-05,
+      "loss": 0.9686,
+      "step": 13540
+    },
+    {
+      "epoch": 0.5481586402266289,
+      "grad_norm": 1.1137951612472534,
+      "learning_rate": 4.610778443113773e-05,
+      "loss": 0.9793,
+      "step": 13545
+    },
+    {
+      "epoch": 0.5483609874544719,
+      "grad_norm": 1.1605082750320435,
+      "learning_rate": 4.608713607268222e-05,
+      "loss": 0.9889,
+      "step": 13550
+    },
+    {
+      "epoch": 0.5485633346823149,
+      "grad_norm": 1.1585845947265625,
+      "learning_rate": 4.606648771422672e-05,
+      "loss": 1.0128,
+      "step": 13555
+    },
+    {
+      "epoch": 0.5487656819101578,
+      "grad_norm": 1.0791916847229004,
+      "learning_rate": 4.604583935577122e-05,
+      "loss": 0.9795,
+      "step": 13560
+    },
+    {
+      "epoch": 0.5489680291380008,
+      "grad_norm": 1.255650520324707,
+      "learning_rate": 4.6025190997315715e-05,
+      "loss": 0.9925,
+      "step": 13565
+    },
+    {
+      "epoch": 0.5491703763658438,
+      "grad_norm": 1.1521662473678589,
+      "learning_rate": 4.600454263886021e-05,
+      "loss": 1.0126,
+      "step": 13570
+    },
+    {
+      "epoch": 0.5493727235936867,
+      "grad_norm": 1.1036776304244995,
+      "learning_rate": 4.598389428040471e-05,
+      "loss": 1.0461,
+      "step": 13575
+    },
+    {
+      "epoch": 0.5495750708215298,
+      "grad_norm": 1.139861822128296,
+      "learning_rate": 4.596324592194921e-05,
+      "loss": 0.9958,
+      "step": 13580
+    },
+    {
+      "epoch": 0.5497774180493727,
+      "grad_norm": 1.1681941747665405,
+      "learning_rate": 4.5942597563493704e-05,
+      "loss": 0.9831,
+      "step": 13585
+    },
+    {
+      "epoch": 0.5499797652772157,
+      "grad_norm": 1.1199458837509155,
+      "learning_rate": 4.59219492050382e-05,
+      "loss": 1.0242,
+      "step": 13590
+    },
+    {
+      "epoch": 0.5501821125050587,
+      "grad_norm": 1.1740652322769165,
+      "learning_rate": 4.59013008465827e-05,
+      "loss": 0.9959,
+      "step": 13595
+    },
+    {
+      "epoch": 0.5503844597329016,
+      "grad_norm": 1.0760498046875,
+      "learning_rate": 4.5880652488127196e-05,
+      "loss": 0.9597,
+      "step": 13600
+    },
+    {
+      "epoch": 0.5505868069607447,
+      "grad_norm": 1.0998018980026245,
+      "learning_rate": 4.5860004129671693e-05,
+      "loss": 1.0328,
+      "step": 13605
+    },
+    {
+      "epoch": 0.5507891541885876,
+      "grad_norm": 1.2137548923492432,
+      "learning_rate": 4.583935577121619e-05,
+      "loss": 1.0383,
+      "step": 13610
+    },
+    {
+      "epoch": 0.5509915014164306,
+      "grad_norm": 1.1961240768432617,
+      "learning_rate": 4.581870741276069e-05,
+      "loss": 0.9823,
+      "step": 13615
+    },
+    {
+      "epoch": 0.5511938486442736,
+      "grad_norm": 1.2069047689437866,
+      "learning_rate": 4.5798059054305185e-05,
+      "loss": 1.0187,
+      "step": 13620
+    },
+    {
+      "epoch": 0.5513961958721165,
+      "grad_norm": 1.1343979835510254,
+      "learning_rate": 4.577741069584968e-05,
+      "loss": 1.0032,
+      "step": 13625
+    },
+    {
+      "epoch": 0.5515985430999595,
+      "grad_norm": 1.1684554815292358,
+      "learning_rate": 4.575676233739418e-05,
+      "loss": 1.0025,
+      "step": 13630
+    },
+    {
+      "epoch": 0.5518008903278026,
+      "grad_norm": 1.1609210968017578,
+      "learning_rate": 4.573611397893868e-05,
+      "loss": 1.0775,
+      "step": 13635
+    },
+    {
+      "epoch": 0.5520032375556455,
+      "grad_norm": 1.1721374988555908,
+      "learning_rate": 4.5715465620483175e-05,
+      "loss": 0.9471,
+      "step": 13640
+    },
+    {
+      "epoch": 0.5522055847834885,
+      "grad_norm": 1.2115263938903809,
+      "learning_rate": 4.569481726202767e-05,
+      "loss": 1.0247,
+      "step": 13645
+    },
+    {
+      "epoch": 0.5524079320113314,
+      "grad_norm": 1.1860140562057495,
+      "learning_rate": 4.567416890357217e-05,
+      "loss": 1.0428,
+      "step": 13650
+    },
+    {
+      "epoch": 0.5526102792391744,
+      "grad_norm": 1.2308109998703003,
+      "learning_rate": 4.5653520545116666e-05,
+      "loss": 1.0178,
+      "step": 13655
+    },
+    {
+      "epoch": 0.5528126264670175,
+      "grad_norm": 1.2241133451461792,
+      "learning_rate": 4.5632872186661164e-05,
+      "loss": 1.0624,
+      "step": 13660
+    },
+    {
+      "epoch": 0.5530149736948604,
+      "grad_norm": 1.2302842140197754,
+      "learning_rate": 4.561222382820566e-05,
+      "loss": 0.9859,
+      "step": 13665
+    },
+    {
+      "epoch": 0.5532173209227034,
+      "grad_norm": 1.160462737083435,
+      "learning_rate": 4.559157546975016e-05,
+      "loss": 0.9593,
+      "step": 13670
+    },
+    {
+      "epoch": 0.5534196681505463,
+      "grad_norm": 1.130302906036377,
+      "learning_rate": 4.5570927111294656e-05,
+      "loss": 1.0348,
+      "step": 13675
+    },
+    {
+      "epoch": 0.5536220153783893,
+      "grad_norm": 1.1454074382781982,
+      "learning_rate": 4.555027875283915e-05,
+      "loss": 1.0393,
+      "step": 13680
+    },
+    {
+      "epoch": 0.5538243626062322,
+      "grad_norm": 1.1717177629470825,
+      "learning_rate": 4.552963039438365e-05,
+      "loss": 1.0357,
+      "step": 13685
+    },
+    {
+      "epoch": 0.5540267098340753,
+      "grad_norm": 1.2080857753753662,
+      "learning_rate": 4.550898203592814e-05,
+      "loss": 1.0632,
+      "step": 13690
+    },
+    {
+      "epoch": 0.5542290570619183,
+      "grad_norm": 1.1342897415161133,
+      "learning_rate": 4.5488333677472645e-05,
+      "loss": 0.9912,
+      "step": 13695
+    },
+    {
+      "epoch": 0.5544314042897612,
+      "grad_norm": 1.120079517364502,
+      "learning_rate": 4.546768531901714e-05,
+      "loss": 1.0341,
+      "step": 13700
+    },
+    {
+      "epoch": 0.5546337515176042,
+      "grad_norm": 1.1466180086135864,
+      "learning_rate": 4.544703696056164e-05,
+      "loss": 1.0419,
+      "step": 13705
+    },
+    {
+      "epoch": 0.5548360987454471,
+      "grad_norm": 1.118067741394043,
+      "learning_rate": 4.542638860210614e-05,
+      "loss": 1.0625,
+      "step": 13710
+    },
+    {
+      "epoch": 0.5550384459732902,
+      "grad_norm": 1.2374582290649414,
+      "learning_rate": 4.540574024365063e-05,
+      "loss": 1.021,
+      "step": 13715
+    },
+    {
+      "epoch": 0.5552407932011332,
+      "grad_norm": 1.3281618356704712,
+      "learning_rate": 4.538509188519513e-05,
+      "loss": 0.9659,
+      "step": 13720
+    },
+    {
+      "epoch": 0.5554431404289761,
+      "grad_norm": 1.2413415908813477,
+      "learning_rate": 4.536444352673962e-05,
+      "loss": 1.0318,
+      "step": 13725
+    },
+    {
+      "epoch": 0.5556454876568191,
+      "grad_norm": 1.1204394102096558,
+      "learning_rate": 4.5343795168284126e-05,
+      "loss": 0.9587,
+      "step": 13730
+    },
+    {
+      "epoch": 0.555847834884662,
+      "grad_norm": 1.3070437908172607,
+      "learning_rate": 4.532314680982862e-05,
+      "loss": 1.0923,
+      "step": 13735
+    },
+    {
+      "epoch": 0.556050182112505,
+      "grad_norm": 1.3235679864883423,
+      "learning_rate": 4.5302498451373114e-05,
+      "loss": 1.0561,
+      "step": 13740
+    },
+    {
+      "epoch": 0.5562525293403481,
+      "grad_norm": 1.1827317476272583,
+      "learning_rate": 4.528185009291762e-05,
+      "loss": 1.0212,
+      "step": 13745
+    },
+    {
+      "epoch": 0.556454876568191,
+      "grad_norm": 1.1174043416976929,
+      "learning_rate": 4.526120173446211e-05,
+      "loss": 1.0995,
+      "step": 13750
+    },
+    {
+      "epoch": 0.556657223796034,
+      "grad_norm": 1.128941535949707,
+      "learning_rate": 4.524055337600661e-05,
+      "loss": 1.0136,
+      "step": 13755
+    },
+    {
+      "epoch": 0.556859571023877,
+      "grad_norm": 1.1892149448394775,
+      "learning_rate": 4.521990501755111e-05,
+      "loss": 1.0523,
+      "step": 13760
+    },
+    {
+      "epoch": 0.5570619182517199,
+      "grad_norm": 1.0689456462860107,
+      "learning_rate": 4.519925665909561e-05,
+      "loss": 1.0007,
+      "step": 13765
+    },
+    {
+      "epoch": 0.557264265479563,
+      "grad_norm": 1.2329188585281372,
+      "learning_rate": 4.5178608300640104e-05,
+      "loss": 1.0157,
+      "step": 13770
+    },
+    {
+      "epoch": 0.5574666127074059,
+      "grad_norm": 1.202246069908142,
+      "learning_rate": 4.5157959942184595e-05,
+      "loss": 1.0,
+      "step": 13775
+    },
+    {
+      "epoch": 0.5576689599352489,
+      "grad_norm": 1.1702176332473755,
+      "learning_rate": 4.51373115837291e-05,
+      "loss": 1.0001,
+      "step": 13780
+    },
+    {
+      "epoch": 0.5578713071630919,
+      "grad_norm": 1.073132038116455,
+      "learning_rate": 4.511666322527359e-05,
+      "loss": 1.012,
+      "step": 13785
+    },
+    {
+      "epoch": 0.5580736543909348,
+      "grad_norm": 1.2309406995773315,
+      "learning_rate": 4.5096014866818093e-05,
+      "loss": 1.0682,
+      "step": 13790
+    },
+    {
+      "epoch": 0.5582760016187778,
+      "grad_norm": 1.0988270044326782,
+      "learning_rate": 4.507536650836259e-05,
+      "loss": 1.0334,
+      "step": 13795
+    },
+    {
+      "epoch": 0.5584783488466208,
+      "grad_norm": 1.1950405836105347,
+      "learning_rate": 4.505471814990708e-05,
+      "loss": 0.9594,
+      "step": 13800
+    },
+    {
+      "epoch": 0.5586806960744638,
+      "grad_norm": 1.2276102304458618,
+      "learning_rate": 4.5034069791451585e-05,
+      "loss": 0.9512,
+      "step": 13805
+    },
+    {
+      "epoch": 0.5588830433023068,
+      "grad_norm": 1.2122454643249512,
+      "learning_rate": 4.5013421432996076e-05,
+      "loss": 0.9706,
+      "step": 13810
+    },
+    {
+      "epoch": 0.5590853905301497,
+      "grad_norm": 1.043339490890503,
+      "learning_rate": 4.499277307454058e-05,
+      "loss": 0.9137,
+      "step": 13815
+    },
+    {
+      "epoch": 0.5592877377579927,
+      "grad_norm": 1.2296773195266724,
+      "learning_rate": 4.497212471608507e-05,
+      "loss": 1.0002,
+      "step": 13820
+    },
+    {
+      "epoch": 0.5594900849858357,
+      "grad_norm": 1.1996498107910156,
+      "learning_rate": 4.495147635762957e-05,
+      "loss": 0.9988,
+      "step": 13825
+    },
+    {
+      "epoch": 0.5596924322136787,
+      "grad_norm": 1.2913368940353394,
+      "learning_rate": 4.493082799917407e-05,
+      "loss": 1.0247,
+      "step": 13830
+    },
+    {
+      "epoch": 0.5598947794415217,
+      "grad_norm": 1.0832664966583252,
+      "learning_rate": 4.491017964071856e-05,
+      "loss": 0.962,
+      "step": 13835
+    },
+    {
+      "epoch": 0.5600971266693646,
+      "grad_norm": 1.238834261894226,
+      "learning_rate": 4.4889531282263066e-05,
+      "loss": 0.9866,
+      "step": 13840
+    },
+    {
+      "epoch": 0.5602994738972076,
+      "grad_norm": 1.069636344909668,
+      "learning_rate": 4.486888292380756e-05,
+      "loss": 0.9986,
+      "step": 13845
+    },
+    {
+      "epoch": 0.5605018211250505,
+      "grad_norm": 1.164320707321167,
+      "learning_rate": 4.4848234565352054e-05,
+      "loss": 0.9916,
+      "step": 13850
+    },
+    {
+      "epoch": 0.5607041683528936,
+      "grad_norm": 1.1442533731460571,
+      "learning_rate": 4.482758620689655e-05,
+      "loss": 1.0067,
+      "step": 13855
+    },
+    {
+      "epoch": 0.5609065155807366,
+      "grad_norm": 1.1331062316894531,
+      "learning_rate": 4.480693784844105e-05,
+      "loss": 0.9219,
+      "step": 13860
+    },
+    {
+      "epoch": 0.5611088628085795,
+      "grad_norm": 1.0536301136016846,
+      "learning_rate": 4.478628948998555e-05,
+      "loss": 1.0004,
+      "step": 13865
+    },
+    {
+      "epoch": 0.5613112100364225,
+      "grad_norm": 1.2179696559906006,
+      "learning_rate": 4.4765641131530043e-05,
+      "loss": 1.0181,
+      "step": 13870
+    },
+    {
+      "epoch": 0.5615135572642654,
+      "grad_norm": 1.1594337224960327,
+      "learning_rate": 4.474499277307454e-05,
+      "loss": 0.9524,
+      "step": 13875
+    },
+    {
+      "epoch": 0.5617159044921085,
+      "grad_norm": 1.311442494392395,
+      "learning_rate": 4.472434441461904e-05,
+      "loss": 1.0565,
+      "step": 13880
+    },
+    {
+      "epoch": 0.5619182517199515,
+      "grad_norm": 1.1441835165023804,
+      "learning_rate": 4.4703696056163535e-05,
+      "loss": 0.9707,
+      "step": 13885
+    },
+    {
+      "epoch": 0.5621205989477944,
+      "grad_norm": 2.9216387271881104,
+      "learning_rate": 4.468304769770803e-05,
+      "loss": 1.0217,
+      "step": 13890
+    },
+    {
+      "epoch": 0.5623229461756374,
+      "grad_norm": 1.1903290748596191,
+      "learning_rate": 4.466239933925253e-05,
+      "loss": 1.0293,
+      "step": 13895
+    },
+    {
+      "epoch": 0.5625252934034803,
+      "grad_norm": 1.1540318727493286,
+      "learning_rate": 4.4641750980797034e-05,
+      "loss": 0.9939,
+      "step": 13900
+    },
+    {
+      "epoch": 0.5627276406313233,
+      "grad_norm": 1.2955759763717651,
+      "learning_rate": 4.4621102622341525e-05,
+      "loss": 0.979,
+      "step": 13905
+    },
+    {
+      "epoch": 0.5629299878591664,
+      "grad_norm": 1.2096494436264038,
+      "learning_rate": 4.460045426388602e-05,
+      "loss": 0.9966,
+      "step": 13910
+    },
+    {
+      "epoch": 0.5631323350870093,
+      "grad_norm": 1.1765676736831665,
+      "learning_rate": 4.457980590543052e-05,
+      "loss": 0.998,
+      "step": 13915
+    },
+    {
+      "epoch": 0.5633346823148523,
+      "grad_norm": 1.1927794218063354,
+      "learning_rate": 4.4559157546975016e-05,
+      "loss": 1.0851,
+      "step": 13920
+    },
+    {
+      "epoch": 0.5635370295426952,
+      "grad_norm": 1.1235953569412231,
+      "learning_rate": 4.453850918851952e-05,
+      "loss": 0.9892,
+      "step": 13925
+    },
+    {
+      "epoch": 0.5637393767705382,
+      "grad_norm": 1.101344347000122,
+      "learning_rate": 4.451786083006401e-05,
+      "loss": 0.9586,
+      "step": 13930
+    },
+    {
+      "epoch": 0.5639417239983813,
+      "grad_norm": 1.2288463115692139,
+      "learning_rate": 4.449721247160851e-05,
+      "loss": 0.9964,
+      "step": 13935
+    },
+    {
+      "epoch": 0.5641440712262242,
+      "grad_norm": 1.2437199354171753,
+      "learning_rate": 4.4476564113153006e-05,
+      "loss": 1.0403,
+      "step": 13940
+    },
+    {
+      "epoch": 0.5643464184540672,
+      "grad_norm": 1.0780836343765259,
+      "learning_rate": 4.44559157546975e-05,
+      "loss": 1.0403,
+      "step": 13945
+    },
+    {
+      "epoch": 0.5645487656819101,
+      "grad_norm": 1.2592697143554688,
+      "learning_rate": 4.4435267396242e-05,
+      "loss": 1.0537,
+      "step": 13950
+    },
+    {
+      "epoch": 0.5647511129097531,
+      "grad_norm": 1.2178301811218262,
+      "learning_rate": 4.44146190377865e-05,
+      "loss": 1.0226,
+      "step": 13955
+    },
+    {
+      "epoch": 0.5649534601375961,
+      "grad_norm": 1.1477197408676147,
+      "learning_rate": 4.4393970679330995e-05,
+      "loss": 0.9814,
+      "step": 13960
+    },
+    {
+      "epoch": 0.5651558073654391,
+      "grad_norm": 1.1132920980453491,
+      "learning_rate": 4.437332232087549e-05,
+      "loss": 0.9875,
+      "step": 13965
+    },
+    {
+      "epoch": 0.5653581545932821,
+      "grad_norm": 1.2448670864105225,
+      "learning_rate": 4.435267396241999e-05,
+      "loss": 1.0072,
+      "step": 13970
+    },
+    {
+      "epoch": 0.565560501821125,
+      "grad_norm": 1.215716004371643,
+      "learning_rate": 4.433202560396449e-05,
+      "loss": 0.9826,
+      "step": 13975
+    },
+    {
+      "epoch": 0.565762849048968,
+      "grad_norm": 1.107932686805725,
+      "learning_rate": 4.4311377245508984e-05,
+      "loss": 0.9905,
+      "step": 13980
+    },
+    {
+      "epoch": 0.565965196276811,
+      "grad_norm": 1.1962240934371948,
+      "learning_rate": 4.429072888705348e-05,
+      "loss": 0.9612,
+      "step": 13985
+    },
+    {
+      "epoch": 0.566167543504654,
+      "grad_norm": 1.2366198301315308,
+      "learning_rate": 4.427008052859798e-05,
+      "loss": 0.968,
+      "step": 13990
+    },
+    {
+      "epoch": 0.566369890732497,
+      "grad_norm": 1.2942988872528076,
+      "learning_rate": 4.4249432170142476e-05,
+      "loss": 0.9525,
+      "step": 13995
+    },
+    {
+      "epoch": 0.56657223796034,
+      "grad_norm": 1.1409313678741455,
+      "learning_rate": 4.422878381168697e-05,
+      "loss": 1.0099,
+      "step": 14000
+    },
+    {
+      "epoch": 0.5667745851881829,
+      "grad_norm": 1.1154898405075073,
+      "learning_rate": 4.420813545323147e-05,
+      "loss": 1.0386,
+      "step": 14005
+    },
+    {
+      "epoch": 0.5669769324160259,
+      "grad_norm": 1.2833094596862793,
+      "learning_rate": 4.418748709477597e-05,
+      "loss": 0.9946,
+      "step": 14010
+    },
+    {
+      "epoch": 0.5671792796438688,
+      "grad_norm": 1.1053303480148315,
+      "learning_rate": 4.4166838736320465e-05,
+      "loss": 1.0222,
+      "step": 14015
+    },
+    {
+      "epoch": 0.5673816268717119,
+      "grad_norm": 1.2782630920410156,
+      "learning_rate": 4.414619037786496e-05,
+      "loss": 0.963,
+      "step": 14020
+    },
+    {
+      "epoch": 0.5675839740995549,
+      "grad_norm": 1.139122486114502,
+      "learning_rate": 4.412554201940946e-05,
+      "loss": 0.949,
+      "step": 14025
+    },
+    {
+      "epoch": 0.5677863213273978,
+      "grad_norm": 1.188439965248108,
+      "learning_rate": 4.410489366095396e-05,
+      "loss": 1.0107,
+      "step": 14030
+    },
+    {
+      "epoch": 0.5679886685552408,
+      "grad_norm": 1.1669381856918335,
+      "learning_rate": 4.4084245302498454e-05,
+      "loss": 0.9895,
+      "step": 14035
+    },
+    {
+      "epoch": 0.5681910157830837,
+      "grad_norm": 1.2100615501403809,
+      "learning_rate": 4.406359694404295e-05,
+      "loss": 0.9938,
+      "step": 14040
+    },
+    {
+      "epoch": 0.5683933630109268,
+      "grad_norm": 1.2322107553482056,
+      "learning_rate": 4.404294858558745e-05,
+      "loss": 1.0095,
+      "step": 14045
+    },
+    {
+      "epoch": 0.5685957102387698,
+      "grad_norm": 1.2963682413101196,
+      "learning_rate": 4.4022300227131946e-05,
+      "loss": 1.0473,
+      "step": 14050
+    },
+    {
+      "epoch": 0.5687980574666127,
+      "grad_norm": 1.2882816791534424,
+      "learning_rate": 4.4001651868676443e-05,
+      "loss": 0.9555,
+      "step": 14055
+    },
+    {
+      "epoch": 0.5690004046944557,
+      "grad_norm": 1.2242742776870728,
+      "learning_rate": 4.398100351022094e-05,
+      "loss": 0.9798,
+      "step": 14060
+    },
+    {
+      "epoch": 0.5692027519222986,
+      "grad_norm": 1.204200029373169,
+      "learning_rate": 4.396035515176544e-05,
+      "loss": 0.9869,
+      "step": 14065
+    },
+    {
+      "epoch": 0.5694050991501416,
+      "grad_norm": 1.1855534315109253,
+      "learning_rate": 4.393970679330993e-05,
+      "loss": 1.0247,
+      "step": 14070
+    },
+    {
+      "epoch": 0.5696074463779847,
+      "grad_norm": 1.0950124263763428,
+      "learning_rate": 4.391905843485443e-05,
+      "loss": 0.9799,
+      "step": 14075
+    },
+    {
+      "epoch": 0.5698097936058276,
+      "grad_norm": 1.1773353815078735,
+      "learning_rate": 4.389841007639893e-05,
+      "loss": 0.9764,
+      "step": 14080
+    },
+    {
+      "epoch": 0.5700121408336706,
+      "grad_norm": 1.2206952571868896,
+      "learning_rate": 4.387776171794343e-05,
+      "loss": 0.9877,
+      "step": 14085
+    },
+    {
+      "epoch": 0.5702144880615135,
+      "grad_norm": 1.174280047416687,
+      "learning_rate": 4.3857113359487924e-05,
+      "loss": 0.9823,
+      "step": 14090
+    },
+    {
+      "epoch": 0.5704168352893565,
+      "grad_norm": 1.272633671760559,
+      "learning_rate": 4.3836465001032415e-05,
+      "loss": 1.021,
+      "step": 14095
+    },
+    {
+      "epoch": 0.5706191825171996,
+      "grad_norm": 1.3584730625152588,
+      "learning_rate": 4.381581664257692e-05,
+      "loss": 1.0188,
+      "step": 14100
+    },
+    {
+      "epoch": 0.5708215297450425,
+      "grad_norm": 1.1411398649215698,
+      "learning_rate": 4.379516828412141e-05,
+      "loss": 0.9952,
+      "step": 14105
+    },
+    {
+      "epoch": 0.5710238769728855,
+      "grad_norm": 1.174756646156311,
+      "learning_rate": 4.3774519925665914e-05,
+      "loss": 1.0058,
+      "step": 14110
+    },
+    {
+      "epoch": 0.5712262242007284,
+      "grad_norm": 1.2773780822753906,
+      "learning_rate": 4.375387156721041e-05,
+      "loss": 0.9758,
+      "step": 14115
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 1.2203837633132935,
+      "learning_rate": 4.373322320875491e-05,
+      "loss": 1.0931,
+      "step": 14120
+    },
+    {
+      "epoch": 0.5716309186564144,
+      "grad_norm": 1.1752070188522339,
+      "learning_rate": 4.3712574850299406e-05,
+      "loss": 1.0002,
+      "step": 14125
+    },
+    {
+      "epoch": 0.5718332658842574,
+      "grad_norm": 1.1498125791549683,
+      "learning_rate": 4.3691926491843896e-05,
+      "loss": 0.9446,
+      "step": 14130
+    },
+    {
+      "epoch": 0.5720356131121004,
+      "grad_norm": 1.1896262168884277,
+      "learning_rate": 4.36712781333884e-05,
+      "loss": 1.0061,
+      "step": 14135
+    },
+    {
+      "epoch": 0.5722379603399433,
+      "grad_norm": 1.2011770009994507,
+      "learning_rate": 4.36506297749329e-05,
+      "loss": 1.0251,
+      "step": 14140
+    },
+    {
+      "epoch": 0.5724403075677863,
+      "grad_norm": 1.150048851966858,
+      "learning_rate": 4.3629981416477395e-05,
+      "loss": 1.025,
+      "step": 14145
+    },
+    {
+      "epoch": 0.5726426547956293,
+      "grad_norm": 1.1370536088943481,
+      "learning_rate": 4.360933305802189e-05,
+      "loss": 1.037,
+      "step": 14150
+    },
+    {
+      "epoch": 0.5728450020234723,
+      "grad_norm": 1.1036863327026367,
+      "learning_rate": 4.358868469956638e-05,
+      "loss": 0.9489,
+      "step": 14155
+    },
+    {
+      "epoch": 0.5730473492513153,
+      "grad_norm": 1.1035983562469482,
+      "learning_rate": 4.356803634111089e-05,
+      "loss": 0.9553,
+      "step": 14160
+    },
+    {
+      "epoch": 0.5732496964791582,
+      "grad_norm": 1.1876813173294067,
+      "learning_rate": 4.354738798265538e-05,
+      "loss": 1.0022,
+      "step": 14165
+    },
+    {
+      "epoch": 0.5734520437070012,
+      "grad_norm": 1.1236966848373413,
+      "learning_rate": 4.352673962419988e-05,
+      "loss": 0.9287,
+      "step": 14170
+    },
+    {
+      "epoch": 0.5736543909348442,
+      "grad_norm": 1.1244624853134155,
+      "learning_rate": 4.350609126574438e-05,
+      "loss": 0.9893,
+      "step": 14175
+    },
+    {
+      "epoch": 0.5738567381626871,
+      "grad_norm": 1.2417951822280884,
+      "learning_rate": 4.348544290728887e-05,
+      "loss": 1.0084,
+      "step": 14180
+    },
+    {
+      "epoch": 0.5740590853905302,
+      "grad_norm": 1.1004974842071533,
+      "learning_rate": 4.346479454883337e-05,
+      "loss": 1.0059,
+      "step": 14185
+    },
+    {
+      "epoch": 0.5742614326183731,
+      "grad_norm": 1.3367033004760742,
+      "learning_rate": 4.3444146190377864e-05,
+      "loss": 1.0446,
+      "step": 14190
+    },
+    {
+      "epoch": 0.5744637798462161,
+      "grad_norm": 1.2608211040496826,
+      "learning_rate": 4.342349783192237e-05,
+      "loss": 1.0296,
+      "step": 14195
+    },
+    {
+      "epoch": 0.5746661270740591,
+      "grad_norm": 1.2179604768753052,
+      "learning_rate": 4.340284947346686e-05,
+      "loss": 1.0519,
+      "step": 14200
+    },
+    {
+      "epoch": 0.574868474301902,
+      "grad_norm": 1.0648919343948364,
+      "learning_rate": 4.3382201115011356e-05,
+      "loss": 0.9879,
+      "step": 14205
+    },
+    {
+      "epoch": 0.5750708215297451,
+      "grad_norm": 1.1244484186172485,
+      "learning_rate": 4.336155275655586e-05,
+      "loss": 1.0345,
+      "step": 14210
+    },
+    {
+      "epoch": 0.575273168757588,
+      "grad_norm": 1.156739592552185,
+      "learning_rate": 4.334090439810035e-05,
+      "loss": 0.9823,
+      "step": 14215
+    },
+    {
+      "epoch": 0.575475515985431,
+      "grad_norm": 1.3055468797683716,
+      "learning_rate": 4.3320256039644854e-05,
+      "loss": 0.9907,
+      "step": 14220
+    },
+    {
+      "epoch": 0.575677863213274,
+      "grad_norm": 1.2654321193695068,
+      "learning_rate": 4.3299607681189345e-05,
+      "loss": 0.9654,
+      "step": 14225
+    },
+    {
+      "epoch": 0.5758802104411169,
+      "grad_norm": 1.2418204545974731,
+      "learning_rate": 4.327895932273385e-05,
+      "loss": 0.9957,
+      "step": 14230
+    },
+    {
+      "epoch": 0.5760825576689599,
+      "grad_norm": 1.2722078561782837,
+      "learning_rate": 4.325831096427834e-05,
+      "loss": 1.0861,
+      "step": 14235
+    },
+    {
+      "epoch": 0.576284904896803,
+      "grad_norm": 1.4188235998153687,
+      "learning_rate": 4.323766260582284e-05,
+      "loss": 0.9552,
+      "step": 14240
+    },
+    {
+      "epoch": 0.5764872521246459,
+      "grad_norm": 1.1043928861618042,
+      "learning_rate": 4.321701424736734e-05,
+      "loss": 0.9975,
+      "step": 14245
+    },
+    {
+      "epoch": 0.5766895993524889,
+      "grad_norm": 1.1651086807250977,
+      "learning_rate": 4.319636588891183e-05,
+      "loss": 0.9782,
+      "step": 14250
+    },
+    {
+      "epoch": 0.5768919465803318,
+      "grad_norm": 1.0888679027557373,
+      "learning_rate": 4.3175717530456335e-05,
+      "loss": 1.007,
+      "step": 14255
+    },
+    {
+      "epoch": 0.5770942938081748,
+      "grad_norm": 1.3613406419754028,
+      "learning_rate": 4.3155069172000826e-05,
+      "loss": 1.0085,
+      "step": 14260
+    },
+    {
+      "epoch": 0.5772966410360179,
+      "grad_norm": 1.208199381828308,
+      "learning_rate": 4.313442081354532e-05,
+      "loss": 1.019,
+      "step": 14265
+    },
+    {
+      "epoch": 0.5774989882638608,
+      "grad_norm": 1.3016966581344604,
+      "learning_rate": 4.311377245508982e-05,
+      "loss": 1.0349,
+      "step": 14270
+    },
+    {
+      "epoch": 0.5777013354917038,
+      "grad_norm": 1.1568187475204468,
+      "learning_rate": 4.309312409663432e-05,
+      "loss": 1.047,
+      "step": 14275
+    },
+    {
+      "epoch": 0.5779036827195467,
+      "grad_norm": 1.1511410474777222,
+      "learning_rate": 4.307247573817882e-05,
+      "loss": 0.9789,
+      "step": 14280
+    },
+    {
+      "epoch": 0.5781060299473897,
+      "grad_norm": 1.2285101413726807,
+      "learning_rate": 4.305182737972331e-05,
+      "loss": 0.9664,
+      "step": 14285
+    },
+    {
+      "epoch": 0.5783083771752326,
+      "grad_norm": 1.274605393409729,
+      "learning_rate": 4.303117902126781e-05,
+      "loss": 1.0009,
+      "step": 14290
+    },
+    {
+      "epoch": 0.5785107244030757,
+      "grad_norm": 1.226155400276184,
+      "learning_rate": 4.301053066281231e-05,
+      "loss": 0.9941,
+      "step": 14295
+    },
+    {
+      "epoch": 0.5787130716309187,
+      "grad_norm": 1.3813090324401855,
+      "learning_rate": 4.2989882304356804e-05,
+      "loss": 1.0203,
+      "step": 14300
+    },
+    {
+      "epoch": 0.5789154188587616,
+      "grad_norm": 1.242066502571106,
+      "learning_rate": 4.296923394590131e-05,
+      "loss": 0.9806,
+      "step": 14305
+    },
+    {
+      "epoch": 0.5791177660866046,
+      "grad_norm": 1.194819450378418,
+      "learning_rate": 4.29485855874458e-05,
+      "loss": 0.9885,
+      "step": 14310
+    },
+    {
+      "epoch": 0.5793201133144475,
+      "grad_norm": 1.0743086338043213,
+      "learning_rate": 4.2927937228990296e-05,
+      "loss": 0.9234,
+      "step": 14315
+    },
+    {
+      "epoch": 0.5795224605422906,
+      "grad_norm": 1.1583209037780762,
+      "learning_rate": 4.290728887053479e-05,
+      "loss": 0.9309,
+      "step": 14320
+    },
+    {
+      "epoch": 0.5797248077701336,
+      "grad_norm": 1.2013063430786133,
+      "learning_rate": 4.288664051207929e-05,
+      "loss": 0.9968,
+      "step": 14325
+    },
+    {
+      "epoch": 0.5799271549979765,
+      "grad_norm": 1.1978249549865723,
+      "learning_rate": 4.286599215362379e-05,
+      "loss": 0.9526,
+      "step": 14330
+    },
+    {
+      "epoch": 0.5801295022258195,
+      "grad_norm": 1.2646806240081787,
+      "learning_rate": 4.2845343795168285e-05,
+      "loss": 0.9877,
+      "step": 14335
+    },
+    {
+      "epoch": 0.5803318494536625,
+      "grad_norm": 1.271074891090393,
+      "learning_rate": 4.282469543671278e-05,
+      "loss": 1.0116,
+      "step": 14340
+    },
+    {
+      "epoch": 0.5805341966815054,
+      "grad_norm": 1.1461739540100098,
+      "learning_rate": 4.280404707825728e-05,
+      "loss": 1.0277,
+      "step": 14345
+    },
+    {
+      "epoch": 0.5807365439093485,
+      "grad_norm": 1.1434636116027832,
+      "learning_rate": 4.278339871980178e-05,
+      "loss": 1.0785,
+      "step": 14350
+    },
+    {
+      "epoch": 0.5809388911371914,
+      "grad_norm": 1.194696307182312,
+      "learning_rate": 4.2762750361346274e-05,
+      "loss": 1.0005,
+      "step": 14355
+    },
+    {
+      "epoch": 0.5811412383650344,
+      "grad_norm": 1.2170504331588745,
+      "learning_rate": 4.274210200289077e-05,
+      "loss": 0.9586,
+      "step": 14360
+    },
+    {
+      "epoch": 0.5813435855928774,
+      "grad_norm": 1.2026551961898804,
+      "learning_rate": 4.272145364443527e-05,
+      "loss": 0.9914,
+      "step": 14365
+    },
+    {
+      "epoch": 0.5815459328207203,
+      "grad_norm": 1.0865665674209595,
+      "learning_rate": 4.2700805285979766e-05,
+      "loss": 0.9974,
+      "step": 14370
+    },
+    {
+      "epoch": 0.5817482800485634,
+      "grad_norm": 1.2037372589111328,
+      "learning_rate": 4.2680156927524264e-05,
+      "loss": 1.0531,
+      "step": 14375
+    },
+    {
+      "epoch": 0.5819506272764063,
+      "grad_norm": 1.1456353664398193,
+      "learning_rate": 4.265950856906876e-05,
+      "loss": 0.9728,
+      "step": 14380
+    },
+    {
+      "epoch": 0.5821529745042493,
+      "grad_norm": 1.1749718189239502,
+      "learning_rate": 4.263886021061326e-05,
+      "loss": 1.0342,
+      "step": 14385
+    },
+    {
+      "epoch": 0.5823553217320923,
+      "grad_norm": 1.1799213886260986,
+      "learning_rate": 4.2618211852157756e-05,
+      "loss": 0.9889,
+      "step": 14390
+    },
+    {
+      "epoch": 0.5825576689599352,
+      "grad_norm": 1.0045866966247559,
+      "learning_rate": 4.259756349370225e-05,
+      "loss": 0.9254,
+      "step": 14395
+    },
+    {
+      "epoch": 0.5827600161877782,
+      "grad_norm": 1.3624835014343262,
+      "learning_rate": 4.257691513524675e-05,
+      "loss": 1.0147,
+      "step": 14400
+    },
+    {
+      "epoch": 0.5829623634156212,
+      "grad_norm": 1.0491089820861816,
+      "learning_rate": 4.255626677679125e-05,
+      "loss": 0.9083,
+      "step": 14405
+    },
+    {
+      "epoch": 0.5831647106434642,
+      "grad_norm": 1.3208736181259155,
+      "learning_rate": 4.2535618418335745e-05,
+      "loss": 1.026,
+      "step": 14410
+    },
+    {
+      "epoch": 0.5833670578713072,
+      "grad_norm": 1.1491566896438599,
+      "learning_rate": 4.251497005988024e-05,
+      "loss": 0.953,
+      "step": 14415
+    },
+    {
+      "epoch": 0.5835694050991501,
+      "grad_norm": 1.2377055883407593,
+      "learning_rate": 4.249432170142474e-05,
+      "loss": 0.9683,
+      "step": 14420
+    },
+    {
+      "epoch": 0.5837717523269931,
+      "grad_norm": 1.1501821279525757,
+      "learning_rate": 4.2473673342969237e-05,
+      "loss": 1.0097,
+      "step": 14425
+    },
+    {
+      "epoch": 0.5839740995548361,
+      "grad_norm": 1.0455905199050903,
+      "learning_rate": 4.2453024984513734e-05,
+      "loss": 1.0379,
+      "step": 14430
+    },
+    {
+      "epoch": 0.5841764467826791,
+      "grad_norm": 1.1751887798309326,
+      "learning_rate": 4.243237662605823e-05,
+      "loss": 1.0322,
+      "step": 14435
+    },
+    {
+      "epoch": 0.5843787940105221,
+      "grad_norm": 1.1225756406784058,
+      "learning_rate": 4.241172826760273e-05,
+      "loss": 1.0131,
+      "step": 14440
+    },
+    {
+      "epoch": 0.584581141238365,
+      "grad_norm": 1.4358786344528198,
+      "learning_rate": 4.2391079909147226e-05,
+      "loss": 0.9935,
+      "step": 14445
+    },
+    {
+      "epoch": 0.584783488466208,
+      "grad_norm": 1.1218318939208984,
+      "learning_rate": 4.2370431550691716e-05,
+      "loss": 1.0259,
+      "step": 14450
+    },
+    {
+      "epoch": 0.5849858356940509,
+      "grad_norm": 1.2656817436218262,
+      "learning_rate": 4.234978319223622e-05,
+      "loss": 1.0967,
+      "step": 14455
+    },
+    {
+      "epoch": 0.585188182921894,
+      "grad_norm": 1.195020318031311,
+      "learning_rate": 4.232913483378072e-05,
+      "loss": 1.0293,
+      "step": 14460
+    },
+    {
+      "epoch": 0.585390530149737,
+      "grad_norm": 1.267820119857788,
+      "learning_rate": 4.2308486475325215e-05,
+      "loss": 1.0329,
+      "step": 14465
+    },
+    {
+      "epoch": 0.5855928773775799,
+      "grad_norm": 1.2683968544006348,
+      "learning_rate": 4.228783811686971e-05,
+      "loss": 0.9337,
+      "step": 14470
+    },
+    {
+      "epoch": 0.5857952246054229,
+      "grad_norm": 1.218563199043274,
+      "learning_rate": 4.226718975841421e-05,
+      "loss": 0.9589,
+      "step": 14475
+    },
+    {
+      "epoch": 0.5859975718332658,
+      "grad_norm": 1.3326445817947388,
+      "learning_rate": 4.224654139995871e-05,
+      "loss": 0.933,
+      "step": 14480
+    },
+    {
+      "epoch": 0.5861999190611089,
+      "grad_norm": 1.155537724494934,
+      "learning_rate": 4.22258930415032e-05,
+      "loss": 1.1038,
+      "step": 14485
+    },
+    {
+      "epoch": 0.5864022662889519,
+      "grad_norm": 1.1963613033294678,
+      "learning_rate": 4.22052446830477e-05,
+      "loss": 0.9746,
+      "step": 14490
+    },
+    {
+      "epoch": 0.5866046135167948,
+      "grad_norm": 1.1540827751159668,
+      "learning_rate": 4.21845963245922e-05,
+      "loss": 1.0504,
+      "step": 14495
+    },
+    {
+      "epoch": 0.5868069607446378,
+      "grad_norm": 1.1261135339736938,
+      "learning_rate": 4.2163947966136696e-05,
+      "loss": 0.9379,
+      "step": 14500
+    },
+    {
+      "epoch": 0.5870093079724807,
+      "grad_norm": 1.1809544563293457,
+      "learning_rate": 4.214329960768119e-05,
+      "loss": 0.9444,
+      "step": 14505
+    },
+    {
+      "epoch": 0.5872116552003237,
+      "grad_norm": 1.1642730236053467,
+      "learning_rate": 4.2122651249225684e-05,
+      "loss": 0.977,
+      "step": 14510
+    },
+    {
+      "epoch": 0.5874140024281668,
+      "grad_norm": 1.2623203992843628,
+      "learning_rate": 4.210200289077019e-05,
+      "loss": 1.0256,
+      "step": 14515
+    },
+    {
+      "epoch": 0.5876163496560097,
+      "grad_norm": 1.1620464324951172,
+      "learning_rate": 4.208135453231468e-05,
+      "loss": 1.0281,
+      "step": 14520
+    },
+    {
+      "epoch": 0.5878186968838527,
+      "grad_norm": 1.2273989915847778,
+      "learning_rate": 4.206070617385918e-05,
+      "loss": 0.986,
+      "step": 14525
+    },
+    {
+      "epoch": 0.5880210441116956,
+      "grad_norm": 1.172258973121643,
+      "learning_rate": 4.204005781540368e-05,
+      "loss": 1.013,
+      "step": 14530
+    },
+    {
+      "epoch": 0.5882233913395386,
+      "grad_norm": 1.1386834383010864,
+      "learning_rate": 4.201940945694817e-05,
+      "loss": 1.0119,
+      "step": 14535
+    },
+    {
+      "epoch": 0.5884257385673817,
+      "grad_norm": 1.1152602434158325,
+      "learning_rate": 4.1998761098492674e-05,
+      "loss": 0.9738,
+      "step": 14540
+    },
+    {
+      "epoch": 0.5886280857952246,
+      "grad_norm": 1.1315892934799194,
+      "learning_rate": 4.1978112740037165e-05,
+      "loss": 0.9592,
+      "step": 14545
+    },
+    {
+      "epoch": 0.5888304330230676,
+      "grad_norm": 1.2412919998168945,
+      "learning_rate": 4.195746438158167e-05,
+      "loss": 1.0437,
+      "step": 14550
+    },
+    {
+      "epoch": 0.5890327802509105,
+      "grad_norm": 1.120381474494934,
+      "learning_rate": 4.1936816023126166e-05,
+      "loss": 1.0226,
+      "step": 14555
+    },
+    {
+      "epoch": 0.5892351274787535,
+      "grad_norm": 1.1512612104415894,
+      "learning_rate": 4.191616766467066e-05,
+      "loss": 1.0483,
+      "step": 14560
+    },
+    {
+      "epoch": 0.5894374747065965,
+      "grad_norm": 1.3168336153030396,
+      "learning_rate": 4.189551930621516e-05,
+      "loss": 0.9527,
+      "step": 14565
+    },
+    {
+      "epoch": 0.5896398219344395,
+      "grad_norm": 1.2016961574554443,
+      "learning_rate": 4.187487094775965e-05,
+      "loss": 0.9641,
+      "step": 14570
+    },
+    {
+      "epoch": 0.5898421691622825,
+      "grad_norm": 1.213405966758728,
+      "learning_rate": 4.1854222589304156e-05,
+      "loss": 1.0915,
+      "step": 14575
+    },
+    {
+      "epoch": 0.5900445163901255,
+      "grad_norm": 1.1711812019348145,
+      "learning_rate": 4.1833574230848646e-05,
+      "loss": 0.9716,
+      "step": 14580
+    },
+    {
+      "epoch": 0.5902468636179684,
+      "grad_norm": 1.2479557991027832,
+      "learning_rate": 4.181292587239315e-05,
+      "loss": 0.9745,
+      "step": 14585
+    },
+    {
+      "epoch": 0.5904492108458114,
+      "grad_norm": 1.2024391889572144,
+      "learning_rate": 4.179227751393765e-05,
+      "loss": 0.9723,
+      "step": 14590
+    },
+    {
+      "epoch": 0.5906515580736544,
+      "grad_norm": 1.2010310888290405,
+      "learning_rate": 4.177162915548214e-05,
+      "loss": 0.9534,
+      "step": 14595
+    },
+    {
+      "epoch": 0.5908539053014974,
+      "grad_norm": 1.2918951511383057,
+      "learning_rate": 4.175098079702664e-05,
+      "loss": 0.9641,
+      "step": 14600
+    },
+    {
+      "epoch": 0.5910562525293404,
+      "grad_norm": 1.25871741771698,
+      "learning_rate": 4.173033243857113e-05,
+      "loss": 0.9994,
+      "step": 14605
+    },
+    {
+      "epoch": 0.5912585997571833,
+      "grad_norm": 1.1251121759414673,
+      "learning_rate": 4.1709684080115637e-05,
+      "loss": 0.9179,
+      "step": 14610
+    },
+    {
+      "epoch": 0.5914609469850263,
+      "grad_norm": 1.3321064710617065,
+      "learning_rate": 4.168903572166013e-05,
+      "loss": 1.0003,
+      "step": 14615
+    },
+    {
+      "epoch": 0.5916632942128692,
+      "grad_norm": 1.2580547332763672,
+      "learning_rate": 4.1668387363204624e-05,
+      "loss": 1.0116,
+      "step": 14620
+    },
+    {
+      "epoch": 0.5918656414407123,
+      "grad_norm": 1.182379126548767,
+      "learning_rate": 4.164773900474913e-05,
+      "loss": 0.9547,
+      "step": 14625
+    },
+    {
+      "epoch": 0.5920679886685553,
+      "grad_norm": 1.2749547958374023,
+      "learning_rate": 4.162709064629362e-05,
+      "loss": 1.008,
+      "step": 14630
+    },
+    {
+      "epoch": 0.5922703358963982,
+      "grad_norm": 1.2136083841323853,
+      "learning_rate": 4.160644228783812e-05,
+      "loss": 1.1016,
+      "step": 14635
+    },
+    {
+      "epoch": 0.5924726831242412,
+      "grad_norm": 1.152838110923767,
+      "learning_rate": 4.1585793929382614e-05,
+      "loss": 0.984,
+      "step": 14640
+    },
+    {
+      "epoch": 0.5926750303520841,
+      "grad_norm": 1.245336890220642,
+      "learning_rate": 4.156514557092711e-05,
+      "loss": 1.0255,
+      "step": 14645
+    },
+    {
+      "epoch": 0.5928773775799272,
+      "grad_norm": 1.218671441078186,
+      "learning_rate": 4.154449721247161e-05,
+      "loss": 1.0192,
+      "step": 14650
+    },
+    {
+      "epoch": 0.5930797248077702,
+      "grad_norm": 1.1206589937210083,
+      "learning_rate": 4.1523848854016105e-05,
+      "loss": 0.9731,
+      "step": 14655
+    },
+    {
+      "epoch": 0.5932820720356131,
+      "grad_norm": 1.1717655658721924,
+      "learning_rate": 4.150320049556061e-05,
+      "loss": 0.9849,
+      "step": 14660
+    },
+    {
+      "epoch": 0.5934844192634561,
+      "grad_norm": 1.2015700340270996,
+      "learning_rate": 4.14825521371051e-05,
+      "loss": 1.0351,
+      "step": 14665
+    },
+    {
+      "epoch": 0.593686766491299,
+      "grad_norm": 1.153236746788025,
+      "learning_rate": 4.14619037786496e-05,
+      "loss": 0.989,
+      "step": 14670
+    },
+    {
+      "epoch": 0.5938891137191421,
+      "grad_norm": 1.1946407556533813,
+      "learning_rate": 4.1441255420194095e-05,
+      "loss": 1.0056,
+      "step": 14675
+    },
+    {
+      "epoch": 0.5940914609469851,
+      "grad_norm": 1.1855626106262207,
+      "learning_rate": 4.142060706173859e-05,
+      "loss": 0.981,
+      "step": 14680
+    },
+    {
+      "epoch": 0.594293808174828,
+      "grad_norm": 0.9957739114761353,
+      "learning_rate": 4.1399958703283096e-05,
+      "loss": 1.0083,
+      "step": 14685
+    },
+    {
+      "epoch": 0.594496155402671,
+      "grad_norm": 1.2123115062713623,
+      "learning_rate": 4.1379310344827587e-05,
+      "loss": 0.926,
+      "step": 14690
+    },
+    {
+      "epoch": 0.5946985026305139,
+      "grad_norm": 1.103893756866455,
+      "learning_rate": 4.135866198637209e-05,
+      "loss": 1.0038,
+      "step": 14695
+    },
+    {
+      "epoch": 0.5949008498583569,
+      "grad_norm": 1.1878453493118286,
+      "learning_rate": 4.133801362791658e-05,
+      "loss": 1.0305,
+      "step": 14700
+    },
+    {
+      "epoch": 0.5951031970862,
+      "grad_norm": 1.1317086219787598,
+      "learning_rate": 4.131736526946108e-05,
+      "loss": 0.9945,
+      "step": 14705
+    },
+    {
+      "epoch": 0.5953055443140429,
+      "grad_norm": 1.0596208572387695,
+      "learning_rate": 4.1296716911005576e-05,
+      "loss": 1.047,
+      "step": 14710
+    },
+    {
+      "epoch": 0.5955078915418859,
+      "grad_norm": 1.2252612113952637,
+      "learning_rate": 4.127606855255007e-05,
+      "loss": 1.0509,
+      "step": 14715
+    },
+    {
+      "epoch": 0.5957102387697288,
+      "grad_norm": 1.1817820072174072,
+      "learning_rate": 4.125542019409458e-05,
+      "loss": 0.9825,
+      "step": 14720
+    },
+    {
+      "epoch": 0.5959125859975718,
+      "grad_norm": 1.2887251377105713,
+      "learning_rate": 4.123477183563907e-05,
+      "loss": 0.9911,
+      "step": 14725
+    },
+    {
+      "epoch": 0.5961149332254149,
+      "grad_norm": 1.3245701789855957,
+      "learning_rate": 4.1214123477183565e-05,
+      "loss": 0.9956,
+      "step": 14730
+    },
+    {
+      "epoch": 0.5963172804532578,
+      "grad_norm": 1.1964930295944214,
+      "learning_rate": 4.119347511872806e-05,
+      "loss": 0.9505,
+      "step": 14735
+    },
+    {
+      "epoch": 0.5965196276811008,
+      "grad_norm": 1.1000622510910034,
+      "learning_rate": 4.117282676027256e-05,
+      "loss": 1.0079,
+      "step": 14740
+    },
+    {
+      "epoch": 0.5967219749089437,
+      "grad_norm": 1.2138142585754395,
+      "learning_rate": 4.115217840181706e-05,
+      "loss": 0.9815,
+      "step": 14745
+    },
+    {
+      "epoch": 0.5969243221367867,
+      "grad_norm": 1.1650316715240479,
+      "learning_rate": 4.1131530043361554e-05,
+      "loss": 1.0002,
+      "step": 14750
+    },
+    {
+      "epoch": 0.5971266693646297,
+      "grad_norm": 1.162232518196106,
+      "learning_rate": 4.111088168490605e-05,
+      "loss": 0.9867,
+      "step": 14755
+    },
+    {
+      "epoch": 0.5973290165924727,
+      "grad_norm": 1.2270444631576538,
+      "learning_rate": 4.109023332645055e-05,
+      "loss": 1.0356,
+      "step": 14760
+    },
+    {
+      "epoch": 0.5975313638203157,
+      "grad_norm": 1.1669378280639648,
+      "learning_rate": 4.1069584967995046e-05,
+      "loss": 0.9689,
+      "step": 14765
+    },
+    {
+      "epoch": 0.5977337110481586,
+      "grad_norm": 1.0744789838790894,
+      "learning_rate": 4.104893660953954e-05,
+      "loss": 0.9858,
+      "step": 14770
+    },
+    {
+      "epoch": 0.5979360582760016,
+      "grad_norm": 1.1895402669906616,
+      "learning_rate": 4.102828825108404e-05,
+      "loss": 0.9566,
+      "step": 14775
+    },
+    {
+      "epoch": 0.5981384055038446,
+      "grad_norm": 1.338990569114685,
+      "learning_rate": 4.100763989262854e-05,
+      "loss": 1.0136,
+      "step": 14780
+    },
+    {
+      "epoch": 0.5983407527316876,
+      "grad_norm": 1.1293729543685913,
+      "learning_rate": 4.0986991534173035e-05,
+      "loss": 0.9983,
+      "step": 14785
+    },
+    {
+      "epoch": 0.5985430999595306,
+      "grad_norm": 1.154767632484436,
+      "learning_rate": 4.096634317571753e-05,
+      "loss": 0.9983,
+      "step": 14790
+    },
+    {
+      "epoch": 0.5987454471873735,
+      "grad_norm": 1.1595851182937622,
+      "learning_rate": 4.094569481726203e-05,
+      "loss": 0.9666,
+      "step": 14795
+    },
+    {
+      "epoch": 0.5989477944152165,
+      "grad_norm": 1.1495463848114014,
+      "learning_rate": 4.092504645880653e-05,
+      "loss": 1.0128,
+      "step": 14800
+    },
+    {
+      "epoch": 0.5991501416430595,
+      "grad_norm": 1.178675651550293,
+      "learning_rate": 4.0904398100351024e-05,
+      "loss": 0.9879,
+      "step": 14805
+    },
+    {
+      "epoch": 0.5993524888709024,
+      "grad_norm": 1.1787465810775757,
+      "learning_rate": 4.088374974189552e-05,
+      "loss": 0.9369,
+      "step": 14810
+    },
+    {
+      "epoch": 0.5995548360987455,
+      "grad_norm": 1.1427485942840576,
+      "learning_rate": 4.086310138344002e-05,
+      "loss": 0.9885,
+      "step": 14815
+    },
+    {
+      "epoch": 0.5997571833265885,
+      "grad_norm": 1.3166379928588867,
+      "learning_rate": 4.0842453024984516e-05,
+      "loss": 0.9942,
+      "step": 14820
+    },
+    {
+      "epoch": 0.5999595305544314,
+      "grad_norm": 1.1612681150436401,
+      "learning_rate": 4.0821804666529014e-05,
+      "loss": 0.9906,
+      "step": 14825
+    },
+    {
+      "epoch": 0.6001618777822744,
+      "grad_norm": 1.1479135751724243,
+      "learning_rate": 4.080115630807351e-05,
+      "loss": 0.9998,
+      "step": 14830
+    },
+    {
+      "epoch": 0.6003642250101173,
+      "grad_norm": 1.1357676982879639,
+      "learning_rate": 4.078050794961801e-05,
+      "loss": 0.905,
+      "step": 14835
+    },
+    {
+      "epoch": 0.6005665722379604,
+      "grad_norm": 1.254703402519226,
+      "learning_rate": 4.0759859591162505e-05,
+      "loss": 0.9856,
+      "step": 14840
+    },
+    {
+      "epoch": 0.6007689194658034,
+      "grad_norm": 1.1657435894012451,
+      "learning_rate": 4.0739211232707e-05,
+      "loss": 0.92,
+      "step": 14845
+    },
+    {
+      "epoch": 0.6009712666936463,
+      "grad_norm": 1.1573985815048218,
+      "learning_rate": 4.07185628742515e-05,
+      "loss": 0.9758,
+      "step": 14850
+    },
+    {
+      "epoch": 0.6011736139214893,
+      "grad_norm": 1.1526132822036743,
+      "learning_rate": 4.0697914515796e-05,
+      "loss": 1.0268,
+      "step": 14855
+    },
+    {
+      "epoch": 0.6013759611493322,
+      "grad_norm": 1.119799256324768,
+      "learning_rate": 4.0677266157340495e-05,
+      "loss": 0.9355,
+      "step": 14860
+    },
+    {
+      "epoch": 0.6015783083771752,
+      "grad_norm": 1.231797695159912,
+      "learning_rate": 4.0656617798884985e-05,
+      "loss": 0.9955,
+      "step": 14865
+    },
+    {
+      "epoch": 0.6017806556050183,
+      "grad_norm": 1.1451510190963745,
+      "learning_rate": 4.063596944042949e-05,
+      "loss": 0.9976,
+      "step": 14870
+    },
+    {
+      "epoch": 0.6019830028328612,
+      "grad_norm": 1.0147970914840698,
+      "learning_rate": 4.0615321081973987e-05,
+      "loss": 0.9881,
+      "step": 14875
+    },
+    {
+      "epoch": 0.6021853500607042,
+      "grad_norm": 1.0572047233581543,
+      "learning_rate": 4.0594672723518484e-05,
+      "loss": 1.0732,
+      "step": 14880
+    },
+    {
+      "epoch": 0.6023876972885471,
+      "grad_norm": 1.1836466789245605,
+      "learning_rate": 4.057402436506298e-05,
+      "loss": 0.9699,
+      "step": 14885
+    },
+    {
+      "epoch": 0.6025900445163901,
+      "grad_norm": 1.2634001970291138,
+      "learning_rate": 4.055337600660747e-05,
+      "loss": 0.981,
+      "step": 14890
+    },
+    {
+      "epoch": 0.6027923917442332,
+      "grad_norm": 1.1793850660324097,
+      "learning_rate": 4.0532727648151976e-05,
+      "loss": 0.9694,
+      "step": 14895
+    },
+    {
+      "epoch": 0.6029947389720761,
+      "grad_norm": 1.241394281387329,
+      "learning_rate": 4.0512079289696466e-05,
+      "loss": 1.027,
+      "step": 14900
+    },
+    {
+      "epoch": 0.6031970861999191,
+      "grad_norm": 1.2651069164276123,
+      "learning_rate": 4.049143093124097e-05,
+      "loss": 1.002,
+      "step": 14905
+    },
+    {
+      "epoch": 0.603399433427762,
+      "grad_norm": 1.171766757965088,
+      "learning_rate": 4.047078257278547e-05,
+      "loss": 1.0042,
+      "step": 14910
+    },
+    {
+      "epoch": 0.603601780655605,
+      "grad_norm": 1.2115875482559204,
+      "learning_rate": 4.045013421432996e-05,
+      "loss": 1.0144,
+      "step": 14915
+    },
+    {
+      "epoch": 0.603804127883448,
+      "grad_norm": 1.08839750289917,
+      "learning_rate": 4.042948585587446e-05,
+      "loss": 1.0083,
+      "step": 14920
+    },
+    {
+      "epoch": 0.604006475111291,
+      "grad_norm": 1.0900975465774536,
+      "learning_rate": 4.040883749741895e-05,
+      "loss": 1.0039,
+      "step": 14925
+    },
+    {
+      "epoch": 0.604208822339134,
+      "grad_norm": 1.8475017547607422,
+      "learning_rate": 4.038818913896346e-05,
+      "loss": 0.9778,
+      "step": 14930
+    },
+    {
+      "epoch": 0.6044111695669769,
+      "grad_norm": 1.1439038515090942,
+      "learning_rate": 4.0367540780507954e-05,
+      "loss": 0.9927,
+      "step": 14935
+    },
+    {
+      "epoch": 0.6046135167948199,
+      "grad_norm": 1.2487175464630127,
+      "learning_rate": 4.034689242205245e-05,
+      "loss": 1.0361,
+      "step": 14940
+    },
+    {
+      "epoch": 0.6048158640226629,
+      "grad_norm": 1.2005751132965088,
+      "learning_rate": 4.032624406359695e-05,
+      "loss": 0.99,
+      "step": 14945
+    },
+    {
+      "epoch": 0.6050182112505059,
+      "grad_norm": 1.0894343852996826,
+      "learning_rate": 4.030559570514144e-05,
+      "loss": 1.0457,
+      "step": 14950
+    },
+    {
+      "epoch": 0.6052205584783489,
+      "grad_norm": 1.2973085641860962,
+      "learning_rate": 4.028494734668594e-05,
+      "loss": 1.018,
+      "step": 14955
+    },
+    {
+      "epoch": 0.6054229057061918,
+      "grad_norm": 1.1314729452133179,
+      "learning_rate": 4.0264298988230434e-05,
+      "loss": 1.0048,
+      "step": 14960
+    },
+    {
+      "epoch": 0.6056252529340348,
+      "grad_norm": 1.1631298065185547,
+      "learning_rate": 4.024365062977494e-05,
+      "loss": 0.9561,
+      "step": 14965
+    },
+    {
+      "epoch": 0.6058276001618778,
+      "grad_norm": 1.217316746711731,
+      "learning_rate": 4.0223002271319435e-05,
+      "loss": 0.9906,
+      "step": 14970
+    },
+    {
+      "epoch": 0.6060299473897207,
+      "grad_norm": 1.2196394205093384,
+      "learning_rate": 4.0202353912863926e-05,
+      "loss": 0.9855,
+      "step": 14975
+    },
+    {
+      "epoch": 0.6062322946175638,
+      "grad_norm": 1.2208998203277588,
+      "learning_rate": 4.018170555440843e-05,
+      "loss": 0.9476,
+      "step": 14980
+    },
+    {
+      "epoch": 0.6064346418454067,
+      "grad_norm": 1.222133994102478,
+      "learning_rate": 4.016105719595292e-05,
+      "loss": 0.9434,
+      "step": 14985
+    },
+    {
+      "epoch": 0.6066369890732497,
+      "grad_norm": 1.2885711193084717,
+      "learning_rate": 4.0140408837497424e-05,
+      "loss": 0.9793,
+      "step": 14990
+    },
+    {
+      "epoch": 0.6068393363010927,
+      "grad_norm": 1.108764410018921,
+      "learning_rate": 4.0119760479041915e-05,
+      "loss": 0.9479,
+      "step": 14995
+    },
+    {
+      "epoch": 0.6070416835289356,
+      "grad_norm": 1.2786108255386353,
+      "learning_rate": 4.009911212058641e-05,
+      "loss": 0.9701,
+      "step": 15000
+    },
+    {
+      "epoch": 0.6072440307567787,
+      "grad_norm": 1.1678630113601685,
+      "learning_rate": 4.0078463762130916e-05,
+      "loss": 1.0102,
+      "step": 15005
+    },
+    {
+      "epoch": 0.6074463779846216,
+      "grad_norm": 1.2520824670791626,
+      "learning_rate": 4.005781540367541e-05,
+      "loss": 0.9828,
+      "step": 15010
+    },
+    {
+      "epoch": 0.6076487252124646,
+      "grad_norm": 1.2107146978378296,
+      "learning_rate": 4.003716704521991e-05,
+      "loss": 1.0024,
+      "step": 15015
+    },
+    {
+      "epoch": 0.6078510724403076,
+      "grad_norm": 1.2790188789367676,
+      "learning_rate": 4.00165186867644e-05,
+      "loss": 1.0017,
+      "step": 15020
+    },
+    {
+      "epoch": 0.6080534196681505,
+      "grad_norm": 1.1601837873458862,
+      "learning_rate": 3.99958703283089e-05,
+      "loss": 0.9687,
+      "step": 15025
+    },
+    {
+      "epoch": 0.6082557668959935,
+      "grad_norm": 1.2074768543243408,
+      "learning_rate": 3.9975221969853396e-05,
+      "loss": 1.0258,
+      "step": 15030
+    },
+    {
+      "epoch": 0.6084581141238365,
+      "grad_norm": 1.3172718286514282,
+      "learning_rate": 3.995457361139789e-05,
+      "loss": 0.9482,
+      "step": 15035
+    },
+    {
+      "epoch": 0.6086604613516795,
+      "grad_norm": 1.1703674793243408,
+      "learning_rate": 3.99339252529424e-05,
+      "loss": 1.0559,
+      "step": 15040
+    },
+    {
+      "epoch": 0.6088628085795225,
+      "grad_norm": 1.2748117446899414,
+      "learning_rate": 3.991327689448689e-05,
+      "loss": 0.9715,
+      "step": 15045
+    },
+    {
+      "epoch": 0.6090651558073654,
+      "grad_norm": 1.1625258922576904,
+      "learning_rate": 3.989262853603139e-05,
+      "loss": 0.9783,
+      "step": 15050
+    },
+    {
+      "epoch": 0.6092675030352084,
+      "grad_norm": 1.2481300830841064,
+      "learning_rate": 3.987198017757588e-05,
+      "loss": 1.0329,
+      "step": 15055
+    },
+    {
+      "epoch": 0.6094698502630514,
+      "grad_norm": 1.2934519052505493,
+      "learning_rate": 3.985133181912038e-05,
+      "loss": 0.978,
+      "step": 15060
+    },
+    {
+      "epoch": 0.6096721974908944,
+      "grad_norm": 1.2320512533187866,
+      "learning_rate": 3.9830683460664884e-05,
+      "loss": 0.9465,
+      "step": 15065
+    },
+    {
+      "epoch": 0.6098745447187374,
+      "grad_norm": 1.1613794565200806,
+      "learning_rate": 3.9810035102209374e-05,
+      "loss": 0.942,
+      "step": 15070
+    },
+    {
+      "epoch": 0.6100768919465803,
+      "grad_norm": 1.1811312437057495,
+      "learning_rate": 3.978938674375388e-05,
+      "loss": 0.975,
+      "step": 15075
+    },
+    {
+      "epoch": 0.6102792391744233,
+      "grad_norm": 1.1655839681625366,
+      "learning_rate": 3.976873838529837e-05,
+      "loss": 0.9918,
+      "step": 15080
+    },
+    {
+      "epoch": 0.6104815864022662,
+      "grad_norm": 1.145727515220642,
+      "learning_rate": 3.9748090026842866e-05,
+      "loss": 1.0017,
+      "step": 15085
+    },
+    {
+      "epoch": 0.6106839336301093,
+      "grad_norm": 1.175281047821045,
+      "learning_rate": 3.9727441668387364e-05,
+      "loss": 1.0345,
+      "step": 15090
+    },
+    {
+      "epoch": 0.6108862808579523,
+      "grad_norm": 1.318772554397583,
+      "learning_rate": 3.970679330993186e-05,
+      "loss": 1.0113,
+      "step": 15095
+    },
+    {
+      "epoch": 0.6110886280857952,
+      "grad_norm": 1.195495843887329,
+      "learning_rate": 3.9686144951476365e-05,
+      "loss": 0.996,
+      "step": 15100
+    },
+    {
+      "epoch": 0.6112909753136382,
+      "grad_norm": 1.170361876487732,
+      "learning_rate": 3.9665496593020855e-05,
+      "loss": 1.0214,
+      "step": 15105
+    },
+    {
+      "epoch": 0.6114933225414811,
+      "grad_norm": 1.2344578504562378,
+      "learning_rate": 3.964484823456535e-05,
+      "loss": 0.981,
+      "step": 15110
+    },
+    {
+      "epoch": 0.6116956697693242,
+      "grad_norm": 1.1404913663864136,
+      "learning_rate": 3.962419987610985e-05,
+      "loss": 1.0143,
+      "step": 15115
+    },
+    {
+      "epoch": 0.6118980169971672,
+      "grad_norm": 1.2335190773010254,
+      "learning_rate": 3.960355151765435e-05,
+      "loss": 0.9336,
+      "step": 15120
+    },
+    {
+      "epoch": 0.6121003642250101,
+      "grad_norm": 1.2277600765228271,
+      "learning_rate": 3.9582903159198845e-05,
+      "loss": 0.9821,
+      "step": 15125
+    },
+    {
+      "epoch": 0.6123027114528531,
+      "grad_norm": 1.2261604070663452,
+      "learning_rate": 3.956225480074334e-05,
+      "loss": 0.981,
+      "step": 15130
+    },
+    {
+      "epoch": 0.612505058680696,
+      "grad_norm": 1.1750844717025757,
+      "learning_rate": 3.954160644228784e-05,
+      "loss": 0.9391,
+      "step": 15135
+    },
+    {
+      "epoch": 0.612707405908539,
+      "grad_norm": 1.1132014989852905,
+      "learning_rate": 3.9520958083832336e-05,
+      "loss": 0.9887,
+      "step": 15140
+    },
+    {
+      "epoch": 0.6129097531363821,
+      "grad_norm": 1.1465575695037842,
+      "learning_rate": 3.9500309725376834e-05,
+      "loss": 0.9454,
+      "step": 15145
+    },
+    {
+      "epoch": 0.613112100364225,
+      "grad_norm": 1.1274523735046387,
+      "learning_rate": 3.947966136692133e-05,
+      "loss": 1.0278,
+      "step": 15150
+    },
+    {
+      "epoch": 0.613314447592068,
+      "grad_norm": 1.1466535329818726,
+      "learning_rate": 3.945901300846583e-05,
+      "loss": 0.8991,
+      "step": 15155
+    },
+    {
+      "epoch": 0.613516794819911,
+      "grad_norm": 1.2201199531555176,
+      "learning_rate": 3.9438364650010326e-05,
+      "loss": 1.0018,
+      "step": 15160
+    },
+    {
+      "epoch": 0.6137191420477539,
+      "grad_norm": 1.1986336708068848,
+      "learning_rate": 3.941771629155482e-05,
+      "loss": 1.0362,
+      "step": 15165
+    },
+    {
+      "epoch": 0.613921489275597,
+      "grad_norm": 1.1385431289672852,
+      "learning_rate": 3.939706793309932e-05,
+      "loss": 0.978,
+      "step": 15170
+    },
+    {
+      "epoch": 0.6141238365034399,
+      "grad_norm": 1.144865870475769,
+      "learning_rate": 3.937641957464382e-05,
+      "loss": 0.9379,
+      "step": 15175
+    },
+    {
+      "epoch": 0.6143261837312829,
+      "grad_norm": 1.3139859437942505,
+      "learning_rate": 3.9355771216188315e-05,
+      "loss": 1.0175,
+      "step": 15180
+    },
+    {
+      "epoch": 0.6145285309591259,
+      "grad_norm": 1.0443328619003296,
+      "learning_rate": 3.933512285773281e-05,
+      "loss": 1.0142,
+      "step": 15185
+    },
+    {
+      "epoch": 0.6147308781869688,
+      "grad_norm": 1.1730111837387085,
+      "learning_rate": 3.931447449927731e-05,
+      "loss": 0.9804,
+      "step": 15190
+    },
+    {
+      "epoch": 0.6149332254148118,
+      "grad_norm": 1.1835631132125854,
+      "learning_rate": 3.929382614082181e-05,
+      "loss": 0.9471,
+      "step": 15195
+    },
+    {
+      "epoch": 0.6151355726426548,
+      "grad_norm": 1.3823308944702148,
+      "learning_rate": 3.9273177782366304e-05,
+      "loss": 1.0063,
+      "step": 15200
+    },
+    {
+      "epoch": 0.6153379198704978,
+      "grad_norm": 1.0838249921798706,
+      "learning_rate": 3.92525294239108e-05,
+      "loss": 1.0202,
+      "step": 15205
+    },
+    {
+      "epoch": 0.6155402670983408,
+      "grad_norm": 1.152218222618103,
+      "learning_rate": 3.92318810654553e-05,
+      "loss": 1.0241,
+      "step": 15210
+    },
+    {
+      "epoch": 0.6157426143261837,
+      "grad_norm": 1.018296241760254,
+      "learning_rate": 3.9211232706999796e-05,
+      "loss": 0.9714,
+      "step": 15215
+    },
+    {
+      "epoch": 0.6159449615540267,
+      "grad_norm": 1.2379239797592163,
+      "learning_rate": 3.919058434854429e-05,
+      "loss": 0.9989,
+      "step": 15220
+    },
+    {
+      "epoch": 0.6161473087818697,
+      "grad_norm": 1.1574537754058838,
+      "learning_rate": 3.916993599008879e-05,
+      "loss": 1.0686,
+      "step": 15225
+    },
+    {
+      "epoch": 0.6163496560097127,
+      "grad_norm": 1.1106170415878296,
+      "learning_rate": 3.914928763163329e-05,
+      "loss": 0.9782,
+      "step": 15230
+    },
+    {
+      "epoch": 0.6165520032375557,
+      "grad_norm": 1.272621989250183,
+      "learning_rate": 3.9128639273177785e-05,
+      "loss": 1.0206,
+      "step": 15235
+    },
+    {
+      "epoch": 0.6167543504653986,
+      "grad_norm": 1.2506707906723022,
+      "learning_rate": 3.910799091472228e-05,
+      "loss": 0.9813,
+      "step": 15240
+    },
+    {
+      "epoch": 0.6169566976932416,
+      "grad_norm": 1.2709102630615234,
+      "learning_rate": 3.908734255626677e-05,
+      "loss": 1.0325,
+      "step": 15245
+    },
+    {
+      "epoch": 0.6171590449210845,
+      "grad_norm": 1.1361140012741089,
+      "learning_rate": 3.906669419781128e-05,
+      "loss": 1.0265,
+      "step": 15250
+    },
+    {
+      "epoch": 0.6173613921489276,
+      "grad_norm": 1.2748886346817017,
+      "learning_rate": 3.9046045839355774e-05,
+      "loss": 0.9605,
+      "step": 15255
+    },
+    {
+      "epoch": 0.6175637393767706,
+      "grad_norm": 1.3143259286880493,
+      "learning_rate": 3.902539748090027e-05,
+      "loss": 1.0045,
+      "step": 15260
+    },
+    {
+      "epoch": 0.6177660866046135,
+      "grad_norm": 1.2562155723571777,
+      "learning_rate": 3.900474912244477e-05,
+      "loss": 1.0042,
+      "step": 15265
+    },
+    {
+      "epoch": 0.6179684338324565,
+      "grad_norm": 1.3425461053848267,
+      "learning_rate": 3.8984100763989266e-05,
+      "loss": 1.0258,
+      "step": 15270
+    },
+    {
+      "epoch": 0.6181707810602994,
+      "grad_norm": 1.1353882551193237,
+      "learning_rate": 3.8963452405533764e-05,
+      "loss": 1.0268,
+      "step": 15275
+    },
+    {
+      "epoch": 0.6183731282881425,
+      "grad_norm": 1.1892770528793335,
+      "learning_rate": 3.8942804047078254e-05,
+      "loss": 0.9912,
+      "step": 15280
+    },
+    {
+      "epoch": 0.6185754755159855,
+      "grad_norm": 1.1367006301879883,
+      "learning_rate": 3.892215568862276e-05,
+      "loss": 0.9799,
+      "step": 15285
+    },
+    {
+      "epoch": 0.6187778227438284,
+      "grad_norm": 1.2165271043777466,
+      "learning_rate": 3.8901507330167255e-05,
+      "loss": 1.0506,
+      "step": 15290
+    },
+    {
+      "epoch": 0.6189801699716714,
+      "grad_norm": 1.2636433839797974,
+      "learning_rate": 3.888085897171175e-05,
+      "loss": 0.9692,
+      "step": 15295
+    },
+    {
+      "epoch": 0.6191825171995143,
+      "grad_norm": 1.301344871520996,
+      "learning_rate": 3.886021061325625e-05,
+      "loss": 1.0132,
+      "step": 15300
+    },
+    {
+      "epoch": 0.6193848644273573,
+      "grad_norm": 1.1484495401382446,
+      "learning_rate": 3.883956225480074e-05,
+      "loss": 0.9806,
+      "step": 15305
+    },
+    {
+      "epoch": 0.6195872116552004,
+      "grad_norm": 1.274308681488037,
+      "learning_rate": 3.8818913896345245e-05,
+      "loss": 0.9839,
+      "step": 15310
+    },
+    {
+      "epoch": 0.6197895588830433,
+      "grad_norm": 1.2011617422103882,
+      "learning_rate": 3.879826553788974e-05,
+      "loss": 1.0091,
+      "step": 15315
+    },
+    {
+      "epoch": 0.6199919061108863,
+      "grad_norm": 1.217674970626831,
+      "learning_rate": 3.877761717943424e-05,
+      "loss": 1.0073,
+      "step": 15320
+    },
+    {
+      "epoch": 0.6201942533387292,
+      "grad_norm": 1.2425445318222046,
+      "learning_rate": 3.8756968820978736e-05,
+      "loss": 1.0315,
+      "step": 15325
+    },
+    {
+      "epoch": 0.6203966005665722,
+      "grad_norm": 1.2316184043884277,
+      "learning_rate": 3.873632046252323e-05,
+      "loss": 0.9831,
+      "step": 15330
+    },
+    {
+      "epoch": 0.6205989477944153,
+      "grad_norm": 1.172351598739624,
+      "learning_rate": 3.871567210406773e-05,
+      "loss": 0.9583,
+      "step": 15335
+    },
+    {
+      "epoch": 0.6208012950222582,
+      "grad_norm": 1.210734248161316,
+      "learning_rate": 3.869502374561222e-05,
+      "loss": 0.9854,
+      "step": 15340
+    },
+    {
+      "epoch": 0.6210036422501012,
+      "grad_norm": 1.279604434967041,
+      "learning_rate": 3.8674375387156726e-05,
+      "loss": 1.0296,
+      "step": 15345
+    },
+    {
+      "epoch": 0.6212059894779441,
+      "grad_norm": 1.3448833227157593,
+      "learning_rate": 3.865372702870122e-05,
+      "loss": 1.0159,
+      "step": 15350
+    },
+    {
+      "epoch": 0.6214083367057871,
+      "grad_norm": 1.362277626991272,
+      "learning_rate": 3.8633078670245713e-05,
+      "loss": 0.9623,
+      "step": 15355
+    },
+    {
+      "epoch": 0.6216106839336301,
+      "grad_norm": 1.2423343658447266,
+      "learning_rate": 3.861243031179022e-05,
+      "loss": 1.0087,
+      "step": 15360
+    },
+    {
+      "epoch": 0.6218130311614731,
+      "grad_norm": 1.1315988302230835,
+      "learning_rate": 3.859178195333471e-05,
+      "loss": 1.0209,
+      "step": 15365
+    },
+    {
+      "epoch": 0.6220153783893161,
+      "grad_norm": 1.128307580947876,
+      "learning_rate": 3.857113359487921e-05,
+      "loss": 0.9881,
+      "step": 15370
+    },
+    {
+      "epoch": 0.622217725617159,
+      "grad_norm": 1.2249513864517212,
+      "learning_rate": 3.85504852364237e-05,
+      "loss": 1.0157,
+      "step": 15375
+    },
+    {
+      "epoch": 0.622420072845002,
+      "grad_norm": 1.1540697813034058,
+      "learning_rate": 3.85298368779682e-05,
+      "loss": 0.985,
+      "step": 15380
+    },
+    {
+      "epoch": 0.622622420072845,
+      "grad_norm": 1.1913505792617798,
+      "learning_rate": 3.8509188519512704e-05,
+      "loss": 0.9908,
+      "step": 15385
+    },
+    {
+      "epoch": 0.622824767300688,
+      "grad_norm": 1.106575608253479,
+      "learning_rate": 3.8488540161057195e-05,
+      "loss": 1.058,
+      "step": 15390
+    },
+    {
+      "epoch": 0.623027114528531,
+      "grad_norm": 1.201764464378357,
+      "learning_rate": 3.84678918026017e-05,
+      "loss": 0.998,
+      "step": 15395
+    },
+    {
+      "epoch": 0.623229461756374,
+      "grad_norm": 1.1951693296432495,
+      "learning_rate": 3.844724344414619e-05,
+      "loss": 0.9374,
+      "step": 15400
+    },
+    {
+      "epoch": 0.6234318089842169,
+      "grad_norm": 1.1139637231826782,
+      "learning_rate": 3.842659508569069e-05,
+      "loss": 1.056,
+      "step": 15405
+    },
+    {
+      "epoch": 0.6236341562120599,
+      "grad_norm": 1.175554633140564,
+      "learning_rate": 3.8405946727235184e-05,
+      "loss": 1.0014,
+      "step": 15410
+    },
+    {
+      "epoch": 0.6238365034399028,
+      "grad_norm": 1.187445044517517,
+      "learning_rate": 3.838529836877968e-05,
+      "loss": 1.0286,
+      "step": 15415
+    },
+    {
+      "epoch": 0.6240388506677459,
+      "grad_norm": 1.302847981452942,
+      "learning_rate": 3.8364650010324185e-05,
+      "loss": 1.0182,
+      "step": 15420
+    },
+    {
+      "epoch": 0.6242411978955889,
+      "grad_norm": 1.1043428182601929,
+      "learning_rate": 3.8344001651868676e-05,
+      "loss": 0.9983,
+      "step": 15425
+    },
+    {
+      "epoch": 0.6244435451234318,
+      "grad_norm": 1.2310084104537964,
+      "learning_rate": 3.832335329341318e-05,
+      "loss": 1.0609,
+      "step": 15430
+    },
+    {
+      "epoch": 0.6246458923512748,
+      "grad_norm": 1.2543423175811768,
+      "learning_rate": 3.830270493495767e-05,
+      "loss": 1.0207,
+      "step": 15435
+    },
+    {
+      "epoch": 0.6248482395791177,
+      "grad_norm": 1.2612595558166504,
+      "learning_rate": 3.828205657650217e-05,
+      "loss": 0.9168,
+      "step": 15440
+    },
+    {
+      "epoch": 0.6250505868069608,
+      "grad_norm": 1.121643304824829,
+      "learning_rate": 3.8261408218046665e-05,
+      "loss": 1.0358,
+      "step": 15445
+    },
+    {
+      "epoch": 0.6252529340348038,
+      "grad_norm": 1.254227876663208,
+      "learning_rate": 3.824075985959116e-05,
+      "loss": 0.9918,
+      "step": 15450
+    },
+    {
+      "epoch": 0.6254552812626467,
+      "grad_norm": 1.2365626096725464,
+      "learning_rate": 3.8220111501135666e-05,
+      "loss": 0.9593,
+      "step": 15455
+    },
+    {
+      "epoch": 0.6256576284904897,
+      "grad_norm": 1.1530671119689941,
+      "learning_rate": 3.819946314268016e-05,
+      "loss": 1.0402,
+      "step": 15460
+    },
+    {
+      "epoch": 0.6258599757183326,
+      "grad_norm": 1.200811505317688,
+      "learning_rate": 3.8178814784224654e-05,
+      "loss": 0.9741,
+      "step": 15465
+    },
+    {
+      "epoch": 0.6260623229461756,
+      "grad_norm": 1.215122103691101,
+      "learning_rate": 3.815816642576915e-05,
+      "loss": 1.0084,
+      "step": 15470
+    },
+    {
+      "epoch": 0.6262646701740187,
+      "grad_norm": 1.1331044435501099,
+      "learning_rate": 3.813751806731365e-05,
+      "loss": 0.9991,
+      "step": 15475
+    },
+    {
+      "epoch": 0.6264670174018616,
+      "grad_norm": 1.1087418794631958,
+      "learning_rate": 3.811686970885815e-05,
+      "loss": 0.9812,
+      "step": 15480
+    },
+    {
+      "epoch": 0.6266693646297046,
+      "grad_norm": 1.2703540325164795,
+      "learning_rate": 3.809622135040264e-05,
+      "loss": 0.9741,
+      "step": 15485
+    },
+    {
+      "epoch": 0.6268717118575475,
+      "grad_norm": 1.2571314573287964,
+      "learning_rate": 3.807557299194714e-05,
+      "loss": 0.9684,
+      "step": 15490
+    },
+    {
+      "epoch": 0.6270740590853905,
+      "grad_norm": 1.2576099634170532,
+      "learning_rate": 3.805492463349164e-05,
+      "loss": 1.0615,
+      "step": 15495
+    },
+    {
+      "epoch": 0.6272764063132336,
+      "grad_norm": 1.0620274543762207,
+      "learning_rate": 3.8034276275036135e-05,
+      "loss": 0.92,
+      "step": 15500
+    },
+    {
+      "epoch": 0.6274787535410765,
+      "grad_norm": 1.2089036703109741,
+      "learning_rate": 3.801362791658063e-05,
+      "loss": 1.0057,
+      "step": 15505
+    },
+    {
+      "epoch": 0.6276811007689195,
+      "grad_norm": 1.563827633857727,
+      "learning_rate": 3.799297955812513e-05,
+      "loss": 0.978,
+      "step": 15510
+    },
+    {
+      "epoch": 0.6278834479967624,
+      "grad_norm": 1.2251412868499756,
+      "learning_rate": 3.7972331199669634e-05,
+      "loss": 0.9571,
+      "step": 15515
+    },
+    {
+      "epoch": 0.6280857952246054,
+      "grad_norm": 1.2349053621292114,
+      "learning_rate": 3.7951682841214124e-05,
+      "loss": 1.0135,
+      "step": 15520
+    },
+    {
+      "epoch": 0.6282881424524484,
+      "grad_norm": 1.2551171779632568,
+      "learning_rate": 3.793103448275862e-05,
+      "loss": 0.979,
+      "step": 15525
+    },
+    {
+      "epoch": 0.6284904896802914,
+      "grad_norm": 1.2055236101150513,
+      "learning_rate": 3.791038612430312e-05,
+      "loss": 1.0239,
+      "step": 15530
+    },
+    {
+      "epoch": 0.6286928369081344,
+      "grad_norm": 1.3006428480148315,
+      "learning_rate": 3.7889737765847616e-05,
+      "loss": 1.006,
+      "step": 15535
+    },
+    {
+      "epoch": 0.6288951841359773,
+      "grad_norm": 1.236907958984375,
+      "learning_rate": 3.7869089407392113e-05,
+      "loss": 0.9948,
+      "step": 15540
+    },
+    {
+      "epoch": 0.6290975313638203,
+      "grad_norm": 1.089967131614685,
+      "learning_rate": 3.784844104893661e-05,
+      "loss": 0.9968,
+      "step": 15545
+    },
+    {
+      "epoch": 0.6292998785916633,
+      "grad_norm": 1.1794949769973755,
+      "learning_rate": 3.782779269048111e-05,
+      "loss": 0.9917,
+      "step": 15550
+    },
+    {
+      "epoch": 0.6295022258195063,
+      "grad_norm": 1.2905328273773193,
+      "learning_rate": 3.7807144332025605e-05,
+      "loss": 1.0404,
+      "step": 15555
+    },
+    {
+      "epoch": 0.6297045730473493,
+      "grad_norm": 1.2904390096664429,
+      "learning_rate": 3.77864959735701e-05,
+      "loss": 1.0027,
+      "step": 15560
+    },
+    {
+      "epoch": 0.6299069202751922,
+      "grad_norm": 1.1533604860305786,
+      "learning_rate": 3.77658476151146e-05,
+      "loss": 0.9814,
+      "step": 15565
+    },
+    {
+      "epoch": 0.6301092675030352,
+      "grad_norm": 1.242135763168335,
+      "learning_rate": 3.77451992566591e-05,
+      "loss": 0.9952,
+      "step": 15570
+    },
+    {
+      "epoch": 0.6303116147308782,
+      "grad_norm": 1.1621266603469849,
+      "learning_rate": 3.7724550898203595e-05,
+      "loss": 0.917,
+      "step": 15575
+    },
+    {
+      "epoch": 0.6305139619587211,
+      "grad_norm": 1.2329875230789185,
+      "learning_rate": 3.770390253974809e-05,
+      "loss": 0.9992,
+      "step": 15580
+    },
+    {
+      "epoch": 0.6307163091865642,
+      "grad_norm": 1.028205156326294,
+      "learning_rate": 3.768325418129259e-05,
+      "loss": 0.9792,
+      "step": 15585
+    },
+    {
+      "epoch": 0.6309186564144071,
+      "grad_norm": 1.2253775596618652,
+      "learning_rate": 3.7662605822837086e-05,
+      "loss": 0.9968,
+      "step": 15590
+    },
+    {
+      "epoch": 0.6311210036422501,
+      "grad_norm": 1.1776041984558105,
+      "learning_rate": 3.7641957464381584e-05,
+      "loss": 0.935,
+      "step": 15595
+    },
+    {
+      "epoch": 0.6313233508700931,
+      "grad_norm": 1.4083956480026245,
+      "learning_rate": 3.762130910592608e-05,
+      "loss": 0.9615,
+      "step": 15600
+    },
+    {
+      "epoch": 0.631525698097936,
+      "grad_norm": 1.0517611503601074,
+      "learning_rate": 3.760066074747058e-05,
+      "loss": 0.965,
+      "step": 15605
+    },
+    {
+      "epoch": 0.6317280453257791,
+      "grad_norm": 1.0910779237747192,
+      "learning_rate": 3.7580012389015076e-05,
+      "loss": 0.9823,
+      "step": 15610
+    },
+    {
+      "epoch": 0.631930392553622,
+      "grad_norm": 1.2836335897445679,
+      "learning_rate": 3.755936403055957e-05,
+      "loss": 0.9022,
+      "step": 15615
+    },
+    {
+      "epoch": 0.632132739781465,
+      "grad_norm": 1.1814792156219482,
+      "learning_rate": 3.753871567210407e-05,
+      "loss": 1.0359,
+      "step": 15620
+    },
+    {
+      "epoch": 0.632335087009308,
+      "grad_norm": 1.1960265636444092,
+      "learning_rate": 3.751806731364857e-05,
+      "loss": 1.0093,
+      "step": 15625
+    },
+    {
+      "epoch": 0.6325374342371509,
+      "grad_norm": 1.1587327718734741,
+      "learning_rate": 3.7497418955193065e-05,
+      "loss": 0.9805,
+      "step": 15630
+    },
+    {
+      "epoch": 0.6327397814649939,
+      "grad_norm": 1.1636066436767578,
+      "learning_rate": 3.747677059673756e-05,
+      "loss": 0.9903,
+      "step": 15635
+    },
+    {
+      "epoch": 0.632942128692837,
+      "grad_norm": 1.295119047164917,
+      "learning_rate": 3.745612223828206e-05,
+      "loss": 1.0165,
+      "step": 15640
+    },
+    {
+      "epoch": 0.6331444759206799,
+      "grad_norm": 1.1014918088912964,
+      "learning_rate": 3.743547387982656e-05,
+      "loss": 1.0494,
+      "step": 15645
+    },
+    {
+      "epoch": 0.6333468231485229,
+      "grad_norm": 1.0557477474212646,
+      "learning_rate": 3.7414825521371054e-05,
+      "loss": 1.0426,
+      "step": 15650
+    },
+    {
+      "epoch": 0.6335491703763658,
+      "grad_norm": 1.279184103012085,
+      "learning_rate": 3.739417716291555e-05,
+      "loss": 1.0112,
+      "step": 15655
+    },
+    {
+      "epoch": 0.6337515176042088,
+      "grad_norm": 1.2059326171875,
+      "learning_rate": 3.737352880446004e-05,
+      "loss": 0.9797,
+      "step": 15660
+    },
+    {
+      "epoch": 0.6339538648320518,
+      "grad_norm": 1.2009028196334839,
+      "learning_rate": 3.7352880446004546e-05,
+      "loss": 0.9616,
+      "step": 15665
+    },
+    {
+      "epoch": 0.6341562120598948,
+      "grad_norm": 1.128832459449768,
+      "learning_rate": 3.733223208754904e-05,
+      "loss": 0.9949,
+      "step": 15670
+    },
+    {
+      "epoch": 0.6343585592877378,
+      "grad_norm": 1.143537163734436,
+      "learning_rate": 3.731158372909354e-05,
+      "loss": 0.9549,
+      "step": 15675
+    },
+    {
+      "epoch": 0.6345609065155807,
+      "grad_norm": 1.2537928819656372,
+      "learning_rate": 3.729093537063804e-05,
+      "loss": 1.0404,
+      "step": 15680
+    },
+    {
+      "epoch": 0.6347632537434237,
+      "grad_norm": 1.2879520654678345,
+      "learning_rate": 3.727028701218253e-05,
+      "loss": 1.0428,
+      "step": 15685
+    },
+    {
+      "epoch": 0.6349656009712666,
+      "grad_norm": 1.1847171783447266,
+      "learning_rate": 3.724963865372703e-05,
+      "loss": 0.9985,
+      "step": 15690
+    },
+    {
+      "epoch": 0.6351679481991097,
+      "grad_norm": 1.1616815328598022,
+      "learning_rate": 3.722899029527153e-05,
+      "loss": 0.9958,
+      "step": 15695
+    },
+    {
+      "epoch": 0.6353702954269527,
+      "grad_norm": 1.2530462741851807,
+      "learning_rate": 3.720834193681603e-05,
+      "loss": 1.0043,
+      "step": 15700
+    },
+    {
+      "epoch": 0.6355726426547956,
+      "grad_norm": 1.148319125175476,
+      "learning_rate": 3.7187693578360524e-05,
+      "loss": 0.9363,
+      "step": 15705
+    },
+    {
+      "epoch": 0.6357749898826386,
+      "grad_norm": 1.1484460830688477,
+      "learning_rate": 3.7167045219905015e-05,
+      "loss": 1.0185,
+      "step": 15710
+    },
+    {
+      "epoch": 0.6359773371104815,
+      "grad_norm": 1.283219575881958,
+      "learning_rate": 3.714639686144952e-05,
+      "loss": 0.9784,
+      "step": 15715
+    },
+    {
+      "epoch": 0.6361796843383246,
+      "grad_norm": 1.1322284936904907,
+      "learning_rate": 3.712574850299401e-05,
+      "loss": 0.997,
+      "step": 15720
+    },
+    {
+      "epoch": 0.6363820315661676,
+      "grad_norm": 1.0504428148269653,
+      "learning_rate": 3.7105100144538513e-05,
+      "loss": 1.0123,
+      "step": 15725
+    },
+    {
+      "epoch": 0.6365843787940105,
+      "grad_norm": 1.1896229982376099,
+      "learning_rate": 3.708445178608301e-05,
+      "loss": 0.9817,
+      "step": 15730
+    },
+    {
+      "epoch": 0.6367867260218535,
+      "grad_norm": 1.301744818687439,
+      "learning_rate": 3.706380342762751e-05,
+      "loss": 1.021,
+      "step": 15735
+    },
+    {
+      "epoch": 0.6369890732496964,
+      "grad_norm": 1.1301766633987427,
+      "learning_rate": 3.7043155069172005e-05,
+      "loss": 1.0073,
+      "step": 15740
+    },
+    {
+      "epoch": 0.6371914204775394,
+      "grad_norm": 1.283570408821106,
+      "learning_rate": 3.7022506710716496e-05,
+      "loss": 0.9424,
+      "step": 15745
+    },
+    {
+      "epoch": 0.6373937677053825,
+      "grad_norm": 1.2190791368484497,
+      "learning_rate": 3.7001858352261e-05,
+      "loss": 0.9692,
+      "step": 15750
+    },
+    {
+      "epoch": 0.6375961149332254,
+      "grad_norm": 1.1517695188522339,
+      "learning_rate": 3.698120999380549e-05,
+      "loss": 0.9715,
+      "step": 15755
+    },
+    {
+      "epoch": 0.6377984621610684,
+      "grad_norm": 1.2132381200790405,
+      "learning_rate": 3.6960561635349995e-05,
+      "loss": 1.0206,
+      "step": 15760
+    },
+    {
+      "epoch": 0.6380008093889113,
+      "grad_norm": 1.2588800191879272,
+      "learning_rate": 3.693991327689449e-05,
+      "loss": 0.9958,
+      "step": 15765
+    },
+    {
+      "epoch": 0.6382031566167543,
+      "grad_norm": 1.2574586868286133,
+      "learning_rate": 3.691926491843898e-05,
+      "loss": 1.0069,
+      "step": 15770
+    },
+    {
+      "epoch": 0.6384055038445974,
+      "grad_norm": 1.2218068838119507,
+      "learning_rate": 3.6898616559983486e-05,
+      "loss": 1.0338,
+      "step": 15775
+    },
+    {
+      "epoch": 0.6386078510724403,
+      "grad_norm": 1.2366846799850464,
+      "learning_rate": 3.687796820152798e-05,
+      "loss": 0.9635,
+      "step": 15780
+    },
+    {
+      "epoch": 0.6388101983002833,
+      "grad_norm": 1.1799321174621582,
+      "learning_rate": 3.685731984307248e-05,
+      "loss": 0.9578,
+      "step": 15785
+    },
+    {
+      "epoch": 0.6390125455281263,
+      "grad_norm": 1.0359095335006714,
+      "learning_rate": 3.683667148461697e-05,
+      "loss": 0.9935,
+      "step": 15790
+    },
+    {
+      "epoch": 0.6392148927559692,
+      "grad_norm": 1.11602783203125,
+      "learning_rate": 3.681602312616147e-05,
+      "loss": 1.0288,
+      "step": 15795
+    },
+    {
+      "epoch": 0.6394172399838122,
+      "grad_norm": 1.1980412006378174,
+      "learning_rate": 3.679537476770597e-05,
+      "loss": 1.0089,
+      "step": 15800
+    },
+    {
+      "epoch": 0.6396195872116552,
+      "grad_norm": 1.22096848487854,
+      "learning_rate": 3.6774726409250463e-05,
+      "loss": 0.9314,
+      "step": 15805
+    },
+    {
+      "epoch": 0.6398219344394982,
+      "grad_norm": 1.2322947978973389,
+      "learning_rate": 3.675407805079497e-05,
+      "loss": 1.0157,
+      "step": 15810
+    },
+    {
+      "epoch": 0.6400242816673412,
+      "grad_norm": 1.2284080982208252,
+      "learning_rate": 3.673342969233946e-05,
+      "loss": 0.9889,
+      "step": 15815
+    },
+    {
+      "epoch": 0.6402266288951841,
+      "grad_norm": 1.2398412227630615,
+      "learning_rate": 3.6712781333883955e-05,
+      "loss": 0.9665,
+      "step": 15820
+    },
+    {
+      "epoch": 0.6404289761230271,
+      "grad_norm": 1.1424891948699951,
+      "learning_rate": 3.669213297542845e-05,
+      "loss": 1.009,
+      "step": 15825
+    },
+    {
+      "epoch": 0.6406313233508701,
+      "grad_norm": 1.210742473602295,
+      "learning_rate": 3.667148461697295e-05,
+      "loss": 1.0247,
+      "step": 15830
+    },
+    {
+      "epoch": 0.6408336705787131,
+      "grad_norm": 1.292371392250061,
+      "learning_rate": 3.6650836258517454e-05,
+      "loss": 0.9835,
+      "step": 15835
+    },
+    {
+      "epoch": 0.6410360178065561,
+      "grad_norm": 1.166009545326233,
+      "learning_rate": 3.6630187900061944e-05,
+      "loss": 0.9475,
+      "step": 15840
+    },
+    {
+      "epoch": 0.641238365034399,
+      "grad_norm": 1.2181354761123657,
+      "learning_rate": 3.660953954160644e-05,
+      "loss": 0.9647,
+      "step": 15845
+    },
+    {
+      "epoch": 0.641440712262242,
+      "grad_norm": 1.1937371492385864,
+      "learning_rate": 3.658889118315094e-05,
+      "loss": 1.0044,
+      "step": 15850
+    },
+    {
+      "epoch": 0.6416430594900849,
+      "grad_norm": 1.2044020891189575,
+      "learning_rate": 3.6568242824695436e-05,
+      "loss": 0.9652,
+      "step": 15855
+    },
+    {
+      "epoch": 0.641845406717928,
+      "grad_norm": 1.1667225360870361,
+      "learning_rate": 3.654759446623994e-05,
+      "loss": 1.0825,
+      "step": 15860
+    },
+    {
+      "epoch": 0.642047753945771,
+      "grad_norm": 1.1601094007492065,
+      "learning_rate": 3.652694610778443e-05,
+      "loss": 0.989,
+      "step": 15865
+    },
+    {
+      "epoch": 0.6422501011736139,
+      "grad_norm": 1.0793761014938354,
+      "learning_rate": 3.6506297749328935e-05,
+      "loss": 0.9882,
+      "step": 15870
+    },
+    {
+      "epoch": 0.6424524484014569,
+      "grad_norm": 1.2860372066497803,
+      "learning_rate": 3.6485649390873426e-05,
+      "loss": 0.9862,
+      "step": 15875
+    },
+    {
+      "epoch": 0.6426547956292998,
+      "grad_norm": 1.1094404458999634,
+      "learning_rate": 3.646500103241792e-05,
+      "loss": 0.9718,
+      "step": 15880
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 1.167775273323059,
+      "learning_rate": 3.644435267396242e-05,
+      "loss": 0.9491,
+      "step": 15885
+    },
+    {
+      "epoch": 0.6430594900849859,
+      "grad_norm": 1.2569758892059326,
+      "learning_rate": 3.642370431550692e-05,
+      "loss": 0.9739,
+      "step": 15890
+    },
+    {
+      "epoch": 0.6432618373128288,
+      "grad_norm": 1.122658371925354,
+      "learning_rate": 3.640305595705142e-05,
+      "loss": 1.0148,
+      "step": 15895
+    },
+    {
+      "epoch": 0.6434641845406718,
+      "grad_norm": 1.1850612163543701,
+      "learning_rate": 3.638240759859591e-05,
+      "loss": 1.0014,
+      "step": 15900
+    },
+    {
+      "epoch": 0.6436665317685147,
+      "grad_norm": 1.2730854749679565,
+      "learning_rate": 3.636175924014041e-05,
+      "loss": 0.9741,
+      "step": 15905
+    },
+    {
+      "epoch": 0.6438688789963577,
+      "grad_norm": 1.190268635749817,
+      "learning_rate": 3.634111088168491e-05,
+      "loss": 1.0011,
+      "step": 15910
+    },
+    {
+      "epoch": 0.6440712262242008,
+      "grad_norm": 1.2439221143722534,
+      "learning_rate": 3.6320462523229404e-05,
+      "loss": 0.9201,
+      "step": 15915
+    },
+    {
+      "epoch": 0.6442735734520437,
+      "grad_norm": 1.273469090461731,
+      "learning_rate": 3.62998141647739e-05,
+      "loss": 1.0155,
+      "step": 15920
+    },
+    {
+      "epoch": 0.6444759206798867,
+      "grad_norm": 1.2768328189849854,
+      "learning_rate": 3.62791658063184e-05,
+      "loss": 1.0094,
+      "step": 15925
+    },
+    {
+      "epoch": 0.6446782679077296,
+      "grad_norm": 1.1529165506362915,
+      "learning_rate": 3.6258517447862896e-05,
+      "loss": 1.0457,
+      "step": 15930
+    },
+    {
+      "epoch": 0.6448806151355726,
+      "grad_norm": 1.2379252910614014,
+      "learning_rate": 3.623786908940739e-05,
+      "loss": 0.9919,
+      "step": 15935
+    },
+    {
+      "epoch": 0.6450829623634157,
+      "grad_norm": 1.090659260749817,
+      "learning_rate": 3.621722073095189e-05,
+      "loss": 1.0371,
+      "step": 15940
+    },
+    {
+      "epoch": 0.6452853095912586,
+      "grad_norm": 1.0815500020980835,
+      "learning_rate": 3.619657237249639e-05,
+      "loss": 0.9732,
+      "step": 15945
+    },
+    {
+      "epoch": 0.6454876568191016,
+      "grad_norm": 1.257379412651062,
+      "learning_rate": 3.6175924014040885e-05,
+      "loss": 0.9469,
+      "step": 15950
+    },
+    {
+      "epoch": 0.6456900040469445,
+      "grad_norm": 1.1595783233642578,
+      "learning_rate": 3.615527565558538e-05,
+      "loss": 1.0029,
+      "step": 15955
+    },
+    {
+      "epoch": 0.6458923512747875,
+      "grad_norm": 1.131295919418335,
+      "learning_rate": 3.613462729712988e-05,
+      "loss": 1.0195,
+      "step": 15960
+    },
+    {
+      "epoch": 0.6460946985026305,
+      "grad_norm": 1.174517035484314,
+      "learning_rate": 3.611397893867438e-05,
+      "loss": 0.9715,
+      "step": 15965
+    },
+    {
+      "epoch": 0.6462970457304735,
+      "grad_norm": 1.2053883075714111,
+      "learning_rate": 3.6093330580218874e-05,
+      "loss": 0.971,
+      "step": 15970
+    },
+    {
+      "epoch": 0.6464993929583165,
+      "grad_norm": 1.260723352432251,
+      "learning_rate": 3.607268222176337e-05,
+      "loss": 0.9793,
+      "step": 15975
+    },
+    {
+      "epoch": 0.6467017401861594,
+      "grad_norm": 1.025206446647644,
+      "learning_rate": 3.605203386330787e-05,
+      "loss": 0.9617,
+      "step": 15980
+    },
+    {
+      "epoch": 0.6469040874140024,
+      "grad_norm": 1.1306520700454712,
+      "learning_rate": 3.6031385504852366e-05,
+      "loss": 0.9441,
+      "step": 15985
+    },
+    {
+      "epoch": 0.6471064346418454,
+      "grad_norm": 1.1041451692581177,
+      "learning_rate": 3.6010737146396863e-05,
+      "loss": 1.0318,
+      "step": 15990
+    },
+    {
+      "epoch": 0.6473087818696884,
+      "grad_norm": 1.3494863510131836,
+      "learning_rate": 3.599008878794136e-05,
+      "loss": 0.9976,
+      "step": 15995
+    },
+    {
+      "epoch": 0.6475111290975314,
+      "grad_norm": 1.1904017925262451,
+      "learning_rate": 3.596944042948586e-05,
+      "loss": 1.0042,
+      "step": 16000
+    },
+    {
+      "epoch": 0.6477134763253743,
+      "grad_norm": 1.2178298234939575,
+      "learning_rate": 3.5948792071030355e-05,
+      "loss": 1.0431,
+      "step": 16005
+    },
+    {
+      "epoch": 0.6479158235532173,
+      "grad_norm": 1.104743242263794,
+      "learning_rate": 3.592814371257485e-05,
+      "loss": 0.9755,
+      "step": 16010
+    },
+    {
+      "epoch": 0.6481181707810603,
+      "grad_norm": 1.3872283697128296,
+      "learning_rate": 3.590749535411935e-05,
+      "loss": 0.9946,
+      "step": 16015
+    },
+    {
+      "epoch": 0.6483205180089032,
+      "grad_norm": 1.269563913345337,
+      "learning_rate": 3.588684699566385e-05,
+      "loss": 1.007,
+      "step": 16020
+    },
+    {
+      "epoch": 0.6485228652367463,
+      "grad_norm": 1.172032356262207,
+      "learning_rate": 3.5866198637208344e-05,
+      "loss": 0.9503,
+      "step": 16025
+    },
+    {
+      "epoch": 0.6487252124645893,
+      "grad_norm": 1.1208676099777222,
+      "learning_rate": 3.584555027875284e-05,
+      "loss": 0.9644,
+      "step": 16030
+    },
+    {
+      "epoch": 0.6489275596924322,
+      "grad_norm": 1.1338741779327393,
+      "learning_rate": 3.582490192029734e-05,
+      "loss": 0.9777,
+      "step": 16035
+    },
+    {
+      "epoch": 0.6491299069202752,
+      "grad_norm": 1.1406617164611816,
+      "learning_rate": 3.580425356184183e-05,
+      "loss": 0.9968,
+      "step": 16040
+    },
+    {
+      "epoch": 0.6493322541481181,
+      "grad_norm": 1.2349566221237183,
+      "learning_rate": 3.5783605203386334e-05,
+      "loss": 0.9214,
+      "step": 16045
+    },
+    {
+      "epoch": 0.6495346013759612,
+      "grad_norm": 1.1992546319961548,
+      "learning_rate": 3.576295684493083e-05,
+      "loss": 0.9601,
+      "step": 16050
+    },
+    {
+      "epoch": 0.6497369486038042,
+      "grad_norm": 1.1756788492202759,
+      "learning_rate": 3.574230848647533e-05,
+      "loss": 0.9593,
+      "step": 16055
+    },
+    {
+      "epoch": 0.6499392958316471,
+      "grad_norm": 1.2423843145370483,
+      "learning_rate": 3.5721660128019826e-05,
+      "loss": 1.0072,
+      "step": 16060
+    },
+    {
+      "epoch": 0.6501416430594901,
+      "grad_norm": 1.2211461067199707,
+      "learning_rate": 3.5701011769564316e-05,
+      "loss": 0.9781,
+      "step": 16065
+    },
+    {
+      "epoch": 0.650343990287333,
+      "grad_norm": 1.2080109119415283,
+      "learning_rate": 3.568036341110882e-05,
+      "loss": 0.9469,
+      "step": 16070
+    },
+    {
+      "epoch": 0.650546337515176,
+      "grad_norm": 1.2381408214569092,
+      "learning_rate": 3.565971505265332e-05,
+      "loss": 0.9369,
+      "step": 16075
+    },
+    {
+      "epoch": 0.6507486847430191,
+      "grad_norm": 1.21284019947052,
+      "learning_rate": 3.5639066694197815e-05,
+      "loss": 1.0199,
+      "step": 16080
+    },
+    {
+      "epoch": 0.650951031970862,
+      "grad_norm": 1.2571697235107422,
+      "learning_rate": 3.561841833574231e-05,
+      "loss": 0.9258,
+      "step": 16085
+    },
+    {
+      "epoch": 0.651153379198705,
+      "grad_norm": 1.1796058416366577,
+      "learning_rate": 3.559776997728681e-05,
+      "loss": 0.9726,
+      "step": 16090
+    },
+    {
+      "epoch": 0.6513557264265479,
+      "grad_norm": 1.285386085510254,
+      "learning_rate": 3.557712161883131e-05,
+      "loss": 0.9783,
+      "step": 16095
+    },
+    {
+      "epoch": 0.6515580736543909,
+      "grad_norm": 1.1672552824020386,
+      "learning_rate": 3.55564732603758e-05,
+      "loss": 0.9669,
+      "step": 16100
+    },
+    {
+      "epoch": 0.651760420882234,
+      "grad_norm": 1.1819517612457275,
+      "learning_rate": 3.55358249019203e-05,
+      "loss": 0.936,
+      "step": 16105
+    },
+    {
+      "epoch": 0.6519627681100769,
+      "grad_norm": 1.2727749347686768,
+      "learning_rate": 3.55151765434648e-05,
+      "loss": 0.9648,
+      "step": 16110
+    },
+    {
+      "epoch": 0.6521651153379199,
+      "grad_norm": 1.2024821043014526,
+      "learning_rate": 3.5494528185009296e-05,
+      "loss": 0.9441,
+      "step": 16115
+    },
+    {
+      "epoch": 0.6523674625657628,
+      "grad_norm": 1.1411044597625732,
+      "learning_rate": 3.547387982655379e-05,
+      "loss": 1.0269,
+      "step": 16120
+    },
+    {
+      "epoch": 0.6525698097936058,
+      "grad_norm": 1.1624733209609985,
+      "learning_rate": 3.5453231468098284e-05,
+      "loss": 0.989,
+      "step": 16125
+    },
+    {
+      "epoch": 0.6527721570214488,
+      "grad_norm": 1.2555545568466187,
+      "learning_rate": 3.543258310964279e-05,
+      "loss": 1.0641,
+      "step": 16130
+    },
+    {
+      "epoch": 0.6529745042492918,
+      "grad_norm": 1.3429213762283325,
+      "learning_rate": 3.541193475118728e-05,
+      "loss": 0.9779,
+      "step": 16135
+    },
+    {
+      "epoch": 0.6531768514771348,
+      "grad_norm": 1.2503955364227295,
+      "learning_rate": 3.539128639273178e-05,
+      "loss": 0.9847,
+      "step": 16140
+    },
+    {
+      "epoch": 0.6533791987049777,
+      "grad_norm": 1.2091625928878784,
+      "learning_rate": 3.537063803427628e-05,
+      "loss": 1.0222,
+      "step": 16145
+    },
+    {
+      "epoch": 0.6535815459328207,
+      "grad_norm": 1.1267467737197876,
+      "learning_rate": 3.534998967582077e-05,
+      "loss": 0.979,
+      "step": 16150
+    },
+    {
+      "epoch": 0.6537838931606637,
+      "grad_norm": 1.1542922258377075,
+      "learning_rate": 3.5329341317365274e-05,
+      "loss": 0.9906,
+      "step": 16155
+    },
+    {
+      "epoch": 0.6539862403885067,
+      "grad_norm": 1.2290841341018677,
+      "learning_rate": 3.5308692958909765e-05,
+      "loss": 0.9737,
+      "step": 16160
+    },
+    {
+      "epoch": 0.6541885876163497,
+      "grad_norm": 1.2749171257019043,
+      "learning_rate": 3.528804460045427e-05,
+      "loss": 0.9836,
+      "step": 16165
+    },
+    {
+      "epoch": 0.6543909348441926,
+      "grad_norm": 1.2310715913772583,
+      "learning_rate": 3.526739624199876e-05,
+      "loss": 1.0134,
+      "step": 16170
+    },
+    {
+      "epoch": 0.6545932820720356,
+      "grad_norm": 1.1757038831710815,
+      "learning_rate": 3.5246747883543257e-05,
+      "loss": 0.9772,
+      "step": 16175
+    },
+    {
+      "epoch": 0.6547956292998786,
+      "grad_norm": 1.2128708362579346,
+      "learning_rate": 3.522609952508776e-05,
+      "loss": 1.0178,
+      "step": 16180
+    },
+    {
+      "epoch": 0.6549979765277215,
+      "grad_norm": 1.2154395580291748,
+      "learning_rate": 3.520545116663225e-05,
+      "loss": 0.9508,
+      "step": 16185
+    },
+    {
+      "epoch": 0.6552003237555646,
+      "grad_norm": 1.1376159191131592,
+      "learning_rate": 3.5184802808176755e-05,
+      "loss": 0.9483,
+      "step": 16190
+    },
+    {
+      "epoch": 0.6554026709834075,
+      "grad_norm": 1.3748664855957031,
+      "learning_rate": 3.5164154449721246e-05,
+      "loss": 1.0099,
+      "step": 16195
+    },
+    {
+      "epoch": 0.6556050182112505,
+      "grad_norm": 1.1054377555847168,
+      "learning_rate": 3.514350609126575e-05,
+      "loss": 0.9903,
+      "step": 16200
+    },
+    {
+      "epoch": 0.6558073654390935,
+      "grad_norm": 1.235331416130066,
+      "learning_rate": 3.512285773281024e-05,
+      "loss": 1.0068,
+      "step": 16205
+    },
+    {
+      "epoch": 0.6560097126669364,
+      "grad_norm": 1.1809173822402954,
+      "learning_rate": 3.510220937435474e-05,
+      "loss": 1.0089,
+      "step": 16210
+    },
+    {
+      "epoch": 0.6562120598947795,
+      "grad_norm": 1.1006264686584473,
+      "learning_rate": 3.508156101589924e-05,
+      "loss": 0.9753,
+      "step": 16215
+    },
+    {
+      "epoch": 0.6564144071226224,
+      "grad_norm": 1.0573210716247559,
+      "learning_rate": 3.506091265744373e-05,
+      "loss": 0.9545,
+      "step": 16220
+    },
+    {
+      "epoch": 0.6566167543504654,
+      "grad_norm": 1.215930700302124,
+      "learning_rate": 3.5040264298988236e-05,
+      "loss": 1.0469,
+      "step": 16225
+    },
+    {
+      "epoch": 0.6568191015783084,
+      "grad_norm": 1.1159571409225464,
+      "learning_rate": 3.501961594053273e-05,
+      "loss": 0.9961,
+      "step": 16230
+    },
+    {
+      "epoch": 0.6570214488061513,
+      "grad_norm": 1.2668739557266235,
+      "learning_rate": 3.4998967582077224e-05,
+      "loss": 0.9515,
+      "step": 16235
+    },
+    {
+      "epoch": 0.6572237960339944,
+      "grad_norm": 1.1844149827957153,
+      "learning_rate": 3.497831922362173e-05,
+      "loss": 1.0669,
+      "step": 16240
+    },
+    {
+      "epoch": 0.6574261432618373,
+      "grad_norm": 1.2933884859085083,
+      "learning_rate": 3.495767086516622e-05,
+      "loss": 1.0009,
+      "step": 16245
+    },
+    {
+      "epoch": 0.6576284904896803,
+      "grad_norm": 1.1424107551574707,
+      "learning_rate": 3.493702250671072e-05,
+      "loss": 0.9989,
+      "step": 16250
+    },
+    {
+      "epoch": 0.6578308377175233,
+      "grad_norm": 1.184024691581726,
+      "learning_rate": 3.491637414825521e-05,
+      "loss": 1.0101,
+      "step": 16255
+    },
+    {
+      "epoch": 0.6580331849453662,
+      "grad_norm": 1.1887484788894653,
+      "learning_rate": 3.489572578979971e-05,
+      "loss": 0.9734,
+      "step": 16260
+    },
+    {
+      "epoch": 0.6582355321732092,
+      "grad_norm": 1.2848747968673706,
+      "learning_rate": 3.487507743134421e-05,
+      "loss": 1.0054,
+      "step": 16265
+    },
+    {
+      "epoch": 0.6584378794010523,
+      "grad_norm": 1.1876070499420166,
+      "learning_rate": 3.4854429072888705e-05,
+      "loss": 0.9477,
+      "step": 16270
+    },
+    {
+      "epoch": 0.6586402266288952,
+      "grad_norm": 1.2064857482910156,
+      "learning_rate": 3.483378071443321e-05,
+      "loss": 0.9756,
+      "step": 16275
+    },
+    {
+      "epoch": 0.6588425738567382,
+      "grad_norm": 1.1177282333374023,
+      "learning_rate": 3.48131323559777e-05,
+      "loss": 0.9524,
+      "step": 16280
+    },
+    {
+      "epoch": 0.6590449210845811,
+      "grad_norm": 1.2096445560455322,
+      "learning_rate": 3.47924839975222e-05,
+      "loss": 1.0441,
+      "step": 16285
+    },
+    {
+      "epoch": 0.6592472683124241,
+      "grad_norm": 1.1764529943466187,
+      "learning_rate": 3.4771835639066694e-05,
+      "loss": 0.9431,
+      "step": 16290
+    },
+    {
+      "epoch": 0.6594496155402672,
+      "grad_norm": 1.1835061311721802,
+      "learning_rate": 3.475118728061119e-05,
+      "loss": 0.9521,
+      "step": 16295
+    },
+    {
+      "epoch": 0.6596519627681101,
+      "grad_norm": 1.1990066766738892,
+      "learning_rate": 3.473053892215569e-05,
+      "loss": 1.0344,
+      "step": 16300
+    },
+    {
+      "epoch": 0.6598543099959531,
+      "grad_norm": 1.2333585023880005,
+      "learning_rate": 3.4709890563700186e-05,
+      "loss": 1.0388,
+      "step": 16305
+    },
+    {
+      "epoch": 0.660056657223796,
+      "grad_norm": 1.1921073198318481,
+      "learning_rate": 3.4689242205244684e-05,
+      "loss": 0.958,
+      "step": 16310
+    },
+    {
+      "epoch": 0.660259004451639,
+      "grad_norm": 1.2137154340744019,
+      "learning_rate": 3.466859384678918e-05,
+      "loss": 1.024,
+      "step": 16315
+    },
+    {
+      "epoch": 0.660461351679482,
+      "grad_norm": 1.1804203987121582,
+      "learning_rate": 3.464794548833368e-05,
+      "loss": 1.0156,
+      "step": 16320
+    },
+    {
+      "epoch": 0.660663698907325,
+      "grad_norm": 1.2474133968353271,
+      "learning_rate": 3.4627297129878176e-05,
+      "loss": 0.9971,
+      "step": 16325
+    },
+    {
+      "epoch": 0.660866046135168,
+      "grad_norm": 1.1229432821273804,
+      "learning_rate": 3.460664877142267e-05,
+      "loss": 0.9787,
+      "step": 16330
+    },
+    {
+      "epoch": 0.6610683933630109,
+      "grad_norm": 1.2054779529571533,
+      "learning_rate": 3.458600041296717e-05,
+      "loss": 0.9773,
+      "step": 16335
+    },
+    {
+      "epoch": 0.6612707405908539,
+      "grad_norm": 1.221892237663269,
+      "learning_rate": 3.456535205451167e-05,
+      "loss": 0.9425,
+      "step": 16340
+    },
+    {
+      "epoch": 0.6614730878186968,
+      "grad_norm": 1.3123283386230469,
+      "learning_rate": 3.4544703696056165e-05,
+      "loss": 1.0315,
+      "step": 16345
+    },
+    {
+      "epoch": 0.6616754350465399,
+      "grad_norm": 1.3071208000183105,
+      "learning_rate": 3.452405533760066e-05,
+      "loss": 1.0287,
+      "step": 16350
+    },
+    {
+      "epoch": 0.6618777822743829,
+      "grad_norm": 1.1664862632751465,
+      "learning_rate": 3.450340697914516e-05,
+      "loss": 1.0081,
+      "step": 16355
+    },
+    {
+      "epoch": 0.6620801295022258,
+      "grad_norm": 1.2433518171310425,
+      "learning_rate": 3.4482758620689657e-05,
+      "loss": 1.0246,
+      "step": 16360
+    },
+    {
+      "epoch": 0.6622824767300688,
+      "grad_norm": 1.2965155839920044,
+      "learning_rate": 3.4462110262234154e-05,
+      "loss": 0.9523,
+      "step": 16365
+    },
+    {
+      "epoch": 0.6624848239579118,
+      "grad_norm": 1.2354497909545898,
+      "learning_rate": 3.444146190377865e-05,
+      "loss": 1.0091,
+      "step": 16370
+    },
+    {
+      "epoch": 0.6626871711857547,
+      "grad_norm": 1.1297180652618408,
+      "learning_rate": 3.442081354532315e-05,
+      "loss": 1.0102,
+      "step": 16375
+    },
+    {
+      "epoch": 0.6628895184135978,
+      "grad_norm": 1.1552640199661255,
+      "learning_rate": 3.4400165186867646e-05,
+      "loss": 0.976,
+      "step": 16380
+    },
+    {
+      "epoch": 0.6630918656414407,
+      "grad_norm": 1.1431080102920532,
+      "learning_rate": 3.437951682841214e-05,
+      "loss": 0.9939,
+      "step": 16385
+    },
+    {
+      "epoch": 0.6632942128692837,
+      "grad_norm": 1.3781251907348633,
+      "learning_rate": 3.435886846995664e-05,
+      "loss": 0.98,
+      "step": 16390
+    },
+    {
+      "epoch": 0.6634965600971267,
+      "grad_norm": 1.2489134073257446,
+      "learning_rate": 3.433822011150114e-05,
+      "loss": 0.9347,
+      "step": 16395
+    },
+    {
+      "epoch": 0.6636989073249696,
+      "grad_norm": 1.084082841873169,
+      "learning_rate": 3.4317571753045635e-05,
+      "loss": 0.9716,
+      "step": 16400
+    },
+    {
+      "epoch": 0.6639012545528127,
+      "grad_norm": 1.173354148864746,
+      "learning_rate": 3.429692339459013e-05,
+      "loss": 0.9983,
+      "step": 16405
+    },
+    {
+      "epoch": 0.6641036017806556,
+      "grad_norm": 1.2579299211502075,
+      "learning_rate": 3.427627503613463e-05,
+      "loss": 0.9981,
+      "step": 16410
+    },
+    {
+      "epoch": 0.6643059490084986,
+      "grad_norm": 1.2689281702041626,
+      "learning_rate": 3.425562667767913e-05,
+      "loss": 1.0367,
+      "step": 16415
+    },
+    {
+      "epoch": 0.6645082962363416,
+      "grad_norm": 1.1956695318222046,
+      "learning_rate": 3.423497831922362e-05,
+      "loss": 1.0017,
+      "step": 16420
+    },
+    {
+      "epoch": 0.6647106434641845,
+      "grad_norm": 1.1953344345092773,
+      "learning_rate": 3.421432996076812e-05,
+      "loss": 0.9572,
+      "step": 16425
+    },
+    {
+      "epoch": 0.6649129906920275,
+      "grad_norm": 1.1050218343734741,
+      "learning_rate": 3.419368160231262e-05,
+      "loss": 0.9739,
+      "step": 16430
+    },
+    {
+      "epoch": 0.6651153379198705,
+      "grad_norm": 1.2522861957550049,
+      "learning_rate": 3.4173033243857116e-05,
+      "loss": 0.9683,
+      "step": 16435
+    },
+    {
+      "epoch": 0.6653176851477135,
+      "grad_norm": 1.4148693084716797,
+      "learning_rate": 3.415238488540161e-05,
+      "loss": 0.9713,
+      "step": 16440
+    },
+    {
+      "epoch": 0.6655200323755565,
+      "grad_norm": 1.3018653392791748,
+      "learning_rate": 3.413173652694611e-05,
+      "loss": 0.9295,
+      "step": 16445
+    },
+    {
+      "epoch": 0.6657223796033994,
+      "grad_norm": 1.2819592952728271,
+      "learning_rate": 3.411108816849061e-05,
+      "loss": 0.973,
+      "step": 16450
+    },
+    {
+      "epoch": 0.6659247268312424,
+      "grad_norm": 1.1259995698928833,
+      "learning_rate": 3.4090439810035105e-05,
+      "loss": 0.9863,
+      "step": 16455
+    },
+    {
+      "epoch": 0.6661270740590854,
+      "grad_norm": 1.1015294790267944,
+      "learning_rate": 3.40697914515796e-05,
+      "loss": 1.0036,
+      "step": 16460
+    },
+    {
+      "epoch": 0.6663294212869284,
+      "grad_norm": 1.012439489364624,
+      "learning_rate": 3.40491430931241e-05,
+      "loss": 1.0279,
+      "step": 16465
+    },
+    {
+      "epoch": 0.6665317685147714,
+      "grad_norm": 1.157087802886963,
+      "learning_rate": 3.40284947346686e-05,
+      "loss": 1.0178,
+      "step": 16470
+    },
+    {
+      "epoch": 0.6667341157426143,
+      "grad_norm": 1.246281623840332,
+      "learning_rate": 3.4007846376213094e-05,
+      "loss": 1.0249,
+      "step": 16475
+    },
+    {
+      "epoch": 0.6669364629704573,
+      "grad_norm": 1.208592414855957,
+      "learning_rate": 3.3987198017757585e-05,
+      "loss": 0.9562,
+      "step": 16480
+    },
+    {
+      "epoch": 0.6671388101983002,
+      "grad_norm": 1.1817739009857178,
+      "learning_rate": 3.396654965930209e-05,
+      "loss": 1.0095,
+      "step": 16485
+    },
+    {
+      "epoch": 0.6673411574261433,
+      "grad_norm": 1.2455251216888428,
+      "learning_rate": 3.3945901300846586e-05,
+      "loss": 1.0175,
+      "step": 16490
+    },
+    {
+      "epoch": 0.6675435046539863,
+      "grad_norm": 1.1381094455718994,
+      "learning_rate": 3.3925252942391084e-05,
+      "loss": 1.0171,
+      "step": 16495
+    },
+    {
+      "epoch": 0.6677458518818292,
+      "grad_norm": 1.1386624574661255,
+      "learning_rate": 3.390460458393558e-05,
+      "loss": 1.0674,
+      "step": 16500
+    },
+    {
+      "epoch": 0.6679481991096722,
+      "grad_norm": 1.1103817224502563,
+      "learning_rate": 3.388395622548007e-05,
+      "loss": 0.9728,
+      "step": 16505
+    },
+    {
+      "epoch": 0.6681505463375151,
+      "grad_norm": 1.268618106842041,
+      "learning_rate": 3.3863307867024575e-05,
+      "loss": 1.006,
+      "step": 16510
+    },
+    {
+      "epoch": 0.6683528935653582,
+      "grad_norm": 1.1923140287399292,
+      "learning_rate": 3.3842659508569066e-05,
+      "loss": 0.9763,
+      "step": 16515
+    },
+    {
+      "epoch": 0.6685552407932012,
+      "grad_norm": 1.3593353033065796,
+      "learning_rate": 3.382201115011357e-05,
+      "loss": 1.0071,
+      "step": 16520
+    },
+    {
+      "epoch": 0.6687575880210441,
+      "grad_norm": 1.3046915531158447,
+      "learning_rate": 3.380136279165807e-05,
+      "loss": 1.0135,
+      "step": 16525
+    },
+    {
+      "epoch": 0.6689599352488871,
+      "grad_norm": 1.217355489730835,
+      "learning_rate": 3.378071443320256e-05,
+      "loss": 0.9486,
+      "step": 16530
+    },
+    {
+      "epoch": 0.66916228247673,
+      "grad_norm": 1.1550461053848267,
+      "learning_rate": 3.376006607474706e-05,
+      "loss": 0.9181,
+      "step": 16535
+    },
+    {
+      "epoch": 0.669364629704573,
+      "grad_norm": 1.2547290325164795,
+      "learning_rate": 3.373941771629155e-05,
+      "loss": 1.0339,
+      "step": 16540
+    },
+    {
+      "epoch": 0.6695669769324161,
+      "grad_norm": 1.3131098747253418,
+      "learning_rate": 3.3718769357836057e-05,
+      "loss": 0.9527,
+      "step": 16545
+    },
+    {
+      "epoch": 0.669769324160259,
+      "grad_norm": 1.1808663606643677,
+      "learning_rate": 3.369812099938055e-05,
+      "loss": 1.0197,
+      "step": 16550
+    },
+    {
+      "epoch": 0.669971671388102,
+      "grad_norm": 1.25763738155365,
+      "learning_rate": 3.367747264092505e-05,
+      "loss": 0.9644,
+      "step": 16555
+    },
+    {
+      "epoch": 0.6701740186159449,
+      "grad_norm": 1.1556329727172852,
+      "learning_rate": 3.365682428246955e-05,
+      "loss": 0.9838,
+      "step": 16560
+    },
+    {
+      "epoch": 0.6703763658437879,
+      "grad_norm": 1.2575442790985107,
+      "learning_rate": 3.363617592401404e-05,
+      "loss": 0.9805,
+      "step": 16565
+    },
+    {
+      "epoch": 0.670578713071631,
+      "grad_norm": 1.1899510622024536,
+      "learning_rate": 3.361552756555854e-05,
+      "loss": 0.9748,
+      "step": 16570
+    },
+    {
+      "epoch": 0.6707810602994739,
+      "grad_norm": 1.3026821613311768,
+      "learning_rate": 3.3594879207103034e-05,
+      "loss": 0.9706,
+      "step": 16575
+    },
+    {
+      "epoch": 0.6709834075273169,
+      "grad_norm": 1.2790369987487793,
+      "learning_rate": 3.357423084864754e-05,
+      "loss": 1.0038,
+      "step": 16580
+    },
+    {
+      "epoch": 0.6711857547551598,
+      "grad_norm": 1.2287906408309937,
+      "learning_rate": 3.355358249019203e-05,
+      "loss": 1.0712,
+      "step": 16585
+    },
+    {
+      "epoch": 0.6713881019830028,
+      "grad_norm": 1.1447017192840576,
+      "learning_rate": 3.3532934131736525e-05,
+      "loss": 1.0111,
+      "step": 16590
+    },
+    {
+      "epoch": 0.6715904492108458,
+      "grad_norm": 1.4328875541687012,
+      "learning_rate": 3.351228577328103e-05,
+      "loss": 1.0095,
+      "step": 16595
+    },
+    {
+      "epoch": 0.6717927964386888,
+      "grad_norm": 1.0844640731811523,
+      "learning_rate": 3.349163741482552e-05,
+      "loss": 0.9768,
+      "step": 16600
+    },
+    {
+      "epoch": 0.6719951436665318,
+      "grad_norm": 1.2308974266052246,
+      "learning_rate": 3.3470989056370024e-05,
+      "loss": 1.094,
+      "step": 16605
+    },
+    {
+      "epoch": 0.6721974908943747,
+      "grad_norm": 1.1745285987854004,
+      "learning_rate": 3.3450340697914515e-05,
+      "loss": 1.0043,
+      "step": 16610
+    },
+    {
+      "epoch": 0.6723998381222177,
+      "grad_norm": 1.2722959518432617,
+      "learning_rate": 3.342969233945901e-05,
+      "loss": 0.9747,
+      "step": 16615
+    },
+    {
+      "epoch": 0.6726021853500607,
+      "grad_norm": 1.318652868270874,
+      "learning_rate": 3.3409043981003516e-05,
+      "loss": 1.0501,
+      "step": 16620
+    },
+    {
+      "epoch": 0.6728045325779037,
+      "grad_norm": 1.0574398040771484,
+      "learning_rate": 3.3388395622548007e-05,
+      "loss": 0.9792,
+      "step": 16625
+    },
+    {
+      "epoch": 0.6730068798057467,
+      "grad_norm": 1.241368055343628,
+      "learning_rate": 3.336774726409251e-05,
+      "loss": 0.9847,
+      "step": 16630
+    },
+    {
+      "epoch": 0.6732092270335897,
+      "grad_norm": 1.2177891731262207,
+      "learning_rate": 3.3347098905637e-05,
+      "loss": 1.0106,
+      "step": 16635
+    },
+    {
+      "epoch": 0.6734115742614326,
+      "grad_norm": 1.1439677476882935,
+      "learning_rate": 3.33264505471815e-05,
+      "loss": 1.0239,
+      "step": 16640
+    },
+    {
+      "epoch": 0.6736139214892756,
+      "grad_norm": 1.3419727087020874,
+      "learning_rate": 3.3305802188725996e-05,
+      "loss": 0.9621,
+      "step": 16645
+    },
+    {
+      "epoch": 0.6738162687171185,
+      "grad_norm": 1.226435661315918,
+      "learning_rate": 3.328515383027049e-05,
+      "loss": 1.0139,
+      "step": 16650
+    },
+    {
+      "epoch": 0.6740186159449616,
+      "grad_norm": 1.2880449295043945,
+      "learning_rate": 3.3264505471815e-05,
+      "loss": 0.9398,
+      "step": 16655
+    },
+    {
+      "epoch": 0.6742209631728046,
+      "grad_norm": 1.1530487537384033,
+      "learning_rate": 3.324385711335949e-05,
+      "loss": 1.005,
+      "step": 16660
+    },
+    {
+      "epoch": 0.6744233104006475,
+      "grad_norm": 1.1972562074661255,
+      "learning_rate": 3.322320875490399e-05,
+      "loss": 0.9762,
+      "step": 16665
+    },
+    {
+      "epoch": 0.6746256576284905,
+      "grad_norm": 1.2207541465759277,
+      "learning_rate": 3.320256039644848e-05,
+      "loss": 0.9496,
+      "step": 16670
+    },
+    {
+      "epoch": 0.6748280048563334,
+      "grad_norm": 1.1928716897964478,
+      "learning_rate": 3.318191203799298e-05,
+      "loss": 1.0167,
+      "step": 16675
+    },
+    {
+      "epoch": 0.6750303520841765,
+      "grad_norm": 1.2012739181518555,
+      "learning_rate": 3.316126367953748e-05,
+      "loss": 0.9725,
+      "step": 16680
+    },
+    {
+      "epoch": 0.6752326993120195,
+      "grad_norm": 1.3488842248916626,
+      "learning_rate": 3.3140615321081974e-05,
+      "loss": 0.9615,
+      "step": 16685
+    },
+    {
+      "epoch": 0.6754350465398624,
+      "grad_norm": 1.1975066661834717,
+      "learning_rate": 3.311996696262648e-05,
+      "loss": 0.9524,
+      "step": 16690
+    },
+    {
+      "epoch": 0.6756373937677054,
+      "grad_norm": 1.2017061710357666,
+      "learning_rate": 3.309931860417097e-05,
+      "loss": 0.9714,
+      "step": 16695
+    },
+    {
+      "epoch": 0.6758397409955483,
+      "grad_norm": 1.1276466846466064,
+      "learning_rate": 3.3078670245715466e-05,
+      "loss": 0.9561,
+      "step": 16700
+    },
+    {
+      "epoch": 0.6760420882233913,
+      "grad_norm": 1.1216696500778198,
+      "learning_rate": 3.305802188725996e-05,
+      "loss": 0.9854,
+      "step": 16705
+    },
+    {
+      "epoch": 0.6762444354512344,
+      "grad_norm": 1.143851399421692,
+      "learning_rate": 3.303737352880446e-05,
+      "loss": 1.0552,
+      "step": 16710
+    },
+    {
+      "epoch": 0.6764467826790773,
+      "grad_norm": 1.0941919088363647,
+      "learning_rate": 3.301672517034896e-05,
+      "loss": 1.0036,
+      "step": 16715
+    },
+    {
+      "epoch": 0.6766491299069203,
+      "grad_norm": 1.1506640911102295,
+      "learning_rate": 3.2996076811893455e-05,
+      "loss": 0.9902,
+      "step": 16720
+    },
+    {
+      "epoch": 0.6768514771347632,
+      "grad_norm": 1.1210603713989258,
+      "learning_rate": 3.297542845343795e-05,
+      "loss": 0.9526,
+      "step": 16725
+    },
+    {
+      "epoch": 0.6770538243626062,
+      "grad_norm": 1.1224634647369385,
+      "learning_rate": 3.295478009498245e-05,
+      "loss": 0.9588,
+      "step": 16730
+    },
+    {
+      "epoch": 0.6772561715904493,
+      "grad_norm": 1.122049331665039,
+      "learning_rate": 3.293413173652695e-05,
+      "loss": 1.0602,
+      "step": 16735
+    },
+    {
+      "epoch": 0.6774585188182922,
+      "grad_norm": 1.176449179649353,
+      "learning_rate": 3.2913483378071444e-05,
+      "loss": 0.9677,
+      "step": 16740
+    },
+    {
+      "epoch": 0.6776608660461352,
+      "grad_norm": 1.2817918062210083,
+      "learning_rate": 3.289283501961594e-05,
+      "loss": 1.0685,
+      "step": 16745
+    },
+    {
+      "epoch": 0.6778632132739781,
+      "grad_norm": 1.1789745092391968,
+      "learning_rate": 3.287218666116044e-05,
+      "loss": 0.976,
+      "step": 16750
+    },
+    {
+      "epoch": 0.6780655605018211,
+      "grad_norm": 1.095564842224121,
+      "learning_rate": 3.2851538302704936e-05,
+      "loss": 0.9916,
+      "step": 16755
+    },
+    {
+      "epoch": 0.678267907729664,
+      "grad_norm": 1.1979174613952637,
+      "learning_rate": 3.2830889944249434e-05,
+      "loss": 1.0478,
+      "step": 16760
+    },
+    {
+      "epoch": 0.6784702549575071,
+      "grad_norm": 1.3009766340255737,
+      "learning_rate": 3.281024158579393e-05,
+      "loss": 0.9987,
+      "step": 16765
+    },
+    {
+      "epoch": 0.6786726021853501,
+      "grad_norm": 1.1497248411178589,
+      "learning_rate": 3.278959322733843e-05,
+      "loss": 0.9367,
+      "step": 16770
+    },
+    {
+      "epoch": 0.678874949413193,
+      "grad_norm": 1.120176076889038,
+      "learning_rate": 3.2768944868882925e-05,
+      "loss": 1.019,
+      "step": 16775
+    },
+    {
+      "epoch": 0.679077296641036,
+      "grad_norm": 1.2681643962860107,
+      "learning_rate": 3.274829651042742e-05,
+      "loss": 0.9813,
+      "step": 16780
+    },
+    {
+      "epoch": 0.679279643868879,
+      "grad_norm": 1.1057486534118652,
+      "learning_rate": 3.272764815197192e-05,
+      "loss": 0.9439,
+      "step": 16785
+    },
+    {
+      "epoch": 0.679481991096722,
+      "grad_norm": 1.1760663986206055,
+      "learning_rate": 3.270699979351642e-05,
+      "loss": 0.9532,
+      "step": 16790
+    },
+    {
+      "epoch": 0.679684338324565,
+      "grad_norm": 1.2541499137878418,
+      "learning_rate": 3.2686351435060915e-05,
+      "loss": 0.9787,
+      "step": 16795
+    },
+    {
+      "epoch": 0.6798866855524079,
+      "grad_norm": 1.2193784713745117,
+      "learning_rate": 3.266570307660541e-05,
+      "loss": 0.9828,
+      "step": 16800
+    },
+    {
+      "epoch": 0.6800890327802509,
+      "grad_norm": 1.1541152000427246,
+      "learning_rate": 3.264505471814991e-05,
+      "loss": 1.0061,
+      "step": 16805
+    },
+    {
+      "epoch": 0.6802913800080939,
+      "grad_norm": 1.0843942165374756,
+      "learning_rate": 3.2624406359694407e-05,
+      "loss": 0.9739,
+      "step": 16810
+    },
+    {
+      "epoch": 0.6804937272359368,
+      "grad_norm": 1.2362585067749023,
+      "learning_rate": 3.2603758001238904e-05,
+      "loss": 0.9515,
+      "step": 16815
+    },
+    {
+      "epoch": 0.6806960744637799,
+      "grad_norm": 1.2097837924957275,
+      "learning_rate": 3.25831096427834e-05,
+      "loss": 0.9912,
+      "step": 16820
+    },
+    {
+      "epoch": 0.6808984216916228,
+      "grad_norm": 1.1431483030319214,
+      "learning_rate": 3.25624612843279e-05,
+      "loss": 0.9931,
+      "step": 16825
+    },
+    {
+      "epoch": 0.6811007689194658,
+      "grad_norm": 1.1017531156539917,
+      "learning_rate": 3.2541812925872396e-05,
+      "loss": 0.9941,
+      "step": 16830
+    },
+    {
+      "epoch": 0.6813031161473088,
+      "grad_norm": 1.1678105592727661,
+      "learning_rate": 3.252116456741689e-05,
+      "loss": 0.9737,
+      "step": 16835
+    },
+    {
+      "epoch": 0.6815054633751517,
+      "grad_norm": 1.2099162340164185,
+      "learning_rate": 3.250051620896139e-05,
+      "loss": 0.9884,
+      "step": 16840
+    },
+    {
+      "epoch": 0.6817078106029948,
+      "grad_norm": 1.3309848308563232,
+      "learning_rate": 3.247986785050589e-05,
+      "loss": 1.0069,
+      "step": 16845
+    },
+    {
+      "epoch": 0.6819101578308377,
+      "grad_norm": 1.2974705696105957,
+      "learning_rate": 3.2459219492050385e-05,
+      "loss": 1.0386,
+      "step": 16850
+    },
+    {
+      "epoch": 0.6821125050586807,
+      "grad_norm": 1.214913249015808,
+      "learning_rate": 3.243857113359488e-05,
+      "loss": 1.0237,
+      "step": 16855
+    },
+    {
+      "epoch": 0.6823148522865237,
+      "grad_norm": 1.3377631902694702,
+      "learning_rate": 3.241792277513937e-05,
+      "loss": 1.0285,
+      "step": 16860
+    },
+    {
+      "epoch": 0.6825171995143666,
+      "grad_norm": 1.264298439025879,
+      "learning_rate": 3.239727441668388e-05,
+      "loss": 1.0209,
+      "step": 16865
+    },
+    {
+      "epoch": 0.6827195467422096,
+      "grad_norm": 1.1477164030075073,
+      "learning_rate": 3.2376626058228374e-05,
+      "loss": 0.9898,
+      "step": 16870
+    },
+    {
+      "epoch": 0.6829218939700527,
+      "grad_norm": 1.2462587356567383,
+      "learning_rate": 3.235597769977287e-05,
+      "loss": 0.9754,
+      "step": 16875
+    },
+    {
+      "epoch": 0.6831242411978956,
+      "grad_norm": 1.2337095737457275,
+      "learning_rate": 3.233532934131737e-05,
+      "loss": 0.9265,
+      "step": 16880
+    },
+    {
+      "epoch": 0.6833265884257386,
+      "grad_norm": 1.1416597366333008,
+      "learning_rate": 3.231468098286186e-05,
+      "loss": 1.0265,
+      "step": 16885
+    },
+    {
+      "epoch": 0.6835289356535815,
+      "grad_norm": 1.1606651544570923,
+      "learning_rate": 3.229403262440636e-05,
+      "loss": 0.9875,
+      "step": 16890
+    },
+    {
+      "epoch": 0.6837312828814245,
+      "grad_norm": 1.233574628829956,
+      "learning_rate": 3.2273384265950854e-05,
+      "loss": 0.9783,
+      "step": 16895
+    },
+    {
+      "epoch": 0.6839336301092676,
+      "grad_norm": 1.2189573049545288,
+      "learning_rate": 3.225273590749536e-05,
+      "loss": 1.0409,
+      "step": 16900
+    },
+    {
+      "epoch": 0.6841359773371105,
+      "grad_norm": 1.1458240747451782,
+      "learning_rate": 3.2232087549039855e-05,
+      "loss": 0.9379,
+      "step": 16905
+    },
+    {
+      "epoch": 0.6843383245649535,
+      "grad_norm": 1.2293412685394287,
+      "learning_rate": 3.221143919058435e-05,
+      "loss": 1.0584,
+      "step": 16910
+    },
+    {
+      "epoch": 0.6845406717927964,
+      "grad_norm": 1.1189359426498413,
+      "learning_rate": 3.219079083212885e-05,
+      "loss": 0.9614,
+      "step": 16915
+    },
+    {
+      "epoch": 0.6847430190206394,
+      "grad_norm": 1.1481940746307373,
+      "learning_rate": 3.217014247367334e-05,
+      "loss": 0.9941,
+      "step": 16920
+    },
+    {
+      "epoch": 0.6849453662484823,
+      "grad_norm": 1.26361882686615,
+      "learning_rate": 3.2149494115217844e-05,
+      "loss": 0.9524,
+      "step": 16925
+    },
+    {
+      "epoch": 0.6851477134763254,
+      "grad_norm": 1.1836000680923462,
+      "learning_rate": 3.2128845756762335e-05,
+      "loss": 0.9901,
+      "step": 16930
+    },
+    {
+      "epoch": 0.6853500607041684,
+      "grad_norm": 1.1145380735397339,
+      "learning_rate": 3.210819739830684e-05,
+      "loss": 0.9956,
+      "step": 16935
+    },
+    {
+      "epoch": 0.6855524079320113,
+      "grad_norm": 1.091498851776123,
+      "learning_rate": 3.2087549039851336e-05,
+      "loss": 1.0457,
+      "step": 16940
+    },
+    {
+      "epoch": 0.6857547551598543,
+      "grad_norm": 1.1746047735214233,
+      "learning_rate": 3.206690068139583e-05,
+      "loss": 0.9925,
+      "step": 16945
+    },
+    {
+      "epoch": 0.6859571023876972,
+      "grad_norm": 1.166630506515503,
+      "learning_rate": 3.204625232294033e-05,
+      "loss": 0.9864,
+      "step": 16950
+    },
+    {
+      "epoch": 0.6861594496155403,
+      "grad_norm": 1.2178932428359985,
+      "learning_rate": 3.202560396448482e-05,
+      "loss": 0.9179,
+      "step": 16955
+    },
+    {
+      "epoch": 0.6863617968433833,
+      "grad_norm": 1.2030609846115112,
+      "learning_rate": 3.2004955606029325e-05,
+      "loss": 1.0445,
+      "step": 16960
+    },
+    {
+      "epoch": 0.6865641440712262,
+      "grad_norm": 1.319400668144226,
+      "learning_rate": 3.1984307247573816e-05,
+      "loss": 0.977,
+      "step": 16965
+    },
+    {
+      "epoch": 0.6867664912990692,
+      "grad_norm": 1.1344331502914429,
+      "learning_rate": 3.196365888911831e-05,
+      "loss": 0.9348,
+      "step": 16970
+    },
+    {
+      "epoch": 0.6869688385269122,
+      "grad_norm": 1.2260794639587402,
+      "learning_rate": 3.194301053066282e-05,
+      "loss": 1.0605,
+      "step": 16975
+    },
+    {
+      "epoch": 0.6871711857547551,
+      "grad_norm": 1.1607056856155396,
+      "learning_rate": 3.192236217220731e-05,
+      "loss": 1.0135,
+      "step": 16980
+    },
+    {
+      "epoch": 0.6873735329825982,
+      "grad_norm": 1.086566686630249,
+      "learning_rate": 3.190171381375181e-05,
+      "loss": 0.9789,
+      "step": 16985
+    },
+    {
+      "epoch": 0.6875758802104411,
+      "grad_norm": 1.1593726873397827,
+      "learning_rate": 3.18810654552963e-05,
+      "loss": 0.9674,
+      "step": 16990
+    },
+    {
+      "epoch": 0.6877782274382841,
+      "grad_norm": 1.2725913524627686,
+      "learning_rate": 3.18604170968408e-05,
+      "loss": 0.9664,
+      "step": 16995
+    },
+    {
+      "epoch": 0.687980574666127,
+      "grad_norm": 1.1086002588272095,
+      "learning_rate": 3.1839768738385304e-05,
+      "loss": 0.9972,
+      "step": 17000
+    },
+    {
+      "epoch": 0.68818292189397,
+      "grad_norm": 1.1716859340667725,
+      "learning_rate": 3.1819120379929794e-05,
+      "loss": 0.9771,
+      "step": 17005
+    },
+    {
+      "epoch": 0.6883852691218131,
+      "grad_norm": 1.1765456199645996,
+      "learning_rate": 3.17984720214743e-05,
+      "loss": 0.9971,
+      "step": 17010
+    },
+    {
+      "epoch": 0.688587616349656,
+      "grad_norm": 1.078888177871704,
+      "learning_rate": 3.177782366301879e-05,
+      "loss": 0.9953,
+      "step": 17015
+    },
+    {
+      "epoch": 0.688789963577499,
+      "grad_norm": 1.1807310581207275,
+      "learning_rate": 3.175717530456329e-05,
+      "loss": 1.0347,
+      "step": 17020
+    },
+    {
+      "epoch": 0.688992310805342,
+      "grad_norm": 1.1671158075332642,
+      "learning_rate": 3.1736526946107784e-05,
+      "loss": 0.9966,
+      "step": 17025
+    },
+    {
+      "epoch": 0.6891946580331849,
+      "grad_norm": 1.1963889598846436,
+      "learning_rate": 3.171587858765228e-05,
+      "loss": 0.9959,
+      "step": 17030
+    },
+    {
+      "epoch": 0.6893970052610279,
+      "grad_norm": 1.1303328275680542,
+      "learning_rate": 3.1695230229196785e-05,
+      "loss": 0.9829,
+      "step": 17035
+    },
+    {
+      "epoch": 0.6895993524888709,
+      "grad_norm": 1.2900586128234863,
+      "learning_rate": 3.1674581870741275e-05,
+      "loss": 0.9906,
+      "step": 17040
+    },
+    {
+      "epoch": 0.6898016997167139,
+      "grad_norm": 1.205917477607727,
+      "learning_rate": 3.165393351228578e-05,
+      "loss": 0.9605,
+      "step": 17045
+    },
+    {
+      "epoch": 0.6900040469445569,
+      "grad_norm": 1.1977159976959229,
+      "learning_rate": 3.163328515383027e-05,
+      "loss": 1.015,
+      "step": 17050
+    },
+    {
+      "epoch": 0.6902063941723998,
+      "grad_norm": 1.1795580387115479,
+      "learning_rate": 3.161263679537477e-05,
+      "loss": 0.9695,
+      "step": 17055
+    },
+    {
+      "epoch": 0.6904087414002428,
+      "grad_norm": 1.1956350803375244,
+      "learning_rate": 3.1591988436919265e-05,
+      "loss": 0.9329,
+      "step": 17060
+    },
+    {
+      "epoch": 0.6906110886280858,
+      "grad_norm": 1.2197588682174683,
+      "learning_rate": 3.157134007846376e-05,
+      "loss": 0.9645,
+      "step": 17065
+    },
+    {
+      "epoch": 0.6908134358559288,
+      "grad_norm": 1.2725110054016113,
+      "learning_rate": 3.1550691720008266e-05,
+      "loss": 1.0407,
+      "step": 17070
+    },
+    {
+      "epoch": 0.6910157830837718,
+      "grad_norm": 1.2563390731811523,
+      "learning_rate": 3.1530043361552756e-05,
+      "loss": 0.9709,
+      "step": 17075
+    },
+    {
+      "epoch": 0.6912181303116147,
+      "grad_norm": 1.1819779872894287,
+      "learning_rate": 3.1509395003097254e-05,
+      "loss": 0.9295,
+      "step": 17080
+    },
+    {
+      "epoch": 0.6914204775394577,
+      "grad_norm": 1.2610588073730469,
+      "learning_rate": 3.148874664464175e-05,
+      "loss": 0.974,
+      "step": 17085
+    },
+    {
+      "epoch": 0.6916228247673006,
+      "grad_norm": 1.2875587940216064,
+      "learning_rate": 3.146809828618625e-05,
+      "loss": 0.975,
+      "step": 17090
+    },
+    {
+      "epoch": 0.6918251719951437,
+      "grad_norm": 1.1317520141601562,
+      "learning_rate": 3.1447449927730746e-05,
+      "loss": 1.036,
+      "step": 17095
+    },
+    {
+      "epoch": 0.6920275192229867,
+      "grad_norm": 1.2763997316360474,
+      "learning_rate": 3.142680156927524e-05,
+      "loss": 0.9996,
+      "step": 17100
+    },
+    {
+      "epoch": 0.6922298664508296,
+      "grad_norm": 1.1817160844802856,
+      "learning_rate": 3.140615321081974e-05,
+      "loss": 0.9727,
+      "step": 17105
+    },
+    {
+      "epoch": 0.6924322136786726,
+      "grad_norm": 1.3305127620697021,
+      "learning_rate": 3.138550485236424e-05,
+      "loss": 0.9357,
+      "step": 17110
+    },
+    {
+      "epoch": 0.6926345609065155,
+      "grad_norm": 1.2789640426635742,
+      "learning_rate": 3.1364856493908735e-05,
+      "loss": 1.0553,
+      "step": 17115
+    },
+    {
+      "epoch": 0.6928369081343586,
+      "grad_norm": 1.2852317094802856,
+      "learning_rate": 3.134420813545323e-05,
+      "loss": 0.9732,
+      "step": 17120
+    },
+    {
+      "epoch": 0.6930392553622016,
+      "grad_norm": 1.098250150680542,
+      "learning_rate": 3.132355977699773e-05,
+      "loss": 1.0155,
+      "step": 17125
+    },
+    {
+      "epoch": 0.6932416025900445,
+      "grad_norm": 1.3127838373184204,
+      "learning_rate": 3.130291141854223e-05,
+      "loss": 1.0226,
+      "step": 17130
+    },
+    {
+      "epoch": 0.6934439498178875,
+      "grad_norm": 1.1965336799621582,
+      "learning_rate": 3.1282263060086724e-05,
+      "loss": 1.0328,
+      "step": 17135
+    },
+    {
+      "epoch": 0.6936462970457304,
+      "grad_norm": 1.1536129713058472,
+      "learning_rate": 3.126161470163122e-05,
+      "loss": 0.9851,
+      "step": 17140
+    },
+    {
+      "epoch": 0.6938486442735734,
+      "grad_norm": 1.166200041770935,
+      "learning_rate": 3.124096634317572e-05,
+      "loss": 1.0096,
+      "step": 17145
+    },
+    {
+      "epoch": 0.6940509915014165,
+      "grad_norm": 1.1907601356506348,
+      "learning_rate": 3.1220317984720216e-05,
+      "loss": 0.9566,
+      "step": 17150
+    },
+    {
+      "epoch": 0.6942533387292594,
+      "grad_norm": 1.1805274486541748,
+      "learning_rate": 3.119966962626471e-05,
+      "loss": 1.004,
+      "step": 17155
+    },
+    {
+      "epoch": 0.6944556859571024,
+      "grad_norm": 1.168805718421936,
+      "learning_rate": 3.117902126780921e-05,
+      "loss": 0.9628,
+      "step": 17160
+    },
+    {
+      "epoch": 0.6946580331849453,
+      "grad_norm": 1.1943094730377197,
+      "learning_rate": 3.115837290935371e-05,
+      "loss": 0.9498,
+      "step": 17165
+    },
+    {
+      "epoch": 0.6948603804127883,
+      "grad_norm": 1.1253079175949097,
+      "learning_rate": 3.1137724550898205e-05,
+      "loss": 1.0004,
+      "step": 17170
+    },
+    {
+      "epoch": 0.6950627276406314,
+      "grad_norm": 1.2831798791885376,
+      "learning_rate": 3.11170761924427e-05,
+      "loss": 1.0009,
+      "step": 17175
+    },
+    {
+      "epoch": 0.6952650748684743,
+      "grad_norm": 1.1908453702926636,
+      "learning_rate": 3.10964278339872e-05,
+      "loss": 1.0469,
+      "step": 17180
+    },
+    {
+      "epoch": 0.6954674220963173,
+      "grad_norm": 1.4067375659942627,
+      "learning_rate": 3.10757794755317e-05,
+      "loss": 0.9516,
+      "step": 17185
+    },
+    {
+      "epoch": 0.6956697693241602,
+      "grad_norm": 1.1740498542785645,
+      "learning_rate": 3.1055131117076194e-05,
+      "loss": 0.986,
+      "step": 17190
+    },
+    {
+      "epoch": 0.6958721165520032,
+      "grad_norm": 1.1124236583709717,
+      "learning_rate": 3.103448275862069e-05,
+      "loss": 0.981,
+      "step": 17195
+    },
+    {
+      "epoch": 0.6960744637798462,
+      "grad_norm": 1.2706719636917114,
+      "learning_rate": 3.101383440016519e-05,
+      "loss": 0.9836,
+      "step": 17200
+    },
+    {
+      "epoch": 0.6962768110076892,
+      "grad_norm": 1.1875782012939453,
+      "learning_rate": 3.0993186041709686e-05,
+      "loss": 0.9679,
+      "step": 17205
+    },
+    {
+      "epoch": 0.6964791582355322,
+      "grad_norm": 1.1833171844482422,
+      "learning_rate": 3.0972537683254184e-05,
+      "loss": 0.9984,
+      "step": 17210
+    },
+    {
+      "epoch": 0.6966815054633752,
+      "grad_norm": 1.259817123413086,
+      "learning_rate": 3.095188932479868e-05,
+      "loss": 1.0075,
+      "step": 17215
+    },
+    {
+      "epoch": 0.6968838526912181,
+      "grad_norm": 1.2820730209350586,
+      "learning_rate": 3.093124096634318e-05,
+      "loss": 0.9491,
+      "step": 17220
+    },
+    {
+      "epoch": 0.6970861999190611,
+      "grad_norm": 1.149603247642517,
+      "learning_rate": 3.0910592607887675e-05,
+      "loss": 0.9498,
+      "step": 17225
+    },
+    {
+      "epoch": 0.6972885471469041,
+      "grad_norm": 1.2305572032928467,
+      "learning_rate": 3.088994424943217e-05,
+      "loss": 1.0212,
+      "step": 17230
+    },
+    {
+      "epoch": 0.6974908943747471,
+      "grad_norm": 1.204371690750122,
+      "learning_rate": 3.086929589097667e-05,
+      "loss": 0.9334,
+      "step": 17235
+    },
+    {
+      "epoch": 0.69769324160259,
+      "grad_norm": 1.2756214141845703,
+      "learning_rate": 3.084864753252117e-05,
+      "loss": 0.9952,
+      "step": 17240
+    },
+    {
+      "epoch": 0.697895588830433,
+      "grad_norm": 1.318031668663025,
+      "learning_rate": 3.0827999174065665e-05,
+      "loss": 1.0067,
+      "step": 17245
+    },
+    {
+      "epoch": 0.698097936058276,
+      "grad_norm": 1.2004657983779907,
+      "learning_rate": 3.080735081561016e-05,
+      "loss": 0.9927,
+      "step": 17250
+    },
+    {
+      "epoch": 0.6983002832861189,
+      "grad_norm": 1.1927980184555054,
+      "learning_rate": 3.078670245715466e-05,
+      "loss": 1.0391,
+      "step": 17255
+    },
+    {
+      "epoch": 0.698502630513962,
+      "grad_norm": 1.2973883152008057,
+      "learning_rate": 3.0766054098699156e-05,
+      "loss": 0.9631,
+      "step": 17260
+    },
+    {
+      "epoch": 0.698704977741805,
+      "grad_norm": 1.366729497909546,
+      "learning_rate": 3.0745405740243654e-05,
+      "loss": 0.9519,
+      "step": 17265
+    },
+    {
+      "epoch": 0.6989073249696479,
+      "grad_norm": 1.086188793182373,
+      "learning_rate": 3.072475738178815e-05,
+      "loss": 0.9726,
+      "step": 17270
+    },
+    {
+      "epoch": 0.6991096721974909,
+      "grad_norm": 1.1788781881332397,
+      "learning_rate": 3.070410902333264e-05,
+      "loss": 0.9806,
+      "step": 17275
+    },
+    {
+      "epoch": 0.6993120194253338,
+      "grad_norm": 1.1486843824386597,
+      "learning_rate": 3.0683460664877146e-05,
+      "loss": 0.9807,
+      "step": 17280
+    },
+    {
+      "epoch": 0.6995143666531769,
+      "grad_norm": 1.2319406270980835,
+      "learning_rate": 3.066281230642164e-05,
+      "loss": 0.9255,
+      "step": 17285
+    },
+    {
+      "epoch": 0.6997167138810199,
+      "grad_norm": 1.2333320379257202,
+      "learning_rate": 3.064216394796614e-05,
+      "loss": 0.933,
+      "step": 17290
+    },
+    {
+      "epoch": 0.6999190611088628,
+      "grad_norm": 1.2039984464645386,
+      "learning_rate": 3.062151558951064e-05,
+      "loss": 0.9584,
+      "step": 17295
+    },
+    {
+      "epoch": 0.7001214083367058,
+      "grad_norm": 1.171964406967163,
+      "learning_rate": 3.060086723105513e-05,
+      "loss": 1.0097,
+      "step": 17300
+    },
+    {
+      "epoch": 0.7003237555645487,
+      "grad_norm": 1.22491455078125,
+      "learning_rate": 3.058021887259963e-05,
+      "loss": 0.9734,
+      "step": 17305
+    },
+    {
+      "epoch": 0.7005261027923917,
+      "grad_norm": 1.3244630098342896,
+      "learning_rate": 3.055957051414412e-05,
+      "loss": 1.0062,
+      "step": 17310
+    },
+    {
+      "epoch": 0.7007284500202348,
+      "grad_norm": 1.1864243745803833,
+      "learning_rate": 3.053892215568863e-05,
+      "loss": 0.9506,
+      "step": 17315
+    },
+    {
+      "epoch": 0.7009307972480777,
+      "grad_norm": 1.193135380744934,
+      "learning_rate": 3.0518273797233124e-05,
+      "loss": 0.9727,
+      "step": 17320
+    },
+    {
+      "epoch": 0.7011331444759207,
+      "grad_norm": 1.3358057737350464,
+      "learning_rate": 3.0497625438777615e-05,
+      "loss": 0.9744,
+      "step": 17325
+    },
+    {
+      "epoch": 0.7013354917037636,
+      "grad_norm": 1.1899951696395874,
+      "learning_rate": 3.0476977080322115e-05,
+      "loss": 0.9446,
+      "step": 17330
+    },
+    {
+      "epoch": 0.7015378389316066,
+      "grad_norm": 1.2701103687286377,
+      "learning_rate": 3.0456328721866613e-05,
+      "loss": 1.033,
+      "step": 17335
+    },
+    {
+      "epoch": 0.7017401861594497,
+      "grad_norm": 1.3066133260726929,
+      "learning_rate": 3.0435680363411113e-05,
+      "loss": 1.0006,
+      "step": 17340
+    },
+    {
+      "epoch": 0.7019425333872926,
+      "grad_norm": 1.1671679019927979,
+      "learning_rate": 3.0415032004955607e-05,
+      "loss": 0.9184,
+      "step": 17345
+    },
+    {
+      "epoch": 0.7021448806151356,
+      "grad_norm": 1.250704050064087,
+      "learning_rate": 3.03943836465001e-05,
+      "loss": 1.0011,
+      "step": 17350
+    },
+    {
+      "epoch": 0.7023472278429785,
+      "grad_norm": 1.2362216711044312,
+      "learning_rate": 3.0373735288044602e-05,
+      "loss": 1.0189,
+      "step": 17355
+    },
+    {
+      "epoch": 0.7025495750708215,
+      "grad_norm": 1.1965128183364868,
+      "learning_rate": 3.03530869295891e-05,
+      "loss": 1.0365,
+      "step": 17360
+    },
+    {
+      "epoch": 0.7027519222986645,
+      "grad_norm": 1.0046998262405396,
+      "learning_rate": 3.0332438571133596e-05,
+      "loss": 1.0028,
+      "step": 17365
+    },
+    {
+      "epoch": 0.7029542695265075,
+      "grad_norm": 1.3213443756103516,
+      "learning_rate": 3.0311790212678094e-05,
+      "loss": 0.9042,
+      "step": 17370
+    },
+    {
+      "epoch": 0.7031566167543505,
+      "grad_norm": 1.106751799583435,
+      "learning_rate": 3.0291141854222594e-05,
+      "loss": 1.0172,
+      "step": 17375
+    },
+    {
+      "epoch": 0.7033589639821934,
+      "grad_norm": 1.1988613605499268,
+      "learning_rate": 3.0270493495767088e-05,
+      "loss": 0.9577,
+      "step": 17380
+    },
+    {
+      "epoch": 0.7035613112100364,
+      "grad_norm": 1.223128318786621,
+      "learning_rate": 3.0249845137311582e-05,
+      "loss": 1.0076,
+      "step": 17385
+    },
+    {
+      "epoch": 0.7037636584378794,
+      "grad_norm": 1.166667103767395,
+      "learning_rate": 3.0229196778856083e-05,
+      "loss": 1.0142,
+      "step": 17390
+    },
+    {
+      "epoch": 0.7039660056657224,
+      "grad_norm": 1.2497634887695312,
+      "learning_rate": 3.020854842040058e-05,
+      "loss": 1.0035,
+      "step": 17395
+    },
+    {
+      "epoch": 0.7041683528935654,
+      "grad_norm": 1.2435925006866455,
+      "learning_rate": 3.018790006194508e-05,
+      "loss": 0.9792,
+      "step": 17400
+    },
+    {
+      "epoch": 0.7043707001214083,
+      "grad_norm": 1.2664761543273926,
+      "learning_rate": 3.0167251703489575e-05,
+      "loss": 1.0018,
+      "step": 17405
+    },
+    {
+      "epoch": 0.7045730473492513,
+      "grad_norm": 1.1748237609863281,
+      "learning_rate": 3.014660334503407e-05,
+      "loss": 1.015,
+      "step": 17410
+    },
+    {
+      "epoch": 0.7047753945770943,
+      "grad_norm": 1.3572683334350586,
+      "learning_rate": 3.012595498657857e-05,
+      "loss": 0.9939,
+      "step": 17415
+    },
+    {
+      "epoch": 0.7049777418049372,
+      "grad_norm": 1.2530001401901245,
+      "learning_rate": 3.0105306628123063e-05,
+      "loss": 1.0323,
+      "step": 17420
+    },
+    {
+      "epoch": 0.7051800890327803,
+      "grad_norm": 1.1220327615737915,
+      "learning_rate": 3.0084658269667564e-05,
+      "loss": 1.0245,
+      "step": 17425
+    },
+    {
+      "epoch": 0.7053824362606232,
+      "grad_norm": 1.2212517261505127,
+      "learning_rate": 3.006400991121206e-05,
+      "loss": 1.0247,
+      "step": 17430
+    },
+    {
+      "epoch": 0.7055847834884662,
+      "grad_norm": 1.3376418352127075,
+      "learning_rate": 3.0043361552756555e-05,
+      "loss": 0.9992,
+      "step": 17435
+    },
+    {
+      "epoch": 0.7057871307163092,
+      "grad_norm": 1.1860454082489014,
+      "learning_rate": 3.0022713194301056e-05,
+      "loss": 1.0302,
+      "step": 17440
+    },
+    {
+      "epoch": 0.7059894779441521,
+      "grad_norm": 1.2094342708587646,
+      "learning_rate": 3.000206483584555e-05,
+      "loss": 1.0539,
+      "step": 17445
+    },
+    {
+      "epoch": 0.7061918251719952,
+      "grad_norm": 1.252813458442688,
+      "learning_rate": 2.998141647739005e-05,
+      "loss": 0.9874,
+      "step": 17450
+    },
+    {
+      "epoch": 0.7063941723998381,
+      "grad_norm": 1.2359617948532104,
+      "learning_rate": 2.9960768118934544e-05,
+      "loss": 0.9854,
+      "step": 17455
+    },
+    {
+      "epoch": 0.7065965196276811,
+      "grad_norm": 1.0895074605941772,
+      "learning_rate": 2.994011976047904e-05,
+      "loss": 0.9447,
+      "step": 17460
+    },
+    {
+      "epoch": 0.7067988668555241,
+      "grad_norm": 1.1701818704605103,
+      "learning_rate": 2.9919471402023542e-05,
+      "loss": 0.9801,
+      "step": 17465
+    },
+    {
+      "epoch": 0.707001214083367,
+      "grad_norm": 1.2293882369995117,
+      "learning_rate": 2.9898823043568036e-05,
+      "loss": 0.955,
+      "step": 17470
+    },
+    {
+      "epoch": 0.70720356131121,
+      "grad_norm": 1.1827373504638672,
+      "learning_rate": 2.9878174685112537e-05,
+      "loss": 1.0189,
+      "step": 17475
+    },
+    {
+      "epoch": 0.707405908539053,
+      "grad_norm": 1.2127366065979004,
+      "learning_rate": 2.985752632665703e-05,
+      "loss": 0.9989,
+      "step": 17480
+    },
+    {
+      "epoch": 0.707608255766896,
+      "grad_norm": 1.1264581680297852,
+      "learning_rate": 2.983687796820153e-05,
+      "loss": 0.9531,
+      "step": 17485
+    },
+    {
+      "epoch": 0.707810602994739,
+      "grad_norm": 1.2644881010055542,
+      "learning_rate": 2.9816229609746025e-05,
+      "loss": 1.0012,
+      "step": 17490
+    },
+    {
+      "epoch": 0.7080129502225819,
+      "grad_norm": 1.210516333580017,
+      "learning_rate": 2.9795581251290523e-05,
+      "loss": 0.9253,
+      "step": 17495
+    },
+    {
+      "epoch": 0.7082152974504249,
+      "grad_norm": 1.147161602973938,
+      "learning_rate": 2.9774932892835023e-05,
+      "loss": 1.0003,
+      "step": 17500
+    },
+    {
+      "epoch": 0.708417644678268,
+      "grad_norm": 1.2431273460388184,
+      "learning_rate": 2.9754284534379517e-05,
+      "loss": 0.9843,
+      "step": 17505
+    },
+    {
+      "epoch": 0.7086199919061109,
+      "grad_norm": 1.1441364288330078,
+      "learning_rate": 2.9733636175924018e-05,
+      "loss": 0.974,
+      "step": 17510
+    },
+    {
+      "epoch": 0.7088223391339539,
+      "grad_norm": 1.2725236415863037,
+      "learning_rate": 2.9712987817468512e-05,
+      "loss": 1.0117,
+      "step": 17515
+    },
+    {
+      "epoch": 0.7090246863617968,
+      "grad_norm": 1.130359172821045,
+      "learning_rate": 2.969233945901301e-05,
+      "loss": 0.9799,
+      "step": 17520
+    },
+    {
+      "epoch": 0.7092270335896398,
+      "grad_norm": 1.3188213109970093,
+      "learning_rate": 2.967169110055751e-05,
+      "loss": 0.9573,
+      "step": 17525
+    },
+    {
+      "epoch": 0.7094293808174827,
+      "grad_norm": 1.219534158706665,
+      "learning_rate": 2.9651042742102004e-05,
+      "loss": 0.987,
+      "step": 17530
+    },
+    {
+      "epoch": 0.7096317280453258,
+      "grad_norm": 1.1407498121261597,
+      "learning_rate": 2.9630394383646504e-05,
+      "loss": 0.9485,
+      "step": 17535
+    },
+    {
+      "epoch": 0.7098340752731688,
+      "grad_norm": 1.3663009405136108,
+      "learning_rate": 2.9609746025191e-05,
+      "loss": 0.9618,
+      "step": 17540
+    },
+    {
+      "epoch": 0.7100364225010117,
+      "grad_norm": 1.2362415790557861,
+      "learning_rate": 2.9589097666735492e-05,
+      "loss": 1.0356,
+      "step": 17545
+    },
+    {
+      "epoch": 0.7102387697288547,
+      "grad_norm": 1.11896812915802,
+      "learning_rate": 2.9568449308279993e-05,
+      "loss": 1.0108,
+      "step": 17550
+    },
+    {
+      "epoch": 0.7104411169566976,
+      "grad_norm": 1.3680909872055054,
+      "learning_rate": 2.954780094982449e-05,
+      "loss": 0.9726,
+      "step": 17555
+    },
+    {
+      "epoch": 0.7106434641845407,
+      "grad_norm": 1.1663466691970825,
+      "learning_rate": 2.952715259136899e-05,
+      "loss": 0.9811,
+      "step": 17560
+    },
+    {
+      "epoch": 0.7108458114123837,
+      "grad_norm": 1.2008883953094482,
+      "learning_rate": 2.9506504232913485e-05,
+      "loss": 0.9161,
+      "step": 17565
+    },
+    {
+      "epoch": 0.7110481586402266,
+      "grad_norm": 1.5105302333831787,
+      "learning_rate": 2.948585587445798e-05,
+      "loss": 1.0413,
+      "step": 17570
+    },
+    {
+      "epoch": 0.7112505058680696,
+      "grad_norm": 1.129315733909607,
+      "learning_rate": 2.946520751600248e-05,
+      "loss": 0.9704,
+      "step": 17575
+    },
+    {
+      "epoch": 0.7114528530959126,
+      "grad_norm": 1.1245055198669434,
+      "learning_rate": 2.9444559157546973e-05,
+      "loss": 1.0059,
+      "step": 17580
+    },
+    {
+      "epoch": 0.7116552003237555,
+      "grad_norm": 1.1994887590408325,
+      "learning_rate": 2.9423910799091474e-05,
+      "loss": 1.0028,
+      "step": 17585
+    },
+    {
+      "epoch": 0.7118575475515986,
+      "grad_norm": 1.2163746356964111,
+      "learning_rate": 2.940326244063597e-05,
+      "loss": 0.9749,
+      "step": 17590
+    },
+    {
+      "epoch": 0.7120598947794415,
+      "grad_norm": 1.201568841934204,
+      "learning_rate": 2.9382614082180472e-05,
+      "loss": 1.0046,
+      "step": 17595
+    },
+    {
+      "epoch": 0.7122622420072845,
+      "grad_norm": 1.2677850723266602,
+      "learning_rate": 2.9361965723724966e-05,
+      "loss": 1.0374,
+      "step": 17600
+    },
+    {
+      "epoch": 0.7124645892351275,
+      "grad_norm": 1.3034635782241821,
+      "learning_rate": 2.934131736526946e-05,
+      "loss": 0.9419,
+      "step": 17605
+    },
+    {
+      "epoch": 0.7126669364629704,
+      "grad_norm": 1.3210262060165405,
+      "learning_rate": 2.932066900681396e-05,
+      "loss": 0.9751,
+      "step": 17610
+    },
+    {
+      "epoch": 0.7128692836908135,
+      "grad_norm": 1.225956678390503,
+      "learning_rate": 2.9300020648358454e-05,
+      "loss": 1.042,
+      "step": 17615
+    },
+    {
+      "epoch": 0.7130716309186564,
+      "grad_norm": 1.2935391664505005,
+      "learning_rate": 2.9279372289902955e-05,
+      "loss": 0.9783,
+      "step": 17620
+    },
+    {
+      "epoch": 0.7132739781464994,
+      "grad_norm": 1.1585502624511719,
+      "learning_rate": 2.9258723931447452e-05,
+      "loss": 0.9657,
+      "step": 17625
+    },
+    {
+      "epoch": 0.7134763253743424,
+      "grad_norm": 1.2231570482254028,
+      "learning_rate": 2.9238075572991946e-05,
+      "loss": 0.9845,
+      "step": 17630
+    },
+    {
+      "epoch": 0.7136786726021853,
+      "grad_norm": 1.0635497570037842,
+      "learning_rate": 2.9217427214536447e-05,
+      "loss": 0.9963,
+      "step": 17635
+    },
+    {
+      "epoch": 0.7138810198300283,
+      "grad_norm": 1.2352100610733032,
+      "learning_rate": 2.919677885608094e-05,
+      "loss": 1.0434,
+      "step": 17640
+    },
+    {
+      "epoch": 0.7140833670578713,
+      "grad_norm": 1.2931196689605713,
+      "learning_rate": 2.917613049762544e-05,
+      "loss": 0.9682,
+      "step": 17645
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 1.2205458879470825,
+      "learning_rate": 2.915548213916994e-05,
+      "loss": 0.9681,
+      "step": 17650
+    },
+    {
+      "epoch": 0.7144880615135573,
+      "grad_norm": 1.1368838548660278,
+      "learning_rate": 2.9134833780714433e-05,
+      "loss": 1.0158,
+      "step": 17655
+    },
+    {
+      "epoch": 0.7146904087414002,
+      "grad_norm": 1.108027458190918,
+      "learning_rate": 2.9114185422258933e-05,
+      "loss": 1.0608,
+      "step": 17660
+    },
+    {
+      "epoch": 0.7148927559692432,
+      "grad_norm": 1.2166893482208252,
+      "learning_rate": 2.9093537063803427e-05,
+      "loss": 1.0056,
+      "step": 17665
+    },
+    {
+      "epoch": 0.7150951031970862,
+      "grad_norm": 1.1938624382019043,
+      "learning_rate": 2.9072888705347928e-05,
+      "loss": 0.9431,
+      "step": 17670
+    },
+    {
+      "epoch": 0.7152974504249292,
+      "grad_norm": 1.1462186574935913,
+      "learning_rate": 2.9052240346892422e-05,
+      "loss": 0.9415,
+      "step": 17675
+    },
+    {
+      "epoch": 0.7154997976527722,
+      "grad_norm": 1.102442979812622,
+      "learning_rate": 2.903159198843692e-05,
+      "loss": 0.9445,
+      "step": 17680
+    },
+    {
+      "epoch": 0.7157021448806151,
+      "grad_norm": 1.1912766695022583,
+      "learning_rate": 2.901094362998142e-05,
+      "loss": 0.9873,
+      "step": 17685
+    },
+    {
+      "epoch": 0.7159044921084581,
+      "grad_norm": 1.3539094924926758,
+      "learning_rate": 2.8990295271525914e-05,
+      "loss": 1.0329,
+      "step": 17690
+    },
+    {
+      "epoch": 0.716106839336301,
+      "grad_norm": 1.161729097366333,
+      "learning_rate": 2.8969646913070415e-05,
+      "loss": 0.9554,
+      "step": 17695
+    },
+    {
+      "epoch": 0.7163091865641441,
+      "grad_norm": 1.1892973184585571,
+      "learning_rate": 2.894899855461491e-05,
+      "loss": 0.9904,
+      "step": 17700
+    },
+    {
+      "epoch": 0.7165115337919871,
+      "grad_norm": 1.1171448230743408,
+      "learning_rate": 2.892835019615941e-05,
+      "loss": 0.9547,
+      "step": 17705
+    },
+    {
+      "epoch": 0.71671388101983,
+      "grad_norm": 1.1299140453338623,
+      "learning_rate": 2.8907701837703903e-05,
+      "loss": 1.0529,
+      "step": 17710
+    },
+    {
+      "epoch": 0.716916228247673,
+      "grad_norm": 1.1318342685699463,
+      "learning_rate": 2.88870534792484e-05,
+      "loss": 1.0226,
+      "step": 17715
+    },
+    {
+      "epoch": 0.7171185754755159,
+      "grad_norm": 1.190993070602417,
+      "learning_rate": 2.88664051207929e-05,
+      "loss": 0.9811,
+      "step": 17720
+    },
+    {
+      "epoch": 0.717320922703359,
+      "grad_norm": 1.2098536491394043,
+      "learning_rate": 2.8845756762337395e-05,
+      "loss": 0.9948,
+      "step": 17725
+    },
+    {
+      "epoch": 0.717523269931202,
+      "grad_norm": 1.2147682905197144,
+      "learning_rate": 2.8825108403881896e-05,
+      "loss": 1.0396,
+      "step": 17730
+    },
+    {
+      "epoch": 0.7177256171590449,
+      "grad_norm": 1.2443863153457642,
+      "learning_rate": 2.880446004542639e-05,
+      "loss": 0.9413,
+      "step": 17735
+    },
+    {
+      "epoch": 0.7179279643868879,
+      "grad_norm": 1.1268609762191772,
+      "learning_rate": 2.8783811686970887e-05,
+      "loss": 0.9783,
+      "step": 17740
+    },
+    {
+      "epoch": 0.7181303116147308,
+      "grad_norm": 1.2341854572296143,
+      "learning_rate": 2.8763163328515384e-05,
+      "loss": 0.9627,
+      "step": 17745
+    },
+    {
+      "epoch": 0.7183326588425738,
+      "grad_norm": 1.1590079069137573,
+      "learning_rate": 2.874251497005988e-05,
+      "loss": 1.0087,
+      "step": 17750
+    },
+    {
+      "epoch": 0.7185350060704169,
+      "grad_norm": 1.437095046043396,
+      "learning_rate": 2.8721866611604382e-05,
+      "loss": 0.9936,
+      "step": 17755
+    },
+    {
+      "epoch": 0.7187373532982598,
+      "grad_norm": 1.2555290460586548,
+      "learning_rate": 2.8701218253148876e-05,
+      "loss": 0.9719,
+      "step": 17760
+    },
+    {
+      "epoch": 0.7189397005261028,
+      "grad_norm": 1.1498318910598755,
+      "learning_rate": 2.868056989469337e-05,
+      "loss": 1.0202,
+      "step": 17765
+    },
+    {
+      "epoch": 0.7191420477539457,
+      "grad_norm": 1.3355313539505005,
+      "learning_rate": 2.865992153623787e-05,
+      "loss": 1.0173,
+      "step": 17770
+    },
+    {
+      "epoch": 0.7193443949817887,
+      "grad_norm": 1.366964340209961,
+      "learning_rate": 2.8639273177782368e-05,
+      "loss": 1.0336,
+      "step": 17775
+    },
+    {
+      "epoch": 0.7195467422096318,
+      "grad_norm": 1.0923272371292114,
+      "learning_rate": 2.861862481932687e-05,
+      "loss": 1.0227,
+      "step": 17780
+    },
+    {
+      "epoch": 0.7197490894374747,
+      "grad_norm": 1.246429443359375,
+      "learning_rate": 2.8597976460871362e-05,
+      "loss": 1.0321,
+      "step": 17785
+    },
+    {
+      "epoch": 0.7199514366653177,
+      "grad_norm": 1.1993050575256348,
+      "learning_rate": 2.8577328102415856e-05,
+      "loss": 0.975,
+      "step": 17790
+    },
+    {
+      "epoch": 0.7201537838931606,
+      "grad_norm": 1.178747534751892,
+      "learning_rate": 2.8556679743960357e-05,
+      "loss": 1.0033,
+      "step": 17795
+    },
+    {
+      "epoch": 0.7203561311210036,
+      "grad_norm": 1.3549706935882568,
+      "learning_rate": 2.853603138550485e-05,
+      "loss": 1.0397,
+      "step": 17800
+    },
+    {
+      "epoch": 0.7205584783488467,
+      "grad_norm": 1.1827131509780884,
+      "learning_rate": 2.851538302704935e-05,
+      "loss": 1.0578,
+      "step": 17805
+    },
+    {
+      "epoch": 0.7207608255766896,
+      "grad_norm": 1.2043248414993286,
+      "learning_rate": 2.849473466859385e-05,
+      "loss": 1.002,
+      "step": 17810
+    },
+    {
+      "epoch": 0.7209631728045326,
+      "grad_norm": 1.2782275676727295,
+      "learning_rate": 2.8474086310138343e-05,
+      "loss": 0.9673,
+      "step": 17815
+    },
+    {
+      "epoch": 0.7211655200323756,
+      "grad_norm": 1.095707893371582,
+      "learning_rate": 2.8453437951682844e-05,
+      "loss": 0.9683,
+      "step": 17820
+    },
+    {
+      "epoch": 0.7213678672602185,
+      "grad_norm": 1.1790738105773926,
+      "learning_rate": 2.8432789593227337e-05,
+      "loss": 0.9955,
+      "step": 17825
+    },
+    {
+      "epoch": 0.7215702144880615,
+      "grad_norm": 1.1360628604888916,
+      "learning_rate": 2.8412141234771838e-05,
+      "loss": 1.045,
+      "step": 17830
+    },
+    {
+      "epoch": 0.7217725617159045,
+      "grad_norm": 1.2507461309432983,
+      "learning_rate": 2.8391492876316332e-05,
+      "loss": 1.0048,
+      "step": 17835
+    },
+    {
+      "epoch": 0.7219749089437475,
+      "grad_norm": 1.1951102018356323,
+      "learning_rate": 2.8370844517860833e-05,
+      "loss": 1.059,
+      "step": 17840
+    },
+    {
+      "epoch": 0.7221772561715905,
+      "grad_norm": 1.2759116888046265,
+      "learning_rate": 2.835019615940533e-05,
+      "loss": 1.0177,
+      "step": 17845
+    },
+    {
+      "epoch": 0.7223796033994334,
+      "grad_norm": 1.1330106258392334,
+      "learning_rate": 2.8329547800949824e-05,
+      "loss": 1.0073,
+      "step": 17850
+    },
+    {
+      "epoch": 0.7225819506272764,
+      "grad_norm": 1.2677319049835205,
+      "learning_rate": 2.8308899442494325e-05,
+      "loss": 0.9473,
+      "step": 17855
+    },
+    {
+      "epoch": 0.7227842978551194,
+      "grad_norm": 1.075184941291809,
+      "learning_rate": 2.828825108403882e-05,
+      "loss": 1.0303,
+      "step": 17860
+    },
+    {
+      "epoch": 0.7229866450829624,
+      "grad_norm": 1.1539348363876343,
+      "learning_rate": 2.826760272558332e-05,
+      "loss": 0.9587,
+      "step": 17865
+    },
+    {
+      "epoch": 0.7231889923108054,
+      "grad_norm": 1.0828721523284912,
+      "learning_rate": 2.8246954367127813e-05,
+      "loss": 1.0265,
+      "step": 17870
+    },
+    {
+      "epoch": 0.7233913395386483,
+      "grad_norm": 1.3209598064422607,
+      "learning_rate": 2.822630600867231e-05,
+      "loss": 1.0518,
+      "step": 17875
+    },
+    {
+      "epoch": 0.7235936867664913,
+      "grad_norm": 1.2192697525024414,
+      "learning_rate": 2.820565765021681e-05,
+      "loss": 0.9714,
+      "step": 17880
+    },
+    {
+      "epoch": 0.7237960339943342,
+      "grad_norm": 1.2710083723068237,
+      "learning_rate": 2.8185009291761305e-05,
+      "loss": 1.0182,
+      "step": 17885
+    },
+    {
+      "epoch": 0.7239983812221773,
+      "grad_norm": 1.3627142906188965,
+      "learning_rate": 2.8164360933305806e-05,
+      "loss": 1.0093,
+      "step": 17890
+    },
+    {
+      "epoch": 0.7242007284500203,
+      "grad_norm": 1.2135127782821655,
+      "learning_rate": 2.81437125748503e-05,
+      "loss": 0.9915,
+      "step": 17895
+    },
+    {
+      "epoch": 0.7244030756778632,
+      "grad_norm": 1.2205002307891846,
+      "learning_rate": 2.8123064216394797e-05,
+      "loss": 0.9681,
+      "step": 17900
+    },
+    {
+      "epoch": 0.7246054229057062,
+      "grad_norm": 1.2374382019042969,
+      "learning_rate": 2.8102415857939298e-05,
+      "loss": 1.0354,
+      "step": 17905
+    },
+    {
+      "epoch": 0.7248077701335491,
+      "grad_norm": 1.304052710533142,
+      "learning_rate": 2.808176749948379e-05,
+      "loss": 1.0044,
+      "step": 17910
+    },
+    {
+      "epoch": 0.7250101173613922,
+      "grad_norm": 1.1533074378967285,
+      "learning_rate": 2.8061119141028292e-05,
+      "loss": 1.045,
+      "step": 17915
+    },
+    {
+      "epoch": 0.7252124645892352,
+      "grad_norm": 1.1206939220428467,
+      "learning_rate": 2.8040470782572786e-05,
+      "loss": 0.9843,
+      "step": 17920
+    },
+    {
+      "epoch": 0.7254148118170781,
+      "grad_norm": 1.1591408252716064,
+      "learning_rate": 2.801982242411728e-05,
+      "loss": 0.9715,
+      "step": 17925
+    },
+    {
+      "epoch": 0.7256171590449211,
+      "grad_norm": 1.144925832748413,
+      "learning_rate": 2.799917406566178e-05,
+      "loss": 0.9587,
+      "step": 17930
+    },
+    {
+      "epoch": 0.725819506272764,
+      "grad_norm": 1.1545476913452148,
+      "learning_rate": 2.7978525707206278e-05,
+      "loss": 0.9965,
+      "step": 17935
+    },
+    {
+      "epoch": 0.726021853500607,
+      "grad_norm": 1.253890037536621,
+      "learning_rate": 2.795787734875078e-05,
+      "loss": 0.9684,
+      "step": 17940
+    },
+    {
+      "epoch": 0.7262242007284501,
+      "grad_norm": 1.132378101348877,
+      "learning_rate": 2.7937228990295273e-05,
+      "loss": 0.9984,
+      "step": 17945
+    },
+    {
+      "epoch": 0.726426547956293,
+      "grad_norm": 1.8752743005752563,
+      "learning_rate": 2.7916580631839773e-05,
+      "loss": 0.9577,
+      "step": 17950
+    },
+    {
+      "epoch": 0.726628895184136,
+      "grad_norm": 1.2820810079574585,
+      "learning_rate": 2.7895932273384267e-05,
+      "loss": 0.9291,
+      "step": 17955
+    },
+    {
+      "epoch": 0.7268312424119789,
+      "grad_norm": 1.2255547046661377,
+      "learning_rate": 2.787528391492876e-05,
+      "loss": 1.0168,
+      "step": 17960
+    },
+    {
+      "epoch": 0.7270335896398219,
+      "grad_norm": 1.2439258098602295,
+      "learning_rate": 2.7854635556473262e-05,
+      "loss": 0.952,
+      "step": 17965
+    },
+    {
+      "epoch": 0.727235936867665,
+      "grad_norm": 1.1685419082641602,
+      "learning_rate": 2.783398719801776e-05,
+      "loss": 1.0325,
+      "step": 17970
+    },
+    {
+      "epoch": 0.7274382840955079,
+      "grad_norm": 1.1932899951934814,
+      "learning_rate": 2.781333883956226e-05,
+      "loss": 0.9937,
+      "step": 17975
+    },
+    {
+      "epoch": 0.7276406313233509,
+      "grad_norm": 1.2068145275115967,
+      "learning_rate": 2.7792690481106754e-05,
+      "loss": 0.9389,
+      "step": 17980
+    },
+    {
+      "epoch": 0.7278429785511938,
+      "grad_norm": 1.2051451206207275,
+      "learning_rate": 2.7772042122651248e-05,
+      "loss": 1.0095,
+      "step": 17985
+    },
+    {
+      "epoch": 0.7280453257790368,
+      "grad_norm": 1.140273094177246,
+      "learning_rate": 2.7751393764195748e-05,
+      "loss": 0.9967,
+      "step": 17990
+    },
+    {
+      "epoch": 0.7282476730068798,
+      "grad_norm": 1.3008310794830322,
+      "learning_rate": 2.7730745405740242e-05,
+      "loss": 1.0487,
+      "step": 17995
+    },
+    {
+      "epoch": 0.7284500202347228,
+      "grad_norm": 1.230934977531433,
+      "learning_rate": 2.7710097047284743e-05,
+      "loss": 1.0499,
+      "step": 18000
+    },
+    {
+      "epoch": 0.7286523674625658,
+      "grad_norm": 1.1574170589447021,
+      "learning_rate": 2.768944868882924e-05,
+      "loss": 0.9733,
+      "step": 18005
+    },
+    {
+      "epoch": 0.7288547146904087,
+      "grad_norm": 1.1711808443069458,
+      "learning_rate": 2.7668800330373734e-05,
+      "loss": 1.0224,
+      "step": 18010
+    },
+    {
+      "epoch": 0.7290570619182517,
+      "grad_norm": 1.1061688661575317,
+      "learning_rate": 2.7648151971918235e-05,
+      "loss": 0.9966,
+      "step": 18015
+    },
+    {
+      "epoch": 0.7292594091460947,
+      "grad_norm": 1.1513621807098389,
+      "learning_rate": 2.762750361346273e-05,
+      "loss": 0.9743,
+      "step": 18020
+    },
+    {
+      "epoch": 0.7294617563739377,
+      "grad_norm": 1.2922521829605103,
+      "learning_rate": 2.760685525500723e-05,
+      "loss": 0.9283,
+      "step": 18025
+    },
+    {
+      "epoch": 0.7296641036017807,
+      "grad_norm": 1.3262009620666504,
+      "learning_rate": 2.7586206896551727e-05,
+      "loss": 0.9928,
+      "step": 18030
+    },
+    {
+      "epoch": 0.7298664508296236,
+      "grad_norm": 1.1433358192443848,
+      "learning_rate": 2.756555853809622e-05,
+      "loss": 0.9375,
+      "step": 18035
+    },
+    {
+      "epoch": 0.7300687980574666,
+      "grad_norm": 1.1117414236068726,
+      "learning_rate": 2.754491017964072e-05,
+      "loss": 0.946,
+      "step": 18040
+    },
+    {
+      "epoch": 0.7302711452853096,
+      "grad_norm": 1.0909231901168823,
+      "learning_rate": 2.7524261821185215e-05,
+      "loss": 0.9309,
+      "step": 18045
+    },
+    {
+      "epoch": 0.7304734925131525,
+      "grad_norm": 1.2091577053070068,
+      "learning_rate": 2.7503613462729716e-05,
+      "loss": 0.9294,
+      "step": 18050
+    },
+    {
+      "epoch": 0.7306758397409956,
+      "grad_norm": 1.213517189025879,
+      "learning_rate": 2.748296510427421e-05,
+      "loss": 0.9991,
+      "step": 18055
+    },
+    {
+      "epoch": 0.7308781869688386,
+      "grad_norm": 1.196061134338379,
+      "learning_rate": 2.746231674581871e-05,
+      "loss": 0.9831,
+      "step": 18060
+    },
+    {
+      "epoch": 0.7310805341966815,
+      "grad_norm": 1.3827677965164185,
+      "learning_rate": 2.7441668387363208e-05,
+      "loss": 0.9718,
+      "step": 18065
+    },
+    {
+      "epoch": 0.7312828814245245,
+      "grad_norm": 1.1541972160339355,
+      "learning_rate": 2.74210200289077e-05,
+      "loss": 0.9847,
+      "step": 18070
+    },
+    {
+      "epoch": 0.7314852286523674,
+      "grad_norm": 1.4115660190582275,
+      "learning_rate": 2.7400371670452202e-05,
+      "loss": 1.039,
+      "step": 18075
+    },
+    {
+      "epoch": 0.7316875758802105,
+      "grad_norm": 1.2789621353149414,
+      "learning_rate": 2.7379723311996696e-05,
+      "loss": 0.9885,
+      "step": 18080
+    },
+    {
+      "epoch": 0.7318899231080535,
+      "grad_norm": 1.1122533082962036,
+      "learning_rate": 2.7359074953541197e-05,
+      "loss": 0.9761,
+      "step": 18085
+    },
+    {
+      "epoch": 0.7320922703358964,
+      "grad_norm": 1.1807162761688232,
+      "learning_rate": 2.733842659508569e-05,
+      "loss": 1.0065,
+      "step": 18090
+    },
+    {
+      "epoch": 0.7322946175637394,
+      "grad_norm": 1.2819427251815796,
+      "learning_rate": 2.7317778236630188e-05,
+      "loss": 1.0249,
+      "step": 18095
+    },
+    {
+      "epoch": 0.7324969647915823,
+      "grad_norm": 1.377754807472229,
+      "learning_rate": 2.729712987817469e-05,
+      "loss": 1.001,
+      "step": 18100
+    },
+    {
+      "epoch": 0.7326993120194253,
+      "grad_norm": 1.3264992237091064,
+      "learning_rate": 2.7276481519719183e-05,
+      "loss": 1.0539,
+      "step": 18105
+    },
+    {
+      "epoch": 0.7329016592472684,
+      "grad_norm": 1.3716049194335938,
+      "learning_rate": 2.7255833161263683e-05,
+      "loss": 0.9949,
+      "step": 18110
+    },
+    {
+      "epoch": 0.7331040064751113,
+      "grad_norm": 1.1768916845321655,
+      "learning_rate": 2.7235184802808177e-05,
+      "loss": 1.0183,
+      "step": 18115
+    },
+    {
+      "epoch": 0.7333063537029543,
+      "grad_norm": 1.0700510740280151,
+      "learning_rate": 2.7214536444352675e-05,
+      "loss": 0.9822,
+      "step": 18120
+    },
+    {
+      "epoch": 0.7335087009307972,
+      "grad_norm": 1.0094561576843262,
+      "learning_rate": 2.7193888085897172e-05,
+      "loss": 0.9612,
+      "step": 18125
+    },
+    {
+      "epoch": 0.7337110481586402,
+      "grad_norm": 1.1067116260528564,
+      "learning_rate": 2.717323972744167e-05,
+      "loss": 0.925,
+      "step": 18130
+    },
+    {
+      "epoch": 0.7339133953864833,
+      "grad_norm": 1.5170282125473022,
+      "learning_rate": 2.715259136898617e-05,
+      "loss": 0.9799,
+      "step": 18135
+    },
+    {
+      "epoch": 0.7341157426143262,
+      "grad_norm": 1.2024511098861694,
+      "learning_rate": 2.7131943010530664e-05,
+      "loss": 0.9495,
+      "step": 18140
+    },
+    {
+      "epoch": 0.7343180898421692,
+      "grad_norm": 1.2323782444000244,
+      "learning_rate": 2.7111294652075158e-05,
+      "loss": 0.9456,
+      "step": 18145
+    },
+    {
+      "epoch": 0.7345204370700121,
+      "grad_norm": 1.0937429666519165,
+      "learning_rate": 2.709064629361966e-05,
+      "loss": 1.0044,
+      "step": 18150
+    },
+    {
+      "epoch": 0.7347227842978551,
+      "grad_norm": 1.1940041780471802,
+      "learning_rate": 2.7069997935164156e-05,
+      "loss": 0.9797,
+      "step": 18155
+    },
+    {
+      "epoch": 0.734925131525698,
+      "grad_norm": 1.17705237865448,
+      "learning_rate": 2.7049349576708656e-05,
+      "loss": 0.9811,
+      "step": 18160
+    },
+    {
+      "epoch": 0.7351274787535411,
+      "grad_norm": 1.2329814434051514,
+      "learning_rate": 2.702870121825315e-05,
+      "loss": 0.9981,
+      "step": 18165
+    },
+    {
+      "epoch": 0.7353298259813841,
+      "grad_norm": 1.1110891103744507,
+      "learning_rate": 2.700805285979765e-05,
+      "loss": 0.9597,
+      "step": 18170
+    },
+    {
+      "epoch": 0.735532173209227,
+      "grad_norm": 1.1086004972457886,
+      "learning_rate": 2.6987404501342145e-05,
+      "loss": 1.0896,
+      "step": 18175
+    },
+    {
+      "epoch": 0.73573452043707,
+      "grad_norm": 1.1698123216629028,
+      "learning_rate": 2.696675614288664e-05,
+      "loss": 0.9755,
+      "step": 18180
+    },
+    {
+      "epoch": 0.735936867664913,
+      "grad_norm": 1.23163640499115,
+      "learning_rate": 2.694610778443114e-05,
+      "loss": 0.9895,
+      "step": 18185
+    },
+    {
+      "epoch": 0.736139214892756,
+      "grad_norm": 1.226365327835083,
+      "learning_rate": 2.6925459425975637e-05,
+      "loss": 0.9699,
+      "step": 18190
+    },
+    {
+      "epoch": 0.736341562120599,
+      "grad_norm": 1.3278625011444092,
+      "learning_rate": 2.6904811067520137e-05,
+      "loss": 0.9982,
+      "step": 18195
+    },
+    {
+      "epoch": 0.7365439093484419,
+      "grad_norm": 1.099804401397705,
+      "learning_rate": 2.688416270906463e-05,
+      "loss": 0.9951,
+      "step": 18200
+    },
+    {
+      "epoch": 0.7367462565762849,
+      "grad_norm": 1.2941951751708984,
+      "learning_rate": 2.6863514350609125e-05,
+      "loss": 0.9536,
+      "step": 18205
+    },
+    {
+      "epoch": 0.7369486038041279,
+      "grad_norm": 1.3249821662902832,
+      "learning_rate": 2.6842865992153626e-05,
+      "loss": 0.9945,
+      "step": 18210
+    },
+    {
+      "epoch": 0.7371509510319708,
+      "grad_norm": 1.1776769161224365,
+      "learning_rate": 2.682221763369812e-05,
+      "loss": 1.0315,
+      "step": 18215
+    },
+    {
+      "epoch": 0.7373532982598139,
+      "grad_norm": 1.1420999765396118,
+      "learning_rate": 2.680156927524262e-05,
+      "loss": 1.0363,
+      "step": 18220
+    },
+    {
+      "epoch": 0.7375556454876568,
+      "grad_norm": 1.2246534824371338,
+      "learning_rate": 2.6780920916787118e-05,
+      "loss": 0.9712,
+      "step": 18225
+    },
+    {
+      "epoch": 0.7377579927154998,
+      "grad_norm": 1.2218377590179443,
+      "learning_rate": 2.6760272558331612e-05,
+      "loss": 0.9366,
+      "step": 18230
+    },
+    {
+      "epoch": 0.7379603399433428,
+      "grad_norm": 1.1964023113250732,
+      "learning_rate": 2.6739624199876112e-05,
+      "loss": 1.0244,
+      "step": 18235
+    },
+    {
+      "epoch": 0.7381626871711857,
+      "grad_norm": 1.1200047731399536,
+      "learning_rate": 2.6718975841420606e-05,
+      "loss": 0.9352,
+      "step": 18240
+    },
+    {
+      "epoch": 0.7383650343990288,
+      "grad_norm": 1.3178179264068604,
+      "learning_rate": 2.6698327482965107e-05,
+      "loss": 0.9939,
+      "step": 18245
+    },
+    {
+      "epoch": 0.7385673816268717,
+      "grad_norm": 1.1094086170196533,
+      "learning_rate": 2.66776791245096e-05,
+      "loss": 1.0116,
+      "step": 18250
+    },
+    {
+      "epoch": 0.7387697288547147,
+      "grad_norm": 1.3109591007232666,
+      "learning_rate": 2.6657030766054098e-05,
+      "loss": 0.9971,
+      "step": 18255
+    },
+    {
+      "epoch": 0.7389720760825577,
+      "grad_norm": 1.1998182535171509,
+      "learning_rate": 2.66363824075986e-05,
+      "loss": 1.0246,
+      "step": 18260
+    },
+    {
+      "epoch": 0.7391744233104006,
+      "grad_norm": 1.2484065294265747,
+      "learning_rate": 2.6615734049143093e-05,
+      "loss": 0.9283,
+      "step": 18265
+    },
+    {
+      "epoch": 0.7393767705382436,
+      "grad_norm": 1.207814335823059,
+      "learning_rate": 2.6595085690687593e-05,
+      "loss": 0.9862,
+      "step": 18270
+    },
+    {
+      "epoch": 0.7395791177660866,
+      "grad_norm": 1.3312376737594604,
+      "learning_rate": 2.6574437332232087e-05,
+      "loss": 0.9849,
+      "step": 18275
+    },
+    {
+      "epoch": 0.7397814649939296,
+      "grad_norm": 1.3686442375183105,
+      "learning_rate": 2.6553788973776588e-05,
+      "loss": 1.017,
+      "step": 18280
+    },
+    {
+      "epoch": 0.7399838122217726,
+      "grad_norm": 1.3069236278533936,
+      "learning_rate": 2.6533140615321085e-05,
+      "loss": 0.9818,
+      "step": 18285
+    },
+    {
+      "epoch": 0.7401861594496155,
+      "grad_norm": 1.190818428993225,
+      "learning_rate": 2.651249225686558e-05,
+      "loss": 0.9691,
+      "step": 18290
+    },
+    {
+      "epoch": 0.7403885066774585,
+      "grad_norm": 1.130037784576416,
+      "learning_rate": 2.649184389841008e-05,
+      "loss": 0.9361,
+      "step": 18295
+    },
+    {
+      "epoch": 0.7405908539053015,
+      "grad_norm": 1.138445258140564,
+      "learning_rate": 2.6471195539954574e-05,
+      "loss": 1.0278,
+      "step": 18300
+    },
+    {
+      "epoch": 0.7407932011331445,
+      "grad_norm": 1.204694390296936,
+      "learning_rate": 2.6450547181499075e-05,
+      "loss": 0.9281,
+      "step": 18305
+    },
+    {
+      "epoch": 0.7409955483609875,
+      "grad_norm": 1.0873088836669922,
+      "learning_rate": 2.642989882304357e-05,
+      "loss": 0.9988,
+      "step": 18310
+    },
+    {
+      "epoch": 0.7411978955888304,
+      "grad_norm": 1.3609319925308228,
+      "learning_rate": 2.6409250464588066e-05,
+      "loss": 0.9882,
+      "step": 18315
+    },
+    {
+      "epoch": 0.7414002428166734,
+      "grad_norm": 1.2003790140151978,
+      "learning_rate": 2.6388602106132566e-05,
+      "loss": 0.9606,
+      "step": 18320
+    },
+    {
+      "epoch": 0.7416025900445163,
+      "grad_norm": 1.2015271186828613,
+      "learning_rate": 2.636795374767706e-05,
+      "loss": 0.9433,
+      "step": 18325
+    },
+    {
+      "epoch": 0.7418049372723594,
+      "grad_norm": 1.1974081993103027,
+      "learning_rate": 2.634730538922156e-05,
+      "loss": 0.9562,
+      "step": 18330
+    },
+    {
+      "epoch": 0.7420072845002024,
+      "grad_norm": 1.1607143878936768,
+      "learning_rate": 2.6326657030766055e-05,
+      "loss": 0.9258,
+      "step": 18335
+    },
+    {
+      "epoch": 0.7422096317280453,
+      "grad_norm": 1.2583789825439453,
+      "learning_rate": 2.630600867231055e-05,
+      "loss": 0.9446,
+      "step": 18340
+    },
+    {
+      "epoch": 0.7424119789558883,
+      "grad_norm": 1.2502634525299072,
+      "learning_rate": 2.628536031385505e-05,
+      "loss": 0.9587,
+      "step": 18345
+    },
+    {
+      "epoch": 0.7426143261837312,
+      "grad_norm": 1.1891366243362427,
+      "learning_rate": 2.6264711955399547e-05,
+      "loss": 0.9987,
+      "step": 18350
+    },
+    {
+      "epoch": 0.7428166734115743,
+      "grad_norm": 1.2171727418899536,
+      "learning_rate": 2.6244063596944048e-05,
+      "loss": 0.9488,
+      "step": 18355
+    },
+    {
+      "epoch": 0.7430190206394173,
+      "grad_norm": 1.162196159362793,
+      "learning_rate": 2.622341523848854e-05,
+      "loss": 0.9942,
+      "step": 18360
+    },
+    {
+      "epoch": 0.7432213678672602,
+      "grad_norm": 1.180525302886963,
+      "learning_rate": 2.6202766880033035e-05,
+      "loss": 0.9381,
+      "step": 18365
+    },
+    {
+      "epoch": 0.7434237150951032,
+      "grad_norm": 1.3493324518203735,
+      "learning_rate": 2.6182118521577536e-05,
+      "loss": 1.011,
+      "step": 18370
+    },
+    {
+      "epoch": 0.7436260623229461,
+      "grad_norm": 1.1396207809448242,
+      "learning_rate": 2.616147016312203e-05,
+      "loss": 1.005,
+      "step": 18375
+    },
+    {
+      "epoch": 0.7438284095507891,
+      "grad_norm": 1.1421303749084473,
+      "learning_rate": 2.614082180466653e-05,
+      "loss": 1.0281,
+      "step": 18380
+    },
+    {
+      "epoch": 0.7440307567786322,
+      "grad_norm": 1.2090312242507935,
+      "learning_rate": 2.6120173446211028e-05,
+      "loss": 1.0517,
+      "step": 18385
+    },
+    {
+      "epoch": 0.7442331040064751,
+      "grad_norm": 1.2150853872299194,
+      "learning_rate": 2.6099525087755522e-05,
+      "loss": 1.0239,
+      "step": 18390
+    },
+    {
+      "epoch": 0.7444354512343181,
+      "grad_norm": 1.1943917274475098,
+      "learning_rate": 2.6078876729300023e-05,
+      "loss": 1.0505,
+      "step": 18395
+    },
+    {
+      "epoch": 0.744637798462161,
+      "grad_norm": 1.2696770429611206,
+      "learning_rate": 2.6058228370844516e-05,
+      "loss": 0.9861,
+      "step": 18400
+    },
+    {
+      "epoch": 0.744840145690004,
+      "grad_norm": 1.2047666311264038,
+      "learning_rate": 2.6037580012389017e-05,
+      "loss": 1.029,
+      "step": 18405
+    },
+    {
+      "epoch": 0.7450424929178471,
+      "grad_norm": 1.2024877071380615,
+      "learning_rate": 2.6016931653933514e-05,
+      "loss": 1.0143,
+      "step": 18410
+    },
+    {
+      "epoch": 0.74524484014569,
+      "grad_norm": 1.1922520399093628,
+      "learning_rate": 2.5996283295478012e-05,
+      "loss": 0.9752,
+      "step": 18415
+    },
+    {
+      "epoch": 0.745447187373533,
+      "grad_norm": 1.1806377172470093,
+      "learning_rate": 2.597563493702251e-05,
+      "loss": 1.0268,
+      "step": 18420
+    },
+    {
+      "epoch": 0.745649534601376,
+      "grad_norm": 1.2532659769058228,
+      "learning_rate": 2.5954986578567003e-05,
+      "loss": 0.9856,
+      "step": 18425
+    },
+    {
+      "epoch": 0.7458518818292189,
+      "grad_norm": 1.15655517578125,
+      "learning_rate": 2.5934338220111504e-05,
+      "loss": 0.9504,
+      "step": 18430
+    },
+    {
+      "epoch": 0.7460542290570619,
+      "grad_norm": 1.2378734350204468,
+      "learning_rate": 2.5913689861655998e-05,
+      "loss": 0.9727,
+      "step": 18435
+    },
+    {
+      "epoch": 0.7462565762849049,
+      "grad_norm": 1.1206599473953247,
+      "learning_rate": 2.5893041503200498e-05,
+      "loss": 0.9938,
+      "step": 18440
+    },
+    {
+      "epoch": 0.7464589235127479,
+      "grad_norm": 1.2198963165283203,
+      "learning_rate": 2.5872393144744995e-05,
+      "loss": 1.053,
+      "step": 18445
+    },
+    {
+      "epoch": 0.7466612707405909,
+      "grad_norm": 1.2086889743804932,
+      "learning_rate": 2.585174478628949e-05,
+      "loss": 1.007,
+      "step": 18450
+    },
+    {
+      "epoch": 0.7468636179684338,
+      "grad_norm": 1.107006549835205,
+      "learning_rate": 2.583109642783399e-05,
+      "loss": 0.9291,
+      "step": 18455
+    },
+    {
+      "epoch": 0.7470659651962768,
+      "grad_norm": 1.2523084878921509,
+      "learning_rate": 2.5810448069378484e-05,
+      "loss": 0.9592,
+      "step": 18460
+    },
+    {
+      "epoch": 0.7472683124241198,
+      "grad_norm": 1.2677992582321167,
+      "learning_rate": 2.5789799710922985e-05,
+      "loss": 0.9707,
+      "step": 18465
+    },
+    {
+      "epoch": 0.7474706596519628,
+      "grad_norm": 1.2503540515899658,
+      "learning_rate": 2.576915135246748e-05,
+      "loss": 1.0012,
+      "step": 18470
+    },
+    {
+      "epoch": 0.7476730068798058,
+      "grad_norm": 1.2988989353179932,
+      "learning_rate": 2.5748502994011976e-05,
+      "loss": 0.9533,
+      "step": 18475
+    },
+    {
+      "epoch": 0.7478753541076487,
+      "grad_norm": 1.226706862449646,
+      "learning_rate": 2.5727854635556477e-05,
+      "loss": 0.9822,
+      "step": 18480
+    },
+    {
+      "epoch": 0.7480777013354917,
+      "grad_norm": 1.1212823390960693,
+      "learning_rate": 2.570720627710097e-05,
+      "loss": 0.9749,
+      "step": 18485
+    },
+    {
+      "epoch": 0.7482800485633346,
+      "grad_norm": 1.1801321506500244,
+      "learning_rate": 2.568655791864547e-05,
+      "loss": 0.9993,
+      "step": 18490
+    },
+    {
+      "epoch": 0.7484823957911777,
+      "grad_norm": 1.2597798109054565,
+      "learning_rate": 2.5665909560189965e-05,
+      "loss": 1.0327,
+      "step": 18495
+    },
+    {
+      "epoch": 0.7486847430190207,
+      "grad_norm": 1.2265557050704956,
+      "learning_rate": 2.5645261201734462e-05,
+      "loss": 0.9664,
+      "step": 18500
+    },
+    {
+      "epoch": 0.7488870902468636,
+      "grad_norm": 1.0432270765304565,
+      "learning_rate": 2.562461284327896e-05,
+      "loss": 0.9637,
+      "step": 18505
+    },
+    {
+      "epoch": 0.7490894374747066,
+      "grad_norm": 1.0922579765319824,
+      "learning_rate": 2.5603964484823457e-05,
+      "loss": 1.0562,
+      "step": 18510
+    },
+    {
+      "epoch": 0.7492917847025495,
+      "grad_norm": 1.1779067516326904,
+      "learning_rate": 2.5583316126367958e-05,
+      "loss": 0.9537,
+      "step": 18515
+    },
+    {
+      "epoch": 0.7494941319303926,
+      "grad_norm": 1.1106982231140137,
+      "learning_rate": 2.556266776791245e-05,
+      "loss": 1.0006,
+      "step": 18520
+    },
+    {
+      "epoch": 0.7496964791582356,
+      "grad_norm": 1.337260127067566,
+      "learning_rate": 2.5542019409456952e-05,
+      "loss": 1.0258,
+      "step": 18525
+    },
+    {
+      "epoch": 0.7498988263860785,
+      "grad_norm": 1.141217827796936,
+      "learning_rate": 2.5521371051001446e-05,
+      "loss": 1.0355,
+      "step": 18530
+    },
+    {
+      "epoch": 0.7501011736139215,
+      "grad_norm": 1.1664879322052002,
+      "learning_rate": 2.5500722692545943e-05,
+      "loss": 0.9495,
+      "step": 18535
+    },
+    {
+      "epoch": 0.7503035208417644,
+      "grad_norm": 1.19535231590271,
+      "learning_rate": 2.548007433409044e-05,
+      "loss": 0.9651,
+      "step": 18540
+    },
+    {
+      "epoch": 0.7505058680696074,
+      "grad_norm": 1.162934422492981,
+      "learning_rate": 2.5459425975634938e-05,
+      "loss": 1.0744,
+      "step": 18545
+    },
+    {
+      "epoch": 0.7507082152974505,
+      "grad_norm": 1.2206164598464966,
+      "learning_rate": 2.543877761717944e-05,
+      "loss": 0.9585,
+      "step": 18550
+    },
+    {
+      "epoch": 0.7509105625252934,
+      "grad_norm": 1.2517008781433105,
+      "learning_rate": 2.5418129258723933e-05,
+      "loss": 1.0579,
+      "step": 18555
+    },
+    {
+      "epoch": 0.7511129097531364,
+      "grad_norm": 1.2082105875015259,
+      "learning_rate": 2.5397480900268427e-05,
+      "loss": 1.0415,
+      "step": 18560
+    },
+    {
+      "epoch": 0.7513152569809793,
+      "grad_norm": 1.3701311349868774,
+      "learning_rate": 2.5376832541812927e-05,
+      "loss": 0.9334,
+      "step": 18565
+    },
+    {
+      "epoch": 0.7515176042088223,
+      "grad_norm": 1.1496940851211548,
+      "learning_rate": 2.5356184183357425e-05,
+      "loss": 1.0017,
+      "step": 18570
+    },
+    {
+      "epoch": 0.7517199514366654,
+      "grad_norm": 1.1699267625808716,
+      "learning_rate": 2.5335535824901925e-05,
+      "loss": 1.0257,
+      "step": 18575
+    },
+    {
+      "epoch": 0.7519222986645083,
+      "grad_norm": 1.2081458568572998,
+      "learning_rate": 2.531488746644642e-05,
+      "loss": 0.978,
+      "step": 18580
+    },
+    {
+      "epoch": 0.7521246458923513,
+      "grad_norm": 1.2505255937576294,
+      "learning_rate": 2.5294239107990913e-05,
+      "loss": 1.0217,
+      "step": 18585
+    },
+    {
+      "epoch": 0.7523269931201942,
+      "grad_norm": 1.1666648387908936,
+      "learning_rate": 2.5273590749535414e-05,
+      "loss": 1.0024,
+      "step": 18590
+    },
+    {
+      "epoch": 0.7525293403480372,
+      "grad_norm": 1.2679240703582764,
+      "learning_rate": 2.5252942391079908e-05,
+      "loss": 0.9983,
+      "step": 18595
+    },
+    {
+      "epoch": 0.7527316875758802,
+      "grad_norm": 1.1022908687591553,
+      "learning_rate": 2.5232294032624408e-05,
+      "loss": 0.9505,
+      "step": 18600
+    },
+    {
+      "epoch": 0.7529340348037232,
+      "grad_norm": 1.192799687385559,
+      "learning_rate": 2.5211645674168906e-05,
+      "loss": 1.0234,
+      "step": 18605
+    },
+    {
+      "epoch": 0.7531363820315662,
+      "grad_norm": 1.08863365650177,
+      "learning_rate": 2.51909973157134e-05,
+      "loss": 0.9991,
+      "step": 18610
+    },
+    {
+      "epoch": 0.7533387292594091,
+      "grad_norm": 1.1891133785247803,
+      "learning_rate": 2.51703489572579e-05,
+      "loss": 1.0023,
+      "step": 18615
+    },
+    {
+      "epoch": 0.7535410764872521,
+      "grad_norm": 1.1604515314102173,
+      "learning_rate": 2.5149700598802394e-05,
+      "loss": 1.0229,
+      "step": 18620
+    },
+    {
+      "epoch": 0.7537434237150951,
+      "grad_norm": 1.220996618270874,
+      "learning_rate": 2.5129052240346895e-05,
+      "loss": 0.9946,
+      "step": 18625
+    },
+    {
+      "epoch": 0.7539457709429381,
+      "grad_norm": 1.2904902696609497,
+      "learning_rate": 2.510840388189139e-05,
+      "loss": 0.9669,
+      "step": 18630
+    },
+    {
+      "epoch": 0.7541481181707811,
+      "grad_norm": 1.170644760131836,
+      "learning_rate": 2.508775552343589e-05,
+      "loss": 1.0126,
+      "step": 18635
+    },
+    {
+      "epoch": 0.754350465398624,
+      "grad_norm": 1.1381133794784546,
+      "learning_rate": 2.5067107164980387e-05,
+      "loss": 0.9748,
+      "step": 18640
+    },
+    {
+      "epoch": 0.754552812626467,
+      "grad_norm": 1.2390223741531372,
+      "learning_rate": 2.504645880652488e-05,
+      "loss": 0.9901,
+      "step": 18645
+    },
+    {
+      "epoch": 0.75475515985431,
+      "grad_norm": 1.189243197441101,
+      "learning_rate": 2.502581044806938e-05,
+      "loss": 0.9997,
+      "step": 18650
+    },
+    {
+      "epoch": 0.7549575070821529,
+      "grad_norm": 1.1826403141021729,
+      "learning_rate": 2.5005162089613875e-05,
+      "loss": 0.9858,
+      "step": 18655
+    },
+    {
+      "epoch": 0.755159854309996,
+      "grad_norm": 1.1260607242584229,
+      "learning_rate": 2.4984513731158372e-05,
+      "loss": 1.0355,
+      "step": 18660
+    },
+    {
+      "epoch": 0.755362201537839,
+      "grad_norm": 1.3457492589950562,
+      "learning_rate": 2.4963865372702873e-05,
+      "loss": 1.0246,
+      "step": 18665
+    },
+    {
+      "epoch": 0.7555645487656819,
+      "grad_norm": 1.0560047626495361,
+      "learning_rate": 2.494321701424737e-05,
+      "loss": 0.9743,
+      "step": 18670
+    },
+    {
+      "epoch": 0.7557668959935249,
+      "grad_norm": 1.07358980178833,
+      "learning_rate": 2.4922568655791868e-05,
+      "loss": 0.9708,
+      "step": 18675
+    },
+    {
+      "epoch": 0.7559692432213678,
+      "grad_norm": 1.1762304306030273,
+      "learning_rate": 2.490192029733636e-05,
+      "loss": 1.0239,
+      "step": 18680
+    },
+    {
+      "epoch": 0.7561715904492109,
+      "grad_norm": 1.247923731803894,
+      "learning_rate": 2.488127193888086e-05,
+      "loss": 0.9804,
+      "step": 18685
+    },
+    {
+      "epoch": 0.7563739376770539,
+      "grad_norm": 1.2488702535629272,
+      "learning_rate": 2.4860623580425356e-05,
+      "loss": 0.9402,
+      "step": 18690
+    },
+    {
+      "epoch": 0.7565762849048968,
+      "grad_norm": 1.3600049018859863,
+      "learning_rate": 2.4839975221969854e-05,
+      "loss": 1.0327,
+      "step": 18695
+    },
+    {
+      "epoch": 0.7567786321327398,
+      "grad_norm": 1.2319860458374023,
+      "learning_rate": 2.4819326863514354e-05,
+      "loss": 1.0062,
+      "step": 18700
+    },
+    {
+      "epoch": 0.7569809793605827,
+      "grad_norm": 1.2301151752471924,
+      "learning_rate": 2.4798678505058848e-05,
+      "loss": 0.9879,
+      "step": 18705
+    },
+    {
+      "epoch": 0.7571833265884257,
+      "grad_norm": 1.2247709035873413,
+      "learning_rate": 2.4778030146603345e-05,
+      "loss": 1.0125,
+      "step": 18710
+    },
+    {
+      "epoch": 0.7573856738162688,
+      "grad_norm": 1.2979135513305664,
+      "learning_rate": 2.4757381788147843e-05,
+      "loss": 0.9305,
+      "step": 18715
+    },
+    {
+      "epoch": 0.7575880210441117,
+      "grad_norm": 1.1458330154418945,
+      "learning_rate": 2.473673342969234e-05,
+      "loss": 0.9594,
+      "step": 18720
+    },
+    {
+      "epoch": 0.7577903682719547,
+      "grad_norm": 1.2261348962783813,
+      "learning_rate": 2.4716085071236837e-05,
+      "loss": 1.0207,
+      "step": 18725
+    },
+    {
+      "epoch": 0.7579927154997976,
+      "grad_norm": 1.2570581436157227,
+      "learning_rate": 2.4695436712781335e-05,
+      "loss": 1.0073,
+      "step": 18730
+    },
+    {
+      "epoch": 0.7581950627276406,
+      "grad_norm": 1.1916133165359497,
+      "learning_rate": 2.4674788354325832e-05,
+      "loss": 0.9727,
+      "step": 18735
+    },
+    {
+      "epoch": 0.7583974099554837,
+      "grad_norm": 1.1151262521743774,
+      "learning_rate": 2.465413999587033e-05,
+      "loss": 0.9785,
+      "step": 18740
+    },
+    {
+      "epoch": 0.7585997571833266,
+      "grad_norm": 1.2051528692245483,
+      "learning_rate": 2.4633491637414827e-05,
+      "loss": 0.9576,
+      "step": 18745
+    },
+    {
+      "epoch": 0.7588021044111696,
+      "grad_norm": 1.2141988277435303,
+      "learning_rate": 2.4612843278959324e-05,
+      "loss": 0.951,
+      "step": 18750
+    },
+    {
+      "epoch": 0.7590044516390125,
+      "grad_norm": 1.1166486740112305,
+      "learning_rate": 2.459219492050382e-05,
+      "loss": 1.0134,
+      "step": 18755
+    },
+    {
+      "epoch": 0.7592067988668555,
+      "grad_norm": 1.1245037317276,
+      "learning_rate": 2.457154656204832e-05,
+      "loss": 0.9814,
+      "step": 18760
+    },
+    {
+      "epoch": 0.7594091460946985,
+      "grad_norm": 1.2884442806243896,
+      "learning_rate": 2.4550898203592816e-05,
+      "loss": 0.9701,
+      "step": 18765
+    },
+    {
+      "epoch": 0.7596114933225415,
+      "grad_norm": 1.1675169467926025,
+      "learning_rate": 2.4530249845137313e-05,
+      "loss": 0.9914,
+      "step": 18770
+    },
+    {
+      "epoch": 0.7598138405503845,
+      "grad_norm": 1.2772468328475952,
+      "learning_rate": 2.450960148668181e-05,
+      "loss": 1.0261,
+      "step": 18775
+    },
+    {
+      "epoch": 0.7600161877782274,
+      "grad_norm": 1.1847087144851685,
+      "learning_rate": 2.4488953128226308e-05,
+      "loss": 0.983,
+      "step": 18780
+    },
+    {
+      "epoch": 0.7602185350060704,
+      "grad_norm": 1.2734636068344116,
+      "learning_rate": 2.4468304769770805e-05,
+      "loss": 0.9856,
+      "step": 18785
+    },
+    {
+      "epoch": 0.7604208822339134,
+      "grad_norm": 1.172006368637085,
+      "learning_rate": 2.4447656411315302e-05,
+      "loss": 1.0438,
+      "step": 18790
+    },
+    {
+      "epoch": 0.7606232294617564,
+      "grad_norm": 1.1985136270523071,
+      "learning_rate": 2.44270080528598e-05,
+      "loss": 0.9974,
+      "step": 18795
+    },
+    {
+      "epoch": 0.7608255766895994,
+      "grad_norm": 1.2858589887619019,
+      "learning_rate": 2.4406359694404297e-05,
+      "loss": 0.9356,
+      "step": 18800
+    },
+    {
+      "epoch": 0.7610279239174423,
+      "grad_norm": 1.2080332040786743,
+      "learning_rate": 2.4385711335948794e-05,
+      "loss": 0.9645,
+      "step": 18805
+    },
+    {
+      "epoch": 0.7612302711452853,
+      "grad_norm": 1.2445225715637207,
+      "learning_rate": 2.436506297749329e-05,
+      "loss": 1.0145,
+      "step": 18810
+    },
+    {
+      "epoch": 0.7614326183731283,
+      "grad_norm": 1.3031141757965088,
+      "learning_rate": 2.4344414619037785e-05,
+      "loss": 1.0283,
+      "step": 18815
+    },
+    {
+      "epoch": 0.7616349656009712,
+      "grad_norm": 1.1435433626174927,
+      "learning_rate": 2.4323766260582283e-05,
+      "loss": 0.9968,
+      "step": 18820
+    },
+    {
+      "epoch": 0.7618373128288143,
+      "grad_norm": 1.119293451309204,
+      "learning_rate": 2.4303117902126783e-05,
+      "loss": 1.0004,
+      "step": 18825
+    },
+    {
+      "epoch": 0.7620396600566572,
+      "grad_norm": 1.1149024963378906,
+      "learning_rate": 2.428246954367128e-05,
+      "loss": 1.0003,
+      "step": 18830
+    },
+    {
+      "epoch": 0.7622420072845002,
+      "grad_norm": 1.1140360832214355,
+      "learning_rate": 2.4261821185215778e-05,
+      "loss": 0.9941,
+      "step": 18835
+    },
+    {
+      "epoch": 0.7624443545123432,
+      "grad_norm": 1.1843329668045044,
+      "learning_rate": 2.4241172826760275e-05,
+      "loss": 1.0184,
+      "step": 18840
+    },
+    {
+      "epoch": 0.7626467017401861,
+      "grad_norm": 1.3155696392059326,
+      "learning_rate": 2.422052446830477e-05,
+      "loss": 0.94,
+      "step": 18845
+    },
+    {
+      "epoch": 0.7628490489680292,
+      "grad_norm": 1.4078176021575928,
+      "learning_rate": 2.4199876109849266e-05,
+      "loss": 0.9909,
+      "step": 18850
+    },
+    {
+      "epoch": 0.7630513961958721,
+      "grad_norm": 1.2036023139953613,
+      "learning_rate": 2.4179227751393767e-05,
+      "loss": 1.0051,
+      "step": 18855
+    },
+    {
+      "epoch": 0.7632537434237151,
+      "grad_norm": 1.0942049026489258,
+      "learning_rate": 2.4158579392938264e-05,
+      "loss": 0.9281,
+      "step": 18860
+    },
+    {
+      "epoch": 0.7634560906515581,
+      "grad_norm": 1.1924493312835693,
+      "learning_rate": 2.413793103448276e-05,
+      "loss": 1.004,
+      "step": 18865
+    },
+    {
+      "epoch": 0.763658437879401,
+      "grad_norm": 1.2332738637924194,
+      "learning_rate": 2.4117282676027256e-05,
+      "loss": 0.9553,
+      "step": 18870
+    },
+    {
+      "epoch": 0.763860785107244,
+      "grad_norm": 1.193235993385315,
+      "learning_rate": 2.4096634317571753e-05,
+      "loss": 0.9922,
+      "step": 18875
+    },
+    {
+      "epoch": 0.764063132335087,
+      "grad_norm": 1.0221155881881714,
+      "learning_rate": 2.407598595911625e-05,
+      "loss": 0.9689,
+      "step": 18880
+    },
+    {
+      "epoch": 0.76426547956293,
+      "grad_norm": 1.1542539596557617,
+      "learning_rate": 2.4055337600660747e-05,
+      "loss": 1.0137,
+      "step": 18885
+    },
+    {
+      "epoch": 0.764467826790773,
+      "grad_norm": 1.1557893753051758,
+      "learning_rate": 2.4034689242205248e-05,
+      "loss": 0.9654,
+      "step": 18890
+    },
+    {
+      "epoch": 0.7646701740186159,
+      "grad_norm": 1.3672646284103394,
+      "learning_rate": 2.4014040883749745e-05,
+      "loss": 1.0191,
+      "step": 18895
+    },
+    {
+      "epoch": 0.7648725212464589,
+      "grad_norm": 1.2017077207565308,
+      "learning_rate": 2.399339252529424e-05,
+      "loss": 0.9156,
+      "step": 18900
+    },
+    {
+      "epoch": 0.765074868474302,
+      "grad_norm": 1.2825926542282104,
+      "learning_rate": 2.3972744166838737e-05,
+      "loss": 1.0015,
+      "step": 18905
+    },
+    {
+      "epoch": 0.7652772157021449,
+      "grad_norm": 1.1321334838867188,
+      "learning_rate": 2.3952095808383234e-05,
+      "loss": 0.9771,
+      "step": 18910
+    },
+    {
+      "epoch": 0.7654795629299879,
+      "grad_norm": 1.218597412109375,
+      "learning_rate": 2.393144744992773e-05,
+      "loss": 0.9791,
+      "step": 18915
+    },
+    {
+      "epoch": 0.7656819101578308,
+      "grad_norm": 1.1967525482177734,
+      "learning_rate": 2.391079909147223e-05,
+      "loss": 0.9675,
+      "step": 18920
+    },
+    {
+      "epoch": 0.7658842573856738,
+      "grad_norm": 1.3359050750732422,
+      "learning_rate": 2.3890150733016726e-05,
+      "loss": 1.0157,
+      "step": 18925
+    },
+    {
+      "epoch": 0.7660866046135167,
+      "grad_norm": 1.2007734775543213,
+      "learning_rate": 2.3869502374561223e-05,
+      "loss": 0.9727,
+      "step": 18930
+    },
+    {
+      "epoch": 0.7662889518413598,
+      "grad_norm": 1.2456916570663452,
+      "learning_rate": 2.384885401610572e-05,
+      "loss": 0.9768,
+      "step": 18935
+    },
+    {
+      "epoch": 0.7664912990692028,
+      "grad_norm": 1.2281389236450195,
+      "learning_rate": 2.3828205657650218e-05,
+      "loss": 1.0122,
+      "step": 18940
+    },
+    {
+      "epoch": 0.7666936462970457,
+      "grad_norm": 1.1848171949386597,
+      "learning_rate": 2.3807557299194715e-05,
+      "loss": 1.04,
+      "step": 18945
+    },
+    {
+      "epoch": 0.7668959935248887,
+      "grad_norm": 1.1801064014434814,
+      "learning_rate": 2.3786908940739212e-05,
+      "loss": 0.9442,
+      "step": 18950
+    },
+    {
+      "epoch": 0.7670983407527316,
+      "grad_norm": 1.1730283498764038,
+      "learning_rate": 2.376626058228371e-05,
+      "loss": 1.001,
+      "step": 18955
+    },
+    {
+      "epoch": 0.7673006879805747,
+      "grad_norm": 1.118220567703247,
+      "learning_rate": 2.3745612223828207e-05,
+      "loss": 0.9982,
+      "step": 18960
+    },
+    {
+      "epoch": 0.7675030352084177,
+      "grad_norm": 1.1394208669662476,
+      "learning_rate": 2.3724963865372704e-05,
+      "loss": 1.0106,
+      "step": 18965
+    },
+    {
+      "epoch": 0.7677053824362606,
+      "grad_norm": 1.1889700889587402,
+      "learning_rate": 2.37043155069172e-05,
+      "loss": 1.0927,
+      "step": 18970
+    },
+    {
+      "epoch": 0.7679077296641036,
+      "grad_norm": 1.1230664253234863,
+      "learning_rate": 2.36836671484617e-05,
+      "loss": 1.0148,
+      "step": 18975
+    },
+    {
+      "epoch": 0.7681100768919465,
+      "grad_norm": 1.2890617847442627,
+      "learning_rate": 2.3663018790006196e-05,
+      "loss": 1.0196,
+      "step": 18980
+    },
+    {
+      "epoch": 0.7683124241197895,
+      "grad_norm": 1.144126057624817,
+      "learning_rate": 2.3642370431550693e-05,
+      "loss": 0.9868,
+      "step": 18985
+    },
+    {
+      "epoch": 0.7685147713476326,
+      "grad_norm": 1.1859136819839478,
+      "learning_rate": 2.362172207309519e-05,
+      "loss": 0.9755,
+      "step": 18990
+    },
+    {
+      "epoch": 0.7687171185754755,
+      "grad_norm": 1.289074182510376,
+      "learning_rate": 2.3601073714639688e-05,
+      "loss": 1.0042,
+      "step": 18995
+    },
+    {
+      "epoch": 0.7689194658033185,
+      "grad_norm": 1.0631299018859863,
+      "learning_rate": 2.3580425356184185e-05,
+      "loss": 0.9353,
+      "step": 19000
+    },
+    {
+      "epoch": 0.7691218130311614,
+      "grad_norm": 1.1243079900741577,
+      "learning_rate": 2.3559776997728683e-05,
+      "loss": 1.0122,
+      "step": 19005
+    },
+    {
+      "epoch": 0.7693241602590044,
+      "grad_norm": 1.182533860206604,
+      "learning_rate": 2.3539128639273176e-05,
+      "loss": 0.9918,
+      "step": 19010
+    },
+    {
+      "epoch": 0.7695265074868475,
+      "grad_norm": 1.144896149635315,
+      "learning_rate": 2.3518480280817677e-05,
+      "loss": 0.9295,
+      "step": 19015
+    },
+    {
+      "epoch": 0.7697288547146904,
+      "grad_norm": 1.1534823179244995,
+      "learning_rate": 2.3497831922362174e-05,
+      "loss": 0.9665,
+      "step": 19020
+    },
+    {
+      "epoch": 0.7699312019425334,
+      "grad_norm": 1.4073677062988281,
+      "learning_rate": 2.3477183563906672e-05,
+      "loss": 0.9633,
+      "step": 19025
+    },
+    {
+      "epoch": 0.7701335491703764,
+      "grad_norm": 1.3116756677627563,
+      "learning_rate": 2.345653520545117e-05,
+      "loss": 0.9783,
+      "step": 19030
+    },
+    {
+      "epoch": 0.7703358963982193,
+      "grad_norm": 1.2318789958953857,
+      "learning_rate": 2.3435886846995663e-05,
+      "loss": 0.9311,
+      "step": 19035
+    },
+    {
+      "epoch": 0.7705382436260623,
+      "grad_norm": 1.1697779893875122,
+      "learning_rate": 2.341523848854016e-05,
+      "loss": 0.913,
+      "step": 19040
+    },
+    {
+      "epoch": 0.7707405908539053,
+      "grad_norm": 1.149557113647461,
+      "learning_rate": 2.339459013008466e-05,
+      "loss": 0.944,
+      "step": 19045
+    },
+    {
+      "epoch": 0.7709429380817483,
+      "grad_norm": 1.1746495962142944,
+      "learning_rate": 2.3373941771629158e-05,
+      "loss": 0.9823,
+      "step": 19050
+    },
+    {
+      "epoch": 0.7711452853095913,
+      "grad_norm": 1.2647314071655273,
+      "learning_rate": 2.3353293413173656e-05,
+      "loss": 0.9882,
+      "step": 19055
+    },
+    {
+      "epoch": 0.7713476325374342,
+      "grad_norm": 1.1291464567184448,
+      "learning_rate": 2.333264505471815e-05,
+      "loss": 1.0234,
+      "step": 19060
+    },
+    {
+      "epoch": 0.7715499797652772,
+      "grad_norm": 1.2369909286499023,
+      "learning_rate": 2.3311996696262647e-05,
+      "loss": 0.9598,
+      "step": 19065
+    },
+    {
+      "epoch": 0.7717523269931202,
+      "grad_norm": 1.0962897539138794,
+      "learning_rate": 2.3291348337807144e-05,
+      "loss": 0.9624,
+      "step": 19070
+    },
+    {
+      "epoch": 0.7719546742209632,
+      "grad_norm": 1.1354165077209473,
+      "learning_rate": 2.327069997935164e-05,
+      "loss": 1.0069,
+      "step": 19075
+    },
+    {
+      "epoch": 0.7721570214488062,
+      "grad_norm": 1.2026914358139038,
+      "learning_rate": 2.3250051620896142e-05,
+      "loss": 1.0439,
+      "step": 19080
+    },
+    {
+      "epoch": 0.7723593686766491,
+      "grad_norm": 1.22047758102417,
+      "learning_rate": 2.322940326244064e-05,
+      "loss": 0.9916,
+      "step": 19085
+    },
+    {
+      "epoch": 0.7725617159044921,
+      "grad_norm": 1.1651899814605713,
+      "learning_rate": 2.3208754903985133e-05,
+      "loss": 1.0401,
+      "step": 19090
+    },
+    {
+      "epoch": 0.772764063132335,
+      "grad_norm": 1.252793550491333,
+      "learning_rate": 2.318810654552963e-05,
+      "loss": 0.976,
+      "step": 19095
+    },
+    {
+      "epoch": 0.7729664103601781,
+      "grad_norm": 1.2411421537399292,
+      "learning_rate": 2.3167458187074128e-05,
+      "loss": 0.9926,
+      "step": 19100
+    },
+    {
+      "epoch": 0.7731687575880211,
+      "grad_norm": 1.2860257625579834,
+      "learning_rate": 2.3146809828618625e-05,
+      "loss": 1.0801,
+      "step": 19105
+    },
+    {
+      "epoch": 0.773371104815864,
+      "grad_norm": 1.345198392868042,
+      "learning_rate": 2.3126161470163122e-05,
+      "loss": 0.9809,
+      "step": 19110
+    },
+    {
+      "epoch": 0.773573452043707,
+      "grad_norm": 1.2429323196411133,
+      "learning_rate": 2.310551311170762e-05,
+      "loss": 0.9996,
+      "step": 19115
+    },
+    {
+      "epoch": 0.7737757992715499,
+      "grad_norm": 1.1630768775939941,
+      "learning_rate": 2.3084864753252117e-05,
+      "loss": 0.9374,
+      "step": 19120
+    },
+    {
+      "epoch": 0.773978146499393,
+      "grad_norm": 1.2421936988830566,
+      "learning_rate": 2.3064216394796614e-05,
+      "loss": 0.9434,
+      "step": 19125
+    },
+    {
+      "epoch": 0.774180493727236,
+      "grad_norm": 1.2803579568862915,
+      "learning_rate": 2.304356803634111e-05,
+      "loss": 0.9821,
+      "step": 19130
+    },
+    {
+      "epoch": 0.7743828409550789,
+      "grad_norm": 1.2965850830078125,
+      "learning_rate": 2.302291967788561e-05,
+      "loss": 0.9934,
+      "step": 19135
+    },
+    {
+      "epoch": 0.7745851881829219,
+      "grad_norm": 1.1569753885269165,
+      "learning_rate": 2.3002271319430106e-05,
+      "loss": 0.9648,
+      "step": 19140
+    },
+    {
+      "epoch": 0.7747875354107648,
+      "grad_norm": 1.2425038814544678,
+      "learning_rate": 2.2981622960974603e-05,
+      "loss": 0.9707,
+      "step": 19145
+    },
+    {
+      "epoch": 0.7749898826386078,
+      "grad_norm": 1.103983759880066,
+      "learning_rate": 2.29609746025191e-05,
+      "loss": 0.9614,
+      "step": 19150
+    },
+    {
+      "epoch": 0.7751922298664509,
+      "grad_norm": 1.1858563423156738,
+      "learning_rate": 2.2940326244063598e-05,
+      "loss": 1.0251,
+      "step": 19155
+    },
+    {
+      "epoch": 0.7753945770942938,
+      "grad_norm": 1.1213321685791016,
+      "learning_rate": 2.2919677885608095e-05,
+      "loss": 0.9825,
+      "step": 19160
+    },
+    {
+      "epoch": 0.7755969243221368,
+      "grad_norm": 1.2138493061065674,
+      "learning_rate": 2.2899029527152593e-05,
+      "loss": 1.0044,
+      "step": 19165
+    },
+    {
+      "epoch": 0.7757992715499797,
+      "grad_norm": 1.0400367975234985,
+      "learning_rate": 2.287838116869709e-05,
+      "loss": 0.9729,
+      "step": 19170
+    },
+    {
+      "epoch": 0.7760016187778227,
+      "grad_norm": 1.2685893774032593,
+      "learning_rate": 2.2857732810241587e-05,
+      "loss": 0.9559,
+      "step": 19175
+    },
+    {
+      "epoch": 0.7762039660056658,
+      "grad_norm": 1.23403799533844,
+      "learning_rate": 2.2837084451786085e-05,
+      "loss": 0.9808,
+      "step": 19180
+    },
+    {
+      "epoch": 0.7764063132335087,
+      "grad_norm": 1.1738874912261963,
+      "learning_rate": 2.2816436093330582e-05,
+      "loss": 0.9195,
+      "step": 19185
+    },
+    {
+      "epoch": 0.7766086604613517,
+      "grad_norm": 1.0791808366775513,
+      "learning_rate": 2.279578773487508e-05,
+      "loss": 0.9871,
+      "step": 19190
+    },
+    {
+      "epoch": 0.7768110076891946,
+      "grad_norm": 1.1705849170684814,
+      "learning_rate": 2.2775139376419576e-05,
+      "loss": 1.0212,
+      "step": 19195
+    },
+    {
+      "epoch": 0.7770133549170376,
+      "grad_norm": 1.12641179561615,
+      "learning_rate": 2.275449101796407e-05,
+      "loss": 0.9875,
+      "step": 19200
+    },
+    {
+      "epoch": 0.7772157021448806,
+      "grad_norm": 1.144472360610962,
+      "learning_rate": 2.273384265950857e-05,
+      "loss": 1.0125,
+      "step": 19205
+    },
+    {
+      "epoch": 0.7774180493727236,
+      "grad_norm": 1.0404157638549805,
+      "learning_rate": 2.271319430105307e-05,
+      "loss": 0.9465,
+      "step": 19210
+    },
+    {
+      "epoch": 0.7776203966005666,
+      "grad_norm": 1.2508095502853394,
+      "learning_rate": 2.2692545942597566e-05,
+      "loss": 1.0029,
+      "step": 19215
+    },
+    {
+      "epoch": 0.7778227438284095,
+      "grad_norm": 1.2936362028121948,
+      "learning_rate": 2.2671897584142063e-05,
+      "loss": 0.9797,
+      "step": 19220
+    },
+    {
+      "epoch": 0.7780250910562525,
+      "grad_norm": 1.134230375289917,
+      "learning_rate": 2.2651249225686557e-05,
+      "loss": 0.9835,
+      "step": 19225
+    },
+    {
+      "epoch": 0.7782274382840955,
+      "grad_norm": 1.2599493265151978,
+      "learning_rate": 2.2630600867231054e-05,
+      "loss": 0.9651,
+      "step": 19230
+    },
+    {
+      "epoch": 0.7784297855119385,
+      "grad_norm": 1.1519733667373657,
+      "learning_rate": 2.2609952508775555e-05,
+      "loss": 1.0254,
+      "step": 19235
+    },
+    {
+      "epoch": 0.7786321327397815,
+      "grad_norm": 1.3085072040557861,
+      "learning_rate": 2.2589304150320052e-05,
+      "loss": 0.9717,
+      "step": 19240
+    },
+    {
+      "epoch": 0.7788344799676244,
+      "grad_norm": 1.1786202192306519,
+      "learning_rate": 2.256865579186455e-05,
+      "loss": 1.0672,
+      "step": 19245
+    },
+    {
+      "epoch": 0.7790368271954674,
+      "grad_norm": 1.0811524391174316,
+      "learning_rate": 2.2548007433409047e-05,
+      "loss": 0.9968,
+      "step": 19250
+    },
+    {
+      "epoch": 0.7792391744233104,
+      "grad_norm": 1.196601152420044,
+      "learning_rate": 2.252735907495354e-05,
+      "loss": 1.0126,
+      "step": 19255
+    },
+    {
+      "epoch": 0.7794415216511533,
+      "grad_norm": 1.213605523109436,
+      "learning_rate": 2.2506710716498038e-05,
+      "loss": 0.9326,
+      "step": 19260
+    },
+    {
+      "epoch": 0.7796438688789964,
+      "grad_norm": 1.202684760093689,
+      "learning_rate": 2.2486062358042535e-05,
+      "loss": 0.9132,
+      "step": 19265
+    },
+    {
+      "epoch": 0.7798462161068394,
+      "grad_norm": 1.096232533454895,
+      "learning_rate": 2.2465413999587036e-05,
+      "loss": 0.974,
+      "step": 19270
+    },
+    {
+      "epoch": 0.7800485633346823,
+      "grad_norm": 1.2750041484832764,
+      "learning_rate": 2.2444765641131533e-05,
+      "loss": 0.9448,
+      "step": 19275
+    },
+    {
+      "epoch": 0.7802509105625253,
+      "grad_norm": 1.304419755935669,
+      "learning_rate": 2.2424117282676027e-05,
+      "loss": 0.9944,
+      "step": 19280
+    },
+    {
+      "epoch": 0.7804532577903682,
+      "grad_norm": 1.141446828842163,
+      "learning_rate": 2.2403468924220524e-05,
+      "loss": 1.0035,
+      "step": 19285
+    },
+    {
+      "epoch": 0.7806556050182113,
+      "grad_norm": 1.2574546337127686,
+      "learning_rate": 2.2382820565765022e-05,
+      "loss": 1.0331,
+      "step": 19290
+    },
+    {
+      "epoch": 0.7808579522460543,
+      "grad_norm": 1.16201913356781,
+      "learning_rate": 2.236217220730952e-05,
+      "loss": 0.9733,
+      "step": 19295
+    },
+    {
+      "epoch": 0.7810602994738972,
+      "grad_norm": 1.1824009418487549,
+      "learning_rate": 2.2341523848854016e-05,
+      "loss": 0.9713,
+      "step": 19300
+    },
+    {
+      "epoch": 0.7812626467017402,
+      "grad_norm": 1.146752953529358,
+      "learning_rate": 2.2320875490398517e-05,
+      "loss": 1.039,
+      "step": 19305
+    },
+    {
+      "epoch": 0.7814649939295831,
+      "grad_norm": 1.1276365518569946,
+      "learning_rate": 2.230022713194301e-05,
+      "loss": 1.0075,
+      "step": 19310
+    },
+    {
+      "epoch": 0.7816673411574262,
+      "grad_norm": 1.2805054187774658,
+      "learning_rate": 2.2279578773487508e-05,
+      "loss": 1.0806,
+      "step": 19315
+    },
+    {
+      "epoch": 0.7818696883852692,
+      "grad_norm": 1.2467832565307617,
+      "learning_rate": 2.2258930415032005e-05,
+      "loss": 1.0169,
+      "step": 19320
+    },
+    {
+      "epoch": 0.7820720356131121,
+      "grad_norm": 1.325809121131897,
+      "learning_rate": 2.2238282056576503e-05,
+      "loss": 1.0255,
+      "step": 19325
+    },
+    {
+      "epoch": 0.7822743828409551,
+      "grad_norm": 1.2889117002487183,
+      "learning_rate": 2.2217633698121e-05,
+      "loss": 1.0637,
+      "step": 19330
+    },
+    {
+      "epoch": 0.782476730068798,
+      "grad_norm": 1.1927183866500854,
+      "learning_rate": 2.2196985339665497e-05,
+      "loss": 0.997,
+      "step": 19335
+    },
+    {
+      "epoch": 0.782679077296641,
+      "grad_norm": 1.2247055768966675,
+      "learning_rate": 2.2176336981209995e-05,
+      "loss": 1.0583,
+      "step": 19340
+    },
+    {
+      "epoch": 0.7828814245244841,
+      "grad_norm": 1.1576895713806152,
+      "learning_rate": 2.2155688622754492e-05,
+      "loss": 0.9673,
+      "step": 19345
+    },
+    {
+      "epoch": 0.783083771752327,
+      "grad_norm": 1.1851730346679688,
+      "learning_rate": 2.213504026429899e-05,
+      "loss": 1.0247,
+      "step": 19350
+    },
+    {
+      "epoch": 0.78328611898017,
+      "grad_norm": 1.150207757949829,
+      "learning_rate": 2.2114391905843487e-05,
+      "loss": 0.9963,
+      "step": 19355
+    },
+    {
+      "epoch": 0.7834884662080129,
+      "grad_norm": 1.1914957761764526,
+      "learning_rate": 2.2093743547387984e-05,
+      "loss": 1.0087,
+      "step": 19360
+    },
+    {
+      "epoch": 0.7836908134358559,
+      "grad_norm": 1.1222381591796875,
+      "learning_rate": 2.207309518893248e-05,
+      "loss": 0.9735,
+      "step": 19365
+    },
+    {
+      "epoch": 0.783893160663699,
+      "grad_norm": 1.2687716484069824,
+      "learning_rate": 2.205244683047698e-05,
+      "loss": 0.9979,
+      "step": 19370
+    },
+    {
+      "epoch": 0.7840955078915419,
+      "grad_norm": 1.1201457977294922,
+      "learning_rate": 2.2031798472021476e-05,
+      "loss": 0.9835,
+      "step": 19375
+    },
+    {
+      "epoch": 0.7842978551193849,
+      "grad_norm": 1.0393489599227905,
+      "learning_rate": 2.2011150113565973e-05,
+      "loss": 0.9536,
+      "step": 19380
+    },
+    {
+      "epoch": 0.7845002023472278,
+      "grad_norm": 1.2747719287872314,
+      "learning_rate": 2.199050175511047e-05,
+      "loss": 0.9486,
+      "step": 19385
+    },
+    {
+      "epoch": 0.7847025495750708,
+      "grad_norm": 1.3020176887512207,
+      "learning_rate": 2.1969853396654964e-05,
+      "loss": 1.0086,
+      "step": 19390
+    },
+    {
+      "epoch": 0.7849048968029138,
+      "grad_norm": 1.2747538089752197,
+      "learning_rate": 2.1949205038199465e-05,
+      "loss": 1.017,
+      "step": 19395
+    },
+    {
+      "epoch": 0.7851072440307568,
+      "grad_norm": 1.2838488817214966,
+      "learning_rate": 2.1928556679743962e-05,
+      "loss": 0.9527,
+      "step": 19400
+    },
+    {
+      "epoch": 0.7853095912585998,
+      "grad_norm": 1.147567629814148,
+      "learning_rate": 2.190790832128846e-05,
+      "loss": 0.9879,
+      "step": 19405
+    },
+    {
+      "epoch": 0.7855119384864427,
+      "grad_norm": 1.268541932106018,
+      "learning_rate": 2.1887259962832957e-05,
+      "loss": 0.973,
+      "step": 19410
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 1.1716783046722412,
+      "learning_rate": 2.1866611604377454e-05,
+      "loss": 0.9145,
+      "step": 19415
+    },
+    {
+      "epoch": 0.7859166329421287,
+      "grad_norm": 1.1512025594711304,
+      "learning_rate": 2.1845963245921948e-05,
+      "loss": 0.9786,
+      "step": 19420
+    },
+    {
+      "epoch": 0.7861189801699717,
+      "grad_norm": 1.2119401693344116,
+      "learning_rate": 2.182531488746645e-05,
+      "loss": 1.0029,
+      "step": 19425
+    },
+    {
+      "epoch": 0.7863213273978147,
+      "grad_norm": 1.2084400653839111,
+      "learning_rate": 2.1804666529010946e-05,
+      "loss": 0.9445,
+      "step": 19430
+    },
+    {
+      "epoch": 0.7865236746256576,
+      "grad_norm": 1.188834309577942,
+      "learning_rate": 2.1784018170555443e-05,
+      "loss": 1.0071,
+      "step": 19435
+    },
+    {
+      "epoch": 0.7867260218535006,
+      "grad_norm": 1.2610461711883545,
+      "learning_rate": 2.176336981209994e-05,
+      "loss": 1.0147,
+      "step": 19440
+    },
+    {
+      "epoch": 0.7869283690813436,
+      "grad_norm": 1.2548645734786987,
+      "learning_rate": 2.1742721453644435e-05,
+      "loss": 1.0113,
+      "step": 19445
+    },
+    {
+      "epoch": 0.7871307163091865,
+      "grad_norm": 1.177480936050415,
+      "learning_rate": 2.1722073095188932e-05,
+      "loss": 0.9405,
+      "step": 19450
+    },
+    {
+      "epoch": 0.7873330635370296,
+      "grad_norm": 1.2200291156768799,
+      "learning_rate": 2.170142473673343e-05,
+      "loss": 1.0175,
+      "step": 19455
+    },
+    {
+      "epoch": 0.7875354107648725,
+      "grad_norm": 1.196225643157959,
+      "learning_rate": 2.168077637827793e-05,
+      "loss": 1.0179,
+      "step": 19460
+    },
+    {
+      "epoch": 0.7877377579927155,
+      "grad_norm": 1.069580078125,
+      "learning_rate": 2.1660128019822427e-05,
+      "loss": 1.0087,
+      "step": 19465
+    },
+    {
+      "epoch": 0.7879401052205585,
+      "grad_norm": 1.1265087127685547,
+      "learning_rate": 2.1639479661366924e-05,
+      "loss": 0.969,
+      "step": 19470
+    },
+    {
+      "epoch": 0.7881424524484014,
+      "grad_norm": 1.1495747566223145,
+      "learning_rate": 2.161883130291142e-05,
+      "loss": 1.0225,
+      "step": 19475
+    },
+    {
+      "epoch": 0.7883447996762445,
+      "grad_norm": 1.1369599103927612,
+      "learning_rate": 2.1598182944455916e-05,
+      "loss": 0.9334,
+      "step": 19480
+    },
+    {
+      "epoch": 0.7885471469040874,
+      "grad_norm": 1.2167320251464844,
+      "learning_rate": 2.1577534586000413e-05,
+      "loss": 0.9984,
+      "step": 19485
+    },
+    {
+      "epoch": 0.7887494941319304,
+      "grad_norm": 1.1752022504806519,
+      "learning_rate": 2.155688622754491e-05,
+      "loss": 0.9765,
+      "step": 19490
+    },
+    {
+      "epoch": 0.7889518413597734,
+      "grad_norm": 1.2124340534210205,
+      "learning_rate": 2.153623786908941e-05,
+      "loss": 0.9868,
+      "step": 19495
+    },
+    {
+      "epoch": 0.7891541885876163,
+      "grad_norm": 1.1635890007019043,
+      "learning_rate": 2.1515589510633905e-05,
+      "loss": 0.9795,
+      "step": 19500
+    },
+    {
+      "epoch": 0.7893565358154593,
+      "grad_norm": 1.2324472665786743,
+      "learning_rate": 2.1494941152178402e-05,
+      "loss": 0.9402,
+      "step": 19505
+    },
+    {
+      "epoch": 0.7895588830433024,
+      "grad_norm": 1.2834147214889526,
+      "learning_rate": 2.14742927937229e-05,
+      "loss": 1.0143,
+      "step": 19510
+    },
+    {
+      "epoch": 0.7897612302711453,
+      "grad_norm": 1.2503693103790283,
+      "learning_rate": 2.1453644435267397e-05,
+      "loss": 0.9469,
+      "step": 19515
+    },
+    {
+      "epoch": 0.7899635774989883,
+      "grad_norm": 1.3614864349365234,
+      "learning_rate": 2.1432996076811894e-05,
+      "loss": 1.0108,
+      "step": 19520
+    },
+    {
+      "epoch": 0.7901659247268312,
+      "grad_norm": 1.2642440795898438,
+      "learning_rate": 2.141234771835639e-05,
+      "loss": 1.0366,
+      "step": 19525
+    },
+    {
+      "epoch": 0.7903682719546742,
+      "grad_norm": 1.1245198249816895,
+      "learning_rate": 2.139169935990089e-05,
+      "loss": 0.9645,
+      "step": 19530
+    },
+    {
+      "epoch": 0.7905706191825173,
+      "grad_norm": 1.3378099203109741,
+      "learning_rate": 2.1371051001445386e-05,
+      "loss": 1.0262,
+      "step": 19535
+    },
+    {
+      "epoch": 0.7907729664103602,
+      "grad_norm": 1.1884162425994873,
+      "learning_rate": 2.1350402642989883e-05,
+      "loss": 1.0256,
+      "step": 19540
+    },
+    {
+      "epoch": 0.7909753136382032,
+      "grad_norm": 1.0942997932434082,
+      "learning_rate": 2.132975428453438e-05,
+      "loss": 1.0105,
+      "step": 19545
+    },
+    {
+      "epoch": 0.7911776608660461,
+      "grad_norm": 1.1098393201828003,
+      "learning_rate": 2.1309105926078878e-05,
+      "loss": 0.9206,
+      "step": 19550
+    },
+    {
+      "epoch": 0.7913800080938891,
+      "grad_norm": 1.1302059888839722,
+      "learning_rate": 2.1288457567623375e-05,
+      "loss": 0.9331,
+      "step": 19555
+    },
+    {
+      "epoch": 0.791582355321732,
+      "grad_norm": 1.3323559761047363,
+      "learning_rate": 2.1267809209167872e-05,
+      "loss": 1.0318,
+      "step": 19560
+    },
+    {
+      "epoch": 0.7917847025495751,
+      "grad_norm": 1.1884725093841553,
+      "learning_rate": 2.124716085071237e-05,
+      "loss": 1.0405,
+      "step": 19565
+    },
+    {
+      "epoch": 0.7919870497774181,
+      "grad_norm": 1.2075164318084717,
+      "learning_rate": 2.1226512492256867e-05,
+      "loss": 1.0126,
+      "step": 19570
+    },
+    {
+      "epoch": 0.792189397005261,
+      "grad_norm": 1.1292778253555298,
+      "learning_rate": 2.1205864133801364e-05,
+      "loss": 1.0286,
+      "step": 19575
+    },
+    {
+      "epoch": 0.792391744233104,
+      "grad_norm": 1.2976049184799194,
+      "learning_rate": 2.1185215775345858e-05,
+      "loss": 1.0342,
+      "step": 19580
+    },
+    {
+      "epoch": 0.792594091460947,
+      "grad_norm": 1.214419960975647,
+      "learning_rate": 2.116456741689036e-05,
+      "loss": 1.0012,
+      "step": 19585
+    },
+    {
+      "epoch": 0.79279643868879,
+      "grad_norm": 1.1389003992080688,
+      "learning_rate": 2.1143919058434856e-05,
+      "loss": 1.0274,
+      "step": 19590
+    },
+    {
+      "epoch": 0.792998785916633,
+      "grad_norm": 1.192901372909546,
+      "learning_rate": 2.1123270699979353e-05,
+      "loss": 0.9447,
+      "step": 19595
+    },
+    {
+      "epoch": 0.7932011331444759,
+      "grad_norm": 1.3499056100845337,
+      "learning_rate": 2.110262234152385e-05,
+      "loss": 0.9936,
+      "step": 19600
+    },
+    {
+      "epoch": 0.7934034803723189,
+      "grad_norm": 1.1274263858795166,
+      "learning_rate": 2.1081973983068348e-05,
+      "loss": 0.9792,
+      "step": 19605
+    },
+    {
+      "epoch": 0.7936058276001619,
+      "grad_norm": 1.1653188467025757,
+      "learning_rate": 2.1061325624612842e-05,
+      "loss": 1.0546,
+      "step": 19610
+    },
+    {
+      "epoch": 0.7938081748280048,
+      "grad_norm": 1.2700755596160889,
+      "learning_rate": 2.104067726615734e-05,
+      "loss": 0.9734,
+      "step": 19615
+    },
+    {
+      "epoch": 0.7940105220558479,
+      "grad_norm": 1.2344692945480347,
+      "learning_rate": 2.102002890770184e-05,
+      "loss": 1.038,
+      "step": 19620
+    },
+    {
+      "epoch": 0.7942128692836908,
+      "grad_norm": 1.2672911882400513,
+      "learning_rate": 2.0999380549246337e-05,
+      "loss": 1.0092,
+      "step": 19625
+    },
+    {
+      "epoch": 0.7944152165115338,
+      "grad_norm": 1.196106195449829,
+      "learning_rate": 2.0978732190790835e-05,
+      "loss": 0.9841,
+      "step": 19630
+    },
+    {
+      "epoch": 0.7946175637393768,
+      "grad_norm": 1.1073206663131714,
+      "learning_rate": 2.095808383233533e-05,
+      "loss": 0.9167,
+      "step": 19635
+    },
+    {
+      "epoch": 0.7948199109672197,
+      "grad_norm": 1.2175101041793823,
+      "learning_rate": 2.0937435473879826e-05,
+      "loss": 0.9759,
+      "step": 19640
+    },
+    {
+      "epoch": 0.7950222581950628,
+      "grad_norm": 1.1923753023147583,
+      "learning_rate": 2.0916787115424323e-05,
+      "loss": 0.9698,
+      "step": 19645
+    },
+    {
+      "epoch": 0.7952246054229057,
+      "grad_norm": 1.3313026428222656,
+      "learning_rate": 2.0896138756968824e-05,
+      "loss": 1.0561,
+      "step": 19650
+    },
+    {
+      "epoch": 0.7954269526507487,
+      "grad_norm": 1.2001488208770752,
+      "learning_rate": 2.087549039851332e-05,
+      "loss": 0.9968,
+      "step": 19655
+    },
+    {
+      "epoch": 0.7956292998785917,
+      "grad_norm": 1.2324854135513306,
+      "learning_rate": 2.0854842040057818e-05,
+      "loss": 1.0547,
+      "step": 19660
+    },
+    {
+      "epoch": 0.7958316471064346,
+      "grad_norm": 1.3095812797546387,
+      "learning_rate": 2.0834193681602312e-05,
+      "loss": 0.9453,
+      "step": 19665
+    },
+    {
+      "epoch": 0.7960339943342776,
+      "grad_norm": 1.228178858757019,
+      "learning_rate": 2.081354532314681e-05,
+      "loss": 0.9482,
+      "step": 19670
+    },
+    {
+      "epoch": 0.7962363415621206,
+      "grad_norm": 1.1888363361358643,
+      "learning_rate": 2.0792896964691307e-05,
+      "loss": 1.0083,
+      "step": 19675
+    },
+    {
+      "epoch": 0.7964386887899636,
+      "grad_norm": 1.1231098175048828,
+      "learning_rate": 2.0772248606235804e-05,
+      "loss": 1.0199,
+      "step": 19680
+    },
+    {
+      "epoch": 0.7966410360178066,
+      "grad_norm": 1.189549207687378,
+      "learning_rate": 2.0751600247780305e-05,
+      "loss": 0.9818,
+      "step": 19685
+    },
+    {
+      "epoch": 0.7968433832456495,
+      "grad_norm": 1.248259425163269,
+      "learning_rate": 2.07309518893248e-05,
+      "loss": 0.9371,
+      "step": 19690
+    },
+    {
+      "epoch": 0.7970457304734925,
+      "grad_norm": 1.1903223991394043,
+      "learning_rate": 2.0710303530869296e-05,
+      "loss": 1.0038,
+      "step": 19695
+    },
+    {
+      "epoch": 0.7972480777013355,
+      "grad_norm": 1.0861550569534302,
+      "learning_rate": 2.0689655172413793e-05,
+      "loss": 0.9521,
+      "step": 19700
+    },
+    {
+      "epoch": 0.7974504249291785,
+      "grad_norm": 1.2282209396362305,
+      "learning_rate": 2.066900681395829e-05,
+      "loss": 0.9781,
+      "step": 19705
+    },
+    {
+      "epoch": 0.7976527721570215,
+      "grad_norm": 1.270195722579956,
+      "learning_rate": 2.0648358455502788e-05,
+      "loss": 1.0046,
+      "step": 19710
+    },
+    {
+      "epoch": 0.7978551193848644,
+      "grad_norm": 1.261534333229065,
+      "learning_rate": 2.062771009704729e-05,
+      "loss": 0.9783,
+      "step": 19715
+    },
+    {
+      "epoch": 0.7980574666127074,
+      "grad_norm": 1.1097315549850464,
+      "learning_rate": 2.0607061738591782e-05,
+      "loss": 0.9546,
+      "step": 19720
+    },
+    {
+      "epoch": 0.7982598138405503,
+      "grad_norm": 1.145125389099121,
+      "learning_rate": 2.058641338013628e-05,
+      "loss": 1.0244,
+      "step": 19725
+    },
+    {
+      "epoch": 0.7984621610683934,
+      "grad_norm": 1.2401931285858154,
+      "learning_rate": 2.0565765021680777e-05,
+      "loss": 0.9883,
+      "step": 19730
+    },
+    {
+      "epoch": 0.7986645082962364,
+      "grad_norm": 1.247721791267395,
+      "learning_rate": 2.0545116663225274e-05,
+      "loss": 1.0114,
+      "step": 19735
+    },
+    {
+      "epoch": 0.7988668555240793,
+      "grad_norm": 1.2011133432388306,
+      "learning_rate": 2.052446830476977e-05,
+      "loss": 0.989,
+      "step": 19740
+    },
+    {
+      "epoch": 0.7990692027519223,
+      "grad_norm": 1.2275869846343994,
+      "learning_rate": 2.050381994631427e-05,
+      "loss": 0.9969,
+      "step": 19745
+    },
+    {
+      "epoch": 0.7992715499797652,
+      "grad_norm": 1.237228512763977,
+      "learning_rate": 2.0483171587858766e-05,
+      "loss": 0.9614,
+      "step": 19750
+    },
+    {
+      "epoch": 0.7994738972076083,
+      "grad_norm": 1.1042230129241943,
+      "learning_rate": 2.0462523229403264e-05,
+      "loss": 1.0287,
+      "step": 19755
+    },
+    {
+      "epoch": 0.7996762444354513,
+      "grad_norm": 1.2946480512619019,
+      "learning_rate": 2.044187487094776e-05,
+      "loss": 0.994,
+      "step": 19760
+    },
+    {
+      "epoch": 0.7998785916632942,
+      "grad_norm": 1.187788963317871,
+      "learning_rate": 2.0421226512492258e-05,
+      "loss": 0.915,
+      "step": 19765
+    },
+    {
+      "epoch": 0.8000809388911372,
+      "grad_norm": 1.2385916709899902,
+      "learning_rate": 2.0400578154036755e-05,
+      "loss": 1.0061,
+      "step": 19770
+    },
+    {
+      "epoch": 0.8002832861189801,
+      "grad_norm": 1.1602866649627686,
+      "learning_rate": 2.0379929795581253e-05,
+      "loss": 0.9703,
+      "step": 19775
+    },
+    {
+      "epoch": 0.8004856333468231,
+      "grad_norm": 1.199293613433838,
+      "learning_rate": 2.035928143712575e-05,
+      "loss": 0.9906,
+      "step": 19780
+    },
+    {
+      "epoch": 0.8006879805746662,
+      "grad_norm": 1.4394046068191528,
+      "learning_rate": 2.0338633078670247e-05,
+      "loss": 0.9859,
+      "step": 19785
+    },
+    {
+      "epoch": 0.8008903278025091,
+      "grad_norm": 1.1568101644515991,
+      "learning_rate": 2.0317984720214745e-05,
+      "loss": 1.0067,
+      "step": 19790
+    },
+    {
+      "epoch": 0.8010926750303521,
+      "grad_norm": 1.117028832435608,
+      "learning_rate": 2.0297336361759242e-05,
+      "loss": 0.9826,
+      "step": 19795
+    },
+    {
+      "epoch": 0.801295022258195,
+      "grad_norm": 1.211776852607727,
+      "learning_rate": 2.0276688003303736e-05,
+      "loss": 1.0046,
+      "step": 19800
+    },
+    {
+      "epoch": 0.801497369486038,
+      "grad_norm": 1.2188165187835693,
+      "learning_rate": 2.0256039644848233e-05,
+      "loss": 1.0403,
+      "step": 19805
+    },
+    {
+      "epoch": 0.8016997167138811,
+      "grad_norm": 1.2479783296585083,
+      "learning_rate": 2.0235391286392734e-05,
+      "loss": 1.0141,
+      "step": 19810
+    },
+    {
+      "epoch": 0.801902063941724,
+      "grad_norm": 1.2970749139785767,
+      "learning_rate": 2.021474292793723e-05,
+      "loss": 0.9122,
+      "step": 19815
+    },
+    {
+      "epoch": 0.802104411169567,
+      "grad_norm": 1.168317198753357,
+      "learning_rate": 2.019409456948173e-05,
+      "loss": 0.9999,
+      "step": 19820
+    },
+    {
+      "epoch": 0.80230675839741,
+      "grad_norm": 1.144818663597107,
+      "learning_rate": 2.0173446211026226e-05,
+      "loss": 0.9321,
+      "step": 19825
+    },
+    {
+      "epoch": 0.8025091056252529,
+      "grad_norm": 1.1624873876571655,
+      "learning_rate": 2.015279785257072e-05,
+      "loss": 0.9882,
+      "step": 19830
+    },
+    {
+      "epoch": 0.8027114528530959,
+      "grad_norm": 1.2556883096694946,
+      "learning_rate": 2.0132149494115217e-05,
+      "loss": 1.0018,
+      "step": 19835
+    },
+    {
+      "epoch": 0.8029138000809389,
+      "grad_norm": 1.088374137878418,
+      "learning_rate": 2.0111501135659718e-05,
+      "loss": 0.9678,
+      "step": 19840
+    },
+    {
+      "epoch": 0.8031161473087819,
+      "grad_norm": 1.1550886631011963,
+      "learning_rate": 2.0090852777204215e-05,
+      "loss": 0.9705,
+      "step": 19845
+    },
+    {
+      "epoch": 0.8033184945366248,
+      "grad_norm": 1.2255133390426636,
+      "learning_rate": 2.0070204418748712e-05,
+      "loss": 1.0041,
+      "step": 19850
+    },
+    {
+      "epoch": 0.8035208417644678,
+      "grad_norm": 1.132543921470642,
+      "learning_rate": 2.0049556060293206e-05,
+      "loss": 0.9867,
+      "step": 19855
+    },
+    {
+      "epoch": 0.8037231889923108,
+      "grad_norm": 1.3151130676269531,
+      "learning_rate": 2.0028907701837703e-05,
+      "loss": 0.9502,
+      "step": 19860
+    },
+    {
+      "epoch": 0.8039255362201538,
+      "grad_norm": 1.1357910633087158,
+      "learning_rate": 2.00082593433822e-05,
+      "loss": 0.9613,
+      "step": 19865
+    },
+    {
+      "epoch": 0.8041278834479968,
+      "grad_norm": 1.2936217784881592,
+      "learning_rate": 1.9987610984926698e-05,
+      "loss": 0.9899,
+      "step": 19870
+    },
+    {
+      "epoch": 0.8043302306758398,
+      "grad_norm": 1.5135611295700073,
+      "learning_rate": 1.99669626264712e-05,
+      "loss": 0.9504,
+      "step": 19875
+    },
+    {
+      "epoch": 0.8045325779036827,
+      "grad_norm": 1.2707353830337524,
+      "learning_rate": 1.9946314268015696e-05,
+      "loss": 0.9918,
+      "step": 19880
+    },
+    {
+      "epoch": 0.8047349251315257,
+      "grad_norm": 1.180137276649475,
+      "learning_rate": 1.992566590956019e-05,
+      "loss": 1.0152,
+      "step": 19885
+    },
+    {
+      "epoch": 0.8049372723593686,
+      "grad_norm": 1.2100902795791626,
+      "learning_rate": 1.9905017551104687e-05,
+      "loss": 1.0173,
+      "step": 19890
+    },
+    {
+      "epoch": 0.8051396195872117,
+      "grad_norm": 1.2103663682937622,
+      "learning_rate": 1.9884369192649184e-05,
+      "loss": 0.9532,
+      "step": 19895
+    },
+    {
+      "epoch": 0.8053419668150547,
+      "grad_norm": 1.119215965270996,
+      "learning_rate": 1.9863720834193682e-05,
+      "loss": 1.0026,
+      "step": 19900
+    },
+    {
+      "epoch": 0.8055443140428976,
+      "grad_norm": 1.3070436716079712,
+      "learning_rate": 1.9843072475738182e-05,
+      "loss": 1.0098,
+      "step": 19905
+    },
+    {
+      "epoch": 0.8057466612707406,
+      "grad_norm": 1.1795378923416138,
+      "learning_rate": 1.9822424117282676e-05,
+      "loss": 1.0136,
+      "step": 19910
+    },
+    {
+      "epoch": 0.8059490084985835,
+      "grad_norm": 1.3635504245758057,
+      "learning_rate": 1.9801775758827174e-05,
+      "loss": 0.9739,
+      "step": 19915
+    },
+    {
+      "epoch": 0.8061513557264266,
+      "grad_norm": 1.170520305633545,
+      "learning_rate": 1.978112740037167e-05,
+      "loss": 1.0263,
+      "step": 19920
+    },
+    {
+      "epoch": 0.8063537029542696,
+      "grad_norm": 1.30668044090271,
+      "learning_rate": 1.9760479041916168e-05,
+      "loss": 0.9868,
+      "step": 19925
+    },
+    {
+      "epoch": 0.8065560501821125,
+      "grad_norm": 1.2427819967269897,
+      "learning_rate": 1.9739830683460666e-05,
+      "loss": 1.0359,
+      "step": 19930
+    },
+    {
+      "epoch": 0.8067583974099555,
+      "grad_norm": 1.1163017749786377,
+      "learning_rate": 1.9719182325005163e-05,
+      "loss": 0.9936,
+      "step": 19935
+    },
+    {
+      "epoch": 0.8069607446377984,
+      "grad_norm": 1.2274726629257202,
+      "learning_rate": 1.969853396654966e-05,
+      "loss": 0.9545,
+      "step": 19940
+    },
+    {
+      "epoch": 0.8071630918656414,
+      "grad_norm": 1.4208333492279053,
+      "learning_rate": 1.9677885608094157e-05,
+      "loss": 1.0062,
+      "step": 19945
+    },
+    {
+      "epoch": 0.8073654390934845,
+      "grad_norm": 1.2495779991149902,
+      "learning_rate": 1.9657237249638655e-05,
+      "loss": 1.0276,
+      "step": 19950
+    },
+    {
+      "epoch": 0.8075677863213274,
+      "grad_norm": 1.2706964015960693,
+      "learning_rate": 1.9636588891183152e-05,
+      "loss": 0.9762,
+      "step": 19955
+    },
+    {
+      "epoch": 0.8077701335491704,
+      "grad_norm": 1.169280767440796,
+      "learning_rate": 1.961594053272765e-05,
+      "loss": 0.9751,
+      "step": 19960
+    },
+    {
+      "epoch": 0.8079724807770133,
+      "grad_norm": 1.1972851753234863,
+      "learning_rate": 1.9595292174272147e-05,
+      "loss": 1.0169,
+      "step": 19965
+    },
+    {
+      "epoch": 0.8081748280048563,
+      "grad_norm": 1.202343225479126,
+      "learning_rate": 1.9574643815816644e-05,
+      "loss": 0.9836,
+      "step": 19970
+    },
+    {
+      "epoch": 0.8083771752326994,
+      "grad_norm": 1.2811102867126465,
+      "learning_rate": 1.955399545736114e-05,
+      "loss": 0.9891,
+      "step": 19975
+    },
+    {
+      "epoch": 0.8085795224605423,
+      "grad_norm": 1.1969845294952393,
+      "learning_rate": 1.953334709890564e-05,
+      "loss": 1.0155,
+      "step": 19980
+    },
+    {
+      "epoch": 0.8087818696883853,
+      "grad_norm": 1.2557622194290161,
+      "learning_rate": 1.9512698740450136e-05,
+      "loss": 0.9148,
+      "step": 19985
+    },
+    {
+      "epoch": 0.8089842169162282,
+      "grad_norm": 1.2543156147003174,
+      "learning_rate": 1.9492050381994633e-05,
+      "loss": 0.9626,
+      "step": 19990
+    },
+    {
+      "epoch": 0.8091865641440712,
+      "grad_norm": 1.1905945539474487,
+      "learning_rate": 1.9471402023539127e-05,
+      "loss": 1.0148,
+      "step": 19995
+    },
+    {
+      "epoch": 0.8093889113719142,
+      "grad_norm": 1.2504193782806396,
+      "learning_rate": 1.9450753665083628e-05,
+      "loss": 0.9622,
+      "step": 20000
+    },
+    {
+      "epoch": 0.8095912585997572,
+      "grad_norm": 1.1896506547927856,
+      "learning_rate": 1.9430105306628125e-05,
+      "loss": 0.9763,
+      "step": 20005
+    },
+    {
+      "epoch": 0.8097936058276002,
+      "grad_norm": 1.116180658340454,
+      "learning_rate": 1.9409456948172622e-05,
+      "loss": 0.9942,
+      "step": 20010
+    },
+    {
+      "epoch": 0.8099959530554431,
+      "grad_norm": 1.131353735923767,
+      "learning_rate": 1.938880858971712e-05,
+      "loss": 0.9795,
+      "step": 20015
+    },
+    {
+      "epoch": 0.8101983002832861,
+      "grad_norm": 1.2329113483428955,
+      "learning_rate": 1.9368160231261613e-05,
+      "loss": 1.0307,
+      "step": 20020
+    },
+    {
+      "epoch": 0.8104006475111291,
+      "grad_norm": 1.2695358991622925,
+      "learning_rate": 1.934751187280611e-05,
+      "loss": 1.039,
+      "step": 20025
+    },
+    {
+      "epoch": 0.8106029947389721,
+      "grad_norm": 1.142235517501831,
+      "learning_rate": 1.932686351435061e-05,
+      "loss": 1.0011,
+      "step": 20030
+    },
+    {
+      "epoch": 0.8108053419668151,
+      "grad_norm": 1.2903319597244263,
+      "learning_rate": 1.930621515589511e-05,
+      "loss": 0.981,
+      "step": 20035
+    },
+    {
+      "epoch": 0.811007689194658,
+      "grad_norm": 1.3007261753082275,
+      "learning_rate": 1.9285566797439606e-05,
+      "loss": 0.9658,
+      "step": 20040
+    },
+    {
+      "epoch": 0.811210036422501,
+      "grad_norm": 1.1206128597259521,
+      "learning_rate": 1.92649184389841e-05,
+      "loss": 1.0229,
+      "step": 20045
+    },
+    {
+      "epoch": 0.811412383650344,
+      "grad_norm": 1.1064411401748657,
+      "learning_rate": 1.9244270080528597e-05,
+      "loss": 1.0258,
+      "step": 20050
+    },
+    {
+      "epoch": 0.8116147308781869,
+      "grad_norm": 1.0990204811096191,
+      "learning_rate": 1.9223621722073095e-05,
+      "loss": 1.0013,
+      "step": 20055
+    },
+    {
+      "epoch": 0.81181707810603,
+      "grad_norm": 1.1460708379745483,
+      "learning_rate": 1.9202973363617592e-05,
+      "loss": 0.96,
+      "step": 20060
+    },
+    {
+      "epoch": 0.812019425333873,
+      "grad_norm": 1.2765713930130005,
+      "learning_rate": 1.9182325005162093e-05,
+      "loss": 1.0415,
+      "step": 20065
+    },
+    {
+      "epoch": 0.8122217725617159,
+      "grad_norm": 1.1970674991607666,
+      "learning_rate": 1.916167664670659e-05,
+      "loss": 1.0275,
+      "step": 20070
+    },
+    {
+      "epoch": 0.8124241197895589,
+      "grad_norm": 1.2413722276687622,
+      "learning_rate": 1.9141028288251084e-05,
+      "loss": 1.022,
+      "step": 20075
+    },
+    {
+      "epoch": 0.8126264670174018,
+      "grad_norm": 1.1680649518966675,
+      "learning_rate": 1.912037992979558e-05,
+      "loss": 0.9867,
+      "step": 20080
+    },
+    {
+      "epoch": 0.8128288142452449,
+      "grad_norm": 1.2049671411514282,
+      "learning_rate": 1.909973157134008e-05,
+      "loss": 1.0504,
+      "step": 20085
+    },
+    {
+      "epoch": 0.8130311614730878,
+      "grad_norm": 1.235776662826538,
+      "learning_rate": 1.9079083212884576e-05,
+      "loss": 1.0066,
+      "step": 20090
+    },
+    {
+      "epoch": 0.8132335087009308,
+      "grad_norm": 1.249600887298584,
+      "learning_rate": 1.9058434854429076e-05,
+      "loss": 0.995,
+      "step": 20095
+    },
+    {
+      "epoch": 0.8134358559287738,
+      "grad_norm": 1.2152076959609985,
+      "learning_rate": 1.903778649597357e-05,
+      "loss": 1.0192,
+      "step": 20100
+    },
+    {
+      "epoch": 0.8136382031566167,
+      "grad_norm": 1.2431395053863525,
+      "learning_rate": 1.9017138137518068e-05,
+      "loss": 0.9872,
+      "step": 20105
+    },
+    {
+      "epoch": 0.8138405503844597,
+      "grad_norm": 1.156506061553955,
+      "learning_rate": 1.8996489779062565e-05,
+      "loss": 1.0307,
+      "step": 20110
+    },
+    {
+      "epoch": 0.8140428976123028,
+      "grad_norm": 1.1493124961853027,
+      "learning_rate": 1.8975841420607062e-05,
+      "loss": 0.9804,
+      "step": 20115
+    },
+    {
+      "epoch": 0.8142452448401457,
+      "grad_norm": 1.266171932220459,
+      "learning_rate": 1.895519306215156e-05,
+      "loss": 0.9843,
+      "step": 20120
+    },
+    {
+      "epoch": 0.8144475920679887,
+      "grad_norm": 1.1572659015655518,
+      "learning_rate": 1.8934544703696057e-05,
+      "loss": 1.0514,
+      "step": 20125
+    },
+    {
+      "epoch": 0.8146499392958316,
+      "grad_norm": 1.2233699560165405,
+      "learning_rate": 1.8913896345240554e-05,
+      "loss": 0.9665,
+      "step": 20130
+    },
+    {
+      "epoch": 0.8148522865236746,
+      "grad_norm": 1.1813246011734009,
+      "learning_rate": 1.889324798678505e-05,
+      "loss": 0.9978,
+      "step": 20135
+    },
+    {
+      "epoch": 0.8150546337515177,
+      "grad_norm": 1.1056386232376099,
+      "learning_rate": 1.887259962832955e-05,
+      "loss": 0.9947,
+      "step": 20140
+    },
+    {
+      "epoch": 0.8152569809793606,
+      "grad_norm": 1.2665637731552124,
+      "learning_rate": 1.8851951269874046e-05,
+      "loss": 1.0013,
+      "step": 20145
+    },
+    {
+      "epoch": 0.8154593282072036,
+      "grad_norm": 1.3943125009536743,
+      "learning_rate": 1.8831302911418543e-05,
+      "loss": 0.9596,
+      "step": 20150
+    },
+    {
+      "epoch": 0.8156616754350465,
+      "grad_norm": 1.2365728616714478,
+      "learning_rate": 1.881065455296304e-05,
+      "loss": 0.9706,
+      "step": 20155
+    },
+    {
+      "epoch": 0.8158640226628895,
+      "grad_norm": 1.1528480052947998,
+      "learning_rate": 1.8790006194507538e-05,
+      "loss": 0.9637,
+      "step": 20160
+    },
+    {
+      "epoch": 0.8160663698907324,
+      "grad_norm": 1.5409117937088013,
+      "learning_rate": 1.8769357836052035e-05,
+      "loss": 0.9236,
+      "step": 20165
+    },
+    {
+      "epoch": 0.8162687171185755,
+      "grad_norm": 1.258923888206482,
+      "learning_rate": 1.8748709477596532e-05,
+      "loss": 0.9726,
+      "step": 20170
+    },
+    {
+      "epoch": 0.8164710643464185,
+      "grad_norm": 1.1251558065414429,
+      "learning_rate": 1.872806111914103e-05,
+      "loss": 1.0228,
+      "step": 20175
+    },
+    {
+      "epoch": 0.8166734115742614,
+      "grad_norm": 1.2459688186645508,
+      "learning_rate": 1.8707412760685527e-05,
+      "loss": 0.9896,
+      "step": 20180
+    },
+    {
+      "epoch": 0.8168757588021044,
+      "grad_norm": 1.2140971422195435,
+      "learning_rate": 1.868676440223002e-05,
+      "loss": 1.0325,
+      "step": 20185
+    },
+    {
+      "epoch": 0.8170781060299473,
+      "grad_norm": 1.1849210262298584,
+      "learning_rate": 1.866611604377452e-05,
+      "loss": 1.0246,
+      "step": 20190
+    },
+    {
+      "epoch": 0.8172804532577904,
+      "grad_norm": 1.1096030473709106,
+      "learning_rate": 1.864546768531902e-05,
+      "loss": 1.0264,
+      "step": 20195
+    },
+    {
+      "epoch": 0.8174828004856334,
+      "grad_norm": 1.3153741359710693,
+      "learning_rate": 1.8624819326863516e-05,
+      "loss": 0.9969,
+      "step": 20200
+    },
+    {
+      "epoch": 0.8176851477134763,
+      "grad_norm": 1.021003246307373,
+      "learning_rate": 1.8604170968408013e-05,
+      "loss": 0.992,
+      "step": 20205
+    },
+    {
+      "epoch": 0.8178874949413193,
+      "grad_norm": 1.2519673109054565,
+      "learning_rate": 1.8583522609952507e-05,
+      "loss": 1.0064,
+      "step": 20210
+    },
+    {
+      "epoch": 0.8180898421691623,
+      "grad_norm": 1.209879994392395,
+      "learning_rate": 1.8562874251497005e-05,
+      "loss": 0.9879,
+      "step": 20215
+    },
+    {
+      "epoch": 0.8182921893970052,
+      "grad_norm": 1.195630669593811,
+      "learning_rate": 1.8542225893041505e-05,
+      "loss": 1.0022,
+      "step": 20220
+    },
+    {
+      "epoch": 0.8184945366248483,
+      "grad_norm": 1.342789888381958,
+      "learning_rate": 1.8521577534586003e-05,
+      "loss": 0.9606,
+      "step": 20225
+    },
+    {
+      "epoch": 0.8186968838526912,
+      "grad_norm": 1.3297576904296875,
+      "learning_rate": 1.85009291761305e-05,
+      "loss": 1.0015,
+      "step": 20230
+    },
+    {
+      "epoch": 0.8188992310805342,
+      "grad_norm": 1.3977317810058594,
+      "learning_rate": 1.8480280817674997e-05,
+      "loss": 0.9668,
+      "step": 20235
+    },
+    {
+      "epoch": 0.8191015783083772,
+      "grad_norm": 1.2348957061767578,
+      "learning_rate": 1.845963245921949e-05,
+      "loss": 0.9717,
+      "step": 20240
+    },
+    {
+      "epoch": 0.8193039255362201,
+      "grad_norm": 1.2327253818511963,
+      "learning_rate": 1.843898410076399e-05,
+      "loss": 1.018,
+      "step": 20245
+    },
+    {
+      "epoch": 0.8195062727640632,
+      "grad_norm": 1.0961917638778687,
+      "learning_rate": 1.8418335742308486e-05,
+      "loss": 1.0093,
+      "step": 20250
+    },
+    {
+      "epoch": 0.8197086199919061,
+      "grad_norm": 1.3516782522201538,
+      "learning_rate": 1.8397687383852986e-05,
+      "loss": 1.0199,
+      "step": 20255
+    },
+    {
+      "epoch": 0.8199109672197491,
+      "grad_norm": 1.3711910247802734,
+      "learning_rate": 1.8377039025397484e-05,
+      "loss": 1.0221,
+      "step": 20260
+    },
+    {
+      "epoch": 0.8201133144475921,
+      "grad_norm": 1.2947044372558594,
+      "learning_rate": 1.8356390666941978e-05,
+      "loss": 0.9439,
+      "step": 20265
+    },
+    {
+      "epoch": 0.820315661675435,
+      "grad_norm": 1.1903367042541504,
+      "learning_rate": 1.8335742308486475e-05,
+      "loss": 1.0293,
+      "step": 20270
+    },
+    {
+      "epoch": 0.820518008903278,
+      "grad_norm": 1.2941244840621948,
+      "learning_rate": 1.8315093950030972e-05,
+      "loss": 0.9934,
+      "step": 20275
+    },
+    {
+      "epoch": 0.820720356131121,
+      "grad_norm": 1.1714632511138916,
+      "learning_rate": 1.829444559157547e-05,
+      "loss": 1.0082,
+      "step": 20280
+    },
+    {
+      "epoch": 0.820922703358964,
+      "grad_norm": 1.4111402034759521,
+      "learning_rate": 1.827379723311997e-05,
+      "loss": 0.9716,
+      "step": 20285
+    },
+    {
+      "epoch": 0.821125050586807,
+      "grad_norm": 1.1551733016967773,
+      "learning_rate": 1.8253148874664468e-05,
+      "loss": 0.9969,
+      "step": 20290
+    },
+    {
+      "epoch": 0.8213273978146499,
+      "grad_norm": 1.2137131690979004,
+      "learning_rate": 1.823250051620896e-05,
+      "loss": 0.9523,
+      "step": 20295
+    },
+    {
+      "epoch": 0.8215297450424929,
+      "grad_norm": 1.05117666721344,
+      "learning_rate": 1.821185215775346e-05,
+      "loss": 1.0274,
+      "step": 20300
+    },
+    {
+      "epoch": 0.821732092270336,
+      "grad_norm": 1.204932451248169,
+      "learning_rate": 1.8191203799297956e-05,
+      "loss": 0.9704,
+      "step": 20305
+    },
+    {
+      "epoch": 0.8219344394981789,
+      "grad_norm": 1.1883883476257324,
+      "learning_rate": 1.8170555440842453e-05,
+      "loss": 1.0267,
+      "step": 20310
+    },
+    {
+      "epoch": 0.8221367867260219,
+      "grad_norm": 1.2452350854873657,
+      "learning_rate": 1.814990708238695e-05,
+      "loss": 0.9956,
+      "step": 20315
+    },
+    {
+      "epoch": 0.8223391339538648,
+      "grad_norm": 1.1953701972961426,
+      "learning_rate": 1.8129258723931448e-05,
+      "loss": 1.0129,
+      "step": 20320
+    },
+    {
+      "epoch": 0.8225414811817078,
+      "grad_norm": 1.3015085458755493,
+      "learning_rate": 1.8108610365475945e-05,
+      "loss": 1.0441,
+      "step": 20325
+    },
+    {
+      "epoch": 0.8227438284095507,
+      "grad_norm": 1.2774007320404053,
+      "learning_rate": 1.8087962007020443e-05,
+      "loss": 1.0372,
+      "step": 20330
+    },
+    {
+      "epoch": 0.8229461756373938,
+      "grad_norm": 1.350160837173462,
+      "learning_rate": 1.806731364856494e-05,
+      "loss": 0.9749,
+      "step": 20335
+    },
+    {
+      "epoch": 0.8231485228652368,
+      "grad_norm": 1.2322993278503418,
+      "learning_rate": 1.8046665290109437e-05,
+      "loss": 0.9913,
+      "step": 20340
+    },
+    {
+      "epoch": 0.8233508700930797,
+      "grad_norm": 1.2659082412719727,
+      "learning_rate": 1.8026016931653934e-05,
+      "loss": 0.9835,
+      "step": 20345
+    },
+    {
+      "epoch": 0.8235532173209227,
+      "grad_norm": 1.2427709102630615,
+      "learning_rate": 1.8005368573198432e-05,
+      "loss": 0.9948,
+      "step": 20350
+    },
+    {
+      "epoch": 0.8237555645487656,
+      "grad_norm": 1.1547057628631592,
+      "learning_rate": 1.798472021474293e-05,
+      "loss": 1.0473,
+      "step": 20355
+    },
+    {
+      "epoch": 0.8239579117766087,
+      "grad_norm": 1.237739086151123,
+      "learning_rate": 1.7964071856287426e-05,
+      "loss": 0.9942,
+      "step": 20360
+    },
+    {
+      "epoch": 0.8241602590044517,
+      "grad_norm": 1.1475869417190552,
+      "learning_rate": 1.7943423497831924e-05,
+      "loss": 1.0079,
+      "step": 20365
+    },
+    {
+      "epoch": 0.8243626062322946,
+      "grad_norm": 1.3094673156738281,
+      "learning_rate": 1.792277513937642e-05,
+      "loss": 0.9764,
+      "step": 20370
+    },
+    {
+      "epoch": 0.8245649534601376,
+      "grad_norm": 1.2824714183807373,
+      "learning_rate": 1.7902126780920915e-05,
+      "loss": 1.0004,
+      "step": 20375
+    },
+    {
+      "epoch": 0.8247673006879805,
+      "grad_norm": 1.0989863872528076,
+      "learning_rate": 1.7881478422465415e-05,
+      "loss": 1.0083,
+      "step": 20380
+    },
+    {
+      "epoch": 0.8249696479158235,
+      "grad_norm": 1.2977863550186157,
+      "learning_rate": 1.7860830064009913e-05,
+      "loss": 0.9982,
+      "step": 20385
+    },
+    {
+      "epoch": 0.8251719951436666,
+      "grad_norm": 1.4900519847869873,
+      "learning_rate": 1.784018170555441e-05,
+      "loss": 0.9782,
+      "step": 20390
+    },
+    {
+      "epoch": 0.8253743423715095,
+      "grad_norm": 1.1205878257751465,
+      "learning_rate": 1.7819533347098907e-05,
+      "loss": 1.0262,
+      "step": 20395
+    },
+    {
+      "epoch": 0.8255766895993525,
+      "grad_norm": 1.1933003664016724,
+      "learning_rate": 1.7798884988643405e-05,
+      "loss": 0.9792,
+      "step": 20400
+    },
+    {
+      "epoch": 0.8257790368271954,
+      "grad_norm": 1.175807237625122,
+      "learning_rate": 1.77782366301879e-05,
+      "loss": 1.0551,
+      "step": 20405
+    },
+    {
+      "epoch": 0.8259813840550384,
+      "grad_norm": 1.477211356163025,
+      "learning_rate": 1.77575882717324e-05,
+      "loss": 0.9905,
+      "step": 20410
+    },
+    {
+      "epoch": 0.8261837312828815,
+      "grad_norm": 1.0942533016204834,
+      "learning_rate": 1.7736939913276897e-05,
+      "loss": 0.9547,
+      "step": 20415
+    },
+    {
+      "epoch": 0.8263860785107244,
+      "grad_norm": 1.2155954837799072,
+      "learning_rate": 1.7716291554821394e-05,
+      "loss": 0.999,
+      "step": 20420
+    },
+    {
+      "epoch": 0.8265884257385674,
+      "grad_norm": 1.205565333366394,
+      "learning_rate": 1.769564319636589e-05,
+      "loss": 0.9823,
+      "step": 20425
+    },
+    {
+      "epoch": 0.8267907729664103,
+      "grad_norm": 1.3723074197769165,
+      "learning_rate": 1.7674994837910385e-05,
+      "loss": 0.9774,
+      "step": 20430
+    },
+    {
+      "epoch": 0.8269931201942533,
+      "grad_norm": 1.211984395980835,
+      "learning_rate": 1.7654346479454882e-05,
+      "loss": 0.967,
+      "step": 20435
+    },
+    {
+      "epoch": 0.8271954674220963,
+      "grad_norm": 1.1118062734603882,
+      "learning_rate": 1.763369812099938e-05,
+      "loss": 0.9869,
+      "step": 20440
+    },
+    {
+      "epoch": 0.8273978146499393,
+      "grad_norm": 1.2070207595825195,
+      "learning_rate": 1.761304976254388e-05,
+      "loss": 1.0136,
+      "step": 20445
+    },
+    {
+      "epoch": 0.8276001618777823,
+      "grad_norm": 1.2678449153900146,
+      "learning_rate": 1.7592401404088378e-05,
+      "loss": 1.0071,
+      "step": 20450
+    },
+    {
+      "epoch": 0.8278025091056253,
+      "grad_norm": 1.0911531448364258,
+      "learning_rate": 1.7571753045632875e-05,
+      "loss": 1.0103,
+      "step": 20455
+    },
+    {
+      "epoch": 0.8280048563334682,
+      "grad_norm": 1.1416620016098022,
+      "learning_rate": 1.755110468717737e-05,
+      "loss": 0.9129,
+      "step": 20460
+    },
+    {
+      "epoch": 0.8282072035613112,
+      "grad_norm": 1.2879242897033691,
+      "learning_rate": 1.7530456328721866e-05,
+      "loss": 1.0324,
+      "step": 20465
+    },
+    {
+      "epoch": 0.8284095507891542,
+      "grad_norm": 1.1336830854415894,
+      "learning_rate": 1.7509807970266363e-05,
+      "loss": 0.9955,
+      "step": 20470
+    },
+    {
+      "epoch": 0.8286118980169972,
+      "grad_norm": 1.1080355644226074,
+      "learning_rate": 1.7489159611810864e-05,
+      "loss": 0.9883,
+      "step": 20475
+    },
+    {
+      "epoch": 0.8288142452448402,
+      "grad_norm": 1.2331814765930176,
+      "learning_rate": 1.746851125335536e-05,
+      "loss": 0.991,
+      "step": 20480
+    },
+    {
+      "epoch": 0.8290165924726831,
+      "grad_norm": 1.0507975816726685,
+      "learning_rate": 1.7447862894899855e-05,
+      "loss": 1.0184,
+      "step": 20485
+    },
+    {
+      "epoch": 0.8292189397005261,
+      "grad_norm": 1.3183051347732544,
+      "learning_rate": 1.7427214536444353e-05,
+      "loss": 0.9714,
+      "step": 20490
+    },
+    {
+      "epoch": 0.829421286928369,
+      "grad_norm": 1.249200463294983,
+      "learning_rate": 1.740656617798885e-05,
+      "loss": 0.9824,
+      "step": 20495
+    },
+    {
+      "epoch": 0.8296236341562121,
+      "grad_norm": 1.0716665983200073,
+      "learning_rate": 1.7385917819533347e-05,
+      "loss": 0.9555,
+      "step": 20500
+    },
+    {
+      "epoch": 0.8298259813840551,
+      "grad_norm": 1.250664472579956,
+      "learning_rate": 1.7365269461077845e-05,
+      "loss": 0.9864,
+      "step": 20505
+    },
+    {
+      "epoch": 0.830028328611898,
+      "grad_norm": 1.2126374244689941,
+      "learning_rate": 1.7344621102622342e-05,
+      "loss": 1.0792,
+      "step": 20510
+    },
+    {
+      "epoch": 0.830230675839741,
+      "grad_norm": 1.1063995361328125,
+      "learning_rate": 1.732397274416684e-05,
+      "loss": 1.0379,
+      "step": 20515
+    },
+    {
+      "epoch": 0.8304330230675839,
+      "grad_norm": 1.1586008071899414,
+      "learning_rate": 1.7303324385711336e-05,
+      "loss": 1.0363,
+      "step": 20520
+    },
+    {
+      "epoch": 0.830635370295427,
+      "grad_norm": 1.195989727973938,
+      "learning_rate": 1.7282676027255834e-05,
+      "loss": 0.9894,
+      "step": 20525
+    },
+    {
+      "epoch": 0.83083771752327,
+      "grad_norm": 1.1733986139297485,
+      "learning_rate": 1.726202766880033e-05,
+      "loss": 0.977,
+      "step": 20530
+    },
+    {
+      "epoch": 0.8310400647511129,
+      "grad_norm": 1.2116700410842896,
+      "learning_rate": 1.7241379310344828e-05,
+      "loss": 0.9803,
+      "step": 20535
+    },
+    {
+      "epoch": 0.8312424119789559,
+      "grad_norm": 1.207283616065979,
+      "learning_rate": 1.7220730951889326e-05,
+      "loss": 1.0284,
+      "step": 20540
+    },
+    {
+      "epoch": 0.8314447592067988,
+      "grad_norm": 1.0571120977401733,
+      "learning_rate": 1.7200082593433823e-05,
+      "loss": 1.0076,
+      "step": 20545
+    },
+    {
+      "epoch": 0.8316471064346418,
+      "grad_norm": 1.2281038761138916,
+      "learning_rate": 1.717943423497832e-05,
+      "loss": 1.0208,
+      "step": 20550
+    },
+    {
+      "epoch": 0.8318494536624849,
+      "grad_norm": 1.186851143836975,
+      "learning_rate": 1.7158785876522817e-05,
+      "loss": 1.0347,
+      "step": 20555
+    },
+    {
+      "epoch": 0.8320518008903278,
+      "grad_norm": 1.3567386865615845,
+      "learning_rate": 1.7138137518067315e-05,
+      "loss": 0.9777,
+      "step": 20560
+    },
+    {
+      "epoch": 0.8322541481181708,
+      "grad_norm": 1.3898897171020508,
+      "learning_rate": 1.711748915961181e-05,
+      "loss": 0.9936,
+      "step": 20565
+    },
+    {
+      "epoch": 0.8324564953460137,
+      "grad_norm": 1.253336787223816,
+      "learning_rate": 1.709684080115631e-05,
+      "loss": 0.966,
+      "step": 20570
+    },
+    {
+      "epoch": 0.8326588425738567,
+      "grad_norm": 1.2542872428894043,
+      "learning_rate": 1.7076192442700807e-05,
+      "loss": 0.9025,
+      "step": 20575
+    },
+    {
+      "epoch": 0.8328611898016998,
+      "grad_norm": 1.243143916130066,
+      "learning_rate": 1.7055544084245304e-05,
+      "loss": 0.93,
+      "step": 20580
+    },
+    {
+      "epoch": 0.8330635370295427,
+      "grad_norm": 1.2384790182113647,
+      "learning_rate": 1.70348957257898e-05,
+      "loss": 0.9604,
+      "step": 20585
+    },
+    {
+      "epoch": 0.8332658842573857,
+      "grad_norm": 1.2404310703277588,
+      "learning_rate": 1.70142473673343e-05,
+      "loss": 0.9954,
+      "step": 20590
+    },
+    {
+      "epoch": 0.8334682314852286,
+      "grad_norm": 1.1246404647827148,
+      "learning_rate": 1.6993599008878792e-05,
+      "loss": 0.9574,
+      "step": 20595
+    },
+    {
+      "epoch": 0.8336705787130716,
+      "grad_norm": 1.0322741270065308,
+      "learning_rate": 1.6972950650423293e-05,
+      "loss": 1.0129,
+      "step": 20600
+    },
+    {
+      "epoch": 0.8338729259409146,
+      "grad_norm": 1.1479917764663696,
+      "learning_rate": 1.695230229196779e-05,
+      "loss": 0.9937,
+      "step": 20605
+    },
+    {
+      "epoch": 0.8340752731687576,
+      "grad_norm": 1.127181053161621,
+      "learning_rate": 1.6931653933512288e-05,
+      "loss": 1.0409,
+      "step": 20610
+    },
+    {
+      "epoch": 0.8342776203966006,
+      "grad_norm": 1.1959785223007202,
+      "learning_rate": 1.6911005575056785e-05,
+      "loss": 0.9722,
+      "step": 20615
+    },
+    {
+      "epoch": 0.8344799676244435,
+      "grad_norm": 1.1904857158660889,
+      "learning_rate": 1.689035721660128e-05,
+      "loss": 0.9778,
+      "step": 20620
+    },
+    {
+      "epoch": 0.8346823148522865,
+      "grad_norm": 1.2734626531600952,
+      "learning_rate": 1.6869708858145776e-05,
+      "loss": 0.9056,
+      "step": 20625
+    },
+    {
+      "epoch": 0.8348846620801295,
+      "grad_norm": 1.2978413105010986,
+      "learning_rate": 1.6849060499690274e-05,
+      "loss": 1.001,
+      "step": 20630
+    },
+    {
+      "epoch": 0.8350870093079725,
+      "grad_norm": 1.2909311056137085,
+      "learning_rate": 1.6828412141234774e-05,
+      "loss": 0.9382,
+      "step": 20635
+    },
+    {
+      "epoch": 0.8352893565358155,
+      "grad_norm": 1.1042921543121338,
+      "learning_rate": 1.680776378277927e-05,
+      "loss": 0.967,
+      "step": 20640
+    },
+    {
+      "epoch": 0.8354917037636584,
+      "grad_norm": 1.1582454442977905,
+      "learning_rate": 1.678711542432377e-05,
+      "loss": 1.0093,
+      "step": 20645
+    },
+    {
+      "epoch": 0.8356940509915014,
+      "grad_norm": 1.2365312576293945,
+      "learning_rate": 1.6766467065868263e-05,
+      "loss": 1.0098,
+      "step": 20650
+    },
+    {
+      "epoch": 0.8358963982193444,
+      "grad_norm": 1.275153398513794,
+      "learning_rate": 1.674581870741276e-05,
+      "loss": 0.9795,
+      "step": 20655
+    },
+    {
+      "epoch": 0.8360987454471873,
+      "grad_norm": 1.328636884689331,
+      "learning_rate": 1.6725170348957257e-05,
+      "loss": 0.9655,
+      "step": 20660
+    },
+    {
+      "epoch": 0.8363010926750304,
+      "grad_norm": 1.2965084314346313,
+      "learning_rate": 1.6704521990501758e-05,
+      "loss": 0.9852,
+      "step": 20665
+    },
+    {
+      "epoch": 0.8365034399028733,
+      "grad_norm": 1.1821112632751465,
+      "learning_rate": 1.6683873632046255e-05,
+      "loss": 0.9975,
+      "step": 20670
+    },
+    {
+      "epoch": 0.8367057871307163,
+      "grad_norm": 1.143494725227356,
+      "learning_rate": 1.666322527359075e-05,
+      "loss": 0.9518,
+      "step": 20675
+    },
+    {
+      "epoch": 0.8369081343585593,
+      "grad_norm": 1.1797096729278564,
+      "learning_rate": 1.6642576915135247e-05,
+      "loss": 1.0115,
+      "step": 20680
+    },
+    {
+      "epoch": 0.8371104815864022,
+      "grad_norm": 1.2627525329589844,
+      "learning_rate": 1.6621928556679744e-05,
+      "loss": 1.006,
+      "step": 20685
+    },
+    {
+      "epoch": 0.8373128288142453,
+      "grad_norm": 1.2649866342544556,
+      "learning_rate": 1.660128019822424e-05,
+      "loss": 0.9186,
+      "step": 20690
+    },
+    {
+      "epoch": 0.8375151760420882,
+      "grad_norm": 1.2155084609985352,
+      "learning_rate": 1.658063183976874e-05,
+      "loss": 0.9853,
+      "step": 20695
+    },
+    {
+      "epoch": 0.8377175232699312,
+      "grad_norm": 1.1376574039459229,
+      "learning_rate": 1.655998348131324e-05,
+      "loss": 1.0875,
+      "step": 20700
+    },
+    {
+      "epoch": 0.8379198704977742,
+      "grad_norm": 1.1920067071914673,
+      "learning_rate": 1.6539335122857733e-05,
+      "loss": 0.9851,
+      "step": 20705
+    },
+    {
+      "epoch": 0.8381222177256171,
+      "grad_norm": 1.170380711555481,
+      "learning_rate": 1.651868676440223e-05,
+      "loss": 0.9402,
+      "step": 20710
+    },
+    {
+      "epoch": 0.8383245649534601,
+      "grad_norm": 1.1304261684417725,
+      "learning_rate": 1.6498038405946728e-05,
+      "loss": 0.9497,
+      "step": 20715
+    },
+    {
+      "epoch": 0.8385269121813032,
+      "grad_norm": 1.155203938484192,
+      "learning_rate": 1.6477390047491225e-05,
+      "loss": 0.9433,
+      "step": 20720
+    },
+    {
+      "epoch": 0.8387292594091461,
+      "grad_norm": 1.1753532886505127,
+      "learning_rate": 1.6456741689035722e-05,
+      "loss": 1.0132,
+      "step": 20725
+    },
+    {
+      "epoch": 0.8389316066369891,
+      "grad_norm": 1.2378957271575928,
+      "learning_rate": 1.643609333058022e-05,
+      "loss": 1.0111,
+      "step": 20730
+    },
+    {
+      "epoch": 0.839133953864832,
+      "grad_norm": 1.0762790441513062,
+      "learning_rate": 1.6415444972124717e-05,
+      "loss": 1.0016,
+      "step": 20735
+    },
+    {
+      "epoch": 0.839336301092675,
+      "grad_norm": 1.2411805391311646,
+      "learning_rate": 1.6394796613669214e-05,
+      "loss": 1.0326,
+      "step": 20740
+    },
+    {
+      "epoch": 0.839538648320518,
+      "grad_norm": 1.2038969993591309,
+      "learning_rate": 1.637414825521371e-05,
+      "loss": 1.0516,
+      "step": 20745
+    },
+    {
+      "epoch": 0.839740995548361,
+      "grad_norm": 1.1077581644058228,
+      "learning_rate": 1.635349989675821e-05,
+      "loss": 1.0015,
+      "step": 20750
+    },
+    {
+      "epoch": 0.839943342776204,
+      "grad_norm": 1.1759651899337769,
+      "learning_rate": 1.6332851538302706e-05,
+      "loss": 1.0217,
+      "step": 20755
+    },
+    {
+      "epoch": 0.8401456900040469,
+      "grad_norm": 1.138208270072937,
+      "learning_rate": 1.6312203179847203e-05,
+      "loss": 1.0095,
+      "step": 20760
+    },
+    {
+      "epoch": 0.8403480372318899,
+      "grad_norm": 1.1641725301742554,
+      "learning_rate": 1.62915548213917e-05,
+      "loss": 1.0087,
+      "step": 20765
+    },
+    {
+      "epoch": 0.8405503844597328,
+      "grad_norm": 1.243086814880371,
+      "learning_rate": 1.6270906462936198e-05,
+      "loss": 0.9965,
+      "step": 20770
+    },
+    {
+      "epoch": 0.8407527316875759,
+      "grad_norm": 1.2246869802474976,
+      "learning_rate": 1.6250258104480695e-05,
+      "loss": 1.0346,
+      "step": 20775
+    },
+    {
+      "epoch": 0.8409550789154189,
+      "grad_norm": 1.2264448404312134,
+      "learning_rate": 1.6229609746025192e-05,
+      "loss": 0.9911,
+      "step": 20780
+    },
+    {
+      "epoch": 0.8411574261432618,
+      "grad_norm": 1.58490788936615,
+      "learning_rate": 1.6208961387569686e-05,
+      "loss": 0.9854,
+      "step": 20785
+    },
+    {
+      "epoch": 0.8413597733711048,
+      "grad_norm": 1.1928730010986328,
+      "learning_rate": 1.6188313029114187e-05,
+      "loss": 1.0452,
+      "step": 20790
+    },
+    {
+      "epoch": 0.8415621205989477,
+      "grad_norm": 1.259055495262146,
+      "learning_rate": 1.6167664670658684e-05,
+      "loss": 0.9524,
+      "step": 20795
+    },
+    {
+      "epoch": 0.8417644678267908,
+      "grad_norm": 1.2072827816009521,
+      "learning_rate": 1.614701631220318e-05,
+      "loss": 0.9834,
+      "step": 20800
+    },
+    {
+      "epoch": 0.8419668150546338,
+      "grad_norm": 1.2875561714172363,
+      "learning_rate": 1.612636795374768e-05,
+      "loss": 0.9549,
+      "step": 20805
+    },
+    {
+      "epoch": 0.8421691622824767,
+      "grad_norm": 1.1406989097595215,
+      "learning_rate": 1.6105719595292176e-05,
+      "loss": 0.966,
+      "step": 20810
+    },
+    {
+      "epoch": 0.8423715095103197,
+      "grad_norm": 1.243126630783081,
+      "learning_rate": 1.608507123683667e-05,
+      "loss": 0.9518,
+      "step": 20815
+    },
+    {
+      "epoch": 0.8425738567381627,
+      "grad_norm": 1.1952537298202515,
+      "learning_rate": 1.6064422878381167e-05,
+      "loss": 1.0327,
+      "step": 20820
+    },
+    {
+      "epoch": 0.8427762039660056,
+      "grad_norm": 1.1402138471603394,
+      "learning_rate": 1.6043774519925668e-05,
+      "loss": 1.0137,
+      "step": 20825
+    },
+    {
+      "epoch": 0.8429785511938487,
+      "grad_norm": 1.1675366163253784,
+      "learning_rate": 1.6023126161470165e-05,
+      "loss": 0.9981,
+      "step": 20830
+    },
+    {
+      "epoch": 0.8431808984216916,
+      "grad_norm": 1.215316891670227,
+      "learning_rate": 1.6002477803014663e-05,
+      "loss": 1.0721,
+      "step": 20835
+    },
+    {
+      "epoch": 0.8433832456495346,
+      "grad_norm": 1.1594030857086182,
+      "learning_rate": 1.5981829444559157e-05,
+      "loss": 1.1031,
+      "step": 20840
+    },
+    {
+      "epoch": 0.8435855928773776,
+      "grad_norm": 1.2180711030960083,
+      "learning_rate": 1.5961181086103654e-05,
+      "loss": 0.9981,
+      "step": 20845
+    },
+    {
+      "epoch": 0.8437879401052205,
+      "grad_norm": 1.2870091199874878,
+      "learning_rate": 1.594053272764815e-05,
+      "loss": 0.9359,
+      "step": 20850
+    },
+    {
+      "epoch": 0.8439902873330636,
+      "grad_norm": 1.3216265439987183,
+      "learning_rate": 1.5919884369192652e-05,
+      "loss": 0.9999,
+      "step": 20855
+    },
+    {
+      "epoch": 0.8441926345609065,
+      "grad_norm": 1.2268571853637695,
+      "learning_rate": 1.589923601073715e-05,
+      "loss": 0.9365,
+      "step": 20860
+    },
+    {
+      "epoch": 0.8443949817887495,
+      "grad_norm": 1.215065836906433,
+      "learning_rate": 1.5878587652281646e-05,
+      "loss": 1.018,
+      "step": 20865
+    },
+    {
+      "epoch": 0.8445973290165925,
+      "grad_norm": 1.2267355918884277,
+      "learning_rate": 1.585793929382614e-05,
+      "loss": 0.9637,
+      "step": 20870
+    },
+    {
+      "epoch": 0.8447996762444354,
+      "grad_norm": 1.1260260343551636,
+      "learning_rate": 1.5837290935370638e-05,
+      "loss": 0.971,
+      "step": 20875
+    },
+    {
+      "epoch": 0.8450020234722785,
+      "grad_norm": 1.1754131317138672,
+      "learning_rate": 1.5816642576915135e-05,
+      "loss": 1.0398,
+      "step": 20880
+    },
+    {
+      "epoch": 0.8452043707001214,
+      "grad_norm": 1.0897783041000366,
+      "learning_rate": 1.5795994218459632e-05,
+      "loss": 0.9756,
+      "step": 20885
+    },
+    {
+      "epoch": 0.8454067179279644,
+      "grad_norm": 1.2261866331100464,
+      "learning_rate": 1.5775345860004133e-05,
+      "loss": 0.9294,
+      "step": 20890
+    },
+    {
+      "epoch": 0.8456090651558074,
+      "grad_norm": 1.1631525754928589,
+      "learning_rate": 1.5754697501548627e-05,
+      "loss": 1.0199,
+      "step": 20895
+    },
+    {
+      "epoch": 0.8458114123836503,
+      "grad_norm": 1.2241733074188232,
+      "learning_rate": 1.5734049143093124e-05,
+      "loss": 0.9786,
+      "step": 20900
+    },
+    {
+      "epoch": 0.8460137596114933,
+      "grad_norm": 1.2011727094650269,
+      "learning_rate": 1.571340078463762e-05,
+      "loss": 0.9845,
+      "step": 20905
+    },
+    {
+      "epoch": 0.8462161068393363,
+      "grad_norm": 1.3325366973876953,
+      "learning_rate": 1.569275242618212e-05,
+      "loss": 0.9528,
+      "step": 20910
+    },
+    {
+      "epoch": 0.8464184540671793,
+      "grad_norm": 1.1038151979446411,
+      "learning_rate": 1.5672104067726616e-05,
+      "loss": 0.9831,
+      "step": 20915
+    },
+    {
+      "epoch": 0.8466208012950223,
+      "grad_norm": 1.2316639423370361,
+      "learning_rate": 1.5651455709271113e-05,
+      "loss": 0.9599,
+      "step": 20920
+    },
+    {
+      "epoch": 0.8468231485228652,
+      "grad_norm": 1.2244083881378174,
+      "learning_rate": 1.563080735081561e-05,
+      "loss": 0.9444,
+      "step": 20925
+    },
+    {
+      "epoch": 0.8470254957507082,
+      "grad_norm": 1.1958881616592407,
+      "learning_rate": 1.5610158992360108e-05,
+      "loss": 0.9704,
+      "step": 20930
+    },
+    {
+      "epoch": 0.8472278429785512,
+      "grad_norm": 1.1687911748886108,
+      "learning_rate": 1.5589510633904605e-05,
+      "loss": 0.9594,
+      "step": 20935
+    },
+    {
+      "epoch": 0.8474301902063942,
+      "grad_norm": 1.2375454902648926,
+      "learning_rate": 1.5568862275449103e-05,
+      "loss": 0.9012,
+      "step": 20940
+    },
+    {
+      "epoch": 0.8476325374342372,
+      "grad_norm": 1.1065729856491089,
+      "learning_rate": 1.55482139169936e-05,
+      "loss": 0.9231,
+      "step": 20945
+    },
+    {
+      "epoch": 0.8478348846620801,
+      "grad_norm": 1.2403284311294556,
+      "learning_rate": 1.5527565558538097e-05,
+      "loss": 0.9543,
+      "step": 20950
+    },
+    {
+      "epoch": 0.8480372318899231,
+      "grad_norm": 1.197769284248352,
+      "learning_rate": 1.5506917200082594e-05,
+      "loss": 0.9887,
+      "step": 20955
+    },
+    {
+      "epoch": 0.848239579117766,
+      "grad_norm": 1.2262988090515137,
+      "learning_rate": 1.5486268841627092e-05,
+      "loss": 1.0221,
+      "step": 20960
+    },
+    {
+      "epoch": 0.8484419263456091,
+      "grad_norm": 1.2083128690719604,
+      "learning_rate": 1.546562048317159e-05,
+      "loss": 1.0031,
+      "step": 20965
+    },
+    {
+      "epoch": 0.8486442735734521,
+      "grad_norm": 1.160172462463379,
+      "learning_rate": 1.5444972124716086e-05,
+      "loss": 0.9793,
+      "step": 20970
+    },
+    {
+      "epoch": 0.848846620801295,
+      "grad_norm": 1.2570322751998901,
+      "learning_rate": 1.5424323766260584e-05,
+      "loss": 0.9949,
+      "step": 20975
+    },
+    {
+      "epoch": 0.849048968029138,
+      "grad_norm": 1.1473708152770996,
+      "learning_rate": 1.540367540780508e-05,
+      "loss": 0.9765,
+      "step": 20980
+    },
+    {
+      "epoch": 0.8492513152569809,
+      "grad_norm": 1.172593593597412,
+      "learning_rate": 1.5383027049349578e-05,
+      "loss": 0.954,
+      "step": 20985
+    },
+    {
+      "epoch": 0.849453662484824,
+      "grad_norm": 1.0129601955413818,
+      "learning_rate": 1.5362378690894076e-05,
+      "loss": 0.9232,
+      "step": 20990
+    },
+    {
+      "epoch": 0.849656009712667,
+      "grad_norm": 1.2935516834259033,
+      "learning_rate": 1.5341730332438573e-05,
+      "loss": 1.0266,
+      "step": 20995
+    },
+    {
+      "epoch": 0.8498583569405099,
+      "grad_norm": 1.2592288255691528,
+      "learning_rate": 1.532108197398307e-05,
+      "loss": 0.9743,
+      "step": 21000
+    },
+    {
+      "epoch": 0.8500607041683529,
+      "grad_norm": 1.1387940645217896,
+      "learning_rate": 1.5300433615527564e-05,
+      "loss": 0.9626,
+      "step": 21005
+    },
+    {
+      "epoch": 0.8502630513961958,
+      "grad_norm": 1.2443729639053345,
+      "learning_rate": 1.527978525707206e-05,
+      "loss": 1.0068,
+      "step": 21010
+    },
+    {
+      "epoch": 0.8504653986240388,
+      "grad_norm": 1.1891158819198608,
+      "learning_rate": 1.5259136898616562e-05,
+      "loss": 0.9737,
+      "step": 21015
+    },
+    {
+      "epoch": 0.8506677458518819,
+      "grad_norm": 1.2353068590164185,
+      "learning_rate": 1.5238488540161058e-05,
+      "loss": 1.0029,
+      "step": 21020
+    },
+    {
+      "epoch": 0.8508700930797248,
+      "grad_norm": 1.1497776508331299,
+      "learning_rate": 1.5217840181705557e-05,
+      "loss": 1.048,
+      "step": 21025
+    },
+    {
+      "epoch": 0.8510724403075678,
+      "grad_norm": 1.2351295948028564,
+      "learning_rate": 1.519719182325005e-05,
+      "loss": 0.9483,
+      "step": 21030
+    },
+    {
+      "epoch": 0.8512747875354107,
+      "grad_norm": 1.2149288654327393,
+      "learning_rate": 1.517654346479455e-05,
+      "loss": 0.9573,
+      "step": 21035
+    },
+    {
+      "epoch": 0.8514771347632537,
+      "grad_norm": 1.1933867931365967,
+      "learning_rate": 1.5155895106339047e-05,
+      "loss": 0.9847,
+      "step": 21040
+    },
+    {
+      "epoch": 0.8516794819910968,
+      "grad_norm": 1.0731860399246216,
+      "learning_rate": 1.5135246747883544e-05,
+      "loss": 0.9594,
+      "step": 21045
+    },
+    {
+      "epoch": 0.8518818292189397,
+      "grad_norm": 1.1704195737838745,
+      "learning_rate": 1.5114598389428041e-05,
+      "loss": 1.0353,
+      "step": 21050
+    },
+    {
+      "epoch": 0.8520841764467827,
+      "grad_norm": 1.1087898015975952,
+      "learning_rate": 1.509395003097254e-05,
+      "loss": 0.9491,
+      "step": 21055
+    },
+    {
+      "epoch": 0.8522865236746257,
+      "grad_norm": 1.2999986410140991,
+      "learning_rate": 1.5073301672517034e-05,
+      "loss": 0.9467,
+      "step": 21060
+    },
+    {
+      "epoch": 0.8524888709024686,
+      "grad_norm": 1.2045687437057495,
+      "learning_rate": 1.5052653314061532e-05,
+      "loss": 0.976,
+      "step": 21065
+    },
+    {
+      "epoch": 0.8526912181303116,
+      "grad_norm": 1.1971813440322876,
+      "learning_rate": 1.503200495560603e-05,
+      "loss": 1.0275,
+      "step": 21070
+    },
+    {
+      "epoch": 0.8528935653581546,
+      "grad_norm": 1.2469074726104736,
+      "learning_rate": 1.5011356597150528e-05,
+      "loss": 0.9444,
+      "step": 21075
+    },
+    {
+      "epoch": 0.8530959125859976,
+      "grad_norm": 1.1796278953552246,
+      "learning_rate": 1.4990708238695025e-05,
+      "loss": 1.0424,
+      "step": 21080
+    },
+    {
+      "epoch": 0.8532982598138406,
+      "grad_norm": 1.1762430667877197,
+      "learning_rate": 1.497005988023952e-05,
+      "loss": 1.0121,
+      "step": 21085
+    },
+    {
+      "epoch": 0.8535006070416835,
+      "grad_norm": 1.522106647491455,
+      "learning_rate": 1.4949411521784018e-05,
+      "loss": 0.9712,
+      "step": 21090
+    },
+    {
+      "epoch": 0.8537029542695265,
+      "grad_norm": 1.2373191118240356,
+      "learning_rate": 1.4928763163328515e-05,
+      "loss": 0.9827,
+      "step": 21095
+    },
+    {
+      "epoch": 0.8539053014973695,
+      "grad_norm": 1.1276428699493408,
+      "learning_rate": 1.4908114804873013e-05,
+      "loss": 1.001,
+      "step": 21100
+    },
+    {
+      "epoch": 0.8541076487252125,
+      "grad_norm": 1.234139084815979,
+      "learning_rate": 1.4887466446417512e-05,
+      "loss": 0.9759,
+      "step": 21105
+    },
+    {
+      "epoch": 0.8543099959530555,
+      "grad_norm": 1.157067894935608,
+      "learning_rate": 1.4866818087962009e-05,
+      "loss": 0.9726,
+      "step": 21110
+    },
+    {
+      "epoch": 0.8545123431808984,
+      "grad_norm": 1.2965073585510254,
+      "learning_rate": 1.4846169729506505e-05,
+      "loss": 0.969,
+      "step": 21115
+    },
+    {
+      "epoch": 0.8547146904087414,
+      "grad_norm": 1.2545359134674072,
+      "learning_rate": 1.4825521371051002e-05,
+      "loss": 0.9854,
+      "step": 21120
+    },
+    {
+      "epoch": 0.8549170376365843,
+      "grad_norm": 1.1478519439697266,
+      "learning_rate": 1.48048730125955e-05,
+      "loss": 0.9751,
+      "step": 21125
+    },
+    {
+      "epoch": 0.8551193848644274,
+      "grad_norm": 1.2274898290634155,
+      "learning_rate": 1.4784224654139996e-05,
+      "loss": 0.9848,
+      "step": 21130
+    },
+    {
+      "epoch": 0.8553217320922704,
+      "grad_norm": 1.1130954027175903,
+      "learning_rate": 1.4763576295684495e-05,
+      "loss": 0.9941,
+      "step": 21135
+    },
+    {
+      "epoch": 0.8555240793201133,
+      "grad_norm": 1.1170850992202759,
+      "learning_rate": 1.474292793722899e-05,
+      "loss": 1.0035,
+      "step": 21140
+    },
+    {
+      "epoch": 0.8557264265479563,
+      "grad_norm": 1.3653924465179443,
+      "learning_rate": 1.4722279578773487e-05,
+      "loss": 1.0266,
+      "step": 21145
+    },
+    {
+      "epoch": 0.8559287737757992,
+      "grad_norm": 1.193739414215088,
+      "learning_rate": 1.4701631220317986e-05,
+      "loss": 1.0277,
+      "step": 21150
+    },
+    {
+      "epoch": 0.8561311210036423,
+      "grad_norm": 1.2416847944259644,
+      "learning_rate": 1.4680982861862483e-05,
+      "loss": 1.0254,
+      "step": 21155
+    },
+    {
+      "epoch": 0.8563334682314853,
+      "grad_norm": 1.1676281690597534,
+      "learning_rate": 1.466033450340698e-05,
+      "loss": 0.9555,
+      "step": 21160
+    },
+    {
+      "epoch": 0.8565358154593282,
+      "grad_norm": 1.2324869632720947,
+      "learning_rate": 1.4639686144951478e-05,
+      "loss": 0.9764,
+      "step": 21165
+    },
+    {
+      "epoch": 0.8567381626871712,
+      "grad_norm": 1.2595820426940918,
+      "learning_rate": 1.4619037786495973e-05,
+      "loss": 1.0357,
+      "step": 21170
+    },
+    {
+      "epoch": 0.8569405099150141,
+      "grad_norm": 1.2351970672607422,
+      "learning_rate": 1.459838942804047e-05,
+      "loss": 1.0074,
+      "step": 21175
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 1.1445361375808716,
+      "learning_rate": 1.457774106958497e-05,
+      "loss": 0.9785,
+      "step": 21180
+    },
+    {
+      "epoch": 0.8573452043707002,
+      "grad_norm": 1.1912777423858643,
+      "learning_rate": 1.4557092711129467e-05,
+      "loss": 1.0694,
+      "step": 21185
+    },
+    {
+      "epoch": 0.8575475515985431,
+      "grad_norm": 1.3794888257980347,
+      "learning_rate": 1.4536444352673964e-05,
+      "loss": 1.008,
+      "step": 21190
+    },
+    {
+      "epoch": 0.8577498988263861,
+      "grad_norm": 1.210392713546753,
+      "learning_rate": 1.451579599421846e-05,
+      "loss": 1.0009,
+      "step": 21195
+    },
+    {
+      "epoch": 0.857952246054229,
+      "grad_norm": 1.1218934059143066,
+      "learning_rate": 1.4495147635762957e-05,
+      "loss": 0.9867,
+      "step": 21200
+    },
+    {
+      "epoch": 0.858154593282072,
+      "grad_norm": 1.0614824295043945,
+      "learning_rate": 1.4474499277307454e-05,
+      "loss": 0.872,
+      "step": 21205
+    },
+    {
+      "epoch": 0.8583569405099151,
+      "grad_norm": 1.2585033178329468,
+      "learning_rate": 1.4453850918851952e-05,
+      "loss": 0.9982,
+      "step": 21210
+    },
+    {
+      "epoch": 0.858559287737758,
+      "grad_norm": 1.2803871631622314,
+      "learning_rate": 1.443320256039645e-05,
+      "loss": 0.9609,
+      "step": 21215
+    },
+    {
+      "epoch": 0.858761634965601,
+      "grad_norm": 1.095826268196106,
+      "learning_rate": 1.4412554201940948e-05,
+      "loss": 1.0214,
+      "step": 21220
+    },
+    {
+      "epoch": 0.8589639821934439,
+      "grad_norm": 1.105059027671814,
+      "learning_rate": 1.4391905843485443e-05,
+      "loss": 0.964,
+      "step": 21225
+    },
+    {
+      "epoch": 0.8591663294212869,
+      "grad_norm": 1.4007976055145264,
+      "learning_rate": 1.437125748502994e-05,
+      "loss": 0.9648,
+      "step": 21230
+    },
+    {
+      "epoch": 0.8593686766491299,
+      "grad_norm": 1.1132739782333374,
+      "learning_rate": 1.4350609126574438e-05,
+      "loss": 0.9983,
+      "step": 21235
+    },
+    {
+      "epoch": 0.8595710238769729,
+      "grad_norm": 1.2380614280700684,
+      "learning_rate": 1.4329960768118935e-05,
+      "loss": 0.987,
+      "step": 21240
+    },
+    {
+      "epoch": 0.8597733711048159,
+      "grad_norm": 1.3055659532546997,
+      "learning_rate": 1.4309312409663434e-05,
+      "loss": 1.0415,
+      "step": 21245
+    },
+    {
+      "epoch": 0.8599757183326588,
+      "grad_norm": 1.1960859298706055,
+      "learning_rate": 1.4288664051207928e-05,
+      "loss": 0.995,
+      "step": 21250
+    },
+    {
+      "epoch": 0.8601780655605018,
+      "grad_norm": 1.2132443189620972,
+      "learning_rate": 1.4268015692752425e-05,
+      "loss": 1.005,
+      "step": 21255
+    },
+    {
+      "epoch": 0.8603804127883448,
+      "grad_norm": 1.3091709613800049,
+      "learning_rate": 1.4247367334296924e-05,
+      "loss": 0.9947,
+      "step": 21260
+    },
+    {
+      "epoch": 0.8605827600161878,
+      "grad_norm": 1.2340102195739746,
+      "learning_rate": 1.4226718975841422e-05,
+      "loss": 0.9763,
+      "step": 21265
+    },
+    {
+      "epoch": 0.8607851072440308,
+      "grad_norm": 1.1634894609451294,
+      "learning_rate": 1.4206070617385919e-05,
+      "loss": 0.9707,
+      "step": 21270
+    },
+    {
+      "epoch": 0.8609874544718737,
+      "grad_norm": 1.237231731414795,
+      "learning_rate": 1.4185422258930416e-05,
+      "loss": 0.9842,
+      "step": 21275
+    },
+    {
+      "epoch": 0.8611898016997167,
+      "grad_norm": 1.1723443269729614,
+      "learning_rate": 1.4164773900474912e-05,
+      "loss": 0.9692,
+      "step": 21280
+    },
+    {
+      "epoch": 0.8613921489275597,
+      "grad_norm": 1.1599555015563965,
+      "learning_rate": 1.414412554201941e-05,
+      "loss": 0.9125,
+      "step": 21285
+    },
+    {
+      "epoch": 0.8615944961554026,
+      "grad_norm": 1.109898567199707,
+      "learning_rate": 1.4123477183563907e-05,
+      "loss": 0.9263,
+      "step": 21290
+    },
+    {
+      "epoch": 0.8617968433832457,
+      "grad_norm": 1.231009602546692,
+      "learning_rate": 1.4102828825108406e-05,
+      "loss": 0.9312,
+      "step": 21295
+    },
+    {
+      "epoch": 0.8619991906110887,
+      "grad_norm": 1.1627553701400757,
+      "learning_rate": 1.4082180466652903e-05,
+      "loss": 1.0614,
+      "step": 21300
+    },
+    {
+      "epoch": 0.8622015378389316,
+      "grad_norm": 1.1727001667022705,
+      "learning_rate": 1.4061532108197398e-05,
+      "loss": 1.0312,
+      "step": 21305
+    },
+    {
+      "epoch": 0.8624038850667746,
+      "grad_norm": 1.2657217979431152,
+      "learning_rate": 1.4040883749741896e-05,
+      "loss": 0.947,
+      "step": 21310
+    },
+    {
+      "epoch": 0.8626062322946175,
+      "grad_norm": 1.2353177070617676,
+      "learning_rate": 1.4020235391286393e-05,
+      "loss": 0.9976,
+      "step": 21315
+    },
+    {
+      "epoch": 0.8628085795224606,
+      "grad_norm": 1.23910653591156,
+      "learning_rate": 1.399958703283089e-05,
+      "loss": 0.9671,
+      "step": 21320
+    },
+    {
+      "epoch": 0.8630109267503036,
+      "grad_norm": 1.2736573219299316,
+      "learning_rate": 1.397893867437539e-05,
+      "loss": 1.0209,
+      "step": 21325
+    },
+    {
+      "epoch": 0.8632132739781465,
+      "grad_norm": 1.2034873962402344,
+      "learning_rate": 1.3958290315919887e-05,
+      "loss": 1.0165,
+      "step": 21330
+    },
+    {
+      "epoch": 0.8634156212059895,
+      "grad_norm": 1.2223907709121704,
+      "learning_rate": 1.393764195746438e-05,
+      "loss": 0.9769,
+      "step": 21335
+    },
+    {
+      "epoch": 0.8636179684338324,
+      "grad_norm": 1.176323652267456,
+      "learning_rate": 1.391699359900888e-05,
+      "loss": 1.0038,
+      "step": 21340
+    },
+    {
+      "epoch": 0.8638203156616754,
+      "grad_norm": 1.2671642303466797,
+      "learning_rate": 1.3896345240553377e-05,
+      "loss": 1.0339,
+      "step": 21345
+    },
+    {
+      "epoch": 0.8640226628895185,
+      "grad_norm": 1.3125969171524048,
+      "learning_rate": 1.3875696882097874e-05,
+      "loss": 0.9145,
+      "step": 21350
+    },
+    {
+      "epoch": 0.8642250101173614,
+      "grad_norm": 1.3143863677978516,
+      "learning_rate": 1.3855048523642371e-05,
+      "loss": 1.0101,
+      "step": 21355
+    },
+    {
+      "epoch": 0.8644273573452044,
+      "grad_norm": 1.1887516975402832,
+      "learning_rate": 1.3834400165186867e-05,
+      "loss": 0.9995,
+      "step": 21360
+    },
+    {
+      "epoch": 0.8646297045730473,
+      "grad_norm": 1.230217456817627,
+      "learning_rate": 1.3813751806731364e-05,
+      "loss": 1.0089,
+      "step": 21365
+    },
+    {
+      "epoch": 0.8648320518008903,
+      "grad_norm": 1.3523839712142944,
+      "learning_rate": 1.3793103448275863e-05,
+      "loss": 0.9322,
+      "step": 21370
+    },
+    {
+      "epoch": 0.8650343990287334,
+      "grad_norm": 1.1450783014297485,
+      "learning_rate": 1.377245508982036e-05,
+      "loss": 0.9545,
+      "step": 21375
+    },
+    {
+      "epoch": 0.8652367462565763,
+      "grad_norm": 1.2504395246505737,
+      "learning_rate": 1.3751806731364858e-05,
+      "loss": 0.9773,
+      "step": 21380
+    },
+    {
+      "epoch": 0.8654390934844193,
+      "grad_norm": 1.2486871480941772,
+      "learning_rate": 1.3731158372909355e-05,
+      "loss": 0.9868,
+      "step": 21385
+    },
+    {
+      "epoch": 0.8656414407122622,
+      "grad_norm": 1.1254518032073975,
+      "learning_rate": 1.371051001445385e-05,
+      "loss": 1.024,
+      "step": 21390
+    },
+    {
+      "epoch": 0.8658437879401052,
+      "grad_norm": 1.1728692054748535,
+      "learning_rate": 1.3689861655998348e-05,
+      "loss": 0.9564,
+      "step": 21395
+    },
+    {
+      "epoch": 0.8660461351679482,
+      "grad_norm": 1.2263543605804443,
+      "learning_rate": 1.3669213297542845e-05,
+      "loss": 0.9728,
+      "step": 21400
+    },
+    {
+      "epoch": 0.8662484823957912,
+      "grad_norm": 1.131757140159607,
+      "learning_rate": 1.3648564939087344e-05,
+      "loss": 0.9845,
+      "step": 21405
+    },
+    {
+      "epoch": 0.8664508296236342,
+      "grad_norm": 1.223541021347046,
+      "learning_rate": 1.3627916580631842e-05,
+      "loss": 0.9843,
+      "step": 21410
+    },
+    {
+      "epoch": 0.8666531768514771,
+      "grad_norm": 1.2180362939834595,
+      "learning_rate": 1.3607268222176337e-05,
+      "loss": 1.0511,
+      "step": 21415
+    },
+    {
+      "epoch": 0.8668555240793201,
+      "grad_norm": 1.2126442193984985,
+      "learning_rate": 1.3586619863720835e-05,
+      "loss": 0.964,
+      "step": 21420
+    },
+    {
+      "epoch": 0.867057871307163,
+      "grad_norm": 1.350108027458191,
+      "learning_rate": 1.3565971505265332e-05,
+      "loss": 1.0417,
+      "step": 21425
+    },
+    {
+      "epoch": 0.8672602185350061,
+      "grad_norm": 1.229135274887085,
+      "learning_rate": 1.354532314680983e-05,
+      "loss": 1.0019,
+      "step": 21430
+    },
+    {
+      "epoch": 0.8674625657628491,
+      "grad_norm": 1.244516372680664,
+      "learning_rate": 1.3524674788354328e-05,
+      "loss": 0.9827,
+      "step": 21435
+    },
+    {
+      "epoch": 0.867664912990692,
+      "grad_norm": 1.2456797361373901,
+      "learning_rate": 1.3504026429898825e-05,
+      "loss": 0.9963,
+      "step": 21440
+    },
+    {
+      "epoch": 0.867867260218535,
+      "grad_norm": 1.1829853057861328,
+      "learning_rate": 1.348337807144332e-05,
+      "loss": 1.0354,
+      "step": 21445
+    },
+    {
+      "epoch": 0.868069607446378,
+      "grad_norm": 1.233486533164978,
+      "learning_rate": 1.3462729712987818e-05,
+      "loss": 0.964,
+      "step": 21450
+    },
+    {
+      "epoch": 0.8682719546742209,
+      "grad_norm": 1.2854801416397095,
+      "learning_rate": 1.3442081354532316e-05,
+      "loss": 0.9566,
+      "step": 21455
+    },
+    {
+      "epoch": 0.868474301902064,
+      "grad_norm": 1.2866089344024658,
+      "learning_rate": 1.3421432996076813e-05,
+      "loss": 0.9384,
+      "step": 21460
+    },
+    {
+      "epoch": 0.8686766491299069,
+      "grad_norm": 1.1400288343429565,
+      "learning_rate": 1.340078463762131e-05,
+      "loss": 1.0305,
+      "step": 21465
+    },
+    {
+      "epoch": 0.8688789963577499,
+      "grad_norm": 1.1516262292861938,
+      "learning_rate": 1.3380136279165806e-05,
+      "loss": 1.0067,
+      "step": 21470
+    },
+    {
+      "epoch": 0.8690813435855929,
+      "grad_norm": 1.212130069732666,
+      "learning_rate": 1.3359487920710303e-05,
+      "loss": 0.9806,
+      "step": 21475
+    },
+    {
+      "epoch": 0.8692836908134358,
+      "grad_norm": 1.2694216966629028,
+      "learning_rate": 1.33388395622548e-05,
+      "loss": 1.0292,
+      "step": 21480
+    },
+    {
+      "epoch": 0.8694860380412789,
+      "grad_norm": 1.2387278079986572,
+      "learning_rate": 1.33181912037993e-05,
+      "loss": 0.9744,
+      "step": 21485
+    },
+    {
+      "epoch": 0.8696883852691218,
+      "grad_norm": 1.2714673280715942,
+      "learning_rate": 1.3297542845343797e-05,
+      "loss": 1.0123,
+      "step": 21490
+    },
+    {
+      "epoch": 0.8698907324969648,
+      "grad_norm": 1.1535682678222656,
+      "learning_rate": 1.3276894486888294e-05,
+      "loss": 0.9502,
+      "step": 21495
+    },
+    {
+      "epoch": 0.8700930797248078,
+      "grad_norm": 1.1316004991531372,
+      "learning_rate": 1.325624612843279e-05,
+      "loss": 0.9473,
+      "step": 21500
+    },
+    {
+      "epoch": 0.8702954269526507,
+      "grad_norm": 1.2753766775131226,
+      "learning_rate": 1.3235597769977287e-05,
+      "loss": 1.0167,
+      "step": 21505
+    },
+    {
+      "epoch": 0.8704977741804937,
+      "grad_norm": 1.2434911727905273,
+      "learning_rate": 1.3214949411521784e-05,
+      "loss": 1.0,
+      "step": 21510
+    },
+    {
+      "epoch": 0.8707001214083367,
+      "grad_norm": 1.237639307975769,
+      "learning_rate": 1.3194301053066283e-05,
+      "loss": 0.9892,
+      "step": 21515
+    },
+    {
+      "epoch": 0.8709024686361797,
+      "grad_norm": 1.1442434787750244,
+      "learning_rate": 1.317365269461078e-05,
+      "loss": 0.968,
+      "step": 21520
+    },
+    {
+      "epoch": 0.8711048158640227,
+      "grad_norm": 1.21522057056427,
+      "learning_rate": 1.3153004336155274e-05,
+      "loss": 1.043,
+      "step": 21525
+    },
+    {
+      "epoch": 0.8713071630918656,
+      "grad_norm": 1.2270630598068237,
+      "learning_rate": 1.3132355977699773e-05,
+      "loss": 0.9797,
+      "step": 21530
+    },
+    {
+      "epoch": 0.8715095103197086,
+      "grad_norm": 1.1119053363800049,
+      "learning_rate": 1.311170761924427e-05,
+      "loss": 0.8997,
+      "step": 21535
+    },
+    {
+      "epoch": 0.8717118575475516,
+      "grad_norm": 1.2777276039123535,
+      "learning_rate": 1.3091059260788768e-05,
+      "loss": 1.0123,
+      "step": 21540
+    },
+    {
+      "epoch": 0.8719142047753946,
+      "grad_norm": 1.17526113986969,
+      "learning_rate": 1.3070410902333265e-05,
+      "loss": 1.0459,
+      "step": 21545
+    },
+    {
+      "epoch": 0.8721165520032376,
+      "grad_norm": 1.2038923501968384,
+      "learning_rate": 1.3049762543877761e-05,
+      "loss": 1.0099,
+      "step": 21550
+    },
+    {
+      "epoch": 0.8723188992310805,
+      "grad_norm": 1.293413519859314,
+      "learning_rate": 1.3029114185422258e-05,
+      "loss": 1.0036,
+      "step": 21555
+    },
+    {
+      "epoch": 0.8725212464589235,
+      "grad_norm": 1.163986325263977,
+      "learning_rate": 1.3008465826966757e-05,
+      "loss": 0.9912,
+      "step": 21560
+    },
+    {
+      "epoch": 0.8727235936867664,
+      "grad_norm": 1.279579520225525,
+      "learning_rate": 1.2987817468511255e-05,
+      "loss": 1.0417,
+      "step": 21565
+    },
+    {
+      "epoch": 0.8729259409146095,
+      "grad_norm": 1.1215415000915527,
+      "learning_rate": 1.2967169110055752e-05,
+      "loss": 0.9581,
+      "step": 21570
+    },
+    {
+      "epoch": 0.8731282881424525,
+      "grad_norm": 1.1366301774978638,
+      "learning_rate": 1.2946520751600249e-05,
+      "loss": 0.9994,
+      "step": 21575
+    },
+    {
+      "epoch": 0.8733306353702954,
+      "grad_norm": 1.1679660081863403,
+      "learning_rate": 1.2925872393144745e-05,
+      "loss": 0.9659,
+      "step": 21580
+    },
+    {
+      "epoch": 0.8735329825981384,
+      "grad_norm": 1.1928454637527466,
+      "learning_rate": 1.2905224034689242e-05,
+      "loss": 1.0112,
+      "step": 21585
+    },
+    {
+      "epoch": 0.8737353298259813,
+      "grad_norm": 1.2413065433502197,
+      "learning_rate": 1.288457567623374e-05,
+      "loss": 0.9623,
+      "step": 21590
+    },
+    {
+      "epoch": 0.8739376770538244,
+      "grad_norm": 1.2356163263320923,
+      "learning_rate": 1.2863927317778238e-05,
+      "loss": 1.0146,
+      "step": 21595
+    },
+    {
+      "epoch": 0.8741400242816674,
+      "grad_norm": 1.1208093166351318,
+      "learning_rate": 1.2843278959322736e-05,
+      "loss": 0.9413,
+      "step": 21600
+    },
+    {
+      "epoch": 0.8743423715095103,
+      "grad_norm": 1.1954208612442017,
+      "learning_rate": 1.2822630600867231e-05,
+      "loss": 1.0004,
+      "step": 21605
+    },
+    {
+      "epoch": 0.8745447187373533,
+      "grad_norm": 1.0967018604278564,
+      "learning_rate": 1.2801982242411728e-05,
+      "loss": 0.9661,
+      "step": 21610
+    },
+    {
+      "epoch": 0.8747470659651962,
+      "grad_norm": 1.3068009614944458,
+      "learning_rate": 1.2781333883956226e-05,
+      "loss": 0.9971,
+      "step": 21615
+    },
+    {
+      "epoch": 0.8749494131930392,
+      "grad_norm": 1.2453967332839966,
+      "learning_rate": 1.2760685525500723e-05,
+      "loss": 0.9947,
+      "step": 21620
+    },
+    {
+      "epoch": 0.8751517604208823,
+      "grad_norm": 1.2303450107574463,
+      "learning_rate": 1.274003716704522e-05,
+      "loss": 0.9848,
+      "step": 21625
+    },
+    {
+      "epoch": 0.8753541076487252,
+      "grad_norm": 1.1697107553482056,
+      "learning_rate": 1.271938880858972e-05,
+      "loss": 0.9864,
+      "step": 21630
+    },
+    {
+      "epoch": 0.8755564548765682,
+      "grad_norm": 1.1332446336746216,
+      "learning_rate": 1.2698740450134213e-05,
+      "loss": 0.9674,
+      "step": 21635
+    },
+    {
+      "epoch": 0.8757588021044111,
+      "grad_norm": 1.082084059715271,
+      "learning_rate": 1.2678092091678712e-05,
+      "loss": 0.9757,
+      "step": 21640
+    },
+    {
+      "epoch": 0.8759611493322541,
+      "grad_norm": 1.266560673713684,
+      "learning_rate": 1.265744373322321e-05,
+      "loss": 0.9536,
+      "step": 21645
+    },
+    {
+      "epoch": 0.8761634965600972,
+      "grad_norm": 1.2842439413070679,
+      "learning_rate": 1.2636795374767707e-05,
+      "loss": 1.0134,
+      "step": 21650
+    },
+    {
+      "epoch": 0.8763658437879401,
+      "grad_norm": 1.2898311614990234,
+      "learning_rate": 1.2616147016312204e-05,
+      "loss": 1.0112,
+      "step": 21655
+    },
+    {
+      "epoch": 0.8765681910157831,
+      "grad_norm": 1.2124890089035034,
+      "learning_rate": 1.25954986578567e-05,
+      "loss": 1.0626,
+      "step": 21660
+    },
+    {
+      "epoch": 0.876770538243626,
+      "grad_norm": 1.126383900642395,
+      "learning_rate": 1.2574850299401197e-05,
+      "loss": 0.9958,
+      "step": 21665
+    },
+    {
+      "epoch": 0.876972885471469,
+      "grad_norm": 1.1929011344909668,
+      "learning_rate": 1.2554201940945694e-05,
+      "loss": 1.0046,
+      "step": 21670
+    },
+    {
+      "epoch": 0.877175232699312,
+      "grad_norm": 1.068507194519043,
+      "learning_rate": 1.2533553582490193e-05,
+      "loss": 0.9997,
+      "step": 21675
+    },
+    {
+      "epoch": 0.877377579927155,
+      "grad_norm": 1.112916111946106,
+      "learning_rate": 1.251290522403469e-05,
+      "loss": 1.0405,
+      "step": 21680
+    },
+    {
+      "epoch": 0.877579927154998,
+      "grad_norm": 1.1991050243377686,
+      "learning_rate": 1.2492256865579186e-05,
+      "loss": 0.9949,
+      "step": 21685
+    },
+    {
+      "epoch": 0.877782274382841,
+      "grad_norm": 1.2340871095657349,
+      "learning_rate": 1.2471608507123685e-05,
+      "loss": 0.9546,
+      "step": 21690
+    },
+    {
+      "epoch": 0.8779846216106839,
+      "grad_norm": 1.2195053100585938,
+      "learning_rate": 1.245096014866818e-05,
+      "loss": 0.9761,
+      "step": 21695
+    },
+    {
+      "epoch": 0.8781869688385269,
+      "grad_norm": 1.2020443677902222,
+      "learning_rate": 1.2430311790212678e-05,
+      "loss": 0.9838,
+      "step": 21700
+    },
+    {
+      "epoch": 0.8783893160663699,
+      "grad_norm": 1.375296950340271,
+      "learning_rate": 1.2409663431757177e-05,
+      "loss": 0.9381,
+      "step": 21705
+    },
+    {
+      "epoch": 0.8785916632942129,
+      "grad_norm": 1.1228500604629517,
+      "learning_rate": 1.2389015073301673e-05,
+      "loss": 1.0171,
+      "step": 21710
+    },
+    {
+      "epoch": 0.8787940105220559,
+      "grad_norm": 1.1400810480117798,
+      "learning_rate": 1.236836671484617e-05,
+      "loss": 0.9277,
+      "step": 21715
+    },
+    {
+      "epoch": 0.8789963577498988,
+      "grad_norm": 1.3855520486831665,
+      "learning_rate": 1.2347718356390667e-05,
+      "loss": 0.9902,
+      "step": 21720
+    },
+    {
+      "epoch": 0.8791987049777418,
+      "grad_norm": 1.1449639797210693,
+      "learning_rate": 1.2327069997935165e-05,
+      "loss": 0.947,
+      "step": 21725
+    },
+    {
+      "epoch": 0.8794010522055847,
+      "grad_norm": 1.1496288776397705,
+      "learning_rate": 1.2306421639479662e-05,
+      "loss": 0.9917,
+      "step": 21730
+    },
+    {
+      "epoch": 0.8796033994334278,
+      "grad_norm": 1.1176990270614624,
+      "learning_rate": 1.228577328102416e-05,
+      "loss": 0.9002,
+      "step": 21735
+    },
+    {
+      "epoch": 0.8798057466612708,
+      "grad_norm": 1.286231279373169,
+      "learning_rate": 1.2265124922568657e-05,
+      "loss": 1.051,
+      "step": 21740
+    },
+    {
+      "epoch": 0.8800080938891137,
+      "grad_norm": 1.270430564880371,
+      "learning_rate": 1.2244476564113154e-05,
+      "loss": 1.0009,
+      "step": 21745
+    },
+    {
+      "epoch": 0.8802104411169567,
+      "grad_norm": 1.1725939512252808,
+      "learning_rate": 1.2223828205657651e-05,
+      "loss": 1.0042,
+      "step": 21750
+    },
+    {
+      "epoch": 0.8804127883447996,
+      "grad_norm": 1.2585035562515259,
+      "learning_rate": 1.2203179847202148e-05,
+      "loss": 1.0237,
+      "step": 21755
+    },
+    {
+      "epoch": 0.8806151355726427,
+      "grad_norm": 1.2797117233276367,
+      "learning_rate": 1.2182531488746646e-05,
+      "loss": 0.9502,
+      "step": 21760
+    },
+    {
+      "epoch": 0.8808174828004857,
+      "grad_norm": 1.1111143827438354,
+      "learning_rate": 1.2161883130291141e-05,
+      "loss": 0.9753,
+      "step": 21765
+    },
+    {
+      "epoch": 0.8810198300283286,
+      "grad_norm": 1.2845194339752197,
+      "learning_rate": 1.214123477183564e-05,
+      "loss": 0.9072,
+      "step": 21770
+    },
+    {
+      "epoch": 0.8812221772561716,
+      "grad_norm": 1.2153072357177734,
+      "learning_rate": 1.2120586413380138e-05,
+      "loss": 1.0065,
+      "step": 21775
+    },
+    {
+      "epoch": 0.8814245244840145,
+      "grad_norm": 1.1618738174438477,
+      "learning_rate": 1.2099938054924633e-05,
+      "loss": 1.0596,
+      "step": 21780
+    },
+    {
+      "epoch": 0.8816268717118575,
+      "grad_norm": 1.3450782299041748,
+      "learning_rate": 1.2079289696469132e-05,
+      "loss": 0.9763,
+      "step": 21785
+    },
+    {
+      "epoch": 0.8818292189397006,
+      "grad_norm": 1.1920745372772217,
+      "learning_rate": 1.2058641338013628e-05,
+      "loss": 1.0137,
+      "step": 21790
+    },
+    {
+      "epoch": 0.8820315661675435,
+      "grad_norm": 1.39998459815979,
+      "learning_rate": 1.2037992979558125e-05,
+      "loss": 0.9687,
+      "step": 21795
+    },
+    {
+      "epoch": 0.8822339133953865,
+      "grad_norm": 1.3136975765228271,
+      "learning_rate": 1.2017344621102624e-05,
+      "loss": 0.9613,
+      "step": 21800
+    },
+    {
+      "epoch": 0.8824362606232294,
+      "grad_norm": 1.1904999017715454,
+      "learning_rate": 1.199669626264712e-05,
+      "loss": 1.0179,
+      "step": 21805
+    },
+    {
+      "epoch": 0.8826386078510724,
+      "grad_norm": 1.0881669521331787,
+      "learning_rate": 1.1976047904191617e-05,
+      "loss": 0.9737,
+      "step": 21810
+    },
+    {
+      "epoch": 0.8828409550789155,
+      "grad_norm": 1.222800374031067,
+      "learning_rate": 1.1955399545736114e-05,
+      "loss": 1.0169,
+      "step": 21815
+    },
+    {
+      "epoch": 0.8830433023067584,
+      "grad_norm": 1.233533501625061,
+      "learning_rate": 1.1934751187280612e-05,
+      "loss": 1.0014,
+      "step": 21820
+    },
+    {
+      "epoch": 0.8832456495346014,
+      "grad_norm": 1.1993024349212646,
+      "learning_rate": 1.1914102828825109e-05,
+      "loss": 1.0282,
+      "step": 21825
+    },
+    {
+      "epoch": 0.8834479967624443,
+      "grad_norm": 1.3259748220443726,
+      "learning_rate": 1.1893454470369606e-05,
+      "loss": 1.0201,
+      "step": 21830
+    },
+    {
+      "epoch": 0.8836503439902873,
+      "grad_norm": 1.1832479238510132,
+      "learning_rate": 1.1872806111914103e-05,
+      "loss": 1.0084,
+      "step": 21835
+    },
+    {
+      "epoch": 0.8838526912181303,
+      "grad_norm": 1.4280952215194702,
+      "learning_rate": 1.18521577534586e-05,
+      "loss": 1.0573,
+      "step": 21840
+    },
+    {
+      "epoch": 0.8840550384459733,
+      "grad_norm": 1.2184512615203857,
+      "learning_rate": 1.1831509395003098e-05,
+      "loss": 0.9729,
+      "step": 21845
+    },
+    {
+      "epoch": 0.8842573856738163,
+      "grad_norm": 1.2103275060653687,
+      "learning_rate": 1.1810861036547595e-05,
+      "loss": 1.0354,
+      "step": 21850
+    },
+    {
+      "epoch": 0.8844597329016592,
+      "grad_norm": 1.17412269115448,
+      "learning_rate": 1.1790212678092093e-05,
+      "loss": 1.0041,
+      "step": 21855
+    },
+    {
+      "epoch": 0.8846620801295022,
+      "grad_norm": 1.2611116170883179,
+      "learning_rate": 1.1769564319636588e-05,
+      "loss": 1.0013,
+      "step": 21860
+    },
+    {
+      "epoch": 0.8848644273573452,
+      "grad_norm": 1.2616214752197266,
+      "learning_rate": 1.1748915961181087e-05,
+      "loss": 0.9902,
+      "step": 21865
+    },
+    {
+      "epoch": 0.8850667745851882,
+      "grad_norm": 1.2103784084320068,
+      "learning_rate": 1.1728267602725585e-05,
+      "loss": 1.0034,
+      "step": 21870
+    },
+    {
+      "epoch": 0.8852691218130312,
+      "grad_norm": 1.2653135061264038,
+      "learning_rate": 1.170761924427008e-05,
+      "loss": 0.9487,
+      "step": 21875
+    },
+    {
+      "epoch": 0.8854714690408741,
+      "grad_norm": 1.273098349571228,
+      "learning_rate": 1.1686970885814579e-05,
+      "loss": 0.9933,
+      "step": 21880
+    },
+    {
+      "epoch": 0.8856738162687171,
+      "grad_norm": 1.221599817276001,
+      "learning_rate": 1.1666322527359075e-05,
+      "loss": 1.0322,
+      "step": 21885
+    },
+    {
+      "epoch": 0.8858761634965601,
+      "grad_norm": 1.2934221029281616,
+      "learning_rate": 1.1645674168903572e-05,
+      "loss": 1.0064,
+      "step": 21890
+    },
+    {
+      "epoch": 0.886078510724403,
+      "grad_norm": 1.3288832902908325,
+      "learning_rate": 1.1625025810448071e-05,
+      "loss": 0.9758,
+      "step": 21895
+    },
+    {
+      "epoch": 0.8862808579522461,
+      "grad_norm": 1.1124895811080933,
+      "learning_rate": 1.1604377451992567e-05,
+      "loss": 0.9687,
+      "step": 21900
+    },
+    {
+      "epoch": 0.886483205180089,
+      "grad_norm": 1.308106780052185,
+      "learning_rate": 1.1583729093537064e-05,
+      "loss": 1.0038,
+      "step": 21905
+    },
+    {
+      "epoch": 0.886685552407932,
+      "grad_norm": 1.2233226299285889,
+      "learning_rate": 1.1563080735081561e-05,
+      "loss": 0.9868,
+      "step": 21910
+    },
+    {
+      "epoch": 0.886887899635775,
+      "grad_norm": 1.2253919839859009,
+      "learning_rate": 1.1542432376626059e-05,
+      "loss": 1.0083,
+      "step": 21915
+    },
+    {
+      "epoch": 0.8870902468636179,
+      "grad_norm": 1.287880778312683,
+      "learning_rate": 1.1521784018170556e-05,
+      "loss": 0.9998,
+      "step": 21920
+    },
+    {
+      "epoch": 0.887292594091461,
+      "grad_norm": 1.310227870941162,
+      "learning_rate": 1.1501135659715053e-05,
+      "loss": 1.0066,
+      "step": 21925
+    },
+    {
+      "epoch": 0.887494941319304,
+      "grad_norm": 1.285409688949585,
+      "learning_rate": 1.148048730125955e-05,
+      "loss": 0.9458,
+      "step": 21930
+    },
+    {
+      "epoch": 0.8876972885471469,
+      "grad_norm": 1.2010833024978638,
+      "learning_rate": 1.1459838942804048e-05,
+      "loss": 0.9599,
+      "step": 21935
+    },
+    {
+      "epoch": 0.8878996357749899,
+      "grad_norm": 1.2600255012512207,
+      "learning_rate": 1.1439190584348545e-05,
+      "loss": 1.0022,
+      "step": 21940
+    },
+    {
+      "epoch": 0.8881019830028328,
+      "grad_norm": 1.1470943689346313,
+      "learning_rate": 1.1418542225893042e-05,
+      "loss": 0.974,
+      "step": 21945
+    },
+    {
+      "epoch": 0.8883043302306758,
+      "grad_norm": 1.2417147159576416,
+      "learning_rate": 1.139789386743754e-05,
+      "loss": 0.9644,
+      "step": 21950
+    },
+    {
+      "epoch": 0.8885066774585189,
+      "grad_norm": 1.3502781391143799,
+      "learning_rate": 1.1377245508982035e-05,
+      "loss": 0.9731,
+      "step": 21955
+    },
+    {
+      "epoch": 0.8887090246863618,
+      "grad_norm": 1.2079834938049316,
+      "learning_rate": 1.1356597150526534e-05,
+      "loss": 0.9691,
+      "step": 21960
+    },
+    {
+      "epoch": 0.8889113719142048,
+      "grad_norm": 1.176963210105896,
+      "learning_rate": 1.1335948792071031e-05,
+      "loss": 0.9759,
+      "step": 21965
+    },
+    {
+      "epoch": 0.8891137191420477,
+      "grad_norm": 1.263867974281311,
+      "learning_rate": 1.1315300433615527e-05,
+      "loss": 0.9369,
+      "step": 21970
+    },
+    {
+      "epoch": 0.8893160663698907,
+      "grad_norm": 1.1320356130599976,
+      "learning_rate": 1.1294652075160026e-05,
+      "loss": 0.9553,
+      "step": 21975
+    },
+    {
+      "epoch": 0.8895184135977338,
+      "grad_norm": 1.204413652420044,
+      "learning_rate": 1.1274003716704523e-05,
+      "loss": 1.0072,
+      "step": 21980
+    },
+    {
+      "epoch": 0.8897207608255767,
+      "grad_norm": 1.1795121431350708,
+      "learning_rate": 1.1253355358249019e-05,
+      "loss": 0.984,
+      "step": 21985
+    },
+    {
+      "epoch": 0.8899231080534197,
+      "grad_norm": 1.2841216325759888,
+      "learning_rate": 1.1232706999793518e-05,
+      "loss": 0.9346,
+      "step": 21990
+    },
+    {
+      "epoch": 0.8901254552812626,
+      "grad_norm": 1.228280782699585,
+      "learning_rate": 1.1212058641338014e-05,
+      "loss": 0.9995,
+      "step": 21995
+    },
+    {
+      "epoch": 0.8903278025091056,
+      "grad_norm": 1.1534844636917114,
+      "learning_rate": 1.1191410282882511e-05,
+      "loss": 0.9765,
+      "step": 22000
+    },
+    {
+      "epoch": 0.8905301497369486,
+      "grad_norm": 1.3221265077590942,
+      "learning_rate": 1.1170761924427008e-05,
+      "loss": 0.9167,
+      "step": 22005
+    },
+    {
+      "epoch": 0.8907324969647916,
+      "grad_norm": 1.1518945693969727,
+      "learning_rate": 1.1150113565971505e-05,
+      "loss": 0.9555,
+      "step": 22010
+    },
+    {
+      "epoch": 0.8909348441926346,
+      "grad_norm": 1.3317203521728516,
+      "learning_rate": 1.1129465207516003e-05,
+      "loss": 0.975,
+      "step": 22015
+    },
+    {
+      "epoch": 0.8911371914204775,
+      "grad_norm": 1.214125394821167,
+      "learning_rate": 1.11088168490605e-05,
+      "loss": 0.9902,
+      "step": 22020
+    },
+    {
+      "epoch": 0.8913395386483205,
+      "grad_norm": 1.1284027099609375,
+      "learning_rate": 1.1088168490604997e-05,
+      "loss": 0.9531,
+      "step": 22025
+    },
+    {
+      "epoch": 0.8915418858761635,
+      "grad_norm": 1.1702163219451904,
+      "learning_rate": 1.1067520132149495e-05,
+      "loss": 0.9817,
+      "step": 22030
+    },
+    {
+      "epoch": 0.8917442331040065,
+      "grad_norm": 1.3232663869857788,
+      "learning_rate": 1.1046871773693992e-05,
+      "loss": 1.0207,
+      "step": 22035
+    },
+    {
+      "epoch": 0.8919465803318495,
+      "grad_norm": 1.061571717262268,
+      "learning_rate": 1.102622341523849e-05,
+      "loss": 1.0115,
+      "step": 22040
+    },
+    {
+      "epoch": 0.8921489275596924,
+      "grad_norm": 1.2872508764266968,
+      "learning_rate": 1.1005575056782987e-05,
+      "loss": 0.9768,
+      "step": 22045
+    },
+    {
+      "epoch": 0.8923512747875354,
+      "grad_norm": 1.2478415966033936,
+      "learning_rate": 1.0984926698327482e-05,
+      "loss": 0.9753,
+      "step": 22050
+    },
+    {
+      "epoch": 0.8925536220153784,
+      "grad_norm": 1.207758903503418,
+      "learning_rate": 1.0964278339871981e-05,
+      "loss": 0.9842,
+      "step": 22055
+    },
+    {
+      "epoch": 0.8927559692432213,
+      "grad_norm": 1.3182350397109985,
+      "learning_rate": 1.0943629981416478e-05,
+      "loss": 0.9764,
+      "step": 22060
+    },
+    {
+      "epoch": 0.8929583164710644,
+      "grad_norm": 1.3606094121932983,
+      "learning_rate": 1.0922981622960974e-05,
+      "loss": 1.009,
+      "step": 22065
+    },
+    {
+      "epoch": 0.8931606636989073,
+      "grad_norm": 1.20542311668396,
+      "learning_rate": 1.0902333264505473e-05,
+      "loss": 0.9638,
+      "step": 22070
+    },
+    {
+      "epoch": 0.8933630109267503,
+      "grad_norm": 1.2064385414123535,
+      "learning_rate": 1.088168490604997e-05,
+      "loss": 0.9451,
+      "step": 22075
+    },
+    {
+      "epoch": 0.8935653581545933,
+      "grad_norm": 1.1925119161605835,
+      "learning_rate": 1.0861036547594466e-05,
+      "loss": 0.985,
+      "step": 22080
+    },
+    {
+      "epoch": 0.8937677053824362,
+      "grad_norm": 1.3347357511520386,
+      "learning_rate": 1.0840388189138965e-05,
+      "loss": 0.9557,
+      "step": 22085
+    },
+    {
+      "epoch": 0.8939700526102793,
+      "grad_norm": 1.3416115045547485,
+      "learning_rate": 1.0819739830683462e-05,
+      "loss": 1.0071,
+      "step": 22090
+    },
+    {
+      "epoch": 0.8941723998381222,
+      "grad_norm": 1.234076738357544,
+      "learning_rate": 1.0799091472227958e-05,
+      "loss": 0.9865,
+      "step": 22095
+    },
+    {
+      "epoch": 0.8943747470659652,
+      "grad_norm": 1.362473487854004,
+      "learning_rate": 1.0778443113772455e-05,
+      "loss": 0.9891,
+      "step": 22100
+    },
+    {
+      "epoch": 0.8945770942938082,
+      "grad_norm": 1.2424629926681519,
+      "learning_rate": 1.0757794755316952e-05,
+      "loss": 0.9554,
+      "step": 22105
+    },
+    {
+      "epoch": 0.8947794415216511,
+      "grad_norm": 1.1338199377059937,
+      "learning_rate": 1.073714639686145e-05,
+      "loss": 0.9994,
+      "step": 22110
+    },
+    {
+      "epoch": 0.8949817887494941,
+      "grad_norm": 1.2993570566177368,
+      "learning_rate": 1.0716498038405947e-05,
+      "loss": 0.9799,
+      "step": 22115
+    },
+    {
+      "epoch": 0.8951841359773371,
+      "grad_norm": 1.1700685024261475,
+      "learning_rate": 1.0695849679950444e-05,
+      "loss": 1.0206,
+      "step": 22120
+    },
+    {
+      "epoch": 0.8953864832051801,
+      "grad_norm": 1.2185229063034058,
+      "learning_rate": 1.0675201321494942e-05,
+      "loss": 1.016,
+      "step": 22125
+    },
+    {
+      "epoch": 0.8955888304330231,
+      "grad_norm": 1.333257794380188,
+      "learning_rate": 1.0654552963039439e-05,
+      "loss": 0.9748,
+      "step": 22130
+    },
+    {
+      "epoch": 0.895791177660866,
+      "grad_norm": 1.2211995124816895,
+      "learning_rate": 1.0633904604583936e-05,
+      "loss": 0.9494,
+      "step": 22135
+    },
+    {
+      "epoch": 0.895993524888709,
+      "grad_norm": 1.155814528465271,
+      "learning_rate": 1.0613256246128433e-05,
+      "loss": 0.9847,
+      "step": 22140
+    },
+    {
+      "epoch": 0.896195872116552,
+      "grad_norm": 1.1218489408493042,
+      "learning_rate": 1.0592607887672929e-05,
+      "loss": 1.0325,
+      "step": 22145
+    },
+    {
+      "epoch": 0.896398219344395,
+      "grad_norm": 1.223176121711731,
+      "learning_rate": 1.0571959529217428e-05,
+      "loss": 1.0434,
+      "step": 22150
+    },
+    {
+      "epoch": 0.896600566572238,
+      "grad_norm": 1.2436774969100952,
+      "learning_rate": 1.0551311170761925e-05,
+      "loss": 0.9439,
+      "step": 22155
+    },
+    {
+      "epoch": 0.8968029138000809,
+      "grad_norm": 1.2586272954940796,
+      "learning_rate": 1.0530662812306421e-05,
+      "loss": 0.9873,
+      "step": 22160
+    },
+    {
+      "epoch": 0.8970052610279239,
+      "grad_norm": 1.3155723810195923,
+      "learning_rate": 1.051001445385092e-05,
+      "loss": 0.9646,
+      "step": 22165
+    },
+    {
+      "epoch": 0.8972076082557668,
+      "grad_norm": 1.1274899244308472,
+      "learning_rate": 1.0489366095395417e-05,
+      "loss": 1.0054,
+      "step": 22170
+    },
+    {
+      "epoch": 0.8974099554836099,
+      "grad_norm": 1.1293829679489136,
+      "learning_rate": 1.0468717736939913e-05,
+      "loss": 0.9475,
+      "step": 22175
+    },
+    {
+      "epoch": 0.8976123027114529,
+      "grad_norm": 1.286868691444397,
+      "learning_rate": 1.0448069378484412e-05,
+      "loss": 0.9467,
+      "step": 22180
+    },
+    {
+      "epoch": 0.8978146499392958,
+      "grad_norm": 1.1316075325012207,
+      "learning_rate": 1.0427421020028909e-05,
+      "loss": 0.9477,
+      "step": 22185
+    },
+    {
+      "epoch": 0.8980169971671388,
+      "grad_norm": 1.0803990364074707,
+      "learning_rate": 1.0406772661573405e-05,
+      "loss": 0.9651,
+      "step": 22190
+    },
+    {
+      "epoch": 0.8982193443949817,
+      "grad_norm": 1.2619987726211548,
+      "learning_rate": 1.0386124303117902e-05,
+      "loss": 1.0046,
+      "step": 22195
+    },
+    {
+      "epoch": 0.8984216916228248,
+      "grad_norm": 1.2577831745147705,
+      "learning_rate": 1.03654759446624e-05,
+      "loss": 1.0353,
+      "step": 22200
+    },
+    {
+      "epoch": 0.8986240388506678,
+      "grad_norm": 1.1763091087341309,
+      "learning_rate": 1.0344827586206897e-05,
+      "loss": 1.0354,
+      "step": 22205
+    },
+    {
+      "epoch": 0.8988263860785107,
+      "grad_norm": 1.2089554071426392,
+      "learning_rate": 1.0324179227751394e-05,
+      "loss": 0.9622,
+      "step": 22210
+    },
+    {
+      "epoch": 0.8990287333063537,
+      "grad_norm": 1.1004221439361572,
+      "learning_rate": 1.0303530869295891e-05,
+      "loss": 0.9883,
+      "step": 22215
+    },
+    {
+      "epoch": 0.8992310805341966,
+      "grad_norm": 1.1818112134933472,
+      "learning_rate": 1.0282882510840389e-05,
+      "loss": 1.0136,
+      "step": 22220
+    },
+    {
+      "epoch": 0.8994334277620396,
+      "grad_norm": 1.1124714612960815,
+      "learning_rate": 1.0262234152384886e-05,
+      "loss": 0.9735,
+      "step": 22225
+    },
+    {
+      "epoch": 0.8996357749898827,
+      "grad_norm": 1.2221177816390991,
+      "learning_rate": 1.0241585793929383e-05,
+      "loss": 0.9817,
+      "step": 22230
+    },
+    {
+      "epoch": 0.8998381222177256,
+      "grad_norm": 1.1876314878463745,
+      "learning_rate": 1.022093743547388e-05,
+      "loss": 0.9975,
+      "step": 22235
+    },
+    {
+      "epoch": 0.9000404694455686,
+      "grad_norm": 1.233405351638794,
+      "learning_rate": 1.0200289077018378e-05,
+      "loss": 0.9749,
+      "step": 22240
+    },
+    {
+      "epoch": 0.9002428166734115,
+      "grad_norm": 1.1681106090545654,
+      "learning_rate": 1.0179640718562875e-05,
+      "loss": 0.8572,
+      "step": 22245
+    },
+    {
+      "epoch": 0.9004451639012545,
+      "grad_norm": 1.2301884889602661,
+      "learning_rate": 1.0158992360107372e-05,
+      "loss": 0.9903,
+      "step": 22250
+    },
+    {
+      "epoch": 0.9006475111290976,
+      "grad_norm": 1.1902940273284912,
+      "learning_rate": 1.0138344001651868e-05,
+      "loss": 1.0435,
+      "step": 22255
+    },
+    {
+      "epoch": 0.9008498583569405,
+      "grad_norm": 1.2263290882110596,
+      "learning_rate": 1.0117695643196367e-05,
+      "loss": 0.9561,
+      "step": 22260
+    },
+    {
+      "epoch": 0.9010522055847835,
+      "grad_norm": 1.2020978927612305,
+      "learning_rate": 1.0097047284740864e-05,
+      "loss": 0.9672,
+      "step": 22265
+    },
+    {
+      "epoch": 0.9012545528126265,
+      "grad_norm": 1.1788512468338013,
+      "learning_rate": 1.007639892628536e-05,
+      "loss": 0.945,
+      "step": 22270
+    },
+    {
+      "epoch": 0.9014569000404694,
+      "grad_norm": 1.294374704360962,
+      "learning_rate": 1.0055750567829859e-05,
+      "loss": 0.9965,
+      "step": 22275
+    },
+    {
+      "epoch": 0.9016592472683124,
+      "grad_norm": 1.2951778173446655,
+      "learning_rate": 1.0035102209374356e-05,
+      "loss": 0.9357,
+      "step": 22280
+    },
+    {
+      "epoch": 0.9018615944961554,
+      "grad_norm": 1.068011999130249,
+      "learning_rate": 1.0014453850918852e-05,
+      "loss": 0.9796,
+      "step": 22285
+    },
+    {
+      "epoch": 0.9020639417239984,
+      "grad_norm": 1.2775375843048096,
+      "learning_rate": 9.993805492463349e-06,
+      "loss": 0.9574,
+      "step": 22290
+    },
+    {
+      "epoch": 0.9022662889518414,
+      "grad_norm": 1.3473182916641235,
+      "learning_rate": 9.973157134007848e-06,
+      "loss": 0.9608,
+      "step": 22295
+    },
+    {
+      "epoch": 0.9024686361796843,
+      "grad_norm": 1.080181360244751,
+      "learning_rate": 9.952508775552344e-06,
+      "loss": 1.0191,
+      "step": 22300
+    },
+    {
+      "epoch": 0.9026709834075273,
+      "grad_norm": 1.2121566534042358,
+      "learning_rate": 9.931860417096841e-06,
+      "loss": 1.0055,
+      "step": 22305
+    },
+    {
+      "epoch": 0.9028733306353703,
+      "grad_norm": 1.2975614070892334,
+      "learning_rate": 9.911212058641338e-06,
+      "loss": 1.0079,
+      "step": 22310
+    },
+    {
+      "epoch": 0.9030756778632133,
+      "grad_norm": 1.1123826503753662,
+      "learning_rate": 9.890563700185835e-06,
+      "loss": 1.0012,
+      "step": 22315
+    },
+    {
+      "epoch": 0.9032780250910563,
+      "grad_norm": 1.1919848918914795,
+      "learning_rate": 9.869915341730333e-06,
+      "loss": 0.9943,
+      "step": 22320
+    },
+    {
+      "epoch": 0.9034803723188992,
+      "grad_norm": 1.231441617012024,
+      "learning_rate": 9.84926698327483e-06,
+      "loss": 1.0508,
+      "step": 22325
+    },
+    {
+      "epoch": 0.9036827195467422,
+      "grad_norm": 1.2481560707092285,
+      "learning_rate": 9.828618624819327e-06,
+      "loss": 1.0405,
+      "step": 22330
+    },
+    {
+      "epoch": 0.9038850667745851,
+      "grad_norm": 1.2379087209701538,
+      "learning_rate": 9.807970266363825e-06,
+      "loss": 1.0126,
+      "step": 22335
+    },
+    {
+      "epoch": 0.9040874140024282,
+      "grad_norm": 1.2807567119598389,
+      "learning_rate": 9.787321907908322e-06,
+      "loss": 1.0009,
+      "step": 22340
+    },
+    {
+      "epoch": 0.9042897612302712,
+      "grad_norm": 1.19586181640625,
+      "learning_rate": 9.76667354945282e-06,
+      "loss": 0.9922,
+      "step": 22345
+    },
+    {
+      "epoch": 0.9044921084581141,
+      "grad_norm": 1.257541298866272,
+      "learning_rate": 9.746025190997317e-06,
+      "loss": 1.0157,
+      "step": 22350
+    },
+    {
+      "epoch": 0.9046944556859571,
+      "grad_norm": 1.1988418102264404,
+      "learning_rate": 9.725376832541814e-06,
+      "loss": 0.9513,
+      "step": 22355
+    },
+    {
+      "epoch": 0.9048968029138,
+      "grad_norm": 1.268105149269104,
+      "learning_rate": 9.704728474086311e-06,
+      "loss": 1.0328,
+      "step": 22360
+    },
+    {
+      "epoch": 0.9050991501416431,
+      "grad_norm": 1.1456495523452759,
+      "learning_rate": 9.684080115630807e-06,
+      "loss": 1.0015,
+      "step": 22365
+    },
+    {
+      "epoch": 0.9053014973694861,
+      "grad_norm": 1.2869234085083008,
+      "learning_rate": 9.663431757175306e-06,
+      "loss": 1.0053,
+      "step": 22370
+    },
+    {
+      "epoch": 0.905503844597329,
+      "grad_norm": 1.133236050605774,
+      "learning_rate": 9.642783398719803e-06,
+      "loss": 0.9603,
+      "step": 22375
+    },
+    {
+      "epoch": 0.905706191825172,
+      "grad_norm": 1.200329303741455,
+      "learning_rate": 9.622135040264299e-06,
+      "loss": 0.9874,
+      "step": 22380
+    },
+    {
+      "epoch": 0.9059085390530149,
+      "grad_norm": 1.145104169845581,
+      "learning_rate": 9.601486681808796e-06,
+      "loss": 0.9957,
+      "step": 22385
+    },
+    {
+      "epoch": 0.9061108862808579,
+      "grad_norm": 1.1849706172943115,
+      "learning_rate": 9.580838323353295e-06,
+      "loss": 1.0053,
+      "step": 22390
+    },
+    {
+      "epoch": 0.906313233508701,
+      "grad_norm": 1.083436131477356,
+      "learning_rate": 9.56018996489779e-06,
+      "loss": 0.9284,
+      "step": 22395
+    },
+    {
+      "epoch": 0.9065155807365439,
+      "grad_norm": 1.2961195707321167,
+      "learning_rate": 9.539541606442288e-06,
+      "loss": 1.0143,
+      "step": 22400
+    },
+    {
+      "epoch": 0.9067179279643869,
+      "grad_norm": 1.2798185348510742,
+      "learning_rate": 9.518893247986785e-06,
+      "loss": 0.9898,
+      "step": 22405
+    },
+    {
+      "epoch": 0.9069202751922298,
+      "grad_norm": 1.2063452005386353,
+      "learning_rate": 9.498244889531282e-06,
+      "loss": 1.013,
+      "step": 22410
+    },
+    {
+      "epoch": 0.9071226224200728,
+      "grad_norm": 1.2208327054977417,
+      "learning_rate": 9.47759653107578e-06,
+      "loss": 0.9956,
+      "step": 22415
+    },
+    {
+      "epoch": 0.9073249696479159,
+      "grad_norm": 2.438999652862549,
+      "learning_rate": 9.456948172620277e-06,
+      "loss": 1.0163,
+      "step": 22420
+    },
+    {
+      "epoch": 0.9075273168757588,
+      "grad_norm": 1.2716456651687622,
+      "learning_rate": 9.436299814164774e-06,
+      "loss": 1.0141,
+      "step": 22425
+    },
+    {
+      "epoch": 0.9077296641036018,
+      "grad_norm": 1.2127629518508911,
+      "learning_rate": 9.415651455709272e-06,
+      "loss": 0.982,
+      "step": 22430
+    },
+    {
+      "epoch": 0.9079320113314447,
+      "grad_norm": 1.2110321521759033,
+      "learning_rate": 9.395003097253769e-06,
+      "loss": 0.9764,
+      "step": 22435
+    },
+    {
+      "epoch": 0.9081343585592877,
+      "grad_norm": 1.1984643936157227,
+      "learning_rate": 9.374354738798266e-06,
+      "loss": 0.9687,
+      "step": 22440
+    },
+    {
+      "epoch": 0.9083367057871308,
+      "grad_norm": 1.256187915802002,
+      "learning_rate": 9.353706380342763e-06,
+      "loss": 0.9754,
+      "step": 22445
+    },
+    {
+      "epoch": 0.9085390530149737,
+      "grad_norm": 1.2054247856140137,
+      "learning_rate": 9.33305802188726e-06,
+      "loss": 0.9847,
+      "step": 22450
+    },
+    {
+      "epoch": 0.9087414002428167,
+      "grad_norm": 1.2009963989257812,
+      "learning_rate": 9.312409663431758e-06,
+      "loss": 0.9836,
+      "step": 22455
+    },
+    {
+      "epoch": 0.9089437474706596,
+      "grad_norm": 1.1755682229995728,
+      "learning_rate": 9.291761304976254e-06,
+      "loss": 1.0035,
+      "step": 22460
+    },
+    {
+      "epoch": 0.9091460946985026,
+      "grad_norm": 1.2057689428329468,
+      "learning_rate": 9.271112946520753e-06,
+      "loss": 0.9971,
+      "step": 22465
+    },
+    {
+      "epoch": 0.9093484419263456,
+      "grad_norm": 1.2346726655960083,
+      "learning_rate": 9.25046458806525e-06,
+      "loss": 0.9771,
+      "step": 22470
+    },
+    {
+      "epoch": 0.9095507891541886,
+      "grad_norm": 1.230569839477539,
+      "learning_rate": 9.229816229609746e-06,
+      "loss": 0.9842,
+      "step": 22475
+    },
+    {
+      "epoch": 0.9097531363820316,
+      "grad_norm": 1.242022156715393,
+      "learning_rate": 9.209167871154243e-06,
+      "loss": 0.9792,
+      "step": 22480
+    },
+    {
+      "epoch": 0.9099554836098745,
+      "grad_norm": 1.308518648147583,
+      "learning_rate": 9.188519512698742e-06,
+      "loss": 0.9571,
+      "step": 22485
+    },
+    {
+      "epoch": 0.9101578308377175,
+      "grad_norm": 1.1113145351409912,
+      "learning_rate": 9.167871154243237e-06,
+      "loss": 0.9663,
+      "step": 22490
+    },
+    {
+      "epoch": 0.9103601780655605,
+      "grad_norm": 1.3198206424713135,
+      "learning_rate": 9.147222795787735e-06,
+      "loss": 1.0134,
+      "step": 22495
+    },
+    {
+      "epoch": 0.9105625252934035,
+      "grad_norm": 1.285475492477417,
+      "learning_rate": 9.126574437332234e-06,
+      "loss": 0.9834,
+      "step": 22500
+    },
+    {
+      "epoch": 0.9107648725212465,
+      "grad_norm": 1.3377816677093506,
+      "learning_rate": 9.10592607887673e-06,
+      "loss": 1.0289,
+      "step": 22505
+    },
+    {
+      "epoch": 0.9109672197490895,
+      "grad_norm": 1.1217929124832153,
+      "learning_rate": 9.085277720421227e-06,
+      "loss": 0.9706,
+      "step": 22510
+    },
+    {
+      "epoch": 0.9111695669769324,
+      "grad_norm": 1.3376717567443848,
+      "learning_rate": 9.064629361965724e-06,
+      "loss": 1.0767,
+      "step": 22515
+    },
+    {
+      "epoch": 0.9113719142047754,
+      "grad_norm": 1.2571460008621216,
+      "learning_rate": 9.043981003510221e-06,
+      "loss": 0.9529,
+      "step": 22520
+    },
+    {
+      "epoch": 0.9115742614326183,
+      "grad_norm": 1.1524591445922852,
+      "learning_rate": 9.023332645054719e-06,
+      "loss": 0.9613,
+      "step": 22525
+    },
+    {
+      "epoch": 0.9117766086604614,
+      "grad_norm": 1.3369122743606567,
+      "learning_rate": 9.002684286599216e-06,
+      "loss": 1.009,
+      "step": 22530
+    },
+    {
+      "epoch": 0.9119789558883044,
+      "grad_norm": 1.2407011985778809,
+      "learning_rate": 8.982035928143713e-06,
+      "loss": 0.9616,
+      "step": 22535
+    },
+    {
+      "epoch": 0.9121813031161473,
+      "grad_norm": 1.1766000986099243,
+      "learning_rate": 8.96138756968821e-06,
+      "loss": 1.0649,
+      "step": 22540
+    },
+    {
+      "epoch": 0.9123836503439903,
+      "grad_norm": 1.2149438858032227,
+      "learning_rate": 8.940739211232708e-06,
+      "loss": 0.9535,
+      "step": 22545
+    },
+    {
+      "epoch": 0.9125859975718332,
+      "grad_norm": 1.3015145063400269,
+      "learning_rate": 8.920090852777205e-06,
+      "loss": 0.961,
+      "step": 22550
+    },
+    {
+      "epoch": 0.9127883447996763,
+      "grad_norm": 1.2790932655334473,
+      "learning_rate": 8.899442494321702e-06,
+      "loss": 1.0255,
+      "step": 22555
+    },
+    {
+      "epoch": 0.9129906920275193,
+      "grad_norm": 1.3546644449234009,
+      "learning_rate": 8.8787941358662e-06,
+      "loss": 1.0645,
+      "step": 22560
+    },
+    {
+      "epoch": 0.9131930392553622,
+      "grad_norm": 1.135512113571167,
+      "learning_rate": 8.858145777410697e-06,
+      "loss": 0.9898,
+      "step": 22565
+    },
+    {
+      "epoch": 0.9133953864832052,
+      "grad_norm": 1.1631687879562378,
+      "learning_rate": 8.837497418955193e-06,
+      "loss": 0.9646,
+      "step": 22570
+    },
+    {
+      "epoch": 0.9135977337110481,
+      "grad_norm": 1.1153877973556519,
+      "learning_rate": 8.81684906049969e-06,
+      "loss": 0.9689,
+      "step": 22575
+    },
+    {
+      "epoch": 0.9138000809388911,
+      "grad_norm": 1.1794513463974,
+      "learning_rate": 8.796200702044189e-06,
+      "loss": 1.0014,
+      "step": 22580
+    },
+    {
+      "epoch": 0.9140024281667342,
+      "grad_norm": 1.1592798233032227,
+      "learning_rate": 8.775552343588684e-06,
+      "loss": 0.9923,
+      "step": 22585
+    },
+    {
+      "epoch": 0.9142047753945771,
+      "grad_norm": 1.2975784540176392,
+      "learning_rate": 8.754903985133182e-06,
+      "loss": 0.9922,
+      "step": 22590
+    },
+    {
+      "epoch": 0.9144071226224201,
+      "grad_norm": 1.25165593624115,
+      "learning_rate": 8.73425562667768e-06,
+      "loss": 0.9499,
+      "step": 22595
+    },
+    {
+      "epoch": 0.914609469850263,
+      "grad_norm": 1.1658289432525635,
+      "learning_rate": 8.713607268222176e-06,
+      "loss": 0.963,
+      "step": 22600
+    },
+    {
+      "epoch": 0.914811817078106,
+      "grad_norm": 1.2364903688430786,
+      "learning_rate": 8.692958909766674e-06,
+      "loss": 0.9883,
+      "step": 22605
+    },
+    {
+      "epoch": 0.9150141643059491,
+      "grad_norm": 1.1751139163970947,
+      "learning_rate": 8.672310551311171e-06,
+      "loss": 0.9371,
+      "step": 22610
+    },
+    {
+      "epoch": 0.915216511533792,
+      "grad_norm": 1.239706039428711,
+      "learning_rate": 8.651662192855668e-06,
+      "loss": 0.986,
+      "step": 22615
+    },
+    {
+      "epoch": 0.915418858761635,
+      "grad_norm": 1.1792747974395752,
+      "learning_rate": 8.631013834400165e-06,
+      "loss": 0.9937,
+      "step": 22620
+    },
+    {
+      "epoch": 0.9156212059894779,
+      "grad_norm": 1.2934255599975586,
+      "learning_rate": 8.610365475944663e-06,
+      "loss": 1.0028,
+      "step": 22625
+    },
+    {
+      "epoch": 0.9158235532173209,
+      "grad_norm": 1.1981898546218872,
+      "learning_rate": 8.58971711748916e-06,
+      "loss": 0.9849,
+      "step": 22630
+    },
+    {
+      "epoch": 0.9160259004451639,
+      "grad_norm": 1.1199761629104614,
+      "learning_rate": 8.569068759033657e-06,
+      "loss": 0.9912,
+      "step": 22635
+    },
+    {
+      "epoch": 0.9162282476730069,
+      "grad_norm": 1.2269060611724854,
+      "learning_rate": 8.548420400578155e-06,
+      "loss": 1.002,
+      "step": 22640
+    },
+    {
+      "epoch": 0.9164305949008499,
+      "grad_norm": 1.19041907787323,
+      "learning_rate": 8.527772042122652e-06,
+      "loss": 0.9154,
+      "step": 22645
+    },
+    {
+      "epoch": 0.9166329421286928,
+      "grad_norm": 1.2220449447631836,
+      "learning_rate": 8.50712368366715e-06,
+      "loss": 0.9968,
+      "step": 22650
+    },
+    {
+      "epoch": 0.9168352893565358,
+      "grad_norm": 1.2628720998764038,
+      "learning_rate": 8.486475325211647e-06,
+      "loss": 1.062,
+      "step": 22655
+    },
+    {
+      "epoch": 0.9170376365843788,
+      "grad_norm": 1.2970205545425415,
+      "learning_rate": 8.465826966756144e-06,
+      "loss": 0.9405,
+      "step": 22660
+    },
+    {
+      "epoch": 0.9172399838122218,
+      "grad_norm": 1.187365174293518,
+      "learning_rate": 8.44517860830064e-06,
+      "loss": 0.9738,
+      "step": 22665
+    },
+    {
+      "epoch": 0.9174423310400648,
+      "grad_norm": 1.2985329627990723,
+      "learning_rate": 8.424530249845137e-06,
+      "loss": 1.0047,
+      "step": 22670
+    },
+    {
+      "epoch": 0.9176446782679077,
+      "grad_norm": 1.143072247505188,
+      "learning_rate": 8.403881891389636e-06,
+      "loss": 0.9638,
+      "step": 22675
+    },
+    {
+      "epoch": 0.9178470254957507,
+      "grad_norm": 1.1585158109664917,
+      "learning_rate": 8.383233532934131e-06,
+      "loss": 0.9175,
+      "step": 22680
+    },
+    {
+      "epoch": 0.9180493727235937,
+      "grad_norm": 1.1999739408493042,
+      "learning_rate": 8.362585174478629e-06,
+      "loss": 1.0941,
+      "step": 22685
+    },
+    {
+      "epoch": 0.9182517199514366,
+      "grad_norm": 1.2880604267120361,
+      "learning_rate": 8.341936816023128e-06,
+      "loss": 0.9812,
+      "step": 22690
+    },
+    {
+      "epoch": 0.9184540671792797,
+      "grad_norm": 1.1725088357925415,
+      "learning_rate": 8.321288457567623e-06,
+      "loss": 0.9726,
+      "step": 22695
+    },
+    {
+      "epoch": 0.9186564144071226,
+      "grad_norm": 1.2655597925186157,
+      "learning_rate": 8.30064009911212e-06,
+      "loss": 0.955,
+      "step": 22700
+    },
+    {
+      "epoch": 0.9188587616349656,
+      "grad_norm": 1.1834826469421387,
+      "learning_rate": 8.27999174065662e-06,
+      "loss": 0.9409,
+      "step": 22705
+    },
+    {
+      "epoch": 0.9190611088628086,
+      "grad_norm": 1.135432243347168,
+      "learning_rate": 8.259343382201115e-06,
+      "loss": 0.9644,
+      "step": 22710
+    },
+    {
+      "epoch": 0.9192634560906515,
+      "grad_norm": 1.1146769523620605,
+      "learning_rate": 8.238695023745612e-06,
+      "loss": 0.9816,
+      "step": 22715
+    },
+    {
+      "epoch": 0.9194658033184946,
+      "grad_norm": 1.2879973649978638,
+      "learning_rate": 8.21804666529011e-06,
+      "loss": 1.0648,
+      "step": 22720
+    },
+    {
+      "epoch": 0.9196681505463375,
+      "grad_norm": 1.2010254859924316,
+      "learning_rate": 8.197398306834607e-06,
+      "loss": 0.9895,
+      "step": 22725
+    },
+    {
+      "epoch": 0.9198704977741805,
+      "grad_norm": 1.1273995637893677,
+      "learning_rate": 8.176749948379104e-06,
+      "loss": 1.0098,
+      "step": 22730
+    },
+    {
+      "epoch": 0.9200728450020235,
+      "grad_norm": 1.1119916439056396,
+      "learning_rate": 8.156101589923602e-06,
+      "loss": 1.0431,
+      "step": 22735
+    },
+    {
+      "epoch": 0.9202751922298664,
+      "grad_norm": 1.2416926622390747,
+      "learning_rate": 8.135453231468099e-06,
+      "loss": 1.04,
+      "step": 22740
+    },
+    {
+      "epoch": 0.9204775394577094,
+      "grad_norm": 1.101529598236084,
+      "learning_rate": 8.114804873012596e-06,
+      "loss": 0.9656,
+      "step": 22745
+    },
+    {
+      "epoch": 0.9206798866855525,
+      "grad_norm": 1.2064924240112305,
+      "learning_rate": 8.094156514557094e-06,
+      "loss": 0.9936,
+      "step": 22750
+    },
+    {
+      "epoch": 0.9208822339133954,
+      "grad_norm": 1.202167272567749,
+      "learning_rate": 8.07350815610159e-06,
+      "loss": 0.9247,
+      "step": 22755
+    },
+    {
+      "epoch": 0.9210845811412384,
+      "grad_norm": 1.2607907056808472,
+      "learning_rate": 8.052859797646088e-06,
+      "loss": 1.0179,
+      "step": 22760
+    },
+    {
+      "epoch": 0.9212869283690813,
+      "grad_norm": 1.1044496297836304,
+      "learning_rate": 8.032211439190584e-06,
+      "loss": 0.9685,
+      "step": 22765
+    },
+    {
+      "epoch": 0.9214892755969243,
+      "grad_norm": 1.3169591426849365,
+      "learning_rate": 8.011563080735083e-06,
+      "loss": 1.0145,
+      "step": 22770
+    },
+    {
+      "epoch": 0.9216916228247674,
+      "grad_norm": 1.160158634185791,
+      "learning_rate": 7.990914722279578e-06,
+      "loss": 0.9693,
+      "step": 22775
+    },
+    {
+      "epoch": 0.9218939700526103,
+      "grad_norm": 1.3280168771743774,
+      "learning_rate": 7.970266363824076e-06,
+      "loss": 0.9799,
+      "step": 22780
+    },
+    {
+      "epoch": 0.9220963172804533,
+      "grad_norm": 1.2774527072906494,
+      "learning_rate": 7.949618005368575e-06,
+      "loss": 0.9861,
+      "step": 22785
+    },
+    {
+      "epoch": 0.9222986645082962,
+      "grad_norm": 1.2364764213562012,
+      "learning_rate": 7.92896964691307e-06,
+      "loss": 0.929,
+      "step": 22790
+    },
+    {
+      "epoch": 0.9225010117361392,
+      "grad_norm": 1.221522569656372,
+      "learning_rate": 7.908321288457568e-06,
+      "loss": 0.9077,
+      "step": 22795
+    },
+    {
+      "epoch": 0.9227033589639821,
+      "grad_norm": 1.2038472890853882,
+      "learning_rate": 7.887672930002066e-06,
+      "loss": 1.0101,
+      "step": 22800
+    },
+    {
+      "epoch": 0.9229057061918252,
+      "grad_norm": 1.1088597774505615,
+      "learning_rate": 7.867024571546562e-06,
+      "loss": 0.9654,
+      "step": 22805
+    },
+    {
+      "epoch": 0.9231080534196682,
+      "grad_norm": 1.1928927898406982,
+      "learning_rate": 7.84637621309106e-06,
+      "loss": 0.9374,
+      "step": 22810
+    },
+    {
+      "epoch": 0.9233104006475111,
+      "grad_norm": 1.1673182249069214,
+      "learning_rate": 7.825727854635557e-06,
+      "loss": 1.0107,
+      "step": 22815
+    },
+    {
+      "epoch": 0.9235127478753541,
+      "grad_norm": 1.2621965408325195,
+      "learning_rate": 7.805079496180054e-06,
+      "loss": 0.9653,
+      "step": 22820
+    },
+    {
+      "epoch": 0.923715095103197,
+      "grad_norm": 1.2311797142028809,
+      "learning_rate": 7.784431137724551e-06,
+      "loss": 0.9723,
+      "step": 22825
+    },
+    {
+      "epoch": 0.9239174423310401,
+      "grad_norm": 1.1744613647460938,
+      "learning_rate": 7.763782779269049e-06,
+      "loss": 0.9745,
+      "step": 22830
+    },
+    {
+      "epoch": 0.9241197895588831,
+      "grad_norm": 1.2949620485305786,
+      "learning_rate": 7.743134420813546e-06,
+      "loss": 0.9944,
+      "step": 22835
+    },
+    {
+      "epoch": 0.924322136786726,
+      "grad_norm": 1.1407514810562134,
+      "learning_rate": 7.722486062358043e-06,
+      "loss": 0.9649,
+      "step": 22840
+    },
+    {
+      "epoch": 0.924524484014569,
+      "grad_norm": 1.2529412508010864,
+      "learning_rate": 7.70183770390254e-06,
+      "loss": 1.0489,
+      "step": 22845
+    },
+    {
+      "epoch": 0.924726831242412,
+      "grad_norm": 1.30082106590271,
+      "learning_rate": 7.681189345447038e-06,
+      "loss": 0.9873,
+      "step": 22850
+    },
+    {
+      "epoch": 0.9249291784702549,
+      "grad_norm": 1.1744364500045776,
+      "learning_rate": 7.660540986991535e-06,
+      "loss": 0.955,
+      "step": 22855
+    },
+    {
+      "epoch": 0.925131525698098,
+      "grad_norm": 1.1827532052993774,
+      "learning_rate": 7.63989262853603e-06,
+      "loss": 0.9811,
+      "step": 22860
+    },
+    {
+      "epoch": 0.9253338729259409,
+      "grad_norm": 1.402942419052124,
+      "learning_rate": 7.619244270080529e-06,
+      "loss": 1.0014,
+      "step": 22865
+    },
+    {
+      "epoch": 0.9255362201537839,
+      "grad_norm": 1.2489560842514038,
+      "learning_rate": 7.598595911625025e-06,
+      "loss": 0.9448,
+      "step": 22870
+    },
+    {
+      "epoch": 0.9257385673816269,
+      "grad_norm": 1.270635962486267,
+      "learning_rate": 7.577947553169523e-06,
+      "loss": 0.9908,
+      "step": 22875
+    },
+    {
+      "epoch": 0.9259409146094698,
+      "grad_norm": 1.2608011960983276,
+      "learning_rate": 7.557299194714021e-06,
+      "loss": 0.9864,
+      "step": 22880
+    },
+    {
+      "epoch": 0.9261432618373129,
+      "grad_norm": 1.2094098329544067,
+      "learning_rate": 7.536650836258517e-06,
+      "loss": 0.965,
+      "step": 22885
+    },
+    {
+      "epoch": 0.9263456090651558,
+      "grad_norm": 1.3155937194824219,
+      "learning_rate": 7.516002477803015e-06,
+      "loss": 0.9344,
+      "step": 22890
+    },
+    {
+      "epoch": 0.9265479562929988,
+      "grad_norm": 1.2582231760025024,
+      "learning_rate": 7.495354119347513e-06,
+      "loss": 0.9776,
+      "step": 22895
+    },
+    {
+      "epoch": 0.9267503035208418,
+      "grad_norm": 1.2340329885482788,
+      "learning_rate": 7.474705760892009e-06,
+      "loss": 1.0597,
+      "step": 22900
+    },
+    {
+      "epoch": 0.9269526507486847,
+      "grad_norm": 1.1896069049835205,
+      "learning_rate": 7.454057402436506e-06,
+      "loss": 1.0278,
+      "step": 22905
+    },
+    {
+      "epoch": 0.9271549979765277,
+      "grad_norm": 1.203372836112976,
+      "learning_rate": 7.4334090439810045e-06,
+      "loss": 1.0645,
+      "step": 22910
+    },
+    {
+      "epoch": 0.9273573452043707,
+      "grad_norm": 1.2288404703140259,
+      "learning_rate": 7.412760685525501e-06,
+      "loss": 0.9511,
+      "step": 22915
+    },
+    {
+      "epoch": 0.9275596924322137,
+      "grad_norm": 1.1734936237335205,
+      "learning_rate": 7.392112327069998e-06,
+      "loss": 0.9714,
+      "step": 22920
+    },
+    {
+      "epoch": 0.9277620396600567,
+      "grad_norm": 1.307991623878479,
+      "learning_rate": 7.371463968614495e-06,
+      "loss": 0.9713,
+      "step": 22925
+    },
+    {
+      "epoch": 0.9279643868878996,
+      "grad_norm": 1.3006573915481567,
+      "learning_rate": 7.350815610158993e-06,
+      "loss": 1.0358,
+      "step": 22930
+    },
+    {
+      "epoch": 0.9281667341157426,
+      "grad_norm": 1.288576364517212,
+      "learning_rate": 7.33016725170349e-06,
+      "loss": 1.0295,
+      "step": 22935
+    },
+    {
+      "epoch": 0.9283690813435856,
+      "grad_norm": 1.1439498662948608,
+      "learning_rate": 7.3095188932479866e-06,
+      "loss": 1.0337,
+      "step": 22940
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 1.1806100606918335,
+      "learning_rate": 7.288870534792485e-06,
+      "loss": 0.9687,
+      "step": 22945
+    },
+    {
+      "epoch": 0.9287737757992716,
+      "grad_norm": 1.2200596332550049,
+      "learning_rate": 7.268222176336982e-06,
+      "loss": 1.0372,
+      "step": 22950
+    },
+    {
+      "epoch": 0.9289761230271145,
+      "grad_norm": 1.3245278596878052,
+      "learning_rate": 7.2475738178814785e-06,
+      "loss": 1.0367,
+      "step": 22955
+    },
+    {
+      "epoch": 0.9291784702549575,
+      "grad_norm": 1.2013788223266602,
+      "learning_rate": 7.226925459425976e-06,
+      "loss": 0.9974,
+      "step": 22960
+    },
+    {
+      "epoch": 0.9293808174828004,
+      "grad_norm": 1.2571405172348022,
+      "learning_rate": 7.206277100970474e-06,
+      "loss": 0.9595,
+      "step": 22965
+    },
+    {
+      "epoch": 0.9295831647106435,
+      "grad_norm": 1.1836342811584473,
+      "learning_rate": 7.18562874251497e-06,
+      "loss": 1.0234,
+      "step": 22970
+    },
+    {
+      "epoch": 0.9297855119384865,
+      "grad_norm": 1.231541633605957,
+      "learning_rate": 7.164980384059468e-06,
+      "loss": 1.0239,
+      "step": 22975
+    },
+    {
+      "epoch": 0.9299878591663294,
+      "grad_norm": 1.2750858068466187,
+      "learning_rate": 7.144332025603964e-06,
+      "loss": 0.9454,
+      "step": 22980
+    },
+    {
+      "epoch": 0.9301902063941724,
+      "grad_norm": 1.145346999168396,
+      "learning_rate": 7.123683667148462e-06,
+      "loss": 1.0459,
+      "step": 22985
+    },
+    {
+      "epoch": 0.9303925536220153,
+      "grad_norm": 1.2808715105056763,
+      "learning_rate": 7.1030353086929595e-06,
+      "loss": 0.9821,
+      "step": 22990
+    },
+    {
+      "epoch": 0.9305949008498584,
+      "grad_norm": 1.3871564865112305,
+      "learning_rate": 7.082386950237456e-06,
+      "loss": 1.0195,
+      "step": 22995
+    },
+    {
+      "epoch": 0.9307972480777014,
+      "grad_norm": 1.1897286176681519,
+      "learning_rate": 7.061738591781953e-06,
+      "loss": 1.0062,
+      "step": 23000
+    },
+    {
+      "epoch": 0.9309995953055443,
+      "grad_norm": 1.2897855043411255,
+      "learning_rate": 7.0410902333264514e-06,
+      "loss": 0.9726,
+      "step": 23005
+    },
+    {
+      "epoch": 0.9312019425333873,
+      "grad_norm": 1.158328652381897,
+      "learning_rate": 7.020441874870948e-06,
+      "loss": 0.9388,
+      "step": 23010
+    },
+    {
+      "epoch": 0.9314042897612302,
+      "grad_norm": 1.2545818090438843,
+      "learning_rate": 6.999793516415445e-06,
+      "loss": 1.073,
+      "step": 23015
+    },
+    {
+      "epoch": 0.9316066369890732,
+      "grad_norm": 1.1400245428085327,
+      "learning_rate": 6.979145157959943e-06,
+      "loss": 0.98,
+      "step": 23020
+    },
+    {
+      "epoch": 0.9318089842169163,
+      "grad_norm": 1.1022324562072754,
+      "learning_rate": 6.95849679950444e-06,
+      "loss": 1.0085,
+      "step": 23025
+    },
+    {
+      "epoch": 0.9320113314447592,
+      "grad_norm": 1.2606849670410156,
+      "learning_rate": 6.937848441048937e-06,
+      "loss": 1.0295,
+      "step": 23030
+    },
+    {
+      "epoch": 0.9322136786726022,
+      "grad_norm": 1.250056266784668,
+      "learning_rate": 6.9172000825934335e-06,
+      "loss": 0.9885,
+      "step": 23035
+    },
+    {
+      "epoch": 0.9324160259004451,
+      "grad_norm": 1.1324617862701416,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.9748,
+      "step": 23040
+    },
+    {
+      "epoch": 0.9326183731282881,
+      "grad_norm": 1.2661421298980713,
+      "learning_rate": 6.875903365682429e-06,
+      "loss": 0.9775,
+      "step": 23045
+    },
+    {
+      "epoch": 0.9328207203561312,
+      "grad_norm": 1.3545794486999512,
+      "learning_rate": 6.855255007226925e-06,
+      "loss": 1.0137,
+      "step": 23050
+    },
+    {
+      "epoch": 0.9330230675839741,
+      "grad_norm": 1.1687790155410767,
+      "learning_rate": 6.834606648771423e-06,
+      "loss": 1.0448,
+      "step": 23055
+    },
+    {
+      "epoch": 0.9332254148118171,
+      "grad_norm": 1.1806179285049438,
+      "learning_rate": 6.813958290315921e-06,
+      "loss": 0.988,
+      "step": 23060
+    },
+    {
+      "epoch": 0.93342776203966,
+      "grad_norm": 1.2068326473236084,
+      "learning_rate": 6.793309931860417e-06,
+      "loss": 0.9895,
+      "step": 23065
+    },
+    {
+      "epoch": 0.933630109267503,
+      "grad_norm": 1.2086092233657837,
+      "learning_rate": 6.772661573404915e-06,
+      "loss": 1.0398,
+      "step": 23070
+    },
+    {
+      "epoch": 0.933832456495346,
+      "grad_norm": 1.2049978971481323,
+      "learning_rate": 6.752013214949413e-06,
+      "loss": 1.0056,
+      "step": 23075
+    },
+    {
+      "epoch": 0.934034803723189,
+      "grad_norm": 1.3573262691497803,
+      "learning_rate": 6.731364856493909e-06,
+      "loss": 0.9986,
+      "step": 23080
+    },
+    {
+      "epoch": 0.934237150951032,
+      "grad_norm": 1.2404375076293945,
+      "learning_rate": 6.7107164980384065e-06,
+      "loss": 0.9761,
+      "step": 23085
+    },
+    {
+      "epoch": 0.934439498178875,
+      "grad_norm": 1.2016634941101074,
+      "learning_rate": 6.690068139582903e-06,
+      "loss": 0.9445,
+      "step": 23090
+    },
+    {
+      "epoch": 0.9346418454067179,
+      "grad_norm": 1.174824833869934,
+      "learning_rate": 6.6694197811274e-06,
+      "loss": 0.9682,
+      "step": 23095
+    },
+    {
+      "epoch": 0.9348441926345609,
+      "grad_norm": 1.2408758401870728,
+      "learning_rate": 6.648771422671898e-06,
+      "loss": 0.9577,
+      "step": 23100
+    },
+    {
+      "epoch": 0.9350465398624039,
+      "grad_norm": 1.238100528717041,
+      "learning_rate": 6.628123064216395e-06,
+      "loss": 0.9912,
+      "step": 23105
+    },
+    {
+      "epoch": 0.9352488870902469,
+      "grad_norm": 1.2727497816085815,
+      "learning_rate": 6.607474705760892e-06,
+      "loss": 0.9427,
+      "step": 23110
+    },
+    {
+      "epoch": 0.9354512343180899,
+      "grad_norm": 1.2046585083007812,
+      "learning_rate": 6.58682634730539e-06,
+      "loss": 1.0214,
+      "step": 23115
+    },
+    {
+      "epoch": 0.9356535815459328,
+      "grad_norm": 1.1359328031539917,
+      "learning_rate": 6.566177988849887e-06,
+      "loss": 0.9797,
+      "step": 23120
+    },
+    {
+      "epoch": 0.9358559287737758,
+      "grad_norm": 1.1375129222869873,
+      "learning_rate": 6.545529630394384e-06,
+      "loss": 1.0091,
+      "step": 23125
+    },
+    {
+      "epoch": 0.9360582760016187,
+      "grad_norm": 1.1417800188064575,
+      "learning_rate": 6.5248812719388805e-06,
+      "loss": 0.9363,
+      "step": 23130
+    },
+    {
+      "epoch": 0.9362606232294618,
+      "grad_norm": 1.1177046298980713,
+      "learning_rate": 6.504232913483379e-06,
+      "loss": 1.0011,
+      "step": 23135
+    },
+    {
+      "epoch": 0.9364629704573048,
+      "grad_norm": 1.1646171808242798,
+      "learning_rate": 6.483584555027876e-06,
+      "loss": 1.0475,
+      "step": 23140
+    },
+    {
+      "epoch": 0.9366653176851477,
+      "grad_norm": 1.3040295839309692,
+      "learning_rate": 6.462936196572372e-06,
+      "loss": 0.9766,
+      "step": 23145
+    },
+    {
+      "epoch": 0.9368676649129907,
+      "grad_norm": 1.1640188694000244,
+      "learning_rate": 6.44228783811687e-06,
+      "loss": 1.0204,
+      "step": 23150
+    },
+    {
+      "epoch": 0.9370700121408336,
+      "grad_norm": 1.2561315298080444,
+      "learning_rate": 6.421639479661368e-06,
+      "loss": 0.9952,
+      "step": 23155
+    },
+    {
+      "epoch": 0.9372723593686767,
+      "grad_norm": 1.1376277208328247,
+      "learning_rate": 6.400991121205864e-06,
+      "loss": 1.002,
+      "step": 23160
+    },
+    {
+      "epoch": 0.9374747065965197,
+      "grad_norm": 1.2699038982391357,
+      "learning_rate": 6.3803427627503615e-06,
+      "loss": 1.0086,
+      "step": 23165
+    },
+    {
+      "epoch": 0.9376770538243626,
+      "grad_norm": 1.3435354232788086,
+      "learning_rate": 6.35969440429486e-06,
+      "loss": 0.9559,
+      "step": 23170
+    },
+    {
+      "epoch": 0.9378794010522056,
+      "grad_norm": 1.3385918140411377,
+      "learning_rate": 6.339046045839356e-06,
+      "loss": 0.9236,
+      "step": 23175
+    },
+    {
+      "epoch": 0.9380817482800485,
+      "grad_norm": 1.1412479877471924,
+      "learning_rate": 6.3183976873838534e-06,
+      "loss": 0.9457,
+      "step": 23180
+    },
+    {
+      "epoch": 0.9382840955078915,
+      "grad_norm": 1.2468445301055908,
+      "learning_rate": 6.29774932892835e-06,
+      "loss": 1.0548,
+      "step": 23185
+    },
+    {
+      "epoch": 0.9384864427357346,
+      "grad_norm": 1.169424057006836,
+      "learning_rate": 6.277100970472847e-06,
+      "loss": 0.9992,
+      "step": 23190
+    },
+    {
+      "epoch": 0.9386887899635775,
+      "grad_norm": 1.1182925701141357,
+      "learning_rate": 6.256452612017345e-06,
+      "loss": 0.965,
+      "step": 23195
+    },
+    {
+      "epoch": 0.9388911371914205,
+      "grad_norm": 1.1153651475906372,
+      "learning_rate": 6.235804253561843e-06,
+      "loss": 0.9414,
+      "step": 23200
+    },
+    {
+      "epoch": 0.9390934844192634,
+      "grad_norm": 1.1454898118972778,
+      "learning_rate": 6.215155895106339e-06,
+      "loss": 0.9536,
+      "step": 23205
+    },
+    {
+      "epoch": 0.9392958316471064,
+      "grad_norm": 1.1971384286880493,
+      "learning_rate": 6.194507536650836e-06,
+      "loss": 0.9632,
+      "step": 23210
+    },
+    {
+      "epoch": 0.9394981788749495,
+      "grad_norm": 1.2329403162002563,
+      "learning_rate": 6.173859178195334e-06,
+      "loss": 1.0392,
+      "step": 23215
+    },
+    {
+      "epoch": 0.9397005261027924,
+      "grad_norm": 1.113457202911377,
+      "learning_rate": 6.153210819739831e-06,
+      "loss": 0.9807,
+      "step": 23220
+    },
+    {
+      "epoch": 0.9399028733306354,
+      "grad_norm": 1.0697240829467773,
+      "learning_rate": 6.132562461284328e-06,
+      "loss": 0.9878,
+      "step": 23225
+    },
+    {
+      "epoch": 0.9401052205584783,
+      "grad_norm": 1.0905632972717285,
+      "learning_rate": 6.1119141028288255e-06,
+      "loss": 1.0008,
+      "step": 23230
+    },
+    {
+      "epoch": 0.9403075677863213,
+      "grad_norm": 1.2598730325698853,
+      "learning_rate": 6.091265744373323e-06,
+      "loss": 0.9747,
+      "step": 23235
+    },
+    {
+      "epoch": 0.9405099150141643,
+      "grad_norm": 1.3575570583343506,
+      "learning_rate": 6.07061738591782e-06,
+      "loss": 0.9928,
+      "step": 23240
+    },
+    {
+      "epoch": 0.9407122622420073,
+      "grad_norm": 1.1281064748764038,
+      "learning_rate": 6.049969027462317e-06,
+      "loss": 0.9636,
+      "step": 23245
+    },
+    {
+      "epoch": 0.9409146094698503,
+      "grad_norm": 1.2783358097076416,
+      "learning_rate": 6.029320669006814e-06,
+      "loss": 0.9372,
+      "step": 23250
+    },
+    {
+      "epoch": 0.9411169566976932,
+      "grad_norm": 1.2213220596313477,
+      "learning_rate": 6.008672310551312e-06,
+      "loss": 0.955,
+      "step": 23255
+    },
+    {
+      "epoch": 0.9413193039255362,
+      "grad_norm": 1.3591943979263306,
+      "learning_rate": 5.9880239520958085e-06,
+      "loss": 1.0201,
+      "step": 23260
+    },
+    {
+      "epoch": 0.9415216511533792,
+      "grad_norm": 1.3074491024017334,
+      "learning_rate": 5.967375593640306e-06,
+      "loss": 1.0069,
+      "step": 23265
+    },
+    {
+      "epoch": 0.9417239983812222,
+      "grad_norm": 1.256322979927063,
+      "learning_rate": 5.946727235184803e-06,
+      "loss": 0.9375,
+      "step": 23270
+    },
+    {
+      "epoch": 0.9419263456090652,
+      "grad_norm": 1.349203109741211,
+      "learning_rate": 5.9260788767293e-06,
+      "loss": 0.9614,
+      "step": 23275
+    },
+    {
+      "epoch": 0.9421286928369081,
+      "grad_norm": 1.1078602075576782,
+      "learning_rate": 5.905430518273798e-06,
+      "loss": 0.9919,
+      "step": 23280
+    },
+    {
+      "epoch": 0.9423310400647511,
+      "grad_norm": 1.2093766927719116,
+      "learning_rate": 5.884782159818294e-06,
+      "loss": 1.0046,
+      "step": 23285
+    },
+    {
+      "epoch": 0.9425333872925941,
+      "grad_norm": 1.3467715978622437,
+      "learning_rate": 5.864133801362792e-06,
+      "loss": 1.0882,
+      "step": 23290
+    },
+    {
+      "epoch": 0.942735734520437,
+      "grad_norm": 1.3225382566452026,
+      "learning_rate": 5.8434854429072896e-06,
+      "loss": 0.9484,
+      "step": 23295
+    },
+    {
+      "epoch": 0.9429380817482801,
+      "grad_norm": 1.210727334022522,
+      "learning_rate": 5.822837084451786e-06,
+      "loss": 1.0186,
+      "step": 23300
+    },
+    {
+      "epoch": 0.943140428976123,
+      "grad_norm": 1.1972028017044067,
+      "learning_rate": 5.802188725996283e-06,
+      "loss": 0.9711,
+      "step": 23305
+    },
+    {
+      "epoch": 0.943342776203966,
+      "grad_norm": 1.1847857236862183,
+      "learning_rate": 5.781540367540781e-06,
+      "loss": 1.015,
+      "step": 23310
+    },
+    {
+      "epoch": 0.943545123431809,
+      "grad_norm": 1.140752911567688,
+      "learning_rate": 5.760892009085278e-06,
+      "loss": 0.9548,
+      "step": 23315
+    },
+    {
+      "epoch": 0.9437474706596519,
+      "grad_norm": 1.1426345109939575,
+      "learning_rate": 5.740243650629775e-06,
+      "loss": 1.0281,
+      "step": 23320
+    },
+    {
+      "epoch": 0.943949817887495,
+      "grad_norm": 1.2390629053115845,
+      "learning_rate": 5.7195952921742725e-06,
+      "loss": 1.0057,
+      "step": 23325
+    },
+    {
+      "epoch": 0.944152165115338,
+      "grad_norm": 1.1874552965164185,
+      "learning_rate": 5.69894693371877e-06,
+      "loss": 1.0439,
+      "step": 23330
+    },
+    {
+      "epoch": 0.9443545123431809,
+      "grad_norm": 1.174991488456726,
+      "learning_rate": 5.678298575263267e-06,
+      "loss": 0.9842,
+      "step": 23335
+    },
+    {
+      "epoch": 0.9445568595710239,
+      "grad_norm": 1.2053190469741821,
+      "learning_rate": 5.6576502168077635e-06,
+      "loss": 0.995,
+      "step": 23340
+    },
+    {
+      "epoch": 0.9447592067988668,
+      "grad_norm": 1.2012766599655151,
+      "learning_rate": 5.637001858352262e-06,
+      "loss": 0.9801,
+      "step": 23345
+    },
+    {
+      "epoch": 0.9449615540267098,
+      "grad_norm": 1.151880145072937,
+      "learning_rate": 5.616353499896759e-06,
+      "loss": 0.9699,
+      "step": 23350
+    },
+    {
+      "epoch": 0.9451639012545529,
+      "grad_norm": 1.3854516744613647,
+      "learning_rate": 5.5957051414412554e-06,
+      "loss": 1.023,
+      "step": 23355
+    },
+    {
+      "epoch": 0.9453662484823958,
+      "grad_norm": 1.3277459144592285,
+      "learning_rate": 5.575056782985753e-06,
+      "loss": 0.9664,
+      "step": 23360
+    },
+    {
+      "epoch": 0.9455685957102388,
+      "grad_norm": 1.2124056816101074,
+      "learning_rate": 5.55440842453025e-06,
+      "loss": 0.9919,
+      "step": 23365
+    },
+    {
+      "epoch": 0.9457709429380817,
+      "grad_norm": 1.1614476442337036,
+      "learning_rate": 5.533760066074747e-06,
+      "loss": 0.9156,
+      "step": 23370
+    },
+    {
+      "epoch": 0.9459732901659247,
+      "grad_norm": 1.1627764701843262,
+      "learning_rate": 5.513111707619245e-06,
+      "loss": 1.0013,
+      "step": 23375
+    },
+    {
+      "epoch": 0.9461756373937678,
+      "grad_norm": 1.1513303518295288,
+      "learning_rate": 5.492463349163741e-06,
+      "loss": 0.9588,
+      "step": 23380
+    },
+    {
+      "epoch": 0.9463779846216107,
+      "grad_norm": 1.2399709224700928,
+      "learning_rate": 5.471814990708239e-06,
+      "loss": 1.0207,
+      "step": 23385
+    },
+    {
+      "epoch": 0.9465803318494537,
+      "grad_norm": 1.1727218627929688,
+      "learning_rate": 5.4511666322527365e-06,
+      "loss": 1.002,
+      "step": 23390
+    },
+    {
+      "epoch": 0.9467826790772966,
+      "grad_norm": 1.1629681587219238,
+      "learning_rate": 5.430518273797233e-06,
+      "loss": 0.9888,
+      "step": 23395
+    },
+    {
+      "epoch": 0.9469850263051396,
+      "grad_norm": 1.2181727886199951,
+      "learning_rate": 5.409869915341731e-06,
+      "loss": 0.9615,
+      "step": 23400
+    },
+    {
+      "epoch": 0.9471873735329825,
+      "grad_norm": 1.2515959739685059,
+      "learning_rate": 5.3892215568862275e-06,
+      "loss": 0.9624,
+      "step": 23405
+    },
+    {
+      "epoch": 0.9473897207608256,
+      "grad_norm": 1.210911750793457,
+      "learning_rate": 5.368573198430725e-06,
+      "loss": 1.018,
+      "step": 23410
+    },
+    {
+      "epoch": 0.9475920679886686,
+      "grad_norm": 1.2728792428970337,
+      "learning_rate": 5.347924839975222e-06,
+      "loss": 0.9122,
+      "step": 23415
+    },
+    {
+      "epoch": 0.9477944152165115,
+      "grad_norm": 1.1684070825576782,
+      "learning_rate": 5.3272764815197194e-06,
+      "loss": 0.9388,
+      "step": 23420
+    },
+    {
+      "epoch": 0.9479967624443545,
+      "grad_norm": 1.2218414545059204,
+      "learning_rate": 5.306628123064217e-06,
+      "loss": 0.9215,
+      "step": 23425
+    },
+    {
+      "epoch": 0.9481991096721974,
+      "grad_norm": 1.234809160232544,
+      "learning_rate": 5.285979764608714e-06,
+      "loss": 0.9451,
+      "step": 23430
+    },
+    {
+      "epoch": 0.9484014569000405,
+      "grad_norm": 1.2986630201339722,
+      "learning_rate": 5.2653314061532105e-06,
+      "loss": 0.9546,
+      "step": 23435
+    },
+    {
+      "epoch": 0.9486038041278835,
+      "grad_norm": 1.1890127658843994,
+      "learning_rate": 5.244683047697709e-06,
+      "loss": 0.9962,
+      "step": 23440
+    },
+    {
+      "epoch": 0.9488061513557264,
+      "grad_norm": 1.1800432205200195,
+      "learning_rate": 5.224034689242206e-06,
+      "loss": 0.9911,
+      "step": 23445
+    },
+    {
+      "epoch": 0.9490084985835694,
+      "grad_norm": 1.3987566232681274,
+      "learning_rate": 5.203386330786702e-06,
+      "loss": 0.9909,
+      "step": 23450
+    },
+    {
+      "epoch": 0.9492108458114124,
+      "grad_norm": 1.1060357093811035,
+      "learning_rate": 5.1827379723312e-06,
+      "loss": 0.9269,
+      "step": 23455
+    },
+    {
+      "epoch": 0.9494131930392553,
+      "grad_norm": 1.1613500118255615,
+      "learning_rate": 5.162089613875697e-06,
+      "loss": 0.9696,
+      "step": 23460
+    },
+    {
+      "epoch": 0.9496155402670984,
+      "grad_norm": 1.2415356636047363,
+      "learning_rate": 5.141441255420194e-06,
+      "loss": 0.9699,
+      "step": 23465
+    },
+    {
+      "epoch": 0.9498178874949413,
+      "grad_norm": 1.0619957447052002,
+      "learning_rate": 5.1207928969646916e-06,
+      "loss": 0.9851,
+      "step": 23470
+    },
+    {
+      "epoch": 0.9500202347227843,
+      "grad_norm": 1.276246428489685,
+      "learning_rate": 5.100144538509189e-06,
+      "loss": 1.0033,
+      "step": 23475
+    },
+    {
+      "epoch": 0.9502225819506273,
+      "grad_norm": 1.1619969606399536,
+      "learning_rate": 5.079496180053686e-06,
+      "loss": 0.9582,
+      "step": 23480
+    },
+    {
+      "epoch": 0.9504249291784702,
+      "grad_norm": 1.3111114501953125,
+      "learning_rate": 5.0588478215981835e-06,
+      "loss": 1.0037,
+      "step": 23485
+    },
+    {
+      "epoch": 0.9506272764063133,
+      "grad_norm": 1.2124131917953491,
+      "learning_rate": 5.03819946314268e-06,
+      "loss": 1.0154,
+      "step": 23490
+    },
+    {
+      "epoch": 0.9508296236341562,
+      "grad_norm": 1.3681647777557373,
+      "learning_rate": 5.017551104687178e-06,
+      "loss": 0.9945,
+      "step": 23495
+    },
+    {
+      "epoch": 0.9510319708619992,
+      "grad_norm": 1.2329882383346558,
+      "learning_rate": 4.9969027462316745e-06,
+      "loss": 0.9812,
+      "step": 23500
+    },
+    {
+      "epoch": 0.9512343180898422,
+      "grad_norm": 1.2132128477096558,
+      "learning_rate": 4.976254387776172e-06,
+      "loss": 0.9996,
+      "step": 23505
+    },
+    {
+      "epoch": 0.9514366653176851,
+      "grad_norm": 1.2256234884262085,
+      "learning_rate": 4.955606029320669e-06,
+      "loss": 0.979,
+      "step": 23510
+    },
+    {
+      "epoch": 0.9516390125455281,
+      "grad_norm": 1.2053040266036987,
+      "learning_rate": 4.934957670865166e-06,
+      "loss": 0.9884,
+      "step": 23515
+    },
+    {
+      "epoch": 0.9518413597733711,
+      "grad_norm": 1.1597155332565308,
+      "learning_rate": 4.914309312409664e-06,
+      "loss": 0.9326,
+      "step": 23520
+    },
+    {
+      "epoch": 0.9520437070012141,
+      "grad_norm": 1.3003026247024536,
+      "learning_rate": 4.893660953954161e-06,
+      "loss": 1.0425,
+      "step": 23525
+    },
+    {
+      "epoch": 0.9522460542290571,
+      "grad_norm": 1.2210971117019653,
+      "learning_rate": 4.873012595498658e-06,
+      "loss": 0.9526,
+      "step": 23530
+    },
+    {
+      "epoch": 0.9524484014569,
+      "grad_norm": 1.221471905708313,
+      "learning_rate": 4.8523642370431556e-06,
+      "loss": 0.9584,
+      "step": 23535
+    },
+    {
+      "epoch": 0.952650748684743,
+      "grad_norm": 1.158247470855713,
+      "learning_rate": 4.831715878587653e-06,
+      "loss": 1.0348,
+      "step": 23540
+    },
+    {
+      "epoch": 0.952853095912586,
+      "grad_norm": 1.1828258037567139,
+      "learning_rate": 4.811067520132149e-06,
+      "loss": 0.9556,
+      "step": 23545
+    },
+    {
+      "epoch": 0.953055443140429,
+      "grad_norm": 1.292490005493164,
+      "learning_rate": 4.7904191616766475e-06,
+      "loss": 1.0561,
+      "step": 23550
+    },
+    {
+      "epoch": 0.953257790368272,
+      "grad_norm": 1.8282053470611572,
+      "learning_rate": 4.769770803221144e-06,
+      "loss": 0.9852,
+      "step": 23555
+    },
+    {
+      "epoch": 0.9534601375961149,
+      "grad_norm": 1.2209551334381104,
+      "learning_rate": 4.749122444765641e-06,
+      "loss": 0.9675,
+      "step": 23560
+    },
+    {
+      "epoch": 0.9536624848239579,
+      "grad_norm": 1.193122148513794,
+      "learning_rate": 4.7284740863101385e-06,
+      "loss": 0.9932,
+      "step": 23565
+    },
+    {
+      "epoch": 0.9538648320518008,
+      "grad_norm": 1.1373188495635986,
+      "learning_rate": 4.707825727854636e-06,
+      "loss": 0.9883,
+      "step": 23570
+    },
+    {
+      "epoch": 0.9540671792796439,
+      "grad_norm": 1.124559760093689,
+      "learning_rate": 4.687177369399133e-06,
+      "loss": 0.9834,
+      "step": 23575
+    },
+    {
+      "epoch": 0.9542695265074869,
+      "grad_norm": 1.2636216878890991,
+      "learning_rate": 4.66652901094363e-06,
+      "loss": 0.9846,
+      "step": 23580
+    },
+    {
+      "epoch": 0.9544718737353298,
+      "grad_norm": 1.1844679117202759,
+      "learning_rate": 4.645880652488127e-06,
+      "loss": 1.0643,
+      "step": 23585
+    },
+    {
+      "epoch": 0.9546742209631728,
+      "grad_norm": 1.1882669925689697,
+      "learning_rate": 4.625232294032625e-06,
+      "loss": 1.0647,
+      "step": 23590
+    },
+    {
+      "epoch": 0.9548765681910157,
+      "grad_norm": 1.2093403339385986,
+      "learning_rate": 4.6045839355771214e-06,
+      "loss": 0.9515,
+      "step": 23595
+    },
+    {
+      "epoch": 0.9550789154188588,
+      "grad_norm": 1.1817821264266968,
+      "learning_rate": 4.583935577121619e-06,
+      "loss": 1.02,
+      "step": 23600
+    },
+    {
+      "epoch": 0.9552812626467018,
+      "grad_norm": 1.1955592632293701,
+      "learning_rate": 4.563287218666117e-06,
+      "loss": 1.0052,
+      "step": 23605
+    },
+    {
+      "epoch": 0.9554836098745447,
+      "grad_norm": 1.329101800918579,
+      "learning_rate": 4.542638860210613e-06,
+      "loss": 1.0078,
+      "step": 23610
+    },
+    {
+      "epoch": 0.9556859571023877,
+      "grad_norm": 1.2466204166412354,
+      "learning_rate": 4.521990501755111e-06,
+      "loss": 0.9645,
+      "step": 23615
+    },
+    {
+      "epoch": 0.9558883043302306,
+      "grad_norm": 1.1778624057769775,
+      "learning_rate": 4.501342143299608e-06,
+      "loss": 0.9081,
+      "step": 23620
+    },
+    {
+      "epoch": 0.9560906515580736,
+      "grad_norm": 1.2266567945480347,
+      "learning_rate": 4.480693784844105e-06,
+      "loss": 0.9674,
+      "step": 23625
+    },
+    {
+      "epoch": 0.9562929987859167,
+      "grad_norm": 1.1299916505813599,
+      "learning_rate": 4.4600454263886025e-06,
+      "loss": 0.9353,
+      "step": 23630
+    },
+    {
+      "epoch": 0.9564953460137596,
+      "grad_norm": 1.2154983282089233,
+      "learning_rate": 4.4393970679331e-06,
+      "loss": 0.9808,
+      "step": 23635
+    },
+    {
+      "epoch": 0.9566976932416026,
+      "grad_norm": 1.2375404834747314,
+      "learning_rate": 4.418748709477596e-06,
+      "loss": 1.0393,
+      "step": 23640
+    },
+    {
+      "epoch": 0.9569000404694455,
+      "grad_norm": 1.2174794673919678,
+      "learning_rate": 4.398100351022094e-06,
+      "loss": 1.0399,
+      "step": 23645
+    },
+    {
+      "epoch": 0.9571023876972885,
+      "grad_norm": 1.078600525856018,
+      "learning_rate": 4.377451992566591e-06,
+      "loss": 0.9643,
+      "step": 23650
+    },
+    {
+      "epoch": 0.9573047349251316,
+      "grad_norm": 1.3520904779434204,
+      "learning_rate": 4.356803634111088e-06,
+      "loss": 0.9751,
+      "step": 23655
+    },
+    {
+      "epoch": 0.9575070821529745,
+      "grad_norm": 1.3326326608657837,
+      "learning_rate": 4.3361552756555855e-06,
+      "loss": 0.9358,
+      "step": 23660
+    },
+    {
+      "epoch": 0.9577094293808175,
+      "grad_norm": 1.2878941297531128,
+      "learning_rate": 4.315506917200083e-06,
+      "loss": 0.9768,
+      "step": 23665
+    },
+    {
+      "epoch": 0.9579117766086604,
+      "grad_norm": 1.404470682144165,
+      "learning_rate": 4.29485855874458e-06,
+      "loss": 1.011,
+      "step": 23670
+    },
+    {
+      "epoch": 0.9581141238365034,
+      "grad_norm": 1.32426917552948,
+      "learning_rate": 4.274210200289077e-06,
+      "loss": 0.9365,
+      "step": 23675
+    },
+    {
+      "epoch": 0.9583164710643464,
+      "grad_norm": 1.253981590270996,
+      "learning_rate": 4.253561841833575e-06,
+      "loss": 0.9478,
+      "step": 23680
+    },
+    {
+      "epoch": 0.9585188182921894,
+      "grad_norm": 1.2419568300247192,
+      "learning_rate": 4.232913483378072e-06,
+      "loss": 1.0353,
+      "step": 23685
+    },
+    {
+      "epoch": 0.9587211655200324,
+      "grad_norm": 1.22864830493927,
+      "learning_rate": 4.212265124922568e-06,
+      "loss": 0.9522,
+      "step": 23690
+    },
+    {
+      "epoch": 0.9589235127478754,
+      "grad_norm": 1.2241621017456055,
+      "learning_rate": 4.191616766467066e-06,
+      "loss": 0.955,
+      "step": 23695
+    },
+    {
+      "epoch": 0.9591258599757183,
+      "grad_norm": 1.1657403707504272,
+      "learning_rate": 4.170968408011564e-06,
+      "loss": 1.0208,
+      "step": 23700
+    },
+    {
+      "epoch": 0.9593282072035613,
+      "grad_norm": 1.2061777114868164,
+      "learning_rate": 4.15032004955606e-06,
+      "loss": 0.9613,
+      "step": 23705
+    },
+    {
+      "epoch": 0.9595305544314043,
+      "grad_norm": 1.3198875188827515,
+      "learning_rate": 4.1296716911005576e-06,
+      "loss": 0.9877,
+      "step": 23710
+    },
+    {
+      "epoch": 0.9597329016592473,
+      "grad_norm": 1.3837649822235107,
+      "learning_rate": 4.109023332645055e-06,
+      "loss": 1.0236,
+      "step": 23715
+    },
+    {
+      "epoch": 0.9599352488870903,
+      "grad_norm": 1.0314319133758545,
+      "learning_rate": 4.088374974189552e-06,
+      "loss": 0.9813,
+      "step": 23720
+    },
+    {
+      "epoch": 0.9601375961149332,
+      "grad_norm": 1.3057734966278076,
+      "learning_rate": 4.0677266157340495e-06,
+      "loss": 0.9563,
+      "step": 23725
+    },
+    {
+      "epoch": 0.9603399433427762,
+      "grad_norm": 1.2327723503112793,
+      "learning_rate": 4.047078257278547e-06,
+      "loss": 0.9636,
+      "step": 23730
+    },
+    {
+      "epoch": 0.9605422905706191,
+      "grad_norm": 1.1845272779464722,
+      "learning_rate": 4.026429898823044e-06,
+      "loss": 1.0171,
+      "step": 23735
+    },
+    {
+      "epoch": 0.9607446377984622,
+      "grad_norm": 1.091482162475586,
+      "learning_rate": 4.005781540367541e-06,
+      "loss": 1.0038,
+      "step": 23740
+    },
+    {
+      "epoch": 0.9609469850263052,
+      "grad_norm": 1.1078810691833496,
+      "learning_rate": 3.985133181912038e-06,
+      "loss": 0.9537,
+      "step": 23745
+    },
+    {
+      "epoch": 0.9611493322541481,
+      "grad_norm": 1.3094723224639893,
+      "learning_rate": 3.964484823456535e-06,
+      "loss": 0.9207,
+      "step": 23750
+    },
+    {
+      "epoch": 0.9613516794819911,
+      "grad_norm": 1.218976616859436,
+      "learning_rate": 3.943836465001033e-06,
+      "loss": 0.9726,
+      "step": 23755
+    },
+    {
+      "epoch": 0.961554026709834,
+      "grad_norm": 1.2138930559158325,
+      "learning_rate": 3.92318810654553e-06,
+      "loss": 1.0626,
+      "step": 23760
+    },
+    {
+      "epoch": 0.9617563739376771,
+      "grad_norm": 1.143302321434021,
+      "learning_rate": 3.902539748090027e-06,
+      "loss": 0.9807,
+      "step": 23765
+    },
+    {
+      "epoch": 0.9619587211655201,
+      "grad_norm": 1.1158267259597778,
+      "learning_rate": 3.881891389634524e-06,
+      "loss": 0.9533,
+      "step": 23770
+    },
+    {
+      "epoch": 0.962161068393363,
+      "grad_norm": 1.12842857837677,
+      "learning_rate": 3.861243031179022e-06,
+      "loss": 0.9658,
+      "step": 23775
+    },
+    {
+      "epoch": 0.962363415621206,
+      "grad_norm": 1.2312090396881104,
+      "learning_rate": 3.840594672723519e-06,
+      "loss": 0.9806,
+      "step": 23780
+    },
+    {
+      "epoch": 0.9625657628490489,
+      "grad_norm": 1.1958683729171753,
+      "learning_rate": 3.819946314268015e-06,
+      "loss": 0.9881,
+      "step": 23785
+    },
+    {
+      "epoch": 0.9627681100768919,
+      "grad_norm": 1.2550935745239258,
+      "learning_rate": 3.7992979558125126e-06,
+      "loss": 1.006,
+      "step": 23790
+    },
+    {
+      "epoch": 0.962970457304735,
+      "grad_norm": 1.1829735040664673,
+      "learning_rate": 3.7786495973570103e-06,
+      "loss": 1.0019,
+      "step": 23795
+    },
+    {
+      "epoch": 0.9631728045325779,
+      "grad_norm": 1.201974630355835,
+      "learning_rate": 3.7580012389015076e-06,
+      "loss": 0.9364,
+      "step": 23800
+    },
+    {
+      "epoch": 0.9633751517604209,
+      "grad_norm": 1.1843925714492798,
+      "learning_rate": 3.7373528804460045e-06,
+      "loss": 0.9625,
+      "step": 23805
+    },
+    {
+      "epoch": 0.9635774989882638,
+      "grad_norm": 1.2058708667755127,
+      "learning_rate": 3.7167045219905022e-06,
+      "loss": 0.9961,
+      "step": 23810
+    },
+    {
+      "epoch": 0.9637798462161068,
+      "grad_norm": 1.3171906471252441,
+      "learning_rate": 3.696056163534999e-06,
+      "loss": 1.0389,
+      "step": 23815
+    },
+    {
+      "epoch": 0.9639821934439499,
+      "grad_norm": 1.1943846940994263,
+      "learning_rate": 3.6754078050794964e-06,
+      "loss": 1.0027,
+      "step": 23820
+    },
+    {
+      "epoch": 0.9641845406717928,
+      "grad_norm": 1.185172438621521,
+      "learning_rate": 3.6547594466239933e-06,
+      "loss": 0.9723,
+      "step": 23825
+    },
+    {
+      "epoch": 0.9643868878996358,
+      "grad_norm": 1.1887675523757935,
+      "learning_rate": 3.634111088168491e-06,
+      "loss": 0.9918,
+      "step": 23830
+    },
+    {
+      "epoch": 0.9645892351274787,
+      "grad_norm": 1.2175261974334717,
+      "learning_rate": 3.613462729712988e-06,
+      "loss": 0.9914,
+      "step": 23835
+    },
+    {
+      "epoch": 0.9647915823553217,
+      "grad_norm": 1.08931303024292,
+      "learning_rate": 3.592814371257485e-06,
+      "loss": 0.9862,
+      "step": 23840
+    },
+    {
+      "epoch": 0.9649939295831647,
+      "grad_norm": 1.2798324823379517,
+      "learning_rate": 3.572166012801982e-06,
+      "loss": 0.979,
+      "step": 23845
+    },
+    {
+      "epoch": 0.9651962768110077,
+      "grad_norm": 1.1880285739898682,
+      "learning_rate": 3.5515176543464798e-06,
+      "loss": 0.9834,
+      "step": 23850
+    },
+    {
+      "epoch": 0.9653986240388507,
+      "grad_norm": 1.332466959953308,
+      "learning_rate": 3.5308692958909766e-06,
+      "loss": 1.019,
+      "step": 23855
+    },
+    {
+      "epoch": 0.9656009712666936,
+      "grad_norm": 1.1690194606781006,
+      "learning_rate": 3.510220937435474e-06,
+      "loss": 1.0431,
+      "step": 23860
+    },
+    {
+      "epoch": 0.9658033184945366,
+      "grad_norm": 1.1507174968719482,
+      "learning_rate": 3.4895725789799717e-06,
+      "loss": 1.0088,
+      "step": 23865
+    },
+    {
+      "epoch": 0.9660056657223796,
+      "grad_norm": 1.1203949451446533,
+      "learning_rate": 3.4689242205244685e-06,
+      "loss": 0.9703,
+      "step": 23870
+    },
+    {
+      "epoch": 0.9662080129502226,
+      "grad_norm": 1.2323098182678223,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 1.0264,
+      "step": 23875
+    },
+    {
+      "epoch": 0.9664103601780656,
+      "grad_norm": 1.2957608699798584,
+      "learning_rate": 3.4276275036134627e-06,
+      "loss": 0.9298,
+      "step": 23880
+    },
+    {
+      "epoch": 0.9666127074059085,
+      "grad_norm": 1.2158442735671997,
+      "learning_rate": 3.4069791451579604e-06,
+      "loss": 0.971,
+      "step": 23885
+    },
+    {
+      "epoch": 0.9668150546337515,
+      "grad_norm": 1.1678370237350464,
+      "learning_rate": 3.3863307867024573e-06,
+      "loss": 1.0172,
+      "step": 23890
+    },
+    {
+      "epoch": 0.9670174018615945,
+      "grad_norm": 1.0520416498184204,
+      "learning_rate": 3.3656824282469546e-06,
+      "loss": 1.0031,
+      "step": 23895
+    },
+    {
+      "epoch": 0.9672197490894374,
+      "grad_norm": 1.2183130979537964,
+      "learning_rate": 3.3450340697914515e-06,
+      "loss": 0.9367,
+      "step": 23900
+    },
+    {
+      "epoch": 0.9674220963172805,
+      "grad_norm": 1.152599811553955,
+      "learning_rate": 3.324385711335949e-06,
+      "loss": 0.9871,
+      "step": 23905
+    },
+    {
+      "epoch": 0.9676244435451234,
+      "grad_norm": 1.1636003255844116,
+      "learning_rate": 3.303737352880446e-06,
+      "loss": 1.0386,
+      "step": 23910
+    },
+    {
+      "epoch": 0.9678267907729664,
+      "grad_norm": 1.1853296756744385,
+      "learning_rate": 3.2830889944249434e-06,
+      "loss": 1.0082,
+      "step": 23915
+    },
+    {
+      "epoch": 0.9680291380008094,
+      "grad_norm": 1.2935311794281006,
+      "learning_rate": 3.2624406359694402e-06,
+      "loss": 1.0014,
+      "step": 23920
+    },
+    {
+      "epoch": 0.9682314852286523,
+      "grad_norm": 1.0872714519500732,
+      "learning_rate": 3.241792277513938e-06,
+      "loss": 1.0006,
+      "step": 23925
+    },
+    {
+      "epoch": 0.9684338324564954,
+      "grad_norm": 1.1798073053359985,
+      "learning_rate": 3.221143919058435e-06,
+      "loss": 1.0605,
+      "step": 23930
+    },
+    {
+      "epoch": 0.9686361796843383,
+      "grad_norm": 1.2358653545379639,
+      "learning_rate": 3.200495560602932e-06,
+      "loss": 0.9557,
+      "step": 23935
+    },
+    {
+      "epoch": 0.9688385269121813,
+      "grad_norm": 1.3411706686019897,
+      "learning_rate": 3.17984720214743e-06,
+      "loss": 1.0711,
+      "step": 23940
+    },
+    {
+      "epoch": 0.9690408741400243,
+      "grad_norm": 1.0622667074203491,
+      "learning_rate": 3.1591988436919267e-06,
+      "loss": 1.0514,
+      "step": 23945
+    },
+    {
+      "epoch": 0.9692432213678672,
+      "grad_norm": 1.4040849208831787,
+      "learning_rate": 3.1385504852364236e-06,
+      "loss": 0.9848,
+      "step": 23950
+    },
+    {
+      "epoch": 0.9694455685957103,
+      "grad_norm": 1.2281724214553833,
+      "learning_rate": 3.1179021267809213e-06,
+      "loss": 1.0132,
+      "step": 23955
+    },
+    {
+      "epoch": 0.9696479158235533,
+      "grad_norm": 1.2061702013015747,
+      "learning_rate": 3.097253768325418e-06,
+      "loss": 0.9329,
+      "step": 23960
+    },
+    {
+      "epoch": 0.9698502630513962,
+      "grad_norm": 1.2600946426391602,
+      "learning_rate": 3.0766054098699155e-06,
+      "loss": 1.011,
+      "step": 23965
+    },
+    {
+      "epoch": 0.9700526102792392,
+      "grad_norm": 1.2559062242507935,
+      "learning_rate": 3.0559570514144128e-06,
+      "loss": 0.9967,
+      "step": 23970
+    },
+    {
+      "epoch": 0.9702549575070821,
+      "grad_norm": 1.216749906539917,
+      "learning_rate": 3.03530869295891e-06,
+      "loss": 1.0337,
+      "step": 23975
+    },
+    {
+      "epoch": 0.9704573047349251,
+      "grad_norm": 1.1401817798614502,
+      "learning_rate": 3.014660334503407e-06,
+      "loss": 0.9632,
+      "step": 23980
+    },
+    {
+      "epoch": 0.9706596519627682,
+      "grad_norm": 1.1647950410842896,
+      "learning_rate": 2.9940119760479042e-06,
+      "loss": 0.9736,
+      "step": 23985
+    },
+    {
+      "epoch": 0.9708619991906111,
+      "grad_norm": 1.3680803775787354,
+      "learning_rate": 2.9733636175924015e-06,
+      "loss": 1.0719,
+      "step": 23990
+    },
+    {
+      "epoch": 0.9710643464184541,
+      "grad_norm": 1.322601079940796,
+      "learning_rate": 2.952715259136899e-06,
+      "loss": 0.9928,
+      "step": 23995
+    },
+    {
+      "epoch": 0.971266693646297,
+      "grad_norm": 1.2319669723510742,
+      "learning_rate": 2.932066900681396e-06,
+      "loss": 0.959,
+      "step": 24000
+    },
+    {
+      "epoch": 0.97146904087414,
+      "grad_norm": 1.271428108215332,
+      "learning_rate": 2.911418542225893e-06,
+      "loss": 0.9896,
+      "step": 24005
+    },
+    {
+      "epoch": 0.9716713881019831,
+      "grad_norm": 1.088760256767273,
+      "learning_rate": 2.8907701837703903e-06,
+      "loss": 0.9785,
+      "step": 24010
+    },
+    {
+      "epoch": 0.971873735329826,
+      "grad_norm": 1.2278944253921509,
+      "learning_rate": 2.8701218253148876e-06,
+      "loss": 0.9492,
+      "step": 24015
+    },
+    {
+      "epoch": 0.972076082557669,
+      "grad_norm": 1.3369373083114624,
+      "learning_rate": 2.849473466859385e-06,
+      "loss": 1.0074,
+      "step": 24020
+    },
+    {
+      "epoch": 0.9722784297855119,
+      "grad_norm": 1.1910151243209839,
+      "learning_rate": 2.8288251084038818e-06,
+      "loss": 0.938,
+      "step": 24025
+    },
+    {
+      "epoch": 0.9724807770133549,
+      "grad_norm": 1.1880441904067993,
+      "learning_rate": 2.8081767499483795e-06,
+      "loss": 1.0394,
+      "step": 24030
+    },
+    {
+      "epoch": 0.9726831242411978,
+      "grad_norm": 1.1743971109390259,
+      "learning_rate": 2.7875283914928764e-06,
+      "loss": 1.0416,
+      "step": 24035
+    },
+    {
+      "epoch": 0.9728854714690409,
+      "grad_norm": 1.2691752910614014,
+      "learning_rate": 2.7668800330373737e-06,
+      "loss": 0.9933,
+      "step": 24040
+    },
+    {
+      "epoch": 0.9730878186968839,
+      "grad_norm": 1.296424388885498,
+      "learning_rate": 2.7462316745818705e-06,
+      "loss": 1.0496,
+      "step": 24045
+    },
+    {
+      "epoch": 0.9732901659247268,
+      "grad_norm": 1.0720136165618896,
+      "learning_rate": 2.7255833161263683e-06,
+      "loss": 0.9824,
+      "step": 24050
+    },
+    {
+      "epoch": 0.9734925131525698,
+      "grad_norm": 1.2150864601135254,
+      "learning_rate": 2.7049349576708655e-06,
+      "loss": 0.9835,
+      "step": 24055
+    },
+    {
+      "epoch": 0.9736948603804128,
+      "grad_norm": 1.1679117679595947,
+      "learning_rate": 2.6842865992153624e-06,
+      "loss": 1.0128,
+      "step": 24060
+    },
+    {
+      "epoch": 0.9738972076082558,
+      "grad_norm": 1.2796560525894165,
+      "learning_rate": 2.6636382407598597e-06,
+      "loss": 0.9919,
+      "step": 24065
+    },
+    {
+      "epoch": 0.9740995548360988,
+      "grad_norm": 1.163400411605835,
+      "learning_rate": 2.642989882304357e-06,
+      "loss": 0.9694,
+      "step": 24070
+    },
+    {
+      "epoch": 0.9743019020639417,
+      "grad_norm": 1.3392345905303955,
+      "learning_rate": 2.6223415238488543e-06,
+      "loss": 0.9879,
+      "step": 24075
+    },
+    {
+      "epoch": 0.9745042492917847,
+      "grad_norm": 2.100224018096924,
+      "learning_rate": 2.601693165393351e-06,
+      "loss": 1.0275,
+      "step": 24080
+    },
+    {
+      "epoch": 0.9747065965196277,
+      "grad_norm": 1.163314938545227,
+      "learning_rate": 2.5810448069378485e-06,
+      "loss": 1.007,
+      "step": 24085
+    },
+    {
+      "epoch": 0.9749089437474706,
+      "grad_norm": 1.1846504211425781,
+      "learning_rate": 2.5603964484823458e-06,
+      "loss": 0.9755,
+      "step": 24090
+    },
+    {
+      "epoch": 0.9751112909753137,
+      "grad_norm": 1.2549328804016113,
+      "learning_rate": 2.539748090026843e-06,
+      "loss": 0.9705,
+      "step": 24095
+    },
+    {
+      "epoch": 0.9753136382031566,
+      "grad_norm": 1.228094220161438,
+      "learning_rate": 2.51909973157134e-06,
+      "loss": 1.0204,
+      "step": 24100
+    },
+    {
+      "epoch": 0.9755159854309996,
+      "grad_norm": 1.2853307723999023,
+      "learning_rate": 2.4984513731158372e-06,
+      "loss": 0.9794,
+      "step": 24105
+    },
+    {
+      "epoch": 0.9757183326588426,
+      "grad_norm": 1.2282099723815918,
+      "learning_rate": 2.4778030146603345e-06,
+      "loss": 1.059,
+      "step": 24110
+    },
+    {
+      "epoch": 0.9759206798866855,
+      "grad_norm": 1.2545039653778076,
+      "learning_rate": 2.457154656204832e-06,
+      "loss": 0.9809,
+      "step": 24115
+    },
+    {
+      "epoch": 0.9761230271145286,
+      "grad_norm": 1.141188383102417,
+      "learning_rate": 2.436506297749329e-06,
+      "loss": 0.9898,
+      "step": 24120
+    },
+    {
+      "epoch": 0.9763253743423715,
+      "grad_norm": 1.217935562133789,
+      "learning_rate": 2.4158579392938264e-06,
+      "loss": 1.0491,
+      "step": 24125
+    },
+    {
+      "epoch": 0.9765277215702145,
+      "grad_norm": 1.1647645235061646,
+      "learning_rate": 2.3952095808383237e-06,
+      "loss": 1.0066,
+      "step": 24130
+    },
+    {
+      "epoch": 0.9767300687980575,
+      "grad_norm": 1.2348576784133911,
+      "learning_rate": 2.3745612223828206e-06,
+      "loss": 1.058,
+      "step": 24135
+    },
+    {
+      "epoch": 0.9769324160259004,
+      "grad_norm": 1.3031387329101562,
+      "learning_rate": 2.353912863927318e-06,
+      "loss": 0.9852,
+      "step": 24140
+    },
+    {
+      "epoch": 0.9771347632537434,
+      "grad_norm": 1.1181423664093018,
+      "learning_rate": 2.333264505471815e-06,
+      "loss": 0.943,
+      "step": 24145
+    },
+    {
+      "epoch": 0.9773371104815864,
+      "grad_norm": 1.23284113407135,
+      "learning_rate": 2.3126161470163125e-06,
+      "loss": 0.989,
+      "step": 24150
+    },
+    {
+      "epoch": 0.9775394577094294,
+      "grad_norm": 1.132831335067749,
+      "learning_rate": 2.2919677885608094e-06,
+      "loss": 0.9345,
+      "step": 24155
+    },
+    {
+      "epoch": 0.9777418049372724,
+      "grad_norm": 1.279350757598877,
+      "learning_rate": 2.2713194301053067e-06,
+      "loss": 0.9246,
+      "step": 24160
+    },
+    {
+      "epoch": 0.9779441521651153,
+      "grad_norm": 1.315624475479126,
+      "learning_rate": 2.250671071649804e-06,
+      "loss": 1.0077,
+      "step": 24165
+    },
+    {
+      "epoch": 0.9781464993929583,
+      "grad_norm": 1.2561452388763428,
+      "learning_rate": 2.2300227131943013e-06,
+      "loss": 1.0589,
+      "step": 24170
+    },
+    {
+      "epoch": 0.9783488466208013,
+      "grad_norm": 1.9925191402435303,
+      "learning_rate": 2.209374354738798e-06,
+      "loss": 1.0016,
+      "step": 24175
+    },
+    {
+      "epoch": 0.9785511938486443,
+      "grad_norm": 1.1933507919311523,
+      "learning_rate": 2.1887259962832954e-06,
+      "loss": 0.9412,
+      "step": 24180
+    },
+    {
+      "epoch": 0.9787535410764873,
+      "grad_norm": 1.336834192276001,
+      "learning_rate": 2.1680776378277927e-06,
+      "loss": 0.9794,
+      "step": 24185
+    },
+    {
+      "epoch": 0.9789558883043302,
+      "grad_norm": 1.2687063217163086,
+      "learning_rate": 2.14742927937229e-06,
+      "loss": 1.0073,
+      "step": 24190
+    },
+    {
+      "epoch": 0.9791582355321732,
+      "grad_norm": 1.1436597108840942,
+      "learning_rate": 2.1267809209167873e-06,
+      "loss": 0.9973,
+      "step": 24195
+    },
+    {
+      "epoch": 0.9793605827600161,
+      "grad_norm": 1.2121599912643433,
+      "learning_rate": 2.106132562461284e-06,
+      "loss": 0.9389,
+      "step": 24200
+    },
+    {
+      "epoch": 0.9795629299878592,
+      "grad_norm": 1.1339515447616577,
+      "learning_rate": 2.085484204005782e-06,
+      "loss": 0.974,
+      "step": 24205
+    },
+    {
+      "epoch": 0.9797652772157022,
+      "grad_norm": 1.2546398639678955,
+      "learning_rate": 2.0648358455502788e-06,
+      "loss": 0.99,
+      "step": 24210
+    },
+    {
+      "epoch": 0.9799676244435451,
+      "grad_norm": 1.3780025243759155,
+      "learning_rate": 2.044187487094776e-06,
+      "loss": 1.0159,
+      "step": 24215
+    },
+    {
+      "epoch": 0.9801699716713881,
+      "grad_norm": 1.1643260717391968,
+      "learning_rate": 2.0235391286392734e-06,
+      "loss": 0.9544,
+      "step": 24220
+    },
+    {
+      "epoch": 0.980372318899231,
+      "grad_norm": 1.1889039278030396,
+      "learning_rate": 2.0028907701837707e-06,
+      "loss": 0.992,
+      "step": 24225
+    },
+    {
+      "epoch": 0.9805746661270741,
+      "grad_norm": 1.2637370824813843,
+      "learning_rate": 1.9822424117282676e-06,
+      "loss": 0.9729,
+      "step": 24230
+    },
+    {
+      "epoch": 0.9807770133549171,
+      "grad_norm": 1.2505862712860107,
+      "learning_rate": 1.961594053272765e-06,
+      "loss": 1.0078,
+      "step": 24235
+    },
+    {
+      "epoch": 0.98097936058276,
+      "grad_norm": 1.205688714981079,
+      "learning_rate": 1.940945694817262e-06,
+      "loss": 0.9961,
+      "step": 24240
+    },
+    {
+      "epoch": 0.981181707810603,
+      "grad_norm": 1.193853497505188,
+      "learning_rate": 1.9202973363617594e-06,
+      "loss": 0.9857,
+      "step": 24245
+    },
+    {
+      "epoch": 0.981384055038446,
+      "grad_norm": 1.2576864957809448,
+      "learning_rate": 1.8996489779062563e-06,
+      "loss": 1.0902,
+      "step": 24250
+    },
+    {
+      "epoch": 0.9815864022662889,
+      "grad_norm": 1.190581202507019,
+      "learning_rate": 1.8790006194507538e-06,
+      "loss": 0.9813,
+      "step": 24255
+    },
+    {
+      "epoch": 0.981788749494132,
+      "grad_norm": 1.1921381950378418,
+      "learning_rate": 1.8583522609952511e-06,
+      "loss": 0.957,
+      "step": 24260
+    },
+    {
+      "epoch": 0.9819910967219749,
+      "grad_norm": 1.3175832033157349,
+      "learning_rate": 1.8377039025397482e-06,
+      "loss": 0.9555,
+      "step": 24265
+    },
+    {
+      "epoch": 0.9821934439498179,
+      "grad_norm": 1.2868329286575317,
+      "learning_rate": 1.8170555440842455e-06,
+      "loss": 1.0136,
+      "step": 24270
+    },
+    {
+      "epoch": 0.9823957911776608,
+      "grad_norm": 1.213088870048523,
+      "learning_rate": 1.7964071856287426e-06,
+      "loss": 0.9871,
+      "step": 24275
+    },
+    {
+      "epoch": 0.9825981384055038,
+      "grad_norm": 1.3250539302825928,
+      "learning_rate": 1.7757588271732399e-06,
+      "loss": 0.96,
+      "step": 24280
+    },
+    {
+      "epoch": 0.9828004856333469,
+      "grad_norm": 1.2626069784164429,
+      "learning_rate": 1.755110468717737e-06,
+      "loss": 0.9576,
+      "step": 24285
+    },
+    {
+      "epoch": 0.9830028328611898,
+      "grad_norm": 1.156372308731079,
+      "learning_rate": 1.7344621102622343e-06,
+      "loss": 0.9399,
+      "step": 24290
+    },
+    {
+      "epoch": 0.9832051800890328,
+      "grad_norm": 1.209682822227478,
+      "learning_rate": 1.7138137518067314e-06,
+      "loss": 0.9698,
+      "step": 24295
+    },
+    {
+      "epoch": 0.9834075273168758,
+      "grad_norm": 1.3231494426727295,
+      "learning_rate": 1.6931653933512286e-06,
+      "loss": 0.9901,
+      "step": 24300
+    },
+    {
+      "epoch": 0.9836098745447187,
+      "grad_norm": 1.1929429769515991,
+      "learning_rate": 1.6725170348957257e-06,
+      "loss": 0.99,
+      "step": 24305
+    },
+    {
+      "epoch": 0.9838122217725617,
+      "grad_norm": 1.1918138265609741,
+      "learning_rate": 1.651868676440223e-06,
+      "loss": 1.035,
+      "step": 24310
+    },
+    {
+      "epoch": 0.9840145690004047,
+      "grad_norm": 1.1524841785430908,
+      "learning_rate": 1.6312203179847201e-06,
+      "loss": 0.9669,
+      "step": 24315
+    },
+    {
+      "epoch": 0.9842169162282477,
+      "grad_norm": 1.2520349025726318,
+      "learning_rate": 1.6105719595292174e-06,
+      "loss": 0.9889,
+      "step": 24320
+    },
+    {
+      "epoch": 0.9844192634560907,
+      "grad_norm": 1.2931617498397827,
+      "learning_rate": 1.589923601073715e-06,
+      "loss": 1.0173,
+      "step": 24325
+    },
+    {
+      "epoch": 0.9846216106839336,
+      "grad_norm": 1.1261266469955444,
+      "learning_rate": 1.5692752426182118e-06,
+      "loss": 0.9884,
+      "step": 24330
+    },
+    {
+      "epoch": 0.9848239579117766,
+      "grad_norm": 1.300706386566162,
+      "learning_rate": 1.548626884162709e-06,
+      "loss": 0.9546,
+      "step": 24335
+    },
+    {
+      "epoch": 0.9850263051396196,
+      "grad_norm": 1.229163408279419,
+      "learning_rate": 1.5279785257072064e-06,
+      "loss": 0.9687,
+      "step": 24340
+    },
+    {
+      "epoch": 0.9852286523674626,
+      "grad_norm": 1.2274683713912964,
+      "learning_rate": 1.5073301672517035e-06,
+      "loss": 1.0014,
+      "step": 24345
+    },
+    {
+      "epoch": 0.9854309995953056,
+      "grad_norm": 1.395676612854004,
+      "learning_rate": 1.4866818087962008e-06,
+      "loss": 1.0205,
+      "step": 24350
+    },
+    {
+      "epoch": 0.9856333468231485,
+      "grad_norm": 1.0325241088867188,
+      "learning_rate": 1.466033450340698e-06,
+      "loss": 0.9521,
+      "step": 24355
+    },
+    {
+      "epoch": 0.9858356940509915,
+      "grad_norm": 1.2387555837631226,
+      "learning_rate": 1.4453850918851952e-06,
+      "loss": 0.9868,
+      "step": 24360
+    },
+    {
+      "epoch": 0.9860380412788344,
+      "grad_norm": 1.153342604637146,
+      "learning_rate": 1.4247367334296924e-06,
+      "loss": 0.9706,
+      "step": 24365
+    },
+    {
+      "epoch": 0.9862403885066775,
+      "grad_norm": 1.3121017217636108,
+      "learning_rate": 1.4040883749741897e-06,
+      "loss": 1.0178,
+      "step": 24370
+    },
+    {
+      "epoch": 0.9864427357345205,
+      "grad_norm": 1.3076057434082031,
+      "learning_rate": 1.3834400165186868e-06,
+      "loss": 1.0004,
+      "step": 24375
+    },
+    {
+      "epoch": 0.9866450829623634,
+      "grad_norm": 1.2197092771530151,
+      "learning_rate": 1.3627916580631841e-06,
+      "loss": 1.0053,
+      "step": 24380
+    },
+    {
+      "epoch": 0.9868474301902064,
+      "grad_norm": 1.2461926937103271,
+      "learning_rate": 1.3421432996076812e-06,
+      "loss": 0.966,
+      "step": 24385
+    },
+    {
+      "epoch": 0.9870497774180493,
+      "grad_norm": 1.2928906679153442,
+      "learning_rate": 1.3214949411521785e-06,
+      "loss": 0.94,
+      "step": 24390
+    },
+    {
+      "epoch": 0.9872521246458924,
+      "grad_norm": 1.213357925415039,
+      "learning_rate": 1.3008465826966756e-06,
+      "loss": 1.0062,
+      "step": 24395
+    },
+    {
+      "epoch": 0.9874544718737354,
+      "grad_norm": 1.2100154161453247,
+      "learning_rate": 1.2801982242411729e-06,
+      "loss": 0.9827,
+      "step": 24400
+    },
+    {
+      "epoch": 0.9876568191015783,
+      "grad_norm": 1.1670055389404297,
+      "learning_rate": 1.25954986578567e-06,
+      "loss": 0.9837,
+      "step": 24405
+    },
+    {
+      "epoch": 0.9878591663294213,
+      "grad_norm": 1.1580208539962769,
+      "learning_rate": 1.2389015073301673e-06,
+      "loss": 0.9521,
+      "step": 24410
+    },
+    {
+      "epoch": 0.9880615135572642,
+      "grad_norm": 1.1305630207061768,
+      "learning_rate": 1.2182531488746646e-06,
+      "loss": 1.0028,
+      "step": 24415
+    },
+    {
+      "epoch": 0.9882638607851072,
+      "grad_norm": 1.2192747592926025,
+      "learning_rate": 1.1976047904191619e-06,
+      "loss": 0.9609,
+      "step": 24420
+    },
+    {
+      "epoch": 0.9884662080129503,
+      "grad_norm": 1.265036702156067,
+      "learning_rate": 1.176956431963659e-06,
+      "loss": 0.9222,
+      "step": 24425
+    },
+    {
+      "epoch": 0.9886685552407932,
+      "grad_norm": 1.2127940654754639,
+      "learning_rate": 1.1563080735081562e-06,
+      "loss": 0.9798,
+      "step": 24430
+    },
+    {
+      "epoch": 0.9888709024686362,
+      "grad_norm": 1.2852286100387573,
+      "learning_rate": 1.1356597150526533e-06,
+      "loss": 1.0294,
+      "step": 24435
+    },
+    {
+      "epoch": 0.9890732496964791,
+      "grad_norm": 1.1611634492874146,
+      "learning_rate": 1.1150113565971506e-06,
+      "loss": 0.9352,
+      "step": 24440
+    },
+    {
+      "epoch": 0.9892755969243221,
+      "grad_norm": 1.1146661043167114,
+      "learning_rate": 1.0943629981416477e-06,
+      "loss": 1.0145,
+      "step": 24445
+    },
+    {
+      "epoch": 0.9894779441521652,
+      "grad_norm": 1.195049524307251,
+      "learning_rate": 1.073714639686145e-06,
+      "loss": 1.0102,
+      "step": 24450
+    },
+    {
+      "epoch": 0.9896802913800081,
+      "grad_norm": 1.1704808473587036,
+      "learning_rate": 1.053066281230642e-06,
+      "loss": 1.024,
+      "step": 24455
+    },
+    {
+      "epoch": 0.9898826386078511,
+      "grad_norm": 1.1268508434295654,
+      "learning_rate": 1.0324179227751394e-06,
+      "loss": 1.014,
+      "step": 24460
+    },
+    {
+      "epoch": 0.990084985835694,
+      "grad_norm": 1.3437390327453613,
+      "learning_rate": 1.0117695643196367e-06,
+      "loss": 1.0233,
+      "step": 24465
+    },
+    {
+      "epoch": 0.990287333063537,
+      "grad_norm": 1.204566478729248,
+      "learning_rate": 9.911212058641338e-07,
+      "loss": 1.0272,
+      "step": 24470
+    },
+    {
+      "epoch": 0.99048968029138,
+      "grad_norm": 1.1397830247879028,
+      "learning_rate": 9.70472847408631e-07,
+      "loss": 0.9275,
+      "step": 24475
+    },
+    {
+      "epoch": 0.990692027519223,
+      "grad_norm": 1.2095698118209839,
+      "learning_rate": 9.498244889531282e-07,
+      "loss": 0.9429,
+      "step": 24480
+    },
+    {
+      "epoch": 0.990894374747066,
+      "grad_norm": 1.149962067604065,
+      "learning_rate": 9.291761304976256e-07,
+      "loss": 1.035,
+      "step": 24485
+    },
+    {
+      "epoch": 0.991096721974909,
+      "grad_norm": 1.2392507791519165,
+      "learning_rate": 9.085277720421228e-07,
+      "loss": 1.0362,
+      "step": 24490
+    },
+    {
+      "epoch": 0.9912990692027519,
+      "grad_norm": 1.2684184312820435,
+      "learning_rate": 8.878794135866199e-07,
+      "loss": 0.9692,
+      "step": 24495
+    },
+    {
+      "epoch": 0.9915014164305949,
+      "grad_norm": 1.1505355834960938,
+      "learning_rate": 8.672310551311171e-07,
+      "loss": 1.0345,
+      "step": 24500
+    },
+    {
+      "epoch": 0.9917037636584379,
+      "grad_norm": 1.2026902437210083,
+      "learning_rate": 8.465826966756143e-07,
+      "loss": 0.9959,
+      "step": 24505
+    },
+    {
+      "epoch": 0.9919061108862809,
+      "grad_norm": 1.0985236167907715,
+      "learning_rate": 8.259343382201115e-07,
+      "loss": 0.9768,
+      "step": 24510
+    },
+    {
+      "epoch": 0.9921084581141238,
+      "grad_norm": 1.1708601713180542,
+      "learning_rate": 8.052859797646087e-07,
+      "loss": 1.0179,
+      "step": 24515
+    },
+    {
+      "epoch": 0.9923108053419668,
+      "grad_norm": 1.1726040840148926,
+      "learning_rate": 7.846376213091059e-07,
+      "loss": 0.9639,
+      "step": 24520
+    },
+    {
+      "epoch": 0.9925131525698098,
+      "grad_norm": 1.179294228553772,
+      "learning_rate": 7.639892628536032e-07,
+      "loss": 0.9699,
+      "step": 24525
+    },
+    {
+      "epoch": 0.9927154997976527,
+      "grad_norm": 1.2063021659851074,
+      "learning_rate": 7.433409043981004e-07,
+      "loss": 0.9642,
+      "step": 24530
+    },
+    {
+      "epoch": 0.9929178470254958,
+      "grad_norm": 1.2376971244812012,
+      "learning_rate": 7.226925459425976e-07,
+      "loss": 0.961,
+      "step": 24535
+    },
+    {
+      "epoch": 0.9931201942533388,
+      "grad_norm": 1.3272475004196167,
+      "learning_rate": 7.020441874870949e-07,
+      "loss": 0.9758,
+      "step": 24540
+    },
+    {
+      "epoch": 0.9933225414811817,
+      "grad_norm": 1.1692842245101929,
+      "learning_rate": 6.813958290315921e-07,
+      "loss": 1.0068,
+      "step": 24545
+    },
+    {
+      "epoch": 0.9935248887090247,
+      "grad_norm": 1.1925163269042969,
+      "learning_rate": 6.607474705760893e-07,
+      "loss": 1.0023,
+      "step": 24550
+    },
+    {
+      "epoch": 0.9937272359368676,
+      "grad_norm": 1.2250268459320068,
+      "learning_rate": 6.400991121205864e-07,
+      "loss": 1.0437,
+      "step": 24555
+    },
+    {
+      "epoch": 0.9939295831647107,
+      "grad_norm": 1.1780215501785278,
+      "learning_rate": 6.194507536650836e-07,
+      "loss": 0.9488,
+      "step": 24560
+    },
+    {
+      "epoch": 0.9941319303925537,
+      "grad_norm": 1.102590799331665,
+      "learning_rate": 5.988023952095809e-07,
+      "loss": 0.9794,
+      "step": 24565
+    },
+    {
+      "epoch": 0.9943342776203966,
+      "grad_norm": 1.2239985466003418,
+      "learning_rate": 5.781540367540781e-07,
+      "loss": 0.9694,
+      "step": 24570
+    },
+    {
+      "epoch": 0.9945366248482396,
+      "grad_norm": 1.2165189981460571,
+      "learning_rate": 5.575056782985753e-07,
+      "loss": 0.999,
+      "step": 24575
+    },
+    {
+      "epoch": 0.9947389720760825,
+      "grad_norm": 1.164983868598938,
+      "learning_rate": 5.368573198430725e-07,
+      "loss": 1.0329,
+      "step": 24580
+    },
+    {
+      "epoch": 0.9949413193039255,
+      "grad_norm": 1.333021640777588,
+      "learning_rate": 5.162089613875697e-07,
+      "loss": 1.048,
+      "step": 24585
+    },
+    {
+      "epoch": 0.9951436665317686,
+      "grad_norm": 1.3476011753082275,
+      "learning_rate": 4.955606029320669e-07,
+      "loss": 0.9855,
+      "step": 24590
+    },
+    {
+      "epoch": 0.9953460137596115,
+      "grad_norm": 1.2735216617584229,
+      "learning_rate": 4.749122444765641e-07,
+      "loss": 1.0284,
+      "step": 24595
+    },
+    {
+      "epoch": 0.9955483609874545,
+      "grad_norm": 1.1732523441314697,
+      "learning_rate": 4.542638860210614e-07,
+      "loss": 1.0321,
+      "step": 24600
+    },
+    {
+      "epoch": 0.9957507082152974,
+      "grad_norm": 1.0077344179153442,
+      "learning_rate": 4.3361552756555857e-07,
+      "loss": 0.9924,
+      "step": 24605
+    },
+    {
+      "epoch": 0.9959530554431404,
+      "grad_norm": 1.174232840538025,
+      "learning_rate": 4.1296716911005576e-07,
+      "loss": 0.9537,
+      "step": 24610
+    },
+    {
+      "epoch": 0.9961554026709835,
+      "grad_norm": 1.2381494045257568,
+      "learning_rate": 3.9231881065455295e-07,
+      "loss": 0.9593,
+      "step": 24615
+    },
+    {
+      "epoch": 0.9963577498988264,
+      "grad_norm": 1.3087583780288696,
+      "learning_rate": 3.716704521990502e-07,
+      "loss": 0.967,
+      "step": 24620
+    },
+    {
+      "epoch": 0.9965600971266694,
+      "grad_norm": 1.222768783569336,
+      "learning_rate": 3.5102209374354744e-07,
+      "loss": 1.0279,
+      "step": 24625
+    },
+    {
+      "epoch": 0.9967624443545123,
+      "grad_norm": 1.2564250230789185,
+      "learning_rate": 3.3037373528804463e-07,
+      "loss": 1.034,
+      "step": 24630
+    },
+    {
+      "epoch": 0.9969647915823553,
+      "grad_norm": 1.0893826484680176,
+      "learning_rate": 3.097253768325418e-07,
+      "loss": 0.9938,
+      "step": 24635
+    },
+    {
+      "epoch": 0.9971671388101983,
+      "grad_norm": 1.1113886833190918,
+      "learning_rate": 2.8907701837703906e-07,
+      "loss": 0.9568,
+      "step": 24640
+    },
+    {
+      "epoch": 0.9973694860380413,
+      "grad_norm": 1.1226166486740112,
+      "learning_rate": 2.6842865992153625e-07,
+      "loss": 1.0333,
+      "step": 24645
+    },
+    {
+      "epoch": 0.9975718332658843,
+      "grad_norm": 1.3192880153656006,
+      "learning_rate": 2.4778030146603344e-07,
+      "loss": 0.913,
+      "step": 24650
+    },
+    {
+      "epoch": 0.9977741804937272,
+      "grad_norm": 1.1434147357940674,
+      "learning_rate": 2.271319430105307e-07,
+      "loss": 1.0118,
+      "step": 24655
+    },
+    {
+      "epoch": 0.9979765277215702,
+      "grad_norm": 1.2090116739273071,
+      "learning_rate": 2.0648358455502788e-07,
+      "loss": 0.9657,
+      "step": 24660
+    },
+    {
+      "epoch": 0.9981788749494132,
+      "grad_norm": 1.2893544435501099,
+      "learning_rate": 1.858352260995251e-07,
+      "loss": 0.9526,
+      "step": 24665
+    },
+    {
+      "epoch": 0.9983812221772562,
+      "grad_norm": 1.1249070167541504,
+      "learning_rate": 1.6518686764402231e-07,
+      "loss": 0.9653,
+      "step": 24670
+    },
+    {
+      "epoch": 0.9985835694050992,
+      "grad_norm": 1.1773556470870972,
+      "learning_rate": 1.4453850918851953e-07,
+      "loss": 1.0098,
+      "step": 24675
+    },
+    {
+      "epoch": 0.9987859166329421,
+      "grad_norm": 1.2133076190948486,
+      "learning_rate": 1.2389015073301672e-07,
+      "loss": 1.0156,
+      "step": 24680
+    },
+    {
+      "epoch": 0.9989882638607851,
+      "grad_norm": 1.2049704790115356,
+      "learning_rate": 1.0324179227751394e-07,
+      "loss": 0.9781,
+      "step": 24685
+    },
+    {
+      "epoch": 0.9991906110886281,
+      "grad_norm": 1.1690733432769775,
+      "learning_rate": 8.259343382201116e-08,
+      "loss": 0.9829,
+      "step": 24690
+    },
+    {
+      "epoch": 0.999392958316471,
+      "grad_norm": 1.3740712404251099,
+      "learning_rate": 6.194507536650836e-08,
+      "loss": 0.9842,
+      "step": 24695
+    },
+    {
+      "epoch": 0.9995953055443141,
+      "grad_norm": 1.194843053817749,
+      "learning_rate": 4.129671691100558e-08,
+      "loss": 1.0084,
+      "step": 24700
+    },
+    {
+      "epoch": 0.999797652772157,
+      "grad_norm": 1.3295986652374268,
+      "learning_rate": 2.064835845550279e-08,
+      "loss": 0.9322,
+      "step": 24705
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1768206357955933,
+      "learning_rate": 0.0,
+      "loss": 0.9516,
+      "step": 24710
+    },
+    {
+      "epoch": 1.0,
+      "step": 24710,
+      "total_flos": 4.050169848999287e+19,
+      "train_loss": 1.0059279920214441,
+      "train_runtime": 51238.2737,
+      "train_samples_per_second": 15.432,
+      "train_steps_per_second": 0.482
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 24710,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.050169848999287e+19,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}