diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14983 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999531550100716,
+  "eval_steps": 500,
+  "global_step": 10673,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 9.368997985665433e-05,
+      "grad_norm": 59.75,
+      "learning_rate": 1.8726591760299626e-07,
+      "loss": 18.5002,
+      "step": 1
+    },
+    {
+      "epoch": 0.00046844989928327167,
+      "grad_norm": 61.75,
+      "learning_rate": 9.363295880149814e-07,
+      "loss": 18.7036,
+      "step": 5
+    },
+    {
+      "epoch": 0.0009368997985665433,
+      "grad_norm": 70.0,
+      "learning_rate": 1.8726591760299627e-06,
+      "loss": 17.5667,
+      "step": 10
+    },
+    {
+      "epoch": 0.001405349697849815,
+      "grad_norm": 62.0,
+      "learning_rate": 2.808988764044944e-06,
+      "loss": 18.7767,
+      "step": 15
+    },
+    {
+      "epoch": 0.0018737995971330867,
+      "grad_norm": 60.75,
+      "learning_rate": 3.7453183520599255e-06,
+      "loss": 18.3964,
+      "step": 20
+    },
+    {
+      "epoch": 0.0023422494964163583,
+      "grad_norm": 57.5,
+      "learning_rate": 4.6816479400749066e-06,
+      "loss": 17.2933,
+      "step": 25
+    },
+    {
+      "epoch": 0.00281069939569963,
+      "grad_norm": 57.0,
+      "learning_rate": 5.617977528089888e-06,
+      "loss": 16.126,
+      "step": 30
+    },
+    {
+      "epoch": 0.0032791492949829017,
+      "grad_norm": 50.75,
+      "learning_rate": 6.554307116104869e-06,
+      "loss": 16.2407,
+      "step": 35
+    },
+    {
+      "epoch": 0.0037475991942661734,
+      "grad_norm": 51.25,
+      "learning_rate": 7.490636704119851e-06,
+      "loss": 15.0747,
+      "step": 40
+    },
+    {
+      "epoch": 0.004216049093549445,
+      "grad_norm": 39.75,
+      "learning_rate": 8.426966292134832e-06,
+      "loss": 13.6143,
+      "step": 45
+    },
+    {
+      "epoch": 0.004684498992832717,
+      "grad_norm": 36.5,
+      "learning_rate": 9.363295880149813e-06,
+      "loss": 12.2398,
+      "step": 50
+    },
+    {
+      "epoch": 0.005152948892115988,
+      "grad_norm": 36.75,
+      "learning_rate": 1.0299625468164795e-05,
+      "loss": 10.2035,
+      "step": 55
+    },
+    {
+      "epoch": 0.00562139879139926,
+      "grad_norm": 36.5,
+      "learning_rate": 1.1235955056179776e-05,
+      "loss": 10.1946,
+      "step": 60
+    },
+    {
+      "epoch": 0.006089848690682531,
+      "grad_norm": 26.0,
+      "learning_rate": 1.2172284644194758e-05,
+      "loss": 8.0871,
+      "step": 65
+    },
+    {
+      "epoch": 0.006558298589965803,
+      "grad_norm": 19.5,
+      "learning_rate": 1.3108614232209737e-05,
+      "loss": 7.436,
+      "step": 70
+    },
+    {
+      "epoch": 0.007026748489249075,
+      "grad_norm": 11.25,
+      "learning_rate": 1.4044943820224721e-05,
+      "loss": 6.5085,
+      "step": 75
+    },
+    {
+      "epoch": 0.007495198388532347,
+      "grad_norm": 9.875,
+      "learning_rate": 1.4981273408239702e-05,
+      "loss": 5.8836,
+      "step": 80
+    },
+    {
+      "epoch": 0.007963648287815619,
+      "grad_norm": 7.375,
+      "learning_rate": 1.591760299625468e-05,
+      "loss": 5.2192,
+      "step": 85
+    },
+    {
+      "epoch": 0.00843209818709889,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.6853932584269665e-05,
+      "loss": 4.9297,
+      "step": 90
+    },
+    {
+      "epoch": 0.008900548086382161,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.7790262172284646e-05,
+      "loss": 4.1697,
+      "step": 95
+    },
+    {
+      "epoch": 0.009368997985665433,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.8726591760299626e-05,
+      "loss": 4.0899,
+      "step": 100
+    },
+    {
+      "epoch": 0.009837447884948705,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.9662921348314607e-05,
+      "loss": 4.0412,
+      "step": 105
+    },
+    {
+      "epoch": 0.010305897784231976,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.059925093632959e-05,
+      "loss": 3.6698,
+      "step": 110
+    },
+    {
+      "epoch": 0.010774347683515248,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.153558052434457e-05,
+      "loss": 3.8425,
+      "step": 115
+    },
+    {
+      "epoch": 0.01124279758279852,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.2471910112359552e-05,
+      "loss": 3.5762,
+      "step": 120
+    },
+    {
+      "epoch": 0.011711247482081792,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.3408239700374533e-05,
+      "loss": 3.8157,
+      "step": 125
+    },
+    {
+      "epoch": 0.012179697381365063,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.4344569288389517e-05,
+      "loss": 3.6434,
+      "step": 130
+    },
+    {
+      "epoch": 0.012648147280648335,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.5280898876404497e-05,
+      "loss": 3.3979,
+      "step": 135
+    },
+    {
+      "epoch": 0.013116597179931607,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.6217228464419475e-05,
+      "loss": 3.2866,
+      "step": 140
+    },
+    {
+      "epoch": 0.013585047079214877,
+      "grad_norm": 4.78125,
+      "learning_rate": 2.715355805243446e-05,
+      "loss": 2.9153,
+      "step": 145
+    },
+    {
+      "epoch": 0.01405349697849815,
+      "grad_norm": 7.875,
+      "learning_rate": 2.8089887640449443e-05,
+      "loss": 2.8557,
+      "step": 150
+    },
+    {
+      "epoch": 0.014521946877781421,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.902621722846442e-05,
+      "loss": 2.3448,
+      "step": 155
+    },
+    {
+      "epoch": 0.014990396777064693,
+      "grad_norm": 5.21875,
+      "learning_rate": 2.9962546816479404e-05,
+      "loss": 1.7144,
+      "step": 160
+    },
+    {
+      "epoch": 0.015458846676347964,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.089887640449438e-05,
+      "loss": 1.47,
+      "step": 165
+    },
+    {
+      "epoch": 0.015927296575631238,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.183520599250936e-05,
+      "loss": 1.4025,
+      "step": 170
+    },
+    {
+      "epoch": 0.016395746474914506,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.277153558052435e-05,
+      "loss": 1.3861,
+      "step": 175
+    },
+    {
+      "epoch": 0.01686419637419778,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.370786516853933e-05,
+      "loss": 1.384,
+      "step": 180
+    },
+    {
+      "epoch": 0.01733264627348105,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.464419475655431e-05,
+      "loss": 1.3646,
+      "step": 185
+    },
+    {
+      "epoch": 0.017801096172764323,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.558052434456929e-05,
+      "loss": 1.368,
+      "step": 190
+    },
+    {
+      "epoch": 0.018269546072047595,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.651685393258427e-05,
+      "loss": 1.2881,
+      "step": 195
+    },
+    {
+      "epoch": 0.018737995971330867,
+      "grad_norm": 4.0,
+      "learning_rate": 3.745318352059925e-05,
+      "loss": 1.278,
+      "step": 200
+    },
+    {
+      "epoch": 0.01920644587061414,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.838951310861423e-05,
+      "loss": 1.3055,
+      "step": 205
+    },
+    {
+      "epoch": 0.01967489576989741,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.9325842696629214e-05,
+      "loss": 1.2798,
+      "step": 210
+    },
+    {
+      "epoch": 0.02014334566918068,
+      "grad_norm": 1.3984375,
+      "learning_rate": 4.0262172284644194e-05,
+      "loss": 1.2411,
+      "step": 215
+    },
+    {
+      "epoch": 0.02061179556846395,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.119850187265918e-05,
+      "loss": 1.2734,
+      "step": 220
+    },
+    {
+      "epoch": 0.021080245467747224,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.2134831460674156e-05,
+      "loss": 1.2793,
+      "step": 225
+    },
+    {
+      "epoch": 0.021548695367030496,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.307116104868914e-05,
+      "loss": 1.2023,
+      "step": 230
+    },
+    {
+      "epoch": 0.022017145266313768,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.4007490636704124e-05,
+      "loss": 1.2116,
+      "step": 235
+    },
+    {
+      "epoch": 0.02248559516559704,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.4943820224719104e-05,
+      "loss": 1.1648,
+      "step": 240
+    },
+    {
+      "epoch": 0.022954045064880312,
+      "grad_norm": 1.375,
+      "learning_rate": 4.5880149812734085e-05,
+      "loss": 1.1978,
+      "step": 245
+    },
+    {
+      "epoch": 0.023422494964163584,
+      "grad_norm": 1.7265625,
+      "learning_rate": 4.6816479400749066e-05,
+      "loss": 1.2017,
+      "step": 250
+    },
+    {
+      "epoch": 0.023890944863446853,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.7752808988764046e-05,
+      "loss": 1.1734,
+      "step": 255
+    },
+    {
+      "epoch": 0.024359394762730125,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.8689138576779034e-05,
+      "loss": 1.1892,
+      "step": 260
+    },
+    {
+      "epoch": 0.024827844662013397,
+      "grad_norm": 1.6796875,
+      "learning_rate": 4.962546816479401e-05,
+      "loss": 1.3598,
+      "step": 265
+    },
+    {
+      "epoch": 0.02529629456129667,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.0561797752808995e-05,
+      "loss": 1.1311,
+      "step": 270
+    },
+    {
+      "epoch": 0.02576474446057994,
+      "grad_norm": 1.96875,
+      "learning_rate": 5.1498127340823975e-05,
+      "loss": 1.1556,
+      "step": 275
+    },
+    {
+      "epoch": 0.026233194359863213,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.243445692883895e-05,
+      "loss": 1.1625,
+      "step": 280
+    },
+    {
+      "epoch": 0.026701644259146486,
+      "grad_norm": 1.890625,
+      "learning_rate": 5.337078651685393e-05,
+      "loss": 1.1606,
+      "step": 285
+    },
+    {
+      "epoch": 0.027170094158429754,
+      "grad_norm": 5.625,
+      "learning_rate": 5.430711610486892e-05,
+      "loss": 1.153,
+      "step": 290
+    },
+    {
+      "epoch": 0.027638544057713026,
+      "grad_norm": 2.875,
+      "learning_rate": 5.52434456928839e-05,
+      "loss": 1.1565,
+      "step": 295
+    },
+    {
+      "epoch": 0.0281069939569963,
+      "grad_norm": 6.21875,
+      "learning_rate": 5.6179775280898885e-05,
+      "loss": 1.1315,
+      "step": 300
+    },
+    {
+      "epoch": 0.02857544385627957,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.711610486891385e-05,
+      "loss": 1.1626,
+      "step": 305
+    },
+    {
+      "epoch": 0.029043893755562843,
+      "grad_norm": 5.15625,
+      "learning_rate": 5.805243445692884e-05,
+      "loss": 1.0912,
+      "step": 310
+    },
+    {
+      "epoch": 0.029512343654846115,
+      "grad_norm": 9.125,
+      "learning_rate": 5.898876404494382e-05,
+      "loss": 1.1472,
+      "step": 315
+    },
+    {
+      "epoch": 0.029980793554129387,
+      "grad_norm": 5.875,
+      "learning_rate": 5.992509363295881e-05,
+      "loss": 1.1274,
+      "step": 320
+    },
+    {
+      "epoch": 0.03044924345341266,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.086142322097379e-05,
+      "loss": 1.1388,
+      "step": 325
+    },
+    {
+      "epoch": 0.030917693352695928,
+      "grad_norm": 1.90625,
+      "learning_rate": 6.179775280898876e-05,
+      "loss": 1.1032,
+      "step": 330
+    },
+    {
+      "epoch": 0.0313861432519792,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.273408239700374e-05,
+      "loss": 1.1108,
+      "step": 335
+    },
+    {
+      "epoch": 0.031854593151262475,
+      "grad_norm": 1.84375,
+      "learning_rate": 6.367041198501872e-05,
+      "loss": 1.0866,
+      "step": 340
+    },
+    {
+      "epoch": 0.032323043050545744,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.460674157303372e-05,
+      "loss": 1.1008,
+      "step": 345
+    },
+    {
+      "epoch": 0.03279149294982901,
+      "grad_norm": 1.9140625,
+      "learning_rate": 6.55430711610487e-05,
+      "loss": 1.1271,
+      "step": 350
+    },
+    {
+      "epoch": 0.03325994284911229,
+      "grad_norm": 2.90625,
+      "learning_rate": 6.647940074906367e-05,
+      "loss": 1.1526,
+      "step": 355
+    },
+    {
+      "epoch": 0.03372839274839556,
+      "grad_norm": 5.125,
+      "learning_rate": 6.741573033707866e-05,
+      "loss": 1.1009,
+      "step": 360
+    },
+    {
+      "epoch": 0.03419684264767883,
+      "grad_norm": 4.21875,
+      "learning_rate": 6.835205992509364e-05,
+      "loss": 1.1185,
+      "step": 365
+    },
+    {
+      "epoch": 0.0346652925469621,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.928838951310862e-05,
+      "loss": 1.1078,
+      "step": 370
+    },
+    {
+      "epoch": 0.035133742446245377,
+      "grad_norm": 10.4375,
+      "learning_rate": 7.02247191011236e-05,
+      "loss": 1.0713,
+      "step": 375
+    },
+    {
+      "epoch": 0.035602192345528645,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.116104868913858e-05,
+      "loss": 1.1002,
+      "step": 380
+    },
+    {
+      "epoch": 0.03607064224481192,
+      "grad_norm": 9.0625,
+      "learning_rate": 7.209737827715356e-05,
+      "loss": 1.0704,
+      "step": 385
+    },
+    {
+      "epoch": 0.03653909214409519,
+      "grad_norm": 4.5625,
+      "learning_rate": 7.303370786516854e-05,
+      "loss": 1.1168,
+      "step": 390
+    },
+    {
+      "epoch": 0.03700754204337846,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.397003745318352e-05,
+      "loss": 1.0697,
+      "step": 395
+    },
+    {
+      "epoch": 0.037475991942661734,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.49063670411985e-05,
+      "loss": 1.0654,
+      "step": 400
+    },
+    {
+      "epoch": 0.037944441841945,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.584269662921349e-05,
+      "loss": 1.0849,
+      "step": 405
+    },
+    {
+      "epoch": 0.03841289174122828,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.677902621722847e-05,
+      "loss": 1.0929,
+      "step": 410
+    },
+    {
+      "epoch": 0.038881341640511546,
+      "grad_norm": 4.9375,
+      "learning_rate": 7.771535580524345e-05,
+      "loss": 1.0614,
+      "step": 415
+    },
+    {
+      "epoch": 0.03934979153979482,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.865168539325843e-05,
+      "loss": 1.0803,
+      "step": 420
+    },
+    {
+      "epoch": 0.03981824143907809,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.958801498127341e-05,
+      "loss": 1.0723,
+      "step": 425
+    },
+    {
+      "epoch": 0.04028669133836136,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.052434456928839e-05,
+      "loss": 1.087,
+      "step": 430
+    },
+    {
+      "epoch": 0.040755141237644635,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.146067415730337e-05,
+      "loss": 1.0719,
+      "step": 435
+    },
+    {
+      "epoch": 0.0412235911369279,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.239700374531836e-05,
+      "loss": 1.1398,
+      "step": 440
+    },
+    {
+      "epoch": 0.04169204103621118,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 1.0681,
+      "step": 445
+    },
+    {
+      "epoch": 0.04216049093549445,
+      "grad_norm": 7.5625,
+      "learning_rate": 8.426966292134831e-05,
+      "loss": 1.1118,
+      "step": 450
+    },
+    {
+      "epoch": 0.04262894083477772,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.520599250936329e-05,
+      "loss": 1.0653,
+      "step": 455
+    },
+    {
+      "epoch": 0.04309739073406099,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.614232209737829e-05,
+      "loss": 1.0697,
+      "step": 460
+    },
+    {
+      "epoch": 0.04356584063334426,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.707865168539327e-05,
+      "loss": 1.0489,
+      "step": 465
+    },
+    {
+      "epoch": 0.044034290532627536,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.801498127340825e-05,
+      "loss": 1.0835,
+      "step": 470
+    },
+    {
+      "epoch": 0.044502740431910805,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.895131086142321e-05,
+      "loss": 1.0403,
+      "step": 475
+    },
+    {
+      "epoch": 0.04497119033119408,
+      "grad_norm": 4.03125,
+      "learning_rate": 8.988764044943821e-05,
+      "loss": 1.0366,
+      "step": 480
+    },
+    {
+      "epoch": 0.04543964023047735,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.082397003745319e-05,
+      "loss": 1.0804,
+      "step": 485
+    },
+    {
+      "epoch": 0.045908090129760624,
+      "grad_norm": 6.21875,
+      "learning_rate": 9.176029962546817e-05,
+      "loss": 1.0536,
+      "step": 490
+    },
+    {
+      "epoch": 0.04637654002904389,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.269662921348315e-05,
+      "loss": 1.0541,
+      "step": 495
+    },
+    {
+      "epoch": 0.04684498992832717,
+      "grad_norm": 4.875,
+      "learning_rate": 9.363295880149813e-05,
+      "loss": 1.037,
+      "step": 500
+    },
+    {
+      "epoch": 0.04731343982761044,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.456928838951311e-05,
+      "loss": 1.0434,
+      "step": 505
+    },
+    {
+      "epoch": 0.047781889726893706,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.550561797752809e-05,
+      "loss": 1.0342,
+      "step": 510
+    },
+    {
+      "epoch": 0.04825033962617698,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.644194756554307e-05,
+      "loss": 1.0589,
+      "step": 515
+    },
+    {
+      "epoch": 0.04871878952546025,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.737827715355807e-05,
+      "loss": 1.0573,
+      "step": 520
+    },
+    {
+      "epoch": 0.049187239424743526,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.831460674157303e-05,
+      "loss": 1.0491,
+      "step": 525
+    },
+    {
+      "epoch": 0.049655689324026794,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.925093632958801e-05,
+      "loss": 1.0508,
+      "step": 530
+    },
+    {
+      "epoch": 0.05012413922331007,
+      "grad_norm": 3.484375,
+      "learning_rate": 0.000100187265917603,
+      "loss": 1.0257,
+      "step": 535
+    },
+    {
+      "epoch": 0.05059258912259334,
+      "grad_norm": 6.34375,
+      "learning_rate": 0.00010112359550561799,
+      "loss": 1.0485,
+      "step": 540
+    },
+    {
+      "epoch": 0.05106103902187661,
+      "grad_norm": 5.0,
+      "learning_rate": 0.00010205992509363296,
+      "loss": 1.0738,
+      "step": 545
+    },
+    {
+      "epoch": 0.05152948892115988,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00010299625468164795,
+      "loss": 1.05,
+      "step": 550
+    },
+    {
+      "epoch": 0.05199793882044315,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00010393258426966293,
+      "loss": 1.1056,
+      "step": 555
+    },
+    {
+      "epoch": 0.05246638871972643,
+      "grad_norm": 4.3125,
+      "learning_rate": 0.0001048689138576779,
+      "loss": 1.0542,
+      "step": 560
+    },
+    {
+      "epoch": 0.052934838619009696,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00010580524344569289,
+      "loss": 1.0249,
+      "step": 565
+    },
+    {
+      "epoch": 0.05340328851829297,
+      "grad_norm": 4.5625,
+      "learning_rate": 0.00010674157303370786,
+      "loss": 1.0088,
+      "step": 570
+    },
+    {
+      "epoch": 0.05387173841757624,
+      "grad_norm": 4.84375,
+      "learning_rate": 0.00010767790262172285,
+      "loss": 1.0343,
+      "step": 575
+    },
+    {
+      "epoch": 0.05434018831685951,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00010861423220973783,
+      "loss": 1.053,
+      "step": 580
+    },
+    {
+      "epoch": 0.054808638216142784,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001095505617977528,
+      "loss": 1.0745,
+      "step": 585
+    },
+    {
+      "epoch": 0.05527708811542605,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001104868913857678,
+      "loss": 1.0639,
+      "step": 590
+    },
+    {
+      "epoch": 0.05574553801470933,
+      "grad_norm": 13.6875,
+      "learning_rate": 0.00011142322097378278,
+      "loss": 1.0186,
+      "step": 595
+    },
+    {
+      "epoch": 0.0562139879139926,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011235955056179777,
+      "loss": 1.037,
+      "step": 600
+    },
+    {
+      "epoch": 0.05668243781327587,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00011329588014981274,
+      "loss": 1.0343,
+      "step": 605
+    },
+    {
+      "epoch": 0.05715088771255914,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001142322097378277,
+      "loss": 1.0305,
+      "step": 610
+    },
+    {
+      "epoch": 0.05761933761184242,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001151685393258427,
+      "loss": 1.0014,
+      "step": 615
+    },
+    {
+      "epoch": 0.058087787511125685,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00011610486891385768,
+      "loss": 1.0608,
+      "step": 620
+    },
+    {
+      "epoch": 0.058556237410408954,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.00011704119850187267,
+      "loss": 1.0366,
+      "step": 625
+    },
+    {
+      "epoch": 0.05902468730969223,
+      "grad_norm": 3.203125,
+      "learning_rate": 0.00011797752808988764,
+      "loss": 1.043,
+      "step": 630
+    },
+    {
+      "epoch": 0.0594931372089755,
+      "grad_norm": 4.21875,
+      "learning_rate": 0.00011891385767790262,
+      "loss": 1.0977,
+      "step": 635
+    },
+    {
+      "epoch": 0.059961587108258774,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00011985018726591762,
+      "loss": 1.0167,
+      "step": 640
+    },
+    {
+      "epoch": 0.06043003700754204,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00012078651685393258,
+      "loss": 1.0229,
+      "step": 645
+    },
+    {
+      "epoch": 0.06089848690682532,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00012172284644194758,
+      "loss": 1.045,
+      "step": 650
+    },
+    {
+      "epoch": 0.06136693680610859,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00012265917602996256,
+      "loss": 1.0012,
+      "step": 655
+    },
+    {
+      "epoch": 0.061835386705391855,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00012359550561797752,
+      "loss": 1.0454,
+      "step": 660
+    },
+    {
+      "epoch": 0.06230383660467513,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00012453183520599252,
+      "loss": 1.0195,
+      "step": 665
+    },
+    {
+      "epoch": 0.0627722865039584,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012546816479400749,
+      "loss": 1.0344,
+      "step": 670
+    },
+    {
+      "epoch": 0.06324073640324167,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.00012640449438202248,
+      "loss": 0.9961,
+      "step": 675
+    },
+    {
+      "epoch": 0.06370918630252495,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00012734082397003745,
+      "loss": 1.1804,
+      "step": 680
+    },
+    {
+      "epoch": 0.06417763620180822,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00012827715355805244,
+      "loss": 1.0536,
+      "step": 685
+    },
+    {
+      "epoch": 0.06464608610109149,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00012921348314606744,
+      "loss": 0.9982,
+      "step": 690
+    },
+    {
+      "epoch": 0.06511453600037476,
+      "grad_norm": 3.65625,
+      "learning_rate": 0.0001301498127340824,
+      "loss": 1.026,
+      "step": 695
+    },
+    {
+      "epoch": 0.06558298589965803,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001310861423220974,
+      "loss": 1.0377,
+      "step": 700
+    },
+    {
+      "epoch": 0.06605143579894131,
+      "grad_norm": 3.3125,
+      "learning_rate": 0.00013202247191011236,
+      "loss": 1.0295,
+      "step": 705
+    },
+    {
+      "epoch": 0.06651988569822458,
+      "grad_norm": 3.1875,
+      "learning_rate": 0.00013295880149812733,
+      "loss": 1.0156,
+      "step": 710
+    },
+    {
+      "epoch": 0.06698833559750784,
+      "grad_norm": 3.0625,
+      "learning_rate": 0.00013389513108614233,
+      "loss": 0.9841,
+      "step": 715
+    },
+    {
+      "epoch": 0.06745678549679111,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00013483146067415732,
+      "loss": 1.0149,
+      "step": 720
+    },
+    {
+      "epoch": 0.0679252353960744,
+      "grad_norm": 3.78125,
+      "learning_rate": 0.00013576779026217231,
+      "loss": 1.0258,
+      "step": 725
+    },
+    {
+      "epoch": 0.06839368529535766,
+      "grad_norm": 6.15625,
+      "learning_rate": 0.00013670411985018728,
+      "loss": 1.022,
+      "step": 730
+    },
+    {
+      "epoch": 0.06886213519464093,
+      "grad_norm": 3.375,
+      "learning_rate": 0.00013764044943820225,
+      "loss": 1.0308,
+      "step": 735
+    },
+    {
+      "epoch": 0.0693305850939242,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013857677902621724,
+      "loss": 0.9978,
+      "step": 740
+    },
+    {
+      "epoch": 0.06979903499320747,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001395131086142322,
+      "loss": 1.0429,
+      "step": 745
+    },
+    {
+      "epoch": 0.07026748489249075,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001404494382022472,
+      "loss": 1.0349,
+      "step": 750
+    },
+    {
+      "epoch": 0.07073593479177402,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014138576779026217,
+      "loss": 0.9822,
+      "step": 755
+    },
+    {
+      "epoch": 0.07120438469105729,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014232209737827716,
+      "loss": 1.0195,
+      "step": 760
+    },
+    {
+      "epoch": 0.07167283459034056,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014325842696629216,
+      "loss": 1.0603,
+      "step": 765
+    },
+    {
+      "epoch": 0.07214128448962384,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00014419475655430713,
+      "loss": 1.0338,
+      "step": 770
+    },
+    {
+      "epoch": 0.07260973438890711,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014513108614232212,
+      "loss": 0.994,
+      "step": 775
+    },
+    {
+      "epoch": 0.07307818428819038,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001460674157303371,
+      "loss": 0.9962,
+      "step": 780
+    },
+    {
+      "epoch": 0.07354663418747365,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00014700374531835205,
+      "loss": 1.0253,
+      "step": 785
+    },
+    {
+      "epoch": 0.07401508408675692,
+      "grad_norm": 3.515625,
+      "learning_rate": 0.00014794007490636705,
+      "loss": 1.016,
+      "step": 790
+    },
+    {
+      "epoch": 0.0744835339860402,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00014887640449438202,
+      "loss": 1.0569,
+      "step": 795
+    },
+    {
+      "epoch": 0.07495198388532347,
+      "grad_norm": 3.75,
+      "learning_rate": 0.000149812734082397,
+      "loss": 1.0709,
+      "step": 800
+    },
+    {
+      "epoch": 0.07542043378460674,
+      "grad_norm": 4.9375,
+      "learning_rate": 0.000150749063670412,
+      "loss": 0.9931,
+      "step": 805
+    },
+    {
+      "epoch": 0.07588888368389,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015168539325842697,
+      "loss": 1.0313,
+      "step": 810
+    },
+    {
+      "epoch": 0.07635733358317327,
+      "grad_norm": 3.265625,
+      "learning_rate": 0.00015262172284644197,
+      "loss": 1.0261,
+      "step": 815
+    },
+    {
+      "epoch": 0.07682578348245656,
+      "grad_norm": 5.0625,
+      "learning_rate": 0.00015355805243445693,
+      "loss": 1.0057,
+      "step": 820
+    },
+    {
+      "epoch": 0.07729423338173982,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015449438202247193,
+      "loss": 1.0323,
+      "step": 825
+    },
+    {
+      "epoch": 0.07776268328102309,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001554307116104869,
+      "loss": 1.0045,
+      "step": 830
+    },
+    {
+      "epoch": 0.07823113318030636,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015636704119850186,
+      "loss": 1.0209,
+      "step": 835
+    },
+    {
+      "epoch": 0.07869958307958964,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00015730337078651685,
+      "loss": 1.0035,
+      "step": 840
+    },
+    {
+      "epoch": 0.07916803297887291,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00015823970037453185,
+      "loss": 0.9975,
+      "step": 845
+    },
+    {
+      "epoch": 0.07963648287815618,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015917602996254682,
+      "loss": 1.0231,
+      "step": 850
+    },
+    {
+      "epoch": 0.08010493277743945,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001601123595505618,
+      "loss": 0.9831,
+      "step": 855
+    },
+    {
+      "epoch": 0.08057338267672272,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016104868913857678,
+      "loss": 1.0326,
+      "step": 860
+    },
+    {
+      "epoch": 0.081041832576006,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016198501872659177,
+      "loss": 1.0352,
+      "step": 865
+    },
+    {
+      "epoch": 0.08151028247528927,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00016292134831460674,
+      "loss": 1.0035,
+      "step": 870
+    },
+    {
+      "epoch": 0.08197873237457254,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016385767790262173,
+      "loss": 0.9943,
+      "step": 875
+    },
+    {
+      "epoch": 0.0824471822738558,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016479400749063673,
+      "loss": 0.99,
+      "step": 880
+    },
+    {
+      "epoch": 0.08291563217313909,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0001657303370786517,
+      "loss": 1.0166,
+      "step": 885
+    },
+    {
+      "epoch": 0.08338408207242236,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 1.0082,
+      "step": 890
+    },
+    {
+      "epoch": 0.08385253197170563,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00016760299625468166,
+      "loss": 1.0033,
+      "step": 895
+    },
+    {
+      "epoch": 0.0843209818709889,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016853932584269662,
+      "loss": 0.9985,
+      "step": 900
+    },
+    {
+      "epoch": 0.08478943177027216,
+      "grad_norm": 3.609375,
+      "learning_rate": 0.00016947565543071162,
+      "loss": 1.0126,
+      "step": 905
+    },
+    {
+      "epoch": 0.08525788166955545,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00017041198501872658,
+      "loss": 0.9762,
+      "step": 910
+    },
+    {
+      "epoch": 0.08572633156883872,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00017134831460674158,
+      "loss": 1.0369,
+      "step": 915
+    },
+    {
+      "epoch": 0.08619478146812198,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00017228464419475657,
+      "loss": 1.0006,
+      "step": 920
+    },
+    {
+      "epoch": 0.08666323136740525,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00017322097378277154,
+      "loss": 0.9776,
+      "step": 925
+    },
+    {
+      "epoch": 0.08713168126668852,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00017415730337078653,
+      "loss": 0.9939,
+      "step": 930
+    },
+    {
+      "epoch": 0.0876001311659718,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001750936329588015,
+      "loss": 0.9966,
+      "step": 935
+    },
+    {
+      "epoch": 0.08806858106525507,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.0001760299625468165,
+      "loss": 1.023,
+      "step": 940
+    },
+    {
+      "epoch": 0.08853703096453834,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00017696629213483146,
+      "loss": 1.0315,
+      "step": 945
+    },
+    {
+      "epoch": 0.08900548086382161,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00017790262172284643,
+      "loss": 0.9933,
+      "step": 950
+    },
+    {
+      "epoch": 0.08947393076310489,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00017883895131086142,
+      "loss": 0.9714,
+      "step": 955
+    },
+    {
+      "epoch": 0.08994238066238816,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017977528089887642,
+      "loss": 1.0209,
+      "step": 960
+    },
+    {
+      "epoch": 0.09041083056167143,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.0001807116104868914,
+      "loss": 0.9626,
+      "step": 965
+    },
+    {
+      "epoch": 0.0908792804609547,
+      "grad_norm": 3.234375,
+      "learning_rate": 0.00018164794007490638,
+      "loss": 1.0271,
+      "step": 970
+    },
+    {
+      "epoch": 0.09134773036023797,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00018258426966292135,
+      "loss": 0.9851,
+      "step": 975
+    },
+    {
+      "epoch": 0.09181618025952125,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00018352059925093634,
+      "loss": 1.0166,
+      "step": 980
+    },
+    {
+      "epoch": 0.09228463015880452,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001844569288389513,
+      "loss": 0.9921,
+      "step": 985
+    },
+    {
+      "epoch": 0.09275308005808779,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001853932584269663,
+      "loss": 1.0042,
+      "step": 990
+    },
+    {
+      "epoch": 0.09322152995737105,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001863295880149813,
+      "loss": 0.9728,
+      "step": 995
+    },
+    {
+      "epoch": 0.09368997985665434,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00018726591760299626,
+      "loss": 1.0043,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0941584297559376,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00018820224719101126,
+      "loss": 0.9723,
+      "step": 1005
+    },
+    {
+      "epoch": 0.09462687965522087,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00018913857677902622,
+      "loss": 1.002,
+      "step": 1010
+    },
+    {
+      "epoch": 0.09509532955450414,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019007490636704122,
+      "loss": 1.0008,
+      "step": 1015
+    },
+    {
+      "epoch": 0.09556377945378741,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019101123595505618,
+      "loss": 1.0241,
+      "step": 1020
+    },
+    {
+      "epoch": 0.0960322293530707,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00019194756554307115,
+      "loss": 0.9787,
+      "step": 1025
+    },
+    {
+      "epoch": 0.09650067925235396,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00019288389513108615,
+      "loss": 1.0108,
+      "step": 1030
+    },
+    {
+      "epoch": 0.09696912915163723,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00019382022471910114,
+      "loss": 1.0067,
+      "step": 1035
+    },
+    {
+      "epoch": 0.0974375790509205,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00019475655430711613,
+      "loss": 0.9861,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09790602895020377,
+      "grad_norm": 3.90625,
+      "learning_rate": 0.0001956928838951311,
+      "loss": 1.0238,
+      "step": 1045
+    },
+    {
+      "epoch": 0.09837447884948705,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00019662921348314607,
+      "loss": 0.9925,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09884292874877032,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019756554307116106,
+      "loss": 0.9541,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09931137864805359,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019850187265917603,
+      "loss": 1.1368,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09977982854733686,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019943820224719102,
+      "loss": 0.9589,
+      "step": 1065
+    },
+    {
+      "epoch": 0.10024827844662014,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0001999999786038819,
+      "loss": 0.9608,
+      "step": 1070
+    },
+    {
+      "epoch": 0.10071672834590341,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00019999973789765796,
+      "loss": 0.9608,
+      "step": 1075
+    },
+    {
+      "epoch": 0.10118517824518668,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001999992297407084,
+      "loss": 1.0051,
+      "step": 1080
+    },
+    {
+      "epoch": 0.10165362814446995,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00019999845413439226,
+      "loss": 1.0035,
+      "step": 1085
+    },
+    {
+      "epoch": 0.10212207804375321,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00019999741108078393,
+      "loss": 0.982,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1025905279430365,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00019999610058267302,
+      "loss": 0.9796,
+      "step": 1095
+    },
+    {
+      "epoch": 0.10305897784231977,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019999452264356452,
+      "loss": 1.0045,
+      "step": 1100
+    },
+    {
+      "epoch": 0.10352742774160303,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019999267726767862,
+      "loss": 1.0186,
+      "step": 1105
+    },
+    {
+      "epoch": 0.1039958776408863,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00019999056445995083,
+      "loss": 0.988,
+      "step": 1110
+    },
+    {
+      "epoch": 0.10446432754016959,
+      "grad_norm": 4.34375,
+      "learning_rate": 0.00019998818422603187,
+      "loss": 1.0081,
+      "step": 1115
+    },
+    {
+      "epoch": 0.10493277743945285,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00019998553657228773,
+      "loss": 1.0179,
+      "step": 1120
+    },
+    {
+      "epoch": 0.10540122733873612,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00019998262150579957,
+      "loss": 0.9541,
+      "step": 1125
+    },
+    {
+      "epoch": 0.10586967723801939,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001999794390343638,
+      "loss": 0.9809,
+      "step": 1130
+    },
+    {
+      "epoch": 0.10633812713730266,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019997598916649196,
+      "loss": 0.9732,
+      "step": 1135
+    },
+    {
+      "epoch": 0.10680657703658594,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001999722719114108,
+      "loss": 1.0038,
+      "step": 1140
+    },
+    {
+      "epoch": 0.10727502693586921,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00019996828727906213,
+      "loss": 0.9543,
+      "step": 1145
+    },
+    {
+      "epoch": 0.10774347683515248,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00019996403528010296,
+      "loss": 0.9554,
+      "step": 1150
+    },
+    {
+      "epoch": 0.10821192673443575,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019995951592590528,
+      "loss": 0.9916,
+      "step": 1155
+    },
+    {
+      "epoch": 0.10868037663371902,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001999547292285562,
+      "loss": 1.0122,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1091488265330023,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019994967520085776,
+      "loss": 0.9566,
+      "step": 1165
+    },
+    {
+      "epoch": 0.10961727643228557,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00019994435385632707,
+      "loss": 0.9811,
+      "step": 1170
+    },
+    {
+      "epoch": 0.11008572633156884,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00019993876520919615,
+      "loss": 0.9731,
+      "step": 1175
+    },
+    {
+      "epoch": 0.1105541762308521,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019993290927441192,
+      "loss": 0.9967,
+      "step": 1180
+    },
+    {
+      "epoch": 0.11102262613013539,
+      "grad_norm": 2.875,
+      "learning_rate": 0.0001999267860676361,
+      "loss": 0.9995,
+      "step": 1185
+    },
+    {
+      "epoch": 0.11149107602941866,
+      "grad_norm": 3.953125,
+      "learning_rate": 0.00019992039560524534,
+      "loss": 0.984,
+      "step": 1190
+    },
+    {
+      "epoch": 0.11195952592870193,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019991373790433103,
+      "loss": 0.9865,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1124279758279852,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00019990681298269927,
+      "loss": 0.9805,
+      "step": 1200
+    },
+    {
+      "epoch": 0.11289642572726846,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00019989962085887087,
+      "loss": 0.9892,
+      "step": 1205
+    },
+    {
+      "epoch": 0.11336487562655174,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019989216155208125,
+      "loss": 0.9501,
+      "step": 1210
+    },
+    {
+      "epoch": 0.11383332552583501,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00019988443508228045,
+      "loss": 0.9623,
+      "step": 1215
+    },
+    {
+      "epoch": 0.11430177542511828,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019987644147013303,
+      "loss": 0.9741,
+      "step": 1220
+    },
+    {
+      "epoch": 0.11477022532440155,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.000199868180737018,
+      "loss": 0.9947,
+      "step": 1225
+    },
+    {
+      "epoch": 0.11523867522368483,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00019985965290502883,
+      "loss": 0.9716,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1157071251229681,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001998508579969733,
+      "loss": 0.9408,
+      "step": 1235
+    },
+    {
+      "epoch": 0.11617557502225137,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0001998417960363735,
+      "loss": 1.018,
+      "step": 1240
+    },
+    {
+      "epoch": 0.11664402492153464,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019983246704746588,
+      "loss": 0.9403,
+      "step": 1245
+    },
+    {
+      "epoch": 0.11711247482081791,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00019982287105520087,
+      "loss": 0.9866,
+      "step": 1250
+    },
+    {
+      "epoch": 0.11758092472010119,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019981300808524303,
+      "loss": 1.0338,
+      "step": 1255
+    },
+    {
+      "epoch": 0.11804937461938446,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019980287816397113,
+      "loss": 1.0068,
+      "step": 1260
+    },
+    {
+      "epoch": 0.11851782451866773,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019979248131847774,
+      "loss": 0.969,
+      "step": 1265
+    },
+    {
+      "epoch": 0.118986274417951,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00019978181757656938,
+      "loss": 1.0264,
+      "step": 1270
+    },
+    {
+      "epoch": 0.11945472431723426,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019977088696676636,
+      "loss": 0.9888,
+      "step": 1275
+    },
+    {
+      "epoch": 0.11992317421651755,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019975968951830276,
+      "loss": 0.9585,
+      "step": 1280
+    },
+    {
+      "epoch": 0.12039162411580082,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00019974822526112632,
+      "loss": 0.9908,
+      "step": 1285
+    },
+    {
+      "epoch": 0.12086007401508408,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019973649422589835,
+      "loss": 0.9777,
+      "step": 1290
+    },
+    {
+      "epoch": 0.12132852391436735,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00019972449644399372,
+      "loss": 0.9487,
+      "step": 1295
+    },
+    {
+      "epoch": 0.12179697381365064,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001997122319475006,
+      "loss": 1.0256,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1222654237129339,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0001996997007692206,
+      "loss": 0.9939,
+      "step": 1305
+    },
+    {
+      "epoch": 0.12273387361221717,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001996869029426685,
+      "loss": 0.9643,
+      "step": 1310
+    },
+    {
+      "epoch": 0.12320232351150044,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001996738385020723,
+      "loss": 0.9645,
+      "step": 1315
+    },
+    {
+      "epoch": 0.12367077341078371,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019966050748237307,
+      "loss": 0.9948,
+      "step": 1320
+    },
+    {
+      "epoch": 0.12413922331006699,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019964690991922473,
+      "loss": 0.9915,
+      "step": 1325
+    },
+    {
+      "epoch": 0.12460767320935026,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001996330458489942,
+      "loss": 0.942,
+      "step": 1330
+    },
+    {
+      "epoch": 0.12507612310863353,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00019961891530876114,
+      "loss": 0.9838,
+      "step": 1335
+    },
+    {
+      "epoch": 0.1255445730079168,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019960451833631789,
+      "loss": 0.9848,
+      "step": 1340
+    },
+    {
+      "epoch": 0.12601302290720007,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001995898549701693,
+      "loss": 0.9989,
+      "step": 1345
+    },
+    {
+      "epoch": 0.12648147280648334,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00019957492524953283,
+      "loss": 0.9635,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1269499227057666,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019955972921433817,
+      "loss": 0.9777,
+      "step": 1355
+    },
+    {
+      "epoch": 0.1274183726050499,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019954426690522734,
+      "loss": 1.0064,
+      "step": 1360
+    },
+    {
+      "epoch": 0.12788682250433317,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019952853836355456,
+      "loss": 0.9731,
+      "step": 1365
+    },
+    {
+      "epoch": 0.12835527240361644,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019951254363138597,
+      "loss": 0.9619,
+      "step": 1370
+    },
+    {
+      "epoch": 0.1288237223028997,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00019949628275149977,
+      "loss": 0.9523,
+      "step": 1375
+    },
+    {
+      "epoch": 0.12929217220218298,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019947975576738584,
+      "loss": 0.9971,
+      "step": 1380
+    },
+    {
+      "epoch": 0.12976062210146624,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00019946296272324593,
+      "loss": 0.9941,
+      "step": 1385
+    },
+    {
+      "epoch": 0.1302290720007495,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019944590366399322,
+      "loss": 0.9658,
+      "step": 1390
+    },
+    {
+      "epoch": 0.13069752190003278,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00019942857863525237,
+      "loss": 0.9469,
+      "step": 1395
+    },
+    {
+      "epoch": 0.13116597179931605,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00019941098768335951,
+      "loss": 0.9712,
+      "step": 1400
+    },
+    {
+      "epoch": 0.13163442169859935,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019939313085536183,
+      "loss": 0.9528,
+      "step": 1405
+    },
+    {
+      "epoch": 0.13210287159788262,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019937500819901767,
+      "loss": 0.9763,
+      "step": 1410
+    },
+    {
+      "epoch": 0.13257132149716588,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019935661976279635,
+      "loss": 0.953,
+      "step": 1415
+    },
+    {
+      "epoch": 0.13303977139644915,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019933796559587797,
+      "loss": 0.982,
+      "step": 1420
+    },
+    {
+      "epoch": 0.13350822129573242,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00019931904574815347,
+      "loss": 0.9518,
+      "step": 1425
+    },
+    {
+      "epoch": 0.1339766711950157,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00019929986027022412,
+      "loss": 0.9832,
+      "step": 1430
+    },
+    {
+      "epoch": 0.13444512109429896,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019928040921340183,
+      "loss": 0.9502,
+      "step": 1435
+    },
+    {
+      "epoch": 0.13491357099358223,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00019926069262970875,
+      "loss": 0.991,
+      "step": 1440
+    },
+    {
+      "epoch": 0.1353820208928655,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019924071057187708,
+      "loss": 0.9673,
+      "step": 1445
+    },
+    {
+      "epoch": 0.1358504707921488,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019922046309334918,
+      "loss": 0.9941,
+      "step": 1450
+    },
+    {
+      "epoch": 0.13631892069143206,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001991999502482772,
+      "loss": 0.9716,
+      "step": 1455
+    },
+    {
+      "epoch": 0.13678737059071533,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00019917917209152306,
+      "loss": 0.9603,
+      "step": 1460
+    },
+    {
+      "epoch": 0.1372558204899986,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019915812867865823,
+      "loss": 0.9475,
+      "step": 1465
+    },
+    {
+      "epoch": 0.13772427038928187,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00019913682006596358,
+      "loss": 0.9458,
+      "step": 1470
+    },
+    {
+      "epoch": 0.13819272028856514,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00019911524631042934,
+      "loss": 0.9505,
+      "step": 1475
+    },
+    {
+      "epoch": 0.1386611701878484,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019909340746975478,
+      "loss": 0.953,
+      "step": 1480
+    },
+    {
+      "epoch": 0.13912962008713167,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00019907130360234825,
+      "loss": 0.9743,
+      "step": 1485
+    },
+    {
+      "epoch": 0.13959806998641494,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00019904893476732684,
+      "loss": 0.9259,
+      "step": 1490
+    },
+    {
+      "epoch": 0.14006651988569824,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001990263010245163,
+      "loss": 0.9878,
+      "step": 1495
+    },
+    {
+      "epoch": 0.1405349697849815,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019900340243445092,
+      "loss": 0.9407,
+      "step": 1500
+    },
+    {
+      "epoch": 0.14100341968426477,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001989802390583733,
+      "loss": 0.9487,
+      "step": 1505
+    },
+    {
+      "epoch": 0.14147186958354804,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00019895681095823418,
+      "loss": 0.9912,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1419403194828313,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00019893311819669244,
+      "loss": 0.9411,
+      "step": 1515
+    },
+    {
+      "epoch": 0.14240876938211458,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001989091608371146,
+      "loss": 0.9697,
+      "step": 1520
+    },
+    {
+      "epoch": 0.14287721928139785,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00019888493894357505,
+      "loss": 0.917,
+      "step": 1525
+    },
+    {
+      "epoch": 0.14334566918068112,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019886045258085553,
+      "loss": 0.9603,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1438141190799644,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001988357018144452,
+      "loss": 0.9652,
+      "step": 1535
+    },
+    {
+      "epoch": 0.14428256897924768,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001988106867105403,
+      "loss": 1.1053,
+      "step": 1540
+    },
+    {
+      "epoch": 0.14475101887853095,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00019878540733604414,
+      "loss": 1.0219,
+      "step": 1545
+    },
+    {
+      "epoch": 0.14521946877781422,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00019875986375856672,
+      "loss": 0.9356,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1456879186770975,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00019873405604642472,
+      "loss": 0.9499,
+      "step": 1555
+    },
+    {
+      "epoch": 0.14615636857638076,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00019870798426864123,
+      "loss": 0.9407,
+      "step": 1560
+    },
+    {
+      "epoch": 0.14662481847566403,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001986816484949456,
+      "loss": 0.9597,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1470932683749473,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00019865504879577322,
+      "loss": 0.9411,
+      "step": 1570
+    },
+    {
+      "epoch": 0.14756171827423056,
+      "grad_norm": 3.625,
+      "learning_rate": 0.00019862818524226538,
+      "loss": 1.0596,
+      "step": 1575
+    },
+    {
+      "epoch": 0.14803016817351383,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019860105790626902,
+      "loss": 0.9631,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1484986180727971,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001985736668603366,
+      "loss": 0.9317,
+      "step": 1585
+    },
+    {
+      "epoch": 0.1489670679720804,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00019854601217772586,
+      "loss": 0.9542,
+      "step": 1590
+    },
+    {
+      "epoch": 0.14943551787136367,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00019851809393239963,
+      "loss": 0.9504,
+      "step": 1595
+    },
+    {
+      "epoch": 0.14990396777064693,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001984899121990257,
+      "loss": 0.9589,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1503724176699302,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019846146705297646,
+      "loss": 0.9818,
+      "step": 1605
+    },
+    {
+      "epoch": 0.15084086756921347,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0001984327585703289,
+      "loss": 0.9372,
+      "step": 1610
+    },
+    {
+      "epoch": 0.15130931746849674,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019840378682786422,
+      "loss": 0.9715,
+      "step": 1615
+    },
+    {
+      "epoch": 0.15177776736778,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001983745519030678,
+      "loss": 0.9831,
+      "step": 1620
+    },
+    {
+      "epoch": 0.15224621726706328,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001983450538741289,
+      "loss": 0.9464,
+      "step": 1625
+    },
+    {
+      "epoch": 0.15271466716634655,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019831529281994032,
+      "loss": 0.9777,
+      "step": 1630
+    },
+    {
+      "epoch": 0.15318311706562984,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0001982852688200985,
+      "loss": 0.9461,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1536515669649131,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00019825498195490303,
+      "loss": 0.9725,
+      "step": 1640
+    },
+    {
+      "epoch": 0.15412001686419638,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019822443230535654,
+      "loss": 0.9799,
+      "step": 1645
+    },
+    {
+      "epoch": 0.15458846676347965,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019819361995316456,
+      "loss": 0.9815,
+      "step": 1650
+    },
+    {
+      "epoch": 0.15505691666276292,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019816254498073512,
+      "loss": 0.931,
+      "step": 1655
+    },
+    {
+      "epoch": 0.15552536656204619,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019813120747117868,
+      "loss": 0.9635,
+      "step": 1660
+    },
+    {
+      "epoch": 0.15599381646132945,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019809960750830788,
+      "loss": 0.9625,
+      "step": 1665
+    },
+    {
+      "epoch": 0.15646226636061272,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019806774517663726,
+      "loss": 0.9563,
+      "step": 1670
+    },
+    {
+      "epoch": 0.156930716259896,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001980356205613831,
+      "loss": 0.9396,
+      "step": 1675
+    },
+    {
+      "epoch": 0.1573991661591793,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001980032337484631,
+      "loss": 0.9367,
+      "step": 1680
+    },
+    {
+      "epoch": 0.15786761605846256,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019797058482449626,
+      "loss": 0.9413,
+      "step": 1685
+    },
+    {
+      "epoch": 0.15833606595774583,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001979376738768026,
+      "loss": 0.9907,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1588045158570291,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019790450099340298,
+      "loss": 0.9707,
+      "step": 1695
+    },
+    {
+      "epoch": 0.15927296575631236,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0001978710662630187,
+      "loss": 0.9337,
+      "step": 1700
+    },
+    {
+      "epoch": 0.15974141565559563,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019783736977507137,
+      "loss": 0.9568,
+      "step": 1705
+    },
+    {
+      "epoch": 0.1602098655548789,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019780341161968279,
+      "loss": 0.9621,
+      "step": 1710
+    },
+    {
+      "epoch": 0.16067831545416217,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00019776919188767452,
+      "loss": 0.9956,
+      "step": 1715
+    },
+    {
+      "epoch": 0.16114676535344544,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019773471067056772,
+      "loss": 0.9403,
+      "step": 1720
+    },
+    {
+      "epoch": 0.16161521525272873,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019769996806058291,
+      "loss": 0.9364,
+      "step": 1725
+    },
+    {
+      "epoch": 0.162083665152012,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00019766496415063965,
+      "loss": 0.9565,
+      "step": 1730
+    },
+    {
+      "epoch": 0.16255211505129527,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019762969903435647,
+      "loss": 0.9566,
+      "step": 1735
+    },
+    {
+      "epoch": 0.16302056495057854,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019759417280605036,
+      "loss": 0.9432,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1634890148498618,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019755838556073682,
+      "loss": 0.945,
+      "step": 1745
+    },
+    {
+      "epoch": 0.16395746474914508,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001975223373941292,
+      "loss": 0.9445,
+      "step": 1750
+    },
+    {
+      "epoch": 0.16442591464842835,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001974860284026389,
+      "loss": 0.9567,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1648943645477116,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019744945868337493,
+      "loss": 0.9604,
+      "step": 1760
+    },
+    {
+      "epoch": 0.16536281444699488,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001974126283341434,
+      "loss": 0.9302,
+      "step": 1765
+    },
+    {
+      "epoch": 0.16583126434627818,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019737553745344765,
+      "loss": 0.9507,
+      "step": 1770
+    },
+    {
+      "epoch": 0.16629971424556145,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001973381861404878,
+      "loss": 0.9551,
+      "step": 1775
+    },
+    {
+      "epoch": 0.16676816414484472,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019730057449516043,
+      "loss": 0.939,
+      "step": 1780
+    },
+    {
+      "epoch": 0.16723661404412798,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001972627026180584,
+      "loss": 0.9205,
+      "step": 1785
+    },
+    {
+      "epoch": 0.16770506394341125,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019722457061047062,
+      "loss": 0.9681,
+      "step": 1790
+    },
+    {
+      "epoch": 0.16817351384269452,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00019718617857438173,
+      "loss": 0.9651,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1686419637419778,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001971475266124717,
+      "loss": 0.9494,
+      "step": 1800
+    },
+    {
+      "epoch": 0.16911041364126106,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001971086148281158,
+      "loss": 0.9315,
+      "step": 1805
+    },
+    {
+      "epoch": 0.16957886354054433,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019706944332538415,
+      "loss": 0.959,
+      "step": 1810
+    },
+    {
+      "epoch": 0.17004731343982762,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019703001220904148,
+      "loss": 0.9456,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1705157633391109,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019699032158454686,
+      "loss": 0.9194,
+      "step": 1820
+    },
+    {
+      "epoch": 0.17098421323839416,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019695037155805344,
+      "loss": 0.9861,
+      "step": 1825
+    },
+    {
+      "epoch": 0.17145266313767743,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00019691016223640818,
+      "loss": 0.9428,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1719211130369607,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019686969372715142,
+      "loss": 0.9658,
+      "step": 1835
+    },
+    {
+      "epoch": 0.17238956293624397,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001968289661385168,
+      "loss": 0.957,
+      "step": 1840
+    },
+    {
+      "epoch": 0.17285801283552724,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00019678797957943083,
+      "loss": 0.9351,
+      "step": 1845
+    },
+    {
+      "epoch": 0.1733264627348105,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00019674673415951265,
+      "loss": 0.9439,
+      "step": 1850
+    },
+    {
+      "epoch": 0.17379491263409377,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019670522998907375,
+      "loss": 0.9431,
+      "step": 1855
+    },
+    {
+      "epoch": 0.17426336253337704,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00019666346717911757,
+      "loss": 0.9579,
+      "step": 1860
+    },
+    {
+      "epoch": 0.17473181243266034,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00019662144584133934,
+      "loss": 0.9198,
+      "step": 1865
+    },
+    {
+      "epoch": 0.1752002623319436,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019657916608812579,
+      "loss": 0.9394,
+      "step": 1870
+    },
+    {
+      "epoch": 0.17566871223122688,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00019653662803255468,
+      "loss": 0.9744,
+      "step": 1875
+    },
+    {
+      "epoch": 0.17613716213051014,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00019649383178839468,
+      "loss": 0.9383,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1766056120297934,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001964507774701049,
+      "loss": 0.9408,
+      "step": 1885
+    },
+    {
+      "epoch": 0.17707406192907668,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019640746519283475,
+      "loss": 0.9479,
+      "step": 1890
+    },
+    {
+      "epoch": 0.17754251182835995,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00019636389507242356,
+      "loss": 0.9514,
+      "step": 1895
+    },
+    {
+      "epoch": 0.17801096172764322,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00019632006722540023,
+      "loss": 0.9277,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1784794116269265,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019627598176898294,
+      "loss": 0.9309,
+      "step": 1905
+    },
+    {
+      "epoch": 0.17894786152620978,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019623163882107888,
+      "loss": 0.9363,
+      "step": 1910
+    },
+    {
+      "epoch": 0.17941631142549305,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019618703850028392,
+      "loss": 0.9387,
+      "step": 1915
+    },
+    {
+      "epoch": 0.17988476132477632,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019614218092588225,
+      "loss": 0.9706,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1803532112240596,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019609706621784607,
+      "loss": 0.9215,
+      "step": 1925
+    },
+    {
+      "epoch": 0.18082166112334286,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001960516944968353,
+      "loss": 0.926,
+      "step": 1930
+    },
+    {
+      "epoch": 0.18129011102262613,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001960060658841973,
+      "loss": 0.9295,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1817585609219094,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001959601805019664,
+      "loss": 0.9429,
+      "step": 1940
+    },
+    {
+      "epoch": 0.18222701082119266,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019591403847286372,
+      "loss": 0.9609,
+      "step": 1945
+    },
+    {
+      "epoch": 0.18269546072047593,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001958676399202968,
+      "loss": 0.9205,
+      "step": 1950
+    },
+    {
+      "epoch": 0.18316391061975923,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001958209849683592,
+      "loss": 0.9242,
+      "step": 1955
+    },
+    {
+      "epoch": 0.1836323605190425,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001957740737418303,
+      "loss": 0.9499,
+      "step": 1960
+    },
+    {
+      "epoch": 0.18410081041832577,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001957269063661748,
+      "loss": 0.9658,
+      "step": 1965
+    },
+    {
+      "epoch": 0.18456926031760904,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019567948296754254,
+      "loss": 0.9494,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1850377102168923,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019563180367276808,
+      "loss": 0.9382,
+      "step": 1975
+    },
+    {
+      "epoch": 0.18550616011617557,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019558386860937034,
+      "loss": 0.9361,
+      "step": 1980
+    },
+    {
+      "epoch": 0.18597461001545884,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00019553567790555238,
+      "loss": 0.9412,
+      "step": 1985
+    },
+    {
+      "epoch": 0.1864430599147421,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019548723169020087,
+      "loss": 0.9626,
+      "step": 1990
+    },
+    {
+      "epoch": 0.18691150981402538,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00019543853009288596,
+      "loss": 0.9385,
+      "step": 1995
+    },
+    {
+      "epoch": 0.18737995971330867,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019538957324386074,
+      "loss": 0.9474,
+      "step": 2000
+    },
+    {
+      "epoch": 0.18784840961259194,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00019534036127406097,
+      "loss": 0.9497,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1883168595118752,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019529089431510484,
+      "loss": 0.9346,
+      "step": 2010
+    },
+    {
+      "epoch": 0.18878530941115848,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019524117249929238,
+      "loss": 0.944,
+      "step": 2015
+    },
+    {
+      "epoch": 0.18925375931044175,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019519119595960528,
+      "loss": 0.9091,
+      "step": 2020
+    },
+    {
+      "epoch": 0.18972220920972502,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019514096482970659,
+      "loss": 0.9136,
+      "step": 2025
+    },
+    {
+      "epoch": 0.1901906591090083,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001950904792439401,
+      "loss": 0.9716,
+      "step": 2030
+    },
+    {
+      "epoch": 0.19065910900829156,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019503973933733025,
+      "loss": 0.9657,
+      "step": 2035
+    },
+    {
+      "epoch": 0.19112755890757482,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019498874524558171,
+      "loss": 0.948,
+      "step": 2040
+    },
+    {
+      "epoch": 0.19159600880685812,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00019493749710507881,
+      "loss": 0.9562,
+      "step": 2045
+    },
+    {
+      "epoch": 0.1920644587061414,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001948859950528855,
+      "loss": 0.9681,
+      "step": 2050
+    },
+    {
+      "epoch": 0.19253290860542466,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019483423922674468,
+      "loss": 0.9417,
+      "step": 2055
+    },
+    {
+      "epoch": 0.19300135850470793,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019478222976507815,
+      "loss": 0.938,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1934698084039912,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00019472996680698588,
+      "loss": 0.966,
+      "step": 2065
+    },
+    {
+      "epoch": 0.19393825830327446,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019467745049224592,
+      "loss": 0.944,
+      "step": 2070
+    },
+    {
+      "epoch": 0.19440670820255773,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019462468096131388,
+      "loss": 0.9288,
+      "step": 2075
+    },
+    {
+      "epoch": 0.194875158101841,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019457165835532267,
+      "loss": 0.9343,
+      "step": 2080
+    },
+    {
+      "epoch": 0.19534360800112427,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00019451838281608197,
+      "loss": 0.9434,
+      "step": 2085
+    },
+    {
+      "epoch": 0.19581205790040754,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019446485448607796,
+      "loss": 0.9183,
+      "step": 2090
+    },
+    {
+      "epoch": 0.19628050779969083,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019441107350847298,
+      "loss": 0.9252,
+      "step": 2095
+    },
+    {
+      "epoch": 0.1967489576989741,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019435704002710496,
+      "loss": 0.9565,
+      "step": 2100
+    },
+    {
+      "epoch": 0.19721740759825737,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00019430275418648725,
+      "loss": 0.9324,
+      "step": 2105
+    },
+    {
+      "epoch": 0.19768585749754064,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001942482161318081,
+      "loss": 0.9295,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1981543073968239,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019419342600893028,
+      "loss": 0.9554,
+      "step": 2115
+    },
+    {
+      "epoch": 0.19862275729610718,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019413838396439083,
+      "loss": 0.9814,
+      "step": 2120
+    },
+    {
+      "epoch": 0.19909120719539045,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019408309014540045,
+      "loss": 0.9376,
+      "step": 2125
+    },
+    {
+      "epoch": 0.19955965709467371,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001940275446998432,
+      "loss": 0.9205,
+      "step": 2130
+    },
+    {
+      "epoch": 0.20002810699395698,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00019397174777627623,
+      "loss": 0.9365,
+      "step": 2135
+    },
+    {
+      "epoch": 0.20049655689324028,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00019391569952392917,
+      "loss": 0.9213,
+      "step": 2140
+    },
+    {
+      "epoch": 0.20096500679252355,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00019385940009270387,
+      "loss": 0.9562,
+      "step": 2145
+    },
+    {
+      "epoch": 0.20143345669180682,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019380284963317398,
+      "loss": 0.9498,
+      "step": 2150
+    },
+    {
+      "epoch": 0.20190190659109009,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019374604829658452,
+      "loss": 0.9581,
+      "step": 2155
+    },
+    {
+      "epoch": 0.20237035649037335,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019368899623485147,
+      "loss": 0.9337,
+      "step": 2160
+    },
+    {
+      "epoch": 0.20283880638965662,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00019363169360056133,
+      "loss": 0.9724,
+      "step": 2165
+    },
+    {
+      "epoch": 0.2033072562889399,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00019357414054697089,
+      "loss": 0.959,
+      "step": 2170
+    },
+    {
+      "epoch": 0.20377570618822316,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00019351633722800658,
+      "loss": 0.9478,
+      "step": 2175
+    },
+    {
+      "epoch": 0.20424415608750643,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001934582837982642,
+      "loss": 0.9777,
+      "step": 2180
+    },
+    {
+      "epoch": 0.20471260598678973,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019339998041300848,
+      "loss": 0.9455,
+      "step": 2185
+    },
+    {
+      "epoch": 0.205181055886073,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00019334142722817266,
+      "loss": 0.9345,
+      "step": 2190
+    },
+    {
+      "epoch": 0.20564950578535626,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00019328262440035802,
+      "loss": 0.9202,
+      "step": 2195
+    },
+    {
+      "epoch": 0.20611795568463953,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00019322357208683362,
+      "loss": 0.9401,
+      "step": 2200
+    },
+    {
+      "epoch": 0.2065864055839228,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019316427044553574,
+      "loss": 0.93,
+      "step": 2205
+    },
+    {
+      "epoch": 0.20705485548320607,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001931047196350674,
+      "loss": 0.9293,
+      "step": 2210
+    },
+    {
+      "epoch": 0.20752330538248934,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001930449198146981,
+      "loss": 0.9321,
+      "step": 2215
+    },
+    {
+      "epoch": 0.2079917552817726,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019298487114436332,
+      "loss": 0.9288,
+      "step": 2220
+    },
+    {
+      "epoch": 0.20846020518105587,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019292457378466412,
+      "loss": 0.9641,
+      "step": 2225
+    },
+    {
+      "epoch": 0.20892865508033917,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019286402789686662,
+      "loss": 0.9317,
+      "step": 2230
+    },
+    {
+      "epoch": 0.20939710497962244,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019280323364290167,
+      "loss": 0.8955,
+      "step": 2235
+    },
+    {
+      "epoch": 0.2098655548789057,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00019274219118536434,
+      "loss": 0.9284,
+      "step": 2240
+    },
+    {
+      "epoch": 0.21033400477818898,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0001926809006875136,
+      "loss": 0.9328,
+      "step": 2245
+    },
+    {
+      "epoch": 0.21080245467747225,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019261936231327172,
+      "loss": 0.9412,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2112709045767555,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019255757622722397,
+      "loss": 0.9362,
+      "step": 2255
+    },
+    {
+      "epoch": 0.21173935447603878,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019249554259461813,
+      "loss": 0.9391,
+      "step": 2260
+    },
+    {
+      "epoch": 0.21220780437532205,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019243326158136406,
+      "loss": 0.9099,
+      "step": 2265
+    },
+    {
+      "epoch": 0.21267625427460532,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00019237073335403318,
+      "loss": 0.9659,
+      "step": 2270
+    },
+    {
+      "epoch": 0.21314470417388862,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001923079580798581,
+      "loss": 0.9235,
+      "step": 2275
+    },
+    {
+      "epoch": 0.21361315407317188,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019224493592673224,
+      "loss": 0.9115,
+      "step": 2280
+    },
+    {
+      "epoch": 0.21408160397245515,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019218166706320923,
+      "loss": 0.945,
+      "step": 2285
+    },
+    {
+      "epoch": 0.21455005387173842,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001921181516585026,
+      "loss": 0.9474,
+      "step": 2290
+    },
+    {
+      "epoch": 0.2150185037710217,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001920543898824851,
+      "loss": 0.9415,
+      "step": 2295
+    },
+    {
+      "epoch": 0.21548695367030496,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019199038190568856,
+      "loss": 0.937,
+      "step": 2300
+    },
+    {
+      "epoch": 0.21595540356958823,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019192612789930324,
+      "loss": 0.9182,
+      "step": 2305
+    },
+    {
+      "epoch": 0.2164238534688715,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001918616280351774,
+      "loss": 0.9502,
+      "step": 2310
+    },
+    {
+      "epoch": 0.21689230336815477,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001917968824858168,
+      "loss": 0.8982,
+      "step": 2315
+    },
+    {
+      "epoch": 0.21736075326743803,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00019173189142438442,
+      "loss": 0.9086,
+      "step": 2320
+    },
+    {
+      "epoch": 0.21782920316672133,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00019166665502469971,
+      "loss": 0.9208,
+      "step": 2325
+    },
+    {
+      "epoch": 0.2182976530660046,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019160117346123836,
+      "loss": 0.9101,
+      "step": 2330
+    },
+    {
+      "epoch": 0.21876610296528787,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019153544690913177,
+      "loss": 0.9423,
+      "step": 2335
+    },
+    {
+      "epoch": 0.21923455286457114,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001914694755441665,
+      "loss": 0.9293,
+      "step": 2340
+    },
+    {
+      "epoch": 0.2197030027638544,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.000191403259542784,
+      "loss": 0.9133,
+      "step": 2345
+    },
+    {
+      "epoch": 0.22017145266313767,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001913367990820798,
+      "loss": 0.9275,
+      "step": 2350
+    },
+    {
+      "epoch": 0.22063990256242094,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019127009433980342,
+      "loss": 0.9211,
+      "step": 2355
+    },
+    {
+      "epoch": 0.2211083524617042,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00019120314549435762,
+      "loss": 0.9112,
+      "step": 2360
+    },
+    {
+      "epoch": 0.22157680236098748,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.000191135952724798,
+      "loss": 0.9264,
+      "step": 2365
+    },
+    {
+      "epoch": 0.22204525226027078,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019106851621083267,
+      "loss": 0.9337,
+      "step": 2370
+    },
+    {
+      "epoch": 0.22251370215955404,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001910008361328215,
+      "loss": 0.9163,
+      "step": 2375
+    },
+    {
+      "epoch": 0.2229821520588373,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019093291267177582,
+      "loss": 0.9088,
+      "step": 2380
+    },
+    {
+      "epoch": 0.22345060195812058,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019086474600935792,
+      "loss": 0.9099,
+      "step": 2385
+    },
+    {
+      "epoch": 0.22391905185740385,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019079633632788044,
+      "loss": 0.9336,
+      "step": 2390
+    },
+    {
+      "epoch": 0.22438750175668712,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019072768381030617,
+      "loss": 0.9155,
+      "step": 2395
+    },
+    {
+      "epoch": 0.2248559516559704,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019065878864024714,
+      "loss": 0.9177,
+      "step": 2400
+    },
+    {
+      "epoch": 0.22532440155525366,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001905896510019645,
+      "loss": 0.9393,
+      "step": 2405
+    },
+    {
+      "epoch": 0.22579285145453692,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00019052027108036788,
+      "loss": 0.9375,
+      "step": 2410
+    },
+    {
+      "epoch": 0.22626130135382022,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00019045064906101484,
+      "loss": 0.9118,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2267297512531035,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00019038078513011048,
+      "loss": 0.9439,
+      "step": 2420
+    },
+    {
+      "epoch": 0.22719820115238676,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019031067947450686,
+      "loss": 0.9321,
+      "step": 2425
+    },
+    {
+      "epoch": 0.22766665105167003,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019024033228170258,
+      "loss": 0.9307,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2281351009509533,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019016974373984216,
+      "loss": 0.9267,
+      "step": 2435
+    },
+    {
+      "epoch": 0.22860355085023656,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00019009891403771572,
+      "loss": 0.9317,
+      "step": 2440
+    },
+    {
+      "epoch": 0.22907200074951983,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.0001900278433647583,
+      "loss": 0.9446,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2295404506488031,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001899565319110494,
+      "loss": 0.9063,
+      "step": 2450
+    },
+    {
+      "epoch": 0.23000890054808637,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018988497986731256,
+      "loss": 0.934,
+      "step": 2455
+    },
+    {
+      "epoch": 0.23047735044736967,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018981318742491472,
+      "loss": 0.926,
+      "step": 2460
+    },
+    {
+      "epoch": 0.23094580034665294,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018974115477586576,
+      "loss": 0.9133,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2314142502459362,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018966888211281808,
+      "loss": 0.9142,
+      "step": 2470
+    },
+    {
+      "epoch": 0.23188270014521947,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018959636962906593,
+      "loss": 0.9286,
+      "step": 2475
+    },
+    {
+      "epoch": 0.23235115004450274,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00018952361751854496,
+      "loss": 0.9026,
+      "step": 2480
+    },
+    {
+      "epoch": 0.232819599943786,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001894506259758318,
+      "loss": 0.9356,
+      "step": 2485
+    },
+    {
+      "epoch": 0.23328804984306928,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018937739519614324,
+      "loss": 0.9393,
+      "step": 2490
+    },
+    {
+      "epoch": 0.23375649974235255,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00018930392537533616,
+      "loss": 0.9234,
+      "step": 2495
+    },
+    {
+      "epoch": 0.23422494964163582,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018923021670990661,
+      "loss": 0.9256,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2346933995409191,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018915626939698945,
+      "loss": 0.9491,
+      "step": 2505
+    },
+    {
+      "epoch": 0.23516184944020238,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018908208363435785,
+      "loss": 0.9285,
+      "step": 2510
+    },
+    {
+      "epoch": 0.23563029933948565,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001890076596204227,
+      "loss": 0.9372,
+      "step": 2515
+    },
+    {
+      "epoch": 0.23609874923876892,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018893299755423203,
+      "loss": 0.9148,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2365671991380522,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00018885809763547067,
+      "loss": 0.9355,
+      "step": 2525
+    },
+    {
+      "epoch": 0.23703564903733546,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00018878296006445946,
+      "loss": 0.923,
+      "step": 2530
+    },
+    {
+      "epoch": 0.23750409893661872,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018870758504215495,
+      "loss": 0.9276,
+      "step": 2535
+    },
+    {
+      "epoch": 0.237972548835902,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001886319727701487,
+      "loss": 0.8998,
+      "step": 2540
+    },
+    {
+      "epoch": 0.23844099873518526,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001885561234506668,
+      "loss": 0.9627,
+      "step": 2545
+    },
+    {
+      "epoch": 0.23890944863446853,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018848003728656932,
+      "loss": 0.9846,
+      "step": 2550
+    },
+    {
+      "epoch": 0.23937789853375183,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018840371448134984,
+      "loss": 0.9108,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2398463484330351,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018832715523913477,
+      "loss": 0.9386,
+      "step": 2560
+    },
+    {
+      "epoch": 0.24031479833231836,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018825035976468288,
+      "loss": 0.9481,
+      "step": 2565
+    },
+    {
+      "epoch": 0.24078324823160163,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001881733282633848,
+      "loss": 0.9623,
+      "step": 2570
+    },
+    {
+      "epoch": 0.2412516981308849,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001880960609412623,
+      "loss": 0.9104,
+      "step": 2575
+    },
+    {
+      "epoch": 0.24172014803016817,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018801855800496803,
+      "loss": 1.0972,
+      "step": 2580
+    },
+    {
+      "epoch": 0.24218859792945144,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018794081966178465,
+      "loss": 0.8949,
+      "step": 2585
+    },
+    {
+      "epoch": 0.2426570478287347,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001878628461196245,
+      "loss": 0.9526,
+      "step": 2590
+    },
+    {
+      "epoch": 0.24312549772801798,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00018778463758702886,
+      "loss": 0.9415,
+      "step": 2595
+    },
+    {
+      "epoch": 0.24359394762730127,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018770619427316762,
+      "loss": 0.894,
+      "step": 2600
+    },
+    {
+      "epoch": 0.24406239752658454,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018762751638783855,
+      "loss": 0.9079,
+      "step": 2605
+    },
+    {
+      "epoch": 0.2445308474258678,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00018754860414146675,
+      "loss": 0.9391,
+      "step": 2610
+    },
+    {
+      "epoch": 0.24499929732515108,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00018746945774510417,
+      "loss": 0.9077,
+      "step": 2615
+    },
+    {
+      "epoch": 0.24546774722443435,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00018739007741042896,
+      "loss": 0.9298,
+      "step": 2620
+    },
+    {
+      "epoch": 0.24593619712371761,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.000187310463349745,
+      "loss": 0.8918,
+      "step": 2625
+    },
+    {
+      "epoch": 0.24640464702300088,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018723061577598113,
+      "loss": 0.9158,
+      "step": 2630
+    },
+    {
+      "epoch": 0.24687309692228415,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018715053490269093,
+      "loss": 0.9127,
+      "step": 2635
+    },
+    {
+      "epoch": 0.24734154682156742,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001870702209440518,
+      "loss": 0.9032,
+      "step": 2640
+    },
+    {
+      "epoch": 0.24780999672085072,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018698967411486455,
+      "loss": 0.9363,
+      "step": 2645
+    },
+    {
+      "epoch": 0.24827844662013399,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00018690889463055283,
+      "loss": 0.9133,
+      "step": 2650
+    },
+    {
+      "epoch": 0.24874689651941725,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018682788270716255,
+      "loss": 0.9094,
+      "step": 2655
+    },
+    {
+      "epoch": 0.24921534641870052,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018674663856136126,
+      "loss": 0.9445,
+      "step": 2660
+    },
+    {
+      "epoch": 0.2496837963179838,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018666516241043753,
+      "loss": 0.9238,
+      "step": 2665
+    },
+    {
+      "epoch": 0.25015224621726706,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00018658345447230054,
+      "loss": 0.9355,
+      "step": 2670
+    },
+    {
+      "epoch": 0.25062069611655036,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00018650151496547933,
+      "loss": 0.9529,
+      "step": 2675
+    },
+    {
+      "epoch": 0.2510891460158336,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018641934410912225,
+      "loss": 0.9202,
+      "step": 2680
+    },
+    {
+      "epoch": 0.2515575959151169,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018633694212299645,
+      "loss": 0.8886,
+      "step": 2685
+    },
+    {
+      "epoch": 0.25202604581440013,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001862543092274872,
+      "loss": 0.9449,
+      "step": 2690
+    },
+    {
+      "epoch": 0.25249449571368343,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001861714456435974,
+      "loss": 0.9299,
+      "step": 2695
+    },
+    {
+      "epoch": 0.25296294561296667,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018608835159294684,
+      "loss": 0.9357,
+      "step": 2700
+    },
+    {
+      "epoch": 0.25343139551224997,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018600502729777176,
+      "loss": 0.9285,
+      "step": 2705
+    },
+    {
+      "epoch": 0.2538998454115332,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018592147298092423,
+      "loss": 0.9083,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2543682953108165,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018583768886587136,
+      "loss": 0.9,
+      "step": 2715
+    },
+    {
+      "epoch": 0.2548367452100998,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018575367517669502,
+      "loss": 0.905,
+      "step": 2720
+    },
+    {
+      "epoch": 0.25530519510938304,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00018566943213809103,
+      "loss": 0.9207,
+      "step": 2725
+    },
+    {
+      "epoch": 0.25577364500866634,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00018558495997536854,
+      "loss": 0.8881,
+      "step": 2730
+    },
+    {
+      "epoch": 0.2562420949079496,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018550025891444968,
+      "loss": 0.9189,
+      "step": 2735
+    },
+    {
+      "epoch": 0.2567105448072329,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001854153291818685,
+      "loss": 0.9398,
+      "step": 2740
+    },
+    {
+      "epoch": 0.2571789947065161,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00018533017100477083,
+      "loss": 0.9478,
+      "step": 2745
+    },
+    {
+      "epoch": 0.2576474446057994,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018524478461091348,
+      "loss": 0.9454,
+      "step": 2750
+    },
+    {
+      "epoch": 0.25811589450508265,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018515917022866353,
+      "loss": 0.9035,
+      "step": 2755
+    },
+    {
+      "epoch": 0.25858434440436595,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001850733280869979,
+      "loss": 0.9178,
+      "step": 2760
+    },
+    {
+      "epoch": 0.25905279430364925,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00018498725841550257,
+      "loss": 0.931,
+      "step": 2765
+    },
+    {
+      "epoch": 0.2595212442029325,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018490096144437214,
+      "loss": 0.9373,
+      "step": 2770
+    },
+    {
+      "epoch": 0.2599896941022158,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001848144374044091,
+      "loss": 0.9457,
+      "step": 2775
+    },
+    {
+      "epoch": 0.260458144001499,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001847276865270232,
+      "loss": 0.8974,
+      "step": 2780
+    },
+    {
+      "epoch": 0.2609265939007823,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018464070904423093,
+      "loss": 0.9225,
+      "step": 2785
+    },
+    {
+      "epoch": 0.26139504380006556,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018455350518865478,
+      "loss": 0.9621,
+      "step": 2790
+    },
+    {
+      "epoch": 0.26186349369934886,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018446607519352273,
+      "loss": 0.8968,
+      "step": 2795
+    },
+    {
+      "epoch": 0.2623319435986321,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018437841929266752,
+      "loss": 0.9271,
+      "step": 2800
+    },
+    {
+      "epoch": 0.2628003934979154,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018429053772052613,
+      "loss": 0.918,
+      "step": 2805
+    },
+    {
+      "epoch": 0.2632688433971987,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001842024307121391,
+      "loss": 0.9287,
+      "step": 2810
+    },
+    {
+      "epoch": 0.26373729329648193,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018411409850314984,
+      "loss": 0.9226,
+      "step": 2815
+    },
+    {
+      "epoch": 0.26420574319576523,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018402554132980413,
+      "loss": 0.9207,
+      "step": 2820
+    },
+    {
+      "epoch": 0.26467419309504847,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018393675942894935,
+      "loss": 0.8867,
+      "step": 2825
+    },
+    {
+      "epoch": 0.26514264299433177,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00018384775303803398,
+      "loss": 0.9316,
+      "step": 2830
+    },
+    {
+      "epoch": 0.265611092893615,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018375852239510695,
+      "loss": 0.9316,
+      "step": 2835
+    },
+    {
+      "epoch": 0.2660795427928983,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001836690677388168,
+      "loss": 0.9294,
+      "step": 2840
+    },
+    {
+      "epoch": 0.26654799269218155,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001835793893084113,
+      "loss": 0.9246,
+      "step": 2845
+    },
+    {
+      "epoch": 0.26701644259146484,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018348948734373674,
+      "loss": 0.9205,
+      "step": 2850
+    },
+    {
+      "epoch": 0.26748489249074814,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018339936208523717,
+      "loss": 0.8891,
+      "step": 2855
+    },
+    {
+      "epoch": 0.2679533423900314,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001833090137739539,
+      "loss": 0.9092,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2684217922893147,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001832184426515247,
+      "loss": 0.9424,
+      "step": 2865
+    },
+    {
+      "epoch": 0.2688902421885979,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00018312764896018347,
+      "loss": 0.9164,
+      "step": 2870
+    },
+    {
+      "epoch": 0.2693586920878812,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018303663294275908,
+      "loss": 0.9174,
+      "step": 2875
+    },
+    {
+      "epoch": 0.26982714198716445,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018294539484267527,
+      "loss": 0.9424,
+      "step": 2880
+    },
+    {
+      "epoch": 0.27029559188644775,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018285393490394963,
+      "loss": 0.911,
+      "step": 2885
+    },
+    {
+      "epoch": 0.270764041785731,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018276225337119305,
+      "loss": 0.943,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2712324916850143,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001826703504896091,
+      "loss": 0.9203,
+      "step": 2895
+    },
+    {
+      "epoch": 0.2717009415842976,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001825782265049933,
+      "loss": 0.9013,
+      "step": 2900
+    },
+    {
+      "epoch": 0.2721693914835808,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018248588166373266,
+      "loss": 0.9295,
+      "step": 2905
+    },
+    {
+      "epoch": 0.2726378413828641,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0001823933162128047,
+      "loss": 0.9222,
+      "step": 2910
+    },
+    {
+      "epoch": 0.27310629128214736,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001823005303997771,
+      "loss": 0.8786,
+      "step": 2915
+    },
+    {
+      "epoch": 0.27357474118143066,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018220752447280675,
+      "loss": 0.9336,
+      "step": 2920
+    },
+    {
+      "epoch": 0.2740431910807139,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001821142986806394,
+      "loss": 0.8944,
+      "step": 2925
+    },
+    {
+      "epoch": 0.2745116409799972,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001820208532726088,
+      "loss": 0.9404,
+      "step": 2930
+    },
+    {
+      "epoch": 0.27498009087928044,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001819271884986359,
+      "loss": 0.9117,
+      "step": 2935
+    },
+    {
+      "epoch": 0.27544854077856373,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018183330460922855,
+      "loss": 0.9196,
+      "step": 2940
+    },
+    {
+      "epoch": 0.27591699067784703,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018173920185548056,
+      "loss": 0.914,
+      "step": 2945
+    },
+    {
+      "epoch": 0.27638544057713027,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001816448804890711,
+      "loss": 0.9055,
+      "step": 2950
+    },
+    {
+      "epoch": 0.27685389047641357,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018155034076226394,
+      "loss": 0.9059,
+      "step": 2955
+    },
+    {
+      "epoch": 0.2773223403756968,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.000181455582927907,
+      "loss": 0.9109,
+      "step": 2960
+    },
+    {
+      "epoch": 0.2777907902749801,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001813606072394314,
+      "loss": 0.9409,
+      "step": 2965
+    },
+    {
+      "epoch": 0.27825924017426334,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001812654139508511,
+      "loss": 0.9086,
+      "step": 2970
+    },
+    {
+      "epoch": 0.27872769007354664,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018117000331676172,
+      "loss": 0.9079,
+      "step": 2975
+    },
+    {
+      "epoch": 0.2791961399728299,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001810743755923405,
+      "loss": 0.9151,
+      "step": 2980
+    },
+    {
+      "epoch": 0.2796645898721132,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00018097853103334505,
+      "loss": 0.9239,
+      "step": 2985
+    },
+    {
+      "epoch": 0.2801330397713965,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018088246989611312,
+      "loss": 0.931,
+      "step": 2990
+    },
+    {
+      "epoch": 0.2806014896706797,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00018078619243756152,
+      "loss": 0.9391,
+      "step": 2995
+    },
+    {
+      "epoch": 0.281069939569963,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018068969891518573,
+      "loss": 0.9131,
+      "step": 3000
+    },
+    {
+      "epoch": 0.28153838946924625,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.000180592989587059,
+      "loss": 0.9385,
+      "step": 3005
+    },
+    {
+      "epoch": 0.28200683936852955,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00018049606471183188,
+      "loss": 0.9403,
+      "step": 3010
+    },
+    {
+      "epoch": 0.2824752892678128,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00018039892454873126,
+      "loss": 0.9198,
+      "step": 3015
+    },
+    {
+      "epoch": 0.2829437391670961,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018030156935755994,
+      "loss": 0.9111,
+      "step": 3020
+    },
+    {
+      "epoch": 0.2834121890663793,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018020399939869573,
+      "loss": 0.9265,
+      "step": 3025
+    },
+    {
+      "epoch": 0.2838806389656626,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0001801062149330909,
+      "loss": 0.9019,
+      "step": 3030
+    },
+    {
+      "epoch": 0.2843490888649459,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001800082162222714,
+      "loss": 0.9214,
+      "step": 3035
+    },
+    {
+      "epoch": 0.28481753876422916,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00017991000352833618,
+      "loss": 0.8984,
+      "step": 3040
+    },
+    {
+      "epoch": 0.28528598866351246,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001798115771139565,
+      "loss": 0.9474,
+      "step": 3045
+    },
+    {
+      "epoch": 0.2857544385627957,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017971293724237516,
+      "loss": 0.8959,
+      "step": 3050
+    },
+    {
+      "epoch": 0.286222888462079,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017961408417740594,
+      "loss": 0.938,
+      "step": 3055
+    },
+    {
+      "epoch": 0.28669133836136224,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001795150181834328,
+      "loss": 0.915,
+      "step": 3060
+    },
+    {
+      "epoch": 0.28715978826064553,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001794157395254091,
+      "loss": 0.9051,
+      "step": 3065
+    },
+    {
+      "epoch": 0.2876282381599288,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001793162484688571,
+      "loss": 0.9939,
+      "step": 3070
+    },
+    {
+      "epoch": 0.28809668805921207,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017921654527986694,
+      "loss": 0.9213,
+      "step": 3075
+    },
+    {
+      "epoch": 0.28856513795849537,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001791166302250963,
+      "loss": 0.9166,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2890335878577786,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017901650357176943,
+      "loss": 0.9027,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2895020377570619,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017891616558767646,
+      "loss": 0.9376,
+      "step": 3090
+    },
+    {
+      "epoch": 0.28997048765634514,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017881561654117277,
+      "loss": 1.0183,
+      "step": 3095
+    },
+    {
+      "epoch": 0.29043893755562844,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001787148567011782,
+      "loss": 0.9606,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2909073874549117,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001786138863371764,
+      "loss": 0.9098,
+      "step": 3105
+    },
+    {
+      "epoch": 0.291375837354195,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017851270571921405,
+      "loss": 0.9015,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2918442872534782,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017841131511790015,
+      "loss": 0.8943,
+      "step": 3115
+    },
+    {
+      "epoch": 0.2923127371527615,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0001783097148044053,
+      "loss": 0.907,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2927811870520448,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00017820790505046103,
+      "loss": 0.9288,
+      "step": 3125
+    },
+    {
+      "epoch": 0.29324963695132805,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001781058861283589,
+      "loss": 0.9369,
+      "step": 3130
+    },
+    {
+      "epoch": 0.29371808685061135,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00017800365831095006,
+      "loss": 0.9314,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2941865367498946,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017790122187164422,
+      "loss": 0.9102,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2946549866491779,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001777985770844091,
+      "loss": 0.9221,
+      "step": 3145
+    },
+    {
+      "epoch": 0.2951234365484611,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00017769572422376964,
+      "loss": 0.9235,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2955918864477444,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001775926635648073,
+      "loss": 0.9155,
+      "step": 3155
+    },
+    {
+      "epoch": 0.29606033634702766,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00017748939538315929,
+      "loss": 0.9374,
+      "step": 3160
+    },
+    {
+      "epoch": 0.29652878624631096,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00017738591995501783,
+      "loss": 0.9158,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2969972361455942,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017728223755712942,
+      "loss": 0.9084,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2974656860448775,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001771783484667941,
+      "loss": 0.8989,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2979341359441608,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00017707425296186474,
+      "loss": 0.9225,
+      "step": 3180
+    },
+    {
+      "epoch": 0.29840258584344403,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017696995132074626,
+      "loss": 0.918,
+      "step": 3185
+    },
+    {
+      "epoch": 0.29887103574272733,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001768654438223949,
+      "loss": 0.9197,
+      "step": 3190
+    },
+    {
+      "epoch": 0.29933948564201057,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017676073074631741,
+      "loss": 0.9206,
+      "step": 3195
+    },
+    {
+      "epoch": 0.29980793554129387,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017665581237257043,
+      "loss": 0.9458,
+      "step": 3200
+    },
+    {
+      "epoch": 0.3002763854405771,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00017655068898175965,
+      "loss": 1.0539,
+      "step": 3205
+    },
+    {
+      "epoch": 0.3007448353398604,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001764453608550391,
+      "loss": 0.9123,
+      "step": 3210
+    },
+    {
+      "epoch": 0.30121328523914365,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017633982827411032,
+      "loss": 0.9205,
+      "step": 3215
+    },
+    {
+      "epoch": 0.30168173513842694,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001762340915212217,
+      "loss": 0.9407,
+      "step": 3220
+    },
+    {
+      "epoch": 0.30215018503771024,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017612815087916773,
+      "loss": 0.9268,
+      "step": 3225
+    },
+    {
+      "epoch": 0.3026186349369935,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00017602200663128815,
+      "loss": 0.9035,
+      "step": 3230
+    },
+    {
+      "epoch": 0.3030870848362768,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017591565906146722,
+      "loss": 0.9534,
+      "step": 3235
+    },
+    {
+      "epoch": 0.30355553473556,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017580910845413315,
+      "loss": 0.9116,
+      "step": 3240
+    },
+    {
+      "epoch": 0.3040239846348433,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017570235509425694,
+      "loss": 0.9181,
+      "step": 3245
+    },
+    {
+      "epoch": 0.30449243453412655,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017559539926735205,
+      "loss": 0.918,
+      "step": 3250
+    },
+    {
+      "epoch": 0.30496088443340985,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001754882412594733,
+      "loss": 0.9134,
+      "step": 3255
+    },
+    {
+      "epoch": 0.3054293343326931,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00017538088135721637,
+      "loss": 0.9195,
+      "step": 3260
+    },
+    {
+      "epoch": 0.3058977842319764,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00017527331984771682,
+      "loss": 0.9192,
+      "step": 3265
+    },
+    {
+      "epoch": 0.3063662341312597,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017516555701864946,
+      "loss": 0.9033,
+      "step": 3270
+    },
+    {
+      "epoch": 0.3068346840305429,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001750575931582275,
+      "loss": 0.8793,
+      "step": 3275
+    },
+    {
+      "epoch": 0.3073031339298262,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017494942855520183,
+      "loss": 0.925,
+      "step": 3280
+    },
+    {
+      "epoch": 0.30777158382910946,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017484106349886024,
+      "loss": 0.9184,
+      "step": 3285
+    },
+    {
+      "epoch": 0.30824003372839276,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017473249827902655,
+      "loss": 0.9181,
+      "step": 3290
+    },
+    {
+      "epoch": 0.308708483627676,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017462373318606007,
+      "loss": 0.8968,
+      "step": 3295
+    },
+    {
+      "epoch": 0.3091769335269593,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017451476851085458,
+      "loss": 0.9257,
+      "step": 3300
+    },
+    {
+      "epoch": 0.30964538342624254,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017440560454483762,
+      "loss": 0.9103,
+      "step": 3305
+    },
+    {
+      "epoch": 0.31011383332552583,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001742962415799698,
+      "loss": 0.8803,
+      "step": 3310
+    },
+    {
+      "epoch": 0.31058228322480913,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00017418667990874392,
+      "loss": 0.9336,
+      "step": 3315
+    },
+    {
+      "epoch": 0.31105073312409237,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017407691982418425,
+      "loss": 0.8923,
+      "step": 3320
+    },
+    {
+      "epoch": 0.31151918302337567,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001739669616198457,
+      "loss": 0.9206,
+      "step": 3325
+    },
+    {
+      "epoch": 0.3119876329226589,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00017385680558981302,
+      "loss": 0.9134,
+      "step": 3330
+    },
+    {
+      "epoch": 0.3124560828219422,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017374645202870013,
+      "loss": 0.9002,
+      "step": 3335
+    },
+    {
+      "epoch": 0.31292453272122545,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017363590123164918,
+      "loss": 0.8985,
+      "step": 3340
+    },
+    {
+      "epoch": 0.31339298262050874,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00017352515349432983,
+      "loss": 0.9056,
+      "step": 3345
+    },
+    {
+      "epoch": 0.313861432519792,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00017341420911293854,
+      "loss": 0.9401,
+      "step": 3350
+    },
+    {
+      "epoch": 0.3143298824190753,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001733030683841976,
+      "loss": 0.8958,
+      "step": 3355
+    },
+    {
+      "epoch": 0.3147983323183586,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017319173160535448,
+      "loss": 0.901,
+      "step": 3360
+    },
+    {
+      "epoch": 0.3152667822176418,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.000173080199074181,
+      "loss": 0.943,
+      "step": 3365
+    },
+    {
+      "epoch": 0.3157352321169251,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001729684710889725,
+      "loss": 0.9003,
+      "step": 3370
+    },
+    {
+      "epoch": 0.31620368201620835,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001728565479485471,
+      "loss": 0.9535,
+      "step": 3375
+    },
+    {
+      "epoch": 0.31667213191549165,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017274442995224474,
+      "loss": 0.9193,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3171405818147749,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001726321173999267,
+      "loss": 0.9472,
+      "step": 3385
+    },
+    {
+      "epoch": 0.3176090317140582,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00017251961059197446,
+      "loss": 0.9295,
+      "step": 3390
+    },
+    {
+      "epoch": 0.31807748161334143,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0001724069098292891,
+      "loss": 0.8673,
+      "step": 3395
+    },
+    {
+      "epoch": 0.3185459315126247,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00017229401541329038,
+      "loss": 0.9175,
+      "step": 3400
+    },
+    {
+      "epoch": 0.319014381411908,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017218092764591608,
+      "loss": 0.9012,
+      "step": 3405
+    },
+    {
+      "epoch": 0.31948283131119126,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.000172067646829621,
+      "loss": 0.9051,
+      "step": 3410
+    },
+    {
+      "epoch": 0.31995128121047456,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00017195417326737635,
+      "loss": 0.9079,
+      "step": 3415
+    },
+    {
+      "epoch": 0.3204197311097578,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017184050726266873,
+      "loss": 0.9397,
+      "step": 3420
+    },
+    {
+      "epoch": 0.3208881810090411,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017172664911949953,
+      "loss": 0.9497,
+      "step": 3425
+    },
+    {
+      "epoch": 0.32135663090832434,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017161259914238398,
+      "loss": 0.9032,
+      "step": 3430
+    },
+    {
+      "epoch": 0.32182508080760763,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017149835763635035,
+      "loss": 0.9019,
+      "step": 3435
+    },
+    {
+      "epoch": 0.3222935307068909,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001713839249069392,
+      "loss": 0.8877,
+      "step": 3440
+    },
+    {
+      "epoch": 0.32276198060617417,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017126930126020246,
+      "loss": 0.9301,
+      "step": 3445
+    },
+    {
+      "epoch": 0.32323043050545747,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017115448700270278,
+      "loss": 0.9423,
+      "step": 3450
+    },
+    {
+      "epoch": 0.3236988804047407,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017103948244151244,
+      "loss": 0.9034,
+      "step": 3455
+    },
+    {
+      "epoch": 0.324167330304024,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017092428788421287,
+      "loss": 0.8691,
+      "step": 3460
+    },
+    {
+      "epoch": 0.32463578020330724,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017080890363889347,
+      "loss": 0.8703,
+      "step": 3465
+    },
+    {
+      "epoch": 0.32510423010259054,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001706933300141511,
+      "loss": 0.9055,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3255726800018738,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017057756731908908,
+      "loss": 0.8936,
+      "step": 3475
+    },
+    {
+      "epoch": 0.3260411299011571,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017046161586331632,
+      "loss": 0.9462,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3265095798004403,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00017034547595694673,
+      "loss": 0.909,
+      "step": 3485
+    },
+    {
+      "epoch": 0.3269780296997236,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001702291479105981,
+      "loss": 0.8918,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3274464795990069,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017011263203539149,
+      "loss": 0.9313,
+      "step": 3495
+    },
+    {
+      "epoch": 0.32791492949829015,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001699959286429502,
+      "loss": 0.9146,
+      "step": 3500
+    },
+    {
+      "epoch": 0.32838337939757345,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00016987903804539917,
+      "loss": 0.9142,
+      "step": 3505
+    },
+    {
+      "epoch": 0.3288518292968567,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016976196055536392,
+      "loss": 0.9114,
+      "step": 3510
+    },
+    {
+      "epoch": 0.32932027919614,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016964469648596996,
+      "loss": 1.0432,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3297887290954232,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016952724615084164,
+      "loss": 0.9197,
+      "step": 3520
+    },
+    {
+      "epoch": 0.3302571789947065,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016940960986410158,
+      "loss": 0.8996,
+      "step": 3525
+    },
+    {
+      "epoch": 0.33072562889398976,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001692917879403697,
+      "loss": 0.8883,
+      "step": 3530
+    },
+    {
+      "epoch": 0.33119407879327306,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016917378069476242,
+      "loss": 0.8973,
+      "step": 3535
+    },
+    {
+      "epoch": 0.33166252869255636,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001690555884428918,
+      "loss": 0.939,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3321309785918396,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016893721150086473,
+      "loss": 0.9165,
+      "step": 3545
+    },
+    {
+      "epoch": 0.3325994284911229,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016881865018528197,
+      "loss": 0.8865,
+      "step": 3550
+    },
+    {
+      "epoch": 0.33306787839040614,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001686999048132375,
+      "loss": 0.8801,
+      "step": 3555
+    },
+    {
+      "epoch": 0.33353632828968943,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001685809757023175,
+      "loss": 0.8828,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3340047781889727,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001684618631705996,
+      "loss": 0.8856,
+      "step": 3565
+    },
+    {
+      "epoch": 0.33447322808825597,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001683425675366519,
+      "loss": 0.892,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3349416779875392,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016822308911953232,
+      "loss": 0.9194,
+      "step": 3575
+    },
+    {
+      "epoch": 0.3354101278868225,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016810342823878762,
+      "loss": 0.902,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3358785777861058,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00016798358521445247,
+      "loss": 0.886,
+      "step": 3585
+    },
+    {
+      "epoch": 0.33634702768538904,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016786356036704874,
+      "loss": 0.9143,
+      "step": 3590
+    },
+    {
+      "epoch": 0.33681547758467234,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016774335401758468,
+      "loss": 0.8993,
+      "step": 3595
+    },
+    {
+      "epoch": 0.3372839274839556,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001676229664875538,
+      "loss": 0.8811,
+      "step": 3600
+    },
+    {
+      "epoch": 0.3377523773832389,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016750239809893433,
+      "loss": 0.9153,
+      "step": 3605
+    },
+    {
+      "epoch": 0.3382208272825221,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016738164917418812,
+      "loss": 0.8942,
+      "step": 3610
+    },
+    {
+      "epoch": 0.3386892771818054,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001672607200362599,
+      "loss": 0.8946,
+      "step": 3615
+    },
+    {
+      "epoch": 0.33915772708108866,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016713961100857635,
+      "loss": 0.9014,
+      "step": 3620
+    },
+    {
+      "epoch": 0.33962617698037195,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016701832241504532,
+      "loss": 0.9209,
+      "step": 3625
+    },
+    {
+      "epoch": 0.34009462687965525,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016689685458005487,
+      "loss": 0.9042,
+      "step": 3630
+    },
+    {
+      "epoch": 0.3405630767789385,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016677520782847248,
+      "loss": 0.8944,
+      "step": 3635
+    },
+    {
+      "epoch": 0.3410315266782218,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016665338248564407,
+      "loss": 0.9111,
+      "step": 3640
+    },
+    {
+      "epoch": 0.341499976577505,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00016653137887739328,
+      "loss": 0.8778,
+      "step": 3645
+    },
+    {
+      "epoch": 0.3419684264767883,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00016640919733002053,
+      "loss": 0.8687,
+      "step": 3650
+    },
+    {
+      "epoch": 0.34243687637607156,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001662868381703021,
+      "loss": 0.9047,
+      "step": 3655
+    },
+    {
+      "epoch": 0.34290532627535486,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016616430172548932,
+      "loss": 0.8688,
+      "step": 3660
+    },
+    {
+      "epoch": 0.3433737761746381,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001660415883233076,
+      "loss": 0.8891,
+      "step": 3665
+    },
+    {
+      "epoch": 0.3438422260739214,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016591869829195573,
+      "loss": 0.89,
+      "step": 3670
+    },
+    {
+      "epoch": 0.34431067597320464,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00016579563196010487,
+      "loss": 0.9202,
+      "step": 3675
+    },
+    {
+      "epoch": 0.34477912587248793,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016567238965689766,
+      "loss": 0.862,
+      "step": 3680
+    },
+    {
+      "epoch": 0.34524757577177123,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016554897171194738,
+      "loss": 0.9204,
+      "step": 3685
+    },
+    {
+      "epoch": 0.34571602567105447,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001654253784553371,
+      "loss": 0.9066,
+      "step": 3690
+    },
+    {
+      "epoch": 0.34618447557033777,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016530161021761875,
+      "loss": 0.9057,
+      "step": 3695
+    },
+    {
+      "epoch": 0.346652925469621,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016517766732981228,
+      "loss": 0.9034,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3471213753689043,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00016505355012340464,
+      "loss": 0.9083,
+      "step": 3705
+    },
+    {
+      "epoch": 0.34758982526818755,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001649292589303491,
+      "loss": 0.9135,
+      "step": 3710
+    },
+    {
+      "epoch": 0.34805827516747084,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016480479408306428,
+      "loss": 0.8928,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3485267250667541,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00016468015591443314,
+      "loss": 0.9567,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3489951749660374,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016455534475780218,
+      "loss": 0.8747,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3494636248653207,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016443036094698074,
+      "loss": 0.9174,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3499320747646039,
+      "grad_norm": 5.0,
+      "learning_rate": 0.00016430520481623977,
+      "loss": 1.063,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3504005246638872,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016417987670031106,
+      "loss": 0.8845,
+      "step": 3740
+    },
+    {
+      "epoch": 0.35086897456317045,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016405437693438652,
+      "loss": 0.9233,
+      "step": 3745
+    },
+    {
+      "epoch": 0.35133742446245375,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016392870585411702,
+      "loss": 0.9112,
+      "step": 3750
+    },
+    {
+      "epoch": 0.351805874361737,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001638028637956117,
+      "loss": 0.9126,
+      "step": 3755
+    },
+    {
+      "epoch": 0.3522743242610203,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016367685109543688,
+      "loss": 0.8878,
+      "step": 3760
+    },
+    {
+      "epoch": 0.35274277416030353,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016355066809061537,
+      "loss": 0.8896,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3532112240595868,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016342431511862538,
+      "loss": 0.9075,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3536796739588701,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016329779251739982,
+      "loss": 0.8908,
+      "step": 3775
+    },
+    {
+      "epoch": 0.35414812385815336,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001631711006253251,
+      "loss": 0.8959,
+      "step": 3780
+    },
+    {
+      "epoch": 0.35461657375743666,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00016304423978124054,
+      "loss": 0.8979,
+      "step": 3785
+    },
+    {
+      "epoch": 0.3550850236567199,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00016291721032443723,
+      "loss": 0.9007,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3555534735560032,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016279001259465734,
+      "loss": 0.928,
+      "step": 3795
+    },
+    {
+      "epoch": 0.35602192345528644,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00016266264693209293,
+      "loss": 0.8735,
+      "step": 3800
+    },
+    {
+      "epoch": 0.35649037335456973,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00016253511367738535,
+      "loss": 0.909,
+      "step": 3805
+    },
+    {
+      "epoch": 0.356958823253853,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016240741317162408,
+      "loss": 0.8885,
+      "step": 3810
+    },
+    {
+      "epoch": 0.35742727315313627,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016227954575634594,
+      "loss": 0.9338,
+      "step": 3815
+    },
+    {
+      "epoch": 0.35789572305241957,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016215151177353417,
+      "loss": 0.9067,
+      "step": 3820
+    },
+    {
+      "epoch": 0.3583641729517028,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016202331156561748,
+      "loss": 0.9244,
+      "step": 3825
+    },
+    {
+      "epoch": 0.3588326228509861,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001618949454754691,
+      "loss": 0.907,
+      "step": 3830
+    },
+    {
+      "epoch": 0.35930107275026935,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00016176641384640607,
+      "loss": 0.8841,
+      "step": 3835
+    },
+    {
+      "epoch": 0.35976952264955264,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00016163771702218803,
+      "loss": 0.9133,
+      "step": 3840
+    },
+    {
+      "epoch": 0.3602379725488359,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001615088553470164,
+      "loss": 0.9089,
+      "step": 3845
+    },
+    {
+      "epoch": 0.3607064224481192,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00016137982916553367,
+      "loss": 0.8985,
+      "step": 3850
+    },
+    {
+      "epoch": 0.3611748723474024,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00016125063882282216,
+      "loss": 0.9012,
+      "step": 3855
+    },
+    {
+      "epoch": 0.3616433222466857,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001611212846644033,
+      "loss": 0.9211,
+      "step": 3860
+    },
+    {
+      "epoch": 0.362111772145969,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001609917670362366,
+      "loss": 0.905,
+      "step": 3865
+    },
+    {
+      "epoch": 0.36258022204525225,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016086208628471877,
+      "loss": 0.948,
+      "step": 3870
+    },
+    {
+      "epoch": 0.36304867194453555,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00016073224275668294,
+      "loss": 0.8925,
+      "step": 3875
+    },
+    {
+      "epoch": 0.3635171218438188,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016060223679939742,
+      "loss": 0.8875,
+      "step": 3880
+    },
+    {
+      "epoch": 0.3639855717431021,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016047206876056497,
+      "loss": 0.8999,
+      "step": 3885
+    },
+    {
+      "epoch": 0.36445402164238533,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016034173898832189,
+      "loss": 0.8879,
+      "step": 3890
+    },
+    {
+      "epoch": 0.3649224715416686,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00016021124783123706,
+      "loss": 0.8873,
+      "step": 3895
+    },
+    {
+      "epoch": 0.36539092144095187,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001600805956383109,
+      "loss": 0.8845,
+      "step": 3900
+    },
+    {
+      "epoch": 0.36585937134023516,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00015994978275897452,
+      "loss": 0.9079,
+      "step": 3905
+    },
+    {
+      "epoch": 0.36632782123951846,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001598188095430889,
+      "loss": 0.9029,
+      "step": 3910
+    },
+    {
+      "epoch": 0.3667962711388017,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015968767634094374,
+      "loss": 0.9574,
+      "step": 3915
+    },
+    {
+      "epoch": 0.367264721038085,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001595563835032567,
+      "loss": 0.8818,
+      "step": 3920
+    },
+    {
+      "epoch": 0.36773317093736824,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001594249313811723,
+      "loss": 0.9133,
+      "step": 3925
+    },
+    {
+      "epoch": 0.36820162083665153,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015929332032626112,
+      "loss": 0.9072,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3686700707359348,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00015916155069051882,
+      "loss": 1.1787,
+      "step": 3935
+    },
+    {
+      "epoch": 0.36913852063521807,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0001590296228263652,
+      "loss": 0.8941,
+      "step": 3940
+    },
+    {
+      "epoch": 0.3696069705345013,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00015889753708664314,
+      "loss": 0.9176,
+      "step": 3945
+    },
+    {
+      "epoch": 0.3700754204337846,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015876529382461793,
+      "loss": 0.8835,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3705438703330679,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00015863289339397598,
+      "loss": 0.8881,
+      "step": 3955
+    },
+    {
+      "epoch": 0.37101232023235114,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015850033614882417,
+      "loss": 0.9039,
+      "step": 3960
+    },
+    {
+      "epoch": 0.37148077013163444,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001583676224436887,
+      "loss": 1.0914,
+      "step": 3965
+    },
+    {
+      "epoch": 0.3719492200309177,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015823475263351434,
+      "loss": 0.8746,
+      "step": 3970
+    },
+    {
+      "epoch": 0.372417669930201,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00015810172707366325,
+      "loss": 0.9153,
+      "step": 3975
+    },
+    {
+      "epoch": 0.3728861198294842,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00015796854611991419,
+      "loss": 0.9237,
+      "step": 3980
+    },
+    {
+      "epoch": 0.3733545697287675,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015783521012846153,
+      "loss": 0.8913,
+      "step": 3985
+    },
+    {
+      "epoch": 0.37382301962805076,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015770171945591422,
+      "loss": 0.9,
+      "step": 3990
+    },
+    {
+      "epoch": 0.37429146952733405,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015756807445929511,
+      "loss": 0.8885,
+      "step": 3995
+    },
+    {
+      "epoch": 0.37475991942661735,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00015743427549603947,
+      "loss": 0.8945,
+      "step": 4000
+    },
+    {
+      "epoch": 0.3752283693259006,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001573003229239947,
+      "loss": 0.8817,
+      "step": 4005
+    },
+    {
+      "epoch": 0.3756968192251839,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015716621710141877,
+      "loss": 0.9255,
+      "step": 4010
+    },
+    {
+      "epoch": 0.3761652691244671,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015703195838697966,
+      "loss": 0.939,
+      "step": 4015
+    },
+    {
+      "epoch": 0.3766337190237504,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00015689754713975416,
+      "loss": 0.8979,
+      "step": 4020
+    },
+    {
+      "epoch": 0.37710216892303366,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015676298371922713,
+      "loss": 0.8675,
+      "step": 4025
+    },
+    {
+      "epoch": 0.37757061882231696,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001566282684852903,
+      "loss": 0.9282,
+      "step": 4030
+    },
+    {
+      "epoch": 0.3780390687216002,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015649340179824158,
+      "loss": 0.9367,
+      "step": 4035
+    },
+    {
+      "epoch": 0.3785075186208835,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015635838401878373,
+      "loss": 0.8955,
+      "step": 4040
+    },
+    {
+      "epoch": 0.3789759685201668,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001562232155080238,
+      "loss": 0.9024,
+      "step": 4045
+    },
+    {
+      "epoch": 0.37944441841945004,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015608789662747188,
+      "loss": 0.8621,
+      "step": 4050
+    },
+    {
+      "epoch": 0.37991286831873333,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00015595242773904026,
+      "loss": 0.8734,
+      "step": 4055
+    },
+    {
+      "epoch": 0.3803813182180166,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001558168092050424,
+      "loss": 1.0425,
+      "step": 4060
+    },
+    {
+      "epoch": 0.38084976811729987,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015568104138819202,
+      "loss": 0.9059,
+      "step": 4065
+    },
+    {
+      "epoch": 0.3813182180165831,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00015554512465160207,
+      "loss": 0.8936,
+      "step": 4070
+    },
+    {
+      "epoch": 0.3817866679158664,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001554090593587838,
+      "loss": 0.8847,
+      "step": 4075
+    },
+    {
+      "epoch": 0.38225511781514965,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00015527284587364579,
+      "loss": 0.8846,
+      "step": 4080
+    },
+    {
+      "epoch": 0.38272356771443294,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001551364845604929,
+      "loss": 0.9156,
+      "step": 4085
+    },
+    {
+      "epoch": 0.38319201761371624,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015499997578402548,
+      "loss": 0.9226,
+      "step": 4090
+    },
+    {
+      "epoch": 0.3836604675129995,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015486331990933818,
+      "loss": 0.9058,
+      "step": 4095
+    },
+    {
+      "epoch": 0.3841289174122828,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015472651730191904,
+      "loss": 0.9001,
+      "step": 4100
+    },
+    {
+      "epoch": 0.384597367311566,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015458956832764865,
+      "loss": 0.8936,
+      "step": 4105
+    },
+    {
+      "epoch": 0.3850658172108493,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015445247335279895,
+      "loss": 0.916,
+      "step": 4110
+    },
+    {
+      "epoch": 0.38553426711013256,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00015431523274403244,
+      "loss": 0.9164,
+      "step": 4115
+    },
+    {
+      "epoch": 0.38600271700941585,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015417784686840103,
+      "loss": 0.8978,
+      "step": 4120
+    },
+    {
+      "epoch": 0.3864711669086991,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015404031609334527,
+      "loss": 0.9038,
+      "step": 4125
+    },
+    {
+      "epoch": 0.3869396168079824,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001539026407866931,
+      "loss": 0.914,
+      "step": 4130
+    },
+    {
+      "epoch": 0.38740806670726563,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015376482131665924,
+      "loss": 0.9174,
+      "step": 4135
+    },
+    {
+      "epoch": 0.3878765166065489,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00015362685805184366,
+      "loss": 0.9186,
+      "step": 4140
+    },
+    {
+      "epoch": 0.3883449665058322,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015348875136123117,
+      "loss": 0.9066,
+      "step": 4145
+    },
+    {
+      "epoch": 0.38881341640511546,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00015335050161419003,
+      "loss": 0.9188,
+      "step": 4150
+    },
+    {
+      "epoch": 0.38928186630439876,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001532121091804713,
+      "loss": 0.9164,
+      "step": 4155
+    },
+    {
+      "epoch": 0.389750316203682,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001530735744302073,
+      "loss": 0.8988,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3902187661029653,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015293489773391138,
+      "loss": 0.9127,
+      "step": 4165
+    },
+    {
+      "epoch": 0.39068721600224854,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00015279607946247625,
+      "loss": 0.8976,
+      "step": 4170
+    },
+    {
+      "epoch": 0.39115566590153183,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015265711998717338,
+      "loss": 0.895,
+      "step": 4175
+    },
+    {
+      "epoch": 0.3916241158008151,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001525180196796519,
+      "loss": 0.9064,
+      "step": 4180
+    },
+    {
+      "epoch": 0.39209256570009837,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015237877891193747,
+      "loss": 0.8892,
+      "step": 4185
+    },
+    {
+      "epoch": 0.39256101559938167,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00015223939805643162,
+      "loss": 0.8649,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3930294654986649,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015209987748591043,
+      "loss": 0.889,
+      "step": 4195
+    },
+    {
+      "epoch": 0.3934979153979482,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015196021757352357,
+      "loss": 0.882,
+      "step": 4200
+    },
+    {
+      "epoch": 0.39396636529723145,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015182041869279354,
+      "loss": 0.8787,
+      "step": 4205
+    },
+    {
+      "epoch": 0.39443481519651474,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015168048121761438,
+      "loss": 0.8888,
+      "step": 4210
+    },
+    {
+      "epoch": 0.394903265095798,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00015154040552225095,
+      "loss": 0.9049,
+      "step": 4215
+    },
+    {
+      "epoch": 0.3953717149950813,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001514001919813376,
+      "loss": 0.9112,
+      "step": 4220
+    },
+    {
+      "epoch": 0.3958401648943645,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015125984096987755,
+      "loss": 0.9035,
+      "step": 4225
+    },
+    {
+      "epoch": 0.3963086147936478,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001511193528632415,
+      "loss": 0.8673,
+      "step": 4230
+    },
+    {
+      "epoch": 0.3967770646929311,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015097872803716692,
+      "loss": 0.9297,
+      "step": 4235
+    },
+    {
+      "epoch": 0.39724551459221435,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015083796686775694,
+      "loss": 0.9039,
+      "step": 4240
+    },
+    {
+      "epoch": 0.39771396449149765,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015069706973147926,
+      "loss": 0.8993,
+      "step": 4245
+    },
+    {
+      "epoch": 0.3981824143907809,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015055603700516533,
+      "loss": 0.8693,
+      "step": 4250
+    },
+    {
+      "epoch": 0.3986508642900642,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015041486906600918,
+      "loss": 0.8812,
+      "step": 4255
+    },
+    {
+      "epoch": 0.39911931418934743,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015027356629156646,
+      "loss": 0.8836,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3995877640886307,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00015013212905975345,
+      "loss": 0.881,
+      "step": 4265
+    },
+    {
+      "epoch": 0.40005621398791397,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014999055774884604,
+      "loss": 0.8774,
+      "step": 4270
+    },
+    {
+      "epoch": 0.40052466388719726,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001498488527374788,
+      "loss": 0.9092,
+      "step": 4275
+    },
+    {
+      "epoch": 0.40099311378648056,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001497070144046437,
+      "loss": 0.8907,
+      "step": 4280
+    },
+    {
+      "epoch": 0.4014615636857638,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014956504312968943,
+      "loss": 0.9038,
+      "step": 4285
+    },
+    {
+      "epoch": 0.4019300135850471,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014942293929232019,
+      "loss": 1.023,
+      "step": 4290
+    },
+    {
+      "epoch": 0.40239846348433034,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014928070327259476,
+      "loss": 0.8826,
+      "step": 4295
+    },
+    {
+      "epoch": 0.40286691338361363,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001491383354509253,
+      "loss": 0.8968,
+      "step": 4300
+    },
+    {
+      "epoch": 0.4033353632828969,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014899583620807668,
+      "loss": 0.8791,
+      "step": 4305
+    },
+    {
+      "epoch": 0.40380381318218017,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001488532059251651,
+      "loss": 0.9137,
+      "step": 4310
+    },
+    {
+      "epoch": 0.4042722630814634,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014871044498365735,
+      "loss": 0.8906,
+      "step": 4315
+    },
+    {
+      "epoch": 0.4047407129807467,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00014856755376536953,
+      "loss": 0.908,
+      "step": 4320
+    },
+    {
+      "epoch": 0.40520916288003,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.0001484245326524663,
+      "loss": 0.8763,
+      "step": 4325
+    },
+    {
+      "epoch": 0.40567761277931325,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001482813820274596,
+      "loss": 0.9505,
+      "step": 4330
+    },
+    {
+      "epoch": 0.40614606267859654,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001481381022732079,
+      "loss": 0.8893,
+      "step": 4335
+    },
+    {
+      "epoch": 0.4066145125778798,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001479946937729149,
+      "loss": 0.9082,
+      "step": 4340
+    },
+    {
+      "epoch": 0.4070829624771631,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00014785115691012864,
+      "loss": 0.9011,
+      "step": 4345
+    },
+    {
+      "epoch": 0.4075514123764463,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001477074920687406,
+      "loss": 0.8987,
+      "step": 4350
+    },
+    {
+      "epoch": 0.4080198622757296,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0001475636996329844,
+      "loss": 0.8585,
+      "step": 4355
+    },
+    {
+      "epoch": 0.40848831217501286,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00014741977998743497,
+      "loss": 0.8986,
+      "step": 4360
+    },
+    {
+      "epoch": 0.40895676207429615,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00014727573351700743,
+      "loss": 0.9062,
+      "step": 4365
+    },
+    {
+      "epoch": 0.40942521197357945,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00014713156060695615,
+      "loss": 0.9112,
+      "step": 4370
+    },
+    {
+      "epoch": 0.4098936618728627,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001469872616428736,
+      "loss": 0.9111,
+      "step": 4375
+    },
+    {
+      "epoch": 0.410362111772146,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001468428370106895,
+      "loss": 0.8804,
+      "step": 4380
+    },
+    {
+      "epoch": 0.41083056167142923,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00014669828709666947,
+      "loss": 0.8998,
+      "step": 4385
+    },
+    {
+      "epoch": 0.4112990115707125,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001465536122874144,
+      "loss": 0.89,
+      "step": 4390
+    },
+    {
+      "epoch": 0.41176746146999577,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00014640881296985913,
+      "loss": 0.8845,
+      "step": 4395
+    },
+    {
+      "epoch": 0.41223591136927906,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014626388953127144,
+      "loss": 0.8634,
+      "step": 4400
+    },
+    {
+      "epoch": 0.4127043612685623,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00014611884235925119,
+      "loss": 0.9032,
+      "step": 4405
+    },
+    {
+      "epoch": 0.4131728111678456,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001459736718417291,
+      "loss": 0.9157,
+      "step": 4410
+    },
+    {
+      "epoch": 0.4136412610671289,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001458283783669658,
+      "loss": 0.903,
+      "step": 4415
+    },
+    {
+      "epoch": 0.41410971096641214,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001456829623235507,
+      "loss": 0.8987,
+      "step": 4420
+    },
+    {
+      "epoch": 0.41457816086569543,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014553742410040117,
+      "loss": 0.9064,
+      "step": 4425
+    },
+    {
+      "epoch": 0.4150466107649787,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001453917640867612,
+      "loss": 0.8856,
+      "step": 4430
+    },
+    {
+      "epoch": 0.41551506066426197,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00014524598267220062,
+      "loss": 0.9017,
+      "step": 4435
+    },
+    {
+      "epoch": 0.4159835105635452,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00014510008024661388,
+      "loss": 0.8495,
+      "step": 4440
+    },
+    {
+      "epoch": 0.4164519604628285,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014495405720021907,
+      "loss": 0.8981,
+      "step": 4445
+    },
+    {
+      "epoch": 0.41692041036211175,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014480791392355695,
+      "loss": 1.005,
+      "step": 4450
+    },
+    {
+      "epoch": 0.41738886026139504,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00014466165080748982,
+      "loss": 0.8967,
+      "step": 4455
+    },
+    {
+      "epoch": 0.41785731016067834,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014451526824320042,
+      "loss": 0.8819,
+      "step": 4460
+    },
+    {
+      "epoch": 0.4183257600599616,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00014436876662219102,
+      "loss": 0.9043,
+      "step": 4465
+    },
+    {
+      "epoch": 0.4187942099592449,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0001442221463362823,
+      "loss": 0.9211,
+      "step": 4470
+    },
+    {
+      "epoch": 0.4192626598585281,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00014407540777761231,
+      "loss": 0.906,
+      "step": 4475
+    },
+    {
+      "epoch": 0.4197311097578114,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.0001439285513386354,
+      "loss": 0.8962,
+      "step": 4480
+    },
+    {
+      "epoch": 0.42019955965709466,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014378157741212126,
+      "loss": 0.9116,
+      "step": 4485
+    },
+    {
+      "epoch": 0.42066800955637795,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001436344863911537,
+      "loss": 0.8853,
+      "step": 4490
+    },
+    {
+      "epoch": 0.4211364594556612,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001434872786691298,
+      "loss": 0.9244,
+      "step": 4495
+    },
+    {
+      "epoch": 0.4216049093549445,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00014333995463975863,
+      "loss": 0.9226,
+      "step": 4500
+    },
+    {
+      "epoch": 0.4220733592542278,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014319251469706052,
+      "loss": 0.8893,
+      "step": 4505
+    },
+    {
+      "epoch": 0.422541809153511,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014304495923536564,
+      "loss": 0.8685,
+      "step": 4510
+    },
+    {
+      "epoch": 0.4230102590527943,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014289728864931315,
+      "loss": 0.8663,
+      "step": 4515
+    },
+    {
+      "epoch": 0.42347870895207756,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001427495033338502,
+      "loss": 0.8739,
+      "step": 4520
+    },
+    {
+      "epoch": 0.42394715885136086,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00014260160368423068,
+      "loss": 0.8792,
+      "step": 4525
+    },
+    {
+      "epoch": 0.4244156087506441,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00014245359009601436,
+      "loss": 0.8911,
+      "step": 4530
+    },
+    {
+      "epoch": 0.4248840586499274,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001423054629650656,
+      "loss": 0.9243,
+      "step": 4535
+    },
+    {
+      "epoch": 0.42535250854921064,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00014215722268755268,
+      "loss": 0.9,
+      "step": 4540
+    },
+    {
+      "epoch": 0.42582095844849394,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001420088696599462,
+      "loss": 0.8846,
+      "step": 4545
+    },
+    {
+      "epoch": 0.42628940834777723,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00014186040427901847,
+      "loss": 0.8815,
+      "step": 4550
+    },
+    {
+      "epoch": 0.4267578582470605,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00014171182694184233,
+      "loss": 0.9076,
+      "step": 4555
+    },
+    {
+      "epoch": 0.42722630814634377,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014156313804578993,
+      "loss": 0.9037,
+      "step": 4560
+    },
+    {
+      "epoch": 0.427694758045627,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014141433798853188,
+      "loss": 0.9161,
+      "step": 4565
+    },
+    {
+      "epoch": 0.4281632079449103,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014126542716803603,
+      "loss": 0.8896,
+      "step": 4570
+    },
+    {
+      "epoch": 0.42863165784419355,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00014111640598256652,
+      "loss": 0.8899,
+      "step": 4575
+    },
+    {
+      "epoch": 0.42910010774347684,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0001409672748306826,
+      "loss": 0.8595,
+      "step": 4580
+    },
+    {
+      "epoch": 0.4295685576427601,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014081803411123767,
+      "loss": 0.8784,
+      "step": 4585
+    },
+    {
+      "epoch": 0.4300370075420434,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00014066868422337822,
+      "loss": 0.8833,
+      "step": 4590
+    },
+    {
+      "epoch": 0.4305054574413267,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0001405192255665426,
+      "loss": 0.8583,
+      "step": 4595
+    },
+    {
+      "epoch": 0.4309739073406099,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00014036965854046014,
+      "loss": 0.9033,
+      "step": 4600
+    },
+    {
+      "epoch": 0.4314423572398932,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014021998354514997,
+      "loss": 0.9071,
+      "step": 4605
+    },
+    {
+      "epoch": 0.43191080713917646,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00014007020098092,
+      "loss": 0.9111,
+      "step": 4610
+    },
+    {
+      "epoch": 0.43237925703845975,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00013992031124836587,
+      "loss": 0.9257,
+      "step": 4615
+    },
+    {
+      "epoch": 0.432847706937743,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013977031474836978,
+      "loss": 0.9058,
+      "step": 4620
+    },
+    {
+      "epoch": 0.4333161568370263,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00013962021188209948,
+      "loss": 0.8888,
+      "step": 4625
+    },
+    {
+      "epoch": 0.43378460673630953,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013947000305100725,
+      "loss": 0.882,
+      "step": 4630
+    },
+    {
+      "epoch": 0.4342530566355928,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00013931968865682883,
+      "loss": 0.8791,
+      "step": 4635
+    },
+    {
+      "epoch": 0.43472150653487607,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013916926910158208,
+      "loss": 0.9085,
+      "step": 4640
+    },
+    {
+      "epoch": 0.43518995643415936,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013901874478756634,
+      "loss": 0.8917,
+      "step": 4645
+    },
+    {
+      "epoch": 0.43565840633344266,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013886811611736098,
+      "loss": 0.8787,
+      "step": 4650
+    },
+    {
+      "epoch": 0.4361268562327259,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013871738349382457,
+      "loss": 0.8935,
+      "step": 4655
+    },
+    {
+      "epoch": 0.4365953061320092,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013856654732009364,
+      "loss": 0.9202,
+      "step": 4660
+    },
+    {
+      "epoch": 0.43706375603129244,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001384156079995817,
+      "loss": 0.8838,
+      "step": 4665
+    },
+    {
+      "epoch": 0.43753220593057573,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00013826456593597807,
+      "loss": 0.9288,
+      "step": 4670
+    },
+    {
+      "epoch": 0.438000655829859,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.000138113421533247,
+      "loss": 0.8848,
+      "step": 4675
+    },
+    {
+      "epoch": 0.43846910572914227,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013796217519562625,
+      "loss": 0.9011,
+      "step": 4680
+    },
+    {
+      "epoch": 0.4389375556284255,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013781082732762637,
+      "loss": 0.897,
+      "step": 4685
+    },
+    {
+      "epoch": 0.4394060055277088,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00013765937833402935,
+      "loss": 0.8621,
+      "step": 4690
+    },
+    {
+      "epoch": 0.4398744554269921,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00013750782861988775,
+      "loss": 0.892,
+      "step": 4695
+    },
+    {
+      "epoch": 0.44034290532627535,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00013735617859052336,
+      "loss": 0.9176,
+      "step": 4700
+    },
+    {
+      "epoch": 0.44081135522555864,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013720442865152642,
+      "loss": 0.9059,
+      "step": 4705
+    },
+    {
+      "epoch": 0.4412798051248419,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00013705257920875423,
+      "loss": 0.9004,
+      "step": 4710
+    },
+    {
+      "epoch": 0.4417482550241252,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00013690063066833038,
+      "loss": 0.923,
+      "step": 4715
+    },
+    {
+      "epoch": 0.4422167049234084,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00013674858343664333,
+      "loss": 0.8823,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4426851548226917,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00013659643792034563,
+      "loss": 0.8835,
+      "step": 4725
+    },
+    {
+      "epoch": 0.44315360472197496,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001364441945263526,
+      "loss": 0.938,
+      "step": 4730
+    },
+    {
+      "epoch": 0.44362205462125825,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013629185366184138,
+      "loss": 0.875,
+      "step": 4735
+    },
+    {
+      "epoch": 0.44409050452054155,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013613941573424978,
+      "loss": 0.8917,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4445589544198248,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001359868811512752,
+      "loss": 0.8446,
+      "step": 4745
+    },
+    {
+      "epoch": 0.4450274043191081,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001358342503208736,
+      "loss": 0.8666,
+      "step": 4750
+    },
+    {
+      "epoch": 0.44549585421839133,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013568152365125828,
+      "loss": 0.9304,
+      "step": 4755
+    },
+    {
+      "epoch": 0.4459643041176746,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00013552870155089888,
+      "loss": 0.8906,
+      "step": 4760
+    },
+    {
+      "epoch": 0.44643275401695787,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00013537578442852035,
+      "loss": 0.8833,
+      "step": 4765
+    },
+    {
+      "epoch": 0.44690120391624116,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00013522277269310164,
+      "loss": 0.9079,
+      "step": 4770
+    },
+    {
+      "epoch": 0.4473696538155244,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00013506966675387488,
+      "loss": 0.9203,
+      "step": 4775
+    },
+    {
+      "epoch": 0.4478381037148077,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00013491646702032406,
+      "loss": 0.8886,
+      "step": 4780
+    },
+    {
+      "epoch": 0.448306553614091,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013476317390218406,
+      "loss": 0.8876,
+      "step": 4785
+    },
+    {
+      "epoch": 0.44877500351337424,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013460978780943948,
+      "loss": 0.8907,
+      "step": 4790
+    },
+    {
+      "epoch": 0.44924345341265753,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001344563091523237,
+      "loss": 0.8603,
+      "step": 4795
+    },
+    {
+      "epoch": 0.4497119033119408,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001343027383413175,
+      "loss": 0.8745,
+      "step": 4800
+    },
+    {
+      "epoch": 0.45018035321122407,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001341490757871482,
+      "loss": 0.8741,
+      "step": 4805
+    },
+    {
+      "epoch": 0.4506488031105073,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001339953219007885,
+      "loss": 0.907,
+      "step": 4810
+    },
+    {
+      "epoch": 0.4511172530097906,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013384147709345542,
+      "loss": 0.8919,
+      "step": 4815
+    },
+    {
+      "epoch": 0.45158570290907385,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013368754177660904,
+      "loss": 0.8807,
+      "step": 4820
+    },
+    {
+      "epoch": 0.45205415280835715,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001335335163619516,
+      "loss": 0.8906,
+      "step": 4825
+    },
+    {
+      "epoch": 0.45252260270764044,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00013337940126142622,
+      "loss": 1.0339,
+      "step": 4830
+    },
+    {
+      "epoch": 0.4529910526069237,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00013322519688721598,
+      "loss": 0.9284,
+      "step": 4835
+    },
+    {
+      "epoch": 0.453459502506207,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00013307090365174263,
+      "loss": 0.9172,
+      "step": 4840
+    },
+    {
+      "epoch": 0.4539279524054902,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00013291652196766564,
+      "loss": 0.8695,
+      "step": 4845
+    },
+    {
+      "epoch": 0.4543964023047735,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00013276205224788107,
+      "loss": 0.8932,
+      "step": 4850
+    },
+    {
+      "epoch": 0.45486485220405676,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00013260749490552037,
+      "loss": 0.881,
+      "step": 4855
+    },
+    {
+      "epoch": 0.45533330210334005,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00013245285035394936,
+      "loss": 0.893,
+      "step": 4860
+    },
+    {
+      "epoch": 0.4558017520026233,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013229811900676707,
+      "loss": 0.8971,
+      "step": 4865
+    },
+    {
+      "epoch": 0.4562702019019066,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001321433012778047,
+      "loss": 0.8746,
+      "step": 4870
+    },
+    {
+      "epoch": 0.4567386518011899,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013198839758112453,
+      "loss": 0.878,
+      "step": 4875
+    },
+    {
+      "epoch": 0.45720710170047313,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013183340833101864,
+      "loss": 0.8936,
+      "step": 4880
+    },
+    {
+      "epoch": 0.4576755515997564,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.000131678333942008,
+      "loss": 0.8799,
+      "step": 4885
+    },
+    {
+      "epoch": 0.45814400149903967,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001315231748288413,
+      "loss": 0.8842,
+      "step": 4890
+    },
+    {
+      "epoch": 0.45861245139832296,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00013136793140649383,
+      "loss": 0.8941,
+      "step": 4895
+    },
+    {
+      "epoch": 0.4590809012976062,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013121260409016628,
+      "loss": 0.879,
+      "step": 4900
+    },
+    {
+      "epoch": 0.4595493511968895,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001310571932952838,
+      "loss": 1.0178,
+      "step": 4905
+    },
+    {
+      "epoch": 0.46001780109617274,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.8989,
+      "step": 4910
+    },
+    {
+      "epoch": 0.46048625099545604,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001307461229326697,
+      "loss": 0.8916,
+      "step": 4915
+    },
+    {
+      "epoch": 0.46095470089473933,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013059046419690027,
+      "loss": 0.8907,
+      "step": 4920
+    },
+    {
+      "epoch": 0.4614231507940226,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00013043472364649786,
+      "loss": 0.9318,
+      "step": 4925
+    },
+    {
+      "epoch": 0.46189160069330587,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001302789016979929,
+      "loss": 0.8905,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4623600505925891,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001301229987681334,
+      "loss": 0.8817,
+      "step": 4935
+    },
+    {
+      "epoch": 0.4628285004918724,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001299670152738839,
+      "loss": 0.8732,
+      "step": 4940
+    },
+    {
+      "epoch": 0.46329695039115565,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012981095163242453,
+      "loss": 0.8962,
+      "step": 4945
+    },
+    {
+      "epoch": 0.46376540029043894,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012965480826114973,
+      "loss": 0.8704,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4642338501897222,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001294985855776672,
+      "loss": 0.8947,
+      "step": 4955
+    },
+    {
+      "epoch": 0.4647023000890055,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012934228399979662,
+      "loss": 0.8959,
+      "step": 4960
+    },
+    {
+      "epoch": 0.4651707499882888,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00012918590394556893,
+      "loss": 0.8834,
+      "step": 4965
+    },
+    {
+      "epoch": 0.465639199887572,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001290294458332247,
+      "loss": 0.8817,
+      "step": 4970
+    },
+    {
+      "epoch": 0.4661076497868553,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00012887291008121352,
+      "loss": 0.8666,
+      "step": 4975
+    },
+    {
+      "epoch": 0.46657609968613856,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00012871629710819234,
+      "loss": 0.9036,
+      "step": 4980
+    },
+    {
+      "epoch": 0.46704454958542185,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00012855960733302492,
+      "loss": 0.8676,
+      "step": 4985
+    },
+    {
+      "epoch": 0.4675129994847051,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00012840284117478023,
+      "loss": 0.875,
+      "step": 4990
+    },
+    {
+      "epoch": 0.4679814493839884,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001282459990527316,
+      "loss": 0.8944,
+      "step": 4995
+    },
+    {
+      "epoch": 0.46844989928327163,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001280890813863555,
+      "loss": 0.8764,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4689183491825549,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012793208859533052,
+      "loss": 0.876,
+      "step": 5005
+    },
+    {
+      "epoch": 0.4693867990818382,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012777502109953613,
+      "loss": 0.9109,
+      "step": 5010
+    },
+    {
+      "epoch": 0.46985524898112146,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00012761787931905153,
+      "loss": 0.8625,
+      "step": 5015
+    },
+    {
+      "epoch": 0.47032369888040476,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00012746066367415467,
+      "loss": 0.8682,
+      "step": 5020
+    },
+    {
+      "epoch": 0.470792148779688,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012730337458532106,
+      "loss": 0.894,
+      "step": 5025
+    },
+    {
+      "epoch": 0.4712605986789713,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001271460124732226,
+      "loss": 0.8831,
+      "step": 5030
+    },
+    {
+      "epoch": 0.47172904857825454,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001269885777587265,
+      "loss": 0.8789,
+      "step": 5035
+    },
+    {
+      "epoch": 0.47219749847753784,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012683107086289413,
+      "loss": 0.8636,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4726659483768211,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012667349220697998,
+      "loss": 0.9318,
+      "step": 5045
+    },
+    {
+      "epoch": 0.4731343982761044,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012651584221243033,
+      "loss": 0.907,
+      "step": 5050
+    },
+    {
+      "epoch": 0.47360284817538767,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00012635812130088248,
+      "loss": 0.877,
+      "step": 5055
+    },
+    {
+      "epoch": 0.4740712980746709,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001262003298941631,
+      "loss": 0.8606,
+      "step": 5060
+    },
+    {
+      "epoch": 0.4745397479739542,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012604246841428767,
+      "loss": 0.8781,
+      "step": 5065
+    },
+    {
+      "epoch": 0.47500819787323745,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012588453728345892,
+      "loss": 0.8644,
+      "step": 5070
+    },
+    {
+      "epoch": 0.47547664777252074,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00012572653692406594,
+      "loss": 0.9029,
+      "step": 5075
+    },
+    {
+      "epoch": 0.475945097671804,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001255684677586829,
+      "loss": 0.8548,
+      "step": 5080
+    },
+    {
+      "epoch": 0.4764135475710873,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012541033021006806,
+      "loss": 1.0169,
+      "step": 5085
+    },
+    {
+      "epoch": 0.4768819974703705,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012525212470116258,
+      "loss": 0.8897,
+      "step": 5090
+    },
+    {
+      "epoch": 0.4773504473696538,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00012509385165508934,
+      "loss": 0.8701,
+      "step": 5095
+    },
+    {
+      "epoch": 0.47781889726893706,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00012493551149515183,
+      "loss": 0.866,
+      "step": 5100
+    },
+    {
+      "epoch": 0.47828734716822036,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00012477710464483307,
+      "loss": 0.8859,
+      "step": 5105
+    },
+    {
+      "epoch": 0.47875579706750365,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012461863152779445,
+      "loss": 0.8764,
+      "step": 5110
+    },
+    {
+      "epoch": 0.4792242469667869,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001244600925678746,
+      "loss": 0.8771,
+      "step": 5115
+    },
+    {
+      "epoch": 0.4796926968660702,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012430148818908822,
+      "loss": 0.8593,
+      "step": 5120
+    },
+    {
+      "epoch": 0.48016114676535343,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012414281881562497,
+      "loss": 0.8778,
+      "step": 5125
+    },
+    {
+      "epoch": 0.4806295966646367,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012398408487184837,
+      "loss": 0.8781,
+      "step": 5130
+    },
+    {
+      "epoch": 0.48109804656391997,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012382528678229465,
+      "loss": 0.8822,
+      "step": 5135
+    },
+    {
+      "epoch": 0.48156649646320326,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012366642497167146,
+      "loss": 0.8901,
+      "step": 5140
+    },
+    {
+      "epoch": 0.4820349463624865,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012350749986485708,
+      "loss": 0.8575,
+      "step": 5145
+    },
+    {
+      "epoch": 0.4825033962617698,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012334851188689893,
+      "loss": 0.9169,
+      "step": 5150
+    },
+    {
+      "epoch": 0.4829718461610531,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012318946146301263,
+      "loss": 0.8796,
+      "step": 5155
+    },
+    {
+      "epoch": 0.48344029606033634,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00012303034901858085,
+      "loss": 0.8967,
+      "step": 5160
+    },
+    {
+      "epoch": 0.48390874595961964,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00012287117497915203,
+      "loss": 0.8841,
+      "step": 5165
+    },
+    {
+      "epoch": 0.4843771958589029,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00012271193977043943,
+      "loss": 1.0036,
+      "step": 5170
+    },
+    {
+      "epoch": 0.4848456457581862,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00012255264381831992,
+      "loss": 0.9065,
+      "step": 5175
+    },
+    {
+      "epoch": 0.4853140956574694,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012239328754883277,
+      "loss": 0.9108,
+      "step": 5180
+    },
+    {
+      "epoch": 0.4857825455567527,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00012223387138817854,
+      "loss": 0.9077,
+      "step": 5185
+    },
+    {
+      "epoch": 0.48625099545603595,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012207439576271813,
+      "loss": 0.8911,
+      "step": 5190
+    },
+    {
+      "epoch": 0.48671944535531925,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012191486109897137,
+      "loss": 0.8984,
+      "step": 5195
+    },
+    {
+      "epoch": 0.48718789525460254,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00012175526782361588,
+      "loss": 0.903,
+      "step": 5200
+    },
+    {
+      "epoch": 0.4876563451538858,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00012159561636348624,
+      "loss": 0.8632,
+      "step": 5205
+    },
+    {
+      "epoch": 0.4881247950531691,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00012143590714557253,
+      "loss": 0.8835,
+      "step": 5210
+    },
+    {
+      "epoch": 0.4885932449524523,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00012127614059701939,
+      "loss": 0.9025,
+      "step": 5215
+    },
+    {
+      "epoch": 0.4890616948517356,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00012111631714512465,
+      "loss": 0.8723,
+      "step": 5220
+    },
+    {
+      "epoch": 0.48953014475101886,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012095643721733846,
+      "loss": 0.8803,
+      "step": 5225
+    },
+    {
+      "epoch": 0.48999859465030215,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012079650124126196,
+      "loss": 0.9091,
+      "step": 5230
+    },
+    {
+      "epoch": 0.4904670445495854,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00012063650964464619,
+      "loss": 0.887,
+      "step": 5235
+    },
+    {
+      "epoch": 0.4909354944488687,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00012047646285539098,
+      "loss": 0.877,
+      "step": 5240
+    },
+    {
+      "epoch": 0.491403944348152,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00012031636130154371,
+      "loss": 0.906,
+      "step": 5245
+    },
+    {
+      "epoch": 0.49187239424743523,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001201562054112983,
+      "loss": 0.8813,
+      "step": 5250
+    },
+    {
+      "epoch": 0.4923408441467185,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00011999599561299398,
+      "loss": 0.8678,
+      "step": 5255
+    },
+    {
+      "epoch": 0.49280929404600177,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001198357323351141,
+      "loss": 0.8975,
+      "step": 5260
+    },
+    {
+      "epoch": 0.49327774394528506,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011967541600628509,
+      "loss": 0.8598,
+      "step": 5265
+    },
+    {
+      "epoch": 0.4937461938445683,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001195150470552753,
+      "loss": 0.8887,
+      "step": 5270
+    },
+    {
+      "epoch": 0.4942146437438516,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00011935462591099374,
+      "loss": 0.8972,
+      "step": 5275
+    },
+    {
+      "epoch": 0.49468309364313484,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011919415300248907,
+      "loss": 0.8584,
+      "step": 5280
+    },
+    {
+      "epoch": 0.49515154354241814,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00011903362875894835,
+      "loss": 0.8616,
+      "step": 5285
+    },
+    {
+      "epoch": 0.49561999344170143,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011887305360969595,
+      "loss": 0.873,
+      "step": 5290
+    },
+    {
+      "epoch": 0.4960884433409847,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011871242798419244,
+      "loss": 0.8739,
+      "step": 5295
+    },
+    {
+      "epoch": 0.49655689324026797,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00011855175231203333,
+      "loss": 0.8845,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4970253431395512,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00011839102702294798,
+      "loss": 0.9028,
+      "step": 5305
+    },
+    {
+      "epoch": 0.4974937930388345,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00011823025254679847,
+      "loss": 0.8753,
+      "step": 5310
+    },
+    {
+      "epoch": 0.49796224293811775,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011806942931357847,
+      "loss": 0.879,
+      "step": 5315
+    },
+    {
+      "epoch": 0.49843069283740105,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011790855775341199,
+      "loss": 0.8764,
+      "step": 5320
+    },
+    {
+      "epoch": 0.4988991427366843,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00011774763829655225,
+      "loss": 0.8722,
+      "step": 5325
+    },
+    {
+      "epoch": 0.4993675926359676,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011758667137338073,
+      "loss": 0.8959,
+      "step": 5330
+    },
+    {
+      "epoch": 0.4998360425352509,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00011742565741440572,
+      "loss": 0.8866,
+      "step": 5335
+    },
+    {
+      "epoch": 0.5003044924345341,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011726459685026135,
+      "loss": 1.0222,
+      "step": 5340
+    },
+    {
+      "epoch": 0.5007729423338174,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00011710349011170638,
+      "loss": 0.8995,
+      "step": 5345
+    },
+    {
+      "epoch": 0.5012413922331007,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011694233762962312,
+      "loss": 0.8826,
+      "step": 5350
+    },
+    {
+      "epoch": 0.5017098421323839,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00011678113983501616,
+      "loss": 0.8776,
+      "step": 5355
+    },
+    {
+      "epoch": 0.5021782920316672,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001166198971590113,
+      "loss": 0.8749,
+      "step": 5360
+    },
+    {
+      "epoch": 0.5026467419309505,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011645861003285439,
+      "loss": 0.8867,
+      "step": 5365
+    },
+    {
+      "epoch": 0.5031151918302338,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011629727888791011,
+      "loss": 0.868,
+      "step": 5370
+    },
+    {
+      "epoch": 0.503583641729517,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011613590415566098,
+      "loss": 0.8676,
+      "step": 5375
+    },
+    {
+      "epoch": 0.5040520916288003,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00011597448626770595,
+      "loss": 0.86,
+      "step": 5380
+    },
+    {
+      "epoch": 0.5045205415280836,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011581302565575951,
+      "loss": 0.8729,
+      "step": 5385
+    },
+    {
+      "epoch": 0.5049889914273669,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00011565152275165034,
+      "loss": 0.873,
+      "step": 5390
+    },
+    {
+      "epoch": 0.5054574413266502,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011548997798732031,
+      "loss": 0.8759,
+      "step": 5395
+    },
+    {
+      "epoch": 0.5059258912259333,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011532839179482315,
+      "loss": 0.8717,
+      "step": 5400
+    },
+    {
+      "epoch": 0.5063943411252166,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011516676460632343,
+      "loss": 0.9097,
+      "step": 5405
+    },
+    {
+      "epoch": 0.5068627910244999,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011500509685409542,
+      "loss": 0.8884,
+      "step": 5410
+    },
+    {
+      "epoch": 0.5073312409237832,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001148433889705218,
+      "loss": 0.8796,
+      "step": 5415
+    },
+    {
+      "epoch": 0.5077996908230664,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001146816413880926,
+      "loss": 0.8702,
+      "step": 5420
+    },
+    {
+      "epoch": 0.5082681407223497,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00011451985453940406,
+      "loss": 0.875,
+      "step": 5425
+    },
+    {
+      "epoch": 0.508736590621633,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001143580288571574,
+      "loss": 0.8904,
+      "step": 5430
+    },
+    {
+      "epoch": 0.5092050405209163,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011419616477415775,
+      "loss": 0.889,
+      "step": 5435
+    },
+    {
+      "epoch": 0.5096734904201996,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011403426272331286,
+      "loss": 0.8519,
+      "step": 5440
+    },
+    {
+      "epoch": 0.5101419403194828,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001138723231376321,
+      "loss": 0.8653,
+      "step": 5445
+    },
+    {
+      "epoch": 0.5106103902187661,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011371034645022518,
+      "loss": 0.903,
+      "step": 5450
+    },
+    {
+      "epoch": 0.5110788401180494,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00011354833309430113,
+      "loss": 0.8613,
+      "step": 5455
+    },
+    {
+      "epoch": 0.5115472900173327,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001133862835031669,
+      "loss": 0.8777,
+      "step": 5460
+    },
+    {
+      "epoch": 0.5120157399166159,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00011322419811022644,
+      "loss": 0.8916,
+      "step": 5465
+    },
+    {
+      "epoch": 0.5124841898158992,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011306207734897951,
+      "loss": 0.8502,
+      "step": 5470
+    },
+    {
+      "epoch": 0.5129526397151825,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00011289992165302035,
+      "loss": 0.8633,
+      "step": 5475
+    },
+    {
+      "epoch": 0.5134210896144658,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011273773145603668,
+      "loss": 0.8879,
+      "step": 5480
+    },
+    {
+      "epoch": 0.513889539513749,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00011257550719180852,
+      "loss": 0.8895,
+      "step": 5485
+    },
+    {
+      "epoch": 0.5143579894130322,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011241324929420695,
+      "loss": 0.8604,
+      "step": 5490
+    },
+    {
+      "epoch": 0.5148264393123155,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00011225095819719306,
+      "loss": 0.8618,
+      "step": 5495
+    },
+    {
+      "epoch": 0.5152948892115988,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011208863433481672,
+      "loss": 0.9259,
+      "step": 5500
+    },
+    {
+      "epoch": 0.5157633391108821,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011192627814121533,
+      "loss": 0.8885,
+      "step": 5505
+    },
+    {
+      "epoch": 0.5162317890101653,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011176389005061294,
+      "loss": 0.8766,
+      "step": 5510
+    },
+    {
+      "epoch": 0.5167002389094486,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011160147049731878,
+      "loss": 0.8666,
+      "step": 5515
+    },
+    {
+      "epoch": 0.5171686888087319,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011143901991572627,
+      "loss": 0.8706,
+      "step": 5520
+    },
+    {
+      "epoch": 0.5176371387080152,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011127653874031178,
+      "loss": 0.891,
+      "step": 5525
+    },
+    {
+      "epoch": 0.5181055886072985,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011111402740563354,
+      "loss": 0.8846,
+      "step": 5530
+    },
+    {
+      "epoch": 0.5185740385065817,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001109514863463305,
+      "loss": 0.8769,
+      "step": 5535
+    },
+    {
+      "epoch": 0.519042488405865,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00011078891599712094,
+      "loss": 0.8819,
+      "step": 5540
+    },
+    {
+      "epoch": 0.5195109383051483,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011062631679280165,
+      "loss": 0.8946,
+      "step": 5545
+    },
+    {
+      "epoch": 0.5199793882044316,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001104636891682465,
+      "loss": 0.884,
+      "step": 5550
+    },
+    {
+      "epoch": 0.5204478381037148,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00011030103355840538,
+      "loss": 0.8788,
+      "step": 5555
+    },
+    {
+      "epoch": 0.520916288002998,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00011013835039830305,
+      "loss": 0.9092,
+      "step": 5560
+    },
+    {
+      "epoch": 0.5213847379022813,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00010997564012303791,
+      "loss": 0.8494,
+      "step": 5565
+    },
+    {
+      "epoch": 0.5218531878015646,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00010981290316778097,
+      "loss": 0.8551,
+      "step": 5570
+    },
+    {
+      "epoch": 0.5223216377008479,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010965013996777451,
+      "loss": 0.8804,
+      "step": 5575
+    },
+    {
+      "epoch": 0.5227900876001311,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010948735095833104,
+      "loss": 0.874,
+      "step": 5580
+    },
+    {
+      "epoch": 0.5232585374994144,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010932453657483206,
+      "loss": 0.8647,
+      "step": 5585
+    },
+    {
+      "epoch": 0.5237269873986977,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.000109161697252727,
+      "loss": 0.8544,
+      "step": 5590
+    },
+    {
+      "epoch": 0.524195437297981,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00010899883342753192,
+      "loss": 0.8719,
+      "step": 5595
+    },
+    {
+      "epoch": 0.5246638871972642,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00010883594553482848,
+      "loss": 0.8891,
+      "step": 5600
+    },
+    {
+      "epoch": 0.5251323370965475,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00010867303401026263,
+      "loss": 0.8787,
+      "step": 5605
+    },
+    {
+      "epoch": 0.5256007869958308,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00010851009928954356,
+      "loss": 0.8583,
+      "step": 5610
+    },
+    {
+      "epoch": 0.5260692368951141,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010834714180844259,
+      "loss": 0.8872,
+      "step": 5615
+    },
+    {
+      "epoch": 0.5265376867943974,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00010818416200279174,
+      "loss": 0.8636,
+      "step": 5620
+    },
+    {
+      "epoch": 0.5270061366936806,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010802116030848284,
+      "loss": 0.8783,
+      "step": 5625
+    },
+    {
+      "epoch": 0.5274745865929639,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010785813716146625,
+      "loss": 0.8731,
+      "step": 5630
+    },
+    {
+      "epoch": 0.5279430364922472,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010769509299774976,
+      "loss": 0.886,
+      "step": 5635
+    },
+    {
+      "epoch": 0.5284114863915305,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010753202825339725,
+      "loss": 0.8794,
+      "step": 5640
+    },
+    {
+      "epoch": 0.5288799362908136,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001073689433645277,
+      "loss": 0.8998,
+      "step": 5645
+    },
+    {
+      "epoch": 0.5293483861900969,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.000107205838767314,
+      "loss": 0.8788,
+      "step": 5650
+    },
+    {
+      "epoch": 0.5298168360893802,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00010704271489798172,
+      "loss": 0.8976,
+      "step": 5655
+    },
+    {
+      "epoch": 0.5302852859886635,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010687957219280798,
+      "loss": 0.8837,
+      "step": 5660
+    },
+    {
+      "epoch": 0.5307537358879468,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010671641108812022,
+      "loss": 0.8613,
+      "step": 5665
+    },
+    {
+      "epoch": 0.53122218578723,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010655323202029524,
+      "loss": 0.8839,
+      "step": 5670
+    },
+    {
+      "epoch": 0.5316906356865133,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001063900354257577,
+      "loss": 0.8867,
+      "step": 5675
+    },
+    {
+      "epoch": 0.5321590855857966,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00010622682174097927,
+      "loss": 0.8875,
+      "step": 5680
+    },
+    {
+      "epoch": 0.5326275354850799,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010606359140247723,
+      "loss": 0.8742,
+      "step": 5685
+    },
+    {
+      "epoch": 0.5330959853843631,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00010590034484681348,
+      "loss": 0.8676,
+      "step": 5690
+    },
+    {
+      "epoch": 0.5335644352836464,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010573708251059331,
+      "loss": 1.0158,
+      "step": 5695
+    },
+    {
+      "epoch": 0.5340328851829297,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00010557380483046406,
+      "loss": 0.8818,
+      "step": 5700
+    },
+    {
+      "epoch": 0.534501335082213,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001054105122431143,
+      "loss": 0.8693,
+      "step": 5705
+    },
+    {
+      "epoch": 0.5349697849814963,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00010524720518527231,
+      "loss": 0.8966,
+      "step": 5710
+    },
+    {
+      "epoch": 0.5354382348807795,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010508388409370526,
+      "loss": 0.9026,
+      "step": 5715
+    },
+    {
+      "epoch": 0.5359066847800628,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00010492054940521763,
+      "loss": 0.8762,
+      "step": 5720
+    },
+    {
+      "epoch": 0.536375134679346,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010475720155665043,
+      "loss": 0.8968,
+      "step": 5725
+    },
+    {
+      "epoch": 0.5368435845786294,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001045938409848798,
+      "loss": 0.8619,
+      "step": 5730
+    },
+    {
+      "epoch": 0.5373120344779125,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00010443046812681594,
+      "loss": 0.8887,
+      "step": 5735
+    },
+    {
+      "epoch": 0.5377804843771958,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010426708341940183,
+      "loss": 0.8597,
+      "step": 5740
+    },
+    {
+      "epoch": 0.5382489342764791,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010410368729961225,
+      "loss": 0.8761,
+      "step": 5745
+    },
+    {
+      "epoch": 0.5387173841757624,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00010394028020445244,
+      "loss": 0.8793,
+      "step": 5750
+    },
+    {
+      "epoch": 0.5391858340750457,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00010377686257095702,
+      "loss": 0.8988,
+      "step": 5755
+    },
+    {
+      "epoch": 0.5396542839743289,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00010361343483618877,
+      "loss": 0.864,
+      "step": 5760
+    },
+    {
+      "epoch": 0.5401227338736122,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00010344999743723749,
+      "loss": 0.8618,
+      "step": 5765
+    },
+    {
+      "epoch": 0.5405911837728955,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010328655081121883,
+      "loss": 0.8766,
+      "step": 5770
+    },
+    {
+      "epoch": 0.5410596336721788,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010312309539527318,
+      "loss": 0.8545,
+      "step": 5775
+    },
+    {
+      "epoch": 0.541528083571462,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001029596316265643,
+      "loss": 0.8839,
+      "step": 5780
+    },
+    {
+      "epoch": 0.5419965334707453,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010279615994227843,
+      "loss": 0.8966,
+      "step": 5785
+    },
+    {
+      "epoch": 0.5424649833700286,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00010263268077962289,
+      "loss": 0.8905,
+      "step": 5790
+    },
+    {
+      "epoch": 0.5429334332693119,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010246919457582507,
+      "loss": 0.8528,
+      "step": 5795
+    },
+    {
+      "epoch": 0.5434018831685952,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001023057017681311,
+      "loss": 0.844,
+      "step": 5800
+    },
+    {
+      "epoch": 0.5438703330678784,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00010214220279380489,
+      "loss": 0.8905,
+      "step": 5805
+    },
+    {
+      "epoch": 0.5443387829671616,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010197869809012673,
+      "loss": 0.8604,
+      "step": 5810
+    },
+    {
+      "epoch": 0.544807232866445,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001018151880943923,
+      "loss": 0.9017,
+      "step": 5815
+    },
+    {
+      "epoch": 0.5452756827657282,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001016516732439114,
+      "loss": 0.9017,
+      "step": 5820
+    },
+    {
+      "epoch": 0.5457441326650114,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010148815397600686,
+      "loss": 0.8484,
+      "step": 5825
+    },
+    {
+      "epoch": 0.5462125825642947,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010132463072801328,
+      "loss": 0.8643,
+      "step": 5830
+    },
+    {
+      "epoch": 0.546681032463578,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00010116110393727592,
+      "loss": 0.8603,
+      "step": 5835
+    },
+    {
+      "epoch": 0.5471494823628613,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010099757404114952,
+      "loss": 0.8969,
+      "step": 5840
+    },
+    {
+      "epoch": 0.5476179322621446,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001008340414769971,
+      "loss": 0.8644,
+      "step": 5845
+    },
+    {
+      "epoch": 0.5480863821614278,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010067050668218885,
+      "loss": 0.8508,
+      "step": 5850
+    },
+    {
+      "epoch": 0.5485548320607111,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010050697009410095,
+      "loss": 0.8716,
+      "step": 5855
+    },
+    {
+      "epoch": 0.5490232819599944,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010034343215011426,
+      "loss": 0.8546,
+      "step": 5860
+    },
+    {
+      "epoch": 0.5494917318592777,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010017989328761339,
+      "loss": 0.8651,
+      "step": 5865
+    },
+    {
+      "epoch": 0.5499601817585609,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00010001635394398534,
+      "loss": 0.8669,
+      "step": 5870
+    },
+    {
+      "epoch": 0.5504286316578442,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.985281455661846e-05,
+      "loss": 0.868,
+      "step": 5875
+    },
+    {
+      "epoch": 0.5508970815571275,
+      "grad_norm": 0.625,
+      "learning_rate": 9.968927556290115e-05,
+      "loss": 0.8629,
+      "step": 5880
+    },
+    {
+      "epoch": 0.5513655314564108,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.952573740022081e-05,
+      "loss": 0.8788,
+      "step": 5885
+    },
+    {
+      "epoch": 0.5518339813556941,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.93622005059625e-05,
+      "loss": 0.8598,
+      "step": 5890
+    },
+    {
+      "epoch": 0.5523024312549772,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.919866531750807e-05,
+      "loss": 0.8892,
+      "step": 5895
+    },
+    {
+      "epoch": 0.5527708811542605,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.903513227223467e-05,
+      "loss": 0.8647,
+      "step": 5900
+    },
+    {
+      "epoch": 0.5532393310535438,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.887160180751376e-05,
+      "loss": 0.8844,
+      "step": 5905
+    },
+    {
+      "epoch": 0.5537077809528271,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.870807436070994e-05,
+      "loss": 0.8746,
+      "step": 5910
+    },
+    {
+      "epoch": 0.5541762308521103,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.854455036917966e-05,
+      "loss": 0.8988,
+      "step": 5915
+    },
+    {
+      "epoch": 0.5546446807513936,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.838103027027016e-05,
+      "loss": 0.8698,
+      "step": 5920
+    },
+    {
+      "epoch": 0.5551131306506769,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.821751450131835e-05,
+      "loss": 0.8662,
+      "step": 5925
+    },
+    {
+      "epoch": 0.5555815805499602,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.805400349964942e-05,
+      "loss": 0.8566,
+      "step": 5930
+    },
+    {
+      "epoch": 0.5560500304492435,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.789049770257588e-05,
+      "loss": 0.8628,
+      "step": 5935
+    },
+    {
+      "epoch": 0.5565184803485267,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.772699754739631e-05,
+      "loss": 0.8937,
+      "step": 5940
+    },
+    {
+      "epoch": 0.55698693024781,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.756350347139426e-05,
+      "loss": 0.8857,
+      "step": 5945
+    },
+    {
+      "epoch": 0.5574553801470933,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.740001591183691e-05,
+      "loss": 0.8391,
+      "step": 5950
+    },
+    {
+      "epoch": 0.5579238300463766,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.723653530597411e-05,
+      "loss": 0.8685,
+      "step": 5955
+    },
+    {
+      "epoch": 0.5583922799456598,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.707306209103706e-05,
+      "loss": 0.8784,
+      "step": 5960
+    },
+    {
+      "epoch": 0.5588607298449431,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.690959670423724e-05,
+      "loss": 0.8867,
+      "step": 5965
+    },
+    {
+      "epoch": 0.5593291797442264,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.674613958276506e-05,
+      "loss": 0.9051,
+      "step": 5970
+    },
+    {
+      "epoch": 0.5597976296435097,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.658269116378902e-05,
+      "loss": 0.8998,
+      "step": 5975
+    },
+    {
+      "epoch": 0.560266079542793,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.641925188445423e-05,
+      "loss": 0.87,
+      "step": 5980
+    },
+    {
+      "epoch": 0.5607345294420761,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.625582218188131e-05,
+      "loss": 0.8709,
+      "step": 5985
+    },
+    {
+      "epoch": 0.5612029793413594,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.609240249316543e-05,
+      "loss": 0.8752,
+      "step": 5990
+    },
+    {
+      "epoch": 0.5616714292406427,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.59289932553748e-05,
+      "loss": 0.8497,
+      "step": 5995
+    },
+    {
+      "epoch": 0.562139879139926,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.576559490554978e-05,
+      "loss": 0.8478,
+      "step": 6000
+    },
+    {
+      "epoch": 0.5626083290392092,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.560220788070165e-05,
+      "loss": 0.8729,
+      "step": 6005
+    },
+    {
+      "epoch": 0.5630767789384925,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.543883261781125e-05,
+      "loss": 0.8664,
+      "step": 6010
+    },
+    {
+      "epoch": 0.5635452288377758,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.52754695538281e-05,
+      "loss": 0.8771,
+      "step": 6015
+    },
+    {
+      "epoch": 0.5640136787370591,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.511211912566899e-05,
+      "loss": 0.8805,
+      "step": 6020
+    },
+    {
+      "epoch": 0.5644821286363424,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.494878177021704e-05,
+      "loss": 0.8739,
+      "step": 6025
+    },
+    {
+      "epoch": 0.5649505785356256,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.478545792432028e-05,
+      "loss": 0.8677,
+      "step": 6030
+    },
+    {
+      "epoch": 0.5654190284349089,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.462214802479074e-05,
+      "loss": 0.8532,
+      "step": 6035
+    },
+    {
+      "epoch": 0.5658874783341922,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.4458852508403e-05,
+      "loss": 0.8711,
+      "step": 6040
+    },
+    {
+      "epoch": 0.5663559282334755,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.429557181189332e-05,
+      "loss": 0.8325,
+      "step": 6045
+    },
+    {
+      "epoch": 0.5668243781327587,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.413230637195817e-05,
+      "loss": 0.8963,
+      "step": 6050
+    },
+    {
+      "epoch": 0.567292828032042,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.396905662525336e-05,
+      "loss": 0.8704,
+      "step": 6055
+    },
+    {
+      "epoch": 0.5677612779313252,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.380582300839264e-05,
+      "loss": 0.8896,
+      "step": 6060
+    },
+    {
+      "epoch": 0.5682297278306085,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.364260595794663e-05,
+      "loss": 0.8826,
+      "step": 6065
+    },
+    {
+      "epoch": 0.5686981777298918,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.347940591044171e-05,
+      "loss": 0.8602,
+      "step": 6070
+    },
+    {
+      "epoch": 0.569166627629175,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.331622330235869e-05,
+      "loss": 0.9071,
+      "step": 6075
+    },
+    {
+      "epoch": 0.5696350775284583,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.315305857013181e-05,
+      "loss": 0.8597,
+      "step": 6080
+    },
+    {
+      "epoch": 0.5701035274277416,
+      "grad_norm": 0.625,
+      "learning_rate": 9.29899121501475e-05,
+      "loss": 0.9042,
+      "step": 6085
+    },
+    {
+      "epoch": 0.5705719773270249,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.282678447874314e-05,
+      "loss": 0.8935,
+      "step": 6090
+    },
+    {
+      "epoch": 0.5710404272263081,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.266367599220601e-05,
+      "loss": 0.855,
+      "step": 6095
+    },
+    {
+      "epoch": 0.5715088771255914,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.25005871267721e-05,
+      "loss": 0.8672,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5719773270248747,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.233751831862491e-05,
+      "loss": 0.9126,
+      "step": 6105
+    },
+    {
+      "epoch": 0.572445776924158,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.217447000389429e-05,
+      "loss": 0.8768,
+      "step": 6110
+    },
+    {
+      "epoch": 0.5729142268234413,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.20114426186553e-05,
+      "loss": 0.8743,
+      "step": 6115
+    },
+    {
+      "epoch": 0.5733826767227245,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.184843659892701e-05,
+      "loss": 0.8724,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5738511266220078,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.168545238067134e-05,
+      "loss": 0.8537,
+      "step": 6125
+    },
+    {
+      "epoch": 0.5743195765212911,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.152249039979184e-05,
+      "loss": 0.8569,
+      "step": 6130
+    },
+    {
+      "epoch": 0.5747880264205744,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.135955109213275e-05,
+      "loss": 0.8927,
+      "step": 6135
+    },
+    {
+      "epoch": 0.5752564763198575,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.119663489347753e-05,
+      "loss": 0.8405,
+      "step": 6140
+    },
+    {
+      "epoch": 0.5757249262191408,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.103374223954785e-05,
+      "loss": 0.858,
+      "step": 6145
+    },
+    {
+      "epoch": 0.5761933761184241,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.08708735660025e-05,
+      "loss": 0.8859,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5766618260177074,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.070802930843602e-05,
+      "loss": 0.8602,
+      "step": 6155
+    },
+    {
+      "epoch": 0.5771302759169907,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.054520990237774e-05,
+      "loss": 0.857,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5775987258162739,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.038241578329048e-05,
+      "loss": 0.8562,
+      "step": 6165
+    },
+    {
+      "epoch": 0.5780671757155572,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.02196473865694e-05,
+      "loss": 0.8498,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5785356256148405,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.005690514754088e-05,
+      "loss": 0.8696,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5790040755141238,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.989418950146143e-05,
+      "loss": 0.8728,
+      "step": 6180
+    },
+    {
+      "epoch": 0.579472525413407,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.973150088351633e-05,
+      "loss": 0.8778,
+      "step": 6185
+    },
+    {
+      "epoch": 0.5799409753126903,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.956883972881859e-05,
+      "loss": 0.8915,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5804094252119736,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.940620647240783e-05,
+      "loss": 0.8785,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5808778751112569,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.9243601549249e-05,
+      "loss": 0.8806,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5813463250105402,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.908102539423131e-05,
+      "loss": 0.8693,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5818147749098234,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.891847844216693e-05,
+      "loss": 0.8833,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5822832248091067,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.875596112779008e-05,
+      "loss": 0.8786,
+      "step": 6215
+    },
+    {
+      "epoch": 0.58275167470839,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.85934738857556e-05,
+      "loss": 1.011,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5832201246076733,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.843101715063792e-05,
+      "loss": 1.006,
+      "step": 6225
+    },
+    {
+      "epoch": 0.5836885745069564,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.826859135692995e-05,
+      "loss": 0.8741,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5841570244062397,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.810619693904173e-05,
+      "loss": 0.925,
+      "step": 6235
+    },
+    {
+      "epoch": 0.584625474305523,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.794383433129952e-05,
+      "loss": 0.8497,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5850939242048063,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.77815039679444e-05,
+      "loss": 0.8662,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5855623741040896,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.761920628313119e-05,
+      "loss": 0.8964,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5860308240033728,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.745694171092742e-05,
+      "loss": 0.897,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5864992739026561,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.729471068531199e-05,
+      "loss": 0.8566,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5869677238019394,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.71325136401741e-05,
+      "loss": 0.9079,
+      "step": 6265
+    },
+    {
+      "epoch": 0.5874361737012227,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.697035100931203e-05,
+      "loss": 0.8806,
+      "step": 6270
+    },
+    {
+      "epoch": 0.5879046236005059,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.680822322643209e-05,
+      "loss": 0.873,
+      "step": 6275
+    },
+    {
+      "epoch": 0.5883730734997892,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.664613072514735e-05,
+      "loss": 0.8524,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5888415233990725,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.64840739389765e-05,
+      "loss": 0.8885,
+      "step": 6285
+    },
+    {
+      "epoch": 0.5893099732983558,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.632205330134269e-05,
+      "loss": 0.8719,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5897784231976391,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.616006924557248e-05,
+      "loss": 0.863,
+      "step": 6295
+    },
+    {
+      "epoch": 0.5902468730969223,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.599812220489453e-05,
+      "loss": 0.8987,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5907153229962055,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.583621261243847e-05,
+      "loss": 0.8778,
+      "step": 6305
+    },
+    {
+      "epoch": 0.5911837728954888,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.567434090123387e-05,
+      "loss": 0.893,
+      "step": 6310
+    },
+    {
+      "epoch": 0.5916522227947721,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.55125075042089e-05,
+      "loss": 0.9027,
+      "step": 6315
+    },
+    {
+      "epoch": 0.5921206726940553,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.53507128541893e-05,
+      "loss": 0.8564,
+      "step": 6320
+    },
+    {
+      "epoch": 0.5925891225933386,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.518895738389719e-05,
+      "loss": 0.869,
+      "step": 6325
+    },
+    {
+      "epoch": 0.5930575724926219,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.502724152594983e-05,
+      "loss": 0.8895,
+      "step": 6330
+    },
+    {
+      "epoch": 0.5935260223919052,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.48655657128586e-05,
+      "loss": 0.8549,
+      "step": 6335
+    },
+    {
+      "epoch": 0.5939944722911884,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.470393037702783e-05,
+      "loss": 0.8571,
+      "step": 6340
+    },
+    {
+      "epoch": 0.5944629221904717,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.454233595075347e-05,
+      "loss": 0.9031,
+      "step": 6345
+    },
+    {
+      "epoch": 0.594931372089755,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.438078286622211e-05,
+      "loss": 0.8523,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5953998219890383,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.421927155550984e-05,
+      "loss": 0.8472,
+      "step": 6355
+    },
+    {
+      "epoch": 0.5958682718883216,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.405780245058093e-05,
+      "loss": 0.8302,
+      "step": 6360
+    },
+    {
+      "epoch": 0.5963367217876048,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.38963759832868e-05,
+      "loss": 0.8678,
+      "step": 6365
+    },
+    {
+      "epoch": 0.5968051716868881,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.373499258536483e-05,
+      "loss": 0.8584,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5972736215861714,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.357365268843727e-05,
+      "loss": 0.863,
+      "step": 6375
+    },
+    {
+      "epoch": 0.5977420714854547,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.341235672400992e-05,
+      "loss": 0.8899,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5982105213847378,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.325110512347115e-05,
+      "loss": 0.8759,
+      "step": 6385
+    },
+    {
+      "epoch": 0.5986789712840211,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.308989831809068e-05,
+      "loss": 0.8537,
+      "step": 6390
+    },
+    {
+      "epoch": 0.5991474211833044,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.292873673901839e-05,
+      "loss": 0.865,
+      "step": 6395
+    },
+    {
+      "epoch": 0.5996158710825877,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.276762081728327e-05,
+      "loss": 0.8583,
+      "step": 6400
+    },
+    {
+      "epoch": 0.600084320981871,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.260655098379214e-05,
+      "loss": 0.8602,
+      "step": 6405
+    },
+    {
+      "epoch": 0.6005527708811542,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.244552766932852e-05,
+      "loss": 0.8685,
+      "step": 6410
+    },
+    {
+      "epoch": 0.6010212207804375,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.228455130455156e-05,
+      "loss": 0.8539,
+      "step": 6415
+    },
+    {
+      "epoch": 0.6014896706797208,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.21236223199949e-05,
+      "loss": 0.8566,
+      "step": 6420
+    },
+    {
+      "epoch": 0.6019581205790041,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.196274114606539e-05,
+      "loss": 0.8589,
+      "step": 6425
+    },
+    {
+      "epoch": 0.6024265704782873,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.180190821304199e-05,
+      "loss": 0.8537,
+      "step": 6430
+    },
+    {
+      "epoch": 0.6028950203775706,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.164112395107472e-05,
+      "loss": 0.8657,
+      "step": 6435
+    },
+    {
+      "epoch": 0.6033634702768539,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.148038879018334e-05,
+      "loss": 0.8321,
+      "step": 6440
+    },
+    {
+      "epoch": 0.6038319201761372,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.13197031602564e-05,
+      "loss": 0.8817,
+      "step": 6445
+    },
+    {
+      "epoch": 0.6043003700754205,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.11590674910498e-05,
+      "loss": 0.8418,
+      "step": 6450
+    },
+    {
+      "epoch": 0.6047688199747037,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.099848221218602e-05,
+      "loss": 0.879,
+      "step": 6455
+    },
+    {
+      "epoch": 0.605237269873987,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.083794775315265e-05,
+      "loss": 0.8397,
+      "step": 6460
+    },
+    {
+      "epoch": 0.6057057197732703,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.067746454330137e-05,
+      "loss": 0.891,
+      "step": 6465
+    },
+    {
+      "epoch": 0.6061741696725536,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.051703301184684e-05,
+      "loss": 0.8681,
+      "step": 6470
+    },
+    {
+      "epoch": 0.6066426195718367,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.035665358786546e-05,
+      "loss": 0.8485,
+      "step": 6475
+    },
+    {
+      "epoch": 0.60711106947112,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.019632670029433e-05,
+      "loss": 0.8705,
+      "step": 6480
+    },
+    {
+      "epoch": 0.6075795193704033,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.003605277793e-05,
+      "loss": 0.8835,
+      "step": 6485
+    },
+    {
+      "epoch": 0.6080479692696866,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.987583224942728e-05,
+      "loss": 0.8667,
+      "step": 6490
+    },
+    {
+      "epoch": 0.6085164191689699,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.971566554329831e-05,
+      "loss": 0.8857,
+      "step": 6495
+    },
+    {
+      "epoch": 0.6089848690682531,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.955555308791125e-05,
+      "loss": 0.84,
+      "step": 6500
+    },
+    {
+      "epoch": 0.6094533189675364,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.939549531148914e-05,
+      "loss": 0.8466,
+      "step": 6505
+    },
+    {
+      "epoch": 0.6099217688668197,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.923549264210872e-05,
+      "loss": 0.8594,
+      "step": 6510
+    },
+    {
+      "epoch": 0.610390218766103,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.907554550769952e-05,
+      "loss": 0.8864,
+      "step": 6515
+    },
+    {
+      "epoch": 0.6108586686653862,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.891565433604234e-05,
+      "loss": 0.8643,
+      "step": 6520
+    },
+    {
+      "epoch": 0.6113271185646695,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.875581955476849e-05,
+      "loss": 0.8533,
+      "step": 6525
+    },
+    {
+      "epoch": 0.6117955684639528,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.859604159135828e-05,
+      "loss": 0.8311,
+      "step": 6530
+    },
+    {
+      "epoch": 0.6122640183632361,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.843632087314021e-05,
+      "loss": 0.8928,
+      "step": 6535
+    },
+    {
+      "epoch": 0.6127324682625194,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.82766578272896e-05,
+      "loss": 0.8723,
+      "step": 6540
+    },
+    {
+      "epoch": 0.6132009181618026,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.811705288082754e-05,
+      "loss": 0.8435,
+      "step": 6545
+    },
+    {
+      "epoch": 0.6136693680610859,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.795750646061974e-05,
+      "loss": 0.8795,
+      "step": 6550
+    },
+    {
+      "epoch": 0.6141378179603691,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.779801899337537e-05,
+      "loss": 0.879,
+      "step": 6555
+    },
+    {
+      "epoch": 0.6146062678596524,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.763859090564598e-05,
+      "loss": 0.8712,
+      "step": 6560
+    },
+    {
+      "epoch": 0.6150747177589356,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.747922262382424e-05,
+      "loss": 0.8458,
+      "step": 6565
+    },
+    {
+      "epoch": 0.6155431676582189,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.731991457414287e-05,
+      "loss": 0.8514,
+      "step": 6570
+    },
+    {
+      "epoch": 0.6160116175575022,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.716066718267352e-05,
+      "loss": 0.8556,
+      "step": 6575
+    },
+    {
+      "epoch": 0.6164800674567855,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.700148087532563e-05,
+      "loss": 0.8541,
+      "step": 6580
+    },
+    {
+      "epoch": 0.6169485173560688,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.684235607784526e-05,
+      "loss": 0.8539,
+      "step": 6585
+    },
+    {
+      "epoch": 0.617416967255352,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.668329321581392e-05,
+      "loss": 0.8759,
+      "step": 6590
+    },
+    {
+      "epoch": 0.6178854171546353,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.652429271464755e-05,
+      "loss": 0.8643,
+      "step": 6595
+    },
+    {
+      "epoch": 0.6183538670539186,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.636535499959521e-05,
+      "loss": 0.8598,
+      "step": 6600
+    },
+    {
+      "epoch": 0.6188223169532019,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.620648049573815e-05,
+      "loss": 0.8576,
+      "step": 6605
+    },
+    {
+      "epoch": 0.6192907668524851,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.604766962798842e-05,
+      "loss": 0.9068,
+      "step": 6610
+    },
+    {
+      "epoch": 0.6197592167517684,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.588892282108798e-05,
+      "loss": 0.8765,
+      "step": 6615
+    },
+    {
+      "epoch": 0.6202276666510517,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.573024049960743e-05,
+      "loss": 0.8731,
+      "step": 6620
+    },
+    {
+      "epoch": 0.620696116550335,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.557162308794493e-05,
+      "loss": 0.8686,
+      "step": 6625
+    },
+    {
+      "epoch": 0.6211645664496183,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.541307101032497e-05,
+      "loss": 0.8672,
+      "step": 6630
+    },
+    {
+      "epoch": 0.6216330163489014,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.525458469079737e-05,
+      "loss": 0.9871,
+      "step": 6635
+    },
+    {
+      "epoch": 0.6221014662481847,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.509616455323606e-05,
+      "loss": 0.8472,
+      "step": 6640
+    },
+    {
+      "epoch": 0.622569916147468,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.493781102133796e-05,
+      "loss": 0.8612,
+      "step": 6645
+    },
+    {
+      "epoch": 0.6230383660467513,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.47795245186218e-05,
+      "loss": 0.8473,
+      "step": 6650
+    },
+    {
+      "epoch": 0.6235068159460345,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.462130546842711e-05,
+      "loss": 0.8701,
+      "step": 6655
+    },
+    {
+      "epoch": 0.6239752658453178,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.446315429391304e-05,
+      "loss": 0.8779,
+      "step": 6660
+    },
+    {
+      "epoch": 0.6244437157446011,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.430507141805714e-05,
+      "loss": 0.8669,
+      "step": 6665
+    },
+    {
+      "epoch": 0.6249121656438844,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.414705726365427e-05,
+      "loss": 0.8459,
+      "step": 6670
+    },
+    {
+      "epoch": 0.6253806155431677,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.398911225331561e-05,
+      "loss": 0.8663,
+      "step": 6675
+    },
+    {
+      "epoch": 0.6258490654424509,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.383123680946729e-05,
+      "loss": 0.8371,
+      "step": 6680
+    },
+    {
+      "epoch": 0.6263175153417342,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.367343135434952e-05,
+      "loss": 0.8637,
+      "step": 6685
+    },
+    {
+      "epoch": 0.6267859652410175,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.351569631001512e-05,
+      "loss": 0.8644,
+      "step": 6690
+    },
+    {
+      "epoch": 0.6272544151403008,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.33580320983288e-05,
+      "loss": 0.8658,
+      "step": 6695
+    },
+    {
+      "epoch": 0.627722865039584,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.32004391409657e-05,
+      "loss": 0.8808,
+      "step": 6700
+    },
+    {
+      "epoch": 0.6281913149388673,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.304291785941048e-05,
+      "loss": 0.8874,
+      "step": 6705
+    },
+    {
+      "epoch": 0.6286597648381506,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.288546867495603e-05,
+      "loss": 0.8561,
+      "step": 6710
+    },
+    {
+      "epoch": 0.6291282147374339,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.27280920087024e-05,
+      "loss": 0.877,
+      "step": 6715
+    },
+    {
+      "epoch": 0.6295966646367172,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.25707882815558e-05,
+      "loss": 0.9023,
+      "step": 6720
+    },
+    {
+      "epoch": 0.6300651145360003,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.241355791422728e-05,
+      "loss": 0.8688,
+      "step": 6725
+    },
+    {
+      "epoch": 0.6305335644352836,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.225640132723165e-05,
+      "loss": 0.865,
+      "step": 6730
+    },
+    {
+      "epoch": 0.6310020143345669,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.209931894088648e-05,
+      "loss": 0.8624,
+      "step": 6735
+    },
+    {
+      "epoch": 0.6314704642338502,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.194231117531084e-05,
+      "loss": 0.8862,
+      "step": 6740
+    },
+    {
+      "epoch": 0.6319389141331334,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.178537845042428e-05,
+      "loss": 0.8646,
+      "step": 6745
+    },
+    {
+      "epoch": 0.6324073640324167,
+      "grad_norm": 0.625,
+      "learning_rate": 7.162852118594554e-05,
+      "loss": 0.8422,
+      "step": 6750
+    },
+    {
+      "epoch": 0.6328758139317,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.147173980139166e-05,
+      "loss": 0.874,
+      "step": 6755
+    },
+    {
+      "epoch": 0.6333442638309833,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.131503471607668e-05,
+      "loss": 0.8584,
+      "step": 6760
+    },
+    {
+      "epoch": 0.6338127137302666,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.115840634911064e-05,
+      "loss": 0.8536,
+      "step": 6765
+    },
+    {
+      "epoch": 0.6342811636295498,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.100185511939822e-05,
+      "loss": 0.8542,
+      "step": 6770
+    },
+    {
+      "epoch": 0.6347496135288331,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.084538144563802e-05,
+      "loss": 0.8574,
+      "step": 6775
+    },
+    {
+      "epoch": 0.6352180634281164,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.068898574632103e-05,
+      "loss": 0.8519,
+      "step": 6780
+    },
+    {
+      "epoch": 0.6356865133273997,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.053266843972987e-05,
+      "loss": 0.8891,
+      "step": 6785
+    },
+    {
+      "epoch": 0.6361549632266829,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.037642994393736e-05,
+      "loss": 0.8495,
+      "step": 6790
+    },
+    {
+      "epoch": 0.6366234131259662,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.022027067680558e-05,
+      "loss": 0.8409,
+      "step": 6795
+    },
+    {
+      "epoch": 0.6370918630252494,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.006419105598474e-05,
+      "loss": 0.8634,
+      "step": 6800
+    },
+    {
+      "epoch": 0.6375603129245327,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.9908191498912e-05,
+      "loss": 0.8375,
+      "step": 6805
+    },
+    {
+      "epoch": 0.638028762823816,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.975227242281042e-05,
+      "loss": 0.8665,
+      "step": 6810
+    },
+    {
+      "epoch": 0.6384972127230992,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.959643424468774e-05,
+      "loss": 0.8747,
+      "step": 6815
+    },
+    {
+      "epoch": 0.6389656626223825,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.944067738133544e-05,
+      "loss": 0.8916,
+      "step": 6820
+    },
+    {
+      "epoch": 0.6394341125216658,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.928500224932746e-05,
+      "loss": 0.8789,
+      "step": 6825
+    },
+    {
+      "epoch": 0.6399025624209491,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.912940926501914e-05,
+      "loss": 0.8581,
+      "step": 6830
+    },
+    {
+      "epoch": 0.6403710123202323,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.897389884454618e-05,
+      "loss": 0.9861,
+      "step": 6835
+    },
+    {
+      "epoch": 0.6408394622195156,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.881847140382337e-05,
+      "loss": 0.8445,
+      "step": 6840
+    },
+    {
+      "epoch": 0.6413079121187989,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.86631273585437e-05,
+      "loss": 0.8399,
+      "step": 6845
+    },
+    {
+      "epoch": 0.6417763620180822,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.850786712417693e-05,
+      "loss": 0.8553,
+      "step": 6850
+    },
+    {
+      "epoch": 0.6422448119173655,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.835269111596883e-05,
+      "loss": 0.8622,
+      "step": 6855
+    },
+    {
+      "epoch": 0.6427132618166487,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.819759974893983e-05,
+      "loss": 0.8592,
+      "step": 6860
+    },
+    {
+      "epoch": 0.643181711715932,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.804259343788401e-05,
+      "loss": 0.8471,
+      "step": 6865
+    },
+    {
+      "epoch": 0.6436501616152153,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.788767259736797e-05,
+      "loss": 0.8955,
+      "step": 6870
+    },
+    {
+      "epoch": 0.6441186115144986,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.773283764172968e-05,
+      "loss": 0.8356,
+      "step": 6875
+    },
+    {
+      "epoch": 0.6445870614137817,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.757808898507747e-05,
+      "loss": 0.8782,
+      "step": 6880
+    },
+    {
+      "epoch": 0.645055511313065,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.742342704128885e-05,
+      "loss": 0.8937,
+      "step": 6885
+    },
+    {
+      "epoch": 0.6455239612123483,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.726885222400933e-05,
+      "loss": 0.8786,
+      "step": 6890
+    },
+    {
+      "epoch": 0.6459924111116316,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.71143649466515e-05,
+      "loss": 0.8774,
+      "step": 6895
+    },
+    {
+      "epoch": 0.6464608610109149,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.695996562239378e-05,
+      "loss": 0.8905,
+      "step": 6900
+    },
+    {
+      "epoch": 0.6469293109101981,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.680565466417939e-05,
+      "loss": 0.8554,
+      "step": 6905
+    },
+    {
+      "epoch": 0.6473977608094814,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.665143248471512e-05,
+      "loss": 0.8392,
+      "step": 6910
+    },
+    {
+      "epoch": 0.6478662107087647,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.649729949647048e-05,
+      "loss": 0.8728,
+      "step": 6915
+    },
+    {
+      "epoch": 0.648334660608048,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.634325611167627e-05,
+      "loss": 0.8674,
+      "step": 6920
+    },
+    {
+      "epoch": 0.6488031105073312,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.618930274232382e-05,
+      "loss": 0.8614,
+      "step": 6925
+    },
+    {
+      "epoch": 0.6492715604066145,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.603543980016349e-05,
+      "loss": 0.863,
+      "step": 6930
+    },
+    {
+      "epoch": 0.6497400103058978,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.588166769670399e-05,
+      "loss": 0.8457,
+      "step": 6935
+    },
+    {
+      "epoch": 0.6502084602051811,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.572798684321095e-05,
+      "loss": 0.8615,
+      "step": 6940
+    },
+    {
+      "epoch": 0.6506769101044644,
+      "grad_norm": 0.625,
+      "learning_rate": 6.557439765070607e-05,
+      "loss": 0.859,
+      "step": 6945
+    },
+    {
+      "epoch": 0.6511453600037476,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.542090052996582e-05,
+      "loss": 0.8606,
+      "step": 6950
+    },
+    {
+      "epoch": 0.6516138099030309,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.526749589152039e-05,
+      "loss": 0.866,
+      "step": 6955
+    },
+    {
+      "epoch": 0.6520822598023142,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.511418414565273e-05,
+      "loss": 0.8903,
+      "step": 6960
+    },
+    {
+      "epoch": 0.6525507097015975,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.496096570239731e-05,
+      "loss": 0.8673,
+      "step": 6965
+    },
+    {
+      "epoch": 0.6530191596008806,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.480784097153895e-05,
+      "loss": 0.8825,
+      "step": 6970
+    },
+    {
+      "epoch": 0.6534876095001639,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.4654810362612e-05,
+      "loss": 0.85,
+      "step": 6975
+    },
+    {
+      "epoch": 0.6539560593994472,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.450187428489898e-05,
+      "loss": 0.9843,
+      "step": 6980
+    },
+    {
+      "epoch": 0.6544245092987305,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.434903314742964e-05,
+      "loss": 0.8613,
+      "step": 6985
+    },
+    {
+      "epoch": 0.6548929591980138,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.419628735897972e-05,
+      "loss": 0.8594,
+      "step": 6990
+    },
+    {
+      "epoch": 0.655361409097297,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.404363732807009e-05,
+      "loss": 0.8643,
+      "step": 6995
+    },
+    {
+      "epoch": 0.6558298589965803,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.389108346296537e-05,
+      "loss": 0.8643,
+      "step": 7000
+    },
+    {
+      "epoch": 0.6562983088958636,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.373862617167314e-05,
+      "loss": 0.8376,
+      "step": 7005
+    },
+    {
+      "epoch": 0.6567667587951469,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.358626586194247e-05,
+      "loss": 0.8656,
+      "step": 7010
+    },
+    {
+      "epoch": 0.6572352086944301,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.343400294126328e-05,
+      "loss": 0.8631,
+      "step": 7015
+    },
+    {
+      "epoch": 0.6577036585937134,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.328183781686487e-05,
+      "loss": 0.8717,
+      "step": 7020
+    },
+    {
+      "epoch": 0.6581721084929967,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.312977089571507e-05,
+      "loss": 0.8714,
+      "step": 7025
+    },
+    {
+      "epoch": 0.65864055839228,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.297780258451904e-05,
+      "loss": 0.8613,
+      "step": 7030
+    },
+    {
+      "epoch": 0.6591090082915633,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.282593328971814e-05,
+      "loss": 0.8567,
+      "step": 7035
+    },
+    {
+      "epoch": 0.6595774581908465,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.267416341748902e-05,
+      "loss": 0.8709,
+      "step": 7040
+    },
+    {
+      "epoch": 0.6600459080901298,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.252249337374238e-05,
+      "loss": 0.8628,
+      "step": 7045
+    },
+    {
+      "epoch": 0.660514357989413,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.237092356412187e-05,
+      "loss": 0.8699,
+      "step": 7050
+    },
+    {
+      "epoch": 0.6609828078886963,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.221945439400308e-05,
+      "loss": 0.8392,
+      "step": 7055
+    },
+    {
+      "epoch": 0.6614512577879795,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.206808626849254e-05,
+      "loss": 0.8342,
+      "step": 7060
+    },
+    {
+      "epoch": 0.6619197076872628,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.19168195924264e-05,
+      "loss": 0.8653,
+      "step": 7065
+    },
+    {
+      "epoch": 0.6623881575865461,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.176565477036961e-05,
+      "loss": 0.881,
+      "step": 7070
+    },
+    {
+      "epoch": 0.6628566074858294,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.161459220661462e-05,
+      "loss": 0.8736,
+      "step": 7075
+    },
+    {
+      "epoch": 0.6633250573851127,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.146363230518036e-05,
+      "loss": 0.8709,
+      "step": 7080
+    },
+    {
+      "epoch": 0.6637935072843959,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.131277546981136e-05,
+      "loss": 0.8563,
+      "step": 7085
+    },
+    {
+      "epoch": 0.6642619571836792,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.116202210397621e-05,
+      "loss": 0.8636,
+      "step": 7090
+    },
+    {
+      "epoch": 0.6647304070829625,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.101137261086707e-05,
+      "loss": 0.8489,
+      "step": 7095
+    },
+    {
+      "epoch": 0.6651988569822458,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.086082739339808e-05,
+      "loss": 0.8412,
+      "step": 7100
+    },
+    {
+      "epoch": 0.665667306881529,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.071038685420463e-05,
+      "loss": 0.8515,
+      "step": 7105
+    },
+    {
+      "epoch": 0.6661357567808123,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.056005139564206e-05,
+      "loss": 0.846,
+      "step": 7110
+    },
+    {
+      "epoch": 0.6666042066800956,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.040982141978469e-05,
+      "loss": 0.878,
+      "step": 7115
+    },
+    {
+      "epoch": 0.6670726565793789,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.025969732842476e-05,
+      "loss": 0.8628,
+      "step": 7120
+    },
+    {
+      "epoch": 0.6675411064786622,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.0109679523071306e-05,
+      "loss": 0.8544,
+      "step": 7125
+    },
+    {
+      "epoch": 0.6680095563779453,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.995976840494904e-05,
+      "loss": 0.8405,
+      "step": 7130
+    },
+    {
+      "epoch": 0.6684780062772286,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.9809964374997396e-05,
+      "loss": 0.8664,
+      "step": 7135
+    },
+    {
+      "epoch": 0.6689464561765119,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.966026783386941e-05,
+      "loss": 0.8669,
+      "step": 7140
+    },
+    {
+      "epoch": 0.6694149060757952,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.9510679181930604e-05,
+      "loss": 0.8692,
+      "step": 7145
+    },
+    {
+      "epoch": 0.6698833559750784,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.936119881925798e-05,
+      "loss": 0.8288,
+      "step": 7150
+    },
+    {
+      "epoch": 0.6703518058743617,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.921182714563889e-05,
+      "loss": 0.8463,
+      "step": 7155
+    },
+    {
+      "epoch": 0.670820255773645,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.9062564560569976e-05,
+      "loss": 0.8557,
+      "step": 7160
+    },
+    {
+      "epoch": 0.6712887056729283,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.891341146325623e-05,
+      "loss": 0.8691,
+      "step": 7165
+    },
+    {
+      "epoch": 0.6717571555722116,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.876436825260967e-05,
+      "loss": 0.8596,
+      "step": 7170
+    },
+    {
+      "epoch": 0.6722256054714948,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.86154353272485e-05,
+      "loss": 0.9965,
+      "step": 7175
+    },
+    {
+      "epoch": 0.6726940553707781,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.8466613085495944e-05,
+      "loss": 0.8696,
+      "step": 7180
+    },
+    {
+      "epoch": 0.6731625052700614,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.831790192537923e-05,
+      "loss": 0.8722,
+      "step": 7185
+    },
+    {
+      "epoch": 0.6736309551693447,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.816930224462851e-05,
+      "loss": 0.8875,
+      "step": 7190
+    },
+    {
+      "epoch": 0.6740994050686279,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.802081444067568e-05,
+      "loss": 0.888,
+      "step": 7195
+    },
+    {
+      "epoch": 0.6745678549679112,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.7872438910653527e-05,
+      "loss": 0.8824,
+      "step": 7200
+    },
+    {
+      "epoch": 0.6750363048671945,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.772417605139455e-05,
+      "loss": 0.8589,
+      "step": 7205
+    },
+    {
+      "epoch": 0.6755047547664778,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.757602625942986e-05,
+      "loss": 0.8611,
+      "step": 7210
+    },
+    {
+      "epoch": 0.675973204665761,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.742798993098813e-05,
+      "loss": 0.8933,
+      "step": 7215
+    },
+    {
+      "epoch": 0.6764416545650442,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.7280067461994704e-05,
+      "loss": 0.8925,
+      "step": 7220
+    },
+    {
+      "epoch": 0.6769101044643275,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.713225924807034e-05,
+      "loss": 0.8408,
+      "step": 7225
+    },
+    {
+      "epoch": 0.6773785543636108,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.6984565684530165e-05,
+      "loss": 0.8312,
+      "step": 7230
+    },
+    {
+      "epoch": 0.6778470042628941,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.6836987166382725e-05,
+      "loss": 0.8491,
+      "step": 7235
+    },
+    {
+      "epoch": 0.6783154541621773,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.668952408832895e-05,
+      "loss": 0.8495,
+      "step": 7240
+    },
+    {
+      "epoch": 0.6787839040614606,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.6542176844760896e-05,
+      "loss": 0.8628,
+      "step": 7245
+    },
+    {
+      "epoch": 0.6792523539607439,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.6394945829760815e-05,
+      "loss": 0.8639,
+      "step": 7250
+    },
+    {
+      "epoch": 0.6797208038600272,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.62478314371002e-05,
+      "loss": 0.9912,
+      "step": 7255
+    },
+    {
+      "epoch": 0.6801892537593105,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.610083406023865e-05,
+      "loss": 0.8715,
+      "step": 7260
+    },
+    {
+      "epoch": 0.6806577036585937,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.595395409232265e-05,
+      "loss": 0.8925,
+      "step": 7265
+    },
+    {
+      "epoch": 0.681126153557877,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.58071919261848e-05,
+      "loss": 0.8934,
+      "step": 7270
+    },
+    {
+      "epoch": 0.6815946034571603,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.566054795434268e-05,
+      "loss": 0.8599,
+      "step": 7275
+    },
+    {
+      "epoch": 0.6820630533564436,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.551402256899757e-05,
+      "loss": 0.8519,
+      "step": 7280
+    },
+    {
+      "epoch": 0.6825315032557268,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.536761616203381e-05,
+      "loss": 0.8797,
+      "step": 7285
+    },
+    {
+      "epoch": 0.68299995315501,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.522132912501732e-05,
+      "loss": 0.8693,
+      "step": 7290
+    },
+    {
+      "epoch": 0.6834684030542933,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.5075161849194966e-05,
+      "loss": 0.8494,
+      "step": 7295
+    },
+    {
+      "epoch": 0.6839368529535766,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.492911472549314e-05,
+      "loss": 0.838,
+      "step": 7300
+    },
+    {
+      "epoch": 0.6844053028528598,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.4783188144516975e-05,
+      "loss": 0.8484,
+      "step": 7305
+    },
+    {
+      "epoch": 0.6848737527521431,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.463738249654924e-05,
+      "loss": 0.8885,
+      "step": 7310
+    },
+    {
+      "epoch": 0.6853422026514264,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.4491698171549156e-05,
+      "loss": 0.8772,
+      "step": 7315
+    },
+    {
+      "epoch": 0.6858106525507097,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.4346135559151555e-05,
+      "loss": 0.8939,
+      "step": 7320
+    },
+    {
+      "epoch": 0.686279102449993,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.420069504866576e-05,
+      "loss": 0.8538,
+      "step": 7325
+    },
+    {
+      "epoch": 0.6867475523492762,
+      "grad_norm": 0.625,
+      "learning_rate": 5.405537702907445e-05,
+      "loss": 0.8417,
+      "step": 7330
+    },
+    {
+      "epoch": 0.6872160022485595,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.391018188903271e-05,
+      "loss": 0.8472,
+      "step": 7335
+    },
+    {
+      "epoch": 0.6876844521478428,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.376511001686704e-05,
+      "loss": 0.8479,
+      "step": 7340
+    },
+    {
+      "epoch": 0.6881529020471261,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.362016180057425e-05,
+      "loss": 0.8684,
+      "step": 7345
+    },
+    {
+      "epoch": 0.6886213519464093,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.347533762782044e-05,
+      "loss": 0.8443,
+      "step": 7350
+    },
+    {
+      "epoch": 0.6890898018456926,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.333063788593986e-05,
+      "loss": 0.8753,
+      "step": 7355
+    },
+    {
+      "epoch": 0.6895582517449759,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.318606296193405e-05,
+      "loss": 0.8848,
+      "step": 7360
+    },
+    {
+      "epoch": 0.6900267016442592,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.304161324247077e-05,
+      "loss": 0.8622,
+      "step": 7365
+    },
+    {
+      "epoch": 0.6904951515435425,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.289728911388279e-05,
+      "loss": 0.8593,
+      "step": 7370
+    },
+    {
+      "epoch": 0.6909636014428256,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.275309096216704e-05,
+      "loss": 0.8212,
+      "step": 7375
+    },
+    {
+      "epoch": 0.6914320513421089,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.260901917298355e-05,
+      "loss": 0.8615,
+      "step": 7380
+    },
+    {
+      "epoch": 0.6919005012413922,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.246507413165445e-05,
+      "loss": 0.9007,
+      "step": 7385
+    },
+    {
+      "epoch": 0.6923689511406755,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.232125622316267e-05,
+      "loss": 0.8551,
+      "step": 7390
+    },
+    {
+      "epoch": 0.6928374010399587,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.2177565832151355e-05,
+      "loss": 0.8723,
+      "step": 7395
+    },
+    {
+      "epoch": 0.693305850939242,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.2034003342922514e-05,
+      "loss": 0.8591,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6937743008385253,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.189056913943604e-05,
+      "loss": 0.877,
+      "step": 7405
+    },
+    {
+      "epoch": 0.6942427507378086,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.17472636053087e-05,
+      "loss": 0.8619,
+      "step": 7410
+    },
+    {
+      "epoch": 0.6947112006370919,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.160408712381327e-05,
+      "loss": 0.8525,
+      "step": 7415
+    },
+    {
+      "epoch": 0.6951796505363751,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.146104007787731e-05,
+      "loss": 0.832,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6956481004356584,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.131812285008208e-05,
+      "loss": 0.8796,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6961165503349417,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.117533582266183e-05,
+      "loss": 0.8522,
+      "step": 7430
+    },
+    {
+      "epoch": 0.696585000234225,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.103267937750251e-05,
+      "loss": 0.8844,
+      "step": 7435
+    },
+    {
+      "epoch": 0.6970534501335082,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.089015389614076e-05,
+      "loss": 0.8797,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6975219000327915,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.074775975976307e-05,
+      "loss": 0.8726,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6979903499320748,
+      "grad_norm": 0.734375,
+      "learning_rate": 5.0605497349204525e-05,
+      "loss": 0.8703,
+      "step": 7450
+    },
+    {
+      "epoch": 0.6984587998313581,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.046336704494804e-05,
+      "loss": 0.8829,
+      "step": 7455
+    },
+    {
+      "epoch": 0.6989272497306414,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.032136922712307e-05,
+      "loss": 0.8909,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6993956996299245,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.017950427550485e-05,
+      "loss": 0.8634,
+      "step": 7465
+    },
+    {
+      "epoch": 0.6998641495292078,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.003777256951321e-05,
+      "loss": 0.8671,
+      "step": 7470
+    },
+    {
+      "epoch": 0.7003325994284911,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.989617448821166e-05,
+      "loss": 0.8632,
+      "step": 7475
+    },
+    {
+      "epoch": 0.7008010493277744,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.9754710410306196e-05,
+      "loss": 0.8988,
+      "step": 7480
+    },
+    {
+      "epoch": 0.7012694992270576,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.961338071414462e-05,
+      "loss": 0.845,
+      "step": 7485
+    },
+    {
+      "epoch": 0.7017379491263409,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.9472185777715155e-05,
+      "loss": 0.8852,
+      "step": 7490
+    },
+    {
+      "epoch": 0.7022063990256242,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.933112597864564e-05,
+      "loss": 0.8198,
+      "step": 7495
+    },
+    {
+      "epoch": 0.7026748489249075,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.9190201694202574e-05,
+      "loss": 0.8643,
+      "step": 7500
+    },
+    {
+      "epoch": 0.7031432988241908,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.904941330128996e-05,
+      "loss": 0.8516,
+      "step": 7505
+    },
+    {
+      "epoch": 0.703611748723474,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.890876117644839e-05,
+      "loss": 0.8518,
+      "step": 7510
+    },
+    {
+      "epoch": 0.7040801986227573,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.876824569585391e-05,
+      "loss": 0.8658,
+      "step": 7515
+    },
+    {
+      "epoch": 0.7045486485220406,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.862786723531722e-05,
+      "loss": 0.8338,
+      "step": 7520
+    },
+    {
+      "epoch": 0.7050170984213239,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.848762617028254e-05,
+      "loss": 0.8643,
+      "step": 7525
+    },
+    {
+      "epoch": 0.7054855483206071,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.8347522875826576e-05,
+      "loss": 0.8682,
+      "step": 7530
+    },
+    {
+      "epoch": 0.7059539982198904,
+      "grad_norm": 0.625,
+      "learning_rate": 4.8207557726657536e-05,
+      "loss": 0.8499,
+      "step": 7535
+    },
+    {
+      "epoch": 0.7064224481191737,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.8067731097114264e-05,
+      "loss": 0.86,
+      "step": 7540
+    },
+    {
+      "epoch": 0.706890898018457,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.79280433611651e-05,
+      "loss": 0.8588,
+      "step": 7545
+    },
+    {
+      "epoch": 0.7073593479177402,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.778849489240681e-05,
+      "loss": 0.8547,
+      "step": 7550
+    },
+    {
+      "epoch": 0.7078277978170234,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.764908606406382e-05,
+      "loss": 0.864,
+      "step": 7555
+    },
+    {
+      "epoch": 0.7082962477163067,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.7509817248987046e-05,
+      "loss": 0.8515,
+      "step": 7560
+    },
+    {
+      "epoch": 0.70876469761559,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.737068881965289e-05,
+      "loss": 0.9113,
+      "step": 7565
+    },
+    {
+      "epoch": 0.7092331475148733,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.723170114816228e-05,
+      "loss": 0.8639,
+      "step": 7570
+    },
+    {
+      "epoch": 0.7097015974141565,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.7092854606239775e-05,
+      "loss": 0.8589,
+      "step": 7575
+    },
+    {
+      "epoch": 0.7101700473134398,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.6954149565232464e-05,
+      "loss": 0.823,
+      "step": 7580
+    },
+    {
+      "epoch": 0.7106384972127231,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.681558639610888e-05,
+      "loss": 0.8547,
+      "step": 7585
+    },
+    {
+      "epoch": 0.7111069471120064,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.667716546945824e-05,
+      "loss": 0.8447,
+      "step": 7590
+    },
+    {
+      "epoch": 0.7115753970112897,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.653888715548932e-05,
+      "loss": 0.8617,
+      "step": 7595
+    },
+    {
+      "epoch": 0.7120438469105729,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.6400751824029366e-05,
+      "loss": 0.8794,
+      "step": 7600
+    },
+    {
+      "epoch": 0.7125122968098562,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.626275984452337e-05,
+      "loss": 0.8515,
+      "step": 7605
+    },
+    {
+      "epoch": 0.7129807467091395,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.612491158603278e-05,
+      "loss": 0.8772,
+      "step": 7610
+    },
+    {
+      "epoch": 0.7134491966084228,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.59872074172348e-05,
+      "loss": 0.8956,
+      "step": 7615
+    },
+    {
+      "epoch": 0.713917646507706,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.584964770642112e-05,
+      "loss": 0.8616,
+      "step": 7620
+    },
+    {
+      "epoch": 0.7143860964069892,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.571223282149718e-05,
+      "loss": 0.8765,
+      "step": 7625
+    },
+    {
+      "epoch": 0.7148545463062725,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.5574963129981063e-05,
+      "loss": 0.8697,
+      "step": 7630
+    },
+    {
+      "epoch": 0.7153229962055558,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.543783899900255e-05,
+      "loss": 0.8554,
+      "step": 7635
+    },
+    {
+      "epoch": 0.7157914461048391,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.5300860795302005e-05,
+      "loss": 0.8471,
+      "step": 7640
+    },
+    {
+      "epoch": 0.7162598960041223,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.516402888522966e-05,
+      "loss": 0.8595,
+      "step": 7645
+    },
+    {
+      "epoch": 0.7167283459034056,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.502734363474436e-05,
+      "loss": 0.8475,
+      "step": 7650
+    },
+    {
+      "epoch": 0.7171967958026889,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.489080540941279e-05,
+      "loss": 0.8655,
+      "step": 7655
+    },
+    {
+      "epoch": 0.7176652457019722,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.475441457440836e-05,
+      "loss": 0.8455,
+      "step": 7660
+    },
+    {
+      "epoch": 0.7181336956012554,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.461817149451028e-05,
+      "loss": 0.8461,
+      "step": 7665
+    },
+    {
+      "epoch": 0.7186021455005387,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.448207653410271e-05,
+      "loss": 0.8912,
+      "step": 7670
+    },
+    {
+      "epoch": 0.719070595399822,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.434613005717345e-05,
+      "loss": 0.8614,
+      "step": 7675
+    },
+    {
+      "epoch": 0.7195390452991053,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.4210332427313336e-05,
+      "loss": 0.8799,
+      "step": 7680
+    },
+    {
+      "epoch": 0.7200074951983886,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.4074684007715104e-05,
+      "loss": 0.8748,
+      "step": 7685
+    },
+    {
+      "epoch": 0.7204759450976718,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.393918516117236e-05,
+      "loss": 0.8428,
+      "step": 7690
+    },
+    {
+      "epoch": 0.7209443949969551,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.380383625007866e-05,
+      "loss": 0.8637,
+      "step": 7695
+    },
+    {
+      "epoch": 0.7214128448962384,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.3668637636426626e-05,
+      "loss": 0.8726,
+      "step": 7700
+    },
+    {
+      "epoch": 0.7218812947955217,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.353358968180692e-05,
+      "loss": 0.8501,
+      "step": 7705
+    },
+    {
+      "epoch": 0.7223497446948048,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.3398692747407154e-05,
+      "loss": 0.8696,
+      "step": 7710
+    },
+    {
+      "epoch": 0.7228181945940881,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.326394719401115e-05,
+      "loss": 0.8518,
+      "step": 7715
+    },
+    {
+      "epoch": 0.7232866444933714,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.312935338199783e-05,
+      "loss": 0.8656,
+      "step": 7720
+    },
+    {
+      "epoch": 0.7237550943926547,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.299491167134023e-05,
+      "loss": 0.8786,
+      "step": 7725
+    },
+    {
+      "epoch": 0.724223544291938,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.28606224216046e-05,
+      "loss": 0.8635,
+      "step": 7730
+    },
+    {
+      "epoch": 0.7246919941912212,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.272648599194948e-05,
+      "loss": 0.8453,
+      "step": 7735
+    },
+    {
+      "epoch": 0.7251604440905045,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.2592502741124706e-05,
+      "loss": 0.8648,
+      "step": 7740
+    },
+    {
+      "epoch": 0.7256288939897878,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.245867302747032e-05,
+      "loss": 0.8661,
+      "step": 7745
+    },
+    {
+      "epoch": 0.7260973438890711,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.232499720891584e-05,
+      "loss": 0.8901,
+      "step": 7750
+    },
+    {
+      "epoch": 0.7265657937883543,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.219147564297918e-05,
+      "loss": 0.8346,
+      "step": 7755
+    },
+    {
+      "epoch": 0.7270342436876376,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.20581086867656e-05,
+      "loss": 0.8532,
+      "step": 7760
+    },
+    {
+      "epoch": 0.7275026935869209,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.192489669696701e-05,
+      "loss": 0.8474,
+      "step": 7765
+    },
+    {
+      "epoch": 0.7279711434862042,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.179184002986069e-05,
+      "loss": 0.8478,
+      "step": 7770
+    },
+    {
+      "epoch": 0.7284395933854875,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.1658939041308674e-05,
+      "loss": 0.8956,
+      "step": 7775
+    },
+    {
+      "epoch": 0.7289080432847707,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.1526194086756475e-05,
+      "loss": 0.8706,
+      "step": 7780
+    },
+    {
+      "epoch": 0.729376493184054,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.1393605521232414e-05,
+      "loss": 0.8881,
+      "step": 7785
+    },
+    {
+      "epoch": 0.7298449430833372,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.12611736993465e-05,
+      "loss": 0.8509,
+      "step": 7790
+    },
+    {
+      "epoch": 0.7303133929826205,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.112889897528955e-05,
+      "loss": 0.8266,
+      "step": 7795
+    },
+    {
+      "epoch": 0.7307818428819037,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.0996781702832156e-05,
+      "loss": 0.8849,
+      "step": 7800
+    },
+    {
+      "epoch": 0.731250292781187,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.0864822235323895e-05,
+      "loss": 0.859,
+      "step": 7805
+    },
+    {
+      "epoch": 0.7317187426804703,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.073302092569219e-05,
+      "loss": 0.8548,
+      "step": 7810
+    },
+    {
+      "epoch": 0.7321871925797536,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.060137812644158e-05,
+      "loss": 0.8576,
+      "step": 7815
+    },
+    {
+      "epoch": 0.7326556424790369,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.046989418965255e-05,
+      "loss": 0.8251,
+      "step": 7820
+    },
+    {
+      "epoch": 0.7331240923783201,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.033856946698079e-05,
+      "loss": 0.8535,
+      "step": 7825
+    },
+    {
+      "epoch": 0.7335925422776034,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.020740430965619e-05,
+      "loss": 0.8719,
+      "step": 7830
+    },
+    {
+      "epoch": 0.7340609921768867,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.007639906848177e-05,
+      "loss": 1.0032,
+      "step": 7835
+    },
+    {
+      "epoch": 0.73452944207617,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.9945554093832934e-05,
+      "loss": 0.8566,
+      "step": 7840
+    },
+    {
+      "epoch": 0.7349978919754532,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.981486973565648e-05,
+      "loss": 0.8843,
+      "step": 7845
+    },
+    {
+      "epoch": 0.7354663418747365,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.968434634346956e-05,
+      "loss": 0.887,
+      "step": 7850
+    },
+    {
+      "epoch": 0.7359347917740198,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.955398426635879e-05,
+      "loss": 0.8646,
+      "step": 7855
+    },
+    {
+      "epoch": 0.7364032416733031,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.942378385297946e-05,
+      "loss": 0.8363,
+      "step": 7860
+    },
+    {
+      "epoch": 0.7368716915725864,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.929374545155445e-05,
+      "loss": 0.8445,
+      "step": 7865
+    },
+    {
+      "epoch": 0.7373401414718695,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.9163869409873275e-05,
+      "loss": 0.8709,
+      "step": 7870
+    },
+    {
+      "epoch": 0.7378085913711528,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.9034156075291274e-05,
+      "loss": 0.8639,
+      "step": 7875
+    },
+    {
+      "epoch": 0.7382770412704361,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.890460579472866e-05,
+      "loss": 0.8563,
+      "step": 7880
+    },
+    {
+      "epoch": 0.7387454911697194,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.877521891466947e-05,
+      "loss": 0.8573,
+      "step": 7885
+    },
+    {
+      "epoch": 0.7392139410690026,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.8645995781160715e-05,
+      "loss": 0.8599,
+      "step": 7890
+    },
+    {
+      "epoch": 0.7396823909682859,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.851693673981156e-05,
+      "loss": 0.8644,
+      "step": 7895
+    },
+    {
+      "epoch": 0.7401508408675692,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.838804213579227e-05,
+      "loss": 0.8637,
+      "step": 7900
+    },
+    {
+      "epoch": 0.7406192907668525,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.825931231383324e-05,
+      "loss": 0.8653,
+      "step": 7905
+    },
+    {
+      "epoch": 0.7410877406661358,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.813074761822426e-05,
+      "loss": 0.8737,
+      "step": 7910
+    },
+    {
+      "epoch": 0.741556190565419,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.8002348392813425e-05,
+      "loss": 0.8605,
+      "step": 7915
+    },
+    {
+      "epoch": 0.7420246404647023,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.787411498100635e-05,
+      "loss": 0.8502,
+      "step": 7920
+    },
+    {
+      "epoch": 0.7424930903639856,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.7746047725765066e-05,
+      "loss": 0.8986,
+      "step": 7925
+    },
+    {
+      "epoch": 0.7429615402632689,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.761814696960726e-05,
+      "loss": 0.8629,
+      "step": 7930
+    },
+    {
+      "epoch": 0.7434299901625521,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.749041305460539e-05,
+      "loss": 0.8515,
+      "step": 7935
+    },
+    {
+      "epoch": 0.7438984400618354,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.736284632238556e-05,
+      "loss": 0.8472,
+      "step": 7940
+    },
+    {
+      "epoch": 0.7443668899611187,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.7235447114126856e-05,
+      "loss": 0.834,
+      "step": 7945
+    },
+    {
+      "epoch": 0.744835339860402,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.7108215770560284e-05,
+      "loss": 0.876,
+      "step": 7950
+    },
+    {
+      "epoch": 0.7453037897596853,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.698115263196791e-05,
+      "loss": 0.8666,
+      "step": 7955
+    },
+    {
+      "epoch": 0.7457722396589684,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.685425803818184e-05,
+      "loss": 0.8773,
+      "step": 7960
+    },
+    {
+      "epoch": 0.7462406895582517,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.672753232858356e-05,
+      "loss": 0.85,
+      "step": 7965
+    },
+    {
+      "epoch": 0.746709139457535,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.66009758421027e-05,
+      "loss": 0.8638,
+      "step": 7970
+    },
+    {
+      "epoch": 0.7471775893568183,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.6474588917216476e-05,
+      "loss": 0.836,
+      "step": 7975
+    },
+    {
+      "epoch": 0.7476460392561015,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.6348371891948455e-05,
+      "loss": 0.8471,
+      "step": 7980
+    },
+    {
+      "epoch": 0.7481144891553848,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.6222325103867885e-05,
+      "loss": 0.8684,
+      "step": 7985
+    },
+    {
+      "epoch": 0.7485829390546681,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.6096448890088743e-05,
+      "loss": 0.8594,
+      "step": 7990
+    },
+    {
+      "epoch": 0.7490513889539514,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.597074358726869e-05,
+      "loss": 0.8713,
+      "step": 7995
+    },
+    {
+      "epoch": 0.7495198388532347,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.584520953160839e-05,
+      "loss": 0.8725,
+      "step": 8000
+    },
+    {
+      "epoch": 0.7499882887525179,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.57198470588505e-05,
+      "loss": 0.8481,
+      "step": 8005
+    },
+    {
+      "epoch": 0.7504567386518012,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.55946565042787e-05,
+      "loss": 0.85,
+      "step": 8010
+    },
+    {
+      "epoch": 0.7509251885510845,
+      "grad_norm": 0.625,
+      "learning_rate": 3.546963820271689e-05,
+      "loss": 0.8442,
+      "step": 8015
+    },
+    {
+      "epoch": 0.7513936384503678,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.534479248852833e-05,
+      "loss": 0.8686,
+      "step": 8020
+    },
+    {
+      "epoch": 0.751862088349651,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.522011969561469e-05,
+      "loss": 0.8741,
+      "step": 8025
+    },
+    {
+      "epoch": 0.7523305382489343,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.509562015741509e-05,
+      "loss": 0.876,
+      "step": 8030
+    },
+    {
+      "epoch": 0.7527989881482176,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.4971294206905326e-05,
+      "loss": 0.838,
+      "step": 8035
+    },
+    {
+      "epoch": 0.7532674380475008,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.484714217659698e-05,
+      "loss": 0.8595,
+      "step": 8040
+    },
+    {
+      "epoch": 0.7537358879467841,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.4723164398536376e-05,
+      "loss": 0.8687,
+      "step": 8045
+    },
+    {
+      "epoch": 0.7542043378460673,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.459936120430384e-05,
+      "loss": 0.8403,
+      "step": 8050
+    },
+    {
+      "epoch": 0.7546727877453506,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.447573292501279e-05,
+      "loss": 0.8502,
+      "step": 8055
+    },
+    {
+      "epoch": 0.7551412376446339,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.435227989130888e-05,
+      "loss": 0.8602,
+      "step": 8060
+    },
+    {
+      "epoch": 0.7556096875439172,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.422900243336893e-05,
+      "loss": 0.91,
+      "step": 8065
+    },
+    {
+      "epoch": 0.7560781374432004,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.410590088090029e-05,
+      "loss": 0.8619,
+      "step": 8070
+    },
+    {
+      "epoch": 0.7565465873424837,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.398297556313983e-05,
+      "loss": 0.8591,
+      "step": 8075
+    },
+    {
+      "epoch": 0.757015037241767,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.386022680885309e-05,
+      "loss": 0.873,
+      "step": 8080
+    },
+    {
+      "epoch": 0.7574834871410503,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.373765494633333e-05,
+      "loss": 0.8759,
+      "step": 8085
+    },
+    {
+      "epoch": 0.7579519370403336,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.361526030340072e-05,
+      "loss": 0.8043,
+      "step": 8090
+    },
+    {
+      "epoch": 0.7584203869396168,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.3493043207401496e-05,
+      "loss": 0.8367,
+      "step": 8095
+    },
+    {
+      "epoch": 0.7588888368389001,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.3371003985207085e-05,
+      "loss": 0.8603,
+      "step": 8100
+    },
+    {
+      "epoch": 0.7593572867381834,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.3249142963213045e-05,
+      "loss": 0.8679,
+      "step": 8105
+    },
+    {
+      "epoch": 0.7598257366374667,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.312746046733844e-05,
+      "loss": 0.8993,
+      "step": 8110
+    },
+    {
+      "epoch": 0.7602941865367498,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.30059568230249e-05,
+      "loss": 0.8778,
+      "step": 8115
+    },
+    {
+      "epoch": 0.7607626364360331,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.288463235523557e-05,
+      "loss": 0.8611,
+      "step": 8120
+    },
+    {
+      "epoch": 0.7612310863353164,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.276348738845454e-05,
+      "loss": 0.8965,
+      "step": 8125
+    },
+    {
+      "epoch": 0.7616995362345997,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.264252224668569e-05,
+      "loss": 0.8374,
+      "step": 8130
+    },
+    {
+      "epoch": 0.762167986133883,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.252173725345208e-05,
+      "loss": 0.8614,
+      "step": 8135
+    },
+    {
+      "epoch": 0.7626364360331662,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.2401132731794824e-05,
+      "loss": 0.8493,
+      "step": 8140
+    },
+    {
+      "epoch": 0.7631048859324495,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.228070900427249e-05,
+      "loss": 0.8571,
+      "step": 8145
+    },
+    {
+      "epoch": 0.7635733358317328,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.216046639296008e-05,
+      "loss": 0.8627,
+      "step": 8150
+    },
+    {
+      "epoch": 0.7640417857310161,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.204040521944809e-05,
+      "loss": 0.8788,
+      "step": 8155
+    },
+    {
+      "epoch": 0.7645102356302993,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.1920525804841886e-05,
+      "loss": 0.8318,
+      "step": 8160
+    },
+    {
+      "epoch": 0.7649786855295826,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.180082846976071e-05,
+      "loss": 0.8791,
+      "step": 8165
+    },
+    {
+      "epoch": 0.7654471354288659,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.168131353433676e-05,
+      "loss": 0.8471,
+      "step": 8170
+    },
+    {
+      "epoch": 0.7659155853281492,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.1561981318214396e-05,
+      "loss": 0.8715,
+      "step": 8175
+    },
+    {
+      "epoch": 0.7663840352274325,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.1442832140549374e-05,
+      "loss": 0.8531,
+      "step": 8180
+    },
+    {
+      "epoch": 0.7668524851267157,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.132386632000789e-05,
+      "loss": 0.8577,
+      "step": 8185
+    },
+    {
+      "epoch": 0.767320935025999,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.120508417476567e-05,
+      "loss": 0.8718,
+      "step": 8190
+    },
+    {
+      "epoch": 0.7677893849252823,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.10864860225073e-05,
+      "loss": 0.8602,
+      "step": 8195
+    },
+    {
+      "epoch": 0.7682578348245656,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.0968072180425244e-05,
+      "loss": 0.8472,
+      "step": 8200
+    },
+    {
+      "epoch": 0.7687262847238487,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.084984296521898e-05,
+      "loss": 0.859,
+      "step": 8205
+    },
+    {
+      "epoch": 0.769194734623132,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.0731798693094213e-05,
+      "loss": 0.8413,
+      "step": 8210
+    },
+    {
+      "epoch": 0.7696631845224153,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.061393967976205e-05,
+      "loss": 0.8662,
+      "step": 8215
+    },
+    {
+      "epoch": 0.7701316344216986,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.0496266240438154e-05,
+      "loss": 0.8721,
+      "step": 8220
+    },
+    {
+      "epoch": 0.7706000843209819,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.0378778689841736e-05,
+      "loss": 0.8383,
+      "step": 8225
+    },
+    {
+      "epoch": 0.7710685342202651,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.0261477342194967e-05,
+      "loss": 0.8414,
+      "step": 8230
+    },
+    {
+      "epoch": 0.7715369841195484,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.014436251122197e-05,
+      "loss": 0.8352,
+      "step": 8235
+    },
+    {
+      "epoch": 0.7720054340188317,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.0027434510148055e-05,
+      "loss": 0.8286,
+      "step": 8240
+    },
+    {
+      "epoch": 0.772473883918115,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.9910693651698827e-05,
+      "loss": 0.8791,
+      "step": 8245
+    },
+    {
+      "epoch": 0.7729423338173982,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.979414024809931e-05,
+      "loss": 0.8561,
+      "step": 8250
+    },
+    {
+      "epoch": 0.7734107837166815,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.9677774611073307e-05,
+      "loss": 0.8361,
+      "step": 8255
+    },
+    {
+      "epoch": 0.7738792336159648,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.9561597051842384e-05,
+      "loss": 0.8456,
+      "step": 8260
+    },
+    {
+      "epoch": 0.7743476835152481,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.9445607881125024e-05,
+      "loss": 0.8525,
+      "step": 8265
+    },
+    {
+      "epoch": 0.7748161334145313,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.9329807409135945e-05,
+      "loss": 0.8702,
+      "step": 8270
+    },
+    {
+      "epoch": 0.7752845833138146,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.921419594558521e-05,
+      "loss": 0.8843,
+      "step": 8275
+    },
+    {
+      "epoch": 0.7757530332130979,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.909877379967725e-05,
+      "loss": 0.8471,
+      "step": 8280
+    },
+    {
+      "epoch": 0.7762214831123811,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.8983541280110306e-05,
+      "loss": 0.8709,
+      "step": 8285
+    },
+    {
+      "epoch": 0.7766899330116644,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.8868498695075343e-05,
+      "loss": 0.8505,
+      "step": 8290
+    },
+    {
+      "epoch": 0.7771583829109476,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.8753646352255437e-05,
+      "loss": 0.836,
+      "step": 8295
+    },
+    {
+      "epoch": 0.7776268328102309,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.8638984558824777e-05,
+      "loss": 0.8435,
+      "step": 8300
+    },
+    {
+      "epoch": 0.7780952827095142,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.8524513621447992e-05,
+      "loss": 0.8428,
+      "step": 8305
+    },
+    {
+      "epoch": 0.7785637326087975,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.8410233846279255e-05,
+      "loss": 0.8552,
+      "step": 8310
+    },
+    {
+      "epoch": 0.7790321825080807,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.8296145538961404e-05,
+      "loss": 0.8468,
+      "step": 8315
+    },
+    {
+      "epoch": 0.779500632407364,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.818224900462527e-05,
+      "loss": 0.8536,
+      "step": 8320
+    },
+    {
+      "epoch": 0.7799690823066473,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.80685445478888e-05,
+      "loss": 0.8225,
+      "step": 8325
+    },
+    {
+      "epoch": 0.7804375322059306,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.7955032472856146e-05,
+      "loss": 0.8995,
+      "step": 8330
+    },
+    {
+      "epoch": 0.7809059821052139,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.7841713083116938e-05,
+      "loss": 1.0004,
+      "step": 8335
+    },
+    {
+      "epoch": 0.7813744320044971,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.772858668174555e-05,
+      "loss": 0.8579,
+      "step": 8340
+    },
+    {
+      "epoch": 0.7818428819037804,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.761565357130016e-05,
+      "loss": 0.89,
+      "step": 8345
+    },
+    {
+      "epoch": 0.7823113318030637,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.7502914053821936e-05,
+      "loss": 0.8625,
+      "step": 8350
+    },
+    {
+      "epoch": 0.782779781702347,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.7390368430834368e-05,
+      "loss": 0.883,
+      "step": 8355
+    },
+    {
+      "epoch": 0.7832482316016302,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.7278017003342306e-05,
+      "loss": 0.8934,
+      "step": 8360
+    },
+    {
+      "epoch": 0.7837166815009134,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.716586007183133e-05,
+      "loss": 0.8615,
+      "step": 8365
+    },
+    {
+      "epoch": 0.7841851314001967,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.7053897936266615e-05,
+      "loss": 0.8772,
+      "step": 8370
+    },
+    {
+      "epoch": 0.78465358129948,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.6942130896092553e-05,
+      "loss": 0.8452,
+      "step": 8375
+    },
+    {
+      "epoch": 0.7851220311987633,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.6830559250231725e-05,
+      "loss": 0.8703,
+      "step": 8380
+    },
+    {
+      "epoch": 0.7855904810980465,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.6719183297084026e-05,
+      "loss": 0.8448,
+      "step": 8385
+    },
+    {
+      "epoch": 0.7860589309973298,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.6608003334526055e-05,
+      "loss": 0.8319,
+      "step": 8390
+    },
+    {
+      "epoch": 0.7865273808966131,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.6497019659910216e-05,
+      "loss": 0.87,
+      "step": 8395
+    },
+    {
+      "epoch": 0.7869958307958964,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.6386232570063964e-05,
+      "loss": 0.8662,
+      "step": 8400
+    },
+    {
+      "epoch": 0.7874642806951796,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.627564236128891e-05,
+      "loss": 0.8614,
+      "step": 8405
+    },
+    {
+      "epoch": 0.7879327305944629,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.616524932936012e-05,
+      "loss": 0.8524,
+      "step": 8410
+    },
+    {
+      "epoch": 0.7884011804937462,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.6055053769525372e-05,
+      "loss": 0.837,
+      "step": 8415
+    },
+    {
+      "epoch": 0.7888696303930295,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.5945055976504275e-05,
+      "loss": 0.894,
+      "step": 8420
+    },
+    {
+      "epoch": 0.7893380802923128,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.583525624448745e-05,
+      "loss": 0.8656,
+      "step": 8425
+    },
+    {
+      "epoch": 0.789806530191596,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.5725654867135872e-05,
+      "loss": 0.8296,
+      "step": 8430
+    },
+    {
+      "epoch": 0.7902749800908793,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.561625213758001e-05,
+      "loss": 0.8475,
+      "step": 8435
+    },
+    {
+      "epoch": 0.7907434299901626,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.5507048348418983e-05,
+      "loss": 0.8399,
+      "step": 8440
+    },
+    {
+      "epoch": 0.7912118798894459,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.5398043791719928e-05,
+      "loss": 0.8458,
+      "step": 8445
+    },
+    {
+      "epoch": 0.791680329788729,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.528923875901703e-05,
+      "loss": 0.8844,
+      "step": 8450
+    },
+    {
+      "epoch": 0.7921487796880123,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.518063354131097e-05,
+      "loss": 0.8779,
+      "step": 8455
+    },
+    {
+      "epoch": 0.7926172295872956,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.507222842906788e-05,
+      "loss": 0.8833,
+      "step": 8460
+    },
+    {
+      "epoch": 0.7930856794865789,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.496402371221882e-05,
+      "loss": 0.8489,
+      "step": 8465
+    },
+    {
+      "epoch": 0.7935541293858622,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.485601968015887e-05,
+      "loss": 0.8629,
+      "step": 8470
+    },
+    {
+      "epoch": 0.7940225792851454,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.4748216621746302e-05,
+      "loss": 0.8429,
+      "step": 8475
+    },
+    {
+      "epoch": 0.7944910291844287,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.4640614825301955e-05,
+      "loss": 0.8517,
+      "step": 8480
+    },
+    {
+      "epoch": 0.794959479083712,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.453321457860841e-05,
+      "loss": 0.8591,
+      "step": 8485
+    },
+    {
+      "epoch": 0.7954279289829953,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.4426016168909117e-05,
+      "loss": 0.8764,
+      "step": 8490
+    },
+    {
+      "epoch": 0.7958963788822785,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.4319019882907734e-05,
+      "loss": 0.8394,
+      "step": 8495
+    },
+    {
+      "epoch": 0.7963648287815618,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.421222600676737e-05,
+      "loss": 0.8574,
+      "step": 8500
+    },
+    {
+      "epoch": 0.7968332786808451,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.4105634826109812e-05,
+      "loss": 0.8843,
+      "step": 8505
+    },
+    {
+      "epoch": 0.7973017285801284,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.3999246626014626e-05,
+      "loss": 0.8522,
+      "step": 8510
+    },
+    {
+      "epoch": 0.7977701784794117,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.3893061691018616e-05,
+      "loss": 0.844,
+      "step": 8515
+    },
+    {
+      "epoch": 0.7982386283786949,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.3787080305114884e-05,
+      "loss": 0.8395,
+      "step": 8520
+    },
+    {
+      "epoch": 0.7987070782779782,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.3681302751752232e-05,
+      "loss": 0.8556,
+      "step": 8525
+    },
+    {
+      "epoch": 0.7991755281772615,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.3575729313834118e-05,
+      "loss": 0.8485,
+      "step": 8530
+    },
+    {
+      "epoch": 0.7996439780765447,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.3470360273718262e-05,
+      "loss": 0.8715,
+      "step": 8535
+    },
+    {
+      "epoch": 0.8001124279758279,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.3365195913215654e-05,
+      "loss": 0.861,
+      "step": 8540
+    },
+    {
+      "epoch": 0.8005808778751112,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.3260236513589927e-05,
+      "loss": 0.8597,
+      "step": 8545
+    },
+    {
+      "epoch": 0.8010493277743945,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.3155482355556425e-05,
+      "loss": 0.8451,
+      "step": 8550
+    },
+    {
+      "epoch": 0.8015177776736778,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.3050933719281663e-05,
+      "loss": 0.8659,
+      "step": 8555
+    },
+    {
+      "epoch": 0.8019862275729611,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.294659088438249e-05,
+      "loss": 0.8382,
+      "step": 8560
+    },
+    {
+      "epoch": 0.8024546774722443,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.2842454129925284e-05,
+      "loss": 0.8526,
+      "step": 8565
+    },
+    {
+      "epoch": 0.8029231273715276,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.2738523734425242e-05,
+      "loss": 0.8859,
+      "step": 8570
+    },
+    {
+      "epoch": 0.8033915772708109,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.2634799975845733e-05,
+      "loss": 0.884,
+      "step": 8575
+    },
+    {
+      "epoch": 0.8038600271700942,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.2531283131597447e-05,
+      "loss": 0.8675,
+      "step": 8580
+    },
+    {
+      "epoch": 0.8043284770693774,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.2427973478537612e-05,
+      "loss": 0.8808,
+      "step": 8585
+    },
+    {
+      "epoch": 0.8047969269686607,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.2324871292969385e-05,
+      "loss": 0.8342,
+      "step": 8590
+    },
+    {
+      "epoch": 0.805265376867944,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.222197685064107e-05,
+      "loss": 0.8497,
+      "step": 8595
+    },
+    {
+      "epoch": 0.8057338267672273,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.2119290426745254e-05,
+      "loss": 0.8725,
+      "step": 8600
+    },
+    {
+      "epoch": 0.8062022766665106,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.201681229591831e-05,
+      "loss": 0.8519,
+      "step": 8605
+    },
+    {
+      "epoch": 0.8066707265657937,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.1914542732239407e-05,
+      "loss": 0.856,
+      "step": 8610
+    },
+    {
+      "epoch": 0.807139176465077,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.1812482009229995e-05,
+      "loss": 0.8556,
+      "step": 8615
+    },
+    {
+      "epoch": 0.8076076263643603,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.1710630399852884e-05,
+      "loss": 0.8398,
+      "step": 8620
+    },
+    {
+      "epoch": 0.8080760762636436,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.160898817651169e-05,
+      "loss": 0.8692,
+      "step": 8625
+    },
+    {
+      "epoch": 0.8085445261629268,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.1507555611050013e-05,
+      "loss": 0.8461,
+      "step": 8630
+    },
+    {
+      "epoch": 0.8090129760622101,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.140633297475063e-05,
+      "loss": 0.8698,
+      "step": 8635
+    },
+    {
+      "epoch": 0.8094814259614934,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.1305320538334983e-05,
+      "loss": 0.9851,
+      "step": 8640
+    },
+    {
+      "epoch": 0.8099498758607767,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.1204518571962273e-05,
+      "loss": 0.9966,
+      "step": 8645
+    },
+    {
+      "epoch": 0.81041832576006,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.110392734522877e-05,
+      "loss": 0.8402,
+      "step": 8650
+    },
+    {
+      "epoch": 0.8108867756593432,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.1003547127167145e-05,
+      "loss": 0.8469,
+      "step": 8655
+    },
+    {
+      "epoch": 0.8113552255586265,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.0903378186245713e-05,
+      "loss": 0.858,
+      "step": 8660
+    },
+    {
+      "epoch": 0.8118236754579098,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.08034207903678e-05,
+      "loss": 0.8404,
+      "step": 8665
+    },
+    {
+      "epoch": 0.8122921253571931,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.070367520687081e-05,
+      "loss": 0.8561,
+      "step": 8670
+    },
+    {
+      "epoch": 0.8127605752564763,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.060414170252576e-05,
+      "loss": 1.0098,
+      "step": 8675
+    },
+    {
+      "epoch": 0.8132290251557596,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.050482054353643e-05,
+      "loss": 0.8824,
+      "step": 8680
+    },
+    {
+      "epoch": 0.8136974750550429,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.0405711995538725e-05,
+      "loss": 0.8625,
+      "step": 8685
+    },
+    {
+      "epoch": 0.8141659249543262,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.0306816323599763e-05,
+      "loss": 0.8539,
+      "step": 8690
+    },
+    {
+      "epoch": 0.8146343748536095,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.02081337922175e-05,
+      "loss": 0.8683,
+      "step": 8695
+    },
+    {
+      "epoch": 0.8151028247528926,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.010966466531976e-05,
+      "loss": 0.86,
+      "step": 8700
+    },
+    {
+      "epoch": 0.8155712746521759,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.0011409206263654e-05,
+      "loss": 0.8658,
+      "step": 8705
+    },
+    {
+      "epoch": 0.8160397245514592,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.991336767783476e-05,
+      "loss": 0.8606,
+      "step": 8710
+    },
+    {
+      "epoch": 0.8165081744507425,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9815540342246562e-05,
+      "loss": 0.9975,
+      "step": 8715
+    },
+    {
+      "epoch": 0.8169766243500257,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9717927461139705e-05,
+      "loss": 0.881,
+      "step": 8720
+    },
+    {
+      "epoch": 0.817445074249309,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9620529295581192e-05,
+      "loss": 0.8489,
+      "step": 8725
+    },
+    {
+      "epoch": 0.8179135241485923,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.95233461060638e-05,
+      "loss": 0.8271,
+      "step": 8730
+    },
+    {
+      "epoch": 0.8183819740478756,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.9426378152505352e-05,
+      "loss": 0.8778,
+      "step": 8735
+    },
+    {
+      "epoch": 0.8188504239471589,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9329625694248075e-05,
+      "loss": 0.8552,
+      "step": 8740
+    },
+    {
+      "epoch": 0.8193188738464421,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9233088990057734e-05,
+      "loss": 0.8524,
+      "step": 8745
+    },
+    {
+      "epoch": 0.8197873237457254,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9136768298123165e-05,
+      "loss": 0.8551,
+      "step": 8750
+    },
+    {
+      "epoch": 0.8202557736450087,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.904066387605543e-05,
+      "loss": 0.8538,
+      "step": 8755
+    },
+    {
+      "epoch": 0.820724223544292,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.8944775980887152e-05,
+      "loss": 0.8867,
+      "step": 8760
+    },
+    {
+      "epoch": 0.8211926734435752,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.884910486907191e-05,
+      "loss": 0.8367,
+      "step": 8765
+    },
+    {
+      "epoch": 0.8216611233428585,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8753650796483403e-05,
+      "loss": 0.8553,
+      "step": 8770
+    },
+    {
+      "epoch": 0.8221295732421418,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.8658414018414983e-05,
+      "loss": 0.8617,
+      "step": 8775
+    },
+    {
+      "epoch": 0.822598023141425,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8563394789578692e-05,
+      "loss": 0.8509,
+      "step": 8780
+    },
+    {
+      "epoch": 0.8230664730407083,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8468593364104856e-05,
+      "loss": 0.8399,
+      "step": 8785
+    },
+    {
+      "epoch": 0.8235349229399915,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.8374009995541274e-05,
+      "loss": 0.8335,
+      "step": 8790
+    },
+    {
+      "epoch": 0.8240033728392748,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8279644936852447e-05,
+      "loss": 0.8516,
+      "step": 8795
+    },
+    {
+      "epoch": 0.8244718227385581,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8185498440419092e-05,
+      "loss": 0.8404,
+      "step": 8800
+    },
+    {
+      "epoch": 0.8249402726378414,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.80915707580374e-05,
+      "loss": 0.8458,
+      "step": 8805
+    },
+    {
+      "epoch": 0.8254087225371246,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7997862140918265e-05,
+      "loss": 0.8718,
+      "step": 8810
+    },
+    {
+      "epoch": 0.8258771724364079,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.790437283968669e-05,
+      "loss": 0.871,
+      "step": 8815
+    },
+    {
+      "epoch": 0.8263456223356912,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7811103104381176e-05,
+      "loss": 0.8513,
+      "step": 8820
+    },
+    {
+      "epoch": 0.8268140722349745,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7718053184452975e-05,
+      "loss": 0.8573,
+      "step": 8825
+    },
+    {
+      "epoch": 0.8272825221342578,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7625223328765394e-05,
+      "loss": 0.8435,
+      "step": 8830
+    },
+    {
+      "epoch": 0.827750972033541,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.753261378559322e-05,
+      "loss": 0.8535,
+      "step": 8835
+    },
+    {
+      "epoch": 0.8282194219328243,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7440224802621995e-05,
+      "loss": 0.8452,
+      "step": 8840
+    },
+    {
+      "epoch": 0.8286878718321076,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.734805662694743e-05,
+      "loss": 0.8796,
+      "step": 8845
+    },
+    {
+      "epoch": 0.8291563217313909,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7256109505074525e-05,
+      "loss": 0.8524,
+      "step": 8850
+    },
+    {
+      "epoch": 0.829624771630674,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.7164383682917208e-05,
+      "loss": 0.8433,
+      "step": 8855
+    },
+    {
+      "epoch": 0.8300932215299573,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.7072879405797516e-05,
+      "loss": 0.8526,
+      "step": 8860
+    },
+    {
+      "epoch": 0.8305616714292406,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.6981596918444953e-05,
+      "loss": 0.8537,
+      "step": 8865
+    },
+    {
+      "epoch": 0.8310301213285239,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6890536464995788e-05,
+      "loss": 0.863,
+      "step": 8870
+    },
+    {
+      "epoch": 0.8314985712278072,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6799698288992526e-05,
+      "loss": 0.8806,
+      "step": 8875
+    },
+    {
+      "epoch": 0.8319670211270904,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.67090826333832e-05,
+      "loss": 0.8733,
+      "step": 8880
+    },
+    {
+      "epoch": 0.8324354710263737,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6618689740520642e-05,
+      "loss": 0.8297,
+      "step": 8885
+    },
+    {
+      "epoch": 0.832903920925657,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6528519852161916e-05,
+      "loss": 0.8539,
+      "step": 8890
+    },
+    {
+      "epoch": 0.8333723708249403,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.643857320946769e-05,
+      "loss": 0.8509,
+      "step": 8895
+    },
+    {
+      "epoch": 0.8338408207242235,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.6348850053001564e-05,
+      "loss": 0.8596,
+      "step": 8900
+    },
+    {
+      "epoch": 0.8343092706235068,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6259350622729362e-05,
+      "loss": 0.8681,
+      "step": 8905
+    },
+    {
+      "epoch": 0.8347777205227901,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6170075158018605e-05,
+      "loss": 0.9096,
+      "step": 8910
+    },
+    {
+      "epoch": 0.8352461704220734,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6081023897637826e-05,
+      "loss": 0.8656,
+      "step": 8915
+    },
+    {
+      "epoch": 0.8357146203213567,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.5992197079755834e-05,
+      "loss": 0.842,
+      "step": 8920
+    },
+    {
+      "epoch": 0.8361830702206399,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.590359494194128e-05,
+      "loss": 0.8458,
+      "step": 8925
+    },
+    {
+      "epoch": 0.8366515201199232,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5815217721161792e-05,
+      "loss": 0.8653,
+      "step": 8930
+    },
+    {
+      "epoch": 0.8371199700192065,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.5727065653783558e-05,
+      "loss": 0.85,
+      "step": 8935
+    },
+    {
+      "epoch": 0.8375884199184898,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.563913897557049e-05,
+      "loss": 0.8852,
+      "step": 8940
+    },
+    {
+      "epoch": 0.8380568698177729,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5551437921683766e-05,
+      "loss": 0.8552,
+      "step": 8945
+    },
+    {
+      "epoch": 0.8385253197170562,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5463962726681125e-05,
+      "loss": 0.8604,
+      "step": 8950
+    },
+    {
+      "epoch": 0.8389937696163395,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5376713624516193e-05,
+      "loss": 0.8476,
+      "step": 8955
+    },
+    {
+      "epoch": 0.8394622195156228,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.528969084853794e-05,
+      "loss": 0.8465,
+      "step": 8960
+    },
+    {
+      "epoch": 0.8399306694149061,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.5202894631490072e-05,
+      "loss": 0.8677,
+      "step": 8965
+    },
+    {
+      "epoch": 0.8403991193141893,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5116325205510262e-05,
+      "loss": 0.8704,
+      "step": 8970
+    },
+    {
+      "epoch": 0.8408675692134726,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5029982802129661e-05,
+      "loss": 0.8784,
+      "step": 8975
+    },
+    {
+      "epoch": 0.8413360191127559,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4943867652272292e-05,
+      "loss": 0.9825,
+      "step": 8980
+    },
+    {
+      "epoch": 0.8418044690120392,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4857979986254333e-05,
+      "loss": 0.8586,
+      "step": 8985
+    },
+    {
+      "epoch": 0.8422729189113224,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.4772320033783605e-05,
+      "loss": 0.8557,
+      "step": 8990
+    },
+    {
+      "epoch": 0.8427413688106057,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4686888023958822e-05,
+      "loss": 0.8477,
+      "step": 8995
+    },
+    {
+      "epoch": 0.843209818709889,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4601684185269149e-05,
+      "loss": 0.8458,
+      "step": 9000
+    },
+    {
+      "epoch": 0.8436782686091723,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.4516708745593477e-05,
+      "loss": 0.8444,
+      "step": 9005
+    },
+    {
+      "epoch": 0.8441467185084556,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4431961932199811e-05,
+      "loss": 0.8613,
+      "step": 9010
+    },
+    {
+      "epoch": 0.8446151684077388,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.434744397174469e-05,
+      "loss": 0.8436,
+      "step": 9015
+    },
+    {
+      "epoch": 0.845083618307022,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.426315509027264e-05,
+      "loss": 0.8626,
+      "step": 9020
+    },
+    {
+      "epoch": 0.8455520682063054,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.417909551321549e-05,
+      "loss": 0.8186,
+      "step": 9025
+    },
+    {
+      "epoch": 0.8460205181055886,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4095265465391727e-05,
+      "loss": 0.8445,
+      "step": 9030
+    },
+    {
+      "epoch": 0.8464889680048718,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.401166517100604e-05,
+      "loss": 0.8795,
+      "step": 9035
+    },
+    {
+      "epoch": 0.8469574179041551,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3928294853648627e-05,
+      "loss": 0.8581,
+      "step": 9040
+    },
+    {
+      "epoch": 0.8474258678034384,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3845154736294574e-05,
+      "loss": 0.8539,
+      "step": 9045
+    },
+    {
+      "epoch": 0.8478943177027217,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.376224504130329e-05,
+      "loss": 0.8707,
+      "step": 9050
+    },
+    {
+      "epoch": 0.848362767602005,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3679565990417942e-05,
+      "loss": 0.8364,
+      "step": 9055
+    },
+    {
+      "epoch": 0.8488312175012882,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.359711780476488e-05,
+      "loss": 0.8418,
+      "step": 9060
+    },
+    {
+      "epoch": 0.8492996674005715,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3514900704852895e-05,
+      "loss": 0.8515,
+      "step": 9065
+    },
+    {
+      "epoch": 0.8497681172998548,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3432914910572813e-05,
+      "loss": 0.8587,
+      "step": 9070
+    },
+    {
+      "epoch": 0.8502365671991381,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.3351160641196835e-05,
+      "loss": 0.8613,
+      "step": 9075
+    },
+    {
+      "epoch": 0.8507050170984213,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.326963811537787e-05,
+      "loss": 0.8201,
+      "step": 9080
+    },
+    {
+      "epoch": 0.8511734669977046,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3188347551149139e-05,
+      "loss": 0.8405,
+      "step": 9085
+    },
+    {
+      "epoch": 0.8516419168969879,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3107289165923353e-05,
+      "loss": 0.8412,
+      "step": 9090
+    },
+    {
+      "epoch": 0.8521103667962712,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.3026463176492376e-05,
+      "loss": 0.8518,
+      "step": 9095
+    },
+    {
+      "epoch": 0.8525788166955545,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2945869799026423e-05,
+      "loss": 0.8523,
+      "step": 9100
+    },
+    {
+      "epoch": 0.8530472665948376,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2865509249073649e-05,
+      "loss": 0.8493,
+      "step": 9105
+    },
+    {
+      "epoch": 0.853515716494121,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.278538174155951e-05,
+      "loss": 0.8818,
+      "step": 9110
+    },
+    {
+      "epoch": 0.8539841663934042,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2705487490786138e-05,
+      "loss": 0.8731,
+      "step": 9115
+    },
+    {
+      "epoch": 0.8544526162926875,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.2625826710431866e-05,
+      "loss": 0.8304,
+      "step": 9120
+    },
+    {
+      "epoch": 0.8549210661919707,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2546399613550597e-05,
+      "loss": 0.8529,
+      "step": 9125
+    },
+    {
+      "epoch": 0.855389516091254,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2467206412571208e-05,
+      "loss": 0.8409,
+      "step": 9130
+    },
+    {
+      "epoch": 0.8558579659905373,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2388247319297053e-05,
+      "loss": 0.8343,
+      "step": 9135
+    },
+    {
+      "epoch": 0.8563264158898206,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.230952254490535e-05,
+      "loss": 0.9583,
+      "step": 9140
+    },
+    {
+      "epoch": 0.8567948657891039,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2231032299946643e-05,
+      "loss": 0.8788,
+      "step": 9145
+    },
+    {
+      "epoch": 0.8572633156883871,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.2152776794344223e-05,
+      "loss": 0.832,
+      "step": 9150
+    },
+    {
+      "epoch": 0.8577317655876704,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.2074756237393514e-05,
+      "loss": 0.8818,
+      "step": 9155
+    },
+    {
+      "epoch": 0.8582002154869537,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1996970837761645e-05,
+      "loss": 0.864,
+      "step": 9160
+    },
+    {
+      "epoch": 0.858668665386237,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1919420803486791e-05,
+      "loss": 0.8819,
+      "step": 9165
+    },
+    {
+      "epoch": 0.8591371152855202,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.1842106341977588e-05,
+      "loss": 0.8678,
+      "step": 9170
+    },
+    {
+      "epoch": 0.8596055651848035,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1765027660012672e-05,
+      "loss": 0.8634,
+      "step": 9175
+    },
+    {
+      "epoch": 0.8600740150840868,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.1688184963740078e-05,
+      "loss": 0.8707,
+      "step": 9180
+    },
+    {
+      "epoch": 0.8605424649833701,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.161157845867672e-05,
+      "loss": 0.8717,
+      "step": 9185
+    },
+    {
+      "epoch": 0.8610109148826534,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1535208349707748e-05,
+      "loss": 0.8843,
+      "step": 9190
+    },
+    {
+      "epoch": 0.8614793647819365,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1459074841086136e-05,
+      "loss": 0.8742,
+      "step": 9195
+    },
+    {
+      "epoch": 0.8619478146812198,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1383178136432049e-05,
+      "loss": 0.8853,
+      "step": 9200
+    },
+    {
+      "epoch": 0.8624162645805031,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.130751843873229e-05,
+      "loss": 0.8491,
+      "step": 9205
+    },
+    {
+      "epoch": 0.8628847144797864,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1232095950339783e-05,
+      "loss": 0.9069,
+      "step": 9210
+    },
+    {
+      "epoch": 0.8633531643790696,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1156910872973059e-05,
+      "loss": 0.8875,
+      "step": 9215
+    },
+    {
+      "epoch": 0.8638216142783529,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.1081963407715711e-05,
+      "loss": 0.8397,
+      "step": 9220
+    },
+    {
+      "epoch": 0.8642900641776362,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1007253755015755e-05,
+      "loss": 0.8587,
+      "step": 9225
+    },
+    {
+      "epoch": 0.8647585140769195,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.0932782114685258e-05,
+      "loss": 0.8695,
+      "step": 9230
+    },
+    {
+      "epoch": 0.8652269639762027,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0858548685899683e-05,
+      "loss": 0.8453,
+      "step": 9235
+    },
+    {
+      "epoch": 0.865695413875486,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0784553667197373e-05,
+      "loss": 0.8721,
+      "step": 9240
+    },
+    {
+      "epoch": 0.8661638637747693,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.0710797256479089e-05,
+      "loss": 0.8571,
+      "step": 9245
+    },
+    {
+      "epoch": 0.8666323136740526,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.063727965100737e-05,
+      "loss": 0.8583,
+      "step": 9250
+    },
+    {
+      "epoch": 0.8671007635733359,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.0564001047406135e-05,
+      "loss": 0.8438,
+      "step": 9255
+    },
+    {
+      "epoch": 0.8675692134726191,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0490961641660036e-05,
+      "loss": 0.8507,
+      "step": 9260
+    },
+    {
+      "epoch": 0.8680376633719024,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0418161629113987e-05,
+      "loss": 0.8789,
+      "step": 9265
+    },
+    {
+      "epoch": 0.8685061132711857,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.0345601204472699e-05,
+      "loss": 0.882,
+      "step": 9270
+    },
+    {
+      "epoch": 0.868974563170469,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0273280561800058e-05,
+      "loss": 0.8675,
+      "step": 9275
+    },
+    {
+      "epoch": 0.8694430130697521,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.0201199894518621e-05,
+      "loss": 0.8368,
+      "step": 9280
+    },
+    {
+      "epoch": 0.8699114629690354,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.01293593954092e-05,
+      "loss": 0.8722,
+      "step": 9285
+    },
+    {
+      "epoch": 0.8703799128683187,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.0057759256610222e-05,
+      "loss": 0.8477,
+      "step": 9290
+    },
+    {
+      "epoch": 0.870848362767602,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.986399669617252e-06,
+      "loss": 0.8723,
+      "step": 9295
+    },
+    {
+      "epoch": 0.8713168126668853,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.915280825282547e-06,
+      "loss": 0.8331,
+      "step": 9300
+    },
+    {
+      "epoch": 0.8717852625661685,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.844402913814477e-06,
+      "loss": 0.876,
+      "step": 9305
+    },
+    {
+      "epoch": 0.8722537124654518,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.773766124777028e-06,
+      "loss": 0.8313,
+      "step": 9310
+    },
+    {
+      "epoch": 0.8727221623647351,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.703370647089282e-06,
+      "loss": 0.8557,
+      "step": 9315
+    },
+    {
+      "epoch": 0.8731906122640184,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.633216669024958e-06,
+      "loss": 0.8799,
+      "step": 9320
+    },
+    {
+      "epoch": 0.8736590621633016,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.56330437821189e-06,
+      "loss": 0.8421,
+      "step": 9325
+    },
+    {
+      "epoch": 0.8741275120625849,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.493633961631476e-06,
+      "loss": 0.8514,
+      "step": 9330
+    },
+    {
+      "epoch": 0.8745959619618682,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.424205605618219e-06,
+      "loss": 0.8492,
+      "step": 9335
+    },
+    {
+      "epoch": 0.8750644118611515,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.355019495859286e-06,
+      "loss": 0.865,
+      "step": 9340
+    },
+    {
+      "epoch": 0.8755328617604348,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.286075817393901e-06,
+      "loss": 0.8825,
+      "step": 9345
+    },
+    {
+      "epoch": 0.876001311659718,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.217374754612906e-06,
+      "loss": 0.8685,
+      "step": 9350
+    },
+    {
+      "epoch": 0.8764697615590012,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.148916491258286e-06,
+      "loss": 0.8457,
+      "step": 9355
+    },
+    {
+      "epoch": 0.8769382114582845,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.08070121042265e-06,
+      "loss": 0.8657,
+      "step": 9360
+    },
+    {
+      "epoch": 0.8774066613575678,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.012729094548733e-06,
+      "loss": 0.9798,
+      "step": 9365
+    },
+    {
+      "epoch": 0.877875111256851,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.945000325428898e-06,
+      "loss": 0.8415,
+      "step": 9370
+    },
+    {
+      "epoch": 0.8783435611561343,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.877515084204735e-06,
+      "loss": 0.8643,
+      "step": 9375
+    },
+    {
+      "epoch": 0.8788120110554176,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.810273551366499e-06,
+      "loss": 0.8601,
+      "step": 9380
+    },
+    {
+      "epoch": 0.8792804609547009,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.74327590675259e-06,
+      "loss": 0.8531,
+      "step": 9385
+    },
+    {
+      "epoch": 0.8797489108539842,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.676522329549186e-06,
+      "loss": 0.8621,
+      "step": 9390
+    },
+    {
+      "epoch": 0.8802173607532674,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.610012998289696e-06,
+      "loss": 0.8623,
+      "step": 9395
+    },
+    {
+      "epoch": 0.8806858106525507,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.543748090854242e-06,
+      "loss": 0.8502,
+      "step": 9400
+    },
+    {
+      "epoch": 0.881154260551834,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.477727784469302e-06,
+      "loss": 0.8303,
+      "step": 9405
+    },
+    {
+      "epoch": 0.8816227104511173,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.411952255707079e-06,
+      "loss": 0.8424,
+      "step": 9410
+    },
+    {
+      "epoch": 0.8820911603504005,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.346421680485217e-06,
+      "loss": 0.8754,
+      "step": 9415
+    },
+    {
+      "epoch": 0.8825596102496838,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.281136234066133e-06,
+      "loss": 0.877,
+      "step": 9420
+    },
+    {
+      "epoch": 0.8830280601489671,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.21609609105669e-06,
+      "loss": 0.8639,
+      "step": 9425
+    },
+    {
+      "epoch": 0.8834965100482504,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.151301425407699e-06,
+      "loss": 1.0118,
+      "step": 9430
+    },
+    {
+      "epoch": 0.8839649599475337,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.086752410413412e-06,
+      "loss": 0.846,
+      "step": 9435
+    },
+    {
+      "epoch": 0.8844334098468168,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.022449218711082e-06,
+      "loss": 0.9843,
+      "step": 9440
+    },
+    {
+      "epoch": 0.8849018597461001,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.958392022280525e-06,
+      "loss": 0.8849,
+      "step": 9445
+    },
+    {
+      "epoch": 0.8853703096453834,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.8945809924436e-06,
+      "loss": 0.8647,
+      "step": 9450
+    },
+    {
+      "epoch": 0.8858387595446667,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.831016299863847e-06,
+      "loss": 0.8792,
+      "step": 9455
+    },
+    {
+      "epoch": 0.8863072094439499,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.767698114545919e-06,
+      "loss": 0.8314,
+      "step": 9460
+    },
+    {
+      "epoch": 0.8867756593432332,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.704626605835218e-06,
+      "loss": 0.8854,
+      "step": 9465
+    },
+    {
+      "epoch": 0.8872441092425165,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.641801942417415e-06,
+      "loss": 0.8566,
+      "step": 9470
+    },
+    {
+      "epoch": 0.8877125591417998,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.579224292317921e-06,
+      "loss": 0.8811,
+      "step": 9475
+    },
+    {
+      "epoch": 0.8881810090410831,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.5168938229015825e-06,
+      "loss": 0.8842,
+      "step": 9480
+    },
+    {
+      "epoch": 0.8886494589403663,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.4548107008721395e-06,
+      "loss": 0.8628,
+      "step": 9485
+    },
+    {
+      "epoch": 0.8891179088396496,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.392975092271781e-06,
+      "loss": 0.8632,
+      "step": 9490
+    },
+    {
+      "epoch": 0.8895863587389329,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.331387162480708e-06,
+      "loss": 0.8568,
+      "step": 9495
+    },
+    {
+      "epoch": 0.8900548086382162,
+      "grad_norm": 0.625,
+      "learning_rate": 7.2700470762167125e-06,
+      "loss": 0.8797,
+      "step": 9500
+    },
+    {
+      "epoch": 0.8905232585374994,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.208954997534767e-06,
+      "loss": 0.8497,
+      "step": 9505
+    },
+    {
+      "epoch": 0.8909917084367827,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.1481110898264705e-06,
+      "loss": 0.8354,
+      "step": 9510
+    },
+    {
+      "epoch": 0.891460158336066,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.087515515819742e-06,
+      "loss": 0.8595,
+      "step": 9515
+    },
+    {
+      "epoch": 0.8919286082353493,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.0271684375783134e-06,
+      "loss": 0.8524,
+      "step": 9520
+    },
+    {
+      "epoch": 0.8923970581346325,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.967070016501298e-06,
+      "loss": 0.8456,
+      "step": 9525
+    },
+    {
+      "epoch": 0.8928655080339157,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.907220413322768e-06,
+      "loss": 0.8684,
+      "step": 9530
+    },
+    {
+      "epoch": 0.893333957933199,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.847619788111348e-06,
+      "loss": 0.8543,
+      "step": 9535
+    },
+    {
+      "epoch": 0.8938024078324823,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.788268300269785e-06,
+      "loss": 0.8336,
+      "step": 9540
+    },
+    {
+      "epoch": 0.8942708577317656,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.729166108534435e-06,
+      "loss": 0.8674,
+      "step": 9545
+    },
+    {
+      "epoch": 0.8947393076310488,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.6703133709749895e-06,
+      "loss": 0.8718,
+      "step": 9550
+    },
+    {
+      "epoch": 0.8952077575303321,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.61171024499393e-06,
+      "loss": 0.8599,
+      "step": 9555
+    },
+    {
+      "epoch": 0.8956762074296154,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.553356887326135e-06,
+      "loss": 0.8693,
+      "step": 9560
+    },
+    {
+      "epoch": 0.8961446573288987,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.495253454038508e-06,
+      "loss": 0.8403,
+      "step": 9565
+    },
+    {
+      "epoch": 0.896613107228182,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.4374001005295e-06,
+      "loss": 0.8618,
+      "step": 9570
+    },
+    {
+      "epoch": 0.8970815571274652,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.379796981528752e-06,
+      "loss": 0.8476,
+      "step": 9575
+    },
+    {
+      "epoch": 0.8975500070267485,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.3224442510965934e-06,
+      "loss": 0.8465,
+      "step": 9580
+    },
+    {
+      "epoch": 0.8980184569260318,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.265342062623736e-06,
+      "loss": 0.8695,
+      "step": 9585
+    },
+    {
+      "epoch": 0.8984869068253151,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.208490568830805e-06,
+      "loss": 0.8419,
+      "step": 9590
+    },
+    {
+      "epoch": 0.8989553567245983,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.151889921767939e-06,
+      "loss": 0.8749,
+      "step": 9595
+    },
+    {
+      "epoch": 0.8994238066238815,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.095540272814337e-06,
+      "loss": 0.8456,
+      "step": 9600
+    },
+    {
+      "epoch": 0.8998922565231648,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.039441772677989e-06,
+      "loss": 0.8403,
+      "step": 9605
+    },
+    {
+      "epoch": 0.9003607064224481,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.9835945713951016e-06,
+      "loss": 0.8151,
+      "step": 9610
+    },
+    {
+      "epoch": 0.9008291563217314,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.927998818329827e-06,
+      "loss": 0.8659,
+      "step": 9615
+    },
+    {
+      "epoch": 0.9012976062210146,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.872654662173782e-06,
+      "loss": 0.8467,
+      "step": 9620
+    },
+    {
+      "epoch": 0.9017660561202979,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.817562250945707e-06,
+      "loss": 0.8416,
+      "step": 9625
+    },
+    {
+      "epoch": 0.9022345060195812,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.762721731991061e-06,
+      "loss": 0.9013,
+      "step": 9630
+    },
+    {
+      "epoch": 0.9027029559188645,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.708133251981573e-06,
+      "loss": 0.8782,
+      "step": 9635
+    },
+    {
+      "epoch": 0.9031714058181477,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.653796956914914e-06,
+      "loss": 0.8547,
+      "step": 9640
+    },
+    {
+      "epoch": 0.903639855717431,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.599712992114303e-06,
+      "loss": 0.841,
+      "step": 9645
+    },
+    {
+      "epoch": 0.9041083056167143,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.5458815022280566e-06,
+      "loss": 0.8368,
+      "step": 9650
+    },
+    {
+      "epoch": 0.9045767555159976,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.492302631229251e-06,
+      "loss": 0.8432,
+      "step": 9655
+    },
+    {
+      "epoch": 0.9050452054152809,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.438976522415351e-06,
+      "loss": 0.8256,
+      "step": 9660
+    },
+    {
+      "epoch": 0.9055136553145641,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.385903318407814e-06,
+      "loss": 0.8407,
+      "step": 9665
+    },
+    {
+      "epoch": 0.9059821052138474,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.333083161151653e-06,
+      "loss": 0.8563,
+      "step": 9670
+    },
+    {
+      "epoch": 0.9064505551131307,
+      "grad_norm": 0.77734375,
+      "learning_rate": 5.280516191915142e-06,
+      "loss": 0.8367,
+      "step": 9675
+    },
+    {
+      "epoch": 0.906919005012414,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.228202551289407e-06,
+      "loss": 0.8634,
+      "step": 9680
+    },
+    {
+      "epoch": 0.9073874549116971,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.176142379188009e-06,
+      "loss": 0.8403,
+      "step": 9685
+    },
+    {
+      "epoch": 0.9078559048109804,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.124335814846614e-06,
+      "loss": 0.8543,
+      "step": 9690
+    },
+    {
+      "epoch": 0.9083243547102637,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.072782996822634e-06,
+      "loss": 0.8842,
+      "step": 9695
+    },
+    {
+      "epoch": 0.908792804609547,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.021484062994819e-06,
+      "loss": 0.8307,
+      "step": 9700
+    },
+    {
+      "epoch": 0.9092612545088303,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.970439150562867e-06,
+      "loss": 0.8947,
+      "step": 9705
+    },
+    {
+      "epoch": 0.9097297044081135,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.91964839604716e-06,
+      "loss": 0.8414,
+      "step": 9710
+    },
+    {
+      "epoch": 0.9101981543073968,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.869111935288273e-06,
+      "loss": 0.8414,
+      "step": 9715
+    },
+    {
+      "epoch": 0.9106666042066801,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.818829903446731e-06,
+      "loss": 0.8286,
+      "step": 9720
+    },
+    {
+      "epoch": 0.9111350541059634,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.768802435002506e-06,
+      "loss": 0.8573,
+      "step": 9725
+    },
+    {
+      "epoch": 0.9116035040052466,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.7190296637547705e-06,
+      "loss": 0.845,
+      "step": 9730
+    },
+    {
+      "epoch": 0.9120719539045299,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.6695117228215405e-06,
+      "loss": 0.8924,
+      "step": 9735
+    },
+    {
+      "epoch": 0.9125404038038132,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.620248744639233e-06,
+      "loss": 0.8385,
+      "step": 9740
+    },
+    {
+      "epoch": 0.9130088537030965,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.5712408609623895e-06,
+      "loss": 0.8617,
+      "step": 9745
+    },
+    {
+      "epoch": 0.9134773036023798,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.522488202863307e-06,
+      "loss": 0.8635,
+      "step": 9750
+    },
+    {
+      "epoch": 0.913945753501663,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.4739909007316614e-06,
+      "loss": 0.8691,
+      "step": 9755
+    },
+    {
+      "epoch": 0.9144142034009463,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.425749084274166e-06,
+      "loss": 0.8787,
+      "step": 9760
+    },
+    {
+      "epoch": 0.9148826533002296,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.37776288251428e-06,
+      "loss": 0.8434,
+      "step": 9765
+    },
+    {
+      "epoch": 0.9153511031995128,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.33003242379173e-06,
+      "loss": 0.865,
+      "step": 9770
+    },
+    {
+      "epoch": 0.915819553098796,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.2825578357623685e-06,
+      "loss": 0.8321,
+      "step": 9775
+    },
+    {
+      "epoch": 0.9162880029980793,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.235339245397607e-06,
+      "loss": 0.8723,
+      "step": 9780
+    },
+    {
+      "epoch": 0.9167564528973626,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.188376778984271e-06,
+      "loss": 0.8666,
+      "step": 9785
+    },
+    {
+      "epoch": 0.9172249027966459,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.141670562124145e-06,
+      "loss": 0.8793,
+      "step": 9790
+    },
+    {
+      "epoch": 0.9176933526959292,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.0952207197336505e-06,
+      "loss": 0.8508,
+      "step": 9795
+    },
+    {
+      "epoch": 0.9181618025952124,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.049027376043568e-06,
+      "loss": 0.8862,
+      "step": 9800
+    },
+    {
+      "epoch": 0.9186302524944957,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.003090654598663e-06,
+      "loss": 0.8892,
+      "step": 9805
+    },
+    {
+      "epoch": 0.919098702393779,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.957410678257356e-06,
+      "loss": 0.8437,
+      "step": 9810
+    },
+    {
+      "epoch": 0.9195671522930623,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.911987569191367e-06,
+      "loss": 0.8285,
+      "step": 9815
+    },
+    {
+      "epoch": 0.9200356021923455,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.86682144888546e-06,
+      "loss": 0.8639,
+      "step": 9820
+    },
+    {
+      "epoch": 0.9205040520916288,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.821912438137087e-06,
+      "loss": 0.8581,
+      "step": 9825
+    },
+    {
+      "epoch": 0.9209725019909121,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.7772606570560144e-06,
+      "loss": 0.8247,
+      "step": 9830
+    },
+    {
+      "epoch": 0.9214409518901954,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.7328662250640802e-06,
+      "loss": 0.8383,
+      "step": 9835
+    },
+    {
+      "epoch": 0.9219094017894787,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.688729260894841e-06,
+      "loss": 0.8371,
+      "step": 9840
+    },
+    {
+      "epoch": 0.9223778516887619,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.6448498825932044e-06,
+      "loss": 0.8343,
+      "step": 9845
+    },
+    {
+      "epoch": 0.9228463015880451,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.601228207515206e-06,
+      "loss": 0.8403,
+      "step": 9850
+    },
+    {
+      "epoch": 0.9233147514873284,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.5578643523276443e-06,
+      "loss": 0.8676,
+      "step": 9855
+    },
+    {
+      "epoch": 0.9237832013866117,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.51475843300777e-06,
+      "loss": 0.8966,
+      "step": 9860
+    },
+    {
+      "epoch": 0.9242516512858949,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.4719105648429726e-06,
+      "loss": 0.8625,
+      "step": 9865
+    },
+    {
+      "epoch": 0.9247201011851782,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.429320862430485e-06,
+      "loss": 0.8733,
+      "step": 9870
+    },
+    {
+      "epoch": 0.9251885510844615,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.3869894396770794e-06,
+      "loss": 0.8656,
+      "step": 9875
+    },
+    {
+      "epoch": 0.9256570009837448,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.3449164097987593e-06,
+      "loss": 0.8613,
+      "step": 9880
+    },
+    {
+      "epoch": 0.9261254508830281,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.3031018853204255e-06,
+      "loss": 0.8558,
+      "step": 9885
+    },
+    {
+      "epoch": 0.9265939007823113,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.261545978075631e-06,
+      "loss": 0.8414,
+      "step": 9890
+    },
+    {
+      "epoch": 0.9270623506815946,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.220248799206227e-06,
+      "loss": 0.8681,
+      "step": 9895
+    },
+    {
+      "epoch": 0.9275308005808779,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.1792104591621296e-06,
+      "loss": 0.8521,
+      "step": 9900
+    },
+    {
+      "epoch": 0.9279992504801612,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.1384310677009533e-06,
+      "loss": 0.8701,
+      "step": 9905
+    },
+    {
+      "epoch": 0.9284677003794444,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.0979107338877544e-06,
+      "loss": 0.8667,
+      "step": 9910
+    },
+    {
+      "epoch": 0.9289361502787277,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.0576495660947667e-06,
+      "loss": 0.8552,
+      "step": 9915
+    },
+    {
+      "epoch": 0.929404600178011,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.017647672001045e-06,
+      "loss": 0.8674,
+      "step": 9920
+    },
+    {
+      "epoch": 0.9298730500772943,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.97790515859222e-06,
+      "loss": 0.8786,
+      "step": 9925
+    },
+    {
+      "epoch": 0.9303414999765776,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.9384221321602124e-06,
+      "loss": 0.8628,
+      "step": 9930
+    },
+    {
+      "epoch": 0.9308099498758607,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.8991986983029518e-06,
+      "loss": 0.8541,
+      "step": 9935
+    },
+    {
+      "epoch": 0.931278399775144,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.8602349619240577e-06,
+      "loss": 0.8647,
+      "step": 9940
+    },
+    {
+      "epoch": 0.9317468496744273,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.8215310272325936e-06,
+      "loss": 0.883,
+      "step": 9945
+    },
+    {
+      "epoch": 0.9322152995737106,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.783086997742812e-06,
+      "loss": 0.8388,
+      "step": 9950
+    },
+    {
+      "epoch": 0.9326837494729938,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.744902976273789e-06,
+      "loss": 0.8612,
+      "step": 9955
+    },
+    {
+      "epoch": 0.9331521993722771,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.7069790649492554e-06,
+      "loss": 0.859,
+      "step": 9960
+    },
+    {
+      "epoch": 0.9336206492715604,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.6693153651972445e-06,
+      "loss": 0.8376,
+      "step": 9965
+    },
+    {
+      "epoch": 0.9340890991708437,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.631911977749868e-06,
+      "loss": 0.879,
+      "step": 9970
+    },
+    {
+      "epoch": 0.934557549070127,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.594769002642994e-06,
+      "loss": 0.8423,
+      "step": 9975
+    },
+    {
+      "epoch": 0.9350259989694102,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.5578865392160587e-06,
+      "loss": 0.8501,
+      "step": 9980
+    },
+    {
+      "epoch": 0.9354944488686935,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.521264686111746e-06,
+      "loss": 0.8467,
+      "step": 9985
+    },
+    {
+      "epoch": 0.9359628987679768,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.484903541275696e-06,
+      "loss": 0.8587,
+      "step": 9990
+    },
+    {
+      "epoch": 0.9364313486672601,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.4488032019563402e-06,
+      "loss": 0.8517,
+      "step": 9995
+    },
+    {
+      "epoch": 0.9368997985665433,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.4129637647045343e-06,
+      "loss": 0.8644,
+      "step": 10000
+    },
+    {
+      "epoch": 0.9373682484658266,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.37738532537336e-06,
+      "loss": 0.8182,
+      "step": 10005
+    },
+    {
+      "epoch": 0.9378366983651099,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.342067979117868e-06,
+      "loss": 0.8404,
+      "step": 10010
+    },
+    {
+      "epoch": 0.9383051482643932,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.307011820394811e-06,
+      "loss": 0.8284,
+      "step": 10015
+    },
+    {
+      "epoch": 0.9387735981636764,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.2722169429623796e-06,
+      "loss": 0.842,
+      "step": 10020
+    },
+    {
+      "epoch": 0.9392420480629596,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.237683439879956e-06,
+      "loss": 0.8749,
+      "step": 10025
+    },
+    {
+      "epoch": 0.9397104979622429,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.2034114035079156e-06,
+      "loss": 0.8509,
+      "step": 10030
+    },
+    {
+      "epoch": 0.9401789478615262,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.169400925507281e-06,
+      "loss": 0.8514,
+      "step": 10035
+    },
+    {
+      "epoch": 0.9406473977608095,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.13565209683958e-06,
+      "loss": 0.8871,
+      "step": 10040
+    },
+    {
+      "epoch": 0.9411158476600927,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.1021650077665324e-06,
+      "loss": 0.8399,
+      "step": 10045
+    },
+    {
+      "epoch": 0.941584297559376,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.0689397478498295e-06,
+      "loss": 0.8421,
+      "step": 10050
+    },
+    {
+      "epoch": 0.9420527474586593,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.035976405950901e-06,
+      "loss": 0.8468,
+      "step": 10055
+    },
+    {
+      "epoch": 0.9425211973579426,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.0032750702306923e-06,
+      "loss": 0.8717,
+      "step": 10060
+    },
+    {
+      "epoch": 0.9429896472572259,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9708358281493753e-06,
+      "loss": 0.8816,
+      "step": 10065
+    },
+    {
+      "epoch": 0.9434580971565091,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9386587664661615e-06,
+      "loss": 0.8588,
+      "step": 10070
+    },
+    {
+      "epoch": 0.9439265470557924,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9067439712390778e-06,
+      "loss": 0.8395,
+      "step": 10075
+    },
+    {
+      "epoch": 0.9443949969550757,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.8750915278246905e-06,
+      "loss": 0.9044,
+      "step": 10080
+    },
+    {
+      "epoch": 0.944863446854359,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8437015208779274e-06,
+      "loss": 0.8487,
+      "step": 10085
+    },
+    {
+      "epoch": 0.9453318967536422,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.8125740343517773e-06,
+      "loss": 0.8761,
+      "step": 10090
+    },
+    {
+      "epoch": 0.9458003466529254,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.7817091514971684e-06,
+      "loss": 0.8454,
+      "step": 10095
+    },
+    {
+      "epoch": 0.9462687965522087,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.751106954862669e-06,
+      "loss": 0.8976,
+      "step": 10100
+    },
+    {
+      "epoch": 0.946737246451492,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.7207675262942868e-06,
+      "loss": 0.8612,
+      "step": 10105
+    },
+    {
+      "epoch": 0.9472056963507753,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.6906909469352584e-06,
+      "loss": 0.8359,
+      "step": 10110
+    },
+    {
+      "epoch": 0.9476741462500585,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6608772972258268e-06,
+      "loss": 0.8584,
+      "step": 10115
+    },
+    {
+      "epoch": 0.9481425961493418,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6313266569030094e-06,
+      "loss": 0.8528,
+      "step": 10120
+    },
+    {
+      "epoch": 0.9486110460486251,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.6020391050004413e-06,
+      "loss": 0.8418,
+      "step": 10125
+    },
+    {
+      "epoch": 0.9490794959479084,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.573014719848087e-06,
+      "loss": 0.8983,
+      "step": 10130
+    },
+    {
+      "epoch": 0.9495479458471916,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.5442535790720635e-06,
+      "loss": 0.8467,
+      "step": 10135
+    },
+    {
+      "epoch": 0.9500163957464749,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.5157557595944616e-06,
+      "loss": 0.8263,
+      "step": 10140
+    },
+    {
+      "epoch": 0.9504848456457582,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4875213376331132e-06,
+      "loss": 0.9152,
+      "step": 10145
+    },
+    {
+      "epoch": 0.9509532955450415,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4595503887013695e-06,
+      "loss": 0.8345,
+      "step": 10150
+    },
+    {
+      "epoch": 0.9514217454443248,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4318429876079454e-06,
+      "loss": 0.8724,
+      "step": 10155
+    },
+    {
+      "epoch": 0.951890195343608,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4043992084566748e-06,
+      "loss": 0.8628,
+      "step": 10160
+    },
+    {
+      "epoch": 0.9523586452428913,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.377219124646356e-06,
+      "loss": 0.8688,
+      "step": 10165
+    },
+    {
+      "epoch": 0.9528270951421746,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.350302808870474e-06,
+      "loss": 0.878,
+      "step": 10170
+    },
+    {
+      "epoch": 0.9532955450414579,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3236503331171102e-06,
+      "loss": 0.8659,
+      "step": 10175
+    },
+    {
+      "epoch": 0.953763994940741,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2972617686687116e-06,
+      "loss": 0.8511,
+      "step": 10180
+    },
+    {
+      "epoch": 0.9542324448400243,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2711371861018451e-06,
+      "loss": 0.8434,
+      "step": 10185
+    },
+    {
+      "epoch": 0.9547008947393076,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2452766552870977e-06,
+      "loss": 0.85,
+      "step": 10190
+    },
+    {
+      "epoch": 0.9551693446385909,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.2196802453888211e-06,
+      "loss": 0.8661,
+      "step": 10195
+    },
+    {
+      "epoch": 0.9556377945378741,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.1943480248649884e-06,
+      "loss": 0.8296,
+      "step": 10200
+    },
+    {
+      "epoch": 0.9561062444371574,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1692800614669707e-06,
+      "loss": 0.8898,
+      "step": 10205
+    },
+    {
+      "epoch": 0.9565746943364407,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.1444764222393934e-06,
+      "loss": 0.8616,
+      "step": 10210
+    },
+    {
+      "epoch": 0.957043144235724,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.119937173519947e-06,
+      "loss": 0.8209,
+      "step": 10215
+    },
+    {
+      "epoch": 0.9575115941350073,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.0956623809392107e-06,
+      "loss": 0.8521,
+      "step": 10220
+    },
+    {
+      "epoch": 0.9579800440342905,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0716521094204513e-06,
+      "loss": 0.8661,
+      "step": 10225
+    },
+    {
+      "epoch": 0.9584484939335738,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.04790642317949e-06,
+      "loss": 0.8529,
+      "step": 10230
+    },
+    {
+      "epoch": 0.9589169438328571,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0244253857245034e-06,
+      "loss": 0.8605,
+      "step": 10235
+    },
+    {
+      "epoch": 0.9593853937321404,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0012090598558677e-06,
+      "loss": 0.8326,
+      "step": 10240
+    },
+    {
+      "epoch": 0.9598538436314236,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.782575076659584e-07,
+      "loss": 0.8343,
+      "step": 10245
+    },
+    {
+      "epoch": 0.9603222935307069,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.555707905390398e-07,
+      "loss": 0.8287,
+      "step": 10250
+    },
+    {
+      "epoch": 0.9607907434299902,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.331489691510765e-07,
+      "loss": 0.8548,
+      "step": 10255
+    },
+    {
+      "epoch": 0.9612591933292735,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.109921034695213e-07,
+      "loss": 0.8439,
+      "step": 10260
+    },
+    {
+      "epoch": 0.9617276432285568,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.891002527532388e-07,
+      "loss": 0.9948,
+      "step": 10265
+    },
+    {
+      "epoch": 0.9621960931278399,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.674734755522939e-07,
+      "loss": 0.8555,
+      "step": 10270
+    },
+    {
+      "epoch": 0.9626645430271232,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.461118297078296e-07,
+      "loss": 0.85,
+      "step": 10275
+    },
+    {
+      "epoch": 0.9631329929264065,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.25015372351845e-07,
+      "loss": 0.8547,
+      "step": 10280
+    },
+    {
+      "epoch": 0.9636014428256898,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.041841599071287e-07,
+      "loss": 0.8644,
+      "step": 10285
+    },
+    {
+      "epoch": 0.964069892724973,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.836182480870701e-07,
+      "loss": 0.8666,
+      "step": 10290
+    },
+    {
+      "epoch": 0.9645383426242563,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.633176918954599e-07,
+      "loss": 0.8699,
+      "step": 10295
+    },
+    {
+      "epoch": 0.9650067925235396,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.432825456264559e-07,
+      "loss": 0.8444,
+      "step": 10300
+    },
+    {
+      "epoch": 0.9654752424228229,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.235128628643284e-07,
+      "loss": 0.8359,
+      "step": 10305
+    },
+    {
+      "epoch": 0.9659436923221062,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.040086964833825e-07,
+      "loss": 0.8682,
+      "step": 10310
+    },
+    {
+      "epoch": 0.9664121422213894,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.8477009864778e-07,
+      "loss": 0.877,
+      "step": 10315
+    },
+    {
+      "epoch": 0.9668805921206727,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.657971208114288e-07,
+      "loss": 0.878,
+      "step": 10320
+    },
+    {
+      "epoch": 0.967349042019956,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.470898137178494e-07,
+      "loss": 0.8769,
+      "step": 10325
+    },
+    {
+      "epoch": 0.9678174919192393,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.286482273999639e-07,
+      "loss": 0.847,
+      "step": 10330
+    },
+    {
+      "epoch": 0.9682859418185225,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.104724111800852e-07,
+      "loss": 0.8791,
+      "step": 10335
+    },
+    {
+      "epoch": 0.9687543917178058,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.925624136696839e-07,
+      "loss": 0.986,
+      "step": 10340
+    },
+    {
+      "epoch": 0.969222841617089,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.749182827693101e-07,
+      "loss": 0.8673,
+      "step": 10345
+    },
+    {
+      "epoch": 0.9696912915163723,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.575400656684271e-07,
+      "loss": 0.8658,
+      "step": 10350
+    },
+    {
+      "epoch": 0.9701597414156556,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.404278088453451e-07,
+      "loss": 0.8949,
+      "step": 10355
+    },
+    {
+      "epoch": 0.9706281913149388,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.235815580670322e-07,
+      "loss": 0.8512,
+      "step": 10360
+    },
+    {
+      "epoch": 0.9710966412142221,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.070013583890254e-07,
+      "loss": 0.852,
+      "step": 10365
+    },
+    {
+      "epoch": 0.9715650911135054,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.906872541552976e-07,
+      "loss": 0.8552,
+      "step": 10370
+    },
+    {
+      "epoch": 0.9720335410127887,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.746392889981577e-07,
+      "loss": 0.8644,
+      "step": 10375
+    },
+    {
+      "epoch": 0.9725019909120719,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.588575058381284e-07,
+      "loss": 0.8886,
+      "step": 10380
+    },
+    {
+      "epoch": 0.9729704408113552,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.433419468837907e-07,
+      "loss": 0.8543,
+      "step": 10385
+    },
+    {
+      "epoch": 0.9734388907106385,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.280926536317509e-07,
+      "loss": 0.8447,
+      "step": 10390
+    },
+    {
+      "epoch": 0.9739073406099218,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.131096668664625e-07,
+      "loss": 0.8641,
+      "step": 10395
+    },
+    {
+      "epoch": 0.9743757905092051,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.9839302666011546e-07,
+      "loss": 0.8413,
+      "step": 10400
+    },
+    {
+      "epoch": 0.9748442404084883,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.8394277237259193e-07,
+      "loss": 0.8857,
+      "step": 10405
+    },
+    {
+      "epoch": 0.9753126903077716,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.697589426512993e-07,
+      "loss": 0.8598,
+      "step": 10410
+    },
+    {
+      "epoch": 0.9757811402070549,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.5584157543109285e-07,
+      "loss": 0.8419,
+      "step": 10415
+    },
+    {
+      "epoch": 0.9762495901063382,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.421907079341646e-07,
+      "loss": 0.8247,
+      "step": 10420
+    },
+    {
+      "epoch": 0.9767180400056213,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.288063766699545e-07,
+      "loss": 0.8495,
+      "step": 10425
+    },
+    {
+      "epoch": 0.9771864899049046,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.156886174350615e-07,
+      "loss": 0.8556,
+      "step": 10430
+    },
+    {
+      "epoch": 0.9776549398041879,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.028374653130994e-07,
+      "loss": 0.8362,
+      "step": 10435
+    },
+    {
+      "epoch": 0.9781233897034712,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.902529546746635e-07,
+      "loss": 0.8553,
+      "step": 10440
+    },
+    {
+      "epoch": 0.9785918396027545,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.7793511917723057e-07,
+      "loss": 0.8601,
+      "step": 10445
+    },
+    {
+      "epoch": 0.9790602895020377,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.658839917649925e-07,
+      "loss": 0.8611,
+      "step": 10450
+    },
+    {
+      "epoch": 0.979528739401321,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.5409960466887817e-07,
+      "loss": 0.8485,
+      "step": 10455
+    },
+    {
+      "epoch": 0.9799971893006043,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.4258198940639853e-07,
+      "loss": 0.8384,
+      "step": 10460
+    },
+    {
+      "epoch": 0.9804656391998876,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.3133117678157956e-07,
+      "loss": 0.8731,
+      "step": 10465
+    },
+    {
+      "epoch": 0.9809340890991708,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.203471968848847e-07,
+      "loss": 0.8821,
+      "step": 10470
+    },
+    {
+      "epoch": 0.9814025389984541,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.0963007909311494e-07,
+      "loss": 0.8541,
+      "step": 10475
+    },
+    {
+      "epoch": 0.9818709888977374,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.991798520693644e-07,
+      "loss": 0.8547,
+      "step": 10480
+    },
+    {
+      "epoch": 0.9823394387970207,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.889965437629093e-07,
+      "loss": 0.8453,
+      "step": 10485
+    },
+    {
+      "epoch": 0.982807888696304,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.790801814091414e-07,
+      "loss": 0.8649,
+      "step": 10490
+    },
+    {
+      "epoch": 0.9832763385955872,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.6943079152953457e-07,
+      "loss": 0.8464,
+      "step": 10495
+    },
+    {
+      "epoch": 0.9837447884948705,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6004839993151167e-07,
+      "loss": 0.8536,
+      "step": 10500
+    },
+    {
+      "epoch": 0.9842132383941538,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.5093303170841121e-07,
+      "loss": 0.8611,
+      "step": 10505
+    },
+    {
+      "epoch": 0.984681688293437,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.420847112394097e-07,
+      "loss": 0.8606,
+      "step": 10510
+    },
+    {
+      "epoch": 0.9851501381927202,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.3350346218947707e-07,
+      "loss": 0.8608,
+      "step": 10515
+    },
+    {
+      "epoch": 0.9856185880920035,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2518930750928803e-07,
+      "loss": 0.8629,
+      "step": 10520
+    },
+    {
+      "epoch": 0.9860870379912868,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1714226943515538e-07,
+      "loss": 0.8474,
+      "step": 10525
+    },
+    {
+      "epoch": 0.9865554878905701,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.093623694890078e-07,
+      "loss": 0.8758,
+      "step": 10530
+    },
+    {
+      "epoch": 0.9870239377898534,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0184962847828994e-07,
+      "loss": 0.8533,
+      "step": 10535
+    },
+    {
+      "epoch": 0.9874923876891366,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.460406649594022e-08,
+      "loss": 0.8661,
+      "step": 10540
+    },
+    {
+      "epoch": 0.9879608375884199,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.762570292032423e-08,
+      "loss": 0.883,
+      "step": 10545
+    },
+    {
+      "epoch": 0.9884292874877032,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.0914556415157e-08,
+      "loss": 0.8615,
+      "step": 10550
+    },
+    {
+      "epoch": 0.9888977373869865,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.447064492951405e-08,
+      "loss": 0.8503,
+      "step": 10555
+    },
+    {
+      "epoch": 0.9893661872862697,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.829398569770939e-08,
+      "loss": 0.8334,
+      "step": 10560
+    },
+    {
+      "epoch": 0.989834637185553,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.238459523932871e-08,
+      "loss": 0.8601,
+      "step": 10565
+    },
+    {
+      "epoch": 0.9903030870848363,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.674248935910731e-08,
+      "loss": 0.8565,
+      "step": 10570
+    },
+    {
+      "epoch": 0.9907715369841196,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.136768314694118e-08,
+      "loss": 0.8497,
+      "step": 10575
+    },
+    {
+      "epoch": 0.9912399868834029,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.626019097782042e-08,
+      "loss": 0.8954,
+      "step": 10580
+    },
+    {
+      "epoch": 0.991708436782686,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.1420026511806984e-08,
+      "loss": 0.9051,
+      "step": 10585
+    },
+    {
+      "epoch": 0.9921768866819693,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.6847202694001434e-08,
+      "loss": 0.8412,
+      "step": 10590
+    },
+    {
+      "epoch": 0.9926453365812526,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.254173175447628e-08,
+      "loss": 0.8516,
+      "step": 10595
+    },
+    {
+      "epoch": 0.9931137864805359,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.8503625208275986e-08,
+      "loss": 0.8652,
+      "step": 10600
+    },
+    {
+      "epoch": 0.9935822363798191,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.473289385538369e-08,
+      "loss": 0.8294,
+      "step": 10605
+    },
+    {
+      "epoch": 0.9940506862791024,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.122954778065456e-08,
+      "loss": 0.8417,
+      "step": 10610
+    },
+    {
+      "epoch": 0.9945191361783857,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7993596353849118e-08,
+      "loss": 0.8383,
+      "step": 10615
+    },
+    {
+      "epoch": 0.994987586077669,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.502504822957773e-08,
+      "loss": 0.8529,
+      "step": 10620
+    },
+    {
+      "epoch": 0.9954560359769523,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.2323911347245087e-08,
+      "loss": 0.8473,
+      "step": 10625
+    },
+    {
+      "epoch": 0.9959244858762355,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.890192931083508e-09,
+      "loss": 0.8372,
+      "step": 10630
+    },
+    {
+      "epoch": 0.9963929357755188,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.723899490108544e-09,
+      "loss": 0.8238,
+      "step": 10635
+    },
+    {
+      "epoch": 0.9968613856748021,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.825036818118967e-09,
+      "loss": 0.8652,
+      "step": 10640
+    },
+    {
+      "epoch": 0.9973298355740854,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.193609993630166e-09,
+      "loss": 0.8583,
+      "step": 10645
+    },
+    {
+      "epoch": 0.9977982854733686,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.8296233799185444e-09,
+      "loss": 0.8498,
+      "step": 10650
+    },
+    {
+      "epoch": 0.9982667353726519,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.733080624999328e-09,
+      "loss": 0.8679,
+      "step": 10655
+    },
+    {
+      "epoch": 0.9987351852719352,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.039846615710446e-10,
+      "loss": 0.8622,
+      "step": 10660
+    },
+    {
+      "epoch": 0.9992036351712185,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.423377070932432e-10,
+      "loss": 0.841,
+      "step": 10665
+    },
+    {
+      "epoch": 0.9996720850705018,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.8141263664369664e-11,
+      "loss": 0.8429,
+      "step": 10670
+    },
+    {
+      "epoch": 0.9999531550100716,
+      "eval_loss": 0.8574451804161072,
+      "eval_runtime": 961.4713,
+      "eval_samples_per_second": 4.677,
+      "eval_steps_per_second": 1.17,
+      "step": 10673
+    },
+    {
+      "epoch": 0.9999531550100716,
+      "step": 10673,
+      "total_flos": 4.092531867595571e+18,
+      "train_loss": 1.0194565681556524,
+      "train_runtime": 58622.8041,
+      "train_samples_per_second": 1.457,
+      "train_steps_per_second": 0.182
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 10673,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 4.092531867595571e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}