{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 648,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07032967032967033,
      "grad_norm": 0.015869140625,
      "learning_rate": 4.972222222222223e-05,
      "loss": 0.6071998119354248,
      "step": 5
    },
    {
      "epoch": 0.14065934065934066,
      "grad_norm": 0.01129150390625,
      "learning_rate": 4.937500000000001e-05,
      "loss": 0.6054113388061524,
      "step": 10
    },
    {
      "epoch": 0.210989010989011,
      "grad_norm": 0.01287841796875,
      "learning_rate": 4.902777777777778e-05,
      "loss": 0.6566670417785645,
      "step": 15
    },
    {
      "epoch": 0.2813186813186813,
      "grad_norm": 0.01220703125,
      "learning_rate": 4.8680555555555554e-05,
      "loss": 0.6298727989196777,
      "step": 20
    },
    {
      "epoch": 0.3516483516483517,
      "grad_norm": 0.013916015625,
      "learning_rate": 4.8333333333333334e-05,
      "loss": 0.6060850143432617,
      "step": 25
    },
    {
      "epoch": 0.421978021978022,
      "grad_norm": 0.01153564453125,
      "learning_rate": 4.7986111111111113e-05,
      "loss": 0.5983633995056152,
      "step": 30
    },
    {
      "epoch": 0.49230769230769234,
      "grad_norm": 0.00970458984375,
      "learning_rate": 4.7638888888888887e-05,
      "loss": 0.5858489036560058,
      "step": 35
    },
    {
      "epoch": 0.5626373626373626,
      "grad_norm": 0.01385498046875,
      "learning_rate": 4.7291666666666666e-05,
      "loss": 0.6647251129150391,
      "step": 40
    },
    {
      "epoch": 0.6329670329670329,
      "grad_norm": 0.01177978515625,
      "learning_rate": 4.6944444444444446e-05,
      "loss": 0.606710147857666,
      "step": 45
    },
    {
      "epoch": 0.7032967032967034,
      "grad_norm": 0.011962890625,
      "learning_rate": 4.6597222222222226e-05,
      "loss": 0.5918601512908935,
      "step": 50
    },
    {
      "epoch": 0.7736263736263737,
      "grad_norm": 0.01129150390625,
      "learning_rate": 4.6250000000000006e-05,
      "loss": 0.6076582431793213,
      "step": 55
    },
    {
      "epoch": 0.843956043956044,
      "grad_norm": 0.00958251953125,
      "learning_rate": 4.590277777777778e-05,
      "loss": 0.5553756713867187,
      "step": 60
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 0.01348876953125,
      "learning_rate": 4.555555555555556e-05,
      "loss": 0.6189921379089356,
      "step": 65
    },
    {
      "epoch": 0.9846153846153847,
      "grad_norm": 0.01220703125,
      "learning_rate": 4.520833333333334e-05,
      "loss": 0.5911207675933838,
      "step": 70
    },
    {
      "epoch": 1.0421978021978022,
      "grad_norm": 0.00958251953125,
      "learning_rate": 4.486111111111111e-05,
      "loss": 0.5420764923095703,
      "step": 75
    },
    {
      "epoch": 1.1125274725274725,
      "grad_norm": 0.01153564453125,
      "learning_rate": 4.4513888888888885e-05,
      "loss": 0.4824223518371582,
      "step": 80
    },
    {
      "epoch": 1.1828571428571428,
      "grad_norm": 0.01434326171875,
      "learning_rate": 4.4166666666666665e-05,
      "loss": 0.5018988132476807,
      "step": 85
    },
    {
      "epoch": 1.2531868131868131,
      "grad_norm": 0.013671875,
      "learning_rate": 4.3819444444444445e-05,
      "loss": 0.5336177825927735,
      "step": 90
    },
    {
      "epoch": 1.3235164835164834,
      "grad_norm": 0.01263427734375,
      "learning_rate": 4.3472222222222225e-05,
      "loss": 0.4850775241851807,
      "step": 95
    },
    {
      "epoch": 1.393846153846154,
      "grad_norm": 0.0126953125,
      "learning_rate": 4.3125000000000005e-05,
      "loss": 0.5138346195220947,
      "step": 100
    },
    {
      "epoch": 1.4641758241758243,
      "grad_norm": 0.01397705078125,
      "learning_rate": 4.277777777777778e-05,
      "loss": 0.5070078849792481,
      "step": 105
    },
    {
      "epoch": 1.5345054945054946,
      "grad_norm": 0.0118408203125,
      "learning_rate": 4.243055555555556e-05,
      "loss": 0.5198071479797364,
      "step": 110
    },
    {
      "epoch": 1.6048351648351649,
      "grad_norm": 0.0137939453125,
      "learning_rate": 4.208333333333334e-05,
      "loss": 0.5197066783905029,
      "step": 115
    },
    {
      "epoch": 1.6751648351648352,
      "grad_norm": 0.00970458984375,
      "learning_rate": 4.173611111111112e-05,
      "loss": 0.4952798366546631,
      "step": 120
    },
    {
      "epoch": 1.7454945054945055,
      "grad_norm": 0.01190185546875,
      "learning_rate": 4.138888888888889e-05,
      "loss": 0.5169046878814697,
      "step": 125
    },
    {
      "epoch": 1.8158241758241758,
      "grad_norm": 0.01324462890625,
      "learning_rate": 4.104166666666667e-05,
      "loss": 0.5308417320251465,
      "step": 130
    },
    {
      "epoch": 1.8861538461538463,
      "grad_norm": 0.0146484375,
      "learning_rate": 4.0694444444444444e-05,
      "loss": 0.46563105583190917,
      "step": 135
    },
    {
      "epoch": 1.9564835164835164,
      "grad_norm": 0.020751953125,
      "learning_rate": 4.0347222222222223e-05,
      "loss": 0.521865701675415,
      "step": 140
    },
    {
      "epoch": 2.014065934065934,
      "grad_norm": 0.01312255859375,
      "learning_rate": 4e-05,
      "loss": 0.44051213264465333,
      "step": 145
    },
    {
      "epoch": 2.0843956043956045,
      "grad_norm": 0.01263427734375,
      "learning_rate": 3.9652777777777776e-05,
      "loss": 0.43095035552978517,
      "step": 150
    },
    {
      "epoch": 2.1547252747252745,
      "grad_norm": 0.020263671875,
      "learning_rate": 3.9305555555555556e-05,
      "loss": 0.4152026653289795,
      "step": 155
    },
    {
      "epoch": 2.225054945054945,
      "grad_norm": 0.0120849609375,
      "learning_rate": 3.8958333333333336e-05,
      "loss": 0.43182148933410647,
      "step": 160
    },
    {
      "epoch": 2.2953846153846156,
      "grad_norm": 0.0162353515625,
      "learning_rate": 3.8611111111111116e-05,
      "loss": 0.4308079719543457,
      "step": 165
    },
    {
      "epoch": 2.3657142857142857,
      "grad_norm": 0.01300048828125,
      "learning_rate": 3.826388888888889e-05,
      "loss": 0.40520739555358887,
      "step": 170
    },
    {
      "epoch": 2.436043956043956,
      "grad_norm": 0.015380859375,
      "learning_rate": 3.791666666666667e-05,
      "loss": 0.4348421096801758,
      "step": 175
    },
    {
      "epoch": 2.5063736263736263,
      "grad_norm": 0.0166015625,
      "learning_rate": 3.756944444444445e-05,
      "loss": 0.3993671894073486,
      "step": 180
    },
    {
      "epoch": 2.576703296703297,
      "grad_norm": 0.01611328125,
      "learning_rate": 3.722222222222222e-05,
      "loss": 0.456925630569458,
      "step": 185
    },
    {
      "epoch": 2.647032967032967,
      "grad_norm": 0.0206298828125,
      "learning_rate": 3.6875e-05,
      "loss": 0.4396049976348877,
      "step": 190
    },
    {
      "epoch": 2.7173626373626374,
      "grad_norm": 0.014404296875,
      "learning_rate": 3.6527777777777775e-05,
      "loss": 0.41503162384033204,
      "step": 195
    },
    {
      "epoch": 2.787692307692308,
      "grad_norm": 0.01446533203125,
      "learning_rate": 3.6180555555555555e-05,
      "loss": 0.36976258754730223,
      "step": 200
    },
    {
      "epoch": 2.858021978021978,
      "grad_norm": 0.0166015625,
      "learning_rate": 3.5833333333333335e-05,
      "loss": 0.4395163059234619,
      "step": 205
    },
    {
      "epoch": 2.9283516483516485,
      "grad_norm": 0.0162353515625,
      "learning_rate": 3.5486111111111115e-05,
      "loss": 0.4275169849395752,
      "step": 210
    },
    {
      "epoch": 2.9986813186813186,
      "grad_norm": 0.01544189453125,
      "learning_rate": 3.513888888888889e-05,
      "loss": 0.42092361450195315,
      "step": 215
    },
    {
      "epoch": 3.056263736263736,
      "grad_norm": 0.01531982421875,
      "learning_rate": 3.479166666666667e-05,
      "loss": 0.3451509475708008,
      "step": 220
    },
    {
      "epoch": 3.1265934065934067,
      "grad_norm": 0.0205078125,
      "learning_rate": 3.444444444444445e-05,
      "loss": 0.31665565967559817,
      "step": 225
    },
    {
      "epoch": 3.1969230769230768,
      "grad_norm": 0.0223388671875,
      "learning_rate": 3.409722222222223e-05,
      "loss": 0.35940887928009035,
      "step": 230
    },
    {
      "epoch": 3.2672527472527473,
      "grad_norm": 0.0211181640625,
      "learning_rate": 3.375000000000001e-05,
      "loss": 0.3280967712402344,
      "step": 235
    },
    {
      "epoch": 3.337582417582418,
      "grad_norm": 0.018310546875,
      "learning_rate": 3.340277777777778e-05,
      "loss": 0.33126680850982665,
      "step": 240
    },
    {
      "epoch": 3.407912087912088,
      "grad_norm": 0.0220947265625,
      "learning_rate": 3.3055555555555553e-05,
      "loss": 0.2990967512130737,
      "step": 245
    },
    {
      "epoch": 3.4782417582417584,
      "grad_norm": 0.03369140625,
      "learning_rate": 3.270833333333333e-05,
      "loss": 0.34647085666656496,
      "step": 250
    },
    {
      "epoch": 3.5485714285714285,
      "grad_norm": 0.01611328125,
      "learning_rate": 3.236111111111111e-05,
      "loss": 0.3715843915939331,
      "step": 255
    },
    {
      "epoch": 3.618901098901099,
      "grad_norm": 0.0194091796875,
      "learning_rate": 3.2013888888888886e-05,
      "loss": 0.3007366418838501,
      "step": 260
    },
    {
      "epoch": 3.689230769230769,
      "grad_norm": 0.023193359375,
      "learning_rate": 3.1666666666666666e-05,
      "loss": 0.3250709533691406,
      "step": 265
    },
    {
      "epoch": 3.7595604395604396,
      "grad_norm": 0.021484375,
      "learning_rate": 3.1319444444444446e-05,
      "loss": 0.33896522521972655,
      "step": 270
    },
    {
      "epoch": 3.82989010989011,
      "grad_norm": 0.0235595703125,
      "learning_rate": 3.0972222222222226e-05,
      "loss": 0.2959084749221802,
      "step": 275
    },
    {
      "epoch": 3.90021978021978,
      "grad_norm": 0.0152587890625,
      "learning_rate": 3.0625000000000006e-05,
      "loss": 0.37686474323272706,
      "step": 280
    },
    {
      "epoch": 3.9705494505494503,
      "grad_norm": 0.0220947265625,
      "learning_rate": 3.0277777777777776e-05,
      "loss": 0.31239821910858157,
      "step": 285
    },
    {
      "epoch": 4.028131868131868,
      "grad_norm": 0.0247802734375,
      "learning_rate": 2.9930555555555555e-05,
      "loss": 0.2946542024612427,
      "step": 290
    },
    {
      "epoch": 4.098461538461539,
      "grad_norm": 0.017822265625,
      "learning_rate": 2.9583333333333335e-05,
      "loss": 0.2789935111999512,
      "step": 295
    },
    {
      "epoch": 4.168791208791209,
      "grad_norm": 0.022216796875,
      "learning_rate": 2.9236111111111115e-05,
      "loss": 0.23769619464874267,
      "step": 300
    },
    {
      "epoch": 4.239120879120879,
      "grad_norm": 0.0198974609375,
      "learning_rate": 2.8888888888888888e-05,
      "loss": 0.23825483322143554,
      "step": 305
    },
    {
      "epoch": 4.309450549450549,
      "grad_norm": 0.0240478515625,
      "learning_rate": 2.8541666666666668e-05,
      "loss": 0.28074493408203127,
      "step": 310
    },
    {
      "epoch": 4.37978021978022,
      "grad_norm": 0.0230712890625,
      "learning_rate": 2.8194444444444445e-05,
      "loss": 0.23039891719818115,
      "step": 315
    },
    {
      "epoch": 4.45010989010989,
      "grad_norm": 0.0216064453125,
      "learning_rate": 2.7847222222222224e-05,
      "loss": 0.21666464805603028,
      "step": 320
    },
    {
      "epoch": 4.52043956043956,
      "grad_norm": 0.02294921875,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.2543968677520752,
      "step": 325
    },
    {
      "epoch": 4.590769230769231,
      "grad_norm": 0.027099609375,
      "learning_rate": 2.7152777777777777e-05,
      "loss": 0.23768765926361085,
      "step": 330
    },
    {
      "epoch": 4.661098901098901,
      "grad_norm": 0.0240478515625,
      "learning_rate": 2.6805555555555557e-05,
      "loss": 0.2496436357498169,
      "step": 335
    },
    {
      "epoch": 4.731428571428571,
      "grad_norm": 0.027099609375,
      "learning_rate": 2.6458333333333334e-05,
      "loss": 0.24581174850463866,
      "step": 340
    },
    {
      "epoch": 4.801758241758241,
      "grad_norm": 0.034912109375,
      "learning_rate": 2.6111111111111114e-05,
      "loss": 0.25378801822662356,
      "step": 345
    },
    {
      "epoch": 4.872087912087912,
      "grad_norm": 0.03857421875,
      "learning_rate": 2.5763888888888887e-05,
      "loss": 0.2762407064437866,
      "step": 350
    },
    {
      "epoch": 4.9424175824175824,
      "grad_norm": 0.02001953125,
      "learning_rate": 2.5416666666666667e-05,
      "loss": 0.24667825698852539,
      "step": 355
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.038818359375,
      "learning_rate": 2.5069444444444447e-05,
      "loss": 0.2256716012954712,
      "step": 360
    },
    {
      "epoch": 5.07032967032967,
      "grad_norm": 0.02490234375,
      "learning_rate": 2.4722222222222223e-05,
      "loss": 0.1889257550239563,
      "step": 365
    },
    {
      "epoch": 5.140659340659341,
      "grad_norm": 0.0400390625,
      "learning_rate": 2.4375e-05,
      "loss": 0.17583080530166625,
      "step": 370
    },
    {
      "epoch": 5.210989010989011,
      "grad_norm": 0.02197265625,
      "learning_rate": 2.402777777777778e-05,
      "loss": 0.1796499252319336,
      "step": 375
    },
    {
      "epoch": 5.281318681318681,
      "grad_norm": 0.0257568359375,
      "learning_rate": 2.3680555555555556e-05,
      "loss": 0.16900094747543334,
      "step": 380
    },
    {
      "epoch": 5.351648351648351,
      "grad_norm": 0.0255126953125,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 0.18438608646392823,
      "step": 385
    },
    {
      "epoch": 5.421978021978022,
      "grad_norm": 0.0294189453125,
      "learning_rate": 2.2986111111111112e-05,
      "loss": 0.1891573190689087,
      "step": 390
    },
    {
      "epoch": 5.492307692307692,
      "grad_norm": 0.031982421875,
      "learning_rate": 2.263888888888889e-05,
      "loss": 0.194422447681427,
      "step": 395
    },
    {
      "epoch": 5.562637362637362,
      "grad_norm": 0.0272216796875,
      "learning_rate": 2.229166666666667e-05,
      "loss": 0.18956830501556396,
      "step": 400
    },
    {
      "epoch": 5.6329670329670325,
      "grad_norm": 0.0478515625,
      "learning_rate": 2.1944444444444445e-05,
      "loss": 0.1821696162223816,
      "step": 405
    },
    {
      "epoch": 5.7032967032967035,
      "grad_norm": 0.0267333984375,
      "learning_rate": 2.1597222222222225e-05,
      "loss": 0.18890082836151123,
      "step": 410
    },
    {
      "epoch": 5.7736263736263735,
      "grad_norm": 0.0235595703125,
      "learning_rate": 2.125e-05,
      "loss": 0.18818626403808594,
      "step": 415
    },
    {
      "epoch": 5.843956043956044,
      "grad_norm": 0.0269775390625,
      "learning_rate": 2.0902777777777778e-05,
      "loss": 0.19587171077728271,
      "step": 420
    },
    {
      "epoch": 5.914285714285715,
      "grad_norm": 0.029052734375,
      "learning_rate": 2.0555555555555555e-05,
      "loss": 0.16554845571517945,
      "step": 425
    },
    {
      "epoch": 5.984615384615385,
      "grad_norm": 0.0250244140625,
      "learning_rate": 2.0208333333333334e-05,
      "loss": 0.18133629560470582,
      "step": 430
    },
    {
      "epoch": 6.042197802197802,
      "grad_norm": 0.0262451171875,
      "learning_rate": 1.986111111111111e-05,
      "loss": 0.19384570121765138,
      "step": 435
    },
    {
      "epoch": 6.112527472527472,
      "grad_norm": 0.0169677734375,
      "learning_rate": 1.951388888888889e-05,
      "loss": 0.12632473707199096,
      "step": 440
    },
    {
      "epoch": 6.182857142857143,
      "grad_norm": 0.0211181640625,
      "learning_rate": 1.9166666666666667e-05,
      "loss": 0.1315527081489563,
      "step": 445
    },
    {
      "epoch": 6.253186813186813,
      "grad_norm": 0.0247802734375,
      "learning_rate": 1.8819444444444444e-05,
      "loss": 0.1730587124824524,
      "step": 450
    },
    {
      "epoch": 6.323516483516483,
      "grad_norm": 0.02880859375,
      "learning_rate": 1.8472222222222224e-05,
      "loss": 0.14160635471343994,
      "step": 455
    },
    {
      "epoch": 6.3938461538461535,
      "grad_norm": 0.0216064453125,
      "learning_rate": 1.8125e-05,
      "loss": 0.13952556848526002,
      "step": 460
    },
    {
      "epoch": 6.4641758241758245,
      "grad_norm": 0.025146484375,
      "learning_rate": 1.777777777777778e-05,
      "loss": 0.13826658725738525,
      "step": 465
    },
    {
      "epoch": 6.534505494505495,
      "grad_norm": 0.0264892578125,
      "learning_rate": 1.7430555555555556e-05,
      "loss": 0.13782529830932616,
      "step": 470
    },
    {
      "epoch": 6.604835164835165,
      "grad_norm": 0.0302734375,
      "learning_rate": 1.7083333333333333e-05,
      "loss": 0.13006110191345216,
      "step": 475
    },
    {
      "epoch": 6.675164835164836,
      "grad_norm": 0.02490234375,
      "learning_rate": 1.673611111111111e-05,
      "loss": 0.16382263898849486,
      "step": 480
    },
    {
      "epoch": 6.745494505494506,
      "grad_norm": 0.019775390625,
      "learning_rate": 1.638888888888889e-05,
      "loss": 0.13489021062850953,
      "step": 485
    },
    {
      "epoch": 6.815824175824176,
      "grad_norm": 0.0235595703125,
      "learning_rate": 1.604166666666667e-05,
      "loss": 0.1541507601737976,
      "step": 490
    },
    {
      "epoch": 6.886153846153846,
      "grad_norm": 0.0277099609375,
      "learning_rate": 1.5694444444444446e-05,
      "loss": 0.13053619861602783,
      "step": 495
    },
    {
      "epoch": 6.956483516483517,
      "grad_norm": 0.02783203125,
      "learning_rate": 1.5347222222222226e-05,
      "loss": 0.12665542364120483,
      "step": 500
    },
    {
      "epoch": 7.014065934065934,
      "grad_norm": 0.0211181640625,
      "learning_rate": 1.5e-05,
      "loss": 0.11801940202713013,
      "step": 505
    },
    {
      "epoch": 7.0843956043956045,
      "grad_norm": 0.0213623046875,
      "learning_rate": 1.4652777777777779e-05,
      "loss": 0.13198474645614625,
      "step": 510
    },
    {
      "epoch": 7.1547252747252745,
      "grad_norm": 0.017333984375,
      "learning_rate": 1.4305555555555555e-05,
      "loss": 0.11638516187667847,
      "step": 515
    },
    {
      "epoch": 7.225054945054945,
      "grad_norm": 0.0172119140625,
      "learning_rate": 1.3958333333333335e-05,
      "loss": 0.12684643268585205,
      "step": 520
    },
    {
      "epoch": 7.295384615384616,
      "grad_norm": 0.0228271484375,
      "learning_rate": 1.3611111111111111e-05,
      "loss": 0.11121166944503784,
      "step": 525
    },
    {
      "epoch": 7.365714285714286,
      "grad_norm": 0.015625,
      "learning_rate": 1.326388888888889e-05,
      "loss": 0.11003491878509522,
      "step": 530
    },
    {
      "epoch": 7.436043956043956,
      "grad_norm": 0.01458740234375,
      "learning_rate": 1.2916666666666668e-05,
      "loss": 0.11681123971939086,
      "step": 535
    },
    {
      "epoch": 7.506373626373627,
      "grad_norm": 0.016845703125,
      "learning_rate": 1.2569444444444444e-05,
      "loss": 0.10627522468566894,
      "step": 540
    },
    {
      "epoch": 7.576703296703297,
      "grad_norm": 0.015869140625,
      "learning_rate": 1.2222222222222222e-05,
      "loss": 0.10095436573028564,
      "step": 545
    },
    {
      "epoch": 7.647032967032967,
      "grad_norm": 0.02392578125,
      "learning_rate": 1.1875e-05,
      "loss": 0.13068802356719972,
      "step": 550
    },
    {
      "epoch": 7.717362637362637,
      "grad_norm": 0.0196533203125,
      "learning_rate": 1.1527777777777779e-05,
      "loss": 0.10702927112579345,
      "step": 555
    },
    {
      "epoch": 7.787692307692308,
      "grad_norm": 0.019775390625,
      "learning_rate": 1.1180555555555557e-05,
      "loss": 0.09682444334030152,
      "step": 560
    },
    {
      "epoch": 7.858021978021978,
      "grad_norm": 0.0181884765625,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.1158555030822754,
      "step": 565
    },
    {
      "epoch": 7.928351648351648,
      "grad_norm": 0.01904296875,
      "learning_rate": 1.0486111111111112e-05,
      "loss": 0.12258208990097046,
      "step": 570
    },
    {
      "epoch": 7.998681318681319,
      "grad_norm": 0.0294189453125,
      "learning_rate": 1.013888888888889e-05,
      "loss": 0.11514959335327149,
      "step": 575
    },
    {
      "epoch": 8.056263736263736,
      "grad_norm": 0.01806640625,
      "learning_rate": 9.791666666666666e-06,
      "loss": 0.12059261798858642,
      "step": 580
    },
    {
      "epoch": 8.126593406593406,
      "grad_norm": 0.015380859375,
      "learning_rate": 9.444444444444445e-06,
      "loss": 0.10077341794967651,
      "step": 585
    },
    {
      "epoch": 8.196923076923078,
      "grad_norm": 0.0191650390625,
      "learning_rate": 9.097222222222223e-06,
      "loss": 0.1014246940612793,
      "step": 590
    },
    {
      "epoch": 8.267252747252748,
      "grad_norm": 0.044921875,
      "learning_rate": 8.75e-06,
      "loss": 0.11813064813613891,
      "step": 595
    },
    {
      "epoch": 8.337582417582418,
      "grad_norm": 0.0186767578125,
      "learning_rate": 8.402777777777779e-06,
      "loss": 0.11902542114257812,
      "step": 600
    },
    {
      "epoch": 8.407912087912088,
      "grad_norm": 0.0164794921875,
      "learning_rate": 8.055555555555557e-06,
      "loss": 0.09582686424255371,
      "step": 605
    },
    {
      "epoch": 8.478241758241758,
      "grad_norm": 0.01446533203125,
      "learning_rate": 7.708333333333334e-06,
      "loss": 0.10142171382904053,
      "step": 610
    },
    {
      "epoch": 8.548571428571428,
      "grad_norm": 0.01336669921875,
      "learning_rate": 7.361111111111112e-06,
      "loss": 0.09789245128631592,
      "step": 615
    },
    {
      "epoch": 8.618901098901098,
      "grad_norm": 0.0150146484375,
      "learning_rate": 7.013888888888889e-06,
      "loss": 0.10006954669952392,
      "step": 620
    },
    {
      "epoch": 8.68923076923077,
      "grad_norm": 0.019775390625,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.11400750875473023,
      "step": 625
    },
    {
      "epoch": 8.75956043956044,
      "grad_norm": 0.0208740234375,
      "learning_rate": 6.319444444444445e-06,
      "loss": 0.10895116329193115,
      "step": 630
    },
    {
      "epoch": 8.82989010989011,
      "grad_norm": 0.037841796875,
      "learning_rate": 5.972222222222223e-06,
      "loss": 0.09994078874588012,
      "step": 635
    },
    {
      "epoch": 8.90021978021978,
      "grad_norm": 0.01611328125,
      "learning_rate": 5.625e-06,
      "loss": 0.09743757843971253,
      "step": 640
    },
    {
      "epoch": 8.97054945054945,
      "grad_norm": 0.0164794921875,
      "learning_rate": 5.277777777777778e-06,
      "loss": 0.10574103593826294,
      "step": 645
    }
  ],
  "logging_steps": 5,
  "max_steps": 720,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.6626503286521856e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}