{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1205,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02074688796680498,
      "grad_norm": 24.019546508789062,
      "learning_rate": 9.95850622406639e-06,
      "loss": 4.1829,
      "step": 5
    },
    {
      "epoch": 0.04149377593360996,
      "grad_norm": 8.10768985748291,
      "learning_rate": 9.91701244813278e-06,
      "loss": 0.3168,
      "step": 10
    },
    {
      "epoch": 0.06224066390041494,
      "grad_norm": 6.1697587966918945,
      "learning_rate": 9.875518672199172e-06,
      "loss": 0.2449,
      "step": 15
    },
    {
      "epoch": 0.08298755186721991,
      "grad_norm": 3.5985426902770996,
      "learning_rate": 9.83402489626556e-06,
      "loss": 0.1747,
      "step": 20
    },
    {
      "epoch": 0.1037344398340249,
      "grad_norm": 7.316824436187744,
      "learning_rate": 9.792531120331951e-06,
      "loss": 0.1978,
      "step": 25
    },
    {
      "epoch": 0.12448132780082988,
      "grad_norm": 4.547542572021484,
      "learning_rate": 9.751037344398341e-06,
      "loss": 0.143,
      "step": 30
    },
    {
      "epoch": 0.14522821576763487,
      "grad_norm": 4.3512043952941895,
      "learning_rate": 9.709543568464731e-06,
      "loss": 0.1289,
      "step": 35
    },
    {
      "epoch": 0.16597510373443983,
      "grad_norm": 4.462110996246338,
      "learning_rate": 9.66804979253112e-06,
      "loss": 0.1096,
      "step": 40
    },
    {
      "epoch": 0.18672199170124482,
      "grad_norm": 4.343535423278809,
      "learning_rate": 9.62655601659751e-06,
      "loss": 0.1213,
      "step": 45
    },
    {
      "epoch": 0.2074688796680498,
      "grad_norm": 3.772761821746826,
      "learning_rate": 9.585062240663902e-06,
      "loss": 0.1172,
      "step": 50
    },
    {
      "epoch": 0.22821576763485477,
      "grad_norm": 2.0513064861297607,
      "learning_rate": 9.543568464730292e-06,
      "loss": 0.0802,
      "step": 55
    },
    {
      "epoch": 0.24896265560165975,
      "grad_norm": 4.065267086029053,
      "learning_rate": 9.502074688796682e-06,
      "loss": 0.129,
      "step": 60
    },
    {
      "epoch": 0.2697095435684647,
      "grad_norm": 6.929261207580566,
      "learning_rate": 9.460580912863071e-06,
      "loss": 0.0843,
      "step": 65
    },
    {
      "epoch": 0.29045643153526973,
      "grad_norm": 1.971679449081421,
      "learning_rate": 9.419087136929461e-06,
      "loss": 0.1294,
      "step": 70
    },
    {
      "epoch": 0.3112033195020747,
      "grad_norm": 2.9084889888763428,
      "learning_rate": 9.377593360995851e-06,
      "loss": 0.0824,
      "step": 75
    },
    {
      "epoch": 0.33195020746887965,
      "grad_norm": 3.351242780685425,
      "learning_rate": 9.33609958506224e-06,
      "loss": 0.0684,
      "step": 80
    },
    {
      "epoch": 0.35269709543568467,
      "grad_norm": 3.296483039855957,
      "learning_rate": 9.294605809128632e-06,
      "loss": 0.0566,
      "step": 85
    },
    {
      "epoch": 0.37344398340248963,
      "grad_norm": 2.073906421661377,
      "learning_rate": 9.253112033195022e-06,
      "loss": 0.0494,
      "step": 90
    },
    {
      "epoch": 0.3941908713692946,
      "grad_norm": 1.1432421207427979,
      "learning_rate": 9.211618257261412e-06,
      "loss": 0.0627,
      "step": 95
    },
    {
      "epoch": 0.4149377593360996,
      "grad_norm": 3.387030601501465,
      "learning_rate": 9.170124481327802e-06,
      "loss": 0.0826,
      "step": 100
    },
    {
      "epoch": 0.43568464730290457,
      "grad_norm": 2.8472816944122314,
      "learning_rate": 9.128630705394191e-06,
      "loss": 0.0966,
      "step": 105
    },
    {
      "epoch": 0.45643153526970953,
      "grad_norm": 3.8003361225128174,
      "learning_rate": 9.087136929460581e-06,
      "loss": 0.0517,
      "step": 110
    },
    {
      "epoch": 0.47717842323651455,
      "grad_norm": 3.382927417755127,
      "learning_rate": 9.045643153526971e-06,
      "loss": 0.0937,
      "step": 115
    },
    {
      "epoch": 0.4979253112033195,
      "grad_norm": 2.5705699920654297,
      "learning_rate": 9.004149377593362e-06,
      "loss": 0.0898,
      "step": 120
    },
    {
      "epoch": 0.5186721991701245,
      "grad_norm": 1.917487621307373,
      "learning_rate": 8.962655601659752e-06,
      "loss": 0.0868,
      "step": 125
    },
    {
      "epoch": 0.5394190871369294,
      "grad_norm": 3.543142318725586,
      "learning_rate": 8.921161825726142e-06,
      "loss": 0.0527,
      "step": 130
    },
    {
      "epoch": 0.5601659751037344,
      "grad_norm": 2.3501718044281006,
      "learning_rate": 8.879668049792532e-06,
      "loss": 0.0521,
      "step": 135
    },
    {
      "epoch": 0.5809128630705395,
      "grad_norm": 2.6340670585632324,
      "learning_rate": 8.838174273858922e-06,
      "loss": 0.0643,
      "step": 140
    },
    {
      "epoch": 0.6016597510373444,
      "grad_norm": 4.2217230796813965,
      "learning_rate": 8.796680497925311e-06,
      "loss": 0.0928,
      "step": 145
    },
    {
      "epoch": 0.6224066390041494,
      "grad_norm": 2.5034725666046143,
      "learning_rate": 8.755186721991701e-06,
      "loss": 0.0576,
      "step": 150
    },
    {
      "epoch": 0.6431535269709544,
      "grad_norm": 2.9784247875213623,
      "learning_rate": 8.713692946058093e-06,
      "loss": 0.0651,
      "step": 155
    },
    {
      "epoch": 0.6639004149377593,
      "grad_norm": 1.975595474243164,
      "learning_rate": 8.672199170124482e-06,
      "loss": 0.0461,
      "step": 160
    },
    {
      "epoch": 0.6846473029045643,
      "grad_norm": 3.185384511947632,
      "learning_rate": 8.630705394190872e-06,
      "loss": 0.0778,
      "step": 165
    },
    {
      "epoch": 0.7053941908713693,
      "grad_norm": 3.2054567337036133,
      "learning_rate": 8.589211618257262e-06,
      "loss": 0.0491,
      "step": 170
    },
    {
      "epoch": 0.7261410788381742,
      "grad_norm": 1.9956356287002563,
      "learning_rate": 8.547717842323652e-06,
      "loss": 0.0487,
      "step": 175
    },
    {
      "epoch": 0.7468879668049793,
      "grad_norm": 4.406678199768066,
      "learning_rate": 8.506224066390042e-06,
      "loss": 0.0817,
      "step": 180
    },
    {
      "epoch": 0.7676348547717843,
      "grad_norm": 2.956308603286743,
      "learning_rate": 8.464730290456431e-06,
      "loss": 0.064,
      "step": 185
    },
    {
      "epoch": 0.7883817427385892,
      "grad_norm": 2.3644497394561768,
      "learning_rate": 8.423236514522823e-06,
      "loss": 0.057,
      "step": 190
    },
    {
      "epoch": 0.8091286307053942,
      "grad_norm": 5.967751502990723,
      "learning_rate": 8.381742738589213e-06,
      "loss": 0.0562,
      "step": 195
    },
    {
      "epoch": 0.8298755186721992,
      "grad_norm": 1.9066091775894165,
      "learning_rate": 8.340248962655602e-06,
      "loss": 0.0531,
      "step": 200
    },
    {
      "epoch": 0.8506224066390041,
      "grad_norm": 3.8539278507232666,
      "learning_rate": 8.298755186721992e-06,
      "loss": 0.069,
      "step": 205
    },
    {
      "epoch": 0.8713692946058091,
      "grad_norm": 2.3997936248779297,
      "learning_rate": 8.257261410788382e-06,
      "loss": 0.0515,
      "step": 210
    },
    {
      "epoch": 0.8921161825726142,
      "grad_norm": 2.316134452819824,
      "learning_rate": 8.215767634854772e-06,
      "loss": 0.0477,
      "step": 215
    },
    {
      "epoch": 0.9128630705394191,
      "grad_norm": 1.7633816003799438,
      "learning_rate": 8.174273858921162e-06,
      "loss": 0.0444,
      "step": 220
    },
    {
      "epoch": 0.9336099585062241,
      "grad_norm": 1.517957091331482,
      "learning_rate": 8.132780082987553e-06,
      "loss": 0.0521,
      "step": 225
    },
    {
      "epoch": 0.9543568464730291,
      "grad_norm": 2.498708486557007,
      "learning_rate": 8.091286307053943e-06,
      "loss": 0.0694,
      "step": 230
    },
    {
      "epoch": 0.975103734439834,
      "grad_norm": 1.0663732290267944,
      "learning_rate": 8.049792531120333e-06,
      "loss": 0.0477,
      "step": 235
    },
    {
      "epoch": 0.995850622406639,
      "grad_norm": 2.276492118835449,
      "learning_rate": 8.008298755186722e-06,
      "loss": 0.0557,
      "step": 240
    },
    {
      "epoch": 1.016597510373444,
      "grad_norm": 2.307063341140747,
      "learning_rate": 7.966804979253112e-06,
      "loss": 0.0409,
      "step": 245
    },
    {
      "epoch": 1.037344398340249,
      "grad_norm": 1.3632686138153076,
      "learning_rate": 7.925311203319502e-06,
      "loss": 0.0219,
      "step": 250
    },
    {
      "epoch": 1.058091286307054,
      "grad_norm": 2.1472036838531494,
      "learning_rate": 7.883817427385892e-06,
      "loss": 0.0348,
      "step": 255
    },
    {
      "epoch": 1.0788381742738589,
      "grad_norm": 2.039602279663086,
      "learning_rate": 7.842323651452283e-06,
      "loss": 0.0402,
      "step": 260
    },
    {
      "epoch": 1.099585062240664,
      "grad_norm": 1.9459106922149658,
      "learning_rate": 7.800829875518673e-06,
      "loss": 0.0361,
      "step": 265
    },
    {
      "epoch": 1.120331950207469,
      "grad_norm": 2.2730486392974854,
      "learning_rate": 7.759336099585063e-06,
      "loss": 0.0328,
      "step": 270
    },
    {
      "epoch": 1.1410788381742738,
      "grad_norm": 1.9750105142593384,
      "learning_rate": 7.717842323651453e-06,
      "loss": 0.0396,
      "step": 275
    },
    {
      "epoch": 1.161825726141079,
      "grad_norm": 0.9928601980209351,
      "learning_rate": 7.676348547717844e-06,
      "loss": 0.0318,
      "step": 280
    },
    {
      "epoch": 1.1825726141078838,
      "grad_norm": 1.7180863618850708,
      "learning_rate": 7.634854771784232e-06,
      "loss": 0.031,
      "step": 285
    },
    {
      "epoch": 1.2033195020746887,
      "grad_norm": 2.4514567852020264,
      "learning_rate": 7.593360995850623e-06,
      "loss": 0.0281,
      "step": 290
    },
    {
      "epoch": 1.2240663900414939,
      "grad_norm": 2.0262625217437744,
      "learning_rate": 7.5518672199170125e-06,
      "loss": 0.032,
      "step": 295
    },
    {
      "epoch": 1.2448132780082988,
      "grad_norm": 0.6366137266159058,
      "learning_rate": 7.510373443983403e-06,
      "loss": 0.02,
      "step": 300
    },
    {
      "epoch": 1.2655601659751037,
      "grad_norm": 1.343851089477539,
      "learning_rate": 7.468879668049793e-06,
      "loss": 0.0207,
      "step": 305
    },
    {
      "epoch": 1.2863070539419086,
      "grad_norm": 2.311086416244507,
      "learning_rate": 7.427385892116184e-06,
      "loss": 0.0284,
      "step": 310
    },
    {
      "epoch": 1.3070539419087137,
      "grad_norm": 0.7572755813598633,
      "learning_rate": 7.385892116182573e-06,
      "loss": 0.0285,
      "step": 315
    },
    {
      "epoch": 1.3278008298755186,
      "grad_norm": 1.4651453495025635,
      "learning_rate": 7.344398340248964e-06,
      "loss": 0.0286,
      "step": 320
    },
    {
      "epoch": 1.3485477178423237,
      "grad_norm": 0.34843432903289795,
      "learning_rate": 7.302904564315353e-06,
      "loss": 0.0301,
      "step": 325
    },
    {
      "epoch": 1.3692946058091287,
      "grad_norm": 3.212113857269287,
      "learning_rate": 7.261410788381743e-06,
      "loss": 0.0467,
      "step": 330
    },
    {
      "epoch": 1.3900414937759336,
      "grad_norm": 2.2156147956848145,
      "learning_rate": 7.219917012448133e-06,
      "loss": 0.0281,
      "step": 335
    },
    {
      "epoch": 1.4107883817427385,
      "grad_norm": 1.4567726850509644,
      "learning_rate": 7.178423236514523e-06,
      "loss": 0.0387,
      "step": 340
    },
    {
      "epoch": 1.4315352697095436,
      "grad_norm": 2.2478785514831543,
      "learning_rate": 7.136929460580914e-06,
      "loss": 0.0373,
      "step": 345
    },
    {
      "epoch": 1.4522821576763485,
      "grad_norm": 1.5350704193115234,
      "learning_rate": 7.0954356846473036e-06,
      "loss": 0.0239,
      "step": 350
    },
    {
      "epoch": 1.4730290456431536,
      "grad_norm": 2.211454391479492,
      "learning_rate": 7.053941908713694e-06,
      "loss": 0.0176,
      "step": 355
    },
    {
      "epoch": 1.4937759336099585,
      "grad_norm": 1.7876611948013306,
      "learning_rate": 7.012448132780083e-06,
      "loss": 0.0236,
      "step": 360
    },
    {
      "epoch": 1.5145228215767634,
      "grad_norm": 1.034866452217102,
      "learning_rate": 6.970954356846473e-06,
      "loss": 0.0268,
      "step": 365
    },
    {
      "epoch": 1.5352697095435683,
      "grad_norm": 2.368783473968506,
      "learning_rate": 6.9294605809128635e-06,
      "loss": 0.0358,
      "step": 370
    },
    {
      "epoch": 1.5560165975103735,
      "grad_norm": 2.605402708053589,
      "learning_rate": 6.887966804979253e-06,
      "loss": 0.0484,
      "step": 375
    },
    {
      "epoch": 1.5767634854771784,
      "grad_norm": 2.10164475440979,
      "learning_rate": 6.846473029045644e-06,
      "loss": 0.0375,
      "step": 380
    },
    {
      "epoch": 1.5975103734439835,
      "grad_norm": 1.2445701360702515,
      "learning_rate": 6.804979253112034e-06,
      "loss": 0.0236,
      "step": 385
    },
    {
      "epoch": 1.6182572614107884,
      "grad_norm": 3.539241313934326,
      "learning_rate": 6.763485477178424e-06,
      "loss": 0.032,
      "step": 390
    },
    {
      "epoch": 1.6390041493775933,
      "grad_norm": 2.5593557357788086,
      "learning_rate": 6.721991701244814e-06,
      "loss": 0.039,
      "step": 395
    },
    {
      "epoch": 1.6597510373443982,
      "grad_norm": 1.5487468242645264,
      "learning_rate": 6.680497925311203e-06,
      "loss": 0.0456,
      "step": 400
    },
    {
      "epoch": 1.6804979253112033,
      "grad_norm": 1.9492735862731934,
      "learning_rate": 6.639004149377594e-06,
      "loss": 0.0269,
      "step": 405
    },
    {
      "epoch": 1.7012448132780082,
      "grad_norm": 1.5356554985046387,
      "learning_rate": 6.5975103734439835e-06,
      "loss": 0.039,
      "step": 410
    },
    {
      "epoch": 1.7219917012448134,
      "grad_norm": 1.8745402097702026,
      "learning_rate": 6.556016597510374e-06,
      "loss": 0.0398,
      "step": 415
    },
    {
      "epoch": 1.7427385892116183,
      "grad_norm": 1.7888027429580688,
      "learning_rate": 6.514522821576764e-06,
      "loss": 0.0587,
      "step": 420
    },
    {
      "epoch": 1.7634854771784232,
      "grad_norm": 1.2012983560562134,
      "learning_rate": 6.4730290456431546e-06,
      "loss": 0.0134,
      "step": 425
    },
    {
      "epoch": 1.784232365145228,
      "grad_norm": 2.0974526405334473,
      "learning_rate": 6.431535269709544e-06,
      "loss": 0.0275,
      "step": 430
    },
    {
      "epoch": 1.8049792531120332,
      "grad_norm": 2.5113155841827393,
      "learning_rate": 6.390041493775933e-06,
      "loss": 0.0467,
      "step": 435
    },
    {
      "epoch": 1.8257261410788381,
      "grad_norm": 2.2300188541412354,
      "learning_rate": 6.348547717842324e-06,
      "loss": 0.0328,
      "step": 440
    },
    {
      "epoch": 1.8464730290456433,
      "grad_norm": 1.998130202293396,
      "learning_rate": 6.307053941908714e-06,
      "loss": 0.038,
      "step": 445
    },
    {
      "epoch": 1.8672199170124482,
      "grad_norm": 2.1886332035064697,
      "learning_rate": 6.265560165975104e-06,
      "loss": 0.0319,
      "step": 450
    },
    {
      "epoch": 1.887966804979253,
      "grad_norm": 2.2429630756378174,
      "learning_rate": 6.224066390041494e-06,
      "loss": 0.0416,
      "step": 455
    },
    {
      "epoch": 1.908713692946058,
      "grad_norm": 3.2857184410095215,
      "learning_rate": 6.182572614107885e-06,
      "loss": 0.0402,
      "step": 460
    },
    {
      "epoch": 1.929460580912863,
      "grad_norm": 1.8334687948226929,
      "learning_rate": 6.1410788381742745e-06,
      "loss": 0.052,
      "step": 465
    },
    {
      "epoch": 1.950207468879668,
      "grad_norm": 0.998278021812439,
      "learning_rate": 6.099585062240665e-06,
      "loss": 0.0329,
      "step": 470
    },
    {
      "epoch": 1.9709543568464731,
      "grad_norm": 1.0715217590332031,
      "learning_rate": 6.058091286307054e-06,
      "loss": 0.0171,
      "step": 475
    },
    {
      "epoch": 1.991701244813278,
      "grad_norm": 2.7846434116363525,
      "learning_rate": 6.016597510373444e-06,
      "loss": 0.0454,
      "step": 480
    },
    {
      "epoch": 2.012448132780083,
      "grad_norm": 1.6487773656845093,
      "learning_rate": 5.9751037344398345e-06,
      "loss": 0.0283,
      "step": 485
    },
    {
      "epoch": 2.033195020746888,
      "grad_norm": 1.9173800945281982,
      "learning_rate": 5.933609958506224e-06,
      "loss": 0.0156,
      "step": 490
    },
    {
      "epoch": 2.0539419087136928,
      "grad_norm": 4.356449604034424,
      "learning_rate": 5.892116182572615e-06,
      "loss": 0.0357,
      "step": 495
    },
    {
      "epoch": 2.074688796680498,
      "grad_norm": 1.495988368988037,
      "learning_rate": 5.850622406639005e-06,
      "loss": 0.0112,
      "step": 500
    },
    {
      "epoch": 2.095435684647303,
      "grad_norm": 1.7308920621871948,
      "learning_rate": 5.809128630705395e-06,
      "loss": 0.0204,
      "step": 505
    },
    {
      "epoch": 2.116182572614108,
      "grad_norm": 2.1147842407226562,
      "learning_rate": 5.767634854771784e-06,
      "loss": 0.0202,
      "step": 510
    },
    {
      "epoch": 2.136929460580913,
      "grad_norm": 1.7742974758148193,
      "learning_rate": 5.726141078838174e-06,
      "loss": 0.0246,
      "step": 515
    },
    {
      "epoch": 2.1576763485477177,
      "grad_norm": 0.8728792667388916,
      "learning_rate": 5.684647302904565e-06,
      "loss": 0.0404,
      "step": 520
    },
    {
      "epoch": 2.1784232365145226,
      "grad_norm": 1.270905613899231,
      "learning_rate": 5.6431535269709545e-06,
      "loss": 0.013,
      "step": 525
    },
    {
      "epoch": 2.199170124481328,
      "grad_norm": 1.3951818943023682,
      "learning_rate": 5.601659751037345e-06,
      "loss": 0.0171,
      "step": 530
    },
    {
      "epoch": 2.219917012448133,
      "grad_norm": 1.7056889533996582,
      "learning_rate": 5.560165975103735e-06,
      "loss": 0.0178,
      "step": 535
    },
    {
      "epoch": 2.240663900414938,
      "grad_norm": 0.9322335720062256,
      "learning_rate": 5.5186721991701256e-06,
      "loss": 0.0132,
      "step": 540
    },
    {
      "epoch": 2.2614107883817427,
      "grad_norm": 1.3173096179962158,
      "learning_rate": 5.477178423236515e-06,
      "loss": 0.0236,
      "step": 545
    },
    {
      "epoch": 2.2821576763485476,
      "grad_norm": 1.268327236175537,
      "learning_rate": 5.435684647302904e-06,
      "loss": 0.0125,
      "step": 550
    },
    {
      "epoch": 2.3029045643153525,
      "grad_norm": 1.9482656717300415,
      "learning_rate": 5.394190871369295e-06,
      "loss": 0.022,
      "step": 555
    },
    {
      "epoch": 2.323651452282158,
      "grad_norm": 0.6962999701499939,
      "learning_rate": 5.352697095435685e-06,
      "loss": 0.0128,
      "step": 560
    },
    {
      "epoch": 2.3443983402489628,
      "grad_norm": 0.6124351024627686,
      "learning_rate": 5.311203319502075e-06,
      "loss": 0.0137,
      "step": 565
    },
    {
      "epoch": 2.3651452282157677,
      "grad_norm": 2.1399357318878174,
      "learning_rate": 5.269709543568465e-06,
      "loss": 0.0173,
      "step": 570
    },
    {
      "epoch": 2.3858921161825726,
      "grad_norm": 1.7154192924499512,
      "learning_rate": 5.228215767634856e-06,
      "loss": 0.02,
      "step": 575
    },
    {
      "epoch": 2.4066390041493775,
      "grad_norm": 3.543788194656372,
      "learning_rate": 5.1867219917012455e-06,
      "loss": 0.0333,
      "step": 580
    },
    {
      "epoch": 2.4273858921161824,
      "grad_norm": 1.5438944101333618,
      "learning_rate": 5.1452282157676345e-06,
      "loss": 0.0256,
      "step": 585
    },
    {
      "epoch": 2.4481327800829877,
      "grad_norm": 1.4733091592788696,
      "learning_rate": 5.103734439834025e-06,
      "loss": 0.0232,
      "step": 590
    },
    {
      "epoch": 2.4688796680497926,
      "grad_norm": 0.5347623825073242,
      "learning_rate": 5.062240663900415e-06,
      "loss": 0.0233,
      "step": 595
    },
    {
      "epoch": 2.4896265560165975,
      "grad_norm": 0.7176046967506409,
      "learning_rate": 5.0207468879668055e-06,
      "loss": 0.0235,
      "step": 600
    },
    {
      "epoch": 2.5103734439834025,
      "grad_norm": 1.4965711832046509,
      "learning_rate": 4.979253112033195e-06,
      "loss": 0.0192,
      "step": 605
    },
    {
      "epoch": 2.5311203319502074,
      "grad_norm": 1.411422610282898,
      "learning_rate": 4.937759336099586e-06,
      "loss": 0.0111,
      "step": 610
    },
    {
      "epoch": 2.5518672199170123,
      "grad_norm": 1.307692050933838,
      "learning_rate": 4.896265560165976e-06,
      "loss": 0.0268,
      "step": 615
    },
    {
      "epoch": 2.572614107883817,
      "grad_norm": 1.2126792669296265,
      "learning_rate": 4.8547717842323655e-06,
      "loss": 0.0119,
      "step": 620
    },
    {
      "epoch": 2.5933609958506225,
      "grad_norm": 0.5013401508331299,
      "learning_rate": 4.813278008298755e-06,
      "loss": 0.0151,
      "step": 625
    },
    {
      "epoch": 2.6141078838174274,
      "grad_norm": 0.564896821975708,
      "learning_rate": 4.771784232365146e-06,
      "loss": 0.0253,
      "step": 630
    },
    {
      "epoch": 2.6348547717842323,
      "grad_norm": 2.106884241104126,
      "learning_rate": 4.730290456431536e-06,
      "loss": 0.0274,
      "step": 635
    },
    {
      "epoch": 2.6556016597510372,
      "grad_norm": 1.5650396347045898,
      "learning_rate": 4.6887966804979255e-06,
      "loss": 0.0186,
      "step": 640
    },
    {
      "epoch": 2.6763485477178426,
      "grad_norm": 1.710783839225769,
      "learning_rate": 4.647302904564316e-06,
      "loss": 0.0158,
      "step": 645
    },
    {
      "epoch": 2.6970954356846475,
      "grad_norm": 1.9014557600021362,
      "learning_rate": 4.605809128630706e-06,
      "loss": 0.0337,
      "step": 650
    },
    {
      "epoch": 2.7178423236514524,
      "grad_norm": 1.1175166368484497,
      "learning_rate": 4.564315352697096e-06,
      "loss": 0.0094,
      "step": 655
    },
    {
      "epoch": 2.7385892116182573,
      "grad_norm": 1.1134710311889648,
      "learning_rate": 4.5228215767634855e-06,
      "loss": 0.0236,
      "step": 660
    },
    {
      "epoch": 2.759336099585062,
      "grad_norm": 1.3698573112487793,
      "learning_rate": 4.481327800829876e-06,
      "loss": 0.0271,
      "step": 665
    },
    {
      "epoch": 2.780082987551867,
      "grad_norm": 1.1698119640350342,
      "learning_rate": 4.439834024896266e-06,
      "loss": 0.0074,
      "step": 670
    },
    {
      "epoch": 2.800829875518672,
      "grad_norm": 1.1777998208999634,
      "learning_rate": 4.398340248962656e-06,
      "loss": 0.0082,
      "step": 675
    },
    {
      "epoch": 2.821576763485477,
      "grad_norm": 1.6402781009674072,
      "learning_rate": 4.356846473029046e-06,
      "loss": 0.0134,
      "step": 680
    },
    {
      "epoch": 2.8423236514522823,
      "grad_norm": 1.0831212997436523,
      "learning_rate": 4.315352697095436e-06,
      "loss": 0.0253,
      "step": 685
    },
    {
      "epoch": 2.863070539419087,
      "grad_norm": 2.4165401458740234,
      "learning_rate": 4.273858921161826e-06,
      "loss": 0.023,
      "step": 690
    },
    {
      "epoch": 2.883817427385892,
      "grad_norm": 2.052992820739746,
      "learning_rate": 4.232365145228216e-06,
      "loss": 0.0206,
      "step": 695
    },
    {
      "epoch": 2.904564315352697,
      "grad_norm": 2.2030189037323,
      "learning_rate": 4.190871369294606e-06,
      "loss": 0.0119,
      "step": 700
    },
    {
      "epoch": 2.9253112033195023,
      "grad_norm": 2.6307854652404785,
      "learning_rate": 4.149377593360996e-06,
      "loss": 0.0241,
      "step": 705
    },
    {
      "epoch": 2.9460580912863072,
      "grad_norm": 1.8030495643615723,
      "learning_rate": 4.107883817427386e-06,
      "loss": 0.0237,
      "step": 710
    },
    {
      "epoch": 2.966804979253112,
      "grad_norm": 1.1759185791015625,
      "learning_rate": 4.0663900414937765e-06,
      "loss": 0.011,
      "step": 715
    },
    {
      "epoch": 2.987551867219917,
      "grad_norm": 1.6851730346679688,
      "learning_rate": 4.024896265560166e-06,
      "loss": 0.024,
      "step": 720
    },
    {
      "epoch": 3.008298755186722,
      "grad_norm": 2.09350848197937,
      "learning_rate": 3.983402489626556e-06,
      "loss": 0.0176,
      "step": 725
    },
    {
      "epoch": 3.029045643153527,
      "grad_norm": 0.5645933151245117,
      "learning_rate": 3.941908713692946e-06,
      "loss": 0.0076,
      "step": 730
    },
    {
      "epoch": 3.0497925311203318,
      "grad_norm": 1.5976221561431885,
      "learning_rate": 3.9004149377593365e-06,
      "loss": 0.0069,
      "step": 735
    },
    {
      "epoch": 3.070539419087137,
      "grad_norm": 0.7920781373977661,
      "learning_rate": 3.858921161825726e-06,
      "loss": 0.0098,
      "step": 740
    },
    {
      "epoch": 3.091286307053942,
      "grad_norm": 1.1054158210754395,
      "learning_rate": 3.817427385892116e-06,
      "loss": 0.0115,
      "step": 745
    },
    {
      "epoch": 3.112033195020747,
      "grad_norm": 1.7089980840682983,
      "learning_rate": 3.7759336099585063e-06,
      "loss": 0.0104,
      "step": 750
    },
    {
      "epoch": 3.132780082987552,
      "grad_norm": 0.7278354167938232,
      "learning_rate": 3.7344398340248965e-06,
      "loss": 0.0082,
      "step": 755
    },
    {
      "epoch": 3.1535269709543567,
      "grad_norm": 1.3134007453918457,
      "learning_rate": 3.6929460580912867e-06,
      "loss": 0.0245,
      "step": 760
    },
    {
      "epoch": 3.1742738589211617,
      "grad_norm": 1.0408704280853271,
      "learning_rate": 3.6514522821576765e-06,
      "loss": 0.0194,
      "step": 765
    },
    {
      "epoch": 3.195020746887967,
      "grad_norm": 1.0514349937438965,
      "learning_rate": 3.6099585062240667e-06,
      "loss": 0.0081,
      "step": 770
    },
    {
      "epoch": 3.215767634854772,
      "grad_norm": 1.0959088802337646,
      "learning_rate": 3.568464730290457e-06,
      "loss": 0.0102,
      "step": 775
    },
    {
      "epoch": 3.236514522821577,
      "grad_norm": 0.6447562575340271,
      "learning_rate": 3.526970954356847e-06,
      "loss": 0.013,
      "step": 780
    },
    {
      "epoch": 3.2572614107883817,
      "grad_norm": 1.8117377758026123,
      "learning_rate": 3.4854771784232365e-06,
      "loss": 0.0084,
      "step": 785
    },
    {
      "epoch": 3.2780082987551866,
      "grad_norm": 1.082808256149292,
      "learning_rate": 3.4439834024896267e-06,
      "loss": 0.0135,
      "step": 790
    },
    {
      "epoch": 3.2987551867219915,
      "grad_norm": 1.179017186164856,
      "learning_rate": 3.402489626556017e-06,
      "loss": 0.0162,
      "step": 795
    },
    {
      "epoch": 3.3195020746887964,
      "grad_norm": 1.0479273796081543,
      "learning_rate": 3.360995850622407e-06,
      "loss": 0.007,
      "step": 800
    },
    {
      "epoch": 3.340248962655602,
      "grad_norm": 2.8797686100006104,
      "learning_rate": 3.319502074688797e-06,
      "loss": 0.0153,
      "step": 805
    },
    {
      "epoch": 3.3609958506224067,
      "grad_norm": 1.5719329118728638,
      "learning_rate": 3.278008298755187e-06,
      "loss": 0.0212,
      "step": 810
    },
    {
      "epoch": 3.3817427385892116,
      "grad_norm": 0.8850643634796143,
      "learning_rate": 3.2365145228215773e-06,
      "loss": 0.0146,
      "step": 815
    },
    {
      "epoch": 3.4024896265560165,
      "grad_norm": 0.3604283928871155,
      "learning_rate": 3.1950207468879666e-06,
      "loss": 0.0045,
      "step": 820
    },
    {
      "epoch": 3.4232365145228214,
      "grad_norm": 1.391987919807434,
      "learning_rate": 3.153526970954357e-06,
      "loss": 0.0132,
      "step": 825
    },
    {
      "epoch": 3.4439834024896268,
      "grad_norm": 1.9832934141159058,
      "learning_rate": 3.112033195020747e-06,
      "loss": 0.0112,
      "step": 830
    },
    {
      "epoch": 3.4647302904564317,
      "grad_norm": 1.6392576694488525,
      "learning_rate": 3.0705394190871373e-06,
      "loss": 0.0126,
      "step": 835
    },
    {
      "epoch": 3.4854771784232366,
      "grad_norm": 1.6327385902404785,
      "learning_rate": 3.029045643153527e-06,
      "loss": 0.0141,
      "step": 840
    },
    {
      "epoch": 3.5062240663900415,
      "grad_norm": 0.5029474496841431,
      "learning_rate": 2.9875518672199173e-06,
      "loss": 0.0084,
      "step": 845
    },
    {
      "epoch": 3.5269709543568464,
      "grad_norm": 0.7686113715171814,
      "learning_rate": 2.9460580912863075e-06,
      "loss": 0.0073,
      "step": 850
    },
    {
      "epoch": 3.5477178423236513,
      "grad_norm": 0.3315852880477905,
      "learning_rate": 2.9045643153526977e-06,
      "loss": 0.0121,
      "step": 855
    },
    {
      "epoch": 3.568464730290456,
      "grad_norm": 0.39139634370803833,
      "learning_rate": 2.863070539419087e-06,
      "loss": 0.0094,
      "step": 860
    },
    {
      "epoch": 3.5892116182572615,
      "grad_norm": 0.7294654846191406,
      "learning_rate": 2.8215767634854773e-06,
      "loss": 0.0102,
      "step": 865
    },
    {
      "epoch": 3.6099585062240664,
      "grad_norm": 2.1243228912353516,
      "learning_rate": 2.7800829875518675e-06,
      "loss": 0.0116,
      "step": 870
    },
    {
      "epoch": 3.6307053941908713,
      "grad_norm": 1.2980406284332275,
      "learning_rate": 2.7385892116182577e-06,
      "loss": 0.0183,
      "step": 875
    },
    {
      "epoch": 3.6514522821576763,
      "grad_norm": 1.274969458580017,
      "learning_rate": 2.6970954356846475e-06,
      "loss": 0.0079,
      "step": 880
    },
    {
      "epoch": 3.6721991701244816,
      "grad_norm": 0.6039006114006042,
      "learning_rate": 2.6556016597510377e-06,
      "loss": 0.0092,
      "step": 885
    },
    {
      "epoch": 3.6929460580912865,
      "grad_norm": 0.6272030472755432,
      "learning_rate": 2.614107883817428e-06,
      "loss": 0.0132,
      "step": 890
    },
    {
      "epoch": 3.7136929460580914,
      "grad_norm": 1.31027352809906,
      "learning_rate": 2.5726141078838172e-06,
      "loss": 0.0113,
      "step": 895
    },
    {
      "epoch": 3.7344398340248963,
      "grad_norm": 0.6987027525901794,
      "learning_rate": 2.5311203319502074e-06,
      "loss": 0.0163,
      "step": 900
    },
    {
      "epoch": 3.7551867219917012,
      "grad_norm": 1.6568554639816284,
      "learning_rate": 2.4896265560165977e-06,
      "loss": 0.0148,
      "step": 905
    },
    {
      "epoch": 3.775933609958506,
      "grad_norm": 0.27225854992866516,
      "learning_rate": 2.448132780082988e-06,
      "loss": 0.0134,
      "step": 910
    },
    {
      "epoch": 3.796680497925311,
      "grad_norm": 0.7029077410697937,
      "learning_rate": 2.4066390041493776e-06,
      "loss": 0.0162,
      "step": 915
    },
    {
      "epoch": 3.817427385892116,
      "grad_norm": 1.0866529941558838,
      "learning_rate": 2.365145228215768e-06,
      "loss": 0.0149,
      "step": 920
    },
    {
      "epoch": 3.8381742738589213,
      "grad_norm": 1.3447505235671997,
      "learning_rate": 2.323651452282158e-06,
      "loss": 0.0175,
      "step": 925
    },
    {
      "epoch": 3.858921161825726,
      "grad_norm": 0.5825247764587402,
      "learning_rate": 2.282157676348548e-06,
      "loss": 0.0116,
      "step": 930
    },
    {
      "epoch": 3.879668049792531,
      "grad_norm": 0.18176421523094177,
      "learning_rate": 2.240663900414938e-06,
      "loss": 0.0144,
      "step": 935
    },
    {
      "epoch": 3.900414937759336,
      "grad_norm": 1.488423466682434,
      "learning_rate": 2.199170124481328e-06,
      "loss": 0.013,
      "step": 940
    },
    {
      "epoch": 3.921161825726141,
      "grad_norm": 0.989838719367981,
      "learning_rate": 2.157676348547718e-06,
      "loss": 0.0089,
      "step": 945
    },
    {
      "epoch": 3.9419087136929463,
      "grad_norm": 1.9489572048187256,
      "learning_rate": 2.116182572614108e-06,
      "loss": 0.0117,
      "step": 950
    },
    {
      "epoch": 3.962655601659751,
      "grad_norm": 0.9473742246627808,
      "learning_rate": 2.074688796680498e-06,
      "loss": 0.0194,
      "step": 955
    },
    {
      "epoch": 3.983402489626556,
      "grad_norm": 0.42608770728111267,
      "learning_rate": 2.0331950207468883e-06,
      "loss": 0.0091,
      "step": 960
    },
    {
      "epoch": 4.004149377593361,
      "grad_norm": 1.3207372426986694,
      "learning_rate": 1.991701244813278e-06,
      "loss": 0.0062,
      "step": 965
    },
    {
      "epoch": 4.024896265560166,
      "grad_norm": 0.7345901131629944,
      "learning_rate": 1.9502074688796682e-06,
      "loss": 0.0042,
      "step": 970
    },
    {
      "epoch": 4.045643153526971,
      "grad_norm": 0.773199737071991,
      "learning_rate": 1.908713692946058e-06,
      "loss": 0.0126,
      "step": 975
    },
    {
      "epoch": 4.066390041493776,
      "grad_norm": 1.3770180940628052,
      "learning_rate": 1.8672199170124482e-06,
      "loss": 0.0139,
      "step": 980
    },
    {
      "epoch": 4.087136929460581,
      "grad_norm": 1.1472896337509155,
      "learning_rate": 1.8257261410788382e-06,
      "loss": 0.0123,
      "step": 985
    },
    {
      "epoch": 4.1078838174273855,
      "grad_norm": 1.593767762184143,
      "learning_rate": 1.7842323651452284e-06,
      "loss": 0.008,
      "step": 990
    },
    {
      "epoch": 4.12863070539419,
      "grad_norm": 1.0951461791992188,
      "learning_rate": 1.7427385892116182e-06,
      "loss": 0.0068,
      "step": 995
    },
    {
      "epoch": 4.149377593360996,
      "grad_norm": 0.7573041915893555,
      "learning_rate": 1.7012448132780084e-06,
      "loss": 0.0052,
      "step": 1000
    },
    {
      "epoch": 4.170124481327801,
      "grad_norm": 0.18247643113136292,
      "learning_rate": 1.6597510373443984e-06,
      "loss": 0.0047,
      "step": 1005
    },
    {
      "epoch": 4.190871369294606,
      "grad_norm": 1.1145219802856445,
      "learning_rate": 1.6182572614107886e-06,
      "loss": 0.0048,
      "step": 1010
    },
    {
      "epoch": 4.211618257261411,
      "grad_norm": 0.7710907459259033,
      "learning_rate": 1.5767634854771784e-06,
      "loss": 0.0044,
      "step": 1015
    },
    {
      "epoch": 4.232365145228216,
      "grad_norm": 0.5366164445877075,
      "learning_rate": 1.5352697095435686e-06,
      "loss": 0.0153,
      "step": 1020
    },
    {
      "epoch": 4.253112033195021,
      "grad_norm": 1.7926034927368164,
      "learning_rate": 1.4937759336099586e-06,
      "loss": 0.0163,
      "step": 1025
    },
    {
      "epoch": 4.273858921161826,
      "grad_norm": 0.5849465131759644,
      "learning_rate": 1.4522821576763488e-06,
      "loss": 0.004,
      "step": 1030
    },
    {
      "epoch": 4.2946058091286305,
      "grad_norm": 0.5547875165939331,
      "learning_rate": 1.4107883817427386e-06,
      "loss": 0.0042,
      "step": 1035
    },
    {
      "epoch": 4.3153526970954355,
      "grad_norm": 0.7352296710014343,
      "learning_rate": 1.3692946058091288e-06,
      "loss": 0.0071,
      "step": 1040
    },
    {
      "epoch": 4.33609958506224,
      "grad_norm": 0.5947510600090027,
      "learning_rate": 1.3278008298755188e-06,
      "loss": 0.008,
      "step": 1045
    },
    {
      "epoch": 4.356846473029045,
      "grad_norm": 0.5303782820701599,
      "learning_rate": 1.2863070539419086e-06,
      "loss": 0.0055,
      "step": 1050
    },
    {
      "epoch": 4.377593360995851,
      "grad_norm": 0.8492805361747742,
      "learning_rate": 1.2448132780082988e-06,
      "loss": 0.005,
      "step": 1055
    },
    {
      "epoch": 4.398340248962656,
      "grad_norm": 1.19044029712677,
      "learning_rate": 1.2033195020746888e-06,
      "loss": 0.013,
      "step": 1060
    },
    {
      "epoch": 4.419087136929461,
      "grad_norm": 0.9511287808418274,
      "learning_rate": 1.161825726141079e-06,
      "loss": 0.0127,
      "step": 1065
    },
    {
      "epoch": 4.439834024896266,
      "grad_norm": 0.23320679366588593,
      "learning_rate": 1.120331950207469e-06,
      "loss": 0.0059,
      "step": 1070
    },
    {
      "epoch": 4.460580912863071,
      "grad_norm": 1.2488551139831543,
      "learning_rate": 1.078838174273859e-06,
      "loss": 0.0081,
      "step": 1075
    },
    {
      "epoch": 4.481327800829876,
      "grad_norm": 1.710477352142334,
      "learning_rate": 1.037344398340249e-06,
      "loss": 0.0076,
      "step": 1080
    },
    {
      "epoch": 4.5020746887966805,
      "grad_norm": 0.8615099191665649,
      "learning_rate": 9.95850622406639e-07,
      "loss": 0.012,
      "step": 1085
    },
    {
      "epoch": 4.522821576763485,
      "grad_norm": 1.6006779670715332,
      "learning_rate": 9.54356846473029e-07,
      "loss": 0.0098,
      "step": 1090
    },
    {
      "epoch": 4.54356846473029,
      "grad_norm": 0.5129662752151489,
      "learning_rate": 9.128630705394191e-07,
      "loss": 0.0129,
      "step": 1095
    },
    {
      "epoch": 4.564315352697095,
      "grad_norm": 0.8513924479484558,
      "learning_rate": 8.713692946058091e-07,
      "loss": 0.0089,
      "step": 1100
    },
    {
      "epoch": 4.5850622406639,
      "grad_norm": 0.8673096895217896,
      "learning_rate": 8.298755186721992e-07,
      "loss": 0.0055,
      "step": 1105
    },
    {
      "epoch": 4.605809128630705,
      "grad_norm": 1.8209664821624756,
      "learning_rate": 7.883817427385892e-07,
      "loss": 0.0138,
      "step": 1110
    },
    {
      "epoch": 4.62655601659751,
      "grad_norm": 0.4811139702796936,
      "learning_rate": 7.468879668049793e-07,
      "loss": 0.0049,
      "step": 1115
    },
    {
      "epoch": 4.647302904564316,
      "grad_norm": 0.6291790008544922,
      "learning_rate": 7.053941908713693e-07,
      "loss": 0.0103,
      "step": 1120
    },
    {
      "epoch": 4.668049792531121,
      "grad_norm": 1.172275185585022,
      "learning_rate": 6.639004149377594e-07,
      "loss": 0.0066,
      "step": 1125
    },
    {
      "epoch": 4.6887966804979255,
      "grad_norm": 0.7825120091438293,
      "learning_rate": 6.224066390041494e-07,
      "loss": 0.0106,
      "step": 1130
    },
    {
      "epoch": 4.70954356846473,
      "grad_norm": 0.1047833114862442,
      "learning_rate": 5.809128630705395e-07,
      "loss": 0.005,
      "step": 1135
    },
    {
      "epoch": 4.730290456431535,
      "grad_norm": 1.2170310020446777,
      "learning_rate": 5.394190871369295e-07,
      "loss": 0.0089,
      "step": 1140
    },
    {
      "epoch": 4.75103734439834,
      "grad_norm": 0.9620624780654907,
      "learning_rate": 4.979253112033195e-07,
      "loss": 0.0058,
      "step": 1145
    },
    {
      "epoch": 4.771784232365145,
      "grad_norm": 1.189684510231018,
      "learning_rate": 4.5643153526970956e-07,
      "loss": 0.0051,
      "step": 1150
    },
    {
      "epoch": 4.79253112033195,
      "grad_norm": 1.586207628250122,
      "learning_rate": 4.149377593360996e-07,
      "loss": 0.0125,
      "step": 1155
    },
    {
      "epoch": 4.813278008298755,
      "grad_norm": 0.47333458065986633,
      "learning_rate": 3.7344398340248966e-07,
      "loss": 0.0096,
      "step": 1160
    },
    {
      "epoch": 4.83402489626556,
      "grad_norm": 0.5717390775680542,
      "learning_rate": 3.319502074688797e-07,
      "loss": 0.0053,
      "step": 1165
    },
    {
      "epoch": 4.854771784232365,
      "grad_norm": 0.6358240842819214,
      "learning_rate": 2.9045643153526976e-07,
      "loss": 0.0054,
      "step": 1170
    },
    {
      "epoch": 4.875518672199171,
      "grad_norm": 0.7031179070472717,
      "learning_rate": 2.4896265560165975e-07,
      "loss": 0.0159,
      "step": 1175
    },
    {
      "epoch": 4.8962655601659755,
      "grad_norm": 1.2947261333465576,
      "learning_rate": 2.074688796680498e-07,
      "loss": 0.01,
      "step": 1180
    },
    {
      "epoch": 4.91701244813278,
      "grad_norm": 0.93119877576828,
      "learning_rate": 1.6597510373443985e-07,
      "loss": 0.0073,
      "step": 1185
    },
    {
      "epoch": 4.937759336099585,
      "grad_norm": 1.0012972354888916,
      "learning_rate": 1.2448132780082988e-07,
      "loss": 0.0055,
      "step": 1190
    },
    {
      "epoch": 4.95850622406639,
      "grad_norm": 1.2632884979248047,
      "learning_rate": 8.298755186721993e-08,
      "loss": 0.0075,
      "step": 1195
    },
    {
      "epoch": 4.979253112033195,
      "grad_norm": 3.3445241451263428,
      "learning_rate": 4.1493775933609963e-08,
      "loss": 0.0119,
      "step": 1200
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.40131083130836487,
      "learning_rate": 0.0,
      "loss": 0.0107,
      "step": 1205
    }
  ],
  "logging_steps": 5,
  "max_steps": 1205,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.92084187136e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}