{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 550,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01818181818181818,
      "grad_norm": 251.0,
      "learning_rate": 3.636363636363636e-06,
      "loss": 46.8778,
      "step": 1
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 239.0,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 46.8224,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 106.5,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 39.5877,
      "step": 10
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 19.75,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 28.2487,
      "step": 15
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 15.8125,
      "learning_rate": 7.272727272727273e-05,
      "loss": 23.0407,
      "step": 20
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 5.5,
      "learning_rate": 9.090909090909092e-05,
      "loss": 20.515,
      "step": 25
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 4.1875,
      "learning_rate": 0.00010909090909090909,
      "loss": 19.5202,
      "step": 30
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 5.75,
      "learning_rate": 0.00012727272727272728,
      "loss": 18.2599,
      "step": 35
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 10.8125,
      "learning_rate": 0.00014545454545454546,
      "loss": 16.923,
      "step": 40
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 21.25,
      "learning_rate": 0.00016363636363636366,
      "loss": 14.3261,
      "step": 45
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 29.625,
      "learning_rate": 0.00018181818181818183,
      "loss": 9.0683,
      "step": 50
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.625,
      "learning_rate": 0.0002,
      "loss": 3.5798,
      "step": 55
    },
    {
      "epoch": 1.0,
      "eval_loss": 4.059033393859863,
      "eval_runtime": 0.2418,
      "eval_samples_per_second": 41.356,
      "eval_steps_per_second": 4.136,
      "step": 55
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 10.125,
      "learning_rate": 0.00019994965423831854,
      "loss": 2.5621,
      "step": 60
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 7.4375,
      "learning_rate": 0.00019979866764718843,
      "loss": 2.0701,
      "step": 65
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 20.125,
      "learning_rate": 0.00019954719225730847,
      "loss": 1.8568,
      "step": 70
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.96875,
      "learning_rate": 0.00019919548128307954,
      "loss": 1.6836,
      "step": 75
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 1.078125,
      "learning_rate": 0.00019874388886763944,
      "loss": 1.5757,
      "step": 80
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 1.171875,
      "learning_rate": 0.00019819286972627066,
      "loss": 1.52,
      "step": 85
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 1.3125,
      "learning_rate": 0.00019754297868854073,
      "loss": 1.4757,
      "step": 90
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 3.46875,
      "learning_rate": 0.00019679487013963564,
      "loss": 1.4221,
      "step": 95
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00019594929736144976,
      "loss": 1.4012,
      "step": 100
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 1.671875,
      "learning_rate": 0.00019500711177409454,
      "loss": 1.3672,
      "step": 105
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00019396926207859084,
      "loss": 1.346,
      "step": 110
    },
    {
      "epoch": 2.0,
      "eval_loss": 2.6943764686584473,
      "eval_runtime": 0.2333,
      "eval_samples_per_second": 42.871,
      "eval_steps_per_second": 4.287,
      "step": 110
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 1.484375,
      "learning_rate": 0.00019283679330160726,
      "loss": 1.3092,
      "step": 115
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00019161084574320696,
      "loss": 1.3052,
      "step": 120
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.625,
      "learning_rate": 0.00019029265382866214,
      "loss": 1.2724,
      "step": 125
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00018888354486549237,
      "loss": 1.2612,
      "step": 130
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00018738493770697852,
      "loss": 1.2532,
      "step": 135
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 4.25,
      "learning_rate": 0.00018579834132349772,
      "loss": 1.2422,
      "step": 140
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00018412535328311814,
      "loss": 1.2256,
      "step": 145
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 2.421875,
      "learning_rate": 0.0001823676581429833,
      "loss": 1.2271,
      "step": 150
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00018052702575310588,
      "loss": 1.2114,
      "step": 155
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00017860530947427875,
      "loss": 1.2068,
      "step": 160
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.6875,
      "learning_rate": 0.0001766044443118978,
      "loss": 1.1944,
      "step": 165
    },
    {
      "epoch": 3.0,
      "eval_loss": 2.6026840209960938,
      "eval_runtime": 0.2422,
      "eval_samples_per_second": 41.281,
      "eval_steps_per_second": 4.128,
      "step": 165
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0001745264449675755,
      "loss": 1.194,
      "step": 170
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 3.515625,
      "learning_rate": 0.00017237340381050703,
      "loss": 1.1668,
      "step": 175
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 1.59375,
      "learning_rate": 0.00017014748877063214,
      "loss": 1.1693,
      "step": 180
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 2.109375,
      "learning_rate": 0.00016785094115571322,
      "loss": 1.167,
      "step": 185
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00016548607339452853,
      "loss": 1.1601,
      "step": 190
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00016305526670845226,
      "loss": 1.1343,
      "step": 195
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 1.265625,
      "learning_rate": 0.00016056096871376667,
      "loss": 1.1379,
      "step": 200
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00015800569095711982,
      "loss": 1.1339,
      "step": 205
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00015539200638661104,
      "loss": 1.1393,
      "step": 210
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.7421875,
      "learning_rate": 0.00015272254676105025,
      "loss": 1.1015,
      "step": 215
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00015000000000000001,
      "loss": 1.1119,
      "step": 220
    },
    {
      "epoch": 4.0,
      "eval_loss": 2.5778679847717285,
      "eval_runtime": 0.2336,
      "eval_samples_per_second": 42.816,
      "eval_steps_per_second": 4.282,
      "step": 220
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 1.5625,
      "learning_rate": 0.0001472271074772683,
      "loss": 1.113,
      "step": 225
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00014440666126057744,
      "loss": 1.0963,
      "step": 230
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00014154150130018866,
      "loss": 1.104,
      "step": 235
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00013863451256931287,
      "loss": 1.0957,
      "step": 240
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 1.9375,
      "learning_rate": 0.00013568862215918717,
      "loss": 1.0835,
      "step": 245
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00013270679633174218,
      "loss": 1.0802,
      "step": 250
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0001296920375328275,
      "loss": 1.0762,
      "step": 255
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00012664738136900348,
      "loss": 1.0773,
      "step": 260
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00012357589355094275,
      "loss": 1.082,
      "step": 265
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00012048066680651908,
      "loss": 1.0687,
      "step": 270
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00011736481776669306,
      "loss": 1.0741,
      "step": 275
    },
    {
      "epoch": 5.0,
      "eval_loss": 2.5495071411132812,
      "eval_runtime": 0.2341,
      "eval_samples_per_second": 42.713,
      "eval_steps_per_second": 4.271,
      "step": 275
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00011423148382732853,
      "loss": 1.0546,
      "step": 280
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00011108381999010111,
      "loss": 1.0566,
      "step": 285
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00010792499568567884,
      "loss": 1.0477,
      "step": 290
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.890625,
      "learning_rate": 0.00010475819158237425,
      "loss": 1.0483,
      "step": 295
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.546875,
      "learning_rate": 0.00010158659638348081,
      "loss": 1.0542,
      "step": 300
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.57421875,
      "learning_rate": 9.84134036165192e-05,
      "loss": 1.0437,
      "step": 305
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.76953125,
      "learning_rate": 9.524180841762577e-05,
      "loss": 1.0542,
      "step": 310
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.546875,
      "learning_rate": 9.207500431432115e-05,
      "loss": 1.0424,
      "step": 315
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.494140625,
      "learning_rate": 8.891618000989891e-05,
      "loss": 1.0356,
      "step": 320
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.58203125,
      "learning_rate": 8.57685161726715e-05,
      "loss": 1.0432,
      "step": 325
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.55078125,
      "learning_rate": 8.263518223330697e-05,
      "loss": 1.0435,
      "step": 330
    },
    {
      "epoch": 6.0,
      "eval_loss": 2.5500569343566895,
      "eval_runtime": 0.2352,
      "eval_samples_per_second": 42.524,
      "eval_steps_per_second": 4.252,
      "step": 330
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.59375,
      "learning_rate": 7.951933319348095e-05,
      "loss": 1.0185,
      "step": 335
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.58203125,
      "learning_rate": 7.642410644905726e-05,
      "loss": 1.0249,
      "step": 340
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.6171875,
      "learning_rate": 7.335261863099651e-05,
      "loss": 1.0206,
      "step": 345
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.6015625,
      "learning_rate": 7.030796246717255e-05,
      "loss": 1.0209,
      "step": 350
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.5859375,
      "learning_rate": 6.729320366825784e-05,
      "loss": 1.0285,
      "step": 355
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.75,
      "learning_rate": 6.431137784081282e-05,
      "loss": 1.0206,
      "step": 360
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.5859375,
      "learning_rate": 6.136548743068713e-05,
      "loss": 1.0245,
      "step": 365
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.5078125,
      "learning_rate": 5.845849869981137e-05,
      "loss": 1.0174,
      "step": 370
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.75,
      "learning_rate": 5.559333873942259e-05,
      "loss": 1.0206,
      "step": 375
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.828125,
      "learning_rate": 5.277289252273174e-05,
      "loss": 1.0244,
      "step": 380
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.83203125,
      "learning_rate": 5.000000000000002e-05,
      "loss": 1.0191,
      "step": 385
    },
    {
      "epoch": 7.0,
      "eval_loss": 2.5536184310913086,
      "eval_runtime": 0.2341,
      "eval_samples_per_second": 42.722,
      "eval_steps_per_second": 4.272,
      "step": 385
    },
    {
      "epoch": 7.090909090909091,
      "grad_norm": 0.7734375,
      "learning_rate": 4.727745323894976e-05,
      "loss": 1.011,
      "step": 390
    },
    {
      "epoch": 7.181818181818182,
      "grad_norm": 0.61328125,
      "learning_rate": 4.4607993613388976e-05,
      "loss": 1.0087,
      "step": 395
    },
    {
      "epoch": 7.2727272727272725,
      "grad_norm": 0.6875,
      "learning_rate": 4.19943090428802e-05,
      "loss": 1.0019,
      "step": 400
    },
    {
      "epoch": 7.363636363636363,
      "grad_norm": 0.7890625,
      "learning_rate": 3.943903128623335e-05,
      "loss": 1.0112,
      "step": 405
    },
    {
      "epoch": 7.454545454545454,
      "grad_norm": 0.58203125,
      "learning_rate": 3.694473329154778e-05,
      "loss": 1.0075,
      "step": 410
    },
    {
      "epoch": 7.545454545454545,
      "grad_norm": 0.5703125,
      "learning_rate": 3.45139266054715e-05,
      "loss": 1.0023,
      "step": 415
    },
    {
      "epoch": 7.636363636363637,
      "grad_norm": 0.8046875,
      "learning_rate": 3.21490588442868e-05,
      "loss": 1.0112,
      "step": 420
    },
    {
      "epoch": 7.7272727272727275,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9852511229367865e-05,
      "loss": 1.0069,
      "step": 425
    },
    {
      "epoch": 7.818181818181818,
      "grad_norm": 0.45703125,
      "learning_rate": 2.7626596189492983e-05,
      "loss": 1.002,
      "step": 430
    },
    {
      "epoch": 7.909090909090909,
      "grad_norm": 0.55078125,
      "learning_rate": 2.5473555032424533e-05,
      "loss": 1.0076,
      "step": 435
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.478515625,
      "learning_rate": 2.339555568810221e-05,
      "loss": 0.9965,
      "step": 440
    },
    {
      "epoch": 8.0,
      "eval_loss": 2.5604465007781982,
      "eval_runtime": 0.24,
      "eval_samples_per_second": 41.665,
      "eval_steps_per_second": 4.167,
      "step": 440
    },
    {
      "epoch": 8.090909090909092,
      "grad_norm": 0.61328125,
      "learning_rate": 2.139469052572127e-05,
      "loss": 1.0043,
      "step": 445
    },
    {
      "epoch": 8.181818181818182,
      "grad_norm": 0.58984375,
      "learning_rate": 1.947297424689414e-05,
      "loss": 0.9997,
      "step": 450
    },
    {
      "epoch": 8.272727272727273,
      "grad_norm": 0.57421875,
      "learning_rate": 1.763234185701673e-05,
      "loss": 0.9942,
      "step": 455
    },
    {
      "epoch": 8.363636363636363,
      "grad_norm": 0.5625,
      "learning_rate": 1.587464671688187e-05,
      "loss": 0.9952,
      "step": 460
    },
    {
      "epoch": 8.454545454545455,
      "grad_norm": 0.453125,
      "learning_rate": 1.4201658676502294e-05,
      "loss": 1.0037,
      "step": 465
    },
    {
      "epoch": 8.545454545454545,
      "grad_norm": 0.419921875,
      "learning_rate": 1.2615062293021507e-05,
      "loss": 1.0051,
      "step": 470
    },
    {
      "epoch": 8.636363636363637,
      "grad_norm": 0.484375,
      "learning_rate": 1.1116455134507664e-05,
      "loss": 0.9905,
      "step": 475
    },
    {
      "epoch": 8.727272727272727,
      "grad_norm": 0.44140625,
      "learning_rate": 9.707346171337894e-06,
      "loss": 0.9977,
      "step": 480
    },
    {
      "epoch": 8.818181818181818,
      "grad_norm": 0.43359375,
      "learning_rate": 8.38915425679304e-06,
      "loss": 0.9894,
      "step": 485
    },
    {
      "epoch": 8.909090909090908,
      "grad_norm": 0.470703125,
      "learning_rate": 7.163206698392744e-06,
      "loss": 1.005,
      "step": 490
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.42578125,
      "learning_rate": 6.030737921409169e-06,
      "loss": 0.9986,
      "step": 495
    },
    {
      "epoch": 9.0,
      "eval_loss": 2.5596535205841064,
      "eval_runtime": 0.2334,
      "eval_samples_per_second": 42.846,
      "eval_steps_per_second": 4.285,
      "step": 495
    },
    {
      "epoch": 9.090909090909092,
      "grad_norm": 0.455078125,
      "learning_rate": 4.992888225905468e-06,
      "loss": 0.9957,
      "step": 500
    },
    {
      "epoch": 9.181818181818182,
      "grad_norm": 0.47265625,
      "learning_rate": 4.050702638550275e-06,
      "loss": 1.0036,
      "step": 505
    },
    {
      "epoch": 9.272727272727273,
      "grad_norm": 0.462890625,
      "learning_rate": 3.2051298603643753e-06,
      "loss": 0.9985,
      "step": 510
    },
    {
      "epoch": 9.363636363636363,
      "grad_norm": 0.458984375,
      "learning_rate": 2.4570213114592954e-06,
      "loss": 0.9961,
      "step": 515
    },
    {
      "epoch": 9.454545454545455,
      "grad_norm": 0.443359375,
      "learning_rate": 1.8071302737293295e-06,
      "loss": 1.0066,
      "step": 520
    },
    {
      "epoch": 9.545454545454545,
      "grad_norm": 0.43359375,
      "learning_rate": 1.2561111323605712e-06,
      "loss": 0.9996,
      "step": 525
    },
    {
      "epoch": 9.636363636363637,
      "grad_norm": 0.439453125,
      "learning_rate": 8.04518716920466e-07,
      "loss": 0.9941,
      "step": 530
    },
    {
      "epoch": 9.727272727272727,
      "grad_norm": 0.419921875,
      "learning_rate": 4.5280774269154115e-07,
      "loss": 0.9898,
      "step": 535
    },
    {
      "epoch": 9.818181818181818,
      "grad_norm": 0.419921875,
      "learning_rate": 2.0133235281156736e-07,
      "loss": 0.9888,
      "step": 540
    },
    {
      "epoch": 9.909090909090908,
      "grad_norm": 0.4453125,
      "learning_rate": 5.0345761681491746e-08,
      "loss": 0.9953,
      "step": 545
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.41796875,
      "learning_rate": 0.0,
      "loss": 0.9948,
      "step": 550
    },
    {
      "epoch": 10.0,
      "eval_loss": 2.5602283477783203,
      "eval_runtime": 0.2428,
      "eval_samples_per_second": 41.182,
      "eval_steps_per_second": 4.118,
      "step": 550
    },
    {
      "epoch": 10.0,
      "step": 550,
      "total_flos": 1.6777423328808796e+18,
      "train_loss": 3.202145513187755,
      "train_runtime": 1331.6624,
      "train_samples_per_second": 26.313,
      "train_steps_per_second": 0.413
    }
  ],
  "logging_steps": 5,
  "max_steps": 550,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 100,
  "total_flos": 1.6777423328808796e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}