{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.96,
  "eval_steps": 500,
  "global_step": 2400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 14.472631454467773,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 5.6365,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": NaN,
      "learning_rate": 6e-06,
      "loss": 7.9556,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.6599702835083,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 6.5246,
      "step": 15
    },
    {
      "epoch": 0.01,
      "grad_norm": 40.07120895385742,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 7.7818,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.662575721740723,
      "learning_rate": 2.1e-05,
      "loss": 5.5899,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": NaN,
      "learning_rate": 2.5e-05,
      "loss": 7.1589,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.6021246910095215,
      "learning_rate": 3e-05,
      "loss": 4.9643,
      "step": 35
    },
    {
      "epoch": 0.02,
      "grad_norm": 7.397390365600586,
      "learning_rate": 3.5e-05,
      "loss": 5.944,
      "step": 40
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.122846603393555,
      "learning_rate": 4e-05,
      "loss": 4.2179,
      "step": 45
    },
    {
      "epoch": 0.02,
      "grad_norm": 7.520490646362305,
      "learning_rate": 4.5e-05,
      "loss": 3.9576,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.957700729370117,
      "learning_rate": 5e-05,
      "loss": 4.6752,
      "step": 55
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.252644538879395,
      "learning_rate": 4.999948617395915e-05,
      "loss": 4.586,
      "step": 60
    },
    {
      "epoch": 0.03,
      "grad_norm": 14.29404354095459,
      "learning_rate": 4.9997944716957985e-05,
      "loss": 4.2009,
      "step": 65
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.03260612487793,
      "learning_rate": 4.9995375692359755e-05,
      "loss": 4.7383,
      "step": 70
    },
    {
      "epoch": 0.03,
      "grad_norm": 7.339493751525879,
      "learning_rate": 4.9991779205767e-05,
      "loss": 4.0294,
      "step": 75
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.51794958114624,
      "learning_rate": 4.99871554050172e-05,
      "loss": 3.8225,
      "step": 80
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.195987701416016,
      "learning_rate": 4.9981504480176696e-05,
      "loss": 4.1299,
      "step": 85
    },
    {
      "epoch": 0.04,
      "grad_norm": 14.785408973693848,
      "learning_rate": 4.997482666353287e-05,
      "loss": 4.1754,
      "step": 90
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.586732864379883,
      "learning_rate": 4.996712222958461e-05,
      "loss": 3.2669,
      "step": 95
    },
    {
      "epoch": 0.04,
      "grad_norm": 25.427532196044922,
      "learning_rate": 4.9958391495031026e-05,
      "loss": 4.1074,
      "step": 100
    },
    {
      "epoch": 0.04,
      "grad_norm": 7.9541730880737305,
      "learning_rate": 4.994863481875841e-05,
      "loss": 4.2565,
      "step": 105
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.371864318847656,
      "learning_rate": 4.993785260182552e-05,
      "loss": 3.3137,
      "step": 110
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.015960693359375,
      "learning_rate": 4.992604528744705e-05,
      "loss": 3.4324,
      "step": 115
    },
    {
      "epoch": 0.05,
      "grad_norm": 6.063784599304199,
      "learning_rate": 4.991321336097546e-05,
      "loss": 3.3621,
      "step": 120
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.260740280151367,
      "learning_rate": 4.989935734988098e-05,
      "loss": 4.7443,
      "step": 125
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.744063377380371,
      "learning_rate": 4.9884477823729956e-05,
      "loss": 4.085,
      "step": 130
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.156418800354004,
      "learning_rate": 4.986857539416144e-05,
      "loss": 3.5387,
      "step": 135
    },
    {
      "epoch": 0.06,
      "grad_norm": 7.296212673187256,
      "learning_rate": 4.9851650714862006e-05,
      "loss": 3.348,
      "step": 140
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.042045593261719,
      "learning_rate": 4.983370448153896e-05,
      "loss": 3.5621,
      "step": 145
    },
    {
      "epoch": 0.06,
      "grad_norm": 12.138426780700684,
      "learning_rate": 4.981473743189163e-05,
      "loss": 3.4145,
      "step": 150
    },
    {
      "epoch": 0.06,
      "grad_norm": 12.3366060256958,
      "learning_rate": 4.979475034558115e-05,
      "loss": 3.9396,
      "step": 155
    },
    {
      "epoch": 0.06,
      "grad_norm": 6.696476936340332,
      "learning_rate": 4.977374404419837e-05,
      "loss": 3.5372,
      "step": 160
    },
    {
      "epoch": 0.07,
      "grad_norm": 6.023372173309326,
      "learning_rate": 4.975171939123005e-05,
      "loss": 3.2419,
      "step": 165
    },
    {
      "epoch": 0.07,
      "grad_norm": 7.861164093017578,
      "learning_rate": 4.9728677292023405e-05,
      "loss": 3.7565,
      "step": 170
    },
    {
      "epoch": 0.07,
      "grad_norm": 6.757652282714844,
      "learning_rate": 4.970461869374889e-05,
      "loss": 3.2182,
      "step": 175
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.434078216552734,
      "learning_rate": 4.967954458536126e-05,
      "loss": 3.229,
      "step": 180
    },
    {
      "epoch": 0.07,
      "grad_norm": 27.395694732666016,
      "learning_rate": 4.965345599755887e-05,
      "loss": 3.903,
      "step": 185
    },
    {
      "epoch": 0.08,
      "grad_norm": 6.09558629989624,
      "learning_rate": 4.962635400274142e-05,
      "loss": 3.1898,
      "step": 190
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.992712497711182,
      "learning_rate": 4.959823971496574e-05,
      "loss": 3.2185,
      "step": 195
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.1479411125183105,
      "learning_rate": 4.95691142899001e-05,
      "loss": 3.2792,
      "step": 200
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.431621551513672,
      "learning_rate": 4.9538978924776634e-05,
      "loss": 3.1418,
      "step": 205
    },
    {
      "epoch": 0.08,
      "grad_norm": 23.00255584716797,
      "learning_rate": 4.9507834858342186e-05,
      "loss": 3.9976,
      "step": 210
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.47463321685791,
      "learning_rate": 4.9475683370807326e-05,
      "loss": 3.4157,
      "step": 215
    },
    {
      "epoch": 0.09,
      "grad_norm": 18.375465393066406,
      "learning_rate": 4.9442525783793794e-05,
      "loss": 3.5929,
      "step": 220
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.78403091430664,
      "learning_rate": 4.940836346028011e-05,
      "loss": 3.1698,
      "step": 225
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.213769912719727,
      "learning_rate": 4.937319780454559e-05,
      "loss": 3.9643,
      "step": 230
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.919612884521484,
      "learning_rate": 4.933703026211262e-05,
      "loss": 3.6494,
      "step": 235
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.598107814788818,
      "learning_rate": 4.9299862319687204e-05,
      "loss": 3.5917,
      "step": 240
    },
    {
      "epoch": 0.1,
      "grad_norm": 37.90977096557617,
      "learning_rate": 4.926169550509787e-05,
      "loss": 3.7991,
      "step": 245
    },
    {
      "epoch": 0.1,
      "grad_norm": 12.698101997375488,
      "learning_rate": 4.9222531387232885e-05,
      "loss": 3.5116,
      "step": 250
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.609795570373535,
      "learning_rate": 4.9182371575975736e-05,
      "loss": 3.4126,
      "step": 255
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.108003616333008,
      "learning_rate": 4.914121772213898e-05,
      "loss": 3.2962,
      "step": 260
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.94726848602295,
      "learning_rate": 4.909907151739633e-05,
      "loss": 3.6953,
      "step": 265
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.730710983276367,
      "learning_rate": 4.905593469421323e-05,
      "loss": 3.4841,
      "step": 270
    },
    {
      "epoch": 0.11,
      "grad_norm": 6.423217296600342,
      "learning_rate": 4.9011809025775486e-05,
      "loss": 3.1157,
      "step": 275
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.736145973205566,
      "learning_rate": 4.8966696325916515e-05,
      "loss": 3.244,
      "step": 280
    },
    {
      "epoch": 0.11,
      "grad_norm": 21.597434997558594,
      "learning_rate": 4.892059844904272e-05,
      "loss": 3.172,
      "step": 285
    },
    {
      "epoch": 0.12,
      "grad_norm": 23.540430068969727,
      "learning_rate": 4.887351729005726e-05,
      "loss": 3.1889,
      "step": 290
    },
    {
      "epoch": 0.12,
      "grad_norm": 11.286609649658203,
      "learning_rate": 4.882545478428218e-05,
      "loss": 3.5211,
      "step": 295
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.204074382781982,
      "learning_rate": 4.877641290737884e-05,
      "loss": 3.255,
      "step": 300
    },
    {
      "epoch": 0.12,
      "grad_norm": 25.35500717163086,
      "learning_rate": 4.8726393675266716e-05,
      "loss": 3.4832,
      "step": 305
    },
    {
      "epoch": 0.12,
      "grad_norm": 6.76464319229126,
      "learning_rate": 4.8675399144040537e-05,
      "loss": 2.9075,
      "step": 310
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.19654655456543,
      "learning_rate": 4.862343140988573e-05,
      "loss": 3.0107,
      "step": 315
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.545323848724365,
      "learning_rate": 4.8570492608992325e-05,
      "loss": 2.7479,
      "step": 320
    },
    {
      "epoch": 0.13,
      "grad_norm": 15.334161758422852,
      "learning_rate": 4.851658491746707e-05,
      "loss": 2.8654,
      "step": 325
    },
    {
      "epoch": 0.13,
      "grad_norm": 14.28708553314209,
      "learning_rate": 4.846171055124401e-05,
      "loss": 3.1851,
      "step": 330
    },
    {
      "epoch": 0.13,
      "grad_norm": 22.403162002563477,
      "learning_rate": 4.8405871765993433e-05,
      "loss": 2.8862,
      "step": 335
    },
    {
      "epoch": 0.14,
      "grad_norm": 5.840334415435791,
      "learning_rate": 4.834907085702908e-05,
      "loss": 2.968,
      "step": 340
    },
    {
      "epoch": 0.14,
      "grad_norm": 12.307807922363281,
      "learning_rate": 4.829131015921385e-05,
      "loss": 2.9458,
      "step": 345
    },
    {
      "epoch": 0.14,
      "grad_norm": 12.92584228515625,
      "learning_rate": 4.82325920468638e-05,
      "loss": 3.4169,
      "step": 350
    },
    {
      "epoch": 0.14,
      "grad_norm": 13.389299392700195,
      "learning_rate": 4.817291893365055e-05,
      "loss": 3.1601,
      "step": 355
    },
    {
      "epoch": 0.14,
      "grad_norm": 16.73059844970703,
      "learning_rate": 4.8112293272502043e-05,
      "loss": 3.2846,
      "step": 360
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.608989715576172,
      "learning_rate": 4.805071755550177e-05,
      "loss": 3.0091,
      "step": 365
    },
    {
      "epoch": 0.15,
      "grad_norm": 5.731265544891357,
      "learning_rate": 4.7988194313786275e-05,
      "loss": 3.3688,
      "step": 370
    },
    {
      "epoch": 0.15,
      "grad_norm": 11.152880668640137,
      "learning_rate": 4.7924726117441135e-05,
      "loss": 3.3177,
      "step": 375
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.759363174438477,
      "learning_rate": 4.7860315575395316e-05,
      "loss": 2.9184,
      "step": 380
    },
    {
      "epoch": 0.15,
      "grad_norm": 37.01634216308594,
      "learning_rate": 4.7794965335313926e-05,
      "loss": 3.1494,
      "step": 385
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.318065643310547,
      "learning_rate": 4.772867808348938e-05,
      "loss": 3.7067,
      "step": 390
    },
    {
      "epoch": 0.16,
      "grad_norm": 14.883944511413574,
      "learning_rate": 4.766145654473095e-05,
      "loss": 2.9832,
      "step": 395
    },
    {
      "epoch": 0.16,
      "grad_norm": 35.67652130126953,
      "learning_rate": 4.759330348225284e-05,
      "loss": 3.6382,
      "step": 400
    },
    {
      "epoch": 0.16,
      "grad_norm": 6.075866222381592,
      "learning_rate": 4.752422169756048e-05,
      "loss": 3.6166,
      "step": 405
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.807673454284668,
      "learning_rate": 4.745421403033548e-05,
      "loss": 2.896,
      "step": 410
    },
    {
      "epoch": 0.17,
      "grad_norm": 17.647350311279297,
      "learning_rate": 4.738328335831883e-05,
      "loss": 2.9218,
      "step": 415
    },
    {
      "epoch": 0.17,
      "grad_norm": 35.758544921875,
      "learning_rate": 4.731143259719265e-05,
      "loss": 3.0896,
      "step": 420
    },
    {
      "epoch": 0.17,
      "grad_norm": 19.15152931213379,
      "learning_rate": 4.72386647004603e-05,
      "loss": 3.0875,
      "step": 425
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.721521377563477,
      "learning_rate": 4.716498265932501e-05,
      "loss": 2.833,
      "step": 430
    },
    {
      "epoch": 0.17,
      "grad_norm": 36.97370147705078,
      "learning_rate": 4.709038950256688e-05,
      "loss": 3.3553,
      "step": 435
    },
    {
      "epoch": 0.18,
      "grad_norm": 18.96316146850586,
      "learning_rate": 4.701488829641845e-05,
      "loss": 2.8189,
      "step": 440
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.662888526916504,
      "learning_rate": 4.693848214443858e-05,
      "loss": 3.4443,
      "step": 445
    },
    {
      "epoch": 0.18,
      "grad_norm": 18.23220443725586,
      "learning_rate": 4.686117418738489e-05,
      "loss": 3.5717,
      "step": 450
    },
    {
      "epoch": 0.18,
      "grad_norm": 39.92061996459961,
      "learning_rate": 4.678296760308474e-05,
      "loss": 3.2364,
      "step": 455
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.42786693572998,
      "learning_rate": 4.6703865606304465e-05,
      "loss": 2.9595,
      "step": 460
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.904202938079834,
      "learning_rate": 4.662387144861734e-05,
      "loss": 3.1582,
      "step": 465
    },
    {
      "epoch": 0.19,
      "grad_norm": 26.168556213378906,
      "learning_rate": 4.6542988418269876e-05,
      "loss": 3.3465,
      "step": 470
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.387030601501465,
      "learning_rate": 4.6461219840046654e-05,
      "loss": 3.0016,
      "step": 475
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.973118782043457,
      "learning_rate": 4.637856907513366e-05,
      "loss": 3.2299,
      "step": 480
    },
    {
      "epoch": 0.19,
      "grad_norm": 31.876556396484375,
      "learning_rate": 4.629503952098011e-05,
      "loss": 3.3682,
      "step": 485
    },
    {
      "epoch": 0.2,
      "grad_norm": 23.903682708740234,
      "learning_rate": 4.6210634611158816e-05,
      "loss": 2.7361,
      "step": 490
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.048882007598877,
      "learning_rate": 4.612535781522504e-05,
      "loss": 2.8006,
      "step": 495
    },
    {
      "epoch": 0.2,
      "grad_norm": 16.777807235717773,
      "learning_rate": 4.6039212638573833e-05,
      "loss": 2.7768,
      "step": 500
    },
    {
      "epoch": 0.2,
      "grad_norm": 35.86275100708008,
      "learning_rate": 4.595220262229601e-05,
      "loss": 3.2371,
      "step": 505
    },
    {
      "epoch": 0.2,
      "grad_norm": 18.5112361907959,
      "learning_rate": 4.586433134303257e-05,
      "loss": 3.9903,
      "step": 510
    },
    {
      "epoch": 0.21,
      "grad_norm": 15.84546947479248,
      "learning_rate": 4.5775602412827604e-05,
      "loss": 3.2624,
      "step": 515
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.097062110900879,
      "learning_rate": 4.5686019478979915e-05,
      "loss": 3.036,
      "step": 520
    },
    {
      "epoch": 0.21,
      "grad_norm": 12.135428428649902,
      "learning_rate": 4.559558622389304e-05,
      "loss": 3.0802,
      "step": 525
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.016368865966797,
      "learning_rate": 4.55043063649239e-05,
      "loss": 2.718,
      "step": 530
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.902190685272217,
      "learning_rate": 4.5412183654229965e-05,
      "loss": 2.9923,
      "step": 535
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.023144721984863,
      "learning_rate": 4.531922187861507e-05,
      "loss": 3.2273,
      "step": 540
    },
    {
      "epoch": 0.22,
      "grad_norm": 12.821276664733887,
      "learning_rate": 4.522542485937369e-05,
      "loss": 3.0269,
      "step": 545
    },
    {
      "epoch": 0.22,
      "grad_norm": 11.519277572631836,
      "learning_rate": 4.51307964521339e-05,
      "loss": 2.9414,
      "step": 550
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.272705554962158,
      "learning_rate": 4.503534054669892e-05,
      "loss": 3.1369,
      "step": 555
    },
    {
      "epoch": 0.22,
      "grad_norm": 15.541712760925293,
      "learning_rate": 4.493906106688712e-05,
      "loss": 3.3594,
      "step": 560
    },
    {
      "epoch": 0.23,
      "grad_norm": 29.176177978515625,
      "learning_rate": 4.484196197037082e-05,
      "loss": 3.7387,
      "step": 565
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.822291374206543,
      "learning_rate": 4.474404724851356e-05,
      "loss": 3.0015,
      "step": 570
    },
    {
      "epoch": 0.23,
      "grad_norm": 13.846043586730957,
      "learning_rate": 4.4645320926206064e-05,
      "loss": 2.8736,
      "step": 575
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.161740303039551,
      "learning_rate": 4.454578706170075e-05,
      "loss": 3.1031,
      "step": 580
    },
    {
      "epoch": 0.23,
      "grad_norm": 12.11698055267334,
      "learning_rate": 4.444544974644493e-05,
      "loss": 3.1832,
      "step": 585
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.144786834716797,
      "learning_rate": 4.434431310491267e-05,
      "loss": 2.7414,
      "step": 590
    },
    {
      "epoch": 0.24,
      "grad_norm": 14.706486701965332,
      "learning_rate": 4.4242381294435154e-05,
      "loss": 2.7289,
      "step": 595
    },
    {
      "epoch": 0.24,
      "grad_norm": 11.251163482666016,
      "learning_rate": 4.413965850502987e-05,
      "loss": 3.4787,
      "step": 600
    },
    {
      "epoch": 0.24,
      "grad_norm": 11.005367279052734,
      "learning_rate": 4.4036148959228365e-05,
      "loss": 3.1987,
      "step": 605
    },
    {
      "epoch": 0.24,
      "grad_norm": 13.336194038391113,
      "learning_rate": 4.393185691190264e-05,
      "loss": 2.9382,
      "step": 610
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.259475708007812,
      "learning_rate": 4.382678665009028e-05,
      "loss": 3.1349,
      "step": 615
    },
    {
      "epoch": 0.25,
      "grad_norm": 20.902462005615234,
      "learning_rate": 4.372094249281821e-05,
      "loss": 3.1365,
      "step": 620
    },
    {
      "epoch": 0.25,
      "grad_norm": 13.907161712646484,
      "learning_rate": 4.3614328790925177e-05,
      "loss": 3.3361,
      "step": 625
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.845438480377197,
      "learning_rate": 4.350694992688289e-05,
      "loss": 3.1903,
      "step": 630
    },
    {
      "epoch": 0.25,
      "grad_norm": 14.879186630249023,
      "learning_rate": 4.3398810314615876e-05,
      "loss": 3.5133,
      "step": 635
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.454853057861328,
      "learning_rate": 4.3289914399320034e-05,
      "loss": 3.0928,
      "step": 640
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.692648887634277,
      "learning_rate": 4.318026665727993e-05,
      "loss": 3.1065,
      "step": 645
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.677508354187012,
      "learning_rate": 4.306987159568479e-05,
      "loss": 2.6484,
      "step": 650
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.058112621307373,
      "learning_rate": 4.2958733752443195e-05,
      "loss": 2.6531,
      "step": 655
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.029059886932373,
      "learning_rate": 4.284685769599658e-05,
      "loss": 3.3165,
      "step": 660
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.306894302368164,
      "learning_rate": 4.273424802513145e-05,
      "loss": 3.2707,
      "step": 665
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.9906392097473145,
      "learning_rate": 4.262090936879029e-05,
      "loss": 3.0822,
      "step": 670
    },
    {
      "epoch": 0.27,
      "grad_norm": 10.101268768310547,
      "learning_rate": 4.250684638588138e-05,
      "loss": 3.3994,
      "step": 675
    },
    {
      "epoch": 0.27,
      "grad_norm": 13.360941886901855,
      "learning_rate": 4.239206376508717e-05,
      "loss": 3.2358,
      "step": 680
    },
    {
      "epoch": 0.27,
      "grad_norm": 25.73331642150879,
      "learning_rate": 4.227656622467162e-05,
      "loss": 3.1359,
      "step": 685
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.391608715057373,
      "learning_rate": 4.216035851228626e-05,
      "loss": 3.1343,
      "step": 690
    },
    {
      "epoch": 0.28,
      "grad_norm": 21.201011657714844,
      "learning_rate": 4.204344540477499e-05,
      "loss": 3.0807,
      "step": 695
    },
    {
      "epoch": 0.28,
      "grad_norm": 12.530668258666992,
      "learning_rate": 4.192583170797774e-05,
      "loss": 2.9729,
      "step": 700
    },
    {
      "epoch": 0.28,
      "grad_norm": 17.209762573242188,
      "learning_rate": 4.180752225653292e-05,
      "loss": 2.9911,
      "step": 705
    },
    {
      "epoch": 0.28,
      "grad_norm": 16.309785842895508,
      "learning_rate": 4.16885219136787e-05,
      "loss": 3.0701,
      "step": 710
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.244511127471924,
      "learning_rate": 4.1568835571053075e-05,
      "loss": 2.8169,
      "step": 715
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.627272605895996,
      "learning_rate": 4.144846814849282e-05,
      "loss": 3.0518,
      "step": 720
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.72758674621582,
      "learning_rate": 4.132742459383122e-05,
      "loss": 3.1135,
      "step": 725
    },
    {
      "epoch": 0.29,
      "grad_norm": 14.121967315673828,
      "learning_rate": 4.120570988269472e-05,
      "loss": 2.9481,
      "step": 730
    },
    {
      "epoch": 0.29,
      "grad_norm": 71.68965148925781,
      "learning_rate": 4.108332901829836e-05,
      "loss": 3.7509,
      "step": 735
    },
    {
      "epoch": 0.3,
      "grad_norm": 22.303184509277344,
      "learning_rate": 4.096028703124014e-05,
      "loss": 2.9864,
      "step": 740
    },
    {
      "epoch": 0.3,
      "grad_norm": 11.789464950561523,
      "learning_rate": 4.083658897929426e-05,
      "loss": 2.7573,
      "step": 745
    },
    {
      "epoch": 0.3,
      "grad_norm": 21.576683044433594,
      "learning_rate": 4.071223994720309e-05,
      "loss": 3.4365,
      "step": 750
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.004244804382324,
      "learning_rate": 4.058724504646834e-05,
      "loss": 3.1543,
      "step": 755
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.743468284606934,
      "learning_rate": 4.046160941514079e-05,
      "loss": 2.638,
      "step": 760
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.981857299804688,
      "learning_rate": 4.033533821760917e-05,
      "loss": 2.4701,
      "step": 765
    },
    {
      "epoch": 0.31,
      "grad_norm": 11.454618453979492,
      "learning_rate": 4.0208436644387834e-05,
      "loss": 3.4152,
      "step": 770
    },
    {
      "epoch": 0.31,
      "grad_norm": 22.861881256103516,
      "learning_rate": 4.008090991190341e-05,
      "loss": 3.0526,
      "step": 775
    },
    {
      "epoch": 0.31,
      "grad_norm": 34.58321762084961,
      "learning_rate": 3.9952763262280405e-05,
      "loss": 3.1283,
      "step": 780
    },
    {
      "epoch": 0.31,
      "grad_norm": 12.163683891296387,
      "learning_rate": 3.982400196312564e-05,
      "loss": 2.8686,
      "step": 785
    },
    {
      "epoch": 0.32,
      "grad_norm": 34.78618240356445,
      "learning_rate": 3.969463130731183e-05,
      "loss": 2.782,
      "step": 790
    },
    {
      "epoch": 0.32,
      "grad_norm": 11.429190635681152,
      "learning_rate": 3.95646566127599e-05,
      "loss": 3.1518,
      "step": 795
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.942728519439697,
      "learning_rate": 3.943408322222049e-05,
      "loss": 3.1183,
      "step": 800
    },
    {
      "epoch": 0.32,
      "grad_norm": 6.556036949157715,
      "learning_rate": 3.9302916503054246e-05,
      "loss": 2.9241,
      "step": 805
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.161313533782959,
      "learning_rate": 3.917116184701125e-05,
      "loss": 2.8394,
      "step": 810
    },
    {
      "epoch": 0.33,
      "grad_norm": 17.556371688842773,
      "learning_rate": 3.903882467000937e-05,
      "loss": 2.667,
      "step": 815
    },
    {
      "epoch": 0.33,
      "grad_norm": 6.0064167976379395,
      "learning_rate": 3.8905910411911625e-05,
      "loss": 2.8605,
      "step": 820
    },
    {
      "epoch": 0.33,
      "grad_norm": 13.329277038574219,
      "learning_rate": 3.8772424536302564e-05,
      "loss": 2.8765,
      "step": 825
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.7756876945495605,
      "learning_rate": 3.8638372530263715e-05,
      "loss": 2.6558,
      "step": 830
    },
    {
      "epoch": 0.33,
      "grad_norm": 6.505970478057861,
      "learning_rate": 3.850375990414801e-05,
      "loss": 2.6178,
      "step": 835
    },
    {
      "epoch": 0.34,
      "grad_norm": 29.54509925842285,
      "learning_rate": 3.836859219135324e-05,
      "loss": 2.9655,
      "step": 840
    },
    {
      "epoch": 0.34,
      "grad_norm": 12.92197322845459,
      "learning_rate": 3.823287494809469e-05,
      "loss": 2.7825,
      "step": 845
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.073705673217773,
      "learning_rate": 3.8096613753176634e-05,
      "loss": 3.3055,
      "step": 850
    },
    {
      "epoch": 0.34,
      "grad_norm": 23.254878997802734,
      "learning_rate": 3.7959814207763135e-05,
      "loss": 2.906,
      "step": 855
    },
    {
      "epoch": 0.34,
      "grad_norm": 16.555452346801758,
      "learning_rate": 3.782248193514766e-05,
      "loss": 3.2928,
      "step": 860
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.429189682006836,
      "learning_rate": 3.7684622580522055e-05,
      "loss": 2.6259,
      "step": 865
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.762368202209473,
      "learning_rate": 3.7546241810744445e-05,
      "loss": 2.7576,
      "step": 870
    },
    {
      "epoch": 0.35,
      "grad_norm": 11.004511833190918,
      "learning_rate": 3.740734531410626e-05,
      "loss": 2.6714,
      "step": 875
    },
    {
      "epoch": 0.35,
      "grad_norm": 56.73052978515625,
      "learning_rate": 3.726793880009845e-05,
      "loss": 3.542,
      "step": 880
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.330306053161621,
      "learning_rate": 3.7128027999176803e-05,
      "loss": 3.0011,
      "step": 885
    },
    {
      "epoch": 0.36,
      "grad_norm": 12.130693435668945,
      "learning_rate": 3.698761866252635e-05,
      "loss": 3.0935,
      "step": 890
    },
    {
      "epoch": 0.36,
      "grad_norm": 12.528786659240723,
      "learning_rate": 3.6846716561824965e-05,
      "loss": 3.3687,
      "step": 895
    },
    {
      "epoch": 0.36,
      "grad_norm": 14.122909545898438,
      "learning_rate": 3.670532748900615e-05,
      "loss": 2.5591,
      "step": 900
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.046327590942383,
      "learning_rate": 3.656345725602089e-05,
      "loss": 2.3845,
      "step": 905
    },
    {
      "epoch": 0.36,
      "grad_norm": 22.46470832824707,
      "learning_rate": 3.642111169459879e-05,
      "loss": 2.8321,
      "step": 910
    },
    {
      "epoch": 0.37,
      "grad_norm": 10.554203987121582,
      "learning_rate": 3.6278296656008366e-05,
      "loss": 2.4067,
      "step": 915
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.757329940795898,
      "learning_rate": 3.6135018010816477e-05,
      "loss": 2.7003,
      "step": 920
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.956037521362305,
      "learning_rate": 3.599128164864706e-05,
      "loss": 2.7975,
      "step": 925
    },
    {
      "epoch": 0.37,
      "grad_norm": 15.920394897460938,
      "learning_rate": 3.5847093477938956e-05,
      "loss": 2.7154,
      "step": 930
    },
    {
      "epoch": 0.37,
      "grad_norm": 15.14090633392334,
      "learning_rate": 3.570245942570315e-05,
      "loss": 3.3547,
      "step": 935
    },
    {
      "epoch": 0.38,
      "grad_norm": 17.72629165649414,
      "learning_rate": 3.5557385437279e-05,
      "loss": 2.7852,
      "step": 940
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.857475757598877,
      "learning_rate": 3.5411877476089975e-05,
      "loss": 2.73,
      "step": 945
    },
    {
      "epoch": 0.38,
      "grad_norm": 22.337295532226562,
      "learning_rate": 3.526594152339845e-05,
      "loss": 3.1076,
      "step": 950
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.523070335388184,
      "learning_rate": 3.5119583578059846e-05,
      "loss": 3.1857,
      "step": 955
    },
    {
      "epoch": 0.38,
      "grad_norm": 11.635588645935059,
      "learning_rate": 3.497280965627605e-05,
      "loss": 2.8584,
      "step": 960
    },
    {
      "epoch": 0.39,
      "grad_norm": 27.29017448425293,
      "learning_rate": 3.4825625791348096e-05,
      "loss": 3.3567,
      "step": 965
    },
    {
      "epoch": 0.39,
      "grad_norm": 25.24538230895996,
      "learning_rate": 3.467803803342821e-05,
      "loss": 3.1713,
      "step": 970
    },
    {
      "epoch": 0.39,
      "grad_norm": 11.2062406539917,
      "learning_rate": 3.4530052449271044e-05,
      "loss": 2.8941,
      "step": 975
    },
    {
      "epoch": 0.39,
      "grad_norm": 20.25462532043457,
      "learning_rate": 3.438167512198436e-05,
      "loss": 3.1291,
      "step": 980
    },
    {
      "epoch": 0.39,
      "grad_norm": 13.0928373336792,
      "learning_rate": 3.4232912150778914e-05,
      "loss": 2.4672,
      "step": 985
    },
    {
      "epoch": 0.4,
      "grad_norm": 21.326560974121094,
      "learning_rate": 3.408376965071779e-05,
      "loss": 3.0489,
      "step": 990
    },
    {
      "epoch": 0.4,
      "grad_norm": 54.281578063964844,
      "learning_rate": 3.393425375246503e-05,
      "loss": 3.0251,
      "step": 995
    },
    {
      "epoch": 0.4,
      "grad_norm": 18.1798038482666,
      "learning_rate": 3.378437060203357e-05,
      "loss": 3.0152,
      "step": 1000
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.757742881774902,
      "learning_rate": 3.363412636053269e-05,
      "loss": 2.5223,
      "step": 1005
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.601003170013428,
      "learning_rate": 3.348352720391469e-05,
      "loss": 2.8158,
      "step": 1010
    },
    {
      "epoch": 0.41,
      "grad_norm": 16.87445068359375,
      "learning_rate": 3.3332579322721046e-05,
      "loss": 3.3828,
      "step": 1015
    },
    {
      "epoch": 0.41,
      "grad_norm": 11.726461410522461,
      "learning_rate": 3.318128892182792e-05,
      "loss": 2.5411,
      "step": 1020
    },
    {
      "epoch": 0.41,
      "grad_norm": 11.669478416442871,
      "learning_rate": 3.3029662220191144e-05,
      "loss": 2.6686,
      "step": 1025
    },
    {
      "epoch": 0.41,
      "grad_norm": 58.972007751464844,
      "learning_rate": 3.2877705450590526e-05,
      "loss": 3.3026,
      "step": 1030
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.434502124786377,
      "learning_rate": 3.272542485937369e-05,
      "loss": 2.6652,
      "step": 1035
    },
    {
      "epoch": 0.42,
      "grad_norm": 22.020599365234375,
      "learning_rate": 3.2572826706199305e-05,
      "loss": 3.1553,
      "step": 1040
    },
    {
      "epoch": 0.42,
      "grad_norm": 17.60317611694336,
      "learning_rate": 3.2419917263779766e-05,
      "loss": 3.2016,
      "step": 1045
    },
    {
      "epoch": 0.42,
      "grad_norm": 15.09745979309082,
      "learning_rate": 3.2266702817623346e-05,
      "loss": 2.6242,
      "step": 1050
    },
    {
      "epoch": 0.42,
      "grad_norm": 11.325228691101074,
      "learning_rate": 3.211318966577581e-05,
      "loss": 2.9649,
      "step": 1055
    },
    {
      "epoch": 0.42,
      "grad_norm": 16.898090362548828,
      "learning_rate": 3.195938411856159e-05,
      "loss": 2.4623,
      "step": 1060
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.761795043945312,
      "learning_rate": 3.180529249832428e-05,
      "loss": 2.4152,
      "step": 1065
    },
    {
      "epoch": 0.43,
      "grad_norm": 11.047560691833496,
      "learning_rate": 3.165092113916688e-05,
      "loss": 3.0119,
      "step": 1070
    },
    {
      "epoch": 0.43,
      "grad_norm": 11.61601734161377,
      "learning_rate": 3.149627638669132e-05,
      "loss": 2.4781,
      "step": 1075
    },
    {
      "epoch": 0.43,
      "grad_norm": 21.56399154663086,
      "learning_rate": 3.1341364597737686e-05,
      "loss": 3.4005,
      "step": 1080
    },
    {
      "epoch": 0.43,
      "grad_norm": 15.666382789611816,
      "learning_rate": 3.118619214012286e-05,
      "loss": 2.9385,
      "step": 1085
    },
    {
      "epoch": 0.44,
      "grad_norm": 11.866064071655273,
      "learning_rate": 3.1030765392378816e-05,
      "loss": 3.0096,
      "step": 1090
    },
    {
      "epoch": 0.44,
      "grad_norm": 11.25349235534668,
      "learning_rate": 3.0875090743490384e-05,
      "loss": 2.0702,
      "step": 1095
    },
    {
      "epoch": 0.44,
      "grad_norm": 10.070486068725586,
      "learning_rate": 3.071917459263264e-05,
      "loss": 3.0652,
      "step": 1100
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.047021389007568,
      "learning_rate": 3.056302334890786e-05,
      "loss": 2.8617,
      "step": 1105
    },
    {
      "epoch": 0.44,
      "grad_norm": 11.59208869934082,
      "learning_rate": 3.040664343108209e-05,
      "loss": 2.8697,
      "step": 1110
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.361946105957031,
      "learning_rate": 3.0250041267321232e-05,
      "loss": 2.7646,
      "step": 1115
    },
    {
      "epoch": 0.45,
      "grad_norm": 11.60400390625,
      "learning_rate": 3.0093223294926892e-05,
      "loss": 2.8238,
      "step": 1120
    },
    {
      "epoch": 0.45,
      "grad_norm": 19.50318145751953,
      "learning_rate": 2.993619596007168e-05,
      "loss": 2.8857,
      "step": 1125
    },
    {
      "epoch": 0.45,
      "grad_norm": 24.245025634765625,
      "learning_rate": 2.9778965717534313e-05,
      "loss": 2.949,
      "step": 1130
    },
    {
      "epoch": 0.45,
      "grad_norm": 14.323025703430176,
      "learning_rate": 2.962153903043422e-05,
      "loss": 2.6237,
      "step": 1135
    },
    {
      "epoch": 0.46,
      "grad_norm": 13.869138717651367,
      "learning_rate": 2.9463922369965917e-05,
      "loss": 2.5021,
      "step": 1140
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.667675018310547,
      "learning_rate": 2.9306122215132976e-05,
      "loss": 2.9346,
      "step": 1145
    },
    {
      "epoch": 0.46,
      "grad_norm": 22.312841415405273,
      "learning_rate": 2.91481450524817e-05,
      "loss": 2.7172,
      "step": 1150
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.218385696411133,
      "learning_rate": 2.8989997375834482e-05,
      "loss": 2.4864,
      "step": 1155
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.388011455535889,
      "learning_rate": 2.8831685686022897e-05,
      "loss": 2.7519,
      "step": 1160
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.172524452209473,
      "learning_rate": 2.8673216490620452e-05,
      "loss": 2.701,
      "step": 1165
    },
    {
      "epoch": 0.47,
      "grad_norm": 11.60141372680664,
      "learning_rate": 2.8514596303675073e-05,
      "loss": 3.1608,
      "step": 1170
    },
    {
      "epoch": 0.47,
      "grad_norm": 21.537139892578125,
      "learning_rate": 2.8355831645441388e-05,
      "loss": 2.9307,
      "step": 1175
    },
    {
      "epoch": 0.47,
      "grad_norm": 12.06334400177002,
      "learning_rate": 2.8196929042112652e-05,
      "loss": 2.9414,
      "step": 1180
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.39703369140625,
      "learning_rate": 2.8037895025552512e-05,
      "loss": 2.938,
      "step": 1185
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.78774642944336,
      "learning_rate": 2.787873613302649e-05,
      "loss": 2.5751,
      "step": 1190
    },
    {
      "epoch": 0.48,
      "grad_norm": 24.190916061401367,
      "learning_rate": 2.7719458906933277e-05,
      "loss": 2.328,
      "step": 1195
    },
    {
      "epoch": 0.48,
      "grad_norm": 17.80084228515625,
      "learning_rate": 2.7560069894535784e-05,
      "loss": 2.7258,
      "step": 1200
    },
    {
      "epoch": 0.48,
      "grad_norm": 17.03345489501953,
      "learning_rate": 2.7400575647692046e-05,
      "loss": 2.8104,
      "step": 1205
    },
    {
      "epoch": 0.48,
      "grad_norm": 36.040897369384766,
      "learning_rate": 2.724098272258584e-05,
      "loss": 2.9138,
      "step": 1210
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.431939125061035,
      "learning_rate": 2.7081297679457236e-05,
      "loss": 2.3052,
      "step": 1215
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.720209121704102,
      "learning_rate": 2.692152708233292e-05,
      "loss": 2.3984,
      "step": 1220
    },
    {
      "epoch": 0.49,
      "grad_norm": 13.637066841125488,
      "learning_rate": 2.676167749875635e-05,
      "loss": 2.8592,
      "step": 1225
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.968180656433105,
      "learning_rate": 2.6601755499517826e-05,
      "loss": 2.4064,
      "step": 1230
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.793657302856445,
      "learning_rate": 2.6441767658384366e-05,
      "loss": 3.0291,
      "step": 1235
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.77736759185791,
      "learning_rate": 2.628172055182948e-05,
      "loss": 2.3978,
      "step": 1240
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.85341739654541,
      "learning_rate": 2.6121620758762877e-05,
      "loss": 2.5431,
      "step": 1245
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.554092407226562,
      "learning_rate": 2.596147486025996e-05,
      "loss": 2.8736,
      "step": 1250
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.345808029174805,
      "learning_rate": 2.5801289439291388e-05,
      "loss": 2.6837,
      "step": 1255
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.449942588806152,
      "learning_rate": 2.564107108045239e-05,
      "loss": 3.1686,
      "step": 1260
    },
    {
      "epoch": 0.51,
      "grad_norm": 32.239585876464844,
      "learning_rate": 2.5480826369692178e-05,
      "loss": 3.0723,
      "step": 1265
    },
    {
      "epoch": 0.51,
      "grad_norm": 36.09556579589844,
      "learning_rate": 2.5320561894043184e-05,
      "loss": 2.6983,
      "step": 1270
    },
    {
      "epoch": 0.51,
      "grad_norm": 42.91587829589844,
      "learning_rate": 2.5160284241350278e-05,
      "loss": 2.6336,
      "step": 1275
    },
    {
      "epoch": 0.51,
      "grad_norm": 20.82798194885254,
      "learning_rate": 2.5e-05,
      "loss": 2.8635,
      "step": 1280
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.868647575378418,
      "learning_rate": 2.4839715758649724e-05,
      "loss": 2.8232,
      "step": 1285
    },
    {
      "epoch": 0.52,
      "grad_norm": 12.662830352783203,
      "learning_rate": 2.467943810595682e-05,
      "loss": 2.8228,
      "step": 1290
    },
    {
      "epoch": 0.52,
      "grad_norm": 15.0082368850708,
      "learning_rate": 2.4519173630307825e-05,
      "loss": 2.6371,
      "step": 1295
    },
    {
      "epoch": 0.52,
      "grad_norm": 15.185354232788086,
      "learning_rate": 2.4358928919547616e-05,
      "loss": 2.6336,
      "step": 1300
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.559114456176758,
      "learning_rate": 2.419871056070862e-05,
      "loss": 2.4209,
      "step": 1305
    },
    {
      "epoch": 0.52,
      "grad_norm": 13.428425788879395,
      "learning_rate": 2.403852513974004e-05,
      "loss": 2.6897,
      "step": 1310
    },
    {
      "epoch": 0.53,
      "grad_norm": 38.91737747192383,
      "learning_rate": 2.3878379241237136e-05,
      "loss": 2.569,
      "step": 1315
    },
    {
      "epoch": 0.53,
      "grad_norm": 15.211456298828125,
      "learning_rate": 2.3718279448170525e-05,
      "loss": 2.4856,
      "step": 1320
    },
    {
      "epoch": 0.53,
      "grad_norm": 11.391417503356934,
      "learning_rate": 2.3558232341615643e-05,
      "loss": 2.5527,
      "step": 1325
    },
    {
      "epoch": 0.53,
      "grad_norm": 21.21587371826172,
      "learning_rate": 2.339824450048218e-05,
      "loss": 3.5899,
      "step": 1330
    },
    {
      "epoch": 0.53,
      "grad_norm": 17.66887664794922,
      "learning_rate": 2.323832250124365e-05,
      "loss": 2.4669,
      "step": 1335
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.46285629272461,
      "learning_rate": 2.3078472917667092e-05,
      "loss": 2.4984,
      "step": 1340
    },
    {
      "epoch": 0.54,
      "grad_norm": 14.3226318359375,
      "learning_rate": 2.291870232054277e-05,
      "loss": 2.8999,
      "step": 1345
    },
    {
      "epoch": 0.54,
      "grad_norm": 47.95698547363281,
      "learning_rate": 2.2759017277414166e-05,
      "loss": 2.7827,
      "step": 1350
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.691919326782227,
      "learning_rate": 2.2599424352307957e-05,
      "loss": 2.9201,
      "step": 1355
    },
    {
      "epoch": 0.54,
      "grad_norm": 15.748466491699219,
      "learning_rate": 2.243993010546422e-05,
      "loss": 2.3433,
      "step": 1360
    },
    {
      "epoch": 0.55,
      "grad_norm": 18.585126876831055,
      "learning_rate": 2.2280541093066732e-05,
      "loss": 2.7896,
      "step": 1365
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.304043769836426,
      "learning_rate": 2.212126386697352e-05,
      "loss": 2.781,
      "step": 1370
    },
    {
      "epoch": 0.55,
      "grad_norm": 14.388626098632812,
      "learning_rate": 2.196210497444749e-05,
      "loss": 2.9678,
      "step": 1375
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.608166694641113,
      "learning_rate": 2.1803070957887347e-05,
      "loss": 3.0607,
      "step": 1380
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.090882301330566,
      "learning_rate": 2.164416835455862e-05,
      "loss": 3.6207,
      "step": 1385
    },
    {
      "epoch": 0.56,
      "grad_norm": 13.572842597961426,
      "learning_rate": 2.1485403696324936e-05,
      "loss": 2.5546,
      "step": 1390
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.186910629272461,
      "learning_rate": 2.1326783509379554e-05,
      "loss": 2.9678,
      "step": 1395
    },
    {
      "epoch": 0.56,
      "grad_norm": 50.014766693115234,
      "learning_rate": 2.11683143139771e-05,
      "loss": 3.1836,
      "step": 1400
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.37592077255249,
      "learning_rate": 2.1010002624165527e-05,
      "loss": 2.659,
      "step": 1405
    },
    {
      "epoch": 0.56,
      "grad_norm": 22.413625717163086,
      "learning_rate": 2.0851854947518313e-05,
      "loss": 2.6929,
      "step": 1410
    },
    {
      "epoch": 0.57,
      "grad_norm": 11.021233558654785,
      "learning_rate": 2.069387778486703e-05,
      "loss": 2.7117,
      "step": 1415
    },
    {
      "epoch": 0.57,
      "grad_norm": 56.379844665527344,
      "learning_rate": 2.0536077630034086e-05,
      "loss": 3.4033,
      "step": 1420
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.377793312072754,
      "learning_rate": 2.0378460969565782e-05,
      "loss": 2.4798,
      "step": 1425
    },
    {
      "epoch": 0.57,
      "grad_norm": 6.888087749481201,
      "learning_rate": 2.02210342824657e-05,
      "loss": 2.6521,
      "step": 1430
    },
    {
      "epoch": 0.57,
      "grad_norm": 6.764181613922119,
      "learning_rate": 2.0063804039928324e-05,
      "loss": 2.6824,
      "step": 1435
    },
    {
      "epoch": 0.58,
      "grad_norm": 55.408729553222656,
      "learning_rate": 1.9906776705073114e-05,
      "loss": 3.5755,
      "step": 1440
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.327556610107422,
      "learning_rate": 1.9749958732678767e-05,
      "loss": 2.3398,
      "step": 1445
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.360859870910645,
      "learning_rate": 1.9593356568917913e-05,
      "loss": 2.5018,
      "step": 1450
    },
    {
      "epoch": 0.58,
      "grad_norm": 32.50775146484375,
      "learning_rate": 1.9436976651092144e-05,
      "loss": 2.8295,
      "step": 1455
    },
    {
      "epoch": 0.58,
      "grad_norm": 23.467443466186523,
      "learning_rate": 1.928082540736737e-05,
      "loss": 2.6034,
      "step": 1460
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.357468605041504,
      "learning_rate": 1.9124909256509622e-05,
      "loss": 3.0481,
      "step": 1465
    },
    {
      "epoch": 0.59,
      "grad_norm": 16.312400817871094,
      "learning_rate": 1.8969234607621186e-05,
      "loss": 2.9409,
      "step": 1470
    },
    {
      "epoch": 0.59,
      "grad_norm": 15.405855178833008,
      "learning_rate": 1.8813807859877147e-05,
      "loss": 2.6141,
      "step": 1475
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.774744987487793,
      "learning_rate": 1.865863540226232e-05,
      "loss": 2.5919,
      "step": 1480
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.001652717590332,
      "learning_rate": 1.8503723613308683e-05,
      "loss": 3.043,
      "step": 1485
    },
    {
      "epoch": 0.6,
      "grad_norm": 32.264766693115234,
      "learning_rate": 1.8349078860833123e-05,
      "loss": 2.711,
      "step": 1490
    },
    {
      "epoch": 0.6,
      "grad_norm": 15.870736122131348,
      "learning_rate": 1.8194707501675724e-05,
      "loss": 2.4717,
      "step": 1495
    },
    {
      "epoch": 0.6,
      "grad_norm": 15.47707748413086,
      "learning_rate": 1.8040615881438425e-05,
      "loss": 2.9706,
      "step": 1500
    },
    {
      "epoch": 0.6,
      "grad_norm": 12.465653419494629,
      "learning_rate": 1.7886810334224192e-05,
      "loss": 3.0024,
      "step": 1505
    },
    {
      "epoch": 0.6,
      "grad_norm": 20.99846839904785,
      "learning_rate": 1.7733297182376663e-05,
      "loss": 3.3546,
      "step": 1510
    },
    {
      "epoch": 0.61,
      "grad_norm": 37.41334533691406,
      "learning_rate": 1.7580082736220237e-05,
      "loss": 2.924,
      "step": 1515
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.230709552764893,
      "learning_rate": 1.74271732938007e-05,
      "loss": 2.7253,
      "step": 1520
    },
    {
      "epoch": 0.61,
      "grad_norm": 15.803852081298828,
      "learning_rate": 1.7274575140626318e-05,
      "loss": 2.8601,
      "step": 1525
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.981109142303467,
      "learning_rate": 1.7122294549409484e-05,
      "loss": 2.6754,
      "step": 1530
    },
    {
      "epoch": 0.61,
      "grad_norm": 21.247787475585938,
      "learning_rate": 1.6970337779808862e-05,
      "loss": 2.5148,
      "step": 1535
    },
    {
      "epoch": 0.62,
      "grad_norm": 13.401627540588379,
      "learning_rate": 1.6818711078172077e-05,
      "loss": 2.5783,
      "step": 1540
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.62679386138916,
      "learning_rate": 1.666742067727896e-05,
      "loss": 2.5611,
      "step": 1545
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.567328929901123,
      "learning_rate": 1.6516472796085315e-05,
      "loss": 2.3179,
      "step": 1550
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.707379341125488,
      "learning_rate": 1.6365873639467315e-05,
      "loss": 2.3441,
      "step": 1555
    },
    {
      "epoch": 0.62,
      "grad_norm": 38.05426788330078,
      "learning_rate": 1.621562939796643e-05,
      "loss": 2.562,
      "step": 1560
    },
    {
      "epoch": 0.63,
      "grad_norm": 18.448406219482422,
      "learning_rate": 1.6065746247534984e-05,
      "loss": 2.6616,
      "step": 1565
    },
    {
      "epoch": 0.63,
      "grad_norm": 12.998920440673828,
      "learning_rate": 1.5916230349282215e-05,
      "loss": 2.9305,
      "step": 1570
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.341577053070068,
      "learning_rate": 1.5767087849221096e-05,
      "loss": 2.6997,
      "step": 1575
    },
    {
      "epoch": 0.63,
      "grad_norm": 11.726909637451172,
      "learning_rate": 1.561832487801565e-05,
      "loss": 2.9281,
      "step": 1580
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.877985000610352,
      "learning_rate": 1.5469947550728958e-05,
      "loss": 3.7181,
      "step": 1585
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.845093727111816,
      "learning_rate": 1.53219619665718e-05,
      "loss": 3.1277,
      "step": 1590
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.064024925231934,
      "learning_rate": 1.5174374208651912e-05,
      "loss": 2.8159,
      "step": 1595
    },
    {
      "epoch": 0.64,
      "grad_norm": 12.611387252807617,
      "learning_rate": 1.502719034372396e-05,
      "loss": 2.9906,
      "step": 1600
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.44251537322998,
      "learning_rate": 1.4880416421940155e-05,
      "loss": 2.7055,
      "step": 1605
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.020173072814941,
      "learning_rate": 1.4734058476601553e-05,
      "loss": 2.7852,
      "step": 1610
    },
    {
      "epoch": 0.65,
      "grad_norm": 18.893342971801758,
      "learning_rate": 1.458812252391003e-05,
      "loss": 2.7677,
      "step": 1615
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.38496208190918,
      "learning_rate": 1.444261456272101e-05,
      "loss": 2.7732,
      "step": 1620
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.147873401641846,
      "learning_rate": 1.4297540574296869e-05,
      "loss": 2.2796,
      "step": 1625
    },
    {
      "epoch": 0.65,
      "grad_norm": 18.142108917236328,
      "learning_rate": 1.4152906522061048e-05,
      "loss": 3.1047,
      "step": 1630
    },
    {
      "epoch": 0.65,
      "grad_norm": 16.73578643798828,
      "learning_rate": 1.400871835135295e-05,
      "loss": 2.9558,
      "step": 1635
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.667417526245117,
      "learning_rate": 1.386498198918352e-05,
      "loss": 2.5587,
      "step": 1640
    },
    {
      "epoch": 0.66,
      "grad_norm": 29.769210815429688,
      "learning_rate": 1.3721703343991633e-05,
      "loss": 3.2146,
      "step": 1645
    },
    {
      "epoch": 0.66,
      "grad_norm": 27.924272537231445,
      "learning_rate": 1.3578888305401207e-05,
      "loss": 2.8355,
      "step": 1650
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.13482666015625,
      "learning_rate": 1.3436542743979125e-05,
      "loss": 2.8218,
      "step": 1655
    },
    {
      "epoch": 0.66,
      "grad_norm": 16.816307067871094,
      "learning_rate": 1.329467251099386e-05,
      "loss": 2.5636,
      "step": 1660
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.0423583984375,
      "learning_rate": 1.3153283438175034e-05,
      "loss": 2.3591,
      "step": 1665
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.223281860351562,
      "learning_rate": 1.3012381337473656e-05,
      "loss": 3.0024,
      "step": 1670
    },
    {
      "epoch": 0.67,
      "grad_norm": 16.49390983581543,
      "learning_rate": 1.2871972000823196e-05,
      "loss": 2.9023,
      "step": 1675
    },
    {
      "epoch": 0.67,
      "grad_norm": 7.545689582824707,
      "learning_rate": 1.2732061199901562e-05,
      "loss": 2.8308,
      "step": 1680
    },
    {
      "epoch": 0.67,
      "grad_norm": 16.10993003845215,
      "learning_rate": 1.2592654685893757e-05,
      "loss": 2.6319,
      "step": 1685
    },
    {
      "epoch": 0.68,
      "grad_norm": 17.161388397216797,
      "learning_rate": 1.2453758189255568e-05,
      "loss": 2.6535,
      "step": 1690
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.898162841796875,
      "learning_rate": 1.231537741947795e-05,
      "loss": 3.3956,
      "step": 1695
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.259682655334473,
      "learning_rate": 1.217751806485235e-05,
      "loss": 2.5027,
      "step": 1700
    },
    {
      "epoch": 0.68,
      "grad_norm": 13.10878849029541,
      "learning_rate": 1.2040185792236874e-05,
      "loss": 2.4352,
      "step": 1705
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.2070722579956055,
      "learning_rate": 1.1903386246823361e-05,
      "loss": 2.8799,
      "step": 1710
    },
    {
      "epoch": 0.69,
      "grad_norm": 12.991480827331543,
      "learning_rate": 1.1767125051905315e-05,
      "loss": 3.0858,
      "step": 1715
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.183439254760742,
      "learning_rate": 1.1631407808646758e-05,
      "loss": 2.767,
      "step": 1720
    },
    {
      "epoch": 0.69,
      "grad_norm": 44.539459228515625,
      "learning_rate": 1.1496240095852001e-05,
      "loss": 2.9168,
      "step": 1725
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.825093269348145,
      "learning_rate": 1.1361627469736285e-05,
      "loss": 2.6493,
      "step": 1730
    },
    {
      "epoch": 0.69,
      "grad_norm": 13.887325286865234,
      "learning_rate": 1.122757546369744e-05,
      "loss": 2.8159,
      "step": 1735
    },
    {
      "epoch": 0.7,
      "grad_norm": 6.512905120849609,
      "learning_rate": 1.1094089588088383e-05,
      "loss": 3.4574,
      "step": 1740
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.677343368530273,
      "learning_rate": 1.096117532999063e-05,
      "loss": 2.7936,
      "step": 1745
    },
    {
      "epoch": 0.7,
      "grad_norm": 29.443038940429688,
      "learning_rate": 1.082883815298876e-05,
      "loss": 2.7603,
      "step": 1750
    },
    {
      "epoch": 0.7,
      "grad_norm": 6.62708044052124,
      "learning_rate": 1.0697083496945765e-05,
      "loss": 2.6984,
      "step": 1755
    },
    {
      "epoch": 0.7,
      "grad_norm": 11.575033187866211,
      "learning_rate": 1.0565916777779519e-05,
      "loss": 2.7648,
      "step": 1760
    },
    {
      "epoch": 0.71,
      "grad_norm": 19.933650970458984,
      "learning_rate": 1.0435343387240098e-05,
      "loss": 2.9776,
      "step": 1765
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.603243827819824,
      "learning_rate": 1.0305368692688174e-05,
      "loss": 2.4419,
      "step": 1770
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.084098815917969,
      "learning_rate": 1.0175998036874356e-05,
      "loss": 2.7489,
      "step": 1775
    },
    {
      "epoch": 0.71,
      "grad_norm": 20.572294235229492,
      "learning_rate": 1.0047236737719601e-05,
      "loss": 3.0075,
      "step": 1780
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.42055606842041,
      "learning_rate": 9.919090088096589e-06,
      "loss": 2.7706,
      "step": 1785
    },
    {
      "epoch": 0.72,
      "grad_norm": 33.90031051635742,
      "learning_rate": 9.791563355612172e-06,
      "loss": 3.6592,
      "step": 1790
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.972926139831543,
      "learning_rate": 9.664661782390841e-06,
      "loss": 2.4741,
      "step": 1795
    },
    {
      "epoch": 0.72,
      "grad_norm": 34.49540328979492,
      "learning_rate": 9.538390584859214e-06,
      "loss": 2.8379,
      "step": 1800
    },
    {
      "epoch": 0.72,
      "grad_norm": 17.665390014648438,
      "learning_rate": 9.412754953531663e-06,
      "loss": 3.0774,
      "step": 1805
    },
    {
      "epoch": 0.72,
      "grad_norm": 24.203062057495117,
      "learning_rate": 9.287760052796909e-06,
      "loss": 2.6639,
      "step": 1810
    },
    {
      "epoch": 0.73,
      "grad_norm": 15.584200859069824,
      "learning_rate": 9.163411020705762e-06,
      "loss": 2.75,
      "step": 1815
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.29335880279541,
      "learning_rate": 9.039712968759864e-06,
      "loss": 2.9486,
      "step": 1820
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.088394165039062,
      "learning_rate": 8.916670981701655e-06,
      "loss": 2.251,
      "step": 1825
    },
    {
      "epoch": 0.73,
      "grad_norm": 12.336640357971191,
      "learning_rate": 8.794290117305296e-06,
      "loss": 2.9527,
      "step": 1830
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.988898277282715,
      "learning_rate": 8.672575406168782e-06,
      "loss": 2.6327,
      "step": 1835
    },
    {
      "epoch": 0.74,
      "grad_norm": 11.5975980758667,
      "learning_rate": 8.551531851507186e-06,
      "loss": 2.6512,
      "step": 1840
    },
    {
      "epoch": 0.74,
      "grad_norm": 23.88221549987793,
      "learning_rate": 8.431164428946927e-06,
      "loss": 2.6542,
      "step": 1845
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.243393898010254,
      "learning_rate": 8.3114780863213e-06,
      "loss": 3.0325,
      "step": 1850
    },
    {
      "epoch": 0.74,
      "grad_norm": 13.184893608093262,
      "learning_rate": 8.192477743467078e-06,
      "loss": 2.7061,
      "step": 1855
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.766141414642334,
      "learning_rate": 8.07416829202227e-06,
      "loss": 2.2925,
      "step": 1860
    },
    {
      "epoch": 0.75,
      "grad_norm": 30.14096450805664,
      "learning_rate": 7.956554595225016e-06,
      "loss": 3.4068,
      "step": 1865
    },
    {
      "epoch": 0.75,
      "grad_norm": 16.042314529418945,
      "learning_rate": 7.839641487713745e-06,
      "loss": 2.388,
      "step": 1870
    },
    {
      "epoch": 0.75,
      "grad_norm": 16.486698150634766,
      "learning_rate": 7.723433775328384e-06,
      "loss": 3.549,
      "step": 1875
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.384564399719238,
      "learning_rate": 7.607936234912841e-06,
      "loss": 3.186,
      "step": 1880
    },
    {
      "epoch": 0.75,
      "grad_norm": 13.724809646606445,
      "learning_rate": 7.493153614118634e-06,
      "loss": 2.6417,
      "step": 1885
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.79148006439209,
      "learning_rate": 7.379090631209712e-06,
      "loss": 2.6296,
      "step": 1890
    },
    {
      "epoch": 0.76,
      "grad_norm": 14.533072471618652,
      "learning_rate": 7.265751974868554e-06,
      "loss": 2.5174,
      "step": 1895
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.014676094055176,
      "learning_rate": 7.153142304003418e-06,
      "loss": 2.5473,
      "step": 1900
    },
    {
      "epoch": 0.76,
      "grad_norm": 14.28827953338623,
      "learning_rate": 7.041266247556813e-06,
      "loss": 2.8404,
      "step": 1905
    },
    {
      "epoch": 0.76,
      "grad_norm": 13.483999252319336,
      "learning_rate": 6.930128404315214e-06,
      "loss": 2.6145,
      "step": 1910
    },
    {
      "epoch": 0.77,
      "grad_norm": 22.0334529876709,
      "learning_rate": 6.819733342720066e-06,
      "loss": 2.7059,
      "step": 1915
    },
    {
      "epoch": 0.77,
      "grad_norm": 42.69429397583008,
      "learning_rate": 6.7100856006799665e-06,
      "loss": 3.0099,
      "step": 1920
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.883504867553711,
      "learning_rate": 6.601189685384126e-06,
      "loss": 2.9242,
      "step": 1925
    },
    {
      "epoch": 0.77,
      "grad_norm": 17.86440658569336,
      "learning_rate": 6.493050073117116e-06,
      "loss": 3.022,
      "step": 1930
    },
    {
      "epoch": 0.77,
      "grad_norm": 27.365087509155273,
      "learning_rate": 6.385671209074828e-06,
      "loss": 2.9742,
      "step": 1935
    },
    {
      "epoch": 0.78,
      "grad_norm": 13.641215324401855,
      "learning_rate": 6.279057507181796e-06,
      "loss": 2.4529,
      "step": 1940
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.5147123336792,
      "learning_rate": 6.173213349909729e-06,
      "loss": 2.5926,
      "step": 1945
    },
    {
      "epoch": 0.78,
      "grad_norm": 13.009632110595703,
      "learning_rate": 6.068143088097372e-06,
      "loss": 2.4284,
      "step": 1950
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.745450973510742,
      "learning_rate": 5.9638510407716394e-06,
      "loss": 2.7295,
      "step": 1955
    },
    {
      "epoch": 0.78,
      "grad_norm": 7.102616310119629,
      "learning_rate": 5.860341494970131e-06,
      "loss": 2.3844,
      "step": 1960
    },
    {
      "epoch": 0.79,
      "grad_norm": 19.79552459716797,
      "learning_rate": 5.757618705564849e-06,
      "loss": 2.6637,
      "step": 1965
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.288678169250488,
      "learning_rate": 5.655686895087329e-06,
      "loss": 2.6376,
      "step": 1970
    },
    {
      "epoch": 0.79,
      "grad_norm": 5.515965938568115,
      "learning_rate": 5.554550253555066e-06,
      "loss": 2.7672,
      "step": 1975
    },
    {
      "epoch": 0.79,
      "grad_norm": 22.45775604248047,
      "learning_rate": 5.454212938299255e-06,
      "loss": 2.9637,
      "step": 1980
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.804770469665527,
      "learning_rate": 5.354679073793942e-06,
      "loss": 2.4708,
      "step": 1985
    },
    {
      "epoch": 0.8,
      "grad_norm": 14.247252464294434,
      "learning_rate": 5.255952751486443e-06,
      "loss": 2.6551,
      "step": 1990
    },
    {
      "epoch": 0.8,
      "grad_norm": 7.674140930175781,
      "learning_rate": 5.158038029629195e-06,
      "loss": 2.7633,
      "step": 1995
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.93137264251709,
      "learning_rate": 5.060938933112891e-06,
      "loss": 2.7034,
      "step": 2000
    },
    {
      "epoch": 0.8,
      "grad_norm": 11.792275428771973,
      "learning_rate": 4.9646594533010875e-06,
      "loss": 2.8564,
      "step": 2005
    },
    {
      "epoch": 0.8,
      "grad_norm": 13.885851860046387,
      "learning_rate": 4.869203547866097e-06,
      "loss": 2.6809,
      "step": 2010
    },
    {
      "epoch": 0.81,
      "grad_norm": 21.813217163085938,
      "learning_rate": 4.7745751406263165e-06,
      "loss": 3.0319,
      "step": 2015
    },
    {
      "epoch": 0.81,
      "grad_norm": 12.183605194091797,
      "learning_rate": 4.680778121384935e-06,
      "loss": 2.5685,
      "step": 2020
    },
    {
      "epoch": 0.81,
      "grad_norm": 26.5369930267334,
      "learning_rate": 4.587816345770032e-06,
      "loss": 2.9539,
      "step": 2025
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.17782735824585,
      "learning_rate": 4.495693635076101e-06,
      "loss": 2.3644,
      "step": 2030
    },
    {
      "epoch": 0.81,
      "grad_norm": 21.22880744934082,
      "learning_rate": 4.404413776106958e-06,
      "loss": 3.0965,
      "step": 2035
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.561256408691406,
      "learning_rate": 4.313980521020092e-06,
      "loss": 3.032,
      "step": 2040
    },
    {
      "epoch": 0.82,
      "grad_norm": 7.701003074645996,
      "learning_rate": 4.224397587172402e-06,
      "loss": 2.7757,
      "step": 2045
    },
    {
      "epoch": 0.82,
      "grad_norm": 17.345849990844727,
      "learning_rate": 4.135668656967434e-06,
      "loss": 2.3689,
      "step": 2050
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.013896942138672,
      "learning_rate": 4.047797377703985e-06,
      "loss": 2.512,
      "step": 2055
    },
    {
      "epoch": 0.82,
      "grad_norm": 6.709656715393066,
      "learning_rate": 3.9607873614261715e-06,
      "loss": 3.0421,
      "step": 2060
    },
    {
      "epoch": 0.83,
      "grad_norm": 13.18295669555664,
      "learning_rate": 3.8746421847749765e-06,
      "loss": 2.4467,
      "step": 2065
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.89919662475586,
      "learning_rate": 3.789365388841193e-06,
      "loss": 2.322,
      "step": 2070
    },
    {
      "epoch": 0.83,
      "grad_norm": 20.111215591430664,
      "learning_rate": 3.7049604790198976e-06,
      "loss": 3.2308,
      "step": 2075
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.88414192199707,
      "learning_rate": 3.621430924866348e-06,
      "loss": 2.5939,
      "step": 2080
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.166717529296875,
      "learning_rate": 3.5387801599533475e-06,
      "loss": 2.9439,
      "step": 2085
    },
    {
      "epoch": 0.84,
      "grad_norm": 31.1898193359375,
      "learning_rate": 3.4570115817301243e-06,
      "loss": 3.2292,
      "step": 2090
    },
    {
      "epoch": 0.84,
      "grad_norm": 26.66686248779297,
      "learning_rate": 3.3761285513826625e-06,
      "loss": 2.9332,
      "step": 2095
    },
    {
      "epoch": 0.84,
      "grad_norm": 4.934303283691406,
      "learning_rate": 3.296134393695538e-06,
      "loss": 2.7759,
      "step": 2100
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.340413093566895,
      "learning_rate": 3.217032396915265e-06,
      "loss": 2.3886,
      "step": 2105
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.123784065246582,
      "learning_rate": 3.1388258126151093e-06,
      "loss": 2.6871,
      "step": 2110
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.508092880249023,
      "learning_rate": 3.06151785556143e-06,
      "loss": 3.0332,
      "step": 2115
    },
    {
      "epoch": 0.85,
      "grad_norm": 6.981659412384033,
      "learning_rate": 2.98511170358155e-06,
      "loss": 2.6556,
      "step": 2120
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.004157066345215,
      "learning_rate": 2.9096104974331184e-06,
      "loss": 2.7863,
      "step": 2125
    },
    {
      "epoch": 0.85,
      "grad_norm": 15.677950859069824,
      "learning_rate": 2.8350173406749973e-06,
      "loss": 2.7697,
      "step": 2130
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.333318710327148,
      "learning_rate": 2.7613352995397078e-06,
      "loss": 2.6726,
      "step": 2135
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.832099437713623,
      "learning_rate": 2.688567402807357e-06,
      "loss": 2.8863,
      "step": 2140
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.78770637512207,
      "learning_rate": 2.6167166416811746e-06,
      "loss": 2.8584,
      "step": 2145
    },
    {
      "epoch": 0.86,
      "grad_norm": 16.921499252319336,
      "learning_rate": 2.545785969664524e-06,
      "loss": 2.9597,
      "step": 2150
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.275824546813965,
      "learning_rate": 2.475778302439524e-06,
      "loss": 2.1545,
      "step": 2155
    },
    {
      "epoch": 0.86,
      "grad_norm": 20.255998611450195,
      "learning_rate": 2.4066965177471645e-06,
      "loss": 2.5138,
      "step": 2160
    },
    {
      "epoch": 0.87,
      "grad_norm": 6.863336086273193,
      "learning_rate": 2.338543455269046e-06,
      "loss": 2.5736,
      "step": 2165
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.660894393920898,
      "learning_rate": 2.271321916510627e-06,
      "loss": 2.5524,
      "step": 2170
    },
    {
      "epoch": 0.87,
      "grad_norm": 13.18586254119873,
      "learning_rate": 2.205034664686076e-06,
      "loss": 2.5716,
      "step": 2175
    },
    {
      "epoch": 0.87,
      "grad_norm": 11.016261100769043,
      "learning_rate": 2.1396844246046903e-06,
      "loss": 3.3971,
      "step": 2180
    },
    {
      "epoch": 0.87,
      "grad_norm": 23.928808212280273,
      "learning_rate": 2.075273882558873e-06,
      "loss": 2.7627,
      "step": 2185
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.036683082580566,
      "learning_rate": 2.0118056862137357e-06,
      "loss": 2.7452,
      "step": 2190
    },
    {
      "epoch": 0.88,
      "grad_norm": 15.620336532592773,
      "learning_rate": 1.949282444498238e-06,
      "loss": 2.7892,
      "step": 2195
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.829482078552246,
      "learning_rate": 1.8877067274979648e-06,
      "loss": 2.7427,
      "step": 2200
    },
    {
      "epoch": 0.88,
      "grad_norm": 24.735841751098633,
      "learning_rate": 1.827081066349459e-06,
      "loss": 2.3911,
      "step": 2205
    },
    {
      "epoch": 0.88,
      "grad_norm": 15.522488594055176,
      "learning_rate": 1.767407953136202e-06,
      "loss": 2.7896,
      "step": 2210
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.789400100708008,
      "learning_rate": 1.7086898407861485e-06,
      "loss": 2.6058,
      "step": 2215
    },
    {
      "epoch": 0.89,
      "grad_norm": 12.961084365844727,
      "learning_rate": 1.6509291429709223e-06,
      "loss": 2.1545,
      "step": 2220
    },
    {
      "epoch": 0.89,
      "grad_norm": 42.19524002075195,
      "learning_rate": 1.59412823400657e-06,
      "loss": 2.7638,
      "step": 2225
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.88717269897461,
      "learning_rate": 1.538289448755989e-06,
      "loss": 3.258,
      "step": 2230
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.746241569519043,
      "learning_rate": 1.483415082532938e-06,
      "loss": 2.4654,
      "step": 2235
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.284168243408203,
      "learning_rate": 1.4295073910076757e-06,
      "loss": 2.8934,
      "step": 2240
    },
    {
      "epoch": 0.9,
      "grad_norm": 6.290571212768555,
      "learning_rate": 1.3765685901142716e-06,
      "loss": 2.8615,
      "step": 2245
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.758255004882812,
      "learning_rate": 1.3246008559594709e-06,
      "loss": 2.4915,
      "step": 2250
    },
    {
      "epoch": 0.9,
      "grad_norm": 64.08142852783203,
      "learning_rate": 1.273606324733284e-06,
      "loss": 2.6644,
      "step": 2255
    },
    {
      "epoch": 0.9,
      "grad_norm": 6.796569347381592,
      "learning_rate": 1.2235870926211619e-06,
      "loss": 3.0118,
      "step": 2260
    },
    {
      "epoch": 0.91,
      "grad_norm": 48.14067459106445,
      "learning_rate": 1.1745452157178206e-06,
      "loss": 2.5428,
      "step": 2265
    },
    {
      "epoch": 0.91,
      "grad_norm": 78.56604766845703,
      "learning_rate": 1.1264827099427417e-06,
      "loss": 3.2121,
      "step": 2270
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.145052909851074,
      "learning_rate": 1.0794015509572818e-06,
      "loss": 2.7715,
      "step": 2275
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.410415649414062,
      "learning_rate": 1.0333036740834856e-06,
      "loss": 2.9311,
      "step": 2280
    },
    {
      "epoch": 0.91,
      "grad_norm": 26.370338439941406,
      "learning_rate": 9.881909742245177e-07,
      "loss": 2.8867,
      "step": 2285
    },
    {
      "epoch": 0.92,
      "grad_norm": 19.1849365234375,
      "learning_rate": 9.440653057867815e-07,
      "loss": 2.559,
      "step": 2290
    },
    {
      "epoch": 0.92,
      "grad_norm": 25.568355560302734,
      "learning_rate": 9.009284826036691e-07,
      "loss": 2.5787,
      "step": 2295
    },
    {
      "epoch": 0.92,
      "grad_norm": 14.307110786437988,
      "learning_rate": 8.587822778610283e-07,
      "loss": 2.8058,
      "step": 2300
    },
    {
      "epoch": 0.92,
      "grad_norm": 25.297014236450195,
      "learning_rate": 8.176284240242638e-07,
      "loss": 2.8921,
      "step": 2305
    },
    {
      "epoch": 0.92,
      "grad_norm": 19.239059448242188,
      "learning_rate": 7.774686127671183e-07,
      "loss": 2.6122,
      "step": 2310
    },
    {
      "epoch": 0.93,
      "grad_norm": 11.51448917388916,
      "learning_rate": 7.383044949021339e-07,
      "loss": 2.2361,
      "step": 2315
    },
    {
      "epoch": 0.93,
      "grad_norm": 6.253133773803711,
      "learning_rate": 7.00137680312804e-07,
      "loss": 2.5475,
      "step": 2320
    },
    {
      "epoch": 0.93,
      "grad_norm": 17.918170928955078,
      "learning_rate": 6.62969737887384e-07,
      "loss": 2.8589,
      "step": 2325
    },
    {
      "epoch": 0.93,
      "grad_norm": 15.065885543823242,
      "learning_rate": 6.268021954544096e-07,
      "loss": 2.9153,
      "step": 2330
    },
    {
      "epoch": 0.93,
      "grad_norm": 13.842984199523926,
      "learning_rate": 5.916365397198975e-07,
      "loss": 2.894,
      "step": 2335
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.621698379516602,
      "learning_rate": 5.574742162062163e-07,
      "loss": 3.1294,
      "step": 2340
    },
    {
      "epoch": 0.94,
      "grad_norm": 13.219593048095703,
      "learning_rate": 5.243166291926782e-07,
      "loss": 2.6543,
      "step": 2345
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.599324226379395,
      "learning_rate": 4.921651416578188e-07,
      "loss": 2.3677,
      "step": 2350
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.731278419494629,
      "learning_rate": 4.6102107522336403e-07,
      "loss": 2.7634,
      "step": 2355
    },
    {
      "epoch": 0.94,
      "grad_norm": 43.908077239990234,
      "learning_rate": 4.308857100999042e-07,
      "loss": 2.8106,
      "step": 2360
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.233963966369629,
      "learning_rate": 4.0176028503425835e-07,
      "loss": 2.5252,
      "step": 2365
    },
    {
      "epoch": 0.95,
      "grad_norm": 13.496291160583496,
      "learning_rate": 3.7364599725858153e-07,
      "loss": 2.2908,
      "step": 2370
    },
    {
      "epoch": 0.95,
      "grad_norm": 11.278969764709473,
      "learning_rate": 3.465440024411265e-07,
      "loss": 2.5045,
      "step": 2375
    },
    {
      "epoch": 0.95,
      "grad_norm": 13.121251106262207,
      "learning_rate": 3.204554146387456e-07,
      "loss": 2.4661,
      "step": 2380
    },
    {
      "epoch": 0.95,
      "grad_norm": 17.475078582763672,
      "learning_rate": 2.9538130625110796e-07,
      "loss": 2.9907,
      "step": 2385
    },
    {
      "epoch": 0.96,
      "grad_norm": 6.753970146179199,
      "learning_rate": 2.7132270797659563e-07,
      "loss": 2.4326,
      "step": 2390
    },
    {
      "epoch": 0.96,
      "grad_norm": 18.052387237548828,
      "learning_rate": 2.482806087699546e-07,
      "loss": 3.0056,
      "step": 2395
    },
    {
      "epoch": 0.96,
      "grad_norm": 11.094266891479492,
      "learning_rate": 2.262559558016325e-07,
      "loss": 2.7058,
      "step": 2400
    }
  ],
  "logging_steps": 5,
  "max_steps": 2500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 200,
  "total_flos": 1.3202552777146368e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}