{
  "best_global_step": 10176,
  "best_metric": 0.5032461881637573,
  "best_model_checkpoint": "saves_multiple/prefix-tuning/llama-3-8b-instruct/train_codealpacapy_789_1767616167/checkpoint-10176",
  "epoch": 20.0,
  "eval_steps": 3392,
  "global_step": 33920,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00294811320754717,
      "grad_norm": 89.86528015136719,
      "learning_rate": 1.179245283018868e-08,
      "loss": 6.8675,
      "num_input_tokens_seen": 3616,
      "step": 5
    },
    {
      "epoch": 0.00589622641509434,
      "grad_norm": 73.07382202148438,
      "learning_rate": 2.6533018867924528e-08,
      "loss": 6.7515,
      "num_input_tokens_seen": 6592,
      "step": 10
    },
    {
      "epoch": 0.00884433962264151,
      "grad_norm": 82.73033905029297,
      "learning_rate": 4.127358490566038e-08,
      "loss": 6.7148,
      "num_input_tokens_seen": 10016,
      "step": 15
    },
    {
      "epoch": 0.01179245283018868,
      "grad_norm": 97.4770736694336,
      "learning_rate": 5.601415094339623e-08,
      "loss": 6.954,
      "num_input_tokens_seen": 12832,
      "step": 20
    },
    {
      "epoch": 0.01474056603773585,
      "grad_norm": 88.6093978881836,
      "learning_rate": 7.075471698113208e-08,
      "loss": 6.9647,
      "num_input_tokens_seen": 16960,
      "step": 25
    },
    {
      "epoch": 0.01768867924528302,
      "grad_norm": 71.94873809814453,
      "learning_rate": 8.549528301886793e-08,
      "loss": 6.5597,
      "num_input_tokens_seen": 20192,
      "step": 30
    },
    {
      "epoch": 0.020636792452830188,
      "grad_norm": 83.42150115966797,
      "learning_rate": 1.0023584905660378e-07,
      "loss": 6.9421,
      "num_input_tokens_seen": 22688,
      "step": 35
    },
    {
      "epoch": 0.02358490566037736,
      "grad_norm": 78.6876220703125,
      "learning_rate": 1.1497641509433962e-07,
      "loss": 6.5901,
      "num_input_tokens_seen": 25696,
      "step": 40
    },
    {
      "epoch": 0.02653301886792453,
      "grad_norm": 78.13228607177734,
      "learning_rate": 1.297169811320755e-07,
      "loss": 6.9363,
      "num_input_tokens_seen": 29088,
      "step": 45
    },
    {
      "epoch": 0.0294811320754717,
      "grad_norm": 65.15396118164062,
      "learning_rate": 1.4445754716981135e-07,
      "loss": 6.7409,
      "num_input_tokens_seen": 31936,
      "step": 50
    },
    {
      "epoch": 0.03242924528301887,
      "grad_norm": 71.11312103271484,
      "learning_rate": 1.591981132075472e-07,
      "loss": 6.3382,
      "num_input_tokens_seen": 35104,
      "step": 55
    },
    {
      "epoch": 0.03537735849056604,
      "grad_norm": 61.796791076660156,
      "learning_rate": 1.7393867924528304e-07,
      "loss": 6.3629,
      "num_input_tokens_seen": 37728,
      "step": 60
    },
    {
      "epoch": 0.038325471698113206,
      "grad_norm": 67.4417953491211,
      "learning_rate": 1.886792452830189e-07,
      "loss": 6.5626,
      "num_input_tokens_seen": 40448,
      "step": 65
    },
    {
      "epoch": 0.041273584905660375,
      "grad_norm": 54.12550354003906,
      "learning_rate": 2.0341981132075473e-07,
      "loss": 6.2755,
      "num_input_tokens_seen": 43232,
      "step": 70
    },
    {
      "epoch": 0.044221698113207544,
      "grad_norm": 68.11248779296875,
      "learning_rate": 2.1816037735849058e-07,
      "loss": 6.3254,
      "num_input_tokens_seen": 46688,
      "step": 75
    },
    {
      "epoch": 0.04716981132075472,
      "grad_norm": 61.58484649658203,
      "learning_rate": 2.3290094339622643e-07,
      "loss": 6.1288,
      "num_input_tokens_seen": 49248,
      "step": 80
    },
    {
      "epoch": 0.05011792452830189,
      "grad_norm": 62.620296478271484,
      "learning_rate": 2.476415094339623e-07,
      "loss": 6.0135,
      "num_input_tokens_seen": 52800,
      "step": 85
    },
    {
      "epoch": 0.05306603773584906,
      "grad_norm": 55.138328552246094,
      "learning_rate": 2.6238207547169815e-07,
      "loss": 5.6154,
      "num_input_tokens_seen": 56928,
      "step": 90
    },
    {
      "epoch": 0.05601415094339623,
      "grad_norm": 69.87220764160156,
      "learning_rate": 2.7712264150943397e-07,
      "loss": 5.726,
      "num_input_tokens_seen": 61152,
      "step": 95
    },
    {
      "epoch": 0.0589622641509434,
      "grad_norm": 73.03726196289062,
      "learning_rate": 2.9186320754716984e-07,
      "loss": 5.5123,
      "num_input_tokens_seen": 63712,
      "step": 100
    },
    {
      "epoch": 0.061910377358490566,
      "grad_norm": 56.68369674682617,
      "learning_rate": 3.0660377358490567e-07,
      "loss": 5.5062,
      "num_input_tokens_seen": 67104,
      "step": 105
    },
    {
      "epoch": 0.06485849056603774,
      "grad_norm": 52.18683624267578,
      "learning_rate": 3.213443396226416e-07,
      "loss": 5.4964,
      "num_input_tokens_seen": 70272,
      "step": 110
    },
    {
      "epoch": 0.06780660377358491,
      "grad_norm": 65.91887664794922,
      "learning_rate": 3.3608490566037736e-07,
      "loss": 5.2631,
      "num_input_tokens_seen": 73248,
      "step": 115
    },
    {
      "epoch": 0.07075471698113207,
      "grad_norm": 58.06288528442383,
      "learning_rate": 3.508254716981133e-07,
      "loss": 5.3147,
      "num_input_tokens_seen": 76512,
      "step": 120
    },
    {
      "epoch": 0.07370283018867925,
      "grad_norm": 65.29225158691406,
      "learning_rate": 3.6556603773584905e-07,
      "loss": 5.1288,
      "num_input_tokens_seen": 79136,
      "step": 125
    },
    {
      "epoch": 0.07665094339622641,
      "grad_norm": 50.588375091552734,
      "learning_rate": 3.80306603773585e-07,
      "loss": 5.2264,
      "num_input_tokens_seen": 81984,
      "step": 130
    },
    {
      "epoch": 0.07959905660377359,
      "grad_norm": 38.79189682006836,
      "learning_rate": 3.9504716981132075e-07,
      "loss": 5.3419,
      "num_input_tokens_seen": 84576,
      "step": 135
    },
    {
      "epoch": 0.08254716981132075,
      "grad_norm": 45.83066940307617,
      "learning_rate": 4.097877358490567e-07,
      "loss": 4.7733,
      "num_input_tokens_seen": 87360,
      "step": 140
    },
    {
      "epoch": 0.08549528301886793,
      "grad_norm": 44.60704803466797,
      "learning_rate": 4.2452830188679244e-07,
      "loss": 4.5885,
      "num_input_tokens_seen": 90336,
      "step": 145
    },
    {
      "epoch": 0.08844339622641509,
      "grad_norm": 35.25266647338867,
      "learning_rate": 4.3926886792452837e-07,
      "loss": 4.6357,
      "num_input_tokens_seen": 94848,
      "step": 150
    },
    {
      "epoch": 0.09139150943396226,
      "grad_norm": 32.92588424682617,
      "learning_rate": 4.5400943396226414e-07,
      "loss": 4.4656,
      "num_input_tokens_seen": 98528,
      "step": 155
    },
    {
      "epoch": 0.09433962264150944,
      "grad_norm": 40.72951889038086,
      "learning_rate": 4.6875000000000006e-07,
      "loss": 4.2239,
      "num_input_tokens_seen": 101632,
      "step": 160
    },
    {
      "epoch": 0.0972877358490566,
      "grad_norm": 32.97052764892578,
      "learning_rate": 4.834905660377359e-07,
      "loss": 4.7877,
      "num_input_tokens_seen": 104768,
      "step": 165
    },
    {
      "epoch": 0.10023584905660378,
      "grad_norm": 39.745235443115234,
      "learning_rate": 4.982311320754717e-07,
      "loss": 4.54,
      "num_input_tokens_seen": 107648,
      "step": 170
    },
    {
      "epoch": 0.10318396226415094,
      "grad_norm": 34.851219177246094,
      "learning_rate": 5.129716981132076e-07,
      "loss": 4.4622,
      "num_input_tokens_seen": 110368,
      "step": 175
    },
    {
      "epoch": 0.10613207547169812,
      "grad_norm": 31.43199920654297,
      "learning_rate": 5.277122641509435e-07,
      "loss": 4.4651,
      "num_input_tokens_seen": 113600,
      "step": 180
    },
    {
      "epoch": 0.10908018867924528,
      "grad_norm": 33.36195373535156,
      "learning_rate": 5.424528301886793e-07,
      "loss": 4.4753,
      "num_input_tokens_seen": 116512,
      "step": 185
    },
    {
      "epoch": 0.11202830188679246,
      "grad_norm": 36.767112731933594,
      "learning_rate": 5.571933962264151e-07,
      "loss": 3.977,
      "num_input_tokens_seen": 119968,
      "step": 190
    },
    {
      "epoch": 0.11497641509433962,
      "grad_norm": 47.47268295288086,
      "learning_rate": 5.71933962264151e-07,
      "loss": 4.86,
      "num_input_tokens_seen": 123712,
      "step": 195
    },
    {
      "epoch": 0.1179245283018868,
      "grad_norm": 43.8890495300293,
      "learning_rate": 5.866745283018868e-07,
      "loss": 4.3363,
      "num_input_tokens_seen": 127040,
      "step": 200
    },
    {
      "epoch": 0.12087264150943396,
      "grad_norm": 28.31290054321289,
      "learning_rate": 6.014150943396227e-07,
      "loss": 4.1935,
      "num_input_tokens_seen": 130912,
      "step": 205
    },
    {
      "epoch": 0.12382075471698113,
      "grad_norm": 38.36419677734375,
      "learning_rate": 6.161556603773585e-07,
      "loss": 3.8832,
      "num_input_tokens_seen": 133472,
      "step": 210
    },
    {
      "epoch": 0.1267688679245283,
      "grad_norm": 39.80194854736328,
      "learning_rate": 6.308962264150945e-07,
      "loss": 3.8396,
      "num_input_tokens_seen": 137088,
      "step": 215
    },
    {
      "epoch": 0.12971698113207547,
      "grad_norm": 33.31916046142578,
      "learning_rate": 6.456367924528302e-07,
      "loss": 3.8179,
      "num_input_tokens_seen": 139872,
      "step": 220
    },
    {
      "epoch": 0.13266509433962265,
      "grad_norm": 29.037561416625977,
      "learning_rate": 6.603773584905661e-07,
      "loss": 3.9771,
      "num_input_tokens_seen": 143200,
      "step": 225
    },
    {
      "epoch": 0.13561320754716982,
      "grad_norm": 21.75967788696289,
      "learning_rate": 6.75117924528302e-07,
      "loss": 3.7566,
      "num_input_tokens_seen": 146816,
      "step": 230
    },
    {
      "epoch": 0.13856132075471697,
      "grad_norm": 33.47700119018555,
      "learning_rate": 6.898584905660379e-07,
      "loss": 3.6073,
      "num_input_tokens_seen": 150336,
      "step": 235
    },
    {
      "epoch": 0.14150943396226415,
      "grad_norm": 40.323577880859375,
      "learning_rate": 7.045990566037736e-07,
      "loss": 3.8868,
      "num_input_tokens_seen": 152736,
      "step": 240
    },
    {
      "epoch": 0.14445754716981132,
      "grad_norm": 35.50404357910156,
      "learning_rate": 7.193396226415095e-07,
      "loss": 3.8677,
      "num_input_tokens_seen": 155872,
      "step": 245
    },
    {
      "epoch": 0.1474056603773585,
      "grad_norm": 24.8291015625,
      "learning_rate": 7.340801886792454e-07,
      "loss": 3.6325,
      "num_input_tokens_seen": 158432,
      "step": 250
    },
    {
      "epoch": 0.15035377358490565,
      "grad_norm": 22.234479904174805,
      "learning_rate": 7.488207547169812e-07,
      "loss": 3.7341,
      "num_input_tokens_seen": 161248,
      "step": 255
    },
    {
      "epoch": 0.15330188679245282,
      "grad_norm": 39.227413177490234,
      "learning_rate": 7.63561320754717e-07,
      "loss": 4.1007,
      "num_input_tokens_seen": 163968,
      "step": 260
    },
    {
      "epoch": 0.15625,
      "grad_norm": 27.64311981201172,
      "learning_rate": 7.783018867924529e-07,
      "loss": 3.3688,
      "num_input_tokens_seen": 167040,
      "step": 265
    },
    {
      "epoch": 0.15919811320754718,
      "grad_norm": 24.19176483154297,
      "learning_rate": 7.930424528301888e-07,
      "loss": 3.7551,
      "num_input_tokens_seen": 170208,
      "step": 270
    },
    {
      "epoch": 0.16214622641509435,
      "grad_norm": 30.01333999633789,
      "learning_rate": 8.077830188679246e-07,
      "loss": 3.5074,
      "num_input_tokens_seen": 173280,
      "step": 275
    },
    {
      "epoch": 0.1650943396226415,
      "grad_norm": 37.98541259765625,
      "learning_rate": 8.225235849056605e-07,
      "loss": 4.2056,
      "num_input_tokens_seen": 176064,
      "step": 280
    },
    {
      "epoch": 0.16804245283018868,
      "grad_norm": 23.61954116821289,
      "learning_rate": 8.372641509433963e-07,
      "loss": 3.6523,
      "num_input_tokens_seen": 178848,
      "step": 285
    },
    {
      "epoch": 0.17099056603773585,
      "grad_norm": 27.868898391723633,
      "learning_rate": 8.520047169811321e-07,
      "loss": 3.0967,
      "num_input_tokens_seen": 181824,
      "step": 290
    },
    {
      "epoch": 0.17393867924528303,
      "grad_norm": 22.540626525878906,
      "learning_rate": 8.66745283018868e-07,
      "loss": 3.1963,
      "num_input_tokens_seen": 184832,
      "step": 295
    },
    {
      "epoch": 0.17688679245283018,
      "grad_norm": 29.870311737060547,
      "learning_rate": 8.814858490566039e-07,
      "loss": 3.324,
      "num_input_tokens_seen": 187872,
      "step": 300
    },
    {
      "epoch": 0.17983490566037735,
      "grad_norm": 35.44401550292969,
      "learning_rate": 8.962264150943397e-07,
      "loss": 3.5791,
      "num_input_tokens_seen": 190816,
      "step": 305
    },
    {
      "epoch": 0.18278301886792453,
      "grad_norm": 22.058258056640625,
      "learning_rate": 9.109669811320755e-07,
      "loss": 3.1697,
      "num_input_tokens_seen": 194752,
      "step": 310
    },
    {
      "epoch": 0.1857311320754717,
      "grad_norm": 17.901290893554688,
      "learning_rate": 9.257075471698114e-07,
      "loss": 3.1868,
      "num_input_tokens_seen": 197920,
      "step": 315
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 30.084461212158203,
      "learning_rate": 9.404481132075473e-07,
      "loss": 3.4802,
      "num_input_tokens_seen": 200640,
      "step": 320
    },
    {
      "epoch": 0.19162735849056603,
      "grad_norm": 27.698747634887695,
      "learning_rate": 9.551886792452833e-07,
      "loss": 2.8794,
      "num_input_tokens_seen": 203584,
      "step": 325
    },
    {
      "epoch": 0.1945754716981132,
      "grad_norm": 17.27296257019043,
      "learning_rate": 9.699292452830188e-07,
      "loss": 3.2775,
      "num_input_tokens_seen": 206912,
      "step": 330
    },
    {
      "epoch": 0.19752358490566038,
      "grad_norm": 24.726953506469727,
      "learning_rate": 9.846698113207548e-07,
      "loss": 3.1551,
      "num_input_tokens_seen": 209760,
      "step": 335
    },
    {
      "epoch": 0.20047169811320756,
      "grad_norm": 15.377513885498047,
      "learning_rate": 9.994103773584906e-07,
      "loss": 2.9782,
      "num_input_tokens_seen": 214400,
      "step": 340
    },
    {
      "epoch": 0.2034198113207547,
      "grad_norm": 18.147586822509766,
      "learning_rate": 1.0141509433962265e-06,
      "loss": 2.9856,
      "num_input_tokens_seen": 217024,
      "step": 345
    },
    {
      "epoch": 0.20636792452830188,
      "grad_norm": 28.250747680664062,
      "learning_rate": 1.0288915094339623e-06,
      "loss": 3.3291,
      "num_input_tokens_seen": 220192,
      "step": 350
    },
    {
      "epoch": 0.20931603773584906,
      "grad_norm": 18.272212982177734,
      "learning_rate": 1.043632075471698e-06,
      "loss": 2.7321,
      "num_input_tokens_seen": 223488,
      "step": 355
    },
    {
      "epoch": 0.21226415094339623,
      "grad_norm": 24.706192016601562,
      "learning_rate": 1.058372641509434e-06,
      "loss": 2.5192,
      "num_input_tokens_seen": 226496,
      "step": 360
    },
    {
      "epoch": 0.21521226415094338,
      "grad_norm": 18.464101791381836,
      "learning_rate": 1.07311320754717e-06,
      "loss": 2.9645,
      "num_input_tokens_seen": 229600,
      "step": 365
    },
    {
      "epoch": 0.21816037735849056,
      "grad_norm": 16.27822494506836,
      "learning_rate": 1.0878537735849056e-06,
      "loss": 2.5926,
      "num_input_tokens_seen": 232512,
      "step": 370
    },
    {
      "epoch": 0.22110849056603774,
      "grad_norm": 17.1165771484375,
      "learning_rate": 1.1025943396226416e-06,
      "loss": 3.0474,
      "num_input_tokens_seen": 236000,
      "step": 375
    },
    {
      "epoch": 0.2240566037735849,
      "grad_norm": 24.01494598388672,
      "learning_rate": 1.1173349056603773e-06,
      "loss": 3.1436,
      "num_input_tokens_seen": 239072,
      "step": 380
    },
    {
      "epoch": 0.2270047169811321,
      "grad_norm": 19.342748641967773,
      "learning_rate": 1.1320754716981133e-06,
      "loss": 2.7559,
      "num_input_tokens_seen": 242112,
      "step": 385
    },
    {
      "epoch": 0.22995283018867924,
      "grad_norm": 20.03110122680664,
      "learning_rate": 1.1468160377358493e-06,
      "loss": 2.9626,
      "num_input_tokens_seen": 245440,
      "step": 390
    },
    {
      "epoch": 0.2329009433962264,
      "grad_norm": 24.672164916992188,
      "learning_rate": 1.1615566037735849e-06,
      "loss": 3.0031,
      "num_input_tokens_seen": 247648,
      "step": 395
    },
    {
      "epoch": 0.2358490566037736,
      "grad_norm": 19.929767608642578,
      "learning_rate": 1.1762971698113208e-06,
      "loss": 2.794,
      "num_input_tokens_seen": 250816,
      "step": 400
    },
    {
      "epoch": 0.23879716981132076,
      "grad_norm": 41.975830078125,
      "learning_rate": 1.1910377358490568e-06,
      "loss": 2.7938,
      "num_input_tokens_seen": 253792,
      "step": 405
    },
    {
      "epoch": 0.2417452830188679,
      "grad_norm": 24.103593826293945,
      "learning_rate": 1.2057783018867926e-06,
      "loss": 2.802,
      "num_input_tokens_seen": 256704,
      "step": 410
    },
    {
      "epoch": 0.2446933962264151,
      "grad_norm": 25.269426345825195,
      "learning_rate": 1.2205188679245284e-06,
      "loss": 2.5662,
      "num_input_tokens_seen": 259456,
      "step": 415
    },
    {
      "epoch": 0.24764150943396226,
      "grad_norm": 15.406280517578125,
      "learning_rate": 1.2352594339622641e-06,
      "loss": 2.4915,
      "num_input_tokens_seen": 263456,
      "step": 420
    },
    {
      "epoch": 0.2505896226415094,
      "grad_norm": 20.058250427246094,
      "learning_rate": 1.25e-06,
      "loss": 2.6517,
      "num_input_tokens_seen": 266400,
      "step": 425
    },
    {
      "epoch": 0.2535377358490566,
      "grad_norm": 17.920419692993164,
      "learning_rate": 1.264740566037736e-06,
      "loss": 2.7024,
      "num_input_tokens_seen": 270080,
      "step": 430
    },
    {
      "epoch": 0.25648584905660377,
      "grad_norm": 17.289133071899414,
      "learning_rate": 1.2794811320754718e-06,
      "loss": 2.535,
      "num_input_tokens_seen": 274016,
      "step": 435
    },
    {
      "epoch": 0.25943396226415094,
      "grad_norm": 25.444141387939453,
      "learning_rate": 1.2942216981132078e-06,
      "loss": 2.7175,
      "num_input_tokens_seen": 276640,
      "step": 440
    },
    {
      "epoch": 0.2623820754716981,
      "grad_norm": 19.11880111694336,
      "learning_rate": 1.3089622641509436e-06,
      "loss": 2.5607,
      "num_input_tokens_seen": 279712,
      "step": 445
    },
    {
      "epoch": 0.2653301886792453,
      "grad_norm": 26.445039749145508,
      "learning_rate": 1.3237028301886792e-06,
      "loss": 2.8851,
      "num_input_tokens_seen": 282144,
      "step": 450
    },
    {
      "epoch": 0.26827830188679247,
      "grad_norm": 17.771923065185547,
      "learning_rate": 1.3384433962264151e-06,
      "loss": 2.3583,
      "num_input_tokens_seen": 285344,
      "step": 455
    },
    {
      "epoch": 0.27122641509433965,
      "grad_norm": 23.78785514831543,
      "learning_rate": 1.353183962264151e-06,
      "loss": 2.683,
      "num_input_tokens_seen": 287840,
      "step": 460
    },
    {
      "epoch": 0.27417452830188677,
      "grad_norm": 23.026071548461914,
      "learning_rate": 1.3679245283018869e-06,
      "loss": 2.4999,
      "num_input_tokens_seen": 290816,
      "step": 465
    },
    {
      "epoch": 0.27712264150943394,
      "grad_norm": 19.9582462310791,
      "learning_rate": 1.3826650943396229e-06,
      "loss": 2.506,
      "num_input_tokens_seen": 294560,
      "step": 470
    },
    {
      "epoch": 0.2800707547169811,
      "grad_norm": 15.065876007080078,
      "learning_rate": 1.3974056603773586e-06,
      "loss": 2.419,
      "num_input_tokens_seen": 297376,
      "step": 475
    },
    {
      "epoch": 0.2830188679245283,
      "grad_norm": 22.934001922607422,
      "learning_rate": 1.4121462264150946e-06,
      "loss": 2.2965,
      "num_input_tokens_seen": 301184,
      "step": 480
    },
    {
      "epoch": 0.28596698113207547,
      "grad_norm": 19.639673233032227,
      "learning_rate": 1.4268867924528304e-06,
      "loss": 2.8833,
      "num_input_tokens_seen": 304704,
      "step": 485
    },
    {
      "epoch": 0.28891509433962265,
      "grad_norm": 19.375682830810547,
      "learning_rate": 1.4416273584905664e-06,
      "loss": 2.6632,
      "num_input_tokens_seen": 307904,
      "step": 490
    },
    {
      "epoch": 0.2918632075471698,
      "grad_norm": 26.016510009765625,
      "learning_rate": 1.456367924528302e-06,
      "loss": 2.4549,
      "num_input_tokens_seen": 311456,
      "step": 495
    },
    {
      "epoch": 0.294811320754717,
      "grad_norm": 16.236494064331055,
      "learning_rate": 1.4711084905660377e-06,
      "loss": 2.5223,
      "num_input_tokens_seen": 315040,
      "step": 500
    },
    {
      "epoch": 0.2977594339622642,
      "grad_norm": 18.02973747253418,
      "learning_rate": 1.4858490566037737e-06,
      "loss": 2.2151,
      "num_input_tokens_seen": 317312,
      "step": 505
    },
    {
      "epoch": 0.3007075471698113,
      "grad_norm": 20.331663131713867,
      "learning_rate": 1.5005896226415096e-06,
      "loss": 2.0399,
      "num_input_tokens_seen": 323264,
      "step": 510
    },
    {
      "epoch": 0.30365566037735847,
      "grad_norm": 22.76030731201172,
      "learning_rate": 1.5153301886792454e-06,
      "loss": 2.2013,
      "num_input_tokens_seen": 326752,
      "step": 515
    },
    {
      "epoch": 0.30660377358490565,
      "grad_norm": 16.37139129638672,
      "learning_rate": 1.5300707547169814e-06,
      "loss": 2.1582,
      "num_input_tokens_seen": 330240,
      "step": 520
    },
    {
      "epoch": 0.3095518867924528,
      "grad_norm": 15.143797874450684,
      "learning_rate": 1.5448113207547172e-06,
      "loss": 2.1456,
      "num_input_tokens_seen": 334080,
      "step": 525
    },
    {
      "epoch": 0.3125,
      "grad_norm": 16.825674057006836,
      "learning_rate": 1.5595518867924531e-06,
      "loss": 2.143,
      "num_input_tokens_seen": 337088,
      "step": 530
    },
    {
      "epoch": 0.3154481132075472,
      "grad_norm": 19.779773712158203,
      "learning_rate": 1.574292452830189e-06,
      "loss": 1.9832,
      "num_input_tokens_seen": 339776,
      "step": 535
    },
    {
      "epoch": 0.31839622641509435,
      "grad_norm": 17.15123748779297,
      "learning_rate": 1.5890330188679245e-06,
      "loss": 2.3305,
      "num_input_tokens_seen": 342016,
      "step": 540
    },
    {
      "epoch": 0.32134433962264153,
      "grad_norm": 26.49778938293457,
      "learning_rate": 1.6037735849056604e-06,
      "loss": 2.0534,
      "num_input_tokens_seen": 344928,
      "step": 545
    },
    {
      "epoch": 0.3242924528301887,
      "grad_norm": 25.75377655029297,
      "learning_rate": 1.6185141509433964e-06,
      "loss": 1.9088,
      "num_input_tokens_seen": 347104,
      "step": 550
    },
    {
      "epoch": 0.3272405660377358,
      "grad_norm": 20.026758193969727,
      "learning_rate": 1.6332547169811322e-06,
      "loss": 2.038,
      "num_input_tokens_seen": 350176,
      "step": 555
    },
    {
      "epoch": 0.330188679245283,
      "grad_norm": 17.94679069519043,
      "learning_rate": 1.6479952830188682e-06,
      "loss": 2.1059,
      "num_input_tokens_seen": 353120,
      "step": 560
    },
    {
      "epoch": 0.3331367924528302,
      "grad_norm": 13.637446403503418,
      "learning_rate": 1.662735849056604e-06,
      "loss": 2.179,
      "num_input_tokens_seen": 360352,
      "step": 565
    },
    {
      "epoch": 0.33608490566037735,
      "grad_norm": 17.031158447265625,
      "learning_rate": 1.67747641509434e-06,
      "loss": 1.9942,
      "num_input_tokens_seen": 363520,
      "step": 570
    },
    {
      "epoch": 0.33903301886792453,
      "grad_norm": 17.807815551757812,
      "learning_rate": 1.6922169811320757e-06,
      "loss": 2.1322,
      "num_input_tokens_seen": 366720,
      "step": 575
    },
    {
      "epoch": 0.3419811320754717,
      "grad_norm": 21.22860336303711,
      "learning_rate": 1.7069575471698112e-06,
      "loss": 1.9788,
      "num_input_tokens_seen": 369920,
      "step": 580
    },
    {
      "epoch": 0.3449292452830189,
      "grad_norm": 18.07029914855957,
      "learning_rate": 1.7216981132075472e-06,
      "loss": 2.0873,
      "num_input_tokens_seen": 373120,
      "step": 585
    },
    {
      "epoch": 0.34787735849056606,
      "grad_norm": 18.018644332885742,
      "learning_rate": 1.736438679245283e-06,
      "loss": 2.3829,
      "num_input_tokens_seen": 375648,
      "step": 590
    },
    {
      "epoch": 0.35082547169811323,
      "grad_norm": 15.573846817016602,
      "learning_rate": 1.751179245283019e-06,
      "loss": 2.1779,
      "num_input_tokens_seen": 378720,
      "step": 595
    },
    {
      "epoch": 0.35377358490566035,
      "grad_norm": 19.868303298950195,
      "learning_rate": 1.765919811320755e-06,
      "loss": 2.246,
      "num_input_tokens_seen": 381728,
      "step": 600
    },
    {
      "epoch": 0.35672169811320753,
      "grad_norm": 20.972476959228516,
      "learning_rate": 1.7806603773584907e-06,
      "loss": 2.0427,
      "num_input_tokens_seen": 384960,
      "step": 605
    },
    {
      "epoch": 0.3596698113207547,
      "grad_norm": 16.662214279174805,
      "learning_rate": 1.7954009433962267e-06,
      "loss": 2.067,
      "num_input_tokens_seen": 387680,
      "step": 610
    },
    {
      "epoch": 0.3626179245283019,
      "grad_norm": 21.03258514404297,
      "learning_rate": 1.8101415094339625e-06,
      "loss": 1.7835,
      "num_input_tokens_seen": 390528,
      "step": 615
    },
    {
      "epoch": 0.36556603773584906,
      "grad_norm": 22.329349517822266,
      "learning_rate": 1.8248820754716984e-06,
      "loss": 2.1513,
      "num_input_tokens_seen": 393248,
      "step": 620
    },
    {
      "epoch": 0.36851415094339623,
      "grad_norm": 20.41299057006836,
      "learning_rate": 1.839622641509434e-06,
      "loss": 2.1367,
      "num_input_tokens_seen": 396480,
      "step": 625
    },
    {
      "epoch": 0.3714622641509434,
      "grad_norm": 17.944154739379883,
      "learning_rate": 1.8543632075471698e-06,
      "loss": 1.77,
      "num_input_tokens_seen": 400064,
      "step": 630
    },
    {
      "epoch": 0.3744103773584906,
      "grad_norm": 17.88338279724121,
      "learning_rate": 1.8691037735849057e-06,
      "loss": 1.6152,
      "num_input_tokens_seen": 403072,
      "step": 635
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 13.2199125289917,
      "learning_rate": 1.8838443396226417e-06,
      "loss": 1.6665,
      "num_input_tokens_seen": 407136,
      "step": 640
    },
    {
      "epoch": 0.3803066037735849,
      "grad_norm": 29.0507755279541,
      "learning_rate": 1.8985849056603775e-06,
      "loss": 2.1445,
      "num_input_tokens_seen": 410368,
      "step": 645
    },
    {
      "epoch": 0.38325471698113206,
      "grad_norm": 10.977303504943848,
      "learning_rate": 1.9133254716981133e-06,
      "loss": 1.6732,
      "num_input_tokens_seen": 414016,
      "step": 650
    },
    {
      "epoch": 0.38620283018867924,
      "grad_norm": 12.714446067810059,
      "learning_rate": 1.9280660377358494e-06,
      "loss": 1.933,
      "num_input_tokens_seen": 416992,
      "step": 655
    },
    {
      "epoch": 0.3891509433962264,
      "grad_norm": 13.606237411499023,
      "learning_rate": 1.9428066037735852e-06,
      "loss": 1.9287,
      "num_input_tokens_seen": 420640,
      "step": 660
    },
    {
      "epoch": 0.3920990566037736,
      "grad_norm": 13.961576461791992,
      "learning_rate": 1.957547169811321e-06,
      "loss": 1.8563,
      "num_input_tokens_seen": 425280,
      "step": 665
    },
    {
      "epoch": 0.39504716981132076,
      "grad_norm": 12.401579856872559,
      "learning_rate": 1.9722877358490568e-06,
      "loss": 1.8198,
      "num_input_tokens_seen": 429248,
      "step": 670
    },
    {
      "epoch": 0.39799528301886794,
      "grad_norm": 15.778531074523926,
      "learning_rate": 1.9870283018867925e-06,
      "loss": 2.0283,
      "num_input_tokens_seen": 432544,
      "step": 675
    },
    {
      "epoch": 0.4009433962264151,
      "grad_norm": 12.1795072555542,
      "learning_rate": 2.0017688679245283e-06,
      "loss": 1.6347,
      "num_input_tokens_seen": 436064,
      "step": 680
    },
    {
      "epoch": 0.40389150943396224,
      "grad_norm": 25.07501220703125,
      "learning_rate": 2.0165094339622645e-06,
      "loss": 1.4524,
      "num_input_tokens_seen": 439136,
      "step": 685
    },
    {
      "epoch": 0.4068396226415094,
      "grad_norm": 16.530725479125977,
      "learning_rate": 2.0312500000000002e-06,
      "loss": 1.6501,
      "num_input_tokens_seen": 442304,
      "step": 690
    },
    {
      "epoch": 0.4097877358490566,
      "grad_norm": 14.54784107208252,
      "learning_rate": 2.045990566037736e-06,
      "loss": 1.4968,
      "num_input_tokens_seen": 446176,
      "step": 695
    },
    {
      "epoch": 0.41273584905660377,
      "grad_norm": 11.189791679382324,
      "learning_rate": 2.0607311320754718e-06,
      "loss": 1.5344,
      "num_input_tokens_seen": 449376,
      "step": 700
    },
    {
      "epoch": 0.41568396226415094,
      "grad_norm": 17.549291610717773,
      "learning_rate": 2.075471698113208e-06,
      "loss": 1.6429,
      "num_input_tokens_seen": 452992,
      "step": 705
    },
    {
      "epoch": 0.4186320754716981,
      "grad_norm": 18.932756423950195,
      "learning_rate": 2.0902122641509437e-06,
      "loss": 1.61,
      "num_input_tokens_seen": 455552,
      "step": 710
    },
    {
      "epoch": 0.4215801886792453,
      "grad_norm": 17.907318115234375,
      "learning_rate": 2.1049528301886795e-06,
      "loss": 1.5535,
      "num_input_tokens_seen": 458656,
      "step": 715
    },
    {
      "epoch": 0.42452830188679247,
      "grad_norm": 13.344645500183105,
      "learning_rate": 2.1196933962264153e-06,
      "loss": 1.3666,
      "num_input_tokens_seen": 461248,
      "step": 720
    },
    {
      "epoch": 0.42747641509433965,
      "grad_norm": 16.79029083251953,
      "learning_rate": 2.134433962264151e-06,
      "loss": 1.445,
      "num_input_tokens_seen": 464224,
      "step": 725
    },
    {
      "epoch": 0.43042452830188677,
      "grad_norm": 14.347173690795898,
      "learning_rate": 2.149174528301887e-06,
      "loss": 1.3696,
      "num_input_tokens_seen": 466976,
      "step": 730
    },
    {
      "epoch": 0.43337264150943394,
      "grad_norm": 17.674001693725586,
      "learning_rate": 2.163915094339623e-06,
      "loss": 1.3111,
      "num_input_tokens_seen": 469856,
      "step": 735
    },
    {
      "epoch": 0.4363207547169811,
      "grad_norm": 12.729475975036621,
      "learning_rate": 2.1786556603773588e-06,
      "loss": 1.3116,
      "num_input_tokens_seen": 473152,
      "step": 740
    },
    {
      "epoch": 0.4392688679245283,
      "grad_norm": 41.43748092651367,
      "learning_rate": 2.1933962264150945e-06,
      "loss": 1.4243,
      "num_input_tokens_seen": 476448,
      "step": 745
    },
    {
      "epoch": 0.44221698113207547,
      "grad_norm": 11.455280303955078,
      "learning_rate": 2.2081367924528303e-06,
      "loss": 1.2545,
      "num_input_tokens_seen": 479904,
      "step": 750
    },
    {
      "epoch": 0.44516509433962265,
      "grad_norm": 14.249820709228516,
      "learning_rate": 2.2228773584905665e-06,
      "loss": 1.6581,
      "num_input_tokens_seen": 482592,
      "step": 755
    },
    {
      "epoch": 0.4481132075471698,
      "grad_norm": 12.392987251281738,
      "learning_rate": 2.237617924528302e-06,
      "loss": 1.0979,
      "num_input_tokens_seen": 485984,
      "step": 760
    },
    {
      "epoch": 0.451061320754717,
      "grad_norm": 14.623025894165039,
      "learning_rate": 2.252358490566038e-06,
      "loss": 1.0051,
      "num_input_tokens_seen": 488448,
      "step": 765
    },
    {
      "epoch": 0.4540094339622642,
      "grad_norm": 15.857216835021973,
      "learning_rate": 2.267099056603774e-06,
      "loss": 1.2801,
      "num_input_tokens_seen": 491424,
      "step": 770
    },
    {
      "epoch": 0.4569575471698113,
      "grad_norm": 12.599034309387207,
      "learning_rate": 2.2818396226415096e-06,
      "loss": 1.0663,
      "num_input_tokens_seen": 494272,
      "step": 775
    },
    {
      "epoch": 0.45990566037735847,
      "grad_norm": 17.8023624420166,
      "learning_rate": 2.2965801886792453e-06,
      "loss": 1.08,
      "num_input_tokens_seen": 497440,
      "step": 780
    },
    {
      "epoch": 0.46285377358490565,
      "grad_norm": 8.497467994689941,
      "learning_rate": 2.3113207547169815e-06,
      "loss": 0.869,
      "num_input_tokens_seen": 501248,
      "step": 785
    },
    {
      "epoch": 0.4658018867924528,
      "grad_norm": 16.955551147460938,
      "learning_rate": 2.3260613207547173e-06,
      "loss": 1.1645,
      "num_input_tokens_seen": 505024,
      "step": 790
    },
    {
      "epoch": 0.46875,
      "grad_norm": 17.2665958404541,
      "learning_rate": 2.340801886792453e-06,
      "loss": 1.0635,
      "num_input_tokens_seen": 507616,
      "step": 795
    },
    {
      "epoch": 0.4716981132075472,
      "grad_norm": 11.953463554382324,
      "learning_rate": 2.355542452830189e-06,
      "loss": 1.1205,
      "num_input_tokens_seen": 510336,
      "step": 800
    },
    {
      "epoch": 0.47464622641509435,
      "grad_norm": 9.689386367797852,
      "learning_rate": 2.3702830188679246e-06,
      "loss": 0.8438,
      "num_input_tokens_seen": 513120,
      "step": 805
    },
    {
      "epoch": 0.47759433962264153,
      "grad_norm": 9.641800880432129,
      "learning_rate": 2.3850235849056604e-06,
      "loss": 0.7925,
      "num_input_tokens_seen": 516992,
      "step": 810
    },
    {
      "epoch": 0.4805424528301887,
      "grad_norm": 8.856449127197266,
      "learning_rate": 2.3997641509433966e-06,
      "loss": 0.8317,
      "num_input_tokens_seen": 520288,
      "step": 815
    },
    {
      "epoch": 0.4834905660377358,
      "grad_norm": 8.702609062194824,
      "learning_rate": 2.4145047169811323e-06,
      "loss": 0.8217,
      "num_input_tokens_seen": 523360,
      "step": 820
    },
    {
      "epoch": 0.486438679245283,
      "grad_norm": 6.97243070602417,
      "learning_rate": 2.429245283018868e-06,
      "loss": 0.8265,
      "num_input_tokens_seen": 527328,
      "step": 825
    },
    {
      "epoch": 0.4893867924528302,
      "grad_norm": 7.049951076507568,
      "learning_rate": 2.443985849056604e-06,
      "loss": 0.7062,
      "num_input_tokens_seen": 531008,
      "step": 830
    },
    {
      "epoch": 0.49233490566037735,
      "grad_norm": 11.048083305358887,
      "learning_rate": 2.45872641509434e-06,
      "loss": 0.8725,
      "num_input_tokens_seen": 535360,
      "step": 835
    },
    {
      "epoch": 0.49528301886792453,
      "grad_norm": 10.5791015625,
      "learning_rate": 2.473466981132076e-06,
      "loss": 0.7335,
      "num_input_tokens_seen": 539072,
      "step": 840
    },
    {
      "epoch": 0.4982311320754717,
      "grad_norm": 7.280594348907471,
      "learning_rate": 2.4882075471698116e-06,
      "loss": 0.7797,
      "num_input_tokens_seen": 542240,
      "step": 845
    },
    {
      "epoch": 0.5011792452830188,
      "grad_norm": 6.1966471672058105,
      "learning_rate": 2.5029481132075474e-06,
      "loss": 0.8844,
      "num_input_tokens_seen": 545760,
      "step": 850
    },
    {
      "epoch": 0.504127358490566,
      "grad_norm": 7.5716872215271,
      "learning_rate": 2.517688679245283e-06,
      "loss": 0.6857,
      "num_input_tokens_seen": 548768,
      "step": 855
    },
    {
      "epoch": 0.5070754716981132,
      "grad_norm": 7.930427074432373,
      "learning_rate": 2.532429245283019e-06,
      "loss": 0.7115,
      "num_input_tokens_seen": 551584,
      "step": 860
    },
    {
      "epoch": 0.5100235849056604,
      "grad_norm": 9.235867500305176,
      "learning_rate": 2.547169811320755e-06,
      "loss": 0.8221,
      "num_input_tokens_seen": 555488,
      "step": 865
    },
    {
      "epoch": 0.5129716981132075,
      "grad_norm": 7.014967918395996,
      "learning_rate": 2.561910377358491e-06,
      "loss": 0.7976,
      "num_input_tokens_seen": 557952,
      "step": 870
    },
    {
      "epoch": 0.5159198113207547,
      "grad_norm": 8.66892147064209,
      "learning_rate": 2.5766509433962266e-06,
      "loss": 0.8701,
      "num_input_tokens_seen": 561248,
      "step": 875
    },
    {
      "epoch": 0.5188679245283019,
      "grad_norm": 9.413871765136719,
      "learning_rate": 2.5913915094339624e-06,
      "loss": 1.2207,
      "num_input_tokens_seen": 563392,
      "step": 880
    },
    {
      "epoch": 0.5218160377358491,
      "grad_norm": 8.08803653717041,
      "learning_rate": 2.6061320754716986e-06,
      "loss": 0.659,
      "num_input_tokens_seen": 567264,
      "step": 885
    },
    {
      "epoch": 0.5247641509433962,
      "grad_norm": 5.434665203094482,
      "learning_rate": 2.6208726415094343e-06,
      "loss": 0.6798,
      "num_input_tokens_seen": 569856,
      "step": 890
    },
    {
      "epoch": 0.5277122641509434,
      "grad_norm": 5.352683067321777,
      "learning_rate": 2.63561320754717e-06,
      "loss": 0.7935,
      "num_input_tokens_seen": 573856,
      "step": 895
    },
    {
      "epoch": 0.5306603773584906,
      "grad_norm": 5.7827959060668945,
      "learning_rate": 2.650353773584906e-06,
      "loss": 0.7004,
      "num_input_tokens_seen": 577760,
      "step": 900
    },
    {
      "epoch": 0.5336084905660378,
      "grad_norm": 10.55976676940918,
      "learning_rate": 2.665094339622642e-06,
      "loss": 0.934,
      "num_input_tokens_seen": 581216,
      "step": 905
    },
    {
      "epoch": 0.5365566037735849,
      "grad_norm": 5.149581432342529,
      "learning_rate": 2.679834905660378e-06,
      "loss": 0.6503,
      "num_input_tokens_seen": 584096,
      "step": 910
    },
    {
      "epoch": 0.5395047169811321,
      "grad_norm": 4.295581340789795,
      "learning_rate": 2.694575471698113e-06,
      "loss": 0.5426,
      "num_input_tokens_seen": 587392,
      "step": 915
    },
    {
      "epoch": 0.5424528301886793,
      "grad_norm": 6.303553581237793,
      "learning_rate": 2.709316037735849e-06,
      "loss": 0.926,
      "num_input_tokens_seen": 590208,
      "step": 920
    },
    {
      "epoch": 0.5454009433962265,
      "grad_norm": 9.305906295776367,
      "learning_rate": 2.724056603773585e-06,
      "loss": 0.6543,
      "num_input_tokens_seen": 592896,
      "step": 925
    },
    {
      "epoch": 0.5483490566037735,
      "grad_norm": 7.1254191398620605,
      "learning_rate": 2.738797169811321e-06,
      "loss": 0.9589,
      "num_input_tokens_seen": 595712,
      "step": 930
    },
    {
      "epoch": 0.5512971698113207,
      "grad_norm": 4.709741115570068,
      "learning_rate": 2.7535377358490567e-06,
      "loss": 0.6873,
      "num_input_tokens_seen": 599872,
      "step": 935
    },
    {
      "epoch": 0.5542452830188679,
      "grad_norm": 6.462931156158447,
      "learning_rate": 2.7682783018867925e-06,
      "loss": 0.6607,
      "num_input_tokens_seen": 602464,
      "step": 940
    },
    {
      "epoch": 0.5571933962264151,
      "grad_norm": 9.48462200164795,
      "learning_rate": 2.7830188679245286e-06,
      "loss": 0.6769,
      "num_input_tokens_seen": 604672,
      "step": 945
    },
    {
      "epoch": 0.5601415094339622,
      "grad_norm": 5.455216884613037,
      "learning_rate": 2.7977594339622644e-06,
      "loss": 0.6459,
      "num_input_tokens_seen": 608864,
      "step": 950
    },
    {
      "epoch": 0.5630896226415094,
      "grad_norm": 8.060654640197754,
      "learning_rate": 2.8125e-06,
      "loss": 0.7928,
      "num_input_tokens_seen": 611872,
      "step": 955
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 5.41253662109375,
      "learning_rate": 2.827240566037736e-06,
      "loss": 0.7992,
      "num_input_tokens_seen": 615776,
      "step": 960
    },
    {
      "epoch": 0.5689858490566038,
      "grad_norm": 16.779678344726562,
      "learning_rate": 2.841981132075472e-06,
      "loss": 0.5546,
      "num_input_tokens_seen": 618784,
      "step": 965
    },
    {
      "epoch": 0.5719339622641509,
      "grad_norm": 3.1124043464660645,
      "learning_rate": 2.856721698113208e-06,
      "loss": 0.5638,
      "num_input_tokens_seen": 622624,
      "step": 970
    },
    {
      "epoch": 0.5748820754716981,
      "grad_norm": 5.431405544281006,
      "learning_rate": 2.8714622641509437e-06,
      "loss": 0.5669,
      "num_input_tokens_seen": 625920,
      "step": 975
    },
    {
      "epoch": 0.5778301886792453,
      "grad_norm": 7.875941753387451,
      "learning_rate": 2.8862028301886794e-06,
      "loss": 0.8048,
      "num_input_tokens_seen": 630240,
      "step": 980
    },
    {
      "epoch": 0.5807783018867925,
      "grad_norm": 5.937239646911621,
      "learning_rate": 2.9009433962264156e-06,
      "loss": 0.6055,
      "num_input_tokens_seen": 633312,
      "step": 985
    },
    {
      "epoch": 0.5837264150943396,
      "grad_norm": 4.949699878692627,
      "learning_rate": 2.9156839622641514e-06,
      "loss": 0.8674,
      "num_input_tokens_seen": 636960,
      "step": 990
    },
    {
      "epoch": 0.5866745283018868,
      "grad_norm": 8.501602172851562,
      "learning_rate": 2.930424528301887e-06,
      "loss": 0.7378,
      "num_input_tokens_seen": 639840,
      "step": 995
    },
    {
      "epoch": 0.589622641509434,
      "grad_norm": 4.403671741485596,
      "learning_rate": 2.9451650943396225e-06,
      "loss": 0.6401,
      "num_input_tokens_seen": 642496,
      "step": 1000
    },
    {
      "epoch": 0.5925707547169812,
      "grad_norm": 10.234060287475586,
      "learning_rate": 2.9599056603773587e-06,
      "loss": 0.5653,
      "num_input_tokens_seen": 644992,
      "step": 1005
    },
    {
      "epoch": 0.5955188679245284,
      "grad_norm": 6.24685001373291,
      "learning_rate": 2.9746462264150945e-06,
      "loss": 0.6589,
      "num_input_tokens_seen": 647808,
      "step": 1010
    },
    {
      "epoch": 0.5984669811320755,
      "grad_norm": 10.167947769165039,
      "learning_rate": 2.9893867924528302e-06,
      "loss": 0.6326,
      "num_input_tokens_seen": 651008,
      "step": 1015
    },
    {
      "epoch": 0.6014150943396226,
      "grad_norm": 7.105341911315918,
      "learning_rate": 3.004127358490566e-06,
      "loss": 0.7067,
      "num_input_tokens_seen": 653952,
      "step": 1020
    },
    {
      "epoch": 0.6043632075471698,
      "grad_norm": 2.9866106510162354,
      "learning_rate": 3.018867924528302e-06,
      "loss": 0.5974,
      "num_input_tokens_seen": 656960,
      "step": 1025
    },
    {
      "epoch": 0.6073113207547169,
      "grad_norm": 3.5141289234161377,
      "learning_rate": 3.033608490566038e-06,
      "loss": 0.7784,
      "num_input_tokens_seen": 660256,
      "step": 1030
    },
    {
      "epoch": 0.6102594339622641,
      "grad_norm": 9.157804489135742,
      "learning_rate": 3.0483490566037737e-06,
      "loss": 0.6353,
      "num_input_tokens_seen": 663712,
      "step": 1035
    },
    {
      "epoch": 0.6132075471698113,
      "grad_norm": 3.199897527694702,
      "learning_rate": 3.0630896226415095e-06,
      "loss": 0.6312,
      "num_input_tokens_seen": 666976,
      "step": 1040
    },
    {
      "epoch": 0.6161556603773585,
      "grad_norm": 4.349030017852783,
      "learning_rate": 3.0778301886792457e-06,
      "loss": 0.6387,
      "num_input_tokens_seen": 669856,
      "step": 1045
    },
    {
      "epoch": 0.6191037735849056,
      "grad_norm": 4.497307300567627,
      "learning_rate": 3.0925707547169815e-06,
      "loss": 0.9133,
      "num_input_tokens_seen": 675936,
      "step": 1050
    },
    {
      "epoch": 0.6220518867924528,
      "grad_norm": 6.537468910217285,
      "learning_rate": 3.1073113207547172e-06,
      "loss": 0.6145,
      "num_input_tokens_seen": 678560,
      "step": 1055
    },
    {
      "epoch": 0.625,
      "grad_norm": 5.812407970428467,
      "learning_rate": 3.122051886792453e-06,
      "loss": 0.6483,
      "num_input_tokens_seen": 681632,
      "step": 1060
    },
    {
      "epoch": 0.6279481132075472,
      "grad_norm": 6.83676290512085,
      "learning_rate": 3.136792452830189e-06,
      "loss": 0.5978,
      "num_input_tokens_seen": 685568,
      "step": 1065
    },
    {
      "epoch": 0.6308962264150944,
      "grad_norm": 7.661698341369629,
      "learning_rate": 3.151533018867925e-06,
      "loss": 0.822,
      "num_input_tokens_seen": 688800,
      "step": 1070
    },
    {
      "epoch": 0.6338443396226415,
      "grad_norm": 3.8842620849609375,
      "learning_rate": 3.1662735849056607e-06,
      "loss": 0.5648,
      "num_input_tokens_seen": 691200,
      "step": 1075
    },
    {
      "epoch": 0.6367924528301887,
      "grad_norm": 5.587656497955322,
      "learning_rate": 3.181014150943397e-06,
      "loss": 0.6118,
      "num_input_tokens_seen": 694656,
      "step": 1080
    },
    {
      "epoch": 0.6397405660377359,
      "grad_norm": 5.511290550231934,
      "learning_rate": 3.1957547169811327e-06,
      "loss": 0.5276,
      "num_input_tokens_seen": 697792,
      "step": 1085
    },
    {
      "epoch": 0.6426886792452831,
      "grad_norm": 3.294952869415283,
      "learning_rate": 3.210495283018868e-06,
      "loss": 0.7128,
      "num_input_tokens_seen": 701056,
      "step": 1090
    },
    {
      "epoch": 0.6456367924528302,
      "grad_norm": 4.073520660400391,
      "learning_rate": 3.225235849056604e-06,
      "loss": 0.5852,
      "num_input_tokens_seen": 704032,
      "step": 1095
    },
    {
      "epoch": 0.6485849056603774,
      "grad_norm": 10.344172477722168,
      "learning_rate": 3.2399764150943396e-06,
      "loss": 0.676,
      "num_input_tokens_seen": 706688,
      "step": 1100
    },
    {
      "epoch": 0.6515330188679245,
      "grad_norm": 3.4547371864318848,
      "learning_rate": 3.2547169811320758e-06,
      "loss": 0.5548,
      "num_input_tokens_seen": 710080,
      "step": 1105
    },
    {
      "epoch": 0.6544811320754716,
      "grad_norm": 6.965508460998535,
      "learning_rate": 3.2694575471698115e-06,
      "loss": 0.742,
      "num_input_tokens_seen": 713184,
      "step": 1110
    },
    {
      "epoch": 0.6574292452830188,
      "grad_norm": 4.547361373901367,
      "learning_rate": 3.2841981132075473e-06,
      "loss": 0.5102,
      "num_input_tokens_seen": 716064,
      "step": 1115
    },
    {
      "epoch": 0.660377358490566,
      "grad_norm": 4.807929039001465,
      "learning_rate": 3.298938679245283e-06,
      "loss": 0.7134,
      "num_input_tokens_seen": 718976,
      "step": 1120
    },
    {
      "epoch": 0.6633254716981132,
      "grad_norm": 4.483486175537109,
      "learning_rate": 3.3136792452830192e-06,
      "loss": 0.6749,
      "num_input_tokens_seen": 722080,
      "step": 1125
    },
    {
      "epoch": 0.6662735849056604,
      "grad_norm": 3.9090898036956787,
      "learning_rate": 3.328419811320755e-06,
      "loss": 0.6619,
      "num_input_tokens_seen": 724928,
      "step": 1130
    },
    {
      "epoch": 0.6692216981132075,
      "grad_norm": 9.003461837768555,
      "learning_rate": 3.3431603773584908e-06,
      "loss": 0.6966,
      "num_input_tokens_seen": 727904,
      "step": 1135
    },
    {
      "epoch": 0.6721698113207547,
      "grad_norm": 5.236022472381592,
      "learning_rate": 3.3579009433962266e-06,
      "loss": 0.6879,
      "num_input_tokens_seen": 731584,
      "step": 1140
    },
    {
      "epoch": 0.6751179245283019,
      "grad_norm": 1.9928230047225952,
      "learning_rate": 3.3726415094339627e-06,
      "loss": 0.693,
      "num_input_tokens_seen": 735392,
      "step": 1145
    },
    {
      "epoch": 0.6780660377358491,
      "grad_norm": 11.479316711425781,
      "learning_rate": 3.3873820754716985e-06,
      "loss": 0.6215,
      "num_input_tokens_seen": 738528,
      "step": 1150
    },
    {
      "epoch": 0.6810141509433962,
      "grad_norm": 3.963622570037842,
      "learning_rate": 3.4021226415094343e-06,
      "loss": 0.7539,
      "num_input_tokens_seen": 741920,
      "step": 1155
    },
    {
      "epoch": 0.6839622641509434,
      "grad_norm": 7.670055389404297,
      "learning_rate": 3.4168632075471705e-06,
      "loss": 0.554,
      "num_input_tokens_seen": 744160,
      "step": 1160
    },
    {
      "epoch": 0.6869103773584906,
      "grad_norm": 3.6673624515533447,
      "learning_rate": 3.4316037735849062e-06,
      "loss": 0.4836,
      "num_input_tokens_seen": 748928,
      "step": 1165
    },
    {
      "epoch": 0.6898584905660378,
      "grad_norm": 5.805579662322998,
      "learning_rate": 3.446344339622642e-06,
      "loss": 0.7141,
      "num_input_tokens_seen": 752544,
      "step": 1170
    },
    {
      "epoch": 0.6928066037735849,
      "grad_norm": 3.821148633956909,
      "learning_rate": 3.4610849056603778e-06,
      "loss": 0.6592,
      "num_input_tokens_seen": 756256,
      "step": 1175
    },
    {
      "epoch": 0.6957547169811321,
      "grad_norm": 2.6932156085968018,
      "learning_rate": 3.475825471698113e-06,
      "loss": 0.6145,
      "num_input_tokens_seen": 763008,
      "step": 1180
    },
    {
      "epoch": 0.6987028301886793,
      "grad_norm": 3.2346081733703613,
      "learning_rate": 3.4905660377358493e-06,
      "loss": 0.5423,
      "num_input_tokens_seen": 765632,
      "step": 1185
    },
    {
      "epoch": 0.7016509433962265,
      "grad_norm": 4.860165596008301,
      "learning_rate": 3.505306603773585e-06,
      "loss": 0.7523,
      "num_input_tokens_seen": 770272,
      "step": 1190
    },
    {
      "epoch": 0.7045990566037735,
      "grad_norm": 3.772733211517334,
      "learning_rate": 3.520047169811321e-06,
      "loss": 0.5646,
      "num_input_tokens_seen": 774592,
      "step": 1195
    },
    {
      "epoch": 0.7075471698113207,
      "grad_norm": 8.273346900939941,
      "learning_rate": 3.5347877358490566e-06,
      "loss": 0.5814,
      "num_input_tokens_seen": 778688,
      "step": 1200
    },
    {
      "epoch": 0.7104952830188679,
      "grad_norm": 2.619436264038086,
      "learning_rate": 3.549528301886793e-06,
      "loss": 0.4878,
      "num_input_tokens_seen": 781888,
      "step": 1205
    },
    {
      "epoch": 0.7134433962264151,
      "grad_norm": 4.2157135009765625,
      "learning_rate": 3.5642688679245286e-06,
      "loss": 0.5565,
      "num_input_tokens_seen": 785376,
      "step": 1210
    },
    {
      "epoch": 0.7163915094339622,
      "grad_norm": 11.360536575317383,
      "learning_rate": 3.5790094339622643e-06,
      "loss": 0.7106,
      "num_input_tokens_seen": 788096,
      "step": 1215
    },
    {
      "epoch": 0.7193396226415094,
      "grad_norm": 11.200027465820312,
      "learning_rate": 3.59375e-06,
      "loss": 0.6209,
      "num_input_tokens_seen": 791712,
      "step": 1220
    },
    {
      "epoch": 0.7222877358490566,
      "grad_norm": 3.0180084705352783,
      "learning_rate": 3.6084905660377363e-06,
      "loss": 0.639,
      "num_input_tokens_seen": 794848,
      "step": 1225
    },
    {
      "epoch": 0.7252358490566038,
      "grad_norm": 2.8974008560180664,
      "learning_rate": 3.623231132075472e-06,
      "loss": 0.6861,
      "num_input_tokens_seen": 797984,
      "step": 1230
    },
    {
      "epoch": 0.7281839622641509,
      "grad_norm": 1.9853707551956177,
      "learning_rate": 3.637971698113208e-06,
      "loss": 0.7276,
      "num_input_tokens_seen": 801632,
      "step": 1235
    },
    {
      "epoch": 0.7311320754716981,
      "grad_norm": 3.9424211978912354,
      "learning_rate": 3.652712264150944e-06,
      "loss": 0.5143,
      "num_input_tokens_seen": 804864,
      "step": 1240
    },
    {
      "epoch": 0.7340801886792453,
      "grad_norm": 3.006974458694458,
      "learning_rate": 3.66745283018868e-06,
      "loss": 0.4857,
      "num_input_tokens_seen": 807712,
      "step": 1245
    },
    {
      "epoch": 0.7370283018867925,
      "grad_norm": 3.4118599891662598,
      "learning_rate": 3.6821933962264156e-06,
      "loss": 0.4577,
      "num_input_tokens_seen": 810752,
      "step": 1250
    },
    {
      "epoch": 0.7399764150943396,
      "grad_norm": 3.1147165298461914,
      "learning_rate": 3.6969339622641513e-06,
      "loss": 0.5785,
      "num_input_tokens_seen": 814528,
      "step": 1255
    },
    {
      "epoch": 0.7429245283018868,
      "grad_norm": 7.25778341293335,
      "learning_rate": 3.7116745283018875e-06,
      "loss": 0.4593,
      "num_input_tokens_seen": 817152,
      "step": 1260
    },
    {
      "epoch": 0.745872641509434,
      "grad_norm": 5.163593292236328,
      "learning_rate": 3.726415094339623e-06,
      "loss": 0.7069,
      "num_input_tokens_seen": 820960,
      "step": 1265
    },
    {
      "epoch": 0.7488207547169812,
      "grad_norm": 8.123295783996582,
      "learning_rate": 3.7411556603773586e-06,
      "loss": 0.5835,
      "num_input_tokens_seen": 824000,
      "step": 1270
    },
    {
      "epoch": 0.7517688679245284,
      "grad_norm": 8.283679962158203,
      "learning_rate": 3.7558962264150944e-06,
      "loss": 0.4915,
      "num_input_tokens_seen": 826336,
      "step": 1275
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 6.3880839347839355,
      "learning_rate": 3.77063679245283e-06,
      "loss": 0.7592,
      "num_input_tokens_seen": 829440,
      "step": 1280
    },
    {
      "epoch": 0.7576650943396226,
      "grad_norm": 4.055119037628174,
      "learning_rate": 3.7853773584905664e-06,
      "loss": 0.4897,
      "num_input_tokens_seen": 833664,
      "step": 1285
    },
    {
      "epoch": 0.7606132075471698,
      "grad_norm": 6.377405166625977,
      "learning_rate": 3.800117924528302e-06,
      "loss": 0.5461,
      "num_input_tokens_seen": 836224,
      "step": 1290
    },
    {
      "epoch": 0.7635613207547169,
      "grad_norm": 5.201138496398926,
      "learning_rate": 3.814858490566038e-06,
      "loss": 0.5122,
      "num_input_tokens_seen": 840032,
      "step": 1295
    },
    {
      "epoch": 0.7665094339622641,
      "grad_norm": 9.94583797454834,
      "learning_rate": 3.829599056603774e-06,
      "loss": 0.7124,
      "num_input_tokens_seen": 842752,
      "step": 1300
    },
    {
      "epoch": 0.7694575471698113,
      "grad_norm": 4.232667922973633,
      "learning_rate": 3.8443396226415094e-06,
      "loss": 0.6074,
      "num_input_tokens_seen": 846144,
      "step": 1305
    },
    {
      "epoch": 0.7724056603773585,
      "grad_norm": 3.183858871459961,
      "learning_rate": 3.859080188679246e-06,
      "loss": 0.4356,
      "num_input_tokens_seen": 849504,
      "step": 1310
    },
    {
      "epoch": 0.7753537735849056,
      "grad_norm": 4.755903244018555,
      "learning_rate": 3.873820754716982e-06,
      "loss": 0.7228,
      "num_input_tokens_seen": 852736,
      "step": 1315
    },
    {
      "epoch": 0.7783018867924528,
      "grad_norm": 6.60689115524292,
      "learning_rate": 3.888561320754717e-06,
      "loss": 0.7182,
      "num_input_tokens_seen": 855360,
      "step": 1320
    },
    {
      "epoch": 0.78125,
      "grad_norm": 13.475582122802734,
      "learning_rate": 3.903301886792453e-06,
      "loss": 0.6794,
      "num_input_tokens_seen": 858080,
      "step": 1325
    },
    {
      "epoch": 0.7841981132075472,
      "grad_norm": 3.7765071392059326,
      "learning_rate": 3.9180424528301895e-06,
      "loss": 0.6622,
      "num_input_tokens_seen": 861088,
      "step": 1330
    },
    {
      "epoch": 0.7871462264150944,
      "grad_norm": 6.179556846618652,
      "learning_rate": 3.932783018867925e-06,
      "loss": 0.5885,
      "num_input_tokens_seen": 864352,
      "step": 1335
    },
    {
      "epoch": 0.7900943396226415,
      "grad_norm": 5.685624599456787,
      "learning_rate": 3.947523584905661e-06,
      "loss": 0.4256,
      "num_input_tokens_seen": 867360,
      "step": 1340
    },
    {
      "epoch": 0.7930424528301887,
      "grad_norm": 5.997613430023193,
      "learning_rate": 3.962264150943396e-06,
      "loss": 0.4454,
      "num_input_tokens_seen": 870496,
      "step": 1345
    },
    {
      "epoch": 0.7959905660377359,
      "grad_norm": 3.444251775741577,
      "learning_rate": 3.977004716981133e-06,
      "loss": 0.4796,
      "num_input_tokens_seen": 874112,
      "step": 1350
    },
    {
      "epoch": 0.7989386792452831,
      "grad_norm": 3.882439136505127,
      "learning_rate": 3.991745283018868e-06,
      "loss": 0.7278,
      "num_input_tokens_seen": 877824,
      "step": 1355
    },
    {
      "epoch": 0.8018867924528302,
      "grad_norm": 5.999035358428955,
      "learning_rate": 4.006485849056604e-06,
      "loss": 0.5762,
      "num_input_tokens_seen": 880928,
      "step": 1360
    },
    {
      "epoch": 0.8048349056603774,
      "grad_norm": 3.7236106395721436,
      "learning_rate": 4.0212264150943395e-06,
      "loss": 0.617,
      "num_input_tokens_seen": 884512,
      "step": 1365
    },
    {
      "epoch": 0.8077830188679245,
      "grad_norm": 4.463553428649902,
      "learning_rate": 4.035966981132076e-06,
      "loss": 0.6512,
      "num_input_tokens_seen": 887712,
      "step": 1370
    },
    {
      "epoch": 0.8107311320754716,
      "grad_norm": 3.5415256023406982,
      "learning_rate": 4.050707547169812e-06,
      "loss": 0.6061,
      "num_input_tokens_seen": 890816,
      "step": 1375
    },
    {
      "epoch": 0.8136792452830188,
      "grad_norm": 4.134223461151123,
      "learning_rate": 4.065448113207547e-06,
      "loss": 0.6259,
      "num_input_tokens_seen": 894112,
      "step": 1380
    },
    {
      "epoch": 0.816627358490566,
      "grad_norm": 3.6033999919891357,
      "learning_rate": 4.080188679245283e-06,
      "loss": 0.4749,
      "num_input_tokens_seen": 896800,
      "step": 1385
    },
    {
      "epoch": 0.8195754716981132,
      "grad_norm": 3.1838014125823975,
      "learning_rate": 4.094929245283019e-06,
      "loss": 0.599,
      "num_input_tokens_seen": 900032,
      "step": 1390
    },
    {
      "epoch": 0.8225235849056604,
      "grad_norm": 5.712179660797119,
      "learning_rate": 4.109669811320755e-06,
      "loss": 0.513,
      "num_input_tokens_seen": 902688,
      "step": 1395
    },
    {
      "epoch": 0.8254716981132075,
      "grad_norm": 5.961308002471924,
      "learning_rate": 4.124410377358491e-06,
      "loss": 0.6217,
      "num_input_tokens_seen": 906240,
      "step": 1400
    },
    {
      "epoch": 0.8284198113207547,
      "grad_norm": 7.362982749938965,
      "learning_rate": 4.1391509433962265e-06,
      "loss": 0.631,
      "num_input_tokens_seen": 912256,
      "step": 1405
    },
    {
      "epoch": 0.8313679245283019,
      "grad_norm": 4.178365707397461,
      "learning_rate": 4.153891509433963e-06,
      "loss": 0.5917,
      "num_input_tokens_seen": 916896,
      "step": 1410
    },
    {
      "epoch": 0.8343160377358491,
      "grad_norm": 6.264890670776367,
      "learning_rate": 4.168632075471699e-06,
      "loss": 0.5361,
      "num_input_tokens_seen": 920000,
      "step": 1415
    },
    {
      "epoch": 0.8372641509433962,
      "grad_norm": 3.1701138019561768,
      "learning_rate": 4.183372641509434e-06,
      "loss": 0.4451,
      "num_input_tokens_seen": 922464,
      "step": 1420
    },
    {
      "epoch": 0.8402122641509434,
      "grad_norm": 7.009893894195557,
      "learning_rate": 4.19811320754717e-06,
      "loss": 0.4705,
      "num_input_tokens_seen": 925120,
      "step": 1425
    },
    {
      "epoch": 0.8431603773584906,
      "grad_norm": 11.524538040161133,
      "learning_rate": 4.212853773584907e-06,
      "loss": 0.5439,
      "num_input_tokens_seen": 927520,
      "step": 1430
    },
    {
      "epoch": 0.8461084905660378,
      "grad_norm": 5.140844821929932,
      "learning_rate": 4.227594339622642e-06,
      "loss": 0.5198,
      "num_input_tokens_seen": 932352,
      "step": 1435
    },
    {
      "epoch": 0.8490566037735849,
      "grad_norm": 1.8821611404418945,
      "learning_rate": 4.242334905660378e-06,
      "loss": 0.458,
      "num_input_tokens_seen": 935456,
      "step": 1440
    },
    {
      "epoch": 0.8520047169811321,
      "grad_norm": 2.9974260330200195,
      "learning_rate": 4.2570754716981135e-06,
      "loss": 0.6592,
      "num_input_tokens_seen": 938464,
      "step": 1445
    },
    {
      "epoch": 0.8549528301886793,
      "grad_norm": 2.694298267364502,
      "learning_rate": 4.271816037735849e-06,
      "loss": 0.5344,
      "num_input_tokens_seen": 941952,
      "step": 1450
    },
    {
      "epoch": 0.8579009433962265,
      "grad_norm": 2.9159343242645264,
      "learning_rate": 4.286556603773585e-06,
      "loss": 0.384,
      "num_input_tokens_seen": 945024,
      "step": 1455
    },
    {
      "epoch": 0.8608490566037735,
      "grad_norm": 7.395695209503174,
      "learning_rate": 4.301297169811321e-06,
      "loss": 0.6832,
      "num_input_tokens_seen": 948512,
      "step": 1460
    },
    {
      "epoch": 0.8637971698113207,
      "grad_norm": 2.5327279567718506,
      "learning_rate": 4.3160377358490565e-06,
      "loss": 0.4049,
      "num_input_tokens_seen": 952448,
      "step": 1465
    },
    {
      "epoch": 0.8667452830188679,
      "grad_norm": 2.4279379844665527,
      "learning_rate": 4.330778301886793e-06,
      "loss": 0.5575,
      "num_input_tokens_seen": 956256,
      "step": 1470
    },
    {
      "epoch": 0.8696933962264151,
      "grad_norm": 2.793804168701172,
      "learning_rate": 4.345518867924529e-06,
      "loss": 0.6182,
      "num_input_tokens_seen": 959264,
      "step": 1475
    },
    {
      "epoch": 0.8726415094339622,
      "grad_norm": 2.7391209602355957,
      "learning_rate": 4.360259433962264e-06,
      "loss": 0.5621,
      "num_input_tokens_seen": 962112,
      "step": 1480
    },
    {
      "epoch": 0.8755896226415094,
      "grad_norm": 3.7803988456726074,
      "learning_rate": 4.3750000000000005e-06,
      "loss": 0.8105,
      "num_input_tokens_seen": 965056,
      "step": 1485
    },
    {
      "epoch": 0.8785377358490566,
      "grad_norm": 4.963263034820557,
      "learning_rate": 4.389740566037737e-06,
      "loss": 0.4849,
      "num_input_tokens_seen": 967840,
      "step": 1490
    },
    {
      "epoch": 0.8814858490566038,
      "grad_norm": 6.25520658493042,
      "learning_rate": 4.404481132075472e-06,
      "loss": 0.5174,
      "num_input_tokens_seen": 972192,
      "step": 1495
    },
    {
      "epoch": 0.8844339622641509,
      "grad_norm": 4.5875678062438965,
      "learning_rate": 4.419221698113208e-06,
      "loss": 0.6129,
      "num_input_tokens_seen": 974496,
      "step": 1500
    },
    {
      "epoch": 0.8873820754716981,
      "grad_norm": 2.510812759399414,
      "learning_rate": 4.4339622641509435e-06,
      "loss": 0.5123,
      "num_input_tokens_seen": 978272,
      "step": 1505
    },
    {
      "epoch": 0.8903301886792453,
      "grad_norm": 5.295869827270508,
      "learning_rate": 4.44870283018868e-06,
      "loss": 0.5929,
      "num_input_tokens_seen": 982112,
      "step": 1510
    },
    {
      "epoch": 0.8932783018867925,
      "grad_norm": 3.9332737922668457,
      "learning_rate": 4.463443396226416e-06,
      "loss": 0.5728,
      "num_input_tokens_seen": 984480,
      "step": 1515
    },
    {
      "epoch": 0.8962264150943396,
      "grad_norm": 8.096114158630371,
      "learning_rate": 4.478183962264151e-06,
      "loss": 0.5342,
      "num_input_tokens_seen": 987680,
      "step": 1520
    },
    {
      "epoch": 0.8991745283018868,
      "grad_norm": 5.009399890899658,
      "learning_rate": 4.4929245283018875e-06,
      "loss": 0.7493,
      "num_input_tokens_seen": 991488,
      "step": 1525
    },
    {
      "epoch": 0.902122641509434,
      "grad_norm": 2.821342706680298,
      "learning_rate": 4.507665094339623e-06,
      "loss": 0.5322,
      "num_input_tokens_seen": 994208,
      "step": 1530
    },
    {
      "epoch": 0.9050707547169812,
      "grad_norm": 2.15200138092041,
      "learning_rate": 4.522405660377359e-06,
      "loss": 0.6577,
      "num_input_tokens_seen": 998624,
      "step": 1535
    },
    {
      "epoch": 0.9080188679245284,
      "grad_norm": 3.3473775386810303,
      "learning_rate": 4.537146226415094e-06,
      "loss": 0.502,
      "num_input_tokens_seen": 1001664,
      "step": 1540
    },
    {
      "epoch": 0.9109669811320755,
      "grad_norm": 3.7360363006591797,
      "learning_rate": 4.5518867924528305e-06,
      "loss": 0.443,
      "num_input_tokens_seen": 1004512,
      "step": 1545
    },
    {
      "epoch": 0.9139150943396226,
      "grad_norm": 4.945830821990967,
      "learning_rate": 4.566627358490566e-06,
      "loss": 0.6752,
      "num_input_tokens_seen": 1007584,
      "step": 1550
    },
    {
      "epoch": 0.9168632075471698,
      "grad_norm": 4.642165660858154,
      "learning_rate": 4.581367924528302e-06,
      "loss": 0.5425,
      "num_input_tokens_seen": 1011648,
      "step": 1555
    },
    {
      "epoch": 0.9198113207547169,
      "grad_norm": 8.137739181518555,
      "learning_rate": 4.596108490566038e-06,
      "loss": 0.5954,
      "num_input_tokens_seen": 1014528,
      "step": 1560
    },
    {
      "epoch": 0.9227594339622641,
      "grad_norm": 2.7659952640533447,
      "learning_rate": 4.610849056603774e-06,
      "loss": 0.4147,
      "num_input_tokens_seen": 1016992,
      "step": 1565
    },
    {
      "epoch": 0.9257075471698113,
      "grad_norm": 1.9279541969299316,
      "learning_rate": 4.62558962264151e-06,
      "loss": 0.6068,
      "num_input_tokens_seen": 1020128,
      "step": 1570
    },
    {
      "epoch": 0.9286556603773585,
      "grad_norm": 3.245415687561035,
      "learning_rate": 4.640330188679246e-06,
      "loss": 0.4828,
      "num_input_tokens_seen": 1024160,
      "step": 1575
    },
    {
      "epoch": 0.9316037735849056,
      "grad_norm": 4.243259429931641,
      "learning_rate": 4.655070754716981e-06,
      "loss": 0.6708,
      "num_input_tokens_seen": 1027392,
      "step": 1580
    },
    {
      "epoch": 0.9345518867924528,
      "grad_norm": 5.531893253326416,
      "learning_rate": 4.6698113207547175e-06,
      "loss": 0.5619,
      "num_input_tokens_seen": 1030400,
      "step": 1585
    },
    {
      "epoch": 0.9375,
      "grad_norm": 3.5415561199188232,
      "learning_rate": 4.684551886792454e-06,
      "loss": 0.5815,
      "num_input_tokens_seen": 1033184,
      "step": 1590
    },
    {
      "epoch": 0.9404481132075472,
      "grad_norm": 2.082848310470581,
      "learning_rate": 4.699292452830189e-06,
      "loss": 0.5376,
      "num_input_tokens_seen": 1036960,
      "step": 1595
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 3.1812143325805664,
      "learning_rate": 4.714033018867925e-06,
      "loss": 0.7485,
      "num_input_tokens_seen": 1040384,
      "step": 1600
    },
    {
      "epoch": 0.9463443396226415,
      "grad_norm": 4.451993465423584,
      "learning_rate": 4.728773584905661e-06,
      "loss": 0.8619,
      "num_input_tokens_seen": 1043488,
      "step": 1605
    },
    {
      "epoch": 0.9492924528301887,
      "grad_norm": 6.560683727264404,
      "learning_rate": 4.743514150943397e-06,
      "loss": 0.5719,
      "num_input_tokens_seen": 1046720,
      "step": 1610
    },
    {
      "epoch": 0.9522405660377359,
      "grad_norm": 14.34493350982666,
      "learning_rate": 4.758254716981133e-06,
      "loss": 0.7161,
      "num_input_tokens_seen": 1049600,
      "step": 1615
    },
    {
      "epoch": 0.9551886792452831,
      "grad_norm": 3.3773386478424072,
      "learning_rate": 4.772995283018868e-06,
      "loss": 0.5739,
      "num_input_tokens_seen": 1052864,
      "step": 1620
    },
    {
      "epoch": 0.9581367924528302,
      "grad_norm": 6.103526592254639,
      "learning_rate": 4.787735849056604e-06,
      "loss": 0.5377,
      "num_input_tokens_seen": 1056384,
      "step": 1625
    },
    {
      "epoch": 0.9610849056603774,
      "grad_norm": 5.53052282333374,
      "learning_rate": 4.80247641509434e-06,
      "loss": 0.6151,
      "num_input_tokens_seen": 1059456,
      "step": 1630
    },
    {
      "epoch": 0.9640330188679245,
      "grad_norm": 19.24129867553711,
      "learning_rate": 4.817216981132076e-06,
      "loss": 1.1561,
      "num_input_tokens_seen": 1062048,
      "step": 1635
    },
    {
      "epoch": 0.9669811320754716,
      "grad_norm": 2.529651403427124,
      "learning_rate": 4.831957547169811e-06,
      "loss": 0.565,
      "num_input_tokens_seen": 1066240,
      "step": 1640
    },
    {
      "epoch": 0.9699292452830188,
      "grad_norm": 1.8959143161773682,
      "learning_rate": 4.8466981132075476e-06,
      "loss": 0.5479,
      "num_input_tokens_seen": 1068960,
      "step": 1645
    },
    {
      "epoch": 0.972877358490566,
      "grad_norm": 4.285905838012695,
      "learning_rate": 4.861438679245283e-06,
      "loss": 0.4857,
      "num_input_tokens_seen": 1071520,
      "step": 1650
    },
    {
      "epoch": 0.9758254716981132,
      "grad_norm": 2.052827835083008,
      "learning_rate": 4.876179245283019e-06,
      "loss": 0.5331,
      "num_input_tokens_seen": 1075392,
      "step": 1655
    },
    {
      "epoch": 0.9787735849056604,
      "grad_norm": 5.942412853240967,
      "learning_rate": 4.890919811320755e-06,
      "loss": 0.563,
      "num_input_tokens_seen": 1078240,
      "step": 1660
    },
    {
      "epoch": 0.9817216981132075,
      "grad_norm": 3.5858325958251953,
      "learning_rate": 4.905660377358491e-06,
      "loss": 0.5465,
      "num_input_tokens_seen": 1080992,
      "step": 1665
    },
    {
      "epoch": 0.9846698113207547,
      "grad_norm": 7.38860559463501,
      "learning_rate": 4.920400943396227e-06,
      "loss": 0.5848,
      "num_input_tokens_seen": 1083776,
      "step": 1670
    },
    {
      "epoch": 0.9876179245283019,
      "grad_norm": 2.6235125064849854,
      "learning_rate": 4.935141509433963e-06,
      "loss": 0.4852,
      "num_input_tokens_seen": 1087264,
      "step": 1675
    },
    {
      "epoch": 0.9905660377358491,
      "grad_norm": 3.0774269104003906,
      "learning_rate": 4.949882075471698e-06,
      "loss": 0.7078,
      "num_input_tokens_seen": 1091104,
      "step": 1680
    },
    {
      "epoch": 0.9935141509433962,
      "grad_norm": 5.3851165771484375,
      "learning_rate": 4.9646226415094346e-06,
      "loss": 0.5759,
      "num_input_tokens_seen": 1094528,
      "step": 1685
    },
    {
      "epoch": 0.9964622641509434,
      "grad_norm": 3.981779098510742,
      "learning_rate": 4.979363207547171e-06,
      "loss": 0.4698,
      "num_input_tokens_seen": 1098048,
      "step": 1690
    },
    {
      "epoch": 0.9994103773584906,
      "grad_norm": 14.195940971374512,
      "learning_rate": 4.994103773584906e-06,
      "loss": 0.6148,
      "num_input_tokens_seen": 1101472,
      "step": 1695
    },
    {
      "epoch": 1.0023584905660377,
      "grad_norm": 4.040242671966553,
      "learning_rate": 5.0088443396226414e-06,
      "loss": 0.4706,
      "num_input_tokens_seen": 1103824,
      "step": 1700
    },
    {
      "epoch": 1.005306603773585,
      "grad_norm": 2.6402554512023926,
      "learning_rate": 5.023584905660378e-06,
      "loss": 0.6003,
      "num_input_tokens_seen": 1107376,
      "step": 1705
    },
    {
      "epoch": 1.008254716981132,
      "grad_norm": 2.9185636043548584,
      "learning_rate": 5.038325471698113e-06,
      "loss": 0.5051,
      "num_input_tokens_seen": 1110224,
      "step": 1710
    },
    {
      "epoch": 1.0112028301886793,
      "grad_norm": 6.817721366882324,
      "learning_rate": 5.05306603773585e-06,
      "loss": 0.5182,
      "num_input_tokens_seen": 1113488,
      "step": 1715
    },
    {
      "epoch": 1.0141509433962264,
      "grad_norm": 3.6822052001953125,
      "learning_rate": 5.067806603773585e-06,
      "loss": 0.5499,
      "num_input_tokens_seen": 1116720,
      "step": 1720
    },
    {
      "epoch": 1.0170990566037736,
      "grad_norm": 2.5823006629943848,
      "learning_rate": 5.0825471698113216e-06,
      "loss": 0.37,
      "num_input_tokens_seen": 1123088,
      "step": 1725
    },
    {
      "epoch": 1.0200471698113207,
      "grad_norm": 2.5223937034606934,
      "learning_rate": 5.097287735849057e-06,
      "loss": 0.6016,
      "num_input_tokens_seen": 1125744,
      "step": 1730
    },
    {
      "epoch": 1.022995283018868,
      "grad_norm": 7.602838039398193,
      "learning_rate": 5.112028301886793e-06,
      "loss": 0.6575,
      "num_input_tokens_seen": 1128976,
      "step": 1735
    },
    {
      "epoch": 1.025943396226415,
      "grad_norm": 3.577951192855835,
      "learning_rate": 5.1267688679245284e-06,
      "loss": 0.6354,
      "num_input_tokens_seen": 1131472,
      "step": 1740
    },
    {
      "epoch": 1.0288915094339623,
      "grad_norm": 1.8606327772140503,
      "learning_rate": 5.1415094339622655e-06,
      "loss": 0.4534,
      "num_input_tokens_seen": 1134800,
      "step": 1745
    },
    {
      "epoch": 1.0318396226415094,
      "grad_norm": 3.1388463973999023,
      "learning_rate": 5.156250000000001e-06,
      "loss": 0.5906,
      "num_input_tokens_seen": 1137872,
      "step": 1750
    },
    {
      "epoch": 1.0347877358490567,
      "grad_norm": 2.9761009216308594,
      "learning_rate": 5.170990566037736e-06,
      "loss": 0.6358,
      "num_input_tokens_seen": 1140848,
      "step": 1755
    },
    {
      "epoch": 1.0377358490566038,
      "grad_norm": 4.402033805847168,
      "learning_rate": 5.185731132075472e-06,
      "loss": 0.5054,
      "num_input_tokens_seen": 1143952,
      "step": 1760
    },
    {
      "epoch": 1.040683962264151,
      "grad_norm": 3.383977174758911,
      "learning_rate": 5.200471698113208e-06,
      "loss": 0.5509,
      "num_input_tokens_seen": 1147312,
      "step": 1765
    },
    {
      "epoch": 1.0436320754716981,
      "grad_norm": 4.202183723449707,
      "learning_rate": 5.215212264150944e-06,
      "loss": 0.5797,
      "num_input_tokens_seen": 1150160,
      "step": 1770
    },
    {
      "epoch": 1.0465801886792452,
      "grad_norm": 4.749207019805908,
      "learning_rate": 5.229952830188679e-06,
      "loss": 0.5194,
      "num_input_tokens_seen": 1152560,
      "step": 1775
    },
    {
      "epoch": 1.0495283018867925,
      "grad_norm": 5.422912120819092,
      "learning_rate": 5.2446933962264154e-06,
      "loss": 0.5713,
      "num_input_tokens_seen": 1155376,
      "step": 1780
    },
    {
      "epoch": 1.0524764150943395,
      "grad_norm": 4.885079383850098,
      "learning_rate": 5.259433962264151e-06,
      "loss": 0.6607,
      "num_input_tokens_seen": 1158640,
      "step": 1785
    },
    {
      "epoch": 1.0554245283018868,
      "grad_norm": 3.0483570098876953,
      "learning_rate": 5.274174528301888e-06,
      "loss": 0.487,
      "num_input_tokens_seen": 1162128,
      "step": 1790
    },
    {
      "epoch": 1.0583726415094339,
      "grad_norm": 4.946349143981934,
      "learning_rate": 5.288915094339623e-06,
      "loss": 0.4955,
      "num_input_tokens_seen": 1165360,
      "step": 1795
    },
    {
      "epoch": 1.0613207547169812,
      "grad_norm": 4.3377299308776855,
      "learning_rate": 5.303655660377359e-06,
      "loss": 0.5888,
      "num_input_tokens_seen": 1168816,
      "step": 1800
    },
    {
      "epoch": 1.0642688679245282,
      "grad_norm": 6.346924781799316,
      "learning_rate": 5.318396226415095e-06,
      "loss": 0.6547,
      "num_input_tokens_seen": 1171184,
      "step": 1805
    },
    {
      "epoch": 1.0672169811320755,
      "grad_norm": 5.077792644500732,
      "learning_rate": 5.333136792452831e-06,
      "loss": 0.4823,
      "num_input_tokens_seen": 1173360,
      "step": 1810
    },
    {
      "epoch": 1.0701650943396226,
      "grad_norm": 2.660081624984741,
      "learning_rate": 5.347877358490566e-06,
      "loss": 0.7328,
      "num_input_tokens_seen": 1177232,
      "step": 1815
    },
    {
      "epoch": 1.0731132075471699,
      "grad_norm": 4.4345173835754395,
      "learning_rate": 5.362617924528302e-06,
      "loss": 0.6148,
      "num_input_tokens_seen": 1179792,
      "step": 1820
    },
    {
      "epoch": 1.076061320754717,
      "grad_norm": 4.01794958114624,
      "learning_rate": 5.377358490566038e-06,
      "loss": 0.4578,
      "num_input_tokens_seen": 1182640,
      "step": 1825
    },
    {
      "epoch": 1.0790094339622642,
      "grad_norm": 2.5963144302368164,
      "learning_rate": 5.392099056603775e-06,
      "loss": 0.4679,
      "num_input_tokens_seen": 1186512,
      "step": 1830
    },
    {
      "epoch": 1.0819575471698113,
      "grad_norm": 5.90488338470459,
      "learning_rate": 5.40683962264151e-06,
      "loss": 0.6515,
      "num_input_tokens_seen": 1188944,
      "step": 1835
    },
    {
      "epoch": 1.0849056603773586,
      "grad_norm": 1.8786847591400146,
      "learning_rate": 5.4215801886792455e-06,
      "loss": 0.5948,
      "num_input_tokens_seen": 1192048,
      "step": 1840
    },
    {
      "epoch": 1.0878537735849056,
      "grad_norm": 3.625117301940918,
      "learning_rate": 5.436320754716982e-06,
      "loss": 0.7595,
      "num_input_tokens_seen": 1195696,
      "step": 1845
    },
    {
      "epoch": 1.0908018867924527,
      "grad_norm": 3.2458336353302,
      "learning_rate": 5.451061320754717e-06,
      "loss": 0.4481,
      "num_input_tokens_seen": 1198000,
      "step": 1850
    },
    {
      "epoch": 1.09375,
      "grad_norm": 5.10528564453125,
      "learning_rate": 5.465801886792453e-06,
      "loss": 0.75,
      "num_input_tokens_seen": 1200784,
      "step": 1855
    },
    {
      "epoch": 1.0966981132075473,
      "grad_norm": 8.545928001403809,
      "learning_rate": 5.4805424528301886e-06,
      "loss": 0.6616,
      "num_input_tokens_seen": 1203344,
      "step": 1860
    },
    {
      "epoch": 1.0996462264150944,
      "grad_norm": 4.458759784698486,
      "learning_rate": 5.495283018867925e-06,
      "loss": 0.6007,
      "num_input_tokens_seen": 1205872,
      "step": 1865
    },
    {
      "epoch": 1.1025943396226414,
      "grad_norm": 2.931718587875366,
      "learning_rate": 5.51002358490566e-06,
      "loss": 0.5431,
      "num_input_tokens_seen": 1210032,
      "step": 1870
    },
    {
      "epoch": 1.1055424528301887,
      "grad_norm": 2.582240104675293,
      "learning_rate": 5.524764150943397e-06,
      "loss": 0.6471,
      "num_input_tokens_seen": 1213104,
      "step": 1875
    },
    {
      "epoch": 1.1084905660377358,
      "grad_norm": 2.3504467010498047,
      "learning_rate": 5.5395047169811325e-06,
      "loss": 0.431,
      "num_input_tokens_seen": 1216176,
      "step": 1880
    },
    {
      "epoch": 1.111438679245283,
      "grad_norm": 2.5821731090545654,
      "learning_rate": 5.554245283018869e-06,
      "loss": 0.4977,
      "num_input_tokens_seen": 1218864,
      "step": 1885
    },
    {
      "epoch": 1.1143867924528301,
      "grad_norm": 2.777735948562622,
      "learning_rate": 5.568985849056604e-06,
      "loss": 0.6474,
      "num_input_tokens_seen": 1221712,
      "step": 1890
    },
    {
      "epoch": 1.1173349056603774,
      "grad_norm": 3.6763858795166016,
      "learning_rate": 5.58372641509434e-06,
      "loss": 0.7935,
      "num_input_tokens_seen": 1225040,
      "step": 1895
    },
    {
      "epoch": 1.1202830188679245,
      "grad_norm": 3.1717238426208496,
      "learning_rate": 5.5984669811320755e-06,
      "loss": 0.7487,
      "num_input_tokens_seen": 1228336,
      "step": 1900
    },
    {
      "epoch": 1.1232311320754718,
      "grad_norm": 4.433318138122559,
      "learning_rate": 5.613207547169813e-06,
      "loss": 0.5584,
      "num_input_tokens_seen": 1230960,
      "step": 1905
    },
    {
      "epoch": 1.1261792452830188,
      "grad_norm": 3.0675339698791504,
      "learning_rate": 5.627948113207548e-06,
      "loss": 0.4734,
      "num_input_tokens_seen": 1233968,
      "step": 1910
    },
    {
      "epoch": 1.1291273584905661,
      "grad_norm": 3.6336021423339844,
      "learning_rate": 5.642688679245284e-06,
      "loss": 0.5711,
      "num_input_tokens_seen": 1237552,
      "step": 1915
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 3.4513232707977295,
      "learning_rate": 5.6574292452830195e-06,
      "loss": 0.5364,
      "num_input_tokens_seen": 1240720,
      "step": 1920
    },
    {
      "epoch": 1.1350235849056605,
      "grad_norm": 2.3215417861938477,
      "learning_rate": 5.672169811320756e-06,
      "loss": 0.5174,
      "num_input_tokens_seen": 1243632,
      "step": 1925
    },
    {
      "epoch": 1.1379716981132075,
      "grad_norm": 2.0179061889648438,
      "learning_rate": 5.686910377358491e-06,
      "loss": 0.4183,
      "num_input_tokens_seen": 1247472,
      "step": 1930
    },
    {
      "epoch": 1.1409198113207548,
      "grad_norm": 5.417402744293213,
      "learning_rate": 5.701650943396226e-06,
      "loss": 0.5346,
      "num_input_tokens_seen": 1250256,
      "step": 1935
    },
    {
      "epoch": 1.1438679245283019,
      "grad_norm": 2.1405327320098877,
      "learning_rate": 5.7163915094339625e-06,
      "loss": 0.442,
      "num_input_tokens_seen": 1255952,
      "step": 1940
    },
    {
      "epoch": 1.146816037735849,
      "grad_norm": 2.379966974258423,
      "learning_rate": 5.731132075471698e-06,
      "loss": 0.5008,
      "num_input_tokens_seen": 1259248,
      "step": 1945
    },
    {
      "epoch": 1.1497641509433962,
      "grad_norm": 4.080671310424805,
      "learning_rate": 5.745872641509435e-06,
      "loss": 0.6445,
      "num_input_tokens_seen": 1261776,
      "step": 1950
    },
    {
      "epoch": 1.1527122641509433,
      "grad_norm": 3.1824378967285156,
      "learning_rate": 5.76061320754717e-06,
      "loss": 0.4763,
      "num_input_tokens_seen": 1265328,
      "step": 1955
    },
    {
      "epoch": 1.1556603773584906,
      "grad_norm": 4.271039009094238,
      "learning_rate": 5.7753537735849065e-06,
      "loss": 0.5789,
      "num_input_tokens_seen": 1268976,
      "step": 1960
    },
    {
      "epoch": 1.1586084905660377,
      "grad_norm": 5.472548484802246,
      "learning_rate": 5.790094339622642e-06,
      "loss": 0.6046,
      "num_input_tokens_seen": 1272592,
      "step": 1965
    },
    {
      "epoch": 1.161556603773585,
      "grad_norm": 3.577645778656006,
      "learning_rate": 5.804834905660378e-06,
      "loss": 0.5593,
      "num_input_tokens_seen": 1275632,
      "step": 1970
    },
    {
      "epoch": 1.164504716981132,
      "grad_norm": 4.544173240661621,
      "learning_rate": 5.819575471698113e-06,
      "loss": 0.6391,
      "num_input_tokens_seen": 1278128,
      "step": 1975
    },
    {
      "epoch": 1.1674528301886793,
      "grad_norm": 2.189887285232544,
      "learning_rate": 5.8343160377358495e-06,
      "loss": 0.5317,
      "num_input_tokens_seen": 1281008,
      "step": 1980
    },
    {
      "epoch": 1.1704009433962264,
      "grad_norm": 2.237506866455078,
      "learning_rate": 5.849056603773585e-06,
      "loss": 0.5157,
      "num_input_tokens_seen": 1284112,
      "step": 1985
    },
    {
      "epoch": 1.1733490566037736,
      "grad_norm": 3.1468422412872314,
      "learning_rate": 5.863797169811322e-06,
      "loss": 0.5331,
      "num_input_tokens_seen": 1287184,
      "step": 1990
    },
    {
      "epoch": 1.1762971698113207,
      "grad_norm": 4.18142032623291,
      "learning_rate": 5.878537735849057e-06,
      "loss": 0.544,
      "num_input_tokens_seen": 1289744,
      "step": 1995
    },
    {
      "epoch": 1.179245283018868,
      "grad_norm": 4.273591995239258,
      "learning_rate": 5.8932783018867934e-06,
      "loss": 0.5721,
      "num_input_tokens_seen": 1292400,
      "step": 2000
    },
    {
      "epoch": 1.182193396226415,
      "grad_norm": 2.5647614002227783,
      "learning_rate": 5.908018867924529e-06,
      "loss": 0.5827,
      "num_input_tokens_seen": 1295024,
      "step": 2005
    },
    {
      "epoch": 1.1851415094339623,
      "grad_norm": 2.451751232147217,
      "learning_rate": 5.922759433962265e-06,
      "loss": 0.4873,
      "num_input_tokens_seen": 1298416,
      "step": 2010
    },
    {
      "epoch": 1.1880896226415094,
      "grad_norm": 4.439822673797607,
      "learning_rate": 5.9375e-06,
      "loss": 0.4047,
      "num_input_tokens_seen": 1301744,
      "step": 2015
    },
    {
      "epoch": 1.1910377358490567,
      "grad_norm": 2.208664655685425,
      "learning_rate": 5.952240566037736e-06,
      "loss": 0.5501,
      "num_input_tokens_seen": 1305808,
      "step": 2020
    },
    {
      "epoch": 1.1939858490566038,
      "grad_norm": 2.6934683322906494,
      "learning_rate": 5.966981132075472e-06,
      "loss": 0.4435,
      "num_input_tokens_seen": 1309072,
      "step": 2025
    },
    {
      "epoch": 1.196933962264151,
      "grad_norm": 3.69163179397583,
      "learning_rate": 5.981721698113207e-06,
      "loss": 0.4874,
      "num_input_tokens_seen": 1312272,
      "step": 2030
    },
    {
      "epoch": 1.1998820754716981,
      "grad_norm": 4.4948015213012695,
      "learning_rate": 5.996462264150944e-06,
      "loss": 0.5866,
      "num_input_tokens_seen": 1315120,
      "step": 2035
    },
    {
      "epoch": 1.2028301886792452,
      "grad_norm": 6.474124908447266,
      "learning_rate": 6.01120283018868e-06,
      "loss": 0.7569,
      "num_input_tokens_seen": 1318384,
      "step": 2040
    },
    {
      "epoch": 1.2057783018867925,
      "grad_norm": 3.1053080558776855,
      "learning_rate": 6.025943396226416e-06,
      "loss": 0.4666,
      "num_input_tokens_seen": 1321392,
      "step": 2045
    },
    {
      "epoch": 1.2087264150943395,
      "grad_norm": 2.250727891921997,
      "learning_rate": 6.040683962264151e-06,
      "loss": 0.5319,
      "num_input_tokens_seen": 1324080,
      "step": 2050
    },
    {
      "epoch": 1.2116745283018868,
      "grad_norm": 2.664297580718994,
      "learning_rate": 6.055424528301887e-06,
      "loss": 0.6366,
      "num_input_tokens_seen": 1327472,
      "step": 2055
    },
    {
      "epoch": 1.2146226415094339,
      "grad_norm": 2.7943365573883057,
      "learning_rate": 6.070165094339623e-06,
      "loss": 0.5207,
      "num_input_tokens_seen": 1331120,
      "step": 2060
    },
    {
      "epoch": 1.2175707547169812,
      "grad_norm": 3.0744247436523438,
      "learning_rate": 6.08490566037736e-06,
      "loss": 0.4973,
      "num_input_tokens_seen": 1334352,
      "step": 2065
    },
    {
      "epoch": 1.2205188679245282,
      "grad_norm": 2.0054001808166504,
      "learning_rate": 6.099646226415095e-06,
      "loss": 0.5436,
      "num_input_tokens_seen": 1337488,
      "step": 2070
    },
    {
      "epoch": 1.2234669811320755,
      "grad_norm": 2.0574240684509277,
      "learning_rate": 6.114386792452831e-06,
      "loss": 0.6138,
      "num_input_tokens_seen": 1340880,
      "step": 2075
    },
    {
      "epoch": 1.2264150943396226,
      "grad_norm": 2.7400941848754883,
      "learning_rate": 6.129127358490567e-06,
      "loss": 0.5281,
      "num_input_tokens_seen": 1343824,
      "step": 2080
    },
    {
      "epoch": 1.2293632075471699,
      "grad_norm": 3.660130739212036,
      "learning_rate": 6.143867924528303e-06,
      "loss": 0.4836,
      "num_input_tokens_seen": 1347952,
      "step": 2085
    },
    {
      "epoch": 1.232311320754717,
      "grad_norm": 4.123905658721924,
      "learning_rate": 6.158608490566038e-06,
      "loss": 0.5088,
      "num_input_tokens_seen": 1350832,
      "step": 2090
    },
    {
      "epoch": 1.2352594339622642,
      "grad_norm": 2.054478168487549,
      "learning_rate": 6.173349056603774e-06,
      "loss": 0.4171,
      "num_input_tokens_seen": 1353584,
      "step": 2095
    },
    {
      "epoch": 1.2382075471698113,
      "grad_norm": 3.449887275695801,
      "learning_rate": 6.18808962264151e-06,
      "loss": 0.5914,
      "num_input_tokens_seen": 1356688,
      "step": 2100
    },
    {
      "epoch": 1.2411556603773586,
      "grad_norm": 1.8288284540176392,
      "learning_rate": 6.202830188679245e-06,
      "loss": 0.6217,
      "num_input_tokens_seen": 1361424,
      "step": 2105
    },
    {
      "epoch": 1.2441037735849056,
      "grad_norm": 1.8224109411239624,
      "learning_rate": 6.217570754716982e-06,
      "loss": 0.4945,
      "num_input_tokens_seen": 1364208,
      "step": 2110
    },
    {
      "epoch": 1.2470518867924527,
      "grad_norm": 2.7721822261810303,
      "learning_rate": 6.232311320754717e-06,
      "loss": 0.5517,
      "num_input_tokens_seen": 1366672,
      "step": 2115
    },
    {
      "epoch": 1.25,
      "grad_norm": 2.4963698387145996,
      "learning_rate": 6.2470518867924536e-06,
      "loss": 0.682,
      "num_input_tokens_seen": 1370960,
      "step": 2120
    },
    {
      "epoch": 1.2529481132075473,
      "grad_norm": 2.3604891300201416,
      "learning_rate": 6.261792452830189e-06,
      "loss": 0.7526,
      "num_input_tokens_seen": 1374576,
      "step": 2125
    },
    {
      "epoch": 1.2558962264150944,
      "grad_norm": 4.6909661293029785,
      "learning_rate": 6.276533018867925e-06,
      "loss": 0.5547,
      "num_input_tokens_seen": 1378224,
      "step": 2130
    },
    {
      "epoch": 1.2588443396226414,
      "grad_norm": 3.194629430770874,
      "learning_rate": 6.2912735849056604e-06,
      "loss": 0.5242,
      "num_input_tokens_seen": 1381456,
      "step": 2135
    },
    {
      "epoch": 1.2617924528301887,
      "grad_norm": 3.015857219696045,
      "learning_rate": 6.306014150943397e-06,
      "loss": 0.6006,
      "num_input_tokens_seen": 1384880,
      "step": 2140
    },
    {
      "epoch": 1.2647405660377358,
      "grad_norm": 1.8358039855957031,
      "learning_rate": 6.320754716981132e-06,
      "loss": 0.4812,
      "num_input_tokens_seen": 1387632,
      "step": 2145
    },
    {
      "epoch": 1.267688679245283,
      "grad_norm": 3.8806540966033936,
      "learning_rate": 6.335495283018869e-06,
      "loss": 0.4428,
      "num_input_tokens_seen": 1391408,
      "step": 2150
    },
    {
      "epoch": 1.2706367924528301,
      "grad_norm": 2.5210306644439697,
      "learning_rate": 6.350235849056604e-06,
      "loss": 0.6868,
      "num_input_tokens_seen": 1394000,
      "step": 2155
    },
    {
      "epoch": 1.2735849056603774,
      "grad_norm": 3.6482436656951904,
      "learning_rate": 6.3649764150943406e-06,
      "loss": 0.6333,
      "num_input_tokens_seen": 1396080,
      "step": 2160
    },
    {
      "epoch": 1.2765330188679245,
      "grad_norm": 4.942153453826904,
      "learning_rate": 6.379716981132076e-06,
      "loss": 0.4526,
      "num_input_tokens_seen": 1399344,
      "step": 2165
    },
    {
      "epoch": 1.2794811320754718,
      "grad_norm": 5.084027290344238,
      "learning_rate": 6.394457547169812e-06,
      "loss": 0.6076,
      "num_input_tokens_seen": 1402896,
      "step": 2170
    },
    {
      "epoch": 1.2824292452830188,
      "grad_norm": 2.880201578140259,
      "learning_rate": 6.4091981132075474e-06,
      "loss": 0.406,
      "num_input_tokens_seen": 1406256,
      "step": 2175
    },
    {
      "epoch": 1.2853773584905661,
      "grad_norm": 5.455946445465088,
      "learning_rate": 6.423938679245284e-06,
      "loss": 0.45,
      "num_input_tokens_seen": 1409968,
      "step": 2180
    },
    {
      "epoch": 1.2883254716981132,
      "grad_norm": 3.8955984115600586,
      "learning_rate": 6.438679245283019e-06,
      "loss": 0.5698,
      "num_input_tokens_seen": 1413424,
      "step": 2185
    },
    {
      "epoch": 1.2912735849056602,
      "grad_norm": 3.811736583709717,
      "learning_rate": 6.453419811320756e-06,
      "loss": 0.4061,
      "num_input_tokens_seen": 1416464,
      "step": 2190
    },
    {
      "epoch": 1.2942216981132075,
      "grad_norm": 3.696631908416748,
      "learning_rate": 6.468160377358491e-06,
      "loss": 0.5561,
      "num_input_tokens_seen": 1419024,
      "step": 2195
    },
    {
      "epoch": 1.2971698113207548,
      "grad_norm": 7.697360038757324,
      "learning_rate": 6.482900943396227e-06,
      "loss": 0.5436,
      "num_input_tokens_seen": 1421712,
      "step": 2200
    },
    {
      "epoch": 1.3001179245283019,
      "grad_norm": 6.749332427978516,
      "learning_rate": 6.497641509433963e-06,
      "loss": 0.4913,
      "num_input_tokens_seen": 1425360,
      "step": 2205
    },
    {
      "epoch": 1.303066037735849,
      "grad_norm": 2.4801857471466064,
      "learning_rate": 6.512382075471698e-06,
      "loss": 0.5063,
      "num_input_tokens_seen": 1429200,
      "step": 2210
    },
    {
      "epoch": 1.3060141509433962,
      "grad_norm": 4.060877799987793,
      "learning_rate": 6.5271226415094344e-06,
      "loss": 0.5533,
      "num_input_tokens_seen": 1432144,
      "step": 2215
    },
    {
      "epoch": 1.3089622641509435,
      "grad_norm": 2.647620439529419,
      "learning_rate": 6.54186320754717e-06,
      "loss": 0.5037,
      "num_input_tokens_seen": 1434416,
      "step": 2220
    },
    {
      "epoch": 1.3119103773584906,
      "grad_norm": 3.3790409564971924,
      "learning_rate": 6.556603773584907e-06,
      "loss": 0.6792,
      "num_input_tokens_seen": 1438512,
      "step": 2225
    },
    {
      "epoch": 1.3148584905660377,
      "grad_norm": 1.7855783700942993,
      "learning_rate": 6.571344339622641e-06,
      "loss": 0.5291,
      "num_input_tokens_seen": 1441840,
      "step": 2230
    },
    {
      "epoch": 1.317806603773585,
      "grad_norm": 2.3709962368011475,
      "learning_rate": 6.586084905660378e-06,
      "loss": 0.4882,
      "num_input_tokens_seen": 1449232,
      "step": 2235
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 1.5067445039749146,
      "learning_rate": 6.600825471698114e-06,
      "loss": 0.4222,
      "num_input_tokens_seen": 1452336,
      "step": 2240
    },
    {
      "epoch": 1.3237028301886793,
      "grad_norm": 1.9331886768341064,
      "learning_rate": 6.61556603773585e-06,
      "loss": 0.4299,
      "num_input_tokens_seen": 1455184,
      "step": 2245
    },
    {
      "epoch": 1.3266509433962264,
      "grad_norm": 2.3233721256256104,
      "learning_rate": 6.630306603773585e-06,
      "loss": 0.5345,
      "num_input_tokens_seen": 1458448,
      "step": 2250
    },
    {
      "epoch": 1.3295990566037736,
      "grad_norm": 2.396291494369507,
      "learning_rate": 6.645047169811321e-06,
      "loss": 0.6097,
      "num_input_tokens_seen": 1461872,
      "step": 2255
    },
    {
      "epoch": 1.3325471698113207,
      "grad_norm": 1.6488536596298218,
      "learning_rate": 6.659787735849057e-06,
      "loss": 0.4878,
      "num_input_tokens_seen": 1465360,
      "step": 2260
    },
    {
      "epoch": 1.335495283018868,
      "grad_norm": 3.878967046737671,
      "learning_rate": 6.674528301886794e-06,
      "loss": 0.5682,
      "num_input_tokens_seen": 1468048,
      "step": 2265
    },
    {
      "epoch": 1.338443396226415,
      "grad_norm": 3.165295124053955,
      "learning_rate": 6.689268867924529e-06,
      "loss": 0.5191,
      "num_input_tokens_seen": 1471728,
      "step": 2270
    },
    {
      "epoch": 1.3413915094339623,
      "grad_norm": 2.476186513900757,
      "learning_rate": 6.704009433962265e-06,
      "loss": 0.5771,
      "num_input_tokens_seen": 1474832,
      "step": 2275
    },
    {
      "epoch": 1.3443396226415094,
      "grad_norm": 3.1725950241088867,
      "learning_rate": 6.718750000000001e-06,
      "loss": 0.4749,
      "num_input_tokens_seen": 1477392,
      "step": 2280
    },
    {
      "epoch": 1.3472877358490565,
      "grad_norm": 2.118039131164551,
      "learning_rate": 6.733490566037736e-06,
      "loss": 0.5764,
      "num_input_tokens_seen": 1480144,
      "step": 2285
    },
    {
      "epoch": 1.3502358490566038,
      "grad_norm": 3.6634762287139893,
      "learning_rate": 6.748231132075472e-06,
      "loss": 0.5651,
      "num_input_tokens_seen": 1484016,
      "step": 2290
    },
    {
      "epoch": 1.353183962264151,
      "grad_norm": 2.3556525707244873,
      "learning_rate": 6.7629716981132076e-06,
      "loss": 0.3681,
      "num_input_tokens_seen": 1487120,
      "step": 2295
    },
    {
      "epoch": 1.3561320754716981,
      "grad_norm": 2.9869425296783447,
      "learning_rate": 6.777712264150944e-06,
      "loss": 0.5036,
      "num_input_tokens_seen": 1490352,
      "step": 2300
    },
    {
      "epoch": 1.3590801886792452,
      "grad_norm": 3.2929322719573975,
      "learning_rate": 6.792452830188679e-06,
      "loss": 0.7121,
      "num_input_tokens_seen": 1493520,
      "step": 2305
    },
    {
      "epoch": 1.3620283018867925,
      "grad_norm": 3.118351936340332,
      "learning_rate": 6.807193396226416e-06,
      "loss": 0.6066,
      "num_input_tokens_seen": 1496432,
      "step": 2310
    },
    {
      "epoch": 1.3649764150943398,
      "grad_norm": 8.196702003479004,
      "learning_rate": 6.8219339622641515e-06,
      "loss": 0.6051,
      "num_input_tokens_seen": 1498736,
      "step": 2315
    },
    {
      "epoch": 1.3679245283018868,
      "grad_norm": 4.0915656089782715,
      "learning_rate": 6.836674528301888e-06,
      "loss": 0.3932,
      "num_input_tokens_seen": 1501136,
      "step": 2320
    },
    {
      "epoch": 1.3708726415094339,
      "grad_norm": 3.9975690841674805,
      "learning_rate": 6.851415094339623e-06,
      "loss": 0.5328,
      "num_input_tokens_seen": 1503824,
      "step": 2325
    },
    {
      "epoch": 1.3738207547169812,
      "grad_norm": 2.0557429790496826,
      "learning_rate": 6.866155660377359e-06,
      "loss": 0.4687,
      "num_input_tokens_seen": 1506608,
      "step": 2330
    },
    {
      "epoch": 1.3767688679245282,
      "grad_norm": 2.1996331214904785,
      "learning_rate": 6.8808962264150946e-06,
      "loss": 0.5752,
      "num_input_tokens_seen": 1509456,
      "step": 2335
    },
    {
      "epoch": 1.3797169811320755,
      "grad_norm": 6.477305889129639,
      "learning_rate": 6.895636792452831e-06,
      "loss": 0.4768,
      "num_input_tokens_seen": 1512176,
      "step": 2340
    },
    {
      "epoch": 1.3826650943396226,
      "grad_norm": 4.112719535827637,
      "learning_rate": 6.910377358490566e-06,
      "loss": 0.3502,
      "num_input_tokens_seen": 1514960,
      "step": 2345
    },
    {
      "epoch": 1.3856132075471699,
      "grad_norm": 2.5176162719726562,
      "learning_rate": 6.925117924528303e-06,
      "loss": 0.5708,
      "num_input_tokens_seen": 1519280,
      "step": 2350
    },
    {
      "epoch": 1.388561320754717,
      "grad_norm": 1.7241085767745972,
      "learning_rate": 6.9398584905660385e-06,
      "loss": 0.5526,
      "num_input_tokens_seen": 1523376,
      "step": 2355
    },
    {
      "epoch": 1.3915094339622642,
      "grad_norm": 7.234544277191162,
      "learning_rate": 6.954599056603775e-06,
      "loss": 0.6993,
      "num_input_tokens_seen": 1526864,
      "step": 2360
    },
    {
      "epoch": 1.3944575471698113,
      "grad_norm": 2.258788824081421,
      "learning_rate": 6.96933962264151e-06,
      "loss": 0.6257,
      "num_input_tokens_seen": 1529680,
      "step": 2365
    },
    {
      "epoch": 1.3974056603773586,
      "grad_norm": 2.660125970840454,
      "learning_rate": 6.984080188679245e-06,
      "loss": 0.4141,
      "num_input_tokens_seen": 1533040,
      "step": 2370
    },
    {
      "epoch": 1.4003537735849056,
      "grad_norm": 5.061126708984375,
      "learning_rate": 6.9988207547169815e-06,
      "loss": 0.5721,
      "num_input_tokens_seen": 1535344,
      "step": 2375
    },
    {
      "epoch": 1.4033018867924527,
      "grad_norm": 2.6086678504943848,
      "learning_rate": 7.013561320754717e-06,
      "loss": 0.7886,
      "num_input_tokens_seen": 1538256,
      "step": 2380
    },
    {
      "epoch": 1.40625,
      "grad_norm": 3.3688604831695557,
      "learning_rate": 7.028301886792454e-06,
      "loss": 0.4302,
      "num_input_tokens_seen": 1542416,
      "step": 2385
    },
    {
      "epoch": 1.4091981132075473,
      "grad_norm": 2.5950000286102295,
      "learning_rate": 7.043042452830188e-06,
      "loss": 0.5975,
      "num_input_tokens_seen": 1545872,
      "step": 2390
    },
    {
      "epoch": 1.4121462264150944,
      "grad_norm": 2.9001176357269287,
      "learning_rate": 7.0577830188679255e-06,
      "loss": 0.6648,
      "num_input_tokens_seen": 1550032,
      "step": 2395
    },
    {
      "epoch": 1.4150943396226414,
      "grad_norm": 2.5995712280273438,
      "learning_rate": 7.072523584905661e-06,
      "loss": 0.6227,
      "num_input_tokens_seen": 1553136,
      "step": 2400
    },
    {
      "epoch": 1.4180424528301887,
      "grad_norm": 2.611138343811035,
      "learning_rate": 7.087264150943397e-06,
      "loss": 0.4454,
      "num_input_tokens_seen": 1555888,
      "step": 2405
    },
    {
      "epoch": 1.4209905660377358,
      "grad_norm": 3.2860090732574463,
      "learning_rate": 7.102004716981132e-06,
      "loss": 0.464,
      "num_input_tokens_seen": 1558384,
      "step": 2410
    },
    {
      "epoch": 1.423938679245283,
      "grad_norm": 3.546232223510742,
      "learning_rate": 7.1167452830188685e-06,
      "loss": 0.5115,
      "num_input_tokens_seen": 1561488,
      "step": 2415
    },
    {
      "epoch": 1.4268867924528301,
      "grad_norm": 6.739683151245117,
      "learning_rate": 7.131485849056604e-06,
      "loss": 0.5805,
      "num_input_tokens_seen": 1564080,
      "step": 2420
    },
    {
      "epoch": 1.4298349056603774,
      "grad_norm": 2.0999245643615723,
      "learning_rate": 7.146226415094341e-06,
      "loss": 0.4481,
      "num_input_tokens_seen": 1567024,
      "step": 2425
    },
    {
      "epoch": 1.4327830188679245,
      "grad_norm": 2.5809998512268066,
      "learning_rate": 7.160966981132076e-06,
      "loss": 0.5088,
      "num_input_tokens_seen": 1570096,
      "step": 2430
    },
    {
      "epoch": 1.4357311320754718,
      "grad_norm": 3.5544846057891846,
      "learning_rate": 7.1757075471698125e-06,
      "loss": 0.4573,
      "num_input_tokens_seen": 1572688,
      "step": 2435
    },
    {
      "epoch": 1.4386792452830188,
      "grad_norm": 3.3734724521636963,
      "learning_rate": 7.190448113207548e-06,
      "loss": 0.594,
      "num_input_tokens_seen": 1577168,
      "step": 2440
    },
    {
      "epoch": 1.4416273584905661,
      "grad_norm": 2.0420730113983154,
      "learning_rate": 7.205188679245284e-06,
      "loss": 0.4288,
      "num_input_tokens_seen": 1580400,
      "step": 2445
    },
    {
      "epoch": 1.4445754716981132,
      "grad_norm": 2.819896697998047,
      "learning_rate": 7.219929245283019e-06,
      "loss": 0.5405,
      "num_input_tokens_seen": 1583344,
      "step": 2450
    },
    {
      "epoch": 1.4475235849056602,
      "grad_norm": 2.1295816898345947,
      "learning_rate": 7.2346698113207555e-06,
      "loss": 0.583,
      "num_input_tokens_seen": 1586768,
      "step": 2455
    },
    {
      "epoch": 1.4504716981132075,
      "grad_norm": 2.5006332397460938,
      "learning_rate": 7.249410377358491e-06,
      "loss": 0.5313,
      "num_input_tokens_seen": 1590160,
      "step": 2460
    },
    {
      "epoch": 1.4534198113207548,
      "grad_norm": 1.7471357583999634,
      "learning_rate": 7.264150943396226e-06,
      "loss": 0.5071,
      "num_input_tokens_seen": 1593392,
      "step": 2465
    },
    {
      "epoch": 1.4563679245283019,
      "grad_norm": 2.186572313308716,
      "learning_rate": 7.278891509433963e-06,
      "loss": 0.5387,
      "num_input_tokens_seen": 1596432,
      "step": 2470
    },
    {
      "epoch": 1.459316037735849,
      "grad_norm": 5.838025093078613,
      "learning_rate": 7.293632075471699e-06,
      "loss": 0.4679,
      "num_input_tokens_seen": 1599120,
      "step": 2475
    },
    {
      "epoch": 1.4622641509433962,
      "grad_norm": 3.7279281616210938,
      "learning_rate": 7.308372641509435e-06,
      "loss": 0.594,
      "num_input_tokens_seen": 1602160,
      "step": 2480
    },
    {
      "epoch": 1.4652122641509435,
      "grad_norm": 3.973458766937256,
      "learning_rate": 7.32311320754717e-06,
      "loss": 0.4877,
      "num_input_tokens_seen": 1604848,
      "step": 2485
    },
    {
      "epoch": 1.4681603773584906,
      "grad_norm": 2.3970954418182373,
      "learning_rate": 7.337853773584906e-06,
      "loss": 0.5393,
      "num_input_tokens_seen": 1608080,
      "step": 2490
    },
    {
      "epoch": 1.4711084905660377,
      "grad_norm": 2.5397684574127197,
      "learning_rate": 7.352594339622642e-06,
      "loss": 0.4643,
      "num_input_tokens_seen": 1610576,
      "step": 2495
    },
    {
      "epoch": 1.474056603773585,
      "grad_norm": 2.2447030544281006,
      "learning_rate": 7.367334905660378e-06,
      "loss": 0.4887,
      "num_input_tokens_seen": 1613744,
      "step": 2500
    },
    {
      "epoch": 1.477004716981132,
      "grad_norm": 2.397318124771118,
      "learning_rate": 7.382075471698113e-06,
      "loss": 0.4703,
      "num_input_tokens_seen": 1616848,
      "step": 2505
    },
    {
      "epoch": 1.4799528301886793,
      "grad_norm": 3.7728283405303955,
      "learning_rate": 7.39681603773585e-06,
      "loss": 0.6162,
      "num_input_tokens_seen": 1622064,
      "step": 2510
    },
    {
      "epoch": 1.4829009433962264,
      "grad_norm": 3.109128713607788,
      "learning_rate": 7.411556603773586e-06,
      "loss": 0.4801,
      "num_input_tokens_seen": 1624880,
      "step": 2515
    },
    {
      "epoch": 1.4858490566037736,
      "grad_norm": 5.143686771392822,
      "learning_rate": 7.426297169811322e-06,
      "loss": 0.569,
      "num_input_tokens_seen": 1627664,
      "step": 2520
    },
    {
      "epoch": 1.4887971698113207,
      "grad_norm": 2.025239944458008,
      "learning_rate": 7.441037735849057e-06,
      "loss": 0.3637,
      "num_input_tokens_seen": 1630512,
      "step": 2525
    },
    {
      "epoch": 1.491745283018868,
      "grad_norm": 1.8764909505844116,
      "learning_rate": 7.455778301886793e-06,
      "loss": 0.4793,
      "num_input_tokens_seen": 1633488,
      "step": 2530
    },
    {
      "epoch": 1.494693396226415,
      "grad_norm": 4.40326452255249,
      "learning_rate": 7.470518867924529e-06,
      "loss": 0.6004,
      "num_input_tokens_seen": 1636208,
      "step": 2535
    },
    {
      "epoch": 1.4976415094339623,
      "grad_norm": 2.668639898300171,
      "learning_rate": 7.485259433962266e-06,
      "loss": 0.477,
      "num_input_tokens_seen": 1639664,
      "step": 2540
    },
    {
      "epoch": 1.5005896226415094,
      "grad_norm": 2.34299635887146,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.5321,
      "num_input_tokens_seen": 1642320,
      "step": 2545
    },
    {
      "epoch": 1.5035377358490565,
      "grad_norm": 3.1393096446990967,
      "learning_rate": 7.5147405660377355e-06,
      "loss": 0.4773,
      "num_input_tokens_seen": 1645648,
      "step": 2550
    },
    {
      "epoch": 1.5064858490566038,
      "grad_norm": 1.8545193672180176,
      "learning_rate": 7.5294811320754726e-06,
      "loss": 0.4372,
      "num_input_tokens_seen": 1650096,
      "step": 2555
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 4.3445563316345215,
      "learning_rate": 7.544221698113208e-06,
      "loss": 0.5729,
      "num_input_tokens_seen": 1653424,
      "step": 2560
    },
    {
      "epoch": 1.5123820754716981,
      "grad_norm": 1.6782920360565186,
      "learning_rate": 7.558962264150944e-06,
      "loss": 0.5918,
      "num_input_tokens_seen": 1658864,
      "step": 2565
    },
    {
      "epoch": 1.5153301886792452,
      "grad_norm": 1.7342463731765747,
      "learning_rate": 7.5737028301886795e-06,
      "loss": 0.5201,
      "num_input_tokens_seen": 1661616,
      "step": 2570
    },
    {
      "epoch": 1.5182783018867925,
      "grad_norm": 3.153855800628662,
      "learning_rate": 7.588443396226416e-06,
      "loss": 0.4874,
      "num_input_tokens_seen": 1665008,
      "step": 2575
    },
    {
      "epoch": 1.5212264150943398,
      "grad_norm": 7.022671699523926,
      "learning_rate": 7.603183962264151e-06,
      "loss": 0.7007,
      "num_input_tokens_seen": 1668816,
      "step": 2580
    },
    {
      "epoch": 1.5241745283018868,
      "grad_norm": 2.978729724884033,
      "learning_rate": 7.617924528301888e-06,
      "loss": 0.424,
      "num_input_tokens_seen": 1671312,
      "step": 2585
    },
    {
      "epoch": 1.5271226415094339,
      "grad_norm": 1.6254667043685913,
      "learning_rate": 7.632665094339623e-06,
      "loss": 0.5506,
      "num_input_tokens_seen": 1674480,
      "step": 2590
    },
    {
      "epoch": 1.5300707547169812,
      "grad_norm": 2.831998825073242,
      "learning_rate": 7.64740566037736e-06,
      "loss": 0.5584,
      "num_input_tokens_seen": 1678160,
      "step": 2595
    },
    {
      "epoch": 1.5330188679245285,
      "grad_norm": 2.198988437652588,
      "learning_rate": 7.662146226415095e-06,
      "loss": 0.5155,
      "num_input_tokens_seen": 1680848,
      "step": 2600
    },
    {
      "epoch": 1.5359669811320755,
      "grad_norm": 2.986589193344116,
      "learning_rate": 7.676886792452832e-06,
      "loss": 0.6604,
      "num_input_tokens_seen": 1683792,
      "step": 2605
    },
    {
      "epoch": 1.5389150943396226,
      "grad_norm": 3.1608846187591553,
      "learning_rate": 7.691627358490567e-06,
      "loss": 0.4826,
      "num_input_tokens_seen": 1687344,
      "step": 2610
    },
    {
      "epoch": 1.5418632075471699,
      "grad_norm": 2.4701035022735596,
      "learning_rate": 7.706367924528303e-06,
      "loss": 0.8183,
      "num_input_tokens_seen": 1690768,
      "step": 2615
    },
    {
      "epoch": 1.544811320754717,
      "grad_norm": 3.2683584690093994,
      "learning_rate": 7.721108490566038e-06,
      "loss": 0.6016,
      "num_input_tokens_seen": 1694384,
      "step": 2620
    },
    {
      "epoch": 1.547759433962264,
      "grad_norm": 2.939972400665283,
      "learning_rate": 7.735849056603775e-06,
      "loss": 0.6535,
      "num_input_tokens_seen": 1697616,
      "step": 2625
    },
    {
      "epoch": 1.5507075471698113,
      "grad_norm": 2.2205686569213867,
      "learning_rate": 7.75058962264151e-06,
      "loss": 0.5965,
      "num_input_tokens_seen": 1700432,
      "step": 2630
    },
    {
      "epoch": 1.5536556603773586,
      "grad_norm": 2.2982678413391113,
      "learning_rate": 7.765330188679246e-06,
      "loss": 0.6191,
      "num_input_tokens_seen": 1704144,
      "step": 2635
    },
    {
      "epoch": 1.5566037735849056,
      "grad_norm": 2.514500617980957,
      "learning_rate": 7.780070754716981e-06,
      "loss": 0.5422,
      "num_input_tokens_seen": 1707440,
      "step": 2640
    },
    {
      "epoch": 1.5595518867924527,
      "grad_norm": 2.701277732849121,
      "learning_rate": 7.794811320754716e-06,
      "loss": 0.5785,
      "num_input_tokens_seen": 1710736,
      "step": 2645
    },
    {
      "epoch": 1.5625,
      "grad_norm": 2.746455669403076,
      "learning_rate": 7.809551886792453e-06,
      "loss": 0.551,
      "num_input_tokens_seen": 1713520,
      "step": 2650
    },
    {
      "epoch": 1.5654481132075473,
      "grad_norm": 3.969853401184082,
      "learning_rate": 7.824292452830189e-06,
      "loss": 0.6729,
      "num_input_tokens_seen": 1716368,
      "step": 2655
    },
    {
      "epoch": 1.5683962264150944,
      "grad_norm": 1.5365447998046875,
      "learning_rate": 7.839033018867926e-06,
      "loss": 0.5188,
      "num_input_tokens_seen": 1719536,
      "step": 2660
    },
    {
      "epoch": 1.5713443396226414,
      "grad_norm": 2.6098976135253906,
      "learning_rate": 7.853773584905661e-06,
      "loss": 0.52,
      "num_input_tokens_seen": 1722704,
      "step": 2665
    },
    {
      "epoch": 1.5742924528301887,
      "grad_norm": 1.8020899295806885,
      "learning_rate": 7.868514150943397e-06,
      "loss": 0.7307,
      "num_input_tokens_seen": 1725744,
      "step": 2670
    },
    {
      "epoch": 1.577240566037736,
      "grad_norm": 1.9730396270751953,
      "learning_rate": 7.883254716981132e-06,
      "loss": 0.6149,
      "num_input_tokens_seen": 1728784,
      "step": 2675
    },
    {
      "epoch": 1.580188679245283,
      "grad_norm": 3.0413737297058105,
      "learning_rate": 7.897995283018869e-06,
      "loss": 0.4312,
      "num_input_tokens_seen": 1732272,
      "step": 2680
    },
    {
      "epoch": 1.5831367924528301,
      "grad_norm": 2.783889055252075,
      "learning_rate": 7.912735849056604e-06,
      "loss": 0.4476,
      "num_input_tokens_seen": 1735184,
      "step": 2685
    },
    {
      "epoch": 1.5860849056603774,
      "grad_norm": 3.5608842372894287,
      "learning_rate": 7.927476415094341e-06,
      "loss": 0.7685,
      "num_input_tokens_seen": 1739184,
      "step": 2690
    },
    {
      "epoch": 1.5890330188679245,
      "grad_norm": 2.287456750869751,
      "learning_rate": 7.942216981132077e-06,
      "loss": 0.4656,
      "num_input_tokens_seen": 1741488,
      "step": 2695
    },
    {
      "epoch": 1.5919811320754715,
      "grad_norm": 2.2944023609161377,
      "learning_rate": 7.956957547169812e-06,
      "loss": 0.6838,
      "num_input_tokens_seen": 1744208,
      "step": 2700
    },
    {
      "epoch": 1.5949292452830188,
      "grad_norm": 3.7054927349090576,
      "learning_rate": 7.971698113207547e-06,
      "loss": 0.7089,
      "num_input_tokens_seen": 1746992,
      "step": 2705
    },
    {
      "epoch": 1.5978773584905661,
      "grad_norm": 2.686279296875,
      "learning_rate": 7.986438679245284e-06,
      "loss": 0.5308,
      "num_input_tokens_seen": 1750800,
      "step": 2710
    },
    {
      "epoch": 1.6008254716981132,
      "grad_norm": 5.939743518829346,
      "learning_rate": 8.00117924528302e-06,
      "loss": 0.4863,
      "num_input_tokens_seen": 1753360,
      "step": 2715
    },
    {
      "epoch": 1.6037735849056602,
      "grad_norm": 4.226158142089844,
      "learning_rate": 8.015919811320757e-06,
      "loss": 0.4906,
      "num_input_tokens_seen": 1755920,
      "step": 2720
    },
    {
      "epoch": 1.6067216981132075,
      "grad_norm": 2.548553705215454,
      "learning_rate": 8.030660377358492e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 1758928,
      "step": 2725
    },
    {
      "epoch": 1.6096698113207548,
      "grad_norm": 2.39188814163208,
      "learning_rate": 8.045400943396227e-06,
      "loss": 0.4886,
      "num_input_tokens_seen": 1762256,
      "step": 2730
    },
    {
      "epoch": 1.6126179245283019,
      "grad_norm": 1.7440837621688843,
      "learning_rate": 8.060141509433963e-06,
      "loss": 0.4286,
      "num_input_tokens_seen": 1765680,
      "step": 2735
    },
    {
      "epoch": 1.615566037735849,
      "grad_norm": 1.3706635236740112,
      "learning_rate": 8.074882075471698e-06,
      "loss": 0.6001,
      "num_input_tokens_seen": 1770032,
      "step": 2740
    },
    {
      "epoch": 1.6185141509433962,
      "grad_norm": 2.354753255844116,
      "learning_rate": 8.089622641509435e-06,
      "loss": 0.5991,
      "num_input_tokens_seen": 1773040,
      "step": 2745
    },
    {
      "epoch": 1.6214622641509435,
      "grad_norm": 2.7358005046844482,
      "learning_rate": 8.10436320754717e-06,
      "loss": 0.4854,
      "num_input_tokens_seen": 1776368,
      "step": 2750
    },
    {
      "epoch": 1.6244103773584906,
      "grad_norm": 3.2367818355560303,
      "learning_rate": 8.119103773584906e-06,
      "loss": 0.5802,
      "num_input_tokens_seen": 1779792,
      "step": 2755
    },
    {
      "epoch": 1.6273584905660377,
      "grad_norm": 2.6300456523895264,
      "learning_rate": 8.133844339622641e-06,
      "loss": 0.6235,
      "num_input_tokens_seen": 1782832,
      "step": 2760
    },
    {
      "epoch": 1.630306603773585,
      "grad_norm": 2.775857448577881,
      "learning_rate": 8.148584905660378e-06,
      "loss": 0.5864,
      "num_input_tokens_seen": 1785648,
      "step": 2765
    },
    {
      "epoch": 1.6332547169811322,
      "grad_norm": 5.58476448059082,
      "learning_rate": 8.163325471698114e-06,
      "loss": 0.6352,
      "num_input_tokens_seen": 1789040,
      "step": 2770
    },
    {
      "epoch": 1.6362028301886793,
      "grad_norm": 4.446688652038574,
      "learning_rate": 8.17806603773585e-06,
      "loss": 0.5508,
      "num_input_tokens_seen": 1792464,
      "step": 2775
    },
    {
      "epoch": 1.6391509433962264,
      "grad_norm": 2.235281229019165,
      "learning_rate": 8.192806603773586e-06,
      "loss": 0.516,
      "num_input_tokens_seen": 1798384,
      "step": 2780
    },
    {
      "epoch": 1.6420990566037736,
      "grad_norm": 4.668322563171387,
      "learning_rate": 8.207547169811321e-06,
      "loss": 0.5726,
      "num_input_tokens_seen": 1801904,
      "step": 2785
    },
    {
      "epoch": 1.6450471698113207,
      "grad_norm": 3.1163153648376465,
      "learning_rate": 8.222287735849057e-06,
      "loss": 0.4287,
      "num_input_tokens_seen": 1805424,
      "step": 2790
    },
    {
      "epoch": 1.6479952830188678,
      "grad_norm": 2.273127794265747,
      "learning_rate": 8.237028301886794e-06,
      "loss": 0.4436,
      "num_input_tokens_seen": 1808848,
      "step": 2795
    },
    {
      "epoch": 1.650943396226415,
      "grad_norm": 5.872840881347656,
      "learning_rate": 8.251768867924529e-06,
      "loss": 0.5815,
      "num_input_tokens_seen": 1812176,
      "step": 2800
    },
    {
      "epoch": 1.6538915094339623,
      "grad_norm": 3.0808072090148926,
      "learning_rate": 8.266509433962266e-06,
      "loss": 0.6075,
      "num_input_tokens_seen": 1815792,
      "step": 2805
    },
    {
      "epoch": 1.6568396226415094,
      "grad_norm": 2.117415428161621,
      "learning_rate": 8.281250000000001e-06,
      "loss": 0.6579,
      "num_input_tokens_seen": 1818800,
      "step": 2810
    },
    {
      "epoch": 1.6597877358490565,
      "grad_norm": 1.6309970617294312,
      "learning_rate": 8.295990566037737e-06,
      "loss": 0.5709,
      "num_input_tokens_seen": 1823248,
      "step": 2815
    },
    {
      "epoch": 1.6627358490566038,
      "grad_norm": 1.3913352489471436,
      "learning_rate": 8.310731132075472e-06,
      "loss": 0.4967,
      "num_input_tokens_seen": 1826160,
      "step": 2820
    },
    {
      "epoch": 1.665683962264151,
      "grad_norm": 2.879456043243408,
      "learning_rate": 8.325471698113207e-06,
      "loss": 0.5214,
      "num_input_tokens_seen": 1833712,
      "step": 2825
    },
    {
      "epoch": 1.6686320754716981,
      "grad_norm": 1.6060512065887451,
      "learning_rate": 8.340212264150944e-06,
      "loss": 0.4471,
      "num_input_tokens_seen": 1836624,
      "step": 2830
    },
    {
      "epoch": 1.6715801886792452,
      "grad_norm": 1.9861119985580444,
      "learning_rate": 8.35495283018868e-06,
      "loss": 0.4239,
      "num_input_tokens_seen": 1839312,
      "step": 2835
    },
    {
      "epoch": 1.6745283018867925,
      "grad_norm": 3.6057701110839844,
      "learning_rate": 8.369693396226415e-06,
      "loss": 0.5561,
      "num_input_tokens_seen": 1842032,
      "step": 2840
    },
    {
      "epoch": 1.6774764150943398,
      "grad_norm": 1.9394910335540771,
      "learning_rate": 8.38443396226415e-06,
      "loss": 0.4991,
      "num_input_tokens_seen": 1845488,
      "step": 2845
    },
    {
      "epoch": 1.6804245283018868,
      "grad_norm": 4.218761920928955,
      "learning_rate": 8.399174528301888e-06,
      "loss": 0.507,
      "num_input_tokens_seen": 1847632,
      "step": 2850
    },
    {
      "epoch": 1.6833726415094339,
      "grad_norm": 4.475637435913086,
      "learning_rate": 8.413915094339623e-06,
      "loss": 0.4864,
      "num_input_tokens_seen": 1850864,
      "step": 2855
    },
    {
      "epoch": 1.6863207547169812,
      "grad_norm": 3.6459853649139404,
      "learning_rate": 8.42865566037736e-06,
      "loss": 0.5504,
      "num_input_tokens_seen": 1853712,
      "step": 2860
    },
    {
      "epoch": 1.6892688679245285,
      "grad_norm": 2.4770374298095703,
      "learning_rate": 8.443396226415095e-06,
      "loss": 0.6989,
      "num_input_tokens_seen": 1857424,
      "step": 2865
    },
    {
      "epoch": 1.6922169811320755,
      "grad_norm": 4.621109485626221,
      "learning_rate": 8.45813679245283e-06,
      "loss": 0.5041,
      "num_input_tokens_seen": 1860144,
      "step": 2870
    },
    {
      "epoch": 1.6951650943396226,
      "grad_norm": 2.2250049114227295,
      "learning_rate": 8.472877358490566e-06,
      "loss": 0.5142,
      "num_input_tokens_seen": 1864208,
      "step": 2875
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 3.32865571975708,
      "learning_rate": 8.487617924528303e-06,
      "loss": 0.3848,
      "num_input_tokens_seen": 1867760,
      "step": 2880
    },
    {
      "epoch": 1.701061320754717,
      "grad_norm": 1.76797354221344,
      "learning_rate": 8.502358490566038e-06,
      "loss": 0.4738,
      "num_input_tokens_seen": 1870864,
      "step": 2885
    },
    {
      "epoch": 1.704009433962264,
      "grad_norm": 4.9690093994140625,
      "learning_rate": 8.517099056603775e-06,
      "loss": 0.5635,
      "num_input_tokens_seen": 1873680,
      "step": 2890
    },
    {
      "epoch": 1.7069575471698113,
      "grad_norm": 2.9682931900024414,
      "learning_rate": 8.53183962264151e-06,
      "loss": 0.5679,
      "num_input_tokens_seen": 1876912,
      "step": 2895
    },
    {
      "epoch": 1.7099056603773586,
      "grad_norm": 3.38326096534729,
      "learning_rate": 8.546580188679246e-06,
      "loss": 0.5031,
      "num_input_tokens_seen": 1879664,
      "step": 2900
    },
    {
      "epoch": 1.7128537735849056,
      "grad_norm": 2.794459104537964,
      "learning_rate": 8.561320754716981e-06,
      "loss": 0.4279,
      "num_input_tokens_seen": 1882000,
      "step": 2905
    },
    {
      "epoch": 1.7158018867924527,
      "grad_norm": 1.2686933279037476,
      "learning_rate": 8.576061320754717e-06,
      "loss": 0.5265,
      "num_input_tokens_seen": 1885552,
      "step": 2910
    },
    {
      "epoch": 1.71875,
      "grad_norm": 2.1500463485717773,
      "learning_rate": 8.590801886792454e-06,
      "loss": 0.3661,
      "num_input_tokens_seen": 1888496,
      "step": 2915
    },
    {
      "epoch": 1.7216981132075473,
      "grad_norm": 2.751096487045288,
      "learning_rate": 8.605542452830189e-06,
      "loss": 0.6477,
      "num_input_tokens_seen": 1892208,
      "step": 2920
    },
    {
      "epoch": 1.7246462264150944,
      "grad_norm": 2.753817319869995,
      "learning_rate": 8.620283018867926e-06,
      "loss": 0.4268,
      "num_input_tokens_seen": 1896304,
      "step": 2925
    },
    {
      "epoch": 1.7275943396226414,
      "grad_norm": 2.5147366523742676,
      "learning_rate": 8.635023584905662e-06,
      "loss": 0.4188,
      "num_input_tokens_seen": 1899728,
      "step": 2930
    },
    {
      "epoch": 1.7305424528301887,
      "grad_norm": 1.7920103073120117,
      "learning_rate": 8.649764150943397e-06,
      "loss": 0.4529,
      "num_input_tokens_seen": 1903376,
      "step": 2935
    },
    {
      "epoch": 1.733490566037736,
      "grad_norm": 2.2412922382354736,
      "learning_rate": 8.664504716981132e-06,
      "loss": 0.4898,
      "num_input_tokens_seen": 1906224,
      "step": 2940
    },
    {
      "epoch": 1.736438679245283,
      "grad_norm": 2.4478485584259033,
      "learning_rate": 8.67924528301887e-06,
      "loss": 0.587,
      "num_input_tokens_seen": 1909392,
      "step": 2945
    },
    {
      "epoch": 1.7393867924528301,
      "grad_norm": 2.370374917984009,
      "learning_rate": 8.693985849056605e-06,
      "loss": 0.5503,
      "num_input_tokens_seen": 1912272,
      "step": 2950
    },
    {
      "epoch": 1.7423349056603774,
      "grad_norm": 1.546656608581543,
      "learning_rate": 8.70872641509434e-06,
      "loss": 0.4254,
      "num_input_tokens_seen": 1915440,
      "step": 2955
    },
    {
      "epoch": 1.7452830188679245,
      "grad_norm": 4.168776988983154,
      "learning_rate": 8.723466981132075e-06,
      "loss": 0.5284,
      "num_input_tokens_seen": 1918608,
      "step": 2960
    },
    {
      "epoch": 1.7482311320754715,
      "grad_norm": 1.6780552864074707,
      "learning_rate": 8.738207547169812e-06,
      "loss": 0.4305,
      "num_input_tokens_seen": 1921712,
      "step": 2965
    },
    {
      "epoch": 1.7511792452830188,
      "grad_norm": 2.34070086479187,
      "learning_rate": 8.752948113207548e-06,
      "loss": 0.4144,
      "num_input_tokens_seen": 1924592,
      "step": 2970
    },
    {
      "epoch": 1.7541273584905661,
      "grad_norm": 2.8589274883270264,
      "learning_rate": 8.767688679245285e-06,
      "loss": 0.6317,
      "num_input_tokens_seen": 1928528,
      "step": 2975
    },
    {
      "epoch": 1.7570754716981132,
      "grad_norm": 2.163041830062866,
      "learning_rate": 8.78242924528302e-06,
      "loss": 0.5465,
      "num_input_tokens_seen": 1931984,
      "step": 2980
    },
    {
      "epoch": 1.7600235849056602,
      "grad_norm": 2.2423150539398193,
      "learning_rate": 8.797169811320755e-06,
      "loss": 0.3586,
      "num_input_tokens_seen": 1935376,
      "step": 2985
    },
    {
      "epoch": 1.7629716981132075,
      "grad_norm": 1.8588203191757202,
      "learning_rate": 8.81191037735849e-06,
      "loss": 0.3888,
      "num_input_tokens_seen": 1938896,
      "step": 2990
    },
    {
      "epoch": 1.7659198113207548,
      "grad_norm": 5.774896144866943,
      "learning_rate": 8.826650943396226e-06,
      "loss": 0.602,
      "num_input_tokens_seen": 1942000,
      "step": 2995
    },
    {
      "epoch": 1.7688679245283019,
      "grad_norm": 1.9162452220916748,
      "learning_rate": 8.841391509433963e-06,
      "loss": 0.4144,
      "num_input_tokens_seen": 1945488,
      "step": 3000
    },
    {
      "epoch": 1.771816037735849,
      "grad_norm": 1.8354791402816772,
      "learning_rate": 8.856132075471698e-06,
      "loss": 0.5617,
      "num_input_tokens_seen": 1948240,
      "step": 3005
    },
    {
      "epoch": 1.7747641509433962,
      "grad_norm": 3.4695446491241455,
      "learning_rate": 8.870872641509435e-06,
      "loss": 0.5936,
      "num_input_tokens_seen": 1951728,
      "step": 3010
    },
    {
      "epoch": 1.7777122641509435,
      "grad_norm": 1.1574922800064087,
      "learning_rate": 8.88561320754717e-06,
      "loss": 0.4449,
      "num_input_tokens_seen": 1954960,
      "step": 3015
    },
    {
      "epoch": 1.7806603773584906,
      "grad_norm": 5.160063743591309,
      "learning_rate": 8.900353773584906e-06,
      "loss": 0.5818,
      "num_input_tokens_seen": 1957104,
      "step": 3020
    },
    {
      "epoch": 1.7836084905660377,
      "grad_norm": 2.358757257461548,
      "learning_rate": 8.915094339622642e-06,
      "loss": 0.6199,
      "num_input_tokens_seen": 1960336,
      "step": 3025
    },
    {
      "epoch": 1.786556603773585,
      "grad_norm": 2.360243082046509,
      "learning_rate": 8.929834905660379e-06,
      "loss": 0.6032,
      "num_input_tokens_seen": 1963248,
      "step": 3030
    },
    {
      "epoch": 1.7895047169811322,
      "grad_norm": 5.553832054138184,
      "learning_rate": 8.944575471698114e-06,
      "loss": 0.4761,
      "num_input_tokens_seen": 1966096,
      "step": 3035
    },
    {
      "epoch": 1.7924528301886793,
      "grad_norm": 2.3416900634765625,
      "learning_rate": 8.959316037735851e-06,
      "loss": 0.501,
      "num_input_tokens_seen": 1969392,
      "step": 3040
    },
    {
      "epoch": 1.7954009433962264,
      "grad_norm": 3.0691497325897217,
      "learning_rate": 8.974056603773586e-06,
      "loss": 0.5202,
      "num_input_tokens_seen": 1972464,
      "step": 3045
    },
    {
      "epoch": 1.7983490566037736,
      "grad_norm": 2.455448865890503,
      "learning_rate": 8.988797169811322e-06,
      "loss": 0.3911,
      "num_input_tokens_seen": 1975344,
      "step": 3050
    },
    {
      "epoch": 1.8012971698113207,
      "grad_norm": 2.4416370391845703,
      "learning_rate": 9.003537735849057e-06,
      "loss": 0.4793,
      "num_input_tokens_seen": 1978576,
      "step": 3055
    },
    {
      "epoch": 1.8042452830188678,
      "grad_norm": 2.080230474472046,
      "learning_rate": 9.018278301886794e-06,
      "loss": 0.5078,
      "num_input_tokens_seen": 1981360,
      "step": 3060
    },
    {
      "epoch": 1.807193396226415,
      "grad_norm": 4.55208683013916,
      "learning_rate": 9.03301886792453e-06,
      "loss": 0.3658,
      "num_input_tokens_seen": 1983760,
      "step": 3065
    },
    {
      "epoch": 1.8101415094339623,
      "grad_norm": 3.9553723335266113,
      "learning_rate": 9.047759433962265e-06,
      "loss": 0.7416,
      "num_input_tokens_seen": 1986256,
      "step": 3070
    },
    {
      "epoch": 1.8130896226415094,
      "grad_norm": 2.119357109069824,
      "learning_rate": 9.0625e-06,
      "loss": 0.5441,
      "num_input_tokens_seen": 1989904,
      "step": 3075
    },
    {
      "epoch": 1.8160377358490565,
      "grad_norm": 2.308850049972534,
      "learning_rate": 9.077240566037735e-06,
      "loss": 0.619,
      "num_input_tokens_seen": 1993936,
      "step": 3080
    },
    {
      "epoch": 1.8189858490566038,
      "grad_norm": 3.171581745147705,
      "learning_rate": 9.091981132075472e-06,
      "loss": 0.4596,
      "num_input_tokens_seen": 1996432,
      "step": 3085
    },
    {
      "epoch": 1.821933962264151,
      "grad_norm": 1.6337982416152954,
      "learning_rate": 9.106721698113208e-06,
      "loss": 0.4186,
      "num_input_tokens_seen": 1999280,
      "step": 3090
    },
    {
      "epoch": 1.8248820754716981,
      "grad_norm": 2.101667642593384,
      "learning_rate": 9.121462264150945e-06,
      "loss": 0.4818,
      "num_input_tokens_seen": 2002832,
      "step": 3095
    },
    {
      "epoch": 1.8278301886792452,
      "grad_norm": 2.055734157562256,
      "learning_rate": 9.13620283018868e-06,
      "loss": 0.5643,
      "num_input_tokens_seen": 2005712,
      "step": 3100
    },
    {
      "epoch": 1.8307783018867925,
      "grad_norm": 3.1485414505004883,
      "learning_rate": 9.150943396226416e-06,
      "loss": 0.5532,
      "num_input_tokens_seen": 2008624,
      "step": 3105
    },
    {
      "epoch": 1.8337264150943398,
      "grad_norm": 4.019582748413086,
      "learning_rate": 9.165683962264151e-06,
      "loss": 0.5208,
      "num_input_tokens_seen": 2011600,
      "step": 3110
    },
    {
      "epoch": 1.8366745283018868,
      "grad_norm": 2.672015905380249,
      "learning_rate": 9.180424528301888e-06,
      "loss": 0.4749,
      "num_input_tokens_seen": 2014384,
      "step": 3115
    },
    {
      "epoch": 1.8396226415094339,
      "grad_norm": 1.9083237648010254,
      "learning_rate": 9.195165094339623e-06,
      "loss": 0.5216,
      "num_input_tokens_seen": 2017200,
      "step": 3120
    },
    {
      "epoch": 1.8425707547169812,
      "grad_norm": 1.8153343200683594,
      "learning_rate": 9.20990566037736e-06,
      "loss": 0.7735,
      "num_input_tokens_seen": 2020464,
      "step": 3125
    },
    {
      "epoch": 1.8455188679245285,
      "grad_norm": 1.7426409721374512,
      "learning_rate": 9.224646226415096e-06,
      "loss": 0.4905,
      "num_input_tokens_seen": 2023440,
      "step": 3130
    },
    {
      "epoch": 1.8484669811320755,
      "grad_norm": 2.128248453140259,
      "learning_rate": 9.239386792452831e-06,
      "loss": 0.4478,
      "num_input_tokens_seen": 2026128,
      "step": 3135
    },
    {
      "epoch": 1.8514150943396226,
      "grad_norm": 2.089012622833252,
      "learning_rate": 9.254127358490566e-06,
      "loss": 0.5238,
      "num_input_tokens_seen": 2029712,
      "step": 3140
    },
    {
      "epoch": 1.8543632075471699,
      "grad_norm": 2.7038159370422363,
      "learning_rate": 9.268867924528303e-06,
      "loss": 0.5221,
      "num_input_tokens_seen": 2033136,
      "step": 3145
    },
    {
      "epoch": 1.857311320754717,
      "grad_norm": 2.5019233226776123,
      "learning_rate": 9.283608490566039e-06,
      "loss": 0.4254,
      "num_input_tokens_seen": 2035760,
      "step": 3150
    },
    {
      "epoch": 1.860259433962264,
      "grad_norm": 3.9849936962127686,
      "learning_rate": 9.298349056603774e-06,
      "loss": 0.6782,
      "num_input_tokens_seen": 2038352,
      "step": 3155
    },
    {
      "epoch": 1.8632075471698113,
      "grad_norm": 2.1318023204803467,
      "learning_rate": 9.31308962264151e-06,
      "loss": 0.387,
      "num_input_tokens_seen": 2042672,
      "step": 3160
    },
    {
      "epoch": 1.8661556603773586,
      "grad_norm": 3.1579315662384033,
      "learning_rate": 9.327830188679245e-06,
      "loss": 0.4952,
      "num_input_tokens_seen": 2045968,
      "step": 3165
    },
    {
      "epoch": 1.8691037735849056,
      "grad_norm": 2.8541390895843506,
      "learning_rate": 9.342570754716982e-06,
      "loss": 0.7022,
      "num_input_tokens_seen": 2049168,
      "step": 3170
    },
    {
      "epoch": 1.8720518867924527,
      "grad_norm": 1.9304672479629517,
      "learning_rate": 9.357311320754717e-06,
      "loss": 0.4869,
      "num_input_tokens_seen": 2053072,
      "step": 3175
    },
    {
      "epoch": 1.875,
      "grad_norm": 4.3858323097229,
      "learning_rate": 9.372051886792454e-06,
      "loss": 0.3994,
      "num_input_tokens_seen": 2056304,
      "step": 3180
    },
    {
      "epoch": 1.8779481132075473,
      "grad_norm": 3.155311107635498,
      "learning_rate": 9.38679245283019e-06,
      "loss": 0.4822,
      "num_input_tokens_seen": 2059536,
      "step": 3185
    },
    {
      "epoch": 1.8808962264150944,
      "grad_norm": 2.017629623413086,
      "learning_rate": 9.401533018867925e-06,
      "loss": 0.4975,
      "num_input_tokens_seen": 2064144,
      "step": 3190
    },
    {
      "epoch": 1.8838443396226414,
      "grad_norm": 2.525209903717041,
      "learning_rate": 9.41627358490566e-06,
      "loss": 0.5129,
      "num_input_tokens_seen": 2067408,
      "step": 3195
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 1.7578011751174927,
      "learning_rate": 9.431014150943397e-06,
      "loss": 0.4737,
      "num_input_tokens_seen": 2070704,
      "step": 3200
    },
    {
      "epoch": 1.889740566037736,
      "grad_norm": 3.9810099601745605,
      "learning_rate": 9.445754716981133e-06,
      "loss": 0.441,
      "num_input_tokens_seen": 2073616,
      "step": 3205
    },
    {
      "epoch": 1.892688679245283,
      "grad_norm": 2.2784247398376465,
      "learning_rate": 9.46049528301887e-06,
      "loss": 0.544,
      "num_input_tokens_seen": 2076400,
      "step": 3210
    },
    {
      "epoch": 1.8956367924528301,
      "grad_norm": 2.026459217071533,
      "learning_rate": 9.475235849056605e-06,
      "loss": 0.475,
      "num_input_tokens_seen": 2079696,
      "step": 3215
    },
    {
      "epoch": 1.8985849056603774,
      "grad_norm": 2.511600971221924,
      "learning_rate": 9.48997641509434e-06,
      "loss": 0.4464,
      "num_input_tokens_seen": 2082544,
      "step": 3220
    },
    {
      "epoch": 1.9015330188679245,
      "grad_norm": 1.5244249105453491,
      "learning_rate": 9.504716981132076e-06,
      "loss": 0.4779,
      "num_input_tokens_seen": 2086224,
      "step": 3225
    },
    {
      "epoch": 1.9044811320754715,
      "grad_norm": 2.415357828140259,
      "learning_rate": 9.519457547169813e-06,
      "loss": 0.5285,
      "num_input_tokens_seen": 2090576,
      "step": 3230
    },
    {
      "epoch": 1.9074292452830188,
      "grad_norm": 1.4155315160751343,
      "learning_rate": 9.534198113207548e-06,
      "loss": 0.5682,
      "num_input_tokens_seen": 2093872,
      "step": 3235
    },
    {
      "epoch": 1.9103773584905661,
      "grad_norm": 1.4652818441390991,
      "learning_rate": 9.548938679245285e-06,
      "loss": 0.6335,
      "num_input_tokens_seen": 2097008,
      "step": 3240
    },
    {
      "epoch": 1.9133254716981132,
      "grad_norm": 2.690863609313965,
      "learning_rate": 9.56367924528302e-06,
      "loss": 0.5357,
      "num_input_tokens_seen": 2100528,
      "step": 3245
    },
    {
      "epoch": 1.9162735849056602,
      "grad_norm": 2.950446844100952,
      "learning_rate": 9.578419811320756e-06,
      "loss": 0.4578,
      "num_input_tokens_seen": 2103408,
      "step": 3250
    },
    {
      "epoch": 1.9192216981132075,
      "grad_norm": 2.7474937438964844,
      "learning_rate": 9.593160377358491e-06,
      "loss": 0.3838,
      "num_input_tokens_seen": 2106192,
      "step": 3255
    },
    {
      "epoch": 1.9221698113207548,
      "grad_norm": 2.445533275604248,
      "learning_rate": 9.607900943396226e-06,
      "loss": 0.4452,
      "num_input_tokens_seen": 2108560,
      "step": 3260
    },
    {
      "epoch": 1.9251179245283019,
      "grad_norm": 2.1544995307922363,
      "learning_rate": 9.622641509433963e-06,
      "loss": 0.7153,
      "num_input_tokens_seen": 2112144,
      "step": 3265
    },
    {
      "epoch": 1.928066037735849,
      "grad_norm": 1.2092195749282837,
      "learning_rate": 9.637382075471699e-06,
      "loss": 0.5192,
      "num_input_tokens_seen": 2115984,
      "step": 3270
    },
    {
      "epoch": 1.9310141509433962,
      "grad_norm": 7.6603684425354,
      "learning_rate": 9.652122641509434e-06,
      "loss": 0.3719,
      "num_input_tokens_seen": 2118672,
      "step": 3275
    },
    {
      "epoch": 1.9339622641509435,
      "grad_norm": 3.199583053588867,
      "learning_rate": 9.66686320754717e-06,
      "loss": 0.5046,
      "num_input_tokens_seen": 2121424,
      "step": 3280
    },
    {
      "epoch": 1.9369103773584906,
      "grad_norm": 6.672671794891357,
      "learning_rate": 9.681603773584907e-06,
      "loss": 0.7532,
      "num_input_tokens_seen": 2124304,
      "step": 3285
    },
    {
      "epoch": 1.9398584905660377,
      "grad_norm": 3.298612117767334,
      "learning_rate": 9.696344339622642e-06,
      "loss": 0.4101,
      "num_input_tokens_seen": 2128176,
      "step": 3290
    },
    {
      "epoch": 1.942806603773585,
      "grad_norm": 1.818219542503357,
      "learning_rate": 9.711084905660379e-06,
      "loss": 0.4387,
      "num_input_tokens_seen": 2131376,
      "step": 3295
    },
    {
      "epoch": 1.9457547169811322,
      "grad_norm": 2.0271811485290527,
      "learning_rate": 9.725825471698114e-06,
      "loss": 0.4956,
      "num_input_tokens_seen": 2135792,
      "step": 3300
    },
    {
      "epoch": 1.9487028301886793,
      "grad_norm": 4.204527854919434,
      "learning_rate": 9.74056603773585e-06,
      "loss": 0.4782,
      "num_input_tokens_seen": 2139856,
      "step": 3305
    },
    {
      "epoch": 1.9516509433962264,
      "grad_norm": 3.2647504806518555,
      "learning_rate": 9.755306603773585e-06,
      "loss": 0.4646,
      "num_input_tokens_seen": 2143056,
      "step": 3310
    },
    {
      "epoch": 1.9545990566037736,
      "grad_norm": 2.4099462032318115,
      "learning_rate": 9.770047169811322e-06,
      "loss": 0.4821,
      "num_input_tokens_seen": 2147056,
      "step": 3315
    },
    {
      "epoch": 1.9575471698113207,
      "grad_norm": 2.584202527999878,
      "learning_rate": 9.784787735849057e-06,
      "loss": 0.4031,
      "num_input_tokens_seen": 2150352,
      "step": 3320
    },
    {
      "epoch": 1.9604952830188678,
      "grad_norm": 1.9154154062271118,
      "learning_rate": 9.799528301886794e-06,
      "loss": 0.5812,
      "num_input_tokens_seen": 2154480,
      "step": 3325
    },
    {
      "epoch": 1.963443396226415,
      "grad_norm": 3.4489235877990723,
      "learning_rate": 9.81426886792453e-06,
      "loss": 0.469,
      "num_input_tokens_seen": 2157008,
      "step": 3330
    },
    {
      "epoch": 1.9663915094339623,
      "grad_norm": 2.319289207458496,
      "learning_rate": 9.829009433962265e-06,
      "loss": 0.6576,
      "num_input_tokens_seen": 2159888,
      "step": 3335
    },
    {
      "epoch": 1.9693396226415094,
      "grad_norm": 2.3519744873046875,
      "learning_rate": 9.84375e-06,
      "loss": 0.6357,
      "num_input_tokens_seen": 2164176,
      "step": 3340
    },
    {
      "epoch": 1.9722877358490565,
      "grad_norm": 2.0769708156585693,
      "learning_rate": 9.858490566037736e-06,
      "loss": 0.4919,
      "num_input_tokens_seen": 2167792,
      "step": 3345
    },
    {
      "epoch": 1.9752358490566038,
      "grad_norm": 3.293250322341919,
      "learning_rate": 9.873231132075473e-06,
      "loss": 0.6322,
      "num_input_tokens_seen": 2171152,
      "step": 3350
    },
    {
      "epoch": 1.978183962264151,
      "grad_norm": 1.4895752668380737,
      "learning_rate": 9.887971698113208e-06,
      "loss": 0.4727,
      "num_input_tokens_seen": 2174384,
      "step": 3355
    },
    {
      "epoch": 1.9811320754716981,
      "grad_norm": 2.5311203002929688,
      "learning_rate": 9.902712264150945e-06,
      "loss": 0.6018,
      "num_input_tokens_seen": 2178128,
      "step": 3360
    },
    {
      "epoch": 1.9840801886792452,
      "grad_norm": 2.6364150047302246,
      "learning_rate": 9.917452830188679e-06,
      "loss": 0.4728,
      "num_input_tokens_seen": 2180912,
      "step": 3365
    },
    {
      "epoch": 1.9870283018867925,
      "grad_norm": 1.3706297874450684,
      "learning_rate": 9.932193396226416e-06,
      "loss": 0.4938,
      "num_input_tokens_seen": 2185200,
      "step": 3370
    },
    {
      "epoch": 1.9899764150943398,
      "grad_norm": 2.933133363723755,
      "learning_rate": 9.946933962264151e-06,
      "loss": 0.5574,
      "num_input_tokens_seen": 2188496,
      "step": 3375
    },
    {
      "epoch": 1.9929245283018868,
      "grad_norm": 2.091433048248291,
      "learning_rate": 9.961674528301888e-06,
      "loss": 0.4865,
      "num_input_tokens_seen": 2191728,
      "step": 3380
    },
    {
      "epoch": 1.9958726415094339,
      "grad_norm": 1.8813694715499878,
      "learning_rate": 9.976415094339624e-06,
      "loss": 0.7135,
      "num_input_tokens_seen": 2195056,
      "step": 3385
    },
    {
      "epoch": 1.9988207547169812,
      "grad_norm": 2.4984889030456543,
      "learning_rate": 9.991155660377359e-06,
      "loss": 0.6441,
      "num_input_tokens_seen": 2198256,
      "step": 3390
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5279628038406372,
      "eval_runtime": 19.4296,
      "eval_samples_per_second": 87.289,
      "eval_steps_per_second": 21.822,
      "num_input_tokens_seen": 2199016,
      "step": 3392
    },
    {
      "epoch": 2.0017688679245285,
      "grad_norm": 1.9459149837493896,
      "learning_rate": 9.999999894098275e-06,
      "loss": 0.8648,
      "num_input_tokens_seen": 2201448,
      "step": 3395
    },
    {
      "epoch": 2.0047169811320753,
      "grad_norm": 2.6821510791778564,
      "learning_rate": 9.99999870270391e-06,
      "loss": 0.5955,
      "num_input_tokens_seen": 2204648,
      "step": 3400
    },
    {
      "epoch": 2.0076650943396226,
      "grad_norm": 2.064089059829712,
      "learning_rate": 9.999996187538341e-06,
      "loss": 0.4368,
      "num_input_tokens_seen": 2207432,
      "step": 3405
    },
    {
      "epoch": 2.01061320754717,
      "grad_norm": 1.349547028541565,
      "learning_rate": 9.999992348602233e-06,
      "loss": 0.573,
      "num_input_tokens_seen": 2211368,
      "step": 3410
    },
    {
      "epoch": 2.013561320754717,
      "grad_norm": 1.8440810441970825,
      "learning_rate": 9.999987185896598e-06,
      "loss": 0.5475,
      "num_input_tokens_seen": 2214920,
      "step": 3415
    },
    {
      "epoch": 2.016509433962264,
      "grad_norm": 1.5056023597717285,
      "learning_rate": 9.99998069942281e-06,
      "loss": 0.5476,
      "num_input_tokens_seen": 2218408,
      "step": 3420
    },
    {
      "epoch": 2.0194575471698113,
      "grad_norm": 1.439417839050293,
      "learning_rate": 9.999972889182583e-06,
      "loss": 0.4938,
      "num_input_tokens_seen": 2221288,
      "step": 3425
    },
    {
      "epoch": 2.0224056603773586,
      "grad_norm": 3.4988367557525635,
      "learning_rate": 9.999963755177984e-06,
      "loss": 0.5379,
      "num_input_tokens_seen": 2224552,
      "step": 3430
    },
    {
      "epoch": 2.025353773584906,
      "grad_norm": 2.609703779220581,
      "learning_rate": 9.999953297411434e-06,
      "loss": 0.5322,
      "num_input_tokens_seen": 2227592,
      "step": 3435
    },
    {
      "epoch": 2.0283018867924527,
      "grad_norm": 1.456058382987976,
      "learning_rate": 9.999941515885699e-06,
      "loss": 0.5205,
      "num_input_tokens_seen": 2231176,
      "step": 3440
    },
    {
      "epoch": 2.03125,
      "grad_norm": 2.209665060043335,
      "learning_rate": 9.999928410603897e-06,
      "loss": 0.4926,
      "num_input_tokens_seen": 2233864,
      "step": 3445
    },
    {
      "epoch": 2.0341981132075473,
      "grad_norm": 2.570680856704712,
      "learning_rate": 9.999913981569502e-06,
      "loss": 0.6631,
      "num_input_tokens_seen": 2238088,
      "step": 3450
    },
    {
      "epoch": 2.037146226415094,
      "grad_norm": 2.0984439849853516,
      "learning_rate": 9.999898228786332e-06,
      "loss": 0.6138,
      "num_input_tokens_seen": 2240584,
      "step": 3455
    },
    {
      "epoch": 2.0400943396226414,
      "grad_norm": 1.5368170738220215,
      "learning_rate": 9.999881152258557e-06,
      "loss": 0.4688,
      "num_input_tokens_seen": 2243784,
      "step": 3460
    },
    {
      "epoch": 2.0430424528301887,
      "grad_norm": 2.0577118396759033,
      "learning_rate": 9.999862751990697e-06,
      "loss": 0.4889,
      "num_input_tokens_seen": 2246664,
      "step": 3465
    },
    {
      "epoch": 2.045990566037736,
      "grad_norm": 1.9750739336013794,
      "learning_rate": 9.999843027987628e-06,
      "loss": 0.4847,
      "num_input_tokens_seen": 2249480,
      "step": 3470
    },
    {
      "epoch": 2.048938679245283,
      "grad_norm": 2.2435483932495117,
      "learning_rate": 9.999821980254567e-06,
      "loss": 0.495,
      "num_input_tokens_seen": 2252744,
      "step": 3475
    },
    {
      "epoch": 2.05188679245283,
      "grad_norm": 1.469650149345398,
      "learning_rate": 9.99979960879709e-06,
      "loss": 0.5093,
      "num_input_tokens_seen": 2257544,
      "step": 3480
    },
    {
      "epoch": 2.0548349056603774,
      "grad_norm": 1.9048575162887573,
      "learning_rate": 9.99977591362112e-06,
      "loss": 0.4957,
      "num_input_tokens_seen": 2260488,
      "step": 3485
    },
    {
      "epoch": 2.0577830188679247,
      "grad_norm": 2.7564568519592285,
      "learning_rate": 9.999750894732927e-06,
      "loss": 0.5078,
      "num_input_tokens_seen": 2263592,
      "step": 3490
    },
    {
      "epoch": 2.0607311320754715,
      "grad_norm": 1.9651620388031006,
      "learning_rate": 9.999724552139136e-06,
      "loss": 0.5142,
      "num_input_tokens_seen": 2267848,
      "step": 3495
    },
    {
      "epoch": 2.063679245283019,
      "grad_norm": 1.9440065622329712,
      "learning_rate": 9.999696885846724e-06,
      "loss": 0.4588,
      "num_input_tokens_seen": 2271176,
      "step": 3500
    },
    {
      "epoch": 2.066627358490566,
      "grad_norm": 2.861860990524292,
      "learning_rate": 9.999667895863012e-06,
      "loss": 0.4286,
      "num_input_tokens_seen": 2274152,
      "step": 3505
    },
    {
      "epoch": 2.0695754716981134,
      "grad_norm": 3.934751033782959,
      "learning_rate": 9.99963758219568e-06,
      "loss": 0.4219,
      "num_input_tokens_seen": 2277192,
      "step": 3510
    },
    {
      "epoch": 2.0725235849056602,
      "grad_norm": 1.5243632793426514,
      "learning_rate": 9.999605944852749e-06,
      "loss": 0.4887,
      "num_input_tokens_seen": 2280328,
      "step": 3515
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 2.311307191848755,
      "learning_rate": 9.999572983842599e-06,
      "loss": 0.4703,
      "num_input_tokens_seen": 2283176,
      "step": 3520
    },
    {
      "epoch": 2.078419811320755,
      "grad_norm": 2.3871772289276123,
      "learning_rate": 9.999538699173951e-06,
      "loss": 0.5303,
      "num_input_tokens_seen": 2285896,
      "step": 3525
    },
    {
      "epoch": 2.081367924528302,
      "grad_norm": 2.8624587059020996,
      "learning_rate": 9.99950309085589e-06,
      "loss": 0.486,
      "num_input_tokens_seen": 2288456,
      "step": 3530
    },
    {
      "epoch": 2.084316037735849,
      "grad_norm": 2.711564540863037,
      "learning_rate": 9.999466158897835e-06,
      "loss": 0.555,
      "num_input_tokens_seen": 2291464,
      "step": 3535
    },
    {
      "epoch": 2.0872641509433962,
      "grad_norm": 2.40155029296875,
      "learning_rate": 9.999427903309569e-06,
      "loss": 0.535,
      "num_input_tokens_seen": 2295240,
      "step": 3540
    },
    {
      "epoch": 2.0902122641509435,
      "grad_norm": 2.0475780963897705,
      "learning_rate": 9.99938832410122e-06,
      "loss": 0.5354,
      "num_input_tokens_seen": 2297832,
      "step": 3545
    },
    {
      "epoch": 2.0931603773584904,
      "grad_norm": 4.172753810882568,
      "learning_rate": 9.999347421283267e-06,
      "loss": 0.4647,
      "num_input_tokens_seen": 2300968,
      "step": 3550
    },
    {
      "epoch": 2.0961084905660377,
      "grad_norm": 2.2608513832092285,
      "learning_rate": 9.999305194866538e-06,
      "loss": 0.4257,
      "num_input_tokens_seen": 2305448,
      "step": 3555
    },
    {
      "epoch": 2.099056603773585,
      "grad_norm": 2.219630241394043,
      "learning_rate": 9.99926164486221e-06,
      "loss": 0.5487,
      "num_input_tokens_seen": 2308232,
      "step": 3560
    },
    {
      "epoch": 2.1020047169811322,
      "grad_norm": 2.11130690574646,
      "learning_rate": 9.99921677128182e-06,
      "loss": 0.5799,
      "num_input_tokens_seen": 2311048,
      "step": 3565
    },
    {
      "epoch": 2.104952830188679,
      "grad_norm": 1.635263204574585,
      "learning_rate": 9.99917057413724e-06,
      "loss": 0.4405,
      "num_input_tokens_seen": 2313928,
      "step": 3570
    },
    {
      "epoch": 2.1079009433962264,
      "grad_norm": 2.077345371246338,
      "learning_rate": 9.99912305344071e-06,
      "loss": 0.4233,
      "num_input_tokens_seen": 2316616,
      "step": 3575
    },
    {
      "epoch": 2.1108490566037736,
      "grad_norm": 3.6981558799743652,
      "learning_rate": 9.999074209204803e-06,
      "loss": 0.6846,
      "num_input_tokens_seen": 2319944,
      "step": 3580
    },
    {
      "epoch": 2.113797169811321,
      "grad_norm": 1.8095779418945312,
      "learning_rate": 9.999024041442455e-06,
      "loss": 0.4746,
      "num_input_tokens_seen": 2323304,
      "step": 3585
    },
    {
      "epoch": 2.1167452830188678,
      "grad_norm": 1.711809754371643,
      "learning_rate": 9.998972550166948e-06,
      "loss": 0.5261,
      "num_input_tokens_seen": 2326344,
      "step": 3590
    },
    {
      "epoch": 2.119693396226415,
      "grad_norm": 2.5615408420562744,
      "learning_rate": 9.998919735391915e-06,
      "loss": 0.5102,
      "num_input_tokens_seen": 2329000,
      "step": 3595
    },
    {
      "epoch": 2.1226415094339623,
      "grad_norm": 1.4711406230926514,
      "learning_rate": 9.998865597131336e-06,
      "loss": 0.5236,
      "num_input_tokens_seen": 2332616,
      "step": 3600
    },
    {
      "epoch": 2.1255896226415096,
      "grad_norm": 1.6700644493103027,
      "learning_rate": 9.998810135399545e-06,
      "loss": 0.4671,
      "num_input_tokens_seen": 2335656,
      "step": 3605
    },
    {
      "epoch": 2.1285377358490565,
      "grad_norm": 2.259582996368408,
      "learning_rate": 9.99875335021123e-06,
      "loss": 0.4835,
      "num_input_tokens_seen": 2338472,
      "step": 3610
    },
    {
      "epoch": 2.1314858490566038,
      "grad_norm": 2.257763385772705,
      "learning_rate": 9.998695241581423e-06,
      "loss": 0.4836,
      "num_input_tokens_seen": 2342920,
      "step": 3615
    },
    {
      "epoch": 2.134433962264151,
      "grad_norm": 1.191511631011963,
      "learning_rate": 9.998635809525504e-06,
      "loss": 0.394,
      "num_input_tokens_seen": 2346280,
      "step": 3620
    },
    {
      "epoch": 2.137382075471698,
      "grad_norm": 2.248244524002075,
      "learning_rate": 9.998575054059212e-06,
      "loss": 0.4852,
      "num_input_tokens_seen": 2349992,
      "step": 3625
    },
    {
      "epoch": 2.140330188679245,
      "grad_norm": 2.108295202255249,
      "learning_rate": 9.998512975198633e-06,
      "loss": 0.5365,
      "num_input_tokens_seen": 2353352,
      "step": 3630
    },
    {
      "epoch": 2.1432783018867925,
      "grad_norm": 2.636305570602417,
      "learning_rate": 9.998449572960202e-06,
      "loss": 0.4973,
      "num_input_tokens_seen": 2356296,
      "step": 3635
    },
    {
      "epoch": 2.1462264150943398,
      "grad_norm": 1.4999414682388306,
      "learning_rate": 9.998384847360705e-06,
      "loss": 0.4126,
      "num_input_tokens_seen": 2360040,
      "step": 3640
    },
    {
      "epoch": 2.1491745283018866,
      "grad_norm": 2.142404794692993,
      "learning_rate": 9.998318798417276e-06,
      "loss": 0.4278,
      "num_input_tokens_seen": 2362760,
      "step": 3645
    },
    {
      "epoch": 2.152122641509434,
      "grad_norm": 3.1819965839385986,
      "learning_rate": 9.998251426147403e-06,
      "loss": 0.4322,
      "num_input_tokens_seen": 2365832,
      "step": 3650
    },
    {
      "epoch": 2.155070754716981,
      "grad_norm": 1.5676376819610596,
      "learning_rate": 9.998182730568927e-06,
      "loss": 0.4741,
      "num_input_tokens_seen": 2369576,
      "step": 3655
    },
    {
      "epoch": 2.1580188679245285,
      "grad_norm": 1.6961256265640259,
      "learning_rate": 9.998112711700028e-06,
      "loss": 0.4669,
      "num_input_tokens_seen": 2372072,
      "step": 3660
    },
    {
      "epoch": 2.1609669811320753,
      "grad_norm": 2.0038411617279053,
      "learning_rate": 9.99804136955925e-06,
      "loss": 0.6617,
      "num_input_tokens_seen": 2375304,
      "step": 3665
    },
    {
      "epoch": 2.1639150943396226,
      "grad_norm": 1.1959389448165894,
      "learning_rate": 9.99796870416548e-06,
      "loss": 0.522,
      "num_input_tokens_seen": 2379464,
      "step": 3670
    },
    {
      "epoch": 2.16686320754717,
      "grad_norm": 2.827949047088623,
      "learning_rate": 9.997894715537953e-06,
      "loss": 0.5468,
      "num_input_tokens_seen": 2382184,
      "step": 3675
    },
    {
      "epoch": 2.169811320754717,
      "grad_norm": 2.715874195098877,
      "learning_rate": 9.997819403696263e-06,
      "loss": 0.6294,
      "num_input_tokens_seen": 2385576,
      "step": 3680
    },
    {
      "epoch": 2.172759433962264,
      "grad_norm": 3.3291869163513184,
      "learning_rate": 9.997742768660345e-06,
      "loss": 0.6342,
      "num_input_tokens_seen": 2389448,
      "step": 3685
    },
    {
      "epoch": 2.1757075471698113,
      "grad_norm": 2.3228235244750977,
      "learning_rate": 9.99766481045049e-06,
      "loss": 0.4661,
      "num_input_tokens_seen": 2392392,
      "step": 3690
    },
    {
      "epoch": 2.1786556603773586,
      "grad_norm": 3.941999673843384,
      "learning_rate": 9.997585529087338e-06,
      "loss": 0.4986,
      "num_input_tokens_seen": 2395592,
      "step": 3695
    },
    {
      "epoch": 2.1816037735849054,
      "grad_norm": 1.8424291610717773,
      "learning_rate": 9.997504924591878e-06,
      "loss": 0.5166,
      "num_input_tokens_seen": 2399048,
      "step": 3700
    },
    {
      "epoch": 2.1845518867924527,
      "grad_norm": 1.5930850505828857,
      "learning_rate": 9.99742299698545e-06,
      "loss": 0.492,
      "num_input_tokens_seen": 2402248,
      "step": 3705
    },
    {
      "epoch": 2.1875,
      "grad_norm": 2.8926682472229004,
      "learning_rate": 9.997339746289749e-06,
      "loss": 0.5609,
      "num_input_tokens_seen": 2405128,
      "step": 3710
    },
    {
      "epoch": 2.1904481132075473,
      "grad_norm": 2.070101261138916,
      "learning_rate": 9.997255172526812e-06,
      "loss": 0.516,
      "num_input_tokens_seen": 2408424,
      "step": 3715
    },
    {
      "epoch": 2.1933962264150946,
      "grad_norm": 1.6226454973220825,
      "learning_rate": 9.99716927571903e-06,
      "loss": 0.554,
      "num_input_tokens_seen": 2412392,
      "step": 3720
    },
    {
      "epoch": 2.1963443396226414,
      "grad_norm": 1.412214756011963,
      "learning_rate": 9.997082055889147e-06,
      "loss": 0.6442,
      "num_input_tokens_seen": 2416328,
      "step": 3725
    },
    {
      "epoch": 2.1992924528301887,
      "grad_norm": 1.4222919940948486,
      "learning_rate": 9.996993513060252e-06,
      "loss": 0.3671,
      "num_input_tokens_seen": 2419304,
      "step": 3730
    },
    {
      "epoch": 2.202240566037736,
      "grad_norm": 1.3161742687225342,
      "learning_rate": 9.996903647255789e-06,
      "loss": 0.485,
      "num_input_tokens_seen": 2422344,
      "step": 3735
    },
    {
      "epoch": 2.205188679245283,
      "grad_norm": 1.8893210887908936,
      "learning_rate": 9.99681245849955e-06,
      "loss": 0.4206,
      "num_input_tokens_seen": 2425864,
      "step": 3740
    },
    {
      "epoch": 2.20813679245283,
      "grad_norm": 2.6138792037963867,
      "learning_rate": 9.996719946815679e-06,
      "loss": 0.655,
      "num_input_tokens_seen": 2429704,
      "step": 3745
    },
    {
      "epoch": 2.2110849056603774,
      "grad_norm": 2.2000458240509033,
      "learning_rate": 9.996626112228665e-06,
      "loss": 0.4695,
      "num_input_tokens_seen": 2433576,
      "step": 3750
    },
    {
      "epoch": 2.2140330188679247,
      "grad_norm": 1.525758147239685,
      "learning_rate": 9.996530954763355e-06,
      "loss": 0.519,
      "num_input_tokens_seen": 2437512,
      "step": 3755
    },
    {
      "epoch": 2.2169811320754715,
      "grad_norm": 2.091216564178467,
      "learning_rate": 9.99643447444494e-06,
      "loss": 0.4459,
      "num_input_tokens_seen": 2440712,
      "step": 3760
    },
    {
      "epoch": 2.219929245283019,
      "grad_norm": 2.564833402633667,
      "learning_rate": 9.996336671298965e-06,
      "loss": 0.4464,
      "num_input_tokens_seen": 2444424,
      "step": 3765
    },
    {
      "epoch": 2.222877358490566,
      "grad_norm": 2.4853029251098633,
      "learning_rate": 9.996237545351323e-06,
      "loss": 0.4619,
      "num_input_tokens_seen": 2447816,
      "step": 3770
    },
    {
      "epoch": 2.2258254716981134,
      "grad_norm": 1.6591962575912476,
      "learning_rate": 9.996137096628259e-06,
      "loss": 0.5607,
      "num_input_tokens_seen": 2451496,
      "step": 3775
    },
    {
      "epoch": 2.2287735849056602,
      "grad_norm": 1.8022717237472534,
      "learning_rate": 9.996035325156366e-06,
      "loss": 0.7482,
      "num_input_tokens_seen": 2454440,
      "step": 3780
    },
    {
      "epoch": 2.2317216981132075,
      "grad_norm": 2.1976938247680664,
      "learning_rate": 9.995932230962589e-06,
      "loss": 0.4976,
      "num_input_tokens_seen": 2457384,
      "step": 3785
    },
    {
      "epoch": 2.234669811320755,
      "grad_norm": 1.662193775177002,
      "learning_rate": 9.995827814074223e-06,
      "loss": 0.4101,
      "num_input_tokens_seen": 2461160,
      "step": 3790
    },
    {
      "epoch": 2.237617924528302,
      "grad_norm": 2.510040521621704,
      "learning_rate": 9.995722074518913e-06,
      "loss": 0.4334,
      "num_input_tokens_seen": 2464392,
      "step": 3795
    },
    {
      "epoch": 2.240566037735849,
      "grad_norm": 1.7488176822662354,
      "learning_rate": 9.99561501232465e-06,
      "loss": 0.4139,
      "num_input_tokens_seen": 2467560,
      "step": 3800
    },
    {
      "epoch": 2.2435141509433962,
      "grad_norm": 4.796572208404541,
      "learning_rate": 9.995506627519786e-06,
      "loss": 0.5709,
      "num_input_tokens_seen": 2470440,
      "step": 3805
    },
    {
      "epoch": 2.2464622641509435,
      "grad_norm": 2.0850396156311035,
      "learning_rate": 9.995396920133012e-06,
      "loss": 0.4935,
      "num_input_tokens_seen": 2473800,
      "step": 3810
    },
    {
      "epoch": 2.2494103773584904,
      "grad_norm": 1.6042383909225464,
      "learning_rate": 9.995285890193373e-06,
      "loss": 0.5516,
      "num_input_tokens_seen": 2477512,
      "step": 3815
    },
    {
      "epoch": 2.2523584905660377,
      "grad_norm": 6.291835308074951,
      "learning_rate": 9.995173537730267e-06,
      "loss": 0.487,
      "num_input_tokens_seen": 2480360,
      "step": 3820
    },
    {
      "epoch": 2.255306603773585,
      "grad_norm": 1.6865124702453613,
      "learning_rate": 9.99505986277344e-06,
      "loss": 0.4248,
      "num_input_tokens_seen": 2485160,
      "step": 3825
    },
    {
      "epoch": 2.2582547169811322,
      "grad_norm": 1.4869468212127686,
      "learning_rate": 9.994944865352986e-06,
      "loss": 0.4505,
      "num_input_tokens_seen": 2489096,
      "step": 3830
    },
    {
      "epoch": 2.261202830188679,
      "grad_norm": 1.4185194969177246,
      "learning_rate": 9.994828545499351e-06,
      "loss": 0.4142,
      "num_input_tokens_seen": 2492488,
      "step": 3835
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 2.127901792526245,
      "learning_rate": 9.994710903243334e-06,
      "loss": 0.5139,
      "num_input_tokens_seen": 2495496,
      "step": 3840
    },
    {
      "epoch": 2.2670990566037736,
      "grad_norm": 2.151571273803711,
      "learning_rate": 9.994591938616079e-06,
      "loss": 0.4007,
      "num_input_tokens_seen": 2498696,
      "step": 3845
    },
    {
      "epoch": 2.270047169811321,
      "grad_norm": 3.052119731903076,
      "learning_rate": 9.994471651649082e-06,
      "loss": 0.4862,
      "num_input_tokens_seen": 2501864,
      "step": 3850
    },
    {
      "epoch": 2.2729952830188678,
      "grad_norm": 1.174790859222412,
      "learning_rate": 9.99435004237419e-06,
      "loss": 0.5627,
      "num_input_tokens_seen": 2505320,
      "step": 3855
    },
    {
      "epoch": 2.275943396226415,
      "grad_norm": 1.9417650699615479,
      "learning_rate": 9.9942271108236e-06,
      "loss": 0.5009,
      "num_input_tokens_seen": 2509000,
      "step": 3860
    },
    {
      "epoch": 2.2788915094339623,
      "grad_norm": 2.1368889808654785,
      "learning_rate": 9.994102857029859e-06,
      "loss": 0.4754,
      "num_input_tokens_seen": 2512232,
      "step": 3865
    },
    {
      "epoch": 2.2818396226415096,
      "grad_norm": 1.9912519454956055,
      "learning_rate": 9.993977281025862e-06,
      "loss": 0.3918,
      "num_input_tokens_seen": 2515656,
      "step": 3870
    },
    {
      "epoch": 2.2847877358490565,
      "grad_norm": 2.3768060207366943,
      "learning_rate": 9.993850382844858e-06,
      "loss": 0.4676,
      "num_input_tokens_seen": 2519016,
      "step": 3875
    },
    {
      "epoch": 2.2877358490566038,
      "grad_norm": 1.8841452598571777,
      "learning_rate": 9.993722162520443e-06,
      "loss": 0.4687,
      "num_input_tokens_seen": 2522120,
      "step": 3880
    },
    {
      "epoch": 2.290683962264151,
      "grad_norm": 3.544928789138794,
      "learning_rate": 9.993592620086564e-06,
      "loss": 0.4946,
      "num_input_tokens_seen": 2525320,
      "step": 3885
    },
    {
      "epoch": 2.293632075471698,
      "grad_norm": 1.6380153894424438,
      "learning_rate": 9.993461755577518e-06,
      "loss": 0.4276,
      "num_input_tokens_seen": 2529000,
      "step": 3890
    },
    {
      "epoch": 2.296580188679245,
      "grad_norm": 2.5875229835510254,
      "learning_rate": 9.99332956902795e-06,
      "loss": 0.3813,
      "num_input_tokens_seen": 2531816,
      "step": 3895
    },
    {
      "epoch": 2.2995283018867925,
      "grad_norm": 2.4318575859069824,
      "learning_rate": 9.993196060472859e-06,
      "loss": 0.5726,
      "num_input_tokens_seen": 2535048,
      "step": 3900
    },
    {
      "epoch": 2.3024764150943398,
      "grad_norm": 1.7257623672485352,
      "learning_rate": 9.993061229947591e-06,
      "loss": 0.5083,
      "num_input_tokens_seen": 2538952,
      "step": 3905
    },
    {
      "epoch": 2.3054245283018866,
      "grad_norm": 2.1718592643737793,
      "learning_rate": 9.992925077487845e-06,
      "loss": 0.4939,
      "num_input_tokens_seen": 2541896,
      "step": 3910
    },
    {
      "epoch": 2.308372641509434,
      "grad_norm": 1.5677859783172607,
      "learning_rate": 9.992787603129666e-06,
      "loss": 0.3787,
      "num_input_tokens_seen": 2545128,
      "step": 3915
    },
    {
      "epoch": 2.311320754716981,
      "grad_norm": 2.1155221462249756,
      "learning_rate": 9.99264880690945e-06,
      "loss": 0.5485,
      "num_input_tokens_seen": 2548104,
      "step": 3920
    },
    {
      "epoch": 2.3142688679245285,
      "grad_norm": 1.7156829833984375,
      "learning_rate": 9.992508688863947e-06,
      "loss": 0.5236,
      "num_input_tokens_seen": 2551656,
      "step": 3925
    },
    {
      "epoch": 2.3172169811320753,
      "grad_norm": 2.048093318939209,
      "learning_rate": 9.99236724903025e-06,
      "loss": 0.4295,
      "num_input_tokens_seen": 2554568,
      "step": 3930
    },
    {
      "epoch": 2.3201650943396226,
      "grad_norm": 4.021775245666504,
      "learning_rate": 9.992224487445809e-06,
      "loss": 0.4579,
      "num_input_tokens_seen": 2557352,
      "step": 3935
    },
    {
      "epoch": 2.32311320754717,
      "grad_norm": 2.552809000015259,
      "learning_rate": 9.99208040414842e-06,
      "loss": 0.3956,
      "num_input_tokens_seen": 2560840,
      "step": 3940
    },
    {
      "epoch": 2.326061320754717,
      "grad_norm": 2.230499744415283,
      "learning_rate": 9.99193499917623e-06,
      "loss": 0.4911,
      "num_input_tokens_seen": 2563848,
      "step": 3945
    },
    {
      "epoch": 2.329009433962264,
      "grad_norm": 4.351468086242676,
      "learning_rate": 9.991788272567735e-06,
      "loss": 0.5443,
      "num_input_tokens_seen": 2566824,
      "step": 3950
    },
    {
      "epoch": 2.3319575471698113,
      "grad_norm": 2.6456329822540283,
      "learning_rate": 9.991640224361781e-06,
      "loss": 0.6015,
      "num_input_tokens_seen": 2570536,
      "step": 3955
    },
    {
      "epoch": 2.3349056603773586,
      "grad_norm": 2.1100380420684814,
      "learning_rate": 9.991490854597563e-06,
      "loss": 0.4838,
      "num_input_tokens_seen": 2574056,
      "step": 3960
    },
    {
      "epoch": 2.3378537735849054,
      "grad_norm": 1.8235100507736206,
      "learning_rate": 9.991340163314632e-06,
      "loss": 0.5011,
      "num_input_tokens_seen": 2577032,
      "step": 3965
    },
    {
      "epoch": 2.3408018867924527,
      "grad_norm": 2.9909300804138184,
      "learning_rate": 9.991188150552878e-06,
      "loss": 0.4493,
      "num_input_tokens_seen": 2579496,
      "step": 3970
    },
    {
      "epoch": 2.34375,
      "grad_norm": 1.8704965114593506,
      "learning_rate": 9.991034816352553e-06,
      "loss": 0.5441,
      "num_input_tokens_seen": 2582504,
      "step": 3975
    },
    {
      "epoch": 2.3466981132075473,
      "grad_norm": 2.164410352706909,
      "learning_rate": 9.990880160754248e-06,
      "loss": 0.3348,
      "num_input_tokens_seen": 2585672,
      "step": 3980
    },
    {
      "epoch": 2.3496462264150946,
      "grad_norm": 3.2801589965820312,
      "learning_rate": 9.990724183798914e-06,
      "loss": 0.4905,
      "num_input_tokens_seen": 2589256,
      "step": 3985
    },
    {
      "epoch": 2.3525943396226414,
      "grad_norm": 1.4063498973846436,
      "learning_rate": 9.990566885527841e-06,
      "loss": 0.5862,
      "num_input_tokens_seen": 2592808,
      "step": 3990
    },
    {
      "epoch": 2.3555424528301887,
      "grad_norm": 2.939521312713623,
      "learning_rate": 9.99040826598268e-06,
      "loss": 0.4608,
      "num_input_tokens_seen": 2595560,
      "step": 3995
    },
    {
      "epoch": 2.358490566037736,
      "grad_norm": 1.1937437057495117,
      "learning_rate": 9.990248325205418e-06,
      "loss": 0.5417,
      "num_input_tokens_seen": 2599336,
      "step": 4000
    },
    {
      "epoch": 2.361438679245283,
      "grad_norm": 1.6047275066375732,
      "learning_rate": 9.990087063238408e-06,
      "loss": 0.4528,
      "num_input_tokens_seen": 2602600,
      "step": 4005
    },
    {
      "epoch": 2.36438679245283,
      "grad_norm": 2.811349630355835,
      "learning_rate": 9.989924480124342e-06,
      "loss": 0.4306,
      "num_input_tokens_seen": 2604872,
      "step": 4010
    },
    {
      "epoch": 2.3673349056603774,
      "grad_norm": 3.646364688873291,
      "learning_rate": 9.989760575906264e-06,
      "loss": 0.6126,
      "num_input_tokens_seen": 2608008,
      "step": 4015
    },
    {
      "epoch": 2.3702830188679247,
      "grad_norm": 3.818903923034668,
      "learning_rate": 9.98959535062757e-06,
      "loss": 0.4856,
      "num_input_tokens_seen": 2610792,
      "step": 4020
    },
    {
      "epoch": 2.3732311320754715,
      "grad_norm": 1.2776048183441162,
      "learning_rate": 9.989428804332002e-06,
      "loss": 0.476,
      "num_input_tokens_seen": 2615848,
      "step": 4025
    },
    {
      "epoch": 2.376179245283019,
      "grad_norm": 3.175480842590332,
      "learning_rate": 9.989260937063656e-06,
      "loss": 0.6023,
      "num_input_tokens_seen": 2619144,
      "step": 4030
    },
    {
      "epoch": 2.379127358490566,
      "grad_norm": 5.5899882316589355,
      "learning_rate": 9.989091748866973e-06,
      "loss": 0.4991,
      "num_input_tokens_seen": 2621544,
      "step": 4035
    },
    {
      "epoch": 2.3820754716981134,
      "grad_norm": 1.2565748691558838,
      "learning_rate": 9.988921239786748e-06,
      "loss": 0.4419,
      "num_input_tokens_seen": 2626632,
      "step": 4040
    },
    {
      "epoch": 2.3850235849056602,
      "grad_norm": 2.4391162395477295,
      "learning_rate": 9.988749409868124e-06,
      "loss": 0.6003,
      "num_input_tokens_seen": 2629384,
      "step": 4045
    },
    {
      "epoch": 2.3879716981132075,
      "grad_norm": 1.3083816766738892,
      "learning_rate": 9.988576259156593e-06,
      "loss": 0.5472,
      "num_input_tokens_seen": 2632360,
      "step": 4050
    },
    {
      "epoch": 2.390919811320755,
      "grad_norm": 5.211008548736572,
      "learning_rate": 9.988401787697996e-06,
      "loss": 0.5304,
      "num_input_tokens_seen": 2634888,
      "step": 4055
    },
    {
      "epoch": 2.393867924528302,
      "grad_norm": 3.3532748222351074,
      "learning_rate": 9.98822599553853e-06,
      "loss": 0.5403,
      "num_input_tokens_seen": 2637576,
      "step": 4060
    },
    {
      "epoch": 2.396816037735849,
      "grad_norm": 2.183716058731079,
      "learning_rate": 9.988048882724732e-06,
      "loss": 0.5243,
      "num_input_tokens_seen": 2639944,
      "step": 4065
    },
    {
      "epoch": 2.3997641509433962,
      "grad_norm": 2.396538257598877,
      "learning_rate": 9.987870449303497e-06,
      "loss": 0.5451,
      "num_input_tokens_seen": 2642632,
      "step": 4070
    },
    {
      "epoch": 2.4027122641509435,
      "grad_norm": 1.6799817085266113,
      "learning_rate": 9.98769069532206e-06,
      "loss": 0.5105,
      "num_input_tokens_seen": 2645608,
      "step": 4075
    },
    {
      "epoch": 2.4056603773584904,
      "grad_norm": 1.8873518705368042,
      "learning_rate": 9.98750962082802e-06,
      "loss": 0.4891,
      "num_input_tokens_seen": 2648840,
      "step": 4080
    },
    {
      "epoch": 2.4086084905660377,
      "grad_norm": 1.5017181634902954,
      "learning_rate": 9.987327225869312e-06,
      "loss": 0.3471,
      "num_input_tokens_seen": 2651880,
      "step": 4085
    },
    {
      "epoch": 2.411556603773585,
      "grad_norm": 2.2586872577667236,
      "learning_rate": 9.987143510494225e-06,
      "loss": 0.6183,
      "num_input_tokens_seen": 2654728,
      "step": 4090
    },
    {
      "epoch": 2.4145047169811322,
      "grad_norm": 1.6855944395065308,
      "learning_rate": 9.9869584747514e-06,
      "loss": 0.4256,
      "num_input_tokens_seen": 2658248,
      "step": 4095
    },
    {
      "epoch": 2.417452830188679,
      "grad_norm": 1.6320115327835083,
      "learning_rate": 9.986772118689828e-06,
      "loss": 0.52,
      "num_input_tokens_seen": 2661352,
      "step": 4100
    },
    {
      "epoch": 2.4204009433962264,
      "grad_norm": 2.474821090698242,
      "learning_rate": 9.986584442358845e-06,
      "loss": 0.4299,
      "num_input_tokens_seen": 2663848,
      "step": 4105
    },
    {
      "epoch": 2.4233490566037736,
      "grad_norm": 1.0483802556991577,
      "learning_rate": 9.98639544580814e-06,
      "loss": 0.552,
      "num_input_tokens_seen": 2668200,
      "step": 4110
    },
    {
      "epoch": 2.426297169811321,
      "grad_norm": 1.590315341949463,
      "learning_rate": 9.98620512908775e-06,
      "loss": 0.6617,
      "num_input_tokens_seen": 2671400,
      "step": 4115
    },
    {
      "epoch": 2.4292452830188678,
      "grad_norm": 2.136136054992676,
      "learning_rate": 9.986013492248064e-06,
      "loss": 0.477,
      "num_input_tokens_seen": 2674216,
      "step": 4120
    },
    {
      "epoch": 2.432193396226415,
      "grad_norm": 2.109036684036255,
      "learning_rate": 9.985820535339817e-06,
      "loss": 0.5226,
      "num_input_tokens_seen": 2677608,
      "step": 4125
    },
    {
      "epoch": 2.4351415094339623,
      "grad_norm": 1.8890572786331177,
      "learning_rate": 9.985626258414093e-06,
      "loss": 0.4376,
      "num_input_tokens_seen": 2681480,
      "step": 4130
    },
    {
      "epoch": 2.4380896226415096,
      "grad_norm": 2.367567539215088,
      "learning_rate": 9.985430661522333e-06,
      "loss": 0.5329,
      "num_input_tokens_seen": 2684040,
      "step": 4135
    },
    {
      "epoch": 2.4410377358490565,
      "grad_norm": 1.465188980102539,
      "learning_rate": 9.98523374471632e-06,
      "loss": 0.4598,
      "num_input_tokens_seen": 2686920,
      "step": 4140
    },
    {
      "epoch": 2.4439858490566038,
      "grad_norm": 2.3793163299560547,
      "learning_rate": 9.985035508048186e-06,
      "loss": 0.4934,
      "num_input_tokens_seen": 2689416,
      "step": 4145
    },
    {
      "epoch": 2.446933962264151,
      "grad_norm": 1.7063579559326172,
      "learning_rate": 9.984835951570417e-06,
      "loss": 0.4133,
      "num_input_tokens_seen": 2692008,
      "step": 4150
    },
    {
      "epoch": 2.449882075471698,
      "grad_norm": 2.593193531036377,
      "learning_rate": 9.984635075335847e-06,
      "loss": 0.5169,
      "num_input_tokens_seen": 2695208,
      "step": 4155
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 2.0366175174713135,
      "learning_rate": 9.984432879397659e-06,
      "loss": 0.6697,
      "num_input_tokens_seen": 2698792,
      "step": 4160
    },
    {
      "epoch": 2.4557783018867925,
      "grad_norm": 2.029085159301758,
      "learning_rate": 9.984229363809383e-06,
      "loss": 0.3922,
      "num_input_tokens_seen": 2701672,
      "step": 4165
    },
    {
      "epoch": 2.4587264150943398,
      "grad_norm": 2.5646262168884277,
      "learning_rate": 9.984024528624904e-06,
      "loss": 0.5126,
      "num_input_tokens_seen": 2704104,
      "step": 4170
    },
    {
      "epoch": 2.4616745283018866,
      "grad_norm": 1.8263518810272217,
      "learning_rate": 9.98381837389845e-06,
      "loss": 0.4909,
      "num_input_tokens_seen": 2706888,
      "step": 4175
    },
    {
      "epoch": 2.464622641509434,
      "grad_norm": 1.3725404739379883,
      "learning_rate": 9.983610899684601e-06,
      "loss": 0.3983,
      "num_input_tokens_seen": 2709832,
      "step": 4180
    },
    {
      "epoch": 2.467570754716981,
      "grad_norm": 1.8950510025024414,
      "learning_rate": 9.983402106038292e-06,
      "loss": 0.5877,
      "num_input_tokens_seen": 2712744,
      "step": 4185
    },
    {
      "epoch": 2.4705188679245285,
      "grad_norm": 1.7748279571533203,
      "learning_rate": 9.983191993014793e-06,
      "loss": 0.5549,
      "num_input_tokens_seen": 2715880,
      "step": 4190
    },
    {
      "epoch": 2.4734669811320753,
      "grad_norm": 2.4430863857269287,
      "learning_rate": 9.982980560669742e-06,
      "loss": 0.4963,
      "num_input_tokens_seen": 2718728,
      "step": 4195
    },
    {
      "epoch": 2.4764150943396226,
      "grad_norm": 1.9876513481140137,
      "learning_rate": 9.98276780905911e-06,
      "loss": 0.493,
      "num_input_tokens_seen": 2721416,
      "step": 4200
    },
    {
      "epoch": 2.47936320754717,
      "grad_norm": 2.3236005306243896,
      "learning_rate": 9.982553738239225e-06,
      "loss": 0.5433,
      "num_input_tokens_seen": 2724616,
      "step": 4205
    },
    {
      "epoch": 2.482311320754717,
      "grad_norm": 1.5325548648834229,
      "learning_rate": 9.982338348266766e-06,
      "loss": 0.5065,
      "num_input_tokens_seen": 2728616,
      "step": 4210
    },
    {
      "epoch": 2.485259433962264,
      "grad_norm": 1.5223034620285034,
      "learning_rate": 9.982121639198756e-06,
      "loss": 0.4354,
      "num_input_tokens_seen": 2731304,
      "step": 4215
    },
    {
      "epoch": 2.4882075471698113,
      "grad_norm": 1.5591561794281006,
      "learning_rate": 9.98190361109257e-06,
      "loss": 0.4296,
      "num_input_tokens_seen": 2734056,
      "step": 4220
    },
    {
      "epoch": 2.4911556603773586,
      "grad_norm": 1.1476281881332397,
      "learning_rate": 9.981684264005934e-06,
      "loss": 0.4849,
      "num_input_tokens_seen": 2740264,
      "step": 4225
    },
    {
      "epoch": 2.4941037735849054,
      "grad_norm": 1.2741329669952393,
      "learning_rate": 9.981463597996917e-06,
      "loss": 0.5212,
      "num_input_tokens_seen": 2744904,
      "step": 4230
    },
    {
      "epoch": 2.4970518867924527,
      "grad_norm": 2.631089925765991,
      "learning_rate": 9.981241613123944e-06,
      "loss": 0.3931,
      "num_input_tokens_seen": 2747816,
      "step": 4235
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.291221022605896,
      "learning_rate": 9.981018309445785e-06,
      "loss": 0.6621,
      "num_input_tokens_seen": 2752008,
      "step": 4240
    },
    {
      "epoch": 2.5029481132075473,
      "grad_norm": 1.6372578144073486,
      "learning_rate": 9.980793687021564e-06,
      "loss": 0.4746,
      "num_input_tokens_seen": 2754888,
      "step": 4245
    },
    {
      "epoch": 2.5058962264150946,
      "grad_norm": 1.4942965507507324,
      "learning_rate": 9.980567745910746e-06,
      "loss": 0.5424,
      "num_input_tokens_seen": 2758120,
      "step": 4250
    },
    {
      "epoch": 2.5088443396226414,
      "grad_norm": 2.124634265899658,
      "learning_rate": 9.980340486173155e-06,
      "loss": 0.4718,
      "num_input_tokens_seen": 2762088,
      "step": 4255
    },
    {
      "epoch": 2.5117924528301887,
      "grad_norm": 1.8445073366165161,
      "learning_rate": 9.980111907868954e-06,
      "loss": 0.6956,
      "num_input_tokens_seen": 2765160,
      "step": 4260
    },
    {
      "epoch": 2.514740566037736,
      "grad_norm": 1.6645914316177368,
      "learning_rate": 9.979882011058662e-06,
      "loss": 0.6489,
      "num_input_tokens_seen": 2767528,
      "step": 4265
    },
    {
      "epoch": 2.517688679245283,
      "grad_norm": 1.5878193378448486,
      "learning_rate": 9.979650795803146e-06,
      "loss": 0.4934,
      "num_input_tokens_seen": 2770376,
      "step": 4270
    },
    {
      "epoch": 2.52063679245283,
      "grad_norm": 2.089817523956299,
      "learning_rate": 9.979418262163621e-06,
      "loss": 0.4161,
      "num_input_tokens_seen": 2773288,
      "step": 4275
    },
    {
      "epoch": 2.5235849056603774,
      "grad_norm": 1.8481394052505493,
      "learning_rate": 9.979184410201652e-06,
      "loss": 0.6219,
      "num_input_tokens_seen": 2777032,
      "step": 4280
    },
    {
      "epoch": 2.5265330188679247,
      "grad_norm": 2.6019442081451416,
      "learning_rate": 9.97894923997915e-06,
      "loss": 0.4961,
      "num_input_tokens_seen": 2780264,
      "step": 4285
    },
    {
      "epoch": 2.5294811320754715,
      "grad_norm": 1.4647784233093262,
      "learning_rate": 9.97871275155838e-06,
      "loss": 0.4916,
      "num_input_tokens_seen": 2784296,
      "step": 4290
    },
    {
      "epoch": 2.532429245283019,
      "grad_norm": 2.039560556411743,
      "learning_rate": 9.978474945001949e-06,
      "loss": 0.5784,
      "num_input_tokens_seen": 2787208,
      "step": 4295
    },
    {
      "epoch": 2.535377358490566,
      "grad_norm": 3.0612335205078125,
      "learning_rate": 9.978235820372822e-06,
      "loss": 0.6709,
      "num_input_tokens_seen": 2790920,
      "step": 4300
    },
    {
      "epoch": 2.538325471698113,
      "grad_norm": 3.5970215797424316,
      "learning_rate": 9.977995377734307e-06,
      "loss": 0.4058,
      "num_input_tokens_seen": 2794024,
      "step": 4305
    },
    {
      "epoch": 2.5412735849056602,
      "grad_norm": 1.8760595321655273,
      "learning_rate": 9.977753617150061e-06,
      "loss": 0.408,
      "num_input_tokens_seen": 2797128,
      "step": 4310
    },
    {
      "epoch": 2.5442216981132075,
      "grad_norm": 1.4838526248931885,
      "learning_rate": 9.977510538684094e-06,
      "loss": 0.5278,
      "num_input_tokens_seen": 2800488,
      "step": 4315
    },
    {
      "epoch": 2.547169811320755,
      "grad_norm": 2.8355698585510254,
      "learning_rate": 9.977266142400757e-06,
      "loss": 0.494,
      "num_input_tokens_seen": 2804136,
      "step": 4320
    },
    {
      "epoch": 2.550117924528302,
      "grad_norm": 2.464789390563965,
      "learning_rate": 9.977020428364759e-06,
      "loss": 0.4716,
      "num_input_tokens_seen": 2807272,
      "step": 4325
    },
    {
      "epoch": 2.553066037735849,
      "grad_norm": 3.1387414932250977,
      "learning_rate": 9.976773396641154e-06,
      "loss": 0.7032,
      "num_input_tokens_seen": 2810280,
      "step": 4330
    },
    {
      "epoch": 2.5560141509433962,
      "grad_norm": 2.9196722507476807,
      "learning_rate": 9.976525047295342e-06,
      "loss": 0.4907,
      "num_input_tokens_seen": 2813032,
      "step": 4335
    },
    {
      "epoch": 2.5589622641509435,
      "grad_norm": 1.7306594848632812,
      "learning_rate": 9.976275380393077e-06,
      "loss": 0.5707,
      "num_input_tokens_seen": 2816328,
      "step": 4340
    },
    {
      "epoch": 2.5619103773584904,
      "grad_norm": 2.9307610988616943,
      "learning_rate": 9.976024396000459e-06,
      "loss": 0.6039,
      "num_input_tokens_seen": 2819112,
      "step": 4345
    },
    {
      "epoch": 2.5648584905660377,
      "grad_norm": 3.583786964416504,
      "learning_rate": 9.975772094183935e-06,
      "loss": 0.4497,
      "num_input_tokens_seen": 2822024,
      "step": 4350
    },
    {
      "epoch": 2.567806603773585,
      "grad_norm": 1.8196063041687012,
      "learning_rate": 9.975518475010306e-06,
      "loss": 0.531,
      "num_input_tokens_seen": 2825416,
      "step": 4355
    },
    {
      "epoch": 2.5707547169811322,
      "grad_norm": 1.8711715936660767,
      "learning_rate": 9.975263538546717e-06,
      "loss": 0.5175,
      "num_input_tokens_seen": 2828776,
      "step": 4360
    },
    {
      "epoch": 2.5737028301886795,
      "grad_norm": 1.4455981254577637,
      "learning_rate": 9.975007284860664e-06,
      "loss": 0.5199,
      "num_input_tokens_seen": 2831656,
      "step": 4365
    },
    {
      "epoch": 2.5766509433962264,
      "grad_norm": 4.318639278411865,
      "learning_rate": 9.974749714019993e-06,
      "loss": 0.4904,
      "num_input_tokens_seen": 2839112,
      "step": 4370
    },
    {
      "epoch": 2.5795990566037736,
      "grad_norm": 1.335287094116211,
      "learning_rate": 9.974490826092894e-06,
      "loss": 0.5114,
      "num_input_tokens_seen": 2842216,
      "step": 4375
    },
    {
      "epoch": 2.5825471698113205,
      "grad_norm": 2.080702304840088,
      "learning_rate": 9.974230621147907e-06,
      "loss": 0.3929,
      "num_input_tokens_seen": 2846120,
      "step": 4380
    },
    {
      "epoch": 2.5854952830188678,
      "grad_norm": 1.5224361419677734,
      "learning_rate": 9.973969099253928e-06,
      "loss": 0.5239,
      "num_input_tokens_seen": 2849064,
      "step": 4385
    },
    {
      "epoch": 2.588443396226415,
      "grad_norm": 2.187376022338867,
      "learning_rate": 9.973706260480194e-06,
      "loss": 0.411,
      "num_input_tokens_seen": 2852328,
      "step": 4390
    },
    {
      "epoch": 2.5913915094339623,
      "grad_norm": 1.840004324913025,
      "learning_rate": 9.97344210489629e-06,
      "loss": 0.4819,
      "num_input_tokens_seen": 2854568,
      "step": 4395
    },
    {
      "epoch": 2.5943396226415096,
      "grad_norm": 2.168484687805176,
      "learning_rate": 9.973176632572158e-06,
      "loss": 0.621,
      "num_input_tokens_seen": 2857480,
      "step": 4400
    },
    {
      "epoch": 2.5972877358490565,
      "grad_norm": 2.8307125568389893,
      "learning_rate": 9.972909843578076e-06,
      "loss": 0.5375,
      "num_input_tokens_seen": 2860904,
      "step": 4405
    },
    {
      "epoch": 2.6002358490566038,
      "grad_norm": 1.6940546035766602,
      "learning_rate": 9.972641737984681e-06,
      "loss": 0.4312,
      "num_input_tokens_seen": 2864264,
      "step": 4410
    },
    {
      "epoch": 2.603183962264151,
      "grad_norm": 1.5629909038543701,
      "learning_rate": 9.972372315862956e-06,
      "loss": 0.5625,
      "num_input_tokens_seen": 2867464,
      "step": 4415
    },
    {
      "epoch": 2.606132075471698,
      "grad_norm": 1.3941233158111572,
      "learning_rate": 9.97210157728423e-06,
      "loss": 0.4072,
      "num_input_tokens_seen": 2870728,
      "step": 4420
    },
    {
      "epoch": 2.609080188679245,
      "grad_norm": 1.1116681098937988,
      "learning_rate": 9.971829522320185e-06,
      "loss": 0.3643,
      "num_input_tokens_seen": 2873544,
      "step": 4425
    },
    {
      "epoch": 2.6120283018867925,
      "grad_norm": 2.121894121170044,
      "learning_rate": 9.971556151042843e-06,
      "loss": 0.4723,
      "num_input_tokens_seen": 2877192,
      "step": 4430
    },
    {
      "epoch": 2.6149764150943398,
      "grad_norm": 1.8785570859909058,
      "learning_rate": 9.971281463524588e-06,
      "loss": 0.4761,
      "num_input_tokens_seen": 2880072,
      "step": 4435
    },
    {
      "epoch": 2.617924528301887,
      "grad_norm": 4.999566078186035,
      "learning_rate": 9.971005459838136e-06,
      "loss": 0.6623,
      "num_input_tokens_seen": 2883368,
      "step": 4440
    },
    {
      "epoch": 2.620872641509434,
      "grad_norm": 1.0499082803726196,
      "learning_rate": 9.970728140056567e-06,
      "loss": 0.3762,
      "num_input_tokens_seen": 2886568,
      "step": 4445
    },
    {
      "epoch": 2.623820754716981,
      "grad_norm": 1.9464884996414185,
      "learning_rate": 9.9704495042533e-06,
      "loss": 0.4214,
      "num_input_tokens_seen": 2889448,
      "step": 4450
    },
    {
      "epoch": 2.6267688679245285,
      "grad_norm": 1.219513177871704,
      "learning_rate": 9.970169552502105e-06,
      "loss": 0.4576,
      "num_input_tokens_seen": 2892776,
      "step": 4455
    },
    {
      "epoch": 2.6297169811320753,
      "grad_norm": 3.2738211154937744,
      "learning_rate": 9.969888284877102e-06,
      "loss": 0.4151,
      "num_input_tokens_seen": 2895624,
      "step": 4460
    },
    {
      "epoch": 2.6326650943396226,
      "grad_norm": 4.9643707275390625,
      "learning_rate": 9.969605701452757e-06,
      "loss": 0.5927,
      "num_input_tokens_seen": 2898504,
      "step": 4465
    },
    {
      "epoch": 2.63561320754717,
      "grad_norm": 1.2316710948944092,
      "learning_rate": 9.969321802303882e-06,
      "loss": 0.4243,
      "num_input_tokens_seen": 2901096,
      "step": 4470
    },
    {
      "epoch": 2.638561320754717,
      "grad_norm": 1.1700246334075928,
      "learning_rate": 9.969036587505644e-06,
      "loss": 0.407,
      "num_input_tokens_seen": 2904200,
      "step": 4475
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 1.8527474403381348,
      "learning_rate": 9.968750057133555e-06,
      "loss": 0.4222,
      "num_input_tokens_seen": 2906792,
      "step": 4480
    },
    {
      "epoch": 2.6444575471698113,
      "grad_norm": 1.7738817930221558,
      "learning_rate": 9.968462211263474e-06,
      "loss": 0.6755,
      "num_input_tokens_seen": 2909928,
      "step": 4485
    },
    {
      "epoch": 2.6474056603773586,
      "grad_norm": 2.128181219100952,
      "learning_rate": 9.96817304997161e-06,
      "loss": 0.4376,
      "num_input_tokens_seen": 2912424,
      "step": 4490
    },
    {
      "epoch": 2.6503537735849054,
      "grad_norm": 2.218435287475586,
      "learning_rate": 9.967882573334519e-06,
      "loss": 0.5268,
      "num_input_tokens_seen": 2916232,
      "step": 4495
    },
    {
      "epoch": 2.6533018867924527,
      "grad_norm": 2.7213187217712402,
      "learning_rate": 9.967590781429106e-06,
      "loss": 0.5083,
      "num_input_tokens_seen": 2919784,
      "step": 4500
    },
    {
      "epoch": 2.65625,
      "grad_norm": 2.1042895317077637,
      "learning_rate": 9.967297674332625e-06,
      "loss": 0.6505,
      "num_input_tokens_seen": 2922344,
      "step": 4505
    },
    {
      "epoch": 2.6591981132075473,
      "grad_norm": 1.6468579769134521,
      "learning_rate": 9.967003252122675e-06,
      "loss": 0.3304,
      "num_input_tokens_seen": 2925544,
      "step": 4510
    },
    {
      "epoch": 2.6621462264150946,
      "grad_norm": 0.9303098917007446,
      "learning_rate": 9.96670751487721e-06,
      "loss": 0.4523,
      "num_input_tokens_seen": 2928936,
      "step": 4515
    },
    {
      "epoch": 2.6650943396226414,
      "grad_norm": 2.4703571796417236,
      "learning_rate": 9.966410462674525e-06,
      "loss": 0.4193,
      "num_input_tokens_seen": 2931272,
      "step": 4520
    },
    {
      "epoch": 2.6680424528301887,
      "grad_norm": 2.83332896232605,
      "learning_rate": 9.966112095593264e-06,
      "loss": 0.6164,
      "num_input_tokens_seen": 2934472,
      "step": 4525
    },
    {
      "epoch": 2.670990566037736,
      "grad_norm": 1.9801992177963257,
      "learning_rate": 9.965812413712425e-06,
      "loss": 0.5096,
      "num_input_tokens_seen": 2937320,
      "step": 4530
    },
    {
      "epoch": 2.673938679245283,
      "grad_norm": 1.8392788171768188,
      "learning_rate": 9.965511417111346e-06,
      "loss": 0.5661,
      "num_input_tokens_seen": 2939944,
      "step": 4535
    },
    {
      "epoch": 2.67688679245283,
      "grad_norm": 2.839603900909424,
      "learning_rate": 9.96520910586972e-06,
      "loss": 0.5798,
      "num_input_tokens_seen": 2942856,
      "step": 4540
    },
    {
      "epoch": 2.6798349056603774,
      "grad_norm": 2.552626371383667,
      "learning_rate": 9.964905480067585e-06,
      "loss": 0.4988,
      "num_input_tokens_seen": 2946088,
      "step": 4545
    },
    {
      "epoch": 2.6827830188679247,
      "grad_norm": 2.2877109050750732,
      "learning_rate": 9.964600539785328e-06,
      "loss": 0.558,
      "num_input_tokens_seen": 2949416,
      "step": 4550
    },
    {
      "epoch": 2.6857311320754715,
      "grad_norm": 2.274007558822632,
      "learning_rate": 9.96429428510368e-06,
      "loss": 0.5706,
      "num_input_tokens_seen": 2951848,
      "step": 4555
    },
    {
      "epoch": 2.688679245283019,
      "grad_norm": 2.6039319038391113,
      "learning_rate": 9.963986716103724e-06,
      "loss": 0.5849,
      "num_input_tokens_seen": 2955176,
      "step": 4560
    },
    {
      "epoch": 2.691627358490566,
      "grad_norm": 1.8845059871673584,
      "learning_rate": 9.963677832866893e-06,
      "loss": 0.4655,
      "num_input_tokens_seen": 2958184,
      "step": 4565
    },
    {
      "epoch": 2.694575471698113,
      "grad_norm": 3.9907684326171875,
      "learning_rate": 9.963367635474962e-06,
      "loss": 0.5336,
      "num_input_tokens_seen": 2960808,
      "step": 4570
    },
    {
      "epoch": 2.6975235849056602,
      "grad_norm": 1.0427062511444092,
      "learning_rate": 9.96305612401006e-06,
      "loss": 0.6029,
      "num_input_tokens_seen": 2963688,
      "step": 4575
    },
    {
      "epoch": 2.7004716981132075,
      "grad_norm": 3.7978429794311523,
      "learning_rate": 9.96274329855466e-06,
      "loss": 0.6143,
      "num_input_tokens_seen": 2967048,
      "step": 4580
    },
    {
      "epoch": 2.703419811320755,
      "grad_norm": 1.6718559265136719,
      "learning_rate": 9.962429159191583e-06,
      "loss": 0.5427,
      "num_input_tokens_seen": 2970344,
      "step": 4585
    },
    {
      "epoch": 2.706367924528302,
      "grad_norm": 4.659627437591553,
      "learning_rate": 9.962113706003997e-06,
      "loss": 0.4906,
      "num_input_tokens_seen": 2972904,
      "step": 4590
    },
    {
      "epoch": 2.709316037735849,
      "grad_norm": 2.2885563373565674,
      "learning_rate": 9.961796939075424e-06,
      "loss": 0.427,
      "num_input_tokens_seen": 2977448,
      "step": 4595
    },
    {
      "epoch": 2.7122641509433962,
      "grad_norm": 2.0075228214263916,
      "learning_rate": 9.961478858489728e-06,
      "loss": 0.5281,
      "num_input_tokens_seen": 2980424,
      "step": 4600
    },
    {
      "epoch": 2.7152122641509435,
      "grad_norm": 3.092205047607422,
      "learning_rate": 9.961159464331119e-06,
      "loss": 0.4089,
      "num_input_tokens_seen": 2984648,
      "step": 4605
    },
    {
      "epoch": 2.7181603773584904,
      "grad_norm": 2.331362724304199,
      "learning_rate": 9.960838756684161e-06,
      "loss": 0.5276,
      "num_input_tokens_seen": 2987464,
      "step": 4610
    },
    {
      "epoch": 2.7211084905660377,
      "grad_norm": 2.5184526443481445,
      "learning_rate": 9.960516735633764e-06,
      "loss": 0.5784,
      "num_input_tokens_seen": 2991784,
      "step": 4615
    },
    {
      "epoch": 2.724056603773585,
      "grad_norm": 1.3405195474624634,
      "learning_rate": 9.960193401265181e-06,
      "loss": 0.5552,
      "num_input_tokens_seen": 2994632,
      "step": 4620
    },
    {
      "epoch": 2.7270047169811322,
      "grad_norm": 6.233210563659668,
      "learning_rate": 9.959868753664018e-06,
      "loss": 0.4671,
      "num_input_tokens_seen": 2997576,
      "step": 4625
    },
    {
      "epoch": 2.7299528301886795,
      "grad_norm": 2.0221831798553467,
      "learning_rate": 9.959542792916227e-06,
      "loss": 0.4876,
      "num_input_tokens_seen": 3000072,
      "step": 4630
    },
    {
      "epoch": 2.7329009433962264,
      "grad_norm": 3.3484041690826416,
      "learning_rate": 9.959215519108108e-06,
      "loss": 0.4682,
      "num_input_tokens_seen": 3002728,
      "step": 4635
    },
    {
      "epoch": 2.7358490566037736,
      "grad_norm": 2.4793059825897217,
      "learning_rate": 9.958886932326306e-06,
      "loss": 0.4611,
      "num_input_tokens_seen": 3005960,
      "step": 4640
    },
    {
      "epoch": 2.7387971698113205,
      "grad_norm": 2.1480765342712402,
      "learning_rate": 9.958557032657817e-06,
      "loss": 0.4639,
      "num_input_tokens_seen": 3009768,
      "step": 4645
    },
    {
      "epoch": 2.7417452830188678,
      "grad_norm": 2.344350576400757,
      "learning_rate": 9.958225820189984e-06,
      "loss": 0.3783,
      "num_input_tokens_seen": 3012264,
      "step": 4650
    },
    {
      "epoch": 2.744693396226415,
      "grad_norm": 2.2079074382781982,
      "learning_rate": 9.957893295010495e-06,
      "loss": 0.5527,
      "num_input_tokens_seen": 3015016,
      "step": 4655
    },
    {
      "epoch": 2.7476415094339623,
      "grad_norm": 2.898836374282837,
      "learning_rate": 9.957559457207391e-06,
      "loss": 0.5313,
      "num_input_tokens_seen": 3017512,
      "step": 4660
    },
    {
      "epoch": 2.7505896226415096,
      "grad_norm": 1.1877919435501099,
      "learning_rate": 9.957224306869053e-06,
      "loss": 0.4974,
      "num_input_tokens_seen": 3020648,
      "step": 4665
    },
    {
      "epoch": 2.7535377358490565,
      "grad_norm": 1.4231140613555908,
      "learning_rate": 9.956887844084216e-06,
      "loss": 0.5161,
      "num_input_tokens_seen": 3023528,
      "step": 4670
    },
    {
      "epoch": 2.7564858490566038,
      "grad_norm": 2.597628593444824,
      "learning_rate": 9.956550068941958e-06,
      "loss": 0.4282,
      "num_input_tokens_seen": 3026632,
      "step": 4675
    },
    {
      "epoch": 2.759433962264151,
      "grad_norm": 3.8640692234039307,
      "learning_rate": 9.95621098153171e-06,
      "loss": 0.4683,
      "num_input_tokens_seen": 3029640,
      "step": 4680
    },
    {
      "epoch": 2.762382075471698,
      "grad_norm": 2.300905466079712,
      "learning_rate": 9.955870581943243e-06,
      "loss": 0.4374,
      "num_input_tokens_seen": 3033608,
      "step": 4685
    },
    {
      "epoch": 2.765330188679245,
      "grad_norm": 1.9141736030578613,
      "learning_rate": 9.955528870266681e-06,
      "loss": 0.468,
      "num_input_tokens_seen": 3037256,
      "step": 4690
    },
    {
      "epoch": 2.7682783018867925,
      "grad_norm": 1.636860966682434,
      "learning_rate": 9.955185846592495e-06,
      "loss": 0.39,
      "num_input_tokens_seen": 3040744,
      "step": 4695
    },
    {
      "epoch": 2.7712264150943398,
      "grad_norm": 1.6407124996185303,
      "learning_rate": 9.9548415110115e-06,
      "loss": 0.5782,
      "num_input_tokens_seen": 3044264,
      "step": 4700
    },
    {
      "epoch": 2.774174528301887,
      "grad_norm": 8.080162048339844,
      "learning_rate": 9.95449586361486e-06,
      "loss": 0.635,
      "num_input_tokens_seen": 3046728,
      "step": 4705
    },
    {
      "epoch": 2.777122641509434,
      "grad_norm": 6.949060440063477,
      "learning_rate": 9.954148904494085e-06,
      "loss": 0.5243,
      "num_input_tokens_seen": 3049448,
      "step": 4710
    },
    {
      "epoch": 2.780070754716981,
      "grad_norm": 6.240416049957275,
      "learning_rate": 9.95380063374104e-06,
      "loss": 0.4741,
      "num_input_tokens_seen": 3052040,
      "step": 4715
    },
    {
      "epoch": 2.7830188679245285,
      "grad_norm": 2.5601089000701904,
      "learning_rate": 9.953451051447927e-06,
      "loss": 0.4623,
      "num_input_tokens_seen": 3054920,
      "step": 4720
    },
    {
      "epoch": 2.7859669811320753,
      "grad_norm": 1.6536729335784912,
      "learning_rate": 9.953100157707299e-06,
      "loss": 0.4972,
      "num_input_tokens_seen": 3058056,
      "step": 4725
    },
    {
      "epoch": 2.7889150943396226,
      "grad_norm": 5.624151706695557,
      "learning_rate": 9.952747952612056e-06,
      "loss": 0.3922,
      "num_input_tokens_seen": 3060712,
      "step": 4730
    },
    {
      "epoch": 2.79186320754717,
      "grad_norm": 1.63435697555542,
      "learning_rate": 9.952394436255451e-06,
      "loss": 0.356,
      "num_input_tokens_seen": 3064232,
      "step": 4735
    },
    {
      "epoch": 2.794811320754717,
      "grad_norm": 1.378397822380066,
      "learning_rate": 9.952039608731072e-06,
      "loss": 0.3803,
      "num_input_tokens_seen": 3068136,
      "step": 4740
    },
    {
      "epoch": 2.797759433962264,
      "grad_norm": 1.5862846374511719,
      "learning_rate": 9.951683470132868e-06,
      "loss": 0.4493,
      "num_input_tokens_seen": 3071592,
      "step": 4745
    },
    {
      "epoch": 2.8007075471698113,
      "grad_norm": 1.2516865730285645,
      "learning_rate": 9.951326020555122e-06,
      "loss": 0.4291,
      "num_input_tokens_seen": 3075848,
      "step": 4750
    },
    {
      "epoch": 2.8036556603773586,
      "grad_norm": 4.0720438957214355,
      "learning_rate": 9.950967260092473e-06,
      "loss": 0.537,
      "num_input_tokens_seen": 3079592,
      "step": 4755
    },
    {
      "epoch": 2.8066037735849054,
      "grad_norm": 3.3032169342041016,
      "learning_rate": 9.950607188839905e-06,
      "loss": 0.5017,
      "num_input_tokens_seen": 3082760,
      "step": 4760
    },
    {
      "epoch": 2.8095518867924527,
      "grad_norm": 1.6673450469970703,
      "learning_rate": 9.950245806892749e-06,
      "loss": 0.5066,
      "num_input_tokens_seen": 3086248,
      "step": 4765
    },
    {
      "epoch": 2.8125,
      "grad_norm": 3.304537057876587,
      "learning_rate": 9.94988311434668e-06,
      "loss": 0.5738,
      "num_input_tokens_seen": 3090120,
      "step": 4770
    },
    {
      "epoch": 2.8154481132075473,
      "grad_norm": 3.7510125637054443,
      "learning_rate": 9.949519111297723e-06,
      "loss": 0.3983,
      "num_input_tokens_seen": 3092520,
      "step": 4775
    },
    {
      "epoch": 2.8183962264150946,
      "grad_norm": 1.461686134338379,
      "learning_rate": 9.949153797842252e-06,
      "loss": 0.5851,
      "num_input_tokens_seen": 3095752,
      "step": 4780
    },
    {
      "epoch": 2.8213443396226414,
      "grad_norm": 2.055211067199707,
      "learning_rate": 9.948787174076982e-06,
      "loss": 0.5535,
      "num_input_tokens_seen": 3099432,
      "step": 4785
    },
    {
      "epoch": 2.8242924528301887,
      "grad_norm": 2.4384799003601074,
      "learning_rate": 9.948419240098978e-06,
      "loss": 0.4986,
      "num_input_tokens_seen": 3102984,
      "step": 4790
    },
    {
      "epoch": 2.827240566037736,
      "grad_norm": 2.744708776473999,
      "learning_rate": 9.948049996005657e-06,
      "loss": 0.4133,
      "num_input_tokens_seen": 3105608,
      "step": 4795
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 2.0244076251983643,
      "learning_rate": 9.947679441894773e-06,
      "loss": 0.5952,
      "num_input_tokens_seen": 3108680,
      "step": 4800
    },
    {
      "epoch": 2.83313679245283,
      "grad_norm": 3.9313180446624756,
      "learning_rate": 9.947307577864433e-06,
      "loss": 0.5994,
      "num_input_tokens_seen": 3111272,
      "step": 4805
    },
    {
      "epoch": 2.8360849056603774,
      "grad_norm": 1.9512778520584106,
      "learning_rate": 9.94693440401309e-06,
      "loss": 0.3822,
      "num_input_tokens_seen": 3114088,
      "step": 4810
    },
    {
      "epoch": 2.8390330188679247,
      "grad_norm": 1.7983839511871338,
      "learning_rate": 9.946559920439545e-06,
      "loss": 0.5028,
      "num_input_tokens_seen": 3118088,
      "step": 4815
    },
    {
      "epoch": 2.8419811320754715,
      "grad_norm": 1.6929305791854858,
      "learning_rate": 9.946184127242942e-06,
      "loss": 0.4163,
      "num_input_tokens_seen": 3121032,
      "step": 4820
    },
    {
      "epoch": 2.844929245283019,
      "grad_norm": 2.0308361053466797,
      "learning_rate": 9.945807024522774e-06,
      "loss": 0.4552,
      "num_input_tokens_seen": 3124136,
      "step": 4825
    },
    {
      "epoch": 2.847877358490566,
      "grad_norm": 1.5933799743652344,
      "learning_rate": 9.945428612378881e-06,
      "loss": 0.4218,
      "num_input_tokens_seen": 3127272,
      "step": 4830
    },
    {
      "epoch": 2.850825471698113,
      "grad_norm": 1.700826644897461,
      "learning_rate": 9.94504889091145e-06,
      "loss": 0.5818,
      "num_input_tokens_seen": 3130344,
      "step": 4835
    },
    {
      "epoch": 2.8537735849056602,
      "grad_norm": 2.701007843017578,
      "learning_rate": 9.944667860221013e-06,
      "loss": 0.5648,
      "num_input_tokens_seen": 3137736,
      "step": 4840
    },
    {
      "epoch": 2.8567216981132075,
      "grad_norm": 1.6669749021530151,
      "learning_rate": 9.944285520408448e-06,
      "loss": 0.495,
      "num_input_tokens_seen": 3140552,
      "step": 4845
    },
    {
      "epoch": 2.859669811320755,
      "grad_norm": 2.916809558868408,
      "learning_rate": 9.943901871574984e-06,
      "loss": 0.4891,
      "num_input_tokens_seen": 3143080,
      "step": 4850
    },
    {
      "epoch": 2.862617924528302,
      "grad_norm": 1.7121535539627075,
      "learning_rate": 9.943516913822192e-06,
      "loss": 0.3545,
      "num_input_tokens_seen": 3146792,
      "step": 4855
    },
    {
      "epoch": 2.865566037735849,
      "grad_norm": 1.4963500499725342,
      "learning_rate": 9.943130647251994e-06,
      "loss": 0.3978,
      "num_input_tokens_seen": 3149960,
      "step": 4860
    },
    {
      "epoch": 2.8685141509433962,
      "grad_norm": 1.8962477445602417,
      "learning_rate": 9.94274307196665e-06,
      "loss": 0.4051,
      "num_input_tokens_seen": 3152776,
      "step": 4865
    },
    {
      "epoch": 2.8714622641509435,
      "grad_norm": 2.227431535720825,
      "learning_rate": 9.942354188068778e-06,
      "loss": 0.4424,
      "num_input_tokens_seen": 3155912,
      "step": 4870
    },
    {
      "epoch": 2.8744103773584904,
      "grad_norm": 2.128610610961914,
      "learning_rate": 9.941963995661333e-06,
      "loss": 0.4484,
      "num_input_tokens_seen": 3158952,
      "step": 4875
    },
    {
      "epoch": 2.8773584905660377,
      "grad_norm": 3.696868658065796,
      "learning_rate": 9.941572494847622e-06,
      "loss": 0.4713,
      "num_input_tokens_seen": 3161320,
      "step": 4880
    },
    {
      "epoch": 2.880306603773585,
      "grad_norm": 3.5287575721740723,
      "learning_rate": 9.941179685731297e-06,
      "loss": 0.54,
      "num_input_tokens_seen": 3163976,
      "step": 4885
    },
    {
      "epoch": 2.8832547169811322,
      "grad_norm": 3.711376428604126,
      "learning_rate": 9.940785568416354e-06,
      "loss": 0.548,
      "num_input_tokens_seen": 3166824,
      "step": 4890
    },
    {
      "epoch": 2.8862028301886795,
      "grad_norm": 1.6392271518707275,
      "learning_rate": 9.940390143007137e-06,
      "loss": 0.4028,
      "num_input_tokens_seen": 3170664,
      "step": 4895
    },
    {
      "epoch": 2.8891509433962264,
      "grad_norm": 1.8397538661956787,
      "learning_rate": 9.939993409608339e-06,
      "loss": 0.4181,
      "num_input_tokens_seen": 3174696,
      "step": 4900
    },
    {
      "epoch": 2.8920990566037736,
      "grad_norm": 1.8542345762252808,
      "learning_rate": 9.939595368324996e-06,
      "loss": 0.5419,
      "num_input_tokens_seen": 3177480,
      "step": 4905
    },
    {
      "epoch": 2.8950471698113205,
      "grad_norm": 2.0663814544677734,
      "learning_rate": 9.93919601926249e-06,
      "loss": 0.474,
      "num_input_tokens_seen": 3180552,
      "step": 4910
    },
    {
      "epoch": 2.8979952830188678,
      "grad_norm": 1.6036221981048584,
      "learning_rate": 9.938795362526552e-06,
      "loss": 0.5263,
      "num_input_tokens_seen": 3184008,
      "step": 4915
    },
    {
      "epoch": 2.900943396226415,
      "grad_norm": 1.6337107419967651,
      "learning_rate": 9.938393398223255e-06,
      "loss": 0.5663,
      "num_input_tokens_seen": 3187176,
      "step": 4920
    },
    {
      "epoch": 2.9038915094339623,
      "grad_norm": 1.594206690788269,
      "learning_rate": 9.937990126459024e-06,
      "loss": 0.4039,
      "num_input_tokens_seen": 3190792,
      "step": 4925
    },
    {
      "epoch": 2.9068396226415096,
      "grad_norm": 2.1525278091430664,
      "learning_rate": 9.937585547340624e-06,
      "loss": 0.6243,
      "num_input_tokens_seen": 3193288,
      "step": 4930
    },
    {
      "epoch": 2.9097877358490565,
      "grad_norm": 1.9240288734436035,
      "learning_rate": 9.937179660975174e-06,
      "loss": 0.5905,
      "num_input_tokens_seen": 3196392,
      "step": 4935
    },
    {
      "epoch": 2.9127358490566038,
      "grad_norm": 2.1942858695983887,
      "learning_rate": 9.936772467470127e-06,
      "loss": 0.4342,
      "num_input_tokens_seen": 3199720,
      "step": 4940
    },
    {
      "epoch": 2.915683962264151,
      "grad_norm": 3.07348370552063,
      "learning_rate": 9.936363966933294e-06,
      "loss": 0.4653,
      "num_input_tokens_seen": 3203208,
      "step": 4945
    },
    {
      "epoch": 2.918632075471698,
      "grad_norm": 2.4538733959198,
      "learning_rate": 9.935954159472828e-06,
      "loss": 0.6107,
      "num_input_tokens_seen": 3207080,
      "step": 4950
    },
    {
      "epoch": 2.921580188679245,
      "grad_norm": 2.252593755722046,
      "learning_rate": 9.935543045197222e-06,
      "loss": 0.5328,
      "num_input_tokens_seen": 3211368,
      "step": 4955
    },
    {
      "epoch": 2.9245283018867925,
      "grad_norm": 1.5019252300262451,
      "learning_rate": 9.935130624215326e-06,
      "loss": 0.573,
      "num_input_tokens_seen": 3214568,
      "step": 4960
    },
    {
      "epoch": 2.9274764150943398,
      "grad_norm": 1.9371074438095093,
      "learning_rate": 9.934716896636329e-06,
      "loss": 0.5996,
      "num_input_tokens_seen": 3217896,
      "step": 4965
    },
    {
      "epoch": 2.930424528301887,
      "grad_norm": 1.4860302209854126,
      "learning_rate": 9.934301862569764e-06,
      "loss": 0.3983,
      "num_input_tokens_seen": 3221768,
      "step": 4970
    },
    {
      "epoch": 2.933372641509434,
      "grad_norm": 1.5999526977539062,
      "learning_rate": 9.933885522125517e-06,
      "loss": 0.4026,
      "num_input_tokens_seen": 3224424,
      "step": 4975
    },
    {
      "epoch": 2.936320754716981,
      "grad_norm": 3.1636881828308105,
      "learning_rate": 9.933467875413813e-06,
      "loss": 0.5272,
      "num_input_tokens_seen": 3227432,
      "step": 4980
    },
    {
      "epoch": 2.9392688679245285,
      "grad_norm": 3.6409122943878174,
      "learning_rate": 9.933048922545227e-06,
      "loss": 0.4597,
      "num_input_tokens_seen": 3230472,
      "step": 4985
    },
    {
      "epoch": 2.9422169811320753,
      "grad_norm": 1.8508120775222778,
      "learning_rate": 9.932628663630679e-06,
      "loss": 0.4544,
      "num_input_tokens_seen": 3233192,
      "step": 4990
    },
    {
      "epoch": 2.9451650943396226,
      "grad_norm": 2.8060081005096436,
      "learning_rate": 9.932207098781432e-06,
      "loss": 0.418,
      "num_input_tokens_seen": 3237416,
      "step": 4995
    },
    {
      "epoch": 2.94811320754717,
      "grad_norm": 2.33608078956604,
      "learning_rate": 9.931784228109102e-06,
      "loss": 0.6054,
      "num_input_tokens_seen": 3240328,
      "step": 5000
    },
    {
      "epoch": 2.951061320754717,
      "grad_norm": 2.568861722946167,
      "learning_rate": 9.93136005172564e-06,
      "loss": 0.5062,
      "num_input_tokens_seen": 3243720,
      "step": 5005
    },
    {
      "epoch": 2.954009433962264,
      "grad_norm": 1.675516128540039,
      "learning_rate": 9.930934569743354e-06,
      "loss": 0.5073,
      "num_input_tokens_seen": 3246472,
      "step": 5010
    },
    {
      "epoch": 2.9569575471698113,
      "grad_norm": 1.0177593231201172,
      "learning_rate": 9.930507782274888e-06,
      "loss": 0.5207,
      "num_input_tokens_seen": 3249512,
      "step": 5015
    },
    {
      "epoch": 2.9599056603773586,
      "grad_norm": 2.6800403594970703,
      "learning_rate": 9.930079689433236e-06,
      "loss": 0.4093,
      "num_input_tokens_seen": 3252136,
      "step": 5020
    },
    {
      "epoch": 2.9628537735849054,
      "grad_norm": 1.6077488660812378,
      "learning_rate": 9.92965029133174e-06,
      "loss": 0.4399,
      "num_input_tokens_seen": 3255400,
      "step": 5025
    },
    {
      "epoch": 2.9658018867924527,
      "grad_norm": 2.364382028579712,
      "learning_rate": 9.929219588084084e-06,
      "loss": 0.4156,
      "num_input_tokens_seen": 3259048,
      "step": 5030
    },
    {
      "epoch": 2.96875,
      "grad_norm": 1.3019481897354126,
      "learning_rate": 9.9287875798043e-06,
      "loss": 0.5042,
      "num_input_tokens_seen": 3262472,
      "step": 5035
    },
    {
      "epoch": 2.9716981132075473,
      "grad_norm": 1.8296897411346436,
      "learning_rate": 9.92835426660676e-06,
      "loss": 0.4362,
      "num_input_tokens_seen": 3265192,
      "step": 5040
    },
    {
      "epoch": 2.9746462264150946,
      "grad_norm": 1.8502777814865112,
      "learning_rate": 9.927919648606188e-06,
      "loss": 0.4586,
      "num_input_tokens_seen": 3268968,
      "step": 5045
    },
    {
      "epoch": 2.9775943396226414,
      "grad_norm": 2.1320178508758545,
      "learning_rate": 9.927483725917652e-06,
      "loss": 0.557,
      "num_input_tokens_seen": 3271624,
      "step": 5050
    },
    {
      "epoch": 2.9805424528301887,
      "grad_norm": 8.295109748840332,
      "learning_rate": 9.927046498656562e-06,
      "loss": 0.448,
      "num_input_tokens_seen": 3275944,
      "step": 5055
    },
    {
      "epoch": 2.983490566037736,
      "grad_norm": 2.03755784034729,
      "learning_rate": 9.926607966938679e-06,
      "loss": 0.5426,
      "num_input_tokens_seen": 3278728,
      "step": 5060
    },
    {
      "epoch": 2.986438679245283,
      "grad_norm": 3.6545090675354004,
      "learning_rate": 9.926168130880103e-06,
      "loss": 0.6271,
      "num_input_tokens_seen": 3282600,
      "step": 5065
    },
    {
      "epoch": 2.98938679245283,
      "grad_norm": 4.340933799743652,
      "learning_rate": 9.925726990597283e-06,
      "loss": 0.4411,
      "num_input_tokens_seen": 3287080,
      "step": 5070
    },
    {
      "epoch": 2.9923349056603774,
      "grad_norm": 1.2086628675460815,
      "learning_rate": 9.925284546207015e-06,
      "loss": 0.5322,
      "num_input_tokens_seen": 3290312,
      "step": 5075
    },
    {
      "epoch": 2.9952830188679247,
      "grad_norm": 1.689041256904602,
      "learning_rate": 9.924840797826436e-06,
      "loss": 0.4084,
      "num_input_tokens_seen": 3292968,
      "step": 5080
    },
    {
      "epoch": 2.9982311320754715,
      "grad_norm": 1.6213297843933105,
      "learning_rate": 9.924395745573029e-06,
      "loss": 0.3309,
      "num_input_tokens_seen": 3299304,
      "step": 5085
    },
    {
      "epoch": 3.001179245283019,
      "grad_norm": 2.1960971355438232,
      "learning_rate": 9.923949389564629e-06,
      "loss": 0.489,
      "num_input_tokens_seen": 3302080,
      "step": 5090
    },
    {
      "epoch": 3.004127358490566,
      "grad_norm": 1.2712323665618896,
      "learning_rate": 9.923501729919404e-06,
      "loss": 0.4496,
      "num_input_tokens_seen": 3304864,
      "step": 5095
    },
    {
      "epoch": 3.0070754716981134,
      "grad_norm": 3.5135300159454346,
      "learning_rate": 9.923052766755878e-06,
      "loss": 0.5469,
      "num_input_tokens_seen": 3307456,
      "step": 5100
    },
    {
      "epoch": 3.0100235849056602,
      "grad_norm": 1.1356046199798584,
      "learning_rate": 9.922602500192914e-06,
      "loss": 0.4928,
      "num_input_tokens_seen": 3310368,
      "step": 5105
    },
    {
      "epoch": 3.0129716981132075,
      "grad_norm": 1.3970695734024048,
      "learning_rate": 9.922150930349725e-06,
      "loss": 0.459,
      "num_input_tokens_seen": 3314112,
      "step": 5110
    },
    {
      "epoch": 3.015919811320755,
      "grad_norm": 2.0845797061920166,
      "learning_rate": 9.921698057345863e-06,
      "loss": 0.5458,
      "num_input_tokens_seen": 3316544,
      "step": 5115
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 2.403240919113159,
      "learning_rate": 9.921243881301229e-06,
      "loss": 0.5698,
      "num_input_tokens_seen": 3319744,
      "step": 5120
    },
    {
      "epoch": 3.021816037735849,
      "grad_norm": 2.587275505065918,
      "learning_rate": 9.920788402336068e-06,
      "loss": 0.4696,
      "num_input_tokens_seen": 3322592,
      "step": 5125
    },
    {
      "epoch": 3.0247641509433962,
      "grad_norm": 3.190228223800659,
      "learning_rate": 9.92033162057097e-06,
      "loss": 0.5509,
      "num_input_tokens_seen": 3325408,
      "step": 5130
    },
    {
      "epoch": 3.0277122641509435,
      "grad_norm": 1.9525765180587769,
      "learning_rate": 9.919873536126869e-06,
      "loss": 0.4337,
      "num_input_tokens_seen": 3328672,
      "step": 5135
    },
    {
      "epoch": 3.0306603773584904,
      "grad_norm": 1.5256085395812988,
      "learning_rate": 9.919414149125046e-06,
      "loss": 0.5101,
      "num_input_tokens_seen": 3331616,
      "step": 5140
    },
    {
      "epoch": 3.0336084905660377,
      "grad_norm": 1.285293698310852,
      "learning_rate": 9.918953459687126e-06,
      "loss": 0.4275,
      "num_input_tokens_seen": 3334976,
      "step": 5145
    },
    {
      "epoch": 3.036556603773585,
      "grad_norm": 2.126821279525757,
      "learning_rate": 9.918491467935078e-06,
      "loss": 0.8169,
      "num_input_tokens_seen": 3338784,
      "step": 5150
    },
    {
      "epoch": 3.0395047169811322,
      "grad_norm": 4.069960117340088,
      "learning_rate": 9.918028173991218e-06,
      "loss": 0.5354,
      "num_input_tokens_seen": 3342944,
      "step": 5155
    },
    {
      "epoch": 3.042452830188679,
      "grad_norm": 1.3570291996002197,
      "learning_rate": 9.917563577978202e-06,
      "loss": 0.4581,
      "num_input_tokens_seen": 3345792,
      "step": 5160
    },
    {
      "epoch": 3.0454009433962264,
      "grad_norm": 1.7390260696411133,
      "learning_rate": 9.917097680019035e-06,
      "loss": 0.5009,
      "num_input_tokens_seen": 3349088,
      "step": 5165
    },
    {
      "epoch": 3.0483490566037736,
      "grad_norm": 4.012695789337158,
      "learning_rate": 9.916630480237066e-06,
      "loss": 0.4512,
      "num_input_tokens_seen": 3352064,
      "step": 5170
    },
    {
      "epoch": 3.051297169811321,
      "grad_norm": 1.7745596170425415,
      "learning_rate": 9.916161978755988e-06,
      "loss": 0.554,
      "num_input_tokens_seen": 3354656,
      "step": 5175
    },
    {
      "epoch": 3.0542452830188678,
      "grad_norm": 1.7283419370651245,
      "learning_rate": 9.915692175699838e-06,
      "loss": 0.5184,
      "num_input_tokens_seen": 3357632,
      "step": 5180
    },
    {
      "epoch": 3.057193396226415,
      "grad_norm": 1.2922356128692627,
      "learning_rate": 9.915221071193e-06,
      "loss": 0.4226,
      "num_input_tokens_seen": 3361152,
      "step": 5185
    },
    {
      "epoch": 3.0601415094339623,
      "grad_norm": 2.053941011428833,
      "learning_rate": 9.914748665360199e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 3364992,
      "step": 5190
    },
    {
      "epoch": 3.0630896226415096,
      "grad_norm": 2.5685923099517822,
      "learning_rate": 9.914274958326507e-06,
      "loss": 0.4547,
      "num_input_tokens_seen": 3367808,
      "step": 5195
    },
    {
      "epoch": 3.0660377358490565,
      "grad_norm": 3.49634051322937,
      "learning_rate": 9.913799950217341e-06,
      "loss": 0.4794,
      "num_input_tokens_seen": 3370880,
      "step": 5200
    },
    {
      "epoch": 3.0689858490566038,
      "grad_norm": 2.105185031890869,
      "learning_rate": 9.91332364115846e-06,
      "loss": 0.4417,
      "num_input_tokens_seen": 3375776,
      "step": 5205
    },
    {
      "epoch": 3.071933962264151,
      "grad_norm": 2.797377824783325,
      "learning_rate": 9.912846031275972e-06,
      "loss": 0.5206,
      "num_input_tokens_seen": 3378880,
      "step": 5210
    },
    {
      "epoch": 3.074882075471698,
      "grad_norm": 1.2068809270858765,
      "learning_rate": 9.912367120696322e-06,
      "loss": 0.5486,
      "num_input_tokens_seen": 3382304,
      "step": 5215
    },
    {
      "epoch": 3.077830188679245,
      "grad_norm": 1.6742191314697266,
      "learning_rate": 9.911886909546307e-06,
      "loss": 0.441,
      "num_input_tokens_seen": 3386176,
      "step": 5220
    },
    {
      "epoch": 3.0807783018867925,
      "grad_norm": 3.0663743019104004,
      "learning_rate": 9.911405397953063e-06,
      "loss": 0.5304,
      "num_input_tokens_seen": 3389632,
      "step": 5225
    },
    {
      "epoch": 3.0837264150943398,
      "grad_norm": 2.718446731567383,
      "learning_rate": 9.910922586044073e-06,
      "loss": 0.51,
      "num_input_tokens_seen": 3392608,
      "step": 5230
    },
    {
      "epoch": 3.0866745283018866,
      "grad_norm": 1.5254319906234741,
      "learning_rate": 9.910438473947163e-06,
      "loss": 0.3314,
      "num_input_tokens_seen": 3396064,
      "step": 5235
    },
    {
      "epoch": 3.089622641509434,
      "grad_norm": 1.861487865447998,
      "learning_rate": 9.909953061790506e-06,
      "loss": 0.4556,
      "num_input_tokens_seen": 3399136,
      "step": 5240
    },
    {
      "epoch": 3.092570754716981,
      "grad_norm": 1.7614611387252808,
      "learning_rate": 9.909466349702613e-06,
      "loss": 0.4585,
      "num_input_tokens_seen": 3401920,
      "step": 5245
    },
    {
      "epoch": 3.0955188679245285,
      "grad_norm": 1.975260853767395,
      "learning_rate": 9.908978337812348e-06,
      "loss": 0.375,
      "num_input_tokens_seen": 3405600,
      "step": 5250
    },
    {
      "epoch": 3.0984669811320753,
      "grad_norm": 1.1806602478027344,
      "learning_rate": 9.908489026248909e-06,
      "loss": 0.4913,
      "num_input_tokens_seen": 3407936,
      "step": 5255
    },
    {
      "epoch": 3.1014150943396226,
      "grad_norm": 1.5589649677276611,
      "learning_rate": 9.907998415141846e-06,
      "loss": 0.4027,
      "num_input_tokens_seen": 3412544,
      "step": 5260
    },
    {
      "epoch": 3.10436320754717,
      "grad_norm": 1.746651291847229,
      "learning_rate": 9.907506504621052e-06,
      "loss": 0.4925,
      "num_input_tokens_seen": 3415200,
      "step": 5265
    },
    {
      "epoch": 3.107311320754717,
      "grad_norm": 3.0600879192352295,
      "learning_rate": 9.907013294816759e-06,
      "loss": 0.4726,
      "num_input_tokens_seen": 3418400,
      "step": 5270
    },
    {
      "epoch": 3.110259433962264,
      "grad_norm": 2.1018898487091064,
      "learning_rate": 9.906518785859548e-06,
      "loss": 0.5124,
      "num_input_tokens_seen": 3421088,
      "step": 5275
    },
    {
      "epoch": 3.1132075471698113,
      "grad_norm": 1.76198148727417,
      "learning_rate": 9.906022977880344e-06,
      "loss": 0.4177,
      "num_input_tokens_seen": 3425184,
      "step": 5280
    },
    {
      "epoch": 3.1161556603773586,
      "grad_norm": 1.1819761991500854,
      "learning_rate": 9.905525871010412e-06,
      "loss": 0.5843,
      "num_input_tokens_seen": 3427904,
      "step": 5285
    },
    {
      "epoch": 3.119103773584906,
      "grad_norm": 3.1766910552978516,
      "learning_rate": 9.905027465381363e-06,
      "loss": 0.5202,
      "num_input_tokens_seen": 3430464,
      "step": 5290
    },
    {
      "epoch": 3.1220518867924527,
      "grad_norm": 1.5864715576171875,
      "learning_rate": 9.904527761125155e-06,
      "loss": 0.4577,
      "num_input_tokens_seen": 3433760,
      "step": 5295
    },
    {
      "epoch": 3.125,
      "grad_norm": 1.4166820049285889,
      "learning_rate": 9.904026758374083e-06,
      "loss": 0.448,
      "num_input_tokens_seen": 3437088,
      "step": 5300
    },
    {
      "epoch": 3.1279481132075473,
      "grad_norm": 3.061554193496704,
      "learning_rate": 9.903524457260794e-06,
      "loss": 0.4661,
      "num_input_tokens_seen": 3440384,
      "step": 5305
    },
    {
      "epoch": 3.1308962264150946,
      "grad_norm": 3.6893649101257324,
      "learning_rate": 9.90302085791827e-06,
      "loss": 0.4262,
      "num_input_tokens_seen": 3443616,
      "step": 5310
    },
    {
      "epoch": 3.1338443396226414,
      "grad_norm": 1.5704944133758545,
      "learning_rate": 9.902515960479844e-06,
      "loss": 0.4515,
      "num_input_tokens_seen": 3446368,
      "step": 5315
    },
    {
      "epoch": 3.1367924528301887,
      "grad_norm": 1.8313522338867188,
      "learning_rate": 9.902009765079188e-06,
      "loss": 0.452,
      "num_input_tokens_seen": 3449472,
      "step": 5320
    },
    {
      "epoch": 3.139740566037736,
      "grad_norm": 1.4776679277420044,
      "learning_rate": 9.90150227185032e-06,
      "loss": 0.4189,
      "num_input_tokens_seen": 3452576,
      "step": 5325
    },
    {
      "epoch": 3.142688679245283,
      "grad_norm": 1.9030383825302124,
      "learning_rate": 9.900993480927603e-06,
      "loss": 0.4892,
      "num_input_tokens_seen": 3456000,
      "step": 5330
    },
    {
      "epoch": 3.14563679245283,
      "grad_norm": 5.217904090881348,
      "learning_rate": 9.90048339244574e-06,
      "loss": 0.4883,
      "num_input_tokens_seen": 3458432,
      "step": 5335
    },
    {
      "epoch": 3.1485849056603774,
      "grad_norm": 1.8816336393356323,
      "learning_rate": 9.899972006539776e-06,
      "loss": 0.525,
      "num_input_tokens_seen": 3461440,
      "step": 5340
    },
    {
      "epoch": 3.1515330188679247,
      "grad_norm": 4.05851411819458,
      "learning_rate": 9.899459323345106e-06,
      "loss": 0.5811,
      "num_input_tokens_seen": 3464096,
      "step": 5345
    },
    {
      "epoch": 3.1544811320754715,
      "grad_norm": 2.0512566566467285,
      "learning_rate": 9.898945342997467e-06,
      "loss": 0.5928,
      "num_input_tokens_seen": 3466464,
      "step": 5350
    },
    {
      "epoch": 3.157429245283019,
      "grad_norm": 1.5012661218643188,
      "learning_rate": 9.898430065632933e-06,
      "loss": 0.278,
      "num_input_tokens_seen": 3468928,
      "step": 5355
    },
    {
      "epoch": 3.160377358490566,
      "grad_norm": 2.1698708534240723,
      "learning_rate": 9.897913491387929e-06,
      "loss": 0.5056,
      "num_input_tokens_seen": 3471648,
      "step": 5360
    },
    {
      "epoch": 3.1633254716981134,
      "grad_norm": 2.1641149520874023,
      "learning_rate": 9.897395620399219e-06,
      "loss": 0.5614,
      "num_input_tokens_seen": 3475712,
      "step": 5365
    },
    {
      "epoch": 3.1662735849056602,
      "grad_norm": 2.8378162384033203,
      "learning_rate": 9.896876452803913e-06,
      "loss": 0.4764,
      "num_input_tokens_seen": 3478912,
      "step": 5370
    },
    {
      "epoch": 3.1692216981132075,
      "grad_norm": 2.2220747470855713,
      "learning_rate": 9.896355988739461e-06,
      "loss": 0.4267,
      "num_input_tokens_seen": 3482400,
      "step": 5375
    },
    {
      "epoch": 3.172169811320755,
      "grad_norm": 1.1470500230789185,
      "learning_rate": 9.895834228343658e-06,
      "loss": 0.495,
      "num_input_tokens_seen": 3485600,
      "step": 5380
    },
    {
      "epoch": 3.175117924528302,
      "grad_norm": 1.809543251991272,
      "learning_rate": 9.895311171754644e-06,
      "loss": 0.4203,
      "num_input_tokens_seen": 3488416,
      "step": 5385
    },
    {
      "epoch": 3.178066037735849,
      "grad_norm": 1.3974146842956543,
      "learning_rate": 9.8947868191109e-06,
      "loss": 0.4349,
      "num_input_tokens_seen": 3493376,
      "step": 5390
    },
    {
      "epoch": 3.1810141509433962,
      "grad_norm": 1.1885560750961304,
      "learning_rate": 9.894261170551249e-06,
      "loss": 0.509,
      "num_input_tokens_seen": 3497472,
      "step": 5395
    },
    {
      "epoch": 3.1839622641509435,
      "grad_norm": 2.136662483215332,
      "learning_rate": 9.893734226214861e-06,
      "loss": 0.4144,
      "num_input_tokens_seen": 3499680,
      "step": 5400
    },
    {
      "epoch": 3.1869103773584904,
      "grad_norm": 3.299881935119629,
      "learning_rate": 9.893205986241246e-06,
      "loss": 0.4382,
      "num_input_tokens_seen": 3502144,
      "step": 5405
    },
    {
      "epoch": 3.1898584905660377,
      "grad_norm": 1.4974356889724731,
      "learning_rate": 9.892676450770257e-06,
      "loss": 0.4355,
      "num_input_tokens_seen": 3505056,
      "step": 5410
    },
    {
      "epoch": 3.192806603773585,
      "grad_norm": 2.129697799682617,
      "learning_rate": 9.892145619942092e-06,
      "loss": 0.4764,
      "num_input_tokens_seen": 3507808,
      "step": 5415
    },
    {
      "epoch": 3.1957547169811322,
      "grad_norm": 2.4239165782928467,
      "learning_rate": 9.891613493897289e-06,
      "loss": 0.4584,
      "num_input_tokens_seen": 3510880,
      "step": 5420
    },
    {
      "epoch": 3.198702830188679,
      "grad_norm": 1.7504124641418457,
      "learning_rate": 9.891080072776733e-06,
      "loss": 0.4791,
      "num_input_tokens_seen": 3513696,
      "step": 5425
    },
    {
      "epoch": 3.2016509433962264,
      "grad_norm": 2.734243154525757,
      "learning_rate": 9.890545356721649e-06,
      "loss": 0.5276,
      "num_input_tokens_seen": 3516416,
      "step": 5430
    },
    {
      "epoch": 3.2045990566037736,
      "grad_norm": 1.8177443742752075,
      "learning_rate": 9.890009345873603e-06,
      "loss": 0.5143,
      "num_input_tokens_seen": 3519072,
      "step": 5435
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 1.631929874420166,
      "learning_rate": 9.889472040374509e-06,
      "loss": 0.4423,
      "num_input_tokens_seen": 3522016,
      "step": 5440
    },
    {
      "epoch": 3.2104952830188678,
      "grad_norm": 2.4058399200439453,
      "learning_rate": 9.88893344036662e-06,
      "loss": 0.4209,
      "num_input_tokens_seen": 3525440,
      "step": 5445
    },
    {
      "epoch": 3.213443396226415,
      "grad_norm": 1.6914806365966797,
      "learning_rate": 9.888393545992531e-06,
      "loss": 0.4773,
      "num_input_tokens_seen": 3528800,
      "step": 5450
    },
    {
      "epoch": 3.2163915094339623,
      "grad_norm": 1.7088629007339478,
      "learning_rate": 9.887852357395184e-06,
      "loss": 0.494,
      "num_input_tokens_seen": 3531808,
      "step": 5455
    },
    {
      "epoch": 3.2193396226415096,
      "grad_norm": 3.0460543632507324,
      "learning_rate": 9.88730987471786e-06,
      "loss": 0.4976,
      "num_input_tokens_seen": 3534400,
      "step": 5460
    },
    {
      "epoch": 3.2222877358490565,
      "grad_norm": 2.6446404457092285,
      "learning_rate": 9.886766098104183e-06,
      "loss": 0.474,
      "num_input_tokens_seen": 3536992,
      "step": 5465
    },
    {
      "epoch": 3.2252358490566038,
      "grad_norm": 1.8440783023834229,
      "learning_rate": 9.886221027698122e-06,
      "loss": 0.3967,
      "num_input_tokens_seen": 3540928,
      "step": 5470
    },
    {
      "epoch": 3.228183962264151,
      "grad_norm": 2.2672481536865234,
      "learning_rate": 9.885674663643983e-06,
      "loss": 0.4166,
      "num_input_tokens_seen": 3543936,
      "step": 5475
    },
    {
      "epoch": 3.231132075471698,
      "grad_norm": 1.4382073879241943,
      "learning_rate": 9.885127006086423e-06,
      "loss": 0.3951,
      "num_input_tokens_seen": 3547200,
      "step": 5480
    },
    {
      "epoch": 3.234080188679245,
      "grad_norm": 3.0813891887664795,
      "learning_rate": 9.884578055170434e-06,
      "loss": 0.5032,
      "num_input_tokens_seen": 3550080,
      "step": 5485
    },
    {
      "epoch": 3.2370283018867925,
      "grad_norm": 3.0672738552093506,
      "learning_rate": 9.884027811041353e-06,
      "loss": 0.406,
      "num_input_tokens_seen": 3553280,
      "step": 5490
    },
    {
      "epoch": 3.2399764150943398,
      "grad_norm": 4.433968544006348,
      "learning_rate": 9.883476273844861e-06,
      "loss": 0.4918,
      "num_input_tokens_seen": 3557280,
      "step": 5495
    },
    {
      "epoch": 3.2429245283018866,
      "grad_norm": 1.950089693069458,
      "learning_rate": 9.882923443726977e-06,
      "loss": 0.4258,
      "num_input_tokens_seen": 3561024,
      "step": 5500
    },
    {
      "epoch": 3.245872641509434,
      "grad_norm": 2.0157511234283447,
      "learning_rate": 9.882369320834068e-06,
      "loss": 0.4516,
      "num_input_tokens_seen": 3563968,
      "step": 5505
    },
    {
      "epoch": 3.248820754716981,
      "grad_norm": 1.4880813360214233,
      "learning_rate": 9.88181390531284e-06,
      "loss": 0.5196,
      "num_input_tokens_seen": 3567776,
      "step": 5510
    },
    {
      "epoch": 3.2517688679245285,
      "grad_norm": 6.786020278930664,
      "learning_rate": 9.88125719731034e-06,
      "loss": 0.4333,
      "num_input_tokens_seen": 3570624,
      "step": 5515
    },
    {
      "epoch": 3.2547169811320753,
      "grad_norm": 1.9518582820892334,
      "learning_rate": 9.880699196973962e-06,
      "loss": 0.4359,
      "num_input_tokens_seen": 3573632,
      "step": 5520
    },
    {
      "epoch": 3.2576650943396226,
      "grad_norm": 2.120213031768799,
      "learning_rate": 9.880139904451436e-06,
      "loss": 0.4875,
      "num_input_tokens_seen": 3576768,
      "step": 5525
    },
    {
      "epoch": 3.26061320754717,
      "grad_norm": 2.36179780960083,
      "learning_rate": 9.879579319890838e-06,
      "loss": 0.5002,
      "num_input_tokens_seen": 3579616,
      "step": 5530
    },
    {
      "epoch": 3.263561320754717,
      "grad_norm": 1.9991211891174316,
      "learning_rate": 9.879017443440584e-06,
      "loss": 0.416,
      "num_input_tokens_seen": 3582784,
      "step": 5535
    },
    {
      "epoch": 3.266509433962264,
      "grad_norm": 2.359377384185791,
      "learning_rate": 9.878454275249436e-06,
      "loss": 0.5169,
      "num_input_tokens_seen": 3585792,
      "step": 5540
    },
    {
      "epoch": 3.2694575471698113,
      "grad_norm": 2.206561326980591,
      "learning_rate": 9.877889815466493e-06,
      "loss": 0.4126,
      "num_input_tokens_seen": 3588832,
      "step": 5545
    },
    {
      "epoch": 3.2724056603773586,
      "grad_norm": 1.7862672805786133,
      "learning_rate": 9.877324064241198e-06,
      "loss": 0.5485,
      "num_input_tokens_seen": 3591328,
      "step": 5550
    },
    {
      "epoch": 3.2753537735849054,
      "grad_norm": 1.6798213720321655,
      "learning_rate": 9.876757021723338e-06,
      "loss": 0.4379,
      "num_input_tokens_seen": 3594752,
      "step": 5555
    },
    {
      "epoch": 3.2783018867924527,
      "grad_norm": 1.7132596969604492,
      "learning_rate": 9.876188688063038e-06,
      "loss": 0.3631,
      "num_input_tokens_seen": 3599552,
      "step": 5560
    },
    {
      "epoch": 3.28125,
      "grad_norm": 8.030023574829102,
      "learning_rate": 9.875619063410768e-06,
      "loss": 0.4891,
      "num_input_tokens_seen": 3604256,
      "step": 5565
    },
    {
      "epoch": 3.2841981132075473,
      "grad_norm": 8.043830871582031,
      "learning_rate": 9.875048147917339e-06,
      "loss": 0.4978,
      "num_input_tokens_seen": 3607488,
      "step": 5570
    },
    {
      "epoch": 3.2871462264150946,
      "grad_norm": 1.950074553489685,
      "learning_rate": 9.874475941733902e-06,
      "loss": 0.4556,
      "num_input_tokens_seen": 3610528,
      "step": 5575
    },
    {
      "epoch": 3.2900943396226414,
      "grad_norm": 2.684434413909912,
      "learning_rate": 9.873902445011952e-06,
      "loss": 0.6185,
      "num_input_tokens_seen": 3613600,
      "step": 5580
    },
    {
      "epoch": 3.2930424528301887,
      "grad_norm": 1.1247355937957764,
      "learning_rate": 9.873327657903324e-06,
      "loss": 0.4347,
      "num_input_tokens_seen": 3618240,
      "step": 5585
    },
    {
      "epoch": 3.295990566037736,
      "grad_norm": 2.886599540710449,
      "learning_rate": 9.872751580560194e-06,
      "loss": 0.3619,
      "num_input_tokens_seen": 3620672,
      "step": 5590
    },
    {
      "epoch": 3.298938679245283,
      "grad_norm": 1.4169235229492188,
      "learning_rate": 9.872174213135084e-06,
      "loss": 0.4113,
      "num_input_tokens_seen": 3624064,
      "step": 5595
    },
    {
      "epoch": 3.30188679245283,
      "grad_norm": 2.2366435527801514,
      "learning_rate": 9.871595555780855e-06,
      "loss": 0.4402,
      "num_input_tokens_seen": 3628448,
      "step": 5600
    },
    {
      "epoch": 3.3048349056603774,
      "grad_norm": 1.3526092767715454,
      "learning_rate": 9.871015608650705e-06,
      "loss": 0.4504,
      "num_input_tokens_seen": 3632288,
      "step": 5605
    },
    {
      "epoch": 3.3077830188679247,
      "grad_norm": 1.901037573814392,
      "learning_rate": 9.870434371898182e-06,
      "loss": 0.4235,
      "num_input_tokens_seen": 3635328,
      "step": 5610
    },
    {
      "epoch": 3.3107311320754715,
      "grad_norm": 1.9982895851135254,
      "learning_rate": 9.869851845677165e-06,
      "loss": 0.5372,
      "num_input_tokens_seen": 3639168,
      "step": 5615
    },
    {
      "epoch": 3.313679245283019,
      "grad_norm": 3.351459503173828,
      "learning_rate": 9.869268030141886e-06,
      "loss": 0.5745,
      "num_input_tokens_seen": 3642496,
      "step": 5620
    },
    {
      "epoch": 3.316627358490566,
      "grad_norm": 1.8345142602920532,
      "learning_rate": 9.86868292544691e-06,
      "loss": 0.4747,
      "num_input_tokens_seen": 3646464,
      "step": 5625
    },
    {
      "epoch": 3.3195754716981134,
      "grad_norm": 2.1672616004943848,
      "learning_rate": 9.868096531747149e-06,
      "loss": 0.4115,
      "num_input_tokens_seen": 3649440,
      "step": 5630
    },
    {
      "epoch": 3.3225235849056602,
      "grad_norm": 1.8300174474716187,
      "learning_rate": 9.867508849197848e-06,
      "loss": 0.496,
      "num_input_tokens_seen": 3652448,
      "step": 5635
    },
    {
      "epoch": 3.3254716981132075,
      "grad_norm": 1.8519386053085327,
      "learning_rate": 9.866919877954602e-06,
      "loss": 0.4341,
      "num_input_tokens_seen": 3655648,
      "step": 5640
    },
    {
      "epoch": 3.328419811320755,
      "grad_norm": 2.922100305557251,
      "learning_rate": 9.866329618173344e-06,
      "loss": 0.4464,
      "num_input_tokens_seen": 3658816,
      "step": 5645
    },
    {
      "epoch": 3.331367924528302,
      "grad_norm": 3.1896846294403076,
      "learning_rate": 9.865738070010346e-06,
      "loss": 0.4079,
      "num_input_tokens_seen": 3661440,
      "step": 5650
    },
    {
      "epoch": 3.334316037735849,
      "grad_norm": 2.095775842666626,
      "learning_rate": 9.865145233622223e-06,
      "loss": 0.4978,
      "num_input_tokens_seen": 3664992,
      "step": 5655
    },
    {
      "epoch": 3.3372641509433962,
      "grad_norm": 1.3995699882507324,
      "learning_rate": 9.864551109165935e-06,
      "loss": 0.4924,
      "num_input_tokens_seen": 3668896,
      "step": 5660
    },
    {
      "epoch": 3.3402122641509435,
      "grad_norm": 3.110442876815796,
      "learning_rate": 9.863955696798773e-06,
      "loss": 0.627,
      "num_input_tokens_seen": 3671776,
      "step": 5665
    },
    {
      "epoch": 3.3431603773584904,
      "grad_norm": 2.1564478874206543,
      "learning_rate": 9.863358996678378e-06,
      "loss": 0.4239,
      "num_input_tokens_seen": 3676704,
      "step": 5670
    },
    {
      "epoch": 3.3461084905660377,
      "grad_norm": 2.0902209281921387,
      "learning_rate": 9.86276100896273e-06,
      "loss": 0.3924,
      "num_input_tokens_seen": 3679968,
      "step": 5675
    },
    {
      "epoch": 3.349056603773585,
      "grad_norm": 8.25231647491455,
      "learning_rate": 9.862161733810147e-06,
      "loss": 0.4844,
      "num_input_tokens_seen": 3683520,
      "step": 5680
    },
    {
      "epoch": 3.3520047169811322,
      "grad_norm": 1.5295838117599487,
      "learning_rate": 9.86156117137929e-06,
      "loss": 0.3406,
      "num_input_tokens_seen": 3690880,
      "step": 5685
    },
    {
      "epoch": 3.354952830188679,
      "grad_norm": 1.22914457321167,
      "learning_rate": 9.860959321829159e-06,
      "loss": 0.42,
      "num_input_tokens_seen": 3693792,
      "step": 5690
    },
    {
      "epoch": 3.3579009433962264,
      "grad_norm": 1.5589091777801514,
      "learning_rate": 9.860356185319102e-06,
      "loss": 0.4466,
      "num_input_tokens_seen": 3697120,
      "step": 5695
    },
    {
      "epoch": 3.3608490566037736,
      "grad_norm": 2.799814224243164,
      "learning_rate": 9.859751762008796e-06,
      "loss": 0.5086,
      "num_input_tokens_seen": 3700128,
      "step": 5700
    },
    {
      "epoch": 3.363797169811321,
      "grad_norm": 1.8474314212799072,
      "learning_rate": 9.859146052058266e-06,
      "loss": 0.4978,
      "num_input_tokens_seen": 3703392,
      "step": 5705
    },
    {
      "epoch": 3.3667452830188678,
      "grad_norm": 2.669536590576172,
      "learning_rate": 9.858539055627876e-06,
      "loss": 0.3944,
      "num_input_tokens_seen": 3708128,
      "step": 5710
    },
    {
      "epoch": 3.369693396226415,
      "grad_norm": 2.0454375743865967,
      "learning_rate": 9.857930772878333e-06,
      "loss": 0.491,
      "num_input_tokens_seen": 3712288,
      "step": 5715
    },
    {
      "epoch": 3.3726415094339623,
      "grad_norm": 3.2120742797851562,
      "learning_rate": 9.857321203970682e-06,
      "loss": 0.4874,
      "num_input_tokens_seen": 3715104,
      "step": 5720
    },
    {
      "epoch": 3.3755896226415096,
      "grad_norm": 1.434944748878479,
      "learning_rate": 9.856710349066307e-06,
      "loss": 0.4363,
      "num_input_tokens_seen": 3717632,
      "step": 5725
    },
    {
      "epoch": 3.3785377358490565,
      "grad_norm": 1.7580788135528564,
      "learning_rate": 9.856098208326937e-06,
      "loss": 0.3648,
      "num_input_tokens_seen": 3720960,
      "step": 5730
    },
    {
      "epoch": 3.3814858490566038,
      "grad_norm": 2.608097553253174,
      "learning_rate": 9.855484781914639e-06,
      "loss": 0.4547,
      "num_input_tokens_seen": 3724416,
      "step": 5735
    },
    {
      "epoch": 3.384433962264151,
      "grad_norm": 2.233640432357788,
      "learning_rate": 9.854870069991817e-06,
      "loss": 0.3591,
      "num_input_tokens_seen": 3727584,
      "step": 5740
    },
    {
      "epoch": 3.387382075471698,
      "grad_norm": 1.8698911666870117,
      "learning_rate": 9.854254072721222e-06,
      "loss": 0.4979,
      "num_input_tokens_seen": 3730464,
      "step": 5745
    },
    {
      "epoch": 3.390330188679245,
      "grad_norm": 2.263829469680786,
      "learning_rate": 9.853636790265938e-06,
      "loss": 0.4637,
      "num_input_tokens_seen": 3733824,
      "step": 5750
    },
    {
      "epoch": 3.3932783018867925,
      "grad_norm": 1.6877143383026123,
      "learning_rate": 9.853018222789397e-06,
      "loss": 0.4153,
      "num_input_tokens_seen": 3736576,
      "step": 5755
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 2.035196304321289,
      "learning_rate": 9.852398370455367e-06,
      "loss": 0.417,
      "num_input_tokens_seen": 3739232,
      "step": 5760
    },
    {
      "epoch": 3.3991745283018866,
      "grad_norm": 3.0141561031341553,
      "learning_rate": 9.851777233427955e-06,
      "loss": 0.5006,
      "num_input_tokens_seen": 3742208,
      "step": 5765
    },
    {
      "epoch": 3.402122641509434,
      "grad_norm": 1.3382726907730103,
      "learning_rate": 9.85115481187161e-06,
      "loss": 0.5655,
      "num_input_tokens_seen": 3745120,
      "step": 5770
    },
    {
      "epoch": 3.405070754716981,
      "grad_norm": 2.0243446826934814,
      "learning_rate": 9.850531105951123e-06,
      "loss": 0.4984,
      "num_input_tokens_seen": 3747776,
      "step": 5775
    },
    {
      "epoch": 3.4080188679245285,
      "grad_norm": 1.361023187637329,
      "learning_rate": 9.84990611583162e-06,
      "loss": 0.3771,
      "num_input_tokens_seen": 3751680,
      "step": 5780
    },
    {
      "epoch": 3.4109669811320753,
      "grad_norm": 3.93633770942688,
      "learning_rate": 9.849279841678572e-06,
      "loss": 0.4668,
      "num_input_tokens_seen": 3755200,
      "step": 5785
    },
    {
      "epoch": 3.4139150943396226,
      "grad_norm": 2.0003154277801514,
      "learning_rate": 9.848652283657785e-06,
      "loss": 0.4216,
      "num_input_tokens_seen": 3758592,
      "step": 5790
    },
    {
      "epoch": 3.41686320754717,
      "grad_norm": 1.3412904739379883,
      "learning_rate": 9.848023441935411e-06,
      "loss": 0.6104,
      "num_input_tokens_seen": 3761984,
      "step": 5795
    },
    {
      "epoch": 3.419811320754717,
      "grad_norm": 1.2809398174285889,
      "learning_rate": 9.847393316677935e-06,
      "loss": 0.4056,
      "num_input_tokens_seen": 3765472,
      "step": 5800
    },
    {
      "epoch": 3.422759433962264,
      "grad_norm": 2.1033318042755127,
      "learning_rate": 9.846761908052188e-06,
      "loss": 0.407,
      "num_input_tokens_seen": 3768192,
      "step": 5805
    },
    {
      "epoch": 3.4257075471698113,
      "grad_norm": 1.4269148111343384,
      "learning_rate": 9.846129216225338e-06,
      "loss": 0.4225,
      "num_input_tokens_seen": 3771584,
      "step": 5810
    },
    {
      "epoch": 3.4286556603773586,
      "grad_norm": 5.247527599334717,
      "learning_rate": 9.845495241364892e-06,
      "loss": 0.4565,
      "num_input_tokens_seen": 3774112,
      "step": 5815
    },
    {
      "epoch": 3.4316037735849054,
      "grad_norm": 1.5690234899520874,
      "learning_rate": 9.844859983638696e-06,
      "loss": 0.6708,
      "num_input_tokens_seen": 3776928,
      "step": 5820
    },
    {
      "epoch": 3.4345518867924527,
      "grad_norm": 1.6847877502441406,
      "learning_rate": 9.844223443214942e-06,
      "loss": 0.485,
      "num_input_tokens_seen": 3779616,
      "step": 5825
    },
    {
      "epoch": 3.4375,
      "grad_norm": 2.640075922012329,
      "learning_rate": 9.843585620262153e-06,
      "loss": 0.5085,
      "num_input_tokens_seen": 3782016,
      "step": 5830
    },
    {
      "epoch": 3.4404481132075473,
      "grad_norm": 1.2328760623931885,
      "learning_rate": 9.842946514949197e-06,
      "loss": 0.464,
      "num_input_tokens_seen": 3784800,
      "step": 5835
    },
    {
      "epoch": 3.4433962264150946,
      "grad_norm": 3.0534818172454834,
      "learning_rate": 9.842306127445279e-06,
      "loss": 0.4552,
      "num_input_tokens_seen": 3788064,
      "step": 5840
    },
    {
      "epoch": 3.4463443396226414,
      "grad_norm": 2.8548316955566406,
      "learning_rate": 9.841664457919944e-06,
      "loss": 0.533,
      "num_input_tokens_seen": 3790976,
      "step": 5845
    },
    {
      "epoch": 3.4492924528301887,
      "grad_norm": 1.388437271118164,
      "learning_rate": 9.841021506543079e-06,
      "loss": 0.4185,
      "num_input_tokens_seen": 3795296,
      "step": 5850
    },
    {
      "epoch": 3.452240566037736,
      "grad_norm": 2.8997609615325928,
      "learning_rate": 9.840377273484904e-06,
      "loss": 0.5208,
      "num_input_tokens_seen": 3797920,
      "step": 5855
    },
    {
      "epoch": 3.455188679245283,
      "grad_norm": 1.398000955581665,
      "learning_rate": 9.839731758915986e-06,
      "loss": 0.347,
      "num_input_tokens_seen": 3801184,
      "step": 5860
    },
    {
      "epoch": 3.45813679245283,
      "grad_norm": 1.5464544296264648,
      "learning_rate": 9.839084963007226e-06,
      "loss": 0.5086,
      "num_input_tokens_seen": 3804544,
      "step": 5865
    },
    {
      "epoch": 3.4610849056603774,
      "grad_norm": 1.622169852256775,
      "learning_rate": 9.838436885929868e-06,
      "loss": 0.6002,
      "num_input_tokens_seen": 3807136,
      "step": 5870
    },
    {
      "epoch": 3.4640330188679247,
      "grad_norm": 2.614377737045288,
      "learning_rate": 9.837787527855492e-06,
      "loss": 0.5204,
      "num_input_tokens_seen": 3810304,
      "step": 5875
    },
    {
      "epoch": 3.4669811320754715,
      "grad_norm": 2.542466878890991,
      "learning_rate": 9.837136888956017e-06,
      "loss": 0.3899,
      "num_input_tokens_seen": 3814144,
      "step": 5880
    },
    {
      "epoch": 3.469929245283019,
      "grad_norm": 2.07767915725708,
      "learning_rate": 9.836484969403705e-06,
      "loss": 0.5318,
      "num_input_tokens_seen": 3816512,
      "step": 5885
    },
    {
      "epoch": 3.472877358490566,
      "grad_norm": 1.9065115451812744,
      "learning_rate": 9.835831769371152e-06,
      "loss": 0.3979,
      "num_input_tokens_seen": 3819712,
      "step": 5890
    },
    {
      "epoch": 3.4758254716981134,
      "grad_norm": 2.994140386581421,
      "learning_rate": 9.835177289031298e-06,
      "loss": 0.6733,
      "num_input_tokens_seen": 3822944,
      "step": 5895
    },
    {
      "epoch": 3.4787735849056602,
      "grad_norm": 2.2046685218811035,
      "learning_rate": 9.834521528557419e-06,
      "loss": 0.5089,
      "num_input_tokens_seen": 3825920,
      "step": 5900
    },
    {
      "epoch": 3.4817216981132075,
      "grad_norm": 2.466369867324829,
      "learning_rate": 9.833864488123128e-06,
      "loss": 0.4965,
      "num_input_tokens_seen": 3828672,
      "step": 5905
    },
    {
      "epoch": 3.484669811320755,
      "grad_norm": 1.5901148319244385,
      "learning_rate": 9.83320616790238e-06,
      "loss": 0.4464,
      "num_input_tokens_seen": 3831904,
      "step": 5910
    },
    {
      "epoch": 3.487617924528302,
      "grad_norm": 3.3060686588287354,
      "learning_rate": 9.832546568069472e-06,
      "loss": 0.4727,
      "num_input_tokens_seen": 3835072,
      "step": 5915
    },
    {
      "epoch": 3.490566037735849,
      "grad_norm": 3.6388742923736572,
      "learning_rate": 9.831885688799031e-06,
      "loss": 0.4594,
      "num_input_tokens_seen": 3837664,
      "step": 5920
    },
    {
      "epoch": 3.4935141509433962,
      "grad_norm": 1.4763870239257812,
      "learning_rate": 9.83122353026603e-06,
      "loss": 0.6165,
      "num_input_tokens_seen": 3840416,
      "step": 5925
    },
    {
      "epoch": 3.4964622641509435,
      "grad_norm": 1.8185480833053589,
      "learning_rate": 9.830560092645778e-06,
      "loss": 0.436,
      "num_input_tokens_seen": 3843552,
      "step": 5930
    },
    {
      "epoch": 3.4994103773584904,
      "grad_norm": 1.5732786655426025,
      "learning_rate": 9.829895376113923e-06,
      "loss": 0.4683,
      "num_input_tokens_seen": 3846848,
      "step": 5935
    },
    {
      "epoch": 3.5023584905660377,
      "grad_norm": 2.5171713829040527,
      "learning_rate": 9.829229380846452e-06,
      "loss": 0.4139,
      "num_input_tokens_seen": 3849824,
      "step": 5940
    },
    {
      "epoch": 3.505306603773585,
      "grad_norm": 2.1614503860473633,
      "learning_rate": 9.82856210701969e-06,
      "loss": 0.4649,
      "num_input_tokens_seen": 3852864,
      "step": 5945
    },
    {
      "epoch": 3.5082547169811322,
      "grad_norm": 1.4068658351898193,
      "learning_rate": 9.827893554810298e-06,
      "loss": 0.3965,
      "num_input_tokens_seen": 3855680,
      "step": 5950
    },
    {
      "epoch": 3.5112028301886795,
      "grad_norm": 1.128760814666748,
      "learning_rate": 9.827223724395281e-06,
      "loss": 0.4659,
      "num_input_tokens_seen": 3858304,
      "step": 5955
    },
    {
      "epoch": 3.5141509433962264,
      "grad_norm": 2.2031826972961426,
      "learning_rate": 9.82655261595198e-06,
      "loss": 0.4574,
      "num_input_tokens_seen": 3861760,
      "step": 5960
    },
    {
      "epoch": 3.5170990566037736,
      "grad_norm": 2.0494754314422607,
      "learning_rate": 9.825880229658073e-06,
      "loss": 0.4459,
      "num_input_tokens_seen": 3864480,
      "step": 5965
    },
    {
      "epoch": 3.5200471698113205,
      "grad_norm": 1.826555848121643,
      "learning_rate": 9.825206565691576e-06,
      "loss": 0.4653,
      "num_input_tokens_seen": 3867520,
      "step": 5970
    },
    {
      "epoch": 3.5229952830188678,
      "grad_norm": 2.6331803798675537,
      "learning_rate": 9.824531624230844e-06,
      "loss": 0.4051,
      "num_input_tokens_seen": 3871040,
      "step": 5975
    },
    {
      "epoch": 3.525943396226415,
      "grad_norm": 1.9716696739196777,
      "learning_rate": 9.823855405454573e-06,
      "loss": 0.4543,
      "num_input_tokens_seen": 3874208,
      "step": 5980
    },
    {
      "epoch": 3.5288915094339623,
      "grad_norm": 1.7854045629501343,
      "learning_rate": 9.823177909541795e-06,
      "loss": 0.4649,
      "num_input_tokens_seen": 3876992,
      "step": 5985
    },
    {
      "epoch": 3.5318396226415096,
      "grad_norm": 2.4777956008911133,
      "learning_rate": 9.822499136671877e-06,
      "loss": 0.5075,
      "num_input_tokens_seen": 3880384,
      "step": 5990
    },
    {
      "epoch": 3.5347877358490565,
      "grad_norm": 1.5444972515106201,
      "learning_rate": 9.82181908702453e-06,
      "loss": 0.4901,
      "num_input_tokens_seen": 3882848,
      "step": 5995
    },
    {
      "epoch": 3.5377358490566038,
      "grad_norm": 2.353781223297119,
      "learning_rate": 9.821137760779797e-06,
      "loss": 0.4271,
      "num_input_tokens_seen": 3885312,
      "step": 6000
    },
    {
      "epoch": 3.540683962264151,
      "grad_norm": 1.3769530057907104,
      "learning_rate": 9.820455158118065e-06,
      "loss": 0.5173,
      "num_input_tokens_seen": 3892768,
      "step": 6005
    },
    {
      "epoch": 3.543632075471698,
      "grad_norm": 1.7759851217269897,
      "learning_rate": 9.819771279220053e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 3895168,
      "step": 6010
    },
    {
      "epoch": 3.546580188679245,
      "grad_norm": 1.799825668334961,
      "learning_rate": 9.819086124266825e-06,
      "loss": 0.5705,
      "num_input_tokens_seen": 3898112,
      "step": 6015
    },
    {
      "epoch": 3.5495283018867925,
      "grad_norm": 1.684653401374817,
      "learning_rate": 9.818399693439778e-06,
      "loss": 0.5889,
      "num_input_tokens_seen": 3900800,
      "step": 6020
    },
    {
      "epoch": 3.5524764150943398,
      "grad_norm": 2.592290163040161,
      "learning_rate": 9.817711986920644e-06,
      "loss": 0.4752,
      "num_input_tokens_seen": 3903904,
      "step": 6025
    },
    {
      "epoch": 3.555424528301887,
      "grad_norm": 2.7377283573150635,
      "learning_rate": 9.817023004891497e-06,
      "loss": 0.5578,
      "num_input_tokens_seen": 3906688,
      "step": 6030
    },
    {
      "epoch": 3.558372641509434,
      "grad_norm": 1.8757731914520264,
      "learning_rate": 9.816332747534752e-06,
      "loss": 0.4855,
      "num_input_tokens_seen": 3909408,
      "step": 6035
    },
    {
      "epoch": 3.561320754716981,
      "grad_norm": 2.2931602001190186,
      "learning_rate": 9.815641215033153e-06,
      "loss": 0.4991,
      "num_input_tokens_seen": 3912288,
      "step": 6040
    },
    {
      "epoch": 3.5642688679245285,
      "grad_norm": 1.4961023330688477,
      "learning_rate": 9.814948407569789e-06,
      "loss": 0.3295,
      "num_input_tokens_seen": 3915008,
      "step": 6045
    },
    {
      "epoch": 3.5672169811320753,
      "grad_norm": 1.7093939781188965,
      "learning_rate": 9.814254325328082e-06,
      "loss": 0.4991,
      "num_input_tokens_seen": 3918272,
      "step": 6050
    },
    {
      "epoch": 3.5701650943396226,
      "grad_norm": 3.028801679611206,
      "learning_rate": 9.813558968491794e-06,
      "loss": 0.532,
      "num_input_tokens_seen": 3921440,
      "step": 6055
    },
    {
      "epoch": 3.57311320754717,
      "grad_norm": 1.6072719097137451,
      "learning_rate": 9.812862337245024e-06,
      "loss": 0.5289,
      "num_input_tokens_seen": 3924480,
      "step": 6060
    },
    {
      "epoch": 3.576061320754717,
      "grad_norm": 2.3451802730560303,
      "learning_rate": 9.812164431772208e-06,
      "loss": 0.374,
      "num_input_tokens_seen": 3927232,
      "step": 6065
    },
    {
      "epoch": 3.579009433962264,
      "grad_norm": 0.8173542022705078,
      "learning_rate": 9.81146525225812e-06,
      "loss": 0.4632,
      "num_input_tokens_seen": 3930304,
      "step": 6070
    },
    {
      "epoch": 3.5819575471698113,
      "grad_norm": 1.9361385107040405,
      "learning_rate": 9.810764798887868e-06,
      "loss": 0.5122,
      "num_input_tokens_seen": 3933152,
      "step": 6075
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 1.5390702486038208,
      "learning_rate": 9.810063071846905e-06,
      "loss": 0.4071,
      "num_input_tokens_seen": 3936416,
      "step": 6080
    },
    {
      "epoch": 3.5878537735849054,
      "grad_norm": 6.409247875213623,
      "learning_rate": 9.809360071321013e-06,
      "loss": 0.4468,
      "num_input_tokens_seen": 3939840,
      "step": 6085
    },
    {
      "epoch": 3.5908018867924527,
      "grad_norm": 1.1480450630187988,
      "learning_rate": 9.808655797496314e-06,
      "loss": 0.582,
      "num_input_tokens_seen": 3943552,
      "step": 6090
    },
    {
      "epoch": 3.59375,
      "grad_norm": 2.425117254257202,
      "learning_rate": 9.807950250559268e-06,
      "loss": 0.4549,
      "num_input_tokens_seen": 3946976,
      "step": 6095
    },
    {
      "epoch": 3.5966981132075473,
      "grad_norm": 1.615707516670227,
      "learning_rate": 9.807243430696673e-06,
      "loss": 0.4793,
      "num_input_tokens_seen": 3950208,
      "step": 6100
    },
    {
      "epoch": 3.5996462264150946,
      "grad_norm": 2.107515335083008,
      "learning_rate": 9.806535338095661e-06,
      "loss": 0.4244,
      "num_input_tokens_seen": 3952576,
      "step": 6105
    },
    {
      "epoch": 3.6025943396226414,
      "grad_norm": 1.871767282485962,
      "learning_rate": 9.805825972943706e-06,
      "loss": 0.4756,
      "num_input_tokens_seen": 3956800,
      "step": 6110
    },
    {
      "epoch": 3.6055424528301887,
      "grad_norm": 2.161963701248169,
      "learning_rate": 9.80511533542861e-06,
      "loss": 0.4378,
      "num_input_tokens_seen": 3960192,
      "step": 6115
    },
    {
      "epoch": 3.608490566037736,
      "grad_norm": 1.5533031225204468,
      "learning_rate": 9.80440342573852e-06,
      "loss": 0.6779,
      "num_input_tokens_seen": 3962944,
      "step": 6120
    },
    {
      "epoch": 3.611438679245283,
      "grad_norm": 0.9668486714363098,
      "learning_rate": 9.803690244061919e-06,
      "loss": 0.4569,
      "num_input_tokens_seen": 3966112,
      "step": 6125
    },
    {
      "epoch": 3.61438679245283,
      "grad_norm": 1.1196751594543457,
      "learning_rate": 9.802975790587621e-06,
      "loss": 0.4201,
      "num_input_tokens_seen": 3969088,
      "step": 6130
    },
    {
      "epoch": 3.6173349056603774,
      "grad_norm": 4.276897430419922,
      "learning_rate": 9.802260065504783e-06,
      "loss": 0.4702,
      "num_input_tokens_seen": 3971712,
      "step": 6135
    },
    {
      "epoch": 3.6202830188679247,
      "grad_norm": 2.0963168144226074,
      "learning_rate": 9.801543069002897e-06,
      "loss": 0.5603,
      "num_input_tokens_seen": 3974752,
      "step": 6140
    },
    {
      "epoch": 3.6232311320754715,
      "grad_norm": 1.8043767213821411,
      "learning_rate": 9.80082480127179e-06,
      "loss": 0.5512,
      "num_input_tokens_seen": 3978176,
      "step": 6145
    },
    {
      "epoch": 3.626179245283019,
      "grad_norm": 1.880252480506897,
      "learning_rate": 9.800105262501628e-06,
      "loss": 0.3866,
      "num_input_tokens_seen": 3984736,
      "step": 6150
    },
    {
      "epoch": 3.629127358490566,
      "grad_norm": 2.3579792976379395,
      "learning_rate": 9.799384452882907e-06,
      "loss": 0.6604,
      "num_input_tokens_seen": 3988192,
      "step": 6155
    },
    {
      "epoch": 3.632075471698113,
      "grad_norm": 1.8629984855651855,
      "learning_rate": 9.798662372606469e-06,
      "loss": 0.5318,
      "num_input_tokens_seen": 3991360,
      "step": 6160
    },
    {
      "epoch": 3.6350235849056602,
      "grad_norm": 1.5217335224151611,
      "learning_rate": 9.797939021863487e-06,
      "loss": 0.529,
      "num_input_tokens_seen": 3994592,
      "step": 6165
    },
    {
      "epoch": 3.6379716981132075,
      "grad_norm": 1.5038437843322754,
      "learning_rate": 9.797214400845472e-06,
      "loss": 0.4798,
      "num_input_tokens_seen": 3998368,
      "step": 6170
    },
    {
      "epoch": 3.640919811320755,
      "grad_norm": 2.702582359313965,
      "learning_rate": 9.796488509744269e-06,
      "loss": 0.5561,
      "num_input_tokens_seen": 4000992,
      "step": 6175
    },
    {
      "epoch": 3.643867924528302,
      "grad_norm": 1.2524915933609009,
      "learning_rate": 9.79576134875206e-06,
      "loss": 0.4873,
      "num_input_tokens_seen": 4004864,
      "step": 6180
    },
    {
      "epoch": 3.646816037735849,
      "grad_norm": 1.857940673828125,
      "learning_rate": 9.795032918061367e-06,
      "loss": 0.4352,
      "num_input_tokens_seen": 4008768,
      "step": 6185
    },
    {
      "epoch": 3.6497641509433962,
      "grad_norm": 3.7594597339630127,
      "learning_rate": 9.794303217865041e-06,
      "loss": 0.5952,
      "num_input_tokens_seen": 4012352,
      "step": 6190
    },
    {
      "epoch": 3.6527122641509435,
      "grad_norm": 1.0669738054275513,
      "learning_rate": 9.79357224835628e-06,
      "loss": 0.4572,
      "num_input_tokens_seen": 4015872,
      "step": 6195
    },
    {
      "epoch": 3.6556603773584904,
      "grad_norm": 2.8727505207061768,
      "learning_rate": 9.792840009728605e-06,
      "loss": 0.4506,
      "num_input_tokens_seen": 4018784,
      "step": 6200
    },
    {
      "epoch": 3.6586084905660377,
      "grad_norm": 1.580567717552185,
      "learning_rate": 9.79210650217588e-06,
      "loss": 0.4596,
      "num_input_tokens_seen": 4021664,
      "step": 6205
    },
    {
      "epoch": 3.661556603773585,
      "grad_norm": 3.3019986152648926,
      "learning_rate": 9.791371725892307e-06,
      "loss": 0.4957,
      "num_input_tokens_seen": 4024608,
      "step": 6210
    },
    {
      "epoch": 3.6645047169811322,
      "grad_norm": 1.528113603591919,
      "learning_rate": 9.79063568107242e-06,
      "loss": 0.4616,
      "num_input_tokens_seen": 4028096,
      "step": 6215
    },
    {
      "epoch": 3.6674528301886795,
      "grad_norm": 1.8987077474594116,
      "learning_rate": 9.78989836791109e-06,
      "loss": 0.4325,
      "num_input_tokens_seen": 4030912,
      "step": 6220
    },
    {
      "epoch": 3.6704009433962264,
      "grad_norm": 3.2653772830963135,
      "learning_rate": 9.789159786603524e-06,
      "loss": 0.4884,
      "num_input_tokens_seen": 4033856,
      "step": 6225
    },
    {
      "epoch": 3.6733490566037736,
      "grad_norm": 1.7995320558547974,
      "learning_rate": 9.788419937345263e-06,
      "loss": 0.5701,
      "num_input_tokens_seen": 4037152,
      "step": 6230
    },
    {
      "epoch": 3.6762971698113205,
      "grad_norm": 1.3462026119232178,
      "learning_rate": 9.787678820332188e-06,
      "loss": 0.5166,
      "num_input_tokens_seen": 4040160,
      "step": 6235
    },
    {
      "epoch": 3.6792452830188678,
      "grad_norm": 1.7173432111740112,
      "learning_rate": 9.78693643576051e-06,
      "loss": 0.3783,
      "num_input_tokens_seen": 4043712,
      "step": 6240
    },
    {
      "epoch": 3.682193396226415,
      "grad_norm": 2.4574198722839355,
      "learning_rate": 9.786192783826782e-06,
      "loss": 0.4336,
      "num_input_tokens_seen": 4046400,
      "step": 6245
    },
    {
      "epoch": 3.6851415094339623,
      "grad_norm": 1.7036266326904297,
      "learning_rate": 9.785447864727887e-06,
      "loss": 0.4464,
      "num_input_tokens_seen": 4049344,
      "step": 6250
    },
    {
      "epoch": 3.6880896226415096,
      "grad_norm": 3.591498374938965,
      "learning_rate": 9.784701678661045e-06,
      "loss": 0.4167,
      "num_input_tokens_seen": 4051904,
      "step": 6255
    },
    {
      "epoch": 3.6910377358490565,
      "grad_norm": 1.9236555099487305,
      "learning_rate": 9.783954225823813e-06,
      "loss": 0.4074,
      "num_input_tokens_seen": 4056544,
      "step": 6260
    },
    {
      "epoch": 3.6939858490566038,
      "grad_norm": 1.823262333869934,
      "learning_rate": 9.783205506414082e-06,
      "loss": 0.4619,
      "num_input_tokens_seen": 4059904,
      "step": 6265
    },
    {
      "epoch": 3.696933962264151,
      "grad_norm": 1.145943522453308,
      "learning_rate": 9.782455520630079e-06,
      "loss": 0.3843,
      "num_input_tokens_seen": 4063008,
      "step": 6270
    },
    {
      "epoch": 3.699882075471698,
      "grad_norm": 1.5212962627410889,
      "learning_rate": 9.781704268670364e-06,
      "loss": 0.4875,
      "num_input_tokens_seen": 4066592,
      "step": 6275
    },
    {
      "epoch": 3.702830188679245,
      "grad_norm": 2.6717371940612793,
      "learning_rate": 9.780951750733837e-06,
      "loss": 0.4308,
      "num_input_tokens_seen": 4070400,
      "step": 6280
    },
    {
      "epoch": 3.7057783018867925,
      "grad_norm": 3.7853822708129883,
      "learning_rate": 9.780197967019728e-06,
      "loss": 0.947,
      "num_input_tokens_seen": 4073024,
      "step": 6285
    },
    {
      "epoch": 3.7087264150943398,
      "grad_norm": 2.667567491531372,
      "learning_rate": 9.779442917727608e-06,
      "loss": 0.513,
      "num_input_tokens_seen": 4076960,
      "step": 6290
    },
    {
      "epoch": 3.711674528301887,
      "grad_norm": 1.2131506204605103,
      "learning_rate": 9.778686603057377e-06,
      "loss": 0.5146,
      "num_input_tokens_seen": 4080192,
      "step": 6295
    },
    {
      "epoch": 3.714622641509434,
      "grad_norm": 2.547410488128662,
      "learning_rate": 9.777929023209271e-06,
      "loss": 0.54,
      "num_input_tokens_seen": 4083648,
      "step": 6300
    },
    {
      "epoch": 3.717570754716981,
      "grad_norm": 1.6496330499649048,
      "learning_rate": 9.777170178383866e-06,
      "loss": 0.4153,
      "num_input_tokens_seen": 4086912,
      "step": 6305
    },
    {
      "epoch": 3.7205188679245285,
      "grad_norm": 1.992221713066101,
      "learning_rate": 9.776410068782068e-06,
      "loss": 0.5538,
      "num_input_tokens_seen": 4089952,
      "step": 6310
    },
    {
      "epoch": 3.7234669811320753,
      "grad_norm": 1.5100078582763672,
      "learning_rate": 9.775648694605118e-06,
      "loss": 0.5112,
      "num_input_tokens_seen": 4092832,
      "step": 6315
    },
    {
      "epoch": 3.7264150943396226,
      "grad_norm": 1.653036117553711,
      "learning_rate": 9.774886056054593e-06,
      "loss": 0.513,
      "num_input_tokens_seen": 4096096,
      "step": 6320
    },
    {
      "epoch": 3.72936320754717,
      "grad_norm": 1.3575177192687988,
      "learning_rate": 9.774122153332408e-06,
      "loss": 0.5656,
      "num_input_tokens_seen": 4099392,
      "step": 6325
    },
    {
      "epoch": 3.732311320754717,
      "grad_norm": 1.742328405380249,
      "learning_rate": 9.773356986640807e-06,
      "loss": 0.4177,
      "num_input_tokens_seen": 4102208,
      "step": 6330
    },
    {
      "epoch": 3.735259433962264,
      "grad_norm": 1.3526649475097656,
      "learning_rate": 9.772590556182373e-06,
      "loss": 0.4901,
      "num_input_tokens_seen": 4105536,
      "step": 6335
    },
    {
      "epoch": 3.7382075471698113,
      "grad_norm": 2.335632801055908,
      "learning_rate": 9.77182286216002e-06,
      "loss": 0.3595,
      "num_input_tokens_seen": 4107872,
      "step": 6340
    },
    {
      "epoch": 3.7411556603773586,
      "grad_norm": 2.9306087493896484,
      "learning_rate": 9.771053904776998e-06,
      "loss": 0.5386,
      "num_input_tokens_seen": 4112224,
      "step": 6345
    },
    {
      "epoch": 3.7441037735849054,
      "grad_norm": 1.5365896224975586,
      "learning_rate": 9.770283684236891e-06,
      "loss": 0.3396,
      "num_input_tokens_seen": 4114976,
      "step": 6350
    },
    {
      "epoch": 3.7470518867924527,
      "grad_norm": 2.7016565799713135,
      "learning_rate": 9.769512200743623e-06,
      "loss": 0.485,
      "num_input_tokens_seen": 4118016,
      "step": 6355
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.7176512479782104,
      "learning_rate": 9.768739454501444e-06,
      "loss": 0.3437,
      "num_input_tokens_seen": 4120864,
      "step": 6360
    },
    {
      "epoch": 3.7529481132075473,
      "grad_norm": 1.3826419115066528,
      "learning_rate": 9.76796544571494e-06,
      "loss": 0.5654,
      "num_input_tokens_seen": 4123744,
      "step": 6365
    },
    {
      "epoch": 3.7558962264150946,
      "grad_norm": 1.5360344648361206,
      "learning_rate": 9.767190174589036e-06,
      "loss": 0.5866,
      "num_input_tokens_seen": 4127840,
      "step": 6370
    },
    {
      "epoch": 3.7588443396226414,
      "grad_norm": 1.5367670059204102,
      "learning_rate": 9.76641364132899e-06,
      "loss": 0.5143,
      "num_input_tokens_seen": 4130816,
      "step": 6375
    },
    {
      "epoch": 3.7617924528301887,
      "grad_norm": 1.238193154335022,
      "learning_rate": 9.765635846140389e-06,
      "loss": 0.5089,
      "num_input_tokens_seen": 4134048,
      "step": 6380
    },
    {
      "epoch": 3.764740566037736,
      "grad_norm": 2.0220513343811035,
      "learning_rate": 9.764856789229157e-06,
      "loss": 0.577,
      "num_input_tokens_seen": 4136352,
      "step": 6385
    },
    {
      "epoch": 3.767688679245283,
      "grad_norm": 1.29502272605896,
      "learning_rate": 9.764076470801557e-06,
      "loss": 0.4982,
      "num_input_tokens_seen": 4139520,
      "step": 6390
    },
    {
      "epoch": 3.77063679245283,
      "grad_norm": 1.8571369647979736,
      "learning_rate": 9.763294891064182e-06,
      "loss": 0.4272,
      "num_input_tokens_seen": 4142880,
      "step": 6395
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 1.571900486946106,
      "learning_rate": 9.762512050223951e-06,
      "loss": 0.4168,
      "num_input_tokens_seen": 4146208,
      "step": 6400
    },
    {
      "epoch": 3.7765330188679247,
      "grad_norm": 2.0116019248962402,
      "learning_rate": 9.761727948488132e-06,
      "loss": 0.6065,
      "num_input_tokens_seen": 4149952,
      "step": 6405
    },
    {
      "epoch": 3.7794811320754715,
      "grad_norm": 2.9846463203430176,
      "learning_rate": 9.760942586064315e-06,
      "loss": 0.5186,
      "num_input_tokens_seen": 4153472,
      "step": 6410
    },
    {
      "epoch": 3.782429245283019,
      "grad_norm": 1.6074416637420654,
      "learning_rate": 9.760155963160431e-06,
      "loss": 0.5088,
      "num_input_tokens_seen": 4156544,
      "step": 6415
    },
    {
      "epoch": 3.785377358490566,
      "grad_norm": 1.9176452159881592,
      "learning_rate": 9.759368079984741e-06,
      "loss": 0.4798,
      "num_input_tokens_seen": 4159552,
      "step": 6420
    },
    {
      "epoch": 3.788325471698113,
      "grad_norm": 2.3324851989746094,
      "learning_rate": 9.758578936745839e-06,
      "loss": 0.5358,
      "num_input_tokens_seen": 4162976,
      "step": 6425
    },
    {
      "epoch": 3.7912735849056602,
      "grad_norm": 4.24580717086792,
      "learning_rate": 9.757788533652656e-06,
      "loss": 0.4832,
      "num_input_tokens_seen": 4165632,
      "step": 6430
    },
    {
      "epoch": 3.7942216981132075,
      "grad_norm": 1.6063214540481567,
      "learning_rate": 9.756996870914454e-06,
      "loss": 0.7236,
      "num_input_tokens_seen": 4168672,
      "step": 6435
    },
    {
      "epoch": 3.797169811320755,
      "grad_norm": 1.9359186887741089,
      "learning_rate": 9.756203948740828e-06,
      "loss": 0.5171,
      "num_input_tokens_seen": 4172416,
      "step": 6440
    },
    {
      "epoch": 3.800117924528302,
      "grad_norm": 2.55765438079834,
      "learning_rate": 9.755409767341709e-06,
      "loss": 0.4446,
      "num_input_tokens_seen": 4175360,
      "step": 6445
    },
    {
      "epoch": 3.803066037735849,
      "grad_norm": 1.3530970811843872,
      "learning_rate": 9.75461432692736e-06,
      "loss": 0.5474,
      "num_input_tokens_seen": 4178080,
      "step": 6450
    },
    {
      "epoch": 3.8060141509433962,
      "grad_norm": 2.2595372200012207,
      "learning_rate": 9.753817627708375e-06,
      "loss": 0.4336,
      "num_input_tokens_seen": 4180608,
      "step": 6455
    },
    {
      "epoch": 3.8089622641509435,
      "grad_norm": 1.5271704196929932,
      "learning_rate": 9.753019669895686e-06,
      "loss": 0.4139,
      "num_input_tokens_seen": 4184288,
      "step": 6460
    },
    {
      "epoch": 3.8119103773584904,
      "grad_norm": 1.8866549730300903,
      "learning_rate": 9.752220453700556e-06,
      "loss": 0.3792,
      "num_input_tokens_seen": 4187072,
      "step": 6465
    },
    {
      "epoch": 3.8148584905660377,
      "grad_norm": 3.7236785888671875,
      "learning_rate": 9.75141997933458e-06,
      "loss": 0.4856,
      "num_input_tokens_seen": 4190368,
      "step": 6470
    },
    {
      "epoch": 3.817806603773585,
      "grad_norm": 3.1844546794891357,
      "learning_rate": 9.750618247009685e-06,
      "loss": 0.6154,
      "num_input_tokens_seen": 4194080,
      "step": 6475
    },
    {
      "epoch": 3.8207547169811322,
      "grad_norm": 1.4923067092895508,
      "learning_rate": 9.749815256938138e-06,
      "loss": 0.4674,
      "num_input_tokens_seen": 4196864,
      "step": 6480
    },
    {
      "epoch": 3.8237028301886795,
      "grad_norm": 2.7273061275482178,
      "learning_rate": 9.749011009332529e-06,
      "loss": 0.411,
      "num_input_tokens_seen": 4200224,
      "step": 6485
    },
    {
      "epoch": 3.8266509433962264,
      "grad_norm": 1.0455515384674072,
      "learning_rate": 9.748205504405787e-06,
      "loss": 0.4397,
      "num_input_tokens_seen": 4203328,
      "step": 6490
    },
    {
      "epoch": 3.8295990566037736,
      "grad_norm": 1.6606463193893433,
      "learning_rate": 9.747398742371177e-06,
      "loss": 0.3786,
      "num_input_tokens_seen": 4206080,
      "step": 6495
    },
    {
      "epoch": 3.8325471698113205,
      "grad_norm": 3.2325146198272705,
      "learning_rate": 9.746590723442289e-06,
      "loss": 0.572,
      "num_input_tokens_seen": 4209504,
      "step": 6500
    },
    {
      "epoch": 3.8354952830188678,
      "grad_norm": 2.8882548809051514,
      "learning_rate": 9.745781447833049e-06,
      "loss": 0.5103,
      "num_input_tokens_seen": 4213056,
      "step": 6505
    },
    {
      "epoch": 3.838443396226415,
      "grad_norm": 1.5680185556411743,
      "learning_rate": 9.74497091575772e-06,
      "loss": 0.5141,
      "num_input_tokens_seen": 4217792,
      "step": 6510
    },
    {
      "epoch": 3.8413915094339623,
      "grad_norm": 2.075716257095337,
      "learning_rate": 9.744159127430888e-06,
      "loss": 0.5331,
      "num_input_tokens_seen": 4220608,
      "step": 6515
    },
    {
      "epoch": 3.8443396226415096,
      "grad_norm": 5.686759948730469,
      "learning_rate": 9.743346083067482e-06,
      "loss": 0.6582,
      "num_input_tokens_seen": 4223136,
      "step": 6520
    },
    {
      "epoch": 3.8472877358490565,
      "grad_norm": 2.042863607406616,
      "learning_rate": 9.742531782882758e-06,
      "loss": 0.5323,
      "num_input_tokens_seen": 4226496,
      "step": 6525
    },
    {
      "epoch": 3.8502358490566038,
      "grad_norm": 1.4008901119232178,
      "learning_rate": 9.741716227092305e-06,
      "loss": 0.393,
      "num_input_tokens_seen": 4230272,
      "step": 6530
    },
    {
      "epoch": 3.853183962264151,
      "grad_norm": 1.7817617654800415,
      "learning_rate": 9.740899415912048e-06,
      "loss": 0.5422,
      "num_input_tokens_seen": 4233568,
      "step": 6535
    },
    {
      "epoch": 3.856132075471698,
      "grad_norm": 1.4794615507125854,
      "learning_rate": 9.740081349558236e-06,
      "loss": 0.3238,
      "num_input_tokens_seen": 4236160,
      "step": 6540
    },
    {
      "epoch": 3.859080188679245,
      "grad_norm": 1.5089831352233887,
      "learning_rate": 9.739262028247459e-06,
      "loss": 0.4991,
      "num_input_tokens_seen": 4240224,
      "step": 6545
    },
    {
      "epoch": 3.8620283018867925,
      "grad_norm": 2.0099644660949707,
      "learning_rate": 9.738441452196633e-06,
      "loss": 0.3429,
      "num_input_tokens_seen": 4243936,
      "step": 6550
    },
    {
      "epoch": 3.8649764150943398,
      "grad_norm": 7.151350021362305,
      "learning_rate": 9.737619621623013e-06,
      "loss": 0.4667,
      "num_input_tokens_seen": 4247392,
      "step": 6555
    },
    {
      "epoch": 3.867924528301887,
      "grad_norm": 1.7722485065460205,
      "learning_rate": 9.73679653674418e-06,
      "loss": 0.3913,
      "num_input_tokens_seen": 4250464,
      "step": 6560
    },
    {
      "epoch": 3.870872641509434,
      "grad_norm": 3.0613813400268555,
      "learning_rate": 9.735972197778047e-06,
      "loss": 0.3519,
      "num_input_tokens_seen": 4252736,
      "step": 6565
    },
    {
      "epoch": 3.873820754716981,
      "grad_norm": 1.268274188041687,
      "learning_rate": 9.735146604942867e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 4256384,
      "step": 6570
    },
    {
      "epoch": 3.8767688679245285,
      "grad_norm": 1.786982536315918,
      "learning_rate": 9.734319758457214e-06,
      "loss": 0.4372,
      "num_input_tokens_seen": 4259680,
      "step": 6575
    },
    {
      "epoch": 3.8797169811320753,
      "grad_norm": 2.8540430068969727,
      "learning_rate": 9.733491658540001e-06,
      "loss": 0.5211,
      "num_input_tokens_seen": 4262528,
      "step": 6580
    },
    {
      "epoch": 3.8826650943396226,
      "grad_norm": 3.382352352142334,
      "learning_rate": 9.732662305410474e-06,
      "loss": 0.4548,
      "num_input_tokens_seen": 4264896,
      "step": 6585
    },
    {
      "epoch": 3.88561320754717,
      "grad_norm": 2.9004528522491455,
      "learning_rate": 9.731831699288203e-06,
      "loss": 0.4422,
      "num_input_tokens_seen": 4267680,
      "step": 6590
    },
    {
      "epoch": 3.888561320754717,
      "grad_norm": 1.9682813882827759,
      "learning_rate": 9.730999840393096e-06,
      "loss": 0.4622,
      "num_input_tokens_seen": 4270528,
      "step": 6595
    },
    {
      "epoch": 3.891509433962264,
      "grad_norm": 4.559301853179932,
      "learning_rate": 9.730166728945391e-06,
      "loss": 0.4506,
      "num_input_tokens_seen": 4274368,
      "step": 6600
    },
    {
      "epoch": 3.8944575471698113,
      "grad_norm": 1.5222504138946533,
      "learning_rate": 9.72933236516566e-06,
      "loss": 0.6306,
      "num_input_tokens_seen": 4277504,
      "step": 6605
    },
    {
      "epoch": 3.8974056603773586,
      "grad_norm": 1.3649005889892578,
      "learning_rate": 9.728496749274806e-06,
      "loss": 0.416,
      "num_input_tokens_seen": 4280640,
      "step": 6610
    },
    {
      "epoch": 3.9003537735849054,
      "grad_norm": 1.4548304080963135,
      "learning_rate": 9.727659881494054e-06,
      "loss": 0.3906,
      "num_input_tokens_seen": 4284736,
      "step": 6615
    },
    {
      "epoch": 3.9033018867924527,
      "grad_norm": 1.2267910242080688,
      "learning_rate": 9.726821762044975e-06,
      "loss": 0.4456,
      "num_input_tokens_seen": 4287616,
      "step": 6620
    },
    {
      "epoch": 3.90625,
      "grad_norm": 1.856896996498108,
      "learning_rate": 9.725982391149465e-06,
      "loss": 0.4997,
      "num_input_tokens_seen": 4290176,
      "step": 6625
    },
    {
      "epoch": 3.9091981132075473,
      "grad_norm": 1.7604360580444336,
      "learning_rate": 9.725141769029747e-06,
      "loss": 0.4563,
      "num_input_tokens_seen": 4293664,
      "step": 6630
    },
    {
      "epoch": 3.9121462264150946,
      "grad_norm": 2.1496195793151855,
      "learning_rate": 9.72429989590838e-06,
      "loss": 0.5332,
      "num_input_tokens_seen": 4296384,
      "step": 6635
    },
    {
      "epoch": 3.9150943396226414,
      "grad_norm": 1.8641494512557983,
      "learning_rate": 9.723456772008257e-06,
      "loss": 0.5098,
      "num_input_tokens_seen": 4299424,
      "step": 6640
    },
    {
      "epoch": 3.9180424528301887,
      "grad_norm": 1.8206076622009277,
      "learning_rate": 9.722612397552598e-06,
      "loss": 0.5602,
      "num_input_tokens_seen": 4303200,
      "step": 6645
    },
    {
      "epoch": 3.920990566037736,
      "grad_norm": 1.431689739227295,
      "learning_rate": 9.72176677276495e-06,
      "loss": 0.3824,
      "num_input_tokens_seen": 4306624,
      "step": 6650
    },
    {
      "epoch": 3.923938679245283,
      "grad_norm": 3.1154963970184326,
      "learning_rate": 9.7209198978692e-06,
      "loss": 0.5918,
      "num_input_tokens_seen": 4309344,
      "step": 6655
    },
    {
      "epoch": 3.92688679245283,
      "grad_norm": 1.9133939743041992,
      "learning_rate": 9.720071773089564e-06,
      "loss": 0.6328,
      "num_input_tokens_seen": 4313312,
      "step": 6660
    },
    {
      "epoch": 3.9298349056603774,
      "grad_norm": 2.8461544513702393,
      "learning_rate": 9.71922239865058e-06,
      "loss": 0.4219,
      "num_input_tokens_seen": 4316064,
      "step": 6665
    },
    {
      "epoch": 3.9327830188679247,
      "grad_norm": 3.026637315750122,
      "learning_rate": 9.718371774777131e-06,
      "loss": 0.6176,
      "num_input_tokens_seen": 4319168,
      "step": 6670
    },
    {
      "epoch": 3.9357311320754715,
      "grad_norm": 1.955230474472046,
      "learning_rate": 9.717519901694416e-06,
      "loss": 0.4927,
      "num_input_tokens_seen": 4322048,
      "step": 6675
    },
    {
      "epoch": 3.938679245283019,
      "grad_norm": 2.4277732372283936,
      "learning_rate": 9.716666779627978e-06,
      "loss": 0.5061,
      "num_input_tokens_seen": 4325056,
      "step": 6680
    },
    {
      "epoch": 3.941627358490566,
      "grad_norm": 1.1962436437606812,
      "learning_rate": 9.715812408803681e-06,
      "loss": 0.4777,
      "num_input_tokens_seen": 4328064,
      "step": 6685
    },
    {
      "epoch": 3.944575471698113,
      "grad_norm": 1.4843961000442505,
      "learning_rate": 9.714956789447726e-06,
      "loss": 0.3679,
      "num_input_tokens_seen": 4333568,
      "step": 6690
    },
    {
      "epoch": 3.9475235849056602,
      "grad_norm": 0.9601065516471863,
      "learning_rate": 9.71409992178664e-06,
      "loss": 0.4515,
      "num_input_tokens_seen": 4337056,
      "step": 6695
    },
    {
      "epoch": 3.9504716981132075,
      "grad_norm": 1.7847347259521484,
      "learning_rate": 9.713241806047282e-06,
      "loss": 0.6464,
      "num_input_tokens_seen": 4341408,
      "step": 6700
    },
    {
      "epoch": 3.953419811320755,
      "grad_norm": 1.7445002794265747,
      "learning_rate": 9.712382442456845e-06,
      "loss": 0.506,
      "num_input_tokens_seen": 4344320,
      "step": 6705
    },
    {
      "epoch": 3.956367924528302,
      "grad_norm": 1.4445054531097412,
      "learning_rate": 9.711521831242846e-06,
      "loss": 0.4787,
      "num_input_tokens_seen": 4346816,
      "step": 6710
    },
    {
      "epoch": 3.959316037735849,
      "grad_norm": 1.3715189695358276,
      "learning_rate": 9.710659972633137e-06,
      "loss": 0.4754,
      "num_input_tokens_seen": 4349344,
      "step": 6715
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 1.4292443990707397,
      "learning_rate": 9.709796866855899e-06,
      "loss": 0.5114,
      "num_input_tokens_seen": 4352480,
      "step": 6720
    },
    {
      "epoch": 3.9652122641509435,
      "grad_norm": 2.624915838241577,
      "learning_rate": 9.70893251413964e-06,
      "loss": 0.5724,
      "num_input_tokens_seen": 4355712,
      "step": 6725
    },
    {
      "epoch": 3.9681603773584904,
      "grad_norm": 3.1062917709350586,
      "learning_rate": 9.708066914713205e-06,
      "loss": 0.4123,
      "num_input_tokens_seen": 4361472,
      "step": 6730
    },
    {
      "epoch": 3.9711084905660377,
      "grad_norm": 1.7217196226119995,
      "learning_rate": 9.707200068805764e-06,
      "loss": 0.4529,
      "num_input_tokens_seen": 4364448,
      "step": 6735
    },
    {
      "epoch": 3.974056603773585,
      "grad_norm": 2.182603359222412,
      "learning_rate": 9.706331976646817e-06,
      "loss": 0.4613,
      "num_input_tokens_seen": 4367936,
      "step": 6740
    },
    {
      "epoch": 3.9770047169811322,
      "grad_norm": 2.4130778312683105,
      "learning_rate": 9.705462638466197e-06,
      "loss": 0.4913,
      "num_input_tokens_seen": 4370464,
      "step": 6745
    },
    {
      "epoch": 3.9799528301886795,
      "grad_norm": 1.815284013748169,
      "learning_rate": 9.704592054494065e-06,
      "loss": 0.426,
      "num_input_tokens_seen": 4373312,
      "step": 6750
    },
    {
      "epoch": 3.9829009433962264,
      "grad_norm": 1.5012723207473755,
      "learning_rate": 9.703720224960909e-06,
      "loss": 0.4549,
      "num_input_tokens_seen": 4376096,
      "step": 6755
    },
    {
      "epoch": 3.9858490566037736,
      "grad_norm": 1.7465699911117554,
      "learning_rate": 9.702847150097552e-06,
      "loss": 0.5038,
      "num_input_tokens_seen": 4378400,
      "step": 6760
    },
    {
      "epoch": 3.9887971698113205,
      "grad_norm": 1.1225504875183105,
      "learning_rate": 9.701972830135143e-06,
      "loss": 0.4792,
      "num_input_tokens_seen": 4381024,
      "step": 6765
    },
    {
      "epoch": 3.9917452830188678,
      "grad_norm": 1.626244306564331,
      "learning_rate": 9.701097265305164e-06,
      "loss": 0.4765,
      "num_input_tokens_seen": 4384672,
      "step": 6770
    },
    {
      "epoch": 3.994693396226415,
      "grad_norm": 1.26576828956604,
      "learning_rate": 9.700220455839422e-06,
      "loss": 0.4046,
      "num_input_tokens_seen": 4389312,
      "step": 6775
    },
    {
      "epoch": 3.9976415094339623,
      "grad_norm": 2.3714599609375,
      "learning_rate": 9.69934240197006e-06,
      "loss": 0.4641,
      "num_input_tokens_seen": 4391840,
      "step": 6780
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.5044654011726379,
      "eval_runtime": 19.5525,
      "eval_samples_per_second": 86.741,
      "eval_steps_per_second": 21.685,
      "num_input_tokens_seen": 4393784,
      "step": 6784
    },
    {
      "epoch": 4.00058962264151,
      "grad_norm": 1.8960009813308716,
      "learning_rate": 9.698463103929542e-06,
      "loss": 0.4924,
      "num_input_tokens_seen": 4394776,
      "step": 6785
    },
    {
      "epoch": 4.003537735849057,
      "grad_norm": 1.7191420793533325,
      "learning_rate": 9.697582561950669e-06,
      "loss": 0.4186,
      "num_input_tokens_seen": 4397176,
      "step": 6790
    },
    {
      "epoch": 4.006485849056604,
      "grad_norm": 1.9263477325439453,
      "learning_rate": 9.696700776266568e-06,
      "loss": 0.4368,
      "num_input_tokens_seen": 4400600,
      "step": 6795
    },
    {
      "epoch": 4.009433962264151,
      "grad_norm": 1.723443627357483,
      "learning_rate": 9.695817747110694e-06,
      "loss": 0.4399,
      "num_input_tokens_seen": 4403832,
      "step": 6800
    },
    {
      "epoch": 4.012382075471698,
      "grad_norm": 1.616361141204834,
      "learning_rate": 9.694933474716831e-06,
      "loss": 0.5603,
      "num_input_tokens_seen": 4407032,
      "step": 6805
    },
    {
      "epoch": 4.015330188679245,
      "grad_norm": 2.4167463779449463,
      "learning_rate": 9.6940479593191e-06,
      "loss": 0.5234,
      "num_input_tokens_seen": 4409624,
      "step": 6810
    },
    {
      "epoch": 4.0182783018867925,
      "grad_norm": 3.79972767829895,
      "learning_rate": 9.693161201151942e-06,
      "loss": 0.6186,
      "num_input_tokens_seen": 4411608,
      "step": 6815
    },
    {
      "epoch": 4.02122641509434,
      "grad_norm": 1.9145468473434448,
      "learning_rate": 9.692273200450128e-06,
      "loss": 0.4131,
      "num_input_tokens_seen": 4416152,
      "step": 6820
    },
    {
      "epoch": 4.024174528301887,
      "grad_norm": 2.1308188438415527,
      "learning_rate": 9.69138395744876e-06,
      "loss": 0.5655,
      "num_input_tokens_seen": 4419608,
      "step": 6825
    },
    {
      "epoch": 4.027122641509434,
      "grad_norm": 1.6613397598266602,
      "learning_rate": 9.690493472383274e-06,
      "loss": 0.6673,
      "num_input_tokens_seen": 4424024,
      "step": 6830
    },
    {
      "epoch": 4.030070754716981,
      "grad_norm": 1.1935826539993286,
      "learning_rate": 9.689601745489423e-06,
      "loss": 0.5432,
      "num_input_tokens_seen": 4427224,
      "step": 6835
    },
    {
      "epoch": 4.033018867924528,
      "grad_norm": 2.0552687644958496,
      "learning_rate": 9.6887087770033e-06,
      "loss": 0.2811,
      "num_input_tokens_seen": 4430392,
      "step": 6840
    },
    {
      "epoch": 4.035966981132075,
      "grad_norm": 1.5242539644241333,
      "learning_rate": 9.687814567161322e-06,
      "loss": 0.6101,
      "num_input_tokens_seen": 4433784,
      "step": 6845
    },
    {
      "epoch": 4.038915094339623,
      "grad_norm": 2.7782156467437744,
      "learning_rate": 9.686919116200232e-06,
      "loss": 0.5511,
      "num_input_tokens_seen": 4436152,
      "step": 6850
    },
    {
      "epoch": 4.04186320754717,
      "grad_norm": 1.9204893112182617,
      "learning_rate": 9.686022424357108e-06,
      "loss": 0.6128,
      "num_input_tokens_seen": 4440344,
      "step": 6855
    },
    {
      "epoch": 4.044811320754717,
      "grad_norm": 2.980475425720215,
      "learning_rate": 9.685124491869353e-06,
      "loss": 0.3997,
      "num_input_tokens_seen": 4443800,
      "step": 6860
    },
    {
      "epoch": 4.0477594339622645,
      "grad_norm": 2.1587772369384766,
      "learning_rate": 9.684225318974696e-06,
      "loss": 0.5155,
      "num_input_tokens_seen": 4446872,
      "step": 6865
    },
    {
      "epoch": 4.050707547169812,
      "grad_norm": 2.0021283626556396,
      "learning_rate": 9.683324905911197e-06,
      "loss": 0.5297,
      "num_input_tokens_seen": 4449944,
      "step": 6870
    },
    {
      "epoch": 4.053655660377358,
      "grad_norm": 1.8095752000808716,
      "learning_rate": 9.682423252917245e-06,
      "loss": 0.4301,
      "num_input_tokens_seen": 4453144,
      "step": 6875
    },
    {
      "epoch": 4.056603773584905,
      "grad_norm": 1.6005330085754395,
      "learning_rate": 9.681520360231557e-06,
      "loss": 0.4333,
      "num_input_tokens_seen": 4456376,
      "step": 6880
    },
    {
      "epoch": 4.059551886792453,
      "grad_norm": 2.4501426219940186,
      "learning_rate": 9.680616228093178e-06,
      "loss": 0.4722,
      "num_input_tokens_seen": 4459192,
      "step": 6885
    },
    {
      "epoch": 4.0625,
      "grad_norm": 1.9325991868972778,
      "learning_rate": 9.67971085674148e-06,
      "loss": 0.3204,
      "num_input_tokens_seen": 4462680,
      "step": 6890
    },
    {
      "epoch": 4.065448113207547,
      "grad_norm": 1.1685845851898193,
      "learning_rate": 9.678804246416164e-06,
      "loss": 0.5313,
      "num_input_tokens_seen": 4466584,
      "step": 6895
    },
    {
      "epoch": 4.068396226415095,
      "grad_norm": 1.71962571144104,
      "learning_rate": 9.677896397357259e-06,
      "loss": 0.3538,
      "num_input_tokens_seen": 4469144,
      "step": 6900
    },
    {
      "epoch": 4.071344339622642,
      "grad_norm": 1.138940453529358,
      "learning_rate": 9.676987309805121e-06,
      "loss": 0.4821,
      "num_input_tokens_seen": 4472504,
      "step": 6905
    },
    {
      "epoch": 4.074292452830188,
      "grad_norm": 2.0202205181121826,
      "learning_rate": 9.67607698400044e-06,
      "loss": 0.5017,
      "num_input_tokens_seen": 4475992,
      "step": 6910
    },
    {
      "epoch": 4.0772405660377355,
      "grad_norm": 3.115718126296997,
      "learning_rate": 9.67516542018422e-06,
      "loss": 0.3733,
      "num_input_tokens_seen": 4479096,
      "step": 6915
    },
    {
      "epoch": 4.080188679245283,
      "grad_norm": 2.2014999389648438,
      "learning_rate": 9.67425261859781e-06,
      "loss": 0.5393,
      "num_input_tokens_seen": 4483256,
      "step": 6920
    },
    {
      "epoch": 4.08313679245283,
      "grad_norm": 2.086264133453369,
      "learning_rate": 9.673338579482871e-06,
      "loss": 0.4186,
      "num_input_tokens_seen": 4486328,
      "step": 6925
    },
    {
      "epoch": 4.086084905660377,
      "grad_norm": 3.357290029525757,
      "learning_rate": 9.672423303081404e-06,
      "loss": 0.455,
      "num_input_tokens_seen": 4489688,
      "step": 6930
    },
    {
      "epoch": 4.089033018867925,
      "grad_norm": 1.8953956365585327,
      "learning_rate": 9.67150678963573e-06,
      "loss": 0.4466,
      "num_input_tokens_seen": 4492440,
      "step": 6935
    },
    {
      "epoch": 4.091981132075472,
      "grad_norm": 2.0223329067230225,
      "learning_rate": 9.670589039388501e-06,
      "loss": 0.3764,
      "num_input_tokens_seen": 4495416,
      "step": 6940
    },
    {
      "epoch": 4.094929245283019,
      "grad_norm": 1.5100703239440918,
      "learning_rate": 9.669670052582695e-06,
      "loss": 0.4313,
      "num_input_tokens_seen": 4498968,
      "step": 6945
    },
    {
      "epoch": 4.097877358490566,
      "grad_norm": 1.9401711225509644,
      "learning_rate": 9.668749829461617e-06,
      "loss": 0.4706,
      "num_input_tokens_seen": 4502072,
      "step": 6950
    },
    {
      "epoch": 4.100825471698113,
      "grad_norm": 1.9601502418518066,
      "learning_rate": 9.667828370268898e-06,
      "loss": 0.4298,
      "num_input_tokens_seen": 4506616,
      "step": 6955
    },
    {
      "epoch": 4.10377358490566,
      "grad_norm": 1.4643375873565674,
      "learning_rate": 9.666905675248505e-06,
      "loss": 0.444,
      "num_input_tokens_seen": 4510136,
      "step": 6960
    },
    {
      "epoch": 4.1067216981132075,
      "grad_norm": 1.9149112701416016,
      "learning_rate": 9.66598174464472e-06,
      "loss": 0.5714,
      "num_input_tokens_seen": 4513048,
      "step": 6965
    },
    {
      "epoch": 4.109669811320755,
      "grad_norm": 2.1186249256134033,
      "learning_rate": 9.665056578702157e-06,
      "loss": 0.4442,
      "num_input_tokens_seen": 4515992,
      "step": 6970
    },
    {
      "epoch": 4.112617924528302,
      "grad_norm": 1.8142814636230469,
      "learning_rate": 9.66413017766576e-06,
      "loss": 0.5312,
      "num_input_tokens_seen": 4518808,
      "step": 6975
    },
    {
      "epoch": 4.115566037735849,
      "grad_norm": 1.2643468379974365,
      "learning_rate": 9.663202541780799e-06,
      "loss": 0.4055,
      "num_input_tokens_seen": 4522296,
      "step": 6980
    },
    {
      "epoch": 4.118514150943396,
      "grad_norm": 1.7322078943252563,
      "learning_rate": 9.662273671292866e-06,
      "loss": 0.379,
      "num_input_tokens_seen": 4525048,
      "step": 6985
    },
    {
      "epoch": 4.121462264150943,
      "grad_norm": 2.230083703994751,
      "learning_rate": 9.661343566447886e-06,
      "loss": 0.3832,
      "num_input_tokens_seen": 4528120,
      "step": 6990
    },
    {
      "epoch": 4.12441037735849,
      "grad_norm": 1.5664139986038208,
      "learning_rate": 9.660412227492107e-06,
      "loss": 0.424,
      "num_input_tokens_seen": 4531544,
      "step": 6995
    },
    {
      "epoch": 4.127358490566038,
      "grad_norm": 1.4593831300735474,
      "learning_rate": 9.659479654672106e-06,
      "loss": 0.5496,
      "num_input_tokens_seen": 4534328,
      "step": 7000
    },
    {
      "epoch": 4.130306603773585,
      "grad_norm": 3.1449389457702637,
      "learning_rate": 9.658545848234784e-06,
      "loss": 0.415,
      "num_input_tokens_seen": 4537560,
      "step": 7005
    },
    {
      "epoch": 4.133254716981132,
      "grad_norm": 2.358154773712158,
      "learning_rate": 9.657610808427372e-06,
      "loss": 0.6456,
      "num_input_tokens_seen": 4541656,
      "step": 7010
    },
    {
      "epoch": 4.1362028301886795,
      "grad_norm": 1.3525947332382202,
      "learning_rate": 9.656674535497425e-06,
      "loss": 0.4374,
      "num_input_tokens_seen": 4544600,
      "step": 7015
    },
    {
      "epoch": 4.139150943396227,
      "grad_norm": 2.71488094329834,
      "learning_rate": 9.655737029692827e-06,
      "loss": 0.4637,
      "num_input_tokens_seen": 4547960,
      "step": 7020
    },
    {
      "epoch": 4.142099056603773,
      "grad_norm": 1.2057307958602905,
      "learning_rate": 9.654798291261785e-06,
      "loss": 0.4467,
      "num_input_tokens_seen": 4552088,
      "step": 7025
    },
    {
      "epoch": 4.1450471698113205,
      "grad_norm": 2.2159042358398438,
      "learning_rate": 9.653858320452833e-06,
      "loss": 0.3966,
      "num_input_tokens_seen": 4555608,
      "step": 7030
    },
    {
      "epoch": 4.147995283018868,
      "grad_norm": 1.660851001739502,
      "learning_rate": 9.652917117514836e-06,
      "loss": 0.3962,
      "num_input_tokens_seen": 4562424,
      "step": 7035
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 1.5226880311965942,
      "learning_rate": 9.651974682696975e-06,
      "loss": 0.4599,
      "num_input_tokens_seen": 4565496,
      "step": 7040
    },
    {
      "epoch": 4.153891509433962,
      "grad_norm": 2.0455563068389893,
      "learning_rate": 9.651031016248773e-06,
      "loss": 0.503,
      "num_input_tokens_seen": 4570136,
      "step": 7045
    },
    {
      "epoch": 4.15683962264151,
      "grad_norm": 1.9866580963134766,
      "learning_rate": 9.65008611842006e-06,
      "loss": 0.4408,
      "num_input_tokens_seen": 4573176,
      "step": 7050
    },
    {
      "epoch": 4.159787735849057,
      "grad_norm": 1.632804036140442,
      "learning_rate": 9.64913998946101e-06,
      "loss": 0.42,
      "num_input_tokens_seen": 4577048,
      "step": 7055
    },
    {
      "epoch": 4.162735849056604,
      "grad_norm": 2.786975383758545,
      "learning_rate": 9.648192629622109e-06,
      "loss": 0.4097,
      "num_input_tokens_seen": 4580152,
      "step": 7060
    },
    {
      "epoch": 4.165683962264151,
      "grad_norm": 1.9070000648498535,
      "learning_rate": 9.647244039154178e-06,
      "loss": 0.4921,
      "num_input_tokens_seen": 4583352,
      "step": 7065
    },
    {
      "epoch": 4.168632075471698,
      "grad_norm": 1.9420804977416992,
      "learning_rate": 9.64629421830836e-06,
      "loss": 0.3693,
      "num_input_tokens_seen": 4586168,
      "step": 7070
    },
    {
      "epoch": 4.171580188679245,
      "grad_norm": 4.734982967376709,
      "learning_rate": 9.64534316733612e-06,
      "loss": 0.4861,
      "num_input_tokens_seen": 4588504,
      "step": 7075
    },
    {
      "epoch": 4.1745283018867925,
      "grad_norm": 1.4429336786270142,
      "learning_rate": 9.644390886489258e-06,
      "loss": 0.5556,
      "num_input_tokens_seen": 4591512,
      "step": 7080
    },
    {
      "epoch": 4.17747641509434,
      "grad_norm": 1.2224425077438354,
      "learning_rate": 9.643437376019893e-06,
      "loss": 0.4395,
      "num_input_tokens_seen": 4595256,
      "step": 7085
    },
    {
      "epoch": 4.180424528301887,
      "grad_norm": 2.3365044593811035,
      "learning_rate": 9.64248263618047e-06,
      "loss": 0.4126,
      "num_input_tokens_seen": 4598424,
      "step": 7090
    },
    {
      "epoch": 4.183372641509434,
      "grad_norm": 2.056363821029663,
      "learning_rate": 9.64152666722376e-06,
      "loss": 0.4468,
      "num_input_tokens_seen": 4601720,
      "step": 7095
    },
    {
      "epoch": 4.186320754716981,
      "grad_norm": 1.8034464120864868,
      "learning_rate": 9.640569469402863e-06,
      "loss": 0.4578,
      "num_input_tokens_seen": 4604440,
      "step": 7100
    },
    {
      "epoch": 4.189268867924528,
      "grad_norm": 1.503327488899231,
      "learning_rate": 9.639611042971198e-06,
      "loss": 0.4154,
      "num_input_tokens_seen": 4607160,
      "step": 7105
    },
    {
      "epoch": 4.192216981132075,
      "grad_norm": 1.4929511547088623,
      "learning_rate": 9.638651388182514e-06,
      "loss": 0.4015,
      "num_input_tokens_seen": 4610040,
      "step": 7110
    },
    {
      "epoch": 4.195165094339623,
      "grad_norm": 1.6327120065689087,
      "learning_rate": 9.637690505290884e-06,
      "loss": 0.5095,
      "num_input_tokens_seen": 4614648,
      "step": 7115
    },
    {
      "epoch": 4.19811320754717,
      "grad_norm": 1.4670945405960083,
      "learning_rate": 9.636728394550705e-06,
      "loss": 0.4305,
      "num_input_tokens_seen": 4617048,
      "step": 7120
    },
    {
      "epoch": 4.201061320754717,
      "grad_norm": 3.1780879497528076,
      "learning_rate": 9.6357650562167e-06,
      "loss": 0.5954,
      "num_input_tokens_seen": 4620088,
      "step": 7125
    },
    {
      "epoch": 4.2040094339622645,
      "grad_norm": 1.9226174354553223,
      "learning_rate": 9.634800490543918e-06,
      "loss": 0.4122,
      "num_input_tokens_seen": 4623384,
      "step": 7130
    },
    {
      "epoch": 4.206957547169812,
      "grad_norm": 1.7896754741668701,
      "learning_rate": 9.633834697787731e-06,
      "loss": 0.4658,
      "num_input_tokens_seen": 4626776,
      "step": 7135
    },
    {
      "epoch": 4.209905660377358,
      "grad_norm": 4.949166774749756,
      "learning_rate": 9.632867678203836e-06,
      "loss": 0.4873,
      "num_input_tokens_seen": 4629336,
      "step": 7140
    },
    {
      "epoch": 4.212853773584905,
      "grad_norm": 3.7040059566497803,
      "learning_rate": 9.631899432048258e-06,
      "loss": 0.39,
      "num_input_tokens_seen": 4631608,
      "step": 7145
    },
    {
      "epoch": 4.215801886792453,
      "grad_norm": 1.5375421047210693,
      "learning_rate": 9.630929959577343e-06,
      "loss": 0.3576,
      "num_input_tokens_seen": 4634488,
      "step": 7150
    },
    {
      "epoch": 4.21875,
      "grad_norm": 1.6688631772994995,
      "learning_rate": 9.629959261047764e-06,
      "loss": 0.3634,
      "num_input_tokens_seen": 4637816,
      "step": 7155
    },
    {
      "epoch": 4.221698113207547,
      "grad_norm": 2.2813541889190674,
      "learning_rate": 9.628987336716513e-06,
      "loss": 0.3255,
      "num_input_tokens_seen": 4640600,
      "step": 7160
    },
    {
      "epoch": 4.224646226415095,
      "grad_norm": 1.7395983934402466,
      "learning_rate": 9.628014186840918e-06,
      "loss": 0.5205,
      "num_input_tokens_seen": 4643384,
      "step": 7165
    },
    {
      "epoch": 4.227594339622642,
      "grad_norm": 2.593214511871338,
      "learning_rate": 9.62703981167862e-06,
      "loss": 0.3616,
      "num_input_tokens_seen": 4647064,
      "step": 7170
    },
    {
      "epoch": 4.230542452830188,
      "grad_norm": 2.103597640991211,
      "learning_rate": 9.626064211487592e-06,
      "loss": 0.4426,
      "num_input_tokens_seen": 4650040,
      "step": 7175
    },
    {
      "epoch": 4.2334905660377355,
      "grad_norm": 2.2585341930389404,
      "learning_rate": 9.625087386526125e-06,
      "loss": 0.385,
      "num_input_tokens_seen": 4653272,
      "step": 7180
    },
    {
      "epoch": 4.236438679245283,
      "grad_norm": 2.754847764968872,
      "learning_rate": 9.624109337052839e-06,
      "loss": 0.5728,
      "num_input_tokens_seen": 4656696,
      "step": 7185
    },
    {
      "epoch": 4.23938679245283,
      "grad_norm": 1.2481683492660522,
      "learning_rate": 9.623130063326678e-06,
      "loss": 0.3768,
      "num_input_tokens_seen": 4660408,
      "step": 7190
    },
    {
      "epoch": 4.242334905660377,
      "grad_norm": 1.9704242944717407,
      "learning_rate": 9.622149565606909e-06,
      "loss": 0.4171,
      "num_input_tokens_seen": 4664440,
      "step": 7195
    },
    {
      "epoch": 4.245283018867925,
      "grad_norm": 1.770379900932312,
      "learning_rate": 9.621167844153122e-06,
      "loss": 0.4296,
      "num_input_tokens_seen": 4667928,
      "step": 7200
    },
    {
      "epoch": 4.248231132075472,
      "grad_norm": 1.8996632099151611,
      "learning_rate": 9.620184899225231e-06,
      "loss": 0.4818,
      "num_input_tokens_seen": 4671320,
      "step": 7205
    },
    {
      "epoch": 4.251179245283019,
      "grad_norm": 0.9687104225158691,
      "learning_rate": 9.619200731083477e-06,
      "loss": 0.3576,
      "num_input_tokens_seen": 4674424,
      "step": 7210
    },
    {
      "epoch": 4.254127358490566,
      "grad_norm": 1.376307487487793,
      "learning_rate": 9.618215339988422e-06,
      "loss": 0.5534,
      "num_input_tokens_seen": 4681208,
      "step": 7215
    },
    {
      "epoch": 4.257075471698113,
      "grad_norm": 4.639444351196289,
      "learning_rate": 9.617228726200951e-06,
      "loss": 0.545,
      "num_input_tokens_seen": 4684216,
      "step": 7220
    },
    {
      "epoch": 4.26002358490566,
      "grad_norm": 3.300452709197998,
      "learning_rate": 9.616240889982277e-06,
      "loss": 0.4163,
      "num_input_tokens_seen": 4687160,
      "step": 7225
    },
    {
      "epoch": 4.2629716981132075,
      "grad_norm": 2.8267719745635986,
      "learning_rate": 9.61525183159393e-06,
      "loss": 0.6327,
      "num_input_tokens_seen": 4689400,
      "step": 7230
    },
    {
      "epoch": 4.265919811320755,
      "grad_norm": 1.7364627122879028,
      "learning_rate": 9.614261551297774e-06,
      "loss": 0.4312,
      "num_input_tokens_seen": 4692184,
      "step": 7235
    },
    {
      "epoch": 4.268867924528302,
      "grad_norm": 2.1914637088775635,
      "learning_rate": 9.613270049355983e-06,
      "loss": 0.4191,
      "num_input_tokens_seen": 4694872,
      "step": 7240
    },
    {
      "epoch": 4.271816037735849,
      "grad_norm": 1.2099653482437134,
      "learning_rate": 9.612277326031065e-06,
      "loss": 0.5514,
      "num_input_tokens_seen": 4697880,
      "step": 7245
    },
    {
      "epoch": 4.274764150943396,
      "grad_norm": 4.182252883911133,
      "learning_rate": 9.611283381585848e-06,
      "loss": 0.4871,
      "num_input_tokens_seen": 4701272,
      "step": 7250
    },
    {
      "epoch": 4.277712264150943,
      "grad_norm": 2.3102054595947266,
      "learning_rate": 9.61028821628348e-06,
      "loss": 0.4649,
      "num_input_tokens_seen": 4703992,
      "step": 7255
    },
    {
      "epoch": 4.28066037735849,
      "grad_norm": 5.551423072814941,
      "learning_rate": 9.609291830387439e-06,
      "loss": 0.4104,
      "num_input_tokens_seen": 4706392,
      "step": 7260
    },
    {
      "epoch": 4.283608490566038,
      "grad_norm": 1.2465821504592896,
      "learning_rate": 9.608294224161523e-06,
      "loss": 0.4997,
      "num_input_tokens_seen": 4710424,
      "step": 7265
    },
    {
      "epoch": 4.286556603773585,
      "grad_norm": 0.9108287692070007,
      "learning_rate": 9.607295397869847e-06,
      "loss": 0.4287,
      "num_input_tokens_seen": 4713656,
      "step": 7270
    },
    {
      "epoch": 4.289504716981132,
      "grad_norm": 2.199532985687256,
      "learning_rate": 9.60629535177686e-06,
      "loss": 0.6398,
      "num_input_tokens_seen": 4716472,
      "step": 7275
    },
    {
      "epoch": 4.2924528301886795,
      "grad_norm": 1.8738750219345093,
      "learning_rate": 9.605294086147325e-06,
      "loss": 0.3126,
      "num_input_tokens_seen": 4719544,
      "step": 7280
    },
    {
      "epoch": 4.295400943396227,
      "grad_norm": 1.8453953266143799,
      "learning_rate": 9.604291601246333e-06,
      "loss": 0.4055,
      "num_input_tokens_seen": 4722552,
      "step": 7285
    },
    {
      "epoch": 4.298349056603773,
      "grad_norm": 1.3559203147888184,
      "learning_rate": 9.603287897339299e-06,
      "loss": 0.4399,
      "num_input_tokens_seen": 4726072,
      "step": 7290
    },
    {
      "epoch": 4.3012971698113205,
      "grad_norm": 1.426222801208496,
      "learning_rate": 9.602282974691953e-06,
      "loss": 0.5326,
      "num_input_tokens_seen": 4728760,
      "step": 7295
    },
    {
      "epoch": 4.304245283018868,
      "grad_norm": 2.5854411125183105,
      "learning_rate": 9.601276833570355e-06,
      "loss": 0.6139,
      "num_input_tokens_seen": 4731896,
      "step": 7300
    },
    {
      "epoch": 4.307193396226415,
      "grad_norm": 2.2705347537994385,
      "learning_rate": 9.600269474240885e-06,
      "loss": 0.4413,
      "num_input_tokens_seen": 4735032,
      "step": 7305
    },
    {
      "epoch": 4.310141509433962,
      "grad_norm": 1.2622761726379395,
      "learning_rate": 9.599260896970246e-06,
      "loss": 0.3624,
      "num_input_tokens_seen": 4737976,
      "step": 7310
    },
    {
      "epoch": 4.31308962264151,
      "grad_norm": 1.4245573282241821,
      "learning_rate": 9.598251102025463e-06,
      "loss": 0.6836,
      "num_input_tokens_seen": 4741752,
      "step": 7315
    },
    {
      "epoch": 4.316037735849057,
      "grad_norm": 1.744040608406067,
      "learning_rate": 9.597240089673882e-06,
      "loss": 0.4957,
      "num_input_tokens_seen": 4744920,
      "step": 7320
    },
    {
      "epoch": 4.318985849056604,
      "grad_norm": 3.520472764968872,
      "learning_rate": 9.596227860183175e-06,
      "loss": 0.5162,
      "num_input_tokens_seen": 4748120,
      "step": 7325
    },
    {
      "epoch": 4.321933962264151,
      "grad_norm": 1.7938311100006104,
      "learning_rate": 9.595214413821334e-06,
      "loss": 0.5123,
      "num_input_tokens_seen": 4751384,
      "step": 7330
    },
    {
      "epoch": 4.324882075471698,
      "grad_norm": 1.4901649951934814,
      "learning_rate": 9.59419975085667e-06,
      "loss": 0.4141,
      "num_input_tokens_seen": 4754680,
      "step": 7335
    },
    {
      "epoch": 4.327830188679245,
      "grad_norm": 1.6399708986282349,
      "learning_rate": 9.593183871557826e-06,
      "loss": 0.5011,
      "num_input_tokens_seen": 4757592,
      "step": 7340
    },
    {
      "epoch": 4.3307783018867925,
      "grad_norm": 1.6061509847640991,
      "learning_rate": 9.592166776193754e-06,
      "loss": 0.3509,
      "num_input_tokens_seen": 4760824,
      "step": 7345
    },
    {
      "epoch": 4.33372641509434,
      "grad_norm": 2.11625599861145,
      "learning_rate": 9.591148465033738e-06,
      "loss": 0.3448,
      "num_input_tokens_seen": 4763832,
      "step": 7350
    },
    {
      "epoch": 4.336674528301887,
      "grad_norm": 3.6507515907287598,
      "learning_rate": 9.590128938347378e-06,
      "loss": 0.528,
      "num_input_tokens_seen": 4768120,
      "step": 7355
    },
    {
      "epoch": 4.339622641509434,
      "grad_norm": 1.9081450700759888,
      "learning_rate": 9.589108196404599e-06,
      "loss": 0.5039,
      "num_input_tokens_seen": 4772248,
      "step": 7360
    },
    {
      "epoch": 4.342570754716981,
      "grad_norm": 2.3687076568603516,
      "learning_rate": 9.588086239475649e-06,
      "loss": 0.4537,
      "num_input_tokens_seen": 4775480,
      "step": 7365
    },
    {
      "epoch": 4.345518867924528,
      "grad_norm": 1.7219486236572266,
      "learning_rate": 9.587063067831092e-06,
      "loss": 0.42,
      "num_input_tokens_seen": 4779064,
      "step": 7370
    },
    {
      "epoch": 4.348466981132075,
      "grad_norm": 1.8348382711410522,
      "learning_rate": 9.586038681741818e-06,
      "loss": 0.5178,
      "num_input_tokens_seen": 4782936,
      "step": 7375
    },
    {
      "epoch": 4.351415094339623,
      "grad_norm": 1.4297137260437012,
      "learning_rate": 9.58501308147904e-06,
      "loss": 0.2615,
      "num_input_tokens_seen": 4790104,
      "step": 7380
    },
    {
      "epoch": 4.35436320754717,
      "grad_norm": 1.884039044380188,
      "learning_rate": 9.583986267314288e-06,
      "loss": 0.4766,
      "num_input_tokens_seen": 4794040,
      "step": 7385
    },
    {
      "epoch": 4.357311320754717,
      "grad_norm": 1.4150986671447754,
      "learning_rate": 9.582958239519416e-06,
      "loss": 0.3273,
      "num_input_tokens_seen": 4796632,
      "step": 7390
    },
    {
      "epoch": 4.3602594339622645,
      "grad_norm": 1.0310558080673218,
      "learning_rate": 9.581928998366597e-06,
      "loss": 0.3732,
      "num_input_tokens_seen": 4800024,
      "step": 7395
    },
    {
      "epoch": 4.363207547169811,
      "grad_norm": 1.96884286403656,
      "learning_rate": 9.58089854412833e-06,
      "loss": 0.4298,
      "num_input_tokens_seen": 4803576,
      "step": 7400
    },
    {
      "epoch": 4.366155660377358,
      "grad_norm": 1.815252423286438,
      "learning_rate": 9.579866877077431e-06,
      "loss": 0.4089,
      "num_input_tokens_seen": 4806872,
      "step": 7405
    },
    {
      "epoch": 4.369103773584905,
      "grad_norm": 5.339573860168457,
      "learning_rate": 9.578833997487038e-06,
      "loss": 0.4271,
      "num_input_tokens_seen": 4810104,
      "step": 7410
    },
    {
      "epoch": 4.372051886792453,
      "grad_norm": 2.2845335006713867,
      "learning_rate": 9.57779990563061e-06,
      "loss": 0.4562,
      "num_input_tokens_seen": 4813208,
      "step": 7415
    },
    {
      "epoch": 4.375,
      "grad_norm": 2.1580443382263184,
      "learning_rate": 9.576764601781928e-06,
      "loss": 0.4268,
      "num_input_tokens_seen": 4816408,
      "step": 7420
    },
    {
      "epoch": 4.377948113207547,
      "grad_norm": 1.6397452354431152,
      "learning_rate": 9.575728086215093e-06,
      "loss": 0.4195,
      "num_input_tokens_seen": 4819672,
      "step": 7425
    },
    {
      "epoch": 4.380896226415095,
      "grad_norm": 2.246169328689575,
      "learning_rate": 9.574690359204527e-06,
      "loss": 0.3731,
      "num_input_tokens_seen": 4823320,
      "step": 7430
    },
    {
      "epoch": 4.383844339622642,
      "grad_norm": 2.190944194793701,
      "learning_rate": 9.573651421024972e-06,
      "loss": 0.4957,
      "num_input_tokens_seen": 4827096,
      "step": 7435
    },
    {
      "epoch": 4.386792452830189,
      "grad_norm": 1.9642006158828735,
      "learning_rate": 9.572611271951494e-06,
      "loss": 0.4508,
      "num_input_tokens_seen": 4830392,
      "step": 7440
    },
    {
      "epoch": 4.3897405660377355,
      "grad_norm": 2.2617075443267822,
      "learning_rate": 9.571569912259473e-06,
      "loss": 0.5749,
      "num_input_tokens_seen": 4834264,
      "step": 7445
    },
    {
      "epoch": 4.392688679245283,
      "grad_norm": 3.0819618701934814,
      "learning_rate": 9.570527342224614e-06,
      "loss": 0.3967,
      "num_input_tokens_seen": 4836440,
      "step": 7450
    },
    {
      "epoch": 4.39563679245283,
      "grad_norm": 3.3338279724121094,
      "learning_rate": 9.569483562122945e-06,
      "loss": 0.3698,
      "num_input_tokens_seen": 4839384,
      "step": 7455
    },
    {
      "epoch": 4.398584905660377,
      "grad_norm": 3.592940330505371,
      "learning_rate": 9.568438572230811e-06,
      "loss": 0.5102,
      "num_input_tokens_seen": 4842008,
      "step": 7460
    },
    {
      "epoch": 4.401533018867925,
      "grad_norm": 1.640785813331604,
      "learning_rate": 9.567392372824873e-06,
      "loss": 0.3514,
      "num_input_tokens_seen": 4845208,
      "step": 7465
    },
    {
      "epoch": 4.404481132075472,
      "grad_norm": 2.051419973373413,
      "learning_rate": 9.566344964182123e-06,
      "loss": 0.4172,
      "num_input_tokens_seen": 4848472,
      "step": 7470
    },
    {
      "epoch": 4.407429245283019,
      "grad_norm": 1.9847784042358398,
      "learning_rate": 9.565296346579862e-06,
      "loss": 0.4211,
      "num_input_tokens_seen": 4851128,
      "step": 7475
    },
    {
      "epoch": 4.410377358490566,
      "grad_norm": 2.213564157485962,
      "learning_rate": 9.564246520295719e-06,
      "loss": 0.4287,
      "num_input_tokens_seen": 4853944,
      "step": 7480
    },
    {
      "epoch": 4.413325471698113,
      "grad_norm": 1.7926193475723267,
      "learning_rate": 9.563195485607638e-06,
      "loss": 0.4678,
      "num_input_tokens_seen": 4856920,
      "step": 7485
    },
    {
      "epoch": 4.41627358490566,
      "grad_norm": 2.0932881832122803,
      "learning_rate": 9.562143242793885e-06,
      "loss": 0.4113,
      "num_input_tokens_seen": 4859864,
      "step": 7490
    },
    {
      "epoch": 4.4192216981132075,
      "grad_norm": 1.685060977935791,
      "learning_rate": 9.561089792133048e-06,
      "loss": 0.4422,
      "num_input_tokens_seen": 4864056,
      "step": 7495
    },
    {
      "epoch": 4.422169811320755,
      "grad_norm": 1.647123098373413,
      "learning_rate": 9.560035133904031e-06,
      "loss": 0.4473,
      "num_input_tokens_seen": 4867096,
      "step": 7500
    },
    {
      "epoch": 4.425117924528302,
      "grad_norm": 1.4020861387252808,
      "learning_rate": 9.55897926838606e-06,
      "loss": 0.5118,
      "num_input_tokens_seen": 4870328,
      "step": 7505
    },
    {
      "epoch": 4.428066037735849,
      "grad_norm": 1.7762866020202637,
      "learning_rate": 9.55792219585868e-06,
      "loss": 0.4472,
      "num_input_tokens_seen": 4872952,
      "step": 7510
    },
    {
      "epoch": 4.431014150943396,
      "grad_norm": 2.5061898231506348,
      "learning_rate": 9.556863916601754e-06,
      "loss": 0.349,
      "num_input_tokens_seen": 4876216,
      "step": 7515
    },
    {
      "epoch": 4.433962264150943,
      "grad_norm": 2.0743680000305176,
      "learning_rate": 9.555804430895467e-06,
      "loss": 0.4663,
      "num_input_tokens_seen": 4880120,
      "step": 7520
    },
    {
      "epoch": 4.43691037735849,
      "grad_norm": 2.0216317176818848,
      "learning_rate": 9.554743739020325e-06,
      "loss": 0.4927,
      "num_input_tokens_seen": 4882808,
      "step": 7525
    },
    {
      "epoch": 4.439858490566038,
      "grad_norm": 1.6926360130310059,
      "learning_rate": 9.553681841257146e-06,
      "loss": 0.443,
      "num_input_tokens_seen": 4886136,
      "step": 7530
    },
    {
      "epoch": 4.442806603773585,
      "grad_norm": 1.4780491590499878,
      "learning_rate": 9.552618737887073e-06,
      "loss": 0.6842,
      "num_input_tokens_seen": 4888824,
      "step": 7535
    },
    {
      "epoch": 4.445754716981132,
      "grad_norm": 2.6374220848083496,
      "learning_rate": 9.55155442919157e-06,
      "loss": 0.6197,
      "num_input_tokens_seen": 4892952,
      "step": 7540
    },
    {
      "epoch": 4.4487028301886795,
      "grad_norm": 1.688773512840271,
      "learning_rate": 9.550488915452416e-06,
      "loss": 0.6377,
      "num_input_tokens_seen": 4896856,
      "step": 7545
    },
    {
      "epoch": 4.451650943396227,
      "grad_norm": 1.656474232673645,
      "learning_rate": 9.54942219695171e-06,
      "loss": 0.4152,
      "num_input_tokens_seen": 4899800,
      "step": 7550
    },
    {
      "epoch": 4.454599056603773,
      "grad_norm": 1.3265234231948853,
      "learning_rate": 9.54835427397187e-06,
      "loss": 0.4917,
      "num_input_tokens_seen": 4903480,
      "step": 7555
    },
    {
      "epoch": 4.4575471698113205,
      "grad_norm": 1.5433731079101562,
      "learning_rate": 9.547285146795634e-06,
      "loss": 0.4888,
      "num_input_tokens_seen": 4906968,
      "step": 7560
    },
    {
      "epoch": 4.460495283018868,
      "grad_norm": 2.0978927612304688,
      "learning_rate": 9.546214815706059e-06,
      "loss": 0.3808,
      "num_input_tokens_seen": 4910072,
      "step": 7565
    },
    {
      "epoch": 4.463443396226415,
      "grad_norm": 2.1644480228424072,
      "learning_rate": 9.545143280986518e-06,
      "loss": 0.5375,
      "num_input_tokens_seen": 4912920,
      "step": 7570
    },
    {
      "epoch": 4.466391509433962,
      "grad_norm": 2.5181288719177246,
      "learning_rate": 9.544070542920703e-06,
      "loss": 0.5108,
      "num_input_tokens_seen": 4915736,
      "step": 7575
    },
    {
      "epoch": 4.46933962264151,
      "grad_norm": 2.6596500873565674,
      "learning_rate": 9.542996601792629e-06,
      "loss": 0.5391,
      "num_input_tokens_seen": 4919480,
      "step": 7580
    },
    {
      "epoch": 4.472287735849057,
      "grad_norm": 1.4276882410049438,
      "learning_rate": 9.541921457886624e-06,
      "loss": 0.3837,
      "num_input_tokens_seen": 4923864,
      "step": 7585
    },
    {
      "epoch": 4.475235849056604,
      "grad_norm": 1.8625987768173218,
      "learning_rate": 9.54084511148734e-06,
      "loss": 0.4639,
      "num_input_tokens_seen": 4926712,
      "step": 7590
    },
    {
      "epoch": 4.478183962264151,
      "grad_norm": 4.745550155639648,
      "learning_rate": 9.539767562879742e-06,
      "loss": 0.5297,
      "num_input_tokens_seen": 4929976,
      "step": 7595
    },
    {
      "epoch": 4.481132075471698,
      "grad_norm": 1.4697104692459106,
      "learning_rate": 9.538688812349118e-06,
      "loss": 0.4022,
      "num_input_tokens_seen": 4932888,
      "step": 7600
    },
    {
      "epoch": 4.484080188679245,
      "grad_norm": 2.126694917678833,
      "learning_rate": 9.537608860181069e-06,
      "loss": 0.3933,
      "num_input_tokens_seen": 4935480,
      "step": 7605
    },
    {
      "epoch": 4.4870283018867925,
      "grad_norm": 1.6938835382461548,
      "learning_rate": 9.536527706661519e-06,
      "loss": 0.4805,
      "num_input_tokens_seen": 4938296,
      "step": 7610
    },
    {
      "epoch": 4.48997641509434,
      "grad_norm": 1.8029478788375854,
      "learning_rate": 9.535445352076707e-06,
      "loss": 0.4333,
      "num_input_tokens_seen": 4940888,
      "step": 7615
    },
    {
      "epoch": 4.492924528301887,
      "grad_norm": 3.5273706912994385,
      "learning_rate": 9.534361796713191e-06,
      "loss": 0.5336,
      "num_input_tokens_seen": 4944216,
      "step": 7620
    },
    {
      "epoch": 4.495872641509434,
      "grad_norm": 1.4992239475250244,
      "learning_rate": 9.533277040857847e-06,
      "loss": 0.5254,
      "num_input_tokens_seen": 4947000,
      "step": 7625
    },
    {
      "epoch": 4.498820754716981,
      "grad_norm": 0.8911965489387512,
      "learning_rate": 9.53219108479787e-06,
      "loss": 0.549,
      "num_input_tokens_seen": 4950936,
      "step": 7630
    },
    {
      "epoch": 4.501768867924528,
      "grad_norm": 2.646056652069092,
      "learning_rate": 9.53110392882077e-06,
      "loss": 0.4674,
      "num_input_tokens_seen": 4954232,
      "step": 7635
    },
    {
      "epoch": 4.504716981132075,
      "grad_norm": 1.6748520135879517,
      "learning_rate": 9.530015573214378e-06,
      "loss": 0.403,
      "num_input_tokens_seen": 4957240,
      "step": 7640
    },
    {
      "epoch": 4.507665094339623,
      "grad_norm": 2.1035094261169434,
      "learning_rate": 9.528926018266837e-06,
      "loss": 0.455,
      "num_input_tokens_seen": 4959736,
      "step": 7645
    },
    {
      "epoch": 4.51061320754717,
      "grad_norm": 2.0281143188476562,
      "learning_rate": 9.527835264266617e-06,
      "loss": 0.4339,
      "num_input_tokens_seen": 4962808,
      "step": 7650
    },
    {
      "epoch": 4.513561320754717,
      "grad_norm": 2.790592908859253,
      "learning_rate": 9.526743311502496e-06,
      "loss": 0.6195,
      "num_input_tokens_seen": 4965816,
      "step": 7655
    },
    {
      "epoch": 4.5165094339622645,
      "grad_norm": 1.6419509649276733,
      "learning_rate": 9.525650160263573e-06,
      "loss": 0.5869,
      "num_input_tokens_seen": 4968632,
      "step": 7660
    },
    {
      "epoch": 4.519457547169811,
      "grad_norm": 2.310368537902832,
      "learning_rate": 9.524555810839267e-06,
      "loss": 0.4635,
      "num_input_tokens_seen": 4971256,
      "step": 7665
    },
    {
      "epoch": 4.522405660377358,
      "grad_norm": 1.3960034847259521,
      "learning_rate": 9.523460263519309e-06,
      "loss": 0.4805,
      "num_input_tokens_seen": 4973912,
      "step": 7670
    },
    {
      "epoch": 4.525353773584905,
      "grad_norm": 2.9373786449432373,
      "learning_rate": 9.522363518593753e-06,
      "loss": 0.5568,
      "num_input_tokens_seen": 4977240,
      "step": 7675
    },
    {
      "epoch": 4.528301886792453,
      "grad_norm": 4.02360200881958,
      "learning_rate": 9.521265576352963e-06,
      "loss": 0.4514,
      "num_input_tokens_seen": 4981304,
      "step": 7680
    },
    {
      "epoch": 4.53125,
      "grad_norm": 1.7672818899154663,
      "learning_rate": 9.520166437087628e-06,
      "loss": 0.3404,
      "num_input_tokens_seen": 4985144,
      "step": 7685
    },
    {
      "epoch": 4.534198113207547,
      "grad_norm": 1.774185299873352,
      "learning_rate": 9.519066101088748e-06,
      "loss": 0.4701,
      "num_input_tokens_seen": 4988696,
      "step": 7690
    },
    {
      "epoch": 4.537146226415095,
      "grad_norm": 1.6668163537979126,
      "learning_rate": 9.51796456864764e-06,
      "loss": 0.5564,
      "num_input_tokens_seen": 4991640,
      "step": 7695
    },
    {
      "epoch": 4.540094339622642,
      "grad_norm": 0.9049614667892456,
      "learning_rate": 9.516861840055942e-06,
      "loss": 0.4793,
      "num_input_tokens_seen": 4995320,
      "step": 7700
    },
    {
      "epoch": 4.543042452830189,
      "grad_norm": 2.1913983821868896,
      "learning_rate": 9.515757915605604e-06,
      "loss": 0.3615,
      "num_input_tokens_seen": 4998648,
      "step": 7705
    },
    {
      "epoch": 4.5459905660377355,
      "grad_norm": 2.076321840286255,
      "learning_rate": 9.514652795588899e-06,
      "loss": 0.4801,
      "num_input_tokens_seen": 5001368,
      "step": 7710
    },
    {
      "epoch": 4.548938679245283,
      "grad_norm": 2.4531936645507812,
      "learning_rate": 9.513546480298405e-06,
      "loss": 0.6201,
      "num_input_tokens_seen": 5004120,
      "step": 7715
    },
    {
      "epoch": 4.55188679245283,
      "grad_norm": 2.691516399383545,
      "learning_rate": 9.512438970027032e-06,
      "loss": 0.3904,
      "num_input_tokens_seen": 5006872,
      "step": 7720
    },
    {
      "epoch": 4.554834905660377,
      "grad_norm": 2.081662654876709,
      "learning_rate": 9.511330265067992e-06,
      "loss": 0.4263,
      "num_input_tokens_seen": 5009912,
      "step": 7725
    },
    {
      "epoch": 4.557783018867925,
      "grad_norm": 0.9323896169662476,
      "learning_rate": 9.510220365714822e-06,
      "loss": 0.3538,
      "num_input_tokens_seen": 5012568,
      "step": 7730
    },
    {
      "epoch": 4.560731132075472,
      "grad_norm": 1.6719551086425781,
      "learning_rate": 9.509109272261373e-06,
      "loss": 0.5729,
      "num_input_tokens_seen": 5015704,
      "step": 7735
    },
    {
      "epoch": 4.563679245283019,
      "grad_norm": 1.6565715074539185,
      "learning_rate": 9.50799698500181e-06,
      "loss": 0.4034,
      "num_input_tokens_seen": 5019256,
      "step": 7740
    },
    {
      "epoch": 4.566627358490566,
      "grad_norm": 5.476583003997803,
      "learning_rate": 9.506883504230618e-06,
      "loss": 0.5284,
      "num_input_tokens_seen": 5022456,
      "step": 7745
    },
    {
      "epoch": 4.569575471698113,
      "grad_norm": 2.54777193069458,
      "learning_rate": 9.505768830242593e-06,
      "loss": 0.3273,
      "num_input_tokens_seen": 5026552,
      "step": 7750
    },
    {
      "epoch": 4.57252358490566,
      "grad_norm": 2.7710564136505127,
      "learning_rate": 9.504652963332852e-06,
      "loss": 0.4946,
      "num_input_tokens_seen": 5028952,
      "step": 7755
    },
    {
      "epoch": 4.5754716981132075,
      "grad_norm": 1.4145680665969849,
      "learning_rate": 9.503535903796825e-06,
      "loss": 0.4966,
      "num_input_tokens_seen": 5031928,
      "step": 7760
    },
    {
      "epoch": 4.578419811320755,
      "grad_norm": 3.0558133125305176,
      "learning_rate": 9.50241765193026e-06,
      "loss": 0.5708,
      "num_input_tokens_seen": 5034904,
      "step": 7765
    },
    {
      "epoch": 4.581367924528302,
      "grad_norm": 2.137226104736328,
      "learning_rate": 9.501298208029214e-06,
      "loss": 0.384,
      "num_input_tokens_seen": 5038168,
      "step": 7770
    },
    {
      "epoch": 4.584316037735849,
      "grad_norm": 1.5513615608215332,
      "learning_rate": 9.500177572390071e-06,
      "loss": 0.3979,
      "num_input_tokens_seen": 5042104,
      "step": 7775
    },
    {
      "epoch": 4.587264150943396,
      "grad_norm": 1.179875373840332,
      "learning_rate": 9.49905574530952e-06,
      "loss": 0.2812,
      "num_input_tokens_seen": 5044568,
      "step": 7780
    },
    {
      "epoch": 4.590212264150943,
      "grad_norm": 1.7562060356140137,
      "learning_rate": 9.497932727084571e-06,
      "loss": 0.6183,
      "num_input_tokens_seen": 5047608,
      "step": 7785
    },
    {
      "epoch": 4.59316037735849,
      "grad_norm": 1.7781366109848022,
      "learning_rate": 9.496808518012545e-06,
      "loss": 0.4442,
      "num_input_tokens_seen": 5050520,
      "step": 7790
    },
    {
      "epoch": 4.596108490566038,
      "grad_norm": 2.058344841003418,
      "learning_rate": 9.495683118391087e-06,
      "loss": 0.3237,
      "num_input_tokens_seen": 5053784,
      "step": 7795
    },
    {
      "epoch": 4.599056603773585,
      "grad_norm": 1.5891942977905273,
      "learning_rate": 9.494556528518146e-06,
      "loss": 0.4518,
      "num_input_tokens_seen": 5056536,
      "step": 7800
    },
    {
      "epoch": 4.602004716981132,
      "grad_norm": 2.261741876602173,
      "learning_rate": 9.493428748691995e-06,
      "loss": 0.3935,
      "num_input_tokens_seen": 5059416,
      "step": 7805
    },
    {
      "epoch": 4.6049528301886795,
      "grad_norm": 1.8467870950698853,
      "learning_rate": 9.492299779211215e-06,
      "loss": 0.4111,
      "num_input_tokens_seen": 5061944,
      "step": 7810
    },
    {
      "epoch": 4.607900943396227,
      "grad_norm": 1.4411492347717285,
      "learning_rate": 9.49116962037471e-06,
      "loss": 0.525,
      "num_input_tokens_seen": 5065944,
      "step": 7815
    },
    {
      "epoch": 4.610849056603773,
      "grad_norm": 2.7703564167022705,
      "learning_rate": 9.490038272481691e-06,
      "loss": 0.5985,
      "num_input_tokens_seen": 5069176,
      "step": 7820
    },
    {
      "epoch": 4.6137971698113205,
      "grad_norm": 1.6263762712478638,
      "learning_rate": 9.488905735831689e-06,
      "loss": 0.604,
      "num_input_tokens_seen": 5071960,
      "step": 7825
    },
    {
      "epoch": 4.616745283018868,
      "grad_norm": 4.091060638427734,
      "learning_rate": 9.487772010724548e-06,
      "loss": 0.4532,
      "num_input_tokens_seen": 5074200,
      "step": 7830
    },
    {
      "epoch": 4.619693396226415,
      "grad_norm": 2.6607766151428223,
      "learning_rate": 9.486637097460425e-06,
      "loss": 0.7262,
      "num_input_tokens_seen": 5077400,
      "step": 7835
    },
    {
      "epoch": 4.622641509433962,
      "grad_norm": 1.0779554843902588,
      "learning_rate": 9.485500996339793e-06,
      "loss": 0.4057,
      "num_input_tokens_seen": 5080600,
      "step": 7840
    },
    {
      "epoch": 4.62558962264151,
      "grad_norm": 1.5746004581451416,
      "learning_rate": 9.484363707663443e-06,
      "loss": 0.4295,
      "num_input_tokens_seen": 5083192,
      "step": 7845
    },
    {
      "epoch": 4.628537735849057,
      "grad_norm": 3.1038403511047363,
      "learning_rate": 9.483225231732474e-06,
      "loss": 0.5089,
      "num_input_tokens_seen": 5085432,
      "step": 7850
    },
    {
      "epoch": 4.631485849056604,
      "grad_norm": 1.5756696462631226,
      "learning_rate": 9.482085568848302e-06,
      "loss": 0.3299,
      "num_input_tokens_seen": 5089080,
      "step": 7855
    },
    {
      "epoch": 4.634433962264151,
      "grad_norm": 3.2194788455963135,
      "learning_rate": 9.480944719312659e-06,
      "loss": 0.5555,
      "num_input_tokens_seen": 5092344,
      "step": 7860
    },
    {
      "epoch": 4.637382075471698,
      "grad_norm": 2.3263063430786133,
      "learning_rate": 9.47980268342759e-06,
      "loss": 0.4657,
      "num_input_tokens_seen": 5095704,
      "step": 7865
    },
    {
      "epoch": 4.640330188679245,
      "grad_norm": 2.179025411605835,
      "learning_rate": 9.478659461495456e-06,
      "loss": 0.4611,
      "num_input_tokens_seen": 5099224,
      "step": 7870
    },
    {
      "epoch": 4.6432783018867925,
      "grad_norm": 2.0089900493621826,
      "learning_rate": 9.477515053818926e-06,
      "loss": 0.4831,
      "num_input_tokens_seen": 5102200,
      "step": 7875
    },
    {
      "epoch": 4.64622641509434,
      "grad_norm": 1.658553957939148,
      "learning_rate": 9.476369460700988e-06,
      "loss": 0.3358,
      "num_input_tokens_seen": 5104536,
      "step": 7880
    },
    {
      "epoch": 4.649174528301887,
      "grad_norm": 2.693556785583496,
      "learning_rate": 9.475222682444944e-06,
      "loss": 0.7577,
      "num_input_tokens_seen": 5107032,
      "step": 7885
    },
    {
      "epoch": 4.652122641509434,
      "grad_norm": 2.0095341205596924,
      "learning_rate": 9.474074719354406e-06,
      "loss": 0.5682,
      "num_input_tokens_seen": 5110168,
      "step": 7890
    },
    {
      "epoch": 4.655070754716981,
      "grad_norm": 2.0023090839385986,
      "learning_rate": 9.472925571733306e-06,
      "loss": 0.3855,
      "num_input_tokens_seen": 5113656,
      "step": 7895
    },
    {
      "epoch": 4.658018867924528,
      "grad_norm": 1.0972715616226196,
      "learning_rate": 9.471775239885883e-06,
      "loss": 0.4378,
      "num_input_tokens_seen": 5116952,
      "step": 7900
    },
    {
      "epoch": 4.660966981132075,
      "grad_norm": 2.578223705291748,
      "learning_rate": 9.470623724116693e-06,
      "loss": 0.4666,
      "num_input_tokens_seen": 5122904,
      "step": 7905
    },
    {
      "epoch": 4.663915094339623,
      "grad_norm": 2.69887113571167,
      "learning_rate": 9.469471024730606e-06,
      "loss": 0.3711,
      "num_input_tokens_seen": 5125624,
      "step": 7910
    },
    {
      "epoch": 4.66686320754717,
      "grad_norm": 2.000084638595581,
      "learning_rate": 9.4683171420328e-06,
      "loss": 0.3965,
      "num_input_tokens_seen": 5129272,
      "step": 7915
    },
    {
      "epoch": 4.669811320754717,
      "grad_norm": 3.294778823852539,
      "learning_rate": 9.467162076328776e-06,
      "loss": 0.5869,
      "num_input_tokens_seen": 5131864,
      "step": 7920
    },
    {
      "epoch": 4.6727594339622645,
      "grad_norm": 1.3698630332946777,
      "learning_rate": 9.466005827924337e-06,
      "loss": 0.3682,
      "num_input_tokens_seen": 5135960,
      "step": 7925
    },
    {
      "epoch": 4.675707547169811,
      "grad_norm": 1.7085334062576294,
      "learning_rate": 9.46484839712561e-06,
      "loss": 0.4231,
      "num_input_tokens_seen": 5139256,
      "step": 7930
    },
    {
      "epoch": 4.678655660377358,
      "grad_norm": 1.6142829656600952,
      "learning_rate": 9.463689784239026e-06,
      "loss": 0.3807,
      "num_input_tokens_seen": 5142808,
      "step": 7935
    },
    {
      "epoch": 4.681603773584905,
      "grad_norm": 1.5779871940612793,
      "learning_rate": 9.462529989571334e-06,
      "loss": 0.3933,
      "num_input_tokens_seen": 5145624,
      "step": 7940
    },
    {
      "epoch": 4.684551886792453,
      "grad_norm": 1.9351084232330322,
      "learning_rate": 9.461369013429595e-06,
      "loss": 0.6099,
      "num_input_tokens_seen": 5148312,
      "step": 7945
    },
    {
      "epoch": 4.6875,
      "grad_norm": 1.7036817073822021,
      "learning_rate": 9.460206856121183e-06,
      "loss": 0.5334,
      "num_input_tokens_seen": 5151416,
      "step": 7950
    },
    {
      "epoch": 4.690448113207547,
      "grad_norm": 1.4140456914901733,
      "learning_rate": 9.459043517953786e-06,
      "loss": 0.3711,
      "num_input_tokens_seen": 5154264,
      "step": 7955
    },
    {
      "epoch": 4.693396226415095,
      "grad_norm": 1.374755859375,
      "learning_rate": 9.457878999235396e-06,
      "loss": 0.4638,
      "num_input_tokens_seen": 5157304,
      "step": 7960
    },
    {
      "epoch": 4.696344339622642,
      "grad_norm": 1.6814062595367432,
      "learning_rate": 9.45671330027433e-06,
      "loss": 0.4323,
      "num_input_tokens_seen": 5160440,
      "step": 7965
    },
    {
      "epoch": 4.699292452830189,
      "grad_norm": 1.5154306888580322,
      "learning_rate": 9.45554642137921e-06,
      "loss": 0.396,
      "num_input_tokens_seen": 5163160,
      "step": 7970
    },
    {
      "epoch": 4.7022405660377355,
      "grad_norm": 1.9687484502792358,
      "learning_rate": 9.454378362858974e-06,
      "loss": 0.4732,
      "num_input_tokens_seen": 5166680,
      "step": 7975
    },
    {
      "epoch": 4.705188679245283,
      "grad_norm": 4.005729675292969,
      "learning_rate": 9.453209125022867e-06,
      "loss": 0.4756,
      "num_input_tokens_seen": 5170232,
      "step": 7980
    },
    {
      "epoch": 4.70813679245283,
      "grad_norm": 2.8925564289093018,
      "learning_rate": 9.452038708180453e-06,
      "loss": 0.5534,
      "num_input_tokens_seen": 5173432,
      "step": 7985
    },
    {
      "epoch": 4.711084905660377,
      "grad_norm": 2.0067808628082275,
      "learning_rate": 9.450867112641603e-06,
      "loss": 0.437,
      "num_input_tokens_seen": 5176472,
      "step": 7990
    },
    {
      "epoch": 4.714033018867925,
      "grad_norm": 1.341781735420227,
      "learning_rate": 9.449694338716506e-06,
      "loss": 0.3222,
      "num_input_tokens_seen": 5179288,
      "step": 7995
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 2.212693452835083,
      "learning_rate": 9.448520386715653e-06,
      "loss": 0.3737,
      "num_input_tokens_seen": 5182168,
      "step": 8000
    },
    {
      "epoch": 4.719929245283019,
      "grad_norm": 1.5724740028381348,
      "learning_rate": 9.447345256949855e-06,
      "loss": 0.4714,
      "num_input_tokens_seen": 5185432,
      "step": 8005
    },
    {
      "epoch": 4.722877358490566,
      "grad_norm": 3.3736894130706787,
      "learning_rate": 9.446168949730234e-06,
      "loss": 0.463,
      "num_input_tokens_seen": 5189080,
      "step": 8010
    },
    {
      "epoch": 4.725825471698113,
      "grad_norm": 1.4543675184249878,
      "learning_rate": 9.444991465368223e-06,
      "loss": 0.4056,
      "num_input_tokens_seen": 5192536,
      "step": 8015
    },
    {
      "epoch": 4.72877358490566,
      "grad_norm": 1.6838704347610474,
      "learning_rate": 9.443812804175562e-06,
      "loss": 0.4763,
      "num_input_tokens_seen": 5195448,
      "step": 8020
    },
    {
      "epoch": 4.7317216981132075,
      "grad_norm": 1.0720560550689697,
      "learning_rate": 9.44263296646431e-06,
      "loss": 0.5474,
      "num_input_tokens_seen": 5198520,
      "step": 8025
    },
    {
      "epoch": 4.734669811320755,
      "grad_norm": 1.3163431882858276,
      "learning_rate": 9.441451952546835e-06,
      "loss": 0.4416,
      "num_input_tokens_seen": 5201688,
      "step": 8030
    },
    {
      "epoch": 4.737617924528302,
      "grad_norm": 3.060107707977295,
      "learning_rate": 9.440269762735814e-06,
      "loss": 0.4113,
      "num_input_tokens_seen": 5206488,
      "step": 8035
    },
    {
      "epoch": 4.740566037735849,
      "grad_norm": 1.0321770906448364,
      "learning_rate": 9.439086397344236e-06,
      "loss": 0.3342,
      "num_input_tokens_seen": 5209976,
      "step": 8040
    },
    {
      "epoch": 4.743514150943396,
      "grad_norm": 2.7166030406951904,
      "learning_rate": 9.437901856685404e-06,
      "loss": 0.515,
      "num_input_tokens_seen": 5213368,
      "step": 8045
    },
    {
      "epoch": 4.746462264150943,
      "grad_norm": 2.19297456741333,
      "learning_rate": 9.436716141072925e-06,
      "loss": 0.4555,
      "num_input_tokens_seen": 5216440,
      "step": 8050
    },
    {
      "epoch": 4.74941037735849,
      "grad_norm": 1.983540654182434,
      "learning_rate": 9.435529250820732e-06,
      "loss": 0.4969,
      "num_input_tokens_seen": 5219384,
      "step": 8055
    },
    {
      "epoch": 4.752358490566038,
      "grad_norm": 2.2066287994384766,
      "learning_rate": 9.43434118624305e-06,
      "loss": 0.4702,
      "num_input_tokens_seen": 5222936,
      "step": 8060
    },
    {
      "epoch": 4.755306603773585,
      "grad_norm": 1.5345041751861572,
      "learning_rate": 9.433151947654428e-06,
      "loss": 0.4469,
      "num_input_tokens_seen": 5228024,
      "step": 8065
    },
    {
      "epoch": 4.758254716981132,
      "grad_norm": 1.8425235748291016,
      "learning_rate": 9.431961535369724e-06,
      "loss": 0.424,
      "num_input_tokens_seen": 5232056,
      "step": 8070
    },
    {
      "epoch": 4.7612028301886795,
      "grad_norm": 2.69683575630188,
      "learning_rate": 9.430769949704103e-06,
      "loss": 0.4087,
      "num_input_tokens_seen": 5235128,
      "step": 8075
    },
    {
      "epoch": 4.764150943396227,
      "grad_norm": 3.0805227756500244,
      "learning_rate": 9.42957719097304e-06,
      "loss": 0.4395,
      "num_input_tokens_seen": 5240152,
      "step": 8080
    },
    {
      "epoch": 4.767099056603773,
      "grad_norm": 2.219154119491577,
      "learning_rate": 9.42838325949233e-06,
      "loss": 0.4209,
      "num_input_tokens_seen": 5242808,
      "step": 8085
    },
    {
      "epoch": 4.7700471698113205,
      "grad_norm": 2.8536109924316406,
      "learning_rate": 9.427188155578062e-06,
      "loss": 0.4342,
      "num_input_tokens_seen": 5245784,
      "step": 8090
    },
    {
      "epoch": 4.772995283018868,
      "grad_norm": 2.6823031902313232,
      "learning_rate": 9.42599187954665e-06,
      "loss": 0.473,
      "num_input_tokens_seen": 5249336,
      "step": 8095
    },
    {
      "epoch": 4.775943396226415,
      "grad_norm": 1.3099194765090942,
      "learning_rate": 9.424794431714814e-06,
      "loss": 0.5086,
      "num_input_tokens_seen": 5252888,
      "step": 8100
    },
    {
      "epoch": 4.778891509433962,
      "grad_norm": 2.2919366359710693,
      "learning_rate": 9.423595812399581e-06,
      "loss": 0.4848,
      "num_input_tokens_seen": 5255608,
      "step": 8105
    },
    {
      "epoch": 4.78183962264151,
      "grad_norm": 4.366831302642822,
      "learning_rate": 9.422396021918296e-06,
      "loss": 0.4549,
      "num_input_tokens_seen": 5258520,
      "step": 8110
    },
    {
      "epoch": 4.784787735849057,
      "grad_norm": 1.227454662322998,
      "learning_rate": 9.421195060588602e-06,
      "loss": 0.3889,
      "num_input_tokens_seen": 5262136,
      "step": 8115
    },
    {
      "epoch": 4.787735849056604,
      "grad_norm": 1.3544565439224243,
      "learning_rate": 9.419992928728461e-06,
      "loss": 0.3917,
      "num_input_tokens_seen": 5265208,
      "step": 8120
    },
    {
      "epoch": 4.790683962264151,
      "grad_norm": 1.1124553680419922,
      "learning_rate": 9.418789626656144e-06,
      "loss": 0.4186,
      "num_input_tokens_seen": 5267640,
      "step": 8125
    },
    {
      "epoch": 4.793632075471698,
      "grad_norm": 1.5058207511901855,
      "learning_rate": 9.417585154690229e-06,
      "loss": 0.4937,
      "num_input_tokens_seen": 5271256,
      "step": 8130
    },
    {
      "epoch": 4.796580188679245,
      "grad_norm": 2.836780309677124,
      "learning_rate": 9.416379513149605e-06,
      "loss": 0.49,
      "num_input_tokens_seen": 5274712,
      "step": 8135
    },
    {
      "epoch": 4.7995283018867925,
      "grad_norm": 2.23384952545166,
      "learning_rate": 9.415172702353471e-06,
      "loss": 0.4374,
      "num_input_tokens_seen": 5278200,
      "step": 8140
    },
    {
      "epoch": 4.80247641509434,
      "grad_norm": 2.481684923171997,
      "learning_rate": 9.413964722621339e-06,
      "loss": 0.5538,
      "num_input_tokens_seen": 5281240,
      "step": 8145
    },
    {
      "epoch": 4.805424528301887,
      "grad_norm": 1.6993961334228516,
      "learning_rate": 9.41275557427302e-06,
      "loss": 0.4506,
      "num_input_tokens_seen": 5284152,
      "step": 8150
    },
    {
      "epoch": 4.808372641509434,
      "grad_norm": 2.4399161338806152,
      "learning_rate": 9.411545257628646e-06,
      "loss": 0.4808,
      "num_input_tokens_seen": 5287416,
      "step": 8155
    },
    {
      "epoch": 4.811320754716981,
      "grad_norm": 1.7393821477890015,
      "learning_rate": 9.410333773008653e-06,
      "loss": 0.5183,
      "num_input_tokens_seen": 5290904,
      "step": 8160
    },
    {
      "epoch": 4.814268867924528,
      "grad_norm": 2.380504608154297,
      "learning_rate": 9.409121120733784e-06,
      "loss": 0.417,
      "num_input_tokens_seen": 5293176,
      "step": 8165
    },
    {
      "epoch": 4.817216981132075,
      "grad_norm": 2.621217966079712,
      "learning_rate": 9.4079073011251e-06,
      "loss": 0.4088,
      "num_input_tokens_seen": 5295928,
      "step": 8170
    },
    {
      "epoch": 4.820165094339623,
      "grad_norm": 2.031277894973755,
      "learning_rate": 9.406692314503956e-06,
      "loss": 0.6052,
      "num_input_tokens_seen": 5298872,
      "step": 8175
    },
    {
      "epoch": 4.82311320754717,
      "grad_norm": 1.3462141752243042,
      "learning_rate": 9.405476161192033e-06,
      "loss": 0.537,
      "num_input_tokens_seen": 5301752,
      "step": 8180
    },
    {
      "epoch": 4.826061320754717,
      "grad_norm": 1.7538155317306519,
      "learning_rate": 9.40425884151131e-06,
      "loss": 0.3868,
      "num_input_tokens_seen": 5304312,
      "step": 8185
    },
    {
      "epoch": 4.8290094339622645,
      "grad_norm": 1.824692726135254,
      "learning_rate": 9.403040355784076e-06,
      "loss": 0.5868,
      "num_input_tokens_seen": 5307832,
      "step": 8190
    },
    {
      "epoch": 4.831957547169811,
      "grad_norm": 1.1856783628463745,
      "learning_rate": 9.401820704332932e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 5310936,
      "step": 8195
    },
    {
      "epoch": 4.834905660377358,
      "grad_norm": 1.680632472038269,
      "learning_rate": 9.400599887480786e-06,
      "loss": 0.4198,
      "num_input_tokens_seen": 5314552,
      "step": 8200
    },
    {
      "epoch": 4.837853773584905,
      "grad_norm": 6.946560382843018,
      "learning_rate": 9.399377905550854e-06,
      "loss": 0.5542,
      "num_input_tokens_seen": 5317528,
      "step": 8205
    },
    {
      "epoch": 4.840801886792453,
      "grad_norm": 2.2071330547332764,
      "learning_rate": 9.398154758866662e-06,
      "loss": 0.5131,
      "num_input_tokens_seen": 5322104,
      "step": 8210
    },
    {
      "epoch": 4.84375,
      "grad_norm": 1.6472985744476318,
      "learning_rate": 9.396930447752041e-06,
      "loss": 0.4153,
      "num_input_tokens_seen": 5324696,
      "step": 8215
    },
    {
      "epoch": 4.846698113207547,
      "grad_norm": 2.071059465408325,
      "learning_rate": 9.395704972531137e-06,
      "loss": 0.4766,
      "num_input_tokens_seen": 5327448,
      "step": 8220
    },
    {
      "epoch": 4.849646226415095,
      "grad_norm": 1.4296942949295044,
      "learning_rate": 9.394478333528396e-06,
      "loss": 0.5814,
      "num_input_tokens_seen": 5330456,
      "step": 8225
    },
    {
      "epoch": 4.852594339622642,
      "grad_norm": 1.9545646905899048,
      "learning_rate": 9.393250531068576e-06,
      "loss": 0.4702,
      "num_input_tokens_seen": 5333592,
      "step": 8230
    },
    {
      "epoch": 4.855542452830189,
      "grad_norm": 2.5077311992645264,
      "learning_rate": 9.392021565476744e-06,
      "loss": 0.5707,
      "num_input_tokens_seen": 5336600,
      "step": 8235
    },
    {
      "epoch": 4.8584905660377355,
      "grad_norm": 1.8199974298477173,
      "learning_rate": 9.390791437078274e-06,
      "loss": 0.4591,
      "num_input_tokens_seen": 5339256,
      "step": 8240
    },
    {
      "epoch": 4.861438679245283,
      "grad_norm": 2.710249900817871,
      "learning_rate": 9.38956014619885e-06,
      "loss": 0.4619,
      "num_input_tokens_seen": 5342360,
      "step": 8245
    },
    {
      "epoch": 4.86438679245283,
      "grad_norm": 1.871763825416565,
      "learning_rate": 9.388327693164456e-06,
      "loss": 0.5275,
      "num_input_tokens_seen": 5345464,
      "step": 8250
    },
    {
      "epoch": 4.867334905660377,
      "grad_norm": 1.51548433303833,
      "learning_rate": 9.387094078301395e-06,
      "loss": 0.3909,
      "num_input_tokens_seen": 5348472,
      "step": 8255
    },
    {
      "epoch": 4.870283018867925,
      "grad_norm": 1.794756531715393,
      "learning_rate": 9.385859301936269e-06,
      "loss": 0.5077,
      "num_input_tokens_seen": 5351384,
      "step": 8260
    },
    {
      "epoch": 4.873231132075472,
      "grad_norm": 2.8901262283325195,
      "learning_rate": 9.38462336439599e-06,
      "loss": 0.4459,
      "num_input_tokens_seen": 5354296,
      "step": 8265
    },
    {
      "epoch": 4.876179245283019,
      "grad_norm": 2.530895233154297,
      "learning_rate": 9.383386266007779e-06,
      "loss": 0.4199,
      "num_input_tokens_seen": 5357048,
      "step": 8270
    },
    {
      "epoch": 4.879127358490566,
      "grad_norm": 2.8682310581207275,
      "learning_rate": 9.382148007099164e-06,
      "loss": 0.5449,
      "num_input_tokens_seen": 5360472,
      "step": 8275
    },
    {
      "epoch": 4.882075471698113,
      "grad_norm": 2.8128411769866943,
      "learning_rate": 9.380908587997977e-06,
      "loss": 0.4827,
      "num_input_tokens_seen": 5363096,
      "step": 8280
    },
    {
      "epoch": 4.88502358490566,
      "grad_norm": 1.258516550064087,
      "learning_rate": 9.37966800903236e-06,
      "loss": 0.4164,
      "num_input_tokens_seen": 5366808,
      "step": 8285
    },
    {
      "epoch": 4.8879716981132075,
      "grad_norm": 2.251322031021118,
      "learning_rate": 9.378426270530762e-06,
      "loss": 0.5117,
      "num_input_tokens_seen": 5370904,
      "step": 8290
    },
    {
      "epoch": 4.890919811320755,
      "grad_norm": 1.6408435106277466,
      "learning_rate": 9.37718337282194e-06,
      "loss": 0.4502,
      "num_input_tokens_seen": 5373752,
      "step": 8295
    },
    {
      "epoch": 4.893867924528302,
      "grad_norm": 1.9437096118927002,
      "learning_rate": 9.375939316234956e-06,
      "loss": 0.486,
      "num_input_tokens_seen": 5377112,
      "step": 8300
    },
    {
      "epoch": 4.896816037735849,
      "grad_norm": 6.631422519683838,
      "learning_rate": 9.374694101099178e-06,
      "loss": 0.538,
      "num_input_tokens_seen": 5381336,
      "step": 8305
    },
    {
      "epoch": 4.899764150943396,
      "grad_norm": 2.17789888381958,
      "learning_rate": 9.373447727744282e-06,
      "loss": 0.4826,
      "num_input_tokens_seen": 5383896,
      "step": 8310
    },
    {
      "epoch": 4.902712264150943,
      "grad_norm": 4.4914374351501465,
      "learning_rate": 9.372200196500253e-06,
      "loss": 0.5268,
      "num_input_tokens_seen": 5386808,
      "step": 8315
    },
    {
      "epoch": 4.90566037735849,
      "grad_norm": 2.0658421516418457,
      "learning_rate": 9.37095150769738e-06,
      "loss": 0.4636,
      "num_input_tokens_seen": 5389528,
      "step": 8320
    },
    {
      "epoch": 4.908608490566038,
      "grad_norm": 1.7967795133590698,
      "learning_rate": 9.369701661666255e-06,
      "loss": 0.3935,
      "num_input_tokens_seen": 5392152,
      "step": 8325
    },
    {
      "epoch": 4.911556603773585,
      "grad_norm": 1.4333566427230835,
      "learning_rate": 9.368450658737782e-06,
      "loss": 0.3913,
      "num_input_tokens_seen": 5395864,
      "step": 8330
    },
    {
      "epoch": 4.914504716981132,
      "grad_norm": 1.8090747594833374,
      "learning_rate": 9.367198499243173e-06,
      "loss": 0.5288,
      "num_input_tokens_seen": 5399480,
      "step": 8335
    },
    {
      "epoch": 4.9174528301886795,
      "grad_norm": 1.5592643022537231,
      "learning_rate": 9.365945183513938e-06,
      "loss": 0.4333,
      "num_input_tokens_seen": 5402712,
      "step": 8340
    },
    {
      "epoch": 4.920400943396227,
      "grad_norm": 2.4790749549865723,
      "learning_rate": 9.3646907118819e-06,
      "loss": 0.4675,
      "num_input_tokens_seen": 5405688,
      "step": 8345
    },
    {
      "epoch": 4.923349056603773,
      "grad_norm": 1.7325142621994019,
      "learning_rate": 9.363435084679185e-06,
      "loss": 0.4722,
      "num_input_tokens_seen": 5408600,
      "step": 8350
    },
    {
      "epoch": 4.9262971698113205,
      "grad_norm": 1.8452893495559692,
      "learning_rate": 9.362178302238227e-06,
      "loss": 0.3898,
      "num_input_tokens_seen": 5411352,
      "step": 8355
    },
    {
      "epoch": 4.929245283018868,
      "grad_norm": 3.392467975616455,
      "learning_rate": 9.360920364891762e-06,
      "loss": 0.4771,
      "num_input_tokens_seen": 5414232,
      "step": 8360
    },
    {
      "epoch": 4.932193396226415,
      "grad_norm": 1.914250373840332,
      "learning_rate": 9.359661272972836e-06,
      "loss": 0.349,
      "num_input_tokens_seen": 5417208,
      "step": 8365
    },
    {
      "epoch": 4.935141509433962,
      "grad_norm": 2.329913854598999,
      "learning_rate": 9.3584010268148e-06,
      "loss": 0.5036,
      "num_input_tokens_seen": 5420088,
      "step": 8370
    },
    {
      "epoch": 4.93808962264151,
      "grad_norm": 2.5466718673706055,
      "learning_rate": 9.357139626751308e-06,
      "loss": 0.553,
      "num_input_tokens_seen": 5424216,
      "step": 8375
    },
    {
      "epoch": 4.941037735849057,
      "grad_norm": 2.1277756690979004,
      "learning_rate": 9.355877073116321e-06,
      "loss": 0.3522,
      "num_input_tokens_seen": 5427128,
      "step": 8380
    },
    {
      "epoch": 4.943985849056604,
      "grad_norm": 2.7849209308624268,
      "learning_rate": 9.354613366244108e-06,
      "loss": 0.5505,
      "num_input_tokens_seen": 5430232,
      "step": 8385
    },
    {
      "epoch": 4.946933962264151,
      "grad_norm": 1.127979040145874,
      "learning_rate": 9.353348506469236e-06,
      "loss": 0.542,
      "num_input_tokens_seen": 5433304,
      "step": 8390
    },
    {
      "epoch": 4.949882075471698,
      "grad_norm": 1.8880505561828613,
      "learning_rate": 9.352082494126586e-06,
      "loss": 0.7613,
      "num_input_tokens_seen": 5435928,
      "step": 8395
    },
    {
      "epoch": 4.952830188679245,
      "grad_norm": 1.9169583320617676,
      "learning_rate": 9.350815329551341e-06,
      "loss": 0.4426,
      "num_input_tokens_seen": 5438328,
      "step": 8400
    },
    {
      "epoch": 4.9557783018867925,
      "grad_norm": 1.3793284893035889,
      "learning_rate": 9.349547013078986e-06,
      "loss": 0.4278,
      "num_input_tokens_seen": 5442200,
      "step": 8405
    },
    {
      "epoch": 4.95872641509434,
      "grad_norm": 2.038081645965576,
      "learning_rate": 9.348277545045312e-06,
      "loss": 0.5534,
      "num_input_tokens_seen": 5445784,
      "step": 8410
    },
    {
      "epoch": 4.961674528301887,
      "grad_norm": 2.0116100311279297,
      "learning_rate": 9.347006925786418e-06,
      "loss": 0.3865,
      "num_input_tokens_seen": 5449432,
      "step": 8415
    },
    {
      "epoch": 4.964622641509434,
      "grad_norm": 1.3580315113067627,
      "learning_rate": 9.34573515563871e-06,
      "loss": 0.405,
      "num_input_tokens_seen": 5452344,
      "step": 8420
    },
    {
      "epoch": 4.967570754716981,
      "grad_norm": 1.3283438682556152,
      "learning_rate": 9.344462234938885e-06,
      "loss": 0.4168,
      "num_input_tokens_seen": 5455928,
      "step": 8425
    },
    {
      "epoch": 4.970518867924528,
      "grad_norm": 2.202380895614624,
      "learning_rate": 9.343188164023962e-06,
      "loss": 0.4582,
      "num_input_tokens_seen": 5458744,
      "step": 8430
    },
    {
      "epoch": 4.973466981132075,
      "grad_norm": 2.3253579139709473,
      "learning_rate": 9.341912943231256e-06,
      "loss": 0.5907,
      "num_input_tokens_seen": 5462168,
      "step": 8435
    },
    {
      "epoch": 4.976415094339623,
      "grad_norm": 2.6678855419158936,
      "learning_rate": 9.340636572898383e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 5465912,
      "step": 8440
    },
    {
      "epoch": 4.97936320754717,
      "grad_norm": 1.51052987575531,
      "learning_rate": 9.339359053363272e-06,
      "loss": 0.45,
      "num_input_tokens_seen": 5470200,
      "step": 8445
    },
    {
      "epoch": 4.982311320754717,
      "grad_norm": 2.0643162727355957,
      "learning_rate": 9.338080384964148e-06,
      "loss": 0.4067,
      "num_input_tokens_seen": 5473208,
      "step": 8450
    },
    {
      "epoch": 4.9852594339622645,
      "grad_norm": 1.1337122917175293,
      "learning_rate": 9.336800568039548e-06,
      "loss": 0.4809,
      "num_input_tokens_seen": 5476984,
      "step": 8455
    },
    {
      "epoch": 4.988207547169811,
      "grad_norm": 3.24387264251709,
      "learning_rate": 9.335519602928307e-06,
      "loss": 0.5012,
      "num_input_tokens_seen": 5480088,
      "step": 8460
    },
    {
      "epoch": 4.991155660377358,
      "grad_norm": 2.8743481636047363,
      "learning_rate": 9.334237489969565e-06,
      "loss": 0.3783,
      "num_input_tokens_seen": 5483000,
      "step": 8465
    },
    {
      "epoch": 4.994103773584905,
      "grad_norm": 2.0913660526275635,
      "learning_rate": 9.332954229502768e-06,
      "loss": 0.435,
      "num_input_tokens_seen": 5486424,
      "step": 8470
    },
    {
      "epoch": 4.997051886792453,
      "grad_norm": 1.8667025566101074,
      "learning_rate": 9.331669821867665e-06,
      "loss": 0.4861,
      "num_input_tokens_seen": 5489912,
      "step": 8475
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.8194719552993774,
      "learning_rate": 9.33038426740431e-06,
      "loss": 0.4287,
      "num_input_tokens_seen": 5493176,
      "step": 8480
    },
    {
      "epoch": 5.002948113207547,
      "grad_norm": 1.1562401056289673,
      "learning_rate": 9.329097566453055e-06,
      "loss": 0.4251,
      "num_input_tokens_seen": 5496248,
      "step": 8485
    },
    {
      "epoch": 5.005896226415095,
      "grad_norm": 2.2610092163085938,
      "learning_rate": 9.327809719354564e-06,
      "loss": 0.6222,
      "num_input_tokens_seen": 5499224,
      "step": 8490
    },
    {
      "epoch": 5.008844339622642,
      "grad_norm": 1.6574430465698242,
      "learning_rate": 9.326520726449795e-06,
      "loss": 0.4886,
      "num_input_tokens_seen": 5502712,
      "step": 8495
    },
    {
      "epoch": 5.011792452830188,
      "grad_norm": 2.2673237323760986,
      "learning_rate": 9.32523058808002e-06,
      "loss": 0.461,
      "num_input_tokens_seen": 5505464,
      "step": 8500
    },
    {
      "epoch": 5.0147405660377355,
      "grad_norm": 1.8861371278762817,
      "learning_rate": 9.323939304586806e-06,
      "loss": 0.4218,
      "num_input_tokens_seen": 5508152,
      "step": 8505
    },
    {
      "epoch": 5.017688679245283,
      "grad_norm": 1.6875883340835571,
      "learning_rate": 9.322646876312025e-06,
      "loss": 0.5548,
      "num_input_tokens_seen": 5511128,
      "step": 8510
    },
    {
      "epoch": 5.02063679245283,
      "grad_norm": 1.408738613128662,
      "learning_rate": 9.321353303597854e-06,
      "loss": 0.4642,
      "num_input_tokens_seen": 5514104,
      "step": 8515
    },
    {
      "epoch": 5.023584905660377,
      "grad_norm": 0.9452056884765625,
      "learning_rate": 9.320058586786771e-06,
      "loss": 0.3888,
      "num_input_tokens_seen": 5516920,
      "step": 8520
    },
    {
      "epoch": 5.026533018867925,
      "grad_norm": 2.2927565574645996,
      "learning_rate": 9.31876272622156e-06,
      "loss": 0.3252,
      "num_input_tokens_seen": 5520312,
      "step": 8525
    },
    {
      "epoch": 5.029481132075472,
      "grad_norm": 1.386008620262146,
      "learning_rate": 9.317465722245305e-06,
      "loss": 0.4532,
      "num_input_tokens_seen": 5523064,
      "step": 8530
    },
    {
      "epoch": 5.032429245283019,
      "grad_norm": 2.169170618057251,
      "learning_rate": 9.316167575201391e-06,
      "loss": 0.4553,
      "num_input_tokens_seen": 5526488,
      "step": 8535
    },
    {
      "epoch": 5.035377358490566,
      "grad_norm": 2.963007688522339,
      "learning_rate": 9.31486828543351e-06,
      "loss": 0.4434,
      "num_input_tokens_seen": 5529464,
      "step": 8540
    },
    {
      "epoch": 5.038325471698113,
      "grad_norm": 1.8185256719589233,
      "learning_rate": 9.313567853285656e-06,
      "loss": 0.4577,
      "num_input_tokens_seen": 5533464,
      "step": 8545
    },
    {
      "epoch": 5.04127358490566,
      "grad_norm": 1.047972321510315,
      "learning_rate": 9.31226627910212e-06,
      "loss": 0.3938,
      "num_input_tokens_seen": 5536920,
      "step": 8550
    },
    {
      "epoch": 5.0442216981132075,
      "grad_norm": 1.9704906940460205,
      "learning_rate": 9.310963563227504e-06,
      "loss": 0.3727,
      "num_input_tokens_seen": 5540024,
      "step": 8555
    },
    {
      "epoch": 5.047169811320755,
      "grad_norm": 6.603094577789307,
      "learning_rate": 9.309659706006704e-06,
      "loss": 0.4393,
      "num_input_tokens_seen": 5543000,
      "step": 8560
    },
    {
      "epoch": 5.050117924528302,
      "grad_norm": 0.9627794623374939,
      "learning_rate": 9.308354707784925e-06,
      "loss": 0.4046,
      "num_input_tokens_seen": 5546104,
      "step": 8565
    },
    {
      "epoch": 5.053066037735849,
      "grad_norm": 2.4410505294799805,
      "learning_rate": 9.307048568907669e-06,
      "loss": 0.6028,
      "num_input_tokens_seen": 5549720,
      "step": 8570
    },
    {
      "epoch": 5.056014150943396,
      "grad_norm": 1.5390472412109375,
      "learning_rate": 9.30574128972074e-06,
      "loss": 0.3768,
      "num_input_tokens_seen": 5553016,
      "step": 8575
    },
    {
      "epoch": 5.058962264150943,
      "grad_norm": 1.1072760820388794,
      "learning_rate": 9.304432870570247e-06,
      "loss": 0.3863,
      "num_input_tokens_seen": 5557656,
      "step": 8580
    },
    {
      "epoch": 5.06191037735849,
      "grad_norm": 3.6769871711730957,
      "learning_rate": 9.303123311802605e-06,
      "loss": 0.4833,
      "num_input_tokens_seen": 5560408,
      "step": 8585
    },
    {
      "epoch": 5.064858490566038,
      "grad_norm": 1.8371472358703613,
      "learning_rate": 9.301812613764516e-06,
      "loss": 0.5578,
      "num_input_tokens_seen": 5563288,
      "step": 8590
    },
    {
      "epoch": 5.067806603773585,
      "grad_norm": 1.1125918626785278,
      "learning_rate": 9.300500776803001e-06,
      "loss": 0.3125,
      "num_input_tokens_seen": 5567640,
      "step": 8595
    },
    {
      "epoch": 5.070754716981132,
      "grad_norm": 1.8751815557479858,
      "learning_rate": 9.29918780126537e-06,
      "loss": 0.5536,
      "num_input_tokens_seen": 5570424,
      "step": 8600
    },
    {
      "epoch": 5.0737028301886795,
      "grad_norm": 2.5737476348876953,
      "learning_rate": 9.297873687499239e-06,
      "loss": 0.4737,
      "num_input_tokens_seen": 5572504,
      "step": 8605
    },
    {
      "epoch": 5.076650943396227,
      "grad_norm": 1.3179229497909546,
      "learning_rate": 9.296558435852528e-06,
      "loss": 0.3437,
      "num_input_tokens_seen": 5575832,
      "step": 8610
    },
    {
      "epoch": 5.079599056603773,
      "grad_norm": 2.599494457244873,
      "learning_rate": 9.295242046673454e-06,
      "loss": 0.431,
      "num_input_tokens_seen": 5579352,
      "step": 8615
    },
    {
      "epoch": 5.0825471698113205,
      "grad_norm": 3.1603479385375977,
      "learning_rate": 9.293924520310535e-06,
      "loss": 0.5456,
      "num_input_tokens_seen": 5582488,
      "step": 8620
    },
    {
      "epoch": 5.085495283018868,
      "grad_norm": 1.7839007377624512,
      "learning_rate": 9.292605857112595e-06,
      "loss": 0.507,
      "num_input_tokens_seen": 5585592,
      "step": 8625
    },
    {
      "epoch": 5.088443396226415,
      "grad_norm": 1.7319002151489258,
      "learning_rate": 9.291286057428755e-06,
      "loss": 0.3966,
      "num_input_tokens_seen": 5588408,
      "step": 8630
    },
    {
      "epoch": 5.091391509433962,
      "grad_norm": 1.2949438095092773,
      "learning_rate": 9.289965121608436e-06,
      "loss": 0.4309,
      "num_input_tokens_seen": 5591608,
      "step": 8635
    },
    {
      "epoch": 5.09433962264151,
      "grad_norm": 3.1135716438293457,
      "learning_rate": 9.288643050001362e-06,
      "loss": 0.4842,
      "num_input_tokens_seen": 5594712,
      "step": 8640
    },
    {
      "epoch": 5.097287735849057,
      "grad_norm": 1.6551392078399658,
      "learning_rate": 9.287319842957557e-06,
      "loss": 0.456,
      "num_input_tokens_seen": 5598680,
      "step": 8645
    },
    {
      "epoch": 5.100235849056604,
      "grad_norm": 2.2060294151306152,
      "learning_rate": 9.285995500827348e-06,
      "loss": 0.4641,
      "num_input_tokens_seen": 5602488,
      "step": 8650
    },
    {
      "epoch": 5.103183962264151,
      "grad_norm": 2.826575756072998,
      "learning_rate": 9.284670023961355e-06,
      "loss": 0.4722,
      "num_input_tokens_seen": 5605112,
      "step": 8655
    },
    {
      "epoch": 5.106132075471698,
      "grad_norm": 2.0179758071899414,
      "learning_rate": 9.28334341271051e-06,
      "loss": 0.6024,
      "num_input_tokens_seen": 5607928,
      "step": 8660
    },
    {
      "epoch": 5.109080188679245,
      "grad_norm": 3.5897486209869385,
      "learning_rate": 9.282015667426036e-06,
      "loss": 0.5425,
      "num_input_tokens_seen": 5610872,
      "step": 8665
    },
    {
      "epoch": 5.1120283018867925,
      "grad_norm": 1.6922554969787598,
      "learning_rate": 9.280686788459461e-06,
      "loss": 0.53,
      "num_input_tokens_seen": 5613688,
      "step": 8670
    },
    {
      "epoch": 5.11497641509434,
      "grad_norm": 0.9314220547676086,
      "learning_rate": 9.279356776162606e-06,
      "loss": 0.3285,
      "num_input_tokens_seen": 5616920,
      "step": 8675
    },
    {
      "epoch": 5.117924528301887,
      "grad_norm": 1.9192028045654297,
      "learning_rate": 9.278025630887607e-06,
      "loss": 0.5637,
      "num_input_tokens_seen": 5619384,
      "step": 8680
    },
    {
      "epoch": 5.120872641509434,
      "grad_norm": 3.061061382293701,
      "learning_rate": 9.27669335298688e-06,
      "loss": 0.5038,
      "num_input_tokens_seen": 5622968,
      "step": 8685
    },
    {
      "epoch": 5.123820754716981,
      "grad_norm": 2.89691162109375,
      "learning_rate": 9.275359942813158e-06,
      "loss": 0.6274,
      "num_input_tokens_seen": 5625976,
      "step": 8690
    },
    {
      "epoch": 5.126768867924528,
      "grad_norm": 1.5589925050735474,
      "learning_rate": 9.274025400719466e-06,
      "loss": 0.429,
      "num_input_tokens_seen": 5629144,
      "step": 8695
    },
    {
      "epoch": 5.129716981132075,
      "grad_norm": 2.9099485874176025,
      "learning_rate": 9.27268972705913e-06,
      "loss": 0.4567,
      "num_input_tokens_seen": 5631800,
      "step": 8700
    },
    {
      "epoch": 5.132665094339623,
      "grad_norm": 2.544900894165039,
      "learning_rate": 9.271352922185772e-06,
      "loss": 0.5844,
      "num_input_tokens_seen": 5635192,
      "step": 8705
    },
    {
      "epoch": 5.13561320754717,
      "grad_norm": 1.7576837539672852,
      "learning_rate": 9.270014986453321e-06,
      "loss": 0.6766,
      "num_input_tokens_seen": 5639224,
      "step": 8710
    },
    {
      "epoch": 5.138561320754717,
      "grad_norm": 1.5105658769607544,
      "learning_rate": 9.268675920215999e-06,
      "loss": 0.4567,
      "num_input_tokens_seen": 5642616,
      "step": 8715
    },
    {
      "epoch": 5.1415094339622645,
      "grad_norm": 2.8113317489624023,
      "learning_rate": 9.26733572382833e-06,
      "loss": 0.5791,
      "num_input_tokens_seen": 5644888,
      "step": 8720
    },
    {
      "epoch": 5.144457547169812,
      "grad_norm": 1.2989709377288818,
      "learning_rate": 9.265994397645137e-06,
      "loss": 0.3816,
      "num_input_tokens_seen": 5647928,
      "step": 8725
    },
    {
      "epoch": 5.147405660377358,
      "grad_norm": 1.6023590564727783,
      "learning_rate": 9.264651942021543e-06,
      "loss": 0.4191,
      "num_input_tokens_seen": 5650904,
      "step": 8730
    },
    {
      "epoch": 5.150353773584905,
      "grad_norm": 1.8267196416854858,
      "learning_rate": 9.263308357312966e-06,
      "loss": 0.4008,
      "num_input_tokens_seen": 5653432,
      "step": 8735
    },
    {
      "epoch": 5.153301886792453,
      "grad_norm": 2.0687100887298584,
      "learning_rate": 9.26196364387513e-06,
      "loss": 0.5377,
      "num_input_tokens_seen": 5661816,
      "step": 8740
    },
    {
      "epoch": 5.15625,
      "grad_norm": 2.468324661254883,
      "learning_rate": 9.26061780206405e-06,
      "loss": 0.4693,
      "num_input_tokens_seen": 5664632,
      "step": 8745
    },
    {
      "epoch": 5.159198113207547,
      "grad_norm": 1.1225111484527588,
      "learning_rate": 9.259270832236043e-06,
      "loss": 0.4154,
      "num_input_tokens_seen": 5669144,
      "step": 8750
    },
    {
      "epoch": 5.162146226415095,
      "grad_norm": 1.359034538269043,
      "learning_rate": 9.257922734747729e-06,
      "loss": 0.4854,
      "num_input_tokens_seen": 5671960,
      "step": 8755
    },
    {
      "epoch": 5.165094339622642,
      "grad_norm": 1.427986979484558,
      "learning_rate": 9.256573509956018e-06,
      "loss": 0.622,
      "num_input_tokens_seen": 5676536,
      "step": 8760
    },
    {
      "epoch": 5.168042452830188,
      "grad_norm": 2.6347427368164062,
      "learning_rate": 9.255223158218127e-06,
      "loss": 0.4503,
      "num_input_tokens_seen": 5679736,
      "step": 8765
    },
    {
      "epoch": 5.1709905660377355,
      "grad_norm": 2.5362133979797363,
      "learning_rate": 9.253871679891566e-06,
      "loss": 0.4899,
      "num_input_tokens_seen": 5682072,
      "step": 8770
    },
    {
      "epoch": 5.173938679245283,
      "grad_norm": 1.4989339113235474,
      "learning_rate": 9.252519075334143e-06,
      "loss": 0.3629,
      "num_input_tokens_seen": 5685112,
      "step": 8775
    },
    {
      "epoch": 5.17688679245283,
      "grad_norm": 2.064222812652588,
      "learning_rate": 9.251165344903969e-06,
      "loss": 0.5706,
      "num_input_tokens_seen": 5689176,
      "step": 8780
    },
    {
      "epoch": 5.179834905660377,
      "grad_norm": 1.260477066040039,
      "learning_rate": 9.249810488959448e-06,
      "loss": 0.4035,
      "num_input_tokens_seen": 5692088,
      "step": 8785
    },
    {
      "epoch": 5.182783018867925,
      "grad_norm": 2.0889360904693604,
      "learning_rate": 9.248454507859285e-06,
      "loss": 0.4567,
      "num_input_tokens_seen": 5695672,
      "step": 8790
    },
    {
      "epoch": 5.185731132075472,
      "grad_norm": 2.279883623123169,
      "learning_rate": 9.247097401962482e-06,
      "loss": 0.3555,
      "num_input_tokens_seen": 5700568,
      "step": 8795
    },
    {
      "epoch": 5.188679245283019,
      "grad_norm": 1.3687242269515991,
      "learning_rate": 9.245739171628335e-06,
      "loss": 0.4842,
      "num_input_tokens_seen": 5703576,
      "step": 8800
    },
    {
      "epoch": 5.191627358490566,
      "grad_norm": 1.8149954080581665,
      "learning_rate": 9.244379817216447e-06,
      "loss": 0.4286,
      "num_input_tokens_seen": 5706808,
      "step": 8805
    },
    {
      "epoch": 5.194575471698113,
      "grad_norm": 2.84779691696167,
      "learning_rate": 9.243019339086708e-06,
      "loss": 0.5396,
      "num_input_tokens_seen": 5710488,
      "step": 8810
    },
    {
      "epoch": 5.19752358490566,
      "grad_norm": 2.6494760513305664,
      "learning_rate": 9.241657737599313e-06,
      "loss": 0.4203,
      "num_input_tokens_seen": 5713784,
      "step": 8815
    },
    {
      "epoch": 5.2004716981132075,
      "grad_norm": 1.5766946077346802,
      "learning_rate": 9.240295013114752e-06,
      "loss": 0.4473,
      "num_input_tokens_seen": 5716760,
      "step": 8820
    },
    {
      "epoch": 5.203419811320755,
      "grad_norm": 2.0030324459075928,
      "learning_rate": 9.238931165993811e-06,
      "loss": 0.4547,
      "num_input_tokens_seen": 5720088,
      "step": 8825
    },
    {
      "epoch": 5.206367924528302,
      "grad_norm": 2.159402847290039,
      "learning_rate": 9.237566196597577e-06,
      "loss": 0.5655,
      "num_input_tokens_seen": 5723224,
      "step": 8830
    },
    {
      "epoch": 5.209316037735849,
      "grad_norm": 1.8406461477279663,
      "learning_rate": 9.236200105287427e-06,
      "loss": 0.5138,
      "num_input_tokens_seen": 5726072,
      "step": 8835
    },
    {
      "epoch": 5.212264150943396,
      "grad_norm": 1.809219241142273,
      "learning_rate": 9.234832892425042e-06,
      "loss": 0.4849,
      "num_input_tokens_seen": 5728568,
      "step": 8840
    },
    {
      "epoch": 5.215212264150943,
      "grad_norm": 1.9482074975967407,
      "learning_rate": 9.2334645583724e-06,
      "loss": 0.5108,
      "num_input_tokens_seen": 5732312,
      "step": 8845
    },
    {
      "epoch": 5.21816037735849,
      "grad_norm": 1.6626592874526978,
      "learning_rate": 9.23209510349177e-06,
      "loss": 0.4527,
      "num_input_tokens_seen": 5735000,
      "step": 8850
    },
    {
      "epoch": 5.221108490566038,
      "grad_norm": 2.1777360439300537,
      "learning_rate": 9.230724528145722e-06,
      "loss": 0.4812,
      "num_input_tokens_seen": 5739352,
      "step": 8855
    },
    {
      "epoch": 5.224056603773585,
      "grad_norm": 1.4084886312484741,
      "learning_rate": 9.229352832697122e-06,
      "loss": 0.3846,
      "num_input_tokens_seen": 5742296,
      "step": 8860
    },
    {
      "epoch": 5.227004716981132,
      "grad_norm": 3.515516757965088,
      "learning_rate": 9.22798001750913e-06,
      "loss": 0.5448,
      "num_input_tokens_seen": 5744888,
      "step": 8865
    },
    {
      "epoch": 5.2299528301886795,
      "grad_norm": 1.4127227067947388,
      "learning_rate": 9.226606082945209e-06,
      "loss": 0.4207,
      "num_input_tokens_seen": 5747960,
      "step": 8870
    },
    {
      "epoch": 5.232900943396227,
      "grad_norm": 1.5133355855941772,
      "learning_rate": 9.225231029369112e-06,
      "loss": 0.5126,
      "num_input_tokens_seen": 5751032,
      "step": 8875
    },
    {
      "epoch": 5.235849056603773,
      "grad_norm": 3.2830655574798584,
      "learning_rate": 9.22385485714489e-06,
      "loss": 0.5448,
      "num_input_tokens_seen": 5753816,
      "step": 8880
    },
    {
      "epoch": 5.2387971698113205,
      "grad_norm": 2.594119071960449,
      "learning_rate": 9.222477566636889e-06,
      "loss": 0.5866,
      "num_input_tokens_seen": 5756696,
      "step": 8885
    },
    {
      "epoch": 5.241745283018868,
      "grad_norm": 1.843015193939209,
      "learning_rate": 9.221099158209757e-06,
      "loss": 0.4351,
      "num_input_tokens_seen": 5759768,
      "step": 8890
    },
    {
      "epoch": 5.244693396226415,
      "grad_norm": 2.0978856086730957,
      "learning_rate": 9.219719632228429e-06,
      "loss": 0.4601,
      "num_input_tokens_seen": 5762616,
      "step": 8895
    },
    {
      "epoch": 5.247641509433962,
      "grad_norm": 1.689556360244751,
      "learning_rate": 9.218338989058141e-06,
      "loss": 0.5404,
      "num_input_tokens_seen": 5765528,
      "step": 8900
    },
    {
      "epoch": 5.25058962264151,
      "grad_norm": 3.0072169303894043,
      "learning_rate": 9.21695722906443e-06,
      "loss": 0.4101,
      "num_input_tokens_seen": 5768312,
      "step": 8905
    },
    {
      "epoch": 5.253537735849057,
      "grad_norm": 1.9534430503845215,
      "learning_rate": 9.215574352613115e-06,
      "loss": 0.4297,
      "num_input_tokens_seen": 5771384,
      "step": 8910
    },
    {
      "epoch": 5.256485849056604,
      "grad_norm": 2.127387046813965,
      "learning_rate": 9.214190360070323e-06,
      "loss": 0.5337,
      "num_input_tokens_seen": 5774776,
      "step": 8915
    },
    {
      "epoch": 5.259433962264151,
      "grad_norm": 1.5937278270721436,
      "learning_rate": 9.212805251802471e-06,
      "loss": 0.4727,
      "num_input_tokens_seen": 5777176,
      "step": 8920
    },
    {
      "epoch": 5.262382075471698,
      "grad_norm": 1.4854357242584229,
      "learning_rate": 9.211419028176273e-06,
      "loss": 0.3462,
      "num_input_tokens_seen": 5780568,
      "step": 8925
    },
    {
      "epoch": 5.265330188679245,
      "grad_norm": 2.084650993347168,
      "learning_rate": 9.210031689558738e-06,
      "loss": 0.3379,
      "num_input_tokens_seen": 5783736,
      "step": 8930
    },
    {
      "epoch": 5.2682783018867925,
      "grad_norm": 4.8332343101501465,
      "learning_rate": 9.208643236317166e-06,
      "loss": 0.7999,
      "num_input_tokens_seen": 5786904,
      "step": 8935
    },
    {
      "epoch": 5.27122641509434,
      "grad_norm": 2.179795026779175,
      "learning_rate": 9.20725366881916e-06,
      "loss": 0.3698,
      "num_input_tokens_seen": 5789944,
      "step": 8940
    },
    {
      "epoch": 5.274174528301887,
      "grad_norm": 2.5404276847839355,
      "learning_rate": 9.205862987432614e-06,
      "loss": 0.4693,
      "num_input_tokens_seen": 5793592,
      "step": 8945
    },
    {
      "epoch": 5.277122641509434,
      "grad_norm": 0.9461173415184021,
      "learning_rate": 9.204471192525715e-06,
      "loss": 0.3237,
      "num_input_tokens_seen": 5796984,
      "step": 8950
    },
    {
      "epoch": 5.280070754716981,
      "grad_norm": 2.061725378036499,
      "learning_rate": 9.203078284466949e-06,
      "loss": 0.5252,
      "num_input_tokens_seen": 5800632,
      "step": 8955
    },
    {
      "epoch": 5.283018867924528,
      "grad_norm": 2.1438491344451904,
      "learning_rate": 9.201684263625091e-06,
      "loss": 0.5651,
      "num_input_tokens_seen": 5803480,
      "step": 8960
    },
    {
      "epoch": 5.285966981132075,
      "grad_norm": 3.5597341060638428,
      "learning_rate": 9.200289130369218e-06,
      "loss": 0.3642,
      "num_input_tokens_seen": 5805880,
      "step": 8965
    },
    {
      "epoch": 5.288915094339623,
      "grad_norm": 2.3908681869506836,
      "learning_rate": 9.198892885068693e-06,
      "loss": 0.499,
      "num_input_tokens_seen": 5808504,
      "step": 8970
    },
    {
      "epoch": 5.29186320754717,
      "grad_norm": 0.8922989964485168,
      "learning_rate": 9.197495528093182e-06,
      "loss": 0.4083,
      "num_input_tokens_seen": 5812056,
      "step": 8975
    },
    {
      "epoch": 5.294811320754717,
      "grad_norm": 1.8601710796356201,
      "learning_rate": 9.196097059812639e-06,
      "loss": 0.4115,
      "num_input_tokens_seen": 5814904,
      "step": 8980
    },
    {
      "epoch": 5.2977594339622645,
      "grad_norm": 2.5759003162384033,
      "learning_rate": 9.194697480597316e-06,
      "loss": 0.4593,
      "num_input_tokens_seen": 5817848,
      "step": 8985
    },
    {
      "epoch": 5.300707547169811,
      "grad_norm": 1.4591844081878662,
      "learning_rate": 9.193296790817755e-06,
      "loss": 0.425,
      "num_input_tokens_seen": 5822488,
      "step": 8990
    },
    {
      "epoch": 5.303655660377358,
      "grad_norm": 1.8231233358383179,
      "learning_rate": 9.1918949908448e-06,
      "loss": 0.425,
      "num_input_tokens_seen": 5825400,
      "step": 8995
    },
    {
      "epoch": 5.306603773584905,
      "grad_norm": 2.373771905899048,
      "learning_rate": 9.190492081049578e-06,
      "loss": 0.5651,
      "num_input_tokens_seen": 5828504,
      "step": 9000
    },
    {
      "epoch": 5.309551886792453,
      "grad_norm": 2.07875394821167,
      "learning_rate": 9.189088061803517e-06,
      "loss": 0.3738,
      "num_input_tokens_seen": 5831192,
      "step": 9005
    },
    {
      "epoch": 5.3125,
      "grad_norm": 5.226472854614258,
      "learning_rate": 9.187682933478337e-06,
      "loss": 0.4794,
      "num_input_tokens_seen": 5833656,
      "step": 9010
    },
    {
      "epoch": 5.315448113207547,
      "grad_norm": 1.8028384447097778,
      "learning_rate": 9.186276696446054e-06,
      "loss": 0.4201,
      "num_input_tokens_seen": 5837048,
      "step": 9015
    },
    {
      "epoch": 5.318396226415095,
      "grad_norm": 1.4793426990509033,
      "learning_rate": 9.184869351078974e-06,
      "loss": 0.399,
      "num_input_tokens_seen": 5840824,
      "step": 9020
    },
    {
      "epoch": 5.321344339622642,
      "grad_norm": 3.41335391998291,
      "learning_rate": 9.183460897749697e-06,
      "loss": 0.3788,
      "num_input_tokens_seen": 5844248,
      "step": 9025
    },
    {
      "epoch": 5.324292452830189,
      "grad_norm": 2.357424736022949,
      "learning_rate": 9.182051336831117e-06,
      "loss": 0.5755,
      "num_input_tokens_seen": 5847576,
      "step": 9030
    },
    {
      "epoch": 5.3272405660377355,
      "grad_norm": 1.5002765655517578,
      "learning_rate": 9.180640668696424e-06,
      "loss": 0.4641,
      "num_input_tokens_seen": 5849944,
      "step": 9035
    },
    {
      "epoch": 5.330188679245283,
      "grad_norm": 1.5500833988189697,
      "learning_rate": 9.179228893719094e-06,
      "loss": 0.3552,
      "num_input_tokens_seen": 5852696,
      "step": 9040
    },
    {
      "epoch": 5.33313679245283,
      "grad_norm": 2.4839935302734375,
      "learning_rate": 9.177816012272904e-06,
      "loss": 0.4241,
      "num_input_tokens_seen": 5855288,
      "step": 9045
    },
    {
      "epoch": 5.336084905660377,
      "grad_norm": 2.7999989986419678,
      "learning_rate": 9.17640202473192e-06,
      "loss": 0.5412,
      "num_input_tokens_seen": 5857816,
      "step": 9050
    },
    {
      "epoch": 5.339033018867925,
      "grad_norm": 1.8636738061904907,
      "learning_rate": 9.1749869314705e-06,
      "loss": 0.3425,
      "num_input_tokens_seen": 5860088,
      "step": 9055
    },
    {
      "epoch": 5.341981132075472,
      "grad_norm": 10.658321380615234,
      "learning_rate": 9.173570732863295e-06,
      "loss": 0.4483,
      "num_input_tokens_seen": 5863704,
      "step": 9060
    },
    {
      "epoch": 5.344929245283019,
      "grad_norm": 1.9576623439788818,
      "learning_rate": 9.172153429285254e-06,
      "loss": 0.4861,
      "num_input_tokens_seen": 5867128,
      "step": 9065
    },
    {
      "epoch": 5.347877358490566,
      "grad_norm": 2.20103120803833,
      "learning_rate": 9.17073502111161e-06,
      "loss": 0.3708,
      "num_input_tokens_seen": 5869944,
      "step": 9070
    },
    {
      "epoch": 5.350825471698113,
      "grad_norm": 1.8489491939544678,
      "learning_rate": 9.169315508717895e-06,
      "loss": 0.4347,
      "num_input_tokens_seen": 5873176,
      "step": 9075
    },
    {
      "epoch": 5.35377358490566,
      "grad_norm": 2.422964572906494,
      "learning_rate": 9.167894892479932e-06,
      "loss": 0.4442,
      "num_input_tokens_seen": 5876088,
      "step": 9080
    },
    {
      "epoch": 5.3567216981132075,
      "grad_norm": 2.9419193267822266,
      "learning_rate": 9.16647317277383e-06,
      "loss": 0.5333,
      "num_input_tokens_seen": 5879224,
      "step": 9085
    },
    {
      "epoch": 5.359669811320755,
      "grad_norm": 1.595181941986084,
      "learning_rate": 9.165050349976002e-06,
      "loss": 0.3965,
      "num_input_tokens_seen": 5883800,
      "step": 9090
    },
    {
      "epoch": 5.362617924528302,
      "grad_norm": 3.180962324142456,
      "learning_rate": 9.16362642446314e-06,
      "loss": 0.49,
      "num_input_tokens_seen": 5886584,
      "step": 9095
    },
    {
      "epoch": 5.365566037735849,
      "grad_norm": 2.2751801013946533,
      "learning_rate": 9.162201396612242e-06,
      "loss": 0.5259,
      "num_input_tokens_seen": 5889656,
      "step": 9100
    },
    {
      "epoch": 5.368514150943396,
      "grad_norm": 2.1908555030822754,
      "learning_rate": 9.160775266800583e-06,
      "loss": 0.4489,
      "num_input_tokens_seen": 5893528,
      "step": 9105
    },
    {
      "epoch": 5.371462264150943,
      "grad_norm": 2.3496134281158447,
      "learning_rate": 9.159348035405742e-06,
      "loss": 0.4113,
      "num_input_tokens_seen": 5896760,
      "step": 9110
    },
    {
      "epoch": 5.37441037735849,
      "grad_norm": 1.3614768981933594,
      "learning_rate": 9.157919702805582e-06,
      "loss": 0.5484,
      "num_input_tokens_seen": 5900472,
      "step": 9115
    },
    {
      "epoch": 5.377358490566038,
      "grad_norm": 2.5091919898986816,
      "learning_rate": 9.156490269378262e-06,
      "loss": 0.364,
      "num_input_tokens_seen": 5903288,
      "step": 9120
    },
    {
      "epoch": 5.380306603773585,
      "grad_norm": 2.161557197570801,
      "learning_rate": 9.15505973550223e-06,
      "loss": 0.4917,
      "num_input_tokens_seen": 5906296,
      "step": 9125
    },
    {
      "epoch": 5.383254716981132,
      "grad_norm": 1.875571370124817,
      "learning_rate": 9.153628101556223e-06,
      "loss": 0.5355,
      "num_input_tokens_seen": 5909464,
      "step": 9130
    },
    {
      "epoch": 5.3862028301886795,
      "grad_norm": 1.462533712387085,
      "learning_rate": 9.152195367919277e-06,
      "loss": 0.3807,
      "num_input_tokens_seen": 5913080,
      "step": 9135
    },
    {
      "epoch": 5.389150943396227,
      "grad_norm": 1.9098997116088867,
      "learning_rate": 9.150761534970713e-06,
      "loss": 0.4184,
      "num_input_tokens_seen": 5916120,
      "step": 9140
    },
    {
      "epoch": 5.392099056603773,
      "grad_norm": 3.1211917400360107,
      "learning_rate": 9.149326603090144e-06,
      "loss": 0.5096,
      "num_input_tokens_seen": 5918968,
      "step": 9145
    },
    {
      "epoch": 5.3950471698113205,
      "grad_norm": 1.4267536401748657,
      "learning_rate": 9.147890572657471e-06,
      "loss": 0.3719,
      "num_input_tokens_seen": 5922488,
      "step": 9150
    },
    {
      "epoch": 5.397995283018868,
      "grad_norm": 1.7294998168945312,
      "learning_rate": 9.146453444052895e-06,
      "loss": 0.4336,
      "num_input_tokens_seen": 5925560,
      "step": 9155
    },
    {
      "epoch": 5.400943396226415,
      "grad_norm": 1.6649200916290283,
      "learning_rate": 9.145015217656899e-06,
      "loss": 0.4159,
      "num_input_tokens_seen": 5928792,
      "step": 9160
    },
    {
      "epoch": 5.403891509433962,
      "grad_norm": 1.8008865118026733,
      "learning_rate": 9.14357589385026e-06,
      "loss": 0.4074,
      "num_input_tokens_seen": 5931768,
      "step": 9165
    },
    {
      "epoch": 5.40683962264151,
      "grad_norm": 1.295956015586853,
      "learning_rate": 9.142135473014046e-06,
      "loss": 0.4106,
      "num_input_tokens_seen": 5935096,
      "step": 9170
    },
    {
      "epoch": 5.409787735849057,
      "grad_norm": 1.0460937023162842,
      "learning_rate": 9.140693955529614e-06,
      "loss": 0.2991,
      "num_input_tokens_seen": 5938104,
      "step": 9175
    },
    {
      "epoch": 5.412735849056604,
      "grad_norm": 2.473194122314453,
      "learning_rate": 9.13925134177861e-06,
      "loss": 0.527,
      "num_input_tokens_seen": 5940984,
      "step": 9180
    },
    {
      "epoch": 5.415683962264151,
      "grad_norm": 2.9737188816070557,
      "learning_rate": 9.137807632142977e-06,
      "loss": 0.5117,
      "num_input_tokens_seen": 5945464,
      "step": 9185
    },
    {
      "epoch": 5.418632075471698,
      "grad_norm": 2.822744131088257,
      "learning_rate": 9.136362827004937e-06,
      "loss": 0.506,
      "num_input_tokens_seen": 5948632,
      "step": 9190
    },
    {
      "epoch": 5.421580188679245,
      "grad_norm": 2.2328429222106934,
      "learning_rate": 9.134916926747015e-06,
      "loss": 0.4406,
      "num_input_tokens_seen": 5951480,
      "step": 9195
    },
    {
      "epoch": 5.4245283018867925,
      "grad_norm": 4.993492603302002,
      "learning_rate": 9.133469931752016e-06,
      "loss": 0.4329,
      "num_input_tokens_seen": 5954776,
      "step": 9200
    },
    {
      "epoch": 5.42747641509434,
      "grad_norm": 2.270702838897705,
      "learning_rate": 9.132021842403035e-06,
      "loss": 0.5478,
      "num_input_tokens_seen": 5959160,
      "step": 9205
    },
    {
      "epoch": 5.430424528301887,
      "grad_norm": 1.959314227104187,
      "learning_rate": 9.130572659083465e-06,
      "loss": 0.3742,
      "num_input_tokens_seen": 5962968,
      "step": 9210
    },
    {
      "epoch": 5.433372641509434,
      "grad_norm": 1.3733329772949219,
      "learning_rate": 9.129122382176982e-06,
      "loss": 0.5388,
      "num_input_tokens_seen": 5966584,
      "step": 9215
    },
    {
      "epoch": 5.436320754716981,
      "grad_norm": 0.4984431862831116,
      "learning_rate": 9.127671012067554e-06,
      "loss": 0.4402,
      "num_input_tokens_seen": 5971864,
      "step": 9220
    },
    {
      "epoch": 5.439268867924528,
      "grad_norm": 1.6374444961547852,
      "learning_rate": 9.126218549139434e-06,
      "loss": 0.3722,
      "num_input_tokens_seen": 5975448,
      "step": 9225
    },
    {
      "epoch": 5.442216981132075,
      "grad_norm": 1.4114530086517334,
      "learning_rate": 9.124764993777171e-06,
      "loss": 0.3574,
      "num_input_tokens_seen": 5979224,
      "step": 9230
    },
    {
      "epoch": 5.445165094339623,
      "grad_norm": 1.800034761428833,
      "learning_rate": 9.1233103463656e-06,
      "loss": 0.4364,
      "num_input_tokens_seen": 5983960,
      "step": 9235
    },
    {
      "epoch": 5.44811320754717,
      "grad_norm": 2.8349766731262207,
      "learning_rate": 9.121854607289842e-06,
      "loss": 0.6284,
      "num_input_tokens_seen": 5988280,
      "step": 9240
    },
    {
      "epoch": 5.451061320754717,
      "grad_norm": 2.9296252727508545,
      "learning_rate": 9.120397776935314e-06,
      "loss": 0.4437,
      "num_input_tokens_seen": 5991544,
      "step": 9245
    },
    {
      "epoch": 5.4540094339622645,
      "grad_norm": 2.7815840244293213,
      "learning_rate": 9.118939855687717e-06,
      "loss": 0.3893,
      "num_input_tokens_seen": 5994904,
      "step": 9250
    },
    {
      "epoch": 5.456957547169811,
      "grad_norm": 1.6625012159347534,
      "learning_rate": 9.117480843933043e-06,
      "loss": 0.4437,
      "num_input_tokens_seen": 5997624,
      "step": 9255
    },
    {
      "epoch": 5.459905660377358,
      "grad_norm": 2.166471004486084,
      "learning_rate": 9.116020742057567e-06,
      "loss": 0.4195,
      "num_input_tokens_seen": 6001080,
      "step": 9260
    },
    {
      "epoch": 5.462853773584905,
      "grad_norm": 1.6249734163284302,
      "learning_rate": 9.114559550447863e-06,
      "loss": 0.4941,
      "num_input_tokens_seen": 6003256,
      "step": 9265
    },
    {
      "epoch": 5.465801886792453,
      "grad_norm": 2.3121440410614014,
      "learning_rate": 9.113097269490784e-06,
      "loss": 0.464,
      "num_input_tokens_seen": 6006392,
      "step": 9270
    },
    {
      "epoch": 5.46875,
      "grad_norm": 1.6128886938095093,
      "learning_rate": 9.111633899573476e-06,
      "loss": 0.5623,
      "num_input_tokens_seen": 6009752,
      "step": 9275
    },
    {
      "epoch": 5.471698113207547,
      "grad_norm": 2.204542636871338,
      "learning_rate": 9.110169441083374e-06,
      "loss": 0.6267,
      "num_input_tokens_seen": 6012760,
      "step": 9280
    },
    {
      "epoch": 5.474646226415095,
      "grad_norm": 1.5607671737670898,
      "learning_rate": 9.108703894408198e-06,
      "loss": 0.4881,
      "num_input_tokens_seen": 6015128,
      "step": 9285
    },
    {
      "epoch": 5.477594339622642,
      "grad_norm": 2.3435556888580322,
      "learning_rate": 9.107237259935959e-06,
      "loss": 0.4386,
      "num_input_tokens_seen": 6018104,
      "step": 9290
    },
    {
      "epoch": 5.480542452830189,
      "grad_norm": 2.7039620876312256,
      "learning_rate": 9.105769538054954e-06,
      "loss": 0.547,
      "num_input_tokens_seen": 6021720,
      "step": 9295
    },
    {
      "epoch": 5.4834905660377355,
      "grad_norm": 3.427825450897217,
      "learning_rate": 9.104300729153768e-06,
      "loss": 0.6021,
      "num_input_tokens_seen": 6029080,
      "step": 9300
    },
    {
      "epoch": 5.486438679245283,
      "grad_norm": 1.492091178894043,
      "learning_rate": 9.102830833621277e-06,
      "loss": 0.3712,
      "num_input_tokens_seen": 6032600,
      "step": 9305
    },
    {
      "epoch": 5.48938679245283,
      "grad_norm": 2.1670589447021484,
      "learning_rate": 9.101359851846639e-06,
      "loss": 0.4273,
      "num_input_tokens_seen": 6036312,
      "step": 9310
    },
    {
      "epoch": 5.492334905660377,
      "grad_norm": 1.6973830461502075,
      "learning_rate": 9.099887784219305e-06,
      "loss": 0.3989,
      "num_input_tokens_seen": 6039000,
      "step": 9315
    },
    {
      "epoch": 5.495283018867925,
      "grad_norm": 3.4042747020721436,
      "learning_rate": 9.098414631129012e-06,
      "loss": 0.4293,
      "num_input_tokens_seen": 6042552,
      "step": 9320
    },
    {
      "epoch": 5.498231132075472,
      "grad_norm": 1.2347184419631958,
      "learning_rate": 9.09694039296578e-06,
      "loss": 0.4525,
      "num_input_tokens_seen": 6045720,
      "step": 9325
    },
    {
      "epoch": 5.501179245283019,
      "grad_norm": 2.2299232482910156,
      "learning_rate": 9.095465070119924e-06,
      "loss": 0.6768,
      "num_input_tokens_seen": 6048248,
      "step": 9330
    },
    {
      "epoch": 5.504127358490566,
      "grad_norm": 2.3229727745056152,
      "learning_rate": 9.09398866298204e-06,
      "loss": 0.4644,
      "num_input_tokens_seen": 6051064,
      "step": 9335
    },
    {
      "epoch": 5.507075471698113,
      "grad_norm": 1.82985520362854,
      "learning_rate": 9.09251117194301e-06,
      "loss": 0.4084,
      "num_input_tokens_seen": 6053784,
      "step": 9340
    },
    {
      "epoch": 5.51002358490566,
      "grad_norm": 3.6390230655670166,
      "learning_rate": 9.091032597394012e-06,
      "loss": 0.3638,
      "num_input_tokens_seen": 6056632,
      "step": 9345
    },
    {
      "epoch": 5.5129716981132075,
      "grad_norm": 1.4377864599227905,
      "learning_rate": 9.089552939726503e-06,
      "loss": 0.4021,
      "num_input_tokens_seen": 6059672,
      "step": 9350
    },
    {
      "epoch": 5.515919811320755,
      "grad_norm": 2.7718324661254883,
      "learning_rate": 9.088072199332227e-06,
      "loss": 0.4558,
      "num_input_tokens_seen": 6062680,
      "step": 9355
    },
    {
      "epoch": 5.518867924528302,
      "grad_norm": 1.0528831481933594,
      "learning_rate": 9.08659037660322e-06,
      "loss": 0.4617,
      "num_input_tokens_seen": 6067032,
      "step": 9360
    },
    {
      "epoch": 5.521816037735849,
      "grad_norm": 2.3537251949310303,
      "learning_rate": 9.085107471931797e-06,
      "loss": 0.3634,
      "num_input_tokens_seen": 6070232,
      "step": 9365
    },
    {
      "epoch": 5.524764150943396,
      "grad_norm": 1.4396177530288696,
      "learning_rate": 9.083623485710564e-06,
      "loss": 0.3246,
      "num_input_tokens_seen": 6075000,
      "step": 9370
    },
    {
      "epoch": 5.527712264150943,
      "grad_norm": 1.6855981349945068,
      "learning_rate": 9.082138418332416e-06,
      "loss": 0.4679,
      "num_input_tokens_seen": 6079160,
      "step": 9375
    },
    {
      "epoch": 5.53066037735849,
      "grad_norm": 2.9409139156341553,
      "learning_rate": 9.080652270190527e-06,
      "loss": 0.4679,
      "num_input_tokens_seen": 6081912,
      "step": 9380
    },
    {
      "epoch": 5.533608490566038,
      "grad_norm": 2.543610095977783,
      "learning_rate": 9.079165041678363e-06,
      "loss": 0.4295,
      "num_input_tokens_seen": 6085112,
      "step": 9385
    },
    {
      "epoch": 5.536556603773585,
      "grad_norm": 1.7721068859100342,
      "learning_rate": 9.077676733189675e-06,
      "loss": 0.4419,
      "num_input_tokens_seen": 6089496,
      "step": 9390
    },
    {
      "epoch": 5.539504716981132,
      "grad_norm": 1.6604127883911133,
      "learning_rate": 9.076187345118496e-06,
      "loss": 0.48,
      "num_input_tokens_seen": 6092536,
      "step": 9395
    },
    {
      "epoch": 5.5424528301886795,
      "grad_norm": 2.8636796474456787,
      "learning_rate": 9.074696877859152e-06,
      "loss": 0.3562,
      "num_input_tokens_seen": 6095096,
      "step": 9400
    },
    {
      "epoch": 5.545400943396227,
      "grad_norm": 1.9150499105453491,
      "learning_rate": 9.073205331806248e-06,
      "loss": 0.4344,
      "num_input_tokens_seen": 6098008,
      "step": 9405
    },
    {
      "epoch": 5.548349056603773,
      "grad_norm": 1.5624960660934448,
      "learning_rate": 9.071712707354676e-06,
      "loss": 0.4718,
      "num_input_tokens_seen": 6100632,
      "step": 9410
    },
    {
      "epoch": 5.5512971698113205,
      "grad_norm": 2.363429546356201,
      "learning_rate": 9.070219004899618e-06,
      "loss": 0.4123,
      "num_input_tokens_seen": 6103448,
      "step": 9415
    },
    {
      "epoch": 5.554245283018868,
      "grad_norm": 2.751859426498413,
      "learning_rate": 9.068724224836538e-06,
      "loss": 0.5257,
      "num_input_tokens_seen": 6106520,
      "step": 9420
    },
    {
      "epoch": 5.557193396226415,
      "grad_norm": 2.3989901542663574,
      "learning_rate": 9.067228367561182e-06,
      "loss": 0.5707,
      "num_input_tokens_seen": 6109624,
      "step": 9425
    },
    {
      "epoch": 5.560141509433962,
      "grad_norm": 2.241774559020996,
      "learning_rate": 9.06573143346959e-06,
      "loss": 0.47,
      "num_input_tokens_seen": 6114232,
      "step": 9430
    },
    {
      "epoch": 5.56308962264151,
      "grad_norm": 1.4339046478271484,
      "learning_rate": 9.064233422958078e-06,
      "loss": 0.3819,
      "num_input_tokens_seen": 6117880,
      "step": 9435
    },
    {
      "epoch": 5.566037735849057,
      "grad_norm": 2.1893632411956787,
      "learning_rate": 9.062734336423248e-06,
      "loss": 0.3726,
      "num_input_tokens_seen": 6120600,
      "step": 9440
    },
    {
      "epoch": 5.568985849056604,
      "grad_norm": 3.5702106952667236,
      "learning_rate": 9.061234174261998e-06,
      "loss": 0.4138,
      "num_input_tokens_seen": 6123800,
      "step": 9445
    },
    {
      "epoch": 5.571933962264151,
      "grad_norm": 1.2458642721176147,
      "learning_rate": 9.059732936871493e-06,
      "loss": 0.42,
      "num_input_tokens_seen": 6126776,
      "step": 9450
    },
    {
      "epoch": 5.574882075471698,
      "grad_norm": 0.8931108713150024,
      "learning_rate": 9.058230624649198e-06,
      "loss": 0.3575,
      "num_input_tokens_seen": 6130424,
      "step": 9455
    },
    {
      "epoch": 5.577830188679245,
      "grad_norm": 1.6845754384994507,
      "learning_rate": 9.056727237992856e-06,
      "loss": 0.3276,
      "num_input_tokens_seen": 6133784,
      "step": 9460
    },
    {
      "epoch": 5.5807783018867925,
      "grad_norm": 4.332545280456543,
      "learning_rate": 9.055222777300493e-06,
      "loss": 0.4511,
      "num_input_tokens_seen": 6136440,
      "step": 9465
    },
    {
      "epoch": 5.58372641509434,
      "grad_norm": 3.006453037261963,
      "learning_rate": 9.053717242970423e-06,
      "loss": 0.6658,
      "num_input_tokens_seen": 6140472,
      "step": 9470
    },
    {
      "epoch": 5.586674528301887,
      "grad_norm": 2.192384719848633,
      "learning_rate": 9.052210635401244e-06,
      "loss": 0.3246,
      "num_input_tokens_seen": 6144888,
      "step": 9475
    },
    {
      "epoch": 5.589622641509434,
      "grad_norm": 3.1968889236450195,
      "learning_rate": 9.050702954991833e-06,
      "loss": 0.5342,
      "num_input_tokens_seen": 6148376,
      "step": 9480
    },
    {
      "epoch": 5.592570754716981,
      "grad_norm": 2.357351303100586,
      "learning_rate": 9.049194202141358e-06,
      "loss": 0.4809,
      "num_input_tokens_seen": 6151288,
      "step": 9485
    },
    {
      "epoch": 5.595518867924528,
      "grad_norm": 1.5530521869659424,
      "learning_rate": 9.047684377249267e-06,
      "loss": 0.4388,
      "num_input_tokens_seen": 6154648,
      "step": 9490
    },
    {
      "epoch": 5.598466981132075,
      "grad_norm": 3.208279848098755,
      "learning_rate": 9.046173480715292e-06,
      "loss": 0.4148,
      "num_input_tokens_seen": 6157656,
      "step": 9495
    },
    {
      "epoch": 5.601415094339623,
      "grad_norm": 1.304370403289795,
      "learning_rate": 9.044661512939451e-06,
      "loss": 0.4711,
      "num_input_tokens_seen": 6161528,
      "step": 9500
    },
    {
      "epoch": 5.60436320754717,
      "grad_norm": 2.089005947113037,
      "learning_rate": 9.043148474322043e-06,
      "loss": 0.4698,
      "num_input_tokens_seen": 6164472,
      "step": 9505
    },
    {
      "epoch": 5.607311320754717,
      "grad_norm": 1.6335675716400146,
      "learning_rate": 9.04163436526365e-06,
      "loss": 0.339,
      "num_input_tokens_seen": 6168696,
      "step": 9510
    },
    {
      "epoch": 5.6102594339622645,
      "grad_norm": 1.9193450212478638,
      "learning_rate": 9.040119186165142e-06,
      "loss": 0.3629,
      "num_input_tokens_seen": 6173528,
      "step": 9515
    },
    {
      "epoch": 5.613207547169811,
      "grad_norm": 2.2783281803131104,
      "learning_rate": 9.038602937427665e-06,
      "loss": 0.6024,
      "num_input_tokens_seen": 6176376,
      "step": 9520
    },
    {
      "epoch": 5.616155660377358,
      "grad_norm": 2.619225025177002,
      "learning_rate": 9.037085619452658e-06,
      "loss": 0.5244,
      "num_input_tokens_seen": 6179544,
      "step": 9525
    },
    {
      "epoch": 5.619103773584905,
      "grad_norm": 2.1083433628082275,
      "learning_rate": 9.035567232641833e-06,
      "loss": 0.4901,
      "num_input_tokens_seen": 6184728,
      "step": 9530
    },
    {
      "epoch": 5.622051886792453,
      "grad_norm": 2.2404136657714844,
      "learning_rate": 9.03404777739719e-06,
      "loss": 0.4162,
      "num_input_tokens_seen": 6187480,
      "step": 9535
    },
    {
      "epoch": 5.625,
      "grad_norm": 1.3799391984939575,
      "learning_rate": 9.032527254121013e-06,
      "loss": 0.4714,
      "num_input_tokens_seen": 6190200,
      "step": 9540
    },
    {
      "epoch": 5.627948113207547,
      "grad_norm": 4.560179710388184,
      "learning_rate": 9.031005663215867e-06,
      "loss": 0.5155,
      "num_input_tokens_seen": 6192664,
      "step": 9545
    },
    {
      "epoch": 5.630896226415095,
      "grad_norm": 2.1808433532714844,
      "learning_rate": 9.029483005084595e-06,
      "loss": 0.6237,
      "num_input_tokens_seen": 6196632,
      "step": 9550
    },
    {
      "epoch": 5.633844339622642,
      "grad_norm": 1.9176064729690552,
      "learning_rate": 9.027959280130337e-06,
      "loss": 0.5093,
      "num_input_tokens_seen": 6199864,
      "step": 9555
    },
    {
      "epoch": 5.636792452830189,
      "grad_norm": 2.649477243423462,
      "learning_rate": 9.026434488756496e-06,
      "loss": 0.3987,
      "num_input_tokens_seen": 6202296,
      "step": 9560
    },
    {
      "epoch": 5.6397405660377355,
      "grad_norm": 1.3461639881134033,
      "learning_rate": 9.024908631366774e-06,
      "loss": 0.5024,
      "num_input_tokens_seen": 6205528,
      "step": 9565
    },
    {
      "epoch": 5.642688679245283,
      "grad_norm": 2.539677858352661,
      "learning_rate": 9.023381708365143e-06,
      "loss": 0.4099,
      "num_input_tokens_seen": 6207736,
      "step": 9570
    },
    {
      "epoch": 5.64563679245283,
      "grad_norm": 2.058901786804199,
      "learning_rate": 9.021853720155866e-06,
      "loss": 0.5301,
      "num_input_tokens_seen": 6210808,
      "step": 9575
    },
    {
      "epoch": 5.648584905660377,
      "grad_norm": 1.5368379354476929,
      "learning_rate": 9.020324667143483e-06,
      "loss": 0.3507,
      "num_input_tokens_seen": 6213720,
      "step": 9580
    },
    {
      "epoch": 5.651533018867925,
      "grad_norm": 1.8812307119369507,
      "learning_rate": 9.018794549732819e-06,
      "loss": 0.4232,
      "num_input_tokens_seen": 6217560,
      "step": 9585
    },
    {
      "epoch": 5.654481132075472,
      "grad_norm": 2.2317006587982178,
      "learning_rate": 9.017263368328977e-06,
      "loss": 0.3911,
      "num_input_tokens_seen": 6220440,
      "step": 9590
    },
    {
      "epoch": 5.657429245283019,
      "grad_norm": 2.402195453643799,
      "learning_rate": 9.015731123337344e-06,
      "loss": 0.3826,
      "num_input_tokens_seen": 6223480,
      "step": 9595
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 1.6339701414108276,
      "learning_rate": 9.01419781516359e-06,
      "loss": 0.4212,
      "num_input_tokens_seen": 6226584,
      "step": 9600
    },
    {
      "epoch": 5.663325471698113,
      "grad_norm": 1.5165494680404663,
      "learning_rate": 9.012663444213664e-06,
      "loss": 0.4306,
      "num_input_tokens_seen": 6229624,
      "step": 9605
    },
    {
      "epoch": 5.66627358490566,
      "grad_norm": 1.4612891674041748,
      "learning_rate": 9.011128010893797e-06,
      "loss": 0.3966,
      "num_input_tokens_seen": 6233208,
      "step": 9610
    },
    {
      "epoch": 5.6692216981132075,
      "grad_norm": 1.9169846773147583,
      "learning_rate": 9.009591515610503e-06,
      "loss": 0.5612,
      "num_input_tokens_seen": 6235832,
      "step": 9615
    },
    {
      "epoch": 5.672169811320755,
      "grad_norm": 1.484295129776001,
      "learning_rate": 9.008053958770575e-06,
      "loss": 0.3846,
      "num_input_tokens_seen": 6240248,
      "step": 9620
    },
    {
      "epoch": 5.675117924528302,
      "grad_norm": 3.2434659004211426,
      "learning_rate": 9.006515340781087e-06,
      "loss": 0.5502,
      "num_input_tokens_seen": 6242488,
      "step": 9625
    },
    {
      "epoch": 5.678066037735849,
      "grad_norm": 3.244037628173828,
      "learning_rate": 9.004975662049396e-06,
      "loss": 0.4808,
      "num_input_tokens_seen": 6245144,
      "step": 9630
    },
    {
      "epoch": 5.681014150943396,
      "grad_norm": 1.607658863067627,
      "learning_rate": 9.003434922983138e-06,
      "loss": 0.4549,
      "num_input_tokens_seen": 6248568,
      "step": 9635
    },
    {
      "epoch": 5.683962264150943,
      "grad_norm": 1.2633109092712402,
      "learning_rate": 9.00189312399023e-06,
      "loss": 0.3643,
      "num_input_tokens_seen": 6252696,
      "step": 9640
    },
    {
      "epoch": 5.68691037735849,
      "grad_norm": 0.22020544111728668,
      "learning_rate": 9.00035026547887e-06,
      "loss": 0.2967,
      "num_input_tokens_seen": 6258968,
      "step": 9645
    },
    {
      "epoch": 5.689858490566038,
      "grad_norm": 1.9949334859848022,
      "learning_rate": 8.998806347857537e-06,
      "loss": 0.5792,
      "num_input_tokens_seen": 6262168,
      "step": 9650
    },
    {
      "epoch": 5.692806603773585,
      "grad_norm": 2.0265707969665527,
      "learning_rate": 8.99726137153499e-06,
      "loss": 0.4239,
      "num_input_tokens_seen": 6265048,
      "step": 9655
    },
    {
      "epoch": 5.695754716981132,
      "grad_norm": 1.2970401048660278,
      "learning_rate": 8.995715336920266e-06,
      "loss": 0.3157,
      "num_input_tokens_seen": 6268344,
      "step": 9660
    },
    {
      "epoch": 5.6987028301886795,
      "grad_norm": 1.4972827434539795,
      "learning_rate": 8.994168244422687e-06,
      "loss": 0.3977,
      "num_input_tokens_seen": 6271160,
      "step": 9665
    },
    {
      "epoch": 5.701650943396227,
      "grad_norm": 1.3918825387954712,
      "learning_rate": 8.992620094451852e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 6273848,
      "step": 9670
    },
    {
      "epoch": 5.704599056603773,
      "grad_norm": 2.091466188430786,
      "learning_rate": 8.991070887417639e-06,
      "loss": 0.3935,
      "num_input_tokens_seen": 6277208,
      "step": 9675
    },
    {
      "epoch": 5.7075471698113205,
      "grad_norm": 1.7068040370941162,
      "learning_rate": 8.989520623730208e-06,
      "loss": 0.3228,
      "num_input_tokens_seen": 6279992,
      "step": 9680
    },
    {
      "epoch": 5.710495283018868,
      "grad_norm": 1.8999199867248535,
      "learning_rate": 8.987969303799998e-06,
      "loss": 0.4142,
      "num_input_tokens_seen": 6282648,
      "step": 9685
    },
    {
      "epoch": 5.713443396226415,
      "grad_norm": 2.4997217655181885,
      "learning_rate": 8.986416928037728e-06,
      "loss": 0.4865,
      "num_input_tokens_seen": 6289624,
      "step": 9690
    },
    {
      "epoch": 5.716391509433962,
      "grad_norm": 2.494603395462036,
      "learning_rate": 8.984863496854395e-06,
      "loss": 0.5184,
      "num_input_tokens_seen": 6292728,
      "step": 9695
    },
    {
      "epoch": 5.71933962264151,
      "grad_norm": 1.3285424709320068,
      "learning_rate": 8.983309010661279e-06,
      "loss": 0.5166,
      "num_input_tokens_seen": 6296152,
      "step": 9700
    },
    {
      "epoch": 5.722287735849057,
      "grad_norm": 10.052237510681152,
      "learning_rate": 8.981753469869934e-06,
      "loss": 0.3714,
      "num_input_tokens_seen": 6299672,
      "step": 9705
    },
    {
      "epoch": 5.725235849056604,
      "grad_norm": 2.934919595718384,
      "learning_rate": 8.980196874892198e-06,
      "loss": 0.4046,
      "num_input_tokens_seen": 6302872,
      "step": 9710
    },
    {
      "epoch": 5.728183962264151,
      "grad_norm": 1.1267192363739014,
      "learning_rate": 8.978639226140184e-06,
      "loss": 0.4483,
      "num_input_tokens_seen": 6306072,
      "step": 9715
    },
    {
      "epoch": 5.731132075471698,
      "grad_norm": 2.421384334564209,
      "learning_rate": 8.977080524026289e-06,
      "loss": 0.4791,
      "num_input_tokens_seen": 6308888,
      "step": 9720
    },
    {
      "epoch": 5.734080188679245,
      "grad_norm": 1.370994210243225,
      "learning_rate": 8.975520768963186e-06,
      "loss": 0.4772,
      "num_input_tokens_seen": 6311608,
      "step": 9725
    },
    {
      "epoch": 5.7370283018867925,
      "grad_norm": 1.6979355812072754,
      "learning_rate": 8.973959961363825e-06,
      "loss": 0.5246,
      "num_input_tokens_seen": 6314776,
      "step": 9730
    },
    {
      "epoch": 5.73997641509434,
      "grad_norm": 1.556896448135376,
      "learning_rate": 8.972398101641438e-06,
      "loss": 0.3243,
      "num_input_tokens_seen": 6317368,
      "step": 9735
    },
    {
      "epoch": 5.742924528301887,
      "grad_norm": 1.7408573627471924,
      "learning_rate": 8.970835190209532e-06,
      "loss": 0.5601,
      "num_input_tokens_seen": 6320312,
      "step": 9740
    },
    {
      "epoch": 5.745872641509434,
      "grad_norm": 1.7415698766708374,
      "learning_rate": 8.969271227481899e-06,
      "loss": 0.3997,
      "num_input_tokens_seen": 6323352,
      "step": 9745
    },
    {
      "epoch": 5.748820754716981,
      "grad_norm": 1.8456833362579346,
      "learning_rate": 8.967706213872599e-06,
      "loss": 0.4268,
      "num_input_tokens_seen": 6326712,
      "step": 9750
    },
    {
      "epoch": 5.751768867924528,
      "grad_norm": 5.567098617553711,
      "learning_rate": 8.966140149795981e-06,
      "loss": 0.5958,
      "num_input_tokens_seen": 6330232,
      "step": 9755
    },
    {
      "epoch": 5.754716981132075,
      "grad_norm": 1.1320762634277344,
      "learning_rate": 8.964573035666663e-06,
      "loss": 0.3903,
      "num_input_tokens_seen": 6333400,
      "step": 9760
    },
    {
      "epoch": 5.757665094339623,
      "grad_norm": 1.9049339294433594,
      "learning_rate": 8.96300487189955e-06,
      "loss": 0.3961,
      "num_input_tokens_seen": 6336536,
      "step": 9765
    },
    {
      "epoch": 5.76061320754717,
      "grad_norm": 2.0106327533721924,
      "learning_rate": 8.961435658909816e-06,
      "loss": 0.4343,
      "num_input_tokens_seen": 6339608,
      "step": 9770
    },
    {
      "epoch": 5.763561320754717,
      "grad_norm": 2.4691953659057617,
      "learning_rate": 8.959865397112918e-06,
      "loss": 0.2953,
      "num_input_tokens_seen": 6343192,
      "step": 9775
    },
    {
      "epoch": 5.7665094339622645,
      "grad_norm": 2.6089956760406494,
      "learning_rate": 8.95829408692459e-06,
      "loss": 0.4312,
      "num_input_tokens_seen": 6345688,
      "step": 9780
    },
    {
      "epoch": 5.769457547169811,
      "grad_norm": 1.4212642908096313,
      "learning_rate": 8.956721728760845e-06,
      "loss": 0.3793,
      "num_input_tokens_seen": 6348536,
      "step": 9785
    },
    {
      "epoch": 5.772405660377358,
      "grad_norm": 1.5054311752319336,
      "learning_rate": 8.95514832303797e-06,
      "loss": 0.438,
      "num_input_tokens_seen": 6351160,
      "step": 9790
    },
    {
      "epoch": 5.775353773584905,
      "grad_norm": 1.3486381769180298,
      "learning_rate": 8.953573870172528e-06,
      "loss": 0.3359,
      "num_input_tokens_seen": 6353912,
      "step": 9795
    },
    {
      "epoch": 5.778301886792453,
      "grad_norm": 1.4680273532867432,
      "learning_rate": 8.951998370581368e-06,
      "loss": 0.4258,
      "num_input_tokens_seen": 6356952,
      "step": 9800
    },
    {
      "epoch": 5.78125,
      "grad_norm": 1.6259901523590088,
      "learning_rate": 8.950421824681605e-06,
      "loss": 0.3197,
      "num_input_tokens_seen": 6359768,
      "step": 9805
    },
    {
      "epoch": 5.784198113207547,
      "grad_norm": 1.122023582458496,
      "learning_rate": 8.948844232890638e-06,
      "loss": 0.4644,
      "num_input_tokens_seen": 6362808,
      "step": 9810
    },
    {
      "epoch": 5.787146226415095,
      "grad_norm": 3.4413599967956543,
      "learning_rate": 8.947265595626144e-06,
      "loss": 0.4096,
      "num_input_tokens_seen": 6365016,
      "step": 9815
    },
    {
      "epoch": 5.790094339622642,
      "grad_norm": 1.481494665145874,
      "learning_rate": 8.945685913306071e-06,
      "loss": 0.3899,
      "num_input_tokens_seen": 6368920,
      "step": 9820
    },
    {
      "epoch": 5.793042452830189,
      "grad_norm": 3.47196102142334,
      "learning_rate": 8.944105186348646e-06,
      "loss": 0.4821,
      "num_input_tokens_seen": 6371896,
      "step": 9825
    },
    {
      "epoch": 5.7959905660377355,
      "grad_norm": 2.5161726474761963,
      "learning_rate": 8.942523415172377e-06,
      "loss": 0.4529,
      "num_input_tokens_seen": 6375064,
      "step": 9830
    },
    {
      "epoch": 5.798938679245283,
      "grad_norm": 1.4407978057861328,
      "learning_rate": 8.94094060019604e-06,
      "loss": 0.3753,
      "num_input_tokens_seen": 6377656,
      "step": 9835
    },
    {
      "epoch": 5.80188679245283,
      "grad_norm": 1.9354358911514282,
      "learning_rate": 8.939356741838696e-06,
      "loss": 0.4731,
      "num_input_tokens_seen": 6380664,
      "step": 9840
    },
    {
      "epoch": 5.804834905660377,
      "grad_norm": 2.7214608192443848,
      "learning_rate": 8.937771840519677e-06,
      "loss": 0.4714,
      "num_input_tokens_seen": 6383544,
      "step": 9845
    },
    {
      "epoch": 5.807783018867925,
      "grad_norm": 2.3314332962036133,
      "learning_rate": 8.936185896658593e-06,
      "loss": 0.4399,
      "num_input_tokens_seen": 6386360,
      "step": 9850
    },
    {
      "epoch": 5.810731132075472,
      "grad_norm": 1.6079890727996826,
      "learning_rate": 8.934598910675329e-06,
      "loss": 0.4657,
      "num_input_tokens_seen": 6389368,
      "step": 9855
    },
    {
      "epoch": 5.813679245283019,
      "grad_norm": 1.7420051097869873,
      "learning_rate": 8.933010882990044e-06,
      "loss": 0.333,
      "num_input_tokens_seen": 6392984,
      "step": 9860
    },
    {
      "epoch": 5.816627358490566,
      "grad_norm": 1.5512326955795288,
      "learning_rate": 8.93142181402318e-06,
      "loss": 0.4007,
      "num_input_tokens_seen": 6395544,
      "step": 9865
    },
    {
      "epoch": 5.819575471698113,
      "grad_norm": 4.516857147216797,
      "learning_rate": 8.929831704195445e-06,
      "loss": 0.5258,
      "num_input_tokens_seen": 6398904,
      "step": 9870
    },
    {
      "epoch": 5.82252358490566,
      "grad_norm": 1.9389543533325195,
      "learning_rate": 8.928240553927831e-06,
      "loss": 0.3428,
      "num_input_tokens_seen": 6402200,
      "step": 9875
    },
    {
      "epoch": 5.8254716981132075,
      "grad_norm": 2.971418619155884,
      "learning_rate": 8.926648363641602e-06,
      "loss": 0.4366,
      "num_input_tokens_seen": 6404632,
      "step": 9880
    },
    {
      "epoch": 5.828419811320755,
      "grad_norm": 1.6788448095321655,
      "learning_rate": 8.925055133758294e-06,
      "loss": 0.5091,
      "num_input_tokens_seen": 6408664,
      "step": 9885
    },
    {
      "epoch": 5.831367924528302,
      "grad_norm": 2.351213216781616,
      "learning_rate": 8.923460864699723e-06,
      "loss": 0.407,
      "num_input_tokens_seen": 6411192,
      "step": 9890
    },
    {
      "epoch": 5.834316037735849,
      "grad_norm": 1.8426318168640137,
      "learning_rate": 8.921865556887979e-06,
      "loss": 0.5726,
      "num_input_tokens_seen": 6414136,
      "step": 9895
    },
    {
      "epoch": 5.837264150943396,
      "grad_norm": 2.2010984420776367,
      "learning_rate": 8.920269210745426e-06,
      "loss": 0.3625,
      "num_input_tokens_seen": 6417464,
      "step": 9900
    },
    {
      "epoch": 5.840212264150943,
      "grad_norm": 1.759907603263855,
      "learning_rate": 8.918671826694704e-06,
      "loss": 0.3641,
      "num_input_tokens_seen": 6420120,
      "step": 9905
    },
    {
      "epoch": 5.84316037735849,
      "grad_norm": 2.040309429168701,
      "learning_rate": 8.91707340515873e-06,
      "loss": 0.4124,
      "num_input_tokens_seen": 6422872,
      "step": 9910
    },
    {
      "epoch": 5.846108490566038,
      "grad_norm": 2.0204644203186035,
      "learning_rate": 8.915473946560688e-06,
      "loss": 0.4509,
      "num_input_tokens_seen": 6425880,
      "step": 9915
    },
    {
      "epoch": 5.849056603773585,
      "grad_norm": 3.8190884590148926,
      "learning_rate": 8.913873451324044e-06,
      "loss": 0.3789,
      "num_input_tokens_seen": 6429272,
      "step": 9920
    },
    {
      "epoch": 5.852004716981132,
      "grad_norm": 2.0160346031188965,
      "learning_rate": 8.912271919872538e-06,
      "loss": 0.3784,
      "num_input_tokens_seen": 6432664,
      "step": 9925
    },
    {
      "epoch": 5.8549528301886795,
      "grad_norm": 1.8670519590377808,
      "learning_rate": 8.910669352630176e-06,
      "loss": 0.4664,
      "num_input_tokens_seen": 6435704,
      "step": 9930
    },
    {
      "epoch": 5.857900943396227,
      "grad_norm": 2.226574659347534,
      "learning_rate": 8.909065750021253e-06,
      "loss": 0.4922,
      "num_input_tokens_seen": 6439320,
      "step": 9935
    },
    {
      "epoch": 5.860849056603773,
      "grad_norm": 2.538410186767578,
      "learning_rate": 8.907461112470323e-06,
      "loss": 0.5056,
      "num_input_tokens_seen": 6442872,
      "step": 9940
    },
    {
      "epoch": 5.8637971698113205,
      "grad_norm": 1.6750158071517944,
      "learning_rate": 8.905855440402225e-06,
      "loss": 0.5232,
      "num_input_tokens_seen": 6447608,
      "step": 9945
    },
    {
      "epoch": 5.866745283018868,
      "grad_norm": 1.2640960216522217,
      "learning_rate": 8.904248734242065e-06,
      "loss": 0.4903,
      "num_input_tokens_seen": 6450136,
      "step": 9950
    },
    {
      "epoch": 5.869693396226415,
      "grad_norm": 2.4400651454925537,
      "learning_rate": 8.902640994415226e-06,
      "loss": 0.5541,
      "num_input_tokens_seen": 6453144,
      "step": 9955
    },
    {
      "epoch": 5.872641509433962,
      "grad_norm": 2.150719165802002,
      "learning_rate": 8.901032221347364e-06,
      "loss": 0.5325,
      "num_input_tokens_seen": 6455896,
      "step": 9960
    },
    {
      "epoch": 5.87558962264151,
      "grad_norm": 1.9698145389556885,
      "learning_rate": 8.899422415464409e-06,
      "loss": 0.4464,
      "num_input_tokens_seen": 6460216,
      "step": 9965
    },
    {
      "epoch": 5.878537735849057,
      "grad_norm": 2.236675500869751,
      "learning_rate": 8.897811577192565e-06,
      "loss": 0.3815,
      "num_input_tokens_seen": 6463192,
      "step": 9970
    },
    {
      "epoch": 5.881485849056604,
      "grad_norm": 1.6076894998550415,
      "learning_rate": 8.896199706958306e-06,
      "loss": 0.4085,
      "num_input_tokens_seen": 6466712,
      "step": 9975
    },
    {
      "epoch": 5.884433962264151,
      "grad_norm": 1.5956227779388428,
      "learning_rate": 8.894586805188384e-06,
      "loss": 0.3563,
      "num_input_tokens_seen": 6470968,
      "step": 9980
    },
    {
      "epoch": 5.887382075471698,
      "grad_norm": 2.3943910598754883,
      "learning_rate": 8.892972872309821e-06,
      "loss": 0.4507,
      "num_input_tokens_seen": 6473912,
      "step": 9985
    },
    {
      "epoch": 5.890330188679245,
      "grad_norm": 2.3881325721740723,
      "learning_rate": 8.89135790874991e-06,
      "loss": 0.3315,
      "num_input_tokens_seen": 6477752,
      "step": 9990
    },
    {
      "epoch": 5.8932783018867925,
      "grad_norm": 1.5394755601882935,
      "learning_rate": 8.889741914936224e-06,
      "loss": 0.4117,
      "num_input_tokens_seen": 6480792,
      "step": 9995
    },
    {
      "epoch": 5.89622641509434,
      "grad_norm": 1.3843072652816772,
      "learning_rate": 8.888124891296602e-06,
      "loss": 0.4567,
      "num_input_tokens_seen": 6485464,
      "step": 10000
    },
    {
      "epoch": 5.899174528301887,
      "grad_norm": 2.165414810180664,
      "learning_rate": 8.886506838259156e-06,
      "loss": 0.5199,
      "num_input_tokens_seen": 6488856,
      "step": 10005
    },
    {
      "epoch": 5.902122641509434,
      "grad_norm": 1.3554251194000244,
      "learning_rate": 8.884887756252279e-06,
      "loss": 0.4807,
      "num_input_tokens_seen": 6492248,
      "step": 10010
    },
    {
      "epoch": 5.905070754716981,
      "grad_norm": 1.0664602518081665,
      "learning_rate": 8.88326764570462e-06,
      "loss": 0.5169,
      "num_input_tokens_seen": 6494936,
      "step": 10015
    },
    {
      "epoch": 5.908018867924528,
      "grad_norm": 2.1176981925964355,
      "learning_rate": 8.88164650704512e-06,
      "loss": 0.462,
      "num_input_tokens_seen": 6498040,
      "step": 10020
    },
    {
      "epoch": 5.910966981132075,
      "grad_norm": 2.909895420074463,
      "learning_rate": 8.880024340702978e-06,
      "loss": 0.3933,
      "num_input_tokens_seen": 6500728,
      "step": 10025
    },
    {
      "epoch": 5.913915094339623,
      "grad_norm": 1.4781135320663452,
      "learning_rate": 8.878401147107667e-06,
      "loss": 0.3799,
      "num_input_tokens_seen": 6505016,
      "step": 10030
    },
    {
      "epoch": 5.91686320754717,
      "grad_norm": 2.190244197845459,
      "learning_rate": 8.87677692668894e-06,
      "loss": 0.5751,
      "num_input_tokens_seen": 6507512,
      "step": 10035
    },
    {
      "epoch": 5.919811320754717,
      "grad_norm": 1.982753038406372,
      "learning_rate": 8.875151679876813e-06,
      "loss": 0.4715,
      "num_input_tokens_seen": 6511192,
      "step": 10040
    },
    {
      "epoch": 5.9227594339622645,
      "grad_norm": 1.8954122066497803,
      "learning_rate": 8.873525407101577e-06,
      "loss": 0.4628,
      "num_input_tokens_seen": 6514520,
      "step": 10045
    },
    {
      "epoch": 5.925707547169811,
      "grad_norm": 5.981013298034668,
      "learning_rate": 8.871898108793796e-06,
      "loss": 0.4068,
      "num_input_tokens_seen": 6516984,
      "step": 10050
    },
    {
      "epoch": 5.928655660377358,
      "grad_norm": 3.197829246520996,
      "learning_rate": 8.870269785384304e-06,
      "loss": 0.4232,
      "num_input_tokens_seen": 6520664,
      "step": 10055
    },
    {
      "epoch": 5.931603773584905,
      "grad_norm": 1.5067697763442993,
      "learning_rate": 8.868640437304206e-06,
      "loss": 0.5699,
      "num_input_tokens_seen": 6524088,
      "step": 10060
    },
    {
      "epoch": 5.934551886792453,
      "grad_norm": 1.789751410484314,
      "learning_rate": 8.86701006498488e-06,
      "loss": 0.3852,
      "num_input_tokens_seen": 6527640,
      "step": 10065
    },
    {
      "epoch": 5.9375,
      "grad_norm": 1.9011822938919067,
      "learning_rate": 8.865378668857972e-06,
      "loss": 0.5657,
      "num_input_tokens_seen": 6531352,
      "step": 10070
    },
    {
      "epoch": 5.940448113207547,
      "grad_norm": 2.9439730644226074,
      "learning_rate": 8.863746249355404e-06,
      "loss": 0.487,
      "num_input_tokens_seen": 6534904,
      "step": 10075
    },
    {
      "epoch": 5.943396226415095,
      "grad_norm": 1.9183653593063354,
      "learning_rate": 8.862112806909365e-06,
      "loss": 0.3789,
      "num_input_tokens_seen": 6538168,
      "step": 10080
    },
    {
      "epoch": 5.946344339622642,
      "grad_norm": 1.985841989517212,
      "learning_rate": 8.860478341952314e-06,
      "loss": 0.4045,
      "num_input_tokens_seen": 6541464,
      "step": 10085
    },
    {
      "epoch": 5.949292452830189,
      "grad_norm": 2.9855377674102783,
      "learning_rate": 8.858842854916985e-06,
      "loss": 0.6543,
      "num_input_tokens_seen": 6543800,
      "step": 10090
    },
    {
      "epoch": 5.9522405660377355,
      "grad_norm": 1.3530082702636719,
      "learning_rate": 8.85720634623638e-06,
      "loss": 0.4514,
      "num_input_tokens_seen": 6546680,
      "step": 10095
    },
    {
      "epoch": 5.955188679245283,
      "grad_norm": 7.453511714935303,
      "learning_rate": 8.855568816343769e-06,
      "loss": 0.4684,
      "num_input_tokens_seen": 6549624,
      "step": 10100
    },
    {
      "epoch": 5.95813679245283,
      "grad_norm": 2.0546066761016846,
      "learning_rate": 8.8539302656727e-06,
      "loss": 0.4148,
      "num_input_tokens_seen": 6552568,
      "step": 10105
    },
    {
      "epoch": 5.961084905660377,
      "grad_norm": 2.2903871536254883,
      "learning_rate": 8.852290694656983e-06,
      "loss": 0.4048,
      "num_input_tokens_seen": 6556024,
      "step": 10110
    },
    {
      "epoch": 5.964033018867925,
      "grad_norm": 2.2785627841949463,
      "learning_rate": 8.8506501037307e-06,
      "loss": 0.4407,
      "num_input_tokens_seen": 6558840,
      "step": 10115
    },
    {
      "epoch": 5.966981132075472,
      "grad_norm": 2.4536187648773193,
      "learning_rate": 8.849008493328209e-06,
      "loss": 0.4829,
      "num_input_tokens_seen": 6561944,
      "step": 10120
    },
    {
      "epoch": 5.969929245283019,
      "grad_norm": 2.5769920349121094,
      "learning_rate": 8.847365863884131e-06,
      "loss": 0.4471,
      "num_input_tokens_seen": 6565528,
      "step": 10125
    },
    {
      "epoch": 5.972877358490566,
      "grad_norm": 2.2029621601104736,
      "learning_rate": 8.845722215833359e-06,
      "loss": 0.4365,
      "num_input_tokens_seen": 6568696,
      "step": 10130
    },
    {
      "epoch": 5.975825471698113,
      "grad_norm": 1.9262051582336426,
      "learning_rate": 8.844077549611056e-06,
      "loss": 0.3339,
      "num_input_tokens_seen": 6571576,
      "step": 10135
    },
    {
      "epoch": 5.97877358490566,
      "grad_norm": 2.1555092334747314,
      "learning_rate": 8.842431865652654e-06,
      "loss": 0.3827,
      "num_input_tokens_seen": 6574872,
      "step": 10140
    },
    {
      "epoch": 5.9817216981132075,
      "grad_norm": 1.0991811752319336,
      "learning_rate": 8.840785164393858e-06,
      "loss": 0.3358,
      "num_input_tokens_seen": 6577880,
      "step": 10145
    },
    {
      "epoch": 5.984669811320755,
      "grad_norm": 1.7313475608825684,
      "learning_rate": 8.839137446270634e-06,
      "loss": 0.4305,
      "num_input_tokens_seen": 6580312,
      "step": 10150
    },
    {
      "epoch": 5.987617924528302,
      "grad_norm": 1.6000044345855713,
      "learning_rate": 8.837488711719226e-06,
      "loss": 0.3998,
      "num_input_tokens_seen": 6583096,
      "step": 10155
    },
    {
      "epoch": 5.990566037735849,
      "grad_norm": 2.221111536026001,
      "learning_rate": 8.835838961176143e-06,
      "loss": 0.5087,
      "num_input_tokens_seen": 6587480,
      "step": 10160
    },
    {
      "epoch": 5.993514150943396,
      "grad_norm": 2.8927290439605713,
      "learning_rate": 8.834188195078164e-06,
      "loss": 0.4242,
      "num_input_tokens_seen": 6590072,
      "step": 10165
    },
    {
      "epoch": 5.996462264150943,
      "grad_norm": 2.037487506866455,
      "learning_rate": 8.832536413862337e-06,
      "loss": 0.512,
      "num_input_tokens_seen": 6593432,
      "step": 10170
    },
    {
      "epoch": 5.99941037735849,
      "grad_norm": 1.8480627536773682,
      "learning_rate": 8.830883617965976e-06,
      "loss": 0.4362,
      "num_input_tokens_seen": 6595896,
      "step": 10175
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.5032461881637573,
      "eval_runtime": 19.4778,
      "eval_samples_per_second": 87.073,
      "eval_steps_per_second": 21.768,
      "num_input_tokens_seen": 6595976,
      "step": 10176
    },
    {
      "epoch": 6.002358490566038,
      "grad_norm": 1.8884270191192627,
      "learning_rate": 8.829229807826665e-06,
      "loss": 0.4333,
      "num_input_tokens_seen": 6599304,
      "step": 10180
    },
    {
      "epoch": 6.005306603773585,
      "grad_norm": 1.790036916732788,
      "learning_rate": 8.82757498388226e-06,
      "loss": 0.4795,
      "num_input_tokens_seen": 6602088,
      "step": 10185
    },
    {
      "epoch": 6.008254716981132,
      "grad_norm": 2.586909532546997,
      "learning_rate": 8.825919146570884e-06,
      "loss": 0.4409,
      "num_input_tokens_seen": 6605192,
      "step": 10190
    },
    {
      "epoch": 6.0112028301886795,
      "grad_norm": 2.689540147781372,
      "learning_rate": 8.824262296330925e-06,
      "loss": 0.5094,
      "num_input_tokens_seen": 6607752,
      "step": 10195
    },
    {
      "epoch": 6.014150943396227,
      "grad_norm": 3.800062656402588,
      "learning_rate": 8.822604433601041e-06,
      "loss": 0.4251,
      "num_input_tokens_seen": 6610920,
      "step": 10200
    },
    {
      "epoch": 6.017099056603773,
      "grad_norm": 4.283565044403076,
      "learning_rate": 8.820945558820158e-06,
      "loss": 0.3849,
      "num_input_tokens_seen": 6613096,
      "step": 10205
    },
    {
      "epoch": 6.0200471698113205,
      "grad_norm": 2.2254891395568848,
      "learning_rate": 8.81928567242747e-06,
      "loss": 0.5283,
      "num_input_tokens_seen": 6616104,
      "step": 10210
    },
    {
      "epoch": 6.022995283018868,
      "grad_norm": 1.999754548072815,
      "learning_rate": 8.817624774862443e-06,
      "loss": 0.5469,
      "num_input_tokens_seen": 6619688,
      "step": 10215
    },
    {
      "epoch": 6.025943396226415,
      "grad_norm": 1.807440996170044,
      "learning_rate": 8.815962866564803e-06,
      "loss": 0.4119,
      "num_input_tokens_seen": 6623240,
      "step": 10220
    },
    {
      "epoch": 6.028891509433962,
      "grad_norm": 3.4922518730163574,
      "learning_rate": 8.814299947974547e-06,
      "loss": 0.4695,
      "num_input_tokens_seen": 6627144,
      "step": 10225
    },
    {
      "epoch": 6.03183962264151,
      "grad_norm": 1.324137806892395,
      "learning_rate": 8.812636019531942e-06,
      "loss": 0.3754,
      "num_input_tokens_seen": 6630312,
      "step": 10230
    },
    {
      "epoch": 6.034787735849057,
      "grad_norm": 2.662675619125366,
      "learning_rate": 8.810971081677517e-06,
      "loss": 0.472,
      "num_input_tokens_seen": 6633608,
      "step": 10235
    },
    {
      "epoch": 6.037735849056604,
      "grad_norm": 4.891324996948242,
      "learning_rate": 8.809305134852076e-06,
      "loss": 0.4936,
      "num_input_tokens_seen": 6636552,
      "step": 10240
    },
    {
      "epoch": 6.040683962264151,
      "grad_norm": 2.2346508502960205,
      "learning_rate": 8.807638179496684e-06,
      "loss": 0.4162,
      "num_input_tokens_seen": 6640296,
      "step": 10245
    },
    {
      "epoch": 6.043632075471698,
      "grad_norm": 1.3740437030792236,
      "learning_rate": 8.805970216052673e-06,
      "loss": 0.5397,
      "num_input_tokens_seen": 6642792,
      "step": 10250
    },
    {
      "epoch": 6.046580188679245,
      "grad_norm": 2.3150084018707275,
      "learning_rate": 8.804301244961645e-06,
      "loss": 0.4385,
      "num_input_tokens_seen": 6645256,
      "step": 10255
    },
    {
      "epoch": 6.0495283018867925,
      "grad_norm": 1.7232710123062134,
      "learning_rate": 8.802631266665465e-06,
      "loss": 0.2921,
      "num_input_tokens_seen": 6648392,
      "step": 10260
    },
    {
      "epoch": 6.05247641509434,
      "grad_norm": 1.324416160583496,
      "learning_rate": 8.80096028160627e-06,
      "loss": 0.3906,
      "num_input_tokens_seen": 6652648,
      "step": 10265
    },
    {
      "epoch": 6.055424528301887,
      "grad_norm": 1.9490495920181274,
      "learning_rate": 8.799288290226457e-06,
      "loss": 0.4094,
      "num_input_tokens_seen": 6655304,
      "step": 10270
    },
    {
      "epoch": 6.058372641509434,
      "grad_norm": 1.2971248626708984,
      "learning_rate": 8.797615292968698e-06,
      "loss": 0.4868,
      "num_input_tokens_seen": 6658824,
      "step": 10275
    },
    {
      "epoch": 6.061320754716981,
      "grad_norm": 1.7486923933029175,
      "learning_rate": 8.795941290275923e-06,
      "loss": 0.3606,
      "num_input_tokens_seen": 6662760,
      "step": 10280
    },
    {
      "epoch": 6.064268867924528,
      "grad_norm": 2.7549540996551514,
      "learning_rate": 8.79426628259133e-06,
      "loss": 0.3798,
      "num_input_tokens_seen": 6665256,
      "step": 10285
    },
    {
      "epoch": 6.067216981132075,
      "grad_norm": 4.4255547523498535,
      "learning_rate": 8.792590270358389e-06,
      "loss": 0.3093,
      "num_input_tokens_seen": 6668168,
      "step": 10290
    },
    {
      "epoch": 6.070165094339623,
      "grad_norm": 2.4917612075805664,
      "learning_rate": 8.790913254020827e-06,
      "loss": 0.3387,
      "num_input_tokens_seen": 6671144,
      "step": 10295
    },
    {
      "epoch": 6.07311320754717,
      "grad_norm": 2.9512252807617188,
      "learning_rate": 8.789235234022643e-06,
      "loss": 0.5039,
      "num_input_tokens_seen": 6674472,
      "step": 10300
    },
    {
      "epoch": 6.076061320754717,
      "grad_norm": 2.485963821411133,
      "learning_rate": 8.787556210808101e-06,
      "loss": 0.3802,
      "num_input_tokens_seen": 6676936,
      "step": 10305
    },
    {
      "epoch": 6.0790094339622645,
      "grad_norm": 2.0318233966827393,
      "learning_rate": 8.78587618482173e-06,
      "loss": 0.331,
      "num_input_tokens_seen": 6680936,
      "step": 10310
    },
    {
      "epoch": 6.081957547169812,
      "grad_norm": 3.1216938495635986,
      "learning_rate": 8.78419515650832e-06,
      "loss": 0.3687,
      "num_input_tokens_seen": 6684904,
      "step": 10315
    },
    {
      "epoch": 6.084905660377358,
      "grad_norm": 1.3701589107513428,
      "learning_rate": 8.782513126312934e-06,
      "loss": 0.3154,
      "num_input_tokens_seen": 6689512,
      "step": 10320
    },
    {
      "epoch": 6.087853773584905,
      "grad_norm": 8.70892333984375,
      "learning_rate": 8.780830094680897e-06,
      "loss": 0.3385,
      "num_input_tokens_seen": 6693096,
      "step": 10325
    },
    {
      "epoch": 6.090801886792453,
      "grad_norm": 1.8383123874664307,
      "learning_rate": 8.779146062057797e-06,
      "loss": 0.4776,
      "num_input_tokens_seen": 6696424,
      "step": 10330
    },
    {
      "epoch": 6.09375,
      "grad_norm": 3.061941385269165,
      "learning_rate": 8.777461028889492e-06,
      "loss": 0.4865,
      "num_input_tokens_seen": 6699304,
      "step": 10335
    },
    {
      "epoch": 6.096698113207547,
      "grad_norm": 1.8246631622314453,
      "learning_rate": 8.775774995622097e-06,
      "loss": 0.485,
      "num_input_tokens_seen": 6703528,
      "step": 10340
    },
    {
      "epoch": 6.099646226415095,
      "grad_norm": 2.930114507675171,
      "learning_rate": 8.774087962702e-06,
      "loss": 0.4399,
      "num_input_tokens_seen": 6706600,
      "step": 10345
    },
    {
      "epoch": 6.102594339622642,
      "grad_norm": 1.332486867904663,
      "learning_rate": 8.772399930575849e-06,
      "loss": 0.4655,
      "num_input_tokens_seen": 6709896,
      "step": 10350
    },
    {
      "epoch": 6.105542452830188,
      "grad_norm": 1.5618077516555786,
      "learning_rate": 8.77071089969056e-06,
      "loss": 0.3742,
      "num_input_tokens_seen": 6713000,
      "step": 10355
    },
    {
      "epoch": 6.1084905660377355,
      "grad_norm": 2.7329437732696533,
      "learning_rate": 8.769020870493309e-06,
      "loss": 0.4949,
      "num_input_tokens_seen": 6716360,
      "step": 10360
    },
    {
      "epoch": 6.111438679245283,
      "grad_norm": 2.764397382736206,
      "learning_rate": 8.767329843431537e-06,
      "loss": 0.4722,
      "num_input_tokens_seen": 6719240,
      "step": 10365
    },
    {
      "epoch": 6.11438679245283,
      "grad_norm": 2.2062032222747803,
      "learning_rate": 8.765637818952954e-06,
      "loss": 0.3072,
      "num_input_tokens_seen": 6722760,
      "step": 10370
    },
    {
      "epoch": 6.117334905660377,
      "grad_norm": 2.098085880279541,
      "learning_rate": 8.76394479750553e-06,
      "loss": 0.4115,
      "num_input_tokens_seen": 6725512,
      "step": 10375
    },
    {
      "epoch": 6.120283018867925,
      "grad_norm": 2.484745979309082,
      "learning_rate": 8.762250779537499e-06,
      "loss": 0.4068,
      "num_input_tokens_seen": 6728328,
      "step": 10380
    },
    {
      "epoch": 6.123231132075472,
      "grad_norm": 2.7383410930633545,
      "learning_rate": 8.760555765497358e-06,
      "loss": 0.4672,
      "num_input_tokens_seen": 6731016,
      "step": 10385
    },
    {
      "epoch": 6.126179245283019,
      "grad_norm": 1.8066169023513794,
      "learning_rate": 8.758859755833873e-06,
      "loss": 0.5263,
      "num_input_tokens_seen": 6734472,
      "step": 10390
    },
    {
      "epoch": 6.129127358490566,
      "grad_norm": 2.633483648300171,
      "learning_rate": 8.757162750996066e-06,
      "loss": 0.4463,
      "num_input_tokens_seen": 6737576,
      "step": 10395
    },
    {
      "epoch": 6.132075471698113,
      "grad_norm": 1.1439383029937744,
      "learning_rate": 8.755464751433229e-06,
      "loss": 0.5408,
      "num_input_tokens_seen": 6742728,
      "step": 10400
    },
    {
      "epoch": 6.13502358490566,
      "grad_norm": 1.785364031791687,
      "learning_rate": 8.753765757594915e-06,
      "loss": 0.3282,
      "num_input_tokens_seen": 6745608,
      "step": 10405
    },
    {
      "epoch": 6.1379716981132075,
      "grad_norm": 2.2832536697387695,
      "learning_rate": 8.752065769930938e-06,
      "loss": 0.5617,
      "num_input_tokens_seen": 6748648,
      "step": 10410
    },
    {
      "epoch": 6.140919811320755,
      "grad_norm": 2.200327157974243,
      "learning_rate": 8.75036478889138e-06,
      "loss": 0.4338,
      "num_input_tokens_seen": 6751624,
      "step": 10415
    },
    {
      "epoch": 6.143867924528302,
      "grad_norm": 1.4451042413711548,
      "learning_rate": 8.748662814926576e-06,
      "loss": 0.4031,
      "num_input_tokens_seen": 6756040,
      "step": 10420
    },
    {
      "epoch": 6.146816037735849,
      "grad_norm": 2.2571699619293213,
      "learning_rate": 8.746959848487139e-06,
      "loss": 0.4989,
      "num_input_tokens_seen": 6759368,
      "step": 10425
    },
    {
      "epoch": 6.149764150943396,
      "grad_norm": 2.058870792388916,
      "learning_rate": 8.745255890023934e-06,
      "loss": 0.6646,
      "num_input_tokens_seen": 6763016,
      "step": 10430
    },
    {
      "epoch": 6.152712264150943,
      "grad_norm": 1.8947391510009766,
      "learning_rate": 8.74355093998809e-06,
      "loss": 0.4823,
      "num_input_tokens_seen": 6765896,
      "step": 10435
    },
    {
      "epoch": 6.15566037735849,
      "grad_norm": 2.1396284103393555,
      "learning_rate": 8.741844998831001e-06,
      "loss": 0.408,
      "num_input_tokens_seen": 6769576,
      "step": 10440
    },
    {
      "epoch": 6.158608490566038,
      "grad_norm": 1.2809884548187256,
      "learning_rate": 8.740138067004323e-06,
      "loss": 0.2713,
      "num_input_tokens_seen": 6775944,
      "step": 10445
    },
    {
      "epoch": 6.161556603773585,
      "grad_norm": 4.602686882019043,
      "learning_rate": 8.738430144959973e-06,
      "loss": 0.474,
      "num_input_tokens_seen": 6778440,
      "step": 10450
    },
    {
      "epoch": 6.164504716981132,
      "grad_norm": 2.0612168312072754,
      "learning_rate": 8.73672123315013e-06,
      "loss": 0.3267,
      "num_input_tokens_seen": 6781512,
      "step": 10455
    },
    {
      "epoch": 6.1674528301886795,
      "grad_norm": 2.12876033782959,
      "learning_rate": 8.735011332027234e-06,
      "loss": 0.4408,
      "num_input_tokens_seen": 6785032,
      "step": 10460
    },
    {
      "epoch": 6.170400943396227,
      "grad_norm": 1.4347407817840576,
      "learning_rate": 8.733300442043993e-06,
      "loss": 0.346,
      "num_input_tokens_seen": 6787912,
      "step": 10465
    },
    {
      "epoch": 6.173349056603773,
      "grad_norm": 1.7403318881988525,
      "learning_rate": 8.73158856365337e-06,
      "loss": 0.4325,
      "num_input_tokens_seen": 6790344,
      "step": 10470
    },
    {
      "epoch": 6.1762971698113205,
      "grad_norm": 1.4935781955718994,
      "learning_rate": 8.729875697308592e-06,
      "loss": 0.4866,
      "num_input_tokens_seen": 6796584,
      "step": 10475
    },
    {
      "epoch": 6.179245283018868,
      "grad_norm": 1.4369373321533203,
      "learning_rate": 8.728161843463148e-06,
      "loss": 0.4595,
      "num_input_tokens_seen": 6799752,
      "step": 10480
    },
    {
      "epoch": 6.182193396226415,
      "grad_norm": 2.3729093074798584,
      "learning_rate": 8.726447002570791e-06,
      "loss": 0.5149,
      "num_input_tokens_seen": 6802728,
      "step": 10485
    },
    {
      "epoch": 6.185141509433962,
      "grad_norm": 3.0586867332458496,
      "learning_rate": 8.724731175085526e-06,
      "loss": 0.3798,
      "num_input_tokens_seen": 6805448,
      "step": 10490
    },
    {
      "epoch": 6.18808962264151,
      "grad_norm": 2.5647048950195312,
      "learning_rate": 8.723014361461633e-06,
      "loss": 0.3547,
      "num_input_tokens_seen": 6811240,
      "step": 10495
    },
    {
      "epoch": 6.191037735849057,
      "grad_norm": 2.412602186203003,
      "learning_rate": 8.72129656215364e-06,
      "loss": 0.4874,
      "num_input_tokens_seen": 6814888,
      "step": 10500
    },
    {
      "epoch": 6.193985849056604,
      "grad_norm": 2.3946287631988525,
      "learning_rate": 8.719577777616347e-06,
      "loss": 0.4456,
      "num_input_tokens_seen": 6818920,
      "step": 10505
    },
    {
      "epoch": 6.196933962264151,
      "grad_norm": 3.360201358795166,
      "learning_rate": 8.717858008304804e-06,
      "loss": 0.4067,
      "num_input_tokens_seen": 6821992,
      "step": 10510
    },
    {
      "epoch": 6.199882075471698,
      "grad_norm": 2.600752353668213,
      "learning_rate": 8.71613725467433e-06,
      "loss": 0.4351,
      "num_input_tokens_seen": 6824776,
      "step": 10515
    },
    {
      "epoch": 6.202830188679245,
      "grad_norm": 3.2504830360412598,
      "learning_rate": 8.714415517180506e-06,
      "loss": 0.5059,
      "num_input_tokens_seen": 6827880,
      "step": 10520
    },
    {
      "epoch": 6.2057783018867925,
      "grad_norm": 2.574557065963745,
      "learning_rate": 8.712692796279164e-06,
      "loss": 0.369,
      "num_input_tokens_seen": 6831048,
      "step": 10525
    },
    {
      "epoch": 6.20872641509434,
      "grad_norm": 2.1246700286865234,
      "learning_rate": 8.710969092426401e-06,
      "loss": 0.356,
      "num_input_tokens_seen": 6833832,
      "step": 10530
    },
    {
      "epoch": 6.211674528301887,
      "grad_norm": 1.973631739616394,
      "learning_rate": 8.70924440607858e-06,
      "loss": 0.4139,
      "num_input_tokens_seen": 6837320,
      "step": 10535
    },
    {
      "epoch": 6.214622641509434,
      "grad_norm": 2.808867931365967,
      "learning_rate": 8.707518737692315e-06,
      "loss": 0.416,
      "num_input_tokens_seen": 6839720,
      "step": 10540
    },
    {
      "epoch": 6.217570754716981,
      "grad_norm": 1.37572181224823,
      "learning_rate": 8.705792087724485e-06,
      "loss": 0.479,
      "num_input_tokens_seen": 6842472,
      "step": 10545
    },
    {
      "epoch": 6.220518867924528,
      "grad_norm": 1.8922019004821777,
      "learning_rate": 8.704064456632231e-06,
      "loss": 0.3282,
      "num_input_tokens_seen": 6846504,
      "step": 10550
    },
    {
      "epoch": 6.223466981132075,
      "grad_norm": 3.93414044380188,
      "learning_rate": 8.702335844872946e-06,
      "loss": 0.393,
      "num_input_tokens_seen": 6850152,
      "step": 10555
    },
    {
      "epoch": 6.226415094339623,
      "grad_norm": 1.872454285621643,
      "learning_rate": 8.700606252904293e-06,
      "loss": 0.3672,
      "num_input_tokens_seen": 6853288,
      "step": 10560
    },
    {
      "epoch": 6.22936320754717,
      "grad_norm": 2.0965211391448975,
      "learning_rate": 8.698875681184183e-06,
      "loss": 0.4562,
      "num_input_tokens_seen": 6856264,
      "step": 10565
    },
    {
      "epoch": 6.232311320754717,
      "grad_norm": 2.128549575805664,
      "learning_rate": 8.697144130170797e-06,
      "loss": 0.4213,
      "num_input_tokens_seen": 6859784,
      "step": 10570
    },
    {
      "epoch": 6.2352594339622645,
      "grad_norm": 4.146857261657715,
      "learning_rate": 8.695411600322568e-06,
      "loss": 0.6326,
      "num_input_tokens_seen": 6862920,
      "step": 10575
    },
    {
      "epoch": 6.238207547169812,
      "grad_norm": 4.283497333526611,
      "learning_rate": 8.693678092098191e-06,
      "loss": 0.4922,
      "num_input_tokens_seen": 6865800,
      "step": 10580
    },
    {
      "epoch": 6.241155660377358,
      "grad_norm": 2.1207242012023926,
      "learning_rate": 8.691943605956621e-06,
      "loss": 0.3562,
      "num_input_tokens_seen": 6868488,
      "step": 10585
    },
    {
      "epoch": 6.244103773584905,
      "grad_norm": 1.651052474975586,
      "learning_rate": 8.690208142357069e-06,
      "loss": 0.4646,
      "num_input_tokens_seen": 6871240,
      "step": 10590
    },
    {
      "epoch": 6.247051886792453,
      "grad_norm": 2.0461831092834473,
      "learning_rate": 8.68847170175901e-06,
      "loss": 0.4046,
      "num_input_tokens_seen": 6873672,
      "step": 10595
    },
    {
      "epoch": 6.25,
      "grad_norm": 1.762420892715454,
      "learning_rate": 8.686734284622168e-06,
      "loss": 0.3399,
      "num_input_tokens_seen": 6876616,
      "step": 10600
    },
    {
      "epoch": 6.252948113207547,
      "grad_norm": 2.2347593307495117,
      "learning_rate": 8.684995891406537e-06,
      "loss": 0.3754,
      "num_input_tokens_seen": 6879912,
      "step": 10605
    },
    {
      "epoch": 6.255896226415095,
      "grad_norm": 3.583677053451538,
      "learning_rate": 8.683256522572362e-06,
      "loss": 0.4204,
      "num_input_tokens_seen": 6882696,
      "step": 10610
    },
    {
      "epoch": 6.258844339622642,
      "grad_norm": 2.1330151557922363,
      "learning_rate": 8.68151617858015e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 6886472,
      "step": 10615
    },
    {
      "epoch": 6.261792452830189,
      "grad_norm": 1.530730128288269,
      "learning_rate": 8.67977485989066e-06,
      "loss": 0.6139,
      "num_input_tokens_seen": 6889096,
      "step": 10620
    },
    {
      "epoch": 6.2647405660377355,
      "grad_norm": 1.8143107891082764,
      "learning_rate": 8.67803256696492e-06,
      "loss": 0.5099,
      "num_input_tokens_seen": 6894024,
      "step": 10625
    },
    {
      "epoch": 6.267688679245283,
      "grad_norm": 1.3023273944854736,
      "learning_rate": 8.676289300264205e-06,
      "loss": 0.4852,
      "num_input_tokens_seen": 6897448,
      "step": 10630
    },
    {
      "epoch": 6.27063679245283,
      "grad_norm": 1.2358739376068115,
      "learning_rate": 8.674545060250054e-06,
      "loss": 0.486,
      "num_input_tokens_seen": 6900456,
      "step": 10635
    },
    {
      "epoch": 6.273584905660377,
      "grad_norm": 2.1106762886047363,
      "learning_rate": 8.672799847384263e-06,
      "loss": 0.485,
      "num_input_tokens_seen": 6904488,
      "step": 10640
    },
    {
      "epoch": 6.276533018867925,
      "grad_norm": 1.2958908081054688,
      "learning_rate": 8.671053662128883e-06,
      "loss": 0.4871,
      "num_input_tokens_seen": 6907464,
      "step": 10645
    },
    {
      "epoch": 6.279481132075472,
      "grad_norm": 1.6628706455230713,
      "learning_rate": 8.669306504946223e-06,
      "loss": 0.4673,
      "num_input_tokens_seen": 6910632,
      "step": 10650
    },
    {
      "epoch": 6.282429245283019,
      "grad_norm": 2.9708359241485596,
      "learning_rate": 8.667558376298854e-06,
      "loss": 0.4947,
      "num_input_tokens_seen": 6913960,
      "step": 10655
    },
    {
      "epoch": 6.285377358490566,
      "grad_norm": 2.472585678100586,
      "learning_rate": 8.665809276649597e-06,
      "loss": 0.4608,
      "num_input_tokens_seen": 6917448,
      "step": 10660
    },
    {
      "epoch": 6.288325471698113,
      "grad_norm": 1.4607027769088745,
      "learning_rate": 8.664059206461537e-06,
      "loss": 0.3771,
      "num_input_tokens_seen": 6921544,
      "step": 10665
    },
    {
      "epoch": 6.29127358490566,
      "grad_norm": 2.2980363368988037,
      "learning_rate": 8.662308166198009e-06,
      "loss": 0.4496,
      "num_input_tokens_seen": 6923976,
      "step": 10670
    },
    {
      "epoch": 6.2942216981132075,
      "grad_norm": 1.1476317644119263,
      "learning_rate": 8.660556156322611e-06,
      "loss": 0.3666,
      "num_input_tokens_seen": 6927112,
      "step": 10675
    },
    {
      "epoch": 6.297169811320755,
      "grad_norm": 1.4175493717193604,
      "learning_rate": 8.658803177299196e-06,
      "loss": 0.4179,
      "num_input_tokens_seen": 6930216,
      "step": 10680
    },
    {
      "epoch": 6.300117924528302,
      "grad_norm": 1.5638962984085083,
      "learning_rate": 8.65704922959187e-06,
      "loss": 0.3862,
      "num_input_tokens_seen": 6933768,
      "step": 10685
    },
    {
      "epoch": 6.303066037735849,
      "grad_norm": 2.984652042388916,
      "learning_rate": 8.655294313664998e-06,
      "loss": 0.3897,
      "num_input_tokens_seen": 6936200,
      "step": 10690
    },
    {
      "epoch": 6.306014150943396,
      "grad_norm": 1.5672595500946045,
      "learning_rate": 8.653538429983204e-06,
      "loss": 0.2759,
      "num_input_tokens_seen": 6940040,
      "step": 10695
    },
    {
      "epoch": 6.308962264150943,
      "grad_norm": 2.0619451999664307,
      "learning_rate": 8.651781579011366e-06,
      "loss": 0.3691,
      "num_input_tokens_seen": 6944168,
      "step": 10700
    },
    {
      "epoch": 6.31191037735849,
      "grad_norm": 1.9197971820831299,
      "learning_rate": 8.650023761214615e-06,
      "loss": 0.3979,
      "num_input_tokens_seen": 6946728,
      "step": 10705
    },
    {
      "epoch": 6.314858490566038,
      "grad_norm": 1.2783067226409912,
      "learning_rate": 8.648264977058344e-06,
      "loss": 0.3159,
      "num_input_tokens_seen": 6949448,
      "step": 10710
    },
    {
      "epoch": 6.317806603773585,
      "grad_norm": 2.89835262298584,
      "learning_rate": 8.646505227008197e-06,
      "loss": 0.5989,
      "num_input_tokens_seen": 6952360,
      "step": 10715
    },
    {
      "epoch": 6.320754716981132,
      "grad_norm": 2.4917800426483154,
      "learning_rate": 8.644744511530074e-06,
      "loss": 0.3966,
      "num_input_tokens_seen": 6955208,
      "step": 10720
    },
    {
      "epoch": 6.3237028301886795,
      "grad_norm": 1.640005111694336,
      "learning_rate": 8.642982831090135e-06,
      "loss": 0.4536,
      "num_input_tokens_seen": 6958408,
      "step": 10725
    },
    {
      "epoch": 6.326650943396227,
      "grad_norm": 2.6420016288757324,
      "learning_rate": 8.64122018615479e-06,
      "loss": 0.4061,
      "num_input_tokens_seen": 6960936,
      "step": 10730
    },
    {
      "epoch": 6.329599056603773,
      "grad_norm": 3.230102777481079,
      "learning_rate": 8.63945657719071e-06,
      "loss": 0.4779,
      "num_input_tokens_seen": 6964488,
      "step": 10735
    },
    {
      "epoch": 6.3325471698113205,
      "grad_norm": 3.558576822280884,
      "learning_rate": 8.637692004664816e-06,
      "loss": 0.3827,
      "num_input_tokens_seen": 6967688,
      "step": 10740
    },
    {
      "epoch": 6.335495283018868,
      "grad_norm": 3.073190450668335,
      "learning_rate": 8.635926469044284e-06,
      "loss": 0.4454,
      "num_input_tokens_seen": 6970536,
      "step": 10745
    },
    {
      "epoch": 6.338443396226415,
      "grad_norm": 3.213703155517578,
      "learning_rate": 8.63415997079655e-06,
      "loss": 0.4242,
      "num_input_tokens_seen": 6972968,
      "step": 10750
    },
    {
      "epoch": 6.341391509433962,
      "grad_norm": 1.688247561454773,
      "learning_rate": 8.6323925103893e-06,
      "loss": 0.4628,
      "num_input_tokens_seen": 6976392,
      "step": 10755
    },
    {
      "epoch": 6.34433962264151,
      "grad_norm": 3.455904722213745,
      "learning_rate": 8.63062408829048e-06,
      "loss": 0.4137,
      "num_input_tokens_seen": 6979784,
      "step": 10760
    },
    {
      "epoch": 6.347287735849057,
      "grad_norm": 2.5560660362243652,
      "learning_rate": 8.628854704968285e-06,
      "loss": 0.3222,
      "num_input_tokens_seen": 6982024,
      "step": 10765
    },
    {
      "epoch": 6.350235849056604,
      "grad_norm": 3.274998903274536,
      "learning_rate": 8.62708436089117e-06,
      "loss": 0.3032,
      "num_input_tokens_seen": 6984296,
      "step": 10770
    },
    {
      "epoch": 6.353183962264151,
      "grad_norm": 1.3598414659500122,
      "learning_rate": 8.625313056527836e-06,
      "loss": 0.445,
      "num_input_tokens_seen": 6988520,
      "step": 10775
    },
    {
      "epoch": 6.356132075471698,
      "grad_norm": 1.5927530527114868,
      "learning_rate": 8.623540792347244e-06,
      "loss": 0.4258,
      "num_input_tokens_seen": 6992488,
      "step": 10780
    },
    {
      "epoch": 6.359080188679245,
      "grad_norm": 2.100494384765625,
      "learning_rate": 8.621767568818614e-06,
      "loss": 0.3887,
      "num_input_tokens_seen": 6995400,
      "step": 10785
    },
    {
      "epoch": 6.3620283018867925,
      "grad_norm": 2.858480930328369,
      "learning_rate": 8.619993386411409e-06,
      "loss": 0.4355,
      "num_input_tokens_seen": 6998184,
      "step": 10790
    },
    {
      "epoch": 6.36497641509434,
      "grad_norm": 1.1012829542160034,
      "learning_rate": 8.618218245595356e-06,
      "loss": 0.2707,
      "num_input_tokens_seen": 7001320,
      "step": 10795
    },
    {
      "epoch": 6.367924528301887,
      "grad_norm": 1.6342871189117432,
      "learning_rate": 8.616442146840427e-06,
      "loss": 0.5452,
      "num_input_tokens_seen": 7004488,
      "step": 10800
    },
    {
      "epoch": 6.370872641509434,
      "grad_norm": 2.5880258083343506,
      "learning_rate": 8.614665090616854e-06,
      "loss": 0.459,
      "num_input_tokens_seen": 7007560,
      "step": 10805
    },
    {
      "epoch": 6.373820754716981,
      "grad_norm": 1.2599232196807861,
      "learning_rate": 8.61288707739512e-06,
      "loss": 0.4868,
      "num_input_tokens_seen": 7010056,
      "step": 10810
    },
    {
      "epoch": 6.376768867924528,
      "grad_norm": 2.044738292694092,
      "learning_rate": 8.611108107645963e-06,
      "loss": 0.4159,
      "num_input_tokens_seen": 7014152,
      "step": 10815
    },
    {
      "epoch": 6.379716981132075,
      "grad_norm": 1.9258297681808472,
      "learning_rate": 8.609328181840368e-06,
      "loss": 0.3833,
      "num_input_tokens_seen": 7017608,
      "step": 10820
    },
    {
      "epoch": 6.382665094339623,
      "grad_norm": 2.4400839805603027,
      "learning_rate": 8.607547300449585e-06,
      "loss": 0.4639,
      "num_input_tokens_seen": 7020584,
      "step": 10825
    },
    {
      "epoch": 6.38561320754717,
      "grad_norm": 2.009814977645874,
      "learning_rate": 8.605765463945105e-06,
      "loss": 0.5657,
      "num_input_tokens_seen": 7023432,
      "step": 10830
    },
    {
      "epoch": 6.388561320754717,
      "grad_norm": 1.5503244400024414,
      "learning_rate": 8.603982672798678e-06,
      "loss": 0.427,
      "num_input_tokens_seen": 7026312,
      "step": 10835
    },
    {
      "epoch": 6.3915094339622645,
      "grad_norm": 4.281957626342773,
      "learning_rate": 8.602198927482309e-06,
      "loss": 0.4072,
      "num_input_tokens_seen": 7029480,
      "step": 10840
    },
    {
      "epoch": 6.394457547169811,
      "grad_norm": 2.373079776763916,
      "learning_rate": 8.600414228468245e-06,
      "loss": 0.3507,
      "num_input_tokens_seen": 7032008,
      "step": 10845
    },
    {
      "epoch": 6.397405660377358,
      "grad_norm": 6.679205417633057,
      "learning_rate": 8.598628576229e-06,
      "loss": 0.3852,
      "num_input_tokens_seen": 7035112,
      "step": 10850
    },
    {
      "epoch": 6.400353773584905,
      "grad_norm": 2.1905996799468994,
      "learning_rate": 8.596841971237328e-06,
      "loss": 0.4662,
      "num_input_tokens_seen": 7037704,
      "step": 10855
    },
    {
      "epoch": 6.403301886792453,
      "grad_norm": 2.206113338470459,
      "learning_rate": 8.595054413966246e-06,
      "loss": 0.6341,
      "num_input_tokens_seen": 7041320,
      "step": 10860
    },
    {
      "epoch": 6.40625,
      "grad_norm": 3.423949718475342,
      "learning_rate": 8.593265904889011e-06,
      "loss": 0.4647,
      "num_input_tokens_seen": 7043816,
      "step": 10865
    },
    {
      "epoch": 6.409198113207547,
      "grad_norm": 2.0427162647247314,
      "learning_rate": 8.591476444479141e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 7046472,
      "step": 10870
    },
    {
      "epoch": 6.412146226415095,
      "grad_norm": 2.6608948707580566,
      "learning_rate": 8.589686033210407e-06,
      "loss": 0.3814,
      "num_input_tokens_seen": 7049160,
      "step": 10875
    },
    {
      "epoch": 6.415094339622642,
      "grad_norm": 1.4788219928741455,
      "learning_rate": 8.587894671556823e-06,
      "loss": 0.4464,
      "num_input_tokens_seen": 7052968,
      "step": 10880
    },
    {
      "epoch": 6.418042452830189,
      "grad_norm": 1.2031573057174683,
      "learning_rate": 8.586102359992663e-06,
      "loss": 0.3879,
      "num_input_tokens_seen": 7056200,
      "step": 10885
    },
    {
      "epoch": 6.4209905660377355,
      "grad_norm": 2.5390360355377197,
      "learning_rate": 8.584309098992447e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 7059272,
      "step": 10890
    },
    {
      "epoch": 6.423938679245283,
      "grad_norm": 3.6403615474700928,
      "learning_rate": 8.58251488903095e-06,
      "loss": 0.4445,
      "num_input_tokens_seen": 7061640,
      "step": 10895
    },
    {
      "epoch": 6.42688679245283,
      "grad_norm": 2.051042318344116,
      "learning_rate": 8.580719730583196e-06,
      "loss": 0.4514,
      "num_input_tokens_seen": 7064808,
      "step": 10900
    },
    {
      "epoch": 6.429834905660377,
      "grad_norm": 1.552900791168213,
      "learning_rate": 8.578923624124462e-06,
      "loss": 0.3385,
      "num_input_tokens_seen": 7068488,
      "step": 10905
    },
    {
      "epoch": 6.432783018867925,
      "grad_norm": 1.8002092838287354,
      "learning_rate": 8.577126570130273e-06,
      "loss": 0.4301,
      "num_input_tokens_seen": 7071688,
      "step": 10910
    },
    {
      "epoch": 6.435731132075472,
      "grad_norm": 1.6404025554656982,
      "learning_rate": 8.575328569076408e-06,
      "loss": 0.3625,
      "num_input_tokens_seen": 7074952,
      "step": 10915
    },
    {
      "epoch": 6.438679245283019,
      "grad_norm": 1.2198235988616943,
      "learning_rate": 8.573529621438896e-06,
      "loss": 0.4451,
      "num_input_tokens_seen": 7079016,
      "step": 10920
    },
    {
      "epoch": 6.441627358490566,
      "grad_norm": 2.0937435626983643,
      "learning_rate": 8.571729727694015e-06,
      "loss": 0.4458,
      "num_input_tokens_seen": 7082024,
      "step": 10925
    },
    {
      "epoch": 6.444575471698113,
      "grad_norm": 2.129958152770996,
      "learning_rate": 8.569928888318298e-06,
      "loss": 0.4793,
      "num_input_tokens_seen": 7085288,
      "step": 10930
    },
    {
      "epoch": 6.44752358490566,
      "grad_norm": 2.426199436187744,
      "learning_rate": 8.56812710378852e-06,
      "loss": 0.4732,
      "num_input_tokens_seen": 7088840,
      "step": 10935
    },
    {
      "epoch": 6.4504716981132075,
      "grad_norm": 1.6548619270324707,
      "learning_rate": 8.566324374581714e-06,
      "loss": 0.3366,
      "num_input_tokens_seen": 7092264,
      "step": 10940
    },
    {
      "epoch": 6.453419811320755,
      "grad_norm": 1.6809580326080322,
      "learning_rate": 8.564520701175158e-06,
      "loss": 0.395,
      "num_input_tokens_seen": 7095240,
      "step": 10945
    },
    {
      "epoch": 6.456367924528302,
      "grad_norm": 2.127981424331665,
      "learning_rate": 8.562716084046387e-06,
      "loss": 0.3325,
      "num_input_tokens_seen": 7098024,
      "step": 10950
    },
    {
      "epoch": 6.459316037735849,
      "grad_norm": 1.5793728828430176,
      "learning_rate": 8.560910523673177e-06,
      "loss": 0.3989,
      "num_input_tokens_seen": 7100616,
      "step": 10955
    },
    {
      "epoch": 6.462264150943396,
      "grad_norm": 2.976280450820923,
      "learning_rate": 8.55910402053356e-06,
      "loss": 0.4667,
      "num_input_tokens_seen": 7103528,
      "step": 10960
    },
    {
      "epoch": 6.465212264150943,
      "grad_norm": 2.1601593494415283,
      "learning_rate": 8.557296575105814e-06,
      "loss": 0.4691,
      "num_input_tokens_seen": 7106696,
      "step": 10965
    },
    {
      "epoch": 6.46816037735849,
      "grad_norm": 2.384552478790283,
      "learning_rate": 8.555488187868469e-06,
      "loss": 0.4396,
      "num_input_tokens_seen": 7109640,
      "step": 10970
    },
    {
      "epoch": 6.471108490566038,
      "grad_norm": 2.947829008102417,
      "learning_rate": 8.5536788593003e-06,
      "loss": 0.3763,
      "num_input_tokens_seen": 7112744,
      "step": 10975
    },
    {
      "epoch": 6.474056603773585,
      "grad_norm": 1.8344777822494507,
      "learning_rate": 8.55186858988034e-06,
      "loss": 0.4425,
      "num_input_tokens_seen": 7115816,
      "step": 10980
    },
    {
      "epoch": 6.477004716981132,
      "grad_norm": 2.5219709873199463,
      "learning_rate": 8.550057380087863e-06,
      "loss": 0.4564,
      "num_input_tokens_seen": 7118920,
      "step": 10985
    },
    {
      "epoch": 6.4799528301886795,
      "grad_norm": 2.3257055282592773,
      "learning_rate": 8.548245230402396e-06,
      "loss": 0.3585,
      "num_input_tokens_seen": 7121384,
      "step": 10990
    },
    {
      "epoch": 6.482900943396227,
      "grad_norm": 2.9642763137817383,
      "learning_rate": 8.546432141303711e-06,
      "loss": 0.5061,
      "num_input_tokens_seen": 7124040,
      "step": 10995
    },
    {
      "epoch": 6.485849056603773,
      "grad_norm": 2.1347403526306152,
      "learning_rate": 8.544618113271833e-06,
      "loss": 0.4566,
      "num_input_tokens_seen": 7127560,
      "step": 11000
    },
    {
      "epoch": 6.4887971698113205,
      "grad_norm": 2.393862247467041,
      "learning_rate": 8.542803146787032e-06,
      "loss": 0.3842,
      "num_input_tokens_seen": 7130568,
      "step": 11005
    },
    {
      "epoch": 6.491745283018868,
      "grad_norm": 1.7894411087036133,
      "learning_rate": 8.54098724232983e-06,
      "loss": 0.5719,
      "num_input_tokens_seen": 7133800,
      "step": 11010
    },
    {
      "epoch": 6.494693396226415,
      "grad_norm": 1.7923866510391235,
      "learning_rate": 8.539170400380994e-06,
      "loss": 0.5441,
      "num_input_tokens_seen": 7137672,
      "step": 11015
    },
    {
      "epoch": 6.497641509433962,
      "grad_norm": 2.274230718612671,
      "learning_rate": 8.537352621421542e-06,
      "loss": 0.5342,
      "num_input_tokens_seen": 7141224,
      "step": 11020
    },
    {
      "epoch": 6.50058962264151,
      "grad_norm": 2.941741466522217,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.3884,
      "num_input_tokens_seen": 7143624,
      "step": 11025
    },
    {
      "epoch": 6.503537735849057,
      "grad_norm": 2.164994239807129,
      "learning_rate": 8.533714254396096e-06,
      "loss": 0.5298,
      "num_input_tokens_seen": 7147912,
      "step": 11030
    },
    {
      "epoch": 6.506485849056604,
      "grad_norm": 2.347034215927124,
      "learning_rate": 8.531893667293375e-06,
      "loss": 0.4851,
      "num_input_tokens_seen": 7151784,
      "step": 11035
    },
    {
      "epoch": 6.509433962264151,
      "grad_norm": 1.9671571254730225,
      "learning_rate": 8.530072145106585e-06,
      "loss": 0.3579,
      "num_input_tokens_seen": 7154824,
      "step": 11040
    },
    {
      "epoch": 6.512382075471698,
      "grad_norm": 2.377157688140869,
      "learning_rate": 8.528249688317978e-06,
      "loss": 0.5009,
      "num_input_tokens_seen": 7157832,
      "step": 11045
    },
    {
      "epoch": 6.515330188679245,
      "grad_norm": 2.0505211353302,
      "learning_rate": 8.526426297410062e-06,
      "loss": 0.4518,
      "num_input_tokens_seen": 7161288,
      "step": 11050
    },
    {
      "epoch": 6.5182783018867925,
      "grad_norm": 1.893911361694336,
      "learning_rate": 8.524601972865586e-06,
      "loss": 0.4475,
      "num_input_tokens_seen": 7164872,
      "step": 11055
    },
    {
      "epoch": 6.52122641509434,
      "grad_norm": 1.4119459390640259,
      "learning_rate": 8.522776715167548e-06,
      "loss": 0.5432,
      "num_input_tokens_seen": 7168200,
      "step": 11060
    },
    {
      "epoch": 6.524174528301887,
      "grad_norm": 2.177902936935425,
      "learning_rate": 8.520950524799192e-06,
      "loss": 0.3679,
      "num_input_tokens_seen": 7171560,
      "step": 11065
    },
    {
      "epoch": 6.527122641509434,
      "grad_norm": 2.2474365234375,
      "learning_rate": 8.51912340224401e-06,
      "loss": 0.3652,
      "num_input_tokens_seen": 7174408,
      "step": 11070
    },
    {
      "epoch": 6.530070754716981,
      "grad_norm": 3.083353281021118,
      "learning_rate": 8.51729534798574e-06,
      "loss": 0.5057,
      "num_input_tokens_seen": 7177928,
      "step": 11075
    },
    {
      "epoch": 6.533018867924528,
      "grad_norm": 3.896127462387085,
      "learning_rate": 8.515466362508369e-06,
      "loss": 0.5592,
      "num_input_tokens_seen": 7181000,
      "step": 11080
    },
    {
      "epoch": 6.535966981132075,
      "grad_norm": 2.2390053272247314,
      "learning_rate": 8.513636446296125e-06,
      "loss": 0.5232,
      "num_input_tokens_seen": 7184616,
      "step": 11085
    },
    {
      "epoch": 6.538915094339623,
      "grad_norm": 2.2274248600006104,
      "learning_rate": 8.51180559983349e-06,
      "loss": 0.3686,
      "num_input_tokens_seen": 7187880,
      "step": 11090
    },
    {
      "epoch": 6.54186320754717,
      "grad_norm": 2.558276414871216,
      "learning_rate": 8.50997382360519e-06,
      "loss": 0.4952,
      "num_input_tokens_seen": 7190600,
      "step": 11095
    },
    {
      "epoch": 6.544811320754717,
      "grad_norm": 1.6795471906661987,
      "learning_rate": 8.508141118096191e-06,
      "loss": 0.4085,
      "num_input_tokens_seen": 7194536,
      "step": 11100
    },
    {
      "epoch": 6.5477594339622645,
      "grad_norm": 1.3980463743209839,
      "learning_rate": 8.506307483791712e-06,
      "loss": 0.4134,
      "num_input_tokens_seen": 7197768,
      "step": 11105
    },
    {
      "epoch": 6.550707547169811,
      "grad_norm": 3.0726377964019775,
      "learning_rate": 8.504472921177215e-06,
      "loss": 0.3831,
      "num_input_tokens_seen": 7200840,
      "step": 11110
    },
    {
      "epoch": 6.553655660377358,
      "grad_norm": 1.1849699020385742,
      "learning_rate": 8.502637430738409e-06,
      "loss": 0.4159,
      "num_input_tokens_seen": 7203688,
      "step": 11115
    },
    {
      "epoch": 6.556603773584905,
      "grad_norm": 2.232661485671997,
      "learning_rate": 8.500801012961248e-06,
      "loss": 0.607,
      "num_input_tokens_seen": 7207368,
      "step": 11120
    },
    {
      "epoch": 6.559551886792453,
      "grad_norm": 1.8268146514892578,
      "learning_rate": 8.49896366833193e-06,
      "loss": 0.4591,
      "num_input_tokens_seen": 7210120,
      "step": 11125
    },
    {
      "epoch": 6.5625,
      "grad_norm": 3.191988945007324,
      "learning_rate": 8.497125397336903e-06,
      "loss": 0.6076,
      "num_input_tokens_seen": 7212968,
      "step": 11130
    },
    {
      "epoch": 6.565448113207547,
      "grad_norm": 2.0926408767700195,
      "learning_rate": 8.495286200462854e-06,
      "loss": 0.4578,
      "num_input_tokens_seen": 7215496,
      "step": 11135
    },
    {
      "epoch": 6.568396226415095,
      "grad_norm": 1.189227819442749,
      "learning_rate": 8.49344607819672e-06,
      "loss": 0.378,
      "num_input_tokens_seen": 7218440,
      "step": 11140
    },
    {
      "epoch": 6.571344339622642,
      "grad_norm": 5.262424945831299,
      "learning_rate": 8.49160503102568e-06,
      "loss": 0.3459,
      "num_input_tokens_seen": 7220488,
      "step": 11145
    },
    {
      "epoch": 6.574292452830189,
      "grad_norm": 1.3843646049499512,
      "learning_rate": 8.489763059437161e-06,
      "loss": 0.3791,
      "num_input_tokens_seen": 7223112,
      "step": 11150
    },
    {
      "epoch": 6.5772405660377355,
      "grad_norm": 1.6262997388839722,
      "learning_rate": 8.487920163918833e-06,
      "loss": 0.42,
      "num_input_tokens_seen": 7226472,
      "step": 11155
    },
    {
      "epoch": 6.580188679245283,
      "grad_norm": 1.9971288442611694,
      "learning_rate": 8.486076344958607e-06,
      "loss": 0.5262,
      "num_input_tokens_seen": 7229640,
      "step": 11160
    },
    {
      "epoch": 6.58313679245283,
      "grad_norm": 1.4269130229949951,
      "learning_rate": 8.484231603044647e-06,
      "loss": 0.3978,
      "num_input_tokens_seen": 7232200,
      "step": 11165
    },
    {
      "epoch": 6.586084905660377,
      "grad_norm": 2.1329102516174316,
      "learning_rate": 8.482385938665352e-06,
      "loss": 0.4053,
      "num_input_tokens_seen": 7234984,
      "step": 11170
    },
    {
      "epoch": 6.589033018867925,
      "grad_norm": 2.977388620376587,
      "learning_rate": 8.480539352309373e-06,
      "loss": 0.4477,
      "num_input_tokens_seen": 7238472,
      "step": 11175
    },
    {
      "epoch": 6.591981132075472,
      "grad_norm": 2.075018882751465,
      "learning_rate": 8.478691844465598e-06,
      "loss": 0.3814,
      "num_input_tokens_seen": 7241672,
      "step": 11180
    },
    {
      "epoch": 6.594929245283019,
      "grad_norm": 2.544499397277832,
      "learning_rate": 8.476843415623168e-06,
      "loss": 0.5468,
      "num_input_tokens_seen": 7246312,
      "step": 11185
    },
    {
      "epoch": 6.597877358490566,
      "grad_norm": 3.320934534072876,
      "learning_rate": 8.474994066271458e-06,
      "loss": 0.5309,
      "num_input_tokens_seen": 7249224,
      "step": 11190
    },
    {
      "epoch": 6.600825471698113,
      "grad_norm": 4.3731160163879395,
      "learning_rate": 8.473143796900089e-06,
      "loss": 0.5036,
      "num_input_tokens_seen": 7251848,
      "step": 11195
    },
    {
      "epoch": 6.60377358490566,
      "grad_norm": 1.6850727796554565,
      "learning_rate": 8.471292607998936e-06,
      "loss": 0.4318,
      "num_input_tokens_seen": 7255016,
      "step": 11200
    },
    {
      "epoch": 6.6067216981132075,
      "grad_norm": 2.166804790496826,
      "learning_rate": 8.469440500058104e-06,
      "loss": 0.4773,
      "num_input_tokens_seen": 7258440,
      "step": 11205
    },
    {
      "epoch": 6.609669811320755,
      "grad_norm": 2.89401912689209,
      "learning_rate": 8.467587473567945e-06,
      "loss": 0.4792,
      "num_input_tokens_seen": 7260936,
      "step": 11210
    },
    {
      "epoch": 6.612617924528302,
      "grad_norm": 2.7684266567230225,
      "learning_rate": 8.46573352901906e-06,
      "loss": 0.373,
      "num_input_tokens_seen": 7263784,
      "step": 11215
    },
    {
      "epoch": 6.615566037735849,
      "grad_norm": 1.921210765838623,
      "learning_rate": 8.463878666902286e-06,
      "loss": 0.4919,
      "num_input_tokens_seen": 7266792,
      "step": 11220
    },
    {
      "epoch": 6.618514150943396,
      "grad_norm": 1.2632955312728882,
      "learning_rate": 8.462022887708706e-06,
      "loss": 0.5436,
      "num_input_tokens_seen": 7270952,
      "step": 11225
    },
    {
      "epoch": 6.621462264150943,
      "grad_norm": 2.168422222137451,
      "learning_rate": 8.460166191929646e-06,
      "loss": 0.5313,
      "num_input_tokens_seen": 7273512,
      "step": 11230
    },
    {
      "epoch": 6.62441037735849,
      "grad_norm": 1.8100581169128418,
      "learning_rate": 8.458308580056675e-06,
      "loss": 0.401,
      "num_input_tokens_seen": 7276456,
      "step": 11235
    },
    {
      "epoch": 6.627358490566038,
      "grad_norm": 3.9232161045074463,
      "learning_rate": 8.456450052581602e-06,
      "loss": 0.3488,
      "num_input_tokens_seen": 7279368,
      "step": 11240
    },
    {
      "epoch": 6.630306603773585,
      "grad_norm": 2.128993034362793,
      "learning_rate": 8.45459060999648e-06,
      "loss": 0.4542,
      "num_input_tokens_seen": 7282664,
      "step": 11245
    },
    {
      "epoch": 6.633254716981132,
      "grad_norm": 2.5209434032440186,
      "learning_rate": 8.452730252793608e-06,
      "loss": 0.4163,
      "num_input_tokens_seen": 7285704,
      "step": 11250
    },
    {
      "epoch": 6.6362028301886795,
      "grad_norm": 1.6096867322921753,
      "learning_rate": 8.450868981465519e-06,
      "loss": 0.3363,
      "num_input_tokens_seen": 7288808,
      "step": 11255
    },
    {
      "epoch": 6.639150943396227,
      "grad_norm": 1.6681891679763794,
      "learning_rate": 8.449006796504997e-06,
      "loss": 0.3001,
      "num_input_tokens_seen": 7292584,
      "step": 11260
    },
    {
      "epoch": 6.642099056603773,
      "grad_norm": 2.264650583267212,
      "learning_rate": 8.44714369840506e-06,
      "loss": 0.5152,
      "num_input_tokens_seen": 7295656,
      "step": 11265
    },
    {
      "epoch": 6.6450471698113205,
      "grad_norm": 2.233914613723755,
      "learning_rate": 8.445279687658973e-06,
      "loss": 0.4898,
      "num_input_tokens_seen": 7298600,
      "step": 11270
    },
    {
      "epoch": 6.647995283018868,
      "grad_norm": 1.1914734840393066,
      "learning_rate": 8.44341476476024e-06,
      "loss": 0.3471,
      "num_input_tokens_seen": 7301544,
      "step": 11275
    },
    {
      "epoch": 6.650943396226415,
      "grad_norm": 1.775597095489502,
      "learning_rate": 8.441548930202608e-06,
      "loss": 0.4783,
      "num_input_tokens_seen": 7305128,
      "step": 11280
    },
    {
      "epoch": 6.653891509433962,
      "grad_norm": 2.0608415603637695,
      "learning_rate": 8.439682184480065e-06,
      "loss": 0.5176,
      "num_input_tokens_seen": 7308328,
      "step": 11285
    },
    {
      "epoch": 6.65683962264151,
      "grad_norm": 2.694744825363159,
      "learning_rate": 8.437814528086837e-06,
      "loss": 0.4839,
      "num_input_tokens_seen": 7311048,
      "step": 11290
    },
    {
      "epoch": 6.659787735849057,
      "grad_norm": 1.9932403564453125,
      "learning_rate": 8.435945961517398e-06,
      "loss": 0.4593,
      "num_input_tokens_seen": 7314184,
      "step": 11295
    },
    {
      "epoch": 6.662735849056604,
      "grad_norm": 3.18202805519104,
      "learning_rate": 8.434076485266458e-06,
      "loss": 0.2913,
      "num_input_tokens_seen": 7318056,
      "step": 11300
    },
    {
      "epoch": 6.665683962264151,
      "grad_norm": 2.470616102218628,
      "learning_rate": 8.432206099828969e-06,
      "loss": 0.3547,
      "num_input_tokens_seen": 7320808,
      "step": 11305
    },
    {
      "epoch": 6.668632075471698,
      "grad_norm": 1.5146888494491577,
      "learning_rate": 8.430334805700122e-06,
      "loss": 0.413,
      "num_input_tokens_seen": 7324040,
      "step": 11310
    },
    {
      "epoch": 6.671580188679245,
      "grad_norm": 3.0046467781066895,
      "learning_rate": 8.428462603375351e-06,
      "loss": 0.3933,
      "num_input_tokens_seen": 7327112,
      "step": 11315
    },
    {
      "epoch": 6.6745283018867925,
      "grad_norm": 2.4517860412597656,
      "learning_rate": 8.426589493350332e-06,
      "loss": 0.3622,
      "num_input_tokens_seen": 7330088,
      "step": 11320
    },
    {
      "epoch": 6.67747641509434,
      "grad_norm": 4.337218761444092,
      "learning_rate": 8.424715476120976e-06,
      "loss": 0.4997,
      "num_input_tokens_seen": 7333032,
      "step": 11325
    },
    {
      "epoch": 6.680424528301887,
      "grad_norm": 1.2546024322509766,
      "learning_rate": 8.422840552183437e-06,
      "loss": 0.3764,
      "num_input_tokens_seen": 7336616,
      "step": 11330
    },
    {
      "epoch": 6.683372641509434,
      "grad_norm": 1.6593834161758423,
      "learning_rate": 8.420964722034111e-06,
      "loss": 0.4901,
      "num_input_tokens_seen": 7339720,
      "step": 11335
    },
    {
      "epoch": 6.686320754716981,
      "grad_norm": 1.9273672103881836,
      "learning_rate": 8.41908798616963e-06,
      "loss": 0.3712,
      "num_input_tokens_seen": 7342664,
      "step": 11340
    },
    {
      "epoch": 6.689268867924528,
      "grad_norm": 1.940942645072937,
      "learning_rate": 8.41721034508687e-06,
      "loss": 0.305,
      "num_input_tokens_seen": 7346664,
      "step": 11345
    },
    {
      "epoch": 6.692216981132075,
      "grad_norm": 2.7236220836639404,
      "learning_rate": 8.415331799282942e-06,
      "loss": 0.4021,
      "num_input_tokens_seen": 7349288,
      "step": 11350
    },
    {
      "epoch": 6.695165094339623,
      "grad_norm": 2.3041553497314453,
      "learning_rate": 8.413452349255205e-06,
      "loss": 0.4018,
      "num_input_tokens_seen": 7352104,
      "step": 11355
    },
    {
      "epoch": 6.69811320754717,
      "grad_norm": 2.787003755569458,
      "learning_rate": 8.411571995501245e-06,
      "loss": 0.4364,
      "num_input_tokens_seen": 7355880,
      "step": 11360
    },
    {
      "epoch": 6.701061320754717,
      "grad_norm": 3.6605989933013916,
      "learning_rate": 8.409690738518895e-06,
      "loss": 0.7467,
      "num_input_tokens_seen": 7358760,
      "step": 11365
    },
    {
      "epoch": 6.7040094339622645,
      "grad_norm": 2.30658221244812,
      "learning_rate": 8.407808578806229e-06,
      "loss": 0.5125,
      "num_input_tokens_seen": 7361928,
      "step": 11370
    },
    {
      "epoch": 6.706957547169811,
      "grad_norm": 1.7174558639526367,
      "learning_rate": 8.405925516861555e-06,
      "loss": 0.3889,
      "num_input_tokens_seen": 7364712,
      "step": 11375
    },
    {
      "epoch": 6.709905660377358,
      "grad_norm": 1.7337075471878052,
      "learning_rate": 8.40404155318342e-06,
      "loss": 0.4185,
      "num_input_tokens_seen": 7367464,
      "step": 11380
    },
    {
      "epoch": 6.712853773584905,
      "grad_norm": 1.1992436647415161,
      "learning_rate": 8.402156688270613e-06,
      "loss": 0.3965,
      "num_input_tokens_seen": 7370184,
      "step": 11385
    },
    {
      "epoch": 6.715801886792453,
      "grad_norm": 1.7200978994369507,
      "learning_rate": 8.400270922622162e-06,
      "loss": 0.4838,
      "num_input_tokens_seen": 7373736,
      "step": 11390
    },
    {
      "epoch": 6.71875,
      "grad_norm": 2.1535794734954834,
      "learning_rate": 8.398384256737328e-06,
      "loss": 0.4318,
      "num_input_tokens_seen": 7377000,
      "step": 11395
    },
    {
      "epoch": 6.721698113207547,
      "grad_norm": 3.597007989883423,
      "learning_rate": 8.396496691115619e-06,
      "loss": 0.3868,
      "num_input_tokens_seen": 7381192,
      "step": 11400
    },
    {
      "epoch": 6.724646226415095,
      "grad_norm": 2.960001230239868,
      "learning_rate": 8.39460822625677e-06,
      "loss": 0.4256,
      "num_input_tokens_seen": 7383880,
      "step": 11405
    },
    {
      "epoch": 6.727594339622642,
      "grad_norm": 1.9399093389511108,
      "learning_rate": 8.392718862660765e-06,
      "loss": 0.4938,
      "num_input_tokens_seen": 7388232,
      "step": 11410
    },
    {
      "epoch": 6.730542452830189,
      "grad_norm": 2.3211140632629395,
      "learning_rate": 8.390828600827818e-06,
      "loss": 0.4145,
      "num_input_tokens_seen": 7390856,
      "step": 11415
    },
    {
      "epoch": 6.7334905660377355,
      "grad_norm": 2.8020994663238525,
      "learning_rate": 8.388937441258385e-06,
      "loss": 0.4008,
      "num_input_tokens_seen": 7394536,
      "step": 11420
    },
    {
      "epoch": 6.736438679245283,
      "grad_norm": 2.4525487422943115,
      "learning_rate": 8.387045384453162e-06,
      "loss": 0.368,
      "num_input_tokens_seen": 7399848,
      "step": 11425
    },
    {
      "epoch": 6.73938679245283,
      "grad_norm": 1.4748603105545044,
      "learning_rate": 8.385152430913073e-06,
      "loss": 0.4302,
      "num_input_tokens_seen": 7403688,
      "step": 11430
    },
    {
      "epoch": 6.742334905660377,
      "grad_norm": 2.1369242668151855,
      "learning_rate": 8.383258581139288e-06,
      "loss": 0.4899,
      "num_input_tokens_seen": 7407240,
      "step": 11435
    },
    {
      "epoch": 6.745283018867925,
      "grad_norm": 1.7670793533325195,
      "learning_rate": 8.381363835633213e-06,
      "loss": 0.3973,
      "num_input_tokens_seen": 7410344,
      "step": 11440
    },
    {
      "epoch": 6.748231132075472,
      "grad_norm": 1.0834170579910278,
      "learning_rate": 8.379468194896492e-06,
      "loss": 0.3473,
      "num_input_tokens_seen": 7413832,
      "step": 11445
    },
    {
      "epoch": 6.751179245283019,
      "grad_norm": 1.8641489744186401,
      "learning_rate": 8.377571659431e-06,
      "loss": 0.4683,
      "num_input_tokens_seen": 7417224,
      "step": 11450
    },
    {
      "epoch": 6.754127358490566,
      "grad_norm": 2.936763048171997,
      "learning_rate": 8.375674229738855e-06,
      "loss": 0.394,
      "num_input_tokens_seen": 7420392,
      "step": 11455
    },
    {
      "epoch": 6.757075471698113,
      "grad_norm": 2.0061302185058594,
      "learning_rate": 8.37377590632241e-06,
      "loss": 0.3335,
      "num_input_tokens_seen": 7423304,
      "step": 11460
    },
    {
      "epoch": 6.76002358490566,
      "grad_norm": 1.6961445808410645,
      "learning_rate": 8.371876689684253e-06,
      "loss": 0.4938,
      "num_input_tokens_seen": 7426632,
      "step": 11465
    },
    {
      "epoch": 6.7629716981132075,
      "grad_norm": 1.4886960983276367,
      "learning_rate": 8.369976580327211e-06,
      "loss": 0.3641,
      "num_input_tokens_seen": 7430024,
      "step": 11470
    },
    {
      "epoch": 6.765919811320755,
      "grad_norm": 2.337693214416504,
      "learning_rate": 8.368075578754345e-06,
      "loss": 0.4047,
      "num_input_tokens_seen": 7433224,
      "step": 11475
    },
    {
      "epoch": 6.768867924528302,
      "grad_norm": 1.8670382499694824,
      "learning_rate": 8.366173685468952e-06,
      "loss": 0.6143,
      "num_input_tokens_seen": 7436776,
      "step": 11480
    },
    {
      "epoch": 6.771816037735849,
      "grad_norm": 2.4310710430145264,
      "learning_rate": 8.364270900974572e-06,
      "loss": 0.4373,
      "num_input_tokens_seen": 7439976,
      "step": 11485
    },
    {
      "epoch": 6.774764150943396,
      "grad_norm": 1.808047890663147,
      "learning_rate": 8.362367225774968e-06,
      "loss": 0.3838,
      "num_input_tokens_seen": 7442632,
      "step": 11490
    },
    {
      "epoch": 6.777712264150943,
      "grad_norm": 1.7529487609863281,
      "learning_rate": 8.360462660374153e-06,
      "loss": 0.4344,
      "num_input_tokens_seen": 7446792,
      "step": 11495
    },
    {
      "epoch": 6.78066037735849,
      "grad_norm": 1.240815281867981,
      "learning_rate": 8.358557205276365e-06,
      "loss": 0.3665,
      "num_input_tokens_seen": 7450600,
      "step": 11500
    },
    {
      "epoch": 6.783608490566038,
      "grad_norm": 1.58860182762146,
      "learning_rate": 8.356650860986083e-06,
      "loss": 0.5751,
      "num_input_tokens_seen": 7453960,
      "step": 11505
    },
    {
      "epoch": 6.786556603773585,
      "grad_norm": 1.909836769104004,
      "learning_rate": 8.354743628008017e-06,
      "loss": 0.4724,
      "num_input_tokens_seen": 7456648,
      "step": 11510
    },
    {
      "epoch": 6.789504716981132,
      "grad_norm": 1.8623963594436646,
      "learning_rate": 8.35283550684712e-06,
      "loss": 0.3599,
      "num_input_tokens_seen": 7459336,
      "step": 11515
    },
    {
      "epoch": 6.7924528301886795,
      "grad_norm": 1.3750662803649902,
      "learning_rate": 8.350926498008572e-06,
      "loss": 0.5034,
      "num_input_tokens_seen": 7466056,
      "step": 11520
    },
    {
      "epoch": 6.795400943396227,
      "grad_norm": 1.4876707792282104,
      "learning_rate": 8.34901660199779e-06,
      "loss": 0.4807,
      "num_input_tokens_seen": 7469128,
      "step": 11525
    },
    {
      "epoch": 6.798349056603773,
      "grad_norm": 2.1745986938476562,
      "learning_rate": 8.347105819320432e-06,
      "loss": 0.4768,
      "num_input_tokens_seen": 7472200,
      "step": 11530
    },
    {
      "epoch": 6.8012971698113205,
      "grad_norm": 1.6291277408599854,
      "learning_rate": 8.34519415048238e-06,
      "loss": 0.3481,
      "num_input_tokens_seen": 7475016,
      "step": 11535
    },
    {
      "epoch": 6.804245283018868,
      "grad_norm": 3.7543129920959473,
      "learning_rate": 8.343281595989761e-06,
      "loss": 0.514,
      "num_input_tokens_seen": 7478088,
      "step": 11540
    },
    {
      "epoch": 6.807193396226415,
      "grad_norm": 2.5987114906311035,
      "learning_rate": 8.341368156348933e-06,
      "loss": 0.4199,
      "num_input_tokens_seen": 7480648,
      "step": 11545
    },
    {
      "epoch": 6.810141509433962,
      "grad_norm": 2.654249906539917,
      "learning_rate": 8.339453832066482e-06,
      "loss": 0.6212,
      "num_input_tokens_seen": 7483368,
      "step": 11550
    },
    {
      "epoch": 6.81308962264151,
      "grad_norm": 1.698182463645935,
      "learning_rate": 8.337538623649237e-06,
      "loss": 0.4278,
      "num_input_tokens_seen": 7487464,
      "step": 11555
    },
    {
      "epoch": 6.816037735849057,
      "grad_norm": 3.123490810394287,
      "learning_rate": 8.33562253160426e-06,
      "loss": 0.4935,
      "num_input_tokens_seen": 7491368,
      "step": 11560
    },
    {
      "epoch": 6.818985849056604,
      "grad_norm": 1.6551620960235596,
      "learning_rate": 8.33370555643884e-06,
      "loss": 0.2806,
      "num_input_tokens_seen": 7493736,
      "step": 11565
    },
    {
      "epoch": 6.821933962264151,
      "grad_norm": 4.777719020843506,
      "learning_rate": 8.331787698660507e-06,
      "loss": 0.4537,
      "num_input_tokens_seen": 7496168,
      "step": 11570
    },
    {
      "epoch": 6.824882075471698,
      "grad_norm": 3.0781102180480957,
      "learning_rate": 8.32986895877702e-06,
      "loss": 0.3524,
      "num_input_tokens_seen": 7499272,
      "step": 11575
    },
    {
      "epoch": 6.827830188679245,
      "grad_norm": 2.1195828914642334,
      "learning_rate": 8.327949337296378e-06,
      "loss": 0.4358,
      "num_input_tokens_seen": 7502824,
      "step": 11580
    },
    {
      "epoch": 6.8307783018867925,
      "grad_norm": 2.6140520572662354,
      "learning_rate": 8.326028834726803e-06,
      "loss": 0.5368,
      "num_input_tokens_seen": 7507624,
      "step": 11585
    },
    {
      "epoch": 6.83372641509434,
      "grad_norm": 1.4449853897094727,
      "learning_rate": 8.324107451576762e-06,
      "loss": 0.3634,
      "num_input_tokens_seen": 7510280,
      "step": 11590
    },
    {
      "epoch": 6.836674528301887,
      "grad_norm": 1.7970553636550903,
      "learning_rate": 8.322185188354947e-06,
      "loss": 0.3421,
      "num_input_tokens_seen": 7513000,
      "step": 11595
    },
    {
      "epoch": 6.839622641509434,
      "grad_norm": 2.0951409339904785,
      "learning_rate": 8.320262045570284e-06,
      "loss": 0.4016,
      "num_input_tokens_seen": 7516072,
      "step": 11600
    },
    {
      "epoch": 6.842570754716981,
      "grad_norm": 1.809987187385559,
      "learning_rate": 8.318338023731937e-06,
      "loss": 0.4442,
      "num_input_tokens_seen": 7519656,
      "step": 11605
    },
    {
      "epoch": 6.845518867924528,
      "grad_norm": 2.418710708618164,
      "learning_rate": 8.316413123349296e-06,
      "loss": 0.4445,
      "num_input_tokens_seen": 7522472,
      "step": 11610
    },
    {
      "epoch": 6.848466981132075,
      "grad_norm": 2.124987840652466,
      "learning_rate": 8.314487344931987e-06,
      "loss": 0.3753,
      "num_input_tokens_seen": 7525608,
      "step": 11615
    },
    {
      "epoch": 6.851415094339623,
      "grad_norm": 1.6468546390533447,
      "learning_rate": 8.31256068898987e-06,
      "loss": 0.5203,
      "num_input_tokens_seen": 7529192,
      "step": 11620
    },
    {
      "epoch": 6.85436320754717,
      "grad_norm": 1.5583869218826294,
      "learning_rate": 8.310633156033032e-06,
      "loss": 0.4563,
      "num_input_tokens_seen": 7532360,
      "step": 11625
    },
    {
      "epoch": 6.857311320754717,
      "grad_norm": 4.606747627258301,
      "learning_rate": 8.3087047465718e-06,
      "loss": 0.5496,
      "num_input_tokens_seen": 7534600,
      "step": 11630
    },
    {
      "epoch": 6.8602594339622645,
      "grad_norm": 1.7678351402282715,
      "learning_rate": 8.306775461116727e-06,
      "loss": 0.3456,
      "num_input_tokens_seen": 7537704,
      "step": 11635
    },
    {
      "epoch": 6.863207547169811,
      "grad_norm": 2.533665418624878,
      "learning_rate": 8.304845300178597e-06,
      "loss": 0.3805,
      "num_input_tokens_seen": 7540104,
      "step": 11640
    },
    {
      "epoch": 6.866155660377358,
      "grad_norm": 1.9612932205200195,
      "learning_rate": 8.302914264268433e-06,
      "loss": 0.3221,
      "num_input_tokens_seen": 7543272,
      "step": 11645
    },
    {
      "epoch": 6.869103773584905,
      "grad_norm": 2.357779026031494,
      "learning_rate": 8.300982353897482e-06,
      "loss": 0.3577,
      "num_input_tokens_seen": 7546696,
      "step": 11650
    },
    {
      "epoch": 6.872051886792453,
      "grad_norm": 2.2127816677093506,
      "learning_rate": 8.299049569577226e-06,
      "loss": 0.5766,
      "num_input_tokens_seen": 7549640,
      "step": 11655
    },
    {
      "epoch": 6.875,
      "grad_norm": 2.7079415321350098,
      "learning_rate": 8.297115911819379e-06,
      "loss": 0.4999,
      "num_input_tokens_seen": 7553256,
      "step": 11660
    },
    {
      "epoch": 6.877948113207547,
      "grad_norm": 1.3917986154556274,
      "learning_rate": 8.295181381135884e-06,
      "loss": 0.3438,
      "num_input_tokens_seen": 7556360,
      "step": 11665
    },
    {
      "epoch": 6.880896226415095,
      "grad_norm": 2.5484814643859863,
      "learning_rate": 8.293245978038917e-06,
      "loss": 0.4162,
      "num_input_tokens_seen": 7558856,
      "step": 11670
    },
    {
      "epoch": 6.883844339622642,
      "grad_norm": 1.6055843830108643,
      "learning_rate": 8.291309703040884e-06,
      "loss": 0.6197,
      "num_input_tokens_seen": 7562344,
      "step": 11675
    },
    {
      "epoch": 6.886792452830189,
      "grad_norm": 1.4938790798187256,
      "learning_rate": 8.289372556654422e-06,
      "loss": 0.4202,
      "num_input_tokens_seen": 7565320,
      "step": 11680
    },
    {
      "epoch": 6.8897405660377355,
      "grad_norm": 2.1401171684265137,
      "learning_rate": 8.287434539392401e-06,
      "loss": 0.3971,
      "num_input_tokens_seen": 7570024,
      "step": 11685
    },
    {
      "epoch": 6.892688679245283,
      "grad_norm": 1.9436986446380615,
      "learning_rate": 8.285495651767916e-06,
      "loss": 0.4488,
      "num_input_tokens_seen": 7573512,
      "step": 11690
    },
    {
      "epoch": 6.89563679245283,
      "grad_norm": 1.845267415046692,
      "learning_rate": 8.283555894294297e-06,
      "loss": 0.4232,
      "num_input_tokens_seen": 7576200,
      "step": 11695
    },
    {
      "epoch": 6.898584905660377,
      "grad_norm": 1.4661033153533936,
      "learning_rate": 8.281615267485105e-06,
      "loss": 0.547,
      "num_input_tokens_seen": 7579272,
      "step": 11700
    },
    {
      "epoch": 6.901533018867925,
      "grad_norm": 1.9230133295059204,
      "learning_rate": 8.279673771854127e-06,
      "loss": 0.4136,
      "num_input_tokens_seen": 7583624,
      "step": 11705
    },
    {
      "epoch": 6.904481132075472,
      "grad_norm": 2.9017415046691895,
      "learning_rate": 8.277731407915386e-06,
      "loss": 0.5737,
      "num_input_tokens_seen": 7586600,
      "step": 11710
    },
    {
      "epoch": 6.907429245283019,
      "grad_norm": 1.6453189849853516,
      "learning_rate": 8.275788176183126e-06,
      "loss": 0.4657,
      "num_input_tokens_seen": 7589896,
      "step": 11715
    },
    {
      "epoch": 6.910377358490566,
      "grad_norm": 3.3939297199249268,
      "learning_rate": 8.273844077171827e-06,
      "loss": 0.5554,
      "num_input_tokens_seen": 7593736,
      "step": 11720
    },
    {
      "epoch": 6.913325471698113,
      "grad_norm": 2.8460118770599365,
      "learning_rate": 8.271899111396202e-06,
      "loss": 0.5073,
      "num_input_tokens_seen": 7597480,
      "step": 11725
    },
    {
      "epoch": 6.91627358490566,
      "grad_norm": 1.4349445104599,
      "learning_rate": 8.269953279371185e-06,
      "loss": 0.2946,
      "num_input_tokens_seen": 7600296,
      "step": 11730
    },
    {
      "epoch": 6.9192216981132075,
      "grad_norm": 1.3972558975219727,
      "learning_rate": 8.268006581611945e-06,
      "loss": 0.2978,
      "num_input_tokens_seen": 7603400,
      "step": 11735
    },
    {
      "epoch": 6.922169811320755,
      "grad_norm": 1.6703873872756958,
      "learning_rate": 8.266059018633878e-06,
      "loss": 0.5103,
      "num_input_tokens_seen": 7605864,
      "step": 11740
    },
    {
      "epoch": 6.925117924528302,
      "grad_norm": 4.630005359649658,
      "learning_rate": 8.264110590952609e-06,
      "loss": 0.4277,
      "num_input_tokens_seen": 7608520,
      "step": 11745
    },
    {
      "epoch": 6.928066037735849,
      "grad_norm": 3.0758166313171387,
      "learning_rate": 8.262161299083993e-06,
      "loss": 0.4518,
      "num_input_tokens_seen": 7612744,
      "step": 11750
    },
    {
      "epoch": 6.931014150943396,
      "grad_norm": 2.1483089923858643,
      "learning_rate": 8.260211143544117e-06,
      "loss": 0.5166,
      "num_input_tokens_seen": 7615368,
      "step": 11755
    },
    {
      "epoch": 6.933962264150943,
      "grad_norm": 1.3902380466461182,
      "learning_rate": 8.258260124849288e-06,
      "loss": 0.3659,
      "num_input_tokens_seen": 7618216,
      "step": 11760
    },
    {
      "epoch": 6.93691037735849,
      "grad_norm": 1.0176910161972046,
      "learning_rate": 8.256308243516048e-06,
      "loss": 0.5365,
      "num_input_tokens_seen": 7621928,
      "step": 11765
    },
    {
      "epoch": 6.939858490566038,
      "grad_norm": 2.2771852016448975,
      "learning_rate": 8.254355500061168e-06,
      "loss": 0.4997,
      "num_input_tokens_seen": 7625192,
      "step": 11770
    },
    {
      "epoch": 6.942806603773585,
      "grad_norm": 1.4895063638687134,
      "learning_rate": 8.252401895001643e-06,
      "loss": 0.3612,
      "num_input_tokens_seen": 7627816,
      "step": 11775
    },
    {
      "epoch": 6.945754716981132,
      "grad_norm": 1.723905324935913,
      "learning_rate": 8.2504474288547e-06,
      "loss": 0.3947,
      "num_input_tokens_seen": 7631208,
      "step": 11780
    },
    {
      "epoch": 6.9487028301886795,
      "grad_norm": 2.9647772312164307,
      "learning_rate": 8.248492102137791e-06,
      "loss": 0.5121,
      "num_input_tokens_seen": 7634504,
      "step": 11785
    },
    {
      "epoch": 6.951650943396227,
      "grad_norm": 2.3933310508728027,
      "learning_rate": 8.246535915368596e-06,
      "loss": 0.4324,
      "num_input_tokens_seen": 7637800,
      "step": 11790
    },
    {
      "epoch": 6.954599056603773,
      "grad_norm": 2.775925397872925,
      "learning_rate": 8.24457886906503e-06,
      "loss": 0.3427,
      "num_input_tokens_seen": 7640648,
      "step": 11795
    },
    {
      "epoch": 6.9575471698113205,
      "grad_norm": 2.656634569168091,
      "learning_rate": 8.242620963745222e-06,
      "loss": 0.4018,
      "num_input_tokens_seen": 7643656,
      "step": 11800
    },
    {
      "epoch": 6.960495283018868,
      "grad_norm": 2.7303309440612793,
      "learning_rate": 8.240662199927538e-06,
      "loss": 0.5462,
      "num_input_tokens_seen": 7647304,
      "step": 11805
    },
    {
      "epoch": 6.963443396226415,
      "grad_norm": 2.167649745941162,
      "learning_rate": 8.238702578130573e-06,
      "loss": 0.4064,
      "num_input_tokens_seen": 7650408,
      "step": 11810
    },
    {
      "epoch": 6.966391509433962,
      "grad_norm": 2.1612186431884766,
      "learning_rate": 8.23674209887314e-06,
      "loss": 0.3881,
      "num_input_tokens_seen": 7653256,
      "step": 11815
    },
    {
      "epoch": 6.96933962264151,
      "grad_norm": 2.2191073894500732,
      "learning_rate": 8.234780762674288e-06,
      "loss": 0.475,
      "num_input_tokens_seen": 7655976,
      "step": 11820
    },
    {
      "epoch": 6.972287735849057,
      "grad_norm": 1.968300461769104,
      "learning_rate": 8.232818570053286e-06,
      "loss": 0.4451,
      "num_input_tokens_seen": 7659816,
      "step": 11825
    },
    {
      "epoch": 6.975235849056604,
      "grad_norm": 2.4325780868530273,
      "learning_rate": 8.230855521529637e-06,
      "loss": 0.3499,
      "num_input_tokens_seen": 7662952,
      "step": 11830
    },
    {
      "epoch": 6.978183962264151,
      "grad_norm": 2.100184440612793,
      "learning_rate": 8.228891617623064e-06,
      "loss": 0.3129,
      "num_input_tokens_seen": 7665640,
      "step": 11835
    },
    {
      "epoch": 6.981132075471698,
      "grad_norm": 2.527059316635132,
      "learning_rate": 8.22692685885352e-06,
      "loss": 0.4339,
      "num_input_tokens_seen": 7668968,
      "step": 11840
    },
    {
      "epoch": 6.984080188679245,
      "grad_norm": 1.9728819131851196,
      "learning_rate": 8.224961245741183e-06,
      "loss": 0.5077,
      "num_input_tokens_seen": 7673096,
      "step": 11845
    },
    {
      "epoch": 6.9870283018867925,
      "grad_norm": 2.7721810340881348,
      "learning_rate": 8.222994778806457e-06,
      "loss": 0.6117,
      "num_input_tokens_seen": 7677000,
      "step": 11850
    },
    {
      "epoch": 6.98997641509434,
      "grad_norm": 2.02433443069458,
      "learning_rate": 8.221027458569972e-06,
      "loss": 0.3843,
      "num_input_tokens_seen": 7684904,
      "step": 11855
    },
    {
      "epoch": 6.992924528301887,
      "grad_norm": 1.922201156616211,
      "learning_rate": 8.219059285552586e-06,
      "loss": 0.4583,
      "num_input_tokens_seen": 7687368,
      "step": 11860
    },
    {
      "epoch": 6.995872641509434,
      "grad_norm": 0.9193456768989563,
      "learning_rate": 8.21709026027538e-06,
      "loss": 0.5703,
      "num_input_tokens_seen": 7689992,
      "step": 11865
    },
    {
      "epoch": 6.998820754716981,
      "grad_norm": 1.1019582748413086,
      "learning_rate": 8.215120383259664e-06,
      "loss": 0.4093,
      "num_input_tokens_seen": 7693000,
      "step": 11870
    },
    {
      "epoch": 7.001768867924528,
      "grad_norm": 5.491140842437744,
      "learning_rate": 8.21314965502697e-06,
      "loss": 0.3819,
      "num_input_tokens_seen": 7695504,
      "step": 11875
    },
    {
      "epoch": 7.004716981132075,
      "grad_norm": 1.7004413604736328,
      "learning_rate": 8.211178076099056e-06,
      "loss": 0.4064,
      "num_input_tokens_seen": 7699856,
      "step": 11880
    },
    {
      "epoch": 7.007665094339623,
      "grad_norm": 1.4783356189727783,
      "learning_rate": 8.209205646997909e-06,
      "loss": 0.4591,
      "num_input_tokens_seen": 7703376,
      "step": 11885
    },
    {
      "epoch": 7.01061320754717,
      "grad_norm": 2.372725486755371,
      "learning_rate": 8.207232368245735e-06,
      "loss": 0.4007,
      "num_input_tokens_seen": 7705904,
      "step": 11890
    },
    {
      "epoch": 7.013561320754717,
      "grad_norm": 1.5856095552444458,
      "learning_rate": 8.205258240364968e-06,
      "loss": 0.4429,
      "num_input_tokens_seen": 7709072,
      "step": 11895
    },
    {
      "epoch": 7.0165094339622645,
      "grad_norm": 1.5518101453781128,
      "learning_rate": 8.203283263878268e-06,
      "loss": 0.3275,
      "num_input_tokens_seen": 7712208,
      "step": 11900
    },
    {
      "epoch": 7.019457547169812,
      "grad_norm": 1.581186294555664,
      "learning_rate": 8.201307439308518e-06,
      "loss": 0.4339,
      "num_input_tokens_seen": 7716080,
      "step": 11905
    },
    {
      "epoch": 7.022405660377358,
      "grad_norm": 1.6093270778656006,
      "learning_rate": 8.199330767178828e-06,
      "loss": 0.3321,
      "num_input_tokens_seen": 7718832,
      "step": 11910
    },
    {
      "epoch": 7.025353773584905,
      "grad_norm": 2.5458157062530518,
      "learning_rate": 8.19735324801253e-06,
      "loss": 0.3349,
      "num_input_tokens_seen": 7721296,
      "step": 11915
    },
    {
      "epoch": 7.028301886792453,
      "grad_norm": 2.0911364555358887,
      "learning_rate": 8.195374882333178e-06,
      "loss": 0.4194,
      "num_input_tokens_seen": 7724400,
      "step": 11920
    },
    {
      "epoch": 7.03125,
      "grad_norm": 2.3258485794067383,
      "learning_rate": 8.193395670664555e-06,
      "loss": 0.4176,
      "num_input_tokens_seen": 7727824,
      "step": 11925
    },
    {
      "epoch": 7.034198113207547,
      "grad_norm": 1.8924334049224854,
      "learning_rate": 8.191415613530667e-06,
      "loss": 0.5135,
      "num_input_tokens_seen": 7731216,
      "step": 11930
    },
    {
      "epoch": 7.037146226415095,
      "grad_norm": 2.173513412475586,
      "learning_rate": 8.189434711455739e-06,
      "loss": 0.4719,
      "num_input_tokens_seen": 7734096,
      "step": 11935
    },
    {
      "epoch": 7.040094339622642,
      "grad_norm": 1.689131498336792,
      "learning_rate": 8.187452964964226e-06,
      "loss": 0.2987,
      "num_input_tokens_seen": 7737808,
      "step": 11940
    },
    {
      "epoch": 7.043042452830188,
      "grad_norm": 1.7066715955734253,
      "learning_rate": 8.185470374580805e-06,
      "loss": 0.39,
      "num_input_tokens_seen": 7741296,
      "step": 11945
    },
    {
      "epoch": 7.0459905660377355,
      "grad_norm": 1.1549220085144043,
      "learning_rate": 8.183486940830371e-06,
      "loss": 0.2453,
      "num_input_tokens_seen": 7744496,
      "step": 11950
    },
    {
      "epoch": 7.048938679245283,
      "grad_norm": 3.9331917762756348,
      "learning_rate": 8.18150266423805e-06,
      "loss": 0.3667,
      "num_input_tokens_seen": 7747312,
      "step": 11955
    },
    {
      "epoch": 7.05188679245283,
      "grad_norm": 1.950658917427063,
      "learning_rate": 8.179517545329188e-06,
      "loss": 0.3964,
      "num_input_tokens_seen": 7751216,
      "step": 11960
    },
    {
      "epoch": 7.054834905660377,
      "grad_norm": 3.5418615341186523,
      "learning_rate": 8.177531584629353e-06,
      "loss": 0.3819,
      "num_input_tokens_seen": 7754288,
      "step": 11965
    },
    {
      "epoch": 7.057783018867925,
      "grad_norm": 2.1454973220825195,
      "learning_rate": 8.175544782664335e-06,
      "loss": 0.4214,
      "num_input_tokens_seen": 7757712,
      "step": 11970
    },
    {
      "epoch": 7.060731132075472,
      "grad_norm": 1.5547668933868408,
      "learning_rate": 8.173557139960151e-06,
      "loss": 0.4478,
      "num_input_tokens_seen": 7760560,
      "step": 11975
    },
    {
      "epoch": 7.063679245283019,
      "grad_norm": 2.3159146308898926,
      "learning_rate": 8.17156865704304e-06,
      "loss": 0.3966,
      "num_input_tokens_seen": 7763952,
      "step": 11980
    },
    {
      "epoch": 7.066627358490566,
      "grad_norm": 2.9877121448516846,
      "learning_rate": 8.169579334439453e-06,
      "loss": 0.4169,
      "num_input_tokens_seen": 7767248,
      "step": 11985
    },
    {
      "epoch": 7.069575471698113,
      "grad_norm": 2.966515064239502,
      "learning_rate": 8.16758917267608e-06,
      "loss": 0.366,
      "num_input_tokens_seen": 7770256,
      "step": 11990
    },
    {
      "epoch": 7.07252358490566,
      "grad_norm": 2.451122760772705,
      "learning_rate": 8.165598172279822e-06,
      "loss": 0.4085,
      "num_input_tokens_seen": 7773872,
      "step": 11995
    },
    {
      "epoch": 7.0754716981132075,
      "grad_norm": 1.7509138584136963,
      "learning_rate": 8.163606333777804e-06,
      "loss": 0.5345,
      "num_input_tokens_seen": 7776624,
      "step": 12000
    },
    {
      "epoch": 7.078419811320755,
      "grad_norm": 3.2252087593078613,
      "learning_rate": 8.161613657697374e-06,
      "loss": 0.2125,
      "num_input_tokens_seen": 7779664,
      "step": 12005
    },
    {
      "epoch": 7.081367924528302,
      "grad_norm": 2.299471139907837,
      "learning_rate": 8.159620144566103e-06,
      "loss": 0.4096,
      "num_input_tokens_seen": 7782736,
      "step": 12010
    },
    {
      "epoch": 7.084316037735849,
      "grad_norm": 3.1591343879699707,
      "learning_rate": 8.157625794911782e-06,
      "loss": 0.5271,
      "num_input_tokens_seen": 7785712,
      "step": 12015
    },
    {
      "epoch": 7.087264150943396,
      "grad_norm": 5.675323963165283,
      "learning_rate": 8.155630609262424e-06,
      "loss": 0.4118,
      "num_input_tokens_seen": 7788560,
      "step": 12020
    },
    {
      "epoch": 7.090212264150943,
      "grad_norm": 2.422335386276245,
      "learning_rate": 8.153634588146262e-06,
      "loss": 0.4674,
      "num_input_tokens_seen": 7792400,
      "step": 12025
    },
    {
      "epoch": 7.09316037735849,
      "grad_norm": 2.724100112915039,
      "learning_rate": 8.15163773209175e-06,
      "loss": 0.4452,
      "num_input_tokens_seen": 7795216,
      "step": 12030
    },
    {
      "epoch": 7.096108490566038,
      "grad_norm": 1.7143603563308716,
      "learning_rate": 8.149640041627566e-06,
      "loss": 0.3451,
      "num_input_tokens_seen": 7798768,
      "step": 12035
    },
    {
      "epoch": 7.099056603773585,
      "grad_norm": 1.8391895294189453,
      "learning_rate": 8.147641517282608e-06,
      "loss": 0.4302,
      "num_input_tokens_seen": 7801744,
      "step": 12040
    },
    {
      "epoch": 7.102004716981132,
      "grad_norm": 3.46140193939209,
      "learning_rate": 8.145642159585992e-06,
      "loss": 0.4714,
      "num_input_tokens_seen": 7805392,
      "step": 12045
    },
    {
      "epoch": 7.1049528301886795,
      "grad_norm": 3.2186472415924072,
      "learning_rate": 8.143641969067057e-06,
      "loss": 0.4057,
      "num_input_tokens_seen": 7808080,
      "step": 12050
    },
    {
      "epoch": 7.107900943396227,
      "grad_norm": 3.0942511558532715,
      "learning_rate": 8.141640946255362e-06,
      "loss": 0.3925,
      "num_input_tokens_seen": 7811056,
      "step": 12055
    },
    {
      "epoch": 7.110849056603773,
      "grad_norm": 1.8903888463974,
      "learning_rate": 8.139639091680687e-06,
      "loss": 0.4396,
      "num_input_tokens_seen": 7814224,
      "step": 12060
    },
    {
      "epoch": 7.1137971698113205,
      "grad_norm": 2.0970606803894043,
      "learning_rate": 8.137636405873031e-06,
      "loss": 0.3799,
      "num_input_tokens_seen": 7817552,
      "step": 12065
    },
    {
      "epoch": 7.116745283018868,
      "grad_norm": 3.1271135807037354,
      "learning_rate": 8.135632889362614e-06,
      "loss": 0.4502,
      "num_input_tokens_seen": 7820752,
      "step": 12070
    },
    {
      "epoch": 7.119693396226415,
      "grad_norm": 2.478670597076416,
      "learning_rate": 8.133628542679879e-06,
      "loss": 0.5559,
      "num_input_tokens_seen": 7823280,
      "step": 12075
    },
    {
      "epoch": 7.122641509433962,
      "grad_norm": 2.745588779449463,
      "learning_rate": 8.131623366355478e-06,
      "loss": 0.4455,
      "num_input_tokens_seen": 7825872,
      "step": 12080
    },
    {
      "epoch": 7.12558962264151,
      "grad_norm": 2.5139214992523193,
      "learning_rate": 8.129617360920297e-06,
      "loss": 0.363,
      "num_input_tokens_seen": 7828976,
      "step": 12085
    },
    {
      "epoch": 7.128537735849057,
      "grad_norm": 1.203438401222229,
      "learning_rate": 8.12761052690543e-06,
      "loss": 0.3647,
      "num_input_tokens_seen": 7832016,
      "step": 12090
    },
    {
      "epoch": 7.131485849056604,
      "grad_norm": 0.3967999219894409,
      "learning_rate": 8.125602864842197e-06,
      "loss": 0.4818,
      "num_input_tokens_seen": 7837712,
      "step": 12095
    },
    {
      "epoch": 7.134433962264151,
      "grad_norm": 6.743397235870361,
      "learning_rate": 8.123594375262135e-06,
      "loss": 0.5076,
      "num_input_tokens_seen": 7840976,
      "step": 12100
    },
    {
      "epoch": 7.137382075471698,
      "grad_norm": 2.0678324699401855,
      "learning_rate": 8.121585058697e-06,
      "loss": 0.4425,
      "num_input_tokens_seen": 7844464,
      "step": 12105
    },
    {
      "epoch": 7.140330188679245,
      "grad_norm": 3.1672308444976807,
      "learning_rate": 8.119574915678767e-06,
      "loss": 0.4599,
      "num_input_tokens_seen": 7847216,
      "step": 12110
    },
    {
      "epoch": 7.1432783018867925,
      "grad_norm": 1.40207040309906,
      "learning_rate": 8.117563946739632e-06,
      "loss": 0.4588,
      "num_input_tokens_seen": 7850896,
      "step": 12115
    },
    {
      "epoch": 7.14622641509434,
      "grad_norm": 2.056417226791382,
      "learning_rate": 8.115552152412006e-06,
      "loss": 0.3934,
      "num_input_tokens_seen": 7854320,
      "step": 12120
    },
    {
      "epoch": 7.149174528301887,
      "grad_norm": 1.7066417932510376,
      "learning_rate": 8.11353953322852e-06,
      "loss": 0.3038,
      "num_input_tokens_seen": 7857392,
      "step": 12125
    },
    {
      "epoch": 7.152122641509434,
      "grad_norm": 3.348076343536377,
      "learning_rate": 8.111526089722024e-06,
      "loss": 0.4046,
      "num_input_tokens_seen": 7860240,
      "step": 12130
    },
    {
      "epoch": 7.155070754716981,
      "grad_norm": 1.4375241994857788,
      "learning_rate": 8.109511822425586e-06,
      "loss": 0.3995,
      "num_input_tokens_seen": 7863440,
      "step": 12135
    },
    {
      "epoch": 7.158018867924528,
      "grad_norm": 2.234502077102661,
      "learning_rate": 8.107496731872491e-06,
      "loss": 0.452,
      "num_input_tokens_seen": 7866576,
      "step": 12140
    },
    {
      "epoch": 7.160966981132075,
      "grad_norm": 2.4165091514587402,
      "learning_rate": 8.105480818596243e-06,
      "loss": 0.4832,
      "num_input_tokens_seen": 7869616,
      "step": 12145
    },
    {
      "epoch": 7.163915094339623,
      "grad_norm": 1.608115553855896,
      "learning_rate": 8.103464083130566e-06,
      "loss": 0.3292,
      "num_input_tokens_seen": 7872880,
      "step": 12150
    },
    {
      "epoch": 7.16686320754717,
      "grad_norm": 3.1230015754699707,
      "learning_rate": 8.101446526009397e-06,
      "loss": 0.4842,
      "num_input_tokens_seen": 7876176,
      "step": 12155
    },
    {
      "epoch": 7.169811320754717,
      "grad_norm": 2.069788932800293,
      "learning_rate": 8.099428147766894e-06,
      "loss": 0.5619,
      "num_input_tokens_seen": 7879664,
      "step": 12160
    },
    {
      "epoch": 7.1727594339622645,
      "grad_norm": 1.5261521339416504,
      "learning_rate": 8.097408948937431e-06,
      "loss": 0.509,
      "num_input_tokens_seen": 7883152,
      "step": 12165
    },
    {
      "epoch": 7.175707547169812,
      "grad_norm": 2.1414318084716797,
      "learning_rate": 8.095388930055599e-06,
      "loss": 0.4625,
      "num_input_tokens_seen": 7886480,
      "step": 12170
    },
    {
      "epoch": 7.178655660377358,
      "grad_norm": 1.481910228729248,
      "learning_rate": 8.093368091656209e-06,
      "loss": 0.339,
      "num_input_tokens_seen": 7889744,
      "step": 12175
    },
    {
      "epoch": 7.181603773584905,
      "grad_norm": 3.0812828540802,
      "learning_rate": 8.091346434274284e-06,
      "loss": 0.3869,
      "num_input_tokens_seen": 7892624,
      "step": 12180
    },
    {
      "epoch": 7.184551886792453,
      "grad_norm": 2.4070494174957275,
      "learning_rate": 8.089323958445068e-06,
      "loss": 0.4494,
      "num_input_tokens_seen": 7895504,
      "step": 12185
    },
    {
      "epoch": 7.1875,
      "grad_norm": 2.408076763153076,
      "learning_rate": 8.08730066470402e-06,
      "loss": 0.4999,
      "num_input_tokens_seen": 7898960,
      "step": 12190
    },
    {
      "epoch": 7.190448113207547,
      "grad_norm": 1.3094491958618164,
      "learning_rate": 8.085276553586814e-06,
      "loss": 0.4379,
      "num_input_tokens_seen": 7902128,
      "step": 12195
    },
    {
      "epoch": 7.193396226415095,
      "grad_norm": 2.889855146408081,
      "learning_rate": 8.083251625629345e-06,
      "loss": 0.4205,
      "num_input_tokens_seen": 7905264,
      "step": 12200
    },
    {
      "epoch": 7.196344339622642,
      "grad_norm": 2.232811689376831,
      "learning_rate": 8.08122588136772e-06,
      "loss": 0.5062,
      "num_input_tokens_seen": 7908912,
      "step": 12205
    },
    {
      "epoch": 7.199292452830188,
      "grad_norm": 2.1486170291900635,
      "learning_rate": 8.079199321338262e-06,
      "loss": 0.2765,
      "num_input_tokens_seen": 7911920,
      "step": 12210
    },
    {
      "epoch": 7.2022405660377355,
      "grad_norm": 1.5640469789505005,
      "learning_rate": 8.077171946077516e-06,
      "loss": 0.2635,
      "num_input_tokens_seen": 7914768,
      "step": 12215
    },
    {
      "epoch": 7.205188679245283,
      "grad_norm": 4.163769721984863,
      "learning_rate": 8.075143756122232e-06,
      "loss": 0.3997,
      "num_input_tokens_seen": 7917488,
      "step": 12220
    },
    {
      "epoch": 7.20813679245283,
      "grad_norm": 2.1627330780029297,
      "learning_rate": 8.073114752009388e-06,
      "loss": 0.3103,
      "num_input_tokens_seen": 7920176,
      "step": 12225
    },
    {
      "epoch": 7.211084905660377,
      "grad_norm": 2.5614118576049805,
      "learning_rate": 8.071084934276168e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 7922960,
      "step": 12230
    },
    {
      "epoch": 7.214033018867925,
      "grad_norm": 1.9293445348739624,
      "learning_rate": 8.069054303459976e-06,
      "loss": 0.5033,
      "num_input_tokens_seen": 7925712,
      "step": 12235
    },
    {
      "epoch": 7.216981132075472,
      "grad_norm": 2.6839752197265625,
      "learning_rate": 8.06702286009843e-06,
      "loss": 0.5113,
      "num_input_tokens_seen": 7928400,
      "step": 12240
    },
    {
      "epoch": 7.219929245283019,
      "grad_norm": 3.1859865188598633,
      "learning_rate": 8.064990604729363e-06,
      "loss": 0.3979,
      "num_input_tokens_seen": 7931088,
      "step": 12245
    },
    {
      "epoch": 7.222877358490566,
      "grad_norm": 2.232417106628418,
      "learning_rate": 8.062957537890827e-06,
      "loss": 0.3949,
      "num_input_tokens_seen": 7934480,
      "step": 12250
    },
    {
      "epoch": 7.225825471698113,
      "grad_norm": 2.3280370235443115,
      "learning_rate": 8.060923660121081e-06,
      "loss": 0.4404,
      "num_input_tokens_seen": 7936848,
      "step": 12255
    },
    {
      "epoch": 7.22877358490566,
      "grad_norm": 1.7205384969711304,
      "learning_rate": 8.058888971958603e-06,
      "loss": 0.4101,
      "num_input_tokens_seen": 7940304,
      "step": 12260
    },
    {
      "epoch": 7.2317216981132075,
      "grad_norm": 1.6537336111068726,
      "learning_rate": 8.056853473942085e-06,
      "loss": 0.4341,
      "num_input_tokens_seen": 7943408,
      "step": 12265
    },
    {
      "epoch": 7.234669811320755,
      "grad_norm": 2.9873087406158447,
      "learning_rate": 8.054817166610438e-06,
      "loss": 0.4459,
      "num_input_tokens_seen": 7946160,
      "step": 12270
    },
    {
      "epoch": 7.237617924528302,
      "grad_norm": 3.341125249862671,
      "learning_rate": 8.052780050502781e-06,
      "loss": 0.5256,
      "num_input_tokens_seen": 7949232,
      "step": 12275
    },
    {
      "epoch": 7.240566037735849,
      "grad_norm": 4.269935131072998,
      "learning_rate": 8.050742126158448e-06,
      "loss": 0.3546,
      "num_input_tokens_seen": 7952560,
      "step": 12280
    },
    {
      "epoch": 7.243514150943396,
      "grad_norm": 3.00346040725708,
      "learning_rate": 8.04870339411699e-06,
      "loss": 0.5044,
      "num_input_tokens_seen": 7955632,
      "step": 12285
    },
    {
      "epoch": 7.246462264150943,
      "grad_norm": 2.011465072631836,
      "learning_rate": 8.046663854918166e-06,
      "loss": 0.5547,
      "num_input_tokens_seen": 7958224,
      "step": 12290
    },
    {
      "epoch": 7.24941037735849,
      "grad_norm": 2.582021713256836,
      "learning_rate": 8.044623509101959e-06,
      "loss": 0.3798,
      "num_input_tokens_seen": 7961424,
      "step": 12295
    },
    {
      "epoch": 7.252358490566038,
      "grad_norm": 1.3962986469268799,
      "learning_rate": 8.042582357208557e-06,
      "loss": 0.4542,
      "num_input_tokens_seen": 7965296,
      "step": 12300
    },
    {
      "epoch": 7.255306603773585,
      "grad_norm": 1.421162486076355,
      "learning_rate": 8.04054039977836e-06,
      "loss": 0.4324,
      "num_input_tokens_seen": 7968656,
      "step": 12305
    },
    {
      "epoch": 7.258254716981132,
      "grad_norm": 1.6082948446273804,
      "learning_rate": 8.038497637351992e-06,
      "loss": 0.4217,
      "num_input_tokens_seen": 7971728,
      "step": 12310
    },
    {
      "epoch": 7.2612028301886795,
      "grad_norm": 2.3717334270477295,
      "learning_rate": 8.036454070470276e-06,
      "loss": 0.474,
      "num_input_tokens_seen": 7974480,
      "step": 12315
    },
    {
      "epoch": 7.264150943396227,
      "grad_norm": 2.243896245956421,
      "learning_rate": 8.03440969967426e-06,
      "loss": 0.2662,
      "num_input_tokens_seen": 7980336,
      "step": 12320
    },
    {
      "epoch": 7.267099056603773,
      "grad_norm": 1.9699573516845703,
      "learning_rate": 8.032364525505198e-06,
      "loss": 0.4218,
      "num_input_tokens_seen": 7984784,
      "step": 12325
    },
    {
      "epoch": 7.2700471698113205,
      "grad_norm": 2.6575653553009033,
      "learning_rate": 8.030318548504561e-06,
      "loss": 0.3654,
      "num_input_tokens_seen": 7988176,
      "step": 12330
    },
    {
      "epoch": 7.272995283018868,
      "grad_norm": 3.364875078201294,
      "learning_rate": 8.028271769214026e-06,
      "loss": 0.3446,
      "num_input_tokens_seen": 7990832,
      "step": 12335
    },
    {
      "epoch": 7.275943396226415,
      "grad_norm": 2.072842836380005,
      "learning_rate": 8.02622418817549e-06,
      "loss": 0.4088,
      "num_input_tokens_seen": 7994416,
      "step": 12340
    },
    {
      "epoch": 7.278891509433962,
      "grad_norm": 3.3514645099639893,
      "learning_rate": 8.024175805931056e-06,
      "loss": 0.4715,
      "num_input_tokens_seen": 7996944,
      "step": 12345
    },
    {
      "epoch": 7.28183962264151,
      "grad_norm": 2.416222095489502,
      "learning_rate": 8.022126623023045e-06,
      "loss": 0.3937,
      "num_input_tokens_seen": 8000720,
      "step": 12350
    },
    {
      "epoch": 7.284787735849057,
      "grad_norm": 1.8750841617584229,
      "learning_rate": 8.020076639993987e-06,
      "loss": 0.4814,
      "num_input_tokens_seen": 8004176,
      "step": 12355
    },
    {
      "epoch": 7.287735849056604,
      "grad_norm": 2.365260362625122,
      "learning_rate": 8.018025857386622e-06,
      "loss": 0.4077,
      "num_input_tokens_seen": 8007792,
      "step": 12360
    },
    {
      "epoch": 7.290683962264151,
      "grad_norm": 1.6801178455352783,
      "learning_rate": 8.015974275743905e-06,
      "loss": 0.3925,
      "num_input_tokens_seen": 8010096,
      "step": 12365
    },
    {
      "epoch": 7.293632075471698,
      "grad_norm": 1.7054762840270996,
      "learning_rate": 8.013921895609e-06,
      "loss": 0.3889,
      "num_input_tokens_seen": 8012912,
      "step": 12370
    },
    {
      "epoch": 7.296580188679245,
      "grad_norm": 2.8592233657836914,
      "learning_rate": 8.011868717525283e-06,
      "loss": 0.4484,
      "num_input_tokens_seen": 8015792,
      "step": 12375
    },
    {
      "epoch": 7.2995283018867925,
      "grad_norm": 3.758345603942871,
      "learning_rate": 8.009814742036343e-06,
      "loss": 0.484,
      "num_input_tokens_seen": 8018160,
      "step": 12380
    },
    {
      "epoch": 7.30247641509434,
      "grad_norm": 1.4077646732330322,
      "learning_rate": 8.007759969685979e-06,
      "loss": 0.4461,
      "num_input_tokens_seen": 8020784,
      "step": 12385
    },
    {
      "epoch": 7.305424528301887,
      "grad_norm": 2.8454792499542236,
      "learning_rate": 8.005704401018199e-06,
      "loss": 0.3684,
      "num_input_tokens_seen": 8023760,
      "step": 12390
    },
    {
      "epoch": 7.308372641509434,
      "grad_norm": 5.520720481872559,
      "learning_rate": 8.003648036577226e-06,
      "loss": 0.5465,
      "num_input_tokens_seen": 8026192,
      "step": 12395
    },
    {
      "epoch": 7.311320754716981,
      "grad_norm": 1.4981896877288818,
      "learning_rate": 8.00159087690749e-06,
      "loss": 0.4972,
      "num_input_tokens_seen": 8029200,
      "step": 12400
    },
    {
      "epoch": 7.314268867924528,
      "grad_norm": 2.0746161937713623,
      "learning_rate": 7.999532922553635e-06,
      "loss": 0.4226,
      "num_input_tokens_seen": 8033520,
      "step": 12405
    },
    {
      "epoch": 7.317216981132075,
      "grad_norm": 1.3934060335159302,
      "learning_rate": 7.997474174060508e-06,
      "loss": 0.4365,
      "num_input_tokens_seen": 8037904,
      "step": 12410
    },
    {
      "epoch": 7.320165094339623,
      "grad_norm": 2.3764402866363525,
      "learning_rate": 7.995414631973179e-06,
      "loss": 0.3386,
      "num_input_tokens_seen": 8040464,
      "step": 12415
    },
    {
      "epoch": 7.32311320754717,
      "grad_norm": 1.3803070783615112,
      "learning_rate": 7.993354296836914e-06,
      "loss": 0.4132,
      "num_input_tokens_seen": 8044560,
      "step": 12420
    },
    {
      "epoch": 7.326061320754717,
      "grad_norm": 2.1555306911468506,
      "learning_rate": 7.991293169197198e-06,
      "loss": 0.3288,
      "num_input_tokens_seen": 8047696,
      "step": 12425
    },
    {
      "epoch": 7.3290094339622645,
      "grad_norm": 4.788967609405518,
      "learning_rate": 7.989231249599725e-06,
      "loss": 0.3708,
      "num_input_tokens_seen": 8050704,
      "step": 12430
    },
    {
      "epoch": 7.331957547169811,
      "grad_norm": 1.6810853481292725,
      "learning_rate": 7.987168538590395e-06,
      "loss": 0.3634,
      "num_input_tokens_seen": 8053520,
      "step": 12435
    },
    {
      "epoch": 7.334905660377358,
      "grad_norm": 2.509845495223999,
      "learning_rate": 7.985105036715322e-06,
      "loss": 0.454,
      "num_input_tokens_seen": 8056752,
      "step": 12440
    },
    {
      "epoch": 7.337853773584905,
      "grad_norm": 2.3521809577941895,
      "learning_rate": 7.983040744520823e-06,
      "loss": 0.3447,
      "num_input_tokens_seen": 8059472,
      "step": 12445
    },
    {
      "epoch": 7.340801886792453,
      "grad_norm": 1.484307050704956,
      "learning_rate": 7.980975662553432e-06,
      "loss": 0.3323,
      "num_input_tokens_seen": 8063056,
      "step": 12450
    },
    {
      "epoch": 7.34375,
      "grad_norm": 1.2351588010787964,
      "learning_rate": 7.978909791359888e-06,
      "loss": 0.4036,
      "num_input_tokens_seen": 8065712,
      "step": 12455
    },
    {
      "epoch": 7.346698113207547,
      "grad_norm": 1.3613877296447754,
      "learning_rate": 7.976843131487136e-06,
      "loss": 0.4557,
      "num_input_tokens_seen": 8069360,
      "step": 12460
    },
    {
      "epoch": 7.349646226415095,
      "grad_norm": 1.7863596677780151,
      "learning_rate": 7.974775683482337e-06,
      "loss": 0.4704,
      "num_input_tokens_seen": 8072176,
      "step": 12465
    },
    {
      "epoch": 7.352594339622642,
      "grad_norm": 2.5328752994537354,
      "learning_rate": 7.972707447892855e-06,
      "loss": 0.4658,
      "num_input_tokens_seen": 8075120,
      "step": 12470
    },
    {
      "epoch": 7.355542452830189,
      "grad_norm": 1.4671952724456787,
      "learning_rate": 7.970638425266264e-06,
      "loss": 0.4068,
      "num_input_tokens_seen": 8078352,
      "step": 12475
    },
    {
      "epoch": 7.3584905660377355,
      "grad_norm": 1.831052541732788,
      "learning_rate": 7.968568616150349e-06,
      "loss": 0.3879,
      "num_input_tokens_seen": 8082160,
      "step": 12480
    },
    {
      "epoch": 7.361438679245283,
      "grad_norm": 2.628215789794922,
      "learning_rate": 7.966498021093096e-06,
      "loss": 0.3274,
      "num_input_tokens_seen": 8084592,
      "step": 12485
    },
    {
      "epoch": 7.36438679245283,
      "grad_norm": 1.7561180591583252,
      "learning_rate": 7.96442664064271e-06,
      "loss": 0.4798,
      "num_input_tokens_seen": 8087568,
      "step": 12490
    },
    {
      "epoch": 7.367334905660377,
      "grad_norm": 3.076904773712158,
      "learning_rate": 7.962354475347593e-06,
      "loss": 0.4655,
      "num_input_tokens_seen": 8090736,
      "step": 12495
    },
    {
      "epoch": 7.370283018867925,
      "grad_norm": 2.6705069541931152,
      "learning_rate": 7.960281525756364e-06,
      "loss": 0.3177,
      "num_input_tokens_seen": 8093392,
      "step": 12500
    },
    {
      "epoch": 7.373231132075472,
      "grad_norm": 1.308343529701233,
      "learning_rate": 7.95820779241784e-06,
      "loss": 0.3665,
      "num_input_tokens_seen": 8097296,
      "step": 12505
    },
    {
      "epoch": 7.376179245283019,
      "grad_norm": 1.359108328819275,
      "learning_rate": 7.956133275881055e-06,
      "loss": 0.5105,
      "num_input_tokens_seen": 8101232,
      "step": 12510
    },
    {
      "epoch": 7.379127358490566,
      "grad_norm": 3.7662196159362793,
      "learning_rate": 7.954057976695244e-06,
      "loss": 0.4463,
      "num_input_tokens_seen": 8104112,
      "step": 12515
    },
    {
      "epoch": 7.382075471698113,
      "grad_norm": 1.980501413345337,
      "learning_rate": 7.951981895409854e-06,
      "loss": 0.3958,
      "num_input_tokens_seen": 8107120,
      "step": 12520
    },
    {
      "epoch": 7.38502358490566,
      "grad_norm": 1.544133186340332,
      "learning_rate": 7.949905032574534e-06,
      "loss": 0.4523,
      "num_input_tokens_seen": 8110064,
      "step": 12525
    },
    {
      "epoch": 7.3879716981132075,
      "grad_norm": 1.8327921628952026,
      "learning_rate": 7.947827388739145e-06,
      "loss": 0.4538,
      "num_input_tokens_seen": 8113584,
      "step": 12530
    },
    {
      "epoch": 7.390919811320755,
      "grad_norm": 1.929289698600769,
      "learning_rate": 7.945748964453747e-06,
      "loss": 0.4275,
      "num_input_tokens_seen": 8116912,
      "step": 12535
    },
    {
      "epoch": 7.393867924528302,
      "grad_norm": 3.635775327682495,
      "learning_rate": 7.943669760268618e-06,
      "loss": 0.3736,
      "num_input_tokens_seen": 8120592,
      "step": 12540
    },
    {
      "epoch": 7.396816037735849,
      "grad_norm": 4.127130508422852,
      "learning_rate": 7.941589776734232e-06,
      "loss": 0.4472,
      "num_input_tokens_seen": 8123248,
      "step": 12545
    },
    {
      "epoch": 7.399764150943396,
      "grad_norm": 2.4001150131225586,
      "learning_rate": 7.939509014401277e-06,
      "loss": 0.4514,
      "num_input_tokens_seen": 8125968,
      "step": 12550
    },
    {
      "epoch": 7.402712264150943,
      "grad_norm": 2.562689781188965,
      "learning_rate": 7.93742747382064e-06,
      "loss": 0.4609,
      "num_input_tokens_seen": 8128848,
      "step": 12555
    },
    {
      "epoch": 7.40566037735849,
      "grad_norm": 1.669224739074707,
      "learning_rate": 7.93534515554342e-06,
      "loss": 0.4813,
      "num_input_tokens_seen": 8132272,
      "step": 12560
    },
    {
      "epoch": 7.408608490566038,
      "grad_norm": 1.4592849016189575,
      "learning_rate": 7.933262060120918e-06,
      "loss": 0.4222,
      "num_input_tokens_seen": 8136336,
      "step": 12565
    },
    {
      "epoch": 7.411556603773585,
      "grad_norm": 2.4850730895996094,
      "learning_rate": 7.931178188104646e-06,
      "loss": 0.4608,
      "num_input_tokens_seen": 8139088,
      "step": 12570
    },
    {
      "epoch": 7.414504716981132,
      "grad_norm": 1.570600152015686,
      "learning_rate": 7.929093540046317e-06,
      "loss": 0.4249,
      "num_input_tokens_seen": 8142384,
      "step": 12575
    },
    {
      "epoch": 7.4174528301886795,
      "grad_norm": 2.4422314167022705,
      "learning_rate": 7.927008116497848e-06,
      "loss": 0.3257,
      "num_input_tokens_seen": 8145648,
      "step": 12580
    },
    {
      "epoch": 7.420400943396227,
      "grad_norm": 2.5866360664367676,
      "learning_rate": 7.924921918011366e-06,
      "loss": 0.4683,
      "num_input_tokens_seen": 8148624,
      "step": 12585
    },
    {
      "epoch": 7.423349056603773,
      "grad_norm": 2.9114606380462646,
      "learning_rate": 7.9228349451392e-06,
      "loss": 0.4653,
      "num_input_tokens_seen": 8151568,
      "step": 12590
    },
    {
      "epoch": 7.4262971698113205,
      "grad_norm": 2.020949602127075,
      "learning_rate": 7.920747198433884e-06,
      "loss": 0.4002,
      "num_input_tokens_seen": 8154960,
      "step": 12595
    },
    {
      "epoch": 7.429245283018868,
      "grad_norm": 2.234898090362549,
      "learning_rate": 7.91865867844816e-06,
      "loss": 0.464,
      "num_input_tokens_seen": 8158160,
      "step": 12600
    },
    {
      "epoch": 7.432193396226415,
      "grad_norm": 2.1939659118652344,
      "learning_rate": 7.916569385734976e-06,
      "loss": 0.4743,
      "num_input_tokens_seen": 8160720,
      "step": 12605
    },
    {
      "epoch": 7.435141509433962,
      "grad_norm": 4.47189474105835,
      "learning_rate": 7.914479320847474e-06,
      "loss": 0.4212,
      "num_input_tokens_seen": 8163344,
      "step": 12610
    },
    {
      "epoch": 7.43808962264151,
      "grad_norm": 2.550306797027588,
      "learning_rate": 7.912388484339012e-06,
      "loss": 0.3463,
      "num_input_tokens_seen": 8166512,
      "step": 12615
    },
    {
      "epoch": 7.441037735849057,
      "grad_norm": 1.4921849966049194,
      "learning_rate": 7.910296876763147e-06,
      "loss": 0.3615,
      "num_input_tokens_seen": 8169904,
      "step": 12620
    },
    {
      "epoch": 7.443985849056604,
      "grad_norm": 1.605087399482727,
      "learning_rate": 7.90820449867364e-06,
      "loss": 0.3838,
      "num_input_tokens_seen": 8172496,
      "step": 12625
    },
    {
      "epoch": 7.446933962264151,
      "grad_norm": 3.52890944480896,
      "learning_rate": 7.90611135062446e-06,
      "loss": 0.3801,
      "num_input_tokens_seen": 8175440,
      "step": 12630
    },
    {
      "epoch": 7.449882075471698,
      "grad_norm": 2.5069262981414795,
      "learning_rate": 7.904017433169775e-06,
      "loss": 0.3295,
      "num_input_tokens_seen": 8178736,
      "step": 12635
    },
    {
      "epoch": 7.452830188679245,
      "grad_norm": 1.9273746013641357,
      "learning_rate": 7.901922746863957e-06,
      "loss": 0.475,
      "num_input_tokens_seen": 8182000,
      "step": 12640
    },
    {
      "epoch": 7.4557783018867925,
      "grad_norm": 2.50080943107605,
      "learning_rate": 7.899827292261589e-06,
      "loss": 0.3616,
      "num_input_tokens_seen": 8188944,
      "step": 12645
    },
    {
      "epoch": 7.45872641509434,
      "grad_norm": 2.4265153408050537,
      "learning_rate": 7.897731069917444e-06,
      "loss": 0.4074,
      "num_input_tokens_seen": 8191696,
      "step": 12650
    },
    {
      "epoch": 7.461674528301887,
      "grad_norm": 3.1523895263671875,
      "learning_rate": 7.895634080386512e-06,
      "loss": 0.5514,
      "num_input_tokens_seen": 8195120,
      "step": 12655
    },
    {
      "epoch": 7.464622641509434,
      "grad_norm": 2.018810749053955,
      "learning_rate": 7.893536324223977e-06,
      "loss": 0.4568,
      "num_input_tokens_seen": 8198064,
      "step": 12660
    },
    {
      "epoch": 7.467570754716981,
      "grad_norm": 2.693422555923462,
      "learning_rate": 7.89143780198523e-06,
      "loss": 0.4378,
      "num_input_tokens_seen": 8201840,
      "step": 12665
    },
    {
      "epoch": 7.470518867924528,
      "grad_norm": 1.9207382202148438,
      "learning_rate": 7.889338514225862e-06,
      "loss": 0.34,
      "num_input_tokens_seen": 8204688,
      "step": 12670
    },
    {
      "epoch": 7.473466981132075,
      "grad_norm": 1.533157467842102,
      "learning_rate": 7.887238461501671e-06,
      "loss": 0.3372,
      "num_input_tokens_seen": 8208688,
      "step": 12675
    },
    {
      "epoch": 7.476415094339623,
      "grad_norm": 2.368145704269409,
      "learning_rate": 7.885137644368654e-06,
      "loss": 0.546,
      "num_input_tokens_seen": 8211376,
      "step": 12680
    },
    {
      "epoch": 7.47936320754717,
      "grad_norm": 1.6940363645553589,
      "learning_rate": 7.883036063383012e-06,
      "loss": 0.3774,
      "num_input_tokens_seen": 8214928,
      "step": 12685
    },
    {
      "epoch": 7.482311320754717,
      "grad_norm": 3.0962588787078857,
      "learning_rate": 7.880933719101148e-06,
      "loss": 0.4289,
      "num_input_tokens_seen": 8217456,
      "step": 12690
    },
    {
      "epoch": 7.4852594339622645,
      "grad_norm": 3.097425699234009,
      "learning_rate": 7.878830612079664e-06,
      "loss": 0.4078,
      "num_input_tokens_seen": 8223824,
      "step": 12695
    },
    {
      "epoch": 7.488207547169811,
      "grad_norm": 1.990559458732605,
      "learning_rate": 7.876726742875369e-06,
      "loss": 0.3297,
      "num_input_tokens_seen": 8227440,
      "step": 12700
    },
    {
      "epoch": 7.491155660377358,
      "grad_norm": 4.292287826538086,
      "learning_rate": 7.874622112045269e-06,
      "loss": 0.3321,
      "num_input_tokens_seen": 8231280,
      "step": 12705
    },
    {
      "epoch": 7.494103773584905,
      "grad_norm": 3.015432834625244,
      "learning_rate": 7.872516720146578e-06,
      "loss": 0.3686,
      "num_input_tokens_seen": 8235952,
      "step": 12710
    },
    {
      "epoch": 7.497051886792453,
      "grad_norm": 4.499696731567383,
      "learning_rate": 7.870410567736705e-06,
      "loss": 0.5077,
      "num_input_tokens_seen": 8239344,
      "step": 12715
    },
    {
      "epoch": 7.5,
      "grad_norm": 2.511276960372925,
      "learning_rate": 7.868303655373264e-06,
      "loss": 0.5336,
      "num_input_tokens_seen": 8241840,
      "step": 12720
    },
    {
      "epoch": 7.502948113207547,
      "grad_norm": 2.4848721027374268,
      "learning_rate": 7.866195983614066e-06,
      "loss": 0.4121,
      "num_input_tokens_seen": 8244656,
      "step": 12725
    },
    {
      "epoch": 7.505896226415095,
      "grad_norm": 1.5181188583374023,
      "learning_rate": 7.864087553017133e-06,
      "loss": 0.4674,
      "num_input_tokens_seen": 8248080,
      "step": 12730
    },
    {
      "epoch": 7.508844339622642,
      "grad_norm": 1.6813366413116455,
      "learning_rate": 7.861978364140674e-06,
      "loss": 0.496,
      "num_input_tokens_seen": 8251888,
      "step": 12735
    },
    {
      "epoch": 7.511792452830189,
      "grad_norm": 2.269415855407715,
      "learning_rate": 7.859868417543109e-06,
      "loss": 0.3372,
      "num_input_tokens_seen": 8255824,
      "step": 12740
    },
    {
      "epoch": 7.5147405660377355,
      "grad_norm": 2.046665668487549,
      "learning_rate": 7.857757713783055e-06,
      "loss": 0.4088,
      "num_input_tokens_seen": 8258416,
      "step": 12745
    },
    {
      "epoch": 7.517688679245283,
      "grad_norm": 2.821319103240967,
      "learning_rate": 7.855646253419331e-06,
      "loss": 0.3956,
      "num_input_tokens_seen": 8261136,
      "step": 12750
    },
    {
      "epoch": 7.52063679245283,
      "grad_norm": 1.4070239067077637,
      "learning_rate": 7.853534037010952e-06,
      "loss": 0.3713,
      "num_input_tokens_seen": 8263952,
      "step": 12755
    },
    {
      "epoch": 7.523584905660377,
      "grad_norm": 1.8820338249206543,
      "learning_rate": 7.851421065117142e-06,
      "loss": 0.4236,
      "num_input_tokens_seen": 8266768,
      "step": 12760
    },
    {
      "epoch": 7.526533018867925,
      "grad_norm": 1.8540452718734741,
      "learning_rate": 7.849307338297314e-06,
      "loss": 0.3891,
      "num_input_tokens_seen": 8270704,
      "step": 12765
    },
    {
      "epoch": 7.529481132075472,
      "grad_norm": 2.2364583015441895,
      "learning_rate": 7.847192857111087e-06,
      "loss": 0.3254,
      "num_input_tokens_seen": 8273296,
      "step": 12770
    },
    {
      "epoch": 7.532429245283019,
      "grad_norm": 2.971019744873047,
      "learning_rate": 7.845077622118282e-06,
      "loss": 0.6156,
      "num_input_tokens_seen": 8276240,
      "step": 12775
    },
    {
      "epoch": 7.535377358490566,
      "grad_norm": 2.238123893737793,
      "learning_rate": 7.842961633878916e-06,
      "loss": 0.37,
      "num_input_tokens_seen": 8280272,
      "step": 12780
    },
    {
      "epoch": 7.538325471698113,
      "grad_norm": 1.8315377235412598,
      "learning_rate": 7.840844892953204e-06,
      "loss": 0.5005,
      "num_input_tokens_seen": 8285008,
      "step": 12785
    },
    {
      "epoch": 7.54127358490566,
      "grad_norm": 3.6664979457855225,
      "learning_rate": 7.838727399901562e-06,
      "loss": 0.3399,
      "num_input_tokens_seen": 8288048,
      "step": 12790
    },
    {
      "epoch": 7.5442216981132075,
      "grad_norm": 2.8654849529266357,
      "learning_rate": 7.836609155284607e-06,
      "loss": 0.4184,
      "num_input_tokens_seen": 8290832,
      "step": 12795
    },
    {
      "epoch": 7.547169811320755,
      "grad_norm": 2.197002410888672,
      "learning_rate": 7.834490159663154e-06,
      "loss": 0.4064,
      "num_input_tokens_seen": 8294160,
      "step": 12800
    },
    {
      "epoch": 7.550117924528302,
      "grad_norm": 2.1981256008148193,
      "learning_rate": 7.832370413598215e-06,
      "loss": 0.4407,
      "num_input_tokens_seen": 8298128,
      "step": 12805
    },
    {
      "epoch": 7.553066037735849,
      "grad_norm": 3.358893632888794,
      "learning_rate": 7.830249917651003e-06,
      "loss": 0.4747,
      "num_input_tokens_seen": 8301232,
      "step": 12810
    },
    {
      "epoch": 7.556014150943396,
      "grad_norm": 4.121964454650879,
      "learning_rate": 7.828128672382926e-06,
      "loss": 0.4804,
      "num_input_tokens_seen": 8304048,
      "step": 12815
    },
    {
      "epoch": 7.558962264150943,
      "grad_norm": 1.7607474327087402,
      "learning_rate": 7.826006678355596e-06,
      "loss": 0.2859,
      "num_input_tokens_seen": 8306768,
      "step": 12820
    },
    {
      "epoch": 7.56191037735849,
      "grad_norm": 3.2271945476531982,
      "learning_rate": 7.823883936130817e-06,
      "loss": 0.4703,
      "num_input_tokens_seen": 8310128,
      "step": 12825
    },
    {
      "epoch": 7.564858490566038,
      "grad_norm": 2.7114875316619873,
      "learning_rate": 7.821760446270597e-06,
      "loss": 0.4295,
      "num_input_tokens_seen": 8312688,
      "step": 12830
    },
    {
      "epoch": 7.567806603773585,
      "grad_norm": 1.9505242109298706,
      "learning_rate": 7.819636209337136e-06,
      "loss": 0.5627,
      "num_input_tokens_seen": 8316400,
      "step": 12835
    },
    {
      "epoch": 7.570754716981132,
      "grad_norm": 3.3648715019226074,
      "learning_rate": 7.817511225892838e-06,
      "loss": 0.3545,
      "num_input_tokens_seen": 8319632,
      "step": 12840
    },
    {
      "epoch": 7.5737028301886795,
      "grad_norm": 2.6900668144226074,
      "learning_rate": 7.8153854965003e-06,
      "loss": 0.4576,
      "num_input_tokens_seen": 8322832,
      "step": 12845
    },
    {
      "epoch": 7.576650943396227,
      "grad_norm": 3.5517659187316895,
      "learning_rate": 7.813259021722319e-06,
      "loss": 0.4715,
      "num_input_tokens_seen": 8325776,
      "step": 12850
    },
    {
      "epoch": 7.579599056603773,
      "grad_norm": 1.6185506582260132,
      "learning_rate": 7.811131802121885e-06,
      "loss": 0.5162,
      "num_input_tokens_seen": 8328976,
      "step": 12855
    },
    {
      "epoch": 7.5825471698113205,
      "grad_norm": 1.3505847454071045,
      "learning_rate": 7.809003838262193e-06,
      "loss": 0.3908,
      "num_input_tokens_seen": 8332304,
      "step": 12860
    },
    {
      "epoch": 7.585495283018868,
      "grad_norm": 2.5270612239837646,
      "learning_rate": 7.806875130706628e-06,
      "loss": 0.3638,
      "num_input_tokens_seen": 8334992,
      "step": 12865
    },
    {
      "epoch": 7.588443396226415,
      "grad_norm": 2.4500558376312256,
      "learning_rate": 7.804745680018775e-06,
      "loss": 0.4028,
      "num_input_tokens_seen": 8338640,
      "step": 12870
    },
    {
      "epoch": 7.591391509433962,
      "grad_norm": 1.613358736038208,
      "learning_rate": 7.802615486762418e-06,
      "loss": 0.3868,
      "num_input_tokens_seen": 8341232,
      "step": 12875
    },
    {
      "epoch": 7.59433962264151,
      "grad_norm": 3.0669944286346436,
      "learning_rate": 7.800484551501528e-06,
      "loss": 0.4441,
      "num_input_tokens_seen": 8344240,
      "step": 12880
    },
    {
      "epoch": 7.597287735849057,
      "grad_norm": 3.5213003158569336,
      "learning_rate": 7.798352874800285e-06,
      "loss": 0.3633,
      "num_input_tokens_seen": 8346992,
      "step": 12885
    },
    {
      "epoch": 7.600235849056604,
      "grad_norm": 2.40199613571167,
      "learning_rate": 7.79622045722306e-06,
      "loss": 0.3467,
      "num_input_tokens_seen": 8350384,
      "step": 12890
    },
    {
      "epoch": 7.603183962264151,
      "grad_norm": 2.691133975982666,
      "learning_rate": 7.794087299334416e-06,
      "loss": 0.3159,
      "num_input_tokens_seen": 8353104,
      "step": 12895
    },
    {
      "epoch": 7.606132075471698,
      "grad_norm": 1.869766354560852,
      "learning_rate": 7.79195340169912e-06,
      "loss": 0.5152,
      "num_input_tokens_seen": 8356496,
      "step": 12900
    },
    {
      "epoch": 7.609080188679245,
      "grad_norm": 1.246387243270874,
      "learning_rate": 7.789818764882127e-06,
      "loss": 0.3081,
      "num_input_tokens_seen": 8360848,
      "step": 12905
    },
    {
      "epoch": 7.6120283018867925,
      "grad_norm": 1.527164101600647,
      "learning_rate": 7.78768338944859e-06,
      "loss": 0.4141,
      "num_input_tokens_seen": 8363792,
      "step": 12910
    },
    {
      "epoch": 7.61497641509434,
      "grad_norm": 3.421034097671509,
      "learning_rate": 7.785547275963865e-06,
      "loss": 0.4684,
      "num_input_tokens_seen": 8367856,
      "step": 12915
    },
    {
      "epoch": 7.617924528301887,
      "grad_norm": 2.489356279373169,
      "learning_rate": 7.783410424993492e-06,
      "loss": 0.4154,
      "num_input_tokens_seen": 8370704,
      "step": 12920
    },
    {
      "epoch": 7.620872641509434,
      "grad_norm": 2.670596122741699,
      "learning_rate": 7.781272837103213e-06,
      "loss": 0.5075,
      "num_input_tokens_seen": 8373392,
      "step": 12925
    },
    {
      "epoch": 7.623820754716981,
      "grad_norm": 2.120415210723877,
      "learning_rate": 7.779134512858964e-06,
      "loss": 0.3793,
      "num_input_tokens_seen": 8376592,
      "step": 12930
    },
    {
      "epoch": 7.626768867924528,
      "grad_norm": 1.4733853340148926,
      "learning_rate": 7.776995452826876e-06,
      "loss": 0.3727,
      "num_input_tokens_seen": 8381040,
      "step": 12935
    },
    {
      "epoch": 7.629716981132075,
      "grad_norm": 2.355581045150757,
      "learning_rate": 7.774855657573274e-06,
      "loss": 0.4578,
      "num_input_tokens_seen": 8384176,
      "step": 12940
    },
    {
      "epoch": 7.632665094339623,
      "grad_norm": 1.365619421005249,
      "learning_rate": 7.772715127664676e-06,
      "loss": 0.4152,
      "num_input_tokens_seen": 8388016,
      "step": 12945
    },
    {
      "epoch": 7.63561320754717,
      "grad_norm": 2.699767589569092,
      "learning_rate": 7.7705738636678e-06,
      "loss": 0.6102,
      "num_input_tokens_seen": 8390768,
      "step": 12950
    },
    {
      "epoch": 7.638561320754717,
      "grad_norm": 2.1799399852752686,
      "learning_rate": 7.768431866149552e-06,
      "loss": 0.4705,
      "num_input_tokens_seen": 8394192,
      "step": 12955
    },
    {
      "epoch": 7.6415094339622645,
      "grad_norm": 2.1256563663482666,
      "learning_rate": 7.766289135677035e-06,
      "loss": 0.5063,
      "num_input_tokens_seen": 8398192,
      "step": 12960
    },
    {
      "epoch": 7.644457547169811,
      "grad_norm": 3.273649215698242,
      "learning_rate": 7.764145672817549e-06,
      "loss": 0.4645,
      "num_input_tokens_seen": 8401232,
      "step": 12965
    },
    {
      "epoch": 7.647405660377358,
      "grad_norm": 2.1191494464874268,
      "learning_rate": 7.762001478138583e-06,
      "loss": 0.5305,
      "num_input_tokens_seen": 8405232,
      "step": 12970
    },
    {
      "epoch": 7.650353773584905,
      "grad_norm": 1.678210973739624,
      "learning_rate": 7.759856552207822e-06,
      "loss": 0.3705,
      "num_input_tokens_seen": 8408400,
      "step": 12975
    },
    {
      "epoch": 7.653301886792453,
      "grad_norm": 1.8566797971725464,
      "learning_rate": 7.757710895593144e-06,
      "loss": 0.4158,
      "num_input_tokens_seen": 8411632,
      "step": 12980
    },
    {
      "epoch": 7.65625,
      "grad_norm": 1.6825487613677979,
      "learning_rate": 7.755564508862623e-06,
      "loss": 0.3722,
      "num_input_tokens_seen": 8414320,
      "step": 12985
    },
    {
      "epoch": 7.659198113207547,
      "grad_norm": 1.4958350658416748,
      "learning_rate": 7.753417392584522e-06,
      "loss": 0.4575,
      "num_input_tokens_seen": 8418000,
      "step": 12990
    },
    {
      "epoch": 7.662146226415095,
      "grad_norm": 2.7413485050201416,
      "learning_rate": 7.751269547327298e-06,
      "loss": 0.4784,
      "num_input_tokens_seen": 8421200,
      "step": 12995
    },
    {
      "epoch": 7.665094339622642,
      "grad_norm": 3.1365866661071777,
      "learning_rate": 7.749120973659606e-06,
      "loss": 0.3631,
      "num_input_tokens_seen": 8424240,
      "step": 13000
    },
    {
      "epoch": 7.668042452830189,
      "grad_norm": 2.382375717163086,
      "learning_rate": 7.746971672150286e-06,
      "loss": 0.5604,
      "num_input_tokens_seen": 8428208,
      "step": 13005
    },
    {
      "epoch": 7.6709905660377355,
      "grad_norm": 2.0644516944885254,
      "learning_rate": 7.74482164336838e-06,
      "loss": 0.4732,
      "num_input_tokens_seen": 8430608,
      "step": 13010
    },
    {
      "epoch": 7.673938679245283,
      "grad_norm": 2.514810800552368,
      "learning_rate": 7.742670887883111e-06,
      "loss": 0.4703,
      "num_input_tokens_seen": 8433264,
      "step": 13015
    },
    {
      "epoch": 7.67688679245283,
      "grad_norm": 2.008574962615967,
      "learning_rate": 7.740519406263905e-06,
      "loss": 0.4423,
      "num_input_tokens_seen": 8436016,
      "step": 13020
    },
    {
      "epoch": 7.679834905660377,
      "grad_norm": 2.5653746128082275,
      "learning_rate": 7.738367199080376e-06,
      "loss": 0.3783,
      "num_input_tokens_seen": 8439248,
      "step": 13025
    },
    {
      "epoch": 7.682783018867925,
      "grad_norm": 3.3298492431640625,
      "learning_rate": 7.73621426690233e-06,
      "loss": 0.5467,
      "num_input_tokens_seen": 8442224,
      "step": 13030
    },
    {
      "epoch": 7.685731132075472,
      "grad_norm": 2.5645480155944824,
      "learning_rate": 7.734060610299764e-06,
      "loss": 0.4357,
      "num_input_tokens_seen": 8445264,
      "step": 13035
    },
    {
      "epoch": 7.688679245283019,
      "grad_norm": 2.7726404666900635,
      "learning_rate": 7.731906229842869e-06,
      "loss": 0.3846,
      "num_input_tokens_seen": 8447568,
      "step": 13040
    },
    {
      "epoch": 7.691627358490566,
      "grad_norm": 1.6250959634780884,
      "learning_rate": 7.729751126102023e-06,
      "loss": 0.3704,
      "num_input_tokens_seen": 8450768,
      "step": 13045
    },
    {
      "epoch": 7.694575471698113,
      "grad_norm": 1.0875568389892578,
      "learning_rate": 7.727595299647805e-06,
      "loss": 0.3896,
      "num_input_tokens_seen": 8454416,
      "step": 13050
    },
    {
      "epoch": 7.69752358490566,
      "grad_norm": 1.6727712154388428,
      "learning_rate": 7.725438751050973e-06,
      "loss": 0.3535,
      "num_input_tokens_seen": 8457584,
      "step": 13055
    },
    {
      "epoch": 7.7004716981132075,
      "grad_norm": 4.630952835083008,
      "learning_rate": 7.723281480882489e-06,
      "loss": 0.3884,
      "num_input_tokens_seen": 8459824,
      "step": 13060
    },
    {
      "epoch": 7.703419811320755,
      "grad_norm": 1.7784769535064697,
      "learning_rate": 7.721123489713494e-06,
      "loss": 0.4161,
      "num_input_tokens_seen": 8462352,
      "step": 13065
    },
    {
      "epoch": 7.706367924528302,
      "grad_norm": 2.4034059047698975,
      "learning_rate": 7.718964778115328e-06,
      "loss": 0.469,
      "num_input_tokens_seen": 8465648,
      "step": 13070
    },
    {
      "epoch": 7.709316037735849,
      "grad_norm": 2.1052565574645996,
      "learning_rate": 7.716805346659519e-06,
      "loss": 0.3418,
      "num_input_tokens_seen": 8468336,
      "step": 13075
    },
    {
      "epoch": 7.712264150943396,
      "grad_norm": 2.844937324523926,
      "learning_rate": 7.714645195917788e-06,
      "loss": 0.3159,
      "num_input_tokens_seen": 8470640,
      "step": 13080
    },
    {
      "epoch": 7.715212264150943,
      "grad_norm": 3.4066762924194336,
      "learning_rate": 7.712484326462038e-06,
      "loss": 0.4131,
      "num_input_tokens_seen": 8473648,
      "step": 13085
    },
    {
      "epoch": 7.71816037735849,
      "grad_norm": 4.817399501800537,
      "learning_rate": 7.710322738864375e-06,
      "loss": 0.5728,
      "num_input_tokens_seen": 8477328,
      "step": 13090
    },
    {
      "epoch": 7.721108490566038,
      "grad_norm": 1.8981860876083374,
      "learning_rate": 7.708160433697085e-06,
      "loss": 0.392,
      "num_input_tokens_seen": 8479888,
      "step": 13095
    },
    {
      "epoch": 7.724056603773585,
      "grad_norm": 2.1933913230895996,
      "learning_rate": 7.705997411532649e-06,
      "loss": 0.4069,
      "num_input_tokens_seen": 8482736,
      "step": 13100
    },
    {
      "epoch": 7.727004716981132,
      "grad_norm": 2.6387219429016113,
      "learning_rate": 7.703833672943735e-06,
      "loss": 0.4693,
      "num_input_tokens_seen": 8486192,
      "step": 13105
    },
    {
      "epoch": 7.7299528301886795,
      "grad_norm": 2.6244845390319824,
      "learning_rate": 7.701669218503206e-06,
      "loss": 0.3088,
      "num_input_tokens_seen": 8489328,
      "step": 13110
    },
    {
      "epoch": 7.732900943396227,
      "grad_norm": 3.123095989227295,
      "learning_rate": 7.699504048784106e-06,
      "loss": 0.4162,
      "num_input_tokens_seen": 8491856,
      "step": 13115
    },
    {
      "epoch": 7.735849056603773,
      "grad_norm": 4.861481666564941,
      "learning_rate": 7.697338164359675e-06,
      "loss": 0.3958,
      "num_input_tokens_seen": 8494352,
      "step": 13120
    },
    {
      "epoch": 7.7387971698113205,
      "grad_norm": 2.337632656097412,
      "learning_rate": 7.69517156580334e-06,
      "loss": 0.4448,
      "num_input_tokens_seen": 8496624,
      "step": 13125
    },
    {
      "epoch": 7.741745283018868,
      "grad_norm": 2.8294284343719482,
      "learning_rate": 7.693004253688716e-06,
      "loss": 0.3693,
      "num_input_tokens_seen": 8499088,
      "step": 13130
    },
    {
      "epoch": 7.744693396226415,
      "grad_norm": 2.383037805557251,
      "learning_rate": 7.690836228589613e-06,
      "loss": 0.6573,
      "num_input_tokens_seen": 8505904,
      "step": 13135
    },
    {
      "epoch": 7.747641509433962,
      "grad_norm": 2.898552894592285,
      "learning_rate": 7.688667491080019e-06,
      "loss": 0.285,
      "num_input_tokens_seen": 8508464,
      "step": 13140
    },
    {
      "epoch": 7.75058962264151,
      "grad_norm": 2.2031452655792236,
      "learning_rate": 7.686498041734121e-06,
      "loss": 0.3846,
      "num_input_tokens_seen": 8510864,
      "step": 13145
    },
    {
      "epoch": 7.753537735849057,
      "grad_norm": 1.3785383701324463,
      "learning_rate": 7.684327881126285e-06,
      "loss": 0.3635,
      "num_input_tokens_seen": 8514576,
      "step": 13150
    },
    {
      "epoch": 7.756485849056604,
      "grad_norm": 1.7527812719345093,
      "learning_rate": 7.682157009831078e-06,
      "loss": 0.5596,
      "num_input_tokens_seen": 8517808,
      "step": 13155
    },
    {
      "epoch": 7.759433962264151,
      "grad_norm": 2.66363787651062,
      "learning_rate": 7.67998542842324e-06,
      "loss": 0.3903,
      "num_input_tokens_seen": 8521104,
      "step": 13160
    },
    {
      "epoch": 7.762382075471698,
      "grad_norm": 1.732384204864502,
      "learning_rate": 7.677813137477711e-06,
      "loss": 0.4033,
      "num_input_tokens_seen": 8525488,
      "step": 13165
    },
    {
      "epoch": 7.765330188679245,
      "grad_norm": 3.9355521202087402,
      "learning_rate": 7.675640137569614e-06,
      "loss": 0.5448,
      "num_input_tokens_seen": 8528656,
      "step": 13170
    },
    {
      "epoch": 7.7682783018867925,
      "grad_norm": 3.288640022277832,
      "learning_rate": 7.673466429274257e-06,
      "loss": 0.3807,
      "num_input_tokens_seen": 8531568,
      "step": 13175
    },
    {
      "epoch": 7.77122641509434,
      "grad_norm": 1.9667081832885742,
      "learning_rate": 7.671292013167143e-06,
      "loss": 0.4057,
      "num_input_tokens_seen": 8536048,
      "step": 13180
    },
    {
      "epoch": 7.774174528301887,
      "grad_norm": 3.6332590579986572,
      "learning_rate": 7.669116889823955e-06,
      "loss": 0.5118,
      "num_input_tokens_seen": 8539216,
      "step": 13185
    },
    {
      "epoch": 7.777122641509434,
      "grad_norm": 2.237518548965454,
      "learning_rate": 7.666941059820567e-06,
      "loss": 0.3247,
      "num_input_tokens_seen": 8542512,
      "step": 13190
    },
    {
      "epoch": 7.780070754716981,
      "grad_norm": 1.4073950052261353,
      "learning_rate": 7.66476452373304e-06,
      "loss": 0.4065,
      "num_input_tokens_seen": 8545616,
      "step": 13195
    },
    {
      "epoch": 7.783018867924528,
      "grad_norm": 1.6775416135787964,
      "learning_rate": 7.66258728213762e-06,
      "loss": 0.3713,
      "num_input_tokens_seen": 8547856,
      "step": 13200
    },
    {
      "epoch": 7.785966981132075,
      "grad_norm": 2.031903028488159,
      "learning_rate": 7.66040933561074e-06,
      "loss": 0.3654,
      "num_input_tokens_seen": 8551280,
      "step": 13205
    },
    {
      "epoch": 7.788915094339623,
      "grad_norm": 2.69401216506958,
      "learning_rate": 7.658230684729027e-06,
      "loss": 0.4468,
      "num_input_tokens_seen": 8553584,
      "step": 13210
    },
    {
      "epoch": 7.79186320754717,
      "grad_norm": 2.3409533500671387,
      "learning_rate": 7.656051330069282e-06,
      "loss": 0.3702,
      "num_input_tokens_seen": 8556720,
      "step": 13215
    },
    {
      "epoch": 7.794811320754717,
      "grad_norm": 2.5068938732147217,
      "learning_rate": 7.6538712722085e-06,
      "loss": 0.4082,
      "num_input_tokens_seen": 8559376,
      "step": 13220
    },
    {
      "epoch": 7.7977594339622645,
      "grad_norm": 2.311683177947998,
      "learning_rate": 7.651690511723862e-06,
      "loss": 0.4313,
      "num_input_tokens_seen": 8562288,
      "step": 13225
    },
    {
      "epoch": 7.800707547169811,
      "grad_norm": 1.2867538928985596,
      "learning_rate": 7.64950904919273e-06,
      "loss": 0.483,
      "num_input_tokens_seen": 8565712,
      "step": 13230
    },
    {
      "epoch": 7.803655660377358,
      "grad_norm": 1.8231614828109741,
      "learning_rate": 7.647326885192662e-06,
      "loss": 0.3798,
      "num_input_tokens_seen": 8568720,
      "step": 13235
    },
    {
      "epoch": 7.806603773584905,
      "grad_norm": 2.3899593353271484,
      "learning_rate": 7.645144020301392e-06,
      "loss": 0.3106,
      "num_input_tokens_seen": 8571952,
      "step": 13240
    },
    {
      "epoch": 7.809551886792453,
      "grad_norm": 3.299260139465332,
      "learning_rate": 7.64296045509684e-06,
      "loss": 0.4952,
      "num_input_tokens_seen": 8575504,
      "step": 13245
    },
    {
      "epoch": 7.8125,
      "grad_norm": 2.8804168701171875,
      "learning_rate": 7.64077619015712e-06,
      "loss": 0.3108,
      "num_input_tokens_seen": 8578640,
      "step": 13250
    },
    {
      "epoch": 7.815448113207547,
      "grad_norm": 2.802535057067871,
      "learning_rate": 7.638591226060519e-06,
      "loss": 0.4449,
      "num_input_tokens_seen": 8581712,
      "step": 13255
    },
    {
      "epoch": 7.818396226415095,
      "grad_norm": 1.7208733558654785,
      "learning_rate": 7.636405563385522e-06,
      "loss": 0.4121,
      "num_input_tokens_seen": 8585200,
      "step": 13260
    },
    {
      "epoch": 7.821344339622642,
      "grad_norm": 2.3679420948028564,
      "learning_rate": 7.634219202710789e-06,
      "loss": 0.4351,
      "num_input_tokens_seen": 8589136,
      "step": 13265
    },
    {
      "epoch": 7.824292452830189,
      "grad_norm": 1.9945547580718994,
      "learning_rate": 7.632032144615168e-06,
      "loss": 0.4801,
      "num_input_tokens_seen": 8592176,
      "step": 13270
    },
    {
      "epoch": 7.8272405660377355,
      "grad_norm": 1.8825629949569702,
      "learning_rate": 7.629844389677695e-06,
      "loss": 0.587,
      "num_input_tokens_seen": 8595440,
      "step": 13275
    },
    {
      "epoch": 7.830188679245283,
      "grad_norm": 1.3767181634902954,
      "learning_rate": 7.627655938477586e-06,
      "loss": 0.3658,
      "num_input_tokens_seen": 8598800,
      "step": 13280
    },
    {
      "epoch": 7.83313679245283,
      "grad_norm": 2.5362038612365723,
      "learning_rate": 7.6254667915942415e-06,
      "loss": 0.4386,
      "num_input_tokens_seen": 8602960,
      "step": 13285
    },
    {
      "epoch": 7.836084905660377,
      "grad_norm": 2.7800021171569824,
      "learning_rate": 7.62327694960725e-06,
      "loss": 0.5015,
      "num_input_tokens_seen": 8606480,
      "step": 13290
    },
    {
      "epoch": 7.839033018867925,
      "grad_norm": 1.898898720741272,
      "learning_rate": 7.621086413096379e-06,
      "loss": 0.4288,
      "num_input_tokens_seen": 8609392,
      "step": 13295
    },
    {
      "epoch": 7.841981132075472,
      "grad_norm": 2.4820542335510254,
      "learning_rate": 7.618895182641584e-06,
      "loss": 0.4514,
      "num_input_tokens_seen": 8612112,
      "step": 13300
    },
    {
      "epoch": 7.844929245283019,
      "grad_norm": 1.5255407094955444,
      "learning_rate": 7.6167032588230035e-06,
      "loss": 0.2797,
      "num_input_tokens_seen": 8615760,
      "step": 13305
    },
    {
      "epoch": 7.847877358490566,
      "grad_norm": 2.092479705810547,
      "learning_rate": 7.614510642220958e-06,
      "loss": 0.4234,
      "num_input_tokens_seen": 8620432,
      "step": 13310
    },
    {
      "epoch": 7.850825471698113,
      "grad_norm": 2.20448899269104,
      "learning_rate": 7.612317333415951e-06,
      "loss": 0.3601,
      "num_input_tokens_seen": 8623792,
      "step": 13315
    },
    {
      "epoch": 7.85377358490566,
      "grad_norm": 2.4925730228424072,
      "learning_rate": 7.610123332988673e-06,
      "loss": 0.3912,
      "num_input_tokens_seen": 8626896,
      "step": 13320
    },
    {
      "epoch": 7.8567216981132075,
      "grad_norm": 2.289379596710205,
      "learning_rate": 7.607928641519992e-06,
      "loss": 0.4199,
      "num_input_tokens_seen": 8629872,
      "step": 13325
    },
    {
      "epoch": 7.859669811320755,
      "grad_norm": 2.6080520153045654,
      "learning_rate": 7.605733259590964e-06,
      "loss": 0.3529,
      "num_input_tokens_seen": 8633840,
      "step": 13330
    },
    {
      "epoch": 7.862617924528302,
      "grad_norm": 3.086143732070923,
      "learning_rate": 7.603537187782826e-06,
      "loss": 0.3691,
      "num_input_tokens_seen": 8636400,
      "step": 13335
    },
    {
      "epoch": 7.865566037735849,
      "grad_norm": 1.820942997932434,
      "learning_rate": 7.601340426676996e-06,
      "loss": 0.4282,
      "num_input_tokens_seen": 8639216,
      "step": 13340
    },
    {
      "epoch": 7.868514150943396,
      "grad_norm": 2.011366844177246,
      "learning_rate": 7.599142976855077e-06,
      "loss": 0.3954,
      "num_input_tokens_seen": 8642512,
      "step": 13345
    },
    {
      "epoch": 7.871462264150943,
      "grad_norm": 4.917770862579346,
      "learning_rate": 7.596944838898854e-06,
      "loss": 0.4398,
      "num_input_tokens_seen": 8646032,
      "step": 13350
    },
    {
      "epoch": 7.87441037735849,
      "grad_norm": 1.186747670173645,
      "learning_rate": 7.594746013390293e-06,
      "loss": 0.3151,
      "num_input_tokens_seen": 8649328,
      "step": 13355
    },
    {
      "epoch": 7.877358490566038,
      "grad_norm": 1.9084423780441284,
      "learning_rate": 7.59254650091154e-06,
      "loss": 0.3877,
      "num_input_tokens_seen": 8653456,
      "step": 13360
    },
    {
      "epoch": 7.880306603773585,
      "grad_norm": 2.1774566173553467,
      "learning_rate": 7.59034630204493e-06,
      "loss": 0.3974,
      "num_input_tokens_seen": 8657072,
      "step": 13365
    },
    {
      "epoch": 7.883254716981132,
      "grad_norm": 2.4961490631103516,
      "learning_rate": 7.588145417372972e-06,
      "loss": 0.3544,
      "num_input_tokens_seen": 8659632,
      "step": 13370
    },
    {
      "epoch": 7.8862028301886795,
      "grad_norm": 5.203574180603027,
      "learning_rate": 7.585943847478361e-06,
      "loss": 0.3724,
      "num_input_tokens_seen": 8662832,
      "step": 13375
    },
    {
      "epoch": 7.889150943396227,
      "grad_norm": 3.4522061347961426,
      "learning_rate": 7.583741592943971e-06,
      "loss": 0.3694,
      "num_input_tokens_seen": 8667152,
      "step": 13380
    },
    {
      "epoch": 7.892099056603773,
      "grad_norm": 2.76965594291687,
      "learning_rate": 7.581538654352859e-06,
      "loss": 0.432,
      "num_input_tokens_seen": 8672016,
      "step": 13385
    },
    {
      "epoch": 7.8950471698113205,
      "grad_norm": 2.9468908309936523,
      "learning_rate": 7.579335032288262e-06,
      "loss": 0.5895,
      "num_input_tokens_seen": 8674992,
      "step": 13390
    },
    {
      "epoch": 7.897995283018868,
      "grad_norm": 6.9595441818237305,
      "learning_rate": 7.577130727333598e-06,
      "loss": 0.3974,
      "num_input_tokens_seen": 8678512,
      "step": 13395
    },
    {
      "epoch": 7.900943396226415,
      "grad_norm": 2.647261142730713,
      "learning_rate": 7.5749257400724695e-06,
      "loss": 0.3451,
      "num_input_tokens_seen": 8682512,
      "step": 13400
    },
    {
      "epoch": 7.903891509433962,
      "grad_norm": 2.0578885078430176,
      "learning_rate": 7.572720071088653e-06,
      "loss": 0.5362,
      "num_input_tokens_seen": 8685648,
      "step": 13405
    },
    {
      "epoch": 7.90683962264151,
      "grad_norm": 1.4284003973007202,
      "learning_rate": 7.570513720966108e-06,
      "loss": 0.3544,
      "num_input_tokens_seen": 8688880,
      "step": 13410
    },
    {
      "epoch": 7.909787735849057,
      "grad_norm": 2.39734148979187,
      "learning_rate": 7.56830669028898e-06,
      "loss": 0.5079,
      "num_input_tokens_seen": 8691888,
      "step": 13415
    },
    {
      "epoch": 7.912735849056604,
      "grad_norm": 2.218555450439453,
      "learning_rate": 7.566098979641588e-06,
      "loss": 0.4372,
      "num_input_tokens_seen": 8695792,
      "step": 13420
    },
    {
      "epoch": 7.915683962264151,
      "grad_norm": 2.1806070804595947,
      "learning_rate": 7.563890589608427e-06,
      "loss": 0.483,
      "num_input_tokens_seen": 8699536,
      "step": 13425
    },
    {
      "epoch": 7.918632075471698,
      "grad_norm": 2.2735090255737305,
      "learning_rate": 7.561681520774187e-06,
      "loss": 0.3369,
      "num_input_tokens_seen": 8703568,
      "step": 13430
    },
    {
      "epoch": 7.921580188679245,
      "grad_norm": 5.623096942901611,
      "learning_rate": 7.559471773723721e-06,
      "loss": 0.5029,
      "num_input_tokens_seen": 8706032,
      "step": 13435
    },
    {
      "epoch": 7.9245283018867925,
      "grad_norm": 1.5507570505142212,
      "learning_rate": 7.557261349042073e-06,
      "loss": 0.4178,
      "num_input_tokens_seen": 8709264,
      "step": 13440
    },
    {
      "epoch": 7.92747641509434,
      "grad_norm": 1.7256579399108887,
      "learning_rate": 7.555050247314464e-06,
      "loss": 0.3665,
      "num_input_tokens_seen": 8711664,
      "step": 13445
    },
    {
      "epoch": 7.930424528301887,
      "grad_norm": 1.7717788219451904,
      "learning_rate": 7.552838469126289e-06,
      "loss": 0.3794,
      "num_input_tokens_seen": 8716016,
      "step": 13450
    },
    {
      "epoch": 7.933372641509434,
      "grad_norm": 1.8598086833953857,
      "learning_rate": 7.550626015063125e-06,
      "loss": 0.2606,
      "num_input_tokens_seen": 8720624,
      "step": 13455
    },
    {
      "epoch": 7.936320754716981,
      "grad_norm": 2.6038191318511963,
      "learning_rate": 7.548412885710734e-06,
      "loss": 0.423,
      "num_input_tokens_seen": 8723760,
      "step": 13460
    },
    {
      "epoch": 7.939268867924528,
      "grad_norm": 2.748284339904785,
      "learning_rate": 7.546199081655048e-06,
      "loss": 0.4045,
      "num_input_tokens_seen": 8727408,
      "step": 13465
    },
    {
      "epoch": 7.942216981132075,
      "grad_norm": 1.8891379833221436,
      "learning_rate": 7.54398460348218e-06,
      "loss": 0.4119,
      "num_input_tokens_seen": 8730928,
      "step": 13470
    },
    {
      "epoch": 7.945165094339623,
      "grad_norm": 4.2176833152771,
      "learning_rate": 7.541769451778425e-06,
      "loss": 0.3757,
      "num_input_tokens_seen": 8734576,
      "step": 13475
    },
    {
      "epoch": 7.94811320754717,
      "grad_norm": 2.1716833114624023,
      "learning_rate": 7.5395536271302536e-06,
      "loss": 0.4207,
      "num_input_tokens_seen": 8737680,
      "step": 13480
    },
    {
      "epoch": 7.951061320754717,
      "grad_norm": 3.765096664428711,
      "learning_rate": 7.5373371301243136e-06,
      "loss": 0.4364,
      "num_input_tokens_seen": 8740720,
      "step": 13485
    },
    {
      "epoch": 7.9540094339622645,
      "grad_norm": 1.4774380922317505,
      "learning_rate": 7.535119961347433e-06,
      "loss": 0.3959,
      "num_input_tokens_seen": 8743376,
      "step": 13490
    },
    {
      "epoch": 7.956957547169811,
      "grad_norm": 1.8260458707809448,
      "learning_rate": 7.532902121386618e-06,
      "loss": 0.4258,
      "num_input_tokens_seen": 8746288,
      "step": 13495
    },
    {
      "epoch": 7.959905660377358,
      "grad_norm": 1.9186124801635742,
      "learning_rate": 7.530683610829051e-06,
      "loss": 0.4111,
      "num_input_tokens_seen": 8749328,
      "step": 13500
    },
    {
      "epoch": 7.962853773584905,
      "grad_norm": 1.8387463092803955,
      "learning_rate": 7.5284644302620906e-06,
      "loss": 0.3881,
      "num_input_tokens_seen": 8752848,
      "step": 13505
    },
    {
      "epoch": 7.965801886792453,
      "grad_norm": 2.781062126159668,
      "learning_rate": 7.526244580273274e-06,
      "loss": 0.5052,
      "num_input_tokens_seen": 8755536,
      "step": 13510
    },
    {
      "epoch": 7.96875,
      "grad_norm": 2.3882007598876953,
      "learning_rate": 7.524024061450318e-06,
      "loss": 0.5549,
      "num_input_tokens_seen": 8758544,
      "step": 13515
    },
    {
      "epoch": 7.971698113207547,
      "grad_norm": 1.552590012550354,
      "learning_rate": 7.521802874381115e-06,
      "loss": 0.4673,
      "num_input_tokens_seen": 8763440,
      "step": 13520
    },
    {
      "epoch": 7.974646226415095,
      "grad_norm": 2.0306084156036377,
      "learning_rate": 7.519581019653731e-06,
      "loss": 0.5401,
      "num_input_tokens_seen": 8765744,
      "step": 13525
    },
    {
      "epoch": 7.977594339622642,
      "grad_norm": 1.6203892230987549,
      "learning_rate": 7.517358497856413e-06,
      "loss": 0.4805,
      "num_input_tokens_seen": 8770032,
      "step": 13530
    },
    {
      "epoch": 7.980542452830189,
      "grad_norm": 2.608941078186035,
      "learning_rate": 7.515135309577584e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 8772944,
      "step": 13535
    },
    {
      "epoch": 7.9834905660377355,
      "grad_norm": 1.9129103422164917,
      "learning_rate": 7.5129114554058425e-06,
      "loss": 0.4114,
      "num_input_tokens_seen": 8776368,
      "step": 13540
    },
    {
      "epoch": 7.986438679245283,
      "grad_norm": 2.3815484046936035,
      "learning_rate": 7.510686935929963e-06,
      "loss": 0.4277,
      "num_input_tokens_seen": 8778928,
      "step": 13545
    },
    {
      "epoch": 7.98938679245283,
      "grad_norm": 1.9033836126327515,
      "learning_rate": 7.5084617517388965e-06,
      "loss": 0.3374,
      "num_input_tokens_seen": 8782416,
      "step": 13550
    },
    {
      "epoch": 7.992334905660377,
      "grad_norm": 2.0506348609924316,
      "learning_rate": 7.506235903421771e-06,
      "loss": 0.3356,
      "num_input_tokens_seen": 8785552,
      "step": 13555
    },
    {
      "epoch": 7.995283018867925,
      "grad_norm": 2.5448858737945557,
      "learning_rate": 7.504009391567889e-06,
      "loss": 0.4648,
      "num_input_tokens_seen": 8788880,
      "step": 13560
    },
    {
      "epoch": 7.998231132075472,
      "grad_norm": 2.7136919498443604,
      "learning_rate": 7.501782216766729e-06,
      "loss": 0.4142,
      "num_input_tokens_seen": 8791632,
      "step": 13565
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.5134048461914062,
      "eval_runtime": 19.5116,
      "eval_samples_per_second": 86.922,
      "eval_steps_per_second": 21.731,
      "num_input_tokens_seen": 8792592,
      "step": 13568
    },
    {
      "epoch": 8.00117924528302,
      "grad_norm": 2.516296148300171,
      "learning_rate": 7.499554379607944e-06,
      "loss": 0.4887,
      "num_input_tokens_seen": 8793872,
      "step": 13570
    },
    {
      "epoch": 8.004127358490566,
      "grad_norm": 2.8204474449157715,
      "learning_rate": 7.497325880681365e-06,
      "loss": 0.3734,
      "num_input_tokens_seen": 8796112,
      "step": 13575
    },
    {
      "epoch": 8.007075471698114,
      "grad_norm": 1.8124593496322632,
      "learning_rate": 7.495096720576994e-06,
      "loss": 0.3161,
      "num_input_tokens_seen": 8799056,
      "step": 13580
    },
    {
      "epoch": 8.01002358490566,
      "grad_norm": 1.3523786067962646,
      "learning_rate": 7.492866899885017e-06,
      "loss": 0.389,
      "num_input_tokens_seen": 8802064,
      "step": 13585
    },
    {
      "epoch": 8.012971698113208,
      "grad_norm": 2.2226576805114746,
      "learning_rate": 7.490636419195782e-06,
      "loss": 0.4226,
      "num_input_tokens_seen": 8805776,
      "step": 13590
    },
    {
      "epoch": 8.015919811320755,
      "grad_norm": 2.519725799560547,
      "learning_rate": 7.488405279099821e-06,
      "loss": 0.3893,
      "num_input_tokens_seen": 8809232,
      "step": 13595
    },
    {
      "epoch": 8.018867924528301,
      "grad_norm": 2.864819049835205,
      "learning_rate": 7.48617348018784e-06,
      "loss": 0.437,
      "num_input_tokens_seen": 8812592,
      "step": 13600
    },
    {
      "epoch": 8.02181603773585,
      "grad_norm": 2.8350892066955566,
      "learning_rate": 7.4839410230507134e-06,
      "loss": 0.2982,
      "num_input_tokens_seen": 8816016,
      "step": 13605
    },
    {
      "epoch": 8.024764150943396,
      "grad_norm": 2.509840965270996,
      "learning_rate": 7.481707908279496e-06,
      "loss": 0.6154,
      "num_input_tokens_seen": 8819216,
      "step": 13610
    },
    {
      "epoch": 8.027712264150944,
      "grad_norm": 1.725878357887268,
      "learning_rate": 7.4794741364654144e-06,
      "loss": 0.3963,
      "num_input_tokens_seen": 8822896,
      "step": 13615
    },
    {
      "epoch": 8.03066037735849,
      "grad_norm": 2.8784067630767822,
      "learning_rate": 7.477239708199871e-06,
      "loss": 0.5062,
      "num_input_tokens_seen": 8825968,
      "step": 13620
    },
    {
      "epoch": 8.033608490566039,
      "grad_norm": 1.741591453552246,
      "learning_rate": 7.475004624074434e-06,
      "loss": 0.3441,
      "num_input_tokens_seen": 8829136,
      "step": 13625
    },
    {
      "epoch": 8.036556603773585,
      "grad_norm": 1.9035379886627197,
      "learning_rate": 7.4727688846808595e-06,
      "loss": 0.4366,
      "num_input_tokens_seen": 8832336,
      "step": 13630
    },
    {
      "epoch": 8.039504716981131,
      "grad_norm": 2.1268346309661865,
      "learning_rate": 7.4705324906110654e-06,
      "loss": 0.3607,
      "num_input_tokens_seen": 8835280,
      "step": 13635
    },
    {
      "epoch": 8.04245283018868,
      "grad_norm": 2.0918631553649902,
      "learning_rate": 7.4682954424571466e-06,
      "loss": 0.4471,
      "num_input_tokens_seen": 8838704,
      "step": 13640
    },
    {
      "epoch": 8.045400943396226,
      "grad_norm": 1.9511775970458984,
      "learning_rate": 7.466057740811372e-06,
      "loss": 0.369,
      "num_input_tokens_seen": 8842128,
      "step": 13645
    },
    {
      "epoch": 8.048349056603774,
      "grad_norm": 3.9508564472198486,
      "learning_rate": 7.463819386266182e-06,
      "loss": 0.3228,
      "num_input_tokens_seen": 8845104,
      "step": 13650
    },
    {
      "epoch": 8.05129716981132,
      "grad_norm": 1.81194269657135,
      "learning_rate": 7.461580379414191e-06,
      "loss": 0.4263,
      "num_input_tokens_seen": 8847632,
      "step": 13655
    },
    {
      "epoch": 8.054245283018869,
      "grad_norm": 3.1810190677642822,
      "learning_rate": 7.459340720848187e-06,
      "loss": 0.4048,
      "num_input_tokens_seen": 8851376,
      "step": 13660
    },
    {
      "epoch": 8.057193396226415,
      "grad_norm": 1.8540452718734741,
      "learning_rate": 7.457100411161128e-06,
      "loss": 0.4759,
      "num_input_tokens_seen": 8854512,
      "step": 13665
    },
    {
      "epoch": 8.060141509433961,
      "grad_norm": 1.6614478826522827,
      "learning_rate": 7.454859450946144e-06,
      "loss": 0.323,
      "num_input_tokens_seen": 8859152,
      "step": 13670
    },
    {
      "epoch": 8.06308962264151,
      "grad_norm": 1.78848135471344,
      "learning_rate": 7.4526178407965396e-06,
      "loss": 0.3942,
      "num_input_tokens_seen": 8862416,
      "step": 13675
    },
    {
      "epoch": 8.066037735849056,
      "grad_norm": 2.1573431491851807,
      "learning_rate": 7.450375581305794e-06,
      "loss": 0.3696,
      "num_input_tokens_seen": 8865584,
      "step": 13680
    },
    {
      "epoch": 8.068985849056604,
      "grad_norm": 3.028982162475586,
      "learning_rate": 7.448132673067552e-06,
      "loss": 0.3782,
      "num_input_tokens_seen": 8868816,
      "step": 13685
    },
    {
      "epoch": 8.07193396226415,
      "grad_norm": 2.0230441093444824,
      "learning_rate": 7.445889116675634e-06,
      "loss": 0.302,
      "num_input_tokens_seen": 8872208,
      "step": 13690
    },
    {
      "epoch": 8.074882075471699,
      "grad_norm": 2.646719455718994,
      "learning_rate": 7.443644912724031e-06,
      "loss": 0.3684,
      "num_input_tokens_seen": 8875568,
      "step": 13695
    },
    {
      "epoch": 8.077830188679245,
      "grad_norm": 2.165346145629883,
      "learning_rate": 7.441400061806907e-06,
      "loss": 0.4165,
      "num_input_tokens_seen": 8878416,
      "step": 13700
    },
    {
      "epoch": 8.080778301886792,
      "grad_norm": 1.8164491653442383,
      "learning_rate": 7.439154564518592e-06,
      "loss": 0.3421,
      "num_input_tokens_seen": 8882064,
      "step": 13705
    },
    {
      "epoch": 8.08372641509434,
      "grad_norm": 3.0796375274658203,
      "learning_rate": 7.436908421453597e-06,
      "loss": 0.3826,
      "num_input_tokens_seen": 8885520,
      "step": 13710
    },
    {
      "epoch": 8.086674528301886,
      "grad_norm": 1.6505786180496216,
      "learning_rate": 7.434661633206593e-06,
      "loss": 0.3292,
      "num_input_tokens_seen": 8888432,
      "step": 13715
    },
    {
      "epoch": 8.089622641509434,
      "grad_norm": 1.8145411014556885,
      "learning_rate": 7.4324142003724286e-06,
      "loss": 0.3554,
      "num_input_tokens_seen": 8892816,
      "step": 13720
    },
    {
      "epoch": 8.09257075471698,
      "grad_norm": 2.1488571166992188,
      "learning_rate": 7.430166123546122e-06,
      "loss": 0.3919,
      "num_input_tokens_seen": 8896432,
      "step": 13725
    },
    {
      "epoch": 8.095518867924529,
      "grad_norm": 2.9397010803222656,
      "learning_rate": 7.427917403322862e-06,
      "loss": 0.3347,
      "num_input_tokens_seen": 8900336,
      "step": 13730
    },
    {
      "epoch": 8.098466981132075,
      "grad_norm": 3.356501817703247,
      "learning_rate": 7.425668040298003e-06,
      "loss": 0.4796,
      "num_input_tokens_seen": 8903600,
      "step": 13735
    },
    {
      "epoch": 8.101415094339623,
      "grad_norm": 1.3694499731063843,
      "learning_rate": 7.4234180350670785e-06,
      "loss": 0.3295,
      "num_input_tokens_seen": 8907024,
      "step": 13740
    },
    {
      "epoch": 8.10436320754717,
      "grad_norm": 1.695972204208374,
      "learning_rate": 7.421167388225785e-06,
      "loss": 0.3315,
      "num_input_tokens_seen": 8910128,
      "step": 13745
    },
    {
      "epoch": 8.107311320754716,
      "grad_norm": 2.586124897003174,
      "learning_rate": 7.41891610036999e-06,
      "loss": 0.3348,
      "num_input_tokens_seen": 8913488,
      "step": 13750
    },
    {
      "epoch": 8.110259433962264,
      "grad_norm": 1.5732455253601074,
      "learning_rate": 7.416664172095732e-06,
      "loss": 0.3721,
      "num_input_tokens_seen": 8917744,
      "step": 13755
    },
    {
      "epoch": 8.11320754716981,
      "grad_norm": 2.4583451747894287,
      "learning_rate": 7.414411603999221e-06,
      "loss": 0.5525,
      "num_input_tokens_seen": 8920688,
      "step": 13760
    },
    {
      "epoch": 8.116155660377359,
      "grad_norm": 1.973357081413269,
      "learning_rate": 7.4121583966768295e-06,
      "loss": 0.4105,
      "num_input_tokens_seen": 8923440,
      "step": 13765
    },
    {
      "epoch": 8.119103773584905,
      "grad_norm": 2.3065457344055176,
      "learning_rate": 7.409904550725109e-06,
      "loss": 0.3453,
      "num_input_tokens_seen": 8926128,
      "step": 13770
    },
    {
      "epoch": 8.122051886792454,
      "grad_norm": 2.0598936080932617,
      "learning_rate": 7.407650066740771e-06,
      "loss": 0.4317,
      "num_input_tokens_seen": 8930000,
      "step": 13775
    },
    {
      "epoch": 8.125,
      "grad_norm": 2.0531005859375,
      "learning_rate": 7.405394945320702e-06,
      "loss": 0.2862,
      "num_input_tokens_seen": 8933008,
      "step": 13780
    },
    {
      "epoch": 8.127948113207546,
      "grad_norm": 1.3269680738449097,
      "learning_rate": 7.403139187061955e-06,
      "loss": 0.3989,
      "num_input_tokens_seen": 8936720,
      "step": 13785
    },
    {
      "epoch": 8.130896226415095,
      "grad_norm": 1.7618627548217773,
      "learning_rate": 7.400882792561752e-06,
      "loss": 0.379,
      "num_input_tokens_seen": 8939216,
      "step": 13790
    },
    {
      "epoch": 8.133844339622641,
      "grad_norm": 2.1491124629974365,
      "learning_rate": 7.39862576241748e-06,
      "loss": 0.4615,
      "num_input_tokens_seen": 8942672,
      "step": 13795
    },
    {
      "epoch": 8.13679245283019,
      "grad_norm": 2.167327404022217,
      "learning_rate": 7.396368097226703e-06,
      "loss": 0.3998,
      "num_input_tokens_seen": 8945520,
      "step": 13800
    },
    {
      "epoch": 8.139740566037736,
      "grad_norm": 3.149949789047241,
      "learning_rate": 7.394109797587144e-06,
      "loss": 0.4428,
      "num_input_tokens_seen": 8947824,
      "step": 13805
    },
    {
      "epoch": 8.142688679245284,
      "grad_norm": 3.937403917312622,
      "learning_rate": 7.3918508640966956e-06,
      "loss": 0.359,
      "num_input_tokens_seen": 8950672,
      "step": 13810
    },
    {
      "epoch": 8.14563679245283,
      "grad_norm": 2.5507447719573975,
      "learning_rate": 7.389591297353424e-06,
      "loss": 0.3789,
      "num_input_tokens_seen": 8954288,
      "step": 13815
    },
    {
      "epoch": 8.148584905660377,
      "grad_norm": 1.7594860792160034,
      "learning_rate": 7.3873310979555565e-06,
      "loss": 0.3179,
      "num_input_tokens_seen": 8957584,
      "step": 13820
    },
    {
      "epoch": 8.151533018867925,
      "grad_norm": 2.6432125568389893,
      "learning_rate": 7.385070266501495e-06,
      "loss": 0.4622,
      "num_input_tokens_seen": 8960912,
      "step": 13825
    },
    {
      "epoch": 8.154481132075471,
      "grad_norm": 1.976227879524231,
      "learning_rate": 7.382808803589798e-06,
      "loss": 0.6382,
      "num_input_tokens_seen": 8963856,
      "step": 13830
    },
    {
      "epoch": 8.15742924528302,
      "grad_norm": 2.618175983428955,
      "learning_rate": 7.380546709819204e-06,
      "loss": 0.4028,
      "num_input_tokens_seen": 8966896,
      "step": 13835
    },
    {
      "epoch": 8.160377358490566,
      "grad_norm": 1.952171802520752,
      "learning_rate": 7.378283985788608e-06,
      "loss": 0.3679,
      "num_input_tokens_seen": 8969584,
      "step": 13840
    },
    {
      "epoch": 8.163325471698114,
      "grad_norm": 1.5458275079727173,
      "learning_rate": 7.376020632097076e-06,
      "loss": 0.3557,
      "num_input_tokens_seen": 8972880,
      "step": 13845
    },
    {
      "epoch": 8.16627358490566,
      "grad_norm": 4.809352397918701,
      "learning_rate": 7.373756649343841e-06,
      "loss": 0.6835,
      "num_input_tokens_seen": 8976560,
      "step": 13850
    },
    {
      "epoch": 8.169221698113208,
      "grad_norm": 5.2060546875,
      "learning_rate": 7.371492038128305e-06,
      "loss": 0.3784,
      "num_input_tokens_seen": 8980240,
      "step": 13855
    },
    {
      "epoch": 8.172169811320755,
      "grad_norm": 2.221069097518921,
      "learning_rate": 7.36922679905003e-06,
      "loss": 0.4172,
      "num_input_tokens_seen": 8982928,
      "step": 13860
    },
    {
      "epoch": 8.175117924528301,
      "grad_norm": 3.1071274280548096,
      "learning_rate": 7.366960932708749e-06,
      "loss": 0.5061,
      "num_input_tokens_seen": 8986192,
      "step": 13865
    },
    {
      "epoch": 8.17806603773585,
      "grad_norm": 3.1158182621002197,
      "learning_rate": 7.364694439704361e-06,
      "loss": 0.4539,
      "num_input_tokens_seen": 8988880,
      "step": 13870
    },
    {
      "epoch": 8.181014150943396,
      "grad_norm": 2.4018447399139404,
      "learning_rate": 7.3624273206369264e-06,
      "loss": 0.3914,
      "num_input_tokens_seen": 8991920,
      "step": 13875
    },
    {
      "epoch": 8.183962264150944,
      "grad_norm": 3.582373857498169,
      "learning_rate": 7.360159576106681e-06,
      "loss": 0.3641,
      "num_input_tokens_seen": 8995312,
      "step": 13880
    },
    {
      "epoch": 8.18691037735849,
      "grad_norm": 1.9357610940933228,
      "learning_rate": 7.357891206714014e-06,
      "loss": 0.4014,
      "num_input_tokens_seen": 8998192,
      "step": 13885
    },
    {
      "epoch": 8.189858490566039,
      "grad_norm": 1.968258261680603,
      "learning_rate": 7.355622213059487e-06,
      "loss": 0.4183,
      "num_input_tokens_seen": 9001168,
      "step": 13890
    },
    {
      "epoch": 8.192806603773585,
      "grad_norm": 1.9877065420150757,
      "learning_rate": 7.353352595743829e-06,
      "loss": 0.3558,
      "num_input_tokens_seen": 9004272,
      "step": 13895
    },
    {
      "epoch": 8.195754716981131,
      "grad_norm": 2.2955875396728516,
      "learning_rate": 7.351082355367928e-06,
      "loss": 0.4085,
      "num_input_tokens_seen": 9007024,
      "step": 13900
    },
    {
      "epoch": 8.19870283018868,
      "grad_norm": 3.1864359378814697,
      "learning_rate": 7.34881149253284e-06,
      "loss": 0.3243,
      "num_input_tokens_seen": 9009712,
      "step": 13905
    },
    {
      "epoch": 8.201650943396226,
      "grad_norm": 2.2280707359313965,
      "learning_rate": 7.346540007839787e-06,
      "loss": 0.3791,
      "num_input_tokens_seen": 9012656,
      "step": 13910
    },
    {
      "epoch": 8.204599056603774,
      "grad_norm": 2.877849578857422,
      "learning_rate": 7.344267901890154e-06,
      "loss": 0.4378,
      "num_input_tokens_seen": 9015952,
      "step": 13915
    },
    {
      "epoch": 8.20754716981132,
      "grad_norm": 2.790161371231079,
      "learning_rate": 7.341995175285491e-06,
      "loss": 0.3096,
      "num_input_tokens_seen": 9020784,
      "step": 13920
    },
    {
      "epoch": 8.210495283018869,
      "grad_norm": 2.2974822521209717,
      "learning_rate": 7.339721828627512e-06,
      "loss": 0.3492,
      "num_input_tokens_seen": 9024208,
      "step": 13925
    },
    {
      "epoch": 8.213443396226415,
      "grad_norm": 2.510770082473755,
      "learning_rate": 7.337447862518096e-06,
      "loss": 0.3698,
      "num_input_tokens_seen": 9027888,
      "step": 13930
    },
    {
      "epoch": 8.216391509433961,
      "grad_norm": 1.7848647832870483,
      "learning_rate": 7.335173277559282e-06,
      "loss": 0.3208,
      "num_input_tokens_seen": 9032720,
      "step": 13935
    },
    {
      "epoch": 8.21933962264151,
      "grad_norm": 1.7342000007629395,
      "learning_rate": 7.332898074353281e-06,
      "loss": 0.4838,
      "num_input_tokens_seen": 9036432,
      "step": 13940
    },
    {
      "epoch": 8.222287735849056,
      "grad_norm": 1.7802433967590332,
      "learning_rate": 7.330622253502461e-06,
      "loss": 0.3286,
      "num_input_tokens_seen": 9039536,
      "step": 13945
    },
    {
      "epoch": 8.225235849056604,
      "grad_norm": 2.0966176986694336,
      "learning_rate": 7.3283458156093534e-06,
      "loss": 0.3955,
      "num_input_tokens_seen": 9043088,
      "step": 13950
    },
    {
      "epoch": 8.22818396226415,
      "grad_norm": 3.45351243019104,
      "learning_rate": 7.326068761276657e-06,
      "loss": 0.3479,
      "num_input_tokens_seen": 9046512,
      "step": 13955
    },
    {
      "epoch": 8.231132075471699,
      "grad_norm": 2.9453279972076416,
      "learning_rate": 7.323791091107231e-06,
      "loss": 0.5204,
      "num_input_tokens_seen": 9050224,
      "step": 13960
    },
    {
      "epoch": 8.234080188679245,
      "grad_norm": 2.2220704555511475,
      "learning_rate": 7.3215128057040986e-06,
      "loss": 0.5182,
      "num_input_tokens_seen": 9053008,
      "step": 13965
    },
    {
      "epoch": 8.237028301886792,
      "grad_norm": 1.3440194129943848,
      "learning_rate": 7.319233905670447e-06,
      "loss": 0.3348,
      "num_input_tokens_seen": 9056304,
      "step": 13970
    },
    {
      "epoch": 8.23997641509434,
      "grad_norm": 5.198568820953369,
      "learning_rate": 7.316954391609622e-06,
      "loss": 0.5263,
      "num_input_tokens_seen": 9059344,
      "step": 13975
    },
    {
      "epoch": 8.242924528301886,
      "grad_norm": 2.9340994358062744,
      "learning_rate": 7.314674264125137e-06,
      "loss": 0.3388,
      "num_input_tokens_seen": 9065104,
      "step": 13980
    },
    {
      "epoch": 8.245872641509434,
      "grad_norm": 2.07985782623291,
      "learning_rate": 7.312393523820665e-06,
      "loss": 0.4021,
      "num_input_tokens_seen": 9068464,
      "step": 13985
    },
    {
      "epoch": 8.24882075471698,
      "grad_norm": 2.283367395401001,
      "learning_rate": 7.310112171300041e-06,
      "loss": 0.4539,
      "num_input_tokens_seen": 9071632,
      "step": 13990
    },
    {
      "epoch": 8.251768867924529,
      "grad_norm": 4.592324256896973,
      "learning_rate": 7.307830207167263e-06,
      "loss": 0.4845,
      "num_input_tokens_seen": 9074448,
      "step": 13995
    },
    {
      "epoch": 8.254716981132075,
      "grad_norm": 2.212526559829712,
      "learning_rate": 7.305547632026493e-06,
      "loss": 0.4548,
      "num_input_tokens_seen": 9077712,
      "step": 14000
    },
    {
      "epoch": 8.257665094339623,
      "grad_norm": 2.51802921295166,
      "learning_rate": 7.3032644464820515e-06,
      "loss": 0.3736,
      "num_input_tokens_seen": 9080688,
      "step": 14005
    },
    {
      "epoch": 8.26061320754717,
      "grad_norm": 2.7788310050964355,
      "learning_rate": 7.30098065113842e-06,
      "loss": 0.3136,
      "num_input_tokens_seen": 9083216,
      "step": 14010
    },
    {
      "epoch": 8.263561320754716,
      "grad_norm": 2.6357245445251465,
      "learning_rate": 7.298696246600244e-06,
      "loss": 0.429,
      "num_input_tokens_seen": 9085776,
      "step": 14015
    },
    {
      "epoch": 8.266509433962264,
      "grad_norm": 3.7076528072357178,
      "learning_rate": 7.2964112334723315e-06,
      "loss": 0.2634,
      "num_input_tokens_seen": 9088752,
      "step": 14020
    },
    {
      "epoch": 8.26945754716981,
      "grad_norm": 2.941669464111328,
      "learning_rate": 7.294125612359647e-06,
      "loss": 0.5551,
      "num_input_tokens_seen": 9092784,
      "step": 14025
    },
    {
      "epoch": 8.272405660377359,
      "grad_norm": 2.526041030883789,
      "learning_rate": 7.291839383867318e-06,
      "loss": 0.4611,
      "num_input_tokens_seen": 9094928,
      "step": 14030
    },
    {
      "epoch": 8.275353773584905,
      "grad_norm": 2.4105820655822754,
      "learning_rate": 7.289552548600638e-06,
      "loss": 0.3708,
      "num_input_tokens_seen": 9098096,
      "step": 14035
    },
    {
      "epoch": 8.278301886792454,
      "grad_norm": 3.9648332595825195,
      "learning_rate": 7.287265107165052e-06,
      "loss": 0.3428,
      "num_input_tokens_seen": 9101232,
      "step": 14040
    },
    {
      "epoch": 8.28125,
      "grad_norm": 3.679180145263672,
      "learning_rate": 7.284977060166171e-06,
      "loss": 0.5674,
      "num_input_tokens_seen": 9104176,
      "step": 14045
    },
    {
      "epoch": 8.284198113207546,
      "grad_norm": 2.4938042163848877,
      "learning_rate": 7.282688408209766e-06,
      "loss": 0.3713,
      "num_input_tokens_seen": 9106736,
      "step": 14050
    },
    {
      "epoch": 8.287146226415095,
      "grad_norm": 2.7398900985717773,
      "learning_rate": 7.2803991519017655e-06,
      "loss": 0.5083,
      "num_input_tokens_seen": 9110544,
      "step": 14055
    },
    {
      "epoch": 8.290094339622641,
      "grad_norm": 2.113032817840576,
      "learning_rate": 7.2781092918482634e-06,
      "loss": 0.2947,
      "num_input_tokens_seen": 9113648,
      "step": 14060
    },
    {
      "epoch": 8.29304245283019,
      "grad_norm": 1.0597474575042725,
      "learning_rate": 7.275818828655508e-06,
      "loss": 0.3377,
      "num_input_tokens_seen": 9116528,
      "step": 14065
    },
    {
      "epoch": 8.295990566037736,
      "grad_norm": 4.143691539764404,
      "learning_rate": 7.27352776292991e-06,
      "loss": 0.4538,
      "num_input_tokens_seen": 9120848,
      "step": 14070
    },
    {
      "epoch": 8.298938679245284,
      "grad_norm": 8.346025466918945,
      "learning_rate": 7.271236095278036e-06,
      "loss": 0.4329,
      "num_input_tokens_seen": 9123984,
      "step": 14075
    },
    {
      "epoch": 8.30188679245283,
      "grad_norm": 3.13061785697937,
      "learning_rate": 7.2689438263066195e-06,
      "loss": 0.5491,
      "num_input_tokens_seen": 9126608,
      "step": 14080
    },
    {
      "epoch": 8.304834905660377,
      "grad_norm": 2.2725422382354736,
      "learning_rate": 7.266650956622546e-06,
      "loss": 0.4517,
      "num_input_tokens_seen": 9130608,
      "step": 14085
    },
    {
      "epoch": 8.307783018867925,
      "grad_norm": 3.481440544128418,
      "learning_rate": 7.2643574868328625e-06,
      "loss": 0.5255,
      "num_input_tokens_seen": 9134512,
      "step": 14090
    },
    {
      "epoch": 8.310731132075471,
      "grad_norm": 1.2125924825668335,
      "learning_rate": 7.262063417544776e-06,
      "loss": 0.3543,
      "num_input_tokens_seen": 9138064,
      "step": 14095
    },
    {
      "epoch": 8.31367924528302,
      "grad_norm": 2.2578959465026855,
      "learning_rate": 7.25976874936565e-06,
      "loss": 0.6029,
      "num_input_tokens_seen": 9141776,
      "step": 14100
    },
    {
      "epoch": 8.316627358490566,
      "grad_norm": 1.6879754066467285,
      "learning_rate": 7.257473482903009e-06,
      "loss": 0.2643,
      "num_input_tokens_seen": 9144432,
      "step": 14105
    },
    {
      "epoch": 8.319575471698114,
      "grad_norm": 1.7362356185913086,
      "learning_rate": 7.255177618764534e-06,
      "loss": 0.409,
      "num_input_tokens_seen": 9147312,
      "step": 14110
    },
    {
      "epoch": 8.32252358490566,
      "grad_norm": 2.787691831588745,
      "learning_rate": 7.252881157558065e-06,
      "loss": 0.5026,
      "num_input_tokens_seen": 9150480,
      "step": 14115
    },
    {
      "epoch": 8.325471698113208,
      "grad_norm": 2.3342607021331787,
      "learning_rate": 7.250584099891602e-06,
      "loss": 0.4103,
      "num_input_tokens_seen": 9153968,
      "step": 14120
    },
    {
      "epoch": 8.328419811320755,
      "grad_norm": 2.520944833755493,
      "learning_rate": 7.248286446373296e-06,
      "loss": 0.554,
      "num_input_tokens_seen": 9156880,
      "step": 14125
    },
    {
      "epoch": 8.331367924528301,
      "grad_norm": 2.500777006149292,
      "learning_rate": 7.245988197611466e-06,
      "loss": 0.4301,
      "num_input_tokens_seen": 9163792,
      "step": 14130
    },
    {
      "epoch": 8.33431603773585,
      "grad_norm": 1.3618745803833008,
      "learning_rate": 7.2436893542145805e-06,
      "loss": 0.3521,
      "num_input_tokens_seen": 9166640,
      "step": 14135
    },
    {
      "epoch": 8.337264150943396,
      "grad_norm": 2.802316427230835,
      "learning_rate": 7.241389916791269e-06,
      "loss": 0.4314,
      "num_input_tokens_seen": 9169136,
      "step": 14140
    },
    {
      "epoch": 8.340212264150944,
      "grad_norm": 2.524799108505249,
      "learning_rate": 7.239089885950317e-06,
      "loss": 0.5296,
      "num_input_tokens_seen": 9171920,
      "step": 14145
    },
    {
      "epoch": 8.34316037735849,
      "grad_norm": 1.9692679643630981,
      "learning_rate": 7.236789262300667e-06,
      "loss": 0.4428,
      "num_input_tokens_seen": 9175408,
      "step": 14150
    },
    {
      "epoch": 8.346108490566039,
      "grad_norm": 2.2480204105377197,
      "learning_rate": 7.23448804645142e-06,
      "loss": 0.4305,
      "num_input_tokens_seen": 9178768,
      "step": 14155
    },
    {
      "epoch": 8.349056603773585,
      "grad_norm": 2.5239980220794678,
      "learning_rate": 7.232186239011834e-06,
      "loss": 0.2996,
      "num_input_tokens_seen": 9182768,
      "step": 14160
    },
    {
      "epoch": 8.352004716981131,
      "grad_norm": 2.2471437454223633,
      "learning_rate": 7.2298838405913195e-06,
      "loss": 0.4188,
      "num_input_tokens_seen": 9185552,
      "step": 14165
    },
    {
      "epoch": 8.35495283018868,
      "grad_norm": 2.3808095455169678,
      "learning_rate": 7.227580851799448e-06,
      "loss": 0.4338,
      "num_input_tokens_seen": 9189136,
      "step": 14170
    },
    {
      "epoch": 8.357900943396226,
      "grad_norm": 2.9382011890411377,
      "learning_rate": 7.2252772732459455e-06,
      "loss": 0.5862,
      "num_input_tokens_seen": 9193680,
      "step": 14175
    },
    {
      "epoch": 8.360849056603774,
      "grad_norm": 3.2330410480499268,
      "learning_rate": 7.222973105540696e-06,
      "loss": 0.3224,
      "num_input_tokens_seen": 9197392,
      "step": 14180
    },
    {
      "epoch": 8.36379716981132,
      "grad_norm": 1.991824746131897,
      "learning_rate": 7.2206683492937345e-06,
      "loss": 0.2833,
      "num_input_tokens_seen": 9201136,
      "step": 14185
    },
    {
      "epoch": 8.366745283018869,
      "grad_norm": 1.8838733434677124,
      "learning_rate": 7.218363005115259e-06,
      "loss": 0.473,
      "num_input_tokens_seen": 9205168,
      "step": 14190
    },
    {
      "epoch": 8.369693396226415,
      "grad_norm": 3.2756264209747314,
      "learning_rate": 7.216057073615617e-06,
      "loss": 0.4017,
      "num_input_tokens_seen": 9207664,
      "step": 14195
    },
    {
      "epoch": 8.372641509433961,
      "grad_norm": 1.8922127485275269,
      "learning_rate": 7.21375055540531e-06,
      "loss": 0.4961,
      "num_input_tokens_seen": 9210704,
      "step": 14200
    },
    {
      "epoch": 8.37558962264151,
      "grad_norm": 3.891885280609131,
      "learning_rate": 7.211443451095007e-06,
      "loss": 0.3089,
      "num_input_tokens_seen": 9213008,
      "step": 14205
    },
    {
      "epoch": 8.378537735849056,
      "grad_norm": 2.488679885864258,
      "learning_rate": 7.2091357612955185e-06,
      "loss": 0.2917,
      "num_input_tokens_seen": 9217296,
      "step": 14210
    },
    {
      "epoch": 8.381485849056604,
      "grad_norm": 1.5056397914886475,
      "learning_rate": 7.206827486617816e-06,
      "loss": 0.3914,
      "num_input_tokens_seen": 9220144,
      "step": 14215
    },
    {
      "epoch": 8.38443396226415,
      "grad_norm": 2.7477002143859863,
      "learning_rate": 7.204518627673026e-06,
      "loss": 0.5562,
      "num_input_tokens_seen": 9222768,
      "step": 14220
    },
    {
      "epoch": 8.387382075471699,
      "grad_norm": 1.9385287761688232,
      "learning_rate": 7.202209185072428e-06,
      "loss": 0.3409,
      "num_input_tokens_seen": 9225584,
      "step": 14225
    },
    {
      "epoch": 8.390330188679245,
      "grad_norm": 1.5278455018997192,
      "learning_rate": 7.199899159427457e-06,
      "loss": 0.5616,
      "num_input_tokens_seen": 9228720,
      "step": 14230
    },
    {
      "epoch": 8.393278301886792,
      "grad_norm": 2.3166656494140625,
      "learning_rate": 7.1975885513497035e-06,
      "loss": 0.5072,
      "num_input_tokens_seen": 9231568,
      "step": 14235
    },
    {
      "epoch": 8.39622641509434,
      "grad_norm": 1.9920352697372437,
      "learning_rate": 7.195277361450909e-06,
      "loss": 0.4673,
      "num_input_tokens_seen": 9234288,
      "step": 14240
    },
    {
      "epoch": 8.399174528301886,
      "grad_norm": 3.150083541870117,
      "learning_rate": 7.192965590342973e-06,
      "loss": 0.3058,
      "num_input_tokens_seen": 9237680,
      "step": 14245
    },
    {
      "epoch": 8.402122641509434,
      "grad_norm": 1.9845867156982422,
      "learning_rate": 7.190653238637945e-06,
      "loss": 0.3488,
      "num_input_tokens_seen": 9240176,
      "step": 14250
    },
    {
      "epoch": 8.40507075471698,
      "grad_norm": 2.242931365966797,
      "learning_rate": 7.18834030694803e-06,
      "loss": 0.3967,
      "num_input_tokens_seen": 9243760,
      "step": 14255
    },
    {
      "epoch": 8.408018867924529,
      "grad_norm": 2.8369717597961426,
      "learning_rate": 7.186026795885589e-06,
      "loss": 0.4473,
      "num_input_tokens_seen": 9247728,
      "step": 14260
    },
    {
      "epoch": 8.410966981132075,
      "grad_norm": 2.9116504192352295,
      "learning_rate": 7.183712706063133e-06,
      "loss": 0.2858,
      "num_input_tokens_seen": 9250800,
      "step": 14265
    },
    {
      "epoch": 8.413915094339623,
      "grad_norm": 1.8056211471557617,
      "learning_rate": 7.1813980380933255e-06,
      "loss": 0.389,
      "num_input_tokens_seen": 9253456,
      "step": 14270
    },
    {
      "epoch": 8.41686320754717,
      "grad_norm": 1.6194428205490112,
      "learning_rate": 7.179082792588986e-06,
      "loss": 0.3332,
      "num_input_tokens_seen": 9256816,
      "step": 14275
    },
    {
      "epoch": 8.419811320754716,
      "grad_norm": 2.2651350498199463,
      "learning_rate": 7.176766970163087e-06,
      "loss": 0.4483,
      "num_input_tokens_seen": 9260304,
      "step": 14280
    },
    {
      "epoch": 8.422759433962264,
      "grad_norm": 2.3212268352508545,
      "learning_rate": 7.17445057142875e-06,
      "loss": 0.3433,
      "num_input_tokens_seen": 9263888,
      "step": 14285
    },
    {
      "epoch": 8.42570754716981,
      "grad_norm": 1.842368245124817,
      "learning_rate": 7.172133596999253e-06,
      "loss": 0.3282,
      "num_input_tokens_seen": 9267792,
      "step": 14290
    },
    {
      "epoch": 8.428655660377359,
      "grad_norm": 3.1013693809509277,
      "learning_rate": 7.1698160474880255e-06,
      "loss": 0.3805,
      "num_input_tokens_seen": 9270640,
      "step": 14295
    },
    {
      "epoch": 8.431603773584905,
      "grad_norm": 2.4580461978912354,
      "learning_rate": 7.167497923508648e-06,
      "loss": 0.4827,
      "num_input_tokens_seen": 9273008,
      "step": 14300
    },
    {
      "epoch": 8.434551886792454,
      "grad_norm": 2.3935301303863525,
      "learning_rate": 7.165179225674854e-06,
      "loss": 0.3495,
      "num_input_tokens_seen": 9275504,
      "step": 14305
    },
    {
      "epoch": 8.4375,
      "grad_norm": 2.2293548583984375,
      "learning_rate": 7.1628599546005276e-06,
      "loss": 0.3531,
      "num_input_tokens_seen": 9279120,
      "step": 14310
    },
    {
      "epoch": 8.440448113207546,
      "grad_norm": 2.5160346031188965,
      "learning_rate": 7.160540110899708e-06,
      "loss": 0.3819,
      "num_input_tokens_seen": 9281936,
      "step": 14315
    },
    {
      "epoch": 8.443396226415095,
      "grad_norm": 4.10056209564209,
      "learning_rate": 7.158219695186582e-06,
      "loss": 0.5638,
      "num_input_tokens_seen": 9285232,
      "step": 14320
    },
    {
      "epoch": 8.446344339622641,
      "grad_norm": 3.935758590698242,
      "learning_rate": 7.1558987080754905e-06,
      "loss": 0.3912,
      "num_input_tokens_seen": 9288912,
      "step": 14325
    },
    {
      "epoch": 8.44929245283019,
      "grad_norm": 2.4816513061523438,
      "learning_rate": 7.1535771501809245e-06,
      "loss": 0.4141,
      "num_input_tokens_seen": 9291728,
      "step": 14330
    },
    {
      "epoch": 8.452240566037736,
      "grad_norm": 1.7855618000030518,
      "learning_rate": 7.151255022117527e-06,
      "loss": 0.6632,
      "num_input_tokens_seen": 9295504,
      "step": 14335
    },
    {
      "epoch": 8.455188679245284,
      "grad_norm": 2.274231195449829,
      "learning_rate": 7.148932324500091e-06,
      "loss": 0.449,
      "num_input_tokens_seen": 9299696,
      "step": 14340
    },
    {
      "epoch": 8.45813679245283,
      "grad_norm": 2.336519241333008,
      "learning_rate": 7.146609057943559e-06,
      "loss": 0.3826,
      "num_input_tokens_seen": 9302736,
      "step": 14345
    },
    {
      "epoch": 8.461084905660377,
      "grad_norm": 2.4922261238098145,
      "learning_rate": 7.14428522306303e-06,
      "loss": 0.3941,
      "num_input_tokens_seen": 9305264,
      "step": 14350
    },
    {
      "epoch": 8.464033018867925,
      "grad_norm": 1.7522164583206177,
      "learning_rate": 7.141960820473745e-06,
      "loss": 0.3256,
      "num_input_tokens_seen": 9308528,
      "step": 14355
    },
    {
      "epoch": 8.466981132075471,
      "grad_norm": 1.857146143913269,
      "learning_rate": 7.139635850791102e-06,
      "loss": 0.3366,
      "num_input_tokens_seen": 9311152,
      "step": 14360
    },
    {
      "epoch": 8.46992924528302,
      "grad_norm": 2.6034483909606934,
      "learning_rate": 7.137310314630647e-06,
      "loss": 0.3819,
      "num_input_tokens_seen": 9314192,
      "step": 14365
    },
    {
      "epoch": 8.472877358490566,
      "grad_norm": 1.8049371242523193,
      "learning_rate": 7.134984212608074e-06,
      "loss": 0.3512,
      "num_input_tokens_seen": 9317488,
      "step": 14370
    },
    {
      "epoch": 8.475825471698114,
      "grad_norm": 2.3048341274261475,
      "learning_rate": 7.13265754533923e-06,
      "loss": 0.4491,
      "num_input_tokens_seen": 9323312,
      "step": 14375
    },
    {
      "epoch": 8.47877358490566,
      "grad_norm": 2.250073194503784,
      "learning_rate": 7.130330313440109e-06,
      "loss": 0.5057,
      "num_input_tokens_seen": 9326064,
      "step": 14380
    },
    {
      "epoch": 8.481721698113208,
      "grad_norm": 1.931333065032959,
      "learning_rate": 7.128002517526856e-06,
      "loss": 0.4391,
      "num_input_tokens_seen": 9328944,
      "step": 14385
    },
    {
      "epoch": 8.484669811320755,
      "grad_norm": 3.4964406490325928,
      "learning_rate": 7.1256741582157654e-06,
      "loss": 0.4389,
      "num_input_tokens_seen": 9331824,
      "step": 14390
    },
    {
      "epoch": 8.487617924528301,
      "grad_norm": 2.4218876361846924,
      "learning_rate": 7.123345236123282e-06,
      "loss": 0.3276,
      "num_input_tokens_seen": 9334480,
      "step": 14395
    },
    {
      "epoch": 8.49056603773585,
      "grad_norm": 6.5609002113342285,
      "learning_rate": 7.121015751865994e-06,
      "loss": 0.3797,
      "num_input_tokens_seen": 9337360,
      "step": 14400
    },
    {
      "epoch": 8.493514150943396,
      "grad_norm": 3.4688684940338135,
      "learning_rate": 7.118685706060645e-06,
      "loss": 0.3176,
      "num_input_tokens_seen": 9340720,
      "step": 14405
    },
    {
      "epoch": 8.496462264150944,
      "grad_norm": 1.6824052333831787,
      "learning_rate": 7.116355099324126e-06,
      "loss": 0.3725,
      "num_input_tokens_seen": 9344176,
      "step": 14410
    },
    {
      "epoch": 8.49941037735849,
      "grad_norm": 1.8989145755767822,
      "learning_rate": 7.114023932273471e-06,
      "loss": 0.3717,
      "num_input_tokens_seen": 9346512,
      "step": 14415
    },
    {
      "epoch": 8.502358490566039,
      "grad_norm": 1.462363839149475,
      "learning_rate": 7.111692205525871e-06,
      "loss": 0.452,
      "num_input_tokens_seen": 9349456,
      "step": 14420
    },
    {
      "epoch": 8.505306603773585,
      "grad_norm": 2.6258811950683594,
      "learning_rate": 7.109359919698658e-06,
      "loss": 0.354,
      "num_input_tokens_seen": 9353904,
      "step": 14425
    },
    {
      "epoch": 8.508254716981131,
      "grad_norm": 3.4083025455474854,
      "learning_rate": 7.107027075409316e-06,
      "loss": 0.3716,
      "num_input_tokens_seen": 9357072,
      "step": 14430
    },
    {
      "epoch": 8.51120283018868,
      "grad_norm": 2.000176429748535,
      "learning_rate": 7.104693673275475e-06,
      "loss": 0.3785,
      "num_input_tokens_seen": 9360240,
      "step": 14435
    },
    {
      "epoch": 8.514150943396226,
      "grad_norm": 3.12658953666687,
      "learning_rate": 7.1023597139149115e-06,
      "loss": 0.4113,
      "num_input_tokens_seen": 9362960,
      "step": 14440
    },
    {
      "epoch": 8.517099056603774,
      "grad_norm": 2.35103178024292,
      "learning_rate": 7.100025197945555e-06,
      "loss": 0.4335,
      "num_input_tokens_seen": 9365808,
      "step": 14445
    },
    {
      "epoch": 8.52004716981132,
      "grad_norm": 3.551767587661743,
      "learning_rate": 7.097690125985476e-06,
      "loss": 0.485,
      "num_input_tokens_seen": 9368624,
      "step": 14450
    },
    {
      "epoch": 8.522995283018869,
      "grad_norm": 2.01930570602417,
      "learning_rate": 7.095354498652895e-06,
      "loss": 0.3414,
      "num_input_tokens_seen": 9372720,
      "step": 14455
    },
    {
      "epoch": 8.525943396226415,
      "grad_norm": 2.372648000717163,
      "learning_rate": 7.093018316566182e-06,
      "loss": 0.3542,
      "num_input_tokens_seen": 9376464,
      "step": 14460
    },
    {
      "epoch": 8.528891509433961,
      "grad_norm": 3.630378007888794,
      "learning_rate": 7.0906815803438465e-06,
      "loss": 0.5274,
      "num_input_tokens_seen": 9379536,
      "step": 14465
    },
    {
      "epoch": 8.53183962264151,
      "grad_norm": 2.125091552734375,
      "learning_rate": 7.088344290604554e-06,
      "loss": 0.3211,
      "num_input_tokens_seen": 9382064,
      "step": 14470
    },
    {
      "epoch": 8.534787735849056,
      "grad_norm": 2.072416305541992,
      "learning_rate": 7.086006447967111e-06,
      "loss": 0.4091,
      "num_input_tokens_seen": 9384688,
      "step": 14475
    },
    {
      "epoch": 8.537735849056604,
      "grad_norm": 2.4276156425476074,
      "learning_rate": 7.08366805305047e-06,
      "loss": 0.4688,
      "num_input_tokens_seen": 9388784,
      "step": 14480
    },
    {
      "epoch": 8.54068396226415,
      "grad_norm": 2.430649757385254,
      "learning_rate": 7.08132910647373e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 9391984,
      "step": 14485
    },
    {
      "epoch": 8.543632075471699,
      "grad_norm": 2.187922477722168,
      "learning_rate": 7.078989608856142e-06,
      "loss": 0.3714,
      "num_input_tokens_seen": 9395120,
      "step": 14490
    },
    {
      "epoch": 8.546580188679245,
      "grad_norm": 1.6468505859375,
      "learning_rate": 7.076649560817092e-06,
      "loss": 0.3612,
      "num_input_tokens_seen": 9397936,
      "step": 14495
    },
    {
      "epoch": 8.549528301886792,
      "grad_norm": 2.1999032497406006,
      "learning_rate": 7.0743089629761245e-06,
      "loss": 0.3152,
      "num_input_tokens_seen": 9401136,
      "step": 14500
    },
    {
      "epoch": 8.55247641509434,
      "grad_norm": 3.011399269104004,
      "learning_rate": 7.071967815952917e-06,
      "loss": 0.3625,
      "num_input_tokens_seen": 9404176,
      "step": 14505
    },
    {
      "epoch": 8.555424528301886,
      "grad_norm": 5.8787007331848145,
      "learning_rate": 7.0696261203673e-06,
      "loss": 0.3404,
      "num_input_tokens_seen": 9406832,
      "step": 14510
    },
    {
      "epoch": 8.558372641509434,
      "grad_norm": 3.3123013973236084,
      "learning_rate": 7.067283876839249e-06,
      "loss": 0.3752,
      "num_input_tokens_seen": 9409680,
      "step": 14515
    },
    {
      "epoch": 8.56132075471698,
      "grad_norm": 1.8370920419692993,
      "learning_rate": 7.064941085988884e-06,
      "loss": 0.3328,
      "num_input_tokens_seen": 9413392,
      "step": 14520
    },
    {
      "epoch": 8.564268867924529,
      "grad_norm": 3.067153215408325,
      "learning_rate": 7.062597748436464e-06,
      "loss": 0.5133,
      "num_input_tokens_seen": 9415856,
      "step": 14525
    },
    {
      "epoch": 8.567216981132075,
      "grad_norm": 4.301210403442383,
      "learning_rate": 7.060253864802402e-06,
      "loss": 0.3893,
      "num_input_tokens_seen": 9418320,
      "step": 14530
    },
    {
      "epoch": 8.570165094339622,
      "grad_norm": 2.2240097522735596,
      "learning_rate": 7.05790943570725e-06,
      "loss": 0.4739,
      "num_input_tokens_seen": 9422288,
      "step": 14535
    },
    {
      "epoch": 8.57311320754717,
      "grad_norm": 2.1056318283081055,
      "learning_rate": 7.055564461771704e-06,
      "loss": 0.3782,
      "num_input_tokens_seen": 9425232,
      "step": 14540
    },
    {
      "epoch": 8.576061320754716,
      "grad_norm": 2.200350046157837,
      "learning_rate": 7.053218943616611e-06,
      "loss": 0.4333,
      "num_input_tokens_seen": 9427984,
      "step": 14545
    },
    {
      "epoch": 8.579009433962264,
      "grad_norm": 1.7421069145202637,
      "learning_rate": 7.050872881862952e-06,
      "loss": 0.4267,
      "num_input_tokens_seen": 9431536,
      "step": 14550
    },
    {
      "epoch": 8.58195754716981,
      "grad_norm": 3.8501367568969727,
      "learning_rate": 7.04852627713186e-06,
      "loss": 0.3534,
      "num_input_tokens_seen": 9434192,
      "step": 14555
    },
    {
      "epoch": 8.584905660377359,
      "grad_norm": 1.9749646186828613,
      "learning_rate": 7.0461791300446055e-06,
      "loss": 0.3462,
      "num_input_tokens_seen": 9437008,
      "step": 14560
    },
    {
      "epoch": 8.587853773584905,
      "grad_norm": 2.226174831390381,
      "learning_rate": 7.043831441222611e-06,
      "loss": 0.4317,
      "num_input_tokens_seen": 9440560,
      "step": 14565
    },
    {
      "epoch": 8.590801886792454,
      "grad_norm": 4.325654029846191,
      "learning_rate": 7.04148321128743e-06,
      "loss": 0.4177,
      "num_input_tokens_seen": 9444656,
      "step": 14570
    },
    {
      "epoch": 8.59375,
      "grad_norm": 3.2384448051452637,
      "learning_rate": 7.039134440860773e-06,
      "loss": 0.3875,
      "num_input_tokens_seen": 9447792,
      "step": 14575
    },
    {
      "epoch": 8.596698113207546,
      "grad_norm": 6.5120978355407715,
      "learning_rate": 7.036785130564484e-06,
      "loss": 0.4802,
      "num_input_tokens_seen": 9450352,
      "step": 14580
    },
    {
      "epoch": 8.599646226415095,
      "grad_norm": 1.9734976291656494,
      "learning_rate": 7.0344352810205544e-06,
      "loss": 0.4775,
      "num_input_tokens_seen": 9453456,
      "step": 14585
    },
    {
      "epoch": 8.602594339622641,
      "grad_norm": 2.8035216331481934,
      "learning_rate": 7.032084892851115e-06,
      "loss": 0.4006,
      "num_input_tokens_seen": 9456592,
      "step": 14590
    },
    {
      "epoch": 8.60554245283019,
      "grad_norm": 3.527529001235962,
      "learning_rate": 7.0297339666784425e-06,
      "loss": 0.441,
      "num_input_tokens_seen": 9460016,
      "step": 14595
    },
    {
      "epoch": 8.608490566037736,
      "grad_norm": 1.9069772958755493,
      "learning_rate": 7.0273825031249556e-06,
      "loss": 0.4098,
      "num_input_tokens_seen": 9462672,
      "step": 14600
    },
    {
      "epoch": 8.611438679245284,
      "grad_norm": 3.436617851257324,
      "learning_rate": 7.025030502813213e-06,
      "loss": 0.3287,
      "num_input_tokens_seen": 9465424,
      "step": 14605
    },
    {
      "epoch": 8.61438679245283,
      "grad_norm": 1.3035259246826172,
      "learning_rate": 7.022677966365917e-06,
      "loss": 0.3847,
      "num_input_tokens_seen": 9468848,
      "step": 14610
    },
    {
      "epoch": 8.617334905660378,
      "grad_norm": 2.4510796070098877,
      "learning_rate": 7.020324894405913e-06,
      "loss": 0.4464,
      "num_input_tokens_seen": 9471952,
      "step": 14615
    },
    {
      "epoch": 8.620283018867925,
      "grad_norm": 1.8900837898254395,
      "learning_rate": 7.017971287556185e-06,
      "loss": 0.3384,
      "num_input_tokens_seen": 9475120,
      "step": 14620
    },
    {
      "epoch": 8.623231132075471,
      "grad_norm": 2.435685396194458,
      "learning_rate": 7.015617146439863e-06,
      "loss": 0.4726,
      "num_input_tokens_seen": 9478096,
      "step": 14625
    },
    {
      "epoch": 8.62617924528302,
      "grad_norm": 1.7181696891784668,
      "learning_rate": 7.0132624716802125e-06,
      "loss": 0.3128,
      "num_input_tokens_seen": 9481424,
      "step": 14630
    },
    {
      "epoch": 8.629127358490566,
      "grad_norm": 3.3300063610076904,
      "learning_rate": 7.0109072639006474e-06,
      "loss": 0.4419,
      "num_input_tokens_seen": 9485584,
      "step": 14635
    },
    {
      "epoch": 8.632075471698114,
      "grad_norm": 2.594794750213623,
      "learning_rate": 7.008551523724717e-06,
      "loss": 0.4082,
      "num_input_tokens_seen": 9489040,
      "step": 14640
    },
    {
      "epoch": 8.63502358490566,
      "grad_norm": 2.888981580734253,
      "learning_rate": 7.006195251776116e-06,
      "loss": 0.4466,
      "num_input_tokens_seen": 9491600,
      "step": 14645
    },
    {
      "epoch": 8.637971698113208,
      "grad_norm": 3.2458131313323975,
      "learning_rate": 7.003838448678674e-06,
      "loss": 0.3519,
      "num_input_tokens_seen": 9494544,
      "step": 14650
    },
    {
      "epoch": 8.640919811320755,
      "grad_norm": 2.9466657638549805,
      "learning_rate": 7.0014811150563675e-06,
      "loss": 0.3865,
      "num_input_tokens_seen": 9497680,
      "step": 14655
    },
    {
      "epoch": 8.643867924528301,
      "grad_norm": 1.8602238893508911,
      "learning_rate": 6.999123251533311e-06,
      "loss": 0.4549,
      "num_input_tokens_seen": 9500848,
      "step": 14660
    },
    {
      "epoch": 8.64681603773585,
      "grad_norm": 2.015270233154297,
      "learning_rate": 6.996764858733756e-06,
      "loss": 0.4161,
      "num_input_tokens_seen": 9503536,
      "step": 14665
    },
    {
      "epoch": 8.649764150943396,
      "grad_norm": 1.8981071710586548,
      "learning_rate": 6.994405937282099e-06,
      "loss": 0.3221,
      "num_input_tokens_seen": 9506576,
      "step": 14670
    },
    {
      "epoch": 8.652712264150944,
      "grad_norm": 1.2834800481796265,
      "learning_rate": 6.9920464878028745e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 9510352,
      "step": 14675
    },
    {
      "epoch": 8.65566037735849,
      "grad_norm": 1.8085260391235352,
      "learning_rate": 6.989686510920758e-06,
      "loss": 0.4596,
      "num_input_tokens_seen": 9513520,
      "step": 14680
    },
    {
      "epoch": 8.658608490566039,
      "grad_norm": 1.7667748928070068,
      "learning_rate": 6.9873260072605634e-06,
      "loss": 0.3675,
      "num_input_tokens_seen": 9517936,
      "step": 14685
    },
    {
      "epoch": 8.661556603773585,
      "grad_norm": 2.645887613296509,
      "learning_rate": 6.984964977447243e-06,
      "loss": 0.3723,
      "num_input_tokens_seen": 9521552,
      "step": 14690
    },
    {
      "epoch": 8.664504716981131,
      "grad_norm": 3.1044046878814697,
      "learning_rate": 6.982603422105889e-06,
      "loss": 0.3333,
      "num_input_tokens_seen": 9525232,
      "step": 14695
    },
    {
      "epoch": 8.66745283018868,
      "grad_norm": 2.3391268253326416,
      "learning_rate": 6.980241341861736e-06,
      "loss": 0.4001,
      "num_input_tokens_seen": 9528656,
      "step": 14700
    },
    {
      "epoch": 8.670400943396226,
      "grad_norm": 2.4977922439575195,
      "learning_rate": 6.977878737340153e-06,
      "loss": 0.3871,
      "num_input_tokens_seen": 9531792,
      "step": 14705
    },
    {
      "epoch": 8.673349056603774,
      "grad_norm": 2.74855637550354,
      "learning_rate": 6.97551560916665e-06,
      "loss": 0.3653,
      "num_input_tokens_seen": 9536144,
      "step": 14710
    },
    {
      "epoch": 8.67629716981132,
      "grad_norm": 2.3061106204986572,
      "learning_rate": 6.973151957966875e-06,
      "loss": 0.3935,
      "num_input_tokens_seen": 9538992,
      "step": 14715
    },
    {
      "epoch": 8.679245283018869,
      "grad_norm": 2.335095167160034,
      "learning_rate": 6.970787784366616e-06,
      "loss": 0.4318,
      "num_input_tokens_seen": 9541808,
      "step": 14720
    },
    {
      "epoch": 8.682193396226415,
      "grad_norm": 2.234062910079956,
      "learning_rate": 6.968423088991797e-06,
      "loss": 0.4184,
      "num_input_tokens_seen": 9545296,
      "step": 14725
    },
    {
      "epoch": 8.685141509433961,
      "grad_norm": 10.694931030273438,
      "learning_rate": 6.966057872468481e-06,
      "loss": 0.4971,
      "num_input_tokens_seen": 9547856,
      "step": 14730
    },
    {
      "epoch": 8.68808962264151,
      "grad_norm": 2.506686210632324,
      "learning_rate": 6.963692135422872e-06,
      "loss": 0.4839,
      "num_input_tokens_seen": 9551728,
      "step": 14735
    },
    {
      "epoch": 8.691037735849056,
      "grad_norm": 1.7478139400482178,
      "learning_rate": 6.961325878481305e-06,
      "loss": 0.4696,
      "num_input_tokens_seen": 9555632,
      "step": 14740
    },
    {
      "epoch": 8.693985849056604,
      "grad_norm": 2.1073927879333496,
      "learning_rate": 6.958959102270259e-06,
      "loss": 0.4333,
      "num_input_tokens_seen": 9559120,
      "step": 14745
    },
    {
      "epoch": 8.69693396226415,
      "grad_norm": 4.196281909942627,
      "learning_rate": 6.95659180741635e-06,
      "loss": 0.5263,
      "num_input_tokens_seen": 9562320,
      "step": 14750
    },
    {
      "epoch": 8.699882075471699,
      "grad_norm": 2.2231242656707764,
      "learning_rate": 6.954223994546326e-06,
      "loss": 0.3752,
      "num_input_tokens_seen": 9565424,
      "step": 14755
    },
    {
      "epoch": 8.702830188679245,
      "grad_norm": 2.6729226112365723,
      "learning_rate": 6.951855664287077e-06,
      "loss": 0.4491,
      "num_input_tokens_seen": 9568048,
      "step": 14760
    },
    {
      "epoch": 8.705778301886792,
      "grad_norm": 1.792404294013977,
      "learning_rate": 6.9494868172656304e-06,
      "loss": 0.3452,
      "num_input_tokens_seen": 9571344,
      "step": 14765
    },
    {
      "epoch": 8.70872641509434,
      "grad_norm": 2.8292236328125,
      "learning_rate": 6.947117454109146e-06,
      "loss": 0.3381,
      "num_input_tokens_seen": 9575280,
      "step": 14770
    },
    {
      "epoch": 8.711674528301886,
      "grad_norm": 3.337578773498535,
      "learning_rate": 6.944747575444924e-06,
      "loss": 0.3027,
      "num_input_tokens_seen": 9578096,
      "step": 14775
    },
    {
      "epoch": 8.714622641509434,
      "grad_norm": 1.9752352237701416,
      "learning_rate": 6.942377181900399e-06,
      "loss": 0.492,
      "num_input_tokens_seen": 9581488,
      "step": 14780
    },
    {
      "epoch": 8.71757075471698,
      "grad_norm": 1.6526000499725342,
      "learning_rate": 6.940006274103146e-06,
      "loss": 0.2818,
      "num_input_tokens_seen": 9584272,
      "step": 14785
    },
    {
      "epoch": 8.720518867924529,
      "grad_norm": 2.6907360553741455,
      "learning_rate": 6.93763485268087e-06,
      "loss": 0.4601,
      "num_input_tokens_seen": 9587728,
      "step": 14790
    },
    {
      "epoch": 8.723466981132075,
      "grad_norm": 3.3725733757019043,
      "learning_rate": 6.935262918261416e-06,
      "loss": 0.41,
      "num_input_tokens_seen": 9590320,
      "step": 14795
    },
    {
      "epoch": 8.726415094339622,
      "grad_norm": 2.4489083290100098,
      "learning_rate": 6.932890471472764e-06,
      "loss": 0.4022,
      "num_input_tokens_seen": 9594192,
      "step": 14800
    },
    {
      "epoch": 8.72936320754717,
      "grad_norm": 2.6559722423553467,
      "learning_rate": 6.930517512943029e-06,
      "loss": 0.4994,
      "num_input_tokens_seen": 9597168,
      "step": 14805
    },
    {
      "epoch": 8.732311320754716,
      "grad_norm": 2.265758514404297,
      "learning_rate": 6.928144043300463e-06,
      "loss": 0.3715,
      "num_input_tokens_seen": 9600880,
      "step": 14810
    },
    {
      "epoch": 8.735259433962264,
      "grad_norm": 4.035518646240234,
      "learning_rate": 6.925770063173451e-06,
      "loss": 0.3733,
      "num_input_tokens_seen": 9603856,
      "step": 14815
    },
    {
      "epoch": 8.73820754716981,
      "grad_norm": 4.968781471252441,
      "learning_rate": 6.923395573190514e-06,
      "loss": 0.4056,
      "num_input_tokens_seen": 9607472,
      "step": 14820
    },
    {
      "epoch": 8.741155660377359,
      "grad_norm": 1.875494360923767,
      "learning_rate": 6.921020573980313e-06,
      "loss": 0.3833,
      "num_input_tokens_seen": 9611152,
      "step": 14825
    },
    {
      "epoch": 8.744103773584905,
      "grad_norm": 2.4260075092315674,
      "learning_rate": 6.918645066171634e-06,
      "loss": 0.4283,
      "num_input_tokens_seen": 9614160,
      "step": 14830
    },
    {
      "epoch": 8.747051886792454,
      "grad_norm": 1.546822428703308,
      "learning_rate": 6.916269050393404e-06,
      "loss": 0.4695,
      "num_input_tokens_seen": 9617968,
      "step": 14835
    },
    {
      "epoch": 8.75,
      "grad_norm": 1.9541709423065186,
      "learning_rate": 6.913892527274686e-06,
      "loss": 0.3564,
      "num_input_tokens_seen": 9621168,
      "step": 14840
    },
    {
      "epoch": 8.752948113207546,
      "grad_norm": 2.4407219886779785,
      "learning_rate": 6.9115154974446716e-06,
      "loss": 0.4931,
      "num_input_tokens_seen": 9625296,
      "step": 14845
    },
    {
      "epoch": 8.755896226415095,
      "grad_norm": 2.314478874206543,
      "learning_rate": 6.909137961532692e-06,
      "loss": 0.3628,
      "num_input_tokens_seen": 9628272,
      "step": 14850
    },
    {
      "epoch": 8.758844339622641,
      "grad_norm": 2.126309633255005,
      "learning_rate": 6.906759920168209e-06,
      "loss": 0.3496,
      "num_input_tokens_seen": 9632144,
      "step": 14855
    },
    {
      "epoch": 8.76179245283019,
      "grad_norm": 2.924417734146118,
      "learning_rate": 6.90438137398082e-06,
      "loss": 0.4448,
      "num_input_tokens_seen": 9635184,
      "step": 14860
    },
    {
      "epoch": 8.764740566037736,
      "grad_norm": 1.6406629085540771,
      "learning_rate": 6.902002323600252e-06,
      "loss": 0.4557,
      "num_input_tokens_seen": 9638192,
      "step": 14865
    },
    {
      "epoch": 8.767688679245284,
      "grad_norm": 1.9555389881134033,
      "learning_rate": 6.899622769656373e-06,
      "loss": 0.3841,
      "num_input_tokens_seen": 9641488,
      "step": 14870
    },
    {
      "epoch": 8.77063679245283,
      "grad_norm": 2.220515012741089,
      "learning_rate": 6.897242712779179e-06,
      "loss": 0.4754,
      "num_input_tokens_seen": 9644464,
      "step": 14875
    },
    {
      "epoch": 8.773584905660378,
      "grad_norm": 2.2768032550811768,
      "learning_rate": 6.894862153598802e-06,
      "loss": 0.4153,
      "num_input_tokens_seen": 9647056,
      "step": 14880
    },
    {
      "epoch": 8.776533018867925,
      "grad_norm": 3.098088264465332,
      "learning_rate": 6.892481092745502e-06,
      "loss": 0.3333,
      "num_input_tokens_seen": 9650288,
      "step": 14885
    },
    {
      "epoch": 8.779481132075471,
      "grad_norm": 2.588426113128662,
      "learning_rate": 6.890099530849677e-06,
      "loss": 0.4751,
      "num_input_tokens_seen": 9653168,
      "step": 14890
    },
    {
      "epoch": 8.78242924528302,
      "grad_norm": 2.31150484085083,
      "learning_rate": 6.887717468541855e-06,
      "loss": 0.3672,
      "num_input_tokens_seen": 9656144,
      "step": 14895
    },
    {
      "epoch": 8.785377358490566,
      "grad_norm": 1.4794188737869263,
      "learning_rate": 6.885334906452696e-06,
      "loss": 0.4142,
      "num_input_tokens_seen": 9659344,
      "step": 14900
    },
    {
      "epoch": 8.788325471698114,
      "grad_norm": 2.016788959503174,
      "learning_rate": 6.882951845212997e-06,
      "loss": 0.407,
      "num_input_tokens_seen": 9662064,
      "step": 14905
    },
    {
      "epoch": 8.79127358490566,
      "grad_norm": 1.7586179971694946,
      "learning_rate": 6.880568285453682e-06,
      "loss": 0.2786,
      "num_input_tokens_seen": 9665328,
      "step": 14910
    },
    {
      "epoch": 8.794221698113208,
      "grad_norm": 1.8773727416992188,
      "learning_rate": 6.878184227805807e-06,
      "loss": 0.4073,
      "num_input_tokens_seen": 9668144,
      "step": 14915
    },
    {
      "epoch": 8.797169811320755,
      "grad_norm": 2.511514663696289,
      "learning_rate": 6.8757996729005645e-06,
      "loss": 0.4186,
      "num_input_tokens_seen": 9670864,
      "step": 14920
    },
    {
      "epoch": 8.800117924528301,
      "grad_norm": 2.7883009910583496,
      "learning_rate": 6.8734146213692756e-06,
      "loss": 0.3302,
      "num_input_tokens_seen": 9674032,
      "step": 14925
    },
    {
      "epoch": 8.80306603773585,
      "grad_norm": 1.8956962823867798,
      "learning_rate": 6.87102907384339e-06,
      "loss": 0.5789,
      "num_input_tokens_seen": 9677360,
      "step": 14930
    },
    {
      "epoch": 8.806014150943396,
      "grad_norm": 2.021355152130127,
      "learning_rate": 6.868643030954494e-06,
      "loss": 0.3219,
      "num_input_tokens_seen": 9680816,
      "step": 14935
    },
    {
      "epoch": 8.808962264150944,
      "grad_norm": 2.3948845863342285,
      "learning_rate": 6.866256493334302e-06,
      "loss": 0.4108,
      "num_input_tokens_seen": 9684752,
      "step": 14940
    },
    {
      "epoch": 8.81191037735849,
      "grad_norm": 2.524641752243042,
      "learning_rate": 6.863869461614659e-06,
      "loss": 0.2473,
      "num_input_tokens_seen": 9687184,
      "step": 14945
    },
    {
      "epoch": 8.814858490566039,
      "grad_norm": 2.1057863235473633,
      "learning_rate": 6.861481936427545e-06,
      "loss": 0.3865,
      "num_input_tokens_seen": 9690128,
      "step": 14950
    },
    {
      "epoch": 8.817806603773585,
      "grad_norm": 3.023084878921509,
      "learning_rate": 6.859093918405067e-06,
      "loss": 0.3939,
      "num_input_tokens_seen": 9693648,
      "step": 14955
    },
    {
      "epoch": 8.820754716981131,
      "grad_norm": 3.087796211242676,
      "learning_rate": 6.856705408179458e-06,
      "loss": 0.2719,
      "num_input_tokens_seen": 9697232,
      "step": 14960
    },
    {
      "epoch": 8.82370283018868,
      "grad_norm": 3.5719711780548096,
      "learning_rate": 6.854316406383093e-06,
      "loss": 0.5107,
      "num_input_tokens_seen": 9700496,
      "step": 14965
    },
    {
      "epoch": 8.826650943396226,
      "grad_norm": 2.3841392993927,
      "learning_rate": 6.8519269136484665e-06,
      "loss": 0.3266,
      "num_input_tokens_seen": 9704016,
      "step": 14970
    },
    {
      "epoch": 8.829599056603774,
      "grad_norm": 3.048090696334839,
      "learning_rate": 6.849536930608208e-06,
      "loss": 0.3476,
      "num_input_tokens_seen": 9706832,
      "step": 14975
    },
    {
      "epoch": 8.83254716981132,
      "grad_norm": 2.127800226211548,
      "learning_rate": 6.847146457895078e-06,
      "loss": 0.3953,
      "num_input_tokens_seen": 9709520,
      "step": 14980
    },
    {
      "epoch": 8.835495283018869,
      "grad_norm": 2.397918701171875,
      "learning_rate": 6.8447554961419615e-06,
      "loss": 0.4021,
      "num_input_tokens_seen": 9712720,
      "step": 14985
    },
    {
      "epoch": 8.838443396226415,
      "grad_norm": 5.099693775177002,
      "learning_rate": 6.842364045981876e-06,
      "loss": 0.287,
      "num_input_tokens_seen": 9716176,
      "step": 14990
    },
    {
      "epoch": 8.841391509433961,
      "grad_norm": 2.5417520999908447,
      "learning_rate": 6.83997210804797e-06,
      "loss": 0.2943,
      "num_input_tokens_seen": 9718384,
      "step": 14995
    },
    {
      "epoch": 8.84433962264151,
      "grad_norm": 2.505199909210205,
      "learning_rate": 6.837579682973519e-06,
      "loss": 0.502,
      "num_input_tokens_seen": 9723152,
      "step": 15000
    },
    {
      "epoch": 8.847287735849056,
      "grad_norm": 2.4862959384918213,
      "learning_rate": 6.835186771391926e-06,
      "loss": 0.3301,
      "num_input_tokens_seen": 9726096,
      "step": 15005
    },
    {
      "epoch": 8.850235849056604,
      "grad_norm": 3.3595376014709473,
      "learning_rate": 6.8327933739367266e-06,
      "loss": 0.3897,
      "num_input_tokens_seen": 9729584,
      "step": 15010
    },
    {
      "epoch": 8.85318396226415,
      "grad_norm": 1.6100256443023682,
      "learning_rate": 6.830399491241584e-06,
      "loss": 0.4855,
      "num_input_tokens_seen": 9733232,
      "step": 15015
    },
    {
      "epoch": 8.856132075471699,
      "grad_norm": 4.8826823234558105,
      "learning_rate": 6.828005123940287e-06,
      "loss": 0.5845,
      "num_input_tokens_seen": 9735856,
      "step": 15020
    },
    {
      "epoch": 8.859080188679245,
      "grad_norm": 3.1842479705810547,
      "learning_rate": 6.825610272666754e-06,
      "loss": 0.3739,
      "num_input_tokens_seen": 9739344,
      "step": 15025
    },
    {
      "epoch": 8.862028301886792,
      "grad_norm": 2.967012643814087,
      "learning_rate": 6.823214938055034e-06,
      "loss": 0.3331,
      "num_input_tokens_seen": 9741936,
      "step": 15030
    },
    {
      "epoch": 8.86497641509434,
      "grad_norm": 2.761531114578247,
      "learning_rate": 6.8208191207393e-06,
      "loss": 0.5196,
      "num_input_tokens_seen": 9744880,
      "step": 15035
    },
    {
      "epoch": 8.867924528301886,
      "grad_norm": 3.8684113025665283,
      "learning_rate": 6.818422821353859e-06,
      "loss": 0.3253,
      "num_input_tokens_seen": 9746960,
      "step": 15040
    },
    {
      "epoch": 8.870872641509434,
      "grad_norm": 3.620842695236206,
      "learning_rate": 6.816026040533139e-06,
      "loss": 0.4161,
      "num_input_tokens_seen": 9750608,
      "step": 15045
    },
    {
      "epoch": 8.87382075471698,
      "grad_norm": 9.281991958618164,
      "learning_rate": 6.8136287789116966e-06,
      "loss": 0.3636,
      "num_input_tokens_seen": 9753040,
      "step": 15050
    },
    {
      "epoch": 8.876768867924529,
      "grad_norm": 2.736816644668579,
      "learning_rate": 6.81123103712422e-06,
      "loss": 0.4642,
      "num_input_tokens_seen": 9755600,
      "step": 15055
    },
    {
      "epoch": 8.879716981132075,
      "grad_norm": 2.5733249187469482,
      "learning_rate": 6.808832815805518e-06,
      "loss": 0.4704,
      "num_input_tokens_seen": 9758416,
      "step": 15060
    },
    {
      "epoch": 8.882665094339622,
      "grad_norm": 1.4421311616897583,
      "learning_rate": 6.806434115590534e-06,
      "loss": 0.4297,
      "num_input_tokens_seen": 9761104,
      "step": 15065
    },
    {
      "epoch": 8.88561320754717,
      "grad_norm": 2.465843677520752,
      "learning_rate": 6.804034937114332e-06,
      "loss": 0.4696,
      "num_input_tokens_seen": 9763504,
      "step": 15070
    },
    {
      "epoch": 8.888561320754716,
      "grad_norm": 3.3365492820739746,
      "learning_rate": 6.8016352810121064e-06,
      "loss": 0.4642,
      "num_input_tokens_seen": 9768112,
      "step": 15075
    },
    {
      "epoch": 8.891509433962264,
      "grad_norm": 4.240110874176025,
      "learning_rate": 6.799235147919176e-06,
      "loss": 0.4772,
      "num_input_tokens_seen": 9771504,
      "step": 15080
    },
    {
      "epoch": 8.89445754716981,
      "grad_norm": 2.6051106452941895,
      "learning_rate": 6.796834538470985e-06,
      "loss": 0.394,
      "num_input_tokens_seen": 9774256,
      "step": 15085
    },
    {
      "epoch": 8.897405660377359,
      "grad_norm": 3.4066224098205566,
      "learning_rate": 6.794433453303106e-06,
      "loss": 0.4905,
      "num_input_tokens_seen": 9777296,
      "step": 15090
    },
    {
      "epoch": 8.900353773584905,
      "grad_norm": 1.8003243207931519,
      "learning_rate": 6.792031893051238e-06,
      "loss": 0.4865,
      "num_input_tokens_seen": 9781008,
      "step": 15095
    },
    {
      "epoch": 8.903301886792454,
      "grad_norm": 2.004033088684082,
      "learning_rate": 6.789629858351201e-06,
      "loss": 0.482,
      "num_input_tokens_seen": 9785008,
      "step": 15100
    },
    {
      "epoch": 8.90625,
      "grad_norm": 1.9640419483184814,
      "learning_rate": 6.787227349838946e-06,
      "loss": 0.3556,
      "num_input_tokens_seen": 9788048,
      "step": 15105
    },
    {
      "epoch": 8.909198113207546,
      "grad_norm": 3.1417391300201416,
      "learning_rate": 6.784824368150548e-06,
      "loss": 0.343,
      "num_input_tokens_seen": 9790864,
      "step": 15110
    },
    {
      "epoch": 8.912146226415095,
      "grad_norm": 1.444929838180542,
      "learning_rate": 6.7824209139222076e-06,
      "loss": 0.3409,
      "num_input_tokens_seen": 9794224,
      "step": 15115
    },
    {
      "epoch": 8.915094339622641,
      "grad_norm": 5.501895904541016,
      "learning_rate": 6.780016987790248e-06,
      "loss": 0.4687,
      "num_input_tokens_seen": 9796912,
      "step": 15120
    },
    {
      "epoch": 8.91804245283019,
      "grad_norm": 3.44594144821167,
      "learning_rate": 6.7776125903911194e-06,
      "loss": 0.4176,
      "num_input_tokens_seen": 9800016,
      "step": 15125
    },
    {
      "epoch": 8.920990566037736,
      "grad_norm": 2.288827419281006,
      "learning_rate": 6.775207722361396e-06,
      "loss": 0.3502,
      "num_input_tokens_seen": 9803088,
      "step": 15130
    },
    {
      "epoch": 8.923938679245284,
      "grad_norm": 4.952582836151123,
      "learning_rate": 6.772802384337778e-06,
      "loss": 0.3961,
      "num_input_tokens_seen": 9805392,
      "step": 15135
    },
    {
      "epoch": 8.92688679245283,
      "grad_norm": 1.2954721450805664,
      "learning_rate": 6.770396576957088e-06,
      "loss": 0.2986,
      "num_input_tokens_seen": 9808752,
      "step": 15140
    },
    {
      "epoch": 8.929834905660378,
      "grad_norm": 1.9110122919082642,
      "learning_rate": 6.767990300856274e-06,
      "loss": 0.3434,
      "num_input_tokens_seen": 9814192,
      "step": 15145
    },
    {
      "epoch": 8.932783018867925,
      "grad_norm": 3.520284652709961,
      "learning_rate": 6.765583556672408e-06,
      "loss": 0.4599,
      "num_input_tokens_seen": 9817360,
      "step": 15150
    },
    {
      "epoch": 8.935731132075471,
      "grad_norm": 2.221761703491211,
      "learning_rate": 6.763176345042687e-06,
      "loss": 0.6035,
      "num_input_tokens_seen": 9821840,
      "step": 15155
    },
    {
      "epoch": 8.93867924528302,
      "grad_norm": 2.3436319828033447,
      "learning_rate": 6.760768666604429e-06,
      "loss": 0.4373,
      "num_input_tokens_seen": 9824752,
      "step": 15160
    },
    {
      "epoch": 8.941627358490566,
      "grad_norm": 1.510965347290039,
      "learning_rate": 6.758360521995079e-06,
      "loss": 0.3912,
      "num_input_tokens_seen": 9827440,
      "step": 15165
    },
    {
      "epoch": 8.944575471698114,
      "grad_norm": 2.5338587760925293,
      "learning_rate": 6.755951911852202e-06,
      "loss": 0.3535,
      "num_input_tokens_seen": 9829776,
      "step": 15170
    },
    {
      "epoch": 8.94752358490566,
      "grad_norm": 2.552328109741211,
      "learning_rate": 6.7535428368134885e-06,
      "loss": 0.4324,
      "num_input_tokens_seen": 9832912,
      "step": 15175
    },
    {
      "epoch": 8.950471698113208,
      "grad_norm": 2.1255154609680176,
      "learning_rate": 6.751133297516752e-06,
      "loss": 0.3144,
      "num_input_tokens_seen": 9835664,
      "step": 15180
    },
    {
      "epoch": 8.953419811320755,
      "grad_norm": 1.8284279108047485,
      "learning_rate": 6.748723294599928e-06,
      "loss": 0.4525,
      "num_input_tokens_seen": 9838224,
      "step": 15185
    },
    {
      "epoch": 8.956367924528301,
      "grad_norm": 1.6096783876419067,
      "learning_rate": 6.746312828701075e-06,
      "loss": 0.4709,
      "num_input_tokens_seen": 9844880,
      "step": 15190
    },
    {
      "epoch": 8.95931603773585,
      "grad_norm": 3.1908388137817383,
      "learning_rate": 6.743901900458374e-06,
      "loss": 0.4594,
      "num_input_tokens_seen": 9848048,
      "step": 15195
    },
    {
      "epoch": 8.962264150943396,
      "grad_norm": 1.3699525594711304,
      "learning_rate": 6.741490510510129e-06,
      "loss": 0.314,
      "num_input_tokens_seen": 9851504,
      "step": 15200
    },
    {
      "epoch": 8.965212264150944,
      "grad_norm": 2.532405376434326,
      "learning_rate": 6.7390786594947665e-06,
      "loss": 0.3935,
      "num_input_tokens_seen": 9854512,
      "step": 15205
    },
    {
      "epoch": 8.96816037735849,
      "grad_norm": 1.3666479587554932,
      "learning_rate": 6.7366663480508335e-06,
      "loss": 0.3124,
      "num_input_tokens_seen": 9857776,
      "step": 15210
    },
    {
      "epoch": 8.971108490566039,
      "grad_norm": 2.5150697231292725,
      "learning_rate": 6.734253576817002e-06,
      "loss": 0.4095,
      "num_input_tokens_seen": 9860592,
      "step": 15215
    },
    {
      "epoch": 8.974056603773585,
      "grad_norm": 2.5042033195495605,
      "learning_rate": 6.731840346432061e-06,
      "loss": 0.3618,
      "num_input_tokens_seen": 9864112,
      "step": 15220
    },
    {
      "epoch": 8.977004716981131,
      "grad_norm": 1.8763065338134766,
      "learning_rate": 6.729426657534922e-06,
      "loss": 0.4096,
      "num_input_tokens_seen": 9867408,
      "step": 15225
    },
    {
      "epoch": 8.97995283018868,
      "grad_norm": 4.557149410247803,
      "learning_rate": 6.727012510764624e-06,
      "loss": 0.4422,
      "num_input_tokens_seen": 9870416,
      "step": 15230
    },
    {
      "epoch": 8.982900943396226,
      "grad_norm": 1.7475416660308838,
      "learning_rate": 6.724597906760322e-06,
      "loss": 0.3898,
      "num_input_tokens_seen": 9873808,
      "step": 15235
    },
    {
      "epoch": 8.985849056603774,
      "grad_norm": 2.9267191886901855,
      "learning_rate": 6.722182846161289e-06,
      "loss": 0.3333,
      "num_input_tokens_seen": 9876432,
      "step": 15240
    },
    {
      "epoch": 8.98879716981132,
      "grad_norm": 2.4872186183929443,
      "learning_rate": 6.719767329606926e-06,
      "loss": 0.5457,
      "num_input_tokens_seen": 9880208,
      "step": 15245
    },
    {
      "epoch": 8.991745283018869,
      "grad_norm": 2.355172872543335,
      "learning_rate": 6.717351357736751e-06,
      "loss": 0.3858,
      "num_input_tokens_seen": 9883824,
      "step": 15250
    },
    {
      "epoch": 8.994693396226415,
      "grad_norm": 3.210533380508423,
      "learning_rate": 6.7149349311904025e-06,
      "loss": 0.4527,
      "num_input_tokens_seen": 9887088,
      "step": 15255
    },
    {
      "epoch": 8.997641509433961,
      "grad_norm": 2.8403491973876953,
      "learning_rate": 6.712518050607642e-06,
      "loss": 0.4182,
      "num_input_tokens_seen": 9890096,
      "step": 15260
    },
    {
      "epoch": 9.00058962264151,
      "grad_norm": 5.331981182098389,
      "learning_rate": 6.710100716628345e-06,
      "loss": 0.4226,
      "num_input_tokens_seen": 9892392,
      "step": 15265
    },
    {
      "epoch": 9.003537735849056,
      "grad_norm": 1.870200753211975,
      "learning_rate": 6.707682929892513e-06,
      "loss": 0.3785,
      "num_input_tokens_seen": 9895336,
      "step": 15270
    },
    {
      "epoch": 9.006485849056604,
      "grad_norm": 3.221512794494629,
      "learning_rate": 6.705264691040266e-06,
      "loss": 0.2644,
      "num_input_tokens_seen": 9899112,
      "step": 15275
    },
    {
      "epoch": 9.00943396226415,
      "grad_norm": 2.183077812194824,
      "learning_rate": 6.7028460007118435e-06,
      "loss": 0.3757,
      "num_input_tokens_seen": 9901672,
      "step": 15280
    },
    {
      "epoch": 9.012382075471699,
      "grad_norm": 3.7173452377319336,
      "learning_rate": 6.700426859547602e-06,
      "loss": 0.4682,
      "num_input_tokens_seen": 9904584,
      "step": 15285
    },
    {
      "epoch": 9.015330188679245,
      "grad_norm": 2.0448191165924072,
      "learning_rate": 6.6980072681880224e-06,
      "loss": 0.3479,
      "num_input_tokens_seen": 9908072,
      "step": 15290
    },
    {
      "epoch": 9.018278301886792,
      "grad_norm": 3.91854190826416,
      "learning_rate": 6.695587227273699e-06,
      "loss": 0.3979,
      "num_input_tokens_seen": 9910600,
      "step": 15295
    },
    {
      "epoch": 9.02122641509434,
      "grad_norm": 3.018773078918457,
      "learning_rate": 6.69316673744535e-06,
      "loss": 0.2801,
      "num_input_tokens_seen": 9913960,
      "step": 15300
    },
    {
      "epoch": 9.024174528301886,
      "grad_norm": 3.606492042541504,
      "learning_rate": 6.6907457993438115e-06,
      "loss": 0.319,
      "num_input_tokens_seen": 9918056,
      "step": 15305
    },
    {
      "epoch": 9.027122641509434,
      "grad_norm": 1.379388451576233,
      "learning_rate": 6.688324413610036e-06,
      "loss": 0.4012,
      "num_input_tokens_seen": 9922472,
      "step": 15310
    },
    {
      "epoch": 9.03007075471698,
      "grad_norm": 2.1807289123535156,
      "learning_rate": 6.685902580885094e-06,
      "loss": 0.2461,
      "num_input_tokens_seen": 9924808,
      "step": 15315
    },
    {
      "epoch": 9.033018867924529,
      "grad_norm": 1.7359849214553833,
      "learning_rate": 6.6834803018101794e-06,
      "loss": 0.3197,
      "num_input_tokens_seen": 9927912,
      "step": 15320
    },
    {
      "epoch": 9.035966981132075,
      "grad_norm": 1.8075523376464844,
      "learning_rate": 6.681057577026599e-06,
      "loss": 0.4709,
      "num_input_tokens_seen": 9931208,
      "step": 15325
    },
    {
      "epoch": 9.038915094339623,
      "grad_norm": 3.041438102722168,
      "learning_rate": 6.6786344071757795e-06,
      "loss": 0.4223,
      "num_input_tokens_seen": 9933992,
      "step": 15330
    },
    {
      "epoch": 9.04186320754717,
      "grad_norm": 2.5706114768981934,
      "learning_rate": 6.676210792899267e-06,
      "loss": 0.3314,
      "num_input_tokens_seen": 9937160,
      "step": 15335
    },
    {
      "epoch": 9.044811320754716,
      "grad_norm": 4.427581310272217,
      "learning_rate": 6.6737867348387235e-06,
      "loss": 0.5616,
      "num_input_tokens_seen": 9940296,
      "step": 15340
    },
    {
      "epoch": 9.047759433962264,
      "grad_norm": 2.0331246852874756,
      "learning_rate": 6.671362233635926e-06,
      "loss": 0.3239,
      "num_input_tokens_seen": 9942856,
      "step": 15345
    },
    {
      "epoch": 9.05070754716981,
      "grad_norm": 1.8107304573059082,
      "learning_rate": 6.668937289932775e-06,
      "loss": 0.4752,
      "num_input_tokens_seen": 9946248,
      "step": 15350
    },
    {
      "epoch": 9.053655660377359,
      "grad_norm": 1.4316049814224243,
      "learning_rate": 6.666511904371285e-06,
      "loss": 0.3976,
      "num_input_tokens_seen": 9950152,
      "step": 15355
    },
    {
      "epoch": 9.056603773584905,
      "grad_norm": 3.068105459213257,
      "learning_rate": 6.664086077593586e-06,
      "loss": 0.3528,
      "num_input_tokens_seen": 9953736,
      "step": 15360
    },
    {
      "epoch": 9.059551886792454,
      "grad_norm": 2.1291332244873047,
      "learning_rate": 6.661659810241924e-06,
      "loss": 0.2955,
      "num_input_tokens_seen": 9956904,
      "step": 15365
    },
    {
      "epoch": 9.0625,
      "grad_norm": 1.9136241674423218,
      "learning_rate": 6.659233102958667e-06,
      "loss": 0.3664,
      "num_input_tokens_seen": 9960264,
      "step": 15370
    },
    {
      "epoch": 9.065448113207546,
      "grad_norm": 3.95762038230896,
      "learning_rate": 6.6568059563862965e-06,
      "loss": 0.3535,
      "num_input_tokens_seen": 9963816,
      "step": 15375
    },
    {
      "epoch": 9.068396226415095,
      "grad_norm": 1.693192481994629,
      "learning_rate": 6.654378371167407e-06,
      "loss": 0.4394,
      "num_input_tokens_seen": 9967176,
      "step": 15380
    },
    {
      "epoch": 9.071344339622641,
      "grad_norm": 2.807948112487793,
      "learning_rate": 6.651950347944715e-06,
      "loss": 0.3668,
      "num_input_tokens_seen": 9970248,
      "step": 15385
    },
    {
      "epoch": 9.07429245283019,
      "grad_norm": 11.005834579467773,
      "learning_rate": 6.649521887361048e-06,
      "loss": 0.543,
      "num_input_tokens_seen": 9973608,
      "step": 15390
    },
    {
      "epoch": 9.077240566037736,
      "grad_norm": 1.2776880264282227,
      "learning_rate": 6.647092990059352e-06,
      "loss": 0.2913,
      "num_input_tokens_seen": 9977352,
      "step": 15395
    },
    {
      "epoch": 9.080188679245284,
      "grad_norm": 3.0335516929626465,
      "learning_rate": 6.644663656682689e-06,
      "loss": 0.3855,
      "num_input_tokens_seen": 9980904,
      "step": 15400
    },
    {
      "epoch": 9.08313679245283,
      "grad_norm": 3.1016857624053955,
      "learning_rate": 6.642233887874234e-06,
      "loss": 0.3616,
      "num_input_tokens_seen": 9984264,
      "step": 15405
    },
    {
      "epoch": 9.086084905660377,
      "grad_norm": 2.7972934246063232,
      "learning_rate": 6.639803684277279e-06,
      "loss": 0.3348,
      "num_input_tokens_seen": 9987048,
      "step": 15410
    },
    {
      "epoch": 9.089033018867925,
      "grad_norm": 2.037322521209717,
      "learning_rate": 6.637373046535233e-06,
      "loss": 0.3688,
      "num_input_tokens_seen": 9990440,
      "step": 15415
    },
    {
      "epoch": 9.091981132075471,
      "grad_norm": 2.564774513244629,
      "learning_rate": 6.634941975291617e-06,
      "loss": 0.3879,
      "num_input_tokens_seen": 9993512,
      "step": 15420
    },
    {
      "epoch": 9.09492924528302,
      "grad_norm": 1.6835846900939941,
      "learning_rate": 6.632510471190065e-06,
      "loss": 0.3988,
      "num_input_tokens_seen": 9996424,
      "step": 15425
    },
    {
      "epoch": 9.097877358490566,
      "grad_norm": 2.1318395137786865,
      "learning_rate": 6.630078534874332e-06,
      "loss": 0.3318,
      "num_input_tokens_seen": 10003592,
      "step": 15430
    },
    {
      "epoch": 9.100825471698114,
      "grad_norm": 2.220658540725708,
      "learning_rate": 6.62764616698828e-06,
      "loss": 0.3997,
      "num_input_tokens_seen": 10007976,
      "step": 15435
    },
    {
      "epoch": 9.10377358490566,
      "grad_norm": 2.295783758163452,
      "learning_rate": 6.625213368175889e-06,
      "loss": 0.4081,
      "num_input_tokens_seen": 10011400,
      "step": 15440
    },
    {
      "epoch": 9.106721698113208,
      "grad_norm": 2.6030209064483643,
      "learning_rate": 6.622780139081256e-06,
      "loss": 0.3815,
      "num_input_tokens_seen": 10014920,
      "step": 15445
    },
    {
      "epoch": 9.109669811320755,
      "grad_norm": 3.0926451683044434,
      "learning_rate": 6.620346480348589e-06,
      "loss": 0.2796,
      "num_input_tokens_seen": 10018056,
      "step": 15450
    },
    {
      "epoch": 9.112617924528301,
      "grad_norm": 2.354705810546875,
      "learning_rate": 6.617912392622206e-06,
      "loss": 0.4094,
      "num_input_tokens_seen": 10020360,
      "step": 15455
    },
    {
      "epoch": 9.11556603773585,
      "grad_norm": 1.4432119131088257,
      "learning_rate": 6.615477876546544e-06,
      "loss": 0.3082,
      "num_input_tokens_seen": 10023816,
      "step": 15460
    },
    {
      "epoch": 9.118514150943396,
      "grad_norm": 2.2909271717071533,
      "learning_rate": 6.6130429327661535e-06,
      "loss": 0.418,
      "num_input_tokens_seen": 10027016,
      "step": 15465
    },
    {
      "epoch": 9.121462264150944,
      "grad_norm": 2.360208511352539,
      "learning_rate": 6.610607561925694e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 10029832,
      "step": 15470
    },
    {
      "epoch": 9.12441037735849,
      "grad_norm": 2.820167303085327,
      "learning_rate": 6.608171764669941e-06,
      "loss": 0.3773,
      "num_input_tokens_seen": 10032808,
      "step": 15475
    },
    {
      "epoch": 9.127358490566039,
      "grad_norm": 2.8986973762512207,
      "learning_rate": 6.605735541643783e-06,
      "loss": 0.287,
      "num_input_tokens_seen": 10035272,
      "step": 15480
    },
    {
      "epoch": 9.130306603773585,
      "grad_norm": 2.3509521484375,
      "learning_rate": 6.603298893492219e-06,
      "loss": 0.4018,
      "num_input_tokens_seen": 10037992,
      "step": 15485
    },
    {
      "epoch": 9.133254716981131,
      "grad_norm": 2.4451797008514404,
      "learning_rate": 6.6008618208603625e-06,
      "loss": 0.4262,
      "num_input_tokens_seen": 10040968,
      "step": 15490
    },
    {
      "epoch": 9.13620283018868,
      "grad_norm": 3.512507915496826,
      "learning_rate": 6.598424324393442e-06,
      "loss": 0.3564,
      "num_input_tokens_seen": 10043560,
      "step": 15495
    },
    {
      "epoch": 9.139150943396226,
      "grad_norm": 2.580174684524536,
      "learning_rate": 6.595986404736792e-06,
      "loss": 0.3593,
      "num_input_tokens_seen": 10049064,
      "step": 15500
    },
    {
      "epoch": 9.142099056603774,
      "grad_norm": 2.6218087673187256,
      "learning_rate": 6.5935480625358615e-06,
      "loss": 0.3826,
      "num_input_tokens_seen": 10051656,
      "step": 15505
    },
    {
      "epoch": 9.14504716981132,
      "grad_norm": 3.0856387615203857,
      "learning_rate": 6.591109298436216e-06,
      "loss": 0.3803,
      "num_input_tokens_seen": 10054664,
      "step": 15510
    },
    {
      "epoch": 9.147995283018869,
      "grad_norm": 1.8666086196899414,
      "learning_rate": 6.5886701130835255e-06,
      "loss": 0.3718,
      "num_input_tokens_seen": 10057640,
      "step": 15515
    },
    {
      "epoch": 9.150943396226415,
      "grad_norm": 1.8912171125411987,
      "learning_rate": 6.586230507123574e-06,
      "loss": 0.4599,
      "num_input_tokens_seen": 10060520,
      "step": 15520
    },
    {
      "epoch": 9.153891509433961,
      "grad_norm": 2.085548162460327,
      "learning_rate": 6.583790481202261e-06,
      "loss": 0.3925,
      "num_input_tokens_seen": 10064008,
      "step": 15525
    },
    {
      "epoch": 9.15683962264151,
      "grad_norm": 2.1661086082458496,
      "learning_rate": 6.5813500359655925e-06,
      "loss": 0.3495,
      "num_input_tokens_seen": 10066280,
      "step": 15530
    },
    {
      "epoch": 9.159787735849056,
      "grad_norm": 4.659165382385254,
      "learning_rate": 6.578909172059687e-06,
      "loss": 0.4459,
      "num_input_tokens_seen": 10069640,
      "step": 15535
    },
    {
      "epoch": 9.162735849056604,
      "grad_norm": 5.84096097946167,
      "learning_rate": 6.576467890130772e-06,
      "loss": 0.4976,
      "num_input_tokens_seen": 10073160,
      "step": 15540
    },
    {
      "epoch": 9.16568396226415,
      "grad_norm": 2.220313787460327,
      "learning_rate": 6.574026190825191e-06,
      "loss": 0.4717,
      "num_input_tokens_seen": 10076040,
      "step": 15545
    },
    {
      "epoch": 9.168632075471699,
      "grad_norm": 2.1932499408721924,
      "learning_rate": 6.57158407478939e-06,
      "loss": 0.3581,
      "num_input_tokens_seen": 10078824,
      "step": 15550
    },
    {
      "epoch": 9.171580188679245,
      "grad_norm": 2.465080738067627,
      "learning_rate": 6.569141542669935e-06,
      "loss": 0.5027,
      "num_input_tokens_seen": 10082024,
      "step": 15555
    },
    {
      "epoch": 9.174528301886792,
      "grad_norm": 2.7712137699127197,
      "learning_rate": 6.566698595113492e-06,
      "loss": 0.4936,
      "num_input_tokens_seen": 10085576,
      "step": 15560
    },
    {
      "epoch": 9.17747641509434,
      "grad_norm": 2.649594783782959,
      "learning_rate": 6.564255232766843e-06,
      "loss": 0.3217,
      "num_input_tokens_seen": 10089864,
      "step": 15565
    },
    {
      "epoch": 9.180424528301886,
      "grad_norm": 2.2998831272125244,
      "learning_rate": 6.561811456276881e-06,
      "loss": 0.6089,
      "num_input_tokens_seen": 10092776,
      "step": 15570
    },
    {
      "epoch": 9.183372641509434,
      "grad_norm": 5.5282392501831055,
      "learning_rate": 6.559367266290605e-06,
      "loss": 0.3181,
      "num_input_tokens_seen": 10095880,
      "step": 15575
    },
    {
      "epoch": 9.18632075471698,
      "grad_norm": 3.7671191692352295,
      "learning_rate": 6.556922663455123e-06,
      "loss": 0.4081,
      "num_input_tokens_seen": 10098856,
      "step": 15580
    },
    {
      "epoch": 9.189268867924529,
      "grad_norm": 2.954158306121826,
      "learning_rate": 6.554477648417657e-06,
      "loss": 0.3619,
      "num_input_tokens_seen": 10101896,
      "step": 15585
    },
    {
      "epoch": 9.192216981132075,
      "grad_norm": 2.558978319168091,
      "learning_rate": 6.552032221825535e-06,
      "loss": 0.418,
      "num_input_tokens_seen": 10105608,
      "step": 15590
    },
    {
      "epoch": 9.195165094339623,
      "grad_norm": 2.2820448875427246,
      "learning_rate": 6.549586384326192e-06,
      "loss": 0.4001,
      "num_input_tokens_seen": 10109224,
      "step": 15595
    },
    {
      "epoch": 9.19811320754717,
      "grad_norm": 1.7174211740493774,
      "learning_rate": 6.547140136567176e-06,
      "loss": 0.4492,
      "num_input_tokens_seen": 10112136,
      "step": 15600
    },
    {
      "epoch": 9.201061320754716,
      "grad_norm": 2.6053404808044434,
      "learning_rate": 6.544693479196142e-06,
      "loss": 0.4293,
      "num_input_tokens_seen": 10114824,
      "step": 15605
    },
    {
      "epoch": 9.204009433962264,
      "grad_norm": 3.3035693168640137,
      "learning_rate": 6.542246412860851e-06,
      "loss": 0.4051,
      "num_input_tokens_seen": 10118088,
      "step": 15610
    },
    {
      "epoch": 9.20695754716981,
      "grad_norm": 2.669290781021118,
      "learning_rate": 6.5397989382091754e-06,
      "loss": 0.5238,
      "num_input_tokens_seen": 10121480,
      "step": 15615
    },
    {
      "epoch": 9.209905660377359,
      "grad_norm": 2.979843854904175,
      "learning_rate": 6.537351055889096e-06,
      "loss": 0.4155,
      "num_input_tokens_seen": 10124584,
      "step": 15620
    },
    {
      "epoch": 9.212853773584905,
      "grad_norm": 4.375035762786865,
      "learning_rate": 6.534902766548698e-06,
      "loss": 0.6171,
      "num_input_tokens_seen": 10128200,
      "step": 15625
    },
    {
      "epoch": 9.215801886792454,
      "grad_norm": 2.838348627090454,
      "learning_rate": 6.532454070836176e-06,
      "loss": 0.3433,
      "num_input_tokens_seen": 10130952,
      "step": 15630
    },
    {
      "epoch": 9.21875,
      "grad_norm": 5.070117473602295,
      "learning_rate": 6.530004969399836e-06,
      "loss": 0.4289,
      "num_input_tokens_seen": 10134152,
      "step": 15635
    },
    {
      "epoch": 9.221698113207546,
      "grad_norm": 2.8108487129211426,
      "learning_rate": 6.5275554628880865e-06,
      "loss": 0.3619,
      "num_input_tokens_seen": 10137096,
      "step": 15640
    },
    {
      "epoch": 9.224646226415095,
      "grad_norm": 3.8892087936401367,
      "learning_rate": 6.525105551949444e-06,
      "loss": 0.334,
      "num_input_tokens_seen": 10140040,
      "step": 15645
    },
    {
      "epoch": 9.227594339622641,
      "grad_norm": 3.8637022972106934,
      "learning_rate": 6.522655237232535e-06,
      "loss": 0.3327,
      "num_input_tokens_seen": 10142280,
      "step": 15650
    },
    {
      "epoch": 9.23054245283019,
      "grad_norm": 1.9315299987792969,
      "learning_rate": 6.5202045193860885e-06,
      "loss": 0.295,
      "num_input_tokens_seen": 10146952,
      "step": 15655
    },
    {
      "epoch": 9.233490566037736,
      "grad_norm": 2.6370809078216553,
      "learning_rate": 6.517753399058944e-06,
      "loss": 0.6047,
      "num_input_tokens_seen": 10149992,
      "step": 15660
    },
    {
      "epoch": 9.236438679245284,
      "grad_norm": 1.3185460567474365,
      "learning_rate": 6.515301876900047e-06,
      "loss": 0.3012,
      "num_input_tokens_seen": 10153448,
      "step": 15665
    },
    {
      "epoch": 9.23938679245283,
      "grad_norm": 3.741084098815918,
      "learning_rate": 6.512849953558445e-06,
      "loss": 0.3793,
      "num_input_tokens_seen": 10156200,
      "step": 15670
    },
    {
      "epoch": 9.242334905660377,
      "grad_norm": 2.2015578746795654,
      "learning_rate": 6.510397629683301e-06,
      "loss": 0.4572,
      "num_input_tokens_seen": 10159848,
      "step": 15675
    },
    {
      "epoch": 9.245283018867925,
      "grad_norm": 2.305957794189453,
      "learning_rate": 6.507944905923872e-06,
      "loss": 0.2665,
      "num_input_tokens_seen": 10162440,
      "step": 15680
    },
    {
      "epoch": 9.248231132075471,
      "grad_norm": 1.8725517988204956,
      "learning_rate": 6.505491782929531e-06,
      "loss": 0.3215,
      "num_input_tokens_seen": 10165352,
      "step": 15685
    },
    {
      "epoch": 9.25117924528302,
      "grad_norm": 2.8534202575683594,
      "learning_rate": 6.5030382613497535e-06,
      "loss": 0.39,
      "num_input_tokens_seen": 10168136,
      "step": 15690
    },
    {
      "epoch": 9.254127358490566,
      "grad_norm": 1.6614688634872437,
      "learning_rate": 6.500584341834119e-06,
      "loss": 0.2808,
      "num_input_tokens_seen": 10171336,
      "step": 15695
    },
    {
      "epoch": 9.257075471698114,
      "grad_norm": 1.841503381729126,
      "learning_rate": 6.498130025032312e-06,
      "loss": 0.4337,
      "num_input_tokens_seen": 10174408,
      "step": 15700
    },
    {
      "epoch": 9.26002358490566,
      "grad_norm": 2.4183695316314697,
      "learning_rate": 6.495675311594123e-06,
      "loss": 0.3665,
      "num_input_tokens_seen": 10178184,
      "step": 15705
    },
    {
      "epoch": 9.262971698113208,
      "grad_norm": 1.2928409576416016,
      "learning_rate": 6.493220202169452e-06,
      "loss": 0.3313,
      "num_input_tokens_seen": 10181192,
      "step": 15710
    },
    {
      "epoch": 9.265919811320755,
      "grad_norm": 3.8228158950805664,
      "learning_rate": 6.490764697408295e-06,
      "loss": 0.414,
      "num_input_tokens_seen": 10184552,
      "step": 15715
    },
    {
      "epoch": 9.268867924528301,
      "grad_norm": 3.221595525741577,
      "learning_rate": 6.48830879796076e-06,
      "loss": 0.5547,
      "num_input_tokens_seen": 10187784,
      "step": 15720
    },
    {
      "epoch": 9.27181603773585,
      "grad_norm": 2.9797120094299316,
      "learning_rate": 6.4858525044770546e-06,
      "loss": 0.4552,
      "num_input_tokens_seen": 10190728,
      "step": 15725
    },
    {
      "epoch": 9.274764150943396,
      "grad_norm": 2.3149824142456055,
      "learning_rate": 6.483395817607497e-06,
      "loss": 0.4642,
      "num_input_tokens_seen": 10193768,
      "step": 15730
    },
    {
      "epoch": 9.277712264150944,
      "grad_norm": 2.5972907543182373,
      "learning_rate": 6.480938738002504e-06,
      "loss": 0.3724,
      "num_input_tokens_seen": 10196648,
      "step": 15735
    },
    {
      "epoch": 9.28066037735849,
      "grad_norm": 6.580033779144287,
      "learning_rate": 6.478481266312597e-06,
      "loss": 0.3768,
      "num_input_tokens_seen": 10199144,
      "step": 15740
    },
    {
      "epoch": 9.283608490566039,
      "grad_norm": 3.323174476623535,
      "learning_rate": 6.476023403188403e-06,
      "loss": 0.4721,
      "num_input_tokens_seen": 10202920,
      "step": 15745
    },
    {
      "epoch": 9.286556603773585,
      "grad_norm": 3.702686309814453,
      "learning_rate": 6.473565149280651e-06,
      "loss": 0.4167,
      "num_input_tokens_seen": 10205800,
      "step": 15750
    },
    {
      "epoch": 9.289504716981131,
      "grad_norm": 1.4805412292480469,
      "learning_rate": 6.471106505240175e-06,
      "loss": 0.295,
      "num_input_tokens_seen": 10208936,
      "step": 15755
    },
    {
      "epoch": 9.29245283018868,
      "grad_norm": 2.290114641189575,
      "learning_rate": 6.468647471717914e-06,
      "loss": 0.3771,
      "num_input_tokens_seen": 10212296,
      "step": 15760
    },
    {
      "epoch": 9.295400943396226,
      "grad_norm": 4.031867980957031,
      "learning_rate": 6.466188049364902e-06,
      "loss": 0.3478,
      "num_input_tokens_seen": 10215144,
      "step": 15765
    },
    {
      "epoch": 9.298349056603774,
      "grad_norm": 2.1130707263946533,
      "learning_rate": 6.463728238832288e-06,
      "loss": 0.4547,
      "num_input_tokens_seen": 10218984,
      "step": 15770
    },
    {
      "epoch": 9.30129716981132,
      "grad_norm": 2.8505778312683105,
      "learning_rate": 6.461268040771311e-06,
      "loss": 0.3664,
      "num_input_tokens_seen": 10221864,
      "step": 15775
    },
    {
      "epoch": 9.304245283018869,
      "grad_norm": 3.029327392578125,
      "learning_rate": 6.458807455833326e-06,
      "loss": 0.3644,
      "num_input_tokens_seen": 10224648,
      "step": 15780
    },
    {
      "epoch": 9.307193396226415,
      "grad_norm": 2.3193747997283936,
      "learning_rate": 6.456346484669778e-06,
      "loss": 0.4504,
      "num_input_tokens_seen": 10230568,
      "step": 15785
    },
    {
      "epoch": 9.310141509433961,
      "grad_norm": 3.4131879806518555,
      "learning_rate": 6.4538851279322225e-06,
      "loss": 0.34,
      "num_input_tokens_seen": 10233256,
      "step": 15790
    },
    {
      "epoch": 9.31308962264151,
      "grad_norm": 2.877987861633301,
      "learning_rate": 6.451423386272312e-06,
      "loss": 0.4173,
      "num_input_tokens_seen": 10236424,
      "step": 15795
    },
    {
      "epoch": 9.316037735849056,
      "grad_norm": 1.574031949043274,
      "learning_rate": 6.448961260341806e-06,
      "loss": 0.4185,
      "num_input_tokens_seen": 10238888,
      "step": 15800
    },
    {
      "epoch": 9.318985849056604,
      "grad_norm": 2.0733723640441895,
      "learning_rate": 6.446498750792563e-06,
      "loss": 0.3533,
      "num_input_tokens_seen": 10242440,
      "step": 15805
    },
    {
      "epoch": 9.32193396226415,
      "grad_norm": 3.625948190689087,
      "learning_rate": 6.444035858276538e-06,
      "loss": 0.5565,
      "num_input_tokens_seen": 10245384,
      "step": 15810
    },
    {
      "epoch": 9.324882075471699,
      "grad_norm": 2.1640610694885254,
      "learning_rate": 6.441572583445799e-06,
      "loss": 0.3789,
      "num_input_tokens_seen": 10247848,
      "step": 15815
    },
    {
      "epoch": 9.327830188679245,
      "grad_norm": 3.5182409286499023,
      "learning_rate": 6.439108926952504e-06,
      "loss": 0.3498,
      "num_input_tokens_seen": 10251496,
      "step": 15820
    },
    {
      "epoch": 9.330778301886792,
      "grad_norm": 2.41933012008667,
      "learning_rate": 6.43664488944892e-06,
      "loss": 0.487,
      "num_input_tokens_seen": 10254536,
      "step": 15825
    },
    {
      "epoch": 9.33372641509434,
      "grad_norm": 2.327704906463623,
      "learning_rate": 6.434180471587409e-06,
      "loss": 0.3227,
      "num_input_tokens_seen": 10257512,
      "step": 15830
    },
    {
      "epoch": 9.336674528301886,
      "grad_norm": 1.959113359451294,
      "learning_rate": 6.431715674020438e-06,
      "loss": 0.2742,
      "num_input_tokens_seen": 10260008,
      "step": 15835
    },
    {
      "epoch": 9.339622641509434,
      "grad_norm": 2.04046893119812,
      "learning_rate": 6.429250497400571e-06,
      "loss": 0.4313,
      "num_input_tokens_seen": 10263528,
      "step": 15840
    },
    {
      "epoch": 9.34257075471698,
      "grad_norm": 2.261782646179199,
      "learning_rate": 6.426784942380475e-06,
      "loss": 0.361,
      "num_input_tokens_seen": 10266248,
      "step": 15845
    },
    {
      "epoch": 9.345518867924529,
      "grad_norm": 2.689133405685425,
      "learning_rate": 6.424319009612917e-06,
      "loss": 0.4587,
      "num_input_tokens_seen": 10269480,
      "step": 15850
    },
    {
      "epoch": 9.348466981132075,
      "grad_norm": 1.8222843408584595,
      "learning_rate": 6.421852699750763e-06,
      "loss": 0.349,
      "num_input_tokens_seen": 10272712,
      "step": 15855
    },
    {
      "epoch": 9.351415094339623,
      "grad_norm": 3.9477057456970215,
      "learning_rate": 6.4193860134469775e-06,
      "loss": 0.4243,
      "num_input_tokens_seen": 10276392,
      "step": 15860
    },
    {
      "epoch": 9.35436320754717,
      "grad_norm": 2.530325412750244,
      "learning_rate": 6.416918951354629e-06,
      "loss": 0.4686,
      "num_input_tokens_seen": 10279368,
      "step": 15865
    },
    {
      "epoch": 9.357311320754716,
      "grad_norm": 2.4255268573760986,
      "learning_rate": 6.4144515141268816e-06,
      "loss": 0.4166,
      "num_input_tokens_seen": 10282120,
      "step": 15870
    },
    {
      "epoch": 9.360259433962264,
      "grad_norm": 1.5579471588134766,
      "learning_rate": 6.411983702416999e-06,
      "loss": 0.3096,
      "num_input_tokens_seen": 10285864,
      "step": 15875
    },
    {
      "epoch": 9.36320754716981,
      "grad_norm": 3.168159008026123,
      "learning_rate": 6.409515516878346e-06,
      "loss": 0.4685,
      "num_input_tokens_seen": 10289192,
      "step": 15880
    },
    {
      "epoch": 9.366155660377359,
      "grad_norm": 2.45904803276062,
      "learning_rate": 6.407046958164387e-06,
      "loss": 0.3839,
      "num_input_tokens_seen": 10293224,
      "step": 15885
    },
    {
      "epoch": 9.369103773584905,
      "grad_norm": 3.071268081665039,
      "learning_rate": 6.404578026928679e-06,
      "loss": 0.3327,
      "num_input_tokens_seen": 10296648,
      "step": 15890
    },
    {
      "epoch": 9.372051886792454,
      "grad_norm": 3.862567663192749,
      "learning_rate": 6.402108723824887e-06,
      "loss": 0.4305,
      "num_input_tokens_seen": 10299496,
      "step": 15895
    },
    {
      "epoch": 9.375,
      "grad_norm": 2.356401205062866,
      "learning_rate": 6.399639049506767e-06,
      "loss": 0.3564,
      "num_input_tokens_seen": 10302536,
      "step": 15900
    },
    {
      "epoch": 9.377948113207546,
      "grad_norm": 4.873205184936523,
      "learning_rate": 6.397169004628177e-06,
      "loss": 0.4077,
      "num_input_tokens_seen": 10306120,
      "step": 15905
    },
    {
      "epoch": 9.380896226415095,
      "grad_norm": 2.8827247619628906,
      "learning_rate": 6.39469858984307e-06,
      "loss": 0.281,
      "num_input_tokens_seen": 10309320,
      "step": 15910
    },
    {
      "epoch": 9.383844339622641,
      "grad_norm": 2.0241265296936035,
      "learning_rate": 6.3922278058055024e-06,
      "loss": 0.3287,
      "num_input_tokens_seen": 10312488,
      "step": 15915
    },
    {
      "epoch": 9.38679245283019,
      "grad_norm": 3.223846197128296,
      "learning_rate": 6.389756653169622e-06,
      "loss": 0.4071,
      "num_input_tokens_seen": 10315016,
      "step": 15920
    },
    {
      "epoch": 9.389740566037736,
      "grad_norm": 2.2879984378814697,
      "learning_rate": 6.387285132589678e-06,
      "loss": 0.3656,
      "num_input_tokens_seen": 10318344,
      "step": 15925
    },
    {
      "epoch": 9.392688679245284,
      "grad_norm": 3.623246908187866,
      "learning_rate": 6.3848132447200166e-06,
      "loss": 0.4442,
      "num_input_tokens_seen": 10320648,
      "step": 15930
    },
    {
      "epoch": 9.39563679245283,
      "grad_norm": 2.7222495079040527,
      "learning_rate": 6.38234099021508e-06,
      "loss": 0.441,
      "num_input_tokens_seen": 10323816,
      "step": 15935
    },
    {
      "epoch": 9.398584905660377,
      "grad_norm": 1.8013993501663208,
      "learning_rate": 6.379868369729409e-06,
      "loss": 0.4077,
      "num_input_tokens_seen": 10327080,
      "step": 15940
    },
    {
      "epoch": 9.401533018867925,
      "grad_norm": 4.359516620635986,
      "learning_rate": 6.377395383917642e-06,
      "loss": 0.4185,
      "num_input_tokens_seen": 10329896,
      "step": 15945
    },
    {
      "epoch": 9.404481132075471,
      "grad_norm": 1.0681259632110596,
      "learning_rate": 6.374922033434507e-06,
      "loss": 0.3953,
      "num_input_tokens_seen": 10332968,
      "step": 15950
    },
    {
      "epoch": 9.40742924528302,
      "grad_norm": 2.784475803375244,
      "learning_rate": 6.372448318934842e-06,
      "loss": 0.4386,
      "num_input_tokens_seen": 10335368,
      "step": 15955
    },
    {
      "epoch": 9.410377358490566,
      "grad_norm": 2.7573184967041016,
      "learning_rate": 6.369974241073569e-06,
      "loss": 0.352,
      "num_input_tokens_seen": 10339144,
      "step": 15960
    },
    {
      "epoch": 9.413325471698114,
      "grad_norm": 3.006347179412842,
      "learning_rate": 6.367499800505709e-06,
      "loss": 0.3079,
      "num_input_tokens_seen": 10341928,
      "step": 15965
    },
    {
      "epoch": 9.41627358490566,
      "grad_norm": 4.32913875579834,
      "learning_rate": 6.365024997886384e-06,
      "loss": 0.3494,
      "num_input_tokens_seen": 10344328,
      "step": 15970
    },
    {
      "epoch": 9.419221698113208,
      "grad_norm": 2.692568302154541,
      "learning_rate": 6.362549833870808e-06,
      "loss": 0.3874,
      "num_input_tokens_seen": 10347304,
      "step": 15975
    },
    {
      "epoch": 9.422169811320755,
      "grad_norm": 2.309537410736084,
      "learning_rate": 6.360074309114293e-06,
      "loss": 0.3649,
      "num_input_tokens_seen": 10350472,
      "step": 15980
    },
    {
      "epoch": 9.425117924528301,
      "grad_norm": 2.2164018154144287,
      "learning_rate": 6.357598424272241e-06,
      "loss": 0.4477,
      "num_input_tokens_seen": 10353544,
      "step": 15985
    },
    {
      "epoch": 9.42806603773585,
      "grad_norm": 1.4627455472946167,
      "learning_rate": 6.355122180000156e-06,
      "loss": 0.4029,
      "num_input_tokens_seen": 10357224,
      "step": 15990
    },
    {
      "epoch": 9.431014150943396,
      "grad_norm": 1.6855387687683105,
      "learning_rate": 6.352645576953635e-06,
      "loss": 0.4815,
      "num_input_tokens_seen": 10360200,
      "step": 15995
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 2.528754711151123,
      "learning_rate": 6.350168615788366e-06,
      "loss": 0.3317,
      "num_input_tokens_seen": 10363144,
      "step": 16000
    },
    {
      "epoch": 9.43691037735849,
      "grad_norm": 5.647634983062744,
      "learning_rate": 6.347691297160137e-06,
      "loss": 0.457,
      "num_input_tokens_seen": 10366120,
      "step": 16005
    },
    {
      "epoch": 9.439858490566039,
      "grad_norm": 2.841991424560547,
      "learning_rate": 6.34521362172483e-06,
      "loss": 0.4115,
      "num_input_tokens_seen": 10369704,
      "step": 16010
    },
    {
      "epoch": 9.442806603773585,
      "grad_norm": 3.8996317386627197,
      "learning_rate": 6.342735590138417e-06,
      "loss": 0.4294,
      "num_input_tokens_seen": 10372840,
      "step": 16015
    },
    {
      "epoch": 9.445754716981131,
      "grad_norm": 2.4861814975738525,
      "learning_rate": 6.340257203056972e-06,
      "loss": 0.426,
      "num_input_tokens_seen": 10375912,
      "step": 16020
    },
    {
      "epoch": 9.44870283018868,
      "grad_norm": 3.618143320083618,
      "learning_rate": 6.3377784611366554e-06,
      "loss": 0.6013,
      "num_input_tokens_seen": 10378920,
      "step": 16025
    },
    {
      "epoch": 9.451650943396226,
      "grad_norm": 2.7253310680389404,
      "learning_rate": 6.335299365033726e-06,
      "loss": 0.3479,
      "num_input_tokens_seen": 10382280,
      "step": 16030
    },
    {
      "epoch": 9.454599056603774,
      "grad_norm": 9.370060920715332,
      "learning_rate": 6.3328199154045346e-06,
      "loss": 0.4941,
      "num_input_tokens_seen": 10385000,
      "step": 16035
    },
    {
      "epoch": 9.45754716981132,
      "grad_norm": 2.882301092147827,
      "learning_rate": 6.3303401129055265e-06,
      "loss": 0.3058,
      "num_input_tokens_seen": 10387752,
      "step": 16040
    },
    {
      "epoch": 9.460495283018869,
      "grad_norm": 2.186275005340576,
      "learning_rate": 6.32785995819324e-06,
      "loss": 0.3178,
      "num_input_tokens_seen": 10391016,
      "step": 16045
    },
    {
      "epoch": 9.463443396226415,
      "grad_norm": 2.48620343208313,
      "learning_rate": 6.3253794519243075e-06,
      "loss": 0.4262,
      "num_input_tokens_seen": 10394152,
      "step": 16050
    },
    {
      "epoch": 9.466391509433961,
      "grad_norm": 3.727792739868164,
      "learning_rate": 6.322898594755452e-06,
      "loss": 0.469,
      "num_input_tokens_seen": 10397192,
      "step": 16055
    },
    {
      "epoch": 9.46933962264151,
      "grad_norm": 2.2788448333740234,
      "learning_rate": 6.320417387343492e-06,
      "loss": 0.3217,
      "num_input_tokens_seen": 10400520,
      "step": 16060
    },
    {
      "epoch": 9.472287735849056,
      "grad_norm": 5.935966491699219,
      "learning_rate": 6.3179358303453386e-06,
      "loss": 0.3969,
      "num_input_tokens_seen": 10402632,
      "step": 16065
    },
    {
      "epoch": 9.475235849056604,
      "grad_norm": 2.491203784942627,
      "learning_rate": 6.315453924417995e-06,
      "loss": 0.3761,
      "num_input_tokens_seen": 10405608,
      "step": 16070
    },
    {
      "epoch": 9.47818396226415,
      "grad_norm": 3.3007702827453613,
      "learning_rate": 6.312971670218554e-06,
      "loss": 0.3485,
      "num_input_tokens_seen": 10408360,
      "step": 16075
    },
    {
      "epoch": 9.481132075471699,
      "grad_norm": 2.5922176837921143,
      "learning_rate": 6.3104890684042055e-06,
      "loss": 0.3598,
      "num_input_tokens_seen": 10411336,
      "step": 16080
    },
    {
      "epoch": 9.484080188679245,
      "grad_norm": 3.8339762687683105,
      "learning_rate": 6.308006119632228e-06,
      "loss": 0.4177,
      "num_input_tokens_seen": 10415464,
      "step": 16085
    },
    {
      "epoch": 9.487028301886792,
      "grad_norm": 3.7707295417785645,
      "learning_rate": 6.305522824559993e-06,
      "loss": 0.3504,
      "num_input_tokens_seen": 10418312,
      "step": 16090
    },
    {
      "epoch": 9.48997641509434,
      "grad_norm": 3.868917942047119,
      "learning_rate": 6.303039183844965e-06,
      "loss": 0.528,
      "num_input_tokens_seen": 10421384,
      "step": 16095
    },
    {
      "epoch": 9.492924528301886,
      "grad_norm": 3.7906789779663086,
      "learning_rate": 6.300555198144697e-06,
      "loss": 0.4439,
      "num_input_tokens_seen": 10424808,
      "step": 16100
    },
    {
      "epoch": 9.495872641509434,
      "grad_norm": 2.425246000289917,
      "learning_rate": 6.2980708681168335e-06,
      "loss": 0.4321,
      "num_input_tokens_seen": 10428488,
      "step": 16105
    },
    {
      "epoch": 9.49882075471698,
      "grad_norm": 1.5957778692245483,
      "learning_rate": 6.2955861944191145e-06,
      "loss": 0.3239,
      "num_input_tokens_seen": 10431208,
      "step": 16110
    },
    {
      "epoch": 9.501768867924529,
      "grad_norm": 1.6795305013656616,
      "learning_rate": 6.293101177709367e-06,
      "loss": 0.5241,
      "num_input_tokens_seen": 10434344,
      "step": 16115
    },
    {
      "epoch": 9.504716981132075,
      "grad_norm": 12.445356369018555,
      "learning_rate": 6.290615818645509e-06,
      "loss": 0.3398,
      "num_input_tokens_seen": 10437160,
      "step": 16120
    },
    {
      "epoch": 9.507665094339622,
      "grad_norm": 2.576777696609497,
      "learning_rate": 6.288130117885552e-06,
      "loss": 0.3715,
      "num_input_tokens_seen": 10440136,
      "step": 16125
    },
    {
      "epoch": 9.51061320754717,
      "grad_norm": 2.815791368484497,
      "learning_rate": 6.285644076087594e-06,
      "loss": 0.2405,
      "num_input_tokens_seen": 10442792,
      "step": 16130
    },
    {
      "epoch": 9.513561320754716,
      "grad_norm": 3.782456636428833,
      "learning_rate": 6.283157693909826e-06,
      "loss": 0.5405,
      "num_input_tokens_seen": 10445960,
      "step": 16135
    },
    {
      "epoch": 9.516509433962264,
      "grad_norm": 2.1463921070098877,
      "learning_rate": 6.280670972010528e-06,
      "loss": 0.2869,
      "num_input_tokens_seen": 10449416,
      "step": 16140
    },
    {
      "epoch": 9.51945754716981,
      "grad_norm": 2.2394702434539795,
      "learning_rate": 6.278183911048072e-06,
      "loss": 0.3072,
      "num_input_tokens_seen": 10451624,
      "step": 16145
    },
    {
      "epoch": 9.522405660377359,
      "grad_norm": 5.346293926239014,
      "learning_rate": 6.275696511680915e-06,
      "loss": 0.3775,
      "num_input_tokens_seen": 10455176,
      "step": 16150
    },
    {
      "epoch": 9.525353773584905,
      "grad_norm": 2.4932258129119873,
      "learning_rate": 6.27320877456761e-06,
      "loss": 0.4024,
      "num_input_tokens_seen": 10458120,
      "step": 16155
    },
    {
      "epoch": 9.528301886792454,
      "grad_norm": 3.514224052429199,
      "learning_rate": 6.270720700366793e-06,
      "loss": 0.4334,
      "num_input_tokens_seen": 10462024,
      "step": 16160
    },
    {
      "epoch": 9.53125,
      "grad_norm": 2.2926714420318604,
      "learning_rate": 6.2682322897371974e-06,
      "loss": 0.4206,
      "num_input_tokens_seen": 10466120,
      "step": 16165
    },
    {
      "epoch": 9.534198113207546,
      "grad_norm": 3.7619476318359375,
      "learning_rate": 6.265743543337634e-06,
      "loss": 0.4359,
      "num_input_tokens_seen": 10468840,
      "step": 16170
    },
    {
      "epoch": 9.537146226415095,
      "grad_norm": 1.373587965965271,
      "learning_rate": 6.2632544618270156e-06,
      "loss": 0.4151,
      "num_input_tokens_seen": 10472840,
      "step": 16175
    },
    {
      "epoch": 9.540094339622641,
      "grad_norm": 3.390592575073242,
      "learning_rate": 6.260765045864332e-06,
      "loss": 0.2599,
      "num_input_tokens_seen": 10476328,
      "step": 16180
    },
    {
      "epoch": 9.54304245283019,
      "grad_norm": 1.8729997873306274,
      "learning_rate": 6.258275296108669e-06,
      "loss": 0.3143,
      "num_input_tokens_seen": 10480328,
      "step": 16185
    },
    {
      "epoch": 9.545990566037736,
      "grad_norm": 2.2077841758728027,
      "learning_rate": 6.2557852132191985e-06,
      "loss": 0.3697,
      "num_input_tokens_seen": 10483656,
      "step": 16190
    },
    {
      "epoch": 9.548938679245284,
      "grad_norm": 1.7557297945022583,
      "learning_rate": 6.253294797855182e-06,
      "loss": 0.3735,
      "num_input_tokens_seen": 10487432,
      "step": 16195
    },
    {
      "epoch": 9.55188679245283,
      "grad_norm": 2.0966105461120605,
      "learning_rate": 6.250804050675964e-06,
      "loss": 0.4133,
      "num_input_tokens_seen": 10490568,
      "step": 16200
    },
    {
      "epoch": 9.554834905660378,
      "grad_norm": 3.057141065597534,
      "learning_rate": 6.248312972340984e-06,
      "loss": 0.427,
      "num_input_tokens_seen": 10493544,
      "step": 16205
    },
    {
      "epoch": 9.557783018867925,
      "grad_norm": 3.651118755340576,
      "learning_rate": 6.2458215635097656e-06,
      "loss": 0.4132,
      "num_input_tokens_seen": 10497896,
      "step": 16210
    },
    {
      "epoch": 9.560731132075471,
      "grad_norm": 3.0433781147003174,
      "learning_rate": 6.2433298248419175e-06,
      "loss": 0.3885,
      "num_input_tokens_seen": 10500200,
      "step": 16215
    },
    {
      "epoch": 9.56367924528302,
      "grad_norm": 1.928633213043213,
      "learning_rate": 6.2408377569971405e-06,
      "loss": 0.3776,
      "num_input_tokens_seen": 10502280,
      "step": 16220
    },
    {
      "epoch": 9.566627358490566,
      "grad_norm": 2.1992008686065674,
      "learning_rate": 6.238345360635221e-06,
      "loss": 0.3917,
      "num_input_tokens_seen": 10504936,
      "step": 16225
    },
    {
      "epoch": 9.569575471698114,
      "grad_norm": 3.87103271484375,
      "learning_rate": 6.2358526364160274e-06,
      "loss": 0.3811,
      "num_input_tokens_seen": 10508456,
      "step": 16230
    },
    {
      "epoch": 9.57252358490566,
      "grad_norm": 1.7812739610671997,
      "learning_rate": 6.2333595849995245e-06,
      "loss": 0.3827,
      "num_input_tokens_seen": 10511848,
      "step": 16235
    },
    {
      "epoch": 9.575471698113208,
      "grad_norm": 4.488799095153809,
      "learning_rate": 6.230866207045756e-06,
      "loss": 0.3641,
      "num_input_tokens_seen": 10515016,
      "step": 16240
    },
    {
      "epoch": 9.578419811320755,
      "grad_norm": 4.02508544921875,
      "learning_rate": 6.228372503214853e-06,
      "loss": 0.429,
      "num_input_tokens_seen": 10517864,
      "step": 16245
    },
    {
      "epoch": 9.581367924528301,
      "grad_norm": 11.707338333129883,
      "learning_rate": 6.225878474167035e-06,
      "loss": 0.343,
      "num_input_tokens_seen": 10520840,
      "step": 16250
    },
    {
      "epoch": 9.58431603773585,
      "grad_norm": 2.887800931930542,
      "learning_rate": 6.22338412056261e-06,
      "loss": 0.3593,
      "num_input_tokens_seen": 10523464,
      "step": 16255
    },
    {
      "epoch": 9.587264150943396,
      "grad_norm": 2.6402618885040283,
      "learning_rate": 6.220889443061966e-06,
      "loss": 0.2928,
      "num_input_tokens_seen": 10527560,
      "step": 16260
    },
    {
      "epoch": 9.590212264150944,
      "grad_norm": 4.12689733505249,
      "learning_rate": 6.2183944423255796e-06,
      "loss": 0.3864,
      "num_input_tokens_seen": 10530088,
      "step": 16265
    },
    {
      "epoch": 9.59316037735849,
      "grad_norm": 2.1398379802703857,
      "learning_rate": 6.215899119014015e-06,
      "loss": 0.5339,
      "num_input_tokens_seen": 10533448,
      "step": 16270
    },
    {
      "epoch": 9.596108490566039,
      "grad_norm": 2.064584255218506,
      "learning_rate": 6.2134034737879175e-06,
      "loss": 0.4003,
      "num_input_tokens_seen": 10538216,
      "step": 16275
    },
    {
      "epoch": 9.599056603773585,
      "grad_norm": 2.198011636734009,
      "learning_rate": 6.2109075073080205e-06,
      "loss": 0.449,
      "num_input_tokens_seen": 10541160,
      "step": 16280
    },
    {
      "epoch": 9.602004716981131,
      "grad_norm": 2.720289945602417,
      "learning_rate": 6.2084112202351425e-06,
      "loss": 0.4837,
      "num_input_tokens_seen": 10547464,
      "step": 16285
    },
    {
      "epoch": 9.60495283018868,
      "grad_norm": 2.847867012023926,
      "learning_rate": 6.205914613230186e-06,
      "loss": 0.296,
      "num_input_tokens_seen": 10551112,
      "step": 16290
    },
    {
      "epoch": 9.607900943396226,
      "grad_norm": 3.6384494304656982,
      "learning_rate": 6.203417686954138e-06,
      "loss": 0.4362,
      "num_input_tokens_seen": 10553960,
      "step": 16295
    },
    {
      "epoch": 9.610849056603774,
      "grad_norm": 1.84572172164917,
      "learning_rate": 6.2009204420680706e-06,
      "loss": 0.2968,
      "num_input_tokens_seen": 10556744,
      "step": 16300
    },
    {
      "epoch": 9.61379716981132,
      "grad_norm": 2.8546817302703857,
      "learning_rate": 6.198422879233141e-06,
      "loss": 0.3914,
      "num_input_tokens_seen": 10559912,
      "step": 16305
    },
    {
      "epoch": 9.616745283018869,
      "grad_norm": 2.8703057765960693,
      "learning_rate": 6.1959249991105895e-06,
      "loss": 0.3912,
      "num_input_tokens_seen": 10563336,
      "step": 16310
    },
    {
      "epoch": 9.619693396226415,
      "grad_norm": 2.4584319591522217,
      "learning_rate": 6.19342680236174e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 10567944,
      "step": 16315
    },
    {
      "epoch": 9.622641509433961,
      "grad_norm": 2.4372966289520264,
      "learning_rate": 6.190928289648003e-06,
      "loss": 0.3544,
      "num_input_tokens_seen": 10571080,
      "step": 16320
    },
    {
      "epoch": 9.62558962264151,
      "grad_norm": 3.154097318649292,
      "learning_rate": 6.188429461630866e-06,
      "loss": 0.3826,
      "num_input_tokens_seen": 10574536,
      "step": 16325
    },
    {
      "epoch": 9.628537735849056,
      "grad_norm": 2.413832664489746,
      "learning_rate": 6.185930318971909e-06,
      "loss": 0.3524,
      "num_input_tokens_seen": 10577608,
      "step": 16330
    },
    {
      "epoch": 9.631485849056604,
      "grad_norm": 2.9438533782958984,
      "learning_rate": 6.1834308623327885e-06,
      "loss": 0.4023,
      "num_input_tokens_seen": 10581800,
      "step": 16335
    },
    {
      "epoch": 9.63443396226415,
      "grad_norm": 2.7330729961395264,
      "learning_rate": 6.180931092375247e-06,
      "loss": 0.4599,
      "num_input_tokens_seen": 10585288,
      "step": 16340
    },
    {
      "epoch": 9.637382075471699,
      "grad_norm": 3.288693904876709,
      "learning_rate": 6.1784310097611075e-06,
      "loss": 0.3619,
      "num_input_tokens_seen": 10588296,
      "step": 16345
    },
    {
      "epoch": 9.640330188679245,
      "grad_norm": 3.726372480392456,
      "learning_rate": 6.1759306151522815e-06,
      "loss": 0.4138,
      "num_input_tokens_seen": 10591112,
      "step": 16350
    },
    {
      "epoch": 9.643278301886792,
      "grad_norm": 2.7793691158294678,
      "learning_rate": 6.173429909210755e-06,
      "loss": 0.3452,
      "num_input_tokens_seen": 10594472,
      "step": 16355
    },
    {
      "epoch": 9.64622641509434,
      "grad_norm": 3.935121536254883,
      "learning_rate": 6.170928892598606e-06,
      "loss": 0.3203,
      "num_input_tokens_seen": 10598120,
      "step": 16360
    },
    {
      "epoch": 9.649174528301886,
      "grad_norm": 2.223863363265991,
      "learning_rate": 6.168427565977984e-06,
      "loss": 0.4573,
      "num_input_tokens_seen": 10601448,
      "step": 16365
    },
    {
      "epoch": 9.652122641509434,
      "grad_norm": 2.318922996520996,
      "learning_rate": 6.165925930011129e-06,
      "loss": 0.3383,
      "num_input_tokens_seen": 10604616,
      "step": 16370
    },
    {
      "epoch": 9.65507075471698,
      "grad_norm": 1.9677366018295288,
      "learning_rate": 6.163423985360359e-06,
      "loss": 0.3319,
      "num_input_tokens_seen": 10607880,
      "step": 16375
    },
    {
      "epoch": 9.658018867924529,
      "grad_norm": 1.9093492031097412,
      "learning_rate": 6.160921732688076e-06,
      "loss": 0.3554,
      "num_input_tokens_seen": 10611976,
      "step": 16380
    },
    {
      "epoch": 9.660966981132075,
      "grad_norm": 3.245573043823242,
      "learning_rate": 6.158419172656759e-06,
      "loss": 0.581,
      "num_input_tokens_seen": 10615080,
      "step": 16385
    },
    {
      "epoch": 9.663915094339622,
      "grad_norm": 2.10091233253479,
      "learning_rate": 6.155916305928974e-06,
      "loss": 0.311,
      "num_input_tokens_seen": 10617672,
      "step": 16390
    },
    {
      "epoch": 9.66686320754717,
      "grad_norm": 3.6634411811828613,
      "learning_rate": 6.153413133167366e-06,
      "loss": 0.5337,
      "num_input_tokens_seen": 10621256,
      "step": 16395
    },
    {
      "epoch": 9.669811320754716,
      "grad_norm": 4.1942267417907715,
      "learning_rate": 6.1509096550346596e-06,
      "loss": 0.5261,
      "num_input_tokens_seen": 10624392,
      "step": 16400
    },
    {
      "epoch": 9.672759433962264,
      "grad_norm": 1.925789475440979,
      "learning_rate": 6.148405872193661e-06,
      "loss": 0.3094,
      "num_input_tokens_seen": 10627816,
      "step": 16405
    },
    {
      "epoch": 9.67570754716981,
      "grad_norm": 2.7246458530426025,
      "learning_rate": 6.14590178530726e-06,
      "loss": 0.3404,
      "num_input_tokens_seen": 10632392,
      "step": 16410
    },
    {
      "epoch": 9.678655660377359,
      "grad_norm": 2.5503573417663574,
      "learning_rate": 6.143397395038422e-06,
      "loss": 0.3925,
      "num_input_tokens_seen": 10635304,
      "step": 16415
    },
    {
      "epoch": 9.681603773584905,
      "grad_norm": 2.307769536972046,
      "learning_rate": 6.140892702050196e-06,
      "loss": 0.4985,
      "num_input_tokens_seen": 10638824,
      "step": 16420
    },
    {
      "epoch": 9.684551886792454,
      "grad_norm": 2.14967942237854,
      "learning_rate": 6.138387707005711e-06,
      "loss": 0.308,
      "num_input_tokens_seen": 10642216,
      "step": 16425
    },
    {
      "epoch": 9.6875,
      "grad_norm": 2.352787733078003,
      "learning_rate": 6.135882410568172e-06,
      "loss": 0.4091,
      "num_input_tokens_seen": 10644296,
      "step": 16430
    },
    {
      "epoch": 9.690448113207546,
      "grad_norm": 2.97570538520813,
      "learning_rate": 6.133376813400872e-06,
      "loss": 0.3244,
      "num_input_tokens_seen": 10647464,
      "step": 16435
    },
    {
      "epoch": 9.693396226415095,
      "grad_norm": 2.4866182804107666,
      "learning_rate": 6.130870916167175e-06,
      "loss": 0.3603,
      "num_input_tokens_seen": 10650440,
      "step": 16440
    },
    {
      "epoch": 9.696344339622641,
      "grad_norm": 2.1291606426239014,
      "learning_rate": 6.128364719530528e-06,
      "loss": 0.2936,
      "num_input_tokens_seen": 10653544,
      "step": 16445
    },
    {
      "epoch": 9.69929245283019,
      "grad_norm": 2.499286413192749,
      "learning_rate": 6.125858224154459e-06,
      "loss": 0.3461,
      "num_input_tokens_seen": 10656296,
      "step": 16450
    },
    {
      "epoch": 9.702240566037736,
      "grad_norm": 6.3327765464782715,
      "learning_rate": 6.123351430702576e-06,
      "loss": 0.4026,
      "num_input_tokens_seen": 10659176,
      "step": 16455
    },
    {
      "epoch": 9.705188679245284,
      "grad_norm": 1.4823777675628662,
      "learning_rate": 6.1208443398385575e-06,
      "loss": 0.327,
      "num_input_tokens_seen": 10663240,
      "step": 16460
    },
    {
      "epoch": 9.70813679245283,
      "grad_norm": 1.6525945663452148,
      "learning_rate": 6.118336952226169e-06,
      "loss": 0.3985,
      "num_input_tokens_seen": 10666792,
      "step": 16465
    },
    {
      "epoch": 9.711084905660378,
      "grad_norm": 1.730012059211731,
      "learning_rate": 6.115829268529254e-06,
      "loss": 0.3374,
      "num_input_tokens_seen": 10671816,
      "step": 16470
    },
    {
      "epoch": 9.714033018867925,
      "grad_norm": 2.6013054847717285,
      "learning_rate": 6.1133212894117326e-06,
      "loss": 0.3955,
      "num_input_tokens_seen": 10674568,
      "step": 16475
    },
    {
      "epoch": 9.716981132075471,
      "grad_norm": 2.5503389835357666,
      "learning_rate": 6.1108130155375986e-06,
      "loss": 0.3493,
      "num_input_tokens_seen": 10677320,
      "step": 16480
    },
    {
      "epoch": 9.71992924528302,
      "grad_norm": 5.342703819274902,
      "learning_rate": 6.108304447570933e-06,
      "loss": 0.4612,
      "num_input_tokens_seen": 10680296,
      "step": 16485
    },
    {
      "epoch": 9.722877358490566,
      "grad_norm": 1.486319899559021,
      "learning_rate": 6.105795586175888e-06,
      "loss": 0.2818,
      "num_input_tokens_seen": 10683080,
      "step": 16490
    },
    {
      "epoch": 9.725825471698114,
      "grad_norm": 3.296710968017578,
      "learning_rate": 6.1032864320166954e-06,
      "loss": 0.3428,
      "num_input_tokens_seen": 10686760,
      "step": 16495
    },
    {
      "epoch": 9.72877358490566,
      "grad_norm": 1.9673529863357544,
      "learning_rate": 6.100776985757666e-06,
      "loss": 0.3361,
      "num_input_tokens_seen": 10690440,
      "step": 16500
    },
    {
      "epoch": 9.731721698113208,
      "grad_norm": 7.292530059814453,
      "learning_rate": 6.098267248063186e-06,
      "loss": 0.45,
      "num_input_tokens_seen": 10693448,
      "step": 16505
    },
    {
      "epoch": 9.734669811320755,
      "grad_norm": 4.574216842651367,
      "learning_rate": 6.0957572195977165e-06,
      "loss": 0.3261,
      "num_input_tokens_seen": 10697000,
      "step": 16510
    },
    {
      "epoch": 9.737617924528301,
      "grad_norm": 3.3586864471435547,
      "learning_rate": 6.0932469010258025e-06,
      "loss": 0.4763,
      "num_input_tokens_seen": 10700200,
      "step": 16515
    },
    {
      "epoch": 9.74056603773585,
      "grad_norm": 3.566136360168457,
      "learning_rate": 6.0907362930120594e-06,
      "loss": 0.3361,
      "num_input_tokens_seen": 10703048,
      "step": 16520
    },
    {
      "epoch": 9.743514150943396,
      "grad_norm": 2.450424909591675,
      "learning_rate": 6.088225396221181e-06,
      "loss": 0.3949,
      "num_input_tokens_seen": 10706088,
      "step": 16525
    },
    {
      "epoch": 9.746462264150944,
      "grad_norm": 1.8852745294570923,
      "learning_rate": 6.0857142113179415e-06,
      "loss": 0.5288,
      "num_input_tokens_seen": 10710504,
      "step": 16530
    },
    {
      "epoch": 9.74941037735849,
      "grad_norm": 1.8983570337295532,
      "learning_rate": 6.083202738967182e-06,
      "loss": 0.3625,
      "num_input_tokens_seen": 10713544,
      "step": 16535
    },
    {
      "epoch": 9.752358490566039,
      "grad_norm": 3.543867826461792,
      "learning_rate": 6.0806909798338324e-06,
      "loss": 0.3528,
      "num_input_tokens_seen": 10717160,
      "step": 16540
    },
    {
      "epoch": 9.755306603773585,
      "grad_norm": 1.8580281734466553,
      "learning_rate": 6.0781789345828854e-06,
      "loss": 0.3338,
      "num_input_tokens_seen": 10719688,
      "step": 16545
    },
    {
      "epoch": 9.758254716981131,
      "grad_norm": 2.5257129669189453,
      "learning_rate": 6.0756666038794195e-06,
      "loss": 0.3182,
      "num_input_tokens_seen": 10722632,
      "step": 16550
    },
    {
      "epoch": 9.76120283018868,
      "grad_norm": 1.650136113166809,
      "learning_rate": 6.073153988388586e-06,
      "loss": 0.3836,
      "num_input_tokens_seen": 10726760,
      "step": 16555
    },
    {
      "epoch": 9.764150943396226,
      "grad_norm": 2.710526466369629,
      "learning_rate": 6.070641088775608e-06,
      "loss": 0.5114,
      "num_input_tokens_seen": 10730344,
      "step": 16560
    },
    {
      "epoch": 9.767099056603774,
      "grad_norm": 5.248807907104492,
      "learning_rate": 6.068127905705787e-06,
      "loss": 0.3134,
      "num_input_tokens_seen": 10733064,
      "step": 16565
    },
    {
      "epoch": 9.77004716981132,
      "grad_norm": 3.1660163402557373,
      "learning_rate": 6.065614439844501e-06,
      "loss": 0.3934,
      "num_input_tokens_seen": 10736616,
      "step": 16570
    },
    {
      "epoch": 9.772995283018869,
      "grad_norm": 2.438344955444336,
      "learning_rate": 6.063100691857198e-06,
      "loss": 0.3591,
      "num_input_tokens_seen": 10739272,
      "step": 16575
    },
    {
      "epoch": 9.775943396226415,
      "grad_norm": 3.946263551712036,
      "learning_rate": 6.060586662409407e-06,
      "loss": 0.388,
      "num_input_tokens_seen": 10741576,
      "step": 16580
    },
    {
      "epoch": 9.778891509433961,
      "grad_norm": 2.11598539352417,
      "learning_rate": 6.058072352166724e-06,
      "loss": 0.4577,
      "num_input_tokens_seen": 10744808,
      "step": 16585
    },
    {
      "epoch": 9.78183962264151,
      "grad_norm": 2.549365997314453,
      "learning_rate": 6.055557761794826e-06,
      "loss": 0.3459,
      "num_input_tokens_seen": 10747816,
      "step": 16590
    },
    {
      "epoch": 9.784787735849056,
      "grad_norm": 2.4622557163238525,
      "learning_rate": 6.053042891959462e-06,
      "loss": 0.4734,
      "num_input_tokens_seen": 10751400,
      "step": 16595
    },
    {
      "epoch": 9.787735849056604,
      "grad_norm": 2.846245050430298,
      "learning_rate": 6.050527743326455e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 10754696,
      "step": 16600
    },
    {
      "epoch": 9.79068396226415,
      "grad_norm": 8.916686058044434,
      "learning_rate": 6.048012316561699e-06,
      "loss": 0.4572,
      "num_input_tokens_seen": 10758696,
      "step": 16605
    },
    {
      "epoch": 9.793632075471699,
      "grad_norm": 5.349390983581543,
      "learning_rate": 6.045496612331166e-06,
      "loss": 0.3463,
      "num_input_tokens_seen": 10761576,
      "step": 16610
    },
    {
      "epoch": 9.796580188679245,
      "grad_norm": 2.769982099533081,
      "learning_rate": 6.0429806313009e-06,
      "loss": 0.4639,
      "num_input_tokens_seen": 10765160,
      "step": 16615
    },
    {
      "epoch": 9.799528301886792,
      "grad_norm": 3.2604215145111084,
      "learning_rate": 6.040464374137015e-06,
      "loss": 0.4461,
      "num_input_tokens_seen": 10768424,
      "step": 16620
    },
    {
      "epoch": 9.80247641509434,
      "grad_norm": 2.0285611152648926,
      "learning_rate": 6.0379478415057045e-06,
      "loss": 0.3811,
      "num_input_tokens_seen": 10771816,
      "step": 16625
    },
    {
      "epoch": 9.805424528301886,
      "grad_norm": 1.8681470155715942,
      "learning_rate": 6.035431034073228e-06,
      "loss": 0.2829,
      "num_input_tokens_seen": 10775400,
      "step": 16630
    },
    {
      "epoch": 9.808372641509434,
      "grad_norm": 2.1378979682922363,
      "learning_rate": 6.0329139525059235e-06,
      "loss": 0.4017,
      "num_input_tokens_seen": 10779592,
      "step": 16635
    },
    {
      "epoch": 9.81132075471698,
      "grad_norm": 2.0711488723754883,
      "learning_rate": 6.030396597470198e-06,
      "loss": 0.4182,
      "num_input_tokens_seen": 10782920,
      "step": 16640
    },
    {
      "epoch": 9.814268867924529,
      "grad_norm": 2.5985636711120605,
      "learning_rate": 6.027878969632534e-06,
      "loss": 0.3539,
      "num_input_tokens_seen": 10785864,
      "step": 16645
    },
    {
      "epoch": 9.817216981132075,
      "grad_norm": 3.449113368988037,
      "learning_rate": 6.025361069659482e-06,
      "loss": 0.3732,
      "num_input_tokens_seen": 10788456,
      "step": 16650
    },
    {
      "epoch": 9.820165094339622,
      "grad_norm": 2.1299781799316406,
      "learning_rate": 6.022842898217668e-06,
      "loss": 0.4236,
      "num_input_tokens_seen": 10792776,
      "step": 16655
    },
    {
      "epoch": 9.82311320754717,
      "grad_norm": 3.137458562850952,
      "learning_rate": 6.020324455973788e-06,
      "loss": 0.3541,
      "num_input_tokens_seen": 10796296,
      "step": 16660
    },
    {
      "epoch": 9.826061320754716,
      "grad_norm": 3.0334525108337402,
      "learning_rate": 6.017805743594612e-06,
      "loss": 0.3421,
      "num_input_tokens_seen": 10799400,
      "step": 16665
    },
    {
      "epoch": 9.829009433962264,
      "grad_norm": 3.6649458408355713,
      "learning_rate": 6.0152867617469776e-06,
      "loss": 0.4342,
      "num_input_tokens_seen": 10802888,
      "step": 16670
    },
    {
      "epoch": 9.83195754716981,
      "grad_norm": 3.843834400177002,
      "learning_rate": 6.012767511097799e-06,
      "loss": 0.3578,
      "num_input_tokens_seen": 10805384,
      "step": 16675
    },
    {
      "epoch": 9.834905660377359,
      "grad_norm": 2.0193822383880615,
      "learning_rate": 6.010247992314055e-06,
      "loss": 0.322,
      "num_input_tokens_seen": 10808072,
      "step": 16680
    },
    {
      "epoch": 9.837853773584905,
      "grad_norm": 2.5740532875061035,
      "learning_rate": 6.007728206062802e-06,
      "loss": 0.2867,
      "num_input_tokens_seen": 10810536,
      "step": 16685
    },
    {
      "epoch": 9.840801886792454,
      "grad_norm": 1.839475154876709,
      "learning_rate": 6.005208153011163e-06,
      "loss": 0.3894,
      "num_input_tokens_seen": 10813128,
      "step": 16690
    },
    {
      "epoch": 9.84375,
      "grad_norm": 2.2293407917022705,
      "learning_rate": 6.0026878338263335e-06,
      "loss": 0.325,
      "num_input_tokens_seen": 10816424,
      "step": 16695
    },
    {
      "epoch": 9.846698113207546,
      "grad_norm": 3.2292091846466064,
      "learning_rate": 6.000167249175579e-06,
      "loss": 0.3877,
      "num_input_tokens_seen": 10820168,
      "step": 16700
    },
    {
      "epoch": 9.849646226415095,
      "grad_norm": 3.172013282775879,
      "learning_rate": 5.997646399726236e-06,
      "loss": 0.3035,
      "num_input_tokens_seen": 10823144,
      "step": 16705
    },
    {
      "epoch": 9.852594339622641,
      "grad_norm": 2.164475440979004,
      "learning_rate": 5.995125286145707e-06,
      "loss": 0.4226,
      "num_input_tokens_seen": 10826440,
      "step": 16710
    },
    {
      "epoch": 9.85554245283019,
      "grad_norm": 7.639603137969971,
      "learning_rate": 5.99260390910147e-06,
      "loss": 0.5843,
      "num_input_tokens_seen": 10830088,
      "step": 16715
    },
    {
      "epoch": 9.858490566037736,
      "grad_norm": 2.2512223720550537,
      "learning_rate": 5.990082269261071e-06,
      "loss": 0.3404,
      "num_input_tokens_seen": 10833640,
      "step": 16720
    },
    {
      "epoch": 9.861438679245284,
      "grad_norm": 3.6231324672698975,
      "learning_rate": 5.987560367292123e-06,
      "loss": 0.3228,
      "num_input_tokens_seen": 10837352,
      "step": 16725
    },
    {
      "epoch": 9.86438679245283,
      "grad_norm": 1.971828579902649,
      "learning_rate": 5.985038203862313e-06,
      "loss": 0.3881,
      "num_input_tokens_seen": 10840360,
      "step": 16730
    },
    {
      "epoch": 9.867334905660378,
      "grad_norm": 2.5553300380706787,
      "learning_rate": 5.982515779639393e-06,
      "loss": 0.4153,
      "num_input_tokens_seen": 10843368,
      "step": 16735
    },
    {
      "epoch": 9.870283018867925,
      "grad_norm": 2.179497718811035,
      "learning_rate": 5.979993095291186e-06,
      "loss": 0.3575,
      "num_input_tokens_seen": 10846216,
      "step": 16740
    },
    {
      "epoch": 9.873231132075471,
      "grad_norm": 1.7909815311431885,
      "learning_rate": 5.977470151485582e-06,
      "loss": 0.4121,
      "num_input_tokens_seen": 10849736,
      "step": 16745
    },
    {
      "epoch": 9.87617924528302,
      "grad_norm": 3.3012895584106445,
      "learning_rate": 5.974946948890544e-06,
      "loss": 0.3633,
      "num_input_tokens_seen": 10852456,
      "step": 16750
    },
    {
      "epoch": 9.879127358490566,
      "grad_norm": 4.349846839904785,
      "learning_rate": 5.9724234881740994e-06,
      "loss": 0.5125,
      "num_input_tokens_seen": 10855720,
      "step": 16755
    },
    {
      "epoch": 9.882075471698114,
      "grad_norm": 2.8887221813201904,
      "learning_rate": 5.9698997700043445e-06,
      "loss": 0.4128,
      "num_input_tokens_seen": 10858376,
      "step": 16760
    },
    {
      "epoch": 9.88502358490566,
      "grad_norm": 2.2795894145965576,
      "learning_rate": 5.9673757950494475e-06,
      "loss": 0.4571,
      "num_input_tokens_seen": 10861416,
      "step": 16765
    },
    {
      "epoch": 9.887971698113208,
      "grad_norm": 1.7488058805465698,
      "learning_rate": 5.964851563977639e-06,
      "loss": 0.3264,
      "num_input_tokens_seen": 10865160,
      "step": 16770
    },
    {
      "epoch": 9.890919811320755,
      "grad_norm": 1.8388878107070923,
      "learning_rate": 5.962327077457219e-06,
      "loss": 0.423,
      "num_input_tokens_seen": 10868488,
      "step": 16775
    },
    {
      "epoch": 9.893867924528301,
      "grad_norm": 3.243752956390381,
      "learning_rate": 5.959802336156558e-06,
      "loss": 0.3858,
      "num_input_tokens_seen": 10871752,
      "step": 16780
    },
    {
      "epoch": 9.89681603773585,
      "grad_norm": 3.9024620056152344,
      "learning_rate": 5.957277340744094e-06,
      "loss": 0.3921,
      "num_input_tokens_seen": 10875336,
      "step": 16785
    },
    {
      "epoch": 9.899764150943396,
      "grad_norm": 1.6749022006988525,
      "learning_rate": 5.954752091888326e-06,
      "loss": 0.291,
      "num_input_tokens_seen": 10878312,
      "step": 16790
    },
    {
      "epoch": 9.902712264150944,
      "grad_norm": 2.3935954570770264,
      "learning_rate": 5.952226590257829e-06,
      "loss": 0.4922,
      "num_input_tokens_seen": 10882056,
      "step": 16795
    },
    {
      "epoch": 9.90566037735849,
      "grad_norm": 2.035158395767212,
      "learning_rate": 5.949700836521239e-06,
      "loss": 0.3331,
      "num_input_tokens_seen": 10885960,
      "step": 16800
    },
    {
      "epoch": 9.908608490566039,
      "grad_norm": 2.4384617805480957,
      "learning_rate": 5.947174831347257e-06,
      "loss": 0.3956,
      "num_input_tokens_seen": 10888456,
      "step": 16805
    },
    {
      "epoch": 9.911556603773585,
      "grad_norm": 4.549461841583252,
      "learning_rate": 5.944648575404657e-06,
      "loss": 0.3689,
      "num_input_tokens_seen": 10891432,
      "step": 16810
    },
    {
      "epoch": 9.914504716981131,
      "grad_norm": 7.186270713806152,
      "learning_rate": 5.942122069362276e-06,
      "loss": 0.4656,
      "num_input_tokens_seen": 10893800,
      "step": 16815
    },
    {
      "epoch": 9.91745283018868,
      "grad_norm": 2.1963412761688232,
      "learning_rate": 5.939595313889016e-06,
      "loss": 0.3012,
      "num_input_tokens_seen": 10897896,
      "step": 16820
    },
    {
      "epoch": 9.920400943396226,
      "grad_norm": 1.5590721368789673,
      "learning_rate": 5.937068309653848e-06,
      "loss": 0.3632,
      "num_input_tokens_seen": 10901576,
      "step": 16825
    },
    {
      "epoch": 9.923349056603774,
      "grad_norm": 3.1455917358398438,
      "learning_rate": 5.934541057325807e-06,
      "loss": 0.2995,
      "num_input_tokens_seen": 10904840,
      "step": 16830
    },
    {
      "epoch": 9.92629716981132,
      "grad_norm": 2.5863983631134033,
      "learning_rate": 5.932013557573992e-06,
      "loss": 0.3688,
      "num_input_tokens_seen": 10908872,
      "step": 16835
    },
    {
      "epoch": 9.929245283018869,
      "grad_norm": 4.338284969329834,
      "learning_rate": 5.929485811067572e-06,
      "loss": 0.3488,
      "num_input_tokens_seen": 10913512,
      "step": 16840
    },
    {
      "epoch": 9.932193396226415,
      "grad_norm": 3.3921635150909424,
      "learning_rate": 5.926957818475778e-06,
      "loss": 0.5719,
      "num_input_tokens_seen": 10921320,
      "step": 16845
    },
    {
      "epoch": 9.935141509433961,
      "grad_norm": 2.933530807495117,
      "learning_rate": 5.924429580467905e-06,
      "loss": 0.3337,
      "num_input_tokens_seen": 10923944,
      "step": 16850
    },
    {
      "epoch": 9.93808962264151,
      "grad_norm": 5.7248854637146,
      "learning_rate": 5.921901097713317e-06,
      "loss": 0.4585,
      "num_input_tokens_seen": 10926600,
      "step": 16855
    },
    {
      "epoch": 9.941037735849056,
      "grad_norm": 2.889597177505493,
      "learning_rate": 5.919372370881442e-06,
      "loss": 0.4774,
      "num_input_tokens_seen": 10929832,
      "step": 16860
    },
    {
      "epoch": 9.943985849056604,
      "grad_norm": 2.2742855548858643,
      "learning_rate": 5.916843400641768e-06,
      "loss": 0.347,
      "num_input_tokens_seen": 10932936,
      "step": 16865
    },
    {
      "epoch": 9.94693396226415,
      "grad_norm": 2.562758207321167,
      "learning_rate": 5.914314187663851e-06,
      "loss": 0.3522,
      "num_input_tokens_seen": 10936488,
      "step": 16870
    },
    {
      "epoch": 9.949882075471699,
      "grad_norm": 3.090973138809204,
      "learning_rate": 5.911784732617314e-06,
      "loss": 0.5616,
      "num_input_tokens_seen": 10939080,
      "step": 16875
    },
    {
      "epoch": 9.952830188679245,
      "grad_norm": 2.342137098312378,
      "learning_rate": 5.90925503617184e-06,
      "loss": 0.3074,
      "num_input_tokens_seen": 10942440,
      "step": 16880
    },
    {
      "epoch": 9.955778301886792,
      "grad_norm": 4.709895610809326,
      "learning_rate": 5.9067250989971745e-06,
      "loss": 0.5481,
      "num_input_tokens_seen": 10945832,
      "step": 16885
    },
    {
      "epoch": 9.95872641509434,
      "grad_norm": 1.453339695930481,
      "learning_rate": 5.904194921763133e-06,
      "loss": 0.4649,
      "num_input_tokens_seen": 10949000,
      "step": 16890
    },
    {
      "epoch": 9.961674528301886,
      "grad_norm": 1.7355375289916992,
      "learning_rate": 5.901664505139589e-06,
      "loss": 0.3862,
      "num_input_tokens_seen": 10951688,
      "step": 16895
    },
    {
      "epoch": 9.964622641509434,
      "grad_norm": 2.261704206466675,
      "learning_rate": 5.8991338497964814e-06,
      "loss": 0.452,
      "num_input_tokens_seen": 10954632,
      "step": 16900
    },
    {
      "epoch": 9.96757075471698,
      "grad_norm": 2.5373144149780273,
      "learning_rate": 5.896602956403812e-06,
      "loss": 0.4248,
      "num_input_tokens_seen": 10958536,
      "step": 16905
    },
    {
      "epoch": 9.970518867924529,
      "grad_norm": 2.4781336784362793,
      "learning_rate": 5.894071825631645e-06,
      "loss": 0.4788,
      "num_input_tokens_seen": 10962024,
      "step": 16910
    },
    {
      "epoch": 9.973466981132075,
      "grad_norm": 5.066862106323242,
      "learning_rate": 5.891540458150109e-06,
      "loss": 0.3506,
      "num_input_tokens_seen": 10965096,
      "step": 16915
    },
    {
      "epoch": 9.976415094339622,
      "grad_norm": 4.052128791809082,
      "learning_rate": 5.889008854629395e-06,
      "loss": 0.3975,
      "num_input_tokens_seen": 10967496,
      "step": 16920
    },
    {
      "epoch": 9.97936320754717,
      "grad_norm": 3.6323788166046143,
      "learning_rate": 5.886477015739754e-06,
      "loss": 0.3775,
      "num_input_tokens_seen": 10970728,
      "step": 16925
    },
    {
      "epoch": 9.982311320754716,
      "grad_norm": 3.1155271530151367,
      "learning_rate": 5.883944942151502e-06,
      "loss": 0.389,
      "num_input_tokens_seen": 10974216,
      "step": 16930
    },
    {
      "epoch": 9.985259433962264,
      "grad_norm": 2.793856620788574,
      "learning_rate": 5.88141263453502e-06,
      "loss": 0.295,
      "num_input_tokens_seen": 10980008,
      "step": 16935
    },
    {
      "epoch": 9.98820754716981,
      "grad_norm": 2.6110284328460693,
      "learning_rate": 5.878880093560744e-06,
      "loss": 0.3679,
      "num_input_tokens_seen": 10982728,
      "step": 16940
    },
    {
      "epoch": 9.991155660377359,
      "grad_norm": 2.930337429046631,
      "learning_rate": 5.876347319899173e-06,
      "loss": 0.293,
      "num_input_tokens_seen": 10986184,
      "step": 16945
    },
    {
      "epoch": 9.994103773584905,
      "grad_norm": 1.7575063705444336,
      "learning_rate": 5.873814314220874e-06,
      "loss": 0.2899,
      "num_input_tokens_seen": 10989320,
      "step": 16950
    },
    {
      "epoch": 9.997051886792454,
      "grad_norm": 3.3234057426452637,
      "learning_rate": 5.871281077196469e-06,
      "loss": 0.4297,
      "num_input_tokens_seen": 10992488,
      "step": 16955
    },
    {
      "epoch": 10.0,
      "grad_norm": 18.91790008544922,
      "learning_rate": 5.868747609496643e-06,
      "loss": 0.2648,
      "num_input_tokens_seen": 10994984,
      "step": 16960
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.5319322347640991,
      "eval_runtime": 19.438,
      "eval_samples_per_second": 87.252,
      "eval_steps_per_second": 21.813,
      "num_input_tokens_seen": 10994984,
      "step": 16960
    },
    {
      "epoch": 10.002948113207546,
      "grad_norm": 2.3976619243621826,
      "learning_rate": 5.8662139117921435e-06,
      "loss": 0.2777,
      "num_input_tokens_seen": 10998760,
      "step": 16965
    },
    {
      "epoch": 10.005896226415095,
      "grad_norm": 2.58954119682312,
      "learning_rate": 5.8636799847537785e-06,
      "loss": 0.2574,
      "num_input_tokens_seen": 11002728,
      "step": 16970
    },
    {
      "epoch": 10.008844339622641,
      "grad_norm": 1.6184606552124023,
      "learning_rate": 5.861145829052415e-06,
      "loss": 0.2967,
      "num_input_tokens_seen": 11006696,
      "step": 16975
    },
    {
      "epoch": 10.01179245283019,
      "grad_norm": 2.4136319160461426,
      "learning_rate": 5.858611445358982e-06,
      "loss": 0.3084,
      "num_input_tokens_seen": 11010088,
      "step": 16980
    },
    {
      "epoch": 10.014740566037736,
      "grad_norm": 2.4596643447875977,
      "learning_rate": 5.856076834344468e-06,
      "loss": 0.3399,
      "num_input_tokens_seen": 11013256,
      "step": 16985
    },
    {
      "epoch": 10.017688679245284,
      "grad_norm": 3.105794668197632,
      "learning_rate": 5.853541996679924e-06,
      "loss": 0.3853,
      "num_input_tokens_seen": 11015912,
      "step": 16990
    },
    {
      "epoch": 10.02063679245283,
      "grad_norm": 4.218655109405518,
      "learning_rate": 5.851006933036456e-06,
      "loss": 0.3712,
      "num_input_tokens_seen": 11018120,
      "step": 16995
    },
    {
      "epoch": 10.023584905660377,
      "grad_norm": 3.921227216720581,
      "learning_rate": 5.848471644085236e-06,
      "loss": 0.4116,
      "num_input_tokens_seen": 11021864,
      "step": 17000
    },
    {
      "epoch": 10.026533018867925,
      "grad_norm": 2.621786117553711,
      "learning_rate": 5.84593613049749e-06,
      "loss": 0.2683,
      "num_input_tokens_seen": 11024744,
      "step": 17005
    },
    {
      "epoch": 10.029481132075471,
      "grad_norm": 5.432643890380859,
      "learning_rate": 5.843400392944509e-06,
      "loss": 0.4608,
      "num_input_tokens_seen": 11028104,
      "step": 17010
    },
    {
      "epoch": 10.03242924528302,
      "grad_norm": 3.239006519317627,
      "learning_rate": 5.840864432097639e-06,
      "loss": 0.4455,
      "num_input_tokens_seen": 11031336,
      "step": 17015
    },
    {
      "epoch": 10.035377358490566,
      "grad_norm": 2.492116928100586,
      "learning_rate": 5.838328248628286e-06,
      "loss": 0.3237,
      "num_input_tokens_seen": 11034568,
      "step": 17020
    },
    {
      "epoch": 10.038325471698114,
      "grad_norm": 2.7857556343078613,
      "learning_rate": 5.835791843207916e-06,
      "loss": 0.4046,
      "num_input_tokens_seen": 11037800,
      "step": 17025
    },
    {
      "epoch": 10.04127358490566,
      "grad_norm": 2.594856023788452,
      "learning_rate": 5.833255216508056e-06,
      "loss": 0.373,
      "num_input_tokens_seen": 11040776,
      "step": 17030
    },
    {
      "epoch": 10.044221698113208,
      "grad_norm": 2.9771475791931152,
      "learning_rate": 5.830718369200284e-06,
      "loss": 0.3354,
      "num_input_tokens_seen": 11044200,
      "step": 17035
    },
    {
      "epoch": 10.047169811320755,
      "grad_norm": 3.2946836948394775,
      "learning_rate": 5.828181301956244e-06,
      "loss": 0.2445,
      "num_input_tokens_seen": 11047688,
      "step": 17040
    },
    {
      "epoch": 10.050117924528301,
      "grad_norm": 2.410019636154175,
      "learning_rate": 5.825644015447636e-06,
      "loss": 0.4787,
      "num_input_tokens_seen": 11051112,
      "step": 17045
    },
    {
      "epoch": 10.05306603773585,
      "grad_norm": 1.9521819353103638,
      "learning_rate": 5.823106510346216e-06,
      "loss": 0.3661,
      "num_input_tokens_seen": 11054824,
      "step": 17050
    },
    {
      "epoch": 10.056014150943396,
      "grad_norm": 2.162914991378784,
      "learning_rate": 5.820568787323798e-06,
      "loss": 0.2654,
      "num_input_tokens_seen": 11057992,
      "step": 17055
    },
    {
      "epoch": 10.058962264150944,
      "grad_norm": 2.2270052433013916,
      "learning_rate": 5.818030847052258e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 11061640,
      "step": 17060
    },
    {
      "epoch": 10.06191037735849,
      "grad_norm": 3.3547723293304443,
      "learning_rate": 5.815492690203528e-06,
      "loss": 0.3456,
      "num_input_tokens_seen": 11064936,
      "step": 17065
    },
    {
      "epoch": 10.064858490566039,
      "grad_norm": 3.1068923473358154,
      "learning_rate": 5.812954317449591e-06,
      "loss": 0.2569,
      "num_input_tokens_seen": 11067688,
      "step": 17070
    },
    {
      "epoch": 10.067806603773585,
      "grad_norm": 2.4098923206329346,
      "learning_rate": 5.810415729462496e-06,
      "loss": 0.379,
      "num_input_tokens_seen": 11071336,
      "step": 17075
    },
    {
      "epoch": 10.070754716981131,
      "grad_norm": 4.022672653198242,
      "learning_rate": 5.807876926914344e-06,
      "loss": 0.3339,
      "num_input_tokens_seen": 11073736,
      "step": 17080
    },
    {
      "epoch": 10.07370283018868,
      "grad_norm": 2.6237592697143555,
      "learning_rate": 5.8053379104772935e-06,
      "loss": 0.3464,
      "num_input_tokens_seen": 11076968,
      "step": 17085
    },
    {
      "epoch": 10.076650943396226,
      "grad_norm": 2.1439268589019775,
      "learning_rate": 5.802798680823562e-06,
      "loss": 0.2298,
      "num_input_tokens_seen": 11079720,
      "step": 17090
    },
    {
      "epoch": 10.079599056603774,
      "grad_norm": 2.1471965312957764,
      "learning_rate": 5.8002592386254186e-06,
      "loss": 0.3715,
      "num_input_tokens_seen": 11082376,
      "step": 17095
    },
    {
      "epoch": 10.08254716981132,
      "grad_norm": 2.215940475463867,
      "learning_rate": 5.797719584555192e-06,
      "loss": 0.2342,
      "num_input_tokens_seen": 11086024,
      "step": 17100
    },
    {
      "epoch": 10.085495283018869,
      "grad_norm": 4.280333042144775,
      "learning_rate": 5.795179719285269e-06,
      "loss": 0.3659,
      "num_input_tokens_seen": 11088648,
      "step": 17105
    },
    {
      "epoch": 10.088443396226415,
      "grad_norm": 1.8839508295059204,
      "learning_rate": 5.792639643488086e-06,
      "loss": 0.2774,
      "num_input_tokens_seen": 11091976,
      "step": 17110
    },
    {
      "epoch": 10.091391509433961,
      "grad_norm": 11.447282791137695,
      "learning_rate": 5.7900993578361434e-06,
      "loss": 0.3521,
      "num_input_tokens_seen": 11094376,
      "step": 17115
    },
    {
      "epoch": 10.09433962264151,
      "grad_norm": 4.046978950500488,
      "learning_rate": 5.7875588630019895e-06,
      "loss": 0.5225,
      "num_input_tokens_seen": 11098088,
      "step": 17120
    },
    {
      "epoch": 10.097287735849056,
      "grad_norm": 2.053182601928711,
      "learning_rate": 5.7850181596582335e-06,
      "loss": 0.3393,
      "num_input_tokens_seen": 11100840,
      "step": 17125
    },
    {
      "epoch": 10.100235849056604,
      "grad_norm": 1.9237629175186157,
      "learning_rate": 5.782477248477535e-06,
      "loss": 0.3949,
      "num_input_tokens_seen": 11104200,
      "step": 17130
    },
    {
      "epoch": 10.10318396226415,
      "grad_norm": 2.1432669162750244,
      "learning_rate": 5.779936130132614e-06,
      "loss": 0.4017,
      "num_input_tokens_seen": 11107368,
      "step": 17135
    },
    {
      "epoch": 10.106132075471699,
      "grad_norm": 2.032250165939331,
      "learning_rate": 5.777394805296242e-06,
      "loss": 0.3246,
      "num_input_tokens_seen": 11109960,
      "step": 17140
    },
    {
      "epoch": 10.109080188679245,
      "grad_norm": 3.332411766052246,
      "learning_rate": 5.774853274641243e-06,
      "loss": 0.4415,
      "num_input_tokens_seen": 11113608,
      "step": 17145
    },
    {
      "epoch": 10.112028301886792,
      "grad_norm": 2.86710262298584,
      "learning_rate": 5.772311538840501e-06,
      "loss": 0.3312,
      "num_input_tokens_seen": 11116776,
      "step": 17150
    },
    {
      "epoch": 10.11497641509434,
      "grad_norm": 1.7534639835357666,
      "learning_rate": 5.76976959856695e-06,
      "loss": 0.384,
      "num_input_tokens_seen": 11119944,
      "step": 17155
    },
    {
      "epoch": 10.117924528301886,
      "grad_norm": 3.2390971183776855,
      "learning_rate": 5.767227454493582e-06,
      "loss": 0.4357,
      "num_input_tokens_seen": 11122952,
      "step": 17160
    },
    {
      "epoch": 10.120872641509434,
      "grad_norm": 2.529688596725464,
      "learning_rate": 5.764685107293436e-06,
      "loss": 0.3142,
      "num_input_tokens_seen": 11126120,
      "step": 17165
    },
    {
      "epoch": 10.12382075471698,
      "grad_norm": 2.447035789489746,
      "learning_rate": 5.762142557639614e-06,
      "loss": 0.4302,
      "num_input_tokens_seen": 11129608,
      "step": 17170
    },
    {
      "epoch": 10.126768867924529,
      "grad_norm": 2.554415464401245,
      "learning_rate": 5.759599806205266e-06,
      "loss": 0.2392,
      "num_input_tokens_seen": 11132072,
      "step": 17175
    },
    {
      "epoch": 10.129716981132075,
      "grad_norm": 2.55080509185791,
      "learning_rate": 5.757056853663594e-06,
      "loss": 0.3819,
      "num_input_tokens_seen": 11135816,
      "step": 17180
    },
    {
      "epoch": 10.132665094339623,
      "grad_norm": 2.291837453842163,
      "learning_rate": 5.754513700687858e-06,
      "loss": 0.402,
      "num_input_tokens_seen": 11138440,
      "step": 17185
    },
    {
      "epoch": 10.13561320754717,
      "grad_norm": 5.713946342468262,
      "learning_rate": 5.751970347951369e-06,
      "loss": 0.4642,
      "num_input_tokens_seen": 11141736,
      "step": 17190
    },
    {
      "epoch": 10.138561320754716,
      "grad_norm": 4.290987968444824,
      "learning_rate": 5.7494267961274875e-06,
      "loss": 0.3144,
      "num_input_tokens_seen": 11144040,
      "step": 17195
    },
    {
      "epoch": 10.141509433962264,
      "grad_norm": 2.3957364559173584,
      "learning_rate": 5.746883045889633e-06,
      "loss": 0.3178,
      "num_input_tokens_seen": 11147080,
      "step": 17200
    },
    {
      "epoch": 10.14445754716981,
      "grad_norm": 4.310591220855713,
      "learning_rate": 5.744339097911272e-06,
      "loss": 0.4461,
      "num_input_tokens_seen": 11150216,
      "step": 17205
    },
    {
      "epoch": 10.147405660377359,
      "grad_norm": 2.4974279403686523,
      "learning_rate": 5.741794952865928e-06,
      "loss": 0.4524,
      "num_input_tokens_seen": 11153256,
      "step": 17210
    },
    {
      "epoch": 10.150353773584905,
      "grad_norm": 3.1563196182250977,
      "learning_rate": 5.739250611427173e-06,
      "loss": 0.4415,
      "num_input_tokens_seen": 11156840,
      "step": 17215
    },
    {
      "epoch": 10.153301886792454,
      "grad_norm": 1.2801347970962524,
      "learning_rate": 5.736706074268633e-06,
      "loss": 0.2525,
      "num_input_tokens_seen": 11160104,
      "step": 17220
    },
    {
      "epoch": 10.15625,
      "grad_norm": 2.557077646255493,
      "learning_rate": 5.734161342063984e-06,
      "loss": 0.2876,
      "num_input_tokens_seen": 11163432,
      "step": 17225
    },
    {
      "epoch": 10.159198113207546,
      "grad_norm": 3.0521738529205322,
      "learning_rate": 5.731616415486957e-06,
      "loss": 0.5465,
      "num_input_tokens_seen": 11166600,
      "step": 17230
    },
    {
      "epoch": 10.162146226415095,
      "grad_norm": 2.106093168258667,
      "learning_rate": 5.72907129521133e-06,
      "loss": 0.3955,
      "num_input_tokens_seen": 11170376,
      "step": 17235
    },
    {
      "epoch": 10.165094339622641,
      "grad_norm": 2.5867292881011963,
      "learning_rate": 5.726525981910935e-06,
      "loss": 0.4733,
      "num_input_tokens_seen": 11174088,
      "step": 17240
    },
    {
      "epoch": 10.16804245283019,
      "grad_norm": 1.757140874862671,
      "learning_rate": 5.723980476259658e-06,
      "loss": 0.3206,
      "num_input_tokens_seen": 11178056,
      "step": 17245
    },
    {
      "epoch": 10.170990566037736,
      "grad_norm": 4.324456691741943,
      "learning_rate": 5.721434778931426e-06,
      "loss": 0.2505,
      "num_input_tokens_seen": 11182760,
      "step": 17250
    },
    {
      "epoch": 10.173938679245284,
      "grad_norm": 3.4334652423858643,
      "learning_rate": 5.7188888906002284e-06,
      "loss": 0.3829,
      "num_input_tokens_seen": 11185640,
      "step": 17255
    },
    {
      "epoch": 10.17688679245283,
      "grad_norm": 3.1583480834960938,
      "learning_rate": 5.716342811940098e-06,
      "loss": 0.3257,
      "num_input_tokens_seen": 11188584,
      "step": 17260
    },
    {
      "epoch": 10.179834905660377,
      "grad_norm": 3.157897710800171,
      "learning_rate": 5.713796543625123e-06,
      "loss": 0.5367,
      "num_input_tokens_seen": 11191560,
      "step": 17265
    },
    {
      "epoch": 10.182783018867925,
      "grad_norm": 2.4658913612365723,
      "learning_rate": 5.711250086329435e-06,
      "loss": 0.3579,
      "num_input_tokens_seen": 11194856,
      "step": 17270
    },
    {
      "epoch": 10.185731132075471,
      "grad_norm": 2.952787160873413,
      "learning_rate": 5.7087034407272225e-06,
      "loss": 0.3029,
      "num_input_tokens_seen": 11197672,
      "step": 17275
    },
    {
      "epoch": 10.18867924528302,
      "grad_norm": 1.9605873823165894,
      "learning_rate": 5.70615660749272e-06,
      "loss": 0.332,
      "num_input_tokens_seen": 11200136,
      "step": 17280
    },
    {
      "epoch": 10.191627358490566,
      "grad_norm": 2.2364823818206787,
      "learning_rate": 5.7036095873002106e-06,
      "loss": 0.3448,
      "num_input_tokens_seen": 11203432,
      "step": 17285
    },
    {
      "epoch": 10.194575471698114,
      "grad_norm": 4.518283367156982,
      "learning_rate": 5.701062380824032e-06,
      "loss": 0.3217,
      "num_input_tokens_seen": 11206984,
      "step": 17290
    },
    {
      "epoch": 10.19752358490566,
      "grad_norm": 2.650974750518799,
      "learning_rate": 5.698514988738566e-06,
      "loss": 0.4998,
      "num_input_tokens_seen": 11209800,
      "step": 17295
    },
    {
      "epoch": 10.200471698113208,
      "grad_norm": 2.4174540042877197,
      "learning_rate": 5.6959674117182465e-06,
      "loss": 0.4106,
      "num_input_tokens_seen": 11212648,
      "step": 17300
    },
    {
      "epoch": 10.203419811320755,
      "grad_norm": 2.3372204303741455,
      "learning_rate": 5.693419650437554e-06,
      "loss": 0.4094,
      "num_input_tokens_seen": 11215080,
      "step": 17305
    },
    {
      "epoch": 10.206367924528301,
      "grad_norm": 3.549341917037964,
      "learning_rate": 5.690871705571022e-06,
      "loss": 0.3148,
      "num_input_tokens_seen": 11217992,
      "step": 17310
    },
    {
      "epoch": 10.20931603773585,
      "grad_norm": 1.96096670627594,
      "learning_rate": 5.688323577793229e-06,
      "loss": 0.4145,
      "num_input_tokens_seen": 11222952,
      "step": 17315
    },
    {
      "epoch": 10.212264150943396,
      "grad_norm": 9.781682014465332,
      "learning_rate": 5.685775267778801e-06,
      "loss": 0.3894,
      "num_input_tokens_seen": 11226024,
      "step": 17320
    },
    {
      "epoch": 10.215212264150944,
      "grad_norm": 2.435809373855591,
      "learning_rate": 5.683226776202416e-06,
      "loss": 0.3289,
      "num_input_tokens_seen": 11228968,
      "step": 17325
    },
    {
      "epoch": 10.21816037735849,
      "grad_norm": 3.2173030376434326,
      "learning_rate": 5.680678103738798e-06,
      "loss": 0.3519,
      "num_input_tokens_seen": 11231720,
      "step": 17330
    },
    {
      "epoch": 10.221108490566039,
      "grad_norm": 2.6525421142578125,
      "learning_rate": 5.678129251062717e-06,
      "loss": 0.4507,
      "num_input_tokens_seen": 11234600,
      "step": 17335
    },
    {
      "epoch": 10.224056603773585,
      "grad_norm": 2.3942363262176514,
      "learning_rate": 5.675580218848995e-06,
      "loss": 0.3397,
      "num_input_tokens_seen": 11237384,
      "step": 17340
    },
    {
      "epoch": 10.227004716981131,
      "grad_norm": 2.4219741821289062,
      "learning_rate": 5.673031007772498e-06,
      "loss": 0.4921,
      "num_input_tokens_seen": 11240680,
      "step": 17345
    },
    {
      "epoch": 10.22995283018868,
      "grad_norm": 2.417099714279175,
      "learning_rate": 5.670481618508141e-06,
      "loss": 0.3464,
      "num_input_tokens_seen": 11243624,
      "step": 17350
    },
    {
      "epoch": 10.232900943396226,
      "grad_norm": 3.93424391746521,
      "learning_rate": 5.667932051730887e-06,
      "loss": 0.4171,
      "num_input_tokens_seen": 11246856,
      "step": 17355
    },
    {
      "epoch": 10.235849056603774,
      "grad_norm": 6.446231365203857,
      "learning_rate": 5.6653823081157434e-06,
      "loss": 0.4595,
      "num_input_tokens_seen": 11250376,
      "step": 17360
    },
    {
      "epoch": 10.23879716981132,
      "grad_norm": 2.9331984519958496,
      "learning_rate": 5.662832388337766e-06,
      "loss": 0.3654,
      "num_input_tokens_seen": 11253608,
      "step": 17365
    },
    {
      "epoch": 10.241745283018869,
      "grad_norm": 1.897558331489563,
      "learning_rate": 5.660282293072057e-06,
      "loss": 0.3051,
      "num_input_tokens_seen": 11256872,
      "step": 17370
    },
    {
      "epoch": 10.244693396226415,
      "grad_norm": 3.3896915912628174,
      "learning_rate": 5.657732022993765e-06,
      "loss": 0.3015,
      "num_input_tokens_seen": 11260520,
      "step": 17375
    },
    {
      "epoch": 10.247641509433961,
      "grad_norm": 3.5362935066223145,
      "learning_rate": 5.655181578778085e-06,
      "loss": 0.3589,
      "num_input_tokens_seen": 11263464,
      "step": 17380
    },
    {
      "epoch": 10.25058962264151,
      "grad_norm": 3.654315710067749,
      "learning_rate": 5.65263096110026e-06,
      "loss": 0.4702,
      "num_input_tokens_seen": 11266600,
      "step": 17385
    },
    {
      "epoch": 10.253537735849056,
      "grad_norm": 3.441154718399048,
      "learning_rate": 5.650080170635573e-06,
      "loss": 0.4048,
      "num_input_tokens_seen": 11269992,
      "step": 17390
    },
    {
      "epoch": 10.256485849056604,
      "grad_norm": 3.3018686771392822,
      "learning_rate": 5.647529208059359e-06,
      "loss": 0.3687,
      "num_input_tokens_seen": 11272264,
      "step": 17395
    },
    {
      "epoch": 10.25943396226415,
      "grad_norm": 2.8790812492370605,
      "learning_rate": 5.6449780740469985e-06,
      "loss": 0.3894,
      "num_input_tokens_seen": 11276840,
      "step": 17400
    },
    {
      "epoch": 10.262382075471699,
      "grad_norm": 2.3498823642730713,
      "learning_rate": 5.642426769273912e-06,
      "loss": 0.3012,
      "num_input_tokens_seen": 11280232,
      "step": 17405
    },
    {
      "epoch": 10.265330188679245,
      "grad_norm": 3.363157272338867,
      "learning_rate": 5.63987529441557e-06,
      "loss": 0.3205,
      "num_input_tokens_seen": 11283048,
      "step": 17410
    },
    {
      "epoch": 10.268278301886792,
      "grad_norm": 2.108511209487915,
      "learning_rate": 5.637323650147487e-06,
      "loss": 0.386,
      "num_input_tokens_seen": 11286504,
      "step": 17415
    },
    {
      "epoch": 10.27122641509434,
      "grad_norm": 2.718304395675659,
      "learning_rate": 5.63477183714522e-06,
      "loss": 0.4439,
      "num_input_tokens_seen": 11289352,
      "step": 17420
    },
    {
      "epoch": 10.274174528301886,
      "grad_norm": 2.706268310546875,
      "learning_rate": 5.632219856084373e-06,
      "loss": 0.478,
      "num_input_tokens_seen": 11292232,
      "step": 17425
    },
    {
      "epoch": 10.277122641509434,
      "grad_norm": 3.5479607582092285,
      "learning_rate": 5.6296677076405944e-06,
      "loss": 0.4129,
      "num_input_tokens_seen": 11296424,
      "step": 17430
    },
    {
      "epoch": 10.28007075471698,
      "grad_norm": 3.5629405975341797,
      "learning_rate": 5.627115392489578e-06,
      "loss": 0.3446,
      "num_input_tokens_seen": 11299816,
      "step": 17435
    },
    {
      "epoch": 10.283018867924529,
      "grad_norm": 1.8996363878250122,
      "learning_rate": 5.624562911307058e-06,
      "loss": 0.4149,
      "num_input_tokens_seen": 11306696,
      "step": 17440
    },
    {
      "epoch": 10.285966981132075,
      "grad_norm": 4.50596284866333,
      "learning_rate": 5.622010264768813e-06,
      "loss": 0.3384,
      "num_input_tokens_seen": 11308904,
      "step": 17445
    },
    {
      "epoch": 10.288915094339623,
      "grad_norm": 2.07731556892395,
      "learning_rate": 5.619457453550673e-06,
      "loss": 0.3812,
      "num_input_tokens_seen": 11312264,
      "step": 17450
    },
    {
      "epoch": 10.29186320754717,
      "grad_norm": 3.235184669494629,
      "learning_rate": 5.616904478328502e-06,
      "loss": 0.4531,
      "num_input_tokens_seen": 11315656,
      "step": 17455
    },
    {
      "epoch": 10.294811320754716,
      "grad_norm": 2.632340908050537,
      "learning_rate": 5.6143513397782104e-06,
      "loss": 0.33,
      "num_input_tokens_seen": 11318536,
      "step": 17460
    },
    {
      "epoch": 10.297759433962264,
      "grad_norm": 1.7120022773742676,
      "learning_rate": 5.611798038575755e-06,
      "loss": 0.3115,
      "num_input_tokens_seen": 11321544,
      "step": 17465
    },
    {
      "epoch": 10.30070754716981,
      "grad_norm": 4.436554431915283,
      "learning_rate": 5.609244575397131e-06,
      "loss": 0.3721,
      "num_input_tokens_seen": 11324072,
      "step": 17470
    },
    {
      "epoch": 10.303655660377359,
      "grad_norm": 2.524446964263916,
      "learning_rate": 5.606690950918381e-06,
      "loss": 0.4185,
      "num_input_tokens_seen": 11326984,
      "step": 17475
    },
    {
      "epoch": 10.306603773584905,
      "grad_norm": 2.7242164611816406,
      "learning_rate": 5.604137165815586e-06,
      "loss": 0.2704,
      "num_input_tokens_seen": 11329864,
      "step": 17480
    },
    {
      "epoch": 10.309551886792454,
      "grad_norm": 6.399605751037598,
      "learning_rate": 5.601583220764874e-06,
      "loss": 0.4074,
      "num_input_tokens_seen": 11333000,
      "step": 17485
    },
    {
      "epoch": 10.3125,
      "grad_norm": 2.615506649017334,
      "learning_rate": 5.599029116442409e-06,
      "loss": 0.3292,
      "num_input_tokens_seen": 11335816,
      "step": 17490
    },
    {
      "epoch": 10.315448113207546,
      "grad_norm": 4.508909702301025,
      "learning_rate": 5.596474853524406e-06,
      "loss": 0.3754,
      "num_input_tokens_seen": 11338312,
      "step": 17495
    },
    {
      "epoch": 10.318396226415095,
      "grad_norm": 2.9991962909698486,
      "learning_rate": 5.593920432687115e-06,
      "loss": 0.3275,
      "num_input_tokens_seen": 11340968,
      "step": 17500
    },
    {
      "epoch": 10.321344339622641,
      "grad_norm": 2.3712575435638428,
      "learning_rate": 5.591365854606829e-06,
      "loss": 0.4595,
      "num_input_tokens_seen": 11345000,
      "step": 17505
    },
    {
      "epoch": 10.32429245283019,
      "grad_norm": 2.6174018383026123,
      "learning_rate": 5.588811119959885e-06,
      "loss": 0.2575,
      "num_input_tokens_seen": 11351816,
      "step": 17510
    },
    {
      "epoch": 10.327240566037736,
      "grad_norm": 2.350597858428955,
      "learning_rate": 5.58625622942266e-06,
      "loss": 0.3071,
      "num_input_tokens_seen": 11354568,
      "step": 17515
    },
    {
      "epoch": 10.330188679245284,
      "grad_norm": 2.0773611068725586,
      "learning_rate": 5.58370118367157e-06,
      "loss": 0.4836,
      "num_input_tokens_seen": 11358056,
      "step": 17520
    },
    {
      "epoch": 10.33313679245283,
      "grad_norm": 2.461127758026123,
      "learning_rate": 5.581145983383077e-06,
      "loss": 0.4433,
      "num_input_tokens_seen": 11361352,
      "step": 17525
    },
    {
      "epoch": 10.336084905660377,
      "grad_norm": 1.7160727977752686,
      "learning_rate": 5.57859062923368e-06,
      "loss": 0.5256,
      "num_input_tokens_seen": 11364456,
      "step": 17530
    },
    {
      "epoch": 10.339033018867925,
      "grad_norm": 2.8272247314453125,
      "learning_rate": 5.5760351218999194e-06,
      "loss": 0.4423,
      "num_input_tokens_seen": 11367688,
      "step": 17535
    },
    {
      "epoch": 10.341981132075471,
      "grad_norm": 1.6113783121109009,
      "learning_rate": 5.573479462058379e-06,
      "loss": 0.4153,
      "num_input_tokens_seen": 11371752,
      "step": 17540
    },
    {
      "epoch": 10.34492924528302,
      "grad_norm": 2.6164116859436035,
      "learning_rate": 5.570923650385679e-06,
      "loss": 0.3236,
      "num_input_tokens_seen": 11374824,
      "step": 17545
    },
    {
      "epoch": 10.347877358490566,
      "grad_norm": 1.7940168380737305,
      "learning_rate": 5.568367687558481e-06,
      "loss": 0.337,
      "num_input_tokens_seen": 11377384,
      "step": 17550
    },
    {
      "epoch": 10.350825471698114,
      "grad_norm": 3.096158742904663,
      "learning_rate": 5.56581157425349e-06,
      "loss": 0.3414,
      "num_input_tokens_seen": 11380200,
      "step": 17555
    },
    {
      "epoch": 10.35377358490566,
      "grad_norm": 4.106292724609375,
      "learning_rate": 5.563255311147446e-06,
      "loss": 0.4023,
      "num_input_tokens_seen": 11383560,
      "step": 17560
    },
    {
      "epoch": 10.356721698113208,
      "grad_norm": 3.4292125701904297,
      "learning_rate": 5.560698898917129e-06,
      "loss": 0.2929,
      "num_input_tokens_seen": 11387784,
      "step": 17565
    },
    {
      "epoch": 10.359669811320755,
      "grad_norm": 2.798210620880127,
      "learning_rate": 5.558142338239365e-06,
      "loss": 0.4537,
      "num_input_tokens_seen": 11390888,
      "step": 17570
    },
    {
      "epoch": 10.362617924528301,
      "grad_norm": 2.0550191402435303,
      "learning_rate": 5.55558562979101e-06,
      "loss": 0.3177,
      "num_input_tokens_seen": 11393896,
      "step": 17575
    },
    {
      "epoch": 10.36556603773585,
      "grad_norm": 2.6730730533599854,
      "learning_rate": 5.553028774248964e-06,
      "loss": 0.3296,
      "num_input_tokens_seen": 11397768,
      "step": 17580
    },
    {
      "epoch": 10.368514150943396,
      "grad_norm": 3.0346333980560303,
      "learning_rate": 5.5504717722901665e-06,
      "loss": 0.3164,
      "num_input_tokens_seen": 11400008,
      "step": 17585
    },
    {
      "epoch": 10.371462264150944,
      "grad_norm": 2.8956730365753174,
      "learning_rate": 5.547914624591597e-06,
      "loss": 0.2411,
      "num_input_tokens_seen": 11403528,
      "step": 17590
    },
    {
      "epoch": 10.37441037735849,
      "grad_norm": 2.6545090675354004,
      "learning_rate": 5.545357331830269e-06,
      "loss": 0.298,
      "num_input_tokens_seen": 11406664,
      "step": 17595
    },
    {
      "epoch": 10.377358490566039,
      "grad_norm": 2.4532601833343506,
      "learning_rate": 5.542799894683235e-06,
      "loss": 0.4421,
      "num_input_tokens_seen": 11410568,
      "step": 17600
    },
    {
      "epoch": 10.380306603773585,
      "grad_norm": 3.2629692554473877,
      "learning_rate": 5.540242313827591e-06,
      "loss": 0.4155,
      "num_input_tokens_seen": 11413512,
      "step": 17605
    },
    {
      "epoch": 10.383254716981131,
      "grad_norm": 3.328017234802246,
      "learning_rate": 5.537684589940466e-06,
      "loss": 0.357,
      "num_input_tokens_seen": 11416904,
      "step": 17610
    },
    {
      "epoch": 10.38620283018868,
      "grad_norm": 3.0733582973480225,
      "learning_rate": 5.535126723699029e-06,
      "loss": 0.4146,
      "num_input_tokens_seen": 11419432,
      "step": 17615
    },
    {
      "epoch": 10.389150943396226,
      "grad_norm": 3.6655616760253906,
      "learning_rate": 5.532568715780485e-06,
      "loss": 0.3395,
      "num_input_tokens_seen": 11421992,
      "step": 17620
    },
    {
      "epoch": 10.392099056603774,
      "grad_norm": 1.883724331855774,
      "learning_rate": 5.530010566862077e-06,
      "loss": 0.3951,
      "num_input_tokens_seen": 11425192,
      "step": 17625
    },
    {
      "epoch": 10.39504716981132,
      "grad_norm": 1.7776939868927002,
      "learning_rate": 5.527452277621089e-06,
      "loss": 0.2934,
      "num_input_tokens_seen": 11427464,
      "step": 17630
    },
    {
      "epoch": 10.397995283018869,
      "grad_norm": 2.5193300247192383,
      "learning_rate": 5.524893848734837e-06,
      "loss": 0.4478,
      "num_input_tokens_seen": 11429864,
      "step": 17635
    },
    {
      "epoch": 10.400943396226415,
      "grad_norm": 6.025152683258057,
      "learning_rate": 5.522335280880676e-06,
      "loss": 0.4117,
      "num_input_tokens_seen": 11432328,
      "step": 17640
    },
    {
      "epoch": 10.403891509433961,
      "grad_norm": 3.2512171268463135,
      "learning_rate": 5.519776574735999e-06,
      "loss": 0.3021,
      "num_input_tokens_seen": 11435272,
      "step": 17645
    },
    {
      "epoch": 10.40683962264151,
      "grad_norm": 2.0679171085357666,
      "learning_rate": 5.5172177309782325e-06,
      "loss": 0.3834,
      "num_input_tokens_seen": 11438568,
      "step": 17650
    },
    {
      "epoch": 10.409787735849056,
      "grad_norm": 2.598850727081299,
      "learning_rate": 5.514658750284844e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 11442696,
      "step": 17655
    },
    {
      "epoch": 10.412735849056604,
      "grad_norm": 2.3256072998046875,
      "learning_rate": 5.512099633333332e-06,
      "loss": 0.416,
      "num_input_tokens_seen": 11445704,
      "step": 17660
    },
    {
      "epoch": 10.41568396226415,
      "grad_norm": 3.1155474185943604,
      "learning_rate": 5.509540380801236e-06,
      "loss": 0.3764,
      "num_input_tokens_seen": 11448712,
      "step": 17665
    },
    {
      "epoch": 10.418632075471699,
      "grad_norm": 2.213688373565674,
      "learning_rate": 5.506980993366129e-06,
      "loss": 0.4612,
      "num_input_tokens_seen": 11451624,
      "step": 17670
    },
    {
      "epoch": 10.421580188679245,
      "grad_norm": 3.0261693000793457,
      "learning_rate": 5.504421471705616e-06,
      "loss": 0.4026,
      "num_input_tokens_seen": 11455944,
      "step": 17675
    },
    {
      "epoch": 10.424528301886792,
      "grad_norm": 2.3847849369049072,
      "learning_rate": 5.501861816497346e-06,
      "loss": 0.448,
      "num_input_tokens_seen": 11458824,
      "step": 17680
    },
    {
      "epoch": 10.42747641509434,
      "grad_norm": 3.0768401622772217,
      "learning_rate": 5.499302028418998e-06,
      "loss": 0.3367,
      "num_input_tokens_seen": 11462472,
      "step": 17685
    },
    {
      "epoch": 10.430424528301886,
      "grad_norm": 2.433180570602417,
      "learning_rate": 5.496742108148285e-06,
      "loss": 0.3858,
      "num_input_tokens_seen": 11465096,
      "step": 17690
    },
    {
      "epoch": 10.433372641509434,
      "grad_norm": 3.871994972229004,
      "learning_rate": 5.494182056362959e-06,
      "loss": 0.3142,
      "num_input_tokens_seen": 11468296,
      "step": 17695
    },
    {
      "epoch": 10.43632075471698,
      "grad_norm": 1.9155415296554565,
      "learning_rate": 5.491621873740804e-06,
      "loss": 0.4281,
      "num_input_tokens_seen": 11472168,
      "step": 17700
    },
    {
      "epoch": 10.439268867924529,
      "grad_norm": 1.9214411973953247,
      "learning_rate": 5.4890615609596375e-06,
      "loss": 0.4235,
      "num_input_tokens_seen": 11475304,
      "step": 17705
    },
    {
      "epoch": 10.442216981132075,
      "grad_norm": 3.1445672512054443,
      "learning_rate": 5.486501118697317e-06,
      "loss": 0.3628,
      "num_input_tokens_seen": 11478312,
      "step": 17710
    },
    {
      "epoch": 10.445165094339623,
      "grad_norm": 1.8820666074752808,
      "learning_rate": 5.483940547631727e-06,
      "loss": 0.3808,
      "num_input_tokens_seen": 11481352,
      "step": 17715
    },
    {
      "epoch": 10.44811320754717,
      "grad_norm": 2.5469932556152344,
      "learning_rate": 5.481379848440792e-06,
      "loss": 0.3764,
      "num_input_tokens_seen": 11485224,
      "step": 17720
    },
    {
      "epoch": 10.451061320754716,
      "grad_norm": 2.433492422103882,
      "learning_rate": 5.478819021802468e-06,
      "loss": 0.3097,
      "num_input_tokens_seen": 11488584,
      "step": 17725
    },
    {
      "epoch": 10.454009433962264,
      "grad_norm": 3.33270525932312,
      "learning_rate": 5.476258068394743e-06,
      "loss": 0.4785,
      "num_input_tokens_seen": 11491720,
      "step": 17730
    },
    {
      "epoch": 10.45695754716981,
      "grad_norm": 3.678601026535034,
      "learning_rate": 5.473696988895644e-06,
      "loss": 0.5295,
      "num_input_tokens_seen": 11496488,
      "step": 17735
    },
    {
      "epoch": 10.459905660377359,
      "grad_norm": 2.486236572265625,
      "learning_rate": 5.471135783983224e-06,
      "loss": 0.3225,
      "num_input_tokens_seen": 11499656,
      "step": 17740
    },
    {
      "epoch": 10.462853773584905,
      "grad_norm": 2.122868537902832,
      "learning_rate": 5.4685744543355745e-06,
      "loss": 0.376,
      "num_input_tokens_seen": 11502152,
      "step": 17745
    },
    {
      "epoch": 10.465801886792454,
      "grad_norm": 2.6207542419433594,
      "learning_rate": 5.466013000630819e-06,
      "loss": 0.3407,
      "num_input_tokens_seen": 11506120,
      "step": 17750
    },
    {
      "epoch": 10.46875,
      "grad_norm": 4.257349967956543,
      "learning_rate": 5.463451423547114e-06,
      "loss": 0.5672,
      "num_input_tokens_seen": 11511400,
      "step": 17755
    },
    {
      "epoch": 10.471698113207546,
      "grad_norm": 2.9057013988494873,
      "learning_rate": 5.460889723762647e-06,
      "loss": 0.3737,
      "num_input_tokens_seen": 11514632,
      "step": 17760
    },
    {
      "epoch": 10.474646226415095,
      "grad_norm": 3.08778977394104,
      "learning_rate": 5.458327901955639e-06,
      "loss": 0.3444,
      "num_input_tokens_seen": 11517608,
      "step": 17765
    },
    {
      "epoch": 10.477594339622641,
      "grad_norm": 2.9857399463653564,
      "learning_rate": 5.455765958804344e-06,
      "loss": 0.3598,
      "num_input_tokens_seen": 11521480,
      "step": 17770
    },
    {
      "epoch": 10.48054245283019,
      "grad_norm": 3.257270574569702,
      "learning_rate": 5.4532038949870455e-06,
      "loss": 0.4163,
      "num_input_tokens_seen": 11524392,
      "step": 17775
    },
    {
      "epoch": 10.483490566037736,
      "grad_norm": 2.7306604385375977,
      "learning_rate": 5.450641711182066e-06,
      "loss": 0.3284,
      "num_input_tokens_seen": 11527656,
      "step": 17780
    },
    {
      "epoch": 10.486438679245284,
      "grad_norm": 3.9536032676696777,
      "learning_rate": 5.448079408067748e-06,
      "loss": 0.3133,
      "num_input_tokens_seen": 11530856,
      "step": 17785
    },
    {
      "epoch": 10.48938679245283,
      "grad_norm": 2.777256488800049,
      "learning_rate": 5.4455169863224775e-06,
      "loss": 0.5786,
      "num_input_tokens_seen": 11534088,
      "step": 17790
    },
    {
      "epoch": 10.492334905660377,
      "grad_norm": 4.217455863952637,
      "learning_rate": 5.442954446624664e-06,
      "loss": 0.3806,
      "num_input_tokens_seen": 11537320,
      "step": 17795
    },
    {
      "epoch": 10.495283018867925,
      "grad_norm": 4.009532928466797,
      "learning_rate": 5.440391789652752e-06,
      "loss": 0.4307,
      "num_input_tokens_seen": 11539848,
      "step": 17800
    },
    {
      "epoch": 10.498231132075471,
      "grad_norm": 2.8378283977508545,
      "learning_rate": 5.437829016085216e-06,
      "loss": 0.393,
      "num_input_tokens_seen": 11543016,
      "step": 17805
    },
    {
      "epoch": 10.50117924528302,
      "grad_norm": 1.6951159238815308,
      "learning_rate": 5.435266126600561e-06,
      "loss": 0.3267,
      "num_input_tokens_seen": 11546664,
      "step": 17810
    },
    {
      "epoch": 10.504127358490566,
      "grad_norm": 2.9048895835876465,
      "learning_rate": 5.4327031218773215e-06,
      "loss": 0.3676,
      "num_input_tokens_seen": 11550856,
      "step": 17815
    },
    {
      "epoch": 10.507075471698114,
      "grad_norm": 2.494565963745117,
      "learning_rate": 5.430140002594067e-06,
      "loss": 0.3983,
      "num_input_tokens_seen": 11553800,
      "step": 17820
    },
    {
      "epoch": 10.51002358490566,
      "grad_norm": 2.6736083030700684,
      "learning_rate": 5.4275767694293934e-06,
      "loss": 0.3101,
      "num_input_tokens_seen": 11556552,
      "step": 17825
    },
    {
      "epoch": 10.512971698113208,
      "grad_norm": 2.7683818340301514,
      "learning_rate": 5.425013423061926e-06,
      "loss": 0.337,
      "num_input_tokens_seen": 11559592,
      "step": 17830
    },
    {
      "epoch": 10.515919811320755,
      "grad_norm": 2.4902546405792236,
      "learning_rate": 5.422449964170324e-06,
      "loss": 0.4098,
      "num_input_tokens_seen": 11562600,
      "step": 17835
    },
    {
      "epoch": 10.518867924528301,
      "grad_norm": 2.721428632736206,
      "learning_rate": 5.419886393433275e-06,
      "loss": 0.3646,
      "num_input_tokens_seen": 11565608,
      "step": 17840
    },
    {
      "epoch": 10.52181603773585,
      "grad_norm": 2.0989279747009277,
      "learning_rate": 5.417322711529491e-06,
      "loss": 0.3125,
      "num_input_tokens_seen": 11568328,
      "step": 17845
    },
    {
      "epoch": 10.524764150943396,
      "grad_norm": 1.7996803522109985,
      "learning_rate": 5.4147589191377224e-06,
      "loss": 0.3312,
      "num_input_tokens_seen": 11571816,
      "step": 17850
    },
    {
      "epoch": 10.527712264150944,
      "grad_norm": 1.9102144241333008,
      "learning_rate": 5.412195016936742e-06,
      "loss": 0.4217,
      "num_input_tokens_seen": 11574568,
      "step": 17855
    },
    {
      "epoch": 10.53066037735849,
      "grad_norm": 3.647956371307373,
      "learning_rate": 5.409631005605354e-06,
      "loss": 0.4309,
      "num_input_tokens_seen": 11577736,
      "step": 17860
    },
    {
      "epoch": 10.533608490566039,
      "grad_norm": 3.0900535583496094,
      "learning_rate": 5.407066885822391e-06,
      "loss": 0.4905,
      "num_input_tokens_seen": 11581832,
      "step": 17865
    },
    {
      "epoch": 10.536556603773585,
      "grad_norm": 3.167769432067871,
      "learning_rate": 5.404502658266717e-06,
      "loss": 0.3596,
      "num_input_tokens_seen": 11584616,
      "step": 17870
    },
    {
      "epoch": 10.539504716981131,
      "grad_norm": 2.064368963241577,
      "learning_rate": 5.4019383236172195e-06,
      "loss": 0.3797,
      "num_input_tokens_seen": 11588136,
      "step": 17875
    },
    {
      "epoch": 10.54245283018868,
      "grad_norm": 3.4428231716156006,
      "learning_rate": 5.39937388255282e-06,
      "loss": 0.4108,
      "num_input_tokens_seen": 11591560,
      "step": 17880
    },
    {
      "epoch": 10.545400943396226,
      "grad_norm": 2.90517258644104,
      "learning_rate": 5.3968093357524645e-06,
      "loss": 0.3977,
      "num_input_tokens_seen": 11594472,
      "step": 17885
    },
    {
      "epoch": 10.548349056603774,
      "grad_norm": 4.4778218269348145,
      "learning_rate": 5.3942446838951245e-06,
      "loss": 0.3621,
      "num_input_tokens_seen": 11597032,
      "step": 17890
    },
    {
      "epoch": 10.55129716981132,
      "grad_norm": 3.330404758453369,
      "learning_rate": 5.3916799276598074e-06,
      "loss": 0.444,
      "num_input_tokens_seen": 11600264,
      "step": 17895
    },
    {
      "epoch": 10.554245283018869,
      "grad_norm": 4.485269069671631,
      "learning_rate": 5.3891150677255425e-06,
      "loss": 0.4076,
      "num_input_tokens_seen": 11602952,
      "step": 17900
    },
    {
      "epoch": 10.557193396226415,
      "grad_norm": 2.509303569793701,
      "learning_rate": 5.386550104771384e-06,
      "loss": 0.3241,
      "num_input_tokens_seen": 11605896,
      "step": 17905
    },
    {
      "epoch": 10.560141509433961,
      "grad_norm": 2.2822988033294678,
      "learning_rate": 5.3839850394764205e-06,
      "loss": 0.4033,
      "num_input_tokens_seen": 11609992,
      "step": 17910
    },
    {
      "epoch": 10.56308962264151,
      "grad_norm": 1.9969121217727661,
      "learning_rate": 5.381419872519763e-06,
      "loss": 0.2754,
      "num_input_tokens_seen": 11613832,
      "step": 17915
    },
    {
      "epoch": 10.566037735849056,
      "grad_norm": 2.846660614013672,
      "learning_rate": 5.378854604580549e-06,
      "loss": 0.306,
      "num_input_tokens_seen": 11616776,
      "step": 17920
    },
    {
      "epoch": 10.568985849056604,
      "grad_norm": 2.993920087814331,
      "learning_rate": 5.376289236337946e-06,
      "loss": 0.4092,
      "num_input_tokens_seen": 11621064,
      "step": 17925
    },
    {
      "epoch": 10.57193396226415,
      "grad_norm": 2.359995126724243,
      "learning_rate": 5.373723768471147e-06,
      "loss": 0.3831,
      "num_input_tokens_seen": 11623752,
      "step": 17930
    },
    {
      "epoch": 10.574882075471699,
      "grad_norm": 3.2069509029388428,
      "learning_rate": 5.37115820165937e-06,
      "loss": 0.3528,
      "num_input_tokens_seen": 11627336,
      "step": 17935
    },
    {
      "epoch": 10.577830188679245,
      "grad_norm": 2.997507333755493,
      "learning_rate": 5.368592536581858e-06,
      "loss": 0.5012,
      "num_input_tokens_seen": 11629992,
      "step": 17940
    },
    {
      "epoch": 10.580778301886792,
      "grad_norm": 5.609867572784424,
      "learning_rate": 5.366026773917885e-06,
      "loss": 0.377,
      "num_input_tokens_seen": 11632904,
      "step": 17945
    },
    {
      "epoch": 10.58372641509434,
      "grad_norm": 2.3239808082580566,
      "learning_rate": 5.363460914346746e-06,
      "loss": 0.3208,
      "num_input_tokens_seen": 11635240,
      "step": 17950
    },
    {
      "epoch": 10.586674528301886,
      "grad_norm": 1.7865527868270874,
      "learning_rate": 5.360894958547762e-06,
      "loss": 0.3572,
      "num_input_tokens_seen": 11638760,
      "step": 17955
    },
    {
      "epoch": 10.589622641509434,
      "grad_norm": 1.9999042749404907,
      "learning_rate": 5.358328907200284e-06,
      "loss": 0.3939,
      "num_input_tokens_seen": 11642216,
      "step": 17960
    },
    {
      "epoch": 10.59257075471698,
      "grad_norm": 2.8241078853607178,
      "learning_rate": 5.355762760983682e-06,
      "loss": 0.3454,
      "num_input_tokens_seen": 11645544,
      "step": 17965
    },
    {
      "epoch": 10.595518867924529,
      "grad_norm": 3.0304479598999023,
      "learning_rate": 5.353196520577356e-06,
      "loss": 0.349,
      "num_input_tokens_seen": 11648264,
      "step": 17970
    },
    {
      "epoch": 10.598466981132075,
      "grad_norm": 7.309081554412842,
      "learning_rate": 5.35063018666073e-06,
      "loss": 0.4134,
      "num_input_tokens_seen": 11651656,
      "step": 17975
    },
    {
      "epoch": 10.601415094339622,
      "grad_norm": 2.5791280269622803,
      "learning_rate": 5.3480637599132515e-06,
      "loss": 0.4131,
      "num_input_tokens_seen": 11654408,
      "step": 17980
    },
    {
      "epoch": 10.60436320754717,
      "grad_norm": 2.1070215702056885,
      "learning_rate": 5.34549724101439e-06,
      "loss": 0.4181,
      "num_input_tokens_seen": 11657320,
      "step": 17985
    },
    {
      "epoch": 10.607311320754716,
      "grad_norm": 2.7615444660186768,
      "learning_rate": 5.342930630643646e-06,
      "loss": 0.3647,
      "num_input_tokens_seen": 11661000,
      "step": 17990
    },
    {
      "epoch": 10.610259433962264,
      "grad_norm": 1.932367205619812,
      "learning_rate": 5.340363929480541e-06,
      "loss": 0.3391,
      "num_input_tokens_seen": 11664296,
      "step": 17995
    },
    {
      "epoch": 10.61320754716981,
      "grad_norm": 2.9753618240356445,
      "learning_rate": 5.3377971382046164e-06,
      "loss": 0.4657,
      "num_input_tokens_seen": 11668456,
      "step": 18000
    },
    {
      "epoch": 10.616155660377359,
      "grad_norm": 2.963609218597412,
      "learning_rate": 5.335230257495446e-06,
      "loss": 0.4681,
      "num_input_tokens_seen": 11671848,
      "step": 18005
    },
    {
      "epoch": 10.619103773584905,
      "grad_norm": 1.7239453792572021,
      "learning_rate": 5.3326632880326205e-06,
      "loss": 0.3358,
      "num_input_tokens_seen": 11675592,
      "step": 18010
    },
    {
      "epoch": 10.622051886792454,
      "grad_norm": 1.6939432621002197,
      "learning_rate": 5.3300962304957515e-06,
      "loss": 0.3764,
      "num_input_tokens_seen": 11679272,
      "step": 18015
    },
    {
      "epoch": 10.625,
      "grad_norm": 2.4801759719848633,
      "learning_rate": 5.327529085564487e-06,
      "loss": 0.3134,
      "num_input_tokens_seen": 11682376,
      "step": 18020
    },
    {
      "epoch": 10.627948113207546,
      "grad_norm": 2.8111610412597656,
      "learning_rate": 5.324961853918485e-06,
      "loss": 0.3307,
      "num_input_tokens_seen": 11684808,
      "step": 18025
    },
    {
      "epoch": 10.630896226415095,
      "grad_norm": 7.087552547454834,
      "learning_rate": 5.32239453623743e-06,
      "loss": 0.4067,
      "num_input_tokens_seen": 11687976,
      "step": 18030
    },
    {
      "epoch": 10.633844339622641,
      "grad_norm": 2.232022762298584,
      "learning_rate": 5.3198271332010335e-06,
      "loss": 0.3994,
      "num_input_tokens_seen": 11691240,
      "step": 18035
    },
    {
      "epoch": 10.63679245283019,
      "grad_norm": 2.053398370742798,
      "learning_rate": 5.317259645489024e-06,
      "loss": 0.3096,
      "num_input_tokens_seen": 11694504,
      "step": 18040
    },
    {
      "epoch": 10.639740566037736,
      "grad_norm": 2.7050018310546875,
      "learning_rate": 5.314692073781157e-06,
      "loss": 0.3093,
      "num_input_tokens_seen": 11697192,
      "step": 18045
    },
    {
      "epoch": 10.642688679245284,
      "grad_norm": 4.194362640380859,
      "learning_rate": 5.312124418757207e-06,
      "loss": 0.3491,
      "num_input_tokens_seen": 11704616,
      "step": 18050
    },
    {
      "epoch": 10.64563679245283,
      "grad_norm": 2.3605334758758545,
      "learning_rate": 5.309556681096972e-06,
      "loss": 0.2366,
      "num_input_tokens_seen": 11707560,
      "step": 18055
    },
    {
      "epoch": 10.648584905660378,
      "grad_norm": 3.4487814903259277,
      "learning_rate": 5.306988861480271e-06,
      "loss": 0.4527,
      "num_input_tokens_seen": 11710568,
      "step": 18060
    },
    {
      "epoch": 10.651533018867925,
      "grad_norm": 2.7388601303100586,
      "learning_rate": 5.304420960586946e-06,
      "loss": 0.4128,
      "num_input_tokens_seen": 11713864,
      "step": 18065
    },
    {
      "epoch": 10.654481132075471,
      "grad_norm": 2.7827320098876953,
      "learning_rate": 5.3018529790968606e-06,
      "loss": 0.2428,
      "num_input_tokens_seen": 11716840,
      "step": 18070
    },
    {
      "epoch": 10.65742924528302,
      "grad_norm": 2.315171480178833,
      "learning_rate": 5.299284917689898e-06,
      "loss": 0.4438,
      "num_input_tokens_seen": 11720008,
      "step": 18075
    },
    {
      "epoch": 10.660377358490566,
      "grad_norm": 2.6058754920959473,
      "learning_rate": 5.296716777045962e-06,
      "loss": 0.4942,
      "num_input_tokens_seen": 11723016,
      "step": 18080
    },
    {
      "epoch": 10.663325471698114,
      "grad_norm": 2.4323296546936035,
      "learning_rate": 5.294148557844983e-06,
      "loss": 0.2867,
      "num_input_tokens_seen": 11726696,
      "step": 18085
    },
    {
      "epoch": 10.66627358490566,
      "grad_norm": 2.95430850982666,
      "learning_rate": 5.291580260766904e-06,
      "loss": 0.3444,
      "num_input_tokens_seen": 11730600,
      "step": 18090
    },
    {
      "epoch": 10.669221698113208,
      "grad_norm": 2.324597120285034,
      "learning_rate": 5.289011886491694e-06,
      "loss": 0.332,
      "num_input_tokens_seen": 11733096,
      "step": 18095
    },
    {
      "epoch": 10.672169811320755,
      "grad_norm": 6.7521796226501465,
      "learning_rate": 5.286443435699342e-06,
      "loss": 0.3444,
      "num_input_tokens_seen": 11736552,
      "step": 18100
    },
    {
      "epoch": 10.675117924528301,
      "grad_norm": 7.641258716583252,
      "learning_rate": 5.283874909069855e-06,
      "loss": 0.3358,
      "num_input_tokens_seen": 11739240,
      "step": 18105
    },
    {
      "epoch": 10.67806603773585,
      "grad_norm": 2.901128053665161,
      "learning_rate": 5.281306307283263e-06,
      "loss": 0.3871,
      "num_input_tokens_seen": 11742216,
      "step": 18110
    },
    {
      "epoch": 10.681014150943396,
      "grad_norm": 3.6663949489593506,
      "learning_rate": 5.2787376310196145e-06,
      "loss": 0.393,
      "num_input_tokens_seen": 11745608,
      "step": 18115
    },
    {
      "epoch": 10.683962264150944,
      "grad_norm": 3.09061598777771,
      "learning_rate": 5.276168880958977e-06,
      "loss": 0.5591,
      "num_input_tokens_seen": 11749224,
      "step": 18120
    },
    {
      "epoch": 10.68691037735849,
      "grad_norm": 2.9534897804260254,
      "learning_rate": 5.273600057781437e-06,
      "loss": 0.4111,
      "num_input_tokens_seen": 11752040,
      "step": 18125
    },
    {
      "epoch": 10.689858490566039,
      "grad_norm": 2.6125564575195312,
      "learning_rate": 5.271031162167103e-06,
      "loss": 0.5188,
      "num_input_tokens_seen": 11756008,
      "step": 18130
    },
    {
      "epoch": 10.692806603773585,
      "grad_norm": 3.8049631118774414,
      "learning_rate": 5.268462194796101e-06,
      "loss": 0.4859,
      "num_input_tokens_seen": 11759080,
      "step": 18135
    },
    {
      "epoch": 10.695754716981131,
      "grad_norm": 6.745090484619141,
      "learning_rate": 5.265893156348576e-06,
      "loss": 0.4157,
      "num_input_tokens_seen": 11761864,
      "step": 18140
    },
    {
      "epoch": 10.69870283018868,
      "grad_norm": 3.323242664337158,
      "learning_rate": 5.2633240475046925e-06,
      "loss": 0.3413,
      "num_input_tokens_seen": 11765160,
      "step": 18145
    },
    {
      "epoch": 10.701650943396226,
      "grad_norm": 5.657540798187256,
      "learning_rate": 5.2607548689446305e-06,
      "loss": 0.3168,
      "num_input_tokens_seen": 11767624,
      "step": 18150
    },
    {
      "epoch": 10.704599056603774,
      "grad_norm": 3.4652881622314453,
      "learning_rate": 5.258185621348595e-06,
      "loss": 0.3588,
      "num_input_tokens_seen": 11771880,
      "step": 18155
    },
    {
      "epoch": 10.70754716981132,
      "grad_norm": 1.868206262588501,
      "learning_rate": 5.255616305396801e-06,
      "loss": 0.3543,
      "num_input_tokens_seen": 11775560,
      "step": 18160
    },
    {
      "epoch": 10.710495283018869,
      "grad_norm": 4.919556140899658,
      "learning_rate": 5.253046921769491e-06,
      "loss": 0.4352,
      "num_input_tokens_seen": 11778056,
      "step": 18165
    },
    {
      "epoch": 10.713443396226415,
      "grad_norm": 0.9494245052337646,
      "learning_rate": 5.250477471146916e-06,
      "loss": 0.2055,
      "num_input_tokens_seen": 11780488,
      "step": 18170
    },
    {
      "epoch": 10.716391509433961,
      "grad_norm": 3.066398859024048,
      "learning_rate": 5.2479079542093535e-06,
      "loss": 0.3566,
      "num_input_tokens_seen": 11783048,
      "step": 18175
    },
    {
      "epoch": 10.71933962264151,
      "grad_norm": 2.2293527126312256,
      "learning_rate": 5.245338371637091e-06,
      "loss": 0.3717,
      "num_input_tokens_seen": 11785640,
      "step": 18180
    },
    {
      "epoch": 10.722287735849056,
      "grad_norm": 2.383026599884033,
      "learning_rate": 5.242768724110437e-06,
      "loss": 0.3338,
      "num_input_tokens_seen": 11791016,
      "step": 18185
    },
    {
      "epoch": 10.725235849056604,
      "grad_norm": 3.5561742782592773,
      "learning_rate": 5.240199012309717e-06,
      "loss": 0.4152,
      "num_input_tokens_seen": 11793704,
      "step": 18190
    },
    {
      "epoch": 10.72818396226415,
      "grad_norm": 4.358463287353516,
      "learning_rate": 5.237629236915273e-06,
      "loss": 0.4309,
      "num_input_tokens_seen": 11797224,
      "step": 18195
    },
    {
      "epoch": 10.731132075471699,
      "grad_norm": 2.9588301181793213,
      "learning_rate": 5.235059398607464e-06,
      "loss": 0.3928,
      "num_input_tokens_seen": 11800456,
      "step": 18200
    },
    {
      "epoch": 10.734080188679245,
      "grad_norm": 3.1863646507263184,
      "learning_rate": 5.232489498066665e-06,
      "loss": 0.3871,
      "num_input_tokens_seen": 11802856,
      "step": 18205
    },
    {
      "epoch": 10.737028301886792,
      "grad_norm": 4.237438201904297,
      "learning_rate": 5.229919535973272e-06,
      "loss": 0.3761,
      "num_input_tokens_seen": 11806280,
      "step": 18210
    },
    {
      "epoch": 10.73997641509434,
      "grad_norm": 3.4081969261169434,
      "learning_rate": 5.2273495130076905e-06,
      "loss": 0.3875,
      "num_input_tokens_seen": 11809256,
      "step": 18215
    },
    {
      "epoch": 10.742924528301886,
      "grad_norm": 3.483201265335083,
      "learning_rate": 5.224779429850344e-06,
      "loss": 0.3916,
      "num_input_tokens_seen": 11811592,
      "step": 18220
    },
    {
      "epoch": 10.745872641509434,
      "grad_norm": 4.42886209487915,
      "learning_rate": 5.222209287181677e-06,
      "loss": 0.4673,
      "num_input_tokens_seen": 11814632,
      "step": 18225
    },
    {
      "epoch": 10.74882075471698,
      "grad_norm": 3.2785677909851074,
      "learning_rate": 5.219639085682142e-06,
      "loss": 0.3349,
      "num_input_tokens_seen": 11817448,
      "step": 18230
    },
    {
      "epoch": 10.751768867924529,
      "grad_norm": 2.424466133117676,
      "learning_rate": 5.2170688260322124e-06,
      "loss": 0.427,
      "num_input_tokens_seen": 11822088,
      "step": 18235
    },
    {
      "epoch": 10.754716981132075,
      "grad_norm": 3.4101722240448,
      "learning_rate": 5.214498508912376e-06,
      "loss": 0.5147,
      "num_input_tokens_seen": 11825480,
      "step": 18240
    },
    {
      "epoch": 10.757665094339622,
      "grad_norm": 3.382107734680176,
      "learning_rate": 5.211928135003135e-06,
      "loss": 0.3392,
      "num_input_tokens_seen": 11829128,
      "step": 18245
    },
    {
      "epoch": 10.76061320754717,
      "grad_norm": 2.5446970462799072,
      "learning_rate": 5.209357704985007e-06,
      "loss": 0.364,
      "num_input_tokens_seen": 11832328,
      "step": 18250
    },
    {
      "epoch": 10.763561320754716,
      "grad_norm": 2.5703935623168945,
      "learning_rate": 5.206787219538524e-06,
      "loss": 0.3446,
      "num_input_tokens_seen": 11835720,
      "step": 18255
    },
    {
      "epoch": 10.766509433962264,
      "grad_norm": 4.024535655975342,
      "learning_rate": 5.204216679344234e-06,
      "loss": 0.3987,
      "num_input_tokens_seen": 11839016,
      "step": 18260
    },
    {
      "epoch": 10.76945754716981,
      "grad_norm": 2.1855781078338623,
      "learning_rate": 5.201646085082696e-06,
      "loss": 0.3191,
      "num_input_tokens_seen": 11841864,
      "step": 18265
    },
    {
      "epoch": 10.772405660377359,
      "grad_norm": 3.864346504211426,
      "learning_rate": 5.199075437434491e-06,
      "loss": 0.3462,
      "num_input_tokens_seen": 11845160,
      "step": 18270
    },
    {
      "epoch": 10.775353773584905,
      "grad_norm": 1.7836920022964478,
      "learning_rate": 5.1965047370802046e-06,
      "loss": 0.3853,
      "num_input_tokens_seen": 11849640,
      "step": 18275
    },
    {
      "epoch": 10.778301886792454,
      "grad_norm": 4.165532112121582,
      "learning_rate": 5.193933984700441e-06,
      "loss": 0.3541,
      "num_input_tokens_seen": 11852552,
      "step": 18280
    },
    {
      "epoch": 10.78125,
      "grad_norm": 3.5444257259368896,
      "learning_rate": 5.1913631809758216e-06,
      "loss": 0.4472,
      "num_input_tokens_seen": 11856040,
      "step": 18285
    },
    {
      "epoch": 10.784198113207546,
      "grad_norm": 2.6528265476226807,
      "learning_rate": 5.188792326586973e-06,
      "loss": 0.3807,
      "num_input_tokens_seen": 11859112,
      "step": 18290
    },
    {
      "epoch": 10.787146226415095,
      "grad_norm": 3.58778715133667,
      "learning_rate": 5.186221422214544e-06,
      "loss": 0.4109,
      "num_input_tokens_seen": 11861864,
      "step": 18295
    },
    {
      "epoch": 10.790094339622641,
      "grad_norm": 9.267985343933105,
      "learning_rate": 5.1836504685391885e-06,
      "loss": 0.3955,
      "num_input_tokens_seen": 11864872,
      "step": 18300
    },
    {
      "epoch": 10.79304245283019,
      "grad_norm": 2.1491315364837646,
      "learning_rate": 5.181079466241582e-06,
      "loss": 0.3279,
      "num_input_tokens_seen": 11867112,
      "step": 18305
    },
    {
      "epoch": 10.795990566037736,
      "grad_norm": 2.8585870265960693,
      "learning_rate": 5.178508416002406e-06,
      "loss": 0.3719,
      "num_input_tokens_seen": 11871016,
      "step": 18310
    },
    {
      "epoch": 10.798938679245284,
      "grad_norm": 3.286445140838623,
      "learning_rate": 5.175937318502357e-06,
      "loss": 0.3162,
      "num_input_tokens_seen": 11874376,
      "step": 18315
    },
    {
      "epoch": 10.80188679245283,
      "grad_norm": 2.6500918865203857,
      "learning_rate": 5.173366174422147e-06,
      "loss": 0.4127,
      "num_input_tokens_seen": 11880488,
      "step": 18320
    },
    {
      "epoch": 10.804834905660378,
      "grad_norm": 2.696051597595215,
      "learning_rate": 5.170794984442492e-06,
      "loss": 0.4996,
      "num_input_tokens_seen": 11883816,
      "step": 18325
    },
    {
      "epoch": 10.807783018867925,
      "grad_norm": 2.400271415710449,
      "learning_rate": 5.16822374924413e-06,
      "loss": 0.3603,
      "num_input_tokens_seen": 11887048,
      "step": 18330
    },
    {
      "epoch": 10.810731132075471,
      "grad_norm": 2.4009387493133545,
      "learning_rate": 5.165652469507806e-06,
      "loss": 0.35,
      "num_input_tokens_seen": 11889800,
      "step": 18335
    },
    {
      "epoch": 10.81367924528302,
      "grad_norm": 4.088010787963867,
      "learning_rate": 5.163081145914276e-06,
      "loss": 0.5445,
      "num_input_tokens_seen": 11893448,
      "step": 18340
    },
    {
      "epoch": 10.816627358490566,
      "grad_norm": 2.1316044330596924,
      "learning_rate": 5.160509779144311e-06,
      "loss": 0.3345,
      "num_input_tokens_seen": 11896008,
      "step": 18345
    },
    {
      "epoch": 10.819575471698114,
      "grad_norm": 3.102667808532715,
      "learning_rate": 5.157938369878688e-06,
      "loss": 0.3579,
      "num_input_tokens_seen": 11898696,
      "step": 18350
    },
    {
      "epoch": 10.82252358490566,
      "grad_norm": 2.2511706352233887,
      "learning_rate": 5.155366918798203e-06,
      "loss": 0.3826,
      "num_input_tokens_seen": 11901736,
      "step": 18355
    },
    {
      "epoch": 10.825471698113208,
      "grad_norm": 1.7467173337936401,
      "learning_rate": 5.152795426583654e-06,
      "loss": 0.4059,
      "num_input_tokens_seen": 11904840,
      "step": 18360
    },
    {
      "epoch": 10.828419811320755,
      "grad_norm": 3.6827893257141113,
      "learning_rate": 5.15022389391586e-06,
      "loss": 0.293,
      "num_input_tokens_seen": 11907912,
      "step": 18365
    },
    {
      "epoch": 10.831367924528301,
      "grad_norm": 2.7271828651428223,
      "learning_rate": 5.147652321475642e-06,
      "loss": 0.3076,
      "num_input_tokens_seen": 11910632,
      "step": 18370
    },
    {
      "epoch": 10.83431603773585,
      "grad_norm": 2.1980648040771484,
      "learning_rate": 5.145080709943835e-06,
      "loss": 0.4273,
      "num_input_tokens_seen": 11913832,
      "step": 18375
    },
    {
      "epoch": 10.837264150943396,
      "grad_norm": 1.5186188220977783,
      "learning_rate": 5.142509060001285e-06,
      "loss": 0.2725,
      "num_input_tokens_seen": 11916648,
      "step": 18380
    },
    {
      "epoch": 10.840212264150944,
      "grad_norm": 3.2239136695861816,
      "learning_rate": 5.139937372328847e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 11919176,
      "step": 18385
    },
    {
      "epoch": 10.84316037735849,
      "grad_norm": 2.5006422996520996,
      "learning_rate": 5.1373656476073876e-06,
      "loss": 0.411,
      "num_input_tokens_seen": 11922472,
      "step": 18390
    },
    {
      "epoch": 10.846108490566039,
      "grad_norm": 2.9334912300109863,
      "learning_rate": 5.134793886517779e-06,
      "loss": 0.3787,
      "num_input_tokens_seen": 11925448,
      "step": 18395
    },
    {
      "epoch": 10.849056603773585,
      "grad_norm": 3.9286863803863525,
      "learning_rate": 5.1322220897409105e-06,
      "loss": 0.4192,
      "num_input_tokens_seen": 11929352,
      "step": 18400
    },
    {
      "epoch": 10.852004716981131,
      "grad_norm": 2.9441757202148438,
      "learning_rate": 5.129650257957671e-06,
      "loss": 0.337,
      "num_input_tokens_seen": 11932264,
      "step": 18405
    },
    {
      "epoch": 10.85495283018868,
      "grad_norm": 2.5360002517700195,
      "learning_rate": 5.12707839184897e-06,
      "loss": 0.2995,
      "num_input_tokens_seen": 11936360,
      "step": 18410
    },
    {
      "epoch": 10.857900943396226,
      "grad_norm": 3.34651517868042,
      "learning_rate": 5.124506492095716e-06,
      "loss": 0.3576,
      "num_input_tokens_seen": 11939176,
      "step": 18415
    },
    {
      "epoch": 10.860849056603774,
      "grad_norm": 1.674032211303711,
      "learning_rate": 5.121934559378831e-06,
      "loss": 0.4329,
      "num_input_tokens_seen": 11942440,
      "step": 18420
    },
    {
      "epoch": 10.86379716981132,
      "grad_norm": 4.8111677169799805,
      "learning_rate": 5.1193625943792456e-06,
      "loss": 0.456,
      "num_input_tokens_seen": 11946024,
      "step": 18425
    },
    {
      "epoch": 10.866745283018869,
      "grad_norm": 3.88504958152771,
      "learning_rate": 5.116790597777901e-06,
      "loss": 0.3965,
      "num_input_tokens_seen": 11949896,
      "step": 18430
    },
    {
      "epoch": 10.869693396226415,
      "grad_norm": 1.8373667001724243,
      "learning_rate": 5.11421857025574e-06,
      "loss": 0.2635,
      "num_input_tokens_seen": 11954952,
      "step": 18435
    },
    {
      "epoch": 10.872641509433961,
      "grad_norm": 2.393994092941284,
      "learning_rate": 5.111646512493721e-06,
      "loss": 0.4493,
      "num_input_tokens_seen": 11958280,
      "step": 18440
    },
    {
      "epoch": 10.87558962264151,
      "grad_norm": 2.306305408477783,
      "learning_rate": 5.109074425172806e-06,
      "loss": 0.4255,
      "num_input_tokens_seen": 11961800,
      "step": 18445
    },
    {
      "epoch": 10.878537735849056,
      "grad_norm": 2.0981318950653076,
      "learning_rate": 5.106502308973967e-06,
      "loss": 0.3014,
      "num_input_tokens_seen": 11964360,
      "step": 18450
    },
    {
      "epoch": 10.881485849056604,
      "grad_norm": 2.409668445587158,
      "learning_rate": 5.103930164578184e-06,
      "loss": 0.4169,
      "num_input_tokens_seen": 11968232,
      "step": 18455
    },
    {
      "epoch": 10.88443396226415,
      "grad_norm": 2.9132890701293945,
      "learning_rate": 5.101357992666441e-06,
      "loss": 0.3915,
      "num_input_tokens_seen": 11971208,
      "step": 18460
    },
    {
      "epoch": 10.887382075471699,
      "grad_norm": 5.350638389587402,
      "learning_rate": 5.098785793919733e-06,
      "loss": 0.4493,
      "num_input_tokens_seen": 11974760,
      "step": 18465
    },
    {
      "epoch": 10.890330188679245,
      "grad_norm": 2.8020529747009277,
      "learning_rate": 5.096213569019061e-06,
      "loss": 0.424,
      "num_input_tokens_seen": 11977768,
      "step": 18470
    },
    {
      "epoch": 10.893278301886792,
      "grad_norm": 3.7676844596862793,
      "learning_rate": 5.0936413186454315e-06,
      "loss": 0.4069,
      "num_input_tokens_seen": 11980520,
      "step": 18475
    },
    {
      "epoch": 10.89622641509434,
      "grad_norm": 3.406536102294922,
      "learning_rate": 5.0910690434798584e-06,
      "loss": 0.3037,
      "num_input_tokens_seen": 11983560,
      "step": 18480
    },
    {
      "epoch": 10.899174528301886,
      "grad_norm": 5.035327434539795,
      "learning_rate": 5.088496744203364e-06,
      "loss": 0.4436,
      "num_input_tokens_seen": 11986728,
      "step": 18485
    },
    {
      "epoch": 10.902122641509434,
      "grad_norm": 1.8515726327896118,
      "learning_rate": 5.085924421496976e-06,
      "loss": 0.4801,
      "num_input_tokens_seen": 11989416,
      "step": 18490
    },
    {
      "epoch": 10.90507075471698,
      "grad_norm": 2.2756056785583496,
      "learning_rate": 5.083352076041725e-06,
      "loss": 0.3837,
      "num_input_tokens_seen": 11992552,
      "step": 18495
    },
    {
      "epoch": 10.908018867924529,
      "grad_norm": 2.2534635066986084,
      "learning_rate": 5.080779708518654e-06,
      "loss": 0.3481,
      "num_input_tokens_seen": 11996328,
      "step": 18500
    },
    {
      "epoch": 10.910966981132075,
      "grad_norm": 3.2687461376190186,
      "learning_rate": 5.078207319608807e-06,
      "loss": 0.2785,
      "num_input_tokens_seen": 11999272,
      "step": 18505
    },
    {
      "epoch": 10.913915094339622,
      "grad_norm": 4.036539077758789,
      "learning_rate": 5.075634909993235e-06,
      "loss": 0.2835,
      "num_input_tokens_seen": 12001704,
      "step": 18510
    },
    {
      "epoch": 10.91686320754717,
      "grad_norm": 2.4117202758789062,
      "learning_rate": 5.073062480352995e-06,
      "loss": 0.312,
      "num_input_tokens_seen": 12005000,
      "step": 18515
    },
    {
      "epoch": 10.919811320754716,
      "grad_norm": 3.565826892852783,
      "learning_rate": 5.070490031369149e-06,
      "loss": 0.4246,
      "num_input_tokens_seen": 12007752,
      "step": 18520
    },
    {
      "epoch": 10.922759433962264,
      "grad_norm": 2.5636932849884033,
      "learning_rate": 5.067917563722762e-06,
      "loss": 0.4465,
      "num_input_tokens_seen": 12010344,
      "step": 18525
    },
    {
      "epoch": 10.92570754716981,
      "grad_norm": 2.711946725845337,
      "learning_rate": 5.065345078094907e-06,
      "loss": 0.3983,
      "num_input_tokens_seen": 12014568,
      "step": 18530
    },
    {
      "epoch": 10.928655660377359,
      "grad_norm": 2.6322314739227295,
      "learning_rate": 5.062772575166663e-06,
      "loss": 0.2756,
      "num_input_tokens_seen": 12017128,
      "step": 18535
    },
    {
      "epoch": 10.931603773584905,
      "grad_norm": 3.6295077800750732,
      "learning_rate": 5.0602000556191075e-06,
      "loss": 0.5425,
      "num_input_tokens_seen": 12019880,
      "step": 18540
    },
    {
      "epoch": 10.934551886792454,
      "grad_norm": 1.8766876459121704,
      "learning_rate": 5.0576275201333284e-06,
      "loss": 0.2613,
      "num_input_tokens_seen": 12023112,
      "step": 18545
    },
    {
      "epoch": 10.9375,
      "grad_norm": 2.4803197383880615,
      "learning_rate": 5.055054969390415e-06,
      "loss": 0.3841,
      "num_input_tokens_seen": 12025800,
      "step": 18550
    },
    {
      "epoch": 10.940448113207546,
      "grad_norm": 2.6972923278808594,
      "learning_rate": 5.052482404071461e-06,
      "loss": 0.4051,
      "num_input_tokens_seen": 12029064,
      "step": 18555
    },
    {
      "epoch": 10.943396226415095,
      "grad_norm": 2.5458266735076904,
      "learning_rate": 5.049909824857564e-06,
      "loss": 0.3112,
      "num_input_tokens_seen": 12032200,
      "step": 18560
    },
    {
      "epoch": 10.946344339622641,
      "grad_norm": 2.0103235244750977,
      "learning_rate": 5.047337232429827e-06,
      "loss": 0.3132,
      "num_input_tokens_seen": 12035272,
      "step": 18565
    },
    {
      "epoch": 10.94929245283019,
      "grad_norm": 2.0866539478302,
      "learning_rate": 5.044764627469354e-06,
      "loss": 0.3515,
      "num_input_tokens_seen": 12038120,
      "step": 18570
    },
    {
      "epoch": 10.952240566037736,
      "grad_norm": 3.1262459754943848,
      "learning_rate": 5.042192010657251e-06,
      "loss": 0.4049,
      "num_input_tokens_seen": 12040552,
      "step": 18575
    },
    {
      "epoch": 10.955188679245284,
      "grad_norm": 2.1007652282714844,
      "learning_rate": 5.039619382674632e-06,
      "loss": 0.3101,
      "num_input_tokens_seen": 12044136,
      "step": 18580
    },
    {
      "epoch": 10.95813679245283,
      "grad_norm": 2.296018362045288,
      "learning_rate": 5.0370467442026115e-06,
      "loss": 0.4457,
      "num_input_tokens_seen": 12047400,
      "step": 18585
    },
    {
      "epoch": 10.961084905660378,
      "grad_norm": 2.116945266723633,
      "learning_rate": 5.034474095922304e-06,
      "loss": 0.3398,
      "num_input_tokens_seen": 12051848,
      "step": 18590
    },
    {
      "epoch": 10.964033018867925,
      "grad_norm": 1.8323339223861694,
      "learning_rate": 5.031901438514832e-06,
      "loss": 0.2236,
      "num_input_tokens_seen": 12054856,
      "step": 18595
    },
    {
      "epoch": 10.966981132075471,
      "grad_norm": 5.146734714508057,
      "learning_rate": 5.0293287726613185e-06,
      "loss": 0.3919,
      "num_input_tokens_seen": 12058440,
      "step": 18600
    },
    {
      "epoch": 10.96992924528302,
      "grad_norm": 4.385286808013916,
      "learning_rate": 5.0267560990428836e-06,
      "loss": 0.4109,
      "num_input_tokens_seen": 12061512,
      "step": 18605
    },
    {
      "epoch": 10.972877358490566,
      "grad_norm": 5.312250137329102,
      "learning_rate": 5.024183418340657e-06,
      "loss": 0.2561,
      "num_input_tokens_seen": 12064232,
      "step": 18610
    },
    {
      "epoch": 10.975825471698114,
      "grad_norm": 2.353353261947632,
      "learning_rate": 5.021610731235766e-06,
      "loss": 0.3774,
      "num_input_tokens_seen": 12067656,
      "step": 18615
    },
    {
      "epoch": 10.97877358490566,
      "grad_norm": 3.5858352184295654,
      "learning_rate": 5.01903803840934e-06,
      "loss": 0.3115,
      "num_input_tokens_seen": 12071720,
      "step": 18620
    },
    {
      "epoch": 10.981721698113208,
      "grad_norm": 3.8548624515533447,
      "learning_rate": 5.016465340542514e-06,
      "loss": 0.4796,
      "num_input_tokens_seen": 12074408,
      "step": 18625
    },
    {
      "epoch": 10.984669811320755,
      "grad_norm": 1.7223355770111084,
      "learning_rate": 5.013892638316417e-06,
      "loss": 0.311,
      "num_input_tokens_seen": 12078120,
      "step": 18630
    },
    {
      "epoch": 10.987617924528301,
      "grad_norm": 3.3283917903900146,
      "learning_rate": 5.011319932412182e-06,
      "loss": 0.361,
      "num_input_tokens_seen": 12080680,
      "step": 18635
    },
    {
      "epoch": 10.99056603773585,
      "grad_norm": 1.9583560228347778,
      "learning_rate": 5.008747223510947e-06,
      "loss": 0.3116,
      "num_input_tokens_seen": 12083880,
      "step": 18640
    },
    {
      "epoch": 10.993514150943396,
      "grad_norm": 2.3658690452575684,
      "learning_rate": 5.006174512293849e-06,
      "loss": 0.3393,
      "num_input_tokens_seen": 12087144,
      "step": 18645
    },
    {
      "epoch": 10.996462264150944,
      "grad_norm": 3.0746712684631348,
      "learning_rate": 5.003601799442019e-06,
      "loss": 0.3115,
      "num_input_tokens_seen": 12089928,
      "step": 18650
    },
    {
      "epoch": 10.99941037735849,
      "grad_norm": 2.544196128845215,
      "learning_rate": 5.0010290856366e-06,
      "loss": 0.4358,
      "num_input_tokens_seen": 12092968,
      "step": 18655
    },
    {
      "epoch": 11.002358490566039,
      "grad_norm": 8.492934226989746,
      "learning_rate": 4.998456371558726e-06,
      "loss": 0.3826,
      "num_input_tokens_seen": 12096480,
      "step": 18660
    },
    {
      "epoch": 11.005306603773585,
      "grad_norm": 3.3224849700927734,
      "learning_rate": 4.995883657889531e-06,
      "loss": 0.4989,
      "num_input_tokens_seen": 12099200,
      "step": 18665
    },
    {
      "epoch": 11.008254716981131,
      "grad_norm": 2.6026594638824463,
      "learning_rate": 4.993310945310158e-06,
      "loss": 0.4403,
      "num_input_tokens_seen": 12102528,
      "step": 18670
    },
    {
      "epoch": 11.01120283018868,
      "grad_norm": 2.7927350997924805,
      "learning_rate": 4.9907382345017416e-06,
      "loss": 0.3302,
      "num_input_tokens_seen": 12106176,
      "step": 18675
    },
    {
      "epoch": 11.014150943396226,
      "grad_norm": 2.3090245723724365,
      "learning_rate": 4.988165526145416e-06,
      "loss": 0.378,
      "num_input_tokens_seen": 12109152,
      "step": 18680
    },
    {
      "epoch": 11.017099056603774,
      "grad_norm": 3.0498993396759033,
      "learning_rate": 4.985592820922319e-06,
      "loss": 0.3202,
      "num_input_tokens_seen": 12113216,
      "step": 18685
    },
    {
      "epoch": 11.02004716981132,
      "grad_norm": 3.408412218093872,
      "learning_rate": 4.983020119513586e-06,
      "loss": 0.376,
      "num_input_tokens_seen": 12115904,
      "step": 18690
    },
    {
      "epoch": 11.022995283018869,
      "grad_norm": 2.980828285217285,
      "learning_rate": 4.9804474226003465e-06,
      "loss": 0.3846,
      "num_input_tokens_seen": 12118656,
      "step": 18695
    },
    {
      "epoch": 11.025943396226415,
      "grad_norm": 1.85659921169281,
      "learning_rate": 4.977874730863739e-06,
      "loss": 0.2913,
      "num_input_tokens_seen": 12122048,
      "step": 18700
    },
    {
      "epoch": 11.028891509433961,
      "grad_norm": 3.033390760421753,
      "learning_rate": 4.975302044984889e-06,
      "loss": 0.391,
      "num_input_tokens_seen": 12124608,
      "step": 18705
    },
    {
      "epoch": 11.03183962264151,
      "grad_norm": 1.697177529335022,
      "learning_rate": 4.972729365644931e-06,
      "loss": 0.3698,
      "num_input_tokens_seen": 12127744,
      "step": 18710
    },
    {
      "epoch": 11.034787735849056,
      "grad_norm": 2.8333616256713867,
      "learning_rate": 4.97015669352499e-06,
      "loss": 0.3492,
      "num_input_tokens_seen": 12131808,
      "step": 18715
    },
    {
      "epoch": 11.037735849056604,
      "grad_norm": 4.193706035614014,
      "learning_rate": 4.967584029306194e-06,
      "loss": 0.4043,
      "num_input_tokens_seen": 12134912,
      "step": 18720
    },
    {
      "epoch": 11.04068396226415,
      "grad_norm": 3.901806116104126,
      "learning_rate": 4.965011373669666e-06,
      "loss": 0.3034,
      "num_input_tokens_seen": 12138016,
      "step": 18725
    },
    {
      "epoch": 11.043632075471699,
      "grad_norm": 3.151733160018921,
      "learning_rate": 4.962438727296527e-06,
      "loss": 0.472,
      "num_input_tokens_seen": 12140864,
      "step": 18730
    },
    {
      "epoch": 11.046580188679245,
      "grad_norm": 4.002111434936523,
      "learning_rate": 4.959866090867897e-06,
      "loss": 0.3276,
      "num_input_tokens_seen": 12144096,
      "step": 18735
    },
    {
      "epoch": 11.049528301886792,
      "grad_norm": 3.1241276264190674,
      "learning_rate": 4.957293465064893e-06,
      "loss": 0.375,
      "num_input_tokens_seen": 12147040,
      "step": 18740
    },
    {
      "epoch": 11.05247641509434,
      "grad_norm": 1.9345872402191162,
      "learning_rate": 4.954720850568627e-06,
      "loss": 0.2758,
      "num_input_tokens_seen": 12150272,
      "step": 18745
    },
    {
      "epoch": 11.055424528301886,
      "grad_norm": 1.6182947158813477,
      "learning_rate": 4.952148248060212e-06,
      "loss": 0.2342,
      "num_input_tokens_seen": 12153184,
      "step": 18750
    },
    {
      "epoch": 11.058372641509434,
      "grad_norm": 2.470810651779175,
      "learning_rate": 4.949575658220755e-06,
      "loss": 0.349,
      "num_input_tokens_seen": 12156032,
      "step": 18755
    },
    {
      "epoch": 11.06132075471698,
      "grad_norm": 3.324422836303711,
      "learning_rate": 4.947003081731359e-06,
      "loss": 0.3827,
      "num_input_tokens_seen": 12159136,
      "step": 18760
    },
    {
      "epoch": 11.064268867924529,
      "grad_norm": 1.4348106384277344,
      "learning_rate": 4.944430519273126e-06,
      "loss": 0.2591,
      "num_input_tokens_seen": 12162816,
      "step": 18765
    },
    {
      "epoch": 11.067216981132075,
      "grad_norm": 4.020450592041016,
      "learning_rate": 4.941857971527152e-06,
      "loss": 0.2441,
      "num_input_tokens_seen": 12166688,
      "step": 18770
    },
    {
      "epoch": 11.070165094339623,
      "grad_norm": 3.36336088180542,
      "learning_rate": 4.93928543917453e-06,
      "loss": 0.3741,
      "num_input_tokens_seen": 12169664,
      "step": 18775
    },
    {
      "epoch": 11.07311320754717,
      "grad_norm": 5.7995476722717285,
      "learning_rate": 4.93671292289635e-06,
      "loss": 0.3743,
      "num_input_tokens_seen": 12173216,
      "step": 18780
    },
    {
      "epoch": 11.076061320754716,
      "grad_norm": 1.5349862575531006,
      "learning_rate": 4.934140423373698e-06,
      "loss": 0.2466,
      "num_input_tokens_seen": 12177088,
      "step": 18785
    },
    {
      "epoch": 11.079009433962264,
      "grad_norm": 3.237405300140381,
      "learning_rate": 4.931567941287651e-06,
      "loss": 0.2964,
      "num_input_tokens_seen": 12180352,
      "step": 18790
    },
    {
      "epoch": 11.08195754716981,
      "grad_norm": 3.110607385635376,
      "learning_rate": 4.9289954773192875e-06,
      "loss": 0.3208,
      "num_input_tokens_seen": 12183872,
      "step": 18795
    },
    {
      "epoch": 11.084905660377359,
      "grad_norm": 2.4838240146636963,
      "learning_rate": 4.926423032149677e-06,
      "loss": 0.3876,
      "num_input_tokens_seen": 12187648,
      "step": 18800
    },
    {
      "epoch": 11.087853773584905,
      "grad_norm": 2.476289749145508,
      "learning_rate": 4.923850606459883e-06,
      "loss": 0.4229,
      "num_input_tokens_seen": 12190272,
      "step": 18805
    },
    {
      "epoch": 11.090801886792454,
      "grad_norm": 3.3942759037017822,
      "learning_rate": 4.921278200930972e-06,
      "loss": 0.3632,
      "num_input_tokens_seen": 12193440,
      "step": 18810
    },
    {
      "epoch": 11.09375,
      "grad_norm": 2.0407707691192627,
      "learning_rate": 4.918705816243996e-06,
      "loss": 0.3615,
      "num_input_tokens_seen": 12196448,
      "step": 18815
    },
    {
      "epoch": 11.096698113207546,
      "grad_norm": 1.9773268699645996,
      "learning_rate": 4.916133453080007e-06,
      "loss": 0.4067,
      "num_input_tokens_seen": 12199840,
      "step": 18820
    },
    {
      "epoch": 11.099646226415095,
      "grad_norm": 3.224524974822998,
      "learning_rate": 4.913561112120046e-06,
      "loss": 0.343,
      "num_input_tokens_seen": 12203264,
      "step": 18825
    },
    {
      "epoch": 11.102594339622641,
      "grad_norm": 2.331324577331543,
      "learning_rate": 4.910988794045154e-06,
      "loss": 0.3539,
      "num_input_tokens_seen": 12206432,
      "step": 18830
    },
    {
      "epoch": 11.10554245283019,
      "grad_norm": 3.3474831581115723,
      "learning_rate": 4.9084164995363626e-06,
      "loss": 0.3949,
      "num_input_tokens_seen": 12209920,
      "step": 18835
    },
    {
      "epoch": 11.108490566037736,
      "grad_norm": 3.9454805850982666,
      "learning_rate": 4.905844229274697e-06,
      "loss": 0.2666,
      "num_input_tokens_seen": 12212928,
      "step": 18840
    },
    {
      "epoch": 11.111438679245284,
      "grad_norm": 3.76985764503479,
      "learning_rate": 4.903271983941177e-06,
      "loss": 0.4624,
      "num_input_tokens_seen": 12216352,
      "step": 18845
    },
    {
      "epoch": 11.11438679245283,
      "grad_norm": 2.0271503925323486,
      "learning_rate": 4.900699764216818e-06,
      "loss": 0.3312,
      "num_input_tokens_seen": 12219904,
      "step": 18850
    },
    {
      "epoch": 11.117334905660377,
      "grad_norm": 1.6836637258529663,
      "learning_rate": 4.898127570782622e-06,
      "loss": 0.3023,
      "num_input_tokens_seen": 12222944,
      "step": 18855
    },
    {
      "epoch": 11.120283018867925,
      "grad_norm": 2.591106414794922,
      "learning_rate": 4.895555404319592e-06,
      "loss": 0.3204,
      "num_input_tokens_seen": 12226176,
      "step": 18860
    },
    {
      "epoch": 11.123231132075471,
      "grad_norm": 2.8185653686523438,
      "learning_rate": 4.89298326550872e-06,
      "loss": 0.5495,
      "num_input_tokens_seen": 12229856,
      "step": 18865
    },
    {
      "epoch": 11.12617924528302,
      "grad_norm": 2.9910998344421387,
      "learning_rate": 4.8904111550309876e-06,
      "loss": 0.4039,
      "num_input_tokens_seen": 12233248,
      "step": 18870
    },
    {
      "epoch": 11.129127358490566,
      "grad_norm": 2.2931058406829834,
      "learning_rate": 4.8878390735673755e-06,
      "loss": 0.3697,
      "num_input_tokens_seen": 12236736,
      "step": 18875
    },
    {
      "epoch": 11.132075471698114,
      "grad_norm": 1.9255762100219727,
      "learning_rate": 4.8852670217988505e-06,
      "loss": 0.291,
      "num_input_tokens_seen": 12239552,
      "step": 18880
    },
    {
      "epoch": 11.13502358490566,
      "grad_norm": 2.264495849609375,
      "learning_rate": 4.882695000406377e-06,
      "loss": 0.3432,
      "num_input_tokens_seen": 12243872,
      "step": 18885
    },
    {
      "epoch": 11.137971698113208,
      "grad_norm": 1.8108818531036377,
      "learning_rate": 4.880123010070909e-06,
      "loss": 0.3251,
      "num_input_tokens_seen": 12247552,
      "step": 18890
    },
    {
      "epoch": 11.140919811320755,
      "grad_norm": 5.492410182952881,
      "learning_rate": 4.877551051473388e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 12250784,
      "step": 18895
    },
    {
      "epoch": 11.143867924528301,
      "grad_norm": 3.8166182041168213,
      "learning_rate": 4.874979125294755e-06,
      "loss": 0.3136,
      "num_input_tokens_seen": 12253568,
      "step": 18900
    },
    {
      "epoch": 11.14681603773585,
      "grad_norm": 2.3378689289093018,
      "learning_rate": 4.872407232215937e-06,
      "loss": 0.3682,
      "num_input_tokens_seen": 12256608,
      "step": 18905
    },
    {
      "epoch": 11.149764150943396,
      "grad_norm": 3.479260206222534,
      "learning_rate": 4.8698353729178546e-06,
      "loss": 0.3816,
      "num_input_tokens_seen": 12259552,
      "step": 18910
    },
    {
      "epoch": 11.152712264150944,
      "grad_norm": 2.6635355949401855,
      "learning_rate": 4.867263548081418e-06,
      "loss": 0.3032,
      "num_input_tokens_seen": 12262304,
      "step": 18915
    },
    {
      "epoch": 11.15566037735849,
      "grad_norm": 1.5909487009048462,
      "learning_rate": 4.8646917583875304e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 12266144,
      "step": 18920
    },
    {
      "epoch": 11.158608490566039,
      "grad_norm": 2.7359273433685303,
      "learning_rate": 4.862120004517082e-06,
      "loss": 0.3454,
      "num_input_tokens_seen": 12268864,
      "step": 18925
    },
    {
      "epoch": 11.161556603773585,
      "grad_norm": 6.483883857727051,
      "learning_rate": 4.859548287150956e-06,
      "loss": 0.3815,
      "num_input_tokens_seen": 12272480,
      "step": 18930
    },
    {
      "epoch": 11.164504716981131,
      "grad_norm": 4.142387390136719,
      "learning_rate": 4.8569766069700275e-06,
      "loss": 0.268,
      "num_input_tokens_seen": 12274944,
      "step": 18935
    },
    {
      "epoch": 11.16745283018868,
      "grad_norm": 2.7717807292938232,
      "learning_rate": 4.854404964655158e-06,
      "loss": 0.2887,
      "num_input_tokens_seen": 12277984,
      "step": 18940
    },
    {
      "epoch": 11.170400943396226,
      "grad_norm": 2.746509552001953,
      "learning_rate": 4.8518333608872015e-06,
      "loss": 0.2926,
      "num_input_tokens_seen": 12280992,
      "step": 18945
    },
    {
      "epoch": 11.173349056603774,
      "grad_norm": 4.050553321838379,
      "learning_rate": 4.849261796347002e-06,
      "loss": 0.3366,
      "num_input_tokens_seen": 12285088,
      "step": 18950
    },
    {
      "epoch": 11.17629716981132,
      "grad_norm": 3.866788864135742,
      "learning_rate": 4.846690271715391e-06,
      "loss": 0.4163,
      "num_input_tokens_seen": 12289216,
      "step": 18955
    },
    {
      "epoch": 11.179245283018869,
      "grad_norm": 1.6248743534088135,
      "learning_rate": 4.844118787673191e-06,
      "loss": 0.2741,
      "num_input_tokens_seen": 12292096,
      "step": 18960
    },
    {
      "epoch": 11.182193396226415,
      "grad_norm": 2.7555477619171143,
      "learning_rate": 4.841547344901214e-06,
      "loss": 0.4072,
      "num_input_tokens_seen": 12295232,
      "step": 18965
    },
    {
      "epoch": 11.185141509433961,
      "grad_norm": 2.0514938831329346,
      "learning_rate": 4.838975944080261e-06,
      "loss": 0.383,
      "num_input_tokens_seen": 12299200,
      "step": 18970
    },
    {
      "epoch": 11.18808962264151,
      "grad_norm": 2.4448204040527344,
      "learning_rate": 4.83640458589112e-06,
      "loss": 0.3879,
      "num_input_tokens_seen": 12302784,
      "step": 18975
    },
    {
      "epoch": 11.191037735849056,
      "grad_norm": 3.0353407859802246,
      "learning_rate": 4.833833271014571e-06,
      "loss": 0.3501,
      "num_input_tokens_seen": 12306400,
      "step": 18980
    },
    {
      "epoch": 11.193985849056604,
      "grad_norm": 1.580006718635559,
      "learning_rate": 4.831262000131379e-06,
      "loss": 0.4613,
      "num_input_tokens_seen": 12310560,
      "step": 18985
    },
    {
      "epoch": 11.19693396226415,
      "grad_norm": 2.7753102779388428,
      "learning_rate": 4.828690773922299e-06,
      "loss": 0.3917,
      "num_input_tokens_seen": 12315264,
      "step": 18990
    },
    {
      "epoch": 11.199882075471699,
      "grad_norm": 3.4600205421447754,
      "learning_rate": 4.826119593068074e-06,
      "loss": 0.3527,
      "num_input_tokens_seen": 12318848,
      "step": 18995
    },
    {
      "epoch": 11.202830188679245,
      "grad_norm": 2.520820379257202,
      "learning_rate": 4.8235484582494375e-06,
      "loss": 0.2871,
      "num_input_tokens_seen": 12322208,
      "step": 19000
    },
    {
      "epoch": 11.205778301886792,
      "grad_norm": 3.347960948944092,
      "learning_rate": 4.8209773701471076e-06,
      "loss": 0.3351,
      "num_input_tokens_seen": 12324864,
      "step": 19005
    },
    {
      "epoch": 11.20872641509434,
      "grad_norm": 1.8796011209487915,
      "learning_rate": 4.818406329441789e-06,
      "loss": 0.3444,
      "num_input_tokens_seen": 12328576,
      "step": 19010
    },
    {
      "epoch": 11.211674528301886,
      "grad_norm": 2.391706705093384,
      "learning_rate": 4.815835336814179e-06,
      "loss": 0.3475,
      "num_input_tokens_seen": 12331904,
      "step": 19015
    },
    {
      "epoch": 11.214622641509434,
      "grad_norm": 3.685561180114746,
      "learning_rate": 4.813264392944957e-06,
      "loss": 0.3557,
      "num_input_tokens_seen": 12335040,
      "step": 19020
    },
    {
      "epoch": 11.21757075471698,
      "grad_norm": 3.2088685035705566,
      "learning_rate": 4.8106934985147905e-06,
      "loss": 0.345,
      "num_input_tokens_seen": 12338240,
      "step": 19025
    },
    {
      "epoch": 11.220518867924529,
      "grad_norm": 2.001336097717285,
      "learning_rate": 4.808122654204338e-06,
      "loss": 0.3371,
      "num_input_tokens_seen": 12341824,
      "step": 19030
    },
    {
      "epoch": 11.223466981132075,
      "grad_norm": 3.6166911125183105,
      "learning_rate": 4.805551860694239e-06,
      "loss": 0.3818,
      "num_input_tokens_seen": 12344736,
      "step": 19035
    },
    {
      "epoch": 11.226415094339623,
      "grad_norm": 3.634016275405884,
      "learning_rate": 4.802981118665122e-06,
      "loss": 0.3744,
      "num_input_tokens_seen": 12348640,
      "step": 19040
    },
    {
      "epoch": 11.22936320754717,
      "grad_norm": 3.965724468231201,
      "learning_rate": 4.800410428797604e-06,
      "loss": 0.4307,
      "num_input_tokens_seen": 12351456,
      "step": 19045
    },
    {
      "epoch": 11.232311320754716,
      "grad_norm": 2.8130578994750977,
      "learning_rate": 4.797839791772286e-06,
      "loss": 0.2406,
      "num_input_tokens_seen": 12354560,
      "step": 19050
    },
    {
      "epoch": 11.235259433962264,
      "grad_norm": 3.6034860610961914,
      "learning_rate": 4.795269208269752e-06,
      "loss": 0.3454,
      "num_input_tokens_seen": 12358144,
      "step": 19055
    },
    {
      "epoch": 11.23820754716981,
      "grad_norm": 1.7778923511505127,
      "learning_rate": 4.792698678970579e-06,
      "loss": 0.3459,
      "num_input_tokens_seen": 12360768,
      "step": 19060
    },
    {
      "epoch": 11.241155660377359,
      "grad_norm": 3.3241758346557617,
      "learning_rate": 4.7901282045553245e-06,
      "loss": 0.4164,
      "num_input_tokens_seen": 12364992,
      "step": 19065
    },
    {
      "epoch": 11.244103773584905,
      "grad_norm": 4.894493103027344,
      "learning_rate": 4.787557785704531e-06,
      "loss": 0.4886,
      "num_input_tokens_seen": 12367936,
      "step": 19070
    },
    {
      "epoch": 11.247051886792454,
      "grad_norm": 2.019251585006714,
      "learning_rate": 4.784987423098731e-06,
      "loss": 0.2897,
      "num_input_tokens_seen": 12371392,
      "step": 19075
    },
    {
      "epoch": 11.25,
      "grad_norm": 5.5597357749938965,
      "learning_rate": 4.7824171174184354e-06,
      "loss": 0.3559,
      "num_input_tokens_seen": 12374336,
      "step": 19080
    },
    {
      "epoch": 11.252948113207546,
      "grad_norm": 2.860588550567627,
      "learning_rate": 4.779846869344146e-06,
      "loss": 0.31,
      "num_input_tokens_seen": 12377344,
      "step": 19085
    },
    {
      "epoch": 11.255896226415095,
      "grad_norm": 5.547386169433594,
      "learning_rate": 4.777276679556346e-06,
      "loss": 0.3038,
      "num_input_tokens_seen": 12381568,
      "step": 19090
    },
    {
      "epoch": 11.258844339622641,
      "grad_norm": 4.156414985656738,
      "learning_rate": 4.774706548735507e-06,
      "loss": 0.4088,
      "num_input_tokens_seen": 12384928,
      "step": 19095
    },
    {
      "epoch": 11.26179245283019,
      "grad_norm": 2.699496030807495,
      "learning_rate": 4.77213647756208e-06,
      "loss": 0.379,
      "num_input_tokens_seen": 12389120,
      "step": 19100
    },
    {
      "epoch": 11.264740566037736,
      "grad_norm": 3.169552803039551,
      "learning_rate": 4.769566466716501e-06,
      "loss": 0.4025,
      "num_input_tokens_seen": 12392032,
      "step": 19105
    },
    {
      "epoch": 11.267688679245284,
      "grad_norm": 2.775785446166992,
      "learning_rate": 4.766996516879195e-06,
      "loss": 0.4659,
      "num_input_tokens_seen": 12395776,
      "step": 19110
    },
    {
      "epoch": 11.27063679245283,
      "grad_norm": 2.8730688095092773,
      "learning_rate": 4.764426628730564e-06,
      "loss": 0.2907,
      "num_input_tokens_seen": 12398848,
      "step": 19115
    },
    {
      "epoch": 11.273584905660377,
      "grad_norm": 8.34559440612793,
      "learning_rate": 4.761856802950999e-06,
      "loss": 0.3193,
      "num_input_tokens_seen": 12401600,
      "step": 19120
    },
    {
      "epoch": 11.276533018867925,
      "grad_norm": 3.7528812885284424,
      "learning_rate": 4.759287040220872e-06,
      "loss": 0.4256,
      "num_input_tokens_seen": 12404768,
      "step": 19125
    },
    {
      "epoch": 11.279481132075471,
      "grad_norm": 3.3103067874908447,
      "learning_rate": 4.756717341220538e-06,
      "loss": 0.4363,
      "num_input_tokens_seen": 12408544,
      "step": 19130
    },
    {
      "epoch": 11.28242924528302,
      "grad_norm": 3.998521566390991,
      "learning_rate": 4.7541477066303365e-06,
      "loss": 0.3392,
      "num_input_tokens_seen": 12411488,
      "step": 19135
    },
    {
      "epoch": 11.285377358490566,
      "grad_norm": 1.967769980430603,
      "learning_rate": 4.75157813713059e-06,
      "loss": 0.322,
      "num_input_tokens_seen": 12414464,
      "step": 19140
    },
    {
      "epoch": 11.288325471698114,
      "grad_norm": 3.5130562782287598,
      "learning_rate": 4.7490086334016e-06,
      "loss": 0.3038,
      "num_input_tokens_seen": 12417888,
      "step": 19145
    },
    {
      "epoch": 11.29127358490566,
      "grad_norm": 1.8622406721115112,
      "learning_rate": 4.746439196123659e-06,
      "loss": 0.3642,
      "num_input_tokens_seen": 12421248,
      "step": 19150
    },
    {
      "epoch": 11.294221698113208,
      "grad_norm": 8.694835662841797,
      "learning_rate": 4.743869825977032e-06,
      "loss": 0.4168,
      "num_input_tokens_seen": 12424160,
      "step": 19155
    },
    {
      "epoch": 11.297169811320755,
      "grad_norm": 2.481052875518799,
      "learning_rate": 4.741300523641972e-06,
      "loss": 0.3415,
      "num_input_tokens_seen": 12427072,
      "step": 19160
    },
    {
      "epoch": 11.300117924528301,
      "grad_norm": 6.005927085876465,
      "learning_rate": 4.738731289798715e-06,
      "loss": 0.3539,
      "num_input_tokens_seen": 12430112,
      "step": 19165
    },
    {
      "epoch": 11.30306603773585,
      "grad_norm": 2.7829573154449463,
      "learning_rate": 4.7361621251274744e-06,
      "loss": 0.3864,
      "num_input_tokens_seen": 12433120,
      "step": 19170
    },
    {
      "epoch": 11.306014150943396,
      "grad_norm": 1.8188482522964478,
      "learning_rate": 4.733593030308446e-06,
      "loss": 0.2844,
      "num_input_tokens_seen": 12435872,
      "step": 19175
    },
    {
      "epoch": 11.308962264150944,
      "grad_norm": 3.954718828201294,
      "learning_rate": 4.731024006021814e-06,
      "loss": 0.2958,
      "num_input_tokens_seen": 12438080,
      "step": 19180
    },
    {
      "epoch": 11.31191037735849,
      "grad_norm": 6.261547565460205,
      "learning_rate": 4.728455052947732e-06,
      "loss": 0.3652,
      "num_input_tokens_seen": 12440992,
      "step": 19185
    },
    {
      "epoch": 11.314858490566039,
      "grad_norm": 2.616736888885498,
      "learning_rate": 4.725886171766349e-06,
      "loss": 0.3923,
      "num_input_tokens_seen": 12444256,
      "step": 19190
    },
    {
      "epoch": 11.317806603773585,
      "grad_norm": 2.0334155559539795,
      "learning_rate": 4.723317363157781e-06,
      "loss": 0.4427,
      "num_input_tokens_seen": 12447392,
      "step": 19195
    },
    {
      "epoch": 11.320754716981131,
      "grad_norm": 2.34916615486145,
      "learning_rate": 4.720748627802135e-06,
      "loss": 0.3816,
      "num_input_tokens_seen": 12450528,
      "step": 19200
    },
    {
      "epoch": 11.32370283018868,
      "grad_norm": 3.519145965576172,
      "learning_rate": 4.718179966379492e-06,
      "loss": 0.3789,
      "num_input_tokens_seen": 12453280,
      "step": 19205
    },
    {
      "epoch": 11.326650943396226,
      "grad_norm": 5.2326812744140625,
      "learning_rate": 4.715611379569919e-06,
      "loss": 0.3289,
      "num_input_tokens_seen": 12455936,
      "step": 19210
    },
    {
      "epoch": 11.329599056603774,
      "grad_norm": 2.6825101375579834,
      "learning_rate": 4.713042868053458e-06,
      "loss": 0.2848,
      "num_input_tokens_seen": 12459616,
      "step": 19215
    },
    {
      "epoch": 11.33254716981132,
      "grad_norm": 7.5283942222595215,
      "learning_rate": 4.7104744325101345e-06,
      "loss": 0.4229,
      "num_input_tokens_seen": 12462656,
      "step": 19220
    },
    {
      "epoch": 11.335495283018869,
      "grad_norm": 2.1662869453430176,
      "learning_rate": 4.7079060736199525e-06,
      "loss": 0.3364,
      "num_input_tokens_seen": 12465952,
      "step": 19225
    },
    {
      "epoch": 11.338443396226415,
      "grad_norm": 4.31026554107666,
      "learning_rate": 4.705337792062897e-06,
      "loss": 0.4038,
      "num_input_tokens_seen": 12468192,
      "step": 19230
    },
    {
      "epoch": 11.341391509433961,
      "grad_norm": 3.4834890365600586,
      "learning_rate": 4.702769588518931e-06,
      "loss": 0.3288,
      "num_input_tokens_seen": 12472352,
      "step": 19235
    },
    {
      "epoch": 11.34433962264151,
      "grad_norm": 4.062103271484375,
      "learning_rate": 4.700201463667996e-06,
      "loss": 0.4053,
      "num_input_tokens_seen": 12474880,
      "step": 19240
    },
    {
      "epoch": 11.347287735849056,
      "grad_norm": 2.8031575679779053,
      "learning_rate": 4.697633418190017e-06,
      "loss": 0.4369,
      "num_input_tokens_seen": 12477664,
      "step": 19245
    },
    {
      "epoch": 11.350235849056604,
      "grad_norm": 4.711950778961182,
      "learning_rate": 4.695065452764893e-06,
      "loss": 0.2918,
      "num_input_tokens_seen": 12481536,
      "step": 19250
    },
    {
      "epoch": 11.35318396226415,
      "grad_norm": 2.056403636932373,
      "learning_rate": 4.692497568072505e-06,
      "loss": 0.3654,
      "num_input_tokens_seen": 12484000,
      "step": 19255
    },
    {
      "epoch": 11.356132075471699,
      "grad_norm": 3.6916656494140625,
      "learning_rate": 4.689929764792711e-06,
      "loss": 0.4842,
      "num_input_tokens_seen": 12486656,
      "step": 19260
    },
    {
      "epoch": 11.359080188679245,
      "grad_norm": 2.528005361557007,
      "learning_rate": 4.687362043605349e-06,
      "loss": 0.3562,
      "num_input_tokens_seen": 12490080,
      "step": 19265
    },
    {
      "epoch": 11.362028301886792,
      "grad_norm": 3.424862861633301,
      "learning_rate": 4.684794405190231e-06,
      "loss": 0.4682,
      "num_input_tokens_seen": 12495008,
      "step": 19270
    },
    {
      "epoch": 11.36497641509434,
      "grad_norm": 4.320531368255615,
      "learning_rate": 4.682226850227155e-06,
      "loss": 0.3576,
      "num_input_tokens_seen": 12497824,
      "step": 19275
    },
    {
      "epoch": 11.367924528301886,
      "grad_norm": 2.2795510292053223,
      "learning_rate": 4.6796593793958875e-06,
      "loss": 0.362,
      "num_input_tokens_seen": 12500992,
      "step": 19280
    },
    {
      "epoch": 11.370872641509434,
      "grad_norm": 5.606743335723877,
      "learning_rate": 4.6770919933761815e-06,
      "loss": 0.3312,
      "num_input_tokens_seen": 12503616,
      "step": 19285
    },
    {
      "epoch": 11.37382075471698,
      "grad_norm": 1.7396124601364136,
      "learning_rate": 4.6745246928477615e-06,
      "loss": 0.2942,
      "num_input_tokens_seen": 12507328,
      "step": 19290
    },
    {
      "epoch": 11.376768867924529,
      "grad_norm": 4.91544246673584,
      "learning_rate": 4.671957478490332e-06,
      "loss": 0.4124,
      "num_input_tokens_seen": 12510912,
      "step": 19295
    },
    {
      "epoch": 11.379716981132075,
      "grad_norm": 3.134603500366211,
      "learning_rate": 4.669390350983574e-06,
      "loss": 0.374,
      "num_input_tokens_seen": 12513984,
      "step": 19300
    },
    {
      "epoch": 11.382665094339623,
      "grad_norm": 2.9249908924102783,
      "learning_rate": 4.666823311007145e-06,
      "loss": 0.4595,
      "num_input_tokens_seen": 12518720,
      "step": 19305
    },
    {
      "epoch": 11.38561320754717,
      "grad_norm": 4.506790637969971,
      "learning_rate": 4.664256359240681e-06,
      "loss": 0.3284,
      "num_input_tokens_seen": 12521728,
      "step": 19310
    },
    {
      "epoch": 11.388561320754716,
      "grad_norm": 6.547369003295898,
      "learning_rate": 4.661689496363793e-06,
      "loss": 0.3379,
      "num_input_tokens_seen": 12524416,
      "step": 19315
    },
    {
      "epoch": 11.391509433962264,
      "grad_norm": 3.2802059650421143,
      "learning_rate": 4.659122723056068e-06,
      "loss": 0.294,
      "num_input_tokens_seen": 12527200,
      "step": 19320
    },
    {
      "epoch": 11.39445754716981,
      "grad_norm": 2.9519495964050293,
      "learning_rate": 4.656556039997072e-06,
      "loss": 0.2726,
      "num_input_tokens_seen": 12530656,
      "step": 19325
    },
    {
      "epoch": 11.397405660377359,
      "grad_norm": 2.498974561691284,
      "learning_rate": 4.653989447866345e-06,
      "loss": 0.4249,
      "num_input_tokens_seen": 12538272,
      "step": 19330
    },
    {
      "epoch": 11.400353773584905,
      "grad_norm": 5.044483184814453,
      "learning_rate": 4.651422947343401e-06,
      "loss": 0.2859,
      "num_input_tokens_seen": 12541216,
      "step": 19335
    },
    {
      "epoch": 11.403301886792454,
      "grad_norm": 3.178744316101074,
      "learning_rate": 4.6488565391077355e-06,
      "loss": 0.406,
      "num_input_tokens_seen": 12543776,
      "step": 19340
    },
    {
      "epoch": 11.40625,
      "grad_norm": 3.537202835083008,
      "learning_rate": 4.646290223838815e-06,
      "loss": 0.2594,
      "num_input_tokens_seen": 12546912,
      "step": 19345
    },
    {
      "epoch": 11.409198113207546,
      "grad_norm": 3.7642405033111572,
      "learning_rate": 4.64372400221608e-06,
      "loss": 0.4028,
      "num_input_tokens_seen": 12549984,
      "step": 19350
    },
    {
      "epoch": 11.412146226415095,
      "grad_norm": 1.9804526567459106,
      "learning_rate": 4.641157874918953e-06,
      "loss": 0.3103,
      "num_input_tokens_seen": 12554848,
      "step": 19355
    },
    {
      "epoch": 11.415094339622641,
      "grad_norm": 4.2121500968933105,
      "learning_rate": 4.6385918426268245e-06,
      "loss": 0.3412,
      "num_input_tokens_seen": 12557248,
      "step": 19360
    },
    {
      "epoch": 11.41804245283019,
      "grad_norm": 2.961444616317749,
      "learning_rate": 4.636025906019062e-06,
      "loss": 0.35,
      "num_input_tokens_seen": 12560448,
      "step": 19365
    },
    {
      "epoch": 11.420990566037736,
      "grad_norm": 3.0886390209198,
      "learning_rate": 4.6334600657750115e-06,
      "loss": 0.3307,
      "num_input_tokens_seen": 12564256,
      "step": 19370
    },
    {
      "epoch": 11.423938679245284,
      "grad_norm": 3.4152920246124268,
      "learning_rate": 4.6308943225739855e-06,
      "loss": 0.394,
      "num_input_tokens_seen": 12567808,
      "step": 19375
    },
    {
      "epoch": 11.42688679245283,
      "grad_norm": 2.0100255012512207,
      "learning_rate": 4.628328677095281e-06,
      "loss": 0.3199,
      "num_input_tokens_seen": 12571424,
      "step": 19380
    },
    {
      "epoch": 11.429834905660377,
      "grad_norm": 2.428969144821167,
      "learning_rate": 4.625763130018159e-06,
      "loss": 0.346,
      "num_input_tokens_seen": 12574528,
      "step": 19385
    },
    {
      "epoch": 11.432783018867925,
      "grad_norm": 1.9437004327774048,
      "learning_rate": 4.6231976820218635e-06,
      "loss": 0.3288,
      "num_input_tokens_seen": 12578016,
      "step": 19390
    },
    {
      "epoch": 11.435731132075471,
      "grad_norm": 2.724034547805786,
      "learning_rate": 4.620632333785605e-06,
      "loss": 0.3534,
      "num_input_tokens_seen": 12580544,
      "step": 19395
    },
    {
      "epoch": 11.43867924528302,
      "grad_norm": 6.765618324279785,
      "learning_rate": 4.618067085988569e-06,
      "loss": 0.5237,
      "num_input_tokens_seen": 12582752,
      "step": 19400
    },
    {
      "epoch": 11.441627358490566,
      "grad_norm": 3.3572797775268555,
      "learning_rate": 4.615501939309921e-06,
      "loss": 0.2583,
      "num_input_tokens_seen": 12585856,
      "step": 19405
    },
    {
      "epoch": 11.444575471698114,
      "grad_norm": 3.019223690032959,
      "learning_rate": 4.612936894428791e-06,
      "loss": 0.3347,
      "num_input_tokens_seen": 12589888,
      "step": 19410
    },
    {
      "epoch": 11.44752358490566,
      "grad_norm": 2.4157702922821045,
      "learning_rate": 4.610371952024285e-06,
      "loss": 0.3876,
      "num_input_tokens_seen": 12593024,
      "step": 19415
    },
    {
      "epoch": 11.450471698113208,
      "grad_norm": 3.047930955886841,
      "learning_rate": 4.607807112775485e-06,
      "loss": 0.2837,
      "num_input_tokens_seen": 12596352,
      "step": 19420
    },
    {
      "epoch": 11.453419811320755,
      "grad_norm": 3.379326105117798,
      "learning_rate": 4.605242377361441e-06,
      "loss": 0.3837,
      "num_input_tokens_seen": 12599136,
      "step": 19425
    },
    {
      "epoch": 11.456367924528301,
      "grad_norm": 3.4715375900268555,
      "learning_rate": 4.60267774646118e-06,
      "loss": 0.3779,
      "num_input_tokens_seen": 12602016,
      "step": 19430
    },
    {
      "epoch": 11.45931603773585,
      "grad_norm": 3.1865127086639404,
      "learning_rate": 4.600113220753698e-06,
      "loss": 0.4204,
      "num_input_tokens_seen": 12605088,
      "step": 19435
    },
    {
      "epoch": 11.462264150943396,
      "grad_norm": 2.809321403503418,
      "learning_rate": 4.597548800917964e-06,
      "loss": 0.422,
      "num_input_tokens_seen": 12608256,
      "step": 19440
    },
    {
      "epoch": 11.465212264150944,
      "grad_norm": 3.713116407394409,
      "learning_rate": 4.594984487632919e-06,
      "loss": 0.4594,
      "num_input_tokens_seen": 12611264,
      "step": 19445
    },
    {
      "epoch": 11.46816037735849,
      "grad_norm": 3.454892635345459,
      "learning_rate": 4.592420281577478e-06,
      "loss": 0.3147,
      "num_input_tokens_seen": 12615520,
      "step": 19450
    },
    {
      "epoch": 11.471108490566039,
      "grad_norm": 3.9381823539733887,
      "learning_rate": 4.589856183430521e-06,
      "loss": 0.3261,
      "num_input_tokens_seen": 12618656,
      "step": 19455
    },
    {
      "epoch": 11.474056603773585,
      "grad_norm": 1.8743760585784912,
      "learning_rate": 4.58729219387091e-06,
      "loss": 0.2887,
      "num_input_tokens_seen": 12621056,
      "step": 19460
    },
    {
      "epoch": 11.477004716981131,
      "grad_norm": 3.169360399246216,
      "learning_rate": 4.584728313577468e-06,
      "loss": 0.4086,
      "num_input_tokens_seen": 12623808,
      "step": 19465
    },
    {
      "epoch": 11.47995283018868,
      "grad_norm": 2.4051966667175293,
      "learning_rate": 4.582164543228993e-06,
      "loss": 0.4126,
      "num_input_tokens_seen": 12626624,
      "step": 19470
    },
    {
      "epoch": 11.482900943396226,
      "grad_norm": 4.014972686767578,
      "learning_rate": 4.5796008835042574e-06,
      "loss": 0.4094,
      "num_input_tokens_seen": 12630592,
      "step": 19475
    },
    {
      "epoch": 11.485849056603774,
      "grad_norm": 7.196697235107422,
      "learning_rate": 4.577037335082e-06,
      "loss": 0.3889,
      "num_input_tokens_seen": 12636736,
      "step": 19480
    },
    {
      "epoch": 11.48879716981132,
      "grad_norm": 3.4131386280059814,
      "learning_rate": 4.574473898640933e-06,
      "loss": 0.3556,
      "num_input_tokens_seen": 12640416,
      "step": 19485
    },
    {
      "epoch": 11.491745283018869,
      "grad_norm": 3.8058583736419678,
      "learning_rate": 4.571910574859732e-06,
      "loss": 0.3339,
      "num_input_tokens_seen": 12643008,
      "step": 19490
    },
    {
      "epoch": 11.494693396226415,
      "grad_norm": 1.593924880027771,
      "learning_rate": 4.5693473644170535e-06,
      "loss": 0.3247,
      "num_input_tokens_seen": 12646688,
      "step": 19495
    },
    {
      "epoch": 11.497641509433961,
      "grad_norm": 4.826627731323242,
      "learning_rate": 4.566784267991516e-06,
      "loss": 0.344,
      "num_input_tokens_seen": 12649504,
      "step": 19500
    },
    {
      "epoch": 11.50058962264151,
      "grad_norm": 2.482475757598877,
      "learning_rate": 4.564221286261709e-06,
      "loss": 0.3913,
      "num_input_tokens_seen": 12652608,
      "step": 19505
    },
    {
      "epoch": 11.503537735849056,
      "grad_norm": 2.0480363368988037,
      "learning_rate": 4.5616584199061964e-06,
      "loss": 0.4055,
      "num_input_tokens_seen": 12655904,
      "step": 19510
    },
    {
      "epoch": 11.506485849056604,
      "grad_norm": 2.887347459793091,
      "learning_rate": 4.559095669603506e-06,
      "loss": 0.3999,
      "num_input_tokens_seen": 12659264,
      "step": 19515
    },
    {
      "epoch": 11.50943396226415,
      "grad_norm": 2.948401689529419,
      "learning_rate": 4.556533036032136e-06,
      "loss": 0.3685,
      "num_input_tokens_seen": 12662272,
      "step": 19520
    },
    {
      "epoch": 11.512382075471699,
      "grad_norm": 3.1227221488952637,
      "learning_rate": 4.553970519870557e-06,
      "loss": 0.2932,
      "num_input_tokens_seen": 12665184,
      "step": 19525
    },
    {
      "epoch": 11.515330188679245,
      "grad_norm": 2.526907205581665,
      "learning_rate": 4.551408121797205e-06,
      "loss": 0.3674,
      "num_input_tokens_seen": 12668384,
      "step": 19530
    },
    {
      "epoch": 11.518278301886792,
      "grad_norm": 3.70326828956604,
      "learning_rate": 4.548845842490486e-06,
      "loss": 0.3781,
      "num_input_tokens_seen": 12671584,
      "step": 19535
    },
    {
      "epoch": 11.52122641509434,
      "grad_norm": 2.1793923377990723,
      "learning_rate": 4.5462836826287745e-06,
      "loss": 0.2747,
      "num_input_tokens_seen": 12674688,
      "step": 19540
    },
    {
      "epoch": 11.524174528301886,
      "grad_norm": 3.5786609649658203,
      "learning_rate": 4.543721642890414e-06,
      "loss": 0.3592,
      "num_input_tokens_seen": 12677184,
      "step": 19545
    },
    {
      "epoch": 11.527122641509434,
      "grad_norm": 4.557994365692139,
      "learning_rate": 4.541159723953714e-06,
      "loss": 0.3051,
      "num_input_tokens_seen": 12679424,
      "step": 19550
    },
    {
      "epoch": 11.53007075471698,
      "grad_norm": 6.400177478790283,
      "learning_rate": 4.538597926496955e-06,
      "loss": 0.4345,
      "num_input_tokens_seen": 12682080,
      "step": 19555
    },
    {
      "epoch": 11.533018867924529,
      "grad_norm": 3.2437756061553955,
      "learning_rate": 4.536036251198384e-06,
      "loss": 0.3478,
      "num_input_tokens_seen": 12685248,
      "step": 19560
    },
    {
      "epoch": 11.535966981132075,
      "grad_norm": 3.0821001529693604,
      "learning_rate": 4.5334746987362124e-06,
      "loss": 0.3259,
      "num_input_tokens_seen": 12687936,
      "step": 19565
    },
    {
      "epoch": 11.538915094339622,
      "grad_norm": 3.4812309741973877,
      "learning_rate": 4.530913269788627e-06,
      "loss": 0.3894,
      "num_input_tokens_seen": 12690912,
      "step": 19570
    },
    {
      "epoch": 11.54186320754717,
      "grad_norm": 3.672941207885742,
      "learning_rate": 4.528351965033775e-06,
      "loss": 0.3337,
      "num_input_tokens_seen": 12694016,
      "step": 19575
    },
    {
      "epoch": 11.544811320754716,
      "grad_norm": 2.5540120601654053,
      "learning_rate": 4.525790785149774e-06,
      "loss": 0.3162,
      "num_input_tokens_seen": 12697952,
      "step": 19580
    },
    {
      "epoch": 11.547759433962264,
      "grad_norm": 3.138040781021118,
      "learning_rate": 4.523229730814705e-06,
      "loss": 0.3754,
      "num_input_tokens_seen": 12700608,
      "step": 19585
    },
    {
      "epoch": 11.55070754716981,
      "grad_norm": 2.6405153274536133,
      "learning_rate": 4.52066880270662e-06,
      "loss": 0.3468,
      "num_input_tokens_seen": 12703456,
      "step": 19590
    },
    {
      "epoch": 11.553655660377359,
      "grad_norm": 2.0165600776672363,
      "learning_rate": 4.518108001503536e-06,
      "loss": 0.2514,
      "num_input_tokens_seen": 12705952,
      "step": 19595
    },
    {
      "epoch": 11.556603773584905,
      "grad_norm": 2.762314558029175,
      "learning_rate": 4.515547327883434e-06,
      "loss": 0.303,
      "num_input_tokens_seen": 12708416,
      "step": 19600
    },
    {
      "epoch": 11.559551886792454,
      "grad_norm": 2.6315758228302,
      "learning_rate": 4.512986782524266e-06,
      "loss": 0.3405,
      "num_input_tokens_seen": 12711424,
      "step": 19605
    },
    {
      "epoch": 11.5625,
      "grad_norm": 2.762866735458374,
      "learning_rate": 4.510426366103946e-06,
      "loss": 0.3445,
      "num_input_tokens_seen": 12713984,
      "step": 19610
    },
    {
      "epoch": 11.565448113207546,
      "grad_norm": 3.7692883014678955,
      "learning_rate": 4.5078660793003544e-06,
      "loss": 0.3271,
      "num_input_tokens_seen": 12716864,
      "step": 19615
    },
    {
      "epoch": 11.568396226415095,
      "grad_norm": 2.4588756561279297,
      "learning_rate": 4.505305922791341e-06,
      "loss": 0.3354,
      "num_input_tokens_seen": 12719936,
      "step": 19620
    },
    {
      "epoch": 11.571344339622641,
      "grad_norm": 3.9342236518859863,
      "learning_rate": 4.502745897254716e-06,
      "loss": 0.4032,
      "num_input_tokens_seen": 12724640,
      "step": 19625
    },
    {
      "epoch": 11.57429245283019,
      "grad_norm": 4.444206714630127,
      "learning_rate": 4.500186003368257e-06,
      "loss": 0.4728,
      "num_input_tokens_seen": 12731712,
      "step": 19630
    },
    {
      "epoch": 11.577240566037736,
      "grad_norm": 3.4900963306427,
      "learning_rate": 4.497626241809709e-06,
      "loss": 0.3451,
      "num_input_tokens_seen": 12736224,
      "step": 19635
    },
    {
      "epoch": 11.580188679245284,
      "grad_norm": 4.653920650482178,
      "learning_rate": 4.4950666132567775e-06,
      "loss": 0.3795,
      "num_input_tokens_seen": 12738656,
      "step": 19640
    },
    {
      "epoch": 11.58313679245283,
      "grad_norm": 3.8582394123077393,
      "learning_rate": 4.492507118387136e-06,
      "loss": 0.314,
      "num_input_tokens_seen": 12740864,
      "step": 19645
    },
    {
      "epoch": 11.586084905660378,
      "grad_norm": 2.1426351070404053,
      "learning_rate": 4.4899477578784235e-06,
      "loss": 0.2149,
      "num_input_tokens_seen": 12743296,
      "step": 19650
    },
    {
      "epoch": 11.589033018867925,
      "grad_norm": 4.424365520477295,
      "learning_rate": 4.487388532408239e-06,
      "loss": 0.318,
      "num_input_tokens_seen": 12746144,
      "step": 19655
    },
    {
      "epoch": 11.591981132075471,
      "grad_norm": 2.326800584793091,
      "learning_rate": 4.48482944265415e-06,
      "loss": 0.4556,
      "num_input_tokens_seen": 12749760,
      "step": 19660
    },
    {
      "epoch": 11.59492924528302,
      "grad_norm": 4.066593647003174,
      "learning_rate": 4.482270489293685e-06,
      "loss": 0.2548,
      "num_input_tokens_seen": 12752480,
      "step": 19665
    },
    {
      "epoch": 11.597877358490566,
      "grad_norm": 7.58344841003418,
      "learning_rate": 4.479711673004341e-06,
      "loss": 0.4895,
      "num_input_tokens_seen": 12755296,
      "step": 19670
    },
    {
      "epoch": 11.600825471698114,
      "grad_norm": 2.000216484069824,
      "learning_rate": 4.477152994463575e-06,
      "loss": 0.4689,
      "num_input_tokens_seen": 12758848,
      "step": 19675
    },
    {
      "epoch": 11.60377358490566,
      "grad_norm": 3.601074457168579,
      "learning_rate": 4.474594454348805e-06,
      "loss": 0.3624,
      "num_input_tokens_seen": 12762464,
      "step": 19680
    },
    {
      "epoch": 11.606721698113208,
      "grad_norm": 4.535464763641357,
      "learning_rate": 4.472036053337419e-06,
      "loss": 0.365,
      "num_input_tokens_seen": 12767040,
      "step": 19685
    },
    {
      "epoch": 11.609669811320755,
      "grad_norm": 2.7089684009552,
      "learning_rate": 4.469477792106761e-06,
      "loss": 0.2692,
      "num_input_tokens_seen": 12769664,
      "step": 19690
    },
    {
      "epoch": 11.612617924528301,
      "grad_norm": 2.500110626220703,
      "learning_rate": 4.466919671334146e-06,
      "loss": 0.2194,
      "num_input_tokens_seen": 12773344,
      "step": 19695
    },
    {
      "epoch": 11.61556603773585,
      "grad_norm": 3.0839643478393555,
      "learning_rate": 4.4643616916968456e-06,
      "loss": 0.4167,
      "num_input_tokens_seen": 12777152,
      "step": 19700
    },
    {
      "epoch": 11.618514150943396,
      "grad_norm": 2.420593023300171,
      "learning_rate": 4.461803853872095e-06,
      "loss": 0.2677,
      "num_input_tokens_seen": 12780384,
      "step": 19705
    },
    {
      "epoch": 11.621462264150944,
      "grad_norm": 2.43774676322937,
      "learning_rate": 4.459246158537094e-06,
      "loss": 0.2847,
      "num_input_tokens_seen": 12783296,
      "step": 19710
    },
    {
      "epoch": 11.62441037735849,
      "grad_norm": 3.2235865592956543,
      "learning_rate": 4.456688606369003e-06,
      "loss": 0.3663,
      "num_input_tokens_seen": 12786080,
      "step": 19715
    },
    {
      "epoch": 11.627358490566039,
      "grad_norm": 15.978479385375977,
      "learning_rate": 4.4541311980449446e-06,
      "loss": 0.2909,
      "num_input_tokens_seen": 12788960,
      "step": 19720
    },
    {
      "epoch": 11.630306603773585,
      "grad_norm": 3.642831802368164,
      "learning_rate": 4.4515739342420045e-06,
      "loss": 0.3462,
      "num_input_tokens_seen": 12791552,
      "step": 19725
    },
    {
      "epoch": 11.633254716981131,
      "grad_norm": 3.1080949306488037,
      "learning_rate": 4.4490168156372294e-06,
      "loss": 0.328,
      "num_input_tokens_seen": 12796096,
      "step": 19730
    },
    {
      "epoch": 11.63620283018868,
      "grad_norm": 3.398048162460327,
      "learning_rate": 4.446459842907626e-06,
      "loss": 0.3741,
      "num_input_tokens_seen": 12798624,
      "step": 19735
    },
    {
      "epoch": 11.639150943396226,
      "grad_norm": 1.7972863912582397,
      "learning_rate": 4.443903016730165e-06,
      "loss": 0.4921,
      "num_input_tokens_seen": 12802112,
      "step": 19740
    },
    {
      "epoch": 11.642099056603774,
      "grad_norm": 2.539865732192993,
      "learning_rate": 4.4413463377817775e-06,
      "loss": 0.4462,
      "num_input_tokens_seen": 12806176,
      "step": 19745
    },
    {
      "epoch": 11.64504716981132,
      "grad_norm": 2.45745587348938,
      "learning_rate": 4.438789806739353e-06,
      "loss": 0.3655,
      "num_input_tokens_seen": 12808832,
      "step": 19750
    },
    {
      "epoch": 11.647995283018869,
      "grad_norm": 2.5344152450561523,
      "learning_rate": 4.436233424279746e-06,
      "loss": 0.362,
      "num_input_tokens_seen": 12811456,
      "step": 19755
    },
    {
      "epoch": 11.650943396226415,
      "grad_norm": 2.2544867992401123,
      "learning_rate": 4.433677191079771e-06,
      "loss": 0.3379,
      "num_input_tokens_seen": 12814560,
      "step": 19760
    },
    {
      "epoch": 11.653891509433961,
      "grad_norm": 3.035067081451416,
      "learning_rate": 4.431121107816201e-06,
      "loss": 0.3362,
      "num_input_tokens_seen": 12817600,
      "step": 19765
    },
    {
      "epoch": 11.65683962264151,
      "grad_norm": 2.3624205589294434,
      "learning_rate": 4.4285651751657676e-06,
      "loss": 0.3646,
      "num_input_tokens_seen": 12820544,
      "step": 19770
    },
    {
      "epoch": 11.659787735849056,
      "grad_norm": 3.667090654373169,
      "learning_rate": 4.4260093938051685e-06,
      "loss": 0.3511,
      "num_input_tokens_seen": 12823680,
      "step": 19775
    },
    {
      "epoch": 11.662735849056604,
      "grad_norm": 2.381565809249878,
      "learning_rate": 4.423453764411056e-06,
      "loss": 0.3564,
      "num_input_tokens_seen": 12826496,
      "step": 19780
    },
    {
      "epoch": 11.66568396226415,
      "grad_norm": 3.608917713165283,
      "learning_rate": 4.4208982876600425e-06,
      "loss": 0.3395,
      "num_input_tokens_seen": 12829120,
      "step": 19785
    },
    {
      "epoch": 11.668632075471699,
      "grad_norm": 3.19148325920105,
      "learning_rate": 4.418342964228705e-06,
      "loss": 0.3948,
      "num_input_tokens_seen": 12831648,
      "step": 19790
    },
    {
      "epoch": 11.671580188679245,
      "grad_norm": 4.263964653015137,
      "learning_rate": 4.415787794793574e-06,
      "loss": 0.3757,
      "num_input_tokens_seen": 12834656,
      "step": 19795
    },
    {
      "epoch": 11.674528301886792,
      "grad_norm": 3.0447134971618652,
      "learning_rate": 4.4132327800311414e-06,
      "loss": 0.3713,
      "num_input_tokens_seen": 12837280,
      "step": 19800
    },
    {
      "epoch": 11.67747641509434,
      "grad_norm": 3.986912488937378,
      "learning_rate": 4.41067792061786e-06,
      "loss": 0.3225,
      "num_input_tokens_seen": 12840032,
      "step": 19805
    },
    {
      "epoch": 11.680424528301886,
      "grad_norm": 3.7158613204956055,
      "learning_rate": 4.408123217230139e-06,
      "loss": 0.3915,
      "num_input_tokens_seen": 12842432,
      "step": 19810
    },
    {
      "epoch": 11.683372641509434,
      "grad_norm": 2.474092721939087,
      "learning_rate": 4.4055686705443464e-06,
      "loss": 0.3345,
      "num_input_tokens_seen": 12846464,
      "step": 19815
    },
    {
      "epoch": 11.68632075471698,
      "grad_norm": 3.611208438873291,
      "learning_rate": 4.4030142812368106e-06,
      "loss": 0.3209,
      "num_input_tokens_seen": 12849344,
      "step": 19820
    },
    {
      "epoch": 11.689268867924529,
      "grad_norm": 3.4463675022125244,
      "learning_rate": 4.400460049983817e-06,
      "loss": 0.4117,
      "num_input_tokens_seen": 12852192,
      "step": 19825
    },
    {
      "epoch": 11.692216981132075,
      "grad_norm": 2.5775575637817383,
      "learning_rate": 4.397905977461608e-06,
      "loss": 0.3164,
      "num_input_tokens_seen": 12854848,
      "step": 19830
    },
    {
      "epoch": 11.695165094339622,
      "grad_norm": 2.412855863571167,
      "learning_rate": 4.395352064346387e-06,
      "loss": 0.4168,
      "num_input_tokens_seen": 12858432,
      "step": 19835
    },
    {
      "epoch": 11.69811320754717,
      "grad_norm": 2.529872179031372,
      "learning_rate": 4.392798311314314e-06,
      "loss": 0.4175,
      "num_input_tokens_seen": 12861248,
      "step": 19840
    },
    {
      "epoch": 11.701061320754716,
      "grad_norm": 3.454092264175415,
      "learning_rate": 4.390244719041502e-06,
      "loss": 0.3894,
      "num_input_tokens_seen": 12865152,
      "step": 19845
    },
    {
      "epoch": 11.704009433962264,
      "grad_norm": 4.411010265350342,
      "learning_rate": 4.387691288204029e-06,
      "loss": 0.3506,
      "num_input_tokens_seen": 12868064,
      "step": 19850
    },
    {
      "epoch": 11.70695754716981,
      "grad_norm": 2.877985954284668,
      "learning_rate": 4.385138019477928e-06,
      "loss": 0.3347,
      "num_input_tokens_seen": 12871424,
      "step": 19855
    },
    {
      "epoch": 11.709905660377359,
      "grad_norm": 2.8163540363311768,
      "learning_rate": 4.3825849135391864e-06,
      "loss": 0.2921,
      "num_input_tokens_seen": 12875648,
      "step": 19860
    },
    {
      "epoch": 11.712853773584905,
      "grad_norm": 3.4387645721435547,
      "learning_rate": 4.380031971063748e-06,
      "loss": 0.4168,
      "num_input_tokens_seen": 12879648,
      "step": 19865
    },
    {
      "epoch": 11.715801886792454,
      "grad_norm": 3.22654128074646,
      "learning_rate": 4.37747919272752e-06,
      "loss": 0.4508,
      "num_input_tokens_seen": 12882880,
      "step": 19870
    },
    {
      "epoch": 11.71875,
      "grad_norm": 2.170736789703369,
      "learning_rate": 4.374926579206357e-06,
      "loss": 0.381,
      "num_input_tokens_seen": 12890496,
      "step": 19875
    },
    {
      "epoch": 11.721698113207546,
      "grad_norm": 2.3108739852905273,
      "learning_rate": 4.372374131176075e-06,
      "loss": 0.34,
      "num_input_tokens_seen": 12894016,
      "step": 19880
    },
    {
      "epoch": 11.724646226415095,
      "grad_norm": 2.9883928298950195,
      "learning_rate": 4.369821849312449e-06,
      "loss": 0.4003,
      "num_input_tokens_seen": 12896992,
      "step": 19885
    },
    {
      "epoch": 11.727594339622641,
      "grad_norm": 2.798375129699707,
      "learning_rate": 4.367269734291203e-06,
      "loss": 0.333,
      "num_input_tokens_seen": 12900352,
      "step": 19890
    },
    {
      "epoch": 11.73054245283019,
      "grad_norm": 2.697693109512329,
      "learning_rate": 4.364717786788022e-06,
      "loss": 0.3922,
      "num_input_tokens_seen": 12903488,
      "step": 19895
    },
    {
      "epoch": 11.733490566037736,
      "grad_norm": 4.594666957855225,
      "learning_rate": 4.362166007478545e-06,
      "loss": 0.3817,
      "num_input_tokens_seen": 12905952,
      "step": 19900
    },
    {
      "epoch": 11.736438679245284,
      "grad_norm": 3.6201326847076416,
      "learning_rate": 4.3596143970383665e-06,
      "loss": 0.4337,
      "num_input_tokens_seen": 12909216,
      "step": 19905
    },
    {
      "epoch": 11.73938679245283,
      "grad_norm": 3.1675782203674316,
      "learning_rate": 4.357062956143035e-06,
      "loss": 0.3888,
      "num_input_tokens_seen": 12912416,
      "step": 19910
    },
    {
      "epoch": 11.742334905660378,
      "grad_norm": 2.8922080993652344,
      "learning_rate": 4.354511685468059e-06,
      "loss": 0.3884,
      "num_input_tokens_seen": 12915040,
      "step": 19915
    },
    {
      "epoch": 11.745283018867925,
      "grad_norm": 2.3347671031951904,
      "learning_rate": 4.351960585688894e-06,
      "loss": 0.3782,
      "num_input_tokens_seen": 12918432,
      "step": 19920
    },
    {
      "epoch": 11.748231132075471,
      "grad_norm": 3.013092517852783,
      "learning_rate": 4.349409657480959e-06,
      "loss": 0.3615,
      "num_input_tokens_seen": 12921760,
      "step": 19925
    },
    {
      "epoch": 11.75117924528302,
      "grad_norm": 4.734930992126465,
      "learning_rate": 4.34685890151962e-06,
      "loss": 0.5417,
      "num_input_tokens_seen": 12924896,
      "step": 19930
    },
    {
      "epoch": 11.754127358490566,
      "grad_norm": 5.73673152923584,
      "learning_rate": 4.344308318480201e-06,
      "loss": 0.2828,
      "num_input_tokens_seen": 12927296,
      "step": 19935
    },
    {
      "epoch": 11.757075471698114,
      "grad_norm": 2.4143080711364746,
      "learning_rate": 4.341757909037981e-06,
      "loss": 0.339,
      "num_input_tokens_seen": 12930752,
      "step": 19940
    },
    {
      "epoch": 11.76002358490566,
      "grad_norm": 7.754148483276367,
      "learning_rate": 4.33920767386819e-06,
      "loss": 0.453,
      "num_input_tokens_seen": 12933632,
      "step": 19945
    },
    {
      "epoch": 11.762971698113208,
      "grad_norm": 3.7630932331085205,
      "learning_rate": 4.336657613646017e-06,
      "loss": 0.3208,
      "num_input_tokens_seen": 12936928,
      "step": 19950
    },
    {
      "epoch": 11.765919811320755,
      "grad_norm": 3.3198673725128174,
      "learning_rate": 4.3341077290466e-06,
      "loss": 0.3409,
      "num_input_tokens_seen": 12939680,
      "step": 19955
    },
    {
      "epoch": 11.768867924528301,
      "grad_norm": 2.4636270999908447,
      "learning_rate": 4.331558020745031e-06,
      "loss": 0.3026,
      "num_input_tokens_seen": 12942368,
      "step": 19960
    },
    {
      "epoch": 11.77181603773585,
      "grad_norm": 2.7716333866119385,
      "learning_rate": 4.329008489416357e-06,
      "loss": 0.2678,
      "num_input_tokens_seen": 12945120,
      "step": 19965
    },
    {
      "epoch": 11.774764150943396,
      "grad_norm": 2.849222421646118,
      "learning_rate": 4.326459135735576e-06,
      "loss": 0.331,
      "num_input_tokens_seen": 12947840,
      "step": 19970
    },
    {
      "epoch": 11.777712264150944,
      "grad_norm": 3.6744046211242676,
      "learning_rate": 4.323909960377644e-06,
      "loss": 0.5297,
      "num_input_tokens_seen": 12950624,
      "step": 19975
    },
    {
      "epoch": 11.78066037735849,
      "grad_norm": 3.404038429260254,
      "learning_rate": 4.3213609640174625e-06,
      "loss": 0.7007,
      "num_input_tokens_seen": 12953696,
      "step": 19980
    },
    {
      "epoch": 11.783608490566039,
      "grad_norm": 2.732004165649414,
      "learning_rate": 4.318812147329889e-06,
      "loss": 0.29,
      "num_input_tokens_seen": 12956576,
      "step": 19985
    },
    {
      "epoch": 11.786556603773585,
      "grad_norm": 2.627290725708008,
      "learning_rate": 4.316263510989737e-06,
      "loss": 0.3123,
      "num_input_tokens_seen": 12960192,
      "step": 19990
    },
    {
      "epoch": 11.789504716981131,
      "grad_norm": 2.3037819862365723,
      "learning_rate": 4.313715055671768e-06,
      "loss": 0.2683,
      "num_input_tokens_seen": 12963200,
      "step": 19995
    },
    {
      "epoch": 11.79245283018868,
      "grad_norm": 2.7089223861694336,
      "learning_rate": 4.311166782050694e-06,
      "loss": 0.3618,
      "num_input_tokens_seen": 12966144,
      "step": 20000
    },
    {
      "epoch": 11.795400943396226,
      "grad_norm": 3.2967770099639893,
      "learning_rate": 4.308618690801184e-06,
      "loss": 0.371,
      "num_input_tokens_seen": 12970560,
      "step": 20005
    },
    {
      "epoch": 11.798349056603774,
      "grad_norm": 3.454355239868164,
      "learning_rate": 4.3060707825978564e-06,
      "loss": 0.3449,
      "num_input_tokens_seen": 12973568,
      "step": 20010
    },
    {
      "epoch": 11.80129716981132,
      "grad_norm": 2.5843403339385986,
      "learning_rate": 4.303523058115278e-06,
      "loss": 0.4561,
      "num_input_tokens_seen": 12976224,
      "step": 20015
    },
    {
      "epoch": 11.804245283018869,
      "grad_norm": 2.6961617469787598,
      "learning_rate": 4.300975518027972e-06,
      "loss": 0.3919,
      "num_input_tokens_seen": 12979744,
      "step": 20020
    },
    {
      "epoch": 11.807193396226415,
      "grad_norm": 2.589271068572998,
      "learning_rate": 4.298428163010411e-06,
      "loss": 0.343,
      "num_input_tokens_seen": 12982496,
      "step": 20025
    },
    {
      "epoch": 11.810141509433961,
      "grad_norm": 3.6839356422424316,
      "learning_rate": 4.295880993737016e-06,
      "loss": 0.3412,
      "num_input_tokens_seen": 12985088,
      "step": 20030
    },
    {
      "epoch": 11.81308962264151,
      "grad_norm": 2.8832032680511475,
      "learning_rate": 4.293334010882164e-06,
      "loss": 0.2953,
      "num_input_tokens_seen": 12987296,
      "step": 20035
    },
    {
      "epoch": 11.816037735849056,
      "grad_norm": 4.010433673858643,
      "learning_rate": 4.290787215120178e-06,
      "loss": 0.4728,
      "num_input_tokens_seen": 12990240,
      "step": 20040
    },
    {
      "epoch": 11.818985849056604,
      "grad_norm": 2.4735987186431885,
      "learning_rate": 4.2882406071253315e-06,
      "loss": 0.4037,
      "num_input_tokens_seen": 12994464,
      "step": 20045
    },
    {
      "epoch": 11.82193396226415,
      "grad_norm": 2.5629630088806152,
      "learning_rate": 4.285694187571852e-06,
      "loss": 0.3776,
      "num_input_tokens_seen": 12997824,
      "step": 20050
    },
    {
      "epoch": 11.824882075471699,
      "grad_norm": 4.173508644104004,
      "learning_rate": 4.283147957133917e-06,
      "loss": 0.4327,
      "num_input_tokens_seen": 13000608,
      "step": 20055
    },
    {
      "epoch": 11.827830188679245,
      "grad_norm": 2.714226245880127,
      "learning_rate": 4.280601916485648e-06,
      "loss": 0.3892,
      "num_input_tokens_seen": 13003872,
      "step": 20060
    },
    {
      "epoch": 11.830778301886792,
      "grad_norm": 2.140943765640259,
      "learning_rate": 4.278056066301123e-06,
      "loss": 0.461,
      "num_input_tokens_seen": 13006496,
      "step": 20065
    },
    {
      "epoch": 11.83372641509434,
      "grad_norm": 3.936939001083374,
      "learning_rate": 4.275510407254366e-06,
      "loss": 0.2729,
      "num_input_tokens_seen": 13008864,
      "step": 20070
    },
    {
      "epoch": 11.836674528301886,
      "grad_norm": 5.70048189163208,
      "learning_rate": 4.27296494001935e-06,
      "loss": 0.2924,
      "num_input_tokens_seen": 13012032,
      "step": 20075
    },
    {
      "epoch": 11.839622641509434,
      "grad_norm": 0.5379644632339478,
      "learning_rate": 4.2704196652700005e-06,
      "loss": 0.4296,
      "num_input_tokens_seen": 13018496,
      "step": 20080
    },
    {
      "epoch": 11.84257075471698,
      "grad_norm": 2.2116522789001465,
      "learning_rate": 4.2678745836801895e-06,
      "loss": 0.2407,
      "num_input_tokens_seen": 13021120,
      "step": 20085
    },
    {
      "epoch": 11.845518867924529,
      "grad_norm": 3.5138449668884277,
      "learning_rate": 4.265329695923736e-06,
      "loss": 0.4002,
      "num_input_tokens_seen": 13024704,
      "step": 20090
    },
    {
      "epoch": 11.848466981132075,
      "grad_norm": 3.3818068504333496,
      "learning_rate": 4.262785002674412e-06,
      "loss": 0.4181,
      "num_input_tokens_seen": 13028128,
      "step": 20095
    },
    {
      "epoch": 11.851415094339622,
      "grad_norm": 4.649567604064941,
      "learning_rate": 4.260240504605938e-06,
      "loss": 0.3856,
      "num_input_tokens_seen": 13030944,
      "step": 20100
    },
    {
      "epoch": 11.85436320754717,
      "grad_norm": 2.405526876449585,
      "learning_rate": 4.2576962023919775e-06,
      "loss": 0.315,
      "num_input_tokens_seen": 13035136,
      "step": 20105
    },
    {
      "epoch": 11.857311320754716,
      "grad_norm": 2.394003391265869,
      "learning_rate": 4.255152096706145e-06,
      "loss": 0.292,
      "num_input_tokens_seen": 13038656,
      "step": 20110
    },
    {
      "epoch": 11.860259433962264,
      "grad_norm": 3.161060094833374,
      "learning_rate": 4.252608188222007e-06,
      "loss": 0.3514,
      "num_input_tokens_seen": 13041792,
      "step": 20115
    },
    {
      "epoch": 11.86320754716981,
      "grad_norm": 3.8759140968322754,
      "learning_rate": 4.250064477613071e-06,
      "loss": 0.3615,
      "num_input_tokens_seen": 13045216,
      "step": 20120
    },
    {
      "epoch": 11.866155660377359,
      "grad_norm": 2.6012954711914062,
      "learning_rate": 4.247520965552797e-06,
      "loss": 0.4069,
      "num_input_tokens_seen": 13048064,
      "step": 20125
    },
    {
      "epoch": 11.869103773584905,
      "grad_norm": 7.535912036895752,
      "learning_rate": 4.24497765271459e-06,
      "loss": 0.4118,
      "num_input_tokens_seen": 13050912,
      "step": 20130
    },
    {
      "epoch": 11.872051886792454,
      "grad_norm": 4.390542984008789,
      "learning_rate": 4.242434539771804e-06,
      "loss": 0.4185,
      "num_input_tokens_seen": 13054400,
      "step": 20135
    },
    {
      "epoch": 11.875,
      "grad_norm": 3.679882764816284,
      "learning_rate": 4.239891627397737e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 13057408,
      "step": 20140
    },
    {
      "epoch": 11.877948113207546,
      "grad_norm": 2.9100241661071777,
      "learning_rate": 4.237348916265638e-06,
      "loss": 0.3743,
      "num_input_tokens_seen": 13060512,
      "step": 20145
    },
    {
      "epoch": 11.880896226415095,
      "grad_norm": 3.157566547393799,
      "learning_rate": 4.234806407048702e-06,
      "loss": 0.3186,
      "num_input_tokens_seen": 13063296,
      "step": 20150
    },
    {
      "epoch": 11.883844339622641,
      "grad_norm": 4.066344738006592,
      "learning_rate": 4.232264100420066e-06,
      "loss": 0.3336,
      "num_input_tokens_seen": 13066368,
      "step": 20155
    },
    {
      "epoch": 11.88679245283019,
      "grad_norm": 2.238661050796509,
      "learning_rate": 4.229721997052819e-06,
      "loss": 0.3539,
      "num_input_tokens_seen": 13068928,
      "step": 20160
    },
    {
      "epoch": 11.889740566037736,
      "grad_norm": 1.8770211935043335,
      "learning_rate": 4.227180097619993e-06,
      "loss": 0.3318,
      "num_input_tokens_seen": 13072352,
      "step": 20165
    },
    {
      "epoch": 11.892688679245284,
      "grad_norm": 3.768514633178711,
      "learning_rate": 4.224638402794566e-06,
      "loss": 0.3825,
      "num_input_tokens_seen": 13076032,
      "step": 20170
    },
    {
      "epoch": 11.89563679245283,
      "grad_norm": 2.716832160949707,
      "learning_rate": 4.222096913249464e-06,
      "loss": 0.4205,
      "num_input_tokens_seen": 13079520,
      "step": 20175
    },
    {
      "epoch": 11.898584905660378,
      "grad_norm": 4.5003981590271,
      "learning_rate": 4.219555629657558e-06,
      "loss": 0.4614,
      "num_input_tokens_seen": 13082656,
      "step": 20180
    },
    {
      "epoch": 11.901533018867925,
      "grad_norm": 2.6987879276275635,
      "learning_rate": 4.21701455269166e-06,
      "loss": 0.3034,
      "num_input_tokens_seen": 13085344,
      "step": 20185
    },
    {
      "epoch": 11.904481132075471,
      "grad_norm": 3.9649617671966553,
      "learning_rate": 4.2144736830245356e-06,
      "loss": 0.2722,
      "num_input_tokens_seen": 13088416,
      "step": 20190
    },
    {
      "epoch": 11.90742924528302,
      "grad_norm": 2.8481290340423584,
      "learning_rate": 4.211933021328889e-06,
      "loss": 0.5285,
      "num_input_tokens_seen": 13091648,
      "step": 20195
    },
    {
      "epoch": 11.910377358490566,
      "grad_norm": 3.9942662715911865,
      "learning_rate": 4.20939256827737e-06,
      "loss": 0.253,
      "num_input_tokens_seen": 13094624,
      "step": 20200
    },
    {
      "epoch": 11.913325471698114,
      "grad_norm": 3.428068161010742,
      "learning_rate": 4.206852324542578e-06,
      "loss": 0.3156,
      "num_input_tokens_seen": 13098304,
      "step": 20205
    },
    {
      "epoch": 11.91627358490566,
      "grad_norm": 2.8170766830444336,
      "learning_rate": 4.2043122907970496e-06,
      "loss": 0.3161,
      "num_input_tokens_seen": 13101344,
      "step": 20210
    },
    {
      "epoch": 11.919221698113208,
      "grad_norm": 2.1978559494018555,
      "learning_rate": 4.2017724677132715e-06,
      "loss": 0.3769,
      "num_input_tokens_seen": 13105376,
      "step": 20215
    },
    {
      "epoch": 11.922169811320755,
      "grad_norm": 2.272714138031006,
      "learning_rate": 4.1992328559636734e-06,
      "loss": 0.5112,
      "num_input_tokens_seen": 13108416,
      "step": 20220
    },
    {
      "epoch": 11.925117924528301,
      "grad_norm": 1.999866008758545,
      "learning_rate": 4.196693456220628e-06,
      "loss": 0.3133,
      "num_input_tokens_seen": 13110912,
      "step": 20225
    },
    {
      "epoch": 11.92806603773585,
      "grad_norm": 5.271966934204102,
      "learning_rate": 4.194154269156452e-06,
      "loss": 0.3209,
      "num_input_tokens_seen": 13114816,
      "step": 20230
    },
    {
      "epoch": 11.931014150943396,
      "grad_norm": 2.9798502922058105,
      "learning_rate": 4.191615295443404e-06,
      "loss": 0.3607,
      "num_input_tokens_seen": 13118272,
      "step": 20235
    },
    {
      "epoch": 11.933962264150944,
      "grad_norm": 3.328861713409424,
      "learning_rate": 4.189076535753692e-06,
      "loss": 0.4827,
      "num_input_tokens_seen": 13120992,
      "step": 20240
    },
    {
      "epoch": 11.93691037735849,
      "grad_norm": 2.587646007537842,
      "learning_rate": 4.186537990759464e-06,
      "loss": 0.2651,
      "num_input_tokens_seen": 13124288,
      "step": 20245
    },
    {
      "epoch": 11.939858490566039,
      "grad_norm": 4.353909492492676,
      "learning_rate": 4.183999661132806e-06,
      "loss": 0.3143,
      "num_input_tokens_seen": 13127232,
      "step": 20250
    },
    {
      "epoch": 11.942806603773585,
      "grad_norm": 3.590428352355957,
      "learning_rate": 4.181461547545756e-06,
      "loss": 0.4378,
      "num_input_tokens_seen": 13132224,
      "step": 20255
    },
    {
      "epoch": 11.945754716981131,
      "grad_norm": 3.940999984741211,
      "learning_rate": 4.178923650670289e-06,
      "loss": 0.4276,
      "num_input_tokens_seen": 13135104,
      "step": 20260
    },
    {
      "epoch": 11.94870283018868,
      "grad_norm": 3.0216257572174072,
      "learning_rate": 4.176385971178324e-06,
      "loss": 0.3666,
      "num_input_tokens_seen": 13138816,
      "step": 20265
    },
    {
      "epoch": 11.951650943396226,
      "grad_norm": 3.2505743503570557,
      "learning_rate": 4.1738485097417225e-06,
      "loss": 0.4101,
      "num_input_tokens_seen": 13142688,
      "step": 20270
    },
    {
      "epoch": 11.954599056603774,
      "grad_norm": 2.712519407272339,
      "learning_rate": 4.1713112670322886e-06,
      "loss": 0.3102,
      "num_input_tokens_seen": 13145792,
      "step": 20275
    },
    {
      "epoch": 11.95754716981132,
      "grad_norm": 3.935619592666626,
      "learning_rate": 4.168774243721768e-06,
      "loss": 0.3093,
      "num_input_tokens_seen": 13148576,
      "step": 20280
    },
    {
      "epoch": 11.960495283018869,
      "grad_norm": 3.8606555461883545,
      "learning_rate": 4.166237440481849e-06,
      "loss": 0.4583,
      "num_input_tokens_seen": 13152768,
      "step": 20285
    },
    {
      "epoch": 11.963443396226415,
      "grad_norm": 2.3054466247558594,
      "learning_rate": 4.163700857984162e-06,
      "loss": 0.3823,
      "num_input_tokens_seen": 13156352,
      "step": 20290
    },
    {
      "epoch": 11.966391509433961,
      "grad_norm": 4.4597578048706055,
      "learning_rate": 4.161164496900275e-06,
      "loss": 0.3497,
      "num_input_tokens_seen": 13158912,
      "step": 20295
    },
    {
      "epoch": 11.96933962264151,
      "grad_norm": 4.474064826965332,
      "learning_rate": 4.1586283579017036e-06,
      "loss": 0.2521,
      "num_input_tokens_seen": 13162400,
      "step": 20300
    },
    {
      "epoch": 11.972287735849056,
      "grad_norm": 3.1879305839538574,
      "learning_rate": 4.156092441659901e-06,
      "loss": 0.3688,
      "num_input_tokens_seen": 13165856,
      "step": 20305
    },
    {
      "epoch": 11.975235849056604,
      "grad_norm": 3.43325138092041,
      "learning_rate": 4.153556748846261e-06,
      "loss": 0.445,
      "num_input_tokens_seen": 13168704,
      "step": 20310
    },
    {
      "epoch": 11.97818396226415,
      "grad_norm": 4.662768840789795,
      "learning_rate": 4.15102128013212e-06,
      "loss": 0.3936,
      "num_input_tokens_seen": 13172160,
      "step": 20315
    },
    {
      "epoch": 11.981132075471699,
      "grad_norm": 3.0826010704040527,
      "learning_rate": 4.1484860361887544e-06,
      "loss": 0.3486,
      "num_input_tokens_seen": 13175136,
      "step": 20320
    },
    {
      "epoch": 11.984080188679245,
      "grad_norm": 3.370262622833252,
      "learning_rate": 4.145951017687379e-06,
      "loss": 0.3205,
      "num_input_tokens_seen": 13178592,
      "step": 20325
    },
    {
      "epoch": 11.987028301886792,
      "grad_norm": 8.090503692626953,
      "learning_rate": 4.1434162252991524e-06,
      "loss": 0.4142,
      "num_input_tokens_seen": 13181408,
      "step": 20330
    },
    {
      "epoch": 11.98997641509434,
      "grad_norm": 2.2202465534210205,
      "learning_rate": 4.140881659695173e-06,
      "loss": 0.3023,
      "num_input_tokens_seen": 13183776,
      "step": 20335
    },
    {
      "epoch": 11.992924528301886,
      "grad_norm": 2.4106674194335938,
      "learning_rate": 4.138347321546477e-06,
      "loss": 0.2751,
      "num_input_tokens_seen": 13186496,
      "step": 20340
    },
    {
      "epoch": 11.995872641509434,
      "grad_norm": 4.315904140472412,
      "learning_rate": 4.13581321152404e-06,
      "loss": 0.3534,
      "num_input_tokens_seen": 13189504,
      "step": 20345
    },
    {
      "epoch": 11.99882075471698,
      "grad_norm": 6.822838306427002,
      "learning_rate": 4.133279330298781e-06,
      "loss": 0.304,
      "num_input_tokens_seen": 13192096,
      "step": 20350
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.5534425377845764,
      "eval_runtime": 19.4781,
      "eval_samples_per_second": 87.072,
      "eval_steps_per_second": 21.768,
      "num_input_tokens_seen": 13192800,
      "step": 20352
    },
    {
      "epoch": 12.001768867924529,
      "grad_norm": 6.925686359405518,
      "learning_rate": 4.130745678541555e-06,
      "loss": 0.3839,
      "num_input_tokens_seen": 13194240,
      "step": 20355
    },
    {
      "epoch": 12.004716981132075,
      "grad_norm": 1.717543601989746,
      "learning_rate": 4.128212256923155e-06,
      "loss": 0.3144,
      "num_input_tokens_seen": 13197216,
      "step": 20360
    },
    {
      "epoch": 12.007665094339623,
      "grad_norm": 1.753656268119812,
      "learning_rate": 4.125679066114318e-06,
      "loss": 0.3256,
      "num_input_tokens_seen": 13200640,
      "step": 20365
    },
    {
      "epoch": 12.01061320754717,
      "grad_norm": 4.245390892028809,
      "learning_rate": 4.123146106785717e-06,
      "loss": 0.3403,
      "num_input_tokens_seen": 13203680,
      "step": 20370
    },
    {
      "epoch": 12.013561320754716,
      "grad_norm": 3.201458692550659,
      "learning_rate": 4.1206133796079625e-06,
      "loss": 0.3505,
      "num_input_tokens_seen": 13206976,
      "step": 20375
    },
    {
      "epoch": 12.016509433962264,
      "grad_norm": 2.771188259124756,
      "learning_rate": 4.1180808852516065e-06,
      "loss": 0.3043,
      "num_input_tokens_seen": 13210592,
      "step": 20380
    },
    {
      "epoch": 12.01945754716981,
      "grad_norm": 3.3748929500579834,
      "learning_rate": 4.115548624387136e-06,
      "loss": 0.3385,
      "num_input_tokens_seen": 13213056,
      "step": 20385
    },
    {
      "epoch": 12.022405660377359,
      "grad_norm": 3.949770212173462,
      "learning_rate": 4.113016597684979e-06,
      "loss": 0.43,
      "num_input_tokens_seen": 13216000,
      "step": 20390
    },
    {
      "epoch": 12.025353773584905,
      "grad_norm": 3.8029401302337646,
      "learning_rate": 4.110484805815502e-06,
      "loss": 0.2462,
      "num_input_tokens_seen": 13218912,
      "step": 20395
    },
    {
      "epoch": 12.028301886792454,
      "grad_norm": 5.414029121398926,
      "learning_rate": 4.107953249449005e-06,
      "loss": 0.3128,
      "num_input_tokens_seen": 13221056,
      "step": 20400
    },
    {
      "epoch": 12.03125,
      "grad_norm": 3.0894198417663574,
      "learning_rate": 4.105421929255729e-06,
      "loss": 0.3332,
      "num_input_tokens_seen": 13223616,
      "step": 20405
    },
    {
      "epoch": 12.034198113207546,
      "grad_norm": 3.4658336639404297,
      "learning_rate": 4.102890845905854e-06,
      "loss": 0.3532,
      "num_input_tokens_seen": 13226528,
      "step": 20410
    },
    {
      "epoch": 12.037146226415095,
      "grad_norm": 4.656062126159668,
      "learning_rate": 4.1003600000694935e-06,
      "loss": 0.417,
      "num_input_tokens_seen": 13229664,
      "step": 20415
    },
    {
      "epoch": 12.040094339622641,
      "grad_norm": 1.788862943649292,
      "learning_rate": 4.0978293924167e-06,
      "loss": 0.2785,
      "num_input_tokens_seen": 13233728,
      "step": 20420
    },
    {
      "epoch": 12.04304245283019,
      "grad_norm": 3.640392303466797,
      "learning_rate": 4.095299023617461e-06,
      "loss": 0.3302,
      "num_input_tokens_seen": 13236000,
      "step": 20425
    },
    {
      "epoch": 12.045990566037736,
      "grad_norm": 2.3176677227020264,
      "learning_rate": 4.092768894341707e-06,
      "loss": 0.2207,
      "num_input_tokens_seen": 13238688,
      "step": 20430
    },
    {
      "epoch": 12.048938679245284,
      "grad_norm": 2.863654613494873,
      "learning_rate": 4.090239005259298e-06,
      "loss": 0.3736,
      "num_input_tokens_seen": 13241536,
      "step": 20435
    },
    {
      "epoch": 12.05188679245283,
      "grad_norm": 1.6612244844436646,
      "learning_rate": 4.087709357040033e-06,
      "loss": 0.3053,
      "num_input_tokens_seen": 13244896,
      "step": 20440
    },
    {
      "epoch": 12.054834905660377,
      "grad_norm": 2.1121652126312256,
      "learning_rate": 4.085179950353648e-06,
      "loss": 0.28,
      "num_input_tokens_seen": 13248416,
      "step": 20445
    },
    {
      "epoch": 12.057783018867925,
      "grad_norm": 2.879765748977661,
      "learning_rate": 4.0826507858698135e-06,
      "loss": 0.3316,
      "num_input_tokens_seen": 13250912,
      "step": 20450
    },
    {
      "epoch": 12.060731132075471,
      "grad_norm": 6.691248416900635,
      "learning_rate": 4.080121864258136e-06,
      "loss": 0.4021,
      "num_input_tokens_seen": 13253536,
      "step": 20455
    },
    {
      "epoch": 12.06367924528302,
      "grad_norm": 4.2812042236328125,
      "learning_rate": 4.077593186188161e-06,
      "loss": 0.377,
      "num_input_tokens_seen": 13257152,
      "step": 20460
    },
    {
      "epoch": 12.066627358490566,
      "grad_norm": 2.6416330337524414,
      "learning_rate": 4.075064752329364e-06,
      "loss": 0.3573,
      "num_input_tokens_seen": 13259872,
      "step": 20465
    },
    {
      "epoch": 12.069575471698114,
      "grad_norm": 4.602396488189697,
      "learning_rate": 4.0725365633511605e-06,
      "loss": 0.3466,
      "num_input_tokens_seen": 13263200,
      "step": 20470
    },
    {
      "epoch": 12.07252358490566,
      "grad_norm": 3.0385568141937256,
      "learning_rate": 4.070008619922899e-06,
      "loss": 0.3042,
      "num_input_tokens_seen": 13266720,
      "step": 20475
    },
    {
      "epoch": 12.075471698113208,
      "grad_norm": 3.9934496879577637,
      "learning_rate": 4.067480922713864e-06,
      "loss": 0.3972,
      "num_input_tokens_seen": 13269760,
      "step": 20480
    },
    {
      "epoch": 12.078419811320755,
      "grad_norm": 3.3791260719299316,
      "learning_rate": 4.064953472393273e-06,
      "loss": 0.4476,
      "num_input_tokens_seen": 13273472,
      "step": 20485
    },
    {
      "epoch": 12.081367924528301,
      "grad_norm": 2.6251001358032227,
      "learning_rate": 4.06242626963028e-06,
      "loss": 0.3603,
      "num_input_tokens_seen": 13277024,
      "step": 20490
    },
    {
      "epoch": 12.08431603773585,
      "grad_norm": 3.4249393939971924,
      "learning_rate": 4.059899315093972e-06,
      "loss": 0.2899,
      "num_input_tokens_seen": 13281216,
      "step": 20495
    },
    {
      "epoch": 12.087264150943396,
      "grad_norm": 4.444034099578857,
      "learning_rate": 4.057372609453374e-06,
      "loss": 0.4394,
      "num_input_tokens_seen": 13284416,
      "step": 20500
    },
    {
      "epoch": 12.090212264150944,
      "grad_norm": 3.6405632495880127,
      "learning_rate": 4.054846153377439e-06,
      "loss": 0.3176,
      "num_input_tokens_seen": 13288064,
      "step": 20505
    },
    {
      "epoch": 12.09316037735849,
      "grad_norm": 3.7984604835510254,
      "learning_rate": 4.052319947535058e-06,
      "loss": 0.3931,
      "num_input_tokens_seen": 13291168,
      "step": 20510
    },
    {
      "epoch": 12.096108490566039,
      "grad_norm": 4.967504978179932,
      "learning_rate": 4.049793992595056e-06,
      "loss": 0.3818,
      "num_input_tokens_seen": 13294848,
      "step": 20515
    },
    {
      "epoch": 12.099056603773585,
      "grad_norm": 2.7917587757110596,
      "learning_rate": 4.047268289226187e-06,
      "loss": 0.4332,
      "num_input_tokens_seen": 13298176,
      "step": 20520
    },
    {
      "epoch": 12.102004716981131,
      "grad_norm": 3.6184866428375244,
      "learning_rate": 4.044742838097147e-06,
      "loss": 0.3067,
      "num_input_tokens_seen": 13301504,
      "step": 20525
    },
    {
      "epoch": 12.10495283018868,
      "grad_norm": 2.6465489864349365,
      "learning_rate": 4.0422176398765564e-06,
      "loss": 0.434,
      "num_input_tokens_seen": 13305088,
      "step": 20530
    },
    {
      "epoch": 12.107900943396226,
      "grad_norm": 3.4872076511383057,
      "learning_rate": 4.039692695232975e-06,
      "loss": 0.4516,
      "num_input_tokens_seen": 13308160,
      "step": 20535
    },
    {
      "epoch": 12.110849056603774,
      "grad_norm": 3.1022300720214844,
      "learning_rate": 4.03716800483489e-06,
      "loss": 0.2465,
      "num_input_tokens_seen": 13310848,
      "step": 20540
    },
    {
      "epoch": 12.11379716981132,
      "grad_norm": 10.89215087890625,
      "learning_rate": 4.034643569350726e-06,
      "loss": 0.3554,
      "num_input_tokens_seen": 13314112,
      "step": 20545
    },
    {
      "epoch": 12.116745283018869,
      "grad_norm": 2.191279888153076,
      "learning_rate": 4.032119389448837e-06,
      "loss": 0.3267,
      "num_input_tokens_seen": 13316896,
      "step": 20550
    },
    {
      "epoch": 12.119693396226415,
      "grad_norm": 3.2623863220214844,
      "learning_rate": 4.0295954657975115e-06,
      "loss": 0.352,
      "num_input_tokens_seen": 13319936,
      "step": 20555
    },
    {
      "epoch": 12.122641509433961,
      "grad_norm": 3.4204037189483643,
      "learning_rate": 4.027071799064968e-06,
      "loss": 0.274,
      "num_input_tokens_seen": 13323360,
      "step": 20560
    },
    {
      "epoch": 12.12558962264151,
      "grad_norm": 3.4913723468780518,
      "learning_rate": 4.02454838991936e-06,
      "loss": 0.3231,
      "num_input_tokens_seen": 13327392,
      "step": 20565
    },
    {
      "epoch": 12.128537735849056,
      "grad_norm": 3.1582820415496826,
      "learning_rate": 4.022025239028768e-06,
      "loss": 0.3848,
      "num_input_tokens_seen": 13330784,
      "step": 20570
    },
    {
      "epoch": 12.131485849056604,
      "grad_norm": 6.1729583740234375,
      "learning_rate": 4.0195023470612095e-06,
      "loss": 0.3152,
      "num_input_tokens_seen": 13333216,
      "step": 20575
    },
    {
      "epoch": 12.13443396226415,
      "grad_norm": 3.7841570377349854,
      "learning_rate": 4.016979714684631e-06,
      "loss": 0.4127,
      "num_input_tokens_seen": 13336768,
      "step": 20580
    },
    {
      "epoch": 12.137382075471699,
      "grad_norm": 3.932253360748291,
      "learning_rate": 4.014457342566909e-06,
      "loss": 0.4393,
      "num_input_tokens_seen": 13340832,
      "step": 20585
    },
    {
      "epoch": 12.140330188679245,
      "grad_norm": 9.46184253692627,
      "learning_rate": 4.011935231375853e-06,
      "loss": 0.5079,
      "num_input_tokens_seen": 13344384,
      "step": 20590
    },
    {
      "epoch": 12.143278301886792,
      "grad_norm": 3.0003738403320312,
      "learning_rate": 4.009413381779203e-06,
      "loss": 0.2962,
      "num_input_tokens_seen": 13347776,
      "step": 20595
    },
    {
      "epoch": 12.14622641509434,
      "grad_norm": 2.263674020767212,
      "learning_rate": 4.00689179444463e-06,
      "loss": 0.4231,
      "num_input_tokens_seen": 13350048,
      "step": 20600
    },
    {
      "epoch": 12.149174528301886,
      "grad_norm": 8.765588760375977,
      "learning_rate": 4.004370470039733e-06,
      "loss": 0.4853,
      "num_input_tokens_seen": 13353472,
      "step": 20605
    },
    {
      "epoch": 12.152122641509434,
      "grad_norm": 2.86527156829834,
      "learning_rate": 4.001849409232046e-06,
      "loss": 0.3995,
      "num_input_tokens_seen": 13356544,
      "step": 20610
    },
    {
      "epoch": 12.15507075471698,
      "grad_norm": 3.8956801891326904,
      "learning_rate": 3.9993286126890274e-06,
      "loss": 0.4197,
      "num_input_tokens_seen": 13360576,
      "step": 20615
    },
    {
      "epoch": 12.158018867924529,
      "grad_norm": 3.221691846847534,
      "learning_rate": 3.996808081078074e-06,
      "loss": 0.3659,
      "num_input_tokens_seen": 13363520,
      "step": 20620
    },
    {
      "epoch": 12.160966981132075,
      "grad_norm": 3.047305107116699,
      "learning_rate": 3.9942878150665035e-06,
      "loss": 0.2976,
      "num_input_tokens_seen": 13370496,
      "step": 20625
    },
    {
      "epoch": 12.163915094339623,
      "grad_norm": 2.777491807937622,
      "learning_rate": 3.991767815321569e-06,
      "loss": 0.367,
      "num_input_tokens_seen": 13373536,
      "step": 20630
    },
    {
      "epoch": 12.16686320754717,
      "grad_norm": 2.4379189014434814,
      "learning_rate": 3.9892480825104504e-06,
      "loss": 0.2861,
      "num_input_tokens_seen": 13376640,
      "step": 20635
    },
    {
      "epoch": 12.169811320754716,
      "grad_norm": 1.8146593570709229,
      "learning_rate": 3.986728617300257e-06,
      "loss": 0.2491,
      "num_input_tokens_seen": 13379072,
      "step": 20640
    },
    {
      "epoch": 12.172759433962264,
      "grad_norm": 3.0326008796691895,
      "learning_rate": 3.984209420358031e-06,
      "loss": 0.2781,
      "num_input_tokens_seen": 13382144,
      "step": 20645
    },
    {
      "epoch": 12.17570754716981,
      "grad_norm": 3.025916814804077,
      "learning_rate": 3.981690492350738e-06,
      "loss": 0.4024,
      "num_input_tokens_seen": 13384640,
      "step": 20650
    },
    {
      "epoch": 12.178655660377359,
      "grad_norm": 4.100348472595215,
      "learning_rate": 3.979171833945276e-06,
      "loss": 0.4488,
      "num_input_tokens_seen": 13387872,
      "step": 20655
    },
    {
      "epoch": 12.181603773584905,
      "grad_norm": 3.862233877182007,
      "learning_rate": 3.97665344580847e-06,
      "loss": 0.3436,
      "num_input_tokens_seen": 13391296,
      "step": 20660
    },
    {
      "epoch": 12.184551886792454,
      "grad_norm": 4.9223313331604,
      "learning_rate": 3.974135328607075e-06,
      "loss": 0.3906,
      "num_input_tokens_seen": 13394176,
      "step": 20665
    },
    {
      "epoch": 12.1875,
      "grad_norm": 3.1967029571533203,
      "learning_rate": 3.971617483007773e-06,
      "loss": 0.3097,
      "num_input_tokens_seen": 13399200,
      "step": 20670
    },
    {
      "epoch": 12.190448113207546,
      "grad_norm": 1.887581706047058,
      "learning_rate": 3.969099909677174e-06,
      "loss": 0.4141,
      "num_input_tokens_seen": 13402304,
      "step": 20675
    },
    {
      "epoch": 12.193396226415095,
      "grad_norm": 4.471799850463867,
      "learning_rate": 3.9665826092818165e-06,
      "loss": 0.3892,
      "num_input_tokens_seen": 13405504,
      "step": 20680
    },
    {
      "epoch": 12.196344339622641,
      "grad_norm": 5.75496244430542,
      "learning_rate": 3.964065582488168e-06,
      "loss": 0.3806,
      "num_input_tokens_seen": 13408224,
      "step": 20685
    },
    {
      "epoch": 12.19929245283019,
      "grad_norm": 3.226266860961914,
      "learning_rate": 3.961548829962622e-06,
      "loss": 0.268,
      "num_input_tokens_seen": 13412032,
      "step": 20690
    },
    {
      "epoch": 12.202240566037736,
      "grad_norm": 2.4428610801696777,
      "learning_rate": 3.959032352371496e-06,
      "loss": 0.3986,
      "num_input_tokens_seen": 13415648,
      "step": 20695
    },
    {
      "epoch": 12.205188679245284,
      "grad_norm": 3.5039219856262207,
      "learning_rate": 3.956516150381043e-06,
      "loss": 0.301,
      "num_input_tokens_seen": 13418720,
      "step": 20700
    },
    {
      "epoch": 12.20813679245283,
      "grad_norm": 2.9093995094299316,
      "learning_rate": 3.954000224657436e-06,
      "loss": 0.4052,
      "num_input_tokens_seen": 13422176,
      "step": 20705
    },
    {
      "epoch": 12.211084905660377,
      "grad_norm": 3.4216501712799072,
      "learning_rate": 3.951484575866776e-06,
      "loss": 0.3426,
      "num_input_tokens_seen": 13425216,
      "step": 20710
    },
    {
      "epoch": 12.214033018867925,
      "grad_norm": 2.141162872314453,
      "learning_rate": 3.948969204675096e-06,
      "loss": 0.2867,
      "num_input_tokens_seen": 13427904,
      "step": 20715
    },
    {
      "epoch": 12.216981132075471,
      "grad_norm": 2.684997320175171,
      "learning_rate": 3.946454111748346e-06,
      "loss": 0.371,
      "num_input_tokens_seen": 13431520,
      "step": 20720
    },
    {
      "epoch": 12.21992924528302,
      "grad_norm": 4.017430782318115,
      "learning_rate": 3.943939297752413e-06,
      "loss": 0.387,
      "num_input_tokens_seen": 13435232,
      "step": 20725
    },
    {
      "epoch": 12.222877358490566,
      "grad_norm": 3.7079386711120605,
      "learning_rate": 3.9414247633531e-06,
      "loss": 0.4628,
      "num_input_tokens_seen": 13439328,
      "step": 20730
    },
    {
      "epoch": 12.225825471698114,
      "grad_norm": 4.6619553565979,
      "learning_rate": 3.9389105092161454e-06,
      "loss": 0.3826,
      "num_input_tokens_seen": 13442656,
      "step": 20735
    },
    {
      "epoch": 12.22877358490566,
      "grad_norm": 3.954289197921753,
      "learning_rate": 3.936396536007205e-06,
      "loss": 0.3446,
      "num_input_tokens_seen": 13445792,
      "step": 20740
    },
    {
      "epoch": 12.231721698113208,
      "grad_norm": 4.059054374694824,
      "learning_rate": 3.933882844391866e-06,
      "loss": 0.2907,
      "num_input_tokens_seen": 13448224,
      "step": 20745
    },
    {
      "epoch": 12.234669811320755,
      "grad_norm": 2.92807936668396,
      "learning_rate": 3.931369435035639e-06,
      "loss": 0.3605,
      "num_input_tokens_seen": 13452032,
      "step": 20750
    },
    {
      "epoch": 12.237617924528301,
      "grad_norm": 5.254330635070801,
      "learning_rate": 3.92885630860396e-06,
      "loss": 0.3226,
      "num_input_tokens_seen": 13454720,
      "step": 20755
    },
    {
      "epoch": 12.24056603773585,
      "grad_norm": 4.6958770751953125,
      "learning_rate": 3.926343465762189e-06,
      "loss": 0.3334,
      "num_input_tokens_seen": 13457984,
      "step": 20760
    },
    {
      "epoch": 12.243514150943396,
      "grad_norm": 2.216141700744629,
      "learning_rate": 3.923830907175613e-06,
      "loss": 0.2796,
      "num_input_tokens_seen": 13461824,
      "step": 20765
    },
    {
      "epoch": 12.246462264150944,
      "grad_norm": 3.9454567432403564,
      "learning_rate": 3.921318633509442e-06,
      "loss": 0.3593,
      "num_input_tokens_seen": 13464768,
      "step": 20770
    },
    {
      "epoch": 12.24941037735849,
      "grad_norm": 4.507927417755127,
      "learning_rate": 3.918806645428811e-06,
      "loss": 0.4847,
      "num_input_tokens_seen": 13471744,
      "step": 20775
    },
    {
      "epoch": 12.252358490566039,
      "grad_norm": 4.569079875946045,
      "learning_rate": 3.916294943598781e-06,
      "loss": 0.5581,
      "num_input_tokens_seen": 13475136,
      "step": 20780
    },
    {
      "epoch": 12.255306603773585,
      "grad_norm": 6.464389801025391,
      "learning_rate": 3.913783528684336e-06,
      "loss": 0.4908,
      "num_input_tokens_seen": 13477856,
      "step": 20785
    },
    {
      "epoch": 12.258254716981131,
      "grad_norm": 2.636849880218506,
      "learning_rate": 3.911272401350381e-06,
      "loss": 0.368,
      "num_input_tokens_seen": 13481568,
      "step": 20790
    },
    {
      "epoch": 12.26120283018868,
      "grad_norm": 2.6978964805603027,
      "learning_rate": 3.9087615622617525e-06,
      "loss": 0.327,
      "num_input_tokens_seen": 13484832,
      "step": 20795
    },
    {
      "epoch": 12.264150943396226,
      "grad_norm": 1.90450918674469,
      "learning_rate": 3.906251012083202e-06,
      "loss": 0.382,
      "num_input_tokens_seen": 13488128,
      "step": 20800
    },
    {
      "epoch": 12.267099056603774,
      "grad_norm": 3.022653341293335,
      "learning_rate": 3.9037407514794085e-06,
      "loss": 0.4015,
      "num_input_tokens_seen": 13491264,
      "step": 20805
    },
    {
      "epoch": 12.27004716981132,
      "grad_norm": 2.735560178756714,
      "learning_rate": 3.901230781114976e-06,
      "loss": 0.4088,
      "num_input_tokens_seen": 13493792,
      "step": 20810
    },
    {
      "epoch": 12.272995283018869,
      "grad_norm": 3.5207509994506836,
      "learning_rate": 3.898721101654431e-06,
      "loss": 0.3133,
      "num_input_tokens_seen": 13497056,
      "step": 20815
    },
    {
      "epoch": 12.275943396226415,
      "grad_norm": 3.7781591415405273,
      "learning_rate": 3.896211713762221e-06,
      "loss": 0.4116,
      "num_input_tokens_seen": 13500512,
      "step": 20820
    },
    {
      "epoch": 12.278891509433961,
      "grad_norm": 3.4260458946228027,
      "learning_rate": 3.893702618102715e-06,
      "loss": 0.3267,
      "num_input_tokens_seen": 13503424,
      "step": 20825
    },
    {
      "epoch": 12.28183962264151,
      "grad_norm": 1.753051519393921,
      "learning_rate": 3.891193815340211e-06,
      "loss": 0.34,
      "num_input_tokens_seen": 13506144,
      "step": 20830
    },
    {
      "epoch": 12.284787735849056,
      "grad_norm": 3.157878875732422,
      "learning_rate": 3.888685306138922e-06,
      "loss": 0.3756,
      "num_input_tokens_seen": 13509248,
      "step": 20835
    },
    {
      "epoch": 12.287735849056604,
      "grad_norm": 7.547794818878174,
      "learning_rate": 3.886177091162987e-06,
      "loss": 0.2845,
      "num_input_tokens_seen": 13511776,
      "step": 20840
    },
    {
      "epoch": 12.29068396226415,
      "grad_norm": 2.685332775115967,
      "learning_rate": 3.883669171076468e-06,
      "loss": 0.2469,
      "num_input_tokens_seen": 13514688,
      "step": 20845
    },
    {
      "epoch": 12.293632075471699,
      "grad_norm": 2.9590225219726562,
      "learning_rate": 3.881161546543348e-06,
      "loss": 0.3638,
      "num_input_tokens_seen": 13518816,
      "step": 20850
    },
    {
      "epoch": 12.296580188679245,
      "grad_norm": 3.3302195072174072,
      "learning_rate": 3.8786542182275295e-06,
      "loss": 0.4728,
      "num_input_tokens_seen": 13521984,
      "step": 20855
    },
    {
      "epoch": 12.299528301886792,
      "grad_norm": 2.4053521156311035,
      "learning_rate": 3.87614718679284e-06,
      "loss": 0.3925,
      "num_input_tokens_seen": 13524544,
      "step": 20860
    },
    {
      "epoch": 12.30247641509434,
      "grad_norm": 2.3366892337799072,
      "learning_rate": 3.8736404529030255e-06,
      "loss": 0.4469,
      "num_input_tokens_seen": 13528800,
      "step": 20865
    },
    {
      "epoch": 12.305424528301886,
      "grad_norm": 3.248141050338745,
      "learning_rate": 3.871134017221756e-06,
      "loss": 0.4418,
      "num_input_tokens_seen": 13531904,
      "step": 20870
    },
    {
      "epoch": 12.308372641509434,
      "grad_norm": 5.22628116607666,
      "learning_rate": 3.868627880412621e-06,
      "loss": 0.3166,
      "num_input_tokens_seen": 13536096,
      "step": 20875
    },
    {
      "epoch": 12.31132075471698,
      "grad_norm": 5.912030220031738,
      "learning_rate": 3.86612204313913e-06,
      "loss": 0.3332,
      "num_input_tokens_seen": 13539584,
      "step": 20880
    },
    {
      "epoch": 12.314268867924529,
      "grad_norm": 3.4164626598358154,
      "learning_rate": 3.863616506064714e-06,
      "loss": 0.2978,
      "num_input_tokens_seen": 13544544,
      "step": 20885
    },
    {
      "epoch": 12.317216981132075,
      "grad_norm": 3.1908040046691895,
      "learning_rate": 3.861111269852727e-06,
      "loss": 0.4159,
      "num_input_tokens_seen": 13547520,
      "step": 20890
    },
    {
      "epoch": 12.320165094339623,
      "grad_norm": 4.356841564178467,
      "learning_rate": 3.858606335166439e-06,
      "loss": 0.3453,
      "num_input_tokens_seen": 13550592,
      "step": 20895
    },
    {
      "epoch": 12.32311320754717,
      "grad_norm": 6.33398962020874,
      "learning_rate": 3.8561017026690415e-06,
      "loss": 0.3235,
      "num_input_tokens_seen": 13553120,
      "step": 20900
    },
    {
      "epoch": 12.326061320754716,
      "grad_norm": 3.9568796157836914,
      "learning_rate": 3.8535973730236495e-06,
      "loss": 0.3751,
      "num_input_tokens_seen": 13557152,
      "step": 20905
    },
    {
      "epoch": 12.329009433962264,
      "grad_norm": 3.012545108795166,
      "learning_rate": 3.851093346893294e-06,
      "loss": 0.3842,
      "num_input_tokens_seen": 13560128,
      "step": 20910
    },
    {
      "epoch": 12.33195754716981,
      "grad_norm": 2.8436508178710938,
      "learning_rate": 3.848589624940927e-06,
      "loss": 0.3786,
      "num_input_tokens_seen": 13563552,
      "step": 20915
    },
    {
      "epoch": 12.334905660377359,
      "grad_norm": 4.24506950378418,
      "learning_rate": 3.846086207829417e-06,
      "loss": 0.4048,
      "num_input_tokens_seen": 13566112,
      "step": 20920
    },
    {
      "epoch": 12.337853773584905,
      "grad_norm": 7.192109107971191,
      "learning_rate": 3.843583096221559e-06,
      "loss": 0.3725,
      "num_input_tokens_seen": 13569440,
      "step": 20925
    },
    {
      "epoch": 12.340801886792454,
      "grad_norm": 3.50996470451355,
      "learning_rate": 3.8410802907800596e-06,
      "loss": 0.4032,
      "num_input_tokens_seen": 13572064,
      "step": 20930
    },
    {
      "epoch": 12.34375,
      "grad_norm": 4.070260524749756,
      "learning_rate": 3.838577792167546e-06,
      "loss": 0.2444,
      "num_input_tokens_seen": 13575552,
      "step": 20935
    },
    {
      "epoch": 12.346698113207546,
      "grad_norm": 2.8987295627593994,
      "learning_rate": 3.836075601046569e-06,
      "loss": 0.3554,
      "num_input_tokens_seen": 13579776,
      "step": 20940
    },
    {
      "epoch": 12.349646226415095,
      "grad_norm": 5.88673734664917,
      "learning_rate": 3.833573718079594e-06,
      "loss": 0.3128,
      "num_input_tokens_seen": 13582752,
      "step": 20945
    },
    {
      "epoch": 12.352594339622641,
      "grad_norm": 4.507251262664795,
      "learning_rate": 3.831072143929002e-06,
      "loss": 0.3501,
      "num_input_tokens_seen": 13585408,
      "step": 20950
    },
    {
      "epoch": 12.35554245283019,
      "grad_norm": 5.32176399230957,
      "learning_rate": 3.828570879257098e-06,
      "loss": 0.3107,
      "num_input_tokens_seen": 13588256,
      "step": 20955
    },
    {
      "epoch": 12.358490566037736,
      "grad_norm": 3.02897047996521,
      "learning_rate": 3.826069924726102e-06,
      "loss": 0.4919,
      "num_input_tokens_seen": 13590912,
      "step": 20960
    },
    {
      "epoch": 12.361438679245284,
      "grad_norm": 2.3246731758117676,
      "learning_rate": 3.823569280998154e-06,
      "loss": 0.4387,
      "num_input_tokens_seen": 13594272,
      "step": 20965
    },
    {
      "epoch": 12.36438679245283,
      "grad_norm": 4.7758097648620605,
      "learning_rate": 3.8210689487353095e-06,
      "loss": 0.3796,
      "num_input_tokens_seen": 13597056,
      "step": 20970
    },
    {
      "epoch": 12.367334905660377,
      "grad_norm": 2.7862002849578857,
      "learning_rate": 3.818568928599539e-06,
      "loss": 0.3285,
      "num_input_tokens_seen": 13600032,
      "step": 20975
    },
    {
      "epoch": 12.370283018867925,
      "grad_norm": 2.8088343143463135,
      "learning_rate": 3.816069221252739e-06,
      "loss": 0.3259,
      "num_input_tokens_seen": 13603072,
      "step": 20980
    },
    {
      "epoch": 12.373231132075471,
      "grad_norm": 5.925510406494141,
      "learning_rate": 3.813569827356715e-06,
      "loss": 0.3969,
      "num_input_tokens_seen": 13606176,
      "step": 20985
    },
    {
      "epoch": 12.37617924528302,
      "grad_norm": 2.5150413513183594,
      "learning_rate": 3.811070747573191e-06,
      "loss": 0.2993,
      "num_input_tokens_seen": 13610048,
      "step": 20990
    },
    {
      "epoch": 12.379127358490566,
      "grad_norm": 4.110857009887695,
      "learning_rate": 3.8085719825638098e-06,
      "loss": 0.3072,
      "num_input_tokens_seen": 13613024,
      "step": 20995
    },
    {
      "epoch": 12.382075471698114,
      "grad_norm": 5.449738025665283,
      "learning_rate": 3.806073532990132e-06,
      "loss": 0.36,
      "num_input_tokens_seen": 13616000,
      "step": 21000
    },
    {
      "epoch": 12.38502358490566,
      "grad_norm": 3.5726630687713623,
      "learning_rate": 3.8035753995136333e-06,
      "loss": 0.4654,
      "num_input_tokens_seen": 13619808,
      "step": 21005
    },
    {
      "epoch": 12.387971698113208,
      "grad_norm": 4.222095966339111,
      "learning_rate": 3.801077582795702e-06,
      "loss": 0.3668,
      "num_input_tokens_seen": 13623296,
      "step": 21010
    },
    {
      "epoch": 12.390919811320755,
      "grad_norm": 2.333235263824463,
      "learning_rate": 3.7985800834976494e-06,
      "loss": 0.3666,
      "num_input_tokens_seen": 13626752,
      "step": 21015
    },
    {
      "epoch": 12.393867924528301,
      "grad_norm": 2.914191961288452,
      "learning_rate": 3.7960829022806965e-06,
      "loss": 0.2556,
      "num_input_tokens_seen": 13633728,
      "step": 21020
    },
    {
      "epoch": 12.39681603773585,
      "grad_norm": 3.536841869354248,
      "learning_rate": 3.793586039805984e-06,
      "loss": 0.3031,
      "num_input_tokens_seen": 13636928,
      "step": 21025
    },
    {
      "epoch": 12.399764150943396,
      "grad_norm": 2.3785181045532227,
      "learning_rate": 3.791089496734567e-06,
      "loss": 0.3705,
      "num_input_tokens_seen": 13639936,
      "step": 21030
    },
    {
      "epoch": 12.402712264150944,
      "grad_norm": 1.7534916400909424,
      "learning_rate": 3.7885932737274163e-06,
      "loss": 0.3066,
      "num_input_tokens_seen": 13643456,
      "step": 21035
    },
    {
      "epoch": 12.40566037735849,
      "grad_norm": 3.161525249481201,
      "learning_rate": 3.7860973714454156e-06,
      "loss": 0.3553,
      "num_input_tokens_seen": 13646208,
      "step": 21040
    },
    {
      "epoch": 12.408608490566039,
      "grad_norm": 3.5160117149353027,
      "learning_rate": 3.7836017905493695e-06,
      "loss": 0.3547,
      "num_input_tokens_seen": 13649440,
      "step": 21045
    },
    {
      "epoch": 12.411556603773585,
      "grad_norm": 3.5557234287261963,
      "learning_rate": 3.7811065316999908e-06,
      "loss": 0.3185,
      "num_input_tokens_seen": 13652576,
      "step": 21050
    },
    {
      "epoch": 12.414504716981131,
      "grad_norm": 9.336081504821777,
      "learning_rate": 3.7786115955579105e-06,
      "loss": 0.3016,
      "num_input_tokens_seen": 13655488,
      "step": 21055
    },
    {
      "epoch": 12.41745283018868,
      "grad_norm": 3.977860689163208,
      "learning_rate": 3.7761169827836746e-06,
      "loss": 0.4331,
      "num_input_tokens_seen": 13658016,
      "step": 21060
    },
    {
      "epoch": 12.420400943396226,
      "grad_norm": 3.343499183654785,
      "learning_rate": 3.773622694037743e-06,
      "loss": 0.3708,
      "num_input_tokens_seen": 13660736,
      "step": 21065
    },
    {
      "epoch": 12.423349056603774,
      "grad_norm": 2.0473508834838867,
      "learning_rate": 3.7711287299804865e-06,
      "loss": 0.41,
      "num_input_tokens_seen": 13663424,
      "step": 21070
    },
    {
      "epoch": 12.42629716981132,
      "grad_norm": 2.1616837978363037,
      "learning_rate": 3.768635091272197e-06,
      "loss": 0.2722,
      "num_input_tokens_seen": 13666880,
      "step": 21075
    },
    {
      "epoch": 12.429245283018869,
      "grad_norm": 3.1803224086761475,
      "learning_rate": 3.7661417785730732e-06,
      "loss": 0.2885,
      "num_input_tokens_seen": 13669248,
      "step": 21080
    },
    {
      "epoch": 12.432193396226415,
      "grad_norm": 3.3014774322509766,
      "learning_rate": 3.7636487925432304e-06,
      "loss": 0.3511,
      "num_input_tokens_seen": 13672032,
      "step": 21085
    },
    {
      "epoch": 12.435141509433961,
      "grad_norm": 2.7295193672180176,
      "learning_rate": 3.761156133842697e-06,
      "loss": 0.243,
      "num_input_tokens_seen": 13675328,
      "step": 21090
    },
    {
      "epoch": 12.43808962264151,
      "grad_norm": 1.9225109815597534,
      "learning_rate": 3.7586638031314182e-06,
      "loss": 0.2576,
      "num_input_tokens_seen": 13678624,
      "step": 21095
    },
    {
      "epoch": 12.441037735849056,
      "grad_norm": 3.7665469646453857,
      "learning_rate": 3.7561718010692477e-06,
      "loss": 0.3866,
      "num_input_tokens_seen": 13682848,
      "step": 21100
    },
    {
      "epoch": 12.443985849056604,
      "grad_norm": 3.70892071723938,
      "learning_rate": 3.7536801283159523e-06,
      "loss": 0.3579,
      "num_input_tokens_seen": 13686592,
      "step": 21105
    },
    {
      "epoch": 12.44693396226415,
      "grad_norm": 3.69545316696167,
      "learning_rate": 3.7511887855312155e-06,
      "loss": 0.3045,
      "num_input_tokens_seen": 13690464,
      "step": 21110
    },
    {
      "epoch": 12.449882075471699,
      "grad_norm": 3.516422986984253,
      "learning_rate": 3.74869777337463e-06,
      "loss": 0.3241,
      "num_input_tokens_seen": 13692896,
      "step": 21115
    },
    {
      "epoch": 12.452830188679245,
      "grad_norm": 3.0676331520080566,
      "learning_rate": 3.7462070925057004e-06,
      "loss": 0.3213,
      "num_input_tokens_seen": 13695648,
      "step": 21120
    },
    {
      "epoch": 12.455778301886792,
      "grad_norm": 3.947134494781494,
      "learning_rate": 3.7437167435838472e-06,
      "loss": 0.3452,
      "num_input_tokens_seen": 13699552,
      "step": 21125
    },
    {
      "epoch": 12.45872641509434,
      "grad_norm": 2.520995616912842,
      "learning_rate": 3.741226727268401e-06,
      "loss": 0.3015,
      "num_input_tokens_seen": 13702432,
      "step": 21130
    },
    {
      "epoch": 12.461674528301886,
      "grad_norm": 4.870541095733643,
      "learning_rate": 3.7387370442186027e-06,
      "loss": 0.3351,
      "num_input_tokens_seen": 13705984,
      "step": 21135
    },
    {
      "epoch": 12.464622641509434,
      "grad_norm": 3.026940107345581,
      "learning_rate": 3.736247695093609e-06,
      "loss": 0.3962,
      "num_input_tokens_seen": 13709760,
      "step": 21140
    },
    {
      "epoch": 12.46757075471698,
      "grad_norm": 4.213428020477295,
      "learning_rate": 3.7337586805524838e-06,
      "loss": 0.3707,
      "num_input_tokens_seen": 13713472,
      "step": 21145
    },
    {
      "epoch": 12.470518867924529,
      "grad_norm": 3.8962841033935547,
      "learning_rate": 3.731270001254205e-06,
      "loss": 0.4026,
      "num_input_tokens_seen": 13717024,
      "step": 21150
    },
    {
      "epoch": 12.473466981132075,
      "grad_norm": 3.962623357772827,
      "learning_rate": 3.728781657857661e-06,
      "loss": 0.3043,
      "num_input_tokens_seen": 13719584,
      "step": 21155
    },
    {
      "epoch": 12.476415094339623,
      "grad_norm": 2.913325309753418,
      "learning_rate": 3.726293651021653e-06,
      "loss": 0.277,
      "num_input_tokens_seen": 13722272,
      "step": 21160
    },
    {
      "epoch": 12.47936320754717,
      "grad_norm": 3.868835210800171,
      "learning_rate": 3.7238059814048888e-06,
      "loss": 0.3474,
      "num_input_tokens_seen": 13726560,
      "step": 21165
    },
    {
      "epoch": 12.482311320754716,
      "grad_norm": 4.355440616607666,
      "learning_rate": 3.7213186496659916e-06,
      "loss": 0.3975,
      "num_input_tokens_seen": 13729088,
      "step": 21170
    },
    {
      "epoch": 12.485259433962264,
      "grad_norm": 2.630906820297241,
      "learning_rate": 3.718831656463493e-06,
      "loss": 0.2538,
      "num_input_tokens_seen": 13732928,
      "step": 21175
    },
    {
      "epoch": 12.48820754716981,
      "grad_norm": 9.86556625366211,
      "learning_rate": 3.7163450024558345e-06,
      "loss": 0.3395,
      "num_input_tokens_seen": 13735648,
      "step": 21180
    },
    {
      "epoch": 12.491155660377359,
      "grad_norm": 1.9069106578826904,
      "learning_rate": 3.7138586883013694e-06,
      "loss": 0.311,
      "num_input_tokens_seen": 13738496,
      "step": 21185
    },
    {
      "epoch": 12.494103773584905,
      "grad_norm": 2.3359880447387695,
      "learning_rate": 3.7113727146583573e-06,
      "loss": 0.3057,
      "num_input_tokens_seen": 13741504,
      "step": 21190
    },
    {
      "epoch": 12.497051886792454,
      "grad_norm": 2.7532124519348145,
      "learning_rate": 3.7088870821849748e-06,
      "loss": 0.4008,
      "num_input_tokens_seen": 13746112,
      "step": 21195
    },
    {
      "epoch": 12.5,
      "grad_norm": 2.9931788444519043,
      "learning_rate": 3.706401791539301e-06,
      "loss": 0.3269,
      "num_input_tokens_seen": 13749216,
      "step": 21200
    },
    {
      "epoch": 12.502948113207546,
      "grad_norm": 5.825485706329346,
      "learning_rate": 3.703916843379328e-06,
      "loss": 0.5561,
      "num_input_tokens_seen": 13752032,
      "step": 21205
    },
    {
      "epoch": 12.505896226415095,
      "grad_norm": 3.4613800048828125,
      "learning_rate": 3.7014322383629575e-06,
      "loss": 0.293,
      "num_input_tokens_seen": 13755168,
      "step": 21210
    },
    {
      "epoch": 12.508844339622641,
      "grad_norm": 6.2296953201293945,
      "learning_rate": 3.6989479771479976e-06,
      "loss": 0.4738,
      "num_input_tokens_seen": 13758144,
      "step": 21215
    },
    {
      "epoch": 12.51179245283019,
      "grad_norm": 2.4733147621154785,
      "learning_rate": 3.696464060392169e-06,
      "loss": 0.357,
      "num_input_tokens_seen": 13761216,
      "step": 21220
    },
    {
      "epoch": 12.514740566037736,
      "grad_norm": 3.302097797393799,
      "learning_rate": 3.6939804887530962e-06,
      "loss": 0.271,
      "num_input_tokens_seen": 13763776,
      "step": 21225
    },
    {
      "epoch": 12.517688679245284,
      "grad_norm": 2.722210168838501,
      "learning_rate": 3.6914972628883196e-06,
      "loss": 0.3743,
      "num_input_tokens_seen": 13767040,
      "step": 21230
    },
    {
      "epoch": 12.52063679245283,
      "grad_norm": 4.793549537658691,
      "learning_rate": 3.6890143834552814e-06,
      "loss": 0.304,
      "num_input_tokens_seen": 13769824,
      "step": 21235
    },
    {
      "epoch": 12.523584905660378,
      "grad_norm": 4.992142200469971,
      "learning_rate": 3.6865318511113347e-06,
      "loss": 0.3429,
      "num_input_tokens_seen": 13773536,
      "step": 21240
    },
    {
      "epoch": 12.526533018867925,
      "grad_norm": 1.9273591041564941,
      "learning_rate": 3.684049666513742e-06,
      "loss": 0.311,
      "num_input_tokens_seen": 13777696,
      "step": 21245
    },
    {
      "epoch": 12.529481132075471,
      "grad_norm": 3.892261028289795,
      "learning_rate": 3.6815678303196715e-06,
      "loss": 0.2904,
      "num_input_tokens_seen": 13780480,
      "step": 21250
    },
    {
      "epoch": 12.53242924528302,
      "grad_norm": 3.4857256412506104,
      "learning_rate": 3.6790863431861988e-06,
      "loss": 0.2408,
      "num_input_tokens_seen": 13783392,
      "step": 21255
    },
    {
      "epoch": 12.535377358490566,
      "grad_norm": 2.640446901321411,
      "learning_rate": 3.676605205770311e-06,
      "loss": 0.3529,
      "num_input_tokens_seen": 13786496,
      "step": 21260
    },
    {
      "epoch": 12.538325471698114,
      "grad_norm": 4.524130821228027,
      "learning_rate": 3.674124418728898e-06,
      "loss": 0.3678,
      "num_input_tokens_seen": 13788928,
      "step": 21265
    },
    {
      "epoch": 12.54127358490566,
      "grad_norm": 3.4768717288970947,
      "learning_rate": 3.671643982718759e-06,
      "loss": 0.3164,
      "num_input_tokens_seen": 13791872,
      "step": 21270
    },
    {
      "epoch": 12.544221698113208,
      "grad_norm": 3.7001566886901855,
      "learning_rate": 3.6691638983966017e-06,
      "loss": 0.3005,
      "num_input_tokens_seen": 13794432,
      "step": 21275
    },
    {
      "epoch": 12.547169811320755,
      "grad_norm": 2.7958526611328125,
      "learning_rate": 3.666684166419037e-06,
      "loss": 0.4937,
      "num_input_tokens_seen": 13798144,
      "step": 21280
    },
    {
      "epoch": 12.550117924528301,
      "grad_norm": 4.617746353149414,
      "learning_rate": 3.6642047874425833e-06,
      "loss": 0.2711,
      "num_input_tokens_seen": 13800608,
      "step": 21285
    },
    {
      "epoch": 12.55306603773585,
      "grad_norm": 10.611656188964844,
      "learning_rate": 3.661725762123671e-06,
      "loss": 0.3515,
      "num_input_tokens_seen": 13803456,
      "step": 21290
    },
    {
      "epoch": 12.556014150943396,
      "grad_norm": 3.044088363647461,
      "learning_rate": 3.659247091118631e-06,
      "loss": 0.2378,
      "num_input_tokens_seen": 13805760,
      "step": 21295
    },
    {
      "epoch": 12.558962264150944,
      "grad_norm": 2.9877898693084717,
      "learning_rate": 3.6567687750837027e-06,
      "loss": 0.3853,
      "num_input_tokens_seen": 13809440,
      "step": 21300
    },
    {
      "epoch": 12.56191037735849,
      "grad_norm": 3.056567430496216,
      "learning_rate": 3.6542908146750287e-06,
      "loss": 0.4025,
      "num_input_tokens_seen": 13814688,
      "step": 21305
    },
    {
      "epoch": 12.564858490566039,
      "grad_norm": 2.1147873401641846,
      "learning_rate": 3.6518132105486624e-06,
      "loss": 0.2522,
      "num_input_tokens_seen": 13817152,
      "step": 21310
    },
    {
      "epoch": 12.567806603773585,
      "grad_norm": 4.627195358276367,
      "learning_rate": 3.649335963360559e-06,
      "loss": 0.277,
      "num_input_tokens_seen": 13820224,
      "step": 21315
    },
    {
      "epoch": 12.570754716981131,
      "grad_norm": 2.502901315689087,
      "learning_rate": 3.6468590737665795e-06,
      "loss": 0.2825,
      "num_input_tokens_seen": 13823200,
      "step": 21320
    },
    {
      "epoch": 12.57370283018868,
      "grad_norm": 1.6205308437347412,
      "learning_rate": 3.6443825424224926e-06,
      "loss": 0.3547,
      "num_input_tokens_seen": 13826976,
      "step": 21325
    },
    {
      "epoch": 12.576650943396226,
      "grad_norm": 2.505662441253662,
      "learning_rate": 3.64190636998397e-06,
      "loss": 0.3141,
      "num_input_tokens_seen": 13829440,
      "step": 21330
    },
    {
      "epoch": 12.579599056603774,
      "grad_norm": 4.512836933135986,
      "learning_rate": 3.639430557106588e-06,
      "loss": 0.4114,
      "num_input_tokens_seen": 13833984,
      "step": 21335
    },
    {
      "epoch": 12.58254716981132,
      "grad_norm": 2.9712071418762207,
      "learning_rate": 3.6369551044458314e-06,
      "loss": 0.2728,
      "num_input_tokens_seen": 13837088,
      "step": 21340
    },
    {
      "epoch": 12.585495283018869,
      "grad_norm": 2.4690256118774414,
      "learning_rate": 3.6344800126570846e-06,
      "loss": 0.3546,
      "num_input_tokens_seen": 13839808,
      "step": 21345
    },
    {
      "epoch": 12.588443396226415,
      "grad_norm": 4.813662528991699,
      "learning_rate": 3.6320052823956385e-06,
      "loss": 0.2845,
      "num_input_tokens_seen": 13842496,
      "step": 21350
    },
    {
      "epoch": 12.591391509433961,
      "grad_norm": 3.5223662853240967,
      "learning_rate": 3.6295309143166906e-06,
      "loss": 0.3823,
      "num_input_tokens_seen": 13845152,
      "step": 21355
    },
    {
      "epoch": 12.59433962264151,
      "grad_norm": 3.372483968734741,
      "learning_rate": 3.62705690907534e-06,
      "loss": 0.3495,
      "num_input_tokens_seen": 13848352,
      "step": 21360
    },
    {
      "epoch": 12.597287735849056,
      "grad_norm": 2.820446491241455,
      "learning_rate": 3.624583267326588e-06,
      "loss": 0.3349,
      "num_input_tokens_seen": 13851456,
      "step": 21365
    },
    {
      "epoch": 12.600235849056604,
      "grad_norm": 4.148345470428467,
      "learning_rate": 3.6221099897253454e-06,
      "loss": 0.2104,
      "num_input_tokens_seen": 13855040,
      "step": 21370
    },
    {
      "epoch": 12.60318396226415,
      "grad_norm": 4.422999382019043,
      "learning_rate": 3.619637076926421e-06,
      "loss": 0.4213,
      "num_input_tokens_seen": 13858272,
      "step": 21375
    },
    {
      "epoch": 12.606132075471699,
      "grad_norm": 2.6791725158691406,
      "learning_rate": 3.617164529584528e-06,
      "loss": 0.3785,
      "num_input_tokens_seen": 13860992,
      "step": 21380
    },
    {
      "epoch": 12.609080188679245,
      "grad_norm": 3.4156031608581543,
      "learning_rate": 3.614692348354286e-06,
      "loss": 0.3067,
      "num_input_tokens_seen": 13864896,
      "step": 21385
    },
    {
      "epoch": 12.612028301886792,
      "grad_norm": 2.7312161922454834,
      "learning_rate": 3.612220533890216e-06,
      "loss": 0.3533,
      "num_input_tokens_seen": 13868256,
      "step": 21390
    },
    {
      "epoch": 12.61497641509434,
      "grad_norm": 5.4737653732299805,
      "learning_rate": 3.609749086846741e-06,
      "loss": 0.3565,
      "num_input_tokens_seen": 13872192,
      "step": 21395
    },
    {
      "epoch": 12.617924528301886,
      "grad_norm": 4.880005836486816,
      "learning_rate": 3.607278007878186e-06,
      "loss": 0.3452,
      "num_input_tokens_seen": 13875264,
      "step": 21400
    },
    {
      "epoch": 12.620872641509434,
      "grad_norm": 2.864842653274536,
      "learning_rate": 3.6048072976387817e-06,
      "loss": 0.3416,
      "num_input_tokens_seen": 13878560,
      "step": 21405
    },
    {
      "epoch": 12.62382075471698,
      "grad_norm": 4.015249729156494,
      "learning_rate": 3.6023369567826585e-06,
      "loss": 0.2825,
      "num_input_tokens_seen": 13881536,
      "step": 21410
    },
    {
      "epoch": 12.626768867924529,
      "grad_norm": 2.572659492492676,
      "learning_rate": 3.599866985963848e-06,
      "loss": 0.3408,
      "num_input_tokens_seen": 13884576,
      "step": 21415
    },
    {
      "epoch": 12.629716981132075,
      "grad_norm": 4.112155914306641,
      "learning_rate": 3.5973973858362885e-06,
      "loss": 0.3297,
      "num_input_tokens_seen": 13886976,
      "step": 21420
    },
    {
      "epoch": 12.632665094339622,
      "grad_norm": 3.4175901412963867,
      "learning_rate": 3.594928157053816e-06,
      "loss": 0.3037,
      "num_input_tokens_seen": 13890304,
      "step": 21425
    },
    {
      "epoch": 12.63561320754717,
      "grad_norm": 4.474735736846924,
      "learning_rate": 3.592459300270168e-06,
      "loss": 0.3217,
      "num_input_tokens_seen": 13893184,
      "step": 21430
    },
    {
      "epoch": 12.638561320754716,
      "grad_norm": 2.4459691047668457,
      "learning_rate": 3.589990816138988e-06,
      "loss": 0.4726,
      "num_input_tokens_seen": 13895872,
      "step": 21435
    },
    {
      "epoch": 12.641509433962264,
      "grad_norm": 2.4831316471099854,
      "learning_rate": 3.587522705313816e-06,
      "loss": 0.3095,
      "num_input_tokens_seen": 13898656,
      "step": 21440
    },
    {
      "epoch": 12.64445754716981,
      "grad_norm": 3.3452038764953613,
      "learning_rate": 3.585054968448094e-06,
      "loss": 0.3022,
      "num_input_tokens_seen": 13903616,
      "step": 21445
    },
    {
      "epoch": 12.647405660377359,
      "grad_norm": 3.332927703857422,
      "learning_rate": 3.5825876061951686e-06,
      "loss": 0.2991,
      "num_input_tokens_seen": 13906304,
      "step": 21450
    },
    {
      "epoch": 12.650353773584905,
      "grad_norm": 2.7298052310943604,
      "learning_rate": 3.5801206192082818e-06,
      "loss": 0.331,
      "num_input_tokens_seen": 13908736,
      "step": 21455
    },
    {
      "epoch": 12.653301886792454,
      "grad_norm": 3.445129156112671,
      "learning_rate": 3.577654008140582e-06,
      "loss": 0.289,
      "num_input_tokens_seen": 13911840,
      "step": 21460
    },
    {
      "epoch": 12.65625,
      "grad_norm": 1.7903727293014526,
      "learning_rate": 3.5751877736451123e-06,
      "loss": 0.2897,
      "num_input_tokens_seen": 13915264,
      "step": 21465
    },
    {
      "epoch": 12.659198113207546,
      "grad_norm": 2.5671563148498535,
      "learning_rate": 3.5727219163748205e-06,
      "loss": 0.3214,
      "num_input_tokens_seen": 13918624,
      "step": 21470
    },
    {
      "epoch": 12.662146226415095,
      "grad_norm": 2.2534425258636475,
      "learning_rate": 3.570256436982552e-06,
      "loss": 0.241,
      "num_input_tokens_seen": 13921568,
      "step": 21475
    },
    {
      "epoch": 12.665094339622641,
      "grad_norm": 4.76428747177124,
      "learning_rate": 3.5677913361210536e-06,
      "loss": 0.2582,
      "num_input_tokens_seen": 13924928,
      "step": 21480
    },
    {
      "epoch": 12.66804245283019,
      "grad_norm": 3.297733783721924,
      "learning_rate": 3.565326614442972e-06,
      "loss": 0.4781,
      "num_input_tokens_seen": 13928608,
      "step": 21485
    },
    {
      "epoch": 12.670990566037736,
      "grad_norm": 5.519156455993652,
      "learning_rate": 3.5628622726008523e-06,
      "loss": 0.4454,
      "num_input_tokens_seen": 13931008,
      "step": 21490
    },
    {
      "epoch": 12.673938679245284,
      "grad_norm": 2.6633803844451904,
      "learning_rate": 3.56039831124714e-06,
      "loss": 0.2851,
      "num_input_tokens_seen": 13934208,
      "step": 21495
    },
    {
      "epoch": 12.67688679245283,
      "grad_norm": 4.1311540603637695,
      "learning_rate": 3.557934731034179e-06,
      "loss": 0.2857,
      "num_input_tokens_seen": 13936672,
      "step": 21500
    },
    {
      "epoch": 12.679834905660378,
      "grad_norm": 3.6707775592803955,
      "learning_rate": 3.5554715326142126e-06,
      "loss": 0.3531,
      "num_input_tokens_seen": 13940288,
      "step": 21505
    },
    {
      "epoch": 12.682783018867925,
      "grad_norm": 3.323411464691162,
      "learning_rate": 3.553008716639384e-06,
      "loss": 0.3378,
      "num_input_tokens_seen": 13944256,
      "step": 21510
    },
    {
      "epoch": 12.685731132075471,
      "grad_norm": 5.7508111000061035,
      "learning_rate": 3.5505462837617338e-06,
      "loss": 0.4005,
      "num_input_tokens_seen": 13947104,
      "step": 21515
    },
    {
      "epoch": 12.68867924528302,
      "grad_norm": 3.2883739471435547,
      "learning_rate": 3.5480842346332013e-06,
      "loss": 0.2418,
      "num_input_tokens_seen": 13950080,
      "step": 21520
    },
    {
      "epoch": 12.691627358490566,
      "grad_norm": 2.693885326385498,
      "learning_rate": 3.5456225699056256e-06,
      "loss": 0.4072,
      "num_input_tokens_seen": 13952896,
      "step": 21525
    },
    {
      "epoch": 12.694575471698114,
      "grad_norm": 3.2553627490997314,
      "learning_rate": 3.5431612902307426e-06,
      "loss": 0.3394,
      "num_input_tokens_seen": 13955232,
      "step": 21530
    },
    {
      "epoch": 12.69752358490566,
      "grad_norm": 3.5387723445892334,
      "learning_rate": 3.540700396260186e-06,
      "loss": 0.287,
      "num_input_tokens_seen": 13958208,
      "step": 21535
    },
    {
      "epoch": 12.700471698113208,
      "grad_norm": 2.5404090881347656,
      "learning_rate": 3.538239888645489e-06,
      "loss": 0.2686,
      "num_input_tokens_seen": 13962144,
      "step": 21540
    },
    {
      "epoch": 12.703419811320755,
      "grad_norm": 4.926163196563721,
      "learning_rate": 3.535779768038082e-06,
      "loss": 0.3832,
      "num_input_tokens_seen": 13966272,
      "step": 21545
    },
    {
      "epoch": 12.706367924528301,
      "grad_norm": 2.312826156616211,
      "learning_rate": 3.5333200350892905e-06,
      "loss": 0.3208,
      "num_input_tokens_seen": 13969216,
      "step": 21550
    },
    {
      "epoch": 12.70931603773585,
      "grad_norm": 2.584332227706909,
      "learning_rate": 3.530860690450342e-06,
      "loss": 0.3434,
      "num_input_tokens_seen": 13972640,
      "step": 21555
    },
    {
      "epoch": 12.712264150943396,
      "grad_norm": 4.989273548126221,
      "learning_rate": 3.528401734772357e-06,
      "loss": 0.3771,
      "num_input_tokens_seen": 13975424,
      "step": 21560
    },
    {
      "epoch": 12.715212264150944,
      "grad_norm": 3.295525550842285,
      "learning_rate": 3.5259431687063538e-06,
      "loss": 0.2448,
      "num_input_tokens_seen": 13979168,
      "step": 21565
    },
    {
      "epoch": 12.71816037735849,
      "grad_norm": 3.308976173400879,
      "learning_rate": 3.523484992903249e-06,
      "loss": 0.3491,
      "num_input_tokens_seen": 13982016,
      "step": 21570
    },
    {
      "epoch": 12.721108490566039,
      "grad_norm": 5.160500526428223,
      "learning_rate": 3.5210272080138573e-06,
      "loss": 0.468,
      "num_input_tokens_seen": 13984608,
      "step": 21575
    },
    {
      "epoch": 12.724056603773585,
      "grad_norm": 1.275927186012268,
      "learning_rate": 3.518569814688887e-06,
      "loss": 0.2979,
      "num_input_tokens_seen": 13987040,
      "step": 21580
    },
    {
      "epoch": 12.727004716981131,
      "grad_norm": 2.590627908706665,
      "learning_rate": 3.5161128135789414e-06,
      "loss": 0.2823,
      "num_input_tokens_seen": 13989440,
      "step": 21585
    },
    {
      "epoch": 12.72995283018868,
      "grad_norm": 3.2968063354492188,
      "learning_rate": 3.513656205334525e-06,
      "loss": 0.348,
      "num_input_tokens_seen": 13992928,
      "step": 21590
    },
    {
      "epoch": 12.732900943396226,
      "grad_norm": 6.143603801727295,
      "learning_rate": 3.5111999906060336e-06,
      "loss": 0.3234,
      "num_input_tokens_seen": 13996160,
      "step": 21595
    },
    {
      "epoch": 12.735849056603774,
      "grad_norm": 4.532614231109619,
      "learning_rate": 3.50874417004376e-06,
      "loss": 0.448,
      "num_input_tokens_seen": 13999680,
      "step": 21600
    },
    {
      "epoch": 12.73879716981132,
      "grad_norm": 6.540213584899902,
      "learning_rate": 3.5062887442978956e-06,
      "loss": 0.4142,
      "num_input_tokens_seen": 14003392,
      "step": 21605
    },
    {
      "epoch": 12.741745283018869,
      "grad_norm": 4.015793323516846,
      "learning_rate": 3.503833714018524e-06,
      "loss": 0.4003,
      "num_input_tokens_seen": 14007040,
      "step": 21610
    },
    {
      "epoch": 12.744693396226415,
      "grad_norm": 1.9706329107284546,
      "learning_rate": 3.5013790798556228e-06,
      "loss": 0.3187,
      "num_input_tokens_seen": 14009344,
      "step": 21615
    },
    {
      "epoch": 12.747641509433961,
      "grad_norm": 2.8999273777008057,
      "learning_rate": 3.4989248424590705e-06,
      "loss": 0.3228,
      "num_input_tokens_seen": 14012736,
      "step": 21620
    },
    {
      "epoch": 12.75058962264151,
      "grad_norm": 2.9620847702026367,
      "learning_rate": 3.4964710024786354e-06,
      "loss": 0.3153,
      "num_input_tokens_seen": 14016064,
      "step": 21625
    },
    {
      "epoch": 12.753537735849056,
      "grad_norm": 4.663697242736816,
      "learning_rate": 3.4940175605639813e-06,
      "loss": 0.3486,
      "num_input_tokens_seen": 14018816,
      "step": 21630
    },
    {
      "epoch": 12.756485849056604,
      "grad_norm": 2.6757993698120117,
      "learning_rate": 3.4915645173646694e-06,
      "loss": 0.3361,
      "num_input_tokens_seen": 14021472,
      "step": 21635
    },
    {
      "epoch": 12.75943396226415,
      "grad_norm": 6.0549635887146,
      "learning_rate": 3.489111873530153e-06,
      "loss": 0.2318,
      "num_input_tokens_seen": 14024832,
      "step": 21640
    },
    {
      "epoch": 12.762382075471699,
      "grad_norm": 2.936457633972168,
      "learning_rate": 3.4866596297097776e-06,
      "loss": 0.316,
      "num_input_tokens_seen": 14029216,
      "step": 21645
    },
    {
      "epoch": 12.765330188679245,
      "grad_norm": 3.251178026199341,
      "learning_rate": 3.484207786552789e-06,
      "loss": 0.306,
      "num_input_tokens_seen": 14032384,
      "step": 21650
    },
    {
      "epoch": 12.768278301886792,
      "grad_norm": 3.4207346439361572,
      "learning_rate": 3.4817563447083214e-06,
      "loss": 0.4266,
      "num_input_tokens_seen": 14036224,
      "step": 21655
    },
    {
      "epoch": 12.77122641509434,
      "grad_norm": 4.541597843170166,
      "learning_rate": 3.4793053048254044e-06,
      "loss": 0.3724,
      "num_input_tokens_seen": 14039456,
      "step": 21660
    },
    {
      "epoch": 12.774174528301886,
      "grad_norm": 3.1736819744110107,
      "learning_rate": 3.47685466755296e-06,
      "loss": 0.2684,
      "num_input_tokens_seen": 14042496,
      "step": 21665
    },
    {
      "epoch": 12.777122641509434,
      "grad_norm": 4.240426063537598,
      "learning_rate": 3.474404433539809e-06,
      "loss": 0.2957,
      "num_input_tokens_seen": 14045120,
      "step": 21670
    },
    {
      "epoch": 12.78007075471698,
      "grad_norm": 2.0742833614349365,
      "learning_rate": 3.4719546034346598e-06,
      "loss": 0.3134,
      "num_input_tokens_seen": 14048480,
      "step": 21675
    },
    {
      "epoch": 12.783018867924529,
      "grad_norm": 2.577871561050415,
      "learning_rate": 3.4695051778861125e-06,
      "loss": 0.331,
      "num_input_tokens_seen": 14051360,
      "step": 21680
    },
    {
      "epoch": 12.785966981132075,
      "grad_norm": 2.7128562927246094,
      "learning_rate": 3.4670561575426677e-06,
      "loss": 0.3775,
      "num_input_tokens_seen": 14055584,
      "step": 21685
    },
    {
      "epoch": 12.788915094339622,
      "grad_norm": 2.118516445159912,
      "learning_rate": 3.4646075430527115e-06,
      "loss": 0.3639,
      "num_input_tokens_seen": 14058304,
      "step": 21690
    },
    {
      "epoch": 12.79186320754717,
      "grad_norm": 3.026690721511841,
      "learning_rate": 3.4621593350645236e-06,
      "loss": 0.4653,
      "num_input_tokens_seen": 14062048,
      "step": 21695
    },
    {
      "epoch": 12.794811320754716,
      "grad_norm": 3.4182021617889404,
      "learning_rate": 3.4597115342262817e-06,
      "loss": 0.2921,
      "num_input_tokens_seen": 14065120,
      "step": 21700
    },
    {
      "epoch": 12.797759433962264,
      "grad_norm": 2.1781680583953857,
      "learning_rate": 3.4572641411860484e-06,
      "loss": 0.3422,
      "num_input_tokens_seen": 14068512,
      "step": 21705
    },
    {
      "epoch": 12.80070754716981,
      "grad_norm": 1.8972536325454712,
      "learning_rate": 3.454817156591782e-06,
      "loss": 0.336,
      "num_input_tokens_seen": 14072352,
      "step": 21710
    },
    {
      "epoch": 12.803655660377359,
      "grad_norm": 2.3235831260681152,
      "learning_rate": 3.4523705810913344e-06,
      "loss": 0.2854,
      "num_input_tokens_seen": 14075616,
      "step": 21715
    },
    {
      "epoch": 12.806603773584905,
      "grad_norm": 2.711843967437744,
      "learning_rate": 3.449924415332443e-06,
      "loss": 0.2299,
      "num_input_tokens_seen": 14079584,
      "step": 21720
    },
    {
      "epoch": 12.809551886792454,
      "grad_norm": 2.696559429168701,
      "learning_rate": 3.447478659962745e-06,
      "loss": 0.2709,
      "num_input_tokens_seen": 14083072,
      "step": 21725
    },
    {
      "epoch": 12.8125,
      "grad_norm": 2.1059091091156006,
      "learning_rate": 3.4450333156297625e-06,
      "loss": 0.3242,
      "num_input_tokens_seen": 14086432,
      "step": 21730
    },
    {
      "epoch": 12.815448113207546,
      "grad_norm": 3.616691827774048,
      "learning_rate": 3.44258838298091e-06,
      "loss": 0.3667,
      "num_input_tokens_seen": 14089824,
      "step": 21735
    },
    {
      "epoch": 12.818396226415095,
      "grad_norm": 4.454730987548828,
      "learning_rate": 3.440143862663497e-06,
      "loss": 0.2656,
      "num_input_tokens_seen": 14092384,
      "step": 21740
    },
    {
      "epoch": 12.821344339622641,
      "grad_norm": 1.9241403341293335,
      "learning_rate": 3.4376997553247183e-06,
      "loss": 0.3809,
      "num_input_tokens_seen": 14095360,
      "step": 21745
    },
    {
      "epoch": 12.82429245283019,
      "grad_norm": 4.388096809387207,
      "learning_rate": 3.4352560616116617e-06,
      "loss": 0.2802,
      "num_input_tokens_seen": 14098048,
      "step": 21750
    },
    {
      "epoch": 12.827240566037736,
      "grad_norm": 3.7661051750183105,
      "learning_rate": 3.4328127821713077e-06,
      "loss": 0.2709,
      "num_input_tokens_seen": 14100864,
      "step": 21755
    },
    {
      "epoch": 12.830188679245284,
      "grad_norm": 5.688485145568848,
      "learning_rate": 3.430369917650521e-06,
      "loss": 0.2787,
      "num_input_tokens_seen": 14103776,
      "step": 21760
    },
    {
      "epoch": 12.83313679245283,
      "grad_norm": 8.63870620727539,
      "learning_rate": 3.427927468696066e-06,
      "loss": 0.455,
      "num_input_tokens_seen": 14106720,
      "step": 21765
    },
    {
      "epoch": 12.836084905660378,
      "grad_norm": 4.201144695281982,
      "learning_rate": 3.425485435954588e-06,
      "loss": 0.5156,
      "num_input_tokens_seen": 14110592,
      "step": 21770
    },
    {
      "epoch": 12.839033018867925,
      "grad_norm": 3.368925094604492,
      "learning_rate": 3.4230438200726274e-06,
      "loss": 0.3185,
      "num_input_tokens_seen": 14113568,
      "step": 21775
    },
    {
      "epoch": 12.841981132075471,
      "grad_norm": 4.010909080505371,
      "learning_rate": 3.4206026216966113e-06,
      "loss": 0.434,
      "num_input_tokens_seen": 14116608,
      "step": 21780
    },
    {
      "epoch": 12.84492924528302,
      "grad_norm": 3.860623359680176,
      "learning_rate": 3.418161841472858e-06,
      "loss": 0.3166,
      "num_input_tokens_seen": 14119488,
      "step": 21785
    },
    {
      "epoch": 12.847877358490566,
      "grad_norm": 2.080955743789673,
      "learning_rate": 3.4157214800475746e-06,
      "loss": 0.4681,
      "num_input_tokens_seen": 14122080,
      "step": 21790
    },
    {
      "epoch": 12.850825471698114,
      "grad_norm": 4.032067775726318,
      "learning_rate": 3.4132815380668577e-06,
      "loss": 0.3423,
      "num_input_tokens_seen": 14125152,
      "step": 21795
    },
    {
      "epoch": 12.85377358490566,
      "grad_norm": 4.418211460113525,
      "learning_rate": 3.410842016176691e-06,
      "loss": 0.4043,
      "num_input_tokens_seen": 14127776,
      "step": 21800
    },
    {
      "epoch": 12.856721698113208,
      "grad_norm": 5.2195305824279785,
      "learning_rate": 3.4084029150229503e-06,
      "loss": 0.3776,
      "num_input_tokens_seen": 14130560,
      "step": 21805
    },
    {
      "epoch": 12.859669811320755,
      "grad_norm": 2.390761137008667,
      "learning_rate": 3.4059642352513965e-06,
      "loss": 0.3633,
      "num_input_tokens_seen": 14133696,
      "step": 21810
    },
    {
      "epoch": 12.862617924528301,
      "grad_norm": 3.4236953258514404,
      "learning_rate": 3.4035259775076813e-06,
      "loss": 0.3293,
      "num_input_tokens_seen": 14137856,
      "step": 21815
    },
    {
      "epoch": 12.86556603773585,
      "grad_norm": 2.6092641353607178,
      "learning_rate": 3.401088142437344e-06,
      "loss": 0.2999,
      "num_input_tokens_seen": 14141472,
      "step": 21820
    },
    {
      "epoch": 12.868514150943396,
      "grad_norm": 4.944450855255127,
      "learning_rate": 3.398650730685813e-06,
      "loss": 0.3852,
      "num_input_tokens_seen": 14144672,
      "step": 21825
    },
    {
      "epoch": 12.871462264150944,
      "grad_norm": 4.03083610534668,
      "learning_rate": 3.396213742898401e-06,
      "loss": 0.3864,
      "num_input_tokens_seen": 14147584,
      "step": 21830
    },
    {
      "epoch": 12.87441037735849,
      "grad_norm": 3.2291510105133057,
      "learning_rate": 3.3937771797203134e-06,
      "loss": 0.4615,
      "num_input_tokens_seen": 14151360,
      "step": 21835
    },
    {
      "epoch": 12.877358490566039,
      "grad_norm": 4.421543121337891,
      "learning_rate": 3.391341041796641e-06,
      "loss": 0.3594,
      "num_input_tokens_seen": 14154112,
      "step": 21840
    },
    {
      "epoch": 12.880306603773585,
      "grad_norm": 2.5148632526397705,
      "learning_rate": 3.3889053297723585e-06,
      "loss": 0.4283,
      "num_input_tokens_seen": 14157024,
      "step": 21845
    },
    {
      "epoch": 12.883254716981131,
      "grad_norm": 2.915168046951294,
      "learning_rate": 3.3864700442923342e-06,
      "loss": 0.4444,
      "num_input_tokens_seen": 14159776,
      "step": 21850
    },
    {
      "epoch": 12.88620283018868,
      "grad_norm": 2.9628310203552246,
      "learning_rate": 3.384035186001318e-06,
      "loss": 0.4104,
      "num_input_tokens_seen": 14163072,
      "step": 21855
    },
    {
      "epoch": 12.889150943396226,
      "grad_norm": 2.9128031730651855,
      "learning_rate": 3.381600755543953e-06,
      "loss": 0.3517,
      "num_input_tokens_seen": 14165984,
      "step": 21860
    },
    {
      "epoch": 12.892099056603774,
      "grad_norm": 4.843732833862305,
      "learning_rate": 3.3791667535647615e-06,
      "loss": 0.3593,
      "num_input_tokens_seen": 14169344,
      "step": 21865
    },
    {
      "epoch": 12.89504716981132,
      "grad_norm": 3.573519706726074,
      "learning_rate": 3.3767331807081584e-06,
      "loss": 0.3126,
      "num_input_tokens_seen": 14172096,
      "step": 21870
    },
    {
      "epoch": 12.897995283018869,
      "grad_norm": 3.5519564151763916,
      "learning_rate": 3.374300037618442e-06,
      "loss": 0.3639,
      "num_input_tokens_seen": 14175264,
      "step": 21875
    },
    {
      "epoch": 12.900943396226415,
      "grad_norm": 2.5191917419433594,
      "learning_rate": 3.371867324939796e-06,
      "loss": 0.4057,
      "num_input_tokens_seen": 14178432,
      "step": 21880
    },
    {
      "epoch": 12.903891509433961,
      "grad_norm": 2.2572503089904785,
      "learning_rate": 3.369435043316293e-06,
      "loss": 0.3646,
      "num_input_tokens_seen": 14181504,
      "step": 21885
    },
    {
      "epoch": 12.90683962264151,
      "grad_norm": 3.0702364444732666,
      "learning_rate": 3.36700319339189e-06,
      "loss": 0.3768,
      "num_input_tokens_seen": 14185088,
      "step": 21890
    },
    {
      "epoch": 12.909787735849056,
      "grad_norm": 4.795702934265137,
      "learning_rate": 3.3645717758104286e-06,
      "loss": 0.2757,
      "num_input_tokens_seen": 14187552,
      "step": 21895
    },
    {
      "epoch": 12.912735849056604,
      "grad_norm": 2.6602847576141357,
      "learning_rate": 3.3621407912156383e-06,
      "loss": 0.3912,
      "num_input_tokens_seen": 14190912,
      "step": 21900
    },
    {
      "epoch": 12.91568396226415,
      "grad_norm": 2.257739543914795,
      "learning_rate": 3.3597102402511326e-06,
      "loss": 0.2983,
      "num_input_tokens_seen": 14194144,
      "step": 21905
    },
    {
      "epoch": 12.918632075471699,
      "grad_norm": 2.3334834575653076,
      "learning_rate": 3.3572801235604093e-06,
      "loss": 0.4103,
      "num_input_tokens_seen": 14197920,
      "step": 21910
    },
    {
      "epoch": 12.921580188679245,
      "grad_norm": 2.1643881797790527,
      "learning_rate": 3.3548504417868538e-06,
      "loss": 0.33,
      "num_input_tokens_seen": 14201312,
      "step": 21915
    },
    {
      "epoch": 12.924528301886792,
      "grad_norm": 6.3833160400390625,
      "learning_rate": 3.352421195573734e-06,
      "loss": 0.243,
      "num_input_tokens_seen": 14204768,
      "step": 21920
    },
    {
      "epoch": 12.92747641509434,
      "grad_norm": 2.8357837200164795,
      "learning_rate": 3.3499923855642026e-06,
      "loss": 0.2619,
      "num_input_tokens_seen": 14207712,
      "step": 21925
    },
    {
      "epoch": 12.930424528301886,
      "grad_norm": 4.098731994628906,
      "learning_rate": 3.3475640124012986e-06,
      "loss": 0.4547,
      "num_input_tokens_seen": 14211008,
      "step": 21930
    },
    {
      "epoch": 12.933372641509434,
      "grad_norm": 2.5532901287078857,
      "learning_rate": 3.345136076727945e-06,
      "loss": 0.3312,
      "num_input_tokens_seen": 14213824,
      "step": 21935
    },
    {
      "epoch": 12.93632075471698,
      "grad_norm": 4.6192169189453125,
      "learning_rate": 3.3427085791869453e-06,
      "loss": 0.3263,
      "num_input_tokens_seen": 14216608,
      "step": 21940
    },
    {
      "epoch": 12.939268867924529,
      "grad_norm": 4.828625202178955,
      "learning_rate": 3.3402815204209926e-06,
      "loss": 0.3611,
      "num_input_tokens_seen": 14219552,
      "step": 21945
    },
    {
      "epoch": 12.942216981132075,
      "grad_norm": 3.43560528755188,
      "learning_rate": 3.337854901072659e-06,
      "loss": 0.575,
      "num_input_tokens_seen": 14226560,
      "step": 21950
    },
    {
      "epoch": 12.945165094339622,
      "grad_norm": 3.634141206741333,
      "learning_rate": 3.3354287217844056e-06,
      "loss": 0.34,
      "num_input_tokens_seen": 14229120,
      "step": 21955
    },
    {
      "epoch": 12.94811320754717,
      "grad_norm": 2.5680947303771973,
      "learning_rate": 3.3330029831985712e-06,
      "loss": 0.2913,
      "num_input_tokens_seen": 14232544,
      "step": 21960
    },
    {
      "epoch": 12.951061320754716,
      "grad_norm": 3.2911276817321777,
      "learning_rate": 3.330577685957382e-06,
      "loss": 0.4825,
      "num_input_tokens_seen": 14235328,
      "step": 21965
    },
    {
      "epoch": 12.954009433962264,
      "grad_norm": 1.723159909248352,
      "learning_rate": 3.3281528307029454e-06,
      "loss": 0.3345,
      "num_input_tokens_seen": 14239296,
      "step": 21970
    },
    {
      "epoch": 12.95695754716981,
      "grad_norm": 4.273426532745361,
      "learning_rate": 3.325728418077251e-06,
      "loss": 0.4926,
      "num_input_tokens_seen": 14242528,
      "step": 21975
    },
    {
      "epoch": 12.959905660377359,
      "grad_norm": 4.703918933868408,
      "learning_rate": 3.3233044487221744e-06,
      "loss": 0.3581,
      "num_input_tokens_seen": 14245024,
      "step": 21980
    },
    {
      "epoch": 12.962853773584905,
      "grad_norm": 4.270314693450928,
      "learning_rate": 3.3208809232794715e-06,
      "loss": 0.3123,
      "num_input_tokens_seen": 14247584,
      "step": 21985
    },
    {
      "epoch": 12.965801886792454,
      "grad_norm": 5.81569766998291,
      "learning_rate": 3.3184578423907797e-06,
      "loss": 0.2993,
      "num_input_tokens_seen": 14250304,
      "step": 21990
    },
    {
      "epoch": 12.96875,
      "grad_norm": 3.177396774291992,
      "learning_rate": 3.3160352066976224e-06,
      "loss": 0.3678,
      "num_input_tokens_seen": 14254016,
      "step": 21995
    },
    {
      "epoch": 12.971698113207546,
      "grad_norm": 8.722487449645996,
      "learning_rate": 3.3136130168414003e-06,
      "loss": 0.4029,
      "num_input_tokens_seen": 14256480,
      "step": 22000
    },
    {
      "epoch": 12.974646226415095,
      "grad_norm": 3.227386474609375,
      "learning_rate": 3.311191273463401e-06,
      "loss": 0.2991,
      "num_input_tokens_seen": 14260608,
      "step": 22005
    },
    {
      "epoch": 12.977594339622641,
      "grad_norm": 4.6679158210754395,
      "learning_rate": 3.3087699772047908e-06,
      "loss": 0.2902,
      "num_input_tokens_seen": 14263360,
      "step": 22010
    },
    {
      "epoch": 12.98054245283019,
      "grad_norm": 4.9804582595825195,
      "learning_rate": 3.3063491287066164e-06,
      "loss": 0.3972,
      "num_input_tokens_seen": 14266272,
      "step": 22015
    },
    {
      "epoch": 12.983490566037736,
      "grad_norm": 2.4663562774658203,
      "learning_rate": 3.303928728609811e-06,
      "loss": 0.3901,
      "num_input_tokens_seen": 14269952,
      "step": 22020
    },
    {
      "epoch": 12.986438679245284,
      "grad_norm": 2.858948230743408,
      "learning_rate": 3.3015087775551835e-06,
      "loss": 0.3118,
      "num_input_tokens_seen": 14272448,
      "step": 22025
    },
    {
      "epoch": 12.98938679245283,
      "grad_norm": 5.397950649261475,
      "learning_rate": 3.299089276183427e-06,
      "loss": 0.3854,
      "num_input_tokens_seen": 14274656,
      "step": 22030
    },
    {
      "epoch": 12.992334905660378,
      "grad_norm": 2.403398036956787,
      "learning_rate": 3.2966702251351157e-06,
      "loss": 0.2595,
      "num_input_tokens_seen": 14279840,
      "step": 22035
    },
    {
      "epoch": 12.995283018867925,
      "grad_norm": 6.835172176361084,
      "learning_rate": 3.2942516250507035e-06,
      "loss": 0.4053,
      "num_input_tokens_seen": 14282560,
      "step": 22040
    },
    {
      "epoch": 12.998231132075471,
      "grad_norm": 2.186286449432373,
      "learning_rate": 3.2918334765705227e-06,
      "loss": 0.2942,
      "num_input_tokens_seen": 14286528,
      "step": 22045
    },
    {
      "epoch": 13.00117924528302,
      "grad_norm": 4.584712505340576,
      "learning_rate": 3.289415780334792e-06,
      "loss": 0.3608,
      "num_input_tokens_seen": 14289000,
      "step": 22050
    },
    {
      "epoch": 13.004127358490566,
      "grad_norm": 2.2119505405426025,
      "learning_rate": 3.2869985369836067e-06,
      "loss": 0.3113,
      "num_input_tokens_seen": 14292072,
      "step": 22055
    },
    {
      "epoch": 13.007075471698114,
      "grad_norm": 3.5815820693969727,
      "learning_rate": 3.2845817471569406e-06,
      "loss": 0.3698,
      "num_input_tokens_seen": 14294952,
      "step": 22060
    },
    {
      "epoch": 13.01002358490566,
      "grad_norm": 2.6338090896606445,
      "learning_rate": 3.28216541149465e-06,
      "loss": 0.3064,
      "num_input_tokens_seen": 14298536,
      "step": 22065
    },
    {
      "epoch": 13.012971698113208,
      "grad_norm": 2.6959125995635986,
      "learning_rate": 3.2797495306364707e-06,
      "loss": 0.351,
      "num_input_tokens_seen": 14302152,
      "step": 22070
    },
    {
      "epoch": 13.015919811320755,
      "grad_norm": 6.879735946655273,
      "learning_rate": 3.2773341052220174e-06,
      "loss": 0.3195,
      "num_input_tokens_seen": 14305032,
      "step": 22075
    },
    {
      "epoch": 13.018867924528301,
      "grad_norm": 2.03348445892334,
      "learning_rate": 3.274919135890783e-06,
      "loss": 0.3011,
      "num_input_tokens_seen": 14307944,
      "step": 22080
    },
    {
      "epoch": 13.02181603773585,
      "grad_norm": 5.513121128082275,
      "learning_rate": 3.2725046232821424e-06,
      "loss": 0.3683,
      "num_input_tokens_seen": 14311368,
      "step": 22085
    },
    {
      "epoch": 13.024764150943396,
      "grad_norm": 3.0301663875579834,
      "learning_rate": 3.270090568035348e-06,
      "loss": 0.3627,
      "num_input_tokens_seen": 14313992,
      "step": 22090
    },
    {
      "epoch": 13.027712264150944,
      "grad_norm": 3.015899181365967,
      "learning_rate": 3.2676769707895306e-06,
      "loss": 0.2665,
      "num_input_tokens_seen": 14317160,
      "step": 22095
    },
    {
      "epoch": 13.03066037735849,
      "grad_norm": 4.1829047203063965,
      "learning_rate": 3.2652638321837015e-06,
      "loss": 0.2548,
      "num_input_tokens_seen": 14319784,
      "step": 22100
    },
    {
      "epoch": 13.033608490566039,
      "grad_norm": 4.0082106590271,
      "learning_rate": 3.2628511528567497e-06,
      "loss": 0.316,
      "num_input_tokens_seen": 14323048,
      "step": 22105
    },
    {
      "epoch": 13.036556603773585,
      "grad_norm": 6.1617021560668945,
      "learning_rate": 3.2604389334474407e-06,
      "loss": 0.338,
      "num_input_tokens_seen": 14325448,
      "step": 22110
    },
    {
      "epoch": 13.039504716981131,
      "grad_norm": 3.2333719730377197,
      "learning_rate": 3.2580271745944224e-06,
      "loss": 0.3259,
      "num_input_tokens_seen": 14328392,
      "step": 22115
    },
    {
      "epoch": 13.04245283018868,
      "grad_norm": 5.34116268157959,
      "learning_rate": 3.255615876936217e-06,
      "loss": 0.355,
      "num_input_tokens_seen": 14331080,
      "step": 22120
    },
    {
      "epoch": 13.045400943396226,
      "grad_norm": 3.124183177947998,
      "learning_rate": 3.2532050411112248e-06,
      "loss": 0.3111,
      "num_input_tokens_seen": 14333960,
      "step": 22125
    },
    {
      "epoch": 13.048349056603774,
      "grad_norm": 9.138395309448242,
      "learning_rate": 3.2507946677577274e-06,
      "loss": 0.3393,
      "num_input_tokens_seen": 14336232,
      "step": 22130
    },
    {
      "epoch": 13.05129716981132,
      "grad_norm": 4.844655990600586,
      "learning_rate": 3.2483847575138807e-06,
      "loss": 0.3539,
      "num_input_tokens_seen": 14340200,
      "step": 22135
    },
    {
      "epoch": 13.054245283018869,
      "grad_norm": 3.3141162395477295,
      "learning_rate": 3.245975311017716e-06,
      "loss": 0.3541,
      "num_input_tokens_seen": 14342728,
      "step": 22140
    },
    {
      "epoch": 13.057193396226415,
      "grad_norm": 3.646578550338745,
      "learning_rate": 3.2435663289071486e-06,
      "loss": 0.2779,
      "num_input_tokens_seen": 14345384,
      "step": 22145
    },
    {
      "epoch": 13.060141509433961,
      "grad_norm": 2.5266189575195312,
      "learning_rate": 3.241157811819966e-06,
      "loss": 0.3211,
      "num_input_tokens_seen": 14348680,
      "step": 22150
    },
    {
      "epoch": 13.06308962264151,
      "grad_norm": 2.9248762130737305,
      "learning_rate": 3.2387497603938327e-06,
      "loss": 0.3388,
      "num_input_tokens_seen": 14352360,
      "step": 22155
    },
    {
      "epoch": 13.066037735849056,
      "grad_norm": 3.373574733734131,
      "learning_rate": 3.2363421752662903e-06,
      "loss": 0.4164,
      "num_input_tokens_seen": 14356840,
      "step": 22160
    },
    {
      "epoch": 13.068985849056604,
      "grad_norm": 4.48626708984375,
      "learning_rate": 3.233935057074759e-06,
      "loss": 0.3025,
      "num_input_tokens_seen": 14359976,
      "step": 22165
    },
    {
      "epoch": 13.07193396226415,
      "grad_norm": 2.5882279872894287,
      "learning_rate": 3.2315284064565324e-06,
      "loss": 0.3432,
      "num_input_tokens_seen": 14362856,
      "step": 22170
    },
    {
      "epoch": 13.074882075471699,
      "grad_norm": 2.6869289875030518,
      "learning_rate": 3.2291222240487813e-06,
      "loss": 0.2544,
      "num_input_tokens_seen": 14365992,
      "step": 22175
    },
    {
      "epoch": 13.077830188679245,
      "grad_norm": 2.958498001098633,
      "learning_rate": 3.226716510488554e-06,
      "loss": 0.471,
      "num_input_tokens_seen": 14368776,
      "step": 22180
    },
    {
      "epoch": 13.080778301886792,
      "grad_norm": 3.51739501953125,
      "learning_rate": 3.224311266412773e-06,
      "loss": 0.3627,
      "num_input_tokens_seen": 14373448,
      "step": 22185
    },
    {
      "epoch": 13.08372641509434,
      "grad_norm": 3.116124391555786,
      "learning_rate": 3.2219064924582366e-06,
      "loss": 0.2525,
      "num_input_tokens_seen": 14376200,
      "step": 22190
    },
    {
      "epoch": 13.086674528301886,
      "grad_norm": 3.484769344329834,
      "learning_rate": 3.2195021892616197e-06,
      "loss": 0.2975,
      "num_input_tokens_seen": 14379432,
      "step": 22195
    },
    {
      "epoch": 13.089622641509434,
      "grad_norm": 4.391125202178955,
      "learning_rate": 3.217098357459472e-06,
      "loss": 0.2684,
      "num_input_tokens_seen": 14382152,
      "step": 22200
    },
    {
      "epoch": 13.09257075471698,
      "grad_norm": 3.353858470916748,
      "learning_rate": 3.214694997688217e-06,
      "loss": 0.3273,
      "num_input_tokens_seen": 14385192,
      "step": 22205
    },
    {
      "epoch": 13.095518867924529,
      "grad_norm": 5.470178127288818,
      "learning_rate": 3.2122921105841572e-06,
      "loss": 0.3496,
      "num_input_tokens_seen": 14388648,
      "step": 22210
    },
    {
      "epoch": 13.098466981132075,
      "grad_norm": 2.5384130477905273,
      "learning_rate": 3.2098896967834647e-06,
      "loss": 0.3378,
      "num_input_tokens_seen": 14392040,
      "step": 22215
    },
    {
      "epoch": 13.101415094339623,
      "grad_norm": 3.3287429809570312,
      "learning_rate": 3.2074877569221896e-06,
      "loss": 0.2598,
      "num_input_tokens_seen": 14394312,
      "step": 22220
    },
    {
      "epoch": 13.10436320754717,
      "grad_norm": 3.4986886978149414,
      "learning_rate": 3.205086291636257e-06,
      "loss": 0.4378,
      "num_input_tokens_seen": 14399272,
      "step": 22225
    },
    {
      "epoch": 13.107311320754716,
      "grad_norm": 5.874946117401123,
      "learning_rate": 3.202685301561463e-06,
      "loss": 0.3226,
      "num_input_tokens_seen": 14401896,
      "step": 22230
    },
    {
      "epoch": 13.110259433962264,
      "grad_norm": 1.8571678400039673,
      "learning_rate": 3.200284787333482e-06,
      "loss": 0.331,
      "num_input_tokens_seen": 14405096,
      "step": 22235
    },
    {
      "epoch": 13.11320754716981,
      "grad_norm": 2.469280242919922,
      "learning_rate": 3.1978847495878595e-06,
      "loss": 0.2661,
      "num_input_tokens_seen": 14408104,
      "step": 22240
    },
    {
      "epoch": 13.116155660377359,
      "grad_norm": 3.643972873687744,
      "learning_rate": 3.1954851889600176e-06,
      "loss": 0.2921,
      "num_input_tokens_seen": 14412296,
      "step": 22245
    },
    {
      "epoch": 13.119103773584905,
      "grad_norm": 4.128039360046387,
      "learning_rate": 3.1930861060852485e-06,
      "loss": 0.304,
      "num_input_tokens_seen": 14415176,
      "step": 22250
    },
    {
      "epoch": 13.122051886792454,
      "grad_norm": 4.840898513793945,
      "learning_rate": 3.1906875015987194e-06,
      "loss": 0.2693,
      "num_input_tokens_seen": 14417512,
      "step": 22255
    },
    {
      "epoch": 13.125,
      "grad_norm": 7.506640434265137,
      "learning_rate": 3.188289376135473e-06,
      "loss": 0.4047,
      "num_input_tokens_seen": 14420360,
      "step": 22260
    },
    {
      "epoch": 13.127948113207546,
      "grad_norm": 2.869011878967285,
      "learning_rate": 3.1858917303304213e-06,
      "loss": 0.371,
      "num_input_tokens_seen": 14423912,
      "step": 22265
    },
    {
      "epoch": 13.130896226415095,
      "grad_norm": 2.661799669265747,
      "learning_rate": 3.1834945648183535e-06,
      "loss": 0.3869,
      "num_input_tokens_seen": 14427528,
      "step": 22270
    },
    {
      "epoch": 13.133844339622641,
      "grad_norm": 3.616877317428589,
      "learning_rate": 3.1810978802339283e-06,
      "loss": 0.3838,
      "num_input_tokens_seen": 14430504,
      "step": 22275
    },
    {
      "epoch": 13.13679245283019,
      "grad_norm": 6.373474597930908,
      "learning_rate": 3.1787016772116767e-06,
      "loss": 0.4245,
      "num_input_tokens_seen": 14433480,
      "step": 22280
    },
    {
      "epoch": 13.139740566037736,
      "grad_norm": 2.9753994941711426,
      "learning_rate": 3.1763059563860073e-06,
      "loss": 0.421,
      "num_input_tokens_seen": 14439240,
      "step": 22285
    },
    {
      "epoch": 13.142688679245284,
      "grad_norm": 2.9045495986938477,
      "learning_rate": 3.1739107183911953e-06,
      "loss": 0.3266,
      "num_input_tokens_seen": 14442728,
      "step": 22290
    },
    {
      "epoch": 13.14563679245283,
      "grad_norm": 1.5633748769760132,
      "learning_rate": 3.1715159638613898e-06,
      "loss": 0.294,
      "num_input_tokens_seen": 14446568,
      "step": 22295
    },
    {
      "epoch": 13.148584905660377,
      "grad_norm": 10.079259872436523,
      "learning_rate": 3.1691216934306134e-06,
      "loss": 0.2572,
      "num_input_tokens_seen": 14450088,
      "step": 22300
    },
    {
      "epoch": 13.151533018867925,
      "grad_norm": 4.092028617858887,
      "learning_rate": 3.16672790773276e-06,
      "loss": 0.3811,
      "num_input_tokens_seen": 14454280,
      "step": 22305
    },
    {
      "epoch": 13.154481132075471,
      "grad_norm": 7.153202056884766,
      "learning_rate": 3.164334607401593e-06,
      "loss": 0.3415,
      "num_input_tokens_seen": 14457448,
      "step": 22310
    },
    {
      "epoch": 13.15742924528302,
      "grad_norm": 6.5032172203063965,
      "learning_rate": 3.1619417930707506e-06,
      "loss": 0.3316,
      "num_input_tokens_seen": 14459624,
      "step": 22315
    },
    {
      "epoch": 13.160377358490566,
      "grad_norm": 6.447333812713623,
      "learning_rate": 3.1595494653737408e-06,
      "loss": 0.3712,
      "num_input_tokens_seen": 14462152,
      "step": 22320
    },
    {
      "epoch": 13.163325471698114,
      "grad_norm": 3.3381292819976807,
      "learning_rate": 3.1571576249439408e-06,
      "loss": 0.3882,
      "num_input_tokens_seen": 14465864,
      "step": 22325
    },
    {
      "epoch": 13.16627358490566,
      "grad_norm": 2.895883321762085,
      "learning_rate": 3.1547662724146e-06,
      "loss": 0.3469,
      "num_input_tokens_seen": 14469160,
      "step": 22330
    },
    {
      "epoch": 13.169221698113208,
      "grad_norm": 4.870639324188232,
      "learning_rate": 3.1523754084188436e-06,
      "loss": 0.3856,
      "num_input_tokens_seen": 14472232,
      "step": 22335
    },
    {
      "epoch": 13.172169811320755,
      "grad_norm": 2.854771614074707,
      "learning_rate": 3.149985033589661e-06,
      "loss": 0.57,
      "num_input_tokens_seen": 14475368,
      "step": 22340
    },
    {
      "epoch": 13.175117924528301,
      "grad_norm": 3.4784491062164307,
      "learning_rate": 3.147595148559912e-06,
      "loss": 0.4832,
      "num_input_tokens_seen": 14477768,
      "step": 22345
    },
    {
      "epoch": 13.17806603773585,
      "grad_norm": 3.6239027976989746,
      "learning_rate": 3.1452057539623328e-06,
      "loss": 0.24,
      "num_input_tokens_seen": 14480456,
      "step": 22350
    },
    {
      "epoch": 13.181014150943396,
      "grad_norm": 2.2332870960235596,
      "learning_rate": 3.142816850429523e-06,
      "loss": 0.2263,
      "num_input_tokens_seen": 14484232,
      "step": 22355
    },
    {
      "epoch": 13.183962264150944,
      "grad_norm": 3.5201923847198486,
      "learning_rate": 3.1404284385939552e-06,
      "loss": 0.326,
      "num_input_tokens_seen": 14487848,
      "step": 22360
    },
    {
      "epoch": 13.18691037735849,
      "grad_norm": 2.8372788429260254,
      "learning_rate": 3.138040519087975e-06,
      "loss": 0.4295,
      "num_input_tokens_seen": 14490984,
      "step": 22365
    },
    {
      "epoch": 13.189858490566039,
      "grad_norm": 5.423201560974121,
      "learning_rate": 3.13565309254379e-06,
      "loss": 0.307,
      "num_input_tokens_seen": 14494216,
      "step": 22370
    },
    {
      "epoch": 13.192806603773585,
      "grad_norm": 9.354516983032227,
      "learning_rate": 3.1332661595934845e-06,
      "loss": 0.3491,
      "num_input_tokens_seen": 14496744,
      "step": 22375
    },
    {
      "epoch": 13.195754716981131,
      "grad_norm": 2.7961325645446777,
      "learning_rate": 3.130879720869008e-06,
      "loss": 0.3313,
      "num_input_tokens_seen": 14500616,
      "step": 22380
    },
    {
      "epoch": 13.19870283018868,
      "grad_norm": 2.6342828273773193,
      "learning_rate": 3.1284937770021815e-06,
      "loss": 0.2948,
      "num_input_tokens_seen": 14505128,
      "step": 22385
    },
    {
      "epoch": 13.201650943396226,
      "grad_norm": 3.566983699798584,
      "learning_rate": 3.1261083286246916e-06,
      "loss": 0.3772,
      "num_input_tokens_seen": 14508328,
      "step": 22390
    },
    {
      "epoch": 13.204599056603774,
      "grad_norm": 3.6471006870269775,
      "learning_rate": 3.1237233763680997e-06,
      "loss": 0.3543,
      "num_input_tokens_seen": 14511368,
      "step": 22395
    },
    {
      "epoch": 13.20754716981132,
      "grad_norm": 2.8481600284576416,
      "learning_rate": 3.1213389208638303e-06,
      "loss": 0.2888,
      "num_input_tokens_seen": 14514536,
      "step": 22400
    },
    {
      "epoch": 13.210495283018869,
      "grad_norm": 2.4535484313964844,
      "learning_rate": 3.1189549627431757e-06,
      "loss": 0.3969,
      "num_input_tokens_seen": 14519176,
      "step": 22405
    },
    {
      "epoch": 13.213443396226415,
      "grad_norm": 4.309487819671631,
      "learning_rate": 3.116571502637304e-06,
      "loss": 0.3653,
      "num_input_tokens_seen": 14522120,
      "step": 22410
    },
    {
      "epoch": 13.216391509433961,
      "grad_norm": 3.3491933345794678,
      "learning_rate": 3.1141885411772434e-06,
      "loss": 0.2833,
      "num_input_tokens_seen": 14525032,
      "step": 22415
    },
    {
      "epoch": 13.21933962264151,
      "grad_norm": 4.455045700073242,
      "learning_rate": 3.111806078993893e-06,
      "loss": 0.3373,
      "num_input_tokens_seen": 14528392,
      "step": 22420
    },
    {
      "epoch": 13.222287735849056,
      "grad_norm": 2.8218331336975098,
      "learning_rate": 3.1094241167180223e-06,
      "loss": 0.3185,
      "num_input_tokens_seen": 14531432,
      "step": 22425
    },
    {
      "epoch": 13.225235849056604,
      "grad_norm": 4.556827545166016,
      "learning_rate": 3.1070426549802623e-06,
      "loss": 0.3702,
      "num_input_tokens_seen": 14534824,
      "step": 22430
    },
    {
      "epoch": 13.22818396226415,
      "grad_norm": 4.7786865234375,
      "learning_rate": 3.1046616944111196e-06,
      "loss": 0.3395,
      "num_input_tokens_seen": 14537384,
      "step": 22435
    },
    {
      "epoch": 13.231132075471699,
      "grad_norm": 8.673052787780762,
      "learning_rate": 3.1022812356409606e-06,
      "loss": 0.2441,
      "num_input_tokens_seen": 14542888,
      "step": 22440
    },
    {
      "epoch": 13.234080188679245,
      "grad_norm": 4.209970951080322,
      "learning_rate": 3.0999012793000244e-06,
      "loss": 0.4013,
      "num_input_tokens_seen": 14546696,
      "step": 22445
    },
    {
      "epoch": 13.237028301886792,
      "grad_norm": 5.0148749351501465,
      "learning_rate": 3.097521826018414e-06,
      "loss": 0.4285,
      "num_input_tokens_seen": 14548968,
      "step": 22450
    },
    {
      "epoch": 13.23997641509434,
      "grad_norm": 4.22088098526001,
      "learning_rate": 3.0951428764260973e-06,
      "loss": 0.2819,
      "num_input_tokens_seen": 14552232,
      "step": 22455
    },
    {
      "epoch": 13.242924528301886,
      "grad_norm": 2.843395948410034,
      "learning_rate": 3.092764431152915e-06,
      "loss": 0.3082,
      "num_input_tokens_seen": 14555336,
      "step": 22460
    },
    {
      "epoch": 13.245872641509434,
      "grad_norm": 2.036524772644043,
      "learning_rate": 3.0903864908285693e-06,
      "loss": 0.369,
      "num_input_tokens_seen": 14558472,
      "step": 22465
    },
    {
      "epoch": 13.24882075471698,
      "grad_norm": 4.879626750946045,
      "learning_rate": 3.088009056082629e-06,
      "loss": 0.3362,
      "num_input_tokens_seen": 14561480,
      "step": 22470
    },
    {
      "epoch": 13.251768867924529,
      "grad_norm": 3.78415846824646,
      "learning_rate": 3.0856321275445324e-06,
      "loss": 0.3506,
      "num_input_tokens_seen": 14565128,
      "step": 22475
    },
    {
      "epoch": 13.254716981132075,
      "grad_norm": 3.2688605785369873,
      "learning_rate": 3.0832557058435808e-06,
      "loss": 0.2888,
      "num_input_tokens_seen": 14568456,
      "step": 22480
    },
    {
      "epoch": 13.257665094339623,
      "grad_norm": 3.0079565048217773,
      "learning_rate": 3.0808797916089405e-06,
      "loss": 0.2359,
      "num_input_tokens_seen": 14571976,
      "step": 22485
    },
    {
      "epoch": 13.26061320754717,
      "grad_norm": 5.0469069480896,
      "learning_rate": 3.078504385469647e-06,
      "loss": 0.3692,
      "num_input_tokens_seen": 14575624,
      "step": 22490
    },
    {
      "epoch": 13.263561320754716,
      "grad_norm": 2.8358561992645264,
      "learning_rate": 3.076129488054599e-06,
      "loss": 0.2544,
      "num_input_tokens_seen": 14578504,
      "step": 22495
    },
    {
      "epoch": 13.266509433962264,
      "grad_norm": 2.712393045425415,
      "learning_rate": 3.0737550999925604e-06,
      "loss": 0.4303,
      "num_input_tokens_seen": 14581672,
      "step": 22500
    },
    {
      "epoch": 13.26945754716981,
      "grad_norm": 2.594625949859619,
      "learning_rate": 3.0713812219121604e-06,
      "loss": 0.3794,
      "num_input_tokens_seen": 14584712,
      "step": 22505
    },
    {
      "epoch": 13.272405660377359,
      "grad_norm": 5.980940341949463,
      "learning_rate": 3.0690078544418934e-06,
      "loss": 0.3981,
      "num_input_tokens_seen": 14587336,
      "step": 22510
    },
    {
      "epoch": 13.275353773584905,
      "grad_norm": 6.120868682861328,
      "learning_rate": 3.0666349982101198e-06,
      "loss": 0.3587,
      "num_input_tokens_seen": 14590120,
      "step": 22515
    },
    {
      "epoch": 13.278301886792454,
      "grad_norm": 3.9062137603759766,
      "learning_rate": 3.0642626538450627e-06,
      "loss": 0.2848,
      "num_input_tokens_seen": 14593000,
      "step": 22520
    },
    {
      "epoch": 13.28125,
      "grad_norm": 3.614949941635132,
      "learning_rate": 3.061890821974809e-06,
      "loss": 0.3286,
      "num_input_tokens_seen": 14596040,
      "step": 22525
    },
    {
      "epoch": 13.284198113207546,
      "grad_norm": 6.142546653747559,
      "learning_rate": 3.059519503227313e-06,
      "loss": 0.4712,
      "num_input_tokens_seen": 14599400,
      "step": 22530
    },
    {
      "epoch": 13.287146226415095,
      "grad_norm": 4.470977783203125,
      "learning_rate": 3.057148698230393e-06,
      "loss": 0.2637,
      "num_input_tokens_seen": 14603656,
      "step": 22535
    },
    {
      "epoch": 13.290094339622641,
      "grad_norm": 3.0512924194335938,
      "learning_rate": 3.0547784076117294e-06,
      "loss": 0.3615,
      "num_input_tokens_seen": 14606632,
      "step": 22540
    },
    {
      "epoch": 13.29304245283019,
      "grad_norm": 4.6200971603393555,
      "learning_rate": 3.0524086319988635e-06,
      "loss": 0.3234,
      "num_input_tokens_seen": 14609384,
      "step": 22545
    },
    {
      "epoch": 13.295990566037736,
      "grad_norm": 5.846859455108643,
      "learning_rate": 3.0500393720192074e-06,
      "loss": 0.2761,
      "num_input_tokens_seen": 14612040,
      "step": 22550
    },
    {
      "epoch": 13.298938679245284,
      "grad_norm": 1.7637194395065308,
      "learning_rate": 3.047670628300031e-06,
      "loss": 0.3498,
      "num_input_tokens_seen": 14615176,
      "step": 22555
    },
    {
      "epoch": 13.30188679245283,
      "grad_norm": 1.9200772047042847,
      "learning_rate": 3.0453024014684694e-06,
      "loss": 0.2705,
      "num_input_tokens_seen": 14619656,
      "step": 22560
    },
    {
      "epoch": 13.304834905660377,
      "grad_norm": 4.927657127380371,
      "learning_rate": 3.0429346921515225e-06,
      "loss": 0.2997,
      "num_input_tokens_seen": 14622856,
      "step": 22565
    },
    {
      "epoch": 13.307783018867925,
      "grad_norm": 3.0339417457580566,
      "learning_rate": 3.04056750097605e-06,
      "loss": 0.3049,
      "num_input_tokens_seen": 14625448,
      "step": 22570
    },
    {
      "epoch": 13.310731132075471,
      "grad_norm": 6.12006139755249,
      "learning_rate": 3.0382008285687754e-06,
      "loss": 0.2328,
      "num_input_tokens_seen": 14628808,
      "step": 22575
    },
    {
      "epoch": 13.31367924528302,
      "grad_norm": 2.127723455429077,
      "learning_rate": 3.035834675556287e-06,
      "loss": 0.3321,
      "num_input_tokens_seen": 14632072,
      "step": 22580
    },
    {
      "epoch": 13.316627358490566,
      "grad_norm": 3.22300124168396,
      "learning_rate": 3.0334690425650336e-06,
      "loss": 0.2774,
      "num_input_tokens_seen": 14635336,
      "step": 22585
    },
    {
      "epoch": 13.319575471698114,
      "grad_norm": 2.9933390617370605,
      "learning_rate": 3.031103930221325e-06,
      "loss": 0.3141,
      "num_input_tokens_seen": 14638504,
      "step": 22590
    },
    {
      "epoch": 13.32252358490566,
      "grad_norm": 2.5112521648406982,
      "learning_rate": 3.028739339151338e-06,
      "loss": 0.2638,
      "num_input_tokens_seen": 14642920,
      "step": 22595
    },
    {
      "epoch": 13.325471698113208,
      "grad_norm": 3.2177820205688477,
      "learning_rate": 3.0263752699811067e-06,
      "loss": 0.331,
      "num_input_tokens_seen": 14646440,
      "step": 22600
    },
    {
      "epoch": 13.328419811320755,
      "grad_norm": 2.1074111461639404,
      "learning_rate": 3.0240117233365267e-06,
      "loss": 0.2074,
      "num_input_tokens_seen": 14650216,
      "step": 22605
    },
    {
      "epoch": 13.331367924528301,
      "grad_norm": 5.652377128601074,
      "learning_rate": 3.0216486998433604e-06,
      "loss": 0.2964,
      "num_input_tokens_seen": 14653096,
      "step": 22610
    },
    {
      "epoch": 13.33431603773585,
      "grad_norm": 4.892590045928955,
      "learning_rate": 3.0192862001272273e-06,
      "loss": 0.4175,
      "num_input_tokens_seen": 14656392,
      "step": 22615
    },
    {
      "epoch": 13.337264150943396,
      "grad_norm": 2.714963674545288,
      "learning_rate": 3.0169242248136066e-06,
      "loss": 0.3614,
      "num_input_tokens_seen": 14659560,
      "step": 22620
    },
    {
      "epoch": 13.340212264150944,
      "grad_norm": 4.28176736831665,
      "learning_rate": 3.0145627745278457e-06,
      "loss": 0.4332,
      "num_input_tokens_seen": 14662472,
      "step": 22625
    },
    {
      "epoch": 13.34316037735849,
      "grad_norm": 6.399781703948975,
      "learning_rate": 3.0122018498951478e-06,
      "loss": 0.3727,
      "num_input_tokens_seen": 14664968,
      "step": 22630
    },
    {
      "epoch": 13.346108490566039,
      "grad_norm": 3.0563342571258545,
      "learning_rate": 3.0098414515405765e-06,
      "loss": 0.4043,
      "num_input_tokens_seen": 14668136,
      "step": 22635
    },
    {
      "epoch": 13.349056603773585,
      "grad_norm": 4.5387864112854,
      "learning_rate": 3.0074815800890576e-06,
      "loss": 0.2692,
      "num_input_tokens_seen": 14671816,
      "step": 22640
    },
    {
      "epoch": 13.352004716981131,
      "grad_norm": 3.261756420135498,
      "learning_rate": 3.005122236165378e-06,
      "loss": 0.289,
      "num_input_tokens_seen": 14675208,
      "step": 22645
    },
    {
      "epoch": 13.35495283018868,
      "grad_norm": 3.482823133468628,
      "learning_rate": 3.0027634203941847e-06,
      "loss": 0.3144,
      "num_input_tokens_seen": 14678600,
      "step": 22650
    },
    {
      "epoch": 13.357900943396226,
      "grad_norm": 1.8591809272766113,
      "learning_rate": 3.0004051333999816e-06,
      "loss": 0.3984,
      "num_input_tokens_seen": 14682280,
      "step": 22655
    },
    {
      "epoch": 13.360849056603774,
      "grad_norm": 6.289658069610596,
      "learning_rate": 2.998047375807139e-06,
      "loss": 0.292,
      "num_input_tokens_seen": 14685672,
      "step": 22660
    },
    {
      "epoch": 13.36379716981132,
      "grad_norm": 10.581445693969727,
      "learning_rate": 2.995690148239881e-06,
      "loss": 0.3076,
      "num_input_tokens_seen": 14687880,
      "step": 22665
    },
    {
      "epoch": 13.366745283018869,
      "grad_norm": 3.641179323196411,
      "learning_rate": 2.993333451322293e-06,
      "loss": 0.2617,
      "num_input_tokens_seen": 14691080,
      "step": 22670
    },
    {
      "epoch": 13.369693396226415,
      "grad_norm": 3.5359225273132324,
      "learning_rate": 2.9909772856783242e-06,
      "loss": 0.2284,
      "num_input_tokens_seen": 14694792,
      "step": 22675
    },
    {
      "epoch": 13.372641509433961,
      "grad_norm": 2.672410726547241,
      "learning_rate": 2.988621651931777e-06,
      "loss": 0.3077,
      "num_input_tokens_seen": 14698216,
      "step": 22680
    },
    {
      "epoch": 13.37558962264151,
      "grad_norm": 3.463399887084961,
      "learning_rate": 2.986266550706315e-06,
      "loss": 0.3264,
      "num_input_tokens_seen": 14702088,
      "step": 22685
    },
    {
      "epoch": 13.378537735849056,
      "grad_norm": 1.7056310176849365,
      "learning_rate": 2.9839119826254627e-06,
      "loss": 0.2793,
      "num_input_tokens_seen": 14705448,
      "step": 22690
    },
    {
      "epoch": 13.381485849056604,
      "grad_norm": 5.763820648193359,
      "learning_rate": 2.981557948312602e-06,
      "loss": 0.4087,
      "num_input_tokens_seen": 14709832,
      "step": 22695
    },
    {
      "epoch": 13.38443396226415,
      "grad_norm": 4.25277042388916,
      "learning_rate": 2.9792044483909733e-06,
      "loss": 0.3302,
      "num_input_tokens_seen": 14712552,
      "step": 22700
    },
    {
      "epoch": 13.387382075471699,
      "grad_norm": 4.654058933258057,
      "learning_rate": 2.9768514834836767e-06,
      "loss": 0.3221,
      "num_input_tokens_seen": 14716264,
      "step": 22705
    },
    {
      "epoch": 13.390330188679245,
      "grad_norm": 2.9162657260894775,
      "learning_rate": 2.9744990542136685e-06,
      "loss": 0.3368,
      "num_input_tokens_seen": 14718952,
      "step": 22710
    },
    {
      "epoch": 13.393278301886792,
      "grad_norm": 3.3255317211151123,
      "learning_rate": 2.9721471612037637e-06,
      "loss": 0.2924,
      "num_input_tokens_seen": 14722248,
      "step": 22715
    },
    {
      "epoch": 13.39622641509434,
      "grad_norm": 2.5401339530944824,
      "learning_rate": 2.9697958050766385e-06,
      "loss": 0.4042,
      "num_input_tokens_seen": 14726760,
      "step": 22720
    },
    {
      "epoch": 13.399174528301886,
      "grad_norm": 4.064121246337891,
      "learning_rate": 2.967444986454825e-06,
      "loss": 0.3328,
      "num_input_tokens_seen": 14730408,
      "step": 22725
    },
    {
      "epoch": 13.402122641509434,
      "grad_norm": 2.496427297592163,
      "learning_rate": 2.9650947059607106e-06,
      "loss": 0.28,
      "num_input_tokens_seen": 14733864,
      "step": 22730
    },
    {
      "epoch": 13.40507075471698,
      "grad_norm": 3.6112349033355713,
      "learning_rate": 2.962744964216542e-06,
      "loss": 0.3332,
      "num_input_tokens_seen": 14736840,
      "step": 22735
    },
    {
      "epoch": 13.408018867924529,
      "grad_norm": 4.885488510131836,
      "learning_rate": 2.960395761844425e-06,
      "loss": 0.2732,
      "num_input_tokens_seen": 14739624,
      "step": 22740
    },
    {
      "epoch": 13.410966981132075,
      "grad_norm": 6.689154148101807,
      "learning_rate": 2.95804709946632e-06,
      "loss": 0.3538,
      "num_input_tokens_seen": 14742792,
      "step": 22745
    },
    {
      "epoch": 13.413915094339623,
      "grad_norm": 2.9753081798553467,
      "learning_rate": 2.9556989777040457e-06,
      "loss": 0.4561,
      "num_input_tokens_seen": 14746696,
      "step": 22750
    },
    {
      "epoch": 13.41686320754717,
      "grad_norm": 5.002169609069824,
      "learning_rate": 2.9533513971792776e-06,
      "loss": 0.4754,
      "num_input_tokens_seen": 14749864,
      "step": 22755
    },
    {
      "epoch": 13.419811320754716,
      "grad_norm": 4.472928524017334,
      "learning_rate": 2.9510043585135473e-06,
      "loss": 0.3684,
      "num_input_tokens_seen": 14752936,
      "step": 22760
    },
    {
      "epoch": 13.422759433962264,
      "grad_norm": 2.7680866718292236,
      "learning_rate": 2.948657862328244e-06,
      "loss": 0.3006,
      "num_input_tokens_seen": 14756104,
      "step": 22765
    },
    {
      "epoch": 13.42570754716981,
      "grad_norm": 3.3271522521972656,
      "learning_rate": 2.946311909244613e-06,
      "loss": 0.4325,
      "num_input_tokens_seen": 14758760,
      "step": 22770
    },
    {
      "epoch": 13.428655660377359,
      "grad_norm": 1.801930546760559,
      "learning_rate": 2.9439664998837538e-06,
      "loss": 0.2861,
      "num_input_tokens_seen": 14761576,
      "step": 22775
    },
    {
      "epoch": 13.431603773584905,
      "grad_norm": 4.353911399841309,
      "learning_rate": 2.941621634866626e-06,
      "loss": 0.3497,
      "num_input_tokens_seen": 14764360,
      "step": 22780
    },
    {
      "epoch": 13.434551886792454,
      "grad_norm": 2.6987531185150146,
      "learning_rate": 2.9392773148140406e-06,
      "loss": 0.4771,
      "num_input_tokens_seen": 14768648,
      "step": 22785
    },
    {
      "epoch": 13.4375,
      "grad_norm": 2.2993714809417725,
      "learning_rate": 2.9369335403466676e-06,
      "loss": 0.3608,
      "num_input_tokens_seen": 14771176,
      "step": 22790
    },
    {
      "epoch": 13.440448113207546,
      "grad_norm": 4.593253135681152,
      "learning_rate": 2.9345903120850318e-06,
      "loss": 0.4086,
      "num_input_tokens_seen": 14774312,
      "step": 22795
    },
    {
      "epoch": 13.443396226415095,
      "grad_norm": 2.71877384185791,
      "learning_rate": 2.932247630649512e-06,
      "loss": 0.5348,
      "num_input_tokens_seen": 14778440,
      "step": 22800
    },
    {
      "epoch": 13.446344339622641,
      "grad_norm": 3.4359543323516846,
      "learning_rate": 2.9299054966603424e-06,
      "loss": 0.3216,
      "num_input_tokens_seen": 14781576,
      "step": 22805
    },
    {
      "epoch": 13.44929245283019,
      "grad_norm": 2.5798587799072266,
      "learning_rate": 2.927563910737613e-06,
      "loss": 0.333,
      "num_input_tokens_seen": 14784680,
      "step": 22810
    },
    {
      "epoch": 13.452240566037736,
      "grad_norm": 2.581653118133545,
      "learning_rate": 2.9252228735012722e-06,
      "loss": 0.3368,
      "num_input_tokens_seen": 14788360,
      "step": 22815
    },
    {
      "epoch": 13.455188679245284,
      "grad_norm": 3.938732147216797,
      "learning_rate": 2.9228823855711174e-06,
      "loss": 0.404,
      "num_input_tokens_seen": 14791112,
      "step": 22820
    },
    {
      "epoch": 13.45813679245283,
      "grad_norm": 4.429026126861572,
      "learning_rate": 2.920542447566802e-06,
      "loss": 0.3579,
      "num_input_tokens_seen": 14795208,
      "step": 22825
    },
    {
      "epoch": 13.461084905660377,
      "grad_norm": 2.562857151031494,
      "learning_rate": 2.918203060107837e-06,
      "loss": 0.2834,
      "num_input_tokens_seen": 14797928,
      "step": 22830
    },
    {
      "epoch": 13.464033018867925,
      "grad_norm": 1.7716660499572754,
      "learning_rate": 2.9158642238135813e-06,
      "loss": 0.3414,
      "num_input_tokens_seen": 14801704,
      "step": 22835
    },
    {
      "epoch": 13.466981132075471,
      "grad_norm": 2.798265218734741,
      "learning_rate": 2.913525939303257e-06,
      "loss": 0.2725,
      "num_input_tokens_seen": 14804424,
      "step": 22840
    },
    {
      "epoch": 13.46992924528302,
      "grad_norm": 7.159753799438477,
      "learning_rate": 2.9111882071959317e-06,
      "loss": 0.3759,
      "num_input_tokens_seen": 14808072,
      "step": 22845
    },
    {
      "epoch": 13.472877358490566,
      "grad_norm": 3.0978610515594482,
      "learning_rate": 2.908851028110532e-06,
      "loss": 0.4691,
      "num_input_tokens_seen": 14811656,
      "step": 22850
    },
    {
      "epoch": 13.475825471698114,
      "grad_norm": 3.181598663330078,
      "learning_rate": 2.906514402665834e-06,
      "loss": 0.2588,
      "num_input_tokens_seen": 14814760,
      "step": 22855
    },
    {
      "epoch": 13.47877358490566,
      "grad_norm": 3.5119080543518066,
      "learning_rate": 2.9041783314804705e-06,
      "loss": 0.3372,
      "num_input_tokens_seen": 14817640,
      "step": 22860
    },
    {
      "epoch": 13.481721698113208,
      "grad_norm": 3.1587162017822266,
      "learning_rate": 2.9018428151729238e-06,
      "loss": 0.2702,
      "num_input_tokens_seen": 14820616,
      "step": 22865
    },
    {
      "epoch": 13.484669811320755,
      "grad_norm": 3.0406463146209717,
      "learning_rate": 2.899507854361537e-06,
      "loss": 0.4556,
      "num_input_tokens_seen": 14823368,
      "step": 22870
    },
    {
      "epoch": 13.487617924528301,
      "grad_norm": 4.648748874664307,
      "learning_rate": 2.8971734496644975e-06,
      "loss": 0.3097,
      "num_input_tokens_seen": 14826920,
      "step": 22875
    },
    {
      "epoch": 13.49056603773585,
      "grad_norm": 6.005543231964111,
      "learning_rate": 2.894839601699851e-06,
      "loss": 0.4486,
      "num_input_tokens_seen": 14829864,
      "step": 22880
    },
    {
      "epoch": 13.493514150943396,
      "grad_norm": 4.179430961608887,
      "learning_rate": 2.8925063110854923e-06,
      "loss": 0.2784,
      "num_input_tokens_seen": 14832968,
      "step": 22885
    },
    {
      "epoch": 13.496462264150944,
      "grad_norm": 2.7040319442749023,
      "learning_rate": 2.8901735784391683e-06,
      "loss": 0.3316,
      "num_input_tokens_seen": 14835688,
      "step": 22890
    },
    {
      "epoch": 13.49941037735849,
      "grad_norm": 2.8914451599121094,
      "learning_rate": 2.8878414043784844e-06,
      "loss": 0.2697,
      "num_input_tokens_seen": 14839400,
      "step": 22895
    },
    {
      "epoch": 13.502358490566039,
      "grad_norm": 2.463181734085083,
      "learning_rate": 2.885509789520891e-06,
      "loss": 0.2793,
      "num_input_tokens_seen": 14842632,
      "step": 22900
    },
    {
      "epoch": 13.505306603773585,
      "grad_norm": 4.801540851593018,
      "learning_rate": 2.8831787344836926e-06,
      "loss": 0.3003,
      "num_input_tokens_seen": 14846056,
      "step": 22905
    },
    {
      "epoch": 13.508254716981131,
      "grad_norm": 2.567183256149292,
      "learning_rate": 2.880848239884049e-06,
      "loss": 0.3902,
      "num_input_tokens_seen": 14849512,
      "step": 22910
    },
    {
      "epoch": 13.51120283018868,
      "grad_norm": 3.1962509155273438,
      "learning_rate": 2.8785183063389667e-06,
      "loss": 0.3065,
      "num_input_tokens_seen": 14851976,
      "step": 22915
    },
    {
      "epoch": 13.514150943396226,
      "grad_norm": 4.378251075744629,
      "learning_rate": 2.876188934465306e-06,
      "loss": 0.2945,
      "num_input_tokens_seen": 14854600,
      "step": 22920
    },
    {
      "epoch": 13.517099056603774,
      "grad_norm": 2.893841028213501,
      "learning_rate": 2.8738601248797758e-06,
      "loss": 0.4211,
      "num_input_tokens_seen": 14858600,
      "step": 22925
    },
    {
      "epoch": 13.52004716981132,
      "grad_norm": 5.005002021789551,
      "learning_rate": 2.8715318781989432e-06,
      "loss": 0.2633,
      "num_input_tokens_seen": 14861864,
      "step": 22930
    },
    {
      "epoch": 13.522995283018869,
      "grad_norm": 4.6457061767578125,
      "learning_rate": 2.869204195039219e-06,
      "loss": 0.3399,
      "num_input_tokens_seen": 14865032,
      "step": 22935
    },
    {
      "epoch": 13.525943396226415,
      "grad_norm": 3.416001319885254,
      "learning_rate": 2.8668770760168673e-06,
      "loss": 0.2245,
      "num_input_tokens_seen": 14868040,
      "step": 22940
    },
    {
      "epoch": 13.528891509433961,
      "grad_norm": 3.13700008392334,
      "learning_rate": 2.864550521748003e-06,
      "loss": 0.3613,
      "num_input_tokens_seen": 14871208,
      "step": 22945
    },
    {
      "epoch": 13.53183962264151,
      "grad_norm": 2.7076315879821777,
      "learning_rate": 2.862224532848591e-06,
      "loss": 0.2612,
      "num_input_tokens_seen": 14874024,
      "step": 22950
    },
    {
      "epoch": 13.534787735849056,
      "grad_norm": 3.5947346687316895,
      "learning_rate": 2.8598991099344455e-06,
      "loss": 0.3409,
      "num_input_tokens_seen": 14877256,
      "step": 22955
    },
    {
      "epoch": 13.537735849056604,
      "grad_norm": 4.408897399902344,
      "learning_rate": 2.857574253621236e-06,
      "loss": 0.3262,
      "num_input_tokens_seen": 14881000,
      "step": 22960
    },
    {
      "epoch": 13.54068396226415,
      "grad_norm": 7.712809085845947,
      "learning_rate": 2.855249964524476e-06,
      "loss": 0.4238,
      "num_input_tokens_seen": 14884200,
      "step": 22965
    },
    {
      "epoch": 13.543632075471699,
      "grad_norm": 3.241457939147949,
      "learning_rate": 2.852926243259531e-06,
      "loss": 0.2484,
      "num_input_tokens_seen": 14886920,
      "step": 22970
    },
    {
      "epoch": 13.546580188679245,
      "grad_norm": 3.3857650756835938,
      "learning_rate": 2.850603090441617e-06,
      "loss": 0.4598,
      "num_input_tokens_seen": 14889960,
      "step": 22975
    },
    {
      "epoch": 13.549528301886792,
      "grad_norm": 3.589874267578125,
      "learning_rate": 2.848280506685798e-06,
      "loss": 0.3369,
      "num_input_tokens_seen": 14892488,
      "step": 22980
    },
    {
      "epoch": 13.55247641509434,
      "grad_norm": 5.425849914550781,
      "learning_rate": 2.845958492606986e-06,
      "loss": 0.4072,
      "num_input_tokens_seen": 14895208,
      "step": 22985
    },
    {
      "epoch": 13.555424528301886,
      "grad_norm": 2.1893508434295654,
      "learning_rate": 2.843637048819949e-06,
      "loss": 0.2795,
      "num_input_tokens_seen": 14898344,
      "step": 22990
    },
    {
      "epoch": 13.558372641509434,
      "grad_norm": 2.2990057468414307,
      "learning_rate": 2.8413161759392966e-06,
      "loss": 0.4037,
      "num_input_tokens_seen": 14901832,
      "step": 22995
    },
    {
      "epoch": 13.56132075471698,
      "grad_norm": 3.964258909225464,
      "learning_rate": 2.8389958745794878e-06,
      "loss": 0.3876,
      "num_input_tokens_seen": 14905416,
      "step": 23000
    },
    {
      "epoch": 13.564268867924529,
      "grad_norm": 3.690164804458618,
      "learning_rate": 2.8366761453548366e-06,
      "loss": 0.3415,
      "num_input_tokens_seen": 14909288,
      "step": 23005
    },
    {
      "epoch": 13.567216981132075,
      "grad_norm": 4.505602836608887,
      "learning_rate": 2.8343569888795e-06,
      "loss": 0.3885,
      "num_input_tokens_seen": 14912456,
      "step": 23010
    },
    {
      "epoch": 13.570165094339622,
      "grad_norm": 17.60578727722168,
      "learning_rate": 2.832038405767483e-06,
      "loss": 0.6881,
      "num_input_tokens_seen": 14916360,
      "step": 23015
    },
    {
      "epoch": 13.57311320754717,
      "grad_norm": 4.522889614105225,
      "learning_rate": 2.8297203966326397e-06,
      "loss": 0.2905,
      "num_input_tokens_seen": 14918984,
      "step": 23020
    },
    {
      "epoch": 13.576061320754716,
      "grad_norm": 3.6183762550354004,
      "learning_rate": 2.8274029620886773e-06,
      "loss": 0.2595,
      "num_input_tokens_seen": 14923464,
      "step": 23025
    },
    {
      "epoch": 13.579009433962264,
      "grad_norm": 2.5080015659332275,
      "learning_rate": 2.825086102749144e-06,
      "loss": 0.3632,
      "num_input_tokens_seen": 14926632,
      "step": 23030
    },
    {
      "epoch": 13.58195754716981,
      "grad_norm": 3.4343154430389404,
      "learning_rate": 2.822769819227438e-06,
      "loss": 0.2751,
      "num_input_tokens_seen": 14929768,
      "step": 23035
    },
    {
      "epoch": 13.584905660377359,
      "grad_norm": 5.185078144073486,
      "learning_rate": 2.8204541121368055e-06,
      "loss": 0.33,
      "num_input_tokens_seen": 14933032,
      "step": 23040
    },
    {
      "epoch": 13.587853773584905,
      "grad_norm": 1.8772265911102295,
      "learning_rate": 2.8181389820903402e-06,
      "loss": 0.3865,
      "num_input_tokens_seen": 14936328,
      "step": 23045
    },
    {
      "epoch": 13.590801886792454,
      "grad_norm": 2.8407695293426514,
      "learning_rate": 2.8158244297009814e-06,
      "loss": 0.5547,
      "num_input_tokens_seen": 14942952,
      "step": 23050
    },
    {
      "epoch": 13.59375,
      "grad_norm": 1.7575010061264038,
      "learning_rate": 2.8135104555815196e-06,
      "loss": 0.2467,
      "num_input_tokens_seen": 14946760,
      "step": 23055
    },
    {
      "epoch": 13.596698113207546,
      "grad_norm": 3.291003942489624,
      "learning_rate": 2.811197060344588e-06,
      "loss": 0.3267,
      "num_input_tokens_seen": 14949224,
      "step": 23060
    },
    {
      "epoch": 13.599646226415095,
      "grad_norm": 3.888955593109131,
      "learning_rate": 2.8088842446026677e-06,
      "loss": 0.265,
      "num_input_tokens_seen": 14951624,
      "step": 23065
    },
    {
      "epoch": 13.602594339622641,
      "grad_norm": 2.881679058074951,
      "learning_rate": 2.806572008968087e-06,
      "loss": 0.276,
      "num_input_tokens_seen": 14954856,
      "step": 23070
    },
    {
      "epoch": 13.60554245283019,
      "grad_norm": 3.433443546295166,
      "learning_rate": 2.80426035405302e-06,
      "loss": 0.3728,
      "num_input_tokens_seen": 14957352,
      "step": 23075
    },
    {
      "epoch": 13.608490566037736,
      "grad_norm": 5.286664009094238,
      "learning_rate": 2.8019492804694852e-06,
      "loss": 0.2908,
      "num_input_tokens_seen": 14960168,
      "step": 23080
    },
    {
      "epoch": 13.611438679245284,
      "grad_norm": 3.062563896179199,
      "learning_rate": 2.799638788829354e-06,
      "loss": 0.2557,
      "num_input_tokens_seen": 14962824,
      "step": 23085
    },
    {
      "epoch": 13.61438679245283,
      "grad_norm": 3.17177677154541,
      "learning_rate": 2.7973288797443367e-06,
      "loss": 0.2691,
      "num_input_tokens_seen": 14965608,
      "step": 23090
    },
    {
      "epoch": 13.617334905660378,
      "grad_norm": 3.183286428451538,
      "learning_rate": 2.7950195538259884e-06,
      "loss": 0.2965,
      "num_input_tokens_seen": 14968456,
      "step": 23095
    },
    {
      "epoch": 13.620283018867925,
      "grad_norm": 3.7099404335021973,
      "learning_rate": 2.792710811685719e-06,
      "loss": 0.3534,
      "num_input_tokens_seen": 14971656,
      "step": 23100
    },
    {
      "epoch": 13.623231132075471,
      "grad_norm": 3.6429390907287598,
      "learning_rate": 2.7904026539347743e-06,
      "loss": 0.382,
      "num_input_tokens_seen": 14975144,
      "step": 23105
    },
    {
      "epoch": 13.62617924528302,
      "grad_norm": 4.307990074157715,
      "learning_rate": 2.7880950811842507e-06,
      "loss": 0.3717,
      "num_input_tokens_seen": 14977928,
      "step": 23110
    },
    {
      "epoch": 13.629127358490566,
      "grad_norm": 3.890187978744507,
      "learning_rate": 2.785788094045085e-06,
      "loss": 0.3165,
      "num_input_tokens_seen": 14980744,
      "step": 23115
    },
    {
      "epoch": 13.632075471698114,
      "grad_norm": 3.898811101913452,
      "learning_rate": 2.7834816931280655e-06,
      "loss": 0.4388,
      "num_input_tokens_seen": 14984072,
      "step": 23120
    },
    {
      "epoch": 13.63502358490566,
      "grad_norm": 4.068681716918945,
      "learning_rate": 2.781175879043821e-06,
      "loss": 0.3225,
      "num_input_tokens_seen": 14986824,
      "step": 23125
    },
    {
      "epoch": 13.637971698113208,
      "grad_norm": 3.460627317428589,
      "learning_rate": 2.778870652402825e-06,
      "loss": 0.3576,
      "num_input_tokens_seen": 14990120,
      "step": 23130
    },
    {
      "epoch": 13.640919811320755,
      "grad_norm": 2.3787569999694824,
      "learning_rate": 2.776566013815396e-06,
      "loss": 0.4387,
      "num_input_tokens_seen": 14993736,
      "step": 23135
    },
    {
      "epoch": 13.643867924528301,
      "grad_norm": 5.866184711456299,
      "learning_rate": 2.774261963891698e-06,
      "loss": 0.2765,
      "num_input_tokens_seen": 14997896,
      "step": 23140
    },
    {
      "epoch": 13.64681603773585,
      "grad_norm": 3.163853406906128,
      "learning_rate": 2.771958503241735e-06,
      "loss": 0.2582,
      "num_input_tokens_seen": 15000712,
      "step": 23145
    },
    {
      "epoch": 13.649764150943396,
      "grad_norm": 8.186193466186523,
      "learning_rate": 2.769655632475362e-06,
      "loss": 0.3905,
      "num_input_tokens_seen": 15003784,
      "step": 23150
    },
    {
      "epoch": 13.652712264150944,
      "grad_norm": 3.650017023086548,
      "learning_rate": 2.7673533522022733e-06,
      "loss": 0.3987,
      "num_input_tokens_seen": 15006632,
      "step": 23155
    },
    {
      "epoch": 13.65566037735849,
      "grad_norm": 2.4600830078125,
      "learning_rate": 2.765051663032007e-06,
      "loss": 0.299,
      "num_input_tokens_seen": 15009640,
      "step": 23160
    },
    {
      "epoch": 13.658608490566039,
      "grad_norm": 3.4269964694976807,
      "learning_rate": 2.7627505655739446e-06,
      "loss": 0.3679,
      "num_input_tokens_seen": 15013352,
      "step": 23165
    },
    {
      "epoch": 13.661556603773585,
      "grad_norm": 4.065048694610596,
      "learning_rate": 2.7604500604373097e-06,
      "loss": 0.336,
      "num_input_tokens_seen": 15016264,
      "step": 23170
    },
    {
      "epoch": 13.664504716981131,
      "grad_norm": 3.4052159786224365,
      "learning_rate": 2.7581501482311757e-06,
      "loss": 0.3908,
      "num_input_tokens_seen": 15019048,
      "step": 23175
    },
    {
      "epoch": 13.66745283018868,
      "grad_norm": 2.775585889816284,
      "learning_rate": 2.7558508295644513e-06,
      "loss": 0.4569,
      "num_input_tokens_seen": 15021864,
      "step": 23180
    },
    {
      "epoch": 13.670400943396226,
      "grad_norm": 2.618102788925171,
      "learning_rate": 2.7535521050458922e-06,
      "loss": 0.371,
      "num_input_tokens_seen": 15025288,
      "step": 23185
    },
    {
      "epoch": 13.673349056603774,
      "grad_norm": 3.6667377948760986,
      "learning_rate": 2.7512539752840926e-06,
      "loss": 0.2635,
      "num_input_tokens_seen": 15028904,
      "step": 23190
    },
    {
      "epoch": 13.67629716981132,
      "grad_norm": 3.533849000930786,
      "learning_rate": 2.748956440887497e-06,
      "loss": 0.3033,
      "num_input_tokens_seen": 15032168,
      "step": 23195
    },
    {
      "epoch": 13.679245283018869,
      "grad_norm": 1.8508124351501465,
      "learning_rate": 2.7466595024643843e-06,
      "loss": 0.3178,
      "num_input_tokens_seen": 15036168,
      "step": 23200
    },
    {
      "epoch": 13.682193396226415,
      "grad_norm": 4.262358665466309,
      "learning_rate": 2.744363160622878e-06,
      "loss": 0.3536,
      "num_input_tokens_seen": 15039112,
      "step": 23205
    },
    {
      "epoch": 13.685141509433961,
      "grad_norm": 2.944366455078125,
      "learning_rate": 2.742067415970948e-06,
      "loss": 0.3558,
      "num_input_tokens_seen": 15042056,
      "step": 23210
    },
    {
      "epoch": 13.68808962264151,
      "grad_norm": 4.619057655334473,
      "learning_rate": 2.739772269116402e-06,
      "loss": 0.3095,
      "num_input_tokens_seen": 15045192,
      "step": 23215
    },
    {
      "epoch": 13.691037735849056,
      "grad_norm": 4.544205188751221,
      "learning_rate": 2.7374777206668874e-06,
      "loss": 0.3666,
      "num_input_tokens_seen": 15047560,
      "step": 23220
    },
    {
      "epoch": 13.693985849056604,
      "grad_norm": 2.230792760848999,
      "learning_rate": 2.735183771229898e-06,
      "loss": 0.2771,
      "num_input_tokens_seen": 15050568,
      "step": 23225
    },
    {
      "epoch": 13.69693396226415,
      "grad_norm": 4.520796775817871,
      "learning_rate": 2.732890421412765e-06,
      "loss": 0.3832,
      "num_input_tokens_seen": 15053096,
      "step": 23230
    },
    {
      "epoch": 13.699882075471699,
      "grad_norm": 3.9878127574920654,
      "learning_rate": 2.7305976718226624e-06,
      "loss": 0.3007,
      "num_input_tokens_seen": 15056072,
      "step": 23235
    },
    {
      "epoch": 13.702830188679245,
      "grad_norm": 3.636120319366455,
      "learning_rate": 2.728305523066609e-06,
      "loss": 0.2827,
      "num_input_tokens_seen": 15058824,
      "step": 23240
    },
    {
      "epoch": 13.705778301886792,
      "grad_norm": 2.9513299465179443,
      "learning_rate": 2.726013975751458e-06,
      "loss": 0.4613,
      "num_input_tokens_seen": 15061576,
      "step": 23245
    },
    {
      "epoch": 13.70872641509434,
      "grad_norm": 3.1616530418395996,
      "learning_rate": 2.723723030483908e-06,
      "loss": 0.2518,
      "num_input_tokens_seen": 15064136,
      "step": 23250
    },
    {
      "epoch": 13.711674528301886,
      "grad_norm": 3.0323219299316406,
      "learning_rate": 2.7214326878704953e-06,
      "loss": 0.25,
      "num_input_tokens_seen": 15067560,
      "step": 23255
    },
    {
      "epoch": 13.714622641509434,
      "grad_norm": 4.767219066619873,
      "learning_rate": 2.7191429485175993e-06,
      "loss": 0.3152,
      "num_input_tokens_seen": 15070376,
      "step": 23260
    },
    {
      "epoch": 13.71757075471698,
      "grad_norm": 3.44941782951355,
      "learning_rate": 2.716853813031435e-06,
      "loss": 0.3179,
      "num_input_tokens_seen": 15073160,
      "step": 23265
    },
    {
      "epoch": 13.720518867924529,
      "grad_norm": 2.673731565475464,
      "learning_rate": 2.714565282018066e-06,
      "loss": 0.2896,
      "num_input_tokens_seen": 15076200,
      "step": 23270
    },
    {
      "epoch": 13.723466981132075,
      "grad_norm": 4.5651350021362305,
      "learning_rate": 2.7122773560833877e-06,
      "loss": 0.3583,
      "num_input_tokens_seen": 15078888,
      "step": 23275
    },
    {
      "epoch": 13.726415094339622,
      "grad_norm": 3.0082452297210693,
      "learning_rate": 2.709990035833139e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 15082696,
      "step": 23280
    },
    {
      "epoch": 13.72936320754717,
      "grad_norm": 9.332196235656738,
      "learning_rate": 2.707703321872896e-06,
      "loss": 0.3091,
      "num_input_tokens_seen": 15086312,
      "step": 23285
    },
    {
      "epoch": 13.732311320754716,
      "grad_norm": 3.742955207824707,
      "learning_rate": 2.705417214808079e-06,
      "loss": 0.3854,
      "num_input_tokens_seen": 15088904,
      "step": 23290
    },
    {
      "epoch": 13.735259433962264,
      "grad_norm": 4.24866247177124,
      "learning_rate": 2.703131715243945e-06,
      "loss": 0.441,
      "num_input_tokens_seen": 15091624,
      "step": 23295
    },
    {
      "epoch": 13.73820754716981,
      "grad_norm": 2.1419482231140137,
      "learning_rate": 2.7008468237855855e-06,
      "loss": 0.3484,
      "num_input_tokens_seen": 15095336,
      "step": 23300
    },
    {
      "epoch": 13.741155660377359,
      "grad_norm": 2.8141279220581055,
      "learning_rate": 2.69856254103794e-06,
      "loss": 0.3996,
      "num_input_tokens_seen": 15098344,
      "step": 23305
    },
    {
      "epoch": 13.744103773584905,
      "grad_norm": 5.4054718017578125,
      "learning_rate": 2.6962788676057806e-06,
      "loss": 0.2993,
      "num_input_tokens_seen": 15101288,
      "step": 23310
    },
    {
      "epoch": 13.747051886792454,
      "grad_norm": 3.217081069946289,
      "learning_rate": 2.69399580409372e-06,
      "loss": 0.2901,
      "num_input_tokens_seen": 15105800,
      "step": 23315
    },
    {
      "epoch": 13.75,
      "grad_norm": 3.212996244430542,
      "learning_rate": 2.6917133511062076e-06,
      "loss": 0.3235,
      "num_input_tokens_seen": 15109032,
      "step": 23320
    },
    {
      "epoch": 13.752948113207546,
      "grad_norm": 3.103567600250244,
      "learning_rate": 2.6894315092475342e-06,
      "loss": 0.3291,
      "num_input_tokens_seen": 15111432,
      "step": 23325
    },
    {
      "epoch": 13.755896226415095,
      "grad_norm": 3.7833964824676514,
      "learning_rate": 2.6871502791218245e-06,
      "loss": 0.2896,
      "num_input_tokens_seen": 15115624,
      "step": 23330
    },
    {
      "epoch": 13.758844339622641,
      "grad_norm": 2.5667150020599365,
      "learning_rate": 2.684869661333048e-06,
      "loss": 0.4346,
      "num_input_tokens_seen": 15118760,
      "step": 23335
    },
    {
      "epoch": 13.76179245283019,
      "grad_norm": 3.0616402626037598,
      "learning_rate": 2.6825896564850074e-06,
      "loss": 0.3899,
      "num_input_tokens_seen": 15121672,
      "step": 23340
    },
    {
      "epoch": 13.764740566037736,
      "grad_norm": 3.4474267959594727,
      "learning_rate": 2.6803102651813416e-06,
      "loss": 0.2305,
      "num_input_tokens_seen": 15124936,
      "step": 23345
    },
    {
      "epoch": 13.767688679245284,
      "grad_norm": 2.7881791591644287,
      "learning_rate": 2.6780314880255307e-06,
      "loss": 0.2437,
      "num_input_tokens_seen": 15127880,
      "step": 23350
    },
    {
      "epoch": 13.77063679245283,
      "grad_norm": 10.391433715820312,
      "learning_rate": 2.675753325620891e-06,
      "loss": 0.2919,
      "num_input_tokens_seen": 15130600,
      "step": 23355
    },
    {
      "epoch": 13.773584905660378,
      "grad_norm": 2.9293720722198486,
      "learning_rate": 2.6734757785705727e-06,
      "loss": 0.329,
      "num_input_tokens_seen": 15133832,
      "step": 23360
    },
    {
      "epoch": 13.776533018867925,
      "grad_norm": 3.9149794578552246,
      "learning_rate": 2.6711988474775712e-06,
      "loss": 0.2459,
      "num_input_tokens_seen": 15136488,
      "step": 23365
    },
    {
      "epoch": 13.779481132075471,
      "grad_norm": 3.0782127380371094,
      "learning_rate": 2.668922532944711e-06,
      "loss": 0.3016,
      "num_input_tokens_seen": 15140456,
      "step": 23370
    },
    {
      "epoch": 13.78242924528302,
      "grad_norm": 3.3538429737091064,
      "learning_rate": 2.6666468355746566e-06,
      "loss": 0.323,
      "num_input_tokens_seen": 15143400,
      "step": 23375
    },
    {
      "epoch": 13.785377358490566,
      "grad_norm": 3.5852630138397217,
      "learning_rate": 2.6643717559699073e-06,
      "loss": 0.3041,
      "num_input_tokens_seen": 15147144,
      "step": 23380
    },
    {
      "epoch": 13.788325471698114,
      "grad_norm": 4.575610160827637,
      "learning_rate": 2.662097294732803e-06,
      "loss": 0.3522,
      "num_input_tokens_seen": 15150248,
      "step": 23385
    },
    {
      "epoch": 13.79127358490566,
      "grad_norm": 3.5402190685272217,
      "learning_rate": 2.6598234524655165e-06,
      "loss": 0.37,
      "num_input_tokens_seen": 15153352,
      "step": 23390
    },
    {
      "epoch": 13.794221698113208,
      "grad_norm": 2.5658626556396484,
      "learning_rate": 2.657550229770054e-06,
      "loss": 0.2585,
      "num_input_tokens_seen": 15157416,
      "step": 23395
    },
    {
      "epoch": 13.797169811320755,
      "grad_norm": 2.8765759468078613,
      "learning_rate": 2.655277627248265e-06,
      "loss": 0.2781,
      "num_input_tokens_seen": 15160136,
      "step": 23400
    },
    {
      "epoch": 13.800117924528301,
      "grad_norm": 2.761681079864502,
      "learning_rate": 2.65300564550183e-06,
      "loss": 0.3209,
      "num_input_tokens_seen": 15162952,
      "step": 23405
    },
    {
      "epoch": 13.80306603773585,
      "grad_norm": 4.67083740234375,
      "learning_rate": 2.6507342851322647e-06,
      "loss": 0.5031,
      "num_input_tokens_seen": 15166120,
      "step": 23410
    },
    {
      "epoch": 13.806014150943396,
      "grad_norm": 4.497594356536865,
      "learning_rate": 2.6484635467409233e-06,
      "loss": 0.3444,
      "num_input_tokens_seen": 15169480,
      "step": 23415
    },
    {
      "epoch": 13.808962264150944,
      "grad_norm": 6.131136417388916,
      "learning_rate": 2.64619343092899e-06,
      "loss": 0.3246,
      "num_input_tokens_seen": 15172168,
      "step": 23420
    },
    {
      "epoch": 13.81191037735849,
      "grad_norm": 1.9817763566970825,
      "learning_rate": 2.643923938297492e-06,
      "loss": 0.2903,
      "num_input_tokens_seen": 15174824,
      "step": 23425
    },
    {
      "epoch": 13.814858490566039,
      "grad_norm": 4.248171329498291,
      "learning_rate": 2.6416550694472855e-06,
      "loss": 0.3313,
      "num_input_tokens_seen": 15177672,
      "step": 23430
    },
    {
      "epoch": 13.817806603773585,
      "grad_norm": 3.09131121635437,
      "learning_rate": 2.639386824979063e-06,
      "loss": 0.3732,
      "num_input_tokens_seen": 15180424,
      "step": 23435
    },
    {
      "epoch": 13.820754716981131,
      "grad_norm": 3.996742010116577,
      "learning_rate": 2.6371192054933525e-06,
      "loss": 0.278,
      "num_input_tokens_seen": 15182728,
      "step": 23440
    },
    {
      "epoch": 13.82370283018868,
      "grad_norm": 4.954832077026367,
      "learning_rate": 2.634852211590516e-06,
      "loss": 0.3454,
      "num_input_tokens_seen": 15186632,
      "step": 23445
    },
    {
      "epoch": 13.826650943396226,
      "grad_norm": 3.069941997528076,
      "learning_rate": 2.6325858438707473e-06,
      "loss": 0.3547,
      "num_input_tokens_seen": 15189512,
      "step": 23450
    },
    {
      "epoch": 13.829599056603774,
      "grad_norm": 3.874788522720337,
      "learning_rate": 2.630320102934082e-06,
      "loss": 0.2495,
      "num_input_tokens_seen": 15192264,
      "step": 23455
    },
    {
      "epoch": 13.83254716981132,
      "grad_norm": 5.943281650543213,
      "learning_rate": 2.628054989380382e-06,
      "loss": 0.3896,
      "num_input_tokens_seen": 15195304,
      "step": 23460
    },
    {
      "epoch": 13.835495283018869,
      "grad_norm": 3.496005058288574,
      "learning_rate": 2.625790503809346e-06,
      "loss": 0.3302,
      "num_input_tokens_seen": 15197704,
      "step": 23465
    },
    {
      "epoch": 13.838443396226415,
      "grad_norm": 2.210726737976074,
      "learning_rate": 2.6235266468205067e-06,
      "loss": 0.3241,
      "num_input_tokens_seen": 15201000,
      "step": 23470
    },
    {
      "epoch": 13.841391509433961,
      "grad_norm": 3.3681769371032715,
      "learning_rate": 2.621263419013227e-06,
      "loss": 0.2413,
      "num_input_tokens_seen": 15203528,
      "step": 23475
    },
    {
      "epoch": 13.84433962264151,
      "grad_norm": 2.8725521564483643,
      "learning_rate": 2.619000820986711e-06,
      "loss": 0.3177,
      "num_input_tokens_seen": 15211336,
      "step": 23480
    },
    {
      "epoch": 13.847287735849056,
      "grad_norm": 3.379457712173462,
      "learning_rate": 2.616738853339988e-06,
      "loss": 0.4908,
      "num_input_tokens_seen": 15214440,
      "step": 23485
    },
    {
      "epoch": 13.850235849056604,
      "grad_norm": 3.134315013885498,
      "learning_rate": 2.614477516671926e-06,
      "loss": 0.2564,
      "num_input_tokens_seen": 15218184,
      "step": 23490
    },
    {
      "epoch": 13.85318396226415,
      "grad_norm": 4.375835418701172,
      "learning_rate": 2.612216811581223e-06,
      "loss": 0.3063,
      "num_input_tokens_seen": 15221160,
      "step": 23495
    },
    {
      "epoch": 13.856132075471699,
      "grad_norm": 6.046290874481201,
      "learning_rate": 2.6099567386664095e-06,
      "loss": 0.3508,
      "num_input_tokens_seen": 15224616,
      "step": 23500
    },
    {
      "epoch": 13.859080188679245,
      "grad_norm": 4.415574550628662,
      "learning_rate": 2.60769729852585e-06,
      "loss": 0.3848,
      "num_input_tokens_seen": 15227432,
      "step": 23505
    },
    {
      "epoch": 13.862028301886792,
      "grad_norm": 3.3121731281280518,
      "learning_rate": 2.6054384917577413e-06,
      "loss": 0.5102,
      "num_input_tokens_seen": 15231688,
      "step": 23510
    },
    {
      "epoch": 13.86497641509434,
      "grad_norm": 4.796216011047363,
      "learning_rate": 2.60318031896011e-06,
      "loss": 0.3879,
      "num_input_tokens_seen": 15234760,
      "step": 23515
    },
    {
      "epoch": 13.867924528301886,
      "grad_norm": 3.19700026512146,
      "learning_rate": 2.60092278073082e-06,
      "loss": 0.3478,
      "num_input_tokens_seen": 15238408,
      "step": 23520
    },
    {
      "epoch": 13.870872641509434,
      "grad_norm": 2.408585548400879,
      "learning_rate": 2.5986658776675644e-06,
      "loss": 0.3387,
      "num_input_tokens_seen": 15241576,
      "step": 23525
    },
    {
      "epoch": 13.87382075471698,
      "grad_norm": 2.125886917114258,
      "learning_rate": 2.5964096103678666e-06,
      "loss": 0.2476,
      "num_input_tokens_seen": 15248424,
      "step": 23530
    },
    {
      "epoch": 13.876768867924529,
      "grad_norm": 3.173767566680908,
      "learning_rate": 2.5941539794290833e-06,
      "loss": 0.3458,
      "num_input_tokens_seen": 15251688,
      "step": 23535
    },
    {
      "epoch": 13.879716981132075,
      "grad_norm": 3.18393874168396,
      "learning_rate": 2.5918989854484024e-06,
      "loss": 0.2929,
      "num_input_tokens_seen": 15254728,
      "step": 23540
    },
    {
      "epoch": 13.882665094339622,
      "grad_norm": 4.679882526397705,
      "learning_rate": 2.5896446290228417e-06,
      "loss": 0.2642,
      "num_input_tokens_seen": 15257576,
      "step": 23545
    },
    {
      "epoch": 13.88561320754717,
      "grad_norm": 4.4036760330200195,
      "learning_rate": 2.5873909107492547e-06,
      "loss": 0.3741,
      "num_input_tokens_seen": 15260648,
      "step": 23550
    },
    {
      "epoch": 13.888561320754716,
      "grad_norm": 2.97655987739563,
      "learning_rate": 2.5851378312243224e-06,
      "loss": 0.288,
      "num_input_tokens_seen": 15263176,
      "step": 23555
    },
    {
      "epoch": 13.891509433962264,
      "grad_norm": 4.390466690063477,
      "learning_rate": 2.5828853910445572e-06,
      "loss": 0.3092,
      "num_input_tokens_seen": 15266184,
      "step": 23560
    },
    {
      "epoch": 13.89445754716981,
      "grad_norm": 3.3213284015655518,
      "learning_rate": 2.5806335908063012e-06,
      "loss": 0.5112,
      "num_input_tokens_seen": 15269576,
      "step": 23565
    },
    {
      "epoch": 13.897405660377359,
      "grad_norm": 2.511732816696167,
      "learning_rate": 2.5783824311057293e-06,
      "loss": 0.237,
      "num_input_tokens_seen": 15273672,
      "step": 23570
    },
    {
      "epoch": 13.900353773584905,
      "grad_norm": 4.3642897605896,
      "learning_rate": 2.5761319125388433e-06,
      "loss": 0.1778,
      "num_input_tokens_seen": 15276232,
      "step": 23575
    },
    {
      "epoch": 13.903301886792454,
      "grad_norm": 2.8317089080810547,
      "learning_rate": 2.57388203570148e-06,
      "loss": 0.2317,
      "num_input_tokens_seen": 15279432,
      "step": 23580
    },
    {
      "epoch": 13.90625,
      "grad_norm": 5.735544681549072,
      "learning_rate": 2.5716328011893055e-06,
      "loss": 0.3644,
      "num_input_tokens_seen": 15281960,
      "step": 23585
    },
    {
      "epoch": 13.909198113207546,
      "grad_norm": 3.421290636062622,
      "learning_rate": 2.5693842095978127e-06,
      "loss": 0.2984,
      "num_input_tokens_seen": 15285288,
      "step": 23590
    },
    {
      "epoch": 13.912146226415095,
      "grad_norm": 2.5865888595581055,
      "learning_rate": 2.567136261522325e-06,
      "loss": 0.2631,
      "num_input_tokens_seen": 15288360,
      "step": 23595
    },
    {
      "epoch": 13.915094339622641,
      "grad_norm": 3.2070744037628174,
      "learning_rate": 2.5648889575579985e-06,
      "loss": 0.3785,
      "num_input_tokens_seen": 15291784,
      "step": 23600
    },
    {
      "epoch": 13.91804245283019,
      "grad_norm": 3.3929028511047363,
      "learning_rate": 2.562642298299814e-06,
      "loss": 0.249,
      "num_input_tokens_seen": 15295176,
      "step": 23605
    },
    {
      "epoch": 13.920990566037736,
      "grad_norm": 4.07966423034668,
      "learning_rate": 2.560396284342584e-06,
      "loss": 0.2397,
      "num_input_tokens_seen": 15298152,
      "step": 23610
    },
    {
      "epoch": 13.923938679245284,
      "grad_norm": 3.49914813041687,
      "learning_rate": 2.558150916280954e-06,
      "loss": 0.3726,
      "num_input_tokens_seen": 15301480,
      "step": 23615
    },
    {
      "epoch": 13.92688679245283,
      "grad_norm": 4.157562255859375,
      "learning_rate": 2.555906194709392e-06,
      "loss": 0.3428,
      "num_input_tokens_seen": 15304296,
      "step": 23620
    },
    {
      "epoch": 13.929834905660378,
      "grad_norm": 1.8992701768875122,
      "learning_rate": 2.553662120222199e-06,
      "loss": 0.3194,
      "num_input_tokens_seen": 15307656,
      "step": 23625
    },
    {
      "epoch": 13.932783018867925,
      "grad_norm": 3.724616765975952,
      "learning_rate": 2.5514186934135026e-06,
      "loss": 0.367,
      "num_input_tokens_seen": 15311272,
      "step": 23630
    },
    {
      "epoch": 13.935731132075471,
      "grad_norm": 3.9686431884765625,
      "learning_rate": 2.54917591487726e-06,
      "loss": 0.3858,
      "num_input_tokens_seen": 15315880,
      "step": 23635
    },
    {
      "epoch": 13.93867924528302,
      "grad_norm": 3.09814190864563,
      "learning_rate": 2.5469337852072547e-06,
      "loss": 0.38,
      "num_input_tokens_seen": 15319624,
      "step": 23640
    },
    {
      "epoch": 13.941627358490566,
      "grad_norm": 4.1180620193481445,
      "learning_rate": 2.5446923049971035e-06,
      "loss": 0.42,
      "num_input_tokens_seen": 15322568,
      "step": 23645
    },
    {
      "epoch": 13.944575471698114,
      "grad_norm": 3.999401092529297,
      "learning_rate": 2.5424514748402463e-06,
      "loss": 0.3037,
      "num_input_tokens_seen": 15325384,
      "step": 23650
    },
    {
      "epoch": 13.94752358490566,
      "grad_norm": 6.311122417449951,
      "learning_rate": 2.540211295329953e-06,
      "loss": 0.3648,
      "num_input_tokens_seen": 15328680,
      "step": 23655
    },
    {
      "epoch": 13.950471698113208,
      "grad_norm": 6.677722454071045,
      "learning_rate": 2.5379717670593197e-06,
      "loss": 0.5302,
      "num_input_tokens_seen": 15331656,
      "step": 23660
    },
    {
      "epoch": 13.953419811320755,
      "grad_norm": 5.281729221343994,
      "learning_rate": 2.53573289062127e-06,
      "loss": 0.3816,
      "num_input_tokens_seen": 15334728,
      "step": 23665
    },
    {
      "epoch": 13.956367924528301,
      "grad_norm": 3.0821824073791504,
      "learning_rate": 2.5334946666085605e-06,
      "loss": 0.3008,
      "num_input_tokens_seen": 15337096,
      "step": 23670
    },
    {
      "epoch": 13.95931603773585,
      "grad_norm": 4.383470058441162,
      "learning_rate": 2.531257095613766e-06,
      "loss": 0.3546,
      "num_input_tokens_seen": 15340264,
      "step": 23675
    },
    {
      "epoch": 13.962264150943396,
      "grad_norm": 2.747466564178467,
      "learning_rate": 2.529020178229297e-06,
      "loss": 0.3686,
      "num_input_tokens_seen": 15343496,
      "step": 23680
    },
    {
      "epoch": 13.965212264150944,
      "grad_norm": 4.404281139373779,
      "learning_rate": 2.5267839150473846e-06,
      "loss": 0.3734,
      "num_input_tokens_seen": 15347432,
      "step": 23685
    },
    {
      "epoch": 13.96816037735849,
      "grad_norm": 2.2000181674957275,
      "learning_rate": 2.5245483066600896e-06,
      "loss": 0.3615,
      "num_input_tokens_seen": 15351176,
      "step": 23690
    },
    {
      "epoch": 13.971108490566039,
      "grad_norm": 2.410264015197754,
      "learning_rate": 2.5223133536592996e-06,
      "loss": 0.3726,
      "num_input_tokens_seen": 15354632,
      "step": 23695
    },
    {
      "epoch": 13.974056603773585,
      "grad_norm": 1.766650915145874,
      "learning_rate": 2.520079056636725e-06,
      "loss": 0.2935,
      "num_input_tokens_seen": 15357832,
      "step": 23700
    },
    {
      "epoch": 13.977004716981131,
      "grad_norm": 3.581083059310913,
      "learning_rate": 2.5178454161839106e-06,
      "loss": 0.4072,
      "num_input_tokens_seen": 15361448,
      "step": 23705
    },
    {
      "epoch": 13.97995283018868,
      "grad_norm": 2.1739320755004883,
      "learning_rate": 2.5156124328922195e-06,
      "loss": 0.3138,
      "num_input_tokens_seen": 15364872,
      "step": 23710
    },
    {
      "epoch": 13.982900943396226,
      "grad_norm": 4.068410873413086,
      "learning_rate": 2.513380107352844e-06,
      "loss": 0.2637,
      "num_input_tokens_seen": 15367976,
      "step": 23715
    },
    {
      "epoch": 13.985849056603774,
      "grad_norm": 5.804596900939941,
      "learning_rate": 2.5111484401568014e-06,
      "loss": 0.3949,
      "num_input_tokens_seen": 15372168,
      "step": 23720
    },
    {
      "epoch": 13.98879716981132,
      "grad_norm": 4.564752101898193,
      "learning_rate": 2.508917431894936e-06,
      "loss": 0.2774,
      "num_input_tokens_seen": 15374632,
      "step": 23725
    },
    {
      "epoch": 13.991745283018869,
      "grad_norm": 2.84310245513916,
      "learning_rate": 2.5066870831579144e-06,
      "loss": 0.3368,
      "num_input_tokens_seen": 15377896,
      "step": 23730
    },
    {
      "epoch": 13.994693396226415,
      "grad_norm": 2.841031551361084,
      "learning_rate": 2.504457394536235e-06,
      "loss": 0.3205,
      "num_input_tokens_seen": 15381384,
      "step": 23735
    },
    {
      "epoch": 13.997641509433961,
      "grad_norm": 3.24414324760437,
      "learning_rate": 2.502228366620216e-06,
      "loss": 0.2875,
      "num_input_tokens_seen": 15384200,
      "step": 23740
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.581352174282074,
      "eval_runtime": 19.5437,
      "eval_samples_per_second": 86.78,
      "eval_steps_per_second": 21.695,
      "num_input_tokens_seen": 15387192,
      "step": 23744
    },
    {
      "epoch": 14.00058962264151,
      "grad_norm": 3.9099087715148926,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.3018,
      "num_input_tokens_seen": 15387864,
      "step": 23745
    },
    {
      "epoch": 14.003537735849056,
      "grad_norm": 6.0520501136779785,
      "learning_rate": 2.497772295265561e-06,
      "loss": 0.3271,
      "num_input_tokens_seen": 15391352,
      "step": 23750
    },
    {
      "epoch": 14.006485849056604,
      "grad_norm": 5.352379322052002,
      "learning_rate": 2.4955452530066897e-06,
      "loss": 0.4479,
      "num_input_tokens_seen": 15394104,
      "step": 23755
    },
    {
      "epoch": 14.00943396226415,
      "grad_norm": 3.5568137168884277,
      "learning_rate": 2.4933188738130043e-06,
      "loss": 0.3075,
      "num_input_tokens_seen": 15397176,
      "step": 23760
    },
    {
      "epoch": 14.012382075471699,
      "grad_norm": 2.5683794021606445,
      "learning_rate": 2.49109315827395e-06,
      "loss": 0.3588,
      "num_input_tokens_seen": 15401272,
      "step": 23765
    },
    {
      "epoch": 14.015330188679245,
      "grad_norm": 1.9350239038467407,
      "learning_rate": 2.4888681069787975e-06,
      "loss": 0.3223,
      "num_input_tokens_seen": 15403960,
      "step": 23770
    },
    {
      "epoch": 14.018278301886792,
      "grad_norm": 7.1260857582092285,
      "learning_rate": 2.4866437205166353e-06,
      "loss": 0.4258,
      "num_input_tokens_seen": 15407192,
      "step": 23775
    },
    {
      "epoch": 14.02122641509434,
      "grad_norm": 5.985248565673828,
      "learning_rate": 2.4844199994763803e-06,
      "loss": 0.3614,
      "num_input_tokens_seen": 15412600,
      "step": 23780
    },
    {
      "epoch": 14.024174528301886,
      "grad_norm": 3.752470016479492,
      "learning_rate": 2.482196944446772e-06,
      "loss": 0.3507,
      "num_input_tokens_seen": 15415768,
      "step": 23785
    },
    {
      "epoch": 14.027122641509434,
      "grad_norm": 2.3987860679626465,
      "learning_rate": 2.4799745560163736e-06,
      "loss": 0.2527,
      "num_input_tokens_seen": 15420344,
      "step": 23790
    },
    {
      "epoch": 14.03007075471698,
      "grad_norm": 3.41813325881958,
      "learning_rate": 2.4777528347735707e-06,
      "loss": 0.3616,
      "num_input_tokens_seen": 15423320,
      "step": 23795
    },
    {
      "epoch": 14.033018867924529,
      "grad_norm": 4.608100891113281,
      "learning_rate": 2.4755317813065766e-06,
      "loss": 0.2221,
      "num_input_tokens_seen": 15425720,
      "step": 23800
    },
    {
      "epoch": 14.035966981132075,
      "grad_norm": 2.620258331298828,
      "learning_rate": 2.4733113962034234e-06,
      "loss": 0.2722,
      "num_input_tokens_seen": 15428184,
      "step": 23805
    },
    {
      "epoch": 14.038915094339623,
      "grad_norm": 4.3311004638671875,
      "learning_rate": 2.4710916800519674e-06,
      "loss": 0.2576,
      "num_input_tokens_seen": 15431224,
      "step": 23810
    },
    {
      "epoch": 14.04186320754717,
      "grad_norm": 3.021972417831421,
      "learning_rate": 2.4688726334398883e-06,
      "loss": 0.3235,
      "num_input_tokens_seen": 15434456,
      "step": 23815
    },
    {
      "epoch": 14.044811320754716,
      "grad_norm": 4.470269680023193,
      "learning_rate": 2.466654256954688e-06,
      "loss": 0.4533,
      "num_input_tokens_seen": 15437176,
      "step": 23820
    },
    {
      "epoch": 14.047759433962264,
      "grad_norm": 3.6224608421325684,
      "learning_rate": 2.4644365511836895e-06,
      "loss": 0.2729,
      "num_input_tokens_seen": 15440184,
      "step": 23825
    },
    {
      "epoch": 14.05070754716981,
      "grad_norm": 3.6807749271392822,
      "learning_rate": 2.4622195167140432e-06,
      "loss": 0.3812,
      "num_input_tokens_seen": 15442872,
      "step": 23830
    },
    {
      "epoch": 14.053655660377359,
      "grad_norm": 5.712309837341309,
      "learning_rate": 2.4600031541327173e-06,
      "loss": 0.3657,
      "num_input_tokens_seen": 15445496,
      "step": 23835
    },
    {
      "epoch": 14.056603773584905,
      "grad_norm": 3.9560117721557617,
      "learning_rate": 2.457787464026503e-06,
      "loss": 0.3243,
      "num_input_tokens_seen": 15449240,
      "step": 23840
    },
    {
      "epoch": 14.059551886792454,
      "grad_norm": 6.930364608764648,
      "learning_rate": 2.455572446982014e-06,
      "loss": 0.3516,
      "num_input_tokens_seen": 15452408,
      "step": 23845
    },
    {
      "epoch": 14.0625,
      "grad_norm": 3.83000111579895,
      "learning_rate": 2.453358103585686e-06,
      "loss": 0.3722,
      "num_input_tokens_seen": 15455032,
      "step": 23850
    },
    {
      "epoch": 14.065448113207546,
      "grad_norm": 3.0626509189605713,
      "learning_rate": 2.4511444344237733e-06,
      "loss": 0.2509,
      "num_input_tokens_seen": 15458072,
      "step": 23855
    },
    {
      "epoch": 14.068396226415095,
      "grad_norm": 2.62334942817688,
      "learning_rate": 2.4489314400823567e-06,
      "loss": 0.2949,
      "num_input_tokens_seen": 15461688,
      "step": 23860
    },
    {
      "epoch": 14.071344339622641,
      "grad_norm": 4.834048748016357,
      "learning_rate": 2.446719121147337e-06,
      "loss": 0.4967,
      "num_input_tokens_seen": 15464856,
      "step": 23865
    },
    {
      "epoch": 14.07429245283019,
      "grad_norm": 3.5948657989501953,
      "learning_rate": 2.4445074782044347e-06,
      "loss": 0.3212,
      "num_input_tokens_seen": 15467320,
      "step": 23870
    },
    {
      "epoch": 14.077240566037736,
      "grad_norm": 2.9487783908843994,
      "learning_rate": 2.442296511839191e-06,
      "loss": 0.3569,
      "num_input_tokens_seen": 15470616,
      "step": 23875
    },
    {
      "epoch": 14.080188679245284,
      "grad_norm": 3.3125271797180176,
      "learning_rate": 2.4400862226369687e-06,
      "loss": 0.364,
      "num_input_tokens_seen": 15473496,
      "step": 23880
    },
    {
      "epoch": 14.08313679245283,
      "grad_norm": 2.1748263835906982,
      "learning_rate": 2.4378766111829514e-06,
      "loss": 0.306,
      "num_input_tokens_seen": 15476184,
      "step": 23885
    },
    {
      "epoch": 14.086084905660377,
      "grad_norm": 3.0458405017852783,
      "learning_rate": 2.435667678062142e-06,
      "loss": 0.215,
      "num_input_tokens_seen": 15480312,
      "step": 23890
    },
    {
      "epoch": 14.089033018867925,
      "grad_norm": 4.159156322479248,
      "learning_rate": 2.4334594238593682e-06,
      "loss": 0.3139,
      "num_input_tokens_seen": 15482904,
      "step": 23895
    },
    {
      "epoch": 14.091981132075471,
      "grad_norm": 3.446516275405884,
      "learning_rate": 2.4312518491592727e-06,
      "loss": 0.3674,
      "num_input_tokens_seen": 15486072,
      "step": 23900
    },
    {
      "epoch": 14.09492924528302,
      "grad_norm": 3.058457136154175,
      "learning_rate": 2.429044954546322e-06,
      "loss": 0.4058,
      "num_input_tokens_seen": 15489464,
      "step": 23905
    },
    {
      "epoch": 14.097877358490566,
      "grad_norm": 3.7812583446502686,
      "learning_rate": 2.426838740604799e-06,
      "loss": 0.2495,
      "num_input_tokens_seen": 15492664,
      "step": 23910
    },
    {
      "epoch": 14.100825471698114,
      "grad_norm": 3.814666271209717,
      "learning_rate": 2.4246332079188066e-06,
      "loss": 0.2858,
      "num_input_tokens_seen": 15495544,
      "step": 23915
    },
    {
      "epoch": 14.10377358490566,
      "grad_norm": 4.019274711608887,
      "learning_rate": 2.4224283570722745e-06,
      "loss": 0.2771,
      "num_input_tokens_seen": 15499832,
      "step": 23920
    },
    {
      "epoch": 14.106721698113208,
      "grad_norm": 2.738748073577881,
      "learning_rate": 2.420224188648943e-06,
      "loss": 0.2255,
      "num_input_tokens_seen": 15503128,
      "step": 23925
    },
    {
      "epoch": 14.109669811320755,
      "grad_norm": 3.3637595176696777,
      "learning_rate": 2.418020703232376e-06,
      "loss": 0.4938,
      "num_input_tokens_seen": 15508248,
      "step": 23930
    },
    {
      "epoch": 14.112617924528301,
      "grad_norm": 3.82148814201355,
      "learning_rate": 2.4158179014059556e-06,
      "loss": 0.4226,
      "num_input_tokens_seen": 15512728,
      "step": 23935
    },
    {
      "epoch": 14.11556603773585,
      "grad_norm": 7.406649112701416,
      "learning_rate": 2.413615783752883e-06,
      "loss": 0.2905,
      "num_input_tokens_seen": 15515864,
      "step": 23940
    },
    {
      "epoch": 14.118514150943396,
      "grad_norm": 3.343538761138916,
      "learning_rate": 2.4114143508561767e-06,
      "loss": 0.3375,
      "num_input_tokens_seen": 15520216,
      "step": 23945
    },
    {
      "epoch": 14.121462264150944,
      "grad_norm": 3.037466526031494,
      "learning_rate": 2.4092136032986783e-06,
      "loss": 0.3479,
      "num_input_tokens_seen": 15523288,
      "step": 23950
    },
    {
      "epoch": 14.12441037735849,
      "grad_norm": 5.757579803466797,
      "learning_rate": 2.407013541663043e-06,
      "loss": 0.359,
      "num_input_tokens_seen": 15526584,
      "step": 23955
    },
    {
      "epoch": 14.127358490566039,
      "grad_norm": 3.36002254486084,
      "learning_rate": 2.4048141665317493e-06,
      "loss": 0.3448,
      "num_input_tokens_seen": 15531064,
      "step": 23960
    },
    {
      "epoch": 14.130306603773585,
      "grad_norm": 3.152108669281006,
      "learning_rate": 2.40261547848709e-06,
      "loss": 0.2894,
      "num_input_tokens_seen": 15533944,
      "step": 23965
    },
    {
      "epoch": 14.133254716981131,
      "grad_norm": 1.8747578859329224,
      "learning_rate": 2.400417478111176e-06,
      "loss": 0.1854,
      "num_input_tokens_seen": 15537112,
      "step": 23970
    },
    {
      "epoch": 14.13620283018868,
      "grad_norm": 3.333301544189453,
      "learning_rate": 2.3982201659859387e-06,
      "loss": 0.4079,
      "num_input_tokens_seen": 15539672,
      "step": 23975
    },
    {
      "epoch": 14.139150943396226,
      "grad_norm": 3.4324235916137695,
      "learning_rate": 2.3960235426931237e-06,
      "loss": 0.2226,
      "num_input_tokens_seen": 15542360,
      "step": 23980
    },
    {
      "epoch": 14.142099056603774,
      "grad_norm": 4.777608394622803,
      "learning_rate": 2.3938276088143003e-06,
      "loss": 0.3527,
      "num_input_tokens_seen": 15545400,
      "step": 23985
    },
    {
      "epoch": 14.14504716981132,
      "grad_norm": 3.3481335639953613,
      "learning_rate": 2.391632364930849e-06,
      "loss": 0.2915,
      "num_input_tokens_seen": 15550200,
      "step": 23990
    },
    {
      "epoch": 14.147995283018869,
      "grad_norm": 4.449940204620361,
      "learning_rate": 2.3894378116239706e-06,
      "loss": 0.2669,
      "num_input_tokens_seen": 15553688,
      "step": 23995
    },
    {
      "epoch": 14.150943396226415,
      "grad_norm": 2.536761999130249,
      "learning_rate": 2.387243949474683e-06,
      "loss": 0.2848,
      "num_input_tokens_seen": 15556536,
      "step": 24000
    },
    {
      "epoch": 14.153891509433961,
      "grad_norm": 4.768927574157715,
      "learning_rate": 2.38505077906382e-06,
      "loss": 0.2045,
      "num_input_tokens_seen": 15560152,
      "step": 24005
    },
    {
      "epoch": 14.15683962264151,
      "grad_norm": 4.905303478240967,
      "learning_rate": 2.382858300972031e-06,
      "loss": 0.2928,
      "num_input_tokens_seen": 15563000,
      "step": 24010
    },
    {
      "epoch": 14.159787735849056,
      "grad_norm": 3.070875644683838,
      "learning_rate": 2.380666515779788e-06,
      "loss": 0.4048,
      "num_input_tokens_seen": 15566584,
      "step": 24015
    },
    {
      "epoch": 14.162735849056604,
      "grad_norm": 3.300879716873169,
      "learning_rate": 2.3784754240673734e-06,
      "loss": 0.2808,
      "num_input_tokens_seen": 15570488,
      "step": 24020
    },
    {
      "epoch": 14.16568396226415,
      "grad_norm": 3.139488458633423,
      "learning_rate": 2.3762850264148883e-06,
      "loss": 0.3456,
      "num_input_tokens_seen": 15573720,
      "step": 24025
    },
    {
      "epoch": 14.168632075471699,
      "grad_norm": 2.9105446338653564,
      "learning_rate": 2.374095323402251e-06,
      "loss": 0.2867,
      "num_input_tokens_seen": 15577016,
      "step": 24030
    },
    {
      "epoch": 14.171580188679245,
      "grad_norm": 3.9092280864715576,
      "learning_rate": 2.371906315609193e-06,
      "loss": 0.3012,
      "num_input_tokens_seen": 15579992,
      "step": 24035
    },
    {
      "epoch": 14.174528301886792,
      "grad_norm": 4.6776628494262695,
      "learning_rate": 2.369718003615263e-06,
      "loss": 0.3774,
      "num_input_tokens_seen": 15584024,
      "step": 24040
    },
    {
      "epoch": 14.17747641509434,
      "grad_norm": 2.873121500015259,
      "learning_rate": 2.3675303879998284e-06,
      "loss": 0.2223,
      "num_input_tokens_seen": 15587000,
      "step": 24045
    },
    {
      "epoch": 14.180424528301886,
      "grad_norm": 3.4861552715301514,
      "learning_rate": 2.365343469342068e-06,
      "loss": 0.4585,
      "num_input_tokens_seen": 15593720,
      "step": 24050
    },
    {
      "epoch": 14.183372641509434,
      "grad_norm": 3.3061325550079346,
      "learning_rate": 2.3631572482209803e-06,
      "loss": 0.2434,
      "num_input_tokens_seen": 15596472,
      "step": 24055
    },
    {
      "epoch": 14.18632075471698,
      "grad_norm": 3.057220935821533,
      "learning_rate": 2.3609717252153752e-06,
      "loss": 0.3982,
      "num_input_tokens_seen": 15600856,
      "step": 24060
    },
    {
      "epoch": 14.189268867924529,
      "grad_norm": 3.2140512466430664,
      "learning_rate": 2.35878690090388e-06,
      "loss": 0.2487,
      "num_input_tokens_seen": 15603608,
      "step": 24065
    },
    {
      "epoch": 14.192216981132075,
      "grad_norm": 3.2735297679901123,
      "learning_rate": 2.356602775864935e-06,
      "loss": 0.2462,
      "num_input_tokens_seen": 15606072,
      "step": 24070
    },
    {
      "epoch": 14.195165094339623,
      "grad_norm": 1.9559108018875122,
      "learning_rate": 2.354419350676796e-06,
      "loss": 0.3044,
      "num_input_tokens_seen": 15609624,
      "step": 24075
    },
    {
      "epoch": 14.19811320754717,
      "grad_norm": 2.392279863357544,
      "learning_rate": 2.3522366259175377e-06,
      "loss": 0.3361,
      "num_input_tokens_seen": 15612696,
      "step": 24080
    },
    {
      "epoch": 14.201061320754716,
      "grad_norm": 5.571837425231934,
      "learning_rate": 2.350054602165044e-06,
      "loss": 0.272,
      "num_input_tokens_seen": 15615448,
      "step": 24085
    },
    {
      "epoch": 14.204009433962264,
      "grad_norm": 2.9382777214050293,
      "learning_rate": 2.3478732799970143e-06,
      "loss": 0.3562,
      "num_input_tokens_seen": 15618840,
      "step": 24090
    },
    {
      "epoch": 14.20695754716981,
      "grad_norm": 4.532802104949951,
      "learning_rate": 2.3456926599909646e-06,
      "loss": 0.3301,
      "num_input_tokens_seen": 15621784,
      "step": 24095
    },
    {
      "epoch": 14.209905660377359,
      "grad_norm": 5.3518595695495605,
      "learning_rate": 2.343512742724222e-06,
      "loss": 0.3328,
      "num_input_tokens_seen": 15625144,
      "step": 24100
    },
    {
      "epoch": 14.212853773584905,
      "grad_norm": 3.2569234371185303,
      "learning_rate": 2.341333528773928e-06,
      "loss": 0.3055,
      "num_input_tokens_seen": 15628344,
      "step": 24105
    },
    {
      "epoch": 14.215801886792454,
      "grad_norm": 3.211573839187622,
      "learning_rate": 2.3391550187170427e-06,
      "loss": 0.3769,
      "num_input_tokens_seen": 15632280,
      "step": 24110
    },
    {
      "epoch": 14.21875,
      "grad_norm": 3.5239181518554688,
      "learning_rate": 2.336977213130333e-06,
      "loss": 0.4271,
      "num_input_tokens_seen": 15634872,
      "step": 24115
    },
    {
      "epoch": 14.221698113207546,
      "grad_norm": 4.40362024307251,
      "learning_rate": 2.3348001125903837e-06,
      "loss": 0.3204,
      "num_input_tokens_seen": 15638328,
      "step": 24120
    },
    {
      "epoch": 14.224646226415095,
      "grad_norm": 4.319519519805908,
      "learning_rate": 2.3326237176735905e-06,
      "loss": 0.2991,
      "num_input_tokens_seen": 15641496,
      "step": 24125
    },
    {
      "epoch": 14.227594339622641,
      "grad_norm": 5.2052412033081055,
      "learning_rate": 2.330448028956164e-06,
      "loss": 0.431,
      "num_input_tokens_seen": 15645656,
      "step": 24130
    },
    {
      "epoch": 14.23054245283019,
      "grad_norm": 1.9001232385635376,
      "learning_rate": 2.3282730470141255e-06,
      "loss": 0.2788,
      "num_input_tokens_seen": 15648824,
      "step": 24135
    },
    {
      "epoch": 14.233490566037736,
      "grad_norm": 2.982280969619751,
      "learning_rate": 2.3260987724233143e-06,
      "loss": 0.2677,
      "num_input_tokens_seen": 15651800,
      "step": 24140
    },
    {
      "epoch": 14.236438679245284,
      "grad_norm": 3.0677859783172607,
      "learning_rate": 2.323925205759374e-06,
      "loss": 0.3641,
      "num_input_tokens_seen": 15654072,
      "step": 24145
    },
    {
      "epoch": 14.23938679245283,
      "grad_norm": 2.8951833248138428,
      "learning_rate": 2.3217523475977715e-06,
      "loss": 0.3819,
      "num_input_tokens_seen": 15657016,
      "step": 24150
    },
    {
      "epoch": 14.242334905660377,
      "grad_norm": 2.5703954696655273,
      "learning_rate": 2.3195801985137773e-06,
      "loss": 0.3989,
      "num_input_tokens_seen": 15660024,
      "step": 24155
    },
    {
      "epoch": 14.245283018867925,
      "grad_norm": 4.855828285217285,
      "learning_rate": 2.317408759082478e-06,
      "loss": 0.324,
      "num_input_tokens_seen": 15662680,
      "step": 24160
    },
    {
      "epoch": 14.248231132075471,
      "grad_norm": 3.911728620529175,
      "learning_rate": 2.31523802987877e-06,
      "loss": 0.3008,
      "num_input_tokens_seen": 15666424,
      "step": 24165
    },
    {
      "epoch": 14.25117924528302,
      "grad_norm": 3.1167654991149902,
      "learning_rate": 2.3130680114773637e-06,
      "loss": 0.3744,
      "num_input_tokens_seen": 15669368,
      "step": 24170
    },
    {
      "epoch": 14.254127358490566,
      "grad_norm": 4.284285068511963,
      "learning_rate": 2.310898704452782e-06,
      "loss": 0.3495,
      "num_input_tokens_seen": 15672280,
      "step": 24175
    },
    {
      "epoch": 14.257075471698114,
      "grad_norm": 4.324525833129883,
      "learning_rate": 2.3087301093793584e-06,
      "loss": 0.3292,
      "num_input_tokens_seen": 15674648,
      "step": 24180
    },
    {
      "epoch": 14.26002358490566,
      "grad_norm": 3.0487468242645264,
      "learning_rate": 2.306562226831237e-06,
      "loss": 0.3542,
      "num_input_tokens_seen": 15677720,
      "step": 24185
    },
    {
      "epoch": 14.262971698113208,
      "grad_norm": 4.913305759429932,
      "learning_rate": 2.304395057382374e-06,
      "loss": 0.3112,
      "num_input_tokens_seen": 15683800,
      "step": 24190
    },
    {
      "epoch": 14.265919811320755,
      "grad_norm": 6.0924482345581055,
      "learning_rate": 2.3022286016065354e-06,
      "loss": 0.3221,
      "num_input_tokens_seen": 15687448,
      "step": 24195
    },
    {
      "epoch": 14.268867924528301,
      "grad_norm": 2.696397542953491,
      "learning_rate": 2.300062860077303e-06,
      "loss": 0.3838,
      "num_input_tokens_seen": 15690488,
      "step": 24200
    },
    {
      "epoch": 14.27181603773585,
      "grad_norm": 3.9756016731262207,
      "learning_rate": 2.297897833368064e-06,
      "loss": 0.288,
      "num_input_tokens_seen": 15693272,
      "step": 24205
    },
    {
      "epoch": 14.274764150943396,
      "grad_norm": 4.36544942855835,
      "learning_rate": 2.2957335220520194e-06,
      "loss": 0.2927,
      "num_input_tokens_seen": 15695992,
      "step": 24210
    },
    {
      "epoch": 14.277712264150944,
      "grad_norm": 4.559852123260498,
      "learning_rate": 2.293569926702179e-06,
      "loss": 0.3092,
      "num_input_tokens_seen": 15699224,
      "step": 24215
    },
    {
      "epoch": 14.28066037735849,
      "grad_norm": 4.419114112854004,
      "learning_rate": 2.291407047891366e-06,
      "loss": 0.4026,
      "num_input_tokens_seen": 15701880,
      "step": 24220
    },
    {
      "epoch": 14.283608490566039,
      "grad_norm": 2.8736398220062256,
      "learning_rate": 2.2892448861922075e-06,
      "loss": 0.3749,
      "num_input_tokens_seen": 15705368,
      "step": 24225
    },
    {
      "epoch": 14.286556603773585,
      "grad_norm": 3.318871021270752,
      "learning_rate": 2.2870834421771505e-06,
      "loss": 0.3742,
      "num_input_tokens_seen": 15709336,
      "step": 24230
    },
    {
      "epoch": 14.289504716981131,
      "grad_norm": 5.574361801147461,
      "learning_rate": 2.2849227164184433e-06,
      "loss": 0.3292,
      "num_input_tokens_seen": 15713240,
      "step": 24235
    },
    {
      "epoch": 14.29245283018868,
      "grad_norm": 4.656291961669922,
      "learning_rate": 2.2827627094881473e-06,
      "loss": 0.3362,
      "num_input_tokens_seen": 15715992,
      "step": 24240
    },
    {
      "epoch": 14.295400943396226,
      "grad_norm": 3.5355987548828125,
      "learning_rate": 2.2806034219581364e-06,
      "loss": 0.2435,
      "num_input_tokens_seen": 15718840,
      "step": 24245
    },
    {
      "epoch": 14.298349056603774,
      "grad_norm": 5.335689544677734,
      "learning_rate": 2.278444854400089e-06,
      "loss": 0.3467,
      "num_input_tokens_seen": 15721240,
      "step": 24250
    },
    {
      "epoch": 14.30129716981132,
      "grad_norm": 3.866835355758667,
      "learning_rate": 2.276287007385496e-06,
      "loss": 0.3209,
      "num_input_tokens_seen": 15723832,
      "step": 24255
    },
    {
      "epoch": 14.304245283018869,
      "grad_norm": 3.203112840652466,
      "learning_rate": 2.2741298814856542e-06,
      "loss": 0.3326,
      "num_input_tokens_seen": 15726840,
      "step": 24260
    },
    {
      "epoch": 14.307193396226415,
      "grad_norm": 2.6691479682922363,
      "learning_rate": 2.2719734772716763e-06,
      "loss": 0.3562,
      "num_input_tokens_seen": 15729944,
      "step": 24265
    },
    {
      "epoch": 14.310141509433961,
      "grad_norm": 2.574572801589966,
      "learning_rate": 2.269817795314477e-06,
      "loss": 0.3642,
      "num_input_tokens_seen": 15733144,
      "step": 24270
    },
    {
      "epoch": 14.31308962264151,
      "grad_norm": 3.9222230911254883,
      "learning_rate": 2.2676628361847834e-06,
      "loss": 0.3182,
      "num_input_tokens_seen": 15736312,
      "step": 24275
    },
    {
      "epoch": 14.316037735849056,
      "grad_norm": 4.071510314941406,
      "learning_rate": 2.2655086004531296e-06,
      "loss": 0.3005,
      "num_input_tokens_seen": 15739672,
      "step": 24280
    },
    {
      "epoch": 14.318985849056604,
      "grad_norm": 4.813467502593994,
      "learning_rate": 2.2633550886898583e-06,
      "loss": 0.3011,
      "num_input_tokens_seen": 15742648,
      "step": 24285
    },
    {
      "epoch": 14.32193396226415,
      "grad_norm": 3.6768999099731445,
      "learning_rate": 2.26120230146512e-06,
      "loss": 0.3171,
      "num_input_tokens_seen": 15745560,
      "step": 24290
    },
    {
      "epoch": 14.324882075471699,
      "grad_norm": 3.395296812057495,
      "learning_rate": 2.2590502393488777e-06,
      "loss": 0.2555,
      "num_input_tokens_seen": 15748472,
      "step": 24295
    },
    {
      "epoch": 14.327830188679245,
      "grad_norm": 5.281701564788818,
      "learning_rate": 2.256898902910898e-06,
      "loss": 0.3923,
      "num_input_tokens_seen": 15751000,
      "step": 24300
    },
    {
      "epoch": 14.330778301886792,
      "grad_norm": 3.6691763401031494,
      "learning_rate": 2.2547482927207548e-06,
      "loss": 0.3595,
      "num_input_tokens_seen": 15753400,
      "step": 24305
    },
    {
      "epoch": 14.33372641509434,
      "grad_norm": 7.117686748504639,
      "learning_rate": 2.252598409347833e-06,
      "loss": 0.2451,
      "num_input_tokens_seen": 15756696,
      "step": 24310
    },
    {
      "epoch": 14.336674528301886,
      "grad_norm": 2.705073118209839,
      "learning_rate": 2.250449253361323e-06,
      "loss": 0.271,
      "num_input_tokens_seen": 15759384,
      "step": 24315
    },
    {
      "epoch": 14.339622641509434,
      "grad_norm": 1.7301428318023682,
      "learning_rate": 2.2483008253302214e-06,
      "loss": 0.3123,
      "num_input_tokens_seen": 15762488,
      "step": 24320
    },
    {
      "epoch": 14.34257075471698,
      "grad_norm": 4.328967094421387,
      "learning_rate": 2.246153125823337e-06,
      "loss": 0.3675,
      "num_input_tokens_seen": 15765112,
      "step": 24325
    },
    {
      "epoch": 14.345518867924529,
      "grad_norm": 4.3187994956970215,
      "learning_rate": 2.2440061554092813e-06,
      "loss": 0.3731,
      "num_input_tokens_seen": 15768024,
      "step": 24330
    },
    {
      "epoch": 14.348466981132075,
      "grad_norm": 3.400721311569214,
      "learning_rate": 2.2418599146564714e-06,
      "loss": 0.3676,
      "num_input_tokens_seen": 15771512,
      "step": 24335
    },
    {
      "epoch": 14.351415094339623,
      "grad_norm": 7.100522994995117,
      "learning_rate": 2.239714404133138e-06,
      "loss": 0.359,
      "num_input_tokens_seen": 15774520,
      "step": 24340
    },
    {
      "epoch": 14.35436320754717,
      "grad_norm": 3.196965456008911,
      "learning_rate": 2.2375696244073126e-06,
      "loss": 0.2749,
      "num_input_tokens_seen": 15778072,
      "step": 24345
    },
    {
      "epoch": 14.357311320754716,
      "grad_norm": 2.483426809310913,
      "learning_rate": 2.235425576046834e-06,
      "loss": 0.3052,
      "num_input_tokens_seen": 15780920,
      "step": 24350
    },
    {
      "epoch": 14.360259433962264,
      "grad_norm": 3.707862138748169,
      "learning_rate": 2.233282259619347e-06,
      "loss": 0.3557,
      "num_input_tokens_seen": 15784312,
      "step": 24355
    },
    {
      "epoch": 14.36320754716981,
      "grad_norm": 4.398914813995361,
      "learning_rate": 2.231139675692308e-06,
      "loss": 0.3317,
      "num_input_tokens_seen": 15787544,
      "step": 24360
    },
    {
      "epoch": 14.366155660377359,
      "grad_norm": 4.164758682250977,
      "learning_rate": 2.228997824832973e-06,
      "loss": 0.3249,
      "num_input_tokens_seen": 15790552,
      "step": 24365
    },
    {
      "epoch": 14.369103773584905,
      "grad_norm": 4.202245235443115,
      "learning_rate": 2.226856707608406e-06,
      "loss": 0.2601,
      "num_input_tokens_seen": 15793944,
      "step": 24370
    },
    {
      "epoch": 14.372051886792454,
      "grad_norm": 2.3453006744384766,
      "learning_rate": 2.2247163245854768e-06,
      "loss": 0.309,
      "num_input_tokens_seen": 15797752,
      "step": 24375
    },
    {
      "epoch": 14.375,
      "grad_norm": 2.507506847381592,
      "learning_rate": 2.222576676330862e-06,
      "loss": 0.3472,
      "num_input_tokens_seen": 15801784,
      "step": 24380
    },
    {
      "epoch": 14.377948113207546,
      "grad_norm": 4.721760272979736,
      "learning_rate": 2.2204377634110403e-06,
      "loss": 0.3089,
      "num_input_tokens_seen": 15805272,
      "step": 24385
    },
    {
      "epoch": 14.380896226415095,
      "grad_norm": 3.1602368354797363,
      "learning_rate": 2.218299586392301e-06,
      "loss": 0.2788,
      "num_input_tokens_seen": 15808312,
      "step": 24390
    },
    {
      "epoch": 14.383844339622641,
      "grad_norm": 3.5271778106689453,
      "learning_rate": 2.2161621458407355e-06,
      "loss": 0.2675,
      "num_input_tokens_seen": 15812056,
      "step": 24395
    },
    {
      "epoch": 14.38679245283019,
      "grad_norm": 2.941889524459839,
      "learning_rate": 2.2140254423222398e-06,
      "loss": 0.2108,
      "num_input_tokens_seen": 15815160,
      "step": 24400
    },
    {
      "epoch": 14.389740566037736,
      "grad_norm": 3.111421823501587,
      "learning_rate": 2.2118894764025146e-06,
      "loss": 0.2805,
      "num_input_tokens_seen": 15818296,
      "step": 24405
    },
    {
      "epoch": 14.392688679245284,
      "grad_norm": 4.779598236083984,
      "learning_rate": 2.2097542486470667e-06,
      "loss": 0.3924,
      "num_input_tokens_seen": 15822200,
      "step": 24410
    },
    {
      "epoch": 14.39563679245283,
      "grad_norm": 3.0399110317230225,
      "learning_rate": 2.207619759621205e-06,
      "loss": 0.2398,
      "num_input_tokens_seen": 15825720,
      "step": 24415
    },
    {
      "epoch": 14.398584905660377,
      "grad_norm": 4.355321407318115,
      "learning_rate": 2.205486009890049e-06,
      "loss": 0.4109,
      "num_input_tokens_seen": 15828760,
      "step": 24420
    },
    {
      "epoch": 14.401533018867925,
      "grad_norm": 4.14400053024292,
      "learning_rate": 2.2033530000185146e-06,
      "loss": 0.3028,
      "num_input_tokens_seen": 15831736,
      "step": 24425
    },
    {
      "epoch": 14.404481132075471,
      "grad_norm": 4.990927696228027,
      "learning_rate": 2.2012207305713244e-06,
      "loss": 0.4057,
      "num_input_tokens_seen": 15834872,
      "step": 24430
    },
    {
      "epoch": 14.40742924528302,
      "grad_norm": 4.92526388168335,
      "learning_rate": 2.19908920211301e-06,
      "loss": 0.3383,
      "num_input_tokens_seen": 15838840,
      "step": 24435
    },
    {
      "epoch": 14.410377358490566,
      "grad_norm": 5.906961441040039,
      "learning_rate": 2.196958415207901e-06,
      "loss": 0.371,
      "num_input_tokens_seen": 15842008,
      "step": 24440
    },
    {
      "epoch": 14.413325471698114,
      "grad_norm": 3.0606441497802734,
      "learning_rate": 2.1948283704201312e-06,
      "loss": 0.3174,
      "num_input_tokens_seen": 15844792,
      "step": 24445
    },
    {
      "epoch": 14.41627358490566,
      "grad_norm": 4.6168646812438965,
      "learning_rate": 2.1926990683136383e-06,
      "loss": 0.3959,
      "num_input_tokens_seen": 15847768,
      "step": 24450
    },
    {
      "epoch": 14.419221698113208,
      "grad_norm": 3.2600436210632324,
      "learning_rate": 2.1905705094521685e-06,
      "loss": 0.2954,
      "num_input_tokens_seen": 15850872,
      "step": 24455
    },
    {
      "epoch": 14.422169811320755,
      "grad_norm": 2.2895495891571045,
      "learning_rate": 2.1884426943992635e-06,
      "loss": 0.3304,
      "num_input_tokens_seen": 15853784,
      "step": 24460
    },
    {
      "epoch": 14.425117924528301,
      "grad_norm": 2.678267478942871,
      "learning_rate": 2.1863156237182727e-06,
      "loss": 0.3914,
      "num_input_tokens_seen": 15856760,
      "step": 24465
    },
    {
      "epoch": 14.42806603773585,
      "grad_norm": 2.3484857082366943,
      "learning_rate": 2.1841892979723466e-06,
      "loss": 0.2663,
      "num_input_tokens_seen": 15859416,
      "step": 24470
    },
    {
      "epoch": 14.431014150943396,
      "grad_norm": 2.8720388412475586,
      "learning_rate": 2.1820637177244375e-06,
      "loss": 0.3157,
      "num_input_tokens_seen": 15862584,
      "step": 24475
    },
    {
      "epoch": 14.433962264150944,
      "grad_norm": 2.6550307273864746,
      "learning_rate": 2.179938883537306e-06,
      "loss": 0.2746,
      "num_input_tokens_seen": 15865624,
      "step": 24480
    },
    {
      "epoch": 14.43691037735849,
      "grad_norm": 4.516193866729736,
      "learning_rate": 2.177814795973508e-06,
      "loss": 0.3401,
      "num_input_tokens_seen": 15868952,
      "step": 24485
    },
    {
      "epoch": 14.439858490566039,
      "grad_norm": 5.736800193786621,
      "learning_rate": 2.1756914555954064e-06,
      "loss": 0.4247,
      "num_input_tokens_seen": 15872184,
      "step": 24490
    },
    {
      "epoch": 14.442806603773585,
      "grad_norm": 2.4329371452331543,
      "learning_rate": 2.173568862965164e-06,
      "loss": 0.2521,
      "num_input_tokens_seen": 15876376,
      "step": 24495
    },
    {
      "epoch": 14.445754716981131,
      "grad_norm": 2.0187172889709473,
      "learning_rate": 2.171447018644746e-06,
      "loss": 0.2701,
      "num_input_tokens_seen": 15879768,
      "step": 24500
    },
    {
      "epoch": 14.44870283018868,
      "grad_norm": 6.138628005981445,
      "learning_rate": 2.1693259231959186e-06,
      "loss": 0.3054,
      "num_input_tokens_seen": 15882392,
      "step": 24505
    },
    {
      "epoch": 14.451650943396226,
      "grad_norm": 6.554543972015381,
      "learning_rate": 2.1672055771802545e-06,
      "loss": 0.2157,
      "num_input_tokens_seen": 15885624,
      "step": 24510
    },
    {
      "epoch": 14.454599056603774,
      "grad_norm": 5.6277546882629395,
      "learning_rate": 2.1650859811591224e-06,
      "loss": 0.3136,
      "num_input_tokens_seen": 15888536,
      "step": 24515
    },
    {
      "epoch": 14.45754716981132,
      "grad_norm": 2.74461030960083,
      "learning_rate": 2.1629671356936943e-06,
      "loss": 0.2744,
      "num_input_tokens_seen": 15891736,
      "step": 24520
    },
    {
      "epoch": 14.460495283018869,
      "grad_norm": 3.745542526245117,
      "learning_rate": 2.1608490413449428e-06,
      "loss": 0.2233,
      "num_input_tokens_seen": 15895128,
      "step": 24525
    },
    {
      "epoch": 14.463443396226415,
      "grad_norm": 4.096607208251953,
      "learning_rate": 2.158731698673645e-06,
      "loss": 0.3128,
      "num_input_tokens_seen": 15897848,
      "step": 24530
    },
    {
      "epoch": 14.466391509433961,
      "grad_norm": 3.9122884273529053,
      "learning_rate": 2.1566151082403752e-06,
      "loss": 0.4451,
      "num_input_tokens_seen": 15901368,
      "step": 24535
    },
    {
      "epoch": 14.46933962264151,
      "grad_norm": 2.7544450759887695,
      "learning_rate": 2.154499270605508e-06,
      "loss": 0.2681,
      "num_input_tokens_seen": 15904600,
      "step": 24540
    },
    {
      "epoch": 14.472287735849056,
      "grad_norm": 2.7703285217285156,
      "learning_rate": 2.1523841863292243e-06,
      "loss": 0.2676,
      "num_input_tokens_seen": 15907992,
      "step": 24545
    },
    {
      "epoch": 14.475235849056604,
      "grad_norm": 3.4356801509857178,
      "learning_rate": 2.1502698559714998e-06,
      "loss": 0.2834,
      "num_input_tokens_seen": 15911960,
      "step": 24550
    },
    {
      "epoch": 14.47818396226415,
      "grad_norm": 4.158863544464111,
      "learning_rate": 2.1481562800921125e-06,
      "loss": 0.3466,
      "num_input_tokens_seen": 15916408,
      "step": 24555
    },
    {
      "epoch": 14.481132075471699,
      "grad_norm": 3.4469988346099854,
      "learning_rate": 2.146043459250641e-06,
      "loss": 0.3399,
      "num_input_tokens_seen": 15920184,
      "step": 24560
    },
    {
      "epoch": 14.484080188679245,
      "grad_norm": 5.154344081878662,
      "learning_rate": 2.1439313940064634e-06,
      "loss": 0.2912,
      "num_input_tokens_seen": 15927320,
      "step": 24565
    },
    {
      "epoch": 14.487028301886792,
      "grad_norm": 5.5301947593688965,
      "learning_rate": 2.141820084918756e-06,
      "loss": 0.2724,
      "num_input_tokens_seen": 15930104,
      "step": 24570
    },
    {
      "epoch": 14.48997641509434,
      "grad_norm": 3.5812456607818604,
      "learning_rate": 2.1397095325465013e-06,
      "loss": 0.3435,
      "num_input_tokens_seen": 15933624,
      "step": 24575
    },
    {
      "epoch": 14.492924528301886,
      "grad_norm": 2.623067617416382,
      "learning_rate": 2.1375997374484754e-06,
      "loss": 0.3243,
      "num_input_tokens_seen": 15938136,
      "step": 24580
    },
    {
      "epoch": 14.495872641509434,
      "grad_norm": 5.436469078063965,
      "learning_rate": 2.1354907001832546e-06,
      "loss": 0.3313,
      "num_input_tokens_seen": 15941112,
      "step": 24585
    },
    {
      "epoch": 14.49882075471698,
      "grad_norm": 3.1061270236968994,
      "learning_rate": 2.133382421309217e-06,
      "loss": 0.307,
      "num_input_tokens_seen": 15944056,
      "step": 24590
    },
    {
      "epoch": 14.501768867924529,
      "grad_norm": 3.4456756114959717,
      "learning_rate": 2.131274901384537e-06,
      "loss": 0.3576,
      "num_input_tokens_seen": 15947352,
      "step": 24595
    },
    {
      "epoch": 14.504716981132075,
      "grad_norm": 8.765913963317871,
      "learning_rate": 2.1291681409671896e-06,
      "loss": 0.337,
      "num_input_tokens_seen": 15949816,
      "step": 24600
    },
    {
      "epoch": 14.507665094339622,
      "grad_norm": 2.6575987339019775,
      "learning_rate": 2.12706214061495e-06,
      "loss": 0.2621,
      "num_input_tokens_seen": 15954296,
      "step": 24605
    },
    {
      "epoch": 14.51061320754717,
      "grad_norm": 6.5773210525512695,
      "learning_rate": 2.124956900885391e-06,
      "loss": 0.2029,
      "num_input_tokens_seen": 15957016,
      "step": 24610
    },
    {
      "epoch": 14.513561320754716,
      "grad_norm": 3.0387487411499023,
      "learning_rate": 2.1228524223358833e-06,
      "loss": 0.3605,
      "num_input_tokens_seen": 15960280,
      "step": 24615
    },
    {
      "epoch": 14.516509433962264,
      "grad_norm": 2.546476125717163,
      "learning_rate": 2.120748705523595e-06,
      "loss": 0.2625,
      "num_input_tokens_seen": 15964184,
      "step": 24620
    },
    {
      "epoch": 14.51945754716981,
      "grad_norm": 2.6448769569396973,
      "learning_rate": 2.1186457510054976e-06,
      "loss": 0.3012,
      "num_input_tokens_seen": 15966776,
      "step": 24625
    },
    {
      "epoch": 14.522405660377359,
      "grad_norm": 2.36069655418396,
      "learning_rate": 2.116543559338355e-06,
      "loss": 0.4573,
      "num_input_tokens_seen": 15970552,
      "step": 24630
    },
    {
      "epoch": 14.525353773584905,
      "grad_norm": 2.2556886672973633,
      "learning_rate": 2.1144421310787305e-06,
      "loss": 0.2549,
      "num_input_tokens_seen": 15974680,
      "step": 24635
    },
    {
      "epoch": 14.528301886792454,
      "grad_norm": 6.147180557250977,
      "learning_rate": 2.11234146678299e-06,
      "loss": 0.3256,
      "num_input_tokens_seen": 15978584,
      "step": 24640
    },
    {
      "epoch": 14.53125,
      "grad_norm": 2.8778529167175293,
      "learning_rate": 2.1102415670072907e-06,
      "loss": 0.2769,
      "num_input_tokens_seen": 15981304,
      "step": 24645
    },
    {
      "epoch": 14.534198113207546,
      "grad_norm": 2.7546610832214355,
      "learning_rate": 2.108142432307591e-06,
      "loss": 0.2169,
      "num_input_tokens_seen": 15985656,
      "step": 24650
    },
    {
      "epoch": 14.537146226415095,
      "grad_norm": 3.1699721813201904,
      "learning_rate": 2.1060440632396456e-06,
      "loss": 0.3452,
      "num_input_tokens_seen": 15989784,
      "step": 24655
    },
    {
      "epoch": 14.540094339622641,
      "grad_norm": 5.3679280281066895,
      "learning_rate": 2.103946460359007e-06,
      "loss": 0.3047,
      "num_input_tokens_seen": 15992920,
      "step": 24660
    },
    {
      "epoch": 14.54304245283019,
      "grad_norm": 3.430447578430176,
      "learning_rate": 2.101849624221022e-06,
      "loss": 0.4849,
      "num_input_tokens_seen": 15997400,
      "step": 24665
    },
    {
      "epoch": 14.545990566037736,
      "grad_norm": 3.409475564956665,
      "learning_rate": 2.0997535553808417e-06,
      "loss": 0.4496,
      "num_input_tokens_seen": 15999864,
      "step": 24670
    },
    {
      "epoch": 14.548938679245284,
      "grad_norm": 2.279515266418457,
      "learning_rate": 2.0976582543934064e-06,
      "loss": 0.2762,
      "num_input_tokens_seen": 16002808,
      "step": 24675
    },
    {
      "epoch": 14.55188679245283,
      "grad_norm": 4.949150085449219,
      "learning_rate": 2.0955637218134573e-06,
      "loss": 0.3417,
      "num_input_tokens_seen": 16005784,
      "step": 24680
    },
    {
      "epoch": 14.554834905660378,
      "grad_norm": 3.304025411605835,
      "learning_rate": 2.09346995819553e-06,
      "loss": 0.3226,
      "num_input_tokens_seen": 16009944,
      "step": 24685
    },
    {
      "epoch": 14.557783018867925,
      "grad_norm": 2.9069314002990723,
      "learning_rate": 2.0913769640939553e-06,
      "loss": 0.2578,
      "num_input_tokens_seen": 16013080,
      "step": 24690
    },
    {
      "epoch": 14.560731132075471,
      "grad_norm": 2.5377485752105713,
      "learning_rate": 2.0892847400628674e-06,
      "loss": 0.2846,
      "num_input_tokens_seen": 16017656,
      "step": 24695
    },
    {
      "epoch": 14.56367924528302,
      "grad_norm": 2.834608316421509,
      "learning_rate": 2.0871932866561885e-06,
      "loss": 0.338,
      "num_input_tokens_seen": 16021080,
      "step": 24700
    },
    {
      "epoch": 14.566627358490566,
      "grad_norm": 8.584307670593262,
      "learning_rate": 2.0851026044276405e-06,
      "loss": 0.328,
      "num_input_tokens_seen": 16023736,
      "step": 24705
    },
    {
      "epoch": 14.569575471698114,
      "grad_norm": 4.27847957611084,
      "learning_rate": 2.083012693930741e-06,
      "loss": 0.3355,
      "num_input_tokens_seen": 16027832,
      "step": 24710
    },
    {
      "epoch": 14.57252358490566,
      "grad_norm": 4.531001091003418,
      "learning_rate": 2.0809235557188e-06,
      "loss": 0.4742,
      "num_input_tokens_seen": 16030456,
      "step": 24715
    },
    {
      "epoch": 14.575471698113208,
      "grad_norm": 2.8990285396575928,
      "learning_rate": 2.0788351903449307e-06,
      "loss": 0.4556,
      "num_input_tokens_seen": 16033048,
      "step": 24720
    },
    {
      "epoch": 14.578419811320755,
      "grad_norm": 2.6075918674468994,
      "learning_rate": 2.0767475983620317e-06,
      "loss": 0.2388,
      "num_input_tokens_seen": 16036536,
      "step": 24725
    },
    {
      "epoch": 14.581367924528301,
      "grad_norm": 2.214679718017578,
      "learning_rate": 2.074660780322806e-06,
      "loss": 0.2747,
      "num_input_tokens_seen": 16039160,
      "step": 24730
    },
    {
      "epoch": 14.58431603773585,
      "grad_norm": 2.6124656200408936,
      "learning_rate": 2.0725747367797473e-06,
      "loss": 0.2848,
      "num_input_tokens_seen": 16042616,
      "step": 24735
    },
    {
      "epoch": 14.587264150943396,
      "grad_norm": 3.3564093112945557,
      "learning_rate": 2.070489468285143e-06,
      "loss": 0.4303,
      "num_input_tokens_seen": 16045400,
      "step": 24740
    },
    {
      "epoch": 14.590212264150944,
      "grad_norm": 3.4219553470611572,
      "learning_rate": 2.068404975391077e-06,
      "loss": 0.4057,
      "num_input_tokens_seen": 16049656,
      "step": 24745
    },
    {
      "epoch": 14.59316037735849,
      "grad_norm": 10.377117156982422,
      "learning_rate": 2.0663212586494293e-06,
      "loss": 0.4115,
      "num_input_tokens_seen": 16051960,
      "step": 24750
    },
    {
      "epoch": 14.596108490566039,
      "grad_norm": 2.495537757873535,
      "learning_rate": 2.064238318611869e-06,
      "loss": 0.3143,
      "num_input_tokens_seen": 16055448,
      "step": 24755
    },
    {
      "epoch": 14.599056603773585,
      "grad_norm": 2.5304958820343018,
      "learning_rate": 2.0621561558298693e-06,
      "loss": 0.3966,
      "num_input_tokens_seen": 16058296,
      "step": 24760
    },
    {
      "epoch": 14.602004716981131,
      "grad_norm": 3.282886028289795,
      "learning_rate": 2.0600747708546877e-06,
      "loss": 0.3345,
      "num_input_tokens_seen": 16061368,
      "step": 24765
    },
    {
      "epoch": 14.60495283018868,
      "grad_norm": 3.0502235889434814,
      "learning_rate": 2.0579941642373814e-06,
      "loss": 0.318,
      "num_input_tokens_seen": 16064408,
      "step": 24770
    },
    {
      "epoch": 14.607900943396226,
      "grad_norm": 4.152999401092529,
      "learning_rate": 2.0559143365287993e-06,
      "loss": 0.3438,
      "num_input_tokens_seen": 16067608,
      "step": 24775
    },
    {
      "epoch": 14.610849056603774,
      "grad_norm": 4.377725124359131,
      "learning_rate": 2.0538352882795846e-06,
      "loss": 0.3293,
      "num_input_tokens_seen": 16070808,
      "step": 24780
    },
    {
      "epoch": 14.61379716981132,
      "grad_norm": 1.6774953603744507,
      "learning_rate": 2.051757020040173e-06,
      "loss": 0.295,
      "num_input_tokens_seen": 16073784,
      "step": 24785
    },
    {
      "epoch": 14.616745283018869,
      "grad_norm": 5.085612773895264,
      "learning_rate": 2.0496795323607983e-06,
      "loss": 0.3699,
      "num_input_tokens_seen": 16076888,
      "step": 24790
    },
    {
      "epoch": 14.619693396226415,
      "grad_norm": 4.488544940948486,
      "learning_rate": 2.0476028257914825e-06,
      "loss": 0.265,
      "num_input_tokens_seen": 16079672,
      "step": 24795
    },
    {
      "epoch": 14.622641509433961,
      "grad_norm": 4.12568473815918,
      "learning_rate": 2.0455269008820433e-06,
      "loss": 0.2836,
      "num_input_tokens_seen": 16083096,
      "step": 24800
    },
    {
      "epoch": 14.62558962264151,
      "grad_norm": 4.185266494750977,
      "learning_rate": 2.0434517581820893e-06,
      "loss": 0.3001,
      "num_input_tokens_seen": 16086904,
      "step": 24805
    },
    {
      "epoch": 14.628537735849056,
      "grad_norm": 3.6816823482513428,
      "learning_rate": 2.041377398241025e-06,
      "loss": 0.2848,
      "num_input_tokens_seen": 16090008,
      "step": 24810
    },
    {
      "epoch": 14.631485849056604,
      "grad_norm": 3.2702667713165283,
      "learning_rate": 2.0393038216080433e-06,
      "loss": 0.2598,
      "num_input_tokens_seen": 16092632,
      "step": 24815
    },
    {
      "epoch": 14.63443396226415,
      "grad_norm": 3.7423501014709473,
      "learning_rate": 2.037231028832135e-06,
      "loss": 0.3766,
      "num_input_tokens_seen": 16097432,
      "step": 24820
    },
    {
      "epoch": 14.637382075471699,
      "grad_norm": 4.25266170501709,
      "learning_rate": 2.0351590204620823e-06,
      "loss": 0.3524,
      "num_input_tokens_seen": 16100408,
      "step": 24825
    },
    {
      "epoch": 14.640330188679245,
      "grad_norm": 2.923633098602295,
      "learning_rate": 2.033087797046457e-06,
      "loss": 0.2243,
      "num_input_tokens_seen": 16103992,
      "step": 24830
    },
    {
      "epoch": 14.643278301886792,
      "grad_norm": 2.7631185054779053,
      "learning_rate": 2.031017359133624e-06,
      "loss": 0.2736,
      "num_input_tokens_seen": 16106904,
      "step": 24835
    },
    {
      "epoch": 14.64622641509434,
      "grad_norm": 2.9468002319335938,
      "learning_rate": 2.0289477072717406e-06,
      "loss": 0.2739,
      "num_input_tokens_seen": 16109560,
      "step": 24840
    },
    {
      "epoch": 14.649174528301886,
      "grad_norm": 3.673149347305298,
      "learning_rate": 2.026878842008756e-06,
      "loss": 0.4379,
      "num_input_tokens_seen": 16113112,
      "step": 24845
    },
    {
      "epoch": 14.652122641509434,
      "grad_norm": 2.971555709838867,
      "learning_rate": 2.0248107638924105e-06,
      "loss": 0.2552,
      "num_input_tokens_seen": 16115608,
      "step": 24850
    },
    {
      "epoch": 14.65507075471698,
      "grad_norm": 3.931057929992676,
      "learning_rate": 2.0227434734702386e-06,
      "loss": 0.2856,
      "num_input_tokens_seen": 16118904,
      "step": 24855
    },
    {
      "epoch": 14.658018867924529,
      "grad_norm": 3.615086078643799,
      "learning_rate": 2.020676971289563e-06,
      "loss": 0.3514,
      "num_input_tokens_seen": 16122136,
      "step": 24860
    },
    {
      "epoch": 14.660966981132075,
      "grad_norm": 4.416012287139893,
      "learning_rate": 2.0186112578975005e-06,
      "loss": 0.3686,
      "num_input_tokens_seen": 16124952,
      "step": 24865
    },
    {
      "epoch": 14.663915094339622,
      "grad_norm": 5.290577411651611,
      "learning_rate": 2.016546333840956e-06,
      "loss": 0.2994,
      "num_input_tokens_seen": 16127576,
      "step": 24870
    },
    {
      "epoch": 14.66686320754717,
      "grad_norm": 4.800883769989014,
      "learning_rate": 2.014482199666627e-06,
      "loss": 0.315,
      "num_input_tokens_seen": 16130328,
      "step": 24875
    },
    {
      "epoch": 14.669811320754716,
      "grad_norm": 2.430908679962158,
      "learning_rate": 2.0124188559210017e-06,
      "loss": 0.2423,
      "num_input_tokens_seen": 16133368,
      "step": 24880
    },
    {
      "epoch": 14.672759433962264,
      "grad_norm": 3.04427433013916,
      "learning_rate": 2.0103563031503613e-06,
      "loss": 0.3501,
      "num_input_tokens_seen": 16137848,
      "step": 24885
    },
    {
      "epoch": 14.67570754716981,
      "grad_norm": 5.610107898712158,
      "learning_rate": 2.0082945419007745e-06,
      "loss": 0.4026,
      "num_input_tokens_seen": 16141912,
      "step": 24890
    },
    {
      "epoch": 14.678655660377359,
      "grad_norm": 2.52652907371521,
      "learning_rate": 2.0062335727181007e-06,
      "loss": 0.3409,
      "num_input_tokens_seen": 16145240,
      "step": 24895
    },
    {
      "epoch": 14.681603773584905,
      "grad_norm": 4.356776237487793,
      "learning_rate": 2.004173396147992e-06,
      "loss": 0.2761,
      "num_input_tokens_seen": 16148216,
      "step": 24900
    },
    {
      "epoch": 14.684551886792454,
      "grad_norm": 3.3060922622680664,
      "learning_rate": 2.0021140127358873e-06,
      "loss": 0.3456,
      "num_input_tokens_seen": 16151960,
      "step": 24905
    },
    {
      "epoch": 14.6875,
      "grad_norm": 3.6769204139709473,
      "learning_rate": 2.0000554230270164e-06,
      "loss": 0.3726,
      "num_input_tokens_seen": 16155832,
      "step": 24910
    },
    {
      "epoch": 14.690448113207546,
      "grad_norm": 3.1186230182647705,
      "learning_rate": 1.997997627566401e-06,
      "loss": 0.398,
      "num_input_tokens_seen": 16158584,
      "step": 24915
    },
    {
      "epoch": 14.693396226415095,
      "grad_norm": 2.296654224395752,
      "learning_rate": 1.9959406268988536e-06,
      "loss": 0.296,
      "num_input_tokens_seen": 16162232,
      "step": 24920
    },
    {
      "epoch": 14.696344339622641,
      "grad_norm": 2.71647047996521,
      "learning_rate": 1.9938844215689717e-06,
      "loss": 0.3164,
      "num_input_tokens_seen": 16165528,
      "step": 24925
    },
    {
      "epoch": 14.69929245283019,
      "grad_norm": 3.2003798484802246,
      "learning_rate": 1.991829012121145e-06,
      "loss": 0.2607,
      "num_input_tokens_seen": 16168024,
      "step": 24930
    },
    {
      "epoch": 14.702240566037736,
      "grad_norm": 2.1873128414154053,
      "learning_rate": 1.989774399099552e-06,
      "loss": 0.3719,
      "num_input_tokens_seen": 16171640,
      "step": 24935
    },
    {
      "epoch": 14.705188679245284,
      "grad_norm": 1.7377322912216187,
      "learning_rate": 1.98772058304816e-06,
      "loss": 0.3275,
      "num_input_tokens_seen": 16174712,
      "step": 24940
    },
    {
      "epoch": 14.70813679245283,
      "grad_norm": 2.2202272415161133,
      "learning_rate": 1.9856675645107244e-06,
      "loss": 0.4372,
      "num_input_tokens_seen": 16178136,
      "step": 24945
    },
    {
      "epoch": 14.711084905660378,
      "grad_norm": 3.245070219039917,
      "learning_rate": 1.9836153440307936e-06,
      "loss": 0.3204,
      "num_input_tokens_seen": 16181624,
      "step": 24950
    },
    {
      "epoch": 14.714033018867925,
      "grad_norm": 1.7662087678909302,
      "learning_rate": 1.9815639221517002e-06,
      "loss": 0.2398,
      "num_input_tokens_seen": 16184280,
      "step": 24955
    },
    {
      "epoch": 14.716981132075471,
      "grad_norm": 2.7901220321655273,
      "learning_rate": 1.9795132994165673e-06,
      "loss": 0.307,
      "num_input_tokens_seen": 16186968,
      "step": 24960
    },
    {
      "epoch": 14.71992924528302,
      "grad_norm": 4.7065749168396,
      "learning_rate": 1.977463476368306e-06,
      "loss": 0.3917,
      "num_input_tokens_seen": 16189848,
      "step": 24965
    },
    {
      "epoch": 14.722877358490566,
      "grad_norm": 2.6094372272491455,
      "learning_rate": 1.975414453549614e-06,
      "loss": 0.2668,
      "num_input_tokens_seen": 16192568,
      "step": 24970
    },
    {
      "epoch": 14.725825471698114,
      "grad_norm": 3.654968738555908,
      "learning_rate": 1.9733662315029826e-06,
      "loss": 0.3284,
      "num_input_tokens_seen": 16195224,
      "step": 24975
    },
    {
      "epoch": 14.72877358490566,
      "grad_norm": 5.805337429046631,
      "learning_rate": 1.9713188107706856e-06,
      "loss": 0.3838,
      "num_input_tokens_seen": 16197528,
      "step": 24980
    },
    {
      "epoch": 14.731721698113208,
      "grad_norm": 3.806828737258911,
      "learning_rate": 1.969272191894786e-06,
      "loss": 0.3184,
      "num_input_tokens_seen": 16200728,
      "step": 24985
    },
    {
      "epoch": 14.734669811320755,
      "grad_norm": 2.3026344776153564,
      "learning_rate": 1.967226375417135e-06,
      "loss": 0.4565,
      "num_input_tokens_seen": 16204248,
      "step": 24990
    },
    {
      "epoch": 14.737617924528301,
      "grad_norm": 2.3539252281188965,
      "learning_rate": 1.965181361879372e-06,
      "loss": 0.2488,
      "num_input_tokens_seen": 16208664,
      "step": 24995
    },
    {
      "epoch": 14.74056603773585,
      "grad_norm": 3.1282501220703125,
      "learning_rate": 1.9631371518229214e-06,
      "loss": 0.4672,
      "num_input_tokens_seen": 16215064,
      "step": 25000
    },
    {
      "epoch": 14.743514150943396,
      "grad_norm": 2.0518414974212646,
      "learning_rate": 1.9610937457889975e-06,
      "loss": 0.2411,
      "num_input_tokens_seen": 16217688,
      "step": 25005
    },
    {
      "epoch": 14.746462264150944,
      "grad_norm": 5.06934118270874,
      "learning_rate": 1.9590511443186032e-06,
      "loss": 0.3075,
      "num_input_tokens_seen": 16220408,
      "step": 25010
    },
    {
      "epoch": 14.74941037735849,
      "grad_norm": 4.129754066467285,
      "learning_rate": 1.9570093479525243e-06,
      "loss": 0.3344,
      "num_input_tokens_seen": 16222712,
      "step": 25015
    },
    {
      "epoch": 14.752358490566039,
      "grad_norm": 4.976624488830566,
      "learning_rate": 1.954968357231335e-06,
      "loss": 0.2823,
      "num_input_tokens_seen": 16225272,
      "step": 25020
    },
    {
      "epoch": 14.755306603773585,
      "grad_norm": 6.024097919464111,
      "learning_rate": 1.9529281726953964e-06,
      "loss": 0.4162,
      "num_input_tokens_seen": 16228056,
      "step": 25025
    },
    {
      "epoch": 14.758254716981131,
      "grad_norm": 3.0294981002807617,
      "learning_rate": 1.9508887948848564e-06,
      "loss": 0.4596,
      "num_input_tokens_seen": 16232312,
      "step": 25030
    },
    {
      "epoch": 14.76120283018868,
      "grad_norm": 5.637691497802734,
      "learning_rate": 1.9488502243396475e-06,
      "loss": 0.2734,
      "num_input_tokens_seen": 16235544,
      "step": 25035
    },
    {
      "epoch": 14.764150943396226,
      "grad_norm": 3.0754523277282715,
      "learning_rate": 1.946812461599492e-06,
      "loss": 0.2865,
      "num_input_tokens_seen": 16238712,
      "step": 25040
    },
    {
      "epoch": 14.767099056603774,
      "grad_norm": 4.027085304260254,
      "learning_rate": 1.944775507203897e-06,
      "loss": 0.3474,
      "num_input_tokens_seen": 16242040,
      "step": 25045
    },
    {
      "epoch": 14.77004716981132,
      "grad_norm": 3.877185344696045,
      "learning_rate": 1.942739361692153e-06,
      "loss": 0.3703,
      "num_input_tokens_seen": 16244536,
      "step": 25050
    },
    {
      "epoch": 14.772995283018869,
      "grad_norm": 3.7695186138153076,
      "learning_rate": 1.94070402560334e-06,
      "loss": 0.3802,
      "num_input_tokens_seen": 16247000,
      "step": 25055
    },
    {
      "epoch": 14.775943396226415,
      "grad_norm": 2.250601053237915,
      "learning_rate": 1.93866949947632e-06,
      "loss": 0.2615,
      "num_input_tokens_seen": 16250456,
      "step": 25060
    },
    {
      "epoch": 14.778891509433961,
      "grad_norm": 3.6380181312561035,
      "learning_rate": 1.9366357838497423e-06,
      "loss": 0.3718,
      "num_input_tokens_seen": 16253880,
      "step": 25065
    },
    {
      "epoch": 14.78183962264151,
      "grad_norm": 3.9778547286987305,
      "learning_rate": 1.9346028792620454e-06,
      "loss": 0.307,
      "num_input_tokens_seen": 16257208,
      "step": 25070
    },
    {
      "epoch": 14.784787735849056,
      "grad_norm": 3.691209554672241,
      "learning_rate": 1.9325707862514464e-06,
      "loss": 0.3344,
      "num_input_tokens_seen": 16260344,
      "step": 25075
    },
    {
      "epoch": 14.787735849056604,
      "grad_norm": 3.8926842212677,
      "learning_rate": 1.930539505355952e-06,
      "loss": 0.3476,
      "num_input_tokens_seen": 16263480,
      "step": 25080
    },
    {
      "epoch": 14.79068396226415,
      "grad_norm": 7.206007480621338,
      "learning_rate": 1.9285090371133524e-06,
      "loss": 0.3561,
      "num_input_tokens_seen": 16266296,
      "step": 25085
    },
    {
      "epoch": 14.793632075471699,
      "grad_norm": 1.8304483890533447,
      "learning_rate": 1.9264793820612228e-06,
      "loss": 0.3467,
      "num_input_tokens_seen": 16269944,
      "step": 25090
    },
    {
      "epoch": 14.796580188679245,
      "grad_norm": 3.410874605178833,
      "learning_rate": 1.924450540736921e-06,
      "loss": 0.3542,
      "num_input_tokens_seen": 16273848,
      "step": 25095
    },
    {
      "epoch": 14.799528301886792,
      "grad_norm": 4.209432601928711,
      "learning_rate": 1.922422513677593e-06,
      "loss": 0.3059,
      "num_input_tokens_seen": 16276696,
      "step": 25100
    },
    {
      "epoch": 14.80247641509434,
      "grad_norm": 3.2771594524383545,
      "learning_rate": 1.9203953014201703e-06,
      "loss": 0.2968,
      "num_input_tokens_seen": 16279832,
      "step": 25105
    },
    {
      "epoch": 14.805424528301886,
      "grad_norm": 4.370633125305176,
      "learning_rate": 1.918368904501364e-06,
      "loss": 0.4121,
      "num_input_tokens_seen": 16283000,
      "step": 25110
    },
    {
      "epoch": 14.808372641509434,
      "grad_norm": 4.456775665283203,
      "learning_rate": 1.9163433234576713e-06,
      "loss": 0.3759,
      "num_input_tokens_seen": 16285688,
      "step": 25115
    },
    {
      "epoch": 14.81132075471698,
      "grad_norm": 3.614872455596924,
      "learning_rate": 1.9143185588253733e-06,
      "loss": 0.313,
      "num_input_tokens_seen": 16288952,
      "step": 25120
    },
    {
      "epoch": 14.814268867924529,
      "grad_norm": 2.304779529571533,
      "learning_rate": 1.9122946111405354e-06,
      "loss": 0.3775,
      "num_input_tokens_seen": 16291992,
      "step": 25125
    },
    {
      "epoch": 14.817216981132075,
      "grad_norm": 6.34199857711792,
      "learning_rate": 1.910271480939005e-06,
      "loss": 0.3416,
      "num_input_tokens_seen": 16295192,
      "step": 25130
    },
    {
      "epoch": 14.820165094339622,
      "grad_norm": 4.498339653015137,
      "learning_rate": 1.9082491687564176e-06,
      "loss": 0.3966,
      "num_input_tokens_seen": 16297816,
      "step": 25135
    },
    {
      "epoch": 14.82311320754717,
      "grad_norm": 4.41160774230957,
      "learning_rate": 1.9062276751281872e-06,
      "loss": 0.4432,
      "num_input_tokens_seen": 16300280,
      "step": 25140
    },
    {
      "epoch": 14.826061320754716,
      "grad_norm": 4.732951641082764,
      "learning_rate": 1.9042070005895136e-06,
      "loss": 0.3356,
      "num_input_tokens_seen": 16302808,
      "step": 25145
    },
    {
      "epoch": 14.829009433962264,
      "grad_norm": 3.174856662750244,
      "learning_rate": 1.9021871456753788e-06,
      "loss": 0.2687,
      "num_input_tokens_seen": 16306008,
      "step": 25150
    },
    {
      "epoch": 14.83195754716981,
      "grad_norm": 2.9012668132781982,
      "learning_rate": 1.9001681109205478e-06,
      "loss": 0.3671,
      "num_input_tokens_seen": 16309400,
      "step": 25155
    },
    {
      "epoch": 14.834905660377359,
      "grad_norm": 2.6859686374664307,
      "learning_rate": 1.898149896859567e-06,
      "loss": 0.3652,
      "num_input_tokens_seen": 16312952,
      "step": 25160
    },
    {
      "epoch": 14.837853773584905,
      "grad_norm": 2.681326150894165,
      "learning_rate": 1.8961325040267714e-06,
      "loss": 0.219,
      "num_input_tokens_seen": 16315800,
      "step": 25165
    },
    {
      "epoch": 14.840801886792454,
      "grad_norm": 2.1208364963531494,
      "learning_rate": 1.894115932956272e-06,
      "loss": 0.3133,
      "num_input_tokens_seen": 16319256,
      "step": 25170
    },
    {
      "epoch": 14.84375,
      "grad_norm": 9.923542022705078,
      "learning_rate": 1.8921001841819652e-06,
      "loss": 0.2856,
      "num_input_tokens_seen": 16322040,
      "step": 25175
    },
    {
      "epoch": 14.846698113207546,
      "grad_norm": 3.9546620845794678,
      "learning_rate": 1.8900852582375284e-06,
      "loss": 0.2856,
      "num_input_tokens_seen": 16324664,
      "step": 25180
    },
    {
      "epoch": 14.849646226415095,
      "grad_norm": 3.6156482696533203,
      "learning_rate": 1.8880711556564214e-06,
      "loss": 0.2504,
      "num_input_tokens_seen": 16327064,
      "step": 25185
    },
    {
      "epoch": 14.852594339622641,
      "grad_norm": 3.439997434616089,
      "learning_rate": 1.8860578769718891e-06,
      "loss": 0.326,
      "num_input_tokens_seen": 16329688,
      "step": 25190
    },
    {
      "epoch": 14.85554245283019,
      "grad_norm": 3.2040982246398926,
      "learning_rate": 1.8840454227169525e-06,
      "loss": 0.3346,
      "num_input_tokens_seen": 16332632,
      "step": 25195
    },
    {
      "epoch": 14.858490566037736,
      "grad_norm": 2.231311082839966,
      "learning_rate": 1.882033793424421e-06,
      "loss": 0.346,
      "num_input_tokens_seen": 16335768,
      "step": 25200
    },
    {
      "epoch": 14.861438679245284,
      "grad_norm": 2.3547816276550293,
      "learning_rate": 1.88002298962688e-06,
      "loss": 0.2623,
      "num_input_tokens_seen": 16338200,
      "step": 25205
    },
    {
      "epoch": 14.86438679245283,
      "grad_norm": 16.819061279296875,
      "learning_rate": 1.8780130118566996e-06,
      "loss": 0.4121,
      "num_input_tokens_seen": 16341336,
      "step": 25210
    },
    {
      "epoch": 14.867334905660378,
      "grad_norm": 2.9588966369628906,
      "learning_rate": 1.876003860646029e-06,
      "loss": 0.3828,
      "num_input_tokens_seen": 16344728,
      "step": 25215
    },
    {
      "epoch": 14.870283018867925,
      "grad_norm": 4.993794918060303,
      "learning_rate": 1.8739955365267997e-06,
      "loss": 0.4281,
      "num_input_tokens_seen": 16347672,
      "step": 25220
    },
    {
      "epoch": 14.873231132075471,
      "grad_norm": 4.128047943115234,
      "learning_rate": 1.8719880400307228e-06,
      "loss": 0.2178,
      "num_input_tokens_seen": 16351608,
      "step": 25225
    },
    {
      "epoch": 14.87617924528302,
      "grad_norm": 3.867133140563965,
      "learning_rate": 1.869981371689295e-06,
      "loss": 0.3481,
      "num_input_tokens_seen": 16354968,
      "step": 25230
    },
    {
      "epoch": 14.879127358490566,
      "grad_norm": 3.6296133995056152,
      "learning_rate": 1.867975532033789e-06,
      "loss": 0.3448,
      "num_input_tokens_seen": 16357656,
      "step": 25235
    },
    {
      "epoch": 14.882075471698114,
      "grad_norm": 4.999349594116211,
      "learning_rate": 1.8659705215952589e-06,
      "loss": 0.3128,
      "num_input_tokens_seen": 16360504,
      "step": 25240
    },
    {
      "epoch": 14.88502358490566,
      "grad_norm": 3.692695140838623,
      "learning_rate": 1.8639663409045405e-06,
      "loss": 0.3047,
      "num_input_tokens_seen": 16363960,
      "step": 25245
    },
    {
      "epoch": 14.887971698113208,
      "grad_norm": 7.817957401275635,
      "learning_rate": 1.8619629904922466e-06,
      "loss": 0.291,
      "num_input_tokens_seen": 16367384,
      "step": 25250
    },
    {
      "epoch": 14.890919811320755,
      "grad_norm": 3.6027727127075195,
      "learning_rate": 1.859960470888777e-06,
      "loss": 0.3324,
      "num_input_tokens_seen": 16371032,
      "step": 25255
    },
    {
      "epoch": 14.893867924528301,
      "grad_norm": 6.958672523498535,
      "learning_rate": 1.857958782624306e-06,
      "loss": 0.3295,
      "num_input_tokens_seen": 16373944,
      "step": 25260
    },
    {
      "epoch": 14.89681603773585,
      "grad_norm": 2.4314541816711426,
      "learning_rate": 1.8559579262287886e-06,
      "loss": 0.2927,
      "num_input_tokens_seen": 16376952,
      "step": 25265
    },
    {
      "epoch": 14.899764150943396,
      "grad_norm": 2.971811056137085,
      "learning_rate": 1.8539579022319599e-06,
      "loss": 0.321,
      "num_input_tokens_seen": 16379800,
      "step": 25270
    },
    {
      "epoch": 14.902712264150944,
      "grad_norm": 3.739137649536133,
      "learning_rate": 1.8519587111633357e-06,
      "loss": 0.216,
      "num_input_tokens_seen": 16382616,
      "step": 25275
    },
    {
      "epoch": 14.90566037735849,
      "grad_norm": 2.6922943592071533,
      "learning_rate": 1.8499603535522082e-06,
      "loss": 0.3342,
      "num_input_tokens_seen": 16385368,
      "step": 25280
    },
    {
      "epoch": 14.908608490566039,
      "grad_norm": 3.7466487884521484,
      "learning_rate": 1.8479628299276543e-06,
      "loss": 0.3999,
      "num_input_tokens_seen": 16387480,
      "step": 25285
    },
    {
      "epoch": 14.911556603773585,
      "grad_norm": 3.461775302886963,
      "learning_rate": 1.8459661408185241e-06,
      "loss": 0.511,
      "num_input_tokens_seen": 16390680,
      "step": 25290
    },
    {
      "epoch": 14.914504716981131,
      "grad_norm": 4.695166110992432,
      "learning_rate": 1.8439702867534536e-06,
      "loss": 0.3252,
      "num_input_tokens_seen": 16393816,
      "step": 25295
    },
    {
      "epoch": 14.91745283018868,
      "grad_norm": 4.339036464691162,
      "learning_rate": 1.841975268260851e-06,
      "loss": 0.2765,
      "num_input_tokens_seen": 16396408,
      "step": 25300
    },
    {
      "epoch": 14.920400943396226,
      "grad_norm": 3.118687391281128,
      "learning_rate": 1.8399810858689066e-06,
      "loss": 0.3901,
      "num_input_tokens_seen": 16399064,
      "step": 25305
    },
    {
      "epoch": 14.923349056603774,
      "grad_norm": 3.5195493698120117,
      "learning_rate": 1.8379877401055884e-06,
      "loss": 0.4184,
      "num_input_tokens_seen": 16402200,
      "step": 25310
    },
    {
      "epoch": 14.92629716981132,
      "grad_norm": 3.728576898574829,
      "learning_rate": 1.8359952314986418e-06,
      "loss": 0.3333,
      "num_input_tokens_seen": 16405368,
      "step": 25315
    },
    {
      "epoch": 14.929245283018869,
      "grad_norm": 2.7341785430908203,
      "learning_rate": 1.8340035605755957e-06,
      "loss": 0.3684,
      "num_input_tokens_seen": 16408664,
      "step": 25320
    },
    {
      "epoch": 14.932193396226415,
      "grad_norm": 8.341533660888672,
      "learning_rate": 1.8320127278637518e-06,
      "loss": 0.2853,
      "num_input_tokens_seen": 16411224,
      "step": 25325
    },
    {
      "epoch": 14.935141509433961,
      "grad_norm": 6.173079013824463,
      "learning_rate": 1.830022733890191e-06,
      "loss": 0.3783,
      "num_input_tokens_seen": 16414264,
      "step": 25330
    },
    {
      "epoch": 14.93808962264151,
      "grad_norm": 2.3037502765655518,
      "learning_rate": 1.8280335791817733e-06,
      "loss": 0.3877,
      "num_input_tokens_seen": 16416696,
      "step": 25335
    },
    {
      "epoch": 14.941037735849056,
      "grad_norm": 3.3357090950012207,
      "learning_rate": 1.826045264265136e-06,
      "loss": 0.3077,
      "num_input_tokens_seen": 16420312,
      "step": 25340
    },
    {
      "epoch": 14.943985849056604,
      "grad_norm": 2.6335644721984863,
      "learning_rate": 1.8240577896666928e-06,
      "loss": 0.5305,
      "num_input_tokens_seen": 16423928,
      "step": 25345
    },
    {
      "epoch": 14.94693396226415,
      "grad_norm": 3.3129632472991943,
      "learning_rate": 1.8220711559126382e-06,
      "loss": 0.1836,
      "num_input_tokens_seen": 16427032,
      "step": 25350
    },
    {
      "epoch": 14.949882075471699,
      "grad_norm": 3.680108070373535,
      "learning_rate": 1.8200853635289417e-06,
      "loss": 0.4247,
      "num_input_tokens_seen": 16429496,
      "step": 25355
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 3.7051026821136475,
      "learning_rate": 1.81810041304135e-06,
      "loss": 0.2799,
      "num_input_tokens_seen": 16432792,
      "step": 25360
    },
    {
      "epoch": 14.955778301886792,
      "grad_norm": 3.744547128677368,
      "learning_rate": 1.8161163049753865e-06,
      "loss": 0.3903,
      "num_input_tokens_seen": 16436216,
      "step": 25365
    },
    {
      "epoch": 14.95872641509434,
      "grad_norm": 5.737813949584961,
      "learning_rate": 1.8141330398563533e-06,
      "loss": 0.3335,
      "num_input_tokens_seen": 16439192,
      "step": 25370
    },
    {
      "epoch": 14.961674528301886,
      "grad_norm": 3.1514954566955566,
      "learning_rate": 1.8121506182093268e-06,
      "loss": 0.354,
      "num_input_tokens_seen": 16442968,
      "step": 25375
    },
    {
      "epoch": 14.964622641509434,
      "grad_norm": 4.585535049438477,
      "learning_rate": 1.8101690405591643e-06,
      "loss": 0.3614,
      "num_input_tokens_seen": 16446456,
      "step": 25380
    },
    {
      "epoch": 14.96757075471698,
      "grad_norm": 2.5221903324127197,
      "learning_rate": 1.8081883074304945e-06,
      "loss": 0.3563,
      "num_input_tokens_seen": 16449368,
      "step": 25385
    },
    {
      "epoch": 14.970518867924529,
      "grad_norm": 5.096269607543945,
      "learning_rate": 1.8062084193477275e-06,
      "loss": 0.257,
      "num_input_tokens_seen": 16453432,
      "step": 25390
    },
    {
      "epoch": 14.973466981132075,
      "grad_norm": 5.983310222625732,
      "learning_rate": 1.804229376835046e-06,
      "loss": 0.3215,
      "num_input_tokens_seen": 16457112,
      "step": 25395
    },
    {
      "epoch": 14.976415094339622,
      "grad_norm": 3.554180860519409,
      "learning_rate": 1.8022511804164105e-06,
      "loss": 0.3527,
      "num_input_tokens_seen": 16460024,
      "step": 25400
    },
    {
      "epoch": 14.97936320754717,
      "grad_norm": 6.008820533752441,
      "learning_rate": 1.8002738306155559e-06,
      "loss": 0.3184,
      "num_input_tokens_seen": 16463192,
      "step": 25405
    },
    {
      "epoch": 14.982311320754716,
      "grad_norm": 3.335423707962036,
      "learning_rate": 1.7982973279559935e-06,
      "loss": 0.2711,
      "num_input_tokens_seen": 16465912,
      "step": 25410
    },
    {
      "epoch": 14.985259433962264,
      "grad_norm": 1.990943193435669,
      "learning_rate": 1.7963216729610134e-06,
      "loss": 0.3002,
      "num_input_tokens_seen": 16469944,
      "step": 25415
    },
    {
      "epoch": 14.98820754716981,
      "grad_norm": 6.2117486000061035,
      "learning_rate": 1.7943468661536773e-06,
      "loss": 0.2631,
      "num_input_tokens_seen": 16472088,
      "step": 25420
    },
    {
      "epoch": 14.991155660377359,
      "grad_norm": 3.608092784881592,
      "learning_rate": 1.7923729080568242e-06,
      "loss": 0.2628,
      "num_input_tokens_seen": 16474648,
      "step": 25425
    },
    {
      "epoch": 14.994103773584905,
      "grad_norm": 3.464495897293091,
      "learning_rate": 1.7903997991930683e-06,
      "loss": 0.3301,
      "num_input_tokens_seen": 16478136,
      "step": 25430
    },
    {
      "epoch": 14.997051886792454,
      "grad_norm": 2.572916030883789,
      "learning_rate": 1.7884275400847972e-06,
      "loss": 0.2977,
      "num_input_tokens_seen": 16481560,
      "step": 25435
    },
    {
      "epoch": 15.0,
      "grad_norm": 2.014467716217041,
      "learning_rate": 1.786456131254175e-06,
      "loss": 0.2977,
      "num_input_tokens_seen": 16484352,
      "step": 25440
    },
    {
      "epoch": 15.002948113207546,
      "grad_norm": 2.363060712814331,
      "learning_rate": 1.784485573223143e-06,
      "loss": 0.3424,
      "num_input_tokens_seen": 16487424,
      "step": 25445
    },
    {
      "epoch": 15.005896226415095,
      "grad_norm": 3.582559585571289,
      "learning_rate": 1.782515866513414e-06,
      "loss": 0.281,
      "num_input_tokens_seen": 16489952,
      "step": 25450
    },
    {
      "epoch": 15.008844339622641,
      "grad_norm": 8.8089599609375,
      "learning_rate": 1.7805470116464758e-06,
      "loss": 0.2429,
      "num_input_tokens_seen": 16494240,
      "step": 25455
    },
    {
      "epoch": 15.01179245283019,
      "grad_norm": 2.6631906032562256,
      "learning_rate": 1.7785790091435911e-06,
      "loss": 0.3317,
      "num_input_tokens_seen": 16497696,
      "step": 25460
    },
    {
      "epoch": 15.014740566037736,
      "grad_norm": 5.102656841278076,
      "learning_rate": 1.776611859525796e-06,
      "loss": 0.3829,
      "num_input_tokens_seen": 16500928,
      "step": 25465
    },
    {
      "epoch": 15.017688679245284,
      "grad_norm": 7.4769463539123535,
      "learning_rate": 1.7746455633139042e-06,
      "loss": 0.3942,
      "num_input_tokens_seen": 16503424,
      "step": 25470
    },
    {
      "epoch": 15.02063679245283,
      "grad_norm": 3.878669261932373,
      "learning_rate": 1.7726801210285005e-06,
      "loss": 0.2543,
      "num_input_tokens_seen": 16506560,
      "step": 25475
    },
    {
      "epoch": 15.023584905660377,
      "grad_norm": 2.799777030944824,
      "learning_rate": 1.7707155331899418e-06,
      "loss": 0.2198,
      "num_input_tokens_seen": 16510464,
      "step": 25480
    },
    {
      "epoch": 15.026533018867925,
      "grad_norm": 3.540792465209961,
      "learning_rate": 1.7687518003183645e-06,
      "loss": 0.2918,
      "num_input_tokens_seen": 16513312,
      "step": 25485
    },
    {
      "epoch": 15.029481132075471,
      "grad_norm": 4.596673965454102,
      "learning_rate": 1.766788922933675e-06,
      "loss": 0.2406,
      "num_input_tokens_seen": 16516288,
      "step": 25490
    },
    {
      "epoch": 15.03242924528302,
      "grad_norm": 3.4566328525543213,
      "learning_rate": 1.7648269015555514e-06,
      "loss": 0.3077,
      "num_input_tokens_seen": 16519776,
      "step": 25495
    },
    {
      "epoch": 15.035377358490566,
      "grad_norm": 3.397101402282715,
      "learning_rate": 1.7628657367034474e-06,
      "loss": 0.3105,
      "num_input_tokens_seen": 16522240,
      "step": 25500
    },
    {
      "epoch": 15.038325471698114,
      "grad_norm": 4.1300811767578125,
      "learning_rate": 1.7609054288965922e-06,
      "loss": 0.4418,
      "num_input_tokens_seen": 16525216,
      "step": 25505
    },
    {
      "epoch": 15.04127358490566,
      "grad_norm": 3.5053930282592773,
      "learning_rate": 1.7589459786539847e-06,
      "loss": 0.2512,
      "num_input_tokens_seen": 16528960,
      "step": 25510
    },
    {
      "epoch": 15.044221698113208,
      "grad_norm": 3.9951024055480957,
      "learning_rate": 1.7569873864943975e-06,
      "loss": 0.2418,
      "num_input_tokens_seen": 16531712,
      "step": 25515
    },
    {
      "epoch": 15.047169811320755,
      "grad_norm": 3.4283318519592285,
      "learning_rate": 1.7550296529363764e-06,
      "loss": 0.352,
      "num_input_tokens_seen": 16535168,
      "step": 25520
    },
    {
      "epoch": 15.050117924528301,
      "grad_norm": 4.01094913482666,
      "learning_rate": 1.7530727784982393e-06,
      "loss": 0.3323,
      "num_input_tokens_seen": 16538304,
      "step": 25525
    },
    {
      "epoch": 15.05306603773585,
      "grad_norm": 2.1053922176361084,
      "learning_rate": 1.7511167636980765e-06,
      "loss": 0.274,
      "num_input_tokens_seen": 16541984,
      "step": 25530
    },
    {
      "epoch": 15.056014150943396,
      "grad_norm": 2.6751091480255127,
      "learning_rate": 1.7491616090537539e-06,
      "loss": 0.432,
      "num_input_tokens_seen": 16544896,
      "step": 25535
    },
    {
      "epoch": 15.058962264150944,
      "grad_norm": 2.7928948402404785,
      "learning_rate": 1.7472073150829056e-06,
      "loss": 0.4202,
      "num_input_tokens_seen": 16547776,
      "step": 25540
    },
    {
      "epoch": 15.06191037735849,
      "grad_norm": 4.440313816070557,
      "learning_rate": 1.745253882302939e-06,
      "loss": 0.2632,
      "num_input_tokens_seen": 16551456,
      "step": 25545
    },
    {
      "epoch": 15.064858490566039,
      "grad_norm": 3.447692394256592,
      "learning_rate": 1.743301311231035e-06,
      "loss": 0.2897,
      "num_input_tokens_seen": 16555360,
      "step": 25550
    },
    {
      "epoch": 15.067806603773585,
      "grad_norm": 3.8257691860198975,
      "learning_rate": 1.7413496023841437e-06,
      "loss": 0.3036,
      "num_input_tokens_seen": 16559424,
      "step": 25555
    },
    {
      "epoch": 15.070754716981131,
      "grad_norm": 5.556098461151123,
      "learning_rate": 1.7393987562789876e-06,
      "loss": 0.2501,
      "num_input_tokens_seen": 16562016,
      "step": 25560
    },
    {
      "epoch": 15.07370283018868,
      "grad_norm": 4.5865044593811035,
      "learning_rate": 1.7374487734320655e-06,
      "loss": 0.2472,
      "num_input_tokens_seen": 16566016,
      "step": 25565
    },
    {
      "epoch": 15.076650943396226,
      "grad_norm": 5.108593940734863,
      "learning_rate": 1.7354996543596408e-06,
      "loss": 0.2389,
      "num_input_tokens_seen": 16569728,
      "step": 25570
    },
    {
      "epoch": 15.079599056603774,
      "grad_norm": 4.324222564697266,
      "learning_rate": 1.7335513995777504e-06,
      "loss": 0.2863,
      "num_input_tokens_seen": 16572960,
      "step": 25575
    },
    {
      "epoch": 15.08254716981132,
      "grad_norm": 3.1259002685546875,
      "learning_rate": 1.7316040096022062e-06,
      "loss": 0.3791,
      "num_input_tokens_seen": 16576000,
      "step": 25580
    },
    {
      "epoch": 15.085495283018869,
      "grad_norm": 3.055734157562256,
      "learning_rate": 1.7296574849485863e-06,
      "loss": 0.3095,
      "num_input_tokens_seen": 16579520,
      "step": 25585
    },
    {
      "epoch": 15.088443396226415,
      "grad_norm": 2.953097343444824,
      "learning_rate": 1.7277118261322423e-06,
      "loss": 0.3033,
      "num_input_tokens_seen": 16582240,
      "step": 25590
    },
    {
      "epoch": 15.091391509433961,
      "grad_norm": 3.6887807846069336,
      "learning_rate": 1.7257670336682925e-06,
      "loss": 0.2448,
      "num_input_tokens_seen": 16585120,
      "step": 25595
    },
    {
      "epoch": 15.09433962264151,
      "grad_norm": 3.9091532230377197,
      "learning_rate": 1.7238231080716339e-06,
      "loss": 0.2431,
      "num_input_tokens_seen": 16588096,
      "step": 25600
    },
    {
      "epoch": 15.097287735849056,
      "grad_norm": 2.99155330657959,
      "learning_rate": 1.721880049856927e-06,
      "loss": 0.2404,
      "num_input_tokens_seen": 16591008,
      "step": 25605
    },
    {
      "epoch": 15.100235849056604,
      "grad_norm": 3.3517489433288574,
      "learning_rate": 1.7199378595386046e-06,
      "loss": 0.2666,
      "num_input_tokens_seen": 16593664,
      "step": 25610
    },
    {
      "epoch": 15.10318396226415,
      "grad_norm": 5.9264044761657715,
      "learning_rate": 1.7179965376308705e-06,
      "loss": 0.2411,
      "num_input_tokens_seen": 16595936,
      "step": 25615
    },
    {
      "epoch": 15.106132075471699,
      "grad_norm": 3.3338265419006348,
      "learning_rate": 1.7160560846476976e-06,
      "loss": 0.2922,
      "num_input_tokens_seen": 16599840,
      "step": 25620
    },
    {
      "epoch": 15.109080188679245,
      "grad_norm": 3.0530130863189697,
      "learning_rate": 1.7141165011028277e-06,
      "loss": 0.268,
      "num_input_tokens_seen": 16602656,
      "step": 25625
    },
    {
      "epoch": 15.112028301886792,
      "grad_norm": 3.031653881072998,
      "learning_rate": 1.7121777875097767e-06,
      "loss": 0.2412,
      "num_input_tokens_seen": 16606176,
      "step": 25630
    },
    {
      "epoch": 15.11497641509434,
      "grad_norm": 3.243682384490967,
      "learning_rate": 1.7102399443818268e-06,
      "loss": 0.3284,
      "num_input_tokens_seen": 16609152,
      "step": 25635
    },
    {
      "epoch": 15.117924528301886,
      "grad_norm": 4.154804229736328,
      "learning_rate": 1.7083029722320294e-06,
      "loss": 0.2451,
      "num_input_tokens_seen": 16611392,
      "step": 25640
    },
    {
      "epoch": 15.120872641509434,
      "grad_norm": 3.358961820602417,
      "learning_rate": 1.7063668715732063e-06,
      "loss": 0.3193,
      "num_input_tokens_seen": 16614976,
      "step": 25645
    },
    {
      "epoch": 15.12382075471698,
      "grad_norm": 4.3129801750183105,
      "learning_rate": 1.7044316429179492e-06,
      "loss": 0.2793,
      "num_input_tokens_seen": 16617120,
      "step": 25650
    },
    {
      "epoch": 15.126768867924529,
      "grad_norm": 2.9092257022857666,
      "learning_rate": 1.7024972867786155e-06,
      "loss": 0.2932,
      "num_input_tokens_seen": 16620224,
      "step": 25655
    },
    {
      "epoch": 15.129716981132075,
      "grad_norm": 1.8873252868652344,
      "learning_rate": 1.7005638036673389e-06,
      "loss": 0.2979,
      "num_input_tokens_seen": 16623808,
      "step": 25660
    },
    {
      "epoch": 15.132665094339623,
      "grad_norm": 2.6856284141540527,
      "learning_rate": 1.6986311940960148e-06,
      "loss": 0.2179,
      "num_input_tokens_seen": 16628064,
      "step": 25665
    },
    {
      "epoch": 15.13561320754717,
      "grad_norm": 5.091623306274414,
      "learning_rate": 1.696699458576308e-06,
      "loss": 0.2619,
      "num_input_tokens_seen": 16630368,
      "step": 25670
    },
    {
      "epoch": 15.138561320754716,
      "grad_norm": 5.561787128448486,
      "learning_rate": 1.6947685976196581e-06,
      "loss": 0.4116,
      "num_input_tokens_seen": 16632896,
      "step": 25675
    },
    {
      "epoch": 15.141509433962264,
      "grad_norm": 2.038984537124634,
      "learning_rate": 1.692838611737267e-06,
      "loss": 0.2807,
      "num_input_tokens_seen": 16635424,
      "step": 25680
    },
    {
      "epoch": 15.14445754716981,
      "grad_norm": 4.181225776672363,
      "learning_rate": 1.690909501440106e-06,
      "loss": 0.2991,
      "num_input_tokens_seen": 16639040,
      "step": 25685
    },
    {
      "epoch": 15.147405660377359,
      "grad_norm": 2.7878005504608154,
      "learning_rate": 1.688981267238915e-06,
      "loss": 0.3808,
      "num_input_tokens_seen": 16642432,
      "step": 25690
    },
    {
      "epoch": 15.150353773584905,
      "grad_norm": 4.787045001983643,
      "learning_rate": 1.687053909644204e-06,
      "loss": 0.3054,
      "num_input_tokens_seen": 16645376,
      "step": 25695
    },
    {
      "epoch": 15.153301886792454,
      "grad_norm": 3.218079090118408,
      "learning_rate": 1.685127429166249e-06,
      "loss": 0.34,
      "num_input_tokens_seen": 16649024,
      "step": 25700
    },
    {
      "epoch": 15.15625,
      "grad_norm": 6.299826145172119,
      "learning_rate": 1.683201826315093e-06,
      "loss": 0.3879,
      "num_input_tokens_seen": 16651840,
      "step": 25705
    },
    {
      "epoch": 15.159198113207546,
      "grad_norm": 3.9292805194854736,
      "learning_rate": 1.681277101600548e-06,
      "loss": 0.4072,
      "num_input_tokens_seen": 16655616,
      "step": 25710
    },
    {
      "epoch": 15.162146226415095,
      "grad_norm": 2.0129494667053223,
      "learning_rate": 1.6793532555321939e-06,
      "loss": 0.3146,
      "num_input_tokens_seen": 16659104,
      "step": 25715
    },
    {
      "epoch": 15.165094339622641,
      "grad_norm": 3.2715024948120117,
      "learning_rate": 1.6774302886193744e-06,
      "loss": 0.4059,
      "num_input_tokens_seen": 16662304,
      "step": 25720
    },
    {
      "epoch": 15.16804245283019,
      "grad_norm": 4.118152618408203,
      "learning_rate": 1.6755082013712076e-06,
      "loss": 0.4205,
      "num_input_tokens_seen": 16664736,
      "step": 25725
    },
    {
      "epoch": 15.170990566037736,
      "grad_norm": 2.9087769985198975,
      "learning_rate": 1.6735869942965716e-06,
      "loss": 0.355,
      "num_input_tokens_seen": 16667488,
      "step": 25730
    },
    {
      "epoch": 15.173938679245284,
      "grad_norm": 2.783609628677368,
      "learning_rate": 1.6716666679041155e-06,
      "loss": 0.3138,
      "num_input_tokens_seen": 16670272,
      "step": 25735
    },
    {
      "epoch": 15.17688679245283,
      "grad_norm": 4.978923320770264,
      "learning_rate": 1.6697472227022533e-06,
      "loss": 0.2721,
      "num_input_tokens_seen": 16672896,
      "step": 25740
    },
    {
      "epoch": 15.179834905660377,
      "grad_norm": 4.5063347816467285,
      "learning_rate": 1.6678286591991644e-06,
      "loss": 0.3194,
      "num_input_tokens_seen": 16676352,
      "step": 25745
    },
    {
      "epoch": 15.182783018867925,
      "grad_norm": 4.5320329666137695,
      "learning_rate": 1.665910977902801e-06,
      "loss": 0.3153,
      "num_input_tokens_seen": 16679136,
      "step": 25750
    },
    {
      "epoch": 15.185731132075471,
      "grad_norm": 4.955874919891357,
      "learning_rate": 1.6639941793208747e-06,
      "loss": 0.244,
      "num_input_tokens_seen": 16682720,
      "step": 25755
    },
    {
      "epoch": 15.18867924528302,
      "grad_norm": 3.6400864124298096,
      "learning_rate": 1.6620782639608674e-06,
      "loss": 0.4904,
      "num_input_tokens_seen": 16686368,
      "step": 25760
    },
    {
      "epoch": 15.191627358490566,
      "grad_norm": 4.648626327514648,
      "learning_rate": 1.6601632323300231e-06,
      "loss": 0.3188,
      "num_input_tokens_seen": 16689216,
      "step": 25765
    },
    {
      "epoch": 15.194575471698114,
      "grad_norm": 4.880080223083496,
      "learning_rate": 1.6582490849353595e-06,
      "loss": 0.3838,
      "num_input_tokens_seen": 16691648,
      "step": 25770
    },
    {
      "epoch": 15.19752358490566,
      "grad_norm": 3.067298173904419,
      "learning_rate": 1.6563358222836523e-06,
      "loss": 0.4313,
      "num_input_tokens_seen": 16694624,
      "step": 25775
    },
    {
      "epoch": 15.200471698113208,
      "grad_norm": 4.726259708404541,
      "learning_rate": 1.654423444881445e-06,
      "loss": 0.3448,
      "num_input_tokens_seen": 16697472,
      "step": 25780
    },
    {
      "epoch": 15.203419811320755,
      "grad_norm": 4.610049247741699,
      "learning_rate": 1.652511953235051e-06,
      "loss": 0.2211,
      "num_input_tokens_seen": 16700192,
      "step": 25785
    },
    {
      "epoch": 15.206367924528301,
      "grad_norm": 3.4819183349609375,
      "learning_rate": 1.650601347850544e-06,
      "loss": 0.347,
      "num_input_tokens_seen": 16703904,
      "step": 25790
    },
    {
      "epoch": 15.20931603773585,
      "grad_norm": 3.698302745819092,
      "learning_rate": 1.6486916292337652e-06,
      "loss": 0.2678,
      "num_input_tokens_seen": 16707008,
      "step": 25795
    },
    {
      "epoch": 15.212264150943396,
      "grad_norm": 3.7175419330596924,
      "learning_rate": 1.6467827978903212e-06,
      "loss": 0.213,
      "num_input_tokens_seen": 16710240,
      "step": 25800
    },
    {
      "epoch": 15.215212264150944,
      "grad_norm": 3.8573973178863525,
      "learning_rate": 1.6448748543255827e-06,
      "loss": 0.2903,
      "num_input_tokens_seen": 16712960,
      "step": 25805
    },
    {
      "epoch": 15.21816037735849,
      "grad_norm": 3.8495192527770996,
      "learning_rate": 1.6429677990446845e-06,
      "loss": 0.3713,
      "num_input_tokens_seen": 16716960,
      "step": 25810
    },
    {
      "epoch": 15.221108490566039,
      "grad_norm": 3.5758376121520996,
      "learning_rate": 1.6410616325525319e-06,
      "loss": 0.3981,
      "num_input_tokens_seen": 16721376,
      "step": 25815
    },
    {
      "epoch": 15.224056603773585,
      "grad_norm": 3.8492817878723145,
      "learning_rate": 1.6391563553537875e-06,
      "loss": 0.3407,
      "num_input_tokens_seen": 16725152,
      "step": 25820
    },
    {
      "epoch": 15.227004716981131,
      "grad_norm": 12.630085945129395,
      "learning_rate": 1.6372519679528832e-06,
      "loss": 0.2703,
      "num_input_tokens_seen": 16728352,
      "step": 25825
    },
    {
      "epoch": 15.22995283018868,
      "grad_norm": 3.00203013420105,
      "learning_rate": 1.6353484708540124e-06,
      "loss": 0.3823,
      "num_input_tokens_seen": 16731232,
      "step": 25830
    },
    {
      "epoch": 15.232900943396226,
      "grad_norm": 3.2742528915405273,
      "learning_rate": 1.633445864561135e-06,
      "loss": 0.2447,
      "num_input_tokens_seen": 16734080,
      "step": 25835
    },
    {
      "epoch": 15.235849056603774,
      "grad_norm": 3.7739217281341553,
      "learning_rate": 1.6315441495779726e-06,
      "loss": 0.3251,
      "num_input_tokens_seen": 16737184,
      "step": 25840
    },
    {
      "epoch": 15.23879716981132,
      "grad_norm": 6.072956085205078,
      "learning_rate": 1.6296433264080152e-06,
      "loss": 0.3483,
      "num_input_tokens_seen": 16740032,
      "step": 25845
    },
    {
      "epoch": 15.241745283018869,
      "grad_norm": 4.955546855926514,
      "learning_rate": 1.627743395554513e-06,
      "loss": 0.3123,
      "num_input_tokens_seen": 16743296,
      "step": 25850
    },
    {
      "epoch": 15.244693396226415,
      "grad_norm": 3.7902066707611084,
      "learning_rate": 1.6258443575204802e-06,
      "loss": 0.2522,
      "num_input_tokens_seen": 16745952,
      "step": 25855
    },
    {
      "epoch": 15.247641509433961,
      "grad_norm": 4.783172607421875,
      "learning_rate": 1.6239462128086936e-06,
      "loss": 0.2682,
      "num_input_tokens_seen": 16749248,
      "step": 25860
    },
    {
      "epoch": 15.25058962264151,
      "grad_norm": 3.31650972366333,
      "learning_rate": 1.6220489619216988e-06,
      "loss": 0.3891,
      "num_input_tokens_seen": 16752672,
      "step": 25865
    },
    {
      "epoch": 15.253537735849056,
      "grad_norm": 4.564633846282959,
      "learning_rate": 1.6201526053618e-06,
      "loss": 0.3413,
      "num_input_tokens_seen": 16755168,
      "step": 25870
    },
    {
      "epoch": 15.256485849056604,
      "grad_norm": 5.236965179443359,
      "learning_rate": 1.6182571436310634e-06,
      "loss": 0.3113,
      "num_input_tokens_seen": 16758144,
      "step": 25875
    },
    {
      "epoch": 15.25943396226415,
      "grad_norm": 3.4579315185546875,
      "learning_rate": 1.616362577231324e-06,
      "loss": 0.2658,
      "num_input_tokens_seen": 16761568,
      "step": 25880
    },
    {
      "epoch": 15.262382075471699,
      "grad_norm": 3.1551737785339355,
      "learning_rate": 1.614468906664175e-06,
      "loss": 0.2863,
      "num_input_tokens_seen": 16765088,
      "step": 25885
    },
    {
      "epoch": 15.265330188679245,
      "grad_norm": 5.745308876037598,
      "learning_rate": 1.612576132430974e-06,
      "loss": 0.4516,
      "num_input_tokens_seen": 16768736,
      "step": 25890
    },
    {
      "epoch": 15.268278301886792,
      "grad_norm": 3.0895867347717285,
      "learning_rate": 1.6106842550328406e-06,
      "loss": 0.3344,
      "num_input_tokens_seen": 16772384,
      "step": 25895
    },
    {
      "epoch": 15.27122641509434,
      "grad_norm": 4.953463077545166,
      "learning_rate": 1.6087932749706582e-06,
      "loss": 0.2798,
      "num_input_tokens_seen": 16775200,
      "step": 25900
    },
    {
      "epoch": 15.274174528301886,
      "grad_norm": 2.831902503967285,
      "learning_rate": 1.6069031927450696e-06,
      "loss": 0.2741,
      "num_input_tokens_seen": 16778880,
      "step": 25905
    },
    {
      "epoch": 15.277122641509434,
      "grad_norm": 3.0286240577697754,
      "learning_rate": 1.605014008856486e-06,
      "loss": 0.2082,
      "num_input_tokens_seen": 16781760,
      "step": 25910
    },
    {
      "epoch": 15.28007075471698,
      "grad_norm": 3.792267322540283,
      "learning_rate": 1.6031257238050745e-06,
      "loss": 0.3908,
      "num_input_tokens_seen": 16785216,
      "step": 25915
    },
    {
      "epoch": 15.283018867924529,
      "grad_norm": 5.187530040740967,
      "learning_rate": 1.601238338090768e-06,
      "loss": 0.3284,
      "num_input_tokens_seen": 16788672,
      "step": 25920
    },
    {
      "epoch": 15.285966981132075,
      "grad_norm": 7.648937702178955,
      "learning_rate": 1.5993518522132595e-06,
      "loss": 0.3476,
      "num_input_tokens_seen": 16791872,
      "step": 25925
    },
    {
      "epoch": 15.288915094339623,
      "grad_norm": 5.948746681213379,
      "learning_rate": 1.5974662666720037e-06,
      "loss": 0.2473,
      "num_input_tokens_seen": 16796544,
      "step": 25930
    },
    {
      "epoch": 15.29186320754717,
      "grad_norm": 4.185173511505127,
      "learning_rate": 1.5955815819662162e-06,
      "loss": 0.2892,
      "num_input_tokens_seen": 16799968,
      "step": 25935
    },
    {
      "epoch": 15.294811320754716,
      "grad_norm": 5.134406566619873,
      "learning_rate": 1.5936977985948788e-06,
      "loss": 0.2835,
      "num_input_tokens_seen": 16802304,
      "step": 25940
    },
    {
      "epoch": 15.297759433962264,
      "grad_norm": 5.041017055511475,
      "learning_rate": 1.5918149170567298e-06,
      "loss": 0.311,
      "num_input_tokens_seen": 16805632,
      "step": 25945
    },
    {
      "epoch": 15.30070754716981,
      "grad_norm": 3.995783805847168,
      "learning_rate": 1.5899329378502698e-06,
      "loss": 0.3908,
      "num_input_tokens_seen": 16809120,
      "step": 25950
    },
    {
      "epoch": 15.303655660377359,
      "grad_norm": 5.8126749992370605,
      "learning_rate": 1.588051861473761e-06,
      "loss": 0.3054,
      "num_input_tokens_seen": 16812192,
      "step": 25955
    },
    {
      "epoch": 15.306603773584905,
      "grad_norm": 2.4718923568725586,
      "learning_rate": 1.5861716884252253e-06,
      "loss": 0.2483,
      "num_input_tokens_seen": 16815456,
      "step": 25960
    },
    {
      "epoch": 15.309551886792454,
      "grad_norm": 3.835577964782715,
      "learning_rate": 1.5842924192024489e-06,
      "loss": 0.3683,
      "num_input_tokens_seen": 16819616,
      "step": 25965
    },
    {
      "epoch": 15.3125,
      "grad_norm": 3.414702892303467,
      "learning_rate": 1.5824140543029742e-06,
      "loss": 0.434,
      "num_input_tokens_seen": 16822656,
      "step": 25970
    },
    {
      "epoch": 15.315448113207546,
      "grad_norm": 2.327544689178467,
      "learning_rate": 1.5805365942241092e-06,
      "loss": 0.3456,
      "num_input_tokens_seen": 16826304,
      "step": 25975
    },
    {
      "epoch": 15.318396226415095,
      "grad_norm": 5.959931373596191,
      "learning_rate": 1.5786600394629181e-06,
      "loss": 0.5368,
      "num_input_tokens_seen": 16829344,
      "step": 25980
    },
    {
      "epoch": 15.321344339622641,
      "grad_norm": 5.7479352951049805,
      "learning_rate": 1.5767843905162261e-06,
      "loss": 0.342,
      "num_input_tokens_seen": 16832128,
      "step": 25985
    },
    {
      "epoch": 15.32429245283019,
      "grad_norm": 4.828582763671875,
      "learning_rate": 1.5749096478806209e-06,
      "loss": 0.3357,
      "num_input_tokens_seen": 16834880,
      "step": 25990
    },
    {
      "epoch": 15.327240566037736,
      "grad_norm": 5.86256742477417,
      "learning_rate": 1.5730358120524452e-06,
      "loss": 0.3992,
      "num_input_tokens_seen": 16838400,
      "step": 25995
    },
    {
      "epoch": 15.330188679245284,
      "grad_norm": 3.496412992477417,
      "learning_rate": 1.5711628835278098e-06,
      "loss": 0.334,
      "num_input_tokens_seen": 16841440,
      "step": 26000
    },
    {
      "epoch": 15.33313679245283,
      "grad_norm": 3.9540796279907227,
      "learning_rate": 1.5692908628025782e-06,
      "loss": 0.3023,
      "num_input_tokens_seen": 16844064,
      "step": 26005
    },
    {
      "epoch": 15.336084905660377,
      "grad_norm": 5.800181865692139,
      "learning_rate": 1.5674197503723765e-06,
      "loss": 0.2505,
      "num_input_tokens_seen": 16846624,
      "step": 26010
    },
    {
      "epoch": 15.339033018867925,
      "grad_norm": 3.78959584236145,
      "learning_rate": 1.5655495467325893e-06,
      "loss": 0.2369,
      "num_input_tokens_seen": 16849568,
      "step": 26015
    },
    {
      "epoch": 15.341981132075471,
      "grad_norm": 4.293105125427246,
      "learning_rate": 1.5636802523783613e-06,
      "loss": 0.3186,
      "num_input_tokens_seen": 16851968,
      "step": 26020
    },
    {
      "epoch": 15.34492924528302,
      "grad_norm": 2.93966007232666,
      "learning_rate": 1.5618118678045947e-06,
      "loss": 0.4277,
      "num_input_tokens_seen": 16854592,
      "step": 26025
    },
    {
      "epoch": 15.347877358490566,
      "grad_norm": 2.9354138374328613,
      "learning_rate": 1.5599443935059549e-06,
      "loss": 0.3562,
      "num_input_tokens_seen": 16857280,
      "step": 26030
    },
    {
      "epoch": 15.350825471698114,
      "grad_norm": 6.03622579574585,
      "learning_rate": 1.5580778299768635e-06,
      "loss": 0.2993,
      "num_input_tokens_seen": 16860736,
      "step": 26035
    },
    {
      "epoch": 15.35377358490566,
      "grad_norm": 6.278172492980957,
      "learning_rate": 1.5562121777114997e-06,
      "loss": 0.344,
      "num_input_tokens_seen": 16863328,
      "step": 26040
    },
    {
      "epoch": 15.356721698113208,
      "grad_norm": 4.688877582550049,
      "learning_rate": 1.5543474372038043e-06,
      "loss": 0.3087,
      "num_input_tokens_seen": 16866560,
      "step": 26045
    },
    {
      "epoch": 15.359669811320755,
      "grad_norm": 3.8979151248931885,
      "learning_rate": 1.5524836089474748e-06,
      "loss": 0.4517,
      "num_input_tokens_seen": 16869920,
      "step": 26050
    },
    {
      "epoch": 15.362617924528301,
      "grad_norm": 4.317824840545654,
      "learning_rate": 1.5506206934359664e-06,
      "loss": 0.2922,
      "num_input_tokens_seen": 16873440,
      "step": 26055
    },
    {
      "epoch": 15.36556603773585,
      "grad_norm": 3.872885227203369,
      "learning_rate": 1.5487586911624947e-06,
      "loss": 0.2785,
      "num_input_tokens_seen": 16876480,
      "step": 26060
    },
    {
      "epoch": 15.368514150943396,
      "grad_norm": 5.56221866607666,
      "learning_rate": 1.5468976026200355e-06,
      "loss": 0.2896,
      "num_input_tokens_seen": 16880960,
      "step": 26065
    },
    {
      "epoch": 15.371462264150944,
      "grad_norm": 4.018038272857666,
      "learning_rate": 1.5450374283013187e-06,
      "loss": 0.3199,
      "num_input_tokens_seen": 16885056,
      "step": 26070
    },
    {
      "epoch": 15.37441037735849,
      "grad_norm": 4.470265865325928,
      "learning_rate": 1.5431781686988317e-06,
      "loss": 0.3476,
      "num_input_tokens_seen": 16888064,
      "step": 26075
    },
    {
      "epoch": 15.377358490566039,
      "grad_norm": 2.8734023571014404,
      "learning_rate": 1.5413198243048233e-06,
      "loss": 0.2282,
      "num_input_tokens_seen": 16891328,
      "step": 26080
    },
    {
      "epoch": 15.380306603773585,
      "grad_norm": 2.659227132797241,
      "learning_rate": 1.5394623956112974e-06,
      "loss": 0.3619,
      "num_input_tokens_seen": 16895296,
      "step": 26085
    },
    {
      "epoch": 15.383254716981131,
      "grad_norm": 2.8040459156036377,
      "learning_rate": 1.537605883110015e-06,
      "loss": 0.3068,
      "num_input_tokens_seen": 16898976,
      "step": 26090
    },
    {
      "epoch": 15.38620283018868,
      "grad_norm": 4.415372371673584,
      "learning_rate": 1.5357502872924984e-06,
      "loss": 0.4319,
      "num_input_tokens_seen": 16901312,
      "step": 26095
    },
    {
      "epoch": 15.389150943396226,
      "grad_norm": 5.616690635681152,
      "learning_rate": 1.5338956086500235e-06,
      "loss": 0.3874,
      "num_input_tokens_seen": 16904608,
      "step": 26100
    },
    {
      "epoch": 15.392099056603774,
      "grad_norm": 3.3440606594085693,
      "learning_rate": 1.5320418476736237e-06,
      "loss": 0.3703,
      "num_input_tokens_seen": 16907680,
      "step": 26105
    },
    {
      "epoch": 15.39504716981132,
      "grad_norm": 2.762059450149536,
      "learning_rate": 1.5301890048540912e-06,
      "loss": 0.2853,
      "num_input_tokens_seen": 16910944,
      "step": 26110
    },
    {
      "epoch": 15.397995283018869,
      "grad_norm": 5.327790260314941,
      "learning_rate": 1.5283370806819743e-06,
      "loss": 0.3475,
      "num_input_tokens_seen": 16913536,
      "step": 26115
    },
    {
      "epoch": 15.400943396226415,
      "grad_norm": 3.2494544982910156,
      "learning_rate": 1.5264860756475752e-06,
      "loss": 0.326,
      "num_input_tokens_seen": 16917088,
      "step": 26120
    },
    {
      "epoch": 15.403891509433961,
      "grad_norm": 2.4333677291870117,
      "learning_rate": 1.5246359902409592e-06,
      "loss": 0.2233,
      "num_input_tokens_seen": 16920000,
      "step": 26125
    },
    {
      "epoch": 15.40683962264151,
      "grad_norm": 4.848178386688232,
      "learning_rate": 1.5227868249519423e-06,
      "loss": 0.2829,
      "num_input_tokens_seen": 16923360,
      "step": 26130
    },
    {
      "epoch": 15.409787735849056,
      "grad_norm": 3.1478376388549805,
      "learning_rate": 1.5209385802700999e-06,
      "loss": 0.2888,
      "num_input_tokens_seen": 16926624,
      "step": 26135
    },
    {
      "epoch": 15.412735849056604,
      "grad_norm": 3.025919198989868,
      "learning_rate": 1.5190912566847626e-06,
      "loss": 0.303,
      "num_input_tokens_seen": 16929952,
      "step": 26140
    },
    {
      "epoch": 15.41568396226415,
      "grad_norm": 3.6880218982696533,
      "learning_rate": 1.5172448546850166e-06,
      "loss": 0.3613,
      "num_input_tokens_seen": 16932896,
      "step": 26145
    },
    {
      "epoch": 15.418632075471699,
      "grad_norm": 2.7964258193969727,
      "learning_rate": 1.515399374759704e-06,
      "loss": 0.3396,
      "num_input_tokens_seen": 16935936,
      "step": 26150
    },
    {
      "epoch": 15.421580188679245,
      "grad_norm": 7.185953617095947,
      "learning_rate": 1.513554817397424e-06,
      "loss": 0.3405,
      "num_input_tokens_seen": 16938624,
      "step": 26155
    },
    {
      "epoch": 15.424528301886792,
      "grad_norm": 2.8106064796447754,
      "learning_rate": 1.5117111830865338e-06,
      "loss": 0.3444,
      "num_input_tokens_seen": 16942336,
      "step": 26160
    },
    {
      "epoch": 15.42747641509434,
      "grad_norm": 3.7072534561157227,
      "learning_rate": 1.509868472315142e-06,
      "loss": 0.2858,
      "num_input_tokens_seen": 16945216,
      "step": 26165
    },
    {
      "epoch": 15.430424528301886,
      "grad_norm": 2.9603962898254395,
      "learning_rate": 1.508026685571113e-06,
      "loss": 0.3205,
      "num_input_tokens_seen": 16947968,
      "step": 26170
    },
    {
      "epoch": 15.433372641509434,
      "grad_norm": 4.100912570953369,
      "learning_rate": 1.506185823342069e-06,
      "loss": 0.213,
      "num_input_tokens_seen": 16950688,
      "step": 26175
    },
    {
      "epoch": 15.43632075471698,
      "grad_norm": 5.126577854156494,
      "learning_rate": 1.504345886115386e-06,
      "loss": 0.307,
      "num_input_tokens_seen": 16953344,
      "step": 26180
    },
    {
      "epoch": 15.439268867924529,
      "grad_norm": 4.62950325012207,
      "learning_rate": 1.502506874378193e-06,
      "loss": 0.4162,
      "num_input_tokens_seen": 16956896,
      "step": 26185
    },
    {
      "epoch": 15.442216981132075,
      "grad_norm": 8.995813369750977,
      "learning_rate": 1.5006687886173805e-06,
      "loss": 0.3303,
      "num_input_tokens_seen": 16960160,
      "step": 26190
    },
    {
      "epoch": 15.445165094339623,
      "grad_norm": 3.232882022857666,
      "learning_rate": 1.498831629319587e-06,
      "loss": 0.3004,
      "num_input_tokens_seen": 16963584,
      "step": 26195
    },
    {
      "epoch": 15.44811320754717,
      "grad_norm": 3.921088933944702,
      "learning_rate": 1.4969953969712087e-06,
      "loss": 0.1816,
      "num_input_tokens_seen": 16966656,
      "step": 26200
    },
    {
      "epoch": 15.451061320754716,
      "grad_norm": 3.9826176166534424,
      "learning_rate": 1.4951600920583963e-06,
      "loss": 0.2644,
      "num_input_tokens_seen": 16969952,
      "step": 26205
    },
    {
      "epoch": 15.454009433962264,
      "grad_norm": 2.4077470302581787,
      "learning_rate": 1.493325715067055e-06,
      "loss": 0.3315,
      "num_input_tokens_seen": 16973472,
      "step": 26210
    },
    {
      "epoch": 15.45695754716981,
      "grad_norm": 3.7826709747314453,
      "learning_rate": 1.4914922664828417e-06,
      "loss": 0.2706,
      "num_input_tokens_seen": 16975904,
      "step": 26215
    },
    {
      "epoch": 15.459905660377359,
      "grad_norm": 5.302326202392578,
      "learning_rate": 1.4896597467911732e-06,
      "loss": 0.3419,
      "num_input_tokens_seen": 16979680,
      "step": 26220
    },
    {
      "epoch": 15.462853773584905,
      "grad_norm": 3.971752405166626,
      "learning_rate": 1.4878281564772156e-06,
      "loss": 0.2828,
      "num_input_tokens_seen": 16982528,
      "step": 26225
    },
    {
      "epoch": 15.465801886792454,
      "grad_norm": 5.936448097229004,
      "learning_rate": 1.4859974960258898e-06,
      "loss": 0.2312,
      "num_input_tokens_seen": 16985760,
      "step": 26230
    },
    {
      "epoch": 15.46875,
      "grad_norm": 7.280527591705322,
      "learning_rate": 1.4841677659218723e-06,
      "loss": 0.2516,
      "num_input_tokens_seen": 16988608,
      "step": 26235
    },
    {
      "epoch": 15.471698113207546,
      "grad_norm": 6.714340686798096,
      "learning_rate": 1.4823389666495886e-06,
      "loss": 0.3827,
      "num_input_tokens_seen": 16991424,
      "step": 26240
    },
    {
      "epoch": 15.474646226415095,
      "grad_norm": 3.6696510314941406,
      "learning_rate": 1.4805110986932258e-06,
      "loss": 0.3767,
      "num_input_tokens_seen": 16994336,
      "step": 26245
    },
    {
      "epoch": 15.477594339622641,
      "grad_norm": 2.1313636302948,
      "learning_rate": 1.4786841625367166e-06,
      "loss": 0.2994,
      "num_input_tokens_seen": 16997216,
      "step": 26250
    },
    {
      "epoch": 15.48054245283019,
      "grad_norm": 3.663418769836426,
      "learning_rate": 1.476858158663752e-06,
      "loss": 0.3179,
      "num_input_tokens_seen": 17000448,
      "step": 26255
    },
    {
      "epoch": 15.483490566037736,
      "grad_norm": 4.726825714111328,
      "learning_rate": 1.4750330875577745e-06,
      "loss": 0.2471,
      "num_input_tokens_seen": 17003360,
      "step": 26260
    },
    {
      "epoch": 15.486438679245284,
      "grad_norm": 2.6707003116607666,
      "learning_rate": 1.4732089497019787e-06,
      "loss": 0.2381,
      "num_input_tokens_seen": 17006048,
      "step": 26265
    },
    {
      "epoch": 15.48938679245283,
      "grad_norm": 4.123985290527344,
      "learning_rate": 1.471385745579313e-06,
      "loss": 0.3098,
      "num_input_tokens_seen": 17009280,
      "step": 26270
    },
    {
      "epoch": 15.492334905660377,
      "grad_norm": 5.958913803100586,
      "learning_rate": 1.4695634756724775e-06,
      "loss": 0.2593,
      "num_input_tokens_seen": 17011808,
      "step": 26275
    },
    {
      "epoch": 15.495283018867925,
      "grad_norm": 2.864502191543579,
      "learning_rate": 1.4677421404639281e-06,
      "loss": 0.2761,
      "num_input_tokens_seen": 17015360,
      "step": 26280
    },
    {
      "epoch": 15.498231132075471,
      "grad_norm": 3.8663644790649414,
      "learning_rate": 1.4659217404358706e-06,
      "loss": 0.3248,
      "num_input_tokens_seen": 17018656,
      "step": 26285
    },
    {
      "epoch": 15.50117924528302,
      "grad_norm": 2.9586942195892334,
      "learning_rate": 1.4641022760702627e-06,
      "loss": 0.2912,
      "num_input_tokens_seen": 17022144,
      "step": 26290
    },
    {
      "epoch": 15.504127358490566,
      "grad_norm": 3.1351630687713623,
      "learning_rate": 1.4622837478488172e-06,
      "loss": 0.2913,
      "num_input_tokens_seen": 17025184,
      "step": 26295
    },
    {
      "epoch": 15.507075471698114,
      "grad_norm": 3.06508731842041,
      "learning_rate": 1.4604661562529953e-06,
      "loss": 0.2921,
      "num_input_tokens_seen": 17027776,
      "step": 26300
    },
    {
      "epoch": 15.51002358490566,
      "grad_norm": 5.29071044921875,
      "learning_rate": 1.4586495017640119e-06,
      "loss": 0.279,
      "num_input_tokens_seen": 17030208,
      "step": 26305
    },
    {
      "epoch": 15.512971698113208,
      "grad_norm": 3.380730628967285,
      "learning_rate": 1.4568337848628366e-06,
      "loss": 0.3275,
      "num_input_tokens_seen": 17033888,
      "step": 26310
    },
    {
      "epoch": 15.515919811320755,
      "grad_norm": 4.787678241729736,
      "learning_rate": 1.4550190060301872e-06,
      "loss": 0.2552,
      "num_input_tokens_seen": 17036352,
      "step": 26315
    },
    {
      "epoch": 15.518867924528301,
      "grad_norm": 3.590803861618042,
      "learning_rate": 1.4532051657465335e-06,
      "loss": 0.4441,
      "num_input_tokens_seen": 17039136,
      "step": 26320
    },
    {
      "epoch": 15.52181603773585,
      "grad_norm": 2.5658626556396484,
      "learning_rate": 1.4513922644920985e-06,
      "loss": 0.3171,
      "num_input_tokens_seen": 17043776,
      "step": 26325
    },
    {
      "epoch": 15.524764150943396,
      "grad_norm": 3.0965189933776855,
      "learning_rate": 1.4495803027468552e-06,
      "loss": 0.2924,
      "num_input_tokens_seen": 17048000,
      "step": 26330
    },
    {
      "epoch": 15.527712264150944,
      "grad_norm": 3.077892541885376,
      "learning_rate": 1.4477692809905263e-06,
      "loss": 0.372,
      "num_input_tokens_seen": 17052800,
      "step": 26335
    },
    {
      "epoch": 15.53066037735849,
      "grad_norm": 3.9784533977508545,
      "learning_rate": 1.4459591997025896e-06,
      "loss": 0.3364,
      "num_input_tokens_seen": 17056064,
      "step": 26340
    },
    {
      "epoch": 15.533608490566039,
      "grad_norm": 2.3792176246643066,
      "learning_rate": 1.4441500593622737e-06,
      "loss": 0.269,
      "num_input_tokens_seen": 17059936,
      "step": 26345
    },
    {
      "epoch": 15.536556603773585,
      "grad_norm": 3.9217891693115234,
      "learning_rate": 1.4423418604485539e-06,
      "loss": 0.2778,
      "num_input_tokens_seen": 17062720,
      "step": 26350
    },
    {
      "epoch": 15.539504716981131,
      "grad_norm": 2.6532249450683594,
      "learning_rate": 1.4405346034401597e-06,
      "loss": 0.2556,
      "num_input_tokens_seen": 17065632,
      "step": 26355
    },
    {
      "epoch": 15.54245283018868,
      "grad_norm": 3.9740867614746094,
      "learning_rate": 1.4387282888155695e-06,
      "loss": 0.3431,
      "num_input_tokens_seen": 17068768,
      "step": 26360
    },
    {
      "epoch": 15.545400943396226,
      "grad_norm": 4.074392318725586,
      "learning_rate": 1.436922917053013e-06,
      "loss": 0.3489,
      "num_input_tokens_seen": 17072896,
      "step": 26365
    },
    {
      "epoch": 15.548349056603774,
      "grad_norm": 3.683452844619751,
      "learning_rate": 1.4351184886304686e-06,
      "loss": 0.311,
      "num_input_tokens_seen": 17079392,
      "step": 26370
    },
    {
      "epoch": 15.55129716981132,
      "grad_norm": 4.518885135650635,
      "learning_rate": 1.4333150040256699e-06,
      "loss": 0.3799,
      "num_input_tokens_seen": 17082272,
      "step": 26375
    },
    {
      "epoch": 15.554245283018869,
      "grad_norm": 3.3318097591400146,
      "learning_rate": 1.4315124637160954e-06,
      "loss": 0.3338,
      "num_input_tokens_seen": 17084672,
      "step": 26380
    },
    {
      "epoch": 15.557193396226415,
      "grad_norm": 2.6786231994628906,
      "learning_rate": 1.4297108681789752e-06,
      "loss": 0.3274,
      "num_input_tokens_seen": 17088352,
      "step": 26385
    },
    {
      "epoch": 15.560141509433961,
      "grad_norm": 3.3110344409942627,
      "learning_rate": 1.4279102178912902e-06,
      "loss": 0.3476,
      "num_input_tokens_seen": 17091040,
      "step": 26390
    },
    {
      "epoch": 15.56308962264151,
      "grad_norm": 3.2776594161987305,
      "learning_rate": 1.4261105133297693e-06,
      "loss": 0.2463,
      "num_input_tokens_seen": 17094240,
      "step": 26395
    },
    {
      "epoch": 15.566037735849056,
      "grad_norm": 2.385831356048584,
      "learning_rate": 1.4243117549708913e-06,
      "loss": 0.3588,
      "num_input_tokens_seen": 17098304,
      "step": 26400
    },
    {
      "epoch": 15.568985849056604,
      "grad_norm": 3.0324604511260986,
      "learning_rate": 1.422513943290888e-06,
      "loss": 0.2392,
      "num_input_tokens_seen": 17103104,
      "step": 26405
    },
    {
      "epoch": 15.57193396226415,
      "grad_norm": 7.484667778015137,
      "learning_rate": 1.4207170787657365e-06,
      "loss": 0.3573,
      "num_input_tokens_seen": 17106112,
      "step": 26410
    },
    {
      "epoch": 15.574882075471699,
      "grad_norm": 6.071022033691406,
      "learning_rate": 1.4189211618711646e-06,
      "loss": 0.2764,
      "num_input_tokens_seen": 17108832,
      "step": 26415
    },
    {
      "epoch": 15.577830188679245,
      "grad_norm": 2.4620466232299805,
      "learning_rate": 1.417126193082648e-06,
      "loss": 0.3477,
      "num_input_tokens_seen": 17112096,
      "step": 26420
    },
    {
      "epoch": 15.580778301886792,
      "grad_norm": 3.665377378463745,
      "learning_rate": 1.4153321728754133e-06,
      "loss": 0.33,
      "num_input_tokens_seen": 17114976,
      "step": 26425
    },
    {
      "epoch": 15.58372641509434,
      "grad_norm": 3.7794721126556396,
      "learning_rate": 1.4135391017244338e-06,
      "loss": 0.3241,
      "num_input_tokens_seen": 17118464,
      "step": 26430
    },
    {
      "epoch": 15.586674528301886,
      "grad_norm": 4.58062744140625,
      "learning_rate": 1.4117469801044332e-06,
      "loss": 0.4486,
      "num_input_tokens_seen": 17120896,
      "step": 26435
    },
    {
      "epoch": 15.589622641509434,
      "grad_norm": 2.434398889541626,
      "learning_rate": 1.4099558084898862e-06,
      "loss": 0.3949,
      "num_input_tokens_seen": 17124256,
      "step": 26440
    },
    {
      "epoch": 15.59257075471698,
      "grad_norm": 4.076146602630615,
      "learning_rate": 1.408165587355011e-06,
      "loss": 0.3586,
      "num_input_tokens_seen": 17127360,
      "step": 26445
    },
    {
      "epoch": 15.595518867924529,
      "grad_norm": 4.977184772491455,
      "learning_rate": 1.4063763171737766e-06,
      "loss": 0.4074,
      "num_input_tokens_seen": 17129824,
      "step": 26450
    },
    {
      "epoch": 15.598466981132075,
      "grad_norm": 5.487415790557861,
      "learning_rate": 1.4045879984198996e-06,
      "loss": 0.3656,
      "num_input_tokens_seen": 17132416,
      "step": 26455
    },
    {
      "epoch": 15.601415094339622,
      "grad_norm": 5.024888515472412,
      "learning_rate": 1.4028006315668457e-06,
      "loss": 0.299,
      "num_input_tokens_seen": 17135456,
      "step": 26460
    },
    {
      "epoch": 15.60436320754717,
      "grad_norm": 3.179896354675293,
      "learning_rate": 1.4010142170878261e-06,
      "loss": 0.3415,
      "num_input_tokens_seen": 17138752,
      "step": 26465
    },
    {
      "epoch": 15.607311320754716,
      "grad_norm": 4.436010837554932,
      "learning_rate": 1.3992287554558042e-06,
      "loss": 0.3154,
      "num_input_tokens_seen": 17141792,
      "step": 26470
    },
    {
      "epoch": 15.610259433962264,
      "grad_norm": 3.1925556659698486,
      "learning_rate": 1.3974442471434885e-06,
      "loss": 0.29,
      "num_input_tokens_seen": 17144256,
      "step": 26475
    },
    {
      "epoch": 15.61320754716981,
      "grad_norm": 3.7790679931640625,
      "learning_rate": 1.395660692623334e-06,
      "loss": 0.2286,
      "num_input_tokens_seen": 17147584,
      "step": 26480
    },
    {
      "epoch": 15.616155660377359,
      "grad_norm": 4.011687278747559,
      "learning_rate": 1.3938780923675454e-06,
      "loss": 0.3032,
      "num_input_tokens_seen": 17151584,
      "step": 26485
    },
    {
      "epoch": 15.619103773584905,
      "grad_norm": 5.5867390632629395,
      "learning_rate": 1.3920964468480718e-06,
      "loss": 0.4059,
      "num_input_tokens_seen": 17154976,
      "step": 26490
    },
    {
      "epoch": 15.622051886792454,
      "grad_norm": 5.3249897956848145,
      "learning_rate": 1.3903157565366143e-06,
      "loss": 0.3594,
      "num_input_tokens_seen": 17157824,
      "step": 26495
    },
    {
      "epoch": 15.625,
      "grad_norm": 4.137224197387695,
      "learning_rate": 1.3885360219046172e-06,
      "loss": 0.3229,
      "num_input_tokens_seen": 17160448,
      "step": 26500
    },
    {
      "epoch": 15.627948113207546,
      "grad_norm": 3.8813116550445557,
      "learning_rate": 1.386757243423273e-06,
      "loss": 0.2912,
      "num_input_tokens_seen": 17163296,
      "step": 26505
    },
    {
      "epoch": 15.630896226415095,
      "grad_norm": 4.563551902770996,
      "learning_rate": 1.384979421563521e-06,
      "loss": 0.3491,
      "num_input_tokens_seen": 17166240,
      "step": 26510
    },
    {
      "epoch": 15.633844339622641,
      "grad_norm": 3.3744993209838867,
      "learning_rate": 1.3832025567960465e-06,
      "loss": 0.3445,
      "num_input_tokens_seen": 17169184,
      "step": 26515
    },
    {
      "epoch": 15.63679245283019,
      "grad_norm": 2.020312547683716,
      "learning_rate": 1.3814266495912815e-06,
      "loss": 0.3422,
      "num_input_tokens_seen": 17175712,
      "step": 26520
    },
    {
      "epoch": 15.639740566037736,
      "grad_norm": 5.188023567199707,
      "learning_rate": 1.3796517004194078e-06,
      "loss": 0.4453,
      "num_input_tokens_seen": 17178304,
      "step": 26525
    },
    {
      "epoch": 15.642688679245284,
      "grad_norm": 3.581986427307129,
      "learning_rate": 1.3778777097503476e-06,
      "loss": 0.2802,
      "num_input_tokens_seen": 17181792,
      "step": 26530
    },
    {
      "epoch": 15.64563679245283,
      "grad_norm": 3.4731380939483643,
      "learning_rate": 1.3761046780537757e-06,
      "loss": 0.2883,
      "num_input_tokens_seen": 17184640,
      "step": 26535
    },
    {
      "epoch": 15.648584905660378,
      "grad_norm": 5.270940780639648,
      "learning_rate": 1.3743326057991086e-06,
      "loss": 0.4416,
      "num_input_tokens_seen": 17187424,
      "step": 26540
    },
    {
      "epoch": 15.651533018867925,
      "grad_norm": 2.68180775642395,
      "learning_rate": 1.3725614934555093e-06,
      "loss": 0.3234,
      "num_input_tokens_seen": 17190656,
      "step": 26545
    },
    {
      "epoch": 15.654481132075471,
      "grad_norm": 4.113996505737305,
      "learning_rate": 1.3707913414918882e-06,
      "loss": 0.4242,
      "num_input_tokens_seen": 17193696,
      "step": 26550
    },
    {
      "epoch": 15.65742924528302,
      "grad_norm": 2.4794387817382812,
      "learning_rate": 1.3690221503768996e-06,
      "loss": 0.3191,
      "num_input_tokens_seen": 17196768,
      "step": 26555
    },
    {
      "epoch": 15.660377358490566,
      "grad_norm": 8.390254020690918,
      "learning_rate": 1.3672539205789465e-06,
      "loss": 0.2945,
      "num_input_tokens_seen": 17199424,
      "step": 26560
    },
    {
      "epoch": 15.663325471698114,
      "grad_norm": 3.250436782836914,
      "learning_rate": 1.3654866525661737e-06,
      "loss": 0.3058,
      "num_input_tokens_seen": 17203232,
      "step": 26565
    },
    {
      "epoch": 15.66627358490566,
      "grad_norm": 4.670880317687988,
      "learning_rate": 1.3637203468064741e-06,
      "loss": 0.2567,
      "num_input_tokens_seen": 17206272,
      "step": 26570
    },
    {
      "epoch": 15.669221698113208,
      "grad_norm": 9.53282356262207,
      "learning_rate": 1.3619550037674838e-06,
      "loss": 0.4322,
      "num_input_tokens_seen": 17209056,
      "step": 26575
    },
    {
      "epoch": 15.672169811320755,
      "grad_norm": 2.9615285396575928,
      "learning_rate": 1.3601906239165857e-06,
      "loss": 0.295,
      "num_input_tokens_seen": 17212800,
      "step": 26580
    },
    {
      "epoch": 15.675117924528301,
      "grad_norm": 4.027200222015381,
      "learning_rate": 1.3584272077209048e-06,
      "loss": 0.2056,
      "num_input_tokens_seen": 17215616,
      "step": 26585
    },
    {
      "epoch": 15.67806603773585,
      "grad_norm": 5.155170440673828,
      "learning_rate": 1.3566647556473168e-06,
      "loss": 0.2796,
      "num_input_tokens_seen": 17218880,
      "step": 26590
    },
    {
      "epoch": 15.681014150943396,
      "grad_norm": 2.987861156463623,
      "learning_rate": 1.3549032681624363e-06,
      "loss": 0.2685,
      "num_input_tokens_seen": 17222272,
      "step": 26595
    },
    {
      "epoch": 15.683962264150944,
      "grad_norm": 2.420109987258911,
      "learning_rate": 1.3531427457326252e-06,
      "loss": 0.3155,
      "num_input_tokens_seen": 17225984,
      "step": 26600
    },
    {
      "epoch": 15.68691037735849,
      "grad_norm": 4.3675618171691895,
      "learning_rate": 1.3513831888239893e-06,
      "loss": 0.3747,
      "num_input_tokens_seen": 17229184,
      "step": 26605
    },
    {
      "epoch": 15.689858490566039,
      "grad_norm": 3.131864070892334,
      "learning_rate": 1.3496245979023786e-06,
      "loss": 0.4127,
      "num_input_tokens_seen": 17231456,
      "step": 26610
    },
    {
      "epoch": 15.692806603773585,
      "grad_norm": 3.6683974266052246,
      "learning_rate": 1.3478669734333865e-06,
      "loss": 0.2883,
      "num_input_tokens_seen": 17234272,
      "step": 26615
    },
    {
      "epoch": 15.695754716981131,
      "grad_norm": 3.8379223346710205,
      "learning_rate": 1.3461103158823546e-06,
      "loss": 0.3728,
      "num_input_tokens_seen": 17237888,
      "step": 26620
    },
    {
      "epoch": 15.69870283018868,
      "grad_norm": 2.9344685077667236,
      "learning_rate": 1.3443546257143624e-06,
      "loss": 0.3427,
      "num_input_tokens_seen": 17241920,
      "step": 26625
    },
    {
      "epoch": 15.701650943396226,
      "grad_norm": 2.835749864578247,
      "learning_rate": 1.3425999033942395e-06,
      "loss": 0.3667,
      "num_input_tokens_seen": 17245440,
      "step": 26630
    },
    {
      "epoch": 15.704599056603774,
      "grad_norm": 4.526761054992676,
      "learning_rate": 1.3408461493865549e-06,
      "loss": 0.2687,
      "num_input_tokens_seen": 17247936,
      "step": 26635
    },
    {
      "epoch": 15.70754716981132,
      "grad_norm": 3.2681593894958496,
      "learning_rate": 1.339093364155622e-06,
      "loss": 0.2295,
      "num_input_tokens_seen": 17251552,
      "step": 26640
    },
    {
      "epoch": 15.710495283018869,
      "grad_norm": 3.135870933532715,
      "learning_rate": 1.3373415481654988e-06,
      "loss": 0.3054,
      "num_input_tokens_seen": 17258784,
      "step": 26645
    },
    {
      "epoch": 15.713443396226415,
      "grad_norm": 5.564846515655518,
      "learning_rate": 1.335590701879984e-06,
      "loss": 0.3092,
      "num_input_tokens_seen": 17262400,
      "step": 26650
    },
    {
      "epoch": 15.716391509433961,
      "grad_norm": 6.56998348236084,
      "learning_rate": 1.3338408257626257e-06,
      "loss": 0.2801,
      "num_input_tokens_seen": 17265984,
      "step": 26655
    },
    {
      "epoch": 15.71933962264151,
      "grad_norm": 4.081916809082031,
      "learning_rate": 1.3320919202767086e-06,
      "loss": 0.2957,
      "num_input_tokens_seen": 17268640,
      "step": 26660
    },
    {
      "epoch": 15.722287735849056,
      "grad_norm": 5.035322666168213,
      "learning_rate": 1.3303439858852636e-06,
      "loss": 0.3168,
      "num_input_tokens_seen": 17271552,
      "step": 26665
    },
    {
      "epoch": 15.725235849056604,
      "grad_norm": 3.3555800914764404,
      "learning_rate": 1.3285970230510636e-06,
      "loss": 0.2483,
      "num_input_tokens_seen": 17275328,
      "step": 26670
    },
    {
      "epoch": 15.72818396226415,
      "grad_norm": 2.39517879486084,
      "learning_rate": 1.3268510322366246e-06,
      "loss": 0.3084,
      "num_input_tokens_seen": 17278688,
      "step": 26675
    },
    {
      "epoch": 15.731132075471699,
      "grad_norm": 2.7585034370422363,
      "learning_rate": 1.3251060139042038e-06,
      "loss": 0.2932,
      "num_input_tokens_seen": 17281824,
      "step": 26680
    },
    {
      "epoch": 15.734080188679245,
      "grad_norm": 4.659368991851807,
      "learning_rate": 1.3233619685158056e-06,
      "loss": 0.2745,
      "num_input_tokens_seen": 17284768,
      "step": 26685
    },
    {
      "epoch": 15.737028301886792,
      "grad_norm": 2.8952252864837646,
      "learning_rate": 1.3216188965331712e-06,
      "loss": 0.354,
      "num_input_tokens_seen": 17288288,
      "step": 26690
    },
    {
      "epoch": 15.73997641509434,
      "grad_norm": 4.449194431304932,
      "learning_rate": 1.3198767984177869e-06,
      "loss": 0.2558,
      "num_input_tokens_seen": 17291072,
      "step": 26695
    },
    {
      "epoch": 15.742924528301886,
      "grad_norm": 5.1639323234558105,
      "learning_rate": 1.3181356746308805e-06,
      "loss": 0.33,
      "num_input_tokens_seen": 17294144,
      "step": 26700
    },
    {
      "epoch": 15.745872641509434,
      "grad_norm": 4.881172180175781,
      "learning_rate": 1.3163955256334226e-06,
      "loss": 0.3923,
      "num_input_tokens_seen": 17298048,
      "step": 26705
    },
    {
      "epoch": 15.74882075471698,
      "grad_norm": 3.5677409172058105,
      "learning_rate": 1.3146563518861227e-06,
      "loss": 0.4311,
      "num_input_tokens_seen": 17301920,
      "step": 26710
    },
    {
      "epoch": 15.751768867924529,
      "grad_norm": 5.221879959106445,
      "learning_rate": 1.3129181538494384e-06,
      "loss": 0.2657,
      "num_input_tokens_seen": 17304992,
      "step": 26715
    },
    {
      "epoch": 15.754716981132075,
      "grad_norm": 3.501457929611206,
      "learning_rate": 1.3111809319835622e-06,
      "loss": 0.2922,
      "num_input_tokens_seen": 17308192,
      "step": 26720
    },
    {
      "epoch": 15.757665094339622,
      "grad_norm": 1.9794400930404663,
      "learning_rate": 1.3094446867484335e-06,
      "loss": 0.387,
      "num_input_tokens_seen": 17312032,
      "step": 26725
    },
    {
      "epoch": 15.76061320754717,
      "grad_norm": 4.238003253936768,
      "learning_rate": 1.3077094186037287e-06,
      "loss": 0.2884,
      "num_input_tokens_seen": 17316128,
      "step": 26730
    },
    {
      "epoch": 15.763561320754716,
      "grad_norm": 2.4108469486236572,
      "learning_rate": 1.305975128008869e-06,
      "loss": 0.192,
      "num_input_tokens_seen": 17319264,
      "step": 26735
    },
    {
      "epoch": 15.766509433962264,
      "grad_norm": 2.6997339725494385,
      "learning_rate": 1.304241815423014e-06,
      "loss": 0.3522,
      "num_input_tokens_seen": 17323040,
      "step": 26740
    },
    {
      "epoch": 15.76945754716981,
      "grad_norm": 3.097982406616211,
      "learning_rate": 1.3025094813050655e-06,
      "loss": 0.3648,
      "num_input_tokens_seen": 17325856,
      "step": 26745
    },
    {
      "epoch": 15.772405660377359,
      "grad_norm": 3.4028525352478027,
      "learning_rate": 1.3007781261136675e-06,
      "loss": 0.2611,
      "num_input_tokens_seen": 17329216,
      "step": 26750
    },
    {
      "epoch": 15.775353773584905,
      "grad_norm": 3.656229019165039,
      "learning_rate": 1.299047750307204e-06,
      "loss": 0.3069,
      "num_input_tokens_seen": 17332832,
      "step": 26755
    },
    {
      "epoch": 15.778301886792454,
      "grad_norm": 5.671070575714111,
      "learning_rate": 1.297318354343799e-06,
      "loss": 0.359,
      "num_input_tokens_seen": 17334880,
      "step": 26760
    },
    {
      "epoch": 15.78125,
      "grad_norm": 3.7342817783355713,
      "learning_rate": 1.295589938681317e-06,
      "loss": 0.4671,
      "num_input_tokens_seen": 17337984,
      "step": 26765
    },
    {
      "epoch": 15.784198113207546,
      "grad_norm": 4.051829814910889,
      "learning_rate": 1.2938625037773628e-06,
      "loss": 0.3716,
      "num_input_tokens_seen": 17340544,
      "step": 26770
    },
    {
      "epoch": 15.787146226415095,
      "grad_norm": 5.199432849884033,
      "learning_rate": 1.2921360500892843e-06,
      "loss": 0.4946,
      "num_input_tokens_seen": 17343360,
      "step": 26775
    },
    {
      "epoch": 15.790094339622641,
      "grad_norm": 2.751394748687744,
      "learning_rate": 1.290410578074167e-06,
      "loss": 0.335,
      "num_input_tokens_seen": 17346368,
      "step": 26780
    },
    {
      "epoch": 15.79304245283019,
      "grad_norm": 4.846244812011719,
      "learning_rate": 1.2886860881888362e-06,
      "loss": 0.2913,
      "num_input_tokens_seen": 17349248,
      "step": 26785
    },
    {
      "epoch": 15.795990566037736,
      "grad_norm": 7.798903465270996,
      "learning_rate": 1.2869625808898584e-06,
      "loss": 0.3199,
      "num_input_tokens_seen": 17352288,
      "step": 26790
    },
    {
      "epoch": 15.798938679245284,
      "grad_norm": 5.872567653656006,
      "learning_rate": 1.2852400566335398e-06,
      "loss": 0.3723,
      "num_input_tokens_seen": 17359072,
      "step": 26795
    },
    {
      "epoch": 15.80188679245283,
      "grad_norm": 4.015979766845703,
      "learning_rate": 1.2835185158759244e-06,
      "loss": 0.3173,
      "num_input_tokens_seen": 17362624,
      "step": 26800
    },
    {
      "epoch": 15.804834905660378,
      "grad_norm": 2.5459954738616943,
      "learning_rate": 1.2817979590728009e-06,
      "loss": 0.293,
      "num_input_tokens_seen": 17365344,
      "step": 26805
    },
    {
      "epoch": 15.807783018867925,
      "grad_norm": 3.5922505855560303,
      "learning_rate": 1.2800783866796918e-06,
      "loss": 0.3279,
      "num_input_tokens_seen": 17368544,
      "step": 26810
    },
    {
      "epoch": 15.810731132075471,
      "grad_norm": 3.695155382156372,
      "learning_rate": 1.2783597991518604e-06,
      "loss": 0.2422,
      "num_input_tokens_seen": 17372544,
      "step": 26815
    },
    {
      "epoch": 15.81367924528302,
      "grad_norm": 5.272406578063965,
      "learning_rate": 1.2766421969443131e-06,
      "loss": 0.2617,
      "num_input_tokens_seen": 17375648,
      "step": 26820
    },
    {
      "epoch": 15.816627358490566,
      "grad_norm": 2.7506659030914307,
      "learning_rate": 1.274925580511791e-06,
      "loss": 0.3581,
      "num_input_tokens_seen": 17378784,
      "step": 26825
    },
    {
      "epoch": 15.819575471698114,
      "grad_norm": 3.138550281524658,
      "learning_rate": 1.2732099503087757e-06,
      "loss": 0.2877,
      "num_input_tokens_seen": 17381888,
      "step": 26830
    },
    {
      "epoch": 15.82252358490566,
      "grad_norm": 4.4250335693359375,
      "learning_rate": 1.2714953067894859e-06,
      "loss": 0.2691,
      "num_input_tokens_seen": 17384992,
      "step": 26835
    },
    {
      "epoch": 15.825471698113208,
      "grad_norm": 2.2097361087799072,
      "learning_rate": 1.2697816504078847e-06,
      "loss": 0.3292,
      "num_input_tokens_seen": 17387808,
      "step": 26840
    },
    {
      "epoch": 15.828419811320755,
      "grad_norm": 3.9476780891418457,
      "learning_rate": 1.2680689816176672e-06,
      "loss": 0.2679,
      "num_input_tokens_seen": 17390912,
      "step": 26845
    },
    {
      "epoch": 15.831367924528301,
      "grad_norm": 2.7565553188323975,
      "learning_rate": 1.2663573008722707e-06,
      "loss": 0.3375,
      "num_input_tokens_seen": 17393376,
      "step": 26850
    },
    {
      "epoch": 15.83431603773585,
      "grad_norm": 2.643336534500122,
      "learning_rate": 1.2646466086248698e-06,
      "loss": 0.2257,
      "num_input_tokens_seen": 17396704,
      "step": 26855
    },
    {
      "epoch": 15.837264150943396,
      "grad_norm": 3.610112428665161,
      "learning_rate": 1.2629369053283779e-06,
      "loss": 0.2729,
      "num_input_tokens_seen": 17399808,
      "step": 26860
    },
    {
      "epoch": 15.840212264150944,
      "grad_norm": 4.190951824188232,
      "learning_rate": 1.2612281914354452e-06,
      "loss": 0.3512,
      "num_input_tokens_seen": 17403584,
      "step": 26865
    },
    {
      "epoch": 15.84316037735849,
      "grad_norm": 4.991774082183838,
      "learning_rate": 1.259520467398463e-06,
      "loss": 0.4997,
      "num_input_tokens_seen": 17406144,
      "step": 26870
    },
    {
      "epoch": 15.846108490566039,
      "grad_norm": 4.085575580596924,
      "learning_rate": 1.2578137336695573e-06,
      "loss": 0.4043,
      "num_input_tokens_seen": 17409792,
      "step": 26875
    },
    {
      "epoch": 15.849056603773585,
      "grad_norm": 4.059009075164795,
      "learning_rate": 1.256107990700594e-06,
      "loss": 0.252,
      "num_input_tokens_seen": 17413952,
      "step": 26880
    },
    {
      "epoch": 15.852004716981131,
      "grad_norm": 2.7124078273773193,
      "learning_rate": 1.2544032389431753e-06,
      "loss": 0.3298,
      "num_input_tokens_seen": 17418144,
      "step": 26885
    },
    {
      "epoch": 15.85495283018868,
      "grad_norm": 3.940809965133667,
      "learning_rate": 1.2526994788486418e-06,
      "loss": 0.3366,
      "num_input_tokens_seen": 17422048,
      "step": 26890
    },
    {
      "epoch": 15.857900943396226,
      "grad_norm": 3.2970075607299805,
      "learning_rate": 1.2509967108680697e-06,
      "loss": 0.4605,
      "num_input_tokens_seen": 17424672,
      "step": 26895
    },
    {
      "epoch": 15.860849056603774,
      "grad_norm": 4.291810989379883,
      "learning_rate": 1.249294935452277e-06,
      "loss": 0.2668,
      "num_input_tokens_seen": 17427744,
      "step": 26900
    },
    {
      "epoch": 15.86379716981132,
      "grad_norm": 3.568946599960327,
      "learning_rate": 1.247594153051815e-06,
      "loss": 0.2668,
      "num_input_tokens_seen": 17430976,
      "step": 26905
    },
    {
      "epoch": 15.866745283018869,
      "grad_norm": 2.9104421138763428,
      "learning_rate": 1.2458943641169718e-06,
      "loss": 0.2699,
      "num_input_tokens_seen": 17433344,
      "step": 26910
    },
    {
      "epoch": 15.869693396226415,
      "grad_norm": 5.00881290435791,
      "learning_rate": 1.2441955690977758e-06,
      "loss": 0.2908,
      "num_input_tokens_seen": 17436544,
      "step": 26915
    },
    {
      "epoch": 15.872641509433961,
      "grad_norm": 8.358563423156738,
      "learning_rate": 1.2424977684439898e-06,
      "loss": 0.4079,
      "num_input_tokens_seen": 17439968,
      "step": 26920
    },
    {
      "epoch": 15.87558962264151,
      "grad_norm": 10.848978042602539,
      "learning_rate": 1.2408009626051137e-06,
      "loss": 0.5067,
      "num_input_tokens_seen": 17442688,
      "step": 26925
    },
    {
      "epoch": 15.878537735849056,
      "grad_norm": 4.614745140075684,
      "learning_rate": 1.2391051520303826e-06,
      "loss": 0.2583,
      "num_input_tokens_seen": 17445600,
      "step": 26930
    },
    {
      "epoch": 15.881485849056604,
      "grad_norm": 3.097745895385742,
      "learning_rate": 1.2374103371687723e-06,
      "loss": 0.3479,
      "num_input_tokens_seen": 17449824,
      "step": 26935
    },
    {
      "epoch": 15.88443396226415,
      "grad_norm": 3.0365865230560303,
      "learning_rate": 1.2357165184689906e-06,
      "loss": 0.2764,
      "num_input_tokens_seen": 17452832,
      "step": 26940
    },
    {
      "epoch": 15.887382075471699,
      "grad_norm": 4.625044345855713,
      "learning_rate": 1.2340236963794845e-06,
      "loss": 0.304,
      "num_input_tokens_seen": 17455936,
      "step": 26945
    },
    {
      "epoch": 15.890330188679245,
      "grad_norm": 3.9439210891723633,
      "learning_rate": 1.232331871348435e-06,
      "loss": 0.3409,
      "num_input_tokens_seen": 17459360,
      "step": 26950
    },
    {
      "epoch": 15.893278301886792,
      "grad_norm": 3.2575843334198,
      "learning_rate": 1.2306410438237603e-06,
      "loss": 0.4476,
      "num_input_tokens_seen": 17462720,
      "step": 26955
    },
    {
      "epoch": 15.89622641509434,
      "grad_norm": 3.975658416748047,
      "learning_rate": 1.228951214253113e-06,
      "loss": 0.2576,
      "num_input_tokens_seen": 17465408,
      "step": 26960
    },
    {
      "epoch": 15.899174528301886,
      "grad_norm": 5.0862507820129395,
      "learning_rate": 1.2272623830838854e-06,
      "loss": 0.2221,
      "num_input_tokens_seen": 17469632,
      "step": 26965
    },
    {
      "epoch": 15.902122641509434,
      "grad_norm": 4.854150772094727,
      "learning_rate": 1.2255745507632016e-06,
      "loss": 0.2854,
      "num_input_tokens_seen": 17471808,
      "step": 26970
    },
    {
      "epoch": 15.90507075471698,
      "grad_norm": 5.437078475952148,
      "learning_rate": 1.223887717737922e-06,
      "loss": 0.3039,
      "num_input_tokens_seen": 17474400,
      "step": 26975
    },
    {
      "epoch": 15.908018867924529,
      "grad_norm": 4.545047760009766,
      "learning_rate": 1.2222018844546434e-06,
      "loss": 0.2472,
      "num_input_tokens_seen": 17477376,
      "step": 26980
    },
    {
      "epoch": 15.910966981132075,
      "grad_norm": 3.6365151405334473,
      "learning_rate": 1.2205170513596975e-06,
      "loss": 0.4314,
      "num_input_tokens_seen": 17481120,
      "step": 26985
    },
    {
      "epoch": 15.913915094339622,
      "grad_norm": 3.2545883655548096,
      "learning_rate": 1.2188332188991493e-06,
      "loss": 0.3594,
      "num_input_tokens_seen": 17484128,
      "step": 26990
    },
    {
      "epoch": 15.91686320754717,
      "grad_norm": 2.684576988220215,
      "learning_rate": 1.217150387518804e-06,
      "loss": 0.314,
      "num_input_tokens_seen": 17487776,
      "step": 26995
    },
    {
      "epoch": 15.919811320754716,
      "grad_norm": 3.4362804889678955,
      "learning_rate": 1.2154685576641967e-06,
      "loss": 0.2578,
      "num_input_tokens_seen": 17491488,
      "step": 27000
    },
    {
      "epoch": 15.922759433962264,
      "grad_norm": 6.353302001953125,
      "learning_rate": 1.2137877297805972e-06,
      "loss": 0.4109,
      "num_input_tokens_seen": 17494208,
      "step": 27005
    },
    {
      "epoch": 15.92570754716981,
      "grad_norm": 3.9034817218780518,
      "learning_rate": 1.2121079043130162e-06,
      "loss": 0.3558,
      "num_input_tokens_seen": 17497440,
      "step": 27010
    },
    {
      "epoch": 15.928655660377359,
      "grad_norm": 2.0721793174743652,
      "learning_rate": 1.210429081706192e-06,
      "loss": 0.3408,
      "num_input_tokens_seen": 17501984,
      "step": 27015
    },
    {
      "epoch": 15.931603773584905,
      "grad_norm": 2.0982859134674072,
      "learning_rate": 1.2087512624046005e-06,
      "loss": 0.3122,
      "num_input_tokens_seen": 17505504,
      "step": 27020
    },
    {
      "epoch": 15.934551886792454,
      "grad_norm": 3.774824380874634,
      "learning_rate": 1.2070744468524503e-06,
      "loss": 0.3814,
      "num_input_tokens_seen": 17509152,
      "step": 27025
    },
    {
      "epoch": 15.9375,
      "grad_norm": 4.076981067657471,
      "learning_rate": 1.2053986354936887e-06,
      "loss": 0.2965,
      "num_input_tokens_seen": 17512032,
      "step": 27030
    },
    {
      "epoch": 15.940448113207546,
      "grad_norm": 2.61836576461792,
      "learning_rate": 1.2037238287719916e-06,
      "loss": 0.391,
      "num_input_tokens_seen": 17515776,
      "step": 27035
    },
    {
      "epoch": 15.943396226415095,
      "grad_norm": 3.1491503715515137,
      "learning_rate": 1.2020500271307721e-06,
      "loss": 0.2358,
      "num_input_tokens_seen": 17518944,
      "step": 27040
    },
    {
      "epoch": 15.946344339622641,
      "grad_norm": 3.4753949642181396,
      "learning_rate": 1.200377231013176e-06,
      "loss": 0.2796,
      "num_input_tokens_seen": 17523168,
      "step": 27045
    },
    {
      "epoch": 15.94929245283019,
      "grad_norm": 2.747499942779541,
      "learning_rate": 1.1987054408620825e-06,
      "loss": 0.3057,
      "num_input_tokens_seen": 17526016,
      "step": 27050
    },
    {
      "epoch": 15.952240566037736,
      "grad_norm": 5.436689853668213,
      "learning_rate": 1.197034657120107e-06,
      "loss": 0.354,
      "num_input_tokens_seen": 17528768,
      "step": 27055
    },
    {
      "epoch": 15.955188679245284,
      "grad_norm": 3.1261820793151855,
      "learning_rate": 1.1953648802295964e-06,
      "loss": 0.2668,
      "num_input_tokens_seen": 17531584,
      "step": 27060
    },
    {
      "epoch": 15.95813679245283,
      "grad_norm": 3.1110966205596924,
      "learning_rate": 1.1936961106326307e-06,
      "loss": 0.1863,
      "num_input_tokens_seen": 17534240,
      "step": 27065
    },
    {
      "epoch": 15.961084905660378,
      "grad_norm": 4.700430870056152,
      "learning_rate": 1.1920283487710237e-06,
      "loss": 0.3939,
      "num_input_tokens_seen": 17537536,
      "step": 27070
    },
    {
      "epoch": 15.964033018867925,
      "grad_norm": 2.8127541542053223,
      "learning_rate": 1.1903615950863228e-06,
      "loss": 0.2968,
      "num_input_tokens_seen": 17540896,
      "step": 27075
    },
    {
      "epoch": 15.966981132075471,
      "grad_norm": 5.032728672027588,
      "learning_rate": 1.1886958500198076e-06,
      "loss": 0.3648,
      "num_input_tokens_seen": 17544480,
      "step": 27080
    },
    {
      "epoch": 15.96992924528302,
      "grad_norm": 3.4827985763549805,
      "learning_rate": 1.1870311140124923e-06,
      "loss": 0.2302,
      "num_input_tokens_seen": 17547008,
      "step": 27085
    },
    {
      "epoch": 15.972877358490566,
      "grad_norm": 3.533801794052124,
      "learning_rate": 1.185367387505123e-06,
      "loss": 0.3591,
      "num_input_tokens_seen": 17551488,
      "step": 27090
    },
    {
      "epoch": 15.975825471698114,
      "grad_norm": 2.7736146450042725,
      "learning_rate": 1.1837046709381783e-06,
      "loss": 0.2637,
      "num_input_tokens_seen": 17554176,
      "step": 27095
    },
    {
      "epoch": 15.97877358490566,
      "grad_norm": 3.7022311687469482,
      "learning_rate": 1.1820429647518678e-06,
      "loss": 0.2918,
      "num_input_tokens_seen": 17557568,
      "step": 27100
    },
    {
      "epoch": 15.981721698113208,
      "grad_norm": 2.434502363204956,
      "learning_rate": 1.1803822693861377e-06,
      "loss": 0.481,
      "num_input_tokens_seen": 17561152,
      "step": 27105
    },
    {
      "epoch": 15.984669811320755,
      "grad_norm": 3.5557007789611816,
      "learning_rate": 1.1787225852806639e-06,
      "loss": 0.3369,
      "num_input_tokens_seen": 17564224,
      "step": 27110
    },
    {
      "epoch": 15.987617924528301,
      "grad_norm": 3.782581090927124,
      "learning_rate": 1.177063912874853e-06,
      "loss": 0.3134,
      "num_input_tokens_seen": 17567328,
      "step": 27115
    },
    {
      "epoch": 15.99056603773585,
      "grad_norm": 1.750786304473877,
      "learning_rate": 1.1754062526078487e-06,
      "loss": 0.2961,
      "num_input_tokens_seen": 17570720,
      "step": 27120
    },
    {
      "epoch": 15.993514150943396,
      "grad_norm": 5.049679756164551,
      "learning_rate": 1.1737496049185215e-06,
      "loss": 0.263,
      "num_input_tokens_seen": 17573184,
      "step": 27125
    },
    {
      "epoch": 15.996462264150944,
      "grad_norm": 2.182346820831299,
      "learning_rate": 1.172093970245477e-06,
      "loss": 0.2873,
      "num_input_tokens_seen": 17577376,
      "step": 27130
    },
    {
      "epoch": 15.99941037735849,
      "grad_norm": 2.7458598613739014,
      "learning_rate": 1.1704393490270516e-06,
      "loss": 0.4434,
      "num_input_tokens_seen": 17580096,
      "step": 27135
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.6057673692703247,
      "eval_runtime": 19.5383,
      "eval_samples_per_second": 86.804,
      "eval_steps_per_second": 21.701,
      "num_input_tokens_seen": 17580200,
      "step": 27136
    },
    {
      "epoch": 16.00235849056604,
      "grad_norm": 4.184218883514404,
      "learning_rate": 1.1687857417013126e-06,
      "loss": 0.2547,
      "num_input_tokens_seen": 17582248,
      "step": 27140
    },
    {
      "epoch": 16.005306603773583,
      "grad_norm": 7.956359386444092,
      "learning_rate": 1.1671331487060583e-06,
      "loss": 0.3103,
      "num_input_tokens_seen": 17585224,
      "step": 27145
    },
    {
      "epoch": 16.00825471698113,
      "grad_norm": 3.5101585388183594,
      "learning_rate": 1.1654815704788237e-06,
      "loss": 0.4271,
      "num_input_tokens_seen": 17588424,
      "step": 27150
    },
    {
      "epoch": 16.01120283018868,
      "grad_norm": 10.184524536132812,
      "learning_rate": 1.1638310074568687e-06,
      "loss": 0.3776,
      "num_input_tokens_seen": 17591208,
      "step": 27155
    },
    {
      "epoch": 16.014150943396228,
      "grad_norm": 2.552238941192627,
      "learning_rate": 1.162181460077188e-06,
      "loss": 0.3797,
      "num_input_tokens_seen": 17595368,
      "step": 27160
    },
    {
      "epoch": 16.017099056603772,
      "grad_norm": 4.382515907287598,
      "learning_rate": 1.1605329287765056e-06,
      "loss": 0.3152,
      "num_input_tokens_seen": 17598440,
      "step": 27165
    },
    {
      "epoch": 16.02004716981132,
      "grad_norm": 3.4729714393615723,
      "learning_rate": 1.1588854139912775e-06,
      "loss": 0.3911,
      "num_input_tokens_seen": 17601896,
      "step": 27170
    },
    {
      "epoch": 16.02299528301887,
      "grad_norm": 2.9481418132781982,
      "learning_rate": 1.1572389161576886e-06,
      "loss": 0.2913,
      "num_input_tokens_seen": 17604936,
      "step": 27175
    },
    {
      "epoch": 16.025943396226417,
      "grad_norm": 2.3009066581726074,
      "learning_rate": 1.15559343571166e-06,
      "loss": 0.206,
      "num_input_tokens_seen": 17607848,
      "step": 27180
    },
    {
      "epoch": 16.02889150943396,
      "grad_norm": 4.223967552185059,
      "learning_rate": 1.153948973088837e-06,
      "loss": 0.305,
      "num_input_tokens_seen": 17611592,
      "step": 27185
    },
    {
      "epoch": 16.03183962264151,
      "grad_norm": 2.8828554153442383,
      "learning_rate": 1.1523055287245993e-06,
      "loss": 0.3161,
      "num_input_tokens_seen": 17616328,
      "step": 27190
    },
    {
      "epoch": 16.034787735849058,
      "grad_norm": 3.219686508178711,
      "learning_rate": 1.150663103054056e-06,
      "loss": 0.3286,
      "num_input_tokens_seen": 17619464,
      "step": 27195
    },
    {
      "epoch": 16.037735849056602,
      "grad_norm": 4.3219709396362305,
      "learning_rate": 1.1490216965120438e-06,
      "loss": 0.4307,
      "num_input_tokens_seen": 17622120,
      "step": 27200
    },
    {
      "epoch": 16.04068396226415,
      "grad_norm": 6.118250370025635,
      "learning_rate": 1.147381309533136e-06,
      "loss": 0.2656,
      "num_input_tokens_seen": 17624968,
      "step": 27205
    },
    {
      "epoch": 16.0436320754717,
      "grad_norm": 5.800889015197754,
      "learning_rate": 1.1457419425516287e-06,
      "loss": 0.3267,
      "num_input_tokens_seen": 17627688,
      "step": 27210
    },
    {
      "epoch": 16.046580188679247,
      "grad_norm": 3.6308255195617676,
      "learning_rate": 1.1441035960015544e-06,
      "loss": 0.2771,
      "num_input_tokens_seen": 17630888,
      "step": 27215
    },
    {
      "epoch": 16.04952830188679,
      "grad_norm": 2.9472644329071045,
      "learning_rate": 1.1424662703166716e-06,
      "loss": 0.3036,
      "num_input_tokens_seen": 17633224,
      "step": 27220
    },
    {
      "epoch": 16.05247641509434,
      "grad_norm": 2.712373971939087,
      "learning_rate": 1.1408299659304684e-06,
      "loss": 0.2274,
      "num_input_tokens_seen": 17636680,
      "step": 27225
    },
    {
      "epoch": 16.055424528301888,
      "grad_norm": 2.0542490482330322,
      "learning_rate": 1.1391946832761642e-06,
      "loss": 0.257,
      "num_input_tokens_seen": 17643336,
      "step": 27230
    },
    {
      "epoch": 16.058372641509433,
      "grad_norm": 5.028131484985352,
      "learning_rate": 1.137560422786706e-06,
      "loss": 0.3469,
      "num_input_tokens_seen": 17646632,
      "step": 27235
    },
    {
      "epoch": 16.06132075471698,
      "grad_norm": 5.726007461547852,
      "learning_rate": 1.1359271848947712e-06,
      "loss": 0.2902,
      "num_input_tokens_seen": 17650600,
      "step": 27240
    },
    {
      "epoch": 16.06426886792453,
      "grad_norm": 7.83866024017334,
      "learning_rate": 1.1342949700327688e-06,
      "loss": 0.3543,
      "num_input_tokens_seen": 17654632,
      "step": 27245
    },
    {
      "epoch": 16.067216981132077,
      "grad_norm": 4.05454158782959,
      "learning_rate": 1.1326637786328332e-06,
      "loss": 0.3599,
      "num_input_tokens_seen": 17657416,
      "step": 27250
    },
    {
      "epoch": 16.07016509433962,
      "grad_norm": 3.7377126216888428,
      "learning_rate": 1.1310336111268293e-06,
      "loss": 0.2734,
      "num_input_tokens_seen": 17660456,
      "step": 27255
    },
    {
      "epoch": 16.07311320754717,
      "grad_norm": 2.746337413787842,
      "learning_rate": 1.1294044679463517e-06,
      "loss": 0.2732,
      "num_input_tokens_seen": 17663848,
      "step": 27260
    },
    {
      "epoch": 16.076061320754718,
      "grad_norm": 3.162679433822632,
      "learning_rate": 1.1277763495227207e-06,
      "loss": 0.2547,
      "num_input_tokens_seen": 17667976,
      "step": 27265
    },
    {
      "epoch": 16.079009433962263,
      "grad_norm": 4.2071213722229,
      "learning_rate": 1.1261492562869913e-06,
      "loss": 0.3466,
      "num_input_tokens_seen": 17670696,
      "step": 27270
    },
    {
      "epoch": 16.08195754716981,
      "grad_norm": 1.971596121788025,
      "learning_rate": 1.1245231886699415e-06,
      "loss": 0.2534,
      "num_input_tokens_seen": 17674216,
      "step": 27275
    },
    {
      "epoch": 16.08490566037736,
      "grad_norm": 3.2433395385742188,
      "learning_rate": 1.12289814710208e-06,
      "loss": 0.3007,
      "num_input_tokens_seen": 17677032,
      "step": 27280
    },
    {
      "epoch": 16.087853773584907,
      "grad_norm": 2.7392632961273193,
      "learning_rate": 1.1212741320136433e-06,
      "loss": 0.4816,
      "num_input_tokens_seen": 17681832,
      "step": 27285
    },
    {
      "epoch": 16.090801886792452,
      "grad_norm": 5.187206745147705,
      "learning_rate": 1.1196511438345963e-06,
      "loss": 0.2935,
      "num_input_tokens_seen": 17684264,
      "step": 27290
    },
    {
      "epoch": 16.09375,
      "grad_norm": 10.906784057617188,
      "learning_rate": 1.118029182994631e-06,
      "loss": 0.3287,
      "num_input_tokens_seen": 17687144,
      "step": 27295
    },
    {
      "epoch": 16.096698113207548,
      "grad_norm": 2.6188037395477295,
      "learning_rate": 1.1164082499231704e-06,
      "loss": 0.2935,
      "num_input_tokens_seen": 17689896,
      "step": 27300
    },
    {
      "epoch": 16.099646226415093,
      "grad_norm": 2.7862820625305176,
      "learning_rate": 1.114788345049364e-06,
      "loss": 0.2432,
      "num_input_tokens_seen": 17693608,
      "step": 27305
    },
    {
      "epoch": 16.10259433962264,
      "grad_norm": 2.8436059951782227,
      "learning_rate": 1.1131694688020872e-06,
      "loss": 0.3502,
      "num_input_tokens_seen": 17696264,
      "step": 27310
    },
    {
      "epoch": 16.10554245283019,
      "grad_norm": 3.532498598098755,
      "learning_rate": 1.1115516216099453e-06,
      "loss": 0.2167,
      "num_input_tokens_seen": 17699144,
      "step": 27315
    },
    {
      "epoch": 16.108490566037737,
      "grad_norm": 2.9326250553131104,
      "learning_rate": 1.1099348039012698e-06,
      "loss": 0.2335,
      "num_input_tokens_seen": 17702440,
      "step": 27320
    },
    {
      "epoch": 16.111438679245282,
      "grad_norm": 9.101142883300781,
      "learning_rate": 1.1083190161041202e-06,
      "loss": 0.361,
      "num_input_tokens_seen": 17705320,
      "step": 27325
    },
    {
      "epoch": 16.11438679245283,
      "grad_norm": 3.6656346321105957,
      "learning_rate": 1.1067042586462822e-06,
      "loss": 0.2503,
      "num_input_tokens_seen": 17708520,
      "step": 27330
    },
    {
      "epoch": 16.11733490566038,
      "grad_norm": 3.2562835216522217,
      "learning_rate": 1.1050905319552718e-06,
      "loss": 0.279,
      "num_input_tokens_seen": 17711240,
      "step": 27335
    },
    {
      "epoch": 16.120283018867923,
      "grad_norm": 3.373202323913574,
      "learning_rate": 1.1034778364583293e-06,
      "loss": 0.4532,
      "num_input_tokens_seen": 17714440,
      "step": 27340
    },
    {
      "epoch": 16.12323113207547,
      "grad_norm": 2.6655631065368652,
      "learning_rate": 1.1018661725824231e-06,
      "loss": 0.3622,
      "num_input_tokens_seen": 17720008,
      "step": 27345
    },
    {
      "epoch": 16.12617924528302,
      "grad_norm": 3.3919928073883057,
      "learning_rate": 1.100255540754247e-06,
      "loss": 0.3434,
      "num_input_tokens_seen": 17722952,
      "step": 27350
    },
    {
      "epoch": 16.129127358490567,
      "grad_norm": 1.9298781156539917,
      "learning_rate": 1.0986459414002244e-06,
      "loss": 0.2026,
      "num_input_tokens_seen": 17725256,
      "step": 27355
    },
    {
      "epoch": 16.132075471698112,
      "grad_norm": 4.827513694763184,
      "learning_rate": 1.0970373749465008e-06,
      "loss": 0.3419,
      "num_input_tokens_seen": 17727784,
      "step": 27360
    },
    {
      "epoch": 16.13502358490566,
      "grad_norm": 5.942607402801514,
      "learning_rate": 1.095429841818954e-06,
      "loss": 0.412,
      "num_input_tokens_seen": 17731752,
      "step": 27365
    },
    {
      "epoch": 16.13797169811321,
      "grad_norm": 4.2184672355651855,
      "learning_rate": 1.093823342443185e-06,
      "loss": 0.1976,
      "num_input_tokens_seen": 17735272,
      "step": 27370
    },
    {
      "epoch": 16.140919811320753,
      "grad_norm": 4.215981483459473,
      "learning_rate": 1.0922178772445203e-06,
      "loss": 0.5631,
      "num_input_tokens_seen": 17738280,
      "step": 27375
    },
    {
      "epoch": 16.1438679245283,
      "grad_norm": 3.7190253734588623,
      "learning_rate": 1.0906134466480146e-06,
      "loss": 0.4631,
      "num_input_tokens_seen": 17742024,
      "step": 27380
    },
    {
      "epoch": 16.14681603773585,
      "grad_norm": 5.701833724975586,
      "learning_rate": 1.0890100510784473e-06,
      "loss": 0.2806,
      "num_input_tokens_seen": 17744712,
      "step": 27385
    },
    {
      "epoch": 16.149764150943398,
      "grad_norm": 5.562639236450195,
      "learning_rate": 1.0874076909603227e-06,
      "loss": 0.29,
      "num_input_tokens_seen": 17747976,
      "step": 27390
    },
    {
      "epoch": 16.152712264150942,
      "grad_norm": 3.233381748199463,
      "learning_rate": 1.0858063667178747e-06,
      "loss": 0.4381,
      "num_input_tokens_seen": 17751112,
      "step": 27395
    },
    {
      "epoch": 16.15566037735849,
      "grad_norm": 3.9325456619262695,
      "learning_rate": 1.0842060787750614e-06,
      "loss": 0.323,
      "num_input_tokens_seen": 17754280,
      "step": 27400
    },
    {
      "epoch": 16.15860849056604,
      "grad_norm": 2.2433865070343018,
      "learning_rate": 1.0826068275555652e-06,
      "loss": 0.3865,
      "num_input_tokens_seen": 17757480,
      "step": 27405
    },
    {
      "epoch": 16.161556603773583,
      "grad_norm": 2.6093430519104004,
      "learning_rate": 1.081008613482794e-06,
      "loss": 0.2933,
      "num_input_tokens_seen": 17760552,
      "step": 27410
    },
    {
      "epoch": 16.16450471698113,
      "grad_norm": 3.669687032699585,
      "learning_rate": 1.079411436979883e-06,
      "loss": 0.2513,
      "num_input_tokens_seen": 17763688,
      "step": 27415
    },
    {
      "epoch": 16.16745283018868,
      "grad_norm": 7.580700397491455,
      "learning_rate": 1.0778152984696905e-06,
      "loss": 0.3631,
      "num_input_tokens_seen": 17766824,
      "step": 27420
    },
    {
      "epoch": 16.170400943396228,
      "grad_norm": 2.8181111812591553,
      "learning_rate": 1.0762201983747993e-06,
      "loss": 0.1988,
      "num_input_tokens_seen": 17770856,
      "step": 27425
    },
    {
      "epoch": 16.173349056603772,
      "grad_norm": 2.855576515197754,
      "learning_rate": 1.0746261371175238e-06,
      "loss": 0.2404,
      "num_input_tokens_seen": 17774216,
      "step": 27430
    },
    {
      "epoch": 16.17629716981132,
      "grad_norm": 2.9444730281829834,
      "learning_rate": 1.0730331151198953e-06,
      "loss": 0.4185,
      "num_input_tokens_seen": 17778024,
      "step": 27435
    },
    {
      "epoch": 16.17924528301887,
      "grad_norm": 4.344099521636963,
      "learning_rate": 1.0714411328036733e-06,
      "loss": 0.359,
      "num_input_tokens_seen": 17781224,
      "step": 27440
    },
    {
      "epoch": 16.182193396226417,
      "grad_norm": 3.663296699523926,
      "learning_rate": 1.0698501905903435e-06,
      "loss": 0.4033,
      "num_input_tokens_seen": 17784712,
      "step": 27445
    },
    {
      "epoch": 16.18514150943396,
      "grad_norm": 3.922592878341675,
      "learning_rate": 1.0682602889011134e-06,
      "loss": 0.2135,
      "num_input_tokens_seen": 17787528,
      "step": 27450
    },
    {
      "epoch": 16.18808962264151,
      "grad_norm": 3.746490716934204,
      "learning_rate": 1.0666714281569152e-06,
      "loss": 0.3144,
      "num_input_tokens_seen": 17790312,
      "step": 27455
    },
    {
      "epoch": 16.191037735849058,
      "grad_norm": 3.4107210636138916,
      "learning_rate": 1.0650836087784095e-06,
      "loss": 0.2804,
      "num_input_tokens_seen": 17793480,
      "step": 27460
    },
    {
      "epoch": 16.193985849056602,
      "grad_norm": 3.339214324951172,
      "learning_rate": 1.0634968311859768e-06,
      "loss": 0.5277,
      "num_input_tokens_seen": 17796680,
      "step": 27465
    },
    {
      "epoch": 16.19693396226415,
      "grad_norm": 7.268819332122803,
      "learning_rate": 1.0619110957997237e-06,
      "loss": 0.1966,
      "num_input_tokens_seen": 17799656,
      "step": 27470
    },
    {
      "epoch": 16.1998820754717,
      "grad_norm": 3.207951068878174,
      "learning_rate": 1.06032640303948e-06,
      "loss": 0.2556,
      "num_input_tokens_seen": 17802664,
      "step": 27475
    },
    {
      "epoch": 16.202830188679247,
      "grad_norm": 3.9205098152160645,
      "learning_rate": 1.0587427533248002e-06,
      "loss": 0.2825,
      "num_input_tokens_seen": 17805448,
      "step": 27480
    },
    {
      "epoch": 16.20577830188679,
      "grad_norm": 2.366485595703125,
      "learning_rate": 1.057160147074961e-06,
      "loss": 0.2377,
      "num_input_tokens_seen": 17808648,
      "step": 27485
    },
    {
      "epoch": 16.20872641509434,
      "grad_norm": 3.4176909923553467,
      "learning_rate": 1.0555785847089657e-06,
      "loss": 0.2697,
      "num_input_tokens_seen": 17811720,
      "step": 27490
    },
    {
      "epoch": 16.211674528301888,
      "grad_norm": 2.6589159965515137,
      "learning_rate": 1.0539980666455407e-06,
      "loss": 0.2855,
      "num_input_tokens_seen": 17815560,
      "step": 27495
    },
    {
      "epoch": 16.214622641509433,
      "grad_norm": 3.045243501663208,
      "learning_rate": 1.052418593303134e-06,
      "loss": 0.1615,
      "num_input_tokens_seen": 17819176,
      "step": 27500
    },
    {
      "epoch": 16.21757075471698,
      "grad_norm": 5.7529778480529785,
      "learning_rate": 1.0508401650999178e-06,
      "loss": 0.3392,
      "num_input_tokens_seen": 17822280,
      "step": 27505
    },
    {
      "epoch": 16.22051886792453,
      "grad_norm": 3.213437080383301,
      "learning_rate": 1.0492627824537877e-06,
      "loss": 0.4474,
      "num_input_tokens_seen": 17825384,
      "step": 27510
    },
    {
      "epoch": 16.223466981132077,
      "grad_norm": 4.9864912033081055,
      "learning_rate": 1.0476864457823626e-06,
      "loss": 0.3539,
      "num_input_tokens_seen": 17828744,
      "step": 27515
    },
    {
      "epoch": 16.22641509433962,
      "grad_norm": 4.279682636260986,
      "learning_rate": 1.0461111555029836e-06,
      "loss": 0.3919,
      "num_input_tokens_seen": 17831240,
      "step": 27520
    },
    {
      "epoch": 16.22936320754717,
      "grad_norm": 5.1897993087768555,
      "learning_rate": 1.0445369120327175e-06,
      "loss": 0.3487,
      "num_input_tokens_seen": 17834184,
      "step": 27525
    },
    {
      "epoch": 16.232311320754718,
      "grad_norm": 5.738783836364746,
      "learning_rate": 1.0429637157883516e-06,
      "loss": 0.2915,
      "num_input_tokens_seen": 17837096,
      "step": 27530
    },
    {
      "epoch": 16.235259433962263,
      "grad_norm": 3.0639259815216064,
      "learning_rate": 1.041391567186395e-06,
      "loss": 0.2843,
      "num_input_tokens_seen": 17839688,
      "step": 27535
    },
    {
      "epoch": 16.23820754716981,
      "grad_norm": 4.611005783081055,
      "learning_rate": 1.0398204666430821e-06,
      "loss": 0.3845,
      "num_input_tokens_seen": 17842664,
      "step": 27540
    },
    {
      "epoch": 16.24115566037736,
      "grad_norm": 3.7438132762908936,
      "learning_rate": 1.0382504145743667e-06,
      "loss": 0.1955,
      "num_input_tokens_seen": 17845896,
      "step": 27545
    },
    {
      "epoch": 16.244103773584907,
      "grad_norm": 3.297858715057373,
      "learning_rate": 1.0366814113959294e-06,
      "loss": 0.2827,
      "num_input_tokens_seen": 17849096,
      "step": 27550
    },
    {
      "epoch": 16.247051886792452,
      "grad_norm": 5.185736656188965,
      "learning_rate": 1.0351134575231697e-06,
      "loss": 0.3431,
      "num_input_tokens_seen": 17852232,
      "step": 27555
    },
    {
      "epoch": 16.25,
      "grad_norm": 5.522406101226807,
      "learning_rate": 1.0335465533712098e-06,
      "loss": 0.2803,
      "num_input_tokens_seen": 17856232,
      "step": 27560
    },
    {
      "epoch": 16.252948113207548,
      "grad_norm": 4.038296699523926,
      "learning_rate": 1.031980699354894e-06,
      "loss": 0.2864,
      "num_input_tokens_seen": 17858888,
      "step": 27565
    },
    {
      "epoch": 16.255896226415093,
      "grad_norm": 3.197166681289673,
      "learning_rate": 1.03041589588879e-06,
      "loss": 0.3742,
      "num_input_tokens_seen": 17861960,
      "step": 27570
    },
    {
      "epoch": 16.25884433962264,
      "grad_norm": 5.884000778198242,
      "learning_rate": 1.0288521433871834e-06,
      "loss": 0.3527,
      "num_input_tokens_seen": 17865736,
      "step": 27575
    },
    {
      "epoch": 16.26179245283019,
      "grad_norm": 4.06273078918457,
      "learning_rate": 1.0272894422640866e-06,
      "loss": 0.3554,
      "num_input_tokens_seen": 17869096,
      "step": 27580
    },
    {
      "epoch": 16.264740566037737,
      "grad_norm": 4.964377403259277,
      "learning_rate": 1.0257277929332332e-06,
      "loss": 0.3643,
      "num_input_tokens_seen": 17871176,
      "step": 27585
    },
    {
      "epoch": 16.267688679245282,
      "grad_norm": 2.5143985748291016,
      "learning_rate": 1.0241671958080745e-06,
      "loss": 0.2052,
      "num_input_tokens_seen": 17875208,
      "step": 27590
    },
    {
      "epoch": 16.27063679245283,
      "grad_norm": 3.582308053970337,
      "learning_rate": 1.0226076513017858e-06,
      "loss": 0.3092,
      "num_input_tokens_seen": 17878056,
      "step": 27595
    },
    {
      "epoch": 16.27358490566038,
      "grad_norm": 3.9291625022888184,
      "learning_rate": 1.0210491598272625e-06,
      "loss": 0.3384,
      "num_input_tokens_seen": 17880552,
      "step": 27600
    },
    {
      "epoch": 16.276533018867923,
      "grad_norm": 7.789273262023926,
      "learning_rate": 1.0194917217971229e-06,
      "loss": 0.2999,
      "num_input_tokens_seen": 17884968,
      "step": 27605
    },
    {
      "epoch": 16.27948113207547,
      "grad_norm": 5.061009883880615,
      "learning_rate": 1.0179353376237038e-06,
      "loss": 0.3018,
      "num_input_tokens_seen": 17887688,
      "step": 27610
    },
    {
      "epoch": 16.28242924528302,
      "grad_norm": 7.231390953063965,
      "learning_rate": 1.0163800077190672e-06,
      "loss": 0.3103,
      "num_input_tokens_seen": 17890792,
      "step": 27615
    },
    {
      "epoch": 16.285377358490567,
      "grad_norm": 5.139244556427002,
      "learning_rate": 1.0148257324949916e-06,
      "loss": 0.3584,
      "num_input_tokens_seen": 17893896,
      "step": 27620
    },
    {
      "epoch": 16.288325471698112,
      "grad_norm": 4.217301368713379,
      "learning_rate": 1.0132725123629783e-06,
      "loss": 0.4061,
      "num_input_tokens_seen": 17897320,
      "step": 27625
    },
    {
      "epoch": 16.29127358490566,
      "grad_norm": 4.6523942947387695,
      "learning_rate": 1.0117203477342497e-06,
      "loss": 0.3605,
      "num_input_tokens_seen": 17900552,
      "step": 27630
    },
    {
      "epoch": 16.29422169811321,
      "grad_norm": 4.8696441650390625,
      "learning_rate": 1.0101692390197477e-06,
      "loss": 0.2553,
      "num_input_tokens_seen": 17904136,
      "step": 27635
    },
    {
      "epoch": 16.297169811320753,
      "grad_norm": 3.921967029571533,
      "learning_rate": 1.0086191866301331e-06,
      "loss": 0.4094,
      "num_input_tokens_seen": 17907400,
      "step": 27640
    },
    {
      "epoch": 16.3001179245283,
      "grad_norm": 5.330355167388916,
      "learning_rate": 1.0070701909757918e-06,
      "loss": 0.2499,
      "num_input_tokens_seen": 17909960,
      "step": 27645
    },
    {
      "epoch": 16.30306603773585,
      "grad_norm": 4.334836006164551,
      "learning_rate": 1.0055222524668267e-06,
      "loss": 0.2935,
      "num_input_tokens_seen": 17912872,
      "step": 27650
    },
    {
      "epoch": 16.306014150943398,
      "grad_norm": 4.850845813751221,
      "learning_rate": 1.00397537151306e-06,
      "loss": 0.2906,
      "num_input_tokens_seen": 17916008,
      "step": 27655
    },
    {
      "epoch": 16.308962264150942,
      "grad_norm": 2.891495704650879,
      "learning_rate": 1.002429548524036e-06,
      "loss": 0.26,
      "num_input_tokens_seen": 17918312,
      "step": 27660
    },
    {
      "epoch": 16.31191037735849,
      "grad_norm": 5.309898376464844,
      "learning_rate": 1.0008847839090175e-06,
      "loss": 0.3197,
      "num_input_tokens_seen": 17920872,
      "step": 27665
    },
    {
      "epoch": 16.31485849056604,
      "grad_norm": 4.501356601715088,
      "learning_rate": 9.993410780769862e-07,
      "loss": 0.2548,
      "num_input_tokens_seen": 17923176,
      "step": 27670
    },
    {
      "epoch": 16.317806603773583,
      "grad_norm": 6.054188251495361,
      "learning_rate": 9.977984314366463e-07,
      "loss": 0.2874,
      "num_input_tokens_seen": 17926184,
      "step": 27675
    },
    {
      "epoch": 16.32075471698113,
      "grad_norm": 4.122568607330322,
      "learning_rate": 9.962568443964216e-07,
      "loss": 0.2906,
      "num_input_tokens_seen": 17928648,
      "step": 27680
    },
    {
      "epoch": 16.32370283018868,
      "grad_norm": 2.6499998569488525,
      "learning_rate": 9.947163173644524e-07,
      "loss": 0.4436,
      "num_input_tokens_seen": 17932264,
      "step": 27685
    },
    {
      "epoch": 16.326650943396228,
      "grad_norm": 3.5293831825256348,
      "learning_rate": 9.931768507486007e-07,
      "loss": 0.3146,
      "num_input_tokens_seen": 17935496,
      "step": 27690
    },
    {
      "epoch": 16.329599056603772,
      "grad_norm": 3.6315066814422607,
      "learning_rate": 9.916384449564453e-07,
      "loss": 0.2914,
      "num_input_tokens_seen": 17939336,
      "step": 27695
    },
    {
      "epoch": 16.33254716981132,
      "grad_norm": 5.112858295440674,
      "learning_rate": 9.90101100395287e-07,
      "loss": 0.2616,
      "num_input_tokens_seen": 17942120,
      "step": 27700
    },
    {
      "epoch": 16.33549528301887,
      "grad_norm": 4.220349311828613,
      "learning_rate": 9.885648174721428e-07,
      "loss": 0.3991,
      "num_input_tokens_seen": 17945448,
      "step": 27705
    },
    {
      "epoch": 16.338443396226417,
      "grad_norm": 3.379584550857544,
      "learning_rate": 9.870295965937532e-07,
      "loss": 0.3088,
      "num_input_tokens_seen": 17948200,
      "step": 27710
    },
    {
      "epoch": 16.34139150943396,
      "grad_norm": 3.586156129837036,
      "learning_rate": 9.854954381665727e-07,
      "loss": 0.274,
      "num_input_tokens_seen": 17951336,
      "step": 27715
    },
    {
      "epoch": 16.34433962264151,
      "grad_norm": 3.9054229259490967,
      "learning_rate": 9.83962342596776e-07,
      "loss": 0.3452,
      "num_input_tokens_seen": 17954440,
      "step": 27720
    },
    {
      "epoch": 16.347287735849058,
      "grad_norm": 3.212830066680908,
      "learning_rate": 9.824303102902576e-07,
      "loss": 0.3577,
      "num_input_tokens_seen": 17957832,
      "step": 27725
    },
    {
      "epoch": 16.350235849056602,
      "grad_norm": 3.9516336917877197,
      "learning_rate": 9.808993416526292e-07,
      "loss": 0.3222,
      "num_input_tokens_seen": 17961256,
      "step": 27730
    },
    {
      "epoch": 16.35318396226415,
      "grad_norm": 2.2458508014678955,
      "learning_rate": 9.793694370892204e-07,
      "loss": 0.2765,
      "num_input_tokens_seen": 17965192,
      "step": 27735
    },
    {
      "epoch": 16.3561320754717,
      "grad_norm": 3.3658223152160645,
      "learning_rate": 9.77840597005082e-07,
      "loss": 0.3491,
      "num_input_tokens_seen": 17968104,
      "step": 27740
    },
    {
      "epoch": 16.359080188679247,
      "grad_norm": 3.51892352104187,
      "learning_rate": 9.763128218049806e-07,
      "loss": 0.3304,
      "num_input_tokens_seen": 17970984,
      "step": 27745
    },
    {
      "epoch": 16.36202830188679,
      "grad_norm": 2.6502442359924316,
      "learning_rate": 9.747861118934005e-07,
      "loss": 0.3418,
      "num_input_tokens_seen": 17974728,
      "step": 27750
    },
    {
      "epoch": 16.36497641509434,
      "grad_norm": 2.33235239982605,
      "learning_rate": 9.732604676745443e-07,
      "loss": 0.2932,
      "num_input_tokens_seen": 17978312,
      "step": 27755
    },
    {
      "epoch": 16.367924528301888,
      "grad_norm": 2.5363848209381104,
      "learning_rate": 9.717358895523333e-07,
      "loss": 0.2882,
      "num_input_tokens_seen": 17981480,
      "step": 27760
    },
    {
      "epoch": 16.370872641509433,
      "grad_norm": 6.839925765991211,
      "learning_rate": 9.702123779304074e-07,
      "loss": 0.3896,
      "num_input_tokens_seen": 17985576,
      "step": 27765
    },
    {
      "epoch": 16.37382075471698,
      "grad_norm": 4.320279121398926,
      "learning_rate": 9.686899332121203e-07,
      "loss": 0.3588,
      "num_input_tokens_seen": 17989256,
      "step": 27770
    },
    {
      "epoch": 16.37676886792453,
      "grad_norm": 3.065141201019287,
      "learning_rate": 9.671685558005488e-07,
      "loss": 0.2589,
      "num_input_tokens_seen": 17993288,
      "step": 27775
    },
    {
      "epoch": 16.379716981132077,
      "grad_norm": 3.144134283065796,
      "learning_rate": 9.656482460984828e-07,
      "loss": 0.3404,
      "num_input_tokens_seen": 17996136,
      "step": 27780
    },
    {
      "epoch": 16.38266509433962,
      "grad_norm": 2.59511399269104,
      "learning_rate": 9.641290045084307e-07,
      "loss": 0.3231,
      "num_input_tokens_seen": 17999080,
      "step": 27785
    },
    {
      "epoch": 16.38561320754717,
      "grad_norm": 4.165119171142578,
      "learning_rate": 9.626108314326182e-07,
      "loss": 0.2682,
      "num_input_tokens_seen": 18001800,
      "step": 27790
    },
    {
      "epoch": 16.388561320754718,
      "grad_norm": 5.470855712890625,
      "learning_rate": 9.610937272729881e-07,
      "loss": 0.3382,
      "num_input_tokens_seen": 18005032,
      "step": 27795
    },
    {
      "epoch": 16.391509433962263,
      "grad_norm": 3.1975481510162354,
      "learning_rate": 9.595776924311996e-07,
      "loss": 0.374,
      "num_input_tokens_seen": 18009000,
      "step": 27800
    },
    {
      "epoch": 16.39445754716981,
      "grad_norm": 3.7333438396453857,
      "learning_rate": 9.580627273086313e-07,
      "loss": 0.3306,
      "num_input_tokens_seen": 18012232,
      "step": 27805
    },
    {
      "epoch": 16.39740566037736,
      "grad_norm": 3.779249906539917,
      "learning_rate": 9.565488323063754e-07,
      "loss": 0.3289,
      "num_input_tokens_seen": 18015208,
      "step": 27810
    },
    {
      "epoch": 16.400353773584907,
      "grad_norm": 4.654049396514893,
      "learning_rate": 9.55036007825243e-07,
      "loss": 0.3024,
      "num_input_tokens_seen": 18017320,
      "step": 27815
    },
    {
      "epoch": 16.403301886792452,
      "grad_norm": 3.0314512252807617,
      "learning_rate": 9.535242542657602e-07,
      "loss": 0.2834,
      "num_input_tokens_seen": 18020488,
      "step": 27820
    },
    {
      "epoch": 16.40625,
      "grad_norm": 3.105285167694092,
      "learning_rate": 9.520135720281692e-07,
      "loss": 0.2223,
      "num_input_tokens_seen": 18023496,
      "step": 27825
    },
    {
      "epoch": 16.409198113207548,
      "grad_norm": 6.459606170654297,
      "learning_rate": 9.505039615124318e-07,
      "loss": 0.2621,
      "num_input_tokens_seen": 18026696,
      "step": 27830
    },
    {
      "epoch": 16.412146226415093,
      "grad_norm": 4.999546527862549,
      "learning_rate": 9.489954231182235e-07,
      "loss": 0.2907,
      "num_input_tokens_seen": 18029032,
      "step": 27835
    },
    {
      "epoch": 16.41509433962264,
      "grad_norm": 3.0914456844329834,
      "learning_rate": 9.474879572449352e-07,
      "loss": 0.2769,
      "num_input_tokens_seen": 18032328,
      "step": 27840
    },
    {
      "epoch": 16.41804245283019,
      "grad_norm": 6.170389175415039,
      "learning_rate": 9.459815642916759e-07,
      "loss": 0.3651,
      "num_input_tokens_seen": 18035336,
      "step": 27845
    },
    {
      "epoch": 16.420990566037737,
      "grad_norm": 3.7705042362213135,
      "learning_rate": 9.444762446572692e-07,
      "loss": 0.2735,
      "num_input_tokens_seen": 18037800,
      "step": 27850
    },
    {
      "epoch": 16.423938679245282,
      "grad_norm": 6.436392307281494,
      "learning_rate": 9.429719987402541e-07,
      "loss": 0.3733,
      "num_input_tokens_seen": 18040904,
      "step": 27855
    },
    {
      "epoch": 16.42688679245283,
      "grad_norm": 3.346299409866333,
      "learning_rate": 9.414688269388883e-07,
      "loss": 0.3314,
      "num_input_tokens_seen": 18044232,
      "step": 27860
    },
    {
      "epoch": 16.42983490566038,
      "grad_norm": 2.8014674186706543,
      "learning_rate": 9.3996672965114e-07,
      "loss": 0.3169,
      "num_input_tokens_seen": 18048008,
      "step": 27865
    },
    {
      "epoch": 16.432783018867923,
      "grad_norm": 4.685744285583496,
      "learning_rate": 9.384657072747e-07,
      "loss": 0.4528,
      "num_input_tokens_seen": 18051528,
      "step": 27870
    },
    {
      "epoch": 16.43573113207547,
      "grad_norm": 3.583794593811035,
      "learning_rate": 9.369657602069676e-07,
      "loss": 0.2166,
      "num_input_tokens_seen": 18054728,
      "step": 27875
    },
    {
      "epoch": 16.43867924528302,
      "grad_norm": 4.398697853088379,
      "learning_rate": 9.354668888450608e-07,
      "loss": 0.4301,
      "num_input_tokens_seen": 18058248,
      "step": 27880
    },
    {
      "epoch": 16.441627358490567,
      "grad_norm": 3.6472997665405273,
      "learning_rate": 9.339690935858125e-07,
      "loss": 0.5031,
      "num_input_tokens_seen": 18062792,
      "step": 27885
    },
    {
      "epoch": 16.444575471698112,
      "grad_norm": 4.085649490356445,
      "learning_rate": 9.324723748257697e-07,
      "loss": 0.2471,
      "num_input_tokens_seen": 18065768,
      "step": 27890
    },
    {
      "epoch": 16.44752358490566,
      "grad_norm": 2.411473274230957,
      "learning_rate": 9.309767329611963e-07,
      "loss": 0.3976,
      "num_input_tokens_seen": 18069224,
      "step": 27895
    },
    {
      "epoch": 16.45047169811321,
      "grad_norm": 5.963064670562744,
      "learning_rate": 9.294821683880695e-07,
      "loss": 0.3281,
      "num_input_tokens_seen": 18072328,
      "step": 27900
    },
    {
      "epoch": 16.453419811320753,
      "grad_norm": 4.252490997314453,
      "learning_rate": 9.279886815020816e-07,
      "loss": 0.3406,
      "num_input_tokens_seen": 18075048,
      "step": 27905
    },
    {
      "epoch": 16.4563679245283,
      "grad_norm": 4.514782905578613,
      "learning_rate": 9.264962726986393e-07,
      "loss": 0.3412,
      "num_input_tokens_seen": 18078088,
      "step": 27910
    },
    {
      "epoch": 16.45931603773585,
      "grad_norm": 4.593124866485596,
      "learning_rate": 9.250049423728652e-07,
      "loss": 0.329,
      "num_input_tokens_seen": 18080776,
      "step": 27915
    },
    {
      "epoch": 16.462264150943398,
      "grad_norm": 2.670788288116455,
      "learning_rate": 9.235146909195936e-07,
      "loss": 0.3702,
      "num_input_tokens_seen": 18083944,
      "step": 27920
    },
    {
      "epoch": 16.465212264150942,
      "grad_norm": 3.8087964057922363,
      "learning_rate": 9.220255187333771e-07,
      "loss": 0.2482,
      "num_input_tokens_seen": 18086632,
      "step": 27925
    },
    {
      "epoch": 16.46816037735849,
      "grad_norm": 4.341442584991455,
      "learning_rate": 9.205374262084798e-07,
      "loss": 0.3171,
      "num_input_tokens_seen": 18089384,
      "step": 27930
    },
    {
      "epoch": 16.47110849056604,
      "grad_norm": 3.5641536712646484,
      "learning_rate": 9.190504137388806e-07,
      "loss": 0.345,
      "num_input_tokens_seen": 18092360,
      "step": 27935
    },
    {
      "epoch": 16.474056603773583,
      "grad_norm": 5.259038925170898,
      "learning_rate": 9.175644817182722e-07,
      "loss": 0.2034,
      "num_input_tokens_seen": 18095816,
      "step": 27940
    },
    {
      "epoch": 16.47700471698113,
      "grad_norm": 6.674475193023682,
      "learning_rate": 9.16079630540061e-07,
      "loss": 0.2142,
      "num_input_tokens_seen": 18098184,
      "step": 27945
    },
    {
      "epoch": 16.47995283018868,
      "grad_norm": 2.4105618000030518,
      "learning_rate": 9.145958605973676e-07,
      "loss": 0.302,
      "num_input_tokens_seen": 18101288,
      "step": 27950
    },
    {
      "epoch": 16.482900943396228,
      "grad_norm": 8.32170295715332,
      "learning_rate": 9.131131722830289e-07,
      "loss": 0.2544,
      "num_input_tokens_seen": 18103944,
      "step": 27955
    },
    {
      "epoch": 16.485849056603772,
      "grad_norm": 4.594747543334961,
      "learning_rate": 9.116315659895892e-07,
      "loss": 0.1822,
      "num_input_tokens_seen": 18106440,
      "step": 27960
    },
    {
      "epoch": 16.48879716981132,
      "grad_norm": 5.134743690490723,
      "learning_rate": 9.10151042109314e-07,
      "loss": 0.3226,
      "num_input_tokens_seen": 18109704,
      "step": 27965
    },
    {
      "epoch": 16.49174528301887,
      "grad_norm": 4.9223737716674805,
      "learning_rate": 9.086716010341767e-07,
      "loss": 0.2697,
      "num_input_tokens_seen": 18112232,
      "step": 27970
    },
    {
      "epoch": 16.494693396226417,
      "grad_norm": 2.746464729309082,
      "learning_rate": 9.071932431558655e-07,
      "loss": 0.3069,
      "num_input_tokens_seen": 18116264,
      "step": 27975
    },
    {
      "epoch": 16.49764150943396,
      "grad_norm": 5.26420259475708,
      "learning_rate": 9.057159688657824e-07,
      "loss": 0.2977,
      "num_input_tokens_seen": 18118280,
      "step": 27980
    },
    {
      "epoch": 16.50058962264151,
      "grad_norm": 7.357564926147461,
      "learning_rate": 9.042397785550405e-07,
      "loss": 0.2908,
      "num_input_tokens_seen": 18120584,
      "step": 27985
    },
    {
      "epoch": 16.503537735849058,
      "grad_norm": 3.8258211612701416,
      "learning_rate": 9.027646726144707e-07,
      "loss": 0.3215,
      "num_input_tokens_seen": 18123080,
      "step": 27990
    },
    {
      "epoch": 16.506485849056602,
      "grad_norm": 3.3608338832855225,
      "learning_rate": 9.012906514346115e-07,
      "loss": 0.4349,
      "num_input_tokens_seen": 18125960,
      "step": 27995
    },
    {
      "epoch": 16.50943396226415,
      "grad_norm": 2.7886874675750732,
      "learning_rate": 8.99817715405717e-07,
      "loss": 0.2992,
      "num_input_tokens_seen": 18130344,
      "step": 28000
    },
    {
      "epoch": 16.5123820754717,
      "grad_norm": 3.8767459392547607,
      "learning_rate": 8.983458649177529e-07,
      "loss": 0.3111,
      "num_input_tokens_seen": 18133448,
      "step": 28005
    },
    {
      "epoch": 16.515330188679247,
      "grad_norm": 3.685734272003174,
      "learning_rate": 8.968751003603982e-07,
      "loss": 0.2629,
      "num_input_tokens_seen": 18136328,
      "step": 28010
    },
    {
      "epoch": 16.51827830188679,
      "grad_norm": 2.1930036544799805,
      "learning_rate": 8.95405422123043e-07,
      "loss": 0.3461,
      "num_input_tokens_seen": 18139656,
      "step": 28015
    },
    {
      "epoch": 16.52122641509434,
      "grad_norm": 7.540945529937744,
      "learning_rate": 8.939368305947932e-07,
      "loss": 0.3429,
      "num_input_tokens_seen": 18142280,
      "step": 28020
    },
    {
      "epoch": 16.524174528301888,
      "grad_norm": 4.95533561706543,
      "learning_rate": 8.92469326164464e-07,
      "loss": 0.2498,
      "num_input_tokens_seen": 18144808,
      "step": 28025
    },
    {
      "epoch": 16.527122641509433,
      "grad_norm": 2.9738283157348633,
      "learning_rate": 8.910029092205829e-07,
      "loss": 0.2647,
      "num_input_tokens_seen": 18148232,
      "step": 28030
    },
    {
      "epoch": 16.53007075471698,
      "grad_norm": 3.4636752605438232,
      "learning_rate": 8.895375801513906e-07,
      "loss": 0.3857,
      "num_input_tokens_seen": 18152328,
      "step": 28035
    },
    {
      "epoch": 16.53301886792453,
      "grad_norm": 5.3271989822387695,
      "learning_rate": 8.880733393448377e-07,
      "loss": 0.4298,
      "num_input_tokens_seen": 18156200,
      "step": 28040
    },
    {
      "epoch": 16.535966981132077,
      "grad_norm": 3.843120813369751,
      "learning_rate": 8.866101871885907e-07,
      "loss": 0.3557,
      "num_input_tokens_seen": 18159016,
      "step": 28045
    },
    {
      "epoch": 16.53891509433962,
      "grad_norm": 6.1894121170043945,
      "learning_rate": 8.851481240700249e-07,
      "loss": 0.3646,
      "num_input_tokens_seen": 18162120,
      "step": 28050
    },
    {
      "epoch": 16.54186320754717,
      "grad_norm": 4.563694477081299,
      "learning_rate": 8.836871503762257e-07,
      "loss": 0.3093,
      "num_input_tokens_seen": 18165192,
      "step": 28055
    },
    {
      "epoch": 16.544811320754718,
      "grad_norm": 4.294272422790527,
      "learning_rate": 8.822272664939946e-07,
      "loss": 0.2564,
      "num_input_tokens_seen": 18167848,
      "step": 28060
    },
    {
      "epoch": 16.547759433962263,
      "grad_norm": 4.799142360687256,
      "learning_rate": 8.80768472809842e-07,
      "loss": 0.3213,
      "num_input_tokens_seen": 18171944,
      "step": 28065
    },
    {
      "epoch": 16.55070754716981,
      "grad_norm": 3.3093197345733643,
      "learning_rate": 8.793107697099884e-07,
      "loss": 0.3351,
      "num_input_tokens_seen": 18175048,
      "step": 28070
    },
    {
      "epoch": 16.55365566037736,
      "grad_norm": 3.5079233646392822,
      "learning_rate": 8.778541575803673e-07,
      "loss": 0.2777,
      "num_input_tokens_seen": 18179048,
      "step": 28075
    },
    {
      "epoch": 16.556603773584907,
      "grad_norm": 4.919696807861328,
      "learning_rate": 8.763986368066241e-07,
      "loss": 0.3086,
      "num_input_tokens_seen": 18183336,
      "step": 28080
    },
    {
      "epoch": 16.559551886792452,
      "grad_norm": 5.199502468109131,
      "learning_rate": 8.749442077741138e-07,
      "loss": 0.3902,
      "num_input_tokens_seen": 18186664,
      "step": 28085
    },
    {
      "epoch": 16.5625,
      "grad_norm": 3.955897808074951,
      "learning_rate": 8.734908708679024e-07,
      "loss": 0.286,
      "num_input_tokens_seen": 18190024,
      "step": 28090
    },
    {
      "epoch": 16.565448113207548,
      "grad_norm": 6.916514873504639,
      "learning_rate": 8.72038626472767e-07,
      "loss": 0.3538,
      "num_input_tokens_seen": 18193032,
      "step": 28095
    },
    {
      "epoch": 16.568396226415093,
      "grad_norm": 3.682419538497925,
      "learning_rate": 8.705874749731962e-07,
      "loss": 0.2955,
      "num_input_tokens_seen": 18197128,
      "step": 28100
    },
    {
      "epoch": 16.57134433962264,
      "grad_norm": 3.0711731910705566,
      "learning_rate": 8.691374167533867e-07,
      "loss": 0.3176,
      "num_input_tokens_seen": 18200424,
      "step": 28105
    },
    {
      "epoch": 16.57429245283019,
      "grad_norm": 4.213841915130615,
      "learning_rate": 8.6768845219725e-07,
      "loss": 0.2681,
      "num_input_tokens_seen": 18203496,
      "step": 28110
    },
    {
      "epoch": 16.577240566037737,
      "grad_norm": 4.577976226806641,
      "learning_rate": 8.662405816884056e-07,
      "loss": 0.2631,
      "num_input_tokens_seen": 18205704,
      "step": 28115
    },
    {
      "epoch": 16.580188679245282,
      "grad_norm": 4.047306537628174,
      "learning_rate": 8.647938056101824e-07,
      "loss": 0.2596,
      "num_input_tokens_seen": 18208744,
      "step": 28120
    },
    {
      "epoch": 16.58313679245283,
      "grad_norm": 2.690678358078003,
      "learning_rate": 8.63348124345621e-07,
      "loss": 0.3054,
      "num_input_tokens_seen": 18211464,
      "step": 28125
    },
    {
      "epoch": 16.58608490566038,
      "grad_norm": 5.285005569458008,
      "learning_rate": 8.619035382774716e-07,
      "loss": 0.3314,
      "num_input_tokens_seen": 18214664,
      "step": 28130
    },
    {
      "epoch": 16.589033018867923,
      "grad_norm": 1.0728672742843628,
      "learning_rate": 8.60460047788193e-07,
      "loss": 0.2389,
      "num_input_tokens_seen": 18219368,
      "step": 28135
    },
    {
      "epoch": 16.59198113207547,
      "grad_norm": 3.4245338439941406,
      "learning_rate": 8.590176532599587e-07,
      "loss": 0.3081,
      "num_input_tokens_seen": 18222024,
      "step": 28140
    },
    {
      "epoch": 16.59492924528302,
      "grad_norm": 4.616681098937988,
      "learning_rate": 8.575763550746475e-07,
      "loss": 0.2813,
      "num_input_tokens_seen": 18224968,
      "step": 28145
    },
    {
      "epoch": 16.597877358490567,
      "grad_norm": 2.9476661682128906,
      "learning_rate": 8.56136153613848e-07,
      "loss": 0.2459,
      "num_input_tokens_seen": 18227720,
      "step": 28150
    },
    {
      "epoch": 16.600825471698112,
      "grad_norm": 4.411141395568848,
      "learning_rate": 8.546970492588619e-07,
      "loss": 0.2653,
      "num_input_tokens_seen": 18230280,
      "step": 28155
    },
    {
      "epoch": 16.60377358490566,
      "grad_norm": 4.170223236083984,
      "learning_rate": 8.532590423906973e-07,
      "loss": 0.2843,
      "num_input_tokens_seen": 18233544,
      "step": 28160
    },
    {
      "epoch": 16.60672169811321,
      "grad_norm": 3.0428144931793213,
      "learning_rate": 8.518221333900728e-07,
      "loss": 0.4525,
      "num_input_tokens_seen": 18241288,
      "step": 28165
    },
    {
      "epoch": 16.609669811320753,
      "grad_norm": 13.12922191619873,
      "learning_rate": 8.503863226374148e-07,
      "loss": 0.3913,
      "num_input_tokens_seen": 18244648,
      "step": 28170
    },
    {
      "epoch": 16.6126179245283,
      "grad_norm": 3.9127557277679443,
      "learning_rate": 8.489516105128632e-07,
      "loss": 0.2404,
      "num_input_tokens_seen": 18248072,
      "step": 28175
    },
    {
      "epoch": 16.61556603773585,
      "grad_norm": 3.7335431575775146,
      "learning_rate": 8.475179973962621e-07,
      "loss": 0.3262,
      "num_input_tokens_seen": 18250504,
      "step": 28180
    },
    {
      "epoch": 16.618514150943398,
      "grad_norm": 5.186847686767578,
      "learning_rate": 8.460854836671678e-07,
      "loss": 0.3477,
      "num_input_tokens_seen": 18253768,
      "step": 28185
    },
    {
      "epoch": 16.621462264150942,
      "grad_norm": 3.8033459186553955,
      "learning_rate": 8.446540697048445e-07,
      "loss": 0.4016,
      "num_input_tokens_seen": 18257832,
      "step": 28190
    },
    {
      "epoch": 16.62441037735849,
      "grad_norm": 3.554198980331421,
      "learning_rate": 8.432237558882639e-07,
      "loss": 0.3698,
      "num_input_tokens_seen": 18261608,
      "step": 28195
    },
    {
      "epoch": 16.62735849056604,
      "grad_norm": 4.198512077331543,
      "learning_rate": 8.417945425961083e-07,
      "loss": 0.248,
      "num_input_tokens_seen": 18264584,
      "step": 28200
    },
    {
      "epoch": 16.630306603773583,
      "grad_norm": 2.000370740890503,
      "learning_rate": 8.403664302067688e-07,
      "loss": 0.2652,
      "num_input_tokens_seen": 18267944,
      "step": 28205
    },
    {
      "epoch": 16.63325471698113,
      "grad_norm": 2.8422558307647705,
      "learning_rate": 8.389394190983446e-07,
      "loss": 0.2376,
      "num_input_tokens_seen": 18270600,
      "step": 28210
    },
    {
      "epoch": 16.63620283018868,
      "grad_norm": 5.883741855621338,
      "learning_rate": 8.37513509648642e-07,
      "loss": 0.3233,
      "num_input_tokens_seen": 18274024,
      "step": 28215
    },
    {
      "epoch": 16.639150943396228,
      "grad_norm": 4.237092018127441,
      "learning_rate": 8.360887022351771e-07,
      "loss": 0.2772,
      "num_input_tokens_seen": 18277544,
      "step": 28220
    },
    {
      "epoch": 16.642099056603772,
      "grad_norm": 5.635860443115234,
      "learning_rate": 8.346649972351739e-07,
      "loss": 0.3886,
      "num_input_tokens_seen": 18280360,
      "step": 28225
    },
    {
      "epoch": 16.64504716981132,
      "grad_norm": 5.698032379150391,
      "learning_rate": 8.33242395025563e-07,
      "loss": 0.3511,
      "num_input_tokens_seen": 18283144,
      "step": 28230
    },
    {
      "epoch": 16.64799528301887,
      "grad_norm": 7.139906883239746,
      "learning_rate": 8.318208959829871e-07,
      "loss": 0.3356,
      "num_input_tokens_seen": 18285992,
      "step": 28235
    },
    {
      "epoch": 16.650943396226417,
      "grad_norm": 3.0570781230926514,
      "learning_rate": 8.304005004837929e-07,
      "loss": 0.3277,
      "num_input_tokens_seen": 18289480,
      "step": 28240
    },
    {
      "epoch": 16.65389150943396,
      "grad_norm": 1.6728651523590088,
      "learning_rate": 8.289812089040344e-07,
      "loss": 0.2901,
      "num_input_tokens_seen": 18292200,
      "step": 28245
    },
    {
      "epoch": 16.65683962264151,
      "grad_norm": 4.482764720916748,
      "learning_rate": 8.275630216194785e-07,
      "loss": 0.3468,
      "num_input_tokens_seen": 18295656,
      "step": 28250
    },
    {
      "epoch": 16.659787735849058,
      "grad_norm": 4.2783331871032715,
      "learning_rate": 8.261459390055948e-07,
      "loss": 0.3807,
      "num_input_tokens_seen": 18298440,
      "step": 28255
    },
    {
      "epoch": 16.662735849056602,
      "grad_norm": 3.394105911254883,
      "learning_rate": 8.24729961437562e-07,
      "loss": 0.3249,
      "num_input_tokens_seen": 18303048,
      "step": 28260
    },
    {
      "epoch": 16.66568396226415,
      "grad_norm": 4.446547508239746,
      "learning_rate": 8.233150892902653e-07,
      "loss": 0.3601,
      "num_input_tokens_seen": 18306216,
      "step": 28265
    },
    {
      "epoch": 16.6686320754717,
      "grad_norm": 3.2332546710968018,
      "learning_rate": 8.219013229383005e-07,
      "loss": 0.3401,
      "num_input_tokens_seen": 18308936,
      "step": 28270
    },
    {
      "epoch": 16.671580188679247,
      "grad_norm": 2.705014944076538,
      "learning_rate": 8.204886627559666e-07,
      "loss": 0.3402,
      "num_input_tokens_seen": 18312680,
      "step": 28275
    },
    {
      "epoch": 16.67452830188679,
      "grad_norm": 5.425186634063721,
      "learning_rate": 8.190771091172722e-07,
      "loss": 0.4033,
      "num_input_tokens_seen": 18315592,
      "step": 28280
    },
    {
      "epoch": 16.67747641509434,
      "grad_norm": 3.253239870071411,
      "learning_rate": 8.176666623959323e-07,
      "loss": 0.2224,
      "num_input_tokens_seen": 18319112,
      "step": 28285
    },
    {
      "epoch": 16.680424528301888,
      "grad_norm": 3.420159339904785,
      "learning_rate": 8.162573229653681e-07,
      "loss": 0.3844,
      "num_input_tokens_seen": 18322280,
      "step": 28290
    },
    {
      "epoch": 16.683372641509433,
      "grad_norm": 2.353745698928833,
      "learning_rate": 8.148490911987073e-07,
      "loss": 0.2203,
      "num_input_tokens_seen": 18326632,
      "step": 28295
    },
    {
      "epoch": 16.68632075471698,
      "grad_norm": 3.944317102432251,
      "learning_rate": 8.134419674687876e-07,
      "loss": 0.2986,
      "num_input_tokens_seen": 18330792,
      "step": 28300
    },
    {
      "epoch": 16.68926886792453,
      "grad_norm": 4.503780841827393,
      "learning_rate": 8.120359521481502e-07,
      "loss": 0.3737,
      "num_input_tokens_seen": 18333352,
      "step": 28305
    },
    {
      "epoch": 16.692216981132077,
      "grad_norm": 5.853037357330322,
      "learning_rate": 8.106310456090438e-07,
      "loss": 0.2668,
      "num_input_tokens_seen": 18336776,
      "step": 28310
    },
    {
      "epoch": 16.69516509433962,
      "grad_norm": 3.7868716716766357,
      "learning_rate": 8.092272482234231e-07,
      "loss": 0.3253,
      "num_input_tokens_seen": 18340232,
      "step": 28315
    },
    {
      "epoch": 16.69811320754717,
      "grad_norm": 5.352800369262695,
      "learning_rate": 8.078245603629486e-07,
      "loss": 0.3274,
      "num_input_tokens_seen": 18343336,
      "step": 28320
    },
    {
      "epoch": 16.701061320754718,
      "grad_norm": 2.9771671295166016,
      "learning_rate": 8.0642298239899e-07,
      "loss": 0.3032,
      "num_input_tokens_seen": 18346216,
      "step": 28325
    },
    {
      "epoch": 16.704009433962263,
      "grad_norm": 1.9129494428634644,
      "learning_rate": 8.050225147026202e-07,
      "loss": 0.421,
      "num_input_tokens_seen": 18350056,
      "step": 28330
    },
    {
      "epoch": 16.70695754716981,
      "grad_norm": 4.681399822235107,
      "learning_rate": 8.03623157644619e-07,
      "loss": 0.3582,
      "num_input_tokens_seen": 18353864,
      "step": 28335
    },
    {
      "epoch": 16.70990566037736,
      "grad_norm": 3.7557342052459717,
      "learning_rate": 8.022249115954728e-07,
      "loss": 0.2757,
      "num_input_tokens_seen": 18356456,
      "step": 28340
    },
    {
      "epoch": 16.712853773584907,
      "grad_norm": 3.4892611503601074,
      "learning_rate": 8.008277769253709e-07,
      "loss": 0.2539,
      "num_input_tokens_seen": 18360168,
      "step": 28345
    },
    {
      "epoch": 16.715801886792452,
      "grad_norm": 5.308775424957275,
      "learning_rate": 7.994317540042135e-07,
      "loss": 0.2924,
      "num_input_tokens_seen": 18363304,
      "step": 28350
    },
    {
      "epoch": 16.71875,
      "grad_norm": 4.03567361831665,
      "learning_rate": 7.980368432016017e-07,
      "loss": 0.4348,
      "num_input_tokens_seen": 18367176,
      "step": 28355
    },
    {
      "epoch": 16.721698113207548,
      "grad_norm": 4.8935747146606445,
      "learning_rate": 7.966430448868461e-07,
      "loss": 0.2605,
      "num_input_tokens_seen": 18370024,
      "step": 28360
    },
    {
      "epoch": 16.724646226415093,
      "grad_norm": 5.608789920806885,
      "learning_rate": 7.952503594289601e-07,
      "loss": 0.3925,
      "num_input_tokens_seen": 18372616,
      "step": 28365
    },
    {
      "epoch": 16.72759433962264,
      "grad_norm": 5.694059371948242,
      "learning_rate": 7.93858787196663e-07,
      "loss": 0.728,
      "num_input_tokens_seen": 18375592,
      "step": 28370
    },
    {
      "epoch": 16.73054245283019,
      "grad_norm": 3.8707115650177,
      "learning_rate": 7.92468328558379e-07,
      "loss": 0.3333,
      "num_input_tokens_seen": 18378760,
      "step": 28375
    },
    {
      "epoch": 16.733490566037737,
      "grad_norm": 4.706076145172119,
      "learning_rate": 7.910789838822386e-07,
      "loss": 0.297,
      "num_input_tokens_seen": 18382152,
      "step": 28380
    },
    {
      "epoch": 16.736438679245282,
      "grad_norm": 3.7958977222442627,
      "learning_rate": 7.89690753536076e-07,
      "loss": 0.2498,
      "num_input_tokens_seen": 18384776,
      "step": 28385
    },
    {
      "epoch": 16.73938679245283,
      "grad_norm": 2.99353289604187,
      "learning_rate": 7.883036378874326e-07,
      "loss": 0.2612,
      "num_input_tokens_seen": 18388488,
      "step": 28390
    },
    {
      "epoch": 16.74233490566038,
      "grad_norm": 5.522387981414795,
      "learning_rate": 7.86917637303552e-07,
      "loss": 0.2441,
      "num_input_tokens_seen": 18395336,
      "step": 28395
    },
    {
      "epoch": 16.745283018867923,
      "grad_norm": 2.304823398590088,
      "learning_rate": 7.855327521513851e-07,
      "loss": 0.2976,
      "num_input_tokens_seen": 18399144,
      "step": 28400
    },
    {
      "epoch": 16.74823113207547,
      "grad_norm": 3.3296022415161133,
      "learning_rate": 7.841489827975851e-07,
      "loss": 0.2361,
      "num_input_tokens_seen": 18402024,
      "step": 28405
    },
    {
      "epoch": 16.75117924528302,
      "grad_norm": 3.828923463821411,
      "learning_rate": 7.827663296085109e-07,
      "loss": 0.2886,
      "num_input_tokens_seen": 18405160,
      "step": 28410
    },
    {
      "epoch": 16.754127358490567,
      "grad_norm": 7.052605152130127,
      "learning_rate": 7.813847929502255e-07,
      "loss": 0.2886,
      "num_input_tokens_seen": 18408456,
      "step": 28415
    },
    {
      "epoch": 16.757075471698112,
      "grad_norm": 3.6108710765838623,
      "learning_rate": 7.800043731884982e-07,
      "loss": 0.3251,
      "num_input_tokens_seen": 18411432,
      "step": 28420
    },
    {
      "epoch": 16.76002358490566,
      "grad_norm": 4.544459342956543,
      "learning_rate": 7.786250706888005e-07,
      "loss": 0.264,
      "num_input_tokens_seen": 18414440,
      "step": 28425
    },
    {
      "epoch": 16.76297169811321,
      "grad_norm": 4.231398582458496,
      "learning_rate": 7.772468858163085e-07,
      "loss": 0.2803,
      "num_input_tokens_seen": 18417256,
      "step": 28430
    },
    {
      "epoch": 16.765919811320753,
      "grad_norm": 3.9471640586853027,
      "learning_rate": 7.758698189359026e-07,
      "loss": 0.3616,
      "num_input_tokens_seen": 18423336,
      "step": 28435
    },
    {
      "epoch": 16.7688679245283,
      "grad_norm": 3.7441658973693848,
      "learning_rate": 7.744938704121658e-07,
      "loss": 0.2704,
      "num_input_tokens_seen": 18425960,
      "step": 28440
    },
    {
      "epoch": 16.77181603773585,
      "grad_norm": 4.690322399139404,
      "learning_rate": 7.731190406093892e-07,
      "loss": 0.3218,
      "num_input_tokens_seen": 18428808,
      "step": 28445
    },
    {
      "epoch": 16.774764150943398,
      "grad_norm": 2.7229220867156982,
      "learning_rate": 7.717453298915617e-07,
      "loss": 0.3046,
      "num_input_tokens_seen": 18432040,
      "step": 28450
    },
    {
      "epoch": 16.777712264150942,
      "grad_norm": 3.809911012649536,
      "learning_rate": 7.703727386223825e-07,
      "loss": 0.2658,
      "num_input_tokens_seen": 18434856,
      "step": 28455
    },
    {
      "epoch": 16.78066037735849,
      "grad_norm": 2.9997589588165283,
      "learning_rate": 7.690012671652491e-07,
      "loss": 0.3318,
      "num_input_tokens_seen": 18438280,
      "step": 28460
    },
    {
      "epoch": 16.78360849056604,
      "grad_norm": 4.3696608543396,
      "learning_rate": 7.676309158832651e-07,
      "loss": 0.2817,
      "num_input_tokens_seen": 18442056,
      "step": 28465
    },
    {
      "epoch": 16.786556603773583,
      "grad_norm": 4.953383922576904,
      "learning_rate": 7.662616851392362e-07,
      "loss": 0.3029,
      "num_input_tokens_seen": 18445416,
      "step": 28470
    },
    {
      "epoch": 16.78950471698113,
      "grad_norm": 2.774354934692383,
      "learning_rate": 7.648935752956732e-07,
      "loss": 0.2852,
      "num_input_tokens_seen": 18448584,
      "step": 28475
    },
    {
      "epoch": 16.79245283018868,
      "grad_norm": 5.469350814819336,
      "learning_rate": 7.635265867147867e-07,
      "loss": 0.2405,
      "num_input_tokens_seen": 18452008,
      "step": 28480
    },
    {
      "epoch": 16.795400943396228,
      "grad_norm": 5.0724334716796875,
      "learning_rate": 7.621607197584963e-07,
      "loss": 0.2569,
      "num_input_tokens_seen": 18455016,
      "step": 28485
    },
    {
      "epoch": 16.798349056603772,
      "grad_norm": 6.820105075836182,
      "learning_rate": 7.607959747884186e-07,
      "loss": 0.2778,
      "num_input_tokens_seen": 18458792,
      "step": 28490
    },
    {
      "epoch": 16.80129716981132,
      "grad_norm": 3.7191953659057617,
      "learning_rate": 7.594323521658769e-07,
      "loss": 0.3033,
      "num_input_tokens_seen": 18462152,
      "step": 28495
    },
    {
      "epoch": 16.80424528301887,
      "grad_norm": 2.541574478149414,
      "learning_rate": 7.580698522518958e-07,
      "loss": 0.3054,
      "num_input_tokens_seen": 18465640,
      "step": 28500
    },
    {
      "epoch": 16.807193396226417,
      "grad_norm": 3.278319835662842,
      "learning_rate": 7.567084754072035e-07,
      "loss": 0.2979,
      "num_input_tokens_seen": 18468776,
      "step": 28505
    },
    {
      "epoch": 16.81014150943396,
      "grad_norm": 4.099082946777344,
      "learning_rate": 7.553482219922282e-07,
      "loss": 0.2749,
      "num_input_tokens_seen": 18471304,
      "step": 28510
    },
    {
      "epoch": 16.81308962264151,
      "grad_norm": 5.047877788543701,
      "learning_rate": 7.539890923671061e-07,
      "loss": 0.245,
      "num_input_tokens_seen": 18474344,
      "step": 28515
    },
    {
      "epoch": 16.816037735849058,
      "grad_norm": 2.2043864727020264,
      "learning_rate": 7.526310868916708e-07,
      "loss": 0.3346,
      "num_input_tokens_seen": 18477448,
      "step": 28520
    },
    {
      "epoch": 16.818985849056602,
      "grad_norm": 5.085360527038574,
      "learning_rate": 7.512742059254602e-07,
      "loss": 0.2977,
      "num_input_tokens_seen": 18480712,
      "step": 28525
    },
    {
      "epoch": 16.82193396226415,
      "grad_norm": 6.135254859924316,
      "learning_rate": 7.499184498277151e-07,
      "loss": 0.2522,
      "num_input_tokens_seen": 18484456,
      "step": 28530
    },
    {
      "epoch": 16.8248820754717,
      "grad_norm": 5.805110931396484,
      "learning_rate": 7.485638189573758e-07,
      "loss": 0.3249,
      "num_input_tokens_seen": 18487432,
      "step": 28535
    },
    {
      "epoch": 16.827830188679247,
      "grad_norm": 4.130301475524902,
      "learning_rate": 7.472103136730891e-07,
      "loss": 0.2732,
      "num_input_tokens_seen": 18491400,
      "step": 28540
    },
    {
      "epoch": 16.83077830188679,
      "grad_norm": 3.147540807723999,
      "learning_rate": 7.458579343331996e-07,
      "loss": 0.2974,
      "num_input_tokens_seen": 18494152,
      "step": 28545
    },
    {
      "epoch": 16.83372641509434,
      "grad_norm": 3.2875897884368896,
      "learning_rate": 7.445066812957569e-07,
      "loss": 0.216,
      "num_input_tokens_seen": 18497832,
      "step": 28550
    },
    {
      "epoch": 16.836674528301888,
      "grad_norm": 4.734256744384766,
      "learning_rate": 7.43156554918511e-07,
      "loss": 0.3001,
      "num_input_tokens_seen": 18502376,
      "step": 28555
    },
    {
      "epoch": 16.839622641509433,
      "grad_norm": 5.22360372543335,
      "learning_rate": 7.418075555589132e-07,
      "loss": 0.234,
      "num_input_tokens_seen": 18505896,
      "step": 28560
    },
    {
      "epoch": 16.84257075471698,
      "grad_norm": 3.3590316772460938,
      "learning_rate": 7.404596835741168e-07,
      "loss": 0.2004,
      "num_input_tokens_seen": 18508360,
      "step": 28565
    },
    {
      "epoch": 16.84551886792453,
      "grad_norm": 2.3617475032806396,
      "learning_rate": 7.391129393209751e-07,
      "loss": 0.2454,
      "num_input_tokens_seen": 18511624,
      "step": 28570
    },
    {
      "epoch": 16.848466981132077,
      "grad_norm": 3.9272730350494385,
      "learning_rate": 7.377673231560478e-07,
      "loss": 0.2631,
      "num_input_tokens_seen": 18515176,
      "step": 28575
    },
    {
      "epoch": 16.85141509433962,
      "grad_norm": 4.8118391036987305,
      "learning_rate": 7.364228354355907e-07,
      "loss": 0.3489,
      "num_input_tokens_seen": 18519240,
      "step": 28580
    },
    {
      "epoch": 16.85436320754717,
      "grad_norm": 5.953233242034912,
      "learning_rate": 7.350794765155627e-07,
      "loss": 0.3352,
      "num_input_tokens_seen": 18522600,
      "step": 28585
    },
    {
      "epoch": 16.857311320754718,
      "grad_norm": 3.3850016593933105,
      "learning_rate": 7.337372467516246e-07,
      "loss": 0.3833,
      "num_input_tokens_seen": 18526248,
      "step": 28590
    },
    {
      "epoch": 16.860259433962263,
      "grad_norm": 2.2690467834472656,
      "learning_rate": 7.323961464991369e-07,
      "loss": 0.2752,
      "num_input_tokens_seen": 18531080,
      "step": 28595
    },
    {
      "epoch": 16.86320754716981,
      "grad_norm": 5.738551139831543,
      "learning_rate": 7.310561761131601e-07,
      "loss": 0.2894,
      "num_input_tokens_seen": 18534664,
      "step": 28600
    },
    {
      "epoch": 16.86615566037736,
      "grad_norm": 4.421821117401123,
      "learning_rate": 7.297173359484605e-07,
      "loss": 0.3374,
      "num_input_tokens_seen": 18537864,
      "step": 28605
    },
    {
      "epoch": 16.869103773584907,
      "grad_norm": 4.334507942199707,
      "learning_rate": 7.283796263595e-07,
      "loss": 0.325,
      "num_input_tokens_seen": 18540936,
      "step": 28610
    },
    {
      "epoch": 16.872051886792452,
      "grad_norm": 3.06626033782959,
      "learning_rate": 7.270430477004431e-07,
      "loss": 0.3161,
      "num_input_tokens_seen": 18543496,
      "step": 28615
    },
    {
      "epoch": 16.875,
      "grad_norm": 3.1577205657958984,
      "learning_rate": 7.257076003251545e-07,
      "loss": 0.3355,
      "num_input_tokens_seen": 18546920,
      "step": 28620
    },
    {
      "epoch": 16.877948113207548,
      "grad_norm": 4.027953624725342,
      "learning_rate": 7.243732845871998e-07,
      "loss": 0.3211,
      "num_input_tokens_seen": 18550152,
      "step": 28625
    },
    {
      "epoch": 16.880896226415093,
      "grad_norm": 5.010776996612549,
      "learning_rate": 7.230401008398441e-07,
      "loss": 0.3857,
      "num_input_tokens_seen": 18553448,
      "step": 28630
    },
    {
      "epoch": 16.88384433962264,
      "grad_norm": 2.2837448120117188,
      "learning_rate": 7.217080494360546e-07,
      "loss": 0.3051,
      "num_input_tokens_seen": 18556808,
      "step": 28635
    },
    {
      "epoch": 16.88679245283019,
      "grad_norm": 4.648995399475098,
      "learning_rate": 7.20377130728498e-07,
      "loss": 0.2092,
      "num_input_tokens_seen": 18559368,
      "step": 28640
    },
    {
      "epoch": 16.889740566037737,
      "grad_norm": 5.485760688781738,
      "learning_rate": 7.190473450695407e-07,
      "loss": 0.3685,
      "num_input_tokens_seen": 18562056,
      "step": 28645
    },
    {
      "epoch": 16.892688679245282,
      "grad_norm": 4.545187950134277,
      "learning_rate": 7.177186928112484e-07,
      "loss": 0.3621,
      "num_input_tokens_seen": 18565736,
      "step": 28650
    },
    {
      "epoch": 16.89563679245283,
      "grad_norm": 7.543068885803223,
      "learning_rate": 7.163911743053876e-07,
      "loss": 0.3144,
      "num_input_tokens_seen": 18568456,
      "step": 28655
    },
    {
      "epoch": 16.89858490566038,
      "grad_norm": 3.8706986904144287,
      "learning_rate": 7.150647899034252e-07,
      "loss": 0.2468,
      "num_input_tokens_seen": 18571752,
      "step": 28660
    },
    {
      "epoch": 16.901533018867923,
      "grad_norm": 7.603405952453613,
      "learning_rate": 7.13739539956525e-07,
      "loss": 0.2674,
      "num_input_tokens_seen": 18574568,
      "step": 28665
    },
    {
      "epoch": 16.90448113207547,
      "grad_norm": 2.9606106281280518,
      "learning_rate": 7.124154248155562e-07,
      "loss": 0.2893,
      "num_input_tokens_seen": 18578152,
      "step": 28670
    },
    {
      "epoch": 16.90742924528302,
      "grad_norm": 5.1635894775390625,
      "learning_rate": 7.110924448310813e-07,
      "loss": 0.2776,
      "num_input_tokens_seen": 18580936,
      "step": 28675
    },
    {
      "epoch": 16.910377358490567,
      "grad_norm": 2.994457960128784,
      "learning_rate": 7.097706003533666e-07,
      "loss": 0.3491,
      "num_input_tokens_seen": 18584808,
      "step": 28680
    },
    {
      "epoch": 16.913325471698112,
      "grad_norm": 3.7430410385131836,
      "learning_rate": 7.084498917323751e-07,
      "loss": 0.3242,
      "num_input_tokens_seen": 18588552,
      "step": 28685
    },
    {
      "epoch": 16.91627358490566,
      "grad_norm": 2.873523712158203,
      "learning_rate": 7.071303193177698e-07,
      "loss": 0.2771,
      "num_input_tokens_seen": 18591976,
      "step": 28690
    },
    {
      "epoch": 16.91922169811321,
      "grad_norm": 3.1893057823181152,
      "learning_rate": 7.058118834589133e-07,
      "loss": 0.1691,
      "num_input_tokens_seen": 18596264,
      "step": 28695
    },
    {
      "epoch": 16.922169811320753,
      "grad_norm": 4.254721641540527,
      "learning_rate": 7.044945845048684e-07,
      "loss": 0.2254,
      "num_input_tokens_seen": 18598824,
      "step": 28700
    },
    {
      "epoch": 16.9251179245283,
      "grad_norm": 2.4068586826324463,
      "learning_rate": 7.031784228043948e-07,
      "loss": 0.1989,
      "num_input_tokens_seen": 18601544,
      "step": 28705
    },
    {
      "epoch": 16.92806603773585,
      "grad_norm": 4.738065242767334,
      "learning_rate": 7.01863398705952e-07,
      "loss": 0.3047,
      "num_input_tokens_seen": 18603912,
      "step": 28710
    },
    {
      "epoch": 16.931014150943398,
      "grad_norm": 2.8705801963806152,
      "learning_rate": 7.005495125576983e-07,
      "loss": 0.2753,
      "num_input_tokens_seen": 18606728,
      "step": 28715
    },
    {
      "epoch": 16.933962264150942,
      "grad_norm": 3.899610996246338,
      "learning_rate": 6.99236764707491e-07,
      "loss": 0.2566,
      "num_input_tokens_seen": 18609768,
      "step": 28720
    },
    {
      "epoch": 16.93691037735849,
      "grad_norm": 3.991589069366455,
      "learning_rate": 6.979251555028843e-07,
      "loss": 0.3112,
      "num_input_tokens_seen": 18613992,
      "step": 28725
    },
    {
      "epoch": 16.93985849056604,
      "grad_norm": 3.2520174980163574,
      "learning_rate": 6.966146852911332e-07,
      "loss": 0.3604,
      "num_input_tokens_seen": 18616744,
      "step": 28730
    },
    {
      "epoch": 16.942806603773583,
      "grad_norm": 3.4976673126220703,
      "learning_rate": 6.953053544191923e-07,
      "loss": 0.3024,
      "num_input_tokens_seen": 18619400,
      "step": 28735
    },
    {
      "epoch": 16.94575471698113,
      "grad_norm": 4.331892967224121,
      "learning_rate": 6.939971632337111e-07,
      "loss": 0.3764,
      "num_input_tokens_seen": 18622920,
      "step": 28740
    },
    {
      "epoch": 16.94870283018868,
      "grad_norm": 6.880521774291992,
      "learning_rate": 6.926901120810387e-07,
      "loss": 0.4061,
      "num_input_tokens_seen": 18625640,
      "step": 28745
    },
    {
      "epoch": 16.951650943396228,
      "grad_norm": 3.4801604747772217,
      "learning_rate": 6.91384201307222e-07,
      "loss": 0.3701,
      "num_input_tokens_seen": 18629896,
      "step": 28750
    },
    {
      "epoch": 16.954599056603772,
      "grad_norm": 3.13507342338562,
      "learning_rate": 6.900794312580078e-07,
      "loss": 0.3646,
      "num_input_tokens_seen": 18635176,
      "step": 28755
    },
    {
      "epoch": 16.95754716981132,
      "grad_norm": 3.6265294551849365,
      "learning_rate": 6.887758022788377e-07,
      "loss": 0.324,
      "num_input_tokens_seen": 18638344,
      "step": 28760
    },
    {
      "epoch": 16.96049528301887,
      "grad_norm": 6.023173809051514,
      "learning_rate": 6.874733147148549e-07,
      "loss": 0.3198,
      "num_input_tokens_seen": 18641480,
      "step": 28765
    },
    {
      "epoch": 16.963443396226417,
      "grad_norm": 4.5023627281188965,
      "learning_rate": 6.861719689108987e-07,
      "loss": 0.2978,
      "num_input_tokens_seen": 18644552,
      "step": 28770
    },
    {
      "epoch": 16.96639150943396,
      "grad_norm": 2.8490779399871826,
      "learning_rate": 6.84871765211505e-07,
      "loss": 0.2349,
      "num_input_tokens_seen": 18647464,
      "step": 28775
    },
    {
      "epoch": 16.96933962264151,
      "grad_norm": 3.1556894779205322,
      "learning_rate": 6.835727039609086e-07,
      "loss": 0.3165,
      "num_input_tokens_seen": 18651368,
      "step": 28780
    },
    {
      "epoch": 16.972287735849058,
      "grad_norm": 4.456608772277832,
      "learning_rate": 6.822747855030415e-07,
      "loss": 0.3272,
      "num_input_tokens_seen": 18654120,
      "step": 28785
    },
    {
      "epoch": 16.975235849056602,
      "grad_norm": 3.0523276329040527,
      "learning_rate": 6.809780101815322e-07,
      "loss": 0.222,
      "num_input_tokens_seen": 18656904,
      "step": 28790
    },
    {
      "epoch": 16.97818396226415,
      "grad_norm": 5.021287441253662,
      "learning_rate": 6.796823783397099e-07,
      "loss": 0.364,
      "num_input_tokens_seen": 18659528,
      "step": 28795
    },
    {
      "epoch": 16.9811320754717,
      "grad_norm": 4.4093732833862305,
      "learning_rate": 6.783878903205976e-07,
      "loss": 0.1704,
      "num_input_tokens_seen": 18662248,
      "step": 28800
    },
    {
      "epoch": 16.984080188679247,
      "grad_norm": 3.1740529537200928,
      "learning_rate": 6.77094546466916e-07,
      "loss": 0.2428,
      "num_input_tokens_seen": 18665096,
      "step": 28805
    },
    {
      "epoch": 16.98702830188679,
      "grad_norm": 5.569344997406006,
      "learning_rate": 6.758023471210845e-07,
      "loss": 0.3025,
      "num_input_tokens_seen": 18668296,
      "step": 28810
    },
    {
      "epoch": 16.98997641509434,
      "grad_norm": 2.692049026489258,
      "learning_rate": 6.745112926252162e-07,
      "loss": 0.3564,
      "num_input_tokens_seen": 18672616,
      "step": 28815
    },
    {
      "epoch": 16.992924528301888,
      "grad_norm": 4.392744541168213,
      "learning_rate": 6.732213833211265e-07,
      "loss": 0.3098,
      "num_input_tokens_seen": 18675944,
      "step": 28820
    },
    {
      "epoch": 16.995872641509433,
      "grad_norm": 4.2128095626831055,
      "learning_rate": 6.719326195503218e-07,
      "loss": 0.4068,
      "num_input_tokens_seen": 18679624,
      "step": 28825
    },
    {
      "epoch": 16.99882075471698,
      "grad_norm": 7.917954921722412,
      "learning_rate": 6.706450016540094e-07,
      "loss": 0.3444,
      "num_input_tokens_seen": 18682632,
      "step": 28830
    },
    {
      "epoch": 17.00176886792453,
      "grad_norm": 3.9436824321746826,
      "learning_rate": 6.69358529973092e-07,
      "loss": 0.3027,
      "num_input_tokens_seen": 18685328,
      "step": 28835
    },
    {
      "epoch": 17.004716981132077,
      "grad_norm": 2.6487109661102295,
      "learning_rate": 6.680732048481681e-07,
      "loss": 0.218,
      "num_input_tokens_seen": 18688592,
      "step": 28840
    },
    {
      "epoch": 17.00766509433962,
      "grad_norm": 4.891993522644043,
      "learning_rate": 6.667890266195321e-07,
      "loss": 0.3229,
      "num_input_tokens_seen": 18691504,
      "step": 28845
    },
    {
      "epoch": 17.01061320754717,
      "grad_norm": 3.5471436977386475,
      "learning_rate": 6.655059956271759e-07,
      "loss": 0.3416,
      "num_input_tokens_seen": 18695152,
      "step": 28850
    },
    {
      "epoch": 17.013561320754718,
      "grad_norm": 4.140858173370361,
      "learning_rate": 6.642241122107884e-07,
      "loss": 0.2644,
      "num_input_tokens_seen": 18700880,
      "step": 28855
    },
    {
      "epoch": 17.016509433962263,
      "grad_norm": 3.1776533126831055,
      "learning_rate": 6.629433767097537e-07,
      "loss": 0.321,
      "num_input_tokens_seen": 18703632,
      "step": 28860
    },
    {
      "epoch": 17.01945754716981,
      "grad_norm": 6.1613383293151855,
      "learning_rate": 6.616637894631517e-07,
      "loss": 0.2817,
      "num_input_tokens_seen": 18706416,
      "step": 28865
    },
    {
      "epoch": 17.02240566037736,
      "grad_norm": 6.2424540519714355,
      "learning_rate": 6.603853508097591e-07,
      "loss": 0.2939,
      "num_input_tokens_seen": 18709936,
      "step": 28870
    },
    {
      "epoch": 17.025353773584907,
      "grad_norm": 3.470947742462158,
      "learning_rate": 6.591080610880468e-07,
      "loss": 0.2725,
      "num_input_tokens_seen": 18712752,
      "step": 28875
    },
    {
      "epoch": 17.028301886792452,
      "grad_norm": 5.3195648193359375,
      "learning_rate": 6.578319206361828e-07,
      "loss": 0.2521,
      "num_input_tokens_seen": 18715248,
      "step": 28880
    },
    {
      "epoch": 17.03125,
      "grad_norm": 2.7565531730651855,
      "learning_rate": 6.565569297920327e-07,
      "loss": 0.3168,
      "num_input_tokens_seen": 18718384,
      "step": 28885
    },
    {
      "epoch": 17.034198113207548,
      "grad_norm": 8.494420051574707,
      "learning_rate": 6.552830888931544e-07,
      "loss": 0.2837,
      "num_input_tokens_seen": 18721168,
      "step": 28890
    },
    {
      "epoch": 17.037146226415093,
      "grad_norm": 2.7407338619232178,
      "learning_rate": 6.540103982768031e-07,
      "loss": 0.3538,
      "num_input_tokens_seen": 18724496,
      "step": 28895
    },
    {
      "epoch": 17.04009433962264,
      "grad_norm": 4.085362911224365,
      "learning_rate": 6.527388582799293e-07,
      "loss": 0.2045,
      "num_input_tokens_seen": 18726896,
      "step": 28900
    },
    {
      "epoch": 17.04304245283019,
      "grad_norm": 4.3741607666015625,
      "learning_rate": 6.514684692391782e-07,
      "loss": 0.3317,
      "num_input_tokens_seen": 18729520,
      "step": 28905
    },
    {
      "epoch": 17.045990566037737,
      "grad_norm": 3.737673759460449,
      "learning_rate": 6.501992314908895e-07,
      "loss": 0.3566,
      "num_input_tokens_seen": 18733520,
      "step": 28910
    },
    {
      "epoch": 17.048938679245282,
      "grad_norm": 3.7173397541046143,
      "learning_rate": 6.489311453711017e-07,
      "loss": 0.1928,
      "num_input_tokens_seen": 18737744,
      "step": 28915
    },
    {
      "epoch": 17.05188679245283,
      "grad_norm": 3.9551498889923096,
      "learning_rate": 6.476642112155457e-07,
      "loss": 0.2471,
      "num_input_tokens_seen": 18740752,
      "step": 28920
    },
    {
      "epoch": 17.05483490566038,
      "grad_norm": 8.120649337768555,
      "learning_rate": 6.463984293596476e-07,
      "loss": 0.3387,
      "num_input_tokens_seen": 18743152,
      "step": 28925
    },
    {
      "epoch": 17.057783018867923,
      "grad_norm": 6.5483269691467285,
      "learning_rate": 6.451338001385282e-07,
      "loss": 0.3217,
      "num_input_tokens_seen": 18746064,
      "step": 28930
    },
    {
      "epoch": 17.06073113207547,
      "grad_norm": 3.5575246810913086,
      "learning_rate": 6.438703238870037e-07,
      "loss": 0.2384,
      "num_input_tokens_seen": 18750928,
      "step": 28935
    },
    {
      "epoch": 17.06367924528302,
      "grad_norm": 3.0641305446624756,
      "learning_rate": 6.426080009395846e-07,
      "loss": 0.4204,
      "num_input_tokens_seen": 18754032,
      "step": 28940
    },
    {
      "epoch": 17.066627358490567,
      "grad_norm": 4.734625816345215,
      "learning_rate": 6.413468316304755e-07,
      "loss": 0.402,
      "num_input_tokens_seen": 18757136,
      "step": 28945
    },
    {
      "epoch": 17.069575471698112,
      "grad_norm": 4.00379753112793,
      "learning_rate": 6.400868162935786e-07,
      "loss": 0.2835,
      "num_input_tokens_seen": 18760336,
      "step": 28950
    },
    {
      "epoch": 17.07252358490566,
      "grad_norm": 6.473146438598633,
      "learning_rate": 6.388279552624877e-07,
      "loss": 0.3503,
      "num_input_tokens_seen": 18763504,
      "step": 28955
    },
    {
      "epoch": 17.07547169811321,
      "grad_norm": 5.4554314613342285,
      "learning_rate": 6.37570248870491e-07,
      "loss": 0.2554,
      "num_input_tokens_seen": 18765744,
      "step": 28960
    },
    {
      "epoch": 17.078419811320753,
      "grad_norm": 5.899824142456055,
      "learning_rate": 6.363136974505718e-07,
      "loss": 0.2274,
      "num_input_tokens_seen": 18768976,
      "step": 28965
    },
    {
      "epoch": 17.0813679245283,
      "grad_norm": 3.018824577331543,
      "learning_rate": 6.350583013354078e-07,
      "loss": 0.1852,
      "num_input_tokens_seen": 18771600,
      "step": 28970
    },
    {
      "epoch": 17.08431603773585,
      "grad_norm": 7.096885681152344,
      "learning_rate": 6.338040608573693e-07,
      "loss": 0.278,
      "num_input_tokens_seen": 18773840,
      "step": 28975
    },
    {
      "epoch": 17.087264150943398,
      "grad_norm": 5.422340393066406,
      "learning_rate": 6.325509763485238e-07,
      "loss": 0.4528,
      "num_input_tokens_seen": 18777872,
      "step": 28980
    },
    {
      "epoch": 17.090212264150942,
      "grad_norm": 3.804208278656006,
      "learning_rate": 6.312990481406301e-07,
      "loss": 0.2726,
      "num_input_tokens_seen": 18780560,
      "step": 28985
    },
    {
      "epoch": 17.09316037735849,
      "grad_norm": 4.466885089874268,
      "learning_rate": 6.300482765651411e-07,
      "loss": 0.2952,
      "num_input_tokens_seen": 18783216,
      "step": 28990
    },
    {
      "epoch": 17.09610849056604,
      "grad_norm": 6.320814609527588,
      "learning_rate": 6.28798661953205e-07,
      "loss": 0.2434,
      "num_input_tokens_seen": 18785904,
      "step": 28995
    },
    {
      "epoch": 17.099056603773583,
      "grad_norm": 2.0503106117248535,
      "learning_rate": 6.275502046356618e-07,
      "loss": 0.2932,
      "num_input_tokens_seen": 18788464,
      "step": 29000
    },
    {
      "epoch": 17.10200471698113,
      "grad_norm": 1.246903657913208,
      "learning_rate": 6.263029049430447e-07,
      "loss": 0.2415,
      "num_input_tokens_seen": 18791344,
      "step": 29005
    },
    {
      "epoch": 17.10495283018868,
      "grad_norm": 3.2167093753814697,
      "learning_rate": 6.250567632055832e-07,
      "loss": 0.2942,
      "num_input_tokens_seen": 18794352,
      "step": 29010
    },
    {
      "epoch": 17.107900943396228,
      "grad_norm": 4.980419158935547,
      "learning_rate": 6.238117797532e-07,
      "loss": 0.3439,
      "num_input_tokens_seen": 18797776,
      "step": 29015
    },
    {
      "epoch": 17.110849056603772,
      "grad_norm": 5.311582088470459,
      "learning_rate": 6.225679549155083e-07,
      "loss": 0.3111,
      "num_input_tokens_seen": 18800272,
      "step": 29020
    },
    {
      "epoch": 17.11379716981132,
      "grad_norm": 3.9590821266174316,
      "learning_rate": 6.213252890218163e-07,
      "loss": 0.3825,
      "num_input_tokens_seen": 18803504,
      "step": 29025
    },
    {
      "epoch": 17.11674528301887,
      "grad_norm": 3.029886245727539,
      "learning_rate": 6.200837824011247e-07,
      "loss": 0.3308,
      "num_input_tokens_seen": 18806576,
      "step": 29030
    },
    {
      "epoch": 17.119693396226417,
      "grad_norm": 2.5309221744537354,
      "learning_rate": 6.188434353821282e-07,
      "loss": 0.2316,
      "num_input_tokens_seen": 18809744,
      "step": 29035
    },
    {
      "epoch": 17.12264150943396,
      "grad_norm": 3.4076783657073975,
      "learning_rate": 6.176042482932132e-07,
      "loss": 0.3867,
      "num_input_tokens_seen": 18812688,
      "step": 29040
    },
    {
      "epoch": 17.12558962264151,
      "grad_norm": 4.76323127746582,
      "learning_rate": 6.163662214624616e-07,
      "loss": 0.3551,
      "num_input_tokens_seen": 18814960,
      "step": 29045
    },
    {
      "epoch": 17.128537735849058,
      "grad_norm": 7.127803802490234,
      "learning_rate": 6.151293552176451e-07,
      "loss": 0.2692,
      "num_input_tokens_seen": 18817520,
      "step": 29050
    },
    {
      "epoch": 17.131485849056602,
      "grad_norm": 4.578367233276367,
      "learning_rate": 6.138936498862291e-07,
      "loss": 0.2623,
      "num_input_tokens_seen": 18820400,
      "step": 29055
    },
    {
      "epoch": 17.13443396226415,
      "grad_norm": 6.523787975311279,
      "learning_rate": 6.126591057953729e-07,
      "loss": 0.3677,
      "num_input_tokens_seen": 18823504,
      "step": 29060
    },
    {
      "epoch": 17.1373820754717,
      "grad_norm": 3.247335910797119,
      "learning_rate": 6.114257232719267e-07,
      "loss": 0.2743,
      "num_input_tokens_seen": 18826448,
      "step": 29065
    },
    {
      "epoch": 17.140330188679247,
      "grad_norm": 6.972044944763184,
      "learning_rate": 6.101935026424332e-07,
      "loss": 0.2579,
      "num_input_tokens_seen": 18830224,
      "step": 29070
    },
    {
      "epoch": 17.14327830188679,
      "grad_norm": 3.5693252086639404,
      "learning_rate": 6.089624442331293e-07,
      "loss": 0.2768,
      "num_input_tokens_seen": 18833552,
      "step": 29075
    },
    {
      "epoch": 17.14622641509434,
      "grad_norm": 3.7503905296325684,
      "learning_rate": 6.077325483699432e-07,
      "loss": 0.3207,
      "num_input_tokens_seen": 18837072,
      "step": 29080
    },
    {
      "epoch": 17.149174528301888,
      "grad_norm": 2.187403440475464,
      "learning_rate": 6.065038153784947e-07,
      "loss": 0.3765,
      "num_input_tokens_seen": 18847952,
      "step": 29085
    },
    {
      "epoch": 17.152122641509433,
      "grad_norm": 2.0181925296783447,
      "learning_rate": 6.052762455840955e-07,
      "loss": 0.2233,
      "num_input_tokens_seen": 18851696,
      "step": 29090
    },
    {
      "epoch": 17.15507075471698,
      "grad_norm": 5.954942226409912,
      "learning_rate": 6.040498393117494e-07,
      "loss": 0.2354,
      "num_input_tokens_seen": 18854320,
      "step": 29095
    },
    {
      "epoch": 17.15801886792453,
      "grad_norm": 4.1953301429748535,
      "learning_rate": 6.028245968861551e-07,
      "loss": 0.2232,
      "num_input_tokens_seen": 18858736,
      "step": 29100
    },
    {
      "epoch": 17.160966981132077,
      "grad_norm": 3.4732844829559326,
      "learning_rate": 6.016005186316987e-07,
      "loss": 0.3879,
      "num_input_tokens_seen": 18862448,
      "step": 29105
    },
    {
      "epoch": 17.16391509433962,
      "grad_norm": 2.8293569087982178,
      "learning_rate": 6.003776048724614e-07,
      "loss": 0.322,
      "num_input_tokens_seen": 18865264,
      "step": 29110
    },
    {
      "epoch": 17.16686320754717,
      "grad_norm": 2.849020004272461,
      "learning_rate": 5.991558559322152e-07,
      "loss": 0.3085,
      "num_input_tokens_seen": 18867920,
      "step": 29115
    },
    {
      "epoch": 17.169811320754718,
      "grad_norm": 3.1719582080841064,
      "learning_rate": 5.979352721344223e-07,
      "loss": 0.2814,
      "num_input_tokens_seen": 18870320,
      "step": 29120
    },
    {
      "epoch": 17.172759433962263,
      "grad_norm": 3.2966747283935547,
      "learning_rate": 5.967158538022383e-07,
      "loss": 0.3381,
      "num_input_tokens_seen": 18873360,
      "step": 29125
    },
    {
      "epoch": 17.17570754716981,
      "grad_norm": 5.29883337020874,
      "learning_rate": 5.954976012585078e-07,
      "loss": 0.399,
      "num_input_tokens_seen": 18877776,
      "step": 29130
    },
    {
      "epoch": 17.17865566037736,
      "grad_norm": 3.3464725017547607,
      "learning_rate": 5.942805148257713e-07,
      "loss": 0.2702,
      "num_input_tokens_seen": 18880688,
      "step": 29135
    },
    {
      "epoch": 17.181603773584907,
      "grad_norm": 4.923304080963135,
      "learning_rate": 5.930645948262553e-07,
      "loss": 0.3401,
      "num_input_tokens_seen": 18883152,
      "step": 29140
    },
    {
      "epoch": 17.184551886792452,
      "grad_norm": 3.4681594371795654,
      "learning_rate": 5.918498415818813e-07,
      "loss": 0.277,
      "num_input_tokens_seen": 18886960,
      "step": 29145
    },
    {
      "epoch": 17.1875,
      "grad_norm": 5.231708526611328,
      "learning_rate": 5.906362554142592e-07,
      "loss": 0.2572,
      "num_input_tokens_seen": 18889872,
      "step": 29150
    },
    {
      "epoch": 17.190448113207548,
      "grad_norm": 3.689689874649048,
      "learning_rate": 5.894238366446925e-07,
      "loss": 0.3068,
      "num_input_tokens_seen": 18892560,
      "step": 29155
    },
    {
      "epoch": 17.193396226415093,
      "grad_norm": 3.0281100273132324,
      "learning_rate": 5.882125855941723e-07,
      "loss": 0.2845,
      "num_input_tokens_seen": 18895696,
      "step": 29160
    },
    {
      "epoch": 17.19634433962264,
      "grad_norm": 4.9863786697387695,
      "learning_rate": 5.870025025833842e-07,
      "loss": 0.3405,
      "num_input_tokens_seen": 18898672,
      "step": 29165
    },
    {
      "epoch": 17.19929245283019,
      "grad_norm": 4.750463962554932,
      "learning_rate": 5.857935879327031e-07,
      "loss": 0.5,
      "num_input_tokens_seen": 18901776,
      "step": 29170
    },
    {
      "epoch": 17.202240566037737,
      "grad_norm": 2.8354806900024414,
      "learning_rate": 5.845858419621936e-07,
      "loss": 0.2009,
      "num_input_tokens_seen": 18905232,
      "step": 29175
    },
    {
      "epoch": 17.205188679245282,
      "grad_norm": 7.502776622772217,
      "learning_rate": 5.83379264991612e-07,
      "loss": 0.374,
      "num_input_tokens_seen": 18908208,
      "step": 29180
    },
    {
      "epoch": 17.20813679245283,
      "grad_norm": 3.4745774269104004,
      "learning_rate": 5.821738573404046e-07,
      "loss": 0.2539,
      "num_input_tokens_seen": 18911696,
      "step": 29185
    },
    {
      "epoch": 17.21108490566038,
      "grad_norm": 4.8069167137146,
      "learning_rate": 5.80969619327707e-07,
      "loss": 0.2783,
      "num_input_tokens_seen": 18915024,
      "step": 29190
    },
    {
      "epoch": 17.214033018867923,
      "grad_norm": 2.603271007537842,
      "learning_rate": 5.797665512723488e-07,
      "loss": 0.244,
      "num_input_tokens_seen": 18919024,
      "step": 29195
    },
    {
      "epoch": 17.21698113207547,
      "grad_norm": 3.5695078372955322,
      "learning_rate": 5.785646534928452e-07,
      "loss": 0.1962,
      "num_input_tokens_seen": 18921456,
      "step": 29200
    },
    {
      "epoch": 17.21992924528302,
      "grad_norm": 2.7819414138793945,
      "learning_rate": 5.77363926307406e-07,
      "loss": 0.2536,
      "num_input_tokens_seen": 18924496,
      "step": 29205
    },
    {
      "epoch": 17.222877358490567,
      "grad_norm": 4.4421892166137695,
      "learning_rate": 5.761643700339281e-07,
      "loss": 0.3302,
      "num_input_tokens_seen": 18927376,
      "step": 29210
    },
    {
      "epoch": 17.225825471698112,
      "grad_norm": 5.932154178619385,
      "learning_rate": 5.749659849899985e-07,
      "loss": 0.3994,
      "num_input_tokens_seen": 18930768,
      "step": 29215
    },
    {
      "epoch": 17.22877358490566,
      "grad_norm": 6.423239231109619,
      "learning_rate": 5.737687714928953e-07,
      "loss": 0.3165,
      "num_input_tokens_seen": 18933904,
      "step": 29220
    },
    {
      "epoch": 17.23172169811321,
      "grad_norm": 3.930244207382202,
      "learning_rate": 5.725727298595846e-07,
      "loss": 0.2406,
      "num_input_tokens_seen": 18937648,
      "step": 29225
    },
    {
      "epoch": 17.234669811320753,
      "grad_norm": 1.7958842515945435,
      "learning_rate": 5.71377860406726e-07,
      "loss": 0.2938,
      "num_input_tokens_seen": 18941200,
      "step": 29230
    },
    {
      "epoch": 17.2376179245283,
      "grad_norm": 6.7675676345825195,
      "learning_rate": 5.701841634506655e-07,
      "loss": 0.2746,
      "num_input_tokens_seen": 18944848,
      "step": 29235
    },
    {
      "epoch": 17.24056603773585,
      "grad_norm": 7.620945930480957,
      "learning_rate": 5.689916393074391e-07,
      "loss": 0.2312,
      "num_input_tokens_seen": 18947728,
      "step": 29240
    },
    {
      "epoch": 17.243514150943398,
      "grad_norm": 6.490934371948242,
      "learning_rate": 5.678002882927725e-07,
      "loss": 0.3148,
      "num_input_tokens_seen": 18950672,
      "step": 29245
    },
    {
      "epoch": 17.246462264150942,
      "grad_norm": 4.0157270431518555,
      "learning_rate": 5.666101107220811e-07,
      "loss": 0.296,
      "num_input_tokens_seen": 18953872,
      "step": 29250
    },
    {
      "epoch": 17.24941037735849,
      "grad_norm": 3.0491042137145996,
      "learning_rate": 5.654211069104693e-07,
      "loss": 0.2686,
      "num_input_tokens_seen": 18957168,
      "step": 29255
    },
    {
      "epoch": 17.25235849056604,
      "grad_norm": 5.486398696899414,
      "learning_rate": 5.642332771727321e-07,
      "loss": 0.4083,
      "num_input_tokens_seen": 18960272,
      "step": 29260
    },
    {
      "epoch": 17.255306603773583,
      "grad_norm": 3.89927339553833,
      "learning_rate": 5.630466218233521e-07,
      "loss": 0.3148,
      "num_input_tokens_seen": 18963440,
      "step": 29265
    },
    {
      "epoch": 17.25825471698113,
      "grad_norm": 2.5919885635375977,
      "learning_rate": 5.618611411765007e-07,
      "loss": 0.3423,
      "num_input_tokens_seen": 18966544,
      "step": 29270
    },
    {
      "epoch": 17.26120283018868,
      "grad_norm": 4.298869609832764,
      "learning_rate": 5.606768355460401e-07,
      "loss": 0.2354,
      "num_input_tokens_seen": 18969520,
      "step": 29275
    },
    {
      "epoch": 17.264150943396228,
      "grad_norm": 5.2090349197387695,
      "learning_rate": 5.594937052455191e-07,
      "loss": 0.3082,
      "num_input_tokens_seen": 18972176,
      "step": 29280
    },
    {
      "epoch": 17.267099056603772,
      "grad_norm": 7.558565139770508,
      "learning_rate": 5.583117505881764e-07,
      "loss": 0.234,
      "num_input_tokens_seen": 18974512,
      "step": 29285
    },
    {
      "epoch": 17.27004716981132,
      "grad_norm": 3.579770803451538,
      "learning_rate": 5.571309718869417e-07,
      "loss": 0.2997,
      "num_input_tokens_seen": 18978192,
      "step": 29290
    },
    {
      "epoch": 17.27299528301887,
      "grad_norm": 3.551110029220581,
      "learning_rate": 5.559513694544282e-07,
      "loss": 0.2827,
      "num_input_tokens_seen": 18981648,
      "step": 29295
    },
    {
      "epoch": 17.275943396226417,
      "grad_norm": 3.080820083618164,
      "learning_rate": 5.547729436029442e-07,
      "loss": 0.2536,
      "num_input_tokens_seen": 18984656,
      "step": 29300
    },
    {
      "epoch": 17.27889150943396,
      "grad_norm": 4.549503326416016,
      "learning_rate": 5.535956946444809e-07,
      "loss": 0.3105,
      "num_input_tokens_seen": 18987440,
      "step": 29305
    },
    {
      "epoch": 17.28183962264151,
      "grad_norm": 3.185995101928711,
      "learning_rate": 5.524196228907203e-07,
      "loss": 0.3654,
      "num_input_tokens_seen": 18991312,
      "step": 29310
    },
    {
      "epoch": 17.284787735849058,
      "grad_norm": 5.1740570068359375,
      "learning_rate": 5.512447286530326e-07,
      "loss": 0.3436,
      "num_input_tokens_seen": 18994768,
      "step": 29315
    },
    {
      "epoch": 17.287735849056602,
      "grad_norm": 3.829077959060669,
      "learning_rate": 5.500710122424746e-07,
      "loss": 0.3287,
      "num_input_tokens_seen": 18997456,
      "step": 29320
    },
    {
      "epoch": 17.29068396226415,
      "grad_norm": 4.195986270904541,
      "learning_rate": 5.488984739697961e-07,
      "loss": 0.2913,
      "num_input_tokens_seen": 19000368,
      "step": 29325
    },
    {
      "epoch": 17.2936320754717,
      "grad_norm": 3.344460964202881,
      "learning_rate": 5.477271141454294e-07,
      "loss": 0.4392,
      "num_input_tokens_seen": 19003408,
      "step": 29330
    },
    {
      "epoch": 17.296580188679247,
      "grad_norm": 5.246891498565674,
      "learning_rate": 5.465569330794974e-07,
      "loss": 0.2997,
      "num_input_tokens_seen": 19006128,
      "step": 29335
    },
    {
      "epoch": 17.29952830188679,
      "grad_norm": 5.230263710021973,
      "learning_rate": 5.453879310818105e-07,
      "loss": 0.2913,
      "num_input_tokens_seen": 19009200,
      "step": 29340
    },
    {
      "epoch": 17.30247641509434,
      "grad_norm": 4.099724769592285,
      "learning_rate": 5.442201084618664e-07,
      "loss": 0.2849,
      "num_input_tokens_seen": 19012400,
      "step": 29345
    },
    {
      "epoch": 17.305424528301888,
      "grad_norm": 4.311807632446289,
      "learning_rate": 5.430534655288528e-07,
      "loss": 0.3128,
      "num_input_tokens_seen": 19015440,
      "step": 29350
    },
    {
      "epoch": 17.308372641509433,
      "grad_norm": 3.017354726791382,
      "learning_rate": 5.418880025916428e-07,
      "loss": 0.2996,
      "num_input_tokens_seen": 19019408,
      "step": 29355
    },
    {
      "epoch": 17.31132075471698,
      "grad_norm": 5.0263824462890625,
      "learning_rate": 5.407237199587973e-07,
      "loss": 0.3307,
      "num_input_tokens_seen": 19022768,
      "step": 29360
    },
    {
      "epoch": 17.31426886792453,
      "grad_norm": 3.5152645111083984,
      "learning_rate": 5.395606179385654e-07,
      "loss": 0.3071,
      "num_input_tokens_seen": 19025456,
      "step": 29365
    },
    {
      "epoch": 17.317216981132077,
      "grad_norm": 5.2915544509887695,
      "learning_rate": 5.383986968388833e-07,
      "loss": 0.3337,
      "num_input_tokens_seen": 19028624,
      "step": 29370
    },
    {
      "epoch": 17.32016509433962,
      "grad_norm": 3.9995903968811035,
      "learning_rate": 5.372379569673736e-07,
      "loss": 0.4007,
      "num_input_tokens_seen": 19031760,
      "step": 29375
    },
    {
      "epoch": 17.32311320754717,
      "grad_norm": 3.9049108028411865,
      "learning_rate": 5.360783986313495e-07,
      "loss": 0.3415,
      "num_input_tokens_seen": 19035056,
      "step": 29380
    },
    {
      "epoch": 17.326061320754718,
      "grad_norm": 9.17409610748291,
      "learning_rate": 5.349200221378076e-07,
      "loss": 0.4009,
      "num_input_tokens_seen": 19038224,
      "step": 29385
    },
    {
      "epoch": 17.329009433962263,
      "grad_norm": 5.191521644592285,
      "learning_rate": 5.33762827793432e-07,
      "loss": 0.3381,
      "num_input_tokens_seen": 19041616,
      "step": 29390
    },
    {
      "epoch": 17.33195754716981,
      "grad_norm": 3.5190205574035645,
      "learning_rate": 5.326068159045978e-07,
      "loss": 0.3585,
      "num_input_tokens_seen": 19044624,
      "step": 29395
    },
    {
      "epoch": 17.33490566037736,
      "grad_norm": 4.664684772491455,
      "learning_rate": 5.314519867773621e-07,
      "loss": 0.3447,
      "num_input_tokens_seen": 19047504,
      "step": 29400
    },
    {
      "epoch": 17.337853773584907,
      "grad_norm": 3.6617603302001953,
      "learning_rate": 5.302983407174711e-07,
      "loss": 0.3223,
      "num_input_tokens_seen": 19050416,
      "step": 29405
    },
    {
      "epoch": 17.340801886792452,
      "grad_norm": 6.1026506423950195,
      "learning_rate": 5.291458780303572e-07,
      "loss": 0.3535,
      "num_input_tokens_seen": 19053200,
      "step": 29410
    },
    {
      "epoch": 17.34375,
      "grad_norm": 1.71602201461792,
      "learning_rate": 5.279945990211411e-07,
      "loss": 0.1941,
      "num_input_tokens_seen": 19056176,
      "step": 29415
    },
    {
      "epoch": 17.346698113207548,
      "grad_norm": 3.532514810562134,
      "learning_rate": 5.26844503994628e-07,
      "loss": 0.3619,
      "num_input_tokens_seen": 19059408,
      "step": 29420
    },
    {
      "epoch": 17.349646226415093,
      "grad_norm": 3.2471282482147217,
      "learning_rate": 5.25695593255311e-07,
      "loss": 0.346,
      "num_input_tokens_seen": 19062096,
      "step": 29425
    },
    {
      "epoch": 17.35259433962264,
      "grad_norm": 4.061997413635254,
      "learning_rate": 5.24547867107369e-07,
      "loss": 0.2863,
      "num_input_tokens_seen": 19065264,
      "step": 29430
    },
    {
      "epoch": 17.35554245283019,
      "grad_norm": 3.468367576599121,
      "learning_rate": 5.234013258546672e-07,
      "loss": 0.2549,
      "num_input_tokens_seen": 19069008,
      "step": 29435
    },
    {
      "epoch": 17.358490566037737,
      "grad_norm": 4.713547229766846,
      "learning_rate": 5.222559698007563e-07,
      "loss": 0.4399,
      "num_input_tokens_seen": 19072144,
      "step": 29440
    },
    {
      "epoch": 17.361438679245282,
      "grad_norm": 5.245474815368652,
      "learning_rate": 5.211117992488763e-07,
      "loss": 0.2839,
      "num_input_tokens_seen": 19075408,
      "step": 29445
    },
    {
      "epoch": 17.36438679245283,
      "grad_norm": 3.443026542663574,
      "learning_rate": 5.199688145019505e-07,
      "loss": 0.2826,
      "num_input_tokens_seen": 19079088,
      "step": 29450
    },
    {
      "epoch": 17.36733490566038,
      "grad_norm": 4.614526271820068,
      "learning_rate": 5.188270158625891e-07,
      "loss": 0.3151,
      "num_input_tokens_seen": 19081936,
      "step": 29455
    },
    {
      "epoch": 17.370283018867923,
      "grad_norm": 5.865871906280518,
      "learning_rate": 5.176864036330875e-07,
      "loss": 0.3468,
      "num_input_tokens_seen": 19084624,
      "step": 29460
    },
    {
      "epoch": 17.37323113207547,
      "grad_norm": 2.9791972637176514,
      "learning_rate": 5.165469781154287e-07,
      "loss": 0.337,
      "num_input_tokens_seen": 19088560,
      "step": 29465
    },
    {
      "epoch": 17.37617924528302,
      "grad_norm": 3.3941407203674316,
      "learning_rate": 5.154087396112789e-07,
      "loss": 0.2453,
      "num_input_tokens_seen": 19091408,
      "step": 29470
    },
    {
      "epoch": 17.379127358490567,
      "grad_norm": 4.7604546546936035,
      "learning_rate": 5.142716884219939e-07,
      "loss": 0.3065,
      "num_input_tokens_seen": 19094640,
      "step": 29475
    },
    {
      "epoch": 17.382075471698112,
      "grad_norm": 3.0304105281829834,
      "learning_rate": 5.131358248486118e-07,
      "loss": 0.226,
      "num_input_tokens_seen": 19097520,
      "step": 29480
    },
    {
      "epoch": 17.38502358490566,
      "grad_norm": 2.630188226699829,
      "learning_rate": 5.120011491918564e-07,
      "loss": 0.294,
      "num_input_tokens_seen": 19100208,
      "step": 29485
    },
    {
      "epoch": 17.38797169811321,
      "grad_norm": 3.134455680847168,
      "learning_rate": 5.108676617521402e-07,
      "loss": 0.2374,
      "num_input_tokens_seen": 19104368,
      "step": 29490
    },
    {
      "epoch": 17.390919811320753,
      "grad_norm": 4.772602081298828,
      "learning_rate": 5.097353628295571e-07,
      "loss": 0.2238,
      "num_input_tokens_seen": 19106544,
      "step": 29495
    },
    {
      "epoch": 17.3938679245283,
      "grad_norm": 5.688600063323975,
      "learning_rate": 5.086042527238893e-07,
      "loss": 0.2957,
      "num_input_tokens_seen": 19109520,
      "step": 29500
    },
    {
      "epoch": 17.39681603773585,
      "grad_norm": 3.6793179512023926,
      "learning_rate": 5.074743317346009e-07,
      "loss": 0.2622,
      "num_input_tokens_seen": 19112400,
      "step": 29505
    },
    {
      "epoch": 17.399764150943398,
      "grad_norm": 4.4846062660217285,
      "learning_rate": 5.063456001608458e-07,
      "loss": 0.2935,
      "num_input_tokens_seen": 19115440,
      "step": 29510
    },
    {
      "epoch": 17.402712264150942,
      "grad_norm": 2.92531681060791,
      "learning_rate": 5.052180583014599e-07,
      "loss": 0.2761,
      "num_input_tokens_seen": 19118704,
      "step": 29515
    },
    {
      "epoch": 17.40566037735849,
      "grad_norm": 3.695979118347168,
      "learning_rate": 5.04091706454965e-07,
      "loss": 0.3691,
      "num_input_tokens_seen": 19122576,
      "step": 29520
    },
    {
      "epoch": 17.40860849056604,
      "grad_norm": 3.578866720199585,
      "learning_rate": 5.029665449195665e-07,
      "loss": 0.2729,
      "num_input_tokens_seen": 19126000,
      "step": 29525
    },
    {
      "epoch": 17.411556603773583,
      "grad_norm": 4.741421699523926,
      "learning_rate": 5.018425739931559e-07,
      "loss": 0.2082,
      "num_input_tokens_seen": 19128816,
      "step": 29530
    },
    {
      "epoch": 17.41450471698113,
      "grad_norm": 4.930683135986328,
      "learning_rate": 5.007197939733099e-07,
      "loss": 0.3804,
      "num_input_tokens_seen": 19132592,
      "step": 29535
    },
    {
      "epoch": 17.41745283018868,
      "grad_norm": 4.325804233551025,
      "learning_rate": 4.995982051572895e-07,
      "loss": 0.241,
      "num_input_tokens_seen": 19135120,
      "step": 29540
    },
    {
      "epoch": 17.420400943396228,
      "grad_norm": 3.880765676498413,
      "learning_rate": 4.984778078420405e-07,
      "loss": 0.2482,
      "num_input_tokens_seen": 19138512,
      "step": 29545
    },
    {
      "epoch": 17.423349056603772,
      "grad_norm": 5.827296733856201,
      "learning_rate": 4.973586023241917e-07,
      "loss": 0.3981,
      "num_input_tokens_seen": 19140912,
      "step": 29550
    },
    {
      "epoch": 17.42629716981132,
      "grad_norm": 4.868558883666992,
      "learning_rate": 4.962405889000588e-07,
      "loss": 0.2265,
      "num_input_tokens_seen": 19145840,
      "step": 29555
    },
    {
      "epoch": 17.42924528301887,
      "grad_norm": 2.8885960578918457,
      "learning_rate": 4.951237678656396e-07,
      "loss": 0.2333,
      "num_input_tokens_seen": 19148688,
      "step": 29560
    },
    {
      "epoch": 17.432193396226417,
      "grad_norm": 4.736023902893066,
      "learning_rate": 4.940081395166174e-07,
      "loss": 0.2851,
      "num_input_tokens_seen": 19151728,
      "step": 29565
    },
    {
      "epoch": 17.43514150943396,
      "grad_norm": 4.222837924957275,
      "learning_rate": 4.928937041483606e-07,
      "loss": 0.2764,
      "num_input_tokens_seen": 19154544,
      "step": 29570
    },
    {
      "epoch": 17.43808962264151,
      "grad_norm": 3.4310951232910156,
      "learning_rate": 4.917804620559202e-07,
      "loss": 0.3282,
      "num_input_tokens_seen": 19158192,
      "step": 29575
    },
    {
      "epoch": 17.441037735849058,
      "grad_norm": 4.094305038452148,
      "learning_rate": 4.906684135340317e-07,
      "loss": 0.2933,
      "num_input_tokens_seen": 19160976,
      "step": 29580
    },
    {
      "epoch": 17.443985849056602,
      "grad_norm": 4.427354335784912,
      "learning_rate": 4.89557558877114e-07,
      "loss": 0.2836,
      "num_input_tokens_seen": 19163856,
      "step": 29585
    },
    {
      "epoch": 17.44693396226415,
      "grad_norm": 7.755279541015625,
      "learning_rate": 4.884478983792728e-07,
      "loss": 0.2545,
      "num_input_tokens_seen": 19166224,
      "step": 29590
    },
    {
      "epoch": 17.4498820754717,
      "grad_norm": 3.0252413749694824,
      "learning_rate": 4.873394323342939e-07,
      "loss": 0.2775,
      "num_input_tokens_seen": 19171088,
      "step": 29595
    },
    {
      "epoch": 17.452830188679247,
      "grad_norm": 4.784431457519531,
      "learning_rate": 4.86232161035648e-07,
      "loss": 0.3897,
      "num_input_tokens_seen": 19173936,
      "step": 29600
    },
    {
      "epoch": 17.45577830188679,
      "grad_norm": 2.94970703125,
      "learning_rate": 4.851260847764916e-07,
      "loss": 0.2725,
      "num_input_tokens_seen": 19177104,
      "step": 29605
    },
    {
      "epoch": 17.45872641509434,
      "grad_norm": 5.263039588928223,
      "learning_rate": 4.840212038496622e-07,
      "loss": 0.5587,
      "num_input_tokens_seen": 19180112,
      "step": 29610
    },
    {
      "epoch": 17.461674528301888,
      "grad_norm": 3.670304536819458,
      "learning_rate": 4.82917518547682e-07,
      "loss": 0.2783,
      "num_input_tokens_seen": 19183536,
      "step": 29615
    },
    {
      "epoch": 17.464622641509433,
      "grad_norm": 3.759061813354492,
      "learning_rate": 4.81815029162756e-07,
      "loss": 0.3217,
      "num_input_tokens_seen": 19186832,
      "step": 29620
    },
    {
      "epoch": 17.46757075471698,
      "grad_norm": 2.954901695251465,
      "learning_rate": 4.807137359867725e-07,
      "loss": 0.2694,
      "num_input_tokens_seen": 19190288,
      "step": 29625
    },
    {
      "epoch": 17.47051886792453,
      "grad_norm": 3.1760072708129883,
      "learning_rate": 4.79613639311306e-07,
      "loss": 0.2957,
      "num_input_tokens_seen": 19193488,
      "step": 29630
    },
    {
      "epoch": 17.473466981132077,
      "grad_norm": 3.1998422145843506,
      "learning_rate": 4.785147394276096e-07,
      "loss": 0.3571,
      "num_input_tokens_seen": 19197008,
      "step": 29635
    },
    {
      "epoch": 17.47641509433962,
      "grad_norm": 1.8927911520004272,
      "learning_rate": 4.774170366266223e-07,
      "loss": 0.2284,
      "num_input_tokens_seen": 19200272,
      "step": 29640
    },
    {
      "epoch": 17.47936320754717,
      "grad_norm": 4.129077911376953,
      "learning_rate": 4.763205311989666e-07,
      "loss": 0.2682,
      "num_input_tokens_seen": 19204208,
      "step": 29645
    },
    {
      "epoch": 17.482311320754718,
      "grad_norm": 3.0172502994537354,
      "learning_rate": 4.752252234349458e-07,
      "loss": 0.2821,
      "num_input_tokens_seen": 19207312,
      "step": 29650
    },
    {
      "epoch": 17.485259433962263,
      "grad_norm": 2.145197629928589,
      "learning_rate": 4.7413111362454634e-07,
      "loss": 0.2694,
      "num_input_tokens_seen": 19210352,
      "step": 29655
    },
    {
      "epoch": 17.48820754716981,
      "grad_norm": 2.481473445892334,
      "learning_rate": 4.7303820205744143e-07,
      "loss": 0.3028,
      "num_input_tokens_seen": 19213488,
      "step": 29660
    },
    {
      "epoch": 17.49115566037736,
      "grad_norm": 4.0389885902404785,
      "learning_rate": 4.7194648902298303e-07,
      "loss": 0.2356,
      "num_input_tokens_seen": 19215856,
      "step": 29665
    },
    {
      "epoch": 17.494103773584907,
      "grad_norm": 4.7423906326293945,
      "learning_rate": 4.7085597481020594e-07,
      "loss": 0.2526,
      "num_input_tokens_seen": 19218704,
      "step": 29670
    },
    {
      "epoch": 17.497051886792452,
      "grad_norm": 3.362785577774048,
      "learning_rate": 4.697666597078293e-07,
      "loss": 0.3155,
      "num_input_tokens_seen": 19222192,
      "step": 29675
    },
    {
      "epoch": 17.5,
      "grad_norm": 3.91206693649292,
      "learning_rate": 4.6867854400425237e-07,
      "loss": 0.2868,
      "num_input_tokens_seen": 19224976,
      "step": 29680
    },
    {
      "epoch": 17.502948113207548,
      "grad_norm": 5.70095157623291,
      "learning_rate": 4.6759162798756084e-07,
      "loss": 0.3145,
      "num_input_tokens_seen": 19227728,
      "step": 29685
    },
    {
      "epoch": 17.505896226415093,
      "grad_norm": 4.320539951324463,
      "learning_rate": 4.6650591194551895e-07,
      "loss": 0.3425,
      "num_input_tokens_seen": 19231472,
      "step": 29690
    },
    {
      "epoch": 17.50884433962264,
      "grad_norm": 4.888162136077881,
      "learning_rate": 4.654213961655757e-07,
      "loss": 0.2893,
      "num_input_tokens_seen": 19233968,
      "step": 29695
    },
    {
      "epoch": 17.51179245283019,
      "grad_norm": 5.383718967437744,
      "learning_rate": 4.6433808093486075e-07,
      "loss": 0.2646,
      "num_input_tokens_seen": 19236912,
      "step": 29700
    },
    {
      "epoch": 17.514740566037737,
      "grad_norm": 3.8427276611328125,
      "learning_rate": 4.63255966540187e-07,
      "loss": 0.2815,
      "num_input_tokens_seen": 19239504,
      "step": 29705
    },
    {
      "epoch": 17.517688679245282,
      "grad_norm": 4.325160980224609,
      "learning_rate": 4.62175053268048e-07,
      "loss": 0.4004,
      "num_input_tokens_seen": 19243760,
      "step": 29710
    },
    {
      "epoch": 17.52063679245283,
      "grad_norm": 3.827845573425293,
      "learning_rate": 4.6109534140462045e-07,
      "loss": 0.3425,
      "num_input_tokens_seen": 19247536,
      "step": 29715
    },
    {
      "epoch": 17.52358490566038,
      "grad_norm": 3.490312337875366,
      "learning_rate": 4.6001683123576226e-07,
      "loss": 0.2842,
      "num_input_tokens_seen": 19251440,
      "step": 29720
    },
    {
      "epoch": 17.526533018867923,
      "grad_norm": 3.977437973022461,
      "learning_rate": 4.589395230470145e-07,
      "loss": 0.4989,
      "num_input_tokens_seen": 19255984,
      "step": 29725
    },
    {
      "epoch": 17.52948113207547,
      "grad_norm": 1.9037286043167114,
      "learning_rate": 4.578634171235996e-07,
      "loss": 0.2545,
      "num_input_tokens_seen": 19259600,
      "step": 29730
    },
    {
      "epoch": 17.53242924528302,
      "grad_norm": 3.977597236633301,
      "learning_rate": 4.567885137504202e-07,
      "loss": 0.2335,
      "num_input_tokens_seen": 19262704,
      "step": 29735
    },
    {
      "epoch": 17.535377358490567,
      "grad_norm": 3.7079803943634033,
      "learning_rate": 4.55714813212062e-07,
      "loss": 0.3047,
      "num_input_tokens_seen": 19265584,
      "step": 29740
    },
    {
      "epoch": 17.538325471698112,
      "grad_norm": 2.940687894821167,
      "learning_rate": 4.5464231579279206e-07,
      "loss": 0.2716,
      "num_input_tokens_seen": 19268176,
      "step": 29745
    },
    {
      "epoch": 17.54127358490566,
      "grad_norm": 3.804704427719116,
      "learning_rate": 4.535710217765571e-07,
      "loss": 0.4135,
      "num_input_tokens_seen": 19271664,
      "step": 29750
    },
    {
      "epoch": 17.54422169811321,
      "grad_norm": 5.733976364135742,
      "learning_rate": 4.5250093144698913e-07,
      "loss": 0.3089,
      "num_input_tokens_seen": 19274512,
      "step": 29755
    },
    {
      "epoch": 17.547169811320753,
      "grad_norm": 4.373210906982422,
      "learning_rate": 4.514320450873988e-07,
      "loss": 0.3579,
      "num_input_tokens_seen": 19277552,
      "step": 29760
    },
    {
      "epoch": 17.5501179245283,
      "grad_norm": 5.758703708648682,
      "learning_rate": 4.503643629807769e-07,
      "loss": 0.3383,
      "num_input_tokens_seen": 19280304,
      "step": 29765
    },
    {
      "epoch": 17.55306603773585,
      "grad_norm": 2.9911587238311768,
      "learning_rate": 4.4929788540979844e-07,
      "loss": 0.3035,
      "num_input_tokens_seen": 19283344,
      "step": 29770
    },
    {
      "epoch": 17.556014150943398,
      "grad_norm": 6.03977108001709,
      "learning_rate": 4.4823261265681596e-07,
      "loss": 0.3491,
      "num_input_tokens_seen": 19287568,
      "step": 29775
    },
    {
      "epoch": 17.558962264150942,
      "grad_norm": 3.912541151046753,
      "learning_rate": 4.471685450038671e-07,
      "loss": 0.2969,
      "num_input_tokens_seen": 19290000,
      "step": 29780
    },
    {
      "epoch": 17.56191037735849,
      "grad_norm": 4.028449535369873,
      "learning_rate": 4.4610568273266706e-07,
      "loss": 0.3302,
      "num_input_tokens_seen": 19293744,
      "step": 29785
    },
    {
      "epoch": 17.56485849056604,
      "grad_norm": 2.530616283416748,
      "learning_rate": 4.450440261246142e-07,
      "loss": 0.3312,
      "num_input_tokens_seen": 19297744,
      "step": 29790
    },
    {
      "epoch": 17.567806603773583,
      "grad_norm": 5.323303699493408,
      "learning_rate": 4.439835754607863e-07,
      "loss": 0.5115,
      "num_input_tokens_seen": 19300624,
      "step": 29795
    },
    {
      "epoch": 17.57075471698113,
      "grad_norm": 5.423924922943115,
      "learning_rate": 4.429243310219422e-07,
      "loss": 0.3771,
      "num_input_tokens_seen": 19302960,
      "step": 29800
    },
    {
      "epoch": 17.57370283018868,
      "grad_norm": 2.941300392150879,
      "learning_rate": 4.418662930885215e-07,
      "loss": 0.2682,
      "num_input_tokens_seen": 19306640,
      "step": 29805
    },
    {
      "epoch": 17.576650943396228,
      "grad_norm": 4.27246618270874,
      "learning_rate": 4.408094619406439e-07,
      "loss": 0.1842,
      "num_input_tokens_seen": 19309904,
      "step": 29810
    },
    {
      "epoch": 17.579599056603772,
      "grad_norm": 8.6797456741333,
      "learning_rate": 4.3975383785810954e-07,
      "loss": 0.2215,
      "num_input_tokens_seen": 19313104,
      "step": 29815
    },
    {
      "epoch": 17.58254716981132,
      "grad_norm": 3.4586760997772217,
      "learning_rate": 4.3869942112040096e-07,
      "loss": 0.2432,
      "num_input_tokens_seen": 19316176,
      "step": 29820
    },
    {
      "epoch": 17.58549528301887,
      "grad_norm": 2.593768358230591,
      "learning_rate": 4.3764621200667936e-07,
      "loss": 0.2714,
      "num_input_tokens_seen": 19319952,
      "step": 29825
    },
    {
      "epoch": 17.588443396226417,
      "grad_norm": 3.174407958984375,
      "learning_rate": 4.365942107957849e-07,
      "loss": 0.425,
      "num_input_tokens_seen": 19323184,
      "step": 29830
    },
    {
      "epoch": 17.59139150943396,
      "grad_norm": 2.122898817062378,
      "learning_rate": 4.35543417766241e-07,
      "loss": 0.2973,
      "num_input_tokens_seen": 19326416,
      "step": 29835
    },
    {
      "epoch": 17.59433962264151,
      "grad_norm": 7.462549686431885,
      "learning_rate": 4.3449383319624785e-07,
      "loss": 0.4074,
      "num_input_tokens_seen": 19329136,
      "step": 29840
    },
    {
      "epoch": 17.597287735849058,
      "grad_norm": 6.4795098304748535,
      "learning_rate": 4.3344545736368926e-07,
      "loss": 0.4355,
      "num_input_tokens_seen": 19333360,
      "step": 29845
    },
    {
      "epoch": 17.600235849056602,
      "grad_norm": 5.215278625488281,
      "learning_rate": 4.323982905461266e-07,
      "loss": 0.3608,
      "num_input_tokens_seen": 19338000,
      "step": 29850
    },
    {
      "epoch": 17.60318396226415,
      "grad_norm": 6.626387119293213,
      "learning_rate": 4.313523330208019e-07,
      "loss": 0.3923,
      "num_input_tokens_seen": 19341392,
      "step": 29855
    },
    {
      "epoch": 17.6061320754717,
      "grad_norm": 3.095278024673462,
      "learning_rate": 4.303075850646371e-07,
      "loss": 0.2875,
      "num_input_tokens_seen": 19344176,
      "step": 29860
    },
    {
      "epoch": 17.609080188679247,
      "grad_norm": 5.2225751876831055,
      "learning_rate": 4.2926404695423305e-07,
      "loss": 0.3164,
      "num_input_tokens_seen": 19347056,
      "step": 29865
    },
    {
      "epoch": 17.61202830188679,
      "grad_norm": 3.0153088569641113,
      "learning_rate": 4.282217189658705e-07,
      "loss": 0.2553,
      "num_input_tokens_seen": 19350832,
      "step": 29870
    },
    {
      "epoch": 17.61497641509434,
      "grad_norm": 6.56001615524292,
      "learning_rate": 4.27180601375512e-07,
      "loss": 0.2015,
      "num_input_tokens_seen": 19353520,
      "step": 29875
    },
    {
      "epoch": 17.617924528301888,
      "grad_norm": 3.669807195663452,
      "learning_rate": 4.2614069445879646e-07,
      "loss": 0.4007,
      "num_input_tokens_seen": 19356848,
      "step": 29880
    },
    {
      "epoch": 17.620872641509433,
      "grad_norm": 3.305898904800415,
      "learning_rate": 4.251019984910448e-07,
      "loss": 0.2349,
      "num_input_tokens_seen": 19362064,
      "step": 29885
    },
    {
      "epoch": 17.62382075471698,
      "grad_norm": 3.064575672149658,
      "learning_rate": 4.2406451374725597e-07,
      "loss": 0.413,
      "num_input_tokens_seen": 19365552,
      "step": 29890
    },
    {
      "epoch": 17.62676886792453,
      "grad_norm": 2.9625282287597656,
      "learning_rate": 4.2302824050210855e-07,
      "loss": 0.3457,
      "num_input_tokens_seen": 19369296,
      "step": 29895
    },
    {
      "epoch": 17.629716981132077,
      "grad_norm": 5.0731377601623535,
      "learning_rate": 4.2199317902995974e-07,
      "loss": 0.2933,
      "num_input_tokens_seen": 19371952,
      "step": 29900
    },
    {
      "epoch": 17.63266509433962,
      "grad_norm": 6.150473117828369,
      "learning_rate": 4.209593296048459e-07,
      "loss": 0.3626,
      "num_input_tokens_seen": 19374800,
      "step": 29905
    },
    {
      "epoch": 17.63561320754717,
      "grad_norm": 5.577224254608154,
      "learning_rate": 4.1992669250048524e-07,
      "loss": 0.2452,
      "num_input_tokens_seen": 19378192,
      "step": 29910
    },
    {
      "epoch": 17.638561320754718,
      "grad_norm": 4.872307777404785,
      "learning_rate": 4.188952679902719e-07,
      "loss": 0.3928,
      "num_input_tokens_seen": 19381136,
      "step": 29915
    },
    {
      "epoch": 17.641509433962263,
      "grad_norm": 3.9594192504882812,
      "learning_rate": 4.178650563472797e-07,
      "loss": 0.3125,
      "num_input_tokens_seen": 19384048,
      "step": 29920
    },
    {
      "epoch": 17.64445754716981,
      "grad_norm": 4.457752227783203,
      "learning_rate": 4.168360578442615e-07,
      "loss": 0.3812,
      "num_input_tokens_seen": 19386928,
      "step": 29925
    },
    {
      "epoch": 17.64740566037736,
      "grad_norm": 3.3535971641540527,
      "learning_rate": 4.1580827275365e-07,
      "loss": 0.2211,
      "num_input_tokens_seen": 19390576,
      "step": 29930
    },
    {
      "epoch": 17.650353773584907,
      "grad_norm": 2.8645901679992676,
      "learning_rate": 4.147817013475536e-07,
      "loss": 0.2919,
      "num_input_tokens_seen": 19393648,
      "step": 29935
    },
    {
      "epoch": 17.653301886792452,
      "grad_norm": 2.3596444129943848,
      "learning_rate": 4.1375634389776375e-07,
      "loss": 0.2561,
      "num_input_tokens_seen": 19397232,
      "step": 29940
    },
    {
      "epoch": 17.65625,
      "grad_norm": 3.550565719604492,
      "learning_rate": 4.127322006757478e-07,
      "loss": 0.3986,
      "num_input_tokens_seen": 19400304,
      "step": 29945
    },
    {
      "epoch": 17.659198113207548,
      "grad_norm": 3.004486322402954,
      "learning_rate": 4.1170927195265163e-07,
      "loss": 0.2948,
      "num_input_tokens_seen": 19403984,
      "step": 29950
    },
    {
      "epoch": 17.662146226415093,
      "grad_norm": 5.423156261444092,
      "learning_rate": 4.1068755799930026e-07,
      "loss": 0.3187,
      "num_input_tokens_seen": 19407344,
      "step": 29955
    },
    {
      "epoch": 17.66509433962264,
      "grad_norm": 4.951664447784424,
      "learning_rate": 4.096670590861962e-07,
      "loss": 0.2914,
      "num_input_tokens_seen": 19411248,
      "step": 29960
    },
    {
      "epoch": 17.66804245283019,
      "grad_norm": 3.806375026702881,
      "learning_rate": 4.086477754835211e-07,
      "loss": 0.2986,
      "num_input_tokens_seen": 19414608,
      "step": 29965
    },
    {
      "epoch": 17.670990566037737,
      "grad_norm": 2.9954888820648193,
      "learning_rate": 4.0762970746113517e-07,
      "loss": 0.2853,
      "num_input_tokens_seen": 19419248,
      "step": 29970
    },
    {
      "epoch": 17.673938679245282,
      "grad_norm": 3.79451322555542,
      "learning_rate": 4.0661285528857676e-07,
      "loss": 0.2688,
      "num_input_tokens_seen": 19421808,
      "step": 29975
    },
    {
      "epoch": 17.67688679245283,
      "grad_norm": 4.284478187561035,
      "learning_rate": 4.0559721923506155e-07,
      "loss": 0.3356,
      "num_input_tokens_seen": 19424752,
      "step": 29980
    },
    {
      "epoch": 17.67983490566038,
      "grad_norm": 4.036877155303955,
      "learning_rate": 4.045827995694834e-07,
      "loss": 0.2957,
      "num_input_tokens_seen": 19427632,
      "step": 29985
    },
    {
      "epoch": 17.682783018867923,
      "grad_norm": 4.595147132873535,
      "learning_rate": 4.035695965604142e-07,
      "loss": 0.3997,
      "num_input_tokens_seen": 19430480,
      "step": 29990
    },
    {
      "epoch": 17.68573113207547,
      "grad_norm": 4.039724349975586,
      "learning_rate": 4.0255761047610365e-07,
      "loss": 0.2517,
      "num_input_tokens_seen": 19432848,
      "step": 29995
    },
    {
      "epoch": 17.68867924528302,
      "grad_norm": 3.9354612827301025,
      "learning_rate": 4.0154684158447864e-07,
      "loss": 0.2548,
      "num_input_tokens_seen": 19435792,
      "step": 30000
    },
    {
      "epoch": 17.691627358490567,
      "grad_norm": 3.762646436691284,
      "learning_rate": 4.0053729015314623e-07,
      "loss": 0.3541,
      "num_input_tokens_seen": 19438736,
      "step": 30005
    },
    {
      "epoch": 17.694575471698112,
      "grad_norm": 3.4783201217651367,
      "learning_rate": 3.9952895644938926e-07,
      "loss": 0.2447,
      "num_input_tokens_seen": 19442608,
      "step": 30010
    },
    {
      "epoch": 17.69752358490566,
      "grad_norm": 4.99130392074585,
      "learning_rate": 3.985218407401681e-07,
      "loss": 0.3396,
      "num_input_tokens_seen": 19445328,
      "step": 30015
    },
    {
      "epoch": 17.70047169811321,
      "grad_norm": 4.763376712799072,
      "learning_rate": 3.975159432921205e-07,
      "loss": 0.2921,
      "num_input_tokens_seen": 19447696,
      "step": 30020
    },
    {
      "epoch": 17.703419811320753,
      "grad_norm": 2.9403908252716064,
      "learning_rate": 3.9651126437156294e-07,
      "loss": 0.2997,
      "num_input_tokens_seen": 19450992,
      "step": 30025
    },
    {
      "epoch": 17.7063679245283,
      "grad_norm": 3.839108467102051,
      "learning_rate": 3.9550780424448653e-07,
      "loss": 0.3481,
      "num_input_tokens_seen": 19453872,
      "step": 30030
    },
    {
      "epoch": 17.70931603773585,
      "grad_norm": 2.7291977405548096,
      "learning_rate": 3.9450556317656487e-07,
      "loss": 0.1546,
      "num_input_tokens_seen": 19456848,
      "step": 30035
    },
    {
      "epoch": 17.712264150943398,
      "grad_norm": 3.3704168796539307,
      "learning_rate": 3.935045414331434e-07,
      "loss": 0.332,
      "num_input_tokens_seen": 19460112,
      "step": 30040
    },
    {
      "epoch": 17.715212264150942,
      "grad_norm": 3.9344124794006348,
      "learning_rate": 3.925047392792475e-07,
      "loss": 0.3019,
      "num_input_tokens_seen": 19463568,
      "step": 30045
    },
    {
      "epoch": 17.71816037735849,
      "grad_norm": 3.56523060798645,
      "learning_rate": 3.9150615697957917e-07,
      "loss": 0.2343,
      "num_input_tokens_seen": 19467120,
      "step": 30050
    },
    {
      "epoch": 17.72110849056604,
      "grad_norm": 2.757141590118408,
      "learning_rate": 3.9050879479851753e-07,
      "loss": 0.3436,
      "num_input_tokens_seen": 19470192,
      "step": 30055
    },
    {
      "epoch": 17.724056603773583,
      "grad_norm": 3.2281548976898193,
      "learning_rate": 3.89512653000117e-07,
      "loss": 0.3121,
      "num_input_tokens_seen": 19474224,
      "step": 30060
    },
    {
      "epoch": 17.72700471698113,
      "grad_norm": 4.538227081298828,
      "learning_rate": 3.8851773184811203e-07,
      "loss": 0.3594,
      "num_input_tokens_seen": 19477424,
      "step": 30065
    },
    {
      "epoch": 17.72995283018868,
      "grad_norm": 3.7493202686309814,
      "learning_rate": 3.8752403160591255e-07,
      "loss": 0.2805,
      "num_input_tokens_seen": 19480912,
      "step": 30070
    },
    {
      "epoch": 17.732900943396228,
      "grad_norm": 2.4275574684143066,
      "learning_rate": 3.8653155253660477e-07,
      "loss": 0.3084,
      "num_input_tokens_seen": 19483760,
      "step": 30075
    },
    {
      "epoch": 17.735849056603772,
      "grad_norm": 5.493334770202637,
      "learning_rate": 3.8554029490295073e-07,
      "loss": 0.2977,
      "num_input_tokens_seen": 19486480,
      "step": 30080
    },
    {
      "epoch": 17.73879716981132,
      "grad_norm": 3.7334675788879395,
      "learning_rate": 3.8455025896739164e-07,
      "loss": 0.277,
      "num_input_tokens_seen": 19489232,
      "step": 30085
    },
    {
      "epoch": 17.74174528301887,
      "grad_norm": 3.2470173835754395,
      "learning_rate": 3.8356144499204215e-07,
      "loss": 0.2501,
      "num_input_tokens_seen": 19494448,
      "step": 30090
    },
    {
      "epoch": 17.744693396226417,
      "grad_norm": 3.6990928649902344,
      "learning_rate": 3.8257385323869576e-07,
      "loss": 0.2513,
      "num_input_tokens_seen": 19497136,
      "step": 30095
    },
    {
      "epoch": 17.74764150943396,
      "grad_norm": 1.8260635137557983,
      "learning_rate": 3.815874839688222e-07,
      "loss": 0.2853,
      "num_input_tokens_seen": 19501424,
      "step": 30100
    },
    {
      "epoch": 17.75058962264151,
      "grad_norm": 5.941934108734131,
      "learning_rate": 3.8060233744356634e-07,
      "loss": 0.4046,
      "num_input_tokens_seen": 19504688,
      "step": 30105
    },
    {
      "epoch": 17.753537735849058,
      "grad_norm": 6.903043270111084,
      "learning_rate": 3.796184139237502e-07,
      "loss": 0.3367,
      "num_input_tokens_seen": 19507440,
      "step": 30110
    },
    {
      "epoch": 17.756485849056602,
      "grad_norm": 2.659345865249634,
      "learning_rate": 3.7863571366987206e-07,
      "loss": 0.2873,
      "num_input_tokens_seen": 19510288,
      "step": 30115
    },
    {
      "epoch": 17.75943396226415,
      "grad_norm": 2.724351644515991,
      "learning_rate": 3.776542369421049e-07,
      "loss": 0.3176,
      "num_input_tokens_seen": 19514032,
      "step": 30120
    },
    {
      "epoch": 17.7623820754717,
      "grad_norm": 3.374269723892212,
      "learning_rate": 3.766739840003003e-07,
      "loss": 0.2718,
      "num_input_tokens_seen": 19518352,
      "step": 30125
    },
    {
      "epoch": 17.765330188679247,
      "grad_norm": 4.6791863441467285,
      "learning_rate": 3.756949551039835e-07,
      "loss": 0.4228,
      "num_input_tokens_seen": 19521456,
      "step": 30130
    },
    {
      "epoch": 17.76827830188679,
      "grad_norm": 5.717727184295654,
      "learning_rate": 3.7471715051235757e-07,
      "loss": 0.2736,
      "num_input_tokens_seen": 19524528,
      "step": 30135
    },
    {
      "epoch": 17.77122641509434,
      "grad_norm": 5.584340572357178,
      "learning_rate": 3.7374057048429947e-07,
      "loss": 0.4407,
      "num_input_tokens_seen": 19528240,
      "step": 30140
    },
    {
      "epoch": 17.774174528301888,
      "grad_norm": 3.2555322647094727,
      "learning_rate": 3.7276521527836396e-07,
      "loss": 0.3347,
      "num_input_tokens_seen": 19532656,
      "step": 30145
    },
    {
      "epoch": 17.777122641509433,
      "grad_norm": 3.149345874786377,
      "learning_rate": 3.717910851527784e-07,
      "loss": 0.285,
      "num_input_tokens_seen": 19536944,
      "step": 30150
    },
    {
      "epoch": 17.78007075471698,
      "grad_norm": 3.4980998039245605,
      "learning_rate": 3.708181803654498e-07,
      "loss": 0.2957,
      "num_input_tokens_seen": 19540208,
      "step": 30155
    },
    {
      "epoch": 17.78301886792453,
      "grad_norm": 3.7274231910705566,
      "learning_rate": 3.6984650117395993e-07,
      "loss": 0.3239,
      "num_input_tokens_seen": 19542800,
      "step": 30160
    },
    {
      "epoch": 17.785966981132077,
      "grad_norm": 2.3126394748687744,
      "learning_rate": 3.688760478355635e-07,
      "loss": 0.3017,
      "num_input_tokens_seen": 19545904,
      "step": 30165
    },
    {
      "epoch": 17.78891509433962,
      "grad_norm": 6.029037952423096,
      "learning_rate": 3.679068206071923e-07,
      "loss": 0.2797,
      "num_input_tokens_seen": 19548656,
      "step": 30170
    },
    {
      "epoch": 17.79186320754717,
      "grad_norm": 2.2784581184387207,
      "learning_rate": 3.669388197454532e-07,
      "loss": 0.278,
      "num_input_tokens_seen": 19551344,
      "step": 30175
    },
    {
      "epoch": 17.794811320754718,
      "grad_norm": 4.405045509338379,
      "learning_rate": 3.6597204550662956e-07,
      "loss": 0.2308,
      "num_input_tokens_seen": 19553904,
      "step": 30180
    },
    {
      "epoch": 17.797759433962263,
      "grad_norm": 3.2753841876983643,
      "learning_rate": 3.650064981466772e-07,
      "loss": 0.2923,
      "num_input_tokens_seen": 19558704,
      "step": 30185
    },
    {
      "epoch": 17.80070754716981,
      "grad_norm": 3.0369882583618164,
      "learning_rate": 3.640421779212311e-07,
      "loss": 0.3021,
      "num_input_tokens_seen": 19562640,
      "step": 30190
    },
    {
      "epoch": 17.80365566037736,
      "grad_norm": 1.8494846820831299,
      "learning_rate": 3.630790850855986e-07,
      "loss": 0.2622,
      "num_input_tokens_seen": 19565392,
      "step": 30195
    },
    {
      "epoch": 17.806603773584907,
      "grad_norm": 5.550745010375977,
      "learning_rate": 3.62117219894762e-07,
      "loss": 0.2695,
      "num_input_tokens_seen": 19568784,
      "step": 30200
    },
    {
      "epoch": 17.809551886792452,
      "grad_norm": 2.91770601272583,
      "learning_rate": 3.611565826033797e-07,
      "loss": 0.2563,
      "num_input_tokens_seen": 19572080,
      "step": 30205
    },
    {
      "epoch": 17.8125,
      "grad_norm": 3.7342138290405273,
      "learning_rate": 3.6019717346578445e-07,
      "loss": 0.3902,
      "num_input_tokens_seen": 19574960,
      "step": 30210
    },
    {
      "epoch": 17.815448113207548,
      "grad_norm": 5.581413745880127,
      "learning_rate": 3.5923899273598293e-07,
      "loss": 0.3477,
      "num_input_tokens_seen": 19578480,
      "step": 30215
    },
    {
      "epoch": 17.818396226415093,
      "grad_norm": 3.339996337890625,
      "learning_rate": 3.582820406676596e-07,
      "loss": 0.3101,
      "num_input_tokens_seen": 19581712,
      "step": 30220
    },
    {
      "epoch": 17.82134433962264,
      "grad_norm": 3.4966206550598145,
      "learning_rate": 3.5732631751417056e-07,
      "loss": 0.2506,
      "num_input_tokens_seen": 19584528,
      "step": 30225
    },
    {
      "epoch": 17.82429245283019,
      "grad_norm": 2.935507297515869,
      "learning_rate": 3.563718235285485e-07,
      "loss": 0.2596,
      "num_input_tokens_seen": 19587120,
      "step": 30230
    },
    {
      "epoch": 17.827240566037737,
      "grad_norm": 3.179380416870117,
      "learning_rate": 3.5541855896349844e-07,
      "loss": 0.3276,
      "num_input_tokens_seen": 19590480,
      "step": 30235
    },
    {
      "epoch": 17.830188679245282,
      "grad_norm": 4.291042804718018,
      "learning_rate": 3.544665240714018e-07,
      "loss": 0.3119,
      "num_input_tokens_seen": 19593968,
      "step": 30240
    },
    {
      "epoch": 17.83313679245283,
      "grad_norm": 4.425055503845215,
      "learning_rate": 3.535157191043137e-07,
      "loss": 0.3039,
      "num_input_tokens_seen": 19596592,
      "step": 30245
    },
    {
      "epoch": 17.83608490566038,
      "grad_norm": 7.329898357391357,
      "learning_rate": 3.5256614431396385e-07,
      "loss": 0.3136,
      "num_input_tokens_seen": 19600720,
      "step": 30250
    },
    {
      "epoch": 17.839033018867923,
      "grad_norm": 4.0184245109558105,
      "learning_rate": 3.516177999517578e-07,
      "loss": 0.3151,
      "num_input_tokens_seen": 19603632,
      "step": 30255
    },
    {
      "epoch": 17.84198113207547,
      "grad_norm": 15.732061386108398,
      "learning_rate": 3.50670686268772e-07,
      "loss": 0.2661,
      "num_input_tokens_seen": 19606192,
      "step": 30260
    },
    {
      "epoch": 17.84492924528302,
      "grad_norm": 2.957980155944824,
      "learning_rate": 3.497248035157602e-07,
      "loss": 0.243,
      "num_input_tokens_seen": 19609584,
      "step": 30265
    },
    {
      "epoch": 17.847877358490567,
      "grad_norm": 3.0166685581207275,
      "learning_rate": 3.4878015194314773e-07,
      "loss": 0.1907,
      "num_input_tokens_seen": 19613840,
      "step": 30270
    },
    {
      "epoch": 17.850825471698112,
      "grad_norm": 4.477793216705322,
      "learning_rate": 3.4783673180103617e-07,
      "loss": 0.4074,
      "num_input_tokens_seen": 19616656,
      "step": 30275
    },
    {
      "epoch": 17.85377358490566,
      "grad_norm": 5.837327003479004,
      "learning_rate": 3.468945433391985e-07,
      "loss": 0.3335,
      "num_input_tokens_seen": 19619536,
      "step": 30280
    },
    {
      "epoch": 17.85672169811321,
      "grad_norm": 4.309791564941406,
      "learning_rate": 3.459535868070851e-07,
      "loss": 0.2688,
      "num_input_tokens_seen": 19623088,
      "step": 30285
    },
    {
      "epoch": 17.859669811320753,
      "grad_norm": 3.7904739379882812,
      "learning_rate": 3.450138624538174e-07,
      "loss": 0.4754,
      "num_input_tokens_seen": 19625872,
      "step": 30290
    },
    {
      "epoch": 17.8626179245283,
      "grad_norm": 3.308687925338745,
      "learning_rate": 3.440753705281913e-07,
      "loss": 0.217,
      "num_input_tokens_seen": 19628688,
      "step": 30295
    },
    {
      "epoch": 17.86556603773585,
      "grad_norm": 4.8197760581970215,
      "learning_rate": 3.4313811127867693e-07,
      "loss": 0.3537,
      "num_input_tokens_seen": 19631760,
      "step": 30300
    },
    {
      "epoch": 17.868514150943398,
      "grad_norm": 3.617766857147217,
      "learning_rate": 3.4220208495341745e-07,
      "loss": 0.2606,
      "num_input_tokens_seen": 19634544,
      "step": 30305
    },
    {
      "epoch": 17.871462264150942,
      "grad_norm": 3.656496286392212,
      "learning_rate": 3.412672918002291e-07,
      "loss": 0.2909,
      "num_input_tokens_seen": 19637936,
      "step": 30310
    },
    {
      "epoch": 17.87441037735849,
      "grad_norm": 2.8414199352264404,
      "learning_rate": 3.403337320666045e-07,
      "loss": 0.355,
      "num_input_tokens_seen": 19641072,
      "step": 30315
    },
    {
      "epoch": 17.87735849056604,
      "grad_norm": 3.046764612197876,
      "learning_rate": 3.394014059997064e-07,
      "loss": 0.2173,
      "num_input_tokens_seen": 19644784,
      "step": 30320
    },
    {
      "epoch": 17.880306603773583,
      "grad_norm": 2.499335527420044,
      "learning_rate": 3.3847031384637185e-07,
      "loss": 0.2821,
      "num_input_tokens_seen": 19648272,
      "step": 30325
    },
    {
      "epoch": 17.88325471698113,
      "grad_norm": 4.795069217681885,
      "learning_rate": 3.3754045585311147e-07,
      "loss": 0.2131,
      "num_input_tokens_seen": 19650896,
      "step": 30330
    },
    {
      "epoch": 17.88620283018868,
      "grad_norm": 3.6479551792144775,
      "learning_rate": 3.366118322661094e-07,
      "loss": 0.3221,
      "num_input_tokens_seen": 19654608,
      "step": 30335
    },
    {
      "epoch": 17.889150943396228,
      "grad_norm": 4.775407314300537,
      "learning_rate": 3.3568444333122283e-07,
      "loss": 0.4047,
      "num_input_tokens_seen": 19657968,
      "step": 30340
    },
    {
      "epoch": 17.892099056603772,
      "grad_norm": 5.384330749511719,
      "learning_rate": 3.347582892939816e-07,
      "loss": 0.4542,
      "num_input_tokens_seen": 19660496,
      "step": 30345
    },
    {
      "epoch": 17.89504716981132,
      "grad_norm": 2.140209913253784,
      "learning_rate": 3.338333703995905e-07,
      "loss": 0.2654,
      "num_input_tokens_seen": 19663344,
      "step": 30350
    },
    {
      "epoch": 17.89799528301887,
      "grad_norm": 2.67392897605896,
      "learning_rate": 3.329096868929238e-07,
      "loss": 0.2257,
      "num_input_tokens_seen": 19666320,
      "step": 30355
    },
    {
      "epoch": 17.900943396226417,
      "grad_norm": 4.321944236755371,
      "learning_rate": 3.319872390185325e-07,
      "loss": 0.288,
      "num_input_tokens_seen": 19670160,
      "step": 30360
    },
    {
      "epoch": 17.90389150943396,
      "grad_norm": 3.5879414081573486,
      "learning_rate": 3.3106602702063727e-07,
      "loss": 0.385,
      "num_input_tokens_seen": 19673584,
      "step": 30365
    },
    {
      "epoch": 17.90683962264151,
      "grad_norm": 3.7093677520751953,
      "learning_rate": 3.3014605114313316e-07,
      "loss": 0.3428,
      "num_input_tokens_seen": 19676944,
      "step": 30370
    },
    {
      "epoch": 17.909787735849058,
      "grad_norm": 4.196757793426514,
      "learning_rate": 3.2922731162958744e-07,
      "loss": 0.301,
      "num_input_tokens_seen": 19680528,
      "step": 30375
    },
    {
      "epoch": 17.912735849056602,
      "grad_norm": 3.822705030441284,
      "learning_rate": 3.2830980872324114e-07,
      "loss": 0.3169,
      "num_input_tokens_seen": 19683472,
      "step": 30380
    },
    {
      "epoch": 17.91568396226415,
      "grad_norm": 5.401037216186523,
      "learning_rate": 3.2739354266700775e-07,
      "loss": 0.3087,
      "num_input_tokens_seen": 19686224,
      "step": 30385
    },
    {
      "epoch": 17.9186320754717,
      "grad_norm": 2.887467861175537,
      "learning_rate": 3.264785137034709e-07,
      "loss": 0.3109,
      "num_input_tokens_seen": 19688944,
      "step": 30390
    },
    {
      "epoch": 17.921580188679247,
      "grad_norm": 4.72654390335083,
      "learning_rate": 3.2556472207488977e-07,
      "loss": 0.318,
      "num_input_tokens_seen": 19693328,
      "step": 30395
    },
    {
      "epoch": 17.92452830188679,
      "grad_norm": 4.177557945251465,
      "learning_rate": 3.246521680231934e-07,
      "loss": 0.2405,
      "num_input_tokens_seen": 19695760,
      "step": 30400
    },
    {
      "epoch": 17.92747641509434,
      "grad_norm": 4.061869144439697,
      "learning_rate": 3.2374085178998594e-07,
      "loss": 0.3674,
      "num_input_tokens_seen": 19699152,
      "step": 30405
    },
    {
      "epoch": 17.930424528301888,
      "grad_norm": 9.083334922790527,
      "learning_rate": 3.2283077361654145e-07,
      "loss": 0.3584,
      "num_input_tokens_seen": 19702608,
      "step": 30410
    },
    {
      "epoch": 17.933372641509433,
      "grad_norm": 3.5302107334136963,
      "learning_rate": 3.2192193374380677e-07,
      "loss": 0.2407,
      "num_input_tokens_seen": 19705872,
      "step": 30415
    },
    {
      "epoch": 17.93632075471698,
      "grad_norm": 4.940862655639648,
      "learning_rate": 3.210143324124021e-07,
      "loss": 0.3335,
      "num_input_tokens_seen": 19708720,
      "step": 30420
    },
    {
      "epoch": 17.93926886792453,
      "grad_norm": 11.022581100463867,
      "learning_rate": 3.2010796986261805e-07,
      "loss": 0.4082,
      "num_input_tokens_seen": 19711056,
      "step": 30425
    },
    {
      "epoch": 17.942216981132077,
      "grad_norm": 2.991029977798462,
      "learning_rate": 3.1920284633441713e-07,
      "loss": 0.3263,
      "num_input_tokens_seen": 19714800,
      "step": 30430
    },
    {
      "epoch": 17.94516509433962,
      "grad_norm": 3.5784730911254883,
      "learning_rate": 3.1829896206743704e-07,
      "loss": 0.424,
      "num_input_tokens_seen": 19718064,
      "step": 30435
    },
    {
      "epoch": 17.94811320754717,
      "grad_norm": 4.089239120483398,
      "learning_rate": 3.173963173009825e-07,
      "loss": 0.5125,
      "num_input_tokens_seen": 19720528,
      "step": 30440
    },
    {
      "epoch": 17.951061320754718,
      "grad_norm": 3.777681827545166,
      "learning_rate": 3.164949122740352e-07,
      "loss": 0.2854,
      "num_input_tokens_seen": 19723312,
      "step": 30445
    },
    {
      "epoch": 17.954009433962263,
      "grad_norm": 3.1730504035949707,
      "learning_rate": 3.1559474722524406e-07,
      "loss": 0.3282,
      "num_input_tokens_seen": 19726672,
      "step": 30450
    },
    {
      "epoch": 17.95695754716981,
      "grad_norm": 4.127590179443359,
      "learning_rate": 3.146958223929325e-07,
      "loss": 0.2837,
      "num_input_tokens_seen": 19730096,
      "step": 30455
    },
    {
      "epoch": 17.95990566037736,
      "grad_norm": 2.665682315826416,
      "learning_rate": 3.1379813801509454e-07,
      "loss": 0.4413,
      "num_input_tokens_seen": 19736528,
      "step": 30460
    },
    {
      "epoch": 17.962853773584907,
      "grad_norm": 3.9621176719665527,
      "learning_rate": 3.1290169432939556e-07,
      "loss": 0.2618,
      "num_input_tokens_seen": 19740208,
      "step": 30465
    },
    {
      "epoch": 17.965801886792452,
      "grad_norm": 2.887362241744995,
      "learning_rate": 3.120064915731735e-07,
      "loss": 0.3585,
      "num_input_tokens_seen": 19744464,
      "step": 30470
    },
    {
      "epoch": 17.96875,
      "grad_norm": 3.3809733390808105,
      "learning_rate": 3.1111252998343723e-07,
      "loss": 0.2175,
      "num_input_tokens_seen": 19746704,
      "step": 30475
    },
    {
      "epoch": 17.971698113207548,
      "grad_norm": 14.16340160369873,
      "learning_rate": 3.102198097968662e-07,
      "loss": 0.2951,
      "num_input_tokens_seen": 19749680,
      "step": 30480
    },
    {
      "epoch": 17.974646226415093,
      "grad_norm": 3.1326987743377686,
      "learning_rate": 3.093283312498124e-07,
      "loss": 0.429,
      "num_input_tokens_seen": 19753392,
      "step": 30485
    },
    {
      "epoch": 17.97759433962264,
      "grad_norm": 13.347482681274414,
      "learning_rate": 3.084380945782989e-07,
      "loss": 0.4372,
      "num_input_tokens_seen": 19756912,
      "step": 30490
    },
    {
      "epoch": 17.98054245283019,
      "grad_norm": 4.219482421875,
      "learning_rate": 3.0754910001801866e-07,
      "loss": 0.4069,
      "num_input_tokens_seen": 19759984,
      "step": 30495
    },
    {
      "epoch": 17.983490566037737,
      "grad_norm": 2.8057289123535156,
      "learning_rate": 3.0666134780433786e-07,
      "loss": 0.3285,
      "num_input_tokens_seen": 19764400,
      "step": 30500
    },
    {
      "epoch": 17.986438679245282,
      "grad_norm": 6.187132358551025,
      "learning_rate": 3.0577483817229306e-07,
      "loss": 0.3255,
      "num_input_tokens_seen": 19767504,
      "step": 30505
    },
    {
      "epoch": 17.98938679245283,
      "grad_norm": 4.246895790100098,
      "learning_rate": 3.0488957135659023e-07,
      "loss": 0.2376,
      "num_input_tokens_seen": 19770032,
      "step": 30510
    },
    {
      "epoch": 17.99233490566038,
      "grad_norm": 4.419832229614258,
      "learning_rate": 3.040055475916087e-07,
      "loss": 0.3417,
      "num_input_tokens_seen": 19773168,
      "step": 30515
    },
    {
      "epoch": 17.995283018867923,
      "grad_norm": 4.518278121948242,
      "learning_rate": 3.0312276711139675e-07,
      "loss": 0.2992,
      "num_input_tokens_seen": 19776432,
      "step": 30520
    },
    {
      "epoch": 17.99823113207547,
      "grad_norm": 3.7937986850738525,
      "learning_rate": 3.0224123014967353e-07,
      "loss": 0.3317,
      "num_input_tokens_seen": 19779184,
      "step": 30525
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.623103141784668,
      "eval_runtime": 19.5395,
      "eval_samples_per_second": 86.799,
      "eval_steps_per_second": 21.7,
      "num_input_tokens_seen": 19780824,
      "step": 30528
    },
    {
      "epoch": 18.00117924528302,
      "grad_norm": 3.184617042541504,
      "learning_rate": 3.013609369398324e-07,
      "loss": 0.3308,
      "num_input_tokens_seen": 19782904,
      "step": 30530
    },
    {
      "epoch": 18.004127358490567,
      "grad_norm": 4.540716171264648,
      "learning_rate": 3.004818877149318e-07,
      "loss": 0.2346,
      "num_input_tokens_seen": 19785656,
      "step": 30535
    },
    {
      "epoch": 18.007075471698112,
      "grad_norm": 2.6180341243743896,
      "learning_rate": 2.9960408270770624e-07,
      "loss": 0.3715,
      "num_input_tokens_seen": 19792088,
      "step": 30540
    },
    {
      "epoch": 18.01002358490566,
      "grad_norm": 4.12465238571167,
      "learning_rate": 2.9872752215055755e-07,
      "loss": 0.3519,
      "num_input_tokens_seen": 19795832,
      "step": 30545
    },
    {
      "epoch": 18.01297169811321,
      "grad_norm": 3.7236714363098145,
      "learning_rate": 2.9785220627555844e-07,
      "loss": 0.4365,
      "num_input_tokens_seen": 19798392,
      "step": 30550
    },
    {
      "epoch": 18.015919811320753,
      "grad_norm": 2.8688035011291504,
      "learning_rate": 2.9697813531445295e-07,
      "loss": 0.2164,
      "num_input_tokens_seen": 19801560,
      "step": 30555
    },
    {
      "epoch": 18.0188679245283,
      "grad_norm": 2.3765954971313477,
      "learning_rate": 2.9610530949865433e-07,
      "loss": 0.2254,
      "num_input_tokens_seen": 19804504,
      "step": 30560
    },
    {
      "epoch": 18.02181603773585,
      "grad_norm": 4.367375373840332,
      "learning_rate": 2.952337290592483e-07,
      "loss": 0.2428,
      "num_input_tokens_seen": 19808056,
      "step": 30565
    },
    {
      "epoch": 18.024764150943398,
      "grad_norm": 3.8662877082824707,
      "learning_rate": 2.9436339422698913e-07,
      "loss": 0.2672,
      "num_input_tokens_seen": 19811736,
      "step": 30570
    },
    {
      "epoch": 18.027712264150942,
      "grad_norm": 2.129762649536133,
      "learning_rate": 2.934943052323008e-07,
      "loss": 0.3605,
      "num_input_tokens_seen": 19818616,
      "step": 30575
    },
    {
      "epoch": 18.03066037735849,
      "grad_norm": 2.872866153717041,
      "learning_rate": 2.926264623052799e-07,
      "loss": 0.3239,
      "num_input_tokens_seen": 19822168,
      "step": 30580
    },
    {
      "epoch": 18.03360849056604,
      "grad_norm": 3.270975112915039,
      "learning_rate": 2.9175986567569036e-07,
      "loss": 0.2957,
      "num_input_tokens_seen": 19825080,
      "step": 30585
    },
    {
      "epoch": 18.036556603773583,
      "grad_norm": 3.078059673309326,
      "learning_rate": 2.9089451557296755e-07,
      "loss": 0.2899,
      "num_input_tokens_seen": 19828120,
      "step": 30590
    },
    {
      "epoch": 18.03950471698113,
      "grad_norm": 3.5939903259277344,
      "learning_rate": 2.9003041222621706e-07,
      "loss": 0.2337,
      "num_input_tokens_seen": 19831192,
      "step": 30595
    },
    {
      "epoch": 18.04245283018868,
      "grad_norm": 4.125835418701172,
      "learning_rate": 2.8916755586421375e-07,
      "loss": 0.3754,
      "num_input_tokens_seen": 19833944,
      "step": 30600
    },
    {
      "epoch": 18.045400943396228,
      "grad_norm": 3.1255295276641846,
      "learning_rate": 2.883059467154031e-07,
      "loss": 0.2964,
      "num_input_tokens_seen": 19837464,
      "step": 30605
    },
    {
      "epoch": 18.048349056603772,
      "grad_norm": 2.3321003913879395,
      "learning_rate": 2.8744558500789887e-07,
      "loss": 0.3201,
      "num_input_tokens_seen": 19841496,
      "step": 30610
    },
    {
      "epoch": 18.05129716981132,
      "grad_norm": 7.473489761352539,
      "learning_rate": 2.8658647096948546e-07,
      "loss": 0.2973,
      "num_input_tokens_seen": 19845112,
      "step": 30615
    },
    {
      "epoch": 18.05424528301887,
      "grad_norm": 6.145713806152344,
      "learning_rate": 2.8572860482761813e-07,
      "loss": 0.3835,
      "num_input_tokens_seen": 19847896,
      "step": 30620
    },
    {
      "epoch": 18.057193396226417,
      "grad_norm": 2.8216655254364014,
      "learning_rate": 2.8487198680942017e-07,
      "loss": 0.2693,
      "num_input_tokens_seen": 19850232,
      "step": 30625
    },
    {
      "epoch": 18.06014150943396,
      "grad_norm": 2.735126495361328,
      "learning_rate": 2.840166171416836e-07,
      "loss": 0.3118,
      "num_input_tokens_seen": 19853464,
      "step": 30630
    },
    {
      "epoch": 18.06308962264151,
      "grad_norm": 3.2697463035583496,
      "learning_rate": 2.8316249605087386e-07,
      "loss": 0.2773,
      "num_input_tokens_seen": 19856568,
      "step": 30635
    },
    {
      "epoch": 18.066037735849058,
      "grad_norm": 6.000901222229004,
      "learning_rate": 2.823096237631212e-07,
      "loss": 0.3327,
      "num_input_tokens_seen": 19859352,
      "step": 30640
    },
    {
      "epoch": 18.068985849056602,
      "grad_norm": 4.513558864593506,
      "learning_rate": 2.814580005042283e-07,
      "loss": 0.3141,
      "num_input_tokens_seen": 19862776,
      "step": 30645
    },
    {
      "epoch": 18.07193396226415,
      "grad_norm": 3.6687188148498535,
      "learning_rate": 2.8060762649966435e-07,
      "loss": 0.2816,
      "num_input_tokens_seen": 19865464,
      "step": 30650
    },
    {
      "epoch": 18.0748820754717,
      "grad_norm": 5.941196441650391,
      "learning_rate": 2.797585019745713e-07,
      "loss": 0.3493,
      "num_input_tokens_seen": 19869144,
      "step": 30655
    },
    {
      "epoch": 18.077830188679247,
      "grad_norm": 3.6285245418548584,
      "learning_rate": 2.789106271537584e-07,
      "loss": 0.3182,
      "num_input_tokens_seen": 19872216,
      "step": 30660
    },
    {
      "epoch": 18.08077830188679,
      "grad_norm": 3.287102699279785,
      "learning_rate": 2.780640022617037e-07,
      "loss": 0.253,
      "num_input_tokens_seen": 19875384,
      "step": 30665
    },
    {
      "epoch": 18.08372641509434,
      "grad_norm": 5.913760185241699,
      "learning_rate": 2.772186275225547e-07,
      "loss": 0.2316,
      "num_input_tokens_seen": 19877976,
      "step": 30670
    },
    {
      "epoch": 18.086674528301888,
      "grad_norm": 7.306692123413086,
      "learning_rate": 2.7637450316012836e-07,
      "loss": 0.2856,
      "num_input_tokens_seen": 19880504,
      "step": 30675
    },
    {
      "epoch": 18.089622641509433,
      "grad_norm": 2.528595447540283,
      "learning_rate": 2.755316293979088e-07,
      "loss": 0.2733,
      "num_input_tokens_seen": 19884056,
      "step": 30680
    },
    {
      "epoch": 18.09257075471698,
      "grad_norm": 4.532585620880127,
      "learning_rate": 2.7469000645905295e-07,
      "loss": 0.2351,
      "num_input_tokens_seen": 19887192,
      "step": 30685
    },
    {
      "epoch": 18.09551886792453,
      "grad_norm": 3.4149014949798584,
      "learning_rate": 2.738496345663827e-07,
      "loss": 0.349,
      "num_input_tokens_seen": 19889880,
      "step": 30690
    },
    {
      "epoch": 18.098466981132077,
      "grad_norm": 3.3373048305511475,
      "learning_rate": 2.7301051394239e-07,
      "loss": 0.3506,
      "num_input_tokens_seen": 19893400,
      "step": 30695
    },
    {
      "epoch": 18.10141509433962,
      "grad_norm": 5.236240863800049,
      "learning_rate": 2.72172644809236e-07,
      "loss": 0.2714,
      "num_input_tokens_seen": 19896024,
      "step": 30700
    },
    {
      "epoch": 18.10436320754717,
      "grad_norm": 5.219109535217285,
      "learning_rate": 2.7133602738875e-07,
      "loss": 0.2979,
      "num_input_tokens_seen": 19898808,
      "step": 30705
    },
    {
      "epoch": 18.107311320754718,
      "grad_norm": 4.668962001800537,
      "learning_rate": 2.7050066190242976e-07,
      "loss": 0.3116,
      "num_input_tokens_seen": 19902520,
      "step": 30710
    },
    {
      "epoch": 18.110259433962263,
      "grad_norm": 5.248502254486084,
      "learning_rate": 2.696665485714428e-07,
      "loss": 0.3081,
      "num_input_tokens_seen": 19906104,
      "step": 30715
    },
    {
      "epoch": 18.11320754716981,
      "grad_norm": 3.574465036392212,
      "learning_rate": 2.6883368761662367e-07,
      "loss": 0.3011,
      "num_input_tokens_seen": 19909688,
      "step": 30720
    },
    {
      "epoch": 18.11615566037736,
      "grad_norm": 4.4044389724731445,
      "learning_rate": 2.680020792584759e-07,
      "loss": 0.2883,
      "num_input_tokens_seen": 19913496,
      "step": 30725
    },
    {
      "epoch": 18.119103773584907,
      "grad_norm": 2.897365093231201,
      "learning_rate": 2.6717172371717113e-07,
      "loss": 0.3052,
      "num_input_tokens_seen": 19916376,
      "step": 30730
    },
    {
      "epoch": 18.122051886792452,
      "grad_norm": 4.1605939865112305,
      "learning_rate": 2.663426212125503e-07,
      "loss": 0.2889,
      "num_input_tokens_seen": 19919288,
      "step": 30735
    },
    {
      "epoch": 18.125,
      "grad_norm": 2.3207688331604004,
      "learning_rate": 2.655147719641216e-07,
      "loss": 0.2688,
      "num_input_tokens_seen": 19922392,
      "step": 30740
    },
    {
      "epoch": 18.127948113207548,
      "grad_norm": 2.5117039680480957,
      "learning_rate": 2.646881761910602e-07,
      "loss": 0.3041,
      "num_input_tokens_seen": 19925688,
      "step": 30745
    },
    {
      "epoch": 18.130896226415093,
      "grad_norm": 4.663676738739014,
      "learning_rate": 2.638628341122135e-07,
      "loss": 0.1987,
      "num_input_tokens_seen": 19928504,
      "step": 30750
    },
    {
      "epoch": 18.13384433962264,
      "grad_norm": 3.4758388996124268,
      "learning_rate": 2.6303874594609314e-07,
      "loss": 0.3286,
      "num_input_tokens_seen": 19931320,
      "step": 30755
    },
    {
      "epoch": 18.13679245283019,
      "grad_norm": 5.981193542480469,
      "learning_rate": 2.622159119108797e-07,
      "loss": 0.3351,
      "num_input_tokens_seen": 19935032,
      "step": 30760
    },
    {
      "epoch": 18.139740566037737,
      "grad_norm": 2.7124521732330322,
      "learning_rate": 2.6139433222442226e-07,
      "loss": 0.3301,
      "num_input_tokens_seen": 19939064,
      "step": 30765
    },
    {
      "epoch": 18.142688679245282,
      "grad_norm": 5.188353061676025,
      "learning_rate": 2.6057400710423787e-07,
      "loss": 0.3796,
      "num_input_tokens_seen": 19943192,
      "step": 30770
    },
    {
      "epoch": 18.14563679245283,
      "grad_norm": 5.410067081451416,
      "learning_rate": 2.5975493676751004e-07,
      "loss": 0.2163,
      "num_input_tokens_seen": 19947096,
      "step": 30775
    },
    {
      "epoch": 18.14858490566038,
      "grad_norm": 3.957061290740967,
      "learning_rate": 2.589371214310926e-07,
      "loss": 0.3186,
      "num_input_tokens_seen": 19949848,
      "step": 30780
    },
    {
      "epoch": 18.151533018867923,
      "grad_norm": 2.3861238956451416,
      "learning_rate": 2.581205613115051e-07,
      "loss": 0.3021,
      "num_input_tokens_seen": 19953592,
      "step": 30785
    },
    {
      "epoch": 18.15448113207547,
      "grad_norm": 4.019924163818359,
      "learning_rate": 2.573052566249357e-07,
      "loss": 0.214,
      "num_input_tokens_seen": 19956696,
      "step": 30790
    },
    {
      "epoch": 18.15742924528302,
      "grad_norm": 4.2969770431518555,
      "learning_rate": 2.5649120758723945e-07,
      "loss": 0.2758,
      "num_input_tokens_seen": 19959160,
      "step": 30795
    },
    {
      "epoch": 18.160377358490567,
      "grad_norm": 3.366968870162964,
      "learning_rate": 2.5567841441393906e-07,
      "loss": 0.245,
      "num_input_tokens_seen": 19961624,
      "step": 30800
    },
    {
      "epoch": 18.163325471698112,
      "grad_norm": 4.100030422210693,
      "learning_rate": 2.548668773202245e-07,
      "loss": 0.3915,
      "num_input_tokens_seen": 19964312,
      "step": 30805
    },
    {
      "epoch": 18.16627358490566,
      "grad_norm": 2.510707378387451,
      "learning_rate": 2.5405659652095573e-07,
      "loss": 0.2748,
      "num_input_tokens_seen": 19968216,
      "step": 30810
    },
    {
      "epoch": 18.16922169811321,
      "grad_norm": 5.284879207611084,
      "learning_rate": 2.5324757223065655e-07,
      "loss": 0.3482,
      "num_input_tokens_seen": 19971928,
      "step": 30815
    },
    {
      "epoch": 18.172169811320753,
      "grad_norm": 3.1317989826202393,
      "learning_rate": 2.524398046635207e-07,
      "loss": 0.3633,
      "num_input_tokens_seen": 19975224,
      "step": 30820
    },
    {
      "epoch": 18.1751179245283,
      "grad_norm": 4.010813236236572,
      "learning_rate": 2.51633294033406e-07,
      "loss": 0.3414,
      "num_input_tokens_seen": 19978392,
      "step": 30825
    },
    {
      "epoch": 18.17806603773585,
      "grad_norm": 2.875767230987549,
      "learning_rate": 2.5082804055384214e-07,
      "loss": 0.3579,
      "num_input_tokens_seen": 19981784,
      "step": 30830
    },
    {
      "epoch": 18.181014150943398,
      "grad_norm": 3.387003183364868,
      "learning_rate": 2.50024044438022e-07,
      "loss": 0.4785,
      "num_input_tokens_seen": 19986296,
      "step": 30835
    },
    {
      "epoch": 18.183962264150942,
      "grad_norm": 2.77329421043396,
      "learning_rate": 2.492213058988069e-07,
      "loss": 0.2658,
      "num_input_tokens_seen": 19990168,
      "step": 30840
    },
    {
      "epoch": 18.18691037735849,
      "grad_norm": 2.4165844917297363,
      "learning_rate": 2.4841982514872633e-07,
      "loss": 0.3194,
      "num_input_tokens_seen": 19993368,
      "step": 30845
    },
    {
      "epoch": 18.18985849056604,
      "grad_norm": 5.7375712394714355,
      "learning_rate": 2.4761960239997497e-07,
      "loss": 0.2693,
      "num_input_tokens_seen": 19996536,
      "step": 30850
    },
    {
      "epoch": 18.192806603773583,
      "grad_norm": 3.7313685417175293,
      "learning_rate": 2.4682063786441556e-07,
      "loss": 0.3342,
      "num_input_tokens_seen": 19999384,
      "step": 30855
    },
    {
      "epoch": 18.19575471698113,
      "grad_norm": 3.74692964553833,
      "learning_rate": 2.460229317535778e-07,
      "loss": 0.3718,
      "num_input_tokens_seen": 20003544,
      "step": 30860
    },
    {
      "epoch": 18.19870283018868,
      "grad_norm": 3.850984811782837,
      "learning_rate": 2.4522648427865725e-07,
      "loss": 0.3533,
      "num_input_tokens_seen": 20007160,
      "step": 30865
    },
    {
      "epoch": 18.201650943396228,
      "grad_norm": 3.4327385425567627,
      "learning_rate": 2.444312956505163e-07,
      "loss": 0.307,
      "num_input_tokens_seen": 20009944,
      "step": 30870
    },
    {
      "epoch": 18.204599056603772,
      "grad_norm": 4.595302581787109,
      "learning_rate": 2.4363736607968537e-07,
      "loss": 0.25,
      "num_input_tokens_seen": 20012696,
      "step": 30875
    },
    {
      "epoch": 18.20754716981132,
      "grad_norm": 5.261036396026611,
      "learning_rate": 2.428446957763608e-07,
      "loss": 0.3475,
      "num_input_tokens_seen": 20016440,
      "step": 30880
    },
    {
      "epoch": 18.21049528301887,
      "grad_norm": 3.107084035873413,
      "learning_rate": 2.4205328495040535e-07,
      "loss": 0.2938,
      "num_input_tokens_seen": 20019576,
      "step": 30885
    },
    {
      "epoch": 18.213443396226417,
      "grad_norm": 2.7901835441589355,
      "learning_rate": 2.412631338113486e-07,
      "loss": 0.2229,
      "num_input_tokens_seen": 20022552,
      "step": 30890
    },
    {
      "epoch": 18.21639150943396,
      "grad_norm": 3.441807746887207,
      "learning_rate": 2.404742425683848e-07,
      "loss": 0.2064,
      "num_input_tokens_seen": 20025016,
      "step": 30895
    },
    {
      "epoch": 18.21933962264151,
      "grad_norm": 3.6470301151275635,
      "learning_rate": 2.3968661143037864e-07,
      "loss": 0.2566,
      "num_input_tokens_seen": 20028696,
      "step": 30900
    },
    {
      "epoch": 18.222287735849058,
      "grad_norm": 2.9421472549438477,
      "learning_rate": 2.3890024060585823e-07,
      "loss": 0.2264,
      "num_input_tokens_seen": 20031416,
      "step": 30905
    },
    {
      "epoch": 18.225235849056602,
      "grad_norm": 3.8332433700561523,
      "learning_rate": 2.3811513030301826e-07,
      "loss": 0.2673,
      "num_input_tokens_seen": 20034168,
      "step": 30910
    },
    {
      "epoch": 18.22818396226415,
      "grad_norm": 2.7780628204345703,
      "learning_rate": 2.373312807297201e-07,
      "loss": 0.2981,
      "num_input_tokens_seen": 20036888,
      "step": 30915
    },
    {
      "epoch": 18.2311320754717,
      "grad_norm": 7.128341197967529,
      "learning_rate": 2.3654869209349007e-07,
      "loss": 0.2583,
      "num_input_tokens_seen": 20040472,
      "step": 30920
    },
    {
      "epoch": 18.234080188679247,
      "grad_norm": 3.1946003437042236,
      "learning_rate": 2.357673646015246e-07,
      "loss": 0.2356,
      "num_input_tokens_seen": 20043416,
      "step": 30925
    },
    {
      "epoch": 18.23702830188679,
      "grad_norm": 5.245927333831787,
      "learning_rate": 2.3498729846068103e-07,
      "loss": 0.326,
      "num_input_tokens_seen": 20046584,
      "step": 30930
    },
    {
      "epoch": 18.23997641509434,
      "grad_norm": 2.7990548610687256,
      "learning_rate": 2.342084938774869e-07,
      "loss": 0.2322,
      "num_input_tokens_seen": 20049688,
      "step": 30935
    },
    {
      "epoch": 18.242924528301888,
      "grad_norm": 3.221703290939331,
      "learning_rate": 2.334309510581334e-07,
      "loss": 0.2296,
      "num_input_tokens_seen": 20053144,
      "step": 30940
    },
    {
      "epoch": 18.245872641509433,
      "grad_norm": 3.2126893997192383,
      "learning_rate": 2.3265467020847864e-07,
      "loss": 0.3167,
      "num_input_tokens_seen": 20056696,
      "step": 30945
    },
    {
      "epoch": 18.24882075471698,
      "grad_norm": 4.244466781616211,
      "learning_rate": 2.31879651534046e-07,
      "loss": 0.3952,
      "num_input_tokens_seen": 20059352,
      "step": 30950
    },
    {
      "epoch": 18.25176886792453,
      "grad_norm": 4.1433916091918945,
      "learning_rate": 2.311058952400247e-07,
      "loss": 0.2682,
      "num_input_tokens_seen": 20062552,
      "step": 30955
    },
    {
      "epoch": 18.254716981132077,
      "grad_norm": 5.263598442077637,
      "learning_rate": 2.3033340153127026e-07,
      "loss": 0.3807,
      "num_input_tokens_seen": 20065624,
      "step": 30960
    },
    {
      "epoch": 18.25766509433962,
      "grad_norm": 3.2747700214385986,
      "learning_rate": 2.295621706123041e-07,
      "loss": 0.3361,
      "num_input_tokens_seen": 20069112,
      "step": 30965
    },
    {
      "epoch": 18.26061320754717,
      "grad_norm": 3.6552610397338867,
      "learning_rate": 2.287922026873135e-07,
      "loss": 0.2502,
      "num_input_tokens_seen": 20072728,
      "step": 30970
    },
    {
      "epoch": 18.263561320754718,
      "grad_norm": 5.996986389160156,
      "learning_rate": 2.2802349796014923e-07,
      "loss": 0.2792,
      "num_input_tokens_seen": 20076152,
      "step": 30975
    },
    {
      "epoch": 18.266509433962263,
      "grad_norm": 4.935770034790039,
      "learning_rate": 2.2725605663433013e-07,
      "loss": 0.3415,
      "num_input_tokens_seen": 20078584,
      "step": 30980
    },
    {
      "epoch": 18.26945754716981,
      "grad_norm": 4.7337965965271,
      "learning_rate": 2.264898789130393e-07,
      "loss": 0.2432,
      "num_input_tokens_seen": 20082072,
      "step": 30985
    },
    {
      "epoch": 18.27240566037736,
      "grad_norm": 3.8048934936523438,
      "learning_rate": 2.2572496499912554e-07,
      "loss": 0.2894,
      "num_input_tokens_seen": 20084696,
      "step": 30990
    },
    {
      "epoch": 18.275353773584907,
      "grad_norm": 3.095332145690918,
      "learning_rate": 2.2496131509510354e-07,
      "loss": 0.3861,
      "num_input_tokens_seen": 20088472,
      "step": 30995
    },
    {
      "epoch": 18.278301886792452,
      "grad_norm": 3.4544804096221924,
      "learning_rate": 2.2419892940315268e-07,
      "loss": 0.2948,
      "num_input_tokens_seen": 20091224,
      "step": 31000
    },
    {
      "epoch": 18.28125,
      "grad_norm": 3.085813045501709,
      "learning_rate": 2.2343780812511819e-07,
      "loss": 0.3132,
      "num_input_tokens_seen": 20093976,
      "step": 31005
    },
    {
      "epoch": 18.284198113207548,
      "grad_norm": 3.1549205780029297,
      "learning_rate": 2.2267795146250936e-07,
      "loss": 0.2543,
      "num_input_tokens_seen": 20097112,
      "step": 31010
    },
    {
      "epoch": 18.287146226415093,
      "grad_norm": 2.3207778930664062,
      "learning_rate": 2.2191935961650146e-07,
      "loss": 0.3308,
      "num_input_tokens_seen": 20100184,
      "step": 31015
    },
    {
      "epoch": 18.29009433962264,
      "grad_norm": 2.418443441390991,
      "learning_rate": 2.2116203278793603e-07,
      "loss": 0.3929,
      "num_input_tokens_seen": 20103960,
      "step": 31020
    },
    {
      "epoch": 18.29304245283019,
      "grad_norm": 2.8908724784851074,
      "learning_rate": 2.2040597117731766e-07,
      "loss": 0.2586,
      "num_input_tokens_seen": 20106552,
      "step": 31025
    },
    {
      "epoch": 18.295990566037737,
      "grad_norm": 4.333861351013184,
      "learning_rate": 2.1965117498481793e-07,
      "loss": 0.3949,
      "num_input_tokens_seen": 20109592,
      "step": 31030
    },
    {
      "epoch": 18.298938679245282,
      "grad_norm": 2.9075803756713867,
      "learning_rate": 2.188976444102714e-07,
      "loss": 0.3204,
      "num_input_tokens_seen": 20112600,
      "step": 31035
    },
    {
      "epoch": 18.30188679245283,
      "grad_norm": 4.1141886711120605,
      "learning_rate": 2.181453796531796e-07,
      "loss": 0.6062,
      "num_input_tokens_seen": 20114936,
      "step": 31040
    },
    {
      "epoch": 18.30483490566038,
      "grad_norm": 2.400723457336426,
      "learning_rate": 2.1739438091270658e-07,
      "loss": 0.3742,
      "num_input_tokens_seen": 20118648,
      "step": 31045
    },
    {
      "epoch": 18.307783018867923,
      "grad_norm": 2.0674586296081543,
      "learning_rate": 2.1664464838768329e-07,
      "loss": 0.2729,
      "num_input_tokens_seen": 20122936,
      "step": 31050
    },
    {
      "epoch": 18.31073113207547,
      "grad_norm": 5.481173992156982,
      "learning_rate": 2.1589618227660426e-07,
      "loss": 0.4851,
      "num_input_tokens_seen": 20126904,
      "step": 31055
    },
    {
      "epoch": 18.31367924528302,
      "grad_norm": 2.3838882446289062,
      "learning_rate": 2.151489827776293e-07,
      "loss": 0.4538,
      "num_input_tokens_seen": 20130392,
      "step": 31060
    },
    {
      "epoch": 18.316627358490567,
      "grad_norm": 3.4260311126708984,
      "learning_rate": 2.1440305008858298e-07,
      "loss": 0.2942,
      "num_input_tokens_seen": 20134360,
      "step": 31065
    },
    {
      "epoch": 18.319575471698112,
      "grad_norm": 2.953777551651001,
      "learning_rate": 2.1365838440695397e-07,
      "loss": 0.257,
      "num_input_tokens_seen": 20137848,
      "step": 31070
    },
    {
      "epoch": 18.32252358490566,
      "grad_norm": 2.988136053085327,
      "learning_rate": 2.129149859298957e-07,
      "loss": 0.3086,
      "num_input_tokens_seen": 20141048,
      "step": 31075
    },
    {
      "epoch": 18.32547169811321,
      "grad_norm": 2.679727077484131,
      "learning_rate": 2.1217285485422622e-07,
      "loss": 0.3238,
      "num_input_tokens_seen": 20144920,
      "step": 31080
    },
    {
      "epoch": 18.328419811320753,
      "grad_norm": 4.334149360656738,
      "learning_rate": 2.114319913764268e-07,
      "loss": 0.3099,
      "num_input_tokens_seen": 20148312,
      "step": 31085
    },
    {
      "epoch": 18.3313679245283,
      "grad_norm": 4.904434680938721,
      "learning_rate": 2.10692395692646e-07,
      "loss": 0.3611,
      "num_input_tokens_seen": 20151448,
      "step": 31090
    },
    {
      "epoch": 18.33431603773585,
      "grad_norm": 9.418401718139648,
      "learning_rate": 2.0995406799869444e-07,
      "loss": 0.3297,
      "num_input_tokens_seen": 20153944,
      "step": 31095
    },
    {
      "epoch": 18.337264150943398,
      "grad_norm": 12.985793113708496,
      "learning_rate": 2.0921700849004743e-07,
      "loss": 0.3667,
      "num_input_tokens_seen": 20156312,
      "step": 31100
    },
    {
      "epoch": 18.340212264150942,
      "grad_norm": 4.119394779205322,
      "learning_rate": 2.084812173618439e-07,
      "loss": 0.334,
      "num_input_tokens_seen": 20159384,
      "step": 31105
    },
    {
      "epoch": 18.34316037735849,
      "grad_norm": 4.049107074737549,
      "learning_rate": 2.0774669480888853e-07,
      "loss": 0.3684,
      "num_input_tokens_seen": 20162392,
      "step": 31110
    },
    {
      "epoch": 18.34610849056604,
      "grad_norm": 4.0088419914245605,
      "learning_rate": 2.0701344102564912e-07,
      "loss": 0.2591,
      "num_input_tokens_seen": 20166456,
      "step": 31115
    },
    {
      "epoch": 18.349056603773583,
      "grad_norm": 4.253391265869141,
      "learning_rate": 2.062814562062576e-07,
      "loss": 0.2627,
      "num_input_tokens_seen": 20171320,
      "step": 31120
    },
    {
      "epoch": 18.35200471698113,
      "grad_norm": 5.777119159698486,
      "learning_rate": 2.0555074054451063e-07,
      "loss": 0.2795,
      "num_input_tokens_seen": 20173752,
      "step": 31125
    },
    {
      "epoch": 18.35495283018868,
      "grad_norm": 4.085989952087402,
      "learning_rate": 2.0482129423386843e-07,
      "loss": 0.3218,
      "num_input_tokens_seen": 20176120,
      "step": 31130
    },
    {
      "epoch": 18.357900943396228,
      "grad_norm": 5.244260311126709,
      "learning_rate": 2.040931174674543e-07,
      "loss": 0.4315,
      "num_input_tokens_seen": 20178808,
      "step": 31135
    },
    {
      "epoch": 18.360849056603772,
      "grad_norm": 3.8625576496124268,
      "learning_rate": 2.0336621043805682e-07,
      "loss": 0.3611,
      "num_input_tokens_seen": 20183064,
      "step": 31140
    },
    {
      "epoch": 18.36379716981132,
      "grad_norm": 4.201781749725342,
      "learning_rate": 2.0264057333812704e-07,
      "loss": 0.2477,
      "num_input_tokens_seen": 20187384,
      "step": 31145
    },
    {
      "epoch": 18.36674528301887,
      "grad_norm": 4.399406433105469,
      "learning_rate": 2.0191620635978127e-07,
      "loss": 0.3527,
      "num_input_tokens_seen": 20190232,
      "step": 31150
    },
    {
      "epoch": 18.369693396226417,
      "grad_norm": 4.4104814529418945,
      "learning_rate": 2.0119310969479833e-07,
      "loss": 0.4037,
      "num_input_tokens_seen": 20192632,
      "step": 31155
    },
    {
      "epoch": 18.37264150943396,
      "grad_norm": 3.0440118312835693,
      "learning_rate": 2.004712835346212e-07,
      "loss": 0.3746,
      "num_input_tokens_seen": 20195320,
      "step": 31160
    },
    {
      "epoch": 18.37558962264151,
      "grad_norm": 2.5000534057617188,
      "learning_rate": 1.99750728070357e-07,
      "loss": 0.2961,
      "num_input_tokens_seen": 20198392,
      "step": 31165
    },
    {
      "epoch": 18.378537735849058,
      "grad_norm": 5.213044166564941,
      "learning_rate": 1.9903144349277536e-07,
      "loss": 0.3034,
      "num_input_tokens_seen": 20202072,
      "step": 31170
    },
    {
      "epoch": 18.381485849056602,
      "grad_norm": 3.048816442489624,
      "learning_rate": 1.983134299923095e-07,
      "loss": 0.3098,
      "num_input_tokens_seen": 20205048,
      "step": 31175
    },
    {
      "epoch": 18.38443396226415,
      "grad_norm": 3.6907620429992676,
      "learning_rate": 1.9759668775905737e-07,
      "loss": 0.3028,
      "num_input_tokens_seen": 20207992,
      "step": 31180
    },
    {
      "epoch": 18.3873820754717,
      "grad_norm": 4.674827575683594,
      "learning_rate": 1.9688121698277995e-07,
      "loss": 0.289,
      "num_input_tokens_seen": 20211160,
      "step": 31185
    },
    {
      "epoch": 18.390330188679247,
      "grad_norm": 8.751976013183594,
      "learning_rate": 1.9616701785290015e-07,
      "loss": 0.4632,
      "num_input_tokens_seen": 20213752,
      "step": 31190
    },
    {
      "epoch": 18.39327830188679,
      "grad_norm": 6.195938587188721,
      "learning_rate": 1.954540905585056e-07,
      "loss": 0.4048,
      "num_input_tokens_seen": 20216408,
      "step": 31195
    },
    {
      "epoch": 18.39622641509434,
      "grad_norm": 6.930541038513184,
      "learning_rate": 1.9474243528834757e-07,
      "loss": 0.3283,
      "num_input_tokens_seen": 20219192,
      "step": 31200
    },
    {
      "epoch": 18.399174528301888,
      "grad_norm": 6.087349891662598,
      "learning_rate": 1.9403205223083866e-07,
      "loss": 0.3405,
      "num_input_tokens_seen": 20221688,
      "step": 31205
    },
    {
      "epoch": 18.402122641509433,
      "grad_norm": 3.3476903438568115,
      "learning_rate": 1.9332294157405619e-07,
      "loss": 0.2977,
      "num_input_tokens_seen": 20224984,
      "step": 31210
    },
    {
      "epoch": 18.40507075471698,
      "grad_norm": 6.520979881286621,
      "learning_rate": 1.926151035057411e-07,
      "loss": 0.235,
      "num_input_tokens_seen": 20227832,
      "step": 31215
    },
    {
      "epoch": 18.40801886792453,
      "grad_norm": 4.185240268707275,
      "learning_rate": 1.9190853821329626e-07,
      "loss": 0.2821,
      "num_input_tokens_seen": 20230904,
      "step": 31220
    },
    {
      "epoch": 18.410966981132077,
      "grad_norm": 2.226314067840576,
      "learning_rate": 1.9120324588378757e-07,
      "loss": 0.225,
      "num_input_tokens_seen": 20233464,
      "step": 31225
    },
    {
      "epoch": 18.41391509433962,
      "grad_norm": 3.4488892555236816,
      "learning_rate": 1.9049922670394461e-07,
      "loss": 0.2833,
      "num_input_tokens_seen": 20236952,
      "step": 31230
    },
    {
      "epoch": 18.41686320754717,
      "grad_norm": 3.241908550262451,
      "learning_rate": 1.897964808601588e-07,
      "loss": 0.2386,
      "num_input_tokens_seen": 20242744,
      "step": 31235
    },
    {
      "epoch": 18.419811320754718,
      "grad_norm": 4.545054912567139,
      "learning_rate": 1.8909500853848517e-07,
      "loss": 0.3381,
      "num_input_tokens_seen": 20248088,
      "step": 31240
    },
    {
      "epoch": 18.422759433962263,
      "grad_norm": 4.605509281158447,
      "learning_rate": 1.8839480992464243e-07,
      "loss": 0.2328,
      "num_input_tokens_seen": 20250808,
      "step": 31245
    },
    {
      "epoch": 18.42570754716981,
      "grad_norm": 3.9678492546081543,
      "learning_rate": 1.8769588520401005e-07,
      "loss": 0.2528,
      "num_input_tokens_seen": 20254264,
      "step": 31250
    },
    {
      "epoch": 18.42865566037736,
      "grad_norm": 3.342923879623413,
      "learning_rate": 1.8699823456163279e-07,
      "loss": 0.2176,
      "num_input_tokens_seen": 20257080,
      "step": 31255
    },
    {
      "epoch": 18.431603773584907,
      "grad_norm": 5.343845367431641,
      "learning_rate": 1.8630185818221514e-07,
      "loss": 0.344,
      "num_input_tokens_seen": 20260792,
      "step": 31260
    },
    {
      "epoch": 18.434551886792452,
      "grad_norm": 2.7779510021209717,
      "learning_rate": 1.856067562501268e-07,
      "loss": 0.3092,
      "num_input_tokens_seen": 20264760,
      "step": 31265
    },
    {
      "epoch": 18.4375,
      "grad_norm": 6.484358787536621,
      "learning_rate": 1.8491292894939837e-07,
      "loss": 0.4198,
      "num_input_tokens_seen": 20267352,
      "step": 31270
    },
    {
      "epoch": 18.440448113207548,
      "grad_norm": 4.143075942993164,
      "learning_rate": 1.8422037646372405e-07,
      "loss": 0.3373,
      "num_input_tokens_seen": 20270232,
      "step": 31275
    },
    {
      "epoch": 18.443396226415093,
      "grad_norm": 2.8861963748931885,
      "learning_rate": 1.8352909897645989e-07,
      "loss": 0.277,
      "num_input_tokens_seen": 20273784,
      "step": 31280
    },
    {
      "epoch": 18.44634433962264,
      "grad_norm": 3.0942230224609375,
      "learning_rate": 1.8283909667062448e-07,
      "loss": 0.2493,
      "num_input_tokens_seen": 20276504,
      "step": 31285
    },
    {
      "epoch": 18.44929245283019,
      "grad_norm": 2.7102861404418945,
      "learning_rate": 1.82150369728899e-07,
      "loss": 0.3183,
      "num_input_tokens_seen": 20279384,
      "step": 31290
    },
    {
      "epoch": 18.452240566037737,
      "grad_norm": 2.9654574394226074,
      "learning_rate": 1.814629183336275e-07,
      "loss": 0.3745,
      "num_input_tokens_seen": 20283064,
      "step": 31295
    },
    {
      "epoch": 18.455188679245282,
      "grad_norm": 9.475138664245605,
      "learning_rate": 1.807767426668139e-07,
      "loss": 0.3283,
      "num_input_tokens_seen": 20286104,
      "step": 31300
    },
    {
      "epoch": 18.45813679245283,
      "grad_norm": 3.77170729637146,
      "learning_rate": 1.8009184291012783e-07,
      "loss": 0.2966,
      "num_input_tokens_seen": 20288888,
      "step": 31305
    },
    {
      "epoch": 18.46108490566038,
      "grad_norm": 4.029168128967285,
      "learning_rate": 1.7940821924489926e-07,
      "loss": 0.2846,
      "num_input_tokens_seen": 20292056,
      "step": 31310
    },
    {
      "epoch": 18.464033018867923,
      "grad_norm": 2.897690534591675,
      "learning_rate": 1.7872587185212009e-07,
      "loss": 0.2483,
      "num_input_tokens_seen": 20294520,
      "step": 31315
    },
    {
      "epoch": 18.46698113207547,
      "grad_norm": 2.8484437465667725,
      "learning_rate": 1.7804480091244524e-07,
      "loss": 0.2678,
      "num_input_tokens_seen": 20297304,
      "step": 31320
    },
    {
      "epoch": 18.46992924528302,
      "grad_norm": 4.737104415893555,
      "learning_rate": 1.7736500660619104e-07,
      "loss": 0.3744,
      "num_input_tokens_seen": 20300632,
      "step": 31325
    },
    {
      "epoch": 18.472877358490567,
      "grad_norm": 5.249640941619873,
      "learning_rate": 1.766864891133352e-07,
      "loss": 0.3132,
      "num_input_tokens_seen": 20303384,
      "step": 31330
    },
    {
      "epoch": 18.475825471698112,
      "grad_norm": 2.1939141750335693,
      "learning_rate": 1.7600924861351843e-07,
      "loss": 0.2869,
      "num_input_tokens_seen": 20307128,
      "step": 31335
    },
    {
      "epoch": 18.47877358490566,
      "grad_norm": 4.507983684539795,
      "learning_rate": 1.7533328528604398e-07,
      "loss": 0.2893,
      "num_input_tokens_seen": 20309944,
      "step": 31340
    },
    {
      "epoch": 18.48172169811321,
      "grad_norm": 3.24709153175354,
      "learning_rate": 1.746585993098754e-07,
      "loss": 0.2256,
      "num_input_tokens_seen": 20312920,
      "step": 31345
    },
    {
      "epoch": 18.484669811320753,
      "grad_norm": 2.64257550239563,
      "learning_rate": 1.7398519086363864e-07,
      "loss": 0.3599,
      "num_input_tokens_seen": 20317624,
      "step": 31350
    },
    {
      "epoch": 18.4876179245283,
      "grad_norm": 3.317401647567749,
      "learning_rate": 1.733130601256211e-07,
      "loss": 0.205,
      "num_input_tokens_seen": 20320952,
      "step": 31355
    },
    {
      "epoch": 18.49056603773585,
      "grad_norm": 2.762835741043091,
      "learning_rate": 1.7264220727377323e-07,
      "loss": 0.2851,
      "num_input_tokens_seen": 20324408,
      "step": 31360
    },
    {
      "epoch": 18.493514150943398,
      "grad_norm": 3.775970697402954,
      "learning_rate": 1.7197263248570517e-07,
      "loss": 0.4656,
      "num_input_tokens_seen": 20327384,
      "step": 31365
    },
    {
      "epoch": 18.496462264150942,
      "grad_norm": 3.8494956493377686,
      "learning_rate": 1.7130433593869124e-07,
      "loss": 0.316,
      "num_input_tokens_seen": 20329784,
      "step": 31370
    },
    {
      "epoch": 18.49941037735849,
      "grad_norm": 4.522744655609131,
      "learning_rate": 1.706373178096643e-07,
      "loss": 0.2546,
      "num_input_tokens_seen": 20333208,
      "step": 31375
    },
    {
      "epoch": 18.50235849056604,
      "grad_norm": 5.280894756317139,
      "learning_rate": 1.6997157827522092e-07,
      "loss": 0.3538,
      "num_input_tokens_seen": 20335960,
      "step": 31380
    },
    {
      "epoch": 18.505306603773583,
      "grad_norm": 4.79353141784668,
      "learning_rate": 1.6930711751161843e-07,
      "loss": 0.2162,
      "num_input_tokens_seen": 20338712,
      "step": 31385
    },
    {
      "epoch": 18.50825471698113,
      "grad_norm": 4.737797737121582,
      "learning_rate": 1.6864393569477556e-07,
      "loss": 0.3669,
      "num_input_tokens_seen": 20341784,
      "step": 31390
    },
    {
      "epoch": 18.51120283018868,
      "grad_norm": 2.7449216842651367,
      "learning_rate": 1.6798203300027295e-07,
      "loss": 0.2844,
      "num_input_tokens_seen": 20344792,
      "step": 31395
    },
    {
      "epoch": 18.514150943396228,
      "grad_norm": 3.32682204246521,
      "learning_rate": 1.6732140960335152e-07,
      "loss": 0.2506,
      "num_input_tokens_seen": 20348312,
      "step": 31400
    },
    {
      "epoch": 18.517099056603772,
      "grad_norm": 8.216902732849121,
      "learning_rate": 1.666620656789153e-07,
      "loss": 0.3118,
      "num_input_tokens_seen": 20350936,
      "step": 31405
    },
    {
      "epoch": 18.52004716981132,
      "grad_norm": 3.6136882305145264,
      "learning_rate": 1.660040014015274e-07,
      "loss": 0.322,
      "num_input_tokens_seen": 20353720,
      "step": 31410
    },
    {
      "epoch": 18.52299528301887,
      "grad_norm": 4.153421401977539,
      "learning_rate": 1.6534721694541344e-07,
      "loss": 0.2797,
      "num_input_tokens_seen": 20357400,
      "step": 31415
    },
    {
      "epoch": 18.525943396226417,
      "grad_norm": 5.12685489654541,
      "learning_rate": 1.6469171248445993e-07,
      "loss": 0.2279,
      "num_input_tokens_seen": 20361016,
      "step": 31420
    },
    {
      "epoch": 18.52889150943396,
      "grad_norm": 5.015128135681152,
      "learning_rate": 1.6403748819221464e-07,
      "loss": 0.3143,
      "num_input_tokens_seen": 20364056,
      "step": 31425
    },
    {
      "epoch": 18.53183962264151,
      "grad_norm": 3.353701114654541,
      "learning_rate": 1.6338454424188632e-07,
      "loss": 0.2241,
      "num_input_tokens_seen": 20367480,
      "step": 31430
    },
    {
      "epoch": 18.534787735849058,
      "grad_norm": 4.000748634338379,
      "learning_rate": 1.6273288080634442e-07,
      "loss": 0.3001,
      "num_input_tokens_seen": 20370552,
      "step": 31435
    },
    {
      "epoch": 18.537735849056602,
      "grad_norm": 3.221647262573242,
      "learning_rate": 1.6208249805811982e-07,
      "loss": 0.304,
      "num_input_tokens_seen": 20374328,
      "step": 31440
    },
    {
      "epoch": 18.54068396226415,
      "grad_norm": 4.314115524291992,
      "learning_rate": 1.6143339616940423e-07,
      "loss": 0.3567,
      "num_input_tokens_seen": 20377016,
      "step": 31445
    },
    {
      "epoch": 18.5436320754717,
      "grad_norm": 3.6203670501708984,
      "learning_rate": 1.6078557531205018e-07,
      "loss": 0.2937,
      "num_input_tokens_seen": 20380280,
      "step": 31450
    },
    {
      "epoch": 18.546580188679247,
      "grad_norm": 4.644618511199951,
      "learning_rate": 1.601390356575705e-07,
      "loss": 0.3253,
      "num_input_tokens_seen": 20383128,
      "step": 31455
    },
    {
      "epoch": 18.54952830188679,
      "grad_norm": 3.0155136585235596,
      "learning_rate": 1.5949377737713988e-07,
      "loss": 0.2537,
      "num_input_tokens_seen": 20386968,
      "step": 31460
    },
    {
      "epoch": 18.55247641509434,
      "grad_norm": 4.482812404632568,
      "learning_rate": 1.5884980064159338e-07,
      "loss": 0.2741,
      "num_input_tokens_seen": 20390296,
      "step": 31465
    },
    {
      "epoch": 18.555424528301888,
      "grad_norm": 3.5331993103027344,
      "learning_rate": 1.5820710562142627e-07,
      "loss": 0.3426,
      "num_input_tokens_seen": 20395704,
      "step": 31470
    },
    {
      "epoch": 18.558372641509433,
      "grad_norm": 5.156729698181152,
      "learning_rate": 1.575656924867952e-07,
      "loss": 0.2596,
      "num_input_tokens_seen": 20398456,
      "step": 31475
    },
    {
      "epoch": 18.56132075471698,
      "grad_norm": 3.357870578765869,
      "learning_rate": 1.5692556140751658e-07,
      "loss": 0.2759,
      "num_input_tokens_seen": 20401432,
      "step": 31480
    },
    {
      "epoch": 18.56426886792453,
      "grad_norm": 7.149957180023193,
      "learning_rate": 1.5628671255306706e-07,
      "loss": 0.2994,
      "num_input_tokens_seen": 20404376,
      "step": 31485
    },
    {
      "epoch": 18.567216981132077,
      "grad_norm": 5.325976371765137,
      "learning_rate": 1.556491460925863e-07,
      "loss": 0.3405,
      "num_input_tokens_seen": 20407960,
      "step": 31490
    },
    {
      "epoch": 18.57016509433962,
      "grad_norm": 4.067577362060547,
      "learning_rate": 1.550128621948721e-07,
      "loss": 0.2226,
      "num_input_tokens_seen": 20411384,
      "step": 31495
    },
    {
      "epoch": 18.57311320754717,
      "grad_norm": 2.471334218978882,
      "learning_rate": 1.5437786102838413e-07,
      "loss": 0.2998,
      "num_input_tokens_seen": 20416152,
      "step": 31500
    },
    {
      "epoch": 18.576061320754718,
      "grad_norm": 3.5936286449432373,
      "learning_rate": 1.5374414276124017e-07,
      "loss": 0.2502,
      "num_input_tokens_seen": 20419288,
      "step": 31505
    },
    {
      "epoch": 18.579009433962263,
      "grad_norm": 2.189171075820923,
      "learning_rate": 1.5311170756122095e-07,
      "loss": 0.2329,
      "num_input_tokens_seen": 20422296,
      "step": 31510
    },
    {
      "epoch": 18.58195754716981,
      "grad_norm": 4.700762748718262,
      "learning_rate": 1.5248055559576647e-07,
      "loss": 0.301,
      "num_input_tokens_seen": 20425016,
      "step": 31515
    },
    {
      "epoch": 18.58490566037736,
      "grad_norm": 3.442884922027588,
      "learning_rate": 1.5185068703197526e-07,
      "loss": 0.3006,
      "num_input_tokens_seen": 20427800,
      "step": 31520
    },
    {
      "epoch": 18.587853773584907,
      "grad_norm": 3.3105571269989014,
      "learning_rate": 1.5122210203661004e-07,
      "loss": 0.3105,
      "num_input_tokens_seen": 20431000,
      "step": 31525
    },
    {
      "epoch": 18.590801886792452,
      "grad_norm": 5.1906328201293945,
      "learning_rate": 1.505948007760899e-07,
      "loss": 0.3039,
      "num_input_tokens_seen": 20434200,
      "step": 31530
    },
    {
      "epoch": 18.59375,
      "grad_norm": 5.836415767669678,
      "learning_rate": 1.4996878341649647e-07,
      "loss": 0.2588,
      "num_input_tokens_seen": 20437272,
      "step": 31535
    },
    {
      "epoch": 18.596698113207548,
      "grad_norm": 5.2827467918396,
      "learning_rate": 1.493440501235699e-07,
      "loss": 0.304,
      "num_input_tokens_seen": 20439736,
      "step": 31540
    },
    {
      "epoch": 18.599646226415093,
      "grad_norm": 4.478475093841553,
      "learning_rate": 1.487206010627118e-07,
      "loss": 0.2821,
      "num_input_tokens_seen": 20442520,
      "step": 31545
    },
    {
      "epoch": 18.60259433962264,
      "grad_norm": 4.599395751953125,
      "learning_rate": 1.4809843639898124e-07,
      "loss": 0.3034,
      "num_input_tokens_seen": 20445912,
      "step": 31550
    },
    {
      "epoch": 18.60554245283019,
      "grad_norm": 4.022238254547119,
      "learning_rate": 1.4747755629710093e-07,
      "loss": 0.2825,
      "num_input_tokens_seen": 20448760,
      "step": 31555
    },
    {
      "epoch": 18.608490566037737,
      "grad_norm": 3.2729668617248535,
      "learning_rate": 1.4685796092145045e-07,
      "loss": 0.2637,
      "num_input_tokens_seen": 20451736,
      "step": 31560
    },
    {
      "epoch": 18.611438679245282,
      "grad_norm": 2.7591452598571777,
      "learning_rate": 1.4623965043607135e-07,
      "loss": 0.2858,
      "num_input_tokens_seen": 20455832,
      "step": 31565
    },
    {
      "epoch": 18.61438679245283,
      "grad_norm": 4.196476459503174,
      "learning_rate": 1.4562262500466273e-07,
      "loss": 0.2542,
      "num_input_tokens_seen": 20458776,
      "step": 31570
    },
    {
      "epoch": 18.61733490566038,
      "grad_norm": 4.119078159332275,
      "learning_rate": 1.4500688479058556e-07,
      "loss": 0.2881,
      "num_input_tokens_seen": 20461496,
      "step": 31575
    },
    {
      "epoch": 18.620283018867923,
      "grad_norm": 3.05969500541687,
      "learning_rate": 1.4439242995685943e-07,
      "loss": 0.2269,
      "num_input_tokens_seen": 20465016,
      "step": 31580
    },
    {
      "epoch": 18.62323113207547,
      "grad_norm": 3.0359599590301514,
      "learning_rate": 1.4377926066616364e-07,
      "loss": 0.3842,
      "num_input_tokens_seen": 20468056,
      "step": 31585
    },
    {
      "epoch": 18.62617924528302,
      "grad_norm": 3.1601438522338867,
      "learning_rate": 1.4316737708083783e-07,
      "loss": 0.2733,
      "num_input_tokens_seen": 20472056,
      "step": 31590
    },
    {
      "epoch": 18.629127358490567,
      "grad_norm": 5.68212890625,
      "learning_rate": 1.4255677936288127e-07,
      "loss": 0.3059,
      "num_input_tokens_seen": 20475064,
      "step": 31595
    },
    {
      "epoch": 18.632075471698112,
      "grad_norm": 4.525813102722168,
      "learning_rate": 1.4194746767395184e-07,
      "loss": 0.4629,
      "num_input_tokens_seen": 20478232,
      "step": 31600
    },
    {
      "epoch": 18.63502358490566,
      "grad_norm": 3.816767454147339,
      "learning_rate": 1.4133944217536722e-07,
      "loss": 0.3176,
      "num_input_tokens_seen": 20481144,
      "step": 31605
    },
    {
      "epoch": 18.63797169811321,
      "grad_norm": 3.892221212387085,
      "learning_rate": 1.4073270302810471e-07,
      "loss": 0.3172,
      "num_input_tokens_seen": 20484056,
      "step": 31610
    },
    {
      "epoch": 18.640919811320753,
      "grad_norm": 5.195976734161377,
      "learning_rate": 1.4012725039280084e-07,
      "loss": 0.375,
      "num_input_tokens_seen": 20486904,
      "step": 31615
    },
    {
      "epoch": 18.6438679245283,
      "grad_norm": 3.766498565673828,
      "learning_rate": 1.3952308442975292e-07,
      "loss": 0.239,
      "num_input_tokens_seen": 20489784,
      "step": 31620
    },
    {
      "epoch": 18.64681603773585,
      "grad_norm": 2.096390724182129,
      "learning_rate": 1.3892020529891637e-07,
      "loss": 0.2254,
      "num_input_tokens_seen": 20492568,
      "step": 31625
    },
    {
      "epoch": 18.649764150943398,
      "grad_norm": 3.213268280029297,
      "learning_rate": 1.3831861315990514e-07,
      "loss": 0.2438,
      "num_input_tokens_seen": 20496184,
      "step": 31630
    },
    {
      "epoch": 18.652712264150942,
      "grad_norm": 2.95981502532959,
      "learning_rate": 1.377183081719935e-07,
      "loss": 0.264,
      "num_input_tokens_seen": 20499512,
      "step": 31635
    },
    {
      "epoch": 18.65566037735849,
      "grad_norm": 2.9124112129211426,
      "learning_rate": 1.3711929049411544e-07,
      "loss": 0.2231,
      "num_input_tokens_seen": 20503096,
      "step": 31640
    },
    {
      "epoch": 18.65860849056604,
      "grad_norm": 4.4908833503723145,
      "learning_rate": 1.365215602848624e-07,
      "loss": 0.3041,
      "num_input_tokens_seen": 20505688,
      "step": 31645
    },
    {
      "epoch": 18.661556603773583,
      "grad_norm": 1.8145605325698853,
      "learning_rate": 1.3592511770248727e-07,
      "loss": 0.286,
      "num_input_tokens_seen": 20508984,
      "step": 31650
    },
    {
      "epoch": 18.66450471698113,
      "grad_norm": 4.197218894958496,
      "learning_rate": 1.3532996290490041e-07,
      "loss": 0.3296,
      "num_input_tokens_seen": 20511832,
      "step": 31655
    },
    {
      "epoch": 18.66745283018868,
      "grad_norm": 3.4732112884521484,
      "learning_rate": 1.347360960496713e-07,
      "loss": 0.2986,
      "num_input_tokens_seen": 20514232,
      "step": 31660
    },
    {
      "epoch": 18.670400943396228,
      "grad_norm": 3.644136667251587,
      "learning_rate": 1.3414351729402862e-07,
      "loss": 0.3189,
      "num_input_tokens_seen": 20517112,
      "step": 31665
    },
    {
      "epoch": 18.673349056603772,
      "grad_norm": 5.052508354187012,
      "learning_rate": 1.3355222679486025e-07,
      "loss": 0.4267,
      "num_input_tokens_seen": 20520888,
      "step": 31670
    },
    {
      "epoch": 18.67629716981132,
      "grad_norm": 3.0150301456451416,
      "learning_rate": 1.3296222470871367e-07,
      "loss": 0.2057,
      "num_input_tokens_seen": 20524472,
      "step": 31675
    },
    {
      "epoch": 18.67924528301887,
      "grad_norm": 2.2694106101989746,
      "learning_rate": 1.3237351119179287e-07,
      "loss": 0.2664,
      "num_input_tokens_seen": 20527320,
      "step": 31680
    },
    {
      "epoch": 18.682193396226417,
      "grad_norm": 3.3364198207855225,
      "learning_rate": 1.3178608639996425e-07,
      "loss": 0.2819,
      "num_input_tokens_seen": 20530264,
      "step": 31685
    },
    {
      "epoch": 18.68514150943396,
      "grad_norm": 7.765411853790283,
      "learning_rate": 1.3119995048874957e-07,
      "loss": 0.2909,
      "num_input_tokens_seen": 20532824,
      "step": 31690
    },
    {
      "epoch": 18.68808962264151,
      "grad_norm": 1.902416467666626,
      "learning_rate": 1.3061510361333186e-07,
      "loss": 0.3288,
      "num_input_tokens_seen": 20535928,
      "step": 31695
    },
    {
      "epoch": 18.691037735849058,
      "grad_norm": 6.458583831787109,
      "learning_rate": 1.3003154592855116e-07,
      "loss": 0.2043,
      "num_input_tokens_seen": 20538616,
      "step": 31700
    },
    {
      "epoch": 18.693985849056602,
      "grad_norm": 3.692347764968872,
      "learning_rate": 1.2944927758890668e-07,
      "loss": 0.3419,
      "num_input_tokens_seen": 20542104,
      "step": 31705
    },
    {
      "epoch": 18.69693396226415,
      "grad_norm": 2.4278314113616943,
      "learning_rate": 1.2886829874855733e-07,
      "loss": 0.2991,
      "num_input_tokens_seen": 20545016,
      "step": 31710
    },
    {
      "epoch": 18.6998820754717,
      "grad_norm": 6.540754318237305,
      "learning_rate": 1.2828860956131894e-07,
      "loss": 0.3968,
      "num_input_tokens_seen": 20547736,
      "step": 31715
    },
    {
      "epoch": 18.702830188679247,
      "grad_norm": 4.46596622467041,
      "learning_rate": 1.2771021018066765e-07,
      "loss": 0.3502,
      "num_input_tokens_seen": 20551256,
      "step": 31720
    },
    {
      "epoch": 18.70577830188679,
      "grad_norm": 4.794787406921387,
      "learning_rate": 1.271331007597365e-07,
      "loss": 0.4428,
      "num_input_tokens_seen": 20554520,
      "step": 31725
    },
    {
      "epoch": 18.70872641509434,
      "grad_norm": 6.003110885620117,
      "learning_rate": 1.2655728145131774e-07,
      "loss": 0.3849,
      "num_input_tokens_seen": 20558584,
      "step": 31730
    },
    {
      "epoch": 18.711674528301888,
      "grad_norm": 7.739224910736084,
      "learning_rate": 1.2598275240786105e-07,
      "loss": 0.3085,
      "num_input_tokens_seen": 20560952,
      "step": 31735
    },
    {
      "epoch": 18.714622641509433,
      "grad_norm": 5.926811218261719,
      "learning_rate": 1.254095137814776e-07,
      "loss": 0.3691,
      "num_input_tokens_seen": 20567320,
      "step": 31740
    },
    {
      "epoch": 18.71757075471698,
      "grad_norm": 2.365098237991333,
      "learning_rate": 1.2483756572393368e-07,
      "loss": 0.3066,
      "num_input_tokens_seen": 20570840,
      "step": 31745
    },
    {
      "epoch": 18.72051886792453,
      "grad_norm": 5.172773361206055,
      "learning_rate": 1.242669083866549e-07,
      "loss": 0.2989,
      "num_input_tokens_seen": 20573848,
      "step": 31750
    },
    {
      "epoch": 18.723466981132077,
      "grad_norm": 3.468010425567627,
      "learning_rate": 1.2369754192072537e-07,
      "loss": 0.3452,
      "num_input_tokens_seen": 20578360,
      "step": 31755
    },
    {
      "epoch": 18.72641509433962,
      "grad_norm": 3.2455124855041504,
      "learning_rate": 1.231294664768873e-07,
      "loss": 0.2578,
      "num_input_tokens_seen": 20581592,
      "step": 31760
    },
    {
      "epoch": 18.72936320754717,
      "grad_norm": 10.351655006408691,
      "learning_rate": 1.225626822055409e-07,
      "loss": 0.372,
      "num_input_tokens_seen": 20584184,
      "step": 31765
    },
    {
      "epoch": 18.732311320754718,
      "grad_norm": 3.4635097980499268,
      "learning_rate": 1.2199718925674508e-07,
      "loss": 0.3447,
      "num_input_tokens_seen": 20587736,
      "step": 31770
    },
    {
      "epoch": 18.735259433962263,
      "grad_norm": 5.234659194946289,
      "learning_rate": 1.2143298778021616e-07,
      "loss": 0.3213,
      "num_input_tokens_seen": 20590552,
      "step": 31775
    },
    {
      "epoch": 18.73820754716981,
      "grad_norm": 4.147350311279297,
      "learning_rate": 1.2087007792532967e-07,
      "loss": 0.3039,
      "num_input_tokens_seen": 20594232,
      "step": 31780
    },
    {
      "epoch": 18.74115566037736,
      "grad_norm": 4.884242534637451,
      "learning_rate": 1.203084598411175e-07,
      "loss": 0.2791,
      "num_input_tokens_seen": 20597368,
      "step": 31785
    },
    {
      "epoch": 18.744103773584907,
      "grad_norm": 4.903059959411621,
      "learning_rate": 1.1974813367627124e-07,
      "loss": 0.3614,
      "num_input_tokens_seen": 20600056,
      "step": 31790
    },
    {
      "epoch": 18.747051886792452,
      "grad_norm": 4.174407958984375,
      "learning_rate": 1.1918909957913949e-07,
      "loss": 0.3655,
      "num_input_tokens_seen": 20603224,
      "step": 31795
    },
    {
      "epoch": 18.75,
      "grad_norm": 3.872450113296509,
      "learning_rate": 1.1863135769772827e-07,
      "loss": 0.3398,
      "num_input_tokens_seen": 20606424,
      "step": 31800
    },
    {
      "epoch": 18.752948113207548,
      "grad_norm": 4.770089626312256,
      "learning_rate": 1.1807490817970279e-07,
      "loss": 0.1652,
      "num_input_tokens_seen": 20611160,
      "step": 31805
    },
    {
      "epoch": 18.755896226415093,
      "grad_norm": 3.430833101272583,
      "learning_rate": 1.1751975117238578e-07,
      "loss": 0.318,
      "num_input_tokens_seen": 20614488,
      "step": 31810
    },
    {
      "epoch": 18.75884433962264,
      "grad_norm": 9.050087928771973,
      "learning_rate": 1.1696588682275633e-07,
      "loss": 0.2361,
      "num_input_tokens_seen": 20617336,
      "step": 31815
    },
    {
      "epoch": 18.76179245283019,
      "grad_norm": 4.597966194152832,
      "learning_rate": 1.1641331527745325e-07,
      "loss": 0.3364,
      "num_input_tokens_seen": 20620728,
      "step": 31820
    },
    {
      "epoch": 18.764740566037737,
      "grad_norm": 4.583281993865967,
      "learning_rate": 1.1586203668277229e-07,
      "loss": 0.2669,
      "num_input_tokens_seen": 20623768,
      "step": 31825
    },
    {
      "epoch": 18.767688679245282,
      "grad_norm": 3.1271445751190186,
      "learning_rate": 1.1531205118466615e-07,
      "loss": 0.349,
      "num_input_tokens_seen": 20628472,
      "step": 31830
    },
    {
      "epoch": 18.77063679245283,
      "grad_norm": 3.1824536323547363,
      "learning_rate": 1.1476335892874669e-07,
      "loss": 0.2723,
      "num_input_tokens_seen": 20631672,
      "step": 31835
    },
    {
      "epoch": 18.77358490566038,
      "grad_norm": 4.0486979484558105,
      "learning_rate": 1.1421596006028157e-07,
      "loss": 0.2533,
      "num_input_tokens_seen": 20634616,
      "step": 31840
    },
    {
      "epoch": 18.776533018867923,
      "grad_norm": 2.426064968109131,
      "learning_rate": 1.1366985472419823e-07,
      "loss": 0.2852,
      "num_input_tokens_seen": 20637432,
      "step": 31845
    },
    {
      "epoch": 18.77948113207547,
      "grad_norm": 3.2279410362243652,
      "learning_rate": 1.1312504306507987e-07,
      "loss": 0.3254,
      "num_input_tokens_seen": 20640152,
      "step": 31850
    },
    {
      "epoch": 18.78242924528302,
      "grad_norm": 5.185301303863525,
      "learning_rate": 1.1258152522716725e-07,
      "loss": 0.2748,
      "num_input_tokens_seen": 20642808,
      "step": 31855
    },
    {
      "epoch": 18.785377358490567,
      "grad_norm": 6.498809814453125,
      "learning_rate": 1.1203930135435914e-07,
      "loss": 0.2677,
      "num_input_tokens_seen": 20645208,
      "step": 31860
    },
    {
      "epoch": 18.788325471698112,
      "grad_norm": 5.252915859222412,
      "learning_rate": 1.1149837159021238e-07,
      "loss": 0.1954,
      "num_input_tokens_seen": 20647800,
      "step": 31865
    },
    {
      "epoch": 18.79127358490566,
      "grad_norm": 4.484426498413086,
      "learning_rate": 1.1095873607793961e-07,
      "loss": 0.3481,
      "num_input_tokens_seen": 20650840,
      "step": 31870
    },
    {
      "epoch": 18.79422169811321,
      "grad_norm": 3.9679746627807617,
      "learning_rate": 1.1042039496041212e-07,
      "loss": 0.2875,
      "num_input_tokens_seen": 20654232,
      "step": 31875
    },
    {
      "epoch": 18.797169811320753,
      "grad_norm": 3.434692144393921,
      "learning_rate": 1.0988334838015812e-07,
      "loss": 0.2704,
      "num_input_tokens_seen": 20657816,
      "step": 31880
    },
    {
      "epoch": 18.8001179245283,
      "grad_norm": 4.225030899047852,
      "learning_rate": 1.0934759647936333e-07,
      "loss": 0.1893,
      "num_input_tokens_seen": 20660856,
      "step": 31885
    },
    {
      "epoch": 18.80306603773585,
      "grad_norm": 3.5754191875457764,
      "learning_rate": 1.0881313939986926e-07,
      "loss": 0.3198,
      "num_input_tokens_seen": 20663992,
      "step": 31890
    },
    {
      "epoch": 18.806014150943398,
      "grad_norm": 4.498053073883057,
      "learning_rate": 1.0827997728317662e-07,
      "loss": 0.2857,
      "num_input_tokens_seen": 20666744,
      "step": 31895
    },
    {
      "epoch": 18.808962264150942,
      "grad_norm": 4.979022026062012,
      "learning_rate": 1.0774811027044196e-07,
      "loss": 0.2578,
      "num_input_tokens_seen": 20669624,
      "step": 31900
    },
    {
      "epoch": 18.81191037735849,
      "grad_norm": 2.7765073776245117,
      "learning_rate": 1.0721753850247984e-07,
      "loss": 0.3426,
      "num_input_tokens_seen": 20672824,
      "step": 31905
    },
    {
      "epoch": 18.81485849056604,
      "grad_norm": 4.057370662689209,
      "learning_rate": 1.0668826211976124e-07,
      "loss": 0.2873,
      "num_input_tokens_seen": 20676152,
      "step": 31910
    },
    {
      "epoch": 18.817806603773583,
      "grad_norm": 4.492893695831299,
      "learning_rate": 1.0616028126241407e-07,
      "loss": 0.3644,
      "num_input_tokens_seen": 20679896,
      "step": 31915
    },
    {
      "epoch": 18.82075471698113,
      "grad_norm": 4.472972869873047,
      "learning_rate": 1.0563359607022372e-07,
      "loss": 0.2688,
      "num_input_tokens_seen": 20682616,
      "step": 31920
    },
    {
      "epoch": 18.82370283018868,
      "grad_norm": 2.07848858833313,
      "learning_rate": 1.05108206682632e-07,
      "loss": 0.3325,
      "num_input_tokens_seen": 20685112,
      "step": 31925
    },
    {
      "epoch": 18.826650943396228,
      "grad_norm": 5.044286727905273,
      "learning_rate": 1.0458411323873874e-07,
      "loss": 0.3053,
      "num_input_tokens_seen": 20689080,
      "step": 31930
    },
    {
      "epoch": 18.829599056603772,
      "grad_norm": 2.2466392517089844,
      "learning_rate": 1.0406131587729962e-07,
      "loss": 0.2954,
      "num_input_tokens_seen": 20692568,
      "step": 31935
    },
    {
      "epoch": 18.83254716981132,
      "grad_norm": 3.208451509475708,
      "learning_rate": 1.035398147367278e-07,
      "loss": 0.3064,
      "num_input_tokens_seen": 20696536,
      "step": 31940
    },
    {
      "epoch": 18.83549528301887,
      "grad_norm": 4.30341100692749,
      "learning_rate": 1.030196099550923e-07,
      "loss": 0.2233,
      "num_input_tokens_seen": 20699736,
      "step": 31945
    },
    {
      "epoch": 18.838443396226417,
      "grad_norm": 3.394279956817627,
      "learning_rate": 1.0250070167011905e-07,
      "loss": 0.342,
      "num_input_tokens_seen": 20703128,
      "step": 31950
    },
    {
      "epoch": 18.84139150943396,
      "grad_norm": 4.651595592498779,
      "learning_rate": 1.0198309001919315e-07,
      "loss": 0.285,
      "num_input_tokens_seen": 20706968,
      "step": 31955
    },
    {
      "epoch": 18.84433962264151,
      "grad_norm": 2.102612257003784,
      "learning_rate": 1.0146677513935277e-07,
      "loss": 0.2258,
      "num_input_tokens_seen": 20709784,
      "step": 31960
    },
    {
      "epoch": 18.847287735849058,
      "grad_norm": 4.459384441375732,
      "learning_rate": 1.0095175716729578e-07,
      "loss": 0.4118,
      "num_input_tokens_seen": 20712792,
      "step": 31965
    },
    {
      "epoch": 18.850235849056602,
      "grad_norm": 4.181687355041504,
      "learning_rate": 1.004380362393742e-07,
      "loss": 0.2885,
      "num_input_tokens_seen": 20715832,
      "step": 31970
    },
    {
      "epoch": 18.85318396226415,
      "grad_norm": 3.7447752952575684,
      "learning_rate": 9.99256124915987e-08,
      "loss": 0.2536,
      "num_input_tokens_seen": 20718744,
      "step": 31975
    },
    {
      "epoch": 18.8561320754717,
      "grad_norm": 5.714812278747559,
      "learning_rate": 9.941448605963577e-08,
      "loss": 0.3558,
      "num_input_tokens_seen": 20721528,
      "step": 31980
    },
    {
      "epoch": 18.859080188679247,
      "grad_norm": 4.963948726654053,
      "learning_rate": 9.890465707880715e-08,
      "loss": 0.3333,
      "num_input_tokens_seen": 20725240,
      "step": 31985
    },
    {
      "epoch": 18.86202830188679,
      "grad_norm": 4.254890441894531,
      "learning_rate": 9.839612568409374e-08,
      "loss": 0.3289,
      "num_input_tokens_seen": 20728312,
      "step": 31990
    },
    {
      "epoch": 18.86497641509434,
      "grad_norm": 5.041440963745117,
      "learning_rate": 9.788889201013119e-08,
      "loss": 0.2621,
      "num_input_tokens_seen": 20731512,
      "step": 31995
    },
    {
      "epoch": 18.867924528301888,
      "grad_norm": 4.277742385864258,
      "learning_rate": 9.738295619121097e-08,
      "loss": 0.3329,
      "num_input_tokens_seen": 20735384,
      "step": 32000
    },
    {
      "epoch": 18.870872641509433,
      "grad_norm": 4.624903202056885,
      "learning_rate": 9.687831836128203e-08,
      "loss": 0.4045,
      "num_input_tokens_seen": 20737944,
      "step": 32005
    },
    {
      "epoch": 18.87382075471698,
      "grad_norm": 4.394538402557373,
      "learning_rate": 9.637497865395029e-08,
      "loss": 0.1946,
      "num_input_tokens_seen": 20740888,
      "step": 32010
    },
    {
      "epoch": 18.87676886792453,
      "grad_norm": 2.29673171043396,
      "learning_rate": 9.587293720247526e-08,
      "loss": 0.2666,
      "num_input_tokens_seen": 20743544,
      "step": 32015
    },
    {
      "epoch": 18.879716981132077,
      "grad_norm": 5.029848575592041,
      "learning_rate": 9.537219413977672e-08,
      "loss": 0.3129,
      "num_input_tokens_seen": 20746840,
      "step": 32020
    },
    {
      "epoch": 18.88266509433962,
      "grad_norm": 4.0514702796936035,
      "learning_rate": 9.487274959842696e-08,
      "loss": 0.3388,
      "num_input_tokens_seen": 20749880,
      "step": 32025
    },
    {
      "epoch": 18.88561320754717,
      "grad_norm": 2.396413564682007,
      "learning_rate": 9.437460371065687e-08,
      "loss": 0.2635,
      "num_input_tokens_seen": 20754264,
      "step": 32030
    },
    {
      "epoch": 18.888561320754718,
      "grad_norm": 4.186864852905273,
      "learning_rate": 9.387775660835263e-08,
      "loss": 0.3585,
      "num_input_tokens_seen": 20756952,
      "step": 32035
    },
    {
      "epoch": 18.891509433962263,
      "grad_norm": 2.8169798851013184,
      "learning_rate": 9.338220842305678e-08,
      "loss": 0.2394,
      "num_input_tokens_seen": 20759448,
      "step": 32040
    },
    {
      "epoch": 18.89445754716981,
      "grad_norm": 3.6519625186920166,
      "learning_rate": 9.288795928596661e-08,
      "loss": 0.305,
      "num_input_tokens_seen": 20762488,
      "step": 32045
    },
    {
      "epoch": 18.89740566037736,
      "grad_norm": 2.6769678592681885,
      "learning_rate": 9.239500932793854e-08,
      "loss": 0.2851,
      "num_input_tokens_seen": 20765848,
      "step": 32050
    },
    {
      "epoch": 18.900353773584907,
      "grad_norm": 4.497045516967773,
      "learning_rate": 9.190335867948263e-08,
      "loss": 0.2333,
      "num_input_tokens_seen": 20768376,
      "step": 32055
    },
    {
      "epoch": 18.903301886792452,
      "grad_norm": 2.6710407733917236,
      "learning_rate": 9.141300747076476e-08,
      "loss": 0.2387,
      "num_input_tokens_seen": 20773048,
      "step": 32060
    },
    {
      "epoch": 18.90625,
      "grad_norm": 2.982206344604492,
      "learning_rate": 9.092395583160773e-08,
      "loss": 0.2206,
      "num_input_tokens_seen": 20775992,
      "step": 32065
    },
    {
      "epoch": 18.909198113207548,
      "grad_norm": 4.990902423858643,
      "learning_rate": 9.043620389149021e-08,
      "loss": 0.2578,
      "num_input_tokens_seen": 20779032,
      "step": 32070
    },
    {
      "epoch": 18.912146226415093,
      "grad_norm": 3.0074241161346436,
      "learning_rate": 8.994975177954723e-08,
      "loss": 0.3892,
      "num_input_tokens_seen": 20782328,
      "step": 32075
    },
    {
      "epoch": 18.91509433962264,
      "grad_norm": 3.6410322189331055,
      "learning_rate": 8.946459962456855e-08,
      "loss": 0.2596,
      "num_input_tokens_seen": 20786296,
      "step": 32080
    },
    {
      "epoch": 18.91804245283019,
      "grad_norm": 3.2987256050109863,
      "learning_rate": 8.89807475550003e-08,
      "loss": 0.2679,
      "num_input_tokens_seen": 20790392,
      "step": 32085
    },
    {
      "epoch": 18.920990566037737,
      "grad_norm": 5.421662330627441,
      "learning_rate": 8.849819569894447e-08,
      "loss": 0.321,
      "num_input_tokens_seen": 20794360,
      "step": 32090
    },
    {
      "epoch": 18.923938679245282,
      "grad_norm": 3.790262460708618,
      "learning_rate": 8.801694418415884e-08,
      "loss": 0.2945,
      "num_input_tokens_seen": 20797464,
      "step": 32095
    },
    {
      "epoch": 18.92688679245283,
      "grad_norm": 6.716434955596924,
      "learning_rate": 8.753699313805708e-08,
      "loss": 0.2692,
      "num_input_tokens_seen": 20800408,
      "step": 32100
    },
    {
      "epoch": 18.92983490566038,
      "grad_norm": 3.5119075775146484,
      "learning_rate": 8.705834268770753e-08,
      "loss": 0.3063,
      "num_input_tokens_seen": 20803896,
      "step": 32105
    },
    {
      "epoch": 18.932783018867923,
      "grad_norm": 3.251007556915283,
      "learning_rate": 8.65809929598349e-08,
      "loss": 0.3283,
      "num_input_tokens_seen": 20807768,
      "step": 32110
    },
    {
      "epoch": 18.93573113207547,
      "grad_norm": 3.339484453201294,
      "learning_rate": 8.610494408082037e-08,
      "loss": 0.3373,
      "num_input_tokens_seen": 20810776,
      "step": 32115
    },
    {
      "epoch": 18.93867924528302,
      "grad_norm": 4.647885799407959,
      "learning_rate": 8.563019617669977e-08,
      "loss": 0.3799,
      "num_input_tokens_seen": 20814744,
      "step": 32120
    },
    {
      "epoch": 18.941627358490567,
      "grad_norm": 4.79951810836792,
      "learning_rate": 8.51567493731642e-08,
      "loss": 0.1983,
      "num_input_tokens_seen": 20817784,
      "step": 32125
    },
    {
      "epoch": 18.944575471698112,
      "grad_norm": 4.023779392242432,
      "learning_rate": 8.468460379556176e-08,
      "loss": 0.3209,
      "num_input_tokens_seen": 20821432,
      "step": 32130
    },
    {
      "epoch": 18.94752358490566,
      "grad_norm": 3.8141188621520996,
      "learning_rate": 8.421375956889355e-08,
      "loss": 0.3285,
      "num_input_tokens_seen": 20824824,
      "step": 32135
    },
    {
      "epoch": 18.95047169811321,
      "grad_norm": 3.5322389602661133,
      "learning_rate": 8.374421681781819e-08,
      "loss": 0.3712,
      "num_input_tokens_seen": 20827672,
      "step": 32140
    },
    {
      "epoch": 18.953419811320753,
      "grad_norm": 3.8961331844329834,
      "learning_rate": 8.327597566665013e-08,
      "loss": 0.4096,
      "num_input_tokens_seen": 20830264,
      "step": 32145
    },
    {
      "epoch": 18.9563679245283,
      "grad_norm": 5.177209377288818,
      "learning_rate": 8.280903623935688e-08,
      "loss": 0.2771,
      "num_input_tokens_seen": 20832984,
      "step": 32150
    },
    {
      "epoch": 18.95931603773585,
      "grad_norm": 3.412698268890381,
      "learning_rate": 8.234339865956342e-08,
      "loss": 0.3902,
      "num_input_tokens_seen": 20835960,
      "step": 32155
    },
    {
      "epoch": 18.962264150943398,
      "grad_norm": 2.8963043689727783,
      "learning_rate": 8.187906305054838e-08,
      "loss": 0.3194,
      "num_input_tokens_seen": 20838904,
      "step": 32160
    },
    {
      "epoch": 18.965212264150942,
      "grad_norm": 3.9321789741516113,
      "learning_rate": 8.141602953524841e-08,
      "loss": 0.352,
      "num_input_tokens_seen": 20842168,
      "step": 32165
    },
    {
      "epoch": 18.96816037735849,
      "grad_norm": 3.6824867725372314,
      "learning_rate": 8.095429823625212e-08,
      "loss": 0.3244,
      "num_input_tokens_seen": 20845752,
      "step": 32170
    },
    {
      "epoch": 18.97110849056604,
      "grad_norm": 5.223037242889404,
      "learning_rate": 8.04938692758045e-08,
      "loss": 0.3279,
      "num_input_tokens_seen": 20850136,
      "step": 32175
    },
    {
      "epoch": 18.974056603773583,
      "grad_norm": 3.7398219108581543,
      "learning_rate": 8.003474277580803e-08,
      "loss": 0.3009,
      "num_input_tokens_seen": 20853176,
      "step": 32180
    },
    {
      "epoch": 18.97700471698113,
      "grad_norm": 3.7303824424743652,
      "learning_rate": 7.95769188578166e-08,
      "loss": 0.2908,
      "num_input_tokens_seen": 20856184,
      "step": 32185
    },
    {
      "epoch": 18.97995283018868,
      "grad_norm": 3.5105772018432617,
      "learning_rate": 7.912039764304213e-08,
      "loss": 0.3602,
      "num_input_tokens_seen": 20858904,
      "step": 32190
    },
    {
      "epoch": 18.982900943396228,
      "grad_norm": 3.723850965499878,
      "learning_rate": 7.866517925235017e-08,
      "loss": 0.2596,
      "num_input_tokens_seen": 20861528,
      "step": 32195
    },
    {
      "epoch": 18.985849056603772,
      "grad_norm": 4.448110103607178,
      "learning_rate": 7.821126380626154e-08,
      "loss": 0.2796,
      "num_input_tokens_seen": 20864632,
      "step": 32200
    },
    {
      "epoch": 18.98879716981132,
      "grad_norm": 2.492372989654541,
      "learning_rate": 7.775865142495286e-08,
      "loss": 0.2716,
      "num_input_tokens_seen": 20868696,
      "step": 32205
    },
    {
      "epoch": 18.99174528301887,
      "grad_norm": 3.920567035675049,
      "learning_rate": 7.730734222825442e-08,
      "loss": 0.2402,
      "num_input_tokens_seen": 20871768,
      "step": 32210
    },
    {
      "epoch": 18.994693396226417,
      "grad_norm": 4.387699604034424,
      "learning_rate": 7.68573363356534e-08,
      "loss": 0.3248,
      "num_input_tokens_seen": 20874904,
      "step": 32215
    },
    {
      "epoch": 18.99764150943396,
      "grad_norm": 2.630159616470337,
      "learning_rate": 7.640863386629005e-08,
      "loss": 0.2606,
      "num_input_tokens_seen": 20878488,
      "step": 32220
    },
    {
      "epoch": 19.00058962264151,
      "grad_norm": 2.8753340244293213,
      "learning_rate": 7.59612349389599e-08,
      "loss": 0.2588,
      "num_input_tokens_seen": 20881088,
      "step": 32225
    },
    {
      "epoch": 19.003537735849058,
      "grad_norm": 6.377274513244629,
      "learning_rate": 7.551513967211433e-08,
      "loss": 0.3416,
      "num_input_tokens_seen": 20883904,
      "step": 32230
    },
    {
      "epoch": 19.006485849056602,
      "grad_norm": 3.4332687854766846,
      "learning_rate": 7.507034818385883e-08,
      "loss": 0.2169,
      "num_input_tokens_seen": 20886496,
      "step": 32235
    },
    {
      "epoch": 19.00943396226415,
      "grad_norm": 7.806169509887695,
      "learning_rate": 7.462686059195423e-08,
      "loss": 0.2538,
      "num_input_tokens_seen": 20889280,
      "step": 32240
    },
    {
      "epoch": 19.0123820754717,
      "grad_norm": 3.55637264251709,
      "learning_rate": 7.418467701381548e-08,
      "loss": 0.3138,
      "num_input_tokens_seen": 20892128,
      "step": 32245
    },
    {
      "epoch": 19.015330188679247,
      "grad_norm": 4.571549415588379,
      "learning_rate": 7.374379756651285e-08,
      "loss": 0.2758,
      "num_input_tokens_seen": 20895072,
      "step": 32250
    },
    {
      "epoch": 19.01827830188679,
      "grad_norm": 4.235462665557861,
      "learning_rate": 7.330422236677015e-08,
      "loss": 0.3487,
      "num_input_tokens_seen": 20898240,
      "step": 32255
    },
    {
      "epoch": 19.02122641509434,
      "grad_norm": 3.757370948791504,
      "learning_rate": 7.286595153096765e-08,
      "loss": 0.2456,
      "num_input_tokens_seen": 20900960,
      "step": 32260
    },
    {
      "epoch": 19.024174528301888,
      "grad_norm": 3.0288753509521484,
      "learning_rate": 7.242898517513864e-08,
      "loss": 0.3145,
      "num_input_tokens_seen": 20904480,
      "step": 32265
    },
    {
      "epoch": 19.027122641509433,
      "grad_norm": 4.151926517486572,
      "learning_rate": 7.199332341497333e-08,
      "loss": 0.3807,
      "num_input_tokens_seen": 20907168,
      "step": 32270
    },
    {
      "epoch": 19.03007075471698,
      "grad_norm": 1.8403031826019287,
      "learning_rate": 7.155896636581394e-08,
      "loss": 0.2599,
      "num_input_tokens_seen": 20911072,
      "step": 32275
    },
    {
      "epoch": 19.03301886792453,
      "grad_norm": 3.9090182781219482,
      "learning_rate": 7.112591414265901e-08,
      "loss": 0.3637,
      "num_input_tokens_seen": 20913280,
      "step": 32280
    },
    {
      "epoch": 19.035966981132077,
      "grad_norm": 4.361315727233887,
      "learning_rate": 7.069416686016018e-08,
      "loss": 0.3598,
      "num_input_tokens_seen": 20916352,
      "step": 32285
    },
    {
      "epoch": 19.03891509433962,
      "grad_norm": 4.337821006774902,
      "learning_rate": 7.026372463262488e-08,
      "loss": 0.2956,
      "num_input_tokens_seen": 20919264,
      "step": 32290
    },
    {
      "epoch": 19.04186320754717,
      "grad_norm": 2.553978681564331,
      "learning_rate": 6.983458757401418e-08,
      "loss": 0.2986,
      "num_input_tokens_seen": 20923264,
      "step": 32295
    },
    {
      "epoch": 19.044811320754718,
      "grad_norm": 3.306420087814331,
      "learning_rate": 6.940675579794443e-08,
      "loss": 0.2652,
      "num_input_tokens_seen": 20926208,
      "step": 32300
    },
    {
      "epoch": 19.047759433962263,
      "grad_norm": 3.112884998321533,
      "learning_rate": 6.898022941768612e-08,
      "loss": 0.3195,
      "num_input_tokens_seen": 20929440,
      "step": 32305
    },
    {
      "epoch": 19.05070754716981,
      "grad_norm": 3.95383882522583,
      "learning_rate": 6.855500854616337e-08,
      "loss": 0.3496,
      "num_input_tokens_seen": 20932384,
      "step": 32310
    },
    {
      "epoch": 19.05365566037736,
      "grad_norm": 4.244848251342773,
      "learning_rate": 6.813109329595557e-08,
      "loss": 0.3663,
      "num_input_tokens_seen": 20935904,
      "step": 32315
    },
    {
      "epoch": 19.056603773584907,
      "grad_norm": 4.072212219238281,
      "learning_rate": 6.770848377929573e-08,
      "loss": 0.3089,
      "num_input_tokens_seen": 20939456,
      "step": 32320
    },
    {
      "epoch": 19.059551886792452,
      "grad_norm": 4.920459747314453,
      "learning_rate": 6.728718010807156e-08,
      "loss": 0.317,
      "num_input_tokens_seen": 20941952,
      "step": 32325
    },
    {
      "epoch": 19.0625,
      "grad_norm": 4.775054931640625,
      "learning_rate": 6.68671823938255e-08,
      "loss": 0.3014,
      "num_input_tokens_seen": 20945568,
      "step": 32330
    },
    {
      "epoch": 19.065448113207548,
      "grad_norm": 2.2541146278381348,
      "learning_rate": 6.644849074775361e-08,
      "loss": 0.3416,
      "num_input_tokens_seen": 20949824,
      "step": 32335
    },
    {
      "epoch": 19.068396226415093,
      "grad_norm": 4.025348663330078,
      "learning_rate": 6.603110528070667e-08,
      "loss": 0.3338,
      "num_input_tokens_seen": 20954048,
      "step": 32340
    },
    {
      "epoch": 19.07134433962264,
      "grad_norm": 2.7630727291107178,
      "learning_rate": 6.561502610318849e-08,
      "loss": 0.2985,
      "num_input_tokens_seen": 20956672,
      "step": 32345
    },
    {
      "epoch": 19.07429245283019,
      "grad_norm": 5.928762912750244,
      "learning_rate": 6.520025332535762e-08,
      "loss": 0.2575,
      "num_input_tokens_seen": 20959552,
      "step": 32350
    },
    {
      "epoch": 19.077240566037737,
      "grad_norm": 4.50465726852417,
      "learning_rate": 6.47867870570279e-08,
      "loss": 0.2427,
      "num_input_tokens_seen": 20962496,
      "step": 32355
    },
    {
      "epoch": 19.080188679245282,
      "grad_norm": 3.8173866271972656,
      "learning_rate": 6.437462740766564e-08,
      "loss": 0.3383,
      "num_input_tokens_seen": 20964672,
      "step": 32360
    },
    {
      "epoch": 19.08313679245283,
      "grad_norm": 3.098423480987549,
      "learning_rate": 6.396377448639246e-08,
      "loss": 0.3933,
      "num_input_tokens_seen": 20968416,
      "step": 32365
    },
    {
      "epoch": 19.08608490566038,
      "grad_norm": 3.9741082191467285,
      "learning_rate": 6.3554228401983e-08,
      "loss": 0.4073,
      "num_input_tokens_seen": 20971168,
      "step": 32370
    },
    {
      "epoch": 19.089033018867923,
      "grad_norm": 4.053957939147949,
      "learning_rate": 6.314598926286663e-08,
      "loss": 0.2887,
      "num_input_tokens_seen": 20974336,
      "step": 32375
    },
    {
      "epoch": 19.09198113207547,
      "grad_norm": 0.6818675398826599,
      "learning_rate": 6.273905717712637e-08,
      "loss": 0.3741,
      "num_input_tokens_seen": 20981088,
      "step": 32380
    },
    {
      "epoch": 19.09492924528302,
      "grad_norm": 5.522767066955566,
      "learning_rate": 6.233343225249933e-08,
      "loss": 0.2871,
      "num_input_tokens_seen": 20983936,
      "step": 32385
    },
    {
      "epoch": 19.097877358490567,
      "grad_norm": 3.6612770557403564,
      "learning_rate": 6.192911459637519e-08,
      "loss": 0.2537,
      "num_input_tokens_seen": 20987744,
      "step": 32390
    },
    {
      "epoch": 19.100825471698112,
      "grad_norm": 3.2058916091918945,
      "learning_rate": 6.152610431580052e-08,
      "loss": 0.2271,
      "num_input_tokens_seen": 20990912,
      "step": 32395
    },
    {
      "epoch": 19.10377358490566,
      "grad_norm": 3.6963629722595215,
      "learning_rate": 6.112440151747389e-08,
      "loss": 0.2534,
      "num_input_tokens_seen": 20993728,
      "step": 32400
    },
    {
      "epoch": 19.10672169811321,
      "grad_norm": 2.2284040451049805,
      "learning_rate": 6.072400630774689e-08,
      "loss": 0.2821,
      "num_input_tokens_seen": 20998976,
      "step": 32405
    },
    {
      "epoch": 19.109669811320753,
      "grad_norm": 3.593777656555176,
      "learning_rate": 6.032491879262637e-08,
      "loss": 0.3669,
      "num_input_tokens_seen": 21002112,
      "step": 32410
    },
    {
      "epoch": 19.1126179245283,
      "grad_norm": 4.095821857452393,
      "learning_rate": 5.99271390777717e-08,
      "loss": 0.2991,
      "num_input_tokens_seen": 21005184,
      "step": 32415
    },
    {
      "epoch": 19.11556603773585,
      "grad_norm": 3.335825204849243,
      "learning_rate": 5.953066726849865e-08,
      "loss": 0.2043,
      "num_input_tokens_seen": 21008384,
      "step": 32420
    },
    {
      "epoch": 19.118514150943398,
      "grad_norm": 4.621332168579102,
      "learning_rate": 5.913550346977326e-08,
      "loss": 0.2106,
      "num_input_tokens_seen": 21011040,
      "step": 32425
    },
    {
      "epoch": 19.121462264150942,
      "grad_norm": 3.538419485092163,
      "learning_rate": 5.874164778621683e-08,
      "loss": 0.3276,
      "num_input_tokens_seen": 21014208,
      "step": 32430
    },
    {
      "epoch": 19.12441037735849,
      "grad_norm": 3.1272976398468018,
      "learning_rate": 5.834910032210539e-08,
      "loss": 0.3035,
      "num_input_tokens_seen": 21017120,
      "step": 32435
    },
    {
      "epoch": 19.12735849056604,
      "grad_norm": 3.5983729362487793,
      "learning_rate": 5.795786118136693e-08,
      "loss": 0.2503,
      "num_input_tokens_seen": 21020064,
      "step": 32440
    },
    {
      "epoch": 19.130306603773583,
      "grad_norm": 4.653785228729248,
      "learning_rate": 5.756793046758302e-08,
      "loss": 0.2248,
      "num_input_tokens_seen": 21022656,
      "step": 32445
    },
    {
      "epoch": 19.13325471698113,
      "grad_norm": 3.0955772399902344,
      "learning_rate": 5.7179308283990544e-08,
      "loss": 0.2393,
      "num_input_tokens_seen": 21024896,
      "step": 32450
    },
    {
      "epoch": 19.13620283018868,
      "grad_norm": 1.9963992834091187,
      "learning_rate": 5.679199473347885e-08,
      "loss": 0.2833,
      "num_input_tokens_seen": 21028384,
      "step": 32455
    },
    {
      "epoch": 19.139150943396228,
      "grad_norm": 4.167960166931152,
      "learning_rate": 5.6405989918590366e-08,
      "loss": 0.2244,
      "num_input_tokens_seen": 21031232,
      "step": 32460
    },
    {
      "epoch": 19.142099056603772,
      "grad_norm": 4.21812629699707,
      "learning_rate": 5.6021293941522225e-08,
      "loss": 0.2607,
      "num_input_tokens_seen": 21034048,
      "step": 32465
    },
    {
      "epoch": 19.14504716981132,
      "grad_norm": 4.088929653167725,
      "learning_rate": 5.563790690412352e-08,
      "loss": 0.2278,
      "num_input_tokens_seen": 21036544,
      "step": 32470
    },
    {
      "epoch": 19.14799528301887,
      "grad_norm": 2.144798994064331,
      "learning_rate": 5.525582890789805e-08,
      "loss": 0.2926,
      "num_input_tokens_seen": 21040448,
      "step": 32475
    },
    {
      "epoch": 19.150943396226417,
      "grad_norm": 6.578419208526611,
      "learning_rate": 5.4875060054002115e-08,
      "loss": 0.253,
      "num_input_tokens_seen": 21043776,
      "step": 32480
    },
    {
      "epoch": 19.15389150943396,
      "grad_norm": 3.2856485843658447,
      "learning_rate": 5.4495600443246755e-08,
      "loss": 0.2712,
      "num_input_tokens_seen": 21047008,
      "step": 32485
    },
    {
      "epoch": 19.15683962264151,
      "grad_norm": 2.856227159500122,
      "learning_rate": 5.411745017609493e-08,
      "loss": 0.2801,
      "num_input_tokens_seen": 21050208,
      "step": 32490
    },
    {
      "epoch": 19.159787735849058,
      "grad_norm": 2.929614543914795,
      "learning_rate": 5.374060935266434e-08,
      "loss": 0.1683,
      "num_input_tokens_seen": 21053920,
      "step": 32495
    },
    {
      "epoch": 19.162735849056602,
      "grad_norm": 2.8232169151306152,
      "learning_rate": 5.3365078072724065e-08,
      "loss": 0.2513,
      "num_input_tokens_seen": 21059232,
      "step": 32500
    },
    {
      "epoch": 19.16568396226415,
      "grad_norm": 5.5941901206970215,
      "learning_rate": 5.299085643569846e-08,
      "loss": 0.2611,
      "num_input_tokens_seen": 21062688,
      "step": 32505
    },
    {
      "epoch": 19.1686320754717,
      "grad_norm": 9.296927452087402,
      "learning_rate": 5.261794454066327e-08,
      "loss": 0.3815,
      "num_input_tokens_seen": 21065664,
      "step": 32510
    },
    {
      "epoch": 19.171580188679247,
      "grad_norm": 3.5578577518463135,
      "learning_rate": 5.224634248635008e-08,
      "loss": 0.3453,
      "num_input_tokens_seen": 21069088,
      "step": 32515
    },
    {
      "epoch": 19.17452830188679,
      "grad_norm": 4.509056091308594,
      "learning_rate": 5.187605037114129e-08,
      "loss": 0.3693,
      "num_input_tokens_seen": 21072128,
      "step": 32520
    },
    {
      "epoch": 19.17747641509434,
      "grad_norm": 2.3362414836883545,
      "learning_rate": 5.15070682930735e-08,
      "loss": 0.325,
      "num_input_tokens_seen": 21075200,
      "step": 32525
    },
    {
      "epoch": 19.180424528301888,
      "grad_norm": 2.6404950618743896,
      "learning_rate": 5.113939634983578e-08,
      "loss": 0.3617,
      "num_input_tokens_seen": 21079328,
      "step": 32530
    },
    {
      "epoch": 19.183372641509433,
      "grad_norm": 4.926098346710205,
      "learning_rate": 5.077303463877192e-08,
      "loss": 0.4012,
      "num_input_tokens_seen": 21082592,
      "step": 32535
    },
    {
      "epoch": 19.18632075471698,
      "grad_norm": 4.013546466827393,
      "learning_rate": 5.040798325687601e-08,
      "loss": 0.3216,
      "num_input_tokens_seen": 21085408,
      "step": 32540
    },
    {
      "epoch": 19.18926886792453,
      "grad_norm": 1.9534202814102173,
      "learning_rate": 5.004424230079852e-08,
      "loss": 0.2484,
      "num_input_tokens_seen": 21088544,
      "step": 32545
    },
    {
      "epoch": 19.192216981132077,
      "grad_norm": 4.973866939544678,
      "learning_rate": 4.968181186684129e-08,
      "loss": 0.2451,
      "num_input_tokens_seen": 21092480,
      "step": 32550
    },
    {
      "epoch": 19.19516509433962,
      "grad_norm": 4.048649311065674,
      "learning_rate": 4.932069205095924e-08,
      "loss": 0.366,
      "num_input_tokens_seen": 21096384,
      "step": 32555
    },
    {
      "epoch": 19.19811320754717,
      "grad_norm": 6.3376994132995605,
      "learning_rate": 4.896088294875978e-08,
      "loss": 0.2431,
      "num_input_tokens_seen": 21099776,
      "step": 32560
    },
    {
      "epoch": 19.201061320754718,
      "grad_norm": 3.2446773052215576,
      "learning_rate": 4.8602384655505044e-08,
      "loss": 0.2536,
      "num_input_tokens_seen": 21103264,
      "step": 32565
    },
    {
      "epoch": 19.204009433962263,
      "grad_norm": 4.7571821212768555,
      "learning_rate": 4.824519726610744e-08,
      "loss": 0.3118,
      "num_input_tokens_seen": 21107008,
      "step": 32570
    },
    {
      "epoch": 19.20695754716981,
      "grad_norm": 4.024714469909668,
      "learning_rate": 4.7889320875135206e-08,
      "loss": 0.2161,
      "num_input_tokens_seen": 21109888,
      "step": 32575
    },
    {
      "epoch": 19.20990566037736,
      "grad_norm": 2.4304568767547607,
      "learning_rate": 4.753475557680742e-08,
      "loss": 0.2441,
      "num_input_tokens_seen": 21113440,
      "step": 32580
    },
    {
      "epoch": 19.212853773584907,
      "grad_norm": 4.532256126403809,
      "learning_rate": 4.718150146499734e-08,
      "loss": 0.3763,
      "num_input_tokens_seen": 21116832,
      "step": 32585
    },
    {
      "epoch": 19.215801886792452,
      "grad_norm": 3.766035556793213,
      "learning_rate": 4.682955863323013e-08,
      "loss": 0.2561,
      "num_input_tokens_seen": 21120256,
      "step": 32590
    },
    {
      "epoch": 19.21875,
      "grad_norm": 3.7971408367156982,
      "learning_rate": 4.6478927174684606e-08,
      "loss": 0.2376,
      "num_input_tokens_seen": 21123712,
      "step": 32595
    },
    {
      "epoch": 19.221698113207548,
      "grad_norm": 4.338912487030029,
      "learning_rate": 4.612960718219095e-08,
      "loss": 0.4045,
      "num_input_tokens_seen": 21126624,
      "step": 32600
    },
    {
      "epoch": 19.224646226415093,
      "grad_norm": 4.19905424118042,
      "learning_rate": 4.578159874823407e-08,
      "loss": 0.2975,
      "num_input_tokens_seen": 21129504,
      "step": 32605
    },
    {
      "epoch": 19.22759433962264,
      "grad_norm": 4.14419412612915,
      "learning_rate": 4.5434901964950264e-08,
      "loss": 0.3548,
      "num_input_tokens_seen": 21132864,
      "step": 32610
    },
    {
      "epoch": 19.23054245283019,
      "grad_norm": 8.344884872436523,
      "learning_rate": 4.508951692412944e-08,
      "loss": 0.2953,
      "num_input_tokens_seen": 21135264,
      "step": 32615
    },
    {
      "epoch": 19.233490566037737,
      "grad_norm": 2.757068395614624,
      "learning_rate": 4.4745443717213455e-08,
      "loss": 0.2935,
      "num_input_tokens_seen": 21138816,
      "step": 32620
    },
    {
      "epoch": 19.236438679245282,
      "grad_norm": 5.082851409912109,
      "learning_rate": 4.4402682435296666e-08,
      "loss": 0.2693,
      "num_input_tokens_seen": 21141440,
      "step": 32625
    },
    {
      "epoch": 19.23938679245283,
      "grad_norm": 3.948082447052002,
      "learning_rate": 4.406123316912758e-08,
      "loss": 0.3014,
      "num_input_tokens_seen": 21144864,
      "step": 32630
    },
    {
      "epoch": 19.24233490566038,
      "grad_norm": 2.318446159362793,
      "learning_rate": 4.372109600910612e-08,
      "loss": 0.1997,
      "num_input_tokens_seen": 21148416,
      "step": 32635
    },
    {
      "epoch": 19.245283018867923,
      "grad_norm": 6.14608907699585,
      "learning_rate": 4.338227104528414e-08,
      "loss": 0.275,
      "num_input_tokens_seen": 21150624,
      "step": 32640
    },
    {
      "epoch": 19.24823113207547,
      "grad_norm": 3.8685154914855957,
      "learning_rate": 4.304475836736821e-08,
      "loss": 0.2761,
      "num_input_tokens_seen": 21153248,
      "step": 32645
    },
    {
      "epoch": 19.25117924528302,
      "grad_norm": 3.653660774230957,
      "learning_rate": 4.27085580647163e-08,
      "loss": 0.3171,
      "num_input_tokens_seen": 21156064,
      "step": 32650
    },
    {
      "epoch": 19.254127358490567,
      "grad_norm": 3.14099383354187,
      "learning_rate": 4.237367022633776e-08,
      "loss": 0.3644,
      "num_input_tokens_seen": 21160288,
      "step": 32655
    },
    {
      "epoch": 19.257075471698112,
      "grad_norm": 4.721896648406982,
      "learning_rate": 4.204009494089612e-08,
      "loss": 0.3241,
      "num_input_tokens_seen": 21163008,
      "step": 32660
    },
    {
      "epoch": 19.26002358490566,
      "grad_norm": 4.021838665008545,
      "learning_rate": 4.170783229670739e-08,
      "loss": 0.2718,
      "num_input_tokens_seen": 21166400,
      "step": 32665
    },
    {
      "epoch": 19.26297169811321,
      "grad_norm": 3.704167366027832,
      "learning_rate": 4.137688238173898e-08,
      "loss": 0.4598,
      "num_input_tokens_seen": 21169312,
      "step": 32670
    },
    {
      "epoch": 19.265919811320753,
      "grad_norm": 2.4168527126312256,
      "learning_rate": 4.104724528361137e-08,
      "loss": 0.3079,
      "num_input_tokens_seen": 21172512,
      "step": 32675
    },
    {
      "epoch": 19.2688679245283,
      "grad_norm": 4.582902431488037,
      "learning_rate": 4.071892108959752e-08,
      "loss": 0.4453,
      "num_input_tokens_seen": 21176000,
      "step": 32680
    },
    {
      "epoch": 19.27181603773585,
      "grad_norm": 6.267623424530029,
      "learning_rate": 4.039190988662234e-08,
      "loss": 0.2429,
      "num_input_tokens_seen": 21178944,
      "step": 32685
    },
    {
      "epoch": 19.274764150943398,
      "grad_norm": 8.04636001586914,
      "learning_rate": 4.006621176126435e-08,
      "loss": 0.279,
      "num_input_tokens_seen": 21181376,
      "step": 32690
    },
    {
      "epoch": 19.277712264150942,
      "grad_norm": 4.863682270050049,
      "learning_rate": 3.974182679975236e-08,
      "loss": 0.3476,
      "num_input_tokens_seen": 21184576,
      "step": 32695
    },
    {
      "epoch": 19.28066037735849,
      "grad_norm": 5.062206745147705,
      "learning_rate": 3.941875508796933e-08,
      "loss": 0.319,
      "num_input_tokens_seen": 21187232,
      "step": 32700
    },
    {
      "epoch": 19.28360849056604,
      "grad_norm": 3.79584002494812,
      "learning_rate": 3.909699671145017e-08,
      "loss": 0.3176,
      "num_input_tokens_seen": 21190624,
      "step": 32705
    },
    {
      "epoch": 19.286556603773583,
      "grad_norm": 2.78607177734375,
      "learning_rate": 3.87765517553812e-08,
      "loss": 0.2357,
      "num_input_tokens_seen": 21193856,
      "step": 32710
    },
    {
      "epoch": 19.28950471698113,
      "grad_norm": 3.391711473464966,
      "learning_rate": 3.8457420304601756e-08,
      "loss": 0.3384,
      "num_input_tokens_seen": 21197152,
      "step": 32715
    },
    {
      "epoch": 19.29245283018868,
      "grad_norm": 5.205026626586914,
      "learning_rate": 3.813960244360371e-08,
      "loss": 0.2962,
      "num_input_tokens_seen": 21200704,
      "step": 32720
    },
    {
      "epoch": 19.295400943396228,
      "grad_norm": 5.2049360275268555,
      "learning_rate": 3.7823098256529744e-08,
      "loss": 0.2655,
      "num_input_tokens_seen": 21203936,
      "step": 32725
    },
    {
      "epoch": 19.298349056603772,
      "grad_norm": 5.820584297180176,
      "learning_rate": 3.750790782717673e-08,
      "loss": 0.4276,
      "num_input_tokens_seen": 21206944,
      "step": 32730
    },
    {
      "epoch": 19.30129716981132,
      "grad_norm": 3.673015594482422,
      "learning_rate": 3.719403123899179e-08,
      "loss": 0.3394,
      "num_input_tokens_seen": 21209888,
      "step": 32735
    },
    {
      "epoch": 19.30424528301887,
      "grad_norm": 5.981060981750488,
      "learning_rate": 3.688146857507624e-08,
      "loss": 0.3703,
      "num_input_tokens_seen": 21212960,
      "step": 32740
    },
    {
      "epoch": 19.307193396226417,
      "grad_norm": 3.1982288360595703,
      "learning_rate": 3.657021991818166e-08,
      "loss": 0.3404,
      "num_input_tokens_seen": 21216000,
      "step": 32745
    },
    {
      "epoch": 19.31014150943396,
      "grad_norm": 4.460146427154541,
      "learning_rate": 3.626028535071213e-08,
      "loss": 0.2546,
      "num_input_tokens_seen": 21220096,
      "step": 32750
    },
    {
      "epoch": 19.31308962264151,
      "grad_norm": 4.9204816818237305,
      "learning_rate": 3.59516649547248e-08,
      "loss": 0.2688,
      "num_input_tokens_seen": 21222688,
      "step": 32755
    },
    {
      "epoch": 19.316037735849058,
      "grad_norm": 5.429598331451416,
      "learning_rate": 3.564435881192818e-08,
      "loss": 0.2931,
      "num_input_tokens_seen": 21225024,
      "step": 32760
    },
    {
      "epoch": 19.318985849056602,
      "grad_norm": 3.719420909881592,
      "learning_rate": 3.5338367003682763e-08,
      "loss": 0.2645,
      "num_input_tokens_seen": 21227616,
      "step": 32765
    },
    {
      "epoch": 19.32193396226415,
      "grad_norm": 4.588597297668457,
      "learning_rate": 3.5033689611000954e-08,
      "loss": 0.3418,
      "num_input_tokens_seen": 21230496,
      "step": 32770
    },
    {
      "epoch": 19.3248820754717,
      "grad_norm": 3.7868168354034424,
      "learning_rate": 3.473032671454768e-08,
      "loss": 0.2868,
      "num_input_tokens_seen": 21234720,
      "step": 32775
    },
    {
      "epoch": 19.327830188679247,
      "grad_norm": 3.387484550476074,
      "learning_rate": 3.44282783946398e-08,
      "loss": 0.2912,
      "num_input_tokens_seen": 21238144,
      "step": 32780
    },
    {
      "epoch": 19.33077830188679,
      "grad_norm": 2.3591299057006836,
      "learning_rate": 3.4127544731245575e-08,
      "loss": 0.1854,
      "num_input_tokens_seen": 21242240,
      "step": 32785
    },
    {
      "epoch": 19.33372641509434,
      "grad_norm": 4.697345733642578,
      "learning_rate": 3.38281258039852e-08,
      "loss": 0.214,
      "num_input_tokens_seen": 21244864,
      "step": 32790
    },
    {
      "epoch": 19.336674528301888,
      "grad_norm": 3.70378041267395,
      "learning_rate": 3.353002169213193e-08,
      "loss": 0.3353,
      "num_input_tokens_seen": 21247680,
      "step": 32795
    },
    {
      "epoch": 19.339622641509433,
      "grad_norm": 4.402281761169434,
      "learning_rate": 3.32332324746093e-08,
      "loss": 0.2657,
      "num_input_tokens_seen": 21251136,
      "step": 32800
    },
    {
      "epoch": 19.34257075471698,
      "grad_norm": 2.3096859455108643,
      "learning_rate": 3.2937758229994455e-08,
      "loss": 0.3024,
      "num_input_tokens_seen": 21253952,
      "step": 32805
    },
    {
      "epoch": 19.34551886792453,
      "grad_norm": 2.9201154708862305,
      "learning_rate": 3.2643599036514815e-08,
      "loss": 0.2623,
      "num_input_tokens_seen": 21256992,
      "step": 32810
    },
    {
      "epoch": 19.348466981132077,
      "grad_norm": 4.934100151062012,
      "learning_rate": 3.2350754972050316e-08,
      "loss": 0.2617,
      "num_input_tokens_seen": 21260096,
      "step": 32815
    },
    {
      "epoch": 19.35141509433962,
      "grad_norm": 3.14333176612854,
      "learning_rate": 3.2059226114132815e-08,
      "loss": 0.3987,
      "num_input_tokens_seen": 21263360,
      "step": 32820
    },
    {
      "epoch": 19.35436320754717,
      "grad_norm": 4.45322847366333,
      "learning_rate": 3.1769012539945575e-08,
      "loss": 0.3651,
      "num_input_tokens_seen": 21266400,
      "step": 32825
    },
    {
      "epoch": 19.357311320754718,
      "grad_norm": 3.1754331588745117,
      "learning_rate": 3.1480114326324364e-08,
      "loss": 0.3485,
      "num_input_tokens_seen": 21269664,
      "step": 32830
    },
    {
      "epoch": 19.360259433962263,
      "grad_norm": 3.995297431945801,
      "learning_rate": 3.1192531549756325e-08,
      "loss": 0.2642,
      "num_input_tokens_seen": 21272576,
      "step": 32835
    },
    {
      "epoch": 19.36320754716981,
      "grad_norm": 2.752998113632202,
      "learning_rate": 3.090626428638e-08,
      "loss": 0.1857,
      "num_input_tokens_seen": 21275136,
      "step": 32840
    },
    {
      "epoch": 19.36615566037736,
      "grad_norm": 9.711514472961426,
      "learning_rate": 3.062131261198531e-08,
      "loss": 0.3547,
      "num_input_tokens_seen": 21277952,
      "step": 32845
    },
    {
      "epoch": 19.369103773584907,
      "grad_norm": 3.667126417160034,
      "learning_rate": 3.033767660201525e-08,
      "loss": 0.3375,
      "num_input_tokens_seen": 21280992,
      "step": 32850
    },
    {
      "epoch": 19.372051886792452,
      "grad_norm": 2.414670705795288,
      "learning_rate": 3.005535633156309e-08,
      "loss": 0.3619,
      "num_input_tokens_seen": 21284192,
      "step": 32855
    },
    {
      "epoch": 19.375,
      "grad_norm": 5.321611404418945,
      "learning_rate": 2.977435187537514e-08,
      "loss": 0.277,
      "num_input_tokens_seen": 21286688,
      "step": 32860
    },
    {
      "epoch": 19.377948113207548,
      "grad_norm": 3.6359691619873047,
      "learning_rate": 2.949466330784745e-08,
      "loss": 0.2654,
      "num_input_tokens_seen": 21289824,
      "step": 32865
    },
    {
      "epoch": 19.380896226415093,
      "grad_norm": 8.776988983154297,
      "learning_rate": 2.921629070302967e-08,
      "loss": 0.3032,
      "num_input_tokens_seen": 21293440,
      "step": 32870
    },
    {
      "epoch": 19.38384433962264,
      "grad_norm": 4.13643217086792,
      "learning_rate": 2.893923413462174e-08,
      "loss": 0.3396,
      "num_input_tokens_seen": 21297312,
      "step": 32875
    },
    {
      "epoch": 19.38679245283019,
      "grad_norm": 3.8492178916931152,
      "learning_rate": 2.866349367597554e-08,
      "loss": 0.3808,
      "num_input_tokens_seen": 21300960,
      "step": 32880
    },
    {
      "epoch": 19.389740566037737,
      "grad_norm": 3.2361409664154053,
      "learning_rate": 2.8389069400094893e-08,
      "loss": 0.2523,
      "num_input_tokens_seen": 21303456,
      "step": 32885
    },
    {
      "epoch": 19.392688679245282,
      "grad_norm": 4.051593780517578,
      "learning_rate": 2.811596137963446e-08,
      "loss": 0.2958,
      "num_input_tokens_seen": 21307424,
      "step": 32890
    },
    {
      "epoch": 19.39563679245283,
      "grad_norm": 5.568869590759277,
      "learning_rate": 2.7844169686900844e-08,
      "loss": 0.3172,
      "num_input_tokens_seen": 21310336,
      "step": 32895
    },
    {
      "epoch": 19.39858490566038,
      "grad_norm": 3.626596689224243,
      "learning_rate": 2.7573694393852047e-08,
      "loss": 0.2073,
      "num_input_tokens_seen": 21313088,
      "step": 32900
    },
    {
      "epoch": 19.401533018867923,
      "grad_norm": 3.016526222229004,
      "learning_rate": 2.7304535572098e-08,
      "loss": 0.323,
      "num_input_tokens_seen": 21316992,
      "step": 32905
    },
    {
      "epoch": 19.40448113207547,
      "grad_norm": 8.25583267211914,
      "learning_rate": 2.703669329289893e-08,
      "loss": 0.3104,
      "num_input_tokens_seen": 21320640,
      "step": 32910
    },
    {
      "epoch": 19.40742924528302,
      "grad_norm": 4.020203590393066,
      "learning_rate": 2.6770167627167554e-08,
      "loss": 0.4303,
      "num_input_tokens_seen": 21323552,
      "step": 32915
    },
    {
      "epoch": 19.410377358490567,
      "grad_norm": 5.284704208374023,
      "learning_rate": 2.6504958645467426e-08,
      "loss": 0.4055,
      "num_input_tokens_seen": 21326400,
      "step": 32920
    },
    {
      "epoch": 19.413325471698112,
      "grad_norm": 4.681271553039551,
      "learning_rate": 2.6241066418014605e-08,
      "loss": 0.3562,
      "num_input_tokens_seen": 21330016,
      "step": 32925
    },
    {
      "epoch": 19.41627358490566,
      "grad_norm": 3.499631404876709,
      "learning_rate": 2.5978491014674866e-08,
      "loss": 0.3445,
      "num_input_tokens_seen": 21332640,
      "step": 32930
    },
    {
      "epoch": 19.41922169811321,
      "grad_norm": 3.420973300933838,
      "learning_rate": 2.571723250496705e-08,
      "loss": 0.3195,
      "num_input_tokens_seen": 21335712,
      "step": 32935
    },
    {
      "epoch": 19.422169811320753,
      "grad_norm": 5.230913162231445,
      "learning_rate": 2.5457290958059155e-08,
      "loss": 0.2897,
      "num_input_tokens_seen": 21338240,
      "step": 32940
    },
    {
      "epoch": 19.4251179245283,
      "grad_norm": 3.3544957637786865,
      "learning_rate": 2.51986664427728e-08,
      "loss": 0.2384,
      "num_input_tokens_seen": 21341952,
      "step": 32945
    },
    {
      "epoch": 19.42806603773585,
      "grad_norm": 8.451828002929688,
      "learning_rate": 2.4941359027579883e-08,
      "loss": 0.3004,
      "num_input_tokens_seen": 21345728,
      "step": 32950
    },
    {
      "epoch": 19.431014150943398,
      "grad_norm": 2.829627275466919,
      "learning_rate": 2.468536878060368e-08,
      "loss": 0.2177,
      "num_input_tokens_seen": 21349344,
      "step": 32955
    },
    {
      "epoch": 19.433962264150942,
      "grad_norm": 7.938571453094482,
      "learning_rate": 2.443069576961832e-08,
      "loss": 0.2864,
      "num_input_tokens_seen": 21352192,
      "step": 32960
    },
    {
      "epoch": 19.43691037735849,
      "grad_norm": 4.80275821685791,
      "learning_rate": 2.4177340062049304e-08,
      "loss": 0.2735,
      "num_input_tokens_seen": 21356064,
      "step": 32965
    },
    {
      "epoch": 19.43985849056604,
      "grad_norm": 3.518430233001709,
      "learning_rate": 2.3925301724974647e-08,
      "loss": 0.2494,
      "num_input_tokens_seen": 21358848,
      "step": 32970
    },
    {
      "epoch": 19.442806603773583,
      "grad_norm": 2.726673126220703,
      "learning_rate": 2.367458082512153e-08,
      "loss": 0.2319,
      "num_input_tokens_seen": 21363424,
      "step": 32975
    },
    {
      "epoch": 19.44575471698113,
      "grad_norm": 4.843780517578125,
      "learning_rate": 2.3425177428870737e-08,
      "loss": 0.3458,
      "num_input_tokens_seen": 21365856,
      "step": 32980
    },
    {
      "epoch": 19.44870283018868,
      "grad_norm": 3.56255841255188,
      "learning_rate": 2.3177091602251677e-08,
      "loss": 0.2815,
      "num_input_tokens_seen": 21369664,
      "step": 32985
    },
    {
      "epoch": 19.451650943396228,
      "grad_norm": 3.6355173587799072,
      "learning_rate": 2.2930323410946254e-08,
      "loss": 0.3171,
      "num_input_tokens_seen": 21373088,
      "step": 32990
    },
    {
      "epoch": 19.454599056603772,
      "grad_norm": 3.565376043319702,
      "learning_rate": 2.2684872920287758e-08,
      "loss": 0.358,
      "num_input_tokens_seen": 21376000,
      "step": 32995
    },
    {
      "epoch": 19.45754716981132,
      "grad_norm": 3.5196330547332764,
      "learning_rate": 2.2440740195260323e-08,
      "loss": 0.2495,
      "num_input_tokens_seen": 21378528,
      "step": 33000
    },
    {
      "epoch": 19.46049528301887,
      "grad_norm": 6.900623321533203,
      "learning_rate": 2.219792530049891e-08,
      "loss": 0.2827,
      "num_input_tokens_seen": 21381504,
      "step": 33005
    },
    {
      "epoch": 19.463443396226417,
      "grad_norm": 4.412519454956055,
      "learning_rate": 2.1956428300290434e-08,
      "loss": 0.2914,
      "num_input_tokens_seen": 21385696,
      "step": 33010
    },
    {
      "epoch": 19.46639150943396,
      "grad_norm": 4.681760787963867,
      "learning_rate": 2.1716249258570966e-08,
      "loss": 0.3508,
      "num_input_tokens_seen": 21388992,
      "step": 33015
    },
    {
      "epoch": 19.46933962264151,
      "grad_norm": 3.2968482971191406,
      "learning_rate": 2.1477388238930196e-08,
      "loss": 0.2793,
      "num_input_tokens_seen": 21392288,
      "step": 33020
    },
    {
      "epoch": 19.472287735849058,
      "grad_norm": 2.899982213973999,
      "learning_rate": 2.1239845304606988e-08,
      "loss": 0.2574,
      "num_input_tokens_seen": 21396224,
      "step": 33025
    },
    {
      "epoch": 19.475235849056602,
      "grad_norm": 4.6290082931518555,
      "learning_rate": 2.100362051849214e-08,
      "loss": 0.365,
      "num_input_tokens_seen": 21399424,
      "step": 33030
    },
    {
      "epoch": 19.47818396226415,
      "grad_norm": 4.322054862976074,
      "learning_rate": 2.076871394312674e-08,
      "loss": 0.3655,
      "num_input_tokens_seen": 21403104,
      "step": 33035
    },
    {
      "epoch": 19.4811320754717,
      "grad_norm": 5.071136474609375,
      "learning_rate": 2.0535125640703813e-08,
      "loss": 0.2675,
      "num_input_tokens_seen": 21406272,
      "step": 33040
    },
    {
      "epoch": 19.484080188679247,
      "grad_norm": 4.364553451538086,
      "learning_rate": 2.0302855673066667e-08,
      "loss": 0.2669,
      "num_input_tokens_seen": 21409088,
      "step": 33045
    },
    {
      "epoch": 19.48702830188679,
      "grad_norm": 3.448744058609009,
      "learning_rate": 2.0071904101710004e-08,
      "loss": 0.3124,
      "num_input_tokens_seen": 21412544,
      "step": 33050
    },
    {
      "epoch": 19.48997641509434,
      "grad_norm": 3.4545719623565674,
      "learning_rate": 1.98422709877788e-08,
      "loss": 0.3364,
      "num_input_tokens_seen": 21419936,
      "step": 33055
    },
    {
      "epoch": 19.492924528301888,
      "grad_norm": 2.6808292865753174,
      "learning_rate": 1.961395639206942e-08,
      "loss": 0.4211,
      "num_input_tokens_seen": 21422688,
      "step": 33060
    },
    {
      "epoch": 19.495872641509433,
      "grad_norm": 3.380133867263794,
      "learning_rate": 1.9386960375029628e-08,
      "loss": 0.3929,
      "num_input_tokens_seen": 21425664,
      "step": 33065
    },
    {
      "epoch": 19.49882075471698,
      "grad_norm": 3.105325222015381,
      "learning_rate": 1.9161282996757458e-08,
      "loss": 0.2491,
      "num_input_tokens_seen": 21428928,
      "step": 33070
    },
    {
      "epoch": 19.50176886792453,
      "grad_norm": 5.1926655769348145,
      "learning_rate": 1.8936924317001225e-08,
      "loss": 0.2094,
      "num_input_tokens_seen": 21431904,
      "step": 33075
    },
    {
      "epoch": 19.504716981132077,
      "grad_norm": 5.435290336608887,
      "learning_rate": 1.8713884395162308e-08,
      "loss": 0.3965,
      "num_input_tokens_seen": 21435264,
      "step": 33080
    },
    {
      "epoch": 19.50766509433962,
      "grad_norm": 4.782301425933838,
      "learning_rate": 1.8492163290290132e-08,
      "loss": 0.2651,
      "num_input_tokens_seen": 21438528,
      "step": 33085
    },
    {
      "epoch": 19.51061320754717,
      "grad_norm": 4.320066928863525,
      "learning_rate": 1.827176106108719e-08,
      "loss": 0.4007,
      "num_input_tokens_seen": 21441696,
      "step": 33090
    },
    {
      "epoch": 19.513561320754718,
      "grad_norm": 4.5389299392700195,
      "learning_rate": 1.8052677765905137e-08,
      "loss": 0.3099,
      "num_input_tokens_seen": 21444032,
      "step": 33095
    },
    {
      "epoch": 19.516509433962263,
      "grad_norm": 4.011986255645752,
      "learning_rate": 1.783491346274757e-08,
      "loss": 0.4484,
      "num_input_tokens_seen": 21447328,
      "step": 33100
    },
    {
      "epoch": 19.51945754716981,
      "grad_norm": 2.796394109725952,
      "learning_rate": 1.7618468209268936e-08,
      "loss": 0.2379,
      "num_input_tokens_seen": 21451040,
      "step": 33105
    },
    {
      "epoch": 19.52240566037736,
      "grad_norm": 6.38177490234375,
      "learning_rate": 1.7403342062773943e-08,
      "loss": 0.2742,
      "num_input_tokens_seen": 21453696,
      "step": 33110
    },
    {
      "epoch": 19.525353773584907,
      "grad_norm": 3.9902420043945312,
      "learning_rate": 1.718953508021759e-08,
      "loss": 0.2586,
      "num_input_tokens_seen": 21456992,
      "step": 33115
    },
    {
      "epoch": 19.528301886792452,
      "grad_norm": 3.959474563598633,
      "learning_rate": 1.6977047318206262e-08,
      "loss": 0.4137,
      "num_input_tokens_seen": 21459712,
      "step": 33120
    },
    {
      "epoch": 19.53125,
      "grad_norm": 4.778595924377441,
      "learning_rate": 1.676587883299774e-08,
      "loss": 0.3425,
      "num_input_tokens_seen": 21463200,
      "step": 33125
    },
    {
      "epoch": 19.534198113207548,
      "grad_norm": 3.5875895023345947,
      "learning_rate": 1.655602968049952e-08,
      "loss": 0.2631,
      "num_input_tokens_seen": 21465856,
      "step": 33130
    },
    {
      "epoch": 19.537146226415093,
      "grad_norm": 2.9150989055633545,
      "learning_rate": 1.634749991626938e-08,
      "loss": 0.3503,
      "num_input_tokens_seen": 21469568,
      "step": 33135
    },
    {
      "epoch": 19.54009433962264,
      "grad_norm": 3.298722982406616,
      "learning_rate": 1.6140289595517056e-08,
      "loss": 0.359,
      "num_input_tokens_seen": 21472864,
      "step": 33140
    },
    {
      "epoch": 19.54304245283019,
      "grad_norm": 4.8868327140808105,
      "learning_rate": 1.5934398773102545e-08,
      "loss": 0.4693,
      "num_input_tokens_seen": 21476096,
      "step": 33145
    },
    {
      "epoch": 19.545990566037737,
      "grad_norm": 3.2748353481292725,
      "learning_rate": 1.5729827503536133e-08,
      "loss": 0.3415,
      "num_input_tokens_seen": 21478944,
      "step": 33150
    },
    {
      "epoch": 19.548938679245282,
      "grad_norm": 5.3973798751831055,
      "learning_rate": 1.5526575840978942e-08,
      "loss": 0.32,
      "num_input_tokens_seen": 21481920,
      "step": 33155
    },
    {
      "epoch": 19.55188679245283,
      "grad_norm": 2.890684127807617,
      "learning_rate": 1.532464383924237e-08,
      "loss": 0.4036,
      "num_input_tokens_seen": 21485696,
      "step": 33160
    },
    {
      "epoch": 19.55483490566038,
      "grad_norm": 3.612659215927124,
      "learning_rate": 1.5124031551789208e-08,
      "loss": 0.3953,
      "num_input_tokens_seen": 21490272,
      "step": 33165
    },
    {
      "epoch": 19.557783018867923,
      "grad_norm": 6.92409610748291,
      "learning_rate": 1.4924739031732527e-08,
      "loss": 0.3254,
      "num_input_tokens_seen": 21492960,
      "step": 33170
    },
    {
      "epoch": 19.56073113207547,
      "grad_norm": 7.686550617218018,
      "learning_rate": 1.4726766331835118e-08,
      "loss": 0.874,
      "num_input_tokens_seen": 21496096,
      "step": 33175
    },
    {
      "epoch": 19.56367924528302,
      "grad_norm": 2.9055848121643066,
      "learning_rate": 1.4530113504512278e-08,
      "loss": 0.2856,
      "num_input_tokens_seen": 21499200,
      "step": 33180
    },
    {
      "epoch": 19.566627358490567,
      "grad_norm": 1.863372564315796,
      "learning_rate": 1.4334780601827914e-08,
      "loss": 0.1578,
      "num_input_tokens_seen": 21502400,
      "step": 33185
    },
    {
      "epoch": 19.569575471698112,
      "grad_norm": 3.7829017639160156,
      "learning_rate": 1.4140767675497325e-08,
      "loss": 0.2367,
      "num_input_tokens_seen": 21505088,
      "step": 33190
    },
    {
      "epoch": 19.57252358490566,
      "grad_norm": 4.712259292602539,
      "learning_rate": 1.394807477688609e-08,
      "loss": 0.3398,
      "num_input_tokens_seen": 21508736,
      "step": 33195
    },
    {
      "epoch": 19.57547169811321,
      "grad_norm": 3.9278736114501953,
      "learning_rate": 1.3756701957011177e-08,
      "loss": 0.2447,
      "num_input_tokens_seen": 21511296,
      "step": 33200
    },
    {
      "epoch": 19.578419811320753,
      "grad_norm": 3.9713051319122314,
      "learning_rate": 1.3566649266538723e-08,
      "loss": 0.2206,
      "num_input_tokens_seen": 21516128,
      "step": 33205
    },
    {
      "epoch": 19.5813679245283,
      "grad_norm": 3.8197531700134277,
      "learning_rate": 1.3377916755786257e-08,
      "loss": 0.3311,
      "num_input_tokens_seen": 21519392,
      "step": 33210
    },
    {
      "epoch": 19.58431603773585,
      "grad_norm": 4.363983154296875,
      "learning_rate": 1.3190504474721588e-08,
      "loss": 0.3115,
      "num_input_tokens_seen": 21522592,
      "step": 33215
    },
    {
      "epoch": 19.587264150943398,
      "grad_norm": 2.8366570472717285,
      "learning_rate": 1.3004412472962802e-08,
      "loss": 0.2384,
      "num_input_tokens_seen": 21525440,
      "step": 33220
    },
    {
      "epoch": 19.590212264150942,
      "grad_norm": 4.3116912841796875,
      "learning_rate": 1.2819640799778266e-08,
      "loss": 0.3986,
      "num_input_tokens_seen": 21528064,
      "step": 33225
    },
    {
      "epoch": 19.59316037735849,
      "grad_norm": 2.9635491371154785,
      "learning_rate": 1.2636189504087737e-08,
      "loss": 0.315,
      "num_input_tokens_seen": 21531008,
      "step": 33230
    },
    {
      "epoch": 19.59610849056604,
      "grad_norm": 4.349174976348877,
      "learning_rate": 1.2454058634460142e-08,
      "loss": 0.2583,
      "num_input_tokens_seen": 21533696,
      "step": 33235
    },
    {
      "epoch": 19.599056603773583,
      "grad_norm": 3.0084691047668457,
      "learning_rate": 1.2273248239115798e-08,
      "loss": 0.2664,
      "num_input_tokens_seen": 21537728,
      "step": 33240
    },
    {
      "epoch": 19.60200471698113,
      "grad_norm": 3.969475746154785,
      "learning_rate": 1.2093758365924746e-08,
      "loss": 0.4395,
      "num_input_tokens_seen": 21541088,
      "step": 33245
    },
    {
      "epoch": 19.60495283018868,
      "grad_norm": 2.73935866355896,
      "learning_rate": 1.1915589062408417e-08,
      "loss": 0.2986,
      "num_input_tokens_seen": 21544000,
      "step": 33250
    },
    {
      "epoch": 19.607900943396228,
      "grad_norm": 2.8506274223327637,
      "learning_rate": 1.1738740375736301e-08,
      "loss": 0.3005,
      "num_input_tokens_seen": 21547008,
      "step": 33255
    },
    {
      "epoch": 19.610849056603772,
      "grad_norm": 5.276908874511719,
      "learning_rate": 1.15632123527315e-08,
      "loss": 0.3714,
      "num_input_tokens_seen": 21550656,
      "step": 33260
    },
    {
      "epoch": 19.61379716981132,
      "grad_norm": 4.5072550773620605,
      "learning_rate": 1.1389005039865176e-08,
      "loss": 0.2389,
      "num_input_tokens_seen": 21553760,
      "step": 33265
    },
    {
      "epoch": 19.61674528301887,
      "grad_norm": 3.1581382751464844,
      "learning_rate": 1.1216118483259875e-08,
      "loss": 0.2415,
      "num_input_tokens_seen": 21556288,
      "step": 33270
    },
    {
      "epoch": 19.619693396226417,
      "grad_norm": 6.757745265960693,
      "learning_rate": 1.1044552728687319e-08,
      "loss": 0.3524,
      "num_input_tokens_seen": 21559936,
      "step": 33275
    },
    {
      "epoch": 19.62264150943396,
      "grad_norm": 6.015334129333496,
      "learning_rate": 1.0874307821570618e-08,
      "loss": 0.2608,
      "num_input_tokens_seen": 21562400,
      "step": 33280
    },
    {
      "epoch": 19.62558962264151,
      "grad_norm": 3.3159313201904297,
      "learning_rate": 1.0705383806982606e-08,
      "loss": 0.2745,
      "num_input_tokens_seen": 21565952,
      "step": 33285
    },
    {
      "epoch": 19.628537735849058,
      "grad_norm": 2.9980740547180176,
      "learning_rate": 1.0537780729646952e-08,
      "loss": 0.1728,
      "num_input_tokens_seen": 21568704,
      "step": 33290
    },
    {
      "epoch": 19.631485849056602,
      "grad_norm": 4.4798102378845215,
      "learning_rate": 1.0371498633937605e-08,
      "loss": 0.2685,
      "num_input_tokens_seen": 21571712,
      "step": 33295
    },
    {
      "epoch": 19.63443396226415,
      "grad_norm": 3.6553783416748047,
      "learning_rate": 1.0206537563877683e-08,
      "loss": 0.3173,
      "num_input_tokens_seen": 21575648,
      "step": 33300
    },
    {
      "epoch": 19.6373820754717,
      "grad_norm": 4.080669403076172,
      "learning_rate": 1.0042897563141695e-08,
      "loss": 0.3272,
      "num_input_tokens_seen": 21578912,
      "step": 33305
    },
    {
      "epoch": 19.640330188679247,
      "grad_norm": 5.145360946655273,
      "learning_rate": 9.88057867505443e-09,
      "loss": 0.3164,
      "num_input_tokens_seen": 21582720,
      "step": 33310
    },
    {
      "epoch": 19.64327830188679,
      "grad_norm": 5.8064069747924805,
      "learning_rate": 9.7195809425904e-09,
      "loss": 0.2031,
      "num_input_tokens_seen": 21585664,
      "step": 33315
    },
    {
      "epoch": 19.64622641509434,
      "grad_norm": 5.484375953674316,
      "learning_rate": 9.559904408373844e-09,
      "loss": 0.2807,
      "num_input_tokens_seen": 21588000,
      "step": 33320
    },
    {
      "epoch": 19.649174528301888,
      "grad_norm": 2.817061424255371,
      "learning_rate": 9.401549114680387e-09,
      "loss": 0.3632,
      "num_input_tokens_seen": 21590720,
      "step": 33325
    },
    {
      "epoch": 19.652122641509433,
      "grad_norm": 3.0501863956451416,
      "learning_rate": 9.244515103434826e-09,
      "loss": 0.2499,
      "num_input_tokens_seen": 21594112,
      "step": 33330
    },
    {
      "epoch": 19.65507075471698,
      "grad_norm": 4.67524528503418,
      "learning_rate": 9.08880241621335e-09,
      "loss": 0.26,
      "num_input_tokens_seen": 21597024,
      "step": 33335
    },
    {
      "epoch": 19.65801886792453,
      "grad_norm": 6.6625871658325195,
      "learning_rate": 8.934411094240758e-09,
      "loss": 0.3657,
      "num_input_tokens_seen": 21601056,
      "step": 33340
    },
    {
      "epoch": 19.660966981132077,
      "grad_norm": 2.7029781341552734,
      "learning_rate": 8.781341178393244e-09,
      "loss": 0.2842,
      "num_input_tokens_seen": 21604608,
      "step": 33345
    },
    {
      "epoch": 19.66391509433962,
      "grad_norm": 2.758492946624756,
      "learning_rate": 8.629592709196167e-09,
      "loss": 0.2811,
      "num_input_tokens_seen": 21607840,
      "step": 33350
    },
    {
      "epoch": 19.66686320754717,
      "grad_norm": 2.771836042404175,
      "learning_rate": 8.479165726826277e-09,
      "loss": 0.3416,
      "num_input_tokens_seen": 21611168,
      "step": 33355
    },
    {
      "epoch": 19.669811320754718,
      "grad_norm": 3.20974063873291,
      "learning_rate": 8.330060271109496e-09,
      "loss": 0.4752,
      "num_input_tokens_seen": 21614048,
      "step": 33360
    },
    {
      "epoch": 19.672759433962263,
      "grad_norm": 3.7387025356292725,
      "learning_rate": 8.18227638152258e-09,
      "loss": 0.2371,
      "num_input_tokens_seen": 21617632,
      "step": 33365
    },
    {
      "epoch": 19.67570754716981,
      "grad_norm": 6.300549507141113,
      "learning_rate": 8.035814097191452e-09,
      "loss": 0.3379,
      "num_input_tokens_seen": 21620512,
      "step": 33370
    },
    {
      "epoch": 19.67865566037736,
      "grad_norm": 6.683382511138916,
      "learning_rate": 7.890673456892317e-09,
      "loss": 0.3598,
      "num_input_tokens_seen": 21623712,
      "step": 33375
    },
    {
      "epoch": 19.681603773584907,
      "grad_norm": 3.123974323272705,
      "learning_rate": 7.746854499052215e-09,
      "loss": 0.2438,
      "num_input_tokens_seen": 21628416,
      "step": 33380
    },
    {
      "epoch": 19.684551886792452,
      "grad_norm": 3.494816541671753,
      "learning_rate": 7.604357261747907e-09,
      "loss": 0.2208,
      "num_input_tokens_seen": 21632416,
      "step": 33385
    },
    {
      "epoch": 19.6875,
      "grad_norm": 2.6309428215026855,
      "learning_rate": 7.463181782705886e-09,
      "loss": 0.33,
      "num_input_tokens_seen": 21635392,
      "step": 33390
    },
    {
      "epoch": 19.690448113207548,
      "grad_norm": 3.946633815765381,
      "learning_rate": 7.3233280993034726e-09,
      "loss": 0.2352,
      "num_input_tokens_seen": 21638240,
      "step": 33395
    },
    {
      "epoch": 19.693396226415093,
      "grad_norm": 4.519543170928955,
      "learning_rate": 7.184796248567161e-09,
      "loss": 0.2539,
      "num_input_tokens_seen": 21640832,
      "step": 33400
    },
    {
      "epoch": 19.69634433962264,
      "grad_norm": 5.268245697021484,
      "learning_rate": 7.047586267173723e-09,
      "loss": 0.2306,
      "num_input_tokens_seen": 21643040,
      "step": 33405
    },
    {
      "epoch": 19.69929245283019,
      "grad_norm": 6.1201019287109375,
      "learning_rate": 6.9116981914502114e-09,
      "loss": 0.359,
      "num_input_tokens_seen": 21645696,
      "step": 33410
    },
    {
      "epoch": 19.702240566037737,
      "grad_norm": 3.8279500007629395,
      "learning_rate": 6.7771320573734036e-09,
      "loss": 0.3141,
      "num_input_tokens_seen": 21648160,
      "step": 33415
    },
    {
      "epoch": 19.705188679245282,
      "grad_norm": 2.1621251106262207,
      "learning_rate": 6.6438879005709114e-09,
      "loss": 0.2958,
      "num_input_tokens_seen": 21650944,
      "step": 33420
    },
    {
      "epoch": 19.70813679245283,
      "grad_norm": 3.6929681301116943,
      "learning_rate": 6.511965756318961e-09,
      "loss": 0.3467,
      "num_input_tokens_seen": 21654208,
      "step": 33425
    },
    {
      "epoch": 19.71108490566038,
      "grad_norm": 3.1258127689361572,
      "learning_rate": 6.381365659545169e-09,
      "loss": 0.2303,
      "num_input_tokens_seen": 21657504,
      "step": 33430
    },
    {
      "epoch": 19.714033018867923,
      "grad_norm": 3.9394404888153076,
      "learning_rate": 6.252087644825766e-09,
      "loss": 0.2727,
      "num_input_tokens_seen": 21660224,
      "step": 33435
    },
    {
      "epoch": 19.71698113207547,
      "grad_norm": 4.029761791229248,
      "learning_rate": 6.124131746388373e-09,
      "loss": 0.3234,
      "num_input_tokens_seen": 21663648,
      "step": 33440
    },
    {
      "epoch": 19.71992924528302,
      "grad_norm": 4.205463409423828,
      "learning_rate": 5.997497998109225e-09,
      "loss": 0.2105,
      "num_input_tokens_seen": 21666176,
      "step": 33445
    },
    {
      "epoch": 19.722877358490567,
      "grad_norm": 5.332382678985596,
      "learning_rate": 5.8721864335153925e-09,
      "loss": 0.2467,
      "num_input_tokens_seen": 21669024,
      "step": 33450
    },
    {
      "epoch": 19.725825471698112,
      "grad_norm": 3.1395184993743896,
      "learning_rate": 5.748197085784224e-09,
      "loss": 0.3298,
      "num_input_tokens_seen": 21674144,
      "step": 33455
    },
    {
      "epoch": 19.72877358490566,
      "grad_norm": 5.46707820892334,
      "learning_rate": 5.62552998774113e-09,
      "loss": 0.3387,
      "num_input_tokens_seen": 21677344,
      "step": 33460
    },
    {
      "epoch": 19.73172169811321,
      "grad_norm": 4.002338409423828,
      "learning_rate": 5.504185171864018e-09,
      "loss": 0.4663,
      "num_input_tokens_seen": 21680224,
      "step": 33465
    },
    {
      "epoch": 19.734669811320753,
      "grad_norm": 2.209017753601074,
      "learning_rate": 5.384162670278858e-09,
      "loss": 0.3674,
      "num_input_tokens_seen": 21684032,
      "step": 33470
    },
    {
      "epoch": 19.7376179245283,
      "grad_norm": 4.889404773712158,
      "learning_rate": 5.265462514762454e-09,
      "loss": 0.2577,
      "num_input_tokens_seen": 21688384,
      "step": 33475
    },
    {
      "epoch": 19.74056603773585,
      "grad_norm": 4.534067153930664,
      "learning_rate": 5.148084736740777e-09,
      "loss": 0.3535,
      "num_input_tokens_seen": 21691040,
      "step": 33480
    },
    {
      "epoch": 19.743514150943398,
      "grad_norm": 5.3081746101379395,
      "learning_rate": 5.032029367290081e-09,
      "loss": 0.3007,
      "num_input_tokens_seen": 21693984,
      "step": 33485
    },
    {
      "epoch": 19.746462264150942,
      "grad_norm": 3.8619420528411865,
      "learning_rate": 4.917296437136898e-09,
      "loss": 0.3436,
      "num_input_tokens_seen": 21697952,
      "step": 33490
    },
    {
      "epoch": 19.74941037735849,
      "grad_norm": 3.537752151489258,
      "learning_rate": 4.8038859766569305e-09,
      "loss": 0.2731,
      "num_input_tokens_seen": 21701152,
      "step": 33495
    },
    {
      "epoch": 19.75235849056604,
      "grad_norm": 3.9993393421173096,
      "learning_rate": 4.691798015876714e-09,
      "loss": 0.2433,
      "num_input_tokens_seen": 21704032,
      "step": 33500
    },
    {
      "epoch": 19.755306603773583,
      "grad_norm": 5.744007587432861,
      "learning_rate": 4.581032584470846e-09,
      "loss": 0.3968,
      "num_input_tokens_seen": 21706560,
      "step": 33505
    },
    {
      "epoch": 19.75825471698113,
      "grad_norm": 3.840881109237671,
      "learning_rate": 4.471589711766422e-09,
      "loss": 0.4457,
      "num_input_tokens_seen": 21710496,
      "step": 33510
    },
    {
      "epoch": 19.76120283018868,
      "grad_norm": 3.2303357124328613,
      "learning_rate": 4.363469426737487e-09,
      "loss": 0.4101,
      "num_input_tokens_seen": 21715520,
      "step": 33515
    },
    {
      "epoch": 19.764150943396228,
      "grad_norm": 3.1090261936187744,
      "learning_rate": 4.256671758010588e-09,
      "loss": 0.2873,
      "num_input_tokens_seen": 21718624,
      "step": 33520
    },
    {
      "epoch": 19.767099056603772,
      "grad_norm": 5.330536365509033,
      "learning_rate": 4.151196733859775e-09,
      "loss": 0.2703,
      "num_input_tokens_seen": 21720960,
      "step": 33525
    },
    {
      "epoch": 19.77004716981132,
      "grad_norm": 2.9250435829162598,
      "learning_rate": 4.047044382211041e-09,
      "loss": 0.3181,
      "num_input_tokens_seen": 21724608,
      "step": 33530
    },
    {
      "epoch": 19.77299528301887,
      "grad_norm": 5.328889846801758,
      "learning_rate": 3.94421473063844e-09,
      "loss": 0.3079,
      "num_input_tokens_seen": 21728128,
      "step": 33535
    },
    {
      "epoch": 19.775943396226417,
      "grad_norm": 3.2074201107025146,
      "learning_rate": 3.842707806366863e-09,
      "loss": 0.3063,
      "num_input_tokens_seen": 21730560,
      "step": 33540
    },
    {
      "epoch": 19.77889150943396,
      "grad_norm": 5.771273136138916,
      "learning_rate": 3.742523636270368e-09,
      "loss": 0.377,
      "num_input_tokens_seen": 21733248,
      "step": 33545
    },
    {
      "epoch": 19.78183962264151,
      "grad_norm": 3.6034700870513916,
      "learning_rate": 3.6436622468738468e-09,
      "loss": 0.3199,
      "num_input_tokens_seen": 21736384,
      "step": 33550
    },
    {
      "epoch": 19.784787735849058,
      "grad_norm": 5.59542989730835,
      "learning_rate": 3.546123664350254e-09,
      "loss": 0.2879,
      "num_input_tokens_seen": 21739232,
      "step": 33555
    },
    {
      "epoch": 19.787735849056602,
      "grad_norm": 5.666369915008545,
      "learning_rate": 3.449907914524486e-09,
      "loss": 0.3504,
      "num_input_tokens_seen": 21742560,
      "step": 33560
    },
    {
      "epoch": 19.79068396226415,
      "grad_norm": 2.990844964981079,
      "learning_rate": 3.355015022869501e-09,
      "loss": 0.2659,
      "num_input_tokens_seen": 21746464,
      "step": 33565
    },
    {
      "epoch": 19.7936320754717,
      "grad_norm": 2.284895896911621,
      "learning_rate": 3.261445014508535e-09,
      "loss": 0.3381,
      "num_input_tokens_seen": 21749312,
      "step": 33570
    },
    {
      "epoch": 19.796580188679247,
      "grad_norm": 3.9516685009002686,
      "learning_rate": 3.1691979142145503e-09,
      "loss": 0.4123,
      "num_input_tokens_seen": 21752352,
      "step": 33575
    },
    {
      "epoch": 19.79952830188679,
      "grad_norm": 4.882086753845215,
      "learning_rate": 3.078273746410787e-09,
      "loss": 0.2706,
      "num_input_tokens_seen": 21755008,
      "step": 33580
    },
    {
      "epoch": 19.80247641509434,
      "grad_norm": 5.284690856933594,
      "learning_rate": 2.988672535169657e-09,
      "loss": 0.2457,
      "num_input_tokens_seen": 21757728,
      "step": 33585
    },
    {
      "epoch": 19.805424528301888,
      "grad_norm": 4.267265796661377,
      "learning_rate": 2.9003943042127393e-09,
      "loss": 0.3798,
      "num_input_tokens_seen": 21760704,
      "step": 33590
    },
    {
      "epoch": 19.808372641509433,
      "grad_norm": 7.368278503417969,
      "learning_rate": 2.8134390769135598e-09,
      "loss": 0.3238,
      "num_input_tokens_seen": 21763328,
      "step": 33595
    },
    {
      "epoch": 19.81132075471698,
      "grad_norm": 4.624442100524902,
      "learning_rate": 2.7278068762925935e-09,
      "loss": 0.3033,
      "num_input_tokens_seen": 21766304,
      "step": 33600
    },
    {
      "epoch": 19.81426886792453,
      "grad_norm": 2.7650201320648193,
      "learning_rate": 2.6434977250217043e-09,
      "loss": 0.2934,
      "num_input_tokens_seen": 21771168,
      "step": 33605
    },
    {
      "epoch": 19.817216981132077,
      "grad_norm": 2.7817182540893555,
      "learning_rate": 2.560511645422481e-09,
      "loss": 0.2716,
      "num_input_tokens_seen": 21775296,
      "step": 33610
    },
    {
      "epoch": 19.82016509433962,
      "grad_norm": 3.030304193496704,
      "learning_rate": 2.4788486594656825e-09,
      "loss": 0.2469,
      "num_input_tokens_seen": 21778304,
      "step": 33615
    },
    {
      "epoch": 19.82311320754717,
      "grad_norm": 4.882077693939209,
      "learning_rate": 2.398508788771792e-09,
      "loss": 0.2465,
      "num_input_tokens_seen": 21781312,
      "step": 33620
    },
    {
      "epoch": 19.826061320754718,
      "grad_norm": 9.694612503051758,
      "learning_rate": 2.3194920546110166e-09,
      "loss": 0.2674,
      "num_input_tokens_seen": 21784544,
      "step": 33625
    },
    {
      "epoch": 19.829009433962263,
      "grad_norm": 4.503481864929199,
      "learning_rate": 2.2417984779032896e-09,
      "loss": 0.2673,
      "num_input_tokens_seen": 21787744,
      "step": 33630
    },
    {
      "epoch": 19.83195754716981,
      "grad_norm": 2.7280526161193848,
      "learning_rate": 2.1654280792193782e-09,
      "loss": 0.3076,
      "num_input_tokens_seen": 21790816,
      "step": 33635
    },
    {
      "epoch": 19.83490566037736,
      "grad_norm": 2.7457621097564697,
      "learning_rate": 2.0903808787769987e-09,
      "loss": 0.3614,
      "num_input_tokens_seen": 21795040,
      "step": 33640
    },
    {
      "epoch": 19.837853773584907,
      "grad_norm": 3.928129196166992,
      "learning_rate": 2.0166568964463673e-09,
      "loss": 0.3432,
      "num_input_tokens_seen": 21799072,
      "step": 33645
    },
    {
      "epoch": 19.840801886792452,
      "grad_norm": 3.6234021186828613,
      "learning_rate": 1.9442561517463153e-09,
      "loss": 0.4043,
      "num_input_tokens_seen": 21801888,
      "step": 33650
    },
    {
      "epoch": 19.84375,
      "grad_norm": 3.391779899597168,
      "learning_rate": 1.8731786638442886e-09,
      "loss": 0.3369,
      "num_input_tokens_seen": 21804960,
      "step": 33655
    },
    {
      "epoch": 19.846698113207548,
      "grad_norm": 5.266139030456543,
      "learning_rate": 1.8034244515591214e-09,
      "loss": 0.2302,
      "num_input_tokens_seen": 21807776,
      "step": 33660
    },
    {
      "epoch": 19.849646226415093,
      "grad_norm": 4.888937950134277,
      "learning_rate": 1.7349935333582646e-09,
      "loss": 0.232,
      "num_input_tokens_seen": 21810272,
      "step": 33665
    },
    {
      "epoch": 19.85259433962264,
      "grad_norm": 3.3330070972442627,
      "learning_rate": 1.6678859273594471e-09,
      "loss": 0.3204,
      "num_input_tokens_seen": 21813568,
      "step": 33670
    },
    {
      "epoch": 19.85554245283019,
      "grad_norm": 5.429058074951172,
      "learning_rate": 1.6021016513295683e-09,
      "loss": 0.2795,
      "num_input_tokens_seen": 21816640,
      "step": 33675
    },
    {
      "epoch": 19.858490566037737,
      "grad_norm": 3.849905014038086,
      "learning_rate": 1.5376407226846968e-09,
      "loss": 0.269,
      "num_input_tokens_seen": 21819840,
      "step": 33680
    },
    {
      "epoch": 19.861438679245282,
      "grad_norm": 6.298313140869141,
      "learning_rate": 1.4745031584917357e-09,
      "loss": 0.3895,
      "num_input_tokens_seen": 21823136,
      "step": 33685
    },
    {
      "epoch": 19.86438679245283,
      "grad_norm": 5.676501750946045,
      "learning_rate": 1.4126889754667583e-09,
      "loss": 0.3267,
      "num_input_tokens_seen": 21826912,
      "step": 33690
    },
    {
      "epoch": 19.86733490566038,
      "grad_norm": 3.1536061763763428,
      "learning_rate": 1.3521981899750069e-09,
      "loss": 0.2838,
      "num_input_tokens_seen": 21829568,
      "step": 33695
    },
    {
      "epoch": 19.870283018867923,
      "grad_norm": 1.7288525104522705,
      "learning_rate": 1.293030818032004e-09,
      "loss": 0.2427,
      "num_input_tokens_seen": 21834176,
      "step": 33700
    },
    {
      "epoch": 19.87323113207547,
      "grad_norm": 11.561240196228027,
      "learning_rate": 1.2351868753018858e-09,
      "loss": 0.3549,
      "num_input_tokens_seen": 21840416,
      "step": 33705
    },
    {
      "epoch": 19.87617924528302,
      "grad_norm": 3.2687270641326904,
      "learning_rate": 1.1786663770996242e-09,
      "loss": 0.189,
      "num_input_tokens_seen": 21843360,
      "step": 33710
    },
    {
      "epoch": 19.879127358490567,
      "grad_norm": 2.772833824157715,
      "learning_rate": 1.1234693383893602e-09,
      "loss": 0.3367,
      "num_input_tokens_seen": 21845856,
      "step": 33715
    },
    {
      "epoch": 19.882075471698112,
      "grad_norm": 2.3885951042175293,
      "learning_rate": 1.0695957737844043e-09,
      "loss": 0.3177,
      "num_input_tokens_seen": 21848960,
      "step": 33720
    },
    {
      "epoch": 19.88502358490566,
      "grad_norm": 3.5318005084991455,
      "learning_rate": 1.0170456975483467e-09,
      "loss": 0.2629,
      "num_input_tokens_seen": 21852384,
      "step": 33725
    },
    {
      "epoch": 19.88797169811321,
      "grad_norm": 3.2223477363586426,
      "learning_rate": 9.658191235933922e-10,
      "loss": 0.2781,
      "num_input_tokens_seen": 21856384,
      "step": 33730
    },
    {
      "epoch": 19.890919811320753,
      "grad_norm": 4.62668514251709,
      "learning_rate": 9.159160654825805e-10,
      "loss": 0.2569,
      "num_input_tokens_seen": 21858688,
      "step": 33735
    },
    {
      "epoch": 19.8938679245283,
      "grad_norm": 3.2080185413360596,
      "learning_rate": 8.673365364281205e-10,
      "loss": 0.2172,
      "num_input_tokens_seen": 21861536,
      "step": 33740
    },
    {
      "epoch": 19.89681603773585,
      "grad_norm": 4.424536228179932,
      "learning_rate": 8.200805492913911e-10,
      "loss": 0.3053,
      "num_input_tokens_seen": 21864704,
      "step": 33745
    },
    {
      "epoch": 19.899764150943398,
      "grad_norm": 2.898951768875122,
      "learning_rate": 7.741481165834952e-10,
      "loss": 0.3345,
      "num_input_tokens_seen": 21867936,
      "step": 33750
    },
    {
      "epoch": 19.902712264150942,
      "grad_norm": 5.465450286865234,
      "learning_rate": 7.29539250465261e-10,
      "loss": 0.3744,
      "num_input_tokens_seen": 21870912,
      "step": 33755
    },
    {
      "epoch": 19.90566037735849,
      "grad_norm": 3.5932023525238037,
      "learning_rate": 6.862539627472409e-10,
      "loss": 0.3061,
      "num_input_tokens_seen": 21874816,
      "step": 33760
    },
    {
      "epoch": 19.90860849056604,
      "grad_norm": 3.862584114074707,
      "learning_rate": 6.442922648897121e-10,
      "loss": 0.4425,
      "num_input_tokens_seen": 21877792,
      "step": 33765
    },
    {
      "epoch": 19.911556603773583,
      "grad_norm": 4.988375186920166,
      "learning_rate": 6.036541680015662e-10,
      "loss": 0.4039,
      "num_input_tokens_seen": 21881056,
      "step": 33770
    },
    {
      "epoch": 19.91450471698113,
      "grad_norm": 2.5500500202178955,
      "learning_rate": 5.643396828419745e-10,
      "loss": 0.2908,
      "num_input_tokens_seen": 21884768,
      "step": 33775
    },
    {
      "epoch": 19.91745283018868,
      "grad_norm": 2.6254327297210693,
      "learning_rate": 5.26348819819833e-10,
      "loss": 0.317,
      "num_input_tokens_seen": 21888224,
      "step": 33780
    },
    {
      "epoch": 19.920400943396228,
      "grad_norm": 5.017343044281006,
      "learning_rate": 4.896815889937622e-10,
      "loss": 0.2684,
      "num_input_tokens_seen": 21890912,
      "step": 33785
    },
    {
      "epoch": 19.923349056603772,
      "grad_norm": 5.055856227874756,
      "learning_rate": 4.543380000704423e-10,
      "loss": 0.2966,
      "num_input_tokens_seen": 21893792,
      "step": 33790
    },
    {
      "epoch": 19.92629716981132,
      "grad_norm": 2.656907081604004,
      "learning_rate": 4.203180624084979e-10,
      "loss": 0.343,
      "num_input_tokens_seen": 21897472,
      "step": 33795
    },
    {
      "epoch": 19.92924528301887,
      "grad_norm": 11.166326522827148,
      "learning_rate": 3.876217850146136e-10,
      "loss": 0.3888,
      "num_input_tokens_seen": 21901472,
      "step": 33800
    },
    {
      "epoch": 19.932193396226417,
      "grad_norm": 8.3731689453125,
      "learning_rate": 3.56249176544643e-10,
      "loss": 0.2786,
      "num_input_tokens_seen": 21904768,
      "step": 33805
    },
    {
      "epoch": 19.93514150943396,
      "grad_norm": 4.382896423339844,
      "learning_rate": 3.262002453047197e-10,
      "loss": 0.2787,
      "num_input_tokens_seen": 21907584,
      "step": 33810
    },
    {
      "epoch": 19.93808962264151,
      "grad_norm": 4.405609607696533,
      "learning_rate": 2.974749992512571e-10,
      "loss": 0.2492,
      "num_input_tokens_seen": 21910592,
      "step": 33815
    },
    {
      "epoch": 19.941037735849058,
      "grad_norm": 4.650798320770264,
      "learning_rate": 2.700734459881726e-10,
      "loss": 0.2963,
      "num_input_tokens_seen": 21913600,
      "step": 33820
    },
    {
      "epoch": 19.943985849056602,
      "grad_norm": 3.9886300563812256,
      "learning_rate": 2.439955927713289e-10,
      "loss": 0.3327,
      "num_input_tokens_seen": 21917120,
      "step": 33825
    },
    {
      "epoch": 19.94693396226415,
      "grad_norm": 3.5222599506378174,
      "learning_rate": 2.1924144650409263e-10,
      "loss": 0.3036,
      "num_input_tokens_seen": 21920896,
      "step": 33830
    },
    {
      "epoch": 19.9498820754717,
      "grad_norm": 5.359842777252197,
      "learning_rate": 1.9581101374066546e-10,
      "loss": 0.2649,
      "num_input_tokens_seen": 21923392,
      "step": 33835
    },
    {
      "epoch": 19.952830188679247,
      "grad_norm": 4.192293643951416,
      "learning_rate": 1.7370430068441858e-10,
      "loss": 0.3402,
      "num_input_tokens_seen": 21926176,
      "step": 33840
    },
    {
      "epoch": 19.95577830188679,
      "grad_norm": 3.6082732677459717,
      "learning_rate": 1.529213131878926e-10,
      "loss": 0.3161,
      "num_input_tokens_seen": 21929408,
      "step": 33845
    },
    {
      "epoch": 19.95872641509434,
      "grad_norm": 3.2863805294036865,
      "learning_rate": 1.3346205675335288e-10,
      "loss": 0.3399,
      "num_input_tokens_seen": 21932832,
      "step": 33850
    },
    {
      "epoch": 19.961674528301888,
      "grad_norm": 2.6613755226135254,
      "learning_rate": 1.1532653653334447e-10,
      "loss": 0.208,
      "num_input_tokens_seen": 21935424,
      "step": 33855
    },
    {
      "epoch": 19.964622641509433,
      "grad_norm": 10.14702033996582,
      "learning_rate": 9.85147573284717e-11,
      "loss": 0.328,
      "num_input_tokens_seen": 21938336,
      "step": 33860
    },
    {
      "epoch": 19.96757075471698,
      "grad_norm": 3.8027918338775635,
      "learning_rate": 8.302672359072894e-11,
      "loss": 0.2165,
      "num_input_tokens_seen": 21940832,
      "step": 33865
    },
    {
      "epoch": 19.97051886792453,
      "grad_norm": 2.5610039234161377,
      "learning_rate": 6.886243941961468e-11,
      "loss": 0.2616,
      "num_input_tokens_seen": 21944480,
      "step": 33870
    },
    {
      "epoch": 19.973466981132077,
      "grad_norm": 3.29679012298584,
      "learning_rate": 5.602190856601741e-11,
      "loss": 0.3734,
      "num_input_tokens_seen": 21947648,
      "step": 33875
    },
    {
      "epoch": 19.97641509433962,
      "grad_norm": 4.624830722808838,
      "learning_rate": 4.450513442888493e-11,
      "loss": 0.2279,
      "num_input_tokens_seen": 21950912,
      "step": 33880
    },
    {
      "epoch": 19.97936320754717,
      "grad_norm": 4.902353286743164,
      "learning_rate": 3.4312120057999886e-11,
      "loss": 0.2558,
      "num_input_tokens_seen": 21953696,
      "step": 33885
    },
    {
      "epoch": 19.982311320754718,
      "grad_norm": 3.138077735900879,
      "learning_rate": 2.5442868151204224e-11,
      "loss": 0.233,
      "num_input_tokens_seen": 21959072,
      "step": 33890
    },
    {
      "epoch": 19.985259433962263,
      "grad_norm": 2.608872175216675,
      "learning_rate": 1.7897381057729867e-11,
      "loss": 0.2802,
      "num_input_tokens_seen": 21962208,
      "step": 33895
    },
    {
      "epoch": 19.98820754716981,
      "grad_norm": 3.683717727661133,
      "learning_rate": 1.1675660773757813e-11,
      "loss": 0.2289,
      "num_input_tokens_seen": 21965888,
      "step": 33900
    },
    {
      "epoch": 19.99115566037736,
      "grad_norm": 4.514104843139648,
      "learning_rate": 6.777708947969253e-12,
      "loss": 0.3038,
      "num_input_tokens_seen": 21969088,
      "step": 33905
    },
    {
      "epoch": 19.994103773584907,
      "grad_norm": 2.923308849334717,
      "learning_rate": 3.2035268765495674e-12,
      "loss": 0.3811,
      "num_input_tokens_seen": 21976160,
      "step": 33910
    },
    {
      "epoch": 19.997051886792452,
      "grad_norm": 3.495323896408081,
      "learning_rate": 9.531155059638863e-13,
      "loss": 0.2195,
      "num_input_tokens_seen": 21978592,
      "step": 33915
    },
    {
      "epoch": 20.0,
      "grad_norm": 3.361555576324463,
      "learning_rate": 2.6475431291750342e-14,
      "loss": 0.2477,
      "num_input_tokens_seen": 21981320,
      "step": 33920
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.6283894777297974,
      "eval_runtime": 19.5659,
      "eval_samples_per_second": 86.681,
      "eval_steps_per_second": 21.67,
      "num_input_tokens_seen": 21981320,
      "step": 33920
    },
    {
      "epoch": 20.0,
      "num_input_tokens_seen": 21981320,
      "step": 33920,
      "total_flos": 9.898089047467622e+17,
      "train_loss": 0.46223124027217055,
      "train_runtime": 4034.9032,
      "train_samples_per_second": 33.612,
      "train_steps_per_second": 8.407
    }
  ],
  "logging_steps": 5,
  "max_steps": 33920,
  "num_input_tokens_seen": 21981320,
  "num_train_epochs": 20,
  "save_steps": 3392,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.898089047467622e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}